Fuel Trends Report:
Gasoline 1995 - 2005
United States
Environmental Protection
Agency
-------
Fuel Trends Report:
Gasoline 1995 - 2005
Compliance and Innovative Strategies Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
NOTICE
This technical report does not necessarily represent final EPA decisions or
positions. It is intended to present technical analysis of issues using data
that are currently available. The purpose in the release of such reports is to
facilitate the exchange of technical information and to inform the public of
technical developments.
SER&
United States
Environmental Protection
Agency
EPA420-R-08-002
January 2008
-------
of
Executive Summary iii
Table of Acronyms viii
Background Material 1
General Methodology 6
Summary of Average Estimates 12
RFC Trends 17
Conventional Gasoline Trends 35
Sulfur 41
Reid Vapor Pressure 50
Oxygenates and Oxygen 60
Benzene 88
Aromatics 96
Olefins 106
Distillation Parameters 112
Emission and Emission Performance 123
References 139
Appendices 140
Appendix to Sulfur Chapter 141
Appendix to RVP Chapter 151
Appendix to Oxygenates and Oxygen Chapter 159
Appendix to Benzene Chapter 174
Appendix to Aromatics Chapter 184
Appendix to Olefins Chapter 194
Appendix to Distillation Parameters Chapter 203
Appendix to Emissions and Emissions Performance Chapter 225
Regression Analysis 250
PADD Level Analysis of Reporting Data 256
The primary data analyst and author of this report is Stuart Romanow.
-------
This report presents 1995-2005 clean fuel programs implementation data collected
and analyzed by the Environmental Protection Agency's (EPA's) Office of Transportation and
Air Quality. The data show that significant changes in gasoline composition during this
period resulted in emission reductions often substantially greater than regulatory
requirements. Future reports on fuel trends will be issued periodically.
As a result of the Clean Air Act (CAA) of 1990, EPA adopted clean fuel programs for gasoline. In
1995, EPA implemented the Reformulated Gasoline (RFC) program, designed to reduce emissions of
ozone-causing volatile organic compounds (VOCs) and oxides of nitrogen (NOx), and air toxics such as
benzene and formaldehyde. At the same time, EPA implemented an anti-dumping program, to protect
the emission qualities of conventional gasoline (CG). In 2000, credit for early gasoline sulfur reduction
was provided by EPA's Tier 2 gasoline sulfur program.
These clean fuel programs required gasoline refiners and importers to analyze gasoline, measure
certain emission-related parameters, and submit the data to EPA. These data have limitations, but in
many respects, provide an unparalleled source of information about gasoline property trends since 1995.1
« Gasoline Sulfur Decreases-- Average annual sulfur content in all gasoline dropped from about
300 ppm in 1997 to about 90 ppm in 2005.
« RFC NOx Reductions Exceed Requirements RFC exceeded applicable NOx performance
standards during both Phase I (1998-1999) and Phase II (2000 and beyond).
« RFC Toxics Reductions Exceed Requirements On average, Phase I RFC complied with
Phase II standards, and toxic performance still improved with the transition to Phase II
standards.
« Conventional Gasoline NOx and Toxics Emissions Decreased Between 1998 and 2005,
the summer NOx emissions of conventional gasoline were reduced by 5.7 percent, while summer
exhaust toxics were reduced by 4.7 percent.
« Ethanol Use in RFC Increased and MTBE Use Decreased In the summer of 1996, about
11 percent of the RFC sold contained ethanol while virtually all the remainder contained MTBE.
By the summer of 2005, the ethanol share increased to about 53 percent, with corresponding
decreases in MTBE.
,
As Figure 1 demonstrates, average annual sulfur content in all gasoline dropped from about 300
ppm in 1997 to about 90 ppm in 2005. Early decreases in overall gasoline sulfur content were primarily
due to decreases in RFC sulfur content linked to the phase-in of increasingly stringent RFC NOx emission
performance standards. These NOx emission performance standards did not mandate sulfur reduction,
but lowering sulfur content was one of several property changes important to meeting the RFC NOx
standards. Post-2000 decreases were also due to early Tier 2 sulfur reductions, applicable to both RFC
1 EPA lacks information about certain properties, and has only partial information on others. One important
limitation of the trend analyses in this report is that, with the exception of certain oxygen and oxygenate analyses,
they do not include California gasoline.
in
-------
and CG. Phase-in of Tier 2 sulfur reductions began in 2004, but credit generation for early sulfur
reduction was allowed beginning in 2000.
Annual Average Gasoline Sulfur Content- (parts per million)
Estimated from EPA Reporting System Data
1
5 200
1
A..
^<. ^"^----^l ''"'"\
\ ^V
"X \
x\^.
~"~-B
1997 1998 1999 2000 2001 2002 2003 2004 2005
Year
All Gasoline
--RFG
- - * - - CG
Figure 1
RFG NOx Reductions Exceed Requirements
As Figure 2 demonstrates, RFG exceeded applicable NOx performance standards during both
Phase I (1998-1999) and Phase II (2000 and beyond). The summer NOx performance of Phase I RFG
exceeded the standard by as much as 3.5 percent, while the summer NOx performance of Phase II RFG
exceeded the standard by as much as 4.1 percent.
-------
NOx Performance of Summer RFG-Estimated from EPA Data
(Based on Phase II Complex Model)
2005
"Phase I standard is an
approximation
based on the Phase II
complex model
Figure 2
RFG Toxics Reductions Exceed Requirements
RFG also exceeded toxics performance standards. As Figure 3 demonstrates, the summer toxics
performance of Phase I RFG exceeded the standard by as much as 15.1 percent, while the summer toxics
performance of Phase II RFG exceeded the standard by as much as 12.8 percent. Winter RFG toxics
performance also exceeded standards (See Figure 9 in the RFG Trends section). On average, Phase I
RFG complied with Phase II standards, and toxic performance still improved with the transition to Phase
II standards.
Toxics Performance of Summer RFG - Estimated from EPA Data
^ (Based on Phase II Complex Model)
linegasolin
0 CO 4:
D CTl c
«
re
.0
>n Reduction (fror
S 8 £
55 IU
(A
1 5
S
y " =^»
=^s
1998 1999 2000 2001 2002 2003 2004 2005
Year
Retail Avg.
^ Reporting Avg.
^ Phs II Averaged 3d
Phs I Averaged 3d *
*Phase I standard is an
approximation
based on the Phase II
complex model
Figure 3
-------
Conventional Gasoline NOx and Toxics Emissions Decreased
As Figures 4 and 5 demonstrate, between 1998 and 2005, the summer NOx emissions of
conventional gasoline were reduced by 5.7 percent, while summer exhaust toxics emissions were reduced
by 4.7 percent. Winter emissions also decreased during this period (See Figures 3 and 5 in the
Conventional Gasoline Trends section). These reductions were not required by EPA regulations; instead,
they were a byproduct of Tier 2 sulfur regulation.
1200
NOx Emissions of Summer CG
Estimated from EPA Reporting System Data
1999 2000
2002 2003
2004 2005
Year
Figure 4
Exhaust Toxics Emissions of Summer CG
Estimated from EPA Reporting System
Year
Figure 5
VI
-------
Ethanol Use in RFC Increased and MTBE Use Decreased
The CM required that RFC contain two percent oxygen by weight. MTBE and ethanol were the
primary oxygenates used. Figure 6 shows the increasing use of ethanol in RFC and the decreasing use of
MTBE through 2005. In the summer of 1996, only about 11 percent of the RFC sold contained ethanol
while virtually all the remainder contained MTBE. By the summer of 2005, the ethanol share increased to
about 53 percent, with corresponding decreases in MTBE use.
% of Summer RFG Oxygenated with Ethanol and Ethers (Including Federal RFG Areas in CA)
100%
90%
80%
70%
Figure 6
Gasoline oxygen content has been a topic of considerable interest. Concerns over groundwater
contamination from MTBE resulted in various state laws banning or phasing out its use in gasoline. The
Energy Policy Act of 2005 included a renewable content requirement for gasoline and eliminated the RFG
oxygen content requirement. RFG data for 2006, while not analyzed for this report, show that RFG
suppliers continued to use oxygen in RFG even after the requirement was removed in May of 2006, and
that virtually all of this RFG was ethanol-oxygenated. MTBE use in RFG is currently at near zero levels.
EPA finalized Renewable Fuel Standard program regulations in April 2007 to implement the Energy Policy
Act renewable content requirement. Like RFG, these regulations include new recordkeeping and
reporting requirements designed to track the volume of renewable fuel, including ethanol.
vn
-------
Ta' :
ASTM ASTM International- a voluntary standards development organization
(formerly, American Society for Testing and Materials)
CAA Clean Air Act
CFR Code of Federal Regulations
CG Conventional Gasoline
E200 Percent evaporated at 200 degrees F
E300 Percent evaporated at 300 degrees F
EPA US Environmental Protection Agency
ETBE Ethyl tertiary-Butyl Ether (an oxygenate)
FR Federal Register
MSAT Mobile Source Air Toxics
MTBE Methyl tertiary-Butyl Ether (an oxygenate)
NOx Oxides of nitrogen
PADD Petroleum Administration for Defense District (PADD I - East Coast, PADD II
Midwest, PADD Ill-Gulf Coast, PADD IV- Rocky Mountain, PADD V-West
Coast)
RBOB Reformulated gasoline blendstock for oxygenate blending (i.e. the RFC
blendstock to which an oxygenate is added)
RFC Federal Reformulated Gasoline
RFS Renewable Fuels Standard
RVP Reid Vapor Pressure
T50 50 percent evaporation temperature (degrees F)
T90 90 percent evaporation temperature (degrees F)
TAME Tertiary-Amyl Methyl Ether (an oxygenate)
VOC Volatile Organic Compound
Vlll
-------
EPA has developed and implemented several regulatory programs addressing air quality and
motor vehicle emissions which have significant impacts on gasoline properties and composition. Portions
of these regulations imposed data-related obligations on gasoline refiners, importers and other entities,
requiring that they analyze gasoline to measure certain emission-related parameters, and submit these
data to EPA. Refiners and importers must also submit gasoline volume data. These data, although
intended for compliance evaluation, also provide a unique source of information about trends in emission-
related gasoline properties and performance. This report analyzes these data to quantify trends, and to
examine how these trends relate to regulatory requirements. This chapter includes a brief description of
these regulatory programs, and the data collected and analyzed in this report.
.*'. am
Reformulated gasoline (RFC) is gasoline blended to burn cleaner and reduce smog-forming and
toxic pollutants. In 711(k) of the 1990 Amendments to the Clean Air Act (CAA 1990), Congress required
RFC to be sold in cities with the worst ozone non-attainment problems. In addition, other cities with
significant smog problems may choose to use RFC. RFC is currently used in 17 states and the District of
Columbia. About 30 percent of gasoline sold in the U.S. is reformulated.
In the 1990 Amendments, Congress also required that non-RFG, or conventional gasoline (CG),
sold in the rest of the country become no more polluting than gasoline sold in 1990. This requirement
ensures that refiners do not "dump" into conventional gasoline fuel components that are restricted in RFC
and that cause environmentally harmful emissions.
EPA introduced the RFC program in 1995, as required by the CAA.2 The RFC program
establishes emissions performance standards for volatile organic compounds (VOCs), nitrogen oxides
(NOx), and toxics. These standards are based on percent reductions from the average emissions of these
pollutants in 1990 model year vehicles operated on a specified baseline gasoline. The RFC program also
establishes a maximum benzene standard of 1.0 volume percent, and an oxygen minimum standard of
2.0 weight percent. (The Energy Policy Act of 2005 repealed the oxygen content requirement.) For
conventional gasoline, the program establishes emissions standards for exhaust toxics and NOx designed
to ensure that an individual refinery's or importer's gasoline will not have higher levels of these pollutants
than the refinery's or importer's 1990 gasoline. These standards for conventional gasoline are called the
anti-dumping standards.3 EPA has implemented the RFC program in three phases; the "Simple Model"
program began in 1995 and ended in 1997, the Phase I "Complex Model" RFC program began in 1998
and ended in!999, and the Phase II RFC program began in 2000. Phase II RFC is designed to result in
greater reductions of VOCs, NOx, and toxics emissions.
Refiners and importers of RFC are allowed to comply with each RFC standard either on a per-
gallon or annual average basis. Refiners and importers of conventional gasoline must comply with their
anti-dumping standards on an annual average basis. Refiners of RFC must comply with the RFC
standards separately for each refinery. Refiners of conventional gasoline may comply separately for each
refinery, or they may aggregate their refineries. Importers comply with both the RFC and conventional
gasoline standards for the aggregate of the gasoline they import during the year.
2 Regulations pertaining to the RFG program are in Subpart D-Reformulated Gasoline, 40 C.F.R. pt. 80
3 Subpart E-Anti-Dumping, 40 C.F.R. pt. 80
-------
The emissions performance of gasoline is calculated using a model, called the Complex Model, which
predicts the emissions level of each regulated pollutant based on the measured values of certain gasoline
properties. These properties are: aromatics, olefins, sulfur, Reid Vapor Pressure (RVP), benzene, oxygen
and distillation points. Refiners and importers are required to measure these properties in each batch of
gasoline they produce or import, using a prescribed regulatory test method, and calculate the emissions
level of each pollutant in each batch of gasoline using the Complex Model. The actual emissions level of
each regulated pollutant in the refiner's or importer's gasoline is compared to the emissions standard for
that pollutant to determine if the gasoline is in compliance.
In addition to the emissions requirements for RFC and conventional gasoline under the RFC
program, EPA established limits on the amount of sulfur that may be present in gasoline nationwide
beginning in 2004.4 These gasoline sulfur limits are part of a major program designed to significantly
reduce the emissions from new passenger cars and light trucks. This program, called the "Tier 2"
program, is a comprehensive regulatory initiative that treats vehicles and fuels as a system, combining
requirements for cleaner vehicles with requirements for much lower levels of sulfur in gasoline.
The Tier 2 program phases in a single set of tailpipe emission standards that apply to all
passenger cars and light trucks. To achieve these standards, very clean vehicle emission control
technology is employed. Gasoline with reduced sulfur levels is required under the Tier 2 program to
enable this very clean vehicle emission control technology to be effective. In addition to its beneficial
effects on vehicle emission control systems, the reduction in gasoline sulfur levels required under the Tier
2 program will contribute directly to cleaner air. The Tier 2 program included an initial phase-in period in
which gasoline refiners and importers were required to comply with a company-wide annual average
sulfur standard of 120 parts per million (ppm) in 2004 and 90 ppm 2005. In 2005, refineries and
importers began complying with a 30 ppm annual average sulfur standard.5 The program also limited
the amount of sulfur that may be present in any gallon of gasoline to 300 ppm in 2004 and 2005, and 80
ppm thereafter.
Under the Tier 2 program, refiners and importers were allowed to generate sulfur "credits" prior
to 2004 for reductions from refinery sulfur baselines established under the Tier 2 regulations. Beginning
with 2004, they were allowed to create sulfur credits if they over-complied with the annual average sulfur
standard by producing or importing gasoline with a sulfur level that is lower than the applicable standard.
Refiners and importers that create credits may use them to comply with the sulfur standard in a
subsequent year or they may trade the credits to another refiner or importer who may need them to
meet the sulfur standard.
In addition to the RFC and sulfur programs, EPA's Mobile Source Air Toxics (MSAT) program
establishes requirements for refiners and importers designed to ensure that the average level of toxic air
emissions from gasoline does not increase.6 This program requires refiners and importers to establish an
individual toxics baseline, separately for RFC and conventional gasoline, based on the average toxics
MOC.F.R.'80.195
5 This 30 ppm standard applies to individual refineries and importers. As noted in the subsequent paragraph,
the regulations allow generation and trading of sulfur credits which may be used to meet this standard.
6 Subpart J-Gasoline Toxics, 40 C.F.R. pt 80
-------
performance of their gasoline during the baseline period 1998 to 2000. Refiners and importers also must
establish a total baseline volume based on their volume of gasoline production or imports during the
baseline period. The volume of gasoline produced or imported during the year, up to the refiner's or
importer's baseline volume can be no more polluting than the refiner's or importer's MSAT baseline level
for that type of gasoline (RFC or conventional). Any volume produced or imported in excess of the
refiner's or importer's individual MSAT baseline volume can be no more polluting than the RFC toxics
standard or the refiner's or importer's conventional gasoline anti-dumping toxics baseline level, as
applicable.
EPA has required refiners and importers of RFC to submit compliance reports on a quarterly
basis.7 These reports include information on the volume, properties, and emissions performance for
VOC, NOx and toxics of each individual batch of gasoline produced or imported during the quarter. In
addition, refiners and importers who comply with an RFC standard for a particular pollutant on an annual
average basis must submit an annual report which calculates the refinery's or importers annual average
emissions for that pollutant.
In addition, refiners and importers of conventional gasoline are required to submit an annual
averaging report.8 This report includes information on the volume, properties exhaust toxics, and NOx
emissions of each batch of gasoline produced or imported during the annual averaging period, and a
calculation of the refinery's or importer's annual average for each pollutant. Refiners and importers of
conventional gasoline are allowed to compile composite gasoline samples up to one month for purposes
of batch reporting.
Beginning in 2004, all refiners and importers were also required to submit an annual average
sulfur compliance report, which demonstrates compliance with the gasoline sulfur standards. 9This report
must include data on each batch of RFC and conventional gasoline produced or imported during the
annual averaging period, including batch volume and sulfur content. The sulfur compliance report also
includes information on the amount of sulfur credits that were created or received during the averaging
period, used to demonstrate compliance, transferred to another party, or carried over to the next
averaging period.
Since most of the information required to be reported under the MSAT program is also required
under the RFC program, refiners and importers are not required to submit a separate MSAT compliance
report. The only additional information required to be reported under the MSAT program relates to the
toxics baseline volume and incremental volume of gasoline produced during the year by a refiner or
importer. Refiners and importers are required to include this information in their RFC and conventional
gasoline reports.
All refiners and importers of gasoline sold in the U.S. are required to comply with the reporting
requirements described above, except for refiners and importers of gasoline produced or imported for use
in California. Refiners and importers of California gasoline are exempted from the federal reporting
requirements, and certain other federal enforcement requirements, based on a determination by EPA that
California's emissions standards are as stringent as, or more stringent than, the federal emissions
740C.F.R. '80.75
840C.F.R.'80.105
940C.F.R.'80.370
-------
standards, and California's enforcement program is sufficient to ensure that refiners and importers of
California gasoline meet California's standards.
The information provided by refiners and importers in the compliance reports described above is
retained in a data base by EPA's Office of Transportation and Air Quality. This OTAQ data base is a
unique repository of information regarding gasoline production and importation in the U.S. 10 Much of the
information provided in the individual refiner and importer reports is deemed to be confidential business
information (CBI), which the Agency may not make available to the public. However, the raw data in
EPA's data base may be collectively analyzed to provide statistics on various aspects of gasoline
production and importation, such as the number and distribution of refiners, refineries and importers;
gasoline production and import volumes; gasoline properties; and emissions levels of various pollutants.
For example, this report includes analyses of gasoline property data to show RFC and CG property
averages by year and season, and property levels by gasoline volume percentile.
The information contained in this report relating to temporal and geographic trends in gasoline
properties, emissions performance, and production and import volumes is based on an analysis of the
collective data obtained from the refiner and importer compliance reports contained in EPA's data base.
This data is not available from other sources. The information and analyses contained in this report,
therefore, provides a unique picture of gasoline production and imports in the U.S.
Under the CAA, areas required to use RFC are called RFC "covered areas". Gasoline meeting RFC
emissions standards is required to be sold in each covered area. The most straightforward way to meet
this requirement would be to require that the contribution of each refinery to the gasoline supply of each
covered area would meet the RFC standards, on average. However, such an approach is unworkable in a
fungible gasoline distribution system ~ one in which a refiner pumps his product into a pipeline where it
mingles with other refiners product in unknown proportions before reaching the various covered areas.
Gasoline is supplied to many covered areas in this manner.
As a result, the RFC program provides for a system of "refinery gate averaging", in which each
refiner is responsible only for the compliance, on average, of its own product as it leaves the refinery.
This approach, however, has the potential for geographical skewing, where one covered area receives
very clean gasoline while another gets very dirty gasoline, or temporal skewing, where a covered area
receives clean gasoline for part of the year, but gets dirty gasoline for another part of the year.
10 It should be noted that EPA's database does not contain information on the production of gasoline by
California refiners for use in California nor on the importation of gasoline into California. Within this report, when we
speak of U.S. gasoline, we refer to non-California gasoline as explained in the preceding statement
-------
To ensure that such skewing does not occur, the RFG program requires refiners and importers to
conduct surveys of gasoline quality in RFG covered areas.11 These surveys are conducted by an industry
association according to a statistical sampling plan approved by EPA and involve sampling gasoline at
retail gasoline stations. If the gasoline in an area fails to meet the RFG standard for a particular RFG
pollutant, the standard for that pollutant is made more stringent and the number of surveys that must be
conducted in the following year is increased.
Under the RFG survey program, gasoline samples are taken at retail stations according to the
statistical sampling plan, the samples are tested for the relevant gasoline properties, and the emissions
performance of the gasoline is calculated to determine if the gasoline complies with the RFG standards.
Therefore, the survey data provides much information regarding the quality of gasoline sold at retail
gasoline stations in RFG covered areas. The information contained in this report relating to trends in
gasoline properties and emissions performance at retail gasoline stations is based on the data obtained
from the RFG surveys. As with the RFG and sulfur program reporting data, the RFG survey data is not
available from other sources. This report, therefore, provides a unique picture of gasoline quality at retail
stations in RFG covered areas.12
11 The requirement to conduct surveys is at 40 C.F.R. '80.67(3) and compliance survey requirements are
contained in 40 C.F.R. '80.68
12 In addition to the RFG survey analyses presented in this report, EPA posts survey-based summary
information on RFG properties and emissions performance by area and season at
http://www.epa.gov/otaq/regs/fuels/rfg/properf/rfgperf.htm
-------
Certain gasoline properties affect vehicle emissions, and these properties are regulated directly
through content standards or limits, and/or indirectly through standards controlling the emissions
characteristics of gasoline. EPA collects data on these emission-related parameters for compliance
purposes and, to verify compliance, reviews the data from individual refineries, import facilities and (for
RFC only) geographic areas. In addition to their intended compliance function, these data provide a
means to examine how gasoline has changed over time with respect to these emission-related
parameters. This "Trends" report includes year-by-year analyses of these data, in aggregate, rather than
on a facility-specific basis. EPA analyzed the "batch data" RFC and CG property information that refiners
and importers submit to EPA, as well as the retail property data collected in RFC surveys that are
conducted as a requirement of EPA's regulations.
While these data, in many respects, provide an unparalleled source of information about gasoline
property trends, EPA's data collection is limited to information needed to evaluate compliance with its
regulations. As a result, EPA lacks information about certain properties, and has only partial information
on others. One important limitation of the trend analyses in this report is that, with the exception of
certain oxygen and oxygenates analyses; they do not include California gasoline. EPA regulations
exempted refiners, importers and oxygenate blenders from reporting requirements for this gasoline, and
limited RFC compliance survey requirements to oxygen and oxygenate sampling.
Individual chapters are devoted to a specific emission-related parameter (e.g. gasoline sulfur
content) or to groups of related parameters (e.g. distillation curve properties or Complex Model emissions
calculation outputs). Each of these parameter chapters provides background information and separate
RFC and CG data analyses for the pertinent parameter(s). The parameter chapters each have an
appendix with further graphical and tabular data summaries. Additionally, the report contains overview
chapters summarizing major trends in RFC and CG. In general, material presented in the overview
chapters is expanded upon in individual parameter chapters. Some graphical analyses presented in the
overview chapters are repeated in the parameter chapters and appendices for completeness and
consistency.
EPA separately analyzed RFC and CG data. EPA further categorized RFC and CG data from each
reporting or survey year as "Summer" or "Winter", and separately analyzed these data. (The "Winter"
data in each reporting or survey year includes gasoline produced or sold before and after the "Summer"
season.) Although certain EPA standards and requirements apply to RFC and CG on an annual basis,
emission-related regulations as well as vehicle performance needs impose seasonal requirements which
substantially affect gasoline composition. RFC intended for sale during the summer ozone season must
meet VOC emission performance standards and was required to meet more stringent NOx emission
performance standards. CG intended for sale during the summer season must meet RVP standards
intended to limit evaporative VOC emissions.
These data can be further split and/or aggregated in various other ways to provide potentially
useful analyses, and, to an extent, this was done. Tabular estimates of volume-weighted annual
parameter averages for each product type, and for RFC and CG combined are presented in a summary
chapter following this chapter. Ideally, this "Trends" report would contain exhaustive analyses of the
available data, providing the best available information to a variety of users with diverse interests and
needs. However, there were some significant limitations to the scope and extent of the analysis that EPA
was able to conduct and present in this report. In part, time and resource constraints limited these
analyses. Additionally, the parties that submit these data to EPA have claimed that the data are
-------
confidential business information (CBI), and these data may be protected by CBI regulations. In order to
avoid potential CBI issues, EPA has not included any analyses where gasoline properties or volumes could
be strongly associated with individual companies, facilities or brands.
EPA analyzed "Summer" and "Winter" RFC and CG for each year in which "good" data were
available, through 2005. The earliest year may be 1995,1997 or 1998, and is dependent on the season
and type of gasoline (RFC or CG), as well as the data set (reporting or surveys) and the specific gasoline
property. This initial year is variable largely because the data submission requirements that EPA imposes
are limited to information needed to determine compliance with regulations and standards. EPA began
collecting certain data in 1995 and, for 1998 and later years, the data requirements changed. RFC
surveys began in 1995, as well, and EPA's analysis includes survey-based property estimates from 1995
when data for a given property were collected. While refinery/importer reports on RFC and CG were
submitted from 1995 on, only data from the more mature data management system that was in place for
1997 and later reports has been analyzed here.
Furthermore, although EPA has compared the properties and emissions qualities of conventional
and reformulated gasoline within this data-analysis time span to those of 1990 baseline gasoline, it is not
extrapolating backwards to extend the trend analysis back to!990. This report draws no conclusion about
how well 1990 baseline gasoline represented gasoline supplied in 1990.
Although the report is not intended to forecast trends, EPA has, in various chapters, identified
statutory or regulatory changes that have affected or will affect gasoline properties and emission qualities
beyond 2005.
Year-to-year changes in gasoline properties can occur for a variety of reasons; however factors
associated with EPA's regulatory requirements are likely to be the dominant influences on gasoline
property trends in recent years. The federal RFC program, which affected more than 25 percent of the
gasoline sold outside of California, was expected to have a substantial impact on gasoline properties.
Because of the scope of the RFC program, RFC requirements could potentially affect CG properties as
well (and EPA's regulations include "Anti-Dumping" standards applicable to CG designed to preserve the
emission-related qualities of CG after these RFC standards were implemented.) The Tier 2 Gasoline sulfur
program, which affects both CG and RFC not only controls gasoline sulfur levels but could affect other
gasoline property trends as well. While this report is largely devoted to providing quantitative estimates
of gasoline parameters and how they changed over time, EPA also attempted to identify factors which
affect these parameter values and which may be responsible for trends or changes. This latter analysis
was not intended to extensively evaluate all possible reasons for trends or changes.
The analyses in this report and the interpretation of results are, for the most part, linked to EPA's
regulatory perspective. Although the majority of gasoline is CG, EPA has emphasized analysis and
interpretation of RFC data because it collects both production and geographic-based retail data and
because the changes in RFC properties over time are likely to be the direct result of EPA regulations.
EPA's analysis of RFC data pays particular attention to the changes that occurred between 1999 and 2000
with the transition from Phase I to the current Phase II standards. EPA's analysis also focuses on
differences between RFC properties in 2000, the first year of Phase II standards, and 2005, the last year
of data included in this report. In addition to looking at overall trends in parameter averages during this
2000 to 2005 period, EPA has also looked at differences in parameter values between 2000 and 2005 on
an area-specific basis. Although differences between these two years could be due to non-regulatory
-------
factors, the Tier 2 sulfur requirements as well as changes in oxygenate use could cause or contribute to
differences not only in sulfur and oxygenates, but in other parameters.
The data analysis in this report is predominantly descriptive. Quantitative results are presented
in graphical and tabular format. Determinations that the data display a trend or patterns are largely
based on subjective judgment and identification of a likely cause for the putative change or pattern.
However, EPA has made limited use of linear regression analysis to provide a more objective means of
evaluating and interpreting its estimates. For example, EPA looked at its estimates of RFC parameter
averages to determine if these estimates indicate a statistically significant shift between 1999 and 2000
distinguishable from year-to-year fluctuations and any overall linear trend. EPA also looked at estimates
of CG parameter averages to determine if there was evidence of any shift between 1999 and 2000.
(Such shifts could, but do not necessarily indicate that CG properties changed as a result of RFC standard
changes.) These regression analyses are discussed in an appendix to this report and results are cited in
the body of the report.
EPA based the trend analyses in this report solely on data that it collected. Furthermore, these
analyses were conducted independently from other published analyses. As a result, parameter value and
volume estimates in this report may not always agree perfectly with comparable values published
elsewhere. Several specific consistency-related discussions are included in the report, and EPA found the
trends report estimates to be in reasonably good agreement with other data and analyses. Reasons for
minor differences between reporting system parameter averages published on EPA's website and
reporting system averages contained in this report are identified in the "Summary of Average Estimates"
chapter. The "Oxygenates and Oxygen" chapter compares ethanol and MTBE volume estimates in this
report with estimates that EPA published in support of the Renewable Fuels Standard regulations. The
"PADD Level Analysis of Reporting Data" appendix includes some comparisons of EPA reporting system
gasoline volume totals with publicly-available Energy Information Administration volume data.
Regulatory requirements, such as independent sampling and testing of some samples, help to
ensure the quality of the RFC and Anti-Dumping reporting data that refiners and importers submit to EPA.
Instances of missing, incomplete or incorrect RFC and Anti-Dumping batch reporting data have been
detected and corrected over the course of the data collection program. Furthermore, the reporting
system had been in place for over three years when the 1997 data were submitted and ten years when
the 2005 data were submitted, providing ample time for the industry to gain a thorough understanding of
its requirements.
Based on EPA's review of the data prior to analysis for this report, data errors are believed to
affect only a small proportion of the total data. Thus, data errors were not expected to have a critical
effect on the analysis. However, some additional data screening was performed as part of the analyses
for this report.
In general, data were eliminated from the analyses in two ways. For estimation of most
seasonal average parameter values, data were excluded on a parameter-specific basis. For each
parameter, a rule was established to determine if the data was missing from a batch record. This was
not entirely straightforward, because, for some parameters a value of zero is possible. For these
properties, a blank or null value might indicate a true "zero" value, rather than missing data. It is also
-------
possible that zeros may have been used to represent missing data. For some parameters, additional
parameter-specific exclusion criteria were sometimes added to remove values that were clearly impossible
(e.g. when a parameter value reported as a volume percentage substantially exceeded 100 percent).
EPA applied more stringent screening for certain analyses, including estimation of average "Complex
Model" emissions or emissions performance and determination of parameter value distributions by
gasoline volume percentile. For this more stringent screening, in addition to the parameter-specific data
screens, CG, RFC and RBOB batches were flagged if they had "outlier" values for any parameter and
eliminated from an analysis even if the analysis was for a different parameter. For many parameters,
the limits were set at 10% beyond the acceptable range limits for the Complex Model. For some
parameters, RFC and CG have different acceptable range limits. RFC range limits are considered as
standards applicable to every RFC batch. However, the intent was not to exclude non-complying or "high
emissions" batches from these analyses, but to identify batches where one or more parameters were
likely to have been mis-reported or a batch mis-labeled (e.g. CG identified as RFC, CG blendstock
identified as CG.)
These additional screens excluded only a small proportion of the data from most of the analyses.
The additional data screens for this report removed some erroneous data items, although valid data may
also have been removed. The level of screening did not have much effect on the averages, suggesting
that over-screening did not substantially bias the results. Removal of extreme values for certain
parameters visibly affected the tails of parameter value distributions by volume for the parameters, but
had little effect on median values. Removal of data on a batch basis rather than a parameter basis for
some of these analyses probably removed some erroneous data values for parameters, even when the
parameter value did not meet the outlier criteria for that parameter.
The RFC Survey data which were analyzed are also subject to regulatory requirements which help
ensure the quality of the data submitted to EPA. Additionally, EPA has a Quality Assurance Project Plan in
place for the survey program. The survey regulations provide for exclusion of samples from a specific
parameter average if a sample exceeds a per gallon minimum or maximum standard for that parameter
plus an enforcement tolerance. These violations are rare and do not have much effect on survey
averages. EPA did not apply any further data screening criteria to the survey data for this report,
although several surveys were totally excluded from the analyses.
The analyses excluded a survey conducted in St. Louis, MO beginning March 20, 2000, because a
waiver issued as a result of a pipeline break allowed sale of conventional gasoline. EPA also excluded all
one-week RFC surveys conducted between September 2, 2005 and September 15, 2005 from the
analyses in this report since these summer surveys potentially included a mix of summer and winter
gasoline. EPA allowed early use of winter gasoline in a Hurricane Katrina-related waiver issued August
31, 2005, and property averages from these surveys may not have been representative of RFC intended
for sale during the summer ozone season.13 Although hurricane-related waivers issued in September,
2005 allowed conventional gasoline use in certain RFC areas, the respective timing of these waivers and
surveys made it unlikely that surveys were significantly affected. No surveys were excluded from trends
report analyses in response to these waivers.
In order to develop estimates of property averages from these data, measurements from
individual samples must be combined. These measurements represent different quantities of gasoline,
consequently volume-weighting of both reporting and survey data is appropriate.
13 Ten surveys were excluded from 2005 averages; one in each of 10 areas (see
http://epa.gov/otaq/regs/fuels/rfg/properf/rfgperf.htm)
-------
Although the reporting system was intended to evaluate the compliance of individual facilities
with standards, rather than to determine seasonal property averages for RFC and CG this is a fairly
straightforward process for reporting data. The volume of each batch is reported so that
volume-weighted property averages can be computed from batch reports.
Developing seasonal property average estimates for all RFC based on individual surveys
conducted in different RFC areas throughout the year is more complex. The individual surveys are
designed to give accurate estimates of average RFC properties in a given area during a given one-week
time period. Surveys are categorized as "Summer" if they are conducted between June 1 and September
15, and "Winter" if they are conducted before or after this period. Each area surveyed is surveyed a
minimum of two times during the summer and two times during the winter. EPA has determined
property averages for each survey, and averaged survey averages to estimate seasonal property
averages for each area.
These seasonal area averages do not represent equal volumes of gasoline so an arithmetic
average of these area averages may not accurately estimate an overall seasonal RFC average.
Consequently, EPA has weighted the area averages by estimates of annual gasoline sales in each of the
survey areas in each of the years. These gasoline volume estimates are provided to EPA in the survey
plan that the RFC Survey Association submits each year. These gasoline volume estimates are included
in the survey plan for purposes other than for combining area averages, but EPA believes that they are
accurate enough (even without further adjustment to estimate area-specific seasonal volumes) to ensure
that areas are not substantially misrepresented in the seasonal average computations.
Individual surveys are designed to sample proportionally to gasoline grade within each area.
Thus, each survey property average was expected to represent the grade mix within each area.
However, this "Trends" report includes estimates of seasonal RFC averages by grade based on survey
data. In order to generate these averages, EPA used area-specific estimates of grade percentages
(included in the survey plans), along with gasoline volume estimates in order to estimate gasoline
volumes by grade. EPA also computed individual survey averages by grade, combined them into
seasonal area averages by grade, and used these estimates of gasoline volume by grade as weighting
factors to compute seasonal averages by grade.
Clearly, estimation of average RFC property trends from survey data is more involved than
estimation of average trends from reporting data and requires assumptions and approximations which
affect the accuracy of the estimates. Also, the survey data are statistical samples while the reporting
data are a census. Thus, errors inherent in statistical sampling affect the precision and accuracy of
survey-based estimates. However, both estimates are included in this report because there may be real
differences between retail and reporting property averages (e.g. due to certain downstream blending
operations or downgrading of product) and the potential for some inaccuracies in reporting as well as
retail estimates. Although both data sets were used for RFC trend analyses, in part, because there could
be differences between production and retail gasoline properties, EPA expected that such differences
would be small. Consequently, EPA believes that when the two estimates are in close agreement, it
corroborates the quality of the data and the reliability of the estimates. (In most cases, the two
estimates agree quite well. Issues pertaining to disparities between survey and reporting-based
estimates are discussed in several chapters.) It is obvious that the RFC Survey data provide important
geographic property information and geographic-based analysis of these retail data are included in this
report. There is no comparable source of such information for CG, so CG analysis is generally limited to
seasonal aggregations of reporting data and, for most properties, averages by premium or regular grade
and season. Some PADD-specific analyses of 2004 and 2005 reporting data are included for both CG and
RFC.
10
-------
In addition to estimates of property averages, this report includes analysis of the distributions of
property values by gasoline volume percentile for both CG and RFC, generated from batch reporting data.
Graphical and tabular distribution analyses are included in the appendix to each parameter chapter and,
in some chapters, in the body of the chapter, as well.
As previously noted, data used in these distribution analyses were screened for missing data and
outliers. Additionally, certain CG batches were excluded from the distribution analyses because they were
blendstock batches, rather than finished gasoline. These batches were included in computations of
average parameter values because it was assumed that the finished gasoline that would have resulted
from blending may not have also been reported. Inclusion of blendstock batches in CG average
computations could reasonably be expected to improve the accuracy of parameter average estimates
(and inclusion or exclusion has only minor effects on most estimates). However, inclusion of CG
blendstock batches in property distributions by volume is likely to make them less representative of
finished gasoline. Some properties of certain CG blendstock batches, which can be such things as
oxygenates and butane, differ substantially from finished gasoline (e.g., the RVP of butane is more than
50 psi). Thus, it is probably more appropriate and useful to exclude blendstock batches from these
distribution analyses.
RFC blendstock for oxygenate blending (RBOB) batches not removed for missing data or outliers
were included in the distribution analyses because these batches are "hand blended" with oxygenates
before properties are measured. Although these batch properties may sometimes differ from the
properties of the actual blend, they are representative of finished RFC.
Certain RFC and CG standards are based on Complex Model emissions calculations. The gasoline
property data that EPA collects are the inputs to this model, and EPA also receives certain model outputs
with these data. The current standards, applicable since 2000, are based on the Phase II version of this
model. Standards applicable for 1998 and 1999 were based on the Phase I version of the model. The
CG and RFC reporting and RFC Survey data which EPA received for 1998 and 1999 included emissions
calculations based on the Phase I model. Trend analyses based on a combination of Phase I and Phase
II model calculations would provide little information on changes in the emissions qualities of gasoline
since they would show the effects of the model version change as well as the effects of actual property
changes on emissions. EPA based all analyses on the Phase II model. EPA calculated Phase II Complex
Model results for each 1998 or 1999 batch or survey sample included in these analyses. EPA calculated
average performance by averaging the performance of individual batches or survey samples. Emissions
calculations using average property values for each year, season and gasoline type would probably
provide good approximations of average emissions or emissions performance with significantly less
computational effort. However, EPA's analysis may give slightly different and, arguably, better estimates
because there are some non-linear terms in the complex model equations. Calculating the Phase II
model performance of individual batches also allowed EPA to present emissions distribution by volume
information for 1998 and 1999 batches that can be directly compared to distributions for later years.
EPA's analyses did not include calculated emissions performance for a subset of these batches
representing about three percent of 1998 and 1999 reported RFC and RBOB volume because of omitted
oxygenate information needed for Complex Model input. (In most cases, EPA regulations allowed these
omissions.)
The gasoline data that EPA collected for years prior to 1998 did not include the full set of
property inputs for the Complex Model. Consequently, this report does not include these years in
emissions or emissions performance trend analyses.
11
-------
The tables in this chapter summarize certain gasoline parameter estimates contained in this
report and aggregate Summer/Winter and RFG/CG reporting system estimates. As explained in the
General Methodology Chapter:
» The primary trend analyses separated the data into Summer RFC, Winter RFC, Summer CG and
Winter CG.
» Data were screened on a parameter-specific basis when calculating RFC and CG seasonal
averages.
The gasoline volumes contained in the tables are the total volumes for the batches in each
category, prior to data screening. In some cases, due to data-screening, these volumes differ from the
total volume of the batches used to calculate the volume-weighted average for a given parameter in a
given year and season. The seasonal volumes shown in the tables, rather than parameter-specific
screened volumes, were used to aggregate seasonal averages into the annual and RFC + CG combined
averages.
These tables are presented here without discussion. They are intended as a quick reference, as
well as an aid to those who may want to examine aggregate trends. These tables do not show specific
oxygenates for RFC or CG, and do not show CG oxygen content. The tables do not contain all of the
information summarized in this report. The reader should refer to individual chapters for more detailed
quantitative as well as qualitative information.
The reporting system parameter averages and gasoline volumes in this report, in some cases,
differ from reporting system averages and volumes published on EPA's website. 14 Generally, the
differences are small enough to be of no significance for most purposes. EPA believes that such
differences are not due to computational error but are primarily due to the dynamic nature of the
reporting system database and to slightly different assumptions regarding the inclusion or exclusion of
data.
Even after reporting data for a given year have been received and entered, the reporting system
data for that year may change due to resubmitted and corrected data. These changes can occur some
time after the end of the reporting period. Results generated for this report and for the web posting may
differ in part because they were based on "snapshots" of the data taken at different times.
Different analysts performed the computations for the trends report and web posting and used
slightly different screening and analysis assumptions. For example, although California refineries report
their non-California production, the web posting analyses excluded these data while the trends report
analysis did not. This may well explain why the "unscreened" CG volumes given in this chapter are
slightly higher than the volumes posted on the website.
The tables posted on the website include oxygen and oxygenate concentration averages for CG.
These results may appear inconsistent with some of the CG oxygen and oxygenate analysis in this "trends
report". However, the web analysis computed oxygen and oxygenates averages only for the batches
with an oxygen value greater than zero. The "trends report" analyses included CG batches with zero or
blank oxygen/oxygenate values in any CG oxygen/oxygenate averages. The report noted the uncertainty
and probable error associated with any oxygen/oxygenates analysis of CG reporting data.
14 See http://epa.gov/otaq/regs/fuels/rfg/properf/rfg-params.htm and
http://epa.gov/otaq/regs/fuels/rfg/properf/cg-params.htm
12
-------
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
^omatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (gal)
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
^omatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (gal)
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
^omatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (gal)
1997
2.13
0.66
12.0
22.4
7.60
289
12,522,261,326
2.21
0.63
11.3
19.2
251
14,942,199,473
1997
2.17
0.64
11.6
20.7
268
27,464,460,799
1998
2.13
0.67
10.9
22.8
48.8
82.6
7.60
202
12,837,419,615
RFG
19§8
2.22
0.65
10.8
19.9
56.0
84.9
203
15,091,302,974
1998
2.18
0.66
10.8
21.2
52.7
83.8
203
27,928,722,589
-
19§9
2.11
0.71
11.4
22.1
49.2
82.8
7.60
205
13,005,827,782
-
1999
2.16
0.65
11.3
19.5
56.0
84.6
214
15,085,096,147
-
1999
2.14
0.67
11.3
20.7
52.9
83.8
210
28,090,923,929
2000
2.24
0.59
10.6
19.3
47.7
84.7
6.78
126
12,983,168,478
2000
2.12
0.65
11.8
19.0
56.3
86.1
200
15,831,074,709
2000
2.18
0.62
11.3
19.1
52.4
85.5
167
28,814,243,187
2001
2.21
0.62
11.8
20.1
47.5
84.4
6.79
127
13,230,634,977
2001
2.11
0.64
12.3
19.2
55.9
85.8
185
15,792,238,105
2001
2.15
0.63
12.1
19.6
52.0
85.1
158
29,022,873,082
2002
2.25
0.59
10.8
20.4
47.5
84.4
6.80
124
13,847,971,634
2002
2.09
0.64
11.2
19.4
55.9
85.5
184
16,494,453,697
2002
2.16
0.62
11.0
19.9
52.0
85.0
156
30,342,425,331
2003
2.30
0.61
11.0
20.1
47.9
84.4
6.83
110
13,587,633,643
2003
2.15
0.64
11.0
19.4
56.0
85.1
164
16,679,773,135
2003
2.22
0.63
11.0
19.7
52.4
84.8
139
30,267,406,778
2004
2.56
0.59
11.3
20.1
47.9
83.4
6.87
79
14,243,059,617
2004
2.38
0.63
11.1
19.1
56.2
84.9
101
17,194,370,899
2004
2.46
0.61
11.2
19.6
52.4
84.3
91
31,437,430,516
2005
2.48
0.66
11.9
20.9
48.8
84.1
6.91
69
14,092,489,036
2005
2.36
0.67
11.0
19.6
56.3
85.3
80
18,046,671,196
2005
2.41
0.67
11.4
20.1
53.0
84.8
75
32,139,160,232
Table 1
13
-------
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
Aromatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
Aromatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
-
1995
2.19
0.67
24.3
7.61
2.42
0.59
20.0
Summer
199 6
2.18
0.68
24.8
7.64
- Winter
1996
2.28
0.69
20.7
1997
2.24
0.68
25.5
7.62
1997
2.38
0.67
20.8
199S
2.26
0.68
10.3
26.0
49.4
82.7
7.65
190
1§§8
2.42
0.65
9.4
21.1
56.9
85.1
207
1999
2.26
0.72
10.8
24.9
49.8
83.1
7.62
204
1999
2.32
0.68
10.0
21.2
56.6
84.6
215
2.31
0.60
9.4
19.5
47.9
84.9
6.77
127
2000
2.31
0.65
10.0
18.1
56.6
86.1
192
2001
2.30
0.64
10.3
20.2
47.5
84.5
6.77
122
2001
2.24
0.65
10.7
18.5
56.5
86.1
182
2002
2.32
0.62
10.8
20.5
47.7
84.1
6.79
117
2.24
0.66
10.6
19.3
56.8
85.7
183
2003
2.41
0.65
10.9
20.1
48.0
84.0
6.82
106
200,3
2.35
0.69
10.8
19.1
56.6
85.6
167
2.65
0.65
10.9
21.2
48.1
83.3
6.87
78
2004
2.61
0.69
10.5
19.4
56.8
84.9
102
2005
2.60
0.72
10.8
21.1
48.8
84.3
6.92
69
2005
2.62
0.70
9.5
18.9
56.9
85.6
80
Table 2
14
-------
G
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
Aromatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (gal)
1.13
12.3
27.4
316
39,709,732,667
1.12
11.5
27.5
44.6
80.8
8.31
297
39,993,052,895
1999
1.14
11.7
27.6
45.0
81.1
8.29
301
39,702,111,855
2000
1.13
11.7
28.4
45.0
80.5
8.26
308
38,879,882,521
2001
1.15
12.6
28.3
45.1
81.1
8.25
295
39,517,397,838
2002
1.09
12.0
28.1
44.9
80.6
8.25
290
41,639,053,732
2003
1.13
11.8
27.9
45.2
80.7
8.29
295
44,550,506,550
2004
1.13
11.1
28.1
45.1
80.6
8.29
114
44,009,126,002
2005
1.19
11.8
27.8
45.6
81.6
8.29
102
42,849,893,176
CG
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
Aromatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (gal)
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
Aromatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (gal)
1.12
12.3
25.0
309
44,979,608,785
1.12
12.3
26.1
312
84,689,341,452
1998
1.07
11.2
24.8
49.9
83.2
12.13
281
46,926,442,568
CG
1998
1.09
11.3
26.0
47.5
82.1
10.4
288
86,919,495,463
1999
1.07
11.4
25.0
49.9
83.0
12.04
298
48,323,290,875
-
1999
1.10
11.5
26.1
47.7
82.1
10.4
300
88,025,402,730
2000
1.07
12.0
24.8
50.2
83.4
12.01
284
49,012,746,434
2000
1.10
11.9
26.4
47.9
82.1
10.3
295
87,892,628,955
2001
1.12
12.4
25.3
49.7
83.2
11.91
286
49,737,646,274
2001
1.14
12.5
26.6
47.7
82.3
10.3
290
89,255,044,112
2002
1.06
11.7
25.0
49.9
83.1
11.99
286
50,639,522,581
2002
1.08
11.9
26.4
47.7
82.0
10.3
288
92,278,576,313
2003
1.08
11.4
24.9
50.3
82.8
12.11
250
48,852,848,576
2003
1.11
11.6
26.3
47.8
81.8
10.3
272
93,403,355,126
2004
1.07
11.2
24.6
50.6
83.3
12.17
117
48,333,969,089
2004
1.10
11.2
26.2
48.0
82.1
10.3
116
92,343,095,091
2005
1.13
11.5
24.7
50.7
84.1
12.06
95
49,467,392,607
2005
1.16
11.7
26.1
48.3
82.9
10.3
98
92,317,285,783
Table 3
15
-------
K! n-<
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
Aromatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (gal)
1997
1.02
12.3
26.2
310
52,231,993,993
1.01
11.3
26.4
45.6
81.3
8.13
274
52,830,472,510
1999
1.03
11.6
26.2
46.1
81.5
8.12
278
52,707,939,637
R H i
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
Aromatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (gal)
1997
1.00
12.0
23.6
0
0
294
59,921,808,258
0.97
11.1
23.6
51.4
83.6
262
62,017,745,542
1999
0.97
11.4
23.7
51.4
83.4
278
63,408,387,022
K! n-<
Oxygen (wt%)
Benzene (Vol%)
Olefins (Vol%)
Aromatics (Vol%)
E200 (%)
E300 (%)
RVP (psi)
Sulfur (ppm)
Volume (qal)
1997
1.01
12.1
24.8
302
112,153,802,251
0.99
11.2
24.9
48.7
82.5
267
114,848,218,052
1.00
11.5
24.8
49.0
82.5
278
116,116,326,659
-
1.00
11.4
26.1
45.6
81.5
7.89
262
51,863,050,999
-
0.96
11.9
23.4
51.7
84.1
264
64,843,821,143
-
0.98
11.7
24.6
49.0
82.9
263
116,706,872,142
1.02
12.4
26.2
45.7
81.9
7.89
253
52,748,032,815
1.01
12.4
23.9
51.2
83.8
262
65,529,884,379
1.01
12.4
24.9
48.7
83.0
258
118,277,917,194
0.97
11.7
26.1
45.5
81.5
7.89
248
55,487,025,366
0.96
11.6
23.6
51.4
83.7
261
67,133,976,278
0.96
11.7
24.7
48.7
82.7
255
122,621,001,644
1.01
11.6
26.1
45.8
81.5
7.95
251
58,138,140,193
0.97
11.3
23.5
51.7
83.4
228
65,532,621,711
0.99
11.5
24.7
48.9
82.5
239
123,670,761,904
1.00
11.2
26.1
45.8
81.3
7.94
106
58,252,185,619
0.96
11.2
23.1
52.1
83.8
113
65,528,339,988
0.98
11.2
24.5
49.1
82.6
110
123,780,525,607
1.06
11.8
26.1
46.4
82.3
7.95
94
56,942,382,212
1.01
11.4
23.3
52.2
84.4
91
67,514,063,803
1.03
11.6
24.6
49.5
83.4
92
124,456,446,015
Table 4
16
-------
'' : ".. nds
Changes in the properties and composition of RFC have occurred since RFC was first introduced
in 1995. These parameter changes were primarily due to regulatory requirements that were phased-in
over time. These changes and their regulatory causes are discussed below, grouped into two time
periods. The first period, from 1995 to 2000, begins with the introduction of the federal RFC program
and includes two RFC standard changes; one from 1997 to 1998 and the second from 1999 to 2000.
RFC standards were at their current Phase II level during the second period, from 2000 through 2005,
however other regulatory factors affected RFC properties and composition.
Ill ;-'': ' :;.'N§11
When RFC was introduced in 1995, it was subject to oxygen minimum and benzene maximum
content standards, RVP limits for VOC-controlled (Summer ozone-season) RFC, and a toxics emission
performance standard. Emission performance refers to an emission reduction, in percent, relative to a
baseline gasoline, as determined by a mathematical model which estimates vehicle emissions based on
certain properties of the gasoline being used. The mathematical model which provided the basis for the
toxics performance standard is called the "Simple Model." The Simple Model considered the emissions
effect of four parameters: aromatics, benzene, oxygen (including, for some calculations, the individual
contribution of several specific oxygen-containing organic compounds called oxygenates), and RVP
(Summer only). The toxics emission performance standard along with the RVP, oxygen and benzene
requirements are called the Simple Model standards.
Beginning in 1998, "Complex Model" standards replaced the Simple Model standards. The
Complex Model standards included the same oxygen minimum and benzene maximum content standards
and, like the Simple Model standards, included a performance standard for toxics emissions. In addition,
the Complex Model standards included performance standards for NOx and VOC emissions. The Complex
Model considers the emissions effects of the four Simple Model parameters discussed above, and the
emissions effects of several additional gasoline parameters: sulfur, olefins, and two distillation points,
E200 and E300. The Complex Model standards were introduced in two phases: Phase I in 1998, and
Phase II, which requires more stringent emissions reductions, beginning in 2000.
EPA's evaluation of the RFC property and emissions performance changes is based on data that it
collects to determine compliance with its regulations. The gasoline parameter information that EPA
needed for compliance purposes changed in 1998, and this limits EPA's ability to compare RFC from 1995
through 1997 to RFC in 1998 and later years. (EPA has also restricted its production data analysis in this
report to 1997 and later years). EPA has collected data on the four Simple Model parameters (aromatics,
oxygen, benzene and RVP) since 1995, and there is no evidence of substantial changes in these
properties between 1997 and 1998, with the transition to the Complex Model standards. EPA also has
comprehensive production data on 1997 sulfur and olefin content, two of the four additional Complex
Model parameters. Based on these data, both sulfur and olefin content decreased between 1997 and
1998, and it is clear that the changes in sulfur content were substantial (figures 1 and 2). (See individual
parameter chapters for detailed graphical and tabular summaries of gasoline parameter values.)
According to the Complex Model, lowering gasoline sulfur content reduces NOx, exhaust toxics
and exhaust VOC emissions. Sulfur content has a substantial influence on NOx emission performance. It
is likely that refiners reduced the sulfur content of RFC from 1997 levels, in part, to meet the Complex
Model NOx emissions performance standard. However, it is unlikely that sulfur reductions of the
magnitude seen between 1997 and 1998 were solely due to the imposition of a NOx performance
standard, since Phase I Complex Model RFC was only required to have NOx emission performance equal
to baseline gasoline's with sulfur levels of 338 (Winter) and 339 (Summer) ppm. In addition, the RFC
17
-------
regulations allowed refiners and importers to supply some RFC with NOx performance worse than
baseline gasoline if they complied on average with a slightly more stringent NOx performance
requirement.
The sulfur reductions in 1998 are attributable, at least in part, to a requirement that each
Complex Model input parameter be within specific range limits. These limits were intended to address
the validity of the model, and EPA's regulations prohibit certification of RFC with properties outside of
these model limits (Limits of the model 40 C.F.R. '80.45(f)). The limit for sulfur is 500 ppm;
consequently each batch of RFC was required to have a sulfur content at or below this level beginning in
1998. Analysis of EPA's reporting data shows that about 17% of Summer and 12% of Winter RFC refined
and imported in 1997 exceeded this limit (figures 3 and 4). Thus, even if 1997 RFC on average met
1998 NOx performance standards, a portion of that RFC would have required sulfur reductions to meet
the Complex Model range limit imposed in 1998. EPA did not calculate the average Complex Model
emission performance of its 1997 refiner/importer data since suppliers were not required to report the
E200 and E300 model inputs. However, as shown later in this chapter, EPA calculated the average
emission performance of 1998 and 1999 RFC. EPA found that the NOx performance of both Summer
and Winter RFC was significantly better than the standard, and presumes that average NOx performance
improved from 1997 to 1998. This over-compliance in 1998 and 1999 is likely partially, if not primarily,
due to sulfur reductions needed to comply with the sulfur range limit.
Presumably, the NOx performance standard and the Complex Model sulfur range limit were the
major factors influencing the 1997 to 1998 decrease in RFG's average sulfur content. However, since the
Complex Model recognized the toxics and VOC emissions benefits of sulfur reduction, reducing sulfur
levels below those needed to minimally comply with the NOx standard and sulfur range limit also may
have become an economically desirable option for some producer's by 1998. (As discussed below, the
more stringent standards applicable to Phase II RFC would result in additional sulfur reductions by year
2000.)
Average Summer RFC Sulfur Content
Year
Figure 1
18
-------
Sulfur (ppm
250 -
200 -
150 -
100 -
50 -
Highest RFG Area
Lowest RFG Area
Retail Avg
Production Avg
\
Average Winter RFG Sulfur Content
»
N^-
*
--*'*
i
*
«a-«-
^fc-
»
T
*-*.
!*
' . \ *
1997
251
1998
268
147
207
203
1999
273
150
215
214
2000
255
121
192
200
2001
235
98
182
185
Year
2002
276
81
133
184
2003
247
77
167
164
">
i
2004
144
50
102
101
--*
2005
143
18
80
80
« Highest RFG Area
Lowest RFG Area
-* Retail Avg.
** Production Avg
Figure 2
Percentile Chart of Sulfur Content by Volume-1997 & 1998 Summer RFG
(from Batch Reports)
son
1000
1200
14ilO
Sulfur (ppm)
Figure 3
19
-------
Percentile Chart of Sulfur Content by Volume-1997&1998 Winter RFG
(from Batch Reports)
Sulfur (ppm)
Figure 4
The transition from Phase I to Phase II Complex Model RFG occurred between 1999 and 2000.
The Phase II emission performance standards, particularly the more stringent standards applicable to
VOC-controlled (Summer) RFG caused significant changes in average values for most Complex Model
input parameters. Phase II RFG performance standards required greater reductions of VOC and NOx
emissions in VOC-controlled (Summer) RFG, and greater reductions in toxics emissions on an annual
basis. Figures 5 through 9 show RFG Complex Model performance and applicable performance
standards. Average Summer RFG performance improved between 1999 and 2000 even for toxics, where
Phase I RFG, on average, complied with the Phase II standards. The Winter NOx standard did not
become more stringent with Phase II and, on average, Phase I Winter RFG met Phase II standards for
Winter NOx and annual toxics performance. EPA's two data sources disagree somewhat about the
magnitude of the emissions performance changes for Winter RFG. While both data sources indicate
Winter performance improvements between 1999 and 2000, only the survey estimates suggest an abrupt
improvement. (The Phase II version of the Complex Model differs from the Phase I version. In this
report, to facilitate comparison of Phase I and Phase II RFG, Phase I RFG performance and standards are
estimated using the Phase II model.)
>n Reduction (from baseline gasoline)
%Emissi
VOC Performance of Summer RFG
(Based on Phase II Complex Model)
30 -
Notes:
1 . Approximation
based on phase II
complex model
varies by geographic
weighted average of
standards shown 5
«- Best Area
Worst Area
Retail Avg.
- - - Reporting Avg.
Phs. I Averaged Std (vol wtd) 1,2
Phs. II Averaged Std. (vol wtd) 2
* t t f t T
. /*=
- i t- - 1" ' "t
t^^i
1998 19B9 2000
248 25.5 299
14.9 146 276
1B.6 19.4 28.6
19.3 18. 8 28.6
16.9 17.2
28 1
2001
300
27.1
286
28.5
27 9
2002
30.2
27 1
28 5
282
27.9
2003
30.4
27.3
285
28.3
27 9
2004
30 1
26.6
28.1
28.3
27 9
2005
31.0
26 7
28 2
28.3
27 3
Year
Figure 5
20
-------
"S*
_c
"o
V)
o
V)
SI
£
&
c
o
s
3
V
o:
c
o
n
.2
£
LLJ
$
Best Area
* Worst Area
Retail Avg.
Reporting Avg.
Phs II Averaged Std
Phs 1 Averaged Std *
NOx Performance of Summer RFC
(Based on Phase II Complex Model)
,.
f * '
i ? i ; ^__"
. /T--.^-^ r^°r"
: /* lit*
te-=v=/'
i *
1990 1999 2000 2001 2002 2003 2004 2335
7.3 8.6 13,5 11.3 13.5 14.6 17.0 17.7
3.8 2.5 7.5 7.3 7.2 8.0 8.0 8.4
4.8 4.3 9.8 9.4 9.2 9.7 10.5 10.9
5.0 4.7 9.2 3.4 9.0 9.4 10.5 10.2
8.8 8.8 8.8 6.8 6.8 6.8
1.5 1.5
Ye,n
*Phase I standard is an
complex model
Figure 6
% Emission Reduction (from baseline gasoline)
NOx Performance of Winter RFG
(Based on Phase II Complex Model)
18 -
12 -
8 -
6-
2 -
Best Area
» Worst Area
Retail Avg.
Reporting Avg
Phs II Averaged Std
Phs I Averaged Std *
H
' " " --
^^L_j_^s^~
_ -~
P - i i ; -
1998 1999 2000 2001 2002
10.1 93 114 11.2 11.4
4.9 1.4 5.0 4.4 3.7
6.4 5.7 7.6 7.4 7.2
6.1 5.5 5.9 6.1 6.8
1.5 15 1.5
1.7 1.7
2003
11.5
4.7
7.6
74
1.5
2004
13.1
8.8
10.1
9.7
15
2005
176
7.9
11.6
10.5
1.5
Year
*Phase I standard is
an approximation
based on the Phase II
complex model
Figure 7
21
-------
Toxics Performance of Summer RFC
(Based on Phase II Complex Model)
35 -
20 -
15 -
Best RFG Area
» Worst RFG Area
Retail Avg.
Reporting Avg
Phs [I Averaged Std
Phs I Averaged Std *
^^_
»
1998
31.6
20.9
2B.O
29.1
14
?*
1999
32.6
20.3
27. B
289
14
/'
2000
366
29.7
34.3
34.0
21.5
..
2001
35.1
27.8
33.4
33.0
21.5
4
i
2002
34.7
28.2
33.3
334
21.5
2003
35.3
29.B
334
337
21.5
.
2004
36.1
28.8
32.7
33.5
21.5
1
{
2005
34.4
28.5
32.3
32.0
21.5
"Phase I standard is an
approximation
based on the Phase II
cornolex model
Figure 8
c
8
o>
.E
1
J
c
ija
o
W
C
'«
'£
UJ
25-
Best RFG Area
» Worst RFG Area
Retail Avg.
Reporting Avg.
Phs II Averaged Std
Phs. 1 Averaged Std*
1
199B
28.8
19.8
25.1
25.0
17.5
Toxic;
Basel
-B^*
f
1999
28.4
19.6
24.1
24.6
17.5
> Perfc
lonP
.
'~V~
2000
31.1
22.0
266
25.3
21.5
rman<
lase II
!
2001
29.5
22.4
264
25.3
21 5
Ye
:eof V
Comf
?
i
2002
28.8
22.5
260
25.5
21.5
.
Winter
lexM
i
2003
29.7
21.9
259
25.6
21 5
RFG
odel)
i
i
2004
29.4
223
285
26.8
21.5
j
2005
29.6
22.5
273
26.7
21.5
"approximation
based on phase [I
Figure 9
22
-------
RFC producers adjusted several of the emission-related properties of their RFC blends to produce
Phase II RFC. Figures 10 through 12 show the composite effect of all (1999) Phase I to (2000) Phase II
property changes on Complex model emissions, using average property value estimates from RFC
surveys, as well as the individual effects of changing one property while holding all others constant. This
provides a good indication of the relative contribution of each property change to the overall Phase I to
Phase II emission performance change. (Because of the non-linear mathematical structure of the
Complex Model, the individual property effects do not precisely add up to the composite effect, and the
emission performance of this "average RFC" is not precisely the same as the average performance of all
RFC blends.) Since the performance standards, and consequentially, the emission-related properties
changed more substantially for Summer RFC, only Summer graphs are presented. Each bar is labeled
with the estimated average parameter values for 1999 and 2000.
Effect of RFG Parameter Changes on Complex Model
VOC Emissions - Summer '99 to Summer '00
summer99 to
sumrnerOO composite
E200(%)49.Bto47.9
OLEF1NS (vol%)
10.82 to 9.36
BENZENE (vol%) 0.72
to 0.60
OXTGEN 2.25 to 2.30
E300(%)B3.1toB4.9
SULFUR (ppm) 204 to
127
AROMATICS (vol%)
24.95 to 19.45
RVP (psi) 7.62 to 6.77
-160 -14D -12D -100 -80 -60 -40 -20 0 20
Change in Emissions (mg per mile)
Figure 10
23
-------
Effect of RFC Parameter Changes on Complex Model
NOx Emissions-Summer '99 to Summer '00
Lower emissions (better perforrnanc
-90 -80 -70 -00 -50 -40 -30 -20 -10 0
Change in Emissions (mg per mils)
Figure 11
Effect of RFG Parameter Changes on Complex Model
Toxics Emissions-Summer '99 to Summer "00
-8 -5 -4 -3 -".' -1
Change in Emissions {mg per mile)
Figure 12
24
-------
RVP reductions were the primary means of meeting the more stringent Phase II VOC
performance standards. RVP effects on emissions are considered in the Summer version of the Complex
Model only. The Complex Model estimates both exhaust and non-exhaust VOC emissions and adds them
together. Although reducing RVP lowers exhaust VOC emissions, the RVP parameter primarily affects the
non-exhaust (i.e. evaporative emission) portion of the Complex Model, and is the only gasoline parameter
affecting its non-exhaust VOC estimates. Reductions in fuel aromatics and sulfur content, affecting
exhaust VOC emissions, contributed to total VOC reductions. (Figure 10 shows two sets of bars because
the non-exhaust portion of the Complex Model varies by geographic VOC Control Region. To simplify this
"property effects" analysis EPA used a single set of property averages based on data from both regions.)
Sulfur and aromatics reductions were the primary changes responsible for Phase I to Phase II
NOx performance improvements. Phase I to Phase II sulfur reductions were larger in Summer RFC, and
this sulfur reduction appears to be the most important single parameter contributing to Summer RFC NOx
performance improvement. However, it is clear that reductions in aromatics and olefin content had non-
negligible effects on Summer NOx performance improvement
Several parameter changes contributed to toxics performance improvements from Phase I to
Phase II. This analysis, based on property estimates from RFC Surveys indicates that reductions in
aromatics content were most important, but other parameter changes, such as sulfur and benzene
reductions, contributed significantly to toxics performance improvements. (Since the reporting system
data estimates a smaller 1999 to 2000 change in aromatics content but almost identical sulfur and
benzene changes it is possible that a similar analysis using reporting system data would conclude that the
aromatics content reduction was a significant but not primary contributor to toxics performance
improvement.)
The Complex Model estimates emissions of benzene, formaldehyde, acetaldehyde, 1,3-butadiene
and polycyclic organic matter and adds them together to produce a toxics emission estimate. Exhaust
benzene emissions are the largest component of Complex Model exhaust toxics, and exhaust benzene
reductions were the largest component of the 1999 to 2000 toxics emissions reductions (see the
Emissions chapter for this analysis.) The Summer model considers evaporative benzene emissions, as
well. As one would expect, both exhaust and non-exhaust benzene emissions are directly related to
gasoline benzene content, although other gasoline parameters also affect benzene emissions. There was
a sharp decrease in Summer RFC benzene content between 1999 and 2000, even though the benzene
content standard for RFC did not change (see the Benzene chapter). Additionally, RVP reductions, which
were necessary to meet the more stringent Phase II VOC standards, also contributed to Summer toxics
emission reductions because reducing RVP reduces evaporative benzene emissions.
The graph for parameter change effects on toxics emissions shows the oxygen contribution
change from individual oxygenates because, unlike VOC and NOx Complex Model calculations, toxics
calculations are a function of the type(s) of oxygenate used as well as the oxygen content of the gasoline
formulation. Oxygen and oxygenate changes between 1999 and 2000, and consequentially, their effects
on emissions, were very small.
25
-------
Changes in RFC from 2000 to 2005
Additional regulatory requirements have been imposed on both RFC and conventional gasoline
since the beginning of the Phase II program in 2000. These requirements have and will influence the
properties, composition and emissions performance of RFC.
One direct change was made to the Phase II RFC performance standards. This change involved
a small relaxation in the VOC performance standards applicable to certain ethanol-oxygenated RFC
supplied to the Chicago and Milwaukee areas. While some decline in VOC performance in these areas
was expected as a result of this change the scope and magnitude of this change was limited and
influence on overall average RFC trends is small.
Figure 5, which estimates the overall VOC performance trend, suggests that RFC VOC
performance may have declined very slightly from 2000 to 2005. This estimate of a slight decline could
easily be attributable to imprecision or inaccuracy in the estimates. More areas declined in VOC
performance than improved, but the estimated performance changes in either direction were generally
small (Figure 13). However, it is apparent that the greatest performance declines occurred in the Chicago
and Milwaukee areas. Since VOC performance declines in these areas were expected as a result of the
standard change it is unlikely that the observed declines in VOC performance in these two areas are the
result of statistical or measurement error.
Change in VOC Performance from 2000 to 2005-Summer RFC
Surveys
St. Louis, MO (29.6,31.0)
Rhode Island (27.6,28.4)
Springfield, MA (27 8,28.2)
Atlantic City, NJ (27.8,28.1)
8 Sussex County, DE (27.7,27.9)
g Baltimore, MO (29.3,29.4)
~ Phila.-Wilm, DE-Trenton, NJ (27.9,28.0)
~ Richmond, VA (29.4 ,29 4)
.2
= Poughkeepsie, NY (27.7,27.7)
.= Washington, D.C -area (29 6,29.5)
| Covington, KY (28.0,27 7)
| NY-NJ-Long Is.-CT (28.0,27.7)
| Norfolk-Virginia Beach, VA (29.7,29.3)
| Louisville, KY (28.5,28.0)
Boston-Worcester, MA (28.1 ,27 5)
« Hartford, CT (28.3,27.7)
u Dallas-Fort Worth, TX (29.4 .28.8)
«5 Houston-Galveston, TX (29.9,29.1)
|j Warren County, NJ (28 1,273)
Manchester, NH (28 6,27 4)
Portsmouth-Dover, NH (28.2,27.0)
Miiwaukee-Racine.WI (28.3,26.7)
Chicago-Lake Co., IL, Gary, IN (28.4,26.7)
'
i
c
1
M
M
^H
1HB
^^^^^m
i
^H
^m
^m
^m
^m
^m
^H
^
H
1
H
I !
! -1.5 - -0.5 0 0.5
1.5 2
Change in % Reduction fioin Baseline
Figure 13
26
-------
Perhaps the most significant post-2000 regulatory change is the Tier 2 gasoline sulfur program,
applicable to both RFC and conventional gasoline, which requires refiners to reduce sulfur levels
incrementally. In 2005 (with some exceptions), refiners and importers were required to comply with a 30
ppm annual average sulfur standard and a 300 ppm per gallon cap. The cap is reduced to 80 ppm for
2006 and later years. Both the RFC requirements and the Tier 2 gasoline sulfur requirements have
resulted in substantial reductions in gasoline sulfur levels. However, there is an important difference
between these sulfur reductions.
As indicated above, the RFC program did not mandate sulfur reductions, but set performance
standards based on the Complex Model. RFC producers, through their refining and blending processes,
are able to adjust various Complex Model input properties, including sulfur, in order to achieve the
required emission reductions. As previously noted, lowering gasoline sulfur content reduces NOx, exhaust
toxics and exhaust VOC emissions. Consequently, the more stringent Phase II performance standards for
these pollutants resulted in sulfur reductions from Phase I to Phase II RFC, especially in Summer RFC
(Figures 1 and 2).
The Tier 2 gasoline sulfur standards are required primarily for the purpose of enabling
technology. Gasoline with reduced sulfur content is necessary to enable the emission control systems in
Tier 2 vehicles to be fully effective. Phase-in of very stringent Tier 2 tailpipe emission standards,
applicable to all passenger cars and light trucks, began in 2004 along with a company-wide annual
average sulfur requirement of 120 ppm. The company-wide 120 ppm standard is comparable to the
average sulfur content in Phase II Summer RFC in 2000 and is substantially lower than the sulfur content
of Winter RFC. Thus, Tier 2 sulfur requirements have and will lower RFC sulfur content, and this trend is
apparent in the retail and production data averages shown in Figures 1 and 2. (Information on CG and
CG/RFG combined sulfur averages is available elsewhere in this report.)
Since the Tier 2 sulfur standards reduce sulfur levels below those needed to meet RFC
performance standards, and the Complex Model predicts that emission performance improves with lower
sulfur, the emissions performance of RFC could potentially improve as a result of the Tier 2 requirements.
Thus, emissions of these pollutants from older vehicles, as well as "Tier 2" vehicles could be lowered by
this sulfur reduction. However, RFC producers could also change their RFC formulations such that other
gasoline property changes would offset the Complex Model emission benefits of the required sulfur
reduction. This first effect clearly dominated RFC NOx emission performance, while the strongest
evidence of this latter effect is in RFC VOC performance. More generally, the interaction between the
RFC requirements and the Tier 2 sulfur reductions should be better understood as additional years of
data are analyzed.
The estimates of average NOx, VOC and toxics performance shown in Figures 5 through 9
indicate clear improvements in average NOx performance between 2000 and 2005, but do not show
improvements for VOCs. Summer and winter toxics changes were mixed, with a small decrease in
summer performance and a small increase in winter performance. Unlike VOC (with only a summer
standard) and NOx (where summer and winter standards differ), compliance with RFC toxics performance
standards and other toxics-related regulatory requirements is determined on an annual average basis.
Thus, improvements in Winter RFC toxics performance could have compensated for decreases in Summer
RFC toxics performance, and this may have been a factor influencing the directionally opposite seasonal
toxics performance trends.
EPA compared 2000 to 2005 RFC in the same manner as previously described for 1999 to 2000
RFC. Figures 14 through 18 show the composite effect of all 2000 to 2005 Phase II property changes on
Complex model emissions, using average property value estimates from RFC surveys, as well as the
individual effects of changing one property while holding all others constant.
27
-------
Effect of RFC Parameter Changes on Complex Model
VOC Emissions-Summer '00 to Summer '05
Change in Emissions (tng per mile|
summerOO to
:r05 composite
RVP (psi) 6.77 to 6.92
ARUWATICS (vol^
19.45 to 21.07
E300 (%) 84 9 to B4.3
BENZENE (vol%) 0 60
to D.72
O;:"vGEN 2 JO to 2 59
E20Q(%)47.9to48.8
OLEFINS (vo1%) 9 36
to 10.77
SULFUR (ppm) 127 to
Figure 14
Effect of RFG Parameter Changes on Complex Model
NOx Emissions-Summer '00 to Summer '05
-40 -30 -20 -10 0
Change in Emissions (mg per mile)
AROMATICS (vol%)
19.45 to 21.07
E300(%)34.9tO
843
BENZENE (vol%)
0.60 to 0.72
OXYGEN 2.30 to
2.59
SULFUR (ppm)127
to 69
Figure 15
Effect of RFG Parameter Changes on Complex Model
NOx Emissions-Winter '00 to Winter '05
winterOQ to winter05
composite
AROMATICS (vo\V,
18.15to 18.92
E2DO (%) 56 6 to 56 9
E300(%)86.1 to 85.6
BENZENE |vo!%) 0.65
to 0.70
OXYGEN 2.30 to 2 B1
OLEFINS (vol%) 9.9
to 9.50
SULFUR (ppm) 192 to
-70 -BO -50 -40 -30 -20 -10 0 10
Change in Emissions {mg per mile)
Figure 16
28
-------
Effect of RFG Parameter Changes on Complex Model
Toxics Emissions-Summer '00 to Summer '05
summerOD to
summerOS composite
-05 0 0.5 1 1.5
Change in Emissions {mg per mile)
Effect of RFG Parameter Changes on Complex Model
Toxics Emissions- Winter'OO to Winter '05
.
ij
E
£
|
Lower emiss ens (better
1
[~
-3 -2.5 -2 -1 5 -1 -0.5
-
Z
I
_
_
_
.
1
0 0
5
winterGO to winterOS
composite
MTBE {wt% oxygen)
1.49 to 1.11
AROMATICS (vol%)
181510 1892
BENZENE (vol%) 0.65
to 070
TAME (wt°/o oxygen)
Q.moO.05
E30D(%)aa.1to856
ETBE (wl% oxygen)
0.00 to 0.00
OLEFINS (vol%) 9.96
to 9.50
E200(%)56.6to569
Ethanol (wt% oxygen)
0.68101.46
SULFUR (ppm) 192 to
BO
1.5
Change n Em ssions (mg per mile)
Figure 17
Figure 18
Figure 14 shows that the VOC emission reduction benefits of lower sulfur (from 127 to 69 ppm)
and increased olefin content (which lower exhaust VOC) were offset by an increase in RVP (from 6.77 to
6.92 psi), which primarily increases evaporative VOC, and by other property changes which increase
exhaust VOC. Comparison of the scale of this chart with the chart showing 1999 to 2000 property effects
(figure 10) shows that, according to the Complex Model, the impact of any of the 2000 to 2005 property
changes on VOC emissions is very small compared to the overall Phase I to Phase II emission change,
which was primarily due to an RVP reduction (from 7.62 to 6.77 psi). The sulfur reduction between 2000
and 2005 may have facilitated some "trade off" between evaporative and exhaust VOC emission
reduction, but, through 2005, the Tier 2 regulation has not had a substantial effect on overall RFG VOC
performance (figure 5). Figure 14 also shows VOC reduction due to increase in RFG oxygen content.
This oxygen content increase is related to an increase in ethanol use. (This is discussed later.)
Figures 15 and 16 show that the Tier 2 sulfur requirements have clearly helped RFG's NOx
performance. As expected, the benefit is greater in Winter RFG because of the larger sulfur reduction
(192 to 80 ppm). Summer sulfur NOx benefits have also been offset to a greater extent by other
property changes. However, the overall trend shows improved NOx performance for both Summer and
Winter RFG between 2000 and 2005 and a substantial improvement in Winter performance between 2003
and 2005 (Figures 6 and 7).15
Figures 17 and 18 show that the toxics reduction benefits of lower sulfur have been countered by
emission increases due to increased aromatics and benzene content, resulting in a slight decline in
15 Beginning in 2007, the NOx performance standards were, with some exceptions, eliminated for RFG
refiners and importers. The gasoline sulfur program will become the sole regulatory mechanism used to implement
gasoline NOx requirements. See 82 FR 8427.
29
-------
Summer toxics performance and a slight improvement in Winter performance. As explained earlier, the
toxics graphs show oxygenate-spedfic effects because oxygenate type affects toxics emissions. These
graphs indicate that there is a toxics emission reduction associated with an increase in oxygen coming
from ethanol and a toxics emission increase associated with a decrease in oxygen coming from MTBE and
TAME. It is important to remember that all oxygenates have certain toxics emission reduction benefits
and that the increase in ethanol is related to the decrease in MTBE and TAME. The toxics emission
benefit lost from decreased use of these ethers in RFC is at least partially offset by the benefit gained
from increased use of ethanol in RFC.
Another post-2000 regulatory change affecting RFC (and conventional gasoline) is the Mobile
Source Air Toxic (MSAT) rule (2001). This rule requires refiners and importers to produce or import
gasoline with toxics emissions that are no greater than the toxics emissions of gasoline they produced or
imported during the period 1998-2000. Data collected by EPA indicates that RFC (and conventional
gasoline) often over-complied with toxics emission requirements. The MSAT rule is intended to preserve
this over-compliance in order to ensure that toxics emissions do not increase above current levels. Thus,
much RFC currently is subject to more stringent toxics emission performance requirements than the Phase
II RFC toxics emissions performance requirements. It would be difficult to isolate the effect of the MSAT
rule on changes in the properties and emission performance of RFC, but it is clear that the MSAT rule
could interact with other requirements and economic factors to affect the composition and performance of
RFC.16
An additional regulatory requirement which has significantly impacted the composition of RFC is
the oxygen content requirement. The Clean Air Act (CAA) required that RFC contain 2 weight percent
oxygen, until the Energy Policy Act of 2005 [in '1504(a)] amended the CAA to remove this requirement.17
This requirement was effective when RFC was introduced in 1995. RFC producers satisfied this
requirement by blending certain organic compounds, known as oxygenates into their gasoline. These
oxygenates fall into two general categories; ethers and alcohols. The most common ether used was
Methyl tertiary Butyl Ether (MTBE) although small quantities of other ethers, primarily tertiary-Amyl Methyl
Ether (TAME), were sometimes used. The only alcohol used in any significant quantity was ethanol.
Although the RFC oxygen content requirement did not change during the time
span considered in the report, it significantly affected composition of post-2000 RFC. A number of studies
have detected MTBE in ground water throughout the country; in some instances these contaminated
waters are sources of drinking water. While most of these detections have been well below levels of public
health concern, low levels of MTBE can make drinking water supplies undrinkable due to its offensive taste
and odor. In 1998, the EPA Administrator appointed a Blue Ribbon Panel to investigate the air quality
benefits and water quality concerns associated with the use of oxygenates in gasoline. The Panel, which
issued its report in September 1999, agreed broadly, but not unanimously, that in order to minimize
current and future threats to drinking water, the use of MTBE should be reduced substantially (Blue Ribbon
Panel, 1999). Subsequently, certain states have acted to ban or limit the use of MTBE in gasoline. This
15 EPA has finalized an "MSAT2" regulation to replace the current MSAT requirements with a 0.62 volume
percent average benzene requirement beginning in 2011. See 72 F.R. 8427
17 Removal of the oxygen requirement became effective with regulatory changes on April 24, 2006 for
California gasoline, and May 5, 2006 for gasoline nationwide. See 71 F.R. 8965 and 71 F.R. 26419.
30
-------
has resulted in a decreased use of MTBE and an increased use of ethanol in RFC. This switch from MTBE
to ethanol affects the composition of RFC and potentially affects the Complex Model emissions
performance of RFC. (There are a number of issues relating to the overall emissions impact of ethanol-
oxygenated gasoline, and the ability of the Complex Model or other emission models to address these
impacts. A discussion of these issues is beyond the scope of this report, and, unless otherwise stated,
emissions performance refers to the Complex Model.)
Figures 19 through 22 show 2000 to 2005 changes in average ethanol and MTBE content (in
weight %), for the various RFC areas surveyed. In some areas where large changes occurred, the change
reflects a transition from all-ether to all-ethanol RFC. In other areas, the change reflects some ethanol
usage in 2000 and increased ethanol usage in 2005. In the case of the NY-NJ-Long Island-CT RFC area,
New York and Connecticut enacted MTBE bans, but New Jersey did not. The charts include several areas
in California where RFC was required under federal law. The federal oxygen content requirement, as well
as California's gasoline requirements, applied to this gasoline.
Change in Ethanol Content from 2000 to 2005-Summer
RFG Surveys
Hartford, CT (0.00, 10.33)
Poughkeepsie, NY (0.00, 10.29)
St. Louis, MO (1.35, 10.53)
Louisville, KY (2.21, 10.36)
C ( , ) _
g , ( , ) _
9 1 , ) _
Covington.KY (5.95, 10.49)
§ Springfield, MA (0.01 ,2 47)
Boston-Worcester, MA (0.00,0.33)
§ Rhode Island (0.00,0.32)
| Chicago-Lake Co., IL, Gary, IN (10 12,10.39)
| Milwaukee-Racine, Wl (10.05, 10.29)
j Phila.-Wilm. DE-Trenton. NJ (0.00,0.01)
f Manchester, NH (0.00,0.01)
| Warren County, NJ (0,00,0.01)
Norfolk-Virginia Beach. VA (0.00,0.00)
Dallas-Fort Worth. TX (0.00,0.00)
Baltimore, MD (0 00,0.00)
Houston-Galveston. TX (0.00,0.00)
Richmond. VA (0.00,0.00)
Washington, D.C.-area (0.00,0.00)
Sussex County, DE (0,00,0.00)
Atlantic City, NJ (0.00 ,0.00)
Portsmouth-Dover, NH (0.00,0.00)
^^^^^zr^^
^^m
3
D
J
J
I
1
0 2 4 6 3 ID 12
Change in ethanol weight percent
Change in MTBE Content from 2000 to 2005-Summer
RFG Surveys
Sussex County, DE (9 11, 9.99)
Rhode Island (8 .53, 10.28)
Norfolk-Virginia Beach, VA (10. 33, 10.69)
Baltimore, MD (10.12,10.46)
Milwaukee-Racine, Wl (0.00,0.02)
Richmond, VA(10.53,10.54)
Chicago-Lake Co., IL, Gary. IN (0.01.0.00)
Washington, D.C.-area (10.43,10.17)
Dallas-Fort Worth, TX (10. 16,9.89)
g Houston-Galveston. TX (1 1 .01 ,10.57)
" Phila.-Wilm, DE-Trenton, NJ (1 1 .50,1 1 .04)
* Boston-Worcester, MA (10.59, 10.03)
-* Warren County, NJ (10.94, 10.32)
LU Atlantic City. NJ(1 1.47, 10.80)
S Manchester.NH(11.11,1uM8)
| Portsmouth-Dover, NH(1 1.82, 10.67)
< Springfield, MA (9 90,7.82)
Covington.KY (3.49.0.00)
NY-NJ-Long Is -CT (10.26,4.54)
Louisville, KY(S 16,0.00)
St. Louis, MO (9.07.0.02)
Hartford, CT (9.21, 0.03)
Poughkeepsie. NY(10 21,0.05)
Los Angeles, CA (10.94.0.00)
San Diego, CA(11.32,0.01)
Sacramento Metro, CA (1 1 .47.0.00)
iM
'
^
^^^m
m
o
3
]
^^^
^^^H
^- H
4 -12 -10 -8 -6 -4 -2
;
i li.ni.l- in MTBE weight %
Figure 19
Figure 20
31
-------
Change in Ethanol Con
RFC
Hartford, CT (0.43 ,10.48)
Poughkeepsie. NY (0.71 ,10-45)
Louisville, KY (2.55,10.61)
NY-NJ-Long Is -CT (0.36,7 20)
Sacramento Metro, CA (0.00,5.97)
San Diego, CA (0.00 ,5 92)
Los Angeles, C A (0.00,5.77)
St Louis, MO (4.73,10.30)
Covington.KY (6.76,10.63)
g" Springfield, MA (0.72,2. 99)
Warren County, NJ (0.01,1.67)
§ Milwaukee-Racine, Wi (10.51 ,10.62)
| Boston- Worcester, MA (0.30,0. 36)
"2 Baltimore, MD (000,0.00)
5 Houston-Galveston, TX (0.00,0.00)
Washington, D.C.-area (0.00,0 00)
5 Richmond, VA (0.00,0. 00)
Manchester, NH (0.00,0.00)
Sussex County, DE (D 00,0 00)
Portsmouth- Dover, NH (0.00,0.00)
Dallas-Fort Worth, TX (0,01 ,0.00)
Norfolk-Virginia Beach, VA [0.01 ,0.00)
Chicago-Lake Co., IL, Gary, IN (10.66,10.64)
Rhode Island (0.48,0 44)
Phila.-Wilm, DE-Trenton, NJ (0.18,0.04)
Atlantic City.NJ (054,0.00)
entfrom 2000 to 2005-Winter
Surveys
i
^^
^H
^^
^H
^^^^^^^^
^^^^^:
^^^
I
^H
^^
^H
1
:
2 0 2 4 B 8 10 1
2
Change in ethanol weight percent
Change in MTBE Content from 2000 to 2005-Winter RFC
Surveys
Sussex County, DE (8.12,9.76)
Atlantic City, NJ (9.87, 10.29)
Houston-Galveston. TX (10.52,10.80)
Manchester, NH (10.05,10.32)
Washington, D.C.-area (9.71 ,9.94)
Richmond. VA (9.43 ,9.62)
Boston-Worcester, MA (9.80 ,9.98)
Baltimore, MD (9.56,9.70)
Milwaukee-Racine, WI (0.00,0.00)
g Chicago-Lake Co., IL, Gary, IN (0.00,0.00)
°, Phila.-Wilm, DE-Trenton, NJ (10.34,10.25)
| Dallas-Fort Worth, TX (9.89,9.62)
-' Rhode Island (9.81,9.45)
m Norfolk-Virginia Beach, VA (9.91 3.54)
5 Portsmouth-Dover, NH (10.96,10.51)
1 Springfield, MA (9.39,7.25)
< Warren County, NJ (10.32,8.09)
Covington.KY (3.67,0.00)
St. Louis, MO (5.59 ,0.03)
NY-NJ-Long Is.-CT (9.37,3.24)
Louisville, KY (7.26,0.00)
Hartford, CT (9.01 ,0.09)
oug eepsie, . , .
Sacramento Metro, CA (11.34 ,0.00)
Los Angeles, CA (12.12,0.00)
-1
^
1
1
1
1
^m
^m
i
f
I
^^^
^^^H
4 -12 -10 -8 -6 -
4 -2 [
2 <
Change in MTBE weight %
Figure 21
Figure 22
The increased use of ethanol has resulted in a higher percentage of RFC being oxygenated with
ethanol, a larger fraction of total oxygen in RFC being supplied by ethanol and an increase, on average, in
the total oxygen content of RFC. Figures 23 through 26 show estimates of the fraction of RFC oxygenated
with ethanol, by year and season, through 2005. Figures 27 through 30 show estimates of the total
oxygen (in weight %) in RFC, and the amounts of this oxygen supplied by ethanol and other oxygenates.
The total oxygen content has increased along with the ethanol share because most ethanol-
oxygenated RFC supplied outside of California contains about 10 volume percent ethanol, providing about
3.5 weight percent oxygen, even though RFC was only required to have only 2 weight percent oxygen.
This occurred, in part, because, until a recent legislative change (American Jobs Creation Act, 2004), a
federal excise tax exemption available to ethanol was greatest when ethanol was blended at 10 volume
percent and pro-rated for blending at 5.7 or 7.7 volume percent (see the Oxygenates chapter).
Additionally, MTBE must be used at around 11 volume percent in order to provide the required 2 weight
percent oxygen. Thus, when ethanol is used to replace MTBE, blending at 10 volume percent replaces
much of the lost MTBE volume.
32
-------
% of Summer RFC Oxygenated with Ethanol and Ethers (Excluding CA}
100% -
90% -
80%
70% -
60%
50%
40%
30% -
20%
Mm
MTBETAMh stbei
MM
Figure 23
% of Winter RFG Oxygenated with Ethanol and Ethers (Excluding C A)
1995 1996 1997 !993 1999 2000 2001 2002 2003 2004 2005
MTBETAML ^V
19% 16% 19% 24% 38% 40'
2% 81% 7B%62% GO'
% of Summer RFC Oxygenated with Ethanol and Ethers (Including Federal RFG Areas in CA}
Figure 24
% of Winter RFG Oxygenated with Ethanol and Ethers (Including Federal RFG Areas in C A)
Figure 25
Figure 26
33
-------
Summer RFG Oxygen Content and Source-(Excluding CA)
Summer RFG Oxygen Content and Source-(lncluding Federal RFG Areas in CA)
Figure 27
Figure 28
Winter RFG Oxygen Content and Source-(Excluding CA)
Figure 29
Winter RFG Oxygen Content and Source-(lncluding Federal RFG Areas in CA}
Figure 30
34
-------
Conventional Gasoline Trends
EPA used the gasoline property data that it received from conventional gasoline (CG) refiners and
importers to investigate property trends. EPA's CG analysis covered the period between 1997 (or 1998
for some properties) and 2005. As noted in the General Methodology chapter, while EPA receives retail
survey as well as refiner/importer data for RFC, it receives only refiner/importer data for CG.
Consequently, the CG analysis in this report is less extensive than its RFC analysis.
EPA's Anti-Dumping regulations, affecting CG, and its Mobile Source Air Toxics regulations
affecting both CG and RFC, limited or prevented CG deterioration, but did not necessarily force
substantial changes in CG's emission-related properties or improvements in CG's emission qualities. The
Anti-Dumping standards changed between 1997 and 1998, with the transition to the Complex Model, and
the model for CG exhaust toxics and NOx emission calculations needed to determine CG compliance
changed in 2000 with the transition to the Phase II Complex Model. However, in sharp contrast to
simultaneous changes in RFC requirements that caused significant RFC composition changes, these CG
changes apparently had minor impacts on CG composition. Moreover, the Anti-Dumping regulations
apparently performed as intended, since CG emissbns did not show significant or sustained increases
when the more stringent Phase II RFC regulations took effect.
The Tier 2 gasoline sulfur regulations applicable in 2004 and later years impose the same sulfur
content requirements on CG and RFC. However, CG sulfur content decreased more precipitously since
other regulatory factors affecting RFC had already resulted in sulfur reductions.
To summarize, EPA's analysis shows that certain CG gasoline properties have fluctuated from
year-to-year within the time period addressed in this report. These fluctuations may be linked to both
regulatory and non-regulatory factors. Specific instances where regulatory factors may have influenced
CG properties or emission qualities are identified in various chapters within this report. However, the
sulfur reduction requirement was the only regulatory factor that clearly exerted a large influence on CG
during this time period.
Figure 1 shows EPA's CG gasoline sulfur content estimates for each year between 1997 and 2005.
The large change from 2003 to 2004 is due to the Tier 2 gasoline sulfur requirements and, as explained
in the Sulfur chapter, these requirements may also have motivated some sulfur reductions prior to 2004.
Average Sulfur Content of Conventional Gasoline
(from Batch Reports)
290
Reporting Year
Figure 1
35
-------
Figures 2 through 5 show CG exhaust toxics and NOx average emission levels from 1998 through
2005 calculated with the Phase II Complex Model. CG must comply with Complex -model based exhaust
toxics and NOx "milligram per mile" emission standards. While sulfur reductions were required to enable
new technology vehicles to meet Tier 2 emission standards, the reductions also result in "cleaner" CG
because these sulfur reductions result in lower Complex Model emission levels. This is evident in the
2003 to 2004 emission improvements that occurred concurrent with the large sulfur decreases. This
suggests that CG sulfur reductions have and will reduce emissions in older vehicles as well. 18 These
Complex Model emission calculations are not intended to be accurate estimates of the overall emissions
rates or emissions changes that have occurred in each year as a result of gasoline property changes.
They are indicators of changes in the emissions qualities of gasoline over time without considering
changes in vehicle technology or emission standards.
Since the standards applicable to CG exhaust toxics and NOx are facility-specific and largely
depend on individual 1990 baselines, these graphs do not show CG standards. Instead, the graphs show
the Complex Model emissions rate of the seasonally-appropriate 1990 statutory baseline gasoline to
provide some basis for comparison. These statutory baseline gasoline properties and emission levels
were intended to be representative of the gasoline supplied in 1990. They are used to determine
compliance for refiners and importers without individual baselines, and also factor into CG compliance
determinations for individual refineries and importers supplying CG in excess of their 1990 gasoline
volumes.
The exhaust toxics and NOx emission levels in each season and year are lower than the 1990
baseline levels. Although the 1990 statutory baseline gasoline properties were intended to be
representative of the gasoline supplied in 1990, EPA does not have comprehensive data on CG properties
in 1990 comparable to the data collected in its Anti-Dumping reporting system. EPA has compared
properties and emission values of 1990 statutory baseline gasoline's to its CG estimates here and
elsewhere in this report to provide some frame of reference. While EPA's Anti-Dumping regulations did
not mandate substantial changes in CG composition or emissions characteristics these comparisons are
insufficient to establish that these statutory baseline gasoline properties were or were not representative
of 1990 gasoline. Regulatory changes other than the Anti-Dumping regulations, including volatility
requirements and winter oxygenated gasoline program requirements certainly impacted CG composition
between 1990 and 1997, the earliest year for which CG data were analyzed in this report. Various states
have adopted specific gasoline requirements (e.g., low RVP, low sulfur) and these "boutique" fuels are
included in EPA's CG data. Non- regulatory factors may also have caused changes in conventional
gasoline during this time period.
18 Beginning in 2007, NOx emission standards were, with some exceptions, eliminated for CG refiners and
importers. The gasoline sulfur program will become the sole regulatory mechanism used to implement gasoline NOx
requirements. See 82 FR 8427.
19 The properties of the Summer Baseline gasoline were specified in the Clean Air Act. EPA regulations
specified the properties of Winter Baseline gasoline. The term statutory baseline gasoline is used here to refer to
both the Summer and Winter.
36
-------
Phase II Complex Model Exhaust Toxics Emissions - Summer CG
(Based on Batch Data-Excluding Blendstocks)
Figure 2
Phase II Complex Model Exhaust Toxics Emissions - Winter CG
(Based on Batch Data-Excluding Blendstocks)
125
115 -
105 -
95
1993
1999
2000
2003
2004
2005
110.2
110.4
112.4
110.7
1096
1044
104.8
1990 baseline exh. topics 120.5
Year
Figure 3
37
-------
Phase II Complex Model NOx Emissions - Summer CG
(Based on Batch Data-Excluding Blendstocks)
1340 -
1320 -
1
11280 -
1260 -
1240 -
CG NOx emissions
1990 baseline NOx
+
i *
~ «
*
\
\
\
1998
1325.7
1340.0
1999
1331.4
20DD
1 333 7
20D1
1 332. 9
2D02
13275
2003
13255
2004
1252.6
2005
12505
Year
Figure 4
Phase II Complex Model NOx Emissions -Winter CG
(Based on Batch Data-Excluding Blendstocks)
1350 -
1300
CG NOx emissions
I Sfli.l baseline NOx
Figure 5
38
-------
Comparison with RFC
The data that EPA collects allow a comparison of CG and RFC Complex Model emissions on a
year-by-year basis. This is particularly interesting since CG emissions performance has improved as a
result of Tier 2 sulfur reduction requirements, and the sulfur levels in CG and RFC are converging. While
CG average sulfur levels in 2005 were still higher than RFC levels in that year, they were much lower than
the sulfur levels typically needed to comply with Phase II RFC emissions performance standards.
Figures 6 through 10 compare RFC and CG in several ways using the "percent reduction"
emission performance measures normally applied to RFC. EPA calculated both CG and RFC VOC, total
toxics and NOx emission reductions from 1990 statutory baseline gasoline. CG reductions from baseline
can be compared to RFC reductions from baseline and to RFC performance standards which are specified
as reductions from these 1990 baselines. (Each graph shows the "averaged" RFC performance standard.)
Summer VOC and Summer total toxics calculations and RFC standards depend on a geographic VOC
control region specification. For these two parameters, EPA compared CG to VOC Control Region 2 (i.e.
"Northern") RFC, and calculated CG emissions using the Region 2 version of Complex Model. RFC
emission reductions were also calculated relative to average CG emissions in each year. These results,
shown as dashed lines in each graph, allow direct comparison of RFC with its contemporary CG.
These graphs show that while CG emissions performance is improving relative to RFC, RFC
emission performance remained superior for each pollutant and season combination. Summer CG's VOC
performance has improved because sulfur reductions lower exhaust VOC emissions. (Sulfur reductions
do not lower non-exhaust VOC's, which are directly related to the Reid Vapor Pressure (RVP) parameter.)
Winter CG has met RFG's Winter NOx standard in each year analyzed. Summer CG is approaching RFG's
Summer NOx standard.
Comparison of Summer CG and RFC VOC Emissions
(Calculated from Batch Data with Phase II / VOC Region 2 Model)
0
m
o
3
O
£ 15
"c
H
o
m
Q- 1f|
5 -
/-
/- -
// ^- ..
/
/
/
* j
^^ -+
* * »- * -*
1998 1999 2000 2001 2002 2003 2004 2005
Year
* CGfrorn Baseline
RFGfroni Baseline
* RFGfrom CG
---*-- RFGStd. from Baseline
Figure 6
39
-------
Comparison of Summer CG and RFG Total Toxics Emissions
(Calculated from Batch Data with Phase II /VOC Region 2 Model)
40
35
30
I 25
T3
S. 20
o
I 15
10
5
0
* CG from Baseline
RFG from Baseline
*- RFG from CG
---w-- RFG Std from Baseline
1998 1999 2000 2001 2002 2003 2004 2005
Year
Figure 7
Comparison of Summer CG and RFG NOx Emissions
(Calculated from Batch Data with Phase II Model)
*CG from Baseline
RFG from Baseline
A- RFG from CG
" RFG Std (from Baseline)
1998 1999 2000 2001 2002 2003 2004 2005
Year
Figure 9
O
E
3
TJ
$. 15
I
Q- 1Q
Comparson of Winter CG and RFG Toxics Emiss
(Calculated from Batch Data with Phase II Mod
5"""!_
m~-
-A" ' ~" ^ * - _
*- ^.-,
s »
/
+ . M *.. ^. . *
1998 1999 2000 2001 2002 2003 2004 2005
Year
ons
el)
* CG from Baseline
RFG from Baseline
A- RFG from CG
---*-- RFG Std (from Baseline)
Figure 8
Comparson of Winter CG and RFG NOx Emissions
(Calculated from Batch Data with Phase II Model)
-CGfrom Baseline
-RFGfrom Baseline
RFG from CG
RFG Std (from Baseline)
1998 1999 2000 2001 2002 2003 2004 2005
Year
Figure 10
40
-------
Sulfur in gasoline originates in the crude oil used to produce the gasoline, and the sulfur content
of crude oil varies substantially. (Crude oil with low sulfur content is classified as "sweet crude".)
Historically, high sulfur content in gasoline has been considered undesirable for reasons unrelated to
vehicle emissions. Sulfur oxides formed during combustion may be converted to acids that promote
corrosion of engine parts and exhaust systems. ASTM Standard D4814, first published in 1988, limited
sulfur content in unleaded gasoline to 0.10 mass percent (equivalent to 1000 parts per million) to protect
against engine wear, deterioration of engine oil and corrosion of exhaust system parts.
More recently, gasoline sulfur content has been a concern because of emission-related effects.
Sulfur affects gasoline vehicle emissions primarily because it adversely affects catalytic converters.
Reductions in gasoline sulfur content can reduce NOx, exhaust toxics and exhaust VOC emissions in the
"1990 technology" vehicles used to develop the Complex Model. The phase-in of stringent Tier 2 tailpipe
emission standards began in 2004, along with the phase-in of the Tier 2 gasoline sulfur regulations.
Vehicles designed to meet these standards are particularly sensitive to sulfur and gasoline with reduced
sulfur content is necessary to enable the emission control systems in these vehicles to be fully effective.
Emission-related regulatory requirements have resulted in substantial sulfur reductions.
Gasoline is produced by blending several components, produced from different processes within a
refinery. The component from the Fluid Catalytic Cracking (FCC) unit typically has much higher sulfur
content than other components used in gasoline blending unless it is treated to remove sulfur. In the
past, refineries have largely relied on a process called hydro treating to reduce sulfur, when necessary.
However, conventional hydro treating results in a substantial loss of octane in the FCC gasoline. EPA's
Tier 2 sulfur reduction requirements have helped spur the development and implementation of
modifications or alternatives to conventional hydro treating. These processes allow removal of sulfur
while minimizing economically adverse consequences such as octane or volume loss.
EPA's RFC and Anti-Dumping regulations have imposed both direct and indirect limits on gasoline
sulfur content. The Tier 2 gasoline sulfur regulations also impose direct gasoline sulfur limits on RFC and
CG. The regulatory requirements that have affected gasoline sulfur content are discussed below.
The Simple Model, which was the basis for RFC emission performance standards through 1997,
did not consider the effect of sulfur on emissions, nor did it evaluate NOx emissions performance, which is
sensitive to gasoline sulfur content. However, the annual average level for sulfur was not allowed to
exceed a refinery's or importer's 1990 baseline level. Hence, although the RFC regulations prior to 1998
neither required nor rewarded sulfur reduction, the regulations prevented any substantial increase in
sulfur levels in RFC.
Between 1997 and 1998, EPA transitioned from the Simple Model to the Phase I Complex Model
for assessing the emission performance of RFC. This transition introduced several factors that affected
RFC sulfur content:
RFC was subject to NOx and VOC performance standards, as well as a toxics performance standard
and the new model recognized that lowering gasoline sulfur reduced emissions of all of these
pollutants.
41
-------
» Each RFC batch was required to have a sulfur content at or below a 500 ppm model range limit.
Phase II RFC, introduced in 2000, required additional NOx, toxics and VOC emission reductions.
EPA only required Phase I RFC to meet, on a per-gallon basis, the NOx performance of 1990 baseline
gasoline's with sulfur levels of 338 ( Winter) and 339 (Summer) ppm. However, for Phase II, EPA
required Summer RFC to have NOx emission performance 5.5 percent better on a per gallon basis, (or
6.8 percent better on average), than this Summer baseline gasoline. It is generally accepted that this
more stringent Phase II NOx performance standard necessitated a reduction in average Summer RFC
sulfur levels.
Prior to 1998, EPA prohibited the annual average sulfur level in CG from exceeding 125% of a
refinery's or importer's 1990 baseline. For 1998 and later, EPA prohibited annual average exhaust toxics
and NOx emissions from exceeding the refiner or importer's 1990 baselines for these emissions.
Additionally, the annual average CG sulfur levels were not allowed to exceed the greater of 1000 ppm
(the model range limit for CG) or the refiner or importer's 1990 sulfur baseline.
The Tier 2 sulfur reduction standards, applicable to both RFC and CG, require a phased reduction
in gasoline sulfur content beginning in 2004. The standards applying to most refiners and importers for
2004 are a 120 ppm corporate pool average and a 300 ppm per gallon cap. In 2005, the standards
include a refinery or importer average of 30 ppm as well as a 90 ppm corporate pool average and a 300
ppm per gallon cap. Since 2006, the standards include a 30 ppm refinery or importer average with an 80
ppm per gallon cap. The Tier 2 requirements contain certain exceptions to the above requirements for
small refiners and for "geographic phase-in " areas, so the full impact of the regulation on gasoline sulfur
levels will not be realized until later than 2006. However, the gasoline sulfur regulations also contain
various Averaging Banking and Trading (ABT) provisions, including mechanisms for credit generation for
sulfur reductions which occurred as early as year 2000. Thus, the Tier 2 requirements potentially
impacted sulfur levels prior to the phase in of the standards.
In summary, the Tier 2 requirements will result in substantial reductions in gasoline sulfur
content. The Tier 2 company-wide annual average sulfur requirement of 120 ppm, applicable in 2004, is
comparable to the average sulfur content in Phase II Summer RFC in 2000, before the impact of the Tier
2 sulfur regulations, and substantially lower than the sulfur content of Winter RFC. The Tier 2 sulfur
reduction requirements are more stringent than the sulfur reductions generally needed to comply with
RFC emission performance standards or with Anti-Dumping program regulations applicable to CG.20
Consequently, the Tier 2 requirements became the dominant federal constraint on gasoline sulfur levels
by 2004. (California's gasoline requirements and a few other state regulatory requirements applicable to
specific localities do restrict sulfur more tightly than the 2004 Tier 2 standards.)
Sulfur data for both RFC and CG are currently necessary to determine compliance with EPA's
regulations. Thus, refiners and importers have been required to submit this information to EPA's
RFG/Anti-Dumping reporting system for each batch of gasoline refined or imported. Additionally, since
1998, EPA has received RFC Survey data on the sulfur content of RFC sold at gasoline stations. (Since
sulfur was not part of the Simple Model, Compliance was not needed with survey standards until 1998)
20 In fact, beginning in 2007 the NOx emission performance and emission requirements were eliminated, with certain
exceptions, for RFG and CG since the mandated sulfur reductions made them unnecessary. See 82 FR 8427.
42
-------
Sulfur Trends
Overview-Reformulated Gasoline
Volume-weighted average RFC Sulfur levels, by year, are shown in Figures 1 (Summer) and 2
(Winter). Both reporting system estimates, by reporting year, and survey-based retail estimates, by
survey year are shown. Summer and Winter RFC sulfur levels exhibit clear downward trends. Both
Summer and Winter sulfur levels declined sharply between 1997 and 1998 with the transition from Simple
Model to Complex Model RFC. As noted, the need to comply with a 500 ppm sulfur limit, and the
introduction of new performance standards along with a model that considers the emission effect of sulfur
influenced this decline. EPA determined that about 17% of Summer and 12% of Winter RFC refined and
imported in 1997 exceeded the 500 ppm sulfur limit applicable in 1998 and later years, with some
batches exceeding 1000 ppm. Although EPA did not quantify the extent to which 1997 RFC exceeded the
NOx standard applicable to 1998 RFC, it is reasonable to assume that the 1998 NOx standard contributed
to the need for these 1997 to 1998 sulfur reductions. (Since sulfur has such a strong influence on
Complex Model NOx calculations, the effects of the range limit and the NOx standard on this sulfur
reduction are likely hard to separate.)
Average Summer RFC Sulfur Content
Figure 1
Average Winter RFC Sulfur Content
2
CO
250 -
200 -
150 -
10D -
50 -
Highest RFG Area
Lowest RFG Area
Retail Avg.
Production Avg
T » .
*! *^=*^_
. i
1
1997 1998 1999 2000
268 273 265
147 150 121
207 215 192
251 203 214 200
«
2001
235
98
182
185
Year
:
m
2002
276
81
183
184
IE
=N
'm
2003
247
77
167
164
*
2004
144
50
102
101
-^
2005
143
18
80
80
* Highest RFG Area
Lowest RFG Area
A Retail Avg
-* Production Avg
Figure 2
43
-------
Summer RFC shows a second sharp sulfur decline between 1999 and 2000 with the transition
from Phase I to Phase II RFC. Although Winter RFC sulfur levels declined as well, the change was much
smaller. The Summer sulfur decrease is largely attributable to the more stringent NOx performance
standard. Since the Winter NOx performance standard did not become more stringent, winter sulfur
declines may not have been driven by the need to comply with NOx performance standards. However,
NOx performance did improve, in part, as a result of these sulfur reductions (see the RFC Trends
chapter.)
Both Summer and Winter RFC show decreasing sulfur levels between 2000 and 2004. The first of
the phased-in Tier 2 sulfur standards necessitated sulfur reductions by 2004. Since the Tier 2 regulation,
through credit generation, provided some incentive for sulfur reduction as early as 2000, it may also have
affected sulfur levels in each of the interim years.
Figures 1 and 2 separately analyze changes in sulfur levels, by reporting and survey year, for
Summer and Winter gasoline. Regardless of the production date reported for a batch, a batch was
categorized as Summer gasoline if it was designated VOC-controlled, and Winter gasoline if it was
designated non-VOC-controlled. Survey data were categorized as Summer for surveys conducted
between June 1 and September 15, and Winter for surveys conducted before or after that date. Thus,
for any year, the Winter RFC in this analysis is largely a mix of gasoline produced before and after each
VOC season.
Quarterly averages more precisely show how sulfur levels varied with time. Figure 3 presents
reporting system average sulfur levels calculated by calendar quarter based on the production date
reported for each batch.
Average Sulfur Content by Production Quarter
(from Batch Reports-CG Blendstocks Excluded)
-CG >RFG -^--Combined
277 293 29£ 86 309 299 313 ffl£ 287 SOS OT - 2 I :. '":.' IE ":37 305 302 284 204
767 j40 Mi'ipnc j:i.:: ii-: 3 ID 300214216 (9£ 126 i -' 13£ f88 129 i 39 0 ! [8S 1 '- I BO 132 !8;118114147106 7!
Figure 3
The lowest line, representing the RFC averages, shows that, for the years 1997 and 2003, 4th
quarter sulfur averages were lower than 1st quarter, indicating that sulfur reductions needed to comply
with standard changes in 1998 and 2004 were implemented, in part, during the year preceding the
standard change. The timing of these sulfur reductions was likely influenced by concerns about
"downstream" compliance (i.e., compliance at locations in the distribution system subsequent to the
44
-------
refinery or import facility). The Complex Model sulfur range limit, which drove the 1997 to 1998 sulfur
reduction, was potentially enforceable as a downstream standard since it applied to every batch of RFC,
although EPA exercised enforcement discretion in this matter. Since the Tier 2 regulations also included
various downstream standards for sulfur, refineries and importers may have chosen to reduce sulfur
levels during late 2003 to ensure compliance with these standards, as well as to generate sulfur credits.
Additionally, there was uncertainty associated with the desulphurization technologies which would be
needed to meet the Tier 2 standards economically. Thus, refiner's schedules for selection and
implementation of new desulphurization processes, which were not fully proven technology when the Tier
2 regulations were finalized in 2000, also affected the timing of these sulfur reductions. For example,
this may have made early sulfur reductions infeasible for some refineries even if the ABT program
provided some incentive for such reductions. On the other hand, refineries probably had some incentive
to get these processes operational earlier than needed for downstream compliance in order to ensure that
there were no technical problems.
Figure 3 also shows the cyclical effect of the more stringent Summer NOx emissions performance
standard on Phase II RFC sulfur levels, with quarters 2 and 3 lower than quarters 1 and 4 within each
year of the Phase II RFC program. Even though the Tier 2 requirements presumably limited the rise in
sulfur between the 3rd and 4th quarters of 2003, the cyclical effect is still apparent.
by
Table 1 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These
averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information):
& bf
Season
Summer
Winter
Annual
PADD Average Value
I
II
III
I
II
III
I
II
III
(ppm)
91
78
73
118
93
101
107
87
87
Gasoline Volume
(gal)*
4,792,114,891
1,740,499,436
5,890,920,167
6,489,530,475
2,436,636,058
6,059,647,031
11,281,645,366
4,177,135,494
11,950,567,198
^Volumes exclude batches with missing values
Average Value
(ppm)
73
64
78
90
78
83
84
72
81
for this parameter
Gasoline Volume
(gal)*
4,430,958,220
1,844,913,833
5,690,766,967
7,081,940,665
2,718,751,541
5,766,524,033
11,512,898,885
4,563,665,374
11,457,291,000
Table 1
45
-------
Average Sulfur Content of Summer RFC Sold at Retail Stations-By PADD
Year
Figure 4
Average Sulfur Content of Winter RFC Sold at Retail Stations-By PADD
Figure 5
Additional Analysis and Observations-RFC
Data analyses pertaining to RFC sulfur are contained in the Appendix to this chapter. Several trends or
patterns are highlighted below:
Average sulfur content in premium grade RFC has been consistently lower than in regular RFC.
21
21 As noted, removing sulfur from the FCC gasoline component could result in an octane loss.
Consequently, it may seem counter-intuitive that higher octane premium RFC had lower sulfur content than regular
grade. However, the lower sulfur content in premium RFC depends on the percentages of other blending
components used to meet octane requirements. For example, refiners typically choose to blend more reformate, a
high octane and low sulfur blending component, into premium gasoline.
46
-------
Average sulfur content of Summer and Winter RFC sold at retail outlets decreased between 2000 and
2005 in all areas surveyed in both years.
Overview-Conventional Gasoline
Volume-weighted average CG sulfur levels, by reporting year, are shown in Figure 6. Summer
and Winter averages are shown separately on the same graph. Both Summer and Winter sulfur levels
decreased from 1997 to 1998, with the transition from annual average sulfur content standards to
Complex Model-based annual average exhaust toxics and NOx emission standards. Although both
standards were derived from refinery and importer individual baselines, the 1998 standards may have
acted to reduce gasoline sulfur since they required emissions equal to the refinery or importer's
compliance baseline, while the 1997 standards allowed annual average sulfur content up to 125% of the
refinery or importer's baseline. Additional analysis would be needed to confirm this effect.
Average Sulfur Content of Conventional Gasoline
(from Batch Reports)
Reporting Year
Figure 6
The Winter CG sulfur level decreased between 2002 and 2003, while the Summer level increased
slightly. This divergence of the Summer and Winter averages is a consequence of the timing of sulfur
reductions in advance of the Tier 2 standards. Similarly to RFC, the CG in this analysis was categorized
as Summer gasoline, regardless of production date, if it was designated as volatility-controlled, and
Winter gasoline if it was designated as non-volatility-controlled. Thus, Winter CG is largely a mix of
gasoline produced before and after each volatility control season. The need to comply with downstream
standards in 2004 would not have affected sulfur levels in volatility-controlled CG produced during 2003,
but would likely have affected sulfur levels in non-volatility-controlled gasoline produced late in 2003.
Figure 3 also shows average CG sulfur levels, by production quarter (upper line). CG sulfur content
dropped during 2003 with a large decrease between the 3rd and 4th quarters of 2003. Again, this is
consistent with the opportunity for sulfur credit generation, the existence of downstream standards under
the Tier 2 regulation and the need to implement new sulfur reduction technology.
47
-------
CG by
Table 2 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These
averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information):
2
Season
Summer
Winter
Annual
iw* & ;*
PADD Average Value
I
II
III
I
II
III
I
II
III
(ppm)
125
140
108
138
135
115
131
137
111
ge by
Gasoline Volume
(gal)*
3,752,272,894
11,345,514,515
21,712,599,435
3,480,957,004
13,169,031,421
23,618,950,103
7,233,229,898
24,514,545,936
45,331,549,538
^Volumes exclude batches with missing values
Average Value
(ppm)
108
121
98
110
117
86
109
119
92
for this parameter
Gasoline Volume
(gal)*
3,520,575,616
10,460,517,681
20,882,654,793
4,040,674,006
13,447,545,074
22,432,331,959
7,561,249,622
23,908,062,755
43,314,986,752
Table 2
The sulfur content of Summer Baseline Gasoline, as specified in the Clean Air Act, is 339 parts
per million (ppm), and the sulfur content of Winter baseline gasoline, as specified by EPA's regulations, is
338 ppm. Data analyses pertaining to CG sulfur are contained in the Appendix to this chapter. These
data include tabular and graphical descriptions of CG sulfur content by volume and grade which show:
» In 1997, the first year for which sulfur reporting data were analyzed, the median Summer sulfur
content was 264 ppm and the 1990 baseline gasoline fell between the 60th and 65th percentile.
* In 2005, the last year for which sulfur reporting data were analyzed the median Summer sulfur
content was 76 ppm and the 1990 baseline gasoline fell above the 98th percentile.
» In 1997 the median Winter sulfur content was 253 ppm and the 1990 baseline gasoline fell between
the 60th and 65th percentile.
* In 2005 the median Winter sulfur content was 64 ppm and the 1990 baseline gasoline fell above the
98th percentile.
» The sulfur content of premium grade CG has been lower than that of regular grade in each year and
season.
48
-------
Overview-All Gasoline
For the most part, this report has separately analyzed property trends by gasoline type and
season. The Tier 2 gasoline sulfur regulations do not distinguish between RFC and CG and make no
seasonal distinctions, but prescribe annual average standards. Consequently, it is appropriate to look at
sulfur trends in the entire gasoline pool and on an annual basis. Figure 3 shows quarterly averages for
RFC and CG combined (middle line). Figure 7 shows annual average sulfur levels for CG and RFC
combined. The only year-to-year increase in sulfur content occurred between 1998 and 1999, when both
RFC and CG sulfur content increased. There is no apparent regulatory explanation for these increases.
Annual Average Sulfur Content
(Estimated from Batch Data including CG Blendstocks)
350
I
300 -
250 -
200 -
150 -
100 -
50 -
-All Gasoline (ppm)
Figure 7
Aggregated sulfur content distributions by gasoline volume for 2005 are shown in Table 3, below:
2005 Sulfur Content (ppm) by Volume
(from
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volumefaall:
Batch Reports excluding CG blendstocks)
CG
0
8
16
21
25
30
36
44
52
60
70
80
93
107
125
147
172
201
233
264
478
90,897,009,258
RFG/RBOB
0
9
18
24
29
35
41
46
51
56
61
68
74
82
91
101
114
127
147
178
371
32,128,092,630
ALL
0
8
16
22
26
31
38
45
51
59
67
76
87
98
112
130
152
180
215
254
478
123,025,101,888
Table 3
49
-------
Vg'ipor
RVP relates to gasoline's volatility, a gasoline's ability to change from a liquid to a vapor. More
specifically, RVP is a measure of vapor pressure of gasoline at 100°F, using a prescribed test method.
Gasoline volatility characteristics are important because they affect both vehicle performance and
emissions. RVP affects both exhaust and evaporative emissions; however, RVP is significant from an
emissions standpoint primarily because it directly relates to evaporative emissions of volatile organic
compounds (VOCs). VOC emissions are involved in the atmospheric chemical reactions that form ground
level ozone, an air pollutant, in the presence of sunlight and oxides of nitrogen. VOC emissions, hence
RVP, are of particular significance from an air quality standpoint during the Summer season. RVP is one
of the Complex Model input parameters used to calculate exhaust VOC emissions, and it is the only
parameter affecting the model's evaporative VOC calculations. Since RVP is important to vehicle
performance on a year-round basis, ASTM specifies RVP standards for gasoline.
on
Several regulatory requirements indirectly or directly limit the RVP level in reformulated and
conventional supplied during late spring and summer when ozone is likely to be a problem. Because RVP
is an input parameter to the Complex Model, it is indirectly limited through Complex Model-based
emission standards. RVP is also directly limited through regulations which prescribe maximum levels for
RVP in gasoline (40 C.F.R. '80.27). The current volatility standards have been in effect since 1992,
predating the RFC and Anti-Dumping regulations.
The RVP limits prescribed in the volatility regulations are 9.0 psi or 7.8 psi, depending on the
location and specific month within the regulatory control period (May 1 to September 15 for refineries and
terminals, June 1 to September 15 for retail outlets). These volatility limits do not apply to gasoline sold
in Alaska, Hawaii and US territories, and gasoline containing between 9% and 10% ethanol by volume
may exceed the applicable standard by 1.0 psi.
Although the volatility regulations technically apply to RFC as well as CG, RFC regulations
effectively require RVP levels below those allowed under the volatility regulations. Under the RFC
regulations, refiners and importers must designate RFC produced or imported for use during the VOC
control period (the same period as the regulatory control period for volatility) as VOC-controlled, and all
other RFC as non-VOC-controlled. Different requirements, or "standards," apply to VOC-controlled RFC
depending on whether the gasoline is intended for use in VOC-Control Region 1 (VOC1), which generally
includes southern RFC areas, or VOC-Control Region 2 (VOC2), which includes northern RFC areas.
Refiners and importers must specify whether a VOC-controlled batch is intended for sale in VOC1 or
VOC2. The RVP in VOC-controlled RFC supplied prior to 1998 was directly controlled through specific RVP
standards. RFC for VOC1 was required to meet a 7.2 psi "per gallon" standard, or alternatively a 7.1 psi
"averaged" standard with a 7.4 psi per gallon maximum. RFC for VOC2 was required to meet an 8.1 psi
"per gallon" standard, or an 8.0 psi "averaged" standard with an 8.3 psi per gallon maximum.
Complex Model RFC requirements do not include RVP standards. However, the RVP in the
"Complex Model" VOC-controlled RFC supplied since 1998 is indirectly controlled through emissions
performance standards. RFC designated as VOC-controlled is required to meet certain VOC emission
reduction standards relative to a "statutory baseline" gasoline, representative of gasoline supplied in
1990, with different VOC standards for VOC Control Regions 1 and 2. 22 Although other gasoline
22 The Energy Policy Act of 2005 requires that EPA consolidate these separate standards.
50
-------
parameters also affect Complex Model VOC emission performance, RVP reduction from the statutory
baseline level of 8.7 psi is the primary means to achieve these VOC reduction standards; hence these
VOC standards indirectly, but tightly, limit RVP in RFC. The VOC emissions reduction standards became
more stringent in 2000 with Phase II RFC. The "averaged" VOC standards specify per gallon minimum
VOC reductions as well as average reductions. As a result, the RFC emissions performance standards not
only constrain average RVP levels below those permitted by the volatility regulations, but generally
constrain maximum RVP levels as well.
As discussed above, CG RVP levels are directly limited by EPA's volatility regulations. CG supplied
since 1998 is also subject to refiner/importer baseline-specific anti-dumping exhaust toxics and NOx
emission requirements calculated by the Complex Model. However, while RVP affects both exhaust toxics
and NOx calculations to some extent, CG emission requirements for exhaust toxics and NOx may not have
a significant limiting effect on RVP.
In addition to the federal fuel regulations affecting RVP, certain states and localities have unique
gasoline specifications for RVP and/or other parameters. Such gasoline's have been referred to as
"boutique" fuels, and include gasoline subject to 7.0 psi or 7.2 psi RVP standards.
EPA collects RVP data in order to determine compliance with the RFC and Anti-Dumping
regulations. RVP values are necessary for the Summer Complex Model emission calculation for VOC-
controlled RFC. As a result, refiners and importers have been required, since 1995, to submit RVP data
for VOC-controlled RFC to EPA's RFG/Anti-Dumping reporting system for each batch of gasoline refined or
imported. Additionally, since 1995, EPA has received RFC Survey data on the RVP of RFC sold at retail
outlets during the VOC control period, June 1 to September 15. Refiners and importers are not required
to report the RVP for RFC designated as non-VOC-controlled. This is because the Winter version of the
Complex Model, used to determine compliance for non-VOC-controlled RFC, does not require the RVP of
the gasoline being evaluated as an input. As a result, EPA has only limited RVP data for Winter RFC.
Winter RFC Survey data reported to EPA also do not include RVP values. Since EPA has only limited
Winter RFC RVP data, this report does not include analysis of these data.
EPA's data collection requirements for CG have differed from RFC both temporally and seasonally.
Whereas RVP values have been required for determining RFC compliance since 1995, they were not
required for CG compliance calculations until 1998, with the introduction of Complex Model-based exhaust
toxics and NOx emission standards. Although these standards are for exhaust emissions only, the RVP
value is needed for the Summer Complex Model exhaust calculation. Therefore, the RVP value is needed
to determine compliance for Summer CG. As discussed above, the actual RVP of the gasoline is not used
in Winter Complex Model emission calculations. However, the RVP value is also needed to determine
compliance for Winter CG. EPA's regulations specify that the correct model (Summer or Winter) for
evaluating the emissions performance of CG depends on its RVP value. Therefore, the RVP of each batch
of CG is needed to verify that the correct model was used for the batch. As a result, refiners and
importers have been required, since 1998, to submit RVP data to EPA's reporting system for each CG
batch, whether or not the batch was produced or imported for use during the regulatory control period
specified in EPA's volatility regulations. Thus, this report includes analysis of both Summer and Winter CG
RVP data.
51
-------
RVP Trends
Overview-Reformulated Gasoline
Figure 1 shows volume-weighted average Summer RFC RVP levels, by year. Both reporting
system estimates and survey-based estimates are shown. RVP levels were essentially unchanged with
the transition from Simple Model to Phase I Complex Model RFC standards between 1997 and 1998.
However, RVP decreased sharply between 1999 and 2000, with the transition from Phase I to Phase II
standards. This decrease in RVP was necessary in order to meet Phase II RFG's more stringent VOC
emission performance standards. Figure 2, using survey-based RVP estimates, shows that the Phase I to
Phase II RVP decrease was much greater for VOC Control Region 2 RFC than for VOC Control Region
RFC.
f
7.5D -
Highest RFG Area
Lowest RFG Area
Retail Avg.
Production Avg.
Average Summer RFG Reid Vapor Pressure
#
*
*«=
=4=
i ! i i 5
i i I 1 i '
1995
7.93
699
7.61
1996
798
7.02
7.64
1997
7.95
6.78
7.62
7.60
1998
798
6.99
7.65
7.60
1999
798
695
7.62
7.60
i
2000
6.90
6.43
6.77
6.78
Year
2001
688
6.54
6.77
6.79
2002
6.92
6.39
6.78
6.80
2003
6.93
6.50
6.82
6.83
2004
7.02
8.50
8.87
6.87
2006
704
6.57
6.82
6.81
* Highest RFG Area
Lowest RFG Area
-*-Retail Avg.
x Production Avg.
Figure 1
Average RVP of Summer RFG Sold at Retail Stations-By VOC Control Region
Q_
-*-VOC 1
H^VOC2
\
\
\
»
'
» .
*
^
\
\
A
"adjuster.
/erages in
VOC " RF
o and MihA
- *
Gsold
=*=
a
» * "~
1995
7.03
7.88
1996
7.06
7.88
1997
7.09
7.90
1998
7.04
7.92
1999
7.02
7.92
2000
6.75
6.78
2001
6.73
678
2002
6.77
6.79
2003
6.83
6.81
2004
6.85
6.89
2005
6.92
692
Year
Figure 2
52
-------
Figure 3 shows that the frequency distribution of RVP by volume for all Summer 1999 RFC is
bimodal; i.e. the "Combined" distribution has two distinct peaks. Separate frequency distributions for
VOC1 and VOC2 RFC are also shown in Figure 3. They are clearly different with very little overlap in
range, explaining the bimodal nature of the combined distribution. This RVP difference is a result of a
more stringent VOC reduction standard applicable to VOC1 RFC during Phase I of the RFC program.
Figure 4 depicts the same analysis for Summer 2000 RFC. It is clear that there is little difference
between the RVP distributions for VOC1 and VOC2. Although VOC1 RFC still has a numerically greater
VOC emission reduction standard than VOC2 RFC, the difference is much smaller than in Phase I of the
RFC program. Furthermore, the Complex Model evaluates VOC1 and VOC2 evaporative emissions
differently. Since much of the apparent difference in VOC standards for the two regions is actually due to
this model effect (See the "Emissions" Chapter), the RVP distributions for the two regions are very similar.
Summer 1999 RFC RVP Distribution-Combined and Split by VOC Region
(from Batch Reports)
E
S
Figure 3
The estimated average RVP levels in 2005 RFC are higher than in 2000 RFC. This increase is
small, slightly above 0.1 psi; however both survey and reporting estimates show an upward trend in
average RVP since 2000 (figure 1), and surveys show 2000 to 2005 increases in 21 of the 23 areas
surveyed in both years (see appendix to this chapter.) These RVP increases do not indicate
noncompliance; they are permissible as long as RFC continues to meet emissions performance standards.
The magnitude of the increase in RVP over this time period is small compared to the decrease in
RVP that occurred between Phase I and Phase II. Furthermore, there is no unquestionably clear cause-
effect relationship between regulatory requirements and this RVP change, as there was with the transition
from Phase I to Phase II RFC where RVP reduction was the only feasible way to meet the VOC
performance standard.
53
-------
Summer 2000 RFC RVP Distribution-Combined and Split By VOC Region
(from Batch Reports)
6.2 6.4
76 7.8
RVP (psi)
Figure 4
Nevertheless, there are indications that a cause-effect relationship may exist between the
increase in RVP and the introduction of the mandated Tier 2 sulfur reduction requirements 23 A reduction
in sulfur content, all else constant, would improve VOC emissions performance by reducing Complex
Model exhaust VOC emissions estimates. An increase in RVP, all else constant, would hurt VOC emissions
performance by increasing Complex Model evaporative and exhaust VOC emissions estimates. Thus, VOC
emissions performance could remain unchanged if a sulfur decrease occurred concurrently with an RVP
increase. The existence of some cause-effect relationship between sulfur decreases and RVP increases in
RFC seems likely, particularly because VOC compliance margins have been small since 2000, (although
there has been substantial NOx and toxics over-compliance). This likely cause-effect relationship may be
due, to a large extent, to the way that refiners control gasoline RVP. It is often economically
advantageous for refiners to blend as much of the lighter hydrocarbons, particularly butane, as possible
into gasoline. However, butane raises RVP and, as a result, the need to comply with the VOC standard
limits the amount of butane that may be blended into summer RFC., Consequently, if sulfur reductions
provided additional VOC compliance margin, refiners may be motivated to offset it by raising RVP. The
data indicate an increase in RVP between 2000 and 2005, which is concurrent with a decrease in sulfur
content resulting from the Tier 2 sulfur requirements. As stated elsewhere in this report, these data do
not reflect the final Tier 2 sulfur requirement, and the interaction between the RFC requirements and the
Tier 2 sulfur reductions should be better understood as additional years of data are analyzed.
RFGRVPbyPADD
Table 1 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These
averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information):
23
This is also discussed in the RFG Trends Chapter and in the Emissions Chapter.
54
-------
2004 & 2005 Reporting Average by PADD-RFG RVP (Refiner Batches Only)
2004 2005
Season PADD Average Value Gasoline Volume Average Value Gasoline Volume
(psi) (gal)* (psi) (gal)*
Summer I 6.83 4,792,114,891 6.88 4,431,080,230
II 6.98 1,740,499,436 6.95 1,844,913,833
III 6.88 5,890,920,167 6.93 5,690,766,967
^Volumes exclude batches with missing values for this parameter
Table 1
Figure 5 shows estimates of average levels by PADD in retail RFC. These averages are volume-weighted
averages of the seasonal averages for each area, using gasoline volume estimates supplied in the survey
plans. In 2005, RFC surveys were conducted in 18 PADD I areas, five PADD II areas and two PADD III
areas. Prior to 2000, VOC control region was the primary factor contributing to geographic differences in
RVP. PADD I is a mix of control region 1 and 2 areas. PADD 2 was entirely control region 2 areas until
St. Louis (included in the 1999 surveys) opted into the RFC program. Both PADD 3 survey areas
(Houston and Dallas, TX) are in control region 1. As expected, the PADD to PADD RVP differences
diminished in 2000. The post-2000 upward RVP trend is apparent in each PADD.
Average RVP of Summer RFC Sold at Retail Stations-By PADD
7.01 7.05 7.07 7.02 0.00 0.74 6.74 6.B4 6.02 6.04
Figure 5
Additional Analyses and Observations-RFC
Data analyses pertaining to RFC RVP are contained in the Appendix to this chapter. Several trends or
patterns are highlighted below:
Twenty-one of twenty-three areas surveyed in both years had higher average RVP levels in 2005 than
in 2000.
» An upward shift in RVP since 2002 can also be seen in the distribution trend graph in the appendix.
55
-------
Overview-Conventional Gasoline
Volume-weighted average Summer and Winter CG RVP levels are shown in figure 6. These RVP
values have varied little from year to year since 1998. These averages estimate RVP from
refiner/importer sampling, and could differ from the average RVP of gasoline sold at retail outlets for
various reasons, including downstream blending operations. For example, the volume-weighted CG
averages reported here include batches which are blendstocks such as ethanol intended for blending
downstream of refineries, as well as certain a 11-hydrocarbon blendstocks which do not contain oxygen.
The implicit assumption in a volume-weighted property average is that the property blends linearly with
volume, and this assumption is inaccurate for RVP when ethanol is blended into an a 11-hydrocarbon
blendstock. Although inclusion of blendstocks in the RVP averages could introduce some error, exclusion
of blendstocks from the average calculations does not necessarily provide a better estimate of the
average RVP of all conventional gasoline. (This is less of an issue in analysis of RFC data since reporting
requirements differ from CG requirements, and since retail RFC data are available as well.)
RVP (psi)
11 00
10.00
9.00-
8.00-
- Summer RVP
--Winter RVP
Average RVP of Conventional Gasoline
(from Batch Reports)
- - -
1998
831
12.13
1999
8.29
12.04
2000
826
12.01
2001
825
11.91
2002
825
11.99
2003
8.29
12.11
2004
8.29
12.17
2005
829
12. OB
Reporting Year
Figure 6
Figure 7 compares the RVP distributions by volume of Summer 1998 and Summer 2005 CG,
showing that the distribution as well as the average has changed little. The distributions are multi-modal
with peaks just to the left of the vertical lines marking 7.8 and 9.0 psi, the two federal volatility
standards. The distributions drop sharply at 7.8 and 9.0 psi suggesting that CG is often blended to meet
one of these two standards with only a small compliance margin. Blendstocks have been excluded from
the CG distribution data.
There is an additional smaller and more rounded peak at about 6.8 RVP, probably due to certain
boutique fuels. These boutique fuels are reported asCG, and EPA's reporting system does not distinguish
them from other conventional gasoline. Consequently, "boutique fuel" volume share cannot be explicitly
determined from EPA's data. (This "boutique fuel" portion of the distribution does not represent all
gasoline with unique specifications. Notably, California gasoline has unique fuel specifications affecting
RVP and other parameters, but gasoline intended for sale in California is exempt from EPA's reporting
requirements.) Readers who wish to determine what portion of gasoline volume falls within a specific
RVP range should refer to the percentile curves contained in the appendix to this chapter.
56
-------
The distributions show a small volume of gasoline with RVP greater than 9.0 psi. This does not
necessarily represent noncompliant gasoline, but could include gasoline intended for Alaska, Hawaii and
US territories not subject to EPA volatility regulations. The distributions also contain some ethanol-
oxygenated gasoline eligible for a 1.0 psi RVP allowance. (Ethanol is typically blended downstream of
refineries, so ethanol oxygenated gasoline is under-represented in these CG distributions.) Additionally,
refiners and importers did not always identify their CG batches as volatility-controlled or not. (Current
instructions require that they do so.) In the absence of a designation, a batch produced or imported
between April and September was assumed to be volatility-controlled, and some non-volatility-controlled
CG batches may have been included in the Summer data.
Comparison of Summer 1998 and Summer 2005 CG RVP Distributions
(from Batch Reports excluding Blendstocks)
Figure 7
Figure 8 compares the RVP distributions by volume of Winter 1998 and Winter 2005 CG, again
showing little difference between the two years. These distributions have a number of peaks. These
multi-modal distributions occur because gasoline, even outside of the volatility control season, is produced
to meet certain vapor pressure specifications. These specifications, described in ASTM Standard D 4814,
define six vapor pressure classes, specifying a maximum vapor pressure for each class. The vapor
pressure class limits are shown as vertical lines and labeled across the top axis. The ASTM standard also
defines vapor pressure class requirements by location and month. Outside of the EPA volatility-control
season these requirements are intended primarily to ensure adequate vehicle performance (e.g. a
gasoline used in Maine in January would require a higher RVP than a gasoline used in Florida in January
to ensure adequate cold-start performance.)
For several of the vapor pressure classes, the Winter CG distributions show a peak just to the left
of the class limit, with a sharp drop at the limit. As with the Summer CG, this suggests that refiners and
importers often supply gasoline which meets a vapor pressure class maximum with only a small
compliance margin. The winter distributions, however, also show peaks at points other than immediately
before a volatility class limit. Most notably, both distributions show a peak at about 12.5 psi, midway
between two class limits.
57
-------
Comparison of Winter 1998 and Winter 2005 CG RVP Distributions
(from Batch Reports Excluding Blendstocks)
11.0 13.0
RVP (psi)
Figure 8
In summary, EPA's reporting system data indicate that both Summer and Winter CG RVP levels
show little change since 1998. However, reporting system RVP data for CG may not completely and
accurately reflect all downstream RVP changes. EPA's volatility regulations and ASTM standards appear
to be the major factors controlling CG RVP. While Tier 2 sulfur reductions could theoretically affect
Summer CG RVP levels through CG's Complex Model-based exhaust toxics and NOx standards, the
significance of this effect would depend on the extent to which refiner/importer individual compliance
baselines for these pollutants, rather than volatility standards, limit CG RVP.
CG RVP by PADD
Table 2 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These
averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information). As noted, the primary geographic factors limiting summer CG RVP
are emission-related EPA volatility and state "boutique fuel" regulations. Winter gasoline RVP is related
to location because of ambient temperature-associated drivability concerns. Thus, it is not surprising to
find lower average RVP levels in PADD III production than in PADDs I and II.
2004 & 2005 Reporting Average by PADD-CG RVP (Refiner Batches Only)
Season
Summer
Winter
Annual
2004
PADD Average Value
I
II
III
I
II
III
I
II
III
(psi)
8.40
8.46
8.19
12.66
13.31
11.77
10.26
10.70
9.94
Gasoline Volume
(gal)*
3,744,548,926
11,453,187,269
21,722,168,035
2,891,701,318
9,805,826,594
20,865,873,542
6,636,250,244
21,259,013,863
42,588,041,577
2005
Average Value
(psi)
8.31
8.46
8.22
12.31
13.25
11.73
10.30
10.61
9.88
Gasoline Volume
(gal)*
3,547,001,722
10,554,125,517
20,889,312,566
3,504,605,534
8,565,282,585
18,781,652,701
7,051,607,256
19,119,408,102
39,670,965,267
^Volumes exclude batches with missing values for this parameter
Table 2
58
-------
The RVP of Summer Baseline Gasoline, as specified in the Clean Air Act, is 8.7 psi, and the RVP of Winter
baseline gasoline, as specified by EPA's regulations, is 11.5 psi.24 Data analyses pertaining to CG RVP
are contained in the Appendix to this chapter. These data include tabular and graphical descriptions of
CG RVP by volume which show:
» In 1998, the first year for which RVP reporting data were analyzed, the median Summer RVP was
8.52 psi and the 1990 baseline gasoline fell between the 65th and 70th percentile.
« In 2005, the last year for which RVP reporting data were analyzed, the median Summer RVP was
8.55 psi and the 1990 baseline gasoline fell between the 65th and 70th percentile.
» In 1998 the median Winter RVP was 12.39 psi and the 1990 baseline gasoline fell between the 30th
and 35th percentile.
« In 2005 the median Winter RVP was 12.21 psi and the 1990 baseline gasoline fell between the 35th
and 40th percentile.
24 The 11.5 psi winter baseline value is specified in EPA's regulations even though a default value of 8.7 psi is
used in Complex Model evaluations of winter gasoline in order to "zero out" the effect of RVP on Winter exhaust
emissions.
59
-------
Oxygenates are compounds used in gasoline blending that contain carbon, hydrogen and oxygen.
There are two main classes of oxygenates; ethers and alcohols. Methyl tertiary-butyl ether (MTBE) was
the predominant ether used in gasoline blending. Other ethers, primarily tertiary-Amyl methyl ether
(TAME), were sometimes used, often in combination with MTBE. In the United States, ethanol is the
only alcohol currently used in any significant quantity in gasoline blending, and ethers are no longer used
in significant quantities.
Oxygenates have been blended into gasoline in order to comply with regulatory requirements
intended to reduce air pollutant emissions from gasoline vehicles and engines. The Clean Air Act
required that reformulated gasoline (RFC) contain 2.0 weight percent oxygen. RFC was intended to
reduce emissions of toxics and ozone-forming pollutants. Oxygenates provided direct toxics and exhaust
VOC emission reduction benefits through the effect of oxygen on combustion, as well as indirect benefits
through dilution of, or partial substitution for gasoline blendstocks containing sulfur, aromatics or benzene
(constituents which can adversely impact emissions). Oxygen also directly reduces carbon monoxide
(CO) emissions through its effect on combustion. The Clean Air Act required that states adopt
oxygenated fuel programs for carbon monoxide non-attainment areas during the portion of the year
(typically winter) in which the area is prone to high ambient CO concentrations.
Oxygenates are also used in gasoline blending for reasons unrelated to emission reduction. Both
ethers and alcohols have a high blending octane value, making them at times economically attractive
blending component choices. Ethanol is a renewable fuel which, in the United States, is primarily
produced from corn. Federal tax incentives
have encouraged the use of ethanol in gasoline. These incentives included an excise tax exemption
available for gasoline alcohol blends. The amount of the exemption depended on the ethanol content of
the blend, with the maximum exemption available for 10% by volume ethanol blends and pro-rated
amounts for 7.7% and 5.7% blends (based on 190 proof or 95% pure ethanol). Additionally, blenders
receive a tax credit for blending ethanol into gasoline. The American Jobs Creation Act of 2004, signed
on October 22, 2004 changed the way the excise tax exemption operates. The amount of the exemption
is no longer based on these three blend levels.
The Energy Policy Act of 2005, repealed the RFC oxygen content requirement, effective
immediately in California and 270 days after enactment elsewhere.25 It added a gasoline renewable fuel
content requirement of 4 billion gallons in 2006, incrementally increasing to 7.5 billion gallons in 2012.26
These recent legislative changes will clearly impact the composition of future RFC and CG.
As a prelude to a discussion of this topic, it is useful to review the various units used to express
gasoline oxygen and oxygenate content, and the relationship between these units. Gasoline oxygen
content is expressed as a weight percentage of the gasoline-oxygenate blend. Gasoline oxygenate
content may expressed as either a weight or volume percentage of the gasoline-oxygenate blend. In
25 Removal of the oxygen requirement became effective with regulatory changes on April 24, 2006 for
California gasoline, and May 5, 2006 for gasoline nationwide. See 71 F.R. 8965 and 71 F.R. 26419.
25 EPA finalized a regulation to implement the Renewable Fuel Standard Program on May 1, 2007. See 72 FR
23900.
60
-------
order to provide gasoline meeting a specific oxygen content requirement, oxygenates such as MTBE or
ethanol are blended into a hydrocarbon blendstock. As noted, oxygenates contain hydrogen and carbon
as well as oxygen, and different oxygenates contain different amounts of oxygen. For example, a pound
of pure MTBE contains about 0.18 pounds of oxygen, while a pound of pure ethanol contains about 0.35
pounds of oxygen. Therefore, it would be necessary to blend about 11 pounds of MTBE with 89 pounds
of non-oxygenated blendstock to produce 100 pounds of 2.0 weight percent oxygenated gasoline, which
would contain 11% MTBE, by weight. Alternatively, it would be necessary to blend about 5.7 pounds of
pure ethanol with 94.3 pounds of non-oxygenated blendstock to produce 100 pounds of 2.0 weight
percent oxygenated gasoline, which would contain 5.7% ethanol, by weight.
Since gasoline and oxygenates are liquids, oxygenate concentrations are more commonly
expressed as volume percentages rather than weight percentages. There are no universal weight percent
to volume percent conversion factors because each gasoline blend is a different mixture of hydrocarbons.
In order to convert exactly between a weight and volume oxygenate concentration it is necessary to
know the density of the gasoline-oxygenate blend, which varies from blend to blend. (There are
procedures to measure the density of a gasoline-oxygenate blend and these measurements are included
in the reporting system data submitted to EPA.) However, even if the density of the specific gasoline-
oxygenate blend is not known, a reasonably accurate weight to volume or volume to weight conversion
can be made because the densities of these blends vary over a fairly narrow range.27 A 5.7 weight
percent pure (anhydrous and 200 proof) ethanol blended gasoline would typically contain slightly less
than 5.7 volume percent of pure ethanol. However, a 5.7 volume percent ethanol concentration is often
nominally associated with a 2.0 weight percent oxygen concentration, 7.7 volume percent ethanol with
2.7 weight percent oxygen and 10 volume percent ethanol with 3.5 weight percent oxygen. The 2.0%
and 2.7% oxygen weights and the 10 volume percent ethanol content have significance with respect to
Clean Air Act requirements and EPA regulations.
As noted, (prior to the effective date of the repeal of this provision by the 2005 energy
legislation) the Clean Air Act required that RFC contain 2.0 weight percent oxygen. EPA's regulations
allowed refiners, importers and oxygenate blenders to meet this requirement through compliance with a
2.0 weight percent "per gallon" standard, or a 2.1 weight percent "averaged standard" and a 1.5 weight
percent per gallon minimum. The per gallon minimum applicable to certain suppliers was adjusted as a
consequence of RFC Survey failures for oxygen content, and for 2005 remained at 1.6 weight percent for
certain suppliers and covered areas. These RFC oxygen requirements applied throughout the year. The
Clean Air Act also required an oxygen credit program; i.e. parties using more than the required amount of
oxygen in RFC generate credits which may be transferred to other parties for use in meeting the oxygen
standard.
The Clean Air Act's oxygenated fuel requirements (Section 211(m)) are intended to address CO
non-attainment. States may also have "maintenance" programs. These oxygenated fuel programs are
seasonal and state-specific. The Clean Air Act specified that the oxygen content for the non-attainment
areas requiring such programs be at least 2.7 weight percent. While some areas currently have winter
oxygenated fuel programs a number of areas, including several areas in the RFC program, at one time
had programs but no longer implement them.
27 EPA received RFC Survey oxygenate concentrations as weight percentages and Reporting System
concentrations as volume percentages. For this report, EPA made approximate weight to volume or volume to weight
conversions assuming a specific gravity of 0.745 for Summer RFC and 0.730 for Winter. The specific gravities of
MTBE, ethanol and TAME published in ASTM D4814 are 0.7460, 0.7939and 0.7758, respectively. To convert a weight
percent oxygenate to a volume percent oxygenate multiply the weight percent oxygenate by the gasoline blend
specific gravity and divide by the oxygenate specific gravity.
61
-------
In addition to the requirements for oxygen use, there are several restrictions on the maximum
oxygen content and maximum amount of certain specific oxygenates that can be used. Both the Simple
Model and Complex Model have valid range limits of 4.0 weight percent for oxygen. However, these
range limits had little effect on the maximum oxygen content of gasoline since other requirements, in
place long before the RFC and Anti-Dumping programs, were as restrictive or more restrictive.
The Clean Air Act C211(f)(l)) prohibits fuel and fuel additive manufacturers from first introducing
into commerce, or increasing the concentration in use of any fuel or fuel additive for general use in light
duty motor vehicles which is not substantially similar to that utilized in the certification of motor vehicles
or engines under section 206 of the Act. A manufacturer may apply for a waiver of this provision for a
fuel or fuel additive which is not substantially similar. The Act, however, does not define "substantially
similar", consequently EPA defined "substantially similar" as it applies to unleaded gasoline in an
interpretive rule, last revised in 1991 (56 FR 5352).28
EPA's "substantially similar" definition allows certain alcohols (other than methanol) and ethers
including ethanol and MTBE, provided that the oxygen content does not exceed 2.7 weight percent. This
oxygen weight allows approximately 15% MTBE by volume, but the exact volume of MTBE allowed under
this definition is blend-specific. However, a waiver has also been granted which, under some conditions,
allows 15% MTBE by volume even if the oxygen content weight limit is exceeded. This oxygen weight
percent limit would only allow ethanol blending up to about 7.7 volume percent. However, the
"gasohol" waiver allows use of up to 10% by volume pure ethanol. This waiver is particularly significant
since ethanol is often blended at about 10 volume percent in order to take full advantage of the tax
incentives. This waiver allows approximately 3.7 weight percent oxygen, but again, this is blend-specific.
The definition and several waivers also allow use of methanol as an oxygenate, but the conditions for
methanol use are much more restrictive. Unless it is used with other oxygenates, methanol can only be
used at 0.3% by volume (about 0.16 weight percent oxygen). Additionally, health effects testing
requirements apply to methanol blends containing oxygen at 1.5 weight percent or greater. (Trace
quantities of methanol were often found in MTBE-oxygenated gasoline since methanol was used to
produce MTBE.) A document summarizing 211(f) waiver requests and EPA decisions is available on EPA's
website (EPA, 1995).
Oxygen and oxygenate content data have always been necessary to determine compliance with
RFC standards. The weight percent oxygen content requirement applies to RFC for all years considered
in this report. Additionally, both the Simple and Complex models require the amount of oxygen coming
from specific oxygenates as inputs in order to calculate toxics emissions. Since 1995, RFC refiners and
importers have been required to report weight percent oxygen as well as MTBE, ethanol, ETBE, TAME, t-
butanol and methanol content. (These oxygenates are reported as volume percentages and volume to
weight conversions can be made as needed since the density of each batch is also reported in units of
API gravity.)
One complication with this reporting is that ethanol blending typically does not occur at refineries,
but downstream at terminals. (This is necessary because gasoline is often shipped via pipeline and there
are problems associated with pipeline shipment of ethanol-blended gasoline due to ethanol's affinity for
water, which is present to some extent in pipelines.) Although this is less common, MTBE and other
ethers may also be blended downstream of the refinery. Thus, refiners generally produce and ship
reformulated blendstock for oxygenate blending (RBOB) rather than finished ethanol-oxygenated RFC.
28 n,
Substantially similar" information is available at http://www.epa.gov/otaq/additive.htm
62
-------
Refiners must also designate the type(s) and amount(s) of oxygenates that may be added to the RBOB.
For compliance purposes, a refiner or importer must estimate the properties of the RFC that will be
produced from an RBOB batch by blending the specified type and amount of oxygenate into a sample of
the RBOB and analyzing the sample. The property and volume data which are reported to EPA for RBOB
batches are based on these "hand blends" rather than on analysis of the RFC produced by the actual
blending operations. There are a number of regulatory requirements associated with RBOB and
downstream blending of oxygenates. For example, RFC oxygenate blenders are also required to report
the type and amount of oxygenate added. These regulations are intended to ensure that the RFC
produced from RBOB meets all applicable standards. However, even if these requirements are precisely
and rigorously adhered to, "hand blend" properties may differ from final RFC properties. In certain
instances, the amount of oxygen added downstream may be greater than the amount of oxygen assumed
for reporting purposes. In fact, unless a refiner or importer of RBOB meets certain contractual and
quality assurance requirements addressing downstream blending, EPA's regulations require that such
assumptions be made.
In addition to the reporting system data, EPA receives RFC Survey data which includes total
oxygen content in weight percent as well as concentrations of specific oxygenates, including several not
included in the reporting data. (These oxygenates have been reported to EPA as weight percentages.
Prior to 2004, EPA did not receive density information for each sample.) Until the repeal of the RFC
oxygen requirement, survey sampling for oxygen and oxygenates included RFC sold in federal RFC areas
in California. This RFC is not included in EPA's reporting system data. California RFC surveys sampled
for oxygen and oxygenates only and are no longer conducted. Surveys conducted outside of California
continue to sample for oxygen and oxygenates because they are input parameters for the Complex
Model.
The Anti-Dumping standards applicable to CG do not include an oxygen content standard.
However, Complex Model CG is subject to NOx and exhaust toxics emissions standards, and oxygen and
oxygenate data are necessary to calculate these emissions. Thus, since 1998, refiners and importers
have been required to report CG oxygen and oxygenate data. The CG reporting procedures and
requirements differ from those of RFC. CG blendstocks for oxygenate blending as well as some
oxygenate batches are included. However oxygenates added downstream of the refinery or import
facility may be included in compliance calculations only if the refiner can show that they were added to
the gasoline or blendstock, and thus may not be reported. Additionally, refiners and importers do not
create "hand blends" of blendstocks and oxygenates to determine properties, but report the pre-blend
properties of the blendstocks.
In summary, downstream oxygenate blending, particularly ethanol blending, contributes to
differences between gasoline leaving the refinery or import facility and gasoline sold at retail. This
affects the analysis and interpretation of the data considered in this report, not only for oxygen and
oxygenates, but for other properties as well since they will change with the addition of an oxygenate.
The effect on RFC data is expected to be small because of the requirement to create "hand
blends", and because of other regulatory requirements. (EPA's analyses of RFC batch reporting data in
this chapter and in other chapters include data for both RFC batches and RBOB batches.) Additionally,
EPA has a second source of data from the RFC surveys, which directly measures retail properties.
Consequently, even though ethanol is widely used in RFC, EPA believes its data provide good estimates of
the oxygenate content and the other properties of the RFC that is used in vehicles. Additionally, EPA
believes that it can make credible, although approximate, estimates of various measures of oxygenate
"market share" (e.g. as in the RFC Trends chapter) and of the volume of oxygenates used in RFC.
63
-------
It is possible that a significant amount of the ethanol added to CG is not included in the Anti-
Dumping reporting data submitted to EPA. Although EPA has analyzed these data to estimate oxygen
and oxygenate content, the results probably underestimate of the oxygenate content of CG sold at retail,
and may not accurately reflect trends in CG oxygenate use. However, even if CG oxygenate use is
significantly under-reported, the volume percentage of oxygenates in CG is believed to be small (certainly
in comparison to RFC). Therefore, EPA does not believe that downstream oxygenate blending has
caused estimates of other CG parameters from reporting system data to be substantially unrepresentative
of CG sold at retail.
The General Methodology chapter explained that, for each parameter analysis in this report, a
rule was established to determine if the data was missing from a batch record. It is worth noting that
the decisions pertaining to oxygen and oxygenate data are somewhat unique. Since the RFC considered
in this report was subject to an oxygen content requirement, and the oxygen could come from any of
several reported oxygenates, a zero or blank value for any oxygenate was treated as a zero if there was a
non-zero value for one or more of the other oxygenates; otherwise it was treated as missing data. For
example, a blank ethanol value would be averaged as a zero for batches reporting MTBE, ETBE, TAME or
T-Butanol content, but excluded from an average if these other oxygenates were also reported as blanks
or zero. Since CG may or may not contain oxygen, a reported oxygen or oxygenate value of zero was
treated as zero, and a blank value was treated as a zero if at least one other non-zero property value was
reported (e.g. a blank MTBE value would be assumed zero if sulfur was greater than zero); otherwise it
was treated as missing data.
"- ; ;.'
Since this topic has also been discussed in the RFC Trends chapter, reference will be made to
material contained therein. Readers, particularly those who are unfamiliar with the issues pertaining to
oxygenate use, may wish to read the relevant portion of that chapter before continuing. (Additionally,
some of the oxygen and oxygenate-related graphical and tabular data analysis in the RFC Trends chapter
is not repeated here.)
Although the RFC oxygen content requirement did not change between 1995 and 2004, average
RFC oxygen content increased during that period. During this period the average ethanol content of all
(i.e. ethanol and ether oxygenated) RFC increased while the average MTBE and TAME content of all RFC
decreased.
Figures 1 and 2 show estimates of average RFC oxygen content in Summer and Winter RFC
Both reporting system estimates, by reporting year, and survey-based retail estimates, by survey year,
are shown. The survey estimates exclude California data and, as for other parameters, California
gasoline is exempted from oxygen and oxygenate reporting requirements. The averages shown are
volume-weighted estimates. (See also Figures 25 through 32 in the RFC Trends chapter. These graphs
provide additional information and also include survey-based estimates which include federal RFC areas in
California.) Figures 1 and 2 show that oxygen content estimates from reporting and survey data are
close, but that survey-based estimates of oxygen content have been consistently higher by small
amounts. The direction of these differences suggests that the amount of oxygen added downstream is,
on average, slightly greater than that assumed for refiner and importer compliance calculations. As
noted, this is allowable and consistent with EPA's regulations. EPA's data screening to exclude batches
with missing oxygen or oxygenate data also contributed to lower average oxygen and ethanol content
estimates from the reporting system. The majority of batches with missing data were RBOB; thus likely
64
-------
to be ethanol-oxygenated and also likely to be oxygenated at about 3.5 weight percent oxygen. (RBOB
refiners and importers are not required to report their "hand blend" oxygen and oxygenate content.
Although the majority of RBOB batch reports contained this information, a significant number in each year
did not.) EPA believes that the reporting data analysis probably underestimates average oxygen (and
ethanol) content. However, other factors could have caused or contributed to the differences between
reporting and survey-based estimates, including statistical sampling error or errors in the estimates of
gasoline volumes for individual survey areas.
Average Summer RFG Oxygen Content (Excluding CA)
Oxygen(wt%)
2 50 -
'
Highest RFG Area
Lowest RFG Area
Retail Avg.
Production Avg.
»
1995
2.61
1 98
2.19
>
* "
1996
3.42
1.85
218
*
1997
3.44
1.89
2.24
2.13
i
1993
346
1.95
2.26
213
!
1999
3.54
1.96
226
2.11
i
2009
3 51
2 01
2.31
2 24
Year
2001
3.52
1.96
230
2.21
»
£=^
2002
3.50
1.93
2 32
2 25
£
^
2903
3.57
2 01
2.41
2.30
I
t=!
i
2904
3.59
2 06
285
2.56
t
:
I
2005
3.66
1.82
2.60
2.48
+ Highest RFG Area
Lowest RFG Area
*- Retail Avg.
** Production Avg.
Figure 1
Average Winter RFG Oxygen Content (Excluding CA)
(wt %)
Highest RFG Area
Lowest RFG Area
Retail Avg.
Production Avg.
1998 1999 2000
2002 2003 2004 2005
Figure 2
65
-------
Both graphs show higher oxygen content in 2004 than in any of the preceding years. Average
oxygen content increased due to the increased use of ethanol which, outside of California, was often
blended at 10 volume % to take full advantage of tax incentives, and to replace much of the lost volume
where ethanol replaced ethers as an oxygenate. Figures 3 through 8 shows that average ethanol content
increased substantially from 2003 to 2004, with corresponding decreases in ether content. These changes
were primarily due to MTBE bans in New York and Connecticut which took effect in 2004. Transitions from
ether-oxygenated to ethanol-oxygenated RFG occurred in other areas, as well, at various times within the
period covered by these trend charts. (Figures 21 through 24 in the RFG Trends Chapter identify areas
where large oxygenate use changes occurred between 2000 and 2004). Ethanol content estimates from
surveys are higher than reporting data estimates, and MTBE estimates lower, again consistent with
downstream ethanol blending in excess of the amount assumed for compliance calculations and the
probable bias resulting from screening for missing oxygenate data in batch reports. (The previously stated
caveats relating to comparison of survey and reporting estimates apply. Additionally, reporting data
volume concentrations were converted to weight concentrations to be directly comparable with survey
estimates, and the conversion calculations were approximations using average density values.)
| Ethanol (weight %) |
Average Summer RFG Ethanol Content (Excluding CA)
10 -
2 -
» Highest RFG Area
Lowest RFG Ares
-*- Retail Avg Wt%
-*? Production Avg Approx Wt%
Production Avg. Vol%
Retail Avg. Approx Vol%
^
!
«-
,, '
1995
6 56
0.00
0.98
0.92
k J
1996
9 58
0.00
1.42
1.33
t
1997
9.61
0.00
1.36
099
093
1.28
, * !
1998
9.76
0.00
1.46
1.07
1 00
1.37
1999
10.21
0.00
1.60
1.14
1 07
1.50
* 3
2000
1012
0.00
1.60
1.23
1.15
1.50
E=2
2001
10.14
0.00
1.57
1 17
1.10
1.47
P_j
2002
10 09
0.00
1.81
1.39
1.30
1.70
^
2003
10 29
0.00
1.96
1.60
1.50
1.84
t=l
<
2004 2005
1034 10.53
000 0.00
3.72 3.81
325 310
305 291
3.49 358
Year
Figure 3
Ethanol (Weight %)
Average Winter RFG Ethanol Content (Excluding CA)
8 -
6 -
2 -
Highest RFG Area
Lowest RFG Area
-*- Retail Avg. Wt%
K Production Avg Approx Wt%
Production Avg. Vol%
Retail Avg Approx Vol%
t r r * ?
*
.
i
*
k ^
1995
7.69
000
1 33
1.22
1996
9.75
0.00
1.51
1 39
1997
10.17
0.00
1 83
1.20
1.10
1.68
1998
10.56
000
1 68
1.35
1.24
1 54
,
1999
10.61
0.00
1.85
1.28
1.18
1.70
2000
10.66
0.00
1 96
1.33
1.22
1.81
2001
10.50
0.00
1.81
1 28
1.18
1.66
t -
2002
10.48
000
1 95
1 42
1 31
1.79
^
. -*
<£
2003
10.71
000
252
1 83
1 68
2.31
'
f
* 'i
2004
11.22
0.00
397
2.91
268
3.65
:
2005
10.64
0.00
4.20
301
2.77
3.86
Year
Figure 4
66
-------
Average Summer RFG MTBE Content (Excluding CA)
UJ
I
12
8 -
A
» Highest RFG Area
L'i'V'.jV'ii RFG Area
-*- Retail Avg Wt%
x Production Avg Approx Wt%
Production Avg Vol%
Retail Avg Approx Vol%
* *
1
«
^3
m
i
1995
11 84
1 49
9.07
9.06
.
r
1996
11.28
0.38
8.48
8.47
i
i
1997
11.49
045
9.01
8.86
8.85
9.00
1998
11.70
0.34
8.92
8.86
8.85
8.91
.
1999
11.44
0.00
8.63
8.61
8.60
8.62
,
2000
11 82
0.00
8.82
9.05
9.04
8.80
2001
12.31
0.00
8.82
9.03
9.02
8.81
2002
12.27
0.00
8.59
8.96
8.95
8.58
>
S
S
2003
12.34
0.00
8.69
892
8.91
8.68
fcs
r
!
2004
11.85
0.00
7.09
741
740
7.08
.
2005
11 04
0.00
6.59
7.22
7.21
6.58
Year
Figure 5
| MTBE (weight %) |
Average Winter RFG MTBE Content (Excluding CA)
12 -
10 -
8 -
» Highest RFG Area
Lowest RFG Area
-^-Retail Avg.Wt%
-x Production Avg Approx Wt%
Production Avg Vol%
Retail Avg Approx Vol%
i
i
<
H*^^
1995
14.06
2.22
9.79
9.58
1996
12.95
0.23
869
8.50
,
, *
,
1997
13.32
0.17
891
8.77
8.58
8.72
I
,
.
j
L
1998 1999 2000
13.93 12.48 10.96
0.01 0.00 0.00
946 8.47 823
8.99 8.66 8.39
8.80 8.47 8.21
9 25 8.29 8.06
'
2001
11.21
0.00
817
8.45
8.27
800
'
2002
10.98
0.01
8.05
8.19
8.01
787
.
^
2003
10.88
0.00
7.72
7.93
7.76
7.55
.
£^1
' 1
2004
11.79
0.00
6.39
7.19
7.03
6.26
:
,
2005
1080
0.00
6.10
6.92
6.77
5.97
Year
Figure 6
67
-------
| TAME (Weight %)
Average Summer RFG TAME Content (Excluding CA)
300 -
200 -
1 00 -
Highest RFG Area
Lowest RFG Area
-*- Retail Avg Wt%
-* Production Avg Approx Wt%
Production Avg. Vol%
Retail Avg Approx Vol%
t
' V #
*
t
t
T
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
184 1.73 1.54 173 3.21 2.86 2.34 2.75 2.87 113
0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
067 0.79 0.70 0.76 0.83 0.93 0.92 0.75 0.90 0.38
0.81 0.81 090 0.99 0.94 0.79 0.86 044
0.78 0.78 0.86 0.95 0.90 0.76 0.83 0.42
065 0.75 067 0.73 080 090 0.88 072 0.86 0.36
==*
2005
1.27
0.00
044
0.46
0.44
0.42
Year
Figure 7
Average Winter RFG TAME Content (Excluding CA)
TAME (Weight %)
1.00 -
0.50 -
» Highest RFG Area
Lowest RFG Area
-*- Retail Avg Wt%
-x Production Avg Approx Wt%
Production Avg Vol%
Retail Avg Approx Vol%
*
*
t
»
«
,r -+ * =r=--*-
1995 1996 1997 1998 1999
1.96 1.66 1.56 1.64 2.63
0.01 0.01 0.00 000 0.00
069 079 0.74 0.70 086
0.74 074 0.86
0.70 0.70 0.81
0.65 0.74 0.70 0.66 0.81
-* 4^,
ii
2000
3.21
0.00
079
0.79
0.74
0.74
2001
3.12
000
078
0.75
0.71
0.73
* ^
2002
2.03
000
057
0.63
0.59
0.54
2003
0.98
000
039
0.40
0.38
0.37
2004
1.25
000
0.37
0.35
0.33
0.35
2005
1.47
000
030
0.30
0.28
0.28
Year
Figure 8
Winter oxygenated gasoline requirements were superimposed on RFG requirements in several
areas including portions of Pennsylvania, New Jersey, Washington, DC, Maryland, Virginia, Massachusetts,
New York and Connecticut. Consequently, the oxygen content of some of the winter RFG sold in these
areas was higher than necessary to meet the federal RFG requirement. These areas were re-designated
to CO attainment at various times between 1996 and 2000, and winter oxygenated programs were
eliminated, except, in some cases, as contingency measures (EPA, 2005). The elimination of these
programs contributed to the downward trend seen over a portion of the Winter MTBE and Winter oxygen
content trend lines. Since these areas used little or no ethanol-oxygenated RFG, it is unlikely that ethanol
trends were affected.
68
-------
Transitions from ethers to ethanol, elimination of winter oxygenate programs in RFC areas and
RFC program opt-ins/opt-outs affected specific areas at specific times. Only some RFC refiners, importers
and oxygenate blenders were affected. However, some of the areas affected were major markets, so
these localized changes at times materially affected overall trends. Tables contained in the appendix to
this chapter show seasonal oxygen weight percent, ethanol weight percent and MTBE weight percent
averages estimated from RFC surveys by area and year. The reader can determine where and when
significant changes in oxygen and oxygenate usage occurred.
Although the above factors are likely to have influenced RFC oxygen and oxygenate trends,
additional factors influencing oxygen and oxygenate use should be mentioned. This discussion is not
intended to be an in-depth analysis of the movements of RFC oxygen and oxygenate content over time
and the factors influencing these movements. Rather, it is intended to identify certain factors which
could be considered in such an analysis.
Compliance with averaged standards, and the use of oxygen credits contributed to the wide
range of oxygen content averages in RFC areas. The highest RFC area averages for each year-season
combination with the exception of 1995 occurred in areas where the preponderance of the gasoline was
ethanol blended at about 10 volume percent. The lowest area averages for each year-season
combination occurred in areas where all or virtually all gasoline was ether-oxygenated, primarily MTBE-
blended. Suppliers who blended this high oxygen content ethanol RFC generated credits which could be
made available to producers of ether-oxygenated RFC. These producers, in turn, could use credits to
help meet the 2.1 weight percent averaged oxygen content standard. This sometimes produced a
situation where the average oxygen content in one or more RFC areas dropped below 2.0 weight percent,
the Clean Air Act requirement for RFC. Figures 1 and 2 show that this occurred on a seasonal basis.
This sometimes occurred on an annual basis, as well, triggering RFC Survey failures. EPA's regulations
allowed adjustment of the per gallon minimum oxygen content standard applicable to certain suppliers
and RFC areas in the event of such failures. As previously noted, EPA did, in certain instances, "ratchet"
this standard. Per gallon minimum oxygen requirements became as high as 1.7 weight percent for some
areas and suppliers. However, even when a small oxygen shortfall did occur, the RFC sold in the
affected area met or exceeded all emissions performance requirements. Consequently, EPA elected not
to adjust per gallon minimum standards in response to oxygen survey failures occurring in 1998 and later
years. In accordance with EPA's regulations, a ratchet automatically became less stringent by 0.1 weight
percent if a ratcheted area passes surveys in two consecutive years. As of 2005, ratchets to a 1.6%
minimum were still in place in the Norfolk, VA and Washington, DC areas. Although the ratchets applied
to specific areas, the effects were not as localized as transitions from ether to ethanol or elimination of
winter oxygenated gasoline programs, but were likely to have applied to much of the total RFC supply
(see'80.41(q)). However, while the ratchets applied to a significant quantity of RFC, they may well have
had only a minor or negligible effect on oxygen or oxygenate content since the ratchets did not affect the
average oxygen content standard.
Comparison of Figures 1 and 2 shows that winter oxygen content levels were higher than
summer levels prior to 2000, but that summer levels were equal or higher in 2000 and later years.
Although there may have been other seasonal influences, the winter oxygenated fuel programs in effect
in several areas contributed to higher winter oxygen levels during the initial years of the RFC program.
As these programs were eliminated, other factors were able to influence seasonal oxygen content
differences. Since this seasonal high low reversal coincided with the transition from Phase I to Phase II
RFC, it is possible that the reversal is related to differences between Phase I and Phase II RFC.
The timing of localized changes not only affected these trend lines but probably affected the
variability of oxygen content within a reporting or survey year. For example, MTBE bans took effect in
New York and Connecticut in 2004, but RFC Surveys indicate that the transition from MTBE to ethanol
69
-------
took place at the end of 2003. As noted throughout this report, the winter data for a given year include
RFC produced and/or sampled both before and after the Summer VOC control season, and differences
between gasoline produced before and after the Summer season may have been exacerbated by these
transitions.
Figure 9 shows fluctuations in average RFC oxygen content by production quarter, based on the
production date reported for each batch. Batches produced during the 1st and 4th quarters are
predominantly Winter RFC, while batches produced during the 2nd and 3rd quarters are predominantly
Summer RFC.
Average RFC &RBOB Oxygen Content by Production Quarter
(from Batch Reports)
Year-Quarter
Figure 9
Figures 10 and 11 show the frequency distribution of oxygen concentrations for reporting year
2005 RFC (and RBOB) batches. For both distributions, the portion representing ethanol-oxygenated RFC
and RBOB blended (or assumed blended) at about 10 volume percent is clearly evident between about
3.3 and 4 weight percent oxygen. Very little RFC or RBOB was reported with oxygen content between
approximately 2.7 weight percent, the "substantially similar" upper limit exclusive of any waivers, and
about 3.3 weight percent. Virtually all of the remaining volume fell between approximately 1.5 weight
percent, the per gallon minimum allowed with averaged standards, and approximately 2.7 weight
percent. The batches in this oxygen content range are predominantly ether-oxygenated, however some
ethanol-oxygenated RBOB batches also fall within this range. Both distributions show a peak below 2.0
weight percent.
Several factors contributed to the variability of oxygen content within this 1.5 to 2.7 weight
percent range. Most RFC producers chose to comply with oxygen content standards on an average
rather than per gallon basis. The regulations permitting compliance on average and the use of credits to
meet the oxygen standard provided flexibility in the amount of oxygen a refiner or importer might use in
an individual batch as well as in total. In some cases, refiners and importers may have chosen to
produce some or all of their RFC with a relatively low oxygen content, using credits, while in other cases
they may have chosen to produce higher oxygen content ether-oxygenated RFC, hence not requiring or
70
-------
even creating credits. Additionally, refiners and importers may have chosen to blend MTBE or other
oxygenates in amounts greater than needed to meet the oxygen content requirement because
oxygenates are a good source of octane. Even if a producer did not, on average, exceed the oxygen
content requirement, the producer may have blended premium gasoline with higher oxygen or oxygenate
content than regular grade. EPA's data analyses indicate that, in aggregate, this was true for MTBE.
Table 1 shows volume-weighted average MTBE content (in volume percent), calculated for
regular and premium grade RFC (and RBOB if any) batches where MTBE content was greater than zero
and ethanol content equal to zero. It shows that higher levels of MTBE were used in premium RFC than
in regular RFC in each year and season. These averages may not precisely represent the actual MTBE
levels, by grade, found in retail MTBE-oxygenated RFC for various reasons (e.g. a portion of regular and
premium production is blended at points downstream of refineries to produce mid-grade gasoline.) RFC
Survey data analysis, presented in the appendix to this chapter, also estimates higher MTBE levels in
premium RFC than in regular or mid-grade. Since the survey-based MTBE averages include ethanol-
oxygenated RFC with no MTBE content, the averages underestimate the level of MTBE in MTBE-
oxygenated RFC.
Average MTBE Content (volume %) By Grade for RFC Batches with
Season
Summer
Winter
Grade
PRM
REG
PRM
REG
1997
11.08
9.78
11.52
9.86
1998
10.96
9.87
11.26
10.15
1999
10.65
9.70
10.72
9.68
2000
11.06
10.08
10.08
9.67
2001
11.19
9.99
10.31
9.54
2002
10.99
10.41
10.11
9.54
MTBE but no Ethanol
2003
10.97
10.57
10.33
9.67
2004
11.65
10.85
10.87
9.98
2005
11.69
10.16
10.67
9.67
Table 1
9.00%
8 00% -
7.00% -
6.00% -
IP
i 5 00% -
3?
4 00% -
Frequency Distribution of Oxygen Content-Summer 2005 RFC & RBOB
(from Batch Reports-Data Grouped into 0.05 wt% Intervals)
2 2.5
Oxygen (Wt%)
Figure 10
71
-------
Frequency Distribution of Oxygen Content-Winter 2005 RFG & RBOB
(from Batch Reports-Data Grouped into 0.05 Wt% Intervals)
2 2.5
Oxygen (Wt°i>
Figure 11
Since the "substantially similar" waiver for gasohol allows ethanol blending at any level up to 10
volume percent, an ethanol-oxygenated RFG batch could have contained any amount of oxygen between
approximately the per-gallon minimum and approximately the 4.0 weight percent Complex model
maximum. Analysis of the 2005 reporting data indicates that virtually no batches fell within a portion of
this range. Although downstream oxygenate blenders may in some cases have blended and distributed
RFG with oxygen content within the "empty" portion of the reporting data distribution, analysis of RFG
Survey data confirms that very little of the RFG sold during that period contained oxygen greater than 2.7
weight percent and less than 3.3 weight percent. In comparison to the reporting data analysis, the
survey data analysis also estimates a slightly larger fraction of the total RFG volume containing 3.3 or
more weight percent oxygen, and a slightly smaller fraction containing 2.7 percent or less oxygen. This
last finding is consistent with other indications that more ethanol is blended downstream than is assumed
for refiner and importer RFG compliance reporting.
There are several additional factors which contribute to the shape of the distributions pictured,
and in particular, explain why very little volume falls between about 2.7 and 3.3 weight percent. As
noted, the excise tax incentive as structured until a recent legislative change favored 10 volume percent
ethanol blending and was tied to three discrete volume concentrations (10%, 7.7% and 5.7%) with 7.7%
nominally equivalent to 2.7 weight percent oxygen. EPA's RFG regulations also, in certain circumstances,
provided incentives for using ethanol at 10 volume percent. These incentives included a relaxed VOC
performance standard applicable to 9-10 volume percent ethanol-blended Summer RFG sold in the
Chicago and Milwaukee areas and the oxygen credit program (i.e. the most credits would be generated
when ethanol is blended at around 10 percent). Additionally, when ethanol is blended into a
hydrocarbon blendstockto make gasoline, an RVP increase occurs which does not occur if MTBE or other
ethers are blended to make gasoline. Consequently, low volatility RBOB is needed to produce ethanol-
oxygenated RFG during the Summer to ensure that the finished RFG will meet VOC performance
standards. Thus refining the blendstock needed to produce ethanol-oxygenated RFG is sometimes more
costly than refining the blendstock needed to produce ether-oxygenated RFG. However, the blendstock
needed to produce 10 percent ethanol-oxygenated RFG may not be more expensive to refine than the
blendstock needed to produce 5.7 volume percent ethanol-oxygenated RFG, or RFG with any intermediate
ethanol content. Lastly, EPA's RFG regulations pertaining to downstream oxygenate blending limit the
72
-------
amount of oxygen that refiners or importers can claim to 2.0 weight percent, and in certain circumstances
requires that a refiner or importer assume that only 4.0 volume percent ethanol (about equivalent to the
1.5 weight percent oxygen per gallon minimum) will be blended downstream, unless they meet certain
contractual and quality assurance requirements (see '80.69). While it would normally have been
advantageous for refiners and importers to claim as high an oxygen content as possible in their
compliance calculations, there would have been some trade-off between the benefits of claiming an
oxygen content equivalent to 10 volume percent ethanol, and any costs associated with meeting these
blender oversight requirements. If a refiner or importer elected not to meet these requirements, the
amount of oxygen assumed would be 2 weight percent or less and if he elected to meet these
requirements, he would in all likelihood have chosen to claim an oxygen content equivalent to 10 volume
percent ethanol. Presumably, if refiners or importers produced RBOB for ethanol blending, there was
little economic motivation to claim an ethanol content greater than that required to meet the 5.7 volume
percent threshold but substantially less than 10 volume percent.
The appendix to this chapter contains tabular and graphical percentile (i.e., cumulative
distribution) data for oxygen content, ethanol content and MTBE content based on reporting data. The
ethanol percentile charts clearly show that ethanol content falls into discrete ranges, as would be
expected from the above discussion. An additional factor contributing to the total absence of gasoline
volume in a portions of this distribution is that ethanol is seldom reported as used in combination with
other oxygenates. The MTBE percentile charts show that there are some batches that contain MTBE in
amounts significantly greater than zero but at a concentration insufficient to meet a 1.5 weight percent
per-gallon minimum oxygen standard (about 8 volume percent MTBE). This occurred because it is not
unusual to find MTBE in combination with other oxygenates such as TAME in a given batch of gasoline.
Since there is interest in the absolute volume of oxygenates in gasoline, EPA has included
estimates of non-oxygenated hydrocarbon and oxygenate volume for federal RFC. Table 2 shows volume
estimates calculated from the batch volumes and oxygenate volume percent concentrations contained in
refiner and importer batch reports.
EPA believes that these totals, in general, provide good, but imperfect, estimates of the volumes
of oxygenates used in federal RFC outside of California. 29 Although some reporting errors and omissions
exist, the impact on overall totals, with the exception of the ethanol volume, is likely to be small.
Reliance on the refiner and importer batch data to estimate ethanol volumes is likely to be a more
significant source of error in the ethanol volume estimates, for two reasons. First, as previously
explained, the amount of ethanol blended downstream may exceed the amount assumed by refiners and
importers. Second, although most RBOB refiners and importers report the oxygenate type and volume
percent concentrations used in their "hand blends", some parties omit this data. EPA's regulations
require that refiners and importers include the oxygenate volume in the volume reported for each RBOB
batch but do not explicitly require that they submit the oxygenate type and volume percent
concentration. Thus, EPA's oxygenate volume calculations in table 2 would not have included the
oxygenate volumes for these batches. Consequently, the ethanol volumes contained in the above table
are virtually certain to be lower than the actual volume of ethanol in RFC.
EPA's regulations require that RFC oxygenate blenders report oxygenate volume data. While
these data provide an alternative means to develop an RFC ethanol volume estimate, blenders complying
29The oxygenate volume estimates in this report are based on analysis of data and information submitted to
EPA for purposes other than tracking the absolute volume of oxygenates used in gasoline. Although, in some cases,
the calculated volume estimates are shown to the nearest gallon this level of precision and accuracy is not implied;
these estimates are approximate. EPA has not determined that these volume estimates are the best available
published information on this topic. Readers should be aware that the Department of Energy's Energy Information
Administration collects and publishes official energy statistics from the US government (see http://www.eia.doe.gov/).
73
-------
with the per gallon oxygen content standard do not submit batch reports, complicating analysis of these
blender data. In lieu of blender data analysis, EPA has performed other analyses which quantify the
sensitivity of ethanol and total volume estimates to downstream blending.
Table 3 compares the estimates of ethanol and total RFC volume calculated directly from refiner and
importer batch reports with two alternatives (shown in bold font). The first alternative assumed that
each batch that did not contain oxygenate type and volume percent concentration data was oxygenated
with 4.0 volume percent ethanol. (This "minimum value" assumption has a regulatory basis under
certain circumstances; however in this analysis EPA applied this value to all "omitted data" batches
regardless of batch type or other information.) The amount of ethanol attributed to these omitted data
batches was added to the amount directly calculated from reported volumes and concentrations. The
second alternative assumed that all batches containing at least 3.8 volume percent ethanol and all
batches with omitted oxygenate data were oxygenated with 10 volume percent ethanol. Although this
had a small effect on ethanol totals, it also assumed batches with less than 3.8 volume percent ethanol
contained the directly calculated ethanol volume. The assumptions in the second alternative also have
an effect on the volume for certain batches, and the table also shows the effect on the total RFC volume.
Since the "10 volume %" assumption is more consistent with survey data and other information
than the "minimum value" assumption. EPA expects that the lower reporting-based estimates are
virtually certain to have underestimated actual ethanol usage in RFC. However, while the higher based
estimates are expected to be closer to actual usage, they may possibly overestimate ethanol volume.
RFC survey sampling occasionally found samples outside of California where ethanol was present in
concentrations significantly lower than 10 volume percent. Additionally, even when ethanol is blended at
a nominal 10 volume percent, the actual ethanol concentration is usually slightly less than 10 percent
because the ethanol blended into gasoline is denatured; i.e. it contains a small amount of an impurity
making it unfit for human consumption.
Non-reporting, reporting errors and omissions as well as violation of EPA's regulatory limits on
ethanol in gasoline could have resulted in RFC ethanol volume outside of the estimated ranges.
However, EPA believes that the actual ethanol volume in RFC, in each year, while almost certainly above
the lower reporting-based estimate, would not have significantly exceeded the higher reporting-based
estimate. To further investigate this assumption, EPA also estimated ethanol volume in each year using
the average RFC ethanol content estimates derived from RFC Survey data together with the annual RFC
volume estimates that the RFC Survey Association provides in its survey plans. EPA assumed that 45%
of this volume is VOC-controlled Summer RFC and 55% Winter RFC, since these are approximately the
ratios seen in the reporting data. These survey-based estimates of RFC ethanol volume are also
summarized in Table 3. In each year through 2004 the survey-based estimates fell between the high and
low reporting-based estimates, closer to the higher reporting estimate. In 2005 the survey-based
estimate exceeded the higher reporting-based estimate by less than three percent. These survey-based
approximations, in general, indicate that the actual RFC ethanol volume in each year probably fell
between the low and high reporting-based estimates. Although the 2005 survey-based estimate slightly
exceeded the upper reporting-based estimate, this does not strongly indicate that actual ethanol volume
exceeded the reporting-based estimate, given the approximate nature of these calculations.
As part of the process leading to the finalization of regulations to implement the renewable fuels
program requirements of the Energy Policy Act of 2005, EPA published a Regulatory Impact Analysis
(RIA) addressing the emissions, air quality and economic impacts of the program. 30The RIA contained
estimates of volumes for ethanol and MTBE use, based on historic data from various sources, to represent
baseline conditions. EPA derived the estimates of oxygenate volume in this "trends" report independently
from the analyses supporting its Renewable Fuels Standard (RFS) Regulations.
30The RIA is available at http://epa.gov/otaq/renewablefuels/420r07004.pdf
74
-------
Given the significance of the RFS program it is appropriate for this "trends report" to include
some comparison of estimates to identify any significant disparities. Examination of comparable results
indicates that they are in reasonably good agreement with respect to RFC oxygenate volumes.
Specifically, estimates of oxygenate use in 2004 federal RFC contained in this report can be compared
with 2004 base year estimates in the RFS RIA.
Ethanol volume totals in Table 3 of this chapter can be compared to the sum of the 2004 PADD 1,
2 and 3 RFC ethanol consumption estimates in the RFS regulatory impact analysis (1,230 million gallons).
31 The ethanol volume estimates in the trends report are intended to estimate the volume of ethanol
without considering the volume of any denaturant that may be present in the ethanol. For various other
purposes, including compliance with the RFS, and Internal Revenue Service excise tax regulations, the
volume of denaturant is included. Assuming that 5% of the ethanol consumption estimate in the RFS RIA
is denaturant, it is appropriate to compare the trends report volume estimates to 95% of the RFS
estimate (1,169 million gallons). The comparable trends report ethanol volume estimates are about 78%
(916 MM gallons-low reporting), 93% (1,090 MM gallons-surveys) and 96% (1,120 MM gallons - high
reporting) of this volume. The upper reporting system-based estimate, which assumed that ethanol is
blended into RBOB at the 10 volume % limit, agrees well with both the survey-based estimate in the
trends report and the estimate in the RFS RIA.
Estimates of 2004 RFC MTBE volume shown in table 2 of this chapter can be compared with 2004
estimates in the RFS RIA. The trends report estimate (2,135 MM gallons) can be compared with the total
RFC MTBE volume (1,878 MM gallons) given in the RFS RIA.32 This latter total included 19 MM gallons
used in PADD5, where Arizona's "Cleaner Burning Gasoline" fuel regulations impose requirements similar
to those of federal and California RFC programs in the Phoenix area.
Refiners and importers are required to report volume percent MTBE content and batch volume for
each RFC batch. MTBE was typically blended at the refinery, rather than at terminals, and there is no
requirement to add a denaturant to MTBE. Consequently, calculation of MTBE volume from these batch
report concentrations and volumes should provide a reasonably accurate estimate of the MTBE volume in
non-California gasoline produced as RFC in a given year. California banned MTBE use after January 1,
2004, so reporting system MTBE estimates for 2004 and 2005, unlike prior year estimates, should reflect
nationwide MTBE usage in RFC. The reporting system estimate exceeds the 2004 estimate in the RFS
RIA by less than 14 percent. The RFS RIA noted that the Energy Information Administration reported
2004 MTBE usage in RFC as 2.0 billion gallons, but the RFS RIA analysis reduced this to 1.9 billion
gallons.33 Although the RFS RIA estimate is lower than the trends report estimate, the RIA estimate is
characterized as a "consumption" estimate while the trends report estimate is a refiner/importer estimate,
and MTBE consumption in RFC areas could have been smaller than the MTBE content in gasoline
produced as RFC since some RFC could have been marketed outside of RFC areas. Since EPA's
reporting-based MTBE volume estimates are expected to closely estimate actual MTBE use in RFC, the
trends report does not include a table of survey-based MTBE volume estimates, as it does for ethanol.
However, using the same basic assumptions, the survey-based MTBE use estimate for 2004 RFC is about
2.0 billion gallons.
EPA has a limited basis for estimation of oxygenate volumes in the RFC sold in federal RFC areas
in California. (These are areas where RFC is mandated by the Clean Air Act and federal regulations,
based on ozone air quality.) EPA's regulations (40 CFR 80.81), which exempt refiners importers and
oxygenate blenders of California gasoline from many of the reporting requirements applicable to gasoline
sold outside of California, provided for oxygen surveys in these areas to ensure that the federal RFC
oxygen content standard was met.
31 See RIA table 2.2-5 page 63
32 See RIA table 2.2-3 page 59
33 See footnote 26 page 57 of the RIA.
75
-------
of on
Season
Summer
Winter
Annual
Data
MTBE
Ethanol
ETBE
TAME
T_butanol
Methanol
Hydrocarbons w/o
oxygen
Total Volume
MTBE
Ethanol
ETBE
TAME
T_butanol
Methanol
Hydrocarbons w/o
oxygen
Total Volume
MTBE
Ethanol
ETBE
TAME
T_butanol
Methanol
Hydrocarbons w/o
oxygen
Total Volume
1,071,597,101
116,719,792
163,905
96,796,209
1,852,929
2,409,022
11,209,230,173
12,498,769,130
1,247,140,935
176,044,353
163,829
102,918,788
2,525,600
2,165,515
13,400,206,530
14,931,165,550
2,318,738,035
292,764,145
327,734
199,714,996
4,378,529
4,574,536
24,609,436,703
27,429,934,680
1,125,469,361
126,671,140
57,783
97,298,332
1,662,098
1,842,350
11,484,418,551
12,837,419,615
1,287,536,349
182,154,056
404,759
102,954,626
2,322,076
1,417,924
13,507,944,631
15,084,734,421
2,413,005,710
308,825,195
462,542
200,252,958
3,984,174
3,260,274
24,992,363,182
27,922,154,036
1,089,183,554
135,612,567
76,521
109,093,183
2,129,311
1,366,976
11,665,634,998
13,003,097,110
1,234,898,146
172,692,564
145,423
118,188,700
2,740,995
1,861,992
13,553,712,580
15,084,240,402
2,324,081,700
308,305,131
221,944
227,281,884
4,870,306
3,228,969
25,219,347,578
28,087,337,512
1,135,416,556
144,765,712
529,920
119,311,611
2,306,023
942,867
11,579,895,788
12,983,168,478
1,256,617,034
188,289,969
722,267
113,252,505
2,202,582
1,073,472
14,268,916,880
15,831,074,709
2,392,033,591
333,055,681
1,252,187
232,564,116
4,508,605
2,016,339
25,848,812,668
28,814,243,187
1,154,011,408
140,950,792
550,207
115,382,581
2,700,257
2,006,367
11,814,114,444
13,229,716,057
1,249,894,911
179,198,097
2,353,316
108,920,290
2,538,465
1,915,335
14,238,183,530
15,783,003,943
2,403,906,319
320,148,888
2,903,523
224,302,871
5,238,723
3,921,702
26,052,297,974
29,012,720,000
1,199,903,109
173,837,314
211,833
102,443,754
3,191,915
939,210
12,367,444,499
13,847,971,634
1,269,523,975
207,787,969
1,857,791
93,243,310
2,829,843
1,758,366
14,856,998,466
16,433,999,720
2,469,427,084
381,625,284
2,069,624
195,687,064
6,021,758
2,697,576
27,224,442,965
30,281,971,354
1,175,062,810
186,597,775
777,507
109,992,263
2,769,361
950,143
12,108,710,985
13,584,860,845
1,224,289,829
264,607,450
176,153
59,626,275
2,839,813
1,469,656
15,126,763,959
16,679,773,135
2,399,352,639
451,205,225
953,660
169,618,537
5,609,175
2,419,800
27,235,474,944
30,264,633,980
1,017,763,982
418,865,421
124,041
57,917,198
3,302,839
1,123,509
12,733,380,559
14,232,477,549
1,116,862,067
425,217,760
285,687
51,860,959
2,673,376
798,214
15,596,672,836
17,194,370,899
2,134,626,049
844,083,182
409,728
109,778,157
5,976,215
1,921,723
28,330,053,395
31,426,848,448
970,917,488
392,714,925
1,417,282
59,781,098
2,879,387
816,929
12,663,961,928
14,092,489,036
1,121,563,294
458,144,406
4,288,669
46,435,024
2,788,917
1,007,389
16,410,604,358
18,044,832,058
2,092,480,782
850,859,331
5,705,951
106,216,123
5,668,304
1,824,318
29,074,566,286
32,137,321,094
Table 2
76
-------
Annual Ethanol Volumes (gallons)
Estimated from Batch
concentrations & volumes
All batches with unreported
oxygenate data assumed 4% EtOH
All batches with >= 3.8% ethanol
or unreported oxygenate data
assumed 10% EtOH
Total Gallons (from batch
volumes)
All batches with unreported
oxygenate data assumed 4% EtOH
All batches with >=3.8% ethanol
or unreported oxygenate data
assumed 10% EtOH
Estimated Average Ethanol WT%
Concentration-Summer
Estimated Approximate Ethanol
Vol% Concentration (assume blend
sg=0.745,etoh sg=0.7939)
Estimated Average Ethanol WT%
Concentration-Winter
Estimated Approximate Ethanol
Vol% Concentration (assume blend
sg=0.730,etoh sg=0.7939)
Estimated Annual RFG Total
Gallons-except CA (From RFG
Survey Plans)
Estimated Annual Ethanol Gallons
(assuming 45% Summer RFG/55%
Winter RFG)
of
292,764,145 308,825,195 308,305,131 333,055,681 320,148,888 381,625,284 451,205,225 844,083,182 850,859,331
332,630,715 337,543,715 343,328,264 371,560,055 361,793,029 423,492,651 504,208,308 915,998,927 935,017,799
480,033,112 471,648,802 485,801,666 501,070,555 492,370,440 592,815,198 677,741,183 1,119,661,357 1,143,158,771
27,429,934,680 27,922,154,036 28,087,337,512 28,814,243,187 29,012,720,000 30,281,971,354 30,264,633,980 31,426,848,448 32,137,321,094
27,429,934,680 27,922,154,036 28,087,337,512 28,814,243,187 29,012,720,000 30,281,971,354 30,264,633,980 31,426,848,448 32,137,321,094
27,517,537,221 28,013,181,344 28,177,276,214 28,885,997,127 29,080,831,200 30,388,492,850 30,358,662,231 31,522,637,261 32,219,224,365
1.36
1.28
1.83
1.68
1.46 1.60 1.60 1.57 1.81
1.37 1.50 1.50 1.47 1.70
1.68
1.54
1.85
1.70
1.96
1.81
1.47
1.81
1.66
1.95
1.79
1.96
1.84
2.52
2.31
3.72
3.49
3.97
3.65
3.81
3.58
4.20
3.86
28,352,970,000 28,759,802,000 29,506,498,000 28,922,480,000 29,046,358,000 29,273,367,000 29,979,353,000 30,460,161,000 31,483,686,000
424,864,889 420,987,795 475,895,074 483,112,058 458,271,047 511,926,099 629,712,996 1,090,058,001 1,175,055,482
Table 3
77
-------
EPA has received both Summer and Winter oxygen survey data for three areas, Los Angeles, San
Diego and Sacramento, since 1997, and for San Joaquin since 2003. These survey data provide
estimates of oxygenate concentrations which can be applied to gasoline volume data to estimate
oxygenate volumes. EPA does not collect such volume data, however the RFC survey plans and other
survey-related documents submitted to EPA contain estimates of annual gasoline volumes for these as
well as non-California RFC areas. These estimates were derived from published population and per-
capita gasoline consumption data and were generated to determine the total number of RFC surveys that
are required in a calendar year. EPA has used these volume estimates throughout this report as
weighting factors when averaging survey data from different areas. The gasoline volume estimates for
the California areas have been used with the survey oxygenate concentration averages in order to
generate the oxygenate volume estimates shown in Table 4. These table values should be considered as
"ballpark" estimates, and more accurate estimates of ethanol use in these California areas may be
available elsewhere. The reader is also reminded that these estimates do not include oxygenate use in
California gasoline sold in areas outside of the federal RFC areas. While California's regulations did not
require oxygen in this gasoline, oxygenates were sometimes used.
Estimates of 2004 California RFC ethanol volume shown in table 4 of this report can also be
compared with estimates in the RFS RIA.34 However, the trends report estimate (629 MM gallons)
included only the ethanol volume from RFC in the federally-mandated RFC areas in California where, in
2004, gasoline was required to contain oxygen as well as meet statewide California RFC standards. EPA,
for its statewide estimate in the RFS RIA (853 MM gallons) assumed, based on discussions with California
Air Resources Board staff, that all California gasoline contained 5.7% ethanol, by volume. These two
estimates differ primarily because the applicable areas differed.
Estimated Ethanol
Volumes
Los Angeles, CA
Sacramento Metro, CA
San Diego, CA
San Joaquin, CA
Total
Estimated MTBE Volumes
Los Angeles, CA
Sacramento Metro, CA
San Diego, CA
San Joaquin, CA
Total
Estimated TAME Volumes
Los Angeles, CA
Sacramento Metro, CA
San Diego, CA
San Joaquin, CA
Total
ed
0.0
0.0
0.0
#N/A
0.0
717.8
99.5
124.6
#N/A
941.9
18.3
3.3
6.5
#N/A
28.2
0.0
0.0
0.0
#N/A
0.0
749.8
103.6
128.6
#N/A
982.0
4.2
4.9
1.7
#N/A
10.8
* Estimates based
of In Fc
0.0
0.0
0.0
#N/A
0.0
796.0
105.3
137.9
#N/A
1039.2
8.6
3.2
2.6
#N/A
14.4
on RFC Survey
0.0
0.0
0.0
#N/A
0.0
800.3
106.1
142.1
#N/A
1048.5
10.5
4.7
2.3
#N/A
17.6
oxygenate
57.9
5.4
5.5
#N/A
68.9
670.5
90.6
134.6
#N/A
895.7
11.5
4.5
0.9
#N/A
16.9
measurements
55.0
8.1
6.0
#N/A
69.2
658.9
119.2
122.0
#N/A
900.2
6.1
5.5
0.1
#N/A
11.6
and survey
RFC Au-o« »
364.4
45.2
55.3
42.3
507.3
44.0
46.1
24.9
82.8
197.8
0.2
2.2
0.1
0.1
2.5
404.7
70.5
68.3
85.2
628.7
0.0
0.0
0.1
0.0
0.1
2.0
1.1
0.6
1.8
5.5
plan estimates of gasoline
397.2
59.5
68.0
85.7
610.4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.5
0.0
0.5
volumes
Table 4
See RIA Table 2.2-3 page 59
78
-------
by
Tables 5 through 8 show 2004 and 2005 reporting averages by PADD for PADD I, II and III
refiners. These averages were calculated from batch data, excluding importer batches (see "PADD Level
Analysis" appendix for additional information):
& by
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
Average
Value (wt %)
2.59
3.48
2.36
2.45
3.26
2.09
2.51
3.35
Gasoline Volume
(gal)*
4,726,753,777
1,390,108,766
5,818,619,897
5,896,265,229
1,824,393,952
5,997,046,241
10,623,019,006
3,214,502,718
III 2.22 11,815,666,138
^Volumes exclude batches with missing values
Average
Value (wt %)
2.50
3.50
2.25
2.41
3.24
2.05
2.45
3.35
2.15
for this parameter
Gasoline Volume
(gal)*
4,224,319,396
1,515,851,939
5,620,100,245
6,230,237,499
2,129,892,557
5,739,092,825
10,454,556,895
3,645,744,496
11,359,193,070
Table 5
& by
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
III
Average
Value (vol %)
3.01
9.39
1.75
3.00
8.61
0.98
3.01
8.95
1.36
Gasoline Volume
(gal)*
4,726,753,777
1,390,108,766
5,827,786,523
5,896,265,229
1,824,393,952
5,997,046,241
10,623,019,006
3,214,502,718
11,824,832,764
^Volumes exclude batches with missing values
Average
Value (vol %)
3.14
9.46
1.31
3.02
8.63
0.91
3.07
8.97
1.11
for this parameter
Gasoline Volume
(gal)*
4,224,319,396
1,515,851,939
5,620,730,455
6,227,516,269
2,129,892,557
5,739,092,825
10,451,835,665
3,645,744,496
11,359,823,280
Table 6
79
-------
Season
Summer
Winter
Annual
2004 & 2005
PADD
I
II
III
I
II
III
I
II
III
*"i M
Average Value
(vol %)
8.02
0.00
8.74
7.11
0.00
8.52
7.52
0.00
8.63
^Volumes exclude batches
by
$"fiif*i»» iP»
Gasoline
Volume (gal)*
4,726,753,777
1,390,108,766
5,827,786,523
5,896,265,229
1,824,393,952
5,997,046,241
10,623,019,006
3,214,502,718
11,824,832,764
with missing values
Average Value
(vol %)
7.13
0.00
8.96
6.82
0.00
8.53
6.95
0.00
8.74
for this parameter
Only)
Gasoline
Volume (gal)*
4,224,319,396
1,515,851,939
5,620,730,455
6,227,516,269
2,129,892,557
5,739,092,825
10,451,835,665
3,645,744,496
11,359,823,280
Table 7
Season
Summer
Winter
Annual
&. bf
2P' ;
PADD Average Value Gasoline
(vol %) Volume (gal)*
I 0.14 4,726,753,777
II 0.00 1,390,108,766
III 0.55 5,827,786,523
I 0.05 5,896,265,229
II 0.00 1,824,393,952
III 0.68 5,997,046,241
I 0.09 10,623,019,006
II 0.00 3,214,502,718
III 0.62 11,824,832,764
^Volumes exclude batches with missing values
IE
Average Value
(vol %)
0.28
0.00
0.59
0.09
0.00
0.53
0.17
0.00
0.56
for this parameter
Only)
Gasoline
Volume (gal)*
4,224,319,396
1,515,851,939
5,620,730,455
6,227,516,269
2,129,892,557
5,739,092,825
10,451,835,665
3,645,744,496
11,359,823,280
Table 8
Figures 12 through 17 show estimates of average oxygen, ethanol and MTBE levels by PADD in
retail RFC.35 These averages are volume-weighted averages of the seasonal averages for each area,
using gasoline volume estimates supplied in the survey plans. (Oxygenate concentrations for these retail
data are in weight percent.) In 2005, RFC surveys were conducted in 18 PADD I areas, five PADD II
areas, two PADD III areas and four PADD V areas. It is obvious that virtually all PADD II RFC in more
recent years was ethanol-oxygenated. Increased ethanol use in PADDs I and V during the "trends report"
time period occurred primarily as a result of state MTBE bans. The survey averages show a small
amount of ethanol use in PADD I winter RFC even prior to these bans, but no ethanol use in summer
RFC. EPA regulations contained prohibitions against combining VOC-controlled ethanol-oxygenated RFC
with VOC controlled RFC produced using any other oxygenate during the January 1st through
September 15th period. Thus, limited ethanol use in predominantly MTBE-oxygenated RFC areas would
have been more feasible in winter than summer RFC.
35Since TAME was always found in conjunction with MTBE but at much lower concentrations, PADD-specific
retail averages are not shown. Area-specific RFC Survey average oxygenate concentrations for TAME and other
oxygenates can be found at http://epa.gov/otaq/regs/fuels/rfg/properf/rfgperf.htm.
80
-------
Average Oxygen Content of Summer RFC Sold at Retail Stations-By PADD
2.50 -
2.00 -
1.50 -
1.00 -
0.50 -
DPADD1
PADD2
D PADD3
D PADD5*
I
»
1995
2 11
253
2.13
i
i
1996
200
3.25
i ea
205
I
^
1997
2.07
3.33
.98
2.05
1998
2.Q5
3 34
205
205
"I " Bars denote canqe of RFG ar
-
T-,
1999
2.04
3.22
2.00
2.19
i
3r
I
2000
2.10
3 19
2 07
2.04
j
1
T
2001
209
3 19
208
208
2002
206
3.30
2 12
2 11
i
ea av
JL
2003
213
3.52
2.21
2.D4
erases
i
ii
2004
2.S2
3.54
2.17
2.08
1
£,
2005
2.44
3.61
2.03
2.13
* Federal RFG Areas in CA only
Year
Figure 12
Average Oxygen Content of Winter RFG Sold at Retail Stations-By PADD
2.50 -
1.50 .
1.00 -
0.50 -
DPADD1
PADD2
n PADD3
n PADD5*
1
1
:
:
1
33:
3£
97
K
I
i
1
Z
3
39t
.14
2E
.85
|
1
1
:
3
:
V
2E
24
9:
04
I
2
3
1
;
39E
?;
51
97
15
1
;
3
2
39i
OE
4C
.87
IE
^
2
:
c
:
:
30C
04
40
01
::.-:
X
2
3
:
:
:o
.9"
3:
o;
.1;
i.
J
3
2
:io:
93
43
99
06
200;
20E
3.63
203
20:
200^
246
367
.99
2 14
J,
J
2005
2.46
367
2.03
2.05
* Federal RFS Areas in CA only
Figure 13
81
-------
Average Ethanol Content (WT%) of Summer RFC Sold at Retail Stations-By
PADD
I
0.01
000
0.00
0.00
0.00
0.72
0.91
4.88
597
6.11
* Federal RFC Areas in CA only
Figure 14
Average Ethanol Content (WT%) of Winter RFC Sold at Retail Stations-By
PADD
6.00 -
4.00 -
D PADD1
PADD2
D PADD3
D PADD5*
_,J
ji
1985
0.26
6.86
0.01
J.
1996
0.12
8.72
0.00
^
1997
0.20
8.75
000
000
~i r
"I" Bars
i
denote r
1 1 1
J J Jm i
1998
0.17
9.66
0.00
000
19B9
0.17
8.95
DOO
000
2000
0.26
8.96
0.01
0.00
2001
0.15
8.73
0.00
OS9
nncje of I
n i
2002
0.13
9.40
000
071
FG area
1
I
2003
077
10.42
0.01
4.90
averages
3-
2004
3.04
10.54
O.DO
6.09
=fc
2005
3.34
10.57
0.00
5B5
* Federal RFG Areas in CA only
Year
Figure 15
82
-------
Average MTBE Content (WT%) of Summer RFG Sold at Retail Stations-By
PADD
id nn -. ,
MTBE (wt%)
12.00 -
10.00 -
6.00 -
4.00 -
2.00 -
D PADD1
PADD2
D PADD3
D PADD5'
1
I
1
1995
10.67
2.23
10.03
1
II
ft
1996
10.14
1.22
a.ee
10.B8
rl
1
1997
0.61
29
9.59
0.95
1
::
4 n
II
1996 198
1047 10.;
1.1S .9
10.17 10.;
11.11 11.E
11 -
n?
I
i
9 2000
5 10.45
2 2.2D
G 10.56
9 11.05
T
2001
10.46
215
0.40
9.88
I
:
\
2002
10.49
0.70
10.72
8.71
£
2003
10.62
o.oa
1 1 .33
.90
I
''ill
2004
8.15
0.03
11.13
0.00
]
i
2005
7.65
001
10.23
0.00
* Federal RFG Areas in CA only Year
Figure 16
TBE (wt%)
S
Av
16.00 -
14.00 -
12.00 -
10.00 -
8.00 -
6.00 -
4.00 -
2.00 -
0.00 -
DPADD1
PADD2
n PADD3
D PADD5*
erage MTBE Content (WT%) of Winter RFG Sold at Retail Stations-By PADD
If
n
y
1995
11.68
2.75
9.53
r!
n
i n
1996 1997
10.50 1103
1.1B .12
8.72 9.62
16.67
"I
i n F
. f Irl
1
1
1998 1999
11.42 10.21
0.05 1.53
9.82 9.09
11.38 11.73
I
r
2000
9.84
1.56
10.20
12.02
u
2001
975
1.40
10.05
9.91
|
ri
1
i
a
2002
9.74
0.07
10.11
9.91
r3
I
2003
9.34
0.09
10.37
1.71
!
2004
7.38
0.03
990
0.00
'-I
2065
690
0.01
10.21
0.00
* Federal RFG Areas in CA only Year
Figure 17
Additional Analysis and Observations-RFG
Data analyses pertaining to RFG oxygen and oxygenates are contained in the Appendix to this chapter.
Overview-Conventional Gasoline
For various reasons, analysis of EPA's CG data provides less information about oxygen and
oxygenates use than analysis of EPA's RFG data. As noted, refiners and importers do not always report
83
-------
oxygenates added downstream of refineries and import facilities, so EPA's data are likely to
underestimate ethanol use in CG, possibly to a greater extent than in RFC, and may underestimate ether
use as well. Unlike downstream blenders who add oxygen to RBOB, CG oxygenate blenders are not
subject to reporting requirements. Additionally, since there is no nationwide oxygen content standard
applicable to CG as there was with RFC, EPA cannot determine if refiner and importer CG batch reports
with blank oxygen and oxygenate data fields represent batches with zero oxygen content or batches with
data omissions. Since there was no regulatory requirement for CG surveys as there was with RFC, EPA
does not have retail data to compare or contrast with CG reporting data.
Nonetheless, EPA has computed estimates of CG oxygenate and hydrocarbon volumes from
reporting data. These estimates are contained in Table 9. Since these data are likely to underestimate
oxygenate use, and EPA lacks retail data to corroborate trends, any conclusions regarding CG oxygenate
volumes or trends based on these data should be considered uncertain. Therefore, this report limits its
interpretation of CG oxygenate data to a few observations.
Reporting system ethanol volume estimates for 2004 CG can also be compared with values in the
RFS RIA. Based on batch reporting data, EPA estimated approximately 357 million gallons of ethanol in
CG. The RFS RIA estimate (Tables 2-9) is 1,149 million gallons in CG and 193 MM gallons in winter
oxygenated gasoline. This is consistent with the assumption in the trends report that EPA's Anti-Dumping
reporting data significantly underestimates CG ethanol use.
In the trends report EPA estimated that about 410 MM gallons of MTBE were used in 2004 CG
production. For its 2004 base case analysis, the RFS RIA assumed that MTBE use outside of RFC could
be ignored. The trends report batch data analysis shows that the total MTBE volume used in 2004 CG
was about 19% of that used in RFC.
84
-------
of CG
Season
Summer
Winter
Annual
Data
MTBE
Ethanol
ETBE
TAME
T_butanol
Methanol
Hydrocarbons w/o
oxygen
Total Volume
MTBE
Ethanol
ETBE
TAME
T_butanol
Methanol
Hydrocarbons w/o
oxygen
Total Volume
MTBE
Ethanol
ETBE
TAME
T_butanol
Methanol
Hydrocarbons w/o
oxygen
Total Volume
252,881,317
68,787,726
321,936
27,389,697
978,010
1,469,751
39,636,317,978
39,988,146,413
221,924,891
90,932,967
232,124
25,929,645
679,657
709,534
46,584,731,037
46,925,139,854
474,806,208
159,720,692
554,059
53,319,341
1,657,666
2,179,285
86,221,049,015
86,913,286,267
263,857,150
103,158,524
364,469
15,739,451
1,025,554
440,993
39,315,845,797
39,700,431,939
244,488,419
140,558,756
82,274
15,370,098
680,049
602,052
47,915,978,752
48,317,760,399
508,345,569
243,717,280
446,743
31,109,549
1,705,602
1,043,045
87,231,824,549
88,018,192,338
239,812,977
104,005,119
697,051
29,062,988
1,038,110
677,409
38,498,391,810
38,873,685,463
200,707,656
179,725,078
724,484
21,838,185
685,072
546,135
48,602,598,915
49,006,825,526
440,520,633
283,730,197
1,421,534
50,901,173
1,723,183
1,223,544
87,100,990,725
87,880,510,989
on
249,459,785
93,305,613
1,203,214
34,129,294
1,151,257
1,051,256
39,112,534,538
39,492,834,956
202,568,448
171,860,389
1,351,462
23,721,415
837,127
1,155,641
49,333,020,314
49,734,514,796
452,028,233
265,166,002
2,554,676
57,850,709
1,988,384
2,206,898
88,445,554,851
89,227,349,752
261,046,445
122,469,291
445,999
43,476,191
1,330,605
1,218,595
41,208,233,201
41,638,220,326
202,532,803
183,660,878
496,032
37,451,523
958,779
1,178,733
50,204,198,335
50,630,477,083
463,579,248
306,130,168
942,031
80,927,714
2,289,384
2,397,328
91,412,431,536
92,268,697,409
303,540,761
123,138,938
4,229,741
29,299,178
763,296
635,660
43,984,491,302
44,446,098,876
180,866,801
179,499,685
1,034,938
20,716,356
656,387
654,325
48,288,412,977
48,671,841,470
484,407,562
302,638,623
5,264,680
50,015,533
1,419,683
1,289,985
92,272,904,279
93,117,940,346
220,336,203
151,645,217
4,600,108
20,856,029
604,433
494,447
43,604,900,088
44,003,436,524
189,744,455
205,465,808
1,645,897
18,822,251
555,308
413,736
47,906,967,487
48,323,614,941
410,080,658
357,111,024
6,246,004
39,678,280
1,159,741
908,183
91,511,867,575
92,327,051,465
156,726,080
158,885,968
1,969,397
17,903,319
413,913
213,725
42,513,780,773
42,849,893,176
145,192,111
231,622,489
4,744,327
18,436,905
262,795
282,911
49,058,601,515
49,459,143,051
301,918,191
390,508,457
6,713,724
36,340,223
676,708
496,635
91,572,382,288
92,309,036,227
Table 9
85
-------
Figure 18 shows that reported CG oxygenate volume is less than one percent of reported CG
volume. (RFC oxygenate volume has been about ten percent of RFC volume.) Based on analysis of
reported volumes and concentrations, MTBE volume percentage has decreased since 1998, while ethanol
volume percentage has increased. Decreases in MTBE volume share in CG may be partially attributable to
MTBE bans and phase-outs, which also affected some states or portions of states using CG. (The "non-
geographic" analysis presented here is insufficient to confirm a direct relationship between MTBE bans and
the MTBE volume decrease.) The increase in ethanol volume share estimated from reported volumes and
concentrations suggests increased use of ethanol in CG.
Oxygenate Volume Fractions from CG Reporting Data
1.00% -
0.90% -
0.80% -
0.70% -
| 0.60% -
"o
rc 0 50% -
2
° 040% -
5^
0 30% -
0.20% -
0.10% -
0.00% -
2001
2002
Reporting Year
Figure 18
CG Oxygen and Oxygenates by PADD
Tables 10 through 13 show 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners.
These averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information).
As with aggregate CG oxygen and oxygenate data, it is likely that oxygen and ethanol are under-reported:
2004 & 2005 Reporting Average by PADD-CG Oxygen (Refiner Batches Only)
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
III
2004
Average Value
(wt %)
0.44
0.34
0.17
0.37
0.40
0.14
0.41
0.37
0.15
Gasoline
Volume (gal)
3,752,272,894
11,456,437,187
21,722,608,623
3,484,116,076
13,323,919,987
23,628,091,025
7,236,388,970
24,780,357,174
45,350,699,648
2005
Average Value
(wt %)
0.34
0.34
0.13
0.30
0.33
0.11
0.32
0.33
0.12
Gasoline
Volume (gal)
3,547,001,722
10,566,219,165
20,889,546,279
4,064,064,814
13,589,333,929
22,443,420,127
7,611,066,536
24,155,553,094
43,332,966,406
Table 10
86
-------
&
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
III
by PADD-CG Only)
Average Value
(vol %)
0.82
0.97
0.06
0.83
1.16
0.05
0.82
1.07
0.05
Gasoline
Volume (gal)
3,752,272,894
11,456,437,187
21,722,608,623
3,484,116,076
13,323,919,987
23,628,091,025
7,236,388,970
24,780,357,174
45,350,699,648
Average Value
(vol %)
0.74
0.97
0.05
0.68
0.92
0.07
0.71
0.94
0.06
Gasoline Volume
(gal)
3,547,001,722
10,566,219,165
20,889,546,279
4,064,064,814
13,589,333,929
22,443,420,127
7,611,066,536
24,155,553,094
43,332,966,406
Table 11
& by Only)
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
III
Average Value
(vol %)
0.70
0.00
0.76
0.35
0.00
0.65
0.53
0.00
0.70
Gasoline Volume
(gal)
3,752,272,894
11,456,437,187
21,722,608,623
3,484,116,076
13,323,919,987
23,628,091,025
7,236,388,970
24,780,357,174
45,350,699,648
Average Value
(vol %)
0.39
0.00
0.60
0.30
0.00
0.41
0.34
0.00
0.50
Gasoline Volume
(gal)
3,547,001,722
10,566,219,165
20,889,546,279
4,064,064,814
13,589,333,929
22,444,880,341
7,611,066,536
24,155,553,094
43,334,426,620
Table 12
Season
Summer
Winter
Annual
&
PADD
I
II
III
I
II
III
I
II
III
Average Value
(vol %)
-0.17
0.00
0.04
0.03
0.00
0.02
-0.08
0.00
0.03
by Only)
Gasoline
Volume (gal)
3,752,272,894
11,456,437,187
21,722,608,623
3,484,116,076
13,323,919,987
23,628,091,025
7,236,388,970
24,780,357,174
45,350,699,648
Average Value
(vol %)
0.07
0.00
0.02
0.05
0.00
0.02
0.06
0.00
0.02
Gasoline Volume
(gal)
3,547,001,722
10,566,219,165
20,889,546,279
4,064,064,814
13,589,333,929
22,444,880,341
7,611,066,536
24,155,553,094
43,334,426,620
Table 13
87
-------
Benzene is an organic compound found in gasoline. It is one of a family of compounds known as
aromatics. Benzene may occur in crude oil. Benzene and other aromatics are also produced in a refinery
process called reforming. While gasoline aromatics content, in total, is an important parameter affecting
emissions, gasoline benzene content is also considered separately because gasoline benzene content has a
direct effect on exhaust and evaporative benzene emissions. Benzene emissions are of concern because
benzene is a toxic air pollutant known to be carcinogenic.
Benzene is an important petrochemical. Its uses include manufacture of other compounds used to
make plastics, nylon, dyes and detergents. Some refineries producing gasoline may also extract benzene.
The demand for benzene as a petrochemical, as well regulatory control of gasoline benzene content for
emissions purposes is likely to affect gasoline benzene content trends.
Several regulatory requirements directly or indirectly limit the benzene content in reformulated and
conventional gasoline. RFC benzene content is directly limited through benzene content standards. CG
benzene content is directly limited, but less strictly, through Anti-Dumping standards. RFC requirements
were intended to create a gasoline that was superior, in emissions performance, to the gasoline that was
supplied in 1990. CG "Anti-Dumping" requirements were intended to prevent the quality of CG from
deteriorating as a result of the RFC requirements. Consequently, both RFC and CG are subject to
regulations designed to limit toxics emissions. Since benzene emissions, which are the largest component
of these toxics emissions, increase with increasing gasoline benzene content, limits on toxics emissions
indirectly limit gasoline benzene content. These regulatory requirements are discussed in more detail
below.
RFC must meet a 1.00 volume percent "per gallon" maximum benzene content standard, or
alternatively, on an annual basis, a 0.95 volume percent "averaged" standard with a 1.30 volume percent
per gallon maximum. These benzene content standards have been in effect since 1995. RFC benzene
content has also been indirectly limited through toxics emission performance (i.e. "percent reduction")
standards since 1995. However, the performance standards, unlike the benzene content standards, have
not remained constant.
The Simple Model, which was the basis for the RFC toxics performance standard from 1995
through 1997, included estimates of exhaust and (for Summer) evaporative benzene emissions in its toxics
performance estimate. The Phase I Complex Model determined emissions performance for RFC in 1998
and 1999. The Phase II Complex Model determines emissions performance for RFC in 2000 and beyond.
While there are differences between the Phase I and Phase II Complex Models, both models include
exhaust and, for Summer, evaporative benzene in their toxics estimates. Importantly, Phase II RFC is
required to meet more stringent toxics reduction standards than Phase I RFC. As with the Simple Model,
the Complex Model considers benzene content as one of several gasoline parameters affecting benzene
emissions. The Complex Model also considers the emissions effects of several parameters, such as sulfur
content, which are not included in the Simple Model. Thus, although other factors are involved, the toxics
performance standard limits benzene content
CG benzene content is also both directly and indirectly limited. Annual average benzene levels
may not exceed the greater of the refiner's or importer's 1990 baseline for benzene or the Complex
Model's 4.9 volume percent benzene "acceptable range" maximum limit specified in EPA's regulations. CG
supplied from 1995 through 1997 was required to have annual average exhaust benzene emissions, as
calculated by a mathematical model, no greater than the refiner's or importer's compliance baseline for
88
-------
exhaust benzene emissions (i.e. refiners and importers must meet individualized standards for CG, rather
than generalized standards, as for RFC.) This exhaust benzene emissions model estimated exhaust
benzene as a function of gasoline benzene and aromatics content. CG supplied subsequent to 1997 is
required to have exhaust toxics emissions, based on the Complex Model, which do not exceed the refiner's
or importer's exhaust toxics compliance baseline. These emissions include other exhaust toxic pollutants
in addition to benzene, and several gasoline parameters in addition to benzene affect the benzene portion
of these Complex Model toxics emissions. Again, although other factors are involved, the emission
requirement indirectly limits CG benzene content.
EPA's Mobile Source Air Toxics (MSAT) regulations have imposed additional requirements on both
RFC and CG. These regulations do not regulate benzene content directly or require additional toxic
emission reductions. Rather, the MSAT regulations required refiners and importers, beginning on January
1, 2002, to produce or import gasoline with toxics emissions that are no worse than the toxics emissions of
gasoline they produced or imported during the period 1998 through 2000. EPA's data indicate that some
refiners and importers had over-complied with toxics emissions requirements. The MSAT rule is intended to
preserve this over-compliance to ensure that toxics emissions do not increase above current levels.36
's
Since gasoline benzene content data has been and is currently necessary to determine RFC and
CG compliance, refiners and importers have been required, since 1995, to submit this information to EPA's
RFG/Anti-Dumping reporting system for each batch of gasoline refined or imported. Additionally, since
1995, EPA has received RFC Survey data on the benzene content of RFC sold at gasoline stations. EPA
has analyzed reporting system benzene data from 1997 through 2005, and RFC Survey data from 1995
through 2005.
*T I / 'I 1 I I,
Figures 1 and 2 show Summer and Winter volume-weighted average RFC benzene content levels,
by year. Both the reporting system estimates, by reporting year, and the RFC Survey-based estimates, by
survey year, are shown. The decrease in benzene content between Summer 1999 and Summer 2000 RFC,
as estimated by both data sets, is greater in magnitude than any prior or subsequent year to year change.
The change in benzene content between Summer 1999 and Summer 2000 is also obvious from the
frequency distribution comparison shown in Figure 3, and in the distribution trend chart in the appendix to
this chapter.
It is clear that Summer RFC benzene content changed concurrently with the transition from Phase
I to Phase II RFC. (Regression analysis confirms the subjective conclusion-see "Regression Analysis"
appendix). Although the RFC benzene content standard did not change from Phase I to Phase II RFC, as
35 EPA has finalized an "MSAT2" regulation to replace the current MSAT requirements with a 0.62 volume
percent average benzene requirement beginning in 2011. See 72 F.R. 8427
89
-------
discussed above, the toxics performance requirement became more stringent with Phase II RFC. It is likely
that the more stringent Phase II toxics performance requirement is partially, if not primarily responsible for
the benzene content reduction, since benzene content affects Complex Model toxics emission performance
calculations. Benzene content does not directly affect Complex Model NOx or VOC emission estimates,
however, benzene content may also have been collaterally affected by refining and blending changes
made in order to comply with Phase II NOx and VOC standards.
The timing of the benzene content shift and the relationship between benzene content and toxics
emissions provide strong circumstantial evidence of a cause-effect relationship between the transition to
Phase II RFC and the shift in benzene content. The analysis presented in this report is, however, more
equivocal in supporting this relationship than in supporting other cause-effect relationships. Analysis
presented in the RFC Trends chapter suggests that benzene reduction was not the primary reason for
Phase I to Phase II toxics performance improvement and also shows that, although Summer toxics
performance improved with Phase II, on average, Phase I RFC complied with Phase II toxics requirements.
In contrast, similar analyses suggest that RVP reduction and sulfur reduction were, respectively, the
primary reasons for VOC and NOx performance improvements, and that, on average, Phase I RFC did not
comply with Phase II VOC or Summer NOx performance standards. Moreover, the demand for benzene as
a petrochemical would need to be considered in any rigorous analysis to explain gasoline benzene content
trends.
1
1
i
0.70 -
0
Highest RFG Area
Lowest RFG Area
Retail Avg
Production Avg
Averaged Std
Per Gallon Std
4
*
1995
1.01
0.47
0.67
0.95
1.00
i
.
1996
1.02
0.57
0.68
0.95
1 00
Wera
*
*
1997
1.05
0.55
0.68
0.66
095
1 00
je Su
*
it?**
,
1998
099
0.56
0.68
0.67
095
1 00
mmer
'
=
-------
a
>
ti
c
c
ili
'
070 -
0.60 -
'
Highest RFC Area
Lowest RFC Area
Retail Avg.
Production Avg
Averaged Std.
Per Gallon Std
Average Winter RFG Benzene Cor
i * : :
: : t
* ' ; i * . '
1995 1996 1997 199S 1999 2000 2001 2002
0.98 0.99 O.B9 0.95 095 0.87 0.89 090
0.49 0.57 0.58 0.53 0.56 0.54 0.52 0.56
0.59 0.69 0.67 0.65 068 0.65 0.65 0.66
0.63 0.65 065 0.65 0.64 064
0.95 0.95 0.95 0.95 0.95 0.95 095 0.95
1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
Year
itent
t
*
, -4
i i
2003 2004 2005
0 87 D.9B 0 05
0.62 0.57 0.58
069 0.69 0.70
0.64 0.63 0 67
0.95 0.95 0.95
1.00 1.00 1.00
* Highest RFG Area
Lowest RFG Area
-* Retail Avg.
x Production Avg.
Per Gallon Std
Figure 2
Comparison of Summer 1999 and Summer 2000 RFG Benzene Distributions
0.6 0.8
Benzene (vol%)
Figure 3
The RFG toxics performance standard applies to all RFG and is met on an annual average basis.
However, there was little change in benzene content between Winter 1999 and Winter 2000 RFG, and pre-
2000 Winter benzene content levels are comparable to those in 2000 and beyond. A possible explanation
for this lack of change in Winter benzene content is that benzene reduction is more valuable as toxics
emission reduction strategy when producing Summer RFG, since both exhaust and non-exhaust benzene
emissions are considered in Summer toxics performance. Another hypothetical "regulatory" explanation is
that, if benzene content changes were largely a collateral effect of refining and blending changes made to
meet Phase II VOC and NOx standards, Winter RFG benzene content may not have been affected. VOC
standards do not apply to Winter RFG and the Winter NOx standard did not become more stringent.
Comparison of Figures 1 and 2 shows that for Phase II RFG, Winter benzene levels have been
consistently higher than Summer levels, while prior to 2000, Summer levels have generally been higher.
The survey standard is a 1 percent annual average and, while in some years an RFG area may have
approached or even exceeded this level on a seasonal basis, there have been no survey failures.
91
-------
Both the survey-based and reporting estimates of average benzene concentration indicate
increases between 2000 and 2005. After 2000, both summer and winter survey-based estimates have
been higher than corresponding reporting estimates. It is not apparent why retail benzene concentrations
would be higher than production benzene concentrations. This divergence may be the result of statistical
or other errors inherent in the use of the geographic survey data to estimate overall averages.
RFC Benzene by PADD
Table 1 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These
averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information):
2004 & 2005
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
III
Reporting Averages by PADD-RFG Benzene (Refiner Batches Only)
2004
Average Value
(vol %)
0.55
0.84
0.54
0.65
0.79
0.54
0.61
0.81
0.54
*Volumes exclude batches
Gasoline Volume
(gal)*
4,787,247,511
1,740,499,436
5,890,920,167
6,488,981,913
2,437,732,174
6,059,647,031
11,276,229,424
4,178,231,610
11,950,567,198
with missing values
2005
Average Value
(vol %)
0.62
0.86
0.60
0.69
0.81
0.57
0.66
0.83
0.59
for this parameter
Gasoline Volume
(gal)*
4,430,366,230
1,844,913,833
5,690,766,967
7,081,608,319
2,718,751,541
5,771,776,973
11,511,974,549
4,563,665,374
11,462,543,940
Table 1
Figures 4 and 5 show estimates of average levels by PADD in retail RFC. These averages are
volume-weighted averages of the seasonal averages for each area, using gasoline volume estimates
supplied in the survey plans. In 2005, RFC surveys were conducted in 18 PADD I areas, five PADD II
areas and two PADD III areas. Survey averages show consistently higher benzene levels in PADD II, as
do the reporting averages for the two years analyzed.
Average Benzene Content or Summer RFG Sold at Retail Statlons-By PADD
I
Figure 4
92
-------
Benzene (vol%)
Average Benzene Content of Winter RFG Sold at Retail Stations-By PADD
0 60 -
0.40 -
0.20 -
DPADD1
PADD2
D PADD3
"I" Bars denote range of RFG area averages
1
1995
0.54
0.76
063
I
I
1996
064
082
076
1997
063
0.87
067
I
i
1998
0.60
0.83
0.69
1
I
1999
0.62
O.B4
0.73
1
I
2000
0.62
0.76
0.66
2001
0.61
080
066
i
T
2002
063
0.74
066
I
2003
0.67
0.7B
0.66
-1
2004
069
0.77
057
fj
2005
0.69
0.79
0.62
Year
Figure 5
Additional Analysis and Observations-RFC
Data analyses pertaining to RFG benzene are contained in the Appendix to this chapter.
These include estimates of grade specific averages which show that average benzene content in premium
grade RFG has been consistently lower than in regular RFG.
Overview-Conventional Gasoline
Volume-weighted average CG benzene content levels, by reporting year, are shown in Figure 6.
Summer and Winter averages are shown separately on the same graph. Based on inspection, these
averages do not suggest a clear trend in either the Summer or Winter data, although the 2005 averages
were the highest for the 1997-2005 period. Summer CG benzene content levels have been higher than
winter CG benzene content levels since 1998. The direction of year to year change in Summer and Winter
benzene content levels has been the same between 1999 and 2003, suggesting that the same factors might
be influencing the year to year changes in both.
It is unclear whether regulatory changes have had a substantial influence on year to year changes
in CG benzene content. A change in CG toxics emission requirements occurred between 1997 and 1998
with the transition to a Complex Model-based standard. CG was required to comply with the MSAT toxics
requirements beginning in 2002. While reductions in average benzene content occurred concurrently with
these changes, neither of these regulatory changes was explicitly intended to require a sharp reduction in
toxics emissions. There was little change in CG benzene content between 1999 and 2000 even though RFG
benzene declined, consistent with the intent of the Anti-Dumping regulations. Given the relatively small
changes in benzene content at these transitions and the observed fluctuations in benzene content, it would
be inappropriate, without further evidence and analysis, to ascribe any cause-effect relationship between
these regulatory changes and changes in CG benzene content.
93
-------
Average Benzene Content of Conventional Gasoline
(from Batch Reports)
Winter Benzene
A- Annual Avg
1.14
Reporting Year
Figure 6
CG Benzene by PADD
Table 2 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These averages
were calculated from batch data, excluding importer batches (see "PADD Level Analysis" appendix for
additional information):
2004 & 2005 Reporting Averages by PADD-CG Benzene (Refiner Batches Only)
Season
Summer
Winter
Annual
2004
PADD Average Value
I
II
III
I
II
III
I
II
III
(vol %)
0.96
1.38
0.98
0.93
1.32
0.89
0.95
1.35
0.93
^Volumes exclude batches
Gasoline
Volume (gal)*
3,721,549,054
11,327,130,611
21,712,599,435
3,451,551,536
13,133,321,322
23,618,108,969
7,173,100,590
24,460,451,933
45,330,708,404
with missing values
2005
Average Value
(vol %)
1.12
1.41
1.07
1.18
1.25
1.01
1.15
1.32
1.04
for this parameter
Gasoline
Volume (gal)*
3,517,507,894
10,437,087,141
20,884,656,303
4,025,985,850
13,416,981,375
22,432,331,959
7,543,493,744
23,854,068,516
43,316,988,262
Table 2
Additional Analyses and Observations-CG
The benzene content of Summer baseline gasoline, as specified in the Clean Air Act, is 1.53
volume percent, and the benzene content of Winter baseline gasoline, as specified by EPA's regulations, is
1.64 volume percent.
Data analyses pertaining to CG benzene are contained in the Appendix to this chapter. These data include
tabular and graphical descriptions of CG benzene content by volume and grade which show:
94
-------
» In 1997, the first year for which benzene reporting data were analyzed, the median Summer benzene
content was 0.95 volume percent and the 1990 baseline gasoline fell between the 70th and 75th
percent! le.
» In 2005, the last year for which benzene reporting data were analyzed, the median Summer benzene
content was 1.06 volume percent and the 1990 baseline gasoline fell between the 70th and 75th
percent! le.
» In 1997 the median Winter benzene content was 0.88 volume percent and the 1990 baseline gasoline
fell between the 75th and 80th percentile.
» In 2005 the median Winter benzene content was 0.98 volume percent and the 1990 baseline gasoline
fell between the 80th and 85th percentile.
» The benzene content of premium grade CG has been lower than that of regular grade in each year and
season.
All
Although this report has analyzed benzene trends separately by season and gasoline type,
benzene is regulated on an annual average basis for both RFC and CG. Since the MSAT2 regulations will
eventually supersede both the RFC and Anti-Dumping benzene and toxics requirements, some readers
may also be interested in analyses combining RFC and CG data. Aggregated reporting data averages are
available, in tabular form, in the "Summary of Average Estimates" chapter of this report. Aggregated
benzene content distributions by gasoline volume for 2005 are shown in Table 3, below:
2005 Benzene Content (vol %) by Volume
(from Batch Reports excluding CG blendstocks)
Volume %
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
CG
0.00
0.36
0.47
0.53
0.60
0.66
0.73
0.79
0.86
0.94
1.01
1.10
1.20
1.29
1.39
1.49
1.61
1.77
2.01
2.43
5.20
90,872,058,197
0.01
0.30
0.37
0.41
0.45
0.48
0.51
0.54
0.58
0.61
0.64
0.67
0.70
0.74
0.78
0.82
0.86
0.91
0.98
1.07
1.68
32,132,421,234
ALL
0.00
0.33
0.42
0.48
0.54
0.59
0.64
0.69
0.74
0.80
0.86
0.92
1.00
1.09
1.19
1.32
1.44
1.60
1.84
2.25
5.20
123,004,479,431
Table 3
95
-------
Aromatics are a family of hydrocarbon compounds with chemical properties similar to benzene.
In addition to benzene, other aromatic compounds such as toluene and xylenes are found in gasoline, and
aromatics other than benzene constitute the bulk of the aromatics volume in most gasoline. The main
source of aromatics in gasoline is reformate, a blending component produced in a refinery process units
called reformers. Reformate is a high octane, low sulfur blending component. While these are desirable
characteristics, reformate, with its high aromatics content, also has some undesirable emissions
characteristics. The Complex Model indicates that certain toxics, NOx and exhaust VOC increase with
increasing gasoline aromatics content.
Benzene and other aromatics are petrochemicals which are used for various purposes such as in
the manufacture of plastics and as solvents. Aromatics produced at refineries are sometimes extracted
and marketed as petrochemicals. Thus, the demand for these aromatics as petrochemicals will, to an
extent, affect the aromatics content of gasoline.
The primary regulatory constraints on gasoline aromatics content have been and currently are
indirect limits resulting from model-based emission standards. Prior to 1998, RFC was subject to a toxics
emission performance standard based on the Simple Model. According to this model, exhaust benzene
emissions, one of several toxic emissions incorporated in this standard, increase with gasoline aromatics
content as well as gasoline benzene content. (While benzene is an aromatic compound, benzene content
is also considered as a separate gasoline parameter and additional regulations address gasoline benzene
content. See the Benzene Chapter in this report.) CG supplied from 1995 through 1997 was required to
have annual average exhaust benzene emissions, as calculated by a formula considering both aromatics
and benzene content, no greater than the refiner's or importer's compliance baseline for exhaust benzene.
In 1998 and subsequent years standards based on the Complex Model applied to both RFC and
CG. Complex model RFC is subject to toxics, NOx and VOC emission performance standards. CG
exhaust toxics and NOx emissions cannot exceed the refiner's or importer's compliance baseline for these
emissions. Gasoline aromatics are generally associated with toxics emissions, and gasoline toxics emission
requirements, including the current MSAT requirements, are expected to have a limiting effect on
aromatics content. The Complex Model predicts that aromatics reduction lowers exhaust benzene
emissions. Although it predicts that emissions of certain other toxics may increase slightly with reduced
aromatics, the exhaust benzene emission reduction more than offsets these increases, resulting in a net
toxics emission reduction with an aromatics content reduction.
Toxics emission standards may not have been the only standards limiting aromatics content
Analysis presented in the RFC Trends chapter suggests that aromatics reductions may have helped
significantly to meet the more stringent Phase II RFC Summer NOx performance standard, as well as to
provide some additional VOC reduction. While NOx emission requirements may have, in the past, had
some limiting effect on the aromatics content of RFC and CG, the Tier 2 sulfur reductions, with consequent
substantial NOx emission reductions and over-compliance for both, clearly reduced the limiting effect, if
any, of NOx emission requirements on aromatics content. 37
37 As noted elsewhere, NOx emission requirements were generally eliminated in 2007 and toxics emission
requirements will generally be eliminated in 2011, removing most of the regulatory constraints on aromatics content.
96
-------
Complex Model acceptable range limits directly limit gasoline aromatics content. The acceptable
range limit for RFC is 50 volume percent and the acceptable range limit for CG is 55 volume percent. For
RFC, this range limit is applicable to each batch. For CG, the annual average aromatics level cannot
exceed 55 volume percent or the refiner's or importer's 1990 baseline for aromatics, whichever is greater
EPA's Aromatics Data
EPA collects aromatics data in order to determine compliance with the RFC and Anti-Dumping
regulations. Refiners and importers have been required, since 1995, to submit aromatics content data for
each batch of CG and RFC refined or imported. Additionally, since 1995, EPA has received RFC Survey
data on the aromatics content of RFC sold at retail outlets.
Aromatics Trends
Overview-Reformulated Gasoline
Analysis of aromatics data from EPA's two data sources shows that a decrease in average
aromatics content occurred with the transition from Phase I to Phase II standards. Decreases in aromatics
content for both Summer and Winter RFC are consistent with a more stringent Phase II annual average
toxics emission performance standard. However, estimates from these two sources have been in better
agreement for most other properties, particularly in quantifying Phase I to Phase II changes.
Consequently, there is additional uncertainty in any conclusions regarding aromatics trends. While it is
reasonable to expect some difference in estimates from the two data sources the estimates appear to
differ in a systematic fashion.
sS
3
1
1
Highest RFG Area
Lowest RF6 Area
Retail Avg
Production Avg
+
*
1995
29.3
19.B
24.3
f
*
*
1996
26.5
20.8
24.8
verac
*
_±
1997
27.9
22.4
25.5
224
je Sui
*
*
199E!
28.1
23.B
26.0
22. B
nmer
*
K
**S
1999
272
214
24. eH
22.1
RFG
\:
T"""
i
2000
24.9
18.0
19.5
19.3
Year
Urom
*
2001
24.0
10.7
202
20 1
atics I
*
4
m
2002
23.8
178
205
204
:onte
»
2003
23.5
14.5
261
20.1
nt
-t^
i
2004
25.1
15.9
21 2
20 1
=4
i
2005
25.1
15.4
21.1
20.9
» Highest RFG Area
Lowest RFG Area
-*- Retail Avg.
:.: Prrirlurtinri Avg.
Figure 1
97
-------
f
1
1
o
<
20 -
Highest RFG Area
Lowest RFG Area
Retail Avg.
Production Avg.
*-"~"
i
1995
25.7
17.4
20.0
19
2:
11-
2C
96
1
2
.7
Av
19
22
1E
2D
18
era
K
97
7
2
.8
2
geW
*
t
1999
23.4
17.6
21.1
19.9
nterl
*-
-K
i
1999
22.7
18.4
21.2
185
;FG A
"*
*
2000
20.3
14.1
10.1
190
Year
roma
*
=JJP=
2001
21.4
15.0
18.6
192
tiesC
*
2002
220
16.0
19.3
19.4
onten
i
2003
21 7
14.8
19.1
19.4
t
*
2004
227
14.6
19.4
19 1
*
==₯
2005
222
13.2
18.8
196
Highest RFG Area
Lowest RFG Area
-*- Retail Avg.
-^r- Production Avg
Figure 2
Figure 1 shows Summer averages calculated from reporting system and survey data. Both data
sets show a decrease in aromatics content from 1999 to 2000 greater than any of the other year to year
changes, although the production data indicate a 2.9 volume percent reduction in aromatics content, while
the survey data indicate a 5.4 volume percent reduction. While the survey-based and reporting system-
based averages are in close agreement in 2000 and subsequent years, the survey-based estimates were
higher than reporting system estimates in 1999 and in the two preceding years.
Figure 2 shows Winter averages calculated from the two data sets. While the survey data show a
1999 to 2000 decrease in aromatics content of 3.1 volume percent, clearly greater than any of the other
year-to-year changes, the production data show only a 0.5 volume percent decrease in aromatics content
between 1999 and 2000. As with the Summer estimates, survey-based average estimates were higher
than reporting system estimates in 1999 and in the two preceding years.
Although estimates from the two data sources clearly differ, there is little doubt that an atypical
change in Summer RFG aromatics content occurred concurrent with the transition from Phase I to Phase II
standards. For both data sets, the 1999 to 2000 decrease in aromatics is, subjectively, large compared to
other year-to- year changes. More objectively, regression analysis of both sets of estimates indicates that
a statistically significant shift in aromatics content occurred (see Regression Analysis Appendix). The
timing and magnitude of the decreases do not, in themselves, establish a cause-effect relationship, but
toxics, NOx and VOC Summer RFG emission performance standards all became more stringent, and
aromatics reductions can improve emission performance for all of these pollutants. Analysis presented in
the RFG chapter, based on survey data, confirms that Summer RFG emission performance improved
between 1999 and 2000, and identifies aromatics content reduction as the primary factor for toxics
performance improvement, and as the secondary factor for NOx and VOC improvement.
While estimates from both data sets indicate a reduction in Winter RFG aromatics content between
1999 and 2000, only the survey-based estimates indicate that a statistically significant shift in aromatics
content occurred concurrent with the transition to Phase II standards. Average toxics performance
98
-------
during Phase I, as estimated from both data sets substantially over-complied with the Phase II toxics
performance standard. Thus, while some refineries and importers may have needed to reduce aromatics
and improve winter toxics performance to meet the Phase II toxics standard (EPA's aggregate analysis did
not address this), the more stringent Phase II toxics standard may not have had a large impact on average
Winter RFC aromatics content. The Winter NOx performance standard did not become more stringent
with Phase II, and average Phase I NOx performance substantially over-complied with the Phase II
standard, so it is unlikely that this standard would have caused a change in Winter RFC aromatics content.
Consequently, evidence supporting a cause-effect relationship between this transition and an aromatics
content change is more equivocal.
The General Methodology Chapter explains why EPA analyzed both RFC survey and reporting data
to estimate overall RFC property trends even though survey-based trend estimation required more
approximations and assumptions. The analysis presented here suggests that, for aromatics, there were
some systematic differences between the two estimates prior to 2000, with good agreement in 2000 and
subsequent years. Consequently, the two data sets differ somewhat in their estimate of the magnitude of
the change in aromatics content between Phase I and Phase II RFC. Arguably, there is some rationale for
rejecting survey-based trend estimates in favor of reporting-based estimates if there is any substantial
disparity. However, EPA has not identified a clear problem with either data set and believes that both sets
of aromatics estimates must be considered. There are several reasons for this view:
It appears unlikely that the differences between the survey-based and reporting system estimates
of average aromatics content are solely attributable to some combination of statistical sampling error and
imprecise sales volume estimates. For Summer RFC, the most notable difference between the two
estimates is that the reporting system averages in 1997,1998 and 1999 fell close to or below the lowest
seasonal average in any of the areas sampled. Thus, even extremely inaccurate estimates of sales
volume, by themselves, would not explain the difference between the two averages. The mandatory RFC
areas, which, in general, would be heavily weighted in the average calculation because of their large sales
volumes were also the most frequently sampled in each survey year with multiple one week surveys during
the June 1-September 15 Summer period, (e.g. In 1998 each of the non-California mandatory areas were
surveyed between six and eight time during this period). Thus, the areas that most heavily influenced the
volume-weighted average calculation would also likely have the most accurate seasonal area average
estimates. For Winter RFC, the most notable difference is that the reporting system estimate showed
only a 0.5 volume percent decrease in aromatics content between 1999 and 2000, while the survey-based
estimate indicates a 3.1 volume percent decrease. Area-specific seasonal survey averages for both
Summer and Winter RFC show that aromatics content decreased in each of the 23 areas surveyed in both
1999 and 2000. For Winter RFC, the area-specific survey average decreases ranged from about 1.2 to 5.9
volume percent. Since these paired area data from the surveys strongly indicate a decrease in Winter
aromatics between 1999 and 2000, it is difficult to dismiss the survey data entirely, and conclude with a
high degree of certainty that there was little change in Winter aromatics between 1999 and 2000.38
Although the reporting data are not a statistical sample, the reporting system averages are also
subject to error. A number of different parties of varying capabilities collect and analyze reporting system
samples and submit reports to EPA. Although EPA's regulations include mechanisms to help ensure the
quality of the data, such as record keeping and independent laboratory sampling and analysis
requirements, these mechanisms cannot totally eliminate reporting data errors. In addition to errors in the
sense of mistakes such as incorrect reporting of data or omitting a data item, there are precision and
accuracy issues associated with the measurement of any gasoline parameter (i.e., even when a gasoline
38 It should be noted that sample sizes in individual one-week surveys are smaller in Winter than in Summer,
and in some areas fewer surveys are scheduled during the Winter season. This may result in less precision and
accuracy in Winter than Summer seasonal average estimates based on survey data.
99
-------
sample is carefully analyzed using the correct methods and procedures measured property values can
differ from the "true" property values). A detailed discussion of these issues is beyond the scope of this
report, however it is worth noting that there are some additional aromatic-specific factors that could affect
the quality of the aromatics data from both data sets. EPA's regulations have allowed two methods for
measurement of gasoline aromatics content; Gas Chromatography/Mass Spectrometry (GC-MS) designated
as ASTM standard method D 5769, and Flourescent Indicator Adsorption (FIA), designated as ASTM
standard method D 1319.39
EPA's regulations designated GC-MS as the method for determining aromatics content, but allowed
use of D 1319 if the result is correlated with the designated method. In its "RFC Questions and Answers"
guidance document (EPA 2003) EPA has characterized D 1319 as "highly operator-dependent" and having
"relatively large reproducibility". "Reprodudbility" is a term referring to the difference in test results from
two different operators measuring the same material at two different laboratories or using two different
instruments. Survey samples were analyzed for aromatics using D 1319, but in each year all or most of
these analyses were done at a single laboratory (although not necessarily with the same operator and
instrument), and this laboratory was used since the inception of the RFC Survey program. A larger
number of laboratories analyzed reporting system samples, and the data do not identify which of these
two methods was used to determine the aromatics content of each batch. (Consequently, it would be
difficult to determine if measurement-related issues are important in explaining the difference between
reporting system and survey aromatics estimates.)
Additionally, there may be actual differences between the average properties of gasoline sold at
retail and the average of properties in the gasoline and blendstock that refiners and importers consider for
compliance determination. (These differences can occur as a result of allowable practices such as
downgrading of RFC to CG, downgrading of Summer RFC to Winter RFC, or blending of oxygenates in
amounts greater than assumed for compliance reporting.) Differences may be even more apparent when,
as in this report, gasoline properties are analyzed and compared on a seasonal basis, particularly when
there are large Summer to Winter variations in a property. EPA does not know if real differences between
retail and reporting properties were a major cause for differences in aromatics estimates. However, since
the true retail averages could differ from the true reporting averages, and since both are of interest, it is
generally appropriate to consider RFC data from both sources in trend analyses.
39 From time to time ASTM revises its standards and the complete standard number includes a suffix
indicating the revision (e.g. D1319-02a). EPA regulations specifying test methods (40 CFR 80.46) reference specific
versions of these test methods. Although EPA from time to time updates these regulations to reflect revisions to
ASTM standards, these regulations do not always specify the "active" standard (see http://www.astm.org).
100
-------
Average Aromatics Content by Production Quarter
(from Batch Reports-CG Blendstocks Excluded)
2|3 4
00
Figure 3
Regardless of which data estimates are used, comparison of Figures 1 and 2 shows that aromatics
levels in Summer RFC are generally higher than aromatics levels in Winter RFC. Gasoline composition
varies seasonally because of regulatory emission requirements, as well as vehicle performance
("drivability") requirements associated with gasoline's volatility and distillation characteristics. The need to
meet these seasonal requirements has resulted in a clear seasonal variation in aromatics content for both
RFC and CG. This variation can be seen in Figure 3, which shows average RFC and CG aromatics content,
estimated from reporting system data, by production quarter. Although this yearly cyclical variation in RFC
aromatics diminished with the transition to the Phase II RFC standards in 2000, it is still apparent.
RFC Aromatics by PADD
Table 1 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These
averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information):
2004 & 2005 Reporting Averages by PADD-RFG Aromatics (Refiner Batches Only)
Season
Summer
Winter
Annual
2004
PADD Average Value
I
II
III
I
II
III
I
II
III
(vol %)
21.2
18.9
18.7
20.4
17.3
17.9
20.7
18.0
18.3
^Volumes exclude batches
Gasoline Volume
(gal)*
4,792,114,891
1,740,499,436
5,890,920,167
6,489,530,475
2,438,403,544
6,059,647,031
11,281,645,366
4,178,902,980
11,950,567,198
with missing values
2005
Average Value
(vol %)
21.9
19.1
19.2
20.7
18.4
18.0
21.2
18.7
18.6
for this parameter
Gasoline Volume
(gal)*
4,431,080,230
1,844,913,833
5,690,766,967
7,081,940,665
2,718,751,541
5,766,524,033
11,513,020,895
4,563,665,374
11,457,291,000
Table 1
101
-------
Figures 4 and 5 show estimates of average levels by PADD in retail RFG. These averages are volume-
weighted averages of the seasonal averages for each area, using gasoline volume estimates supplied in
the survey plans. In 2005, RFG surveys were conducted in 18 PADD I areas, five PADD II areas and two
PADD III areas. For Phase II RFG, survey data generally showed the highest aromatics concentrations in
PADD I RFG, as did reporting data for the two years analyzed.
Average Aromatics Content of Summer RFG Sold at Retail Stations-By PADD
1995 1996 1997 1990 1999 2000 2001 2002 2003 2004 2005
Figure 4
omatics (vol%)
<
ft
30 -r
25 -
20 -
15 -
10 -
5 -
0 -
DPADD1
PADD2
n PADD3
verage Aromatics Content of Winter RFC Sold at Retail Stations-By PADD
"I" Ears denote range ot RFG area averages
rn
ri-
1995
20.0
19.0
21 0
ft
1896
20.8
19.4
21.3
rt
n
1997
21
13.9
20.8
i
$
1998
21.6
18.9
21.2
f
|
i
1999
21.5
195
21 9
f
I
r
2000
18.9
15.9
18.0
*
2001
19.0
16.7
18.6
f\
|J|
2002
19.8
17.0
200
fl
rh
2003
19.8
17.2
18.8
*
1
ft
:
2004
20.2
16.3
19.5
I
f-
2005
19.8
18.6
18.4
Year
Figure 5
102
-------
Data analyses pertaining to RFC aromatics content are contained in the Appendix to this chapter. Several
trends or patterns are highlighted below:
» Average aromatics content of premium grade RFC has been higher than aromatics content in regular
or (from survey data) mid-grade RFC in each year and season40
» Analysis of retail data shows that aromatics content decreased from 1999 to 2000 for each grade-
season combination
Figure 6 shows Summer and Winter average CG aromatics levels, by reporting year. Summer and
Winter average aromatics levels peaked in 2000 and 2001 respectively. The largest year-to-year change in
Summer aromatics content occurred between 1999 and 2000, when average aromatics increased by more
than 0.8 volume percent, and the largest Winter change occurred between 2000 and 2001, when average
aromatics increased by more than 0.5 volume percent. Notably, although Summer averages decreased in
each succeeding year through 2003, averages in 2000 and subsequent years were higher than averages in
1997 through 1999. Regression analysis also indicates that a shift in Summer CG aromatics content
occurred between 1999 and 2000. This sustained Summer CG aromatics increase concurrent with a
sustained Summer RFC Phase I to Phase II aromatics decrease suggests, but does not establish that the
transition to Phase II RFC influenced CG aromatics content. Figure 3, which also shows CG aromatics
averages by production quarter, illustrates that average aromatic levels have been consistently highest
during the 2nd quarter of production, where the majority of production volume is Summer gasoline, and
these 2nd quarter averages in 2000 and succeeding years were higher than in each year prior to 2000.
40 The RFG Survey sampling plan estimates the grade market share in each area to ensure that grades are
properly represented in each survey. Errors in estimating the grade mix may contribute to inaccuracies in individual
survey estimates of property averages and, in turn, contribute to inaccuracies in survey-based estimates of overall
average property values. EPA does not believe that incorrect estimates of grade mix are the major cause of
differences between survey-based and reporting system based aromatics estimates, since the same directional
differences are apparent when 1997 through 1999 regular and premium grade survey averages and reporting
averages are compared on a grade by grade basis (see appendix).
103
-------
Average Aromatics Content of Conventional Gasoline
(from Batch Reports)
27 -
w
* summer
winter
* Annual
^
+
»
^ *
- "^" = .
.
1997
27.4
25.9
26 1
1998
27.5
24.8
26.8
1999
27.6
258
26.1
2888
2G.4
248
264
2601
28.3
25.3
26.6
Year
2002
28.1
25.0
26.4
2083
27.9
249
26.3
2804
28.1
24.6
26.2
2005
27.8
247
26.1
Figure 6
Although EPA has no CG data from a second source comparable to the RFC Surveys, the disparity
between RFC Survey and reporting system-based estimates of aromatics raises the level of uncertainty in
conclusions about CG trends based on reporting system data. Since EPA cannot explain why these
differences occurred, and since the differences appear to be somewhat systematic, EPA cannot rule out
the possibility that this is indicative of an issue with the reporting system data that may affect both RFC
and CG.
CG Aromatics by PADD
Table 2 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These
averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information):
2004 & 2005 Reporting Averages by PADD-CG Aromatics (Refiner Batches Only)
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
III
2004
Average Value
(vol %)
28.4
28.7
27.3
24.0
25.1
24.1
26.2
26.8
25.6
Gasoline Volume
(gal)*
3,721,549,054
11,333,095,661
21,712,599,435
3,455,187,610
13,146,020,260
23,618,763,665
7,176,736,664
24,479,115,921
45,331,363,100
^Volumes exclude batches with missing values
2005
Average Value
(vol %)
28.9
28.6
27.0
24.6
24.5
24.5
26.6
26.3
25.7
for this parameter
Gasoline Volume
(gal)*
3,520,575,616
10,443,353,793
20,882,654,793
4,011,434,950
13,428,549,183
22,432,331,959
7,532,010,566
23,871,902,976
43,314,986,752
Table 2
104
-------
The aromatic content of Summer Baseline Gasoline, as specified in the Clean Air Act, is 32.0 volume
percent, and the aromatic content of Winter baseline gasoline, as specified by EPA's regulations, is 26.4
volume percent. Data analyses pertaining to CG aromatics are contained in the Appendix to this chapter.
These data include tabular and graphical descriptions of CG aromatics content by volume and grade which
show:
» In 1997, the first year for which aromatics reporting data were analyzed, the median Summer
aromatics content was 26.9 volume percent and the 1990 baseline gasoline fell between the 75th and
80th percentile.
» In 2005, the last year for which aromatics reporting data were analyzed, the median Summer
aromatics content was 27.4 volume percent and the 1990 baseline gasoline fell between the 75th and
80th percentile.
» In 1997 the median Winter aromatics content was 24.3 volume percent and the 1990 baseline gasoline
fell between the 60th and 65th percentile.
* In 2005 the median Winter aromatics content was 24.2 volume percent and the 1990 baseline gasoline
fell between the 65th and 70th percentile.
» The aromatics content of premium grade CG has been higher than that of regular grade in each year
and season.
105
-------
Olefins are a class of hydrocarbons found in gasoline. Olefins content, in volume percent, is an
input parameter for the Complex Model. The gasoline blending component from the Fluid Catalytic
Cracking (FCC) unit often has a high olefin concentration. Olefins are desirable gasoline constituents
because of their octane characteristics but, according to the Complex Model, NOx and exhaust toxics
emissions increase with olefin content. The model predicts that exhaust VOCs will decrease as olefin
content increases.
EPA's RFC and Anti-Dumping regulations have imposed both direct and indirect limits on gasoline
olefin content. The Simple Model, applicable to RFC through 1997, did not consider the effect of olefins
on emissions. However, the annual average olefin content was not allowed to exceed a refinery's or
importer's 1990 olefins baseline level. Currently, the olefin content of RFC is directly limited by a 25.0
volume percent maximum Complex Model range limit intended to ensure the validity of the model's
emission performance estimates. The olefin content of RFC and CG was indirectly limited through
Complex Model-based limits on RFC and CG emissions since Complex Model NOx and exhaust toxics
increase with increasing gasoline olefin content. Regulatory changes eliminating these emission limits
have or will remove these indirect constraints.
Prior to 1998, the annual average olefins level in CG was not allowed to exceed 125% of a
refinery's or importer's 1990 baseline. For 1998 and later, CG was required to have annual average
Complex Model emissions of exhaust toxics and NOx that did not exceed the refiner or importer's 1990
baselines for these emissions. Additionally, annual average CG olefins levels were not allowed to exceed
the greater of the Complex Model valid range limit for CG olefins (30.0 volume %) or the refiner or
importer's 1990 olefins baseline.
Olefins data for both RFC and CG have been and are currently necessary to determine compliance
with EPA's regulations. Refiners and importers have been required, since 1995, to submit this information
to EPA's RFG/Anti-Dumping reporting system for each batch of gasoline refined or imported. Additionally,
since 1998, EPA has received RFC Survey data on the olefins content of RFC sold at gasoline stations.
(Since olefins content was not part of the Simple Model, it was not needed to determine compliance with
survey standards until 1998).
Figures 1 and 2 show Summer and Winter volume-weighted average RFC olefins, by year.
Although the reporting-based production and survey-based retail trendlines are somewhat different, EPA
has not identified any pervasive problem with either set of olefins data. As with the aromatics estimates,
EPA believes that estimates from both sets of data must be considered unless a clear problem can be
identified in one or the other data sets.
106
-------
^~
_3
^>_
C
O
HighestRFGArea
Lowest RFG Area
Retail Avg
Production Avg
*5£
1997
12.0
Av
.
"--^_
* '
1
1998
12.8
6.6
10.3
10.9
erage
*
~^jt~~-
1999
14.2
6.8
10.8
11.4
Summ
~~*^
J
2000
11.6
4.0
9.4
10.6
erRFG
*
^*^.
***
i
2001
12.7
5.3
10.3
11.8
Year
Oiefin
2002
140
58
10.8
108
Conte
»
.
2003
143
34
10.9
11.0
nt
*
. * '
i,
2004
13.2
27
109
11.3
f
. *
*
*
2005
13.6
3.5
10.8
11.9
» Highest RFG Area
Lowest RFG Area
-A-Retail Avg.
-w Production Avg.
Figure 1
Average Winter RFG Oiefin Content
14 -
12 -
I
c
» a -
5
4 -
HighestRFGArea
Lowest RFG Area
Retail Avg
Production Avg
f
* T ^ 1
x^ *-* ""^ """"-^v
* -
i
*
-A-
i
-4
i
1997 1998 1999 2000 2001 2002 2003
12.0 13.8 12.8 13.9 13.3 13.9
5.8 6.0 5.3 6.0 6.2 5.7
94 10.0 10.0 10.7 10.6 108
11.3 10.8 11.3 11.8 12.3 11.1 11.0
Year
2004
15.1
5.1
10.5
11.1
2005
128
4.7
95
11.0
« HighestRFGArea
Lowest RFG Area
-*- Retail Avg
x Production Avg.
Figure 2
The reporting system data show decreases in average olefin content between 1997 and 1998 for
both Summer and Winter RFG. The 1997 Summer average is higher than in any of the succeeding years
while the 1997 Winter average is not. Although EPA analyzed only a single year of Simple Model RFG
olefins data, there is some basis for concluding that these decreases were in response to EPA's Complex
Model requirements. Close to four percent of both Summer and Winter RFG supplied in 1997 would not
have meet 1998 requirements because it did not meet the Complex Model's 25 volume percent olefin
content range limit restriction, applicable to each RFG batch. (The graphical and tabular data in the
appendix to this chapter show shifts in both "RFG olefins by gasoline volume" distributions between 1997
and 1998.)
Additionally, the Complex Model recognized the emission benefits of olefin reduction, and some
107
-------
suppliers may have reduced olefins in order to meet 1998 NOx and toxics performance standards. The
transition from Simple Model to Complex Model standards affected sulfur content in a similar but much
more substantial manner. (While Summer and Winter olefins content decreased by about 11% and 5%,
sulfur content decreased by about 30% and 19%, respectively.) In addition to the parallel regulatory
factors cited above, olefin reductions may also have occurred as a result of refining and blending processes
used to reduce sulfur content. As noted in the Sulfur Chapter, refineries have, in the past, largely relied on
a process called hydro treating to reduce sulfur, when necessary. In addition to lowering sulfur content,
hydro treating may also have reduced olefins.
The survey-based estimates of Summer olefin content suggest that a downward shift occurred
between 1999 and 2000 with the transition from Phase I to Phase II, superimposed on an upward trend.
Although the reporting data also show a decrease in olefin content between 1999 and 2000, it is not
distinguishable from other year to-year changes. By contrast, EPA concludes that, for Summer RFC, both
data sets clearly show all other Complex Model parameters, except oxygen, shifted between 1999 and
2000. Neither of the Winter estimates suggest that olefin content changed in response to this transition.
Analysis based on RFC Survey property averages presented in the RFC Trends Chapter indicates
that, even for Summer RFC, the 1999 to 2000 olefin reduction provided only a small portion of the total
Phase I to Phase II RFC NOx and toxics emission reduction.
In summary, olefin content reductions apparently had a small role in meeting Phase II standards,
and conversely, the transition from Phase I to Phase II RFC probably had a limited effect on RFC olefins
content.
by
Table 1 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These averages
were calculated from batch data, excluding importer batches (see "PADD Level Analysis" appendix for
additional information):
2«
Season
Summer
Winter
Annual
M ft i
PADD
I
II
III
I
II
III
I
II
t
Average Value
(vol %)
13.0
5.3
11.2
12.9
4.9
11.4
12.9
5.0
III 11.3
^Volumes exclude batches
if Olef
Gasoline Volume
(gal)*
4,792,114,891
1,740,499,436
5,890,920,167
6,489,530,475
2,438,403,544
6,059,647,031
11,281,645,366
4,178,902,980
11,950,567,198
with missing values
ns
Average Value
(vol %)
13.3
4.7
12.5
13.0
5.3
11.0
13.1
5.1
11.7
for this parameter
es
Gasoline
Volume (gal)*
4,431,080,230
1,844,913,833
5,690,766,967
7,081,940,665
2,718,751,541
5,766,524,033
11,513,020,895
4,563,665,374
11,457,291,000
Table 1
Figures 3 and 4 show estimates of average levels by PADD in retail RFC. These averages are volume-
weighted averages of the seasonal averages for each area, using gasoline volume estimates supplied in
108
-------
the survey plans. In 2005, RFC surveys were conducted in 18 PADD I areas, five PADD II areas and two
PADD III areas. The survey data show that olefins content is consistently lowest in PADD II RFC, as do
reporting data for the two years analyzed.
Average Olefln Content of Summer RFG Sold at Retail Stations-By PADD
160
12.0 -
8.0 -
6.0 -
40 -
2.0 -
0.0
IPADD2
rli
10.0
10.5
11.9
12.8
Year
Figure 3
Average Olefln Content of Winter RFG Sold at Retail Stations-By PADD
DPADD1
IPADD2
DPADD3
"I" Ears denote range of RFG area averages
102
6.3
1999
6.7
2000
6.1
9.4
2001
6.9
10.6
2002
7.0
10.6
200,1
115
2005
5.4
12.5
Year
Figure 4
Additional Analysis and Observations-RFC
Data analyses pertaining to RFG olefins are contained in the Appendix to this chapter. Grade
specific analyses show that average olefins content in premium grade RFG has been consistently lower
than in regular RFG.
109
-------
Overview-Conventional Gasoline
Figure 5 shows volume-weighted average CG olefins levels, by reporting year. These levels
decreased from 1997 to 1998, concurrent with the transition from annual average olefin content standards
to Complex Model-based annual average exhaust toxics and NOx emission standards. However, by 2001
both Summer and Winter olefin levels peaked above 1997 values, prior to decreasing through 2004. The
standard change described earlier probably represented an increase in stringency that could have
influenced the 1997 to 1998 olefin content decrease It is not readily apparent if, or how, the latter
movements in CG olefin averages relate to regulatory requirements.
Average Olefins Content of Conventional Gasoline
(from Batch Reports)
Reporting Year
Figure 5
CG Olefins by PADD
Table 2 shows 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These averages
were calculated from batch data, excluding importer batches (see "PADD Level Analysis" appendix for
additional information):
2004 & 2005 Reporting Average by PADD-CG Olefins (Refiner Batches Only)
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
III
2004
Average Value
(vol %)
12.83
8.82
12.20
15.66
8.60
12.18
14.19
8.70
12.19
*Volumes exclude batches
Gasoline Volume
(gal)*
3,721,549,054
11,333,095,661
21,712,599,435
3,455,187,610
13,155,255,287
23,618,763,665
7,176,736,664
24,488,350,948
45,331,363,100
with missing values
2005
Average Value
(vol %)
12.67
9.63
12.80
14.78
9.45
12.07
13.80
9.53
12.42
for this parameter
Gasoline Volume
(gal)*
3,520,575,616
10,443,213,355
20,882,654,793
4,035,235,552
13,434,249,988
22,432,331,959
7,555,811,168
23,877,463,343
43,314,986,752
Table 2
110
-------
The olefin content of Summer Baseline Gasoline, as specified in the Clean Air Act, is 9.2 volume
percent, and the olefin content of Winter baseline gasoline, as specified by EPA's regulations, is 11.9
volume percent. Data analyses pertaining to CG olefins are contained in the Appendix to this chapter.
These data include tabular and graphical descriptions of CG olefins content by volume and grade which
show:
In 1997, the first year for which olefins reporting data were analyzed, the median Summer olefin
content wa
percent! le.
In 2005, the last year for which olefins reporting data were analyzed the median Summer olefin
content wa
percent! le.
content was 11.8 volume percent and the 1990 baseline gasoline fell between the 30th and 35th
content was 11.4 volume percent and the 1990 baseline gasoline fell at approximately the 35th
In 1997, the median Winter olefin content was 11.7 volume percent and the 1990 baseline gasoline fell
between the 50th and 55th percentile.
In 2005, the median Winter olefin content was 10.8 volume percent and the 1990 baseline gasoline fell
between the 55th and 60th percentile.
The olefin content of premium grade CG has been lower than that of regular grade in each year and
season.
111
-------
Gasoline is a mixture of hydrocarbons which have different boiling points. Consequently, as
gasoline is heated, portions of its volume evaporate at different temperatures. The relationship between
temperature and percent of volume evaporated is referred to as a distillation curve or profile, and a test
procedure (ASTM D86) is used to determine the distillation profile of a gasoline sample. Gasoline volatility
characteristics are important because they affect both emissions and vehicle performance. The
relationship between EPA's regulatory requirements and gasoline distillation characteristics will be
discussed later in this chapter. In addition to these emission-related requirements, there are ASTM
specifications that place restrictions on distillation parameters for vehicle performance purposes. ASTM
standard D4814 defines specifications for six different Vapor Pressure/Distillation classes of gasoline which
include limits on T10, T50 and T90 (the temperatures at which 10%, 50% and 90% of the gasoline
evaporates). This standard also specifies which class of gasoline is required for each state and month of
the year. In 1998 a specification was added for the "drivability index" (DI), a formula
(1.5*T10+3.0*T50+1.0*T90) that combines these three distillation points into a number which, if too high,
indicates that the gasoline might cause performance problems. A 2006 revision of the standard, D4814-
06a, amended the formula with an upward adjustment for ethanol content
(1.5T10+3.0T50+1.0T90+2.4 °F*ethanol volume%).
Under the Simple Model standards which applied prior to 1998, the annual average T90 level for
RFC could not exceed the refinery's or importer's 1990 baseline level. The Simple Model, itself, did not
consider the effect of any distillation profile parameters on emissions. The Complex Model, however,
included two distillation parameters, E200 and E300. These parameters are, respectively, the percent of
gasoline evaporated at 200 degrees and 300 degrees Fahrenheit. There is a strong negative correlation
between T50 and E200 and between T90 and E300. These "E' parameters are determined from the same
distillation curve as the "T' parameters. E200 and E300, like other Complex Model parameters, have
acceptable range limits (30% to 70% for E200 and 70% to 100% for E300), which are applicable to each
batch of RFC. Since E200 and E300 are part of the Complex Model, RFC and CG emission requirements,
in theory, could constrain these parameters.
Simple Model CG standards limited the annual average T90 level to 125% of the refiner's or
importer's compliance baseline. For Complex Model CG, annual average E200 and E300 were limited to
the greater of the refiner's or importer's 1990 baseline or the Complex Model range limit. Since gasoline's
distillation characteristics depend on its composition, any changes to or limits on gasoline composition
necessary for compliance with emission requirements potentially affect or constrain these distillation
parameters. Changes to or trends in distillation parameters over time may indicate that such composition
changes occurred.
EPA collects data in order to determine compliance with RFC and Anti-Dumping requirements.
EPA needed T90 data through reporting year 1997, and E200/E300 data in 1998 and subsequent years.
EPA's RFG/Anti-Dumping batch reporting forms have provisions for reporting T50, T90, E200 and E300 for
each gasoline batch, and some parties chose to report distillation parameter information beyond that
needed to determine compliance with standards. EPA has restricted its analysis of batch data distillation
parameters to those that were required for compliance determination. RFC Survey data prior to 1998 did
112
-------
not include any distillation parameter information. For 1998 and later years, the data included not only
E200 and E300 but T50 and T90; consequently survey-based estimates for all four of these parameters will
be included for 1998 and subsequent years.
v. - ' - ds
', ".'.. '"
Figures 1 through 4 show volume-weighted E200 and E300 averages, by year. Both the reporting
system estimates, by reporting year, and the RFC Survey-based estimates, by survey year, are shown.
The reporting and survey based estimates both indicate that the largest year-to-year change in
average Summer E200 was a decrease between 1999 and 2000, concurrent with the transition from Phase
I to Phase II RFC. Regression analysis of both sets of estimates for years 1998 through 2004 indicates
that a statistically significant shift in E200 occurred at that time (see Regression Analysis Appendix).
There was little or no change in Winter E200 between 1999 and 2000.
The largest year-to-year changes in average Summer and Winter E300 were increases between
1999 and 2000, and regression analysis also indicates that statistically significant shifts in E300 occurred.
Additionally, according to both reporting and survey-based estimates, Summer and Winter E300 were
lower in 2005 than in 2000, although increases from 2004 to 2005 may indicate reversal of apparent
downward trends. Regression analysis indicates statistically significant downward linear trends through
2004 in the Summer survey-based estimates, and in both Winter estimates.
113
-------
CS- 48 -
A-J
Highest RFC Area
Lowest RFG Area
Retail Avg
Production Avg
Av
*
^ "
i
1898
so. a
47.1
49 4
48.8
;rage S
=k
1999
52 D
40. 0
498
49.2
jmmer 1
4
^
*
2000
50.3
45.1
47.8
47.7
iFG E2C
I
2001
5D.7
44.1
47.5
47.5
Ye
0 (% Ev
*
2002
50 6
45.0
47.7
47.5
ar
aporate
*
2003
51.7
45.1
480
47.9
0 at 200
f
\
2004
49.7
46 1
48.1
47.9
F)
*
Jl
i
2005
51.8
445
488
48.8
Highest RFG Area
Lowest RFG Area
-*- Retail Avg
-*- Production Avg
Average Winter RFG E200 (% Evaporated at 200 F)
60 -
58 -
56 -
54 -
52
Highest RFG Area
Lowest RFG Area
Retail Avg
Production Avg
t
t
t
'- '.
.
A
1998
62.1
546
56.9
560
1999
60.9
54.8
56.6
560
2000
60.0
53.5
56.6
563
2001
60.1
53.5
56.5
559
2002
60.8
54.4
56.8
560
2003
620
54.3
56.6
560
2004
62.2
548
560
56.2
2005
60.8
54.2
56.9
56.3
» Highest RFG Area
Lowest RFG Area
-*- Retail Avg
w Production Avg
Year
Figure 1
Average Summer RFG E300 {% Evaporated at 300 F)
88 -
86 -
g 84 -
w 82 -
80
Highest RFG Area
Lowest RFG Area
Retail Avg
Production Avg
+
, ,
* * * *
1
T
. *
i . ' * f '
1998 1999 2000 2001 2002 2003 2004
856 858 86.7 85.7 856 87. 1 87.7
81.1 80.7 81.6 81.7 82.0 82.2 80.5
82.7 83.1 84.9 84.5 84.1 84.0 83.3
82.6 82.8 84.7 844 84.4 84.4 83.4
Year
2005
87.9
81 5
84.3
84.1
Highest RFG Area
Lowest RFG Area
-*- Retail Avg
w Production Avg
Figure 2
Average Winter RFG E300 (% Evaporated at 300 F)
86 -
84 -
82 -
Highest RFG Area
Lowest RFG Area
Retail Avg
Production Avg
,
t ' * . * ?
*=
*
_ L**"
*
-
i
.
"
*--
.
1 998
89.0
83.5
85.1
84.9
1999
87.6
82.4
84.6
84.6
2000
90.6
82.8
86.1
86.1
88.9
82.6
80.1
85.8
2002
89.2
81 7
85.7
858
2003
88.3
82.4
856
85.1
2004
89.4
81.9
84.9
849
2005
88.7
82.8
856
85.3
Highest RFG Area
Lowest RFG Area
-*- Retail Avg
w Production Avg
Year
Figure 3
Figure 4
114
-------
Figures 5 and 6 show volume-weighted T50 averages, by year, estimated from RFC Survey data.
The largest year-to-year change Summer T50 was an increase between 1999 and 2000, and
regression analysis indicates that a statistically significant shift in T50 occurred. There was a much smaller
increase in Winter T50 between 1999 and 2000, and regression analysis does not indicate a statistically
significant shift. However, figure 6 shows that Winter T50 was lowest in 2004 and 2005, and regression
analysis through 2004 indicates a statistically significant downward linear trend.
Figures 7 and 8 show volume-weighted T90 averages, by year, estimated from RFC Survey data.
The largest year-to-year changes in both Summer and Winter T90 were decreases between 1999
and 2000, and regression analysis indicates that statistically significant shifts in T90 occurred. Additionally,
both Summer and Winter T90 increased between 2000 and 2004, and regression analysis indicates
statistically significant upward trends through 2004. Decreases from 2004 to 2005 may indicate reversal
of these upward trends.
115
-------
Average Summer RFG T50 (50% Evaporation Temperature)
(degrees
205 -
200 -
195 -
190 -
1QC _
Highest RFG Area
Lowest RFG Area
^'pi.'-iil Average
-------
In general, it is not surprising that changes in Summer RFC distillation parameters occurred
concurrent with the transition from Phase I to Phase II standards. The more stringent Phase II standards
applicable to Summer RFC necessitated a number of refining and blending changes, and these changes
may alter distillation profiles, even if it was not their intended purpose. In addition to the shifts in
distillation parameters, analyses of Summer RFC reporting and survey data concur that statistically
significant shifts in RVP, benzene, aromatics and sulfur occurred between 1999 and 2000.
Phase II requirements for Winter RFC, other than toxics performance, were not more stringent
than Phase I requirements, so there was probably less regulatory pressure for changes in composition, and
hence, less reason to expect significant changes in distillation properties. However, analysis of Winter
survey data indicates statistically significant shifts in E300, T90 and aromatics content between 1999 and
2000. Analysis of reporting data confirms the shift in E300, but does not confirm the shift in aromatics or
identify any other parameters that changed significantly between 1999 and 2000. (Reporting data were
not analyzed for T90 changes.)
Since the two data sets are in agreement with respect to E300 increase, there is little doubt that a
shift occurred between 1999 and 2000, possibly, but not necessarily, as a result of the transition to Phase
II standards. Since there is strong evidence of an E300 change, it is reasonable to assume that there was
some significant change in Winter RFC composition between 1999 and 2000, even if it was not captured in
EPA's data or identified in EPA's analysis (e.g. even if aromatics or olefins content did not change,
proportions of specific aromatics or olefins may have changed).
As noted, the analyses also indicate possible trends in distillation parameters between 2000 and
2004. As explained in the RFC Trends chapter, the major regulatory factors affecting RFC between 2000
and 2005 were the Tier 2 sulfur requirements, MSAT, and state bans on MTBE with consequent increased
use of ethanol. EPA has not undertaken a detailed analysis to identify possible causes for these apparent
trends.
:.- . :--".
Tables 1 and 2 show 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners.
These averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information):
& bf
Season
Summer
Winter
Annual
PADD Average Value
I
II
III
I
II
III
I
II
III
(% evaporated)
47.3
46.6
48.7
55.4
58.4
55.8
52.0
53.5
52.3
^Volumes exclude batches
Volume (gal)*
4,792,114,891
1,740,499,436
5,890,920,167
6,489,530,475
2,438,403,544
6,059,647,031
11,281,645,366
4,178,902,980
11,950,567,198
with missing values
Average Value
(% evaporated)
47.6
46.3
49.9
55.6
56.5
56.7
52.5
52.3
53.3
for this parameter
Volume (gal)
4,431,080,230
1,844,913,833
5,690,766,967
7,081,940,665
2,718,751,541
5,766,524,033
11,513,020,895
4,563,665,374
11,457,291,000
Table 1
117
-------
2004 & 2005
Season
Summer
Winter
Annual
PADD
I
II
III
I
II
III
I
II
III
Reporting Averages by PADD- RFC E300 (Refiner Batches Only)
2004
Average Value
(% evaporated)
83.6
84.7
82.5
85.1
85.7
83.7
84.4
85.3
83.1
^Volumes exclude batches
Volume (gal)*
4,792,114,891
1,740,499,436
5,890,920,167
6,489,530,475
2,438,403,544
6,059,647,031
11,281,645,366
4,178,902,980
11,950,567,198
with missing values
2005
Average Value
(% evaporated)
84.1
84.3
83.2
85.7
84.7
84.5
85.1
84.5
83.9
for this parameter
Volume (gal)*
4,431,080,230
1,844,913,833
5,690,766,967
7,081,940,665
2,718,751,541
5,766,524,033
11,513,020,895
4,563,665,374
11,457,291,000
Table 2
Figures 9 through 12 show estimates of average E200 and E300 levels by PADD in retail RFC. These
averages are volume-weighted averages of the seasonal averages for each area, using gasoline volume
estimates supplied in the survey plans. In 2005, RFC surveys were conducted in 18 PADD I areas, five
PADD II areas and two PADD III areas.
Average E200 of Summer RFC Sold at Retail Stations-By PADD
Year
Figure 9
118
-------
Average E200 of Winter RFC Sold at Retail Stations-By PADD
Year
Figure 10
Average E300 of Summer RFC Sold at Retail Stations-By PADD
Year
Figure 11
119
-------
Average E300 of Winter RFC Sold at Retail Stations-By PADD
Year
Figure 12
Additional Analysis and Observations-RFC
Data analyses pertaining to RFC distillation parameters are contained in the Appendix to this chapter.
Overview-Conventional Gasoline
Figures 13 and 14 show volume-weighted E200 and E300 averages by year. As noted, changes in
distillation parameter averages are indicators of changes in gasoline composition which may or may not be
reflected in the data that EPA collects. EPA's analysis for this report did not extensively look for or identify
specific reasons for the observed changes in CG distillation parameters.
>/oevaporatd@200F)
8
m
Average E200 of Conventional Gasoline
(from Batch Reports)
51.0 -
50.0 -
49.0 -
47.0 -
46.0 -
45.0 -
44.0 -
430 -
42 0
Summer
--Winter
» ___ * ~~~* "
- .
t t " -"" "*
»
1998 1999 2000 2001 2002 2003 2D04 2005
44 6 45 0 450 45.1 448 45.2 45.1 45. 6
499 49.9 50.2 49.7 499 50.3 50. B 50.7
Reporting Year
Figure 13
120
-------
IT
1
1
I
1
83.0 -
Summer
-- Winter
Average E300 of Conventional Gasoline
(from Batch Reports)
_^-~»- _ ^m^"^
- "^ * -"-""'"''^
^*
4- ' """---^ _-*****" ~"""*"*---^_ * *-^"
2^***
1888 iaaa 2000 2001 2002 2003 2BB4 2BB5
SB. a B1.1 80.5 B1.1 Bfl.6 Bfl.7 Bfl.6 B1.B
83.2 83.0 83.4 83.2 83.1 82.8 833 841
Reporting Year
Figure 14
CG Distillation Parameters by PADD
Tables 3 and 4 show 2004 and 2005 reporting averages by PADD for PADD I, II and III refiners. These
averages were calculated from batch data, excluding importer batches (see "PADD Level Analysis"
appendix for additional information):
2004 & 2005 Reporting Averages by PADD-CG E200 (Refiner Batches Only)
Season
Summer
Winter
Annual
2004
PADD Average Value
I
II
III
I
II
III
I
II
III
(% evaporated)
45.5
46.6
44.6
51.5
52.6
49.7
48.4
49.8
47.2
^Volumes exclude batches
Volume (gal)*
3,752,272,894
11,456,335,295
21,716,635,879
3,482,171,831
13,313,565,716
23,627,111,420
7,234,444,725
24,769,901,011
45,343,747,299
with missing values
2005
Average Value
(% evaporated)
45.1
47.5
44.8
51.0
52.7
49.3
48.2
50.4
47.1
for this parameter
Volume (gal)*
3,547,001,722
10,565,279,667
20,889,312,566
4,063,455,417
13,578,145,833
22,444,499,833
7,610,457,139
24,143,425,500
43,333,812,399
Table 3
121
-------
2
Season
Summer
Winter
Annual
004 & ;
PADD Average Value
(% evaporated)
I 82.7
II 81.4
III
I
II
III
I
II
III
79.4
84.4
84.1
81.8
83.5
82.9
80.6
^Volumes exclude batches
s bf ESC
Volume (gal)*
3,752,272,894
11,456,437,187
21,722,168,035
3,482,279,645
13,313,789,786
23,627,111,420
7,234,552,539
24,770,226,973
45,349,279,455
with missing values
Ji
Average Value
(% evaporated)
82.8
81.9
80.5
84.4
84.6
82.6
83.7
83.4
81.6
for this parameter
s
Volume (gal)*
3,547,001,722
10,566,219,165
20,889,312,566
4,063,455,417
13,583,692,059
22,444,499,833
7,610,457,139
24,149,911,224
43,333,812,399
Table 4
Data analyses pertaining to CG distillation parameters are contained in the Appendix to this chapter.
122
-------
Pcrfbrrc
Both RFC and CG must meet certain emission-related standards tied to the Complex Model. This
model produces VOC, NOx and toxics emission estimates for "1990 technology" light duty gasoline vehicles
as a function of the gasoline parameters discussed throughout this report. The model is used to produce
two types of estimates; "milligrams per mile" emission estimates for a given gasoline formulation and
"percent reduction" emission performance estimates which compare the emissions of a given gasoline
formulation to the emissions of the 1990 "statutory" baseline gasoline whose properties are specified in
the Clean Air Act and/or in EPA's regulations. CG Anti-Dumping standards are intended to maintain the
NOx and exhaust toxics emissions qualities of the gasoline that each individual refinery or importer
produced in 1990. RFC standards require VOC (for Summer RFC), total toxics (including evaporative
benzene for Summer RFC) and NOx emission reductions from the specified 1990 baseline gasoline. The
"milligram per mile" Complex Model estimates are used to evaluate CG emission compliance and the
"percent reduction" estimates are used to evaluate RFC compliance.
Quantitative emissions and emissions performance estimates in this report are based on the
Complex Model. The standards applicable to gasoline in years 1998 and 1999 were based on the Phase I
version of this model, however all analyses in this report are based on the currently applicable Phase II
Complex Model. Trend analyses based on a combination of Phase I and Phase II model calculations would
provide little information on changes in the emissions qualities of gasoline since they would show the
effects of the model version change as well as the effects of actual property changes on emissions.
In addition to the Phase I and Phase II model difference, there are multiple versions of the Phase
II model. Different versions of the model as well as different baseline gasoline properties apply to
Summer and Winter gasoline. The Summer model produces different VOC and toxics estimates for VOC
Control Region 1 (Southern) and VOC Control Region 2 (Northern) RFC, because its calculation of the non-
exhaust portion of these emissions varies by region. As with the gasoline property analyses, EPA
categorized data from each reporting or survey year as "Summer" or "Winter", and separately analyzed
these data. However, while direct comparison of seasonal or regional gasoline properties is often
informative, direct comparison of emissions or emission reductions could lead to erroneous conclusions.
For example, it is incorrect to conclude that the differences between Summer and Winter model milligrams
per mile estimates in a given year are due solely or primarily to differences in the emissions characteristics
of Summer and Winter gasoline.
EPA's Complex Model was developed from several thousand emission tests collected from studies
designed to measure the effects of gasoline properties on vehicle emissions. In order to meet Clean Air
Act's RFC requirements this model was based on 1990 vehicle technology, and it has not been updated to
include newer technology vehicles. Although the model's developers used rigorous statistical analysis to
produce a sophisticated emission model, the Complex Model was intended to be used to determine if
gasoline blends complied with emission standards rather than to estimate fleet-wide emission changes
resulting from changes in gasoline.
Section 1506 of The Energy Policy Act of 2005, ("Analyses of Motor Vehicle Fuel Changes") which
amends Section 211 of the Clean Air Act, directs the Administrator of EPA to develop and finalize an
emissions model that reflects the effects of gasoline characteristics or components on emissions from
vehicles in the motor vehicle fleet during calendar year 2007. This section contains additional
requirements addressing the effects of fuel changes on emissions, including a "permeation effects" study
which will provide estimates of evaporative emission increases that could result from use of gasoline with
ethanol content in a motor vehicle. Fulfilling the requirements in this section of the Energy Policy Act
should provide analytical tools to better understand, prospectively and retrospectively, the emission
impacts of gasoline property changes.
123
-------
In summary, the emissions and emissions performance analyses in this chapter and elsewhere in
this report provide some basis for comparing the emissions characteristics of production and retail gasoline
to standards. They are also indicators of changes in the emissions qualities of gasoline over time, without
considering changes in vehicles over time. Results allow some quantitative comparisons, and provide
some means of judging relative emissions qualities and trends. However, the "milligram per mile" and
"percent reduction" values presented here are not intended to and almost certainly do not provide an
accurate estimate of the overall emissions changes that have occurred as a result of gasoline property
changes.
As stated, RFC refineries and importers must meet VOC, total toxics and NOx emissions
performance standards. These standards are specified as percent reductions from emissions with a
Summer or Winter baseline gasoline (i.e. "statutory" baseline gasoline). Refiners and importers may
choose to comply with "per gallon" standards or with "averaged" standards. These averaged standards
require greater emission reductions than per gallon standards, but allow more batch to batch variability in
emissions performance. Suppliers who comply with the averaged VOC performance standard must also
meet a per-gallon minimum performance standard. There are no per-gallon minimums for toxics or NOx.
VOC standards, which are applicable only to Summer RFC, vary by VOC control region and suppliers much
designate the region for each batch. Slightly less stringent VOC standards apply to RFC containinglO
volume percent ethanol supplied to the Chicago and Milwaukee areas.
Suppliers generally elect to comply with averaged standards. EPA's regulations required the RFC
surveys that generated the retail data analyzed in this report as a condition of compliance with averaged
standards. The survey regulations include requirements for VOC, toxics and NOx performance surveys,
and failure criteria for each type of survey.
Current and historic emission performance standards and requirements for RFC compliance are
contained in EPA's regulations (40 C.F.R. '80.41). Where relevant, averaged emission performance
standards are displayed in the graphical and tabular analyses in this report.
CG refineries and importers must meet "milligram per mile" exhaust toxics and NOx emission
standards on an annual average basis. Standards applicable to CG refineries and importers are facility-
specific and largely dependent on the "individual baseline" properties and volumes of the gasoline they
supplied in 1990.
The Mobile Source Air Toxics regulations impose "anti-backsliding" toxics requirements on both
RFC and CG suppliers. RFC MSAT standards are total toxics performance standards and CG standards are
exhaust toxics emissions standards. Both the RFC and CG requirements are facility-specific.
Since 1998, refiner and importer batch reports for both CG and RFC included the Complex Model
outputs necessary for compliance determination, as well as the property information needed to determine
the Complex Model inputs. RFC Survey data submitted since 1998 has also included both Complex Model
outputs and input information. Prior to 1998 neither data source included the full set of Complex Model
input properties. Consequently, although estimates of Complex Model emission performance could have
been derived for earlier years with assumptions about the missing inputs, EPA has restricted the emissions
analyses in this report to 1998 and later.
124
-------
EPA recalculated 1998 and 1999 Complex Model outputs for each batch and survey sample using
the Phase II model to provide a common basis for measurement of emissions trends. Additionally, both
to provide information for further analysis and to verify the accuracy of the reported results, EPA
recalculated Complex Model outputs for 2000 through 2005 CG and RFC batch data. Differences between
average estimates based on reported and calculated values were small. EPA chose to rely on calculated
emission values, rather than reported emission values for all CG analyses. However, EPA found that some
RFC blendstock (RBOB) batch reports did not contain information on the type (presumably ethanol) and
concentration of oxygenate. (While most parties report this information they are not required to do so.)
These batches contained other parameter information as well as reported emission performance values.
Since EPA could not correctly calculate emission performance for these batches but did not want to
exclude them from trend analyses, EPA used reported emission performance values for all batches in
overall RFC trend analyses. Some of the additional analyses in this chapter use calculated RFC emission
or emission performance results for the batches with complete parameter information. These results are
labeled "calculated from batch data".
The General Methodology chapter in this report notes that EPA screened the batch data in this
report and describes, in general, the procedures used to screen gasoline property data. EPA excluded
some data from the emissions analyses as well although the exclusion criteria were somewhat different.
As explained in that chapter certain batches with outlier values for one or more parameters were flagged
and excluded from distribution by volume analyses for all parameters, but not necessarily from average
calculations. Since the Complex Model has multiple parameter inputs, flagged batches were excluded from
average emission and emission performance calculations as well as distribution by volume analyses. EPA
also used maximum and minimum results from its calculated RFC emissions performance results to
estimate a range of reasonable values in order to flag emission outliers that are likely to be erroneous
values.
For gasoline property analyses, CG blendstock batches were included in averages but excluded
from property by volume distributions. CG blendstocks were excluded from all emission analyses. While
EPA's regulations include an "equivalent emissions performance" (EEP) procedure for calculating Complex
Model emissions for these blendstock batches, this procedure is for compliance purposes and inclusion of
these EEP batches in emissions analyses would not necessarily produce better trend estimates.
(Additionally, it would have been difficult for EPA to replicate these EEP calculations.)
: ' rfujj
The VOC and NOx performance standards applicable to Summer RFC and the toxics standard
applicable to both Summer and Winter RFC became more stringent in 2000 with the transition from Phase
I to Phase II RFC standards. Analyses of both reporting and survey data confirmed improvements in
performance concurrent with this transition. Figures 1 through 3 show the Summer RFC performance
trend lines derived from both reporting and survey data. (These performance trend lines were also
presented in the "RFC Trends" chapter). Since Phase I Summer RFC, on average, did not meet either the
averaged or per-gallon Phase II VOC and NOx standards, an improvement in average performance was
necessary. Summer toxics performance clearly improved as well, even though Phase I RFC, on average,
over-complied with the Phase II standard. An analysis presented in the RFC Trends chapter illustrates
one probable reason for this toxics performance improvement. This analysis shows that certain of the RFC
property changes that reduced VOC and/or NOx emissions also reduced toxics emissions. Even if the
toxics standard had not become more stringent with Phase II, the more stringent VOC and NOx standards
would have forced certain property changes beneficial to toxics performance. RFC emission performance
125
-------
standards applied on a refinery or importer-specific basis and, unlike oxygen and benzene content
standards, could not be met through the transfer of credits. Consequently, even though Phase I Summer
RFC on average, over-complied with the Phase II toxics standard, it is probable that performance also
improved because some refiners and importers did not meet Phase II toxics performance standards during
Phase I and would not have met the toxics performance standard solely through property changes
necessary for Phase II VOC and NOx compliance. EPA did not analyze the data on a facility-specific basis,
however EPA did determine that average benzene levels dropped with the transition to Phase II. Benzene
content reductions lower Complex Model toxics emissions but do not directly affect VOC or NOx emissions,
suggesting that some suppliers reduced benzene content in order to improve toxics performance. (See the
Benzene Chapter).
The Summer RFC trend lines also show, on average, significant over-compliance with the current
NOx and toxics performance standards, but very little VOC compliance margin. The survey data shows
the best NOx performance in 2005 and the reporting data in 2004, and the improvement in NOx
performance since 2000 is the result of the mandated Tier 2 gasoline sulfur reductions.41 These
reductions were necessary to enable the emission control systems in "new technology" Tier 2 vehicles to
be fully effective. However, the improvement in Complex Model NOx emission performance suggests that
these sulfur reductions have reduced NOx emissions in older vehicles using RFC as well. Although sulfur
reductions, all else constant, also lower VOC and toxics emissions, the trend lines through 2005 do not
show any performance benefits. Analyses presented in the RFC Trends Chapter demonstrated that the
VOC and toxics emission benefits of sulfur reductions between 2000 and 2005 were offset by changes in
other properties that increased VOC and/or toxics emissions.
The data are more equivocal about Winter RFC performance. Figures 4 and 5 show that both
reporting and survey average estimates indicate better Winter NOx and toxics performance in 2000 than in
1999. However only the survey estimates suggest that the transition from Phase I to Phase II RFC
standards had a strong influence on Winter RFC emissions performance, since the reporting system
performance improvements concurrent with this transition are not markedly different from other year-to-
year performance changes. These trend lines show that Phase I Winter RFC, on average, substantially
over-complied with Phase II Winter NOx and toxics emission "averaged" performance standards.
Additionally, after adjusting the Phase I standards to the Phase II model, it is apparent that for Winter RFC
only the toxics standard, which is applicable on an annual average basis, became more stringent in Phase
II. Consequently, while EPA's aggregate data analysis did not rule out the need for Phase II performance
improvements in Winter RFC, it did not indicate that emission performance improvements were necessary.
The disparity between the reporting and survey estimates leaves some question about the effect
of the Phase II requirements on Winter RFC. Although there is some rationale for relying more heavily
on reporting data to estimate overall averages, there are also reasons to consider survey-based estimates
as well. The issue of disparity between reporting and survey estimates has been discussed in the chapter
on aromatics. Since aromatics content is an important parameter in Complex Model NOx and toxics
emissions calculations, differences between reporting and survey aromatics content estimates are likely to
be strongly related to differences in performance estimates.
While there is some divergence in the middle, the beginning and end-points of the reporting and
survey-based trend estimates are in better agreement for both NOx and toxics. Both NOx lines show
approximately the same performance improvement between 1998 and 2005. The increased rate of
improvement between 2003 and 2005 almost certainly resulted from the phased reduction in gasoline
41A "ratchet" of the RFC NOx standards in 2002 and 2003, resulting from a 2001 NOx survey failure in Sussex
County, DE may have had a minor effect on overall post-2000 NOx performance. These more stringent (by 1%)
average NOx reduction standards applied to refineries that supplied RFG to this area.
126
-------
sulfur standards beginning in 2004. Both toxics estimates concur that toxics performance in 2005 was
better than in 1998 and 1999, and that toxics performance improved from 2003 to 2005.
VOC Performance of Summer RFC
(Based on Phase II Complex Model)
"5"
.E
% Emission Reduction (from baseline gaso
30 -
Notes
1 . Approximation
based on phase I!
complex model
varies by geographic
weighted average of
standards shown ,-
* Best Area
Worst Area
Retail Avg.
- - - Reporting Avg.
Phs. I Averaged Std (vol wtd) 1 .2
Phs. II Averaged Std. (vol wtd) 2
* « T- * ?
Jf T . . i ,
I !/
iaas igaa
24.8 25 5
14.9 14.6
18. B 19.4
19.3 19.8
16.9 17.2
20DD
29.9
27.6
28.6
286
26.1
2001
30.0
27.1
28.6
28.5
27. B
2002
302
27.1
28.5
28.2
27-9
2003
304
27.3
285
283
279
2004
30 1
26.6
28.1
28.3
279
20D5
31.0
26.7
28.2
28 3
278
Year
Figure 1
"S
c
i
I
UJ
NOx Performance of Summer RFC
(Based on Phase II Complex Model)
18 -
12 -
6 -
2 -
°
Best Area
Worst Are i
Retail Avg.
Reporting Avg.
Phs II Averaged Std
Phs 1 Averaned Std *
*
.
T " '
.
T __[/_
J^~
~k~~"~~V
i r
1898
7.3
38
48
5.0
1.5
1999
86
2.5
43
4.7
1.5
2000
13.5
7.5
98
92
6.8
2001
11.3
73
94
8.4
68
2002
13.5
72
92
9.0
6.8
2003
14. B
80
97
9.4
6.8
2004
17.0
8.0
10.5
10.5
6.8
2005
17.7
8.4
10.9
10.2
6.8
Year
"Phase I standard is an
approximation
based on the Phase II
complex model
Figure 2
127
-------
I
Toxics Performance of Summer RFG
(Based on Phase II Complex Model)
35-
U
Best RFG Area
Worst RFG Area
Retail Avg
Reporting Avg.
Phs II Averaged Std
Phs I Averaged Std *
^-^
i
1
1998
31.6
209
26.0
291
14
' ,
.
1399
326
203
27 .8
28.9
14
3
:
2000
36.6
29.7
34.3
34.0
21.5
1
i
2001
351
27.8
33.4
33.0
21.5
..
2002
34.7
26.2
33.3
33.4
21.5
2003
35.3
29.8
33.4
33.7
21.5
2004
361
28.8
32.7
33.5
21.5
1
2005
34.4
28.5
32.3
32.6
21.5
*Phase I standard is an
approximation
based on the Phase II
cornclex model
Figure 3
NOx Performance of Winter RFG
(Based on Phase II Complex Model)
1R -
16 -
B -
Best Area
* Worst Area
Retail Avg.
Reporting Avg.
Phs II Averaged Std
Phs I Averaged Std *
i
F"-=
+
1998
10.1
49
6.4
61
1.7
-ar^
'
1999
9.3
1.4
5.7
5.5
1.7
1
S
-
2000
11.4
50
7.6
5.9
1.5
1
--
2001
11.2
4.4
7.4
6 1
1 5
*
2002
1 1 .4
37
7.2
6.8
1 5
1
^X*
;
2003
11.5
4.7
76
7.4
1.5
_ '
^
2004
13.1
6.8
10.1
9.7
1.5
;
_ J
2005
17.6
7.9
11.6
10.5
1.5
"Phase I standard is
an approximation
based on the Phase II
complex model
Figure 4
128
-------
Toxics Performance of Winter RFC
(Based on Phase II Complex Model)
I
5 -
Best RFC Area
* Worst RFG Area
Retail Avg
Reporting Avg.
Phs. II Averaged Std.
Phs. 1 Averaged Std *
m
t
1998
20.8
19.8
25.1
250
17.5
1
1999
28.4
19.6
24.1
24.6
17.5
1
i
2000
31.1
22.0
286
253
21.5
;
2001
29.5
22.4
26.4
25.3
21.5
|
2002
28.8
225
260
25.5
21.5
1
2003
29.7
21.9
259
25.6
21.5
|
2004
29.4
223
265
268
21.5
_.;
|
2005
29.6
225
27.3
26.7
21.5
'approximation
based on phase II
complex model
Figure 5
Additional Analyses and Observations-RFC
VOC Emissions and Emission Performance
Figure 6 shows the exhaust and non-exhaust components of Complex Model VOC emissions
expressed in milligrams per mile, rather than as percent reductions. The Phase II Complex Model
attributes about 69 percent of Phase I and about 72 percent of Phase II Summer RFG VOC emissions to
exhaust emissions. While both exhaust and non-exhaust emissions were reduced in order to meet Phase
II VOC standards, non-exhaust emission reductions accounted for about 63% of the total reduction. The
Complex Model estimates non-exhaust VOC emissions from gasoline RVP (and the VOC control region
specification). All Complex Model input parameters except benzene affect Complex Model exhaust VOC
emissions. In addition to producing the non-exhaust VOC reductions, RVP reductions contributed to
exhaust VOC reductions. Consequently, RVP reduction was critical to meeting Phase II VOC reduction
standards.
RFG Exhaust and Non-Exhaust VOC Milligrams per Mile Emissions
(Calculated from Batch Data with Phase II Model)
Figure 6
129
-------
The emission estimates in Figures 1 and 6 reflect a composite of RFC designated for VOC Control
Region 1 and VOC Control Region 2. Figure 7 shows VOC emission performance, based on the VOC
control region specified for each batch, and calculated with the appropriate regional version of the Phase II
Complex Model. The Region 2 data include some "adjusted VOC" ethanol-oxygenated RFC supplied to
Chicago and Milwaukee, although this RFC may be under-represented because of data screening. This
graph correctly suggests that there was a substantial difference between Region 1 and Region 2 VOC
performance prior to 2000 and that both improved between 1999 and 2000, narrowing the gap. However,
as explained earlier, it is important to interpret region-to region emission comparisons correctly.
1 %Reduction from 1990 Gasoline
RFG VOC Emissions Performance by VOC Control Region
(Calculated from Batch Data with Region-Appropriate Phase II Model)
30 -
25-
15 -
« Region 1
Region 2
Avg. Std.-Region 1
Avg. Std-Region 2
-- Adjusted Std.-Region 2
/
. /
1998 1999 2008 2001 2002 2083 2884
25.1 25.4 29. B 295 29.3 29.4 29.4
16.2 16.4 27.8 278 27.5 27.5 27.5
29.0 28.0 29.0 29.0 29.0
27.4 27.4 27.4 27.4 27.4
25.4 25.4 25.4 25.4
Year
2005
28.4
27,4
230
27.4
25.4
Figure 7
Since the Complex Model evaluates the non-exhaust portion of VOC emissions differently
depending on the VOC Control Region, these composite emissions reflect, to some degree, a geographic
effect related to temperature, as well as the effects of the gasoline properties on emissions. In order to
remove this geographic effect, non-exhaust VOC emissions for VOC Region 1 batches were recalculated
using the VOC Region 2 model. This allows a side-by-side comparison of the property-related VOC
emission differences between these two types of RFG. Figure 8 shows, on average, that the property-
related emission characteristics of these two types of RFG, which differed during Phase I primarily because
of non-exhaust emissions, became essentially the same after the Phase II standards took effect.
Comparison of F
Calculated from
900 -r
800 -
700 -
§ 600 -
S. 500 -
400 -
j 300-
200 -
100 -
0 -
nExti. VOC-Region1 RFG
Exh. VOC-Region 2 RFG
DNon-exh VOC-Region 1 RFG
n Non-exh VOC-Region 2 RFG
egion 1 and Region 2 Complex Model VOC Emissions
Batch Data with Phase II Region 2 Model-w/o "Adjusted
VOC" RFG
r
-
1998
785
796
284
376
-
1999
778
791
286
379
n
2000
741
741
266
269
1
1
2001
743
735
265
270
2002
744
740
266
269
1
~n
1
2003
737
739
271
270
1
2004
736
737
272
272
_
2005
731
734
276
277
Year
Figure 8
130
-------
The Energy Policy Act of 2005 requires that EPA consolidate the RFC regulations applicable to VOC
Control Region 1 and VOC Control Region 2 by eliminating the less stringent requirements applicable to
VOC Control Region 2 and instead applying the more stringent requirements applicable to VOC Control
Region 1 C1504(c)). The VOC performance standard for VOC Control Region 1 (29.0 percent "averaged"
reduction) is numerically more stringent than the Region 2 standard (27.4. percent except for "adjusted
VOC" gasoline supplied to Chicago and Milwaukee). While there is an apparent 1.6 percent difference in
stringency between the two standards, in fact, as the above analysis suggests, there is little or no real
difference in stringency. To confirm this, EPA evaluated the Complex Model VOC performance of 4194
individual Summer RFC formulations reported in 2004 batch reports, using the Region 1 and Region 2
versions of the Complex Model. These batches were a mix of Region 1 and Region 2 RFC. When all
formulations were evaluated with the Region 1 model the average performance was 29.2 percent, and
when evaluated with the Region 2 model average performance was 27.7 percent. The 1.5 percent
difference between the two performance averages is nearly the same as the difference between the
Region 1 and the unadjusted Region 2 standards, demonstrating that most of this apparent difference in
stringency is due to differences between the Region 1 and Region 2 models.
Figure 9 shows exhaust, non-exhaust and total VOC milligram per mile emissions plotted as
differences from year 2000 emissions. Both exhaust and non-exhaust emissions changes between 2000
and 2005 are clearly very small compared to changes from 1999 to 2000. Exhaust emissions are lower in
2005 than in 2000, while non-exhaust emissions are higher in 2005 than in 2000. In particular, the
exhaust emission decreases between 2002 and 2005 were nearly offset by non-exhaust emission
increases.
Summer RFC VOC "Milligram Per Mile" Emissions Relative to Year 2000
(Calculated from Batch Data)
Figure 9
These opposing movements in exhaust and non-exhaust VOC emissions provide tenuous evidence
of a trend because of their small magnitude and short duration. However, additional factors suggest that
they may represent something other than random fluctuations. The VOC emission performance trend
lines (Figures 1 and 6) show that VOC compliance margins are small especially when compared to NOx
and toxics compliance margins, indicating that RFG's VOC emission performance standards are more
constraining than these other RFC standards. It was suggested in the RVP chapter that suppliers would
strive to increase RVP in Summer RFC, but that RVP was constrained by the VOC performance standard.
Summer RFC sulfur levels decreased continuously from 2002 to 2005 as a result of the Tier 2 gasoline
131
-------
sulfur requirements, and these sulfur reductions were partially responsible for exhaust VOC reductions.
Although these sulfur reductions could have provided additional VOC compliance margin, refiners may
have chosen to take back this additional compliance margin by raising RVP. This is generally consistent
with an analysis, based on 2000 and 2005 RFC Survey property estimates, presented in the RFC Trends
chapter. It showed that VOC emission reductions due primarily to sulfur decreases were offset by VOC
increases due primarily to RVP increases. This is also somewhat consistent with assumptions that EPA's
Mobile 6.2 model makes when modeling the impact of an RFC program. This model assumed an RVP
increase from 6.7 to 6.8 psi in RFC beginning in 2003 as a result of the Tier 2 gasoline sulfur program
(Brzezinski, 2001).
It is important to recognize that even if the Tier 2 reductions in sulfur content do interact with RFC
requirements in ways that affect how refiners meet RFC standards, the resultant property changes may be
small (e.g. RVP increased by slightly more than 0.1 psi between 2000 and 2005, and other factors, such as
the implementation of the Chicago/Milwaukee "adjusted VOC" standards contributed to this increase).
Moreover, even if the sulfur reductions do cause changes in RFC composition, it does not necessarily follow
that RFG's overall emissions impact would be adversely affected.
The Complex Model estimates emissions of exhaust benzene, non-exhaust benzene, acetaldehyde,
formaldehyde, 1, 3 Butadiene and polycyclic organic matter (POM). It sums these estimates to estimate
total toxics and makes no adjustments for the different potencies of these toxic pollutants. The Winter
Complex Model assumes that non-exhaust benzene emissions are zero. Figures 10 and 11 show the
milligram per mile contribution of each of these toxic pollutants to RFC toxics emissions, and exhaust
benzene is the major contributor to both Summer and Winter toxics emissions.
Figures 12 and 13 show each toxics component and total toxics plotted as milligram per mile
changes from the 1998 estimates. The change in exhaust benzene emissions over time was the major
component of the toxics emission trends for both Summer and Winter RFC. Changes in several gasoline
parameters, including but not limited to gasoline benzene content, have had significant effects on RFC
benzene emissions. Analysis based on RFC Survey data presented in the RFC Trends chapter showed
that aromatics, sulfur, and benzene content reductions accounted for much of the Phase I to Phase II
toxics reduction in Summer RFC and that sulfur reductions produced further toxics reductions between
2000 and 2005. These content changes, all else constant, lower exhaust benzene emissions.
Figures 12 and 13 also show that recent increases in acetaldehyde emissions have partially offset
decreases in benzene emissions. These acetaldehyde emission increases probably resulted from
increased use of ethanol in place of MTBE. RFC ethanol use increased substantially between 2002 and
2005 concurrent with these acetaldehyde increases. (See the RVP Trends and Oxygenates chapters.)
The Complex Model, for toxics emissions calculations only, considers not only the amount of oxygen but
the specific oxygenate(s) used. According to the Complex Model, adding any oxygenate to gasoline, all
else constant, will reduce certain toxics, including exhaust benzene. Additionally, use of any oxygenate
may further reduce toxics emissions because of dilution of and/or substitution for gasoline components
that adversely affect toxics emissions. However, the Complex Model predicts that specific oxygenates will
have different effects on certain of the constituent toxics in the total toxics calculation. It predicts that
acetaldehyde emissions increase as ethanol content increases and also as MTBE content decreases. The
model indicates no relationship between ethanol and formaldehyde emissions, but predicts a positive
relationship between MTBE content and formaldehyde emissions. However, EPA's analysis did not find
formaldehyde emission decreases concurrent with the substantial decrease in MTBE use. This lack of
correlation between MTBE reduction and formaldehyde emission reduction can occur since all Complex
Model toxics are functions of multiple parameters.
132
-------
the
The chapter appendix contains additional RFC analyses which treat the Complex Model emission
performance parameters in the same manner as the gasoline property parameters addressed in other
chapters, including grade-to-grade and geographic comparisons. Since these emission performance
comparisons connote that some RFC is better than other RFC, some further discussion is warranted.
There are patterns in grade-to-grade emissions performance. For example, on average, premium
gasoline exhibited consistently better Summer and Winter NOx performance than regular or mid-grade
gasoline. This is probably largely due to lower sulfur and olefins levels in premium RFC (see the sulfur and
olefins chapters). These differences were largely due to refining and blending practices needed to meet
the octane requirements of each gasoline grade. Although the Complex Model may have identified
premium RFC as being, in certain respects, a "cleaner" gasoline than other grades of RFC, it does not
necessarily follow that significant environmental benefits would occur with increased use of premium
gasoline in vehicles that do not need premium gasoline
The grade-to-grade differences may have been affected, to an extent, by some refiner decisions to
market a low sulfur premium grade gasoline in advance of Tier 2 sulfur reduction standards. EPA's
analysis for this report did not isolate and compare the Complex Model emissions performance of gasoline
brand/grade combinations explicitly marketed as "cleaner" to that of other of the same grade. However,
it is quite possible that such produced lower emissions than other of the same grade.
RFC emissions performance varied geographically, even for RFC covered by the same region-
specific standards. Although some areas received better RFC than others, by various Complex Model
emission performance measures, EPA's regulations, in particular its RFC Survey requirements, ensured that
each area received "good" RFC.
133
-------
Contribution of Individual Toxics to Summer RFC Total Toxics Emissions
(Calculated from Batch Data)
Contribution of Individual Toxics to Winter RFG Toxics Emissions
(Calculated from Batch Data)
100
90
80
70
60
50
40
30
20
10
0
Butadiene
n Formaldehyde
n Acetaldehyde
aExh. Benzene
2005
Figure 10
Figure 11
Summer RFG Toxics Milligrams per Mile Relative to Year 1998
(Calculated from Batch Data)
1 -
o -
!-'-
I
i -2 -
8
-3 -
-4 -
1998 1990
jijuC
V
^-^"*~-
^
2oai
^»-
joo;
-s.
-,
^
-*v
/
\
.''-
20LK
"V
'
== *-
/
s
Year
I 1 trite |
Exh. Benzene
Non-exh. Benzene
* Acetaldehyde
Formaldehyde
* Butadiene
POM
Winter RFG Toxics Milligram Per Mile Relative to Year 1998
(Calculated from Batch Data)
Figure 12
Figure 13
134
-------
Figures 14 through 17 show exhaust toxics and NOx emission trend lines for CG. Since CG
standards are facility-specific these graphs, unlike the RFC trend graphs, do not show standards.
However the graphs show the seasonally-appropriate 1990 baseline gasoline emission value, which
provides some basis for comparison. In all cases, average emission levels have been lower than these
baseline emission levels.
For each pollutant-season combination the lowest emission levels were in 2004 or 2005 and the
largest year-to-year changes were emission decreases between 2003 and 2004. These changes were
concurrent with the large gasoline sulfur reductions that resulted from the phase in of Tier 2 gasoline
sulfur standards. EPA's analysis did not investigate the extent to which individual emission parameters
influenced CG emissions, but it is virtually certain that these sulfur reductions were the primary cause of
CG emission reductions.
The Phase II Complex Model exhaust toxics emission rate for 1990 Summer Baseline Gasoline is 80.10
milligrams per mile and the exhaust toxics emission rate for 1990 Winter Baseline Gasoline is 120.55
milligrams per mile. The appendix to this chapter includes tabular and graphical descriptions of CG
emission rates which show:
» In 1998, the first year for which Complex Model emissions were analyzed, the median Summer
exhaust toxics emission rate was 70.6 mg/mi and the 1990 baseline gasoline fell between the 80th and
85th percent! le.
» In 2005, the last year for which Complex Model emissions were analyzed, the median Summer exhaust
toxics emission rate was 66.7 mg/mi and the 1990 baseline gasoline fell between the 85th and 90th
percent! le.
» In 1998, the median Winter exhaust toxics emission rate was 107.5 mg/mi and the 1990 baseline
gasoline fell between the 80th and 85th percentile.
» In 2005, the median Winter exhaust toxics emission rate was 100.8 mg/mi and the 1990 baseline
gasoline fell between the 85th and 90th percentile.
The Phase II Complex Model NOx emission rate for 1990 Summer Baseline Gasoline is 1340.0 milligrams
per mile and the NOx emission rate for 1990 Winter Baseline Gasoline is 1540.0 milligrams per mile.
In 1998, the first year for which Complex Model emissions were analyzed, the median Summer NOx
emission ra
percentile.
emission rate was 1323.5 mg/mi and the 1990 baseline, gasoline fell between the 55th and 60th
In 2005, the last year for which Complex Model emissions were analyzed, the median Summer NOx
emission rate was 1234.0 mg/mi and the 1990 baseline gasoline fell between the 85th and 90th
percentile.
In 1998, the median Winter NOx emission rate was 1489.4 mg/mi and the 1990 baseline gasoline fell
between the 60th and 65th percentile.
In 2005, the median Winter NOx emission rate was 1386.2 mg/mi and the 1990 baseline gasoline fell
135
between the 90th and 95th percentile.
-------
*
Q.
£
2
o
S
I7?n -
CG NOx emissions
1990 baseline NOx
Phase II Complex Model NOx Emissions - Summer CG
(Based on Batch Data-Excluding Blendstocks)
^ . . *-^_ ___
\
\
\
v »
1998 1999 2900 2001 2002 2003 2004 2005
1325.7 1331.4 1333.7 1332.9 1327.5 1325.5 1252.6 1250.5
13400
Year
Phase II Complex Model NOx Emissions - Winter CG
(Based on Batch Data-Excluding Blendstocks)
1350 -
CG NOx emissions 1492.0 1501.6 1500.5 1508.6 1499.7 1483.4 1423.5 1400.9
1990 baseline NOx 1540.0
1990 1999 2000 2001 2002 2003 2004 2005
Figure 14
Milligrams per Mile
Phase II Complex Model Exhaust Toxics Emissions - Summer CG
(Based on Batch Data-Excluding Blendstocks)
74 -
72 -
64 -
CG exh. toxics
1990 baseline exh. toxics
\
>*. ' »
1998 1999 2000 2001 2002 2003 2004 2005
72.5 73.0 73.5 73 6 72.6 73.0 68.7 69.1
80.1
ffeai
Figure 15
Phase II Complex Model Exhaust Toxics Emissions - Winter CG
(Based on Batch Data-Excluding Blendstocks)
1990 baseline exh. toxics
Figure 16
Figure 17
136
-------
To facilitate comparison of RFC and CG Complex Model emission qualities, EPA has evaluated RFC
exhaust toxics and NOx "milligram per mile" emissions, and CG VOC, total toxics and NOx "percent
reduction" emission performance. These results are presented in this chapter in tabular form, together
with the appropriate results for comparison. Graphical comparisons of CG and RFC emissions
performance are also presented in the Conventional Gasoline Trends chapter of this report. Since the
VOC Control Region specification, which is not provided for CG, affects Summer VOC and Summer total
toxics emissions calculations, all CG percent reduction calculations assumed VOC Control Region 2, and the
control-region dependent CG results are compared to VOC Control Region 2 RFC.
It is apparent from these tables that CG, at least since 1998, has been "cleaner" than 1990
baseline gasoline. RFC has provided substantial additional emission benefits relative to CG in each year
and season although improvements in CG emissions have narrowed the gap between RFC and CG
The largest improvements in Summer and Winter CG emission performance for all pollutants
occurred between 2003 and 2004, concurrent with the largest sulfur reductions. NOx performance is
particularly sensitive to sulfur content. However, even though CG sulfur levels in 2004 dropped to levels
comparable to or below those of "pre Tier 2" RFC, the average NOx performance of 2005 Summer CG was
slightly worse than the "averaged" standard for RFC. Winter CG, in each year since 1998, has met the
Winter RFC NOx 1.5% averaged performance standard.
It is not the intent of this report to extrapolate these data to estimate RFC or CG emission levels
when the Tier 2 sulfur reductions are fully implemented. However, CG NOx performance will likely
improve and the gap between RFC and CG NOx performance may decrease, but not necessarily disappear,
as further sulfur reductions occur. Although analysis or detailed discussion of the air quality benefits
associated with these emission reductions is also beyond the scope of this report, it is reasonable to expect
that any incremental RFC NOx reductions beyond the RFC standards or CG performance levels may
translate into tangible air quality benefits. This is clearly the case for Summer RFC in areas that need to
achieve additional NOx reductions in order to attain ambient ozone standards. Winter RFC over-
compliance may also be important because NOx plays a role in secondary particulate formation.
CG in Per
Year
1998
1999
2000
2001
2002
2003
2004
2005
1990 Statutory
Baseline
Emissions
Summer
RFC
Summer
CG
Exhaust toxics
59.0
59.1
55.3
56.0
55.7
55.5
55.7
56.3
72.5
73.0
73.5
73.6
72.6
73.0
68.7
69.1
80.1
S u in in e r S,u in in e r
RFC
NOx
1273
1277
1218
1228
1221
1214
1202
1207
CG
1326
1331
1334
1333
1327
1325
1253
1250
1340
Winter
RFC
Exhaust
90.5
90.8
90.0
89.8
89.7
89.5
88.0
88.0
Winter
CG
toxics
110.2
110.9
110.4
112.4
110.7
109.6
104.4
104.8
120.5
Winter
RFC
Winter
CG
NOx
1446
1456
1450
1445
1436
1427
1392
1380
1492
1502
1500
1509
1500
1483
1423
1410
1540
Table 1
137
-------
Complex
CG in
Summer CG
Year
1998
1999
2000
2001
2002
2003
2004
2005
Phase II RFC
Averaged
Standards
VOC (VOC
Region 2
Model)
5.8
6.2
5.8
6.6
6.2
5.7
7.9
8.6
Summer RFG
VOC (VOC
Region 2
Model and
data)
16.2
16.4
27.8
27.8
27.5
27.5
27.5
27.4
27.4
VOC region 2
Summer CG
Total toxics
(VOC Region
2 Model)
11.0
10.5
9.9
9.5
11.0
10.4
15.4
14.6
Summer RFG Summer CG
Total toxics
(VOC Region
2 Model and
data)
28.9
28.7
33.5
32.4
32.8
32.6
32.2
31.4
21.5
Annual
Standard
NOx
1.1
0.6
0.5
0.5
0.9
1.1
6.5
6.7
Summer RFG
NOx (all data-
see note)
5.0
4.7
9.1
8.3
8.9
9.4
10.3
9.9
6.8
Table 2
CG
in
Winter CG RFG
Year
1998
1999
2000
2001
2002
2003
2004
2005
Phase II RFG Averaged
Standards
8.6
8.0
8.4
6.8
8.2
9.1
13.4
13.1
Toxics
25.0
24.6
25.3
25.5
25.6
25.7
27.0
27.0
21.5
Annual Standard
Winter CG
3.1
2.5
2.6
2.0
2.6
3.7
7.6
8.4
Winter IRii-G
NOx
6.1
5.5
5.9
6.1
6.7
7.3
9.6
10.4
1.5
Table 3
138
-------
American Jobs Creation Act of 2004, Pub. L. No. 108-357,301,104 Stat 1418 (2004)
ASTM D 1319. Standard test method for hydrocarbon types in liquid petroleum products by fluorescent
indicator adsorption. ASTM International
ASTM D 4814. Standard specification for automotive spark-ignition engine fuel. ASTM International
ASTM D 5769. Standard test method for determination of benzene, toluene, and total aromatics in finished
by gas chromatography/mass spectrometry. ASTM International
Blue Ribbon Panel on Oxygenates in Gasoline (1999, September). Achieving clean air and clean water.
Brzezinski, D. (2001, April). Estimating emission effects of RFC gasoline in Mobiles. EPA Report: EPA420-R-
01-028, M6.FUL.005, Retrieved April 15, 2006, from
http://www.epa.gov/otaq/models/mobile6/r01028.pdf
Clean Air Act of 1990, Pub. L. No. 101-549,104 Stat. 2399 (1990)
Energy Information Administration. (2005, June). Petroleum supply annual 2004, volume 1. Retrieved
June 20, 2006 from
http://www.eia.doe.QOv/oil gas/petroleum/data publications/petroleum supply annual/psa volum
el/psa_volumel.html
Energy Policy Act of 2005, Pub. L. No. 109-58,119 Stat 594 (2005)
Regulation of Fuels and Fuel Additives (2004, July). Code of Federal Regulations Title 40, Pt.80
U.S. Environmental Protection Agency (2007, AprilJ Regulatory impact analysis: renewable fuel standard
program. EPA Report EPA420-R-07-004, Retrieved May, 2007 from
http://epa.gov/otaq/renewablefuels/420r07004.pdf
U.S. Environmental Protection Agency. (2005, November). State winter oxygenated fuel program
requirements for attainment or maintenance of CO NAAQS. EPA Report: EPA420-B-05-013,
Retrieved May 15, 2006, from http://www.epa.gov/otaq/regs/fuels/420b05013.pdf
U.S. Environmental Protection Agency. (2003, July). RFC questions and answers consolidated by topic
(Section 2: sampling and testing p.13). EPA Report: EPA420-R-03-039. Retrieved May 15, 2006,
from http://www.epa.gov/otaq/rfg_qa.htm
U.S. Environmental Protection Agency. (1995, August 22). Waiver requests under section 211(f) of the
Clean Air Act. Retrieved May 31, 2006, from
http://www.epa.gov/otaq/regs/fuels/additive/waiver.pdf
139
-------
APPENDICES
140
-------
Appendix to Sulfur Chapter
RFG Sulfur by Gasoline Volume
Sulfur Volume Distribution Trend-Summer RFG
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 1
Summer RFG Sulfur Content (ppm) by Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
(gal):
1997
0
29
52
74
94
111
131
150
178
204
237
265
288
313
343
387
446
541
641
793
1204
12,436,241,790
1998
1
35
61
79
96
113
128
145
162
176
192
206
222
237
255
277
297
321
359
410
499
12,832,805,667
1999
4
42
64
82
103
121
139
153
167
180
192
207
220
237
257
278
300
328
360
413
500
12,998,841,841
2000
1
25
39
51
64
74
84
94
102
112
121
130
141
149
158
167
179
191
211
242
479
12,983,127,318
2001
1
30
42
51
61
71
82
90
99
109
120
129
140
149
158
169
179
195
218
253
488
13,196,166,160
2002
0
26
37
45
54
62
72
80
90
99
110
119
130
140
151
163
178
197
224
268
472
13,835,809,610
2003
1
18
27
37
47
55
62
71
79
89
98
107
116
127
139
150
163
179
201
239
469
13,583,183,155
2004
0
10
17
24
30
37
43
48
54
60
67
74
83
91
100
110
121
132
153
184
331
14,232,658,149
2005
0
6
16
22
28
33
39
44
49
52
57
62
68
75
84
93
105
120
136
164
313
14,083,260,572
141
-------
RFC Sulfur by Gasoline Volume (continued):
Sulfur Volume Distribution Trend-Winter RFC
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 2
Winter RFC Sulfur Content (ppm) by Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
faall:
1997
0
29
51
69
86
101
119
137
160
179
199
219
240
267
294
328
373
440
548
717
2912
14,905,356,077
1998
1
33
53
72
89
106
122
139
153
168
183
197
218
239
262
290
321
353
381
426
499
15,062,572,102
1999
1
36
58
81
104
121
139
155
169
183
197
214
232
249
273
297
328
358
395
429
511
15,079,905,134
2000
1
33
52
69
84
99
112
129
145
163
182
201
220
240
259
281
310
343
384
430
496
15,829,077,693
2001
1
29
45
61
78
91
105
117
131
144
159
174
192
213
240
266
290
323
361
417
500
15,727,561,297
2002
0
29
44
59
77
97
111
123
133
145
159
174
187
209
233
259
290
321
359
404
499
16,429,804,256
2003
1
23
40
54
65
79
94
108
119
127
138
150
163
179
197
220
249
287
334
390
500
16,669,506,697
2004
0
14
22
29
36
43
51
60
69
79
90
101
111
125
136
146
159
173
193
223
337
17,188,867,639
2005
0
11
19
25
31
37
43
49
54
60
66
72
79
88
96
107
119
134
155
192
371
18,044,832,058
142
-------
Percentile Chart of Sulfur by Volume-2005 Summer RFC
(from Batch Reports)
150 200
Sulfur (ppm)
Figure 3
70%
60%
50%
40%
30%
20%
0%
Percentile Chart of Sulfur by Volume-2005 Winter RFC
(from Batch Reports)
150 20D 250
Sulfur (ppm)
Figure 4
143
-------
RFC Sulfur by Grade
Average Sulfur Content of Summer RFC Sold at Retail Stations-By Grade
1
D Avg Regular
I Avg. Mid
a .Avg Premium
1998 19B9 2DOD 2001 2002 2003 2004 2005
Figure 5
Average Sulfur Content of Winter RFG Sold at Retail Stations-By Grade
Sulfur (ppm)
250 -
150 -
100 -
50 -
D Avg. Regular
Avg. Mid
EH Avg Premium
1998 1
227
203
152
--,
999
236
207
42
-
-
2
:
1
"|
JUU 2
11 2
90 1
36 1
"|
J0 1 2
03 1
81 1
22 1
I
J02 2
99
74
28
--,
003
B1
72
05
1
I
2 004
113
95
58
Year
\
1
2005
90
78
49
Figure 6
144
-------
By
S ..,(, , », , ' i
Grade Season Data
PRM s Average 196 151 149 78 85 84 66 55 43
Volume 3,437,323,049 3,651,099,126 3,537,723,033 2,788,523,817 2,765,064,985 2,919,724,883 2,735,295,696 2,385,535,898 2,143,973,091
w Average 163 148 150 139 122 114 94 57 44
Volume 3,875,280,201 4,137,427,813 3,836,358,338 3,419,995,241 3,299,267,883 3,387,457,520 3,103,333,727 2,945,338,123 2,583,694,635
REG s Average 323 222 226 140 138 134 121 85 74
Volume 8,933,744,061 9,007,552,063 9,320,974,363 10,122,448,765 10,373,861,808 10,879,031,865 10,824,904,975 11,783,350,501 11,881,795,879
w Average 278 224 237 217 202 219 180 110 86
Volume 10,810,119,364 10,661,459,665 10,902,774,956 12,246,078,983 12,361,561,882 13,031,122,001 13,476,711,626 14,159,160,294 15,422,172,055
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific
estimates based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting
system may also differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-
grade gasoline is often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume
reported and included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis
excludes some gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as
regular and premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline
marketed as mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if
they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
145
-------
RFC SulfurGeographic
Change in Sulfur Content from 2000 to 2005-Summer RFC Surveys
Manchester. NH (75.52)
Chicago-Lake Co., IL, Gary, IN (102.76)
St Louis. MO (136.107)
Portsmouth-Dover, NH (35,54)
Milwaukee-Racine, Wl (38,57)
Susses County, DE (134,96)
NY-NJ-Long is.-CT (114,61)
Praia.-Wilm DE-Trenton, NJ (130,76)
Boston-Worcester, MA (111,50)
Baltimore, MD (148,82)
Hartford, CT (125,59)
Warren County. NJ (122,55)
Poughkeepsie, NY (117,49)
Norfolk-Virginia Beach, VA (162,91)
Washington, D.C.-area (159,85)
Atlantic City, NJ (130,55)
Dallas-Fort Worth, TX (157,82)
Rhode Island (131,54)
Springfield, MA (135,55)
Richmond, VA (150,78)
Houston-Galveston, TX (151.63)
Louisville, KY (112,16)
Covington, KY (116,10)
'. !i.-i(i'|!> in Snlfni ppm
Change in Sulfur Content from 2000 to 2005-Winter RFG Surveys
Warren County, NJ (125,97)
Louisville, KY (153,102)
NY-NJ-Long Is.-CT (141,86)
Poughkeepsie, NY (167,93)
Hartford, CT (163,87)
Manchester, NH (121,39)
Boston-Worcester, MA (140,52)
Phila.-Wilm, DE-Trenton, NJ (174,74)
Sussex County, DE (225,123)
Baltimore, MD (211,106)
Portsmouth-Dover, NH (148,40)
Springfield, MA (181.68)
Covington. KY (135,18)
Rhode Island (193.56)
St. Louis. MO (212,72)
Richmond, VA (250,107)
Washington, D.C.-area (230,85)
Norfolk-Virginia Beach, VA (244,94)
Houston-Galveston, TX (219,68)
Atlantic City. NJ (207,58)
Chicago-Lake Co., IL, Gary, IN (249,96)
Dallas-Fort Worth, TX (245,81)
Milwaukee-Racine, Wl (265,85)
-200 -180 -160 -140 -120 -100 -80 -BO -40
Change in Siilfin ppm
Figure 7
Figure 8
146
-------
Conventional Gasoline
CG Sulfur by Gasoline Volume
Sulfur Volume Distribution Trend-Summer CG
(from Batch Reports-Excluding Blendstocks)
2000 2001 2002
Reporting Year
Figure 9
Summer CG Sulfur Content (ppm) by Volume (from Batch Reports excluding Blendstocks)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
(gal):
1997
0
19
41
64
89
118
150
178
203
233
264
290
318
349
385
423
487
590
725
849
1081
39,315,348,774
1998
0
27
49
72
100
122
145
172
204
231
254
279
308
335
368
402
449
527
631
791
1084
39,013,838,890
1999
0
23
44
66
92
116
146
179
211
237
264
289
319
358
393
433
476
552
669
836
1060
37,130,899,033
2000
0
30
54
75
97
124
149
177
203
232
261
284
311
341
379
422
490
604
724
859
1081
36,366,550,602
2001
0
27
47
69
92
113
135
156
178
204
232
260
289
318
353
405
463
545
680
840
1045
38,566,421,725
2002
0
21
43
64
85
105
130
156
182
207
232
261
294
326
367
408
466
529
637
810
1089
40,811,809,360
2003
0
15
30
47
63
85
117
150
179
210
240
276
306
344
379
427
484
565
678
832
1096
43,223,878,447
2004
0
8
15
21
27
36
46
57
68
80
94
113
129
146
161
178
198
219
239
269
449
43,512,526,323
2005
0
8
17
22
26
32
38
48
57
66
76
88
100
113
132
153
176
204
233
269
478
42,196,228,805
147
-------
CG Sulfur by Gasoline Volume (continued);
Sulfur Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2000 2001 2002
Reporting Year
Figure 10
Winter CG Sulfur Content
by Volume (from Batch Reports excluding Blendstocks)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
faall:
1997
0
20
42
65
89
112
136
164
194
222
253
281
311
349
388
431
489
579
704
849
1071
44,568,243,930
1998
0
14
38
60
80
102
123
148
172
200
229
258
289
316
353
392
438
511
617
784
1083
46,189,100,491
1999
0
19
43
66
87
113
138
163
187
214
238
270
302
339
381
429
485
565
660
800
1057
47,577,145,369
2000
0
19
40
60
80
100
123
149
176
202
226
253
284
311
353
399
449
515
644
811
1086
47,958,091,470
2001
0
18
39
59
80
105
134
158
178
202
227
254
286
323
364
404
457
524
617
789
1083
48,807,540,
2002
0
20
40
60
79
101
128
156
182
205
227
255
279
313
350
403
458
537
650
811
1079
198 49,661,093,570
2003
0
15
26
40
57
77
98
123
146
172
199
221
246
275
309
345
395
465
586
767
1010
47,455,096,094
2004
0
9
17
24
30
36
46
58
72
86
100
115
133
150
168
185
204
222
244
269
449
47,651,820,292
2005
0
8
15
20
25
29
34
41
49
56
64
75
87
101
118
141
168
199
232
259
478
48,700,780,453
148
-------
Percentile Chart of Sulfur content By Volume-2005 Summer CG
(from Batch Reports-Excluding Blendstocks)
Figure 11
Percentile Chart of Sulfur content By Volume-2005 Winter CG
(from Batch Reports-Excluding Blendstocks)
100%
90%
B0%
70%
» 60%
_3
§ 50%
*o
£ 40%
30%
20%
10%
100
200 300
Sulfur (ppm)
400
500
BOO
Figure 12
149
-------
CG
CG
Grade Season Data
PRM s Average 134 120 118 114 100 85 97 51 51
Volume 6,908,485,473 6,600,159,963 6,231,165,137 4,955,635,124 4,996,778,324 5,177,600,571 5,143,196,797 4,635,507,910 4,138,979,931
w Average 117 103 110 108 101 98 81 50 44
Volume 8,304,496,529 8,193,379,471 7,668,490,252 6,627,924,003 6,554,918,975 6,568,341,755 5,507,050,392 5,077,869,493 4,508,926,779
REG s Average 368 345 361 357 337 332 334 119 107
Volume 28,005,415,849 28,380,712,680 26,617,451,271 27,335,453,049 28,964,605,205 30,992,801,780 33,051,938,207 34,605,072,844 33,851,903,439
w Average 367 330 346 329 330 330 283 123 100
Volume 32,127,505,321 32,117,523,489 33,273,916,900 35,151,238,087 35,939,854,456 36,931,713,041 35,948,070,047 37,501,622,970 39,121,167,000
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific
estimates based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting
system may also differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-
grade gasoline is often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume
reported and included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis
excludes some gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as
regular and premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline
marketed as mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if
they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
150
-------
Appendix to RVP Chapter
RFC RVP by Gasoline Volume
RVP Volume Distribution Trend-Summer RFG
{from Batch Reports)
2000 2DD1 2002 2003 2004 2005
Reporting Year
Figure 1
Summer RFG RVP Content by Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
(gal):
1997
6.03
6.79
6.90
6.99
7.05
7.10
7.18
7.31
7.49
7.68
7.76
7.82
7.88
7.92
7.96
8.00
8.04
8.10
8.16
8.23
9.22
12,448,415,575
1998
5.95
6.79
6.90
6.99
7.05
7.11
7.19
7.31
7.57
7.69
7.77
7.82
7.87
7.91
7.95
7.99
8.03
8.08
8.13
8.20
8.41
12,832,964,637
1999
6.18
6.80
6.94
7.00
7.05
7.11
7.18
7.26
7.45
7.66
7.76
7.84
7.89
7.94
7.98
8.00
8.04
8.08
8.14
8.20
8.65
12,998,841,841
2000
6.34
6.43
6.48
6.53
6.56
6.59
6.63
6.66
6.69
6.72
6.73
6.76
6.79
6.83
6.86
6.91
6.94
7.00
7.07
7.21
7.75
12,983,168,478
2001
6.19
6.44
6.49
6.53
6.57
6.60
6.64
6.66
6.70
6.72
6.74
6.77
6.81
6.84
6.88
6.93
6.98
7.05
7.11
7.24
7.75
13,222,633,468
2002
6.19
6.43
6.49
6.53
6.57
6.60
6.65
6.68
6.72
6.73
6.76
6.79
6.82
6.85
6.89
6.94
6.99
7.05
7.11
7.21
7.76
13,847,971,634
2003
6.35
6.47
6.53
6.57
6.62
6.65
6.68
6.71
6.73
6.76
6.79
6.84
6.86
6.89
6.92
6.97
7.02
7.07
7.13
7.23
7.69
13,584,860,845
2004
6.34
6.47
6.53
6.59
6.65
6.69
6.72
6.75
6.78
6.82
6.85
6.88
6.91
6.94
6.98
7.02
7.06
7.11
7.18
7.29
8.96
14,232,658,149
2005
6.33
6.49
6.57
6.63
6.69
6.73
6.76
6.79
6.83
6.86
6.89
6.92
6.96
6.99
7.02
7.06
7.11
7.17
7.24
7.32
9.95
14,083,382,582
151
-------
RFC RVP by Gasoline Volume (continued):
(Cumulative Distribution for Latest Year Data)
Percentile Chart of RVP by Volume-2005 Summer RFG
(from Batch Reports)
|
_5
*o
z:
/
/
*
+
*
/
*
«
/
*
I
/
7
6 6.5 7 7.5 8 8.5 9 9.5 10 10.5
RVP (psi)
Figure 2
RVP by Grade
Average RVP of Summer RFG Sold at Retail Stations-By Grade
Year
Figure 3
152
-------
ty
and by C
-
Grade
PRM
REG
Season Data
s Average
Volume
s Average
Volume
1§§7
7.59
3,416,852,331
7.61
8,915,975,289
7.56
3,651,258,096
7.61
9,007,552,063
7.60
3,537,723,033
7.59
9,320,974,363
6.73
2,788,564,977
6.79
10,122,448,765
6.73
2,772,816,841
6.81
10,392,283,260
6.71
2,931,886,907
6.82
10,879,031,865
6.74
2,736,973,386
6.85
10,824,904,975
6.77
2,385,535,898
6.88
11,783,350,501
6.77
2,144,095,101
6.94
11,881,795,879
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates based
on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also differ from
actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is often blended from
regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported and included in these tables as
regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes some gasoline and blendstock that was
included in aggregate average estimates. EPA has presented averages for batches labeled as regular and premium gasoline only; excluding batches
labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as mid-grade), mix of grades, or without a grade
label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed above
these volumes are not expected to represent the actual volumes by grade of retail gasoline.
153
-------
RFC RVP-Geographic
Change in RVP from 2C
Portsmouth-Dover, NH (6 66.6.96)
Manchester, NH (672.701)
Louisville, KY (6.67,6.94)
Houston-Galveston, TK (6.67,6 94)
Dallas-Fort Worth, TX (6.80,7.04)
Milwaukee-Racine, Wl (6.75,6.99)
Chicago-Lake Co . IL, Gary, IN (6 71 ,6 94)
Covington.KY (682,703)
_ Warren County, NJ (6.77,6.96)
H Boston-Worcester, MA (6.81 ,6.95)
J Sussex County, DE (643,657)
CL
°e
£ Poughkeepsie, NY (6 82,6.93)
Baltimore, MD (6.80,6.90)
Washington, D.C.-area (6 76.6 86)
Richmond, VA (6.82,6.91)
Phila.-Wilm, DE-Tranton, NJ (6.76,6.61)
Hartford, CT (6.88,6.92)
Atlantic Ctty.NJ (6.77.6.91)
St. Louis, MO (6.72,6.73)
Rhode Island (6 86,6 87)
Springfield, MA (6.90,6 83)
-0
00 to 2005-Summer RFC Surveys
1
:
]
I
^^^^
=
3
=ZJ
^H
D
1
1 0 0.1 0.2 0.3 0
Change in RVP psi
4
Figure 4
154
-------
Conventional Gasoline
CG RVP by Gasoline Volume
RVP Volume Distribution Trend-Summer CG
(from Batch Reports-Excluding Blendstocks)
2001 2002
Reporting Year
Figure 5
Summer CG RVP (psi) by Volume (from Batch Reports excluding Blendstocks)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
(qal):
1998
5.89
6.94
7.39
7.52
7.62
7.73
8.04
8.27
8.38
8.46
8.52
8.58
8.62
8.66
8.71
8.75
8.79
8.84
8.89
8.94
12.10
39,091,550,524
1999
5.83
6.99
7.39
7.50
7.59
7.67
7.76
8.12
8.34
8.44
8.51
8.58
8.62
8.66
8.71
8.74
8.79
8.83
8.88
8.94
11.95
37,194,647,473
2000
5.85
6.96
7.39
7.52
7.60
7.68
7.76
8.05
8.25
8.39
8.47
8.54
8.61
8.65
8.69
8.74
8.79
8.82
8.88
8.95
12.09
36,395,981,388
2001
5.80
6.88
7.36
7.51
7.59
7.67
7.76
8.05
8.24
8.37
8.49
8.55
8.61
8.65
8.70
8.75
8.79
8.84
8.89
8.95
12.06
38,566,835,643
2002
5.96
6.89
7.37
7.52
7.60
7.68
7.79
8.07
8.29
8.40
8.49
8.57
8.62
8.67
8.71
8.75
8.79
8.85
8.90
8.95
12.10
40,819,234,602
2003
5.83
6.89
7.39
7.52
7.62
7.69
7.83
8.17
8.37
8.47
8.54
8.61
8.64
8.69
8.73
8.76
8.81
8.85
8.90
8.95
12.08
43,243,575,045
2004
5.80
6.87
7.36
7.51
7.61
7.69
7.87
8.19
8.36
8.46
8.54
8.60
8.64
8.68
8.72
8.76
8.80
8.85
8.89
8.95
12.07
43,509,104,037
2005
5.80
6.82
7.37
7.53
7.63
7.72
7.92
8.21
8.36
8.46
8.55
8.59
8.63
8.68
8.72
8.76
8.81
8.85
8.90
8.95
12.10
42,196,228,805
155
-------
CG RVP by Gasoline Volume (continued):
RVP Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2001 2002
Reporting Year
Figure 6
Winter CG RVP (psi) by Volume (from Batch Reports excluding Blendstocks)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
5.54
8.56
9.49
10.30
10.82
11.10
11.32
11.59
11.96
12.20
12.39
12.55
12.71
12.89
13.10
13.29
13.46
13.94
14.29
14.64
16.30
46,061,502,458
1999
5.57
8.39
8.89
10.13
10.71
11.04
11.26
11.46
11.87
12.14
12.35
12.51
12.69
12.87
13.07
13.29
13.44
13.93
14.31
14.65
16.00
45,842,624,362
2000
5.40
8.26
8.81
9.91
10.71
11.04
11.23
11.45
11.89
12.19
12.40
12.55
12.71
12.89
13.09
13.29
13.45
13.79
14.25
14.59
15.94
41,249,911,572
2001
5.48
8.14
8.74
9.86
10.52
10.90
11.14
11.35
11.61
11.94
12.20
12.41
12.58
12.77
12.99
13.22
13.41
13.70
14.24
14.59
16.20
48,563,137,130
2002
5.39
8.33
8.82
9.97
10.66
11.00
11.21
11.42
11.76
12.05
12.26
12.44
12.62
12.83
13.04
13.28
13.44
13.76
14.26
14.61
15.78
48,898,338,201
2003
5.35
8.69
9.81
10.51
10.88
11.12
11.29
11.49
11.87
12.09
12.27
12.41
12.58
12.76
12.99
13.23
13.40
13.78
14.26
14.61
16.00
41,635,803,399
2004
5.32
8.76
10.07
10.63
10.93
11.14
11.31
11.53
11.88
12.08
12.23
12.39
12.53
12.69
12.89
13.10
13.30
13.62
14.25
14.64
15.50
39,996,766,942
2005
5.40
8.75
9.81
10.43
10.79
11.01
11.23
11.47
11.79
12.04
12.21
12.36
12.50
12.66
12.83
13.02
13.24
13.48
14.07
14.55
15.35
38,557,421,343
156
-------
Percentile Chart of RVP By Volume-2005 Summer CG
(from Batch Reports-Excluding Blendstocks)
5.5 B.D 6.5 7.0 7.5
85 90 9.5 10.0 10.5 111
RVP (psi)
115 12.0 125
Figure 7
Percentile Chart of RVP By Volume-2005 Winter CG
(from Batch Reports-Excluding Blendstocks)
7.8 9.0 10.0
Figure 8
13.5 16.0
157
-------
CG RVP by
CG RYP and Vol1
Grade Season Data
PRM
REG
w
Average
Volume
Average
Volume
Average
Volume
Average
Volume
8.17 8.16 8.07 8.08 8.10 8.17 8.15 8.14
6,635,803,305 6,252,416,423 4,957,386,104 4,996,778,324 5,181,939,065 5,149,486,549 4,635,507,910 4,138,979,931
12.11 11.98 11.94 11.82 11.93 12.01 11.94 12.00
8,209,902,634 7,399,170,984 5,413,808,077 6,555,892,276 6,542,422,368 5,021,592,048 4,297,707,397 3,716,904,849
8.33 8.29 8.25 8.25 8.27 8.30 8.30 8.30
28,418,412,132 26,656,367,421 27,360,707,607 28,965,019,123 30,995,888,528 33,053,399,345 34,601,650,558 33,851,903,439
12.06 11.97 11.89 11.80 11.87 11.97 12.02 11.94
31,980,799,123 31,791,685,337 29,648,094,807 35,690,792,290 36,194,877,059 31,768,671,300 31,555,556,718 32,128,135,107
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific
estimates based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting
system may also differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-
grade gasoline is often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume
reported and included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis
excludes some gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as
regular and premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline
marketed as mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if
they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
158
-------
Appendix to Oxygenates and Oxygen Chapter
RFG Oxygen Weight Percent by Gasoline Volume (Excluding California)
Oxygen Volume Distribution Trend-Summer RFG
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 1
Summer RFG Oxygen Content (wt%) by Volume (from Batch Reports)
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volums
Cqall:
Reporting
1997
1.20
1.53
1.58
1.62
1.68
1.72
1.78
1.83
1.90
1.96
2.00
2.06
2.10
2.16
2.25
2.34
2.44
2.52
2.65
3.47
3.91
12,212,323,923
Year
1998
1.50
1.62
1.67
1.70
1.74
1.78
1.82
1.87
1.91
1.98
2.02
2.06
2.09
2.13
2.17
2.22
2.32
2.46
2.64
3.47
3.96
12,566,698,539
1999
1.50
1.70
1.72
1.75
1.77
1.79
1.81
1.84
1.87
1.90
1.93
1.97
2.01
2.07
2.13
2.20
2.30
2.45
2.62
3.47
3.82
12,660,919,964
2000
1.50
1.68
1.73
1.76
1.80
1.83
1.88
1.91
1.95
2.01
2.07
2.13
2.20
2.29
2.37
2.44
2.52
2.60
3.42
3.52
3.78
12,569,342,940
2001
1.51
1.68
1.72
1.76
1.79
1.82
1.85
1.88
1.93
1.99
2.05
2.11
2.17
2.23
2.31
2.39
2.47
2.55
3.31
3.53
4.10
12,783,848,128
2002
1.43
1.66
1.74
1.78
1.82
1.86
1.89
1.94
2.00
2.05
2.10
2.16
2.21
2.27
2.35
2.43
2.51
2.58
3.40
3.49
4.00
13,409,581,178
2003
1.44
1.63
1.71
1.78
1.84
1.90
1.94
1.99
2.05
2.10
2.15
2.23
2.29
2.36
2.42
2.48
2.54
2.65
3.48
3.57
4.02
13,185,376,501
2004
1.45
1.69
1.76
1.82
1.88
1.94
2.00
2.08
2.17
2.25
2.35
2.43
2.50
2.58
3.31
3.45
3.49
3.53
3.57
3.64
3.91
13,746,792,069
2005
1.44
1.66
1.72
1.75
1.78
1.82
1.87
1.91
1.98
2.07
2.19
2.33
2.43
2.51
2.67
3.44
3.51
3.54
3.59
3.66
3.95
13,464,126,854
159
-------
RFC Oxygen Weight Percent by Gasoline Volume (Continued):
Oxygen Volume Distribution Trend-Winter RFC
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 2
Winter RFC Oxygen
Content (wt%) by Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volums
(qaO:
1997
1.44
1.59
1.64
1.68
1.72
1.76
1.80
1.86
1.93
1.99
2.04
2.09
2.14
2.21
2.31
2.56
2.70
2.77
2.87
3.53
3.99
14,520,852,742
1998
1.36
1.66
1.70
1.74
1.77
1.80
1.83
1.86
1.90
1.94
2.00
2.05
2.13
2.21
2.32
2.61
2.72
2.78
2.89
3.53
4.02
14,621,530,037
1999
1.50
1.71
1.74
1.76
1.78
1.80
1.82
1.85
1.88
1.91
1.95
1.99
2.03
2.09
2.16
2.26
2.40
2.62
2.82
3.54
3.85
14,604,242,854
2000
1.50
1.68
1.73
1.76
1.79
1.81
1.83
1.85
1.88
1.90
1.93
1.97
2.00
2.05
2.10
2.16
2.26
2.43
2.64
3.56
3.81
15,291,171,680
2001
1.43
1.69
1.72
1.75
1.78
1.80
1.82
1.85
1.87
1.89
1.91
1.94
1.98
2.02
2.08
2.14
2.23
2.39
2.58
3.57
4.02
15,116,796,877
2002
1.45
1.65
1.70
1.73
1.76
1.78
1.80
1.83
1.85
1.87
1.90
1.93
1.97
2.01
2.07
2.13
2.21
2.36
2.66
3.57
3.94
15,826,024,638
2003
1.25
1.59
1.64
1.69
1.73
1.76
1.79
1.81
1.84
1.86
1.90
1.93
1.97
2.02
2.09
2.19
2.35
2.57
3.53
3.62
4.00
15,771,471,139
2004
1.37
1.64
1.70
1.75
1.78
1.82
1.85
1.88
1.91
1.94
1.99
2.07
2.16
2.29
2.44
2.67
3.53
3.58
3.64
3.72
4.04
15,875,951,167
2005
1.46
1.62
1.68
1.72
1.75
1.78
1.81
1.84
1.88
1.90
1.94
1.98
2.06
2.21
2.39
3.40
3.53
3.58
3.65
3.72
3.97
16,565,687,773
160
-------
Percentile Chart of Oxygen Content By Volume-2005 Summer RFC
(from Batch Reports)
1.50 2.00 2.50
Oxygen (wt%)
4.50
Figure 3
0.00
Percentile Chart of Oxygen Content By Volume-2005 Winter RFC
(from Batch Reports)
1.50 2.00 2.50
Oxygen (wt%)
3.00
3.50
4.00
4.50
Figure 4
161
-------
' M"BS ty
f
Year
Volume
%tile
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
50%
55%
70%
75%
80%
85%
90%
95%
100%
Volume
faall:
1997
4.95
5.78
8.21
8.52
8.72
8.99
9.27
9.58
9.94
10.44
10.90
11.30
11.59
11.95
12.57
13.18
13.81
15.71
11,977,238,567
1998
5.11
6.23
8.05
8.95
9.13
9.39
9.67
9.99
10.35
10.78
11.11
11.35
11.57
11.83
12.08
12.64
13.75
16.00
12,553,324,899
4.42
5.86
7.30
8.87
9.36
9.57
9.70
9.83
9.99
10.17
10.39
10.59
10.89
11.29
11.75
12.36
13.52
16.21
12,608,377,199
2000
5.66
6.72
8.17
8.73
9.05
9.36
9.60
9.82
10.10
10.40
10.68
11.11
11.56
12.08
12.92
13.44
14.02
15.17
12,574,035,936
2001
5.18
7.36
8.41
8.80
9.11
9.34
9.58
9.80
10.09
10.35
10.74
11.20
11.64
12.11
12.68
13.25
13.74
15.86
12,783,848,128
0.20
6.55
8.34
8.88
9.17
9.51
9.77
10.04
10.34
10.65
11.08
11.47
11.83
12.24
12.76
13.35
13.91
15.79
13,409,581,178
2003
0.06
5.97
7.91
8.69
9.01
9.37
9.79
10.16
10.55
10.95
11.39
11.74
12.19
12.64
13.04
13.52
13.98
15.62
13,168,085,773
2004
0.05
0.10
6.30
8.52
9.13
9.53
9.92
10.26
10.77
11.25
11.81
12.45
13.05
13.55
14.00
15.12
13,749,448,485
2005
0.04
0.19
7.68
8.65
9.03
9.40
9.62
9.78
10.03
10.30
10.65
11.35
12.20
13.07
13.72
15.24
13,464,757,064
Rl ..
bf
Year
Volume
%tile
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
faall:
1997
0.21
5.06
7.18
8.34
8.64
8.89
9.09
9.33
9.63
10.14
10.73
11.21
11.63
12.22
13.71
14.52
14.94
16.93
14,400,124,642
1998
0.18
5.24
6.94
8.68
9.04
9.30
9.52
9.74
9.98
10.24
10.58
11.09
11.70
12.16
13.49
14.58
14.98
16.86
14,621,530,037
1999
0.47
5.42
6.89
8.72
9.18
9.38
9.53
9.67
9.80
9.97
10.18
10.42
10.75
11.16
11.69
12.66
14.34
16.10
14,594,054,317
2000
0.42
5.55
7.22
8.40
8.88
9.20
9.40
9.60
9.77
9.91
10.07
10.24
10.49
10.83
11.31
11.89
12.92
15.63
15,275,786,534
2001
0.65
5.59
7.99
8.59
8.98
9.24
9.44
9.60
9.76
9.92
10.06
10.24
10.44
10.74
11.18
11.63
12.44
15.30
15,122,255,995
2002
0.07
5.23
7.67
8.56
8.84
9.10
9.26
9.41
9.61
9.77
9.95
10.15
10.36
10.67
11.09
11.55
12.30
15.62
15,825,705,992
2003
0.14
7.36
8.36
8.66
8.89
9.15
9.37
9.53
9.70
9.87
10.04
10.28
10.56
11.00
11.52
12.41
15.02
15,771,471,139
2004
0.05
0.28
7.95
8.56
8.87
9.15
9.39
9.62
9.84
10.06
10.29
10.79
11.39
12.12
13.01
15.13
15,875,951,167
2005
0.04
0.20
7.85
8.45
8.78
9.03
9.23
9.44
9.63
9.85
10.09
10.29
10.55
11.09
12.19
15.79
16,559,495,884
162
-------
100%
Percentile Chart of MTBE Content By Volume-2005 Summer RFG
(from Batch Reports)
Figure 5
30% -
20% :
10% -
Percentile Chart of MTBE Content By Volume-2005 Winter RFG
(from Batch Reports)
8 10
MTBE (v%)
12
18
Figure 6
163
-------
Volume
%tile
57%
58%
59%
70%
75%
80%
81%
82%
83%
84%
85%
86%
87%
88%
89%
90%
95%
100%
Volums
(gal):
Summer
Reporting Year
4.66
5.18 4.87
5.36 5.15
9.40 9.45
10.51 10.69
(v<
4.81
4.96 7.44
5.15 8.95
5.26 9.23
9.47 9.51
10.42 10.10
11,977,238,567 12,553,324,899 12,608,377,199 12,574,035,936
by
2001
0.09
8.59
9.11
9.56
10.85
12,783,848,128
2002
0.03
5.09
5.33
5.50
5.76
9.07
9.23
9.45
10.31
13,409,581,178
2003
0.01
4.70
5.23
5.35
9.10
9.31
9.37
9.42
9.65
10.40
13,168,085,773
2004
5.28
5.56
9.03
9.34
9.44
9.45
9.48
9.49
9.51
9.54
9.56
9.59
9.62
9.64
9.67
9.88
11.00
13,749,448,485
2005
5.38
9.32
9.48
9.50
9.54
9.56
9.58
9.60
9.61
9.63
9.67
9.70
9.72
9.92
10.60
13,464,757,064
Volume
%tile
66%
67%
68%
69%
70%
75%
76%
77%
78%
79%
80%
81%
82%
83%
84%
85%
90%
95%
100%
Volums
(gal):
Reporting Year
1997 1999
5.02 4.69
5.21 4.86 4.60
5.57 5.74 5.47
9.38 9.42 9.45
10.49 10.71 10.39
14,400,124,642 14,621,530,037 14,594,054,317
1 bf
4.52
0.02 4.81
4.92 4.53 4.90
5.55 5.39 5.75
9.48 9.53 9.50
10.24 10.52 10.46
2003
0.08
4.82
4.90
4.98
5.03
5.11
5.29
9.44
9.67
10.70
15,275,786,534 15,122,255,995 15,825,705,992 15,771,471,139
0.02
0.02
0.02
4.85
5.65
9.15
9.26
9.34
9.40
9.43
9.46
9.49
9.51
9.54
9.56
9.70
9.90
10.82
15,875,951,167
2005
0.02
0.02
0.02
0.08
4.80
9.13
9.23
9.31
9.36
9.40
9.44
9.47
9.50
9.53
9.56
9.59
9.73
9.93
10.73
16,559,495,884
164
-------
Percentile Chart of Ethanol Content by Volume-2005 Summer RFC
(from Batch Reports)
Ethanol (v%)
Figure 7
Percentile Chart of Ethanol Content By Volume-2005 Winter RFC
(from Batch Reports)
4 6
Ethanol (v%)
Figure 8
10
12
165
-------
RFC Oxygen Content by Grade
Oxygen (Wt%)
Aver<
4.00
3.50 -
3.00 -
2.50 -
2.00 -
1 50 -
1.00 -
0.50 -
0.00 -
D Avg Regular
Avg Mid
D Avg Premium
ige Oxygen Content of Summer RFC Sold at Retail Stations-By Grade
"I " Bars denote range of RFS area aver ac
ffi
1
l"lf
1995 1996
2.15 2.17
2.18 2.17
2.23 2.22
r- " -
1997 1998 1999 20
2.22 2.25 2.25 2.
2.23 2.24 2.25 2.
2.28 2.29 2.29 2.
1 T I T
-L ^
r-
1 J-
00 2001 2002 2003 200
29 2.28 2.32 2.42 2.6:
33 2.30 2.32 2.42 2.6E
37 2.35 2.31 2.40 2.7:
es
r I
1 2005
2.57
2.67
2.70
Yeni
Figure 9
Average Oxygen Content of Winter RFC Sold at Retail Stations-By Grade
6.00
Year
Figure 10
166
-------
RFC MTBE Content by Grade
JJ.
1
Ul
as
Average MTBE Content (WT%) of Summer RFG Sold at Retail Stations-By
Grade
(Excluding CA)
14.00 -
12.00 -
10.00 -
8.00 -
6.00 -
4.00 -
2.00 -
o.oo 1
D Avg Regular 8
Avg Mid 9
D Avg Premium 1
r-i
r
395 1996 19
56 7.8B 8.
54 8.6B 9
30 9.92 10
n
r " r *
37 199B 999
45 8.48 8.33
J7 9.08 8.77
27 10. 05 9.50
r
I
2000 2001
8.46 8.45
9.05 B.95
9.BB 9.96
I
2002 2003
8.38 8.50
8.66 8.65
9.41 9.50
n M
I
2004 2005
7.04 6.37
7.00 6.37
7.35 766
Year
Figure 11
I.
Average MTBE Content (WT%) of Winter RFC Sold at Retail Stations-By
Grade (Excluding CA)
14.00- '
12.00 -
10.00 -
8.00 -
6.00 -
4.00 -
2.00 -
1
a Avg Regular 9
Avg Mid 1
n Avg Premium 1
.,
395 1996 1E
11 814 8
.56 8.89 9
.09 10.06 10
97
25
41
33
1
1
1998
9.11
9.37
10.42
"I" Bars denote range of RFG area averages
I
19
8.
a:
82
r-
9 2000
3 B.08
3 8.20
3 8.77
l - r
2001 2002
7.92 7.B8
8.44 7.86
8.85 8.67
i
2003 20(
7.53 6.:
7.79 6.E
8.38 6.E
"1
4 2005
1 5.93
8 B.15
7 6.75
Year
Figure 12
These averages include RFG which did not contain MTBE.
167
-------
RFC Ethanol Content by Grade
Average Ethanol Content (WT%) of Summer RFC Sold at Retail Stations-By
Grade (Excluding CA)
Figure 13
Average Ethanol Content (WT%) of Winter RFC Sold at Retail Stations-By
Grade (Excluding CA)
HiiHii
Figure 14
These averages include RFC which did not contain ethanol.
168
-------
RFG
(f CA)
Average of oxygen (weight °/o)
Season Area ^: _ .
Atlantic City, NJ (1)
Baltimore, MD
Boston-Worcester, MA
Chicago-Lake Co., IL, Gary, IN
Covington, KY (3)
CT - remainder (3)
Dallas-Fort Worth, TX
Hartford, CT
Houston-Galveston, TX
Knox Co. and Lincoln Co., ME (2,3)
Lewiston-Auburn, ME (2,3)
Louisville, KY
Manchester, NH
Milwaukee-Racine, WI
Norfolk-Virginia Beach, VA
NY-NJ-Long Is.-CT
Phila.-Wilm, DE-Trenton, NJ
Phoenix, AZ (1,2)
Portland, ME (2)
Portsmouth-Dover, NH (3)
Poughkeepsie, NY (3)
Queen Anne Co. -Kent Co., MD (3)
Rhode Island
Richmond, VA
Springfield, MA
St. Louis, MO (1)
Sussex County, DE (1)
Warren County, NJ (1)
Washington, DC area
Atlantic City, NJ
Baltimore, MD
Boston-Worcester, MA
Chicago-Lake Co., IL, Gary, IN
Covington, KY
CT - remainder
Dallas-Fort Worth, TX
Hartford, CT
Houston-Galveston, TX
Knox Co. and Lincoln Co., ME
Lewiston-Auburn, ME
Louisville, KY
Manchester, NH
Milwaukee-Racine, WI
Norfolk-Virginia Beach, VA
NY-NJ-Long Is.-CT
Phila.-Wilm, DE-Trenton, NJ
Phoenix, AZ
Portland, ME
Portsmouth-Dover, NH
Poughkeepsie, NY
Queen Anne Co. -Kent Co., MD
Rhode Island
Richmond, VA
Springfield, MA
St. Louis, MO
Sussex County, DE
Warren County, NJ
Washington, DC area
Reasons for no data in certain
years:
1 No data prior to opt-in to RFG
2.17
2.07
2.18
2.59
1.98
2.09
2.13
2.17
2.11
2.05
2.08
2.61
2.10
2.08
2.09
2.23
2.11
2.10
2.06
2.26
2.08
2.17
2.09
2.13
2.65
2.30
2.10
3.13
2.03
1.97
2.20
2.00
2.25
1.98
2.11
2.95
2.01
2.61
2.26
2.20
2.11
2.45
2.28
2.25
2.20
2.13
2.11
2.71
1.98
1.87
2.12
3.36
2.02
1.91
2.03
1.87
1.98
3.42
1.88
1.99
2.00
2.07
2.08
2.06
1.85
2.12
1.92
1.81
1.83
1.98
3.53
2.04
1.87
2.06
1.83
2.04
2.91
1.85
2.53
1.85
2.12
1.91
1.95
1.86
1.94
1.87
2.16
1.94
2.15
3.44
2.01
2.02
1.95
2.30
2.08
3.42
1.92
2.11
2.09
1.89
2.10
2.08
1.95
1.95
2.12
1.97
2.04
1.93
1.96
3.56
1.91
2.17
1.92
2.19
1.95
2.52
1.91
2.64
2.01
3.56
2.11
2.41
1.96
1.88
2.00
1.96
1.96
2.08
3.46
2.70
2.01
2.07
2.04
2.04
2.22
2.36
2.11
3.45
1.97
2.03
2.13
2.18
2.15
2.02
2.08
1.95
2.08
1.99
1.90
2.06
3.68
3.32
2.03
1.97
2.06
1.97
2.26
2.16
2.10
3.53
1.87
2.70
2.04
2.26
2.15
2.52
2.05
1.91
2.09
1.91
2.14
1.99
2.01
3.54
3.05
2.01
2.02
2.00
2.27
1.96
3.51
2.03
2.02
2.12
1.98
2.03
2.07
2.00
2.08
2.22
1.98
2.04
2.03
2.14
1.95
2.10
3.69
3.14
1.94
2.10
1.99
2.40
2.10
3.55
1.96
2.19
1.95
2.12
2.25
2.08
1.98
2.18
2.69
1.89
2.36
2.01
2.23
2.06
2.17
3.51
3.06
2.03
2.13
2.11
2.27
2.18
3.49
2.07
2.03
2.20
2.21
2.04
2.07
2.13
2.12
2.20
2.01
2.17
2.11
2.08
1.98
2.14
3.70
3.02
1.99
2.12
2.04
2.24
2.04
3.65
2.00
1.99
2.06
2.24
2.12
2.20
1.94
2.22
2.71
2.00
1.98
1.98
2.32
2.01
2.10
3.52
2.99
2.07
2.10
2.08
2.25
2.17
3.44
1.98
2.06
2.29
2.27
2.06
2.10
1.99
1.96
2.07
2.26
2.14
1.96
1.94
1.98
2.11
3.65
2.99
1.99
2.06
2.06
2.25
2.15
3.51
1.98
1.91
1.94
2.12
1.91
2.00
2.04
1.94
2.03
2.54
1.98
1.94
2.29
2.06
2.09
3.50
2.91
2.10
2.08
2.12
2.15
2.05
2.18
3.49
2.04
1.99
2.26
2.18
1.99
2.20
1.93
1.98
2.01
3.05
2.29
1.99
1.96
1.92
2.05
3.64
3.34
2.05
1.93
2.01
2.04
2.90
2.06
3.61
1.98
1.87
1.90
2.05
1.90
1.83
1.86
1.95
2.11
2.84
1.96
1.98
2.31
2.07
2.08
3.51
3.57
2.08
2.21
2.13
2.20
3.50
2.13
3.50
2.12
2.11
2.31
2.12
2.01
2.02
2.12
2.11
3.54
2.25
2.18
2.09
2.05
1.96
2.05
3.66
3.72
2.76
1.97
2.37
2.08
3.71
2.08
3.69
2.01
2.08
1.94
2.04
2.74
1.93
1.95
1.97
3.46
1.87
1.90
2.01
2.20
2.12
2.14
3.53
3.59
3.53
2.21
3.54
2.13
3.56
2.06
3.51
2.24
2.91
2.21
2.15
3.56
2.10
2.09
2.12
2.16
3.56
2.09
2.15
2.08
1.94
1.98
2.04
3.67
3.90
3.59
1.95
3.60
2.03
3.72
2.01
3.65
2.08
2.96
1.93
2.15
3.68
1.87
2.03
1.99
2.20
3.61
1.80
1.91
2.05
2.02
2.03
2.07
3.61
3.64
3.57
2.01
3.59
2.05
3.60
1.96
3.58
2.07
2.87
2.07
1.98
3.58
1.93
2.10
2.05
2.40
3.66
1.82
1.94
2.00
1.89
1.88
2.03
3.69
3.70
3.64
1.99
3.66
2.06
3.68
1.92
3.69
1.93
3.09
1.88
1.92
3.65
1.84
1.89
1.88
2.38
3.58
1.77
2.07
1.94
2 No data subsequent to opt-out of RFG
3 Not sampled in certain years (smaller
area)
169
-------
Average of Oxygen (Wt%)
Season Area
Los Angeles, CA
Sacramento Metro, CA
San Diego, CA
San Joaquin, CA
Los Angeles, CA
Sacramento Metro, CA
San Diego, CA
San Joaauin, CA
by
Year
2.05 2.05
2.12 2.04
2.03 2.09
2.03
2.03
2.09
i an
dl
2
2
2
2
2
2
04
11
06
15
13
14
irea
(F
2
2
2
2
2
2
20
13
19
16
09
17
2.01
2.18
2.09
2.23
2.14
2.20
2.07
1.99
2.17
2.12
2.13
2.21
in
2.11
2.13
2.07
2.05
2.05
2.12
2.02
2.05
2.09
2.11
2.01
2.03
2.02
2.08
nia)
2.08
2.05
2.08
2.13
2.13
2.17
2.09
2.23
2.12
2.14
2.11
2.19
2.01
2.13
2.08
2.13
170
-------
RFG
Average of MTBE (wt°/o)
Season Area
Atlantic City, NJ
Baltimore, MD
Boston-Worcester, MA
Chicago-Lake Co., IL, Gary,
IN
Covington, KY
CT - remainder
Dallas-Fort Worth, TX
Hartford, CT
Houston-Galveston, TX
Knox Co. and Lincoln Co.,
ME
Lewiston-Auburn, ME
Louisville, KY
Manchester, NH
Milwaukee-Racine, WI
Norfolk-Virginia Beach, VA
NY-NJ-Long Is.-CT
Phila.-Wilm, DE-Trenton, NJ
Phoenix, AZ
Portland, ME
Portsmouth-Dover, NH
Poughkeepsie, NY
Queen Anne Co.-Kent Co.,
MD
Rhode Island
Richmond, VA
Springfield, MA
St. Louis, MO
Sussex County, DE
Warren County, NJ
Washington, DC area
Atlantic City, NJ
Baltimore, MD
Boston-Worcester, MA
Chicago-Lake Co., IL, Gary,
IN
Covington, KY
CT - remainder
Dallas-Fort Worth, TX
Hartford, CT
Houston-Galveston, TX
Knox Co. and Lincoln Co.,
ME
Lewiston-Auburn, ME
Louisville, KY
Manchester, NH
Milwaukee-Racine, WI
Norfolk-Virginia Beach, VA
NY-NJ-Long Is.-CT
Phila.-Wilm, DE-Trenton, NJ
Phoenix, AZ
Portland, ME
Portsmouth-Dover, NH
Poughkeepsie, NY
Queen Anne Co. Kent Co.,
MD
Rhode Island
Richmond, VA
Springfield, MA
St. Louis, MO
Sussex County, DE
Warren County, NJ
Washington, DC area
Year
11.12
10.43
10.14
1.89
7.67
9.23
10.40
10.79
11.42
4.80
10.37
1.49
10.07
10.70
10.68
11.84
11.06
10.90
10.04
11.37
10.25
10.61
9.63
10.47
14.06
11.72
9.60
2.22
7.45
9.10
10.58
9.94
12.13
6.34
10.24
2.46
9.78
12.88
11.16
12.00
10.79
12.89
11.49
10.30
10.74
9.52
9.78
13.18
10.67
9.02
10.37
0.60
10.14
8.93
10.29
8.84
7.87
0.38
8.99
10.43
10.42
11.28
11.21
10.22
8.94
10.82
9.17
8.99
8.21
9.13
0.60
9.98
8.89
9.82
8.56
7.69
0.23
8.69
12.95
9.01
11.07
9.54
9.09
8.42
9.08
8.86
by
11.49
9.31
10.70
0.65
9.93
10.39
9.27
8.12
10.52
0.45
9.51
11.19
10.78
9.99
11.23
10.98
10.04
9.21
11.03
9.67
10.56
9.51
9.20
0.48
9.89
10.20
9.37
7.55
10.22
0.49
9.42
13.32
9.99
0.17
11.51
12.05
9.39
9.12
9.77
9.63
9.55
10.23
0.43
5.63
10.14
9.87
10.38
10.45
11.70
7.19
11.03
0.34
9.68
10.65
11.15
11.58
11.27
10.76
10.32
9.68
10.62
9.85
9.71
9.57
0.07
3.29
9.68
9.61
9.79
10.02
11.76
7.93
10.07
0.01
9.47
13.93
10.68
11.73
11.29
12.50
10.12
9.81
9.95
9.58
11.44
9.51
9.52
0.00
2.52
10.31
9.08
10.26
7.81
10.24
0.00
9.96
10.67
11.28
10.57
10.43
9.96
9.72
9.68
7.91
8.05
10.67
9.59
10.87
9.50
8.96
0.00
2.91
9.43
8.85
10.33
7.83
10.28
0.00
9.92
11.23
10.24
9.98
11.01
9.25
9.53
9.02
5.46
8.90
12.48
9.87
11.47
10.12
10.59
0.01
3.49
10.16
9.21
11.01
8.16
11.11
0.00
10.33
10.26
11.50
11.82
10.21
9.53
10.53
9.90
9.07
9.11
10.94
10.43
9.87
9.58
9.80
0.00
3.67
9.89
9.01
10.52
7.26
10.05
0.00
9.91
9.87
10.34
10.96
9.78
9.81
9.43
9.39
5.59
8.12
10.32
9.71
12.31
9.73
9.96
0.00
3.44
10.38
9.60
10.42
7.40
11.11
0.00
9.62
10.54
12.03
11.75
10.54
9.51
9.65
9.93
9.80
8.85
9.74
9.84
10.07
9.39
9.54
0.00
3.63
9.89
9.36
10.21
7.29
11.17
0.00
9.91
9.85
9.87
11.21
9.54
9.48
9.62
9.61
9.59
4.90
8.04
9.73
12.27
10.30
10.36
0.01
3.29
9.22
10.48
9.27
10.95
7.65
11.23
0.00
10.33
10.26
12.06
11.40
10.13
10.54
9.93
10.12
9.34
0.22
10.19
10.13
10.23
9.61
9.98
0.01
1.81
8.85
9.83
8.85
10.39
4.10
10.68
0.01
9.97
9.68
9.87
10.98
9.20
8.91
9.31
9.83
9.27
2.91
8.80
9.96
12.31
10.33
10.12
0.02
0.00
9.78
11.26
9.24
11.39
0.00
11.17
0.01
10.47
10.46
12.34
11.14
9.59
9.99
10.74
9.74
0.27
11.51
10.73
10.35
10.88
9.84
10.40
0.01
0.01
5.21
10.11
7.33
10.62
0.01
10.77
0.00
10.19
8.55
10.18
10.67
5.14
9.53
9.70
8.93
0.46
10.05
9.95
10.12
11.77
10.76
10.88
0.03
0.00
0.06
11.12
0.06
11.13
0.00
10.47
0.01
11.74
4.89
11.85
11.36
0.20
11.19
11.44
10.99
11.07
0.08
11.46
11.61
10.43
10.48
10.15
10.07
0.00
0.00
0.11
9.60
0.09
10.20
0.00
10.62
0.01
10.88
3.99
10.57
11.79
0.11
9.93
10.59
10.29
8.91
0.16
9.87
10.25
10.42
10.80
10.46
10.03
0.00
0.00
0.04
9.89
0.03
10.57
0.00
10.18
0.02
10.69
4.54
11.04
10.67
0.05
10.41
10.29
10.54
7.82
0.02
9.99
10.32
10.17
10.29
9.70
9.98
0.00
0.00
0.07
9.62
0.09
10.80
0.00
10.32
0.00
9.54
3.24
10.25
10.51
0.10
9.99
9.45
9.62
7.25
0.03
9.76
8.09
9.94
171
-------
Content by in
Average of MTBE (wt%)
Season
Summer
Winter
Area
Los Angeles, CA
Sacramento Metro, CA
San Diego, CA
San Joaquin, CA
Los Angeles, CA
Sacramento Metro, CA
San Diego, CA
San Joaquin, CA
Year
10.85 10
11.56 10
10.55 10
10
10
10
96
85
98
63
85
78
11.14
10.99
11.00
11.37
11.25
11.51
11.94
11.54
11.89
11.82
11.12
11.71
10.94
11.47
11.32
12.12
11.34
11.96
9.76
9.37
10.92
9.66
10.04
11.21
9.56
9.86
10.39
9.86
9.59
10.54
0.42
4.83
2.50
6.08
0.79
2.73
1.68
5.28
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
bf and CA)
Average of Ethanol (wt°/o)
Season Area
Atlantic City, NJ
Baltimore, MD
Boston-Worcester, MA
Chicago-Lake Co., IL, Gary,
IN
Covington, KY
CT - remainder
Dallas-Fort Worth, TX
Hartford, CT
Houston-Galveston, TX
Knox Co. and Lincoln Co.,
ME
Lewiston-Auburn, ME
Louisville, KY
Manchester, NH
Milwaukee-Racine, WI
Norfolk-Virginia Beach, VA
NY-NJ-Long Is.-CT
Phila.-Wilm, DE-Trenton, NJ
Phoenix, AZ
Portland, ME
Portsmouth- Dover, NH
Poughkeepsie, NY
Queen Anne Co.-Kent Co.,
MD
Rhode Island
Richmond, VA
Springfield, MA
St. Louis, MO
Sussex County, DE
Warren County, NJ
Washington, DC area
Atlantic City, NJ
Baltimore, MD
Boston-Worcester, MA
Chicago-Lake Co., IL, Gary,
IN
Covington, KY
CT - remainder
Dallas-Fort Worth, TX
Year
0.00
0.00
0.00
6.33
1.45
0.03
0.00
0.00
0.00
1.36
0.00
6.56
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.15
0.00
0.07
7.69
1.72
0.02
0.00
0.00
0.00
9.29
0.00
0.00
0.00
0.00
1.45
9.58
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.02
0.00
0.00
0.00
0.22
9.75
0.17
0.00
0.00
0.00
0.00
9.57
0.00
0.00
0.00
2.32
0.00
9.61
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.15
0.00
0.10
10.00
0.00
0.00
0.00
9.74
4.83
0.00
0.00
0.01
0.01
0.00
3.01
0.00
9.76
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.18
10.56
7.85
0.23
0.00
0.00
0.01
0.00
10.21
7.46
0.00
0.00
0.00
2.40
0.00
10.10
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2.00
0.00
0.00
0.00
0.16
0.00
0.33
10.61
7.49
0.00
0.00
0.00
0.00
10.12
6.95
0.00
0.00
0.00
2.21
0.00
10.05
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
1.35
0.00
0.00
0.00
0.54
0.00
0.30
10.66
6.76
0.01
0.00
0.00
0.00
10.14
6.78
0.00
0.00
0.00
2.54
0.00
9.91
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1.52
0.00
0.00
0.07
0.00
0.36
10.50
6.69
0.00
0.00
0.00
0.00
10.09
6.64
0.00
0.00
0.00
0.00
1.78
0.00
10.05
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
8.65
0.00
0.00
0.17
0.00
0.18
10.48
8.65
0.41
0.00
0.00
0.00
0.00
10.10
10.29
0.00
0.00
0.00
0.00
10.05
0.00
10.07
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10.04
0.00
0.01
0.00
0.00
0.00
0.30
10.55
10.71
5.15
0.01
0.00
0.00
0.00
10.15
10.34
10.13
0.00
10.16
0.00
10.25
0.00
10.12
0.00
5.78
0.00
0.00
10.14
0.00
0.00
0.00
0.36
10.20
0.00
0.00
0.00
0.00
0.00
0.22
10.57
11.22
10.28
0.00
0.00
0.00
0.33
10.39
10.49
10.27
0.00
10.33
0.00
10.36
0.01
10.29
0.00
5.76
0.01
0.00
10.29
0.00
0.32
0.00
2.47
10.53
0.00
0.01
0.00
0.00
0.00
0.38
10.64
10.63
10.45
0.00
172
-------
Hartford, CT
Houston-Galveston, TX
Knox Co. and Lincoln Co.,
ME
Lewiston-Auburn, ME
Louisville, KY
Manchester, NH
Milwaukee-Racine, WI
Norfolk-Virginia Beach, VA
NY-NJ-Long Is.-CT
Phila.-Wilm, DE-Trenton, NJ
Phoenix, AZ
Portland, ME
Portsmouth- Dover, NH
Poughkeepsie, NY
Queen Anne Co.-Kent Co.,
MD
Rhode Island
Richmond, VA
Springfield, MA
St. Louis, MO
Sussex County, DE
Warren County, NJ
Washington, DC area
0.22
0.00
0.00
1.61
0.00
6.72
0.00
0.41
0.44
0.00
0.00
0.13
0.00
0.24
0.00
0.37
0.00
0.09
0.35
0.00
1.64
8.13
0.00
0.11
0.09
0.00
0.16
0.30
0.00
0.35
0.00
0.51
0.00
2.34
0.00
6.99
0.00
0.35
0.14
10.17
0.00
0.33
0.14
0.00
0.21
0.00
0.35
0.00
0.00
2.01
0.00
10.16
0.00
0.30
0.05
0.00
0.00
0.32
0.25
0.00
0.22
0.00
0.41
0.00
2.69
0.00
10.23
0.00
0.18
0.16
0.00
0.38
0.25
0.00
0.36
4.74
0.00
0.00
0.00
0.43
0.00
2.55
0.00
10.51
0.01
0.36
0.18
0.00
0.71
0.48
0.00
0.72
4.73
0.00
0.01
0.00
0.27
0.00
2.58
0.00
10.10
0.00
0.16
0.10
0.00
0.30
0.00
0.34
0.00
0.27
4.52
0.00
0.01
0.45
0.00
6.13
0.00
10.40
0.00
0.13
0.12
0.00
0.32
0.00
0.35
0.00
0.50
6.54
0.00
0.00
2.72
0.00
10.69
0.03
10.63
0.00
1.33
0.14
0.00
5.05
0.38
0.00
0.59
9.72
0.00
0.00
0.00
10.33
0.00
10.72
0.00
10.49
0.00
6.37
0.00
0.01
10.51
0.00
0.26
0.00
1.44
10.32
0.00
0.00
0.00
10.48
0.00
10.61
0.00
10.62
0.00
7.20
0.04
0.00
10.45
0.01
0.44
0.00
2.99
10.30
0.00
1.67
0.00
Average of Ethanol (wt°/o)
Season
Summer
Winter
Area
Los Angeles, CA
Sacramento
Metro, CA
San Diego, CA
San Joaquin, CA
Los Angeles, CA
Sacramento
Metro, CA
San Diego, CA
San Joaquin, CA
Year
0.01 0
0.00 0
0.00 0
0
0
0
00
00
00
00
00
00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.77
0.61
0.49
1.00
0.63
0.48
1.00
0.77
0.52
0.75
0.65
0.57
5.59
3.38
4.71
2.90
5.36
4.24
4.94
3.21
5.96
5.90
5.94
6.10
6.08
6.08
5.99
6.25
6.09
6.12
6.02
6.25
5.77
5.97
5.92
6.07
173
-------
Appendix to Benzene Chapter
RFG Benzene by Gasoline Volume
Benzene Volume Distribution Trend-Summer RFG
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 1
Summer RFG Benzene Content by Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
(gal):
1997
0.07
0.25
0.32
0.37
0.41
0.44
0.48
0.51
0.55
0.58
0.62
0.67
0.71
0.75
0.80
0.84
0.89
0.95
1.02
1.13
1.30
12,435,941,155
1998
0.05
0.24
0.32
0.38
0.42
0.47
0.50
0.53
0.55
0.59
0.62
0.65
0.70
0.76
0.82
0.88
0.95
1.01
1.06
1.13
1.33
12,831,030,495
1999
0.06
0.29
0.36
0.40
0.43
0.48
0.52
0.57
0.61
0.65
0.68
0.71
0.75
0.80
0.85
0.90
0.97
1.02
1.08
1.15
1.42
12,996,112,177
2000
0.00
0.29
0.34
0.37
0.39
0.42
0.44
0.47
0.49
0.51
0.53
0.55
0.58
0.61
0.66
0.70
0.76
0.83
0.95
1.05
1.28
12,982,665,444
2001
0.01
0.27
0.34
0.38
0.41
0.44
0.46
0.48
0.51
0.53
0.56
0.59
0.62
0.66
0.70
0.76
0.82
0.89
0.98
1.10
1.30
13,222,175,920
2002
0.00
0.25
0.32
0.36
0.39
0.42
0.44
0.47
0.50
0.52
0.55
0.57
0.60
0.63
0.66
0.70
0.75
0.81
0.91
1.04
1.29
13,847,238,818
2003
0.06
0.27
0.33
0.36
0.38
0.41
0.43
0.46
0.49
0.52
0.55
0.59
0.63
0.66
0.71
0.75
0.80
0.86
0.94
1.08
1.30
13,584,238,237
2004
0.00
0.28
0.33
0.37
0.40
0.43
0.45
0.48
0.51
0.54
0.56
0.59
0.61
0.64
0.67
0.71
0.75
0.80
0.88
1.01
1.51
14,227,790,769
2005
0.01
0.30
0.37
0.41
0.45
0.48
0.51
0.54
0.57
0.60
0.64
0.67
0.71
0.74
0.78
0.82
0.86
0.90
0.96
1.06
1.34
14,082,668,582
174
-------
RFG Benzene by Gasoline Volume (continued):
c 0.60 -
M
Benzene Volume Distribution Trend-Winter RFG
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 2
Winter RFG Benzene Content by Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
tail:
1997
0.00
0.21
0.28
0.33
0.36
0.40
0.44
0.47
0.50
0.54
0.58
0.63
0.67
0.72
0.77
0.82
0.87
0.93
1.01
1.11
2.65
14,914,492,944
1998
0.02
0.23
0.31
0.36
0.40
0.43
0.47
0.50
0.53
0.56
0.60
0.65
0.69
0.74
0.79
0.84
0.89
0.95
1.01
1.09
1.32
15,066,250,612
1999
0.01
0.27
0.32
0.36
0.39
0.42
0.45
0.48
0.52
0.56
0.59
0.63
0.67
0.72
0.78
0.83
0.89
0.96
1.04
1.13
1.30
15,082,593,974
2000
0.03
0.28
0.33
0.37
0.40
0.43
0.47
0.50
0.53
0.56
0.60
0.64
0.68
0.72
0.78
0.83
0.89
0.95
1.02
1.11
1.33
15,826,525,899
2001
0.01
0.28
0.34
0.38
0.42
0.45
0.48
0.51
0.54
0.57
0.60
0.63
0.66
0.70
0.74
0.79
0.85
0.93
1.01
1.10
1.30
15,733,483,045
2002
0.01
0.27
0.33
0.37
0.40
0.44
0.48
0.51
0.55
0.58
0.61
0.65
0.68
0.71
0.74
0.78
0.83
0.89
0.98
1.09
1.30
16,429,447,508
2003
0.00
0.27
0.33
0.36
0.39
0.43
0.46
0.50
0.53
0.57
0.61
0.64
0.68
0.72
0.76
0.80
0.87
0.93
1.01
1.11
1.30
16,675,657,569
2004
0.01
0.26
0.31
0.35
0.38
0.42
0.45
0.49
0.52
0.56
0.59
0.62
0.66
0.71
0.75
0.79
0.85
0.91
0.99
1.09
1.31
17,189,415,193
2005
0.01
0.30
0.36
0.40
0.45
0.48
0.51
0.55
0.58
0.61
0.64
0.67
0.70
0.74
0.78
0.82
0.87
0.92
0.98
1.08
1.68
18,049,752,652
175
-------
Percentile Chart of Benzene Content By Volume-2005 Summer RFG
(from Batch Reports)
0.60 0.80
Benzene {v %)
Figure 3
Percentile Chart of Benzene Content By Volume-2005 Winter RFG
(from Batch Reports)
100% -
90% -
80% -
70% -
«> 60% -
| 50% -
e£ 40% -
30% -
20% -
10% -
0% -
0
^^
/
/
/
/
/
f
f
_^/
)0 0.20 0.40 0.60 080 1.00 120 1.40 160 1.80
Benzene {v %}
Figure 4
176
-------
RFC Benzene by Grade
Average Benzene Content of Summer RFG Sold at Retail Stations-By Grade
"o
jg
1 n
I oso -
0.40 -- J-
D 20 -
1995
Q Avg. Regular 0.70
Avg. Mid 0.67
DAvg. Premium 0.58
"I " Bars denote range of RFG area averages
111
1996 1997 199B
0.71 0.72 071
068 0.68 069
0.62 0.60 OBI
1999
0.74
0.71
0.63
T I IT
200D 2DD1 2002
0.62 0.67 0.65
0.59 0.63 OBD
0.53 0.54 0.50
"1 ' "1
2003 2004 200
0.68 0.68 0.7
O.B4 O.B3 0.7
0.53 0.53 OB
--,
5
4
1
2
Year
Figure 5
Average Benzene Content of Winter RFG Sold at Retail Stations-By Grade
"o
>
n
N
I D-6°- Hi
D.40 -
0.20 -
1995
DAvg. Regular 0.63
Avg. Mid 059
D Avg. Premium 0.48
"I" Bars denote range of EFG area averages
-1 1 -,
1996 1997 1998
0.73 0.72 D.68
0.68 0.6E! 0.67
0.59 0.58 0.56
"
1999
0.70
D.68
0.61
1 IT!'
ill
20DO 2001 2002
068 069 D.68
0.67 0.65 0.62
0.57 0.55 0.53
T T
'] 1
2003 20D4 20D
D.73 0.72 0.7
0.70 0.66 06
0.54 0.55 0.5
~|
5
4
3
3
Year
Figure 6
177
-------
by
bf
' ill
a
Grade Season
PRM s
w
REG s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
1997
0.60
3,433,830,960
0.57
3,876,199,245
0.68
8,933,744,061
0.64
10,818,337,187
0.60
3,649,323,954
0.59
4,139,145,511
0.70
9,007,552,063
0.67
10,663,420,477
0.63
3,537,270,189
0.61
3,839,047,178
0.73
9,318,697,543
0.66
10,902,774,956
0.53
2,788,061,943
0.54
3,419,795,153
0.61
10,122,448,765
0.68
12,243,727,277
0.52
2,772,653,293
0.55
3,305,189,631
0.65
10,392,283,260
0.67
12,361,561,882
0.48
2,931,154,091
0.52
3,387,100,772
0.62
10,879,031,865
0.67
13,031,122,001
0.50
2,736,350,778
0.50
3,102,938,885
0.63
10,824,904,975
0.68
13,483,257,340
0.50
2,384,807,492
0.52
2,944,118,191
0.62
11,779,211,527
0.65
14,160,927,780
0.58
2,143,381,101
0.56
2,583,362,289
0.68
11,881,795,879
0.69
15,427,424,995
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates
based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also
differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is
often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported and
included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes
some gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as regular and
premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as
mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a
grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
178
-------
RFG Benzene-Geographic
Change in Benzene Content from 2000 to 2005-Summer RFG
Surveys
Boston-Worcester, MA (0.51.0.77)
Atlantic City, NJ (0.52,0 76)
Portsmouth-Dover, NH (0.62,0.83)
Manchester, NH (0.65,0.87)
Sussex County, DE (0.73,0 92)
Hartford, CT (0.50,0.69)
Springfield, MA (0.53,0.71)
St. Louis, MO (0.55,0.70)
-
Dallas-Fort Worth, TX (0.58,0 73)
Covington, KY (0.82,0.97)
Phila.-Wilm, DE-Trenton, NJ (0.57,0.71)
Rhode Island (0.56,0 69)
NY-NJ-Long Is.-CT (0.52,0.65)
Poughkeepsie, NY (0.54,0.65)
Warren County, NJ (0.52,0.62)
Washington, D.C.-area (0.58,0 68)
Chicago-Lake Co., IL, Gary, IN (0.78,0.84)
Richmond, VA (0.59,0.65)
Baltimore, MD (0.60,0.65)
Houston-Galveston, TX (0.61,0.66)
Louisville, KY (0.88,0 89)
Norfolk-Virginia Beach, VA (0.71,0.67)
Milwaukee-Racine. Wl (0.93,0.84)
-0
15 -01 -0.05 0 0.05 0.1 0.15 0.2 0.25 03
Change in Benzene volume %
Figure 7
Change in Benzene Content
Warren County. NJ (0.60,0.85)
St Louis, MO (0.57,0 80)
Poughkeepsie, NY (0.64,0.77)
Hartford, CT (062 ,0.75)
Rhode Island (0.60,0 72)
NY-NJ-Long is.-CT (0.62,0.73)
Phila.-Wilm, DE-Trenton, NJ (0.60,0.69)
Atlantic City, NJ (0.63 ,0.71)
& Portsmouth-Dover, NH (0.63,0.70)
g
| Washington, D.C.-area (0.54,0. BO)
Sussex County, DE (D.65,0.70)
> Norfolk-Virginia Beach, VA (0.66,0.70)
X Springfield, MA (065,0. 69)
1
co Boston-Worcester, MA (0.65,0. 6B)
| Dallas-Fort Worth, TX (0.63,0.65)
Chicago-Lake Co., IL, Gary, IN (0.78,0.79)
Baltimore. MD (0 60,0.60)
Louisville, KY (0.84,0.82)
Richmond, VA (0.60 ,0.58)
Manchester, NH (0.73,0.68)
Covington, KY (0.77,0.71)
Milwaukee-Racine, Wl (0.87,0.79)
Houston-Galveston, TX (0 69,0 58)
-0
from 2000 to 2005-Winter RFG Surveys
1
^m
^m
^m
^^^M
' i
!
1
=^^
^m
izz:
LZ=
H
D
^
]
15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3
Cliiinge in Benzene volume
Figure 8
179
-------
Conventional Gasoline
CG Benzene by Gasoline Volume
Benzene Volume Distribution Trend-Summer CG
(from Batch Reports-Excluding Blendstocks)
2000 2001 2002
Reporting Year
Figure 9
Summer CG Benzene Content by Volume (from Batch Reports excluding Blendstocks)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volums
(gal):
1997
0.00
0.26
0.37
0.44
0.50
0.57
0.65
0.73
0.79
0.87
0.95
1.05
1.17
1.29
1.42
1.55
1.69
1.85
2.06
2.43
5.33
39,313,521,046
1998
0.00
0.27
0.39
0.46
0.53
0.59
0.66
0.73
0.80
0.88
0.96
1.06
1.15
1.26
1.38
1.52
1.67
1.85
2.09
2.39
5.29
39,075,503,909
1999
0.00
0.26
0.38
0.47
0.54
0.60
0.66
0.73
0.81
0.87
0.95
1.03
1.13
1.24
1.37
1.51
1.65
1.84
2.09
2.43
5.00
37,192,482,168
2000
0.00
0.34
0.44
0.51
0.57
0.62
0.67
0.74
0.81
0.87
0.95
1.03
1.12
1.22
1.34
1.47
1.58
1.74
1.98
2.36
5.25
36,393,283,808
2001
0.03
0.35
0.47
0.55
0.62
0.68
0.73
0.79
0.85
0.92
1.00
1.08
1.17
1.27
1.40
1.54
1.68
1.83
2.07
2.37
5.00
38,566,604,576
2002
0.00
0.34
0.45
0.50
0.54
0.60
0.69
0.76
0.83
0.88
0.94
1.01
1.10
1.19
1.30
1.42
1.57
1.76
1.95
2.28
5.00
40,819,234,602
2003
0.00
0.36
0.46
0.53
0.59
0.65
0.72
0.78
0.85
0.92
1.00
1.08
1.18
1.27
1.37
1.49
1.61
1.77
1.95
2.30
5.00
43,235,647,755
2004
0.00
0.35
0.47
0.55
0.62
0.68
0.73
0.79
0.86
0.93
0.99
1.08
1.15
1.24
1.35
1.49
1.62
1.72
1.94
2.30
5.14
43,503,138,987
2005
0.00
0.37
0.47
0.54
0.60
0.67
0.74
0.80
0.88
0.98
1.06
1.15
1.26
1.37
1.45
1.55
1.66
1.85
2.09
2.45
5.08
42,188,895,941
180
-------
CG Benzene by Gasoline Volume (continued):
Benzene Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2000 2001 2002
Reporting Year
Figure 10
Winter CG Benzene Content by Volume
(from Batch
Reports excluding Blendstocks)
Reporting Year
Volume
%tile
Tiinimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
Cqall:
1997
0.00
0.23
0.34
0.43
0.49
0.54
0.60
0.66
0.74
0.81
0.88
0.99
1.10
1.25
1.40
1.58
1.74
1.93
2.19
2.64
5.23
44,616,143,682
1998
0.00
0.23
0.35
0.43
0.48
0.55
0.60
0.67
0.72
0.79
0.90
1.00
1.10
1.21
1.33
1.43
1.56
1.77
2.01
2.39
5.00
46,277,269,226
1999
0.00
0.26
0.39
0.46
0.52
0.56
0.61
0.67
0.73
0.79
0.87
0.97
1.07
1.18
1.30
1.41
1.56
1.74
2.02
2.39
5.34
47,691,599,779
2000
0.00
0.28
0.39
0.46
0.52
0.56
0.61
0.66
0.73
0.80
0.88
0.96
1.05
1.15
1.25
1.38
1.53
1.72
1.98
2.40
5.28
48,005,753,994
2001
0.00
0.31
0.42
0.50
0.56
0.61
0.67
0.72
0.79
0.86
0.93
1.01
1.10
1.22
1.34
1.47
1.60
1.78
2.09
2.52
5.00
48,802,911,584
2002
0.00
0.32
0.43
0.49
0.55
0.61
0.66
0.72
0.78
0.84
0.91
0.98
1.06
1.15
1.23
1.33
1.45
1.64
1.91
2.30
4.88
49,671,321,830
2003
0.00
0.32
0.44
0.51
0.57
0.62
0.68
0.74
0.81
0.87
0.94
1.02
1.10
1.19
1.28
1.38
1.52
1.67
1.89
2.28
5.22
47,446,594,748
2004
0.00
0.32
0.42
0.49
0.55
0.61
0.67
0.73
0.79
0.85
0.92
0.99
1.08
1.17
1.27
1.39
1.51
1.67
1.92
2.25
5.00
47,637,145,526
2005
0.00
0.35
0.46
0.53
0.60
0.66
0.72
0.79
0.85
0.92
0.98
1.06
1.15
1.25
1.34
1.43
1.54
1.69
1.94
2.42
5.20
48,683,162,256
181
-------
Percentile Chart of Benzene Content By Volume-2005 Summer CG
(from Batch Reports-Excluding Blendstocks)
3.00
Benzene |v %)
Figure 11
Percentile Chart of Benzene Content By Volume-2005 Winter CG
(from Batch Reports-Excluding Blendstocks)
2.00 3.00
Benzene (v %)
Figure 12
182
-------
i
CG le
me bf Yi
I
Grade Season
PRM s
w
REG s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
1997
0.95
6,847,436,919
0.83
8,297,352,383
1.12
28,060,742,477
1.15
32,178,466,105
0.95
6,631,452,976
0.86
8,231,582,167
1.13
28,406,715,846
1.09
32,165,391,015
0.92
6,252,416,423
0.89
7,674,529,390
1.14
26,654,202,116
1.09
33,356,627,836
0.83
4,957,386,104
0.86
6,623,696,787
1.10
27,358,010,027
1.05
35,197,390,669
0.94
4,996,778,324
0.96
6,552,510,604
1.15
28,964,788,056
1.11
35,933,948,416
0.85
5,181,939,065
0.86
6,571,003,043
1.08
30,995,888,528
1.06
36,930,102,215
0.87
5,144,650,585
0.87
5,505,637,848
1.13
33,050,308,019
1.08
35,937,081,377
0.85
4,629,542,860
0.84
5,063,677,945
1.11
34,601,650,558
1.05
37,501,139,752
0.89
4,132,713,279
0.94
4,498,577,769
1.18
33,850,837,227
1.12
39,113,050,332
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates
based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also
differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is
often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported and
included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes
some gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as regular and
premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as
mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a
grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
183
-------
Appendix to Aromatics Chapter
RFG Aromatics by Gasoline Volume
Aromatics Volume Distribution Trend-Summer RFG
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 1
Summer RFG Aromatics Content (Vol%) by Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
(gal):
1997
0.2
11.7
13.9
15.3
16.4
17.3
18.1
19.0
20.0
20.9
21.8
22.6
23.5
24.3
25.1
26.1
27.3
29.1
32.1
35.7
52.4
12,446,593,615
1998
2.6
12.9
14.7
15.8
16.9
17.9
19.2
20.0
20.8
21.4
22.1
22.8
23.5
24.3
25.3
26.6
28.1
30.0
32.4
35.0
50.6
12,832,964,637
1999
0.4
12.0
13.8
15.2
16.1
17.0
18.0
18.8
19.9
20.8
21.6
22.5
23.4
24.2
25.3
26.6
27.8
29.5
31.3
34.0
48.4
12,998,841,841
2000
1.9
10.5
13.2
14.6
15.3
16.0
16.5
17.0
17.5
18.0
18.6
19.2
19.8
20.5
21.2
22.0
23.0
24.5
26.5
29.4
47.8
12,983,168,478
2001
3.1
13.0
14.4
15.2
16.0
16.7
17.4
17.9
18.5
19.0
19.5
20.1
20.7
21.4
22.1
22.9
23.8
24.8
26.3
28.8
42.3
13,222,633,468
2002
0.0
13.3
14.6
15.5
16.2
16.9
17.6
18.2
18.7
19.3
19.9
20.5
21.2
21.9
22.6
23.5
24.3
25.3
26.4
28.4
48.9
13,847,971,634
2003
1.9
12.7
14.1
15.2
16.1
16.8
17.4
18.0
18.5
19.0
19.6
20.1
20.7
21.3
21.9
22.7
23.7
24.9
26.3
28.4
46.4
13,584,860,845
2004
0.4
13.1
14.5
15.5
16.3
17.0
17.7
18.2
18.7
19.2
19.7
20.2
20.8
21.3
21.9
22.7
23.5
24.7
25.9
28.0
40.1
14,232,658,149
2005
2.0
13.7
15.0
16.0
17.0
17.8
18.6
19.1
19.5
19.9
20.3
20.7
21.3
21.9
22.7
23.6
24.7
25.7
27.0
29.1
41.8
14,083,382,582
184
-------
RFGAromatics by Gasoline Volume (continued):
Aromattcs Volume Distribution Trend-Winter RFC
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 2
Winter RFC Aromatics Content by Volume
(from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
faall:
1997
0.0
9.6
11.6
12.7
13.7
14.5
15.3
16.0
16.7
17.4
18.1
18.9
19.7
20.5
21.5
22.7
24.0
26.0
28.8
32.5
50.0
14,914,411,907
1998
1.0
10.8
12.5
13.7
14.5
15.2
16.1
16.8
17.6
18.3
18.9
19.6
20.4
21.1
22.0
22.9
24.5
26.3
29.1
32.8
50.0
15,059,853,293
1999
0.6
10.2
12.1
13.4
14.3
15.1
15.8
16.4
17.1
17.7
18.5
19.2
19.9
20.8
21.8
22.9
24.3
25.9
28.5
32.6
48.7
15,079,861,417
2000
0.9
9.8
12.0
13.2
13.9
14.7
15.3
15.9
16.5
17.1
17.7
18.5
19.3
20.2
21.2
22.3
23.7
25.3
27.5
31.1
46.9
15,829,263,333
2001
0.9
10.2
12.2
13.3
14.2
15.0
15.6
16.2
16.9
17.4
17.9
18.7
19.4
20.2
21.1
22.3
23.6
25.1
27.8
32.1
49.1
15,724,574,173
2002
1.4
11.5
13.1
14.0
14.8
15.5
16.1
16.7
17.2
17.8
18.5
19.1
19.9
20.6
21.3
22.3
23.3
24.7
26.7
30.9
45.8
16,433,999,720
2003
0.6
11.8
13.3
14.3
15.0
15.7
16.4
17.0
17.6
18.2
18.7
19.3
19.9
20.5
21.2
22.1
23.1
24.5
26.3
29.7
46.6
16,679,773,135
2004
0.7
10.8
12.7
13.9
15.0
15.7
16.4
17.0
17.5
18.1
18.6
19.2
19.7
20.3
21.1
21.8
22.8
24.0
25.7
29.0
47.6
17,190,635,125
2005
0.2
11.8
13.7
14.8
15.6
16.3
16.9
17.5
18.1
18.7
19.2
19.7
20.2
20.7
21.3
22.0
22.9
24.2
25.8
28.5
49.1
18,044,832,058
185
-------
Percentile Chart of Aromatics Content by Volume 2005 Summer RFG
(from Batch Reports)
20 25
Aromatics (Volume %)
Figure 3
Percentile Chart of Aromatics Content by Volume 2005 Winter RFG
(from Batch Reports)
Aromatics (Volume %)
Figure 4
186
-------
RFG Aromatics by Grade
>r
Aromatics {vol
Average
40 -T
20 -
15 -
10 -
c
J "
QAvg. Regular
Avg.Mid
DAvg. Premium
' Aromatics
rn
1995 199
22.8 23.^
25.1 25
28.0 28(
Content of
i T r
3 1997
1 23.9
5 261
) 28.6
Summer R
I
I (
1998 199
24.4 23.
26.8 25.
30.0 28.
FG Sold at
Bars donot
r-
9 2000
> 18.6
' 20.0
3 21.9
Year
Ret
2001
19.6
20.6
22.2
ail!
nge
Stations-
of RFC a
b
2002 2
19.9 1
20.9 2
22.7 2
By Grade
"° av°"3'
1
303 2004
9.6 20.8
05 21.5
2.1 22.5
|
2005
20.7
21 1
22.4
Figure 5
Average Aromatics Content of Winter RFG Sold at Retail Stations-By Grade
a Avg. Regular
I Avg.Mid
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
20.3
18.8
21.3
21.3
19.2
21.9
19.8
21.8
16.9
18.6
17.2
19.3
18.4
19.7
19.9
18.7
19.8
18.2
19.2
DAvg. Premium | 24.7 | 26.4 | 24.1 | 25.6 | 25.8 | 22.2 | 22.5 | 22.8 | 23.1 | 22.5 | 21 i
Year
Figure 6
187
-------
By G,
RI
:G
t bf Y
Grade Season
PRM s
w
REG s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
1997
27.0
3,449,496,834
23.3
3,889,379,685
20.6
8,931,922,101
17.8
10,816,109,633
27.4
3,651,258,096
25.0
4,139,316,745
20.9
9,007,552,063
17.9
10,663,420,477
26.5
3,537,723,033
24.4
3,839,207,618
20.5
9,320,974,363
17.9
10,900,737,704
22.5
2,788,564,977
23.5
3,420,180,881
18.4
10,122,448,765
17.7
12,246,078,983
23.2
2,773,110,841
24.6
3,305,514,921
19.3
10,392,283,260
17.7
12,361,561,882
22.4
2,931,886,907
23.3
3,387,457,520
19.8
10,879,031,865
18.4
13,031,122,001
22.6
2,736,973,386
23.1
3,103,333,727
19.4
10,824,904,975
18.6
13,486,978,064
22.4
2,385,535,898
22.7
2,945,338,123
19.7
11,783,350,501
18.4
14,160,927,780
22.8
2,144,095,101
23.3
2,583,694,635
20.6
11,881,795,879
19.0
15,422,172,055
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates based
on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also differ from
actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is often blended from
regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported and included in these tables as
regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes some gasoline and blendstock that was
included in aggregate average estimates. EPA has presented averages for batches labeled as regular and premium gasoline only; excluding batches
labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as mid-grade), mix of grades, or without a grade
label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed above
these volumes are not expected to represent the actual volumes by grade of retail gasoline.
188
-------
RFC Aromatics-Geographic
Change in Aromatics Conte
S
Hartford. CT (17.53 ,22.75)
Rhode Island (18.4223.29)
Springfield, MA (19.00 ,23 52)
Boston-Worcester, MA (19.37,23 80)
Warren County, NJ (19.69,22.80)
Manchester, NH (21.72,23.99)
Portsmouth-Dover, NH (23 15,25 07)
Dallas-Fort Worth, TX (18.64,20.49)
| Washington, D.C.-area (19.44,21 .28)
1 NY-NJ-Long Is.-CT (20.17,21.93)
Houston-Galveston, TX (20.12,21 68)
' Poughkeepsie,NT(21.11,22.47)
f Phila.-Wilm, DE-Trenton, NJ (21.19,22.29)
| Norfolk-Virginia Beach, VA (19.62,20.69)
J Chicago-Laka Co , IL, Gary, IN (17.07,18.05)
Richmond, VA (19.95,20 63)
Milwaukee-Racine, Wl (16.84,17.32)
Baltimore, MD (20.27,20.70)
Atlantic City, NJ (21 .66 ,21 .88)
Sussex County, DE (24.87,23.85)
Louisville, KV (18.00,16.13)
Covington, KY (19.19,15 39)
St. Louis.MO (21 .02,1648)
-
it from 2000 to 2005-Summer RFG
urveys
mmt
mt
mmm
^
i
i
^^
mmm
1
mmm
mmm
mm
i
1
1
mmm
mm
mm
m
m
]
|
i
5-4-20246
CliAliye in AioimiTics vol %
Change in Aromatics Conte
5
Portsmouth-Dover, NH (15.33,19.79)
Rhode Island (17.49.20.84)
Springfield. MA (17.13,20.11)
Manchester, NH (17.87,20.07)
Hartford, CT (17.54,19.62)
St. Louis, MO (20.10,22.16)
Warren County, NJ (18.80,20.66)
Atlantic City, NJ (17.39,18.99)
| NY-NJ-Long Is.-CT (19.47 ,20.82)
;" Dallas-Fort Worth, TX (16.99 ,17. 96)
Baltimore, MD (18.01 ,18.87)
Poughkeepsie, NY (20.18,21.03)
o
| Washington, D.C.-area (17.79,18.61)
| Chicago-Lake Co., IL, Gary, IN (15.09,15.74)
2 Sussex County, DE (20.33,20.95)
' Phila.-Wilm, DE-Trenton, NJ (18.74,19.04)
Richmond, VA (17.73,17.99)
Louisville, KY (14. 11,14. 16)
Boston-Worcester, MA (19.76,19.54)
Houston-Galveston, TX (19.14,18.82)
Milwaukee-Racine, Wl (15.56,15 16)
Norfolk-Virginia Beach, VA (19.36,18.14)
Covington, KY (15.17,13.22)
nt from 2000 to 2005-Winter RFG
urveys
1
mm
t
m
m
mmm
mmm
^^^^^^^^
i
^^mm
mmmmm
^^^
mm
mmt
^
^
mt
n
3-2-1012345
Change in Aiomatics vol %
Figure 7
Figure 8
189
-------
Conventional Gasoline
CGAromatics by Gasoline Volume
Aromatics Volume Distribution Trend-Summer CG
(from Batch Reports-Excluding Blendstocks)
2000 2001
Reporting Year
Figure 9
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume(gal):
Summer
Reporting Year
1997
0.0
16.30
19.00
20.60
21.90
22.90
23.80
24.70
25.50
26.20
26.90
27.70
28.50
29.20
30.10
31.00
32.40
34.00
36.10
39.40
60.20
39,380,364,542
CG Aromatics Content (Vol
1998
0.0
16.60
19.30
20.80
22.10
23.10
24.10
24.90
25.50
26.30
26.90
27.60
28.30
29.10
29.90
30.90
32.10
33.60
35.90
39.70
59.50
39,078,941,326
1999
0.9
17.10
19.40
20.80
22.00
23.10
24.00
24.90
25.80
26.40
27.10
27.90
28.80
29.50
30.40
31.40
32.20
33.50
35.60
39.10
60.50
37,192,967,557
%) by Volume (from Batch Reports excluding Blendstocks)
2000
0.0
18.90
20.60
21.90
23.00
24.00
24.90
25.80
26.50
27.40
28.10
28.80
29.50
30.10
30.80
31.60
32.70
34.10
36.20
39.80
60.00
36,387,851,238
2001
0.0
19.00
20.80
22.10
23.00
23.90
24.60
25.40
26.10
26.80
27.60
28.40
29.20
30.00
30.80
31.70
32.60
34.10
36.30
40.80
59.80
38,542,042,404
2002
0.0
18.10
20.40
21.70
22.70
23.60
24.40
25.30
26.10
26.90
27.80
28.70
29.40
30.20
30.90
31.60
32.50
33.80
35.60
39.50
59.30
40,818,795,156
2003
1.0
18.50
20.30
21.70
22.70
23.60
24.40
25.10
25.90
26.80
27.50
28.30
29.00
29.80
30.70
31.60
32.60
33.70
35.60
39.40
58.30
43,243,575,045
2004
0.0
18.10
20.20
21.70
22.80
23.80
24.60
25.40
26.20
26.90
27.70
28.40
29.10
30.00
30.90
31.90
32.80
34.00
35.80
39.30
60.50
43,509,104,037
2005
0.0
17.60
20.00
21.70
22.90
23.80
24.70
25.40
26.10
26.70
27.40
28.10
28.80
29.60
30.50
31.40
32.40
33.50
35.30
38.60
59.90
42,196,228,805
190
-------
CGAromatics by Gasoline Volume (continued):
Aromatics Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2000 2001 2002
Reporting Year
Figure 10
Winter CG Aromatics Content (Vol %) by Volume (from Batch Reports excluding Blendstocks)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1997
0.0
15.5
17.3
18.5
19.5
20.3
21.1
21.9
22.5
23.4
24.3
25.1
26.0
26.8
27.8
28.7
30.0
31.6
33.5
36.9
60.3
44,627,982,120
1998
0.0
14.5
16.8
18.3
19.5
20.4
21.3
22.1
22.9
23.6
24.3
24.9
25.7
26.6
27.4
28.4
29.5
31.1
33.4
36.7
60.5
46,293,631,981
1999
0.0
15.2
17.1
18.6
19.7
20.6
21.4
22.1
23.0
23.7
24.5
25.2
26.0
26.8
27.7
28.6
29.7
31.2
33.2
36.7
58.0
47,701,715,605
2000
0.0
15.2
17.2
18.6
19.6
20.6
21.5
22.3
23.1
23.8
24.4
25.1
25.7
26.5
27.2
28.2
29.2
30.6
32.8
36.5
60.4
48,024,395,442
2001
0.0
16.2
18.0
19.2
20.2
21.1
21.9
22.7
23.5
24.2
24.8
25.5
26.2
26.9
27.7
28.5
29.7
31.1
33.4
37.0
58.9
48,797,132,426
2002
0.0
16.0
18.0
19.2
20.1
20.9
21.7
22.2
22.9
23.6
24.3
24.9
25.7
26.4
27.2
28.1
29.2
30.7
32.7
37.3
58.8
49,655,135,030
2003
0.0
15.8
17.9
19.2
20.2
21.0
21.8
22.4
23.1
23.8
24.4
25.0
25.7
26.3
27.2
28.0
29.2
30.5
32.6
36.3
59.0
47,469,723,098
2004
0.0
15.2
17.2
18.6
19.8
20.8
21.5
22.1
22.7
23.4
24.0
24.9
25.7
26.4
27.2
28.1
29.1
30.4
32.4
35.7
58.5
47,654,786,710
2005
0.0
15.2
17.8
19.1
20.1
20.9
21.6
22.3
23.0
23.6
24.2
24.8
25.6
26.3
27.1
27.9
28.8
30.1
32.3
35.8
59.2
48,705,329,856
191
-------
Percentile Chart of Aromatics Content By Volume-2005 Summer CG
(from Batch Reports-Excluding Blendstocks)
30 40
Aromatics (v%)
Figure 11
Percentile Chart of Aromatics Content By Volume-2005 Winter CG
(from Batch Reports-Excluding Blendstocks)
30 40
Aromatics (v %)
Figure 12
192
-------
CG
(
Vol / s ' by -
Grade
PRM
REG
Season Data
s Average
Volume
w Average
Volume
s Average
Volume
w Average
Volume
1997
31.8
6,910,973,553
28.2
8,310,678,815
26.3
28,062,618,693
24.3
32,175,385,639
32.6
6,635,803,305
29.7
8,233,384,513
26.4
28,410,709,416
23.8
32,177,203,389
31.6
6,252,416,423
29.6
7,690,219,372
26.6
26,656,367,421
24.1
33,356,627,836
33.1
4,957,386,104
29.1
6,648,441,003
27.6
27,358,607,481
24.0
35,199,495,331
34.7
4,996,778,324
30.3
6,548,540,176
27.4
28,964,604,238
24.6
35,935,271,164
32.4
5,181,939,065
29.8
6,571,003,043
27.4
30,995,449,082
24.3
36,931,713,041
32.2
5,149,486,549
29.5
5,517,761,736
27.4
33,053,399,345
24.4
35,949,009,293
31.2
4,635,507,910
28.3
5,077,869,493
27.5
34,601,650,558
24.1
37,504,589,388
31.3
4,138,979,931
29.5
4,510,104,837
27.3
33,851,903,439
24.3
39,123,690,864
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates based
on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also differ from
actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is often blended from
regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported and included in these tables as
regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes some gasoline and blendstock that was
included in aggregate average estimates. EPA has presented averages for batches labeled as regular and premium gasoline only; excluding batches
labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as mid-grade), mix of grades, or without a grade
label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed above
these volumes are not expected to represent the actual volumes by grade of retail gasoline.
193
-------
Appendix to Olefins Chapter
RFG Olefins Content by Gasoline Volume
Olefin Volume Distribution Trend-Summer RFG
(from Batch Reports)
2000 2001 2002 2003
Reporting Year
Figure 1
Summer RFG Olefins Content (Vol %) by Volume (from Batch Reports)
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
Reporting Year
1997
0.0
2.7
4.4
5.6
6.5
7.1
7.9
8.6
9.5
10.3
11.1
11.9
12.7
13.7
14.8
16.1
17.5
19.0
20.9
23.8
39.7
12,424,923,379
1998
0.3
2.4
3.9
4.8
5.8
6.7
7.4
8.3
9.2
9.9
10.7
11.5
12.1
12.9
13.7
14.6
15.4
16.5
18.1
20.0
28.0
12,832,964,637
1999
0.3
2.8
4.4
5.3
6.2
7.2
8.1
8.9
9.5
10.1
10.8
11.4
12.1
13.0
13.9
14.9
16.1
17.5
19.2
21.3
28.6
12,996,111,169
2000
0.3
1.3
2.4
4.4
5.8
7.0
8.1
9.0
9.9
10.5
11.0
11.6
12.2
12.8
13.6
14.5
15.1
15.9
16.6
17.9
24.8
12,983,168,478
2001
0.3
2.0
3.9
6.4
8.1
9.3
10.0
10.6
11.2
11.6
12.0
12.6
13.2
13.7
14.4
15.1
15.7
16.5
17.5
18.9
24.7
13,222,633,468
2002
0.0
1.8
3.5
5.4
6.8
8.0
8.9
9.5
10.0
10.5
11.0
11.6
12.1
12.6
13.2
13.9
14.6
15.5
16.3
17.7
25.0
13,847,178,590
2003
0.3
1.5
3.2
5.4
6.8
7.8
8.9
9.5
10.1
10.6
11.2
11.8
12.5
13.1
13.8
14.4
15.1
15.9
16.8
18.4
24.9
13,584,860,845
2004
0.0
2.0
4.1
6.0
7.1
8.2
9.1
9.9
10.5
11.1
11.8
12.3
12.9
13.4
14.1
14.7
15.3
16.0
16.9
18.5
24.7
14,232,658,149
2005
0.0
1.8
4.1
5.9
7.1
8.4
9.4
10.1
10.8
11.6
12.3
13.0
13.7
14.4
15.1
15.7
16.5
17.2
18.1
19.4
24.4
14,083,382,582
194
-------
RFG Olefins Content by Gasoline Volume (continued):
Olefin Volume Distribution Trend-Winter RFG
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 2
Winter RFG Olefins Content (Vol %) by Volume (from Batch
Reports)
Reporting Year
Volume %tile
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1997
2.00
3.60
4.70
5.60
6.40
7.00
7.70
8.50
9.40
10.30
11.20
12.00
13.10
14.00
15.20
16.50
18.30
20.30
23.40
44.10
14,918,897,927
1998
2.90
3.90
4.70
5.60
6.40
7.10
7.90
8.60
9.20
9.90
10.60
11.50
12.40
13.30
14.40
15.80
17.40
18.90
21.00
25.00
15,059,853,293
1999
3.10
4.30
5.10
6.10
6.80
7.60
8,40
9.30
9.90
10.70
11.40
12.10
12.90
13.90
14.80
16.10
17.40
19.50
21.70
25.40
15,081,898,669
2000
3.00
4.20
5.20
6.20
7.10
8.00
8.90
9.70
10.60
11.30
12.10
13.10
13.90
14.90
16.00
17.20
18.50
19.90
21.70
25.00
15,829,263,333
2001
2.90
4.40
5.70
6.90
8.00
8.90
9.70
10.50
11.30
12.00
12.70
13.40
14.20
15.10
16.10
17.30
18.90
20.80
22.80
26.90
15,724,574,173
2002
1.90
3.60
4.90
6.00
6.90
7.80
8.60
9.50
10.20
10.90
11.60
12.30
13.20
13.90
14.70
15.60
16.70
18.30
21.00
25.00
16,433,999,720
2003
1.50
3.30
4.50
5.80
6.80
7.70
8.60
9.40
10.10
10.80
11.50
12.20
13.00
13.90
14.70
15.80
17.10
18.90
21.00
27.00
16,679,773,135
2004
1.30
3.60
4.80
5.90
6.80
7.50
8.30
9.20
10.10
10.80
11.60
12.30
13.10
14.00
14.80
15.80
17.10
18.80
21.20
24.80
17,194,948,147
2005
1.40
3.40
4.70
5.90
6.90
7.80
8.60
9.40
10.00
10.60
11.30
12.00
12.70
13.60
14.50
15.50
17.10
19.10
21.30
25.00
18,044,832,058
195
-------
Percentile Chart of Olefin Content by Volume 2005 Summer RFC
(from Batch Reports)
Figure 3
Percentile Chart of Olefin Content by Volume 2005 Winter RFC
(from Batch Reports)
10 15
Olefins (Volume %)
Figure 4
196
-------
RFG Olefins by Grade
Average Olefins Content of Summer RFG Sold at Retail Stations-By Grade
18
16 -
14 -
12 -
10 -
8 -
6 -
4 -
2 -
nAvg Regular
I Avg. Mid
n Avg. Premium
"I" Bars denote range of RFG
11.3
9.9
1999
11.9
7.8
2000
10;
7.2
11.0
9.9
8.0
2002
11.5
10.4
8.0
2003
10.0
7.9
Year
Figure 5
| Olefins (vol %) |
Average Olefins Content of Winter RFG Sold at Retail Stations-By Grade
12 -
10 -
6 -
4 -
2 -
n Avg. Regular
Avg. Mid
n Avg. Premium
"I" Bars denote range of RFG
|
I
1998
10.4
8.9
7.2
1999
108
94
7.4
T
'
2000
10.7
9.5
8.0
-
2001
11.5
103
8.1
-
2002
11.3
101
8.0
2003
11.6
10.1
8.0
area averages
2004
11.1
9.9
8.3
2005
10.2
8.7
7.5
Year
Figure 6
197
-------
by
Grade
PRM
REG
Season
s
w
s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
bf
8.4 7.8 8.1 8.1 9.2 7.8 7.6 8.3 8.6
3,449,496,834 3,651,258,096 3,537,723,033 2,788,564,977 2,773,110,841 2,931,886,907 2,736,973,386 2,385,535,898 2,144,095,101
8.7 8.2 8.4 9.0 9.5 8.1 8.0 8.6 8.8
3,891,638,151 4,139,316,745 3,839,207,618 3,420,180,881 3,305,514,921 3,387,457,520 3,103,333,727 2,945,338,123 2,583,694,635
13.4 12.2 12.6 11.3 12.5 11.6 11.8 12.0 12.5
8,933,744,061 9,007,552,063 9,320,974,363 10,122,448,765 10,392,283,260 10,878,238,821 10,824,904,975 11,783,350,501 11,881,795,879
12.3 11.8 12.4 12.6 13.1 12.0 11.7 11.6 11.4
10,818,337,187 10,663,420,477 10,902,774,956 12,246,078,983 12,361,561,882 13,031,122,001 13,486,978,064 14,160,927,780 15,422,172,055
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates based
on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also differ from
actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is often blended from
regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported, and included in these tables, as
regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes some gasoline and blendstock that was
included in aggregate average estimates. EPA has presented averages for batches labeled as regular and premium gasoline only; excluding batches
labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as mid-grade), mix of grades, or without a grade
label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed above
these volumes are not expected to represent the actual volumes by grade of retail gasoline.
198
-------
RFG Olefins-Geographic
Change In Olefin Content from 2000 to 2005-Summer RFG Surveys
Manchester, NH (7.09,12.93)
Portsmouth-Dover, NH (6.74, 12.13)
Dallas-Fort Worth, TX (9.70,13.61)
Warren County, NJ (1039,12.82)
Washington, D.C.-area (9.78,12 .16)
Phila.-Wilrn, DE-Trenton, NJ (9.73,12.00)
Rhode Island (9.61 ,11. 62)
Baltimore, MD (10 27. 12 14)
g Houston-Galveston,TX(10.22,12.08)
o Boston- Worcester, MA (9.77 ,1 1 .52)
J Norfolk-Virginia Beach, VA (9 69,11 39)
§ Poughkeepsie, NY (11. 25,12.43)
f Springfield, MA (10 11, 11 08)
o
« Richmond, VA (10.66, 11. 58)
| NY-NJ-Longls.-CT (11. 19,12.08)
Chicago-Lake Co., IL, Gary, IN (3.97,4.70)
Milwaukee-Racine, Wl (5.59,5.88)
Hartford, CT (1 1.63, 11. 57)
Sussex County, DE (11.04,10.79)
Atlantic City, NJ (10. 98, 10. 01)
Louisville, KY (5.30,3.46)
St. Louis, MO (10. 22,7. 71)
Covington, KY (B.44,5.53)
1
H
^
I
^^^M
i
"
1
=^
M
zz:
zn
L7Z
zz
L7H
cz
1
en
CZ1
nn
D
zz
zz
zz
zn
zn
ZD
Z3
]
n
]
zz
-4 -3 -2 - 0123456"
Ui.iii'i- ill Olefin Vol %
Change in Olefin Content from 2000 to 2005-Winter RFG Surveys
Dallas-Fort Worth, IX (9. 46,12.83)
Houston-Galvestori, TX (9.30,12.17)
Manchester, NH (7.11,9.91)
Atlantic CSy.NJ (9.32,11.29)
Louisville, KY (6.68,7.96)
Boston-Worcester, MA (9.88,10.62)
Baltimore, MD (10.43,11 03)
Phila .-Wilm, DE-Trenton, NJ (10.04,10.36)
g Portsmouth-Dover, NH (9 46,9.53)
S Chicago-Lake Co., IL, Gary, IN (5.26,5.01)
^ Norfolk-Virginia Beach, VA (10.63,10.32)
§ Richmond, VA (10. 58,9. 92)
| Sussex County, DE (12.03,11. 35)
0
« Milwaukee-Racine, Wl (5.44,4.72)
| Hartford, CT(11. 58,10. 48)
Washington, D.C.-area (10 48,9.16)
Poughkeepsie. NY (11.72,10.06)
Covington, KY (8 30,6.43)
Springfield, MA (11 94,9.65)
Warren County, NJ (12.26,9.78)
St. Louis, MO (8. 95,6.20)
NY-NJ-Long Is. -CT (12.84,9.72)
Rhode Island (11.69,7.81)
1
1
1
1
I
H
^
^
^
^ZZZZ:
^^H
^^
^
1
^H
H
^^
1
3
m
-5-4-3-2-10 1 234
Cli^ncje in Olefin Vol %
Figure 7
Figure 8
199
-------
Conventional Gasoline
CG Olefins Content by Gasoline Volume
Olefins Volume Distribution Trend-Summer CG
(from Batch Reports-Excluding Blendstocks)
2000 2001 2002
Reporting Year
Figure 9
Summer CG Olefins Content (Vol %) by Volume (from Batch Reports excluding Blendstocks)
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
Reporting Year
1997
0.0
1.4
3.1
5.0
6.7
7.6
8.5
9.4
10.3
11.0
11.8
12.6
13.4
14.2
15.1
16.0
17.4
18.6
20.6
24.1
32.9
39,012,171,998
1998
0.0
1.8
3.6
5.2
6.4
7.0
7.9
8.7
9.5
10.3
11.0
11.7
12.5
13.3
14.3
15.4
16.4
17.7
19.3
21.8
32.8
39,086,529,466
1999
0.0
1.9
3.7
5.1
6.3
7.1
7.9
8.9
9.7
10.6
11.4
12.2
13.0
13.9
14.7
15.7
16.7
18.1
19.9
22.6
32.9
37,192,967,557
2000
0.2
2.3
4.3
5.5
6.6
7.4
8.3
9.1
9.7
10.5
11.2
12.1
12.9
13.8
14.7
15.5
16.4
17.5
19.0
21.3
33.0
36,389,951,364
2001
0.0
2.8
4.5
5.9
6.8
7.7
8.7
9.5
10.5
11.3
12.2
13.0
13.8
14.6
15.4
16.2
17.2
18.4
20.0
23.1
33.0
38,541,577,884
2002
0.0
2.2
3.9
5.4
6.6
7.5
8.4
9.2
10.1
11.0
11.7
12.4
13.3
14.1
14.9
15.8
16.9
18.2
19.7
22.2
32.9
40,819,234,602
2003
0.1
2.1
3.9
5.1
6.4
7.2
8.0
8.9
9.8
10.6
11.4
12.2
12.9
13.8
14.7
15.6
16.6
17.8
19.4
22.5
33.0
43,243,575,045
2004
0.1
2.0
3.8
5.0
6.0
6.8
7.6
8.4
9.1
9.8
10.6
11.4
12.1
12.9
13.7
14.8
15.8
17.2
18.7
21.1
32.8
43,497,571,635
2005
0.0
2.5
4.4
5.7
6.7
7.6
8.4
9.2
10.0
10.7
11.4
12.0
12.7
13.6
14.4
15.4
16.5
17.9
19.5
22.3
33.0
42,178,688,849
200
-------
CG Olefins Content by Gasoline Volume (continued);
Olefins Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2000 2001 2002 2003
Reporting Year
Figure 10
Winter CG Olefins Content (Vol %) by Volume (from Batch Reports excluding Blendstocks)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1997
0.0
1.4
3.1
4.8
6.2
7.2
8.2
9.1
9.9
10.8
11.7
12.7
13.5
14.4
15.3
16.2
17.2
18.6
20.4
23.8
33.0
44,301,841,176
1998
0.0
1.2
2.8
4.5
5.7
6.7
7.5
8.3
9.1
9.8
10.6
11.5
12.3
13.1
14.0
14.9
16.1
17.4
19.0
21.8
33.0
46,293,525,301
1999
0.0
1.7
3.4
4.7
5.9
6.9
7.7
8.6
9.3
10.0
10.7
11.5
12.4
13.3
14.2
15.3
16.3
17.7
19.4
22.1
33.0
47,701,715,605
2000
0.0
1.5
3.7
5.1
6.3
7.4
8.2
8.9
9.6
10.4
11.3
12.2
13.0
13.8
14.8
15.8
16.9
18.3
20.0
23.0
33.0
48,024,395,442
2001
0.0
1.8
3.9
5.3
6.5
7.6
8.6
9.5
10.3
11.1
11.9
12.8
13.6
14.5
15.4
16.5
17.5
18.7
20.7
23.8
33.0
48,786,984,011
2002
0.0
1.9
3.5
4.8
6.1
7.1
8.2
9.0
9.8
10.6
11.3
12.0
12.7
13.5
14.3
15.3
16.4
17.8
19.8
22.7
33.0
49,644,581,942
2003
0.0
1.6
3.3
4.7
5.8
6.8
7.6
8.4
9.1
9.9
10.7
11.4
12.2
13.2
14.1
15.0
16.2
17.7
19.6
22.8
33.0
47,465,390,042
2004
0.0
1.7
3.7
4.8
5.6
6.5
7.4
8.2
8.9
9.5
10.3
11.1
11.9
12.8
13.8
14.9
16.2
17.8
19.5
22.9
33.0
47,636,716,042
2005
0.0
1.7
4.1
5.3
6.3
7.2
7.9
8.7
9.4
10.0
10.8
11.4
12.2
12.9
13.7
14.7
15.9
17.3
19.4
23.2
33.0
48,690,992,232
201
-------
CG by
Grade
PRM
REG
Season
s
w
s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
CG by
6.65 5.91 6.33 6.41 6.74 5.80 6.65 6.33 6.86
6,846,643,119 6,635,688,729 6,252,416,423 4,957,386,104 4,996,778,324 5,181,939,065 5,149,486,549 4,623,975,508 4,123,526,493
6.97 6.00 6.22 6.99 7.13 6.63 5.86 6.20 6.42
8,265,699,839 8,233,277,833 7,690,219,372 6,648,441,003 6,551,541,289 6,569,323,421 5,517,761,736 5,069,256,175 4,498,707,927
14.03 13.13 13.35 13.10 13.83 13.38 12.95 12.11 12.80
27,758,756,583 28,418,412,132 26,656,367,421 27,360,707,607 28,964,139,718 30,995,888,528 33,053,399,345 34,601,650,558 33,849,816,921
13.85 12.81 12.80 13.39 13.89 12.98 12.65 12.38 12.44
31,894,223,671 32,177,203,389 33,356,627,836 35,199,495,331 35,928,338,728 36,922,839,575 35,944,676,237 37,495,132,038 39,120,750,150
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates based
on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also differ from
actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is often blended from
regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported, and included in these tables, as
regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes some gasoline and blendstock that was
included in aggregate average estimates. EPA has presented averages for batches labeled as regular and premium gasoline only; excluding batches
labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as mid-grade), mix of grades, or without a grade
label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed above
these volumes are not expected to represent the actual volumes by grade of retail gasoline.
202
-------
Appendix to Distillation Parameters Chapter
RFG E200 by Gasoline Volume
E200 Volume Distribution Trend-Summer RFG
(from Batch Reports)
Figure 1
Summer RFG E200 (%) by Volume
(from Batch Reports)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
30.6
39.1
41.2
42.8
44.0
45.1
46.0
47.1
47.9
48.7
49.3
49.9
50.7
51.2
51.8
52.4
53.2
54.3
55.3
56.8
64.9
12,832,964,637
1999
29.9
39.6
41.9
43.2
44.5
45.4
46.4
47.5
48.3
48.9
49.5
50.1
50.8
51.3
51.9
52.8
53.6
54.7
56.1
57.7
73.7
12,996,111,169
2000
31.3
40.5
41.9
42.7
43.4
44.1
44.7
45.4
46.0
46.6
47.3
47.9
48.5
49.1
49.8
50.7
51.7
52.8
54.1
55.6
65.9
12,983,168,478
2001
32.4
39.9
41.4
42.5
43.4
44.1
44.7
45.3
46.1
46.7
47.4
48.0
48.6
49.2
49.8
50.5
51.2
52.1
53.2
55.3
63.3
13,222,633,468
2002
29.0
39.4
40.9
41.7
42.5
43.1
44.0
45.0
46.0
46.6
47.3
47.9
48.6
49.3
50.0
50.8
51.9
53.2
54.7
56.4
63.1
13,847,971,634
2003
30.1
39.5
41.0
42.1
43.1
44.0
44.8
45.6
46.3
47.0
47.7
48.4
49.2
49.8
50.6
51.3
52.1
53.2
54.8
56.7
63.5
13,584,860,845
2004
31.4
40.0
41.8
42.8
43.7
44.6
45.2
45.8
46.4
47.0
47.6
48.2
48.8
49.4
50.1
50.9
51.8
52.9
54.2
56.0
63.0
14,232,658,149
2005
30.4
40.7
42.1
43.4
44.2
44.9
45.6
46.4
47.2
47.8
48.5
49.2
50.0
50.8
51.7
52.4
53.2
54.3
55.7
57.4
67.9
14,083,382,582
203
-------
RFC E200 by Gasoline Volume (continued):
E200 Volume Distribution Trend-Winter RFC
(from Batch Reports)
2001 2002
Reporting Year
Figure 2
Winter RFC E200 (%) by Volume (from Batch Reports)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
30.3
44.9
47.6
49.6
51.2
52.6
53.7
54.6
55.4
56.1
56.8
57.3
57.9
58.5
59.1
59.8
60.4
61.3
62.4
64.4
70.0
15,059,853,293
1999
31.5
45.2
48.0
49.9
51.3
52.6
53.7
54.5
55.3
56.1
56.7
57.4
58.0
58.5
59.1
59.7
60.4
61.2
62.3
64.2
73.6
15,081,898,669
2000
32.1
46.2
48.7
50.4
51.5
52.7
53.7
54.5
55.3
56.0
56.8
57.5
58.0
58.7
59.3
60.0
60.7
61.5
62.7
64.9
70.0
15,829,263,333
2001
30.9
45.6
47.9
49.6
51.1
52.3
53.4
54.2
55.0
55.8
56.5
57.0
57.6
58.2
58.9
59.6
60.3
61.2
62.4
64.1
71.8
15,724,574,173
2002
31.4
45.3
48.5
50.3
51.7
52.8
53.8
54.6
55.3
56.1
56.7
57.3
57.9
58.5
59.1
59.9
60.6
61.2
62.0
63.8
69.5
16,433,999,720
2003
30.4
45.9
48.6
50.4
51.6
52.6
53.6
54.5
55.3
56.0
56.7
57.3
57.9
58.5
59.1
59.7
60.3
61.0
62.0
63.6
70.0
16,679,773,135
2004
30.5
47.1
49.2
50.8
51.9
53.0
53.8
54.5
55.3
55.9
56.4
57.0
57.7
58.4
59.0
59.6
60.4
61.1
62.2
63.9
69.9
17,190,635,125
2005
31.3
47.0
49.3
50.8
52.0
52.9
53.8
54.7
55.5
56.3
57.0
57.6
58.3
58.7
59.2
59.8
60.4
61.1
62.0
63.2
69.1
18,044,832,058
204
-------
RFC E200 by Gasoline Volume (continued):
(Cumulative Distributions for Latest Year Data)
Percentile Chart of E200 by Volume- 2005 Summer RFC
(from Batch Reports)
E200(%)
Figure 3
Percentile Chart of E200 by Volume- 2005 Winter RFC
(from Batch Reports)
i oar. T
E200(%)
Figure 4
205
-------
RFC E200 by Grade
Average E200 of Summer RFG Sold at Retail Stations-By Grade
60
40 -
20 -
D Avg Regular
Avg Mid
D Avg Premium
"I" Bars denote range of RFG
iaae
49.0
44. B
iaaa
49.2
45.2
2000
47.4
44.2
20D1
47.0
43.7
2002
46.9
42.9
2003
47.1
43.2
2004
47.2
43.7
2005
47.0
43.8
Year
Figure 5
LJ_
T3
I
O
Q.
(0
O
8
iii
70 -
60 -
50 -
40 -
30 -
20 -
10 -
0 -
D Avg Regular
Avg Mid
D Avg Premium
Average E200 of Winter RFG Sold at Retail Stations-By Grade
]
[
iaaa
59.0
56.1
51.6
I
I
1999
58.2
56.0
51.4
T
2000
583
560
51.3
Bars
I
1
2001
58.3
55.3
51.1
di
not
e range
T
1
2002
58.3
560
51.3
of RFC
>
; ac
2003
53
55.8
51.1
ea e
verage
2004
57.9
55.5
52.6
3
2005
53.5
56.0
51.5
Year
Figure 6
206
-------
RFG by,
and bf and
Grade Season
PRM s
w
REG s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
43.95
3,649,195,434
50.76
4,139,316,745
50.73
9,007,552,063
57.95
10,638,539,349
44.24
3,537,723,033
50.73
3,839,207,618
51.09
9,320,974,363
57.83
10,902,774,956
43.75
2,788,564,977
51.14
3,418,369,505
48.76
10,122,448,765
57.72
12,246,078,983
43.20
2,773,110,841
50.47
3,301,129,239
48.60
10,389,949,614
57.31
12,307,517,794
42.44
2,931,886,907
50.58
3,387,457,520
48.88
10,879,031,865
57.22
13,031,122,001
42.97
2,736,973,386
50.64
3,103,333,727
49.13
10,824,904,975
57.27
13,486,978,064
43.09
2,385,535,898
51.29
2,943,555,811
48.86
11,783,350,501
57.21
14,160,927,780
43.08
2,144,095,101
50.91
2,583,694,635
49.90
11,881,795,879
57.18
15,422,172,055
Users of these grade-specific property estimates calculated from RFG and Anti-Dumping data should be aware that grade-specific
estimates based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting
system may also differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-
grade gasoline is often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume
reported and included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis
excludes some gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as
regular and premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline
marketed as mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if
they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
207
-------
RFC E200-Geographic
Change in E200 (% Evaporated @ 200F) from 2000 to 2005-Summer
RFC Surveys
Portsmouth-Dover, NH (47.3,50 7)
Manchester, NH (49.0,51.8)
Boston-Worcester, MA (47 3,50.5)
Houston-Galveston, TX (47.7,50 2)
St Louis, MO (48.2,49.8)
Dallas-Fort Worth, TX (50.3,51 .7)
Rhods Island (49. 1,50.5)
g Baltimore, MD (48.7,50.0)
J- Milwaukee-Racine, Wl (45.1 ,46.2)
§ Washington, D.C.-area (49.4,50.2)
1 Norfolk-Virginia Beach, VA (49.0,49 6)
g- Atlantic City, NJ (48.1 ,48.7)
,- Warren County, NJ (47.5,480)
§ NY-NJ-Long Is -CT (47 3,47 7)
" Chicago-Lake Co , IL, Gary, IN (46.2,46.2)
^ Covington, KY (46 4,46 3)
Richmond, VA (49.2,49.0)
Phila.-Wilm, DE-Trenton, NJ (48.7,48.4)
Poughkeepsie, NY (46.B.46.2)
Louisville. KY (47.3,46.5)
Springfield, MA (49. 4,48 5)
Sussex County, DE (45.5,445)
Hartford, CT (48.6,47.2)
_
1
^m
^^
^M
^^m
m^^m
^^m
z^=
^^
^Z=
.1
^^M
i
^H
^^^
^H
^H
CZ3
^
2 -1 0
D
234
Change ill '* Ev.i|>o[.ite«l (a> 200F
Figure 7
Change in E200 (% Evaporated @ 200F) from 2000 to 2005-Winter
RFC Surveys
Sussex County, DE (53.5,57.3)
Warren County. NJ (54.3.58.0)
Norfolk-Virginia Beach, VA (54.0,57.2)
Dallas-Fort Worth, TX (55.6,58.2)
Houston-Galveston, TX (54.3,56.1)
Richmond, VA (54.9 ,56. 6)
Washington, D.C.-area (54.4,55.9)
^ Rhode Island (57.8,58 7)
| Springfield, MA (57.6,58.4)
§. Louisville, KY (60.0,60.8)
| Baltimore, MD (55.9,56. 6)
§ Phila.-Wilm, DE-Trenton, NJ (55.3,55.7)
Manchester. NH (57.6,57 9)
S NY-NJ-Long Is. -CT (56.6,56, 9)
« Atlantic City, NJ (56.5,56.6)
1 Hartford, CT (57.3,57.0)
Boston-Worcester, MA (57.7,57.3)
Poughkeepsie, NY (54. 7,54. 2)
Covington, KY (58.9,57.9)
Milwaukee-Racine, Wl (59.9,58.7)
St Louis. MO (55.8,54.4)
Chicago-Lake Co., IL, Gary, IN (59.8,58.0)
Portsmouth-Dover. NH (58 3.55.7)
1
^m
^M
^m
^H
I
^H
^H
=
^
]
3-2-1 D
^Zl
^
1
2345
Ch.inye in % Ewlnol.ited £' 200F
Figure 8
208
-------
RFC E300 by Gasoline Volume
E300 Volume Distribution Trend-Summer RFG
(from Batch Reports)
2DD1 2002
Reporting Year
Figure 9
Summer RFG E300 (%) by Volume
(from Batch Reports)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
63.9
76.1
77.7
78.6
79.3
79.8
80.3
80.8
81.3
81.8
82.3
82.7
83.2
83.8
84.4
85.0
85.9
86.7
87.9
89.6
100.0
12,832,964,637
1999
70.2
76.5
78.1
79.0
79.7
80.2
80.7
81.2
81.7
82.2
82.6
83.1
83.5
84.0
84.6
85.1
85.9
86.7
87.7
89.2
100.0
12,996,111,169
2000
74.5
79.2
80.1
80.8
81.3
81.8
82.3
82.8
83.3
83.8
84.3
84.9
85.3
85.8
86.4
87.0
87.8
88.6
89.7
91.1
99.0
12,983,168,478
2001
74.7
79.3
80.1
80.7
81.2
81.8
82.3
82.8
83.3
83.8
84.2
84.7
85.1
85.5
86.0
86.4
86.9
87.6
88.5
89.8
99.9
13,222,633,468
2002
72.4
79.1
80.2
80.8
81.3
81.8
82.3
82.7
83.1
83.6
84.1
84.6
85.1
85.5
86.0
86.5
86.9
87.7
88.7
90.3
99.7
13,847,971,634
2003
74.8
79.3
80.1
80.7
81.3
81.8
82.1
82.5
83.0
83.5
84.0
84.5
85.0
85.6
86.1
86.7
87.2
88.0
88.9
90.1
100.0
13,584,860,845
2004
74.4
78.2
79.2
79.8
80.4
80.9
81.2
81.6
82.0
82.4
82.8
83.3
83.9
84.5
85.1
85.8
86.4
87.3
88.2
89.7
100.0
14,232,658,149
2005
73.8
79.1
79.9
80.5
81.1
81.6
82.0
82.4
82.7
83.1
83.5
83.9
84.4
85.0
85.6
86.3
87.0
87.9
88.9
90.1
100.0
14,083,382,582
209
-------
RFC E300 by Gasoline Volume (continued):
E300 Volume Distribution Trend-Winter RFG
(from Batch Reports)
2001 2002
Reporting Year
Figure 10
Winter RFG E300 (%) by Volume (from Batch Reports)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
64.4
77.6
79.5
80.8
81.7
82.4
82.9
83.3
83.8
84.3
84.9
85.4
85.9
86.5
87.1
87.7
88.2
88.9
89.9
91.3
99.5
15,059,853,293
1999
70.5
77.6
79.2
80.4
81.2
82.0
82.6
83.1
83.6
84.0
84.4
84.9
85.5
86.0
86.6
87.3
88.0
88.8
89.9
91.4
100.0
15,081,898,669
2000
71.0
79.1
80.8
81.9
82.8
83.5
84.0
84.5
84.9
85.4
85.9
86.5
87.0
87.5
88.1
88.8
89.4
90.2
91.2
92.6
99.1
15,829,263,333
2001
67.5
78.7
80.0
81.2
82.0
82.7
83.5
84.0
84.6
85.1
85.7
86.2
86.8
87.5
88.1
88.8
89.5
90.3
91.2
92.7
99.4
15,724,574,173
2002
71.0
78.7
80.3
81.4
82.2
82.9
83.5
84.0
84.5
85.0
85.4
85.8
86.4
87.0
87.6
88.3
89.0
90.1
91.1
92.5
100.0
16,433,999,720
2003
64.1
78.6
80.0
81.0
81.7
82.4
82.9
83.5
84.0
84.5
84.9
85.4
85.9
86.6
87.2
87.9
88.5
89.3
90.3
91.8
100.0
16,679,773,135
2004
69.6
78.1
79.8
80.8
81.5
82.2
82.8
83.3
83.8
84.3
84.8
85.3
85.7
86.2
86.8
87.4
88.3
89.1
90.2
91.8
100.0
17,190,635,125
2005
70.3
79.0
80.5
81.5
82.2
82.9
83.4
83.8
84.3
84.7
85.2
85.7
86.2
86.7
87.2
87.9
88.4
89.2
90.0
91.2
100.0
18,044,832,058
210
-------
Percentile Chart of E300 by Volume- 2005 Summer RFC
(from Batch Reports)
Figure 11
65
Percentile Chart of E300 by Volume- 2005 Winter RFC
(from Batch Reports)
E300{%)
Figure 12
211
-------
RFC E300 by Grade
Average E300 of Summer RFC Sold at Retail Statlons-By Grade
Year
Figure 13
Average E300 of Winter RFC Sold at Retail Stations-By Grade
"I" Bars denote range of RFG area average
Year
Figure 14
212
-------
ifrf '-..,, -
Grade
PRM
REG
Season
s
w
s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
and bf and
83.70 84.14 86.85 86.27 85.86 86.42 85.62 86.05
3,649,195,434 3,537,723,033 2,788,564,977 2,773,110,841 2,931,886,907 2,736,973,386 2,385,535,898 2,144,095,101
85.23 85.72 86.96 86.22 86.99 86.15 86.31 86.68
4,139,316,745 3,839,207,618 3,420,180,881 3,301,129,239 3,387,457,520 3,103,333,727 2,943,555,811 2,583,694,635
82.16 82.38 84.09 83.85 83.98 83.87 82.97 83.72
9,007,552,063 9,320,974,363 10,122,448,765 10,389,949,614 10,879,031,865 10,824,904,975 11,783,350,501 11,881,795,879
84.79 84.24 85.85 85.65 85.11 84.91 84.67 85.12
10,638,539,349 10,902,774,956 12,246,078,983 12,307,517,794 13,031,122,001 13,486,978,064 14,160,927,780 15,422,172,055
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates
based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also
differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is
often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported and
included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes
some gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as regular and
premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as
mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a
grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
213
-------
RFC E300 Geographic
Change in E300 (% Evaporated @ 300F) from 2000 to 2005-Summer
RFG Surveys
St. Louis. MO (83. 2,87.1)
Covington. KY (85.5.87.9)
Louisville, KY (86.2,87.4)
Rhode Island (85.5,86.4)
Atlantic City, NJ (84. 4, 84.8)
NY-NJ-Long Is.-CT (85 0,85.4)
Poughkeepsie, NY (84.6,84.8)
Sussex County, DE (81.6,81.5)
o Warren County, NJ (85.0,84.9)
| Dallas-Fort Worth, TX (83. 8,83.6)
-'- Manchester, NH (BS.8.8B.3)
Springfield, MA (86.1,85.5)
ft
> Phila.-Wilrn, DE-Trenton, NJ (84.4,83.7)
^ Norfolk-Virginia Beach. VA (84.5.83.7)
1 Hartford. CT (86.4,85.6)
Baltimore, MD (84.2,83.3)
Houston-Galveston, TX (84.1 ,83.1)
Washington, D.C-area (84.0,82.6)
Boston-Worcester, MA (86.7,85.3)
Richmond, VA (84.3,82.8)
Portsmouth-Dover, NH (86.1 ,83.6)
Milwaukee-Racine, Wl (85.0,82.3)
Chicago-Lake Co., IL. Gary, IN (85.9,83.1)
i
M
^
^
^
^
^m
^m
1
i
=
1
i ; ]
'- - i
1 -3 -2 -1 0
2345
Ch.inye in % Ev.l|>or.lteil .:S>300F
Change in E300 (% Evaporated @ 300F) from 2000 to 2005-Winter
RFG Surveys
Warren County, NJ [85.6,87.0)
Norfolk-Virginia Beach, VA 184.5,85.4)
Covington.KY (87.7 ,687)
Sussex County, DE (B2.8,83.6)
Houston-Galveston, TX (83.6,84.4)
Poughkeepsie, NY |B6.5,86.9)
Dallas-Fort Worth, TX |B5.4,B5.4)
Rhode Island pB.3 ,88.3)
S Baltimore, MD P5. 6,85. 5)
H
| Louisville, KY[B8. 4,88. 3)
? Phila.-Wilm, DE-Trenton, NJ p5 3,85 1)
| Chicago-Lake Co., IL, Gary, IN p55,85 1)
Springfield, MA p8.4,88.D)
J Milwaukee-Racine , Wl (85.8 ,85.4)
1 Atlantic City, NJ (86.1 ,85.7)
Washington, D C.-area P53.845)
St. Louis, MO P3.6,62.B)
NY-NJ-Long Is.-CT (87.3.86.4)
Hartford, CT p7. 9,86.8)
Boston- Worcester, MA pB 3,87.2)
Richmond, VAP5.2,B4.0)
Manchester, NH [B9 6, 86 8)
Portsmouth-Dover, NH (906,85.8)
-
1
1
I
3 -5 -4 -3 -2
1
I
^
^m
^m
^M
^^
^^
t
1
1 1
^H
i
1
Change n % Evaporated 'tiOOOF
Figure 17
Figure 18
214
-------
RFC T50 Geographic
Change in T50 (50% Evaporat
Summe
Hartford, CT (203. 0,207.6)
Poughkeepsie, NY (206.7 ,21 1 .0)
Sussei County , DE (21 0.7, 21 3.6)
Springfield, MA (201 .6,203.5)
Louisville, KY (205.4,206.9)
Chicago-Lake Co., IL, Gary, IN (208 3,209 7)
Phila.-Wilm, DE-Trenton, NJ (202.8203.7)
Richmond. VA (201 .6,202.1)
§ NY-NJ-Long Is -CT (205. 9,206.1)
8 Covington,KY(208.1,207.7)
IT Warren County, NJ (205.3,204.5)
Milwaukee-Racine, Wl (210.6,209.7)
? Atlantic City, NJ (204. 3,203.1)
S
I- Norfolk-Virginia Beach, VA (202.2,200.5)
1 Washington, D.C.-area (201.3,199.4)
Baltimore. MD (202.6,199 .8)
Rhode Island (202.2,199.2)
Dallas-Fort Worth, TX (199.7,196.2)
St. Louis, MO (204.1 ,199.7)
Boston-Worcester, MA (204.3, 199.0)
Houston-Gakeston, TX (205.2,199.3)
Manchester, NH (202.3, 196. 3)
Portsmouth-Dover, NH (205.5,198.2)
on Temperature) from 2000 to 2005-
RFG Surveys
^
^M
^M
1
^H
1
H
1
^^5
. ]
1
-8-6-4-20246
Ch.inge in T50 (iegiees F
Change in T50 (50% Evaporation Temperature) from 2000 to 2005-
Winter RFG Surveys
Chicago-Lake Co., IL, Gary, IN (160.6,165.0)
Milwaukee-Racine, Wl (159.6,162.4)
Portsmouth-Dover, NH (184.0,185.6)
St. Louis, MO (178. 6, 178. 6)
Atlantic City, NJ (183. 7, 183. 5)
Boston-Worcester, MA (182.9,182 4)
Covington,KY(169.1,167.4)
Phila.-Wilm, DE-Trenton, NJ (186.7,184.5)
S Baltimore, MD (185.4,183.1)
§ Rhode Island (182.0,179.3)
S! Manchester, NH (184. 6,181. 2)
Houston-Galveston, TX (189.0,185.3)
2 Washington, D.C.-area (188.9,184.4)
8
Dallas-Fort Worth, TX (186. 6, 181 .5)
Richmond, VA (187.5,182.4)
Springfield, MA (163 0,176 4)
Norfolk-Virginia Beach. VA (189.7 .1 82. 1 )
Poughkeepsie, NY (188.0,179.6)
NY-NJ-Long Is -CT (183.6,173.0)
Sussex County, DE (190.8,179.3)
Warren County, NJ (189.5,177.4)
Louisville, KY (172. 7,158. 2)
Hartford, CT (183. 5,168. 5)
-2
1
^
^
1
"
^M
^^
^^
^^
^M
1
]
]
|
0 -15 -10 -50 5 10
Change in T50 ileyiees F
Figure 19
Figure 20
215
-------
RFC T90 Geographic
Change in T90 (90% Evaporation Temperature) from 2000 to 2005-
Summe
Milwaukee-Racine, Wl (326. 1,341. 7)
Portsmouth-Dover, NH (315.1 ,330.5)
Chicago-Lake Co. , IL, Gary, IN (323.B.337.7)
Boston-Worcester, MA (315.4,321.9)
Richmond, VA (330.2,336.1)
Sussex County, DE (327.6,333.0)
Manchester. NH (318. 7,322.0)
Washington, D.C. -area (333.4,336.7)
§ Houston-Galveston, IX (329.6,332.9)
S Baltimore, MD (330 3,333 0)
!? Phrla.-Wilm, DE-Trenton, NJ (326.8,328.6)
Norfolk-Virginia Beach. VA 029.9,331. 2)
^ Hartford, CT (318.0,319.0)
^ Springfield, MA (320.6,319.8)
£ Warren County, NJ (324.9,323.9)
it
Poughkaspsie, NY (325.4,3236)
Dallas-Fort Worth, IX (334.2,332.4)
Atlantic City, NJ (325.6,323.2)
NY-NJ-Long Is.-CT (324.1 ,321.4)
Louisville, KY (320.1, 316.0)
Rhode Island (323.4,315.8)
Covington,KY (322.6,314.2)
St. Louis, MO (336. 3,318.8)
-2
r RFG Surveys
^m
^
"
i
i
IM
^H
^H
^^
]
^H
H
H
D
3
3
1
]
0 -15 -10 -5 0 5 10 15 20
Change in T90 degiees F
Change in T90 (90% Evaporat
Winter
Portsmouth-Dover, NH (295.0,322.2)
Manchester, NH (301.4,317.4)
Richmond, VA (325.2,332 7)
Boston-Worcester, MA (307.7,315 2)
Louisville, KY (306.8 ,311.1)
Hartford, CT (31 0.6 ,314.9)
Milwaukee-Racine, Wl (325.5,329.1)
Washington, D.C. -area (326.3,329.7)
§ NY-NJ-Longls-CT(314.1,317.3)
| Springfield, MA (307 3.310.1)
i St. Louis, MO (330.3,332.9)
i Chicago-Lake Co., IL, Gary, IN (327.0,329.0)
Atlantic City, NJ (319. 6,321 .4)
P Dallas-Fort Worth, TX (324 1,325 8)
S Phila.-Wilm. DE-Trenton, NJ (323.3,324.5)
Sussex County, DE (328.5,329.4)
Rhode Island (307.3,308.1)
Houston-Galvsston, TX (329.5,328.9)
Baltimore. MD (324.1 ,323 4)
Norfolk-Virginia Beach, VA (326.2,323.7)
Poughkeepsie, NY (317 4,314.1)
Covington.KY £311. 8,308.5)
Warren County, NJ (323.7,314 6)
-1
ion Temperature) from 2000 to 2005-
RFG Surveys
^m
1
1
^m
czz
^m
^
^
i
^
^
n
3
i
3
H=
5 -10 -5 0 5 10 15 20 25 30
Change in T90 deijiees F
Figure 21
Figure 22
216
-------
Conventional Gasoline
CG E200 by Gasoline Volume
E200 Volume Distribution Trend-Summer CG
(from Batch Reports-Excluding Blendstocks)
2001 2002
Reporting Year
Figure 23
Summer CG E200 (%) by Volume (from Batch
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
Reporting Year
1998
27.0
32.1
35.4
37.8
39.6
40.9
41.9
42.9
43.5
44.3
45.0
45.8
46.4
47.2
47.9
48.9
50.0
51.3
52.9
54.9
68.9
39,086,644,042
1999
27.0
32.6
36.0
38.5
40.0
41.3
42.2
43.1
43.9
44.6
45.3
46.0
46.6
47.3
48.2
49.0
50.0
51.4
52.7
54.9
76.0
37,192,967,557
2000
27.0
34.5
37.3
38.9
40.0
41.1
42.2
43.1
43.8
44.6
45.3
45.9
46.7
47.4
48.1
49.0
49.8
50.9
52.1
54.1
64.6
36,389,951,364
2001
27.0
34.8
37.7
39.3
40.4
41.3
42.1
42.9
43.7
44.4
45.2
45.9
46.5
47.2
47.9
48.9
49.9
50.9
52.4
54.3
69.8
38,542,457,289
Reports excluding Blendstocks)
2002
27.0
34.8
37.6
39.1
40.3
41.2
42.1
42.8
43.5
44.2
44.9
45.6
46.2
47.0
47.9
48.8
49.7
50.5
51.6
53.5
76.4
40,819,234,602
2003
27.0
34.7
37.4
39.0
40.3
41.3
42.2
43.0
43.8
44.5
45.2
45.9
46.5
47.2
47.9
48.7
49.6
50.9
52.3
54.6
76.8
43,243,575,045
2004
27.0
35.0
37.7
39.2
40.3
41.3
42.1
42.8
43.5
44.2
44.8
45.5
46.3
46.9
47.8
48.8
49.8
50.9
52.5
54.4
73.5
43,509,104,037
2005
27.0
35.4
38.1
39.6
40.7
41.7
42.5
43.3
44.0
44.7
45.4
46.1
46.9
47.6
48.4
49.2
50.2
51.5
53.1
55.1
67.0
42,196,228,805
217
-------
CG E200 by Gasoline Volume (continued):
E200 Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2001 2002
Reporting Year
Figure 24
Winter CG E200 (%) by Volume (from Batch Reports Excluding Blendstocks)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
27.0
35.90
39.10
41.70
44.00
45.60
46.90
47.90
48.90
49.70
50.50
51.20
51.90
52.60
53.50
54.50
55.60
56.70
58.10
60.40
75.50
46,293,637,987
1999
27.0
35.90
39.40
42.10
44.20
45.60
46.80
47.90
48.80
49.60
50.40
51.30
51.90
52.60
53.40
54.20
55.30
56.50
57.90
60.20
75.50
47,701,715,605
2000
27.0
37.50
40.90
43.10
44.70
45.80
47.00
48.00
48.90
49.70
50.40
51.20
51.90
52.70
53.50
54.30
55.20
56.40
57.80
59.70
76.30
48,030,024,114
2001
27.0
37.60
40.50
42.70
44.20
45.40
46.50
47.50
48.40
49.20
50.00
50.80
51.60
52.30
53.10
54.00
54.80
55.90
57.20
58.80
76.90
48,810,395,817
2002
27.0
37.80
41.30
43.00
44.50
45.60
46.60
47.60
48.50
49.30
50.10
50.90
51.60
52.30
53.20
54.00
55.00
56.00
57.40
59.00
75.90
49,655,135,030
2003
27.0
38.20
41.50
43.60
45.00
46.20
47.10
48.10
48.80
49.70
50.40
51.10
51.80
52.40
53.20
54.00
55.10
56.00
57.30
59.30
75.40
47,469,723,098
2004
27.0
38.90
42.10
43.90
45.40
46.60
47.50
48.40
49.20
49.90
50.70
51.40
52.00
52.70
53.50
54.40
55.30
56.30
57.50
59.30
76.80
47,654,973,148
2005
27.0
38.70
41.90
43.90
45.40
46.60
47.50
48.40
49.10
49.90
50.60
51.40
52.10
52.90
53.60
54.40
55.40
56.30
57.40
59.20
73.60
48,705,329,856
218
-------
Percentile Chart of E200 by Volume- 2005 Summer CG
(from Batch Reports-Excluding Blendstocks)
30 35 40 45 50 55
65 70
Figure 25
Percentile Chart of E200 by Volume- 2005 Winter CG
(from Batch Reports-Excluding Blendstocks)
45 50 55
E200(%)
60
75
Figure 26
219
-------
CG (by
CG and by and
Grade Season
PRM s
w
REG s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
36.7
6,635,803,305
40.5
8,233,390,519
46.3
28,418,412,132
51.9
32,177,203,389
37.9
6,252,416,423
40.7
7,690,219,372
46.6
26,656,367,421
51.6
33,356,627,836
37.8
4,957,386,104
41.2
6,648,441,003
46.0
27,360,707,607
51.4
35,199,495,331
38.1
4,996,778,324
41.2
6,555,892,276
46.1
28,965,019,123
50.7
35,941,182,455
37.4
5,181,939,065
41.1
6,571,003,043
45.9
30,995,888,528
51.0
36,931,713,041
37.8
5,149,486,549
41.4
5,517,761,736
46.1
33,053,399,345
51.0
35,949,009,293
37.8
4,635,507,910
42.1
5,077,869,493
45.8
34,601,650,558
51.1
37,504,775,826
37.2
4,138,979,931
41.1
4,510,104,837
46.3
33,851,903,439
50.9
39,123,690,864
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates
based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also
differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is
often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported and
included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes
some gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as regular and
premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as
mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a
grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
220
-------
CG E300 by Gasoline Volume
E300 Volume Distribution Trend-Summer CG
(from Batch Reports-Excluding Blendstocks)
2001 2D02
Reporting Year
Figure 27
Summer CG E300 (%) by Volume (from Batch Reports Excluding Blendstocks)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (call:
1998
63.2
73.8
75.4
76.5
77.4
78.1
78.6
79.2
79.7
80.2
80.7
81.1
81.6
82.1
82.8
83.4
84.0
84.9
85.9
87.6
100.0
39,086,644,042
1999
63.0
74.3
75.9
77.1
77.9
78.5
78.9
79.4
79.9
80.4
80.9
81.4
81.8
82.3
82.8
83.4
84.0
84.8
85.9
87.6
98.6
37,192,967,557
2000
63.1
73.1
74.9
76.1
77.0
77.7
78.4
78.9
79.4
79.8
80.3
80.8
81.3
81.9
82.4
83.0
83.7
84.6
85.7
87.5
98.2
36,389,951,364
2001
63.0
74.0
75.8
77.0
77.9
78.5
79.0
79.5
79.9
80.4
80.9
81.2
81.7
82.3
82.9
83.5
84.2
84.9
85.9
87.9
100.0
38,542,457,289
2002
63.1
73.4
75.0
76.2
77.1
77.8
78.3
78.8
79.3
79.8
80.2
80.8
81.3
81.8
82.3
83.0
83.8
84.8
85.9
88.1
99.4
40,819,234,602
2003
64.4
73.7
75.3
76.3
77.1
77.8
78.4
78.9
79.5
80.0
80.4
80.8
81.4
82.1
82.7
83.3
84.0
84.9
86.1
88.2
98.3
43,243,575,045
2004
64.5
73.0
74.8
76.1
77.0
77.7
78.4
79.0
79.6
80.1
80.7
81.2
81.7
82.1
82.6
83.1
83.8
84.7
85.8
87.4
98.2
43,509,104,037
2005
63.6
74.0
75.7
77.0
78.1
78.9
79.5
80.0
80.5
81.0
81.5
82.0
82.5
83.1
83.6
84.2
84.9
85.6
86.8
88.3
100.0
42,196,228,805
221
-------
CG E300 by Gasoline Volume (continued):
E300 Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2000 2001 2002
Reporting Year
Figure 28
Winter CG E300(%) by Volume (from Batch
Reports Excluding Blendstocks)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume Cqall:
1998
63.4
75.3
77.0
78.3
79.2
80.0
80.7
81.4
82.0
82.6
83.1
83.7
84.2
84.8
85.4
86.0
86.8
87.7
88.7
90.3
100.0
46,293,637,987
1999
63.4
74.8
77.0
78.4
79.3
80.1
80.7
81.3
81.8
82.3
82.9
83.3
83.9
84.4
85.0
85.6
86.4
87.4
88.6
90.6
100.0
47,701,715,605
2000
63.4
75.9
77.8
79.2
80.0
80.7
81.3
81.8
82.4
82.8
83.3
83.8
84.4
84.9
85.5
86.2
86.9
87.9
89.0
90.9
100.0
48,030,024,114
2001
63.4
75.8
77.6
78.9
79.7
80.4
81.0
81.5
82.0
82.5
83.0
83.5
83.9
84.5
85.1
85.7
86.4
87.4
88.7
90.9
100.0
48,810,395,817
2002
63.4
75.9
77.6
78.6
79.4
80.1
80.8
81.4
81.9
82.4
82.9
83.5
84.0
84.5
85.1
85.6
86.4
87.2
88.5
90.4
100.0
49,655,135,030
2003
63.4
75.7
77.3
78.3
79.2
79.9
80.5
81.1
81.7
82.1
82.7
83.2
83.7
84.1
84.7
85.3
86.0
87.0
88.2
90.2
100.0
47,469,723,098
2004
63.4
75.5
77.2
78.6
79.6
80.4
81.1
81.7
82.2
82.8
83.2
83.7
84.2
84.7
85.3
85.9
86.6
87.6
88.8
90.4
100.0
47,654,973,148
2005
63.4
76.8
78.5
79.7
80.5
81.2
81.8
82.4
82.9
83.3
83.8
84.2
84.7
85.2
85.7
86.3
87.1
88.1
89.2
90.9
100.0
48,705,329,856
222
-------
Percentile Chart of E300 by Volume- 2005 Summer CG
(from Batch Reports- Excluding Blenclstocks)
E300(%)
Figure 29
Percentile Chart of E300 by Volume- 2005 Winter CG
(from Batch Reports-Excluding Blenclstocks)
Figure 30
223
-------
CG by
Grade Season
PRM s
w
REG s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
C
81.9
6,635,803,305
83.6
8,233,390,519
80.5
28,418,412,132
82.9
32,177,203,389
G E30
82.4
6,252,416,423
83.6
7,690,219,372
80.7
26,656,367,421
82.8
33,356,627,836
0 bf
81.6
4,957,386,104
84.2
6,648,441,003
80.0
27,360,707,607
83.2
35,199,495,331
83.2
4,996,778,324
83.7
6,555,892,276
80.5
28,965,019,123
82.9
35,941,182,455
82.1
5,181,939,065
84.0
6,571,003,043
80.1
30,995,888,528
82.8
36,931,713,041
I
82.8
5,149,486,549
83.7
5,517,761,736
80.2
33,053,399,345
82.5
35,949,009,293
82.4
4,635,507,910
83.5
5,077,869,493
80.2
34,601,650,558
82.9
37,504,775,826
83.2
4,138,979,931
85.2
4,510,104,837
81.2
33,851,903,439
83.6
39,123,690,864
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be aware that grade-specific estimates
based on reporting data may differ from actual retail property values by grade. Although aggregate estimates from the reporting system may also
differ from actual retail property values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is
often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the gasoline volume reported and
included in these tables as regular or premium would be marketed as mid-grade. Additionally, EPA's grade-specific average analysis excludes some
gasoline and blendstock that was included in aggregate average estimates. EPA has presented averages for batches labeled as regular and
premium gasoline only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of gasoline marketed as
mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock batches from grade-specific analyses even if they had a
grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In part, because of factors discussed
above these volumes are not expected to represent the actual volumes by grade of retail gasoline.
224
-------
Appendix to Emissions and Emissions Performance Chapter
RFG VOC Performance by Gasoline Volume
VOC Reduction Volume Distribution Trend-Summer RFG
(from Batch Reports)
2DD1 2002
Reporting Year
Figure 1
Summer RFG VOC Reduction (%) by Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
Cqall:
1998
-5.7
11.0
12.5
13.6
14.4
15.1
15.7
16.4
17.1
17.7
18.5
19.4
20.4
21.5
22.8
24.1
25.0
26.0
27.0
28.1
35.2
12,551,351,739
1999
-2.7
11.1
12.6
13.6
14.4
15.1
16.0
16.8
17.5
18.4
19.3
20.4
21.4
22.5
23.4
24.6
25.6
26.5
27.4
28.6
33.3
12,605,647,535
2000
21.9
25.4
26.1
26.6
27.0
27.2
27.5
27.8
28.0
28.2
28.5
28.7
29.0
29.3
29.5
29.8
30.1
30.5
30.9
31.6
33.9
12,924,625,994
2001
21.4
25.0
25.8
26.4
26.8
27.2
27.5
27.8
28.0
28.3
28.5
28.8
29.0
29.2
29.5
29.7
30.0
30.4
30.8
31.4
33.6
13,215,304,396
2002
20.2
25.0
25.7
26.2
26.6
26.9
27.2
27.5
27.7
28.0
28.2
28.5
28.7
28.9
29.1
29.4
29.7
30.1
30.5
31.1
33.6
13,838,063,327
2003
21.7
25.0
25.8
26.2
26.6
27.0
27.3
27.5
27.8
28.1
28.3
28.5
28.8
29.0
29.3
29.5
29.8
30.1
30.5
31.2
34.8
13,583,809,485
2004
21.5
25.1
25.7
26.2
26.6
27.0
27.3
27.5
27.8
28.0
28.2
28.5
28.7
29.0
29.2
29.5
29.8
30.1
30.6
31.3
34.4
14,230,001,733
2005
22.2
25.0
25.7
26.2
26.5
27.0
27.3
27.6
27.8
28.0
28.2
28.5
28.7
29.0
29.2
29.5
29.8
30.1
30.5
31.1
33.6
14,070,501,006
225
-------
Percentile Chart of VOC Reduction by Volume- 2005 Summer RFG
(from Batch Reports)
28 30 32
VOC Reduction (%)
Figure 2
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
"
24.3
27.0
27.5
27.9
28.1
28.4
28.7
29.0
29.1
29.3
29.5
29.7
29.9
30.1
30.3
30.6
30.8
31.1
31.5
32.1
33.9
5,710,929,463
, \
24.0
26.9
27.5
27.8
28.1
28.4
28.6
28.8
29.0
29.1
29.3
29.5
29.7
29.9
30.2
30.4
30.7
31.0
31.3
31.8
33.6
5,909,480,958
24.6
27.0
27.3
27.6
27.9
28.2
28.5
28.6
28.8
29.0
29.2
29.4
29.6
29.8
30.0
30.2
30.5
30.7
31.1
31.6
33.6
6,017,416,713
r . ' 'tin
23.6
26.9
27.4
27.8
28.1
28.4
28.6
28.8
29.0
29.2
29.4
29.5
29.7
29.9
30.1
30.3
30.5
30.8
31.1
31.6
34.8
6,095,268,147
r;.:: '
25.0
26.9
27.3
27.6
27.9
28.1
28.3
28.6
28.8
29.0
29.2
29.4
29.6
29.8
30.0
30.3
30.5
30.8
31.3
31.9
34.4
6,329,701,695
25.1
27.2
27.6
27.9
28.1
28.4
28.6
28.8
29.0
29.2
29.3
29.5
29.6
29.8
30.1
30.2
30.5
30.7
31.0
31.6
33.6
5,983,085,189
226
-------
2
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume(gal):
21.9
24.8
25.6
26.0
26.4
26.6
26.9
27.1
27.3
27.5
27.7
27.9
28.1
28.3
28.5
28.7
29.0
29.4
29.9
30.7
33.9
7,213,696,531
22.2
25.2
25.9
26.3
26.7
27.0
27.3
27.5
27.7
27.9
28.1
28.3
28.6
28.7
29.0
29.3
29.5
29.8
30.1
30.6
32.8
5,680,076,425
bf
23.4
25.3
25.9
26.2
26.5
26.7
26.9
27.2
27.3
27.5
27.7
27.9
28.1
28.3
28.5
28.8
29.0
29.3
29.6
30.1
32.6
6,070,322,689
23.4
25.5
26.0
26.3
26.5
26.8
27.0
27.2
27.4
27.6
27.8
27.9
28.1
28.3
28.5
28.7
29.0
29.2
29.6
30.0
33.6
5,978,921,094
23.4
25.3
25.8
26.2
26.4
26.7
27.0
27.3
27.4
27.6
27.8
28.0
28.2
28.4
28.6
28.8
29.0
29.3
29.7
30.2
33.1
6,160,483,137
23.3
25.0
25.8
26.2
26.4
26.8
27.0
27.2
27.4
27.6
27.7
27.8
28.1
28.2
28.5
28.7
29.0
29.3
29.7
30.2
32.6
6,247,759,799
2-
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
'
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
₯OC" RF
21.4
23.1
23.9
24.2
24.6
24.9
25.1
25.4
25.6
25.8
26.0
26.3
26.5
26.7
26.8
27.0
27.3
27.6
27.9
28.4
31.1
1,625,747,013
G ₯OC
20.2
23.0
23.7
24.0
24.4
24.8
25.0
25.1
25.4
25.6
25.7
25.9
26.1
26.4
26.6
26.8
27.1
27.5
27.9
28.6
29.8
1,639,100,323 1
(%) bf
21.7
23.5
24.1
24.4
24.6
24.8
25.0
25.1
25.3
25.4
25.6
25.7
25.9
26.1
26.2
26.5
26.7
27.0
27.2
27.7
30.7
,509,620,244 1,
21.5
24.2
24.4
24.6
24.7
24.9
25.1
25.3
25.4
25.5
25.6
25.8
26.0
26.2
26.4
26.8
26.9
27.4
27.9
28.7
32.2
739,816,901
Reports)
22.2
23.8
24.3
24.7
24.9
25.1
25.2
25.5
25.6
25.7
25.9
26.0
26.2
26.3
26.5
26.6
26.9
27.7
28.4
29.3
30.9
1,839,656,018
227
-------
RFC VOC Performance by Grade
Average VOC Performance of Summer RFC Sold at Retail Stations-By Grade
n regular
a prern
"I" Bars denote range o± RFG Area Averages
1998 1999
19.2
17.0
19.9
19.6
2000
28.8
28.5
27.7
2001 2002
28.8
28.5
27.9
28.7
234
27.6
2003
28.7
235
277
2004
28.2
27.6
2005
28.3
27.8
Figure 3
RFG VOC Emissions Performance (% reduction) by Year and Grade-From Reporting Data
Grade Season
PRM s
REG s
Data
Average
Volume
Average
Volume
Reporting Year
2000
28.0
2,775,514,551
28.8
10,076,956,707
2001
28.0
2,772,816,841
28.6
10,377,246,679
2002
27.7
2,930,007,747
28.4
10,871,002,718
2003
27.7
2,735,922,026
28.5
10,822,132,177
2004
27.9
2,385,535,898
28.4
11,770,292,617
2005
27.9
2,141,224,947
28.3
11,862,678,003
Users of these grade-specific property estimates calculated from RFG and Anti-Dumping data should be
aware that grade-specific estimates based on reporting data may differ from actual retail property values by
grade. Although aggregate estimates from the reporting system may also differ from actual retail property
values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is
often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the
gasoline volume reported and included in these tables as regular or premium would be marketed as mid-grade.
Additionally, EPA's grade-specific average analysis excludes some gasoline and blendstock that was included in
aggregate average estimates. EPA has presented averages for batches labeled as regular and premium gasoline
only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of
gasoline marketed as mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock
batches from grade-specific analyses even if they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In
part, because of factors discussed above these volumes are not expected to represent the actual volumes by
grade of retail gasoline.
228
-------
RFG VOC Performance (Geographic)
Average VOC Performance of Summer RFG Sold at Retail Statlons-By PADD
35.0
30.0 -
25.0 -
20.0 -
15.0 -
10.0 -
5.0 -
D PADD 1
IPADD2
DPADD3
"I" Bars denote range of KFG area averages
1998
18.0
15.2
24.7
1999
18.7
17.0
25.1
2000
28.3
28.6
29.7
2001
28 6
27.7
29.8
2002
284
278
293
28.4
28.0
29.4
2004
28.2
27.5
28.7
2005
28.'
27.8
290
Year
Figure 4
Change in VOC Performance from 2000 to 2005-Summer RFG
Surveys
St. Louis, MO (29.6,31.0)
Rhode Island (27 6,28 4)
Springfield, MA (27.8,28.2)
Atlantic City, NJ (27.8,28.1)
8 Sussex County, DE (27.7,27.9)
o Baltimore, MD (29.3,29.4)
8
Phila.-Wilm, DE-Trenton, NJ (27 9,28.0)
~ Richmond, VA (29. 4,29. 4)
--" Poughkeepsie. NY (27.7,27.7)
S Washington, D.C. -area (296,29.5)
J Covington, KY (28.0,27.7)
| NY-NJ-Long Is.-CT (28.0,27.7)
= Norfolk-Virginia Beach, VA (29.7,29.3)
Louisville, KY (28.5,28.0)
= Boston-Worcester, MA (28.1 ,27.5)
°= Hartford, CT (28.3,27.7)
u Dallas-Fort Worth, TO (29. 4,28. 8)
« Houston-Galveston, TO (29. 9,29.1)
fS
| Warren County, NJ (28.1,27.3)
Manchester, NH (28 6,27.4)
Portsmouth-Dover. NH (28.2,27.0)
Milwaukee-Racine, Wl (28.3,26.7)
Chicago-Lake Co., 1L, Gary, IN (28.4,26.7)
1
"
1
1
I
M
_
|
M
^H
^M
^H
^^
^^
1
^^
c=
^^
1
D
]
^^
^m
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Change in % Reduction fiom Baseline
Figure 5
229
-------
RFC Toxics Performance by Gasoline Volume
I 30-
Toxics Reduction Volume Distribution Trend-Summer RFC
(from Batch Reports)
2001 2002
Reporting Year
Figure 6
Summer RFC Toxics Reduction (%) by Volume (from Batch Reports)
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
\/nh i mp
VUIU 1 1 1C
(qal):
Reporting Year
1998
1.3
18.2
20.9
22.8
24.3
25.6
26.6
27.4
28.2
29.0
29.6
30.4
31.2
31.8
32.6
33.1
33.9
34.7
35.8
37.6
45.6
12,551,351,739
1999
0.3
17.7
20.5
22.1
23.6
24.8
25.8
26.7
27.7
28.7
29.4
30.3
31.1
31.8
32.6
33.4
34.2
34.9
36.0
37.7
45.7
12,605,647,535
2000
20.1
27.4
28.9
30.1
30.9
31.6
32.1
32.6
33.2
33.6
34.1
34.6
35.0
35.4
35.8
36.4
37.0
37.6
38.4
39.6
45.0
12,924,625,994
2001
16.8
24.9
27.4
28.8
29.9
30.7
31.4
32.0
32.5
33.0
33.4
33.9
34.4
34.8
35.2
35.7
36.1
36.8
37.5
38.8
44.2
13,215,304,396
2002
6.0
26.2
28.4
29.5
30.5
31.0
31.6
32.1
32.7
33.1
33.6
34.0
34.4
34.8
35.3
35.9
36.6
37.2
38.2
39.6
43.8
13,838,063,327
2003
19.0
26.1
28.0
29.2
30.1
30.7
31.3
32.0
32.6
33.2
33.8
34.4
34.9
35.4
36.0
36.6
37.1
37.8
38.7
40.0
44.9
13,583,809,485
2004
16.9
26.7
28.3
29.5
30.3
30.9
31.6
32.0
32.5
33.1
33.5
34.1
34.6
35.1
35.7
36.2
36.7
37.3
38.1
39.4
45.5
14,230,001,733
2005
16.2
26.2
27.7
28.7
29.4
30.1
30.7
31.1
31.5
32.0
32.5
32.9
33.4
33.9
34.5
35.1
35.8
36.6
37.4
38.8
45.1
14,070,501,006
230
-------
RFC Toxics Performance by Gasoline Volume (continued):
Toxics Reduction Volume Distribution Trend-Winter RFC
(from Batch Reports)
2000 2001 2002 2003
Reporting Year
Figure 7
Winter RFC Toxics Reduction (%) Volume (from Batch Reports)
Reporting Year
Volume
%tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume
1998
-6.9
14.5
17.0
18.3
19.6
20.8
21.9
22.9
23.9
24.8
25.6
26.2
26.8
27.5
28.3
29.0
29.9
30.8
32.1
34.1
41.1
14,617,590,035
1999
-8.7
13.5
15.9
17.6
19.1
20.6
21.9
22.8
23.7
24.5
25.3
26.0
26.9
27.6
28.3
29.1
29.9
30.7
31.8
33.6
39.6
14,589,007,345
2000
2.8
15.4
17.6
19.1
20.3
21.4
22.4
23.1
24.0
24.9
25.7
26.5
27.2
27.9
28.6
29.3
30.1
31.0
32.0
33.5
39.3
15,829,263,333
2001
-2.3
15.4
17.7
19.4
20.8
22.0
23.0
23.8
24.5
25.3
25.9
26.5
27.0
27.7
28.3
28.9
29.5
30.4
31.4
33.0
39.9
15,703,151,227
2002
0.6
16.4
18.5
20.0
21.1
22.2
22.9
23.7
24.4
25.1
25.9
26.6
27.2
27.8
28.5
29.1
29.7
30.4
31.3
32.7
38.7
16,429,670,147
2003
-3.2
16.4
18.9
20.4
21.4
22.3
23.0
23.8
24.5
25.2
25.9
26.5
27.2
27.8
28.5
29.1
29.9
30.6
31.6
33.2
38.5
16,679,773,135
2004
-5.0
19.0
20.5
21.7
22.7
23.6
24.4
25.1
25.8
26.5
27.1
27.7
28.2
28.8
29.4
30.0
30.7
31.4
32.3
33.7
39.9
17,190,635,125
2005
-1.0
19.2
20.7
21.9
22.9
23.7
24.4
25.1
25.7
26.3
26.9
27.5
28.0
28.6
29.1
29.8
30.5
31.2
32.4
33.5
39.0
18,044,155,549
231
-------
Percentile Chart of Toxics Reduction by Volume- 2005 Summer RFG
(from Batch Reports)
Toxics Reduction (%)
Figure 8
Percentile Chart of Toxics Reduction by Volume- 2005 Winter RFG
(from Batch Reports)
15 20 25
Toxics Reduction ( %|
30
40
45
Figure 9
232
-------
RFC Toxics Performance by Grade
Average Toxics Performance of Summer RFC Sold at Retail Stations-By
Grade
n regular
npreni
"I" Bars denote range of RFG Area Average
1998
232
27.9
27.5
1999
27.8
27.8
27.8
2000
34.1
34.4
349
2001
33.6
34.5
2002
33.1
33.5
34.4
2003
33.1
33.6
2004
32.4
32.9
339
2005
32.1
32.4
334
Figure 10
Average Toxics Performance of Winter RFG Sold at Retail Stations-By Grade
p regular
mid
pprern
"I" Bars denote range of RFG Area Averaues
1998 1999 2000
25.5
24.7
24.1
24.2
24.3
23.4
267
26.4
26.0
2002
25.9
26.4
26.0
Figure 11
2003
25.8
25.4
26.1
2004
26.2
26.3
26.0
2005
27.2
27.5
27.3
233
-------
jfrf (?-,-.' . , . -.
Grade
PRM
REG
Season Data
s Average
Volume
w Average
Volume
s Average
Volume
w Average
Volume
(°/o red
34.4
2,775,514,551
24.7
3,418,369,505
33.9
10,076,956,707
25.5
12,246,078,983
bf Y
33.8
2,772,816,841
24.0
3,288,964,179
32.8
10,377,246,679
25.6
12,289,025,746
34.5
2,930,007,747
25.3
3,387,457,520
33.2
10,871,002,718
25.5
12,966,338,451
anc
34.5
2,735,922,026
25.9
3,103,333,727
33.4
10,822,132,177
25.5
13,486,978,064
i R
34.1
2,385,535,898
26.5
2,943,555,811
33.4
11,770,292,617
26.9
14,158,974,318
33.4
2,141,224,947
26.0
2,581,178,988
32.5
11,862,678,003
26.9
15,422,172,055
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be
aware that grade-specific estimates based on reporting data may differ from actual retail property values by
grade. Although aggregate estimates from the reporting system may also differ from actual retail property
values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is
often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the
gasoline volume reported and included in these tables as regular or premium would be marketed as mid-grade.
Additionally, EPA's grade-specific average analysis excludes some gasoline and blendstock that was included in
aggregate average estimates. EPA has presented averages for batches labeled as regular and premium gasoline
only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of
gasoline marketed as mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock
batches from grade-specific analyses even if they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In
part, because of factors discussed above these volumes are not expected to represent the actual volumes by
grade of retail gasoline.
234
-------
RFC Toxics Performance (Geographic)
AverageToxics Performance of Summer RFC Sold at Retail Stations-By
PADD
40.0
35.0 -
30.0 -
25.0
200
15.0
10.0
5.0
0.0
DPADD1
IPADD2
DPADD3
1998
29.4
22.1
27.5
"I" Bars denote range of RPG area averages
1999
29.5
22.8
269
2000
35.0
31.7
34.7
2001
34.1
30.3
34.5
2002
33.8
309
344
2003
33.7
31.2
34.8
2004
32.6
30.5
35.8
32.3
30.8
33.8
Year
Figure 12
AverageToxics Performance of Winter RFC Sold at Retail Stations-By PADD
35 0
30.0 -
25.0
20.0
15.0 -
50 -
0.0
1998
"I" Bars denote range of RFG area averages
1999
2000
2001
2002
2003
2004
2005
D PADD1
26.2
276
27 6
26 o
26.5
26.3
IPADD 2
20.7
20.6
23.1
23.4
23.;
25.4
DPADD3
24.7
26.4
25.7
25.7
26.6
28.6
Year
Figure 13
235
-------
RFC Toxics Performance Geographic (continued):
Change in Toxics Performance from 2000 to 2005-Summer RFC
Covington, KY 00.6,32.0)
Norfolk-Virginia Beach, VA 03.5,34.2)
Milwaukee-Racine, Wl (29.9,30.3)
Louisville, KY 02.4,32.5)
Houston-Galveston, TX 04.1 ,34,2)
& Baltimore, MD 04.4,34.4)
1 Richmond, VA 04.5,34. 4)
2 St Louis, MO 04.0,32.9)
o
Sussex County, DE (29.7,28.5)
" Washington, D.C.-area 049,337)
'5 Atlantic City, NJ 04.4,33.1)
° Chicago-Lake Co., IL, Gary, IN 01.4,300)
~ Dallas-Fort Worth, TX 05.3.33 5)
jj Phila.-Wilm, DE-Trenton, NJ (34 2,32 4)
| Warren County, NJ 05.4,33.2)
f Rhode Island 05 7,33.3)
CL
NY-NJ-Long Is.-CT 04.9,31.8)
| Springfield, MA 05.7 ,32.1)
Manchester, NH 04.4,30.7)
Portsmouth-Dover, NH 04 1 ,30 5)
Poughkeepsie, NY 04.2,30.2)
Boston-Worcester, MA 06.2,32.0)
Hartford, CT 06.5,30.0)
Surveys
i
c
^
^
+*
^^
^^
^^
^^M
h^
^^
i
^m
^m
^m
^m
^m
^m
^m
^m
c=:
M
]
H
H
H
H
H
3
' -6 -5 -4 -3 -2-10 ;
Change in % Reduction fiein Baseline
Change in Toxics Performance from 2000 to 2005-Winter RFG
Milwaukee-Racine, Wl (22.0,26.2)
Houston-Galveston, TX (25.7,29.6)
Norfolk-Virginia Beach, VA (25.2,29.1)
Richmond, VA (26.7,29.6)
_ Chicago-Lake Co., IL, Gary, IN (22.9,25.5)
§ Covington, KY (27.2,29.6)
1 Dallas-Fort Worth, TX (27.2,29.5)
B Washington, D.C.-area (27.5,29.6)
o
$ Sussex County, DE (24.8,26.3)
I
"J Baltimore. MD (27.6,28.9)
'i Atlantic City, NJ (27.6,28.8)
Boston-Worcester, MA (28.2,29. 4)
" Phila.-Wilm, DE-Trenton, NJ (27.5,285)
£ Manchester, NH (29.2,29 5)
| Rhode Island (28.5,28.5)
1 Springfield, MA (28.5,28.2)
o:
Louisville, KY (28.5,26.9)
« Portsmouth-Dover, NH 01 . 1 ,29.2)
St. Louis, MO (24.5,22.5)
NY-NJ-Long Is.-CT (27.4,25.1)
Warren County, NJ (28.1 ,25.7)
Poughkeepsie, NY (26.3,23.3)
Hartford, CT (26.8,24.5)
.
Surveys
IM
^
^
__
L___ _
:
M
^
^
I
^^M
^
-
^^
C=
^^
1 1
HI
n
i
]
5-4-3-2-10 2345
Change in " Reduction fioin Baseline
Figure 14
Figure 15
236
-------
RFC NOx Performance by Gasoline Volume
NOx Reduction Volume Distribution Trend-Summer RFC
(from Batch Reports)
2000 2001 2002
Reporting Year
Figure 16
Summer RFC NOx Reduction (%) by Volume (from Batch Reports)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
-15.2
-2.8
-1.1
0.1
0.8
1.5
2.2
2.9
3.5
4.1
4.8
5.4
6.1
6.8
7.5
8.3
9.3
10.2
11.4
13.1
20.4
12,551,351,739
1999
-16.1
-3.5
-1.8
-0.5
0.3
1.0
1.6
2.3
3.0
3.8
4.4
5.2
6.0
6.8
7.6
8.3
9.0
10.0
11.1
12.4
19.2
12,605,647,535
2000
-5.5
3.1
4.3
5.1
5.8
6.5
7.0
7.6
8.0
8.4
8.9
9.4
9.9
10.5
11.1
11.7
12.3
13.1
14.0
15.5
19.6
12,924,625,994
2001
-11.1
2.4
3.9
5.0
5.6
6.1
6.6
7.0
7.4
7.8
8.2
8.6
8.9
9.3
9.7
10.3
11.0
11.9
13.1
14.6
19.2
13,215,304,396
2002
-5.9
2.7
4.6
5.5
6.1
6.6
7.0
7.5
7.9
8.2
8.7
9.1
9.6
10.1
10.7
11.3
12.1
12.8
14.0
15.7
21.5
13,838,063,327
2003
-8.1
3.4
4.8
5.7
6.3
6.9
7.3
7.7
8.2
8.6
9.1
9.5
10.0
10.5
11.2
11.8
12.6
13.4
14.4
16.0
21.0
13,583,809,485
2004
-6.2
4.1
5.5
6.3
7.1
7.7
8.3
8.9
9.3
9.8
10.3
10.8
11.4
12.0
12.5
13.2
13.8
14.5
15.4
16.6
21.1
14,230,001,733
2005
-4.4
3.1
4.7
5.8
6.6
7.3
7.9
8.4
9.0
9.5
10.0
10.7
11.3
11.9
12.5
13.1
13.8
14.4
15.3
16.6
21.0
14,070,501,006
237
-------
RFC NOx Performance by Gasoline Volume (continued):
NOx Reduction Volume Distribution Trend-Winter RFG
(from Batch Reports)
20D1 2DD2
Reporting Year
Figure 17
Winter RFG NOx Reduction (%) by Volume (from Batch Reports)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
-12.2
-1.5
0.2
1.2
1.8
2.4
3.1
3.7
4.4
5.1
5.8
6.6
7.3
8.1
9.0
9.9
10.7
11.6
12.6
14.0
19.8
14,617,590,035
1999
-14.6
-3.2
-1.0
0.1
1.0
1.8
2.4
3.2
3.9
4.6
5.4
6.2
6.9
7.7
8.4
9.3
10.1
11.1
12.1
13.4
20.2
14,589,007,345
2000
-13.2
-2.2
-0.3
1.0
1.8
2.4
2.9
3.5
4.0
4.7
5.3
6.1
7.0
7.8
8.8
9.6
10.6
11.6
12.7
14.0
20.0
15,829,263,333
2001
-13.4
-2.1
-0.3
0.6
1.5
2.1
2.8
3.4
4.0
4.8
5.7
6.6
7.5
8.3
9.2
10.0
10.8
11.7
12.7
14.4
20.6
15,703,151,227
2002
-10.4
-1.4
0.2
1.2
2.1
3.0
3.7
4.4
5.1
5.7
6.6
7.3
8.2
8.9
9.6
10.5
11.3
12.2
13.2
14.8
19.3
16,429,670,147
2003
-16.0
-0.8
0.6
1.7
2.7
3.5
4.3
5.2
6.0
6.8
7.5
8.3
9.0
9.7
10.3
11.1
11.8
12.7
13.8
15.2
20.4
16,679,773,135
2004
-8.6
0.1
2.6
4.4
5.7
6.9
7.7
8.4
9.1
9.7
10.3
10.9
11.5
11.9
12.4
13.0
13.6
14.2
15.1
16.8
21.5
17,190,635,125
2005
-11.0
1.3
3.6
5.5
7.0
8.2
9.1
9.8
10.3
10.8
11.3
11.7
12.2
12.5
12.9
13.4
13.8
14.3
15.0
16.5
21.1
18,044,155,549
238
-------
Percentile Chart of NOx Reduction by Volume- 2005 Summer RFC
(from Batch Reports)
5 10
NOx Reduction (%)
Figure 18
Percentile Chart of NOx Reduction by Volume- 2005 Winter RFC
(from Batch Reports)
0 5 10
NOx Reduction |«.)
Figure 19
239
-------
RFC NOx Performance by Grade
Average NOx Performance of Summer RFC Sold at Retail Stations-By Grade
Figure 20
Average NOx Performance of Winter RFC Sold at Retail Stations-By Grade
Figure 21
240
-------
NC
rat.
(% by
Grade Season
PRM s
w
REG s
w
Data
Average
Volume
Average
Volume
Average
Volume
Average
Volume
11.8
2,775,514,551
8.9
3,418,369,505
8.4
10,076,956,707
5.1
12,246,078,983
11.0
2,772,816,841
9.2
3,288,964,179
7.7
10,377,246,679
5.3
12,289,025,746
12.0
2,930,007,747
10.4
3,387,457,520
8.2
10,871,002,718
5.8
12,966,338,451
12.5
2,735,922,026
11.0
3,103,333,727
8.7
10,822,132,177
6.6
13,486,978,064
12.8
2,385,535,898
12.4
2,943,555,81 1
9.9
11,770,292,617
9.1
14,158,974,318
13.0
2,141,224,947
12.6
2,581,178,988
9.6
1 1 ,862,678,003
10.1
15,422,172,055
Users of these grade-specific property estimates calculated from RFC and Anti-Dumping data should be
aware that grade-specific estimates based on reporting data may differ from actual retail property values by
grade. Although aggregate estimates from the reporting system may also differ from actual retail property
values, there are several additional reasons why these grade-specific estimates may differ. Mid-grade gasoline is
often blended from regular and premium gasoline at some point downstream of the refinery. Thus, some of the
gasoline volume reported and included in these tables as regular or premium would be marketed as mid-grade.
Additionally, EPA's grade-specific average analysis excludes some gasoline and blendstock that was included in
aggregate average estimates. EPA has presented averages for batches labeled as regular and premium gasoline
only; excluding batches labeled as mid-grade (since these batches may not be a representative sample of
gasoline marketed as mid-grade), mix of grades, or without a grade label. EPA also excluded CG blendstock
batches from grade-specific analyses even if they had a grade designation.
The table shows the total volume, in gallons, for the batches used to calculate each grade average. In
part, because of factors discussed above these volumes are not expected to represent the actual volumes by
grade of retail gasoline.
241
-------
RFC NOx Performance (Geographic)
Average NOx Performance of Summer RFC Sold at Retail Stations-By PADD
'I Bars denote range of FJG area avera.::??
Year
Figure 22
Average NOx Performance of Winter RFG Sold at Retail Stations-By PADD
20 ij
Year
Figure 23
242
-------
RFC NOx Performance-Geographic (continued):
Change in NOx Performar
Covington.KY (10.9,17.7)
Louisville, KY (12.0,17.3)
St. Louis, MO (8.5,12.5)
Atlantic City, NJ (8.6,11.9)
Richmond, VA (7. 9, 10. 3)
o Sussex County, DE (7.5,9.6)
1 Houston-Galveston,TX(8.3,10.3)
5
2 Poughkeepsie, NY (9.1,11.1)
o
jg Springfield, MA (9.5. 11. 2)
"J Norfolk-Virginia Beach, VA (8.4,10.0)
Baltimore, MD |B. 4, 9. 8)
Hartford, CT(9.7,11.D)
NY-NJ-Long Is.-CT (9.5,10.7)
J Washington, D.C. -area (B. 5,9. 4)
| Rhode Island (10.0,10.8)
f Warren County, NJ (9.7,10.2)
*' Phila.-Wilm, DE-Trenton, NJ (9.1 ,9.5)
OS
S Chicago-Lake Co., IL, Gary, IN (13. 2, 13. E)
Milwaukee-Racine, Wl (135,136)
Dallas-Fort Worth, TX (8.6,8.8)
Boston-Worcester, MA (10 5, 10. 7)
Portsmouth-Dover, NH (11.7,9.8)
Manchester, NH (12.0,9.8)
ice from 2000 to 2005-Summer RFG
Surveys
mm
IH
3 -2 -
^m
mi
I
I
: : _l
m^^m
^HH
1
: 1
^mm
mt
zzz
zz:
zz:
mm
D
^
I
1
3
z:
n
n
j
0 1 2 3 4 5 B 7 t
Change in % Reduction from Baseline
Figure 24
Change in NOx Performance from 2000 to 2005-Winter RFC Surveys
Covington, KY (11.2,17.6) |
Milwaukee-Racine, Wl (78,140) I
Norfolk-Virginia Beach, VA (5.3,11.3)
SI. Louis, MO (6.6,12.2)
Washington, D.C-area (6.4,11.9)
Richmond, VA (5.6,11.0)
Rhode Island (7.1,12.5)
Chicago-Lake Co., IL, Gary, IN (8.3,13.3)
Springfield, MA (7.5,12.0)
Houston-Galveston, "C< (6.6,11.0)
Poughkeepsie, NY (7.1,10.7)
Dallas-Fort Worth, TX (6.7,10.2)
Atlantic City, NJ (8.4,11.8) )
Sussex County, DE (5.0,8.4)
NY-NJ-Long la.-CT (7.5,10.8)
Baltimore, MD (6.6,10.1)
Boston-Worcester, MA (9.0,12.2)
Phila.-Wilm, DE-Trenton, NJ (8.5,11.6)
Portsmouth-Dover, NH (10.6,13.4)
Hartford, CT(B.2,11.0)
Louisville, KY (11 4.13.1)
Manchester, NH (11.2,12.0)
Wamn County. NJ (8.8,10.1)
0
I
I
^^^^^^^^"
I
I
I
-i
1 2 3 4 5 B
'." ii,-i!».jT- in % Reduction from Boseline
Figure 25
243
-------
Conventional Gasoline
CG Exhaust Toxics Emissions by Gasoline Volume
Complex Model Exhaust Toxics Volume Distribution Trend-Summer CG
(from Batch Reports-Excluding Blendstocks)
2001 2002
Reporting Year
Figure 26
Summer CG Exhaust Toxics Emissions (mg/mi) by Volume (Calculated from Batch Data)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
49.9
58.9
60.9
62.8
64.2
65.5
66.5
67.5
68.5
69.5
70.6
71.7
72.9
74.0
75.4
77.2
79.5
82.1
86.1
93.2
155.2
38,992,885,793
1999
48.6
57.9
60.7
62.4
63.9
65.1
66.4
67.5
68.8
70.1
71.3
72.7
74.0
75.4
76.7
78.2
80.1
82.7
86.7
95.1
252.3
37,127,053,812
2000
49.6
59.3
61.5
62.9
64.3
65.5
66.5
67.6
68.8
70.1
71.5
73.0
74.5
76.1
77.8
79.5
81.4
83.7
87.2
93.7
255.2
36,355,722,872
2001
47.0
59.1
61.5
63.3
64.6
65.8
66.9
67.9
69.1
70.4
71.7
72.9
74.4
76.0
77.8
79.7
81.9
84.2
87.5
93.1
258.0
38,541,026,687
2002
47.0
59.0
61.2
62.8
64.0
65.0
66.0
67.0
68.1
69.2
70.3
71.7
73.1
74.7
76.4
78.5
80.6
82.9
86.3
92.2
245.8
40,811,369,914
2003
47.0
58.6
60.7
62.3
63.5
64.7
65.8
66.9
68.1
69.6
71.1
72.6
74.4
76.0
77.6
79.3
81.1
83.3
87.0
93.7
159.3
43,215,951,157
2004
46.4
57.9
59.8
61.1
62.0
62.8
63.5
64.3
65.0
65.8
66.6
67.6
68.6
69.7
71.1
72.8
74.7
76.8
79.7
85.0
154.0
43,495,028,871
2005
48.4
58.2
59.8
61.0
61.9
62.7
63.5
64.3
65.0
65.8
66.7
67.8
69.0
70.2
71.5
73.0
75.0
77.4
80.9
86.8
175.8
42,171,355,985
244
-------
CG Exhaust Toxics by Gasoline Volume (continued):
Complex Model Exhaust Toxics Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2001 2002
Reporting Year
Figure 27
Winter CG Exhaust Toxics Emissions (mg/mi) by Volume (Calculated from Batch Data)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
76.4
89.7
93.1
95.6
97.6
99.4
101.0
102.7
104.3
105.7
107.5
109.3
110.9
112.8
114.8
117.1
119.6
123.5
128.7
140.1
408.8
46,167,047,173
1999
73.8
90.9
94.0
96.4
98.4
100.1
101.8
103.2
105.0
106.7
108.5
110.3
111.9
113.7
115.9
118.0
120.8
124.7
129.9
139.9
395.2
47,555,086,633
2000
76.2
89.8
93.3
95.8
97.6
99.4
101.2
102.6
104.1
105.6
107.2
108.7
110.5
112.5
114.7
117.6
121.0
125.4
131.1
140.4
419.4
47,925,321,684
2001
75.5
91.2
94.2
96.5
99.0
100.8
102.4
104.0
105.6
107.2
108.9
110.5
112.1
114.0
116.5
119.4
122.6
127.0
134.7
146.7
414.9
48,770,268,393
2002
75.4
90.4
93.5
95.8
97.8
99.4
101.0
102.6
104.3
105.9
107.5
109.2
111.1
113.1
115.4
117.8
121.1
124.7
130.9
141.9
239.8
49,640,309,828
2003
74.4
90.3
93.1
94.9
96.4
98.1
99.5
101.2
102.9
104.7
106.6
108.4
110.4
112.5
114.7
117.2
120.4
124.0
129.3
139.4
399.7
47,426,711,234
2004
72.9
88.6
91.0
92.7
93.9
94.9
96.0
97.2
98.4
99.7
101.0
102.7
104.4
106.3
108.3
110.2
113.1
116.1
120.9
129.5
276.7
47,615,735,564
2005
73.9
88.5
91.3
92.8
94.0
95.1
96.3
97.4
98.5
99.7
100.8
102.4
104.1
105.8
108.0
110.1
112.7
116.5
121.9
132.0
260.6
48,664,360,993
245
-------
Percentile Chart of Complex Model Exhaust Toxics Emissions By Volume
2005 Summer CG
(Calculated from Batch Data-Excluding Blendstocks)
100 120 140
Exhaust toxics (mg/mi)
Figure 28
Percentile Chart of Complex Model Exhaust Toxics Emissions By Volume
2005 Winter CG
(Calculated from Batch Data-Excluding Blendstocks)
140 190
Exhaust toxics (mg/mi)
Figure 29
246
-------
CG NOx Emissions by Gasoline Volume
Complex Model NOx Emission Volume Distribution Trend-Summer CG
{from Batch Reports-Excluding Blendstocks)
2001 2002
Reporting Year
Figure 30
Summer CG NOX Emissions (mg/mi) by Volume (Calculated from Batch Data)
Reporting Year
Volume %tile
minimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (gal):
1998
1069.8
1163.8
1182.3
1201.4
1224.7
1244.0
1262.8
1280.4
1298.3
1311.3
1323.5
1335.4
1349.3
1365.6
1382.0
1399.4
1418.1
1440.3
1463.3
1494.0
1790.5
38,992,885,793
1999
1053.6
1159.2
1178.2
1199.5
1220.2
1243.2
1265.4
1288.0
1306.5
1320.6
1335.2
1351.7
1364.7
1377.0
1391.7
1406.6
1425.2
1446.8
1470.9
1508.3
1799.1
37,127,053,812
2000
1071.4
1168.0
1191.5
1206.9
1225.9
1248.0
1271.2
1291.9
1308.9
1320.9
1334.2
1345.5
1358.5
1371.6
1389.3
1405.9
1427.8
1449.5
1472.6
1507.4
1765.5
36,355,722,872
2001
1072.8
1163.5
1184.0
1202.0
1221.2
1242.2
1263.5
1283.4
1302.4
1319.5
1334.6
1348.1
1362.3
1379.6
1394.0
1411.5
1432.2
1453.7
1477.3
1511.2
1813.7
38,541,026,687
2002
1064.4
1160.5
1179.9
1196.7
1214.9
1232.3
1254.1
1275.1
1295.1
1313.5
1329.1
1344.3
1357.2
1373.1
1392.6
1411.0
1430.6
1450.5
1472.9
1503.3
1769.8
40,811,369,914
2003
1072.6
1157.1
1172.4
1186.8
1201.1
1221.7
1247.7
1273.0
1295.3
1314.9
1330.1
1343.6
1357.4
1373.1
1388.9
1408.5
1428.3
1450.6
1477.0
1515.9
1782.6
43,215,951,157
2004
1058.0
1154.1
1163.6
1171.7
1180.3
1187.6
1194.8
1204.5
1217.6
1231.3
1245.8
1259.9
1270.4
1283.3
1293.4
1304.5
1316.0
1330.3
1349.6
1384.7
1629.2
43,495,028,871
2005
1052.1
1154.9
1166.4
1175.7
1182.4
1187.9
1195.8
1204.0
1213.6
1223.9
1234.0
1246.2
1259.1
1272.6
1287.2
1301.4
1314.8
1331.0
1353.4
1395.6
1662.1
42,171,355,985
247
-------
CG NOx Emissions by Gasoline Volume (continued):
Complex Model NOx Emission Volume Distribution Trend-Winter CG
(from Batch Reports-Excluding Blendstocks)
2001 2002
Reporting Year
Figure 31
Winter CG NOX Emissions (mg/mi) by Volume (Calculated from Batch Data)
Reporting Year
/olume %tile
ninimum
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Volume (qal):
1998
1207.9
1318.6
1335.1
1352.9
1372.9
1393.0
1413.8
1434.3
1454.4
1473.4
1489.4
1506.0
1524.0
1540.8
1558.3
1576.2
1596.7
1618.6
1645.9
1686.6
2033.8
46,167,047,173
1999
1214.1
1321.9
1339.9
1361.7
1382.2
1403.5
1426.1
1446.9
1464.0
1481.8
1497.5
1514.0
1531.8
1549.8
1568.4
1588.4
1611.5
1633.6
1660.2
1696.9
2048.4
47,555,086,633
2000
1210.4
1317.5
1337.8
1358.7
1378.0
1400.9
1420.7
1444.2
1464.0
1480.9
1496.4
1512.5
1530.2
1551.1
1567.4
1586.2
1607.7
1632.4
1661.5
1709.0
2019.9
47,925,321,684
2001
1202.6
1320.1
1339.0
1358.6
1379.8
1403.7
1428.5
1451.4
1471.9
1487.8
1507.4
1524.4
1542.5
1561.3
1579.7
1599.5
1619.6
1644.1
1672.2
1719.0
2010.7
48,770,268,393
2002
1217.4
1314.0
1334.7
1355.1
1374.4
1397.8
1421.3
1444.7
1462.3
1479.3
1496.1
1511.9
1527.7
1544.0
1564.7
1589.0
1613.9
1637.3
1666.4
1710.6
2005.6
49,640,309,828
2003
1201.5
1313.1
1328.0
1341.8
1356.9
1376.0
1397.4
1421.6
1441.0
1456.8
1472.9
1489.8
1507.8
1524.4
1544.3
1567.0
1592.8
1624.1
1659.0
1708.8
2017.4
47,426,711,234
2004
1204.1
1304.8
1320.6
1329.4
1337.7
1346.9
1356.5
1369.8
1383.6
1397.0
1411.4
1425.5
1440.1
1454.4
1466.2
1480.8
1495.0
1516.3
1543.6
1599.2
1886.7
47,615,735,564
2003
1205.4
1305.4
1319.9
1327.8
1335.1
1342.3
1349.5
1357.2
1365.7
1376.0
1386.2
1398.4
1412.1
1428.7
1444.8
1463.6
1481.2
1503.7
1531.9
1584.2
1897.0
48,664,360,993
248
-------
1000
Percentile Chart of Complex Model NOx Emissions By Volume
2005 Summer CG
(from Batch Reports-Excluding Blendstocks)
1100
1300 1400
NOx (mg/mi)
1500
1600
1700
Figure 32
Percentile Chart of Complex Model NOx Emissions By Volume
2005 Winter CG
(from Batch Reports-Excluding Blendstocks)
1400 1500 1600
NOx (rng/mi)
Figure 33
249
-------
This report has largely focused on how gasoline has changed in response to EPA's regulatory
requirements. In order to supplement more subjective evidence that gasoline composition has or has not
changed concurrent with regulatory changes, EPA has made limited use of regression analysis.
RFC composition was expected to change when the Phase II standards were implemented. Many
of the descriptive analyses presented in this report show property changes between 1999 and 2000.
However it is not always clear that these changes are distinguishable from year-to-year fluctuations and
any overall linear time trend that may be present. EPA fit each set of average estimates to the model:
Y=b0+b1*YEAR+b2*Phs2
where Y is EPA's estimate of the property average. YEAR is the reporting or survey year and Phs2 is a
categorical variable set to 0 for 1998 and 1999 and 1 for 2000 through 2004. (Although EPA's data for
some properties included years prior to 1998, and the report examines trends through 2005, these
regression analyses were limited to years 1998 through 2004.) 42A statistically significant Phs2 term
would provide objective evidence that a sustained shift in a property value occurred concurrent with the
transition from Phase I to Phase II RFC standards. Table 1 shows an estimate of the Phs2 coefficient and
its p-value for each data set. The value of the Phs2 coefficient is an estimate of the magnitude of the
parameter shift and a p-value of 0.050 or less indicates a Phs2 coefficient that can be considered
statistically significant.
42 These regression analyses were done before the 2005 averages were calculated. They were not updated
since the primary intent of this analysis was to provide a more objective basis for evaluating changes between 1999
and 2000.
250
-------
.'"
'. .-...->! .' .!
Summer Benzene
Aromatics
RVP
E200
E300
T50
T90
Sulfur
Oxygen
Olefins
NOx reduction
Toxics reduction
VOC reduction
Winter Benzene
Aromatics
E200
E300
T50
T90
Sulfur
Oxygen
Olefins
NOx reduction
Toxics reduction
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
: n
nn ' -'-.
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
'.-"r'-m
04
0.006
-0.112
0.002
-6.109
0.000
-0.913
0.003
-2.124
0.004
2.451
0.003
4.892
0.004
-9.039
0.030
-51.320
0.340
-0.125
0.028
-1.436
0.002
4.717
0.000
6.617
0.000
9.674
0.072
-0.037
0.001
-3.295
0.245
-0.275
0.002
1.928
0.099
2.826
0.003
-9.838
0.365
29.467
0.206
-0.242
0.512
0.315
0.812
0.330
0.026
2.087
04
0.019
-0.092
0.008
-2.904
0.000
-0.859
0.003
-1.710
0.006
2.311
0.024
-53.790
0.700
-0.050
0.652
-0.320
0.012
3.267
0.001
4.811
0.000
9.168
0.288
0.008
0.132
-0.611
0.872
0.044
0.000
1.726
0.367
29.143
0.191
-0.192
0.104
1.290
0.228
-1.515
0.867
-0.101
Table 1
251
-------
For Summer RFC, both survey (retail) and reporting (production)-based data analyses indicated
that statistically significant shifts occurred in all Complex Model input properties except oxygen and
olefins. Neither analysis indicated a statistically significant shift in oxygen content and only the
survey-based analysis indicated a shift in olefins content. The survey analysis indicated statistically
significant shift in T50 and T90. (These parameters are not reported for all production data, so these
trends were analyzed for surveys only.) Both analyses found that emissions performance for VOCs, NOx
and toxics changed significantly concurrent with the transition.
For Winter RFC, the survey data analysis detected statistically significant shifts in aromatics,
E300, T90 and toxics performance. Reporting data analysis detected a shift in E300 only.
In general, these regression results are consistent with subjective interpretation of the data and
consideration of the underlying cause-effect relationships. This analysis, in some cases, may have failed
to detect shifts that occurred between 1999 and 2000 because important variables or terms were omitted
from the regression model. For example, the Winter NOx emission performance graph based on survey
data gives a distinct impression of a shift (see the RFC Trends and Emissions Chapters). However the
regression analysis did not conclude significance quite possibly because an even larger change occurred
between 2003 and 2004. This latter change occurred as a result of Tier 2 sulfur reductions, and there is
no term in the regression model to account for anything but a linear change over time.
Table 2 shows an estimate of the YEAR coefficient and its p-value for each data set. A
statistically significant (P < 0.050) YEAR coefficient indicates that an upward (positive coefficient) or
downward (negative) linear trend was detected between 1998 and 2004 after correcting for any upward
or downward Phase I to Phase II shift.
252
-------
".:.'.'
Summer Benzene
Aromatics
RVP
E200
E300
T50
T90
Sulfur
Oxygen
Olefins
NOx reduction
Toxics reduction
VOC reduction
Winter Benzene
Aromatics
E200
E300
T50
T90
Sulfur
Oxygen
Olefins
NOx reduction
Toxics reduction
1
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
YEAR
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
P-value
Coefficients
199> ' >"" ;
irt-f j«
0.051
0.013
0.232
0.257
0.047
0.023
0.280
0.091
0.022
-0.340
0.480
-0.128
0.013
1.433
0.043
-10.287
0.044
0.076
0.015
0.389
0.434
0.126
0.017
-0.306
0.559
-0.060
0.027
0.012
0.015
0.301
0.286
0.056
0.007
-0.318
0.038
-0.910
0.009
1.611
0.036
-20.333
0.152
0.064
0.230
0.140
0.192
0.460
0.472
-0.109
RFC
0.894
0.001
0.421
0.121
0.008
0.020
0.162
0.100
0.085
-0.223
0.038
-10.430
0.068
0.068
0.593
0.086
0.116
0.337
0.731
-0.040
0.379
-0.065
0.074
-0.003
0.686
0.032
0.989
-0.001
0.000
-0.288
0.035
-20.325
0.151
0.049
0.157
-0.242
0.027
0.816
0.071
0.311
Table 2
253
-------
For Summer RFC, the analysis detected a statistically significant upward linear trend in RVP and a
downward linear trend in sulfur in both data sets. The analysis found statistically significant downward
E300 and sulfur trends in both Winter RFC data sets. For several other properties only one of the two
data sets indicated a statistically significant trend. Again, this analysis may have failed to detect
systematic increases or decreases between 1998 and 2004 because the regression model chosen does
not adequately describe the trend. For instance, although only the reporting data indicated statistical
significance, there is little doubt (because of required sulfur reductions) that the NOx performance of both
retail and production Winter RFC improved between 1998 and 2004.
Since substantial changes in some RFC properties occurred with Phase II, it is possible that the
transition to Phase II RFC standards also affected CG composition. (EPA's Anti-Dumping regulations were
designed to limit RFG's adverse impact on CG's emission-related qualities.) EPA applied the same
regression model to its reporting data-based CG average estimates in order to detect possible property
value shifts between 1999 and 2000.
Table 3 shows estimates of the Phs2 coefficients and p-values. For Summer CG, aromatics and
RVP indicated statistically significant values. The shift in aromatics was positive, opposite in direction to
the statistically significant negative change in RFC aromatics. While this may indicate that some
aromatics were shifted from RFC to CG production to achieve RFC aromatics reductions, this is uncertain
(see the Aromatics Chapter.) The RVP shifts for CG and RFC were in the same direction. For Winter
CG, only olefins showed a statistically significant value, and the RFC regression analysis did not indicate a
change in Winter olefin content.
Summer Aromatics
Benzene
E200
E300
Olefins
RVP
Sulfur
Winter Aromatics
Benzene
E200
E300
Olefins
RVP
Sulfur
Term
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
Phs2
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
P- value
Coefficient
CG
0.004
0.960
0.955
0.002
0.853
0.043
0.447
-0.274
0.234
0.903
0.027
-0.063
0.309
89.591
0.433
0.299
0.572
0.020
0.556
-0.233
0.388
0.301
0.038
1.281
0.120
-0.208
0.308
75.818
Table 3
254
-------
Table 4 shows estimates of the YEAR coefficients and p-values. None of these are statistically
significant. As previously stated, this regression analysis may not have detected systematic changes
between 1998 and 2004 because the model does not adequately describe the trend. For example, there
is no question that Summer and Winter CG sulfur decreased between 1998 and 2004, but the decreases
were very non-linear.
Summer Aromatics
Benzene
E200
E300
Olefins
RVP
Sulfur
Winter Aromatics
Benzene
E200
E300
Olefins
RVP
Sulfur
Term
Year
Year
Year
Year
Year
Year
Year
Year
Year
Year
Year
Year
Year
Year
1
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
P-value
Coefficient
CG
0.059
-0.099
0.874
-0.001
0.344
0.053
0.959
0.004
0.283
-0.180
0.086
0.009
0.102
-36.705
0.396
-0.074
0.750
-0.003
0.170
0.137
0.431
-0.062
0.073
-0.229
0.128
0.046
0.079
-34.410
Table 4
255
-------
United States gasoline production and supply data are often aggregated into geographic regions
referred to as Petroleum Administration for Defense Districts (PADDs). The US is divided into five PADDs
(PADD I-East Coast, PADD II-Midwest, PADD Ill-Gulf Coast, PADD IV-Rocky Mountain, PADD V-West
Coast).
In the various individual parameter appendices, EPA presented estimates of RFC property averages in
each year and season, aggregated at the PADD level. These estimates were based on the retail data
collected in RFC Surveys. EPA did not have comparable geographic data for CG, and did not include any
"geographic" CG trend analysis in those appendices. This appendix includes a limited PADD-specific
analysis of RFC and CG batch reporting data. It does not examine trends, but separately summarizes
2004 and 2005 reporting data.43These PADD averages also appear in the individual parameter chapters.
Batch records do not explicitly identify a PADD, but each batch record contains a "facility"
identifier, so that a refiner's batch data can be related to a refinery, and the refinery related to a PADD,
based on its location. This PADD-specific analysis excluded those batches where the "facility" was an
importer. Importer batches were included in the aggregate trend analyses presented in the body of the
report.
In order to provide the additional geographic information contained in this PADD-level analysis,
while still maintaining a sufficient degree of aggregation to prevent gasoline properties or volumes from
being strongly associated with individual companies or facilities, EPA has reported gasoline property
averages for PADDs I, II and III only. (The parties that submit these data to EPA have claimed that the
data are confidential business information (CBI), and these data may be protected by CBI regulations.)
Users of these PADD-specific averages should be aware that while these estimates are likely to
provide better information about the properties and emissions qualities of gasoline sold or used within a
PADD than national averages, the properties of gasoline produced in a PADD are likely to differ somewhat
from the properties of gasoline consumed in a PADD. Significant PADD to PADD movement of gasoline
(e.g. via the Colonial Pipeline which transports gasoline from PADD III to PADD I) as well as importation
of finished gasoline account for this difference. PADD-specific retail property averages estimated from
RFC Survey data are included in the parameter chapters.
Tables 1 and 4 show the volumes, in gallons, for the batches allocated to the various PADDs in
each of the two years. The gasoline volume in the column labeled "Other" includes total volume from
batches associated with PADD IV or V refineries and refineries in US territories. The column labeled
"Grand total" is the sum of the volumes in the PADD I through III and "Other" columns. The column
labeled "Aggregate Total" represents the volume of the batches considered in the aggregate analyses
reported in the body of this report and in the individual parameter appendices. The "Aggregate Totals"
are higher because the aggregate analyses included importer as well as refiner batches. The volume
totals in Tables 1 and 4 are totals prior to any data screening done in conjunction with computation of the
averages presented in this report. (The PADD-average tables in the parameter chapters give the post-
screening volumes for each parameter average calculation). PADD I, II and III volume totals are
expected to be approximately the total gasoline production in each of those PADDs, excluding any
gasoline volume that is exported.
43 Resources did not permit PADD-specific analysis of each year's batch data. EPA felt that analysis of more
recent data may be of greater use to a wider audience than analysis of older data.
256
-------
EPA reporting data "Grand Totals" by PADD in Table 1 compare reasonably well to estimates of
finished gasoline production volume minus exports based on data published in the Energy Information
Administration's "Petroleum Supply Annual 2004" (EIA,2005).44 Using the EIA data, EPA calculated
that these volumes were, for PADDs I, II and III, respectively, 18,926,166,000 gallons, 31,976,994,000
gallons and 54,012,336,000 gallons. The purpose of these volume comparisons was not to question
EIA's volume data, but to provide some independent verification of the completeness of EPA's data, and
to detect any gross inconsistencies, possibly indicating a problem with EPA's data or analysis. Although
there were some differences (see footnote), EPA does not believe that they are "gross inconsistencies",
and has not investigated them. EPA did not do a similar comparison of 2005 volumes. The volumes in the
"Grand total" column do not represent total PADD I through V gasoline production because they exclude
PADD V "California" gasoline, and include some volume from refineries outside of the 50 states.
201
type
CG Summer
Winter
CG
Total
RFG Summer
Winter
RFG Total
Grand Total
)4
PADD
I
3,752,272,894
3,484,116,076
7,236,388,970
4,792,114,891
6,489,530,475
11,281,645,366
18,518,034,336
i
II
11,456,437,187
13,323,944,095
24,780,381,282
1,740,499,436
2,438,403,544
4,178,902,980
28,959,284,262
bf
III
21,722,608,623
23,628,091,025
45,350,699,648
5,890,920,167
6,059,647,031
11,950,567,198
57,301,266,846
6,400,078,025
6,809,448,022
13,209,526,047
624,073,296
461,826,582
1,085,899,878
14,295,425,925
43,331,396,729
47,245,599,218
90,576,995,947
13,047,607,790
15,449,407,632
28,497,015,422
119,074,011,369
44,009,126,002
48,333,969,089
92,343,095,091
14,243,059,617
17,194,370,899
31,437,430,516
123,780,525,607
Table 1
Tables 2, 3, 5, and 6 show the volume-weighted parameter averages calculated for PADDs I, II,
and III RFG and CG. In both cases, the data were screened as for the aggregate average computations;
i.e. for estimation of most seasonal average parameter values, data were screened on a parameter-
specific basis primarily to exclude missing data. Additionally, for Complex Model emission average
calculations, batches with "outlier" property values and CG blendstock batches were excluded. Although
EPA's aggregate CG emissions rate averages were based on EPA's recalculated Complex Model emissions
values for each batch, these PADD-level analyses used reported emission values. Further information
relating to data screening, as well as information relating to each parameter can be found in the body of
this report. The negative value for the volume-weighted average TAME content that appears in the CG
average table warrants further explanation. This arises because certain batches, for compliance
calculations, are reported with negative volumes. When a refiner uses previously certified gasoline (PCG)
as a blending component to make other gasoline the PCG batch properties are reported with a negative
volume in order to avoid double-counting this gasoline in calculations. When a batch of certified gasoline
is exported, it is also reported with a negative volume. Consequently, although a negative property
average may be physically impossible, it could occur in volume-weighted average calculations
44EPA used supply and disposition data for Finished Motor Gasoline for PADD I (Table 4 page 34), PADD II
(Table 6, page 36) and PADD III (Table 8, page 38) summing field production and refinery production and
subtracting exports. Volumes were reported in thousand barrels. These tables also subdivided Finished Gasoline into
"Reformulated", "Oxygenated" and "Other" categories. EPA compared the EIA "Reformulated" volumes with its RFG
volumes and the sum of the "Oxygenated" and "Other" volumes with its CG volumes. EIA's PADD II "Reformulated"
volume was about 32% higher than EPA's and its PADD III "Reformulated" volume about 13% lower. The remaining
differences ranged from about 0.4% to under 7 percent. The sum of the compared gasoline category volumes over
the three PADDs as well as the sum of gasoline volume over these PADDs each differed by less than 1 percent.
257
-------
by
Summer
Winter
PADD
I
II
III
I
II
III
Aromatics
21.2
18.9
18.7
20.4
17.3
17.9
Benzene
(v%)
0.55
0.84
0.54
0.65
0.79
0.54
Olefins
(v%)
12.97
5.29
11.24
12.92
4.86
11.40
Oxygen E200 E300 RVP Sulfur
(wt%) (%) (%) (psi) (ppm)
2.59 47.3 83.6 6.83 91
3.48 46.6 84.7 6.98 78
2.36 48.7 82.5 6.88 73
2.45 55.4 85.1 118
3.26 58.4 85.7 93
2.09 55.8 83.7 101
Ethanol
3.01
9.39
1.75
3.00
8.61
0.98
MTBE
8.02
0.00
8.74
7.11
0.00
8.52
ETBE
0.00
0.00
0.00
0.00
0.00
0.00
TAME
0.14
0.00
0.55
0.05
0.00
0.68
T_butanol
0.01
0.00
0.05
0.01
0.00
0.04
Performance
NOx
8.8
13.3
11.3
7.6
12.9
10.3
Toxics
(°/o)
32.8
29.9
35.6
25.0
25.2
29.0
voc
28.0
26.6
29.1
Table 2
by
"' vlex
r,!.- .(mg/mi)
Summer
Winter
PADD
I
II
III
I
II
III
Aromatics
(v%)
28.4
28.7
27.3
24.0
25.1
24.1
Benzene
(v%)
0.96
1.38
0.98
0.93
1.32
0.89
Olefin
s (v%)
12.8
8.8
12.2
15.7
8.6
12.2
Oxygen
(wt%)
0.44
0.34
0.17
0.37
0.40
0.14
E200
(°/o)
45.5
46.6
44.6
51.5
52.6
49.7
E300
(°/o)
82.7
81.4
79.4
84.4
84.1
81.8
RVP
(psi)
8.40
8.46
8.19
12.66
13.31
11.77
Sulfur
(ppm)
125
140
108
138
135
115
Ethanol
(v%)
0.82
0.97
0.06
0.83
1.16
0.05
MTBE
(v%)
0.70
0.00
0.76
0.35
0.00
0.65
ETBE
(v%)
0.09
0.00
0.00
0.01
0.00
0.00
TAME
(v%)
-0.17
0.00
0.04
0.03
0.00
0.02
T butanol
(v%)
0.00
0.00
0.00
0.00
0.00
0.00
Exhaust
toxics
67.7
71.2
67.1
103.7
107.6
101.1
NOx
1275.8
1249.9
1254.5
1477.5
1413.5
1424.9
Table 3
258
-------
by
PADD
CG
CG Total
RFC
RFC Total
Grand Total
Summer
Winter
Summer
Winter
I
3,547,001,722
4,064,064,814
7,611,066,536
4,431,080,230
7,081,940,665
11,513,020,895
19,124,087,431
II
10,568,941,899
13,589,333,929
24,158,275,828
1,844,913,833
2,718,751,541
4,563,665,374
28,721,941,202
III
20,889,546,279
22,444,880,341
43,334,426,620
5,690,766,967
5,766,524,033
11,457,291,000
54,791,717,620
6,061,232,118
6,725,201,537
12,786,433,655
546,882,756
388,731,882
935,614,638
13,722,048,293
41,066,722,018
46,823,480,621
87,890,202,639
12,513,643,786
15,955,948,121
28,469,591,907
116,359,794,546
42,849,893,176
49,467,392,607
92,317,285,783
14,092,489,036
18,046,671,196
32,139,160,232
124,456,446,015
Table 4
by
Summer
Winter
PADD
I
II
III
I
II
III
r
Aromatics
(V<>/0)
21.9
19.1
19.2
20.7
18.4
18.0
Benzene
(V<>/0)
0.62
0.86
0.60
0.69
0.81
0.57
Olefins
(V<>/0)
13.3
4.7
12.5
13.0
5.3
11.0
Oxygen
(wto/o)
2.50
3.50
2.25
2.41
3.24
2.05
E200
(o/o)
47.6
46.3
49.9
55.6
56.5
56.7
E300
(o/o)
84.1
84.3
83.2
85.7
84.7
84.5
RVP
(psi)
6.88
6.95
6.93
Sulfur
(ppm)
73
64
78
90
78
83
Ethanol
(V<>/0)
3.14
9.46
1.31
3.02
8.63
0.91
MTBE
(V<>/0)
7.13
0.00
8.96
6.82
0.00
8.53
ETBE
(V<>/0)
0.01
0.00
0.01
0.04
0.00
0.02
TAME
(V<>/0)
0.28
0.00
0.59
0.09
0.00
0.53
#**'*« 8itrt»*li«« 8 # -"" "^ % ^ «"-iC.. «-«~~ ~» -»-> . -»
l»
T_butanol
(V<>/0)
0.01
0.00
0.04
0.01
0.00
0.03
r
NOx
(o/o)
9.0
13.8
10.0
8.6
13.2
11.3
Toxics
(o/o)
31.9
29.9
34.7
25.1
24.5
29.4
nee
voc
(o/o)
27.9
26.5
29.2
Table 5
by
Summer
Winter
PADD
I
II
III
I
II
III
t
Aromatics
(vo/o)
28.9
28.6
27.0
24.6
24.5
24.5
Benzene
(vo/o)
1.12
1.41
1.07
1.18
1.25
1.01
Olefins
(vo/o)
12.7
9.6
12.8
14.8
9.4
12.1
Oxygen
(wto/o)
0.34
0.34
0.13
0.30
0.33
0.11
E200
(o/o)
45.1
47.5
44.8
51.0
52.7
49.3
E300
(o/o)
82.8
81.9
80.5
84.4
84.6
82.6
RVP
(psi)
8.31
8.46
8.22
12.31
13.25
11.73
Sulfur
(ppm)
108
121
98
110
117
86
Ethanol
(vo/o)
0.74
0.97
0.05
0.68
0.92
0.07
MTBE
(vo/o)
0.39
0.00
0.60
0.30
0.00
0.41
ETBE
(vo/o)
0.00
0.00
0.01
0.00
0.00
0.01
TAME
(vo/o)
0.07
0.00
0.02
0.05
0.00
0.02
1
|tH|/ft11 J
T_butanol
(vo/o)
0.00
0.00
0.00
0.00
0.00
0.00
Exhaust
toxics
69.5
71.2
68.0
107.7
105.6
102.6
NOx
1264.9
1244.9
1253.9
1453.4
1405.3
1405.2
Table 6
259
------- |