Final Regulatory Impact Analysis and
Summary and Analysis of Comments
Phase II Gasoline Volatility Regulations
May 1990
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Mobile Sources
-------
Final Regulatory Impact Analysis and
Summary and Analysis of Comments
Phase II Gasoline Volatility Regulations
May 1990
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Mobile Sources
-------
Table of Contents
Page
Chapter 1 - Introduction 1-1
Chapter 2 - Period, Location, and Levels of Volatility
Control 2-1
I. Synopsis of NPRM Analysis 2-1
II. Synopsis of Phase I Volatility Control 2-1
III. Phase II RVP Control: Summary and Analysis of
Comments 2-2
A. Control Period
B. Location of Volatility Control
1. Equivalent Emissions Analysis
2. Assignment of RVP Class Designations
C. Levels of RVP Control for Classes 'A1, 'B1, and 'C'
Chapter 3 - Environmental Impact ' 3-1
I. Need For Ozone Control 3-1
A. Synopsis of NPRM
B. Summary and Analysis of Comments
1. Ozone Health Effects and Standards
2. Extent of Ozone Problem
3. Neces.sary Control Programs
C. Final Analysis
II. Emission Factors 3-7
A. Synopsis of NPRM
B. Summary and Analysis of Comments
1. Evaporative Emissions Model
2. Fuel Weathering
3. Emission Factor Modeling Inputs
C. Emission Factor Results
III. Emission Inventories 3-23
A. Synopsis of NPRM
B. Summary and Analysis of Comments
C. Emission Inventory Results
IV. Ozone Modeling 3-31
A. Air Quality Projections
1. Synopsis of NPRM
2. Summary and Analysis of Comments
3. Final Analysis
B. Butane and Oxygenate Reactivity and
Oxygenated Blend Environmental Impacts
1. Synopsis of NPRM
2. Summary and Analysis of Comments
-------
V. Effects of RVP Control on Benzene Emissions and 3-44
Health Effects
A. Synopsis of the NPRM
1. Benzene Emissions as a Function of Fuel
Parameters
2. Effect of RVP Control on Fuel Composition
3. Effect of RVP Control on Benzene Emissions
4. Nationwide Benzene Emissions
5. Cancer Incidence Analysis
B. Summary and Analysis of Comments
C. Final Analysis
Chapter 4 - Economic Impact 4-1
I. Refining Costs 4-1
A. Synopsis of NPRM Analysis
B. Summary and Analysis of Comments
1. Feasibility of and Leadtime for
the Second Phase of RVP Controls
2. Cost of RVP Control
C. Refinery Cost of RVP Control
II. Effect of Volatility Control on the Butane and 4-25
Pentane Markets
A. Synopsis of NPRM Analysis
1. Displacement of Butane
2. Economic Effects of Butane Displacement
B. Summary and Analysis of Comments
1. Displacement of Butane
a. Amount Displaced
b. Effects of Replacing Displaced
Butane
2. Economic Effects of Butane Displacement
a. Depression of Butane Prices
b. Alternative Uses for Butane
3. Displacement of Pentane
III. Effect of Volatility Regulations on Imports 4-32
IV. Effect of Volatility Regulations on Increased
Energy Density and Evaporative Emissions Recovery 4-34
V. Effect of Volatility Regulations on Driveability
and Safety 4-35
A. Synopsis of Draft Regulatory Impact Analysis
l. Volatility Increases and Driveability
Problems
2. Driveability Cost Estimation
B. Summary and Analysis of Comments
1. Hot Temperature Driveability
2. Cold Temperature Driveability
3. Cold Temperature Low Volatility Fuel
Safety
4. Hot Temperature Fuel Safety
ii
-------
VI. Enforcement Cost of Volatility Regulations 4-51
i
Chapter 5 - Analysis of Alternatives 5-1
I. Background 5-1
A. Introduction
B. Synopsis of NPRM Methodology
C. Summary and Analysis of Comments
1. Basic Model
D. Synopsis of Phase I Final Analysis
II. Phase II Volatility Control 5-5
A. Summary of Methodology
B. Summary and Analysis of Comments
C. Inputs for C/E Calculations ;
D. Cost-Effectiveness Results
Chapter 6 - Control of Volatility of Alcohol-Blended
Fuels . 6-1
I. Background 6-1
II. Ethanol Blends 6-1
A. Air: Quality Related Issues
1. Summary and Analysis of Comments
B. Economic Issues
1. Summary and Analysis of Comments
III. Methanol Blends 6-4
A. Summary and Analysis of Comments
111
-------
CHAPTER 1
INTRODUCTION
This document comprises most of the technical
documentaitori supporting the second of two phases of national
gasoline volatility control proposed on August 19, 1987 (52 FR
31274). In addition, this document summarizes comments
received on Phase II control and presents EPA's responses to
them. Further documentation for Phase II regulations including
analysis of enforcement comments exists in the Phase I Final
Rule (54 FR 11868) and in the Final Regulatory Impact Analysis
supporting that action. In addition, EPA has assembled
responses to a number of questions regarding enforcement and
implementation of RVP controls (see Docket A-85-21, Document
IV-A-10).
The analyses presented in this document incorporate
several assumptions about characteristics of an ultimate Phase
II program, as described below. If the final program differs
in minor respects from the program assumed here, the
conclusions reached in this document should be taken in that
context.
For purposes of analysis in this document, EPA assumes
that a Phase II gasoline volatility control program will be a
national program (excepting Hawaii and Alaska) reducing
volatility during the months of common ozone violation. The
air quality and economic impact analyses assume a set of Reid
Vapor Pressure (RVP) standards beginning in 1992 and varying by
state according to climate and elevation. The climate-based
analysis is described in Chapter 2.
Following this introduction, .Chapter 2 also addresses
questions of timing, location, and levels of control. Chapters
3 and 4 analyze air quality and economic issues, respectively.
Finally, Chapters 5 and 6 analyze alternatives for alcohol-free
gasoline and alcohol blends, respectively.
A limited supply of individual copies of this document
will be available through Ms. Jackie McManus, U.S. EPA, 2565
Plymouth Road, Ann Arbor, MI 48105 (phone: 313-668-4756).
-------
CHAPTER 2
PERIOD, LOCATION, AND LEVELS OF VOLATILITY CONTROL
Synopsis of NPRM
In the volatility control NPRM (52 FR 31274, August 19,
1987), EPA proposed to limit the volatility level of gasoline
for the period of May 16 through September 15 in order to
control evaporative emissions and reduce high levels of ozone
present in many urban areas across the country. The proposal
called for volatility levels reduced to 9.0 psi RVP in ASTM
Class C areas, with proportional reductions in all other
classes in order to achieve equivalent emissions nationwide.
The current state-by-state ASTM class designations were used in
the proposal with the exception that dual-classification areas
were assigned the lower of the two volatility classes. Hawaii
and Alaska were exempted from the program due to the fact that
they have no ozone" problem and both have independent gasoline
supply networks.
Due to the necessary lead time for this level of
volatility reduction, the control strategy was split into two
phases. The first phase would reduce volatility to levels for
which very little new capital expenditure would be needed.
This first phase was proposed to begin in the summer of 1989
and limit volatility levels to 10.5 psi RVP in Class C areas
with proportional reductions in Classes A and B. The final
level of control (9.0 psi RVP in Class C -areas with
proportional reductions in Classes A and B) was proposed to
begin in the summer of 1992.
II. Synopsis of Phase I Volatility Control
The final rule for the first phase of volatility control,
which took into account the comments in response to the NPRM
pertaining to Phase I control, was published in 1989 (54 FR
11868, March 22, 1989). It was enforced beginning in the
summer of 1989. The three main issues in the rulemaking—
period of control, location of control, and levels of RVP in
Classes A, B, and C are summarized below.
To determine the period of enforcement we took into
account the transition times needed to blend gasoline to and
from the control levels (six weeks at the beginning of the
season and one week at the end) . The period of enforcement for
terminals and all upstream parties is May 1 through September
15. For service stations and other end users, the period is
June 1 through September 15. This later date at the beginning
of the control season allows the gasoline at retail stations
with slower turnover times more time to blend down to
controlled levels. These small stations mav otherwise need
-------
2-2
to try to influence the refiners to begin production of
controlled fuel earlier on .their behalf, with uncertain
results. Due to the short lead time available, the program's
start was delayed for the summer of 1989 only.
Volatility classifications for each state ' or part of a
state were changed slightly from ASTM classifications.
Although EPA proposed in the NPRM to require the lower of the
two RVP classes for states which had dual-classifications
according to the ASTM system, this approach was reexamined for
the Phase I final rule. Based on whether or not the state was
in nonattainment of the ozone standard and on the uncontrolled
diurnal index (UDI) for that state, the more appropriate
classification was determined. The UDI analysis was also done
for single-classification states to determine if any changes to
the ASTM classification system would enhance the workability of
the control program without the loss of nonattainment-area
emission benefits. It was determined that a more relaxed
standard during some months of the control period would be more
appropriate for some states. A list of these states and
months, along with a list of the final Phase I classifications
of each state in each month can be found in the Phase I Final
RIA.
Although EPA had proposed in the NPRM to limit volatility
for Phase I to 10.5, 9.1, and 8.2 psi in Classes C, B, and A,
respectively, the final regulation limited volatility to 10.5,
9.5 and 9.0 psi. The changes for Class B and A areas assured
that new capital equipment, and thus greater lead time, would
not be needed to reduce volatility levels below 9.5 RVP in
Class B areas and below 9.0 RVP in Class A areas.
Ill. Phase II RVP Control: Summary and Analysis of Comments
The second phase of volatility control was proposed to
begin in May of 1992 and to further reduce the levels of
gasoline volatility down to 9.0, 7.8 and 7.0 psi RVP for Class
C, B, and A areas, respectively. Comments received which
pertain to the period, location, and levels of Phase II control
are summarized below with EPA's responses to these comments.
Although EPA addressed all comments dealing with the
state-by-state volatility classifications for the Phase I final
rule, EPA has reviewed those Phase I classifications in greater
detail and the results are presented below in Section III.B.
A. Control Period
The topic of the period during which RVP should be
controlled involves two related issues: l) how long of a
transition period is needed to blend fuel in the distribution
system down to compliance levels, and 2) which months or parts
of months need control. Based on these answers an appropriate
-------
2-3
period of control can be determined. These topics are
addressed below.
The majority of comments pertaining to the control period
dealt with the transition times necessary to blend the gasoline
down to the desired volatility level. These comments are
mentioned in. more detail in Chapter 1 of the Final RIA for
Phase I. These comments also apply to a Phase II program, In
brief, the comments mentioned that because the RVP reduction in
the proposal is greater than the current winter-to-summer
reduction, more transition time will be necessary to blend down
the uncontrolled high RVP fuel to the controlled RVP levels.
An analysis of the transition phase was done in the Phase I
FRIA and average transition times were calculated. The
analysis also showed that, although Phase II reductions are
deeper, transition times for a Phase II program should be
essentially the same as for the Phase I program. Therefore,
the average transition time at the beginning of the control
period should be about six weeks. At the end of the control
period, the average transition time was calculated to be only
one week. These figures include the effect of backmixing in
storage tanks and the average transportation times to the
terminals in pipelines and from terminals to service stations.
Other comments referred to potential problems of safety,
driveability, and increased HC emissions during the transition
period due to low-RVP fuel reaching the consumer early in the
transition period when there may still be cooler temperatures.
EPA does not believe such problems will occur in any general
way during the transition. A more detailed response to these
comments is found in Chapter 4 of this Phase II Final RIA.
The other key topic dealing with the control period is the
length of time the regulated fuel need be required. In Phase
I, the length of the control period was selected based on the
need for ozone control (i.e., during the period of the year
most ozone exceedances occurred). The analysis used to support
this selection is described in the Phase I Final RIA. For
Phase II, the control period was re-evaluated based on updated
ozone exceedance data. In this analysis, 95.5 percent of
nonattainment area ozone exceedances for 1986-88 (excluding CA
and Houston) occur during the Phase I period of May 1 through
September 15. Table 2-1 shows the number of ozone exceedances
for each semimonthly period during those 3 years, indicating
that 1.9 percent of exceedances occurred in late April and 0.8
percent occurred in late September; only 2.1 percent occur in
the nonsummer period from October 1 through April 15.
As shown in the Phase I Final RIA analysis of RVP
transition time for the 70th percentile refinery, since
dispensed RVPs would be near controlled levels a few weeks
earlier than required, substantial control would result in late
April before a May 1 starting date. However, since dispensed
-------
2-4
Table 2-1
1986-88 Nationwide Ozone
Exceedances (Excluding CA and Houston)
Number of • Percentage
Exceedances of Total
January 1-15 0 0.0
January 16-31 2 0.1
February 1-15 1 0.1
February 16-28 1 0.1
March 1-15 5 0.3
March 16-31 6 0.3
April 1-15 2 0.1
April 16-30 35 1.9
May 1-15 26 1.4
May 16-31 . 154 8.2
June 1-15 147 7.8
June 16 - 30 352 18.7
July 1-15 303 16.1
July 16 - 31 379 20.1
August 1-15 284 15 . 1
August 16 - 31 121 6.4
September 1-15 34 1.8
September 16-30 15 -0.8
October 1-15 6 0.4
October 16-31 8 0.4
November 1-15 0 0.0.
November 16-30 l o.l
December 1-15 l O.l
December 16 - 31 l 0 .1
1884 100.0
-------
2-5
RVPs would increase relatively quickly after the control period
ends, the effects of RVP control would extend perhaps a week
after a September 15 ending date. When these conclusions are
superimposed on the distribution of ozone exceedances in Table
2-1, the Phase I control period of May 1 through September 15
continues to be the appropriate period to assure control from
mid-April into late September, when ozone exceedances are
frequent.-
B. Location of Volatility Control
EPA received several comments in response to the NPRM
regarding the location of the various classes of controlled
fuel. These were responded to in detail in the Final RIA for
Phase I and are summarized below.
In the NPRM, as discussed earlier, EPA proposed to enforce
standards during the control period based generally on the ASTM
monthly volatility classes and boundaries. However, EPA
proposed to eliminate all transition (dual-classification)
months, requiring instead the lower of the two volatility
classes during those months. Various responses to this
proposed scenario were received. Marathon and Phillips both
supported using the ASTM schedule. Phillips also commented
that the transition months should be kept. The California Air
Resources Board (GARB), on the other hand, did not agree with
enforcing the ASTM classifications or boundaries as they apply
to California. They requested that the new system should match
California's current RVP requirements.
Other comments were received which dealt with the required
location of the controlled fuel. Some comments stated that
controlled fuel should only be required where it is needed
(i.e., in ozone nonattainment areas). Others stated that
refiners cannot generally segregate fuel between attainment and
nonattainment areas, but rather only over broad geographical
areas. Distribution problems, some comments stated, would
appear as a result of a patchwork of different volatility
controls across the country. ;
With these comments in mind, the "map" for Phase I of
volatility control was developed. The "map" was based mainly
on the ASTM state volatility classifications. ASTM based their
classifications on an early understanding of RVP effects on
driveability. Phase I improved upon this system for air
quality purposes by modeling the effect of RVP on emissions.
For this second phase of volatility control, an improved
modeling analysis was possible, which was based on achieving
equivalent vehicle emissions across the country. This analysis
was the basis for determining state volatility classifications,
and is described below.
-------
2-6
1. Equivalent Emissions Analysis
EPA decided to do a new analysis to improve upon the ASTM
system of RVP designations across the country. MOBILE 4.0 was
available and was used for this analysis, which aimed at
determining an RVP for each state in each month that would
result in equivalent vehicle emissions nationwide. MOBILE4.0
was used to determine RVPs for each state for the months May
through October which would result in emissions equivalent to
the average of those in Phase I Class C areas in July when
using 9.0 RVP fuel.
Temperatures for use in the MOBILE4 runs were generated as
follows: First, temperature data were obtained for all ozone
monitors on days when the ozone level exceeded 0.08 ppm. As
temperature data are not available from the ozone monitor
station, temperatures were taken from the weather station most
representative of the conditions in the immediate vicinity of
the ambient monitor. Ozone observation data were used from the
years 1985-87, the most recent data available at the time.
Then, temperatures from days of high-ozone measurements in each
month at each monitor were matched with the date of the
observation. Using these temperature data, MOBILE4.0 was used
to calculate an emission factor for each of the top ten ozone
days in each month at each monitor. Next, for each month, the
average emission factor at each monitor was calculated. A
state's RVP for each month was then calculated as the RVP which
corresponded to the population-weighted average emission factor
of all monitors in the state.
As some Northwest states had no ozone data, the
temperatures used for these states for May through September
were the 30-year July-average temperatures.
A more detailed explanation of this analysis is given in
the report "Methodology for Estimating the Reid Vapor Pressure
that Results in Equivalent Nationwide Motor Vehicle Hydrocarbon
Emissions."C1]
Due to unusual circumstances, four exceptions to the above
methodology were made. First, due to California's rapidly
changing meteorology from moderate coastal temperatures to arid
inland temperatures, the temperature monitors used in the
calculation of an equivalent emission RVP for California were
rechosen. The computer analysis originally used monitors from
the moderate, coastal areas for areas where inland temperatures
were more appropriate, thereby assigning California
unexpectedly high RVPs throughout the summer, Therefore,
different weather stations expected to have more appropriate
temperatures for the cities involved were used for matching
ozone observations with temperatures.
-------
2-7
There was an anomaly in Arkansas where an extremely low
RVP was assigned to the state in July while no other state
around it was assigned such a low RVP. On closer examination,
we found that this assignment was due to high temperatures in
Little Rock, which is not an ozone nonattainment area. Also,
Memphis, Tennessee, whose CMSA includes the only nonattainment
area in Arkansas, was not used in the RVP calculation as there
were no monitors for the Memphis CMSA in Arkansas. Therefore,
the July RVP of Arkansas was set by simply substituting
Tennessee's calculated July RVP.
Another exception to the equivalent emission analysis was
in New Jersey. Here, a very low RVP was assigned in May and
June when no state around it was as low, while its RVP went
much higher for the hotter summer months of July through
September. We found this to be caused by two or three days
each month of unusually high New York City temperatures
(specifically, the New York City portion of New Jersey). The
New York City population was enough to outweigh those of
Atlantic City and Philadelphia, bringing New Jersey's RVP down
very low. However, contrary to expectations, these high
temperatures did not appear at other New York City locations.
Because of this anomaly, New Jersey was reassigned an RVP,
excluding these extreme days from New York City in May and June.
Finally, the District of Columbia was assigned Virginia's
calculated summertime RVPs, since there were not enough ozone
(and thus temperature) data to form a basis for an independent
calculation of RVP for the District.
The calculated equivalent emission RVPs for each state are
presented in Table 2-2. These values incorporate the changes
described above.
2. Assignment of RVP Class Designations
In order to maintain simplicity, EPA decided against
assigning individual RVP standards to states, but decided
instead to maintain the approach established by ASTM of having
three summertime volatility classes. These classes will be
designated as 'A1, 'B', and 'C1 in order to distinguish them
from ASTM classes A, B, and C.
As part of the equivalent emission analysis described
previously, states were assigned individual RVPs for each month
which would result in emissions in each state comparable to
those of the average Phase I Class 'C' state with 9.0 RVP fuel
in July.tl] The analysis then analytically picked RVP ranges
for Classes 'A', 'B', and 'C1 which would result in the minimum
mean-squared difference between the emission inventory for the
class and that of the baseline Class 'C' areas. The ranges
-------
2-8
Table 2-2
Equivalent Emission RVP Calculations
State May June July Aug Sept
Alabama 9.6 8.2 7.9 7.9 9.1
Arizona 7.0 7.0 7.0 7.0 7.0
Arkansas 9.3 8.6 7.9* 7.9 8.7
California** 8.7 8.0 7.5 7.6 8.2
Colorado 9.1 7.6 7.0 7.4 9.0
Connect icut 9.4 9.5 8.7 8.9 10.2
Delaware 9.5 9.0 8.6 9.1 9.5
Dist. of Columbia*** 8.9 9.3 8.3 8.4 9.0
Florida 8.9 8.1 8.2 7.9 8.5
Georgia 9.4 8.2 7.7 8.2 9.3
Idaho 8.8 8.8 8.8 8.8 8.8
Illinois 9.7 8.8 8.8 8.9 9.6
Indiana 9.6 8.9 9.0 9.3 9.6
Iowa 9.7 8.8 8.4 9.2 9.6
Kansas 9.8 8.5 7.7 8.3 8.6
Kentucky 9.5 9.3 8.6 8.8 9.6
Louisiana 9.5 8.6 8.0 8.1 8.6
Maine 9.8 10.0 9.6 9.3 10.1
Maryland 9.2 8.9 8.4 8.5 9.5
Massachusetts 9.6 9.8 9.6 9.3 9.8
Michigan 9.5 9.4 8.9 9.5 9.9
Minnesota 9.7 8.7 9.2 9.8
Mississippi 9.5 8.4 7.6 7.9 8.9
Missouri 9.6 8.7 8.0 8.8 9.2
Montana 9.8 9.8 9.8 9.8 9.8
Nebraska 9.8 8.5 8.4 9.3 9.4
Nevada 7.7 7.0 7.0 7.0 7.9
New Hampshire 10.0 9.9 9.4 9.5 10.2
New Jersey 8.3 8.2****8.8 8.8 9.7
New Mexico 9.2 7.8 7.0 7.1 8.9
New York 9.1 9.2 8.8 9.3 9.6
North Carolina 9.6 8.6 7.9 8.1 9.6
North Dakota 9.9 9.9 9.9 9.9 9.9
Ohio ' 9.7 9.5 9.2 9.3 9.8
Oklahoma 9.5 9.1 7.9 8.0 8.8
Oregon 8.9 8.5 8.0 7.8 9.1
Pennsylvania 9.5 9.2 9.0 9.3 10.0
Rhode Island 9.3 9.6 9.4 9.3 9.9
South Carolina 9.4 8.3 7.5 8.0 9.4
South Dakota 9.4 9.4 9.4 9.4 9.4
Tennessee 9.6 8.7 7.9 8.1 9.2
Texas 9.2 8.4 7.7 7.6 8.3
-------
2-9
Table 2-2 - (Cont'd)
Equivalent Emission RVP Calculations
State May . June July Aug Sept
Utah 8.9 7.1 7.0 7.0
Vermont 10.0 9.8 9.5 9.3
Virginia 8.9 9.3 8.3 8.4
Washington 10.0 9.6 9.3 9.2
West Virginia 9.6 9.6 8.8 8.9
Wisconsin 9.5 8.5 9.3 9.8 9.6
Wyoming 9.7 9.7 9.7 9.7 9.7
* The only nonattainment area in Arkansas is in Memphis
CMSA. Since Little Rock dominated the calculation of
average temperature, we substituted Tennessee's July RVP.
** California RVP values are adjusted, as described in the
text.
*** Lacking sufficient data specific to Washington DC, we
assigned Virginia's calculated RVPs to the District.
**** Recalculated ignoring the disproportionate effect of New
York City, as described in the text.
-------
2-10
chosen by this analysis were the following:
Class 'A1
Class 'B'
Class 'C'
7.0 psi RVP
7.1-8.5 psi RVP
>_8.6 psi RVP
The midpoints of the RVP ranges were then used as the RVP for
each class of fuel. This resulted in 7.0 and 7.8 psi RVP fuel,
respectively, for Classes 'A1 and 'B'. Class 'C1 fuel was set
at 9.0 ps i RVP.
With the availability of equivalent emission RVPs for each
state in each month (May through September; October data was
not used), it seemed appropriate to assign the states, when
possible, into the class that resulted in the best
approximation to equivalent emissions for them individually.
Therefore, although it meant deviating from the minimum
mean-squared difference, EPA chose to reassign the RV? ranges
per class. The cutoff points for each class were the midpoints
between the proposed standards of 7.0 and 7.8 RVP, and between
7.8 and 9.0 RVP. Thus, the following breakdown was used:
Class 'A1: 7.0-7.3 psi RVP
Class 'B1: 7.4-8.4 psi RVP
Class 'C1 : >_8 . 5 psi RVP
>_8.5 psi RVP
Table 2-3 presents the equivalent emission RVP
designations when this breakdown is applied to the calculated
values presented in Table 2-2. (In order to check the
appropriateness of the 7.0 and 7.8 psi RVPs in Classes 'A' and
'B', respectively, a separate analysis was performed and is
discussed in the next section.)
C. Levels of RVP Control for Classes 'A', '3', and 'C'
In the NPRM, EPA proposed an RVP standard of 9.0 psi in
Class 'C' areas with standards representing proportional RVP
reductions in Classes 'A' and 'B' (the percent reduction in
going from 11.5 to 9.0 in Class 'C1 being applied also to
Classes 'A1 and 'B', to -result in levels of 7.0 and 7.8 psi).
Many of the comments disagreed with this approach and stated
that the RVP levels should be based on equivalent
cost-effectiveness in each class rather than proportional
reductions. Those opposed to proportional volatility
reductions stated that proportional reductions outside of Class
'C' areas are not cost-effective. NRDC commented that the
amount of control in an area should be based on how much need
the area has for control.
Some comments also stated that the final RVP levels should
not create cost burdens that disproportionately affect refiners
(i.e., refiners supplying Classes 'A' and 'B' should not carry
the greatest cost burden). The Petroleum Marketers Association
-------
2-11
Table 2-3
Equivalent Emission Analysis Designations
State May June July Aug Sept
Alabama 9.0 7.8 7.8 7.8 9.0
Arizona 9.0 7.0 7.0 7.0 7.0
Arkansas . 9.0 7.8 7.8 7.8 9.0
California 9.0 7.8 7.8 7.8 7.8
Colorado 9.0 7.8 7.8 7.8 9.0
Connecticut 9.0 9.0 9.0 9.0 9.0
Delaware 9.0 9.0 9.0 9.0 9.0
Dist. of Columbia 9.0 9.0 7.8 7.8 9.0
Florida 9.0 7.8 7.8 7.8 9.0
Georgia 9.0 7.8 7.8 7.8 9.0
Idaho 9.0 9.0 9.0 9.0 9.0
Illinois 9.0 9.0 9.0 9.0 9.0
Indiana 9.0 9.0 9.0 9.0 9.0
Iowa 9.0 9.0 9.0 9.0 9.0
Kansas 9.0 7.8 7.8 7.8 9.0
Kentucky 9.0 9.0 9.0 9.0 9.0
Louisiana 9.0 7.8 7.8 7.8 9.0
Maine 9.0 9.0 9.0 9.0 9.0
Maryland 9.0 9.0 7.8 9.0 9.0
Massachusetts . 9.0 9.0 9.0 9.0 9.0
Michigan 9.0 9.0 9.0 9.0 9.0
Minnesota 9.0 9.0 9.0 9,0 9.0
Mississippi 9.0 7.8 7.8 7.8 9.0
Missouri " 9.0 7.8 7.8 7.8 9.0
Montana 9.0 9.0 9.0 9.0 9.0
Nebraska 9.0 9.0 9.0 9.0 9.0
Nevada 9.0 7.8 7.8 7.8 7.8
New Hampshire 9.0 9.0 9.0 9.0 9.0
New Jersey 9.0 9.0 9.0 9.0 9.0
New Mexico 9.0 7.8 7.8 7.8 9.0
New York 9.0 9.0 9.0 9.0 9.0
North Carolina 9.0 7.8 7.8 7.8 9.0
North Dakota 9.0 9.0 9.0 9.0 9.0
Ohio 9.0 9.0 9.0 9.0 9.0
Oklahoma 9.0 7.8 7.8 7.8 9.0
Oregon 9.0 7.8 7.8 7.8 9.0
Pennsylvania 9.0 9.0 9.0 9.0 9.0
Rhode Island 9.0 9.0 9.0 9.0 9.0
South Carolina 9.0 7.8 7.8 7.8 9.0
South Dakota 9.0 9.0 9.0 9.0 9.0
Tennessee 9.0 7.8 7.8 7.8 9.0
Texas 9.0 7.8 7.8 7.8 7.8
Utah 9.0 7.0 7.0 7.0 7.0
Vermont 9.0 9.0 9.0 9.0 9.0
Virginia 9.0 9.0 7.8 7.8 9.0
-------
2-12
Table 2-3 (cont'd)
Equivalent Emission Analysis Designations
State May June July Aug
Washington 9.0 9.0 9.0 9.0
West Virginia 9.0 9.0 9.0 9.0
Wisconsin 9.0 9.0 9.0 9.0
Wyoming 9.0 9.0 9.0 9.0
-------
2-13
of America stated that the reductions should not be so
stringent as to harm independent sources of supply, regardless
of how the reductions are taken.
From these comments there appear to be at least three
alternatives for determining the RVP level for each class:
equivalent cost-effectiveness in each class, equivalent
emissions across the country, or RVPs based on the extent of
need in the area. EPA has in the past and in this final rule
used cost-effectiveness as one way to compare control programs
among one another (see Chapter 5). For this control program,
by using equivalent cost-effectiveness, necessary control in
some areas would be foregone even though it may be
cost-effective. The option of establishing RVPs based on the
extent of need in the area would be of interest in that
especially bad areas for ozone would receive additional
control. However, in a national fuel regulatory program,
significant variations among local control levels is not
consistent with the more regional nature of the gasoline supply
system. EPA believes that focusing on equivalent emissions
nationwide, as based on climatic conditions and elevation, is
the most appropriate approach. In this case, control in all
three classes is very cost-effective despite the fact that some
classes are relatively lower or higher in cost-effectiveness
(again, see Chapter 5).
Although EPA proposed an approach of proportional
reductions in RVP in each class, the method of equivalent
emissions in each class now appears actually to be more
consistent with our goal. This is due to the fact that use of
proportional RVP reductions focuses on the amount of change in
RVP (and emissions) rather than on the final level of
emissions. Therefore, EPA calculated RVPs for each class based
on the analysis of comparable vehicle emissions across the
country described previously.
In the NPRM, EPA proposed numerical volatility levels of
9.0, 7.8 and 7.0 psi RVP in Classes 'C', 'B', and 'A',
respectively. The appropriateness of these proportional
reduction standards was checked by using the equivalent
emissions analysis described previously. By using the RVP
limits for each class as described above (7.0-7.3 psi for 'A',
7.4-8.4 psi for 'B'), states in Classes 'A' and 'B' in July
were segregated. For each of the two classes, July RVP levels
for each state in the class were population-weighted to
generate a population-weighted average RVP level for both Class
'A1 and Class 'B'. This population-weighting of equivalent
emission RVPs resulted in average levels of 7.0 and 7.8 psi RVP
for Classes 'A' and 'B1, respectively, confirming the
appropriateness of the proposed levels.
-------
2-14 .
References (Chapter 2)
1. "Methodology for Estimating the Reid Vapor Pressure
that Results in Equivalent Nationwide Motor Vehicle Hydrocarbon
Emissions," Mark A. Wolcott, EPA, OAR, QMS, ECTD, TEB and
Dennis F. Kahlbaum, Computer Sciences Corporation, June, 1990.
-------
CHAPTER 3
ENVIRONMENTAL IMPACT
This chapter examines the environmental impact associated
with a second phase of RVP controls as proposed to follow the
Phase I program promulgated in 1989. It discusses the control
of volatile organic compounds (VOC) emissions in order to lower
tropospheric ozone levels; including discussions of the need
for ozone control, the expected emission reduction effect from
RVP controls, and the impact of these controls on ozone
attainment.
I. Need For Ozone Control
A. Synopsis of NPRM
Ozone is a powerful oxidant which is formed in the
troposphere by photochemical reactions of VOC and oxides of
nitrogen (NOx). Ozone 'affects humans by irritating the
respiratory system and reducing lung function. Laboratory
studies suggest that it also may actually damage lung and other
tissues. This damage may impair breathing and reduce immunity
to disease for people in good health, and may be even more
severe for people with pre-existing respiratory diseases. In
plants, oxidation by ozone can impair tissue function and
reduce the yield of some crops. Oxidation by ozone may also
damage materials such as rubber products.
Based on the air quality analysis presented in the NPRM,
many areas of the nation continued to violate the ozone NAAQS.
During the three-year period of 1982-1984, EPA determined that
73 urban areas throughout the nation exceeded the ozone
standard. Twelve of these areas were located in California.
In order to determine the need for future hydrocarbon
control, EPA not only looked at the present state of air
quality with respect to ozone, but also at projected future air
quality trends. Estimates of future air quality without
further controls were made for the 61 non-California urban
nonattainment areas. Since California has separate vehicle
standards, and thus California vehicles were not accounted for
in EPA's vehicle emission model, California was excluded from
the analysis. Current attainment areas were excluded from the
future air quality projections even though some attainment
areas were close to the standard and may have been projected to
become nonattainment areas in the future.
The air quality modeling relied on certain assumptions
regarding emission rates, growth rates, control technologies,
emission standards, and control efficiencies. Based on these
analyses of future air quality, EPA projected that there likely
would be improvements in air quality from 1988 to 1995 due to
the effect of current emission standards for mobile and
stationary sources, but that most of the large urban areas
-------
3-2
modeled will remain in nonattainment throughout this period.
However, without further controls, growth was projected to
offset these improvements and to cause the air quality problem
to worsen after 1995. Based on estimates made by the EPA, VOC
emission reductions of 50 to 80 percent appeared to be
necessary to bring some urban areas into compliance for ozone.
B. Summary and Analysis of Comments
The comments received concerning the need for ozone
control addressed several different aspects of this issue. The
comments can be divided, into three main areas: 1) the health
effects of ozone exposures, 2) the extent of the ozone problem,
and 3) the more general issue of the type of control needed.
1. Ozone Health Effects and Standards
Comments on health effects varied greatly. The Natural
Resources Defense Council (NRDC) stated that at concentrations
of 50 percent of the current ozone NAAQS, suppression of the
immune system has been observed in laboratory animals. The
American Lung Association (ALA) also agreed that the current
ozone standard does not adequately protect public health. ALA
commented that clinical studies have shown that adverse
respiratory health effects result from experimental ozone
exposure at the current standard level. The Northeast States
for Coordinated Air Use Management (NESCAUM) stated that ozone
pollution is one of the most serious and widespread public
health problems in the northeastern part of the country. They
also stated that recent health data strongly suggest that the
existing ozone NAAQS may not be strict enough to protect
public health. Others commented, though, that any health
effects are very short term in nature and that those showing
health effects recover quickly. API commented that the line
between attainment and nonattainment seems arbitrary and that
it is improbable that any real public health consequences are
associated with the transition from nonattainment to
attainment. Related to agricultural effects, NESCAUM claimed
that much terrestrial damage has occurred due to ozone exposure
at ambient concentrations well below the current health
standard.
Commenting on the margin of safety required by the Clean
Air Act, Chrysler felt that it is irrelevant whether or not an
area is projected to be in borderline compliance, as long as
they reach attainment. They felt that there already exists a
significant safety margin built into the current ozone standard
to protect public health. Therefore, Chrysler argued, as long
as compliance occurs, no extra margin of safety is needed.
NRDC, however, stated that recent scientific evidence shows
that the ozone NAAQS will have to be tightened in order to
protect the public health with an adequate margin of safety.
-------
3-3
These comments indicate that additional information may be
needed in order to resolve concerns about the appropriate level
of the ozone standard. This rulemaking, however, is not the
proper forum to attempt such a resolution. The Agency is
already investigating this issue in its periodic (5-year cycle)
review of each NAAQS. Should it be determined that
modifications to the current ozone standard are necessary, they
will be made in a separate rulemaking. At this time, control
decisions have to be based on the current NAAQS standard.
In response to Chrysler's comment, it is true that the
Clean Air Act requires an adequate margin of safety to protect
the public health. Thus, EPA does not base its regulatory
decisions on areas that are in "borderline attainment." This
does not, however, mean that the extent of an area's compliance
is irrelevant. Concerns such as NKDC's on the margin of
safety, as well as concerns about the accuracy of ozone
projections, make the projection of an area to be in borderline
attainment very significant.
2. Extent of Ozone Problem
Regardless of the level of ozone at which an unhealthy
environment exists, the extent of the ozone problem in this
nation is also under debate. Phillips Petroleum commented that
few urban areas frequently or significantly exceed the ozone
standard. It also commented that, with the exception of
Southern California, the ozone problem is largely unnoticed by
the general public since exceedances of the ozone standard are
so few. Other commenters also noted that on an hour-by-hour
basis, most areas are in compliance with the ozone standard
more than 99 percent of the time. API agreed with EPA that
there is an ozone problem; however, it felt this problem has
been overstated by the EPA. Sun Oil also felt that EPA has
overstated the problem. They believe that, except for the Los
Angeles area, most nonattainment areas are in compliance with
the ozone standard during most of the summer ozone season.
MVMA commented that EPA has already determined through its
NAAQS process that an ozone concentration of 0.12 ppm or less
is by definition adequate to protect public health and welfare
from any adverse effects.
Many commenters claimed that EPA has overestimated the
current ozone problem because the method for determining
whether or riot an area was in attainment was erroneous. They
feel that the method does not represent a quantitative view of
ozone exposure and that it overstates exposure duration and
total population exposure. Sohio and API both commented that
determining attainment or nonattainment with a single monitor
is unrealistic and overstates the problem. Sohio feels that
the average of all monitors in the area may be more
representative of typical ozone season air quality. API and
Sohio stated that overall ozone levels are well below the
-------
3-4
standard, but that the current method of reporting ozone
nonattainment data does not show that. Sohio also feels that
since the design value is based on a reading from a single
monitor, it does not show the progress which has been achieved
in reducing ambient ozone levels; therefore, it does not
measure the extent of nonattainment. Sohio commented that
monitors with exceedances may not be in highly populated
areas. It said that in most nonattainment areas, the
population living or working in the area is much lower than the
total population. API and Sohio also stated that under the
current method, one year's data, even if it is
uncharacteristic, is enough to keep an area out of attainment
for a three-year period. One last commenter questioned whether
or not a one-hour standard for ozone was appropriate. They
believe that a more practical long-term solution would be going
to an eight or twenty-four hour average standard.
EPA does not deny that ozone levels in urban areas are
generally below the level of the standard of 0.125 ppm the
majority of the time, nor does it claim that its monitoring
system reports an ozone level to which the entire area is
exposed. Nevertheless, it does maintain that its current
method of determining air quality, with respect to ozone
levels, is reasonable. This is because the Agency considered
these issues in setting the level of the standard. As part of
that process, it was decided that one-hour ozone concentrations
greater than 0.125 ppm, in any part of an area, more than once
per year would be indicative of unacceptable air quality. EPA
will consider comments on modifying the nature of the ozone
standard at the appropriate time in its review of the ozone
NAAQS. However, as noted before, such modifications are
clearly beyond the scope of this rulemaking.
3. Necessary Control Programs
Many commenters stated that because many areas in the
nation still have not reached attainment of the ozone standard,
further control of HC emissions is necessary. Some commenters
argued for volatility controls, some argued for refueling
and/or improved vehicle evaporative controls, and some argued
for both. The Texas Air Control Board felt that onboard
controls would be a very cost effective strategy, as did API
and many other commenters. API, however, does note that before
forcing costly hydrocarbon controls, it would be useful to have
a better understanding of the role of NOx in ozone formation,
the role of transport, and the overall hydrocarbon inventory.
API cited the "National Acid Precipitation Report," which,
according to API, stated that two-thirds of VOC emissions east
of the Mississippi and three-fourths of VOC emissions west of
the Mississippi are from natural sources.
The Conservation Law Foundation of New England (CLF)
agreed with EPA's assumption behind the proposed volatility
-------
3-5
rule that vehicle-based controls alone are not enough to
address the short-term ozone nonattainraent problem. This is
because emission reductions from such controls only occur after
a long period of time needed for fleet turnover. It also felt
that EPA was correct in realizing that emission reductions
beyond those available at zero cost are necessary to achieve
short-term air quality improvement. It felt that the in-use
fuel volatility restrictions are technically feasible and cost
effective, and thus should be adopted.
NRDC and the American Lung Association felt that VOC
reductions beyond those currently required in nonattainment
areas are necessary to protect public health. They felt that
volatility, onboard, and Stage II controls are all necessary to
achieve attainment. NRDC stated that $2000 per ton should not
be a cost-effectiveness ceiling for ozone or carcinogen
controls.
NESCAUM, in commenting on the ozone problem in the
Northeast, stated that long-term attainment of the ozone
standard can only be reached through a region-wide program
based on all available control strategies. It also felt that a
reason for the failure to attain the ozone standard in the
Northeast has been an incomplete understanding of the complex
process of ozone formation. This would lead some states to
underestimate the required reductions to meet the ozone
standard, and to overestimate the ozone reductions obtained
from various control measures.
Some commenters, however, felt that ever-worsening air
quality is not a problem. Therefore, they claim that no
nationwide "crash" program is needed to reduce emissions.
Toyota felt that the effectiveness of current control programs
was being underestimated by EPA.
Clearly there is disagreement as to whether the proposed
controls are necessary. However, none of these comments
provided sufficient rationale to change EPA's position on VOC
control. The Agency remains convinced that cost-effective VOC
controls are the most appropriate ozone strategy at this time.
(It should be noted that ,no cost-effectiveness ceiling of $2000
per ton, or of any other value, has been established by EPA.)
Thus, the argument presented by commenters who argue against
implementing any program that in and of itself only reduces a
small fraction of total VOC, such as the control of refueling
emissions which only account for two percent of the VOC
inventory, is not valid. There are several important factors
in deciding whether or not certain VOC controls are reasonable,
including the cost effectiveness of the controls as well as the
size of the emission reduction. However, the size of the
emission reduction is not the most important factor since the
-benefits of many •smaller programs can account for large
emissions reductions when considered together.
-------
3-6
EPA recognizes that natural VOC emissions can play an
important role in ozone formation, and, in fact, EPA is
investigating this issue. The other issues such as the role of
NOx and of transport in ozone formation are also important and
are being investigated further. Nevertheless, this does not
alter the fact that control of VOC emissions is an effective
means of reducing ozone levels.
C. Final Analysis
As stated previously, the environmental impact analysis
contained in the NPRM was based on ozone data collected from.
1982 to 1984. For this analysis of the second phase of RVP
controls, the most recently available ozone data from 1986 to
1988 was used. The large number of nonattainment areas and the
high levels of ozone show that ozone remains a problem in many
large urban areas. Based on the 1986 to 1988 data, EPA -has
determined that there are currently 101 areas exceeding the
NAAQS ozone standard, with 10 of these located in California.
In estimating the environmental impacts of a second phase
of RVP controls, EPA based its analysis on the same
nonattainment areas which have been considered by the EPA
contractor, E.H. Pechan, in its analysis of the recent Clean
Air Act amendments.[1] In performing their analyses, Pechan
combined the Massachusetts nonattainment areas into one
statewide area, they eliminated the non-urban areas, eliminated
those areas with an ozone design value below 0.125 ppm (the
EKMA model cannot predict changes for such a small ozone
decrease), and eliminated the areas where ozone transport from
other areas is considered to be the cause of nonattainment.
After these adjustments are made, 81 nationwide urban ozone
nonattainment areas remain based on the 1986 to 1988 data, with
10 of these located in California.
It should be noted that in performing this analysis for
the second phase of volatility controls, EPA has modified its
approach to the nonattainment situation in two additional
ways. First, in the time since the first phase of controls was
promulgated, several Northeast states have begun regulating
summer gasoline RVP to 9.0 psi, the same level EPA recommended
in Chapter 2 for Class C areas. Because these states (New
York, New Jersey, Delaware, Massachusetts, Connecticut, Rhode
Island, Vermont, New Hampshire, and Maine) have already started
controlling RVP to the levels EPA would have promulgated with
this second phase of controls, they are not considered in this
analysis. Therefore, the environmental impact analysis, as
well as the economic impact analysis, do not include effects
which would be experienced in these Northeast states.
Second, unlike the analysis for the first phase of RVP
controls, California has been included in this analysis of the
second phase of volatility controls. California currently
-------
3-7
regulates summer gasoline RVP to 9.0 psi. This RVP is lower
than the level which EPA promulgated for Class B areas in the
first phase of controls; therefore, the EPA regulations had no
impact on California. However, the RVP level recommended in
Chapter 2 for California is 7.8 psi for most months. For this
reason, California now will be required to lower their summer
gasoline volatility even further. Thus, the environmental
impacts and economic impacts which will be experienced in
California have been included in this second phase of RVP
controls analysis.
It should be noted that MOBILE4.0 is based on emissions
data from vehicles certified to meet the Federal emissions
standards. However, California vehicles are required to meet a
separate set of standards and often have different vehicle
designs which lead to different emissions characteristics.
Therefore, a special version of MOBILE4.0 which included
California vehicle emission estimates, was used to estimate
California emissions. An explanation of the differences
between the release MOBILE4.0 and special California MOBILE4.0
models are described in a recent EPA memorandum.[2]
After these two adjustments have been taken into account,
EPA has determined that there are 70 non-Northeast
nonattainraent areas, with 10 of these areas located in
California. Table 3-1 contains a list of the non-Northeast
ozone nonattainment areas used for this analysis based on 1986
to 1988 data.
Based on the air quality modeling performed by EPA it is
estimated that with the current Federal motor vehicle program
and Phase I RVP controls, but without the second phase of RVP
controls, 67 of the 70 non-Northeast urban ozone nonattainment
areas would still be in nonattainment for ozone in 1990,
decreasing to 54 areas in 1995, 52 areas in 2000, and
increasing to 58 areas in 2005, and staying at that level in
2010. Therefore, further VOC controls appear to be necessary
to assist most of the current ozone nonattainment areas to
reach attainment in the future.
II. Emission Factors
A. Synopsis of NPRM
As an initial step to estimating the environmental impacts
of the control options, a computer model was used to calculate
vehicle HC emission factors in grams per mile for a range of
potential control programs. The emission factors were
calculated for exhaust, evaporative, and refueling hydrocarbon
(HC) emissions, and were combined with estimates of vehicle
miles travelled (VMT) and stationary source inventories to
obtain VOC inventory projections of future VOC emissions as
described in Section III.A below. The inventory projections in
-------
Table 3-1
Nationwide Urban Ozone Nonattainment Area Input Data For MOBILE4.0
1986-88 Ozone
Nonattainment Area
State of Massachusetts
Greater Connecticut CMSA
Manchester-Nashua, NH
Portland, ME NECMA
Portsmouth-Dover-Roch.,
NH-ME
Providence, RI CMSA
Atlantic City, NJ
Buffalo, NY CMSA
Glens Falls, NY
New York, NY CMSA
Poughkeepsie, NY
Allentown-Bethlehem,
PA-NJ
Altoona, PA
Baltimore, MD
Charleston, WV
Erie, PA
Ha rr i sburg-Lebanon-
Carlisle, PA
Huntington-Ashland,
WV-KY-OH
Johnstown, PA
Lancaster, PA
Norfolk-Va. Beach-
Newport News, VA
Parkersburg-Marietta
WV-OH .
Philadelphia, PA CMSA
Pittsburg, PA CMSA
Reading, PA
Richmond-Petersburg, VA
1987 Avg
In-Use
RVP (psi)
10.8
10.8
10.8
10.8
10.8
10.8
11.3
11.2
11.2
11.2
11.2
11.3
11.3
11.0
11.0
11.7
11.3
Summer Avg
Temperature
(°I
Max
79.1
81.9
77.8
76.7
80.9
79.1
82.9
77.4
79.5
83.1
84.6
83.1
80.9
85.0
84.5
75.7
84.8
••)
Min
62.6
58.7
57.6
54.4
54.7
60.2
63.0
58.7
56.7
65.7
60.2
60.4
57.6
64.3
62.6
58.2
63.2
Top
Day
Max
90.7
93.9
91.7
90.2
90.5
88.0
92.0
88.3
96.0
94.5
87.8
93.1
94.5
96.9
94.1
88.7
95.1
Ten Ozone
Temperature
(°F)
Min
68.6
67.9
61.9
64.1
62.2
63.8
68.1
65.1
61.0
74.1
69.0
68.6
69.2
69.9
66.5
66.2
69.0
1986-88
Ozone Design
Value (ppm)
.170
.140
.140
.160
.180
.160
.140
.130
.130
.220
.140
.130
.130
.180
.140
.130
.140
11.0
11.3
11.3
11.0
11.7
84.6 63.1
83.7
84.9
85.0
58.2
59.4
68.1
84.4 62.9
91.1
95.4
92.8
93.2
93.5
60.5
69.2
67.2
71.5
65.4
OO
.170
,130
.130
,130
,150
11.3
11.3
11.3
11.0
84.9
82.5
84.0
86.7
64.4
63.4
63.8
65.5
95.1
90.2
94.3
93.8
72.0
61.2
68.9
69.5
.200
.150
.140
.140
-------
Table 3-1 (Cont'd)
Nationwide Urban Ozone Nonattainment Area Input Data For MOBILE4.0
1986-88 Ozone
Nonattainment Area
Scranton-Wilkes-Barre,
PA
Sharon, PA
Washington, DC-MD-VA
York, PA
Atlanta, GA
Birmingham, AL
Charlotte-Gastonia-
Rock Hill, NC
Fayetteville, NC
Greensboro-Winston
Sal em-High Point, NC
Greenvi 1 le-Spartanburg ,
SC
Knoxville, TN
Lexington-Fayette, KY
Louisville, KY-IN
Memphis, TN-AR-MS
Miami, FL CMSA
Montgomery, AL
Nashville, TN
Owensboro, KY
Raleigh-Durham, NC
Tampa-St. Petersburg-
Clearwater, FL
Canton, OH
Chicago, IL CMSA
Cincinnati, OH CMSA
Cleveland, OH CMSA
Columbus, Oil
Dayton-Springfield, OH
Detroit, MI CMSA
Grand Rapids, MI
1987 Avg
In-Use
RVP (psi)
11.3
11.3
11.0
11.3
10.6
10.6
10.6
10.6
10.6
10.6
10.6
11.6
11.6
10.5
10.2
10.6
10.6
11.6
10.6
10.2
11.7
11.6
11.7
11.7
11.7
11.7
11.6
11.6
Summer Avg
Temperature
Top Ten Ozone
Day Temperature
(°F) (°F)
Max
80.9
85.3
86.5
85.8
85.8
89.5
87.4
89.2
86.4
86.8
87.1
85.2
86.0
90.3
89.0
90.1
89.0
88.7
86.7
90.1
81.1
82.8
85.4
80.1
83.5
83.3
81.3
81.5
Min
59.1
60.7
67.1
60.3
68.2
68.2
67.3
67.6
65.4
67.7
66.8
64.3
64.8
70.0
75.1
70.3
67.5
64.7
65.5
73.2
59.1
63.2
64.2
59.5
60.5
62.7
61.6
57.8
Max
92.5
91.3
95.2
94.4
94.8
94.9
94.0
94.4
93.7
93.2
97.5
92.8
91.7
94.2
88.8
97.0
95.0
96.7
95.3
89.7
94.2
94.7
94.3
92.3
95.7
93.2
93.2
94.5
Min
66.4
63.5
73.2
67.3
70.5
68.3
69.7
65.8
68.5
68.0
65.9
68.3
65.2
71.4
65.9
67.6
68.3
67.9
68.9
68.9
66.3
70.3
68.7
64.9
65.0
67.1
71.9
67.2
1986-88
Ozone Design
Value (ppm)
.130
.130
.170
.130
.170
.140
.150
.130
.150
.140
.140
.130
.170
.150
.150
.140
.140
.140
.140
.130
.140
.190
.160
.150
,130
,140
,140
,140
-------
Table 3-1 (Cont'd)
Nationwide Urban Ozone Nonattainment Area Input Data For MOBILE4.0
1986-88 Ozone
Nonattainment Area
Lafayette-
West Lafayette, IN
Milwaukee, WI CMSA
Muskegon, MI
Sheboygan, WI
Toledo, OH
Youngstown, OH-PA
Baton Rouge, LA
Beaumont-Port Arthur, TX
Dallas, TX CMSA
El Paso, TX
Houston, TX CMSA
Lake Charles, LA
Kansas City, MO-KS
St. Louis, MO-IL
Salt Lake City-
Ogden, UT
Bakersfield, CA
Fresno, CA
Los Angeles, CA CMSA
Modesto, CA
Phoenix, AZ
Sacramento, CA
San Diego, CA
San Francisco, CA CMSA
Santa Barbara-
Santa Maria-Lompoc, CA
Stockton, CA
Visalia-Tulare-
Porterville, CA
Portland, OR CMSA
1987 Avg
In-Use
RVP (psi)
11.6
11.6
11.6
11.6
11.6
11.7
10.5
9.8
9.8
9.0
9.8
10.5
9.8
10.2
9.7
8.6
8.6
8.6
8.6
8.5
8.6
8.6
8.6
8.6
8.6
8.6
10.8
Summer Avg
Temperature
<°]
Max
84.5
78.5
78.6
79.5
82.3
80.2
90.9
91.5
94.2
94.1
93.2
90.7
85.8
86.9
88.2
95.7
94.9
81.1
91.8
102.8
90.2
81.1
79.5
73.1
91.9
96.8
76.4
?)
Min
60.4
57.2
57.8
57.0
59.1
57.8
71.7
73.1
72.7
68.0
72.0
72.9
67.7
66.9
56.8
63.2
60.4
62.5
57.4
73.8
56.4
58.7
53.3
56.0
57.3
59.7
54.0
Top Ten Ozone
Day Temperature
(°F
Max
96.3
90.9
89.8
90.9
94.1
92.2
88.6
88.2
98.2
91.5
87.3
89.3
94.9
94.3
94.1
98.0
101.8
89.6
96.9
105.8
101.5
87.1
94.9
78.2
95.2
102.7
94.8
) ^-2.3
Min
69.0
67.4
72.2
63.6
63.8
63.8
65.2
65.2
73.4
56.4
61.2
65.8
71.8
73.3
61.7
66.7
67.0
64.5
60.4
81.0
62.5
59.8
60.3
51.2
58.6
69.4
60.6
1986-88
Ozone Design
Value (ppm)
.130
.180
.180
.170
.140
.130
.160
.170
.140
.170
.190
.130
.130
.160
.140
.160
.170
.330
.140
.140
.170
.180
.140
,130
.140
.150
,150
I
I—•
o
-------
3-11
turn were used to determine the total emission reductions from
the various control options.
The version of EPA's computer emissions model used for the
Draft RIA analysis was an improved version of the MOBILES
emission factor model (hereafter referred to as MOBILES.9).
Since evaporative emissions are dependent on temperature,
MOBILES.9 allowed for the use of climatic data reflecting
actual temperatures in the areas modeled. In the Draft RIA
analysis, two different types of city-specific temperatures
were used for the different analyses performed. The first set
was used for projecting the environmental impact of RVP
controls and the second set was used in assessing the economic
impacts of RVP controls.
Environmental impacts were modeled using temperatures from
the design value day between 1982 and 1984. (The design value
is defined to be the fourth highest one-hour ambient ozone
concentration occurring over three years.) EPA used the design
value day temperature in projecting emission inventories and
ozone attainment status as one way of recognizing that the days
on which high ozone levels occur are the days on which ozone
levels must "be reduced in order for an area to meet the ozone
NAAQS. Inventory projections based on the design value day
temperatures were chosen to appropriately compare and rank
various control programs.
Economic: related impacts were based on emission reductions
which used July average temperatures rather than design value
day temperatures. The results of these MOBILES.9 runs were
used to calculate both the evaporative emissions recovery
credit (the value of the fuel saved due to reduced evaporative
emissions losses with lower RVP gasoline) and the attainment
area VOC emissions reduction credit. EPA used the average
temperature approach because the economic savings of recovering
evaporative emissions and reducing attainment area emissions
occur throughout the summer, not just on high ozone days. In
particular, the average temperatures during the month of July
were used because they correspond to the in-use fuel volatility
survey, data used in this study. Based on the July average
temperatures, EPA calculated the emission factors (and then
emission inventories and emission reductions) for the
projection years under various control options. Both the
evaporative recovery credit and the attainment area VOC
emission reductions credit were used in determining the final
cost-effectiveness of RVP controls.
For city specific RVPs to use in the emission factor
model, EPA used in-use survey data which represented average
volatilities for nonalcohol-containing unleaded gasolines in
the summer months. Survey RVP levels from 1983 were used to
develop the 1983 base year inventories. The 1985 survey RVP
-------
3-12
data were used to project inventories in all of the projection
years assuming no RVP controls existed. For these 1985 values,
if the area surveyed had an RVP less than the ASTM limit for
that area, the ASTM limit was used in place of the surveyed RVP
level. This was done based on the assumption that the RVP in
the area would continue to rise until it reached the ASTM
limit. If no RVP data were available for a given nonattainment
area, the RVP of the nearest survey area was used. Also, the
most appropriate RVP value would be that for the month during
which the design value day occurred. Since the volatility
surveys were performed only in July, if the design value day
was not in July, the design value day RVP was determined by the
ASTM class limit for the month in which the design value day
occurred. However, for six areas where the July survey RVP was
higher than the July ASTM class limit for that area, the ASTM
class limit for the month in which the design value day
occurred plus the difference between the July survey RVP and
the July ASTM class RVP limit was used.
In addition to city-specific temperatures and RVPs,
city-specific alcohol-blend market shares were used in the RVP
modeling. EPA estimated that alcohol blends have an RVP that
is 1.0 psi higher than the base gasoline. The market-share
data was also used to evaluate the impact of various
alternative treatments of alcohol blends. In the modeling it
was assumed that the RVP of gasohol was 1.2 psi greater than
that of gasoline (i.e., 1.0 psi RVP effect with a 0.2 psi
average commingling effect). In those scenarios where alcohol
blends were required to meet the same RVP restrictions as
gasoline, the in-use RVP of alcohol blends was assumed to be
0.2 psi greater than gasoline, due solely to commingling.
The above city-specific inputs were incorporated into the
MOBILES.9 model which then calculated HC emission factors for
each of the urban, non-California ozone nonattainment areas
modeled. These nonattainment area emission factors were then
converted into nationwide emission factors based on a
population-weighted average of the individual emission
factors. Fleetwide exhaust, evaporative, and refueling
emission factors were calculated for each gasoline-fueled
vehicle category (LDV, LDT, and HDV).
B. Summary and Analysis of Comments
In general, comments on the evaporative and exhaust
emissions models, used for the NPRM analysis fell into the
following four categories: 1) the validity of the emissions
model correlations developed for predicting vehicle emissions,
2) the effect of weathering of in-use fuel on emissions, 3) the
representativeness of temperature, RVP and tank fill level
inputs used to calculate emission factors, and 4) the magnitude
of the effect of ethanol blends on evaporative emissions.
-------
3-13
These comments are now examined in the light of a new
emissions model developed by EPA. In the time since the NPRM
was released in August 1987, EPA has revised the MOBILES
emissions model and issued MOBILE4.0. The MOBILE4.0 emissions
model, which has been used as the basis for the environmental
impact analysis for today's rule, has incorporated many of the
comments related to modeling evaporative emissions which were
received on the NPRM. Comments on the first three of the areas
listed above are examined in the following section. The effect
of ethanol blends on emissions will be examined in Section IV
of this chapter.
1. Evaporative Emissions Model
As mentioned above, comments were received which
questioned the validity of the evaporative emissions
correlations developed in MOBILES.9. API and -Sun Refining and
Marketing Company expressed concern over the representativeness
of the data base used for emissions modeling. Texaco and API
pointed out that the UDI equations for predicting diurnal
emissions were derived from an EPA data base limited to three
RVP test conditions. SOHIO and ARCO stated that there is
insufficient data to indicate that emissions are reduced as
volatility is dropped below 9.0 psi RVP. As an alternative to
EPA's model, a number of oil and auto companies submitted, and
recommended that EPA use, a model recently developed by Radian,
Inc. under the auspices of the Coordinating Resource Council.
As previously mentioned, EPA's computer model for
calculating calendar-year, fleet-average motor vehicle emission
factors for various gaseous pollutants has been released in an
updated version, MOBILE4.0. The differences in emission
modeling calculations between EPA's.earlier version, MOBILES.9,
and the updated version, MOBILE4.0, are primarily that an
assessment of. vehicle running losses has been added, the effect
of RVP and • temperature on exhaust emissions has been
reassessed, and the form of the uncontrolled diurnal index
(UDI) and hot soak correlations has been changed with new
regression coefficients calculated. However, before explaining
these modeling changes, it is necessary to respond to the
comments submitted by API and Sun Refining and Marketing
Company which questioned the representativeness of the data
base upon which the emissions model is based.
API correctly pointed out that the data base used for
MOBILE4.0 is the emission factor (EF) vehicle testing performed
by EPA. API then claimed that testing equipment and facilities
used by EPA led to an overestimation of evaporative emissions.
For diurnal emissions, API stated that EPA's method of using an
external thermocouple to measure the diurnal temperature range
(60° to 84°F, a rise of 24°F) for the fuel tank resulted in an
actual diurnal temperature rise of between 26°F to 30°F. For
-------
3-14
hot-soak emissions, API stated that EPA's ventilation
conditions in the dynamometer rooms result in unrealistically
high fuel tank temperatures prior to the hot-soak. API claims
that these problems cause EPA to greatly overestimate
evaporative emissions from vehicles.
EPA has studied the accuracy of different methods of fuel
temperature measurement in the past. In one study, EPA
compared fuel temperatures measured with a temperature probe
placed inside a fuel tank versus a thermocouple attached
externally to the tank with an epoxy compound, the current EF
method for measuring fuel temperature. The results of the
testing showed that during a diurnal test, the differance
between the externally measured temperature and the internally
measured fuel temperature was never more than 2°F and in many
cases there was no difference between the two measurements.[3]
Therefore, EPA believes that the external thermocouple method
used for measuring fuel temperature when testing EF vehicles,
provides an accurate measurement of the diurnal test
temperatures and does not lead to an overestimation of diurnal
emissions.
EPA has also investigated the amount of cooling vehicles
receive while operated on the dynamometer versus being driven
on the road on a typical summer day. The result of one EPA
test program showed that for two of the four vehicles tested,
on the road fuel temperature increases were higher than the
fuel temperature rises on the dynamometer. For the remaining
two vehicles, on the road fuel temperature rises were lower
than or equal to dynamometer fuel temperature rises.[4] Based
on these results it appears that the amount of cooling a
vehicle receives while being driven on the road versus on the
dynamometer is highly dependent on the vehicle model. For some
vehicles the amount of cooling received on the dynamometer may
be less than would be experienced in-use, while for other
vehicles, the amount of cooling received on the dynamometer may
be actually more than would be experienced in-use. Due to
limited amount of data on vehicle cooling and the conflicting
results of current data, EPA believes that its use of the EF
data for the MOBILE4.0 emissions model is appropriate. Should
more data on vehicle cooling become available which could
improve the model, EPA will consider those results in future
revisions of the model.
Sun Refining and Marketing presented test data for a
single vehicle showing that it was the difference in the
high-end volatility (T90) of EPA's 9.0 psi and 11.5 psi RVP
test fuels, and not RVP, that caused the exhaust emissions
effect. In contrast, EPA's own comprehensive testing of a
vehicle showed that it was a difference in the amount of
hydrocarbon vapors generated in the tank during the test (which
were greater with the higher RVP fuel), that caused the exhaust
emissions effect.[5] In addition, recent running loss test
-------
3-15
results also indicate that exhaust emissions are strongly
affected by the amount of hydrocarbon vapors generated in the
tank, much more than EPA's emissions factor testing would
indicate.[6] Thus, even if part of the exhaust emissions
effect seen with the emission factor testing is due to
differences in Tgg, the overall exhaust emissions effect
estimated using EPA's emissions factor' testing likely
underestimates the exhaust emissions effect which occurs
in-use. Thus, the emission factor data base will continue to
be used.
Regarding changes to the emissions model, the issue of
temperature and RVP correction factors for exhaust emissions
has been reassessed in MOBILE4.0. Whereas the former
correction factors were based solely on 75° FTP testing, the
new MOBILE4.0 analysis contains correction factors which adjust
exhaust emissions for RVP as well as temperature, by utilizing
new emissions data.[7] In addition, based on available
emissions data for vehicles operated on fuels below 9.0 psi
RVP, EPA has determined that there is no consistent effect of
fuel volatility on exhaust emissions. Therefore, no RVP
exhaust correction factor is applied for RVPs below 9.0 psi.
Also related to exhaust emissions modeling, API provided
comments to the effect that due to the unrepresentative nature
of the FTP and the emission factor test program, EPA's exhaust
emission HC benefit due to RVP control does not apply to
typical in-use vehicle operation, and is thus significantly
overstated. Factors affecting exhaust HC emissions claimed in
support of this comment were high FTP ambient temperatures,
unrepresentative fuel tank heating, canister loading by
excessive diurnal emissions, EF testing bias, EF data
manipulation techniques, and cold temperature, low volatility
driveability problems.
In response to the first point, it is true that the soak
temperatures used by EPA in the initial emission factor testing
are higher than average summer overnight low temperatures in
Class C areas. However, the overnight low temperature is not
representative of all driving that takes place with vehicles
over a 24 hour period. In fact, due to a non-linearity of the
effect of RVP on exhaust emissions with temperatures, it is
more appropriate to err on the high side with temperature than
on the low side if one's goal is to represent the nation as a
whole. This is especially true since EPA's main concern is
high ozone days which also tend to be the days with the highest
temperature. All this aside, with EPA's reanalysis of the
issue, RVP correction factors are now determined as a function
of temperature using data collected at a variety of
temperatures. As a result, this comment should no longer be a
concern.
In response to the second point, even if the fuel tank
heating during the 20 minute FTP were higher than that
-------
3-16
typically seen during in-use operation for 20 minutes, it is
still well within the range found in-use especially on high
ozone days with high solar loading. In addition, since excess
emissions tend to increase faster with increasing temperature,
in order to represent the nation as a whole, an increase in
temperature higher than just the average is more representative
of average canister loading from excess evaporative emissions.
Also, as a final note, since the testing is done consistently
for all fuels and temperature conditions, any
unrepresentativeness in fuel tank heating would have little
effect on the RVP correction factors since they are the ratio
of the exhaust emissions under a given condition to those at
75°F with 9.0 RVP fuel. As a result, the effects of any
possible excessive fuel tank heating for the most part cancels
out of the equation.
In API's third point, they suggested that the diurnal
procedure loads the evaporative canister to an unrepresentative
degree compared to in-use diurnals just prior to exhaust
testing, causing exhaust emissions to increase during the
subsequent testing. While it is true that the initial
conditions of the canister prior to testing may significantly
affect the exhaust emissions measured during the FTP, EPA
believes many vehicles probably experience very high canister
loadings, especially on high temperature days. In any event,
although EPA disagrees with the statement that our test
procedure loads the canister unrepresentatively compared to
in-use, this is not important since for both the case of 75°F
using 9.0 RVP and the case being compared to it, the same
diurnal procedure is followed. Thus, the magnitude of the
diurnal tends to cancel out of the equation used to determine
the exhaust correction factors.
In API's fourth point they suggest that EPA was only able
to show an RVP effect through an -unjustified manipulation of
the data and testing sequence. In actuality, the RVP effect
had been masked by the test order prior to EPA's change. The
testing on low RVP fuels maintained adequate canister capacity
to minimize the effects of the subsequent testing on high RVP
fuels. By switching the order, this effect was eliminated.
However, in this case, one might argue that the measured effect
is smaller than actual, since the canister tends to be
overloaded prior to testing on the low RVP fuels, resulting in
a greater than normal amount of purge hydrocarbon from the
canister, and thus a dampening of the real effect of lowering
RVP. The fact that the RVP effect seen in the EF data is not
simply a result of the testing procedure is supported by other
studies now incorporated into the analysis which all show
similar results.[7]
In their fifth point, API stated that due to
cold-temperature low-volatility driveability problems EPA has
misrepresented the effect of volatility control on exhaust
emissions. In support of this API provided data which they
-------
3-17
claimed demonstrated the fact that the effect of RVP on exhaust
emissions is non-linear with minimum exhaust HC occurring at
approximately 8-9 psi RVP, and that as temperatures are reduced
below standard FTP temperatures, the effect of RVP to reduce
exhaust HC is minimized and in fact is reversed at realistic
Summer (55°F) and Spring/Fall (43°F) average morning lows.
In Chapter 4 of this analysis, the issue of
cold-temperature low-volatility driveability is discussed in
detail. That analysis shows that at temperatures and RVPs
substantially below those mentioned here no significant
driveability effects were encountered. Since the majority of
vehicle travel occurs at temperatures well above the average
morning lows, test temperatures of 75°F or higher may be the
most representative (especially if the non-linearity - of the
effect of volatility on exhaust emissions is taken into
account).
The data provided by API consisted of three studies by
Chevron, one by Exxon for API, two by ATL for API (one
currently in progress), and one by CARB. (The final results of
the second ATL program were not available at the time of this
analysis.) In -addition to the test program results, API
pointed out that CARB recently proposed an in-use summer
gasoline RVP standard of 8.0 psi citing similar emissions
increases at 7.0 psi as a reason for not lowering RVP below 8.0
psi.
It has long been recognized that ambient temperature has a
strong effect on exhaust emissions due mainly to its effect on
fuel distribution, fuel/air mixing, catalyst lightoff, etc.
However, the effect of RVP is a more recently discovered
effect, which results mainly from its effect on the quantity of
evaporative emissions being purged to the engine. Thus, it is
not strictly an RVP effect, but an RVP and temperature effect,
or more accurately a true vapor pressure (TVP) effect.
The CARB exhaust emissions analysis is based on four of
the five test programs which API referenced that measured the
effect of low RVP on exhaust emissions. Their analysis shows
that depending on the age of the vehicle, exhaust emissions
stay the same or decrease as RVP is reduced until around 8.0
psi (at either 80°F or 55°F), and exhaust emissions are
projected to rise slightly by about four to eight percent,
depending upon the ambient temperature, as RVP is decreased
below 8.0 psi. However, when other volatility related
emissions reductions are included, such as evaporative
emissions and stationary source emissions reductions (running
losses were not included in the CARB estimates), the overall
emissions inventory continues to decrease until an RVP of 7.2
psi in 1990 and an RVP of 7.6 psi in 2000. CARB points out
that their analysis is based on the exhaust emissions results
at FTP conditions because FTP temperatures are representative
of average summer temperatures. They did not include the
-------
3-18
effects of lower temperature on emissions in their analysis
because GARB believes that the highest ozone levels typically
occur on days with the highest temperatures.
We agree with CARB's position that high ozone days
typically occur on days with high temperatures. Therefore, at
typical high ozone day temperatures, the reductions in
evaporative and running loss emissions which result from
lowering RVP below 9.0 psi, should more than offset any small
increases in exhaust emissions which might occur. Based on
CARB's analysis, this appears to be true for the Phase II Class
B areas (of which California is one), since EPA's Phase II
Class B RVP limit is 7.8 psi, slightly higher than the 7.6 psi
value indicated in the GARB analysis' to be where the overall
inventory begins to increase as RVP is lowered.
With regard to Phase II Class A areas, -the available
emissions data are not directly applicable since the testing
was not done at temperatures which would be experienced on
typical high ozone days. However, vapor pressure related
phenomenon under Class A RVP and temperatures should be
comparable to Class B RVP and temperatures. Therefore, we
would expect that a similar trend would occur in Class A areas
as exists for Class B areas. In other words, any exhaust
emissions increase at low RVPs (which the previously mentioned
data shows would decrease as the ambient temperature rises)
should be more than offset by decreases in vapor-related
emissions, especially when running losses are included, as RVP
is lowered from the current Class A level of 9.0 psi to 7.0
psi. In addition, the benefits of RVP reduction on evaporative
emissions and running losses would be expected to be greater in
Class A areas than in Class B areas because of the higher
temperatures.
Evidence that driveability problems (presumably the cause
of any exhaust emissions increase at low RVPs) should not be a
serious concern at RVP levels below 9.0 psi is the fact that
vehicle manufacturers have supported these RVP levels in their
comments on the original volatility proposal. Again, it is an
issue of actual vapor pressure conditions in the geographical
area under consideration, not the RVP limit. Based on the
Phase II RVP limit for California, EPA agrees with CARS that
driveability and any exhaust emissions effects should not be a
serious concern with an RVP of 7.8 psi in California. At the
same time, for a Phase II Class A area, an RVP as low as 7.0
psi should not lead to any serious concerns with driveability
or exhaust emissions increases, due to the higher temperatures
which occur in Class A areas compared to Class B areas.
MOBILE4.0 treats running-loss emissions (evaporative
hydrocarbons emitted when the vehicle is in operation)
differently than MOBILES. In MOBILES, running loss emissions
were assumed to be zero. Since 1986, EPA has contracted
Automotive Testing Laboratories, Inc. (ATL), to test vehicles
-------
3-19
to define the quantity of evaporative running losses. The
results from these test programs were analyzed and have been
used for estimating running loss emissions in MOBILE4.0.[7]
With MOBILE4.0, evaporative emissions are calculated by
the same method used for the Draft RIA and the analysis
performed for the Final RIA for the first phase of RVP controls
with minor changes. These changes include the form of the
equation used for estimating diurnal emissions and the
regression coefficients calculated for the equation as well as
the equation and regression coefficients used to calculate hot
soak emissions. The diurnal index equation used in MOPILE4.0
differs from that of MOBILES.9 in that the first order
dependence of the diurnal index (DI) on UDI was found to be
insignificant and subsequently deleted leaving the DI with only
a second order dependence on UDI.[7] In addition, MOBILE4.0
hot soak emissions are adjusted by ambient temperatures as well
as by fuel RVP. The regression coefficients for each were
recalculated from a larger test data base.[7]
The Radian model at this point in its development contains
a number of major disadvantages relative to EPA's which argue
against its use at this time. First, and foremost, the model
is entirely statistical, consisting of correlations of data
versus test parameters. This prevents its use outside of the
test conditions, since no engineering model is used to
demonstrate the validity of extrapolation. Unfortunately, the
range of summer climates occurring in the U.S. is much wider
than the range of test data. Therefore, extrapolations must be
performed in order to model in-use conditions.
Second, some of Radian's statistical techniques result in
the consistent underestimation of measured emissions even at
the test conditions. This arbitrarily leads to a reduction in
the estimated effect of RVP on emissions, which is unacceptable.
Third, the most apparent advantage of the Radian model in
estimating partial diurnals may not be as strong as it
initially appears. Diurnal emissions are a function of
available canister capacity, as well as the quantity of fuel
tank vapors directed to the canister. Recent modeling by EPA
of the interaction of tank vapors, canister loadings, and purge
rates to the engine indicates that canisters are likely
saturated at the end of the day.[8] In addition, EPA's running
loss testing confirms that many vehicles' canisters are 'fully
loaded at today's RVPs on high temperature days typical of high
ozone conditions.[9] Thus, EPA's emission factor test
conditions of a complete diurnal with a purged canister are
almost certainly more accurate than Radian's partial diurnals
with a purged canister (reality being a combination of partial
and full diurnals with a saturated canister).
-------
3-20
2. Fuel Weathering
Comments submitted by Texaco, General Motors, and API
maintained that the effects of weathering were not considered
in the DRIA vehicle evaporative emission analysis. However,
EPA stated in the Draft RIA that fuel weathering was an
important factor in assessing vehicle evaporative emissions
and, in fact, the decrease in vehicle fuel RVP as a result of
the weathering phenomenon was taken into account in the diurnal
evaporative emission analysis based on MOBILES.9. The effect
of weathering and the variability of in-use fuel tank levels on
diurnal emissions were incorporated into the model by
determining a fuel tank fill level which accounted for the
effects of fuel weathering, RVP variability, average in-use
fuel tank level, as well as daily temperature variability.
Fuel weathering was not included in the hot-soak emissions
analysis for the NPRM because neither ambient temperature
effects nor the effect of temperature and RVP variability were
included .in the hot-soak model.
The MOBILE4.0 emissions model now incorporates the effect
of weathering on diurnal emissions as well as hot-soak, running
loss and exhaust emissions. The manner in which weathering has
been taken into account by MOBILE4.0 is more straightforward
than in MOBILES. 9. The MOBILE4.0 correlation for RVP loss due
to weathering, determined as a function of the fuel's initial
RVP and maximum daily temperature, was based on a fuel
weathering model developed by Radian, Inc. for Coordinating
Resource Council. For a more complete discussion of how
MOBILE4.0 incorporates fuel weathering into emissions
calculations, the reader is directed to the EPA memorandum on
MOBILE4.0.[7] It should be noted that for this analysis,
weathered RVP was allowed to drop below 7.0 psi for running
loss emissions determination. (The standard version of
MOBILE4.0 resets RVP to 7.0 psi.)
3. Emission Factor Modeling Inputs
In general, comments were received which related to the
representativeness of the data used to calculate evaporative
emission factors. Several of the areas on which comments were
received included temperature data, RVP values, in-use fuel
tank fill levels and air quality projection data. (A summary
and analysis of comments covering air quality projection data
inputs including emission source growth rates and inventory
data will be presented in Section III of this chapter on
Emission Inventories.)
Some commenters questioned the number of nonattainment
areas that should be included in the analysis. Both the Office
of Management and Budget (OMB) and Multinational Business
Services (MBS) stated that only those areas expected to be in
nonattainment in the future (which EPA projects to decline in
the mid- to late-1990s and then increase after 2000 without
further controls) should be included.
-------
3-21
Texaco objected to the use of temperatures from a single
design value day for estimating emissions. The Motor Vehicle
Manufacturers of America (MVMA) commented that there is no
clear relationship between the design value level and the July
average temperatures used. MVMA feels that better correlation
with temperature could be obtained with an ozone statistic more
robust than the design value. When MVMA repeated the analysis
to calculate inventory projections, it used 30-year daily
average minimum and maximum temperature data for July whereas
EPA used the design value day minimum and maximum temperature
data for each city. MVMA noted that in some cities this
resulted in higher temperatures, while in others it resulted in
lower temperatures.
Commenters also stated that EPA should not have increased
each area's RVP to the ASTM maximum limit if it was
historically below that limit. They stated that the fuel
distribution system was such, that lower limits in adjacent
states often kept fuel in a given state below its maximum level.
In the reanalysis which MVMA had done, they used more
city-specific data than EPA, which they say is preferable.
Whereas EPA used a 1983 implementation date for I/M programs
for all cities, MVMA used city-specific implementation dates.
MVMA also accounted for Stage II controls being implemented in
New York, New Jersey and St. Louis, and used city-specific
model year start dates and vehicle classes.
The choice of 1983 as the base year for the analysis was
thought by some to introduce significant error in the results.
Sun- Oil stated that the summer of 1983 was unseasonably warm
and had an unusually large number of ozone exceedances. It
felt that EPA must develop statistical attainment data using
normal temperature information determined over an extended
period of time (15-30 years). MVMA also commented that EPA
selected an unrepresentative base year design level on which
the air quality calculations were based.
MBS noted that their analysis used a 30-year model of the
fleet, instead of a 20-year model like EPA used. They feel
that this is appropriate because vehicles more than 20 years
old contribute significantly to total fuel consumption.
Chrysler stated that MVMA statistics show that fleet turnover
is slower than that used in the NPRM (i.e., it takes 15 years,
they said, not 13 years, for 90 percent of cars and trucks to
be replaced).
For the second phase of RVP controls, EPA has updated and
improved much of our analysis, which in turn addresses these
comments. For example, the most recent three-year period for
which design values exist at the time of this analysis is 1986
to 1988. Therefore, the nonattainment analysis is based on
these three years and EPA believes that the middle year, 1987,
is the most appropriate year for the base year.
-------
3-22
In regard to temperature inputs, MVMA is correct in noting
that there is no clear relationship between the ozone design
value and July average temperatures. This is not surprising
considering there are many other city-specific factors that
could impact the magnitude of the design value, such as VOC
emissions and VOC/NOx ratios. Therefore, for determining
emissions inventories for future attainment projections, EPA
has based its analysis on the days with the ten highest ozone
exceedances in each nonattainment area, or all of the days with
exceedances if there were fewer than ten exceedances for a
given area. (In some cities where there were two or more days
which had the tenth highest ozone exceedances, all of those
days were included in the analysis. So in some cases there
were cities where more than ten days were included in the
temperature analysis.) Based on actual temperature data for
these days, the minimum and maximum temperatures on each of the
days was determined. A straight average of these temperatures
for both the minimum and maximum temperatures was calculated
and used for the city specific temperature inputs. For
determining emissions recovery credits and attainment area
credits, where the average ozone season emission level is
important, EPA now' uses the average of the 30-year average
daily minimum and maximum temperatures for June, July and
August. The temperatures used for both of these analyses are
contained in Table 3-1.
In regards to city-specific RVP inputs, EPA no longer
assumes each city's RVP is at the ASTM limit. Instead EPA has
used 1987 MVMA summer fuel survey results to determine actual
in-use RVP levels prior to RVP controls. In those cities for
which no RVP survey results are available, the RVP from the
nearest city for which survey results existed (either
geographically or based on likely fuel distribution patterns)
was used. After implementation of the volatility regulations,
we assumed that RVP levels dropped to those specified in the
first phase of RVP controls and then to those RVP levels
specified in today's rule. The city-specific RVPs used in this
analysis are contained in Table 3-1.
Regarding other city-specific inputs, EPA has been able to
account for other city-specific inputs with the MOBILE4.0
emissions model. In addition to temperature and RVP, EPA has
incorporated city-specific I/M programs, city-specific Stage II
programs, as well as city-specific growth rates into this
analysis. The result of this approach is expected to increase
the accuracy of estimating the impact of RVP control on a
nationwide basis.
Regarding the comments on nonattainment areas it has been
Agency policy to include the most recent nonattainment areas in
air quality analyses. Analyses based on future projections of
attainment or nonattainment would be very dependent upon the
assumptions used in making the projections, which could lead to
-------
3-23
significant policy problems. The potential for such problems
can be seen in the large number of ozone nonattainment areas
which occurred in 1988 due to a hotter than normal summer, For
this reason, EPA will base this analysis on the nonattainment
situation during 1986 to 1988, the most recent data available
at the time of the analysis.
Finally, while it is true that a 30-year model of the
vehicle fleet is likely to be more accurate than a 20-year
model, the difference would not be expected to be substantial.
EPA's model accounts for all model-years more than 19 years old
on an aggregate basis. This is reasonable since such vehicles
account for less than one percent of vehicle-miles traveled
(VMT). In addition, since Chrysler did not include information
about the model years involved in the MVMA study showing 90
percent of vehicles being replaced after 15 years", it is not
possible to resolve the difference from EPA's estimate.
However, if for example the MVMA estimate involved different
model years than EPA's, that might explain the two year
difference since vehicle sales vary from year to year due to
economic influences.
C. Emission Factor Results
Tables 3-2 and 3-3 contain the MOBILE4.0 non-Northeast
(labeled "Non-NESCAUM") ozone nonattainment area emission
factors based on the top 10 ozone day temperatures and the
summer average temperatures, respectively. Each table contains
an overall vehicle fleet emission factor, as well as the
emission factors for each vehicle class (i.e., light-duty
gasoline vehicles, light-duty gasoline trucks, heavy-duty
gasoline vehicles and diesel vehicles). In addition, each
table contains total NMHC emission factors as well as the
individual emission components including exhaust NMHC,
evaporative NMHC, refueling NMHC and running loss NMHC. As
described earlier, the emission factors from Table 3-2 (based
on top 10 ozone temperatures) were used in projecting the
environmental impact of RVP control. The emission factors in
Table 3-3 (based on summer average temperatures) were used in
determining the economic credits due to recovered evaporative
emissions and attainment areas emission reductions (see Chapter
5).
III. Emission Inventories
A. Synopsis of NPRM
Using the calculated nationwide emission factors, the NPRM
inventory projections of future VOC emissions were
estimated. Non-California, urban, nonattainment area
inventories were calculated for both mobile and stationary
sources, both as a whole and subdivided into several individual
source categories using assumptions concerning growth rates and
future technological improvements.
-------
LDGV
Table 3-2
MOBILE4 VMT Weighted Emission Factors (g/mi)
Phase II Volatility Option Evaluation
Temperature Profiles: Top 10
Areas Analyzed: Non-NESCAUM
LDGT
HDGV
Base
Phase I
9.5 psi
9.0 psi
P II+0.5
Phase II
Base
Phase I
9.5 psi
9.0 psi
P II+0.5
Phase II
Base
Phase I
9.5 psi
9.0 psi
P I 1+0.5
Phase II
5
5
5
5
5
5
2
2
2
2
2
2
1987
.784
.784
. 784
.784
.784
.784
.362
.362
.362
.362
.362
.362
.541
.541
.541
.54 1
.541
.54 1
4
3
3
3
3
3
1
1
1
1
1
1
1
0
0
0
0
0
1990
502
961
961
961
961
961
748
701
701
701
701
701
1 14
970
970
970
970
970
3
2
2
2
2
1
1
1
0
0
0
0
0
0
0
0
0
0
1995
.230
.735
.425
. 222
. 107
.966
.077
.029
.993
.990
.989
.988
. 729
.618
.553
.465
.424
.381
2
2
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
2000
.690
.215
.914
.733
.626
.501
.801
.753
.718
.713
.712
.712
.554
.458
.403
.326
.288
. 250
2005
2.
2.
1 .
1 .
1 .
1 .
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
61 1
148
856
678
569
447
776
728
693
689
688
688
51 1
421
370
296
259
222
2
2
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
2010
.588
. 130
.842
.665
.554
.432
•
.768
.721
.687
.683
.681
.681
.500
.411
.361
.289
.251
.215
6
6
6
6
6
6
3
3
3
3
3
3
1
1
1
1
1
1
1987
.016
.016
.016
.016
.016
.016
.291
.291
.291
.291
.291
.291
.288
.288
.288
.288
.288
.288
1990 1995
Total
4.483 3.053
4. 156 2.776
4. 156 2.635
4. 156 2.468
4. 156 2.352
4. 156 2.242
Exhaust
2.432 1.540
2.368 1 .470
2.368 1 .425
2.368 1.418
2.368 1.415
2.368 1.414
Evaporat
0.942 0.647
0.832 0.563
0.832 0.525
0.832 0.442
0.832 0.391
0.832 0.352
2000
NMHC
2.455
2.201
2.072
1 .927
1 .824
1 .729
NMHC
. 190
. 1 18
.072
.064
.062
.061
2005
2.
2.
1 .
1 .
1 .
1 .
1 .
1 .
1 .
0.
0.
0.
357
1 12
989
847
743
650
1 18
048
004
997
994
993
2
2
1
1
1
1
1
1
0
0
0
0
2010
.337
.096
.976
.834
.729
.638
. 109
.040
.996
.989
.986
.985
13
13
13
13
13
13
5
5
5
5
5
5
1987
.017
.017
.017
.017
.017
.017
.530
.530
.530
.530
.530
.530
9
8
8
a
a
8
3
3
3
3
3
3
1990
.335
.605
.605
.605
.605
.605
.645
.618
.618
.618
.618
.618
1995
6.
6.
6.
5.
5.
4.
2.
2.
2.
2.
2.
2.
988
390
088
61 1
317
967
508
468
440
437
436
436
6
5
5
4
4
4
2
2
2
2
2
2
2000
.231
.672
.398
.949
.663
.327
. 143
.098
.067
.062
.062
.061
5
5
5
4
4
4
2
1
1
1
1
1
2005
.871
.335
.077
.644
.363
.036
.017
.970
.937
.932
.931
.931
2010
5.855
5.324
5.069
4.637
4.352
4.025
2.011
.964
.932
.927
.926
.926
ive NMHC
0.492
0.420
0.389
0.318
0.274
0.240
0.
0.
0.
0.
0.
0.
474
406
377
309
264
233
0
0
0
0
0
0
.466
.399
.372
.304
.260
.228
4
4
4
4
4
4
.403
.403
.403
.403
.403
.403
2
2
2
2
2
2
.905
.581
.581
.581
.581
.581
•
884
635
494
282
171
078
1
1
1
1
0
0
.558
.330
.205
.018
.919
.838
1
1
1
0
0
0
.366
. 150
.032
.860
.770
.695
1 .364
1 . 149
1 .032
0.860
0.768
0.694
Refuel ing NMHC
Base
Phase I
9.5 psi
9.0 psi
P U + 0.5
Phase II
Base
Phase 1
9.5 psi
9.0 psi
P 11+0.5
Phase I I
0
0
0
0
0
0
1
1
1
1
1
1
. 243
.243
.243
. 243
. 243
. 243
.fa3b
.63B
.638
.638
.638
.638
0
0
0
0
0
0
2 17
207
207
207
207
207
422
OB3
083
083
OB3
083
0
0
0
0
0
0
1
0
0
0
0
0
. 195
. 186
. 178
. 170
. 166
. 158
. L'2b
.902
.700
.596
.528
.439
0
0
0
0
0
0
1
0
0
0
0
0
. 188
. 180
. 172
. 165
. 160
. 152
. 146
.824
.621
.528
.465
.387
0.
0.
0.
0.
0.
0.
, _
0 .
0.
0.
0.
0.
188
179
172
164
160
152
137
820
621
528
463
385
0
0
0
0
0
0
1
0
0
0
0
0
. 188
. 180
. 173
. 164
. 160
. 152
. 132
.818
.622
.528
.462
.384
0
0
0
0
0
0
1
1
1
1
1
1
. 287
. 287
.287
. 287
.287
. 287
. 150
. 150
. 150
. 150
. 150
. 150
0.263 0.248
0.252 0.238
0.252 0.229
0.252 0.219
0.252 0.212
0.252 0.201
Runn i ng
0.846 0.618
0.704 0.506
0.704 0.456
0.704 0.389
0.704 0.334
0.704 0.274
0. 243
0.234
0.226
0.216
0.208
0. 198
Loss
0.530
0.430
0.386
0.328
0.280
0.230
0.
0.
0.
0.
0.
0.
243
234
226
216
208
198
0
0
0
0
0
0
. 244
.234
.226
.216
.209
. 198
0
0
6
0
0
0
.547
.547
.547
.547
.547
.547
0
0
0
0
0
0
.513
.490
.490
.490
.490
.490
0.
0.
0.
0.
0.
0.
479
457
439
419
407
386
0
0
0
0
0
0
.454
.434
.418
.399
.386
.367
0
0
0
0
0
0
.443
.423
.407
.389
.377
.358
0.440
0.420
0.404
0.386
0.374
0.355
NMHC
0.
0.
0.
0.
0.
0.
521
424
382
326
277
227
0
0
0
0
0
0
.518
.422
.382
.325
.275
.226
2
2
2
2
2
2
.536
.536
.536
.536
.536
.536
2
1
1
1
1
1
. 27 1
.916
.916
.916
.916
.916
2.
1 18
830
7 14
473
303
067
2
1
1
1
1
1
.076
.810
. 709
.470
. 295
.061
2
.044
.792
.701
.463
. 285
.052
2.041
1 . 791
1 .701
1 .463
1 .283
1 .050
I
NJ
-C-
-------
Base
Phase I
9.5 psi
9.0 psi
P I 1+0.5
Phase II
Base
Phase I
9.5 psi
9.0 psi
P 11+0.5
Phase II
Table 3-2 (Cont.)
MOBILE4 VMT Weighted Emission Factors (g/m1)
Phase II Volatility Option Evaluation
Temperature Profiles: Top 10
Areas Analyzed: Non-NESCAUM
DSLV
1987 1990 1995 2000 2005 2010
A I I Veh
1987 1990 1995 2000 2005 2010
Total NMHC
2.697 2.250 1.703 1.496 1.452 1.448 5.845 4.495 3.180 2.628 2.534 2.509
5.845 4.016 2.757 2.233 2.155 2.138
5.845 4.016 2.499 1.991 1.924 1.914
5.845 4.016 2.310 1.825 1.763 1.754
5.845 4.016 2.198 1.722 1.660 1.650
5.845 4.016 2.068 1.609 1.550 1.540
2.697 2.250 1.703 1.496 1.452 1.448
2.697 2.250 1.703 1.496 1.452 1.448
2.697 2.250 1.703 1.496 1.452 1.448
2.697 2.250 1.703 1.496 1.452 1.448
2.697 2.250 1.703 1.496 1.452 1.448
Exhaust NMHC
0.572 0.566 0.442 0.443 0.468 0.485 2.520 .866 1.164 0.887 0.855 0.852
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
2.520
2.520
2.520
2.520
.818 1.114 0.837 0.806 0.803
.818 1.079 0.802 0.772 0.77O
.818 1.075 0.797 0.768 0.765
.818 1.073 0.796 0.766 0.764
2.520 1.818 1.073 0.795 0.765 0.763
OJ
I
r-o
Evaporative NMHC
Base
Phase I
9.5 psi
9.0 psi
P 11+0.5
Phase 11
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000.0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
.491 1.067 0.693 0.522 0.481 0.470
.491 0.932 0.591 0.435 0.400 0.391
.491 0.932 0.534 0.388 0.356 0.349
.491 0.932 0.450 0.316 0.288 0.281
.491 0.932 0.407 0.278 0.251 0.244
.491 0,932 0.367 0.242 0.217 0.211
Refuel ing NMHC
Base
Phase 1
9.5 psi
9.0 psi
P 1^0.5
Phase II
Babe
Phase 1
9.5 pbi
9.0 ps i
P 11+0.5
Phase II
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.248 0.222 0.200 0.192 0
0. 248 0.212 0.19
0.248 0.212 0.184 0
0.248 0.212 0.175 0.169 0
0. 248 0.212 0.171 0.164 0
191 0.191
0. 184 0. 182 0. 183
176 0.176 0.176
168 0.168
163 0.163
0.248 0.212 0.162 0.156 0.155 0.155
Running Loss NMHC
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
1.503 1.265 1.049 0.951 0.929 0.916
.503 0.979 0.786 0.701 0.688 0.681
.503 0.979 0.629 0.549 0.542 0.539
.503 0.979 0.536 0.467 0.461 0.459
.503 0.979 0.473 0.409 0.402 0.399
.503 0.979 0.392 0.340 0.334 0.331
-------
Table 3-3
MOBILE4 VMT Weighted Emission Factors (g/mi)
Phase II Volatility Option Evaluation
Temperature Profiles: Summer
Areas Analyzed: Non-NESCAUM
LDGV LDGT HDGV
1987 1990 1995 2000 2005 2010 1987 1990 1995 2000 2005 2010 1987 1990 1995 2000 2005 2010
Total NMHC
Base
Phase I
9.5 psi
9.0 psi
P II+0.5
Phase 1 I
4
4
4
4
4
4
509
509
509
509
509
509
3
3
3
3
3
3
49
1
1
1
1
1
5 2.488 2.083 2.032 2.019 5.215 3.883 2.626 2.102 2.014
2.143 1.752 1.707 1.697 5.215 3.620 2.404 1.900 1.B17
1.989 1.605 1.563 1.554 5.215 3.620 2.313 1.817 1.738
1.823 1.449 1.409 1.400 5.215 3.620 2.168 1.687 1.611
1.748 1.380 1.338 1.328 5.215 3.620 2.087 1.617 1.540
1.645 1.284 1.243 1.233 5.215 3.620 2.001 1.542 1.467
.998
.804
.727
.601
.529
.456
1 1 .383
1 1 .383
1 1 .383
1 1 .383
1 1 .383
1 1 .383
8.030
7 .382
7.382
7 .382
7.382
7.382
5.924
5.373
5.131
4.767
4.518
4.268
5.248 4
4.726 4
4.501 4
4. 159 3
3.921 3
3.684 3
938 4.928
434 4.428
218 4.215
889 3.885
658 3.651
429 3.422
Exhaust NMHC
Base
Phase I
9.5 psi
9.0 psi
P II+0.5
Phase II
Base
Phase I
9.5 psi
9.0 psi
P II+O.b
Phase 11
Base
Phase I
9.5 psi
9.0 psi
P II + 0.5
Phase I I
Base
Phase I
9.5 psi
9.0 psi
P I 1 + 0.5
Phase II
2. 225
2. 225
2. 225
2. 225
2. 225
2.225
.647 1.023 0.780 0.756 0.748
.609 0.981 0.737 0.713 0.706
.609 0.955 0.710 0.686 0.679
.609 0.937 0.691 0.668 0.661
.609 0.943 0.697 0.673 0.666
.609 0.935 0.690 0.666 0.659
.010 0.734 0.493 0.38B 0.367 0.362
.010 0.659 0.433 0.334 0.314 0.309
.010 0.659 0.395 0.299 0.281 0.277
.010 0.659 0.335 0.245 0.227 0.223
.010 0.659 0.297 0.209 0.193 0.188
.010 0.659 0.266 0.180 0.164 0.160
0.243 0.217 0.195 0.188 0.188 0.188
0.243 0.207 0.186 0.180 0.179 0.180
0.243 0.207 0.178 0.172 0.172 0.173
0.243 0.207 0.170 0.165 0.164 0.164
0.243 0.207 0.166 0.160 0.160 0.160
0.243 0.207 0.158 0.152 0.152 0.152
.030 0.897 0.777 0.726 0.722 0.720
.030 G.b36 O.b42 0.501 O.bOl 0.501
.030 0.636 0.461 0.424 0.425 0.426
.030 0.636 0.381 0.349 0.350 0.350
.030 0.636 0.342 0.314 0.313 0.313
1.030 0.636 0.285 0.262 0.262 0.262
3.121 2.314
3.121 2.261
3.121 2.261
3.121 2.261
3.121 2.261
3.121 2.261
1.480 1.157 1.088 1.079
.418 1.092 1.025 1.017
.386 1.058 0.992 0.984
.359 1.030 0.964 0.957
.365 1.036 0.969 0.962
.355 1.026 0.960 0.952
Evaporative NMHC
.070 0.764 0.504 0.365 0.350 0.344
.070 0.696 0.453 0.323 0.310 0.305
.070 0.696 0.428 0.304 0.293 0.289
.070 0.696 0.366 0.252 0.243 0.239
.070 0.696 0.320 0.212 0.203 0.199
.070 0.696 0.288 0.186 0.178 0.175
Refuel ing NMHC
5.272 3.477 2.419 2.083 1.974
5.272 3.450 2.380 2.038 1.926
5.272 3.450 2.354 2.010 1.897
5.272 3.450 2.339 1.992 1.878
5.272 3.450 2.345 1.998 1.885
5.272 3.450 2.338 1.990 1.877
.969
. 922
.893
.874
.880
.872
3.886 2.529 1.608 1.314 1.142 1.141
3.886 2.292 1.432 1.156 0.993 0.992
3.886 2.292 1.327 1.063 0.906 0.906
3.886 2.292 1.152 0.910 0.765 0.766
3.886 2.292 1.037 0.809 0.673 0.672
3.886 2.292 0.952 0.736 0.607 0.606
0.287 0.263 0.248 0.243 0.243 0.244 0.547 0.513 0.479 0.454 0.443 0.440
0.287 0.252
0.287 0.252 0
0.287 0.252 0
0.238 0.234 0.234 0.234
229 0.226 0.226 0.226
219 0.216 0.216 0.216
0.287 0.252 0.212 0.208 0.208 0.209
0.287 0.252 0.201 0.198 0.198 0.198
Running Loss NMHC
0.737 0.541 0.394 0.337 0.332 0.331
0.737 0.412 0.296 0.251 0.249 0.248
0.737 0.412 0.270 0.229 0.228 0.228
0.737 0.412 0.224 0.190 0.189 0.189
0.737 0.412 0.191 0.161 0.159 0.159
0.737 0.412 0.157 0.132 0.131 0.131
0.547 0.490 0.457 0.434 0.423 0.420
0.547 0.490 0.439 0.418 0.407 0.404
0.547 0.490 0.419 0.399 0.389 0.386
0.547 0.490 0.407 0.386 0.377 0.374
0.547 0.490 0.386 0.367 0.358 0.355
.677 1.510 1.418 1.396 1.379 1.379
.677 1.149 1.104 1.098 1.092 1.094
.677 1.149 1.010 1.010 1.008 1.011
.677 1.149 0.856 0.858 0.857 0.860
.677 1.149 0.729 0.727 0.724 0.725
.677 1.149 0.592 0.590 0.588 0.588
I
1X3
-------
Table 3-3 (Cont.)
MOBILE4 VMT Weighted Emission Factors (g/m1)
Phase II Volatility Option Evaluation
Temperature Profiles: Summer
Areas Analyzed: Non-NESCAUM
DSLV
1987 1990 1995 2000 2005 2010
Al1 Veh
1987 1990 1995 2000 2005 2010
Base
Phase I
9.5 psi
9.0 psi
P I 1*0.5
Phase II
2.697 2.250 1.703
2.697 2.250 1.703
2.697 2.250 1.703
2.697 2.250 1.703
2.697 2.250 1.703
2.697 2.250 1.703
.496
.496
.496
.496
.496
.496
.452
.452
.452
.452
.452
.452
.448 4
.448 4
.448 4
.448 4
.448 4
.448 4
Total NMHC
.448 4.699 3.605 2.540 2.111 2.046 2.033
4.699 3.256 2.236 1.828 1.772 1.763
4.699 3.256 2.102 1.704 1.652 1.646
4.699 3.256 1.946 1.561 1.512 1.507
4.699 3.256 1.870 1.492 1.443 1.437
4.699 3.256 1.774 1.405 1.358 1.353
Exhaust NMHC
Base
Phase I
9.5 ps i
9.0 ps i
P I 1+0.5
Phase II
Base
Phase I
9.5 psi
9.0 psi
P 11+0.5
Phase 1 1
Base
Phase 1
9.5 ps i
9.0 psi
P I HO. 5
Phase I I
BdSU
Phase I
9.5 ps i
9.0 ps i
P I 1 + 0.5
Phase 1 I
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
0.572 0.566 0.442 0.443 0.468 0.485
2
2
2
2
2
2
381
381
381
381
381
381
1 . 765
1 .726
1 . 726
1 .726
1 .726
1 .726
.111
.068
.041
.023
.028
.021
0
0
0
0
0
0
.864
.819
.792
.773
.779
.771
0.
0.
0.
0.
0.
0.
834
790
763
744
749
742
0
0
0
0
0
0
831
787
761
742
747
740
Evaporative NMHC
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
1.037 0.743 0.489 0.375 0.351 0.345
1.037 0.670 0.432 0.325 0.303 0.298
1.037 0.670 0.397 0.295 0.275 0.271
0.000 0.000 0.000 0.000 0.000 0.000 1.037 0.670 0.338 0.243 0.225 0.221
0.000 0.000 0.000 0.000 0.000 0.000 1.037 0.670 0.299 0.207 0.190 0.186
0.000 0.000 0.000 0.000 0.000 0.000 1.037 0.670 0.269 0.180 0.164 0.160
Refuel ing NMHC
0.000 0.000 0.000 0.000 0.000 0.000 0.248 0.222 0.200 0.192 0.191 0.191
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.248 0.212 0.191 0.184 0.182 0.183
0.248 0.212 0.184 0.176 0.176 0.176
0.248 0.212 0.17S 0.169 0.168 0.168
0.248 0.
0.248 0.
.212 0.171 0.164 0.163 0.163
.212 0.162 0.156 0.155 0.155
Running Loss .NMHC
0.950 0.801 0.666 0.604 0.592 0.585
0.950 0.575 0.471 0.424 0.418 0.415
0.950 0.575 0.406 0.364 0.360 0.358
0.000 0.000 0.000 0.000 0.000 0.000 0.950 0.575 0.336 0.300 0.297 0.295
0.000 0.000 0.000 0.000 0.000 0.000 0.950 0.575 0.299 0.267 0.263 0.261
0.000 0.000 0.000 0.000 0.000 0.000 0.950 0.575 0.248 0.222 0.219 0.217
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 O.OOQ 0.000 0.000 0.000
U)
I
NJ
—I
-------
3-28
These inventory projections were used to generate
projected emission reductions associated with each control
strategy which in turn were used in ranking the strategies.
The nonattainment area inventories were used as input for the
ozone air quality modeling and for estimating the cost
effectiveness of RVP controls. In modeling the environmental
impacts of RVP control, the analysis used emission factors
based on the ozone design value day conditions. Nationwide
inventory projections for the calculation of economic credits
due to RVP control were made using emission factors based on
July average temperatures for the nonattainment areas.
B. Summary and Analysis of Comments
MVMA objected to some of the adjustments made to the NEDS
inventory for the base year 1983. They felt that the
adjustments tended to decrease the contribution of all
stationary sources to total NMHC emissions, and thus to
increase the apparent importance of motor vehicles. MVMA felt
that this could not be validated with the current knowledge of
emissions of stationary sources. They also disagreed with the
resulting trends in the inventory.
While it may be true that improvements could be made to
the estimation of stationary source emissions, such
improvements would not substantively change the conclusions of
this- analysis. This is because the evaluation of the cost
effectiveness of a given program is performed using absolute
reductions, and thus the estimation of stationary sources is
not central to this analysis. For more information on the
stationary source inventory, the reader is directed to the EPA
report which documents the air quality analysis.[10]
MVMA commented that the growth rates used in the DRIA were
not city-specific, yet should have been, since much of the
other air quality projection data used was city-specific. They
also suggested that growth rates should also be city-specific
since they vary widely from city to city. Another commenter,
J.G. Bathe, stated that EPA's growth rates are too high, and
lead to overestimated emissions.
EPA has recently developed improved growth rate
assumptions. This work includes consideration of city-specific
aspects and this analysis employs the city-specific growth
rates in projecting future inventories.[10]
C. Emission Inventory Results
This section presents the results of the emissions
inventory projections. Table 3-4 presents the 1987 base year
non-Northeast nationwide nonattainment area VOC inventory used
in this analysis. Table 3-5 shows the total VOC inventory
reduction and the percent reduction of total VOC, from the 1987
-------
3-29
Table 3-4
Base Year (1987) Non-Northeast Nationwide
Urban Ozone Nonattainment Area VOC Emissions Inventory
Source Category
Light-Duty Gasoline Vehicles (LDGV)
Light-Duty Gasoline Trucks (LDGT)
Heavy-Duty Gasoline-Fueled Vehicles (HDGV)
Diesel-Powered Vehicles (DSLV)
Area Sources
Point Sources
Mobile Source Subtotal
Stationary Source Subtotal
Gasoline-related Subtotal**
Non-gasoline-related Subtotal * * *
TOTAL EMISSIONS
VOC Emissions*
(1000 tons)
3743.3
912.6
215.1
93. 7
4427.7
1183.4
4964 . 7
5611.1
4871.0
5704.8
10575.8
* Top 10 ozone day emissions multiplied by 365.
** Gasoline-related = LDGV + LDGT + HDGV.
*** Non-gasoline-related = DSLV + Area + Point.
-------
3-30
Table 3-5
Non-Northeast Urban Ozone Nonattainment Areas
Total VOC Emissions Reductions
(xlOOO tons)
Control Scenarios Year
1995 2000 2005 2010
Base to Phase I 446.9 462.9 488.1 520.6
(4.2)* (4.4) (4.6) (4.9)
Phase I to Phase II** 712.5 713.1 759.6 816.7
(6.7) (6.7) (7.2) (7.7)
Non-Northeast Nationwide
Total VOC Emissions Reductions***
(xlOOO tons)
Control Scenario Year
1995 2000 2005 2010
Base to Phase I 826.0 855.5 902.1 962.2
(4.2)* (4.4) (4.6) (4.9)
Phase I to Phase II 1316.9 1318.0 1403.9 1509.4
(6.7) (6.7) (7.2) (7.7)
* Numbers in parentheses are percent reductions of 1987
non-Northeast base year VOC inventory.
** Based on equivalent emissions map, top 10 ozone day
reductions multiplied by 365 for comparison to year round
VOC control programs which include control of non-summer
emissions).
*** Non-Northeast Nationwide emissions were determined from
non-Northeast nonattainment area emissions by assuming
that non-Northeast nonattainment area emissions are
proportional to the fraction of non-Northeast nationwide
VMT in the non-Northeast nonattainment areas (determined
to be 0.541).
-------
3-31
base year, for the first phase of RVP controls as well as the
second phase of RVP controls for both the non-Northeast
nonattainment areas alone and for all non-Northeast areas
nationwide. (The Phase II numbers are based, on the equivalent
emissions map described in Chapter 2.) Table 3-6 breaks down
the VOC inventories for the years 1995 and' 2010 by source
assuming the equivalent emissions map for Phase II.
The results presented in the tables mentioned above were
derived from the daily emission rates (in tons/day) multiplied
by 365 days per year to obtain an equivalent annual emission
reduction. This was done for comparison to other VOC control
programs which count emission reductions outside the summer
period. These non-summer VOC emission reductions are not as
valuable in reducing ozone levels which occur primarily during
the summer, when RVP controls will be in effect.
The results presented in these tables do not contain
emission reductions in the Northeastern section of the
country. As mentioned previously, eight states (Maine, New
Hampshire, Vermont, Massachusetts, Connecticut, Rhode Island,
New York and New Jersey) have begun controlling summertime RVP
to 9.0 psi or are affected by states which do. Therefore these
states and the nonattainment areas within them have been
omitted from this analysis. In order to estimate the
non-Northeast nationwide emissions, inventory from the
non-Northeast nonattainment area emissions inventory emissions
were assumed to be proportional to vehicle miles travelled
(VMT). Therefore, the non-Northeast nonattainment emissions
inventory divided by the non-Northeast VMT fraction of
non-Northeast nationwide VMT (determined to 0.541 based on
Federal Highway Administration estimates of VMT by state)
yields the non-Northeast nationwide emissions inventory results.
IV. Ozone Modeling
A. Air Quality Projections
1. Synopsis of NPRM
The ozone air quality analysis, which predicted future
ozone concentrations, was done using the EKMA computer models.
The VOC emission inventories were used as input to predict the
future ozone concentrations for the non-California urban ozone
nonattainment areas. The model is primarily a nationwide
model. City-specific information was used only as input for
the base year ozone concentration and for the ratios of NMHC to
NOx. Meteorological conditions for the EKMA model are based on
data from one of three cities: 1) Los Angeles - for modeling
California coastal cities, 2) Denver - for modeling cities in
Arizona, Colorado, Nevada, New Mexico, and Utah, and 3) St.
Louis - for modeling all other areas.
-------
3-32
Table 3-6
1995 Non-Northeast Urban Ozone Nonattainment
Area VOC Emissions Inventory
(xlOOO tons)
Emission Source
Base
Control Scenario
Phase I
Phase II
Exhaust
Evaporative
Refueling
Running Loss
Total Mobile Source*
Total Stationary Source
Total Inventory
1177.5
701 .0
202.3
1061 . 1
1119 .2
593 .8
191.9
789. 7
3216.8 2769.9
5412.1 5412.1
8628.9 8182.0
1067.5
365. 1
161 .2
390.0
2057.4
5412. 1
7469.5
The total of evaporative, exhaust, refueling, and running
loss emisisons do not add up exactly to the mobile' source
total due to rounding off during calculations.
-------
3-33
2. Summary and Analysis of Comments
Both MVMA and General Motors (GM) commented on what they
believe is an overly simplistic methodology used in EKMA. MVMA
stated that their analysis showed that even if the ozone design
level for two areas varied little, the per capita data, as well
as the total VOC inventory emissions, could be very different.
They also stated that EPA's EKMA calculations may show too
large of a change in ozone design levels with relatively small
reductions in VOC emissions (as with the proposed volatility
control program). Since EKMA is actually a nationwide model,
GM feels that the air quality projections for individual cities
were oversimplified. They did, however, request that EPA
present the results of each city separately. They commented
that by using such a simplistic modeling approach, EPA has
unnecessarily introduced considerable uncertainty into the air
quality predictions. MVMA felt that by using nationwide
averages in the model, EKMA projections should only be used in
a relative sense, not.an absolute sense. They stated that the
EKMA model was developed in order to quantify the current
understanding of ozone chemistry. No model on ozone formation
has been universally accepted yet. Therefore, MVMA argued, the
EKMA model should be used only as guide to the development of
control strategies in a relative manner and not an absolute way.
The fact that the design values for two areas could be
similar while VOC emissions are very different- is not
surprising. Ozone is a complex pollutant, and is dependent on
many other variables such as NOx emissions. Concerning the
appropriate role of the model, it is agreed that because of . the
simplified nature of this approach, it should not be used in. an
absolute, or city-specific, sense. Rather, its role is to
translate VOC reductions into projected ozone impacts using the
limited available data. For this reason, it is emission
reductions and not ozone levels that EPA uses as the basis for
cost-effectiveness calculations.
As was noted, no model has been universally accepted.
However, it has been recognized that EKMA is a reasonably
accurate method of relating VOC emissions to estimated ozone
impacts. Thus, both the use of EKMA and the role it plays in
EPA analyses are appropriate at this time.
3. Final Analysis
For the air -quality projections contained in this
analysis, a slightly different method was used to determine
nonattainment status. Based on an analysis performed by E.H.
Pechan & Associates for EPA, estimates of the percent reduction
in the base year (1987) inventory which would be required for
each nonattainment area to reach attainment were made. Using
those estimates, the reductions which would be obtained by the
first phase and second phase of RVP controls (along with the
-------
3-34
reductions due to the current Federal motor vehicle control
program) were compared to these percent reduction
requirements. If the reductions from the above programs were
equal to or greater than the reductions required to bring an
area into attainment, that area was projected to be in
attainment of the ozone standard. Further details are
contained in the EPA report documenting the air quality
analysis.[10]
The projected number of non-Northeast urban ozone
nonattainment areas is presented in Table 3-7. The results
project that the second phase of RVP controls will
significantly reduce the number of areas projected to be in
nonattainment in the future.
B. Butane and Oxygenate Reactivity and Oxygenated Blend
Environmental Impacts
1. Synopsis of NPRM
The photochemical reactivity of butane and some oxygenates
(i.e., ethanol and methanol) was discussed in the NPRM.
However, the previously established Agency policy was to treat
all reactive compounds equally. Since the rationale for this
position was not discussed in detail in the NPRM, it would be
useful at this point to review some of the reasons why EPA has
chosen not to consider differential photochemical reactivity in-
its control programs in the past.
First, there is the issue of multi-day pollution episodes,
of which there are two types: one type is where the pollution
remains in an area for a prolonged period due to stagnation,
and the other type is where the pollution is transported from
one area to a different area without being substantially
diluted. With both types of multi-day pollution, the rate at
which a particular species reacts becomes less important since
it has a longer time to react. For example, a slowly reacting
compound may not react completely in a single city, but could
be transported to another city where it would continue to
react. Consequently, reaction rate data showing that attack by
OH radical occurs slowly, or even single-day modeling studies
showing a compound to be less reactive, are not sufficient to
quantify the effect of a particular compound during a multi-day
episode.
Second, the modeling capabilities and chemical data
available are too limited in many respects to completely
address this issue. The available models are only
approximations of what happens in a real airshed, and thus, the
results cannot be considered to exactly describe the actual
effect of any particular compound. Photochemical modeling
requires a great deal of meteorological data to account for
things such as transport and dispersion. Often these data are
-------
3-35
Table 3-7
Projected Number of Non-Northeast
Urban Ozone Nonattainraent Areas*
Projection Year
Control Scenario
Base to Phase I
Phase I to Phase II
1990
67
1995
54
44
2000
52
43
2005
58
46
2010
58
52
The number of modeled Non-Northest urban ozone
nonattainment areas based on the 1986 to 1988 data is 70.
-------
3-36
not completely available, and assumptions are necessary. Other
modeling simplifications are generally included to reduce
computational time (and thus costs). With respect to chemical
data, it is noted that for complicated molecules there can be
too many reaction pathways to model completely. Even for a
relatively simple molecule like methanol, there has been some
question about its reaction mechanisms.[11] This situation can
result in the exclusion of several mechanisms that are deemed
to be insignificant. Thus, predicting the reactivity of a
molecule requires simplifying assumptions, all of which add
some degree of uncertainty to any attempted analysis.
Finally, attempts to regulate while considering reactivity
can easily result in unworkable regulations. As noted before,
future changes in HC to NOx ratios could change the real-world
reactivity of a compound, which combined with the potential for
changes in modeling technology, could make analyses continually
subject to change. Also, regulating on the basis of reactivity
could require careful monitoring and controlling of the
chemical composition of fuels. This would be much more
difficult than controlling a fairly simple parameter such as
RVP. Moreover, since fuels tend to contain a large number of
components, the volume of data necessary to accurately predict
the reactivity of different types of emissions (e.g.,
evaporative emissions) could become overwhelming. While
concerns about the ease of regulating are not adequate
justification for not considering photochemical reactivities,
these concerns do need to be weighed against the potential for
benefits from such consideration.
In the NPRM, the photochemical reactivity of butane was
classified using a 1984 EPA report.[12] This report classifies
compounds as "unreactive," "borderline," or "reactive";
reactive compounds being those which are significantly more
reactive than ethane (based on smog chamber and/or rate data).
Using this system butane was classified as reactive. In
support of this conclusion, it was noted that butane would also
be classified as reactive using the GM scale that was developed
in the mid-1960s.[13] The NPRM analysis went further to say:
"Modeling and smog chamber data also verify the
contribution of butane to ozone formation in the
troposphere. Based on available information, it is EPA's
position that butane is a photochemically reactive
compound which contributes to ground level ozone
formation. Therefore, reductions in butane emissions
through in-use fuel volatility control are expected to
lead to subsequent reductions in ambient ozone levels."
Thus, no numerical consideration of the specific reactivity of
butane relative to average VOC was attempted by EPA.
-------
3-37
The photochemical reactivities of ethanol and methanol
also were discussed briefly in the NPRM; no other oxygenates,
however, were discussed. The only oxygenate which EPA
considered in its air quality analyses was ethanol. For
calculational purposes, ethanol was assumed to be as reactive
as average VOC on a per carbon atom basis, which meant that
ethanol was assumed to be only 62 percent as reactive on a mass
basis. This difference is because ethanol molecules contain
oxygen atoms, and thus, ethanol has a much higher mass to
carbon atom ratio than most VOC. This approach was chosen
since it is the oxidation of carbon that contributes to ozone
formation in the troposphere. It was also noted that the rate
at which ethanol reacts with the OH radical in the atmosphere
(the primary mechanism for oxidizing VOC) is on the same order
as hydrocarbons such as butane or toluene. The air quality
impacts of ethanol blends were estimated based on these
assumptions.
The NPRM discussion also noted that studies had indicated
that methanol is less reactive than typical hydrocarbon
vapors. The studies showed the reactivity of methanol to be 2
to 43 percent as reactive as average VOC on a carbon basis.
These numbers were used to estimate the air quality impacts of
methanol blends.
The increase in emissions of formaldehyde due to
combustion of methanol and ethanol blends was also discussed.
Due to the very high reactivity of formaldehyde, which offsets
the lower reactivity of methanol and ethanol, the reactivity of
exhaust emissions from methanol and ethanol blends was assumed
to be the same as from gasoline.
2. Summary and Analysis of Comments
The Agency received many comments (from Sohio, the Ad Hoc
Ethanol Committee, OFA, the Ohio Farm Bureau, NESCAUM, Sun
Refining Co. and GM) which suggested that EPA ought to consider
the photochemical reactivity of evaporative emissions in this
rule. Most of the comments were with regard to butane (and
other light: paraffins) and oxygenated compounds. The
commenters stated that these compounds are less reactive than
average VOC. This conclusion was based on rate constants for
reaction with OH radical, the results of smog chamber studies,
and the results of single-day computer modeling studies. Since
light paraffins (and oxygenated compounds for oxygenate blends)
make up a large part of evaporative emissions, the commenters
felt that controlling such emissions will not be as effective
in reducing ambient ozone as EPA has suggested.
One of the modeling studies submitted, which was performed
by Systems Applications Inc. (SAI), looked at various control
strategies involving ethanol blends.[14] The model predicted
that even if ethanol blends had higher RVPs than straight
-------
3-38
gasoline, they would still result in less ozone production for
many cities. The report explained that this was due to the
combined effect of ethanol's reactivity and its effect on CO
emissions (which contribute to ozone formation). The report
went further to say that there are two key aspects of ethanol's
chemistry that make it less reactive. First, it noted that the
initial reaction of ethanol with hydroxyl radicals generates
only half as much ozone as the initial reactions of typical
hydrocarbons. Second, it stated that acetaldehyde, the
principle intermediate product, is not highly reactive and can
inhibit ozone formation by reacting with NOx to form
peroxyacetylnitrate (PAN).
Similarly, the report stated reasons why the reactivity of
butane is low. It noted that the low rate constant for the
reaction of OH radical with butane is much lower than the
average value used for VOC in EKMA. It also stated that the
main products of the reaction of OH with butane are
acetaldehyde and methylethyl ketone, and that these products
are "not much more important in further ozone formation than is
butane itself."
Other commenters also addressed the points noted above,
regarding EPA's rationale for not considering reactivity.
NESCAUM noted that there is generally an incomplete
understanding of the photochemical formation of ozone. Sun
noted that HC to NOx ratios can have a significant effect on
reactivities, and thus, for some areas HC control without NOx
control will not result in attainment. They added that .the
mechanisms of transport and dispersal are not well understood.
Sun also noted that because there are processes which scavenge
ozone in the atmosphere, compounds that react to produce ozone
more slowly do not allow concentrations of ozone to reach as
high of a level as other compounds might. They went further to
say that this slow reaction rate also would allow the compound
to be dispersed before it could produce large amounts of ozone
in an urban area.
Much of the evidence presented to show that butane and
many oxygenated compounds react slowly in the atmosphere was
smog chamber data and rate constants for the reaction with OH
radical. This information, while valid, is insufficient to
allow a quantified estimate of the reactivities of these
compounds in real situations, However, Sohio in their comments
stated that EPA should not rely on smog chamber studies to
determine the reactivity of butane since smog chambers are not
representative of real world conditions.
In addition, the SAI report contained qualitative
discussions which are interesting; however, some of the points
raised are misleading. First, the report claims that
acetaldehyde is not highly reactive; however, at least one
modeling study has shown acetaldehyde to be more reactive than
-------
3-39
formaldehyde, which is accepted as a highly reactive
compound.[11] Moreover, the report notes that acetaldehyde
scavenges NOx by reacting with it to form PAN. This would in
fact be of some benefit with respect to ozone levels, but there
are also negative aspects of PAN formation. PAN itself is also
an air pollutant which causes significant health effects, and
it can be transported long distances before it decomposes and
then regenerates the NOx. EPA regulates ambient ozone levels
as a surrogate for all oxidants, assuming that satisfactory
levels of ozone will result in satisfactory levels of all
oxidants. Substitution of PAN for ozone would clearly be
unacceptable. Another misleading aspect is the comparison of
the rate constant for the reaction of OH with butane to the
rate constants for reactions with the theoretical species of
the Carbon Bond Mechanism (CBM) which is used in EKMA. This is
not really appropriate since the butane rate constant
represents only the initial reaction with OH, while the CBM
rate constants represent an average of the reactions of all the
carbon atoms initially present as paraffins. Thus, the CBM
rate constants also account for the products of the initial
reaction, while the butane rate constant does not.
The corresponding photochemical modeling study is more
useful, yet it still does not resolve the concerns noted above,
especially the concern regarding multi-day pollution. Thus,
none of the comments are sufficient to justify reversing
established Agency policy. Also, it should be noted that Sohio
was incorrect in stating that EPA relies solely on smog chamber
studies when considering reactivity.
EPA does not deny that butane and many oxygenated
additives react more slowly than average VOC, or that such
slowly reacting compounds can result in somewhat less ozone
than other compounds. Rather EPA holds that at this time it is
not possible to accurately address reactivity issues such that
consideration would be workable and appropriate.
This precedent was reaffirmed just recently in the
rulemaking that established standards for methanol-fueled
vehicles. That rule regulated organic emissions on a carbon
basis, and did not give any allowance for lower reactivity,
even though there is some evidence that methanol has a very low
photochemical reactivity.
The SAI report submitted by the Ad Hoc Ethanol Committee,
while it does not resolve the concerns noted above, does
suggest that the potential reduction of CO from vehicles fueled
with ethanol blends could impact ozone levels also. Since the
chemistry of CO is fairly simple and well understood,
quantifying the impact of CO reduction does not involve the
same degree of uncertainty which occurs when dealing with
reactivity benefits of other compounds. For this reason, the
Agency is open to considering the effect of CO on ozone levels,
when dealing with alcohol blends.
-------
3-40
EPA recently commissioned a study by SAI which in addition
to other fuel scenarios, attempted to quantify the ozone impact
of oxygenated blends, taking into consideration both VOC and CO
impacts. To combine all of the exhaust, evaporative, and
running loss emission effects into a net effect, EPA provided
SAI with MOBILES.9 based emissions for the scenarios of
interest as defined by fuel type, calendar year, and ambient
temperature. The adjustment factors used were fuel specific
and were based on test data acquired for several oxygenate and
alcohol fuel blends as outlined in the technical report,
"Guidance on Estimating Motor Vehicle Emission Reductions from
the Use of Alternative Fuels and Fuel Blends."[15] Table 3-8
contains the emission effects for vehicles fueled with
ethanol/gasoline blends with a 0.76 psi margin and also a blend
which meets the same RVP as gasoline. The Ad Hoc Ethanol
Committee submitted technical data describing the average
volatility increase for ethanol blended fuels. Their data
resulted in a 0.76 psi RVP increase for a 10 percent ethanol
blend in gasoline rather than the 1 psi increase in RVP used by
the EPA for its analysis. Further data submitted by API also
supports this 0.76 psi RVP increase at 11.5 gasoline RVP.[16]
SAI's study is presented in the report, "A Low-Cost
Application of the Urban Airshed Model to the New York
Metropolitan Area and the City of St. Louis. "[17] As shown in
Table 3-9, there is virtually no change in peak ozone levels
for any oxygenated blend scenario when the urban airshed model
is used. However, the report points out that for the New York
area, the airshed modeling boundary conditions cause
insensitivity in estimating peak ozone concentrations and
complicates obtaining accurate results.
The St. Louis area ozone modeling is not complicated by
boundary conditions. As shown in Table 3-9, the St. Louis
results also predicted no change in peak ozone for 50 percent
use of an ethanol/gasoline blend with a 1 psi RVP allowance.
Results of another study performed by SAI for the California
Renewable Fuels Association (CRFA) indicate an eight to nine
percent reduction of ozone with use of ethanol blends in the
South Coast Air Basin. [18] The difference in results can, at
least in part, be attributed to different assumptions used.
The CRFA study assumed 100 percent use of the ethanol blend in
gasoline vehicles at 8.76 psi, compared to a base case gasoline
of 9 psi. The St. Louis study assumed 50 percent use of the
ethanol blend at 8.8 psi, compared to a base case gasoline of
7.8 psi. The CRFA study also used a 1985 inventory which
attributed a higher percentage of VOC and CO emissions to
mobile sources than the 1995 inventory used for St. Louis.
Therefore, it appears that allowing a 1 psi RVP allowance
for ethanol blends would not contribute to as significant of a
change in .ozone levels as EPA previously thought. Requiring
ethanol blends to meet the same RVP levels as gasoline could
-------
3-41
Table 3-8
Technology-Specific VOC Emission Effects of Blends
(Percent Change From Gasoline)
Total Organic From a Blend With a 0.76 psi RVP Margin
Evaporative Emission Effect
100% Market Share
(no commingling)
Diurnal
Garb
FI
Hot-Soak
Garb
FI
50% Market Share
(maximum commingling)
Diurnal
Garb
FI
Hot-Soak
Garb
FI
Exhaust Emission Effect:
11. 5 psi -Base
RVP + 0.76 psi
+ 80.1
+122.2
+35.3
+20.2
+96.2
+144.9
+39.6
+25.0
9.0 psi Base
RVP + 0.76 psi
+ 41. 1
+42.7
+25.5
+34 . 0
+ 51 . 7
+ 55. 1
+28.5
+ 42.4
Technology
Non-Catalyst -22.8
Open-Loop Catalyst -33.4
Closed Loop -17.2
3.7% Oxygen (10% Ethanol or
5% Methanol/Cosolvent Blends)
CO NOx VOC
+3.8
+4 .0
+8. 1
-4.2
-14 .5
-2.4
-------
3-42
Table 3-8 (Cont'd)
Technology-Specific VOC Emission Effects of Blends
(Percent Change From Gasoline)
Total Organic from a Blend that Meets Gasoline RVP Limits
Evaporative Emission Effect:
100% Market Share
(no commingling) 11.5 psi Base 9.Q psi Base
Diurnal
Garb -9.7 -9.7
FI -9.7 -9.7
Hot-Soak
Garb +14.8 +14.8
FI -5.7 -5.7
50% Market Share
(maximum commingling)
Diurnal
Garb +21.0 -2.4
FI +33.9 -3.3
Hot-Soak
Garb +18.2 +16.2
FI +1.6 +0.3
Exhaust Emission Effect:
3.7% Oxygen (10% Ethanol or
5% Methanol/Cosolvent Blends)
Technology CO NOx VOC
Non-Catalyst -24.5 +3.8 -5.5
Open-Loop Catalyst -34.9 +4.0 -15.6
Closed Loop -21.4 +8.1 -5.1
-------
3-43
Table 3-9
Airshed Peak Ozone Concentrations
Scenario
New York
Scenario 1 (11.5 RVP)
Scenario 2 (9.0 RVP)
Scenario 3 (100% gasohol
at 10.0 RVP)
Scenario 4 (11.5 RVP with
no running losses)
St. Louis
Scenario 1 (10.0 RVP)
Scenario 2 (7.8 RVP)
Scenario 5 (50% gasohol
at 8.8 RVP)
Scenario 7 (100% ETBE blend
at 7.8 RVP)
Scenario 8 (10.0 RVP with
high running losses)
SIP Scenario A
(40% reduction of
Scenario 1 VOC)
SIP Scenario B
(40% VOC reduction where
the most: reactive VOC
are reduced first
Airshed (UAM)
Peak
Ozone
(pphrn) *
17.4
17.4
17.5
17.4
% Change from
Scenarios
15.0
14.5
14.5
14.4
15.4
13.5
-3.3
+2.7
-10 .0
12.3 -18.0
+ 0 .6
+3.3
-0. 7
* Parts per hundred million.
-------
3-44
lead to a decrease in ozone levels, based on the emissions
expected from vehicles fueled with such a blend as presented in
Table 3-8. Policy issues related to ethanol blends are further
discussed in Chapter 6.
For methanol blends, EPA would expect similar ozone
effects to those discussed above for ethanol blends. However,
since methanol blends have been required to meet the same RVP
levels as gasolines since fuel waivers were granted for these
methanol blends, EPA does not believe that any change is
necessary at this time.
V. Effects of RVP Control on Benzene Emissions and Health
Effects
A. Synopsis of the NPRM
1. Benzene Emissions as a Function of Fuel Parameters
Chapter 2 of the Draft Regulatory Impact Analysis modeled
benzene emissions from gasoline vehicles as a function of the
emission type, the vehicle type, and the relevant fuel
parameters. The resulting models were based on theoretical
analysis of the different emission scenarios and the data
available in the literature. Three models were developed for
exhaust benzene emissions as a fraction of total hydrocarbon
emissions depending on whether the vehicle was equipped with an
oxidation catalyst (on pre-1980 vehicles), a 3-way catalyst, or
a 3-way plus oxidation catalyst. Benzene exhaust emissions
were assumed to be a function of the aromatic and benzene
contents of the fuel.
Two sets of models also were developed for benzene
evaporative emissions as a fraction of total hydrocarbon
emissions, one for carbureted vehicles and the other for
fuel-injected vehicles. Since few data were available on
evaporative benzene emissions with which to verify the models,
more than one model was developed for each type of vehicle to
provide a range of estimated evaporative benzene emissions.
Benzene evaporative emissions were assumed to be primarily a
function of the fuel fraction which is benzene, and to a lesser
extent the RVP of the fuel (which affects total HC).
A model of refueling benzene emissions as a fraction of
the volume of fuel dispensed was taken from earlier EPA
work.[19] The model calculated benzene emissions as a function
of the fuel fraction benzene, the temperature of the dispensed
fuel, and the temperature difference between the dispensed and
in-tank fuel. The temperature of the dispensed fuel and the
temperature difference between dispensed and in-tank fuel
should not vary with RVP. As a result, the model was scaled
back such that refueling benzene emissions were assumed to be
strictly a function of fuel benzene content.
-------
3-45
2. Effect of RVP Control on Fuel Composition
Volatility control was projected to result in
approximately a two percent decrease in the butane
concentration of the fuel per psi decrease in RVP. This lost
fuel volume must be replaced with something else, some of which
would likely be benzene or other aromatics. Projections of
fuel aromatic content increases resulting from reduced fuel
volatility were used to estimate both the aromatic and benzene
fuel fractions for various levels of volatility control.
3. Effect of RVP.Control on Benzene Emissions
The models developed for benzene emissions as a function
of fuel parameters were combined with the estimates of fuel
fraction benzene and aromatics to yield estimates for the
benzene emissions as • a function of RVP control. On the basis
of these models, RVP control is expected to result in a greater
fraction of hydrocarbon emissions being benzene.
4. Nationwide Benzene Emissions
The benzene fractions of the hydrocarbon emissions
calculated in Chapter 2 of the Draft RIA and discussed above
were weighted by vehicle registration and catalyst technology
distributions from MOBILES and combined with emission factors
from the Draft RIA to yield exhaust benzene and evaporative
benzene emission factors (mg/mi) as a function of the level of
volatility control (Chapter 3 of the Draft RIA). The refueling
benzene emission rates were similarly combined with the total
refueling emission factors to yield estimates of refueling
benzene emission factors (mg/mi).
The sum of these benzene emission factors were multiplied
by the nationwide VMT estimates for the different classes of
vehicles to yield estimates of the total mass of nationwide
benzene emissions expected with the various levels of
volatility control. The result of this analysis was that
although the benzene emissions as a fraction of total
hydrocarbon emissions tended to increase with volatility
control, the decrease in total hydrocarbon emissions also
resulting from volatility control tended to compensate for this
slightly and resulted in lower total nationwide benzene
emissions.
5. Cancer Incidence Analysis
Risk analyses of both ambient exposures and individual
exposures were applied to the nationwide benzene emission
estimates to yield estimates for the annual number of cancer
incidences expected to result from benzene exposure for varying
levels of volatility control. The impact of fuel volatility
control on the number of nationwide cancer incidences was found
-------
3-46
to be within the accuracy of the modeling techniques. The
Draft RIA concluded that volatility control should have no
significant effect on cancer incidences from benzene.
B. Summary and Analysis of Comments
Volkswagen stated that volatility control will not
necessarily result in an increase in fuel benzene levels.
While this may be true, it is still likely that an increase in
fuel benzene levels will result from volatility control. If
nothing more is done than simply removing butane from gasoline,
the fuel benzene and aromatic concentrations will tend to
increase since they are now a greater proportion of the fuel.
In addition, since butane is a high octane component of
gasoline, its removal will require a substitute also high in
octane. This requirement can be met by using benzene and the
other aromatics. As a result, fuel benzene and aromatic
concentrations very likely will increase with volatility
control. This analysis was supported by Marathon, Phillips,
the Renewable Fuels Association (RFA) and the Ad Hoc Committee
on Ethanol (AHCE) in their comments on the Draft RIA.
The Ohio Farm Bureau along with the AHCE and RFA continued
this line of reasoning by stating that if gasoline volatility
control leads to less use of ethanol in gasoline, ethanol
currently being used to enhance fuel octane will also likely be
replaced with benzene and other aromatics. Although the
resulting increase in benzene would not be large, they feel
that this is further justification for EPA to allow a 1.0 psi
allowance for ethanol blends over the RVP control limits set
for gasoline.
Ethanol is not usually added to gasoline to enhance
octane, and as a result there would not be a great requirement
for additional octane if ethanol were not used. However, the
use of ethanol does dilute the gasoline such that less benzene
exists in ethanol blends. But since the dilution effect is
only approximately 10 percent, and since only approximately
eight percent of the gasoline sold in the U.S. is blended with
ethanol, EPA's treatment of alcohol blends should not have any
significant effect on nationwide benzene-related cancer
incidences.
A final comment involved whether EPA should regulate the
concentration of benzene or aromatics in gasoline. Of those
who commented, VW and NRDC were in favor of EPA establishing
such controls while Marathon, Phillips, and Chrysler do not
believe such regulations are required. Pending the outcome of
the Clean Air Act debates currently underway in Congress, EPA
may consider to regulate the concentration of benzene or
aromatics in gasoline. However, if this is found to be
necessary, EPA will make such a proposal in a separate
rulemaking.
-------
3-47
C. Final Analysis
The comments on the effect of volatility control on
benzene-related cancer incidences were not specific or detailed
enough to challenge the validity of the Draft RIA's analysis of
the topic. However, we have expanded one aspect of the
analysis to incorporate the results of MOBILE4.0 based analyses
and more recent ambient benzene level information.[20 ] The
composite benzene emission factors based on MOBILE4.0 and the
U.S. population and VMT data used to calculate benzene-related
cancer incidences are shown in Table 3-10. The benzene-related
cancer incidences for the varying levels of RVP control are
shown in Table 3-11. The upper -bound values in Table 3-11
result from more recent work by EPA on the topic of benzene
carcinogenicity. The revised analysis reinforces the earlier
conclusion that RVP control should not have any significant
effect on cancer incidences from benzene.-
-------
3-48
Table 3-10
MOBILE4.0 Ben?,ene Emission Factor Data
MOBILE 4.0 Benzene Emission Factors U.S. * VMT * *
Year Exhaust Evaporative Running Loss PopulationlQ^ xlO9
1986 .1021 .0120 .0107 240
1995 .0415 .0044 .0035 260 1936.59
2005 .0315 .0030 .0032 276 2326.33
* Population estimates from U.S. Census Bureau Statistical Abstract.
** From MOBILE4.0 Fuel Consumption Model.
-------
3-49
Table 3-11
Estimated Cancer Incidences Due to Mobile Source Benzene Exposure
Year RVP Exhaust Evap RL Refueling Total Upper Bound
1986 11.5 75 9 8 8 100 155
1995 9.0 44 5 4 7 60 107
2005 9.0 46 45 12 67 114
* Lower bound = low end evaporative emissions, with I/M program
in-use.
-------
3-50
References (Chapter 3)
1. "Ozone Nonattainment Area Analysis - A Comparison of
Bills," prepared by E. H. Pechan & Associates for EPA, January
1990.
2. "California Version of the MOBILE4.0 Model," EPA
memorandum from Dave J. Brzezinski to Phil Lorang, EPA, OAR,
QMS, ECTD, TSS, May 14, 1990.
3. "External Fuel Temperature Measurement," EPA memo
from Daniel Stokes, CES, to Tom Darlington, TE3, April 13, 1987.
4. "Summary of Fuel Temp Increases," Temperature data
collected in support of the Side Fan cooling Test Program, EPA,
August and September 1988.
5.. Schuler, Alan, "Effects of Gasoline Volatility on
the Hydrocarbon Exhaust Emissions From a 1984 Oldsmobile
Cutlass," U.S. EPA, OAR, QMS, ECTD, SDSB, August 1987.
6. Exhaust Emissions on Repeated LA-4s, Running Loss
Test Program, Data Extracted from MICRO Data Base, January 18,
1989.
7. "Reference Material for Phase II Volatility Control
FRM Docket," EPA memo from Celia Shin, TEB, to Tad Wysor, SDSB,
March 26, 1990.
8. Bartus, David B., "PT Evaporative Emissions Model,
Description and Users Guide," U.S. EPA, OAR, QMS, ECTD, SDSB,
September 1988.
9. Running Loss Test Program: Interim Results, U.S.
EPA, OAR, QMS, ECTD, SDSB,. September 16, 1988.
10. "Volatility Regulations for Gasoline and Alcohol
Blends Sold in Calendar Years 1992 and Beyond," EPA memorandum
(with attached report) from Mark Wolcott, TEB, to Chester J.
France, EPA, OAR, OMS, ECTD, SDSB, June 1990.
11. Ito, K., et. al., "Photochemical Reaction of
Alcohol-Fueled Engine Exhaust Gases," 7th International
Symposium on Alcohol Fuels, 1986.
12. Singh, H.B., et. al. , "Reactivity/Volatility
Classification of Selected Organic Chemicals: Existing Data,"
EPA-600/3-84-082, 1984.
13. Caplan, J.D., "Smog Chemistry Points the Way to
Rational Vehicle Emission Control," SAE Transactions, Vol. 74,
1966.
-------
3-51
14. Whitten, Gary Z,, Systems Application, Inc.,
"Evaluation of the Impact of Ethanol/Gasoline Blends on Urban
Ozone Formation," revised final report prepared for Renewable
Fuels Foundation, February 10, 1988.
15. "Guidance on Estimating Motor Vehicle Emission
Reductions from the Use of Alternative Fuels and Fuel Blends."
EPA Technical Report, EPA-AA-TSS-PA-87-4, January 29, 1988
(NTIS tt PB 88 169594/REB).
16. Letter from Dr. Terry F. Yosie, API, to Mr. Charles
L. Gray, Jr., June 28, 1988.
17. "Interim Final Report. A Low-Cost Application of the
Urban Airshed Model to the New York Metropolitan Area and the
City of St. Louis. (Five Cities UAM Study Phase I),"
SYSAPP-89/070. May 15, 1989.
18. Whitten, Gary Z., Systems Application, Inc.,
"Impacts of Ethanol Fuel Use On Air Quality in Los Angeles,"
final report prepared for the California Renewable Fuels
Association, March 8, 1990.
19. Laing, Paul M. , "Factors Influencing Benzene
Emissions from Passenger Car Refueling," SAE 861559, 1986.
20. "Air Toxics Emissions and Health Risks from Motor
Vehicles," Jonathan M. Adler and Penny M. Carey. Presentation
to the Air & Waste Management Association, June, 1989.
-------
CHAPTER 4
ECONOMIC IMPACT
The societal cost of RVP control is composed of several
elements which will be discussed in this chapter. First,
refiners directly incur a cost for producing gasoline with a
lower RVP. Indirectly, two specific aspects of the petroleum
industry, butane sales and purchases and oil and gasoline
imports, will also be affected by volatility regulations.
Second, consumers realize a savings with lower RVP fuel due to
increased fuel energy density (which leads to higher fuel
economy) and decreased evaporative emissions and running loss
emissions. Third, changes in the driveability and safety of
vehicles if they occur, also might have an indirect economic
impact on society. Finally, enforcement of RVP regulations
results in a slight cost to society.
This chapter contains a summary and analysis of economic
comments received on the August 1987 NPRM[l] which relate to
the second phase of RVP controls finalized with this action.
This chapter also contains the projected cost impacts of the
second phase of volatility controls as well as the costs of the
first phase of controls which have been calculated again using
the methodology presented in this chapter.
I. Refining Costs
A. Synopsis of NPRM Analysis
The Draft RIA which accompanied the August 1987 NPRM,
examined the leadtime requirements and economic impacts of
proposed RVP regulations.[2] At that time, EPA proposed that
the second phase of volatility controls take affect in 1992.
This was based on the assumption that three to four years of
leadtime was necessary to meet the lower RVP levels in an
economical manner> (The feasibility of a shorter leadtime was
not analyzed in detail.) The costs were based on refinery
modeling performed by the EPA consultant Bonner and Moore
Management Science, which evaluated the cost of RVP controls at
1 and 2 psi reductions. Based on in-use fuel survey data and
the Bonner and Moore model, an average nationwide cost of
control was determined for several RVP control scenarios.
B. Summary and Analysis of Comments
EPA responded to a number of the economic-related comments
received on the August 1987 NPRM in the January 1989 Final RIA
which supported the first phase of RVP regulations.[3] The
majority of comments received on the August 1987 NPRM were
related to the second phase of RVP regulations finalized with
this action and are dealt with below.
-------
4-2
1. Feasibility of and Leadtime for the Second Phase of
RVP Controls
Very few comments were received on the feasibility of and
the length of leadtime required to meet the second phase of RVP
reductions. API stated that since refiners do not currently
produce 7.0 psi or 8.0 psi RVP fuel, the feasibility must be
demonstrated. One comment received from Texaco claimed that
four years of leadtime are necessary to install additional
refining capacity to replace lost volume and octane. They also
added that four years may be inadequate because of difficulties
in obtaining permits for new facilities. On the other hand,
both the Natural Resources Defense Council and the National
Clean Air Coalition stated that EPA's RVP reduction can be
achieved on a faster timetable, citing RVP controls implemented
by the Northeast States for Coordinated Air Use Management
(NESCAUM) which required 9.0 psi fuel beginning in the summer
of 1989 as opposed to 1992 for the EPA proposal.
Based on the modeling performed by Bonner and Moore, EPA
believes that RVPs at least as low as those being considered in
the second phase of volatility controls are feasible. The
refinery model for California showed that RVP was able to be
reduced to nearly 6.8 psi, slightly lower than the 7.0 psi
limit being considered for Class A areas by EPA in Phase II RVP
controls. It should also be noted that some refiners have
already begun producing gasolines with RVP around 8.0 psi
either as part of a reformulated gasoline program, or in order
to comply with Phase I RVP requirements.
EPA believes that promulgating the second . phase of
controls in 1992 would still allow refiners to meet the RVP
requirements contained in today's rule. At the time the NPRM
was published, the expected generous leadtime made
distinguishing technological feasibility from the most
economical means of achieving RVP reductions unnecessary. Now,
with some of that time having passed, it is important to look
at this issue in more detail. This is because that when such a
distinction is made, it becomes clear that a shorter leadtime
is possible if the most economical response by each refiner is
not seen as the deciding factor. That is, some refiners may
need to modify their short-term operations in order to meet the.
RVP specifications in a slightly less economical manner than
longer-term modifications. However, this is a different issue
from the technological feasibility of meeting the RVP standards
within a given period of time. Due to the different means of
achieving such RVP reductions, EPA will discuss its leadtime
reasoning for Class C fuels separate from Classes A and 3 fuels.
For Class C fuels, the second phase of volatility
regulations would require refiners to meet a 9.0 psi RVP
limit. In order to meet this level of RVP, refiners would not
be able to add the butane which they currently add at the
-------
4-3
refinery. It is possible that -some refiners would be required
to debutanize gasoline even further. To do this, two possible
options for refiners might be to add debutanizing equipment or
to operate their current fractionators in such a manner to
improve the sharpness of their butane removal. These refinery
process changes in which butane is not added or additional
butane is removed will require refiners to make up both the
lost octane and lost volume currently contributed by the butane.
To make up the octane, refiners have several options
available. In the short term they might shift some of the high
octane components which currently are used to produce premium
gasoline into the regular unleaded grade of gasoline. In
addition, they might modify their operations to purchase or
produce more MTBE or other such high octane, low RVP gasoline
components (e.g., toluene, xylenes, and alkylate). Refiners
could increase reformer severity or move to fluidized catalytic
cracking catalysts which produce higher octane. As mentioned,
these are short-term modifications which refiners could perform
to replace the lost octane due to butane removal. For some
refiners, these processes may not be as economically attractive
as other longer-term modifications, but they should allow
refiners to meet the RVP standards within about two years. In
the meantime, they could continue to implement other refinery
modifications, such as building additional alkylation capacity
or MTBE production capacity, which might take a year or two
longer to plan and implement but would ultimately result in
more economically efficient production of RVP-controlled fuel.
To make up the lost volume due to butane removal, refiners
should be able to change operating conditions within about two
years in order to decrease the amount of butane in the final
gasoline product. One way in which refiners could do this
would be to increase their refinery capacity utilization.
Current projections show the U.S. refining industry operating
at about 85-90 percent of capacity.[4] This seems to show that
some refiners may choose to increase their throughput without
building 'additional refining capacity. Another way refiners
could make up lost volume (if octane needs could be met in
other ways) would be to operate their reformers at a lower
severity. For those refiners not able to fully replace the
lost- volume in the short term through increased capacity
utilization or refinery modifications, they should be able to
import finished gasoline. A number of new refineries currently
are being built in oil-producing regions overseas that appear
to be aimed at an export market. Thus, imported gasoline
should be available to supply any extra gasoline volume which
might be needed until domestic refineries can modify their
operations to make up the lost volume, preventing situations of
gasoline shortage.
For Class B and Class A fuels, the second phase of
volatility regulations would require refiners to meet RVP
-------
4-4
limits of 7.8 psi and 7.0 psi, respectively (see Chapter 2).
To meet these levels, refiners would likely have to remove most
of the butanes and some pentanes from gasoline and replace the
lost volume and octane. In order to remove this butane and
pentane, refiners have several options. In the short-term, as
the Bonner and Moore modeling results show, refiners should be
capable of lowering RVP to approximately 8.0 psi without making
any capital investment. To do this, refiners could operate
their fractionators to cut deeper and remove more butane and
pentane. They also could operate their reformers with lower
severity, which would result in less butane. In addition,
within two years refiners should be able to install additional
debutanizers and depentanizers to remove the required amounts
of these components. As for Class C fuel, it appears that a
sufficient supply of octane components exists or could be
produced to meet the gasoline pool octane demands. Again, EPA
believes that any production volume shortfall, which should be
minimized as refiners reoptimize their operations, would be
made up by refining additional crude oil or by importing
increased amounts of finished gasoline.
For the reasons explained above, EPA believes that the
second phase of RVP controls is technologically feasible within
about two years and should allow most refiners to meet the
standards in the most economical way. It is possible that some
refiners may have to choose less economical means of achieving
the required volatility reductions in the short term before
being able to economically optimize their operations. However,
assuming final standards are issued by the summer of 1990,
meeting the required RVP standards for the summer of 1992
appears technologically feasible. In addition, EPA is
considering provisions for states to request more relaxed RVP
standards because of unusual localized economic impacts.
Although about two years in general should provide
sufficient leadtime for refiners to meet the Phase II RVP
standards, EPA believes that less than two years of leadtime
would be insufficient and would not allow refiners to meet the
Phase II RVP standards on a nationwide basis. Two years
appears to be the minimum amount of leadtime in order to
install new equipment and change refinery processes. Less than
two years of leadtime likely would lead to substantial negative
economic impacts.
2. Cost of RVP Control
The main area to which economic comments were addressed
was the Bonner and Moore refinery model itself.[5] Most of the
-------
4-5
comments about the model pointed to specific areas of the model
which commenters believed needed to be corrected or made more
realistic. The American Petroleum Institute (API) adjusted
Bonner and Moore's results for various reasons discussed in
this section and estimated the cost of EPA's volatility
proposal (combined first and second phases of control) to be
$1.5 billion per year. (API also hired a contractor to perform
an independent cost analysis which will be addressed separately
later in this section.) There was also a general comment about
the model which will be addressed first.
A general comment brought against the model was that such
national and regional models oversimplify refinery operations,
overoptimize results, and grossly understate costs. EPA agrees
that a national or regional refinery model does involve a
simplification of the refining industry operations and that
costs will ' vary from one refinery to another. However, we
believe that the Bonner and Moore model contains a level of
sophistication which allows the model to predict cost results
which are representative of average nationwide costs of RVP
controls. In fact, in comments from API, they stated that
Bonner and Moore's modeling results for PADDs I, II, III and V
are similar to models they have used and that the stock balance
changes and operating changes appear reasonable.
In contrast, API stated that the model of PADD. IV
(California) is invalid. API said that Bonner and Moore's
model of California was unnecessarily limited by a certain
volatility parameter (the percent evaporated at 160°F) which
results in most of the pentanes being removed before any of the
butanes are removed. Thus, API claimed, the base case contains
the cost of removing pentanes, and therefore, the RVP control
costs for California are underestimated. EPA agrees with API's
comments on the problems with the percent evaporated at 160bF
contained in Bonner and Moore's California model. We have
eliminated a separate evaluation of California from our
analysis and based our costs on the results from the other
regions which were modeled.
API raised the issue of how the summertime nature of the
volatility control proposal is accounted for in estimating
amortized capital costs. API pointed out that in the Bonner
and Moore report, the amortized capital costs for a given year
were expressed in dollars per day and the costs were spread
over the entire year. In allocating costs to the control
period only,, API claimed that Bonner and Moore multiplied this
daily cost by 152 days (five month control period), and
therefore inadvertently recognized only a fraction of the total
yearly cost. API's comment would be true if in fact the
capital costs presented in the Bonner and Moore report were
actual capital costs. However, it appears that API
-------
4-6
misunderstood that the capital costs presented in the Bonner
and Moore report actually were the costs of meeting the
volatility control requirements, assuming refiners had to meet
the RVP standards over the entire year, not only in the
summertime. In other words, the costs had been artificially
raised by a factor of 2.4 to adjust from the five month control
period to the entire year. Therefore, EPA's methodology of
amortizing the inflated capital costs over the entire year and
then multiplying this cost by 152 days results in the proper
amortized costs for summertime volatility control.
API also mentioned that EPA did not consider that in
states with dual ASTM volatility classifications (states
designated A/B, B/C or C/D, where ASTM recommends either class
of fuel), in-use RVP is at the higher of the two
classifications, and therefore, volatility reductions will cost
more than EPA projected. EPA's current model for estimating
RVP control costs, in which costs can now be calculated by
volatility classification as well as on a nationwide basis, no
longer assumes that pre-control RVP levels are at the ASTM
limits as was the case in NPRM. Instead, the model is now
based on in-use RVPs for 23 cities sampled in 1987. [6] These
samples tended to be somewhat lower than ASTM levels on average.
Based on the in-use summer RVP survey data, the average
in-use RVP of ASTM class A, A/B, B, B/C, and C areas was
determined. Then, based -on each state's current ASTM
classification (which determined an initial in-use RVP based on
in-use survey data) and the new EPA volatility classification
contained in this rule, the actual amount of RVP reduction
needed to meet the RVP standards was determined, and costs were
based on that reduction. Therefore, API's concern about
current RVPs in dual-designation areas should no longer arise,
except for possibly in ASTM class C/D areas. Since no in-use
RVP data were available for areas designated C/D, it was
assumed that these areas were at the average of the allowable
standards (12.5 psi) prior to RVP controls. . If actual
volatility in these areas is higher, EPA's cost estimates for
the first phase of RVP controls would be slightly
underestimated. However, since areas designated C/D by ASTM
only account for approximately three percent of fuel
consumption over the summer control period, the effect on'
overall costs and cost-effectiveness of Phase I RVP controls
would be negligible. Also, there should be no effect on Phase
II costs and cost-effectiveness since the dual classification
system does not now exist i-n the current Phase I volatility
classification system.
API commented that refiners will have to produce gasoline
which has an RVP lower than the standards to assure that
gasoline at the terminal complies with the volatility
regulations due to the lower volatility fuel mixing with the
higher RVP fuel already in the storage tanks. Therefore, they
-------
4-7
said, refinery costs will be greater than EPA estimated. EPA's
analysis of the transition period before volatility regulations
take effect is presented in Chapter 2 of this Final RIA. That
analysis shows that refiners do not need to produce gasoline
lower than the standards to assure compliance at the terminal
if they start producing regulated gasoline six weeks before
enforcement begins (which our cost analysis takes into
account). For example, for an average refinery producing Class
D fuel (13.5 psi) which will be required to meet Class C fuel
requirements (9.0 psi) on May 1, if they produce 9.0 psi for
six weeks prior to the beginning of the control period (instead
of the 11.5 psi gasoline they would normally produce without
volatility regulations), then the RVP of the fuel at the
terminal would be at 9.0 psi by May 1. For this reason we feel
our analysis correctly estimates the effect of the transition
period on the cost of RVP control.
API also commented that because of RVP testing variability
refiners would have to include a margin of compliance of 0.3
psi to assure that their fuels meet the volatility
requirements, and that this margin should be considered as part
of the industry costs. EPA agrees that it is probably
necessary for gasoline refiners to include a margin of
compliance in order to assure that their gasoline meets the
volatility regulations. Although it is not clear to EPA how
much of a margin is required, it seems likely that as refiners
gain more experience with producing regulated RVP gasoline and
the RVP test methods, they will be increasingly able to
minimize this margin of compliance. Therefore, EPA has not
included a margin of compliance in its cost calculations.
API commented that Bonner and Moore overestimated the
blending value of butane at 65 psi, stating that based on their
confidential survey of refiners, a blending value of 61 psi is
more realistic. As API pointed out, a lower butane blending
value would result in a greater amount of projected butane
rejection to achieve a given RVP reduction, and therefore lead
to greater costs. However, EPA believes Bonner and Moore's
analysis is reasonable considering that the butane blending
value used was based on publicly available information. EPA
cannot base its analysis on a summary of information which has
not been provided to us because it is considered confidential.
Therefore, a butane blending value of 65 psi will continue to
be used.
Related to gasoline demand, API pointed out in their
comments that Bonner and Moore used 1990 fuel consumption
numbers for 1992, and since projected fuel consumption will be
higher in 1992 than 1990, total refinery costs were
underestimated. One refiner stated that EPA should increase
-------
4-8
its base gasoline demand by five percent to agree with API's
forecasts. In addition, OMB commented that EPA's MOBILES,0
fuel consumption model underestimates fuel consumption and
therefore costs. For this analysis of the second phase of RVP
controls, EPA has updated its estimate for gasoline demand.
Our current economic and environmental impact analyses now are
based on the MOBILE4.0 fuel consumption figures, which are
higher than the MOBILES.0 fuel consumption figures used in the
earlier analyses.
API and some refiners claimed that EPA did not recognize
any investment or operating costs to fractionate butanes or
pentanes from gasoline and to store them, and therefore
understates the actual costs to refiners. EPA disagrees with
this comment. Bonner and Moore included equipment in the
refinery model such as units to debutanize and depentanize
gasoline as well as pentane splitters. In addition, as will be
addressed in more detail in the next section, EPA believes that
refiners will modify their operations to purchase less butane
from outside sources or find alternative uses for the butane
and pentane which will be displaced from gasoline within the
refinery. Therefore, additional storage facilities are not
likely to be required for butane or pentane, and EPA believes
that the Bonner and Moore model does account for the impacts
which RVP control will have on the handling of butane and
pentane by the refining industry.
One refiner commented that EPA should add back the
motorist fuel savings which occur from the more energy dense
fuel and reduced evaporative emissions, which were credited to
refinery costs instead of to consumers and society. EPA
believes that this commenter misunderstood the way in which the
improved fuel economy and reduced evaporative emissions are
handled. In fact, our refinery cost and cents per gallon cost
estimates do not include these credits, which are discussed
later in this chapter. These credits are applied only when
determining the net societal cost and cost effectiveness of RVP
controls.
Several comments were received which related to the
additional costs which would be incurred for increased RVP
testing of fuels throughout the distribution system (for
documentation and other related activities). API referred to
those costs as "verification" costs, and estimated the cost to
be in the range of $100 to $115 million per year, assuming all
work is contracted out and includes the cost for travel time,
obtaining the sample, testing, and producing a certification of
analysis. Unfortunately, API did not provide any documentation
for their cost estimate. One commenter claimed that because of
the RVP regulations, many distributors and retailers would find
themselves with unsold stocks of cold-weather, high RVP
gasoline. This commenter said that these businesses would be
forced to either violate the regulations or shut down
operations, which would lead to bankruptcy.
-------
4-9
EPA does not foresee the costs or the type of problems
resulting from verifying or complying with the RVP regulations
to be as severe as summarized in the preceding paragraph.
Refiners currently must measure the RVP of gasoline they
produce even without RVP regulations. EPA would expect that
the only additional RVP testing due to this program would be
for oversight programs which certain parties may institute.
This RVP testing would not likely be batch by batch, but more
periodic than existing refinery level programs. Therefore, EPA
believes that there probably will be some additional costs of
verifying that the gasoline being distributed and sold meets
the RVP regulations which cannot be quantified. However, there
should be no significant increase in costs over Phase I
verification costs. In relation to the refinery costs
presented in this chapter, it would not be a significant cost
and should have little, if any, discernible effect on the price
of gasoline.
Regarding the impacts on distributors and retailers, EPA
believes that if these businesses plan ahead properly, they
will not find themselves left with fuel which cannot be
marketed because it has too high of an RVP if they plan ahead
properly. As the beginning of the control period approaches,
these businesses will have to do additional planning in order
to make sure that any gasoline they purchase will not cause
them to be in exceedance of the RVP standards -for their
location. For retailers this should not be a big problem if,
as in the Phase I program, refiners and terminals are required
to start complying with the RVP standards one month in advance
of retail service stations.
One comment received from OMB stated that EPA should use a
more reasonable crude oil cost. For this analysis EPA is using
a cost of $20 per barrel which is slightly below the price
projected by the Department of Energy for crude oil in 1.995 of
$20.60 per barrel and is also very similar to the costs assumed
in the API analysis.[7] Therefore, EPA believes that a $20 per
barrel crude oil cost is reasonable to use for projections into
the foreseeable future.
Several refiners submitted comments which stressed that
RVP control will result in reduced gasoline production of up to
5,6 percent and will also affect their capability to produce
certain types of products such as premium unleaded gasoline and
aviation gasoline. EPA agrees that volatility control to the
proposed levels would have a significant impact on refinery
operations. This analysis takes into account the economic
impact to refiners of RVP control • on reduced gasoline demand
and thus production. Regarding the impact on specific fuel
production effects, EPA acknowledges that RVP control will
inevitably lead to moderate changes in the product slates for
some refineries. However, the refining industry is a highly
developed industry which is capable of change and can respond
to market demands in an efficient manner. Therefore, EPA does
-------
4-10
not expect a major long-term effect on the ability of refiners
to supply the products which the market requires.
In addition to providing the comments on Bonner and
Moore's refinery model which have been discussed above, API
hired a consultant, Turner, Mason & Company, to perform a
similar analysis quantifying the impacts of RVP controls on
refiners. The findings of their study, in and of themselves,
are a comment on EPA's analysis and are addressed below.
Turner and Mason's refinery analysis used a set of 'eight
linear programs and four simulator models to determine the cost
of compliance for meeting the original EPA proposal as well as
an alternative API proposal under several scenarios. The final
cost of compliance also included costs for butane fractionation
and handling which were estimated based on a confidential
survey of refiners. API's nationwide cost for meeting EPA's
combined first phase and second phase proposals (based on
Turner and Mason's model, assuming b-ase gasoline demand) was
determined to be $934 million per year, substantially higher
than EPA's estimate. However, there are several major areas
which differ between API's analysis and EPA's analysis which
EPA believes caused an overestimation of costs by API.
The first area of difference is the length of the
compliance period. API assumes that six months of volatility
control will be necessary to comply with four months of actual
control at retail service stations. As presented in Chapter 2
of this Final RIA, EPA's analysis shows that 5-1/2 months of
actual production of controlled fuel would occur. This
shortened period of compliance would lower costs by
approximately 8 percent.
A second area of difference which leads API to
overestimate costs is that they project in-use RVP will have to
be reduced 3.1 psi to comply with the EPA regulations. EPA
believes that API's estimate of the required RVP reduction is
too high for several reasons. First, API assumed that in-use
gasoline RVP was at current ASTM limits, which is not the
case. Based on 1987 MVMA summer fuel survey data, average
in-use RVPs categorized by ASTM class ranged from 0.2 psi to
0.5 psi below the ASTM limit. For those cities with dual ASTM
classifications, the average RVP ranged from 0.8 psi to 1.0 psi
below the maximum RVP allowed. Therefore API's estimate of the
required RVP reduction appears to be too high.
A third area of difference previously highlighted is that
API has assumed that in-use RVPs will have to be lowered below
the proposed limits to assure compliance at the terminal, due
to mixing with the more volatile gasoline which remains in the
storage tanks. API has included a margin of compliance of 0.3
psi to assure that in-use RVP meet the volatility standards.
For reasons discussed above, EPA has not included a margin of
compliance as part of the costs of complying with the RVP
regulations.
-------
4-11
A fourth area of difference is that API included all of
the nation in their cost analysis. For the EPA cost analysis,
the eight Northeast states have been eliminated because they
have begun their own programs of summertime gasoline RVP
control or are affected by such programs. To put the API
results on the same basis as EPA's analysis, API's nationwide
costs should be reduced by about 16 percent (the approximate
fuel consumption fraction of the eight Northeast states).
The overall effect of these four assumptions has caused
API to overestimate the actual amount of in-use RVP control
required to comply with the second phase of volatility
regulations. Assuming 1987 in-use RVP levels, EPA estimates
that in-use RVP must be reduced from pre-control levels by only
about 2.0 psi to comply with the second phase of regulatory
requirements. This large difference in projected RVP
reductions causes API to overestimate costs by a significant
amount which cannot be quantified easily without rerunning the
API model, but which accounts for a substantial part of the
difference between our two analyses.
Another major difference between API's and EPA's refinery
model is that it appears the API model does not optimize their
refinery operations as well as the EPA model does. The effects
of this can be seen in the amount of incremental crude
purchased in order to meet product demands after volatility
controls are in place and also in how the model refines this
incremental crude into various products. For example, to
obtain a 2.5 psi drop in RVP, the Bonner and Moore model
purchased approximately 280,000 barrels per day of extra crude
oil, whereas, to lower RVP by 2.7 psi the Turner and Mason
model purchased 453,000 barrels per day of crude oil. Closer
examination of the Turner and Mason product slate shows that
almost 55 percent of the incremental crude ends up as residual
oil. This is an extremely high fraction considering that
roughly only five percent of crude oil purchases end up as
residual oil in the base case without RVP controls. Assuming
Turner and Mason's crude oil price of around $19 per barrel and
residual oil value of around $15 per barrel, the Turner and
Mason model is incurring a $4 per barrel loss on a significant
volume of crude oil purchases.
In order to estimate the impact that the extra crude oil
purchase has on API's costs, it is necessary to make some
assumptions. As stated previously, EPA's analysis shows that
to meet the second phase of RVP standards, refiners will have
to reduce RVP approximately 2.0 psi from pre-control levels.
Assuming that incremental crude purchases are proportional to
the amount of RVP reduction, Bonner and Moore's model would
project that refiners would be required to purchase around
224,000 barrels per day of incremental crude oil to reduce RVP
by 2.0 psi. As previously reported, API's analysis projects
that refiners would have to purchase 453,000 barrels per day of
-------
4-12
crude oil to meet a 2.7 psi reduction. However, their cost
estimate for EPA's proposal is based on a 3.1 psi reduction,
which means they are projecting that refiners will purchase
about 520,000 barrels per day of extra crude oil.
Therefore, if API's modeled refinery were as efficient as
EPA's model, and used an in-use RVP reduction of only 2.0 psi
(the reduction which EPA's analysis shows is necessary to meet
the standards) the resulting crude purchases would be lower by
approximately 296,000 barrels per day. At a loss of $4 per
barrel, this results in a daily loss of $1.18 million per day.
Subtracting this amount from API's projected daily cost of RVP
controls, eliminating the eight Northeast states from the
analysis and adjusting for a 5-1/2 month compliance period,
results in a modified API non-Northeast nationwide projection
of $546 million per year. This is actually lower than EPA's
total estimate of $670 million per year for both the first and
second phases of volatility control (which also excludes the
Northeast states).
In order for Turner and Mason's model to make refineries
more efficient, it will probably be necessary for the model to
account for more investment in new refinery equipment.
However, it should be noted that in addition to the Turner,
Mason modeling results, API's cost estimate included a sizeable
amount of investment for butane storage facilities which was
based on a confidential survey of refiners. As explained in
the next section, EPA does not believe that extra butane
storage will need to be added at most refineries. We expect
that refiners will reoptimize production to produce less butane
and find other uses for the butanes which were removed to lower
RVP (such as using them as feedstock for MTBE production or in
some cases as fuel for other units within the refinery) . In
addition, current storage which is used for purchased butane
can be used to store the butane separated out of the reduced
RVP gasoline. Therefore, costs included in the API analysis
for butane storage might better be seen as offsetting any
increased costs assumed to be necessary to improve the
efficiency of their refineries.
In addition to API, three refiners also modeled costs for
their refineries. Since these refiners modeled only their own
operations, it is not possible to compare total refinery costs
estimates but rather only the cost per gallon for reducing
gasoline RVP.
The first refiner, Chevron, used a linear program model
which they normally use in making projections for their own
refineries. The cost of meeting the second phase of RVP
controls using uncontrolled RVP levels as the base, was 3.14
0/gallon for Class A fuel, 2.04 ^/gallon for Class 3 fuel, and
1.21 ^/gallon for Class C fuel. Compared to the costs
determined for this analysis, the Class A fuel cost increase is
-------
4-13
higher than EPA's value of 1.8 ^/gallon, very similar to EPA's
increase of 1.9 ^/gallon for Class B fuel, and slightly lower
than EPA's cost increase of 1.4 ^/gallon for Class C fuel. It
appears that the main disagreement is the cost of compliance
for Class A areas. Unfortunately Chevron did not supply any
detailed information upon which their costs are based, making
it difficult to comment on their analysis. However, as stated
previously, EPA believes its refinery model provides a good
estimate of average refinery costs although costs of RVP
control will vary from refiner to refiner. Also, even with the
higher cost Chevron determined for Class A fuel, EPA would
still consider RVP control as a cost effective means of
reducing VOC emissions.
The second refiner, Sinclair, using a less sophisticated
model, determined the cost of RVP control to be 1.3 0/gallon.
This is actually slightly lower than EPA's average volatility
class- weighted estimate of 1.6 ^/gallon.
The third refiner which submitted their own cost analysis
was Valero Energy Corporation, a company which owns one
refinery and ten natural gas liquids extraction plants. They
modified the Bonner and Moore model based upon their own
operations arid estimated the economic effects of both the first
phase and second phase of volatility controls. Valero
projected the total cost of volatility control to be 6,3
^/gallon -from current operation to the second phase of
controls, substantially higher than Bonner and Moore's
projections. In predicting their impacts, Valero assumed that
butane would fall to fuel value and that ethane and propane
would fall to halfway between the base price and fuel value.
EPA believes that Valero's projected impacts are different
for two reasons. First, Valero is relatively heavily involved
in the NGL industry for a small refiner. Second, Valero's
assumptions about butane prices appear unrealistic to EPA. As
discussed in the following section of this chapter, an analysis
performed by EPA's contractor does not show that butane prices
will drop to fuel value, but rather that the price will decline
by only about 11 percent. In addition, EPA does not foresee an
impact of the magnitude projected by Valero (as well as other
commenters) on ethane and propane prices. An examination of
1989 monthly spot market prices for ethane and propane, as well
as butanes and pentane, shows no identifiable changes in the
prevailing cost trends as a result of the implementation of
Phase I RVP controls (which resulted in significant RVP
decreases, and, therefore, presumably, increased summer butane
supplies). Recent changes in the refining industry also appear
to'bear out EPA's beliefs. MTBE plants are already being built
which utilize field butanes. In addition, ethylene producers
appear to be using increased amounts of butane as well as
ethane and propane.[8] These increases seem to be occurring at
the expense of naptha and residual (which, in turn, can be used
for gasoline production). Therefore, EPA believes Valero's
-------
4-14
cost estimates would come closer to EPA's estimates if more
realistic assumptions were made regarding, the price and uses of
butane and other natural gas liquids, although Valero's costs
still may be slightly higher due to their deep involvement in
the NGL industry.
Several comments were received expressing concern over the
effect which volatility control will have on the refinery
industry as a whole, or on various sectors of the industry.
One general comment was that the net industry effect would
probably result in some refinery closings, reduced gasoline
production and increased gasoline prices. EPA agrees that RVP
controls will result in reduced summertime gasoline production
(which our model already takes into account) and that gasoline
prices will increase slightly. However, this is due to the
fact that motorists will be purchasing gasoline which provides
greater fuel economy and results in fewer evaporative emission
losses. EPA does not expect there to be any refinery closings
because of RVP controls. All refiners will have to incur costs
to reduce RVP, but they will recover their investments through
the higher prices at the pump.
EPA believes that the recent movements made by the
refining industry with regard to reformulated gasoline support
these conclusions. The refining industry as a whole is heavily
involved in developing reformulated fuels which are intended to
produce significant emission reductions from current vehicles.
At this time there are five reformulated gasolines being
marketed, each of which has a lower RVP than EPA's Phase I
requirements. Thus, it appears refiners know that lowered RVP
is a primary component to reduced emissions and they are able
to produce the fuel and find markets for the butanes which are
displaced. In addition, the impacts of volatility control are
much less than the expected impacts of reformulating the entire
gasoline pool, which the refining industry appears to
anticipate.
There were also comments received which stated that RVP
controls would disproportionately burden small refiners,
refiners supplying Class A and B areas, domestic gasoline
producers, and convenience store suppliers. However, there was
little evidence provided to support these conclusions. EPA, in
a separate analysis, has looked at the effects of RVP controls
on small refiners and concluded that they would not be affected
disproportionately (see 54 FR 11883, Section X). As explained
previously in this section, EPA also does not expect that
refiners which supply Class A and B areas to be
disproportionately affected.
Regarding the disadvantage to domestic refiners,
commenters pointed out that since environmental controls are
less stringent in foreign countries and because foreign
refiners will be able to spread the cost of RVP control over
all of their gasoline production, the cost of RV? control for
-------
4-15
gasoline from foreign refiners will be less than the cost for
gasoline from domestic refiners. EPA does not foresee such an
impact on gasoline imports due to volatility controls. The RVP
regulations will not affect the stringency of environmental
control and, therefore, should not increase any existing
difference. Also, it seems unlikely that foreign refiners
would spread the cost of RVP control over their entire
production of gasoline since it would tend to make their
gasoline less competitive with other refiners in their own
country and other countries outside of the United States.
Likewise, EPA does not foresee any negative impacts on
convenience store suppliers.
As already stated, EPA believes our average cost estimates
are meant to incorporate the diversity of effects expected
across the industry. We realize that costs will vary from one
refiner to another; however, we do not expect this variation to
cause a significant disruption within the refinery industry.
In another related comment, one refiner said that since
EPA had exempted Alaska and Hawaii from RVP controls because
gasoline consumption in those states is relatively small, EPA
should also exempt small refiners from RVP controls or set
higher RVP levels than those for larger refineries. This
commenter appears to have misunderstood EPA's reason for
exempting Alaska and Hawaii from RVP controls. EPA has
exempted Alaska and Hawaii from the RVP regulations because
neither state contains any ozone nonattainment areas and also
because each of them has an independent supply of fuel, not
because of the relatively small gasoline consumption in each of
these areas. A similar argument cannot be made for small
refineries. Small refineries exist in all regions of the
country and most, if not all, supply gasoline to states with
nonattainment areas. Therefore, from an air quality standpoint
i't would be counterproductive to allow an exemption or set less
stringent RVP levels for gasoline from small refiners. For
this reason, EPA does not believe it is wise to allow an
exemption or less stringent standards for gasoline produced by
small refiners.
In addition to all of the aforementioned comments which
questioned the accuracy of EPA's projection model of refinery
costs, there were also many comments which supported our
results. As with many of the comments challenging our
analysis, these comments were not supported by any independent
analyses and thus do not reinforce any specific aspects of the
analysis. There is no reason to address them individually.
Finally, one commenter stated that the cost of RVP control
can be considered reasonable only if RVP controls result in a
higher quality fuel. While we do not agree that fuel quality
should be the only deciding factor in determining
reasonableness (see Chapter 5), lower RVP fuel would benefit
-------
4-16
the motorist in several ways that already have been pointed
out. First, lower RVP gasoline contains more energy per gallon
and therefore results in improved fuel economy. Second, there
is less evaporation of the gasoline once the gasoline is pumped
into the vehicle, which results in a greater use of the fuel to
drive the car as opposed to the fuel evaporating and often
being emitted into the atmosphere. In addition, lower RVP fuel
will result in better high-temperature driveability, which will
be discussed later in this chapter. For these reasons, SPA
believes RVP controls will indeed result in a higher quality
motor fuel.
C. Refinery Cost of RVP Control
As explained in the previous section, refiners will need
to make capital investments in a variety of new refining
equipment to meet the second phase of volatility controls
finalized today. The refinery modeling work performed by
Bonner and Moore Management Science included several
improvements made since the original refinery modeling was
performed for the August 1987 NPRM. First, it was possible to
incorporate directly into the model the effects of reduced
gasoline demand under a range of RVP control scenarios. (RVP
control reduces the demand for gasoline because less gasoline
is lost to evaporation, and lower RVP gasoline contains more
energy per unit volume which leads to improved fuel economy.)
Second, the latest modeling was able to estimate the impact of
a drop in the price of butane on raw material purchases made by
refiners and on the demand for products from the natural gas
liquids industry. A final improvement was to extend the range
of the modeling from a two psi drop to a three psi drop in
RVP. This has helped to improve the accuracy of EPA's cost
estimates at lower levels of RVP control.
In addition to these modeling improvements, EPA modified
the model results in two ways to better reflect reality.
First, we excluded the results for California (Region 4 in
Bonner and Moore's model) because, as commenters pointed out,
the base case fuel contained unrealistically high levels of
butane and low levels of pentane. Second, EPA adjusted the
results to represent a more reasonable 10 percent real
after-tax rate of return on investment rather than the higher
15 percent rate included in Bonner and Moore's model.
EPA also has applied the results of the model in a more
sophisticated manner than the analysis contained in the NPRM.
Based on the summer 1987 MVMA fuel survey results, EPA
estimated the pre-control RVP level of fuel in each state by
month based on its ASTM classification and its post-control RVP
level based on the classification contained in the first phase
or second phase of volatility controls. States were then
grouped by volatility classification to determine average
in-use RVPs, and EPA applied a refinery cost to each group.
-------
4-17
(Projections of costs are described below.) This approach
allowed the determination of a separate cost for each of the
control levels (7.0, 7.8 and 9.0 psi RVP) as well as a
nationwide cost, excluding member states of NESCAUM. NESCAUM
states were eliminated from this analysis because EPA considers
all of them to be Class 'C' (9.0 psi RVP) areas which
independently implemented RVP controls to 9.0 psi beginning in
the summer of 1989; therefore, EPA's program should not assume
any benefits or costs for federal.RVP control in these areas.
Table 4-1 contains the results of the Bonner and Moore
modeling for Regions 1, 2 and 3 as described in their study.
The cost estimates represent the additional cost incurred by
the refinery in producing low-RVP fuel (i.e., the cost to the
refinery of producing the control-case volume of reduced-RVP
gasoline less the cost of producing the base-case volume of
base-RVP gasoline). It should be noted that the volume of
gasoline produced in the control cases is less than the base
case because motorists are realizing higher fuel economy on the
reduced-RVP gasoline and are losing less fuel to evaporation as
well.
For the purpose of estimating RVP control costs, it is
necessary to determine the cost of producing the base-case
volume of reduced-RVP gasoline. (Credits for the fuel economy
improvements and emission reductions will be taken later when
determining societal costs.) To determine the base-case volume
costs, the cost reduction to the refinery resulting from lower
gasoline volume requirements was calculated by multiplying the
gasoline volume reduction in each control case by the
pool-average incremental gasoline cost, By adding this value
to the costs shown in Table 4-1, the cost of controlling RVP
(excluding volume reduction effects) as a function of RVP
levels was determined.
Further adjustments were made to these values, First,
costs were adjusted to a crude oil price of $20 per barrel.
Bonner and Moore had evaluated the sensitivity of RVP control
costs to crude oil price with modeling runs in Region 3. Cases
were run under $22, $17, and 527 per barrel crude scenarios.
By interpolation, the cost of RVP control at $20 per barrel was
determined for Region 3, and costs for Regions 1 and 2 were
adjusted proportionally.
Next, an adjustment was made to show the effects which
will occur when refiners have to reduce volatility without
investing in new equipment, which is representative of short
term costs for the first phase of RVP controls. Bonner and
Moore ran a case for Region 3 which determined the cost of RVP
control assuming no investments were made. Based on the
relative cost of volatility control in Region 3 under the
"no-investment" and "investment" scenarios, a proportional
-------
4-18
Table 4-1
Bonner and Moore Costs of RVP Control
(With Investment, $22/bbl Crude,
Excluding
California)
RVP Reduction Level
Region 1 Base
Pool Avg RVP (psi) 11.75
Gasoline Volume (MBPD) 648.94
Direct Refining Cost (M$/D)
Direct Refining Cost (S/bbl1)
Region 2
Pool Avg RVP (psi) 11.01
Gasoline Volume (MBPD) 2075.17
Direct Refining Cost (M$/D)
Direct Refining Cost ($/bbl)
Region 3
Pool Avg RVP (psi) 10.92
Gasoline Volume (MBPD) 3132.91
Direct Refining Cost (M$/D)
Direct Refining Cost ($/bbl)
National
Pool Avg RVP (psi) 11.04
Gasoline Volume (MBPD) 5857.02
Direct Refining Cost (M$/D)
Direct Refining Cost ($/bbl)
1st
10.73
645.52
11.0
0.017
10.05
2064.17
173.4
0.084
9.97
3116.30
370.8
0.118
10.08
5825.99
555.2
0.095
2nd
9.71
642.91
94.6
0.146
9.09
2055.87
654.7
0.320
9.02
3103.77
926.4
0.296
9.12
5802.55
1685.7
0.288
3rd
8.69
640.38
- 259.1
0.399
8.14
2047.78
1339.6
0.646
8.07
3091.55
1554.0
0.496
8.16
5779.71
3152.7
0.538
Denominator is barrels of gasoline produced in base case.
-------
4-19
adjustment was made to "investment" control costs in Regions 1
and 2. The end result of these adjustments were
"no-investment, $20 per barrel crude" RVP control costs defined
for three different levels of RVP control. The national
average costs of RVP control assuming no investment is made are
shown in Figure 4-1.
For, the projection of long-term costs of RVP control (for
1995 and later years), EPA included an additional .adjustment
for the effect of lower butane prices on the cost of RVP
control. As butane prices drop because of RVP controls, other
uses for the excess butane become more economical and can
offset the cost of increased purchases of crude. Bonner and
Moore examined the effect that a reduced butane price would
have on crude purchases. This effect has been included by EPA
only in the long-term prices because refiners may require
several years to modify these operations to take advantage of
the increased butane supply. The end result of these
adjustments were "with investment, $20 per barrel crude, lower
butane price" RVP control costs. The national average costs of
RVP control assuming investments are made are shown in Figure
4-2. To determine costs below the lowest national average RVP
examined by Bonner and Moore, EPA extrapolated the cost curve
to 7.0 psi, as shown in the figure.
Costs taken from these curves were used to estimate the
national cost of RVP controls for the first phase of RVP
controls (from Figure 4-1) and the second phase of RVP controls
(from Figure 4-2). EPA estimated the current RVP of fuel in
each state by month (based on 1987 summer MVMA fuel survey data
for pre-control RVP levels, or on the volatility levels
promulgated in the first phase of controls). The post-control
RVP level was then determined for each state according to the
standards promulgated in the final rule for the first phase of
controls, or recommended in Table 2-3 of Chapter 2 for the
second phase of controls. The gasoline sales volumes were
determined for each state based on Energy Information
Administration estimates of monthly sales of petroleum
products.[9] As previously mentioned, costs were aggregated
according to the volatility class designations developed in
Chapter 2 for use in determining class specific cost
effectiveness. The appropriate RVP control cost was then
applied to the volume of fuel which would have to be produced
in order to meet the volatility standards (about 5-1/2 months
as determined in Chapter 2).
Using the methodology described above, the cost of the
first phase of volatility controls has been calculated again in
order to update the analysis without the NESCAUM states. The
costs are based on the volatility classification map contained
in the first phase of RVP controls (which is different from the
second phase map contained in today's rule). As with the
results of the analysis performed for the Final Rule for the
-------
FIGURE 4-1
RVP Control Costs
($20/bbl crude,w/o Investment, excl. CA)
$
1.5
n
i
t
i
a
I
b
b
I
0.5
0
-0.5
12
11
10
RVP (psi)
I
r-o
o
8
-------
FIGURE 4-2
RVP Control Costs
($20/bbl crude, w/ Investment, excl. CA)
$
n
j
t
i
a
I
b
b
I
0.4
0
12
11
10 9
RVP (psi)
8
7
-------
4-22
first phase of RVP controls, no control costs would be incurred
in Class 'A' areas to meet the 9.0 psi standard because those
areas designated as Class 'A' are already below 9.0 psi. For
fuel sold in areas required to meet the 9.5 psi Class 'B'
standard, an average compliance cost of 0,48 cents per gallon
was calculated. Average compliance costs for those areas
required to meet the 10.5 psi Class 'C' standard would be 0.50
cents per gallon. The total average nationwide refinery cost
of the first phase of volatility controls (excluding NSSCAUM
areas) was calculated to be 0.47 cents per gallon, or $190
million per year. Table 4-2 contains further details of the
Phase I cost calculations.
As stated previously, the costs of meeting the first phase
of volatility controls have been based on the assumption that
no capital investment is made in the short term (see Figure
4-1). However, as time proceeds -and a second phase of
volatility controls becomes effective, refiners will invest in
capital equipment and the cost to meet the RVP levels
promulgated in the first phase of controls will decrease
slightly (see Figure 4-2). However, for the Phase II analysis,
the costs of RVP reduction were based on the RVP reduction
required from Phase I RVP levels, not uncontrolled levels.
Therefore none of the Phase I cost decrease which happens as
capital investment is made has been credited to the cost of the
second phase of volatility controls presented in this analysis.
Table 4-3 presents results for the analysis to lower RVP
from the first phase control levels to those contained in a
Phase II program. (As noted previously, these costs do not
include NESCAUM areas.) For fuel sold in areas designated as
Class 'A1 (7.0 psi), the additional cost to reduce RVP further
will be 1.8 cents per gallon. The average additional
compliance cost for fuel sold in areas classified at Class '3'
(7.8 psi) will be 1.3 cents per gallon. In areas designated as
Class 'C' (9.0 psi), the additional compliance cost averages
0.9 cents per gallon. Nationwide, the overall cost to reduce
RVP from the first phase RVP control levels to the second phase
RVP control levels is 1.1 cents per gallon, or $480 million per
year.
The annual costs cited above for both the first and second
phases of RVP controls include increased operating expenses
plus any amortized capital costs for new equipment. The total
capital cost of the volatility control program (combined first
and second phases) is estimated to be about $670 million.
-------
Phase I
4-23
Table 4-2
RVP Control Costs
1987
EPA Phase I ASTM
Classification Classification
A A
Subtotal A
B B
B B/C
B Cal. B
B A
B A/B
Subtotal B
C C
C Cal. C
C C/D
C B
C B/C
Subtotal C
Percent of
Fuel Sold
4
4
16
4
10
0
2
34
34
0
2
3
3
44
.02
.02
.38
.31
.59
.62
.28
.18
.49
.56
.90
.15
.86
.96
In-Use
RVP
8
8
9
10
8
8
9
9
11
8
12
9
10
11
.60
.60
.98
.68
.60
.60
.67
.50
.27
.60
.50
.98
.68
.18
Post-
Control
RVP
8.
8.
9.
9.
8.
a.
9.
9.
10.
8.
10.
9.
10.
10.
60
60
50
50
60
60
50
21
50
60
50
98
50
44
1990 Cost of Control*
(d/gal) (105$/yr)
0
0 0
0.60
1.42
0
0
0.26
0.48 79
0.55
0
1.09
0
0.15 •
0.50 110
Nationwide (ex. NESCAUM)
83.16
10.40
9.84
0.47
189
Based on no investment, a 5-1/2 month summer control period
and a 1990 MOBILE4.0 fuel • consumption of 9.8 x 1010
gallons gasoline.
-------
4-24
Table 4-3
Phase II RVP Control Costs
EPA Phase I
Classification
A
A
A
Subtotal A
B
B
B
B
B
B
Subtotal B
C
C
C
C
C
C
Subtotal C
Nationwide (ex. NESCAUM)
ASTM
Classification
A
B
A/B
A
A/B
B
B/C
C
Calif
A/B
B
B/C
C
C/D
Calif
Percent of
Fuel Sold
1.90
0.28
0.92
3.10
2.74
0.96
14.86
4.35
7.16
10.14
40.21
0.41
4.40
3.82
27.34
2.90
1.01
39.88
Phase
I
RVP
8.60
9.50
9.00
8.80
8.60
9.50
9.50
9.50
10.50
8.60
9.39
9.50
9.73
10.45
10.50
10.50
8.60
10.35
Phase
II
RVP
7.00
7.00
7.00
7.00
7.80
7.80
7.80
7.80
7.80
7.80
7.80
9.00
9.00
9.00
9.00
9.00
8.50
8.99
1995 Cost of Control*
(izi/qal)
1.67
2.36
1.95
83.19
9.83
8.34
1.82
0.76
1.45
1.45
1.45
2.04
0.76
1.34
0.40
0.55
0.96
0.99
0.99
0
0.91
1.15
28
270
182
480
* Assuming investment, a 5-1/2 month summer control period
and a 1995 MOBILE4.0 Fuel Consumption of 1.0 x 1011
gallons gasoline.
-------
4-25
II. Effect of Volatility Control on the Butane and Pentane
Markets
A. Synopsis of NPRM Analysis
A direct effect of RVP control is to displace butane (and
at low enough RVPs, some pentane) from use by refiners as a
gasoline component. At the time the Draft RIA was prepared,
EPA had available somewhat contradictory assessments of the
quantity of butane likely to be displaced and the resulting
economic effects of that displacement on the natural gas
liquids industry and others. EPA used an average of two
extreme cases in estimating the overall cost of RVP control, A
summary of the original butane analysis follows, The effects
of RVP control on the pentane market was not considered in the
NPRM analysis.
1. Displacement of Butane
EPA reported a figure provided by the Gas Processors
Association (GPA) of 17 million barrels of butane displaced per
year. A more detailed analysis based on Bonner and Moore's
work predicted 5.5 million barrels of excess butane per year
would result, Finally, a Jack Faucett Associates (JFA) report,
which assessed the economic effects of displacing butane,
concluded that significant RVP control would result in large
amounts of excess butane and would lower butane prices to, but
not below, the petrochemical floor price.[10]
2. Economic Effects of Butane Displacement
On the basis of the information available at the time, EPA
decided to use the average of the economic impact of two
scenarios in the overall assessment of RV? control costs. The
first scenario, involving "open" butane marketing, assumed that
refiners will buy only the butane which they choose and that
any losses to the NGL industry will result in increased
revenues for other sectors. The second scenario, characterized
by "fixed" butane marketing, assumed that refiners will buy all
excess butane at pre-RVP-control prices, regardless of its
value to them, thus resulting in no transfers of revenues. EPA
used a simple average of these extremes while stating that the
most likely situation would be close to the "open" scenario
with NGL industry losses serving to increase revenues to users
of butane and for consumers of products involving butane in
their manufacture.
B. Summary and Analysis of Comments
1. Displacement of Butane
a. Amount .Displaced
Several commenters presented estimates of the quantity of
normal butane that would be displaced from use in gasoline as a
-------
4-26
result of the proposed RVP reductions. Updating their earlier
estimates, GPA estimated that 117 thousand barrels per day
(MBPD) would be displaced in 1989 and 280 MBPD in 1992 T Based
on GPA's assumption of 153 days of effective control, these
figures translate into annual totals of 17.8 million barrels
(MMB) of butane in 1989 and 42.8 MMB of butane in 1992.
Enron Liquid Fuels presented butane displacements of 120
MBPD (1989) and 300 MBPD (1992); Unocal also quoted 300 MBPD
for 1992. Using GPA's 153-day assumption, these figures are
equivalent to 18.4 and 45.9 MMB, respectively. ARCO's estimate
of 24 MMB per psi of RVP control translates into 24 MMB (1989)
and 60 MMB (1992). Finally, Amoco stated that EPA
underestimated butane displacement by 56 percent in 1989 and 38
percent in 1992 because EPA's estimate is based on a transition
period of less than 60 days, which Amoco claims is necessary to
meet the RVP regulations.
Bonner and Moore's most recent analysis for EPA projects a
nationwide reduction in butane purchases of about 40 MMB per
year for a 2.5 psi drop in RVP. This is based on the five and
one-half month period of control which EPA has determined is
necessary to result in four months of volatility control at
retail service stations. However, this is not necessarily
synonymous with excess butane supply because, as discussed
below, many refiners are likely to begin to reduce their
internal butane production wherever possible and develop other
internal refinery uses (e.g., isobutane, MTBE, and alkylate).
Thus, some refiners which currently act as net producers would
not be adding that butane to the "open market" pool.
b. Effects of Replacing Displaced Butane
More than one hundred commenters stated that the butane
displaced from motor fuel would need to be made up by importing
additional crude oil. GPA pointed out that two barrels of
imported crude oil would be required for each barrel of
displaced butane; using their numbers, this would be 35.6 MMB
(1989) and 85.6 MMB (1992) of imported crude oil per year.
Commenters also raised the issue of the effect on the U.S.
trade- deficit of such imports. This issue of the effect of RVP
regulations on oil imports will be addressed later in this
chapter.
EPA does not believe that all of the volume of butane
displaced by RVP reductions would likely be replaced by
refining additional imported crude oil. We mentioned above
EPA's expectation (discussed in more detail below) that
refiners would expand their capacity for converting • and
upgrading normal butane into other gasoline components such as
MTBE, alkylate, and isobutane. Such expansion would have the
related effects of reducing the potential surplus of normal
butane (as discussed in the previous section) while also
-------
4-27
producing gasoline components within the refinery with which to
make up all or part of the lost butane volume. All of this
replacement volume would then directly reduce the- need for
additional refining of imported crude oil.
Offsetting any additional importation of crude oil would
be a reduction in the amount of imported butane. As reported
by JFA, normal butane imports between 1982 and 1984 ranged from
17.4 MMB to 21.5 MMB per year. Not all of this imported butane
would necessarily be displaced, due to contractual arrangements
and some cases in which the raw material originated in the
U.S. However, much butane which is currently imported would
clearly lose its market, thus easing any negative effects of
volatility controls on the balance of trade and energy
security. From an overall energy standpoint, in fact, refining
more imported crude oil may be preferable to blending imported
butane into gasoline since so much butane is currently lost by
evaporation to the atmosphere and thus is not available for use
as fuel for the vehicle.
2. Economic Effects of Butane Displacement
Most of the comments on the economic effects of displacing
the use of butane in gasoline fall into two general categories,
as follows: a) other uses for butane will drive the price
down, with a number of resulting negative effects including
broader effects unrelated to the price depression; and b) other
uses for butane are themselves problematic, apart from the
price depression.
a. Depression of Butane Prices
Commenters from the NGL industry generally assumed that
there would be a severe and permanent drop in butane prices as
RVP controls were implemented. GPA made specific estimates of
the degree of price depression, In their analysis of the five
PADD districts, GPA projected normal butane prices would fall
by 1 to 10 cents per gallon (cpg) in 1989 and by 2 to 14 cpg in
the most likely 1992 scenario. The impact of summer RVP
control was found to vary among the PADDs based on butane's
predicted uses and current prices for petrochemical feedstocks
and fuels with which the surplus butane is assumed to compete.
Several other commenters reinforced the conclusion that prices
would reach those of petrochemical feedstocks or industrial
fuels, but they did not offer any quantitative support.
GPA and other commenters attributed a range of negative
effects to the projected drop in butane prices. Most of these
anticipated effects would occur, it was claimed, because of
presumed shifts in butane usage or from reduced revenues to and
closings of gas processing facilities. First, if large
quantities of normal butane become attractive as a
petrochemical feedstock, several commenters concluded that
-------
4-28
butane would compete with other NGLs (primarily ethane and
propane) and depress their value. Another displacement GPA
mentions is the effect on terminals which import butane since
surplus domestic butane would have a serious effect on this
business.
GPA also outlined their position on the effects of the
loss of revenue to the NGL industry. These effects would be
compounded by the fact that nearly 40 percent of gas processing
plants (representing 12 percent of 1986 production) are
operated by smaller, independent companies. GPA predicted an
overall loss in revenues of as much as 50 percent due to the
drop in butane prices. Phillips added that there would be
reduced profits for reinvestments and/or increased consumer
costs as profitability is maintained. Finally, more than one
hundred commenters stated that lower revenues for gas
processors means reduced income for thousands of upstream
interest holders; ARCO and Texaco added that this effect on gas
producers may result in less incentive for domestic gas
exploration and development.
GPA and other commenters said that given the presumed
butane price depression and the vulnerability of independent
gas processors, many gas processors would shut down. These
closings would then have direct .and indirect effects. Small
natural gas producers dependent on these processors would,
according to GPA, need to .choose between laying new pipe to
other facilities or shutting in the gas. Another effect of gas
processing plant closings would be on the supply and service
industry supporting the gas processors. Finally, lost
production of other NGLs resulting from plant closings would,
GPA believes, be replaced by imports of ethane, propane, etc.
GPA and other commenters also stated that there would be a
rippling of effects through society if many gas processors
close. Commenters listed loss of jobs, loss of local, state,
and federal revenues, reduced consumer spending, and possible
increase in bank failures could eventually result from the
presumed drop in butane prices and the closing of processing
plants.
EPA's assessment of the likelihood of severe price
depression differs from that expressed in most of the
comments. The primary difference lies in EPA's assumption of a
dynamic response by the refining industry to the reduced demand
for butane as a gasoline blendstock. Specifically, as
mentioned above, EPA expects that the current growth in
installation of MTBE production capacity would continue and
increase, absorbing large amounts of otherwise surplus butane.
For example, even at today's butane prices U.S. MT3E capacity
has grown from 11,6 to 35.6 million barrels per year between
1984 and 1987.[11] In addition, SPA expects alkylation
capacity as excess butanes become available due to RVP
-------
4-29
controls; ETBE production may also increase. Also, conversion
of some butane into isobutane for use in gasoline would further
utilize surplus normal butane. By' converting or upgrading
butane in these ways for use as gasoline components, refiners
would likely prevent the value of butane from falling
significantly below its current value.
An examination of the monthly spot market prices for
butanes in 1989 (after the Phase I RVP controls took affect)
also supports EPA's position on butane price effects. Even
though refiners reduced RVP levels by significant amounts
during the summer of 1989, and, therefore, summer butane
supplies should have increased significantly, butane prices
showed no identifiable changes in prevailing trends as a result
of Phase I RVP controls.[12]
EPA believes that the trend to use butane for MTBE and
ETBE production would likely increase very significantly if RVP
controls were to make surplus butane available and also if
oxygenated fuels programs grow in the future. It is entirely
possible, particularly during the winter months when current
butane demand should not be affected, that the presence of
expanded MTBE, ETBE, and alkylation capacity may require some
refiners to seek butane on the open market when internally
produced supplies are insufficient. EPA believes that the
commenters have overlooked these emerging high-value uses for
butane in their scenarios of the likely consequences of RVP
control.
The consequences of. the scenario EPA projects are much
less severe than those projected by most commenters. EPA
agrees that without extensive expansion of capacity for
production of MTBE, etc., there likely will be some loss of
summer butane sales for many gas processing facilities.
However, we do not expect the value of butane will drop
significantly even under such a scenario. As part of their
modeling, Bonner and Moore concluded that butane prices would
drop by no more than 11 percent as RVP is lowered. They
reached this conclusion even without allowing their model to
increase the production of MTBE (which uses butane as a
feedstock) which has been increasing over the last few years.
EPA believes the impact on butane prices would be even
less if more areas adopt wintertime oxygenated fuels programs
on their own or as required by the CAA amendments. MTBE
already has become the primary choice for refiners in the areas
currently involved in oxygenated fuels programs. Based on the
method presented in an earlier EPA analysis, EPA estimates that
for an oxygenated fuels program like the one contained in the
Senate version of the CAA amendments (as of February 1990),
-------
4-30
which would require a 3.1 percent average oxygen content for
six months in all current CO nonattainraent areas, approximately
30 MMB per year of butane would be necessary to supply 60
percent of the market (the remainder supplied by ethanol
blends).[13] This amounts to a substantial volume of butane
and a potentially significant market for excess butanes.
Therefore, EPA does not expect a severe impact on the NGL
industry. The following paragraphs expand further on this
issue.
Fully and partially integrated oil and gas companies
should have the ability to shift production priorities
sufficiently to suffer little or no overall impact from a
partial loss of summertime butane demand. However, there may
be a subset of gas processors which under their current
business arrangement could be economically marginal and
vulnerable to a summertime interruption in the cash flow from
butane sales. Even here, EPA expects that the -desire of gas
well owners and interest holders to produce gas would lead to
shifts in those arrangements. Specifically, most gas producers
are likely to be willing to pay for the service of processing
NGLs out of their gas stream if need be, rather than shut the
gas in. Therefore, EPA believes that RVP control should have
little, if any, effect on imports of ethane and propane.
In . conclusion, EPA does not anticipate widespread
shutdowns of independent gas processors, the -shutting in of
gas, nor other related indirect impacts suggested by
commenters. We do anticipate a moderate loss of revenues for
many independent gas processors and, in some cases, gas
production interests as well. The secondary effects of such a
loss of revenue will undoubtedly include some of those
suggested by commenters; again, however, we believe that GPA
and others have seriously overestimated the magnitude of these
effects.
b. Alternative Uses for Butane
GPA and most other commenters on this issue raised
problems that would arise from the displacement of butane in
addition to the depression of prices. As summarized above,
most commenters said that 'the only options for butane would be
for use as a petrochemical feedstock (primarily in ethylene
production) or as industrial fuel. These uses were called
economically "inefficient" and problematic for several reasons.
The first problem would be that in many areas there is
insufficient installed ethylene capacity to absorb all of the
presumed surplus butane, even if other feedstocks are
displaced. In addition, ethylene plants in many locations do
not have the capacity to "crack" and thus utilize normal
butane. Sufficient seasonal storage is only available in the
Gulf Coast area (PADD III), but such storage has the drawback
of interrupting the cash flow, a particular hardship for
-------
4-31
smaller independent gas processors. Commenters asserted that
the use of surplus butane as fuel has its own problems; in
particular, difficulties in seasonal fuel shifting by
industrial customers and the problem of obtaining air quality
permits for new combustion units in refineries. Finally,
commenters stated that simple flaring of excess butane is
wasteful and also constitutes a new source of emissions.
Regarding the conversion of butane into lower-volatility
gasoline components by alkylation or conversion to MTBE, GPA
and other commenters appear not to expect expansion of capacity
for these processes. Unocal gave as a reason that economic
justification for such new facilities would be difficult given
only four to five months per year of usage. GPA pointed out
that if the price must drop in order for such facilities to be
justified, gas processors will still be hurt.
As the above EPA response to the issue of depressing
butane prices makes clear, EPA believes that internal refinery
upgrading of butane to other gasoline components would become a
chief usage for butane displaced from direct blending. We
disagree, therefore, with comments projecting that the NGL
industry would experience severe problems in finding other uses
for butane. As mentioned previously, EPA believes that current
trends within the refining industry show that excess butanes
are finding a. market as feedstock for MTBE and ethylene without
causing any appreciable consequences for other related
industries or markets. Thus, the issues of availability of
petrochemical and fuel markets appear much less relevant in
this context.
EPA does not agree with Unocal that capacity for upgrading
butanes would be operated in most cases only during the
summer. We made the point earlier that this production is
likely to be year-round and in some cases may represent a "new"
market for butane. GPA is correct that to whatever extent
prices needed to drop to justify expanding such upgrading
capacity, the NGL industry would experience a loss of
revenues. However, as discussed earlier, much MTBE production
capacity is already being installed with no RVP-related
reduction in butane prices. 'Again, we expect little impact
from RVP controls on butane prices.
3 . Displacement of Pentane
Comments were received from several commenters regarding
the effects of RVP control on the pentane market. Both API and
Texaco pointed out that to reduce RVP below 9.0 psi, both
butane and pentane will have to be removed. EPA agrees that at
low enough RVPs pentane removal will become necessary for some
refiners; pentane removal was included as part of the RVP
control mechanism in the Bonner and Moore refinerv model.
-------
4-32
Relatively speaking, the effect of RVP controls on pentane
displacement will be much less than the effect on butane
displacement.
According to API, this increased supply of pentane will
result in negative impacts on the refinery industry. They
claim that there is no process for converting saturated
pentanes into a low RVP gasoline component. Chevron said that
butane has more potential than pentanes for isomerization and
alkylation, implying that pentane does have some potential to
be used as feedstock for other refinery processes. API then
points out that no market exists for pentanes outside the Gulf
Coast and that pentanes cannot be burned as part of refinery
gas since they would condense out and cause safety problems.
In addition, API says that pentanes would require a separate
fuel system and furnaces would need to be modified to burn the
pentane. " Chevron claims that the overall effect of RVP
controls will probably result in the value of pentane being
lowered to fuel value on the West Coast.
As with our response to the effect of RVP controls on the
butane market, EPA believes that there may be a small decrease
in the price of pentane. However, we also believe the industry
response to the increased pentane supply will be more dynamic
and different than the response suggested by commenters. The
Bonner and Moore refinery model included two uses for displaced
pentanes. First, it was used as feedstock for steam crackers
to. produce olefins. Second, it was used as boiler fuel within
the refinery. EPA believes that refiners will be able to find
economic uses for any displaced pentanes and no major pentane
market effects should result from RVP controls..
Ill. Effect of Volatility Regulations on Imports
One issue .which has received extensive comments is the
effect volatility controls would have on imports of crude oil
and gasoline. Commenters stated that since RVP regulations
will result in the displacement of domestically produced
discretionary butane from the gasoline pool, purchases of crude
oil will necessarily increase in order to meet a fixed gasoline
energy demand, as will foreign gasoline purchases since
refineries currently are running at near capacity levels.
Commenters estimated that imports would rise by a total of
300,000 to 410,000 barrels per day of crude oil in 1992.
Estimates of the effect on the U.S. trade deficit ranged from
$1.5 to $2.0 billion per year.
EPA agrees that oil and gasoline imports may rise in the
short term before investments can be made to install equipment
necessary to convert butane to MTBE, ET3E, alkylate, etc.
However, as additional processing equipment is installed and as
a price decrease - establishes butane as a competitive
petrochemical feedstock, purchases of additional crude should
-------
4-33
decrease. In order to assess this issue, the maximum quantity
of normal butane (n-butane) rejected from the gasoline pool in
order to achieve the given RVP reduction was estimated for each
of Bonner and Moore's control scenarios using a blending value
of 65 psi for n-butane. As described in the RIA for the first
phase of volatility controls, the energy content of this
rejected butane, less the energy content of recovered
evaporative emissions, was then compared against incremental
crude purchases made in each control scenario. Results of that
analysis showed that, on average, the energy of the incremental
crude purchased exceeded that lost by the displacement of
butane by a factor of approximately 1.8.
Using this information, an analysis of the effect of both
the first phase and second phase of volatility standards on
imports was made. Based on the amount of fuel undergoing a
volatility reduction, a nationwide (excluding NESCAUM) estimate
was made of the maximum quantity of discretionary n-butane
which would have to be rejected. A total of 9.3 million
barrels of butane per year was calculated (assuming a five and
one half month refining period) for the first phase of
volatility controls. However, as presented in Chapter 3, about
272,000 tons of evaporative emissions should be recovered
during the control period, with the energy equivalent of 2.7
million barrels of butane. Thus the energy equivalent of 6.6
million barrels of butane increased by a factor of 1.8 will be
required in incremental crude oil purchases for the first phase
of RVP controls. This totals approximately 7.7 million barrels
per year of crude oil (47,000 barrels per day) or $154 million
per year at $20 per barrel.
For the second phase of volatility controls (incremental
to the first phase), -a nationwide (excluding NESCAUM) total of
25.4 million barrels of rejected butane per year was calculated
for a five and one half month refining period. However, as
presented in 'Chapter 3, about 402,000 tons of evaporative
emissions should be recovered during the control period under
the second phase of RVP standards, with the energy equivalent
of 4.0 million barrels of butane. Thus the energy equivalent
of 21.4 million barrels of butane increased by a factor of 1.8
will be required in incremental crude oil purchases for the
second phase of RVP controls. This totals approximately 25.0
million barrels per year of crude oil (152,000 barrels per day)
or $500 million per year at $20 per barrel.
Therefore, the maximum effect on oil imports of the
combined first and second phases of volatility control would be
expected to be slightly under 200,000 barrels per day.
Compared to annual crude oil imports which have been between
6,000,000 and 8,000,000 barrels per day since 1987 and have
been growing at an annual rate of about 500,000 to 700,000
barrels per day, the impact of RVP controls on imports should
not be major. [14] It should be noted that this is a maximum
-------
4-34
expected impact and is based on the assumption that all of the
light-end hydrocarbons (butanes, pentanes, etc.) which are
removed to lower the vapor pressure to the promulgated RVP
levels (minus the recovered evaporative emissions) are replaced
by imported oil. As previously stated in this chapter, EPA
believes that refiners will use some of the displaced
hydrocarbons as feedstock for low-volatility, high-octane
gasoline components (such as MTBE) and therefore, lower the
expected impact on oil imports. In addition, it appears likely
that butane imports would decrease as RVP controls lessen the
demand for butane as a gasoline additive.
One additional comment on imports concerned the effect
that EPA's regulatory approach to ethanol blends could have on
the quantity of fuel imported. The commenter cited the fact
that every gallon of ethanol produced domestically can mean up
to two fewer gallons of imported oil. If the approach taken
with respect to ethanol blends allows a 1.0 psi exemption from
the standards promulgated for gasoline fuels, EPA would not
expect any significant impact on the current gasohol market,
since it essentially would continue to allow the splash
blending of ethanol into gasoline at a maximum of 10 percent by
volume. If ethanol blends were required to meet the same RVP
levels as gasoline fuels, and as a result ethanol blending
completely came to a standstill (which EPA would not expect),
then based on the current gasohol market penetration of around
seven percent, [15] approximately 90,000 barrels per day of
extra crude oil would need to be imported to replace the
ethanol with gasoline, assuming the 2:1 oil to gasoline ratio
cited by the commenter above. On a relative basis, this is
still much less than the increase in crude oil imports which
has occurred over the last few years.
IV. Effect of Volatility Regulations on Increased Energy
Density and Evaporative Emissions Recovery
The Draft RIA examined both the fuel economy impact
associated with volatility control resulting from changes in
the energy density of the fuel, and the fuel savings resulting
from a decrease in evaporative emissions. Both of these topics
were reanalyzed in detail in the January 1989 FRIA for the
first phase of the volatility control program, and as a result
nothing more will be said here with two minor exceptions.
First, as explained in the January 1989 FRIA, the estimate
of the relationship between fuel volatility and heat of
combustion had been revised slightly from the DRIA. The
updated analysis coupled with the revised volatility
classification map as explained in Chapter 2, results in the
following percent increases in fuel energy density with the
second phase of RVP control. The increase in fuel energy
density is 0.65 percent for Class 'A' areas, 0.51 percent for
-------
4-35
Class 'B' areas, and 0.38 percent for Class 'C' areas. These
results are used to calculate the magnitude of the credit which
is taken for improved fuel economy when determining the cost
effectiveness of RVP control.
The second area requiring a short discussion is API's
comment concerning the possible effect of poorer cold
temperature driveability with lower volatility fuels on the
value of R (percent change in fuel economy for a percent change
in energy density of the fuel). The response to this comment
was that especially for the level of volatility control being
considered in the first phase of RVP controls, there was no
reason to believe that there would be a significant impact on
cold temperature driveability, and if there was an impact, it
would not significantly effect the value of R. Although this
rulemaking calls for greater reductions in fuel volatility than
were contained in the first phase of the program, as will be
explained in Section V of this chapter, no significant cold
temperature driveability impacts are expected to occur. In
addition, as explained in the January 1989 FRIA, data provided
by API on fuels representative of the second phase of
volatility • controls did not show any apparent correlation
between R and cold temperature driveability. As a result, no
impact of cold temperature driveability is incorporated into
the model used to estimate R.
In conclusion, the results arrived at in the January 1989
FRIA for the first phase of RVP controls appear to apply to the
second phase of controls as well. The best value to use for R
at this point in time appears to be 0.85. Based on discussion
in the FRIA for the first phase of RVP controls, the best value
to use for the evaporative emission recovery factor appears to
be l.O. These values are used in Chapter 5 to determine
credits to apply toward the overall cost effectiveness of
volatility control.
V. Effect of Volatility Regulations on Driveability and Safety
A. Synopsis of Draft Regulatory Impact Analysis
1. Volatility Increases and Driveability Problems
In the period from 1974 to 1985, volatility of unleaded
regular gasoline increased by 10 to 20 percent depending on the
area of the country. In 1985, the average nationwide summer
volatility actually surpassed the average ASTM recommended
limits. The result has been that even though vehicles during
the same period have been designed to operate better on higher
volatility fuels, some vehicles have begun to experience
varying levels of vapor lock, fuel foaming and fuel spurting,
causing unacceptable driveability (and safety concerns). This
statement was supported by comments on the 1985 Volatility
Study, information provided by automobile manufacturers, and
-------
4-36
also by a number of studies which looked into the problems of
hot temperature driveability. These studies showed that
volatility levels were indeed at a point where driveability
problems could be expected on some vehicles. At the same time,
other studies demonstrated that for the period and the levels
of RVP control proposed, cold temperature driveability should
not be a significant concern. This last statement was
supported by in-use information from California where
volatility has already been controlled for some time.
2. Driveability Cost Estimation
The cost associated with current driveability problems (or
the cost savings resulting from improved driveability with
volatility control) was estimated by assuming that people whose
vehicles are experiencing driveability problems were willing to
pay an extra 1$ to 30 per gallon for fuel which would avoid
those problems. Temperature and population data for the 10
largest ozone nonattainment cities were used to approximate the
nation as a whole. This information was then used along with
vehicle age distributions, vehicle usage patterns and
information on the fraction of vehicles with unacceptable
driveability (based on Coordinated Research Council (CRC)
testing) to determine the fraction of fuel sold nationwide
which is burned under conditions of unacceptable driveabilicy.
Multiplying this amount of fuel by the 1? to 3<2 per gallon
range yielded a nationwide cost of poor hot-temperature
driveability of up to $78 million per year,
Although EPA did not rely on the following information for
the final cost estimation, GM and Chrysler provided information
on in-use vehicles which also demonstrated that the costs
associated with hot temperature driveability problems were
real. On the basis of their submittals, annual warranty costs
due to hot temperature driveability problems amount to .roughly
$8.1 million annually, and costs associated with vehicle design
modifications to avoid hot temperature driveability problems
have amounted to roughly $124.1 million annually. However, EPA
decided that not all of these design costs could be recovered
if fuel volatility were reduced since they are costs that have
already been incurred to fix the problems. Still, the
estimated cost savings which could result from removal of
certain corrective parts from the vehicle if volatility were
reduced was estimated to be $44 million annually. Since the
accuracy of the methods used to extrapolate the information
provided by Chrysler and GM into nationwide costs was unknown,
these costs were not incorporated into our analysis. However,
they do serve to support the costs calculated above.
-------
4-37
B. Summary and Analysis of Comments
1. Hot Temperature Driveability
The January 1989 FRIA for the first phase of volatility
controls reanalyzed the topic of hot temperature driveability
based on comments received from a number of organizations. As
a result of that analysis, it was concluded that although it is
still apparent that hot temperature driveability problems
exist, and represent a significant cost to society, an accurate
means of quantifying that cost could not be found. In
addition, since the driveability cost benefit proved to have
such a small effect on the cost effectiveness of volatility
control in the DRIA, failure to take a cost benefit for
improved hot: temperature driveability should have little
overall effect on the regulation. These conclusions are
considered to be applicable to the second phase of volatility
controls as well.
2. Cold Temperature Driveability
A great deal of comments were received on the topic of
potential cold temperature driveability problems if fuel
volatility is reduced significantly below the ASTM ratings.
Motor vehicle manufacturers as well as a few other commenters
did not believe that cold temperature driveability would be a
problem at EPA's proposed RVP levels (although they provided no
support for these statements). However, many of the oil
companies stated that if in-use gasoline volatility is brought
down to the levels proposed, there will be significant'
driveability problems (poor acceleration, hesitation, stalling,
difficult starting) in early spring and late fall, with
particular emphasis by some placed on ASTM Class A and B areas
or high-altitude areas. Amoco and API provided a number of
studies on the topic of cold temperature driveability at
reduced fuel volatilities to support their claims.
The study provided by Amoco consisted of the testing of 9
1980 to 1986 vehicles on fuels of 6.4, 8.2, and 10.7 psi RVP at
25°F and 40°F over a modified CRC test procedure. The
combination of the fuel volatility of 6.4 psi, the test
temperature of 25°F, and the test procedure was selected so as
to test under worst case conditions. This combination is not
likely ever to occur under the EPA program being finalized
today which is described in detail in Chapter 2, As is shown
in Tables 4-4 and 4-5, temperatures of 25°F are not common even
for those areas likely to receive 8.2 psi RV? fuel.
The results of the CRC testing showed that starting times,
idle stalls, driving stalls, severe hesitations and overall
driveability demerits all increased at lower fuel RVPs. Amoco
arbitrarily set the level of unacceptable driveability at 160
demerits, a level that was not met by all of the vehicles even
-------
4-38
at the highest RVP tested. This raised questions as to the
condition of the vehicles prior to testing. If all vehicles
which either stalled or exhibited a tendency to backfire on the
highest RVP fuel are eliminated from the data set, only one
carbureted vehicle is left of the total of nine, and even this
vehicle exhibited somewhat poor driveability regardless of the
level of fuel RVP. As a result, although this test data
appears to justify that cold-temperature driveability worsens
with lower volatility fuels, it cannot be used to show that
properly maintained vehicles will experience unacceptable
driveability on those fuels even under the most severe of
conditions.
This conclusion is supported, at least for newer vehicles,
by API comments which cited test data collected in one of their
programs. They tested 12 1984-87 vehicles on 6.5 and 9 psi RV?
fuels at 25°F and 40°F. (Once again, the testing of 6.5 psi
RVP fuel at 25°F represents the worst case as demonstrated by
the data in Tables 4-4 and 4-5.) Although they noted some
worsening of driveability and increasing of starting times on
the lower RVP fuel, they noted no pronounced effects.
API also provided a Chevron Research study which tested 12
1981-3 vehicles on 6.1 and 8.4 psi RVP fuels'at 55°F and 75°F
over the FTP, and also "similar" vehicles on 8.1 and 11.4 psi
RVP fuels at 43°F. The results of the testing demonstrated
statistically significant increases in vehicle driveability
demerits at reduced fuel volatility. However, once again, no
mention was made of the condition of the vehicles prior to
testing, nor was there any information on the types of vehicles
tested (except for age). Also, although the increase in
vehicle driveability demerits was said to be statistically
significant, the magnitude of the increase was small,indicating
that even at fuel volatility levels below EPA's proposed
volatility limits, the driveability of the vehicles was not
unacceptable.
API also had testing performed on six 1983-86 closed loop
feedback vehicles, three of which were carbureted, two throttle
body injected, and one port fuel injected. The testing was
performed over the FTP on 6.5, 8.0, 9.0, and 10.5 psi RVP fuels
at 42°F, 55°F, and 80°F. API found the vehicles tested to be
relatively tolerant of RVP reductions even at low temperatures,
without showing the driveability degradation as seen with
earlier model year vehicles. API also submitted a report which
performed a statistical analysis of this test data. Although a
function was found which correlated the driveability demerits
raised to the second power (and divided by 10,000) with a
function of fuel RVP and test temperature, there is no
theoretical reason why such a function should be
representative. Rather it was probably just an anomaly of the
data which allowed for the correlation. Even if the function
was representative of in-use driveability, the changes in
-------
Table 4-4
Driveability Comparison of Gasoline TVP
In January and April Under EPA's Proposed Regulations
City Altitude
Correction
Factor
Albuquerque
Atlanta
Billings
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Kansas City
Las Vegas
Los Angeles
Miami
Minn/St. Paul
New Orleans
New York City
Philadelphia
Phoenix
St. Louis
San Antonio
San Francisco
Seattle
Washington D.C.
1.159
1.032
1.096
1.001
1.017
1.017
1.015
1.171
1.018
1.024
1.062
1.008
1.002
1.025
1.000
1.001
1.001
1.032
1.013
1.020
1.002
1.002
1.013
Jan 1988
MVMA Reg
Unleaded
Min RVP
12.6
11.9
13.6
10.0
14.1
14.7
11.7
12.4
13.4
13.1
12.3
12.4
9.9
14.0
12.6
11.9
13.3
12.2
13.1
12.2
12.6
12.7
11.9
Jan Jan TVP April 1988 April April TVP
3%' min Controlled 3% min
Temp RVP (-0.5) Temp
7
14
-20
4
-9
-2
13
-12
-4
-2
20
38
40
-22
26
10
7
24
-2
20
32
13
11
2.5
2.5
1.4
1.5
1.7
2.2
2.3
1.6
1.8
1.9
3.1
4.3
3.4
1.3
3.4
2.3
2.4
3.3
1.9
3.0
3.8
2.6
2.3
Min =1.3
Mean = 2.46 '
Sx = 0.8
7,3 (6,5)
8.5
8.5 (7.3)
8.5
8.5
8.5
8.5
8.5 (7.3)
8.5
8.5
7.3 (6.5)
7.3
8.5
8.5
8.5
8.5
8.5
7.3 (6.5)
8.5
7.3 (6.5)
8.5
8.5
8.5
30
35
19
30
26
24
40
20
23
29
38
46
56
20
44
32
29
42
29
41
40
34
34
Min
Mean
Sx
2.3*
2.6
2.0
2.3
2.1
2.0*
2.9
2.1
2.0
2.3
2.5*
2.8*
4.0
1.9
3.1*
2.4
2.3*
2.6*
2.3
2.6*
2.9*
2.5*
2.5
= 1.9
= 2.48
= 0.46
(2.0)
(1.6)
(1.8)
(2.2)
(2.3)
(2.2)
-P-
I
* Those cities with a TVP lower in April than in January, but still well above the minimum
in January.
NOTE: Numbers in parentheses are based on the RVP control program proposed in the NPRM
-------
Table 4-5
Safety Comparison of Gasoline TVP
In January and April Under EPA's Proposed Regulations
City Altitude
Correction
Factor
Albuquerque
Atlanta
Billings
Boston
Chicago
Cleveland
1.015
Denve r
Detroit
Kansas City
Las Vegas
Los Angeles
Miami
Minn/St. Paul
New Orleans
New York City
Philadelphia
Phoenix
St. Louis
San Antonio
San Francisco
Seattle
Washington D.C.
1.159
1.032
1.096
1.001
1.017
1.017
11.7
1.171
1.018
1.024
1.062
1.008
1.002
1.025
1.000
1.001
1.001
1.032
1.013
1.020
1.002
1.002
1.013
Jan 1988
MVMA Reg
Unleaded
Min RVP
12.6
11.9
13.6
10.0
14.1
14.7
8
12.4
13.4
13.1
12.3
12.4
9.9
14.0
12.6
11.9
13.3
12.2
13.1
12.2
12.6
12.7
11.9
Jan
0% min
Temp
-7
-2
-30
-12
-18
-18
2.1
-26
-13
-9
10
33
34
-28
14
0
-2
18
-8
8
29
1
5
Min
Mean
Jan TVP
1.8
1.7
1.0
1.0
1.4
1.5
8.5
1.1
1.5
1.6
2.5
3.9
3.0
1.1
2.6
1.8
1.9
2.9
1.7
2.2
3.6
2.0
2.0
= 1.0
= 2.00
April 1988 April
Controlled 0% min
RVP (-0.5) Temp
7.3 (6.5)
8.5
8.5 (7.3)
8.5
8.5
8.5
31
8.5 (7.3)
8.5
8.5
7.3 (6.5)
7.3
8.5
8.5
8.5
8.5
8.5
7.3 (6.5)
8.5
7.3 (6.5)
8.5
8.5
.8.5
25
30
12
18
19
12
2.4
7
14
23
32
43
50
4
40
24
' 25
39
23
33
38
31
25
Min
Mean
April TVP
1.8 (1
2.4
1.6 (1
1.8
1.8
.7)
-4)
1.5 Dallas
1.6 (1
1.6
2.0
2.2*(1
2.6*
3.5
1.3
2.9
2.0
2.1
2.5*(2
2.0
2.1*(1
2.8*
2.4
2.1
= 1.3
= 2.13
.3)
.9)
.2)
.8)
I
-p-
o
50
Sx = 0.78
Sx = 0.51
Those cities with a TVP lov/er in April than in January,
in January.
but still well above the minimum
NOTE:
Numbers in parentheses are based on the RVP control program proposed in the NPRM
-------
4-41
vehicle driveability with fuel volatility were not substantial
enough to be of concern from a vehicle operational perspective.
API also presented test data collected by Chevron on seven
1973-6 carbureted vehicles in an effort to show that even if
new vehicles can handle the lower volatility fuels, the older
open loop vehicles still on the road may not be able to handle
it. This testing was performed in 1983 on 6.5 and 8.5 psi RVP
fuels at 55°F and 75°F over the FTP cycle (not worst case
testing). The results of the testing showed 4 and 5 times more
engine stalls at 75°F and 55°F respectively on the the 6.5 psi
RVP fuel as compared to the 8.5 psi RVP fuel. Overall,
driveability demerits also increased by large percentages,
These apparently large increases, however, are misleading. The
overall magnitude of the driveability demerits and number of
stalls was small even with the low volatility fuel. In
addition, the vehicles were tested in their original condition
with the exception of the replacement of failed ignition and
emission parts, and setting of the ignition timing to the
manufacturer's specifications. If those vehicles not operating
well on the high RVP fuel are eliminated from the data set, the
remaining data shows only minor changes in vehicle driveability
resulting from the lower RVP fuel.
The last: and most recent report submitted by API on the
topic of cold temperature driveability consisted of testing on
51 1972 to 1988 -vehicles by Mobil Research. Testing was
performed in -early March in Maine at ambient temperatures of
21°F to 30°F on 8.6 and 13.4 psi RVP fuels as well as the fuel
in the vehicles when acquired (averaging approximately 13 psi
RVP). (Under the regulations contained in today's rule it is
highly unlikely that 8.6 psi RVP fuel will be distributed in
Maine during the month of March.) For the entire 51-car fleet,
there were on. average nearly twice as many start stalls, three
times as many driving stalls, and 2.5 times more heavy
hesitations and stumbles on the 8.6 psi RVP fuel than on the
13.4 psi fuel. The majority of this deterioration was due to a
"19-severe-car subfleet" which were defined as those vehicles
which had two or more heavy hesitations and/or stalls on the
low RVP fuel compared to the 13.4 psi fuel. Somewhat
expectedly, however, only one of the 19 vehicles in this
subclass was fuel injected. The 15 fuel-injected vehicles in
the study had only minor increases in starting times,
hesitations, and start stalls, and in fact as a group exhibited
better performance in all categories tested on the 8.6 psi RVP
fuel than the carbureted vehicles did on the 13.4 psi RVP fuel.
The same thing can be said for the 30 closed loop vehicles
as compared to the 21 open loop vehicles. Many of the vehicles
in this Mobil study, as well as those in earlier studies
discussed above, were apparently severely out of tune as they
did not operate acceptably (without stalling, heavy
hesitations, or excessively prolonged starting) even- on the
-------
4-42
13.4 psi RVP fuel. The report stated that the vehicles were
indeed tested in the as received condition. Lowering the
volatility of the fuel in this study appeared to increase the
occurrence of unacceptable driveability problems on a number of
vehicles. However, for properly maintained vehicles, and
especially closed-loop and fuel-injected vehicles, reduced fuel
volatility did not appear to have a major effect on cold
temperature driveability.
The studies on cold temperature driveability discussed
above appear to support the conclusion that any detriment
associated with volatility control to the levels proposed by
EPA in Chapter 2 of this RIA will seldom be unacceptable if
even noticeable to the average consumer. The driveability of
new vehicles, the majority of which are fuel-injected and
closed-loop controlled, appears to be affected very little by
volatility control of the magnitude proposed by EPA. Some,
older, properly maintained, carbureted vehicles appear to have
a slight driveability detriment at cold temperatures and very
low fuel volatilities. Fortunately, the detriment is typically
not large, and such combinations of temperature and RV? are
expected to occur only rarely, if ever, under EPA's volatility
control program (see the analysis below). Some extremely
sensitive vehicles, and especially poorly maintained vehicles
may encounter noticeably poorer driveability on lower RVP fuels
at cold temperatures. However, the preferable solution for
this problem is to tune-up the vehicle, not to operate it out
of tune on a high RVP fuel. Since the vast majority of
vehicles on the road will be closed-loop and/or fuel-injected,
by the time the second stage of EPA's volatility regulations is
proposed take effect in 1992, volatility control is not
expected to have any noticeable fleet-wide effect on
cold-temperature performance.
Statements by all of the major motor vehicle manufacturers
as well as a few 'others in their comments on the DRIA support
this conclusion. All of them agree that the level of
volatility control proposed by EPA is not likely to result in
any significant cold-temperature driveability detriment. Since
it is the motor vehicle manufacturers who will likely have to
respond to customer complaints if cold-temperature driveability
problems result from EPA's regulations, their comments serve to
strengthen our conclusions.
In an effort to further substantiate these conclusions,
the true vapor pressure (TVP) of fuels sold in January was
compared with the TVP of fuels during the summer .volatility
control season. The TVP of the fuel, since it incorporates the
operating temperature of the vehicle, is more accurate than RVP
alone in evaluating the cold-temperature driveability of
vehicles. It was assumed that oil companies currently produce
winter gasoline with a volatility which will insure acceptable
driveability during the month of January. MVMA's winter 1988
-------
4-43
gasoline survey was used to determine the minimum RV? fuel
currently being marketed in each .survey city in the month or
January.[16] These fuel volatilities were combined with
historical temperature data and altitude data corresponding to
those cities to yield estimates for the minimum TVP of the
fuels sold in those cities on some of the coldest'days of the
year.[17] (The three percentile minimum temperatures were
chosen for this analysis to insure evaluation of even the
coldest days of the year.) The TVP curves shown in Figure 4-3
were derived from an API nomograph and used for the TVP
determinations.[18] A similar analysis was then performed for
the volatility control season.
Since questions existed as to the length of the period
needed prior to the regulations effective date to insure
compliance, as well as the in-use RVP required to insure
compliance, the analysis was performed for the month of April
on fuel 0.5 psi below that required in May by EPA's proposed
regulations. The results -of this analysis, as shown in Table
4-4, indicate that even if fuel with an RVP below EPA's control
limits is distributed throughout the month of April, the TVP of
this fuel for the majority of the cities surveyed is above the
TVP for current fuels in January. For those cities where this
is not true, the TVP in April is still well above the minimum
for all cities in January.
As a result of this analysis, it can be said that if a
vehicle currently maintains adequate driveability during the
month of January, it should also maintain adequate driveability
throughout the period of RVP control including likely phase-in
periods. Thus, this analysis serves to support the conclusions
expressed above (based on the test data); that is, that the
level of volatility control proposed by EPA should not result
in any unacceptable increase in cold temperature driveability
problems. Even if a negative driveabifity impact could be
quantified such that a cost to society could be associated with
it, it is likely that, as for the hot-temperature driveability
improvements discussed earlier, the cost would not be large
enough to be of significance to the cost effectiveness of EPA's
volatility control regulations.
3. Cold Temperature Low Volatility Fuel Safety
In addition to comments on vehicle driveability, a number
of comments were also received which raised the issue of fuel
tank -explosivity at cold temperatures on reduced volatility
fuels, In general, fuel volatilities greater than 9.0 psi RVP
were said to provide an adequate level of safety, but
volatilities lower than this were said to result in potential
problems. Various oil companies provided comments stating that
fuels meeting the proposed" 1992 volatility standard may form an
explosive vapor/air mixture during cold temperatures early or
late in the period of control, and that a phase-in period of up
-------
FIGURE 4-3
GASOLINE TRUE VAPOR
PRESSURE vs. TEMPERATURE
(Extrapolated below 0°F)
-70 -60 -50 -40 -30 -20 -10
0 10 20 30 AO 50 60 70 80 90 100 110 120
TEMPERATURE (°F)
-------
4-45
to 2 months and a compliance margin of at least 0.5 psi serve
to compound the problem. The most severe problems were said to
be with fuel delivery trucks, oil storage tanks, and during the
event of refueling.
API provided a study which determined that the restriction
in the fill pipe of current gasoline vehicles is not typically
adequate to prevent a flame from an external ignition source
from propagating down the fill pipe, while Phillips commented
that today's in-tank fuel pumps represent another possible
ignition source. Therefore, if a flammable fuel/air mixture
does exist in the fuel tank, there are potential ignition
sources. On the other hand, NESCAUM, in their analysis of fuel
volatility control for the Northeast states, determined the
safety hazard concerns of the oil industry 'to be
unsubstantiated. The auto manufacturers apparently do not
consider cold, temperature fuel safety to be a problem either,
as they provided no comments on the topic, despite the fact
that they would likely be the parties facing liability claims
if it were a problem.
The potential of a fuel tank explosion on low RVP fuels is
a serious matter, and therefore needs to be addressed
carefully. Little actual data was presented by the oil
companies showing for what areas of the country and for what
times of the year the potential existed for an explosive
fuel/air mixture to exist in fuel tanks. In an effort to
analyze the situation, historical temperature data for a number
of cities across the United States were combined with fuel RVP
information (as in the previous analysis) to yield TVPs for the
situations of current fuels in January, and controlled fuel in
summer. For this analysis, the zero percentile minimum
temperatures were selected, since any chance of a fuel tank
explosion, even on the coldest day of the year, may be
considered unacceptable. The results are shown in Table 4-5.
As can be seen, the TVP of the fuel is typically higher in
April with controlled volatility fuel than it is in January
with current fuels. For those cities for which it is lower,
the TVP is still well above the minimum for all cities. As a
result of this analysis, it can be said that the fuel tank
explosion potential is less in April with controlled RVP fuels
than in January with current winter fuels. Since the maximum
concentration of gasoline in air which is still flammable is
approximately 7.6 volume percent,[19] which translates into a
TVP of approximately 1.1 psi, it can be seen that for a few
cities in the country in.January, there currently may be the
possibility for fuel tanks to become explosive. However,
during the period of volatility control including a phase-in
period, tanks are very unlikely to reach the flammable range,
The situation of vehicle refueling is slightly different
in that air is mixed with the fuel, and equilibrium is not
necessarily achieved above the fuel. This allows explosive
-------
4-46
mixtures to form at higher temperatures. However, since
vehicles are refueled with fuel from underground storage tanks,
the fuel temperature of concern (20°F to 30°F according to API)
should never be reached and flammability should not be a
problem.
A very recent report by NIPER for Phillips Petroleum
Company quantified some of the effects of low RV? on tank vapor
flammability at low temperatures.[20] Phillips' concern
centers on having fuel tanks with low-RVP fuel in the cooler
transition months before and after the summer ozone season.
The NIPER study and additional comments from Phillips claim
that flammability with low-RVP fuel may be reached already at
temperatures in the range of 10°-12°F, which is slightly higher
than our earlier analysis predicted. General Motors has also
responded to the Phillips/NIPER report, questioning the study
and its results.[21]
Some deficiencies in the NIPER report are worth noting.
First, the fuel chosen was not representative of in-use fuels.
Relying on a base gasoline with 3.5-4.7 psi RV? and adding
butane to reach final volatilities of 6.5-9.4 psi deviates from
commercial practice and could distort the results. The
composition of the fuel and the unusual 40°F fuel temperature
increase during driving may have together contributed to the
degree of weathering observed, which was greate-r than the
amount of weathering which appears to be predicted from
existing data. The disproportionate quantity of butane in the
fuel would also probably increase the likelihood of reaching a
flammable condition, since the upper flammability limit of
butane is considerably higher than that for other components of
gasoline (8.5 psi vs. approximately 7.6 psi for a typical
gasoline vapor). Also, the method of sampling might bias the
results toward more likely flammability. The vapors were drawn
from the very top of the tank where the mixture is leanest.
Pulling 2 liters of vapor out of the tank may even cause some
dilution of the sample with ambient air. Finally, Phillips'
graph of the upper flammability limit superimposed on NIPER's
results appears arbitrary and does not closely follow the
plotted points; a curve based on true vapor pressure would
predict somewhat lower temperatures and would appear to be more
consistent with NIPER's data.
Phillips used their graph to derive a critical temperature
for each volatility class. They interpolated the weathering
data to find weathered RVP and found the corresponding
temperature on their graph, which they expect to be the lowest
temperature that assures safe vapor mixtures. To facilitate
comparison we simply took the same RVP values for weathered
fuel and, based on the TVP graph (Figure 4-3) and an upper
flammability limit of 1.1 psi (7.6 volume percent), determined
critical temperatures for each class of fuel (Table 4-6).
To investigate the possibility.of in-use fuel being in the
flammable range, EPA. expanded the analysis of cold temperature
-------
4-47
Table 4-6
Critical Temperatures
Fresh Weathered Phillips
Class RVP RVP Temp.
A 7.0 6.7 15 12
B 7.8 7.1 13 8
C 9.0 7.6 12 5
Based on NIPER's weathering data.
-------
4-48
fuel safety to include historical minimum temperature data for
the continental U.S. in April and September for states that
might reach the necessary low temperatures. Before April we
would not expect vehicles to have fuel that already fully meets
the May volatility requirements, even considering early
deliveries to terminals to meet volatility requirements of
certain locations.[3] Also, by the beginning of October most
or all terminals should have at least their first shipment of
higher volatility fuel, which would sufficiently raise the
volatility of subsequent terminal sales and reduce the
likelihood that low-RVP fuel will still be in use. We used
three-hour minimum temperatures to determine a range of daily
minimum fuel temperatures across each state.[17] The
three-hour minimum temperatures were chosen to represent what
fuel tanks actually experience, although absolute minimums are
only 2-3°F colder. A representative range of sites from each
state included the coldest site for which data was available.
Applying Phillips' conclusions to September conditions,
conservatively basing each state's fuel volatility on the
summer's minimum volatility in July, leads to the conclusion
that no state would be expected to experience temperatures that
would cause a flammability condition (Table 4-7). Applying
Phillips' conclusions to April conditions shows that 3 percent
of the days at selected sites in Michigan, Minnesota and
Montana may reach a flammable condition; EPA's lower critical
temperatures significantly decrease the likelihood of
flammability, although they fall short of complete assurance
(Table 4-8). However, this analysis was .performed using
NIPER's estimates for weathering and upper flammability limit.
Using more conventional data for weathering and upper
flammability limits leaves a very small probability that any
site would reach flammable conditions on any day in April.
The NIPER report further supports EPA's position that
there should be no flammability problem with summer volatility
control, especially since true vapor pressures are not at their
lowest until the uncontrolled winter season. Thus, regardless
of summer volatility control to the levels proposed by EPA, the
worst-case condition will remain uncontrolled winter fuels in
the winter season.
4 . Hot Temperature Fuel Safety
In addition to comments on cold temperature, low-RV? fuel
safety, comments were also received on hot temperature,
high-RVP fuel safety. These comments, however, have already
been addressed in the FRIA for the first phase of volatility
controls and will not be reassessed here. The conclusions of
that analysis were that while reducing the volatility of in-use
gasoline should serve to reduce the problems of fuel spurting
and fuel system overpressurization, it is not possible on the
basis of existing data to determine to what extent historical
-------
4-49
Table 4-7
Historic September Minimum
Temperatures for Selected States
3-Hour Minimum Temperatures
State
OR
NV
UT
CO
KS
MO
VA
MD
MI
MN
MT
ND
ME
0 percentile
- 21-37
15-57
29-39
21-38
33-42
38-41
37-52
37-47
20-33
24-29
17-29
20-26
28-29
3 percentile
29-43
26-57
35-44
27-42
37-48
44-47
43-56
43-52
34-39
27-38
27-33
29-33
30-37
50 percentile
41-54
38-69
52-58
41-57
52-63
59-62
60-69
59-64
47-53
43-52
42-47
40-48
45-52
Phillips*
Critical Temp.
13°F
13°F
15°F
13°F
13°F
13°F
13°F
13°F
12°F
12°F
12°F
12°F
12°F
EPA**
Critical Temo.
8°F
12°F
3°F
8°F
3°F
8°F
8°F
5°F
5°F
5°F
5°F
5°F
* Uses Phillips' curve through NIPER's plot of data/ July
RVPs weathered according to NIPER data.
** Uses EPA TV? analysis, July RVPs weathered according to
NIPER data.
-------
4-50
Table 4-8
Historic April Minimum
Temoeratures for Selected States
3-Hour Minimum Temperatures
State
OR
NV
UT
CO
KS
MO
VA
MD
MI
MN
MT
ND
ME
0 percentile
15-34
12-35
17-28
-2-24
12-22
20-29
25-31
26-27
3-16
-3- 9
-10-16
-2- 2
3-14
3 percentile
16-36
19-41
23-32
15-30
22-30
28-32
30-39
32-35
10-25
11-21
12-30
15-19
16-24
50 percentile
31-44
32-54
35-45
28-42
36-46
44-43
47-51
45-48
30-38
28-35
30-34
31-34
30-35
Phillips*
Critical Temp.
12°F
13°F
12°F
12°F
12°F
12°F
12°F
12°F
12°F
12°F
12°F
12°F
12°F
EPA**
Critical Temp,
8°F
5°F
5°F
5°F
5°F
5°F
5°F
5°F
5°F
5°F
5°F
5°F
* Uses Phillips' curve through NIPER's plot of data, May
equivalent emissions RVP weathered according to NIPER data.
** Uses EPA TVP analysis, May equivalent emissions RVP
weathered according to NIPER data.
-------
4-51
hot-temperature safety problems would have been avoided, or
substantially reduced in severity, if fuel volatility. had been
lower. Neither is it possible to project safety problems into
the future. As a result, no evaluation of possible benefits
due to RVP control was attempted.
VI. Enforcement Cost of Volatility Regulations
The costs of several different enforcement options were
presented in the Draft RIA, Costs for the enforcement options
evaluated were estimated to range from $0.3 to $2.3 million per
year., EPA does not expect that actual enforcement costs will
differ significantly from these levels for either Phase I or
Phase II. Since these enforcement costs are small relative to
the other elements of RVP control (such as refinery costs,
etc.), the cost of enforcement was not included in the cost
effectiveness calculations of Chapter 5.
-------
4-52
References (Chapter 4)
1. 52 Federal Register, p. 31274, August 19, 1987.
2, "Draft Regulatory Impact Analysis, Control of
Gasoline Volatility and Evaporative Hydrocarbon Emissions From
New Motor Vehicles," U.S. EPA, OAR, QMS, July 1987.
3. "Final Regulatory Impact Analysis and Summary and
Analysis of Comments on the NPRM, Interim Control of Gasoline
Volatility," U.S. EPA, OAR, OMS, January 19, 1989.
4. "Petroleum Supply Monthly," various issues, May
through November 1989, DOE/EIA - 0109.
5. "Assessment of Impacts on the Refining and Natural
Gas Liquids Industries of Summer Gasoline Vapor Pressure
Control," prepared by Bonner and Moore Management Science for
EPA, August 24, 1987.
6. "MVMA National Gasoline Survey, Summer Season,"
Motor Vehicle Manufacturers Association, Sairoling Date-July 15,
1987.
7. "Annual Outlook for Oil and Gas 1989,." DOE/EIA -
0517(89).
8. "Octane Week," February 12, 1990.
9. "Petroleum Marketing Annual, 1985, Volume 2," Energy
Information Administration, Office of Oil and Gas, U.S. DOE,
DOE/EIA - 0487 (85)/2, December, 1986.
10. "The Butane Industry: An Overview and Analysis of
the Effects of Gasoline Volatility Control on Prices and
Demand," Jack Faucett Associates report for U.S. EPA, May 30,
1985.
11. Letter from George J, Yogis, Manager, Refinery
Economics and Business Development, ARCO, to Joe Somers, U.S.
EPA, March 23, 1987.
12. Letter from Kevin Hyatt, ENRON Gas Liquids, to Phil
Carlson, U.S. EPA, April 4, 1990.
13. "Calculation of Potential Oxygenated Use Under
Several Possible Conditions," EPA memo from Jonathan Adler to
Phil Lorang, OAR, OMS, ECTD, TSS, February 23, 1990.
14. "Monthly Energy Review" U.S. Department of Energy,
August 1989, DOE/EIA-0035
-------
4-53
15. "Alcohol Outlook," December 1988.
16. "MVMA National Gasoline Survey, Winter Season,"
Motor Vehicle Manufacturers Association, Sampling Date-January
15, 1988.
17.. "A Predictive Study for Defining Limiting
Temperatures and Their Application in Petroleum Product
Specifications," John P. Doner, Coating and Chemical Laboratory
for the U.S. Army, November, 1972, NTIS tt AD 756 420.
18. "Technical Data Book, Petroleum Refining," American
Petroleum Institute, Vol I, 1977.
19. "The Transport of Methanol by Pipeline," U.S.
Department of Transportation, April 1985.
20. Letter with enclosures from Richard I. Robinson,
Phillips Petroleum Company, to William Reilly,• EPA, March 12,
1990, including "Effect of Temperature and Gasoline Volatility
on the Flamraability of Vapor Spaces," William F. Marshall,
National Institute for Petroleum and Energy Research, February,
1990.
'21. Letter from James Pasek, GM, to Richard D. Wilson,
U.S. EPA, with attachment, May 18, 1990.
-------
CHAPTER 5
ANALYSIS OF ALTERNATIVES
I. Background
A. Introduction
Several aspects of the overall proposed RVP control
program make it a very attractive option compared to other
approaches to ozone control. The absolute reductions in VOC
available are larger than any other single program now
available. The program is feasible, and costs, while
sign-ificant, will not likely be discernible from typical price
fluctuations by most consumers. A gasoline RVP program is
further attractive because the costs can be limited to the
summer months, when ozone is a problem. Another attractive
aspect of RVP control is its immediate, total effect on
emissions from all gasoline-powered vehicles of all ages, and
conditions, as well from gasoline-related stationary sources.
In addition to considering these factors, SPA has
performed analysis of the cost effectiveness of today's RV?
control program. EPA has commonly used cost effectiveness
(dollars per ton of emissions reduced) as one tool for
assessing how alternative approaches to control compare to one
another as well as how a control program compares to other
related programs (in this case, VOC control programs). EPA
presents cost-effectiveness results merely to provide
additional comparative information; these results should not be
interpreted as establishing a baseline for cost effective
standards in any context.
For volatility control, it is most useful to evaluate the
incremental cost effectiveness (or the cost effectiveness for
the final step of control) rather than the overall cost
effectiveness. This is due to the fact that as RV? is reduced,
costs increase and emission reductions decrease. Therefore,
the cost-effectiveness value for the total RVP reduction of the
program could theoretically be favorable while the value for
the last increment of control is not. To avoid underestimating
this value, EPA used an incremental cost effectiveness
calculation in the NPRM and in the Phase I final rule; the
Agency will continue to use this approach for this rulemaking.
B. Synopsis of NPRM Methodology
For the Volatility and Evaporative Emissions NPRM, EPA
determined the cost effectiveness of various alternative
combinations of vehicle-based and fuel-based evaporative
emission control programs. The various options included
lowering in-use fuel volatility (for the range of RVPs from 8.0
to 11.5 psi in 0.5 psi increments), matching certification and
-------
5-2
in-use Class C fuel RVPs, and performing sensitivity analyses
using two crude oil prices (per barrel) and with and without
onboard refueling controls. For each option, a C/E range was
calculated based on different assumed values for improved fuel
economy resulting from increased fuel energy density (R-values
from 0.82 to 0.95) .
The cost-effectiveness (C/E) model used in the volatility
NPRM calculated incremental C/E values for each level (0.5 psi
increment) of volatility control. The C/E value of each level
could then be compared to that of the preceding level and to
that of other control programs. Thus, the effect of each
degree of control could be shown.
In obtaining the C/E value, the model calculated a cost
and an emission reduction for each increment. The C/E value
was then the ratio of these two in dollars per ton VOC
reduced. The cost values included both refinery costs and
costs due to vehicle change, Credits were included to account
for control in attainment areas ($250 per ton VOC reduced, as
discussed below), improved driveability, improved fuel economy,
and the utilization of captured evaporative emissions (see
Chapter 4). Costs were calculated based on the assumption that
controls would be in place for the five-month period of May
through September. Emission reductions were calculated for
nonattainment areas over the course of a whole year
(multiplying the five-month reductions by 12/5), thus making
these C/E values comparable to the C/E values of other VOC
control programs which count year-round reductions.
Since the control options were both vehicle- and
fuel-based, analyses were performed for both the short and the
long term. Short-term analyses looked at each of six different
years from 1988 to 2000. The long-term analysis was done for
the year 2010 and assumed that all fleet turnover due to
vehicle changes would be complete by then. As both the short-
and long-term analyses were steady-state in nature (assuming
total fleet turnover . had already occurred), a 33-year
discounted analysis was also performed. This analysis was an
attempt at properly weighting start-up costs of the control
options (i.e., the cost of implementing vehicle controls while
a portion of the fleet is still uncontrolled).
C. Summary and Analysis of Comments
All comments received in response to the NPRM have been
addressed in the Final Regulatory Impact Analysis for Phase I
of volatility control. These comments are briefly summarized
below.
1. Basic Model
One issue of comment was the method used to adjust this
program's cost effectiveness to make it comparable to other
-------
5-3
yearround VOC control programs. Commenters objected to the
adjustment of emission reductions only, not making costs also
yearround.
As was explained in the Phase I FRIA, this adjustment was
made to correct the discrepancy in what the cost effectiveness
is measuring. Because the aim is to reduce summertime ozone
(since that is when over 90 percent of exceedances occur), cost
effectiveness should be measured for the ozone season only.
Since other control programs have historically counted ozone
reductions outside the -ozone season, this volatility control
program needed to expand its emission reductions to a yearround
value in order to be comparable to the other C/E values. This
is equivalent to adjusting other analyses to credit only summer
reductions. The correction used here, however, was more direct.
A second area of comment was the $250/ton credit given for
emission reductions in attainment areas. As stated in the
Phase I FRIA, although some commenters believed this value to
be unjustified, others said an even higher value is
appropriate. EPA has not established $250 per ton over any
other value as appropriate for such reductions. In fact, the
total benefits of reducing ozone levels (e.g., less damage to
crops, materials, and forests), as well as additional direct
benefits of reducing VOC and particulate matter are very likely
in excess of $250 per ton. As a conservative valuation of
these benefits, the value of $250/ton will be retained for this
analysis.
NRDC suggested an alternative method of giving credit to
benefits in attainment areas. Their method was to include
transport regions and borderline nonattainment areas into the
fraction of the country which is currently in nonattainment.
Emission reductions would then be credited in those areas as in
nonattainment areas. Since the transport work being done by
EPA was not yet completed, we felt that the method chosen by
EPA to give credit to attainment areas was the more justified
of the two methods. Therefore, the approach was not changed.
Texaco commented that a' factor must be added to the
cost-effectiveness value to account for fuel weathering between
the marketed fuel and that fuel actually used in the engine.
Since this factor was incorporated into the evaporative
emission factors used in calculating emission reductions, it
did not need to be corrected for again.
2. Comparative C/E Value
The $2,000 per ton guideline suggested as a rough
indicator of acceptable cost effectiveness of VOC control also
received a number of comments. Some commenters claimed that
there was no support for such a guideline. NESCAUM, on the
other hand, pointed out that many other control measures in the
-------
5-4
Northeast have costs much higher than this guideline. In a
study done by OPPE, the average cost of emission control
programs to achieve additional emission reductions above what
could be achieved with the currently available control options
was in the range of $2,000 to $10,000 per ton.
There continues to be no benchmark for "reasonable" cost
effectiveness used by the Agency; it is often of interest,
however, to compare programs among one another using C/E as one
indicator, and thus the analysis has been again performed.
3. Sensitivity of C/E to Various Factors
All comments dealing with the' various inputs to the C/E
model, and not the model itself, were dealt with elsewhere.
NRDC commented that sensitivity runs should be done using
assumed ozone standards of 0.08 and 0.10 ppm. However, since
the volatility program is cost effective even at the current
standard, these runs would not add significantly to the
rulemaking and so were not done.
OMB suggested that the C/E for Class A and B areas be
provided separately from Class C areas. These class-specific
values were determined and provided in the Phase I Final RIA as
well as in Section II.D. below for Phase II.
4. Alternative Programs
Although several comments were received providing various
modifications or substitutions for the regulations proposed by
EPA, they lacked sufficient analysis to justify any deviation
by EPA from its proposals. When partial analyses or results of
analyses were presented, it did not appear that the
alternatives proposed would result in an equal or greater level
of VOC control at an equal or lower cost.
D. Synopsis of Phase I Final Analysis
The Final Rule for Phase I of volatility control gave
limits on the RVP of gasoline for the 1989 through 1991 time
period. It required volatility levels of 10.5 psi in Class C
areas, 9.5 psi in Class B areas, and 9.0 psi in Class A areas.
The classification of states was similar to that of ASTM but
with a few exceptions (mentioned elsewhere in the FRIA). The
cost-effectiveness analysis was similar to the methodology
proposed in the NPRM. However, since it was a short-term
analysis, no 33-year cost-effectiveness analysis was
performed. Also, no vehicle costs were included as these were
applicable to the onboard controls and excess evaporative
emissions portions of the proposal which are now separate
rulemakings. The resulting cost-effectiveness values for Phase
I of volatility control were $236/ton nationwide, $165/ton for
Class C areas, $576/ton for Class B areas, and $0/ton for Class
-------
5-5
A areas since the promulgated RVP level for these areas was
greater than the RVP level which A areas previously had.
II. Phase II Volatility Control
A. Summary of Methodology
The cost-effectiveness analysis for Phase II of volatility
control is very similar in method to that described in the
NPRM. However, no 33-year analysis was done. This is due to
the fact that a long-term analysis is needed when dealing with
fleet turnover (i.e., vehicle changes). However, since the
onboard and excess evaporative emission/running loss issues are
now separate rulemakings, no vehicle costs, changes or turnover
are involved. Cost-effectiveness values were, therefore,
calculated for the single year of 1995 only. The methodology
of the C/E calculation is briefly described below.
Emission reductions (incremental and total) were
calculated (as described in Chapter 3) for high ozone days and
also for summer average temperatures.
Costs were calculated based on the roughly 5-1/2 month
period from mid-March to the first week in September (since, on
average, refiners must begin production of the controlled fuel
six weeks early and can end production one week early as
described in Chapter 2). Costs from Chapter 4 (in dollars per
barrel) were used in the calculation of refining costs.
Credits were then taken for both increased fuel economy and for
the fuel "recovered" due to less evaporation of the fuel. The
fuel recovery credit is equal to the "summer average"
evaporative emission reductions converted to gallons of
gasoline (from gallons of butane as described in Chapter 5 of
the DRIA) and then converting the gallons of gasoline to
dollars based on a value of $.82 per gallon (as described in
Chapter 5 of. the DRIA). The fuel economy credit is calculated
from the summer gas fuel consumption, increased by the percent
increase in fuel economy due to lower volatility fuel, and then
converted to dollars using the same $.82 per gallon value as in
the fuel recovery credit.
The net cost was, then, the refining cost less the two
credits for fuel economy and fuel recovery. A final adjustment
was made to credit the program for emission reductions in
attainment areas. A value of $250 per ton of VOC reduced was
credited in attainment areas, as discussed above.
The cost effectiveness was, thus, the net cost less the
attainment area credit, divided by the yearround nationwide
high ozone day nonattainment area emission reductions outside
the Northeast. Cost-effectiveness values were calculated on a
nationwide and class-specific basis for both the overall Phase
II control (RVP dropping from Phase I to Phase II levels) and
-------
5-6
for the last incremental step of control (9.5 to 9.0 in C
areas, 8.2 to 7.8 in B, 7.4 to 7.0 in A).
B. Summary and Analysis of Comments
None of the comments received on the cost-effectiveness
analysis dealt solely with the Phase II portion of the
volatility control program. Therefore, all the comments were
addressed in the Phase I final rule. When considering Phase II
control, all responses to these comments remain unchanged.
C. Inputs for C/E Calculations
As mentioned above, C/E values were calculated in a manner
similar to that of Phase I. Some of the inputs into the
calculations, however, have changed. These changes are
described below.
No onboard refueling controls or improved vehicle
evaporative emission controls are now assumed since these are
being addressed separately from RVP control. Also, although
various crude oil prices and R-values were examined in the
NPRM, Phase II used a $20/barrel crude oil price and an R-value
of 0.85 (see Chapter 4). Also, when taking into account the
transition time needed to get the controlled fuel in place, a
5-1/2 month period was used in calculating the costs and
emission reductions of this control program in place of the
5-month period assumed earlier.
Finally, because several Northeast states have already put
regulations for 9.0 RVP gasoline into effect (beginning in the
summer of 1989), these states and others affected by their
programs have been excluded from the analyses of costs and
emission reductions. Their exclusion from the analysis is also
reflected in the class-specific nonattainment area VMT (vehicle
miles traveled) fraction and the fuel consumption fractions in
each class, both of which are used in calculating class
specific cost-effectiveness results.
D. Cost-Effectiveness Results
The results of this analysis show that this second
level of volatility control continues to be very a very
attractive approach to VOC control. The values obtained for
emission reductions, net cost, attainment area credit, and the
overall cost-effectiveness values are shown in Table 5-1.
These value are presented both for the entire nation and on a
class-specific basis. For comparison, the nationwide values
for Phase I are presented when using this same method.
-------
5-7
Table 5-1
Cost-Effectiveness Calculation Values
Phase II:
Nationwide
Class A
Class B
Class C
Year-Round
Non-Northeast
Nonattainment
Emission Reductions
(thousand tons)
713
35
390
287
Attainment
Area
Net Cost* Credit
(million $) (millions)
229
7
129
89
99
18
55
32
Cost
Effectiveness
($/ton)
183
-302
189
197
Phase II
(Last Increment)
Nationwide
Class A
Class B
Class C
132
5
71
56
112
5
61
47
20
2
9
9
699
607
721
673
Phase I:
Nationwide
452
63
69
-14
Includes refining cost, fuel economy credit, and fuel recovery credit.
-------
CHAPTER 6
CONTROL OF VOLATILITY OF ALCOHOL-BLENDED FUELS
I. Background
In the volatility proposal, three main options for the
treatment of gasohol were suggested: 1) exempting gasohol
from RVP control, 2) granting a permanent 1.0 psi RVP exemption
above that of gasoline for gasohol, and 3) requiring gasohol's
RVP to equal that of gasoline. Three suboptions to this third
option were also proposed. They were: 1) applying this
requirement nationwide, 2) applying this only in nonattainment
areas, granting a 1.0 psi RVP allowance in attainment areas, or
3) delaying this requirement until 1993, providing a temporary
1.0 psi RVP allowance in the interim. EPA also suggested the
possibility of providing a permanent or temporary allowance to
methanol-blends if it is granted to gasohol.
In the final rule for the first phase of volatility
control, a temporary 1.0 psi RVP allowance was provided for
gasohol. This allowance would continue until the final rule
for Phase II was promulagated, at which time a permanent
decision on the allowance would be made. Methanol blends
received no special treatment for the Phase I program.
For the second phase of volatility control, the comments
received in response to the NPRM have been summarized and are
presented below along with EPA's response to the comments.
These comments are divided into those relating to air quality
issues and those relating to economic issues.
II. Ethanol Blends
A. Air Quality Related Issues
1. Summary and Analysis of Comments
A number of comments dealt with the potential air quality
impacts of the various options. Some stated that no allowance
should be given unless vehicles are able to pass the normal
exhaust and evaporative emission test using the higher
volatility fuel. Also, although gasohol production fills only
a small portion of the total market, it is responsible for a
much larger portion of the market in some states. Several
commenters on this topic agreed that an RVP exemption or
allowance should only be considered on the basis of health and
environmental benefits. To accomplish this, they believe a
study should be done to determine the total effects an
allowance or exemption would have on the environment. A few
commenters suggested that an allowance should only be given if
-------
6-2
gasoline volatility were lowered even further to compensate for
the emissions from the gasohol.
Since gasohol is currently used as an oxygenate to help
attain the CO standard in the winter months (e.g., in
Colorado), there is some concern as to how this rulemaking will
affect the winter use of gasohol. In response to this, Amoco
stated that since this rulemaking is only for the summer, it
will not impact mandated wintertime use of oxygenates. Another
commenter suggested that gasohol use be targeted for the winter
months when the exceedances of the CO standard occur.
Recent studies have indicated that the impact on ozone of
a 1.0 psi RVP allowance for gasohol is less than the Agency
earlier believed. As discussed in Chapter 3, the effect of
gasohol on ozone levels is affected by factors including the
lower relative reactivity of ethanol and the effect of ethanol
on CO emissions (which also contribute to ozone formation).
The most recent modeling performed by SAI for EPA shows that
for 50 percent use of gasohol in St. Louis with a 1 psi
allowance, ozone levels increase by 0 to 0.8 percent.
Therefore, EPA believes that allowing a 1 psi RVP allowance for
ethanol blends would likely result at most in a modest effect
on ambient ozone levels.
B. Economic Issues
1. Summary and Analysis of Comments
Comments regarding the economic implications of the
gasohol RVP control options came from three main groups: those
in favor of a permanent allowance, those in favor of a
temporary allowance, and those opposed to any allowance. There
were also some comments received from parties who had alternate
suggestions on how to deal with gasohol. All of these comments
are summarized below.
Those in favor of a permanent allowance claimed that
without an allowance, ethanol blending would cease, since
special lower-RVP gasoline for blending would not likely be
produced by refiners. Added costs would include distributing
(commingling in pipelines would be a problem and truck, rail,
or barge would likely need to be used, as- the batches are not
large enough for pipelines), storing (many terminals currently
do not have the storage capacity for an additional grade of
product), and the cost of lowering the RVP of the base
gasoline. The end of ethanol blending would also be the end of
a much-needed market for surplus corn, which is now used as a
feedstock in ethanol production. Some refiners also stated
that they would not be able to supply sub-RVP gasoline as
blendstock due to limitations in the refinery operations.
Finally, many commenters stated that not having an allowance
would be inequitable, since ethanol blenders could bear an
added cost of testing the final fuel for volatility.
-------
6-3
Commenters in favor of a temporary allowance advocated
extending the allowance to 1993 when the federal tax credit
expires. According to MVMA, this would give enough lead time
to develop sources of low-RVP base fuel. Conservation Law
Foundation stated that there is no need to extend the allowance
past 1993, since splash-blended gasohol is not expected to be
profitable without the tax credit.
Some of those opposed to any allowance for gasohol felt
that the consequences of no allowance would not be significant
enough to hurt the ethanol industry. SOHIO emphasized that
there is no evidence to show that corn growers would be aided
by the allowance or hurt by its absence. AMOCO commented that
the cost of compliance is only 5-10 percent of the existing
subsidy (the cost being less than one cent per gallon with
subsidies of 8-10 cents per gallon) . API also argued that if
the demand for a lower RVP fuel was there, the market would
supply it.
Suggestions for different alternatives were received from
various comrnenters. A common request was to regulate the
volatility of the base fuel only, in part to eliminate
inequitable extra testing of blends. An additional requirement
could limit the content of ethanol in gasohol for an assurance
that its volatility would not exceed the volatility of gasoline
by more than 1 psi. Sinclair and the National Automobile
Dealers Association (NADA) suggested having a 1-psi allowance
only in ozone attainment areas. However, the National Petroleum
Refiners Association objected to the idea of requiring refiners
to produce fuel that must meet additional specifications
particular to blended fuels.
A number of refiners commented that they would have
difficulty and would likely not provide sub-RVP fuel for use in
gasohol production. This would make it very difficult or
impossible for distibutors to supply gasohol that meets RVP
regulations without an allowance.
If ethanol did not continue in use as a direct blend with
gasoline, alternate markets might be possible. One such future
ethanol market might be in the production of ethyl tertiary
butyl ether (ETBE). Because of its high octane and low
volatility, its potential for production and blending at the
refinery arid for transport by pipeline, and the recent
extension to ETBE of the ethanol blenders' tax credit, ETBE
production may expand. However, such expansion would
presumably be in competition with production of methyl tertiary
butyl ether (MTBE), since isobutylene capacity limits the
production of any butyl ether. The outcome of emerging
competition between ethanol and methanol marketers for access
to the isobutylene for production of ETBE and MTBE is not
certain, but downward pressure in alcohol prices can be
expected. In that case there is a good chance that methanol
would maintain its market, since its prices are more flexible,
-------
6-4
whereas ethanol producers may have to leave the market rather
than lower their price. An additional factor is the current
substantial dependence of methanol producers on the MTBE
market, which raises the stakes for that market and introduces
the possibility of an anti-ethanol strategy. In any case ETBE
is by no means certain to become a market for ethanol
substantial enough to replace the current market for directly
blended ethanol.
The Brazilian need for ethanol as an auto fuel would also
be a potential market, especially since sugar prices in Brazil
are currently inflated. However, this alternative probably
represents at best a limited and short-term market.
The alternative of regulating the volatility of the base
fuel only, with or without a cap on the ethanol content, is
problematic. Once the gasoline is blended, it becomes
impossible to test the volatility of the base fuel. The
alternative of granting an allowance only in attainment areas
would require a massive shift in the gasohol market, largely
from urban to rural areas, and would be by no means certain to
replace the entire lost ethanol market in ozone nonattainment
areas. In addition, EPA does not support a requirement to
produce special fuels for blends, preferring to set standards
and then allow the market to supply the resources for products
to meet them.
From the comments it is clear that there is at this time a
legitimate risk that the fuel ethanol industry would be
jeopardized in the absence of an RVP allowance for ethanol
blends. It appears likely both that the refining industry
would not supply low-RVP base fuel and that significant
alternative markets for fuel ethanol are not certain to develop.
Ill. Methanol Blends
A. Summary and Analysis of Comments
Several comments urged that methanol blends also be
treated differently from gasoline regarding RVP control.
Specifically, commenters suggested that an RVP allowance should
apply to methanol blends, particularly if such special
treatment is afforded to ethanol blends. Other commenters
opposed such an allowance for any blends,
The issues relating to methanol blends are different from
those relating to ethanol blends. Because of the nature of the
waivers granted several methanol blends, reduced RVP base fuel
has always been required when such blends have been marketed
(i.e., splash blending has never been permitted). Introducing
an RVP allowance would represent a significant departure from
current regulatory practice, one which has the potential to
harm air quality. There do not appear to be any compelling
economic or equity reasons for taking such action.
------- |