[[pp. 35479-35528]] Control of Air Pollution From New Motor Vehicles: Proposed Heavy-

Note: EPA no longer updates this information, but it may be useful as a reference or resource.

[Federal Register: June 2, 2000 (Volume 65, Number 107)]
[Proposed Rules]
[Page 35479-35528]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr02jn00-27]

[[pp. 35479-35528]] Control of Air Pollution From New Motor Vehicles: Proposed Heavy-
Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur
Control Requirements

[[Continued from page 35478]]

[[Page 35479]]

fuel sulfur levels higher than the proposed level were adopted. We also
recognize that technology evolution may affect the sulfur level at
which these technologies are enabled.
Therefore, we are evaluating whether or not the proposed program
could benefit from a future reassessment of the control effectiveness
of diesel NO_X exhaust emission control technologies and
associated fuel sulfur requirements. We would expect to conduct such a
reassessment in the 2003 timeframe, though we welcome comment on
whether such a reassessment will be needed and on the appropriate
timing for it. We also welcome comment on the extent to which a review
of NO_X control technology should also include a review of
the appropriate diesel fuel sulfur level for enabling the
NO_X control technology, including consideration of impacts
that a revised fuel requirement would have on PM control technology.
Another possible area for consideration during the reassessment could
be non-conformance penalties (NCPs) and the role they might play in
this program. NCPs would allow engine manufacturers to produce and sell
noncomplying engines under limited circumstances in exchange for paying
a penalty to the government. We welcome comment on the role NCPs may
play.
In conducting the review, we would expect to determine whether or
not there was a need to formally consider a change in the final
regulations adopted for this program. If such a change were determined
to be necessary, we would conduct a formal rulemaking, including
conducting public hearings.

I. Encouraging Innovative Technologies

We encourage comments on approaches that could provide increased
incentives for the development and introduction of clean advanced
engine technologies. Some such approaches have been suggested by
stakeholders or have been a part of other EPA rules. One of these would
be to develop a program for providing a special designation for engines
or vehicles that are significantly below the standards or use specific
innovative propulsion technologies. EPA finalized such a designation,
the ``Blue Sky Series Engine'' program, as a part of the 1998 nonroad
diesel standards final rule. Incorporating such a designation could be
very valuable for use in programs developed by states, municipalities,
or corporations to highlight or reward the purchase and use of
especially clean or innovative vehicles and engines. We request comment
on how we might structure a program like the ``Blue Sky Series''
program in the context of today's proposal, including what criteria we
should use to qualify an engine or vehicles for such a designation.
It has also been suggested that we might adapt the proposed ABT
program described in section VII.C. below to provide extra incentives
for manufacturers that encourage innovative technologies. For example,
manufacturers might get additional credits under the ABT program if
they introduce extra clean models or if they meet future standards
early. We believe our current ABT program, with the proposed revisions
discussed below, should encourage manufacturers to seriously consider
any technologies that can economically reach the very low emission
levels proposed today. Nevertheless, we request comment on the need for
and appropriateness of such additional provisions under the ABT
program.

IV. Diesel Fuel Requirements

As discussed in section III above, we believe that advanced exhaust
emission control technology exists and is being developed that can
reduce emissions of NO_X and PM to very low levels. However,
those exhaust emission control technologies will require changes to
diesel fuel in order to operate efficiently and reach the new engine
emissions standards we are proposing in today's NPRM. This section will
present our proposed changes to diesel fuel that are intended to enable
heavy-duty engines to meet our proposed new emission standards. We will
also describe the extent and applicability of the proposed diesel fuel
program, the means through which we expect refiners to meet the new
diesel fuel standards, and incentives we are providing refiners for
early introduction. The economic and environmental impacts of the
proposed diesel fuel program will be covered in subsequent sections in
combination with the implications of the proposed engine standards.

A. Why Do We Believe New Diesel Fuel Sulfur Controls Are Necessary?

In section III, we discussed our proposed finding that new
standards for heavy-duty engines can be established on the basis of
exhaust emission controls which we believe will be fully viable and
widely available for the 2007 model year. However, we also discussed
our understanding that those exhaust emission control technologies have
a significant and irreversible sensitivity to the sulfur content of the
fuel. Deep sulfur reductions are necessary to enable both the
NO_X and PM emission control technology that we believe
vehicles would need to use to achieve the emission standards we are
proposing today. Since we believe that new standards for heavy-duty
engines are an appropriate next step for reducing ambient pollution,
and it is these very exhaust emission control technologies which
manufacturers are likely to use in order to reach these low emission
levels, we are proposing to reduce the sulfur content of highway diesel
fuel.
Engine manufacturers and representatives of States, and
environmental and public health organizations have expressed general
support for a highway diesel fuel sulfur reduction strategy similar to
the gasoline sulfur reduction program. However, some stakeholders, in
particular refiners, have expressed concern that the sulfur sensitivity
of heavy-duty diesel exhaust emission controls has not been quantified
with a sufficient degree of certainty to provide a basis for setting a
specific low sulfur standard. Although it is likely that the efficiency
of exhaust emission control technology improves with decreasing fuel
sulfur levels all the way down to nominally zero levels, we believe
that it is possible to set a non-zero sulfur standard that sufficiently
enables high-efficiency control technology. The sulfur standard we are
proposing and the associated justification is described in more detail
in section IV.B below.
Sulfur appears to be the only diesel fuel property that must be
changed in order for the prospective exhaust emission control
technologies to operate effectively. Changes in other fuel properties,
such as cetane, aromatics, density, and high-end distillation, might
all provide small emission benefits for engines meeting our proposed
standards, but those benefits would be very small in comparison to the
sulfur standard. They would also not enable new advances in emission
control technology, and so would not likely produce significant step
changes in heavy-duty engine emissions. See section VI.B for a more
complete discussion of non-sulfur property changes for diesel fuel.
Finally, there is also an expectation on the part of some
automobile manufacturers that diesel engines will be used more
frequently in light-duty vehicles in the coming decade. However, any
light-duty diesel vehicles will be required to meet our final Tier 2
standards, which we believe will require the use of the same high
efficiency exhaust emission control technologies envisioned for heavy-
duty applications. Although we are not proposing a change to diesel
fuel specifically for light-duty diesel

[[Page 35480]]

vehicles, it is our expectation that the availability of a low-sulfur
fuel intended primarily to enable heavy-duty engines to meet our
proposed new standards would enable automobile manufacturers to produce
light-duty diesel vehicles that could meet the Tier 2 standards. We
would like comment on whether any other changes to diesel fuel
specifically for light-duty diesel vehicles are necessary, and on the
appropriateness, benefits, and costs of doing so.

B. What New Sulfur Standard Are We Proposing for Diesel Fuel?

We are proposing to require substantial reductions in diesel fuel
sulfur levels nationwide. Our proposal would require that all highway
diesel fuel produced or imported by refiners and importers be subject
to a maximum sulfur level of 15 ppm by weight. The technological need
for low-sulfur diesel fuel and the reasons for our proposed sulfur
standard are discussed in section III above. However, we are also
seeking comment on whether the sulfur standard should be set as high as
50 ppm or as low as 5 ppm, as well as what the associated costs and
benefits would be of a higher or lower level. (See section VI.B. for
further discussion of various sulfur standards.)
We believe our proposed diesel fuel sulfur program balances the
goal of achieving dramatic reductions in emissions from heavy-duty
vehicles with the goal of providing sufficient lead-time for the engine
emission control technology to develop and for the refining industry to
transition to a lower sulfur diesel fuel. Nevertheless, as noted
elsewhere, we are seeking comments on all these issues. We are aware of
diesel fuel industry concerns about their ability to consistently
deliver fuel meeting this low cap requirement. We are also aware that
some engine manufacturers are concerned that even fuel meeting the 15
ppm cap requirement may not adequately enable the exhaust emission
control technologies. In determining the appropriate sulfur level and
scope for our proposed program, we considered the implications of
diesel fuel sulfur on the emission control hardware of both heavy-duty
and light-duty vehicles (that is, light-duty diesel vehicles that are
required to meet our Tier 2 emission standards). Specifically, we
analyzed the degree to which the emission control devices described in
section III, above, may tolerate diesel fuel sulfur. We also evaluated
the environmental implications of sulfur control beyond the expected
NO_X and PM benefits (see section II) and the costs of
controlling fuel sulfur content, and we considered the ability of all
refiners and importers to meet the proposed diesel fuel sulfur standard
at essentially the same time (see section IV.D). We hope to benefit
from further discussion of all of these issues during the public
comment period.
The following sections describe in more detail the standard we are
proposing and the reasons why we are proposing a program that applies
year-round and nationwide.
1. Why Is EPA Proposing a 15 ppm Cap and Not a Higher or Lower Level?
There are five key factors which, when taken together, lead us to
propose that a diesel fuel sulfur cap of 15 ppm is both necessary to
enable the NO_X and PM exhaust emission control technology
(and thereby allow the proposed emission standards to be met), and
appropriate, taking into consideration the challenges involved in
providing low-sulfur fuel. These factors, as discussed in more detail
in sections III and IV.D, are the implications that sulfur levels in
excess of 15 ppm would have for the efficiency, reliability, and fuel
economy impacts of the exhaust emission control systems, and the
feasibility and costs of producing low-sulfur diesel fuel.
The efficiency of emission control technologies at reducing harmful
pollutants is directly impacted by sulfur in diesel fuel. Initial and
long term conversion efficiencies for NO_X, NMHC, CO and
diesel PM emissions are significantly reduced by catalyst poisoning and
catalyst inhibition due to sulfur. NO_X conversion
efficiencies with the NO_X adsorber technology in particular
are dramatically reduced in a very short time due to sulfur poisoning
of the NO_X storage bed. In addition total PM control
efficiency is negatively impacted by the formation of sulfate PM. The
formation of sulfate PM is likely to be in excess of the total PM
standard proposed today, unless diesel fuel sulfur levels are below 15
ppm.
The reliability of the emission control technologies to continue to
function as required under all operating conditions for the life of the
vehicle is also directly impacted by sulfur in diesel fuel. As
discussed in section III, sulfur in diesel fuel can prevent proper
operation and regeneration of both NO_X and PM control
technologies leading to permanent loss in emission control
effectiveness and even catastrophic failure of the systems. We believe
that diesel fuel with sulfur levels less than 15 ppm will be required
to provide a level of reliability for these technologies to allow their
introduction into the marketplace.
The sulfur content of diesel fuel will also affect the fuel economy
of vehicles equipped with NO_X and PM exhaust emission
control technologies. As discussed in detail in section III,
NO_X adsorbers are expected to consume diesel fuel in order
to cleanse themselves of stored sulfates and maintain efficiency. The
larger the amount of sulfur in diesel fuel, the greater this impact on
fuel economy. As sulfur levels increase above 15 ppm the fuel economy
impact transitions from merely noticeable to levels most diesel vehicle
operators would consider unacceptable (see discussion in section III).
Likewise PM trap regeneration is inhibited by sulfur in diesel fuel.
This leads to increased PM loading in the diesel particulate filter,
increased exhaust backpressure, and poorer fuel economy. Thus for both
NO_X and PM technologies the lower the fuel sulfur level the
better the fuel economy of the vehicle.
As a result of these factors, we believe that 15 ppm represents an
upper threshold of diesel fuel sulfur levels that would make these
technologies viable, and are therefore proposing to cap in-use sulfur
levels there. In comments received on the ANPRM, as well as in
subsequent meetings and discussions, however, we have often heard
different points of view on this issue expressed by the vehicle and
engine manufacturers, and by oil refiners.
Some vehicle and engine manufacturers have argued for a maximum cap
on the sulfur content of diesel fuel of 5 ppm, believing that this
level is necessary. As we discuss in section III, however, we believe
that a cap of 15 ppm (likely resulting in an in-use sulfur level 7 to
10 ppm) would be sufficient to ensure the reliability of PM exhaust
emission control technology (avoid potential for irreversible failure)
and enable it to reach the very high efficiencies needed over the wide
range of vehicle operation and conditions that would be needed for the
engines to comply with our proposed standards. Although at the current
stage of development, high efficiency NO_X technology is
extremely sulfur intolerant, work is already underway to develop
capability in the technology to tolerate at least some sulfur in the
fuel. As discussed in section III, however, it is likely that to
maintain the very high operational efficiencies of the emission control
equipment that we believe would be needed to meet the proposed emission
standards, and to avoid a significant fuel economy penalty, the sulfur
level in the fuel would still have to be very low.

[[Page 35481]]

We believe that requiring a cap lower than 15 ppm would not be
necessary to enable the exhaust emission control technology to meet the
very low NO_X and PM emission standards proposed. A cap lower
than 15 ppm would provide little additional emission reduction but
would increase the cost. Consequently, requiring a sulfur cap lower
than that necessary to enable the exhaust emission control technology
to meet the emission standards would be inappropriate. Further
discussion and analysis of alternative sulfur standards is contained in
section VI.
Conversely, many oil refiners have argued for a higher maximum cap
(if any) on the content of sulfur in diesel fuel, typically on the
order of 50 ppm. They argue that the cost of reducing the sulfur level
below a cap of 50 ppm (and average of 30 ppm) becomes prohibitively
high. They further argue that diesel engine exhaust emission control
technology is still in its infancy and will likely develop rapidly over
the next several years to the point where it is much less sulfur
sensitive than the technology of today. As discussed in section III, we
also believe that the diesel engine exhaust emission control technology
will develop rapidly over the coming years, and in particular are
projecting that the sensitivity of NO_X adsorber technology
to fuel sulfur will improve considerably through the development of
techniques to effectively regenerate themselves of stored sulfur
compounds. The Manufacturers of Emission Controls Association (MECA)
recently sent a letter strongly supporting this position, stating ``we
strongly believe that NO_X adsorber technology will be
commercially available in 2007 to help heavy-duty diesel engines meet
the stringent NO_X standards being considered by EPA and that
any current engineering challenges involved with this technology will
be addressed provided that very low sulfur fuel is available.'' \127\
Based on available information and our projections from that
information, we believe that a cap higher than 15 ppm sulfur, and in
particular a cap as high as 50 ppm would not enable the exhaust
emission control technology needed to achieve the proposed emission
standards and furthermore may severely compromise the reliability of
the systems and result in unacceptable fuel economy impacts. In
addition, as discussed in section IV.D below, although we acknowledge
that the cost to desulfurize diesel fuel does increase with more
stringent sulfur levels, we believe that these costs would not be
prohibitively high, and maintain that the environmental benefits of the
program are sufficient to justify the costs of the program at a sulfur
cap level of 15 ppm.
---------------------------------------------------------------------------

\127\ Letter to Carol Browner, Administrator of EPA from Bruce
Bertelsen, Executive Director of Manufacturers of Emission Controls
Association, May 3, 2000.
---------------------------------------------------------------------------

Based on our assessment of the efficiency, reliability, and fuel
economy impacts of sulfur on diesel engine exhaust emission control
technology, and the cost and feasibility factors associated with
reducing the sulfur content of diesel fuel, we propose to adopt 15 ppm
as the appropriate sulfur cap. However, we have analyzed the impacts on
technology enablement, costs, and benefits from controlling fuel sulfur
to a 15 ppm average level with a 25 ppm cap, as well as from capping
fuel sulfur at 5 ppm and 50 ppm. These levels have been put forward by
various stakeholders as either necessary (in the case of a 5 ppm cap)
or adequate (in the case of a 50 ppm cap) for enabling high-efficiency
diesel exhaust emission controls, and so we believe that assessments of
these levels is appropriate. These assessments are discussed in section
VI.B. We request comment on the appropriate level of the highway diesel
fuel sulfur standard, and on our assessment of alternative standards.
2. Why Propose a Cap and Not an Average?
We are proposing a cap on the sulfur content of diesel fuel in
order to protect the vehicle aftertreatment technologies that we expect
would be used to meet the proposed standards for heavy-duty engines and
vehicles. An average standard by itself would not be sufficient to
ensure that sulfur levels higher than those that could be tolerated by
the exhaust emission control technology would not be used in vehicles
for extended periods of time. Consequently, we do not believe that an
average standard can stand by itself and would at minimum have to be
coupled with a cap.
3. Should the Proposed 15 ppm Cap Standard Also Have an Average
Standard?
Although our current 500 ppm sulfur limit for diesel fuel provides
no averaging flexibility, in the years since that limit was set our
motor vehicle fuel regulations have frequently incorporated provisions
allowing regulated industries to average regulated parameters around a
standard, often with a capped upper limit. In fact this approach was
taken in the recently promulgated control of gasoline sulfur levels, in
which we adopted a 30 ppm average level with an 80 ppm cap.
Despite the ability of averaging provisions in some programs to
increase compliance flexibility and in some cases reduce overall costs
while still achieving the environmental objectives, we are not
proposing such provisions for the diesel fuel sulfur standard we are
proposing today. Basing the fuel program around an average sulfur level
could risk failure in meeting the whole objective of sulfur control
(the enablement of sulfur-sensitive technologies) and thereby the
environmental objectives of the program, or else could require the
adoption of a cap so low as to make the average level largely
irrelevant. The exhaust emission control technologies enabled by diesel
sulfur control appear to be far more sensitive to and far less
forgiving of variations in fuel sulfur level than advanced Tier 2
gasoline technologies. Enough is known about the exhaust emission
control technologies to convince us that the proposed sulfur level will
likely represent an enablement threshold level, above which increases
in emissions and potentially system failures could be expected.
Consumption of diesel fuel with sulfur levels above this threshold
could be very problematic.
Some commenters who responded to our diesel fuel ANPRM did express
interest in an averaged fuel sulfur standard, but only from the
viewpoint that the flexibility provided by averaging is generally
desirable, and not with specific solutions to the above-discussed
problems created by this approach. Other commenters opposed an
averaging requirement due to the test burden associated with
demonstrating compliance under such a program. We request specific
suggestions on how to structure a viable averaging requirement in
conjunction with a 15 ppm cap, and whether it would be desirable to do
so. One benefit of having only a cap instead of an average is that it
allows for a simplified enforcement scheme. Imposing an average
standard in addition to the cap would require additional product
sampling, recordkeeping, and reporting requirements to demonstrate
compliance with the standard. Thus, depending on how the program is
structured, the flexibility of an average standard may not be worth the
additional cost and complexity that would result, particularly with a
cap set at 15 ppm.
Some have suggested that it may be possible to set an average
standard of 10 ppm coupled with a higher cap. They

[[Page 35482]]

suggest that a 10 ppm average would achieve essentially the same
average in-use sulfur level as the proposed 15 ppm cap, and that as
long as the cap is sufficiently protective of the exhaust
aftertreatment technology, then the refining and distribution systems
may have greater flexibility in complying with the standard, allowing
for lower costs and less potential for disruptions of fuel supply. We
request comment on whether it would be possible to have a higher cap as
long as the average remained essentially unchanged and if so, what cap
would be appropriate. If such an approach could enable the technology,
we seek comment on the extent to which it would help address the
concerns refiners have raised with very low sulfur levels with respect
to the potential for fuel shortages and price increases.
If an averaged fuel sulfur standard were to be adopted (at any
sulfur level), one added flexibility option that has been suggested to
facilitate it is an averaging, banking and trading program. Because we
believe that the exhaust emission control devices would require ultra-
low sulfur diesel fuel, this flexibility would be focused on the
average component of the standard, rather than on the cap component.
Refineries would have the option to average across batches, to bank
credits for use in the future, and to purchase credits from other
refineries. In addition, under this concept the Agency could offer
additional ``average credits'' at a predetermined price to refineries.
This could provide more certainty about the cost of complying with the
average component of the standard by establishing a ceiling price on
these tradable and bankable credits. These credits could be used for a
refinery to comply with the average requirement; however, refineries'
use of these credits would still be subject to the cap standard. We
request comment on the concept of an averaging, banking, and trading
program in the context of an average standard, including: (1) whether
the additional flexibility of offering additional ``average credits''
at a predetermined price would benefit refineries; and, (2) what the
appropriate predetermined price for EPA-offered ``average credits''
should be.
4. Why We Believe Our Diesel Fuel Sulfur Program Should Be Year-round
and Nationwide
We believe it is necessary for all highway diesel fuel to meet the
proposed 15 ppm sulfur limit at all times. To relax this requirement
would jeopardize many of the environmental benefits of the proposed
program. Although NO_X benefits are only realized in the
summer, PM and air toxics benefits are realized year-round. Moreover,
the exhaust emission control devices require low-sulfur diesel fuel
year-round. The use of highway fuel with a sulfur content greater than
our proposed sulfur standard could damage the emission control
technology of 2007 and later model year vehicles and engines. Once
vehicles are equipped with the new exhaust emission control devices,
they can only be fueled with the low-sulfur fuel. This precludes any
consideration of a seasonal program. In addition, because diesel
vehicles travel across the country transporting goods from region to
region and state to state, low-sulfur diesel fuel will have to be
available nationwide (see discussion in section VI.C. for possible
exceptions. The health effects associated with diesel PM emissions are
not area-specific, nor are the adverse effects of high sulfur diesel on
engines with exhaust emission control. For these reasons, we do not
believe that any regional or seasonal exemptions from the proposed
sulfur requirements would be practical.

C. When Would the New Diesel Sulfur Standard Go Into Effect?

Since the need for low-sulfur diesel is dictated by the
implementation of new engine standards, the proposed sulfur standard
would become effective commensurate with the introduction of the first
heavy-duty engines meeting our proposed standards. As described in
section III.H, the phase-in of the engine standards is proposed to
begin with the 2007 model year. Since light-heavy-duty trucks might be
introduced as early as January 2 of the previous calendar year but are
often introduced beginning about July 1, we are proposing that all
highway diesel fuel sold at retail stations and wholesale purchaser-
consumers meet the proposed sulfur standard by June 1, 2006. We believe
that this one month lead time will be sufficient to provide confidence
that the fuel available for purchase on July 1 will comply with the
proposed sulfur cap. We are also proposing that highway diesel fuel at
the terminal level be required to meet the proposed sulfur standard as
of May 1, 2006, and that highway diesel fuel produced by refiners (and
imported) meet the proposed sulfur standard by April 1, 2006. We
believe these earlier compliance requirements at terminals and
refineries would be necessary to provide an orderly transition to low-
sulfur fuel and to avoid the market disruptions that occurred when the
sulfur level of diesel fuel was lowered to 500 ppm in 1993 with only a
retail compliance date. The three months between April and July should
allow sufficient time for fuel to move through the distribution system,
for existing tankage to transition down to the lower sulfur level that
would be required. It would also ensure that all fuel is complying with
the proposed sulfur standard and is available for use in heavy-duty
engines when 2007 model year engines are introduced to the market. We
request comment on this proposed approach.
We believe that the lead-time issue is particularly important,
because not only would failure to meet the standards at the retail
level cause emission increases from new technology vehicles, but
violations of the standard due to insufficient turnover in the
distribution system could potentially permanently disable the emission
control systems of new technology vehicles and could cause driveability
problems for the operators of such vehicles. We would like to take
comment on these dates for the start of our low-sulfur diesel program,
and in particular on whether the three-month lead time is more than
adequate, adequate, or less than adequate for an orderly transition.
Some parties have suggested that low-sulfur diesel should be
required at the same time as low-sulfur gasoline, in 2004. They point
out that refinery synergies are optimized when refiners are forced to
address both requirements at the same time instead of sequentially. The
earlier introduction of low-sulfur diesel would also provide both
reductions in sulfur dioxide and sulfate PM emissions for the in-use
fleet prior to 2007, and would give engine manufacturers greater
flexibility to make use of sulfur-sensitive technologies such as cooled
EGR.
We do not believe that it is appropriate to require all on-highway
diesel fuel to meet our proposed sulfur standard prior to the
introduction of heavy-duty engines meeting our proposed standards. By
proposing a 2006 start year for the low-sulfur diesel program, we are
giving refiners a long lead-time to begin the planning process for
meeting our proposed requirements. They always have the flexibility to
make a single set of refinery changes prior to 2004 that will allow
them to meet both the low-sulfur gasoline and our proposed low-sulfur
diesel requirements by 2004. Although we are not requiring it, we would
encourage the introduction of highway diesel fuel that meets the
proposed sulfur standard prior to 2006, as discussed in section IV.F.
Finally, some parties have suggested that low-sulfur diesel is
necessary by 2004 to ensure that light-duty vehicles

[[Page 35483]]

can meet our Tier 2 standards using diesel fuel. Although some analysts
have predicted a greater proportion of diesel-powered light-duty
vehicles in the coming decade, we do not believe that they can justify
the introduction of low-sulfur diesel prior to 2006. As discussed in
more detail in section VI.A.2, we believe diesel-powered light-duty
vehicles will not actually need low-sulfur diesel fuel prior to 2006,
given the flexibility offered by the Tier 2 program's bin structure. It
would also appear that light-duty vehicles would not produce lower
emissions using lower-sulfur diesel fuel than they would using
gasoline, since all light-duty must meet the same Tier 2 standards.
There would be no emission benefits associated with introducing low-
sulfur diesel fuel prior to 2006, for use in light-duty vehicles, and
thus it would be difficult to justify the costs. We welcome comments on
requiring low-sulfur diesel fuel prior to 2006 for use in light-duty
vehicles. We also welcome comments on the appropriateness of a 2006
start date for the diesel fuel sulfur standard.

D. Why We Believe the Proposed Diesel Sulfur Standard Is
Technologically Feasible

In addition to evaluating the merits of diesel powered highway
vehicles operating on low-sulfur diesel fuel, we also considered the
ability of refiners to reduce diesel fuel sulfur in essentially every
gallon of highway diesel fuel by mid-2006. Based on this evaluation, we
believe it is technically feasible for refiners to meet the proposed
standards and that it is possible for them to do so in the proposed
time frame. We are summarizing our analysis here and we refer the
reader to the Draft RIA for more details. We welcome comments on all
aspects of this analysis.
1. What Technology Would Refiners Use?
Conventional diesel desulfurization technologies have been
available and in use for many years. Conventional hydrotreating
technology involves combining hydrogen with the distillate (material
falling into the boiling range of diesel fuel) at moderate pressures
and temperatures and flowing the mixture through a fixed bed of
catalyst. EPA required refiners and diesel fuel distributors and
marketers to provide diesel fuel for highway vehicles which does not
exceed 500 ppm by weight in sulfur starting in October 1993. As a
result, most U.S. refiners installed diesel desulfurization units to
reduce their onroad diesel fuel from the pre-control average of about
3000 ppm, to the current average of about 350 ppm.
Based on our review of the literature and discussions with vendors
of catalyst technology and desulfurization technology, the most
difficult challenge to reducing sulfur to extremely low levels via
conventional hydrotreating is the presence of certain aromatic
compounds. These aromatic compounds are referred to as sterically
hindered, because the physical arrangement of the atoms of these
compounds hinders interaction between the sulfur atom and the
catalyst.\128\ One method to desulfurize these compounds is to design
the shape of catalyst surfaces so that these sterically hindered
compounds can more easily approach the catalytic material. Another
approach is to saturate one or more of the aromatic rings present,
which makes the sulfur atom more accessible to the catalytic surface.
---------------------------------------------------------------------------

\128\ Typical compounds which are difficult to desulfurize are
4-methyl, dibenzothiophene and 4,6-dimethyl, dibenzothiophene. The
methyl group(s) attached to the aromatic rings make it very
difficult for the sulfur atom to physically approach the catalyst,
which is essential for the desulfurization process to proceed.
---------------------------------------------------------------------------

Refiners produce diesel fuel from a variety of distillate blending
streams in the refinery. The largest component is straight run
distillate, which comes straight from crude oil, hence the name
straight run. The second largest component is light cycle oil (LCO)
which comes from the fluidized catalytic cracker, or FCC unit. This
unit primarily produces gasoline from material having a higher
molecular weight than either gasoline or diesel fuel, but also produces
a significant amount of distillate. About 62 percent of today's highway
diesel fuel contains some LCO. The third largest component is light
coker gas oil, which comes from the coker, which also produces lighter
molecular weight material from heavier material. Both straight run
distillate and light coker gas oil contain relatively low levels of
sterically hindered compounds. LCO contains a much higher concentration
of sterically hindered compounds. Thus, the difficulty of achieving the
15 ppm sulfur cap being proposed today is primarily a function of the
amount of light cycle oil (LCO) that a refiner processes into its
highway diesel pool.\129\
---------------------------------------------------------------------------

\129\ LCOs are not homogeneous and can vary dramatically in
chemical composition from refiner to refiner. The discussion here
applies to a typical LCO composition.
---------------------------------------------------------------------------

We project that all refiners would be technically capable of
meeting the proposed sulfur cap with extensions of the same
conventional hydrotreating which they are using to meet the current
highway diesel fuel standard. This extension would likely mean adding a
second stage of conventional hydrotreating. In a two-stage process,
hydrogen sulfide is removed from the treated distillate after the first
reactor and fresh hydrogen added prior to the second reactor. This
stripping of the hydrogen sulfide serves two purposes. First and
foremost, it reduces the concentration of hydrogen sulfide throughout
the second reactor. This speeds up the desufurization reactions
substantially. Second, it reduces the concentration of hydrogen sulfide
at the end of the second reactor. This is the point where hydrogen
sulfide can react with the treated distillate, forming new sulfur
compounds (essentially adding sulfur back into the fuel). This process
is termed recombination and low hydrogen sulfide concentrations
decrease it dramatically. Finally, reducing the concentration of
hydrogen sulfide increases the concentration of hydrogen, again
speeding up the desulfurization reactions.
Converting an existing one-stage hydrotreater into a two-stage
hydrotreater would involve adding an additional reactor, a hot hydrogen
sulfide stripper, modifications to the compressor to increase pressure
to the new reactor and possibly a pressure-swing adsorption (PSA) unit
to increase hydrogen purity. Essentially all of the units comprising
the existing hydrotreater would still be used.
We project that all refiners could utilize recently developed, high
activity catalysts, which increase the amount of sulfur which can be
removed relative to the catalysts which were available when the current
desulfurization units were designed and built. The cost of these
advanced catalysts is very modest relative to less active catalysts,
but they would significantly reduce the size of the new reactors
described above. We also project that refiners and technology vendors
could achieve the 15 ppm cap without significant saturation of aromatic
compounds. This will be achieved through the selection of catalysts and
through the control of operating conditions, particularly temperature.
The above projections are based primarily on information received
from a number of refining technology vendors, supported by published
literature, as no operating experience at sulfur levels below 10 ppm
currently exists with this technology on diesel fuel feedstocks typical
of U.S. refiners. All the vendors supplying information to EPA and
others studying diesel fuel desulfurization projected that the 15 ppm
cap can be met using diesel fuel

[[Page 35484]]

hydrotreaters which operate at hydrogen pressures ranging from 600-900
pounds per square inch (psi) and with total reactor volumes of roughly
2-3 times those of current diesel fuel hydrotreaters. A number of oil
refiners informed us that they believe that much larger reactors would
be required. API believes that both higher pressures and larger
reactors will be needed. Either change would increase our projected
costs (described in section V.D.1 below).
Based on our review of the literature, we do not believe that these
extremely large reactors would be required to meet the proposed sulfur
cap. However, 15 ppm sulfur diesel fuel is not yet being produced
commercially from feedstocks typical of the U.S. Thus, we request
comments on the sufficiency of 600-900 psi operating pressures for
diesel fuel hydrotreaters to meet the proposed sulfur cap. We also
request comment on the sufficiency of total reactor volumes which are
2-3 times greater than those currently being utilized under the 500 ppm
sulfur cap in order to meet a 15 ppm cap.
Other options are available to refiners. Some refiners could choose
to add an FCC feed hydrotreater. This improves the yield of high value
products from the FCC unit and reduces the sulfur content of both FCC
naphtha and LCO. FCC naphtha is the primary source of sulfur in
gasoline, for which EPA recently set stringent standards. However,
while hydrotreating the FCC feed reduces the sulfur content of the LCO
produced by the FCC unit, it can increase the concentration of
sterically hindered compounds. Also, FCC feed hydrotreating is much
more costly than distillate hydrotreating or ring opening technology.
Thus, we are not projecting that any refiners would utilize this
technology to meet the proposed diesel fuel sulfur cap.
Refiners could also add a hydrocracker to process their LCO if they
have not already done so. This would increase the production of high
value gasoline with a very low sulfur content. However, hydrocrackers
are very costly to build and operate, so a refiner choosing to do so
would likely do so for reasons beyond removing sulfur from diesel fuel.
In addition to these major technological options, most refiners
would also have to add other more minor units to support the new
desulfurization unit. These units could include hydrogen plants, sulfur
recovery plants, amine plants and sour water scrubbing facilities. All
of these units are already operating in refineries but may have to be
expanded or enlarged.
2. Are These Technologies Commercially Demonstrated?
As mentioned above, conventional diesel desulfurization
technologies have been available and in use for many years. U.S.
refiners have roughly seven years of experience with this technology in
producing highway diesel fuel with less than 500 ppm sulfur. Refiners
in California also have the same length of experience with meeting the
California 500 ppm cap on sulfur and an additional aromatics
standard.\130\ In order to meet both sulfur and aromatics standards,
refineries in California are producing highway and nonroad diesel fuel
with an average sulfur level of 150 ppm.
---------------------------------------------------------------------------

\130\ California allows refiners to use an engine test to
certify an alternative fuel mixture which meets or exceeds the NOx
reducing performance of a 10 volume percent maximum aromatics and a
500 ppm maximum sulfur diesel fuel.
---------------------------------------------------------------------------

Some refiners in Europe are producing a very low-sulfur, low
aromatics diesel fuel for use in the cities in Sweden (Class I Swedish
Diesel) using two-stage hydrotreating. This ``Swedish city diesel'' is
averaging under 10 ppm sulfur and under 10 volume percent aromatics.
While clearly demonstrating the feasibility of consistently producing
diesel fuel with less than 10 ppm sulfur from selected feedstocks,
there are a few differences between the Swedish fuel and typical U.S.
diesel fuel. First, the tight aromatics specification applicable to
Swedish City diesel fuel usually requires the use of ring-opening or
dearomatization catalysts in the second stage of the two-stage
hydrotreating unit. This eases the task of desulfurizing any sterically
hindered compounds present. Second, Swedish Class I diesel fuel also
must meet a tight density specification. This, coupled with the fact
that European diesel fuel contains less LCO than U.S. diesel fuel,
significantly reduces the amount of sterically hindered compounds
present in the feed to the desulfurization unit. Third, it is not clear
whether any refiner is producing a large fraction of their distillate
production to this specification. Thus, the European experience
demonstrates the efficacy of the two-stage process and its ability to
produce very low sulfur diesel fuel. However, doing so without
saturating most of the aromatics present and with heavier feedstock has
only been demonstrated in pilot plants and not commercially.
Europe has adopted a 50 ppm cap sulfur standard for all diesel fuel
which takes effect in 2005. Some countries, including England, have
implemented tax incentives for refiners to produce this fuel sooner.
The great majority of diesel fuel in England already meets the 50 ppm
specification. Refiners have reported no troubles with this technology.
This diesel fuel is being produced in one-stage hydrotreaters. However,
as mentioned above, European diesel fuel contains less LCO than diesel
fuel in the U.S., so the use of one-stage conventional hydrotreating to
meet very low sulfur levels is applicable, but not sufficient to
demonstrate feasibility in the U.S. Germany has also established a tax
incentive, but for diesel fuel containing 10 ppm or less sulfur. One
European technology vendor indicated that they have already licensed
two desulfurization units to German refiners planning to produce diesel
fuel to obtain this tax credit.
Overall, conventional diesel desulfurization ring-opening and
dearomatization technologies have all been installed and are operating
in one or more refineries. Thus, there should not be much concern among
refiners whether these technologies will work reliably in general.
Refiners' primary concern would be focused on the treatment of any LCO
currently being blended into highway diesel fuel. They would be
particularly concerned with the ability to desulfurize this material to
very low sulfur levels using conventional technology and, absent that,
ways to shift this material to other valuable fuel pools or treat it
more severely in available hydrotreaters or hydrocrackers. Of course,
refiners would also be concerned with the reliability of the technology
in complying with a 15 ppm cap day in and day out.
In addition to these more traditional technologies, Energy
Biosystems recently announced the availability of their
biodesulfurization technology for desulfurizing diesel fuel.
Biodesulfurization is a process which uses bacteria which has been
genetically enhanced to biologically remove the sulfur atoms from
petroleum compounds. This process is still being developed and is
expected to begin commercial demonstration in the next couple of years.
At the present time, the goal of the developers is to produce diesel
fuel with less than 50 ppm sulfur. It is not known whether this
technology would be capable of meeting the proposed cap of 15 ppm. This
process has the advantage of operating at ambient temperature and
pressures, and requires no hydrogen. The economics of the process,
however, rely on a market for its by-products, which may limit its
widespread application. Because of

[[Page 35485]]

uncertainties in this technology's ability to achieve the proposed 15
ppm cap, we did not factor it into our cost projections. We request
comment on the availability of this technology in the relevant time
frame for this proposed rulemaking.
3. Are There Unique Concerns for Small Refiners?
We have heard concerns that small refiners would bear
proportionately higher economic burdens if they were required to
produce diesel fuel meeting the same sulfur levels as larger
refineries. The most significant concern expressed to us has been their
more limited ability to obtain the capital necessary to make the
refinery modifications necessary to produce low sulfur diesel fuel
compared to the larger refiner. To address these and other concerns
related to small refiners, we have participated in a review and
evaluation process specific to small businesses under the Small
Business Regulatory Enforcement Flexibility Act (SBREFA). More
information can be found in our response to the Regulatory Flexibility
Act (see section XI.B). In short, we are seeking comment on provisions
that would assist small refiners in addressing unique challenges, as
discussed in section VIII.E.
4. Can Refiners Comply with an April 1, 2006 Start Date?
We believe that our proposal that the program begin on April 1,
2006 would provide more than an adequate amount of time for refiners to
plan their investment, complete the design package and complete the
construction and startup of the new or modified desulfurization unit
and other associated units in their refineries. In response to our
proposed Tier 2 gasoline desulfurization rulemaking, the American
Petroleum Institute (API) commented that 4 years is needed for refiners
to complete this cycle of planning, design, construction and startup.
While we believe 4 years to be more than sufficient, we have initiated
this rulemaking sufficiently early to provide over 5 years of lead
time. We recognize that most refiners will have to make investments in
their refineries to desulfurize their gasoline during this time, so the
additional time from final rule to implementation is expected to be
valuable for refiners. Similarly, by informing refiners now (i.e.,
before they make their gasoline desulfurization investments) of our
proposed highway diesel fuel desulfurization program we hope to allow
refiners to coordinate their investments and produce both low-sulfur
gasoline and low-sulfur onroad diesel at a lower cost. The additional
time between promulgation and implementation is important because of
the number of refiners which are expected to have to make these
investments. Unlike the gasoline sulfur program which really only
affected refineries outside of California, this program would affect
the California refiners as well, in addition to a number of refineries
which produce onroad diesel fuel but no gasoline.\131\ However, the
total capital cost of the investments projected to be required to meet
the proposed diesel fuel sulfur cap is less than that for the Tier 2
gasoline sulfur standards.
---------------------------------------------------------------------------

\131\ By far most of California gasoline meets a 30 ppm
averaging standard, except for a small volume which is exported out
of the state. However, since the California refiners already have
the desulfurization units in place to desulfurize the majority of
their gasoline, they are expected to use those same units to
desulfurize the exported gasoline as well.
---------------------------------------------------------------------------

A particular concern has been raised to the Agency regarding the
capability of the engineering and construction (E&C) industries to be
able to design and build diesel fuel hydrotreaters while at the same
time doing the same for gasoline, as well as accomplishing their other
objectives. We believe that the E&C industry is capable of supplying
the oil refining industry with the equipment necessary to comply with
the proposed diesel fuel sulfur cap on time.\132\ We believe that this
is facilitated by the extended phase-in we allowed regarding compliance
with the Tier 2 gasoline sulfur standards. For example, we project that
only roughly a third of all gasoline-producing refineries outside of
California will be building gasoline desulfurization equipment for
start-up in early 2006 and 2007. Thus, most of the construction related
to gasoline desulfurization will be completed prior to the proposed
implementation of the diesel fuel sulfur cap. Also, low sulfur gasoline
and diesel fuel standards scheduled for Europe and Canada become
effective in 2005. We believe that this precedes the proposed highway
diesel fuel sulfur cap sufficiently to enable the availability of
European equipment fabrication capacity to be available to meet the
needs of the proposed sulfur cap in the U.S. Thus, we do not foresee
any shortage in either E&C industry personnel or equipment fabrication
capacity. We request comment on these findings.
---------------------------------------------------------------------------

\132\ Rykowski, Richard A., ``Implementation of Ultra Low Sulfur
Diesel Fuel: Construction Capacity and Aggregate Capital
Investment,'' EPA Memorandum to the Record, Docket A-99-06.
---------------------------------------------------------------------------

We are aware that the National Petroleum Council (NPC) is
conducting a Refining Study which also addresses this issue. It appears
from a publically available draft final report that the NPC may
conclude otherwise. We plan to consider the findings of this study once
it becomes final.
Another issue related to the feasibility of the April 1, 2006 start
date relates to refiners' ability to hook up their new equipment to
their existing diesel fuel hydrotreaters while still providing the
nation with diesel fuel during the transition. This issue is relevant
since: (1) we expect most refiners to revamp their current equipment,
as opposed to building entirely new equipment and (2) all refiners face
the same April 1, 2006 deadline. We expect that any new equipment
required as part of the revamp would be able to be constructed on-site
while the current equipment is operating. Inter-connecting the new and
old equipment would occur prior to April 2006 when the current
hydrotreater is scheduled to be down for maintenance. Existing
equipment which would require modification, such as compressors and
heat exchangers, would be modified during this time, as well. Diesel
fuel hydrotreaters currently operate roughly two years in between
scheduled maintenance. Thus, there should be at least one and possibly
two scheduled maintenance periods between the time when refiners could
have project designs completed, permits issued, as appropriate, and
April 2006. Under this schedule of refinery maintenance, modifying
current diesel fuel hydrotreaters to meet the proposed sulfur cap
should not impact diesel fuel production. If refiners had to schedule
additional down time in order to complete the revamp, then diesel fuel
production could be affected. We expect that any such shortfall would
be made up by other refiners or the previous build-up of inventory. We
request comment on the ability of the industry to continue to supply
highway diesel fuel while it is modifying equipment in order to comply
with the proposed sulfur cap.
Concerns have also been raised with respect to the refining
industry's ability to raise the capital necessary to make the refinery
modifications necessary to meet a 15 ppm sulfur cap on diesel fuel,
while at the same time expending capital to reduce the sulfur level in
gasoline as a result of the recently promulgated Tier 2 standards. This
has led to concerns that some refiners may refrain from investing to
continue to produce highway diesel fuel, which could cause a shortage
when the program is implemented. As discussed in section IV.B. of the
draft RIA, we have designed these programs in a

[[Page 35486]]

manner which will serve to maximize refiner flexibility and minimize
costs. Furthermore, as discussed in section V.D.1., we believe that
despite the capital cost of desulfurizing their highway diesel fuel,
other options for marketing the distillate streams from their
refineries will be limited. Finally, as discussed in section VI.A., we
are also considering various phase-in approaches for implementing the
low sulfur diesel standard. A phase-in could help spread out the
design, construction, and capital expenditure of refinery modifications
necessary to comply with the proposed diesel fuel sulfur standard. We
request comment on the necessity and ability of a phase-in to address
these concerns.
In summary, we believe that meeting a 15 ppm cap is achievable with
the diesel desulfurization technologies available now. We are confident
that we are providing more than a sufficient amount of time between
when this rule is expected to be finalized and the proposed startup
date of the program. This timing should allow for a smooth transition
of low-sulfur fuel into the marketplace. We request comments on all of
these issues. In particular, we request comment and supporting
information on the challenges refiners would face in competing for
engineering and construction resources and obtaining capital for diesel
fuel sulfur control. We also seek comment with supporting information
on the potential for diesel fuel shortages at the beginning of the
program that some believe might result from individual refinery
decisions to shift all or a portion of their production to other
distillate products or export, and on the ability of the market to self
correct if a shortage does occur.
5. Can a 15 ppm Cap on Sulfur Be Maintained by the Distribution System?
The proposed cap on sulfur content would apply to on-highway diesel
fuel at the refinery gate, and at every point along the distribution
system through to the end-user. The current distribution system for
petroleum distillates currently carries products with sulfur contents
that range from 30 ppm to over 10,000 ppm. The system includes
pipelines, tankers, tanks, and delivery trucks. To date, this system
has not been required to deliver a product with the purity which would
be required under this proposal. Consequently, to ensure the sulfur
standard is not exceeded during the fuel's journey to the end-user, the
refiner would actually produce diesel fuel sufficiently below the cap
to account for its own compliance margin (estimated to be 7 ppm on
average), as well as for test variability and potential downstream
contamination. Under the current sulfur cap of 500 ppm, refiners
typically provide ample margin, producing fuel with roughly 350 ppm
sulfur. With a sulfur cap of 15 ppm, the absolute magnitude of the
margin refiners could provide would obviously be much smaller. In
addition, the impact of contamination in the distribution system would
be potentially much more severe. If the proposed 15 ppm cap on the
sulfur content of on highway diesel fuel were adopted, other products
in the distribution system such as nonroad diesel fuel would have
sulfur concentrations over 200 times that of highway diesel fuel
instead of the 10-fold factor at present. Additives to diesel fuel
added in small amounts downstream which sometimes contain high sulfur
concentrations levels may also become much more of a concern (see
section IV.D.6.c). If as expected, refiners would produce highway
diesel fuel with an average sulfur content of approximately 7 ppm to
comply with the proposed sulfur standard, and variability in measuring
diesel sulfur content is limited to less than +/-4 ppm, downstream
sulfur contamination would need to be limited to less than 3 ppm to
maintain compliance with the proposed 15 ppm cap. Petroleum marketers
and distributors have cautioned that the distribution system is
unfamiliar with limiting sulfur contamination to such a low level.
Current industry practices may need to be modified to control and
limit sulfur contamination in the distribution system. Current
practices which are critical to minimizing contamination and which may
need to be more carefully performed include:

--Properly leveling tank trucks to ensure that they can drain
completely of high-sulfur product prior to being filled with the
proposed diesel fuel.
--Allowing sufficient time for transport tanks to drain of high-sulfur
product prior to being filled with the proposed diesel fuel.
--Purging delivery hoses of higher sulfur product prior to their use to
deliver the proposed diesel fuel.

To adequately limit sulfur contamination, we believe that such
practices would need to be followed each and every time with adequate
care taken to ensure their successful and full completion. Some
distributors may find it necessary to conduct an employee education
program to emphasize their importance. We request comment on our
assessment for each segment in the distribution chain, including tank
trucks, tank wagons, rail tankers, barges, and marine tankers.
As discussed in section V.D.3 of today's document, there may be an
increase in distribution costs associated with an increase in pipeline
interface volumes and the need to sample and test each batch of on
highway diesel fuel at the terminal level for its sulfur content. There
could also be an increase in the occurrence of noncomplying fuel
showing up in the distribution system. As is the case today, this could
cause temporary, local market shortages of fuel meeting the proposed
sulfur cap. This off-specification fuel would also either have to be
downgraded to off-highway, or re-refined, though we have assumed that
the frequency of such occurrence would be low enough as to not impact
the costs of the program noticeably. The potential sources of sulfur
contamination in the distribution system, what controls we believe
would be necessary to ensure downstream compliance with the proposed
sulfur standard, and the costs associated with such controls are
discussed in more detail in the Draft RIA. We request comment on the
challenges that each segment of the distribution chain would face in
controlling sulfur contamination, on the extent that each segment might
reasonably be expected to limit sulfur contamination, and on the
associated costs.
6. What Are the Potential Impacts of the Proposed Sulfur Change on
Lubricity, Other Fuel Properties, and Specialty Fuels?

a. What Is Lubricity and Why Might It be a Concern?

Diesel fuel lubricity properties are depended on by the engine
manufacturers to lubricate and protect moving parts within fuel pumps
and injection systems for reliable performance. Unit injector systems
and in-line pumps, commonly used in heavy-duty engines, are actuated by
cams lubricated with crankcase oil, and have minimal sensitivity to
fuel lubricity. However, rotary and distributor type pumps, commonly
used in light and medium-duty diesel engines, are completely fuel
lubricated, resulting in high sensitivity to fuel lubricity.
Experience has shown that it is very rare for a naturally high-
sulfur fuel to have poor lubricity, although, most studies show
relatively poor overall correlation between sulfur content and
lubricity. Considerable research remains to be performed for a better
understanding of the fuel components most responsible for lubricity.

[[Page 35487]]

Consequently, we are uncertain about the impact of today's proposal on
fuel lubricity. Nevertheless, there is evidence that the typical
process used to remove sulfur from diesel fuel (hydrotreating) can
impact lubricity depending on the severity of the treatment process and
characteristics of the crude. If refiners use hydrotreating to achieve
the proposed sulfur limit, there may be reductions in the concentration
of those components of diesel fuel which contribute to adequate
lubricity. As a result, the lubricity of some batches of fuel may be
reduced compared to today's levels, resulting in an increased need for
the use of lubricity additives in highway diesel fuel.
Blending small amounts of lubricity-enhancing additives increases
the lubricity of poor-lubricity fuels to acceptable levels. At the
present time, it is believed that oil companies are treating diesel
fuel in this way on a batch to batch basis, when poor lubricity fuel is
expected. This practice of treating fuel on an as-needed and voluntary
basis has been effective in ensuring good diesel fuel lubricity for the
diesel heavy-duty vehicle fleet. Our review of the technical literature
\133\ indicates that the U.S. military also uses lubricity-enhancing
additives in its diesel fuel. The U.S. military has found that the
traditional corrosion inhibitor additives that it uses have been highly
effective in reducing fuel system component wear. Consequently, the
U.S. Army now blends MIL-I-25017E corrosion inhibitor additive to all
fuels when poor lubricity is expected, and regularly for Jet A-1, JP-5
and JP-8 fuels. We believe that this practice would continue, with some
portion of the fuel refined to the proposed standard being treated with
lubricity-enhancing additives. For a more detailed discussion of diesel
fuel lubricity and current industry practices, please refer to the
Draft RIA for this proposal. We have included a 0.2 cents per gallon
cost in our calculations to account for the potential increased use of
lubricity additives (see section V.D.2).
---------------------------------------------------------------------------

\133\ See the draft RIA for a more detailed discussion.
---------------------------------------------------------------------------

b. Voluntary Approach for the Maintenance of Fuel Lubricity

If action on fuel lubricity does prove necessary, we believe a
voluntary approach would provide customer protection from engine
failures due to low lubricity, while providing the maximum flexibility
for industry. In a voluntary approach we would encourage, but not
require, fuel producers and distributors to monitor and provide fuel
with adequate lubricity to protect diesel engine fuel systems. This
approach recognizes the uncertainties of measuring fuel lubricity, and
allows flexibility as research produces better information and improved
test methods. The voluntary approach discussed here would be a
continuation of current industry practices for diesel fuel produced to
meet the current Federal and California 500 ppm sulfur diesel fuel
specifications, and benefits from the considerable experience gained
since 1993. The advantage of this approach is avoidance of an
additional regulatory scheme and associated burdens. On the down side,
voluntary measures do not guarantee results. We believe the risk in
this case is small. Refiners and distributors have an incentive to
supply fuel products that will not damage consumer equipment. Even if
occasional batches of poor lubricity fuel are distributed, they would
likely be ``treated'' with residual quantities of good lubricity fuel
in storage tanks, tanker trucks, retail tanks, and vehicle fuel tanks
(even at very low treatment levels lubricity enhancing additives
provide significant protection; see the discussion in the Draft RIA for
this proposal). Further, we expect that the American Society for
Testing and Materials intends to address lubricity in its ASTM D-975
specifications for diesel fuel quality after its concerns about test
issues have been resolved.
We are asking for comments on the alternative of specifying minimum
fuel lubricity, and suggestions for the appropriate lubricity standard
and test method. Under this approach, we would require fuel producers
to monitor and provide minimum lubricity. This would be similar to the
approach of Canada and our understanding of the usage requirements of
the U.S. military. The advantage of this approach is to guarantee the
minimum quality of fuel in the market. On the down side, such a new
specification would need to be tied specifically to emissions or
emission control hardware, and we question whether such a requirement
is appropriate considering the uncertainty about the adequacy of the
existing test methods. The American Society for Testing and Materials
has declined to specify a lubricity standard in its ASTM D-975
specifications for diesel fuel quality until its concerns about test
issues have been resolved. Also, this approach would require an
enforcement scheme and associated compliance burden. Further, we
believe that this approach would probably not be significantly more
effective than the voluntary approach. Refiners and distributors have
an incentive to supply fuel products that will not damage consumer
equipment, and the U.S. commercial market has adequately addressed
similar concerns in the past.
The U.S. Department of Defense (DOD) expressed strong reservations
about the ability of the proposed voluntary approach to ensure adequate
fuel lubricity and requested that EPA establish a uniform requirement
to ensure that diesel fuel introduced into commerce has adequate
lubricity. Absent such a requirement, DOD related that the military
would face a considerable burden to ensure that highway diesel fuel
used in military vehicles provides sufficient lubricity. DOD stated
that since they rely on the commercial market to supply highway diesel
to military users and are currently experiencing lubricity problems in
certain parts of the country during the winter months, a reduction in
diesel sulfur would increase the risk and scope of lubricity problems.
DOD also stated that due to harsher operating conditions, engines used
in their vehicles (especially tactical vehicles) are more vulnerable to
lubricity problems than the same engines operated in commercial
vehicles. In addition, at some U.S. military installations DOD uses
highway diesel fuel in their off highway vehicles as well as their
highway vehicles. We request comment on the unique challenges that our
proposed voluntary approach would place on the military and on the
appropriate means to address DOD's concerns.

c. What Are the Possible Impacts of Potential Changes in Fuel
Properties Other Than Sulfur on the Materials Used in Engines and Fuel
Supply Systems?

With the introduction of low-sulfur diesel fuel in the United
States in 1993, some diesel engine fuel pumps with a Nitrile material
for O-ring seals began to leak. Fuel pumps using a Viton material for
the seals did not experience leakage. The leakage from the Nitrile
seals was determined to be due to low aromatics levels in some low-
sulfur fuel, not the low sulfur levels. In the process of lowering the
sulfur content of some fuel, some of the aromatics had been removed.
Normally, the aromatics in the fuel penetrate the Nitrile material and
cause it to swell, thereby providing a seal with the throttle shaft.
When low-aromatics fuel is used after conventional fuel has been used,
the aromatics already in the swelled O-ring will leach out into the
low-aromatics fuel.

[[Page 35488]]

Subsequently, the Nitrile O-ring will shrink and pull away, thus
causing leaks, or the stress on the O-ring during the leaching process
will cause it to crack and leak. Not all low-sulfur fuels caused this
problem, because the amount and type of aromatics varied. Although
manufacturers have apparently resolved this issue, and we have no
evidence that further desulfurization will cause further changes in O-
ring shape or other concerns, we request comments on this or other
potential impacts of fuel properties on the materials used in engines
and fuel supply systems.

d. What Impact Would the 15 ppm Cap Have on Diesel Performance
Additives?

Our proposal to limit the sulfur content of performance additives
used in diesel fuel to less than 15 ppm (see section VIII) would
require that the use of certain high-sulfur diesel fuel additives be
discontinued. Our review of EPA's Fuel and Fuel Additives database
indicates that alternative additives that perform the same function and
which do not contain sulfur are readily available. Our evaluation
suggests that discontinuing the use of the limited number of diesel
additives with a high sulfur content would not result in significant
increased costs or an undue hardship to additive and fuel manufacturers
(see the draft RIA). We request comment on the difference in price
between high- and low-sulfur performance additives and whether there
are differences in their efficiency. As an alternative to the proposed
15 ppm cap on the sulfur content of performance additives, we are
requesting comment on whether additives not meeting the 15 ppm sulfur
cap should be allowed to be added to diesel fuel downstream in de
minimis amounts, as long as the final blend still meets the 15 ppm cap.

e. What Are the Concerns Regarding the Potential Impact on the
Availability and Quality of Specialty Fuels?

The Department of Defense (DOD) has expressed concerns regarding
the potential impact of today's proposed rule on the availability and
quality of military fuels, especially the aviation fuels JP-5 and JP-8.
DOD is concerned that today's rule might reduce the number of
refineries that produce military fuels by limiting the slate of fuels
that refiners can economically produce or the number of refiners that
continue to produce military fuels. DOD notes that the special flash
point requirement for military JP-5 fuel already limits DOD's supply
base and that the proposed rule may make some refiners opt out of
manufacturing this speciality fuel, which would reduce supply
availability and increase costs. DOD also states that the increased
hydroprocessing severity and other refinery process modifications
necessary to meet the proposed sulfur standard could impact certain
chemical/physical characteristics that are part of their fuel
specifications. DOD relates that previous environmentally-driven
changes to gasoline and diesel specifications have caused a degradation
in the quality of the jet fuel. For example, DOD states that they have
noticed a reduction and continued decline in jet fuel stability.
DOD is also concerned that refiners that currently blend more than
10 percent light cycle oil (LCO) into their highway diesel fuel might
shift some LCO into off-highway distillate fuels. DOD relates that this
would adversely affect the quality of off highway fuels used by the
military such as their naval distillate fuel F-76. DOD states that they
have experienced quality problems with LCO component streams that were
not adequately hydrotreated causing a highly unstable finished product.
Storage stability is an important issue for DOD since military naval
fuel F-76 is often stored for extended periods (longer than six months)
and unstable LCO used to manufacture F-76 could compromise mission
readiness. The potential changes that refiners might make in the way
they process LCO streams and incorporate such streams into their slate
of distillate fuels is discussed in section V.D.1 and in the Draft RIA.
We believe that concerns related to the quality of specialty fuels
can continue to be addressed by actions taken by the manufacturers and
purchasers of such fuels without the need for intervention by EPA. We
also anticipate that demand for such fuels will be sufficient to
encourage their continued availability. We request comment on the
potential impact of today's proposed rule on the quality and
availability of specialty fuels such as those used by the U.S.
military, on what actions might be necessary to mitigate such impacts,
and on the associated costs. Comment is specifically requested on the
need for the military to modify its specifications and/or enhance
enforcement of these specifications to achieve their fuel quality goals
if the proposed sulfur standards are adopted, and on the costs
associated with such changes.

E. Who Would Be Required to Meet This Proposed New Diesel Sulfur
Standard?

As discussed earlier, the highway diesel fuel sulfur content
standard being proposed today is a per-gallon cap of 15 ppm. We believe
that heavy-duty diesel trucks subject to the standards we are proposing
today would require the consistent use of diesel fuel with a sulfur cap
of 15 ppm to avoid the potentially severe emission, performance, and
durability problems that arise from operation on higher-sulfur fuel. On
this basis we believe that the proposed sulfur standard should apply to
the diesel fuel at the point of sale to the ultimate consumer. In other
words, the proposed cap on sulfur content should apply at all points in
the diesel fuel production and distribution system, including the
retail level.
We understand that there are production and distribution practices,
such as blending of additives and winter viscosity improvers such as
kerosene or No. 1 diesel fuel, that could cause the sulfur level of
diesel fuel to vary as it travels from refinery to end-point consumers.
Along with concerns about contamination and test method
reproducibility, these issues suggest that we should include some sort
of tolerance along with our proposed sulfur cap. However, we are
concerned that such tolerances on top of the 15 ppm cap may not be
appropriate given the sensitivity of diesel exhaust emission control
technology to fuel sulfur above the proposed sulfur cap. In practice,
therefore, refiners will likely be required by the downstream
distribution system to produce diesel fuel having a sulfur content
significantly below the proposed sulfur cap to ensure that downstream
practices do not end up producing a retail-level fuel with sulfur
levels higher than the proposed maximum. Thus, all parties in the
distribution system, including refiners and importers, would be
prohibited from selling, storing, transporting, dispensing,
introducing, or causing or allowing the introduction of highway diesel
fuel whose sulfur content exceeds the proposed sulfur cap. The
advantage of such an approach is that, as downstream distribution
practices and sulfur measurement accuracy improves, refiners will be
able to reduce production costs by producing fuel closer to the
proposed sulfur cap. Alternatively, we could enforce the proposed 15
ppm sulfur cap at retail and enforce a lower cap at the refinery level.
This cap would likely have to be less than 10 ppm to allow for
downstream contamination, additive blending, and test method
variability.

[[Page 35489]]

However, we believe it is more appropriate to leave this tolerance to
the market.

F. What Might Be Done To Encourage the Early Introduction of Low-Sulfur
Diesel Fuel?

As discussed in section IV.C, we are proposing that the entire
highway diesel pool be required to meet a lower standard on sulfur
content beginning June 1, 2006.\134\ This should provide certainty that
low-sulfur diesel fuel will be available for model year (MY) 2007
heavy-duty diesel engines by July 1, 2006. If low-sulfur diesel fuel
was available prior to July 1, 2006, engine manufacturers have
indicated that fleet trials might be conducted of the sulfur-sensitive
exhaust emission control equipment intended for use in heavy-duty
vehicles to meet the proposed MY 2007 emissions standards. The
information gained from these trials could be used to improve the
efficiency and durability of such exhaust emission control equipment.
This could lower the cost of the exhaust emission control equipment and
help ensure the smooth implementation of the proposed MY 2007, heavy-
duty standards. If low-sulfur diesel fuel was available earlier than
July 1, 2006, it might also facilitate the early introduction of
sulfur-sensitive exhaust emission control equipment in light-duty
diesel vehicles. Automobile manufacturers expressed interest in using
sulfur-sensitive exhaust emission control equipment in some of their
light-duty vehicles beginning in MY 2004, so that they might benefit
from in-use experience prior to the anticipated use of such equipment
in all MY 2007, light-duty diesel vehicles. In addition, early
availability of some low sulfur diesel fuel would have the added
advantage of allowing the distribution system a chance to develop
experience handling diesel fuel with such a low sulfur level before the
standards would take effect.
---------------------------------------------------------------------------

\134\ This is the proposed retail-level compliance date. The
proposed compliance date at the refinery level is April 1, 2006.
---------------------------------------------------------------------------

We believe that some low-sulfur diesel fuel meeting the proposed 15
ppm sulfur cap would be available in advance of when we are proposing
that it must be produced by refiners. Most refiners will need to
install new equipment to meet the proposed sulfur standard. Since the
technical and construction resources needed for such refinery upgrades
is limited, a number of refiners are likely to have the new
desulfurization equipment installed well in advance of the proposed
compliance date. Refiners who produce low-sulfur diesel early would
want to market it as a premium fuel rather than losing the added value
by selling it as current highway diesel fuel. Some refiners have
already begun programs to market low-sulfur diesel as a premium fuel.
For example, ARCO Products Company recently announced a fleet program
to demonstrate the emissions benefits of its EC--D (emission control)
diesel which has a lower sulfur and aromatics content, and a higher
cetane rating than current highway diesel fuel.\135\ Engine and vehicle
manufacturers are assisting in the overall program design and
implementation of the program. Emission control equipment manufacturers
are supplying exhaust emission control equipment which works more
effectively with low-sulfur fuel. ARCO has also begun marketing diesel
fuel in California with a maximum sulfur content of 15 ppm. This fuel
is being made available, upon request, to operators of urban municipal
fleets retrofitted with catalytic exhaust emission controls in
connection with the California ARB's proposed urban bus program (see
section I.C.6). \136\ Mobil Corporation, Ford Motor Company, Navistar,
and Volkswagen also have a cooperative program underway to evaluate the
emissions benefits of new engine/aftertreatment technologies using a
lower-sulfur diesel fuel (also with reduced polynuclear aromatic
content). We are interested in encouraging additional programs between
refiners and vehicle manufacturers to introduce vehicles equipped with
exhaust emission control technologies which benefit from the use of
low-sulfur diesel fuel prior to the date when we are proposing that
such fuel must be made available.
---------------------------------------------------------------------------

\135\ ARCO Products Company news release dated October 7, 1999,
Docket A-99-06 Item II-G-13.
\136\ ARCO Products Company news release dated December 15,
1999.
---------------------------------------------------------------------------

There are numerous strategies involving voluntary market incentives
that could help promote the early introduction of low-sulfur diesel
fuel. Under existing voluntary emission credits programs, a system
might be created whereby refiners that produce low-sulfur fuel early
could generate emission reduction credits that could then be sold
through a market mechanism to other entities that could use such
credits to meet their emission compliance goals. We welcome comments on
whether additional incentives are needed and feasible to encourage the
early introduction of low-sulfur diesel fuel for use in vehicles
equipped to provide lower emissions with the use of such a fuel. We
also request comments on how such incentives might be structured under
a phase in of low sulfur highway diesel fuel (see section VI.A).

V. Economic Impact

This section discusses the projected economic impact and cost
effectiveness of the proposed emission standards and low-sulfur fuel
requirement. We welcome comment on the estimated cost for research and
development and the necessary lead time to develop these technologies
for heavy-duty vehicles. Additionally we invite the reader to review
all of the underlying cost assumptions made in the accompanying draft
RIA and ask for comment on the validity of these assumptions. Full
details of our cost and cost effectiveness analyses can be found in the
Draft RIA.

A. Cost for Diesel Vehicles To Meet Proposed Emissions Standards

1. Summary of New System and Operating Costs
The technologies described in section III show a good deal of
promise for controlling emissions, but also make clear that much effort
remains to develop and optimize these new technologies for maximum
emission-control effectiveness with minimum negative impacts on engine
performance, durability, and fuel consumption. On the other hand, it
has become clear that manufacturers have a great potential to advance
beyond the current state of understanding by identifying aspects of the
key technologies that contribute most to hardware or operational costs
or other drawbacks and pursuing improvements, simplifications, or
alternatives to limit those burdens. To reflect this investment in
long-term cost savings potential, the cost analysis includes an
estimated $385 million in R&D outlays for heavy-duty engine designs and
$220 million in R&D for catalysts systems giving a total R&D outlay for
improved emission control of more than $600 million. The cost and
technical feasibility analyses accordingly reflect substantial
improvements on the current state of technology due to these future
developments.
Estimated costs are broken into additional hardware costs and life-
cycle operating costs. The incremental hardware costs for new engines
are comprised of variable costs (for hardware and assembly time) and
fixed costs (for R&D, retooling, and certification). Total operating
costs include the estimated incremental cost for low-sulfur diesel
fuel, any expected

[[Page 35490]]

increases in maintenance cost, or fuel consumption costs along with any
decreases in operating cost expected due to low-sulfur fuel. Cost
estimates based on these projected technology packages represent an
expected incremental cost of engines in the 2007 model year. Costs in
subsequent years would be reduced by several factors, as described
below. Separate projected costs were derived for engines used in three
service classes of heavy-duty diesel engines. All costs are presented
in 1999 dollars.
The costs of these new technologies for meeting the proposed 2007
model year standards are itemized in the Draft RIA and summarized in
Table V.A-1. For light heavy-duty vehicles, the cost of a new 2007
model year engine is estimated to increase by $1,688 and operating
costs over a full life-cycle to increase by about $431. For medium
heavy-duty vehicles the cost of a new engine is estimated to increase
by $2,213, with life-cycle operating costs increasing to $826.
Similarly, for heavy heavy-duty engines, the vehicle cost is expected
to increase by $2,768, and estimated additional life-cycle operating
costs are $3,362. The higher incremental increase in operating costs
for the heavy heavy-duty vehicles is due to the larger number of miles
driven over their lifetime (714,000 miles on average) and their
correspondingly high lifetime fuel usage. Emission reductions are also
proportional to VMT and so are significantly higher for heavy heavy-
duty vehicles.
We also believe there are factors that would cause cost impacts to
decrease over time, making it appropriate to distinguish between near-
term and long term costs. Research in the costs of manufacturing has
consistently shown that as manufacturers gain experience in production,
they are able to apply innovations to simplify machining and assembly
operations, use lower cost materials, and reduce the number or
complexity of component parts.\137\ Our analysis, as described in more
detail in the draft RIA, incorporates the effects of this learning
curve by projecting that the variable costs of producing the low-
emitting engines decreases by 20 percent starting with the third year
of production (2009 model year) and by reducing variable costs again by
20 percent starting with the fifth year of production. We invite
comment on this methodology to account for the learning curve phenomena
and also request comment on whether learning is likely to reduce costs
in this industry. Additionally, since fixed costs are assumed to be
recovered over a five-year period, these costs are not included in the
analysis after the first five model years. Finally, manufacturers are
expected to apply ongoing research to make emission controls more
effective and to have lower operating cost over time. However, because
of the uncertainty involved in forecasting the results of this
research, we have conservatively not accounted for it in this analysis.
Table V.A-1 lists the projected costs for each category of vehicle in
the near- and long-term. For the purposes of this analysis, ``near-
term'' costs are those calculated for the 2007 model year and ``long
term'' costs are those calculated for 2012 and later model years.
---------------------------------------------------------------------------

\137\ ``Learning Curves in Manufacturing,'' Linda Argote and
Dennis Epple, Science, February 23, 1990, Vol. 247, pp. 920-924.
---------------------------------------------------------------------------

We welcome comment on the degree to which this program may
influence sales of new heavy-duty vehicles in the early years of the
program, and the resulting impact this would have on our projected
program benefits and costs. Costlier model year 2007 vehicles may
induce some potential purchasers of these vehicles to instead buy 2006
models to save money, or to defer a purchase longer than they otherwise
might have. On the other hand, we would anticipate that the very low
emissions characteristics of these new vehicles would cause many buyers
for whom cleaner diesels would be good for business (for example, urban
transit authorities and touring or shuttle services) to retire older
higher-emitting vehicles early.

Table V.A-1.--Projected Incremental System Cost and Life Cycle Operating Cost for Heavy-Duty Diesel Vehicles
[Net present values in the year of sale, 1999 dollars]
----------------------------------------------------------------------------------------------------------------
Life-cycle
Vehicle class Model year Hardware cost operating
cost*
----------------------------------------------------------------------------------------------------------------
Light heavy-duty.............................. Near term....................... $1,688 $431
Long term....................... 982 413
Medium heavy-duty............................. Near term....................... 2,213 826
Long term....................... 1,188 800
Heavy heavy-duty.............................. Near term....................... 2,768 3,362
Long term....................... 1,572 3,265
Urban Bus..................................... Near term....................... 2,268 3,942
Long term....................... 1,252 3,874
----------------------------------------------------------------------------------------------------------------
* Incremental life-cycle operating costs include the incremental costs to refine and distribute low sulfur
diesel fuel, the service cost of closed crankcase filtration systems, and the lower maintenance costs realized
through the use of low sulfur diesel fuel (see discussion in section V.3).

2. New System Costs for NO_X and PM Emission Control
Several new technologies are projected for complying with the
proposed 2007 model year emission standards. We are projecting that
NO_X adsorbers and catalyzed diesel particulate filters would
be the most likely technologies applied by the industry in order to
meet our proposed emissions standards. The fact that manufacturers
would have several years before implementation of the proposed new
standards ensures that the technologies used to comply with the
standards would develop significantly before reaching production. This
ongoing development could lead to reduced costs in three ways. First,
we expect research will lead to enhanced effectiveness for individual
technologies, allowing manufacturers to use simpler packages of
emission control technologies than we would predict given the current
state of development. Similarly, we anticipate that the continuing
effort to improve the emission control technologies will include
innovations that allow lower-

[[Page 35491]]

cost production. Finally, we believe that manufacturers would focus
research efforts on any drawbacks, such as fuel economy impacts or
maintenance costs, in an effort to minimize or overcome any potential
negative effects.
We anticipate that in order to meet the proposed standards,
industry would introduce a combination of primary technology upgrades
for the 2007 model year. Achieving very low NO_X emissions
will require basic research on NO_X emission control
technologies and improvements in engine management to take advantage of
the exhaust emission control system capabilities. The manufacturers are
expected to take a systems approach to the problem optimizing the
engine and exhaust emission control system to realize the best overall
performance possible. Since most research to date with exhaust emission
control technologies has focused on retrofit programs there remains
room for significant improvements by taking such a systems approach.
The NO_X adsorber technology in particular is expected to
benefit from re-optimization of the engine management system to better
match the NO_X adsorbers performance characteristics. The
majority of the $600 million dollars we have estimated for research is
expected to be spent on developing this synergy between the engine and
NO_X exhaust emission control systems. PM control
technologies are expected to be less sensitive to engine operating
conditions as they have already shown good robustness in retrofit
applications with low-sulfur diesel fuel.
The NO_X adsorber system that we are anticipating would
be applied in 2007 consists of a catalyst which combines traditional
gasoline three-way conversion technology with a newly developed
NO_X storage function, a reductant metering system and a
means to control engine air fuel (A/F) ratio. The NO_X
adsorber catalyst itself is a relatively new device, but is benefitting
in its development from over 20 years of gasoline three-way catalyst
development. In order for it to function properly, a systems approach
that includes a reductant metering system and control of engine A/F
ratio is also necessary. Many of the new air handling and electronic
system technologies developed in order to meet the 2004 heavy-duty
engine standards can be applied to accomplish the NO_X
adsorber control functions as well. Some additional hardware for
exhaust NO_X or O2 sensing and for fuel metering
will likely be required. We have estimated that this additional
hardware will increase new engine costs by approximately $350 for a
heavy heavy-duty diesel engine. The Draft RIA also calculates an
increase in warranty costs for this additional hardware. In total the
new NO_X control technologies required in order to meet the
proposed 2007 emission standards are estimated to increase light heavy-
duty engine costs by $890, medium heavy-duty engine costs by $1,047 and
heavy heavy-duty engine costs by $1,410 in the year 2007. In the year
2012 and beyond the incremental costs are expected to decrease to $570
for a light heavy-duty engine, $670 for a medium heavy-duty engine and
to $902 for a heavy heavy-duty engine.
Catalyzed diesel particulate filters are experiencing widespread
retrofit use in much of Europe as low-sulfur diesel fuel becomes
readily available. These technologies are proving to be robust in their
non-optimized retrofit applications requiring no modification to engine
or vehicle control functions. We therefore anticipate that catalyzed
diesel particulate filters can be integrated with new diesel engines
with only a minimal amount of engine development. We do not anticipate
that additional hardware beyond the diesel particulate filter itself
and an exhaust pressure sensor for OBD will be required in order to
meet the proposed PM standard. We estimate in 2007 that diesel
particulate filter systems will add $633 to the cost of a light heavy-
duty vehicle, $796 to the cost of a medium heavy-duty vehicle and
$1,028 to the cost of a heavy heavy-duty vehicle. By 2012 these costs
are expected to decrease to $389, $491, and $638 respectively. These
cost estimates are comparable to estimates made by the Manufacturers of
Emission Controls Association for these technologies.\138\
---------------------------------------------------------------------------

\138\ Letter from Bruce Bertelsen, Manufacturers of Emission
Controls Association (MECA) to William Charmley, US EPA, December
17, 1998. The letter documents a MECA member survey of expected
diesel particulate filter costs. EPA Air Docket A-99-06.
---------------------------------------------------------------------------

We have proposed to eliminate the exemption that allows turbo-
charged heavy-duty diesel engines to vent crankcase gases directly to
the environment, so called open crankcase systems, and have projected
that manufacturers will rely on engineered closed crankcase ventilation
systems which filter oil from the blow-by gases. We have estimated the
initial cost of these systems in 2007 to be $37, $42, and $49 for
light, medium and heavy heavy-duty diesel engines respectively.
Additionally we expect a portion of the oil filtration system to be a
service replacement oil filter which will be replaced on a 30,000 mile
service interval with a service cost of $10, $12, and $15 for light,
medium, and heavy heavy-duty diesel engines respectively. These cost
are summarized with the other cost for emission controls in Table V.A-1
and are included in the aggregate cost reported in section V.E.
3. Operating Costs Associated With NO_X and PM Control
The Draft RIA assumes that a variety of new technologies will be
introduced to enable heavy-duty vehicles to meet the new emissions
standards we are proposing. Primary among these are advanced emission
control technologies and low-sulfur diesel fuel. The many benefits of
low-sulfur diesel fuel are described in section III, and the
incremental cost for low-sulfur fuel is described in section V.D. The
new emission control technologies are themselves not expected to
introduce additional operating costs in the form of increased fuel
consumption. Operating costs are estimated in the Draft RIA over the
life of the vehicle and are expressed as a net present value (NPV) in
1999 dollars for comparison purposes.
Total operating cost estimates include both the expected increases
in maintenance and fuel costs (both the incremental cost for low-sulfur
fuel and any fuel consumption penalty) due to the emission control
systems application and the predicted decreases in maintenance cost due
to the use of low-sulfur fuel. Today's proposal estimates some increase
in operating costs due to the incremental cost of low-sulfur diesel
fuel but no net increase in fuel consumption with the application of
the new emission control technologies (see discussion in section
III.G). The net increase in operating costs are summarized in Table
V.A-1. While we are using these incremental operating cost estimates
for our cost effectiveness calculations, it is almost certain that the
manufacturers will improve existing technologies or introduce new
technologies in order to offset at least some of the increased
operating costs. We request comment on these operating cost estimates
and on ways in which industry may be able to offset these operating
costs.
We estimate that the low-sulfur diesel fuel we are proposing to
require in order to enable these technologies would have an incremental
cost of approximately $0.044/gallon as discussed in section V.D. The
proposed low-sulfur diesel fuel may also provide additional benefits by
reducing the engine maintenance costs associated with corrosion due to
sulfur in the current diesel fuel. These benefits, which are discussed
further in section V.C and in the draft RIA, include extended oil

[[Page 35492]]

change intervals due to the slower acidification rate of the engine oil
with low-sulfur diesel fuel. Service intervals for the EGR system are
also expected to increase due to lower-sulfur induced corrosion than
will occur with today's higher-sulfur fuel. This lengthening of service
intervals provides a significant savings to the end user. As described
in more detail in the Draft RIA we anticipate that low-sulfur diesel
fuel would provide additional cost savings to the consumer of $153 for
light heavy-duty vehicles, $249 for medium heavy-duty vehicles and $610
for heavy heavy-duty vehicles. The operating costs for replacement
filters in the closed crankcase filtration systems are estimated to be
$48 for light heavy-duty vehicles, $72 for medium heavy-duty vehicles
and $268 for heavy heavy-duty vehicles in 2007 and in the long term are
expected to decrease to $31 for a light heavy-duty vehicle, $46 for a
medium heavy-duty vehicle and $172 for a heavy heavy-duty vehicle.
Factoring the cost savings due to low sulfur diesel fuel into the
additional cost for low-sulfur diesel fuel and the service cost of the
closed crankcase ventilation system yields a net increase in vehicle
operating costs of $431 for a light heavy-duty vehicle, $826 for a
medium heavy-duty vehicle and $3,362 for a heavy heavy-duty vehicle.
These life cycle operating costs are also summarized in Table V.A-1.
The net increase in operating cost can also be expressed as an average
annual operating cost for each class of heavy-duty vehicle. Expressed
as an approximate annual per vehicle cost, the additional operating
cost is estimated as $50 for a light heavy-duty vehicle, $100 for a
medium heavy-duty vehicle, and $400 for a heavy heavy-duty vehicle.

B. Cost for Gasoline Vehicles to Meet Proposed Emissions Standards

1. Summary of New System Costs
To perform a cost analysis for the proposed standards, we first
determined a package of likely technologies that manufacturers could
use to meet the proposed standards and then determined the costs of
those technologies. In making our estimates we have relied on our own
technology assessment which included publicly available information,
such as that developed by California, as well as confidential
information supplied by individual manufacturers, and the results of
our own in-house testing.
In general, we expect that heavy-duty gasoline vehicles would (like
Tier 2 light duty vehicles) be able to meet these standards through
refinements of current emissions control components and systems rather
than through the widespread use of new technology. More specifically,
we anticipate a combination of technology upgrades such as the
following:
Improvements to the catalyst system design, structure, and
formulation, plus an increase in average catalyst size and loading.
Air and fuel system modifications including changes such
as improved oxygen sensors, and calibration changes including improved
precision fuel control and individual cylinder fuel control.
Exhaust system modifications, possibly including air
gapped components, insulation, leak free exhaust systems, and thin wall
exhaust pipes.
Increased use of fully electronic exhaust gas
recirculation (EGR).
Increased use of secondary air injection.
Use of ignition spark retard on engine start-up to improve
upon cold start emission control.
Use of low permeability materials and minor improvements
to designs, such as the use of low-loss connectors, in evaporative
emission control systems.
We expect that the technologies needed to meet these proposed
heavy-duty gasoline standards would be very similar to those required
to meet the Tier 2 standards for vehicles over 8,500 pounds GVWR. Few
heavy-duty gasoline vehicles currently rely on technologies such as
close coupled catalysts and secondary air injection, but we expect they
would do so to in order to meet the proposed 2007 standards.
For each group we developed estimates of both variable costs (for
hardware and assembly time) and fixed costs (for R&D, retooling, and
certification). Cost estimates based on the current projected costs for
our estimated technology packages represent an expected incremental
cost of vehicles in the near-term. For the longer term, we have
identified factors that would cause cost impacts to decrease over time.
First, since fixed costs are assumed to be recovered over a five-year
period, these costs disappear from the analysis after the fifth model
year of production. Second, the analysis incorporates the expectation
that manufacturers and suppliers would apply ongoing research and
manufacturing innovation to making emission controls more effective and
less costly over time. Research in the costs of manufacturing has
consistently shown that as manufacturers gain experience in production
and use, they are able to apply innovations to simplify machining and
assembly operations, use lower cost materials, and reduce the number or
complexity of component parts.\139\ These reductions in production
costs are typically associated with every doubling of production
volume. Our analysis incorporates the effects of this ``learning
curve'' by projecting that a portion of the variable costs of producing
the new vehicles decreases by 20 percent starting with the third year
of production. We applied the learning curve reduction only once since,
with existing technologies, there would be less opportunity for
lowering production costs than would be the case with the adoption of
new technology. We did not apply the learning curve reduction to
precious metal costs, nor did we apply it for the evaporative
standards. We invite comment on this methodology to account for the
learning curve phenomena and also request comment on whether learning
is likely to reduce costs in this industry.
---------------------------------------------------------------------------

\139\ See Chapter V of the final Tier 2 Regulatory Impact
Analysis, contained in Air Docket A-97-10.
---------------------------------------------------------------------------

We have prepared our cost estimates for meeting the new heavy-duty
gasoline standards using a baseline of current technologies for heavy-
duty gasoline vehicles and engines. Finally, we have incorporated what
we believe to be a conservatively high level of R&D spending at
$2,500,000 per engine where no California counterpart exists. We have
included this large R&D effort because calibration and system
optimization is likely to be a critical part of the effort to meet the
standards. However, we believe that the R&D costs may be generous
because the projection probably underestimates the carryover of
knowledge from the development required to meet the light-duty Tier 2
and CARB LEV-II standards.
Table V.B-1 provides our estimates of the per vehicle increase in
purchase price for heavy-duty gasoline vehicles and engines. The near-
term cost estimates in Table V.B-1 are for the first years that
vehicles meeting the standards are sold, prior to cost reductions due
to lower productions costs and the retirement of fixed costs. The long-
term projections take these cost reductions into account. We request
comment on the costs shown in Table V.B-1 and the analysis behind them.

[[Page 35493]]

Table V.B-1.--Projected Incremental System Cost and Life Cycle Operating Cost for Heavy-Duty Gasoline Vehicles
[Net present values in the year of sale, 1999 dollars]
----------------------------------------------------------------------------------------------------------------
Incremental Life-cycle
Vehicle class Model year system cost operating cost
----------------------------------------------------------------------------------------------------------------
Heavy-Duty Gasoline........................... Near term....................... $182 $0
Long term....................... 152 0
----------------------------------------------------------------------------------------------------------------

2. Operating Costs Associated With Meeting the Heavy-Duty Gasoline
Standard
Low sulfur gasoline is a fundamental enabling technology which will
allows heavy-duty gasoline vehicles to meet the very low emission
standards being proposed today. The low sulfur gasoline required under
the Tier 2 proposal will enable advanced exhaust emission control for
heavy-duty vehicles as well. Today's proposal puts no additional
requirements on gasoline sulfur levels and as such should not directly
increase gasoline fuel costs. Additionally, the new technologies being
employed in order to meet the new standards are not expected to
increase fuel consumption for heavy-duty gasoline vehicles. In fact,
there may be some small improvement in fuel economy from the
application of improved fuel and air control systems on these engines.
Therefore, in the absence of changes to gasoline specifications and
with no decrease in fuel economy, we do not expect any increase in
vehicle operating costs.

C. Benefits of Low-Sulfur Diesel Fuel for the Existing Diesel Fleet

We estimate that the proposed low-sulfur diesel fuel would provide
additional benefits to the existing heavy-duty vehicle fleet as soon as
the fuel is introduced. We believe these benefits could offer
significant cost savings to the vehicle owner without the need for
purchasing any new technologies. The Draft RIA has catalogued a variety
of benefits from the proposed low-sulfur diesel fuel. These benefits
are summarized in Table V.C-1.

Table V.C-1.--Components Potentially Affected by Lower Sulfur Levels in
Diesel Fuel
------------------------------------------------------------------------
Effect of lower Potential impact
Affected components sulfur on engine system
------------------------------------------------------------------------
Piston Rings.................... Reduce corrosion Extended engine
wear. life and less
frequent
rebuilds.
Cylinder Liners................. Reduce corrosion Extended engine
wear. life and less
frequent
rebuilds.
Oil Quality..................... Reduce deposits Reduce wear on
and less need for piston ring and
alkaline cylinder liner
additives. and less frequent
oil changes.
Exhaust System (tailpipe)....... Reduces corrosion Less frequent part
wear. replacement.
EGR............................. Reduces corrosion Less frequent part
wear. replacement.
------------------------------------------------------------------------

The actual value of these benefits over the life of the vehicle
would depend upon the length of time that the vehicle operates on low-
sulfur diesel fuel and the degree to which vehicle operators change
engine rebuild patterns to take advantage of these benefits. For a
vehicle near the end of its life in 2007 the benefits would be quite
small. However for vehicles produced in the years immediately preceding
the introduction of low-sulfur fuel the savings would be substantial.
The Draft RIA estimates that a heavy heavy-duty vehicle introduced into
the fleet in 2006 would realize savings of $610 over its life. This
savings could alternatively be expressed in terms of fuel costs as
approximately 1 cent per gallon as discussed in the draft RIA. These
savings would occur without additional new cost to the vehicle owner
beyond the incremental cost of the low-sulfur diesel fuel, although
these savings would require changes to existing maintenance schedules.
Such changes seem likely given the magnitude of the savings and the
nature of the regulated industry.
The maintenance benefits we project come primarily from extended
oil change intervals. We have no quantitative data on how much longer
these intervals might be. Based on discussions with some engine
manufacturers, we believe it is reasonable to assume that engine oil
change intervals will increase by 10 percent for each class of engine
(in both new and existing fleets). We seek comment on this key
assumption and on these projected savings and all of the assumptions
behind them; details of the analysis behind these savings can be found
in the draft RIA contained in the docket for this rule.

D. Cost of Proposed Fuel Change

We estimate that the overall cost associated with lowering the
sulfur cap from the current level of 500 ppm to the 15 ppm level
proposed today will be approximately 4.4 cents per gallon. As discussed
in sections V.A. and V.C., this cost would be offset by a one cent per
gallon savings (or more) from the reduction in vehicle maintenance
savings that result from the use of the cleaner fuel. The fuel cost is
comprised of a number of components associated with refining and
distributing the fuel. The majority of the fuel cost is expected to be
the refining cost which is estimated to be approximately 4.0 cents per
gallon, which includes the cost of producing more volume of diesel fuel
because desulfurization decreases the energy density of the fuel. The
remaining 0.4 cents per gallon in fuel costs is associated with an
anticipated increase in the use of additives to maintain fuel lubricity
at a cost of 0.2 cents per gallon, and an increase in distribution
costs of 0.2 cents per gallon. The increase in distribution costs
comprises 0.1 cents per gallon to distribute the additional volume of
diesel fuel needed to compensate for the decrease in fuel energy
density, and 0.1 cents per gallon to maintain product integrity in the
distribution system. These cost estimates are discussed in more detail
below and in the Draft RIA.

[[Page 35494]]

When the 4.4 cent per gallon cost is applied to the expected low sulfur
diesel fuel sales volume of approximately 40 billion gallons at the
start of the program, it equates to an annual cost of roughly $1.8
billion per year. This fuel cost would be offset by a reduction in
maintenance costs of roughly $0.4 billion per year.
1. Refinery Costs
As explained in Section IV, refiners would have to install capital
equipment to meet the proposed diesel fuel sulfur standard. Presuming
that refiners will want to minimize the cost involved and use
conventional technology, refiners are expected to build onto their
existing desulfurization unit by adding another hydrotreating reactor
and other related equipment.
In our analysis, we estimated the cost of lowering onroad diesel
fuel sulfur levels for a national average refinery starting from the
current national average sulfur level of about 350 ppm down to 7 ppm.
We believe that a refinery's average diesel fuel sulfur level would be
roughly 7 ppm under a 15 ppm cap standard. We then calculated a
national aggregate cost and cents-per-gallon cost. Based on this
analysis we estimate that, on average, individual refiners in the years
2004-05 would be expected to invest about $30 million for capital
equipment and spend about $8 million per year for each refinery to
cover the operating costs associated with these desulfurization units.
Since this average represents a diverse size range of refineries, some
refineries would pay more and others less than this average cost. When
the average per-refinery cost is aggregated for all the onroad diesel
fuel expected to be produced in this country in 2007, we estimate that
the total investment for desulfurizing diesel fuel would be about $1.9,
$2.0, and $0.2 billion in 2004, 2005, and 2006, respectively, as
discussed in section IV.B. Operating costs for these units are expected
to be about $1.1 billion per year.
Using our estimated capital and operating costs we calculated the
average per-gallon cost of reducing diesel fuel sulfur down to meet the
proposed 15 ppm cap standard. Using a capital cost amortization factor
based on a seven percent rate of return on investment before taxes, we
estimated the average national cost for desulfurizing onroad diesel
sulfur to be about 4.0 cents per gallon. This cost is our estimated
cost to society of producing onroad diesel to meet a 15 ppm cap
standard that we used for estimating cost effectiveness.
There is currently no commercial experience in the U.S. and only a
limited amount of information in the public literature on the costs
associated with reducing the sulfur level in diesel fuel to very low
levels on an ongoing operational basis. Experience in Sweden involves
other changes to the fuel as well that would tend to drive up the costs
considerably. The EMA recently commissioned a study by Mathpro of the
economics of controlling the sulfur content of highway and nonroad
diesel fuel to various sulfur levels as low as 2 ppm. Unfortunately,
none of the scenarios modeled in the EMA study are consistent with our
proposal today. Furthermore, some of the assumptions made in the
analysis are inconsistent with our standard assumptions for economic
analysis. For example, Mathpro used a higher rate of return on new
capital than the rate we use. Nevertheless, some insight can be gained
from a broad comparison of Mathpro's and our cost projections. The
proposed sulfur cap for highway diesel fuel is very roughly bracketed
by two Mathpro sulfur control scenarios: (1) a highway diesel fuel
standard of 20 ppm on average with a nonroad diesel fuel standard of
350 ppm on average, and (2) an highway diesel fuel standard of 2 ppm on
average with a nonroad diesel fuel standard of 20 ppm on average.
Mathpro's projected refining costs for these two scenarios range from 4
to just under 6 cents per gallon (citing their costs for revamping
current diesel fuel hydrotreaters with reactors in series, which is
equivalent to our technology projections). Considering that Mathpro
uses a higher rate of return on capital and that both of their
scenarios included controlling nonroad diesel fuel, the two sets of
cost projections appear to be roughly consistent. This serves to give
us some confidence that our cost estimate for a sulfur cap of 15 ppm on
highway diesel fuel is reasonable. This is discussed in further detail
in the Draft RIA.
Although API assisted in the study, API has expressed some concern
about the accuracy of the EMA cost estimates. API highlighted their
concerns on the EMA study in a memo to the Director the Office of
Transportation Air Quality, which is included in the docket.\140\ While
API expressed their belief that the cost outcomes of the EMA study are,
in general, reasonable, they expressed serious concerns about the cost
of producing diesel with sulfur levels below 20 ppm (roughly equivalent
to a 30 ppm cap). API believes that, particularly at extremely low
sulfur levels, the measures needed to be taken would result in
significantly higher costs than estimated by EMA. We request comment on
this assessment.
---------------------------------------------------------------------------

\140\ Edward H. Murphy, API to Margo Oge, US EPA, October
26,1999.
---------------------------------------------------------------------------

We acknowledge that some refiners likely face higher
desulfurization costs than others. This is generally the case with any
fuel quality regulation, since the crude oils processed by, as well as
the configurations and product slates of individual refineries vary
dramatically. As mentioned in section IV, API believes that those
refiners facing higher than average costs may decide to leave the
highway diesel fuel market. They argue this is especially a possibility
if they are faced with a sulfur standard below a 30 ppm average (or 50
ppm cap), which they believe will require very large investments for
high pressure hydrotreating to maintain current highway diesel
production volumes. API also believes that many refiners may reduce
their production of highway diesel fuel, by switching the feedstocks
(i.e., LCO) which are most difficult to desulfurize to other markets,
thus avoiding the higher investments associated with high pressure
hydrotreating. If some refiners reduce highway diesel fuel production,
that could present an opportunity for other refiners, who choose to
make the investment, of higher prices for the new 15 ppm sulfur
product. Whether the potential for higher prices would be sufficient
and be apparent with sufficient leadtime to allow refiners to make an
added investment by the time the proposed rule is effective is
currently unclear.
For example, the refining industry actually overbuilt
desulfurization capacity for the current 500 ppm standard, as evidenced
by the significant use in the off-highway market of diesel fuel
produced to the current highway diesel sulfur standard of 500 ppm. Some
of this overproduction may have been due to limitations in the
distribution system to distribute both highway and off-highway grades
of diesel fuel. Despite the overall market overproduction, a number of
small refiners did decide to switch from the highway diesel fuel market
to the off-highway diesel fuel market, presumably for economic reasons.
Another incentive for refiners to invest in highway diesel fuel
desulfurization equipment is the potential for a growing light-duty
diesel market. Many vehicle manufacturers have announced plans to equip
their light-duty vehicles and, particularly, light-duty trucks with
diesel engines. Refiners may want to ensure their

[[Page 35495]]

presence in this growing and potentially profitable market.
Alternative markets for distillate products are limited in the U.S.
The domestic off-highway diesel fuel and heating oil markets are much
smaller than the highway diesel fuel market. The domestic off-highway
diesel fuel and heating oil markets are currently in balance,
considering the fact that some highway diesel fuel is currently being
sold into these markets. Assuming that the distribution system can be
changed to segregate highway and other distillate fuels more
economically, some amount of current highway diesel fuel production
could switch to these other markets with no loss of highway diesel fuel
supply. In addition, although the off-highway diesel fuel market is
growing, this growth will occur gradually over the next 6 years and not
occur on April 1, 2006. The heating oil market is very seasonal (strong
in the winter and weak in the summer), regional (strong in the
Northeast) and not growing. Thus, overall, we do not see much
opportunity for large domestic producers of highway diesel fuel to be
able to shift their production to these other domestic markets.
Export opportunities for diesel fuel are also limited to some
degree. Japan and Europe will have stringent sulfur caps in place by
2005 and have cetane requirements well beyond the cetane levels of
current U.S. diesel fuel. Asia, while growing in demand for diesel
fuel, has also been the focus of new grassroots refinery production and
again has high cetane requirements. Thus, the primary areas for export
of diesel fuel of average U.S. quality would appear to be Africa and
Latin America.
Refiners have also raised the possibility of exporting some of
their more difficult to desulfurize diesel feedstocks such as LCO to
other distillate markets. While this may be a possibility to some
degree as discussed in Section IV and the draft RIA, the opportunities
to do so appear to be limited. We have not conducted a detailed
analysis of the potential for this exportation. Refiners would have to
hydrotreat this material to lower its sulfur content in order to meet
the European Union 50 ppm sulfur cap (and increase its cetane) in order
for it to be used as a diesel fuel blendstock. Otherwise, its only use
without additional treating would be in heating fuel. With Europe and
developing countries expected to experience increasing demand for non-
diesel, distillate fuel, there may be economic opportunities for
exporting such fuel.
We request comments on the possibility that the proposed sulfur cap
would cause some refiners to abandon the U.S. highway diesel fuel
market or to reduce highway diesel fuel production, as well as on the
impact that this would have on diesel fuel supply and price in the U.S.
We also request comment on whether refiners would likely desire to
shift all their LCO to non-highway diesel fuel markets or just the
heavier portion which contains the most sterically hindered compounds.
We also request comment on the economic viability of alternative
markets for current highway diesel fuel or its more difficult to
desulfurize components. We also request comments on the ability of
overseas refiners providing highway diesel fuel under the proposed
sulfur cap should domestic refiners reduce production. Finally, as
discussed in section VI.A., we are also considering various phase-in
approaches for implementing the low sulfur diesel standard. A phase-in
could help spread out the design, construction, and capital expenditure
of refinery modifications necessary to comply with the proposed diesel
fuel sulfur standard, and in so doing could further minimize any risk
of supply shortages. We request comment on the appropriateness and
ability of a phase-in to address these concerns.
2. Cost of Possibly Needed Lubricity Additives
As discussed in section IV, the refinery processes needed to
achieve the sulfur standard have some potential to degrade the natural
lubricity characteristics of the fuel. Consequently an increase in the
use of lubricity additives for diesel fuel may be anticipated over the
amounts used today. We contacted various producers of lubricity
additives to get their estimates of what costs might be incurred for
this increase in the use of lubricity additives. The cost estimates
varied from 0.1 to 0.5 cents per gallon. This range is to be expected
since the cost will be a strong function of not only the additive type,
but also the assumed treatment rate and the volume of fuel that needs
to be treated, both of which will be, to some extent, a function of the
sulfur cap. As described in more detail in the Draft RIA, we have
included in the fuel cost estimate an average cost of 0.2 cents per
gallon for lubricity additives over the entire pool of low-sulfur
highway diesel fuel. This estimate is comparable to an estimate made by
Mathpro in a study sponsored by the EMA. We request comment on our cost
estimate. In particular, we request comment on whether there may be
unique costs for the military to maintain the lubricity of their
distillate fuels. We request that such comments addressing this issue
include a detailed discussion of the volumes of fuel effected, current
lubricity additive use, and the additional measures that might be
needed (and associated costs) to maintain the appropriate level of fuel
lubricity.
3. Distribution Costs
Under the proposed 15 ppm sulfur cap, we project that distribution
costs would increase by a total of 0.2 cents per gallon as discussed
below.
If the proposed sulfur standard is adopted, there would be a
greater difference between the sulfur content of highway diesel fuel
and other distillate products than presently exists.\141\ For example,
off-highway diesel fuel currently has a sulfur content that is
approximately ten times that of highway diesel. Under the proposed
sulfur standard, off-highway diesel fuel would have a sulfur content
over two hundred times that of highway diesel fuel. This could
potentially make it more difficult to limit the sulfur contamination of
highway diesel fuel with other distillate products as the fuel travels
through the distribution system. As discussed in section IV, standard
industry practices, if followed carefully, should be able to virtually
eliminate the potential contamination. To do so, however, is expected
to result in slightly increased costs in a few different parts of the
distribution system.
---------------------------------------------------------------------------

\141\ Highway diesel fuel currently must have a sulfur content
of no more than 500 ppm and typically has an average sulfur content
of 350 ppm. Off-highway diesel fuel sulfur content is currently
unregulated and is approximately 3,500 ppm on average. The maximum
allowed sulfur content of heating oil is 5,000 ppm. The maximum
allowed sulfur content of kerosene (and jet fuel) is 3,000 ppm.
---------------------------------------------------------------------------

We identified three segments in the distribution system (pipeline
operators, terminal operators, and tank-truck operators) that might
experience increased costs due to increased difficulty in limiting
sulfur contamination under the proposed sulfur standard. As discussed
in the Draft RIA, we estimate that the total increase in diesel
distribution costs associated with adequately limiting sulfur
contamination under today's proposal would be no more than 0.1 cents
per gallon for the distribution system as a whole. The majority of this
increased cost is attributed to the unavoidable mixing of highway
diesel with other products that occurs in pipeline shipments. The
amount of interface (e.g., mixture of a highway diesel batch and a
nonroad diesel batch) that must be downgraded to a lower

[[Page 35496]]

price product is expected to grow with a lower sulfur cap for highway
diesel, resulting in a slightly increased cost for pipeline shipments.
A slight increase in distribution costs is also expected to result at
terminals due to the anticipated need for additional quality assurance
testing at very low sulfur levels. We believe that, although tank-truck
operators may need to more carefully observe current industry practices
used to limit product contamination, this will not result in a
significant increase in costs.
We invite comment on the amount of sulfur contamination which might
be expected from each segment of the distribution system, the measures
that might be taken to limit contamination, and the costs associated
with these measures. We also request comment on the level of sulfur
contamination in the distribution system that might be considered
unavoidable without the imposition of an undue burden on diesel
distributors and how this bears on the question of what sulfur level
the refiner would need to meet at the refinery gate (the compliance
margin) to ensure that highway diesel fuel does not exceed the proposed
cap on sulfur content. Please refer to section IV.E for discussion of
the compliance margin that we anticipate refiners will need to provide.
The energy density of diesel fuel would be decreased as a side
effect of reducing sulfur content to the proposed 15 ppm cap.
Consequently, to meet the same level of consumer demand an increased
volume of diesel fuel would need to move through the distribution
system. The cost of distributing this increased volume of diesel fuel
was calculated within the model that used to evaluate refining costs
(see the Draft RIA). Spread over the total volume of diesel fuel
distributed, the additional cost is estimated at 0.1 cents per gallon.
We request comment on this cost estimate.

E. Aggregate Costs

Using current data for the size and characteristics of the heavy-
duty vehicle fleet and making projections for the future, the diesel
per-engine, gasoline per-vehicle, and per-gallon fuel costs described
above can be used to estimate the total cost to the nation for the
emission standards in any year. Figure V.E-1 portrays the results of
these projections.\142\ All capital costs have been amortized.
---------------------------------------------------------------------------

\142\ Figure V.E-1 is based on the amortized engine, vehicle and
fuel costs as described in the Draft RIA. Actual capital
investments, particularly important for fuels, would occur prior to
and during the initial years of the program.

BILLING CODE 6560-50-P
[GRAPHIC] [TIFF OMITTED] TP02JN00.003

BILLING CODE 6560-50-C
As can be seen from the figure, the annual costs start out at less
than a billion dollars in year 2006 and increase over the phase-in
period to about $2.8 billion in 2015. Thereafter, total annualized
costs are projected to continue increasing due to the effects of
projected growth in engine sales and fuel consumption. The Draft RIA
provides further detail regarding these cost projections.
Future consumption of today's proposed low sulfur diesel fuel may
be influenced by a potential influx of diesel-powered cars and light
trucks into the light-duty fleet. At the present time, virtually all
cars and light trucks being sold are gasoline fueled. However, the
possibility exists that diesels will become more prevalent in the car
and light-duty truck fleet, since automotive companies have announced
their desire to increase their sales of diesel cars and light trucks.
For the Tier 2 rulemaking, the Agency performed a sensitivity analysis
using A.D.Little's ``most likely'' increased growth scenario of diesel
penetration into the light-duty vehicle fleet which culminated in a 9
percent and 24 percent penetration of diesel vehicles in the LDV and
LDT markets,

[[Page 35497]]

respectively, in 2015 (see Tier 2 RIA, Table III.A. 13). Were this
scenario to play out, the increased number of diesel-powered cars and
light-duty trucks would increase the societal costs (those costs, in
total, paid by consumers) for the proposed higher priced diesel fuel
because more diesel fuel would be consumed. However, were more diesel
vehicles to penetrate the light-duty fleet, less gasoline would be
consumed than was estimated in our Tier 2 cost analysis. Also, diesel
vehicles tend to get higher fuel economy. In the end, the effect of
increased dieselization of the light-duty fleet may have little or no
impact on the aggregate costs estimated for today's proposal. While we
have not fully analyzed this light-duty diesel penetration scenario, we
request comment on it and relevant data which would allow us to perform
a sensitivity analysis.

F. Cost Effectiveness

One tool that can be used to assess the value of new standards for
heavy-duty vehicles and engines is cost effectiveness, in which the
costs incurred to reach the standards are compared to the mass of
emission reductions. This analysis results in the calculation of a $/
ton value, the purpose of which is to show that the reductions from the
engine and fuel controls being proposed today are cost effective, in
comparison to alternative means of control. This analysis involves a
comparison of our program not only to past measures, but also to other
potential future measures that could be implemented. Both EPA and
states have already adopted numerous control measures, and remaining
measures tend to be more expensive than those previously employed. As
we and States tend to employ the most cost effective available measures
first, more expensive ones must be adopted to achieve further emission
reductions.
1. What Is the Cost Effectiveness of This Proposed Program?
We have calculated the cost-effectiveness of our proposed diesel
engine/gasoline vehicle/diesel sulfur standards based on two different
approaches. The first considers the net present value of all costs
incurred and emission reductions generated over the life of a single
vehicle meeting our proposed standards. This per-vehicle approach
focuses on the cost-effectiveness of the program from the point of view
of the vehicles and engines which will be used to meet the new
requirements. However, the per-vehicle approach does not capture all of
the costs or emission reductions from our proposed diesel engine/
gasoline vehicle/diesel sulfur program since it does not account for
the use of low sulfur diesel fuel in current diesel engines. Therefore,
we have also calculated an 30-year net present value cost-effectiveness
using the net present value of costs and emission reductions for all
in-use vehicles over a 30-year time frame. The baseline or point of
comparison for this evaluation is the previous set of engine, vehicle,
and diesel sulfur standards (in other words, the applicable 2004 model
year standards).
As described earlier in the discussion of the cost of this program,
the cost of complying with the new standards will decline over time as
manufacturing costs are reduced and amortized capital investments are
recovered. To show the effect of declining cost in the per-vehicle
cost-effectiveness analysis, we have developed both near term and long
term cost-effectiveness values. More specifically, these correspond to
vehicles sold in years one and six of the vehicle and fuel programs.
Chapter VI of the RIA contains a full description of this analysis, and
you should look in that document for more details of the results
summarized here.
The 30-year net present value approach to calculating the cost-
effectiveness of our program involves the net present value of all
nationwide emission reductions and costs for a 30 year period beginning
with the start of the diesel fuel sulfur program and introduction of
model year 2007 vehicles and engines in year 2006. This 30-year
timeframe captures both the early period of the program when very few
vehicles that meet our proposed standards will be in the fleet, and the
later period when essentially all vehicles in the fleet will meet our
proposed standards. We have calculated the 30-year net present value
cost-effectiveness using the net present value of the nationwide
emission reductions and costs for each calender year. These emission
reductions and costs are given for every calendar year in the RIA, in
addition to details of the methodology we used to calculate the 30-year
net present value cost-effectiveness.
Our per-vehicle and 30-year net present value cost-effectiveness
values are given in Tables V.F-1 and V.F-2. Table V.F-1 summarizes the
per-vehicle, net present value lifetime costs, NMHC + NO_X
and PM emission reductions, and resulting cost-effectiveness results
for our proposed diesel engine/gasoline vehicle/diesel sulfur standards
using sales weighted averages of the costs (both near term and long
term) and emission reductions of the various vehicle and engine classes
affected. Table V.F-2 provides the same information from the program
30-year net present value perspective. It includes the net present
value of the 30 year stream of vehicle and fuel costs, NMHC +
NO_X and PM emission reductions, and the resulting 30-year
net present value cost-effectiveness. Diesel fuel costs applicable to
diesel engines have been divided equally between the adsorber and trap,
since low sulfur diesel is intended to enable all technologies to meet
our proposed standards. In addition, since the trap produces reductions
in both PM and hydrocarbons, we have divided the total trap costs
equally between compliance with the proposed PM standard and compliance
with the proposed NMHC standard.
Tables V.F-1 and V.F-2 also display cost-effectiveness values based
on two approaches to account for the reductions in SO2
emissions associated with the reduction in diesel fuel sulfur. While
these reductions are not central to the program and are therefore not
displayed with their own cost-effectiveness, they do represent real
emission reductions due to our program. The first set of cost-
effectiveness numbers in the tables simply ignores these reductions and
bases the cost-effectiveness on only the emission reductions from our
proposed program. The second set accounts for these ancillary
reductions by crediting some of the cost of the program to
SO2. The amount of cost allocated to SO2 is based
on the cost-effectiveness of SO2 emission reductions that
could be obtained from alternative, potential future EPA programs. The
SO2 credit was applied only to the PM calculation, since
SO2 reductions are primarily a means to reduce ambient PM
concentrations.

[[Page 35498]]

Table V.F-1.--Per-Engine Cost Effectiveness of the Proposed Standards for 2007 and Later MY Vehicles
----------------------------------------------------------------------------------------------------------------
Discounted Discounted
Discounted lifetime Discounted lifetime cost
Pollutants lifetime emission lifetime cost effectiveness
vehicle & reductions effectiveness per ton with
fuel costs (tons) per ton SO2 credit ^a
----------------------------------------------------------------------------------------------------------------
Near-term costs ^b:
NO_X+NMHC............................................ $1535 0.8838 $1,736 $1,736
PM.................................................. 872 0.0672 12,977 6,338
Long-term costs:
NO_X+NMHC............................................ 1121 0.8838 1,268 1,268
PM.................................................. 652 0.0672 9,704 3,065
----------------------------------------------------------------------------------------------------------------
\a\ $446 credited to SO2 (at $4800/ton) for PM cost effectiveness.
\b\ As described above, per-engine cost effectiveness does not include any costs or benefits from the existing,
pre-control, fleet of vehicles that would use the low sulfur diesel fuel proposed in this document.

Table V.F-2.--30-year Net Present Value ^a Cost Effectiveness of the Standards
----------------------------------------------------------------------------------------------------------------
30-year
n.p.v. 30-year 30-year
engine, n.p.v. 30-year n.p.v. cost
vehicle, & reduction n.p.v. cost effectiveness
fuel costs (tons) (in effectiveness per ton with
(in millions) per ton SO2 credit ^b
billions)
----------------------------------------------------------------------------------------------------------------
NO_X + NMHC.............................................. $28.9 18.9 $1,531 $1,531
PM...................................................... 8.8 0.79 11,248 1,850
----------------------------------------------------------------------------------------------------------------
^a This cost effectiveness methodology reflects the total fuel costs incurred in the early years of the program
when the fleet is transitioning from pre-control to post-control diesel vehicles. In 2007 10% of highway
diesel fuel is anticipated to be consumed by 2007 MY vehicles. By 2012 this increases to >50% for 2007 and
later MY vehicles.
^b $7.4 billion credited to SO2 (at $4800/ton).

2. Comparison With Other Means of Reducing Emissions
In comparison with other mobile source control programs, we believe
that our program represents a cost effective strategy for generating
substantial NO_X, NMHC, and PM reductions. This can be seen
by comparing the cost effectiveness of today's program with a number of
mobile source standards that EPA has adopted in the past. Table V.F-3
summarizes the cost effectiveness of several past EPA actions for
NO_X+ NMHC. Table V.F-4 summarizes the cost effectiveness of
several past EPA actions for PM.

Table V.F-3.--Cost Effectiveness of Previous Mobile Source Programs for
NO_X+NMHC
------------------------------------------------------------------------
Program $/ton
------------------------------------------------------------------------
Tier 2 vehicle/gasoline sulfur........................ 1,311-2,211
2004 Highway HD diesel................................ 207-405
Nonroad diesel engine................................. 416-660
Tier 1 vehicle........................................ 2,010-2,732
NLEV.................................................. 1,888
Marine SI engines..................................... 1,146-1,806
On-board diagnostics.................................. 2,263
Marine CI engines..................................... 23-172
------------------------------------------------------------------------
Note.--costs adjusted to 1998 dollars.

Table V.F-4.--Cost Effectiveness of Previous Mobile Source Programs for
PM
------------------------------------------------------------------------
Program $/ton
------------------------------------------------------------------------
Marine CI engines..................................... 511-3,797
1996 urban bus........................................ 12,000-19,200
Urban bus retrofit/rebuild............................ 29,600
1994 highway HD diesel................................ 20,450-23,940
------------------------------------------------------------------------
Note.--costs adjusted to 1998 dollars.

We can see from these tables that the cost effectiveness of our
proposed diesel engine/gasoline vehicle/diesel sulfur standards falls
within the range of these other programs for both NO_X+NMHC
and PM. Our proposed program overlaps the range of the recently
promulgated standards for Tier 2 light-duty vehicles and gasoline
sulfur shown in Table V.F-3. Our proposed program also overlaps the
cost-effectiveness of past programs for PM. It is true that some
previous programs have been more cost efficient than the program we are
proposing today. However, it should be expected that the next
generation of standards will be more expensive than the last, since the
least costly means for reducing emissions is generally pursued first.
In evaluating the cost effectiveness of our proposed diesel engine/
gasoline vehicle/diesel sulfur program, we also considered whether our
proposal is cost effective in comparison with possible stationary
source controls. In the context of the Agency's rulemaking which would
have revised the ozone and PM NAAQS,\143\ the Agency compiled a list of
additional known technologies that could be considered in devising new
emission reductions strategies.\144\ Through this broad review, over 50
technologies were identified that could reduce NOx, VOC, or PM. The
cost effectiveness of these technologies averaged approximately $5,000/
ton for VOC, $13,000/ton for NO_X, and $40,000/ton for PM.
Although a $10,000/ton limit was actually used in the air quality
analysis presented in the NAAQS revisions rule, these values clearly
indicate that, not only are future emission control strategies likely
to be more expensive (less cost effective) than past strategies, but
the cost effectiveness of our proposed program falls well

[[Page 35499]]

below the average of those choices, and is near the lower end of the
range of potential future strategies.
---------------------------------------------------------------------------

\143\ This rulemaking was remanded by the DC Circuit Court on
May 14, 1999. However, the analyses completed in support of that
rulemaking are still relevant, since they were designed to
investigate the cost effectiveness of a wide variety of potential
future emission control strategies.
\144\ ``Regulatory Impact Analyses for the Particulate Matter
and Ozone National Ambient Air Quality Standards and Proposed
Regional Haze Rule,'' Appendix B, ``Summary of control measures in
the PM, regional haze, and ozone partial attainment analyses,''
Innovative Strategies and Economics Group, Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC, July 17, 1997.
---------------------------------------------------------------------------

In summary, we believe that the weight of the evidence from
alternative means of providing substantial NO_X+NMHC and PM
emission reductions indicates that our proposed diesel engine/gasoline
vehicle/diesel sulfur program is cost effective. We believe this is
true from the perspective of other mobile source control programs and
from the perspective of other stationary source technologies that might
be considered. We request comment on the cost-effectiveness of this
program.

G. Does the Value of the Benefits Outweigh the Cost of the Proposed
Standards?

In addition to cost-effectiveness, further insight regarding the
merits of the standards can be provided by benefit-cost analysis. The
purpose of this section is to propose the methods to be used in
conducting an analysis of the economic benefits of the final rule for
heavy-duty vehicles and diesel fuel, and to discuss the potential for
economic benefits associated with the rule. While the quantification of
the benefits will not be available until the final rule, it is our
belief that, based on the similarity between today's proposed rule and
Tier 2/gasoline sulfur rule in terms of the costs per ton of emissions
reduced and types of health and welfare benefits expected, the health
and welfare benefits would substantially outweigh the costs.
1. What Is the Purpose of This Benefit-Cost Comparison?
Benefit-cost analysis (BCA) is a useful tool for evaluating the
economic merits of proposed changes in environmental programs and
policies. In its traditional application, BCA estimates the economic
``efficiency'' of proposed changes in public policy by organizing the
various expected consequences and representing those changes in terms
of dollars. Expressing the effects of these policy changes in dollar
terms provides a common basis for measuring and comparing these various
effects. Because improvement in economic efficiency is typically
defined to mean maximization of total wealth spread among all members
of society, traditional BCA must be supplemented with other analyses in
order to gain a full appreciation of the potential merits of new
policies and programs. These other analyses may include such things as
examinations of legal and institutional constraints and effects;
engineering analyses of technology feasibility, performance and cost;
or assessment of the air quality need.
In addition to the economic efficiency focus of most BCAs, the
technique is also limited in its ability to project future economic
consequences of alternative policies in a definitive way. Critical
limitations on the availability, validity, or reliability of data;
limitations in the scope and capabilities of environmental and economic
effect models; and controversies and uncertainties surrounding key
underlying scientific and economic literature all contribute to an
inability to estimate the economic effects of environmental policy
changes in exact and unambiguous terms. Under these circumstances, we
consider it most appropriate to view BCA as a tool to inform, but not
dictate, regulatory decisions such as the ones reflected in today's
proposed rule.
Despite the limitations inherent in BCA of environmental programs,
we consider it useful to analyze the potential benefits of today's
proposed action both in terms of physical changes in human health and
welfare and environmental change, and in terms of the estimated
economic value of those physical changes.
2. What Is Our Overall Approach to the Benefit-Cost Analysis?
The basic question we will seek to answer in the BCA is: ``What are
the net yearly economic benefits to society of the reduction in air
pollutant emissions likely to be achieved by the proposed rule for
heavy-duty vehicles and diesel fuel?'' In designing an analysis to
answer this question, we will model the benefits in a future year
(2030) that is representative of full-implementation of the program. We
will also adopt an analytical structure and sequence similar to that of
the benefit analysis for the Tier 2/gasoline sulfur rulemaking and used
for the ``section 812 studies'' \145\ to estimate the total benefits
and costs of the entire Clean Air Act. Moreover, we will use many of
the same models and assumptions actually used in the section 812
studies, and other Regulatory Impact Analyses (RIA's) prepared by the
Office of Air and Radiation. By adopting the major design elements,
models, and assumptions developed for the section 812 studies and other
RIA's, we will largely rely on methods which have already received
extensive review by the independent Science Advisory Board (SAB), by
the public, and by other federal agencies. In addition to the 2030
analysis, we plan to provide further characterization of the benefits
for the interim period between 2007 and 2030.
---------------------------------------------------------------------------

\145\ The ``section 812 studies'' refers to (1) US EPA, Report
to Congress: The Benefits and Costs of the Clean Air Act, 1970 to
1990, October 1997 (also known as the ``section 812 Retrospective);
and (2) the first in the ongoing series of prospective studies
estimating the total costs and benefits of the Clean Air Act (see
EPA report number: EPA-410-R-99-001, November 1999).
---------------------------------------------------------------------------

3. What Are the Significant Limitations of the Benefit-Cost Analysis?
Every BCA examining the potential effects of a change in
environmental protection requirements is limited to some extent by data
gaps, limitations in model capabilities (such as geographic coverage),
and uncertainties in the underlying scientific and economic studies
used to configure the benefit and cost models. Deficiencies in the
scientific literature often result in the inability to estimate changes
in health and environmental effects, such as potential increases in
premature mortality associated with increased exposure to carbon
monoxide. Deficiencies in the economics literature often result in the
inability to assign economic values even to those health and
environmental outcomes which can be quantified, such as changes in
visibility in residential areas. While these general uncertainties in
the underlying scientific and economics literatures will be discussed
in detail in the RIA for the final action, the key uncertainties are:
The exclusion of potentially significant benefit
categories (e.g., health and ecological benefits of incidentally
controlled hazardous air pollutants),
Errors in measurement and projection for variables such as
population growth,
Variability in the estimated relationships of health and
welfare effects to changes in pollutant concentrations.
In addition to these uncertainties and shortcomings which pervade
all analyses of criteria air pollutant control programs, a number of
limitations apply specifically to a BCA. Though we will use the best
data and models available, we will likely be required to adopt a number
of simplifying assumptions and to use data sets which, while reasonably
close, will not match precisely the conditions and effects expected to
result from implementation of the standards. For example, to estimate
the effects of the program at full implementation we will need to
project vehicle miles traveled and populations in the year 2030. These
assumptions may play a significant role in determining the magnitude of
the benefits estimate. In addition, the emissions data sets which

[[Page 35500]]

will be used for the analysis may not anticipate the emissions
reductions realized by other future actions and by expected near-future
control programs. For example, it is possible that the proposed heavy-
duty vehicle and diesel fuel sulfur standards will not be the governing
vehicle emissions standards in 2030. In the years before 2030, the
benefits from the proposed rule for heavy-duty vehicles and diesel fuel
will be less than in 2030 because the heavy-duty fleet will not be
fully phased in.
The key limitations and uncertainties unique to the BCA of the
final rule, therefore, will include:
Uncertainties in the estimation of future year emissions
inventories and air quality,
Uncertainties associated with the extrapolation of air
quality monitoring data to some unmonitored areas required to better
capture the effects of the standards on affected populations, and
Uncertainties associated with the effect of potential
future actions to limit emissions.
Despite these uncertainties, we believe the BCA will provide a
reasonable indication of the expected economic benefits of the proposed
rule for heavy-duty vehicles and diesel fuel in 2030 under one set of
assumptions. This is because the analysis will focus on estimating the
economic effects of the changes in air quality conditions expected to
result from today's proposed action, rather than focusing on developing
a precise prediction of the absolute levels of air quality likely to
prevail in 2030. An analysis focusing on the changes in air quality can
give useful insights into the likely economic effects of emission
reductions of the magnitude expected to result from today's proposed
rule.
4. How Will the Benefit-Cost Analysis Change From the Tier 2 Benefit-
Cost Analysis?
We will evaluate the economics and scientific literature prior to
conducting the benefit-cost analysis for the final rule. Our final
benefit-cost methodology will reflect the most up to date set of health
and welfare effects and the most current economic valuation methods. In
addition, we will use updated emission inventories. We will also be
evaluating the air quality models used to predict changes in future air
quality for use in the benefits analysis.
5. How Will We Perform the Benefit-Cost Analysis?
The analytical sequence begins with a projection of the mix of
technologies likely to be deployed to comply with the new standards,
and the costs incurred and emissions reductions achieved by these
changes in technology. The proposed rule for heavy-duty vehicles and
diesel fuel has various cost and emission related components. These
components would begin at various times and in some cases would phase
in over time. This means that during the early years of the program
there would not be a consistent match between cost and benefits. This
is especially true for the vehicle control portions of the program,
where the full vehicle cost would be incurred at the time of vehicle
purchase, while the cost for low sulfur diesel fuel along with the
emission reductions and benefits would occur throughout the lifetime of
the vehicle.
To develop a benefit-cost number that is representative of a fleet
of heavy-duty vehicles, we need to have a stable set of cost and
emission reductions to use. This means using a future year where the
fleet is fully turned over and there is a consistent annual cost and
annual emission reduction. For the proposed rule for heavy-duty
vehicles and diesel fuel, this stability would not occur until well
into the future. For this analysis, we selected the year 2030. The
resulting analysis will represent a snapshot of benefits and costs in a
future year in which the heavy-duty fleet consists almost entirely of
heavy-duty vehicles meeting the proposed standards. As such, it depicts
the maximum emission reductions (and resultant benefits) and among the
lowest costs that would be achieved in any one year by the program on a
``per mile'' basis. (Note, however, that net benefits would continue to
grow over time beyond those resulting from this analysis, because of
growth in population and vehicle miles traveled.) Thus, based on the
long-term costs for a fully turned over fleet, the resulting benefit-
cost ratio will be close to its maximum point (for those benefits which
we have been able to value).
To present a BCA, we are designing the cost estimate to reflect
conditions in the same year as the benefit valuation. Costs, therefore,
will be developed for the year 2030 fleet. For this purpose we will use
the long term cost once the capital costs have been recovered and the
manufacturing learning curve reductions have been realized, since this
will be the case in 2030.
We will also make adjustments in the costs to account for the fact
that there is a time difference between when some of the costs are
expended and when the benefits are realized. The vehicle costs are
expended when the vehicle is sold, while the fuel related costs and the
benefits are distributed over the life of the vehicle. We will resolve
this difference by using costs distributed over time such that there is
a constant cost per ton of emissions reduction and such that the net
present value of these distributed costs corresponds to the net present
value of the actual costs.
The resulting adjusted costs will be somewhat greater than the
expected actual annual cost of the program, reflecting the time value
adjustment. Thus, the costs will not represent expected actual annual
costs for 2030. Rather, they will represent an approximation of the
steady-state cost per ton that would likely prevail in that time
period. The benefit cost ratio for the earlier years of the program
would be expected to be lower than that based on these costs, since the
per-vehicle costs are larger in the early years of the program while
the benefits are smaller.
In order to estimate the changes in air quality conditions which
would result from these emissions reductions, we will develop two
separate, year 2030 emissions inventories to be used as inputs to the
air quality models. The first, baseline inventory, will reflect the
best available approximation of the county-by-county emissions for
NO_X, VOC, and SO2 expected to prevail in the year
2030 in the absence of the standards. To generate the second, control
case inventory, we will first estimate the change in vehicle emissions,
by pollutant and by county, expected to be achieved by the 2030 control
scenario described above. We will then take the baseline emissions
inventory and subtract the estimated reduction for each county-
pollutant combination to generate the second, control case emissions
inventory. Taken together, the two resulting emissions inventories will
reflect two alternative states of the world and the differences between
them will represent our best estimate of the reductions in emissions
which would result from our control scenario.
With these two emissions inventories in hand, the next step will be
to ``map'' the county-by-county and pollutant-by-pollutant emission
estimates to the input grid cells of appropriately selected air quality
and deposition models. One such model, called the Urban Airshed Model
(UAM), is designed to estimate the tropospheric ozone concentrations
resulting from a specific inventory of emissions of ozone precursor
pollutants, particularly NO_X and NMHC. Another model, called
the Climatological Regional Dispersion Model Source-Receptor Matrix
model (S-R Matrix), is designed to estimate the changes in ambient
particulate matter and visibility which would result from a specific
set of changes in emissions of primary

[[Page 35501]]

particulate matter and secondary particulate matter precursors, such as
SO2, NO_X, and NMHC. Also, nitrogen loadings to
watersheds can be estimated using factors derived from previous
modeling from the Regional Acid Deposition Model (RADM). By running
both the baseline and control case emissions inventories through models
such as these, we will be able to estimate the expected 2030 air
quality conditions and the changes in air quality conditions which
would result from the emissions reductions expected to be achieved by
the proposed rule for heavy-duty vehicles and diesel fuel.
After developing these two sets of year 2030 air quality profiles,
we will use the same health and environmental effect models used in the
section 812 studies to calculate the differences in human health and
environmental outcomes projected to occur with and without the proposed
standards. Specifically, we will use the Criteria Air Pollutant
Modeling System (CAPMS) to estimate changes in human health outcomes,
and the Agricultural Simulation Model (AGSIM) to estimate changes in
yields of a selected few agricultural crops. In addition, the impacts
of reduced visibility impairment and estimates of the effect of changes
in nitrogen deposition to a selection of sensitive estuaries will be
estimated using slightly modified versions of the methods used in the
section 812 studies. At proposal, we expect that several air quality-
related health and environmental benefits, however, will not be able to
be calculated for the BCA of today's proposed standards. Changes in
human health and environmental effects due to changes in ambient
concentrations of carbon monoxide (CO), gaseous sulfur dioxide
(SO2), gaseous nitrogen dioxide (NO2), and
hazardous air pollutants will likely not be included. In addition, some
health and environmental benefits from changes in ozone and PM may not
be included in our analysis (i.e., commercial forestry benefits).
However, if our review of the economics and scientific literature
reveals new information that will allow us to quantify these effects,
they will be considered for inclusion in the estimate of total benefits
for the final rule. Table IV-X lists the set of effects that we expect
to be able to quantify for the BCA of the final rule, along with those
effects which are known to exist, but that are currently
unquantifiable.
To characterize the total economic value of the reductions in
adverse effects achieved across the lower 48 states, we plan to use the
same set of economic valuation coefficients and models used in the
section 812 studies and the Tier 2 benefits analysis, as approved by
the SAB. The set of coefficients and their sources are listed in the
final Tier 2 RIA. However, any new methods uncovered in our evaluation
of the economic and scientific literature may be incorporated into our
final analysis. The net monetary benefits of the proposed rule for
heavy-duty vehicles and diesel fuel will then be calculated by
subtracting the estimated costs of compliance from the estimated
monetary benefits of the reductions in adverse health and environmental
effects.
The last step of the analysis will be to characterize the
uncertainty surrounding our estimate of benefits. Again, we will follow
the recommendations of the SAB for the presentation of uncertainty.
They recommend that a primary estimate should be presented along with a
description of the uncertainty associated with each endpoint.
Therefore, for the final rule for heavy-duty vehicles and diesel
fuel, the benefit analysis will adopt an approach similar to the
section 812 study and the final Tier 2/gasoline sulfur benefit-cost
analysis. Our analysis will first present our estimate for a primary
set of benefit endpoints followed by a presentation of ``alternative
calculations'' of key health and welfare endpoints to characterize the
uncertainty in this primary set. However, the adoption of a value for
the projected reduction in the risk of premature mortality is the
subject of continuing discussion within the economic and public policy
analysis community within and outside the Administration. In response
to the sensitivity on this issue, we will provide estimates reflecting
two alternative approaches. The first approach--supported by some in
the above community and preferred by EPA--uses a Value of a Statistical
Life (VSL) approach developed for the Clean Air Act section 812
benefit-cost studies. This VSL estimate of $5.9 million (1997$) was
derived from a set of 26 studies identified by EPA using criteria
established in Viscusi (1992), as those most appropriate for
environmental policy analysis applications.

Table V.G-1.--Human Health and Welfare Effects of Pollutants Affected by the Proposed Heavy-duty Vehicle Rule
--------------------------------------------------------------------------------------------------------------------------------------------------------
Alternative quantified and/or
Pollutant Quantified and monetized effects monetized effects Unquantified effects
--------------------------------------------------------------------------------------------------------------------------------------------------------
Ozone Health..................... Minor restricted activity days/acute ...................................... Premature mortality; ^a Increased
respiratory symptoms; Hospital airway responsiveness to stimuli;
admissions--respiratory and Inflammation in the lung; Chronic
cardiovascular; Emergency room visits respiratory damage; Premature aging
for asthma. of the lungs; Acute inflammation and
respiratory cell damage; Increased
susceptibility to respiratory
infection; Non-asthma respiratory
emergency room visits.
Ozone Welfare.................... Decreased worker productivity; ...................................... Decreased yields for commercial
Decreased yields for commercial forests; Decreased yields for fruits
crops. and vegetables.
PM Health........................ Premature mortality; Bronchitis-- ...................................... Infant mortality; Low birth weight;
chronic and acute; Hospital Changes in pulmonary function;
admissions--respiratory and Chronic respiratory diseases other
cardiovascular; Emergency room visits than chronic bronchitis;
for asthma; Lower and upper Morphological changes; Altered host
respiratory illness; Shortness of defense mechanisms; Cancer; Non-
breath; Minor restricted activity asthma respiratory emergency room
days/acute respiratory symptoms; Work visits.
loss days.

[[Page 35502]]

PM Welfare....................... Visibility in California, Visibility in Northeastern, .....................................
Southwestern, and Southeastern Class Northwestern, and Midwestern Class I
I areas. areas; Household soiling.
Nitrogen and Sulfate Deposition ...................................... Costs of nitrogen controls to reduce Impacts of acidic sulfate and nitrate
Welfare. eutrophication in selected eastern deposition on commercial forests;
estuaries. Impacts of acidic deposition to
commercial freshwater fishing;
Impacts of acidic deposition in
terrestrial ecosystems; Impacts of
nitrogen deposition on commercial
fishing, agriculture, and forests;
Impacts of nitrogen deposition on
recreation in estuarine ecosystems;
Reduced existence values for
currently healthy ecosystems.
CO Health........................ ...................................... ...................................... Premature mortality; ^a Behavioral
effects; Hospital admissions--
respiratory, cardiovascular, and
other; Other cardiovascular effects;
Developmental effects; Decreased
time to onset of angina.
HAPS Health...................... ...................................... ...................................... Cancer (benzene, 1,3-butadiene,
formaldehyde, acetaldehyde); Anemia
(benzene); Disruption of production
of blood components (benzene);
Reduction in the number of blood
platelets (benzene); Excessive bone
marrow formation (benzene);
Depression of lymphocyte counts
(benzene); Reproductive and
developmental effects (1,3-
butadiene); Irritation of eyes and
mucus membranes (formaldehyde);
Respiratory irritation
(formaldehyde); Asthma attacks in
asthmatics (formaldehyde).
HAPS Welfare..................... ...................................... ...................................... Direct toxic effects to animals;
Bioaccumlation in the food chain.
--------------------------------------------------------------------------------------------------------------------------------------------------------
^a Premature mortality associated with ozone is not separately included in this analysis. It is assumed that the Pope, et al. C-R function for premature
mortality captures both PM mortality benefits and any mortality benefits associated with other air pollutants.

An alternative, age-adjusted approach is preferred by some others
in the above community both within and outside the Administration. This
approach was also developed for the Section 812 studies and addresses
concerns with applying the VSL estimate--reflecting a valuation derived
mostly from labor market studies involving healthy working-age manual
laborers--to PM-related mortality risks that are primarily associated
with older populations and those with impaired health status. This
alternative approach leads to an estimate of the value of a statistical
life year (VSLY), which is derived directly from the VSL estimate. It
differs only in incorporating an explicit assumption about the number
of life years saved and an implicit assumption that the valuation of
each life year is not affected by age.\146\ The mean VSLY is $360,000
(1997$); combining this number with a mean life expectancy of 14 years
yields an age-adjusted VSL of $3.6 million (1997$).
---------------------------------------------------------------------------

\146\ Specifically, the VSLY estimate is calculated by
amortizing the $5.9 million mean VSL estimate over the 35 years of
life expectancy asssociated with subjects in the labor market
studies. The resulting estimate, using a 5 percent discount rate, is
$360,000 per life-year saved in 1997 dollars. This annual average
value of a life-year is then multiplied times the number of years of
remaining life expectancy for the affected population (in the case
of PM-related premature mortality, the average number of $ life-
years saved is 14).
---------------------------------------------------------------------------

Both approaches are imperfect, and raise difficult methodological
issues which are discussed in depth in the recently published Section
812 Prospective Study, the draft EPA Economic Guidelines, and the peer-
review commentaries prepared in support of each of these documents. For
example, both methodologies embed assumptions (explicit or implicit)
about which there is little or no definitive scientific guidance. In
particular, both methods adopt the assumption that the risk versus
dollars trade-offs revealed by available labor market studies are
applicable to the risk versus dollar trade-offs in an air pollution
context.
EPA currently prefers the VSL approach because, essentially, the
method reflects the direct application of what EPA considers to be the
most reliable estimates for valuation of premature mortality available
in the current economic literature. While there are several differences
between the labor market studies EPA uses to derive a VSL estimate and
the particulate matter air pollution context addressed here, those
differences in the affected populations and the nature of the risks
imply both upward and downward adjustments. For example, adjusting for
age differences may imply the need to adjust the $5.9 million VSL
downward as would adjusting for health differences, but the involuntary
nature of air pollution-related risks and the lower level of risk-
aversion of the

[[Page 35503]]

manual laborers in the labor market studies may imply the need for
upward adjustments. In the absence of a comprehensive and balanced set
of adjustment factors, EPA believes it is reasonable to continue to use
the $5.9 million value while acknowledging the significant limitations
and uncertainties in the available literature. Furthermore, EPA prefers
not to draw distinctions in the monetary value assigned to the lives
saved even if they differ in age, health status, socioeconomic status,
gender or other characteristic of the adult population.
Those who favor the alternative, age-adjusted approach (i.e. the
VSLY approach) emphasize that the value of a statistical life is not a
single number relevant for all situations. Indeed, the VSL estimate of
$5.9 million (1997 dollars) is itself the central tendency of a number
of estimates of the VSL for some rather narrowly defined populations.
When there are significant differences between the population affected
by a particular health risk and the populations used in the labor
market studies--as is the case here--they prefer to adjust the VSL
estimate to reflect those differences. While acknowledging that the
VSLY approach provides an admittedly crude adjustment (for age though
not for other possible differences between the populations), they point
out that it has the advantage of yielding an estimate that is not
presumptively biased. Proponents of adjusting for age differences using
the VSLY approach fully concur that enormous uncertainty remains on
both sides of this estimate--upwards as well as downwards--and that the
populations differ in ways other than age (and therefore life
expectancy). But rather than waiting for all relevant questions to be
answered, they prefer a process of refining estimates by incorporating
new information and evidence as it becomes available.
The presentation of the alternative calculations for certain
endpoints will demonstrate how much the overall benefit estimate might
vary based on the value EPA gives to a parameter (which has some
uncertainty associated with it) underlying the estimates for human
health and environmental effect incidence and the economic valuation of
those effects. These alternative calculations will represent conditions
that are possible to occur, however, EPA has selected the best
supported values based on current scientific literature for use in the
primary estimate. The alternate calculations will include:
Presentation of an estimated confidence interval around
the Primary estimate of benefits to characterize the standard error in
the C-R and valuation studies used in developing benefit estimates for
each endpoint;
Valuing PM-related premature mortality based on a
different C-R study;
Value of avoided premature mortality incidences based on
statistical life years;
Consideration of reversals in chronic bronchitis treated
as lowest severity cases;
Value of visibility changes in all Class I areas;
Value of visibility changes in Eastern U.S. residential
areas;
Value of visibility changes in Western U.S. residential
areas;
Value of reduced household soiling damage; and
Avoided costs of reducing nitrogen loadings in east coast
estuaries.
For instance, the estimate of the relationship between PM exposure
and premature mortality from the study by Dockery, et al. is a
plausible alternative to the Pope, et al. study used for the Primary
estimate of benefits. The SAB has noted that ``the study had better
monitoring with less measurement error than did most other studies''
(EPA-SAB-COUNCIL-ADV-99-012, 1999). The Dockery study had a more
limited geographic scope (and a smaller study population) than the
Pope, et al. study and the Pope study appears more likely to mitigate a
key source of potential confounding. The Dockery study also covered a
broader age category (25 and older compared to 30 and older in the Pope
study) and followed the cohort for a longer period (15 years compared
to 8 years in the Pope study). For these reasons, the Dockery study is
considered to be a plausible alternative estimate of the avoided
premature mortality incidences that are expected to be associated with
the final heavy-duty rule rule. The alternative estimate for mortality
can be substituted for the valuation component in our primary estimate
of mortality benefits to observe how the net benefits of the program
may be influenced by this assumption. Unfortunately, it is not possible
to combine all of the assumptions used in the alternate calculations to
arrive at different total benefit estimates because it is highly
unlikely that the selected combination of alternative values would all
occur simultaneously. Therefore, it will be more appropriate to
consider each alternative calculation individually to assess the
uncertainty in the estimate.
In addition to the estimate for the primary set of endpoints and
alternative calculations of benefits, our RIA for the final rule will
also present an appendix with supplemental benefit estimates and
sensitivity analyses of other key parameters in the benefit analysis
that have greater uncertainty surrounding them due to limitations in
the scientific literature. Supplemental estimates will be presented for
premature mortality associated with short-term exposures to PM and
ozone, asthma attacks, occurrences of moderate or worse asthma
symptoms, and the avoided incidences of premature mortality in infants.
Even with our efforts to fully disclose the uncertainty in our
estimate, this uncertainty presentation method does not provide a
definitive or complete picture of the true range of monetized benefits
estimates. This proposed approach, to be implemented in the BCA for the
final rule, will not reflect important uncertainties in earlier steps
of the analysis, including estimation of compliance technologies and
strategies, emissions reductions and costs associated with those
technologies and strategies, and air quality and deposition changes
achieved by those emissions reductions. Nor does this approach provide
a full accounting of all potential benefits associated with the
proposed rule for heavy-duty vehicles and diesel fuel, due to data or
methodological limitations. Therefore, the uncertainty range will only
be representative of those benefits that we will be able to quantify
and monetize.
6. What Types of Results Will Be Presented in the Benefit-Cost
Analysis?
The BCA for the final rule for heavy-duty vehicles and diesel fuel
will reflect a single year ``snapshot'' of the yearly benefits and
costs expected to be realized once the standards have been fully
implemented and non-compliant vehicles have all been retired. Near-term
costs will be higher than long-run costs as vehicle manufacturers and
oil companies invest in new capital equipment and develop and implement
new technologies. In addition, near-term benefits will be lower than
long-run benefits because it will take a number of years for compliant
heavy-duty vehicles to fully displace older, more polluting vehicles.
However, we will adjust the cost estimates upward to compensate for
some of this discrepancy in the timing of benefits and costs and to
ensure that the long-term benefits and costs are calculated on a
consistent basis. Because of the adjustment process, the cost estimates
should not be interpreted as reflecting the actual costs expected to be
incurred in the year 2030. Actual program costs can be found earlier in
this preamble.
With respect to the benefits, the BCA for the final rule for heavy-
duty vehicles and diesel fuel will follow the

[[Page 35504]]

presentation format used in the Tier 2 BCA, presenting several
different measures of benefits which will be useful to compare and
contrast to the estimated compliance costs. These benefit measures
include (a) the tons of emissions reductions achieved, (b) the
reductions in incidences of adverse health and environmental effects,
and (c) the estimated economic value of those reduced adverse effects.
Calculating the cost per ton of pollutant reduced is particularly
useful for comparing the cost-effectiveness of the new standards or
programs against existing programs or alternative new programs
achieving reductions in the same pollutant or combination of
pollutants. Considering the absolute numbers of avoided adverse health
and environmental effects can also provide valuable insights into the
nature of the health and environmental problem being addressed by the
proposed rule as well as the magnitude of the total public health and
environmental gains potentially achieved. Finally, when considered
along with other important economic dimensions--including environmental
justice, small business financial effects, and other outcomes related
to the distribution of benefits and costs among particular groups--the
direct comparison of quantified economic benefits and economic costs
can provide useful insights into the potential magnitude of the
estimated net economic effect of the rule, keeping in mind the limited
set of effects we expect to be able to monetize.

VI. Alternative Program Options

In the course of developing the proposal, we considered a broad
range of options, many of which were raised by commenters on the ANPRM.
Various options were considered for the best manner to implement a
change to diesel fuel, on how to structure a sulfur standard, on fuel
changes other than sulfur, and on the geographic scope of the program.
This section helps to explain many alternative program options that we
considered in designing today's proposal. In this section, we also are
seeking comment on voluntary phase-in options for implementing the fuel
program (see section VI.A.2), and on issues connected with the use of
JP-8 fuel in highway-going military vehicles (see section VI.D).

A. What Other Fuel Implementation Options Have We Considered?

A broad spectrum of approaches for implementing the fuel program
were either raised by the Agency in the ANPRM, received as public
comments on the ANPRM, or raised by various parties during the
development of this proposal. Below, we discuss some of the options we
have considered, including alternatives on which we are seeking
comment.
1. What Are the Advantages and Disadvantages of a Phase-in Approach to
Implementing the Low Sulfur Fuel Program?
EPA is proposing, as discussed in section IV.C., that the entire
pool of highway diesel fuel be converted to low sulfur diesel fuel all
at once in 2006. In the early years of the program, the use of low
sulfur diesel fuel will result in reductions in the amount of direct
and secondary particulate matter from the existing fleet of heavy-duty
vehicles. Nevertheless, the primary benefit of the fuel change is the
emission reductions that would occur over time from the new vehicle
fleet as a result of the enablement of advanced aftertreatment exhaust
emission control technologies. Consequently, we believe there may be
some advantages, particularly in the early years, to allowing some
flexibility in the program so that not all of the highway diesel fuel
pool must be converted to low sulfur all at once. First, owners of old
vehicles could continue to refuel on higher-sulfur (500 ppm) diesel
fuel, potentially saving money for consumers. Second, we believe a
phase-in approach, if designed properly, has the potential to be
beneficial for refiners, by reducing the fuel production costs in the
early years of the program. This flexibility could reduce operating
costs, if the entire volume of highway fuel does not have to meet the
low sulfur standard. If coupled with averaging, banking and trading
provisions, some refineries may be able to delay desulfurization
investments for several years. Even for refiners planning to
desulfurize their entire highway fuel pool to low sulfur levels at the
beginning of the program, there may be circumstances where the actual
fuel produced is slightly off-spec (i.e., above the low sulfur
standard). A phase-in approach could allow refiners to continue selling
that fuel to the highway market (as 500 ppm fuel), rather than to other
distillate markets. Refiners could also have more flexibility to
continue producing highway diesel (as 500 ppm fuel) during unit
downtime (e.g., turnarounds and upsets).
While a phase-in approach could provide flexibility for refiners
and potentially lower costs for consumers, a number of concerns would
need to be addressed before such an approach could be implemented.
These include: ensuring sufficient availability of the low sulfur fuel
when and where it is needed, minimizing the potential for misfueling,
minimizing the risk of spot outages, and minimizing impacts on the fuel
distribution and retail industries. These issues are discussed further
below. It is not obvious at what level the fuel production and
distribution systems can provide two grades of highway diesel fuel
while minimizing the potential for localized supply shortages and price
spikes, and misfueling problems. For example, we expect that in the
first year of the program only about 10 percent of highway diesel fuel
would be consumed by 2007 model year vehicles requiring the use of low
sulfur fuel. In a perfect world where the distribution system could,
without additional cost, make low sulfur diesel fuel widely available
(in addition to the current 500 ppm fuel), only about 10 percent of the
highway diesel fuel produced by refiners in the first year would then
have to be low sulfur. Unfortunately, since this perfect world does not
exist, the question remains whether, and to what extent, the system can
distribute two grades of highway diesel fuel in a way that takes
advantage of any flexibilities offered, and ensures sufficient supply
of fuel for the new vehicles that need it.
During the process of developing this proposal (including comments
received on the ANPRM), many industry stakeholders (many diesel
distributors, marketers, larger refiners, and end-users such as
truckers and centrally-fueled fleets) have commented on ways to
implement the fuel program. While each stakeholder may have had
different assumptions behind their position (including assumptions
about the structure of a phase-in, and expectations about the resulting
costs and fuel prices), many stakeholders have encouraged EPA to
implement any fuel change all at once, rather than incur the added
distribution costs and marketplace complication of phasing in a new
grade of highway diesel fuel. The following sections discuss some of
the challenges in implementing a phase-in approach.

a. Availability of Low Sulfur Diesel Fuel

Because new vehicles would need to be fueled exclusively with low
sulfur diesel, for a phase-in approach to be workable, low sulfur
diesel fuel would have to be available in all parts of the country. It
is not clear what minimum level of availability would be necessary to
meet the needs of diesel vehicles. The trucking industry has indicated
that a limited number of phased-in fueling locations would not meet the
needs of the trucking industry.

[[Page 35505]]

We seek comment on what level of availability would be appropriate
under a phase-in approach, to ensure that the low sulfur diesel fuel is
available, within a reasonable distance, to all consumers in all parts
of the country. For example, would sufficient availability be achieved
if all major truck stops across the country offered low sulfur fuel, or
if some minimum percentage of diesel retailers in different geographic
areas offered low sulfur fuel? Are there studies on fuel availability
that would serve to inform efforts to assure adequate availability? We
request that commenters consider what fraction of truck stops and other
retail outlets would need to make low sulfur fuel available within any
given area in order to ensure reasonable availability from the public's
perspective.

b. Misfueling

Any phase-in approach would introduce an additional grade of
highway diesel fuel into the market, by allowing both high and low-
sulfur grades to coexist, with a potential for a price differential
between the grades. Many industry stakeholders, including diesel
marketers, truck stop operators, and engine manufacturers, have
commented that misfueling would be significant under a phase-in
approach.\147\ That is, customers with new vehicles that need low-
sulfur fuel might use the higher-sulfur fuel, mistakenly or
deliberately, which could increase emissions and damage the emissions
control technology on the vehicle. Diesel marketers have also raised
the issue that a phase-in system could create incentives for consumers
to tamper with the emission control equipment of new vehicles, if they
believe that will enable them to use a lower priced fuel. Therefore, we
are concerned about the potential for misfueling, as it could reduce
the emission benefits of the program. However, if a phase-in approach
were to work well and misfueling were not an issue, we would expect to
achieve the same environmental benefits as the proposed single fuel
approach.
---------------------------------------------------------------------------

\147\ Comment letters from the Engine Manufacturers Association
(Item II-D-35), National Association of Truck Stop Operators
(Included in Report of the Small Business Advocacy Review (SBAR)
Panel, Appendix B, Page 30), and Petroleum Marketers Association of
America (Included in SBAR Panel Report, Appendix B, Page 38).
---------------------------------------------------------------------------

Some degree of misfueling occurs even today with a single grade of
highway diesel fuel, due to the availability of tax exempt off-highway
diesel fuel. The opportunity for misfueling with off-highway diesel
fuel, however, is somewhat limited by the limited number of highway
diesel refueling locations that market both grades of diesel fuel.
Nevertheless, since off-highway diesel fuel will still be available
even under a complete switch of highway diesel fuel to low sulfur, the
problem of misfueling is not entirely unique to the phase-in approach.
It is, however, true that the greater availability of 500 ppm diesel
fuel alongside the low sulfur fuel will make misfueling easier. Thus,
the appropriate question to ask when considering a phase-in approach is
not ``will people misfuel?'' but ``to what extent?'' and ``how can the
design of the program minimize the potential for misfueling?''
One factor that might encourage misfueling would be the existence
of a price differential between low sulfur diesel fuel and 500 ppm
fuel. For many diesel vehicles, particularly line-haul tractor
trailers, the fuel cost can be as much as 20 percent of annual
operating costs, so drivers have a strong incentive to save on fuel
costs. On the other hand, there are also several factors that might
serve as a deterrent to misfueling. First, the potential risk
associated with voiding a manufacturer emission warranty or damaging
the engine and exhaust system on an expensive vehicle might cause
owners and operators of heavy-duty trucks to be more circumspect in
ensuring that their vehicles are fueled properly. Second, misfueled
vehicles could experience a loss in performance, such as poor
acceleration or even engine stalling (as discussed in section
III.F.1.a). Third, under the proposed regulations it would be unlawful
for any person to misfuel.
Depending on the potential for misfueling, EPA may need to require
that new vehicles be fitted with a unique nozzle interface, with a
corresponding size nozzle for the low-sulfur diesel. This would be
analogous to the nozzle interface approach used to discourage
misfueling in the unleaded gasoline program. However, diesel marketers
have indicated that they do not support the use of unique nozzle
interfaces for the low sulfur fuel, particularly if it would affect
volume delivery. They have expressed the concern that a smaller nozzle
size would reduce the volume of fuel delivered, result in slower
refuelings, and increase wait times at retail stations. Further, based
on our experience with unleaded gasoline,\148\ it is likely that people
intent on misfueling would quickly find ways around a unique nozzle/
nozzle interface. We request comment on ways to structure a unique
nozzle/nozzle interface approach that would discourage misfueling while
avoiding these problems. We also request comment on any alternative
methods that could be used to discourage misfueling.
---------------------------------------------------------------------------

\148\ ``An Analysis of the Factors Leading to the Use of Leading
to the Use of Leaded Gasoline in Automobiles Requiring Unleaded
Gasoline,'' September 29, 1978, Sobotka & Company, Inc. See also
``Motor Vehicle Tampering Survey--1983,'' July 1984, U.S. EPA,
Office of Air and Radiation, Docket A-99-06. See also ``Anti-
Tampering and Anti-Misfueling Programs to Reduce In-Use Emissions
From Motor Vehicles,'' May 25, 1983 (EPA/AA/83-3). Contained in
Docket A-99-06.
---------------------------------------------------------------------------

We invite comment on the potential for misfueling under phase-in
approaches, what factors would influence misfueling, and how the
potential for misfueling might vary under the different phase-in
approaches described in subsection 2 below. We further seek comment on
how these phase-in approaches could be designed to minimize the
potential for misfueling.

c. Distribution System Impacts

While providing flexibility for refiners and potentially lower
costs to consumers, a phase-in approach would rely on the fuel
distribution infrastructure being able to accommodate the second grade
of highway diesel fuel. The economics of modifying the distribution
infrastructure to handle two grades of highway diesel fuel would affect
the extent to which refiners can take advantage of the flexibility, and
consumers enjoy the cost-savings, of a phase-in. There are a vast array
of businesses in the diesel fuel distribution system, encompassing
thousands of companies, including pipelines, bulk terminals, bulk
plants, petroleum marketers (who carry the fuel from bulk terminals and
bulk plants via transport trucks and fuel tank wagons to retail outlets
and fleet customers), fuel oil dealers, service stations, truck stops,
and centrally-fueled fleets (commercial fleets, federal/state/local
government fleets, and farms). Based on available data, the vast
majority of these are small businesses according to the Small Business
Administration's definitions.\149\ These businesses may make
investments and change their practices to accommodate two grades of
highway diesel fuel. The economics of a phase-in could be viewed as
follows: Through intermediate price mark-ups on the product, the system
would distribute some of the cost savings experienced by the refiners
and consumers to those making capital investments. If the potential
cost savings

[[Page 35506]]

were not sufficient to justify such investments, then those investments
would not occur and the entire system would convert to low sulfur
diesel. We seek comment on how the economics of a phase-in would
actually play out.
---------------------------------------------------------------------------

\149\ For more information, see the Report of the Small Business
Advocacy Review Panel, contained in the docket.
---------------------------------------------------------------------------

If the cost savings of a phase-in are substantial, many bulk
terminals and bulk plants may find it economical to add new tank
capacity to accommodate a second grade of highway diesel fuel. However,
if the cost savings of a phase-in are modest, fewer terminal operators
would profit from such investments, since some have commented on the
costs, space constraints, and permitting difficulties associated with
new tankage.\150\ The magnitude of the cost savings also affects the
role of diesel marketers in this market. Some marketers have commented
that if some terminals offer two grades while others offer only one
grade, the costs of transporting fuel would increase since some trucks
would have to travel greater distances to alternate terminals or bulk
plants.\151\ The share of the cost savings that marketers could enjoy
from the mark-up on diesel products would have to at least equal the
higher transport costs for them to offer to handle two grades of fuel.
---------------------------------------------------------------------------

\150\ Letter from Independent Terminal Operators Association,
July 13, 1999 (Item # II-D-80).
\151\ Letter from Petroleum Marketers Association of America,
November 8, 1999, Docket A-99-06.
---------------------------------------------------------------------------

Similarly, many service stations, truck stops, and centrally-fueled
fleets would be faced with a decision of whether to add additional
underground storage tanks to carry the extra grade of diesel fuel.
Retailers with more than one diesel tank, such as many truck stops and
some fleets, could choose to demanifold existing tanks (involving
breaking concrete) in order to dedicate one or more tanks to the new
fuel. Those that find it economical to do so will undertake the
investment and offer two grades, while those that would not find the
investment profitable would forego this option.
Generally we would expect that where businesses could profit from
managing two grades they would do so and provide some 500 ppm diesel to
the market. Thus, the impact to the distribution system of a phase-in
would include costs from new investments, but these could be
compensated by higher profits. Where the costs of handling two fuels in
the distribution system are larger than the cost savings enjoyed by
refineries (and passed down to consumers in lower fuel prices), then
only low sulfur diesel would be offered. Some refiners and distributors
have expressed the concern, however, that these additional investments
would be ``stranded'' after the phase-in period ends. A key question
will be whether each party in the refining/distribution system can
accurately anticipate what the others will do, so as to avoid
unnecessary investments (e.g., if the system should switch over the low
sulfur more quickly than expected). Since the diesel fleet transitions
over relatively quickly (greater than 50 percent of VMT is typically
driven by new diesel vehicles after just 5 years), there may be limited
time to recoup any investment made to handle an additional grade of
highway diesel fuel. We request comment overall on the economics of a
phase-in approach.
In addition to overall impacts on the distribution system, an
additional grade of highway diesel fuel could reduce the flexibility of
the distribution system to carry all grades of fuels that it does
today. This may particularly be a concern with specialty fuels or
segregated shipments of fuel through pipelines that require separate
tankage such as those utilized by the Department of Defense (DOD). DOD
stated that since its specialty fuels (F-76, JP-5, and JP-8) are not
fungible fuels, if today's rule places additional stress on an already
capacity-strained pipeline system, it may limit DOD's ability to
transport adequate volumes of their specialty fuels to meet operational
readiness requirements. Consequently we request comment on this
particular impact on the distribution system in regard to accommodating
a second grade of highway diesel fuel.

d. Uncertainty in the Transition to Low Sulfur

We believe the proposed single fuel approach provides more
certainty to the market for making the large investments needed to
introduce low-sulfur fuel. Yet even under a single fuel approach,
refiners have indicated that there is uncertainty in refiner decisions
to invest or not (or to underinvest) in desulfurization, which could
lead to a risk of supply shortfalls and high prices. Refiners may make
this choice to exit the highway diesel market, or to reduce production
volume of highway diesel fuel, especially if faced with uncertainty
about the ability to recover their investments (see further discussion
in section V.D.1). A phase-in approach could minimize any potential for
such a shortfall in the overall highway diesel fuel supply. Under a
phase-in, the level of uncertainty is different, however, in that since
the highway diesel pool would be split into two grades, refiners would
need to predict in advance the relative demand for each grade.
Under the phase-in flexibility approaches (described in the
following section), the presumption is that the fuel production and
distribution system will react to both the market demand and the
incentive of the various programs to produce and distribute the low
sulfur fuel at reasonable prices to all parts of the country. Turning
any of these approaches into a reality requires embracing the
possibility that the market reacts differently than anticipated. For
example, diesel retailers have indicated that it would be extremely
difficult to predict how retailers would respond to making low sulfur
fuel available, given the many factors that influence retail decisions.
Consequently, refiners might have little certainty about continued
markets for 500 ppm fuel when making their investment decisions and all
of them might choose to convert to low sulfur. Given the lead time
needed for additional desulfurization capacity at refineries to come on
line, it is important for a smooth transition to low sulfur diesel fuel
that predictions of demand be similar to the actual demand. Each of the
phase-in approaches described in the following section is intended to
be designed to allow the market the flexibility to find a lower cost
option than full initial conversion to low sulfur fuel if such a
solution exists, and to default to a full low sulfur program if such a
solution does not exist. Each approach is, however, subject to
different sources of uncertainty. We request comment on the ability of
refiners to accurately predict demand for desulfurization capacity
under a phase-in approach. Commenters should discuss this issue in the
context of the phase-in approaches described below and in the context
of the proposed single fuel approach.

e. Cost Considerations Under a Phase-in Approach

Because it avoids the need to produce all of the fuel to the low
sulfur standard in the first year, a phase-in approach could provide an
opportunity for cost savings to refiners and could significantly lower
overall diesel fuel production costs. Consumers of pre-2007 diesel
vehicles could also realize a savings if the current 500 ppm fuel were
still available and priced lower than the new low sulfur fuel. In a
perfect world with a distribution system capable of distributing a
second grade of highway diesel fuel at no cost, if low sulfur
production could be matched with the demand from new vehicles, the
fraction of highway diesel fuel that would have to be low sulfur would
increase from approximately 9% in

[[Page 35507]]

2007 to approximately 60% in 2012 based on typical fleet turnover
rates. Thus, the amount of low sulfur fuel refiners would have to
produce in the early years of the program could be reduced
significantly, with a corresponding reduction in production costs
theoretically as high as $4 billion, using our estimated per gallon
fuel costs discussed in section IV. This theoretical distribution
system does not exist and there would be a number of important and
potentially significant costs incurred in the distribution system that
could impact these savings. As discussed above, a wide array of
entities in the distribution system, including refiners, bulk
terminals, pipelines, bulk plants, petroleum marketers, fuel oil
dealers service stations, truck stops, and centrally fuelled fleets
would have to make investment decisions in order to distribute a second
grade of highway diesel fuel. We seek comment on the potential cost
savings associated with a phase-in approach, including the potential
costs of managing two grades of highway diesel fuel in the distribution
system, how these costs would vary depending on the relative volumes of
the two grades of highway diesel fuel, the necessary margin for
businesses in the distribution system to find it economic to manage two
grades of highway fuel, and how these cost savings and margins could
vary depending on the range of ways the distribution system might
respond.
2. What Phase-in Options Is EPA Seeking Comment on in Today's Proposal?
In this section, we are requesting comment on three different
phase-in approaches for implementing a program for low sulfur highway
diesel fuel.

a. Refiner Compliance Flexibility

Despite the concerns described above with a phase-in approach for
implementing the diesel fuel sulfur control program, EPA nevertheless
believes that a program, if voluntary, can be devised which can address
these concerns and take advantage of at least some of the benefits a
phase-in approach has to offer. Consequently, as part of our proposed
program for implementing low sulfur highway diesel, as described in
section IV.C, we also are seeking comment on a voluntary option that
would provide compliance flexibilities for refiners, while still
achieving the environmental benefits of the program. In this section,
we describe this refiner compliance flexibility concept and seek
comment on all aspects of its design. We also discuss how this
compliance flexibility relates to the options for small refiner
flexibility (which we're seeking comment on in section VIII.E).

i. Overview of Compliance Flexibility

We are seeking comment on a voluntary compliance flexibility that
would allow refiners to continue producing fuel at the 500 ppm level
for a fraction of their total highway diesel fuel volume in the first
few years of the program. The fraction of 500 ppm fuel allowed to be
produced by refiners would phase-down over a period of several years.
Specifically, we request comment on the appropriate fraction of highway
diesel fuel allowed to be produced as 500 ppm fuel beginning in 2006.
Three possible scenarios are shown in Table VI.A-1 below. The level at
which this flexibility begins would significantly affect its design. We
are seeking comment on a range of production percentages for the 500
ppm fuel. We are particularly interested in the degree to which
percentages of 500 ppm at the higher end of this range could pose
challenges for ensuring sufficient availability of the low sulfur fuel
and minimizing the potential for misfueling. In addition, we request
comment on the extent to which different proportions of 500 ppm fuel
will pose different challenges for the distribution system. Several
issues and implications of setting the 500 ppm production limits at
higher or lower levels are discussed below. We seek comment on our
assumptions and the implications of these issues for the design of such
a compliance flexibility program. Further, we request comment on the
number of years this flexibility should be provided.

Table VI.A-1.--Two Possible Scenarios for Implementing the Compliance Flexibility
----------------------------------------------------------------------------------------------------------------
Percent of highway diesel fuel permitted to be 500 ppm
--------------------------------------------------------------
2006 2007 2008 2009 2010 2011 2012
----------------------------------------------------------------------------------------------------------------
Scenario A....................................... 20 20 10 10 0 0 0
Scenario B....................................... 50 50 30 15 0 0 0
Scenario C....................................... 75 75 60 45 30 15 0
----------------------------------------------------------------------------------------------------------------

We believe this compliance flexibility would be potentially
beneficial for refiners. This flexibility could reduce operating costs,
by not requiring the entire volume of highway fuel to meet the low
sulfur standard. With averaging, banking and trading provisions as a
component of this compliance flexibility (as discussed below), some
refineries may be able to delay desulfurization investments for several
years. Even for refiners planning to desulfurize their entire highway
fuel pool to low sulfur levels at the beginning of the program, there
may be circumstances where the actual fuel produced is slightly off-
spec (i.e., above the low sulfur standard). This flexibility would
allow refiners to continue selling that fuel to the highway market (as
500 ppm fuel), rather than to other distillate markets. Refiners would
also have more flexibility to continue producing highway diesel (as 500
ppm fuel) during unit downtime (e.g., turnarounds and upsets).
This approach would need appropriate safeguards to minimize
contamination of the low sulfur fuel and misfueling. Thus, low sulfur
highway diesel would have to remain a segregated product throughout its
distribution (see further discussion of segregation requirements in
section VI.A.2.a.v). Further, any retail pumps carrying 500 ppm fuel
would have to be prominently labeled to prevent misfueling of 2007 and
later model year vehicles. We seek comment on whether other measures to
discourage misfueling might also be necessary. For example, the use of
a unique refueling nozzle/vehicle nozzle interface could further
discourage misfueling, although we question the need to pursue this
approach if the 500 ppm fuel were in the market in relatively low
volumes and only during the initial years when new vehicles still
comprise a relatively small percent of the fleet. Other issues
regarding the potential for misfueling are discussed in subsection 1
above.
We also propose an averaging, banking and trading (ABT) program as
part of this compliance flexibility. Refiners owning more than one
refinery would be allowed to average their

[[Page 35508]]

production volumes across refineries in determining compliance. This
could provide flexibility for some refining companies to delay making
desulfurization investments at some smaller refineries for several
years. Refiners also could generate credits based on the volume of low
sulfur fuel produced above the required percentage. For example, if a
refinery were required to produce a minimum of 80 percent of its
highway diesel pool as low sulfur in the first year, and that refinery
actually produced 100 percent of its highway diesel as low sulfur that
year, it could generate credits based on the volume of the ``extra'' 20
percent of low sulfur fuel it produced. Those credits could be sold or
traded with another refinery, which could in turn use the credits to
produce a greater percentage of 500 ppm sulfur highway diesel fuel.
More details on how these ABT provisions could be structured are
discussed in section VI.A.2.a.iv below.
We believe a credit trading program may be particularly beneficial
for refiners whose volumes of highway diesel are relatively small. It
is possible that the credits generated by a refiner producing a large
volume of low sulfur diesel could potentially be sufficient to offset a
smaller refiner's entire highway diesel production, thereby enabling a
smaller refiner to comply solely by the use of credits--and avoid
desulfurization investments--for several years.
While we believe that a credit trading program could add meaningful
flexibility under this approach, we are concerned about the potential
for shortfalls in supply of low sulfur highway diesel in those areas
supplied exclusively or primarily by refiners complying by the use of
credits (i.e., producing only 500 ppm fuel). This situation could
potentially occur, for example, in the Rocky Mountain area, or other
areas served primarily by smaller refineries, or areas with relatively
isolated fuel distribution systems. This concern becomes more salient
as the percentage of 500 ppm fuel allowed to be produced increases. If
the flexibility were to begin with 20 percent (of 500 ppm fuel) in the
first year, the likelihood of a supply shortfall would be less likely
than if the program begins with 50 percent (of 500 ppm fuel).
Therefore, we seek comment on the extent to which this situation could
occur and ways to structure the credit trading system to prevent low
sulfur fuel supply shortfalls in any area, perhaps through regional
restrictions in credit trading, or providing incentives for refiners to
supply sufficient volumes of low sulfur fuel. We have been, and will
continue, working with the Western states (for example, through the
Western Governors Association) to discuss the best ways of implementing
the program in that area.
Alternatively, we request comment on a regional approach to
designing a compliance flexibility (for example, different refiner
production levels and/or availability provisions for different areas of
the country). We seek comment on whether and how this compliance
flexibility could be enhanced by such a regional approach, including
information and data that would help us to better understand regional
differences in highway diesel fuel supply, demand and distribution.
Refiners have expressed concern that under some phase-in approaches
it might be difficult for them to recover their capital investments. We
request comment this issue, including how the potential for cost
recovery under a phase-in approach compares with that under the single-
fuel approach, and what the implications are for the optimal production
level of low sulfur diesel under the compliance flexibility approach.
We also invite comment on an alternative in which we simply
establish a minimum production percentage for low sulfur fuel in the
beginning of the program, and allow the market to take over in
determining the appropriate supply and distribution from that point on.
One concern with this approach is that it would perpetuate the
potential for misfueling for as long as two grades of highway fuel
remained in the market. We request comment on how long two grades of
highway diesel would likely coexist in the market under this approach.
Further, the level of this minimum low sulfur production percentage
would have to be carefully designed to assure sufficient availability
throughout the country. If you believe this or other alternative
approaches would make the program more useful, please share your
specific suggestions with us.

ii. What Are the Key Considerations in Designing the Compliance
Flexibility?

A key consideration in designing this compliance flexibility is
whether or not it should be accompanied by a retailer availability
requirement. Under an availability requirement, diesel retailers would
have to offer low sulfur fuel, but would have the flexibility to offer
the 500 ppm fuel as well. We believe the need for an availability
requirement is linked to the refiners' 500 ppm fuel production limits.
At a 500 ppm fuel production limit beginning at 20 percent, our
concerns for lack of availability and misfueling would likely be low
enough not to warrant a retailer availability requirement or additional
misfueling controls such as special nozzles. Our presumption is that if
at least 80 percent of the highway fuel volume is low sulfur (i.e., a
maximum 20 percent is 500 ppm), the low sulfur fuel should be
sufficiently available across the country. Alternatively, if refiners
were allowed to produce some greater proportion of their highway diesel
fuel as 500 ppm fuel in the first few years, there would be a greater
likelihood of low sulfur fuel supply shortfalls, lack of availability,
and misfueling , and there would be a more compelling need to ensure
that some minimum fraction of diesel retailers offered the low sulfur
fuel. We request comment on the level of the 500 ppm fuel production
limit at which concerns about low sulfur shortfalls, lack of
availability, and misfueling would be great enough to warrant imposing
a retailer availability requirement. We ask that commenters also
consider whether they would prefer a ``blended'' program (i.e., a
program with both a production limit on 500 ppm fuel and some form of a
retailer availability requirement) to a program that permits a slightly
lower level of 500 ppm fuel, but with no availability requirement.
In considering this issue, note that the percentage of low sulfur
diesel fuel produced would not necessarily match the availability
level. For example, if 80 percent of the highway fuel pool were low
sulfur, this would not necessarily translate into the low sulfur fuel
being available at 80 percent of retail stations currently selling
diesel fuel. Since large retail stations (e.g., large truck stops) and
centrally-fueled fleets represent a disproportionate share of the
diesel sales volume, it is possible that the percentage of retail
stations offering low sulfur fuel could be much lower than 80 percent
of the diesel retail stations. If this were the case, would there still
be concerns with lack of availability of the low sulfur fuel (e.g.,
even with 20 percent of highway fuel as low sulfur)?
We believe there are merits to designing this compliance
flexibility in a way that avoids the need for a retailer availability
requirement. With no availability requirement, retailers would be free
to choose to sell 500 ppm fuel only, low sulfur fuel only, or both. We
have heard from refiners and diesel marketers that they believe that
retailers, if faced with an availability requirement, would likely
decide not to carry both grades of fuel but, rather, would switch over
to the low sulfur fuel to avoid the expense of installing new tanks and
pumps. If this were true, an

[[Page 35509]]

availability requirement could have the effect of significantly
limiting a refiner's markets for its 500 ppm fuel, thus, limiting the
benefits of the compliance flexibility approach. Nevertheless, we seek
comment on whether an availability requirement for low sulfur diesel
fuel should be a condition for retailers marketing 500 ppm fuel.
We seek comment on whether a retailer availability requirement
would diminish the utility of the compliance flexibility approach, and
at what point in designing this option (e.g., at what 500 ppm fuel
production limit) a retailer availability requirement would become
necessary to encourage sufficient availability of low sulfur fuel.
Since this compliance flexibility is voluntary, we anticipate that
refiners would only produce and market 500 ppm fuel under the allowed
percentages to the extent that the costs of distributing it are offset
by savings elsewhere. The distribution system has only a limited
ability to accommodate a second grade of highway diesel without
incurring significant costs (e.g., installing new tankage). Therefore,
while refiners may be able to reduce the costs of diesel fuel
production if higher percentages of high sulfur diesel fuel are
permitted, they may find it difficult to market 500 ppm fuel in volumes
much above even the 20 percent level, due to distribution system costs.
We request comment on the degree to which the distribution and retail
costs associated with accommodating two grades of highway diesel fuel
depend on the relative volumes of those fuels. For example, how would
the costs incurred in the distribution system vary as the amount of 500
ppm fuel produced by refiners increases from zero to 50 percent, or
even beyond?

iii. How Does This Compliance Flexibility Relate to the Options for
Small Refiner Flexibility?

In section VIII.E., we seek comment on three approaches for small
refiner flexibility. One of these approaches would allow small refiners
to continue selling 500 ppm fuel for an unspecified period of time
(although we seek comment on an appropriate duration for this
flexibility). If the compliance flexibility approach described here
were implemented for the refining industry as a whole, we seek comment
on the best ways to meld this flexibility with approaches for
minimizing the burden on small refiners. For example, we seek comment
on whether it would be appropriate to either relax or remove any 500
ppm production limits for small refiners. In other words, we may
consider allowing small refiners to continue selling their full
production volume of highway diesel as 500 ppm fuel for some period of
time (likely at least as long as the compliance flexibility provided to
the refining industry as a whole, if not for some or an unlimited
number of years beyond that). We request comment on the appropriate
duration of this flexibility for small refiners. Further, we seek
comment on whether small refiners should be allowed to generate and
sell credits under the compliance flexibility's ABT program, even if
small refiners are not required to produce any portion of their highway
fuel as low sulfur diesel. The ABT approach could minimize the burden
on small refiners by allowing them to make some additional profit to
offset their desulfurization investments, thus giving them an incentive
to produce low sulfur highway diesel fuel earlier than they otherwise
would. We seek comment on other ways this compliance flexibility could
be crafted to minimize burden on small refiners and to better meld with
the approaches for small refiner flexibility described in section
VIII.E.
It should be noted that our approach to allow small refiners to
continue selling 500 ppm highway diesel (on which we're seeking public
comment in section VIII.E.1.) does not include a retailer availability
requirement. During the SBREFA process, small refiners expressed
concern that an availability requirement would significantly limit
their potential markets for 500 ppm fuel, since they believe that few
retail outlets would be willing to offer both grades of highway diesel
due to the significant costs of installing new tanks and pumps.
Therefore, if this option for small refiner flexibility is promulgated
in the final rule, we would reconsider its design in light of any
decisions made for compliance flexibilities for the whole refining
industry (e.g., the issue of whether an availability requirement would
be necessary).

iv. How Would the Averaging, Banking and Trading Program Work?

This section discusses in more detail how we envision an averaging,
banking and trading (ABT) program working in conjunction with the
compliance flexibility approach. The goal of the ABT provisions is to
maximize the flexibility provided by the program without diminishing
its environmental benefits. We envision that this ABT program could
apply to the program regardless of the actual level of the minimum
refiner production requirement for low sulfur highway diesel. We
request comment on all aspects of these ABT provisions. If you have
ideas on how these provisions could be structured differently to
enhance the program, please share your specific suggestions with us.

Averaging

Refiners and importers could be allowed to meet the required
minimum percentage of low sulfur fuel production averaged over their
entire corporate highway diesel pool. The minimum required percentage
of low sulfur fuel production under the compliance flexibility would be
determined on an annual average basis, across all refineries owned by
that refiner (or all highway diesel fuel imported by the importer in
the calendar year). Thus, within a given refining company, the volume
of low sulfur fuel produced at one refinery could be below the minimum
required percentage, so long as the volume produced at another refinery
exceeded the minimum percentage by a sufficient amount such that the
minimum required percent of low sulfur volume was met at the corporate
level.

Generating Credits

Beginning in 2006, refineries and importers could generate credits
based on the volume of low sulfur fuel produced above the required
percentage. For example, a refinery produced 10 million gallons of
highway diesel fuel in 2006 and was required to produce a minimum of 80
percent of its highway diesel volume (8 million gallons) as low sulfur
that year. That refinery actually produced 100 percent of its highway
diesel as low sulfur that year. Thus, it could generate credits based
on the volume of the ``extra'' 20 percent of low sulfur fuel it
produced above the required minimal percentage `` that is, 2 million
gallons of credits. Under this program, we do not envision a need to
establish a baseline volume of diesel fuel, since credits would be
generated based on the volume of low sulfur diesel fuel actually
produced above the required percentage.
Credits could be generated in each year that the compliance
flexibility provisions are in place. In other words, if the duration of
the compliance flexibility were for four years (i.e., refiners were
allowed to continue producing some specified percentage of 500 ppm fuel
for four years after the start of the low sulfur program), from 2006
through 2009, credits could be generated in each of those years.
We seek comment on whether there could be circumstances where the
use of low sulfur highway diesel could be shown to demonstrate
environmental benefits significant enough to warrant

[[Page 35510]]

the generation of early credits. To the extent there may be
circumstances that warrant early credit generation, we seek comment on
whether there should be an appropriate discount factor applied to such
credits, to ensure they would be comparable with the environmental
benefits achieved by the use of low sulfur fuel in vehicles meeting
today's proposed standards. See section IV.F.
As an additional aspect to implementing the compliance flexibility
program, we seek comment on whether it would be advantageous for EPA to
offer to sell additional ABT credits to refineries at a predetermined
price. This would provide more certainty about the cost of supplying
low sulfur diesel fuel by establishing a ceiling price on the ABT
credits. We request comment on (1) what should be the appropriate
predetermined price for these ABT credits; (2) whether there should be
a cap on the total number of credits available from EPA to assure
availability of low sulfur diesel; and (3) if there is a cap, whether
credits should be sold on a first-come, first-serve basis.

Using Credits

Refiners and importers would be able to use credits to demonstrate
compliance with the minimum required percentage of low sulfur highway
diesel fuel, if they are unable to meet this requirement with actual
highway diesel fuel production. Although credits would not officially
exist until the end of the calendar year (based on the generating
refinery's actual low sulfur fuel production) there is nothing to
prevent companies from contracting with each other for credit sales
prior to the end of the year, based on anticipated production. The
actual credit transfer would not take place until the end of the year.
All credit transfer transactions would have to be concluded by the last
day of February after the close of the annual compliance period (e.g.,
February 28, 2007 for the 2006 compliance period).
For example, refiners who wish to purchase credits to comply with
the 2006 required percentage of low sulfur fuel could do so based on
the generating refinery's projections of low sulfur fuel production. By
the end of February the following year, both the purchaser and the
seller would need to reconcile the validity of the credits, as well as
their compliance with the required percentages of low sulfur fuel
produced.
We seek comment on allowing an individual refinery that does not
meet the required percentage of low sulfur fuel production in a given
year to carry forward a credit deficit for one year. Under this
provision, the refinery would have to make up the credit deficit and
come into compliance with the required low sulfur production percentage
in the next calendar year, or face penalties. This provision would give
some relief to refiners faced with an unexpected shutdown or that
otherwise were unable to obtain sufficient credits to meet the required
percentage of low sulfur fuel production.
We recognize that there is potential for credits to be generated by
one party and subsequently purchased and used in good faith by another
party, yet later found to have been calculated or created improperly,
or otherwise determined to be invalid. Our preference would be to hold
the credit seller, as opposed to the credit purchaser, liable for the
violation. Generally, we would anticipate enforcing a compliance
shortfall (caused by the good faith purchase of invalid credits)
against a good faith purchaser only in cases where the seller is unable
to recover valid credits to cover the compliance shortfall. Moreover,
in settlement of such cases, we would strongly encourage the seller to
purchase credits to cover the good faith purchaser's credit shortfall.
We believe that any person could act as a broker in facilitating
credit transactions, whether or not such person is a refiner or
importer, so long as the title to the credits are transferred directly
from the generator to the purchaser. Whether credits are transferred
directly from the generator to the purchaser, or through a broker, the
purchaser needs to have sufficient information to fully assess the
likelihood that credits would be valid. Any party that can generate and
hold credits could also resell them, but the credits should not be
resold more than twice. Repeated sales of credits could significantly
reduce the ability to verify the validity of those credits.

How Long Would Credits Last?

The goal of these ABT provisions is to provide refiners additional
flexibility in the early years of the low sulfur fuel program. After
the first few years of the program, there would be a significantly
greater proportion of aftertreatment-equipped vehicles in the fleet. It
would be important to ensure a full transition to the new low sulfur
fuel to prevent misfueling of those vehicles and preserve the
environmental benefits of the program. Therefore, we do not currently
envision allowing credits to be used more than a few years beyond the
compliance flexibility period. We seek comment on whether credit
lifetime should be limited, and if so on the appropriate length of time
credits should be allowed to be used (in other words, the ``lifetime''
of credits).

v. Compliance, Recordkeeping, and Reporting Requirements

This section describes the types of provisions we believe the
regulations would need to include if a compliance flexibility approach
were adopted, to ensure that diesel fuel subject to the 500 ppm sulfur
standard would not be introduced into model year 2007 and later diesel
vehicles.
Refiners and importers of 500 ppm highway diesel fuel would be
required to designate all highway diesel fuel produced as meeting the
500 ppm sulfur standard or meeting the proposed 15 ppm standard. Such
refiners and importers would be required to maintain records regarding
each batch of motor vehicle diesel fuel produced or imported, including
the volume of each batch, and would be required to maintain records,
and to report regarding credits earned and credit transactions.
Reporting would also be required regarding volumes of highway diesel
fuel produced or imported.
All parties in the distribution system that chose to carry 500 ppm
fuel would be required to segregate that fuel from 15 ppm sulfur fuel,
and would be responsible for ensuring that fuel designated as 15 ppm or
500 ppm meets the respective sulfur standards, throughout the
distribution system. Such segregation requirements would likely be
modeled after those of the reformulated gasoline (RFG) program (e.g.,
the RFG program's requirements for product transfer documents,
refiners' designations of the standards to which each batch of fuel
applies, and registration requirements for refiners producing both
highway diesel fuels). However, the RFG program's segregation
provisions are somewhat different, in that they were designed to
segregate RFG from conventional gasoline by geographic area. In the
highway diesel program, the segregation provisions would be much more
widespread, because both grades of highway fuel could be distributed
throughout the country, depending on how refiners choose to take
advantage of the compliance flexibility. We seek comment on the need to
require refiners producing 500 ppm fuel to conduct some form of
downstream quality assurance sampling, similar to the surveys required
under the RFG program.
Further, all parties in the distribution system would be subject to
prohibitions against selling, transporting, storing, or introducing or
causing or allowing the introduction of diesel fuel having a

[[Page 35511]]

sulfur content greater than: (1) the proposed 15 ppm standard into
highway diesel vehicles manufactured in the 2007 model year and beyond;
and (2) 500 ppm into any highway vehicle. Under the proposed
presumptive liability scheme (as discussed in section VIII.A.8), if a
violation is found at any point in the distribution system, all parties
in the distribution system for the fuel in violation are responsible
unless they can establish a defense. Because of our concerns for
contamination and misfueling with having two grades of highway diesel
in the market, we seek comment on whether a refiner should lose its
flexibility to continue producing 500 ppm fuel if it is found liable
for a violation.
All parties handling 500 ppm fuel also would be required to
maintain product transfer documents for five years that indicate to
which highway diesel fuel standard the fuel is subject. Pump labels
would be required at retail outlets and wholesale purchaser-consumer
facilities providing notice regarding the different highway fuel types
and the vehicles they may/may not be used in. As mentioned above,
nozzle requirements might also be considered if the minimum volume
requirement for low sulfur diesel is low enough to warrant it.
The rule would prohibit any refiner from producing more 500 ppm
highway diesel fuel than allotted, and would prohibit any party from
distributing or selling diesel fuel not meeting the proposed 15 ppm
standard unless it is properly designated and accompanied by
appropriate product transfer documents. The rule would also prohibit
any person from introducing or causing or allowing the introduction of
highway diesel fuel not meeting the 15 ppm sulfur standard into any
model year 2007 or later vehicle.
As with any ABT program, we would need refiners to keep appropriate
records, and to file necessary reports, to ensure compliance as well as
the integrity of any credit generation, trading, and use. If this
program is promulgated in the final rule, we would envision that
refiners would likely be required to keep records of key information
pertaining to the ABT program. Beginning the first year that credits
are generated, any refiner for each of its refineries, and any importer
for the highway diesel fuel it imports, would keep information
regarding credits generated, separately kept according to the year of
generation. We envision that refiners would keep records of the
following information, at a minimum, and report such information to EPA
on an annual basis, for any year in which credits are generated,
transferred, or used:
The total volume of highway diesel fuel produced
The total volume of highway diesel fuel produced meeting
the 500 ppm sulfur standard
The total volume of highway diesel fuel produced meeting
the low sulfur standard
The total volume of highway diesel fuel produced
(delineating both 500 ppm fuel and low sulfur fuel) after inclusion of
any credits
The number of credits in the refiner's or importer's
possession at the beginning of the averaging period
The number of credits used
If any credits were obtained from or transferred to other
parties, for each other party, its name, its EPA refiner or importer
registration number, and the number of credits obtained from or
transferred to the other party;
The number of credits in the refiner's or importer's
possession that will carry over into the next averaging period
Contracts or other commercial documents that establish
each transfer of credits from the transferor to the transferee
The calculations used to determine compliance with the
minimum required percentage of low sulfur highway diesel fuel
The calculations used to determine the number of credits
generated

b. Refiner-Ensured Availability

An alternative concept suggested to the Agency to accomplish the
objective of ensuring widespread availability of low sulfur diesel fuel
while still allowing flexibility for producing less than all of the
diesel fuel pool as low sulfur is to have the refiners ensure that it
is widely available. The base program would still be a requirement that
refiners produce only highway diesel fuel which meets the sulfur
standard proposed today. However, refiners could voluntarily choose to
participate in a program where they would be allowed to sell a larger
fraction of their highway diesel fuel as 500 ppm fuel, in exchange for
ensuring that low sulfur diesel fuel is made widely available at the
retail level.
This concept may entail a refinery contracting with, or purchasing
credits from, retailers, who in exchange for incentives from the
refiner, agree to make low sulfur diesel fuel available. This could
mean that the retailer decides to switch over entirely to selling low
sulfur diesel fuel, or that they offer both low sulfur and high sulfur
diesel fuel simultaneously. The retailer would have to make a showing
that: (1) the low sulfur diesel was ``meaningfully'' available; (2)
there was an assured supply chain for obtaining low sulfur diesel fuel;
and (3) the diesel fuels were segregated and properly labeled at the
pumps. ``Meaningfully'' available might mean having dedicated pumps and
tankage for low sulfur diesel with a capacity in the thousands of
gallons range, and operating all year long. To be clear, the contract/
credits would be for making low sulfur diesel available for sale, not
necessarily selling a given volume of low sulfur diesel.
The relief that refiners receive in exchange for providing for low
sulfur availability could be calculated on the basis of the retailer's
total diesel sales volume. For example, the refiner would be permitted
to produce a certain volume of highway diesel fuel at the current 500
ppm cap in proportion to the total diesel sales volume of the retailers
that the refiner contracts with (or purchases credits from). A ratio
could be applied to the retailer's sales volume to ensure sufficient
retail availability.
An example of how this concept might work is as follows: A refinery
producing highway diesel fuel contracts with several truck stops and
service stations to make low sulfur fuel available at their stations.
The refiner would then be permitted to produce 500 ppm grade diesel
fuel in an amount up to the combined diesel sales volume (or some
multiple thereof) for these retailers. The retailers may receive their
low sulfur diesel fuel from this refiner or from other refiners to
comply with the contract.
Under this approach, refiners would likely make arrangements with,
or purchase credits from, the largest retailers (since they have the
largest fuel volumes), in order to minimize transaction costs. Because
the largest 5 percent of diesel retail stations represent 60 percent of
the sales volume, \152\ to achieve any meaningful availability of low
sulfur fuel at retail stations, the program may require a considerably
larger percentage of the sales volume to be targeted by weighting more
heavily credits generated by smaller retail outlets.
---------------------------------------------------------------------------

\152\ Memorandum to Docket A-99-06 from Jeffrey Herzog, EPA,
entitled: ``Diesel Throughput Volume by Percentage of Diesel Fuel
Retailers,'' May 5, 2000.
---------------------------------------------------------------------------

We ask for comment on this concept, on its advantages and
disadvantages compared to other implementation options, on the
percentage of retail outlets that may be sufficient under this concept
to achieve satisfactory low

[[Page 35512]]

sulfur diesel fuel availability, on means of ensuring adequate
geographic distribution of low sulfur diesel fuel throughout the year,
and on the appropriate means of calculating the volumes that refiners
should be permitted to produce as high sulfur in exchange for making
low sulfur available. We also request comment on how such a program
could be implemented and enforced. In particular, we request comment on
the type of recordkeeping and reporting EPA should require in ensuring
a refiner actually has legitimate credits, contracts or other binding
arrangements with retailers to make low sulfur diesel fuel
``meaningfully'' available. We further request comment on whether and
what type of recordkeeping and reporting may be necessary for retailers
and distributors, particularly if the program were structured to allow
retailers to generate and sell credits.

c. Retailer Availability Requirement

One way of ensuring widespread availability of the low sulfur fuel
under a phase-in approach would be to require retailers selling highway
diesel to make available the low-sulfur diesel (i.e., a retailer
availability requirement). Retailers would be free to sell the current
500 ppm sulfur fuel as well, but at a minimum would have to offer the
low sulfur fuel. This approach could either be a stand-alone program
design (i.e., with no refiner production requirement for a minimum
amount of low sulfur diesel), or could be coupled with a refiner
production requirement. Retailers would be responsible for getting low-
sulfur diesel from the distribution system. The premise of this
approach is that the fuel distribution system would react to the market
demands, and supply and distribute the second grade of fuel in all
parts of the country.
In order to turn this premise into a reality, the fundamental
issues associated with a phase-in approach, as discussed in subsection
1 above, would have to be addressed. Consequently, in the context of an
availability requirement, we seek comment on how to resolve the
concerns raised in subsection 1. With regard to the structure of such
an availability requirement, we seek comment on when it should begin,
whether it could be limited to just a fraction of the diesel fuel
retail outlets, and what fraction would constitute acceptable
availability in the marketplace. We specifically request comment on the
merits of limiting an availability requirement to the larger diesel
retailers. Under such an approach, the larger diesel retailers would
have to carry low sulfur diesel, but could also choose to carry the 500
ppm grade as well. Smaller retailers not subject to the availability
requirement would have the flexibility to choose to carry only the low
sulfur grade, only the 500 ppm grade, or both. For example, we seek
comment on the merits of limiting the requirement to only truck stops
selling more than 200,000 gallons of diesel fuel per month, and other
retail outlets selling more than 20,000 gallons of diesel per month, as
suggested by some Panel members during the Small Business Advocacy
Review process. We encourage commenters to consider other appropriate
throughput thresholds, for both truck stops and service stations that
could limit an availability requirement to the larger retailers, while
still ensuring sufficient availability.
While desirable to limit the fraction of retailers subject to an
availability requirement, ensuring sufficient availability is
complicated by the fact that diesel fuel is sold at a portion of all
retail outlets today. \153\ If less than 100 percent of diesel retail
outlets are required to make the new fuel available, how would we
ensure availability in all parts of the country? Commenters should
consider the distribution of diesel fuel outlets around the country,
and the distances between outlets in addressing this issue. How would
the rest of the distribution system respond to supply the low sulfur
fuel to the retail outlets needing to make it available? To help
protect against fuel shortages either nationally or regionally, would
an availability requirement need to be coupled with a production
requirement on refiners to ensure supply of a minimum amount of low-
sulfur diesel fuel? If so, how should such a production requirement be
structured? Conversely, could an availability requirement be coupled
with a production requirement in a way that would allow a larger
percentage of 500 ppm fuel production in the early years? (See the
discussion above in subsection 2.a.ii)
---------------------------------------------------------------------------

\153\ ``Summary Data on Diesel Fuel Retailers,'' Memo to the
docket from Jeffrey Herzog, EPA, March 23, 2000 (Docket item # II-B-
07).
---------------------------------------------------------------------------

With regard to the impacts on the diesel fuel retail and
distribution system, numerous parties in the industry have commented
that managing two grades of highway diesel in the distribution system
would raise their costs. We seek comment on what actions retailers,
centrally fueled fleets, wholesalers, terminals, pipelines, and
refiners would take to manage two grades of highway diesel, and in
particular on the cost impacts resulting from those actions. We
especially seek comment on what cost savings refiners might realize
under such an approach, and whether these savings would be greater than
the costs incurred by the distribution system to distribute a second
grade of highway diesel fuel. In this context, we also seek comment on
how refiners would plan their refinery changes given the uncertainty of
low sulfur diesel demand from retailers under such a phase-in approach.
When would they make their capital investments, and for what volume of
fuel would they plan to build desulfurization capacity? How would they
predict demand in the time frame when they would need to make their
capital investments? How would they adjust to different volumes from
predicted demand levels, and what would be the implications?
Commenters should address this approach from the perspective of the
issues discussed above in subsection A.1 (including misfueling,
distribution system impacts, potential costs, etc). We are also
interested in the implications of such an approach on prices in the
wholesale and retail markets, and on the ability of retailers and
distributors to recover costs under such an approach.
We also invite comment on the merits of applying an averaging,
banking and trading program within the context of a retailer
availability requirement. Such a credit trading program could entail
elements similar to the program described in subsection 2.a.v. for
refiners under the compliance flexibility approach, but would be
tailored specifically to retailers subject to an availability
requirement. Commenters should address how such a credit trading
program might be structured, if they believe it should differ
significantly from the refiner-based approach discussed above.
Finally, the trucking industry and diesel marketers have also
commented that an availability requirement would be administratively
intensive for the Agency to implement and enforce, especially in
verifying actual fuel availability. Therefore, we ask comment on ways
to streamline the enforcement of such a program to avoid unnecessary
burden on both industry and the Agency.
2. Why Is a Regulation Necessary to Implement the Fuel Program?
Some commenters on the ANPRM suggested simply leaving it up to the
market to introduce low-sulfur highway diesel fuel--that is, establish
no regulatory requirements for refiners to produce the fuel and no
requirements for retailers to sell the fuel. The

[[Page 35513]]

commenters' line of reasoning for this suggestion is as follows. The
vehicle and engine manufacturers would be forced by emission standards
to introduce vehicles meeting stringent emission standards. Since the
engines and vehicles would need low-sulfur diesel fuel to meet the
emission standards, then the vehicle purchasers would have to refuel
only with low-sulfur diesel fuel. The fuel production and distribution
system would then respond to the demand and provide the fuel if, when,
and where necessary.
Such an approach raises many of the same issues discussed above
with respect to phase-in approaches (e.g., fuel availability,
misfueling, and uncertainties in the transition to low sulfur). These
concerns, however, would be heightened by the fact that no regulatory
measures would be in place to mitigate them. We seek comment on whether
a market-based approach could adequately ensure availability of the low
sulfur fuel for the vehicles that need it.
3. Why Not Just Require Low-Sulfur Diesel Fuel for Light-Duty Vehicles
and Light-Duty Trucks?
In the ANPRM, we requested and received considerable comment on
focusing the rulemaking effort on providing low-sulfur diesel fuel for
light-duty vehicles and trucks only. By providing a clean grade of
diesel fuel, exhaust emission control technology would be enabled. This
in turn would give light-duty diesel vehicles a much better chance of
meeting the final Tier 2 emission standards. The appeal of a light-duty
only approach is that the program would be relatively small and could
set the stage for future expansion of low-sulfur diesel fuel into the
heavy-duty market if the demand developed.
Based on the comments received on the ANPRM and our own analysis,
however, there appears to be little justification for such a regulatory
approach. First, and most importantly, such an approach would provide
no environmental benefit to justify the costs of the program. Under the
Tier 2 program, all LDVs and LDTs must meet on average a certain
NO_X emission standard. There are a number of emission
standards or ``bins'' that individual vehicles can be certified to, but
an overall fleet average emission standard must still be met.
Consequently, regardless of whether or not the Tier 2 fleet is
comprised of a large number of diesel vehicles, the same overall fleet
average NO_X emission rate will be achieved. The only
anticipated difference would be in particulate emissions where, even
though the emission standards are the same, in-use emissions are
assumed to be somewhat lower for gasoline vehicles than for diesel
vehicles. In contrast, today's proposed program for setting new
emission standards for heavy-duty engines and vehicles in conjunction
with lower sulfur highway diesel fuel would achieve significant
reductions in NO_X and particulate matter, as discussed
further in section II.
Secondly, the comments received on the ANPRM from the fuel
production and distribution system indicated that such an approach
would be very costly. The Engine Manufacturers Association conducted a
study of the cost increase associated with distributing a unique grade
of diesel fuel for just light-duty vehicles and trucks.\154\ The
results of this study indicated that the distribution costs alone
(i.e., not including refiner production costs) for such a fuel could be
3 to 4 cents per gallon. Moreover, this study made some simplifying
assumptions that served to underestimate actual volume of highway
diesel fuel that would have to be produced and the costs. The study
assumed a production volume of 5 percent low sulfur diesel, which is
not realistic because many retailers might choose to switch over
entirely to the low sulfur fuel. Thus, refiners would have to make the
investments to produce a considerably larger volume of low sulfur
diesel fuel than might be required for new light-duty vehicles and
trucks only.
---------------------------------------------------------------------------

\154\ ``Very-Low-Sulfur Diesel Distribution Cost,'' Baker &
O'Brien Inc., for the Engine Manufacturers Association, August 1999.
---------------------------------------------------------------------------

Third, commenters indicated that such an approach may be
impractical. In areas where there are few fuel distribution options
(e.g., areas not served by pipelines, areas with few diesel retail
outlets), the low-sulfur diesel fuel may not be made available or, if
it is, it could only be sold at retail prices considerably higher than
the refiners' cost to produce the fuel. Consumer demand for light-duty
diesel vehicles could be reduced by both unavailability of the low
sulfur fuel and uncertainty about it being available at reasonable
prices.
Finally, a light-duty only approach would appear to be
inappropriate in light of our demonstrated air quality need for
additional emission reductions and the opportunity available with
recent advancements in diesel engine exhaust emission control
technology to obtain these emission reductions from heavy-duty engines.
If the technology necessary to meet very low emission standards for
light-duty diesel vehicles is feasible with the control of diesel fuel
sulfur, and if that same technology is applicable to heavy-duty diesel
vehicles, then we have an obligation under the Clean Air Act to
consider emission standards for heavy-duty vehicles that would be
enabled by that technology as well. Given the air quality need, we
would be remiss in our obligations under section 202(a)(3)(A) of the
Act which requires us to set the most stringent standards feasible for
heavy-duty vehicles, taking into consideration cost and other factors.
EPA can revise such standards, however, based on available information
regarding the effects of air pollutants from heavy-duty engines on
public health or welfare.
4. Why Not Phase-Down the Concentration of Sulfur in Diesel Fuel Over
Time as Was Done With Gasoline in the Tier 2 Program?
There are a number of ways a fuel change can be introduced over
time. The most recent example is in the Tier 2 rulemaking where the
concentration of sulfur in gasoline was phased-down over time. Such an
approach is not workable for diesel fuel, however, due to the demands
of the exhaust emission control technology. As discussed in section
III, the efficiency of both the NO_X and PM exhaust emission
control drops off quickly if the vehicle is operated on sulfur levels
higher than the standard proposed. Thus, the vehicles would be unable
to meet the emission standards, and there would be very little if any
emission benefit to be gained until the end of any such phase-down.
Furthermore, as discussed in section III, in some applications it is
possible that operation on higher sulfur levels may not only cause
permanent damage to the PM trap, but also could result in vehicle
driveability and safety concerns. Consequently, it is imperative that
aftertreatment-equipped vehicles are fueled exclusively with fuel
meeting the proposed low sulfur levels, and that the low sulfur fuel
remain segregated in the distribution system.
This contrasts with the gasoline sulfur control program, where the
impact of sulfur on the exhaust emission control technology was thought
to be less severe and emission benefits accrued even at the phased-down
sulfur levels. Furthermore, if gasoline vehicles are operated on higher
sulfur fuel, no driveability concerns are anticipated; higher sulfur
diesel would have detrimental effects on the driveability of diesel
engines. Thus, in the gasoline sulfur program there was not a need to
require that low sulfur gasoline remain segregated from the remaining
gasoline pool while sulfur levels are being phased-down. Here there is
a need to

[[Page 35514]]

segregate low sulfur highway diesel fuel to ensure the new technology
vehicles are not damaged by higher sulfur levels.

B. What Other Fuel Standards Have We Considered in Developing This
Proposal?

1. What About Setting the 15 ppm Sulfur Level as an Average?
We have considered several potential diesel fuel sulfur
alternatives in developing today's proposed rulemaking, including two
alternatives centered around a 15 ppm sulfur level: a cap at this level
as proposed, and an average at this level with a 25 ppm cap to ensure
that sulfur levels would not exceed a 15 ppm average level by too much.
The analyses of technology enablement, costs, emission reductions, and
cost effectiveness discussed in the preceding sections are based on a
15 ppm cap. In this section we provide the results of these analyses
for the 15 ppm average sulfur level case.

a. Emission Control Technology Enablement Under a 15 ppm Average
Standard

Having a 15 ppm average standard with a 25 ppm cap would increase
uncertainty around the advanced technologies required here and would
therefore be less attractive to diesel engine and vehicle
manufacturers. As discussed at length in Section III, fuel sulfur
adversely impacts the effectiveness of all known and projected exhaust
emission control devices. Despite these adverse effects, it may be
possible that the design, precious metal loading, and application of
exhaust emission control devices could be fundamentally similar under
both a 15 ppm cap and a 15 ppm average. However, we would expect that
the exhaust emission control devices would not operate at the same
level of efficiency as expected under the 15 ppm cap program and there
would be some sacrifice in the durability and reliability of these
devices due to the higher sulfur level.
PM trap regeneration would be compromised due to sulfur's adverse
impacts on the NO to NO2 conversion necessary for completely
passive PM trap regeneration.\155\ Because of this effect, concerns
have been raised that a 15 average/25 cap program would require that
some vehicle applications, particularly lighter applications having
lower operating temperatures, incorporate some form of active PM trap
regeneration strategy. Such an active regeneration strategy could take
the form of a fueling strategy capable of increasing exhaust
temperature as opposed to an electrical heater or some other ``added''
hardware. The active regeneration scheme would likely be incorporated
into the design as a backup, or protective measure, and would not
function at all times. Instead, the active regeneration would kick in
under conditions such as very cold ambient temperature conditions or
extended idles where exhaust temperatures might be too low for too long
to enable passive regeneration. There are also concerns that fuel
economy would be reduced both due to the use of active regeneration and
due to the higher, on average, PM trap backpressure. This would likely
occur due to the slightly higher soot loading, on average, resulting
from less efficient passive trap regeneration. This higher backpressure
would probably occur on all applications, not just the lighter
applications. Nonetheless, we believe that the fuel economy effect
would probably not be greater than one percent.
---------------------------------------------------------------------------

\155\ Cooper and Thoss, Johnson Matthey, SAE 890404.
---------------------------------------------------------------------------

Under a 15 ppm average standard, we would expect the in-use average
sulfur level to be roughly double the in-use average under a 15 ppm cap
program. The higher in-use sulfur level would roughly double in-use PM
emissions. Since an average limit would be in place and be enforced,
and since in-use emissions would be expected to approximate the
average, we might consider allowing engine manufacturers to certify
their engines on diesel fuel meeting the average sulfur level rather
than the cap. If this approach were taken, setting the sulfur standard
at a 15 ppm average instead of a 15 ppm cap would not necessitate an
increase in the PM standard. However, in-use PM emissions would nearly
double due to the increased average fuel sulfur level (when compared to
the 15 ppm cap base case).
Regarding the NO_X adsorber, we believe that a 15
average/25 cap program may have the potential to enable NO_X
adsorber technology, though with increased uncertainty. However, while
the NO_X adsorber would continue to adsorb and subsequently
reduce NO_X despite the higher sulfur fuel, the frequency of
sulfur regeneration events, referred to as desulfation in section III,
would roughly double relative to the rate with a 15 ppm cap. The
increased frequency of desulfation would increase fuel consumption
probably on the order of one percent and would be realized on all
diesel applications equipped with NO_X adsorber
technology.\156\ Additionally, the increased frequency of desulfation
may adversely impact NO_X adsorber durability because the
thermal strain placed on the adsorber during any desulfation event
would increase in frequency. Also, because of the increased frequency
of desulfation events, there would be a corresponding decrease in the
likelihood of being able to perform the desulfation during ideal
operating conditions. This may cause more thermal strain on the
NO_X adsorber and/or less efficient desulfation with a
corresponding increase in fuel usage. The result would be a decrease in
our level of confidence that the NO_X adsorber would be
capable of fulfilling the demands of heavy-duty diesel engines in terms
of fuel consumption and durability.
---------------------------------------------------------------------------

\156\ See section III and Table III.F-2 for more detail on
desulfation and the associated fuel economy impacts.
---------------------------------------------------------------------------

Note that, although the analysis finds that a 15 ppm average/25 ppm
cap standard has potential to be adequate for enabling high-efficiency
exhaust emissions controls, this finding involves a significantly
higher level of uncertainty than the proposed 15 ppm sulfur cap,
because it is based on the assumption that exhaust emission control
designs could be focused on the average fuel sulfur levels.
Manufacturers have commented that the possibility of some in-use fuel
at near-cap levels would necessitate designing to accommodate this
level, and they contend that this would not allow the high-efficiency
technology to be enabled. If so, the technology enablement for this
case would likely be similar to that for the 50 ppm cap case.

b. Vehicle and Operating Costs for Diesel Vehicles To Meet the Proposed
Emissions Standards With a 15 ppm Average Standard

As pointed out above, we believe it may be possible that the
design, precious metal loading, and application of exhaust emission
control devices could be fundamentally similar under both a 15 ppm cap
and a 15 ppm average. Therefore, we believe that having a 15 ppm
average sulfur standard would have no quantifiable impact on the cost
of emission control hardware relative to the costs associated with a 15
ppm cap standard. However, as mentioned, we would expect a one percent
fuel economy decrease (i.e., a one percent increase in fuel
consumption) due to the increased frequency of desulfation of the
NO_X adsorber. This reduction in fuel economy would result in
consumption

[[Page 35515]]

of more fuel and, therefore, higher costs. We have estimated the
discounted lifetime cost of this one percent fuel economy impact at
$108, $207, $755, and $893 for a light, medium, and heavy heavy-duty
diesel, and urban buses, respectively. See the draft RIA for details on
how this cost was calculated.

c. Diesel Fuel Costs Under a 15 ppm Average Standard

Having a 15 ppm average with a 25 ppm cap sulfur standard would be
directionally more attractive to the petroleum industry because of the
slightly higher sulfur levels. Overall, we would expect this approach
to provide more flexibility to refiners and distributors, and
directionally help in addressing concerns that have been expressed
about the difficulties of distributing diesel fuel with very low sulfur
specifications. The cost of meeting a 15 ppm sulfur average at the
refinery (with a 25 ppm cap) would be significantly less than meeting
the proposed cap of 15 ppm. We project that roughly half of all
refiners would be able to meet a 15 ppm average by modifying their
existing one-stage hydrotreating unit by adding a hydrogen sulfide
scrubbing unit, a PSA unit to increase hydrogen purity and a second
reactor. A new, high activity catalyst would also replace today's
catalyst. Refiners who would be capable of meeting a 15 ppm average
with a one-stage unit would likely be those blending low amounts of
light cycle oil (LCO) into their diesel fuel or those having
substantial excess hydrotreating capacity in their current unit. The
remaining refiners would require essentially the same two-stage
hydrotreating unit that would be required to meet the proposed 15 ppm
cap. In all cases, hydrogen consumption would be somewhat less than
that required to meet the proposed 15 ppm cap standard.
As for fuel distribution, under the proposed 15 ppm cap on diesel
sulfur content, we estimate that sulfur contamination in the
distribution system can be adequately controlled at modest additional
cost through the consistent and careful observation of current industry
practices. A 0.2 cent per gallon increase in distribution cost is
anticipated due to the need for an increase in pipeline shipment
interface volumes, increased quality testing at product terminals, and
the need to distribute an increased volume of fuel to meet the same
level of consumer demand due to a reduction in energy density. Having a
15 ppm average standard would mean that the increase in pipeline
interface volumes would likely be somewhat smaller than under the
proposed 15 ppm cap. However, we do not expect that the savings in
interface volumes would be proportional to the difference between the
standards. This is due to the similarity of the alternative standards
with the proposed 15 ppm sulfur cap relative to their comparison with
the sulfur level of other products in the distribution system such as
nonroad diesel fuel (3,400 ppm average sulfur content). Consequently,
we estimate that distribution costs under a 15 ppm average standard
would only be marginally lower (approximately 0.003 cents per gallon
less) than under the proposed 15 ppm cap.
Overall, we project that the average cost of meeting the 15 ppm
average at the refinery would be about 3.0 cents per gallon, about 1.0
cents per gallon less than the corresponding cost for fuel meeting a 15
ppm sulfur cap. Adding the cost of lubricity additives and increase in
distribution costs, the final cost for the 15 ppm average/25 ppm cap
fuel would be 3.4 cents/gallon, as compared to 4.4 cents per gallon
under the proposed 15 ppm cap standard.

d. Emission Reductions Under a 15 ppm Average Standard

As discussed above, we believe that the same basic exhaust emission
control technology could be used to reduce exhaust emissions from HDDEs
even if we required a 15 ppm average rather than a 15 ppm cap. However,
as pointed out above, there would likely be penalties in durability,
fuel consumption, and emissions.
At this higher fuel sulfur level, we believe that the particulate
trap will still result in large reductions of HC, CO, and carbon soot.
We also believe that the 0.2 g/bhp-hr NO_X standard may be
achieved using a NO_X adsorber. Nonetheless, the total PM
reductions would be lower under a 15 ppm average standard. Sulfur in
the fuel impacts the amount of direct sulfate PM in the exhaust gas. We
estimate that a 15 ppm average standard would result in almost double
the total PM emissions as compared to a 15 ppm cap standard because the
15 ppm cap is assumed to result in a 7 ppm in-use average. Table VI.B-1
presents projected nationwide HDDE PM emissions for the baseline and
control case for a 15 ppm average/25 ppm sulfur cap standard along with
the corresponding reductions. For comparison, the same information is
shown for the proposed 15 ppm cap. Refer to the draft RIA for details
of this analysis.

Table VI.B-1.--HDDE PM Emissions With a 15 ppm Average/25 ppm Sulfur Cap
[Thousand short tons]
----------------------------------------------------------------------------------------------------------------
15 ppm average 15 ppm cap (for
------------------ comparison)
Calendar year Baseline -----------------
Controlled Controlled
----------------------------------------------------------------------------------------------------------------
2007...................................................... 100 89 88
2010...................................................... 94 60 59
2015...................................................... 93 33 30
2020...................................................... 98 19 15
2030...................................................... 119 13 8
----------------------------------------------------------------------------------------------------------------

A higher average sulfur level also results in lower SO_X
emission reductions. We assume that the sulfur in the fuel that is not
converted to sulfate PM is converted to SO2. Because we base
SO_X emissions on the amount of sulfur flowing through the
engine, the increase in fuel consumption also negatively impacts
SO_X emissions. Table VI.B-2 presents projected nationwide
HDDE SO_X reductions for a 15 ppm average/25 ppm sulfur cap
standard and for the proposed 15 ppm cap.

[[Page 35516]]

Table VI.B-2.--HDDE SO_X Emission Reductions With a 15 ppm Average/25 ppm
Sulfur Cap
[Thousand short tons]
------------------------------------------------------------------------
15 ppm
Calendar year average 15 ppm cap
------------------------------------------------------------------------
2007.......................................... 86 88
2010.......................................... 91 93
2015.......................................... 99 102
2020.......................................... 107 109
2030.......................................... 120 123
------------------------------------------------------------------------

e. Cost Effectiveness of a 15 ppm Average Standard

The methodology used to determine the cost-effectiveness of a 15
ppm average sulfur standard follows that described in Section V for our
proposed 15 ppm cap standard. The alternative standard of 15 ppm on
average does have impacts on specific values in the calculations,
including lower desulfurization and distribution, lower in-use PM
benefits, and lower SO2 benefits all of which were pointed
out above. Engine costs are assumed not to change under either a 15 ppm
cap or 15 ppm average standard. We have calculated cost-effectiveness
using both the per-vehicle and aggregate approaches, consistent with
our cost-effectiveness presentation in Section V for our proposed
program. The results are shown in Tables VI.B-3 and VI.B-4 which can be
directly compared to Tables V.F-1 and V.F-2, respectively, showing
values for the proposed 15 ppm cap standard. Details of the
calculations are presented in the draft RIA which can be found in the
docket for this rulemaking.

Table VI.B-3.--Per-Vehicle Cost-Effectiveness of a 15 ppm Average/25 ppm Cap Sulfur Standard
----------------------------------------------------------------------------------------------------------------
Discounted
Discounted Discounted Discounted lifetime cost
Pollutants lifetime vehicle lifetime emission lifetime cost effectiveness per
& fuel costs reductions (tons) effectiveness per ton with SO2
ton credit ^a
----------------------------------------------------------------------------------------------------------------
Near-term costs:^b
NO_X + NMHC...................... $1,565 0.88 $1,800 $1,800
PM.............................. 774 0.064 12,100 5,200
Long-term costs:
NO_X + NMHC...................... $1,151 0.88 $1,300 $1,300
PM.............................. 554 0.064 8,700 1,800
----------------------------------------------------------------------------------------------------------------
^a $440 credited to SO2 (at $4800/ton) for PM cost effectiveness.
^b As described above, per-engine cost effectiveness does not include any costs or benefits from the existing,
pre-control, fleet of vehicles that would use the low sulfur diesel fuel proposed in this document.

Table VI.B-4.-- 30-Year Net Present Value Cost-Effectiveness of a 15 ppm Average/25 ppm Cap Sulfur Standard
----------------------------------------------------------------------------------------------------------------
30-year NPV
30-year NPV 30-year NPV 30-year NPV cost
costs reduction cost effectiveness
(billion) (million tons) effectiveness per ton with
per ton SO2 credit ^a
----------------------------------------------------------------------------------------------------------------
NO_X + NMHC...................................... $26.4 18.9 $1,400 $1,400
PM.............................................. $8.0 0.75 $10,700 $1,100
----------------------------------------------------------------------------------------------------------------
^a $7.2 billion credited to SO2 (at $4800/ton).

2. What About a 5 ppm Sulfur Level?
Some diesel engine and automobile manufacturers have expressed
support for a sulfur cap of 5 ppm (sometimes termed ``near-zero'') for
some or all of the highway diesel fuel pool.\157\ They view the
technology solutions envisioned in this rulemaking to be infeasible at
higher fuel sulfur levels. Although the feasibility analysis results of
this proposal lead us to disagree with this conclusion, we have
evaluated the impact that a 5 ppm sulfur cap would have on technology
enablement, vehicle and fuel costs, and emissions reductions. The
results of this analysis are provided below. Analysis details are
provided in the Draft RIA. We encourage comment on our assessment,
preferably accompanied by data and analysis supporting the commenter's
views.
---------------------------------------------------------------------------

\157\ See for example letter from Patrick Charbonneau of
Navistar to Robert Perciasepe of EPA dated July 21, 1999, EPA,
docket A-99-06.
---------------------------------------------------------------------------

Capping diesel fuel sulfur at 5 ppm would clearly strengthen the
viability of new emissions control technologies enabled at 15 ppm,
although we are aware of no additional technologies that this lower
sulfur level would enable. PM traps would emit somewhat less sulfate
PM, but non-sulfate PM emissions and certification test measurement
tolerances would effectively limit the extent to which the standard
could be lowered from the proposed 0.01 g/bhp-hr level at this time.
Given the level of precision implicit in the 0.01 numerical standard,
we would not expect a 5 ppm sulfur cap to result in a lower PM
standard. Nevertheless, there would be an in-use benefit compared to a
15 ppm cap, because the average fuel sulfur would be lower (perhaps 2-3
ppm compared to about 7 ppm) and so new vehicles

[[Page 35517]]

would emit less sulfate PM, providing a projected 86,000 ton per year
PM benefit in these vehicles in 2020, compared to 83,000 tons per year
achieved under a 15 ppm cap. We have assumed that the small margins
involved and the extremely high trapping efficiencies of filters that
are already readily available would give manufacturers no incentive to
take advantage of the lower sulfate emissions to design for higher non-
sulfate emissions under the standard.
Lower sulfate PM emissions in the existing fleet would provide a
105 tons per year additional PM benefit (in 2007 when this benefit
peaks) from adoption of a 5 ppm sulfur cap compared to a 15 ppm cap.
However this is quite small compared to the corresponding 7100 ton per
year existing fleet PM benefit of reducing fuel sulfur from typical
current average levels of around 340 ppm to levels near 15 ppm, which
in turn is a small fraction of the total direct PM emissions benefit of
the 15 ppm cap, most of which comes from enabling PM traps on new
engines (see Figure II.D-2). SO_X and SO_X-derived
secondary PM would also be reduced in about the same small proportion.
The robustness of the PM trap regeneration process would also be
directionally aided by the near zero sulfur fuel, because less of the
catalyst sites that promote regeneration would be blocked by sulfur
poisoning. (This phenomenon is described in section III.F.1.a). In
fact, designers could further increase regeneration robustness by
increasing precious metal loading without fear of inordinate sulfate
production because of the lower fuel sulfur level (though at added
cost). However, we have not quantified this directional benefit or cost
difference because we deem the 15 ppm level adequate for robust
regeneration already.
Five ppm sulfur fuel would also benefit NO_X adsorber
technology. Adsorber desulfation would be needed about four times less
often than that required under a 15 ppm sulfur cap, providing a
projected 1 percent improvement in fuel economy. There may also be a
small gain in NO_X adsorber durability due to the less
frequent thermal cycling built into the desulfation process. However,
available evidence suggests that at any fuel sulfur level under 15 ppm,
these cycles are not likely to be so numerous or severe over the
vehicle life as to seriously constrain durability. NO_X
emissions would not be much affected because the basic NO_X
storage and removal processes would occur in much the same way, and
desulfation events would be programmed to occur frequently enough to
maintain NO_X reduction efficiencies high enough to meet the
standard with a minimum of fuel consumption.
We have not performed an extensive analysis of the refining cost of
meeting a 5 ppm sulfur cap. However, Mathpro, under contract to EMA,
did estimate the refining cost of producing diesel fuel with an average
sulfur level of 2 ppm, a reasonable average under a 5 ppm cap. Mathpro
examined two sets of cases where average on-highway diesel fuel sulfur
levels were reduced from 20 ppm to 2 ppm, one with nonroad diesel fuel
sulfur at 350 ppm (Cases 1 and MP1) and the other with nonroad diesel
fuel sulfur at 20 ppm (Cases 4 and 8). From these cases, Mathpro's
estimated cost of reducing highway diesel fuel sulfur from 20 ppm to 2
ppm ranges from 1.7 to 2.1 cents per gallon. Assuming a linear
relationship between sulfur and cost per gallon in this range, the cost
of reducing average sulfur levels from 7 ppm (that projected under the
proposed 15 ppm cap) to 2 ppm would be 0.7-0.8 cents per gallon.
Although it is possible that the cost per ppm of sulfur reduced would
actually increase as sulfur was reduced, the extent of this increase is
difficult to estimate. Thus, the best cost that we can project at this
time is 0.7-0.8 cents per gallon, incremental to the cost of the 15 ppm
sulfur cap program.
Although we have not attempted to analyze in detail the cost
impacts of distributing a fuel with a cap on sulfur content as low as 5
ppm, the American Petroleum Institute recently had a contractor do
so.\158\ That study estimated that, compared to current costs,
distribution costs would increase by 0.9 to 2.1 cents per gallon if a 5
ppm standard were adopted for the entire highway diesel pool.\159\ The
following reasons were cited for why, as the sulfur specification is
decreased, it becomes more difficult to maintain product purity and
supply:
---------------------------------------------------------------------------

\158\ ``Costs/Impacts of Distributing Potential Ultra Low Sulfur
Diesel, Turner, Mason, & Company Consulting Engineers,'' February
2000. EPA Docket A-99-06, item II-G-49.
\159\ ``Costs/Impacts of Distributing Potential Ultra Low Sulfur
Diesel, Turner, Mason, & Company Consulting Engineers,'' February
2000. EPA Docket A-99-06, item II-G-49.
---------------------------------------------------------------------------

--There is increased difficulty and cost associated with correcting
off-specification batches in the distribution system.
--Measurement accuracy becomes more limiting.
--The pipeline compliance margin becomes more limiting at refineries.
--Supply outages due to off-specification product will become more
common.
--The difference between the sulfur content of highway diesel fuel and
that of abutting higher sulfur products in the pipeline system becomes
larger.

Even with the estimated increase in distribution costs, the report
still concluded that it was probably impractical to attain continuous
supply availability of diesel fuel in all areas and outlets within the
current distribution system at a 5 ppm cap on fuel sulfur content. If
such problems are to be avoided, additional, more costly measures may
be necessary. Should a segregated distribution system be needed to
control contamination, including dedicated pipelines and tank trucks,
the costs would be considerably higher than the 0.9 to 2.1 cents per
gallon estimated in the report.
We too are concerned that the measures which form the basis for the
0.9 to 2.1 cents per gallon cost estimate in the API-sponsored study
may not ensure widespread compliance. Under a 5 ppm standard, sulfur
measurement variability would need to be reduced appreciably from
current tolerances, perhaps to a level of 1 ppm or less, and the test
equipment purchases and quality control steps needed to attain this
could prove costly. Yet the bulk of the impact would come from the
major shift likely to be needed in the practices used to avoid
contamination in the distribution system. Assuming an extremely
demanding maximum sulfur specification of 3 ppm at the refinery gate
and a test variability of 1 ppm, only 1 ppm contamination through the
distribution system could be tolerated, and this would need to be
maintained nationwide and year round in a distribution system that
routinely handles products with sulfur levels of up to several thousand
ppm. Refiners would also need to take additional measures to meet the 3
ppm refinery gate standard that would likely be set by pipeline
operators. Similar to the distribution system, the measures that
refiners would need to take to further reduce sulfur content and limit
process variability are unclear, and might prove quite costly.
The overall cost of a program with a 5 ppm sulfur cap is comprised
of the program's cost in producing and distributing the fuel, offset by
the cost of the projected 1 percent fuel economy gain. As the sulfur
level reaches this very low level, the types of process changes in the
refinery and fuel distribution systems necessary to eliminate
contamination and maintain sufficient process flexibility in the system
become much more uncertain. Consequently, serious concerns have

[[Page 35518]]

been raised concerning the ability to achieve a 5 ppm sulfur cap
without drastic and costly changes to how diesel fuel is produced and
distributed today. Nevertheless, assuming the average of the per gallon
production and distribution cost ranges discussed above, this
corresponds to a net $47.1 billion 30-year NPV cost, compared to $37.7
billion for the 15 ppm sulfur cap proposal. Considering the
NO_X emissions benefits (unchanged from the 15 ppm sulfur cap
case) and the PM emissions benefits (slightly improved), the resulting
aggregate cost effectiveness is projected to be $1900 per ton of
NO_X+NMHC and $4500 per ton of PM (including the
SO2 credit). These compare to $1500 per ton of
NO_X+NMHC and $1900 per ton of PM for the 15 ppm sulfur cap
proposal.
3. What About a 50 ppm Sulfur Level?
The American Petroleum Institute has proposed that we set a sulfur
cap for highway diesel fuel of 50 ppm with a required refinery output
average of 30 ppm, along with other proposal elements.\160\ API's
proposal is based on their assessment of technological need and
viability. Key to API's position is the view that, ``while EPA may set
standards to encourage advanced technology, EPA must not base a sulfur
level on a particular technology the Agency predicts might prove
viable.'' However, we believe that we must set standards in the context
of real technologies that can be expected to be feasible, rather than
as a means of generally encouraging advanced technology. With this in
mind, we have analyzed the impact that a 50 ppm sulfur cap would have
on technology enablement, vehicle and fuel costs, and emissions
reductions. The results of this analysis are provided below. Analysis
details are provided in the Draft RIA. We encourage comment on this
assessment, preferably accompanied by data and analysis supporting the
commenter's views.
---------------------------------------------------------------------------

\160\ Letter from Red Cavaney of API to EPA Administrator Carol
Browner, dated February 7, 2000, EPA docket A-99-06.
---------------------------------------------------------------------------

As discussed in detail in section III.F, we believe that diesel
fuel needs to be desulfurized to the 15 ppm level to enable emission
control technologies capable of meeting the proposed standards. Setting
a fuel sulfur cap of 50 ppm would require that the PM standard be set
at a less stringent level to accommodate the approximate tripling of
sulfate PM production in the trap compared to a 15 ppm cap. However, we
believe increased fuel sulfur would have an even larger effect on
robust trap regeneration than on sulfate production, bringing into
question the very viability of PM traps at the higher sulfur levels. As
discussed in section III.F.1, field experience in Sweden, where below
10 ppm diesel fuel sulfur is readily available, has been good.
Experience has also been good in regions without extended periods of
cold ambient conditions (such as the United Kingdom) using 50 ppm cap
low sulfur fuel. However, field tests in Finland, where colder winter
conditions are sometimes encountered (similar to many parts of the
United States), have revealed a failure rate of 10 percent, due to
insufficient trap regeneration. We believe that failures of the
severity experienced with 50 ppm fuel in Finland would be unacceptable.
These problems could become even more pronounced in light-duty
applications, which tend to involve cooler exhaust streams, making
regeneration more difficult. Field data with such applications is still
sparse.
One means of attempting to resolve these problems is through use of
an active regeneration mechanism, such as electric heaters or fuel
burners. These could potentially introduce additional hardware and fuel
consumption costs. They would also raise reliability concerns, based on
past experience with such approaches. Active regeneration failures in
PM traps would be of more concern than in NO_X exhaust
emission control devices because they involve the potential for
complete exhaust stream plugging, runaway regeneration at very high
temperatures, trap melting, engine stalling, and stranding of motorists
in severe weather. As a result, we do not consider dependence on active
PM trap regeneration to be a sufficient basis for establishing PM trap
feasibility.
NO_X adsorber technology would likely be infeasible with
50 ppm sulfur fuel as well, due to the rapid poisoning of
NO_X storage sites. Desulfation would be needed much more
frequently and with a much higher resulting fuel consumption. Even if
the fuel economy penalty could somehow be justified, we expect that
overly frequent desulfation could cause unacceptable adsorber
durability or driveability problems (because of the difficulty in
timing the desulfation to avoid driving modes in which it might be
noticed by the driver). A less stringent NO_X standard could
help to mitigate these concerns by allowing the NO_X storage
bed to sulfate up to a greater degree before desulfating. However, this
might then cause deeper sulfate penetration into the storage bed and
thus possible long-term degradation because of the difficulty of
removing this deeper sulfate.
Instead, we expect that diesel fuel with an average fuel sulfur
level of 30 ppm and a cap of 50 ppm could enable lean NO_X
catalyst technology (described in section III.E). These devices can
provide modest NO_X reductions and, because of their reliance
on precious metal catalyst, also serve the function of a diesel
oxidation catalyst, removing some of the gaseous hydrocarbons and the
soluble organic fraction of PM. Unfortunately, lean NO_X
catalysts also share the oxidation catalyst's tendency to convert fuel
sulfur into sulfate PM, and do so even more aggressively because they
require higher precious metal loadings to reduce NO_X. They
also require a fairly large addition of diesel fuel to accomplish
NO_X reduction, typically about 4 percent or more of total
fuel consumption. The injected fuel also makes it difficult to achieve
an overall hydrocarbon reduction, despite the potential to convert much
of the engine-out hydrocarbons over the catalyst. Typically, current
lean NO_X catalyst designs actually show a net hydrocarbon
increase.
We have assumed that lean NO_X catalysts could be
developed over time to deliver 20 percent reductions in NO_X
(well beyond their current proven performance over the Federal Test
Procedure) with a net PM reduction of 20 percent and no net increase in
gaseous hydrocarbons with a 4 percent fuel economy penalty. Although
this PM reduction level is below that achieved by current diesel
oxidation catalysts, it represents an ambitious target to designers
attempting to balance NO_X reduction with sulfate production
from the still substantial sulfur in the fuel. We have estimated that
lean NO_X catalysts (including their diesel oxidation
catalyst function) would add an average long term cost of $603 to a
heavy-duty vehicle, inclusive of maintenance savings realized through
the use of low sulfur fuel. This is lower than the cost increase for
technologies enabled by 15 ppm sulfur fuel.
Based on the 20% expected emission reductions, we believe the
appropriate emissions standards at a 30 ppm average / 50 ppm cap diesel
sulfur level would be 1.8 g/bhp-hr NO_X and 0.08 g/hp-hr PM.
Because the enabled technologies do not allow very large emission
reductions and stringent emission standards, it is conceivable that
continued progress in engine design may eventually allow these
standards to be met through improvements in EGR and combustion
optimization, although we cannot outline such a technology path at this
time. It is likely that such a path would still involve a substantial
fuel economy penalty.

[[Page 35519]]

The 50 ppm sulfur cap would therefore result in projected
NO_X and PM emission reductions in 2020 of 540,000 and 17,000
tons per year, respectively, compared to 2.0 million and 83,000 tons
per year for a 15 ppm cap. It should be noted that virtually none of
the PM reduction comes from a reduction in the soot component of PM.
The cost of meeting a 50 ppm sulfur cap at the refinery would be
substantially less costly than meeting the proposed cap of 15 ppm. In
some cases, refiners may be able to meet a 50 ppm cap with only
relatively minor capital investment of a few million dollars for a new
hydrogen sulfide scrubbing unit and a PSA unit to increase hydrogen
purity. New, high activity catalyst would also replace today's
catalyst. In other cases, refiners would also have to add a second
reactor. Finally, some refiners would require essentially the same two-
stage hydrotreating unit that would be required to meet the proposed 15
ppm standard. In all cases, hydrogen consumption would be somewhat less
than that required to meet the proposed 15 ppm standard.
Refiners who would be capable of meeting a 50 ppm cap with only
minor capital investment would likely be those not blending any LCO
into their diesel fuel, or those having substantial excess
hydrotreating capacity in their current unit. We estimate that about 15
percent of on-highway diesel fuel production would fall into this
category. Refiners blending some LCO into their diesel fuel (e.g., 15
percent or less), or with somewhat greater levels of LCO but also
having significant excess current hydrotreating capacity, would likely
be capable of meeting a 50 ppm cap with an additional reactor. We
estimate that about 35 percent of on-highway diesel fuel production
would fall into this category. Finally, about 50 percent of on-highway
diesel fuel production would likely require a two-stage hydrotreating
unit due to their higher LCO fraction or lack of excess current
hydrotreating capacity. Overall, we project that the average cost of
meeting the 50 ppm standard at the refinery would be about 2.3 cents
per gallon, about 1.7 cents per gallon less than the corresponding cost
for fuel meeting a 15 ppm sulfur cap.
It would be slightly less expensive to distribute the 50 ppm sulfur
fuel than the15 ppm sulfur fuel. The pipeline interface between highway
diesel fuel and higher sulfur products that must be sold with the
higher sulfur product to ensure quality of the highway diesel fuel
could be reduced. We estimate the cost savings per gallon of diesel
fuel to be about 0.01 cents.
The overall cost of a program with a 50 ppm sulfur cap with a 30
ppm average is comprised of the hardware cost of lean NO_X
catalyst technology, the cost increase in producing and distributing
the fuel, and the cost of the projected 4% fuel economy loss. This
corresponds to a net $35.4 billion 30-year NPV cost, compared to $37.7
billion for the 15 ppm sulfur cap proposal. Considering the PM and
NO_X emissions benefits, the resulting aggregate cost
effectiveness is projected to be $3600 per ton of NO_X+NMHC
and $56,700 per ton of PM (including the SO2 credit). These
compare to $1500 per ton of NO_X+NMHC and $1900 per ton of PM
for the 15 ppm sulfur cap proposal. The large difference in PM cost
effectiveness is primarily due to the fuel economy penalty and the fact
that none of the fuel cost could be allocated to hydrocarbon control,
because of the lack of a hydrocarbon benefit.
Table VI.B-5 summarizes key emissions and cost impacts of a program
adopting the sulfur levels analyzed. Note that, although the analysis
finds that a 15 ppm average/25 ppm cap standard has potential to be
adequate for enabling high-efficiency exhaust emissions controls, this
finding involves a significantly higher level of uncertainty than the
proposed 15 ppm sulfur cap, because it is based on the assumption that
exhaust emission control designs could be focused on the average fuel
sulfur levels. We believe that the possibility of some in-use fuel at
near-cap levels would necessitate designing to accommodate this level,
and they contend that this would not allow the high-efficiency
technology to be enabled. If so, the technology enablement for this
case would likely be similar to that for the 50 ppm cap case. The
analysis results show that the 50 ppm cap case does not enable high-
efficiency exhaust control technology at all.

Table VI.B-5.--Summary of Emissions and Cost Impacts at Different Fuel Sulfur Levels
--------------------------------------------------------------------------------------------------------------------------------------------------------
2020 emission reductions Cost impacts
(thousand tons/year) ---------------------------------------------------------------
Sulfur level -------------------------------- Fuel Aggregate 30-
Vehicle ^c consumption Fuel ( cents/ yr NPV ($
NO_X PM (percent) gal) billion)
--------------------------------------------------------------------------------------------------------------------------------------------------------
5 ppm cap............................................... 2,020 86 $1,133 -1 ^d 6.0-7.3 ^d 47.1
15 ppm cap.............................................. 2,020 83 1,133 0 4.4 37.7
25 ppm cap w/15 ppm average ^a........................... 2,020 79 1,133 1 3.4 34.5
50 ppm cap w/30 ppm average^b ........................... 538 17 603 4 2.7 35.4
--------------------------------------------------------------------------------------------------------------------------------------------------------
\a\ Note that this sulfur level involves significant increased uncertainty with respect to technology enablement. Manufacturers have commented that the
possibility of some in-use fuel at or near the 25 ppm cap level would necessitate designing to accommodate this level, thus precluding high-efficiency
technology enablement, and making technology for this case similar to that for the 50 ppm cap case.
\b\ This sulfur level is not expected to enable high-efficiency exhaust control technology.
\c\ Costs of added hardware combined with lifetime maintenance cost impacts; figures shown for comparison purposes are long-term costs for heavy heavy-
duty vehicles.
\d\ Fuel cost based on industry analyses of refinery and distribution costs; costs could range much higher depending on fuel segregation measures
required.

We welcome comments on all aspects of these analyses for
alternative fuel sulfur standards, including the technology enablement
assessments, vehicle and fuel costs, emissions reductions, and cost
effectiveness.
4. What Other Fuel Properties Were Considered for Highway Diesel Fuel?
In addition to changes in highway diesel fuel sulfur content, we
also considered changes to other fuel properties such as cetane number,
aromatics, density, or distillation. Each of these fuel properties has
the potential to affect the combustion chemistry within the engine, and
so aid in reducing emissions of regulated pollutants. Indeed, some
manufacturers have made public statements to the effect that an
idealized highway diesel fuel is necessary in order to optimize

[[Page 35520]]

the efficiency of the next generation of heavy-duty diesel vehicles.
The focus of the fuel changes we are proposing today is to enable
diesel engines to meet much more stringent emission standards. As
described earlier in this section, we believe that diesel engines can
meet much more stringent emission standards using advanced exhaust
emission control systems, but the performance of these systems is
dramatically reduced by sulfur. Thus, we have determined that sulfur in
diesel fuel would need to be lowered. It does not appear that other
fuel properties have the same sort of effect on advanced exhaust
emission controls, and as a result we do not believe that changes in
fuel properties other than sulfur are necessary in order for heavy-duty
engines to reach the low emission levels offered by the advanced
exhaust emission controls discussed above. In fact, after conducting a
research study on this topic, industry members concluded that, ``If in
the future, fuel sulfur levels are significantly reduced in order to
enable efficient exhaust emission controls, then it should be
recognized that the exhaust emission control device becomes the primary
driver on tailpipe emissions and that all other fuel properties will
have only minor or secondary effects on the tailpipe emissions.'' \161\
---------------------------------------------------------------------------

\161\ Lee, et al., SAE 982649.
---------------------------------------------------------------------------

Emission reductions can also be achieved through changes in diesel
fuel properties as a direct means for reducing engine-out emissions. In
this approach, it is not the exhaust emission control which is being
``enabled,'' but rather the combustion process itself which is being
optimized. This approach has the advantage that the effects are fleet-
wide and immediate upon introduction of the new fuel, whereas new
engine standards do not produce significant emission reductions until
the fleet turns over. However, regulated changes in diesel fuel
properties may produce emission reductions that disappear over time, if
compliance test fuel is changed concurrently with the changes to in-use
fuel (to assure that such fuel remains representative of in-use fuels).
Manufacturers will redesign their new engines to take advantage of any
benefit a cleaner fuel provides, resulting in engines still meeting the
same emission standards in-use. Consequently, it would only be those
engines sold before the compliance test fuel changes that would be
likely to produce emission benefits, and as these engines drop out of
the fleet, so also would the benefit of changes to diesel fuel.
Even so, it is useful to consider what emission reductions are
achievable through changes to non-sulfur diesel fuel properties. The
non-sulfur fuel properties most often touted as good candidates for
producing emission reductions from heavy-duty engines are cetane number
and aromatics content. According to correlations between these fuel
properties and emissions that have been presented in various published
documents, the effects are rather small. We have estimated that an
increase in cetane number from 44 to 50 would reduce both
NO_X and PM emissions by about 1 percent for the in-use fleet
in calender year 2004.\162\ Likewise a reduction in total aromatics
content from 34 volume percent to 20 volume percent would reduce both
NO_X and PM emissions by about 3 percent. We expect changes
in other fuel properties to produce emission reductions that are no
greater than these effects. These reductions are insignificant in
comparison to the emission benefits projected to result from today's
proposal, and would come at a considerable refining cost. As a result,
at this time we do not believe that it is appropriate to require
changes to non-sulfur diesel fuel properties as a means for producing
reductions in engine-out emissions. There may, however, be performance
or engine design optimization benefits associated with non-sulfur
changes to diesel fuel that could justify their cost. Therefore we
welcome cross-industry collaboration on voluntary diesel fuel
improvements beyond the sulfur reduction proposed in this notice, and
we continue to solicit information on the impact of non-sulfur fuel
changes on exhaust emission control, engine-out emissions, and engine
design and performance.
---------------------------------------------------------------------------

\162\ ``Exhaust emissions as a function of fuel properties for
diesel-powered heavy-duty engines,'' memorandum from David Korotney
to EPA Air Docket A-99-06, September 13, 1999.
---------------------------------------------------------------------------

C. Should Any States or Territories Be Excluded From This Rule?

1. What Are the Anticipated Impacts of Using High-Sulfur Fuel in New
and Emerging Diesel Engine Technologies if Areas Are Excluded From This
Rule?
Section III discusses the technological feasibility of the emission
standards being proposed today and the critical need to have sulfur
levels reduced to 15 ppm for the technology to achieve these emission
standards. The implications to be drawn from section III with regard to
exemptions from the sulfur standards for States and Territories is
fairly straightforward. If vehicles and engines employing these
technologies to achieve the proposed emission standards will be
operated in these states or territories, then low-sulfur diesel fuel
must be available for their use.
Some have suggested allowing persons in Alaska to remove emission
control equipment to enhance the viability of using high-sulfur fuel.
In addressing this issue, we note that, under the Clean Air Act, it is
prohibited in all 50 states to remove emission control equipment from
an engine, unless that equipment is damaged or not properly
functioning, and then is replaced with equivalent properly functioning
equipment.
2. Alaska

a. Why is Alaska Unique?

There are important nationwide environmental and public health
benefits that can be achieved with cleaner diesel engines and fuel,
particularly from reduced particulate emissions, nitrogen oxides, and
air toxics (as further discussed in section II). Therefore, it is also
important to implement this program in Alaska. Any 2007 and later model
year diesel vehicles in Alaska would have to be fueled with low sulfur
highway diesel, or risk potential damage to the aftertreatment
technologies or even the engines themselves. Although the engine
standards proposed today do not have different technology and cost
implications for Alaska as compared to the rest of the country, the low
sulfur fuel program would have different implications (described
below). Therefore, in evaluating the best approach for implementing the
low sulfur fuel program, it is important to consider the extremely
unique factors in Alaska.
Section 211(i)(4) provides that the states of Alaska and Hawaii may
seek an exemption from the 500 ppm sulfur standard in the same manner
as provided in section 325 of the Clean Air Act. Section 325 provides
that upon request of Guam, American Samoa, the Virgin Islands, or the
Commonwealth of the Northern Mariana Islands, EPA may exempt any person
or source, or class of persons or sources, in that territory from any
requirement of the CAA, with some specific exceptions. The requested
exemption could be granted if EPA determines that compliance with such
requirement is not feasible or is unreasonable due to unique
geographical, meteorological, or economic factors of the territory, or
other local factors as EPA considers significant.
Unlike the rest of the nation, Alaska is currently exempt from the
500 ppm

[[Page 35521]]

sulfur standard for highway diesel fuel (as discussed in section c
below). Since the beginning of the 500 ppm highway diesel fuel program,
we have granted Alaska exemptions from meeting the sulfur standard and
dye requirements, because of its unique geographical, meteorological,
air quality, and economic factors. These unique factors are described
in more detail in the Draft Regulatory Impact Analysis contained in the
docket.
Second, in Alaska, unlike in the rest of the country, diesel fuel
consumption for highway use represents only five percent of the State's
total distillate fuel consumption, because of the relatively small
numbers of vehicles in the State. Most of this fuel is produced by
refineries located in Alaska, primarily because of the more severe
cloud point specification needed for the extremely low temperatures
experienced in much of Alaska during the winter. There are four
commercial refineries in Alaska. Only one of these refineries currently
has any desulfurization capacity, which is relatively small.
Consequently, because these refineries would have to reduce sulfur from
uncontrolled levels to meet the proposed 15 ppm standard, these
refineries could incur substantially higher costs than those in the
rest of the nation. Given the very small highway diesel demand,
however, it is doubtful that more than one or two Alaska refineries
would choose to produce low sulfur highway fuel, and these refiners
could even decide to import it from refineries outside of Alaska.
Third, Alaska's highway diesel vehicle fleet is relatively small,
particularly outside the Federal Aid Highway System. The State
estimates that there are less than 9000 diesel vehicles in the entire
State, with less than 600 of these vehicles in all of rural Alaska. The
State also indicates that these vehicles are predominantly older than
the average elsewhere.\163\
---------------------------------------------------------------------------

\163\ See further discussion in the Draft RIA (Chapter VIII).
---------------------------------------------------------------------------

Finally, Alaska's fuel distribution system faces many unique
challenges. Unlike the rest of the country, because of its current
exemption from the 500 ppm sulfur standard, Alaska does not currently
segregate highway diesel fuel from that used for off-road, marine,
heating oil, and other distillate uses. Therefore, the distribution
system costs for segregating a low sulfur grade of diesel for highway
uses will be significant. The existing fuel storage facilities limit
the number of fuel types that can be stored. In addition to significant
obstacles to expanding tankage in Alaska, the cost of constructing
separate storage facilities, and providing separate tanks for
transporting low-sulfur diesel fuel (e.g., by barge or truck), could be
significant. Most of Alaska's communities rely on barge deliveries, and
ice formation on the navigable waters during the winter months
restricts fuel delivery to these areas. Construction costs are 30
percent higher in Alaska than in the lower-48 states, due to higher
costs for freight deliveries, materials, electrical, mechanical, and
labor. There is also a shorter period of time during which construction
can occur, because of seasonal extremes in temperature and the amount
of daily sunlight.

b. What Flexibilities Are We Proposing for Alaska?

Because of the unique circumstances in Alaska, we are proposing an
alternative option for implementing the low sulfur fuel program in
Alaska. We are proposing to provide the State an opportunity to develop
an alternative low sulfur transition plan for Alaska. We would intend
to facilitate the development of this plan by working in close
cooperation with the State and key stakeholders. This plan would need
to ensure that sufficient supplies of low sulfur diesel fuel are
available in Alaska to meet the demand of any new 2007 and later model
year diesel vehicles. Given that Alaska's demand for highway diesel
fuel is very low and only a small number of new diesel vehicles are
introduced each year, it may be possible to develop an alternative
implementation plan for Alaska in the early years of the program that
provides low sulfur diesel only in sufficient quantities to meet the
demand from the small number of new diesel vehicles. This would give
Alaska refiners more flexibility during the transition period because
they would not have to desulfurize the entire highway diesel volume.
Our goal in offering this additional flexibility would be to transition
Alaska into the low sulfur fuel program in a manner that minimizes
costs, while still ensuring that the new vehicles receive the low
sulfur fuel they need. We expect that the transition plan would begin
to be implemented at the same time as the national program, but the
State would have an opportunity to determine what volumes of low sulfur
fuel would need to supplied, and in what timeframes, in different areas
of the State.
At a minimum, such a transition plan would need to: (1) Ensure an
adequate supply (either through production or imports), (2) ensure
sufficient retail availability of low sulfur fuel for new vehicles in
Alaska, (3) address the growth of supply and availability over time as
more new vehicles enter the fleet, (4) include measures to prevent
misfueling, and (5) ensure enforceability. We would anticipate that, to
develop a workable transition plan, the State would likely work in
close cooperation with refiners and other key stakeholders, including
retailers, distributors, truckers, engine manufacturers, environmental
groups, and other interested groups. For example, the State would
likely rely on input from the trucking industry in determining the
expected low sulfur fuel volume needed in Alaska, based on the
anticipated number of new vehicles, and how this volume is expected to
grow during the first few years of the program. Similarly, the State
would likely rely on the Alaska refiners' input regarding plans for
supplying (either through production or imports) low sulfur fuel to
meet the expected demand. Further, the State would likely rely on input
and cooperation from retailers and distributors to determine at which
locations the low sulfur fuel should be made available. Retailers
offering low sulfur fuel would have to take measures to prevent
misfueling, such as pump labeling. All parties in the distribution
system would need to ensure the low sulfur fuel remains segregated and
take measures to prevent sulfur contamination, in the same manner as
described for the national program in section VIII.
If the State anticipates that the primary demand for low sulfur
fuel will be along the highway system (e.g., to address truck traffic
from the lower 48 states) in the early years of the program, then the
initial stages of the transition plan could be focused in these areas.
We believe it would be appropriate for the State to consider an
extended transition schedule for implementing the low sulfur program in
rural Alaska, as part of the state's overall plan, based on when they
anticipate the introduction of a significant number of 2007 and later
model year vehicles in the remote areas.
Under such an approach, the State would be given the opportunity to
develop such a transition plan, as an alternative to the national
program, and submit it to EPA. Our goal would be to help facilitate the
development of the plan, by working closely with the State and the
stakeholder group so they would have an opportunity to address EPA's
concerns in their submittal. We envision that the State would develop
and submit this plan to EPA within about one year of the final diesel
rule. Our goal would be to conduct a rulemaking and publish a final
rule

[[Page 35522]]

promulgating a new regulatory scheme for Alaska, if appropriate. The
goal would be to issue a final rule within one year of Alaska's
submittal of the plan, so that refiners and other affected parties
would have certainty as to their regulatory requirements. We request
comment on the timing for the State to submit such an alternative plan,
and for EPA to conduct the rulemaking action. If the State chose not to
submit an alternative plan, or if the plan did not provide a reasonable
alternative for Alaska as described above, then Alaska would be subject
to the national program.
We seek comment on all aspects of this approach, and on other
approaches that may have merit, to provide additional flexibility in
transitioning the low sulfur fuel program for Alaska.

c. How Do We Propose to Address Alaska's Petition Regarding the 500 ppm
Standard?

Background

On February 12, 1993, Alaska submitted a petition under section 325
of the Act to exempt highway vehicle diesel fuel in Alaska from
paragraphs (1) and (2) of section 211(i) of the Act, except for the
minimum cetane index requirement.\164\ The petition requested that we
temporarily exempt highway vehicle diesel fuel in communities served by
the Federal Aid Highway System from meeting the sulfur content
specified in section 211(i) of the Act and the dye requirement for non-
highway diesel fuel of 40 CFR 80.29, until October 1, 1996. The
petition also requested a permanent exemption from those requirements
for areas of Alaska not reachable by the Federal Aid Highway System--
the remote areas. On March 22, 1994, (59 FR 13610), we granted the
petition based on geographical, meteorological, air quality, and
economic factors unique to Alaska.
---------------------------------------------------------------------------

\164\ Copies of information regarding Alaska's petition for
exemption and subsequent requests by Alaska and actions by EPA are
available in public docket A-96-26.
---------------------------------------------------------------------------

On December 12, 1995, Alaska submitted a petition for a permanent
exemption for all areas of the State served by the Federal Aid Highway
System, that is, those areas covered only by the temporary exemption.
On August 19, 1996, we extended the temporary exemption until October
1, 1998 (61 FR 42812), to give us time to consider comments to that
petition that were subsequently submitted by stakeholders. On April 28,
1998 (63 FR 23241) we proposed to grant the petition for permanent
exemption. Substantial public comments and substantive new information
were submitted in response to the proposal. To give us time to consider
those comments and new information, we extended the temporary exemption
for another nine months until July 1, 1999 (September 16, 1998, 63 FR
49459). During this time period, we started work on a nationwide rule
to consider more stringent diesel fuel requirements, particularly for
the sulfur content (i.e., today's proposed rule). To coordinate the
decision on Alaska's request for a permanent exemption with this
nationwide rule on diesel fuel quality, we extended the temporary
exemption until January 1, 2004 (June 25, 1999 64 FR 34126).

Today's Proposed Action

As mentioned above, Alaska has submitted a petition for a permanent
exemption from the 500 ppm standard for areas not served by the Federal
Aid Highway System. Our goal is to take action on this petition in a
way that minimizes costs through Alaska's transition to the low sulfur
program. The cost of compliance could be reduced if Alaska refiners
were given the flexibility to meet the low sulfur standard in one step,
rather than two steps (i.e., once for the current 500 ppm sulfur
standard in 2004 when the temporary exemption expires, and again for
the proposed 15 ppm standard in 2006). Therefore, we propose to extend
the temporary exemption for the areas of Alaska served by the Federal
Aid Highway System from January 1, 2004 (the current expiration date)
to the proposed effective date for the proposed 15 ppm sulfur standard
(i.e., April 1, 2006 at the refinery level; May 1, 2006 at the terminal
level; and June 1, 2006 at all downstream locations).
As discussed in section b above, we are proposing to allow Alaska
to develop a transition plan for implementing the 15 ppm sulfur
program. During this transition period, it is possible that both 15 ppm
(for proposed 2007 and later model year vehicles) and higher sulfur
(for older vehicles) highway fuels might be available in Alaska. To
avoid the two-step sulfur program described above, we seek comment on
whether we should consider additional extensions to the temporary
exemption of the 500 ppm standard beyond 2006 (e.g., for that portion
of the highway pool that is available for the older technology vehicles
during Alaska's transition period). We would expect that any additional
temporary extensions, if appropriate, would be made in the context of
the separate rulemaking taking action on Alaska's transition plan (as
described in the previous section).
As in previous actions to grant Alaska sulfur exemptions, we would
not base any vehicle or engine recall on emissions exceedences caused
by the use of high-sulfur (>500 ppm) fuel in Alaska during the period
of the temporary sulfur exemption. In addition, manufacturers may have
a reasonable basis for denying emission related warranties where damage
or failures are caused by the use of high-sulfur (>500 ppm) fuel in
Alaska.
Finally, the costs of complying could be reduced significantly if
Alaska were not required to dye the non-highway fuel. Dye contamination
of other fuels, particularly jet fuel, is a serious potential problem.
This is a serious issue in Alaska since the same transport and storage
tanks used for jet fuel are generally also used for other diesel
products, including off-highway diesel products which are required to
be dyed under the current national program. This issue is discussed
further in the Draft RIA (Chapter VIII). Therefore, we also propose to
grant Alaska's request for a permanent exemption from the dye
requirement of 40 CFR 80.29 and 40 CFR 80.446 for the entire State.
We are interested in comments on all aspects of this proposal.
3. American Samoa, Guam, and the Commonwealth of Northern Mariana
Islands

a. Why Are We Considering Excluding American Samoa, Guam, and the
Commonwealth of Northern Mariana Islands?

Prior to the effective date of the current highway diesel sulfur
standard of 500 ppm, the territories of American Samoa, Guam and the
Commonwealth of Northern Mariana Islands (CNMI) petitioned EPA for an
exemption under section 325 of the Act from the sulfur requirement
under section 211(i) of the Act and associated regulations at 40 CFR
80.29. The petitions were based on geographical, meteorological, air
quality, and economic factors unique to those territories. We
subsequently granted the petitions.\165\ With today's proposal we need
to evaluate whether to include or exclude the territories in areas for
which the fuel sulfur standard would apply.
---------------------------------------------------------------------------

\165\ See 57 FR 32010, July 20, 1992 for American Samoa; 57 FR
32010, July 30, 1992 for Guam; and 59 FR 26129, May 19, 1994 for
CNMI.
---------------------------------------------------------------------------

b. What are the Relevant Factors?

The key relevant factors unique to these territories, briefly
discussed below, are discussed in detail in the

[[Page 35523]]

Draft RIA. These U.S. Territories are islands with limited
transportation networks. Consequently among these three territories
there are currently only approximately 1300 registered diesel vehicles.
Diesel fuel consumption in these vehicles represents just a tiny
fraction of the total diesel fuel volume consumed in these places; the
bulk of diesel fuel is burned in marine, nonroad, and stationary
applications. Consequently highway diesel vehicles are believed to have
a negligible impact on the air quality in these territories, which,
with minor exceptions, is very good.
All three of these territories lack internal petroleum supplies and
refining capabilities and rely on long distance imports. Given their
remote location from the U.S. mainland, petroleum products are imported
from east rim nations, particularly Singapore. Although Australia, the
Philippines, and certain other Asian countries have or will soon
require low-sulfur diesel fuel, this requirement is a 500 ppm sulfur
limit, not the proposed 15 ppm sulfur limit. Compliance with low-sulfur
requirements for highway fuel would require construction of separate
storage and handling facilities for a unique grade of diesel fuel for
highway purposes, or importation of low-sulfur diesel fuel for all
purposes, either of which would significantly add to the already high
cost of diesel fuel in territories which rely heavily on United States
support for their economies.

c. What Are the Options and Proposed Provisions for the Territories?

We could include or exclude the territories in the areas for which
the proposed diesel fuel sulfur standard would apply. As in the early
1990's when the 500 ppm sulfur standard was implemented, we believe
that compliance with the proposed 15 ppm sulfur standard would result
in relatively small environmental benefit, but major economic burden.
We are also concerned about the impact to vehicle owners and operators
of running the new and upcoming engine and emission control
technologies using high-sulfur fuel. We believe that for the sulfur
exemption to be viable for vehicle owners and operators, they would
need access to either low-sulfur fuel or vehicles meeting the pre-2007
HDV emission standards that could be run on high-sulfur fuel without
significant engine damage or performance degradation.
We are proposing to exclude American Samoa, Guam and CNMI from the
proposed diesel fuel sulfur requirement of 15 ppm because of the high
economic cost of compliance and minimal air quality benefits. We are
also proposing to exclude, but not prohibit, the territories from the
2007 heavy-duty diesel vehicle and engine emissions standards, and
other requirements associated with those emission standards based on
the increased costs associated with implementing the vehicle and fuel
standards together in these territories. Thus, the territories would
continue to have access to 2006 diesel vehicle and engine technologies.
This exclusion from standards would not apply to gasoline engines and
vehicles because gasoline that complies with our regulations will be
available, and so concerns about damage to engines and emissions
control systems will not exist. As proposed this exclusion from
standards does not apply to light-duty diesel vehicles and trucks
because gasoline vehicles meeting the emission standards and capable of
fulfilling the same function would be available.
We are proposing to continue requiring all diesel motor vehicles
and engines to be certified and labeled to the applicable requirements
(either to the 2006 model year standards and associated requirements,
or to the standards and associated requirements applicable for the
model year of production) and warranted, as otherwise required under
the Clean Air Act and EPA regulations. Special recall and warranty
considerations due to the use of exempted high-sulfur fuel are proposed
to be the same as those proposed for Alaska during its proposed
transition period. To protect against this exclusion being used to
circumvent the emission requirements applicable to the rest of the
United States (i.e., continental United States, Alaska, Hawaii, Puerto
Rico and the U.S. Virgin Islands) after 2006 by routing pre-2007
technology vehicles and engines through one of these territories, we
propose to restrict the importation of vehicles and engines from these
territories into the rest of the United States. After the 2006 model
year, diesel vehicles and engines certified under this exclusion to
meet the 2006 model year emission standards for sale in American Samoa,
Guam and CNMI would not be permitted entry into the rest of the United
States.
We request comment on these exclusions and particularly on whether
it should be extended to light-duty diesel vehicle and truck standards
as well.

D. What About the Use of JP-8 Fuel in Diesel-Equipped Military
Vehicles?

In 1995, EPA issued a letter to the Deputy Under Secretary of
Defense for Environmental Security which concluded that the military
specification fuel known as JP-8 did not meet the definition of diesel
fuel under EPA's regulations and was, therefore, not subject to the
0.05 percent by weight sulfur standard. EPA also determined that
despite the slightly higher sulfur levels, the use of JP-8 in motor
vehicles by the military would not be a violation of EPA regulations as
a matter of policy. This decision was made after careful consideration
of the impact on operational readiness, logistical considerations and
cost for the military. EPA also evaluated data presented by the
military which compared the emissions of vehicles operated on typical
highway diesel and JP-8. These data supported the conclusion that there
would not be a significant adverse environmental consequence from the
limited use of JP-8 fuel. EPA's evaluation of the emissions impact was,
of course, based on the results of tests conducted using vehicles
representative of diesel emission control technology and diesel fuel in
use at that time.
The technical basis for EPA's decision on this matter may be
affected by the prospect of military vehicles equipped with the highly
sulfur sensitive technology that is expected to be used on vehicles and
engines designed to meet the standards for 2007 and beyond. We request
comment from interested parties on how to best deal with this
situation, including comment on the extent to which national security
exemptions pursued under 40 CFR 85.1708 may affect resolution of the
issue.

VII. Requirements for Engine and Vehicle Manufacturers

A. Compliance With Standards and Enforcement

We are not proposing any changes to the enforcement scheme
currently applicable to vehicles and engines under Title II of the CAA.
Thus, they would continue to apply to the vehicles and engines subject
to today's proposed standards. This includes the enforcement provisions
relating to the manufacture, importation and in-use compliance of these
vehicles and engines (see sections 202-208 of the CAA). Manufacturers
are required to obtain a certificate of conformity for their engine
designs prior to introducing them into commerce, and are subject to
Selective Enforcement Audits during production. Although there are

[[Page 35524]]

currently no regulatory requirements for manufacturers to test in-use
engines, they are responsible for the emission performance of their
engines in use. If we determine that a substantial number of properly
maintained and used engines in any engine family is not complying with
the standards in use, then we may require the manufacturer to recall
the engines and remedy the noncompliance. Failure by a manufacturer to
comply with the certification, warranty, reporting, and other
requirements of Title II can result in sanctions including civil
penalties and injunctive relief (see sections 202-208 of the CAA).
Other enforcement provisions regulating persons in addition to
manufacturers would also be applicable to the affected diesel vehicles,
including provisions such as the tampering and defeat device
prohibitions. It is also important to note that, because the CAA
defines manufacturer to include importers, all of these requirements
and prohibitions apply equally to importers.
Consideration has been given to in-use issues that may arise from
use of the new exhaust emission control technology. While it is
believed that the technology is sufficient to ensure that emission
control devices and elements of design will be effective throughout the
useful life of the vehicle, some concern has been expressed regarding
the possibility that instances of driveability or other operational
problems could occur in-use. One example brought up, is the possibility
that a vehicle could experience severe driveability problems if the PM
trap becomes plugged. At this time, however, we are confident that the
technologies will be developed to prevent these types of problems from
occurring provided the vehicle is operated on the appropriate fuel.
Nevertheless, comments are requested on any in-use problems that may
arise as a result of inclusion of exhaust emission control technology.
Your comments should address the nature of the problem, likelihood of
its occurrence and options for ensuring it does not occur.
Another issue related to certification is what (if any) maintenance
we should allow for adsorbers and traps. Our existing regulations
define these to be critical emission-related components, which means
that the amount of maintenance of them that the manufacturer is allowed
to conduct during durability testing (or specify in the maintenance
instructions that it gives to operators) is limited. We believe that
this is appropriate because, as we already noted, we expect that these
technologies will be very durable in use and will last the full useful
life with little or no scheduled maintenance. However, our existing
regulations (40 CFR 86.004-25) would allow a manufacturer to specify
something as drastic as replacement of the adsorber catalyst bed or the
trap filter after as little as 100,000-150,000 miles if there was a
``reasonable likelihood'' that the maintenance would get done. We are
concerned that some manufacturers may underdesign the adsorbers and
traps compared to the level of durability that is achievable. If this
occurred, even if most users replaced their adsorber or trap according
to the manufacturer's schedule, there would certainly be some users
that did not. Therefore, we are proposing to require that these
technologies be designed to last for the full useful life of the
engine. More specifically, the proposed regulations state that
scheduled replacement of the PM filter element or catalyst bed is not
allowed during the useful life. Only cleaning and adjustment will be
allowed as scheduled maintenance.
It may be appropriate to establish non-conformance penalties (NCPs)
for the standards being proposed today. NCPs are monetary penalties
that manufacturers can pay instead of complying with an emission
standard. In order for us to establish NCPs for a specific standard, we
would have to find that: (1) Substantial work will be required to meet
the standard for which the NCP is offered; and (2) there is likely to
be a ``technological laggard'' (i.e., a manufacturer that cannot meet
the standard because of technological (not economic) difficulties and,
without NCPs, might be forced from the marketplace). According to the
CAA (section 206(g)), such NCPs ``shall remove any competitive
disadvantage to manufacturers whose engines or vehicles achieve the
required degree of emission reduction.'' We also must determine
compliance costs so that appropriate penalties can be established. We
have established NCPs in past rulemakings. However, since the
implementation of our averaging, banking and trading program, their use
has been rare. We believe manufacturers have taken advantage of the
averaging, banking and trading program as a preferred alternative to
incurring monetary losses. At this time, we have insufficient
information to evaluate these criteria for heavy-duty engines. While we
believe that substantial work will be required to meet the 2007
standards, we currently have no information indicating that a
technological laggard is likely to exist. Recognizing that it may be
premature for manufacturers to comment on these criteria, since
implementation of these standards is still more than six years away, we
expect to consider NCPs in a future action. We welcome comment on this
approach.
Today's proposal includes PM standards for heavy-duty gasoline
engines. Because gasoline engines have inherently low PM emissions, it
may be appropriate in some cases to waive the requirement to measure PM
emissions. Therefore, we are proposing to maintain the flexibility to
allow manufacturers to certify gasoline engines without measuring PM
emissions, provided they have previous data, analyses, or other
information demonstrating that they comply with the standards. The
flexibility is the same as that allowed for PM emissions from light-
duty gasoline vehicles and for CO emissions from heavy-duty diesel
engines.

B. Certification Fuel

It is well established that measured emissions are affected by the
properties of the fuel used during the test. For this reason, we have
historically specified allowable ranges for test fuel properties such
as cetane and sulfur content. These specifications are intended to
represent most typical fuels that are commercially available in use.
Because today's action is proposing to lower the upper limit for sulfur
content in the field, we are also proposing a new range of allowable
sulfur content for testing that would be 7 to 15 ppm (by weight).
Beginning in the 2007 model year, these specifications would apply to
all emission testing conducted for Certification and Selective
Enforcement Audits, as well as any other laboratory engine testing for
compliance purposes. Because the same in use fuel is used for light-and
heavy-duty highway diesel vehicles, we are also proposing to change the
sulfur specification for light-duty diesel vehicle testing to the same
7 to 15 ppm range, beginning in the 2007 model year. We request comment
on these test fuel specifications. We also request comment regarding
whether the range of allowable test fuel properties should include the
full range of in-use properties or include the most typical range
around the average properties (e.g., 7 to 10 ppm sulfur).

C. Averaging, Banking, and Trading

We are proposing to continue the basic structure of the existing
ABT program for heavy-duty diesel engines. (Note that this includes the
Otto-cycle engine and vehicle ABT programs that were proposed on
October 29, 1999, 64 FR 58472.) This program allows manufacturers to
certify that their

[[Page 35525]]

engine families comply with the applicable standards on average. More
specifically, manufacturers are allowed to certify their engine
families with various family emission limits (FELs), provided the
average of the FELs does not exceed the standard when weighted by the
numbers of engines produced in each family for that model year. To do
this, they generate certification emission credits by producing engine
families that are below the applicable standard. These credits can then
be used to offset the production of engines in engine families that are
certified to have emissions in excess of the applicable standards.
Manufacturers are also allowed to bank these credits for later use or
trade them to other manufacturers. We are proposing some restrictions
to prevent manufacturers from producing very high-emitting engines and
unnecessarily delaying the transition to the new exhaust emission
control technology. These restrictions are described below. We are
continuing this ABT program because we believe that it would provide
the manufacturers significant compliance flexibility. This compliance
flexibility would be a significant factor in the manufacturers' ability
to certify a full line of engines in 2007 and would help to allow
implementation of the new, more stringent standard as soon as
permissible under the CAA. This is especially true given the very low
levels of the proposed standards. In some ways the ABT program is
intended to serve the same purpose as the phase-in for diesel engines.
As is described below, we have proposed some restrictions to make this
program compatible with the phase-in. Thus your comments on this ABT
program should address how it fits with the phase-in, and vice versa.
The existing ABT program includes limits on how high the emissions
from credit-using engines can be. These limits are referred to as FEL
caps. No engine family may be certified above these caps using credits.
These limits provide the manufacturers compliance flexibility while
protecting against the introduction of unnecessarily high-emitting
engines. In today's action, we are proposing to establish lower caps
for those engines that are required to comply with the proposed
standards. Specifically, we are proposing that the engines subject to
the new standards have NO_X emissions no higher than 0.50 g/
bhp-hr, and PM emissions no higher than 0.02 g/bhp-hr. Without this
cap, we are concerned that one or more manufacturer(s) could use the
ABT program to unnecessarily delay the introduction of exhaust emission
control technologies. Allowing this would be contrary to one of the
goals of the phase-in program, which is to allow manufacturers to gain
experience with these technologies on a limited scale before they are
applied to their full production. Similarly, we are proposing FEL caps
of 1.0 g/mi NO_X and 0.03 g/mi PM for chassis-certified
heavy-duty vehicles. We request comment on the need for and the levels
of these FEL caps.
We are proposing separate averaging sets during the phase-in
period. In one set, engines would be certified to the 2.4 g/bhp-hr
NO_X+NMHC standard (which applies for model years 2004-2006),
and would be subject to the restrictions and allowances established for
those model years. In the other set, engines would be certified to the
proposed 0.20 g/bhp-hr NO_X standard, and would be subject to
the restrictions and allowances proposed today. Averaging would not be
allowed between these two sets within the same model year. The reason
for this is similar to that for the low FEL caps. Allowing averaging
between the sets would be contrary to one of the goals of the phase-in
program, which is to allow manufacturers to introduce engines with
ultra-low emission technologies on a limited scale before they are
applied to their full production. We are concerned that manufacturers
could delay the introduction of NO_X aftertreatment
technology, diminishing the projected benefits of the proposed program
during the phase-in. We request comment on the need for this
restriction. As a part of this restriction of cross-set averaging, we
are also proposing that banked NO_X+NMHC and PM credits
generated from 2006 and earlier engines may not be used to comply with
the stricter standards that apply to 2007 and later engines (unless
such credits are generated from engines that meet all of the stricter
standards early). We are also requesting comments on alternatives to
these restrictions, such as only allowing banked credits generated from
engines below some threshold (e.g., 1.5 g/bhp-hr NO_X+NMHC or
0.05 g/bhp-hr PM) to be used for compliance with the 2007 standards.
Under the threshold approach, the credits would be calculated in
reference to the threshold rather than the applicable standard. Your
alternatives should address our two primary concerns: (1) Ensuring that
manufacturers produce engines during the phase-in period that are
equipped with the advanced NO_X aftertreatment controls; and
(2) ensuring that the program produces equivalent or greater emission
reductions during the phase-in period.
We propose to apply these same restrictions to the 2007 chassis-
based standards. This would affect the averaging program that was
proposed previously for model year 2004 (October 29, 1999, 64 FR
58472). We believe that these restrictions are equally necessary for
the chassis-based program, but are also open to alternatives. We are
particularly interested in the possibility of using the Tier 2 pull-
ahead approach that would allow manufacturers to phase in the new
standards on a per-vehicle basis rather than on a total gram basis.
Under this approach, for each ``2007-technology'' vehicle that a
manufacturer introduced before 2007, it could produce one ``2006-
technology'' vehicle in 2007 or later. We recognize that this approach
would be complicated for heavy-duty vehicles because of the different
weight classes, but believe that this problem could be addressed with
appropriate weighting factors (e.g, setting one 14,000 lb vehicle as
equivalent to two 8,500 lb vehicles). While it is less clear that such
an approach would work for the engine programs, we would welcome such
comments.
The Agency continues to be interested in the potential of early
benefits to be gained from retrofitting highway engines. Thus, we are
also asking for comment on various concepts by which manufacturers
could earn credits potentially to be used in a variety of programs. An
example of such credits in the 2007 MY program might include
consideration by EPA of the retiring of retrofit credits in deciding
whether to make a discretionary determination under section 207(c) of
substantial non-conformity. For discussion of related issues, see the
final rule for spark-ignition marine engines (61 FR 52088, 52095,
October 4, 1996), and the final rule for locomotive engines (63 FR
18978, 18988, April 16, 1998). We ask for comment as to what emission
benefits could be achieved by this concept and by what legal authority
such credits could be applied. Such systems would bring existing
highway engines into compliance with the standards being proposed for
new engines, or alternately with some less stringent standards levels
that still achieve large emission reductions. We ask comment on how
such an emissions reduction calculation should be formulated and how
such benefits and resulting credits should be applied. Certification
requirements for such retrofit systems could be developed along the
lines of those adopted in EPA's urban bus retrofit program (58 FR
21359, April 21, 1993). Credits would be

[[Page 35526]]

calculated based on the expected lifetime emissions benefits of the
retrofit systems. Because this benefit depends on the remaining life of
the retrofitted vehicle, and this could vary considerably, any emission
reduction formula would require the certainty to account for this in
the calculation, such as by estimating an average remaining life for
retrofits in each engine family, or by using a vehicle age-dependent
proration factor for each retrofitted system, similar to the approach
taken in the locomotive emissions rule (see Appendix K of the
Regulatory Support Document for the locomotives final rule. 63 FR
18977, April 16, 1998).

D. Chassis Certification

Heavy-duty vehicles under 14,000 pounds can generally be split into
two groupings, complete and incomplete vehicles. Complete vehicles are
those that are manufactured with their cargo carrying container
attached. These vehicles consist almost entirely of pick-up trucks,
vans, and sport utility vehicles. Incomplete vehicles are those chassis
that are manufactured by the primary vehicle manufacturer without their
cargo carrying container attached. These chassis may or may not have a
cab attached. The incomplete chassis are then manufactured into a
variety of vehicles such as recreational vehicles, tow trucks, dump
trucks, and delivery vehicles.
Recently, we proposed to require all complete Otto-cycle vehicles
between 8,500 and 14,000 pounds to be certified to vehicle-based
standards rather than engine-based standards beginning in model year
2004 (October 29, 1999, 64 FR 58472). Under this proposal manufacturers
would test the vehicles in essentially the same manner light-duty
trucks are tested. We continue to believe this approach is reasonable
and are thus proposing to continue it with the more stringent
standards. We request comment regarding the possible mandatory or
voluntary application of this program to complete diesel vehicles under
14,000 pounds.

E. FTP Changes to Accommodate Regeneration of Aftertreatment Devices

It is possible that some of the exhaust emission control devices
used to meet the proposed standard will have discrete regeneration
events that could effect emission characteristics. For example,
NO_X adsorbers and actively regenerated PM traps each
incorporate discrete regenerations. The NO_X adsorber stores
NO_X under normal conditions until the NO_X storage
capacity is nearly full, at which point, the regeneration event is
triggered to purge the stored NO_X and reduce it across a
catalyst. Actively regenerated PM traps incorporate heating devices to
periodically initiate regeneration. In both cases, we would expect that
these regeneration events would be controlled by the engine computer,
and would thus be generally predictable. Even passively regenerating
catalytic PM trap designs can have discrete regeneration events.
Discrete regeneration events can be important because it is
possible for exhaust emissions to increase during the regeneration
process. The regeneration of a NO_X adsorber for instance,
could result in increased particulates, NMHC and NO_X due to
the rich exhaust gas required to purge and reduce the NO_X.
We expect that in most cases, the regeneration events would be
sufficiently frequent to be included in the measured emissions. Our
feasibility analysis projects very frequent regeneration of the
NO_X adsorbers, and continuously regenerating PM traps.
Nevertheless, this issue becomes a regulatory concern because it is
also conceivable that these emission storage devices could be designed
in such a way that a regeneration event would not necessarily occur
over the course of a single heavy-duty FTP cycle, and thus be
unmeasured by the current test procedure. Since these regeneration
events could produce increased emissions during the regeneration
process, it will be important to make sure that regeneration is
captured as part of the certification testing. We seek comment on the
need to measure regeneration emissions as part of each emission test,
and the best method of making such measurements.
In order to verify the emission levels during regeneration, we
propose that the transient FTP applicable for certification be repeated
until a regeneration occurs. The transient FTP will be repeated until a
regeneration event is confirmed. The emissions measured during the
cycle in which the regeneration occurs must be below the applicable
transient cycle standard. For example, if an actively regenerated
heavy-duty PM trap does not regenerate over the cold-soak-hot cycle,
the hot portion of the cycle will be repeated until a regeneration is
observed. The specific hot cycle with the highest emissions would be
used as the representative hot cycle, and its emissions would be
weighted with the cold cycle emissions (as is currently required) to
determine compliance with the composite emission standard for the cold-
soak-hot cycle. We seek comment on the proposed method of capturing
regeneration emissions and whether we should allow the manufacturers to
use the average hot-start emissions rather than the worst case.
This proposal is based on the assumption that the systems would
include a fairly high frequency of regeneration events (e.g., one
regeneration event per hour). We seek comment on the need to capture
regeneration emissions as part of the certification testing if the
regeneration events occur much less frequently. Similarly, we request
comment on the need to measure emissions during desulfurization of the
NO_X adsorber. Would it be appropriate to allow manufacturers
to use a mathematical adjustment of measured emissions to account for
increased emissions during infrequent regeneration or desulfurization
events? For example, if a system required a desulfurization after every
20 transient cycles, and PM emissions increased by 20 percent during
desulfurization, would it be appropriate to adjust measured emissions
upward by one percent (20 percent divided by 20 cycles)?

F. On-Board Diagnostics

OBD systems help ensure continued compliance with emission
standards during in-use operation, and they help mechanics to properly
diagnose and repair malfunctioning vehicles while minimizing the
associated time and effort. We implemented OBD requirements on light-
duty applications in the 1994 model year (58 FR 9468, February 19,
1993). We recently proposed OBD requirements for 8500 to 14,000 pound
heavy-duty gasoline and diesel applications (October 29, 1999, 64 FR
58472). The 8500 to 14,000 pound requirements are scheduled for
implementation in the 2004 model year with a phase-in running through
the 2006 model year; the 2007 model year would be the first year of 100
percent OBD compliance on 8500 to 14,000 pound applications. We are
currently working with industry to develop OBD requirements for the
over 14,000 pound heavy-duty gasoline and diesel engines. Those
requirements will be proposed in a separate rulemaking and are
anticipated to be effective on or before the 2007 model year;
consequently, we are not proposing them here.
As discussed in the October 29, 1999, proposed rule, OBD system
requirements would allow for potential inclusion of heavy-duty vehicles
and engines in inspection/maintenance programs via a simple check of
the OBD system. The OBD system must monitor emission control components
for any malfunction or deterioration that could cause exceedance of
certain emission thresholds. The OBD system also

[[Page 35527]]

notifies the driver when repairs are needed via a dashboard light, or
malfunction indicator light (MIL), when the diagnostic system detects a
problem.
An OBD system is important on heavy-duty vehicles and engines for
many reasons. In the past, heavy-duty diesel engines have relied
primarily on in-cylinder modifications to meet emission standards. For
example, emission standards have been met through changes in injection
timing, piston design, combustion chamber design, use of four valves
per cylinder rather than two valves, and piston ring pack design and
location improvements. In contrast, the proposed 2004 and 2007
standards represent a significant technological challenge that would
require use of EGR and exhaust emission control devices whose
deterioration or malfunction can easily go unnoticed by the driver. The
same argument is true for heavy-duty gasoline vehicles and engines;
while emission control is managed both with engine design elements and
exhaust emission control devices, the latter are the primary emission
control features. Because deterioration and malfunction of these
devices can go unnoticed by the driver, and because their sole purpose
is emissions control, some form of detection is crucial. An OBD system
is well suited to detect such deterioration or malfunction.
Today's proposal does not contain any new OBD requirements. The
vehicles and engines designed to comply with today's proposed emission
standards would be required to comply with the OBD requirements already
in place or proposed for implementation in the 2004 model year (i.e.,
light-duty and heavy-duty through 14,000 pounds). However, because some
of the existing OBD requirements are based on multipliers of the
applicable emission standards, we request comment regarding the effect
of the low levels of the proposed standards on these OBD requirements.
We believe that these requirements will be feasible for these engines.
If you believe that the OBD requirements will not be feasible, you
should include in your comments suggestions for how they should be
revised to make them feasible.
We are also requesting comment regarding whether there are new OBD
requirements that should be adopted for these exhaust emission control
technologies. Comments supporting new requirements should indicate
whether they would be intended only to prevent emission problems, or
would also be intended to prevent performance problems, such as exhaust
emission control plugging.

G. Supplemental Test Procedures

To ensure better control of in-use emissions, we recently proposed
(October 29, 1999, 64 FR 58472) \166\ to add two supplemental sets of
requirements for heavy-duty diesel engines: (1) A supplemental steady-
state test and accompanying limits; and (2) NTE Limits. Both types of
these proposed supplemental emission requirements are expressed as
multiples of the normal duty cycle-weighted emission standards, or FEL
if the engine is certified under the ABT program, whichever is
applicable. For example, the diesel engine NTE limit for NO_X
+ NMHC emissions from 2004 engines would be 1.25 times the 2.4 g/bhp-hr
emission standard, or 1.25 times the applicable FEL. Although we are
not proposing any changes to these requirements, we are requesting
comment on the feasibility of technologies needed to meet the standards
being proposed in this notice, in the context of applying these
multipliers to these new standards.
---------------------------------------------------------------------------

\166\ Today's notice proposes to apply the heavy-duty diesel NTE
and supplemental steady-state test provisions intended to be
finalized as part of the 2004 standards rulemaking. The October 29,
1999 proposal for that rule contained the description of these
provisions. We expect that a number of modifications will be made to
those provisions in the FRM for that rule based on feedback received
during the comment period. While the details of the final provisions
are not yet available, we will provide the necessary information in
the docket for this rule as soon as it becomes available in order to
allow for comment.
---------------------------------------------------------------------------

Like current requirements, these new requirements would apply to
certification, production line testing, and vehicles in actual use. All
existing provisions regarding standards (e.g., warranty, certification,
recall) would be applicable to these new requirements as well. The
steady-state test was proposed because it represents a significant
portion of in-use operation of heavy-duty diesel engines that is not
adequately represented by the FTP. The combination of these
supplemental requirements is intended to provide assurance that engine
emissions achieve the expected level of in-use emissions control over
expected operating regimes in-use. We stated in the previous NPRM that
we believed that compliance with these requirements would not require
manufacturers to add additional emission control technologies, but
would require manufacturers to put forth some effort to better optimize
their engines with respect to emissions over a broader range of
operating conditions. You should read the previous NPRM for more
detail. You should also read the comments that we received in response
to this proposal. In those comments, some engine manufacturers raised
concerns regarding the feasibility of implementing these requirements
in the 2004 model year, in the context of the technologies expected to
be seen in the 2004 time frame (principally cooled EGR, advanced fuel
injection systems, advanced turbo-charging systems).\167\ Many of these
comments question the feasibility of meeting the proposed NTE emission
limits under the high-load regions of the proposed NTE zone,
particularly under conditions of high temperature and/or altitude.
These comments are highlighted here because the resolution of these
issues for the 2004 diesel engine standards, may also be relevant to
today's rulemaking.
---------------------------------------------------------------------------

\167\ See, for example, comments from Engine Manufacturers
Association, Detroit Diesel Corporation, Navistar International
Transportation Corp., Mack Trucks Inc., in EPA Air Docket No. A-98-
32.
---------------------------------------------------------------------------

We plan to apply these requirements with the proposed 2007
standards in the same manner as they would be applied with the 2004
standards, if adopted. There is some concern that certain exhaust
emission control devices, though capable of providing large emission
reductions and performing robustly over a wide range of expected
operating conditions, may have degraded performance in some conditions
included in the NTE or supplemental steady-state testing requirements.
We are thus asking for comments and supporting data related to this
concern. Your comments should address the following questions:

--What is the relative ability of the emission control technologies
being considered in today's action to control emissions over the full
range of speeds and loads typically encountered in actual use? Are
there areas of the map in which the emission controls are significantly
less effective?
--What is the relative need for emission reduction for different areas
of the speed-load map?
--How do the emission control technologies being considered in today's
action perform at different ambient conditions?
--Are the multipliers proposed previously the most appropriate
multipliers for ensuring in-use emissions control on exhaust emission
control-equipped engines?
--Are there other cost effective approaches to controlling in-use
emissions for engines equipped with exhaust emission controls?
--Are the technological issues raised in the 2004 rulemaking equally
applicable to diesel engines featuring

[[Page 35528]]

advanced exhaust emission controls and designed to meet the proposed
2007 standards?

H. Misfueling Concerns

As explained in Section III, the emissions standards contained in
this proposal will likely make it necessary for manufacturers to employ
exhaust emission control devices that require low-sulfur fuel to ensure
proper operation. This proposal therefore restricts the sulfur content
of highway diesel fuel sold in the U.S. There are, however, some
situations in which vehicles requiring low-sulfur fuel may be
accidentally or purposely misfueled with higher-sulfur fuel. Vehicles
operated within the continental U.S. may cross into Canada and Mexico,
countries which have not confirmed that they plan to adopt the same low
sulfur requirements we are proposing here. In addition, high-sulfur
nonroad fuel may illegally be used by some operators to fuel highway
vehicles. Any of these misfueling events could seriously degrade the
emission performance of sulfur-sensitive exhaust emission control
devices, or perhaps destroy their functionality altogether.
There are, however, some factors that help to mitigate concerns
about misfueling. Most operators are very conscious of the need to
ensure proper fueling and maintenance of their vehicles. The fear of
large repair and downtime costs may often outweigh the temptation to
save money through misfueling.
The likelihood of misfueling in Canada and Mexico is lessened by
current cross-border shipment practices and prospects for eventual
harmonization of standards. Canada has historically placed a priority
on harmonization with U.S. vehicle emission standards. They have also
placed a priority on harmonization with U.S. fuels standards, as they
import a significant amount of fuel from the U.S. and do not want to
become a ``dumping ground'' for fuel that does not comply with U.S.
fuel standards. We think it likely therefore that Canada will harmonize
with the U.S. revised engine standards and the fuel sulfur levels
required to support those standards. This will offer vehicle owners the
option of refueling with low-sulfur fuel there. Even if Canada were to
lag the U.S. in mandating low-sulfur fuels, these fuels would likely
become available along major through routes to serve the needs of U.S.
commercial traffic that have the need to purchase it. In addition,
there is less potential for U.S. commercial vehicles needing low-sulfur
fuel to refuel in Canada because Canadian fuel is currently more costly
than U.S. fuel. As a result, most vehicles owners will prefer to
purchase fuel in the U.S., prior to entering Canada, whenever possible.
This is facilitated by large tractor-trailer trucks that can have long
driving ranges--up to 2,000 miles or so--and the fact that most of the
Canadian population lives within 100 miles of the United States/Canada
border.
In Mexico, the entrance of trucks beyond the border commercial zone
has been prohibited since before the conclusion of the North American
Free Trade Agreement in 1994. This prohibition applies in the U.S. as
well, as entrance of trucks into the U.S. beyond the border commerce
zone is also not allowed. Since these prohibitions are contrary to the
intent of the Free Trade Agreement, a timetable was established to
eliminate them.\168\ However, these prohibitions are a point of
contention between the U.S. and Mexico and remain in force at this
time.
---------------------------------------------------------------------------

\168\ See NAFTA, Volume II, Annex I, Reservations for Existing
Measures and Liberalization Commitments, Pages I-M-69 and 70, and
Pages I-U-19 and 20.
---------------------------------------------------------------------------

The NAFTA negotiations included creation of a ``corridor'' where
commercial truck travel occurs, and where Mexico is obligated to
provide ``low-sulfur'' fuel. At the time of the NAFTA negotiations,
``low-sulfur'' fuel was considered 500 ppm, which was the level needed
to address the needs of engines meeting the 1994 emission standards.
The travel prohibition currently in place may be lifted at some point.
At that time, the issue of assuring, for U.S. vehicles, fuel with a
sulfur level needed by the technology that results from this regulation
may need to be addressed.
Even considering these mitigating factors, we believe it is
reasonable to propose two additional measures with very minor costs to
manufacturers and consumers. First, we are proposing a requirement that
heavy-duty vehicle manufacturers notify each purchaser of a model year
2007 or later diesel-fueled vehicle that the vehicle must be fueled
only with the low-sulfur diesel fuel meeting our regulations. We
believe this requirement is necessary to alert vehicle owners to the
need to seek out low-sulfur fuel when operating in areas such as Canada
and Mexico where it may not be widely available. We are also proposing
that model year 2007 and later heavy-duty diesel vehicles must be
equipped by the manufacturer with labels on the dashboard and near the
refueling inlet that say: ``Ultra-Low Sulfur Diesel Fuel Only.'' We
request comment on the need for these measures, alternative suggestions
for wording, whether or not these requirements should exist for only a
limited number of years, and whether any vehicles certified to the new
standards without the need for low-sulfur fuel should be exempted. We
also request comment on whether additional measures are needed to
preclude misfueling, such as requiring that the new technology vehicles
be equipped with refueling inlet restrictors that can only accept
refueling nozzles from pumps that dispense low-sulfur fuel. We would
also need to require that these pumps (or the high-sulfur fuel pumps)
be correspondingly equipped with specialized nozzles or other devices
to complement the vehicle refueling inlet restrictor.

I. Light-Duty Provisions

We are proposing that the heavy-duty vehicle labeling and purchaser
notification requirements discussed in section VII.H be applied to the
light-duty diesel vehicles certified to the final Tier 2 standards as
well, because these vehicles are expected to require the low-sulfur
fuel and so would be equally susceptible to misfueling damage.

J. Correction of NO_X Emissions for Humidity Effects

Engine-out emissions of NO_X are known to be affected
significantly by the amount of moisture in the intake air. The water
absorbs heat which lowers combustion temperatures, and thus lowers
NO_X emissions. Our existing regulations include equations
that give correction factors to eliminate this effect. For example, if
the equation indicated that NO_X emissions measured on a
relatively high humidity day would be about three percent lower than
would be expected with standard humidity, they would be multiplied by
1.03 to correct them to standard conditions. However, these equations
were developed many years ago, based on data from older technology
engines. We are concerned that these equations may not be valid for
engines equipped with catalytic emission controls. It is possible that
with catalytic systems, the effect may be very different. Perhaps with
these newer technologies, the effect will not be significant and
correction factors will not be needed. Therefore, we are requesting
comment regarding the accuracy of the existing equations for engines
equipped with NO_X adsorbers, and the need for such
correction factors for the 2007 standards. To the extent possible, your
comments should address the broader issue of the need for correction
factors for NO_X and other

[[Continued on page 35529]]

Notices For
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994

[[pp. 35479-35528]] Control of Air Pollution From New Motor Vehicles: Proposed Heavy-

Local Navigation