Nf-i-OZ.
     REGULATORY  SUPPORT DOCUMENT
EMISSIONS  STANDARDS  FOR HEAVY-DUTY
           CLEAN-FUEL FLEETS
       U.S. Environmental Protection Agency

   Regulatory Development and Support Division
                  June,  1994

-------

-------
                       Table  of Contents
. 1.  Introduction	  1  -  l

 2.  Technology Assessment Chapter  	  2-1
     2.0.1 Introduction	 .  .  .  .  .  .  .  .  .  .  .2-1
     2.0.2 Gasoline Engines ...	2-1
          2.0.2.1  Fundamentals  of   Gasoline  Engines Which
          Impact  ...... 	  2-1
               NMHC + NOx emissions
               2.0.2.1.1 Air-to-Fuel Ratio  	  ...2-1
               2.0.2.1.2 Ignition Timing	  .  .  .2-2
               2.0.2.1.3 Combustion Chamber Design  .  .  .  .2-3
          2.0.2.2   Current   Gasoline   Emissions   Control
          Technology		.2-3
               2.0.2.2.1 Exhaust Gas Recirculation  .  .  .  .  2  -  4
               2.0.2.2.2 Aftertreatment  Systems    .  .  .  .  .2-4
                    2.0.2.2.2.1 Catalytic Converters  .  .  .2-4
                    2.0.2.2.2.2 Electronic Controls  ....  2  -  5
         . 2.0.2.3    Future    Gasoline   Emissions   Control
          Technology  . . . . . . .   . ".""'.'".• .  .  .  .  .  .  .  .2-5
               2.0.2.3.1    Electrically-Heated    Catalytic
               Converters	  ..  .  .  .2-6
               2.0.2.3.2     Closed-Coupled     Catalytic
               Converters . . . ...   . .  .	2-6
               2.0.2.3.3 Lean-Burn Calibration .  . . .  .  .  .2-6
          2.0.2.4  Expected  Approaches  for  Clean-Fuel   Gasoline
          Engines    	 .........  ..2-7
     2.3 Diesel Bnoinac	  2  -  7
          2.0.3.1   Technical   Background/Fundamentals    of
          Diesel	  2  -  7
               Engines Which Impact NMHC + NOx emissions
               2.0.3.1.1 Fuel System	  .  .  .  .  .2-8
               2.0.3.1.2 Air System  	  ....2-8
          2.0.3.2 Current Diesel Emissions Control Technology  -  9
               2.0.3.2.1 Retarded Injection Timing  .  .  .  .  2  -  9
               2.0.3.2.2 Injection Pressure ........  2  -  9
               2.0.3.2.3    Injector   Nozzle   Holes    and
               Diameters	   2 - 10
               2.0.3.2.4 Intake Air Turbocharging .  .  .  .2-10
               2.0.3.2.5 Aftercooling	   2-10
               2.0.3.2.6 Cylinder Design  	   2-10
          2.0.3.3    Future    Diesel   Emissions    Control
          Technology	   2-11
               2.0.3.3.1 Exhaust Gas Recirculation  ...   2-11
               2.0.3.3.2 Improved Turbocharging  .....   2-12
               2.0.3.3.3 Injection Rate Shaping  .  .  .  .  .2-12
               2,0.3.3.4 Electronic Controls  ......   2-12
               2.0.3.3.5 Aftertreatment Devices  .  .  .  .  .2-13
                    2.0.3.3.5.1 Particulate Trap-Oxidizers 2 - 13
                    2.0.3.3.5.2  Catalytic   Oxidizing   Con-

-------

-------
                    verters	 .   2 - 14
                    2.0.3.3.5.3  Catalytic NOx reduction .   2 - 14
               2.0.3.3.6 Variable Compression Ratio .  . .   2-15
          2.0.3.4  Expected  Approaches  for  Clean-Fuel   Diesel
          Engines 	  .......   2-15
     2.0.4 Alternative Fuel  Technologies  ........   2-15
          2.0.4.1 Methanol	   2-16
          2.0.4.2 Natural Gas	 .   2-16
          .2.0.-4.3 Liquefied  Petroleum Gas ........   2-17
          2.0.4.4 Electric Vehicles  	  . .   2-18
     2 .5 Summary	2-18

3.  Environmental Benefits Chapter  .  . .  .... .  . 3 - i
     3.0.1 Introduction	3-1
     3.0.2 Calculation Method  ............... 3 - 1
     3.0.3 Diacusaion of Data	.  . 3 - 1
          3.0.3.1 Light and  Medium-Heavy Duty Fleet Vehicle
               Demographics  .	  . .  .	3-1
          3.0.3.2 Emission Factor Calculations   	 3-7
               3.0.3.2.1 1998  and  Later Baseline Emission
               Factors	'.	 3 - 7
               3.0.3.2.2 Projections   for   1998   Clean-Fuel
               Fleet Emission  Factora  .— .-•-?	3-10

     3.0.4 Environmental Impacts	3-10

4.  Costs  and Cost  Effectiveness    .	.  .4-1
     4.0.1 Introduction	4-1
     4.0.2 Costs	.	4-1
          4.0.2.1 Operating  Costs  .... 	  ...4-1
          4.0.2.2 Engine Costs  	 	 4-2
               4.0.2.2.1 Otto-Cycle  Heavy-Duty Engines  .  . 4-2
                    4.0.2.2.1.1  Hardware Costs   .  ... .  .4-2
                    4.0.2.2.1.2  Development  Costs  ..... 4 - 3
                    4.0.2.2.1.3  Certification Costs  .  . .". 4 - 3
               4.0.2.2.2 Diesel-Cycle Heavy-Duty Engines  .4-3
                    4.0.2.2.2.1  Hardware Costs   .  .  .  . . .4-3
                    4.0.2.2.2.2  Development  Costs  .  .  . . .4-4
                    4.0.2.2.2.3  Certification Costs  .  . . . 4 - 4
          4.0.2.3 Aggregate  Costs  	 4-4
               4.0.2.3.1 Manufacturer Costs  ...  	 4-5
               4.0.2.3.2 Costs to Users	  . ..4-8
     4.0.3 Cost Effectiveness  .......  	  4-10
     4.0.4 Summary	  4-11

-------

-------
                        List of  Tables


Table 3-1  Heavy-Duty Clean-Fuel Fleet Vehicle Population .3-3
Table  3-2   Vehicle  Miles  Travelled by  Clean-Fuel  Fleet
     LHDVs/MHDVs - .  .	•  •  •  -  •  •  .3-4
Table 3-3   Sales-Weighted 1991 Light  and Medium  Heavy-Duty
     Certification Values .	3-8

Table 3-4  1998 Model Year Baseline Emission Factors  .  .  .3-9
Table 3-5  1998 Heavy-Duty Clean-Fuel Fleet  Vehicle Emission
     Factors  ........ 	  ......   3-10
Table 3-6  Nationwide Emissions  Inventories of Fleets Covered
     by the Clean-Fuel Fleet Program  .	3-12

Table 4-1  Manufacturer Fixed Costs for Heavy-Duty Clean-Fuel
     Engines	  4 - 6
Table 4-2  Costs to Manufacturers . . . .  .  .-.  .  .  .  ... .  4-7
Table 4-3  Costs to Consumers	4-9
Table 4-4  Cost Effectiveness in $/ton (1998 Present Value)4  - 11

-------

-------
                      List  of Figures
Figure 3-1 - Clean-Fuel Fleet Vehicle VMT
3 - 5

-------

-------
                         1.0 Introduction


     This document is intended to provide technical, environmental
and economic analyses of the heavy-duty portion of the Clean-Fuel
Fleet program.  The heavy-duty portion of the fleet program applies
to only light heavy- and medium heavy-duty vehicles  and the engines
designated for use in these vehicles.  EPA is adopting a heavy-duty
clean-fuel  fleet  vehicle  standard of  3.8 g/Bhp-hr  non-methane
hydrocarbon (NMHC) + oxides of nitrogen (NOx).   Credit generating
standards   for  the   fleet  program  have  also   been  adopted.
Technological discussions of NMHC  and NOx  formation and control,
calculations of environmental benefits and an assessment of costs
and cost effectiveness are contained in separate chapters.

     Chapter 2 contains an assessment of technologies available for
reducing NMHC and NOx emissions in conventionally-fueled heavy-duty
vehicles  as well  as technologies  capable  of meeting  the credit
generating  standards.   First the  formations  of NMHC and NOx in
gasoline  engines  and the  impacts  of gasoline  engine  design are
discussed.  Next,  emissions formation and control in diesel engines
are  covered.   Finally,  alternative fuel  technologies  and their
emissions capabilities are presented,.	„___

     Chapter 3 contains a discussion of the environmental benefits
expected from this program.  The chapter begins with a presentation
of the vehicle demographics used  in the  calculations.   This is
followed with a discussion of how new emissions  factors applicable
to the vehicles covered by these standards were  derived.  Finally,
the emissions benefits are presented and discussed.

     Chapter 4 begins with an estimate of the costs  associated with
meeting these standards.  Expected costs for research,, development
and  testing  are  combined  with   certification   costs   and the
anticipated incremental hardware and operating costs to derive both
manufacturer  and  consumer  costs  associated  with  this  program.
These  costs are then combined with the emissions  estimates  from
chapter 3 in a calculation of the 22-year cost effectiveness of the
heavy-duty  portion of the Clean-Fuel Fleet program.

-------

-------
                           Chapter  2
                    Technology Assessment

    2.0.1 Introduction

     Internal combustion engines produce hydrocarbon1 and oxides of
nitrogen  (NOx)  emissions  through a series  of complex  related
processes which also  control the production  of other pollutants
such  as CO  and particulates  as well  as the  power  output  and
efficiency of the  engines.   Hydrocarbon (HC)  emissions generally
result  from the  incomplete combustion  of the  fuel, while  NOx
results  from the  reaction  of oxygen and nitrogen present  in  the
combustion air.  The  reactions which produce  NOx occur much more
rapidly under conditions of high temperature and pressure.

     Engine design and operations which  affect emissions of HC and
NOx will be discussed  for both  heavy-duty  diesel and  gasoline
engines  as  well  as  for  alternative fuel  technologies.    Other
emission control technologies such as  catalytic aftertreatment will
also be  discussed.  The  purposes  of  these  discussions are  to: 1)
provide background  information regarding the  fundamentals  of the
technologies, 2) describe the technologies which are being commonly
used today,  3)  describe innovative technologies which may be used
by clean-fuel fleet vehicles, and 4)  identify the approaches that
are most  likely to be used by manufacturers to comply with  the
clean-fuel  fleet  vehicle  standards.   The  discussions in  this
chapter, will be qualitative in  nature and generally do not include
predictions of emissions reductions likely  to  result from use of a
particular technology.

    2.0.2 Gasoline Engines

       2.0.2.1  Fundamentals Of Gasoline Engines Which Impact NMHC
             + NOx emissions

     NMHC and NOx  emissions from gasoline engines  (i.e.,  spark-
ignited, otto cycle) are impacted by  the following characteristics
of  the  engine  and its  operation:  air-to-fuel ratio,  ignition
timing, and combustion chamber design.  Adjustments or variations
to these  parameters can substantially affect the  performance as
well as the emissions of gasoline engines.

         2.0.2.1.1  Air-to-Fuel Ratio
     1    For  simplicity,  the  term hydrocarbon   (HC)  is  used
throughout this chapter to  describe  general effects in emissions
control.   The term  non-methane hydrocarbon  (NMHC)  is  used for
clarity  when  discussing the  emissions  standards  or  specific
emission levels.   The discussions using the  term  HC,  therefore,
should be considered to be equally applicable to NMHC emissions.

                              2-1

-------

-------
     Emissions from gasoline engines are extremely sensitive to the
ratio  of fuel  and air  in the  combustion  system.    A critical
parameter then in the performance and emissions characteristics of
gasoline engines  is the excess  air  factor designated  as  X.   An
excess  air  factor  of  1.0  indicates the  engine  is  operating at
stoichiometric conditions while X values of more than 1.0 indicate
there  is excess  air  in the  combustion   chamber.   The  current
strategy,  especially with  the  presence  of  three-way catalytic
converters, is to attempt to keep A, at 1.0.

     The  air-to-fuel  ratio,   since  it  affects  nearly all  the
operational parameters of the engine, is a primary design element
in the operation of a. gasoline-fueled engine.  Proper air-to-fuel
ratio  control is  critical in  controlling  both  the  engine  out
emissions of hydrocarbons (HC)  and NOx and, as will be discussed in
a  later  section,  in  the performance  of   catalytic  converter
aftertreatment devices.   In principle, as X is increased (i.e.,
leaner  air-to-fuel  ratio),  HC  emissions  decrease due   to  the
increasing availability  of oxygen for  more  complete combustion.
NOx emissions, on the other hand, show  slight increases  in engine-
out emissions levels at X values  slightly  above  1.0 and  decreasing
levels at higher values of X.         ..'.__.__

     NOx values increase at X values  slightly greater than  1.0 due
to the increased  availability  of oxygen and nitrogen to react to
form NOx.   As X increases, the additional  air  has  a competing
effect on NOx; by adding mass to  absorb the heat of combustion, it
lowers  the  peak  temperatures  and  pressures,  which  leads to  a
decrease in- the rate of formation of NOx.   Also  as X increases, HC
emissions begin to increase slowly at first and then  rapidly as the
lean misfire point  is  approached.    Within  reasonable   limits,
however, the effects of air-to-fuel ratio  on engine-out emissions
of HC are much less than the effects on engine-out NOx  emissions.

     Engines  using lean air-to-fuel  ratios also  tend to  show
greater  thermal  efficiencies due to  factors such  as  lower heat
losses and the ability to use higher compression ratios  (lean fuel
mixtures are less susceptible to autoignition).   At higher values
of X, however,  the power  output per volume can drop off  despite the
increase in thermal efficiency  due to the lower fuel  content.  This
effect can b« off set by either using lean burn only at partial load
conditions oar by turbocharging the intake  mixture  to pack more air
and the 3am*  quantity of  fuel into the cylinder at  a higher  initial
pressure.

         2.0.2.1.2  Ignition Timing

     Control of the timing of  the ignition spark is important in
controlling both NOx and HC emissions as well as in maximizing the
work output  of  the engine.   Delaying the  ignition  as much as
possible reduces  the  amount of NOx produced by reducing both the
      .'-^V.    -'         -
                              2-2

-------

-------
peak temperatures  and pressures of the  combustion cyel|i£ and the
length of time that the gases are exposed to high temperatures and
pressures.   Retarding the ignition timing can also increase the
work output of the engine by reducing the amount of extsrlt work the
piston must  do during the final .stages  of  compression, after the
combustion has already begun.  If the ignition timing isj; retarded
too much, however,  the combustion process begins late, resulting in
somewhat incomplete combustion.   Under these conditions/, the power
output drops and under conditions of greater timing retard the HC
emissions can go up.   Ideally, to minimize the  NOx and HC  emissions
and to  maximize the  fuel  economy of the engine,  the combustion
process needs  to be  optimized so that the ignition timing can be
delayed to as  close  to  the point in time when the piston reaches
top dead center and to have as much of the combustion as possible
take place just after the piston reaches top dead center.

         2.0.2.1.3 Combustion Chamber Design

     Combustion chamber parameters can affect the peak temperatures
and duration of the combustion process, which in turn impact on the
level of NOx  emissions  and  the  amount  of work  produced.   The
primary goals in the  design and operation of the combustion chamber
are  to  maximize  the  amount of  work .arid, power  output by the
combustion process  (which  also  implies  maximum combustion of the
HC) while minimizing the production of NOx,  the amount of wear on
the engine and the likelihood of autoignition  (knocking).

     Reducing the combustion time is an important goal in chamber
design.   Shorter combustion time allows the timing to be  retarded,
which thereby  reduces the formation  of  NOx and HC,  while still
allowing the  combustion to proceed to completion  earlier in the
expansion  stroke  (thereby  capturing  more  of  the  energy  of
combustion).  Decreasing the combustion time can be  accomplished by
making design changes to the combustion chamber which minimize the
distance the flame  front  needs to travel and/or increases  the flame
speed.   Flame speed can be increased by increasing the turbulence
in the chamber. In addition to decreasing emissions and increasing
the percentage  of  the fuel burned near  the optimal point in the
process, "fast-burn"  techniques also reduce the tendency of the
engine to knock by reducing  the amount  of time available for the
unburned fuel to auto-ignite.  Therefore, compression ratios can be
increased which  in turn further  increases the efficiency of the
engine.

      2.0.2.2  Current Gasoline Emissions Control Technology

     All heavy-duty gasoline engines certified for  model  year 1992
were equipped with exhaust gas recirculation  (EGR),  and almost all
were equipped with some form of  catalytic converter arid electronic
engine  controls.   Most  were  equipped  with  three-way   catalytic
converters  (converters  which  are capable  of both reducing NOx
emissions and oxidizing HC.and CO emissions).  These technologies

              .-               2-3         '.'•.'•'. :>

-------

-------
are described below.   It is also worth noting  that  some current
engines  are  already very  close to  complying with  the  NMHC+NOx
emission standard of 3.8 g/Bhp-hr for heavy-duty clean-fuel fleet
vehicles.  Three of the  gasoline-fueled engines families certified
for the  1992 model year had HC+NOx emissions of  4.0 g/Bhp-hr or
less, including one with HC+NOx emissions of 3.6 g/Bhp-hr.

         2.0.2.2.1  Exhaust Gas Recirculation

     From a NOx  control standpoint,  it  is  desirable  to have some
inert  gases  in  the cylinder  to take  up  some of  the  heat  of
combustion.  The recirculation of exhaust gases can provide these
inert gases.  Recirculating exhaust gases is a more effective means
of NOx emissions control than  using  additional  air to absorb the
heat of  combustion  (leaning  out the  air-to-fuel ratio) since the
water and CO2 in the exhaust  gases have high heat capacities which
make the exhaust gases  more effective at absorbing the  excess heat
of combustion.   Thus by using EGR," the peak temperatures can be
reduced further before  the  volume of inert gases  reach the point of
interfering with the combustion process.  Furthermore, recirculated
exhaust  gases  do not  add  the  excess oxygen.  Modest levels of
excess air can  lead to  increases  in  the engine-out levels of NOx
emissions.                         	_____—

     In  some instances, exhaust gas recirculation can lead to  a
reduction in engine-out HC emissions.   However, EGR normally has
very little effect  on  HC emissions.   At high levels of EGR  (i.e.
>25%), the combustion process can become unstable, just as in the
case  of  too much  excess  air.   Under  these conditions,  the HC
emissions begin to  rise sharply.

         2.0.2.2.2  Aftertreatment Systems

               2.0.2.2.2.1  Catalytic Converters

     Catalytic  converters  are  an  important  means  of  further
reducing the  emissions  of gasoline  engines.   As already noted,
heavy-duty  gasoline. engine systems  are  usually  equipped with
catalytic converters which remove HC, CO and NOx  from  the exhaust
stream.  Others use catalysts that are designed only to oxidize the
HC and CO.

     In  three-way  or oxidation/reduction catalysts,  the  HC and CO
are  oxidized to either  less complex  intermediate  products or COa
and  water vapor,  while NOx  is reduced to N2  and  oxygen.   The
reduction of NOx is accomplished by simultaneously  utilizing the
oxygen from the NOx molecules to oxidize the remaining HC and CO in
the exhaust.  This  results in  partial oxidation of the HC and CO.
Further  oxidation may  be  achieved  using  an oxidation  catalyst.
However, if there  is excess oxygen in the exhaust stream,  the HC
will  preferentially react with the free  oxygen rather  than the
oxygen  contained  in the  NOx  molecules thereby  leaving the NOx

                               2-4

-------

-------
 molecules relatively unaffected.  On the other hand,  if there  •
 not sufficient oxygen available  in the exhaust stream, the  HC will
 not be oxidized or will be only partially oxidized due to the lack
 of  oxygen.    Therefore,  the performance  of three-way catalytic
 converters is very  sensitive  to the  air-to-fuel  ratio  in the
 engine.                                                        e

      This problem can be partially overcome by using a catalytic
 converter that consists of two separate beds (a reduction catalyst
 followed by an oxidation catalyst) and injecting extra air  between
 the two beds.  This allows the air-to-fuel ratio to be run slightly
 on the rich side to ensure that good conversion of the NOx  will  be
 achieved in the first bed.  However, this technique  results in a
 decrease in fuel economy because of the slightly rich calibration.
 Furthermore,  the injection of too much air into the second  bed has
 the potential  for cooling the  gases down to  the point that the
 catalytic converter  can lose some of  its effectiveness.

      Each  catalytic   converter  design  has  its  advantages and
 disadvantages.  The  best  approach  for a  given engine design  is
 determined based upon the  composition and  level of the engine-out
 emissions.

                2.0.2.2.2.2   Electronic Controls

      Recent advances  in electronic controls have  made it possible
 to  achieve very tight control over engine parameters  such as the
 air-to-fuel ratio and ignition (spark) timing.  This improvement  in
 the  controllability  of  the  combustion  process  minimizes the
 excursions  of these engines in non-optimized operating ranges  which
 result  in both increased emissions and decreased  performance.

      The most significant  advantage to using electronic controls  is
 that  it provides a means  by which the air-to-fuel ratio  can  be
 adjusted to ensure that the concentration of oxygen in the exhaust
 is  at a level which leads  to  the  optimum efficiency"of a three-way
 catalyst  under  a  broad  range   of  operating conditions.     By
 installing  oxygen sensors in the  exhaust  stream near the catalytic
 converter,  the system can diagnose whether or not the  proper air-
 to-fuel ratio  is being maintained.  Using these feedback signals,
 the engine  computer can then adjust the air-to-fuel ratio rapidly
 to  compensate  if the  exhaust stream does not contain  the optimal
 level of oxygen.

       2.0.2.3 Future Gasoline Emissions Control  Technology

     Due  to   the   1998  4.0  g/Bhp-hr  NOx  standard,   general
 improvements can be expected  to current gasoline emissions control
technology such as  exhaust gas recirculation, catalytic converters
 (likely catalyst improvements), and especially electronic controls.
Additional  emission  control  of  gasoline  engines  may occur  with
 other emission control technologies that  are  under development:

                              2-5

-------

-------
 electrically-heated catalytic  converters, close-coupled catalytic
 converters,  and lean-burn  calibration.

          2.0.2.3.1  Electrically-Heated Catalytic Converters

      Catalytic  converters  require fairly  high temperatures to be
 effective.   Heat is provided by the exhaust  and from the reactions
 of the HC in the catalyst bed.   Following  a  cold start, there  is a
 delay before the catalytic converter becomes effective,  while it
 heats up to  its operating  temperature; once the  converter reaches
 the desired  operating temperature, it performs well  with very  high
 efficiency.  Electrically  heating catalytic converters at vehicle
 start-up  can   reduce  the  delay  before  the  converters  become
 effective.   This approach may also  be of value to the emissions
 control strategy for light  and medium duty vehicles.  Electrically-
 heated  catalytic  converters  are  made   of  metals  rather   than
 ceramics, thus,  these converters  experience less cracking and are
 more  durable than other  catalytic  converters. '   However,   some
 electrically-heated  catalytic  converters   are  susceptible  to
 aluminum  washcoat  degradation.    This  technology is  currently
 available.   Other  catalyst  bed  improvements  are  also available
 which could  improve emission reductions efficiency.

         2.0.2.3.2 Close-Coupled Catalytic Converters

      Another approach that can be used to  minimize the delay is to
 install the converter very  close to the engine, so that the exhaust
 gases contacting the catalyst  are hotter.  However, these close^
 coupled  catalytic  converters  are more  susceptible  to  thermal
 degradation  because they are continually exposed to significantly
 higher temperatures than converters  located further away from the
 engine.   By  installing converters  only  slightly  closer  to  the
 engine than current converters  so that the  exhaust gases contacting
 the catalyst are only  at  a  slightly higher temperature,  smaller
 emission benefits may occur.  Closed-coupled catalytic converters
 will be available for the 1994 model year due to the  California LEV
 standards.   Insulating exhaust pipes may provide similar benefits
 at lower costs.                                      *

         2.0.2.3.3  Lean-Burn Calibration

     While the air-to-fuel ratios of current engines are generally
 calibrated at stoichiometry, there would be potentially significant
benefits with leaner calibrations.  Leaner air-to-fuel  ratios could
 lead to lower engine-out HC, CO, and perhaps  NOx emissions, as  well
 as a decrease in fuel consumption.  On the other hand,  there can be
 a loss of power and problems with  ignition and flame propagation.
Also,  leaner combustion would  lead to leaner exhaust  which would
present the  additional  challenges  of  developing  catalysts  and
oxygen sensors that work well under such conditions.   It is likely
that  lean-burn  calibration would only be  used on engines  with
advanced electronic feedback controls.  The availability  of this

                              2-6

-------

-------
technology is uncertain at this time, for heavy duty engines.

       2.0.2.4  Expected Approaches for Clean-Fuel Gasoline Engines

     Overall,  given that emissions  from some  current gasoline-
fueled engines are already very close to those required by the CFFV
(clean-fuel  fleet   vehicle)   standard,  it   is   expected  that
optimization of existing technologies will be adequate to allow a
significant number  of  gasoline-fueled engines to  comply with the
clean-fuel fleet  standard, and that  dramatic changes will not be
necessary.  Engine changes, such as but not limited to changes in
the EGR system, combustion chamber design improvements,  and tighter
control over the  air-to-fuel ratio, are expected to  provide much of
the necessary reductions  in NMHC+NOx  emissions for many engines.
Slightly  leaner  calibrations  may be used under some conditions;
however, very lean calibrations should not be necessary.

     Additional  reductions  will also  likely come  from upgraded
exhaust aftertreatment systems.  It is expected that all gasoline
engines  in  the Clean  Fuel  Fleet program  will be  equipped with
three-way catalysts; and that these catalysts will be slightly more
effective  than   those currently  being  used,  through  either
optimization of the catalytic  materials,.__ increases  in the catalyst
loading and/or bed size or exhaust pipe insulation.   The Agency
does not currently expect that electrically-heated or close-coupled
catalysts will  be  necessary  in  order for  most  gasoline-fueled
engines to comply with the HD (heavy-duty)  CFFV standards.

     The  docket  contains  further  supporting material  on  the
feasibility of the CFFV NMHC + NOx emission standard for heavy-duty
engines.1                                               ' "

    2.0.3 Diesel Engines

       2.0.3.1  Technical Background/Fundamentals of Diesel Engines
Which Impact NMHC + NOx Emissions

     In  diesel  engines,  NOx  is  formed in  the early  phases of
combustion where temperatures and  pressures  reach a peak.   In
diesel engines, as fuel is injected into the  combustion chamber it
mixes with hot compressed air already present, and after a brief
period known- as ignition delay, this  fuel-air mixture ignites.  In
this premixAd;  burning phase/  the fuel-air  mixture burns  in an
uncontrolled manner, which causes a rapid rise in cylinder pressure
and heat  release, until  the  mixing  controlled combustion phase
(diffusion-controlled  burning)  takes  control of  the combustion
process.     Once   diffusion-controlled  burning   begins,   fuel
essentially burns as  it  is  injected,  allowing partially burned
droplets  and  particulates  to  be  consumed  as   oxygen  becomes
available at local combustion sites.   It is believed that most of
the NOx is formed during the period before diffusion takes control
of combustion.2  Therefore, anything  which  can be done to reduce

                              2-7

-------

-------
the temperatures during this uncontrolled burning phase will tend
to reduce the formation of NOx.

     Hydrocarbons are the  result  of incomplete combustion of the
fuel.  Because of the lean combustion technology used in diesels,
out  hydrocarbon  emissions  tend to  be inherently low.   However,
controls  that reduce  NOx often  have  an adverse  effect  on  HC
emissions since they are inversely related to oxygen availability
and peak conbustion temperature.

         2.0.3.1.1  Fuel System

     The.  fuel  system in  a  diesel . engine   is  responsible  for
controlling both the amount of fuel delivered to the cylinder and
the timing of the fuel delivery.  Since the fuel is injected into
an air mass that is  already  compressed and has consequently been
heated to a temperature above the auto-ignition point for the fuel,
the  injection timing determines  when combustion will  begin and
serves some of  the  same functions  in  diesel  engines  as ignition
(spark) timing does  in gasoline engines.   There  are two types of
injection systems available for diesel engines: direct and indirect
injection.

     Diesel  engines  can   use  either  direct  injection   (DI)  or
indirect  injection   (IDI)   combustion systems.   Direct injection
engines inject the fuel into a  hollow in the piston and have the
air/fuel mixing controlled by  swirling motions in the intake air
and  the  momentum  and  spray characteristics of  the fuel jet.
Indirect injection engines, on the other hand, inject the fuel into
a pre-chamber and accomplish most of the air/fuel mixture through
turbulence created  by the expansion out of   the  chamber.   Since
indirect injection engines do not expose the bulk of the initial
uncontrolled burning to as much oxygen,  they have somewhat lower
NOx emissions rates.3

       The primary disadvantage of engines using  IDI technology is
that they are less fuel efficient than DI engines due to heat and
frictional losses in the pre-chamber.  However, indirect injection
technology has  low  initial costs which make  it  well  suited for
small high-speed diesel engines where the fuel consumption  is least
significant.4    In addition,  IDI  is a  useful option  for small
diesels because it helps address problems with air utilization due
to their small cylinder volume.3  Concerns about the fuel economy
penalties,  however,  are  leading   some  manufacturers  to  use
alternative methods of controlling NOx emissions.

         2.0.3.1.2  Air System

     Air  system improvements also  show  promise for NOx  and HC
control.    By increasing  the  mass  of  gases  contained  in  the
cylinder, the temperature'  rise  can  be decreased which results in
reduced NOx formation.   If some of  the  extra gas is oxygen, the

                              2-8

-------

-------
increased  oxygen  availability  will  result  in  more  complete
combustion reducing HC and particulate emissions.  Cooling of the
intake  air  will  further reduce  the  in-cylinder  temperatures
resulting in reductions in NOx formation.

       2.0.3.2  Current Diesel Emissions Control Technology

     Current emissions  control for  diesel engines  is generally
based  upon optimization  of  engine  parameters  such  as  ignition
timing,  injector nozzle  and  cylinder  design,  and  air  intake.
Nearly  all of the  light and  medium heavy-duty  engine  families
certified  in 1992  used turbochargers with aftercooling.   A very
small number of current light and medium heavy-duty engine families
use electronic controls, but most use mechanical controls.  The use
of mechanical controls is probably  the most significant limitation
of  current  diesel  emission  control.    While   some  mechanical
improvements  are still  being made,  it is  generally true that
emissions from mechanically controlled engines have been optimized
nearly  as  much  as  is  possible  without  the  development  of
dramatically new technologies, or reductions  in  fuel  economy.  For
example, NOx  emissions could  be reduced  from  current levels by
retarding the injection timing, but this would lead to reduced fuel
economy and increased HC and PM emissions .(which would increase the
need for exhaust aftertreatment).  The following  sections discuss
some of the currently technologies which affect diesel emissions.

           2.0.3.2.1  Retarded Injection Timing

     Retarding the injection timing is  a well-known and proven
technique  for  significantly  reducing NOx  emissions.   The primary
mechanism  for  reducing  NOx  emissions by  this  technique  is the
reduction  in  the  duration  of uncontrolled  burning.   However,
retarded timing  has undesirable effects,  particularly when used
alone.     Retarding  the  timing   significantly  will  increase
particulate and hydrocarbon emissions and decrease fuel economy due
to reduced time for the fuel and initial combustion products to mix
with the excess air in order  to burn.5  The  regulatory pressures on
particulate  emissions and the  market pressures  on  fuel economy
combine  to limit the  use of retarded timing  in controlling NOx
emissions.

           2.0.3.2.2   Injection Pressure

     Injection pressure  increases in  DI  engines can be used to
limit  particulate  emissions  and  speed  up  the completion of
injection.   Increased injection pressure  increases the  air/fuel
mixing in  direct  injection  engines  because  of  increased air
entrainment into the fuel spray and higher turbulence therein (less
condensation is likely to occur).   This has little direct effect on
NOx formation; however,  it does lead to a  decrease in particulate
formation.    Furthermore, an injection  pressure  increase  will
shorten the duration of the injection.6  The combination of reduced

                               2-9

-------

-------
particulate formation and shortened injection duration should allow
the  timing  to  be  retarded  without  incurring  other  impacts.
Hydrocarbon  emissions are  insensitive to  injection pressure at
full-load, and only increase slightly under half-load high speed
conditions.7

          2.0.3.2.3   Injector Nozzle  Holes  and Diameters

     Nozzle characteristics are important to optimize for emissions
and performance reasons,  particularly  for direct injection engines.
Decreasing  the  nozzle'" hole diameters  increases  the injection
duration  without increasing  the ignition  delay;  therefore,  the
maximum heat release in the pre-mixed burning phase decreases,  in
addition, reducing the hole size better atomizes  the fuel allowing
more complete burning.   For these reasons,  smoke can be decreased
without  an  increase  in  NOx by  decreasing  the nozzle diameter.2
Changing the number of holes  in the injection nozzle may be  used to
affect  the  spray  characteristics  of the  fuel.    Hydrocarbon
emissions  are only  slightly affected by  the  number of  holes;
however,  CO  and smoke have  been found to  be least sensitive to
injection pressure for a 6-hole  injector.8

          2.0.3.2.4   Intake Air  Turbochacging

     Turbochargers are used to force additional air (mass) into the
combustion chamber allowing  a  shorter ignition delay  and more
complete burning of a given  amount of injected fuel.  Therefore,
turbocharging may be used to reduce HC emissions at  partial loads
which increases  the  capability of retarding timing  (lower NOx).8
As with other lean burn  systems  the addition of additional air to
make  the  mixture  even  leaner  results  in  reduced  cylinder
temperatures.   This change has  a direct  positive  impact  on NOx
emissions.

          2.0.3.2.5  Aftercooling                        •

     If the intake air is turbocharged, the temperature will rise
during compression.   Cooling  the intake air after it  has been
turbocharged will  help  lower  the  in cylinder  temperature,  and
therefore reduce NOx  formation.5  There are several  approaches to
aftercooling, including  air-air  and air-liquid aftercoolers.   By
using air-to-air aftercoolers, some manufacturers have been  able to
reduce the temperature of the turbocharged intake air to about ten
or fifteen degrees Fahrenheit above the ambient air  temperature.

          2.0.3.2.6  Cylinder Design

     The combustion  chamber shape may be optimized in  order to
promote mixing with enhanced turbulence.  Enhanced t'urbulence can
reduce particulate and hydrocarbon emissions through better mixing
but has a tendency to  increase NOx emissions through  the increased
heat release.  A reentrant chamber has a small lip around the top

                             2-10

-------

-------
^•'S-'-j.'- ' ^- ^'•''•''^•::':''/L-^^^':-^L-'
 I'.'.;',1-.' -^-: ^;^7-f'^-^-J^iV^-'7^
                      ";"'"'4^-->V ^-£~*^':"-^£~.-.7-- -s-ra*^:.'' '*-£*:.«>_£*•
                      ' ^ ;VJS~.:V T.-!iJ7—•V-^;' "*V-' - ' .. '"".^-^irS:".-e7''L,"vT?

                                                   ""''
                            . *#W Piston, - This combustioii
                          particulate -emissions  arid?
redtacedy
                        pa*feLcul§ite ^ emissions  to
                        ftg caiv b^-retarded,tb reduce           oa
                    ^
                         leyeJr of > psrticulate  emissions^ or • even;
                                                   '-   • -          ' '   •'
^"."•"^.^V^r".!?"
                   degree
                   -*^^ ^
                   -cli* ®v^
                                  ^
                                  miniinizea the  f orntation ofc pollutants
                                       ,th«-alr and fuel in th^ combustioA
                                          depends oiv bbtJx air:"pafeterns and
                                          Gult: toJ incorporate ^degree of
                                       mixing at all  operating conditions.
                                     ; in ^ the combustion chamber may be used
                                                     NOx^ smoke^ atid  fuel
                                              ts on  engines equipped with a
                                                        ^shpwn   simultaneous
                  ?-n; bQth, partioulate. emissions  and NOx J emissions  by
     allpwing; -the^ engine^Hto ^be^  optimized  for  NQx  or  partlculate
     enpss^ns reductions {at^bperatino- ^nditiojis unde:^ which formation
                                             '                       '        '
                                    BrR        -.Control Technoloqv ™
                                    ^-=:-»-''''1  "'^"lAT.".'-' •' ...-•:'. '- -.-.-''•" "... . •• -,- .' '„ •' ,—• ' ' •' • ^•**-ii/" •••
            r optimization of ;en5|i|ie?parameters such ^as in jection timing,
                 pressure  -d^nfacor^ nozzle  and  cylinder,  design?   air
                                   J^Sitipnal diesel  emission^ control  may
                                      'jj^8*±aa control, technologies; among
                                                      &l6.= --:!*'.H1"'s'^~v':---?>^VTrv:-'-.~:"' ;-.-;:v-:'--.*1?S>:'.--^^t*-f:JK.7.f^5y

-------

-------
    -  Exhaust _gas  recirculation  reduces  the  temperature of- the
 cylinder by,  adding  inert gases to expand  to  do work.   This can
 increase Pit emissions,  especially at full load, due to the lack of
 oxygen  preaentr  in  the  combustion  chamber   which  results  in
 incomplete combustion.   Moderate  EGR rates  (5-15%) have been shown
 to bring NOx down into the 1.9 -  3.4 g/Bhp-hr  range.  This may be -
 done without  increasing particulates significantly or de-rating the
 engine,- _by using  electronically-controlled EGR (see  discussion
 below}.    In  addition NOx reduction engines using EGR have shown
 significantly better fuel economy than engines using timing retard..
 More research must still be done to demonstrate the full potential
 of EGR.9           •

,  >     -   2.0.3^3.2  Improved Turbocharging

     " Two improvements to intake air turbocharging which are being
 developed, may  result  in significant  emissions reductions.   The
 first is the variable geometry turbocharger (VGT) , which allows a
 more optimized flow of air  into  the engine, such as  more air at
 high loads.  VGT would not directly impact NOx emissions, but would
 decrease  particulate   emissions  and  fuel  consumption,  thereby
 allowing further retard of the injection timing.  It  also provides
 a means of controlling  cylinder  pressureja__for  engines using EGR.
 The  second  approach - called turbo  expanding involves  over-
 compressing the intake air,  then  expanding it  to a lower pressure
 after the charge,  is cooled  by the aftercooler.  The  result is a
 reduced intake  temperature,  and thus lower NOx emissions.

        •  2.0.3.3.3  Injection Rate Shaping

      Injection  rate shaping is  a  very  promising technology for
 reducing  NOx   emissions  without   adversely  affecting  other
 performance parameters.- The basic approach is to reduce the amount
 of fuel  injected  in the early phases of  injection so  that" the
 amount of fuel which undergoes premix burning reactions is reduced.
 Preliminary testing^ and even production  system development has
 begun in- earnest over _ the? last- few' years.

    1  Results from the  Ricardtai  HDD  (heavy-duty  diesel)  engine
 research progrant  demonstrate that  rate  shaping may  be  used to
 achieve  vesj-loir NOx levels without a penalty  in fuel economy or
 particulat«|eraissions.10 By injecting the fuel with a "gradual rise
 and a shar|lpeutn  as opposed to a constant  injection,  the NOx is
 reduced be^iiuse  of a  lower  heat release  during   the  pre-mixed
 burning phase where NOx is formed.2 In addition, a clean injection
 cut-off will  reduce smoke and HC  emissions by  avoiding the poorly
 atomized end- of in jectiprw8 "

      - ' , 2 . 0~. 3^ 3 . 4-~  Electronic, Controls
          historical problem with managing the engine  operation
 parameters .is_ the inherent problem that mechanical  controls have

        "                 -   - " 2 - 12

-------

-------
only a limited ability to adjust the operating parameters to match
the optimal conditions for  any given speed and load condition.  An
engine set up to produce low emissions at one set of conditions may
have relatively  high emissions at some other sets  of  operating
conditions.   These  problems are  especially pronounced in  the
smaller diesel engines affected by new  clean-fuel  fleet vehicle
regulations.

     The  advent  of electronic engine control devices will allow
more flexible control of engine operation. Electronic controls are
expected  to be  installed  on essentially  all heavy-duty  diesel
engines  by  1994  in  order  to  both  achieve  improved  engine
performance  and  to assist  in achieving  compliance  with the 1994
emission standards.  Through the use of electronic controls and the
advent of advanced fuel delivery systems, it will become possible
to store  a map of the optimum operation parameters and then to
control the engine operation at  these  conditions over a wide range
of operating modes  thereby  reducing some of  the needs  to make
design  trade-offs  in cases  where  engine  operation  parameters
optimized  for one  operating  condition may not be appropriate for
other operating  conditions.8  Among the parameters  that could be
controlled electronically are injection timing, injection rate, and
EGR rate.                               	

     One example of  the use of electronic control would be to use
it to  significantly  retard injection timing under most operating
conditions  except high loads, where the  need for both particulate
control and high power is significant.  Similarly, EGR rates could
be reduced at high  loads  to decrease  the  impact  on particulate
emissions.   It may also be possible to use exhaust gas sensors  for
oxygen and/or NOx  in a closed-loop control  system to correct  for
in-use  deterioration.  EPA  believes that,   of  the  technological
advances  expected  in the next several years, electronic controls
have the  greatest  potential  for improving the emissions and  fuel
economy of diesel  engines.

          2.0.3.3.5   Aftertreatment Devices

     Aftertreatment  devices  have not  yet come into widespread use
on  diesel  engines.   Up to  the present time,  emissions  control
strategies for most diesel  engines  have relied on technologies
which control engine out emissions.  However, both particulate trap
oxidizer  systems and  flow through catalytic oxidizers  are  under
development and expected to be available for wide-scale  commercial
production  if  necessary  to comply  with the  0.10  g/Bhp-hr  PM
standard  for the 1994 model  year.   In addition,  there  is  active
research  ongoing in  the area of catalytic  converters  capable  of
reducing  NOx emissions  in  lean  exhausts.    These   are  discussed
below.

              2.0.3.3.5.1   Particulate Trap-Oxidizers


                              2-13

-------

-------
     A trap-oxidizer system primarily affects PM and HC emissions.
This method  of aftertreatment consists of  a  durable particulate
filter  (trap) which collects particulate emissions in the exhaust
stream.    Developers  of  these   systems  claim  that  collection
efficiencies  of 80 percent  or greater can be achieved.   Since
collection  of  the  particulates  without  a  system  for  eventual
removal would quickly plug the traps and shut down the engine, some
method of regenerating the filter by burning  off (oxidizing)  the
particulates is required.   The traps must be  regenerated before the
systems become  plugged  to the extent that  back pressures  in the
exhaust  system rise  too  much,  but  too  frequent  regenerations
increase the amount of energy which must be put  into the systems to
effect the regeneration.

     Trap-equipped urban bus engines have been certified recently,
but manufacturers remain reluctant to rely on these systems because
of their  relatively high  costs  and remaining concerns  over the
durability  of  the  systems.    However,  these  systems should  be
available for use prior to 1998,  if needed.

             2.0.3.3.5.2  Catalytic Oxidizing Converters

     Catalytic  oxidizing  converters greatly reduce  HC emissions.
Reductions of  as  much as  40  or  50 percent  of the engine  out HC
emissions may be seen  from the designs likely to be available prior
to 1998.  Since catalytic converters may see widespread use in the
control of particulate emissions,  large  reductions in HC emissions
will result as a corollary benefit.  The  catalytic converter avoids
the problem of regeneration due to its flow-through design and some
believe this  is a simpler, more cost efficient and  more durable
method  of  aftertreatment than particulate  traps.    However,  its
major drawback is sulfate production. This  problem is expected to
be managed  with  the  low  sulfur fuels being  mandated  for  sale
beginning in October of 1993.9

             2.0.3.3.5.3  Catalytic NOx Reduction

     Aftertreatment  devices  to  control  NOx  emissions  in  lean
exhausts are not available  at this time,  although  research on
promising technologies is progressing in Japan and Germany.11  The
only aftertreatment device to be  demonstrated to date involves the
injection of  ammonia  or urea into  the  exhaust stream to consume
NOx.  However,  there are no acceptable methods to ensure that there
will always be  a  supply of ammonia or urea  to keep these devices
operating.  Therefore, actual in-use emissions reductions cannot be
relied on and use of these devices could be  problematic.  There is
also concern about harmful effects  from the emissions of ammonia.

     Passive  flow-through  catalytic  converters  to  reduce  NOx
emissions in  lean-burn  exhausts  are under  investigation.   These
devices  appear to  operate  by simultaneously  reducing NOx  and
oxidizing HC or particulates with the oxygen produced from the NOx.

                              2-14

-------

-------
However, the development of useful devices is still some years away
and availability of these devices for the 1998 model year can not
be assumed.8

         2.0.3.3.6  Variable Compression Ratio

     Increasing the compression ratio in the engine reduces white
smoke and  hydrocarbon  emissions.   There  is,  however,  an optimum
point at which a further increase in the compression ratio causes
an increase in particulate emissions.12  A prototype diesel engine
has been developed that had a variable compression ratio.  NOx and
NMHC emissions  from this  engine were very  low.   This technology
thus may become  available to reduce particulates without increasing
NOx.13

     2.0.3.4  Expected Approaches for Clean-Fuel Diesel Engines

     Diesel engines will need to make changes to comply with the HD
clean-fuel  fleet vehicle  emission standard for NOx and NMHC, but
this should be feasible for a significant number of engines.  It is
expected that electronically controlled EGR will be  necessary, and
that highly optimized  electronic control of injection timing and
rate   shaping  will  also   be   incorporated.     Some  catalytic
aftertreatment may be used to control NMHC and PM emissions,  since
engine-out  emissions of NMHC and  PM could increase  as  a result of
timing  retard.   Although  the  Acurex Report1 done under  contract
with  the  California  Air Resources  Board,  entitled "Technical
Feasibility of  Reducing NOx  and Particulate Emissions  From Heavy-
Duty Engines,"  projects that catalytic  aftertreatment  will reduce
NOx emissions significantly by 2000,  based on our  above assessment
of   catalytic  aftertreatment   and  recent   reports  from  the
Manufacturers  of  Emission Controls Association  (MECA)  that  this
technology will not be available until after 1998, EPA  does not
expect  that catalytic aftertreatment will  be used to reduce NOx
emissions.   It is not clear,  at this time, how fuel economy will be
affected,    since   electronic    controls,   EGR,    and   improved
turbocharging can improve fuel economy, while timing  retard will
have  a  negative  impact.    Manufacturers  will  be  faced with a
decision of how to best trade off improvements to the engines with
increases  in fuel consumption,  in order  to control emissions in the
most cost  effective manner.

     The  docket  contains  further  supporting  material  on the
feasibility of  the CFFV NMHC +  NOx emission  standard for heavy-duty
engines.1

    2.0.4  Alternative Fuel Technologies

     While  alternative   fuel  technologies  are  certainly  viable
 candidates  for use in  clean-fuel  fleet  vehicles,  the  HD  CFFV
 emission standards are not set at a level which will require their
 use.   It will be difficult, however, for most diesel engines, and

                               2-15  .

-------

-------
possibly  some  types of  gasoline engines,  to  reach the  credit-
generating  standards.    Alternative  fuel  technologies  such  as
methanol- and  gaseous-fueled engines are expected to  meet  these
standards while  electric vehicles are  viewed  as being the  only
technology capable of meeting the zero emission vehicle standards.
The docket contains further supporting material on the feasibility
of the credit  level  NMHC + NOx emission standards for heavy-duty
engines.1


       2.0.4.1  Methanol

     Methanol is an attractive fuel from an emissions standpoint.
Its lower flame temperature leads to an inherent reduction in the
formation  of NOx  emissions.   Exhaust emissions  of  NMHC   (more
appropriately  called organics  for  methanol-fueled  engines)  are
generally comparable to those of similar petroleum-fueled engines.
Nevertheless,  methanol-fueled  engines can still  provide  some
benefit with respect to organic emissions  for two reasons.  First,
organic   emissions   from  methanol-fueled  engines,   while  not
necessarily  less  than   those  from  engines  using  conventional
petroleum fuels, tend to be less reactive in the processes which
form ozbne, and thus can  have a less  significant impact on ambient
air quality.   It should  be noted, however, that  this benefit is
difficult to  quantify.    Second,  non-exhaust  (e.g.,  evaporative)
emissions of  organics  are  much lower  than those  from gasoline-
fueled vehicles due to the lower vapor pressure of methanol.

     Both otto-cycle and diesel-cycle methanol-fueled engines and
vehicles  are  under development currently.  While  it is expected
that otto-cycle technology will be  capable of meeting the credit
standards by 1998,  these vehicles are being developed primarily for
light-duty uses.  Diesel-cycle engines, however, because of their
inherent  fuel  economy   advantages,   are  more  likely  to  see
applications  in  heavy-duty  vehicles.   Recently,  Detroit Diesel
certified the  first heavy-duty diesel  cycle methanol engine for
commercial sale.   This engine was recertified for the 1993 model
year as  having NMHC+NOx  equivalent  emissions of 1.8 g/Bhp-hr, CO
emissions of 2.1 g/Bhp-hr and particulate  emissions of  0.03 g/Bhp-
hr.  As can be seen, this engine is already in compliance with the
credit standard.   This  particular  engine  is a heavy heavy-duty
engine intended for use  in urban buses, but the technology should
be transferrable to other engines more likely  to  be used in the
Clean Fuel Fleet program.

       2.0.4.2 . Natural Gas

     Natural  gas,  in either  the form of  compressed natural gas
 (CNG) or  liquified natural gas (LNG),  is  likely to be used  as an
alternative fuel in  some  light  and medium heavy-duty vehicles for
use in the Clean Fuel Fleet program.  Indeed, experimental delivery
vans  converted from gasoline to CNG  are being used by several

                              2 - 16

-------

-------
 delivery  fleets  due  to its  potential  for fuel  cost  savin
 Moreover,  Cummins has  recently certified  a CNG-fueled L-10 2
 engine with the  State  of  California.    Two types of  CNG-fueled
 engines are  being  developed:  stoichiometric  (i.e.,   converted
 gasoline  engines)  and  lean-burn  (e.g.,   the  Cummins  L-10 bus
 engine).

      EPA  analyzed  the  emissions  characteristics  of  CNG-fueled
 vehicles  and  engines   in  an  April   1990  special   report.14
 Stoichiometric engines, which are the type of engines being used by
 delivery vans, were projected  in that report to have  emissions of
 0.88  g/Bhp-hr NMHC+NOx, 7.4  g/Bhp-hr  of  CO,  0.01  g/Bhp-hr of
 particulate  and  0.0006 g/Bhp-hr  of formaldehyde  for  optimized
 engines.   Projections for optimized  lean-burn engines,  which are
 favored for use  in  heavy  heavy-duty applications (greater than
 26,000  Ibs GVWR)  for fuel economy reasons, indicated that NMHC+NOx
 emissions  of  4.06   g/Bhp-hr,   CO   emissions  of   1.5   g/Bhp-hr,
 particulate emissions of 0.05  g/Bhp-hr and  formaldehyde emissions
 of  0.03 g/Bhp-hr can be achieved  using  traditional technology.
 However, the Cummins  L-10 engine that was  recently certified had
 emissions  that  were much  lower than those  projected for such a
 lean-burn  engine: NMHC+NOx emissions of  2.6 g/Bhp-hr,  CO emissions
 of 0.4 g/Bhp-hr,  and particulate emissions of 0.02 g/Bhp-hr.  While
 neither the projected emissions, nor  the emissions from the Cummins
 L-10  engine,  would comply  with all  the  credit  standards, it is
 expected that  the standards could  be met  by  engines using more
 advanced technology.  Although the L-10  engine is  for  heavy heavy-
 duty applications, the L-10  engine demonstrates that it ought to be
 feasible to meet the clean-fuel fleet vehicle standards  with  light
 and medium heavy-duty vehicles.  Lean-burn CNG engines will need to
 improve the control of NOx  and/or NMHC emissions in order to meet
 the  credit  standards,   but it is   likely  that  this   would  be
 accomplished by incorporating the advances in NOx control which are
 being developed to comply with the  4.0  g/Bhp-hr standard  for all
 heavy-duty engines.   Stoichiometric technology will  also  require
 minor advances, such as  slightly higher catalyst loadings, in  order
 to comply with these standards  by 1998.

       2.0.4.3  Liquefied Petroleum Gas

     Liquefied Petroleum Gas  (LPG)  is another gaseous  fuel which
 expected to be used to some extent in the Clean Fuel Fleet program.
 It is a very clean-burning  and economical fuel,  and has been the
most widely used alternative  fuel  for  many years.   Most of the
 incentive  for its use has come from its economic advantages, and
there  is  only  a  limited  amount  of emissions  data  available,
especially for heavy-duty vehicles.  In  general,  emissions from
LPG-fueled engines should be similar to  the emissions  from CNG-
 fueled engines, except for NMHC  which could be slightly higher from
LPG-fueled engines.   This is because  NMHC emissions are generally
due to unburned fuel, and LPG is largely  a  non-methane  fuel (unlike
CNG which  is  comprised mostly, of  methane).  As  with CNG-fueled

                             2-17

-------

-------
vehicles, it is expected that  LPG-fueled heavy-duty engines will
also be  able to  comply with  the  credit standards  with similar
control approaches.  EPA plans to release a report  on LPG fuels and
vehicles in the next few months.

       2.0.4.4  Electric Vehicles

     EPA  anticipates that  electric vehicle  technology will  be
required for a vehicle  to comply with  the zero standards.   While
light  and  medium heavy-duty vehicles  may be better equipped to
handle the bulk and mass of the necessary batteries, the additional
batteries,  limited  driving  range,  and operating  expenses  will
probably  still make electric  vehicle  technology  an  unpopular
choice.  Advances in technology and market incentives for credits,
however, may combine to make such technology viable at some point
in time.

    2.0.5 Summary

     The clean-fuel  fleet vehicle  standards should be achievable
using a wide range  of technologies.   Gasoline-fueled engines should
be able to reach compliance  with the standards through the further
optimization of technologies already in use.  Diesel engines will
significant developmental work, including the introduction of new
technologies such as EGR  for diesels.   EPA also  believes that
optimized electronic controls  for  fuel will be required for all
diesel  engines  in  this  program  but  much  of  this  will  be
incorporated meet the 4.0  g/Bhp-hr NOx  standard.   Optimized  (or in
some cases non-optimized)  alternative fuel technologies capable of
meeting  the  clean-fuel  fleet  vehicle  standards  are  already
available.

     Credit-generating  standards may be achievable in some cases
using  conventional  fuel  technologies  and  will   certainly  be
achievable  using  alternative  fuels.    These standards  might be
achievable with highly advanced gasoline engines using electrically
heated catalytic  converters and possibly even  by diesel engines
using optimized EGR and electronic  fuel control should they become
available.   A methanol  engine which meets all of the  credit
standards  has already  been  certified,  and natural gas-fueled
engines  appear to be promising candidates  also.    Zero-emitting
vehicles will probably  require electric vehicle technology.
                              2 .- 18

-------

-------
                            Chapter 3
                     Environmental Benefits
      3.0.1  Introduction







 vehicles .which ™1 be> aV^SJ?1  +1® been made °f the n^mber of
 of  miles  each of th.«,  -SET??.?^^?  requirements, the number
 emission
     3-0-2 Calculation Matherf
total  number of  vehicle
                                                           by the
heavy-duty (8, 501-1  500
heav^-duty (19,501-26  000
subclassed are the
                                                            light
                                                 (GVW) > "d »ed±u«
                                           subclass«* since  these


light heavy- and       '
    3-°-3  Diseuaaion of
      3.0.3.1  Light
demographics
                      and  medium   heavy-duty   fleet   vehicle
                                                             a:
                             3-1

-------

-------
subclasses  will  be discussed in  order  to provide the background
data necessary to make the emissions inventory calculations and to
interpret the results of the emissions  inventory modeling.

     EPA  has  estimated that,  in  1989,  there  were 305,000 light
heavy-duty  and 668,000 medium heavy-duty vehicles  operating in
fleets of ten or more heavyrduty vehicles  in  areas  affected by the
fleet  provisions.15   Based on  information  available  from EPA's
MOBILE4 emissions model, the light heavy-duty  fleet population is
projected to  grow at a rate  of 2.75 percent  per  year while the
population  of medium heavy-duty  fleet  vehicles is  projected to
decline at a rate of 1.11  percent  per year.16   It is also estimated
that heavy-duty  fleet vehicles are  replaced every six years and
that 80 percent of the vehicles operating  in  fleets of  ten or more
heavy-duty vehicles in affected areas will actually be covered by
the program.   (Twenty percent of such vehicles are assumed not to
be covered  because they are  not  centrally  fueled or  capable of
being centrally fueled or  are exempt under the CAA  fleet program.)
Combining these data and assumptions with  the fact  that 50 percent
of the purchases of affected heavy-duty  fleets will be  required to
be clean-fuel fleet vehicles starting in 1998, it is projected that
8,516  light heavy-duty and 11,124  medium heavy-duty clean-fuel
fleet vehicles will be required  to  be  purchased in 1998.17   The
results of this projection  of  the number  of  light heavy-, medium
heavy- and  total  heavy-duty vehicles affected by  the Clean-Fuel
Fleet Program from 1998 through 2020 are shown in Table 3-1.
                              3-2

-------

-------
       iir
                                   i— CM  CO
                                   CM CM  CM
                                  r*»  co  en  o  CM co  m
                                  PM  CM  CM  eo  eo co  co
	     35,  o>  o  o>  o> o"  o" o"  o" erf  o*
 A  i^*~r-*~-^CMCMCMCMCMeM
                                                                     CM  CM" CM"  evi  co"  co" eo"  ^
        w 7  3
       *- 2  8-
        '" T-  CM CO  3  S  CO  *  W" W" ™  ~~
                                                              of  of co  co" iC »C  
 lii       :   N. CM  r^ co  o    co ^  co_ co  co  co  cv
                                                             CM  U
                                                             oo  a
S, a_cM~-t-" o"  co" .  «  «  •*  « «  •»_ in  N._ 03  o  eo   CM  3 o  in o  in  "  K
          10 Si  2  S  S  S 8  9 ?  9 5  9" S  $  § S  5 S  S S  S  S  S
 I    !3i|*".<^.°"L|sl<®l'>tCM»^O. O>flOr*|s1r»f>1t>.CMCMCMCMCM*
 :  g i  4i{ "^  ^ 2  2 2  2 2  2 2  2 w  *  *  w" °* *  °* °*  * *  •*  °*  °°
 '  oi
   51—
         5
         JI


         f
                EsSZ***2r5'^^MCMT-co^h*oo)o>o>OfCMr>.floo>o>o«j
                       «o  co  m  eo  to  to  to to  CM CM  CM CM  eo" co"  co" CM"  CM" CM" CM  co"  c\
                    co  co  e»  o» of o" o" o" o"  »-" -p-"  —" ^  CM" CM"  CM" to"  to" eo" V
                                                                                     •»  •» in
         Ssltliiis««3So1|a3Kfe22}C-|
         in «o   S  35 o>  S 3
         CM CM  CM CM  PM CM  eo to  to eo  eo  eo  to" to"  to"  eo"  to"  •»" V  co" co"  to" V
         IIJln;u>o>
         S  iS SS  ^ 5|  S g  5 g  3  5  s S  8  3 K  ol 8  9 K  S «  S S
             111 tA 
-------

-------
                                                1
 CO


 a
 (0


 a
 Ul
 IU
 UJ

 111




 I
 O
 >•
 00

 Q
UJ
(0
UJ


5

UJ


o


UJ

• •

^
CO

Ul

00
88-551868888
         S 8 S'
8 5 8
       8 S S 18
co oo
r- CM 'it
           co co
          CM CVJ 5 g

-------

-------
Billions of Miles Travelled
h-H- ru CJ >f»- ui CTI -vi 00
tO

_ 	 .0 	 ""'
-
' *•*++.
**""•-•-•+' 	 *...,.__

	 o
Light-Heavy
• — •&••-•
Medium-Heavy
Total
	 o 	
'.:.:*& 	



,,,,--a
&e_— -B—-^8"
1 1 1 " 1
95 2000 2005 2010 2015
Year
i


2020 2025
Figure 3-1 - Clean-fuel fleet vehicle VMT
                             3-5

-------

-------
      Information about fleet operating practices and 4®f§(;-available
 from  MOBILESa were  combined to  calculate annual  ijjjjjelG miles
 traveled  (VMT)  by the clean-fuel  fleet  vehicle popjjalpion.  For
 simplicity, it was assumed that all fleet vehicles wd;||tin: a given
 class travel the same number of miles per year  regardless  of age.
 Since light heavy-  and medium heavy-duty fleet vehicles are both
 projected to accumulate their average fleet life milesf within six
 years, the annual VMT per vehicle for  each class was calculated by
 averaging the VMT per vehicle projections for the firsfc-.six years
 of each class as published in the User's  Guide  to MOBIJiB^.   It was
 further assumed that the VMT/vehicle data will remain constant from
 year to year.  Using this methodology, it is projected that light
 heavy-duty fleet  vehicles will travel 17,870  miles per year and
 medium heavy-duty fleet vehicles will travel. 36,190 mllea per year.

     The projections of the total number  of vehicle mii&es traveled
 by heavy-duty vehicles affected by the Clean-Fuel Fleefc program are
 shown in Table 3-2 and Figure 3-1.  It can be seen that;^ due to the
 simultaneous  growth in light  heavy-duty fleet vehicles  and the
 decline in the population of medium heavy-duty  fleet vehicles, the
 number of  light heavy-duty vehicles  is  projected  to: surpass the
 number of  medium heavy-duty  vehicles by  approximately  the year
 2008.  The total number of vehicles increases throughout the time
 period evaluated.  However,  due to the fact that medium heavy-duty
 fleet vehicles  accumulate approximately twice  as  many miles per
 vehicle each year as light heavy-duty  fleet vehicles  (see previous
 paragraph), medium heavy-duty  fleet  vehicles together accumulate
 more miles each year than the light heavy-duty  fleet vehicles even
 out to the year 2020.

     In order to understand  some of the complexities of the  results
 which will be shown  later,  it  is important to understand  some of
 the contrasts between light heavy- and medium heavy-duty vehicles
 and between otto-cycle (gasoline) and diesel-cycle engines. As has
 already  been  pointed out,   the  population  of light, heavy-duty
vehicles is  growing  while  the  population of  medium heavy-duty
vehicles is declining.  Furthermore,  it has also been stated that
medium heavy-duty vehicles  travel  about  twice  as  many miles in a
 year as do light heavy-duty vehicles.  Another  important contrast
between light heavy  and medium heavy-duty vehicles is the mix of
 engine types these vehicles use.  It  is estimated that 70  percent
 of the light heavy-duty vehicles use gasoline powered otto-cycle
 engines with the  remainder  being primarily diesel-cycle engines.
For medium heavy-duty  vehicles,   the engine  mix  is  nearly the
 opposite of  the  engine mix  for  light  heavy-duty  vehicles  (30
percent otto-cycle and 70 percent diesel-cycle).16

     The distinction between engine type mixes becomes particularly
 critical when the contrasts in emission factors between gasoline
 and diesel  powered vehicles  are taken into account.  Diesel  engines
 typically emit higher levels of NOx than  do gasoline engines while
 simultaneously emitting lower  levels  of HC due to the fact that

                              3 - 6             .   ' -,:—\

-------

-------
diesel engines use lean-burn modes  of  operation  and consequently
have relatively low engine-out HC emissions levels.  Today's diesel
engines are not equipped with  catalytic  converters  to  reduce  the
engine-out HC and NOx emissions as are gasoline engines^

       3.0.3.2  Emission Factor Calculations

     In  order  to  calculate the  emissions  inventories for  the
desired scenarios, emission factors and  deterioration rates  for
light heavy-duty and medium heavy-duty vehicles are needed.   In
this analysis,  all the vehicles are either  gasoline-  or diesel-
powered so emission factors for only these two fuel types will be
needed.  It is  expected that some alternative-fueled vehicles will
be produced, but that the numbers will be sufficiently small,  and
certification  emission levels and  deterioration characteristics
sufficiently similar, so as  to not substantially impact  the results
of this analysis.

         3.0.3.2.1  1998 and Later Baseline Emission Factors

     Determination of the environmental impacts of  fleet vehicles
requires estimates of emission factors  for,  in-use  vehicles.  For
this analysis,  data  is used  that projects that  in-use 1998 and
later heavy-duty vehicles  will behave similarly to certification
vehicles,  and  that  they  will meet  their  respective standards
(intermediate useful life and  full useful life standards)  in use.
In order for this approach  to  be valid, it would be necessary for
there  to  be an extensive  inspection  and maintenance program, as
well  as  an active  field  enforcement  program,  for   heavy-duty
vehicles.  Both of these seem reasonably likely,  especially in the
areas  that will be  affected by the  fleet program.   As will be
discussed  later, a similar  approach is also being used for heavy-
duty clean-fuel fleet  vehicles.   Since the same approach is being
used for both baseline and clean-fuel fleet emission  factors, these
assumptions  should not  significantly impact  the calculation of
incremental benefits even if the assumptions were to be somewhat in
error.  The 1998 baseline vehicle emission factors are derived from
1991 sales-weighted  certification emissions, taking into  account
the  effects of new  emission  standards.   The 1991  certification
emissions  rates are  shown  in Table  3-3.

     NMHC  estimates   for   pre-1998  vehicles  were   .derived  from
MOBILESa  projections.   Non-methane  hydrocarbon (NMHC)  emission
factors  were  estimated at  95 percent of the total  hydrocarbon
emission  factor for  diesel engines  and at 75 percent of the total
HC emission factor for gasoline engines.14 For gasoline engines the
same deterioration rates for HC emission factors are applied to the
NMHC  emission  factors since  present  catalytic converters  have
little effect  on methane emissions.
                               3-7

-------

-------
    Tabla 3-3 Sales-Weighted 1991 Light and Medium Heavy-Duty
                      Certification Values
Sales-Weighted Heavy-Duty Certification Values
(g/Bhp-hr)
Vehicle class
Light Heavy-Duty Diesel
Medium Heavy-Duty Diesel
Light Heavy-Duty Gasoline
Medium Heavy-Duty Gasoline
THC
0.56
0.41
0.45
0.79
NMHC
(estimate)
0.54
0.39
0.34
0.59
NOx
4.44
5.01
3.87
3.84
     The most significant standard affecting 1998 baseline heavy-
duty vehicles is a reduction in the NOx emission standard from 5.0
to 4.0 g/Bhp-hr  beginning in that model_year.   Light  and medium
heavy-duty diesel  engines currently exceed this level  and will
therefore  require  new  compliance  effort.   Although  the  sales-
weighted certification values for NOx  emissions  from 1991 heavy-
duty gasoline engines are numerically low enough on average to just
meet  the  1998  4.0 g/Bhp-hr NOx standard,  the  relatively  small
compliance margins  and the  fact  that some engine families do not
meet  the  standard indicates that further reductions are needed.
Manufacturers will likely incorporate compliance margins of 10 to
15 percent for  certification  vehicles  to allow for  production
variability and  in-use  operation.   Thus,  in-use  emissions may be
under the standard on average.  For this analysis, however, average
NOx emissions  will be modeled as being  at the  level  of the 4.0
g/Bhp-hr standard at the end of the vehicle's useful life.

     Hydrocarbon standards for 1998  heavy-duty engines  will be the
same as for current engines. Heavy-duty  diesel NMHC  emissions may
drop if catalytic converters achieve a significant penetration into
the light  heavy- and medium heavy-duty  diesel engine  market for
particulate control.   This is because,  in addition to providing
particulate. control,  these catalytic  converters  are expected to
reduce HC emissions. However, it is also  possible that engine-out
HC emissions will  be allowed to  rise somewhat in the presence of
the catalyst in order to provide  more control  flexibility for NOx.
This  would result  in little or no net  reduction of NMHC for 1998
baseline vehicles.   For this analysis, 1998 baseline diesel NMHC
will  be modeled as being at a level equivalent  to  the 1991 NMHC
levels.
                               3-8

-------

-------
     Hydrocarbon emissions from heavy-duty  gasoline  vehicles  are
also projected to comply with the applicable HC standards in use.
Light heavy-duty gasoline vehicles are assumed to comply with the
1.1 g/Bhp-hr HC standard for vehicles 14,000 Ibs GVW and under, and
medium heavy-duty gasoline vehicles are  assumed to comply with the
1.9 g/Bhp-hr HC standard for vehicles over 14,000 Ibs GVW.  Given
the Mobile model vehicle grouping  scheme laid out in the beginning
of this  chapter for light and medium heavy-duty vehicles,  these
assumptions  could  introduce  some  analytical  error  into  the
analysis,  since some light  heavy-duty  gasoline  vehicles  (those
between 14,000 and 19,000  Ibs GVW)  could emit up to the 1.9 g/Bhp-
hr standard.  However, this should not introduce significant error
into the  analysis  since there are few  gasoline vehicles in this
weight range.   For  this analysis,  as with diesels, 1998 baseline
gasoline NMHC  will  be  modeled as  equivalent to 1991 NMHC levels
since some reductions in NOx emissions will be needed in 1998.

     The  deterioration  rates for  heavy-duty  diesel  engines are
based on the assigned deterioration factors for heavy-duty diesel
engines  with  aftertreatment,  and the  deterioration rates  for
gasoline engines are based on the assigned deterioration factors
for light-duty gasoline  trucks with three-way catalysts.18  For this
analysis,   the   estimated   zero-mile	emission    factors   are
back-calculated by using the emission projections estimated above
and the  respective  deterioration  factors.  These estimated zero-
mile emission  factors  and deterioration  rates for 1998 baseline
vehicle  engines are presented in Table  3-4.   The deterioration
factors presented in Table 3-4 are the additive equivalents  of the
multiplicative deterioration factors, divided over the useful life.
       Table 3-4 1998 Model Year Baseline Emission Factors

Vehicle
L-H diesel
M-H di«s«l
L-H gasoline
M-H gasoline
Zero Mile EF
(g/Bhp-hr)
NMHC
0.42
0.30
0.20
0.35
NOx
3.33
3.33
3.33
3.33
Deterioration Rate
(g/Bhp-hr/10, OOOmi)
NMHC
0.011
0.005
0.013
0.022
NOx
0.061
0.036
0.061
0.061
                               3-9

-------

-------
         3.0.3.2.2  Projections for 1998 Clean-Fuel Fleet Emission
Factors

     Emission factors and deterioration rates for clean-fuel fleet
heavy-duty vehicles were estimated using the following methodology.
In  today's Regulatory  Support  Document the  clean-fuel  engine
emission factors are based on  the  projection that  average NMHC +
NOx emissions will comply with a  combined 3.8  g/Bhp-hr NMHC+NOx
standard at  the end of  the vehicle's useful  life instead  of a
combined  3.5  g/Bhp-hr  NMHC + NOx  standard  as proposed.   The
vehicles'  emissions  are  also assumed to deteriorate  in a manner
consistent with baseline vehicles,  and thus  the  same assigned
deterioration factors were used for the clean-fuel fleet vehicles
as for the baseline  vehicles.   However,  since  the end-of-useful-
life projections are different  between these two types of vehicles,
the multiplicative deterioration factors will result in different
additive deterioration rates.

     With a combined NMHC + NOx standard manufacturers can  get the
required reductions  from NMHC and NOx.   Each engine  family is
likely to use a  different  mix,  but in most cases reductions  in both
pollutants are  expected.   This analysis assumes that  the end of
life NMHC level for clean-fuel  engines will, be 30 percent lower in
1998 than  in  1991.   This  is  based on the NMHC reduction expected
from improved catalysts  in  gasoline and diesel applications.  To
estimate  NOx  levels  for  1998  clean-fuel  engines,   the  just-
calculated 1998 NMHC level is subtracted from the NMHC+NOx standard
of  3.8 g/Bhp-hr.    Zero-mile  emissions are  estimated by back-
calculating using the emissions projections estimated above and the
respective deterioration factors.   Table 3-5 contains the projected
zero-mile  in-use  emission  factors  and  deterioration  rates  for
heavy-duty clean-fuel fleet vehicles.
     Based  on  the baseline and clean-fuel fleet vehicle emission
factors  developed above, emission  inventory estimates have been
calculated  for  NMHC  and NOx  emissions  from  heavy-duty  fleet
vehicles.  Historically, carbon monoxide emissions and particulate
emissions from diesel engines have been directly proportional to HC
emissions and  inversely proportional to NOx emissions.  However,
the introduction  of new technologies such as catalytic converters,
rate-shaped electronically-controlled unit injectors  and  exhaust
gas recirculation will change the relationships  between pollutant
emissions  rates  in ways which can  not  be accurately  predicted.
Therefore,  projections of emission inventories will be made only
for NMHC and NOx,  the  primary pollutants targeted by the clean-fuel
fleet  requirements.
                              3-10

-------

-------
   Tabl« 3-5  1998 Heavy-Duty Clean-Fuel Fleet Vehicle Emission
                             Factors

Vehicle
L-H diesel
M-H diesel
L-H gasoline
M-H gasoline
Zero Mile EF
(g/Bhp-hr)
NMHC
0.29
0.21
0.14
0.24
NOx
2.85
2.94
2.97
2.82
Deterioration Rate
(g/Bhp-hr/10, OOOmi)
NMHC
0.008
0.003
0.009
0.016
NOx
0.052
0.032
0.054
0.051
     The  inventories  of NMHC  and NOx emissions  from heavy-duty
clean-fuel  fleet  vehicles  1998  baseline  vehicles  have  been
calculated  for  the years during which these  standards are being
phased in and for every  five years beyond ..that  until the year 2020.
In general, for each  vehicle  subclass these per-vehicle emission
benefits  were  calculated  by  subtracting  the clean-fuel  fleet
vehicle emission factors from the baseline emission factors for the
respective  pollutants  and  then  multiplying  the  result by  the
estimated vehicle miles traveled by all clean-fuel fleet vehicles
in that vehicle subclass during  each  year.   The overall emission
benefits  results  from  light  heavy-duty  and medium heavy-duty
vehicles were  then combined for  each year and are  presented in
Table 3-6 according to engine type.

     As demonstrated  in Table 3-6, the  emission  inventories and
benefits climb rapidly during the first six years as the fleet is
turned over;  after 2003,  however, the emissions  inventories and
benefits become  relatively constant.  After 2003 diesel-cycle NOx
benefits begin  to  decline as  otto-cycle  benefits  increase due to
the gradual replacement of medium heavy-duty vehicles  by light
heavy-duty vehicles which are mostly otto-cycle vehicles.
                              3-11

-------

-------
 Table 3-6 Nationwide Emissions Inventories of Fleets Covered by
                   the  Clean-rFuel Fleet Program

calendar
year
1998
1999
2000
2001
2002
2003
2004
2005
2010
2015
2020
Total
NMHC Emissions Benefit
(tons per year)
Otto-
Cycle
26
60
' 101
149
204
266
265
264
261
260
261
5,251
Diesel-
Cycle
60
124
193
266
345
428
425
422
408
397
387
8,286
Total
86
184
294
415
548
693
689
686
669
657
648
13,537
NOx Emissions Benefit
(tons per year)
Otto-
Cycle ,
130
276
437
615
808
1,018
1,017
1,017
1,020
1,031
1,051
20,748
Diesel-
Cycle
255
523
803
1,095
1,399
1,715
1,703
1,690
1,633
1,583
1,542
33,260
Total
385
798
1,240
1,710
2,207
2,733
2,720
2,707
2,652
2,614
2,593
54,008
     This  analysis generates  emission reductions  based  on  the
assumption that the end of life NMHC level for clean-fuel engines
will be 30 percent lower than the end of life NMHC level for 1998
baseline engines and the remainder of the required reduction comes
from NOx.   There are any number of potential approaches which could
be used to meet the NMHC + NOx levels of the HD CFFV standard.  For
example, another possibility is that otto-cycle engine NMHC levels
increase over 1991 levels under pressure from the 4.0 g/Bhp-hr NOx
standard and diesel-cycle NMHC emissions  decrease  as a result of
the particulate matter control technology discussed in Chapter 2.
Thus, a different  set of NMHC benefits would be expected for clean-
fuel heavy-duty engines.  Assuming the NMHC  + NOx split for 1991
engines the NMHC  benefit for  otto-cycle would  increase,  but for
                              3-12

-------

-------
diesel-cycle engines the  benefits would decrease.   NOX benefits
would essentially be the same.

     After  examining the  sensitivity  of  the  emission  benefit
results  to  the  results  generated  from this  other  reasonable
assumption of emission benefits for HD CFVs,  EPA has concluded that
the impact on cost effectiveness  is not major.  Thus, EPA will use
the emission benefit results generated from the above  scenario that
assumes NMHC levels of clean-fuel engines will be 30 percent lower
than  NMHC  levels  of  1998 conventional HDVs.   Based on  this
scenario, the  22-year total of  emission benefits  for otto-cycle
engines will be 5,300 tons of  NMHC benefits  and 20,700 tons of NOx
benefits.  The 22-year total of emission  benefits for diesel-cycle
engines will be 8,300 tons of NMHC benefits,  and 33,300 tons of NOx
benefits.   Combined benefits  range from 13,500 tons of NMHC and
54,000 tons of NOx.
                              3-13

-------

-------

                           Chapter 4
               Costs  and Cost Effectiveness
    4.0.1  Introduction

     This chapter describes EPA's analysis of the economic impact
of  the  Clean-Fuel  Fleet  program  on heavy-duty  vehicles.    The
purpose of this analysis is to estimate the per-vehicle costs,  the
total cost, and the  cost effectiveness of the program.  To do this,
it is necessary to make several assumptions about how manufacturers
will choose to comply with the program.   The Agency recognizes that
manufacturers may deviate from the control techniques assumed here,
and   that   such  deviations  could  lead   to   different  costs.
Nevertheless,  in the absence  of  better  information,  EPA believes
that  these approaches are  technically  reasonable and  that they
result  in  reasonably  accurate estimates of  the  costs associated
with this program.

      It should also be noted that this  analysis  does not assume
that  all current light-heavy and medium-heavy duty vehicle engine
families will comply with the standards of this program, but rather
that only those engine families which ceroid- comply with relatively
minor changes  would be produced  for the clean-fuel fleet vehicle
program.   This is  important because costs would be significantly
higher  if  it- were necessary that all light-heavy and medium-heavy
duty  vehicle engine families comply with the standards.  However,
this  is not the case, and given the small  size  of the affected
market, there should   be a  significant incentive to modify only
those engine families for  which costs were _ relatively small or
demand  was larger from a nationwide perspective.

    4.0.2  Coata

       4.0.2.1  Operating Costs

      Increased operating  costs  can arise  from  three  sources;
increases  in fuel consumption, increases in maintenance costs  and
decreases   in  engine  life.     Increased  fuel   consumption   has
traditionally been a  potential  problem in  engines  designed to
reduce  NOx emissions.  This reduction in fuel economy is usually a
side effect of retarding the timing of the combustion (i.e., spark
timing  in ••* otto-cycle  engines  or  fuel   injection  timing  in
diesel-cycl« engines).  It is  not  clear at this time,  however,
whether engine manufacturers will need or choose to sacrifice fuel
economy to reduce NOx  emissions.  Historically,  manufacturers have
been reluctant to do  so and have  sought other  technologies which
reduce NOx emissions  without increased fuel consumption.  Up to a
one  percent   increase  in  fuel  consumption  is   possible  if
manufacturers  choose  emission  control  approaches  which have an
 adverse  effect on fuel  consumption.    However,  based  on  the


                               4-1

-------

-------
                                             -• •^;--'-v;:~"'";-\~-^^-^^.t^'54:^":^i;">-^' •^.•r','c:^."','^i
technology  being developed to meet the 4.0 g/Bhp-hr NOx  standard
 (see Chapter  2  of this  document) ,  it  is more  likely that  NMHC+NOx
emissions will be reduced through the addition of emission control
hardware without a fuel economy  penalty.

     Since engines for heavy-duty clean- fuel fleet vehicles are not
expected to have significantly different  designs than  the engines
designed to meet the 1998  heavy-duty NOx . standard, they  are not
expected to be  less  durable or more  expensive to maintain  than
their general-use counterparts.   Therefore, EPA does not anticipate
that heavy-duty clean- fuel engines or vehicles will  be  any  more
expensive   to  operate  than  their   contemporary  general  use
counterparts  and therefore  that  all costs will  be associated  with
engine and emission control modifications.

       4.0.2.2 Engine Costs

     Engine   costs  consist of  four elements:    research  and
development  (R&D)  costs, additional hardware requirements, added
manufacturing costs and engine  certification costs.    Additional
hardware and added manufacturing  are variable costs included in the
price of each engine purchased.  Engine certification and R&D costs
are  fixed costs paid  up-front  by the,_engine manufacturer and
recovered through additional costs added to each engine over  a
period of time.  For purposes of this  analysis,  the certification
costs will be recovered on  a yearly basis, while the R&D costs  will
be amortized  over a period of five years  using a rate of return  of
7  percent.     (In  contrast  to   the  proposed  regulatory  support
document, the rate of return used in this  analysis is 7  percent
instead of  10 percent as recommended by  EPA' s Office  of  Policy,
Planning, and Evaluation for this type of analysis)19

         4.0.2.2.1  Otto-Cycle Heavy-Duty Engines

           4.0.2.2.1.1  Hardware  Costs

     Gasoline (otto-cycle)  heavy-duty  engines are  not  expected  to
require the  development of new  hardware to  meet the  heavy-duty
clean-fuel fleet vehicle emission standard.  However, as discussed
in  Chapter  2,  some will  require  improvements in the  existing
hardware i  Nearly all  otto-cycle  heavy-duty engines currently  have
electronically controlled  fuel  injection and three-way catalytic
converters capable of both reducing NOx emissions and oxidizing  HC
emissions,  with  oxygen  sensors and feedback controls.   Compliance
with the clean-fuel fleet vehicle standards,  however,  may require
higher catalyst  loadings,   which would increase catalyst  costs.
Other improvements or minor modifications may or may not  increase
the manufacturing  costs .   In order to take these potential costs
into account,  engine production cost increases will be estimated  at
$50 per engine.  This is conservative  given that current  NMHC+NOx
certification levels are already close to the standard (see Table


                              4-2

-------

-------
 3-3)  and  the  need  for  all heavy-duty  engines to  meet the  4.0
 g/Bhp-hr NOx standard in 1998.

            4.0.2.2.1.2 Development Costs

      In  order  to  comply with  the heavy-duty  clean-fuel  fleet
 vehicle  standard,  some otto-cycle  heavy-duty engines will  require
 minor improvements  to existing systems  beyond those  needed to
 achieve  compliance with the 1998 4.0 g/Bhp-hr NOx standard.  It is
 expected  that  some  combination  of  better  air/fuel   handling,
 improved catalytic converter technology, exhaust pipe insulation
 and enhanced exhaust gas recirculation will be  the main compliance
 strategies.   In order to make  the design changes necessary,  EPA
 estimates  that the calibrations and other development efforts will
 cost  approximately $30,000 per  engine  family."
20
            4.0.2.2.1.3   Certification  Costs

     To  certify  an  engine  family,  manufacturers must  .perform
emission  tests on  a  representative engine from  that  family and
submit  the results to  EPA;  this  only need occur  once  for each
engine   family  at   the  beginning   of   its   production,   with
certification being carried over in each, subsequent year for about
80 percent of  families.  Recordkeeping and reporting requirements
accompany new  certification  and annual recertification.   EPA has
estimated  that the cost of  certification testing for heavy-duty
gasoline engine families is about $200,000 per family21 and  the cost
for  reporting  and recordkeeping  is  about  $100,000  per family
certified  or   recertified.22    In addition   to  these  costs,
manufacturers  must pay  a  certification  fee of  $12,500  for each
heavy-duty engine  family.  Thus the total certification costs for
each heavy-duty gasoline engine family are projected to be about
$312,500.

          4.0.2.2.2 Diesel-Cycle Heavy-Duty Engines

           4.0.2.2.2.1  Hardware Costs

     Manufacturers  will  apply  refined  or   improved  control
technologies that will allow  compliance without an increase in fuel
consumption.  As was discussed in Chapter 2, it  appears  likely that
further optimization of the technology  that will be available as a
result  of the 1998 4.0 g/Bhp-hr NOx  and 1994  0.10  g/Bhp-hr PM
standards will  allow  manufacturers to meet  the  clean-fuel fleet
vehicle standards for  many  engine families. Technologies which are
expected to be used in some conventional diesel engines to comply
with the 1998 4.0 g/Bhp-hr NOx standard include  improved electronic
engine  controls  and exhaust gas  recirculation,  and some of the
other  technologies discussed  in  Chapter 2.    Particulate  trap
oxidizers and catalytic oxidizing converters (to reduce PM and HC
emissions) should make  a  broad penetration into  the  heavy-duty
diesel engine  market  in order to  facilitate compliance  with the

                              4-3

-------

-------
 0.10  g/Bhp-hr PM standard.  Clean-fuel diesel engines are expected
 to be optimized versions of the  cleanest  of these  conventional
 engines,  and may also include advanced turbocharging and catalytic
 converters  (for NMHC control).

      By basing  the  clean-fuel diesel  engines  on  only the cleanest
 conventional diesel engine families, manufacturers will be able to
 minimize  costs.  EPA estimates that the additional  manufacturing
 cost  of the  improved control system hardware for  heavy-duty clean-
 fuel  fleet  vehicle engines,  which will  improve emissions while
 maintaining  the same fuel consumption, will be about $100 dollars
 per engine more than the costs associated with the  cleanest  1998
 baseline  engines.   This $100 is roughly equivalent to the lifetime
 cost  of  a one percent increase  in fuel  consumption  (i.e., $ 100
 fuel  economy penalty)  for light- and medium-heavy duty  diesel
 vehicles.    Thus,  in order  to  avoid the fuel  economy  penalty,
 manufacturers would likely make  the necessary hardware  changes at
 a  cost ($100) equivalent to the amount the consumer would spend in
 lifetime  operating  costs.

           4.0.2.2.2.2 Development Costs

      As with otto-cycle engines,  diesel-cy.cle engines are expected
 to require additional development beyond that required for engines
 meeting the  1998 4.0 g/Bhp-hr  NOx standard.   It is expected that
 several   additional  calibrations  will  be   required.    These
 calibrations will require design changes  and engine modifications.
 The cost  of  these calibrations and other development  is  estimated
 to be about  $!0a, 000 (in  1992 dollars) for each engine family.20

           4.0.2.2.2.3  Certification Costs

     As with otto-cycle  engines,  HDD engine manufacturers will
 incur  costs  for certification  of engines for  clean-fuel fleet
vehicles  as  well as the associated  reporting  and  recordkeeping
requirements.   EPA  estimates  that these costs are $260, OOO23 per
family for certification  testing and  $78,000  per family  certified
for reporting  and  recordkeeping.   (The estimated reporting and
recordkeeping costs  for diesels are less  than  those for  otto-cycle
engines because information collection for evaporative testing does
not exist for diesels.)22   As was  the case with gasoline engines,
EPA projects that   about 80  percent of  the  families  will be
recertified  each year  using  carryover  provisions  and  that 20
percent will engage  in   new certification and  the   accompanying
testing.   The recordkeeping and reporting costs cited above apply
to each family each year  as does  the certification  fee of $12,500.
Thus  the  total certification  costs  for each heavy-duty diesel
engine family is projected to be about $350,500.

       4.0.2.3  Aggregate  Costs
                              4-4

-------

-------
      As was discussed above, heavy-duty clean-fuel fleet vehicles
 are  not  expected to have  any increases in  operating costs over
 general heavy-duty  vehicles;  thus, essentially  all  the costs of
 this  program should  come  from  increased  engine/vehicle  costs
 Since the number  of clean-fuel heavy-duty  engines required to be
 purchased will be relatively small, the per engine and total costs
 of this portion of  the  Clean-Fuel  Fleet program will be a strong
 function of how many engine families  are  developed and certified
 for this program.   For this analysis,  it is estimated  that a total
 of six light and medium  heavy-duty  otto-cycle and twelve light and
 medium heavy-duty diesel-cycle engine families  will  be certified
 for participation in the Clean-Fuel Fleet program. For comparison,
 in 1991 a total  of nine  otto-cycle  and fourteen diesel-cycle light
 or medium heavy-duty engines were certified.    If  significantly
 fewer engine families are certified, costs will be less than those
 estimated here.

      In addition,  to be conservative this analysis has  assumed that
 ful-|;. certification costs will be incurred  for each engine  family
 certified,  even though most families  will probably be able  to be
 certified based on California test data. In  such cases,  separate
 federal  testing  would  probably   not  be  necessary   and  the
 certification testing costs correspondingly lower.

          4.0.2.3.1  Manufacturer  Costs

      Based on the development costs  projected above, the total cost
 ?l ftnenni°PiHg *+"+  'I"0-'?016 engine families will be approximately
 $180,000;  the  total cost of developing twelve heavy-duty diesel-
 cycle engine families will be approximately $1,200,000.  Similarly,
 the  first-year  certification costs for six otto-cycle and twelve
 diesel-cycle  engine families will  be  $1,200,000 and  $3,120,000,
 respectively.  For subsequent years, this analysis assumes that one
 otto-cycle engine  family and two diesel-cycle engine families will
 be   recertified   with   emission    testing   required   each   year
 ^nSrSni     Y   °  Perce^fc> '•' certification costs  would then  total
 $200,000 per year and  $520,000  per year,  respectively.  Annual
 reporting, recordkeeping, and certification fees  for  all  families
 amount  to $675,000  for  otto-cycle engines  and $1,086,000  for
 diesel-cycle  engines.   The fixed costs  to manufacturers  for
 developing each engine family are presented in Table  4-1.

     The total costs to manufacturers  will  consist of  these  fixed
 costs of developing and certifying each engine family combined with
the  variable costs  of  manufacturing  the  engines.    Using  the
projections of the number of clean-fuel fleet  vehicles  required to
be purchased from Table 3-1  in chapter 3, the yearly variable  costs
to  manufacturers  from  hardware   and   production costs can be
calculated (per  vehicle  production  +  hardware cost  * number of
vehicles).   These  costs  have been analyzed through the year  2020
 (the first 22 years that the standard  is in effect).
                              4-5

-------

-------
 Tabla 4-1   Manufacturer Fixed  Costs  for  Heavy-Duty  Clean-Fuel
 Engines





Otto-
cycle
Diesel
—cycle
Number
of
Families


6

12

Total
Develop.
Costs ($)


180,000

1,200,000

Total
First-Year
Cert.
Costs ($)


1,200,000

3,120,000

Total
Annual
Cert.
Costs
($)

200,000

520,000

Total
Annual
Recording
Reporting
& Fees
Costs ($)
675,000

1,086,000

     In   order  to  calculate  the   total  aggregate  costs  to
manufacturers, all  costs  are discounted-to-the first year of the
standard, 1998.  Research and development costs will be  assumed to
occur  in the  second year before  the standard  goes into effect
(1996).      Initial   costs   for   certifying   (and   fulfilling
reporting/recordkeeping  requirements)  for  all  engine  families
certified as clean-fuel fleet vehicles is assumed to occur in  1997,
with full recertification occurring annually thereafter  for only
one otto-cycle family  and two diesel-cycle families, as described
above.   The assumed chronology for  the incurring  of  costs for
research,  development,  and  testing  (RD&T),   certification and
reporting, and for  the hardware costs is presented in Table 4-2.
The present value costs to manufacturers accrued during the  first
22 years of the standard discounted to 1998  (in 1992 dollars) are
also presented in Table 4-2.   The total  present  value costs to
manufacturers of the first 22 years of the program discounted to
1998  (in  1992  dollars) is approximately  $19.1 million for otto-
cycle  engines  and  approximately  $37.7 million  for diesel-cycle
engines  (using a rate  of return of 7 percent).
                              4-6

-------

-------
Tabl« 4-2  Costs to Manufacturers

Year
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2010
2015
2020
1998 NPV
Otto-Cycle
RD&T/
Cert .
$180,000
$1,875,000
$875,000
$875,000
$875,000
$875,000
$875,000
$875,000
$875,000
$875,000
$875,000
$875,000
$875,000
Hardware


$464,900
$471,300
$477,900
$484,700
$491,800
$499,200
$506,800
$514,800
$558,700
$610,800
$672,000
19,193,200
Diesel-Cycle
RD&T/
Cert .
$1,200,000
$4,206,000
$1,606,000
$1,606,000
$1,606,000
$1,606,000
$1,606,000
$1,606,000
$1,606,000
~~ $1,606, 000
$1,606,000
$1,606,000
$1,606,000
Hardware


$1,034,200
$1,032,500
$1,031,300
$1,030,200
$1,029,500
$1,029,100
$1,029,000
$1,029,100
$1,034,900
$1,049,400
$1,073,300
37,650,400
                                4-7

-------

-------
          4.0.2.3.2  Costs to Users

      Since clean-fuel heavy-duty vehicles are not expected to have
 different fuel consumption or maintenance costs than general use
 light and medium heavy-duty vehicles, the only cost to the consumer
 will be the first cost of purchasing the vehicles.   In addition to
 the cost for hardware changes, consumers will also have to pay for
 the amortized cost  of the  research, development,  and  testing as
 well as for the  retail price mark-up.

      Manufacturers are expected to  recover the development  costs
 and first-year certification costs  over the  first five years of
 engine sales. By amortizing these costs over the predicted  sales
 during the  first five years of the  program, EPA has  calculated that
 the development  costs and first-year certification costs of  these
 engine families  will add an average of  $60 to the manufacturer's
 cost of a clean-fuel heavy-duty  otto-cycle engine and $147 to the
 manufacturer's  cost  of   a  clean-fuel heavy-duty  diesel engine,
 respectively.

     Adding on the estimated additional certification, hardware and
 manufacturing costs  for  both otto-  and diesel-cycle  heavy-duty
 clean-fuel  engines and factoring in an estimated 29 percent retail
 price  mark-up it  is estimated that these engines will  cost  about
 $246 more per otto-cycle engine  and  $477 more per  diesel-cycle
 engine  than for  engines  that would be used in general  heavy-duty
 vehicles  during the first five  years of the program.   During the
 remaining years of the program (from year 2003 to 2020),  clean-fuel
 otto-cycle  engines are estimated  to cost an  additional $147  to
 $178, and clean-fuel diesel-cycle engines are estimated to cost an
 additional  $332 to $338  in 1992 dollars. (These additional  costs
 for  clean-fuel engines are  different from  year to  year  due  to
 variations  in the projected number  of clean-fuel  fleet vehicles
 required  to  be purchased  (see Table  3-1).)

     Table 4-3 presents total consumer cost projections through the
 year 2020, along with the  aggregate cost, expressed in 1992 dollars
 discounted  to the year 1998.   As  the table shows,  costs  in  the
 first  five  years  of the program  are marginally higher  than  in
 successive  years  because development  costs  are being recovered
 during this period.  The  22-year present value costs  of the  fleet
will be $24.8 million for otto-cycle engines and $48.6 million for
diesel-cycle engines.  (The costs to consumers have increased  since
the  proposed regulatory  support document due  to the use  of a 7
percent rate of return instead of  a  10 percent rate of  return.)
                              4-8

-------

-------
Table 4-3  Costs to Consumers
Year
1998
1999
2000
2001
2002
2003
2004
2005
2010
2015
2020
1998 NPV
Otto-Cycle
$2,283,500
$2,299,300
$2,315,800
$2,332,700
$2,350,400
$1,772,700
$1,782,500
$1,792,800
$1,849,500
$1,916,700
$1,995,600
$24,759,200
Diesel-Cycle
$4,925,200
$4,920,700
$4,917,300
$4,914,200
$4,912,400
$3,399,200
$3,398,900
$3,399,300
$3,406,700
$3,425,300
$3,456,300
$48,569,100
                             4-9

-------

-------
    4.0.3  Coat  gfg«ctivanafl«

      The cost effectiveness of the heavy-duty portion of the Clean-
 Fuel   Fleet  program   was  calculated  using   a  22-year  cost
 effectiveness method.   The 22-year cost effectiveness analysis is
 performed  by discounting  both the  total consumer  costs  and the
 total benefits  of the first 22 years  of  the  program to 1998 and
 dividing  the costs by  the benefits.   Since  the costs have been
 estimated  separately for otto-cycle and diesel-cycle engines, their
 cost  effectiveness will be analyzed separately also.  It should be
 noted that for  otto-cycle  engines which have  both NMH
 C and NOx  emissions reductions, the  costs  have been  divided evenly
 between NMHC and NOx.

      The 1998 present value for the 22 years of emission  reductions
 are  calculated from  the  data in  Table  3-6  of chapter  3.   As
 discussed  in Chapter 3, this analysis generates  emission benefits
 for clean-fuel fleet vehicles that may be overly conservative.  The
 22-year present value emission benefit for diesel-cycle  engines is
 4,000  tons of  NMHC  emission  reductions  and  16,000 tons  of NOx
 emission reductions.  For  otto-cycle engines,  the 22-year present
 value benefit is 2,500 tons of NMHC emission  reductions and 9,700
 tons  of NOx emission reductions.     ..  	

     The 22-year cost effectiveness, in dollars per ton,  can now be
 calculated by dividing the  present value costs  by the present value
 emission benefits.   The costs  of the otto-cycle and diesel-cycle
 engines are divided evenly between the NMHC  and NOx  benefits.  The
 22-year cost effectiveness for otto-cycle  and  diesel-cycle engines
 is presented  in Table  4-4.   The 22-year cost  effectiveness for
 otto-cycle engines is  $5,000 per ton of NMHC  emission reductions
 and $1,300 per ton of NOx  emissions reductions.  For diesel-cycle
 engines, the resulting 22-year  cost effectiveness is $6,100 per ton
 of NMHC emission reductions and is $1,500 per ton of NOx emission
 reductions.  The relatively high cost effectiveness of NMHC control
 and low cost effectiveness of NOx control is  a result of the cost
 allocation method used.  Given the nature of the standard (NMHC +
NOx)  and the large number  of possible ways  to split controls and
 control  costs  undue  importance  should  not   be placed  on  the
numerical  values  derived.   A simple reallocation of costs  or
 assumed change in NMHC/NOx control fraction would change the cost
effectiveness value.  Given, the depth of analyses possible at this
time  any  change probably  would  not be meaningful  until  further
 information becomes available.
                              4-10

-------

-------
   Table 4-4  Cost Effectiveness in $/ton (1998 Present Value)

Costs ($)
NMHC (tons)
NOx (tons)
NMHC Cost
Effectiveness
($/ton)
NOx Cost
Effectiveness
($/ton)
Otto-Cycle
25 million
2,500
9,700
5,000
1,300
Diesel-Cycle
49 million
4,000
16,000
6,100
1,500
    4.0.4 Summary

     The total cost of the heavy-duty Clean-Fuel Fleet program to
consumers is estimated to be approximately $7.2 to $7.3 million
per year during the first five years of the program (1992
dollars), and approximately $5.2 to $5.5 million per year after
that.  Estimates of the cost effectiveness of this program range
from $5,000 to $6,100 per ton for NMHC control and range from
$1,300 to $1,500 per ton for NOx control.
                              4-11
                                                                          I

-------

-------