Draft Regulatory Analysis
Heavy-Duty Diesel Particulate Regulations
     Environmental Protection Agency
   Office of Air, Noise, and Radiation
   Mobile Source Air Pollution Control

-------
           Draft Regulatory Analysis
   Heavy-Duty Diesel Particulate Regulations
        Environmental Protection Agency
      Office of Air, Noise, and Radiation
      Mobile Source Air Pollution Control
                 Approved by:
Michael P. Walsh, Deputy Assistant Administrator
    for Mobile Source Air Pollution Control
                          Date:  December  23,  1980

-------
                                -1-







                               NOTE







     This document has  been  prepared  in satisfaction of the Regu-




latory  Analysis  and  the  Urban and Community  Impact  Analysis  re-




quired by Executive Order 12044 and the Economic Impact Assessment




required by Section 317  of  the amended  Clean Air Act.  This docu-




ment  also contains  an  Environmental  Impact  Statement for  the




proposed Rulemaking Action.

-------
                                -11-
                         Table of Contents


Chapter                                                    Page


I.   Summary  .......................   1

     A.   Background  ....... .  .....  ......   1

     B.   Proposed Rulemaking ..........  .....   1

     C.   Heavy-Duty Diesel Characterization
          and Industry Description  ............   1
     D.   Standards and Technology  ............   2

     E.   Environmental Impact  ..............   7

     F.   Economic Impact .................   8

     G.   Cost Effectiveness  ...............   8

     H.   Alternative Actions Considered  .........  10


II.  Introduction .....................  13

     A.   Background of Heavy-Duty Diesel
          Particulate Emission Regulation .........  13

     B.   Description of Particulate Emission
          Control from Heavy-Duty Diesels .... .....  14

          1.   Test Procedure and Instrumentation .....  14

          2.   Emission Standards .............  14

     C.   Organization of the Statement ..........  17


III.  Description of the Product and the Industry  .....  19

     A.   Heavy-Duty Diesel Vehicles  ...........  19

     B.   Heavy-Duty Diesel Engines ..... .......  20

     C.   Structure of the Heavy-Duty Diesel
          Industry  ... .................  26

          1.   Heavy-Duty Diesel Engine
               Manufacturers  ... ............  26

-------
                                -111-
                     Table of Contents (cont'd)


Chapter
          2.   Heavy-Duty Diesel Vehicle
               Manufacturers	  29

     D.   Future Sales of Heavy-Duty Diesels  	  31
IV-  Standards and Technology	40

     A.   Introduction	40

     B.   Trap-Oxidizers	40

     C.   Engine Modifications  	  46

     D.   Particulate-NOx Relationship  	  52

     E.   Rationale for Level of Control	52

          1.   Baseline Level	54

          2.   Choice of Standard	59


V.   Environmental Impact 	 .....  65

     A.   Health Effects of Particulate Matter  	  65

     B.   Health Effects of Diesel Particulate  	  65

          1.   Size-Related Effects 	  66

          2.   Chemical Composition-Related Effects ....  71

     C.   Visibility		 .  71

     D.   Current Ambient Levels of TSP	  75

     E.   Impact of Diesel Particulate Emissions  .....  83

          1.   Emissions	83

          2.   Regional Impact	  87

          3.   Localized Impact	91

     F.   Air Quality Impact of Regulation	95

-------
                                -iv-
                     Table of Contents (cont'd)
Chapter                                                    Page
     G.   Secondary Environmental Impacts
          of Regulation	  97
VI.  Economic Impact	103

     A.   Costs to Vehicle Manufacturers  	 103

          1.   Emission Control System Costs  	 103

          2.   Certification Costs  	 106

          3.   Costs of Selective Enforcement
               Auditing (SEA-)	Ill

          4.   Test Facility Modifications   	 114

     B.   Costs to Users of Heavy-Duty Diesels   ...... 117

     C.   Aggregate Costs — 1986-1990  	 121

     D.   Socio-Economic Impact 	 123

          1.   Impact on Heavy-Duty Engine
               Manufacturers	  . 123

               a.   Capital Expenditures	 123

               b.   Sales of Heavy-Duty Vehicles   	 133

          2.   Impact on Users of Heavy-Duty Diesels   .  .  . 135

          3.   Impact on Urban Areas and Specific
               Communities	136


VII. Cost Effectiveness ....... 	  ...... 138

     A.   1986 Heavy-Duty Diesel
          Particulate Standard  .........  	 138

     B.   Comparison of Strategies  ..... 	 140


VIII.Alternative Actions  	 .... 152

     A.   Control of Stationary Sources 	 152

-------
                                 -v-







                     Table of Contents (cont'd)







Chapter                                                    Page







     B.   Control of Other Mobile Sources 	 152




     C.   Alternative Individual Vehicles 	 154







Appendix I	166







Appendix II	170




     A.   Emission Control System Costs 	 170




     B.   Savings Due to Maintenance Reductions 	 192




     C.   Sensitivity Analyses   	 200




          1.   Learning Curve	201




          2.   Number of Suppliers	201

-------
                                -1-

                             CHAPTER I

                              SUMMARY

A.   Background

     Heavy-duty vehicles powered by diesel  engines are a signifi-
cant source  of particulate emissions, especially  in urban areas.
Currently, diesel engines power a third of the heavy-duty vehicles
sold  in this  country.  By  1995,  though,  it  is  projected  that
diesels will comprise about two-thirds  of heavy-duty vehicle sales.
Over a  third  of these emissions will  occur  in urban areas, where
the total suspended particulate problems are most acute.

     Based on the above and the fact that Congress has required the
control  of  particulate emissions  from these  vehicles  through the
1977 Amendments  to the Clean  Air  Act, EPA  is proposing emission
standards to control particulate emissions from heavy-duty vehicles
powered by diesel  engines.   Also  included  are changes  in the test
equipment and  procedures  currently used to  measure  gaseous emis-
sions from these vehicles.   These  changes  will allow the measure-
ment of  particulate  emissions concurrently with the measurement of
the  currently  regulated gaseous  emissions  without  affecting the
stringency of current gaseous  emission  standards.

B.   Proposed Rulemaking

     Section 202(a)(3)(A)(iii)  of  the  Clean  Air Act,  as amended,
requires the  Administrator  to prescribe particulate emission
standards by the 1981 model year.    It  is under this authority that
EPA  is  now  proposing  a  Federal  heavy-duty diesel particulate
emission standard  for 1986  and  later model year vehicles. The
standard  was  delayed until 1986  due primarily to the  lack of an
adequate  test  procedure.  Also, the  existing hydrocarbon and smoke
standards appeared  capable  of  holding current  particulate levels in
reasonable check in  the absence of  strong  forces  in the opposite
direction.

     The  proposed  changes  to  the  existing  regulations  include:

     1.    The addition of a dilution tunnel and other equipment to
measure particulate emissions,  and

     2.    The  implementation  of an  exhaust  emission  standard for
particulate  matter  from diesel-powered heavy-duty vehicles of 0.25
gram per brake  horsepower-hour  (0.093 gram per megajoule) beginning
with the 1986 model year.

C.   Heavy-Duty Diesel Characterization and Industry Description

     The  particulate  regulations  being proposed apply  to  diesel-
Note:  All  references  referred to in  these  chapters  are shown as

-------
                              -2-
powered  heavy-duty vehicles.   The  heavy-duty vehicle  class
consists  of vehicles  rated at more  than  8,500 pounds  (3,546  kg)
gross vehicle  weight rating (GVWR).   This  vehicle  class  would also
include those  vehicles  under  8,500  pounds  (3,546 kg)  GVWR  which
have a total frontal cross  section of more than 46 square feet (4.3
square meters).

     Currently, about one-third of the heavy-duty  vehicles  sold in
the U.S.  are  powered by diesel engines.  The engines are made
primarily  by  five  U.S. manufacturers whose sales  comprise  97
percent of  domestic  sales; Cummins,  Detroit  Diesel  (CMC),  Cater-
pillar, Mack,  and International Harvester.   The  remaining  three
percent  are  produced  by  a  number  of  foreign  manufacturers.

     Due  primarily  to the  rising  cost  of fuel, the  percentage  of
heavy-duty vehicles  sold  with diesel engines  is  projected  to
increase  dramatically  over  the next  15  years.  By  1995,  EPA
projects  that  diesels will power  nearly  two-thirds of  all  heavy-
duty vehicles  sold in the U.S.  This,  coupled with general  growth,
is expected to increase  sales  of heavy-duty diesels by 166  percent
between 1980 and 1995.

D.   Standards  and Technology

     The  Clean Air Act,  as  amended in August 1977- requires  heavy-
duty diesel particulate emission control  based upon  control  tech-
nology which the Administrator  determines will be available  for  the
model year to  which such standards apply.   Due  consideration must
also be given to cost,  energy, leadtime  and safety.  The 0.25 gram
per  brake  horsepower-hour  (g/BHP-hr)  (0.093  gram per  megajoule
(g/MJ)) proposed standard fulfills these requirements.

     The  level  of this proposed standard was based on:

     1.   An engine-out  particulate emission level of 0.41 g/BHP-hr
(0.153 g/MJ);

     2.   A 60  percent  reduction  in engine-out particulate  emis-
sions from the  application  of  trap-oxidizers;

     3.   Over  the  full  useful life,  an  increase  in particulate
emissions  of  up  to  20  percent due  to  engine and trap-oxidizer
deterioration;  and

     4.   A 12 percent variability in the  particulate  emissions  of
production engines  (used  to  determine  the  effect of a Selective
Enforcement Audit having  a 10 percent  acceptable quality  limit).
These 4 points  are discussed below.

     The  0.41  g/BHP-hr  level  represents  the  level  of particulate
emissions  determined to  be  technologically feasible by 1986  without
the use of aftertreatment devices  (i.e., trap-oxidizers).   Practi-
cally, it  is  the average  of  the  set of  engines  made  up of each

-------
                                -3-
manufacturer's lowest particulate emitting model  tested  by EPA on
No.  2-diesel  fuel.   This  approach  was chosen  from  among several
alternatives because  it  complies  most closely with  the  Clean Air
Act  requirements that the standard "reflect the greatest  degree of
emission reduction  achievable  .  .  .  giving  appropriate  consider-
ation to the  cost  .  . . and  to noise,  energy,  and safety factors
associated  . . .  ."

     Three   other  approaches  to  determining  the  technologically
achievable   level of  engine-out  particulate  emissions  were  con-
sidered.    They were:   1) the  worst  baseline  engine   (highest
particulate  emission level),  2)  the  lowest  particulate emis-
sion level  among the tested engines,  and  3)  the  highest  emission
level among each manufacturer's best engines.

     The reasons why these  alternatives were rejected are  presented
in detail  in  Chapter  IV.   Briefly,  the  major  fault with  options  1
and  3 is that they would  ignore the  emission reduction potential of
engine  modifications  already  incorporated on  many  current  pro-
duction-line engines.  Option  2 was  rejected  because  it would  lead
to  a standard  beyond the  technological  limits of  most  engines.
Implicit  in this  option (2) is  the  judgment  that  all   engines,
regardless  of size or application, have exactly the same  potential
for  achieving  low  particulate  emissions  as the best engine.   EPA
has  not been able  to  absolutely make  this determininat ion.

     The option  chosen,  which based the  feasible  level of engine-
out  particulate  emissions on the average of the emission  levels of
the  lowest-emitting  engine from  each  of  the five major  manufac-
turers, appears  to  best  solve  the problems present  in  each of the
previous three options and  comply with the  applicable congressional
mandate.   The level  of  0.41  g/BHP-hr (0.15 g/MJ) is  a  stringent
level,  requiring the higher-polluting  engines  to incorporate,  to  a
great degree, the demonstrated technology  of  the best engines;  yet
it recognizes some  level of difference between manufacturer's
designs  and  avoids  the problems  associated  with  focusing on
the  single  best engine (Option 2).

     Up until  this  point,  the discussion  of  engine-related  tech-
nology was  restricted  to that  already present  on  existing engines
and  avoided discussing additional engine modifications  which could
also reduce  particulate  emissions.    The   reason  for  this is  the
additional Congressional mandate  related to  heavy-duty diesel
emissions which requires  that  emissions of nitrogen oxides (NOx) be
reduced by  75  percent  from a  pre-controlled  gasoline engine  base-
line (to be proposed  for the  1986 model year).   While  the mandate
referring   to  particulate emissions  calls   for the  greatest  reduc-
tions achievable  considering cost, leadtime, safety and energy,  the
mandate referring to NOx emissions is  more specific,  calling  for  a
set  reduction from a certain  baseline  level.   In  the case of
heavy-duty diesels,  it  is  often  possible  to reduce emissions
of both  pollutants at  the same time. However,  there  are  also
those control techniques which reduce the  emissions  of one pollu-

-------
                                -4-
tant while raising the other.   To date the data available have not
shown that the NOx standard can be met using NOx control  techniques
which  do  not also  increase  particulate emissions.   Thus,  it  is
possible  that  some  NOx  control  techniques  which  increase particu-
late  emissions  (e.g.,  exhaust  gas  recirculation  and retarded
timing) may  be  necessary  to attain  the NOx standard.   Given the
specificity of  the  NOx  mandate and  its stringency,  it   would not
therefore be  reasonable to  rely  on  particulate control  techniques
which would also increase  NOx  emissions.  Thus, no such  techniques
have  been included in the determination of a  technologically
feasible level of engine-out particulate emissions.

     Particulate reductions  from  engine  modifications  not yet used
on  current engines  and  not  adversely  affecting NOx emissions were
not included  in  determining the  technologically feasible level  of
engine-out particulate emissions.  Instead, these  reductions were
reserved to mitigate increases due to NOx control.   The forthcoming
NOx rulemaking will take this into account  and  propose  a standard
which  will be attainable  by  heavy-duty diesels  which are also
complying with the  proposed particulate  standard.  This  allowance
should make  this  proposed particulate standard a  reasonable  stan-
dard in light of the NOx mandate.

     In  addition to  reducing  particulate  emissions  formed  in
the  combustion  process,  additional  reductions are  available
from  the application of aftertreatment  devices,  particularly
trap-oxidizers.    A  trap-oxidizer basically consists of a high-
temperature trapping material  housed  in a  stainless  steel  shell.
Placed in the exhaust, it collects particulate and  periodically (or
continually)  incinerates (oxidizes)  it.   The  incineration  process
usually  requires  a minimum  exhaust  temperature  of  450-500°C  to
begin.  Because such temperatures may  not normally  occur  in  heavy-
duty diesel  exhaust,  exhaust  temperatures  may need to be  artifi-
cially  raised to the necessary level  when regeneration  (i.e.,
incineration) is desired.

     The particulate collection efficiencies of many trap materials
are already very good.  Many materials, such as alumina-coated wire
mesh and metal wool, have shown efficiencies of up to 65 percent.
Slightly-modified ceramic  monolithic  substrates (similar to  those
used in automotive catalysts) have shown collection efficiencies  of
up  to  84 percent.   In  determining  the  technologically-achievable
level  of particulate emissions  with aftertreatment,  60  percent
initial collection  efficiency was used.   This  is  the  same  effi-
ciency which  was  determined to be achievable  in light-duty  appli-
cations (See  45  FR 14496).

     Several  trap-oxidizer  regeneration  approaches  have been
investigated.  The simplest  solution  would be to  continuously
(or near-continuously) oxidize the  particulate, in which case
the trap-oxidizer  would  function much like  a diesel  catalytic
converter.    The problem  with diesel  converters is  simply   in
maintaining the  high-temperature conditions that  ensure  continual

-------
                                -5-
oxidation.  Much  effort  is being expended on  producing  converters
which would function on diesels, and designs have been tested  that
are  close to what  is  needed.   An alternative  is  to oxidize  the
particulate only  occasionally,  when  enough  organic  material  has
been collected by the trap to  aid  the process  and when the  exhaust
temperature is high enough to  initiate oxidation.  Many  approaches
have been suggested to initiate the oxidation process, but  the  most
promising  is  the  addition of  an inlet  air  throttle,  which would
limit the intake air into the combustion  chambers, thus raising the
temperature of the exhaust.  The throttling  would be  periodic,  and
could be  actuated by a combination  of  the odometer reading  and  rack
position, or might have  to be linked to a controller unit  coordi-
nating  several  parameters  such as rack  position,   backpressure,
exhaust  gas  recirculation,  etc.   In  a study using light-duty
diesels,  GM reported  that  over a 1,000-mile  series  of load-up  and
regeneration tests, utilizing throttling  to  initiate oxidation,  the
trap  collection  efficiency  actually increased slightly.   There
appear  to be no  technical problems with  utilizing  throttling  to
initiate oxidation,  and  there  is evidence that throttling may
possibly  reduce engine-out particulate and NOx emissions  slightly.

     Collection efficiencies and  regeneration techniques have
progressed to the point  where the  most  critical issue  is  whether
the  efficiency  and regeneration mechanism can be maintained  over
the  useful  life  of  the  vehicle.   At  this  time, EPA has  limited
trap-oxidizer durability data,  as  researchers have been  reluctant
to  fund durability testing until other,  more  basic questions  were
solved.   The  problems  of durability are  problems which lend them-
selves  to engineering solutions;  no  major  new technology is  re-
quired.   We  are  confident that  the durability questions  will  be
resolved  in the  near future.

     EPA  is  very  confident  that  trap-oxidizers will  be avail-
able  to permit  compliance with  the 1986 standard.   As   discussed
above,  the basic  concept of the trap-oxidizer  is well understood.
The  improvements  that  are  necessary are engineering problems,  and
are more  a function of the resources allocated to the problem  than
any scientific or technical breakthrough.   In the leadtime  analysis
in the  light-duty diesel  situation  (45 FR 14496),  it was determined
that  trap-oxidizers  would  likely be available for  the  1984 model
year.   However,  to ensure  their availability, the  more  stringent
light-duty standard  was  postponed  until 1985.   The delay of  an
additional year should be  sufficient  for the  application of these
devices to heavy-duty diesels.   All of the information gained  from
the  study of trap-oxidizers  on  light-duty  diesels  to  this  date
should  be equally  applicable  to  heavy-duty diesels.   However,
unique  heavy-duty diesel  design  and  operational characteristics
(such  as   long  idling periods  which could  inhibit   regeneration)
indicate that  more time is  needed in order  to optimize their use  on
heavy-duty diesels.  The  fact that  trap-oxidizers are  not  currently
available to  permit  compliance  with  the proposed 1986 heavy-duty
diesel  standard   is  recognized,  but  given   sufficient good faith
effort  by the manufacturers,  60 percent efficient trap-oxidizers

-------
                                -6-
should be  available  in  time to be  incorporated on  the  1986 model
year  fleet.  While  the necessary  particulate collection  effi-
ciencies have  been achieved,  improvements  in  the areas  of  dura-
bility, backpressure build-up  and on-board  incineration  are still
needed.

     Left  to  the  marketplace,  it  is extremely unlikely  that
sufficient pressure would  be brought to  bear  on  the  industry
to  aggressively pursue trap-oxidizer development.  Experience
has  shown  the greatest  emission  control  development work  to
have  taken  place when direct  regulatory incentives were  in place.
Perhaps the best example of this  was the general reluctance by the
light-duty  industry to  pursue  catalyst  technology before Congress
mandated control of  gaseous  emissions  from  those  vehicles.   Since
final trap-oxidizer designs  are not  now available to  success-
fully comply with the 1986 standard,  to the  extent that  the stand-
ard  motivates  the  industry  to aggressively  pursue  research  and
development it, is  a "technology-forcing" standard.    The  term
"technology-forcing"  often  implies that the sought-after  technology
is  completely  unknown or  unforeseeable, but  such  is not  the case
here.  The  basic concept of  the  trap-oxidizer  is  very well under-
stood,  and, as  explained  above,  much development has already
occurred.   Thus, this rulemaking  is  technology-forcing only in the
respect that  it will encourage  a  feasible  control  strategy  that
might otherwise be  ignored.

     The third factor  used  to  determine  the level of control
relates to emissions  deterioration.   Data  indicating  the  degree  of
deterioration of heavy-duty diesel  engines,  with regard  to partic-
ulate  emission  over their  useful  lives, are  not  available.
However,  EPA tests of in-use light-duty diesels having accumulated
an average  48,000 miles  (77,250 kilometers) indicate that  little  if
any increase in engine-out  particulate  emissions occurs.   With the
stability  of heavy-duty  diesel emissions  of other pollutants  and
the  similarity  of  the  general emissions  stability of light-  and
heavy-duty  diesels, it is reasonable to project  that the  engine-out
particulate emissions of heavy-duty diesels will  deteriorate very
little.  Information on the  deterioration of  trap-oxidizer  effi-
ciency is  even more  scarce,  as none  are  currently commercially
available  and  durability tests of  available prototypes  have  been
waiting until  after  collection and  burn-off techniques  were  per-
fected. For the purposes of this  proposed  rulemaking, the combined
engine and  trap-oxidizer deterioration  was estimated  to be no more
than 20 percent.

     In addition  to complying with  EPA's  certification  process
for new engines, heavy-duty  diesel  manufacturers  are also subject
to a Selective  Enforcement Audit (SEA) of their  production engines,
the fourth   point mentioned  above.   As is  the  case for other  reg-
ulated pollutants,  at  least  90  percent  of  the production  engine
must meet  the  proposed  particulate  standard.   This forces  manu-
facturers  to design their emission  control systems to reach  levels
below  the  standard on  the average.   Otherwise,  if the control

-------
                                -7-
system were designed to just  meet  the  standard, only about half the
engines would pass instead of the  required 90 percent.

     To  determine  how  far  a manufacturer  must  design  below the
standard,  two  factors  must be  taken  into  account:   1)  the vari-
ability of the particulate emissions of the production engines of a
given  engine  family,  and 2)  the  small  number of  prototypes  upon
which  the  design  decision  is  made.   Overall,  it  is estimated that
the  10 percent  acceptable quality level  could force manufacturers
to  design  their  engines  to  meet a  particulate  level  22  percent
lower  than  the  standard (0.25 g/BHP-hr)  divided  by the deteriora-
tion factor (1.2),  or 0.16  g/BHP-hr (0.060  g/MJ) if they  were
unable  to  reduce  production  line  variability.  Actual  data  on the
particulate emission  variability  of production  engines  is not
available.  However, this variability was assumed to be similar to
that for  gaseous emissions, or 12  percent  of mean emissions.

E.   Environmental Impact

     Despite significant  gains  made  in  the  control of particulate
emiss.ions from stationary  sources,  there are  many air quality
regions which are not able to meet the primary  National Ambient Air
Quality  Standard  (NAAQS)  for total  suspended particulate  matter
(TSP)  of  75 micrograms per cubic meter  (annual  mean).   As  diesel
vehicles  assume  an  increasing  portion  of  the heavy-duty  vehicle
market,  their  contribution  to  ambient   TSP  levels will  increase
because  diesel  engines emit   approximately  40  times the  amount  of
particulate  that  is emitted  by  gasoline-fueled engines  equipped
with catalytic converters.

     If  the  diesel  fraction  of  heavy-duty vehicle  market  sales  is
assumed  to be 57-69  percent by  1995,  this  standard  will  reduce
particulate emissions from heavy-duty  diesels by 64 percent in 1995
with  respect  to  what  would  be  expected  without  regulation.   Na-
tional  particulate  emissions  in 1995  from  heavy-duty  diesels  will
be  reduced  from approximately 218,000-266,000 metric tons  per year
to  78,000-95,000 metric tons  per year.  Urban particulate emissions
from  these  vehicles  will also decrease  64.3 percent  in  1995  from
79,000-97,000 metric tons per year to  28,000-35,000 metric tons per
year.   This emission  reduction  will reduce  ambient heavy-duty
diesel particulate levels in  large cities (e.g., New York,  Chicago,
Los  Angeles)  from 1.7-7.2 to 0.6-2.6 micrograms per  cubic  meter.
Heavy-duty diesel particulate levels  in  smaller  cities (e.g., St.
Louis,  Pittsburgh,  Phoenix)   will also   decrease  from  1.6-4.9  to
0.6-1.8 micrograms  per  cubic  meter.   Localized levels  which occur
over  and  above  these larger-scale impacts will also decrease  from
4.9-6.0 micrograms per cubic  meter to 1.6-2.0 micrograms per cubic
meter.  These  latter   impacts  could occur  as  far  as 90  meters
from very busy roadways.

     The  above  impacts  clearly  show   the  significant ambient
particulate emission   level  reductions  that are expected  from
these  regulations.   But not  all  types  of  particulate  matter

-------
                                -8-
have  the same  level of  impact on human health.   Small  parti-
cles, which  are  much more likely to  be  deposited  in the alveolar
region and which require much longer periods of time to be cleared
from the respiratory tract,  are believed to be much more deleteri-
ous to human health  on  an equal mass  basis than larger particles.
Thus, control of  diesel  particulate (100 percent is  less  than 15
micrometers in diameter  and  approximately  97  percent  is less than
2.5 micrometers  in  diameter)  is especially important with respect
to human health.  There  is  also particular concern  over the chem-
ical composition of  diesel  particulate  emissions,  as the extract-
able organic fraction  of diesel particulate  has  been  shown to
be mutagenic in short-term bioassays.   EPA is currently performing
a health assessment  to determine the carcinogenic risk (if any) to
human health.   This uncertainty is another  factor  which necessi-
tates priority control of diesel  particulate emissions.

F.   Economic Impact

     The retail price of heavy-duty  diesel vehicles  is expected to
increase by approximately $527-650  in 1986 due  to  the  engine  and
vehicle modifications necessitated by this  regulation.  (All costs
are in terms of  1980  dollars.)  The  retail  price increase of a  new
vehicle mentioned above  is about  0.5-3 percent of the total cost of
a  new  heavy-duty diesel  vehicle.   The  range  of costs is  due  to
possible differences  in  trap-oxidizer systems  which  may be used
on different models.   The trap-oxidizer  system is also expected to
require maintenance  costing about $30 when it  is five  years old.
However,  the  vehicle modifications  involved  in adding  the  trap-
oxidizer will eliminate  the  need to replace the exhaust  pipe  and
muffler throughout the vehicle's life.  This will  save  about $409
in maintenance costs (undiscounted)  during the  vehicle's life.   In
all,  vehicle maintenance costs should decrease  by $197  due  to  the
1986 standard (discounted to  year of  vehicle  purchase).   Overall,
then,  this  regulation will cost  $349-472 per vehicle.  All of these
estimates  include  profit at both  the  manufacturer and  dealer
level.   Overall, the increased cost of owning and operating a
heavy-duty diesel  due  to  this  regulation  will  be  about 0.06
percent.

     Due  to past  and future  increases  in  the  price  of gasoline-
fueled  vehicles  due to  emission  controls  and the negligible
impact  of   this  regulation  on  the cost  of transporting  goods
via heavy-duty diesels,  EPA  expects  no  decrease in  diesel  sales
relative  to  the  sales of gasoline-fueled  vehicles  due  to  aggre-
gate environmental  regulation.   The  aggregate  cost of  this pro-
posed  particulate  standard  over  five years  (1986-1990) will  be
$249-413 million (present  value in 1980) or  $442-731 million
(present  value in 1986).   Two present value reference  points  are
given  because  two different conventions have been used in the past;
the present (1980)  and the year  the standard  is to  be implemented
(1986).

-------
                                -9-
G.   Cost Effectiveness

     The overall  and  marginal cost  effectiveness  of the proposed
1986 heavy-duty  diesel particulate standard  is $1070-1410 per
metric ton  of  particulate  controlled.   (All  costs are in terms of
1980 dollars.)  However, the traditional measure of cost effective-
ness  (dollars  per  metric  ton of  particulate controlled)  can be
made more  relevant  to health improvements by considering only the
inhalable or fine particulate that is controlled.   Based on avail-
able data,  the inhalable  and,  especially, the fine fractions of
suspended particulate may have  the greatest  potential  adverse
health impact.  When  this  is done, the marginal cost effectiveness
for  the  1986  standard  is  $1070-1410 per metric  ton of inhalable
particulate  and   $1070-1550  per  metric  ton  of  fine particulate.
Using any  of  these  three bases  the cost effectiveness of the 1986
diesel  standard  is  consistent  with  that of  stationary  source
and other mobile  source  control strategies which have been adopted
in the past.

     There  is another  step  which  can  be taken  to improve the
measure of  cost  effectiveness  and  that is to  relate it to reduc-
tions  in ambient  pollutant concentrations  instead of emission
reductions.  People's  exposure to  pollutants is directly related to
the  ambient  pollutant concentration  of the air they breathe, but
only  indirectly  related  to the  emissions  from  various  sources.
However, the data necessary to perform such an analysis are diffi-
cult to obtain and not generally available.  Still, to indicate the
potential  effects  such  factors can  have on  a cost-effectiveness
analysis, some rough  calculations  were performed.  Using some rough
indicators  of  a   source's  impact  on air  quality  relative  to its
emissions,  it  was found that diesels  produce between  45  and 188
times  the   ambient  pollutant  concentration as  the  largest  power
plants  (2,920  megawatt  heat  input)  based on  equivalent  emission
rates.    Similarly,  diesels produce between 1.1 and  4.7  times the
ambient pollutant concentration as  smaller  power  plants (73 mega-
watt heat  input), based  on equivalent  emission rates.  Only large-
scale impacts were examined.  Had  localized impacts been included,
the results could have been different.   Similarly, a comparison of
a different stationary source to  diesels could have a much differ-
ent result.  One  can  imagine  the  potential  effects of adding five
to  ten  such  factors   to the cost-effectiveness  analysis.   The
results of the previous paragraph could be made meaningless.  Thus,
while the  cost-effectiveness  of heavy-duty  diesel control  appears
to be consistent  with that of past  EPA  actions,  the use  of cost-
effectiveness  to  compare  different  source  strategies  should  be
taken very cautiously.   The  type  of  factors which  need  to be
included are simply not  available  and could drastically affect the
results. The size of  these  factors  also  shows the need to further
develop the methodology  used  to determine particulate  cost effec-
tiveness before it can really be used to identify strategies which
should  be implemented  from  those which  should not.

     The marginal  cost  effectiveness of  the   1986  standard  could
only be compared with  those  from  a few other  strategies.   Because

-------
                                -10-


the  use  of  a marginal  cost effectiveness  is relatively new,
these  values  are not readily  available  for most existing  control
strategies.   Similarly,  it was  available  for only  one future
control  strategy,  the control  of emissions  from mid-sized  steam
generators (3-73 megawatt  heat  input).   However, more future
control will be needed than the heavy-duty diesel and the  mid-sized
steam  generator  regulations  if  the nation is to meet the national
ambient air quality standard  for  suspended  particulate.   Thus,  the
cost effectiveness of  the 1986 standard should  really  be compared
to  those strategies  which  will  be needed in  the  future, which
haven't yet been  developed and  implemented.   These  strategies will
likely be  more  costly than  those  of  the  past,   since EPA has been
attempting to  implement  the  most  cost effective strategies  first.
This being  the  case,  the cost  effectiveness  of the  1986 standard
would  appear even more cost  effective than  it did against the past
strategies.   This is  all the  more  reason why the  1986 standard
appears to be  a reasonable control strategy.

H.   Alternative Actions  Considered

     Control of  particulate  emissions  from heavy-duty  vehicles  is
required by the Clean Air Act.   Thus,  EPA  does not have  the  author-
ity to forego  control  of  heavy-duty diesel  particulate emissions  in
favor  of other particulate control strategies.   However,  to  demon-
strate that this  action  is  consistent with EPA's overall strategy
for  controlling   particulate  emissions,  other  control  strategies
were examined in the  course  of  this rulemaking.  They  included
further control  of stationary source and  other mobile  sources  of
particulate emissions.    Per engine emission standards for  heavy-
duty diesels of varying stringency were  also considered  as alterna-
tives .

     Averaging concepts are not being  considered in  this heavy-duty
diesel  rulemaking.  This  decision is  based primarily on  the  find-
ings of  the Regulatory Analysis for Light-Duty  Diesel  Particulate
Regulations.    However, EPA  is  planning a detailed  examination  of
averaging concepts for  the mobile  source  area  in  a future  rule-
making. Any decisions for averaging will in part  arise  out  of that
analysis.

     The  alternative  of  further  controlling stationary sources  of
particulate emissions  as a  substitute  for  these regulations  was
rejected  for two  reasons.  First, while stationary  source controls
can mitigate the effects  of  future growth,  they  cannot  be expected
to reduce  TSP concentrations  in  urban areas.   Secondly,  further
control of stationary  sources would not  diminish the high  levels  of
diesel  particulate near  roadways  where significant  adverse  impacts
occur.

     The  control  of other mobile sources was  also considered as  an
alternative to these  regulations.  By 1986, the only class of new
vehicles  emitting significant amounts  of particulate  matter will  be
light-duty diesels.   Since  these  vehicles  have already been con-

-------
                               -11-
trolled to the  furthest  extent  possible,  further  control  is not a
viable  alternative  to  these  heavy-duty  diesel  regulations.

     The alternatives remaining  concern  both  the  level and timing
of an  individual  engine particulate standard.  The  Clean Air Act
requires this  individual  engine  standard to  "reflect the greatest
degree  of  emission control achievable through  the application of
technology which the Administrator  determines  will be available for
the model year to  which  such standards apply."  EPA must  also give
due consideration  to  cost,  energy, and  safety.   The main goal of
our analysis of  alternative levels  and dates,  then,  was to deter-
mine the  level(s)  and timing  of the  standard  which  best complied
with the requirements -of the Act.

     First, the implementation  of a one-step or a two-step standard
was considered.   The  prime  advantage  of  the  one-step standard was
that  the  final  level  of technology (trap-oxidizers)  would be
available in the same year  (1986)  as the revised NOx standards for
heavy-duty diesels.  This would allow manufacturers to design their
engines to meet both standards  simultaneously.  An interim standard
earlier than  1985 would  have  to use the  13-mode  test procedure,
which  would  not be  as  representative  of in-use  particulate emis-
sions as the transient cycle (heavy-duty  engines must certify under
the  transient  test  procedure beginning in  1985).   An  interim
standard  in  1985  would  only  hold  the  line   against  increases in
particulate emissions  at a time when  no such increases  would be
expected  and  divert  valuable  Agency  and industry  resources  from
implementing and meeting  the  1986  standards  (NOx and particulate)
and  shifting them toward  a less  effective  interim particulate
standard.  In  1986 with the coming of the revised NOx  standard, a
particulate standard  will be needed to prevent potential  increases
in  particulate  emissions.   However,  by then the  final  standard
based  on  trap-oxidizers  could  be  implemented  and  no interim stan-
dard would be needed.

     To insure  that  this was  the  case,  the  alternative of a two-
step standard with the  first step occurring in 1986 was considered
in detail.  Under  this  scenario, the  1986 standard would be based
on improved engine design, while the later standard (in this case,
1988) would be based  on the use of trap-oxidizers.  This alterna-
tive would have the  advantage of  allowing  the manufacturers more
time to develop trap-oxidizers  and  also separate this work from the
engine-related   work.    Its  disadvantages were  the added  cost of
recertifying  all engines  in 1988 and delaying air quality improve-
ments  for two more  years.  The effect  of delay on capital  and
trap-oxidizer costs was  examined,  but  no major effects were found
in either direction.   In all,  the  advantages  did  not outweigh the
disadvantages and this alternative  was rejected.  Thus,  a one
step standard was chosen for 1986.

     Second,  the  possible choices  for the level  of  this standard
were considered.   These alternative  levels have already  been
discussed in  the  section on technology  and  will not  be repeated

-------
                               -12-
here.   In  summary,  EPA examined  the various  levels  in light  of
their ability to comply with the primary Clean  Air Act  requirement
that the  standard reflect the greatest reduction  potential  achiev-
able given  the  leadtime  available.   Standards  less  stringent  than
the  proposed  standard  were  not able  to  fulfill  this requirement.
Standards significantly more stringent  than the  proposed standard
carried the risk that  a  large  number  of diesels  would not  be  able
to meet the standard  and  the cost of compliance  (and  non-compliance
penalties) could have  been  excessive.  EPA  did find the proposed
standard to be reasonable with  respect to cost, energy,  and  safety
and to comply with those requirements of Section  202(a)(3)(A)(iii)
of the  Clean Air Act.   Thus,  the  level of 0.25 gram  per brake
horsepower-hour  in  1986 was chosen to  be proposed.

-------
                               -13-


                            CHAPTER II

                           INTRODUCTION

A.   Background of Heavy-Duty Diesel Particulate Emission  Regu-
     lation

     The  regulations  examined  in this  document  are  intended  to
limit the  emission of particulate matter from heavy-duty  diesels.
The regulations  were  mandated  by Congress via the  1977  Amendments
to the  Clean  Air Act  and apply to diesel-powered  heavy-duty  vehi-
cles hereafter  designated  heavy-duty diesels.   Section  202(a)(3)-
(A)(iii) of the Act as amended  states:

     The Administrator shall prescribe  regulations under paragraph
     (1) of this  subsection  applicable  to emissions  of  particulate
     matter  from  classes  or categories  of vehicles manufactured
     during and  after model  year  1981  (or during  any earlier  model
     year, if practicable).  Such regulations  shall  contain stand-
     ards  which  reflect  the greatest degree of  emission reduction
     achievable  through  the application  of  technology  which  the
     Administrator determines will be  available  for  the  model  year
     to which  such standards apply,   giving  appropriate  considera-
     tion  to the cost  of applying such technology within  the period
     of time  available  to  manufacturers and to noise, energy,  and
     safety factors associated  with  the application of  such  tech-
     nology.   Such standards  shall  be promulgated and  shall  take
     effect  as   expeditiously  as  practicable  taking into account
     the period necessary for compliance.

     These  regulations  were necessitated  because  of  the current
national urban  particulate  problem.I/,2/  With current  projections
showing a  doubling of the  penetration  of  diesels into  the heavy-
duty market by  the early 1990's, particulate emissions  from  these
diesel-powered  vehicles  will  become  even  more   of  a  significant
source  of  particulate emissions  in  urban  areas and  a major source
in areas immediately nearby busy roadways.

     While the  Clean  Air Act required  this  standard for  the  1981
model year,  a number of  factors have  caused  EPA to postpone  the
implementation date until 1986.  First and  foremost was  the absence
of a transient  test  for  heavy-duty diesels, which is necessary  to
accurately simulate  in-use particulate  emissions.   This   test
procedure  has  just been developed  and will be  available for  all
1985 diesel  emissions testing ._3_/   Second,  the  leadtime available
for earlier  implementation dates would not  have allowed  any  sub-
stantial reduction from  uncontrolled  levels.   Third,  until  1986 and
the required revision  of the emission standard for nitrogen oxides,
there  were no  outside  forces  tending to  increase  particulate
emissions.   With  the existing  hydrocarbon  and   smoke  standards,
there was  little  reason to  expect that  particulate emissions  would
increase if left uncontrolled.

-------
                                 -14-
B.   Description of Part iculate Emission Control from Heavy-Duty
     Diesels

     1.   Test Procedure and  Instrumentation

     The  test  procedure  under  which  particulate  emissions  will be
determined is essentially the same test procedure currently used to
determine  gaseous  exhaust emissions.    The  test  for  particulate
emissions will  be performed  simultaneously with  the test  for
gaseous pollutants.  Thus,  the  driving cycles, weighting procedure,
etc.,  will  remain  the  same as  currently set  forth  in  the  current
Federal Test Procedure.  The  changes  required include the need for
additional  equipment and  instrumentation to  allow for  the deter-
mination of  the amount  of particulate  matter  being emitted.

     One  significant  change in the  test equipment will be  the
substitution of a dilution tunnel for the current baffle box.   The
baffle  box  causes  a measurable  decrease  in  particulate emissions
from diesels  due  to  particle deposition  on the  baffles.4/   Also,
the  baffle box  does not  provide the amount of-  residence  time
necessary for the organic compounds  in the exhaust  to  come to
equilibrium with  the particulate before  sampling.   The  dilution
tunnel will allow the diesel  exhaust to be diluted with ambient air
with a  minimum of  particle deposition and  allow  reactions  between
the gaseous and particulate phase to  occur  before sampling  as  they
would in real life.

     The other significant  change depends upon which of two  partic-
ulate sample systems  is  used.   If a single dilution system is  used,
then larger volumetric  sampling  systems  will be  required to  lower
the  exhaust  temperature below  the required  125°F (52°C).   If  a
double  dilution system (two-stage  dilution) is  used, then  the
existing sampler  will provide  sufficient flow for  the  first  stage
of dilution and only  a  small  second stage dilution system will  have
to be  added.   A  heat exchanger will be  required with  either  sam-
pling  system to  ensure that the mass  flow rate  of the  exhaust
sample being filtered is always  a constant proportion  of the  mass
flow rate of the  total  diluted  exhaust.  Otherwise,  the particulate
sampling system would overweight  certain portions  of the test  cycle
and  underweight  others.   Existing  gaseous  emission  regulations
already require  the systems used to measure gaseous emissions
to be proportional.

     2.   Emission  Standards

     Heavy-duty vehicles are currently  required  to  meet emission
standards  for  hydrocarbons, carbon monoxide, oxides  of nitrogen and
smoke  (diesels only),  but  no standards  exist for particulate
emissions.   The  current and  future standards  for the gaseous
pollutants  are shown  in Table  II-l.   The  proposed  standard  for
particulate  emissions  from heavy-duty diesels  is 0.25  gram  per
brake horsepower-hour (g/BHP-hr)(0.093 gram per  megajoule  (g/MJ))

-------
                                                   Table  II-l
                                  Heavy-Duty Engine Exhaust Emission Standards
Federal
Year Option
1969 I/
1970-71 I/
1972 I/
1973
1974 21
1975-76
1977-78

1979 A
B
1980-83 A
B
1984 A 4/
B 5/
1986 4/
HC
NR
275
275
275 JV
—
—
—

1.5V
—
1.5 3/
—
1.3
0.5
1.3
CO
NR
1.5
1.5
1.5 I/
40
40
40

25
25
25
25
15.5
15.5
15.5
NOx
NR
NR
NR
NR
—
—
—

—
—
—
—
10.7
9.0
75% 6/
HC+NOx
NR
NR
NR
NR
16
16
16

10 3/
5
10 3/
5
—
—

California
Option HC
275
275
180
—
—
—
A
B 1.0
A 1.5 3/
B
A 1.0
B
0.5


CO NOx
1.5 NR
1.5 NR
1.0 NR
40 2/
40
30
25
25 7.5
25 7.5
25
25
25
25


HC+NOx
NR
NR
NR
16 2/
16
10
5
—
—
5
6
5
4.5


J7    HC =  parts  per million;  CO = % mole volume.  Used for Federal Standards 1970-73 and California Standards
1969-72.

2/    Grams per brake horsepower-hour hereafter.

_3/    Measured on  1979  test  procedure (HFID for HC).  Reduced 0.5  g/BHP-hr when 1978 procedure is used (NDIR for
HC).  NDIR is allowed  in  1979  for all manufacturers, beyond 1980 only for low volume manufacturers seeking
Federal certification.

4/   As measured on transient  test procedure.

5/   Option only available  for diesels.  1979 test procedure.

-------
                                -16-
beginning with  the 1986 model year*.   This level  of control  is
expected  to  require  the  use of  trap-oxidizers  on  most  vehicles.

     With  a market  penetration  for  diesels  of  57-69 percent
by  1995,  these  standards  will  result in a 64  percent  reduction  in
particulate emissions  from heavy-duty diesels  in  1995  with  respect
to  what would  be  expected  without these  regulations.    National
particulate emissions in 1995 from heavy-duty  diesels will  be
reduced  from approximately 218,000-266,000 metric  tons per  year  to
78,000-95,000 metric  tons per  year.   Urban emissions from  these
vehicles will  also  decrease  64 percent   in 1995  from 79,000-97,000
metric  tons per  year  to 28,000-35,000 metric tons per year.   This
emission  reduction  will reduce heavy-duty  ambient  diesel  partic-
ulate  levels  in  large cities  (e.g.,  New  York,  Chicago,  Dallas)
from  1.7-7.2  to  0.6-2.6  micrograms per cubic  meter.   Heavy-duty
diesel  particulate levels  in  smaller  cities (e.g.,  St.  Louis,
Phoenix) will also  decrease  from 1.6-4.9 to  0.6-0.8  micrograms per
cubic meter.   Localized  levels which  occur over  and above  these
larger-scale impacts  will also decrease from  4.9-6.0 to  1.6-2.0
micrograms per  cubic  meter.   These  latter  impacts could occur  as
far as 90 meters away from very busy roadways.   The primary  nation-
al  ambient  air  quality standard (NAAQS) for TSP is  75 micrograms
per cubic meter.

     While  these standards  are projected   to  reduce particulate
emissions  from  heavy-duty  diesels by 64  percent,  particulate
emissions from these vehicles will  still be  about  15 times  greater
than  the particulate  emissions  from a typical  catalyst-equipped
vehicle  powered  by  a gasoline engine.   Thus,  while  the  standards
call  for  significant  control,  they  do  not  call  for control to  a
level attainable by  an alternative  type  of motor  vehicle.

     No standards are being promulgated  at  this  time  to control any
other aspects of diesel particulate besides its  total weight.
While EPA health  effects  studies  performed thus far indicate  that
certain organic materials  present  on the filter used  to  determine
diesel  particulate  mass  emissions  may  present  a  greater health
hazard than the  particulate's  effect on ambient TSP levels,  there
is  currently not  enough data  available on  which  to base  special
control  of these  substances.   It  is possible,  though, that  addi-
tional standards will  be  promulgated in the  future  to control the
emission of any  particularly dangerous   compounds  as more  becomes
known about  their unique effect  on  health.

     The new standard  for particulate  emissions  could affect the
ability   of  heavy-duty  diesels  to  meet the  revised standard for
nitrogen oxide  emissions to  be  proposed for  1986.   To prevent the
situation from  occurring where  two standards  must  be  met,  each
being feasible  alone  but  together  infeasible,  the  influence  of
conflicting  control  technologies  has been  taken  into  account  in
determining  the  level  of the  proposed particulate  standard.

     The accompanying  changes  in  the  test  equipment  are not ex-
pected  to  affect the   stringency  of gaseous  emission standards.

-------
                               -17-
Th e dilution tunnel  should  be  equally effective as the baffle box
in mixing  the exhaust  with the  dilution air  and  the additional
dilution air  should not  affect  the  measurement  of gaseous emis-
sions .

C.   Organization of the Statement

     This statement presents an assessment of  the  environmental and
economic impacts of  the particulate emission  regulations  for
heavy-duty  diesels  which EPA  is  proposing.   It   also  provides  a
description  of  the  information  and  analyses used  to  review  all
reasonable alternative actions  which were  available.

     The remainder  of  this statement  is divided  into  six major
sections.   Chapter  III presents a  brief  description of the manu-
facturers of heavy-duty diesel engines and vehicles  and the market
in which they compete.

     An  analysis  of  available particulate control  technology is
presented  in  Chapter IV.   Potential  emission standards and their
timing are also discussed in detail.

     An  assessment  of  the primary and  secondary  environmental
impacts  attributed  to  these  particulate  regulations is  given in
Chapter  V.   The  degree of control  reflected  by  the standards is
described  and  a  projection of  nationwide and  urban  particulate
emissions   from  heavy-duty diesels  in  1995  is  presented.   The
impacts of  these  regulations on urban and roadside air quality are
also presented.    Secondary  effects on other  air pollutant emis-
sions,  water  pollution  and  noise are  also  discussed in this sec-
tion.

     An examination of  the  cost  of  complying  with the new regula-
tions  is presented in  Chapter VI.   These  costs include those
incurred to install  emission  control  equipment  on  vehicles  and
trucks, costs required  to purchase  new emission testing equipment,
and  the  costs to certify new engines for sale, as well  as  any
increased vehicle  operating  costs which might occur.  Analysis is
made to  determine  aggregate  cost  for  the 1986-1990  time  frame.
Finally,  the impact that this  regulation  will  have on industry and
consumers will be reviewed.

     Chapter VII  will present  a cost effectiveness  analysis of this
action  and  compare  the results  of  this  analysis  with  those per-
formed  on other mobile  source  and stationary  source control stra-
tegies.

     Chapter VIII  will  examine alternative mobile source control
options  including  alternative  per  engine  emission  standards.   It
also will   explain  why the  alternatives  of  achieving  additional
reduction  of  emissions  from  other  mobile sources  or  stationary
sources were not  considered  to  be  acceptable substitute  actions for
these regulations.

-------
                                 -18-

                            References

\J   "National  Air Quality  and  Emissions  Trends  Report,  1976,"
~    OAQPS, OAWM, EPA,  December  1977,  EPA-450/1-77-002.

2/   "National Assessment of the Urban Particulate  Problem,  Volume
~    I:   National  Assessment,"  OAQPS, OAWM,  EPA,  July 1976,  EPA-
     450/3-76-024.

3/   "Gaseous Emission Regulations  for  1984  and  Later Model  Year
     Heavy-Duty Engines,"  Federal Register, Vol.  45,  No. 14 s
     Monday,  January 21,  1980, pp. 4136-4227.

4_/   Black, Frank,  "Comments on  Recommended  Practice for  Measure-
     ment  of Gaseous  and  Particulate  Emissions  from Light-Duty
     Diesel Vehicles,"  ORD, EPA, April 13, 1978.

-------
                               -19-

                           Chapter III

           DESCRIPTION OF THE PRODUCT AND THE INDUSTRY

A.   Heavy-Duty Diesel Vehicles

     The  Clean  Air Act  defines  the  term "heavy-duty  vehicle"  as
".  .  .a  truck,  bus,  or other  vehicle  manufacturered  primarily for
use on the  public  streets,  roads, and  highways  (not  including any
vehicle operated exclusively on  a rail  or rails) which has a gross
vehicle weight  (as determined under  regulations  promulgated by the
Administrator)  in  excess of six thousand pounds.   Such .term in-
cludes any  such vehicle which has special  features  enabling  off-
street or off-highway operation and use."_l_/

     For  the purposes of this  regulation, however, heavy-duty
vehicles are those vehicles which fulfill the above description and
which have  a gross vehicle weight (GVW)  in  excess  of 8500 pounds.
Vehicles with a gross  vehicle weight greater  than  6000 pounds but
less  than  8500  pounds are termed light-duty trucks and  have  been
considered  under  separate  regulations.   This treatment is  in
harmony with  the Clean Air Act  which  states  that  with regards  to
particulate emissions regulations, "the Administrator may base  such
classes  or  categories"  of vehicles  to  be regulated "on gross
vehicle  weight,  horsepower,  or  such other  factors  as may  be ap-
propriate."^/

     Heavy-duty  vehicles  are powered  by two types  of  engines  -
gasoline (spark ignition)  and  diesel  (compression ignition).
Generally both  type  of engines are  treated  equally by EPA regula-
tions.   In  this  instance  it  has been determined  that heavy-duty
gasoline engines are  not significant particulate emitters and that,
under the authority  of Section 202  (a)(3)(A)(iv) of  the  Clean Air
Act cited  above, they would not  be required  to certify  under the
proposed standards.   Thus the  proposed  regulations  apply  to
heavy-duty diesel engines only.

     Traditionally the  industry uses GVW as a basis  for reporting
production and sales  data.   The standard categories are:

          Class                        GVW (pounds)

             I
            II
           III
            IV
             V
            VI
           VII
          VIII                      33,001 and greater

Thus  the  proposed regulations  would  apply to part  of Class  II
0 -
6,001 -
10,001 -
14,001 -
16,001 -
19,501 -
26,001 -
6,000
10,000
14,000
16,000
19,500
26,000
33,000

-------
                                  -20-

diesel vehicles (those with  gross vehicle weights between 8,500  and
10,000 pounds)  and all  of the diesel  vehicles  in Classes  III
through VIII.

     Table III-l  gives  the  total  heavy-duty vehicle (gasoline  and
diesel)  sales  in  the United States during  the  years 1972 through
1978.   These data  include  vehicles  imported from  Canada  sold  in
this country  and  excludes vehicles built here but sold elsewhere.
As mentioned earlier there is a  discrepency  in that while the break
point  for the proposed regulations is at 8,500 pounds the industry
only reports sales for  6,000  to 109000  pound  class.  A  study
based  on 1973 production data found that 5.0 percent of all trucks
of  10,000 pounds GVW or less were in the 8,500 to 10,000 pound  GVW
range.  A similar  study based on  1977  production data  found  this
figure  to be  5.8  percent.   As .it  is  not known whether or not this
change between 1973 and 1977 is  a trend or a more random variation,
an  intermediate value of 5.5 percent  was used to develop the sales
split  shown in Table III-l.

     Table III-2 lists the total heavy-duty diesel vehicle (HDV-D)
sales  in the United  States for the years  1972  through  1978  and
Table  III-3  lists  the diesel percentages of  the  heavy-duty vehi-
cle  market.   Clearly diesels  have dominated  the  very  heavy truck
market  for years,  but  have  played no role  in  the 8,500 to 19,500
GVW  classes.   The basic  tradeoff  involved  is the  higher  initial
cost  of the  diesel  engine versus  its  lower operating  costs,
in  terms of better  fuel  economy  and  less  frequent major main-
tenance.   For example, a diesel  engine that  could be used  in a
27,000  GVW  vehicle  would  cost  anywhere  from $6,000  to  $10,000,
about  three  times  the cost of a gasoline engine for the  same
vehicle  ($2,000 to  $3,500).   But  the diesel  would  yield anywhere
from 25 to 50  percent  lower fuel  consumption (on  a  work output
basis)  and would require overhauling  only about half as often.   In
the past this tradeoff favored  the diesel only in the very largest
and  most  heavily  used  trucks.   As  Tables  III-2 and  III-3 show,
however, diesels now comprise the majority  of  new vehicles  in  the
26,001   to  33,000  pound GVW class  and  are  beginning to make  up a
significant  percentage  of the   19,501  to 26,000  pound  GVW class.
Overall, diesels  comprised 33 percent of the new heavy-duty vehicle
fleet in 1978, up  from  30  percent in  1978.   As fuel economy becomes
more and more important it  is expected  that diesels will continue
to make up a  larger portion  of the heavy-duty vehicle market.   The
introduction of  the diesel into  the  light-duty truck market
indicates that diesels will  begin  to be used  in  the lower heavy-
duty weight  classes.

B.   Heavy-Duty  Diesel  Engines

     A  heavy-duty  diesel  engine is  simply  a diesel  engine  which
powers   a heavy-duty vehicle as defined  in  the  previous  section.
Diesel  engines are reciprocating internal combustion engines which
produce power by confining a combustible mixture  in a small  volume

-------
                                                  Table III-l

                                  U.S. Sales of Trucks and Buses by GVWR (pounds)
                             (U.S. Domestic Factory Sales plus Imports from Canada)
Year
1978
1977
1976
1975
1974
1973
1972
0-*
8,500
3,218,772
2,972,752
2,525,755
1,790,355
2,088,200
2,370,208
1,929,883
8,501-
10,000
187,336
173,017
147,002
104,201
121,535
137,949
112,321
10,001-
14,000
34,014
30,064
43,411
19,497
8,916
52,558
57,803
14,001-
16,000
5,959
3,231
67
6,508
8,120
8,744
10,353
16,001-
19,500
3,982
4,989
8,920
13,916
24,366
37,043
37,492
19,501-
26,000
157,168
160,396
149,293
152,070
215,221
199,481
177,723
26,001-
33,000
41,516
32,249
22,918
24,698
32,364
40,816
40,150
33,000
and over
163,836
148,728
103,098
74,896
160,465
155,814 '
130,328
Yearly
Totals
3,812,583
3,525,426
3,000,466
2,186,141
i
2,659,187 ^
3,002,613
2,496,054
 *    The MVMA does not split sales at 8,500 pounds GVWR, but rather publishes sales for the 0-6,000 and the
 6,001-10,000 pound classes.  The split in the table represents EPA's best estimate.
                                    Total Vehicles Subject to HP Regulations
                                            1978
                                            1977
                                            1976
                                            1975
                                            1974
                                            1973
                                            1972
593,811
552,674
474,709
395,786
570,987
632,405
566,170
Source:  FS-3, MVMA  data.

-------
                                     -22-

                                 Table III-2
                 Total U.S.. Heavy-Duty Diesel Vehicle Sales*
Year
1978
1977
1976
1975
1974
1973
1972
Source
8,500-
10,000
0
0
0
0
0
0
0
: MVMA
10,001-
14,000
0
0
0
0
0
0
0

14,001-
16,000
0
0
0
0
0
296
215

16,001-
19,500
0
0
0
159
41
6
5

19,501-
26,000
*13,148
11,142
6,216
4,803
3,360
3,740
3,704

26,001-
33,000
*25,464
17,997
10,053
10,320
11,700
16,018
12,450

33,001
and over
*155,890
141,294
93,714
62,016
137,908
137,147
116,473

Yearly
Totals
*194,502
170,433
111,481
77,298
153,009
157,207
132,847

*      Includes  1978 diesel  bus  production data  but not  previous
years bus data.   Includes  Canadian  imports but excludes  all  other
imports.

-------
                    -23-
                 Table III-3
Diesels as a Percentage of Heavy-Duty Market
Year
1978
1977
1976
1975
1974
1973
1972
8,500-
10,000
0
0
0
0
0
0
0
10,001-
14,000
0
0
0
0
0
0
0
14,001-
16,000
0
0
0
0
0
3%
2%
16,001-
19,500-
0
0
0
1%
0
0
0
19,501-
26,000
8%
7%
4%
3%
2%
2%
2%
26,001-
33,000
61%
56%
44%
42%
36%
39%
31%
33,001
and over
95%
95%
91%
83%
86%
88%
89%
All HD
Vehicles
33%
31%
23%
20%
27%
25%
23%

-------
                                -24-
between the head of a piston and its surrounding cylinder, causing
the  mixture to burn, 'and allowing the resulting high-pressure
products  of combustion  gases  to  push  the piston.  - This  force
can  then  be used  as a  source  of power.   Diesel  engines  are dif-
ferentiated from gasoline engines in the way  in which ignition is
instigated.  In the gasoline engine a  metered  mixture of gasoline
and  air is  ignited  by the spark  of an electrical discharge (thus
gasoline engines are also called  spark ignition engines).   In the
diesel engine  only  air is compressed  and  heated  in  the cylinder
such  that  when diesel  fuel  is  injected  near-spontaneous ignition
occurs.   Diesel engines  are also  known as  compression ignition
engines.

     Diesel  engine design  is  a  very  large complex field.   The
rest  of  this  section  will attempt  to  explain, as  simply  as pos-
sible,  some basic diesel engine design parameters, especially
those which may be  discussed  later in this analysis.

     One basic  engine  parameter  is the  number of  piston  strokes
per  combustion  cycle.   The  four-stroke  cycle  involves  1)  an
intake  stroke in which  the fresh  air charge  is drawn into  the
combustion cylinder followed by  2)  a  compression  stroke  in which
the  air  is compressed  to a temperature suitable  for combustion.
Late  in  the compression  stroke  the diesel  fuel  is  injected  in-
to  the  cylinder.    Combustion  transpires  producing 3)  the  expan-
sion  stroke as the high  pressure  gases force the piston  down-
ward  transferring  energy  to the  crankshaft.    During 4) the  ex-
haust  stroke the exhaust gases  are  rejected  from the cylinder
due  to  the  upward  movement  of  the  piston.   This  one four-stroke
cycle necessitates  two  revolutions of the crankshaft.

     In a two-stroke engine  the  power  cycle is completed  in just
one  revolution of the  crankshaft. The basic concept involves
avoiding  complete piston  strokes  for intake and exhaust purposes.
As the piston moves to  the  top  of the  cylinder, air is compressed
for ignition.  The  fuel  is  injected initiating  combustion  and  the
piston delivers power during  the  expansion stroke.  Near the end of
the expansion stroke the exhaust  ports  open  and  exhaust gases begin
to be purged.  Also the intake  ports are opened allowing fresh air
to be  blown  into  the  cylinder.    Soon  after the piston  begins  the
following  compression  stroke the  exhaust and  intake ports  are  all
closed.

     The  primary advantage of the two-stroke engine is its greater
horsepower to weight ratio since it  has twice as many powerstrokes
per unit  time  as  the four-stroke  engine.  Its poorer scavenging and
volumetric efficiencies  are its primary drawbacks.

     A second  parameter  of  interest   is  injection timing.    This
refers to the time at which  the  diesel fuel begins to be injected
into the combustion cylinder.   Ideally it  would  be preferable  to

-------
                               -25-

inject the  fuel  instantaneously  when the  piston  is at  top dead
center (TDC)  of  the  cylinder  since at that  time  the  air  is com-
pressed to  its maximum  extent  and combustion conditions  are most
optimum.   But because  it  takes  finite  amounts of time  for the
physical  processes  of injection and  ignition  to take  place,
injection  is always  initiated  before  the  piston reaches  TDC.
The. point where injection  begins  is usually expressed  in  terms of
degrees of crankshaft rotation before  TDC.

     Also  of interest  is  whether  the  engine  utilizes  direct
or indirect fuel injection.   With direct injection the  diesel
fuel  is  introduced directly  into the cylinder head initiating
combustion there.  With  indirect injection the fuel is  intro-
duced, and combustion  is  initiated  in  a  small fuel-rich  ante-
chamber before expanding into  the rest of the cylinder.   Indirect
injection  engines are  also referred   to  as precombustion  chamber
engines  or prechamber  engines.  Indirect  injection has  several
advantages but has  not been  utilized  often  in the heavy-duty
industry due  to a slight fuel economy penalty associated  with its
use.

     Another  important parameter is the  method of introducing
air  into  the combustion  cylinder.   In  a  naturally aspirated
engine the vacuum created  behind  the moving piston  is  utilized
to draw  in the  fresh  air  charge.   Since  a  greater power output
per unit volume of the  cylinder is  possible with greater masses of
air and  fuel involved  in  the combustion  process, many  engines
pressurize  the intake air  which  allows a similar  increase  in the
amount of  fuel  which  can   be  effectively burned.   A turbocharger
combines a  turbine,  driven by engine exhaust gases, with  a com-
pressor which increases the air  flow  into  the  cylinder.   Cooling
the presssurized  air  before it  entered  the  cylinder, called after-
cooling  or  intercoo 1 ing, also  increases the  mass that  can be
accommodated and  thus the power  output.   Both two-stroke and
four-stroke  engines can be either naturally aspirated or  turbo-
charged.    All naturally-aspirated  two-stroke  engines  utilize
low-pressure blowers  to aid in the expulsion  of exhaust  gases and
intake of fresh  air.   However,  these blowers  are only  used to
remove exhaust gases  and   do  not  pressurize  the chamber.   These
engines are called blower-scavenged engines.  The primary tradeoff
involved in all of the options is  the  cost of  the system versus the
increase  in power.

     A final  parameter  of  interest, though more accurately termed
an emission  control  technique  rather  than  simply  an engine para-
meter,  is  exhaust gas  recirculation (EGR).   This involves the
recirculation of  exhaust gases directly  into the intake manifold.
The  exhaust  then  makes up  part  of the  "fresh"  aircharge  to the
cylinders.  The purpose of EGR is to reduce peak temperatures in the
cylinders by  providing  a  mass  which  can absorb  some  of  the heat
released during combustion.  The  lower peak temperatures result in
lower  emissions  of  oxides  of  nitrogen  (NOx).   EGR has  become  a
principal NOx control  strategy.   The  major disadvantage of EGR is

-------
                               -26-
th at  the addition  of the residual gases  often necessitates  a
slightly richer mixture, resulting  in  slightly  greater fuel con-
sumption .

     Two additional characteristics of  heavy-duty  diesel  engines
will  be used  for  classification  purposes:    engine displacement
and maximum  power.   Engine displacement is simply  the volume  of
each cylinder that is swept out by  each piston during  combustion,
i.e.,  from bottom dead center of the cylinder to top dead center,
multiplied  by the number of cylinders in the  engine.  Maximum power
is the  power delivered by the engine shaft at the output end when
the engine  is operated at the optimum speed  for power.   The two are
related in  that  generally the maximum power  increases as  the
displacement  increases.

C.   Structure  of the Heavy-Duty Diesel  Industry

     Unlike  the light-duty  vehicle  industry where  the engine
and vehicle  manufacturers are typically one  and  the same,  heavy-
duty  diesel  vehicles  and  the diesel  engines used in them  are
often  manufactured  by  independent  companies.    The engine/vehi-
cle interconnections are so  marked, in fact,  that  three of  the
major heavy-duty diesel engine  manufacturers  sell  engines  to
every major  heavy-duty diesel vehicle manufacturer.  Because
of this characteristic,  as  well  as because  of the logic of basing
heavy-duty  emissions on  a  useful  engine work  basis,  EPA requires
heavy-duty  engine  certification   rather , than vehicle  certifica-
tion.   This  has  facilitated the performance  of  certification
requirements  by  the engine  manufacturer  and  avoided  the  situa-
tion  where many  vehicle manufacturers might certify the very
same engine resulting in a duplication  of effort.

     The difficulty posed  by the engine  manufacturer/vehicle
manufacturer  matrix is  that  it  makes  any analysis  of  the  heavy-
duty  diesel  industry  that  much more  complicated.   However,  be-
cause  it is  the  engine manufacturer  who bears  the financial
burden  both  for any  design  changes, necessary to meet emissions
standards  and  for  the  facilities and personnel  required  for
certification testing,  it has  been concluded  that the primary
economic impact of  emission  regulations  would affect  the  engine
manufacturers rather than  the  vehicle  manufacturers.    Thus  this
chapter will place  somewhat more emphasis  on  engine manufac-
turers .

     1.   Heavy-Duty Diesel Engine  Manufacturers

    The U.S.  heavy-duty  diesel   engine  market can  be handily
divided  into  two groups of manufacturers with distinctive charac-
teristics.   One group  is  composed of the  five  large, domestic
manufacturers which  seem to  have  a fairly permanent hold on  the
market:  Cummins,  General Motors  (Detroit Diesel. Allison division),
Caterpillar,  Mack,  and  International Harvester.   The  manufacture

-------
                               -27-


and marketing of heavy-duty diesel engines in the  U.S.  is a signif-
icant concern  of each of  these  companies.    Together  these five
firms accounted  for  97.2  percent  of all heavy-duty diesel  engine
sales in the U.S. in  1978 (see Table III-4)-

     The  second  group generally  includes  large  foreign manufac-
turers which  sell a  very  small  fraction of  their  total  produc-
tion  in  the U.S.   The composition of  this  group is  thus much
more likely to be variable  as the  decision of  whether  a  particular
manufacturer will export a  small  number of already-built  engines
or vehicles can  be reversed  in a  relatively short period of time.
As  Table III-4 shows, this group accounted for approximately
2.8  percent of  1978  U.S.  diesel engine sales and was composed
of  Mercedes-Benz, IVECO, Volvo,  and Deutz  Diesel.   It  is  rather
difficult to  predict the composition  of  this  second group  in
future years, but this limitation  will  be considered later in this
analysis.

     The  leading heavy-duty  diesel  engine  manufacturer  is Cum-
mins  Engine  Company.   They  constituted  36.9  percent  of the 1978
market.   Unlike many  of  their  competitors,  Cummins does  not
manufacture  gasoline  engines nor  diesel vehicle chassis.   Cum-
mins  makes  exclusively four-stroke,  direct injection  engines
with  approximately 90  percent of  those  eventually powering  heavy-
duty  vehicles  having  a displacement of  855 cubic  inches.   Their
horsepower output ranges  from approximately 250 to 400 horse-
power which makes  them among  the most powerful truck  engines
produced.   Cummins  makes widespread use  of turbocharging  for
additional  power  (approximately 95  percent)  and uses  intercoolers
on  about half of  its models.   Approximately  two-thirds of Cummins'
total sales result from their NTC-290 and NTC-350  models.  Cummins'
engines  are utilized by  every major  heavy-duty diesel  vehicle
manufacturer, with  International  Harvester  its biggest  customer,
and  power  many  of  the very heaviest  diesel  freight  trucks.

     Detroit  Diesel  Allison  is  a division  of  General Motors
Corporation  (CMC) which is  primarily  involved  in manufacturing
the  engines used in CMC heavy-duty diesel  vehicles.   Detroit
Diesel,  which constituted  almost  one-quarter of the  market  in
1978,  manufactures  two-stroke,   direct   injection  engines   exclu-
sively and  is  the only maker of two-stroke  diesel truck engines.
It  is  believed  that  Detroit  Diesel  may soon market a four-stroke
engine.   Detroit Diesel  has a very broad  range  of displacements
(212  to  736 cubic inches)  and power  settings (170  to 430  horse-
power), thus,  their engines  are  used  in both  lighter and  heavier
heavy-duty vehicles.  About 30  percent  of  these engines  are
blower-scavenged  (naturally  aspirated)  with  the  remaining  turbo-
charged.  Most  of the  turbocharged  engines also include an  inter-
cooler.   Although approximately  one-third  of  all  Detroit  Diesel
engines  are used in  CMC  trucks, every  major heavy-duty diesel
vehicle manufacturer  is a customer of Detroit  Diesel.   Top models
include the 6V-92TA, 8V-92TA, and 6L-71N models.

-------
                             -28-
                    Table  II1-4




1978 U.S.  Heavy-Duty Diesel Engine  Sales  by Manufacturer
Manufacturer
Cummins
Detroit Diesel
Caterpillar
Mack
International
Mercedes
IVECO
Volvo
Deutz
TOTAL
Number
73,872
47,737
30,576
27,504
14,813
2,607
2,397
360
180
200,046
Percentage of Market
36.9%
23.9%
15.3%
13.7%
7.4%
1.3%
1.2%
0.2%
0.1%
100.0%

-------
                                -29-
     Th e  Caterpillar  Tractor  Company,  for  years  the leading
manufacturer  of  heavy  construction  machinery,* has  a  15.3  per-
cent share  of the  1978  heavy-duty  diesel  engine market.   Cater-
pillar  is the only major manufacturer which  produces indirect
injection engines, although  approximately five-sixths  of  its
fleet utilizes direct injection.   All Caterpillar engines  are
four-stroke  engines.   Only  about  one-fifth  of  these  engines  are
turbocharged, with most of these also using an intercooler.   As of
the  1978  model year,  Caterpillar was  also  the  only  company  mar-
keting  a  heavy-duty diesel engine  with exhaust  gas recircula-
tion (EGR) .   The  majority of Caterpillar's engines  are in  the 636
to  638  cubic  inch  size range,  though they do  make some larger
engines.   Caterpillar  sells  a majority  of  its  engines  to the  Ford
Motor Company but  also sells  its  engines to  every other major
diesel vehicle manufacturer.  Its most  popular model  is  its  stan-
dard 3208  engine,  accounting for approximately  70  percent of  its
total sales.

     The  fourth  largest heavy-duty diesel engine  manufacturer  is
Mack Trucks,  Incorporated  with 13.7  percent  of  the 1978 U.S.
market.   All Mack engines are four-stroke, direct injection  engines
with over  95  percent  of  them having a displacement of  672  cubic
inches  and utilizing  a turbocharger.  Half of its engines  are
intercooled as well.   Two models,  the ENDT(B)676 and the ENDT
(B)675,  epitomize the entire Mack fleet and themselves  account  for
over 80 percent of the  total  sales.  All Mack engines are assembled
into Mack heavy-duty vehicles.

     The International  Harvester  Company, which is a major manufac-
turer of  gasoline  engines, and diesel and gasoline heavy-duty
vehicles as well,  had a 7.4 percent share of  the heavy-duty  diesel
engine market  in 1978.   One of its engines,  the DT-466B,  accounted
for  almost  90  percent  of its sales  in  1978  and  thus,  well  repre-
sents its  entire  fleet.   It  is a  four-stroke,  direct  injection,
turbocharged engine  with a  displacement  of  466 cubic  inches  and
about 200  horsepower output  and  is used  in  some of  the  lighter
heavy-duty diesel vehicles.   All International  engines  are used  in
International vehicles.

     2.    Heavy-Duty Diesel Vehicle Manufacturers

     The U.S. heavy-duty diesel  vehicle market  is split among  more
than fifteen vehicle manufacturers.   However,  like  the heavy-duty
diesel engine  market,  most  of the vehicle market  is  concentrated
among a few  large manufacturers.  Nearly  70  percent  of the market
belongs  to the four largest vehicle manufacturers:  IHC,  Ford,  Mack
and CMC.  The seven  largest manufacturers hold nearly 90 percent  of
the market.  The  actual  breakdown of U.S. sales in 1978 is shown  in
Table III-5,   along  with  a  breakdown  of manufacturers  providing


*     This  analysis ignores all diesel  engines  used  in off-road
vehicles.

-------
                               -30-






                         Table III-5




       U.S. Sales of Heavy-Duty Diesel Vehicles - 1978
Vehicle
Manufacturer
IHC
Ford
Mack
CMC
White
Kenworth
Freightliner
Peterbilt
Chevrolet
Others
Total
Source: MVMA,

Cummins
22,331
10,108
2,057
4,667
9, 840
8,987
9,155
5,410
605
712
73,872
FS-5.

Detroit
Diesel
8,304
6,930
356
16,226
3,794
2,833
2,119
1,502
3,749
1,924
47,737
Corrected
Engine
Caterpillar
2,587
18,964
202
1,330
965
2,309
709
2,027
1,071
412
30,576
for imports,
Manufacturer
Mack IHC Others Total
14,813 - 48,035
- 36,002
27,504* - - 30,119
- - - 22,223
14,599
14,129
- 11,983
8,939
5,425
5,544 8,592
27,504 14,813 5,544 200,046
buses, and exports.
Includes 596 engines  produced  by Scania Vabis.

-------
                               -31-
engines for  the  vehicles  being produced.   As can be  seen,  every
vehicle manufacturer buys  engines  from a number of engine  manufac-
turers .  Only two vehicle manufacturers, CMC (including Chevrolet)
and Mack,  also produce a majority  of the engines  which  are used  in
their vehicles.

D.   Future Sales of Heavy-Duty Diesels

     There  are  many factors  which  could  affect  future sales
of heavy-duty diesels.   Soaring  energy costs  are  likely to be
a  major  factor.    Federal  deregulation  of  the  trucking  industry
and  changing  state  weight and  length  limitations  could also
affect future  sales.  When these  factors  are coupled with gen-
eral  uncertainties about  the nation's  economic growth in the
1980's, it becomes  obvious  that predicting  future  sales of  heavy-
duty  diesels  is a  very  difficult task.   Any  such  prediction
is going  to  be  tenuous,  being  based  on a  number  of questionable
assumptions.  Rather than try  to develop  a scenario  that  is
based  on  an  assumption  concerning each  one of  these factors,
each  assumption being  quite questionable, here  we will simply
try to predict the  effect of energy costs  on  the gasoline-diesel
engine split.   The  other  factors  mentioned above will be  essen-
tially ignored  at  this  time due to  the  uncertainty of their
effects.   This exclusion is in itself an assumption concerning the
cumulative effects of these  factors  (that of no effect). Given the
available  information, this assumption  is probably as  good  as any
other.

     The  actual  projection of future sales  will  be performed
in two steps.   First,  a  projection  of total heavy-duty  vehicle
sales  will  be made.   Second,  a  projection of the  split  between
gasoline and  diesel  engines  will  be made.   These  two  projections
will  then be combined to  yield  a scenario of  future sales  of
heavy-duty diesels.

     As mentioned earlier,  this study will  not attempt  to  account
for many  of the  factors  which could  affect heavy-duty  vehicle
sales  in  future  years.   Instead,  the historical growth rate from
1967 will  be projected to continue in the future.  This should not
be  too inaccurate  since  the  time period being  examined  includes
both periods of real growth  in  the  early years and then  a period of
retrenchment due  to  the  1973 oil embargo and the continuing rise of
fuel prices.  It is  reasonable to  expect some growth  in the  1980's
and early 1990's, though it is similarly reasonable to  expect this
growth to  be tempered by other factors,  energy prices being  one of
them.

     A compilation of annual domestic sales  of heavy-duty  vehicles
from U.S.  plants  between 1967 and  1978 is  shown in Table III-6.  A
linear regression of  this data shows  that sales  have grown  on the
average of  10,904 vehicles  per year,  with  1978  sales  (taken from
the  least-squares line) being  510,174 vehicles.   However,  imports
(all from Canada) have averaged about 10% of  these figures  during

-------
                                    -32-
                           Table III-6

                   Domestic Sales of Heavy-Duty
              Vehicles  from U.S. Plants - 1967-1978

               Year                           Sales

               1978                          555,902

               1977                          505,293

               1976                          419,368

               1975                          348,438

               1974                          513,572

               1973                          563,348

               1972                          509,503

               1971                          413,750

               1970                          366,622

               1969                          428,362

               1968                          408,820

               1967                          369,471

Results of Linear Regression:

     S = -340307 + 10903.6Y   where;

     S =  Annual sales  in a given year

     Y =  Last two digits in year

-------
                                -33-
this period.   Thus,  modifying the regression results accordingly,
the annual  growth  rate would be  11,994  vehicles  per year and the
regression  yields  561,191 vehicles  for  1978 sales.    It  will be
assumed that growth  continues  to  be  linear through 1995 at 11,994
vehicles  per  year starting  with  561,191  vehicles  in  1978.   The
resultant sales  for future  years  can be  found in Table III-7.

     To apportion  these  total sales  projections among the various
weight classes, historical data was  used  again.   The breakdown by
class"between 1974  and  1978  was compiled and averaged  and the
following breakdown resulted.

         Class  IIB          8,501  -  10,000*          28.3%
         Class  III         10,001  -  14,000*          5.3%
         Class  IV          14,001  -  16,000#          0.9%
         Class  V           16,001  -  19,500*          2.2%
         Class  VI          19,501  -  26,000*          32.2%
         Class  VII         26,001  -  33,000*          5.9%
         Class  VIII        33,001  and over            25.2%

     This breakdown  was  assumed  to  stay  constant through  1995 and
was used to allocate the projections of  total sales  among the
various classes.   These projected sales within each class  are also
shown in Table  III-7.

     The  last  step remaining is  to  estimate the breakdown between
diesel  and  gasoline  engines.   This  is  a  more  difficult  area to
predict due  to the nation's current  energy problems.   Due to the
wide range of  vehicle  types  which fall into  the category of heavy-
duty vehicles,  separate  projections  of  gas-diesel  split  will be
made for most  classes.  The  need  for  this  is  shown in Table III-3,
which shows the historical gas-diesel split  for each  class.  As can
be  seen,  there  are   large  differences  between  classes,  with the
heavier weight  classes showing the  higher  percentage  of  diesels.
This is  primarily  true because it is these heavier vehicles which
are used  in  line-haul operation  and  have the highest annual mile-
age.  Since the primary advantage  of  the  diesel is  in fuel  economy,
these  are the vehicles  where the diesel  shows the  greatest ad-
vantage.    The actual projections  of gas-diesel split  follow,
beginning with  the  heaviest classes  (VIII) and moving to the
lightest.

     Class VIII (greater than 33,000  pounds  GVWR) has traditionally
been the  class where  diesels have had the greatest penetration, as
shown by  the  figures  in  Table III-3.  The fraction  of  diesels in
this class has been  steadily increasing  over the past  seven years
and  it  is safe to assume that this  should  continue.   It  will be
assumed that this  class  will be  entirely diesel  by  1984, with the
diesel  fraction  increasing  linearly from 0.95 in 1979.  These
projections are shown in Table III-8.

-------
                    -34-
            Table  III-7
      Estimated  HDV Sales  for
1979 through 1995  by GVWR  (pounds)
Year
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
8,500-
10,000
216,520
213,126
209,731
206,337
202,943
199,549
196,154
192,760
189,366
185,971
182,577
179,183
175,788
172,394
169,000
165,605
162,211
10,001-
14,000
40,550
39,914
39,278
38,643
38,007
37,371
36,736
36,100
35,464
34,829
34,193
33,557
32,921
32,286
31,650
31,014
30,379
14,001-
16,000
6,855
6,777
6,670
6,562
6,454
6,346
6,238
6,130
6,022
5,915
5,806
5,698
5,590
5,482
5,375
5,267
5,159
16,001-
19,500
16,832
(16,568
16,304
16,040
15,776
15,512
15,248
14,985
14,721
14,457
14,193
13,929
13,666
13,402
13,138
12,874
12,610
19,501-
26,000
246,359
242,497
238,635
234,772
230,910
227,048
223,186
219,324
215,462
211,600
207,738
203,876
200,014
196,151
192,289
188,427
184,565
26,001-
33,000
45,140
44,433
43,725
43,018
42,310
41,602
40,895
40S187
39,479
38,771
38,064
.37,356
36,648
35,941
35,233
34,526
33,818
33,001
and over
192,803
189,780
186,758
183,735
180,713
17.7,691
174,668
171,645
168,623
165,600
162,578
159,555
156,533
153,510
150,487
147,465
144,442
All HD
Vehicles
765,089
753,095
741,101
729,107
717,113
705,119
693,125
681,131
669,137
657,142
645,149
633,154
621,160
609,166
597,172
585,178
573,184

-------
                                        -35-
                             Table III-8

            Projected Future Diesel Sales as a Fraction of
        Heavy-Duty Sales for 1979 through 1995 by GVWR (pounds)
Year
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
8,500-
19,500
0.30
0.28
0.26
0.24
0.22
0.20
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
0.00
19,501-
26,000 I/
0.15
0.14
0.13
0.12
0.11
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.00
0.00
19,501-
26,000 21
0.71
0.67
0.64
0.60
0.56
0.53
0.49
0.46
0.42
0.39
0.35
0.31
0.28
0.24
0.21
0.17
0.14
26,001-
33,000
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.96
0.92
0.89
0.85
0.81
0.77
0.73
0.69
0.65
33,001
and over
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.99
0.98
0.97
0.96
0.95
All HD
Vehicles
0.63
0.61
0.59
0.57
0.56
0.54
0.52
0.50
0.48
0.46
0.44
0.42
0.40
0.37
0.35
0.34
0.33
I/   School buses

2/   Trucks

-------
                                -36-
     As can be  seen, with the current diesel fraction of class VIII
so near one, the actual year in which the fraction becomes one has
little effect on the projected number of  diesel  Class  VIII sales.
This is convenient, as will become more  evident  later,  because 1)
most of total diesel sales will come from this class and 2) an even
greater percent  of diesel  vehicle-miLes-travelled will  come  from
this class.   While many of  the  following projections  of diesel-
gasoline  splits are very  tenuous,  their effect on the  overall
results are much smaller than the impact of Class VIII.   Because of
this,  the  accuracy of  the  overall  sales  projection will  be  much
greater than it might appear on the surface.

     Vehicles sold in  Class  VII (26,001-33,000  pounds)  have  also
been increasingly  equipped  with diesel engines.   Between 1972 and
1978, diesel usage increased from 31 percent  to  61 percent of the
vehicles.   Due to both  the relatively  low  sales of  this class
(40,000  vehicles  per year)  and  the  push  toward  greater  fuel
economy,  the dieselization of this class is expected to  continue at
its current pace,  reaching  100 percent in  1988.   The actual gaso-
line-diesel split up to 1988 is shown in Table III-8.

     So far, Class VII and VIII vehicles  have been projected  to
switch completely  to the diesel  by the mid to late  1980's.   This
projection is  in  accordance  with many  other  projections  made
elsewhere._3_/,kj, 5/  However, these past studies have also projected
similar levels  of  dieselization for the  lighter  classes  of heavy-
duty vehicles (Class VI and  below).   There are a number of reasons
why  this  is probably  not  going to be  the case.   One reason is
simply economics.   The  diesel"s  primary  advantage  lies  in  fuel
economy,  and its  primary disadvantage  lies  in initial  cost.   The
annual mileage  of  these lighter vehicles  is  far  below  that of the
Class VII and VIII  vehicles, so the advantage of the diesel is  much
less, though the disadvantage is the same.

     A second  reason is primarily  social.   Diesels were  tried in
the 1960's  in  these classes and were not  very successful.6/  Poor
designs and  inappropriate  use  caused a  host of  mechanical prob-
lems.  While the  quality  of future mid-size diesels should be
vastly improved over  those  used in  the  1960's, it may take time to
change prejudices from  the past.

     The  third  reason  for  lower  diesel  sales in  these  classes is
primarily  practical.   First,  diesels  do not  provide  the  same
ability to accelerate  as  their  gasoline  counterparts.   Second  they
are harder  to  start  in the  cold.   Third, an established service
industry  does  not  exist to  service  a  completely  diesel heavy-duty
fleet.   Last,  the  shortage of  diesel  fuel last summer  is still in
people's  minds.   These practical  considerations  will  all cause a
hesistancy to dieselize,  even  though  in  the long  run (20-30 years)
their effect, may be minimal.

-------
                                -37-
     A few manufacturers have predicted that Class VI sales  will  be
35-50  percent  by  1985.6J    Given  1)  the  considerations  mentioned
above,  2) the  fact that  the  50 percent  estimate  came from two
diesel  manufacturers  and the  35  percent  estimate  came  from  a
manufacturer of both gasoline and diesel engines,  and  3)  the  usual
optimism  of  industry projections, the  lower  figure has  been chosen
as  the  best estimate  of  the  gasoline-diesel split  for Class  VI
trucks in 1985.

     About  18  percent  of  1978  Class  VI  sales  were school buses.
Due  to their   suspected  lower annual mileage, these vehicles
are not expected  to dieselize nearly as much as  the  trucks  of this
class.   A  projection  of  10 percent  will be used for  the diesel
fraction of  school bus sales in 1990.   In 1978,  about 10 percent  of
class VI  trucks were  diesel and no school  buses  were  diesel.  The
dieselization  rate  for trucks  is projected  to be  linearly  between
10  percent  in   1978 and  35 percent  in  1985.   The rate  for school
buses is  also  projected to be linearly starting  with  zero  percent
in  1980  growing  to  10 percent  in  1990.   The resulting  gasoline-
diesel splits  for  1979 to  1995  for both  of  these sub-classes are
shown  in  Table  III-8.   Also,  due  to  the nationwide  trend toward
fewer school children,  total sales  of  school buses were assumed  to
remain constant  at  1978 levels.   All growth  in  Class  VI sales was
assumed to be trucks.

     For  the remaining classes  (II-V),  the  rate  of dieselization
should be  less  than that of Class  VI.  As the exact difference  is
difficult to determine, it will be assumed that  the diesel fraction
of  1990   sales  for  these  classes will  be about  the same as  that
projected for light-duty vehicles,  20 percent.^/   It  will simply  be
projected  that  this growth  occurs  linearly  from zero  in 1980 and
continues  at least  until  1995  (30  percent).  These fractions for
the various  years are also shown in Table III-8.

     The  diesel  fractions  of sales  by class  (Table  III-8)  can now
be  coupled  with  the projections  of total heavy-duty  sales  (Table
III-7)  to yield  the  overall diesel  fraction of  sales each  year
(shown in Table III-8) and  the  total  sales  of  heavy-duty  diesels
each  year by  class (Table  III-9).  As  can be  seen,  the  diesel
fraction  of  heavy-duty sales increases  from  33 percent  in  1979  to
63  percent  in   1995.   Total  heavy-duty diesel sales  increase  from
181,080 vehicles in 1979 to 481,120 vehicles in  1995.

-------
                           -38-
                  Table II1-9
     Estimated Diesel Usage in Heavy-Duty
Vehicles for 1979 through 1995 by GVWR (pounds)
Year
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
8,500-
10,000
64,956
59,675
54,530
49,521
44,647
39,910
35,308
30,841
26,511
22,316
18,258
14,335
10,547
6,896
3,380
0
0
10,001-
14,000
12,165
11,176
10,212
9,274
8,362
7,474
6,612
5,776
4,965
4,179
3,419
2,685
1,975
1,291
633
0
0
14,001-
16,000
2,065
1,898
1,734
1,575
1,420
1,269
1,123
981
843
710
581
456
335
219
108
0
0
16,001-
19,500
5,050
4,639
4,239
3,850
3,471
3,102
2,745
2,398
2,061
1,735
1,419
1,114
820
536
263
0
0
19,501-
26,000
158,941
148,233
137,784
127,621
117,737
108,132
98,804
89,755
80,984
72,490
64,275
56,338
48,931
40,852
34,723
27,223
21,878
26,001-
33,000
45,140
44,433
43,725
43,018
42,310
41,602
40,895
40,187
37,979
35,824
33,725
31,678
29,685
27,675
25,720
23,823
21,982
33,001
and over
192,803
189,780
186,758
183,735
180,713
177,691
174,668
171,645
168,623
165,600
162,578
159,555
154,968
150,440
145,972
141,566
137,220
All HD
Vehicles
481,120
459,824
438,982
418,594
398,660
379,181
360,155
341,583
321,966
302,854
284,255
266,161
247,261
227,909
210,799
192,612
181,080

-------
                                -39-


                            References

1.   Clean  Air Act  as  amended  in August,  1977, Section 202(b)
2.   Clean  Air Act  as  amended  in August,  1977, Section 202(a)
3.   "The Impact of  Future  Diesel  Emissions  on the Air Quality of
     Large  Cities,"  PEDCo  Environmental,  Inc.  for  EPA,  February
     1979, EPA.

4.   "Air Quality Assessment  of Particulate  Emissions from Diesel
     Powered Vehicles,"  PEDCo Environmental,  Inc.  for EPA, March
     1978, EPA-450/3-78-038.

5.   "Assessment of  Environmental  Impacts   of  Light-Duty  Vehicle
     Dieselization  (Draft),"  Aerospace  Corp.  for DOT, March 1979.

6.   Szigetly,  William, "Will Diesels  Dominate?," Fleet Specialist,
     Chilton, May/June, 1979.

7.   Summary and  Analysis  of  Comments  on the  Notice of Proposed
     Rulemaking for  the  Control of  Light-Duty  Diesel Particulate
     Emissions   from  1981  and  Later  Model   Year Vehicles,  MSAPC,
     EPA, October  1979.

-------
                                -40-


                            Chapter IV

                     STANDARDS AND TECHNOLOGY

A.   Introduction

     The statutory authority for proposal of this  heavy-duty diesel
particulate  regulation  is Section  202(a) (3) (A) (iii)  of the  Clean
Air Act as amended in August  1977.   It  requires  that  "The  Adminis-
trator  shall prescribe  regulations .. .applicable  to  emissions  of
particulate  matter  from  classes  or  categories  of vehicles  manu-
factured during  or  after model  year 1981  (or  during any  earlier
model  year  if practicable).  Such  regulations  shall  contain  stan-
dards  which reflect  the greatest degree  of  emission  reduction
achievable  through  the  application  of  technology which  the  Ad-
ministrator  determines  will  be  available  for  the model  year  to
which  such  standards  apply,  giving appropriate considerations  to
the  cost of  applying  such  technology with  the period  of time
available to manufacturers and to noise,  energy,  and safety factors
associated with the  application of such  technology.  Such standards
shall  be  promulgated and shall  take effect  as  expeditiously  as
practicable taking  into account the period  necessary  for com-
pliance."  Based  on  the  above  edict, a  standard of  0.25 grams  of
particulate per brake horsepower-hour (g/BHP-hr)  (0.093 grams  per
megajoule (g/MJ)) is  being  proposed  for heavy-duty diesels begin-
ning with the 1986 model year.

     In order to determine typical particulate  emission  levels  from
existing and  future  heavy-duty  diesels,  a  test  program is  being
conducted by EPA  at the Southwest Research Institute.  Table  IV-1
lists the 23 engines included  in  this program and  Table IV-2  shows
findings from  the heavy-duty  transient  cycle  tests  conducted  to
date.  Although the test  program  is not  complete  at this time,  EPA
believes a sufficient  number of engines have been  tested to estab-
lish  a representative  range of  heavy-duty  diesel particulate
emissions .

     In this  chapter, several means of achieving  reduced  particu-
late emissions  (engine  modifications  and after-treatment  devices)
from heavy-duty  diesels  are  discussed  followed by a section  ex-
plaining the rationale behind the proposed level of control.  Much
impetus is  placed on  trap-oxidizer technology since  it is currently
viewed as the most promising  means of  obtaining  substantial  partic-
ulate  control  and a  technique applicable  to  all  engines.   Col-
lection efficiencies  of up to  84 percent have  been  reported  for
prototype trap designs  and  research  is  well under way by  several
firms to refine trap-oxidizers  for  heavy-duty diesel  applications.
B.   Trap-Oxidizers

     Several approaches  to diesel  exhaust  particulate  control exist
today.   In addition to engine modifications,  to  be  discussed later,

-------
                           -41-



                        Table IV-1

         Heavy-Duty Diesel Test Program Engines

        Caterpillar -    1978 3208 Dina Family 3
                         1979 3208 Dina Family 3
                         1979 3208 EGR Family 13
                         1979 3406 DITA Family 16
                         1979 3406 PLTA Family 10

        Cummins -        1976 NTC-350
                         1979 NTCC-350
                         1979 VTB-903 Coach
                         1979 BIG CAM NTC-350
                         1979 NTC-290
                         1979 NH-250

        Detroit Diesel - 1978 6V-92T*
           Allison       1978 8V-71N Coach*
                         1979 6V-92TA lOg
                         1979 6L-71T
                         1979 6V-92TA 6g
                         1979 8V-71TA
                         1979 V8-8.2

        Deutz -          1979 F5L-912

        International -  1979 DTI-466B
           Harvester     1979 DT-466

        Mack -           1979 ETAZ(B) 673A
                         1980 ETSX 676-01
These engines are scheduled to be discontinued by 1982.I/

-------
                                 -42-
                            Table  IV-2

      Heavy-Duty Diesel Test Program Results - Transient  Cycle
Emissions
Particulate
Engine* g/BHP-hr
1)
2)
3)
4)
5)
6)
7)
8)

9)
10)

11)
12)
13)
14)
15)
16)
17)
1978
Caterpillar 3208**
1979 Caterpillar 3406
(Family 10)
1979 Caterpillar 3406
(Family 16)
1976
Cummins NTC-350**
1979 Cummins NTC-350
"Big Cam"
1979
1979
1979
#1
#2
1978
Cummins NTCC-350**
Cummins NTC-290
Cummins VTB-903
Fuel**
Fuel
DDA 6V-92T**
0.79
0.
0.
0.
0.
0.
0.

0.
0.
0.
,37
,52
60
40
39
58

31
37
54
NOx
g/BHP-hr
5.84
5.
8.
8.
7.
4.
8.

5.
6.
7.
,40
41
51
43
91
28

58
33
12
Number
of tests
7
2
2
4
2
2
2

2
3
2
1978 DDA 8V-71N**
#1
#2
1979
#1
#2
1979
1979
1979
1979
1979
1980
Fuel
Fuel
DDA 6V-92TA 6 g.**
Fuel
Fuel***
DDA 6V-92TA 10 g.
DDA 8V-71TA
IHC DTI-466B**
IHC DT-466
Mack ETAZ(B)673A
Mack ETSX-676
Oo
0.
0.
0.
0.
0.
0.
0.
0.
0.
69
79
48
55
54
38
36
53
58
63
5.
5.
5.
5.
8.
7.
5.
5.
6.
5.
33
69
82
83
69
32
56
90
73
15
2
2
2
3
2
2
2
2
2
2
*    Engines operated on #2 fuel except where noted,
**   Bagged NOx.
***  Particulate  mean based on 2 tests.

-------
                                -43-
research  has  also  focused  on  add-on particulate  collection de-
vices.   General  Motors and others,  in their  effort  to develop a
particulate control  strategy  for  light-duty  diesel  engines, have
investigated  the  feasibility  of   several  such devices.4V    These
include disposable traps with paper filters, traps requiring  owner
servicing, and traps  with  regenerative capabilities.  Although each
warrants  further  investigation,  the regenerating traps, (trap-
oxidizers) are currently  the  most  promising and  will  be 'the only
design discussed  here.

     A  trap-oxidizer  basically consists  of  a trapping  substrate
(such as  a metal  or  fiber mesh or a  ceramic monolith) housed in a
stainless  steel  shell  designed  to last  throughout  the  vehicle's
lifetime.   Placed in  the exhaust  line,  a trap-oxidizer collects
particulate  and  incinerates  it  on-board,  eliminating disposal
problems.  Significant backpressure build-up, due to the  collected
particles clogging  the substrate's  air  passageways,  should be
avoided  if incineration  occurs  prior  to substantial particulate
accumulation.   The general  consensus  is that the minimum tempera-
ture  required for combustion  of  the  particulate is approximately
450-500°C.  Because such high•temperatures are not always found in
heavy—duty  diesel exhaust, research  has  been initiated to both
periodically raise  the  exhaust  gas temperature  and  lower the
particulate oxidation  temperature.   In addition to these  measures,
the  use of exhaust  insulating  features,  such as port  liners and
insulated manifolds will reduce heat loss  at  all  times, effectively
raising the exhaust gas temperature.

     Current  trap-oxidizer research centers  on both continual and
periodic  oxidation designs.   Efforts  to develop  a continually-
oxidizing  trap have often involved the application of catalysts to
the  trapping  media  in order  to   lower the  particulate  oxidation
temperature.   Periodic oxidation  involves routinely regenerating
the  trap  by  artificially raising  the exhaust temperature period-
ically  to levels  that foster particulate combustion.

     General  Motors  has  suggested two means  to elevate exhaust
temperatures:   air intake throttling  and the use of  an external
heat  supply.47   Throttling  increases exhaust  temperatures by
restricting  the   air  intake,  thus  increasing the  fuel-air   ratio
and reducing the  dilution  air  available in the  engine.  GM reported
that  the  collection  efficiency  actually increased slightly  over a
1,000-mile load-up and  incineration test  when throttling was used
to initiate incineration every 100  miles.4/

     An  electrical heating  element is also  a  potential source of
the additional heat needed to  incinerate the  collected particulate.
Such  a  technique  could employ a dual  path trap with dual heating
elements and a flip valve  which  would  route a  small fraction of the
exhaust flow to  the side being incinerated and  the rest  to the side
currently  trapping.4V   The  dual path   design has  the advantage of
only requiring a small  amount  of energy to heat the exhaust,  since
only a small fraction of the exhaust is  actually being heated.  Its

-------
                               -44-


disadvantage is  high trap  cost, since  two  full  traps  are es-
sentially needed.

     Traps using  periodic regeneration  would probably  require  a
control  unit  with a  number  of sensors monitoring  various engine
parameters in  order  to  regulate  the  regeneration  process.   The
control  unit  would  detect backpressure build-up  or mileage  since
the  last  burn-up  and trigger  the  incineration process  at desired
times. The control  unit  could be very  similar to those currently
used  for  feedback carburetion control on gasoline engines or  could
possibly be much simpler  in design.

     A trap-oxidizer  system,  of either the  continual  or periodic
regeneration  designs,  would necessitate  durable exhaust piping from
the engine to the trap capable of lasting throughout the vehicle's
useful life,  estimated to be  475,000 miles  or 9 years._5_/  This is
necessary to ensure  that  all the exhaust is  being  routed through
the trap  and not  escaping  through  holes and cracks  in the system.
If  piping  upstream  of the trap  were made  of materials  unable to
last  throughout  the vehicle's (and  the trap's)  lifetime, the
opportunity  to remove  the trap-oxidizer during exhaust  system
servicing would become more of  a  possibility.   Stainless steel is
the  usual candidate  for such applications  and  experience  with
catalyst-equipped  light-duty  vehicles has shown  that  it  will last
the life of the vehicle.

     Thus, a  possible trap-oxidizer system would consist of stain-
less  steel exhaust  piping upstream  of  the  trap,  port  liners and
insulated exhaust manifolds,   the  trap  itself with stainless  steel
shell  and  trapping  media, and  an auxilliary  heat  supply  or air
intake throttle with the  appropriate  control  logic and sensors.  As
mentioned  above  a  catalyzing  material could  potentially provide
continual  particulate incineration.     Such  a  system  conceivably
would not  require the control device associated with periodically
regenerating  traps.

     As  was  the  case  with light-duty diesels,6/ the trap-oxidizer
system should be able  to  function properly throughout the vehicle's
life.   It is acknowledged that  the Class  VII  and VIII  diesels
travel  more  miles  over  their life  (475,000) than the  lighter
heavy-duty classes  and  light-duty diesels  (120,000 and  100,000
respectively) .J^/6/   However,   the  usage characteristics,  such  as
longer trips,  steadier  operation,  etc., of  these largest diesels
should be much  less  stressing on the  trap-oxidizer  on  a per mile
basis  than those  associated  with  lighter  class  diesel  operation.
These advantageous usage patterns coupled with the stainless  steel
construction  of trap-oxidizers should make them capable of filter-
ing exhaust particulate efficiently throughout the vehicle's  life.

     Among the  corporations   actively   pursuing  trap-oxidizer de-
velopment are Corning,  Texaco,  Engelhard,  Matthey-Bishop, and
Imperial Chemical  Industries  Limited  (ICI).  Table IV-3 provides an
overview  of  initial  collection efficiencies  reported   for   their

-------
                                   -45-

                            Table IV-3

                    Trap-Oxidizer Test Results*
Effect on Brake-Specific Emissions(%)
Trap
Texaco A-IR I/
Englehard CST-1 coating
on Texaco trap I/
ICI-Saffil 21
Corning Ex-20 2/
Matthey-Bishop** 7/
Matthey-Bishop*** 8/
Particulate
-59
-49
-32
-84
-58
-61
NOx
0
-1
9
-4
-
-8
HC
-60
-92
2
-48
-90
-90
CO
. -4
-99
-11
-17
-
-94
*     Tests  were  conducted on light-duty diesel  vehicles  over the
FTP cycle, except where indicated.

**   This Matthey-Bishop trap-oxidizer was tested by Matthey-Bishop
on a  taxi for 2,000  miles;  its  efficiency was  tested  at  600-mile
intervals.   All  tests  of other  traps were  conducted  by  EPA and
reflect zero-mile collection efficiencies.

***   After 600 miles, particulate  emissions  increased  by  approxi-
mately 200 percent of baseline levels.

-------
                               -46-
respective  designs.   Nearly  all  of the research performed  so  far
on  the  use of  trap-oxidizers  on light-duty diesels  should  be
applicable   to  heavy-duty  diesels.    However,  heavy-duty  diesels
typically  are  subjected  to  different  operating  conditions  than
light-duty  diesels.    Heavy-duty diesels,  for  example,  are  often
left running for several hours while  the operator rests or  eats a
meal.   These long periods  would not be  conducive  to  trap-oxidizer
regeneration  since the  exhaust temperature is very low during
idling.   Also,  the effect  of frequent high-load operation  which
is  characteristic  of  heavy-duty  diesel operation  could  require
improvements over the  light-duty design.   Thus,  some  additional
leadtime beyond  1985  appears  appropriate for heavy-duty  trap
development.    One extra  year plus the leadtime  remaining  after
promulgation of the heavy-duty diesel  particulate standard  should
be  sufficient  to  optimize  trap  designs for  the  larger  diesels.

C.   Engine Modifications

     A large number  of  engine design and operating variables  could
conceivably affect particulate emission  levels.   Among these  are
timing,  load,  speed,  combustion chamber  design,  fuel injector
design and  orientation, injection pressure,  and  turbocharging.   In
a report prepared by Southwest Research  Institute (SwRI)  for  EPA,
the effect of  several  of these variables on heavy-duty diesel
particulate emissions   was  investigated.9/   Table IV-4  shows  the
effect of  timing,  EGR,  and   indirect   versus  direct   injection  on
the particulate  and  NOx  emissions  of  a  Caterpillar 3406  engine
described in Table IV-5.   Noteworthy  particulate reductions of  23
percent  due to  a 5 degree  timing advance and 21 percent by  indirect
injection were  found.   Although these tests  were run  on the 13-mode
cycle and  not  over  the more  representative transient cycle,  they
indicate the potential  impact of such parameters.

     In   order  to  evaluate the  effect  of turbocharging on  diesel
particulate emissions,  Southwest  chose  a Daimler-Benz OM-352
naturally-aspirated  and an OM-352A  turbocharged engine.  As  can  be
seen from Table IV-5, these engines are quite comparable  except  for
the turbocharging.   The  OM-352  emitted  0.991g/BHP-hr  (0.369  g/MJ)
of  particulate over  two  tests  while  the turbocharged  OM-352A
emitted   an  average  of  0.562 g/BHP-hr   (0.209  g/MJ).9/  Thus,  the
turbocharged  version  emitted 43 percent  less particulate.   It
should be  noted that there is no way to evaluate the effect  that
turbocharging  alone had  on  this emission reduction  since  the
addition of a  turbocharger also  required related  engine modifica-
tions and  adjustments to optimize performance.   For example,
injection  pump  recalibration  and  timing adjustments, two  modifi-
cations  deemed  necessary when converting a  naturally-respirated  to
a turbocharged  engine,10J  will themselves affect emissions.   The  43
percent   reduction  should, therefore,   be interpreted as  the  net
effect of  turbocharging  and  associated adjustments.   However,  as
these modifications  always accompany  turbocharging,   the reduction
can be said to be essentially due  to  turbocharging.   Since  turbo-
chargers are  in wide  use on today's  heavy-duty fleet,  minimal

-------
                         -47-
                      Table IV-4
      Summary of Emission Reduction Potential of
Selected Engine Modifications (Based on 13-Mode Cycle) 9j
Modification
5" timing advance
10° timing retard
EGR
Indirect injection
Effect on Brake-Specific
Particulate
-23
191
166
-21
Emissions (%)
NOx
42
-46
-44
-47

-------
                                     -48-
                                   Table IV-5

                         Description of Heavy-Duty Diesel
                  Engines Used to Evaluate Engine Modifications 9/
Engine Make
Engine Model
Engine Serial
No.
Strokes/Cycle
Cylinder
Arrangement
Displacement
(Liter)
Compression
Ratio
Type Aspiration
Rated Speed (rpm)
Power at Rated
Speed (kw)
Peak Torque
Speed (rpm)
Peak Torque (N-M)
Typical Appli-
Mack
ETAY(B)673A
6F4310
4
1-6
11.01
14.99:1
TC a/
1900
235
1450
1423.8
1C b/
Caterpillar
3406 c/
IA5484
4
1-6
14.63
14.5:1 (16:1)
TC £/
2100
242
1200 (1400)
1375 (1319)
1C b/
Daimler-Benz
OM-352
936-10-125488
4
1-6
5.67
17.0:1
NA a]
2800
96
2000
361
U b/
Daimler-Benz
OM-352A
935-10-01-9653
4
1-6
5.67
16.0:1
TC a/
2800
108
1800
415
U b/
  cation
Typical Fuel Type    DF-2
DF-2
DF-2
DF-2
a/   TC-Turbocharged,  NA-Naturally Aspirated.
b/   IC-Intercity Truck,  Tractor;  U-Urban Truck and Truck-Tractor.
cj   Items in ( ) are  for indirect injection configuration.

-------
                                -49-
research should  be  involved for those manufacturers  choosing  this
method to reduce  particulate  emissions.   However, it is  also  only
available  to  those  vehicles  without  turbochargers.     While  the
leadtime necessary to turbocharge a given naturally-aspirated
engine can  be a  number  of years,  many  manufacturers may  already
have  investigated the turbocharging  of  their  naturally-aspirated
engines  and there may be  time  available before  1986 for manufac-
turers to use this control strategy if they so desire.

     Southwest Reseach also investigated  the reduction potential of
a new  high-pressure injection  system  being developed by American
Bosch,  ll/  The  test engine  was a Mack  ETAY(B) 673A,  which  is
described in Table  IV-5.   Particulate  emissions were  obtained  from
a standard Mack engine which had been run for 1,000 hours, the  same
engine with a new  standard injection  system,  and the same  engine
with a  new high-pressure  injection system.   As was   the  case  with
turbocharging, engine  adjustments  and  modifications were  needed  to
optimize high-pressure injection performance.  However, since again
these  adjustments would  always accompany  high-pressure  injection,
any  effect  can  be  attributed  to  high-pressure  injection  itself.
The  results  outlined  in  Table IV-6  lead  to  several conclusions.
First  and  foremost  are   the  50 percent  particulate  emission  re-
duction  by the high—pressure system compared  to the  1,000-hour  old
standard pump  and the 55  percent  reduction of this   system  versus
a new  standard pump.   Second, these  results  were accompanied  by
increased  fuel  economy,  3.7  percent  relative to the   1,000-hour
standard pump  and 1.1  percent  relative to a new  standard pump.  A
third  result  is  the demonstrated lack of deterioration  in the
particulate emission  rate  due  to  injection system  deterioration.
The  experimental high pressure injection system also  caused a
34 percent increase in NOx emissions.

     There is  also  evidence that  basic engine  modifications, often
made for reasons  other than particulate control, can  in fact  have a
beneficial effect on  particulate emissions.   Two  examples  involve
Cummins  and Caterpillar engines.  First,  two improved  versions  of a
1976  Cummins  NTC-350  engine were  tested  at  SwRI, along with  the
older version.  One of the newer engines  was the California version
(NTCC-350) and the  other  was  the 49-state version (NTC-350).   The
emission results  of all three  engines  are  shown in Table  IV-2.   As
can be seen, both newer engines showed marked reductions  in  partic-
ulate  emissions,  35 and  33 percent, respectively;  NOx emissions  of
these newer engines decreased at the same time by  42  and  13  percent
respectively.   It is also important to point out that  fuel consump-
tion  of the  newer versions  decreased by  7.9  and 4.2 percent
respectively._12_/  This is  an indication that engine modification can
be optimized to  simultaneously  reduce  particulate  and NOx (a point
which will become more readily  important in  later  portions of  this
chapter), without adversely affecting fuel  economy.

     Second,  Caterpillar  has  submitted  data on a 3406  engine
which was  redesigned  for NOx  control ._13_/   There  were a  number  of
improvements made to the  engine, but the  most notable  were separate

-------
                              -50-
                           Table IV-6

           Effect of High-Pressure Injection on Emission
  and Fuel Consumption (2-test averages, based on 13-mode cycle) 9/

                               Particulate     NOx     Fuel Consumption
   Engine Configurations        g/BHP-hr    g/BHP-hr      kg/BHP-hr

High pressure A.  Bosch pump       0.30         9.0           .181
     (10° BTC)

Standard pump - 1000 hours        0.61         6.61          .175
     (21° BTC)

Standard pump - new               0.67          -            .179
     (21° BTC)
     NOx as  N02 by non-dispersive infrared.

-------
                                -51-
circuit aftercooling, high  pressure  injection,  and a small  degree
of retarded timing.   While  NOx emissions were reduced 20  percent,
particulate emissions  (as estimated by  smoke) were reduced 50
percent.   While  smoke  measurements  do  not  always correlate with
particulate emissions,  there  are  two  factors in  this  case which
would support the  smoke  reduction  as  a reasonable indication of a
particulate reduction.  First,  only  one  engine  was involved and a
correlation of smoke and particulate  was  already  available  for that
engine.  Second, the instantaneous  volumetric flow rate was  coupled
with the smoke  reading  (which  is a form of  concentration  measure-
ment) to give the overall smoke measurement more  of a mass  emission
orientation.   Thus,  while  these  data  cannot be  used  to  strictly
state that particulate emissions were reduced 50  percent, it  can be
said  that  particulate  emissions  decreased  and  very likely by a
large  amount.   One  of  the  more notable aspects  of  these  data is
that the unmodified  Caterpillar 3406 is  already one of the  cleaner
engines tested by  SwRI  (see Table IV-2)-   These  data are evidence,
then,  that reductions via  engine  modifications  can be  made even
beyond the lowest values shown  in Table IV-2.

     The leadtime  necessary to make  such modifications  would vary
from engine  to  engine.   In the  two  examples  cited above,  the
modifications  were  occurring  for  reasons  other  than  particulate
control and will be  implemented by  1986 with or without the promul-
gation  of  a particulate standard.   This may not be  the case with
other  engines  and those design features beneficial to particulate
control will have  to be  incorporated specifically  for that reason.
However, it should be true  that most  if not  all heavy-duty  diesel
engines will be undergoing some degree  of redesign  in the next five
years.   The drive for fuel economy  is forcing improvements in
existing  engines  as  well  as  opening  up new markets  for  on-road
diesel engines.   The  latter  should  result in new families of
diesels being designed,  as well as  the  modification of old  designs,
to  power   vehicles  traditionally equipped  with  gasoline  engines.
Due  to this degree of redesign  already  occurring,  it should be much
easier  to  incorporate the necessary design  changes for particulate
control than it would be if existing  designs were not changing over
the  next  5-10 years.  Thus,  it can  be expected  that most engines
will be  able  to incorporate  the  necessary   changes  by  1986.

     In summary,  several  engine modifications have  been shown to
reduce  particulate  emissions by varying degrees.  Advanced  timing
reduced particulate  emissions by 23 percent,  indirect injection by
21 percent, turbocharging by 43 percent and high-pressure injection
by  50 percent.   The  leadtime  necessary  to incorporate timing
changes and turbochargers should be available before 1986 since the
former  is  a relatively simple  adjustment  and the latter  involves
technology  which  is  readily  available.    The Bosch  high-pressure
injection  system   is  still  in  the developmental  stages  and will
require more time to refine. Nevertheless,  such  a  system should be
available as a particulate control  device for use by the  1986 model
year fleet.  Indirect  injection  would require redesign of the
engine, implying significant effort in order  to implement.    As can

-------
                              -52-
be  seen from  Table IV-4,  indirect  injection  not only  reduces
particulate emissions but  NOx  emissions  as well.  These reductions,
however, are usually  associated with a  fuel  economy penalty; 7.5
percent in this test case._9_/

     Engine modifications  incorporated on newer production engines
and from prototypes also  indicate  that particulate emissions can be
reduced without raising NOx emissions. These are probably the most
promising modifications  as  they have already been practically
demonstrated and have  occurred  for reasons  other than particulate
control.  The  latter factor would imply that  these design modifi-
cations would have other benefits connected  with them  besides
particulate reduction.  With the large amount of redesign occurring
in  the  industry at  this  time, these types  of modifications should
be  able  to  be  incorporated by  1986  on  most,  if not all, engines.

D.   Particulate-NOx Relationship

     Section 202(a) (3) (A)( ii)  of  the Clean Air  Act  calls for a 75
percent reduction  in> NOx  emissions for  heavy-duty vehicles (based
on  uncontrolled gasoline  engines).  This  requirement  is relevant
to particulate  control  since,  as discussed in the previous section,
certain  engine modifications   which reduce  particulate  emissions
also increase NOx emissions and vice versa.   Also evident, from the
preceding  section  is  the  fact  that not all  engine modifications
improving particulate emissions have  a  deleterious  effect  on NOx.
Figure IV-1 demonstrates that engines can be built which  emit
relatively  low amounts  of both NOx and  particulate.    As  can be
seen,  at  least four engines  produced by  three  different manufac-
turers have both very low  NOx  and  particulate emissions.

     Since  EPA is   required to  regulate both particulate  and NOx
emissions from heavy-duty  diesels,  and  as  mentioned above certain
control  techniques  which  lower particulate  also raise  NOx   emis-
sions,  the  method  used  to set  the  particulate  standard  should be
designed to affect  the achievability of the mandated NOx reduction
as  little  as  possible.   In  this way,  manufacturers will  not be
placed in  the unfair position  of  complying  with two Congres-
sional  mandates, which  separately may  be  achievable but together
are not.   Thus,  in arriving at the level of the particulate  stan-
dard (a discussion  of  this topic follows this section) EPA will not
consider  the  potential  particulate emission  reduction  of   those
engine  modifications  which could have  an  adverse   impact  on NOx
emissions.   It  should be pointed  out  at this  time that  trap-
oxidizers do not adversely affect  NOx emissions  and  are affected by
the above-mentioned restriction.

E.   Rationale  for  Level of Control
     When determining the  level  of  control,  several  factors must be
considered.   These  include  1)  the  degree of reductions achievable
from existing levels; 2) emission deterioration over the vehicle's
useful  life;  and  3)  the  10  percent  Acceptable  Quality  Level

-------
                        -53-

                  Figure IV-1
        NOx  vs.  Particulate Emissions
           from Table IV-2 Engines
  NOx
Emissions
(g/BHP-hr)
9.0
8.0
7.0
 /- n
 b.U
 5.0
 4.0
 3.0
       0.2


                               0.3
0.4
                                      '   .12
                                        3
                                              ,16

                                                   17
0.5
0.6
0,7
                                                                    0.8
                        Pareiculate Emissions
                             (g/BHP-hr)

-------
                              -54-
(AQL) of  Selective  Enforcement  Auditing (SEA), which includes  the
effect of production line variability.

     Before these factors can be applied, however, a decision must
be made as to the baseline emission  level  from which  the reductions
should be  taken.   This  topic is dealt  with in the  following sub-
section and the 3 points  listed above are discussed  in the  subsec-
tion which follows it.

     1.   Baseline Level

     Section 202(a)(3)(A)(iii) requires  the  Administrator  of EPA to
set  a  particulate standard for heavy-duty  diesels which reflects
"the greatest degree of emission  reduction achievable through  the
application of  technology which  the Administrator determines will
be available  .  .  . ."   Several  options have been considered with
regards to this  edict,  any one of which  could  conceivably  provide a
baseline  from which to  set the  standard.   They  include:   1)  the
worst engine (highest  particulate emission level  from the  Southwest
test  program);  2)  the   lowest  particulate  emission level  of  the
tested engines;   3)  the  highest  emission level among each manufac-
turer's best engines; or  4) the  average emission level of  the  set
of  engines  which  includes  each  manufacturer's  best  engine.

     As can be seen from  the above options, a  wide range  of levels
could  conceivably  be  chosen  as  the baseline  from  which  to  take
reductions.  An  examination of  the options  and  the data in Table
IV-2 reveals  that Option  1 would  result in  the  highest baseline
level,  0.79  g/BHP-hr  (0.29  g/MJ),  and  Option  2 would  result
in the lowest level, 0.31 g/BHP-hr  (0.12 g/MJ).  The  extent  of this
range makes  it  clear  that the determination  of  the proper "base-
line"  level  from the test data  contained  in  Table  IV-2  also  in-
cludes an  evaluation of the  control technology inherently  present
in each of the engines  shown in Table IV-2.   As  such, the question
of which  option  best  conforms  with the  requirements  of  the Clean
Air  Act  includes more,  than  that  of  the  typical baseline,  but
includes discussion of  control  technology  as  well.   This  control
technology can  be distinguished from that  to be discussed later,
when further reductions  are  taken  from the  baseline,  by the fact
that  it  is already present  on existing  engines.   The  control
technology to be  discussed later  will  consist of new devices  and
techniques, both engine-related  and exhaust-related, which  are  not
generally  in  use today.   Thus, when  determining this "baseline"
level,  the  mandate  of  the Act  to  achieve  the greatest   reduction
that  is  technologically  feasible  is  just  as applicable  as  when
determining the  reduction potential  of trap-oxidizers and  the like.
With this  in  mind,  a discussion of the four options will be pre-
sented below.

     Referring to Table  IV-2,  a standard based  on  Option 1 would
reflect the  reductions  achievable   from  the relatively small 1978
Caterpillar 3208's level of ,0.79 g/BHP hr  (0.29 g/MJ).   This
particular engine emits  by far  more particulate  than its counter-

-------
                                -55-
parts  (14 percent more  than the  second  highest engine).   Any
standard based  on this level  could still reflect  the greatest
degree of  control  available from new technology, but  would  ignore
the  demonstrated  potential  of  other  existing  engine designs  to
reduce particulate  emissions  by more  than 50  percent.   It  seems
clear  from the wording of  the Act,  that  this .demonstrated  tech-
nology is  to  be  included  in determining the  level  of  the particu-
late standard.   This  would  require the rejection of  0.79 g/BHP-hr
(0.29 g/MJ) as a viable baseline level.

     It  is  possible that not  all  heavy-duty diesel engines  would
have the same inherent potential for low particulate emissions.   If
a  certain  type of  engine  (e.g.,  a bus  engine)  or a  certain  size
engine (e.g.,  relatively low  power) had  inherently  higher particu-
late emissions than the others and  this  type  or  size  of  engine  was
necessary to  the  market,  then  some allowance may  be  in order.
Certainly, this  is the case  with use  of  No.  1  diesel   fuel.   As
indicated  in  Table IV-2 and  supported  by  past literature, use  of
No.  1  diesel fuel, as  opposed  to use of  No.  2 diesel  fuel, will
reduce  particulate emissions  10-20 percent.   Thus,  it would  be
inappropriate  to use  test  results on  No. 1  fuel  to demonstrate
feasibility  for  another engine  required  to use  No.  2 fuel.   And
while  this particular  problem  does not apply to  the Caterpillar
3208,  this particular  engine is  a relatively  small  engine with
respect  to power  (210 hp) and  it could  be  possible that small
engines have  inherently higher emissions than  larger,  more powerful
engines.   To evaluate  this possibility, the  particulate  emissions
of  those engines  shown  in  Table IV-2  were plotted against  their
maximum  power  outputs  in Figure  IV-2.   As  can be  seen, particulate
emissions  appear to have no correlation at  all with  engine size  and
all  sizes  would appear to  have  nearly  the same potential for  low
particulate  emissions.   Thus, no  allowance appears necessary  for
engine size.

     It  could also be  possible  that  the particular design of  the
3208 may lead to higher particulate emissions  and  some significant
changes  in its design would be necessary to reach  the  lower partic-
ulate  levels  of  other  engines.   The requirements  of Section  202(a)
of the Act would still require a standard based on the  lower  levels
of  demonstrated  technology.   However,   the  provisions of Section
206(g) of  the Act,  providing  non-conformance  penalties for engines
of this  class, would  apply  very  appropriately  in  this  case.   Under
these  provisions,  engines  not meeting  an  emission standard  could
still  be sold if  a pre-determined fee  were  paid  for each  engine
sold.   The fee  schedule  would be  designed to encourage  manufac-
turers  to  meet  the  standard  as  soon as possible.   Thus, even  in
this case, Option 1 should be  rejected.

     The logic of the above argument could  be  applied  to  the  second
highest  emitting engine,  the  third highest,  etc.  until the only
engine  left  would  be  the  best engine  tested so far, the Cummins
VTB-903  at 0.31  g/BHP-hr  (0.12 g/MJ).   this  is essentially  Option
2.   A problem  arises  with this  particular engine because it  was

-------
                     -56-
                Figure IV-2

  Engine Size vs. Particulate Emissions
 (Numbers beside points refer to engine
      listing order from Table IV-2)
500
400
Engine Size

  (rated
Horsepower)
300
200
                                     .3
                      ,2
                                             .4
                8
                                  11
                     14
                                      15
  0.2
                                   0.3
0.4
0.5
0.6
                                                       10
                                                      0.7
0.8
            Particulate Emissions
                 (g/BHP-hr)

-------
                                -57-
tested using  No.  1  fuel.   As discussed  earlier,  this particular
datum  should  not  be  used in  setting  the baseline.   Instead,  an
engine tested on No.  2 fuel  should  be used  and  this could easily be
done.  However, the arguments that would apply for or  against that
engine also apply  to  the VTB-903  and for simplicity it  will be used
as the example.

     In  order  for Option 2  to  be acceptable,  the determination
would  have to be made that all  engines  (or  nearly  all  engines
when  nonconformance  penalties  are considered)  could  incorporate
all  of the pertinent design  features of the best engine.   Ab-
solutely no allowances would be  made for  engine type,  size  or
manufacturer differences.  One might  say  that  the only guaranteed
way  to ensure reaching  that level  would  be to  copy the  best
engine.   While Figure IV-2 shows no real  relationship  between
particulate emissions and engine size and Table IV-2 shows  no
discernable difference between truck and bus engine emissions, the
data  are  simply not  strong  enough to  demonstrate that there  is
absolutely no effect  in this  area.  This option would also leave no
room for differences  between  manufacturers, even if their effect on
particulate emissions  were  quite  small.   From all this  it  would
appear that Option 2  would go beyond the mandate of the Act and set
a standard that may not be achievable by a sizeable portion of the
industry.  Therefore, it should be  rejected.

     One solution to  the  problems  of  Option  2  would be to include
manufacturer  differences  into  the methodology.   In  the  extreme,
rather than a  baseline set by  the  best engine  of those tested, the
baseline would be set by the  highest-emitting engine from among the
best  of  each  manufacturer.   This  is  Option 3 and  would base the
standard  on  Mack's  1979  ETAZ(B)673A  which  emitted  0.58  g/BHP-hr
particulate.   Upon examining Table" IV-2,  it  is apparent  that this
engine's particulate  emission level is well above that  of the other
manufacturers' best engines'  (57  percent higher than the average of
the  other  low-emitting  engines)-   Indeed,  12 of  the 17 engines  on
Table  IV-2 are already at or below this level.  While this option
attempts  to  take  manufacturer differences into  account,  it  would
appear to  go  too  far and ignore  the possibility that  the manufac-
turers  of  the higher-emitting diesels  could produce  engines like
those  of their competitors.   It would seem  impossible to argue that
this  level represented the lowest  achievable level when two-thirds
of  the existing  engines  could  do better,  particularly with the
possibility  of nonconformance penalties  being available.    Thus,
Option 3 should be rejected.

     The last  option  listed,  which would  average  the best  engines
of  each  manufacturer, appears to  be the  most appropriate  as  it
avoids  the  problems  associated with the  other options.    One,  it
avoids  basing  the standard  on a  single  engine  design.    Two,  it
still  appears to  comply with the "greatest  degree  of emission
reduction"  requirement of the  Clean Air  Act.   Following this
procedure, the standard  would  be  based  on further emission reduc-

-------
                            -58-
tions achievable from a 0.41 g/BHP-hr level.*  This is a stringent
level,  currently  achieved  by  only  six  of  the  seventeen  engines
tested so far and only  14 percent  higher than the lowest-emitting
engine  (on  No.  2 fuel).   Three,  it does  take  into  account manu-
facturer differences  by  basing  the  level  on an average (four of the
five  manufacturers'  engines are  below  the  average).    And four,
engines of  different  size  and  type  are included  in  the average.
This  should allow  for any slight  differences in inherent emission
levels due  to  these  factors.   A closer  examination  of each manu-
facturer  indicates  the  feasibility  of  the  0.41  g/BHP-hr  level.

     Caterpillar's 1979  3406  Family 10  is already below  this
level.  Their 3406 Family 16 is slightly above, but should be able
to incorporate  features  of the  Family  10  and  also be able to comply
with  the  proposed  standard.   While the  3208 model listed  is  well
above 0.41  g/BHP-hr  (tests  indicate  0.78 g/BHP-hr) the particular
engine tested was a relatively  old  1978  model.   It should also be
pointed out that another manufacturer's engine with the same rated
horsepower as the  3208's   has been  tested and  found  to  emit  less
than  half as much  particulate over  the  transient  cycle:   the
International  Harvester DTI-466B.   Therefore,  there  should be
no  inherent  reason  why  an  engine of  that  size cannot  reach  the
0.41 g/BHP-hr level.   As can be seen from Figure IV-2, there is no
discernable link between engine  size and  particulate emission
level.

     Three  Cummins engines tested  are below  the 0.41 g/BHP-hr
level.  Newer versions of the  1976 NTC-350, which was found to emit
0.60 g/BHP-hr,  were among those meeting the  0.41 g/BHP-hr require-
ment.   The  Cummins  engines  listed  on  Table  IV-2  further indicate
that  lower  particulate  emissions  can  indeed be  achieved  through
basic engine modifications.

     Of the  Detroit  Diesel  Allison (DDA)  engines  listed  in Table
IV-2, the 8V-71TA is  already below 0.41 g/BHP-hr, a second is close
at  0.48  g/BHP-hr  (the.6V-92TA  6g),  and  two of  those  listed  (the
6V-92T and 8V-71N)  are'l978 models  scheduled  to be discontinued by
1982.JY   EPA believes  technology already proven on the relatively
new  8V-71TA  should  enhance the  ability of  other  DDA engines  to
reach the 0.41  g/BHP-hr  level.

     Neither of  the  two Mack  engines  tested emit less  than  0.41
g/BHP-hr  particulate; their levels were  0.58 and 0.63  g/BHP-hr.
One of the Mack  engines,  however,  has  the  second lowest  NOx emis-
sions of  the  engines  tested.    This  should  aid  them  in  complying
with  any  future NOx  standard  (mandated by the  Clean Air Act).
Since  as mentioned earlier,  some  engine  modifications which
reduce NOx  also  increase  particulate,  e.g.,  retarded  timing,  Mack
may not heed to  rely  on  such techniques to  the same extent as other
*     This  value  reflects the average of those best engines tested
on No.  2  diesel  fuel  only,  since as mentioned  earlier,  only bus
engines can  certify using  the  lower  particulate emitting  No.  1
diesel fuel.

-------
                            -59-
manufacturers.    Thus  when the potential adverse effect on partic-
ulate emissions of NOx control is  considered, Mack's  engines may be
ultimately  in  a .more advantageous  position  than is now apparent.

     If after  a  good  faith effort,  their engines (or those of any
other  manufacturers) are not able  to comply with an  emission
standard,   nonconformance   could  be made available,  as  mentioned
above.

     To summarize Option 4, it would base the level  of the proposed'
standard  on an average  taken from  each manufacturer's  lowest
particulate emitting engine.   By requiring  dirtier engines to
become more like  the  cleaner ones before  reductions  from add-on
devices (trap-oxidizers)  are  considered,  this methodology complies
with requirements  of  the  Clean Air Act  that  the standard reflect
the "greatest degree of emission reduction achievable  . . . ."  The
0.41  g/BHP-hr   level  is  not  excessively  stringent  since  control
technology  exists today whereby  several engines have  already
reached this level (refer to Table IV-2).  By averaging the partic-
ulate  emission levels of the best engines,  Option 4  reflects  a
more  representative  range of  performance  capabilities  than  other
options based  on  the performance  of single engines  while  still
resulting  in a  stringent standard.  All  manufacturers  listed in
Table IV-2  have  at least  1 engine which  is below the 0.41 g/BHP-hr
level  except  Mack.   In  Mack's  case,  the fact that four  other
manufacturers,   each within their  own  design constraints,  have met
this level  should  be  sufficient  evidence that  the  ability to meet
this  level   is  not connected  to  some unique design  feature,  but
indeed can  be attained by all manufacturers.

     2.   Choice of Standard

     Now that the basic engine-out particulate emission level (0.41
g/BHP-hr) has  been established, the process of  choosing a standard
can  continue.   The  next  step is to  determine the greatest  de-
gree  of  reductions  achievable on  prototype  engines  by  applying
control technology not  commonly  found  on  current  engines.   These
techniques  fall  into two categories,  engine  modifications  and
exhaust  aftertreatment  (trap-oxidizers).   As  mentioned  earlier,
reductions  achievable from  engine modifications  which  adversely
affect NOx  emissions  have been excluded from the determination of
a  technologically  achievable  particulate standard.   This  was done
in order  to affect the  achievability  of the mandated NOx standard
as little as possible (refer  to the discussion  of the particulate-
NOx  relationship  earlier  in  this chapter).   In addition  to  fore-
going  this  type of control  technique,  other engine modifications
not commonly found in current engines  which do not adversely affect
NOx  (e.g.,   indirect  injection)  have  also been excluded  from the
methodology  used to set the  particulate  standard.  The forthcoming
NOx  rulemaking  will  consider the  potential of  such techniques and
propose a NOx  standard  which heavy-duty diesels can meet  while at
the  same  time  complying with  a  0.25 g/BHP-hr particulate standard.
Thus, the emission reduction  potential of  engine modifications not

-------
                            -60-
commonly  found  on  today's  engines  have  not  been applied  to the
basic engine-out particulate  emission  level  of 0.41 g/BHP-hr.  The
second category of control techniques not commonly  found on current
engines  is  exhaust  aftertreatment   (trap-oxidizers).   Particulate
reductions from this strategy have been included in the methodology
used to set the level of the standard.

     Based on  results  from prototype  trap-oxidizers,  EPA believes
60 percent efficient trap-oxidizers  will  be  available for applica-
tion on the 1986 model  year heavy-duty diesel fleet (see Section B
and  Table IV-3).   This level  of efficiency  is  the same  as that
determined to  be feasible  for  light-duty diesel  applications for
the  1985  model year fleet._6/  By  mathematically  applying  such  a
device  to the  0.41  g/BHP-hr  baseline, an  emission level  of 0.16
g/BHP-hr  is obtained. This value represents the lowest mean partic-
ulate emission level a  manufacturer  could achieve on new prototype
engines.

     In  order  to  estimate  the  production  line  mean from this
sample  (prototype)  mean when the standard deviation of  the popu-
lation is known,  the z  distribution can be used.  In equation form,
this is represented by:
Where:

     P  = probability
     Vi  = population (production line) mean
     "x  = sample (prototype) mean
     a  = degree of confidence
     zfl = z statistic
     a  = standard deviation of population
     n  = sample size

     Several of these factors deserve clarification.  The sample
prototype mean is 0.16 g/BHP-hr, as determined earlier.  The
degree of confidence that the equation within the brackets in the
above equation is true is represented by a.  A 90 percent con-
fidence level has been chosen for this application.  This level is
believed to be reasonable since not all engine families are audited
and a greater degree of certainty would likely be cost prohibi-
tive.  Based on this 90 percent degree of confidence, the z
statistic can be obtained from statistical tables such as Table A-l
of reference 14; and is 1.28.  No data are available which indicate
the standard deviation of production line particulate emissions (a)
so EPA has assumed it to be 12 percent of the population mean, as
is the case with regards to gaseous emissions from heavy-duty
diesels ._15/  Last, for the purposes of this study, a sample size of
3 was chosen.  This value represents the number of prototypes a
manufacturer might develop on a certain project.   Of course the
number of prototypes to be developed is at the manufacturer's
discretion.  It should be pointed out, however, that as n in-

-------
                                -61-
creases, z decreases; this has the effect of  enhancing  the  likeli-
hood  that  the prototypes  are  indicative of  potential  production-
line  performance.    Incorporating  the above  values,  a  production
line  mean based  on  a prototype mean of 0.16 g/BHP-hr would be
approximately 0.18 g/BHP-hr.

     The remaining factors to be  considered in  the  standard  devel-
opment  process deal with deterioration  (both engine and trap-
oxidizer) and  Selective  Enforcement  Auditing  (SEA).  A  20  percent
increase in particulate  emissions over a vehicle's  useful life due
to deterioration  in  trap-oxidizer collection  efficiency and  engine
wear has  been assumed.    In-use  data on General Motors  light-duty
diesels  show  negligible  increase in particulate emissions with an
average  48,000 miles  accumulated.16/  If this  were  also indicative
of heavy-duty engines,  it  would essentially leave  the full 20
percent factor  for  trap  deterioration alone.  Production line
engines  emitting  a mean  0.18  g/BHP-hr  would  thus,  at  the  end of
their useful lives, emit approximately 0.22  g/BHP-hr.

     To assure that 90 percent of his vehicles pass  SEA,  a manufac-
turer must allow a margin of 1.28 a  from the mean  production-line
emission level, where a  is defined as above.  Thus,  a manufacturer
would design his vehicles so that the mean  production  line parti-
culate  emission level is 0.03 g/BHP-hr below the standard.   Apply-
ing this factor to the deterioration corrected production line
mean, a  standard of 0.25 g/BHP-hr is determined.

     Engines  operating   on No.  1 fuel  (bus engines) may  have an
advantage with regards to  a  margin  for NOx  reductions  since  use of
this  fuel  usually results  in  lower particulate and NOx  emissions
but these engines  are only required  to meet a particulate standard
based on emission results using No.  2 fuel.  This  effectively gives
them  an additional control technique not available to  all  engines
(refer to earlier discussions  of  this topic in  this chapter).
Other methods which could be employed to reduce particulate  without
adversely  affecting  NOx  include  a  tighter  control of  production
line  variability,  technologically  improving the durability of
trap-oxidizer  and/or engine  components,  the  development  of more
efficient trap-oxidizers,  and by  relying on a  less  than  90  percent
confidence  that  a particular  engine family will pass  SEA.   This
latter  point  is  important  since,  if after a good  faith effort a
manufacturer's  engines  are  slightly  above  the 0.41 g/BHP-hr pre-
trap-oxidizer  level,  there can still be a  good chance  of  passing
SEA, only not with a  90  percent confidence.

     As  mentioned  earlier,  the  Agency is in the process of  devel-
oping  a standard  to limit  the  emissions   of  NOx  from  heavy-duty
engines  and  light-duty  trucks.   Since  certain control  techniques
which  lower  NOx  also raise  particulate  emissions and  vice  versa,
the possibility exists whereby heavy-duty diesels  could  be required
to meet  NOx and particulate  standards which alone would  be  techno-
logically feasible,  but  taken  together,  infeasible.  To avoid this
situation no particulate control techniques  which  also  cause a rise

-------
                            -62-
in NOx  emissions  were relied upon  in this  proposal  to determine
the  technologically  achievable level  of  particulate control.
Similarly,  the  forthcoming  NOx  proposal  will demonstrate  that  a
0.25  g/BHP-hr  particulate  standard  can  be achieved while  at  the
same  time complying with  the  proposed NOx  level of control.

     Given  the data available,  the  proposed  0.25  g/BHP-hr (0.093
g/MJ) particulate standard complies  with  the  requirements  of  the
Clean Air Act  as  delineated in Section 202(a)(3)(A)(iii).

-------
                                 -63-
                             References

_!_/   Anderson,  Wayne  S.,  Detroit  Diesel  Allison,  Personal  Com-
     munications  with  Timothy  P.  Cox,   SDSB,  EPA,  June  1979.

_2/   Penninga,  T.,  TAEB, EPA,  "Second Interim  Report  on Status of
     Particulate  Trap  Study,"  Memorandum  to  R.  Stahman,  Chief,
     TAEB, August 28, 1979.

3f   Penninga,  T.,  TAEB, EPA,  "Third Interim Report  on Status of
     Particulate  Trap  Study,"  Memorandum  to  R.  Stahman,  Chief,
     TAEB, EPA, November 6, 1979.

4/   General Motors Reponse to EPA Notice of Proposed Rulemaking on
     Particulate  Regulation  for Light-Duty Diesel  Vehicles,  April
     1979.

5/   Passavant, Glenn W. ,  SDSB,  EPA, "Average Lifetime Periods for
     Light-Duty  Trucks  and  Heavy-Duty  Vehicles,"  November  1979.

(>J   "Regulatory  Analysis,  Light-Duty  Diesel  Particulate  Regula-
       tions," U.S. EPA,  OMSAPC, January 29, 1980.

TJ   Johnson-Mat they  Corporation,  Personal Communications  with S.
     Blacker, EPA, September 24, 1979.

_8_/   Penninga, T.,  TAEB, EPA,  "Preliminary  Report  on Johnson/
     Matthey Peugeot 504 Data," December 26, 1979.

_9_/   Springer,  Karl,  Southwest Research  Institute,  "Characteriza-
     tion of Sulfates, Odor,  Smoke  POM, and  Particulates  from
     Light-  and Heavy-Duty Engines  - Part  IX," EPA-460/3-79-007,
     June  1979.

10/  Springer,  Karl  and Ralph  C.  Stahman,  "Diesel Emission Control
     through  Retrofits," SAE  Paper  750205  presented at Automotive
     Engineering  Congress  and  Exposition,  Detroit,  February 24-25,
     1975.

j.1/  Voss, J. R. and R.  E. Vanderpoel, "The Shuttle Distributor for
     a  Diesel  Fuel Injection Pump,"  SAE Paper  770083  presented at
     SAE  Automotive  Engineering Congress,  Detroit,  February  28  -
     March 4, 1977.

12/  Southwest  Research Institute  Diesel  Baseline  Emissions  Sum-
     mary, EPA, June 1,  1980.

13/  Caterpillar  Tractor  Company,   Presentation  to EPA,  May  14,
     1980.

-------
                                -64-

14/  Lipson,  Charles  and Narrendra J.  Sheth,  Statistical Design and
     Analysis of  Engineering Experiments.  Table A-l,  McGraw Hill,
     1973.

15/  Regulatory Analysis and Environmental Impact of Final Emission
     Regulations  for  1984  and  Later  Model Year Heavy-Duty Engines,
     U.S.  EPA OMSAPC, December 1979.

16/  White,  John T.  and  Gary  T.  Jones,  TAEB, EPA, A Study of
     Exhaust  Emissions  from Twenty  High-Mileage  Oldsmobile Diesel
     Passenger .Cars, March 1980.

-------
                                 -65-



                             CHAPTER V

                        ENVIRONMENTAL IMPACT


A.   Health Effects of Particulate Matter

     Suspended particulate  matter has  long been  recognized  as  a
major pollutant  of our nation's air.   Of the greatest  concern  is
the effect  of  particulate matter  (PM)  on human health.   Research
has shown  that  exposure to PM  is  associated with respiratory  and
pulmonary functions, and that effects of  high PM  levels  range  from
increased discomfort to healthy  persons  and  aggravation  of cardio-
respiratory  symptoms  in  elderly  persons,  to  increased  suscepti-
bility  to   bronchitis,  asthma,   and  pneumonia,  to  increased mor-
tality.    Based  on such  research,  when  the Clean  Air Act Amend-
ments of  1970 mandated the  establishment of National Ambient  Air
Quality Standards  (NAAQS), PM expressed  in terms of  levels  of  total
suspended  particulates  (TSP),  was among  the first six  pollutants
for which  a standard  was  promulgated.  The  primary NAAQS  for TSP,
which are  intended  to provide protection  to  the public health,  are
75 micrograms per cubic meter  (annual  geometric mean) and  260
micrograms  per  cubic  meter (maximum  24-hour concentration, may  be
exceeded once  per year).   The   secondary  NAAQS  for  TSP, which  is
intended to protect the public welfare,  is 150 micrograms  per  cubic
meter  (24-hour   average  to  be  exceeded  only  once per year).

     Since promulgation of the NAAQS for TSP, numerous  reviews have
appeared evaluating the  scientific literature bearing on the
scientific  basis  for  the standards.   For  example,  the  National
Academy  of Sciences has  extensively reviewed all  aspects of  PM,
and the  reader  is  referred  to   the  NAS  document  on Airborne Par-
ticles for  a  detailed  treatment of  the  health and welfare effects
of PM.JY   Also,   EPA  is  currently  conducting a  review of  the cri-
teria and  standards for  particulate  matter.   The scientific con-
sensus that  particle  levels  impact  on  human health  will be  taken
as given here.  The emphasis  of  this section will  be on the contri-
bution of  heavy-duty diesel  particulate emissions to ambient  PM
levels,  and  to  any special health  impacts  that  might result from
diesel particulate matter.

B.   Health Effects of Diesel Particulate

     This  section  will  highlight only those aspects of  the health
effects  of  diesel particulate which  differ from those  of TSP  in.
general.   Much has been learned  in the years  since the  NAAQS  (based
on total mass  of particulate) was promulgated,  and  it is now
accepted by most  scientists  that  some  particulate  emissions  are
more deleterious  than  others,  and  that  some  sources  necessitate
priority control  over  others.   There  are  two  characteristics  of
diesel particulate matter which  place  it  among  the  most  harmful
types of particulate matter.  The  first  is  size  and the  second  is

-------
                                -66-


chemical composition.  These will be discussed below.

     1.   Size-Related Effects

     It  is  now generally  accepted that  size is  one  of the  most
critical  characteristics  of  particulate  matter.    The  size of  a
particle  primarily  affects  three  parameters  which,  in  turn,  help
determine the health effect  of that particle:   total  deposition,  or
how  efficiently the  particles  are  deposited  in the  respiratory
tract;  regional deposition,  or where the particle is deposited  in
the respiratory tract; and clearance time, or  how long  it  takes  to
remove  the  particle  from  the respiratory tract.    When  examining
data  presented,  it  will be  important  to note the  differences  in
deposition between nose and  mouth breathers.   As  the  nasal  passages
are more efficient   in  capturing  large particles than  the  mouth,
the sizes of particles reaching various sections  of the  respiratory
tract depend on how the air  is being inhaled.

     Total  deposition by particle size  for   a  mouth  breather  is
shown in  Figure V-l.   As  can be  seen, the fraction  decreases  with
particle  diameter,  until  about 0.5-0.7 micrometers  when the trend
begins to reverse.

     More important than total deposition, however,  is  the  deposi-
tion  occurring  in selected regions  of the respiratory  tract,
because the health effect of  a particle is dependent on the region
in  which it  is  deposited.    Deposition  in three regions will  be
discussed:   the head,  the  tracheo-bronchial  zone  and  conducting
airways,  and  the  alveolar  zone.   These  regions  are  depicted  in
Figure V-2._2/

     Deposition in  the head  (for nose breathers)  is  highest for
large particles and  negligible  for very  small particles.   Deposi-
tion is close to 100  percent  for  coarse mode  particles  between ten
and  fifteen  micrometers  and  higher in size,  while  deposition  is
less  than 10 percent for fine  mode particles below  one to two
micrometers ,J_/  It  is  clear  that far less deposition in the nasal
passages  and  greater  respiratory tract penetration  occur for both
fine and  coarse mode  particles during  mouth breathing  than  during
nasal breathing.

     Deposition in the tracheo-bronchial region  is very similar  to
that in the head(for both nose and mouth breathers), if deposition
is determined as a fraction,  or  percent, of particles entering the
tracheo-bronchial  region.  Deposition approaches  100  percent  around
eight to  fifteen micrometers  and approaches  10  percent  around one
to two micrometers.

     Deposition in  the alveolar region  is  shown  in  Figure  V-3,
based on the total number of  particles entering  the  mouth  or nose,
not on  the  number  of  particles  entering the alveolar  region.I/
Deposition in this  region  is low  above  five  to  seven  micrometers
because  the  larger  particles have  already  been  captured  by the

-------
                     -67-


                   Figure V-l If
                                     i   i  f  i i i 11    i
     0.1    0.2  aa.  oj-  or  tjo     33     5   r  ID
                 A«rodyr>am»< Dtarnvtor— prrr
Ttotal resp:LratDry tract <3eposition during mouthpiece inhala-
tions as a  function of D (aerodynamic dianetsr in  ym)  except
below 0.5 urn,.where deposition is plotted vs linear diairster.

-------
                                             -68-



                                    Figure V-2 2/
                Uppsr respiratory tract
               Anterior cares
                    Lowsr

                  respiratory

                    tract
                                                V
                                                      3
                                                      •o
                                                      c
                                                      o
                                                      u
                                                                     Bronchus
Diagraimnatic  representation of  the human upper and lower respiratory tract,

-------
                         Figure V-3 I/
1.0
c
JO
E .3
Q
ex
•
a .
.4

.3


.3
.1

0
1 ' ' 1 ' ' ' ' I • t ' ' 1 ' ' l ' | '
••
••».
• v
.^^ *3
' \k.
/ v\
i- 1 * * i \ »
~* / i \ •• -
/ *' A
;> •- \*
- T ' A*"
T / -1 n
T *t
x o'/--^'""~"""v*- r~» 	 AI»«eJoi
- K ,•*" \ A* via -
^i^ ^-^ ^ \ •L* mo*rth
1 — °- * \ .TV--
1 i ^ vl*AA"»
no**- V* i>.
\-\
• 1 1 t f , r , , t f ! I 1 1 I*^«!S !
O.1 0.2. OO- OJ. Otr tO- 99 ST1O M
y,ir«»wr* C^TrTmPt'Pr. UTH Ae»rnr1 vrvwrn r» i^TAtirit^froT*. urn

r \

Dqccsition in the nonciliated alveolar region >r  by percent of aerosol
entering the mouthpiece, as a function of diameter.

-------
                               -70-

nasal  passages  and  the  tracheo-bronchial region.   Deposition
reaches a relative peak around two to five micrometers.  The level
of the peak depends on whether the person is breathing through the
mouth, when deposition  reaches  40-50 percent,  or the nose,  20
percent.

     Depending on chemical composition, particles deposited in any
region of the respiratory tract can  affect  health.   Of particular
concern are those particles that reach the  lung (tracheo-bronchial
region, conducting airways and the alveoli).   The alveolar region
(where gas-exchange takes  place)  is  the most  sensitive  region  of
the respiratory tract.  Moreover,  a significantly longer clearance
time is required  for  particles in  the  alveolar region.  Clearance
time is  the  time  it  typically takes for a  particle  to be removed
from the  region  in question.   In healthy individuals, the clearance
of particles deposited  in the nasal  passages  and the tracheo-
bronchial  region is  usually completed in less  than one  day.jY
Clearance can take somewhat  longer for  those  people with respira-
tory ailments.   In  the alveolar region, clearance  is  measured  in
weeks  unless  the  particle is  very  soluble  in body  fluid,  which
diesel  particulate is  not.  While  the results of studies on humans
are variable, it appears  that a half-time clearance for relatively
insoluble  particles  is  on  the  order  of  five to  nine weeks .J_/

     As a result  of  a review of  the  available information on the
effects of particle size on  deposition  and  health,  EPA has recom-
mended  that  future health  effects research  be  conducted  on  two
size-specific fractions  of PM.2/   One fraction is  labeled  in-
halable particulate (i.e., particles  having  a diameter equal to  or
less than  fifteen micrometers).  This fraction includes  those
particles which  primarily  deposit  in the conducting airways and the
gas-exchange portions  of  the  respiratory tract.   The second frac-
tion is  the fine  particulate  (i.e.,  particles having a diameter
equal  to  or  less than 2.5  micrometers).    This  second cutoff  was
chosen for two  reasons;  1)  this  fraction  includes those particles
which  primarily deposit  in  the gas  exchange  portion  of  the lung
(alveolar),  and  2) due to the  breakdown  of  ambient  particulate  by
size and  chemical composition, there  is  a natural  break  between
fine and  coarse  (diameter  larger than 2.5 micrometers)  particles  at
this size.

     Diesel  particulate  is  very  small in size.   Its mass mean
diameter  varies  between 0.05 and 0.2 micrometers ._3/4/  Essentially
all diesel particles  fall  into the inhalable range and between 94%
and  100% can be  characterized  as  fine  particulate .^M/j)/  Because
of its small size, diesel particulate  belongs  to  thaT category  of
particulate  which  is most  likely to deposit  in  the alveolar region,
thus remaining  in contact with  the  most  sensitive  areas  of  the
respiratory  tract  for  comparatively long periods of time.  Clearly,
diesel  particulate is of more  concern  than  larger particles which
deposit in  the  head  or  tracheo-bronchial  regions and which have
much shorter  clearance  times.   Because of  this,  the  control  of
diesel  particulate and other  fine  and  inhalable  particulate is  of
high priority.

-------
                                -71-
     2.   Chemical Composition-Related  Effects

     In addition to  particle  size,  chemical composition is an
important  factor  in determining the health  effect  of a particle.
There are a  wide  variety  of chemicals  of particular concern,  such
as  fibers  (e.g., asbestos),  toxic elements  (e.g., Be,  Cd,  Pb),
organic matter  (e.g.,  benzo(a)pyrene),  carbon,  and sulfuric acid.

     Diesel  particulate is  primarily carbonaceous, with between 10
and  50  percent of  the particulate by weight being extractable
organic matter ,^_/5_/6_/l_/   This  organic  matter  is definitely muta-
genic in short-term bioassays,_7_/ and EPA is currently performing a
health  assessment  to  determine the carcinogenic risk  of  diesel
particulate  to humans .JJ/  Known or  suspected human carcinogens are
present in  diesel  particulate,  such as benzo(a)pyrene,  which
comprises  about  0.0001  to 0.007 percent by  weight  of diesel  par-
ticulate.^/^/   However, most  of  the mutagenic  response  is being
caused  by  substituted  polycyclic  organic  matter, which  does  not
require metabolic activation.^/ At  this time, no  definitive  state-
ment  can  be made concerning the  complete  effect of  diesel  par-
ticulate on human health.   However, the data available is serious
enough to merit caution and diesel  particulate  should definitely be
numbered among  those  chemical  types of particulate which require
priority control.

C.   Visibility

     Visibility degradation  is  perhaps  the  most  noticeable  effect
of  air  pollution  on today's  society.   In  addition to the adverse
health effects previously delineated, diesel  particulate also plays
a significant  role  in light extinction;  which  is defined,  for the
purposes of  this  study,  as  the process whereby the illuminance of
light  is  reduced while  propagating  through  a medium  (such as
air).9/

     The typical observer can detect an object with  2  percent
contrast against  the background.^/  Expressed mathematically the
distance,  Lv, at  which  a black  object  is just visible is given by:

          3.92
     Lv = T-	
          bext

Where,  3.92 = -In  0.02 and  bext  refers to  the sum of  the col-
lective extinction  coefficients of the four processes responsible
for light  attenuation.^/  These processes are:

     1)   Scattering  by gas molecules, responsible for  the sky's
          blue color;

     2)   Absorption by gas molecules;

     3)   Scattering by small particles;  and,

-------
                                -72-
     4)   Absorption by particles.

Since  diesel  particulate  impacts  directly  on two  of these  four
mechanisms, the  latter  two,  its potential effect on  visibility  is
of  some  concern.   In order to gain  insight  into  the  relative  role
of  diesel particulate  in light  attenuation,  each  of  the  four
components of  the  extinction coefficient  should  be  examined.

     The extinction  coefficient due  to  scattering by  gas  molecules
in  the free atmosphere at  sea  level  is  roughly 1.5  x  10~-> meters"1
(for light at a wavelength of  0.52 micrometers  (green));  values  of
the extinction  coefficient  within a  few  percent  of  this  have
actually been measured.^/  If  light  degradation were  due  solely  to
gas molecule scattering,  then  the visibility  would be  approximately
260 kilometers,  by the aforementioned formula.   Thus,  scattering  by
gas molecules does not play a  major role in visibility degradation.

     Of  the  many gaseous  species  present  in the atmosphere,  only
nitrogen dioxide (N02) is  present  in high  enough  concentrations  to
have a significant light absorption  impact ,_9/  Nitrogen dioxide  is
a  strong absorber  of blue  light  and can  cause the  atmosphere  to
have a reddish-brown haze.  At a  NC>2 concentration  of 0.05  parts
per million,  the National Ambient Air Quality Standard for nitrogen
dioxide,  the extinction coefficient  due to NC>2 absorption would  be
approximately 8.1  x  10"^  meters"1  (based on 0.40 micrometer  wave-
length light  (blue)).10/   In a homogeneous  atmosphere  with  0.05
parts per million NC>2 visibility would be  roughly  41 kilometers due
to  the combined  effect of  NC>2 absorption and  gas  molecule  scat-
tering.  An important caveat to consider is that the  aforementioned
extinction coefficient  was based  on blue light  only; the visual
spectrum, of course,  consists  of other colors as well, colors  which
are not  as affected  by  NC>2 absorption.   For example,  light with a
wavelength of  0.70  micrometers (red)  would  yield  an extinction
coefficient of 8.1 x 10~7  meters"1 at a N02 concentration of  0.05
parts per  million;  less  sensitive  than blue  light  by a  factor  of
100.

     The  scattering of light by particulate is  generally attributed
to  particles whose  size  corresponds  to  the wavelength of incident
light; that is,  sub-micron  particles.   Figure V-4 shows  the  ratio
of  mass  to scatter coefficient as a function of particle radius.
From this  figure, it  follows that  particles  whose radii lie  in the
0.1 to 1.0 micrometer range are the most  efficient at scattering.
Some typical particles  in this  size range  (and  up  to 2.0 micro-
meters in  diameter)  include sulfates and organic  compounds such  as
condensed hydrocarbons and oxidized  organic matter .JY  By contrast
such particles  as  soil and tire  dust,  road  debris,  fly  ash, and
airborne  products  of  rock-crushing  have  little  influence  on
scattering (except  in the case of rare dust storms).\J

     In  addition to  particle size  (and, of course,  concentration),
atmospheric water  vapor  plays an  important  interactive role  in
light  scattering  by  particles.    Relative humidity  in   the  30-60

-------
                   -73-
             Figure V-4 9/
 m
bfia
                                              rIOOOO
                                              rlOOO
                                              HOO   m
                                              -10
   Calculated scattering efficiency for a log normal
   aerosol  size  distribution, geometric standard
   deviation  equal  to 2, as a function of geometric
   mass  mean  radius.  M is the fine particle mass
   concentration, bsp is the scattering coefficient,
   and rgm  is the geometric mass mean radius.  For
   reference, the right hand axis is the mass concen-
   tration  required to give a visual range of 40 km.

-------
                                -74-
percent  range  has little  effect  on  visibility.    However,  at  80
percent  relative humidity,  the  light scattering potential  of
aerosols  is  twice that at  the  30  percent  level.LLf   As  relative
humidity  approaches  88 percent,  the  light  scattering  ability  of
typical  aerosols  is  about  four times  that  for  the  same  aerosol
concentration at  30  percent  relative  humidity._!!_/  This effect  is
due to the hygroscopic  nature of the particles.  As  the air's water
vapor  content  increases,  particles pick  up water  and,  thereby,
increase in size.   Their potential  to  scatter light  is maximized  as
their size approaches the wavelengths  of visible light and  as  they
ultimately become fog droplets.   Of course,  the effect of  relative
humidity varies  depending  on the composition  of the aerosol.

     For  particles with a diameter from  0.05 to 0.2 micrometers,
typical  for diesel  particulate,  the scattering extinction co-
efficient at a concentration  of  5 micrograms  per cubic meter ranges
from  8.3 x 10"^  to  5.0 x 10"? meters "1 .j>y Maximum scattering
occurs from  particles  whose  diameter  is  approximately  0.6 micro-
meters .

     Absorption of light  by  particles  is  approximately  10  percent
of the particle scattering attenuation in clean areas and up to  50
percent  in  urban  areas,  with the  most important contributor being
graphite  carbon.j)/   Thus,  any sub-micron  particles  with  a high
carbon  content will have  a significant impact on visibility.
Heavy-duty  diesel  particulate,  with  its  50 to  90  percent carbon
content,   falls  into  this category ._5_/jj/_7/   The  absorption to mass
ratio for  carbon  is  approximately  7 meters^  per  gram.!2_/  Actual
measurements of the absorption to mass ratio of diesel particulate
approach this value,  verifying the high  carbon content and  implying
that gaseous hydrocarbons bound to  the surface  of diesel  particu-
late play an inconsequential  role  in light absorption.^/   Thus,  at
a concentration of 5 micrograms per cubic meter the light  absorp-
tion coefficient of carbon in diesel  particulate  is roughly 1.8 x
10~5 to 3.2 x 10~5 meters"1.12/

     Since diesel particulate affects  visibility  through both the
scattering and absorbing phenomena, the extinction coefficients  of
each of these processes should be combined when evaluating  the net
visibility impact of diesel  particulate.   Thus,  the visibility  in
an atmosphere permeated with  5 micrograms  per cubic meter of diesel
particulate would be 71-117  kilometers when the scattering effect
of ubiquitous gas  molecules is included.   Although this scenario  is
admittedly ideal,  due  to  such assumptions as a  fixed  5 microgram
per cubic meter heavy-duty diesel particulate level  extending
throughout a hypothetical 71-117 kilometer  line  of  sight,  it does
indicate  the  potential visibility  impact  of  diesel particulate.
Indeed,  such an impact  may already  exist,  as suggested in  a recent
study of Denver's  "Brown Cloud."13/

     Another approach to assessing  the visibility impact of diesel
particulate is  to quantify attenuation on  a per kilometer basis.
This can be done through the  following  relationship:

-------
                                -75-
     I = I0 e ~bextx

Where:

     I    = Intensity of light after attenuation;

     IQ   = initial intensity;

     bext = extinction coefficient;

     X    = distance from observer to object.^/

     Using  the values  of  the extinction  coefficient previously
determined, one  finds that an  atmosphere  permeated  with  5 micro-
grams  per  cubic  meter  attenuates  3.3 to  5.4  percent  of incident
light per kilometer of propagation.   Were that same atmosphere void
of  diesel  particulate,  then  only  1.5  percent of  incident light
would  be attenuated  due  to  inherent  gas molecule  scattering.

     In conclusion, of the four primary mechanisms of  light atten-
uation  in  the atmosphere, heavy-duty diesel  particulate  directly
impacts two:  particle scattering  and absorption.  In an atmosphere
void of  NC>2  and  sub-micron   particulate,  the  hypothetical  visual
range is approximately 260 kilometers. With 5 micrograms per cubic
meter  homogeneously  distributed  throughout  the  same pristine
atmosphere, the visibility  is reduced to 71-117 kilometers, a
reduction  of  55-73  percent.    These figures  reflect  the  maximum
distance at  which  an  object  could  be discerned.   Impairment  can
occur  at substantially smaller  distances.   On  a per kilometer
basis,  for instance, 3.3-5.4 percent of the  incident light is
attenuated  in  an atmosphere  with  a heavy-duty diesel particulate
concentration of 5 micrograms  per  cubic meter.

D.   Current Ambient Levels of TSP

     The primary  NAAQS for TSP of  75 micrograms per  cubic meter
(annual  geometric mean)  is  currently  being exceeded in many
areas of the country.  While  relatively large reductions in ambient
TSP  levels  occurred  between   1971  and   1975,JA/  (particularly at
those  sites which  showed high levels of  TSP),  the  next  two years
have shown  more of  a holding  pattern  than a  continued  downward
trend.15/ Figure V-5  shows the nationwide  averages  of ambient  TSP
levels~~from 1972 through  1977-   The ambient  TSP level exceeded by
25  percent  of the  sites  decreased  from 78  to 71  micrograms  per
cubic meter between  1971 and  1975,  while  in  1977  it  was  still 71
micrograms per cubic meter.  The TSP level  exceeded by  the worst 10
percent  of  the sites  still  managed  to  improve,  however,  through
1977.   This  level  decreased   from  97 to  88 micrograms  per cubic
meter between 1972 and 1975 and then decreased to 84 micrograms per
cubic meter in 1977.

     The high  ambient levels   of  1976 and 1977 were  due  at least
partially to very dry  weather,15_/16_/  In 1977, some sites recorded
levels  of 1000 micrograms per  cubic  meter  for a day or  two and  this

-------
       _76-     Figure  V-5 15_/
                      0-
                                -90.TH PERCENTILE
                                -75TH PERCENTILE
                           «*•
                            «$	25TH PERCENTILE
                                -lOTH PERCEMTILE
                    Figure 3-1. Sample illustration
                    of plotting conventions for
                    box plots.
liiu
I—
=«. 120
P 5
ts "^
|| BO
0- -£j
sy so
_I O
o
h-
20
o
BOX


: i t
•s ^
i? V
— - >» X
-TV
—
! !
197Z 1973

PLOT ANNUAL VALUES
^y.^.

lilt-
i rj ITT
^ K H M —
V T t V '-
,_-
\ ! f r •
1974 1975 1976 1977
YEAR
          Nationwide trends in annual mean total suspended
particutate concentrations from 1972 to 1977 at 2,707 sam-
pling sites.

-------
                                -77-
alone can cause the annual mean to increase 10 percent ._15_/  Figures
V-6 and V-7 show the  ambient  TSP  trends  by region for 1972 through
1977.  The dust storms of 1976 were primarily located in Regions 8,
9, and 10, while those  of  1977  were  primarily located in Region 6.

     The  fraction  of  the nation's  population which  is  exposed to
TSP levels exceeding  the  primary  NAAQS is shown  in  Figure V-8.16/
While the number of people exposed to such levels dropped 9 percent
between 1972 and 1975, this downward trend stopped in 1976 and 1977
when  the  number of people  exposed remained  constant at  about  22
percent of the nation's population.   An  identical trend  is present
for the nation's metropolitan population.  For the last three years
(1975-1977),  27 percent of the nation's metropolitan population has
been  exposed  to ambient  TSP levels  exceeding the  primary  NAAQS.
These people are living in  areas  where the quality of the air they
breathe could be harmful to their health.

     An even greater percentage of  people are  living  in  areas
exceeding the  secondary NAAQS for TSP.   For  example,  in 1975 when
49 million people were living in areas exceeding the primary NAAQS,
89  million  people  were  living  in areas exceeding  the  secondary
NAAQS.   These  people are  living  in  areas  where the air quality
could be  a hazard  to  their  welfare  (i.e., visibility, corrosion of
materials, vegetation, etc.).
                                                            \
     To  examine the  TSP  problem  in  greater  detail, ambient  TSP
trends  are available for  five large metropolitan areas.15/167
These  five  cities, New York,  Chicago, Denver, Cleveland, and  St.
Louis, were largely unaffected by the dry weather of 1976 and 1977
(except possibly St.  Louis), so  this bias should not be present.
The populations  exposed  to TSP levels exceeding  the primary  NAAQS
in these  five metropolitan  areas  are  shown in Table  V-l.   The most
significant  improvements  occurred  in  the  New  York  metropolitan
area.16/   In  1970,  11.2  million people  in metropolitan  New York
lived in  areas  where  the  annual primary  NAAQS was being  exceeded.
By  1976,  all  TSP monitors  had  registered annual means  below this
level.   Thus,  no  one  was  living in areas  exceeding the primary
NAAQS.    The  average   TSP  concentration  in  metropolitan  New York
dropped from 78 micrograms per cubic meter in 1970 to 55  micrograms
per cubic meter in 1976.

     The results for the other four cities were somewhat  different.
Improvements in the number  of people  exposed  to  ambient  TSP  levels
in  excess of  the  primary  NAAQS  have been  made, but  significant
numbers are still exposed.   Denver  is probably in the worst  situa-
tion.^/   While the  percentage  of  people  exposed  to TSP  levels
exceeding the  NAAQS has decreased 9  percent,  a  full three-fourths
of  the population  are still  exposed, to these  excessive  levels.
Likewise, for  Chicago, 64 percent  of  the population  are  still
living in areas where  the TSP levels  violate the  primary NAAQS.16/
Cleveland has  experienced a steady  decrease  in population exposure
to excessive TSP levels since 1972,  though 27 percent of  the people
in the air quality control region are still exposed.15/

-------
                              ''\               A
                                          >         -'
                       U.S.EPA A1B QUALITY CONTROL REGIONS, EASTERN STATES
 60
140
12Q
ioo
 80'
 20
REGION
          1.1.1
     1972 1373  1974 1375 1975 1977
REGION 2
                                     t    I.I    1  _. 1    F
                                                                   r
              1972 1973  1974 1975 1975 1977
                                                           i    ?
             1972 1973 1974 1975 1975 13
                160
                140
                120
               eioo
                 20
                 20
                      I    t    t
                REGION 4 „
                                       t    t
                     1972 1973 1374  1975 1976 1977
                REGION 5 _
                                                     1 .   t    I    I
                              1972 1373 1974 1975 1975  1977
         1972-1977.
                   Regional trends of annual mean total suspended participate concentrations

-------
                  9 D-
                                   Figure V-7  15/
                      U.S.EPAAIR QUALITY CONTROL REGIONS, WESTERN STATES
160
140
12D
 60
 40
 2D
REGIONS _
REGION? -
                II
                                                       t   I    I
     1972 1973-19-74-49-75-1976 1977        1972 1973 1974  1975 1976  1977       1972 1973 1974 1975 1975 1977
                                                                            REGION 10 _
                       1972 1973 1974 1975 1976  1977
                                                            J	!	1	'    '     t.
                                 1972 1973  1974 1975  197S  1977
                                                YEAR
        trations. 1972- 1977.
                               Regional trends of annual mean total suspended participate concen-

-------
  40
5 30
 UJ
' CO
 GS
 c.
 x 2Q
 • • • *•**
  10
                   _L
                              METROPOLITAN
                             NATION
                                   NON-
                              METROPOLITAN
                                    1
    72      73      74      75      76

               YEAR OFTSP EXPOSURE

                Population exposure to annual
     mean TSP in excess of NAAQS (75
                                           77

-------
                                -81-
                                Table V-l

                  Population Exposure to TSP Levels in
                 Violation of the Primary NAAQS 15/16/
Population
(millions)*
1970
1972
1973
1974
1975
1976
1977
New York Chicago Denver Cleveland
17 3.4 1.1 3.4
Percentage of Population Exposed to Levels
Exceeding NAAQS
60% 100% 83%
60%
12% 50%
37%
75% 44%
0% 64% 29%
27%
St. Louis
1.9

69%
46%
48%
43%
60%
62%
*    1970 Census data for the area studied,  usually comprising the
     Air Quality Control Region.

-------
                                -82-
     St. Louis is  the most  interesting  case.   The population
exposed to excessive TSP levels decreased steadily from 69 percent
to 43  percent  between 1972  and  1975.  After that the  exposed
population increased back  to  nearly the 1972  level.   Part of the
reason for this  increase,  which first occurred  in  1976,  may have
been the dry  weather  of that  year.   The precipitation around St.
Louis was "slightly below normal" for  1976 .JJ>/   However, nothing is
mentioned  concerning the weather of  1977 and Region 7  (which
includes St.  Louis) in general  showed  no  signs  of  exceptionally dry
weather in 1977  (see  Figure  V-7).   Thus,  it  would  appear  that at
least  some and perhaps most  of the  increase of 1977 is  due to
factors other than dry weather.

     There  are  two primary  reasons  why ambient TSP  levels  have
dropped significantly  between  1971  and 1975.  Both reasons concern
stationary source  particulate  emissions.   The  first  reason is the
application of  particulate control  technology  to  the  stationary
sources of particulate emissions.   Since 1970,  many of the largest
polluting  industries  have been required  to  control  particulate
emissions.  This  has occurred nationwide  through attempts by states
and  localities  to comply  with the NAAQS  for  TSP  (e.g.,  through
equipping  existing plants with particulate  control devices as
deemed  necessary by  local  TSP levels).   The  second reason is
that many  combustion  sources have  switched  to  cleaner fuels which
result  in  lower  particulate  emissions.    The  combustion  of  coal
produces much  more particulate emissions  than the  combustion of
oil, and  the combustion of  natural  gas  produces even less  par-
ticulate emissions than the combustion of oil.   Thus, many sources
in the early  1970's switched to oil and gas to reduce particulate
emissions,  as well as  sulfur  dioxide emissions.

     While  these  methods  have decreased ambient  TSP levels over the
last seven  to eight years,  there are some inherent problems associ-
ated with both of  them which limit   future reductions.  First, most
of  the  large  reductions in  particulate emissions  possible  from
stationary  sources have already been made.14/  The majority of the
largest polluting plants have already come  under state and federal
standards,  or are under compliance  schedules soon to be completed.
The potential for continued emission reductions has diminished, and
future reductions  will  be  even more  costly.    Since  current  NSPS
are based on  the  best  system of emission reduction which  has been
adequately  demonstrated, (while taking  into  account the  cost  of
such a  system),  the advent  of even greater control  of  currently
controlled  industries  will  not  be widespread, barring major technor
logical breakthroughs.

     Second,  the  trend toward switching to oil  and natural gas from
coal has  already stopped and even  reversed  itself due  to the
shortage of domestic oil and  natural gas.  Thus, any gains made in
the past  from switching  to  cleaner  fuels  will eventually disappear,
and likely reverse themselves  as coal  usage  becomes  more  and more
prominent.

-------
                              -83-
     Finally, growth  in  production  will  enter into the situation.
In any  industry where emission  standards  stay  at  current levels,.
every new  plant not  replacing  an  obsolete plant  will  add  to the
overall  emissions  inventory.   The ability  of the air  to  clean
itself does not increase  with  the nation's  productive capabilities,
so the end result is dirtier air.

     In  conclusion,  while  significant  progress was  made in the
early 1970's  in reducing  ambient  TSP levels,  22  percent of the
national population is  still  exposed to ambient TSP levels in
excess  of the  primary  NAAQS of 75  micrograms per cubic  meter
(annual geometric mean).   And the two strategies which contributed
most  to  the TSP  reductions of  the early 1970's,   application of
emission controls  to the stationary  sources  with the largest
potential  reductions and  fuel-switching from coal  to oil and
natural  gas, clearly will  not be able to  provide  significant new
reductions, especially since the fuel-switching  process will likely
reverse itself and continued economic  growth is  expected to provide
new  sources of  particulate matter.   Therefore,  heretofore un-
controlled particulate  sources   and new major particulate sources
will  need  to be  regulated if  further TSP reductions are  to be
achieved.  The next section will show  the environmental benefits to
be gained  from the  control of light-duty diesel particulate  emis-
sions.

E.   Impact of Diesel Particulate Emissions

     Three different  aspects  of the diesel's  environmental  impact
will be examined here.  First,  the amount of particulate emitted to
the atmosphere will be determined.   Second, the diesel's impact on
large-scale  TSP  levels  will  be examined.   Finally,  the  diesel's
impact  in  localized  areas  where particularly high concentrations
could occur will be  examined.  All  of these impacts will be deter-
mined for  1995,  as  by that time the environmental  benefits  of the
1986 standard will be nearly complete.

     1.   Emissions
     In  order to  determine  the  particulate  emissions  from all
heavy-duty  diesel  vehicles,  two  basic  factors  are needed:   the
amount  of particulate emitted  by each vehicle  per unit distance
traveled and the total distance traveled  by  all heavy-duty diesels.
For  the purpose of evaluating the future  impact of heavy-duty
diesels  as  a particulate  source  the year  1995  will  be  the focal
point.

     Historically,   2.0 grams  per  mile (g/mi)(1.24 grams  per kilo-
meter  (g/km))  has  been used  as the heavy-duty diesel particulate
emission  factor ._17y_18_/_19/   Even  though  this factor was  based on
steady-state  tests,  as  opposed to  more  representative  transient
tests,  it is believed  to be a  good  estimation of  future heavy-duty
diesel  particulate  emission  levels.   Transient cycle test results
from the Southwest  Research Institute program (refer  to Chapter IV)

-------
                                -84-
indicate an average emission rate of approximately 1.5 g/mi for new
engines.  However, this simple average does not consider the caveat
that  larger  diesel  engines, which generally emit  more  particulate
per  unit  distance traveled  than  smaller engines, will  constitute
the  largest  share of the heavy-duty  diesel  fleet.  Also,  as  sug-
gested in Chapter IV, the mandated control of heavy-duty diesel NOx
emissions would  likely  increase  particulate emissions  if a  par-
ticulate standard were not  implemented.   These  two tenets  together
with  some in  use deterioration,  support the use of  the  historical
2.0 g/mi(1.4 g/km) heavy-duty diesel particulate emission factor in
this analysis.

     PEDCo Environmental (based on DOT data)  reported that  in 1974,
1.286 trillion miles  were traveled by  all  motor vehicles  in  this
country; 8.8  percent  of which were by heavy-duty  vehicles  nation-
wide and 3.4  percent  by heavy-duty vehicles in urban areas ,19_/   A
1.5  percent  per  year growth  rate  in  nationwide and urban  vehicle
miles traveled (VMT)  has  been used to extrapolate 1974  VMT  to the
1995 scenario.  These results appear in Table V-2.

     In order  to determine  the  fraction of future heavy-duty  VMT
attributable  to  diesels,  several  factors have been  used.   These
include the sale  projections  outlined  in  Chapter  III; the  standard
EPA  breakdown of annual  heavy-duty VMT by  model year;^0y  the
fraction of  total registration by model  year ',2QJ and the  urban/-
rural VMT split  by  mobile  source category,19_/   The  result  is  that
71-86 percent  of nationwide  heavy-duty VMT and  67-82  percent  of
urban heavy-duty VMT will be  by diesels  in  1995.   Consult  Appendix
I for further details.

     Combining the expected 1995  heavy-duty diesel  VMT with the 2.0
g/mi  (1.248  g/km)  emission factor, 218,000-266,500 metric tons  of
heavy-duty  diesel  exhaust  particulate  will  be  emitted  in  1995
nationwide if no  control  is implemented.  In urban  areas,  79,000-
97,000 metric tons will be emitted in  1995.   These  values are  shown
in Table V-2.  To put things into perspective,  Table  V-3  provides  a
comparison of  current annual emissions  from several major  indus-
trial source categories with estimates  of uncontrolled diesel
emissions  in 1990.   As can  be  seen, heavy-duty and  light-duty
diesels   are  projected  to  be  significant sources  of  particulate
emissions by 1990, if left  uncontrolled.

     It  should be remembered  that heavy-duty gasoline engines  also
emit particulate  (mostly in the form of  lead-salts).  The  recently
promulgated  1984  hydrocarbon and  carbon monoxide emission standards
for  all  heavy-duty vehicles  will result  in the  use  of  unleaded
gasoline in  this  vehicle  class;  effectively phasing out lead-salt
emissions from that source.   Reductions in lead-salt  emissions  will
also transpire in the  1980-1984  time  frame due to the  "capturing"
of part  of  the heavy-duty  gasoline market by   heavy-duty  diesels.
EPA  estimates  that   gasoline-fueled  heavy-duty  vehicles   emitted
approximately 30,000  metric  tons  of   lead-salt  in 1974;  however,
only  13,000  metric tons  could be classified  as   suspendable  (the

-------
                           -85-

                         Table V-2

                  Traffic Characterization
1974 Total VMT 19/

1995 Total VMT

VMT Growth Rate

Heavy-Duty VMT
  Fraction 19/

1995 Heavy-Duty VMT

1995 Heavy-Duty
  Diesel VMT
  Nationwide

1.286 Trillion

1.758 Trillion

1.5 % Per Year

0.088


154 Billion

109-133 Billion
                                                    Urban
1995 Heavy-Duty Partic-   218,000-266,500
  ulate Emission (Metric Tons)
0.694 Trillion

0.949 Trillion

1.5 % Per Year

0.062


58.8 Billion

39.5-48.3 Billion


79,000-97,000

-------
                                 -86-
                               Table V-3

         1975 Emissions from Selected Major Stationary  Source
     Categories and Projected 1990 Emissions from Diesel Vehicles
                                                       1975 Emissions*
 Stationary Sources                                    (tons  per  year)

 Electric Generation Plants                               3,000,000

 Industrial Boilers                                       1,000,000

 Iron and Steel Industry

      Coke Ovens                                          <100,000

      Basic Oxygen Furnaces                                100,000

      Blast Furnaces                                      <100,000

 Kraft Pulp Mills                                          200,000

 Aluminum Industry                                         200,000
                                                      1990 Emissions
 Mobile Source                                        (tons per year)

 Heavy-Duty Diesels                                   215,000-266,500
*    Stationary source data extracted from National Emission Data
System, 1975.

-------
                                -87-
rest-  being  too  large  to remain  suspended in  the atmosphere  for
extended  periods of time).   In  urban  areas approximately  13,000
metric  tons  of lead-salt were emitted  in  1974 (5,600 metric  tons
suspendable).   By 1995,  nationwide lead-salt emission from  heavy-
duty  gasoline-fueled vehicles are  expected  to decrease to  1,700-
3,600 metric  tons nationwide (700-1,500 metric tons  suspendable).
In urban  areas,  projected 1995 lead-salt emissions  from  heavy-duty
gasoline  vehicles  are  600-1,300  metric  tons  (250-540 metric  tons
suspendable).

     2.   Regional Impact

     The  regional,  or  large-scale,  impact  of diesel particulate
emissions is greatest in urban areas.   This is  no  surprise  since  it
is  in  urban  areas where  the greatest  concentration of  vehicles
exist.   As  it is also in urban  areas  where  most  of  the people  of
the nation  live  and  where  most  of the violations of  the NAAQS  for
TSP  occur ,_15_/  it  is  appropriate that  this  section  concentrates
primarily on  the impact of  diesel particulate emissions  in urban
areas.

     Two  studies  have  attempted  to determine the impact of  diesel
particulate emissions  on urban  air. The first study  was performed
by  PEDCo Environmental  for  EPA.^1_/ It  used  ambient  lead concen-
trations coupled with  lead emission factors  to determine the
relationship  between emissions  and air quality for mobile sources
in New  York,  Chicago,  and Los Angeles.   Then, ambient  concentra-
tions of  diesel particulate  in  those cities were calculated using
this  relationship and known  diesel particulate emission  factors.
Ambient  levels  of diesel particulate  were  calculated at 15  actual
TSP monitoring sites  so the calculated levels  could be  directly
compared  to  levels  currently being  measured  at  the same  sites.

     The  second  study  was  conducted by EPA and used  a methodology
similar to  that used  in  the PEDCo  report ,22_/   Ambient diesel
particulate  concentrations  were   estimated from  ambient  lead mea-
surements taken  in  over  35  cities ranging  in population from  less
than  100,000  to over 5,000,000.    The  study  also includes similar
estimates of  ambient  diesel particulate levels  in Chicago and
Toledo  which  were submitted  by  General  Motors during the comment
period  following  the proposal of  the  light-duty diesel particulate
regulation.23/

     Each study used a  different set of  input  data for  emission
factors, VMT  growth,  diesel penetration,  etc.    In order  to  be
comparable,  each had to  be adjusted  to  a common  set of  input
factors.  This  has  already  been done  under  separate  cover  for
convenience.22/  The common set  of input  factors used  was described
in the  previous  subsection on emissions  from uncontrolled  diesels.
The only  difference  was  that growth in VMT was only  assumed to  be
one percent  per year in the  central  city  areas being examined  by
the two studies.

     One additional adjustment was also made to the results  of  the

-------
                                -88-
PEDCo study.  From  PEDCo's  text,  it seemed possible  that  an  error
was made  concerning the automobile's  contribution to ambient  TSP
levels in  New York.  An analysis  of the references  used  by  PEDCo
revealed that an error was indeed made.   A  referenced study,  which
determined  the  auto's  total  contribution  to  ambient  TSP  levels
included reentrained  dust,  but was  taken to refer  only  to  auto-
mobile exhaust emissions.  This  error  caused the New York results
to be overestimated by a factor of  2.66.  Due to the  fact  that  the
Chicago results  were partially based on this erroneous factor, they
were  overestimated by  a factor  of 1.62.   Any use  of  the PEDCo
results  here will  be adjusted  by these factors and  a detailed
discussion of the  adjustments  can  be found.under separate cover.22/

     The results of both studies are shown  in Tables  V-4,  and V-5.
The expected impacts in  New York,  Los Angeles  and Chicago are  about
the  same  whether  determined  by  EPA  (Table V-5)  or  PEDCo  (Table
V-A) .   This finding is not  surprising  since both  studies used
ambient  lead  measurements  as  a  basis,  though  slightly different
methodologies were  used  to  convert  these ambient lead  concentra-
tions into  diesel  particulate concentrations.   The level  found  at
the  first  Chicago  monitor  modeled  by PEDCo  (Table  V-4)   appears
quite out  of line  with  all  the others  and  will be  excluded from
further reference.  It is known that PEDCo  assumed that  automotive
exhaust particulate was  a constant  fraction of TSP throughout  the
city.  If  this particular monitor was  in  a  heavily  industrial area
showing a  very high TSP  level due to  industrial  sources,  of which
Chicago has quite  a few,16/  then the automotive  portion  could have
been overestimated.

     The studies indicate that the  regional impact of uncontrolled
heavy-duty diesel  particulate emissions  in  1995  would be  2-7
micrograms  per  cubic  meter  in the  nation's three largest  cities.
The levels  for  other cities  are  somewhat  lower  and  these  levels
tend  to  decrease  with  decreasing  population,  as  shown in  Table
V-5.  There are exceptions  in  each population  category,  such   as
Phoenix  and  Kansas  City.   The  impact of  heavy-duty diesel par-
ticulate emissions in Phoenix  is  projected to  be 4.0-4.9  micrograms
per cubic  meter while that  in Kansas City  is only projected to  be
1.4-1.7 micrograms  per  cubic  meter. It should  be  noted that  the
regional impacts  in Table V-5 are  based  on National Air  Surveil-
lance Network (NASN) data,  which typically  involve only  one or  two
monitors per city.   Certainly the  small  number of monitors might
explain some of the variability between cities.  However,  being  a
part of the  NASN  system, these  monitors have a much  greater  like-
lihood of  representing  areas  at  least  as  large as a neighborhood
and not  be overly  influenced  by  nearby  sources.   National  Aero-
metric Data  Bank  (NADB)  data  was not  used  because these  monitors
are more  likely to  be located near  large sources of lead  and  may
not represent larger-scale  impacts.  Thus, the  presence  of  a  large
nearby source should not  be  the  cause of this  variability.

     In summary, moderate increases  in heavy-duty diesel  particu-
late levels  (2-7  micrograms  per cubic  meter)  will add  to  already

-------
-89-
Table V-4
Estimated 1995 Ambient Levels of Heavy-Duty Diesel Particulate
at 15 TSP Monitoring Sites in Three Cities 22/
Height
City (meters)
New York* 22.9
22.9
18.3
13.7
7.6
Los Angeles 1.2
7.6
27.4
5.5
18.3
Chicago* 9.5
4.6
4.9
39.9
19.2
Distance
from Road
(meters)
91.5
30.5
15.25
30.5
91.5
N/A
1.8
5.0
17.0
N/A
24.4
30.5
21.3
9.15
3.6
Average
Daily
Traffic
12,100
16,500
26,600
17,900
16,800
15,000
15,000
13,500
18,000
N/A
N/A
4,700
9,400
11,600
25,100
Diesel Particulate Levels
(micrograms per cubic meter)
1.8-2.2
1.9-2.3
2.3-2.8
1.7-2.1
2.2-2.7
4.7-5.7
4.9-6.0
5.9-7.2
5.0-6.1
5.4-6.6
8.5-10.4
4.2-5.1
4.5-5.5
4.3-5.3
3.6-4.4
*    The  levels  shown  include a reduction by a factor of 2.66 (New
York) and  1.62  (Chicago) to  account  for an error  in the original
PEDCo analysis.   See text for further description.

-------


Estimated 1995
-90-
Table
Regional Ambient

V-5
Levels of Heavy-Duty Diesel
Particulate in 39 Cities in 1990 22/*
Population
Category
Over 1
million







500,000 to
1,000,000








250,000 to
500,000










100,000 to
250,000










Under 100,000




City
Chicago
Detroit
Houston
Los Angeles

New York
Philadelphia
Average
Boston
Dallas
Denver
Kansas City, MO
New Orleans
Phoenix
Pittsburgh
San Diego
St. Louis
Average
Atlanta
Birmingham, AL
Cincinnati
Jersey City
Louisville

Oklahoma City
Portland
Sacramento
Tucson
Yonkers, NY
Average
Baton Rouge
Jackson, MS

Kansas City, KA
Mobile, AL
New Haven
Salt Lake City
Spokane
Tor ranee, CA
Trenton, NJ
Waterbury, CT
Average
Anchorage
Helena, MN
Jackson Co. , MS
Average
Particulate Level
(micrograms per cubic meter)
2.7-3.4
5.8-7.0
1.9-2.3
4.0-4.9
5.2-6.4
2.0-2.5
2.6-3.1
2.4-2.9
3.3-4.0
1.7-2.1
5.8-7.2
1.8-2.2
1.4-1.7
2.0-2.5
4.0-4.9
1.6-2.0
2.2-2.7
2.3-2.8
2.6-3.1
2.0-2.5
2.4-2.9
1.6-1.9
2.0-2.5
1.8-2.2
3.2-3.9
1.9-2.3
1.6-1.9
2.0-2.5
1.5-1.8
2.2-2.7
2.0-2.5
1.8-2.2
1.6-1.9
0 8-1 0
V* • V .L • \J
1.2-1.5
1.8-2.2
2.2-2.7
1.9-2.3
1.1-1.3
4.6-5.6
1.7-2.1
3.5-4,2
2.0-2.5
1.9-2.3
0.5-0.7
0.8-1.0
1.1-1.3
Based on data from National Air Monitoring System (NAMS)

-------
                                -91-
excessive  regional  levels of  TSP  and  increase  the  difficulty of
complying with the primary NAAQS for  TSP for  practically all of the
regions which have the very worst TSP violations.  As discussed in
the section  on  health effects,  all of  this  additional particulate
burden will  involve  particles  which  are inhalable,  and nearly all
will  involve  particles  with diameters  less  than 2.5 micrometers,
which  are  thought  to have  the greatest  potential  for  affecting
human health.

     3.   Localized Levels

     Approximately  six  studies  are  available  which  examine  the
localized air quality impact  of diesel particulate emissions.  Here
localized  is defined  to  include areas on an expressway,  beside an
expressway  at  distances  up  to  approximately  91 meters  from  its
edge,  and  in  a  street canyon.   These scenarios represent exposure
to:   people  while  commuting  to and from work; persons employed by
roadside businesses  such as gasoline stations;  families  residing
near  major thoroughfares;  pedestrians  on  busy  streets;  and  oc-
cupants of offices, apartments, etc. which flank busy streets.   As
a survey and analysis of  these  studies  has already been performed,
only  the pertinent  results  along with  short descriptions  shall be
discussed here.24/

     Since each  study  utilized  different diesel penetration
rates  and  emission  factors,  these  variables  were  factored  from
their  respective  results  and replaced by the  standard  set of
conditions, described  earlier,  in order  to  be comparable.  For
heavy-duty vehicles, the diesel  emission  factor is 2.0  grams/
mile.  The low and high  diesel  penetration  estimates are 67 percent
and 82  percent  of urban  miles  traveled by heavy-duty vehicles in
1995,  respectively.   An analysis of  urban traffic characteristics
reveals that 93.8 percent of accumulated miles are from light-duty
vehicles and trucks.  The remainder  are,  for the purposes o.f this
study,  attributable   to  heavy-duty vehicles  (based  on DOT  data,
PEDCol9/)-

     A Southwest Research Institute study evaluated the on-express-
way scenario.^/   Positive aspects  of  this report  include:   the
choice  of dispersion model,   GM' s  line source  model,2_5_/  which
yielded good correlation  with tracer  gas experiments;^/  the study
site,   a  portion of  1-45 at  Joplin   (Houston),  where the  wind is
oriented roughly parallel to  the roadway approximately 15 percent
of  the  time (from  2.75°-25.25° relative  to the  road  at  2.06-8.3
meters/second);  and  the  traffic count was  well documented at 1494
vehicles/hour  for  each  of  6   lanes.  The  results, modified to
comply  with  the  aforementioned  standard  emission factors  and
dieselization rates,  can be found in  Table  V-6.

     From  this study it  can be  seen that commuters on an expressway
with  a  traffic volume  of approximately  9000 vehicles  per hour
may expect exposure  to  heavy-duty diesel  particulate at concen-
trations above  regional  levels  of diesel particulate ranging from

-------
                         -92-
                        Table V-6

         Expected 1995 On-Expressway Heavy-Duty
            Diesel Particulate Concentrations
         	(micrograms per cubic meter)	

  2.06 m/sec     2.06 m/sec    ,8.3 m/sec    8.3 m/sec
  at 2.75°*      at 25.25°     'at 2.75°     at 25.25°
 31.9 - 39.1    20.2 - 24.7   23.0 - 28.2   6.7 - 8.2
Wind speed, and orientation with road.
                        Table V-7

         Expected 1995 Off-Expressway Heavy-Duty
            Diesel Particulate Concentrations
         	(micrograms per cubic meter)	

                24-Hour Max        Annual Geo. Mean

  30 Meters     21.0 - 25.8           7.0 - 8.6
   from Road

  91 Meters     13.7 - 16.8           4.6 - 5.6
   from Road
                        -Table V-8

         Expected Street Canyon Concentrations
         	(micrograms per cubic meter)
                24-hour Max        Annual Geo. Mean

1.8 Meters      15.0 - 18.4           5.0 - 6.1
Above Street

9.1 Meters      12.1 - 14.8           4.0 - 4.9
Above Street

27.4 Meters     7.2 -   8.8           2.4 - 3.0
Above Street

-------
                                -93-
6.7-31.9 micrograms per cubic meter.  These values reflect the low
estimate of dieselization. The  high estimate  of dieselization
yields  concentrations  ranging 'from 8.2-39.1 micrograms  per  cubic
meter.  The  wide range  in expected levels  reflects  the important
role  of  the wind.  Higher on-expressway concentrations result when
lower velocity  wind  approaches  a  trajectory parallel  to the  road.
This  condition allows cumulative dispersion  towards receptors
(people  in  cars)  rather than away  from them as  would  be'the case
for steeper road-wind angles.

     To  characterize  the off-expressway  impact, the Aerospace
Corporation utilized  a  number  of  studies which used monitors
to  construct  roadside  spatial  distributions  of  carbon  monoxide
and tracer  gases.J_8_/   Carbon monoxide is  an especially good sur-
rogate  for  ambient  diesel   particulate  level   projections,  since
motor vehicles are  the predominant  contributors to ambient  CO
levels  and  diesel particulate  disperses  more  like  a  gas than  a
typical large  particle.  Their  approach  involved developing  a
pollutant  concentration index  by  subtracting  background  concen-
tration  from  measured roadside  values and  dividing  the resulting
difference by  the appropriate  source  term.    This  process was
repeated for various distances from the roadway.  A roadside diesel
particulate concentration profile was developed by multiplying the
index values  for  specific  locations  by  the  desired  particulate
source term.  The  7850  vehicle per hour traffic count was based  on
a  24-hour  integration of actual  traffic  flow  on  an 8  lane  urban
freeway in Los Angeles.

     This  approach  should  be   superior  to mathematical  modeling
efforts because it is  based on measured trends  and characteristics
while avoiding  such  assumptions  as constant wind  speed  and atmos-
pheric  stability.   The results,   found in  Table  V-7,  are  given  in
terms of  a  24-hour maximum concentration during one year  and  the
corresponding annual  geometric  mean.   In  order  to obtain 24-hour
maximums, Aerospace  chose values  of the  concentration  index  which
corresponded to the 99.73 percentile ((1  -  1/365) x 100%).  Annual
geometric means  were  then  calculated  by dividing  the  24-hour
maximum values by 3.

     To  confirm this  relationship between  the  two sampling times,
the carbon monoxide records of the 8 cities listed in Table 6-1  of
Air Quality Criteria for Carbon Monoxide were  examined ._2_7_/   A
slightly different divisor of 3.16 was obtained when the geometric
mean of the ratio of 24-hour  maximums  to annual geometric means was
calculated.   Since the range  of  individual  ratios is 2.44 (Chicago)
to 5.0  (Washington B.C.), it  is  concluded  that  the factor used  by
Aerospace is  reasonable and  well  within  the scatter of the  data.

     Following  this  methodology,   persons  approximately  30 meters
from  a  roadway  carrying 7850 vehicles  per hour  could  be exposed
to  annual  mean diesel particulate  concentrations  of  7.0-8.6
micrograms  per  cubic meter  from  both light and  heavy-duty  vehi-
cles.   Similarly,  concentrations  at,a  distance  of about 91 meters

-------
                               -94-
from  the  roadway  fall  in  the 4.6-5.6 microgram  per cubic  meter
range.   As mentioned above,  annual geometric mean  values  are
roughly one-third  of the 24-hour maximum values.

     It is  important  to remember  that  all these  local impacts
consider  only  one source.    The  total  concentration that  people
would be  exposed  to would,  therefore, be the  predicted  localized
value plus  the regional or background  value coming  from  other
roadways  nearby  which  was discussed in  the previous  section.
It is also important to  note that  the 91  meter distance used above
to characterize a localized effect  is further from the  road than
many  of the  'regional1 monitors used  to  develop  the  regional
impacts  shown in Tables  V-4 and V-5.  This  does not  mean that  the
regional  impacts described  in  Tables  V-7 and V-8 are instead
localized  impacts.  The regional  monitors  are located near  road-
ways, but  most are elevated and the roads  are not heavily  travelled
relative  to the expressway examined above.  Rather,  the  large
distances  (91 meters)  at which one  can  still find  single  source
effects  (busy  expressway) is  simply an  example  of  the  extent  of
potential  localized effects.

     Aerospace used  the same  methodology employed in the  off-ex-
pressway study to characterize the  street canyon  impact .J_8_/   Data
collected  from carbon monoxide monitors  at various  heights  above
the  street  were used to determine  the  pollutant concentration
indices.  Although it is  recognized  that mathematical models  are
valuable  tools when  trying  to analyze  pollutant  dispersion,  the
Aerospace  approach is  more appropriate when trying  to  study general
trends  and  situations.  By not  relying on such assumptions  as
constant building  height and wind  velocity  this study relates more
directly  to everyday conditions.    Their  results, modified  to
reflect  the  standardizing  assumptions mentioned  earlier,  are  in
Table V-8.   The  traffic count for the street  canyon scenario  was
936 vehicles per hour.

     When  determining  the potential  impact  of  a particular  concen-
tration, it  is important to consider the length of time  people will
be exposed to that level of pollutant.  People who live  and work in
downtown areas (characterized by  the 9.1 and  27.4 meter receptor
heights) will be  exposed for longer  periods  of time  than  those  who
are merely shopping  (pedestrians).   The impact  to  those living  and
working in the downtown area is,  therefore, greater  than  the
pedestrian impact  under  the conditions of  this study.

     In assessing  the  localized impact from  diesels,  it is  benefi-
cial  to compare  predicted  concentrations to  the  National  Ambient
Air Quality Standards  for particulate.   The  primary  standards  are
75 micrograms per  cubic  meter  for  an annual  geometric mean and  260
micrograms per cubic meter  for a  maximum  24-hour  concentration  not
to be exceeded more than once a year.

     Due  to the  highly-specialized  nature  of the  on-expressway
study (designed to represent a worst case meteorology), no  compar-

-------
                                -95-
isons  of  its maximum  31.9-39.1  microgram per  cubic  meter  diesel
particulate levels  to the standards will be  made.  Conditions
favorable for such  levels  will occur less than  15  percent  of the
time.  However,  it would be useful to note that commuters could be
exposed to these levels up  to  2 hours  per day.

     Approximately 30  meters  from the  roadway,  heavy-duty  diesel
particulate will constitute 8.1-9.9 percent of the 24-hour standard
and  9.3-17.5 percent  of the  annual standard.   At the 91 meter
distance,  diesel  contributions represent  5.3-6.5 percent of  the
24-hour standard and 6.1-7.5 percent  of  the annual standard.   It is
important to remember  that these  numbers  reflect the contribution
from a  single  roadway and, therefore,  do  not  consider  background
levels  from other nearby streets and  highways.

     In the  street  canyon, at the  1.8 meter height, diesels  are
responsible for  5.8-7.1 percent of the 24-hour maximum and 6.7-8.1
percent of  the  annual  standard.   At a height  of 9.1 meters  the
percentages are  4.7-5.7  percent  for  the 24-hour  case and  5.3-6.5
percent for the  annual  case.

     These  analyses clearly   indicate  that uncontrolled heavy-
duty  diesel particulate  emission levels  would  have  significant
air  quality impacts on areas surrounding busy streets and express-
ways.   These localized  impacts  would be  in  addition to the  re-
gional impacts  analyzed  in  the previous section  and  would make it
extremely difficult  for  some areas to  comply with  the NAAQS
standards  for  TSP.   The  health  effects  consequences on  persons
who  live,  work,  and  travel in such  areas  would be  even  greater
than those  expected based  on  TSP impacts,  since the small  size
of  diesel  particulate  makes  it especially hazardous  to human
health.

F.   Air Quality Impact of  Regulation

     Beginning  in 1986, emissions from new heavy-duty diesels  will
be reduced 67 percent from  uncontrolled  levels (as per Chapter  IV)..
The  full  impact  of  this regulation on air quality will not be
realized,  however, until older uncontrolled  trucks  (pre-1986)  are
replaced.   By  1995  particulate  emissions  from combined pre-  and
post-control heavy-duty  diesels will  be reduced 64.3 percent  from
218,000-266,500  metric tons per  year to 77,800-95,100 metric  tons
per  year  nationwide.   Urban  emissions  will similarly be  reduced
64.3 percent  from 79,000-97,000  metric tons  per year to  28,200-
34,600 metric tons per  year.

     Table  V-9  shows  the  ambient  levels both  before and after
regulation  of 15  cities having a population  over 500,000  people.
The data have been taken from  Tables  V-4 and V-5 and the full range
has  been  used  when more than  one estimate was  available.   These
impacts  should be indicative of neighborhood  or  larger scale
impacts in  the  cities mentioned.   Any monitors  modeled by  PEDCo
(Table V-4) which  did not meet  EPA's criteria  for the minimum

-------
                     -96-
                 Table V-9

Large-Scale Air Quality Impact on Regulation of
Heavy-Duty Diesel Particulate Emissions - 1995
Heavy-Duty Diesel Ambient
Particulate Level
Population
Category
Over 1
Million




500,000 to
1,000,000







City
New York
Los Angeles
Chicago
Philadelphia
Houston
Detroit
Dallas
New Orleans
Boston
Denver
Pittsburgh
San Diego
Phoenix
St. Louis
Kansas City, MO
micrograms per
Uncontrolled
1.7
4.7
2.7
2.4
4.0
1.9
5.8
2.0
1.7
1.8
1.6
2.2
4.0
2.3
1.4
- 3.1
- 7.2
- 7.0
- 2.9
- 4.9
- 2.3
- 7.2
- 2.5
- 2.1
- 2.2
- 2.0
- 2.7
- 4.9
- 2.8
- 1.7
cubic meter
Regulated
0.6 - 1.1
1.7 -
1.0 -
0.9 -
1.4 -
0.7 -
2.1 -
0.7 -
0.6 -
0.6 -
0.6 -
0.8 -
1.4 -
0.8 -
0.5 -
2.6
2.5
1.0
1.7
0.8
2o6
0.9
0.7
0.8
0.7
1.0
1.8
1.0
0.6

-------
                                -97-
distance from the roadway were excluded from Table V-9.  As  can be
seen, ambient heavy-duty diesel particulate levels will be reduced
by  3.0-4.6  micrograms  per  cubic meter  in  Los  Angeles and 2.4-3.2
micrograms per  cubic  meter  in  Houston due  to this  regulation.

     The impact  of this regulation on particulate levels in  local-
ized  areas  of particularly  high  concentrations  is  also signifi-
cant.   Table V-10 presents  an overview of  this  impact.   (All
concentrations refer to heavy-duty diesel  contributions only.)  On
the  expressway  the diesel particulate  level will  drop from 31.9-
39.1 micrograms  per  cubic  meter  to 11.4-14.0 micrograms per cubic
meter for the 2.06 meters  per second wind speed -2.75° worst case
scenario.   At  a distance of approximately  30  meters from the
roadway, the  maximum 24-hour particulate  levels  are reduced from
21.0-25.8 to 7.5-9.2  micrograms  per cubic meter.  This  reduction in
the  heavy-duty diesel  particulate  levels  will  benefit such  people
as  service station operators who  spend large  amounts of time
near roadways.  Similarly,  the maximum 24-hour  particulate exposure
level to people  residing  approximately  91  meters  from the roadway
is  reduced  from 13.7-18.8 to 4.9-6.0  micrograms  per cubic   meter.

     Although heavy-duty trucks may  constitute a smaller fraction
of  the  central  business  district  VMT  when  compared to urbanwide
VMT,  other  heavy-duty  diesel vehicles, such as  buses and  garbage
trucks,  are in wide  use  there.   Thus,  people  residing in downtown
areas  are  exposed  to heavy-duty diesel particulate as  well.
Without control  in  1995  maximum 24-hour heavy-duty  levels  will be
12.1-14.8 micrograms  per cubic meter at a  height of  9  meters above
the  street.  These levels  will be  reduced to 4.3-5.3 micrograms per
cubic meter if this proposed  rulemaking  is  promulgated.

     As was mentioned  earlier, uncontrolled heavy-duty diesels are
projected  to  be  a significant source  of  particulate emissions by
1995.   In  terms  of projected reduction potential, however,   heavy-
duty diesel may  be even  more significant.   The annual particulate
emission reductions available from heavy-duty  diesels  are actually
close to  the  total annual emissions  from  some entire industries,
such as the iron and  steel industry (see   Table V-3).  Also,  while
further reductions in  stationary  source emissions  can be expected
to mitigate future increases  in  emissions due to industrial growth,
they  cannot  be  expected  to   significantly reduce  total  emissions
from current levels,  making reductions  from heavy-duty  diesels even
more necessary.

G.   Secondary Environmental  Impacts  of  Regulation

     Five potential  secondary areas  of impact will  be discussed:
energy,  noise, safety,  waste, and water pollution.  No significant
impact is expected in any  of  these  areas.

     The control  technology expected to be used  to meet  the
1986 standard  does  not  appear to  affect fuel  economy- either
positively or negatively.  Thus,  there  should  be  no impact   on the

-------
                                        -98-
                                   Table V-10

                   Heavy-Duty Diesel Particulate Levels With and
               Without  Regulation (micrograms per cubic meter) - 1995

                                  On-Expre s sway

                2.06 m/sec wind   2.06 m/sec    8.3 m/sec    8.3 m/sec
                speed at  2.75°*   at 25.25°     at 2.75°     at 25.25°
Without Control  31.9 - 39.1

With Control     11.4 - 14.0
20.2 - 24.7   23.0 - 28.2   6.7 - 8.2

 7.2 - 8.8     8.2 - 10.1   2.4 - 2.9

Off-Expressway
                30 Meters  from Road
                 91 Meters from Road
Without
Control
With
Control
Without
Control
With
Control
24-Hour Max. Annual Geo. Mean 24-Hour Max. Annual Geo. Mean
21.0 - 25.8 7.0 - 8.6 13.7 - 16.8 4.6 - 5.6
7.5 - 9.2 2.5 - 3.1 4.9 - 6.0 1.6 - 2.0
Street Canyon
1-8 Meters Above Street 9.1 Meters Above Street 27.4 Meters Above Street
24-Hour Annual 24-Hour Annual 24-Hour Annual
Max. Geo. Mean Max. Geo. Mean Max. Geo. Mean
15.0 - 18.4 5.0 - 6.1 12.1 - 14.8 4.0 - 4.9 7.2 - 8.8 2.4 - 3.0
5.4 - 6.6 1.8 - 2.2 4.3 - 5.3 1.4 - 1.7 2.6 - 3.1 0.9 - 1.1
     Wind road  angle.

-------
                                -99-
nation's  energy  resources.   Similarly,  this  control  technology
should not significantly  affect engine noise.

     There are potential  safety implications connected with  the  use
of a trap-oxidizer.   It is  possible that the trap-oxidizer could be
damaged  by extreme temperatures if too much particulate were
captured before  burn  off.   These higher  than  normal temperatures
could also represent a fire hazard if combustible material (such as
leaves) were  in  close proximity  to  the  malfunctioning  trap-oxi-
dizer.   This  scenario, however, is considered unlikely since
heavy-duty diesels  typically  operate  on paved roadways  which  are
essentially free   of  such debris.   To the extent  that vehicle
manufacturers  mount  trap-oxidizers on the vehicles'  side,  as is  the
current practice  for  most heavy-duty diesel  mufflers,   this risk
should be further minimized as this location would be away from  the
ground where combustible  materials might be.

     Of course,  any  design  of a device like this  will need  to
adequately ensure that accidental occurrence such as  those depicted
would not affect  vehicle  safety.

     It is  also  possible;: that these regulations  could have  an
impact on solid waste and water pollution.  While disposable traps
are  not  envisioned  as a  likely  control technology,  if  they were
used to collect  the particulate emissions,  these  traps would need
to be discarded into the  garbage, or burned.  If discarded into  the
garbage  and used as  land fill,  some of the  chemical  compounds
present  in diesel  particulate could leach into the ground  and
pollute the ground  water.   However,  this should not  be  more dif-
ficult to solve than the current problem of disposing of used
engine lubricating oil.  Assuming  a typical heavy-duty diesel
engine oil  replacement period  of  10,000 miles, (6214 km),  a 26.5
liter  engine capacity and an oil having  a specific  gravity  of 0.9,
23.9 kilograms (kg)  of oil must be disposed of every 10,000 miles.
If a trap collected 1.3 g/mi (0.81 g/km), this would produce 13 kg
of particulate plus  the trap every 10,000 miles.   Since  the engine
oil  actually  contains some .particulate  from  the cylinder  and  is
essentially all  organic  matter,  while  the majority of the par-
ticulate matter is carbon, the traps  should be less  of an environ-
mental problem than  the existing oil disposal problem.

-------
                                -100-

                              References

I/   "Airborne Particles," National  Academy  of Sciences, November
~    1977, EPA-600/1-77-053,  PB-276 723.

2/   Miller,  Frederick  J.,  et.  al . ,  "Size  Considerations  for
~~    Establishing  a Standard for Inhalable Particles," JAPCA, Vol.
     29, No.  6,  June  1979,  pp.  610-615.

3/   Groblicki,  P.J.,  and  C.R. Begeman,  "Particle  Size Variation
~~    in Diesel Car Exhaust,"  SAE 790421.

4/   Schreck,  Richard M.,  et.al.,  "Characterization  of  Diesel
~    Exhaust  Particulate Under Different  Engine Load Conditions,"
     Presented at   71st  Annual Meeting of APCA,  June 25-30, 1978.

5/   Hare, Charles T.  and Thomas M.  Baines,  "Characterization of
     Particulate and  Gaseous Emissions from Two Diesel Automobiles
     as Functions  of  Fuel  and Driving  Cycle," SAE 790424.

6/   Braddock, James  N. and Peter A.  Gabele,  "Emission Patterns of
~~    Diesel-Powered   Passenger  Cars  -  Part  II,"  SAE 770168 =

7/   Huisingh, J., et.al., "Application of Bioassay to the Charac-
     terization  of Diesel Particulate Emissions," presented  at the
     Symposium  on Application  of  Short-Term Bioassays  in  the
     Fractionation and Analysis of Complex Environmental Mixtures,
     Williamsburg, VA,  February 21-23,  1978.

&J   Earth, D. S., and  Blacker, S. M. ,  "EPA's Program to Assess the
     Public Health Significance of Diesel Emissions," Presented at
     the APCA National  Meeting, June  28,  1979.

9f   Visibility  Protection for Class  I Areas, The Technical  Basis,
     Washington University,  Seattle,  Prepared  for Council  on
     Environmental Quality,  Washington  D.C.,  August  1978,  Pb-288
     842.

10/  Stern, A.C.,  Air Pollution,  Volume  II,  3rd Edition, Academic
     Press, New  York,  1977,  p.  10.

ll/  Emission of  Sulfur-Bearing  compounds from Motor  Vehicle  And
     Aircraft Engines,  August 1978, EPA-600/9-78-028.

12/  Based on telephone conversation with  Alan P. Waggoner, Univer-
     sity of  Washington, March  1980.

-------
                               -101-
13/  Pierson, William R. and Philip A. Russell,  "Aerosol  Carbon  in
     the  Denver  Area  in November  1973" Atmospheric  Environment,
     Vol. 13, No. 12,  1979,  pp.  1623-1628.

14/  "National  Air  Quality  and  Emissions  Trends  Report,   1975,"
     OAQPS,  OAWM, EPA,  November  1976,  EPA-450/1-76-002.

15/  "National Air  Quality,  Monitoring,  and  Emissions  Trends
     Report,  1977" OAQPS,  OAWM,  EPA,  December 1978,  EPA-450/2-78-
     052.

16/  "National  Air  Quality  and  Emissions  Trends  Report,   1976,"
     OAQPS,  OAWM, EPA,  December  1977,  EPA-450/1-77-002.

17/  "Study of Particulate  Emissions  from Motor Vehicles,"  Report
     to  Congress,  Draft,  Southwest  Research  Institute  for EPA.

18/  "Assessment  of  Environmental  Impacts  of  Light-Duty Vehicle
     Dieselization,"  Draft,  Aerospace Corp. for  DOT, March  1979.

197  "Air Quality Assessment of Particulate Emissions  from Diesel-
     Powered  Vehicles,"  PEDCo  Environmental  Inc.  for  EPA,  March
     1978, EPA 450/3-78-038.

20/  "Mobile Source Emission Factors," EPA, March  1978, EPA-400/9-
     78-005.

21/  "The Impact of Future  Diesel Emissions  on the Air Quality  of
     Large  Cities," PEDCo  Environmental  for the EPA, Contract No.
     68-02-2585,  February 1979.

22/  Reiser,  Daniel,  "An  Investigation  of  Future Ambient  Diesel
     Particulate Levels Occurring  in Large-Scale  Urban Areas," EPA,
     September 1979,  SDSB 79-21.

23/  "General Motors'  Response  to EPA NPRM on Particulate Regula-
     tion  for  Light-Duty  Diesel Vehicles,"  April  19,  1979-

24/  Atkinson, R. Dwight, "Localized  Air Quality  Impacts  of  Diesel
     Particulate Emissions,"  EPA,  November  1979, SDSB  79-31.

257  Chock,  David P.,  "A Simple  Line-Source  Model  for Dispersion
     Near Roadways,"  Atmospheric Environment, Vol.  12,  pp. 823-829,
     1979.

-------
                                -102-
26/  "Dispersion  of Pollutants  Near Highways  -  Data Analysis  and
     Model  Evaluation,"  Environmental  Sciences  Research  Laboratory,
     U.S. EPA,  EPA-600/4-79-011,  February;  1979.

211  "Air   Quality  Criteria  for  Carbon  Monoxide,"  Environmental
     Health Service,   Public  Health Service,  Dept.  of  HEW,  March
     1970,  AP-62.

-------
                               -103-


                            CHAPTER VI

                          ECONOMIC IMPACT

     There  is associated with  nearly  all  emission standards a
cost  of compliance.   In this  chapter, the costs necessary for
compliance with  these  regulations  are examined and  analyzed.
The primary cost  involves  the development and installation of
emission control  technology and  hardware on the diesel vehicles.
Lesser costs are incurred by the emissions testing required for EPA
certification, which  include the  purchase  of new  instrumentation
and equipment  required  for  the measurement  of  particulate  emis-
sions.  All of  these  costs  are borne by the manufacturer, who, in
turn,  passes  them  on  to  the consumer.  The  manufacturer will also
attempt to  make  a profit on his  investment  and  this  will also be
passed on to  the consumer.  A  return  on  the  manufacturer's invest-
ment  is necessary, even  if  the  investment is  for pollution control
equipment.    Finally,   the  consumer also must  bear  any  additional
operating costs that  may result from the proposed  standards.   All
costs presented in the following  sections will be in terms of 1980
dollars.  The economic impact of  alternate approaches can be found
in Chapter  VIII,  Alternative Actions.

A.   Costs  to Vehicle Manufacturers
     1.   Emission Control System Costs

     The technology necessary to meet  the  1986  particulate emission
standard was discussed in  Chapter IV.   Heavy-duty engines  are
expected to be  able  to meet the 1986 standard with trap-oxidizers
along  with incorporating  the  design features  of those  current
engines  with  low particulate emissions.   The  trap-oxidizer repre-
sents additional equipment and will increase  the  cost of the engine
(and vehicle).  The design modifications,  however,  should not raise
production costs,  except  through  the amortization of  new tooling
and engineering costs.  These design features  of the lower partic-
ulate emitting diesels are present on these engines at no apparent
price differential and should be similarly available to others.  It
is  possible  that some of  these heavy-duty vehicles will  be able
to  use  other  techniques  to meet the standard,  but to be conserva-
tive,  this  economic  analysis  will assume that  all  vehicles will
require trap-oxidizers.

     In  summary, EPA  estimates  the average first price increase of
a  trap-oxidizer system  for heavy-duty  vehicles  to be  $521-$632
(1980  dollars).   The cost  of  the  trap  itself represents  about 80
percent  of this  total.   Necessary modifications  to the engine and
exhaust system represent  10 percent of the total  cost.  The remain-
ing costs  are  associated  with  the control system used  to initiate
oxidation of the trapped particulate. The use  of the trap-oxidizer
system  as described  in this  section should also  reduce maintenance
costs  by $197  (1980  dollars,  discounted  back to  year  of vehicle
purchase)  due  to reduced  exhaust  system  maintenance.   A detailed

-------
                               -104-
discussion of  the cost  estimates  for  components  comprising  a
trap-oxidizer system is contained  in Appendix  II  of this document.
It is suggested that Appendix  II be read  to understand the details
of the methodology  behind  the  cost estimates.  This  section (Sec-
tion VI-Al)  will  only  present the costs  estimated for  the whole
trap-oxidizer system  (as  determined  from Appendix  II)  and outline
some of the methodology followed.*

     The  cost analysis covered  the  first five years following
implementation of  the  regulation,  1986-1990.  It was  assumed that
the  trap-oxidizer  and  other components would be  produced  by three
outside  suppliers,  each having  a third  of the  heavy-duty  diesel
engine sales  market.   For  purposes of this  analysis,  a 12 percent
learning curve was used,  which  means that the cost of a  trap-
oxidizer  system will increase  12 percent  each  time  the accumulated
production is halved.

     The  costs  of  the trap-oxidizer  systems  are  shown  in  Table
VI-1, where  it  can be  seen that   costs appear  for  four  groups  of
vehicle classes.  The first group  consists of Classes  IIB,  III,  and
IV.  The  costs estimated for this  group corresponds to the average
size of trap-oxidizer systems  fitted to the three vehicle classes.
The remaining groups consists of Classes  V,  VI,  VII, and VIII.  The
trap-oxidizer system for these  groups  were sized with  a trap to  fit
a  Class  VIII vehicle,  and  it  was assumed  that a  Class  VIII trap
would be  used for Classes V, VI, and VII  vehicles as  well.  Origi-
nally,  it was assumed  that  different  sizes of traps  would be used
for Class V-VI, Class  VII,  and Class  VIII  vehicles.   However,  the
low production volumes  involved with  the  Class V-VI  and  Class  VII
traps caused these traps to be  more expensive than the larger Class
VIII traps,  even  when the  effect  of  trap size on  costs  was taken
into account.  Thus, it was  assumed that  the industry would follow
the  least-cost  approach and  equip the smaller vehicles  with  the
larger traps.

     A range of  costs  are  shown for each  of the two  groups in Table
VI-1.   This  is  due to  possible   variations  in  the  trap-oxidizer
systems that could be  used.   For example,  the cheapest system could
consist of a trap, stainless steel  exhaust pipe, electronic control
unit, sensors,  and a  throttle body  and  switch.    The  next  higher
costing system could be the same as the previous system without  the
electronic control  unit and sensor,  and  with the addition of port
liners,  insulated  exhaust  manifold,  and  mechanical control.   The
most expensive system could  include the combined  components  of  the
first two systems.   This   range  of  costs  will  decrease  in later
years because of an increase in cumulative  production from 1986 to
1990.  The  fleetwide-average cost for each year  is  then  a  sales-
weighted  average  (based on the sales  scenario  in  Table  A-II-1  of
*   Costing methodology  was based  on a  study by LeRoy  Lindgren,
"Cost Estimations  for  Emission Control  Related  Components/Systems
and Cost  Methodology  Description" EPA-460/3-78-002.    See Appendix
II for a more detailed discussion.

-------
                         -105-
                       Table VI-1

        Estimated Costs of Trap-Oxidizer Systems
     At Predicted Production Volumes (1980 dollars)

               Vehicle Class     Vehicle Class Average

1986           IIB, III, IV             551-644
                   V, VI                611-688
                    VII                 652-805
                   VIII                 642-789

Sales Weighted:                         629-756
1987           IIB, III, IV             480-562
                   V, VI                540-632
                    VII                 574-709
                   VIII                 564-695

Sales Weighted:                         552-670
1988           IIB, III, IV             438-513
                   V, VI                497-584
                    VII                 530-649
                   VIII                 520-643

Sales Weighted:                         508-618
1989           IIB, III, IV             410-480
                   V, VI                472-553
                    VII                 502-622
                   VIII                 495-612

Sales Weighted:                         482-586
1990           IIB, III, IV              384-451
                   V, VI                 449-527
                    VII                  480-596
                   VIII                  472-588

Sales Weighted:                          458-559

-------
                             -106-
Appendix II) of  costs for  the two  basic vehicle groups.   The
fleetwide average  cost  in  1986  is  then $629-$756 and  should
decrease to $458-$559  in 1990.

     Sensitivity analyses  were performed to examine the effects of
the  two  assumptions  made  above;  1)  that  three  outside suppliers
would produce the trap-oxidizers and 2) that a 12 percent learning
curve would  apply (see Appendix  II-Section C).    To  examine the
effect of  the  first  assumption, a  new assumption  was  made,  that
each manufacturer would produce his own trap-oxidizers.   This is a
worst case assumption since the production volumes  involved are as
small as  practically  possible.    Also,  from  catalyst  production
experience,  it  is highly unlikely  that   each  manufacturer would
attempt  to produce  his  own  traps,  due  precisely to the small
production volumes, involved.   The  analysis  showed  that trap-oxi-
dizer costs were  not  highly sensitive to  the  number  of supplies.
Under the  worst possible  situation, industry-wide costs would only
increase 5 percent and the cost to the smallest manufacturer would
only  be  23 percent  higher  than the  industry  average.   The  12
percent   learning  curve was  again  used to  adjust  for  changes  in
production volume.

     To  examine  the effect of  the 12 percent learning curve, it was
assumed   that there was  no learning curve  and  that  costs  would be
the  same  at  any  production volume.   This  change did  affect  cost
significantly,  particularly  in the early  years.    In  1986, trap-
oxidizer costs  were about  37-62 percent lower  than  those  shown in
Table VT-1.  However,  by 1990,  the difference had narrowed to 15-40
percent.   The examination  of  a  flat  learning curve in this analysis
stems from the judgment that,  if  the  12 percent learning curve is
in error,  then  it errs by being too steep.  Thus, the percent cost
analysis is  conservative in  this  respect.   If additional  data
becomes   available during  the  comment  period  with  respect  to the
level of  learning curve operating  in  this  area,  the analysis  will
be adjusted accordingly.  However,  at  this time, 12 percent is the
best estimate available.

     2.    Certification Costs

     Certification is  the  process in which  EPA determines whether a
manufacturer's  engines  conform  to  applicable  regulations.   The
engine manufacturer must prove  to EPA  that  its engines are designed
and will be built such that they are  capable of  complying with the
emission  standards  over  their full  useful  life.   Certification
begins by  a  manufacturer  submitting an application  for certifica-
tion  to  EPA  and  is  followed by a  2-step  process which determines
the emissions of  the  engine over its useful  life.

     The  first  step  involves the  determination of preliminary
deterioration factors  for the  regulated pollutants.   These  deteri-
oration  factors  must  be  multiplicative  in nature.   The  engine
manufacturer  may determine these preliminary deterioration  factors
in any  manner  it deems necessary  to  ensure that  the preliminary

-------
                                 -107-
deterioration factors it submits  to EPA  for  certification purposes
are accurate  for  the full useful  life.   Manufacturers must state
that  their  procedures  follow  sound engineering practices  and
specifically  account  for  the  deterioration of EGR, air pumps,  and
catalysts as  well  as other critical  deterioration processes which
the manufacturer may identify.   In addition, when applicable,  the
manufacturers must  state  that the allowable maintenance intervals
were followed in determining the preliminary  deterioration factors.
The manufacturers  would submit preliminary  deterioration factors,
based on the  definition of useful  life,  in each case where current
certification  procedures require testing  of a durability-data
engine.   Beyond  these  requirements,  EPA  would not approve or
disapprove  the  durability  test  procedures  used  by  the  manufac-
turers .

     Step two involves  emission-data engines.  One  to two diesel
engines  will be chosen for each engine family.  These engines would
be operated  for 125 hours  in a procedure designed by  the manufac-
turers before the  emission  test.   The  preliminary deterioration
factor will  be  multipled  by  the 125-hour emission test results to
predict  whether the  emission  data  engines would meet the standards
for their full useful  life.   If the emission-data  engines  are
predicted to pass  the standards over  the  full useful life, then the
engine family  is  granted  certification.   The number of  engines
expected to  undergo both types of testing is  shown in  Table
VI-2.JY

     For  the  purpose of this  cost analysis,  the  following assump-
tions are  reasonable based on  past  practice.   Manufacturers will
certify one  emission control  system per  engine family  resulting in
the need for one  set of preliminary deterioration  factors  per
family.   EPA  will  select two  emission-data engines for each diesel
family, \J since each manufacturer  will develop its own preliminary
deterioration factors.  As a  base  estimate,  EPA has assumed that a
manufacturer  will   follow  the former EPA durability  procedure.2_/
For a  diesel engine,  this procedure covers 1,000 hours  with an
emissions test  each  125 hours  plus tests  associated with scheduled
maintenance.

     In  order to  estimate  certification  costs,  unit  costs  must be
known for each  of  the following:   1)  an emission data engine test
with and  without  particulate  testing,  including  the  required  125
hours of  service   accumulation  plus   the  prototype  engine  and, 2)
preliminary  deterioration  factor  assessment with  and  without
particulate  testing,  which EPA believes  will  be conducted  in  a
manner similar  to  the  former pre-production  durability  testing
procedure.    All  certification test  costs  include  transient  and
smoke testing for  diesel engines.

    The  cost  of both types of  gaseous emission testing for diesels
has already  been determinedjY  and is  shown in Table VI-3 inflated
to 1980 dollars.   The  additional  cost of particulate  testing will
be estimated here.   It is  estimated that  the  additional requirement

-------
                              -108-
                                Table VI-2

                   Heavy-Duty Diesel Certification
          Costs Due to Particulate Regulations (1980 dollars)
                                   1986
                          Estimated Number           Estimated Number of
Manufacturer           of Durability Engines 12/   Emission Data Engines 12/

GM                              9                            18
Cummins                        19                            38
Caterpillar                     9                            18
Mack                            4                             8
IHC                             5                            10
Deutz                           -                             2
Isuzu                           -                             2
Fiat                            -                             2
Mercedes                        3                             6
Mitsubishi                      -                             1
Scania-Vabis                    -                             1
Volvo                           -                             3
Hino                            -                             1
Total                          48                           110

                           1987-1990

Total                           5                            11
Direct Cost per              $975                          $100
Engine Tested Due
to Particulate Testing

Indirect cost per            $145                          $ 15
Engine Tested Due
to Particulate
Measurement
        Total Certification Costs Due to Particulate Regulation

                       1986              $65,000

                       1987-1990         $ 7,000

-------
                                -109-
                             Table VI-3

         Unit Costs of Certification Tests (1980 dollars) I/
        Test
1.   Preliminary deterio-
     ration factor assess-
     ment .27

2.   125-hour emission
     data engine.3/
 Gaseous         Gaseous and
Emissions    Particulate Emissions
 $114,000          $114,975


 $ 21,600          $ 21,700
I/   Includes  transient  and  smoke  emissions.   Source:   Ref.  12/
2j   Assumed manufacturers  follow past EPA procedures,  but  this is
not mandatory.
3/   The manner  in which  the 125-hour break-in period  is  carried
out is at the manufacturers' discretion.

-------
                             -110-
of particulate measurement  will  require one extra  technician  hour
during testing (including weighing of particulate  filter),  and one
extra technician hour  for  data processing. -  Filters are the  only
material that must be renewed  for each  particulate  test.  Assuming
that technicians  cost $30/hour, and  filters cost  about  $3  per  test,
the total incremental cost  is $63  per emissions  test.

     In  addition  to  this direct  cost,  the measurement of  partic-
ulate emissions could increase testing  costs by  affecting the  void
rate occurring with emission testing.   Based on  current EPA exper-
ience in  testing  light-duty diesels  for gaseous emissions, a  void
rate  of 20 percent  on  such tests  is  typical.^/  This rate  is
expected to decrease  in  the future  as  experience  with the diesel
procedure increases.   When  other  disqualifiers are  included (e.g.,
manufacturer  and  administrative  errors,  lack  of correlation  with
previous  tests,  etc.),  the current  overall  retest rate  becomes
about 50 percent.^/    The  addition  of  a particulate measurement
system is expected to increase  this retest rate by  5 percent,  to  an
overall  rate of 55 percent ,_3_/

     The impact of both  the direct  and indirect costs of  testing
for particulate emissions depends on the  number  of  tests  involved.
The direct cost of particulate testing  is $63 per  emissions  test.
In the  course of determining  a  deterioration factor  for a  dura-
bility engine, about ten successful  emission tests  are required.kj
Using an overall void rate of  1.55,  the direct  cost  of  adding
particulate testing  to durability testing would  be  about $975  per
durability engine  (10  x 1.55  x  63).   Only one  emissions  test  is
required for  an emission data  engine.   Again using  a void  rate  of
1.55, the  direct  cost  of  particulate  testing  per emission   data
engine is about $100  (1.55  x 63).  Both  of these  costs  are shown  in
Table VI-2.

     The indirect  costs  of  particulate  testing  can be determined
similarly.    The  addition  of  particulate testing  is  expected  to
increase the  overall void  rate for  emissions  testing from  1.50  to
1.55, or 5 percent based on  the number  of successful tests.  About
2.5  percent  of the  costs  of testing a durability  engine and  1.2
percent  of the costs  of testing an emission-data engine are due  to
emission testing.4/  The rest of  the costs are associated with the
cost of  the engine and  the cost  of  accumulating time on  the  dyna-
mometer.   Thus,  the  addition of particulate testing will increase
the  total  cost of testing  by 0.125 percent  (0.025  x  0.05)  for
durability engines and  0.06 percent  (0.05 x  0.012) for  emission-
data engines.  Using the figures in Table VI-10,   this  translates
into a cost of $145 per durability engine  and  $15 per emission-data
engine.   These  costs  are  shown in the middle  of Table VI-2.

     All that  now remains   is  to  determine the  number of  engines
being certified  under  these  particulate   regulations.   The  years
1986-1990 will be examined  as  these  are the five  years covered  by
the  aggregate  cost  analysis which appears later in this chapter.
Normally,  the  implementation of a new  emission  standard,  such  as

-------
                               -1 li-

the  1986 particulate  standard,  would  require  all  manufacturers
to  recertify  their engines.   However,  in  this  particular situa-
tion,  there  is one  other  action  which  will  already  require  the
recertification of all heavy-duty diesel engines  in 1986.  This is
the  75  percent  reduction NOx standard which Congress has mandated
and which will also be implemented in 1986.  This revised standard
should require most if not  all  heavy-duty diesels to be recertified
in  1986 with or without particulate regulations.   The total number
of  engine  families expected in  this  time   frame has  already  been
determined elsewhere^/ and  are  shown  in Table VI-2.

     In  the  years  following  1986,  far  fewer  engines  should  re-
quire  certification.   As  emission standards  are not  expected to
change between 1987 and 1990, the great majority of engines should
be able to obtain  carryover for these  years.  If  it is assumed that
a heavy-duty diesel engine  is marketed about 10 years before major
redesign, then  about  one-tenth  of the engines  certified  in  1986
should require certification in each year following.   This is  also
shown in Table VI-2.

     The total certification costs  in  each year,  1986-1990, can now
be  calculated  directly  from the figures  in Table  VI-2.   The
certification  costs  due  to these  particulate  regulations  are
$65,000 in 1985 and $7,000  per  year from  1987 to  1990.

     3.   Costs of Selective Enforcement Auditing (SEA)

     The  addition  of  particulate standards  is not expected to
increase  the  number  of  Selective  Enforcement Audit  (SEA)  tests
performed on heavy-duty diesels.    These  engines  would  have  had to
be  audited  for  compliance  with gaseous  emission  standards  in  any
event.   There will be an  increase in the  cost of  these  tests,
however, due to both an increase in the number of voided tests and
an  increase in the number of personnel needed to  perform  each
test. The number  of heavy-duty diesel engine families  which could
be audited each year is 21  for  1986 and  1987, and 22 for 1988,  1989
and  1990.I/   If it is assumed  that all audits  are passed,  then an
average of 12  engines  will  need to  be  tested  in each audit.J_/*  The
cost of testing an engine is about  $l,900,j_/  ($1,750 (1979 dollars)
inflated to  1980 dollars)  with  about  33 percent  of the costs
being due to actual emission testing.47  Using  an overall void  rate
of  1.55, the  total  number  of tests in each year  would  be as shown
in  Table VI-4.  The direct cost  of particulate testing is $63 per
test,  as  discussed in the  previous  section.   The  indirect cost,
*     If an audit is failed, then the need to accurately determine
the average emission  level  of that engine family  could  raise  the
total  number of  engines requiring  testing to 55.   This  would
increase  the economic  impact of  the particulate standard by a
factor  of  4.6  for  that  particular  audit.   However,  since  failed
audits  are  expected  to  be quite rare  (much  less  than 5 percent),
the additional  impact  of failed audits  will be small  (much less
than 20 percent).

-------
                               -112-
due to a higher void rate, is as follows.  The third of the $1,963
per engine  testing  cost  which  is associated  with emission testing
is  increased  by  0.05 out of the former  void rate of  1.50.   (The
overall  void rate  of  1.55 is used rather  than 1.0 because the
number of tests shown  in Table  VI-4 is the actual number of tests
including voids.)   This  fraction comes to $21 per test.  The total
impact of particulate testing is  the sum of the direct and indirect
costs, which  is about  $85  per  engine tested.  Combining this cost
with the estimated number of tests performed  each year (Table VI-4)
yields  the overall cost of SEA testing  each  year  due to  this
regulation.    As can  be  seen  from Table VI-4, the cost rises grad-
ually from  $33,000  per  year  in  1986  to  $35,000  per  year in 1990.

     EPA also expects  that  all manufacturers  will institute a
manufacturer operated production line audit program to measure the
effectiveness  of  the  compliance efforts  and provide  themselves
assurance  of their  ability to  pass a formal EPA audit._!_/   EPA
believes that  in  the first two  years of SEA, 1984 and  19~85,  the
manufacturers may  audit,  on the average,  as much  as 0.6 percent of
their production.    However,  as  they gain greater confidence  in
their SEA compliance efforts  and build engines to achieve the same
emission standards for several  years, this percentage will decline
to  about 0.4  percent by 1986 and remain there through  1990.   The
total cost  of a  production  line audit  for  gaseous  emissions  is
$1,380 per engine  (inflated to  1980 dollars),J./ with  45 percent of
this cost due to actual  emission  testing.4/

     The presence of a  particulate  standard  should not  affect the
number of engines  tested,  but  could  affect  the  cost  of  each test
and the  total number of  tests performed.  Again,  the direct cost of
particulate testing due  to  personnel and equipment needs is $63 per
test  (Section VI-A2).   The indirect cost  is again 3.3 percent
(0.05/1.55)  of 45 percent  of  $1,440 (1380 +  63),  or $21  per
test.  Together the cost of particulate testing  is  about  $85 per
test.

     When this  cost,  the  projected  production  figures  shown  in
Table A-II-1, the  overall void  rate (1.55)  and the above-mentioned
testing  percentages  (0.4 percent  in 1986  and  later years)  are
combined multiplicatively,  the  result  is the  annual  cost  for each
manufacturer.  These annual costs are  shown in Table  VI-5.   As can
be seen,  the total cost  for the entire industry is $103,000 in 1986
and  slowly  increasing  to  $129,000  in 1990.   These  costs  should
continue to increase slowly as  total heavy-duty  diesel production
increases.

     4.    Test Facility Modifications

     These  heavy-duty diesel particulate regulations  will  require
manufacturers to purchase new  equipment to modify their  emis-
sion test cells to  allow for the measurement of  particulate emis-
sions.  It will be  assumed  that  heavy-duty diesel engine manufac-
turers will  anticipate  these  particulate regulations at  the  time

-------
                      -113-
                       Table VI-4

                  Increased Cost of SEA
      Due to Particulate Regulation (1980 dollars)
Year
1986
1987
1988
1989
1990
Number of
of Audits
21
21
22
22
22
Number of
Emission Tests
391
391
409
409
409
Increased
Cost of SEA I/
$33,000
$33,000
$35,000
$35,000
$35,000
I/   Based on a total direct and indirect cost of $85 per
emission test.

-------
                               -114-
th ey  purchase  equipment  for the  new transient test  procedure
becoming effective  in  1984-1985.   In  the  preamble  to the regula-
tions  implementing  the transient test  procedure,^/  EPA announced
that  these  particulate  regulations were being  developed for
the  same  timeframe and  that  the gaseous  emission  test procedure
had  been  modified  to  allow  for  equipment  needed  in particulate
testing which  was also  central to  the measurement of  gaseous
emissions.    In  this  way,  heavy-duty diesel  manufacturers   could
design their testing systems for particulate testing right  from the
start  and  no money would be wasted  on modifying newly-installed
test equipment  before  it  had ever  been used.  Because  of  this, only
that particulate-testing equipment which  is needed  in addition to
the  basic  set-up  required  for gaseous  emission testing  will  be
taken to be additional  requirements  of this  particulate regulation.
This basic  set-up  for  gaseous  emission testing will  include  those
particulate-oriented devices  which  are also  central  to  gaseous
emission testing  and  which  essentially  replace  a  similar device
which  would  not be  allowed when testing  for  particulate (e.g.,
dilution tunnel over mixing  box).

     Table  VI-6  shows  the costs  of  the  additional  test equipment
needed to measure particulate emissions.  The costs  of two systems
are shown;  a single dilution  system and a double dilution system.
Both are allowed under  the  proposed  test  procedure.   If a single
dilution system is  involved,  the major additional  expense is due to
the larger volume sampling  system (CFV or  PDF) required.  A major
additional  cost of  a double  dilution  system  is  due to  the secondary
dilution tunnel and related  measurement  devices.   The  total cost of
the double  dilution system is $34,000 more than the cost of equip-
ment required for gaseous  emission testing,  while  the  total cost of
the single  dilution system is  $61,000  to  $70,000 more  than the cost
of  equipment required  for gaseous  emission testing.   It  will  be
assumed that manufacturers will install the double dilution system
since  it is  the  least  expensive of  the two systems  shown in Table
IV-14.  The  incremental  cost  of  test  equipment  per  modified cell
due to this particulate regulation would  then be $34,000.  A filter
weighing system  for  each test facility  (with  either  type system)
will cost an additional $33,000.

     The expected  number  of  test cells and  facilities  requiring
modification is shown  in  Table VI-7  for each manufacturer._!/  Only
those test  cells designed  for emission testing have been  assumed to
require modification and not those used for  service accumulation.I/
Coupling these  figures  with  the above  costs  yields the overall cost
to each manufacturer,  and  these are  shown in Table VI-5.  The  total
cost  to the industry will  be $2,056,000  (1980 dollars).   The
estimated  number  of  test  cells and facilities includes  those
required for SEA testing.

B.   Costs  to Users  of  Heavy-Duty  Diesels

     Purchasers of heavy-duty  diesels initially  will  have to pay
for the costs  of any  emission control  equipment  used to meet the

-------
                           -115-
                            Table VI-5

               Cost of Self-Audit Programs  Due to
       Particulate Regulation (Thousands of 1980 Dollars)
Manufacturer
Cummins
Detroit Diesel (GM)
Caterpillar
Mack
Internat ional
Deutz
Isuzu
Hino I/
Fiat
Mercedes
Mitsubishi
Scania-Vabis
Volvo
Total
1986
48
40
29
22
11
1
1
0
1
3
0
0
1
157
1987
51
42
31
24
12
1
1
0
1
3
0
0
1
167
1988
54
45
32
25
13
1
1
0
1
3
0
0
1
176
1989
57
47
34
27
13
2
1
0
2
3
0
0
1
187
                                                            1990

                                                             60
                                                             50
                                                             36
                                                             28
                                                             14
                                                              2
                                                              1
                                                              0
                                                              1
                                                              4
                                                              0
                                                              0
                                                              1

                                                             197
I/   Zeros signify less than $500.

-------
                                   -116-
                                 Table  VI-6

                   Cost  of  Modifying  an Emission Test Cell
         for  the Measurement  of  Particulate Emissions (1980 dollars)
 Cost Per  Item  In  Test  Cell

 Constant  Volume Sampler
     CFV-CVS 1/2/
     PDP-CVS 1/2/

 Heat Exchanger 3/

 Secondary Dilution
 Tunnel

 Sampling  Equipment  for
 Tunnel  (Probes,  Filter
 Holders, Flowmeters,  etc.)

 Total Cost per Cell

     CFV-CVS
     PDP-CVS

 Cost per  Item  at Test  Facility

 Microgram Balance

 Weighing  Chamber

 Additional Cost per
 Facility
Single Dilution
   31,000
   40,000

   25,000
    5,000
   61,000
   70,000
   11,000

   22,000

   33,000
Double Dilution


      $0
      $0

    25,000

     1,500


     7,500
    34,000
    34,000
    11,000

    22,000

    33,000
_!_/   Costs which are incremental to the system  specified  in
the 1984 heavy-duty gaseous emissions regulations.
_2/   Size of single dilution tunnel system is 6000CFM.  Size
of double dilution tunnel system is that which  is required by
the 1984 heavy-duty gaseous emission regulations.
_3/   Includes price of water conditioning features  in heat ex-
changer jackets.

-------
                             -117-
particulate emission  standards  plus  the cost of certification and
SEA which  includes  the  cost of new particulate measurement equip-
ment.    The vehicle manufacturers pass  on  these  costs  to the pur-
chaser  by increasing the  "first  cost" or  sticker price  of the
vehicle.

     To calculate these costs, an estimate of the number of heavy-
duty diesels  which  will  be sold each  year  is  needed.   EPA's best
estimate of diesel penetration can be  found  in Section III of this
analysis.   The estimates of total heavy-duty sales were determined
from an  analysis of  1967   to  1978  sales of heavy-duty vehicles.
Projected  sales of all heavy-duty vehicles for 1979 and beyond were
based on  a growth  rate determined  by a regression analysis  of
the 1967  to  1978 data.   It  was estimated that  sales of  Class
IIB, III,  and IV vehicles would  be 20 percent diesel by  1990,
sale of Class  VI  school  buses would  be 10 percent diesel by 1990,
sales  of Class VI trucks would be 35 percent diesel by 1985, sales
of Class VII  vehicles  would be entirely diesel  by 1988, and sales
of Class  VIII vehicles would  be entirely diesel by  1984.   These
diesel  penetration  rates  by  class  were then  combined with  the
projections of overall sales to yield diesel sales by class,  which
are shown  in Table III-9.

     The costs  of this regulation to  users  of  heavy-duty  diesels
can now be calculated and  are  shown  in Table  VI-8.  The  cost  of
test equipment modifications were assumed to occur in 1985  and all
certification costs were assumed to occur during the year prior  to
that model year.  SEA costs were assumed to  occur during the model
year in question.  A 10 percent discount rate was used to determine
the present  value of  these sets of expenditures in  1985.   These
costs  were then amortized over 1986-1990 diesel production to yield
a constant cost per vehicle.  These five years were chosen because
they are  the  first  five years  that  the standard  is  in  effect  and
the  five   years  over  which  the aggregate cost  is determined.
Expenditures  were  assumed   to  occur  on  January  1  of the  year  in
question and  revenues  were assumed  to  occur on  December 31.   The
resulting  cost is $2 per vehicle.

     The cost of emission control hardware are  shown next in Table
VI-8,   (taken  from  Table  VI-1).   Engine modifications  will  cost
$4-16  per  vehicle  throughout  the 5-year period  (see Section D.I.a
of this chapter).   Trap-oxidizer  costs are highest in  1986  at
$629-756 per vehicle and decrease to $458-559 per vehicle in 1990.
Over the  five-year  period,  the sales-weighted  average  is $521-632
per vehicle.

     The users of heavy-duty diesels will also  have  to pay for any
increases  in the costs of maintenance or fuel that occur because  of
this regulation.   No  fuel  penalty  is  expected  from the use  of  a
trap-oxidizer or  from any  engine modification necessitated  by this
regulation.

     As the designs  of trap-oxidizers are becoming better known,  it
seems  reasonable  to  expect some maintenance to be  required.   The

-------
                             -118-
                           Table VI-7

        Certification and SEA Test-Equipment Modification
                Cost by Manufacturer (1980 dollars)
Manufacturer
GM
Cummins
Caterpillar
Mack
IHC
Deutz
Isuzu
Fiat
Mercedes
Mitsubishi
Scania-Vabis
Volvo
Hino

Estimated Number
of Modified Cells
6
12
7
4
4
1
1
1
2
1
1
2
1

Estimated Number.
of Facilities
2
2
2
2
2
1
1
1
1
1
1
1
1
Total
With Cost
of Capital*
Total Cost
$270,000
474,000
304,000
202,000
202,000
67,000
67,000
67,000
101,000
67,000
67,000
101,000
67,000
$2,056,000
$2,344,000
One year at 14 percent interest,

-------
                                -119-
                               Table VI-8

                Cost to  the Consumer of Heavy-Duty Diesel
            Particulate  Regulation  (1986-1990)  (1980 Dollars)

               Test Equipment, Certification  and  SEA Costs

                                                               SEA  and
                    Equipment           Certification        Self  Audit

       1985          2,056,000              65,000
       1986               0                   7,000              190,000
       1987               0                   7,000              200,000
       1988               0                   7,000              211,000
       1989               0                   7,000              222,000
       1990               0                     -                232,000

Total  cost (present value in 1985)_!_/ - $2,725,000
Cost per vehicle  (1986-1990 production)^/ - $2 per vehicle

                         Engine Modifications

                  1986-1990                 $4-16  per vehicle

                         Control Hardware Costs
1986
1987
19«8
1989
1990
629-756
552-670
508-618
482-586
458-559
Sales Weighted Average,  1986 -  1990 _3/     $521-632

                      Operating Costs  (1986 and on)

Maintenance increases (discounted to               $19
  year of vehicle purchase)

Maintenance reductions (discounted to             (-$197)
  year of vehicle purchase)

                          Net Cost to Consumer

                 1986 and on               $349-472
I/   Discount rate at  10 percent.
~2/   Amortization weighted to result in an equal cost per vehicle
over the years of production cited.  Discount rate assumed to be  10
percent.  Expenses are assumed to occur on January 1 of the given
year and revenues are  assumed to be recieved on December 31 of the
given year.
_3/   Based on total diesel sales projections for 1986-1990 shown  in
Table A-II-1.

-------
                               -120-
trap  itself  should  still  be maintenance-free,  but  the  oxidation
control  system may require periodic adjustment.   It  is also  pos-
sible  that  a temperature  sensor may  also need  replacement.    It
is  estimated  that  this  type  of maintenance  would  require about
one hour of labor  and  $10  worth  of  parts  and  occur once throughout
the  life of  the vehicle.   At  a labor rate  of  $20  per hour,  the
total  cost would be  $30.   This maintenance  should  occur  halfway
through  the  vehicle's  life  (approximately  9 years)  and  will  be
assumed  to occur after 5  years  of  vehicle operation.    Discounted
back to year  of  vehicle purchase, this  cost would be  $19.

     However,  the  addition of a  trap-oxidizer  system  is  also
expected to reduce  maintenance  in  two  ways.  One, the  system will
include  a  stainless  steel  exhaust  pipe which  will  eliminate  the
normal need to  replace it.  Two,  the  presence of the  trap itself
should  eliminate the  need for  the muf f ler,^_/6/_7_/ which  in  turn
eliminates the need to replace  the muffler.

     In  order to calculate the  savings  resulting from the elimi-
nation of these two maintenance  items,  two pieces  of data  are
needed  for both  muffler  and  exhaust pipe  replacements;  timing and
costs.   These two  items  are  thoroughly examined in Appendix II  of
this document.

     It was found that an average of 1.27 muffler replacements were
necessary for  each  heavy-duty vehicle.   Using  a  10  percent discount
rate,  the  replacement  rate as  outlined in Appendix  II is  equiva-
lent to  0.61  replacements  at the time  of  vehicle purchase, or  to
0.98  replacements  when the vehicle  is  five years  old.  As no
similar  data  could  be found  which  related specifically to  exhaust
pipe  replacements,  these  findings  will  also be  used  for  exhaust
pipes as well  as  mufflers.

     The aftermarket  costs of mufflers  were estimated  to be $136,
$161,  $164, and  $181  for  Class  IIB-IV,  Class V  and VI, Class  VII,
and  Class VIII  engines,  respectively.   The cost  of  labor  and
incidental parts  was estimated  to be 25 percent  of the  cost of the
muffler  (details  in Appendix  II), so that  the total  cost  of a
muffler  replacement  would  be $170,   $201,  $205,  and  $211,  respec-
tively.   Using  the sales  figures  of  Table  A-II-1 of  Appendix
II, the  sales-weighted average  of these  costs is $211  per muffler
replacement.   Undiscounted, 1.27 muffler replacements would amount
to  $268  per  vehicle.  Using  the actual  schedule of replacements
described  in Appendix II and  a 10 percent  discount  rate,  the
savings  from  eliminating  muffler replacements  would be  $129  per
vehicle,  discounted back  to the  year of  vehicle purchase.

     The total cost of exhaust  pipe replacements  with 25  percent
for labor  and  incidental  parts  becomes  $54,  $85, $105,  and  $136
respectively.   Again,  using  the sales   figures of  Table A-II-1,  a
sales-weighted average cost  is  $111 per replacement.   Using  1.27
replacements per  vehicle,  the undiscounted  savings becomes $141 per
vehicle.    Using the  actual replacement   schedule determined  for

-------
                               -121-
mufflers above and a 10 percent  discount rate, the savings resulted
from  eliminating  exhaust pipe  replacements  is  $68 per vehicle
(discounted to year  of vehicle  purchase).  Adding to this  the
savings  determined for  mufflers above,  the  total maintenance
savings  is  $197 per  vehicle  (discounted  to  the  year  of  vehicle
purchase).  This savings is  shown  in Table VI-8.

     The total cost to the consumer can now be simply added up from
the figures of Table VI-8.   For the  1986 models,  the cost  of
owning and using a heavy-duty diesel should increase $349-$472  per
vehicle due  to  these  regulations.   The range of  the  cost  is pri-
marily due  to  the  possibility  of  different  trap-oxidizer  systems
being used on different models.   The actual cost paid by consumers
will  fall somewhere  between these  two costs, depending on  the
complexity of the  trap-oxidizer  system used  on a given model.

C.   Aggregate Costs—1986-1990

     The aggregate  cost  to  the  nation of complying with the 1986
heavy-duty  diesel  particulate  standards  consists of  the  sum  of
increased costs for new emission  control  devices,  new test  equip-
ment,  additional  certification and SEA costs,  and changes  in
vehicle maintenance  requirements.  The cost  of  the  1986  standard
will be calculated over a period of five years, 1986-1990.

     The aggregate cost to the nation is dependent on the number  of
heavy-duty diesels sold during these time  periods.  Any projection
of this  type  will  by nature be rough, due to the many  social  and
economic  factors  involved.    The sales -projections  used  will  be
those shown  in  Table  A-II-1,  plus and minus  10 percent.  The  per
vehicle cost  of this  regulation  will  be  taken  from Table  VI-8.

     The aggregate cost to the nation based  on these sales  projec-
tions  and  per  vehicle costs  is  shown  in Table VI-9.   As can  be
seen,  two aggregate costs are  presented.  Both are in terms  of 1980
dollars,  but the  present value  in  each  case was  determined  in
different years,  1980  and 1986.   Two different years were  chosen
because there appears  to  be two  conventions which  set  the  present
value  reference point  for the aggregate cost.   Some  analyses have
used one and some the other.  One is the current year of analysis,
or "the present."   In  our case this year is 1980.  The other  is  the
year the regulation becomes  effective, which here is 1986.   The  two
figures are exactly equivalent.   They are related by  the  discount
rate  (10  percent  per annum)  to  the  sixth power  (1986-1980  =  6).
The present value of the  five-year aggregate  cost  in  1980  is
$249-413 million and  relates  most  closely to the cost  to  society
today.  The present value of  the  five-year  aggregate  cost  in 1986
is $442-731 million and relates most closely to the cost  to  society
in the year the regulation becomes effective-1986.

-------
                           -122-
                           Table VI-9
           Aggregate Cost to the Nation of Heavy-Duty
          	Diesel Particulate Regulations	
                             Per
                        Vehicle Cost  Sales
                      Aggregate Costs
                      (1980 Dollars) I/
1986-1990 Model Years
$349-472   1.5-1.8   $249-413 Million _2/
           Million
                                               $442-731 Million 3/
\J   Ten percent  discount  rate used.
2]   Present  value  in 1980.
3/   Present  value  in 1986.

-------
                               -123-


D.   Socio-Economic Impact

     1.    Impact on Heavy-Duty Engine Manufacturers

     This regulation will affect  diesel  engine manufacturers  in  two
ways.  First,  the  engine manufacturers  will be faced with capital
expenditures  for  test  equipment   and  certification,  and also
possibly  for  some  engine redesign  and trap-oxidizer  production.
Second,  the cost of emission control systems could affect  sales  and
in  turn  affect  the  profitability of the  company  and  employment.
These effects will  be investigated below.

     a.    Capital Expenditures

     Capital expenditure  is  money spent for replacing, expanding,
and  improving  business  facilities.    Capital expenditures include
the cost of machinery and equipment  used in production,  the cost of
research  and  development, engineering  and product  launching,  and
the  costs  of certifying  to  EPA  emission  standards.   Capital  ex-
penditures do not  include  operating costs or product  material
costs.   The capital expenditures arising  out  of this  particulate
regulation  fall  into two categories.   First,  the  engine manufac-
turers will  need to modify  their current  test  cells to allow  for
particulate measurements  and  to  test their vehicles  in  1986  for
particulate emissions  when  they  would  have only had  to  test  for
gaseous  emissions  and smoke.   Second, engine manufacturers or
outside  suppliers  must  pay  for tooling, R&D, machinery, land,  and
other capital  costs  involved with producing aftertreatment devices
and possibly in the partial  redesign of  engines.

     Overall, diesel engine  manufacturers  will  have to spend $2.3
million  (1980 dollars) by 1986 to modify their emission test  cells
and  certify 1986 model  year  engines,  including  the cost  of  one
year's borrowing at  14  percent interest.   These initial costs  are
small compared  to  the  initial costs estimated  for  the heavy-duty
gaseous   emission regulations  being  implemented in 1984.  Breakdown
of  the  initial  costs  for  both  regulations are  shown  for each
manufacturer in  Table  VI-10.   (Taken from Tables VI-2 and VI-7 in
this section and from Table  V-DD  in  reference J_/).  EPA has already
determined that the  five   largest manufacturers  (Cummins, GM,
Caterpillar, Mack,  and  IHC)  will  be able to  raise the  capital
involved  for  test  facilities  and certification  costs  of  the 1984
gaseous  emissions regulation.I/  Thus, the  small  additional cost of
this particulate regulation  Tabout  2.5-3.4 percent  of the initial
cost for  the  gaseous emission regulations) should  not be trouble-
some for these manufacturers.  While  it is  possible that the  small
additional  initial  capital  requirements  of this particulate  regu-
lation could be the proverbial "straw that  broke  the camel's back,"
this does  not seem  to be a likely  possibility.   The  amounts of
monies involved  are  simply  very  small,  both absolute and relative
to the 1984 requirements.

     Small-volume  manufacturers  may  experience  a  greater  impact
than the  large  manufacturers on  a  per  vehicle  (sold in the  U.S.)

-------
                              -124-
                            Table  VI-10

           Initial  Investment Required by Heavy-Duty Diesel
            Engine Manufacturers for Diesel  Particulate
           and Gaseous Emission Regulations  (1980 dollars)
Cost for Test Cell
Modification and 1984
Certification due to
Manufacturer Gaseous Regulation I/
Cummins
GM (Detroit Diesel)
Caterpillar
Mack
IHC
Deutz
Isuzu
Fiat
Mercedes
Mitsubishi
Scania Vabis
Volvo
Hino
$17
$1
2
$12
$
$
$
$
$
$
$
$
$
$
6
6
1
1
1
3
1
1
3
1
,477,
,246,
,537,
,908,
,744,
,562,
,562,
,562,
,429,
,065,
,065,
,021,
,'065,
000
000
000
000
000
000
000
000
000
000
000
000
000
Cost for Test Cell
Modification and 1986
Certification due to
Particulate Regulation
$500
$282
$316
$207
$209
$ 67
$ 67
$ 67
$105
$ 67
$ 67
$101
$ 67
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
JY   Costs are the sum of 1979 initial certification costs and 1979
test facility modification costs for certification and SEA, taken
from Table V-DD (p. 117-118) of reference _!_/.  The costs as they
appear in this table are inflated to 1980 dollars, using an 8
percent inflation rate.

-------
                                -125-
basis, but on an absolute basis,  these manufacturers  should  be  able
to raise the small amounts of capital involved (at  most,  $105,000).
All the  small volume  diesel  engine manufacturers are  foreign-based
with  a  small percentage of  their  sales  exported to  the United
States.   On  a worldwide  basis,  their profits should be more  than
sufficient to cover the  initial  investments  for  1986  certification
and test facility modifications of this regulation.

     Small volume truck and bus manufacturers should not  experience
any disadvantage,  since  they will only  see  an increase in  engine
prices  as they  purchase engines and this  increase will not be
significantly larger than the  increase seen  by large truck  and bus
manufacturers.

     The  second  area  of  capital  investments  concerns those  capital
expenditures associated  with  production of  aftertreatment  devices
or engine  redesign.   Looking at aftertreatment devices, the major
capital  expenditures  involved  are tooling,  land and building, and
research  and  development expense.   These costs are  contained in
equation  5 of Appendix  II for calculating the retail price  equiv-
alent.   Tooling expenses consist  of four major components:   one
year  recurring  tooling  expenses  (tool bits,  disposable jigs and
fixtures,  etc.);  three-year non-recurring tooling  expenses  (dies,
etc.); twelve-year machinery and equipment expense; and  twelve-year
launching  costs  (machinery  foundations  and other incidental  set-up
costs) which was assumed  to be  10 percent  of the cost of machinery
and  equipment._!_/   Land   and  buildings needed  for new  production
facilities has  also been included for  calculating  emission  control
hardware costs (see Appendix II), and their cost has been amortized
over  40  years.   In most cases,  however, space in existing  facil-
ities  was assumed to have been made available  for  production
purposes and hence is covered  in  the  overhead  costs.  Research and
development  (R&D)  expenses  associated with  aftertreatment  devices
will  include product  development, engineering, and product  launch-
ing.   Tooling,  land and  building expenses will  be referred to as
simply tooling  costs  (unless  otherwise stated),   and  these costs
will be  calculated  separately from  R&D expenses for the remainder
of this section.

     The  tooling costs   associated with  the  production  of  after-
treatment  devices  will   be  calculated first   for the  trap  itself.
The cost of a trap most  closely represents the cost of a  monolithic
oxidation  catalyst without  the washcoat  and  noble  metals and  this
was the basis of the trap costs determined in Appendix II. Lindgren
has calculated  the  tooling  costs  for  a  1.0 liter (63 cu. in.)
monolithic oxidation catalyst  (pp.  114-117).J_/  Lindgren has
also calculated  the manufacturing costs for  1.0 to  6.5 liter (63
cu. in.  to 400  cu. in.)  monolithic  catalysts (pp. 134, 359-360),
and projects no change of capital  costs with this increase in
catalyst  size.   The  size of  a heavy-duty  trap is  still  larger,
between  10 and  12.8  liters,   but the same assumption that  capital
costs do not increase will  still be made.  This projection  appears
reasonable as the  cost of machinery  and space for land  and  build-

-------
                               -126-
ings should follow the function of  the machine  more  closely than a
simple  doubling  of the  part  size.    To  take this projection  into
account,  however,  a  range will  be  placed  around  the  calculated
capital requirements  to  indicate  that some  variation  is  expected.

     While  the  capital investment  is  not  expected  to  change  sig-
nificantly with  trap  size,  it would change with varying production
volumes of  traps.   A manufacturer of traps  could  not  increase his
production  capacity  adequately  without  investing  in  additional
tooling.   However,  tooling costs  can  be  expected to  decrease (or
increase)  at  a  slower rate than  a  decrease  (or increase)  in  pro-
duction volume.  For  purposes of this  analysis,  it will be assumed
that  tooling costs  decrease by  30 percent when the production
volume  is  halved.   Knowing this,  the tooling costs calculated  at
Lindgren's production volume monolithic oxidation catalyst (p.  115)
can be  modified  to  reflect the  tooling costs of heavy-duty diesel
traps  at   EPA's  estimated  production  volume.    Lindgren's  annual
production volume and  tooling costs for  each part of  the trap are
shown in Table VI-11.  EPA's annual  production volume of heavy-duty
diesel  traps  can be  estimated  by  looking   at  the  average  annual
heavy-duty diesel sales figures  of Table  A-II-1.  Assuming one  trap
for each  Class  IIB-IV vehicle  and two traps for each  Class V-VIII
vehicle,  the  number  of  Class  IIB-IV traps  and Class V-VII  traps
sold  from  1986-1990  would be about   162,000  and  2,897,000,  respec-
tively.   The  five  year total is  roughly 3 million,  and  an annual
average would be about 600,000.  It  was assumed  in  Appendix II  that
a trap sized for a  Class  VIII vehicle would also be fitted to Class
V, VI, and VII vehicles as  well.  Thus,  there will be  two types  of
traps, one to fit Class IIB-IV vehicles and one  to  fit  Class V-VIII
vehicles.   The  total  capital  costs  for trap production  is  the sum
of capital  costs calculated for  producing  traps  for  Class  IIB-IV
vehicles  at an  annual production  of 30,000,  and  for producing
traps for Class  V-VIII vehicles  at an annual  production of 570,000.
These total costs are  shown in  Table VI-11  and  have been inflated
to 1980 dollars. The total tooling  cost  for a  trap is  $8  million
including the cost  of capital (14  percent).

     The  tooling costs  for the  remaining  trap-oxidizer system
components are  shown  in  Table  VI-12.   It  was  assumed  as  before
that tooling costs vary  with  production  volume  but  not  part  size.
Once  again,  Lindgren's  projections were  revised to  account  for
EPA's production volume.   For the stainless  steel  exhaust pipe and
for the muffler, EPA's production  volume  is 400,000 and is based  on
3/4 of the heavy-duty diesel fleet  having a  single exhaust  system,
and 1/4  having  a dual  exhaust  system.   The remaining  components
have one  unit  per  vehicle  with  a  resulting production  volume  of
300,000.   As  discussed in  Appendix  II,  the  components other  than
the trap and control units  would vary with each engine design.  In
Appendix  II  it  was  assumed that ten basic  engine  designs would
cover  the great majority  of  heavy-duty diesel   production,  with
Classes IIB-IV allotted two designs, Classes V  and VI  allotted two
designs, Class VII allotted three designs, and  Class VIII allotted
three  designs.   It  was   further  assumed that  an  equal  number  of

-------
Part
                  -127-

             Table VI-11

   Estimated Tooling Costs of Parts
 to Heavy-Duty Diesel Trap-Qxidizer _!_/

             Lindgren's                   EPA's
Lindgren's    Tooling        EPA's      Tooling
 Economic      Costs       Economic       Cost
 Volume    (1977 dollars)   Volume   (1980 dollars) 2/
Converter
Assembly
Shell
Ring
Inlet Cone
Outlet Cones
Inlet Pipe
Flanges
Mesh
Hardware
Substrate
Vehicle
Assembly
Body
2,000,000
2,000,000
4,000,000
2,000,000
2,000,000
2,000,000
4,000,000
2,000,000
10,000,000
2,000,000
300,000
300,000
4,636,000
636,000
222,000
222,000
222,000
222,000
131,000
222,000
106,000
900,000
516,000
51,600
600,000
600,000
600,000
600,000
600,000
600,000
600,000
600,000
600,000
600,000
600,000
600,000
3,676,000
504,000
123,000
176,000
176,000
176,000
73,000
176,000
36,000
714,000
1,073,000
107,000
Modification

Total

Total with Cost of Capital
                                        7,010,000

                                        8,000,000
JL/   From Lindgren's analysis of a 1.0-liter (63 cu. in.) mono-
lithic oxidation catalyst (p. 115), without the washcoat and noble
metals.
2j   This is the sum of capital costs determined for Class IIB-IV
traps at an annual production of 30,000 and for Class V-VIII
traps at an annual production of 570,000.

-------
                                 -128-
                             Table VI-12

               Estimated  Tooling Costs of Trap-Oxidizer
           System Components  for Heavy-Duty Diesel Vehicles

                             Lindgren's                   EPA's
                Lindgren's     Tooling         EPA's      Tooling
                 Economic       Costs  \J   Production      Cost
                 Volume     (1977 dollars)    Volume   (1980 dollars) 2/
 Port  Liners
 Stainless
 Steel Exhaust
 Pipe

 Insulated
 Exhaust
 Manifold
2,000,000


  400,000


1,000,000



1,000,000
Electronic    2,000,000
Control
Unit  (50%
of Total
NOx and Part.)

Sensors 4/

Throttle
Body Actuator 4/
 8,086,000
  (p.  115)

 1,076,000
  (p.  193)

   733,000
(p.  262,  271)
   516,000
  (p.  178)
                865,000
               (p.  299)
600,000     8,000,000


300,000 3/  2,400,000


400,000 3/  1,600,000



300,000 3/  2,030,000
                300,000
              460,000
                             300,000

                             300,000
                              114,000

                              200,000
_!_/   Pages in Lindgren from which these values are  taken  are  shown
in parenthesis.
_2_/   This column is based on a 30 percent cost reduction  for
halving of production, an 8 percent inflation rate, and includes
the cost of capital (14 percent).
_3_/   Production volume for these units occur for two engine designs
of Class IIB-IV vehicles, two engine designs of Class V and VI
vehicles, three engine designs of Class VII vehicles, and three
engine designs of Class VIII vehicles.  The annual production
volume for components of engines with the same design is  about
15,000 for Class IIB-IV vehicles, 35,000 for Class V and  VI vehi-
cles, 13,000 for Class VII vehicles, and 55,000 for Class VII
vehicles, except for stainless steel exhaust pipes, where the
production volume is 4/3 the amount in each group.  The total
tooling costs is the sum of tooling costs determined for  each
engine design group.
47   No estimates made by Lindgren.   These values assume  ratio of
tooling costs to RPE is same as ratio of tooling costs to RPE for
the ECU.

-------
                               -129-
engines  would  be  produced  according to  each engine  design in  a
vehicle  group.   The  annual  production  for  each  vehicle group  is
approximately 30,000  for Class IIB-IV vehicles, 70,000  for  Class  V
and VI  vehicles,  40,000 for  Class VII vehicles,  and  160,000  for
Class  VIII vehicles,  and the annual production of  each  engine with
the same basic  design would  then be about 15,000  for  Class  IIB-IV
vehicles, 35,000 for  Class V and VI vehicles,  13,000  for  Class  VII
vehicles, and about 55,000 for Class VIII vehicles.   For  stainless
steel  exhaust pipes,  the  production volume for each  engine  design
group  would  be  about  four-thirds times  the engine production with
the same basic  engine design  due to the presence  of  dual exhausts
on one-third of the  engines.   For the remaining trap-oxidizer
system components other than  the  trap and  control  units  the  produc-
tion volume  for each  engine  design group is  equal  to the  engine
production.  The tooling costs  must be determined  separately  at  the
production volume for  each  of  these ten  basic  engine design groups.
The total  cost   for  a  trap-oxidizer  component  is the  sum  of  the
tooling costs determined for each design group.  These  total costs
are shown in Table VI-12 including  a cost  of  capital of  14 percent.

     Lindgren  did  not estimate capital  costs  for  the  sensors
or the  throttle-body  actuator.   It was  assumed  that  the ratio of
tooling costs to retail price  equivalents  (RPE) for these  items  was
proportional to that  for  electronic control units   (ECU).   Both
costs  are known  for the ECU;  the tooling cost is  $406,000 and  the
RPE from Table  A-II-4  is  $37-   (Both  of  these  costs reflect  the
allocation of half the  ECU cost  to  particulate control  and half to
NOx control.)  Based  on the ECU  ratio of  $406,000/$37,  the tooling
costs  for sensors and  throttle body actuator, with RPE's of  $9  and
$16, respectively, would then  be about  $100,000 and  $175,000,
respectively.   With  a cost of capital  of  14 percent, these costs
increase to $114,000 and $200,000,  respectively.

     The tooling costs for  the  complete  trap-oxidizer  system  can be
determined by adding the components for  Systems III and  IV of Table
A-II-5,  as these are  the two systems used  in  calculating  the range
of the system RPE  in  Section A of  this  chapter.  The  total tooling
cost for  the industry is $10.4  to  $14.8  million.   To reflect  the
uncertainty contained  in these projections, an added range of plus
and minus  20 percent  will  be  included,  resulting in  an overall
projection of $9-18 million.

     Next,  the  R  & D expenses  will be estimated for  the trap-
oxidizer system.   Two components are  expected to require R&D,  the
trap and the ECU.  Looking  first  at the  trap  itself, it  is expected
that most of the research and development already  taking place  for
light-duty diesel  traps due  to the 1985 light-duty diesel partic-
ulate  standard  will  apply directly to  heavy-duty  diesel traps so
that the entire  process will  not need to be repeated.  However,  a
trap-oxidizer manufacturer will  still   recover  this  R&D expense
from both light-duty  and heavy-duty  diesel sales.   It  is not
possible to  directly  determine  the cost of  R&D  needed to develop
trap-oxidizers.   The  best estimate comes from past experience with

-------
                               -130-
similar developmental efforts.  The most similar development would
appear to be that of the catalyst, again.  Lindgren estimated that
the cost  of  R&D  for  the monolithic  catalyst  was $25 million (1977
dollars, or $31.5 million (inflated  to  1980 dollars).  This will be
assumed to be the total R&D expense for both light-duty and heavy-
duty  diesel  traps.    The percentage of  R&D  expense  allocated  to
heavy-duty diesel traps  can  be determined by  looking  at  the pro-
jected sales data for both light-duty diesel and heavy-duty diesel
vehicles.   Betweeen  1986  and 1990,  about 2-3 million light-duty
diesel vehicles  with  traps will  be  sold annually,_3_/  and  as pre-
viously determined,  about  600,000 heavy-duty  diesel  traps  will  be
sold  annually.   This means  that  15-25 percent of  all traps sold
between 1986 and  1990 will be heavy-duty diesel traps.  Thus,  the
portion due to heavy-duty diesel traps would be  15 to 25 percent of
$31.5 million,  or about  $5  to  8 million (1980 dollars).

     The  R&D costs associated with  the ECU  are  the last  to  be
determined.  Lindgren estimated that the development  of  ECU's  for
three-way catalysts would cost $6 million  (1977 dollars),  or $7.6
million  (inflated  to  1980   dollars).  Most  of  this knowledge
should  be directly  attributable to  ECU's  use to  control  trap
oxidation.   Thus, the  total  R&D  effort  associated with  ECU's  for
trap-oxidizers  should  be much  less,  roughly,  between 25  and  50
percent of $7.6 million, or  $1.9-3.8 million.   Allocating  half  of
this  to  NOx  control  would leave  $1-2  million  due  to particulate
control.    As there  will  be   only one  ECU per  heavy-duty  diesel,
unlike the nearly two  traps per vehicle,  the heavy-duty fraction of
overall production will  be less  than  15-25  percent and close  to
10-15  percent.  Thus,  the R&D costs  associated  with ECU's  for
heavy-duty ECU's  is  $100,000-300,000  (1980  dollars).    Total  R&D
expenses associated  with the  trap-oxidizer come to about $5 to  $8
million.   When  a  14 percent cost of capital is added, the total R&D
expenditure becomes $6-8 million.

     The  total  tooling  and  R&D  cost  associated  with trap-oxi-
dizers  is about  $15 to  26  million and  would be  borne  by  the
entire  industry  of  heavy-duty diesel  manufacturers   should  they
decide to manufacture  their own trap-oxidizer systems.  However,  as
explained in  Section  A and  Appendix II, it appears that it would be
more economical for outside suppliers (about  three) to manufacture
trap-oxidizers.   Therefore, heavy-duty diesel manufacturers should
not have  to  raise all  of this capital, but  most of it  will  be
raised by  suppliers whose  market  will  be greatly  improved by  the
use of  trap-oxidizers  and  who should  be in good  financial shape
because of it.

     The only capital expenditures  remaining  to be calculated  are
those associated  with engine  modification.   In setting the  partic-
ulate  standard  at 0.25  g/BHP-hr  (0.093  g/MJ), the  baseline  was
taken  from  the  average  of  each   major manufacturer's  best engine
(see Chapter IV  for  details).   Implicit in this  decision is  the
expectation that  manufacturers  can make basic improvements on their
"worst"  engines based  on the existing designs  of  their  "best"

-------
                              -131-
engines.  These improvements  should not increase the cost of these
"worst" engines in the  long  run,  since  the  "best"  engines  do  not
currently cost more because  of  better  particulate emission charac-
teristics.   However,  these  "worst"  engines  will need redesigning
to  some  extent and that  will  require capital  investment  in  the
areas of 1)  design and engineering and 2)  retooling  for produc-
tion.   These  capital   investments will need  to be  recovered  and
this could  raise the  cost  of  these engines  somewhat  for  a  few
years.

     Determining  the  capital  costs  involved  in  such  redesign
is  a difficult task  for  a number  of reasons.   One,  the  modi-
fications necessary will vary from engine  to engine  and  cause
the  resulting costs  to vary  similarly.   Two,  and this may  be
the  dominant  reason,  in  any engine improvement program under-
taken  by a manufacturer,  improvements  are made  in more than
just the area  of particulate emissions.  Once a  decision is made to
redesign an  engine, or  some part of it,  full advantage is taken  and
many changes are incorporated affecting performance, fuel economy,
and other regulated pollutants.  Not  only  can some  of these other
improvements  piggyback a decision  to improve the particulate
emission  characteristics  of  an  engine,  but  design changes  to
improve particulate emissions  can  also  fit  into an  existing  re-
design  problem that was  initiated for  other reasons.

     This situation is particularly likely  today as  there  are
a number of factors causing manufacturers  to  look at basic  engine
design.  One  factor is the  stringent 75 percent  reduction  NOx
standard due  in  1986.   As the  level  of this  standard  is clearly
outlined in  the Clean  Air  Act (Section 202(a)(3)(A)(ii)), manufac-
turers  have  been working toward achieving it for  some time and much
of  this effort centers  around  basic combustion chamber and  injec-
tion design.   Another factor is the drive for improved fuel economy
in  these days of ever-increasing fuel  costs.   Some  of this  effort
centers around the  combustion system, but some of it also involves
more external  devices such as turbocharging, improved aftercooling,
etc., which  can also  have  beneficial results in  the emissions
area.  Last, the need  for improved fuel economy is  also extending
the  application of diesel  engines to a wider range of trucks,
particularly  into the  lighter classes of heavy-duty vehicles.
This is resulting  in the development of totally new diesel engines
for  these vehicles  and the  improvement  of  existing  engines  for
wider application.

     A  good example of this  situation would  be work currently
underway  by  Caterpillar  on  their 3406  engine.   In a meeting
with EPA,  Caterpillar described  some  of the  engine modifica-
tions  they  were considering to reduce NOx  emissions.^/   These
modifications  included  high-pressure injection, a separate-circuit
after-cooler  and piston redesign  as  well  as  many  others.   While
these  modifications  resulted  in  a  20  percent  decrease in  NOx
emissions,  fuel consumption  decreased  1-2  percent and particulate
emissions  decreased 50  percent  (as  indicated  by continuous  smoke

-------
                               -132-
measurements) .    These  modifications  will  be  introduced  between
1983 and 1985.

     As  listed  in Table  IV-2  in Chapter IV,  the  particulate
emissions  from this  engine  are already  quite  low,  0.37  g/BHP-hr
(0.138 g/MJ),  the  fact  that  particulate emissions could be reduced
further via  a NOx reduction program  is  significant.   Albeit,  the
reduction was  measured  via continuous  smoke measurement,  as  Cater-
pillar was not yet able to measure particulate directly.   However,
given  the  facts  that  1) the smoke  measurements  were coupled  with
exhaust flow  rate and integrated over the  test,  2)  the  estimated
particulate emission  of the  unmodified engine correlated  well  with
transient mass measurements  taken  at Southwest Research  Institute
and  3) only  one engine  is  involved,  it would  seem  reasonable
to extend  the decrease  in smoke to  decrease  in particulate mass,
although the  degree of  reduction may  differ somewhat.   Thus,  here
is a  case  where  particulate  reductions resulted from basic  engine
modifications which were not intended for particulate control.   In
cases  like this,  of course,  the cost  of the particulate  reduction
is negligible, as the costs  of  the modifications will be  allocated
elsewhere.

     There are other  examples  of this  happening  that are  known  to
EPA.  EPA has tested two redesigned versions of the Cummins NTC-350
engine,  one California version and one 49-state version, with
neither redesign being  performed for particulate control.   Rather,
improved performance  and  fuel  economy appear to be  the main  pur-
poses  behind  the  redesign, with the  California version also being
designed for  low NOx emissions.   In  both  cases, while  both  NOx
emissions  and  fuel  consumption decreased,  so  did particulate
emissions over EPA's transient  cycle.

     While  these  are  two positive examples  of particulate  reduc-
tions accompanying general  engine improvements, this  is unlikely to
always be the case.   For some  engines, design work will need to be
initiated primarily to reduce  particulate.  Other improvements  may
be able  to accompany  these changes  and minimize  the  costs  of
change-over,  but  the primary  reason for the changes will be partic-
ulate emissions.   At  the present time it  is not possible  to  deter-
mine  exactly  how many engines  will  fall into  this category.
However,  as outlined in  the description of heavy-duty diesel  engine
manufacturers (Chapter III,  Section Cl), most manufactures  rely on
only a small number of basic engine designs for the  great  majority
of their sales.  Given that a number of these basic designs already
have met the  0.41 g/BHP-hr  (0.156  g/MJ)  mark and many will  meet
this through  design  changes  occurring before  1986  for other  rea-
sons,  a reasonable estimate would be that 2-8 basic engine designs
may need work specifically  for  particulate control.

     The  cost  of  each of these programs  is equally difficult  to
estimate.   But again, a reasonable guess would be  $2 million  per
engine design, including research, development and retooling.   This
would put the  total cost  to  the industry at $4-16 million.  As  it

-------
                               -133-
is not possible to determine which engines will require this work,
it  is  also not possible  to allocate this  cost  among the various
manufacturers.  Over the entire industry, however;  this cost would
amount to  $2-10 per engine  produced  over  five years  (1986-1990).

     The total  capital  expenditures  due to this heavy-duty diesel
particulate regulation is  the sum of test  equipment,  certifi-
cation, tooling, R&D, and  engine  redesign  costs.   This amounts to
about $21-44 million.  However,  it is  emphasized again  that outside
suppliers are  expected  to  make  trap-oxidizer  systems and  would be
confronted with the  tooling  costs and R&D expenses totaling about
$15-26 million.  Heavy-duty diesel manufacturers will need to raise
capital for  test  equipment and  certification  and engine redesign,
which amounts to $6-18 million.    Thus,  less than half  of the total
capital expenditures involved with a  diesel particulate regulation
would be borne  by  the  heavy-duty  diesel manufacturers.  This is a
small  amount  when  compared to  the capital expenditures for diesel
engines due  to the  1984  heavy-duty  gaseous emissions regulation,
and the manufacturers should be able  to raise the  money involved.

     b.   Sales of Heavy-Duty Vehicles

     The second  area of  impact  of these  regulations  on  manufac-
turers  occurs  in the  area of  increased vehicle  prices due to
emission  control  hardware.   Cash  flow problems  should not be
significant  since  the  money invested  in  emission  control devices
(e.g.  trap-oxidizers)  is recovered  soon  after  from the  sale
of  controlled  vehicles.   The sticker price increase  due  to  these
devices,  though, could potentially affect sales.   Between  1986 and
1990,  projected  price  increases  are  expected to  average  between
$527-$642  per vehicle.   This   represents  about  0.5-3  percent  of
initial  vehicle  prices  based on a heavy-duty  diesel  cost  of
$16,000-140,000 in   1980  (see  the  following  section,  "Impact  on
Users  of  Heavy-Duty Diesels.")    The credit  for maintenance  cost
brings the net cost  to consumer down  to $349-472.  This real price
increase could  affect  sales in two  ways.   Purchasers  of diesel-
powered vehicles  might  switch  to gasoline-powered vehicles.   Or
some purchasers may decide to wait an  additional year before buying
a new diesel.

     It should  be realized  that  the price of  a gasoline-powered
vehicle will  also  increase  by 1986 due  in part to  the new gaseous
emission standards  being  implemented in 1984  and  the reduced NOx
standard in 1986.   Using the same  cost methodology as that used for
trap-oxidizers, a  catalytic converter  system  and  other  costs  of
compliance for the  1984  standards  are expected  to  raise the  total
cost of gasoline-fueled  vehicles  by  about  $394 in  1979 dollars,_!/
or $444 in 1980 dollars.   Operating costs of $259  (1979 dollars)^/
or  $280  (1980 dollars) for switching to unleaded gasoline and
maintenance savings of $176 (1979  dollars)^/ or $190  (1980 dollars)
for  elimination  of  spark plug and  exhaust  system replacements
brings the  total  net  cost to  $534.    (Lifetime fuel  savings for
heavy-duty gasoline vehicles  were  estimated  to  be $788  (1979

-------
                               -134-
dollars)!/ or  $851  (1980 dollars)  with  respect  to heavy-duty
gasoline vehicles with  1979-level emission controls.   However;  no
fuel  savings  were  projected with respect to  uncontrolled  engines.
Thus,  no  such  savings  should be  incorporated  in  this  analysis).

     The cost  per  vehicle  for  a diesel  to meet the  1984  gaseous
emission standards should be about $195  in  1979  dollars,_l_/  or $211
in  1980  dollars.  It  would  appear-  then, that  the  1984  standards
would give diesels a $323 advantage over gasoline engines.   This is
slightly misleading,  however,  because  the  sources   of  these  costs
are  quite  different.    Over 62  percent  of  the  diesel cost  is the
result of  amortized  (5-year)  one-time  capital investments  in test
equipment and  research  and development ,\J  Only 7   percent of the
gasoline engine cost  is of this  type.   Thus, after  five  years  of
these  price  increases,  the diesel costs  will decrease to  $80 per
engine, while the costs  for gasoline  engines  will only decrease  to
$497 per engine.  Here the difference  is $417  per engine.   As  it  is
unlikely that  the  manufacturers will  actually spread  their  fixed
costs  evenly  over  just these  five   years,   the  actual  difference
between diesel  and  gasoline  engine  costs  will  likely be  between
$323 and $417 per engine and last longer than  five  years.   However,
as  can be  seen,  this  difference  negates most of the diesel engine
price increase expected  from this particulate  regulation.

     In 1986,  it is  likely that  a three-way  catalyst  will  be used
on  heavy-duty  gasoline  vehicles  to meet the  reduced NOx  standard.
A  three-way  catalyst  system includes  the  three-way catalyst,  a
feed-back carburetor,  an electronic  control  unit  system,  and  an
oxygen sensor. The  oxidation  catalyst  and the air  pump  already  on
the vehicle would be replaced.    The net increase from 1984 to 1986
is  expected to be roughly $100-$200.  Thus, the  combined  effect  of
all  emission  regulations will impact  diesel  and gasoline  engines
roughly equally.  Any absolute  decrease in diesel  sales should  be  no
greater than any decrease in sales of  gasoline-powered vehicles and
this  particulate  standard  should  be  no  less acceptable than the
Congressionally-mandated  gaseous emission  standards from  this
standpoint.

     With respect  to the  entire economy,  this  regulation  should
have  no  adverse effect.    If  sales  of  heavy-duty   diesels  should
decrease somewhat due to the  increase  in vehicle prices  resulting
from  this  regulation,  the increase  in  jobs and  sales from the
production  of trap-oxidizers will more  than make up for  any losses
in  the  heavy-duty  industry itself.    Indeed,  given  the  projected
growth  rates  for sales of  heavy-duty diesels,  any reduction  in
sales due  to  this  regulation would only reduce  growth  and should
not  result  in a  real  decrease  in sales.   Thus,   this  regulation
should not  have any  adverse local effects on employment.

     EPA does  not   expect  diesel heavy-duty  vehicle sales or the
heavy-duty industry  in general  to suffer in  the  long run because  of
a  shift  in  the mode  of freight transportation  used.   As  will  be
shown  in the  next  section,  the  impact  of  this  regulation on the

-------
                               -135-
cost of  owning and  operating  a heavy-duty  diesel  is  very  small
(less than one-half a percent).   Such a small increase  should have
little or no effect on  the demand for heavy-duty diesels.

     Thus,  these  regulations  should not  adversely  affect the
heavy-duty  diesel  industry,  either  through unreasonable  capital
requirements or reductions in sales.

     2.    Impact on Users of Heavy-Duty Diesels

     Users of  heavy-duty diesels  will be affected  through higher
initial  vehicle  costs  averaging $527-$650 for   1986  and  on.   The
average  retail price of a new heavy-duty  diesel  truck  in  1979 was
estimated to  be between  $15,000  and  $50,000,_iy  ($16,000 and
$54,000  in  1980  dollars).   In  addition,  as  described  in  the next
section,   new  diesel-powered  buses   cost  approximately  $140,000.
This means  that the average  vehicle sticker price will  increase
0.5-3 percent in  1986 and  beyond.   However,   accompanying this
sticker  price increase  will  be  a reduction in maintenance  costs of
$178 per  vehicle  (discounted to year of vehicle purchase).  This
savings  would reduce  the  overall  impact of this  regulation to
$349-$472 per vehicle.

     Users of  heavy-duty  diesels will have  to  recover their in-
creased   investment  by  increasing the  handling  costs  of  freight.
Operating costs for  intercity trucks in  1975 were about $1.70 per
mile,^_/  or about  $2.50 per mile in 1980 dollars.  The  average
operating  revenue  in  1975  was  about $1.80  per  mile,^/  or  about
$2.65 in 1980 dollars.   The average  life-time of  a  heavy-duty
diesel is about 9 years with a lifetime mileage  of 475,000  miles._9_/
The  total  freight  revenue  per  truck would  then be  $825,000,
discounted back to the year of  purchase,  according  to  the  distri-
bution of mileage  throughout  its  life.  The maximum impact of
this regulation  on  a  Class  VIII vehicle should be  no more than
$472 (the  upper  limit  of the previously  determined range).   Thus,
this  regulation  should only  increase  operating  costs  by 0.06
percent.   This should have  little effect  on  the  trucking  industry.

     The  smaller, Class IIB-VII diesels should have a  smaller
lifetime mileage  due  to different usage  characteristics.  As
described  in  Chapter IV,  these vehicles  are expected  to  be used
much like their gasoline-fueled  counterparts, which  have a  lifetime
mileage  of 114,000  miles. 9_/   The  freight  costs  (on  a  per mile
basis)  on these trucks  were  not  readily available,  but  they should
be higher than that  for the  intercity diesels due to shorter  trips
and more  loading and unloading.  The lower  lifetime mileage  would
tend to decrease lifetime operating  expenses by  about a factor of
four, but a high expense per mile could remove  half or  all of this
difference.  At  most,  the  impact of this regulation on the  oper-
ating costs of these  smaller  diesels  would be four  times that
determined for the  larger diesels.   This would  be  an  increase of
0.24 percent, which  is still quite  small.   Or  the  impact  could be
as  low as that for  the large diesels, 0.06  percent, if the  oper-

-------
                                -136-
ating costs per mile  for  the  smaller diesels  were four times that
for the  larger diesels.   In either case,  the impact of this regu-
lation  on the use  of these vehicles to haul  freight should be
hardly measurable and prove to be  no problem.   Thus, this regula-
tion should not have  an  adverse  impact  on the users of heavy-duty
diesels.

     3.   Impact  on  Urban  Areas and Specific Communities

     The purpose of this  section  is to  identify the socioeconomic
impact of  this  heavy-duty diesel  particulate  regulation  on urban
areas and specific communities.   The analysis has been broken down
into three parts  which will  evaluate the impacts of this regulation
on personal income,  employment, and fiscal condition of urban areas
and specially-affected communities,  respectively.  As will be seen,
no  adverse urban  or community  impacts  are expected from this
regulation.

     a.   Personal  Income

     One important aspect of this  regulation is  its  effect  on the
use of personal  income  in urban  areas in general  and  in specific
localities.   In other  words,  would  this regulation  cause  urban
dwellers to pay  for  a disproportionate  share of  the costs  of
control  or  to  shift  a significant  portion of  their  income  to pay
for heavy-duty  diesel  particulate control?

     Concerning the  direct costs  of this  regulation,  its effect on
low  income groups  and on  urban  dwellers  in general should  be
negligible  since  these  individuals  are not  involved to  any sig-
nificant degree  in  the  purchasing of  heavy-duty  diesels  for
personal or business use.   Businesses located  in  urban areas will
have to pay more  for their diesel-powered trucks, but they will pay
no more  than those located outside urban areas.   The absolute
effect of the price increase for  urban businesses is addressed for
cities purchasing diesel-powered  buses  in  the  section below en-
titled "Fiscal Condition."  As will be seen  there,  the effect of
this regulation on the purchase price of heavy-duty diesels is very
small.  It is  also true that most of the trucks purchased for urban
use are  powered  by gasoline  and  not  diesel  engines,   though any
future shift is expected to  be toward diesels.

     The users  of  heavy-duty diesels will  have to  recover  their
increased investment by increasing  the handling  costs  of freight.
The magnitude of  this  one-time  increase  has been  estimated  to be
less than  0.06-0.3  percent  (Section D.2.  of this  chapter).   This
should  have little effect on  the  trucking industry, urban or
rural. Consumers in all  localities and  of all  income  levels will
have  to  pay for  this increased  operating cost,  since it  will
probably be applied to the  costs  of most  food  and consumer items.
However,  since  transportation  represents  only  a  fraction of the
total cost  of  consumer  goods,  the rise  in prices  should  be even
less than 0.06-0.3 percent,  which itself  is  negligible.   Thus, the

-------
                                -136a-
burden on low income level groups should be negligible and will be
no different from the burden  imposed on higher income  groups.

     It should be  mentioned  at  this  time  that the primary benefit
of this  regulation,  that  of  improved air  quality, will  occur
primarily in urban areas.   As  outlined  in  Chapter V, the largest
concentration of heavy-duty  diesels  and their emissions occurs in
urban areas.  Similarly,  the  greatest  air quality  improvements will
occur in urban areas.  However, as was  seen above  and will be seen
below, the bulk of the  cost  of control will be spread  fairly evenly
between urban and  non-urban areas.  Thus, this regulation actually
tends to favor  urban areas  by providing benefits primarily to urban
areas, where they  are  legitimately  needed  the most,  and spreading
the cost rather  evenly across the whole nation.    This is somewhat
unavoidable since  nearly  all  heavy-duty diesels  enter urban areas
for a  fraction  of  their  travel and  controls  cannot  be  placed on
them only when  they are in  urban areas.

     b.   Employment

     The production  of heavy-duty  diesel  engines in the  U.S.  is
spread across a number of states and  no one locality has more than
one manufacturer.  Of the five major manufacturers, two are located
in Indiana (Columbus  and Fort Wayne), one  in  Michigan,  one in
Pennsylvania and one in Illinois.  The heavy-duty  vehicle industry
is even  less  concentrated.    Given  this,  any  effect  of  this  reg-
ulation on employment would  not be concentrated in  a single city or
even  in  a  single  state,  but be  spread over  a number  of  states.
However,  each manufacturer tends to be located in  or near a single
mid-size city.  Any decrease in  employment for a given manufacturer
would then  affect  a  single  area and the workers  affected would be
primarily urban dwellers,  though a fraction of those affected would
certainly have  commuted into  the city  from rural areas.

     As  was outlined  earlier  (Section  D.l.b.  of  this  chapter),
however,  the negative  effect  of this  regulation on sales should be
negligible. However, there is a general trend toward the increased
use  of  diesel   engines in  heavy-duty vehicles (see  Chapter  III,
section C)  and heavy-duty diesel sales  should actually increase 20
percent  between  1980  and 1990.   Thus, total employment  in the
heavy-duty diesel industry  should increase substantially.   In
addition, new jobs will be  created to  research, develop and produce
emission control equipment  for these vehicles.  Overall, then,  this
regulation should not  have any  adverse  impact  on employment in any
specific localities or  in urban areas  in general.

     c.   Fiscal Condition

     The identification of  specific cities or the types  of cities
that  are likely to incur  an economic burden due  to the costs
associated with  the heavy-duty  diesel particulate  regulation is an
important part  of  this urban analysis.   The  two primary  factors
affecting  the  fiscal  viability  of  cities  have already  been  dis-
cussed:  employment and income,  both personal  and  business.  As was

-------
                              -136b-
seen, neither factor will  be adversely affected by  this regulation.
However,  there is  one possible  way that  this  regulation could
affect cities  which  deserves further  attention.   The increase in
the first cost of  heavy-duty diesel  engines  could affect the larger
cities that support a  large mass  transit system primarily consist-
ing of buses.

     Cities  that  need to  purchase new  buses with heavy-duty
diesel engines  in 1986 and later  years  to upgrade  their  fleets
will  initially  have  to  pay higher  initial  costs  due to  this
regulation. It is  estimated  that  the average first price increase
for  heavy-duty vehicles will be  $521-$632  due  to  this regulation
(see  section  B).    However,  these modifications will  also  reduce
maintenance costs  by $178  over the life of the vehicle (discounted
to year of purchase).   As  this regulation is  not expected to affect
fuel economy, the cost of  owning and operating a heavy-duty diesel
should increase $349-$472  per vehicle  beginning in  1986.

     The  biggest effect on  the  cities will  be  the purchase price
increase  of roughly $520-630 per bus.  However,  this increase only
represents a 0.4 to 0.5 percent  increase in the purchase price of a
intracity transit  bus, which at  the present  time  is approximately
$140,000  per vehicle ._!£/   This  small  increase due  to  this regula-
tion will not  offset  the  fact that buses  are the  best option for
intracity transport  and  should  also not prevent any city  from
buying buses that  needs them.  Likewise, the  effect of this regula-
tion on intracity bus  ridership,  due  to  fare increases,  should be
negligible.

-------
                                -137-
                            References
JY   "Regulatory Analysis  and  Environmental  Impact of  Final  Emis-
     sion  Regulations   for  1984  and  Later  Model  Year  Heavy-Duty
     Engines," December 1979,  OMSAPC,  EPA.

2/   Code of Federal Regulations, Title 40,  §86.077-26.

3J   "Regulatory Analysis  - Light-Duty Diesel Particulate  Regula-
     tion," MSAPC,  OANR, EPA,  January 29,  1980.

4/   "Summary and Analysis of Comments to  the NPRM:   1983 and  Later
     Model  Year Heavy-Duty Engines, Proposed Gaseous Emission
     Regulations," EPA, December 1979.

5J   "Gaseous Emission  Regulations  for 1984  and Later  Model Year
     Heavy-Duty  Engines," EPA,   45FR4136,  January   21,   1980.

6/   Penninga, T.,  TAEB,  EPA,  "Second Interim Report on  Status of
     Particulate  Trap  Study,"  Memorandum  to R.  Stahman,  Chief,
     TAEB, EPA,  August 28, 1979.

T_l   Alson, Jeffrey, SDSB,  EPA,  "Meeting  Between Texaco  and EPA to
     Discuss  Particulate  Trap  Work,"  Memorandum  to  the  Record,
     October 1979.

8/   Passavant,  Glenn  W.  "Average Lifetime Periods  for  Light-Duty
~~    Trucks and Heavy-Duty  Vehicles," EPA,  November  1979, SDSB-79-
     24.

_9/   "Alternative Fuels  and Intercity Trucking," Ryder  Program in
     Transportation  and Escher Technology Associates for U.S.
     Department  of  Energy,  June  1978,   HCP/M3294-01,  pp.  71-72.

10/  Personal communications with the Ann Arbor  Transit  Authority,
     November 6, 1980.

-------
                                -138-


                            Chapter  VII

                        COST EFFECTIVENESS

     Intuitively, cost  effectiveness  is  a  measure of the economic
efficiency  of  an action towards achieving  a  goal.   Historically,
however, the cost effectiveness  of emission  control regulations has
been  expressed  in  such terms  as  "dollars per  ton of  pollutant
controlled."  This  expression  is  a  measure  of the cost  of the
regulation, not  necessarily  its  efficiency.   The presence of this
conflict makes  it awkward to speak in relative  terms  about cost
effectiveness since a low cost-effectiveness value implies a highly
effective regulation.  To escape  this  conflict  and still  follow the
precedent  of  placing cost in the  numerator,  the measure  of cost
effectiveness will be referred to as  the cost-effectiveness ratio,
or C/E ratio.

     Furthermore, air  pollution  control  regulations  have multiple
and  frequently differing  goals  and, therefore, do not  easily lend
themselves to direct comparison  of C/E measures.   In the past, the
principal application of  comparing  C/E  measures  has been the
evaluation of alternative control strategies applicable to the same
source, in the same  time frame,  and with the same objective.  This
markedly simplifies the analysis  and,  as will be  seen below, avoids
many problems.  Nevertheless, a  rough  measure  of  one aspect of the
relative merit of the  proposed   heavy-duty  diesel  standard  can be
achieved  by  comparing  the  C/E measures  of  alternative  diesel
standards  with  other  strategies  designed   to  control  particulate
emissions.   One  area where  EPA  has  adopted regulations  to limit
particulate emissions  is the New  Source  Performance  Standards
(NSPS)  for Stationary  Sources  called  for   by  Section  111  of the
Clean Air  Act.   While  the statutory purposes  and tests in Section
111 are different from  those applicable to this diesel particulate
standard,   a  rough comparison has  been made which  indicates that
this decision  is  not inconsistent with  other  decisions the Agency
has made to control particulate  emissions.

     In this  chapter,   the  C/E  measures for  the level  of  diesel
particulate control will  be calculated  and compared to  those
from other control strategies.  As will be seen,  it is not possible
to take  into account  all of the environmental factors  such as
meteorological conditions,  location,  population  exposures,  etc.,
due  to  a  lack of data.   However,  as  many  of  the factors for
which data are available will be  incorporated.

A.   1985 Heavy-Duty  Diesel Particulate Standard

     The calculation  of the C/E ratio  for  heavy-duty  diesel par-
ticulate control  is  quite  simple.  Most of the necessary  input data
have already  been determined in past  chapters.   The uncontrolled
emission level is  2.0  g/mi   (1.24 g/km).   Under  the 0.25 g/Bhp-hr
(0.093 g/MJ)  standard   the  in-use  emission  level should  be about

-------
                             -139-
0.67 g/mi  (0.42  g/km).   If these levels are assumed to occur over
the entire  life  of  the  vehicle,  the improvement due to regulation
is 1.33 g/mi (0.82 g/km).

     The average life  of a heavy-duty diesel in the  1986-1995
time-frame now has to be determined.  EPA has examined past data in
this area and found  the average  lifetime of heavy-duty vehicles to
be  114,000 miles/8  years (gasoline  engine)  and  475,000 miles/9
years  (diesel  engines)._!_/   While  the differences between the
durability  of  the two  types  of engines may  cause a part  of the
difference  in  lifetime  mileage,  most of the difference  is  due to
the different usage  characteristics  of the vehicles equipped with
each  type  of  engine.    Line-haul  inter-city  trucks  have  been
equipped with  diesel engines,  while short-haul  trucks  have  been
equipped with gasoline engines.

     However, as  outlined  in  Chapter III,  diesels are expected to
start capturing  the  shorter-haul market.   It  is doubtful that the
lifetime mileage  of  these vehicles would  change  with a  switch to
diesels, as the  basic  function  of  the vehicle  wouldn't change.
Thus, as diesels begin to  capture the market from  gasoline engines,
their lifetime mileages  should  decrease, moving  toward the lifetime
mileage  for vehicles with gasoline  engines.    For simplicity,  it
will be  assumed  that  all  Class  VI and lighter heavy-duty vehicles
have an  average lifetime mileage  of 114,000  miles (183,000 kilo-
meters) and that all  Class VII  and  heavier  vehicles have an average
lifetime mileage  of  475,000 miles  (764,000 kilometers).   From the
data in  Table  III-9,  the  fraction  of total diesel sales which are
Class VI or lighter  can be determined.   In  1986, the fraction is
0.33 and in 1995  it  is  0.51.   An  average  would  then  be  0.42.  If
the above  mentioned lifetime  mileages are combined using  this
average  split,  the  average lifetime of heavy-duty diesels between
1985 and 1995 becomes 323,000 miles (520,000 kilometers).  Coupling
this lifetime mileage  with the 1.33 g/mi (0.82 g/km)  emission
reduction yields  a  lifetime  particulate  reduction of 0.430 metric
tons.

     The cost of control  has  been calculated in  Chapter  VI  to be
$349-472 per vehicle.   Thus,  the C/E ratio is $349-472  divided by
0.430  metric tons,  or  $800-1100  per metric ton of particulate
control.   This  is the  C/E ratio  for  emission  reductions arising
from both improved engine  design and  the use of  trap-oxidizers.  It
is possible  to  separate out  the cost  effectiveness  of  the  use of
trap-oxidizers  alone.   This  latter figure could be termed the
incremental cost effectiveness,  while the figure already determined
would be the overall  cost  effectiveness.

     The calculation of the incremental cost effectiveness primar-
ily requires the calculation  of  the  emission reduction and cost of
trap-oxidizers alone.   As -determined above, emissions  of a trap-
oxidizer-equipped vehicle  are  0.67 g/mi (0.42  g/km).   Given  that
the trap  is 60 percent  efficient, the  emissions  from the vehicle
without  the trap would  be 1.67  g/mi  (1.05  g/km).   The emission

-------
                                -140-
reduction  due  to trap-oxidizer  use  is  1.0  g/mi  (0.62  g/km),  or
0.322 metric tons over  the  vehicle's  life.   The cost of trap-oxi-
dizers  can  be  taken  from  Table  VI-8 and  is  $343-454 per vehicle
(includes maintenance costs  and  credits).   The  C/E  ratio  is then
$1,070-1,410 per metric  ton.   This incremental cost  effectiveness
will be used for comparison  purposes  in  the next  section.

B.   Comparison of Strategies

     The purpose of this section is to determine the  C/E ratios of
other particulate control strategies and demonstrates that the C/E
ratio of the heavy-duty  diesel  regulations  is not inconsistent with
those of past strategies. All  of the  C/E ratios  examined should be
incremental  in  nature.   This  is  necessary  because the comparison
must  be made between the  cost of the  last  level of control and
cannot be influenced by  the  costs at  less stringent control levels.

     The incremental C/E ratios  for several stationary sources are
shown in Table  VII-1.   Except  for the industrial boiler category,
all  of  the  C/E  measures shown  represent  the  costs  and  emission
reductions  of  a Federal New Source Performance  Standard  over the
less stringent alternative rejected by the Agency in  selecting the
level of  the standard.    The C/E ratio  for  the  industrial boiler
category represents the  costs and effectiveness of two alternative
control devices which are available.

     The incremental  cost effectiveness  for  the  control of parti-
culate  emissions  from light-duty  diesels  is  also shown  in Table
VII-1.  The control  increment examined was the  1985 standard of 0.2
g/mi (0.12 g/km) for light-duty vehicles  (0.26  g/mi (0.16 g/km) for
light-duty  trucks) over  the  1982  standard  of  0.6 g/mi (0.37 g/km)
for both vehicle classes .J5/

     As mentioned earlier, the most direct  and  easiest  use of
a  cost-effectiveness measure is to  compare various levels  of
control of  a single  source.   In  this  case,  most of the  factors
pertinent  to  the environmental  impact,  such  as  source location,
dispersion characteristics,  and pollutant characteristics,  are the
same for all the levels  considered  and the  "dollar per ton1 measure
is a good relative measure of the cost effectiveness of the various
strategies.  Given  enough  knowledge  and data,  there  is  no reason
that this same kind  of  analysis  cannot  be  used to compare various
strategies   for controlling  different  sources.   The problem is,  of
course,  that the necessary data is  usually very difficult to obtain
and not available.   The  comparisons being made in this section are
not  true comparisons of the  cost effectiveness  of any  of the
strategies being  examined.   The necessary  data is simply not
available.  However, comparisons  such as these  are being made
elsewhere and will  be made  in the future.  The  goal here will be to
make the comparisons, while at the same time  stating clearly the
limitations involved,  insuring  that any use of the results of this
section is  accompanied by  full  knowledge  of  their meaning.

-------
                               -141-


                              Table VII-1

            Incremental Cost Per Ton of Particulate Removed
           for Selected New Stationary Sources (1980 Dollars)

                                   Cost-$/Metric Ton for
                                   Particulate Collected
	Source	           in Incremental Range        Reference

Medium Sized Industrial
  BoilersjV                               $1000                    2

Electric Utility Coal-
  Fired Steam Generator^/                $900-$1000                3

Kraft Recovery FurnaceS/                $1400-$1900                4

Kraft Smelt Tank4/                       $160-$220                 4

Rotary Lime Kiln_5/                      $1200-$1300                5,6

Electric Arc Furnaces
  - Steel6/                                $700                    7
J7   Baghouse  (0.03  lb/106BTU)  versus   cyclone  (0.3  lb/106BTU).
2J   High  efficiency  ESP (0.03 lb/106BTU) versus  lower efficiency
     ESP (0.1 lb/10° BTU).
_3_/   High efficiency ESP  (99.5 percent) versus lower efficiency ESP
~~   (99.0 percent).
4/   Venturi  scrubber  versus  Demister   (80  percent  efficiency).
5/   High  efficiency ESP  (0.3  Ib/ton  limestone)  versus  lower
~~   efficiency ESP  (0.6  Ib/ton limestone) for  500 TPD  plant;
     baghouse (0.3  Ib/ton)  versus  lower  efficiency  ESP for 125 TPD
     plant.
_6/   Direct evacuation with 90 percent efficient canopy hood versus
     direct evacuation with open roof.

-------
                                -142-
     The  strategies  being examined  here all  address  particulate
emissions  on  a  nationwide  scale.   Both diesel  regulations  will
apply to every new diesel  sold  in  the U.S., regardless of where the
vehicle  is  bought  or  used.   Likewise,  the  New Source Performance
Standards (NSPS)  for  stationary sources,also  apply to  all new
significantly  modified  plants  of  a  certain  type nationwide.

     While both the mobile source and stationary source strategies
being examined  control  particulate emissions  into the atmosphere,
there are differences  in their  primary purposes.  An examination of
Title II of the Clean  Air  Act,  particularly Section 202, shows that
the primary purpose of mobile source  regulations  is to protect the
public health and welfare.   The  primary purpose  of the NSPS's,  on
the other  hand,  is  to  reduce inequities in interstate competition
for economic growth, while minimizing emissions through the nation-
wide use  of  the best  available control  technology.   A nationwide
NSPS  prevents  those  states and localities without severe air
pollution problems from having  an  unreasonable advantage in drawing
new plants from areas  where strict controls are required.

     While  the primary purpose  of  the two  types of strategies
differ,  the levels of control they  represent  do have a  common
purpose, that  of  protecting  the  public  health  and welfare.   The
NSPS's  exist  because  some states and localities  require  at least
this level of  control  to  protect  the  public  health and welfare in
their areas.   There are factors that affect the relative stringency
of  the  two types of  standards.   For  example, economics may  be  a
more critical  parameter for NSPS's than mobile source standards and
the requirements  for  the  demonstration  of  technology are stricter
for NSPS's  than  mobile  source  standards.  In  a  rough  sense,  how-
ever,  both represent   control  levels  implemented  to  protect  the
public health  and welfare.

     To take  one rough step toward  making  the measure of  cost
effectiveness  more  relevant  to  health   and  welfare  impacts,  the
basis  of  the  previously cited  'dollar  per  ton'   figures  shall  be
modified  to  reflect the  cost  of  controlling inhalable  and  fine
particulate.   In  Chapter  V,  it was shown that it is these partic-
ulates  that  have  the  greatest potential  for adverse health im-
pact.    Thus,  it  is appropriate to emphasize  the control  of these
particles.  Also,  it  is these  smaller  particles  (inhalable parti-
cles have diameters of less  than 15 micrometers and fine particles
have diameters of less than  2.5 micrometers)  which have the great-
est effect on  visibility,  which is likely one  of the  largest
welfare  effects of diesel  particulate emissions.

     Particle  size data currently available  for  these  sources are
limited  and the  figures presented below should  only be considered
to  be rough  approximations.   The  size  of diesel  particulate has
already  been  discussed in  Chapter V.   All  of  the uncontrolled
diesel particulate is  inhalable (diameter less than 15 micrometers)
and between 94  and  100 percent fine (less  than  2.5 micrometers).
The trap-oxidizer,  however, may be more efficient in trapping large

-------
                                -143-
particles than small ones.  To be conservative, it will be assumed
that all  coarse  particles (diameter greater than 2.5 micrometers)
are captured and burned and that  only  that amount of fine particles
necessary to meet the  1986 standard are  also captured and burned.
Given this assumption and a 60 percent efficient trap-oxidizer, the
result  is that  100% (by weight)  of the additional particulate
controlled by the 1986  standard  is  inhalable and  91-100%  is fine.

     Power  plants  (large  steam  generators)  tend  to  emit  larger
particles than  diesel engines.   EPA has  measured the particle
size  distribution   of   electrostatic  precipitator   (ESP)  effluent
at both  the  previous emission standard of  0.1  pounds  per  million
BTU  (43   nanograms  per  joule)  and the  revised standard  of  0.03
pounds per million  BTU (13 nanograms per joule).  Of the additional
particulate collected  at  the  revised standard,  90-100 percent (by
weight) is inhalable and 20-40 percent is fine.9/

     Medium-sized boilers are  commonly spreader stoker-type boilers
which  emit  coarser   particles  than  pulverized  coal-fired  boilers.
As an approximation,  it is estimated that 70 percent of the partic-
ulate  collected  in  the  incremental  range  between  a  cyclone  and
baghouse  is  inhalable  and 25  percent  is  fine.  For  electric  arc
furnaces, the  particulate removed  by a baghouse  installed  with  a
canopy hood  is  about 90  percent inhalable  and  60  percent  fine.JT/
For a kraft recovery furnace  the incremental particulate collected
by an  ESP in the range  from  99.5  to 99.0 percent is about  100
percent  inhalable and  70 percent  fine.   The differential  quantity
of entrainment collected by a  venturi  scrubber in comparison with  a
demister on a kraft  mill smelt tank is  about 85 percent inhalable
and  55  percent  fine.    High  efficiency collection  versus  medium
efficiency collection  of particulate  from a rotary lime  kiln
captures  particulate that  is  about  80  percent  inhalable  and  50
percent fine.

     Using  these approximations,  the  C/E  ratios  for these six
sources can now  be  placed on  an  inhalable  and  a fine particulate
basis.  The results  are  shown  in Table VII-2.  As can be seen, the
cost effectiveness  of the heavy-duty diesel standard is not incon-
sistent with those  of past Agency actions or with a possible future
Agency action (medium-size industrial boilers).

     It is important to emphasize a point  made earlier, i.e.,  that
in some respects  the  mobile stationary  source  strategies  for
particulate control have certain  differences  in  their  primary
purposes.   Therefore,  selection  of  a measure of effectiveness for
comparison purposes  has  inherent  limitations.   In  spite of these,
however,  a  comparison  may still be useful   to the  degree  that  it
focuses  on one  of  their common  purposes,  protection of  public
health and welfare.

     Up  to  this point,  however,  we have  only  incorporated one
factor which  may improve the  comparability  of the cost-effective-
ness  measures  for   different  source  strategies.   There are  many

-------
                                 -144-
                             Table  VII-2

            Incremental  Cost-Effectiveness  Ratios  of  Particulate
                 Control  Strategies Using Three Measures  of
                 Effectiveness  (1980 Dollars  per Metric Ton)

Controlled Source
Heavy-Duty Diesel
1986 Standard
Light-Duty Diesel -
1985 Standard
Utility Steam Gen-
erators*
Medium-Size Industrial
Boilers
Electric Arc Furnaces
Steel
Lime Kilns
Kraft Pump Mills
Recovery Furnaces
Smelt Tank
Total Particu-
late Basis
1070-1410
2600-3270
900-1000
1000

700
1200-1300
1400-1900
160-220
Inhalable Particu-
late Basis
1070-1410
2600-3270
900-1100
1400

800
1500-1600
1400-1900
200-260
Fine Particu-
late Basis
1070-1550
2600-3600
2900-3300
4200

1100
2400-2500
2100-2600
300-400
*    Assumes that an average of 30 percent of controlled particu-
late matter is fine.  If the full range of the fine fraction is
used (20-40 percent),  then the cost-effectiveness is $2,200-4,900
per metric ton.

-------
                                  -145-
other factors which would  need  to be accounted for before a truly
valid comparison could be made,  such  as  emission dispersion charac-
teristics, "source  location,  chemical  composition  (and resulting
health effects) of the  particulate,  etc.   As these factors cannot
be incorporated at this  time due  to  lack of  data,  even  the compar-
ison  performed  in  Table  VII-2  must be taken  cautiously.   The
incorporation of the factors  mentioned above  could  change the
results  drastically.

     To  indicate  this possibility,   one  rough  calculation  will be
made comparing the air quality  impact of a given rate of emissions
for both types of diesels and  power plants.   Only rough  large-scale
impacts will be considered, so  this will  not  be an  exhaustive
comparison by any measure.   However,  it  will  serve  to highlight the
possible effects that  these missing   factors may have  on  any com-
parison  of the cost  effectiveness  of  different  strategies.

     As  a  rough approximation  of the  relationship of ambient impact
to emission rate, the ratio of  the maximum ground  level concentra-
tion to  the annual emission rate  will be used.  The maximum ground
level concentration  was chosen  as  an indicator  of air  quality
impact  because:  1)   it  was available  for  both  sources,  and 2)
particulate levels near  this maximum  should  occur over  large areas
for both  sources.  From 2), no  localized concentrations of diesel
particulate will be  used in this analysis, only regional concentra-
tions, nor  will  unusually  high  impacts  from power plants  due to
unique topography or  poor design be used.  The  annual emission rate
was chosen as the indicator of emission  levels because it is a good
indicator  of long-term emission  impact.

     EPA has  already analyzed  the  air quality impact of  power
plants  and it will  only be  summarized here.3/  Three sizes of
steam generators were examined along  with stack heights  typical for
those plants.   The dispersion  of emissions were  then  modeled to
determine  the maximum downwind  concentration at ground  level.   The
results  are  shown  in Table VII-3.   As  can  be seen,  the  ratio of
the maximum ground level concentration to the annual emission rate
is larger  for the smaller plants.  This  is primarily due to shorter
stacks.

     The same calculation for  both heavy- and light-duty diesels is
slightly more complicated in that  there  are many individual diesels
in close proximity to  each  other at various concentrations.   No one
source can  be  modeled and  at the same  time, no one  source  has  a
very large impact on air quality.  With diesels,  then,  a geograph-
ical area  must be examined  rather  than a  single vehicle.

     A metropolitan  area would be  appropriate since it represents a
large area  (on  the order of that  affected  by a large power plant,
though  possibly  smaller)  and   it contains  areas  of  high  concen-
trations  (downtown) and  low concentrations  (rural  areas).   Kansas
City will  be chosen for  this  task even  though it appears to have a
smaller diesel  impact relative  to  other cities  its  size.   The

-------
                                 -146-
                              Table VII-3

             Air Quality Impact of Three Steam Generators
                           at Ground Level 3/*
Annual Emission Rate
(metric tons per year)
                                          Plant Size  (Megawatts)
                                        25          300        1000
99.3
1192
3974
Typical Stack Height (meters)
 75
 175
 275
Maximum Ground Level Concentration
(micrograms per cubic meter):

     Annual Mean
     24-Hour Maximum
 0.1
 1.3
 0.1
 1.3
                                                               1.3
Ratio of Maximum Ground Level Concen-
tration to Annual Emission Rate (micro-
grams per cubic meter/metric tons
per year)

     Annual
     24-Hour Maximum
  .0010
  .0131
 0.00008
 0.0011
<0.000025
 0.00033
*    Numbers bracketed (_/)  indicate references at the end of this
chapter.

-------
                                  -147-
necessary data  is  available  for  Kansas  City,  and the metropolitan
area does contain both  urban  and  rural areas.

     The Kansas  City area  examined  here will be  that  examined  by
PEDCo.J_OY   It comprises 660 square  kilometers.   Total vehicle
travel in this  area  in 1974 was  2.85  x 10^ miles per  year.   The
impact  from  light-duty diesels will  be  examined  first,- using
Chapter V  of the  Regulatory Analysis  for  light-duty  diesel  par-
ticulate regulations .JLO/  Using  a 1%  per year  growth  rate,  total
vehicle travel in  1990  will  be  3.34 x  10^ miles per  year.   If the
low  estimate  of diesel-dieselization  is  examined here, 9.57%  of
total vehicle travel will  be by  light-duty diesel in  1990.   At  a
particulate  emission  rate  of  1.0 g/mi  (0.6  g/km),  light-duty
diesels would emit 321  metric tons per year.  Using this scenario,
the  ambient  concentration  at a  typical TSP monitor  would be  1.5
micrograms  per cubic meter  (Table V-7).10/   The  ratio  of  ambient
concentration to the  annual emission rate would be 0.0047 microgram
per  cubic meter  (per)  metric ton per  year.  The maximum  24-hour
impact  for  light-duty diesels is about  3.16  times  the annual
geometric mean (see Chapter  V) ._10/ Thus, the  ratio of  the  24-hour
ambient  concentration to  annual emission rate  would be 0.015
microgram per cubic meter  (per) metric ton per year.  These  results
are summarized in Table VII-4.

     The impact of heavy-duty diesels  will  now be considered  using
the  results  contained  in  Chapter V of this document.   If  the low
estimate of  dieselization  is again assumed, 4.2  percent of  total
vehicle travel will be  by heavy-duty diesel  in 1995.   Total travel
in  the  area   in  1995 would  be 3.51 x  10^  miles.  At  an  emission
rate of  2.0  g/mi  (1.24  g/km),  heavy-duty  diesels would emit  295
metric tons  per year.  Using this scenario,  the ambient concentra-
tion at  a  typical TSP  monitor  would  be 1.4 micrograms per  cubic
meter.  The  ratio  of ambient concentration  to  the annual  emission
rate would  be 0.0047  microgram per cubic meter (per) metric  ton per
year.  These  figures are shown  in Table VII-4.   They are the  same
ratios  as  calculated  above  for  light-duty  diesels  and for  good
reason.   Particulate matter  is emitted  from either type of  diesel
from the same general locations  (i.e., roadways) and any difference
in  overall vehicle concentration  or vehicle emission rate  affects
both  total  emissions  and  ambient  concentration proportionately.
Thus, the ratio  of these two  parameters remains constant.

     A comparison of  the values  in Table VII-4 with those  in Table
VII-3 shows  that the ambient concentrations  per unit  emission  rate
of  diesels  is 4.7 and  188   times that  for  small and  large  steam
generators  on an annual basis, respectively.  On a  24-hour  basis,
the  ambient  concentration  per unit emission  rate for diesels
is  actually 1.1  and 45  times larger than  that  for small  and  large
power plants, respectively.

     As mentioned  earlier,  the above ratios  are  only an extremely
rough  estimate  of the relative  air  quality impacts  of diesels
and  power  plants.  Many  simplifications  were  necessary to be

-------
                                 -148-
                             Table VII-4

                Air Quality Impact of Light-Duty and
      Heavy-Duty Diesels  in the  Kansas City Metropolitan Area
 Total  vehicle miles
   traveled  in area
   per  year

 Fraction of travel
   by light-duty  diesel
   (low estimate  of
   dieselization)

 Emission factor
   (g/mi)

 Annual emissions
   (metric tons
   per  year)
Light-Duty (1990)

   3.34 x 109



   0.0957
   1.0


   321
Heavy-Duty  (1995)

    3.5 x 109



    0.042
    2.0
    295
Maximum  regional
  air  quality  impact
  (micrograms  per
  cubic  meter)

Maximum  24-hour
  average per  year
  (micrograms  per
  cubic  meter)

Ratio of maximum ground
level concentration to
annual emission rate  ,
(micrograms per cubic
meter (per) metric
tons per year):

       Annual

       24-Hour
   1.5
   4.74
   0.0047

   0.015
    1.4
                                  4.42
    0.0047

    0.015

-------
                                 -149-

able  to  make this  comparison  at all.   However, the  results do
indicate the  size  of the  factors  which  may occur if an  extensive
analysis were performed  and how  the  results  of Table VII-2 might
change if other  factors were incorporated.

     If the decision is made to  restrict  the comparison  to only  the
two diesel standards, there is one  last  step which can  be taken to
improve the cost-effectiveness methodology and  that involves  taking
into account the population exposed.  The similarity  between  light-
and heavy-duty  diesels  reduces  the need  for  accurate population
exposure data and allows more general  source characteristics to be
sufficient.   Indeed, there  is little more population  exposure data
for diesel  emissions  available  than  are available for  other
sources.    In a  relative  sense,  however,  the  two  diesel sources
can be compared.

     Due to  the lack  of  available  exposure data,  a very  simple
source characteristic  will be  used to  estimate  population expo-
sure.    This characteristic will  be the  fraction  of total  source
emissions which  are urban.  In other words,  the cost  effectiveness
will now be determined  on an urban basis.  This  is  justified  by  the
fact  that 85%  of  the nation's  population  that lives  in  areas
exceeding the primary NAAQS for TSP are  metropolitan  (Figure V-8).
The practical effect on  the calculation  of the C/E  ratios is  that
the costs  will  be  divided  by the fraction of the emissions which
are urban.   This recognizes that  an emission standard requires  all
vehicles to  reduce  emissions  irrespective of where they  are used,
but it is  those  in  operation  in  urban areas which affect  the most
people.   From Table V-4  about  56 percent of light-duty diesel
emissions  are  urban and  about  36.4  percent  of heavy-duty  diesel
emissions are urban.   Thus, the  C/E ratios of Table  VII-2 need to
be divided by these  fractions.

     These final C/E ratios are shown  in  Table VII-5.   As would be
expected, the difference between  the C/E  ratios of the  two sources
has diminished.    Control of  light-duty  diesels  is  now only about
50-60   percent more  costly  than heavy-duty diesel control, due to
the greater urban impact  of light-duty diesels.

     In  conclusion,   the  heavy-duty  diesel  particulate   standard
appears  to  be  no  less  cost  effective  than other  cost-effective
measures adopted in  the  past  by  EPA,  using  the cost-effectiveness
methodology developed  in  this chapter.  Even after  the incorpora-
tion of urban/rural differences,  the control of heavy-duty diesels
is  still more  cost  effective   than  the  control  of  light-duty
diesels.

-------
                               -ISO-
                            Table VII-5

                C/E Ratios for Heavy- and Light-Duty
             Diesel Particulate Control Only Considering
            Urban Effectiveness (Total Dollars per Metric
            Ton of Particulate Controlled In Urban Areas)
                          Inhalable Basis                Fine Basis
Heavy-Duty Diesel            2900-3800                    2900-4200
1986 Standard
Light-Duty Diesel            4600-5800                    4600-6400
1985 Standard

-------
                                 -151-

                              References

_!_/   Passavant,  Glenn  W. ,  "Average Lifetime Periods for Light-Duty
     Trucks  and  Heavy-Duty Vehicles," Technical Report, SDSB, EPA,
     November 1979, SDSB #79-24.

2J   "Particulate  Emission  Control  Costs   for  Intermediate-Sized
     Boilers,"   Industrial  Cleaning  Institute  for EPA,  -February
     1977-

_3_/   "Electric Utility Steam Generating Units - Background Informa-
      tion  for  Proposed   Particulate Matter  Emission Standards,"
      OAQPS, EPA, July 1978, EPA 450/2-78-006a.

47   "Standards  Support  and Environmental Impact Statement, Volume
     1:  Proposed Standards of  Performance  for  Kraft  Pulp Mills,"
     OAQPS OAWM, EPA,  September  1976.

5J   "Standards  Support  and Environmental Impact Statement, Volume
     1:  Proposed Standards  of  Performance  for  Lime Manufacturing
     Plants,"  OAQPS,  OAWM, EPA,  April  1977,   EPA 450/2-77-007a.

_&_/   Compilation  of Air  Pollutant  Emission Factors,  AP-42,  Sup-
     plement No.  7,  U.S.  Environmental Protection Agency, Research
     Triangle Park, North  Carolina, April 1977.

_7_/   "Background Information for Standards of Performance: Electric
     Arc  Furnaces  in  the Steel  Industry   Volume 1: Proposed Stan-
     dards,"  OAQPS,   OAWM,  EPA,  October  1974,  EPA-450/2-74-017a.

&_/   "Regulatory  Analysis, Light-Duty  Diesel  Particulate Regula-
     tions," MSAPC, OANR,  EPA.

_9_/   Personal  communication with  Jim Abbot,  Industrial   Emissions
     Research Laboratory  Studies,  ORD, EPA,  January  10,  1980,
     unpublished emission  control  test results.

10/  "Air Quality  Assessment  of Particulate Emissions  from Diesel-
     Powered  Vehicles,"  PEDCo  Environmental for EPA,  March 1978,
     EPA-450/3-78-038.

-------
                             -152-


                          CHAPTER VIII

                       ALTERNATIVE ACTIONS

     These  particulate  regulations  for heavy-duty  diesels  were
required  by Congress  in the  1977 Amendments to the Clean  Air
Act.  Nonetheless,  possible control of other sources of particulate
emissions were examined  to  ensure  that these regulations  were
consistent with EPA's program to improve the nation's air quality.
Also, Congress  left it  to  EPA to determine  the actual level of the
emission  standard,  so  many  alternatives were  available  in  this
area.   In the  following  pages  these  alternative  actions  will  be
presented  and discussed.   In  the first  two  sections,  those actions
which would  preclude control of heavy-duty diesels  will  be  pre-
sented.   These would  include  1)  further  control  of stationary
sources, and 2) the control of mobile sources other than heavy-duty
diesels.   Strategies for  controlling fugitive  dust  or reentrained
dust  have been  discussed previously  and  will not  be repeated
here.I/    Finally,  in  the  third  section  specific  alternative
emission  standards to the 0.25  g/BHP-hr  (0.093  g/MJ)  standard for
1986 will  be presented  and discussed.

     The use of an averaging  approach upon which to base the actual
particulate standard was  not  considered  for  this rulemaking.   This
decision  is based  primarily on the  findings  of  the Regulatory
Analysis  for Light-Duty  Diesel  Particulate  Regulations.   A  more
thorough  investigation  of averaging approaches  is  being performed
as part  of the  heavy-duty  NOx standard revision for 1986.

A.   Control of Stationary Sources

     The  majority  of major  urban   areas  have  severe  particulate
non-attainment  problems.    The  need for  reductions  in particulate
emissions   from some  sources  is  clear.   However, these  areas  have
also  demonstrated  that  attainment is  not feasible even after
adoption  of  all  reasonable stationary source  controls.   While new
source  performance  standards  can definitely help to mitigate
increased  emissions  and ambient  impacts  due to industrial growth,
they cannot be  expected to reduce TSP concentrations in urban areas
from current levels  (see Chapter V, _l/_2_/) .   Thus,  it is concluded
that further control of stationary  sources  is  not  a  viable alter-
native to  these heavy-duty diesel regulations.

B.   Control of Other Mobile  Sources
     In addition to considering  further control  of stationary
sources of particulate  emissions  as  an alternative  to controlling
heavy-duty diesels, the  control  of other mobile  sources  was  also
considered.    These alternative  mobile sources  include  gasoline-
powered light-  and  heavy-duty  vehicles,  diesel-powered light-duty
vehicles,  locomotives and  aircraft.

     Light-duty vehicles  and  light-duty trucks powered by  the

-------
                              -153-
gasoline engine  and  using lead  fuel  were  once a very  significant
source  of  particulate emissions.   In 1974,  it  is  estimated that
exhaust emissions  from these vehicles totalled 250,000 metric
tons  of particulate, with 107,000 metric  tons classfiable as
suspended  particulate ._3/   The  great  majority of this  particulate
matter  consisted  of  particles  related to  the lead  and lead  scav-
engers  used  in  the fuel.   Since 1975 though, the majority of new
vehicles have required the use  of unleaded  fuel in order to  prevent
premature catalyst degradation.   With  unleaded  fuel and catalysts,
these vehicles produce less than 3% of the  particulate  emissions of
a  diesel-powered  vehicle.   By  1981,  when more  stringent  gaseous
emission  standards  for  light-duty vehicles  will have  come into
effect,  it  is  expected  that almost all manufacturers will  require
the use of unleaded  fuel  in their vehicles.   Thus,  by 1986, when
these  heavy-duty  diesel  particulate regulations come into  effect,
new gasoline-powered  light-duty vehicles  and  trucks  will be , pro-
ducing  very  low levels of  particulate  emissions.  Thus, control of
these  vehicles does  not  present an  alternative to  controlling
light-duty diesel particulate emissions.

     Light-duty diesels,  even  more than their heavy-duty counter-
parts,  were expected  to  be  a  significant  source  of  particulate
emissions by 1990.  However, EPA has  already  implemented  stringent
particulate standards for these vehicles  and further control  is not
feasible  at this  time._3/   Thus,  further  control  of  particulate
emissions  from  light-duty diesels  is  not  a viable  alternative to
these heavy-duty diesel regulations.   At the  same time, control of
light-duty  diesel  emissions does  not  reduce  the need  for  regula-
ting  heavy-duty  diesels.  The  rationale for  the level  of the
light-duty  standards  was  based  only  on  the  projected impact of
light-duty  emissions.  The light-duty standards  were  not  set at  a
level  to alleviate the  total   diesel  contribution  to  ambient TSP
levels.   Reductions  will  be required from heavy-duty  diesels and
were  assumed in  the  process  of  determining  the light-duty  stan-
dards.  Also, reductions  from heavy-duty  diesels  are necessary from
an air  quality standpoint if the contribution  of diesel  particulate
to  ambient  TSP levels is  to be  reduced as far as  technology and
economics  permit.    Thus,  controlling  particulate  emissions  from
light-duty  diesels   is  not an  alternative  to these  heavy-duty
regulations, but  is  a necessary complement to the  overall mobile
source  scheme for reducing particulate emissions.

     The  contribution  of  heavy-duty  vehicles powered  by  gasoline
engines to total particulate emissions was  also examined.   In  1974,
heavy-duty vehicles (gasoline)  emitted about  30,000 metric  tons of
particulate  (see  Chapter  V).    Because  today's  heavy-duty  trucks
(gasoline) are still  being built for operation on leaded fuel, this
figure  would still be a rough estimate of emissions  in  1978.   While
the  particulate  emission  level  of heavy-duty vehicles (gasoline)
does not compare with the  particulate  emission  level of light- and
heavy-duty diesels, it is  still  significant.   By  1984,  however, it
is  expected  that most  heavy-duty vehicles (gasoline)  will be
equipped with  catalysts  due to  new emission standards which will

-------
                               -154-
come into effect  that  year.   This will require  unleaded  fuel,  and
the particulate  emissions  from these vehicles will  decrease  dras-
tically, as in the light-duty  case.   Thus,  it  appears  that  partic-
ulate emissions will be low from  the  new  vehicles  of this class by
1984, and no further control  will  be required.

     Locomotives  are another  source of particulate  emissions in the
U.S.   In 1975,  locomotives  emitted nearly  45,000 metric  tons  of
particulate.4/  While this  is  not  insignificant,  a  complete  removal
of all locomo~tive particulate  emissions  would only  be a fraction of
the  necessary reductions  of  emissions  from  heavy-duty  diesels.
Also,  reductions in  locomotive  emissions  will not decrease  the
effect  of  automotive diesels  near  the  roadway,  where  the  largest
impacts  will  occur.   Thus, while  locomotive  particulate  emissions
may merit control at some  time in  the future, such  control is  not a
feasible alternative to the proposed heavy-duty diesel  regulations,
either in magnitude  or locality of emissions.

     Finally,   the  control  of  particulate emissions from  aircraft
was  examined  as  a   possible  alternative  to  the  proposed  regula-
tions.   In 1975, civil and  commercial  aircraft  emitted 18,000
metric  tons of particulate.4/  This emission  level is even  less
than that from locomotives and amounts  to only  7-8  percent of the
projected heavy-duty  diesel  emissions  in  1995.   Thus, control  of
aircraft particulate  emissions  is not a viable  alternative to  the
proposed standards for heavy-duty  diesels.

C.   Alternative  Individual Vehicle Standards

     Now  that it has  been   shown  that  a  particulate  standard
for heavy-duty diesels is  necessary  (i.e.,  no  other  alternatives
are preferable),  the  timing and stringency of the  standard is  all
that remains  to  be   discussed.    In  the case of heavy-duty diesel
particulate regulations  the   question  of  timing can be  expanded.
This expansion  involves whether   there  should  be  a single  final
standard or an  interim standard  and  then a final  standard,  to  be
implemented at  a time  when   technology  will have  developed  to  a
point where significant reductions can occur beyond those  available
at  the  earlier  date.   This option of  a  one or two step  standard
will be examined  first.

     While  there are  many  internal factors which  could affect  the
number of steps  and  timing of the standard, there  is  one external
factor which  deserves  mentioning   first.   That  external  factor  is
the stringent NOx  standard  to be proposed for  1986.   With  the
negative interaction  which can occur  between particulate  and  NOx
control, the presence of this  stringent  NOx standard could cause an
increase in particulate emissions  if  a  particulate standard  were
not in effect.   With a particulate standard in  place  by  1986,  the
NOx controls used will be  those having less  of a  deleterious effect
on particulate emissions.   Thus,  it would be  advantageous  to  have
some particulate standard  in  place by 1986 to prevent  an  increase
in particulate emissions which might otherwise  occur.

-------
                               -155-
     Th e  other  factors  affecting  the  timing  of  the  standard
are  internal  to this  rulemaking.   The primary factor  is the
availability of control technology.   The  two general  categories of
control techniques  available  are trap-oxidizers  and engine  modi-
fications.   The availability of trap-oxidizers  for  heavy-duty
diesel use should be able to follow that  for light-duty diesels by
one year,  or 1986.   Most  of  the past and current trap  work has been
performed  on  light-duty  diesels  and not  on heavy-duty diesels.
However,  the results of these efforts should be directly applicable
to the use  of trap-oxidizers on  heavy-duty diesels.   Some  addi-
tional effort will  be  required to optimize  regeneration  on heavy-
duty diesels (due to longer idle  periods) and  to  ensure  that they
can withstand the additional vibrations  and  the frequent  high-load
operation characteristic of heavy-duty diesels.   The leadtime
remaining  after promulgation of  the  heavy-duty  diesel  particulate
standard   should  be  sufficient  for  optimizing  these  systems for
heavy-duty use.

     The  leadtime  necessary for  manufacturers  to improve the
engine-out particulate  emissions  of  their engines  to  0.41 g/BHP-hr
(0.153 g/MJ)  is  more  difficult  to  determine and  will vary from
manufacturer to manufacturer.   The  technology already exists,  as
evidenced  by  the emission  results  of existing  engines  shown  in
Table IV-2.   The necessary  leadtime will consist primarily of the
time  needed to  incorporate  the available technology  into the
engines with relatively high levels of particulate emissions.  As
evidenced  in Chapter IV,  many engines are currently being  redesign-
ed for various reasons  and many  of the necessary modifications for
particulate  control  could  already be in process.   For  the  other
engines,   the  specific  changes  needed for particulate  control may
not be  in process,  but  they could  easily  fit  into  the existing
redesign   program for  that engine and be  completed by  1986.   For
still  others, and  this is expected to be a small number, the
magnitude  of internal design changes  and  the lack  of  proper timing
with  existing  redesign programs  could put  the implementation  of
these changes  out  past 1986.    In these  cases, the  manufacturers
could  always  continue   production  via non-conformance  penalties,
which  in  the  first few years  after 1986  should not seriously
hamper sales.  However,  there are  other alternatives.  If the
manufacturer  (for  the  short-term)  can  improve  production  vari-
ability,  decrease his deterioration  factor,  build  more  prototypes,
or use a  more efficient trap,  he could meet the standard even
though his  engine-out emissions were well above 0.41  g/BHP-hr
(0.153 g/MJ).  Thus,  it would appear  that for the majority of
engines,   the  availability  of  trap-oxidizers will  be the limiting
factor rather  than  the  leadtime  connected  with engine  modifica-
tions.  This implies that the  standard, or  the final  standard of a
two-step  standard,  could  be  implemented in 1986.

     However,  1986  is  also  the  year  that  the more stringent NOx
standard  is to be  implemented.  Since  the transient test pro-
cedure will  not be available until 1985,  an  interim standard  prior
to this  date would  be based  on  the  less  representative  13-mode

-------
                                -156-
steady-state test  and  serve  only to prevent increases  in  particu-
late  emission  levels  at a  time when none  are  expected.   The  one
external event  that  could cause  an increase in particulate  emis-
sions  will not  occur  until 1986  (the application of  more  NOx
control).  Given these  two facts, it does not appear  reasonable to
promulgate a standard before  1986.   With the first year of  imple-
mentation being  1986 and the  final  level of control  being achiev-
able  in  1986,  the  two-step standard approach appears unnecessary.
It  would  only be  useful  if  an  interim  particulate  standard  was
needed  in   1986  to prevent  unnecessary  increases  in  particulate
emissions  due  to  NOx  control until  a final  standard  could  be
implemented based  on trap-oxidizers, or if  such an approach  could
provide a  significant  economic benefit  for the manufacturers  and
the public without adversely affecting  air  quality to an unaccept-
able degree.

     One such alternative,  which will be examined here,  would  be to
implement a  2-step standard  where the   first, interim,  stage  would
take effect in 1986 and be based  on the particulate levels achiev-
able from today's best  engines.   The final  level could  follow some
time  later, e.g., one  or two  years,  and be based  on  further
reductions   through  the  application  of  trap-oxidizers.    A 2-year
delay in the final  standard  will be  used specifically, instead of a
1-year delay primarily  due  to  the cost of  completely recertifying
the heavy-duty diesel fleet after only  one  year.  To  recertify  the
entire fleet would cost about $7.1  million  (1980 dollars,  inflated
from $6.58 million  (1979 dollars)) ._5_/   The entire cost (or  at  least
90  percent  of  it)  would be  due to  the  second particulate  standard
since no other emission standard will be changing in  1987  or  1988.
Normally,  only  about  10 percent  of the engines  go  through full
certification testing each year and the rest obtain carryover from
previous years'  testing.   Thus,  any  delay  in  the  final  standard
will cost $7.1 million  in  certification costs,  but  a 2-year  delay
would allow  that much more  time for trap-oxidizer development  and
separate this work from the engine  modification work  also  underway
in a way that a one-year delay  would not.

     The prime beneficiary  of a 2-year  delay in the final  standard
would be the heavy-duty diesel  manufacturers.   They would have  an
additional  two years to perfect trap-oxidizer designs.  This  would
also delay  most  of the investments needed  to  develop  and produce
trap-oxidizers  for  two years.  The  public might  also  benefit
economically if  this  delay  would reduce the  cost  of  compliance.

     The cost of the delay  would  be borne  primarily  by the public
in  terms  of poorer  air quality  and  its accompanying  health  ef-
fects.  Congress has already made the  decision that mobile source
particulate emissions should  be  controlled  as  much  as technology
allows,,  with due  consideration  being given  to  leadtime,  cost,
noise, safety  and  energy.    A  2-year delay primarily  affects  the
factors of  leadtime  and cost.    A  more detailed analysis of  the
effect of delay on  each  of  these factors is  needed before the  delay
can be accepted or  rejected.

-------
                              -157-
     First, with  respect to  leadtime,  a 2-year delay would  allow
heavy-duty  diesel manufacturers  to  delay introduction  of  trap-
oxidizers until light-duty diesel  manufacturers  have  utilized  these
devices  for  three years.   Thus,  heavy-duty  diesel manufacturers
would  be able to  draw on three  years  of in-use experience (on
light-duty diesels) before having  to  use the  devices.   This  would
tend to follow the history of  the  oxidation and  three-way catalyst,
where  their  use  on heavy-duty  vehicles  followed their  use on
light-duty vehicles by  a  number  of years.  However, the situation
differs  here  in  a number of ways.  One,  Congress mandated the
delays  for  catalyst  use on  heavy-duty vehicles, but  in  Section
202(a) (3) (A) (iii)   of  the  Clean  Air  Act,  Congress  gave  the  same
mandate  for  particulate control to both light- and heavy-duty
vehicles and  made  no  provision for a special  delay to any class.
Two, the manufacturers  of gasoline-fueled light-duty vehicles are
for the most part  the  same as  those who  produce  heavy-duty gasoline
engines.  Thus, they bore the  cost of  developing catalysts for both
classes.   With respect  to diesels,  only  General Motors produces
large amounts  of vehicles/engines  in  both  classes, while the  other
four major  heavy-duty  diesel  manufacturers  produce few,  if  any,
light-duty diesels.  Thus, for the most  part, the  heavy-duty diesel
manufacturers  are not bearing  any  costs  associated with the light-
duty  particulate  standard  and a  delay based  on past precedence
alone does not appear  to be merited.

     The degree of benefit of an additional  2 years  of leadtime
depends  primarily  on  the degree  of difficulty  of developing  trap-
oxidizers for heavy-duty use for  1986.  If the task  is  a reasonable
one  for  1986,  the benefit of  waiting two  years is  not great.  If
completion of the  task is very questionable for  1986, then there is
a  greater  benefit from  delay.    In Chapter  IV,  the technological
argument for  applying trap-oxidizers  to heavy-duty diesels is
based,  in great  part, on their  use on  light-duty diesels.  To
determine the difficulty of development  for heavy-duty  application,
the development for light-duty application must  be examined  and any
pertinent  differences  between the   two  applications  considered.

     In  EPA's analysis of  light-duty diesel  trap-oxidizer avail-
ability, strong evidence  was  found to support  the availability of
trap-oxidizers in  1984.3/6/  However,  to minimize  the economic risk
of  the  0.2 gram  per  mile standard,  the  standard  was  delayed a
year. For our purposes here,  then, it  is reasonable  to  say that the
leadtime available prior to 1984  was likely sufficient  to develop a
light-duty trap-oxidizer  and  that  1985  represents a  certain margin
of  safety.    One  reason  for  this  safety  margin  is  the nonavail-
ability  of  nonconformance penalties  for  most   light-duty diesels.
If  an engine  family or  two could  not  meet  the 0.2 g/mi standard in
1984, then  there  would have  been  no  recourse  for EPA but to  pro-
hibit  their  sale   and  the economic  impact could have  been  quite
large,  depending on the projected  sales of  that family.   The
economic risk  is not nearly as great  for  heavy-duty  diesels due to
the  potential  availability of nonconformance penalties.  Thus, the
leadtime criteria  is  somewhat less crucial for heavy-duty  diesels

-------
                                -158-
th an it was  for  light-duty  diesels,  due simply to the differences
in the economic risks involved.

     With respect to  the  heavy-duty diesel situation, then, there
was enough  leadtime  available  (as  of March,  1980)  to develop and
produce trap-oxidizers for light-duty applications in time for the
1984 model  year,  and certainly  well  before the  beginning  of the
1985 model  year-  Given that  the  heavy-duty diesel model year
begins four months later  than  the  light-duty  model  year,  it would
appear that there is  about  one  full year of leadtime available to
the heavy-duty diesel manufacturers  after the date at which  light-
duty diesel  manufacturers were  expected  to have  a  trap-oxidizer
available to  them.   However,  it  is possible  that  the heavy-duty
diesel  particulate  standard will  not  be  promulgated until late
1981 and  this will  be  a full  year and a half after the  light-
duty regulation was  promulgated.  This delay would more than erase
the extra year of leadtime  between  1984 and 1985, unless the work
performed on  light-duty  trap-oxidizers  prior  to mid-1981  was also
applicable to heavy-duty  applications.

     Trap-oxidizer  research  has  been underway  for  well  over  two
years  and has  centered   primarily  on  light-duty applications.^/
However,  the  earliest  trap  work on  diesels  occured  on heavy-duty
diesels, TJ as prior  to 1977 there were very few light-duty diesels
sold in  the  U.S.  The difficulties  associated with  trap-oxidizer
development  center in three  general areas.  First  and  primary is
the  trapping  efficiency,  as  this sets  the upper  limit  on  the
effectiveness of  the  trap over  its  life.  Second  is the ability to
oxidize the  trapped  particulate, as  this  allows  the trap  to  be
regenerated  and useable for more than a few hundred miles.   Third
is  the durability  of the  trap material, both  with respect  to
structural durability and  to a  continued  efficiency  in  trapping.

     The  first  area, that  of   trapping  efficiency,   is  a similar
problem for both light- and heavy-duty diesels.   As described
in  Chapter  IV of  this document and  the light-duty regulatory
analysis, 3J  the character of the particulate  from both light- and
heavy-duty diesels  is similar,  if not indistinguishable,  given the
degree  of variation  within  a  single  vehicle's  particulates  and
that between  vehicles  in each  class.   Thus,  the  ability to trap
particulate  from  the  exhaust of vehicles in either class  should be
the same  and  traps  developed  for  one class should have  the same
trapping  efficiency on a  vehicle of  the other class   if  sized
properly.  Much of the work  already  performed on trap-ozidizers has
centered on  trapping  efficiency  and  a number of materials have been
found with an efficiency  of at  least 60 percent.*  These materials
(and thus,  the work  performed  in  this  area to  date) should have
equal applicability to heavy-duty applications.
*     See  Chapter  IV for technical  details  here and in the rest of
the discussion of  this  alternative.

-------
                               -159-
     Th e second area of development,  that of  particulate oxidation,
is one where  some differences between light- and heavy-duty appli-
cations could  exist.   The  primary difficulty in  this  area  is  to
keep or cause the  exhaust  temperature to  be sufficiently  high  to
start the oxidation of  the  particulate.   Also,  the oxidation must
occur with enough frequency to keep the maximum temperature of the
oxidation  process low enough to  protect the trap materials.   If too
much particulate is trapped prior to oxidation,  the ability of the
exhaust and outside air to  cool the trap can be overridden and the
temperature of  the trap  can exceed  its  design  limit.  The  most
important  criterion involved in designing such an oxidation system
is the exhaust temperature,  which  is  determined primarily by engine
design and  the  operating conditions  imposed on  it.   The  biggest
problem is  keeping the  exhaust  temperature  high  enough to  begin
oxidation.    Also,  the closer  the  oxidation can  be made  to  be
continuous (i.e., consistently high  temperatures),  the less  over-
heating is a  problem.   The  trap and the exhaust  system can easily
handle  the exhaust  temperatures,  even  at their  maximum.   It
is the  temperatures  of the combusting particulate in the  trap
itself that can cause  structural design limits to be exceeded.   The
temperatures normally occuring  in  light-duty  applications appear  to
be too  low to assure oxidation  at regular intervals under all
feasible operating  conditions.  Thus,  a  number  of techniques  have
been devised  to  raise  the  temperature  of  the exhaust.  Insulating
the  exhaust  system between  the  exhaust  ports and  the trap  is  a
passive system which  raises the  exhaust  temperature  at all times.
Others, such as  intake air  throttling or electrical heating at the
trap, operate  periodically to begin oxidation at regular intervals.

     The available  evidence  indicates  that  the exhaust temperatures
of heavy-duty diesels are higher than those  of light-duty diesels.
One  reason  is that the  horsepower-to-weight  ratios  of heavy-duty
diesels  are  much  lower  than occur with  light-duty diesels.
Because of this, the  former operate  at higher  relative loads  than
the  latter  where  the fuel/air ratios are  higher,  which causes
exhaust temperatures  to  be higher.   Turbocharging,  which  is  more
common on heavy-duty  diesels than  light-duty diesels  can  tend  to
counteract this, but further evidence indicates  that the effect  of
the  fuel/air  ratio is the overriding factor.   Analysis of the
heavy-duty particulate  test procedure has  indicated  that  higher
dilution ratios  are necessary  to  lower  the exhaust temperature  of
heavy-duty diesels  to  less  than  125°F (51.7°C)  than  is necessary
for  light-duty  diesels.^/  This  indicates  that  the heavy-duty
exhaust temperatures are higher, even with turbocharging.   Coupled
with  the  fact  that all  of the  temperature-raising  techniques
currently  being  examined  are  equally applicable  to heavy-duty use
as to light-duty use,  the  problem  of  ensuring periodic oxidation  of
particulate could  actually  be  easier  for  heavy-duty diesels  than
light-duty diesels  if  it  were not  for the  sometimes long (several
hours) periods of  time that heavy-duty diesels are  left to  idle.
This  could  present regeneration  problems  because exhaust  from
idling  engines  is cooler  than that  from  engines  under normal
operating  conditions.   If,  for example, a heavy-duty trap-oxidizer

-------
                               -160-
is very  near  the point where  it  needs  to  be regenerated when the
operator leaves  the vehicle  in  the  idling  mode for several hours,
particulate could build up to the point where the trap would clog.
This  problem, though  not insurmountable  is  a  unique  aspect of
heavy-duty trap-oxidizer applications which must be  addressed
before  they  are  applied to  these vehicles.   Also,  the  high-load
operation  that should  make  it  easier  to initiate regeneration may
also  make  it  easier for  the  trap to  overheat.   Thus,  some addi-
tional effort will be  required  to fully develop  a trap  for heavy-
duty  application  even  after  a  light-duty trap  is available.

     The third technological  area, that  of  trap durability, is also
an important  one  to  examine for  differences in  light-  and heavy-
duty application.  For  one,  the  mileage  life  of a heavy-duty diesel
is much longer  than  that  of  a  light-duty diesel (475,000  vs.
100,000 miles).   However,  in terms  of  time, the  lives of the two
types of vehicles are  about  the same  (9-10 years).   The  fact that
the mileages  are  very different while lifetimes are  the  same
indicates one of the differences  in the usage  patterns  of the two
types of vehicles.  Heavy-duty diesel  driving is  more concentrated
and continual (higher mileage per day).   Heavy-duty use also tends
to occur under more  warmed-up conditions.   This is evidenced by the
vast differences  in the cold start-hot  start weighting of the two
test  procedures  (43/57  for light-duty  and  14/86  for heavy-duty).
While higher  mileages  do  increase  durability problems, frequent
cold-hot operation should be a  more important  factor.   Given that
the lifetimes are the  same  and  that heavy-duty operation tends to
be more  warmed-up,  it  would appear that  trap  durability problems
for heavy-duty diesels should be  no greater than  those  for light-
duty  diesels.   An  additional  factor  would also  be the higher
exhaust  temperatures of  heavy-duty  diesels  mentioned in the pre-
ceding discussion.   These  should allow for  more continual oxidation
which should  definitely help to retain  trap efficiency  and struc-
tural stability.

     In all,  the  problems of developing a  trap-oxidizer  for heavy-
duty application  appear to  be only slightly more difficult than the
task  facing light-duty manufacturers; not  sufficient  to  justify a
three year delay between their  respective applications.   Also, the
work  performed to date  appears  equally  applicable to either class
of vehicle.   Certainly, light-duty diesel manufacturers  might have
more direct experience with trap-oxidizer operation than  do heavy-
duty  diesel  manufacturers at  the  present  time.   However,  this
expertise has been shared  with  the  independent trap suppliers and
can be  easily transferred to  heavy-duty  diesel manufacturers.
Thus,  the  tasks  of developing a heavy-duty trap  oxidizer for 1986
appears  at least  as  accomplishable  as  the task  facing   light-duty
diesel manufacturers  for  1985.

     Besides   leadtime,  the  other prime consideration  is one  of
cost.    Already mentioned  was  the $7.1  million cost of   recertifi-
cation  which  would  occur  whenever  an  emission  standard  is  sub-
stantially revised.   The  1986  standard  avoids  this  by occuring at

-------
                                -161-
the same time  as  the forthcoming revision of the NOx standard.  A
later standard, however, will bear the entire cost of recertifica-
tion.

     On  the  positive  side,  however,  is the  belief that  three
additional  years would allow  some improvements in design that will
reduce the cost  of production.    These  savings  could occur in two
ways.   One,  light-duty  production  experience could  lead  to more
economical  heavy-duty trap production techniques.  Two, light-duty
design experience  could lead  to more  economical  heavy-duty trap
designs.   The one year  delay  should  provide  for  some benefits here.

     The  first  effect,  while aiding  the  production  of heavy-duty
traps, does  so  at  some  expense  to light-duty trap production.  In
other words,  light-duty trap  production  will be  deprived  of its
benefit  from  heavy-duty  experience.   There  should still  be  a
positive  effect of delay,  as  it  is nearly  always more economical to
start on one project  and use  that experience toward the  next,
compared  to  starting both at once.   Also,  heavy-duty production
itself (in 1988)  is derived   from three years'  experience.   While
heavy-duty  traps  in 1988  (under  a  1988 trap-oxidizer  based stan-
dard) might  be  expected  to be cheaper than 1986 traps,  the former
will be  more costly than  a   1988 trap  under  a  1986  trap-oxidizer
based standard.  Thus,  some savings has been obtained by the three
year delay  relative to  the  first year  of trap  introduction under an
earlier  standard.   However, this savings is  obtained by 1) delaying
the benefits  of emission  controls  two years  and 2)  causing  the
eventual  use  of trap-oxidizers to be more  expensive than would have
occurred  in that year if the  standard  was  implemented earlier.  The
same  arguments  hold  for  the  effects  of design improvements.
Unfortunately,  the  available  data do  not  allow the quantification
of the net  savings.  However, an  estimate can be made.  In Chapter
VI, Table VT-1, it  can be  seen  that assuming a  12 percent learning
curve, trap-oxidizer costs decrease  20  percent between  1986  and
1988.  The  effect of a  2-year delay should be less than this since
direct heavy-duty  experience is  not   available.   Thus,   it  would
appear reasonable  to project  that delaying  two  years would reduce
the first  year of  trap-oxidizer  costs  by something less  than  20
percent.

     There  could  also  be  a   positive  effect  of delay on  captial
costs.  One,  the additional  light-duty experience could solve some
of the heavy-duty problems  and reduce  the  total heavy-duty research
effort.   Two, the delay might provide flexibility as  to the source
of  the necessary  capital and reduce  the cost  of  the capital.

     First,  it will be helpful to  examine the  actual  capital
expenditures  which  this  regulation  could  impose on the heavy-duty
diesel industry.   As  discussed in Chapter  VI,  there are  three
sources   of  capital  costs  which  are  related to  this  regulation.
First, there is  the cost  of  test  equipment,  which is  $2 million.
It  will  be borne  directly by the heavy-duty diesel  industry and
will occur prior  to 1986  regardless  of whether  or not  a two-year

-------
                              -162-
delay  is  granted.    Second,  there are  the  costs of  trap-oxidizer
development and  tooling  for  production, which have been  estimated
to  be  $6-8 million  and  $9-18 million,  respectively.   The  former
cost,  as  a capital cost, will  likely  be split between the  heavy-
duty diesel  industry and  the trap  suppliers and  would  at  least
partially be delayed  if  a  two-year  delay in  the  final  particulate
standard were granted.  The latter will almost entirely  be borne  by
independent  suppliers  and  would  almost entirely be  delayed by  a
two-year delay.   Third,  there is the  cost  of engine redesign and
tooling and this could range  between $4 and  $16 million.   This cost
will be  borne  entirely  by  the heavy-duty diesel manufacturers and
would  not  be affected by a delay in the  trap-oxidizer  based stan-
dard.   In all,  the  heavy-duty  diesel  manufacturers and  suppliers
will be  required  to  raise about $21-44 million because of this
regulation  and  between  $15-26  million  would be deferred   if the
final  standard  were delayed two  years.

     As can be  seen from the  size of these capital costs,  the total
requirements are not very large  for  five major engine manufacturers
and their suppliers, and the capital costs which would  be  deferred
by  the two-year  delay are  also  not  significant.   A two-year delay
will defer  capital  expenditures  of  between  $15-26 million for the
manufacturers and suppliers and  would impose  a recertification cost
of  about $7 million on the manufacturers.  These numbers  show that
the cost savings from a  two-year  delay  may not outweigh  the capital
expenditures necessary to meet a model year  1986 standard.   Given
that trap-oxidizers can be available in  1986  and that the  benefits
of  a  two-year  delay do  not  appear  to substantially  outweigh the
costs  of  a model year 1986  standard,  EPA is not proposing  a two-
step approach  at  this  time.    However, EPA  will  reconsider this
approach if additional data warrants such action.

     The  alternatives  remaining  are  1)  the  implementation date
of  the one-step  standard and  2)  the level of  this  standard.
In  analyzing the question  of a  one-  or  two-step  standard  above,
however,  the implementation date  of  the  one-step standard has
been all but determined.   From  the  above analysis the  choice must
be  1986.   That  is the year  the  trap-oxidizer should be  available
and the  year of  the  revised NOx standard.    Thus,  the only real
choice  remaining is that  of the  level of the  standard.

     The methodology used  to  set the  level  of the  proposed  stan-
dard has  been  outlined in  detail   in Chapter  IV.   In  essence,
the  level  is  based  on  1)  an  engine-out  particulate  level  of
0.41 g/BHP-hr  (0.153 g/MJ),   2)  the use of  a trap-oxidizer, and
3)  the reservation  of  certain   engine-related control  techniques
(e.g.,  high-pressure injection)   for  the mitigation of  particulate
increases due to NOx control in  1986.   The  alternatives to setting
the  technologically-achievable   level   of  engine-out   particulate
emissions at 0.41 g/BHP-hr (0.153 g/MJ) were  considered in Chapter
IV  and the logic  for  choosing   this level  can be  found  there   in
detail  also.  It will not  be repeated   here,  except that  the prime
consideration was the Clean Air  Act  mandate  to achieve the  greatest

-------
                               -163-
emission  reduction  possible,  while  taking  into account leadtime,
energy, cost, and safety.

     The  second  factor  is the use  of  a trap-oxidizer as a viable
control technique.   Given that  the device should be available for
use on heavy-duty diesels in  1986  and  that the  analysis  in Chapter
VII  shows  it to  be  a cost-effective  control  technique,  it  would
appear to violate  the  congressional mandate  not to base the stan-
dard on  its  use.  Thus,  the  alternative of  rejecting its  use was
rejected.

     Finally,  the  factor  of  the  1986  NOx  standard  requiring
a  75  percent reduction  from  baseline  levels  must be considered.
The  methodology leading to  the  proposed  level  of particulate
control  reserves  the use  of  some  particulate  control   techniques
for  their possible  use  in reducing  the negative effects of NOx
control.   In complying  with the  mandate of  the Clean Air Act
with respect to  particulate control, one  could also  have  con-
ceivably taken the opposite stand,  and  set  the  particulate standard
based on  every available  control technique and  left no cushion for
increases due to  NOx control.   Which is the proper choice  in this
case?

     It  is  known  that  certain  NOx control  techniques  can  cause
increased  particulate  emissions,   namely  exhaust  gas  recircula-
tion and  retarded  timing.   At  the same  time, other techniques
do not  have  this  trade-off.   This  is  evidenced by  the  fact  that
many of the lowest particulate emitters also have low NOx emissions
(Figure  IV-1)  and the  Cummins  and  Caterpillar  experiences  where
redesigns of certain engines have  reduced both particulate  and NOx
emissions (Chapter IV).   There  is  also the specific congressional
mandate  calling for a  75 percent reduction  in  heavy-duty NOx
emissions from uncontrolled baseline levels.   As this NOx  mandate
is more specific  than the particulate mandate,  it would seem proper
that the  particulate standard impact  the  achievability  of  the  75
percent  NOx  standard as little as possible.   The use of  trap-
oxidizers complies with  this approach as trap-oxidizers do not have
an adverse  effect on NOx emissions.   Also,  the engines used  to
determine the 0.41   g/BHP-hr engine-out particulate  level  had
relatively low  NOx  emissions  as  well as the  lowest particulate
emissions.   However,  this  still  has the effect of setting  a  limit
on  future particulate  increases   since  these  low NOx  levels  are
still far from the level  expected  to be required in 1986.  To rule
out any increases in  particulate  emissions  entirely would appear to
overly restrict  the  ability  of  heavy-duty diesel manufacturers  to
meet the  1986 standard  as well as restrict the  Agency from at-
taining those required  reductions.  Thus,  some allowance  appears
reasonable.    However,  the Agency has  not  yet  determined what NOx
level  is  achievable  by  heavy-duty  diesel and  what  would be  the
effect of various levels  of particulate control.  This information
will be  gathered as  the  Agency  proposes and  promulgates  the NOx
standard.

-------
                              -164-
     At this time then,  it  is  not  possible to quantify the partic-
ulate  allowance  required.   Yet  some allowance  is appropriate
to  balance  the  two  congressional  mandates.   The  present  allow-
ance would  appear reasonable  in this  vein;  that of the exclusion
of  some  control  techniques  from consideration in  setting  a  tech-
nologically-achievable  particulate  standard.    Without  being
able  to  quantify the  effect of  NOx  control at  this time,  one
could  of course  argue for  a larger allowance  to  be safe,  or
for a  smaller  allowance to propose the most stringent particulate
standard  conceivable.    Without  data,  it is difficult  to  defini-
tively argue  against  either view.  However,  the  presence  of
arguments on either side  calling  for  changes  in  the  standard  in
opposite  directions  is  in  itself some evidence of reasonableness.
Thus,  on that  basis,  the  decision was made to  give the  above
mentioned allowance.

     It now  appears  that  the 0.25 g/BHP-hr (0.093  g/MJ)  partic-
ulate standard  is the  best  alternative available.   It is  based  on
trap-oxidizer technology, which does not  affect NOx control  and  on
some of the  best  existing engines with  both low particulate and low
NOx emissions.  The standard also allows  for  some increase  in
engine-out  particulate  levels due  to  further NOx  control and
reserves  particulate  reductions   available from  other control
technologies for negating these  increases.    It  appears  to  be
cost-effective  (Chapter  VII),  to  be a necessary standard  for  the
protection  of both the  public health  and welfare  (Chapter V)  and
to comply fully with all  congressional mandates (Section 202(a)(3)
(A)(ii) and  (iii)).  Thus, it should be proposed.

-------
                                -165-
                            References

_!/   "Summary  and  Analysis  of  Comments  to  Proposed  Particulate
     Regulations  for  Light-Duty  Diesels,"  MSAPC,  EPA,  October,
     1979.

2j   "Impact of  New Source Performance  Standards  on  1985 National
     Emissions  from Stationary  Sources," EPA-450/3-76-017,  April
     1977.

_3_/   "Regulatory  Analysis,  Light-Duty  Diesel  Particulate  Regula-
     tions," MSAPC, OANR, EPA, January 29, 1980.

4_/   "1975  National Emissions Report," OAQPS,  EPA,  May  1978,
     EPA 450/2-78-020.

5/   "Regulatory  Analysis  and Environmental Impact of  Final  Emis-
     sion  Regulations   for  1984  and Later  Model  Year  Heavy-Duty
     Engines," OMSAPC, EPA, December 1979.

6/   45 FR 14496, March 5, 1980.

Tj   Shahed,  Syed  M.,  Personal Communications  with  Richard  A.
     Rykowski, EPA,at  the Symposium on Diesel  Particulate Emission
     Measurement  and Characterization,  May 17-19,  1978,  Ann Arbor,
     Michigan.

8/   Reiser,  Daniel P.,  "Summary  and  Analysis  of Comments to  the
     Draft  Recommended  Practice  for  Measurement  of  Gaseous  and
     Particulate  Emissions  from Heavy-Duty  Diesel  Engines  Under
     Transient  Conditions,"  Technical  Report,   SDSB,  EPA,  August
     1980.

-------
                            Appendix I

     An  estimate  of  the nationwide  fraction,  Fn ,  of  heavy-duty
vehicle-miles traveled (VMT) attributable  to  diesels  in 1995  can be
obtained from the  following  equation:
              20         20
             g Z(ab)£ + h Z
 rn     20         20          20
       g Z(ab)i + h Z(cde)£ + h Z<
Where:

       ai =   fraction  of total  registration of  Class  VII  and
              VIII heavy-duty diesels  (HDD's)  i  years  old;

       bi =   annual mileage  accumulation rate  of Class  VII  and
              VIII HDD's;

       ci =   fraction  of total  registration rate  of  Class  II
              through VI heavy-duty vehicles  i years old;

       di =   annual mileage accumulation rate of Class II  through
              VI heavy-duty vehicles i years  old;

       ei =   diesel sales  fraction of Classes II  through VI  for
              ith model year  (i =  1 being 1995, i =  2 being  1994,
              etc.);

       fi =   gasoline sales fraction of Classes II through VI  for
              ith model year;

       g  =   0.31,  the  fraction  of  total  heavy-duty  sales  from
              Classes VII and VIII  (see Table  III-7);

       h  =    0.69,  the  fraction   of  total  heavy-duty sales  from
              Classes II through VI (see Table III-7);

The  two  items  in  the numerator  represent  diesel  VMT  in  Classes
VII and VIII and diesel VMT in Classes II through  VI,  respectively.
The  third  term  in the denominator  represents the gasoline VMT in
Classes II  through  VI.   Based on discussion  in Chapter III,  it is
assumed  that gasoline - powered  vehicles  constitute  a.  negligible
fraction of Class VII and VIII heavy-duty vehicles.

     Values  for  the  above variables  are  given  in Table A-l.   The
fraction  of  total  registration  and   annual  mileage  accumulation
rate of Class VII  and VIII heavy-duty diesels are  taken  from EPA's
Mobile  Source Emission  Factors document .J_/    Although  they  were
intended to  apply  to all  classes of heavy-duty  diesels,  the record
of  past  heavy-duty diesel sales  (see Table  III-2) indicates  that

-------
                                                Table A-l
                 Heavy-Duty Diesels
                 Classes 7 and 8  I/
Heavy-Duty Vehicles
 Classes 2 thru 6


Model Year
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
1978
1977
1976
I/ From
2/ From
(a)
Fraction Total
Registration
0.077
0.135
0.134
0.131
0.099
0.090
0.082
0.062
0.045
0.033
0.025
0.015
0.013
0.011
0.010
0.008
0.007
0.006
0.005
0.004
Table IV-5 of Mobile
(b)
Annual Mileage
Accumulation Rate
73600
73600
69900
63300
56600
50000
45600
41200
38200
36000
34600
33800
33100
32400
30900
28700
25700
21300
18400
15400
Source Emission F
Table III-5 of Mobile Source Emission
3/ Columns (e) and (f) are
used to estimate
(c) 21
Fraction Total
Registration
0.037
0.070
0.078
0.086
0.075
0.075
0.075
0.068
0.059
0.053
0.044
0.032
0.038
0.036
0.034
0.032
0.030
0.028
0.026
0.024
(d) 21
Annual Mileage
Accumulation Rate
19000
19000
17900
16500
15000
13500
12000
10600
9500
8600
7800
7000
6300
5900
5300
4900
4700
4600
4400
4200
(e) J3/
Diesel Sales
Fraction
0.46
0.43
0.41
0.38
0.36
0.33
0.30
0.28
0.25
0.22
0.19
0.17
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
(f) 3/
Gasoline Sales
Fraction
0.54
0.57
0.59
0.62
0.64
0.67
0.70
0.72
0.75
0.78
0.81
0.83
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
                                                                                                                  ON
                                                                                                                  -J
                                                                                                                  I
                                               capture by  diesels in classes 2 thru 6;
based on Chapter 2 sales estimates.

-------
                                -168-
Classes VII and VIII  constituted  the  great  majority of diesels on
the road at the time of the document's publication-1978.  For this
reason, the values given in Mobile Source Emission Factors for the
fraction of  total diesel  registration  by vehicle  age  and  annual
mileage accumulation rates were  in this study used  for Class VII
and VIII diesels only.

     Values of  these  parameters  as  they apply to heavy-duty gaso-
line engines  are  also taken from  Mobile  Source  Emission 'Factors.
Since sales estimates outlined  in Chapter 3 project the capture by
diesels of portions  of the  existing heavy-duty gasoline market, the
assumption has  been made that  the mileage accumulation and  yearly
registration  fraction  characteristics  of the  heavy-duty gasoline
vehicles (predominantly Classes II  through VI) also apply to diesel
vehicles in those classes.   That  is, diesels in Classes II through
VI will  have useage  characteristics similar to their gasoline
counterparts  rather  than  the  heavier Class VII  and VIII diesels.
This is  apparent  since,  for example,  a Class III  delivery truck
will make  the  same  number of deliveries per day whether  it is
gasoline or diesel powered.

     In order to  add  the diesel  VMT  fraction  from Classes VII and
VIII to the diesel VMT fraction  from Classes  II  through VI, the 2
categories  must be  normalized  for sales  (the 0.31 and  0.69  fac-
tors).    After incorporating  this normalization and  the  values in
Table A-l  into  the aforementioned equation,  the  nationwide frac-
tion, Fn,  of  heavy-duty  VMT  in  1995  due  to  diesels is determined
to be  78.6 percent.   Because sales  projections as well  as  any
assumptions are  subject to  error, this study projects that  a range
of 71.5 to  86.5 percent  of nationwide heavy-duty  VMT in 1995 will
be attributed to diesels; reflecting  a 10 percent margin of  error.

     The same methodology was  followed to  determine the  urban
fraction,  Fu of  heavy-duty VMT in  1995  due to  diesels.    This
value is given by:
 FU
where all parameters  except  the urban  fraction of heavy-duty diesel
VMT (Classes VII and VIII),  j (equal to 0.33), and the urban frac-
tion of  heavy-duty  gasoline VMT  (plus Class  II-VI diesels),  k
(equal  to 0.43), are the  same  as  those used to  determine Fn.   The
urban-rural  breakdown was  obtained  from a PEDCo report based on DOT
data._2/

     The urban  fraction,  Fu,  of  heavy-duty VMT by diesels  was
thus determined  to  be 74.6 percent;  67.1-82.1 percent, allowing for
a 10 percent margin of  error.
20 20
jg Z(ab)^ + kh Z(cde)^
20 20
1 n T I 3 r» j • 4- If Vl ¥ i f* t^ & \ * J
Jg i, \ d U J -t ~ ixi I A^*-U.c^-|
20
H kh Z(cdf)i

-------
                                -169-
                            References

W   "Mobile Source  Emission  Factors," EPA March  1978,  EPA-400/9-
     78-005.

2J   Air Quality  Assessment  of Particulate Emissions  from Diesel-
     Powered Vehicles,  PEDCo  Environmental  for  EPA, March  1978,
     EPA-450/3-78-038.

-------
                                -170-
                           APPENDIX  II

     Appendix  II contains a  detailed  cost analysis for  trap-
oxidizer system components and for potential savings due to elimi-
nation of muffler and exhaust  system maintenance.

A.   Emission Control System Costs

     The technology  necessary  to meet the  1986 particulate emission
standard was discussed in Chapter IV.  Heavy-duty engines  are
expected to be able  to  meet the  1986  standard with trap-oxidizers
along  with incorporating the design  features  of those current
engines  with low  particulate  emissions.   The trap-oxidizer repre-
sents additional  equipment  and  will  increase  the cost of the engine
(and vehicle).   The  design  modifications, however, should not raise
production costs, except  through  the amortization  of  new tooling
and  engineering  costs.   These design  features  of  the  lower par-
ticulate emitting  diesels are  present  on these  engines  at  no
apparent price differential and  should be  similarly  available  to
others.  It is  possible  that some of these heavy-duty vehicles will
be able  to use other techniques to  meet the standard;  however,  to
be conservative,  this economic  analysis  will  assume that  all
vehicles will require trap-oxidizers.

     In  summary,  EPA estimates the average cost  of a trap-oxidizer
system for heavy-duty vehicles  to be $521-$632 (1980 dollars).  The
cost of  the trap itself represents about 80 percent of this total.
Necessary modifications  to the engine  and exhaust system represent
10 percent of the total cost.   The  remaining costs are associated
with the control system used  to initiate  oxidation  of  the trapped
particulate.   The use of the  trap-oxidizer  system as  described  in
this  section  should also  reduce  maintenance costs by  $197  (1980
dollars, discounted back  to  year of  vehicle  purchase) due  to
reduced  exhaust  system  maintenance.   A detailed analysis of  the
cost  estimates  for trap-oxidizer components follows.

     Because  the  costs of trap-oxidizers will likely depend on the
production volume,  the first step in this analysis will  be  to
estimate  heavy-duty  diesel production  volumes   between  1986  and
1990.  This five-year period was  chosen because  it will correspond
with  the period  used to calculate the  aggregate  cost  of  the 1986
standard,  which will be  performed in Section  C of Chapter  VI,
Economic Impact.

     Projections  of  overall heavy-duty diesel production  and  the
breakdown by vehicle class from  1986  to 1990 are  needed  in this
analysis.  These  projections can be  found in Section III,  Descrip-
tion  of  Industry,  in the  discussion  of  future  heavy-duty  diesel
sales.    Table  A-II-1 shows the breakdown of heavy-duty  sales  by
vehicle  class projected  for 1986  to  1990.   As  will  be discussed
later in this section,  nearly all the  costs  of  components  of the
trap-oxidizer system will be  dependent  on engine  size,  which will
be assumed to  be related  to  vehicle  class.   For  the  purposes  of
this  analysis,  it will be assumed that the same basic trap-oxidizer
system can be used within each of  four vehicle groups.   The groups

-------
                                -171-






                               Table A-II-1




                Projected Heavy-Duty Diesel Sales By Class

1986
1987
1988
IIB
22,316
26,511
30,841

4
4
5
III
,179
,965
,776
IV
710
843
981
V
1,735
2,061
2,398
VI
72
80
89
,490
,984
,755
VII
35
37
40
,824
,979
,187
VIII
165,600
168,623
171,645
TOTAL
302,854
321,966
341,583
1989   35,308   6,612   1,123   2,745   98,804   40,895   174,668   360,155




1990   39,910   7,474   1,269   3,102  108,132   41,602   177,691   379,181

-------
                                -172-
are gross vehicle weight dependent and each group  consists  of  one
or more of the  traditional heavy-duty vehicle classes.

                      Gross Vehicle Weight
             Group         (Pounds)            Classes

               1          8,500-16,000        IIB,III, IV
               2         16,001-26,000          V, VI
               3         26,001-33,000           VII
               4         33,001 and over         VIII

Classes III  and  IV were grouped  with  Class  IIB due to  the small
relative sales  of Classes III and IV.   The  same reasoning applies
for grouping Class  V with Class  VI.   Vehicle  Class IIB  in  this
section will always refer to vehicles  in  the  traditional  Class  II
category with  a  weight  above 8,500 pounds  (i.e.,  those  Class  II
vehicles  which  fall  into EPA's heavy-duty vehicle  category).

     In manufacturing, it is  a common  occurrence that  the cost  of
production decreases with experience.   This  experience  is usually
measured  in  terms  of accumulated production.   The  relationship
between  cost  and  accumulated  production  is called  a  learning
curve  and  is  usually  described by   the  logarithmic  function:

                         ,ln(1.0+zK
                          ~
              _
              Cl

Where :

P} and  ?2 = two   different  levels  of  accumulated  production.
        z = the  fraction  or  percentage that  costs are  increased
            each time  the  accumulated production is halved.
GI and  C2 = costs  of  the  item with  total  accumulated produc-
            tion of PI and T?2-

     For the purposes of  this analysis,  z  will be assumed  to  be
0.12, or that the  cost  of a  trap-oxidizer system will  increase  12
percent each time  the  accumulated production is halved.*  Given the
cost at  a  specified  accumulated  production,  a new cost  at  a dif-
ferent  production  can  then be  found using  equation (1).   The effect
of this assumption will  be examined  in Section C of this appendix,
"Sensitivity Analyses,"  where  costs  will be  calculated  assuming
that a  learning  curve  does not apply.

     It is highly unlikely that  each manufacturer will  produce his
own  trap-oxidizer for  three  reasons.   One,  the  area involves
sophisticated   technology that each  manufacturer  cannot  really
afford  to  develop  independently.  Two,  a number  of  firms  have
already developed  an  expertise  in  the area in  response  to light-
duty diesel  particulate  standards  and  have  a head-start   on  the
*    This 12 percent  factor  is the same as that used on page 107 of
the Regulatory Analysis  for the Light-Duty Diesel Particulate
Regulations, which  in  turn  came  from a contractors  meeting  with
LeRoy Lindgren  of  Rath  and  Strong on August 18, 1979.

-------
                                -173-

heavy-duty diesel manufacturers.   Three,  the  production volumes  of
many  heavy-duty  diesel manufacturers  are  too small  to justify
in-house  development  and production  given  that the expertise  and
production capability will likely exist outside.  Thus,  for  costing
purposes,  it  will be  assumed  that there will  be only three  sup-
pliers of trap-oxdizers , each having a third  of  the  market shown  in
Table A-II-1.  However, the  effect  of  this assumption will  be
examined  in Section C  of  this  appendix,  where costs  will  be  deter-
mined  assuming that  each manufacturer  produces his  own  control
systems .

     The  cost  of  each  component  to a trap-oxidizer supplier  will
depend upon his  total production,  and  the  number of  different
components needed.   For items such as traps, throttle  assemblies,
exhaust  pipes,  etc.,  a different  component will  be assumed to  be
needed for each vehicle group.   To facilitate the costing of these
components, each  vehicle  group will be assigned an  average  engine
displacement.    These engine  displacements are 5.7 liters  (350  CID)
(Classes  IIB,  III,  and  IV),  8.2  liters  (500 CID)  (Classes V and
VI),  10.5 liters (640 CID) (Classes VII), and  13.9 liters  (850  CID)
(Class VIII).   Once again assuming that each  trap-oxidizer supplier
shares a third of the  production  as  shown  in  Table A-II-1, using
equation  (1)  the  average cost  to each trap-oxidizer supplier per
vehicle  group  is as follows:

                      j            ln(1.0 +  z)
                                -    ln 2
Where :

     Cij  =    Cost of item in vehicle group  i, year  j.
     Cref =    Cost of  component  at  a  production  volume of Pref.
     Pij  =    Production of vehicle group  i,  year  j.
     Pref =    Reference production volume.

     For other  items  such as thermocouples and electronic control
units,  one type can be used on all heavy-duty  diesel  engine models.
In  this case,  the trap-oxidizer suppliers  will produce  these
components at  one-third  the total  annual engine fleetwide produc-
tion.  The average cost per vehicle group will  then be:

                       APk    , In (1.0 + zK
                              (n~2       }                        (IB)
            P  4=   ,
    Cij  =  Cref x (3 x pref )
Where :

                                            M
     APj  =    Total production in year }  = % Pij
     M    =    Number of vehicle groups.

-------
                                -174-
     Th e fleetwide average cost  is  simply  a  sales-weighted  average
of the costs to each trap-oxidizer supplier and is described by the
equation :

               M
               Z Cij  x Pij
          .   i=l                                                 (2)
     Cave,;, =
Where:

     Cave,j =    Sales-weighted average cost  in year j.

     Equations 1A and  IB  can be substituted  into equation  2.   For
traps  and  other components that  differ among vehicle groups,
equation (2) becomes:
                       M         j                      In 2
               Cref    £(Pi,j     Z    Pik      .                 )   (2A)
     Cave,j
                      . = 1       k=1   3

Similarly, for electronic control  units,  equation  (2)  becomes:
                       M         j                  •    In 2
     Cave i  = Cref x  Z(Pi>J  x  l  (  APk                       >
     Cave.j      .   x .=      x  =(
     If 0.12  is substituted  for z,  then equation  ( 1A)  and  (IB)
become, respectively:

                   j             -0.164                           (3A)
     Cij = Cref x
                   j             -0.164                           (3B)
      ii = Cref x .  ,
Also, equations ( 2A)  and (2B)  become:
                       M         j                -0.164
     Cave,j  -£ref    Z(Pi,j     I   Pik	 ,           )         (4A)
         >J    APj  x i=l      x k=P  3 x Pref ;

-------
                              -175-
                       M         j               -0.164
               Cref    E(Pi,j    2 , APk      .           )         (4B)
     Cave, j  = AT>.  x .  ,      x ,  , ( -=	~—F )
          J    APj    i=l        k=l   3 x Pref
     Once the reference  production  (Pref)  is  chosen  and  the  refer-
ence cost determined,  equations  3A, 3B, 4A,  and  4B  will  allow the
costs  for each  vehicle group and the average  cost over  the  entire
fleet to be determined in any given year.

     Two  final adjustments must be made here  to  determine  the
actual cost to  reference  cost  ratio for each  item.   First,  it  will
be  assumed  that two traps  will  be  required  for  each  Class  V-VIII
vehicle.   The  (Ci,j/Cref)  and (Cave,  j/Cref)  ratios  for each  of
these  traps can be calculated  by multiplying  the results of equa-
tions  2A, 3A, and  4A by 2 to the -0.164 power, or 0.89.  Second,  it
is  expected that  items  other  than  electronic  control  units  and
traps  will  be  manufactured according to each  basic  engine design.
For  purposes  of this  analysis it  is  assumed that  the  heavy-duty
diesel industry consists  of  about ten basic engine designs.  These
can be broken down into  two  designs for Class IIB-IV vehicles,  two
designs  for  Class  V and  VI  vehicles,  three designs for  Class  VII
vehicles, and  three designs for Class VIII vehicles.  Assuming  an
equal  number of  engines per  engine design  within each vehicle
group, the actual  cost to reference  cost ratio for these  components
can be calculated  by dividing  the results  of  equations 2A, 3A,  and
4A  by  2  to the -0.164 power,  or 0.89,   for Class  IIB-VI  vehicles,
and by 3 to the  -0.164  power,  or  0.84,  for Class VII and VIII
vehicles.

     The  production data  shown  in  Table  A-II-1 can  now be  used
directly  to  calculate  (Cave,j/Cref) for the  years  1986-1990 (j =
1-5).  Pref will be set at 300,000 units.  The results  are shown  in
Table A-II-2.   As can be seen,  the cost  of  components such as traps
starts out 83 percent greater for Classes IIB, III, and IV than  the
cost at  an  accumulated  production  of  300,000 units  (1986)  and 4
years  later  is  only  32  percent greater than the cost at 300,000
units.   A  similar  result  occurs  for the  Class V-VIII  traps.
In  1986,  the  cost is  37 percent  greater than the  cost  at  the
reference production and by 1990 the cost is  1 percent  greater  than
Cref for Class  V  and VI traps.   For  electronic  control units,  the
cost in  1986  is 21 percent greater than the  cost  at the  reference
production and by 1985  the cost  is  9 percent less than  Cref.
Similar results occur  for components that  vary with  engine design.

     EPA's original cost  estimates  of the  individual components  of
a  trap-oxidizer system were  taken  from  a study of the costs  of
emission  control   systems.I/   The  formula used  to  determine   the
retail  price equivalent of each item is  shown  below.

Retail             _.   „      _.    .     Fixed
Price       = [[(MD,lreCt1) + (?"eCt) + (Variable)]
                 Material     Labor       _.   ,    ,
Equivalent                               Overhead

-------
                               -176-
                              Table A-II-2

            Values  for the Ratio of  the Actual  Cost  of  a
          Component  to Its Cost  at an Accumulated  Production
                          of  300,000 Units  I/
Electronic Control Units
1986
1.21
1987
1
.07
1988
1.00
1989
0
.95
1990
0.91
Cost Ratios if Component Pro-
duction Equals Vehicle Group
Production
Class
Class
Class
Class
Traps2/
Class
Class
Class
Class
IIB, III, IV
V, VI
VII
VIII

IIB, III, IV
V, VI
VII
VIII
1.83
1.53
1.71
1.32

1.83
1.36
1.52
1.17
1
1
1
1

1
1
1
1
.61
.36
.52
.18

.61
.21
.35
.05
1.48
1.26
1.42
1.10

1.48
1.12
1.26
0.98
1
1
1
1

1
1
1
0
.39
.19
.34
.05

.39
.06
.19
.93
1.32
1.13
1.29
1.01

1.32
1.01
1.15
0.90
Components that Vary with
Engine Design3/
Class
Class
Class
Class
IIB, III, IV
V, VI
VII
VIII
2.06
1.72
2.04
1.57
1
1
1
1
.81
.53
.81
.40
1.66
1.42
1.69
1.31
1
1
1
1
.56
.34
.60
.25
1.48
1.27
1.54
1.20
_!_/   Assumes each supplier shares one-third of the market at
production shown in Table A-II-1.

2f   Assumes 1 trap for each Class IIB-IV vehicle, and two traps
for each Class V-VIII vehicle.

2/   Assumes 2 basic engine designs for Class IIB-IV vehicles, 2
basic engine designs for Class V and VI vehicles, 3 basic engine
designs for Class VII vehicles, and 3 basic engine designs for
Class VIII vehicles.  These components include port liners, stain-
less steel exhaust pipes, insulated exhaust pipe, insulated exhaust
manifold and throttle-body actuators.

-------
                                -177-
                      (n r> Corporate ,.    en 9 Supplier-.,    ..Tooling.
                           Allocation      '   Profit        Expense
    Land &                                   _                Dealer
  f-r.  •-i , •  \i   ri   /r\ n Corporate \    /n _ Corporate\    /,. ,
+ (.Building;] x 11 + v.0.2 A11   ._ •   )  + V.U.2   _  _.   )  + \(J.^
                          Allocation           Profit            „  ...
   Expense                                                     & Profit

  -.Research & .    ,Tooling.                                    f ^
   Development     Expense


     RPE = [(DM + DL + OHX1.4) + TE + LBE](1.8) + RD +  TE         (6)

Direct materials entail those  materials  of  which  a given component
is comprised.  Direct labor  includes  the  cost  of  laborers  directly
involved in  the  fabrication of  a  given  component.   Overhead  in-
cludes<  both  the  fixed  and  variable components of overhead.    The
fixed portion includes  supervisory salaries,  building maintenance,
heat,  power,  lighting,  and other costs which are  substantially
unaffected by production volume while  the variable portion includes
small expendable tools, devices, and  materials  used  in  production,
repairs  and  maintenance made  to  machines  directly   involved,  and
other overhead costs which  tend  to vary  with  production  volume.   A
straight 40 percent  of  the  direct  labor  amount was  used  to deter-
mine all overhead costs.

     A figure of 20  percent  applied to  the  sum of material, labor,
and overhead costs was  used  to determine  corporate allocation.   In
other words,  this  is  the  amount  needed to  cover the  supplier's
support from its  front office.  Also,  to the sum  of  material,
labor,  and overhead  costs,  a  figure  of  20  percent was  applied  to
determine  the  supplier's profit.   Approximately  half  of this  20
percent is used  to  pay  corporate  taxes with  the  remaining  portion
being divided  between  dividend  disbursements  to   stockholders  and
retained  earnings,  which  are  used  to  finance  working capital
requirements (increases in current assets  and/or  decreases  in
current  liabilities) and/or  new  capital expenditures   (long-term
assets).

     Tooling  expense consists  of   four  components:   one  year  re-
curring tooling expenses  (tool bits,  disposable jigs  and  fixtures,
etc.);  three  year  non-recurring  tooling  expenses  (dies, etc.);
twelve  year machinery and  equipment expenses;   and twelve  year
launching costs  (machinery  foundations  and  other  incidental set-up
costs) which was assumed to  be 10  percent of  the  cost of machinery
and equipment.

     The sum  of  the  above costs,  material,  labor, plant  overhead,
tooling expense,  corporate   allocation,  and profit,  makes up  the
price (or,  in the case to   a  division,  transfer  price) which  the
supplier charges  the vehicle  manufacturer  for a  given  component.
At the vehicle assembly level, 20  percent of  this  price  is charged
or allocated for  the vehicle manufacturer's  corporate level support
and 20 percent for corporate profit.  Also, a figure of  40 percent

-------
                                -178-

is applied to the supplier price to account  for  the  dealer's  margin
which includes sales commissions,  overhead,  and  profit.

     There  is  a  need,  in  many  instances,  to  make modifications
to the engine or body to incorporate a  component and to  assemble  it
into a  vehicle.   These costs  have also been accounted for  at  the
division  level  and  transferred to  the  corporate level at vehicle
assembly.

     Lindgren's  study  primarily  focused on  determining the  manu-
facturing  costs  of  emission  control equipment.   Much  effort was
expended  to accurately  determine  the  cost of materials,  labor,
tooling,  etc.    EPA  has available  a number of confidential  cost
estimates from emission-control equipment suppliers  and  these costs
confirm Lindgren's estimates at the vendor  level.

     Less  resources were  available to  Lindgren to  determine  over-
head costs and profit margins  and,  in general,  rules of thumb were
used in equation (6).   These estimates  of overhead  costs and  profit
margins at the corporate and dealer levels would profit  from  a more
detailed  analysis.   Overhead and   profit at  the vendor level  will
not  be  reexamined because  the independent  vendor  estimates  men-
tioned  above confirmed Lindgren's estimates  up  to that level.

     The  first two factors  to  be  examined are those  indicating the
corporate  overhead and  corporate  profit.   Typical  levels  of  over-
head and  profit  can  be obtained  from Moody's Industrial Manual. 2J
For diesel engines,  EPA examined the 1976,  1977, and 1978 financial
data for five manufacturers:  General Motors, Cummins, Caterpillar,
Mack,  and  International  Harvester.   The  corporate  overhead and
profit  (in  terms of the  fraction of the cost  sales)  for  each  of
these manufacturers  in 1976,  1977, and 1978   are  shown  in  Table
A-II-3.   The before-tax corporate profit  (as a percentage of  cost
of sales)  ranged  from  5.5  percent to 17.2  percent  with an average
of about  11.9 percent.   The corporate overhead of  the five  manufac-
turers  (as  a percentage of cost of sales ) over the 3  year  period
ranged  from  8.5  percent to 33.6  percent with  an  average  of  16,7
percent.

     If  Cummins'  corporate overhead is excluded  from  the  latter
range, the range of corporate  overheads is narrowed  to  8.5 percent
to  19.3 percent.   With  the  exclusion of  the Cummins  overhead
figure,  the  two ranges  (profit  and overhead)  are actually  quite
small, considering the variety of  firms involved and the number  of
years being  examined.   The  Cummins overhead figure requires  some
examination.   Cummins  does have  the most  limited  product  line  in
that  they manufacture only diesel engines.   All of the  other
manufacturers also  produce vehicles of some sort  in  addition  to
engines.  This difference could be  the  cause of an  increased level
of overhead.  However,  it  would seem more  likely that  the differ-
ence would be associated with  the method of  accounting  rather than
an  actual difference  in  the  level  of  overhead.   Cummins,   being
solely a  producer of  engines, may  not  have  the complex corporate
structure of a  General Motors  (GM) or International Harvester

-------
                       -179-
                    Table A-II-3

        Corporate Overhead and Profit as a
Fraction of Cost of Sales for Five Manufacturers 2/
1976
Overhead Profit
GM
IHC
Caterpillar
Cummins
Mack
0.117
0.172
0.121
0.308
0.123
0.145
0.075
0.165
0.169
0.055
1977
Overhead Profit
0.109
0.169
0.113
0.335
0.096
0.141
0.074
0.172
0.150
0.067
1978
Overhead Profit
0.117
0.193
0.109
0.336
0.085
0.129
0.056
0.172
0.115
0.099

-------
                                -180-
(IHC).  Much  of  the division level overhead, which  is  assigned  to
the  cost  of sales by GM  or  IHC,  may be assigned to  Cummins'  cor-
porate level.  Equation (5)  recognizes  that  there will  be  overhead
costs  (and  profits) at lower  than corporate levels  and  increased
costs  by  40 percent at that  level to  account for  these costs.
Thus, it would seem likely that some of these overhead costs,  which
are  included here  at  the  vendor or divisional  level,  are  included
in  Cummins'  corporate  overhead figures.   If this  were the  case,
then it  would  be appropriate to exclude  the Cummins  overhead
figures  from  the analysis.  However, to  be  conservative, the
Cummins  figures  will  be included and  weighed equally  with the
others.

     Given  the moderate  size of the range  of overhead and  profit
figures for these  five  manufacturers   (with  the  exception of the
Cummins overhead  figure),  the mean  of the  corporate overhead and
profit figures of all  five manufacturers should adequately  repre-
sent them all.  Thus, 16.7 percent  and  11.9  percent,  or a  sum of  29
percent,  will be used in equation  (5) as appropriate  allocations  of
corporate overhead and profit,  respectively.

     Turning  finally  to dealer overhead and profit,  EPA sees  no
incremental increase in dealer  or  franchise  overhead  as  a  result  of
these  regulations.   No  additional personnel or  engine  servicing
will  be  necessary.   Most heavy-duty  diesel engines  sold in the
United  States  are  not  sold  through  conventional dealers as are
automobiles and  light-duty  trucks; instead,  they are  sold through
either  dealer  franchises  which specialize  in  trucks  or through
manufacturers'  representatives.  The  individual retail price of  a
diesel truck or  bus  may  exceed $50,000 and multiple  unit  sales  to
city  transit  systems,  inter-city  bus companies, or  large  trucking
companies  are quite common.   Admittedly,  dealers might try  to  get  a
small profit on their increased investment in the engine.   However,
this  profit  should be very  small  given the  very  short period  of
investment  (a  few days)  and fall  within other  possible errors  in
estimating  manufacturing  costs or  corporate  overhead and profit.

     Now that  revised  estimates of  corporate and dealer  overhead
and  profit  have  been  developed,   these revised estimates can  be
substituted into equation (5) to form  a  new  costing  equation.  The
new  factor  for  corporate  overhead and  corporate  profit   is  0.29.
Also, since research and  development costs and tooling expense were
included in the cost of sales upon which these  factors  were  based,
the  two costs (RD and TE) should also  be increased by the  overhead
and profit factors in the  costing  equation.   The resulting  equation
is shown below:

     RPE   =  [(DM + DL  +  OH)(1.4)  + TE  + LBE  + RD + TE](1.29)       (7)

     Now that the revised  retail cost methodology is  available, the
next step will be  to  calculate the cost of  the various components
which together form a trap-oxidizer system.   A  standard production
volume of  300,000 units  will  be used for the  time being.  After the
cost of all the components has  been determined,  the ratios  shown  in

-------
                                 -181-

Table  A-II-2  will  be  used  to  calculate  the fleetwide average
costs.

     The major  portion  of the cost of a  trap-oxidizer  is the trap
itself.   The most promising  trap  designs fall close to  that  of  a
monolithic  catalyst.   In some cases,  actual  monolithic substrates
are  being  used with and without washcoat and noble  metals  for
prototype trap  testing ._3_/  In other cases, the trapping  material is
alumina-coated  steel  wool  or saffil fiber.4/  In  either  case,  the
manufacturing  of a  trap out of  these  materials  should  follow
closely to  that of a monolithic catalyst.  .Since no cost  data
are  available for  the other  trap designs, the cost of a similarly-
sized  monlithic catalyst will be  used  to approximate  the cost  of
the  trap.

     The costs  for four trap volumes will  be calculated,  account-
ing  for the different  sizes which will  be required by  different
engine  sizes.   The basis for the sizes is the successful testing of
a  5.3-liter trap fitted  to  an Opel 2100D, which  has a fuel economy
of 31.5 miles per  gallon ._5_/   Extrapolations  of trap size  were made
to  larger   and  smaller  engines  using the ratios  of the  fuel  con-
sumptions of the various engines (vehicles).  Fuel  consumption is  a
good,  available  indicator  of volumetric  flow  through the  trap,
which  should be one  of  the  main  considerations in  sizing  the trap.
The  average fuel economies  of Class  V and VI, Class VII,  and Class
VIII  for  the late 1980's and early  1990's  will  be taken as  8.3,
'7.2, and  6.7 miles per  gallon,  respectively.^/  A fuel economy of
13.0 miles  per   gallon will  be  used for Classes IIB, III, and  IV.
This  last   fuel  economy  was  determined by  interpolating  the  above
fuel  economies  with  their  respective  gross  vehicle weights  along
with the fuel economy and GVW of a standard  light-duty diesel truck
(20 miles per gallon, 7,500 pounds).^/  The  trap volume  for a Class
IIB, III, and IV vehicle is  then calculated to be  12.8  liters (785
cubic   inches)(5.3  liter x  31.5  mpg/13.0  mpg).    For the   other
vehicle classes,  it  is  assumed that  two  traps will be required for
each vehicle.*   The  trap volumes for  a  single trap for  Class V and
VI,  Class VII,  and Class VIII are  10.0 (612), 11.6 (710), and 12.4
(762)  liters (cubic  inches), respectively.

     Lindgren   (p.  145)  has  determined  the  cost  of a monolithic
catalyst as  a function of volume and noble metal content and  put it
in a formula equivalent  to equation (6):

     RPE (Trap)  =  (NM +  $2.52 +  0.101 x V) x 2.52 + $6.00          (8)

Where:

     RPE (Trap)  =  Retail price equivalent of a trap (monolithic
                   catalyst).
 *     Trap-oxidizers will  be  assumed  to  be  in series  unless  a dual
 exhaust  system  is used.

-------
                                  -182-


             NM = Cost of noble  metals  at  manufacturing  level.

              V = Volume of trap in  cubic  inches.


     The multiplicative factor  of 2.52  in  equation  (8) is the
product of  the factors  for  vendor  overhead  and  profit (1.4) and
corporate and  dealer  overhead  and  profit  (1.8).    In  the 'revised
methodology of equation (7),  the first  factor  remains  the  same, but
the  second factor  becomes  1.29.  Also, the  factor  of  1.29  is
applied to the $6.00 cost of research  and  development and  tooling.
Thus, in terms of the  revised methodology  of equation  (7),  equation
(8) becomes:

RPE(Trap) = ((NM + $2.52 + 0.1013 x  V)  x  1.4 + $6.00)  x  1.29        (9)

     The trap  volumes  needed for equation (9)  are already avail-
able, but the noble  metal loadings are  not. At  this point  in  time,
it  is  not known whether  or not diesel  particulate  traps  will
require noble metals.   The purpose of the  noble  metals,  if  present,
would be to  lower the  temperature necessary to ignite  the trapped
particulate  and  possibly  to aid the  oxidation  process  to   reach
carbon dioxide and water-  To cover  the range  of possibilities, two
loadings will  be assumed, one  with no noble metals  and  one with
oxidation-promoting  metals (Pt and Pd)  at  a level  found in current
oxidation catalysts  for gasoline engines, which  is  around   0.012
gram per  cubic inch with  a  2:1  ratio of Pt to  Pd.   Noble  metal
costs  are  currently around  $10.40  per gram  for  Pt  and $3.71 per
gram for Pd ._8_/*   However,  since the Lindgren costs represent  1977
prices and a  general  inflation  rate of 8  percent  per year will  be
used to adjust these Lindgren costs, these  current  1980  noble metal
costs will be  divided  by 1.26  so that when they are adjusted for
inflation later,  they  will represent current  prices.   Using  Lind-
gren' s formula for the cost of the noble metals  (p.  134):

     NM = $8.26 x 0.008 V + $2.94 x  0.004  V +  $0.14 x  0.0012 V

or

     NM = $0.0780 V                                                 (10)

     The  last term  (0.0012V)  accounts  for  manufacturing costs.

     Equation (9) includes the cost  of a washcoat.  However,  if  no
noble metals  are  to be present, the washcoat  should not be neces-
sary and its cost should be  deleted.   From a  breakdown  of  catalyst
costs at various  volumes (Lindgren  p.  360),  it is found  that the
cost of the washcoat is  proportional to the volume of the  catalyst
and represents 10.3 percent  of  the  0.101   term  in  equation (9),  or
*    Prices stated here  are  77  percent  of  the market  prices  quoted
in the  reference  and represent  prices  available  to  larger  volume
buyers.

-------
                                 -183-


0.010 V.   Subtracting  this  and the noble metal cost from equation
(9) yields  the  cost for a  trap  without  washcoat  or noble metals:

     RPE (Trap)  = (($2.52 + 0.091V) x 1.4 +  6.00) x  1.29           (9A)

     Two  final   adjustments  are  needed before  calculating the
costs of  the  traps.   One,  inflation needs  to be  considered.  The
costs that Lindgren  quotes  are from 1977-  An 8 percent per  annum
inflation rate will be used to convert costs  to 1980 costs.   While
this inflation rate  is below  the Consumer Price Index  (CPI)  infla-
tion rate for 1978 and 1979, it is  actually  above the New Car  Price
Index (NCPI)  for these two years. 9_/   The NCPI should  be a better
indicator of  the inflation rate  to be  used  here  even though the
NCPI may reflect some  lowering of  profits to  sell cars in the last
few years.   However,  an eight  percent  inflation rate  is  still
greater than  the  NCPI  for  1978 and  1979  and  thus should take care
of  any  change in pricing  structure.  For  traps,  Lindgren  quotes
costs for 1977,  and these  costs  must be multiplied by a factor of
1.26.   Two, production volume  needs to be  taken into account.
Lindgren  assumed  a  production  volume of  2,000,000  catalysts (p.
115).   The  production volume of  interest here is  300,000  units.
Using equation (1) with z = 0.12,  it  is  found that the cost should
be  a  factor  of   1.36 higher at the lower production volume.   Com-
bining  the  inflation and production factors,  the costs determined
by  equations (9)  and (9A),  should be  increased  by  a  factor of
1.72.

     The necessary  equations  ((9), (9A), and  (10))  are now avail-
able with which  the cost of  the trap can be  determined.  Substi-
tuting  equation  (10)  into  equation  (9)  and  multiplying equations
(9) and (9A) by  1.72:

Trap cost - No noble metals

     RPE (Trap)  =   (($2.52  -i-  0.0909  V) x 1.4  + 6.00) x 1.29 x 1.72

or

     RPE (Trap)  =  $21.10 + 0.282 V                               (11)

Trap cost - With noble metals

     RPE (Trap)  = (($2.59 + 0.101 + 0.0780 x 1.4 + 6.00)  x
                     1.29 x 1.72

or

     RPE (Trap)  = $21.10 + 0.556 V                                (HA)

     Using  equations  (11) and  (11A) the  costs  of the  traps at
various volumes  can now be calculated.   These are shown in  Table
A-II-4.

     Port  liners,  insulated  exhaust  manifolds  and  an   insulated

-------
                                  -184-
                             Table A-II-4

     Estimated Cost of a Trap-Oxidizer  System (1980 Dollars)
     	(Production Volume  = 300,000)  I/	

                       Class IIB,     Class      Class     Class
Item	    III. IV      V.  VI  2/   VII 2/   VIII 2/

Trap 3/
 Without Catalyst         243           388       443       472
 With Catalyst            458           723       832       890

Port Liners                 20            25         29        36

Stainless Steel             30            39         50        66
 Exhaust Pipe 4_/

Insulated Exhaust           69            99       125       172
 Pipe 5_/

Insulated Exhuast           25            36         46        58
 Manifold

Electronic Control          37            37         37        37
 Unit (50% of Total
   NOx and Part.)

Sensors                     9            999

Throttle Body               16            16         16        16
 Actuator

Electro-Mechanical          6            666
 Control

Muffler (Credit)6/        (44)          (52)      (53)      (58)
Ij   "Cost Estimation  for Emission  Control  Related Components/
Systems and Cost Methodology Description,"  Rath  and Strong for EPA,
March 1978, EPA-460/3-78-002.
2j   Cost is for total of two  traps.
_3_/   Costs are shown for an oxidation  catalyst,  12.8 liters for a
Class IIB, III, and IV vehicle,  10.0 liters for  each of the two
traps for a Class V and VI vehicle, 11.6  liters  for each of the two
traps for a Class VII vehicle,  and  12.4  liters  for each of the two
traps for a Class VIII vehicle.
j4_/   Includes credit for steel  exhaust pipe which it replaces; $14
for a Class IIB, III,  and IV vehicle,  $22 for  a  Class V and VI
vehicle, $27 for a Class VII,  and $35  for a Class VIII vehicle.
_5_/   Includes only cost for insulating an exhaust pipe, cost of
exhaust pipe itself is not included.
6_/   Production volume equal to  in-use production and not at
300,000.

-------
                              -185-
exhaust pipe may  also  be necessary to ensure that  the  exhaust  gas
temperature remains  high enough  to  permit  oxidation in  the  trap.
From Lindgren  (p.  195),  the manufacturer's cost  (vendor  cost  plus
research and development  and  tooling)  of port liners for  a  8-cyl-
inder  light-duty  engine is  $11.30.  Taking inflation (26 per-
cent)  and  corporate  and  dealer overhead (29  percent) into  account
would  increase this  to $18.30.  The production volume  assumed  was
400,000 engines.  Using  equation  (1),  with  z  =  0.12,  to convert to
300,000 units results  in  a  cost  increase of 4.8  percent to $19.20,
or $19.  It will be assumed that material costs  for  port liners  are
proportional to  engine  size.    The  engine  size  for a  8-cylinder
vehicle used in Lindgren's  calculations  was 5.20  liters (318  CID) .
The  final  calculated  costs  for  port  liners  for the  four  engine
sizes,  5.7, 8.2, 10.7, and  13.9  liters,  are shown in Table A-II-4.

     The  cost  of  an  insulated  exhaust manifold  has also been
indirectly  determined  by Lindgren  (pp.  171-90).    From Lindgren1s
treatment  of a  thermal reactor,  the cost of  simply  insulating  the
manifold can  be determined.   For a 8-cylinder light-duty  engine,
the manufacturer's cost  of  ceramic  liners and insulation  is  $13.10
(p. 179).   Research  and  development cost of $1.00 per manifold  (p.
180) will  be  assumed  to be  entirely due  to  the thermal  reactor
function and  will  be  assumed  to be  zero  for simply insulating  a
manifold.   Vehicle  assembly  and  engine modifications amount  to
$0.69  for  the  entire  thermal  reactor  (p.  180).   Subtracting  from
this  the  cost  of assembling a  standard  manifold  ($0.56 for  a
6-cylinder  engine,  p.  188)  results  in a negligible  net cost  and
will not  be considered.   It will  be  assumed  that the  cost of  the
manifold itself will not  change.   The  cost  of insulating  should be
multiplied  by  1.29  (see  equation  (7)  to obtain  the retail  price
equivalent, which is $16.90.  The production volume  assumed  was  the
same as in the case  of port  liners  above, or  400,000 units.   Thus,
the conversion  factor  for  inflation  and  production volume is  the
same as  above,  1.32 (1.26  x  1.05).   Taking  this factor into  ac-
count,  the cost  of  insulating an  8-cylinder  manifold in a  light-
duty vehicle  (engine size = 5.2  liter)  in  1979  is  then  $22.30  or
$22.   As  with the  port  liners, the  costs of  insulated exhaust
manifolds  for the larger heavy-duty engines  have been calculated by
taking the ratio of material costs to  engine size.  These  costs  are
shown  in Table A-II-4.

     Looking next  at  the exhaust pipe,  there  are  two levels  at
which  it can be improved.  One, the standard steel material must be
converted  to stainless  steel if the system will  be expected  to  last
the entire life  of a heavy-duty  diesel  vehicle.  There is  no
guarantee that people would  replace a rusted-out  exhaust pipe
before it  developed  holes, which would allow  exhaust  to  bypass  the
trap and also cool the  exhaust, possibly  to  the  point  of preventing
any oxidation from ocurring.   Two,  the exhaust pipe may have  to be
insulated  to keep the exhaust temperature high enough  for  oxidation
to occur.

     The cost of  changing the  exhaust  pipe to stainless  steel  can

-------
                                -186-


be taken from Lindgren.   Lindgren  performed a cost analysis for two
types of exhaust systems, the first system attaching to the single
exhaust manifold of a 6-cylinder  (3.7 liter)  light-duty engine (p.
255), and the second system for  a  V-8  (5.2 liter) light-duty engine
(p.  256).   In  the  case  of heavy-duty diesel  engines, about three-
fourths of the systems  are of the  single, non-branching variety and
one-fourth  are  dual exhaust  systems, where  two entirely separate
exhaust systems are  used.   (The source and  details of  this is
contained  in  Section  B.)   This breakdown  is  assumed  to  apply to
each  vehicle  class as  well  as  the entire  fleet.    From  this, it
would appear that  Lindgren's  cost analysis  for  the 6-cylinder
engine would be most analagous  to that of heavy-duty diesels.  For
those heavy-duty diesels, the cost of  a single  exhaust pipe will be
calculated.   For  those  heavy-duty  diesels with dual-exhaust  sys-
tems, the  cost  will be  doubled  (i.e., two  exhaust  pipes assumed).

     The manufacturing  cost  (DM +  DL +  OH  in  equation  (6))  of a
standard steel exhaust  pipe  is $3.27 (6-cylinder light-duty engine,
3.7  liter) (p.  254).  Tooling costs  are only $0.10 per pipe.  Using
equation (7),  the  retail price  equivalent  of  this  pipe  is $5.94.
The  retail  price equivalent  of  a stainless steel exhaust  pipe is
$15.91  (p. 247  and equation (7)).  The  cost of  converting to
stainless steel is  then $9.07 for a 5.2 liter engine.  The assumed
production in both  cases was 1,000,000.   Using equation (1),  these
costs need  to  be  increased  by  21.8 percent  to convert  to  a pro-
duction of  300,000 units.   They  also need  to  be increased  by 26
percent because  of inflation.  In  total,  then, the  cost  of con-
verting the exhaust pipe  to  stainless steel  is $14 for an 6-cylin-
der  light-duty  engine.    Again,  the  costs for heavy-duty  engines
will  be calculated by prorating material  costs to engine displace-
ment with 75 percent of  the  heavy-duty diesel engines requiring one
exhaust pipe and 25 percent  of heavy-duty diesel engines requiring
two  exhaust  pipes.   These costs  are  shown in  Table A-II-4.

     The cost  of   adding  a   double  wall  to  the exhaust  pipe with
insulation in between is  next to  be determined.   Again from Lind-
gren  (p. 272  and  equation (7)),  the  retail price equivalent of a
double-walled,   stainless  steel,  insulated  pipe is  $38.40  for  a
6-cylinder engine.    Subtracting the  costs of  the stainless  steel
pipe  calculated above  leaves $23.30.   Using  the  same adjustments
for  production volume  and  inflation,  and the same assumptions
concerning engine  size  and  single exhaust-dual exhaust breakdown,
the cost of converting a  stainless steel  exhaust pipe to a double-
walled,  insulated,  stainless  steel  pipe  is shown  in Table  A-II-4.

     The cost  of the oxidation control unit  will include costs for
sensors, thermocouples,  and a throttle for raising the temperature
of the exhaust.  The  estimates  shown  in  Table  A-II-4 are based on
the following.   In his  study, Lindgren solicited  estimates  of the
cost  of an electronic  control unit (ECU)  which  monitored  and
controlled  a  large number  of  sensors  and  controllers  (p.  320).
This  type of ECU should  be  of the same  capacity as  that needed to
control  the oxidation  process of  a  trap-oxidizer system.   The

-------
                                -187-
industry estimate was $45.  Taking this to be a vendor level cost,
the retail  price  equivalent would  be  $57.   Inflating this to 1980
prices, the cost increases to  $73.  However, half of  this cost will
be  alotted  to particulate control and  half  to NOx  control.   The
presence  of the  electronic  control  unit will  allow the  use  of
programmed NOx control systems (e.g.,  timing,  exhaust gas recircu-
lation)  which should provide reductions in NOx  emissions  from
heavy-duty diesels. Thus,  the  cost  of  the unit  due to  diesel
particulate regulations  is $37,  which  is  shown in  Table A-II-4.

     The costs of the sensors, throttle body  and actuator can also
be  taken from  the same Lindgren  table (p.  320).   Allowing for two
thermocouples near  the  trap,  an engine  speed  sensor,  and  a  rack
position sensor,  the vendor cost  at a production volume of 300,000
is  approximately $5.  With three year's inflation,  this cost would
increase to  around  $6.    If equation  (7) is  used  to  calculate the
retail price  equivalent, the  cost  becomes  $9.   The  throttle
switch and body should cost about $10 at the vendor  level (p. 320)
at  a  production  volume  of  300,000  units.   With  inflation and
conversion  to  retail price  equivalents,  the  cost  should  be  $16.
Both costs are shown in Table  A-II-4.

     It may  also  be  possible  that  a  much simpler  control  device
would  suffice in the situation.  If all that was  needed was  a
periodic boost  in exhaust temperature during some  general  engine
condition,  then a controller on the order of  an  automatic choke  or
an  odometer-controlled maintenance light (e.g., EGR  light)  should
be  satisfactory.  For example, if  the throttle  actuator  was keyed
to  the odometer  and rack position,  it could  operate  periodically,
for a  set period  of  time  at a certain rack position.   This type  of
control system would only  require two  or three  sensors  and mechan-
ical  or  electrical  connections  to the  throttle  actuator.    From
sensor  costs  shown  by Lindgren  (p.   320)  and  equation  (7),  this
system should only  cost  about  $16.  This  option has  been included
among  the components shown in  Table A-II-4.

     It is  also  very likely  that  the addition of  a trap  to the
exhaust system would allow the muffler to be  deleted.10/,ll/   This
would  result in a savings  to the consumer,  not  only  initially, but
every  time  the standard   steel exhaust system would  need  replace-
ment.   The reduction in  initial  vehicle sticker  price will  be
examined here, while the  reduction  in  vehicle  operating costs  will
be  examined later in Section  B of  Chapter  VI,  "Costs to  the Users
of Heavy-Duty Diesels."

     Lindgren only  estimated  the  cost  of  mufflers  for  passenger
cars.   Due to this and the fact that  aftermarket muffler costs are
available  for heavy-duty diesels,   it would  appear  to be  more
accurate to  convert  these heavy-duty  aftermarket costs  to  retail
price  equivalents  than  to extrapolate  the light-duty costs  to
heavy-duty.    A  survey of  heavy-duty  diesel dealerships  has  shown
that the cost  of a typical replacement muffler is about $136 (Class
IIB-IV vehicle),  $161 (Class  V and  VI),  $164  (Class  VII),  and $181

-------
                                -188-
(Class VIII).  These aftermarket costs were  taken from an analysis
of  the  replacement costs  of mufflers on heavy-duty  diesels,  the
details of which can be  found in Section  VI-B.   Lindgren estimated
that the aftermarket cost is about  four times the vendor cost_l/ and
this has  been  confirmed  for light-duty exhaust  systems.12J   Thus,
the vendor cost  for a  muffler for a Class IIB-IV vehicle would be
about $34, for a Class V and  VI vehicle  about $40, for a Class VII
vehicle about  $41,  and  for a Class VIII  vehicle  about $45.   Using
equation  (7),  the  retail price  equivalent would  be  $44,  $52,  $53,
and $58  for  a Class IIB-IV,  Class V  and VI, Class  VII,  and Class
VII vehicle,  respectively.   This  would  be  the  savings  resulting
from eliminating the need  for a muffler   on  these vehicles.   These
savings are  shown  in Table A-II-4  and should be  deducted from the
cost  of  trap-oxidizer  systems  calculated  below.   These  muffler
savings are based  on existing in-use production  volumes  and should
not be multiplied by the cost ratios  in Table A-II-2.

     Now that  the  cost  of  all the components has been determined,
the decision needs  to be made concerning  which of these  components
will be  needed  on any given vehicle.   As  this  is inherently a
projection, there  will be  a number of  component  combinations which
may be able to reduce particulate  emissions to required levels, but
it is also possible that they may not.   There will  also  be varia-
tions  between models and manufacturers, as is usually the case with
a system as complex as  a trap-oxidizer.

     Four  basic  combinations  appear  to  have varying degrees  of
probability  in  being  able to trap  and oxidize  diesel particulate
safely and efficiently.  These are shown  in  Table A-II-5.  At this
time,  it does not  appear likely that a simple trap will  be able to
perform adequately by  itself.   Some  additional  features will  be
necessary to  ensure that  the particulate will  be  oxidized  effec-
tively and safely.   Systems  I,  II,  and III  all  include  one  or two
such features.   System I  includes a  trap plus  exhaust  insulating
features to help retain  exhaust temperature  and  promote  oxidation.
It also includes a  throttle to raise  exhaust  temperature controlled
by  an  electro-mechanical  system.    This  control  system would  be
envisioned to be much simpler than  that  for a three-way catalyst or
electronic fuel injection.   The  control system would be more on the
order  of  an  automatic  choke or an odometer-controlled maintenance
light  (e.g.,  EGR).  The insulation  of  the  exhaust pipe  has  been
omitted primarily  because  of  its  cost,   which  is  $69-172.    Road
tests  on a light-duty vehicle (Mercedes-Benz  300D)  have  shown that
the temperature  drop  between the  exhaust  manifold  and  the  trap
inlet  is  only  15-20°C  with an uninsulated   exhaust  pipe.13/   It
would  seem that  this small decrease  in temperature  can  be~made up
elsewhere more  economically;  using,  for  example,  a  throttle.
Actually,  the omission of  an  insulated exhaust pipe  was  one of the
prime  reasons for including a throttle  in this system.

     System II  consists  of a trap, a  throttle and  simple  control
system,  but instead of  insulating  features it will use a coating of
noble  metals  to  promote  oxidation.  System III consists  of a trap,

-------
                                      -189-
System
  I
Trap
(no noble metals)

Stainless Steel
Exhaust Pipe

Port Liners
Insulated Exhaust
Manifold

Throttle Body
and Switch

Mechanical Control
                                     Table A-II-5

                                 Components Included in
                            Potential Trap-Oxidizer Systems
     System
       II
     System
       III
Trap
Trap
(w/noble metals)   (no noble metals)
Stainless Steel
Exhaust Pipe

Throttle Body
and Switch

Mechanical
Control
Stainless Steel
Exhaust Pipe

Electronic
Control Unit

Sensors
                   Throttle Body
                   and Switch
     System
       IV
Trap
(no noble metals)

Stainless Steel
Exhaust Pipe

Port Liners
Insulated Manifold
Exhaust

Sensors
                                       Electronic
                                       Control Unit
                                                            Throttle Body
                                                            and Switch

-------
                              -190-
a  throttle  and a sophisticated control  system,  but uses  no  insu-
lating techniques or catalytic materials.

     Any one or  all three of  these  systems  may be able to trap and
oxidize  diesel particulate  successfully.   However, there  is  some
chance that  a  more  advanced system will be needed,  which leads to
System IV.   System  IV  combines  the  oxidation-promoting features of
Systems  I  and III, consisting  of  a  trap,  throttle,  port  liners,
insulated exhaust manifold  and  sophisticated  control.   This system
should be sufficient in any case, and  represents  an upper bound of
necessary technology.

     The  costs  of  the four  systems  are  shown  in  Table A-II-6.
System I and  III  are the  least  expensive, which  is  to  be expected.
However,  System  II, which  could  be considered  less likely to be
viable than  System  IV,  is more expensive than System  IV.   This is
primarily due  to  three  assumptions  used to estimate the  amount of
catalytic material on  the trap.   One,  it was assumed that Pt and Pd
would be the catalysts  used.  Two, it  was assumed that  the catalyst
loading would  be  that  found on  current oxidation  catalysts,  around
0.012 gram per cubic inch.   Three, it  was assumed that  this loading
would be  needed   throughout  the whole trap.   It is possible  that
expensive catalysts  such  as  Pt  and  Pd may be  avoided and more
inexpensive catalysts, such  as  silver nitrate, may prove  suf-
ficient.  It  is  also  possible that  the  loading  could  be decreased
or that  the  catalyst would  only be needed  near  the  inlet to begin
the  oxidation  process, which  would proceed  thermally  thereafter.
Any  of these  changes would  lower  the  costs of System  II and could
make it competitive with Systems I and III.

     It is not possible to place any probability  on  the possibility
of any  of these  systems  being used.   It  is quite possible  that
System I will  be  used on some models,  particularly  those which may
be relatively  close to  the  1986  standard without a trap-oxidizer.
It is also possible that  some models  will  need  System  IV.   Rather
than  give  the  systems  a probability weighting  which would  have
little basis,  the entire range  of costs  between  Systems III and IV
will  be  used  hereafter,  as  it  does   indicate the  range of costs
which could occur.  The cost  of System IV will be  taken to be the
maximum cost.  It will be assumed that System II will  be used  only
if the catalytic material or  its  loading can  be  changed to make it
economically competitive with  Systems  I and  III.

     The range of system costs  (III-IV) of  Table  A-II-6 can now be
combined with  the actual cost  to  reference  cost ratios of Table
A-II-2 to  yield  the actual  cost  of  trap-oxidizer  systems at  the
production volumes  expected.    The  costs for  the traps  should be
multiplied by  the ratios corresponding to each vehicle  group shown
for  traps in  Table  A-II-2.   The costs  for  the port  liners,  stain-
less  steel exhaust  pipe, insulated  exhaust  pipe,  insulated  exhaust
manifold, and  the throttle  body and actuator  should be multiplied
by the  ratios corresponding  to components  that  vary   with  engine
designs.    The costs for  the  control  units and  sensors  should be

-------
                                  -191-





                                Table A-II-6




      Cost of Four Potential Trap-Oxidizer Systems (1980 dollars) 1/
Vehicle Class
System
I
II
III
IV
IIB, III & IV
296
466
292
336
V, VI
' 458
732
437
501
VII
541
851
502
580
VIII
596
920
542
636
I/   Production Volume = 300,000.

-------
                              -192-
multiplied by  the  ratios  shown in Table A-II-2  for  the electronic
control units.  The results are shown in Table A-II-7.

     A  closer  look at  Table  A-II-7 shows  that  a Class  VII trap-
oxidizer  system costs more  than  a Class VIII trap-oxidizer system,
despite  the  fact  that  a  Class  VII trap-oxidizer  is  smaller  and
requires  less  material to  manufacture.   The costs  of  Class  VII
trap-oxidizer  Systems  III and IV (as   shown  in  Table  A-II-6)  are
$43-59  less  than  the  Class VIII   systems, at  a constant production
volume  of 300,000  for each  group  of  vehicle  classes.   However,  the
lower sales  of Class  VII  vehicles result in  a higher overall cost
of manufacturing Class VII trap-oxidizer systems.  With the exhaust
flow of a Class VII vehicle being  less than that of  a Class VIII
vehicle there  is no reason  that  a Class VIII trap cannot be placed
on  a  Class  VII vehicle.   This would reduce the  cost  of  the Class
VII trap  to  that  of the  Class VIII  trap and also reduce the cost
of  the  Class  VIII  trap by increasing the production  volumes.  For
example, in  1986 the cost ratio  for  both classes combined would be
1.15 for the traps (over  a  production of 300,000)  and the new cost
would be $543.  This is less than the  $673  and $552  costs of traps
for Class VII  and Class  VIII  vehicles, respectively.   This  would
also lower the trap-oxidizer system  cost to  $652-805  for  Class  VII
vehicles and  $642-789  for  Class VIII vehicles.

     Further  analysis  shows that   the same holds  true  for  the Class
V-VI systems.  Overall  it is  less expensive  to  fit Class V and VI
vehicles with  Class VIII  traps than to manufacture traps specifi-
cally  sized  for  the  Class V and  VI  vehicles.  For example,  a
Class V and VI system in  1986  costs  $528, using  the cost ratio  for
traps in Table A-II-2  and  the  cost of traps  in Table A-II-4.  Using
the 1986 cost  estimate  of $543 calculated  above  for  Class VII  and
VIII traps,  a sales weighted average (again,  based on  Table A-II-1)
of  Class  V-VIII  traps  is  $539.    If Class   V and VI  vehicles  are
fitted  with  Class VIII  traps, the  cost  ratio  in  1986  for  traps
would be  1.09  (as shown  in Table A-II-8),  and  the  cost  would be
$514.   This  is less than  the  sales-weighted average  of  $539 cal-
culated above.  This also has  the effect of  lowering  trap-oxidizer
system costs  to $611-688  for Class V and VI  vehicles,  $652-805  for
Class VII vehicles,  and  $642-789 for   Class  VIII  vehicles.   This
reduction in cost  holds true  for every year.  New cost ratios  for
traps  (compared  to  an accumulated  production  of  300,000  units)
grouping  Classes  V-VIII  traps together are  shown  in  Table  VI-8.
The revised  cost  estimates of trap-oxidizer systems   are  shown in
Table A-II-9.   The  fleet-average cost  for   each year is  again  a
sales-weighted average  (based   on  the sales  scenerio  in  Table
A-II-1) of costs for the  four basic  vehicle  groups.   The fleetwide
average cost  in 1986 would then be $629-$756  and  should decrease to
$458-$559 in  1989.

B.   Savings  Due to Maintenance Reductions

     The  addition  of  a  trap-oxidizer   system  is also  expected  to
reduce  maintenance  in  two  ways.   One, the  system  will  include  a

-------
                         -193-
                       Table A-II-7

          Estimated' Costs of Trap-Oxidizer Systems
       At Predicted Production Volumes (1980 dollars)
1986
Sales Weighted

1987
Sales Weighted:

1988
Sales Weighted:

1989
Sales Weighted:

1990
Vehicle Class

IIB, III, IV
    V, VI
     VII
    VIII
IIB, III, IV
   V, VI
    VII
   VIII
IIB, III, IV
   V, VI
    VII
   VIII
IIB, III, IV
   V, VI
    VII
   VIII
IIB, III, IV
   V, VI
    VII
   VIII
Vehicle Class Average

       551-644
       626-730
       811-964
       679-826

       672-806

       480-562
       551-645
       714-850
       601-733

       592-711

       438-513
       507-593
       658-785
       558-681

       547-657

       410-480
       477-559
       624-744
       528-645

       515-619

       384-451
       449-531
       598-714
       505-618
Sales Weighted:
                         489-590

-------
                               -194-
                          Table  A-II-8

       Revised  Values  for  the  Ratio  of the  Actual  Cost  of
      a Component  to Its Cost  at  an  Accumulated  Production
      	Of 300,000  Units I/	

                              1986    1987    1988   1989   1990
 Electronic  Control  Units      1.21    1.07    1.00    0.95    0.91

 Cost Ratios if  Component
 Production  Equals Vehicle
 Group Production

    Class  IIB, III,  IV         1.83    1.61    1.48    1.39    1.32

    Class  V,  VI                1.53    1.36    1.26    1.19    1.13

    Class  VII                 1.71    1.52    1.42    1.34    1.29

    Class  VIII                 1.32    1.18    1.10    1.05    1.01

 Traps_2/

    Class  IIB, III,  IV         1.83    1.61    1.48    1.39    1.32

    Class  V-VIII               1.09    0.97    0.90    0.86    0.83

 Components  that Vary
 with  Engine  Design

    Class  IIB, III,  IV         2.06    1.81    1.66    1.56    1.48

    Class  V,  VI                1.72    1.53    1.42    1.34    1.27

    Class  VII                  2.04    1.81    1.69    1.60    1.54

    Class  VIII                 1.57    1.40    1.31    1.25    1.20
JY   Assumes each supplier shares one-third of the market at
production shown in Table A-II-1.

2J   Class V-VIII production combined.

-------
                         -195-
                       Table A-II-9

      Revised Estimated Costs of Trap-Oxidizer Systems
      At Predicted Production Volumes (1980 dollars) I/
1986
Sales Weighted

1987
Sales Weighted

1988
Sales Weighted;

1989
Sales Weighted

1990
Vehicle Class

IIB, III, IV
    V, VI
     VII
    VIII
IIB, III, IV
   V, VI
    VII
   VIII
IIB, III, IV
   V, VI
    VII
   VIII
IIB, III, IV
   V, VI
    VII
   VIII
IIB, III, IV
   V, VI
    VII
   VIII
Sales Weighted
Vehicle Class Average

       551-644
       611-688
       652-805
       642-789

       629-762

       480-562
       540-632
       574-709
       564-695

       552-670

       438-513
       497-584
       530-649
       520-643

       508-618

       410-480
       472-553
       502-622
       495-612

       482-586

       384-451
       449-527
       480-596
       472-588

       458-559
J_/   Assumes Class VIII traps are fitted to Class V-VIII
vehicles.

-------
                                -196-

stainless steel  exhaust  pipe  which  will eliminate the normal  need
to replace it.  Two, the presence of  the  trap  itself  should  elimi-
nate  the  need for  the muf f ler ,_10_/J.J_/J_2_/  which in turn  eliminates
the need to replace the muffler.

     In  order  to calculate  the savings  resulting from  the  elimi-
nation  of these  two  maintenance  items,  two  pieces  of data are
needed  for  both  muffler  and  exhaust pipe replacements;  timing and
costs.  These two items will  be  examined  below.

     EPA  has performed  a  statistical  analysis   to  determine the
number  and timing  of  muffler  replacements  that normally occur
during  the  lifetime of  a  heavy-duty vehicle.  Muffler  failure
probability as a  function of service time was  obtained from  an SAE
report ,_14/ and this is shown  in Table A-II-10.   It is likely  that
this stucfy only included light-duty vehicles.  However,  it was the
only  study  available  examining  the lifetime  of  exhaust  systems.
The heavy-duty vehicle  scrappage rate  as  a  function of  service  time
was obtained  from an EPA  study._1_5_/   This relationship is  shown  in
Table A-II-11.

     A Monte  Carlo  technique was used  to couple  muffler life with
vehicle  life.   In  this  analysis  a muffler  life  and  a  heavy-duty
vehicle life were randomly chosen according to  their probability  of
occurance.16/   It  was  assumed that  muffler  replacement  was un-
economical  one-half  year before  the truck  life ended.   If the
muffler  life  was equal  to  or  greater  than  the  truck life minus
one-half year, then the truck was assumed to use  only  one  muffler.
If the  muffler  life was  less  than the  truck  life  minus  one-half
year then another muffler  life was randomly chosen until  the  sum  of
the muffler lives  for  that truck  was equal  to  or greater than
the truck  life  minus  one-half  year.   The number of  mufflers re-
quired by 90 random vehicles  was determined.    It  was found that  an
average  of  1.27  muffler replacements  were  necessary  for each
heavy-duty vehicle.   More  specifically, 66  percent of the heavy-
duty vehicles required  at  least  one muffler replacement  which
occurred on the  average after  5  years,  38 percent  required  at  least
two mufflers, the second replacement  occurring after  10 years,  19
percent  require   at  least  three mufflers,  the  third  replacement
occurring after  15 years, 4 percent  required  at least  four  replace-
ments,  the fourth occurring  after  17 years, and 1 percent  required
five mufflers,  the  fifth  replacement  occurring  after  20 years.
These figures were  used  to determine  the discounted  cost of the
muffler savings  in the year  the vehicle was purchased.   Using a  10
percent discount  rate,  these  replacement  rates  are  equivalent  to
0.61  replacement at  the time  of  vehicle purchase, or  to 0.98
replacements  when  the  vehicle  is  five  years  old.  As no similar
data could  be  found  which  related specifically to  exhaust  pipe
replacements, these  findings will also be used  for exhaust  pipes  as
well as mufflers.

     A  survey  of diesel  equipment suppliers  has shown  that the
average cost  of  a replacement  muffler is  $136 for  a  Class  IIB-IV

-------
                      -197-

                  Table A-II-10

          Variation of Aluminized Steel
Mu f fler Failure Probability With Service-Time 13/
Years
1.5
2.5
3.5
4.5
5.5
6.5
7.5
8.5
9.5
Percent Failure
Per Year
2
8
17
18
17
11
9
10
8
Cumulative Percent
Failure
2
10
27
45
62
73
82
92
100

-------
 Years

  1.5
  2.5
  3.5
  4.5
  5.5
  6.5
  7.5
  8.5
  9.5
 10.5
 11.5
 12.5
 13.5
 14.5
 15.5
 16.5
 17.5
 18.5
 19.5
 20.5
 21.5
 22.5
 23.5
 24.5
 25.5
26.5
                            -198-
                   Table A-II-11

             Truck Life Scrappage Rates
         As a Function of Service Time 14/
Percent Scrapped
Per Year
9
7
7
6
7
6
5
5
5
4
4
4
3
4
3
2
3
2
3
2
2
2
2
1
1
1
Cumulative Percent
Failure
9
16
23
29
36
42
47
52
57
61
65
69
72
76
79
81
84
86
89
91
93
95
97
98
99
100

-------
                                 -199-
vehicle,  $161  for  a Class V and  VI  vehicle,  $164 for a Class VII
vehicle,  and  $181  for a Class  VIII  vehicle. 17/   These costs have
already been used in Section A of  Chapter VI  to estimate the retail
price equivalents  of  mufflers  for each  of  the  four basic vehicle
groups.  From  this survey  it was  found that muffler costs are
affected by  two major  factors.    First, the costs depended on
the  physical  muffler  requirements  of  each  engine.   Second, the
costs were related  to  services  rendered by the dealer selling the
muffler.   A dealer providing  services  such  as installation  would
have  a  higher material cost  for these  engine  parts  (including
mufflers) than would  a dealer  who only sells the parts.  From the
range of  costs discovered by the  survey,  a single average cost has
been estimated which should  be a representative aftermarket muffler
cost for each group of vehicle classes.

     For  example,  the  cheapest muffler found for  a Class  IIB-IV
heavy-duty diesel engine with a single exhaust system costs $47 17/
from a dealer who did  not perform any  services.   This same muffler
costs $70  if  sold  from a dealer  that did have services available.
Assuming  that the  average of these  costs would best represent the
actual  prices  paid in-use,  the  typical  cost  would then be  $58,
which will be  used  as  the minimum cost.  The maximum muffler cost
for a Class  IIB-IV heavy-duty  diesel engine  with a  single exhaust
system  was  $146 from  a dealer with  no services, or  $219  from  a
dealer providing services.   The average of these two costs,   $182,
will  again be  used as  the typical  cost.   In  the absence  of  a
breakdown  of  sales by vehicle type  (which  is very difficult  to
obtain)  the average  of the  least  expensive  and most expensive
muffler,  in this  case $120,  will  be  assumed  to be the average
muffler  cost for  a Class IIB-IV vehicle with  a single exhaust
system.   Similarly, the minimum least expensive muffler for a  Class
IIB-IV engine with dual exhaust costs $108-$164,  (average of $136),
while the most  expensive muffler  costs  $188-$282,  (average of
$235).  The average of  these costs is  $185.  As mentioned earlier,
3/4 of heavy-duty  diesel  engines  are expected  to require  a  single
exhaust   system  and 1/4  of  all heavy-duty diesel  engines are ex-
pected to  require  a dual exhaust system.  With  this in mind, the
aftermarket cost of a  Class  IIB-IV muffler would then be $136  ((3/4
x  120) +  (1/4 x 185)).  Applying the same method as above,  after-
market muffler  costs  of $161,  $164,  and  $181 were estimated for
Classes  V and VI, Class  VII, and  Class  VIII engines, respectively.
These costs will be assumed  to  occur  at  in-use production volumes.

     For  the  light-duty market,  the  cost  of  labor  and incidental
parts amount  to 25 percent  of  the  cost  of  the  muffler.JJ!/   This
also holds  for  the heavy-duty  situation,  and  these muffler  costs
need to be increased by 25  percent to represent the  total cost of a
muffler replacement.   These  total costs are  $170, $201,  $205, and
$226, respectively.   Using  the sales figures of  Table  VI-1, the
sales-weighted average of these costs is  $211 per muffler replace-
ment.  Undiscounted, 1.27 muffler  replacements would  amount to $268
per vehicle.   Using the actual schedule of replacements described
above and a  10  percent  discount rate, the  savings from eliminating

-------
                                -200-
muffler replacements would be $129 per vehicle, discounted back  to
the year of vehicle purchase.

     The cost of  steel  exhaust  pipes  has  already been examined  in
Appendix II-A.  In that section, it was assumed that  3/4  of heavy-
duty diesel engines required one non-branching exhaust pipe.   This
fraction corresponds to the fraction of  turbocharged diesel engines
sold by the five  largest manufacturers,  as discussed  in Chapter  3,
the Description of the Industry.  The remaining 1/4 of the engines
(those being naturally-aspirated) are assumed to have dual exhaust
systems.  While a cross-over  pipe  could be used instead  of a  dual
exhaust  system,  this  is highly unlikely  for  heavy-duty diesel
engines due  to  their  large  engine  size.    Thus,  these naturally-
aspirated engines  are assumed to require  two  exhaust  pipes.    This
breakdown of single and dual  exhausts is  assumed  to apply to  each
vehicle  class. Using  the data  presented  in Appendix  II-A, the
retail price equivalents for the four vehicle groups are  $14  (IIB,
III, IV),  $22  (V, VI), $27  (VII), and  $35  (VIII).   From equation
(7) of Section VI-A, the vendor  level costs can be  found  by divid-
ing these  costs  by 1.29.   From Lindgrenl/  aftermarket   costs are
four  times vendor level  costs and  this relationship  has  been
confirmed for light-duty  exhaust  system components .J_2_/    Adding  25
percent for labor and  incidental parts,  the total  cost of exhaust
pipe replacements  becomes  $54,  $85,  $105,  and $136,  respectively.
Again   using  the  sales  figures  of Table  A-II-1,  a sales-weighted
average cost is  $111  per  replacement.

     Using 1.27 replacements per vehicle, the undiscounted savings
become  $141  per  vehicle.   Using the actual  replacement schedule
determined for mufflers above and a  10  percent  discount   rate, the
savings resulting  from  eliminating exhaust pipe replacements is $68
per vehicle  (discounted  to year of vehicle purchase).   Adding  to
this  the  savings  determined  for mufflers above, the  total  main-
tenance savings is $197 per vehicle (discounted to year of vehicle
purchase).

C.   Sensitivity Analyses

     Two assumptions were made  in Section A to  aid  in determining
the control system costs  associated with particulate control.   One,
it was assumed  that a  12  percent learning curve applied.  Two,  it
was assumed  that  the  trap-oxidizer  systems would  be  supplied  by
three   outside suppliers.   The effect of  these  assumptions  on the
final  costs will be examined here by recalculating the costs based
on  two new sets  of assumptions.   One,  rather  than using  a  12
percent learning  curve,  it  will be assumed  that  a  learning  curve
does not apply  and  that  the  costs are  the  same  at  all production
volumes.   The  number  of  outside  suppliers  and/or manufacturers
providing  their  own systems has no  effect  in  this case.   Two,
rather than assuming that three outside  suppliers will produce all
of  the  control systems,  the assumption will  be  made  that  each
manufacturer will   produce  his own control  systems.   Here  the  12
percent learning curve  will be used again  to convert the changes  in
production to changes  in  costs.

-------
                               -201-
      1.   Learning Curve

      To  remove  the  effect  of the  learning curve is actually a
fairly simple  process.   All that needs  to  be done is to return to
Section A and retrieve the costs determined by Lindgren,!/ adjusted
for component  size and  inflation.   These costs will be "the same as
those  shown  in Table A-II-4,  except that  they will  be divided by
the  learning curve  factor  which  was  used  to adjust  the  costs to
300,000 units in the first place.    These component costs, assuming
no learning  curve, are  shown in Table A-II-12,  with the production
volumes originally used by Lindgren.

      As production does not  affect costs in this case, the costs of
the various  components  of Systems  III and IV  (See  Section  A)  can
simply be added  up and  will  apply  to the five years of production,
1986-1990.   These system  costs  for  each  of the four vehicle groups
are  shown  in  Table  A-II-13,  along  with  the costs  determined  in
Section A, which used a  learning curve.   As can be seen,  the costs
are markedly less  without a learning curve;  37-62  percent  in 1986
and 15-40 percent in  1990, depending on  the vehicle class.   Sales-
weighting  all classes,  the  costs   determined  without a  learning
curve are 41 percent  less in 1986  and 20  percent  less in 1990,  or
roughly 30 percent less over the five years,  1986-1990.

     This  difference is  quite  significant  and  occurs  primarily
because the  original production volumes  used  by Lindren were quite
large compared to  the production volumes  projected  for heavy-duty
trap-oxidizer system components.  Due to  these large differences in
production volume,  one would  expect  some  difference in costs  to
occur.  Thus, at least  some  of  the  cost  differential  is certain to
occur.  However,  it is  possible that the learning  curve may not be
as steep as  12  percent  and  that the costs  determined  in  Section A
are overestimated by  something less  than  30  percent.  Due  to  the
lack of more detailed information on the  actual learning curves  for
these types  of components, the more conservative  costs of  Section A
will  be  used with the  knowledge  that  the costs  could  decrease
significantly  as additional  information  becomes  available  after
proposal.

     2.    Number  of Suppliers

     The  evaluation  of  the  effect  of the  number  of  suppliers  on
cost  will  require  returning to  Section  A  of  this  chapter  and
slightly  modifying the  equations used to modify costs  per  produc-
tion  changes.  As  described  in Section A,  Equations  (3A)  and (3B)
give   the ratio  of the  actual  cost  to the  cost  at  some  reference
production  for components which differ  between  vehicle  group  and
those which   don't,  respectively.  In these  two equations,  a  factor
of three is  used  in  the denominator  to  split  the  total production
equally among  three  suppliers.   This factor  will  require  modifi-
cation to  reflect  the  assumption  that  each  manufacturer will  be
producing his  own  trap-oxidizers.    The  actual  adjustment will  be
to remove the factor of  three  from  the denominator  (representing

-------
                               -202-

                        Table A-II-12

                  Emission  Control  System
        Component Costs Assuming No Learning  Curve
Lindgren ' s
Item Production Class IIB-IV Class V,VI Class VII
Trap*
Port Liners
SS Exhaust
Pipe
Insulated
Exhaust Pipe
Insulated
Exhaust
Manifold
Electronic
Control Unit
Sensors
Throttle
Body Actuator
Electro-
Mechanical
Control
Mufflers
2,000,000 179 285 326
400,000 19 24 28
1,000,000 25 32 41
1,000,000 57 81 103
400,000 24 34 44
200,000- 37 37 37
500,000
200,000- 999
500,000
200,000- 16 16 16
500,000
200,000 666
500,000-
(44) (52) (53)
                                                            347

                                                             34

                                                             54


                                                            141


                                                             55



                                                             37


                                                              9


                                                             16
                                                            (58)


Without noble metals.

-------
                -203-
             Table A-II-13

    Trap-Oxidizer System Costs Both
With and Without Use of a Learning Curve

Vehicle Classes
IIB-IV
V-VI
VII
VIII
All
With 12
Learning
1986
551-644
611-688
652-805
642-789
629-562
Percent
Curve
1990
384-451
449-527
480-596
472-588
458-559

Without
Learning Curve
222-265
327-385
376-448
405-494
369-445

-------
                              -204-


one third of the  production)  and  to insert into the numerator the
fraction of total sales belonging to that  particular manufacturer.

     The actual  fractions  of  sales  to be  used  will be taken from
Table  III-4,  representing the distribution  of 1979  sales, and
are shown below:

          Cummins             38.0%
          Detroit Diesel       24.6%
          Caterpillar          15.7%
          Mack                14.1%
          IHC                   7.6%

It has been assumed that  the  1979 market share will hold constant
throughout 1990.   Also,  the foreign  manufacturers have been assumed
to buy  their  traps  from the  other  manufacturers  since their com-
bined sales were  less  than three percent  of  total  sales in  1979.
Their sales have been distributed among the domestic manufacturers
in proportion  to  the latter"s  sales.

     When the above fractions  are used  in  Equations (3A) and (3B)
and  the  calculations  of  Section  A  are  repeated,  actual cost-to-
reference cost ratios for  each manufacturer  are generated.   These
are  shown in  the upper  two  sections  of  Tables A-II-14 through
A-II-18.  The  uppermost  section includes  the factors generated from
Equation (3A)  which apply to electronic control units and sensors.
The  second section (Equation (3B))  applies  to  exhaust  system
components .

     To  determine these  ratios as  they  apply to traps,  the same
modifications  must be  made as was done in  Section A.  That is, the
production  of Classes V-VIII must be  considered together and
multiplied by  two to represent  the use of  one size  trap  on all
those vehicles and  the  use of two  traps per  vehicle.   The ratios
for Classes II-B-IV remain the same  as those in the second section
of the tables.  These  trap ratios  are shown in the third  section of
Tables A-II-14 through A-II-18.

     Lastly,  the  ratios  for those components that vary with engine
design must  be  determined.   In Section A, it was  assumed  that  there
were ten basic engine designs  throughout the  industry.   Here, for
convenience,  a very conservative  assumption will be made that each
manufacturer produces  five basic engine designs,  one in each of the
three lightest groups  of vehicle classes  and two  in  Class  VIII.
This would total  25  engines across  the  industry and  will help to
make this  a  worst-case  analysis.   Since  the  first  three vehicle
groups have exactly one engine per group,  the engine production is
the same  as the  vehicle  group production  and the actual-to-refer-
ence cost ratios  are the same  as  in the  second  section of the five
tables.   In Class VIII, however, engine production is only half the
vehicle  group production,  so the ratios there  are 12  percent
higher than those  in  the  second  section (by  definition of  the 12
percent  learning  curve).   These  engine-dependent ratios are  shown
in the  bottom-most section  of Tables A-II-14 through  A-II-18.

-------
                              -205-



                         Table A-II-14

                            Cummins

      Revised Values for the Ratio of the Actual Cost of
     a Component to Its Cost at an Accumulated Production
     	Of 300,000 Units I/	

                             1986   1987   1988   1989   1990

Electronic Control Units     1.18   1.05   0.98   0.93   0.89

Cost Ratios if Component
Production Equals Vehicle
Group Production

   Class IIB, III, IV        1.79   1.58   1.45   1.36   1.30

   Class V, VI               1.50   1.33   1.23   1.16   1.11

   Class VII                 1.68   1.49   1.39   1.32   1.26

   Class VIII                1.30   1.15   1.08   1.03   0.96

Traps_2/

   Class IIB, III, IV        1.79   1.58   1.45   1.36   1.30

   Class V-VIIL3/            1.07   0.95   0.88   0.84   0.81

Components that Vary
with Engine Design

   Class IIB, III, IV        1.79   1.58   1.45   1.36   1.30

   Class V, VI               1.50   1.33   1.23   1.16   1.11

   Class VII                 1.68   1.49   1.39   1.32   1.26

   Class VIII                1.46   1.29   1.21   1.16   1.08
17Assumes Cummins captures  38.0% of the production shown
Tn Table A-II-1.

2/   Class V-VIII production combined.

3/   Assume two traps per vehicle.

-------
                               -206-


                         Table A-II-15

                         Detroit Diesel

      Revised Values for the Ratio of the Actual Cost of
     a Component to Its Cost at an Accumulated Production
     	Of 300,000 Units I/	

                             1986    1987    1988    1989   1990

Electronic Control Units     1.27    1.12    1.05    0.99   0.95

Cost Ratios if Component
Production Equals Vehicle
Group Production

   Class IIB, III, IV        1.92    1.69    1.55    1.46   1.39

   Class V, VI               1.61    1.42    1.32    1.24   1.19

   Class VII                 1.80    1.60    1.49    1.41   1.35

   Class VIII                1.39    1.24    1.11    1.03   1.00

Traps_2/

   Class IIB, III, IV        1.92    1.69    1.55    1.46   1.39

   Class V-VIII3/            1.15    1.02    0.95    0.90   0.87

Components that Vary
with Engine Design

   Class IIB, III, IV        1.92    1.69    1.55    1.46   1.39

   Class V, VI               1.61    1.42    1.32    1.24   1.19

   Class VII                 1.80    1.60    1.49    1.41   1.35

   Class VIII                1.56    1.39    1.25    1.16   1.12


_!_/   Assumes Detroit Diesel captures 24.6%  of the  production shown
in Table A-II-1.

_2/   Class V-VIII production combined.

_3/   Assume two traps per vehicle.

-------
                              -207-


                         Table A-II-16

                          Caterpillar

      Revised Values for the Ratio of the Actual Cost of
     a Component to Its Cost at an Accumulated Production
     	Of 300,000 Units I/	

                             1986   1987   1988   1989   1990
Electronic Control Units     1.36   1.21   1.13   1.07   1.03

Cost Ratios if Component
Production Equals Vehicle
Group Production

   Class IIB, III, IV        2.07   1.82   1.67   1.57   1.49

   Class V, VI               1.73   1.53   1.42   1.34   1.30

   Class VII                 1.93   1.72   1.60   1.52   1.46

   Class VIII                1.49   1.33   1.25   1.19   1.14

Traps_2/

   Class IIB, III, IV        2.07   1.82   1.67   1.57   1.49

   Class V-VIII_3/            1.23   1.10   1.02   0.97   0.93

Components that Vary
with Engine Design

   Class IIB, III, IV        2.07   1.82   1.67   1.57   1.49

   Class V, VI               1.73   1.53   1.42   1.34   1.30

   Class VII                 1.93   1.72   1.60   1.52   1.46

   Class VIII                1.67   1.49   1.40   1.34   1.28


T7Assumes Caterpillar captures  15.7% of the production shown
Tn Table A-II-1.

2]   Class V-VIII production combined.

3/   Assume two traps per vehicle.

-------
                               -208-


                          Table A-II-17

                              Mack

       Revised Values  for  the Ratio  of the Actual  Cost  of
      a Component to Its Cost at an  Accumulated  Production
     	Of 300,000 Units I/	

                             1986    1987    1988    1989   1990

 Electronic Control Units     1.39    1.24    1.15    1.09   1.05

 Cost Ratios if Component
 Production Equals Vehicle
 Group  Production

   Class IIB, III, IV        2.11    1.86    1.71    1.61   1.53

   Class V, VI               1.77    1.56    1.45    1.37   1.31

   Class VII                 1.98    1.75    1.55    1.49   1.44

   Class VIII                1.53    1.36    1.27    1.21   1.17

 Traps_2/

   Class IIB, III, IV        2.11    1.86    1.71    1.61   1.53

   Class V-VIin/            1.26    1.12    1.04    0.99   0.95

 Components that Vary
 with Engine Design

   Class IIB, III, IV        2.11    1.86    1.71    1.61   1.53

   Class V, VI               1.77    1.56    1.45    1.37   1.31

   Class VII                 1.98    1.75    1.55    1.49   1.44

   Class VIII                1.72    1.53    1.43    1.36   1.31


JY   Assumes Mack captures 14.1% of the production shown in Table
 A-II-1.

_2/   Class V-VIII production combined.

_3/   Assume two traps per vehicle.

-------
                              -209-


                         Table A-II-18

                    International Harvester

      Revised Values for the Ratio of the Actual Cost of
     a Component to Its Cost at an Accumulated Production
     	Of 300,000 Units I/	

                             1986   1987   1988   1989   1990
Electronic Control Units     1.56   1.39   1.29   1.22   1.17

Cost Ratios if Component
Production Equals Vehicle
Group Production

   Class IIB, III, IV        2.37   2.08   1.91   1.80   1.71

   Class V, VI               1.98   1.75   1.62   1.53   1.47

   Class VII                 2.21   1.97   1.83   1.74   1.67

   Class VIII                1.71   1.52   1.42   1.36   1.31

Traps_2/

   Class IIB, III, IV        2.37   2.08   1.91   1.80   1.71

   Class V-VIin/            1-41   1.25   1.17   1.11   1.07

Components that Vary
with Engine Design

   Class IIB, III, IV        2.37   2.08   1.91   1.80   1.71

   Class V, VI               1.98   1.75   1.62   1.53   1.47

   Class VII                 2.21   1.97   1.83   1.74   1.67

   Class VIII                1.92   1.71   1.60   1.53   1.47


T7Assumes IHC  captures  7.6% of the  production shown in Table
A-II-1.

2/   Class V-VIII production combined.

3/   Assume two traps  per  vehicle.

-------
                                -210-
     To determine the ratios  for  the entire industry, the ratios of
the  five  tables  must be  sales-weighted using  the  sales breakdown
shown above.  This has been done  and the results are shown in Table
A-II-19.

     All that  remains  to  be  done to  determine whole trap-oxidizer
system costs  is  to  apply the ratios of Tables  A-II-13 through
A-II-18 to the component costs of Table A-II-4.  This will be done
for  Cummins (largest manufacturer,  least cost),  IHC (smallest
manufacturer,  greatest cost), and the  industry  average.   A  sales-
weighted average across vehicle  groups for each year is also shown
using Equation (2) of Section A.

     The results  are  shown in Table A-II-20.  As can been seen,  the
results are quite close together.  Cummins'  costs  would be about 9
percent below the industry average and  IHC's cost would be about 23
percent higher than the  industry average.  Also, the industry
averages calculated  here  are only 4-5 percent higher  than those
calculated  in  Section A  (Table  A-II-9).   Thus,  the  sensitivity
analysis has shown that  industry  average costs are not sensitive to
the assumption that three  outside suppliers will provide all  of the
trap-oxidizers for  the  industry.  Two,  it has  shown that some
spread could occur between manufacturers  (maximum of  35 percent),
but given  that this  is  a worst  case  spread,  it is  actually quite
reasonable.   Thus,  given  the  small  likelihood of  this  situation
occurring,  the costs  calculated  in Section A  should be indicative
of the actual costs seen in the field  even if the  actual number of
suppliers  differed from three.

-------
                              -211-


                         Table A-II-19

                     Industry-Wide Average

      Revised Values for the Ratio of the Actual Cost of
     a Component to Its Cost at an Accumulated Production
     	Of 300,000 Units I/	

                             1986   1987   1988   1989   1990
Electronic Control Units     1.29   1.18   1.07   1.01   0.97

Cost Ratios if Component
Production Equals Vehicle
Group Production

   Class IIB, III, IV        1.95   1.72   1.58   1.48   1.41

   Class V, VI               1.64   1.45   1.34   1.26   1.21

   Class VII                 1.83   1.62   1.50   1.43   1.37

   Class VIII                1.41   1.26   1.17   1.11   1.06

TrapsjZ/

   Class IIB, III, IV        1.95   1.72   1.58   1.48   1.41

   Class V-VIII3/            1.17   1.04   0.96   0.92   0.88

Components that Vary
with Engine Design

   Class IIB, III, IV        1.95   1.72   1.58   1.48   1.41

   Class V, VI               1.64   1.45   1.34   1.26   1.21

   Class VII                 1.83   1.62   1.50   1.43   1.37

   Class VIII                1.58   1.42   1.31   1.25   1.19
I/   Assumes each manufacturer supplies his own trap-oxidizer,

2/   Class V-VIII production combined.

3/   Assume two traps per vehicle.

-------
                              -212-
                            Table A-II-20

          Revised Estimated Costs of Trap-Oxidizer Systems
           At Predicted Production Volumes (1980 dollars)

                        Industry-Wide
     Vehicle Class         Average         Cummins        IHC

    1986:
     IIB, III,  IV          579-667          500-580      713-819
        V,  VI              650-750          589-681      794-915
         VII               680-817          617-743      830-996
        VIII               683-832          621-758      837-1017

     Sales  Weighted:        664-795          604-724      817-975

    1987:

     IIB, III,  IV          507-585          461-524      621-715
        V,  VI              573-661          518-599      698-805
         VII               599-721          542-654      731-879
        VIII               604-737          544-666      736-897

     Sales  Weighted:        586-702          531-637      717-858

    1988:

     IIB, III,  IV
        V,  VI
         VII
        VIII

     Sales  Weighted:        537-644          491-589      665-795

    1989:
462-533
524-606
548-661
552-675
420-485
488-563
499-603
502-615
567-653
648-747
679-817
685-835
     IIB,  III,  IV          430-497          377-438       532-613
        V,  VI              498-575          451-522       612-706
         VII               522-629          526-572       642-772
        VIII               525-643          476-585       647-791

     Sales  Weighted:        508-610          465-553       625-749

    1990:
     IIB,  III,  IV           408-472          373-431       504-581
        V,  VI              475-548          432-500       588-677
         VII               497-600          453-548       616-694
        VIII               500-611          454-555       621-760

     Sales  Weighted:        483-578          438-525       598-710

Sales Weighted,  1986-1990   550-660          501-600       679-810

-------
                                  -213-
                            References

JY   Lindgren,  Leroy H. ,  "Cost  Estimations  for  Emission  Control
     Related  Components/Systems  and  Cost  Methodology Description,"
     Rath  and  Strong  for  EPA,  March  1978,   EPA-460/3-78-002.

2/   Moody's  Industrial Manual,  1979, Vol. 1, A-I.

3/   Penninga,  Thomas,  TAEB,  EPA, "Diesel Particulate  Trap Study:
     Interim  Report  on  Status  of Study and Effects of Throttling,"
     Memorandum  to  Ralph  C.   Stahman,  Chief,  TAEB,  EPA,  May  18,
     1979.

4-/   Rykowski,  Richard  A.,  SDSB, "Size  Considerations  Concerning
     the  Use  of Trap-Oxidizers  in Light-Duty Diesels," Memorandum
     to  Robert E.  Maxwell,  Chief,  SDSB, EPA,  November 5,  1979.

5/   Dun's Review,  September  1978,  Vol. 112, No.  3,  pp.  124,  125.

_6_/   Interagency  Study  of  Post  - 1980 Goals for  Commercial  Motor
     Vehicles, June  1976, p. 11-12.

1J   "1979 Gas  Mileage  Guide," Second  Edition,  OANR,  OMSAPC,  MVEL,
     EPA, January,  1979.

JJ/   American Metals Market, January, 1980.

_9_/   Personal Communications  with Bureau  of  Labor Statistics.

IP/  Penninga,  T.,  TAEB, EPA, "Second  Interim  Report on  Status  of
     Particulate  Trap  Study,"  Memorandum  to  R.  Stahman,  Chief,
     TAEB, EPA, August 28,  1979.

JJY  Alson, Jeffrey,  SDSB,  EPA,  "Meeting  Between Texaco and EPA to
     Discuss  Particulate  Trap  Work,"  Memorandum  to  the  Record,
     October, 1979.

12/  Regulatory Analysis - "Light-Duty Diesel  Particulate  Regula-
     tion," MSAPC, OANR, EPA, January 29,  1980.

13/  Springer,  Karl J.,  "Investigation of  Diesel-Powered  Vehicle
     Emissions:   VIII.    Removal of Exhaust  Particulate  from Mer-
     cedes  300D Diesel  Car,"  June 1977,  EPA 460/3-77-007,  p.  34.

14/  "Designing  for Automotive  Corrosion  Prevention,"  Society  of
     Automotive Engineers, November, 1978, p. 78.

15/  Passavant, Glenn  W. "Average Lifetime  Periods  for Light-Duty
     Trucks  and Heavy-Duty  Vehicles," EPA,  November  1979,  SDSB-
     79-24.

16/  Lipson  Charles and  Narenda J.  Sheth,  Statistical Design and
     Analysis of  Engineering  Experiments,  McGraw-Hill,  1973.

17/  "Heavy Duty  Mufflers  and Exhaust  System Parts,  Suggested List
     and Resale Price,"  Riker Manufacturing, No. 190-4.

-------