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.
------- |