EPA-AA-IMS/80-8
Derivation of 1981 and Later Light Duty Vehicle Emission Factors for
Low Altitude, Non-California Areas
November, 1980
NOTICE
This report does not necessarily represent the final EPA decisions or posi-
tions. It is intended to present a technical analysis of the issue using data
which are currently available. The purpose in the release of such reports is
to facilitate the exchange of technical information and to inform the public
of technical developments which may form the basis for a final EPA decision,
position or regulatory action.
Inspection and Maintenance Staff
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air, Noise and Radiation
U.S. Environmental Protection Agency
-------
Table of Contents
Page
I. Introduction 1
A. Purpose and Coverage 1
B. Background on the Post-1980 Fleet 1
C. Composition of the Fleet: Technology Types 2
D. Summary of Results 4
II. Description of the Data Base 6
III. General Methodology 10
A. Introduction 10
B. HC/CO Methodology 10
C. NOx Methodology 11
D. Comparison to the Methodology Used in the
Previous Analysis 12
IV. Specific Unit Analyses - HC/CO 13
A. Closed Loop vehicles designed to meet
0.41 g/mi HC and 3.4 g/mi CO standards 13
B. Closed Loop vehicles designed to meet
0.41 g/mi HC and 7.0 g/mi CO standards 25
C. Oxidation Catalyst vehicles designed to meet
0.41 g/mi HC and 3,4 g/mi CO standards 29
D. Oxidation Catalyst vehicles designed to meet
0.41 g/mi HC ammd 7.0 g/mi CO standards 33
V. Specific Unit Analyses - NOx 35
A. Closed Loop vehicles designed to meet
a 1.0 g/mi NOx standard 35
B. Oxidation Catalyst svehicles designed to meet
a 1.0 g/mi NOx standard 45
VI. Composite Emission Factors for the 1981 Federal Fleet 47
VII. Composite Emission Factors for the 1982 Federal Fleet 53
VIII. Composite Emission Factors for the 1983 and Beyond
Federal Fleet 58
IX. Comparison to Previous EPA Emission Factor Estimates 63
-------
I. INTRODUCTION
A. Purpose and Coverage.
The purpose of this document is to describe the methodology used in revis-
ing the emission factor equations for the post-1980 Light Duty vehicle
fleet. The available data will be presented and discussed and the perti-
nent assumptions and analyses will be outlined. Low Altitude,
Non-California, gasoline-fueled Light Duty vehicles will be the only cate-
gory of vehicles covered in this report. Light Duty Trucks, Light Duty
Diesels, California vehicles, and High Altitude vehicles are all covered
under separate analyses, although the other analyses will often use this
document as a source analysis. This document is only concerned with the
non-I/M case. The I/M case will be discussed in a separate analysis and
will result in substantially different emission factor equations. This
document is one contributor to a larger effort designed to revise the
entire Mobile Source Emission Factors Document (EPA-400/ 9-78-005).
B. Background on the Post-1980 Fleet.
The post-1980 Light Duty vehicle fleet merits a separate analysis from the
current fleet for several reasons. Beginning in 1981, the Federal exhaust
emission standard for oxides of nitrogen (NOx) drops from 2.0 g/mi to 1.0
g/mi. The hydrocarbon (HC) standard remains at 0.41 g/mi. in 1981 and the
carbon monoxide (CO) standard drops to 3.4 g/mi for most vehicles. It is
the change in the NOx standard specifically that is of significance. The
effect of this change in the standard will be to lead manufacturers of
most vehicles to adopt a technology which utilizes what has become known
as a Three-Way catalyst. It is called a Three-Way catalyst because it
allows not only the conversion of hydrocarbons (HC) and carbon monoxide
(CO) as with a conventional Oxidation or Two-Way catalyst, it also allows
the catalytic conversion of NOx. Thus it provides a new and previously
unused source of NOx control. To enable the catalyst to perform these
three conversion functions simultaneously, precise control of the air/fuel
ratio is required. This is most often accomplished through the use of an
on-board microprocessor which receives inputs from a variety of sensors
(notably the oxygen sensor, which is located in the exhaust stream and
provides an indication of the air/fuel ratio), processes these inputs con-
tinuously, and then provides an output signal to the carburetor or fuel
injectors to adjust the air/fuel ratio. The system thereby provides a
feedback loop and is therefore also known as a Closed Loop system. A
microprocessor can be designed to control other engine functions as well,
such as spark timing, idle speed, and EGR flow rate. Thus the net effect
of the change in the NOx standard will be to introduce significantly
different technology into the fleet in large measure beginning in 1981.
Due to the differences between this new technology and the more conven-
tional technology of the past several years, it is to be expected that
there will be differences in the in-use emissions performance of the
-------
post-1980 fleet. These differences in in-use performance necessitate a
separate analysis. This fact was recognized in the previous edition of
the Mobile Source Emission Factors Document (EPA-40Q/ 9-78-005) which this
document revises and updates.
Two other considerations that affect the post-1980 fleet need to be raised
here. First, beginning in 1981, the "Parameter Adjustment" regulations
(44 F.R. 2960) will be applied to the fleet. These regulations will cause
limitations in the adjustability of some of the basic parameters of the
engine. In 1981, these regulations will be applied idle mixture and
choke, and in 1982, they will be applied to timing. While these regula-
tions have less of an effect on most Closed Loop vehicles than on most
Oxidation catalyst vehicles, due to the largely self-adjusting nature of
the former, the Parameter Adjustment regulations will nonetheless have an
impact, especially on certain parts of the fleet. The effect of the Para-
meter Adjustment regulations has been taken into account throughout this
analysis.
A second consideration affecting the post-1980 fleet is the presence of
the Clean Air Act Section 202(b)(5) waiver fleet: those cars that re-
ceived a CO waiver from 3.4 g/mi to 7.0 g/mi in 1981 and 1982. In 1981,
and to a lesser extent in 1982, a portion of the fleet will be designed to
meet a 7.0 g/mi CO standard as a result of these waivers. These cars can
be expected to have higher CO emissions in general due to the higher stan-
dard, and the impact of those higher emissions, although small, has been
figured into the fleet-average emissions in this analysis.
C. Composition of the Fleet; Technology Types.
Before discussing the methodology in depth, the projected make up of the
fleet needs to be discussed. In terms of the emission control systems to
be employed there will essentially be three different systems in the
post-1980 fleet, however two of these three are assumed to have similar
emissions performance. Thus, this analysis distinguishes only two tech-
nology types with unigue emissions performance from among the fleet. The
distinguishing characteristics of the two technology types will be briefly
presented below.
1. Closed Loop Vehicles.
By far the largest percentage of the fleet will be comprised of this
technology type. The distinguishing characteristics of this tech-
nology type are feedback control of the air/fuel ratio and the use of
a Three-Way catalyst. In reality, as was alluded to above, this
technology type could be further broken down into two separate tech-
nology types: those vehicles equipped with an Oxidation catalyst,
supplied with air by an air pump, following the Three-Way catalyst
and those vehicles without the additional Oxidation catalyst and air
-------
pump. Three reasons were responsible for not making this further
division of Closed Loop vehicles. First, the manufacturers of most
vehicles have indicated that during significant failure modes of
their Closed Loop vehicles, the air flow from the air pump will be
diverted to the atmosphere (or to the air cleaner for silencing)
instead of to the Oxidation catalyst. This is due to concerns for
catalyst protection. As will be discussed later, it is these signi-
ficant failure modes of Closed Loop vehicles that are assumed to make
large contributions to the overall fleet composites. During a fail-
ure mode of this type, without the benefit of air being supplied to
the Oxidation catalyst, the in-use data indicates that these systems
have HC and CO emission levels as high as systems without the addi-
tional Oxidation catalyst and air pump. Thus, there is no need to
distinguish between these two systems during significant failure
modes.
Second, examination of in-use data from both types of vehicles (with
and without the additional Oxidation catalyst and air pump) did not
reveal a significant and consistent difference between the emissions
of the two systems when operating at other than significant failure
modes.
Third, the presense or absence of an Oxidation catalyst and air pump
is not assumed to effect NOx significantly.
In summary then, the two systems will be treated as one technology
type having a unique emissions performance.
One final point to be made regarding this technology type has to do
with Ford Motor Company vehicles in the 1981-1983 timeframe. Ford
has indicated that it intends to certify a large portion of its
1981-1983 fleet as open loop vehicles equipped with Three-Way cata-
lysts. That is, these vehicles would not employ an on-board micro-
processor with a feedback oxygen sensor but would still have a
Three-Way catalyst to enable some catalytic reduction of NOx. Due
to a lack of any in-use data on these systems at the time of this
analysis, as well as to uncertainty as to the fraction of Ford vehi-
cles which will have open loop systems, these vehicles were included
under the Closed Loop technology type. In 1984, any Ford vehicles
which had been open loop are assumed to go closed loop due to the
advent of the High Altitude regulations.
2. Oxidation Catalyst Vehicles.
Some vehicles, notably small foreign vehicles and vehicles with
unique engine configurations, will be able to meet the 1981 standards
without the catalytic control of NOx provided by Three-Way catalyst
-------
technology. Instead, they will rely on an Oxidation catalyst, air
pump or pulse-air system, and, in most cases, EGR. These vehicles
are expected to comprise a relatively small segment of the fleet.
D. Summary of Results.
The results of this analysis can best be presented by discussing the vari-
ous failure modes expected to occur and by briefly characterizing the per-
formance of each pollutant (HC, CO, NOx). For the majority of the fleet,
which consists of Closed Loop vehicles, the principal failure mode result-
ing in significant emission increases occurs through the loss of the
closed loop capability of the system and a resultant rich mode of oper-
ation. These failure modes, while not initially very numerous, have a
large impact on the overall fleet emissions for HC and CO due to the very
high emissions resulting for those two pollutants. Thus, a relatively
small percentage of the fleet contributes a disproportionately large share
of the final composite emissions for HC and CO. This is especially true
for CO. The data base of in-use vehicles which was relied on in perform-
ing this analysis gave significant indication of this type of behavior for
Closed Loop vehicles. NOx emissions for those cars with an open loop
failure will decrease due to the rich operating condition.
For Oxidation Catalyst vehicles, which are designed to operate open loop,
a more traditional deterioration pattern is assumed to occur. Briefly
stated, the regression methodology used for the 1975-1980 Light Duty fleet
was revised to represent 1981 and later Oxidation Catalyst vehicles. The
revised methodology accounts for the effect of the Parameter Adjustment
regulations and for the fact that the 1981 standards are more stringent
than the 1975-1980 standards.
A graphical comparison of the new emission composites with the emission
composites arrived at in the 1978 analysis (Appendix E of the Mobile
Source Emission Factors Document, EPA-400/9-78-005) will be presented in
Section IX. To generally characterize that comparison however, there is a
slight increase in the HC composite at 50,000 miles, a definite increase
in the CO composite at 50,000 miles, and a slight increase in the NOx com-
posite at 50,000 miles.
The following table presents the new emission factor equations:
-------
Table 1-1
Exhaust Emission Rates for Light Duty Vehicles
For All Areas Except California and High Altitude
New Vehicle Deterioration Rate
Pollutant Model Year Emission Rate (g/mi/10,000 miles)
(g/mi)
HC
HC
HC
CO
CO
CO
NOx
NOx
NOx
81
82
83+
81
82
83+
81
82
83+
0.39
0.39
0.39
5.60
5.21
5.00
0.75
0.75
0.75
0.19
0.19
0.19
2.75
2.76
2.76
0.15
0.15
0.15
-------
II. DESCRIPTION OF THE DATA BASE
A. Introduction.
The data used in this analysis come from a variety of sources. The bulk
of the data comes from EPA testing programs in Los Angeles designed to
test in-use Closed Loop vehicles.[1,2,3]* In addition, data from 50
in-use Closed Loop vehicles tested in Portland, Oregon were used.[4]* For
each set of data, the effect of the Parameter Adjustment regulations was
accounted for by removing from the data base those cars which showed evi-
dence of removal of idle mixture limiting devices where applicable, a mal-
adjustment of idle speed of greater than 200 rpm**, or a timing maladjust-
ment of greater than +5°. In some cases, data from vehicles with evidence
of maladjusted parameters, but which also had evidence of other non
parameter-related problems, were retained in the calculations of the
levels of pollutants not primarily affected by the maladjusted parameter
and/or in the calculation of the incidence of non parameter-related prob-
lems.
A final preliminary consideration which needs to be mentioned at this
point has to do with the methane correction factor for vehicles certified
in California. The California certification process accounts for the fact
that a certain portion of the HC measured in the exhaust is methane
(Cfy). Vehicles being certified in California are allowed to claim a
"credit" for that portion of the HC" exhaust which is methane. The de-
fault credit is 11% methane, however manufacturers can claim more credit
by demonstrating that their vehicles emit a higher fraction of methane.
Since most of the vehicles in the data base were certified in California,
yet this analysis is concerned with non-California vehicles, this differ-
ence needed to be accounted for. Use of the methane credit has the effect
of raising the total .HC design standard from the perspective of this ana-
lysis. For example, vehicles receiving the default methane credit of 11%
can emit up to 0.46 g/mi total HC and still pass California certification
after the credit has been applied. This analysis accounted for this rela-
tive difference in effective design standards by applying a ratio of those
design standards (e.g. 0.41/0.46) to the HC emission levels found in the
data base, except in those cases where the emission levels were judged to
be independent of the design standard. These exceptions will be pointed
out as they occur.
The principal data used for analyzing the emissions performance of the two
technology types will be briefly presented in the following sections.
* Numbers in brackets indicate references listed at the end of the Section.
** At the time the analysis was finalized, idle speed was included in the
Parameter Adjustment regulations as a parameter which would need to be de-
signed to be non-adjustable beginning in 1982. That requirement has since
been, lifted. The ' analysis was not revisted to account for the change
since it was determined to not have a significant impact and due to the
fact that the composite emission factors had already been finalized.
-------
B. Closed Loop Vehicle Data Base.
The vehicles from the following four groups were combined to create an
overall data base to be used in determining the emissions characteristics
of Closed Loop vehicles. An assortment of various systems and engine
sizes are represented. This data base is assumed to give representative
emission values for Closed Loop vehicles. The four groups which make up
the data base represent the most recent and advanced technology put for-
ward by the various manufacturers. The simpler, first generation systems
were excluded from the data base (e.g. 1978/ 1979 Ford Pintos, 1978/1979
Pontiac Sunbirds) due to being judged unrepresentative.
1. 1979 Ford/Mercury 351 CID engine family (3.8WBV2TT95x95)
equipped with Electronic Engine Control II (EEC-II).
a. A total of 97 in-use vehicles were tested. 82 vehicles were
tested by contractors in Los Angeles, and 15 vehicles were
tested in Portland, Oregon. Six vehicles were eliminated due to
Parameter Adjustment concerns. This engine family is equipped
with an Oxidation catalyst following the Three-Way catalyst and
an air pump which supplies air to the Oxidation catalyst.
b. This engine family uses a digitally based microprocessor
which allows very sophisticated control of the engine. It regu-
lates not only the air/fuel ratio, but also spark timing, EGR
flow rate, and the deployment of air flow from the air pump.
c. Of the 15 vehicles tested in Portland, Oregon, one was from
California and was therefore designed to meet the 1979 Californ-
ia emissions .standards (0.41/9.0/1.5). The other vehicles,
while being sold outside of California, were nonetheless assumed
to be designed to meet the 1979 California standards for HC and
C0_ and the" 1979 Federal Standard for NOx (2.0 g/mi). ThUf was
done based on the technical assumption that the 1979 351 CID
engine family was intended to be an in-the-field test of the EEC
system. As such, it was primarily designed to meet the tighter
California standards to give a better indication of how well the
system would perform under the eventual 1981 Federal standards
(0.41/3.4/1.0). Thus the basic system calibration was de-
signed to meet the 1979 California standards rather than the
looser 1979 Federal standards (1.5/15.0/2.0). For the vehicles
sold outside California, however, (i.e. the vehicles tested in
Portland) it is assumed that the calibrations pertaining to NOx
(e.g. spark timing, EGR flow rate) were relaxed due to fuel
economy and driveability concerns. This could easily be done
without significantly affecting the primary HC and CO calibra-
tion. The in-use data from Portland well supports this assump-
tion: the average HC and CO levels from Portland are the same as
or lower than the average HC and CO levels from Los Angeles and
the average Portland NOx levels are higher.
-------
d. These vehicles received a 17% Methane credit.
2.1980 General Motors California X-Body vehicles (01C2XCP
01C2XC. and 02X2NC) equipped with the Computer Controller?
Catalytic Convertor system (C-4). 151 and 171 CID.
a. A total of 92 in-use vehicles were tested, all in the Los
Angeles area. 9 vehicles were eliminated due to Parameter
Adjustment concerns.
b. These vehicles employ a digitally based control system. The
C-4 system in this application primarily controls the air/fuel
ratio. The C-4 system is the basic system projected to be
employed by General Motors throughout the early 1980's.
c. These vehicles received the standard 11% Methane credit.
3. 1979 Toyota Celica Supra vehicles (4M-E) 156 CID.
a. A total of 25 in-use vehicles were tested, all in the Los
Angeles area. Two vehicles were eliminated due to Parameter
Adjustment concerns.
b. These vehicles are fuel injected and employ an analog based
control system.
c. These vehicles received.the standard 11% Methane credit.
4. 1979 VW Audi: 5000. 131 CID.
a. A total of 4 in-use vehicles were tested, all in the Los
Angeles area. No vehicles were eliminated due to Parameter
Adjustment concerns.
b. These vehicles are fuel injected. They do not use EGR.
c. These vehicles received the standard 11% Methane credit.
C. Oxidation Catalyst Vehicle Data Base.
1. Data support for this technology type, which is similar in most
respects to current oxidation-catalyst-and-air-pump technologies, was
obtained from the analysis completed to revise the emission factor
equations for 1975-1980 vehicles. That analysis was based on in-use
data obtained from a wide array of 1975-1979 model year vehicles
tested around the country through EPA's Emission Factor Program. The
specific methodology involved in applying that data to this analysis
will be presented later (Sections IV and V).
-------
D. Miscellaneous Data. 1978-79 Volvos and Saabs equipped with the Lambda
Sond system (4CL, 6CL, BT20CA and BSI20CA). A total of 162 Volvos and
Saabs were tested in programs in Los Angeles and Portland.[1,2,3,4] Data
from these vehicles were used chiefly in determining deterioration rates
for Closed Loop vehicles (see Section IV.A.2.b.).
This fleet has a much wider mileage spread than found in the major Closed
Loop vehicle data base and could therefore be used in determining deteri-
oration rates. All Volvo and Saab vehicles marketed west of the Missis-
sippi have the same engine calibrations, so the Portland and Los Angeles
vehicles could be analyzed together without needing to take different
standards into account.
References for Section II.
1. EPA Contract No. 68-03-2590 with Automotive Environmental Systems, Inc.
Results published in EPA document EPA-460/3-79-004.
2. EPA Contract No. 68-03-2889 with Automotive Environmental Systems, Inc.
Results published in EPA document EPA-46Q/3-80-006.
3. EPA Contract No. 68-03-2774 with Automotive Testing Laboratories, Inc.
Results not published at the time of this report.
4. EPA Contract No. 68-03-2829, Test Group No. 1, with Hamilton Test Systems,
Inc. Results published in EPA document EPA-460/3-80-006.
-------
10
III. GENERAL METHODOLOGY
A. Introduction.
This section presents the general methodology used in constructing the
emission factor model for the post-1980 fleet. Specific methodologies are
presented in detail in Sections IV and V. The HC/CO methodology and the
NOx methodology will be discussed separately. HC and CO have been ana-
lyzed together due to a basic similarity in the in-use failure modes which
result in high HC and CO emission levels, whereas the failure modes to
which NOx is sensitive are relatively independent. The general methodol-
ogy used in this analysis is similar in concept to that used in the 1978
analysis (Appendix E) as will be discussed later in this section.
B. HC/CO Methodology.
1. Unit of Analysis. The basic units of analysis (i.e., the unique
subfleets of similar vehicles for which unique analyses are performed
and which are aggregated later into a fleet-composite analysis) are
individual technology types certified to specific pairs of HC/CO
standards. Thus, each of the two technology types discussed in Sec-
tion I.e. will comprise a separate unit of analysis for each of two
possible combinations of HC/CO standards, for a total of four units
of analysis. The two possible combinations of HC/CO standards are
0.41 g/mi HC with 3.4 g/mi CO, and 0.41 g/mi HC with 7.0 g/mi CO.
This second combination stems from the CO waiver decision.
2. Categories. Each unit of analysis representing Closed Loop vehi-
cles will be divided into categories (i.e. groupings of vehicles
within the overall unit of analysis which have similar emission char-
acteristics). Several different vehicle conditions could be included
under and contribute to the size of a category, but a single set of
emission characteristics for each pollutant describes the category.
Oxidation Catalyst vehicles will not be divided into categories.
3. Emissions Performance of the Categories. The average emission
characteristics of each category for a given pollutant are described
by two parameters:
a. Zero-mile level. The zero-mile level is the average emis-
sion TeverTor~a~pollutant at zero miles.
b. Deterioration rate. The deterioration rate is the amount
of increaseInthe emission level of a pollutant per 10,000
miles.
The unit of measurement for the zero-mile level is g/mi. The unit of
measurement for the deterioration rate is g/mi/10,000 miles. The
deterioration rate should not be confused with the "deterioration
-------
11
factor" which is the ratio of the emission level at 50,000 miles to
the emission level at 4,000 miles. The concept of the deterioration
factor will be used later.
4. Migration between Categories. Within a given unit of analysis,
for example the Closed Loop vehicle technology type designed to meet
a 0.41 g/mi HC standard and a 7.0 g/mi CO standard, the size of the
different categories will change with time. As would be expected,
the general movement over time is from categories representing more
well-maintained vehicles with lower emissions _to categories repre-
senting less well-maintained vehicles or vehicles with some component
failure with correspondingly higher emissions. The growth rates for
the categories are expressed in units of percentage of the total unit
which enters the category per 10,000 miles. Some categories of
course will have negative growth rates. That is, as some categories
grow, others must decrease in size.
5. unit-of-Analysis Composite Emissions. The emissions from each of
the categories within a unit of analysis will be weighted together
every 10,000 miles to arrive at a composite emission characteristic
for that unit. This will be done by weighting the emission levels of
each category at a given mile point by the fraction of the unit of
analysis represented by that category at the given mile point. The
net result will be a table of the unit's emission levels. at each
10,000 mile point. This table will reflect the relative contribution
of the various categories based on the size of the categories as well
as their individual emission levels. Examples of this concept can be
found in Sections IV and V.
6. Fleetwide Composite Emissions. To obtain a fleetwide composite
emission factor for a given model year, the emission composites from
each of the units of analysis will be combined according to a weight-
ing scheme similar to that described above for combining categories
within a unit of analysis., The separate unit-of-analysis emission
composites will be weighted according to the respective fractions of
that model year's fleet represented by those units of analysis.
These fractions were determined by examining manufacturer's state-
ments and certification data. The determination of how the fleet was
broken down into the various units of analysis will be discussed in
depth for each model year in Sections VI through VIII.
C. NDx Methodology.
The NOx methodology is essentially identical in concept to the HC/CO meth-
odology. In the following subsections only the pertinent differences will
be mentioned.
-------
12
1. Unit of Analysis. For NOx, there are only two units of ana-
lysis. There is only one applicable NOx standard, 1.0 g/mi, and NOx
is not assumed to be affected by the different CO standards intro-
duced by the waiver decision. Thus the two units of analysis are:
a. Closed Loop vehicles
b. Oxidation Catalyst vehicles.
2. Categories.
a. Closed loop vehicles will have four categories of operating
condition.
b. Oxidation Catalyst vehicles will not be divided into cate-
gories.
3. Emissions Performance of the Categories. - No difference.
4. Migration between Categories. - No difference.
5. Uhit-of-Analysis Composite Emissions. - No difference.
6. Fleetwide Composite Emissions. - No difference.
D. Comparison to the Methodology used in the Previous Analysis.
1. The methodology used in this analysis is similar to that used in
developing Appendix E (Appendix E of the Mobile Source Emission Fac-
tor Document EPA-400/9-78-005). Appendix E divided a unit of ana-
lysis into different categories with independent emission performance
characteristics (zero-mile levels and deterioration rates) as has
been done in this analysis. Those categories also grew or declined
in size over time, and a unit composite was obtained by weighting the
various categories and adding them together. The chief methodologi-
cal difference between Appendix E and this analysis is that Appendix
E used a single unit of analysis to represent the entire fleet for
all model years after 1980 whereas this analysis has four distinct
units of analysis for HC and CO and two units of analysis for NOx.
Appendix E assumed one technology type; this analysis assumes two
different technology types. Appendix E treated only one pair of HC
and CO standards; this analysis has needed to account for a wider
array of combinations of HC and CO standards due to the waiver de-
cisions. As will be apparent later, the categories defined for this
analysis are also very different from the categories used in Appendix
E, as are the categories' emission levels.
-------
13
IV. SPECIFIC UNIT ANALYSES - HC/CO
A. Closed Loop vehicles designed to meet 0.41 g/mi HC and
3.A g/mi CO standards.
1. Category definitions.
a. Primary. This category is comprised of those vehicles,
designed to operate closed loop, which have either lost their
closed loop capacity, or have had it severely restricted.
These vehicles are assumed to then operate in a rich mode.
Possible contributors to this category include vehicles with
catastrophic oxygen sensor failure, microprocessor failure,
other sensor failures or tampering with the closed loop system.
It was necessary to select from the data base the subset of
vehicles which would be relied upon in predicting the emission
characteristics of this category. This was done by selecting
all vehicles which had CO emission levels greater than 50.0
g/mi, on the premise that any vehicle with CO emissions this
high must be experiencing rich operation due to a failure of the
closed loop control system.
b. Secondary. This category is comprised of all the vehicles
in the fleet not in the Primary category and not in the Mis-
fueling category (see below). Contributors to the Secondary
category include vehicles experiencing general malmaintenance,
failure or degradation of sensors not leading to loss of closed
loop operation, tampering, as well as a large percentage of
vehicles experiencing generally good maintenance with resultant
low emission levels.
As with the Primary category, a subset of the data base fleet
was selected to be used in predicting the emission characteris-
tics of this category. This was done by selecting all vehicles
which had CO emission levels less than 50.0 g/mi.
c. Misfueling. This category is comprised of vehicles which
have been misfueled and have therefore had their emission con-
trol systems damaged. Misfueling refers to the use of leaded
fuel in vehicles equipped with catalysts. The leaded fuel in
effect "poisons" the catalyst and dramatically reduces its abil-
ity to convert HC, CO and NOx to harmless by-products. Misfuel-
ing also damages the oxygen sensor on Closed Loop vehicles. It
affects the output voltage of the sensor as well as hindering
the ability of the sensor to respond quickly to changes in the
composition of the exhaust stream. This normally results in
more rich operation and consequently higher HC and CO emission
-------
14
levels. The interaction of this category with the other two
will be described in a later section dealing with category sizes
(Section IV.A. 3).
2. Emission levels of the categories.
a. Primary Category.
i. Zero mile levels. The average measured HC/CO emission
levels and the average odometer mileage of the vehicles selected
from the data base to represent the Primary category are:
Average HC = 3.85 g/mi (n = 10)
Average CO = 108.rf g/mi (n = 10)
Average Mile = 9,163 miles (n = 10)
(Note: Of the ten vehicles in the data base which met the cri-
teria to be selected to represent the Primary category (greater
than 50.0 g/mi CO) five were from the General Motors X-Body
fleet and five were from the Ford 351 fleet.)
Applying the deterioration rates (see below) to these average
emission levels results in the following back-projected emission
levels at zero miles:
HC = 3.74 g/mi
CO = 107.36 g/mi
(Nate: The reader will note that for the Secondary category
discussed below, the zero-mile CO level was adjusted by a ratio-
ing factor of 3.4/9.0 to account for the fact that the vehicles
in the data base were designed to the California CO emission
standard of 9.0 g/mi rather than to the 3.4 g/mi standard which
applies to this unit of analysis. The CO emission level was not
ratioed to the 3.4 g/mi CO standard for the Primary category
since this category represents vehicles with a major, failure
mode during which CO emissions are assumed to be effectively
independent of the CO design standard. This same logic dictated
that the HC emission levels need not be ratioed to account for
. the various effective HC design standards resulting from the
application of methane credits. See Section II.A. for a more
complete discussion of this issue.)
ii. Deterioration rates. The HC/CO deterioration rates for the
Primary category were obtained by adopting the deterioration
rates developed for the Secondary category. That development
will be presented in the next section. The deterioration rates
-------
15
for the Secondary category were adopted due to a lack of data
describing the deterioration of Primary category vehicles. The
deterioration rates arrived at are:
HC= 0.12 g/mi/10,000 miles
C0= 0.71 g/mi/10,000 miles
b. Secondary Category.
i. Zero-mile levels. The average measured HC/CO emission
levels and the average odometer mileage of the vehicles in the
data base which represent the Secondary category are:
Average HC = 0.32 g/mi (n = 191)*
Average CO = 5.47 g/mi (n = 195)*
Average Miles = 8,064 miles (n = 191)
Applying the deterioration rates (see below) to these average
emission levels to back-project the levels to zero miles and
then applying a ratio of 3.4/9.0 to the CO emission level to
reflect the fact that the vehicles in the data base were de-
signed to meet a 9.0 g/mi CO standard and not the 3.4 g/mi stan-
dard which is the focus of this analysis, results in the follow-
ing Zero-mile emission levels:
HC = 0.23 g/mi
CO = 1.86 g/mT
ii. Deterioration rates.
The development of deterioration rates for the Secondary cate-
gory was an involved process. Before presenting that process in
detail, several points can be made which will give helpful back-
ground on why the deterioration rates were developed as they
were.
First, the vehicles in the Closed Loop vehicle data base do not
have a large enough mileage spread to support a credible regres-
sion analysis (which would yield a deterioration rate and
zero-mile level). The vast majority of vehicles in the data
base have odometer mileages between 6,000 and 12,000 miles.
* The sample sizes used to calculate the average emission levels
for the two pollutants differ due to the fact that Parameter
Adjustment concerns eliminated vehicles for one pollutant but
not for the other. See Section II.A.
-------
16
The next most apparent source of data for examining the deteri-
oration of Closed Loop vehicles is the deterioration observed in
the EPA Certification process. For each engine family with a
unique emission control system produced in a given model year, a
representative vehicle from that family must demonstrate the
ability to meet the applicable emission standards out to 50,000
miles in order to be certified. While the deterioration ob-
served in the Certification process is believed to be a good
measure of the relative durability of a given vehicle's emission
control system, it does not, however, give an accurate picture
of how the vehicle will perform out in the field. It has always
been the case that in-use vehicles deteriorate at a faster rate
than that observed in Certification. This is due to owner mal-
maintenance and tampering, harsh real-world conditions, and
other factors. This is assumed to still be the case in the
post-1980 timeframe.
Given the fact then, that the data base used to determine
zero-mile levels could not be relied upon to predict deteriora-
tion and that the use of Certification .deterioration factors by
themselves would be unrealistic, it was necessary to rely on
other sources of data to determine the deterioration rates. As
was mentioned in Section II.D., there is a fleet of in-use
Closed Loop vehicles, the Volvo/Saab fleet, which does have a
significant mileage spread. This fleet is made up of 162 1978
and 1979 Volvos and Saabs with a mileage spread of between 0 and
30,000 miles. There are, however, substantial reasons why the
in-use deterioration observed for these vehicles could not be
adopted outright to represent the in-use deterioration of the
post-1980 fleet. These reasons center around differences in
technology between the Volvos and Saabs and typical post-1980
vehicles. The Volvos and Saabs are fuel injected as opposed to
carbureted, have an analog based control system as opposed to a
digitally based control system, do not use EGR and have a
European manufacturer. These differences are significant,
especially so for NOx. Since an approach was needed which could
be applied consistently to all three regulated pollutants, (HC,
CO, NOx), the differences in technology needed to be taken into
account.
In view of all of the above-mentioned considerations, the
following procedure was decided upon. First, a regression was
performed on the Volvo/Saab fleet after removing vehicles due to
Parameter Adjustment concerns. Two vehicles with FTP CO emis-
sions greater than 50.0 g/mi were also removed from the fleet
since these vehicles would be classified as Primary category
vehicles. The regression yielded a zero-mile level and a slope
which represent in-use deterioration. Second, the deterioration
-------
17
predicted by Certification for the 1978 and 1979 Volvos and
Saabs was quantified by weighting together the Certification
deterioration factors (d.f. *s) for each of the possible model
year/engine size combinations among the Volvo/Saab fleet. The
weighting technique was based on the number of vehicles in the
fleet which came from the various model year/engine size combin-
ations. There were some HC d.f. 's from some of the model
year/engine size combinations which were less than 1.0 which
were raised to 1.0 for the purposes of calculating the weighted
average. Third, a ratio was formed using the in-use deteriora-
tion and the Certification deterioration. This ratio gives an
indication of how Secondary category Closed Loop vehicles deter-
iorate in the field relative to how prototypes of the same
models deteriorate in the Certification process; the ratio can
be used as an adjustment, or correction factor, to predict
in-use deterioration of Secondary category vehicles given a
figure of Certification deterioration. This ratio was based on
units of the "deterioration factor minus one" (d.f.-l). (The
in-use deterioration regression was easily converted to a d.f.-l
by first finding the emissions at 50,000 miles and 4,000 miles,
taking the ratio of those two figures to obtain a d.f. and then
subtracting one to obtain the d.f.-l). The d.f.-l was used in
the ratio since it is the portion of the d.f. greater than one
which represents the percent increase in a pollutant over 50,000
miles. To use the d.f. by itself in the ratio would have been
misleading since the d.f. is itself a ratio (the emissions at
50,000 miles over the emissions at 4,000 miles).
Finally, the ratio described above, which predicts the rela-
tionship between Certification deterioration and in-use deteri-
oration, was applied to an average Certification d.f.-l for 1981
Closed Loop vehicles in the Certification process (n = 89).
Applying the ratio yields a figure for an in-use d.f.-l which
can be used to represent the in-use deterioration of Secondary
category post-1980 vehicles. The following equation illustrates
this relationship:
x (d.f .-l)ynivn/Saah in-use = (d«f --98+ n-use
(d.f.-l)volvo/Saab Cert.
The in-use d.f.-l arrived at can be converted to a d.f. by add-
ing one. This deterioration factor (d.f.) can then be converted
to a deterioration rate (d.r.) by using the following equation
and the average emission levels at the average mileage (approxi-
mately 8,000 miles) of the Secondary category vehicles found in
the Closed Loop in-use data base.
-------
18
d.f.[in-use] = Emissions at 50.000 miles _ Emissions at 8,000 miles + (4.2)(d.r.)
Emissions at 4,000 miles ~ Emissions at- 8,000 miles-(.4)(d.r.)
The only unknown in this equation is the deterioration rate
(d.r.) which can therefore be easily solved for.
Two exceptions to this procedure need to be mentioned here be-
fore presenting a table of the values used in the calculations
and the deterioration rates arrived at. First, for HC, the
average Volvo/Saab Certification d.f. (1.02) was so low that
using the d.f.-1 to find the ratio between in-use and Certifica-
tion deterioration resulted in an unrealistically high ratio
(40.5). In view of how low the HC d.f. is and how it thereby
unrealistically inflates the in-use/Certification ratio, it was
decided to regard the HC Certification d.f. as an anomaly and to
adopt the ratio observed for CO (5.6) to be used for HC as
well. Given the fact that HC and CO emissions are both affected
in the same way by relatively similar malperformances for Closed
Loop vehicles, this approach is judged to give a better estimate
for the HC ratio.
The second exception to the general procedure outlined above has
to do with the fact that for CO, the vehicles in the data base
were designed under a 9.0 g/mi CO standard whereas this unit of
analysis consists of vehicles designed to meet a 3.4g/mi stan-
dard. The following set of equations was used to translate from
the in-use data point based on 9.0 g/mi CO standard vehicles to
arrive at a slope for vehicles designed to a 3.4 g/mi CO stan-
dard. These equations incorporate the assumption of equal
slopes (or deterioration rates) for vehicles designed to meet
3.4 g/mi or 9.0 g/mi CO standards and the assumption that the
Zero-mile levels for vehicles designed to 3.4 g/mi or 9.0 g/mi
CO standards are in the same ratio as their standards (9.0/3.4).
1. d.f.81 = a + 4.2(s)
01 a - .4(s)
2. Z9 = b - (.8)(s)
3, 23.4 = a - (.8)(s)
4. Z9/Z3t4 = 9.0/3.4
d.f.si = in-use d.f. for post-1980 vehicles (derived
as described above for the general case).
a = CO emissions at 8,000 miles for Secondary vehicles
designed to meet the 3.4 g/mi CO standard (an unknown).
-------
19
s = in-use deterioration rate (unknown).
b = CO emissions at 8,000 miles for Secondary vehicles
designed to meet the 9.0 g/mi CO standard (known from
the in-use data).
19 = Zero-mile CO emissions for Secondary vehicles
designed to meet the 9.0 g/mi CO standard (unknown).
23.4 = Zero-mile CO emissions for Secondary vehicles
designed to meet the 3.4 g/mi CO standard (unknown).
These equations can be combined given the in-use data point (b)
and the calculated d.f. (d.f.31) to arrive at a CO deteriora-
tion rate (s) for vehicles designed to meet the 3.4 g/mi CO
standard.
The following table presents the pertinent values used in calcu-
lating the HC/CO deterioration rates.
Table IV-1
Input Values to the Calculation of Deterioration Rates
HC
CD-
Volvo
In-Use
d.f.-l
0.81
0.96
Volvo
Cert .
d.f.-l
0.02
0.17
Actual
In-Use/
Cert Ratio
40.5
5.65
Ratio Used
for In-Use/
Cert.
5.65
5.65
Average
1981 Cert.
d.f.-l
0.37
0.27
Resultant
d.f.-l
2.09
1.52
Resultant
1981+ d.f.
3.09
2.52
Applying the resultant d.f.'s to the in-use data points found in
the data base as described above gives the following HC/CO
deterioration rates:
HC = 0.12 g/mi/10,000 miles
CO = 0.71 g/mi/10,000 mile's"
-------
20
c. Misfueling Category.
i. Zero-mile levels. The zero-mile levels for the Misfueling
category were obtained in a two-stage process. First, the
amount of damage caused by misfueling was quantified. This was
done by examining the results from two programs which measured a
number of vehicles' emissions before and after extensive mis-
fueling. Both of these programs were performed under contract
to EPA. One was performed by the California Air Resources Board
(CARB)[1] and the other was performed by an independent con-
tractor, Automotive Testing Laboratories Inc. (ATL).[2] These
two programs tested a total of nine Closed Loop vehicles. Three
of the nine vehicles were removed from the analysis due to con-
cerns about engine problems the car might have been experiencing
or due to the vehicle having only had a small amount of leaded
fuel run through it. For the remaining six cars, the average
emission levels before and after misfueling were examined to
determine the average percent increase in emissions due to mis-
fueling. Those average percent increases are:
HC = 364% (n=6)
CO = 128% (n=6)
These figures represent increases due to oxygen sensor damage
and other engine-out effects as well as catalyst damage.
The second stage of the procedure was to add these percent in-
creases onto the Zero-mile emission levels for the Secondary
category. Vehicles in the Primary category are also expected to
be misfueled, at the same proportional rate experienced by the
rest of the fleet, but the effects of misfueling (for HC and CO)
are assumed to be overshadowed by the effects due to experienc-
ing a Primary category failure. The rich operation assumed for
Primary category vehicles, results in essentially zero catalyst
efficiency for HC and CO by itself due to lack of oxygen in the
catalyst bed(s).
-------
21
The resulting Zero-mile levels for the MisfuRji.no oateoory are:
HC = 1.05 a/ml
CO =4.24 g/mi
ii. Deterioration rates. The HC/rn deterioration rates for thp
Misfueling category were ohtained by adoptino the deterioration
rates developed for the Secondary catenory. There is insuffi-
cient data on misfueled Closed Loop vehicles to allow a senarate
analysis of the deterioration nf these vehicles. The develnn-
ment of the deterioration rates for the .Secondary catennrv can
be found in Section TV.A.2.h.
HC = 0.1? o/mi/10.nnn miles
CO = 0.71 o/mi/10,nnn miles
3. Category Size.
The size of a category is described hy two narameters: the size at
zero miles, expressed in percent nf the unit of analysis, and the
growth rate of the category as expressed hv the nercent of thp ori-
ginal unit which "migrates" into the cateoorv oer in ,000 miles.
a. Primary. The size parameters nf the Primary catennrv were
estimated hy considering a number of separate sources of infnr-
nation. First, the incidence of vehicles from the in-nsp vehi-
cle data base which would fall 5nto the Primary catpnorv was
examined. A total of ten Primary cateoorv vehicles was found in
the fleet. The total fleet size is ?n^ vehicles after taHna
Parameter Adjustment concerns into account.* I is i no these f in-
ures results in an incidence of 5.n percent fnr the Primary
cateoory . This is at an averaoe mileane of °,l£? miles.
The second piece of evidence used to determine the size nara-
meters of the Primary cateoory came from InoWno at the f.lppt nf
162 Volvos and Saabs (described in Section TT.n.l^. These
* Note: There are a total, of ?nl cars used in the Mr
and 205 cars used for P.n since some cars were removed due to
Parameter Adjustment concerns for one nollutant hut not for the
other, as discussed previously (Spotion TT.O.I. 9n^ Was uspd a<
the average.
-------
22
vehicles employ the Lambda Sond system: a closed loop, system
with ported fuel injection and a Three-Way catalyst. One of the
distinguishing characteristics of this fleet is that it contains
a wider range of vehicle odometer mileages than the Closed Loop
vehicle data base. This fact was used to examine the growth of
Primary category problems with increasing mileage. The fleet
was divided into two groups based on mileage intervals, and the
incidence of vehicles with some malperformance of the Lambda
Sond system was noted (eg. oxygen sensor failure, defective
electronic control unit). The first group contained vehicles
with between 0 and 10,000 accumulated miles. This group con-
sisted of 64 vehicles, three of which had a Lambda Sond mal-
performance indicated. This translates to an incidence of 4.7
percent. Those vehicles with a Lambda Sond malperformance had
average CO emission levels over 400 percent higher than the
group's overall average. The average mileage of the first group
was approximately 5,800 miles. The second group was made up of
vehicles with between 10,000 and 20,000 accumulated miles. This
group consisted of 65 vehicles, five of which had a Lambda Sond
malperformance. This translates to an incidence of 7.7
percent. Those vehicles with a Lambda Sond malperformance had
average CO emission levels 390 percent higher than the group's
overall average. The average mileage of the second group was
approximately 14,400 miles.
In comparing these two groups of vehicles, the chief conclusion
to be drawn is the trend towards an increase in malperformance
of the closed loop control system with increasing mileage. The
two groups examined are of essentially the same size, the only
difference being between their average mileages. The difference
in the incidence of malperformance for the two groups is 3.0
percent.
TO arrive at the size parameters for the Primary category, the
preceding findings were combined and supplemented with technical
judgment. In applying the data from the in-use data base, a
figure of 5 percent was taken as the incidence of Primary cate-
gory vehicles at 10,000 miles. The 3% difference between the
two mileage groups of Volvos and Saabs is taken as an indication
that there will be a measurable increase in Primary category
failures with increasing mileage. Due to the small sample size
of vehicles, however, and due to differences between the
Volvo/Saab system and the variety of systems to be seen in the
post-1980 fleet, this figure does not necessarily indicate the
definitive growth rate for 1981 and later vehicles. This ana-
lysis assumes that the growth rate will be somewhat lower:
2%/10,000 miles. This is based on the assumption that due to
-------
23
improved assembly line testing, the presence of on-board diag-
nostics, and other reasons, the performance in the field of the
majority of the fleet will be better than what was seen among
the in-use Volvos and Saabs.
Combining this growth rate with the incidence of 5% at 10,000
miles resulted in the following size parameters:
Initial Size = 3%.
Growth Rate = 2%/10,000 miles.
b. Secondary. The size parameters for the Secondary category
were determined by simply taking what was left of the fleet
after establishing the sizes of the Primary category and the
Misfueling category (see below).
The size parameters for the Secondary category are:
Initial Size = 89.24%.
Growth Rate = -1.84%/10,000 miles
c. Misfueling. The size of the Misfueling category was deter-
mined by adopting the rate observed in an EPA covert observation
study performed by the Mobile Sources Enforcement Division.[3]
This study observed the fueling practices of over 22,000
catalyst-equipped vehicles in 36 states and has the largest
sample size of any study of its kind. It observed vehicles of
various model years and manufacturers. While arguments are
often put forward as to trends in the misfueling rate with
regard to vehicle age, model year, engine technology etc., none
of those trends have been substantiated or quantified to the
extent necessary to be used in an analysis of this sort. There-
fore the observed rate of 8% was adopted for this analysis as a
best estimate. The category is not assumed to grow with time.
As was mentioned earlier, however, vehicles which have a Primary
category type of failure and which are misfueled are assumed to
be best represented by the Primary category emission levels.
Thus, the overlap of the Primary and Misfueling categories is
represented by the Primary category. As the. Primary category
grows with time therefore, the percentage of cars represented by
the separate Misfueling category emission levels declines.
This growth of the overlap between the two categories, although
small, has been accounted for and is the source of the negative
growth rate below. Thus, the Misfueling category size para-
meters are:
Initial Size = 7.76%
Growth Rate = -0.16%/10,000 miles
-------
24
0
1
2
3
4
5
6
7
8
9
10
4. Unit-of-Analysis Composite.
The table below presents the composite emissions of Closed Loop vehi-
cles designed to meet standards of 3.4 g/mi CO and 0.41 g/mi HC.
These composites were arrived at by weighting together the various
categories as was described in Section III.B.5. The following tables
present the composites and illustrate how the various categories were
weighted together. The HC and CO values are presented in g/mi at
10,000 mile intervals.
3.
3.
3.
4,
4.
.74
.86
.98
.10
.22
4.34
4.46
4.58
4.70
4.82
4.94
0.03
0.05
0.07
0.09
0.11
0.13
0.15
0.17
0.19
0.21
0.23
Table IV-2
Miles E**(Pri.) S*(Pri.) E**(Sec.)
HC
KSec.)
0.23
0.35
0.47
0.59
0.71
0.83
0.95
1.07
1.19
1.31.
1.43
Emissions
S*(Sec.)
0.89
0.87
0.86
0.84
0.82
0.80
0.78
0.76
0.75
0.73
0.71
E**(Misfuel.)
1.05
1.17
1.29
1.41
1.53
1.65
1.77
1.89
2.01
2.13
2.25
S*(Misfuel.)
0.08
0.08
0.07
0.07
0.07
0.07
0.07
0.07
0.06
0.06 '
0.06
2.29
* S = Size
** E•= Emission Level
Table IV-3
CO Emissions
Miles E(Pri.) S(Pri.) E(Sec.) S(Sec.) E(Misfuel.) 5(Misfuel.) Comp
0
1
2
3
4
5
6
7
8
9
10
107.36
108.0-7
108.77
109.48
110.18
110.89
111.60
112.30
113.01
113.71
114.42
0.03
0.05
0.07
0.09
0.11
0.13
0.15
0.17
0.19
0.21
0.23
1.86
2.57
3.27
3.98
4.68
5.39
6.10
6.80
7.51
8.21
8.92
0.89
0.87
0.86
0.84
0.82
0.80
0.78
0.76
0.75
0.73
0.71
4.24
4.95
5.65
6.36
7.06
7.77
8.48
9.18
9.89
10.59
11.30
0.08
0.08
0.07
0.07
0.07
0.07
0.07
0.07
0.06
0.06
0.06
5.21
8.02
10.83
13.65
16.46
19.27
22.08
24.90
27.71
30.52
33.33
-------
25
These composites yield the following regression equations:
HC = 0.40 + (0.19)(m)
00^5.21 + (2.81)(m)
(m = miles/10,000)
These regression equations were not used in determining the final
fleet composite emissions, although very little difference results if
they are. They are presented for those interested as an aid in com-
paring the emissions performance of various units of analysis.
B. Closed Loop vehicles designed to meet 0.41 g/mi HC and
7.0 g/mi CO standards.
The reader will note that the definitions, rates and levels defined for
this unit of analysis borrow extensively from the previous unit of ana-
lysis. This is due to the fact that the only difference between the two
units of analysis is the CO design standard. This difference will effect
only certain CO emission levels. While the emission control technology
used under these two standards will differ (vehicles designed to meet the
7.0 g/mi CO standard will not use a back-up oxidation catalyst and air
pump in many cases), this difference is not expected to cause much differ-
ence in in-use emissions performance as discussed in Section I.C.I.
1. Category definitions.
The categories for this unit of analysis are defined to be the same
as those defined in Section IV.A.I.
2. Emission levels of the categories.
a. Primary Category.
i. Zero-mile levels. The zero-mile levels are the same as
those determined for the Primary category of the unit represent-
ing Closed Loop vehicles designed to meet 0.41 g/mi HC and 3.4
g/mi CO standards (Section IV.A.2.a):
HC = 3.74 g/mi
CO = 107.36 g/mi
(Note: . The CO emission level is not ratioed to the 7.0 g/mi CO
standard since the Primary category represents vehicles with a
major failure mode whose emissions would be independent of the
design standard. Similarly, the HC emission level is not ratio-
ed to take the different standards resulting from the Methane
correction factor into account, as discussed in Section II.A.)
-------
26
ii. Deterioration rates. The HC/CO deterioration rates are the
same as those determined for Closed Loop vehicles designed to
meet 0.41 g/mi HC and 3.4 g/mi CO standards (Section IV.A.
2.b.). This was done to follow the principle that similar tech-
nologies deteriorate at the same rate. This principle was wide-
ly used in the 1978 analysis (EPA-400/9-78- 005) and has been
widely used in the 1980 revision of that document for both
pre-1981 and post-1980 vehicles. The deterioration rates
arrived at are:
HC = 0.12 g/mi/10,000 miles
CO = 0.71 g/mi/10,000 mile?
b. Secondary Category.
i. Zero-mile levels. The average HC/CO emission levels and the
average mileage of the vehicles in the data base which represent
the Secondary category are:
Average HC = 0.32 g/mi (n = 191)*
Average CO = 5.47 g/mi (n = 195)*
Average Miles = 8,064 miles (n = 191)
Applying the deterioration rates (see below) to these average
emission levels to project them to zero miles and then applying
a ratio of 7.0/9.0 to the CO emission level to reflect the fact
that the vehicles in the data base were designed to meet a 9.0
g/mi CO standard and not a 7.0 g/mi standard, results in the
following Zero mile emission levels:
HC = 0.23 g/mi
CO = 3.83 g/mi
ii. Deterioration rates. The HC/CO deterioration rates are the
same as those determined for Closed Loop vehicles designed to
meet 0.41 g/mi HC and 3.4 g/mi CO standards. (Section
IV.A.2.b.). This follows the principle that similar technolo-
gies deteriorate at the same rate.
HC = 0.12 g/mi/10,000 miles
CO = 0.71 g/mi/10,000 miles
*Thesesamplesizes differ due to the fact that Parameter
Adjustment concerns eliminated vehicles from the calculations
for one pollutant, but not for the other. See Section II.A.
-------
27
c. Misfueling Category.
i. Zero-mile levels. The Zero-mile levels of the Misfueling
category for this unit of analysis (Closed Loop vehicles de-
signed to meet standards of 0.41 g/mi HC and 7.0 g/mi CO) were
developed in the same way as for the previous unit of analysis
(Section IV.A.2.C.). The figures for percent increase due to
misfueling were simply added onto the Zero-mile levels of the
Secondary category. The Zero-mile levels arrived at for the
Misfueling category are:
HC = 1.05 g/mi
CO = 8.73 g/mT
ii. Deterioration rates. The HC/CO deterioration rates for the
Misfueling category are the same as those developed for Closed
Loop vehicles designed to meet standards of 0.41 g/mi HC and 3.4
g/mi CO (Section J.V.A.2.D.). This follows the principle that
similar technologies deteriorate at the same rate. The deteri-
oration rates arrived at are:
HC = 0.12 g/mi/10,000 miles
CO = 0.71 g/mi/10,000 miles'
3. Category Size.
The category size parameters for this unit of analysis are the same
as those developed for the previous unit of analysis (Section
IV.A.3.).
a. Primary Category.
i. Initial Size = 3%.
ii. Growth Rate = 2%/10,000 miles.
b. Secondary Category.
i. Initial Size = 89.24%.
ii. Growth Rate = -1.84%/10,000 miles.
c. Misfueling Category
i. Initial Size = 7.76%.
ii. Growth Rate = -0.16%/10,000 miles.
-------
28
0
1
2
3
4
5
6
7
8
9
10
0
1
2
3
4
5
6
7
8
9
10
4. Unit-of-Analysis Composite.
Tables IV-4 and IV-5 below present the composite emissions' of Closed
Loop vehicles designed to meet standards of 0.41 g/mi HC and 7.0 g/mi
CO. The HC and CO values are presented in units of g/mi at 10,000
mile intervals.
Miles ECPri.
3.
3.
3.
4.
4,
.74
.86
.98
.10
.22
4.34
4.46
4.58
4.70
4.82
4.94
107.36
108.07
108.77
109.48
110.18
110.89
111
112
113
60
30
01
113.71
114.42
SCPri.)
0.03
0.05
0.07
0.09
0.11
.13
.15
.17
0.19
0.21
0.23
Miles ECPri.) S(Pri.)
0.03
0.05
0.07
0.09
0.11
0.13
0.15
0.17
0.19
0.21
0.23
Table IV-4
HC Emissions
ECSec.)
0.23
0.35
0.47
0.59
0.71
0.83
0.95
1.07
1.19
1.31
1.43
SCSec.)
0.89
0.87
0.86
0.84
0.82
0.80
0.78
0.76
0.75
0.73
0.71
Table IV-5
ECMisfuel.)
1.05
1.17
1.29
1.41
1.53
1.65
1.77
1.89
2.01
2.13
2.25
CO Emissions
ECSec.)
3.83
4.54
5.24
5.95
6.65
7.36
8.07
8.77
9.48
10.18
10.89
SCSec.)
0.89
0.87
0.86
0.84
0.82
0.80
0.78
0.76
0.75
0.73
0.71
ECMisfuel.)
8.73
9.44
10.14
10.85
11.55
12.26
12.97
13.67
14.38
15.08
15.79
SCMisfuel.) Comp.
0.08
0.08
0.07
0.07
0.07
0.07
0.07
0.07
0.06
0.06
0.06
0.40
0.59
0.77
0.96
1.15
1.34
1.53
1.72
1.91
2.10
2.29
SCMisfuel.) Comp.
0.08
0.08
0.07
0.07
0.07
0.07
0.07
0.07
0.06
0.06
0.06
These composites yield the following regression equations:
HC = 0.40 + CO..
7.32
10.09
12.85
15.62
18.39
21.16
23.93
26.70
29.47
32.24
35.00
CO = 7.32 + (2.77)(m7
Cm = miles/10,000)
-------
29
C. Oxidation Catalyst vehicles designed to meet
standards of 0.41 g/mi HC and 3.A g/mi CO.
1. Discussion of the analysis.
Since the units of analysis which represent Oxidation Catalyst vehi-
cles are not broken down into categories, the approach used in deter-
mining their emission rates will be presented in a much different
format. As was discussed in Section I.D., Oxidation Catalyst vehi-
cles in the post-1980 timeframe are assumed to have an emissions per-
formance similar to current Oxidation Catalyst vehicles. Oxidation
Catalyst vehicles after 1980 are not assumed to experience signifi-
cant changes in basic emission control technology. They are assumed
to rely on EGR and engine modifications for the control of NOx and to
be equipped with Oxidation catalysts and air pumps for the control of
HC and CO. This assumption largely determined what data could best
be used to predict the emissions for these vehicles. In essence, the
emission rates developed for 1980 Oxidation Catalyst vehicles from
the 1980 revision of the Mobile Source Emission Factors Document
(EPA-400/ 9-78-005) were adopted after two important factors were
taken into account: the effect of the Parameter Adjustment regula-
tions and the difference in CO standards.
First, the effect of the Parameter Adjustment regulations was
accounted for. This was done in the following three step procedure.
First, the data base assembled to analyze the emissions performance
of current technology vehicles (1975-80) was stratified to look at
only those vehicles equipped with Oxidation catalysts and air pumps,
i.e., those vehicles equipped with the control technology most simi-
lar to that which is assumed to be used in the post-1980 timeframe.
Most of these vehicles were also equipped with EGR, but the presence
of EGR was not used as a stratifying criteria since some unique
engine configurations are assumed to meet the 1.0 g/mi NOx standard
without EGR. From this subset of the data base, the average emission
levels for HC, CO and NOx were determined. In the second step,
vehicles were eliminated from the subset of the data base described
above if they showed evidence of maladjustments which can be expected
to be prevented by the Parameter Adjustment regulations. The average
emission levels for HC, CO and NOx were then calculated for these
vehicles. Comparing the average emissions before and after removing
those vehicles affected by the Parameter Adjustment regulations led
-------
30
to the calculation of a "percent benefit" figure due to the regula-
tions, (The "percent benefit" approach was taken rather than per-
forming separate regressions on the subsets of vehicles described
above, due to those subsets having relatively small sample sizes.
For example, there were only 85 vehicles in the subset used to pre-
dict the percent benefit due to the regulations in the 1982 model
year). This procedure was performed in a stepwise fashion to account
for the fact that the Parameter Adjustment regulations are applied to
idle mixture and choke in 1981 and then applied to timing in 1982.
Thus, there is a separate "percent benefit" for both of those years.
The 1981 "percent benefit" reflects the impact of limiting the
adjustability of idle mixture and choke while the 1982 "percent bene-
fit" reflects the impact of limiting the adjustability of timing.
Thus, the total impact of the regulations is staggered between 1981
and 1982. The following table presents the calculated figures of
"percent benefit" for HC and CO.
Table IV-6
Percent Benefit from Parameter Adjustment Regulations
1981 1982*
HC 16.2% 16.7%
CO 25.0% 27.3%
Finally, these figures of percent benefit were applied to the zero
mile levels and deterioration rates determined for the 1980 low alti-
tude, non-CaTTTbrnia Light Duty Vehicle (LDV) fleet. Table IV-7 pre-
sents the zero mile emission levels and deterioration rates for that
fleet before applying any benefits or modifications.
*Cummulative benefit.
-------
31
Table IV-7
1980 Light Duty~Vehicle "Emission Factors
Zero Mile Deterioration Rate
HC 0.29 0.29
CJD 6.14 2.86
(Note: The zero mile levels and deterioration rates from the 1980
LDV fleet presented above incorporate the effect of misfueling at the
same misfueling rate projected for the Closed Loop vehicles of the
post-1980 fleet).
Account must also be taken of the difference in design standards for
CO between 1980 and 1981. The 1980 Federal CO standard is 7.0 g/mi
while it is 3.4 g/mi for this unit of analysis. (The difference in
NOx standards will be discussed in a similar way in Section V.B.I.)
To account for this difference in standards, the zero-mile CO emis-
sion level for the 1980 model year, after being modified as discussed
above to account for the effect of the Parameter Adjustment regula-
tions, was ratioed by a factor of 3.4/7.0. The deterioration rate
was not factored to account for the difference in standards in order
to follow the principle that similar technologies deteriorate at the
same rate. This principle was widely used in the 1978 analysis
(EPA-400/9-78-005) and has been widely used in the 1980 revision for
both pre-1981 and post!980 vehicles. The deterioration rate was
modified to account for the effect of the Parameter Adjustment regu-
lations since those regulations represent a change in technology
which will affect the rate of in-use deterioration.
Table IV-8 presents the final HC and CO zero mile emission levels and
deterioration rates determined for Oxidation Catalyst vehicles for
1981 and 1982-and-beyond (1982+).
-------
33
D. Oxidation Catalyst vehicles designed to meet standards
of 0.41 g/mi HC and 7.0 g/mi CO.
1. Discussion of the analysis.
This unit of analysis represents Oxidation Catalyst vehicles granted
a CO waiver to 7.0 g/mi under Section 202(b)(5) of the Clean Air
Act. The different CO standard is the only difference between this
unit of analysis and the unit of analysis described in the previous
section. This difference in CO standards is not expected to result
in a significant difference in the basic nature of the emission con-
trol technology to be used. Vehicles which receive 'a CO waiver to
7.0 g/mi are expected to simply use less stringent air/fuel control
and less efficient catalysts than vehicles designed to meet the 3.4
g/mi standard. The analysis for these vehicles is therefore essen-
tially the same as that performed for the previous section. The one
exception being that the 1980 Light Duty Vehicle CO emission level
does not need to be ratioed to account for a difference in standards
since there is no difference.
The resultant zero mile emission levels and deterioration rates for
this unit of analysis are presented in Table IV-10.
Table IV-10
Oxidation Catalyst Vehicle Emission Factors (7.0 g/mi CO Standard)
1981 1982+
Zero Mile Deterioration Rate Zero Mile Deterioration Rate
HC 0.24 0.24 0.24 0.24
CO 4.61 2.15 4.47 2.08
2. unit-of-Analysis Composite.
This section presents the composite emissions of Oxidation Catalyst
vehicles designed to meet standards of 0.41 g/mi HC and 7.0 g/mi CO.
The HC and CO values are presented in units of g/mi at 10,000 mile
intervals. Table IV-11 is simply a tabulation of the equations shown
in Table IV-10.
-------
34
Table IV-11
Oxidation Catalyst Vehicle Emission Composites (7.0 g/mi CO Standard)
1981 1982+
Miles
0
1
2
3
4
5
6
7
8
9
10
HC
0.24
0.48
0.72
0.96
.20
.44
.68
.92
.16
.40
1.
1.
1,
1.
2.
2.
2.64
CO
4.61
6.76
8.91
11.06
13.21
15.36
17.51
19.66
21.81
23.96
26.11
HC
0.24
0.48
0.72
0.96
.20
.44
.68
.92
1.
1,
1,
1.
2.16
2.40
2.64
C0_
4.47
6.55
8.63
10.71
12.79
14.87
16.95
19.03
21.11
23.19
25.27
References for Section IV.
1. EPA Contract No. 68-03-2783, Work Effort No. 2 with the California Air
Resources Board.
2. EPA Contract No. 68-03-2693, Work Effort No. 4 with Automotive Testing
Laboratories, Inc.
3. EPA Internal Memorandum; August 2, 1979; from Benjamin R. Jackson, Deputy
Assistant Administrator, Mobile Source and Noise Enforcement, to all Regional
Administrators.
-------
35
V. SPECIFIC UNIT ANALYSES - NOx
A. Closed Loop vehicles designed to meet
a 1.0 g/mi NOx standard.
1. Category definitions.
There are four categories of vehicles in this unit of analysis. Each
category will be presented and briefly characterized.
a. EGR. This category is comprised of those vehicles with
inoperative EGR systems. This includes cases where the EGR sys-
tem is inoperative due to tampering, in-use failure, or manufac-
turing defect. It also includes cases of EGR failure where EGR
operation falls under computer control and is therefore subject
to computer malfunction. Vehicles were selected from the data
base to represent this category on the basis of a diagnostic
assessment of EGR function and measured NOx emissions. Vehicles
diagnosed as having an EGR problem and whose NOx emissions
exceeded their design standard were selected.
b. Misfueling. This category is comprised of vehicles which
have been misfueled, i.e. they have had extensive damage done to
their emission control system through the use of leaded rather
than unleaded gas. As will be discussed below, misfueling is
assumed to overlap with each of the other three NOx categories.
c. Low. This category is made up of vehicles which would fall
into the Primary HC/CO category and which are therefore assumed
to have lower than normal NOx levels. This assumption is based
on the fact that rich operation (as found in Primary category
vehicles) inherently implies lower NOx production in the engine,
as well as the fact that rich operation improves the catalytic
conversion of NOx in a Three-Way catalyst.
d. Secondary. As in the HC/CO analysis, this category includes
the remainder of the fleet.
2. Emission levels of the categories.
Before discussing how the emission levels of the categories were
determined, several preliminary points need to be made. First,
unlike for HC and CO, the vehicles in the Closed Loop Vehicle data
base were designed to meet several different NOx standards. The 1980
General Motors X-Body vehicles were designed to meet a 1.0 g/mi NOx
standard. The other vehicles tested in Los Angeles were 1979 models
(the Ford 351's, Celica Supra's and Audi 5000's) and were therefore
designed to meet a 1.5 g/mi standard. A third group of vehicles are
-------
36
the Ford 351's tested in Portland, Oregon. As was discussed in Sec-
tion U.S., these cars are assumed to have been designed to meet the
Federal 2.0 g/mi NOx standard. Since this analysis is concerned with
predicting NOx emissions under a 1.0 g/mi standard, the NOx emissions
from all vehicles designed under the 1.5 g/mi or 2.0 g/mi standards
were divided by 1.5 or 2.0 respectively before being used in the ana-
lysis. This was done due to the fact that NOx emission levels are
determined by the interaction of timing, air/fuel ratio, catalyst
design, and rate of EGR, all of which are balanced in order to meet
the NOx design standard. A vehicle is therefore assumed to perform
in direct relation to its design standard. The reader may note that
this same principle was used in ratioing the CO emission levels for
Secondary category vehicles to be applicable to the analysis of the
various CO standards.
The second preliminary point that needs to be made has to do specifi-
cally with the Ford 351's from Los Angeles. During the course of the
analysis it became clear that these vehicles had a unique NOx prob-
lem. Of the 73 Los Angeles 351's not in the Low category, 17 vehi-
cles were found to have an EGR problem with NOx emissions above the
standard and an additional 20 vehicles had NOx emissions significant-
ly above standard without indication of an EGR problem. It appears
that the 351 system, which controls the EGR flow rate by command of
the on-board computer, had an operating problem resulting from either
a design or manufacturing flaw. Since these vehicles represent a
substantial fraction of the data base and since the size of the prob-
lem observed with the 351's was judged to be atypical of the 1981 and
later fleet, it was decided to reweight the various categories within
the 351 fleet for data analysis purposes. The internal weighting
from the General Motors X-Body fleet (roughly the same size) was
therefore applied to the Ford 351 fleet to be able to use this data
in a representative way. The internal weighting of the X-Body fleet
(without Low category cars) was: four vehicles in the EGR category,
five cars without an EGR problem but with NOx emissions above stan-
dard and 70 cars with NOx emissions below standard. The average
emission levels from the Ford 351's grouped by these same internal
divisions were reweighted to match the weighting of the X-Body fleet.
The emission characteristics of the various categories will next be
presented.
a. EGR Category.
i. Zero-mile level. The average NOx emission level and the
average mileage of the vehicles in the fleet selected to repre-
sent the EGR category are:
-------
37
Average NOx =1.71 g/mi (n = 9)*
Average Miles = 9,510 miles (n = 9)
The deterioration rate (see below) was then applied to work
backwards to arrive at a Zero-mile level of:
NOx =1.58 g/mi
ii. Deterioration rate. The deterioration rate for the EGR
category was obtained by adopting the deterioration rate devel-
oped for the NOx Secondary category (see below). This was done
due to a lack of data describing the deterioration of EGR cate-
gory vehicles.
NOx =0.13 g/mi/10,000 miles
b. Misfueling Category.
i. Zero-mile level. As was mentioned above, the Misfueling
categoryoverlapswith the other three NOx categories. For
example, some vehicles are assumed to have an EGR problem and to
have been misfueled. As could be expected, each of the possible
overlap situations is assumed to have a unique emissions per-
formance. This might appear to be a different approach than was
used for the HC/CO analysis. For HC/CO the overlap of the'Mis-
fueling category and the Primary category was handled by assum-.
ing the emissions of the overlap vehicles were represented by
the emissions of the Primary category, rather than establishing
a new, distinct category. In that case however, Primary cate-
gory vehicles, due to the rich operation inherent to such vehi-
cles, are already experiencing essentially zero catalytic activ-
ity for HC and CO. This is due to the lack of oxygen in the
catalyst bed(s). Thus, misfueling, which primarily damages the
catalyst, will have a negligible incremental effect on Primary
HC and CO emission levels. NOx, however, is a very different
case. None of the operating modes represented by the other
three NOx categories are assumed to have a comparable effect on
NOx conversion in the catalyst. Misfueling will therefore
incrementally damage the NOx control capability of vehicles in
each of the other categories. In essence then, the Misfueling
category is made up of three sub-categories: Misfueled/EGR,
* In accordance with the explanation above regarding EGR fail-
ures among Ford 351 vehicles in Los Angeles, this sample of "9"
is in fact composed of 4 X-Body vehicles, 17 Ford 351 vehicles
from Los Angeles collectively having the weight of only 4 vehi-
cles and 1 Ford 351 from Portland.
-------
38
Misfueled/Low, Misfueled/Secondary. The emission characteris-
tics of each sub-category were developed by following a common
procedure. That procedure will first be explained below and
then applied to each sub-category.
The procedure for calculating the damage resulting from misfuel-
ing parallels the procedure adopted for HC/CO. The same fleet
of six Closed Loop vehicles described in Section IV.A.2.C was
examined to quantify the percent increase in NOx emissions due
to misfueling. The average percent increase observed was 197%
(n = 6).
This percent increase was then applied across the board to the
Zero-mile NOx levels of the EGR, Low and Secondary categories.
Table V-l presents the resultant Zero-mile levels for each of
the misfueling sub-categories.
Table V-l
Zero-mile Levels of the Misfueling Sub-categories
Sub-Category NOx Zero-mile level
Misfueled/EGR 4.69 g/mi
Misfueled/Low 0.80 g/mi
Misfueled/Secondary 1.72 g/mi
ii. Deterioration rate. The deterioration rate used for each
of the Misfueling sub-categories was obtained by adopting the
deterioration rate determined for the NOx Secondary category
(see below). This implies that vehicles with badly damaged
catalysts will deteriorate at the same rate as vehicles with
relatively undamaged catalysts. While this assumption is not
perfectly accurate, the deterioration rate for the Secondary
category was adopted due to a lack of any data describing the
deterioration of misfueled vehicles and because the stated
deterioration rate includes the deterioration due to engine wear
and other phenomenon rather than just due to catalyst deteriora-
tion. The deterioration rate arrived at is:
NOx = 0.13 g/mi/10,000 miles
c. Low Category.
i. Zero-mile level. The average NOx level and the average
mileage of the vehicles in the data base selected to represent
the Low category are:
-------
39
Average NOx = 0.40 g/mi (n = 9)
Average Miles = 9,539 miles (n = 9)
(Note: One vehicle from the GM X-Body fleet which had high HC/CO
emission levels was not used to calculate the Low category NOx
emission level due to its having an inoperative EGR system. Its
emission levels were therefore not representative of vehicles in
the Low category. Vehicles with this combination of high HC/CO
emissions yet with an inoperative EGR system will be discussed
in the next section on category sizes).
The deterioration rate (see below) was then used to work back-
wards and arrive at a zero mile level of:
NOx = 0.27 g/mi
ii. Deterioration Rate. The deterioration rate for the low
category was obtained by adopting the rate developed for the NOx
Secondary category (see below). This was done due to a lack of
data describing the deterioration of Low category vehicles. The
deterioration rate arrived at is:
NOx =0.13 g/mi/10,000 miles
d. Secondary Category.
i. Zero-mile level. The average NOx level and the average
mileage of the vehicles in the data base selected to represent
the OK category are:
Average NOx = 0.68 g/mi (n = 155)
Average Miles = 7,814 miles (n = 155)
The deterioration rate (see below) was then used to work back-
wards and arrive at a zero mile level of:
NOx = 0.58 g/mi
ii. Deterioration Rate. The deterioration rate for the NOx
Secondary category was obtained in essentially the same way as
was done for the HC/CO Secondary category (Section IV.A.2. b).
It will therefore not be discussed in depth, rather the perti-
nent numbers will be briefly presented. The NOx d.f. obtained
from the in-use 1978-79 Volvo/Saab fleet was 2.56 and the
1978-79 Volvo/ Saab Certification fleet had an average weighted
d.f. of 1.27. Thus the ratio of in-use to certification d.f.-l
values is 5.78 (1.56/.27). The average Certification d.f.-l for
vehicles being certified for the 1981modelyear was 0.16.
-------
40
Applying the Volvo/ Saab ratio to this figure and adding 1
results in a projected 1981 in-use d.f. of 1.92. Applying this
d.f. to the data point found in the data base (0.68 g/mi NOx at
7,814 miles) and converting to a deterioration rate gives a
deterioration rate of:
NOx = 0.13 g/mi/10,000 miles
3. Category Size.
The sizes of the various categories are described by two parameters:
the initial size, expressed in percent, and the growth rate of the
category as expressed in percent of the original unit which "mi-
grates" into the category per 10,000 miles. These two parameters are
developed for each of the four NOx categories.
a. EGR.
Several sources of information contributed to the development of
the size parameters of the EGR category. Each will be presented
separately, followed by a summary discussion.
i. 1980 General Motor's X-Body Fleet.
The first source of information examined was the incidence of
EGR category vehicles in the 1980 X-Body fleet. A total of 5
vehicles was found in the EGR category out of a total fleet size
of 92 vehicles This results in an incidence of 5.4% at an aver-
age mileage of 5,500 miles. It should be noted here that since
we are only looking at the incidence of a particular problem
within the fleet, the effect of the Parameter Adjustment regula-
tions need not be taken into account. This is especially true
for EGR problems, which will not be affected by the Parameter
Adjustment regulations.
ii. 1979 Ford 351 Fleet.
The second source of information examined was the incidence of
EGR category vehicles in the 1979 Ford 351 fleet. The micro-
processor used on this engine family controls the EGR flow rate
through a system of sensors and solenoids. As was previously
discussed, for an unusually large number of the vehicles tested,
the EGR system was found to be malperforming. In many cases,
there was either no voltage signal or an inadequate voltage sig-
nal coming from the microprocessor unit to the EGR-controlling
solenoid. 20 vehicles out of a total fleet size of 97 were
found to be in the EGR category. (This includes vehicles which
were removed from the calculation of NOx emission levels due to
Parameter Adjustment concerns.) This gives an incidence of
20.6% at an average mileage of approximately 11,000 miles.
-------
41
iii. Surveillance Data on Current Technology Vehicles.
The final source of information relevant to this discussion
deals with the magnitude of EGR malperformance among the current
in-use fleet. The information comes from an internal EPA ana-
lysis of data from the EPA Emission Factors Program (EFP) and
the 1978 EPA Tampering Survey.[1] The analysis examined the
initial incidence and growth rate of EGR problems. Data from
vehicles from the 1973 through 1978 model years were included in
the analysis. For purposes of this document however, only the
1977-1978 model year vehicles will be considered. (The earlier
model years show a significantly higher incidence of EGR mal-
performance). These vehicles are equipped with the most recent
EGR systems. It is assumed that these EGR systems are most
representative of the EGR systems to be employed in the
post-1980 fleet. The results of the study which are applicable
to this analysis are presented below:
Table V-2
1977-78 Model Years
-^
Initial EGR
Data Source Failure Rate Growth Rate
Emission Factors Program 2.77% 2.68%/lO,000 miles
1978 Tampering Study 5.15% Q.92%/10,000 miles
The size parameters of the EGR category were obtained through
combining the preceding sources of information with technical
judgement. The 5.4 percent incidence of EGR category vehicles
from the X-Body fleet combined with the MSED Tampering Survey
figure of 5.15 percent were combined to establish a zero mile
incidence of 5 percent for malfunctions in the EGR system it-
self. An additional 2 percent was added onto this base level to
account for the effect of computer malfunctions on computer con-
trolled EGR systems as was evidenced by the high incidence of
this failure mode among the Ford 351 fleet. The growth rate of
the EGR category was obtained by taking the average of growth
rates indicated by the Emission Factors data and the MSED Tamp-
ering Survey data. The average of 1.8%/10,QOO miles was rounded
up to 2%/ 10,000 miles. The overlap of the EGR category with
the Misfueling category was also taken into account as discussed
in the following section. Thus, the size parameters of the EGR
category are:
-------
42
Initial Size = 6.44%
Growth Rate = 1.84V10, OOQm lies
b. Misfueling. As was discussed in Section V.2.b. which pre-
sented emission levels for the NOx Misfueling category, this
category is further broken down into three sub-categories: EGR/
Misfueling, Low/Misfueling, Secondary/Misfueling. These
sub-categories represent the overlap of Misfueling with the
other categories. Several helpful points can be made before
presenting the growth rates of the individual subcategories.
First, this analysis assumes that a given portion of the fleet,
8%, is misfueled at zero-miles and remains misfueled throughout
the analysis. (See Section IV.A.3.C. for a more complete dis-
cussion on why an 8% constant rate was chosen). The analysis
also assumes that this 8% is spread proportionately throughout
the fleet. That is, the number of cars which are misfueled and
in the EGR category is directly proportional to the total number
of EGR category vehicles in the overall fleet, and likewise for
the Low and Secondary categories. It follows therefore that as
the incidence of vehicles in the various NOx categories (EGR,
Low, Secondary) changes with increasing mileage, that the
sub-categories which represent the overlap situations also
change in size in a parallel fashion. This concept can best be
illustrated by considering the 8% of the fleet which makes up
the Misfueling category as a separate fleet. The Misfueling
fleet experiences the same internal changes in its sub-category
sizes as does the total fleet with its categories, and in exact-
ly the same proportion.
The following table presents the size parameters of each of the
Misfueling sub-categories which were calculated in the manner
described above.
Table V-3
Size Parameters of the Misfueling Sub-categories
Initial Size Growth Rate
Misfueled/EGR 0.56% 0.16%/10,000 miles
Misfueled/Low 0.16% 0.1256/10,000 miles
Misfueled/Secondary 7.28% -0.28%/10,000 miles
-------
43
c. Low.
The Low category represents those vehicles with Primary category
HC/CO emission levels and a correspondingly low NOx emission
level. Thus, the size of the Low category is inherently and
directly related to the size of the Primary HC/CO category. One
consideration that needs to be taken into account here, however,
is the overlapping of the Primary HC/CO and the EGR NOx cate-
gories. That is, some vehicles can be expected to have both a
Primary category failure mode and an inoperative EGR system,
with resulting high or medium NOx levels. This phenomenon was
observed in the X-Body fleet and is expected to limit the size
of the Low category. This is especially so since it can logic-
ally be assumed that since tampering contributes to both the
Primary and EGR categories, that there will be instances in
which both the closed loop fuel system and the EGR system are
tampered concurrently. Based on these considerations and
accounting for the overlap of the Low category with the Misfuel-
ing category, the size parameters of the Low category are
assumed to be:
Initial Size = 1.84%
Growth Rate = 1.38%/10,000 miles
d. Secondary.
As with the Secondary category for the HC/CO analysis, the
Secondary category for NOx is simply assumed to represent the
remainder of the fleet. Thus, its size parameters are:
Initial Size = 83.72%
Growth Rate = -3.22%/10,000 miles
A. Unit-of-Analysis Composite.
a. This section presents the composite emissions of Closed Loop
vehicles designed to meet a 1.0 g/mi NOx standard. The NOx com-
posite values are presented in units of g/mi at 10,000 mile
intervals.
-------
Table V-4
NOx Emissions for Closed Loop Vehicles
Miles
0
1
2
3
4
5
6
7
8
9
10
E(EGR)
1.58
1.71
1.84
1.97
2.10
2.23
2.36
2.49
2.62
2.75
2.88
S(EGR)
0.06
0.08
0.10
0.12
0.14
0.16
0.17
0.19
0.21
0.23
0.25
E(Low)
0.27
0.40
0.53
0.66
0.79
0.92
1.05
1.18
1.31
1.44
1.57
S(Low) E(Secondary) S(Secondary)
0.02 0.58 0.84
0.03 0.71 0.80
0.05 0.84 0.77
0.06 0.97 0.74
0.07 1.10 0.71
0.09 1.23 0.68
0.10 1.36 0.64
0.12 1.49 0.61
0.13 1.62 0.58
0.14 1.75 0.55
0.16 1.88 0.51
This table of composite emissions
NOx = 0.74
E(M/EGR)
4.69
4.82
4.95
5.08
5.21
5.34
5.47
5.60
5.73
5.86
5.99
yields
+ (0.15)
S(M/EGR)
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.02
0.02
E(M/Low)
0.80
0.93
1.06
1.19
1.32
1.45
1.58
1.71
1.84
1.97
2.10
S(M/Low)
0.00
0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
E(M/2nd)
1.72
1.85
1.98
2.11
2.24
2.37
2.50
2.63
2.76
2.89
3.02
S(M/2nd)
0.07
0.07
0.07
0.06
0.06
0.06
0.06
0.05
0.05
0.05
0.04
Comp
0.74
0.88
1.03
1.18
1.33
1.48
1.62
1.77
1.92
2.07
2.21
a regression equation of:
(m)
(m = miles/10,000)
-------
45
B. Oxidation Catalyst vehicles designed to meet a 1.0 g/mi
NOx standard.
1. Discussion of the analysis.
The development of the emission rates for this unit of analysis
relies on the same methodology as that developed in Sections IV.C.I.
and IV.D.I. for Oxidation Catalyst vehicles meeting various HC and CO
standards. It will therefore not be presented in detail. In
essence, the zero-mile emission levels and deterioration rates for
1980 Light Duty Vehicles (as developed in the 1980 revision of the
Mobile Source Emission Factors Document (EPA-400/ 9-78-005)) were
adopted after taking into account the effect of the Parameter Adjust-
ment regulations and ratioing the zero mile emission level to account
for the different design standard.
The "percent benefit" figures for NOx attributable to the Parameter
Adjustment regulations are presented in Table 6. They were derived
in the same way as were the HC and CO "percent benefit" figures as
described in Section IV.C.I.
Table V-5
Percent Benefit from the Parameter Adjustment Regulations
1981 1982+
NOx -5,0%* -7.6%*
These figures of "percent benefit" or, more precisely, "percent
penalty" were then applied to the 1980 Light Duty Vehicle zero mile
emission level (1.56 g/mi) and deterioration rate (0.10 g/mi/10,000
mile).
The final step consisted of applying a ratio of (1.0/2.0) to the zero
mile emission level to account for the fact that the 1980 fleet is
designed to meet a 2.0 g/mi NOx standard whereas the post-1980 fleet
will be designed to meet a 1.0 g/mi NOx standard. The resultant
zero mile emission levels and deterioration rates are presented in
Table V-6.
* These figures are negative. They reflect what was observed in the
data, that removing those vehicles with maladjustments which will
most likely be prevented by the Parameter Adjustment regulations
results in an increase in average NOx levels.
-------
46
Table V-6
NOx Emission Factors for Oxidation Catalyst Vehicles
1981 1982+
Zero Mile Deterioration Rate Zero Mile Deterioration Rate
NOx 0.82 0.11 0.84 0.11
2. Unit-of-Analysis Composite.
This section presents the composite NOx emissions of Oxidation Cata-
lyst vehicles designed to meet a 1.0 g/mi NOx standard. The NOx
values are presented in units of g/mi at 10,000 mile intervals.
Table V-7 simply tabulates the equations shown in Table V-6.
Table V-7
NOx Emission Composites for Oxidation Catalyst Vehicles
Miles 1981 NOx 1982+ NOx
0 0.82 0.84
1 0.93 0.95
2 1.04 1.06
3 1.15 1.17
4 1.26 1.28
5 1.37 1.39
6 1.48 1.50
7 1.59 1.61
8 1.70 1.72
9 1.81 1.83
10 1.92 1.94
References for Section V.
1. Internal EPA memorandum; July, 1980; from David Brzezinski/David Hughes,
Inspection/Maintenance Staff to Tom Cackette, Chief, Inspection/Maintenance
Staff. "EGR Tampering/Failure Rates: Comparison of 1978 MSED and FY77 Emis-
sion Factors Surveys".
-------
47
VI. COMPOSITE EMISSION FACTORS FOR THE 1981 FEDERAL FLEET
A. Fraction of Total 1981 New Car Sales Contributed
by Each Unit of Analysis"
The effect of the CO Waiver decisions of 1979 and 1980 for Light Duty
vehicles was to allow certain engine families for some of the manufactur-
ers to certify to a 7.0 g/mi CO standard in 1981 and 1982. (Not all eng-
ine families which were waived in 1981 were also waived for 1982.) The
majority of the waived families are expected to employ Closed Loop tech-
nology in 1981 and 1982, however, several manufacturers are also expected
to certify the waived portions of their Federal fleets with Oxidation
Catalyst systems.
The model's predicted technology mix for the Federal fleet in 1981 and
1982 is 93 percent Closed Loop and 7 percent Oxidation Catalyst, based on
confidential EPA Certification sales estimates and CO Waiver testi-
mony. [1,2,3,4,5]* It is assumed that this technology split is essentially
unaffected by the waiver decisions.
It is predicted that 28 percent of the 1981 Federal fleet has been waived
to the 7.0 g/mi CO standard, based on EPA Certification sales estimates
for 1980 and the CO Waiver decisions for specific engine fami-
lies.[1,2,3,4,5] The emissions model is constructed so that the waived
segment of the fleet is apportioned among the different units of analysis
as predicted from technologies employed in 1981 Certification cars cur-
rently on record with EPA, and CO Waiver testimony.[1,2,3,A,5]
As was discussed in Section I.C.I, the portion of the Ford Motor Co. fleet
which may be produced as an open loop system employing a Three-Way cata-
lyst has been subsumed into the Closed Loop vehicle technology type.
Table VI-1 presents the percentage contribution of each unit of analysis
to the 1981 fleet.
* These references appear at the end of Section VIII.
-------
48
Table VI-1
Percentage Contribution"of the Various Uhits of Analysis
to the 1981 Federal Fleet
Technology Percent of CO Design Percent of
Type Fleet Sales Standard Fleet Sales
Closed Loop 93 3.4 67.0
7.0 26.0
Oxidation Catalyst 7 3.4 5.0
7.0 2.0
B. Composite Emissions versus Mileage.
In the emissions model, the 1981 fleet composite at each mileage point
from zero to 100,000 miles, in intervals of 10,000 miles, is calculated as
a weighted average of the emission levels of each unit of analysis.
Tables VI-2.a., VI-2.b. and VI-2.C. present the sales fractions and emis-
sion levels of each unit of analysis, and the fleet composites, at each
10,000 mile interval.
-------
Table VI-2.a.
ON
HC Emissions by
1981
Federal Fleet
Units of Analysis and Fleet Composites
Oxidation
Closed Loop
Mileage in
Miles/10,000
0
1
2
3
4
5
6
7
8
9
10
at 3.4 CO
Emission.
level in
g/mi
0.40
0.59
0.77
0.96
1.15
1.34
1.53
1.72
1.91
2.10
2.29
Sales
frac.
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
Closed Loop
at 7.0 CO
Emission
level in
g/mi
0.40
0.59
0.77
0.96
1.15
1.34
1.53
1.72
1.91
2.10
2.29
Sales
frac.
0.26
0.26
0.26
0.26
0.26
0.26
0.26
0.26
0.26
0.26
0.26
Catalyst
at 3.4
Emission
level in
g/mi
0.24
0.48
0.72
0.96
1.20
1.44
1.68
1.92
2.16
2.40
2.64
CO
Sales
frac.
0.05
0.05
0.05
0.05
0.05
0.05
0.05
•Q.Q.5
0.05
0.05
0.05
Oxidation
Catalyst
at 7.0
Emission
level in
g/mi
0.24
0.48
0.72
0.96
1.20
1.44
1.68
1.92
2.16
2.40
2.64
CO
Sales
frac.
0.02 -
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
Fleet Composite
0.39
0.58
0.77
0.96
1.16
1.35
1.54
1.73
1.93
2.12
2.31
-------
Table VI-2.b.
o
u-i
CO Emissions by
1981
Federal Fleet
Units of Analysis and Fleet Composites
Oxidation
Closed Loop
Mileage in
Miles/10,000
0
1
2
3
4
5
6
7
8
9
10
at 3.4 CO
Emission
level in
g/mi
5.21
8.02
10.83
13 .65
16.46
19.27
22.08
24.90
27.71
30.52
33.33
Sales
frac.
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
Closed Loop
at 7.0 CO
Emission
level in
g/mi
7.32
10.08
12.85
15.62
18.39
21.16
23.93
26.70
29.47
32.24
35.00
Sales
frac.
0.26
0.26
0.26
0.26
0.26
0.26
0.26
0.26
0.26
0.26
0.26
Catalyst
at 3.4
Emission
level in
g/mi
2.24
4.39
6.54
8.69
10.84
12.99
15.14
17.29
19.44
21.59
23.74
CO
Sales
frac.
0.05
0.05
0.05
0.05
0.05
0.05 •
0.05
0.05
0.05
0.05
0.05
Oxidation
Catalyst
at 7.0. C
Emission
level in
g/mi
4.61
6.76
8.91
11.06
13.21
15.36
17.51
19.66
21.81
23.96
26.11
-u
Sales
frac.
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
Fleet Composite
5.60
8.35
11.11
13.86
16.62
19.37
22.12
24.88
27.63
30.39
33.14
-------
Table VI-2.C.
1981 Federal Fleet
NOx Emissions by Units of Analysis and Fleet Composites
Closed Loop Vehicles
Mileage in
Miles/10, GOO
0
1
2
3
4
5
6
7
8
9
10
Emission
level in
g/mi
0.74
0.88
1.03
1.18
1.33
1.48
1.62
1.77
1.92
2.07
2.21
Sales
frac.
0.93
0.93
0,93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
Oxidation Catalyst Vehicles
Emission
level in
g/mi
0.82
0.93
1.04
1.15
1.26
1.37
1.48
1.59
1.70
1.81
1.92
Sales
frac.
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Fleet Composite
0.75
0.89
1.04
1.18
1.32
1.47
1.62
1.76
1.91
2.06
2.20
-------
52
C. Regression Equations.
The fleet composites presented in Section VLB need to be summarized in
linear form for the purposes of emission factors programming. However,
these emission levels do not constitute an exact linear function of the
mileage points (although they are very close to a linear function).
Therefore, these composite emissions have been linearized by regression
analysis. Table VI-3 presents these linear equations by pollutant for the
1981 federal fleet. In all the emission equations, the constant equals
the Zero-mile emission level in g/mi and the slope equals the regression-
based composite deterioration rate in g/mi/10,000 miles.
Table VI-3
Regression Equations for 1981 Federal Fleet
Pollutant Equation (q/mi)
HC 0.39 + (O.l9)(m)
CO 5.60 + (2.75)(m)
NOx 0.75 + (O.l5)(m)
m = miles/10,000
-------
53
VII. COMPOSITE EMISSION FACTORS FOR THE 1982 FEDERAL FLEET
A. Fraction of Total 1982 New Car Sales Contributed
by Each Unit of Analysis.
Some engine families which received CO waivers in 1981 received 2 year
waivers; these families amount to an estimated 10% of the 1982 Federal
fleet. No engine families received waivers for 1982 that had not previ-
ously received them for 1981. These engine families are accounted for in
the model in the same way as are the waived families for 1981. The pre-
dicted technology split is the same (93% Closed Loop, 7% Oxidation Cata-
lyst) and the waiver decisions are expected to have essentially no impact
on this split.
Table VII-1 presents the contribution of each unit of analysis to the 1982
Federal fleet.
Table VII-1
Percentage Contribution of the Various Units of Analysis
to the 1982 Federal Fleet
Technology Percent of CO Design Percent of
Type Fleet Sales Standard Fleet Sales
Closed Loop 93 3.4 83.6
7.0 9.4
Oxidation Catalyst 7 3.4 6.4
7.0 0.6
B. Composite Emissions versus Mileage.
Tables VII-2.a. through VII-2.C. present composite emissions and sales
fractions for the various units.of analysis and the 1982 Federal fleet
composite, as explained in Section VLB.
-------
Table VII-2.a.
-------
Table VII-2.b.
I/I
CO Emissions by
1982
Federal Fleet
Units of Analysis and Fleet Composites
Oxidation
Closed Loop
Mileage in
Miles/10,000
0
1
2
3
4
5
6
7
8
9
10
at 3.4 CO
Emission
level in
g/mi
5.21
8.02
10.83
13.65
16.46
19.27
22.08
24.90
27.71
30.52
33.33
Sales
frac.
0.836
0.836
0.836
0.836
0.836
0.836
0.836
0.836
0.836
0.836
0.836
Closed Loop
at 7.0 CO
Emission
level in
g/mi
7.32
10.08
12.85
15.62
18.39
21.16
23.93
26.70
29.47
32.24
35.00
Sales
frac.
0.094
0.094
0.094
0.094
0.094
0.094
0.094
0.094
0.094
0.094
0.094
Catalyst
at 3.4
Emission
level in
g/mi
2.17
4.25
6.33
8.41
10.49
12.57
14.65
16.73
18.81
20.89
22.97
CO
Sales
frac.
0.064
0.064
0.064
0.064
0.064
0.064
0.064
0.064
0.064
0.064
0.064
Oxidation
Catalyst
at 7.0
Emission
level in
g/mi
4.47
6.55
8.63
10.71
12.79
14.87
16.95
19.03
21.11
23.19
25.27
CO
Sales
frac.
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
Fleet Composite
5.21
7.97
10.72
13.48
16.24
18.99
21.75
24.51
27.27
30.02
32.78
-------
Table VII-2.C.
1982 Federal Fleet
NOx Emissions by Units of Analysis and Fleet Composites
Mileage in
Miles/10,000
0
1
2
3
4
5
6
7
8
9
10
Closed Loop Vehicles
Emission
level in
g/mi
0.74
0.88
1.03
1.18
1.33
1.48
1.62
1.77
1.92
2.07
2.21
Sales
frac.
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
Oxidation Catalyst Vehicles
Emission
level in
g/mi
0.84
0.95
1.06
1.17
1.28
1.39
1.50
1.61
1.72
1.83
1.94
Sales
frac.
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Fleet Composite
0.75
0.89
1.04
1.18
1.33
1.48
1.62
1.77
1.91
2.06
2.20
-------
57
C. Regression Equations.
Table VII-3 presents the regression equations for the 1982 Federal fleet,
as explained in Section VI.C. The CO equation is different than 1981 due
to the effects of changing CO standards for certain engine families and
due to the impact of the second step of the Parameter Adjustment regula-
tions being implemented; the HC and NOx equations are different due only
to the impact of the second step of the Parameter Adjustment regulations
being implemented.
Table VII-3
Regression Equations for th~1982 Federal Fleet
Pollutant Regression Equation (g/mi)
HC
CO
NOx
0.39 H
5.21 H
0.75 H
k (0.19)(m)
r (2.76)(m)
- (0.15)(m)
m = miles/10,000
-------
58
VIII. COMPOSITE EMISSION FACTORS FOR THE 1983 AND BEYOND FEDERAL FLEET
A. Fraction of Total 1983 New Car Sales Contributed
by Each Unit of Analysis.
Starting in 1983 the entire Federal fleet will be required to certify to
the 3.4 g/mi CO standard. The predicted technology split is the same (93%
Closed Loop, 7% Oxidation Catalyst). Table VIII-1 presents the contribu-
tion of each unit of analysis to the 1983 federal fleet.
Table VIII-1
Percentage Contribution of the Various Units of Analysis
~to the 1983 Federal Fleet
Technology Type CO Design Standard Percent of Fleet Sales
Closed Loop 3.4 93
Oxidation Catalyst 3.4 7
B. Composite Emissions versus Mileage.
Tables VIII-2.a. through VIII-2.C. present composite emissions and sales
fractions for the various units of analysis and the 1983 and Beyond Feder-
al fleet composite.
-------
Table VIII-2.3.
1983 and Beyond Federal Fleet
HC Emissions by Units
Closed Loop at 3.4 CO
Mileage in
Miles/10,000
0
1
2
3
A
5
6
7
8
9
10
Emission
level in
g/mi
0.40
0.59
0.77
0.96
1.15
1.34
1.53
1.72
1.91
2.10
2.29
Sales
frac.
.0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
of Analysis and Fleet Composites
Oxidation Catalyst at 3.4 CO
Emission
level in
g/mi
0.24
0.48
0.72
0.96
1.20
1.44
1.68
1.92
2.16
2.40
2.64
Sales
frac.
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Fleet Composite
0.39
0.58
0.77
0.96
1.16
1.35
1.54
1.73
1.93
2.12
2.31
-------
Table VIII-2.b.
1983 and Beyond Federal Fleet
CO Emissions by Uhits
Closed Loop at 3.4 CO
Mileage in
Miles/10,000
0
1
2
3
4
5
6
7
8
9
10
Emission
level in
g/mi
5.21
8.02
10.83
13.65
16.46
19.27
22.08
24.90
27.71
30.52
33.33
Sales
frac.
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
of Analysis and Fleet Composites
Oxidation Catalyst at 3.4 CO
Emission
level in
g/mi
2.17
4.25
6.33
8.41
10.49
12.57
14.65
16.73
18.81
20.89
22.97
Sales
frac.
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Fleet Composite
5.00
7.76
10.52
13.28
16.04
18.80
21.56
24.32
27.08
29.85
32.61
-------
Table VIII-2.C.
1983 and Beyond Federal Fleet
NOx Emissions by Units
Closed Loop at 3. A CO
Mileage in
Miles/10,000
0
1
2
3
A
5
6
7
8
9
10
Emission
level in
g/mi
0.7A
0.88
1.03
1.18
1.33
1.A8
1.62
1.77
1.92
2.07
2.21
Sales
frac.
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
of Analysis and
Fleet Composites
Oxidation Catalyst at 3. A CO
Emission
level in
g/mi
.84
.95
1.06
1.17
1.28
1.39
1.50
1.61
1.72
1.83
1.9A
Sales
frac.
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Fleet Composite
0.75
0.89
l.OA
1.18
1.33
1.A8
1.62
1.77
1.91
2.06
2.20
-------
62
C. Regression Equations.
Table VIII-3 presents the regression equations for the 1983 Federal fleet,
as explained in Section VI.C. The HC and NOx equations are unchanged from
the 1982 equations; the CO equation is different due to the universality
of the 3.4 g/mi CO standard for the 1983 and Beyond Federal fleet.
Table VIII-3
Regression Equations for 1983 Federal Fleet
Pollutant Regression Equation (g/mi)
HC 0.39 + (0.19)(m)
CO 5.00 + (2.76)(m)
NOx 0.75 + (O.l5)(m)
m = miles/10,000
References for Chapters VI-VIII.
1) Federal Register, Vol. AA, No. 179, 9/13/79, pp. 53376-53A08.
2) Federal Register. Vol. AA, No. 233, 12/3/79, pp. 69A16-69A62.
3) Federal Register, Vol. A5, No. 22, 1/31/80, pp. 7122-7138.
A) Federal Register, Vol. A5, No. 55, 3/19/80, pp. 1791A-17922.
5) Federal Register, Vol. A5, No. 107, 6/2/80, pp. 37360-37371.
-------
63
IX. CONPARISON TO PREVIOUS EPA EMISSION FACTOR ESTIMATES
This section simply compares the emission factor equations obtained from this
analysis with the equations developed in Appendix E to the Mobile Source Emis-
sion Factors Document (EPA-400/9-78-005). The attached figures (Figures
IX.a.-c.) compare only the 1983-and-Beyond equations from this analysis with
the Appendix E equations for HC, CO and NOx.
-------
64
3.0
2.5
«2.0
to
i-l.S
z
cc
h-
jl.O
0.5
0.0
FIGURE IX.fl,
HC
New Analysis
flPPENOJX E
20000 UOOOO 60000
MILES
80000 100000
-------
65
35.0
30.0
25.0
15.0
£10.0
5.0
0.0
FIGURE IX.B
CO •
New Analysis
flPPENDJX E
20000 UOOOO 60000
MILES
80000 100000
-------
66
3.0
2.5
~2.0
x
N.
z
oc
FIGURE IX.C,
NOX
1.0
a
a.
0.5
0.0
New Analysis
flPPENDIX E
I
20000 UOOOO 60000
MILES
80000
100000
------- |