APTD-1544
AUTOMOBILE EXHAUST
EMISSION SURVEILLANCE
A SUMMARY
I .S. KNYIRONMKNTAL PROTKCTION
Office of Air and Water Program-
Office of Mobile Source Air Pollution Control
Certification and Surveillance Division
Ann Arbor, Michigan 48105
-------
APTD-1544
AUTOMOBILE EXHAUST
EMISSION SURVEILLANCE
A SUMMARY
Prepared by
CALSPAN Corporation
Buffalo, New York
Contract No. 68-01-0435
EPA Project Officer: Charles J. Domke
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Mobile Source Air Pollution Control
Certification and Surveillance Division
Ann Arbor, Michigan 48105
May 1973
-------
The APTD (Air Pollution Technical Data) series of reports is issued by
the Office of Air Quality Planning and Standards, Office of Air and
Water Programs, Environmental Protection Agency, to report technical
data of interest to a limited number of readers. Copies of APTD reports
are available free of charge to Federal employees, current contractors
and grantees, and non-profit organizations as supplies permit - from
the Air Pollution Technical Inofrmation Center, Environmental Protection
Agency, Research Triangle Park, North Carolina 27711, or may be obtained,
for a nominal cost, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22151.
This report was furnished to the Environmental Protection Agency by CALSPAN
Corporation, Buffalo, New York, in fulfillment of Contract No. 68-01-0435.
The contents of this report are reproduced herein as received from the
contractor. The opinions, findings, and conclusions expressed are those
of the author and not necessarily those of the Environmental Protection
Agency. Mention of company or product names is not to be considered as an
endorsement by the Environmental Protection Agency.
Publication No. APTD-1544
11
-------
TABLE OF CONTENTS
Section
Title
Page No.
1 SUMMARY, CONCLUSIONS AND BACKGROUND 1
1.1 Summary 2
1.2 Conclusions .... 3
1.3 Background 6
2 RESULTS AND DISCUSSION 8
2.1 Seven-Mode Tests of 1968-1971 Model Vehicles 8
2.1/1 Results for 1968 and 1969 Model Vehicles 9
2.1.2 Results for 1970 Model Vehicles 12
2.1.3 Results for 1971 Model Vehicles 13
2.1.4 Summary of City and Vehicle Make Effects 14
2.1.5 Mileage Effects 17
2.1.5.1 Stabilized Engines 17
2.1.5.2 Low Mileage vs Stabilized Engines 19
2.2 CVS Tests of 1957-1971 Model Vehicles 20
2.2.1 Emissions Data and Results 21
2.2.2 Modal Emission Sequence 22
2.2.3 Evaporative Emissions 23
TABLES 27
FIGURES 57
APPENDIX I Emission Reduction GoalCalculations 67
APPENDIX II The Log-Normal Distribution as a Statistical Model
for Exhaust Emissions ' 68
APPENDIX III Discriminant Function Analysis of Automobile Exhaust
Emissions 72
111
-------
LIST OF TABLES
Table No. Title Page No.
1 Summary of Seven Mode Surveillance Programs by Vehicle
Model Year 8
2 Emission Levels for 1968 Model Vehicles (Kansas City] 29
3 Emission Levels for 1968 Model Vehicles (Houston) 29
4 Emission Levels for 1969 Model Vehicles (Kansas City) 30
5 Emission Levels for 1969 Model Vehicles (Houston) 30
6 Summary of Emissions (Arithmetic Means) for 1968 and 1969
Model Vehicles 31
7 Emission Levels for 1970 Model Vehicles (Kansas City) 32
8 Emission Levels for 1970 Model Vehicles (Houston) 32
9 Emission Levels for 1970 Model Vehicles (Los Angeles) 33
10 Emission Levels for 1970 Model Vehicles (Detroit) 33
11 Emission Levels for 1970 Model Vehicles (Denver) 34
12 Emission Levels for 1970 Model Vehicles (Washington) 34
13 Summary of Emissions (Arithmetic Means) for 1970 Model
Vehicles 35
14 Emission Levels for 1971 Model Vehicles (Phase I)
(Houston) 36
15 Emission Levels for 1971 Model Vehicles (Phase I)
(Los Angeles) 36
16 Emission Levels for 1971 Model Vehicles (Phase I)
(Detroit) 37
17 Emission Levels for 1971 Model Vehicles (Phase. I)
(Denver) 37
18 Emission Levels for 1971 Model Vehicles (Phase II)
(Houston) 38
19 Emission Levels for .1971 Model Vehicles (Phase II)
(Los Angeles) 38
20 Emission Levels for 1971 Model Vehicles (Phase II)
(Detroit) 39
21 Emission Levels for 1971 Model Vehicles (Phase II)
(Denver) 39
22 Summary of Emissions (Arithmetic Means) for 1971 Model
Vehicles., 40
zv
-------
LIST OF TABLES (continued)
Table No. Title Page No.
23 Summary of Seven-Mode Emissions Tests of 1968-1971 Model
Vehicles 40
24 Tabulation of Emission Levels of 1968 Model Vehicles by
Engine CID 41
25 Tabulation of Emission Levels of 1969 Model Vehicles by
Engine CID 41
26 Tabulation of Emission Levels of 1970 Model Vehicles by
Engine CID (Excluding Denver) 42
27 Tabulation of Emission Levels of 1970 Model Vehicles by
Engine CID (Denver Only) 42
28 Manufacturer/Make CID Groupings 43
29 Mileage Effects for 1968 and 1969 Model Vehicles 44
30 Mileage Effects by City for 1971 Model Vehicles 44
31 Mileage Effects by Make for 1971 Model Vehicles (Excluding
Denver) 45
32 Mileage Effects by Make for 1971 Model Vehicles (Denver
Only) 45
33 Six-City Surveillance Program--1957-1971 Model Vehicles 46
34 Composite Emission Levels as Determined by 1972 Test
Procedures (Excluding Denver) 47
35 Composite Emission Levels as Determined by 1972 Test
Procedures (Denver Only) 47
36 Composite Emission Levels as Determined by 1975 Test
Procedures (Excluding Denver) 48
37 Composite Emission Levels as Determined by 1975 Test
Procedures (Denver Only) 48
38 Emission Levels as Affected by Geographic Location--1972
and 1975 CVS Test Procedures 49
39 Emission Levels by Vehicle Weight as Determined by 1975
Test Procedure (Excluding Denver and Denver Only) 50
40 Acceleration/Deceleration Modes and Driving Sequence 51
41 Mean Emissions Data for 32 Acceleration/Deceleration Modes
(Excluding Denver) 52
42 Mean Emission Data for 32 Acceleration/Deceleration Modes
(Denver Only) 53
-------
LIST OF TABLES (continued)
Table No. Title Page No.
43 Mean Emission Data for Five Steady State Modes 54
44 Fuel Evaporative Emissions Using the Enclosure Technique.... 55
VI
-------
LIST OF FIGURES
Figure No. Title Page No.
1 Conformance of 1968 and 1969 Model Test Fleets With
Emission Standards (Cold Start, Seven-Mode, Seven-Cycle
Federal Test Procedure) 57
2 Conformance of 1970 Model Test Fleets with Emission
Standards (Cold Start, Seven-Mode, Seven-Cycle Federal
Test Procedure) 58
3 Histograms of HC and CO Emissions--1970 Model Vehicles
(All Cities Except Denver) 59
4 Histograms of HC and CO Emissions--1970 Model Vehicles
(Denver Only) 60
5 Conformance of 1971 Model Test Fleets With Emission
Standards (Cold Start, Seven-Mode, Seven-Cycle Federal
Test Procedure) 61
6 Mileage Effects (Histogram of Slopes) 1968 Model Vehicles... 62
7 Mileage Effects (Histogram of Slopes) 1969 Model Vehicles... 63
8 Mileage Effects (Histogram of Slopes) 1971 Model Vehicles
Excluding Denver 64
. 9 Mileage Effects (Histogram of Slopes) 1971 Model Vehicles
in Denver 65
10 1975 CVS Emission Levels vs. Cumulative Percentage of Test
Vehicles for Pre-Control, 1968-1969 and 1970-1971 Model
Years (All Cities Excluding Denver) 66
III-l Discriminant Function Analysis of Automobile Exhaust
Emissions 72
III-2 Maximum Likelihood Discrimination 73
II1-3 Two-Way Plot of Emissions for Four Cities (1971 Phase I
Chevrolets) CO vs. HC 75
III-4 Two-Way Plot of Emissions for Four Cities (1971 Phase I
Chevrolets) NOX vs. HC 76
II1-5 Two-Way Plot of Emissions for Four Cities (1971 Phase I
Chevrolets) NOX vs. CO 77
III-6 Discriminant Plane Plot of Emissions for Four Cities (1971
Phase I Chevrolets) 78
III-7 Distribution of Principal Factor (1971 Phase I Chevrolets)
Denver vs. Rest 79
VII
-------
1. SUMMARY, CONCLUSIONS AND BACKGROUND
The purpose of this report is to summarize information on emissions
from light-duty vehicles. State and local agencies, Federal air pollution
officials, automobile manufacturers and concerned citizens can use this report
to estimate the impact that light-duty vehicle emissions have on air quality
and to determine conformity of vehicles to the standards under which they were
certified. Contained in the report are the findings and results of three
exhaust emission surveillance programs conducted by the EPA: (1) the Great
Plains (Two-City) Surveillance Program--1968-1969 Model Year Survey; (2) the
National Surveillance Program--1970 (Six City) and 1971 (Four City) Model Year
Surveys; and (3) a Study of Emissions from Light-Duty Vehicles in Six Cities--
1957-1971 Model Year Survey. The first two programs employed the Federal Seven-
Mode Test Procedure, whereas the third program utilized the 1972 and 1975 Con-
stant Volume Sampling (CVS) Federal Test Procedure.
The Congress, through the enactment of the Clean Air Act of 1963 and
amendments thereto, provided for a national air pollution program to monitor
and control emissions from new motor vehicles. Administrative responsibility
for the air pollution control program is vested with the U.S. Environmental
Protection Agency (EPA). The first nationwide standards, together with the
testing and certification procedures, were issued in 1966 and were applicable
to 1968 model year passenger vehicles and light-duty trucks sold within the
United States. Levels for maximum allowable emissions were imposed initially
on hydrocarbon (HC) and carbon monoxide (CO) effluents only. Hydrocarbons
were restricted to 275 parts per million concentration and carbon monoxide was
restricted to 1.5 percent.* These effluents were measured using the 7-mode
cold start test procedure. More stringent standards were introduced for 1970
and 1971 model vehicles. The Federal standards based on the 7-mode procedure,
expressed in mass equivalents, were 2.2 grams/mile for HC and 23 grams/mile
for CO. In 1972, a change was made to a new test procedure. This procedure
contained a new sampling method, the Constant Volume Sampling Procedure (CVS),
and a new driving sequence. At that time the standards were again strengthened.
The first Federal standards applicable to oxides of nitrogen were promulgated
for 1973 model year light-duty vehicles.
The surveillance studies in this report were done using the test pro-
cedure in effect at the time of the study. Although correlation factors have
been developed to relate the 7-mode and CVS Federal test procedures, they are
based on average results of a sales-weighted sample of many different vehicles.
For this analysis, it is more appropriate to treat the results of tests made
by the two procedures without endeavoring to translate these results from one
procedure to the other.
These were the standards for vehicles with engines greater than 140 cubic
inches displacement. Vehicles with engines which did not exceed 100 cubic
inches displacement were restricted to 410 ppm HC and 2.3 percent CO.
-------
1.1 SUMMARY
Hydrocarbon and carbon monoxide emissions for the vehicles tested
were assessed by comparing their mean emission levels with applicable Federal
standards. Such comparison is consistent with the goals of pre-1972 regula-
tions, which were aimed at insuring that average vehicle emission levels met
standards. NOX emissions, which were not subject to control during the model
years covered by the programs, showed an increase of approximately 35 percent
from measurements on pre-1968 vehicles to measurements on post-1968 vehicles.
In general, hydrocarbon and carbon monoxide emissions showed a significant
downward trend with newer model vehicles. The most recent study, a survey of
1957 to 1971 model year vehicles based on the 1972 CVS Federal Test Procedure,
shows this downward trend in emissions. The following table displays the per-
cent of EPA's goal which has been achieved:
Percent of Emission Reduction Achieved
Model Date Percent of Goal Achieved*
Year Tested HC CO
1971 1971 99% 85%
1970 1971 84% 69%
1969 1971 58% 38%
1968 1971 54% 30%
This table is based on a comparison of actual emissions with the 1972 Federal
standards. The actual emission standards for 1968-1971 vehicles for HC and CO,
in terms of converted 7-mode standards, were less stringent. Thus the use of
ths 1972 Federal emission standards as a point of reference to evaluate the
percentage of the emission reduction goal that has been achieved for the model
years 1968 through 1971 results in a conservative estimate of improvement. In
that context, it should be noted that the imposition of stricter standards
tends to cause a step-function improvement in vehicle exhaust emissions, as
can be seen by examining the difference between 1969 and 1970 model year
vehicles.
In an effort to assess the extent to which local climate, terrain,
driving practices and other geographically differentiated factors affect
emissions, vehicles were sampled in several cities, selected to span the range
of such factors. Only small differences were observed in the emissions mea-
sured in the cities included in the survey, the only notable exception being
Denver. Significantly higher carbon monoxide and hydrocarbon emissions and
lower NOX emissions were observed in Denver than in the other cities, presum-
ably because of the effect of altitude on air-fuel ratios.
See Appendix I for additional information on the development of this table.
-------
The effect of mileage accumulation on vehicle functioning, as reflec-
ted by measurable changes in emission levels, was explored by making two or
more emission measurements on the same vehicle at different points in its
mileage-accumulation history. The effect of mileage accumulation on emission
levels was shown to be highly variable from one vehicle to another. With the
exception of Denver, when emissions increased with mileage, vehicles with
initially low mileages (less than 60 miles) exhibited a greater change in
emission levels per 1000-mile increment of mileage than did vehicles having
initial mileages in the 20,000- to 30,000-mile range.
The test programs were conducted on as-received vehicles and the
test results may reflect to some extent the influence of such variables as
state of maintenance and repair. The statistical distribution of automobile
exhaust emissions for "as received" vehicles is skewed toward the high end
of the distribution curve. If the distribution were symmetrical and fifty
percent of the vehicles met the standard, the mean of all the vehicles would
also meet the standard. This relationship does not apply, however, with a
skewed distribution. If an indication of total mean emissions is desired
with a skewed distribution, the mean emission level of a group of vehicles
must be looked at independently of the percent of these vehicles which con-
form to the standard.
1.2 CONCLUSIONS
Results of the three automobile exhaust emissions surveillance pro-
grams summarized in this report reveal the following:
1. Exhaust emission levels depend on a number of factors
peculiar to a specific vehicle, including its make and
its accumulated mileage.
2. Though statistically significant differences were observed
in the mean emission levels in different cities, Denver
was the only city showing consistent differences of
sufficient magnitude to be of engineering significance.
Carbon monoxide and hydrocarbon emissions tended to be
higher in Denver than in other cities, whereas oxides
of nitrogen tended to be lower. The observed differ-
ences are believed to be attributable to the effect of
altitude on air-fuel ratios. There was some indication
that CO and HC emissions were somewhat lower and NOX
emissions somewhat higher in Los Angeles than in the
other cities, but the differences were not as marked
as for Denver and not judged to be of practical sig-
nificance.
3. Individual vehicles of a particular category show wide
dispersion in exhaust emissions. Consequently, two.
categories of vehicles, such as those tested in two
-------
different cities, may show considerable overlap of
their statistical distributions even though the mean
emissions for the two categories are appreciably differ-
ent. Generalizations with regard to make, city or other
categories of interest, therefore, are often not appli-
cable to comparison of individual vehicles or small
subsets of vehicles drawn from the two categories.
4. In the surveillance tests of 1968 and 1969 year vehicles
in Kansas City and Houston, the following observations
were made, based upon the 7x7 Federal Test Procedure.
a. 46% of the 1968 model year vehicles tested individually
complied with applicable standards for HC, that is, 46%
of the vehicles exhibited emission levels below the
level allowable by_ the standard. Similarly, 35% were
below standards for CO, and 23% were below standards
for both HC and CO. Mean emission levels for HC, CO
and NOX were 352 ppm, 2.14%, and 1275 ppm, respectively.
b. The mean emission levels for both the 1968 and 1969
model year vehicles exceeded the applicable standards
of 275 ppm HC and 1.50% CO.
c. 53% of the 1969 model year vehicles were below appli-
cable standards for HC, 42% were below standards for CO,
and 30% were below standards for both HC and CO. Mean
emission levels for HC, CO and NOX were 303 ppm, 1.86%
and 1453 ppm, respectively.
5. In the surveillance tests of 1970 model year vehicles in Kansas
City, Houston, Los Angeles, Detroit, Denver and Washington, the
following observations were made, based upon the 7x7 Federal
Test Procedure.
a. With Denver vehicles excluded, 32% of the vehicles tested
were below applicable standards for HC, 35% were below
standards for CO, and 19% were below standards for both
HC and CO. In Denver, the corresponding percentages
were 5%, 3%, and 0.3%.
b. In general, the mean emission levels of the vehicles
tested exceeded the applicable standards of 2.2 gins/mi,
and 23.0 gins/mi for HC and CO, respectively.
c. Mean emission levels for HC, CO and NOX for all cities
except Denver were 2.83 gms/mi, 35.3 gms/mi and 4.69
gms/mi, respectively. The corresponding values for
Denver were 4.25 gms/mi, 65.1 gms/mi, and 3.83 gms/mi.
-------
6.
7.
In the surveillance tests of 1971 model year vehicles in
Houston, Los Angeles, Detroit and Denver, the following
observations were made.
a.
Low-mileage (less than 60 odometer miles) and stabilized
engines showed distinct differences in the percentage
of vehicles tested which were below applicable standards
for HC, CO and both HC and CO.
Percent Below Standards for
Total Excluding Denver:
Denver Only:
Low Mileage
Stabilized
Low Mileage
Stabilized
HC_
87
59
36
29
C0_
60
49
2
10
Both
58
38
1
6
b.
Mean emission levels for HC, CO and NOX for low mileage and
stabilized engines were as follows:
HC
Total Excluding Denver:
Denver Only:
Low Mileage
Stabilized
Low Mileage
Stabilized
1.47
2.12
2.63
3.16
Cp_
23.7
29.0
64.0
45.2
Both
3.16
3.74
2.86
3.19
c. Applicable standards of 2.2 gins/mi HC and 23.0 gms/mi CO were,
for the most part, exceeded by the higher mileage and corres-
pondingly older stabilized engines, which would most nearly
represent engines operating in vehicles in general use.
Mileage accumulation has a statistically significant effect on
exhaust emissions, but the effect varies strongly from one vehicle
to another and may be confounded with attendant factors such as
state of maintenance. In most of the cases analyzed, nearly
equal numbers of individual vehicles showed increasing and de-
creasing emission levels with mileage accumulation. On an
aggregated or average basis, however, significant trends were
established. For stabilized engines, these trends were of the
order of 0.5% to 1.0% per thousand miles, with HC and CO increas-
ing and NOX decreasing. Engines having low initial mileages were
shown to exhibit more pronounced changes with mileage, typically
5% per 1000 miles. When all cities except Denver were considered,
all three pollutants for low-mileage engines increased with
mileage. When Denver was considered separately, HC and NOX
increased but CO showed a marked decrease with mileage.
-------
8. Tests of light-duty vehicles, model years 1957 to 1971, show a
consistent downward trend in HC and CO emissions from 1957 to
1971. The trend is highlighted by the following average emission
levels based upon the 1975 CVS-CH Test Procedure for pre-control
and controlled vehicles, excluding Denver and for Denver alone.
Emission Levels (gms/mi) 1975 CVS-CH Test
Procedure
H£ C0_ N°xc
Total Excluding Denver
Pre-Control 8.74 86.5 3.54
Controlled 4.54 54.9 4.79 ,
H.7^'.C'<. -^ " ^
Denver Only
Pre-Control 10.16 126.9 1.89
Controlled 6.46 92.2 2.67
These results are indicative of the effect of emission controls
on air quality.
9. Tests of evaporative emissions using the SHED technique from
126 light-duty vehicles, model years 1957 to 1971, in Los Angeles
show approximately a 30% decrease in evaporative emissions with
the advent of California evaporative controls in 1970. Federal
evaporative emission controls were instituted with model year
1971. This trend is demonstrated by the following average
Diurnal, Hot Soak and Combined evaporative losses for pre-control
vehicles and vehicles employing evaporative controls.
Evaporative Losses Using Enclosure Technique - HC - gms/Test
Diurnal Mean Hot Soak Mean Combined Mean
Pre-Control 25.97 14.67 40.64
Controlled 16.28 10.92 27.20
In addition, evaporative losses for 1971 vehicles in Denver were
found to be approximately two to three times larger than losses
in comparable Los Angeles vehicles. The diurnal, hot soak, and
combined mean evaporative losses per test for Denver vehicles
were 47.2 gms HC, 34.77 gms HC and 81.97 gms HC, respectively.
1.3 BACKGROUND
Under the Clean Air Act, manufacturers are required to submit appli-
cations containing data gathered during both phases of a two-part test program
in order to qualify for certificates of conformity. For model years 1968
-------
through 1971, the first phase of testing provides data on exhaust emissions
which show the performance of the control equipment after the engine has been
broken in, but before substantial mileage has been accumulated. These data
are known as 4,000 mile emission data. The second phase of the test program
provides data on the durability of the emission control system. These data
are known as 50,000 mile durability data. For the vehicle model years treated
in this report, compliance was demonstrated whenever the mean emission level
from a specified sample of emission-data prototypes of each engine displacement,
according to projected sales volume, is within the applicable standard.* This
mean incorporates a deterioration factor determined from a sample of durability-
data prototypes representative of at least 70% of the manufacturer's engine
displacement/transmission options. Test and certification procedures, together
with the appropriate standards, are published in the Federal Register. Inherent
in the method of certification is the fact that mean values for HC or CO near
the standard make it possible for 50% of certification or in-use vehicles to
be above the standard for either pollutant. If both pollutants are considered
together, it is likely that more than 50% of the vehicles will be above stan-
dards. Consequently, the percentage of individual vehicles with both HC and
CO emission levels simultaneously within standards may give an unduly pessi-
mistic picture of the degree of compliance with standards.
EPA has recognized that a realistic assessment of the effectiveness
of Federal air pollution regulations requires the measurement of emissions from
production vehicles in the hands of the motoring public. Accordingly, a series
of exhaust emission surveillance programs has been administered by the EPA dur-
ing the past several years to obtain such definitive information. Test fleets
of consumer-owned vehicles within various major cities were selected by make,
model, engine size, transmission, carburetor and compression ratio in such
proportion as to be representative of the normal production vehicles sold (or
projected to be sold) for that model year in the United States.
The principal objectives of such surveillance programs have been to
establish degree of conformity of in-use production vehicles with certification
levels and to assess the effects on emissions levels resulting from the test
locale (i.e., the influence of climate, topography and urban development),
vehicle mileage accumulation and vehicle make/model/engine differences. (The
National Source Inventory Section (NSIS) and the Office of Land Use Planning
(OLUP) have been closely involved with these efforts.)
For 1972 and subsequent model-year vehicles, every vehicle tested ir, the
certification sample must exhibit emissions below the applicable standard
level.
-------
2. RESULTS AND DISCUSSION
This section of the report is a discussion of the findings and the
results of three surveillance programs which in the aggregate represent tests
of over 5500 light-duty vehicles. These three programs are: (1) the Great
Plains (Two-City) Surveillance Program--1968-1969 Model Survey, (2) the National
Surveillance Program--1970 [Six-City) and 1971 (Four-City) Model Year Surveys,
and (3) "A Study of Emissions from Light-Duty Vehicles in Six Cities" (1957-
1971 model vehicles). The emissions data collected in each of the first two
programs cited above are susceptible to comparison and correlation as they all
were measured using the seven-mode cold-start Federal Test Procedure and NDIR
(non-dispersive infrared) instrumentation for effluent measurement. The last-
cited program made use of both the 1972 (CVS-C) and 1975 (CVS-CH) Federal Test
Procedures and utilized different instrumentation for measurement of HC (a
flame ionization detector, FID) and NOX (a chemiluminescent detector). A
summary and overview of the procedures, sample size, test locations and objec-
tives of these several programs are presented in Tables 1 and 33.
Although numerical results appearing herein have been drawn liberally
from contractors' final reports for each of these programs, additional and
different types of data analyses have been performed to provide additional per-
spectives and insights into the results and, in some cases, to broaden or
refine the application of statistical methodology. In the formulation of
conclusions, caution has been exercised to consider inherent sampling biases,
confounding factors and other limitations that are associated with the data
base. The results of data analysis have been interpreted pragmatically from
the standpoint of the magnitudes of the emissions measurements as well as from
the standpoint of the statistical significance levels associated with the
results.
2.1 SEVEN-MODE TESTS OF 1968-1971 MODEL VEHICLES
This section of the report consolidates the data from two different
surveillance programs which employed identical test procedures and analytical
instrumentation in the collection of exhaust emission data. Composite exhaust
emissions data taken from each vehicle tested were based on the seven-mode,
cold-start test procedure as stipulated by the Federal Register appropriate to
the model year involved. A brief descriptive summary characterizing these two
programs is given in Table 1. It is seen that the project objectives have been
directed to the evaluation of those factors that impinge most vitally on the
ultimate success of realizing the goals of the national clean air program: the
extent to which vehicles on the road meet emission standards, the atmospheric
loading created by vehicle emissions, and the possible effects on emissions of
vehicle mileage and geographic differences.
In all cases, an attempt was made to select the test fleets from the
motoring public according to a strategy that yielded a sample-as nearly repre-
sentative as possible of the vehicles that were in use (or, in the case of new
-------
models, projected to be in use). Vehicle categories were identified and selected
by manufacturer and engine cubic-inch displacement (CID). In the case of the large
manufacturers, such as General Motors, engines employed by different, makes may
be of different design despite sharing the same numerical value of displacement.
In such cases, identification was by manufacturer, make and displacement. In
the case of the 1968 and 1969 models, a total of 27 different CID engines was
tested whereas the number increased to 44 for the 1971 models despite a decrease
in size of the total test fleet from approximately 2,000 to 370 vehicles.
Because of the effect of such limitations on the amount of data available in
each vehicle category, emissions data for the most part in this report are
summarized by manufacturer and make only. Where sufficient data exists,
however, city and CID tabulations are presented.
The Federal Test Procedures applicable to the model years 1968 through
1971 remained substantially unchanged despite the fact that the certification
standards and their dimensional units were revised. For 1968 and 1969, the
standards for emissions were stated on a concentration basis, but in 1970 a
change was made to the more realistic mass basis (grams/mile). Thus, for the
most part, emissions data for 1968 and 1969 are presented in this report on a
concentration basis, whereas data for 1970 and 1971 are presented on a mass
basis. It is to be noted, however, that the change in basis is primarily one
of data processing since the basic direct measurement was, and still is, in
terms of concentration. A functional relationship was developed to calculate
the vehicle exhaust volume per mile with vehicle inertia weight and trans-
mission type as independent variables. By means of this relation, grams-per-mile
values are calculable to afford a comparison of 1968, 1969, 1970 and 1971 model
vehicles on a common basis. Although there were no standards for the oxides of
nitrogen for these model years, the NOX data were nonetheless measured and are
included.
2.1.1 Results for 1968 and 1969 Model Vehicles
Exhaust emissions data for each of the three principal effluents
determined from measurements on 1968-1969 model vehicles are shown by city in
Tables 2 through 5. These data consolidate results from the three phases of
this program during which time each vehicle, on the average, accumulated approx-
imately 15,000 miles. Included in these tables are vehicles produced by American
Motors (AMC), Chrysler (Chry), Ford (FoMoCo), General Motors (GM) and Volkswagen
(VW) representing, in total, a statistical sampling of 90% of the sales for
those model years.
The results are expressed as concentrations and should be assessed
against the certification standards applicable to these models (275 ppm HC and
1.50% CO, except for VW, for which the standards are 410 ppm HC and 2.30% CO).
Oxides of nitrogen are expressed as NOXC, the subscript c denoting correction
for humidity. For each of the vehicle categories listed, the tables give N,
the number of tests, the mean mileage in thousands of miles, and emissions data
in three different forms. First, results are expressed in terms of the percent
of vehicles having emissions below the applicable standards, either for HC, for
-------
CO, or for both HC and CO. Next, results are expressed in terms of arithmetic
mean and standard deviation and, finally, in terms of geometric mean and
standard deviation. The rationale for this form of presentation is presented
below and is elaborated upon in Appendix II.
The frequency distribution of automobile exhaust emissions is governed
by constraints which make it unlikely that the emission measurements will follow
a normal (Gaussian) distribution. In particular, emission concentrations are
necessarily non-negative and are therefore strictly bounded on the low end of
the distribution. At the high end of the distribution, however, emissions are
not subject to such an arbitrary constraint. This fact, and the fact that errors
of measurement tend to be proportional to the concentration being measured, com-
bine to cause the frequency distribution of exhaust measurements to be skewed
toward the high side of the range of emission values. Furthermore, experience
as well as theoretical statistical arguments suggest that the frequency distri-
bution of exhaust emissions is essentially log-normal. In other words, if the
logarithms of the emission quantities are used to compile a frequency distribution
or histogram, the resulting distribution tends to be symmetric and is approximated
by a normal distribution with appropriate mean and standard deviation. These
quantities, computed in logarithmic units, can be transformed back to anti-
logarithms, but the transformed values are not to be confused with the mean
and standard deviation computed from the original data as expressed in percent
concentration, parts per million or grams per mile. Mean values computed from
logarithmically transformed data represent geometric means, whereas mean values
computed from the original data represent arithmetic means.
A word of explanation is in order with regard to the geometric mean
and standard deviation and their interpretation in an emissions context. The
geometric mean is in units of concentration or grains/mile, depending on whether
the analysis deals with measurements expressed in concentration (as would be
the case for 7-mode tests) or mass (as would be the case for CVS tests). The
geometric standard deviation, however, is in reality a ratio and is accordingly
dimensionless. If the geometric mean is multiplied by the geometric standard
deviation, one obtains a quantity which represents approximately the 84th per-
centile of the distribution, in much the same way as one obtains this percent!le
in a normal distribution by adding the standard deviation to the mean. Similarly,
by multiplying the geometric mean by the geometric standard deviation squared,
one obtains approximately the 95th percentile of the distribution in much the
same way as one obtains this percentile in a normal distribution by adding two
standard deviations to the mean. In short, normal distribution probabilities
can be employed directly in the statistical analysis of automobile exhaust
emissions provided the geometric mean and standard deviation are used multi-
plicatively rather than additively.
The interpretation of automobile exhaust emission data and the con-
clusions drawn-from these data depend strongly on the nature of the statistical
analysis to which the data are subjected. Consequently, considerable care must
be taken to prevent conclusions from being artifacts of the analysis and inappro-
priate to the purposes of the investigation. For example, consider the insert,
which depicts two hypothetical distributions of hydrocarbon emissions in parts
10
-------
per million (ppm). Distribution A has a lower geometric mean than Distribution
B, but a greater fraction of the automobiles in Distribution B conforms to the
standard than do those in Distribution A. The apparent anomaly occurs because
of the considerably greater dispersion or scatter of emission measurements in
Distribution A as compared with Distribution B. Thus, if the fraction of
vehicles conforming to standards is the criterion of comparison, a different
result is obtained than if attention is focused only on the difference between
geometric means. For many purposes it is essential to take into consideration
the entire statistical distribution of results rather than a single numerical
index computed from that distribution.
20«-
15
I
Ul
o
U)
ec
"-
I
ui
PC
10
m = ZOO
a = 1.28
10% EXCEED STANDARD
m= ISO
a » 2.53
25% EXCEED STANDARD
50 100 150 200 250 300 350 400
HYDROCARBONS (PARTS PER MILLION)
450
500
550
600
HYPOTHETICAL EMISSION DISTRIBUTIONS
Finally, a note of caution must be injected concerning the purposes
for which the arithmetic means and the geometric means are applicable. Inasmuch
as the model distribution for emission measurements is taken to be log-normal,
the mean and standard deviation of the log-transformed data for a particular
vehicle category are estimates of the population parameters for that category.
If these parameter estimates are known, the entire distribution can be approxi-
mated in a straightforward way. It must be pointed out, however, that the
geometric mean is not appropriate for assessing the total contribution of a
particular make of vehicle to the pollution inventory. For this purpose, the
arithmetic mean is required. For example, the arithmetic mean, expressed in
11
-------
grams per mile, can be multiplied by an estimate of the total mileage for a
given class of vehicles to obtain an estimate of the "impact" of that class of
vehicles on air quality. The geometric mean, on the other hand, would clearly
be inappropriate for such estimation. It is for these reasons that the emission
data contained in this report are presented in several different ways.
Figure 1 presents a graphic comparison of the aggregated emissions
data for Kansas City and Houston in terms of the percent of the test fleet which
exhibited emission levels below the standards set for CO and HC. It is evident
from this figure that there is little difference between the two cities in this
respect. This conclusion is based more on the magnitude of the differences ob-
served than on the statistical levels of significance involved. For example,
suppose the percent below standard level is approximately 50%. Then two groups
of cars can be adjudged significantly different at the 95% level of significance
if the observed difference in fraction below standard is greater than about
1.4/ V~N", where N is the number of cars in each of the test categories. Quite
clearly, if N is sufficiently large, the minimum difference required to reject
the hypothesis of equal performance becomes very small. When as many as 500
vehicles in each category is involved, as it is in this case, the minimum differ-
ence is about 0.06. Strictly speaking, therefore, the difference between Kansas
City and Houston for HC emissions of 1969 model vehicles exceeds this value and
might, therefore, be regarded as significantly different. As a matter of prac-
tical significance, however, this difference is hardly large enough to be of
much engineering concern. A similar conclusion is reached if one compares the
mean emissions levels for the two cities (see Table 6).
A note of caution should be voiced concerning excessive reliance
upon percent conforming to standards. Due to the log-normal nature of emissions
data, a situation can easily arise where 50% of the vehicles tested meet the
standards but the mean value of all cars tested exceeds the applicable stan-
dard. Comparison of Table 6 and Figure 1 reveals that such is the case for
the 1968 and 1969 vehicles tested. Hence, percent conforming figures present
a picture of emissions levels for 1968 and 1969 vehicles which must be tem-
pered with the actual mean emission levels.
2.1.2 Results for 1970 Model Vehicles
Exhaust emissions data for each of the three principal effluents
determined from measurements on 1970 model vehicles tested in six cities are
shown in Tables 7 through 12. As before, the data are reported in terms of
percent of vehicles below standards for CO, HC or both as well as in terms
of means and standard deviations on both the arithmetic and geometric basis.
Figure 2 presents a graphic comparison of the six cities in terms of
the percent of the test fleet which exhibited emission levels below the HC and
CO standards. It is evident from this figure that the city of Denver departs
markedly from the other five cities in this respect.
12
-------
This observation is further substantiated by Table 13, in which the
mean emission levels for the six cities are summarized. Again, only Denver
stands out as distinct from the other cities. Finally, Figure 3 and Figure 4
present a graphical interpretation of both the Denver effect and the essential
log-normal nature of emissions data. As is evident from comparison of both the
HC and CO histograms for all cities except Denver and for Denver only, not only
is the mean emission level higher in Denver but a substantially greater number
of vehicles fail to pass emissions standards. In addition, histograms of the
logarithms of the data are very nearly Gaussian distributions substantiating
the log-normal nature of the data set. When Denver is considered alone, a
smaller number of vehicles is available for analysis than in the case in which
the data from several cities are pooled. This reduction in sample size results
in a greater variation from a true Gaussian distribution. It should be reit-
erated at this point that conclusions reached from percent conforming figures
should be tempered with actual mean emission levels. Even from this viewpoint,
however, Denver's unique nature stands out, as is well evidenced by Figure 2
and Table 13.
2a.3 Results for 1971 Model Vehicles
Exhaust emissions data for each of the three principal effluents
determined from measurements on 1971 model vehicles tested in four cities are
shown in Tables 14 through 21. Tables 14 through 17, denoted Phase 1, summar-
ize exhaust emission data for 1971 model vehicles tested in low mileage condition
(less than 60 odometer miles). Tables 18 through 21, denoted Phase 2, summarize
exhaust emissions data for "stabilized" engines after accumulation of approxi-
mately 5000 miles. Again, the data are reported in terms of percent of vehicles
below standards for CO, HC or both as well as in terms of means and standard
deviations on both the arithmetic and geometric basis.
Figure 5 presents a graphic comparison of the four cities in terms
of the percent of the test fleet which exhibited emission levels below the HC
and CO standards. Again, it is evident that Denver departs markedly from the
other three cities, especially with regard to the low mileage vehicles. A
similar conclusion can be drawn from Table 22, in which mean emissions for the
vehicles tested in the four cities are compared for both low mileage and stab-
ilized engines. The entire tes,t fleet consisted of only 369 vehicles. Fol-
lowing mileage accumulation (4,000 to 7,000 miles), the same vehicles were
retested on the presumption that the engines had stabilized. The principal
effect of this stabilization on emissions is believed to be the result of
"wearing-in" or "seating" of valves and piston rings as well as a stabilization
of combustion chamber deposits.
In order to bring greater perceptiveness to bear on the nature of
apparent differences between geographic locales and between makes and models,
plots of emissions in various data subsets were employed. A hypothetical plot
13
-------
of this type is shown in the insert. The three emittants are plotted along
three coordinate axes, and the data for the two cities show up as two clouds
of points in three-dimensional space. Plots of this type were found to be very
useful in tempering conclusions drawn from statistical analyses based on means
and standard deviations of data subsets. For example, it is very unlikely that
two sets of data will be found without considerable intermingling of points,
even though the means or "centers of gravity" of the two clouds are significantly
different. By the use of discriminant function analysis, as presented in Appen-
dix III, a derived plot can be evolved which maximizes the separation of data
clouds and minimizes their overlap or interpenetration. It was on the basis
of such plots, together with a consideration of the magnitude of differences
among the means for the several cities, that Denver was isolated as the only
city deserving especial note.
X CITY A
O CITYB
2.1.4
CO
Summary of City and Vehicle Make Effects
It is often hazardous to ascribe differences observed in test results
to one particular effect or another, inasmuch as several factors, inextricably
involved, may be operating simultaneously. For example, the apparent difference
in vehicle emissions noted for two cities may be confounded by the fact that
there is a difference in the mean mileage levels between the two test fleets.
There is also the possibility that the test fleets in the two cities are not
statistically identical despite the fact that random selection techniques were
used. Ideally, the same set of vehicles should be tested in different cities in
order to measure city effects unconfounded with extraneous factors.
Another factor of import is the need to recognize the practical
consequences of differences or effects determined by the application of statis-
tical techniques. With a sufficient quantity of data (statistical degrees of
freedom), a finding which is highly significant statistically can be ascertained,
14
-------
but this finding may be of little practical importance to the air pollution
problem. This point has been previously alluded to but is of sufficient impor-
tance to bear further elaboration.
Consider, for example, two categories of vehicles for which the
arithmetic means of their hydrocarbon emissions are 200 ppm and 225 ppm. Assume
that the standard deviation of emission measurements, as computed from measure-
ments on individual vehicles, is 60 ppm. The standard error of the difference
between means is approximately 85 ppm/ V N, where N is the number of vehicles
tested in each category. By standard statistical argument, a difference between
the observed means of the two categories is adjudged significant at the 5%
significance level if the observed difference is approximately twice as large
as the standard error of the difference. In the present instance, this event
occurs if 170 ppm / V N is less than 25 ppm, the observed difference between
the two vehicle categories. Quite clearly, therefore, whether the difference
is declared statistically significant depends on the number of vehicles tested.
If N is greater than about 46, the observed difference of 25 ppm would be
declared to be "statistically significant", whereas it would not be regarded
as significant if N is smaller than 46, say 35.
It is clear, therefore, that by pooling a sufficient amount of data
it is possible to label as statistically significant an effect which may be of
negligible engineering magnitude. More germane is the consideration that if
the difference between two means is no greater than--say--10% of their pooled
mean, it may be of little consequence, for purposes of air quality assessment,
to regard the two quantities as distinct. Moreover, the importance of an
effect does not depend on the number of vehicles tested, and the mere act of
declaring the effect to be statistically significant in no way augments its
practical magnitude.
It is this philosophy which governs many of the decisions made in this
report, particularly as they apply to the pooling of data from various source
categories, such as makes, models and cities. For example, it is this type of
consideration which draws special attention to Denver and causes it to be
treated separately from other cities in the statistical analysis. Though
statistically significant differences in mean emission levels can be established
among at least some of the other cities, these differences are not large enough
to merit separate statistical treatment. For this reason, it is regarded logical
to pool results for all cities other than Denver. Moreover, a physical reason,
namely altitude, can be advanced for the "Denver effect" and is believed to be
a substantive constraint affecting the compliance of that city with emission
standards. Though Los Angeles shows some tendency to lower emissions than the
other cities, this difference is not of sufficient magnitude to affect the
pooled average appreciably. Accordingly, it was considered legitimate to pool
Los Angeles with the other cities, Denver excluded. The ability to pool the
data, however, does not preclude the desirability of sampling from several
cities in the interest of obtaining a more representative and credible cross
section of in-use vehicles.
15
-------
Table 23 summarizes the salient features of the seven-mode tests of.
1968-1971 model vehicles. City effects on emissions were observed only for 1970
and 1971 model year vehicles, since a wider segment of the geographical area
of the United States was included in the surveillance program associated with
these models. Although a number of statistically significant effects can be
shown to occur in one city or another or for one effluent or another on an
isolated basis, only two consistent effects associated with cities predominate.
The first of these is a strong effect associated with Denver, which
in all test data for all vehicles yielded a significantly higher value for HC
and CO emissions than was observed in the other cities and, as might be expected,
a tendency toward lower NOXC. The principal consideration distinguishing Den-
ver is its altitude, which affects carburetion and tends to produce an excess-
ively rich fuel mixture. The second effect is that associated with Los Angeles.
Emission data for this city in general show HC and CO emission levels somewhat
lower than for the other cities tested, and NOXC emissions somewhat higher than
in some other cities with relatively high HC and CO levels. It should be noted
that in addition to small differences in mean emission levels, discriminant
plane analysis (Appendix II) and plots of the emission data were used to deter-
mine this small Los Angeles effect. As has been pointed out, however, the
Los Angeles effect was not considered strong enough to justify special treat-
ment of the data.
City effects associated with NOX are greatly reduced when the humidity
correction is applied to the data to obtain the NOXC levels as used in this
report. This fact substantiates the effectiveness of the humidity correction
in minimizing data variations associated with tests conducted under different
ambient conditions.
Tables 24 and 25 summarize emissions for 1968 and 1969 model year
vehicles according to engine CID for all cases in which at least thirty units
were available for averaging. The averages are taken over the two cities in
which tests were conducted, namely Kansas City and Houston. These tables show
that appreciable differences occur among engines but that these differences
are not solely related to size but rather to design. Both low and high emitters
can be found among both small and large CID engines. In addition, generaliza-
tions are further complicated by the fact that NOX emissions often tend to be
inversely related to CO and HC emissions, so that it is difficult to make broad
statements with regard to total emissions.
Data were also available for comparison of 1970 model vehicles by
engine size. Table 26 presents those cases in which a sufficient number of
vehicles was available to justify comparisons and represents aggregation of
data from Houston, Los Angeles, and Detroit. Table 27 presents similar data
for Denver. In this table, however, it was necessary to reduce the minimum
number of test units to 10 rather than 30 in order to obtain averages for com-
parison of engine size.
Comparison of emissions according to vehicle make involved the pooling
of many sizes of engines within the make, as shown in Table 28. In view of
16
-------
this fact, apparent differences among makes must be viewed with caution, and any
attempt to rank makes in a strictly monotonic sequence should be resisted. This
caution stems largely from the interrelation among statistical significance,
number of vehicles tested, and variability of emission measurements, as pre-
viously noted.
As an illustration of this point, consider the following example,
drawn from 1971 model year vehicles as tested in Phase 1 in Denver (Table 17).
HC Emissions (gms/mi)
Make N Mean Std. Dev.
Cadillac 3 0.67 0.18
Pontiac 5 1.82 0.17
Buick 9 2.08 0.99
Oldsmobile 6 2.13 0.91
Chevrolet 34 2.37 0.55
Volkswagen 23 2.43 0.64
Plymouth 14 2.99 0.72
Ford 24 3.12 1.19
American 20 3.12 1.27
Dodge 3 4.33 1.69
The difference between Pontiac and Chevrolet (0.55 gms/mi) is statistically
significant. However, the difference between Pontiac and Dodge (2.51 gms/mi),
although of greater apparent magnitude, is not adjudged to be statistically
significant because of the small number of Dodge vehicles tested and the large
standard deviation which they exhibit. Thus it is clear that unless ranking is
tempered with much statistical and engineering judgment, spurious conclusions
may result.
2.1.5 Mileage Effects
The term "mileage effects" when used with reference to emissions may
have several connotations and may therefore be subject to misinterpretation.
Ideally an investigation of "mileage effects" should only be concerned with the
deterioration of emission control performance with increasing mileage. Practi-
cally, however, a variety of factors such as the state of engine adjustment or
repair hinder any attempt to isolate this fundamental mileage effect. Emittant
levels were established for each vehicle in an "as received" condition regardless
of its operating condition. Consequently, in the context of this report, mile-
age effects will be used in its broadest sense to describe trends which become
increasingly prominent as the vehicle accumulates mileage or receives inadequate
maintenance.
2.1.5.1 Stabilized Engines
One of the principal purposes of the seven-mode test program on 1968-
1969 vehicles in Houston and Kansas City was to ascertain the presence and mag-
17
-------
nitude of a mileage effect by performing three emissions measurements on each
vehicle candidate: an initial test (Phase I), a subsequent test after approxi-
mately 4,000 miles (Phase II), and a final test after an additional 4,000 miles
(Phase III). The initial tests were performed on well-stabilized engines, the
mean mileage for all vehicles tested during Phase I being approximately 28,000
miles for 1968 and 17,000 miles for 1969 model vehicles.
Two distinct approaches to analysis of the mileage effect are possible.
The first approach consists basically of pooling all the emissions data and per-
forming a linear regression vs mileage. Data from over 1600 applicable emission
tests were processed and these pooled data showed a general trend of increasing
HC and CO (approximately 0.5 to 1.0% per 1000 miles) with a similarly decreasing
trend in NOX.
Such an approach based upon pooled data, however, can frequently mask
several important facets of the "mileage effect", since the analysis provides
only an overall result and cannot yield the underlying distribution derived
from each individual vehicle. In addition, many vehicles could not be retrieved
for Phase II or Phase III and hence could not contribute any information with
regard to change of emissions with mileage. Though it can be argued that differ-
ent vehicles tested at different mileages make possible the inference of mileage
trends, these trends are confounded with numerous other factors which differ-
entiate one vehicle from another, such as maintenance and driving habits. To
avoid these problems, a new analysis was undertaken which performed an emissions
vs mileage regression for each candidate vehicle which survived more than one
test. This set includes vehicles which were available during all three phases,
Phase I and II or Phase I and III. The resultant trends (slopes) were then
averaged and the result tested using a standard t-test at both 95% and 99% sig-
nificance levels to determine if the average slope was significantly different
from zero. In addition, a histogram of the individual vehicle slopes was plotted
to gain more insight into the nature of the mileage effect process. A schematic
diagram of the general technique used is presented in the insert for a hypo-
thetical case.
FLOW DIAGRAM FOR COMPUTATION OF MILEAGE EFFECTS
18
-------
A summary of the significant results for 1968 and 1969 data is pre-
sented in Table 29. Examination of this table reveals a mileage effect of
comparable magnitude to that found using a pooled vehicle technique and includ-
ing non-survivors. A major advantage of the individual surviving vehicle tech-
nique rests in the information provided by the histograms of the individual
slopes. Figures 6 and 7 illustrate typical distributions of slopes character-
istic of 1968 and 1969 data. It should be noted that data for these years
represent well stabilized engines with accumulated mileage over 20,000 miles
and therefore would not be expected to exhibit dramatic mileage effects. For
these particular data sets, the distributions appear almost equally spaced on
both sides of zero. This phenomenon is demonstrated in Table 29 for the case
of HC in 1969. Due to the large sample size of 318 vehicles, the standard
t-test provides a 95% confidence level that the mean mileage effect is differ-
ent from zero. In this particular case, HC deteriorates 0.98% per 1000 miles.
The percentage is computed by dividing the slope (emission increment per 1000
miles) by the extrapolated emission levels at 4000 miles. This mileage point
was chosen as a base since it corresponds to a new but stabilized engine.
Several interpretations of this result are possible. If one is
primarily interested in impact studies in which large numbers of vehicles are
being considered and mean values are important, this increase in HC is most
probably significant. On the other hand, little can be said about the per-
formance of an individual vehicle due to the wide dispersion exhibited by the
slopes. A similar phenomenon exists in virtually all other data sets in which
a mileage effect can be established. No doubt factors such as maintenance
practice, grade of fuel, and other unknowns confound the mileage phenomenon.
For example, a tune-up immediately before testing in Phase II or Phase III
could override or mask normal mileage deterioration. In summation, although
a statistically significant mileage effect can be shown in many data sets,
care should be exercised in the interpretation and use of this data.
2.1.5.2 Low Mileage Versus Stabilized Engines
To explore the possibility that mileage effects on emissions may
exhibit different characteristics at different stages of the mileage accumula-
tion life of a vehicle, emissions data were collected on a fleet of 1971 models
when new (less than 60 miles),and on the same vehicles after they had gone at
least 4,000 miles but less than 7,000 miles. Emissions data for each of 369
vehicles were regressed against mileage and the individual emissions vs mile-
age trends were pooled in the manner described above. These results are
presented in Table 30 for each of the four cities investigated and for all
cities combined except Denver. In contrast to the findings for 1968-1969
vehicles having higher accumulated mileages, HC, CO, and NOX during 1971 tests
were found to increase by several percentage points per 1,000 miles in all
cities except Denver, where a 7.5% decrease in CO and a 2.0% increase in NOX
were observed. Although the precise cause of this phenomenon is unknown,
special carburetor adjustments and other tune-up procedures unique to Denver's
19
-------
high altitude may well be a factor. In terms of absolute emissions levels
the change in emissions per 1,000 miles is usually considerably greater for
vehicles with initially low mileage than for vehicles with initial mileage in
the 20,000 to 30,000 mile range. Figure 8 presents slope histograms for all
cities pooled except Denver. Figure 9 presents corresponding histograms for
Denver only. These histograms are comparable to Figures 6 and 7 for 1968 and
1969 models but show a much more appreciable mileage effect.
Tables 31 and 32 present mileage effects by make for the low mileage
vs stabilized engine data base. In these tables, only those makes are included
which reach significance at at least the 95% level. Table 31 is a composite
of all the 1971 data excluding Denver, whereas Table 32 contains data for
Denver only. Appreciable differences among makes with regard to mileage effects
are observed in these tables.
2.2 CVS TESTS OF 1957-1971 MODEL VEHICLES
This section of the report summarizes results from a single test
program devoted to the measurement of exhaust emissions from 1957 through
1971 model light-duty vehicles in six metropolitan areas. Separate treatment
is afforded these results since, due to the unique orientation and objectives
of this program, test procedures for emissions measurements were employed
which produced numerical data not readily comparable to those discussed in
Section 2.1. In addition, modal emissions data and evaporative emissions
data are also presented.
A brief overview of the program is given in Table 33. The prin-
cipal function was the collection of data from which average emission factors
could be formulated in order to define the contribution of the automobile
population to the nations' air pollution burden. To achieve this objective,
the best available methodology and technology was employed to accurately deter-
mine mass emissions under vehicle operating conditions representative of road
use. Since 1957-1971 model- vehicles comprised more than 95% of the population
as of 1971, a statistically-representative sample of this population was tested
in each of the six cities which were chosen to maximize variations in climate,
terrain, and urban development.
Fifteen makes and/or manufacturers were identified. Since fifteen
model years were also involved and a city vehicle fleet consisted of only
approximately 170 units, the situation which occurred most frequently was that
only one sample of any one make/model year was tested. Consequently to ascer-
tain differences among model years of makes, data aggregation by all vehicles
within a make or by manufacturer was necessary. In addition, any one city sample
did not necessarily include all model years for each make or all makes themselves.
In the case of the early-year models (1957-1960) with a relatively small popula-
tion compared with the late-year models, only two or three makes (one vehicle
20
-------
each) were tested within any one city. Thus, the sample size for any one make/
model year/city category was very small at best and frequently consisted of only
a single datum point or frequently no data at all. Despite these limitations
in the actual size of the sample, the overall sample is very representative
of the overall vehicle population on the road. Data from this program are
particularly useful, therefore, in estimating the overall impact of emissions
on air quality.
2.2.1 Emissions Data and Results
As in the previous programs, all vehicles were tested for exhaust
emissions in an as-received condition. Cold start tests, with only minor
exceptions, were performed in accord with the 1972 (CVS-C) and 1975 (CVS-CH)
Federal Test Procedures (FTP). In actuality, however, only one test per vehicle
was made, since vehicle-diluted exhaust emissions were bagged in such a manner
that vehicle mass emissions could be calculated according to both 1972 and
1975 specifications. In addition to an assessment of the three principal
exhaust gas pollutants, a measurement was also made of carbon dioxide (C02).
In programs concerned with certification and compliance, the emissions
tests are performed using a specified blend of fuel (indolene) in the vehicle.
Since the goal of this effort was to define the actual emissions of vehicles
as operated on the road, whatever fuel was in the vehicle tank was used in all
cases (supplemented by a commercial fuel as required).
Mass emissions data, expressed as arithmetic and geometric means as
well as their respective standard deviations, are summarized in Tables 34 and
35 for the 1972 FTP, and in Tables 36 and 37 for the 1975 FTP. In addition,
the percentage of vehicles with emission levels below the reference levels of
3.4 gins/mi HC, 39.0 gms/mi CO, and 3.0 gms/mi NOXC is also presented. Although
these reference levels do not represent emission standards under which the
vehicles were certified, they are useful in examining time trends for vehicles
of various model years in which no standards or differing standards were in
effect. The results are aggregated by year and by all cities except Denver
and Denver only.
The differences that are inherent in the 1972 and 1975 procedures
reflect the fact that not all trips made by a vehicle originate from a "cold
start" (defined in the Federal Register as a start preceded by a 12-hour, no-
use soak period). The 1972 FTP determines mass emissions from a driving
schedule (LA-4) that comprises two portions, a cold-transient and a cold
stabilized. The 1975 FTP uses the same driving schedule but also adds a "hot
start" hot transient portion. In this latter case, emissions are determined
by combining a weighted average of the cold and hot-transient portions with
the cold-stabilized portion. Since a large fraction of the composite HC and
CO emissions are generated during the fuel-rich, engine warm-up phase of the
cold-transient portion of the driving schedule, the 1975 data (as compared
with'the 1972 data) will show lower levels of emission for these two effluents.
21
-------
The data of Tables 34 through 37 bear out this conclusion. On the other hand,
NOXC levels remain substantially uniform, the observed differences probably
being attributable to experimental sources.
Trends by model year in mean emission levels and percentage of
vehicles below the reference levels are well defined for all effluents. HC
and CO levels show continuing decreases and reflect the impact of increasingly
more stringent emission controls. Consistent with this decrease, there is
observed an increase in NOXC levels which, during this period of time, were
not subject to control. Figure 10, which presents 1975 CVS emission levels vs
the cumulative percentage of test vehicles, emphasizes these model year trends
for the three pollutants.
Table 38 summarizes emissions data for pre-control and control time
periods by geographic allocation. As in the case of the 7-mode test data,
emissions data taken in Denver were found to be clearly different from those
obtained in the five other cities. Consistent with previous findings, high
levels of HC and CO emissions were measured with correspondingly low levels
of NOXC for reasons which have already been discussed. Consequently, the
Denver data were considered separately from the other cities, Los Angeles,
Chicago, Houston, St. Louis and Washington, D.C.
Table 39 presents the data by vehicle weight class as determined
from the vehicle weight classification specified by the Federal Test Procedures.
Five hundred pound increments were used to establish classes from 1500 to 5500
pounds. An increase in NOXC levels with weight is evident but trends in HC
and CO are not as clearly defined.
2.2.2 Modal Emission Sequence
Considerable interest in recent years has centered around the devel-
opment of computer models to describe emission levels during various phases of
vehicle operation. For example, minimization of vehicle emissions could well
become a significant consideration in the design of expressway entrance and
exit ramps and in other highway design applications. For such purposes, it
is necessary to model the impact of emissions as they occur in the situation
of interest. Such modelling can be achieved if there are available emission
factors for a variety of steady states and driving modes. With these potential
applications in mind, a survey of the driving characteristics of the public in
the Los Angeles area was undertaken and a profile of average acceleration/decel-
eration rates were established for use as transitions between all paired com-
binations of the following speeds: 0 mph, 15 mph, 30 mph, 45 mph and 60 mph.
In addition, acceleration/deceleration rates both higher and lower than the
average values were used as transitions only between all paired combinations
of 0 mph, 30 mph, and 60 mph. A total of 32 accel/decel modes was developed.*
"The Construction of Chassis Dynamometer Test Cycles," Vols. I and II, Scott
Research Laboratories, Inc., 1971, San Bernadino, California.
22
-------
A driving cycle comprised of a sequence of the 32 accel/decel modes
with the transitions consisting of short steady-state idles or cruises was
identified. The complete sequence defined a driving schedule of 1054 seconds
duration. Emissions data obtained from vehicles operated over this driving
schedule were compared with composite results calculated from the emissions
data corresponding to the individual steady-state and accel/decel modes which
comprise this driving schedule. The synthesized composite mass emissions cal-
culated in this manner agreed very well with those measured by actual operation
of the vehicle over this driving schedule. There is encouragement, therefore,
to presume that realistic composite emissions for different driving schedules
formulated from various combinations of steady-state and accel/decel modes
can be accurately synthesized.
Table 40 lists modes together with the corresponding time in mode,
average speed, average accel/decel rate, and distance traveled. Emissions data,
taken with the constant volume sampling (CVS) technique, were acquired for each
of these 32 modes. Table 41 summarizes the results for HC, CO, C02, and NOX
for all cities except Denver, and Table 42 shows comparable data for Denver
alone.
To provide useful information on the emissions from vehicles operating
at cruise (steady state) conditions, data were collected at speeds of 0 (idle),
15, 30, 45 and 60 mph. CVS techniques again were used in the acquisition of
these emissions data. The results are shown in Table 43. With minor differ-
ences, HC and CO mean mass emissions at 30 mph, 45 mph and 60 mph are relatively
uniform in all cities. Sizable increases in the levels of O>2 and NOX emissions
occur as the steady-state speed increases from 30 mph to 60 mph. The 15 mph
cruise is identified by unusually high levels of HC, CO and CO. as compared
with the 30 mph cruise condition.
2.2.3 Evaporative Emissions
The venting of fuel vapors from the carburetor front chamber and
vehicle fuel tank constitutes a sizable source of HC emissions in the absence
of appropriate control devices. A determination of the mean evaporative losses
was made for a randomly-selected subset of vehicles in Los Angeles and Denver.
These tests were conducted in accord with the requirements of SAE Procedure J171
and constitute the SHED (Sealed Housing for Evaporative Determinations) technique.
Briefly, two types of losses are measured while the vehicle is enclosed in the
SHED, those associated with the diurnal soak and with the hot soak. The former
involves losses occurring over a one-hour period while fuel (in the vehicle tank)
is raised in temperature as per a prescribed schedule. The latter involves
losses, also over a one-hour period, while the vehicle cools following the com-
pletion of a complete cold-start 1975 Federal Test Procedure.
A word of explanation is in order with regard to the interpretation
of evaporative emissions as measured by the SHED technique. Federal certifica-
tion values, based on the canister technique (SAEJ170), are generally considerably
23
-------
lower than emissions as measured by the SHED procedure. The canister method
is aimed at collecting and weighing diurnal losses, running losses and hot
soak losses from the fuel tank and other parts of the fuel system. The SHED
technique takes account of emissions from the entire fuel system including
emissions, such as gasket leakages, that would be hard to trap. Consequently,
evaporative emissions determined by the SHED technique are not directly com-
parable to those determined by the Federal procedure.
For purposes of the surveillance program, it was advantageous to
employ the SHED rather than the canister technique in conducting evaporative
emissions tests on 1957 to 1971 model vehicles. Vehicles manufactured prior
to the implementation of evaporative controls are not readily amenable to
measurement by the canister system because of the multiplicity of tubes and
connections which would have been required to recover the emissions from
diverse fuel systems. On the other hand, the SHED technique, being performed
in a sealed enclosure, could accommodate the wide variety of vehicles involved
and was felt to be the only reasonable approach by which pre-control and later
model year vehicles could be compared. Moreover, most of the vehicles in the
1957-1971 model years belonged to the pre-control category: only 1971 model
year vehicles and 1970 model year vehicles from California employed evaporative
controls. In addition to providing a uniform system for comparing all vehicles
in the program, the SHED technique was also regarded as providing a realistic
assessment of overall evaporative emissions from the standpoint of their impact
on air quality.
Evaporative emissions for Los Angeles and Denver vehicles are summarized
below. A commercially-available, popular- premium-grade fuel was used for these
tests.
Los Angeles Denver
Number of Vehicles 126 20
Diumal Soak, (HC, grams/test) 23.67 47.20
Hot Soak (HC, grams/test) 13.75 34.78
Combined Losses (HC, grams/test) 37.41 81.98
Mean Differences Between Replicates*
Diurnal Soak (HC, grams/test) 3.99 13.80
Hot Soak (HC, grams/test) 0.27 18.08
Combined (HC, grams/test) 0.01 31.88
Ten replicates on Los Angeles, only one in Denver.
Based on the data means, the diurnal losses are appreciably larger
than the hot soak losses with the Denver losses approximately twice as large
as those in Los Angeles. In addition, a year-by-year breakdown from 1957 to
1971 of the diurnal, hot soak and combined means for the Los Angeles data is
presented in Table 44 and summarized below.
24
-------
MEAN LOS ANGELES EVAPORATIVE EMISSION LEVELS USING SHED TECHNIQUE
HC - gm/test
Pre-Control 1970-1971
Diurnal Mean 25.97 16.28
Hot Soak Mean 14.67 10.92
Combined Loss 40.64 27.20
Although no dramatic trend exists, evaporative losses as measured by
the enclosure technique show smaller evaporative losses for 1970 and 1971 Los
Angeles vehicles employing evaporative control devices than for pre-control
vehicles in that city.
25
-------
EXPLANATION OF TABLES
Table 1 is a summary of the Surveillance Programs employing the 7x7
Federal Test Procedure.
Tables 2 through 5 present vehicle emissions as determined by 7x7
Federal Test Procedure by manufacturer and make. N denotes the number of
vehicles tested, and the average mileage for those vehicles is given, in thou-
sands of miles, in the column headed "Mean Miles (K)". The percent of vehicles
which exhibited emission levels within the applicable standard levels set for
hydrocarbons (HC), carbon-monoxide (CO) and for both HC and CO are given
together with arithmetic and geometric means and standard deviations. In some
vehicle categories, there were missing observations for one or more of the
emittants. The tabulated value of N does not reflect this fact, but it is
taken into account in the computed means and standard deviations.
Table 6 is self-explanatory.
Tables 7 through 12 follow the same format as Tables 2 through 5.
Data were obtained using the 7x7 Federal Test Procedure.
Table 13 is self-explanatory.
Tables 14 through 21 are similar to Tables 2 through 5 except that
in the Phase I tables the mean mileage -is expressed in miles rather than in
thousands of miles. All data were obtained using the 7x7 Federal Test Procedure.
Tables 22 through 32 are self-explanatory.
Table 33 is a summary of Six City Surveillance Program. All data
presented for this program (Tables 34 through 39) were obtained using the 1972
CVS-1 and 1975 CVS-3 tests.
Tables 40 through 44 are self-explanatory.
27
-------
Table 1
SUMMARY OF 7-MODE SURVEILLANCE PROGRAMS BY VEHICLE MODEL YEAR
OBJECTIVES:
BASIC TEST PROCEDURES:
INSTRUMENTATION:
CITIES:
EMISSION STANDARDS:
TOTAL VEHICLES TESTED:
1968-1969 MODEL VEHICLES
EVALUATE MILEAGE AND CITY EFFECTS;
DETERMINE CONFORMITY WITH CERTIFICATION
STANDARDS
7 X 7 FTP IN ACCORDANCE WITH FEDERAL
REGISTER, VOL. 31, NO. 61 (PART II), MARCH
30, 1966; EACH VEHICLE TESTED 3 TIMES,
INITIALLY AND AT EACH OF TWO 4,000-
MILE INTERVALS.
NON-DISPERSIVE INFRARED (NDIR) APPAR-
ATUS.
KANSAS CITY AND HOUSTON
275 ppm HC, 1.50* CO (410 ppm HC, 2.30%
CO FOR VW)
1949
1970 MODEL VEHICLES
DETERMINE CONFORMITY WITH CERTIFICATION
STANDARDS; EVALUATE CITY EFFECTS
7 X 7 FTP IN ACCORDANCE WITH FEDERAL REG-
ISTER, VOL. 33, NO. 106, JUNE 4, 1968 (PART II);
EACH VEHICLE TESTED ONCE (MINIMUM MILEAGE
REQUIREMENT 4,000 Ml.).
NON-DISPERSIVE INFRARED (NDIR) APPARA-
TUS.
KANSAS CITY. HOUSTON, LOS ANGELES.
DETROIT, DENVER, WASHINGTON
2.2 GRAMS/MI. HC, 23.0 GRAMS/MI CO
2181
1971 MODEL VEHICLES
DETERMINE CONFORMITY WITH CERTIFICATION
STANDARDS; EVALUATE CITY EFFECTS,
COMPARE LOW MILEAGE AND STABILIZED ENGINE
EMISSIONS (PAIRED TESTS).
7 X 7 FTP IN ACCORDANCE WITH FEDERAL REGISTER,
VOL. 33, NO. 10B. JUNE 4. 19M (PART III: TWO PAIRED
TESTS PER VEHICLE: LOW MILEAGE ENGINE (80 MILES)
AND STABILIZED ENGINE (4,000-7 ,000 MILES).
NON-DISPERSIVE INFRARED (NDIR) APPARATUS.
HOUSTON. LOS ANGELES, DETROIT AND DENVER
2.2 GRAMS/MI. HC, 23.0 GRAMS/MI. CO.
369
-------
TABLE 2
EMISSION LEVELS FOR 1969 10DEL VEHICLES
KANSAS CITY
MANU-
FACTURER
MAKE
AMC
CHRY CORP
PLYM
DODGt
CHRY
FORD MO CO.
FORD
MERC
LINC
CM
CHEV
PONT
OLDS
BUICK
CADI
VULKS
TOTAL
* K *
* *
* *
1«
6B
33
17
ia
86
79
7
0
276
136
38
37
44
21
50
4SB
MEAN
MILES
(K)
22.1
36.6
35.6
37.6
37.1
33.0
33.3
29.1
U.O
31.8
29.4
30.6
31.6
36.6
39.5
31.1
32.^
* X BELOh * HVORnCARBOl
STANDARDS
* ARIThMET 1C
HC
26
44
48
59
22
47
47
43
0
50
32
74
70
50
90
56
48
CD
33
19
27
0
22
55
53
71
0
29
27
50
24
2
62
66
36
BOTH*
17
13
21
0
11
30
32
14
0
22
17
42
24
2
57
44
24
MEAN
337.
333.
282.
303.
454.
304.
306.
286.
0.
301.
34! .
23B.
266.
297.
197.
426.
320.
SO >
119.
254.
88.
135,
448.
65.
68.
49.
0.
131.
142.
62.
125.
116.
59.
211.
160.
IS PPM
*
GEOMETRIC *
MEAN *
319.
298.
271.
285.
372.
295.
296.
285.
0.
279.
321.
229.
250.
262.
169.
384.
295.
SO
1.40
1.49
1. 34
1.40
1.71
1.27
1.29
1.19
0.0
1.46
1.46
1.33
1.41
1.35
1.35
1.57
1.46
CARBON MONOXIDE (
ARITHMETIC
HEAN *
2.30
2.12
1.73
2.42
2.53
1.83
I.d7
1.38
0.0
2.09
2.21
1.54
1.92
2.67
1.37
1.90
2.04
SO
1.26
0.91
0.53
1.06
l.OS
1.28
1.32
0.41
0.0
1.05
1.23
0.47
0.73
0.80
0.36
0.90
1.07
*
GEOMETRIC
HEAN *
1.91
1.95
1.65
2.25
2.31
1.52
1.54
1.33
0.0
1.87
1.96
1.45
1.76
2.56
1.30
1.67
1.80
SO
1.97
1.51
1.41
1.45
i.sr
1. 84
l.aa
1.30
0.0
1.61
1.63
1.50
1.60
1. 34
1.43
1.75
1.68
NOX PPM 1
ARITHMETIC
MEAN
1536.
1587.
1583.
1494.
1689.
L589.
1604.
1417.
0.
1282.
1267.
1349.
1176.
1245.
1525.
1313.
1389.
SO "
716.
573.
494.
631.
419.
617.
614.
678.
0.
406.
443.
287.
233.
443.
415.
434.
506.
GEOMETRIC '
MEAN
1366.
1452.
1509.
1202.
1630.
1421.
1437.
1255.
0.
1199.
1155.
1319.
1151.
1167.
1456.
1239.
1277.
SO '
1.66
1.61
1.38
2.14
1.34
1.73
1.73
1.75
0.0
1.52
1.67
1.24
1.25
1.45
1.39
1.43
1.58
«« APPLICABLE STANDARDS1HC-275PPN CO-1.51 I410PPM.2.3I FOR VOLKS) ««
TABLE 3
EMISSION LEVELS FOR I960 MODEL VEHICLES
HCLSTON
MANU-
FACTURER
MAKE
AMC
CHUT CORP
PLVM
DODGt
CMRY
FOMO MO CO.
FORD
MERC
LINC
GH
CHEV
PONT
OLDS
BUICK
CADI
VOLKS
TOTAL
t>
*
*
23
83
29
29
25
75
70
5
0
249
115
43
39
42
10
39
469
MEAN
MILES
IKI
35.0
4J. 3
39.9
43.7
36.7
37.5
37.0
44.9
0.0
35.9
35.9
32.4
33.1
40.0
42.5
31.9
36.5
* I BELOk
STANDARDS
HYDROCARBONS PPM
* ARITHMETIC *
HC
13
34
48
24
28
48
49
40
0
47
31
70
44
61
60
46
43
CO
30
28
41
34
4
55
51
100
0
25
30
37
13
5
50
59
33
BOTH*
9
14
21
21
0
33
33
40
0
18
18
33
5
50
44
22
MEAN *
355.
376.
381.
394.
349.
317.
319.
293.
0.
367.
473.
250. i
316.
271.
240.
671.
315.
so
148.
307.
256.
434.
152.
193.
199.
77.
0.
363.
503.
SI.
138.
SO.
179.
680.
372.
*
GEOMETRIC »
MEAN
338.
328.
330.
327.
326.
288.
28B.
285.
0.
310.
377.
245.
296.
266.
205.
4B6.
323.
so
1.33
1.56
1.65
1.62
1.42
1.49
1.50
1.31
0.0
1.64
1.82
1.23
1.45
1.20
1.69
2.09
1.66
CARBON HCNOIl I OE X
*
ARITHMETIC *
MEAN
1.98
2.37
1.86
2.19
3.14
1.70
1.75
0.96
0.0
2.36
2.43
1.63
2.54
2.66
1.97
2.48
2.25
SO *
I. 00
1.49
1.30
0.97
1.89
1.07
1.08
0.35
0.0
1.25
1.46
1.00
1.18
0.80
0.93
1.88
1.34
GEOM6TUC *
MEAN
1.76
2.01
1.55
1.98
2.75
1.42
1.47
0.89
0.0
2.07
2.04
1.65
2.29
2.53
1.61
1.88
1.91
SD
1.67
1.78
1.87
1.59
1.64
1.84
1.83
1.60
0.0
1.70
1.65
1.54
1.62
1.38
1.52
2.16
1.80
NOX PPM i
ARITHMETIC *
MEAN
1265.
1244.
1399.
1257.
1041.
1477.
1480.
1439.
0.
1030.
1049.
1182.
876.
969.
1055.
942.
1145.
SD
382.
516.
574.
524.
458.
657.
677.
324.
0.
462.
533.
515.
363.
228.
264.
453.
534.
CEOMETRIC i
MEAN
1175.
1076.
1232.
1131.
866.
1311.
1304.
1405.
0.
927.
909.
1078.
800.
942.
1023.
800.
1006.
SD
1.60
1.91
1.84
1.67
2.20
1.70
1.73
1.30
0.0
1.62
1.76
1.58
1.57
1.27
1.31
1.91
1.74
APPLICABLE STANDAROS1HC-275PPM CO-1.5J (410PPM.2.3I FOR VOLKSI
29
-------
TABLE 4
EMISSION LEVELS FOR 1969 MODEL VEHICLES
KANSAS CITT
NANU-
FACTIOER
MAKE
ANC
CURT CORP
PLYM
OOOCE
CHRY
FORD NO CO.
FORD
MERC
LEMC
CM
CHEV
PONT
OLOS
1UICK
CADI
VOLKS
TOTAL
ft
*
19
12
28
30
24
113
97
16
0
249
90
7?
41
27
14
31
494
MEAN
MILES
IK)
20.1
21.2
21.5
22.4
19.3
21.6
23.1
12.7
0.0
21.0
20.7
21.7
19.1
24.2
17.*
11. <
20.9
* J BELOM
* STANDARCS
HYDROCARBONS PPM *
ARITHMETIC * GEOMETRIC
HC CO (BOTH* MEAN
16
49
57
40
50
43
38
75
0
65
49
65
78
71
100
81
56
47
37
36
27
50
62
57
94
0
31
21
34
27
37
86
94
44
5
26
29
17
33
35
28
75
0
29
19
29
27
37
86
77
32
383.
328.
293.
337.
359.
322.
332.
261.
0.
269.
307.
270.
241.
235.
1*4.
321.
299.
SD
162.
209.
94.
130.
345.
261.
280.
64.
0.
123.
171.
74.
77.
65.
42.
210.
187.
MEAN *
358.
300.
282.
317.
301.
292.
299.
255.
251.
280.
262.
230.
226.
160.
273.
273.
SO
1.44
1.45
1.31
1.41
1.64
1.44
1.46
1.24
1.42
1.49
1.29
1.36
1.34
1.27
1.74
1.47
CARBON MONOXIDE I *
ARITHMETIC GEOMETRIC *
MEAN
1.36
2.04
1.75
2.51
1.80
1.45
1.55
0.89
1.98
2.10
l.as
2.25
1.99
1.17
1.30
1.81
SD
0.56
1.05
0.58
I. 27
1.02
0.72
0.73
0.29
0.97
0.92
0.71
1.46
0.79
0.36
0.63
0.94
MEAN SO
1.22 1.67
1.79
1.65
2.18
1.55
1.30
1.39
0.85
1.80
1.95
1.72
1.95
1.83
1.11
.68
.44
.76
.73
.61
.60
.35
.0
.54
.44
.50
.68
.56
.42
1.11 l.BB
1.60 1.65
N0« PPM
ARITHMETIC GEOMETRIC
MEAN »
1964.
1613.
1880.
1366.
1609.
1728.
1722.
1769.
0.
1425.
1392.
1484.
1286.
1501.
1578.
1674.
1562.
SO
381.
500.
353.
553.
426.
509.
530.
365.
0.
404.
427.
344.
442.
361.
426.
512.
476.
MEAN
1928.
1515.
1841.
1240.
1590.
1640.
1625.
1733.
0.
1345.
1314.
1437.
1137.
1457.
1518.
1697.
1472.
SD
1.22
1.47
1.25
1.61
1.34
1.42
1.44
1.24
0.0
1.48
1.45
1.31
1.89
1.29
1.35
1.33
1.47
APPLICABLE STANDARDS IHC-275PPN C0-1. 51 I410PPH.2.3S FOR VOLKS I *
TABLE 5
EMISSION LEVELS FOX 1969 MODEL VEHICLES
HOUSTON
NANU-
FACTUtER
MAKE
AMC
CHKT CORP
PLVM
DODGE
CMRV
FORD NO CO.
FORD
MERC
LINC
CM
CNEV
PONT
OLDS
BUICK
CADI
VOLKS
TOTAL
*
h
*
13
73
34
23
16
105
99
6
0
266
114
65
32
31
26
29
4(8
MEAN
MILES
IKI
19.6
25.5
25. B
24.9
26.0
25.9
26.1
22.7
0.0
25. a
27.1
29.7
24.6
19.9
11.4
25.1
25.6
* S BELOM
STAMCARCS
*
* HC
38
49
38
57
63
30
29
33
0
57
49
38
69
81
100
52
49
CO
54
36
29
39
44
49
51
17
0
33
36
20
3*
10
77
69
39
HYDROCARBONS PPN *
*
ARITHMETIC * GEOMETRIC
BOTH* MEAN »
15
29
21
35
38
22
23
0
0
28
29
13
34
10
77
45
28
301.
294.
305.
276.
291.
354.
357.
306.
0.
271.
291.
311.
235.
243.
156.
491.
306.
SO
31.
94.
76.
62.
154.
164.
168.
48.
0.
105.
122.
94.
56.
}5.
21.
334.
152.
MEAN
296.
281.
294.
269.
270.
326.
327.
303.
0.
255.
ZJ4.
300.
228.
240.
155.
44».
283.
SO
1.21
1.36
1.31
1.24
1.54
1.48
1.49
I.J6
0.0
1.40
1.42
1.30
1.29
1.16
1.14
1.52
1.46
CARBON MONOXIOE I «
ARITHMETIC GEOMETRIC
MEAN *
1.52
2.04
2.31
1.71
1.92
1.75
1.65
3.31
0.0
1.95
1.86
2.13
2.00
2.41
1.31
2.10
1.92
SO
0.62
1.07
1.15
0.70
1.25
0.85
0.75
0.99
0.0
0.86
0.77
0.94
0.83
0.94
0.39
1.24
U.92
MEAN SD
1.38 1.59
1.77 .73
2.02
1.55
1.62
1.56
1.50
3.14
0.0
1.7B
1.70
1.96
1.12
2.26
1.26
.73
.61
. 6AROSIHC-275PPH CO-1.5S (410PPM.2.3I FOR VDLKSI »
30
-------
Table 6
SUMMARY OF EMISSIONS (ARITHMETIC MEANS)
FOR 1968 AND 1969 MODEL VEHICLES
1968 MODEL VEHICLES
KANSAS CITY
HOUSTON
1969 MODEL VEHICLES
KANSAS CITY
HOUSTON
HYDROCARBONS
ppm gm/mi
320 3.81
385 4.48
299 3.65
306 3.67
CARBON MONOXIDE
Percent gm/mi
2.04 46.3
2.25 51.3
1.81 42.4
1.92 44.1
NOX
xc
ppm gm/mi
1389 4.69
1145 4.50
1562 5.34
1337 5.33
68-69 STANDARDS: HC-275 ppm, CO-1.5% (HC-410 ppm, CO-2.3% FOR VOLKS)
31
-------
TABLE T
EMISSION LEVELS FCK 1970 MODEL tftHICLES
KANSAS CITY
MANU-
FACTURER
MAKE
ADC
CHRY CORP
PLYM
OOOCE
CNRY
F0»0 MO CO.
FOKO
NWC
LINC
CM
CHEV
PCNT
OLDS
BUICK
CADI
VOUS I NISC
TOTAL
* N
*
* *
20
60
42
13
5
124
115
6
3
171
40
27
24
2S
5
20
395
MEAN
MILES
IKI
10.3
10.1
9.6
10. «
12.7
u.e
13.0
9.4
10.2
11.0
1.»
11.5
10.9
13.9
14. S
0.5
11.2
t BELOM
* STANCARLS
HC
SO
33
36
31
20
23
22
0
100
37
32
59
SO
16
40
65
34
CO
SO
28
JO
l>
0
48
49
SU
0
22
17
48
17
20
20
95
36
*
BOTH*
40
18
24
8
0
15
17
0
0
16
12
37
17
20
65
20
HYDROCARBONS ON /PI I
ARITHME
MEAN
2.44
2.74
2.74
2.!>4
3.27
2.72
2.73
2.«6
1.69
2.66
2.40
2.58
2.27
2.66
2.29
2.21
2.66
TIC
SO
0.78
1.09
1.23
0.49
0.95
0.72
0.72
0.56
0.33
1.11
1.09
1.76
0.77
0.60
0.46
1.21
0.99
(.EOMET
MEAN
2.34
I. 60
2.58
2.49
J.14
2.64
2.65
2.91
1.66
2.47
2.0 1
2.24
2. IS
2.62
2.21
2.07
2.51
RIC *
SO
1.33
1.36
1.39
1.23
1.39
1.29
1.27
1.21
1.24
1.46
1.49
1.65
1.41
1.27
1.23
1.55
1.40
CARBON MONOXIDE GN/MI *
*
ARITHME
MEAN *
31.2
40.9
38.4
40. B
61.7
26.4
26.3
2S.8
29.3
39.1
41.7
28.6
40.9
40.4
33.1
17.3
33.9
TIC
SO *
21.6
23.2
24.0
17.4
22.7
13.8
14.0
14.8
2.6
20.5
18.3
27.3
20.5
17.9
16.1
4.J
19.9
GEONET
MEAN
24. I
34.7
i2.0
37.0
57.8
22.7
22.6
21.9
29.2
34.1
37. T
22.9
3S.9
36.7
28.6
16.1
28. S
RIC «
SO *
2.06
t.ai
1.9S
1.62
1.S2
1.79
1.80
1.91
1.09
1.72
1.60
1.91
1.72
1.97
1.97
1.31
1.83
ARITHME
MEAN *
*.32
5.7»
S.80
9.69
5.40
4.95
5.02
3.96
4.27
4.11
3.97
3.76
3.53
5.19
5.79
4.01
4.63
urn GM/NI
7 1C
sn
1.18
1.54
1.57
1.57
1.39
1.74
1.77
1.12
0.54
1.33
1.15
1.43
1.16
1.19
1.37
0.87
1.59
GtOMET
MEAN
-. 12
9.52
5.58
5.42
i.24
4.62
4.68
}.81
4.25
3.88
3.79
3.52
3.34
5.04
5.63
3.92
4.)4
*
RIC <
so
1.42
1.34
1-.34
1. 38
1.33
1.47
1.48
1.3V
1.13
1.43
1.40
1.44
1.42
1.29
1.32
1.24
1.45
APPLICABLE STANDABDS1HC-2.2 CM/MI CO-23. GM/MI >
TABLE 8
EMISSION LEVELS FOX 1970 HOI)EL VEHICLES
HOUSTON
NANU- *
FACTURER N
MAKE
AMC 20
CHKY COKP 60
PLYM 30
DODGE 26
CHRY 4
FORO HO CO. 118
fORO 107
MERC 9
LINC 2
GM 16V
CHEV 8S
PONT 27
OLDS 22
MICK 24
CADI 8
VOUS t NISC 20
TOTAL 387
MEAN
MILES
IKI
10.7
11.4
12.6
9.9
11.5
14.2
14.4
12.3
10.6
12.9
12.7
13.8
13.3
12.9
12.5
10.6
12. a
* S BELOk « HYDROCARBONS GN/NI CARBON MONOXIDE CM/11 NOX GM/MI
STANDARDS . ,
ARITHMETIC GtOMETHIC ARITHMETIC GEOMETRIC ARITHMETIC GFOMFTRIC
HC CO «BOTH« MfcAN SO « MCAN « SO MEAN SO « MEAN SO MEAN SO MEAN * Sl>
30 SO 20 2.71 u.78 2.61 1.32 26.6 15.5 23.1 1.72 3.28 1.90 4.95 1.46
22 23 13 3.0B 1.72 ^.64 1.43 48.0 35.9 58.9 1.89 6.20 2.17 5.67 1.66
20 30 13 2.78 0. BO 2.68 1.31 39.7 34.1 32.3 1.84 6.54 2.22 5.91 1.78
23 15 12 314S 2.42 >.05 1.56 58.0 37.7 48.0 1.88 5.88 2.25 5.39 .58
25 25 25 2.91 0.91 2.61 1.36 45.1 25.5 40.0 1.73 5.78 0.63 5.76 .11
20 34 13 2.98 1.34 2.81 1.38 37.6 22.2 31.5 1.87 4.65 .83 4.26 .56
20 34 11 3.03 1.38 2.84 1.39 it. 4 22.8 32.1 1.87 4.60 .85 4.?1 .57
31 44 33 2.42 0.52 2.37 1.21 29.3 15.2 25.0 1.92 4.93 .79 4.65 .44
0 U 0 2.93 0.08 2.93 1.03 32.4 2.6 32.4 1. OS 5.95 0.59 5.94 .11
25 18 12 3.02 1.61 2.73 .56 43.6 22.9 37.5 1.80 4.20 .56 3.9) .45
15 14 6 3.38 l.Be J.07 .52 45. S 22.3 39.7 1.76 4.17 .45 3.93 .42
22 30 19 2.87 1.11 2.65 .53 39.4 23.6 32.5 1.95 3.54 .21 3.32 .46
64 41 36 2.41 1.77 1.98 .84 )2.4 19.3 27.6 1.79 3.60 .09 3.44 1.36
25 8 S 2.67 0.61 2.61 .25 53.0 26.3 46.4 1.73 4.35 .60 4.58 1.47
50 0 0 2.27 0.39 2.24 .1) 40.3 10.8 V).2 1.28. 6.61 .91 6.34 1.38
25 70 20 3.05 1.35 <.83 i.4f> 22.9 13. I 21.1 1.50 5.90 0.96 3.79 1.28
24 28 13 3. DO 1.50 2.77 1.47 40.9 25.2 33.8 1.85 4.69 1.88 4.31 1.54
APPLICABLE STANOARDS:HC-2.2 CM/HI CO-23. GM/MI -
32
-------
TABLE 9
EMISSION LEVELS FOR 1470 MOOfcL VEHICLES
LOS ANGELES
MANU-
FACTURER
MAKfc
AMC
CHR» CIJRP
PL»N
COOGE
CHR»
FOUO MO CO.
FORO
MERC
LINC
GH
CHEV
PONT
OLDS
8UICK
CADI
VULKS t NISC
TOTAL
N
*
*
19
57
27
22
8
81
73
6
2
175
90
30
23
23
9
20
352
H£AN
MILES
IK)
7.4
H. 4
7.8
9.2
8.5
7.7
7.7
7.9
7.6
8.4
7.9
9.3
8.9
7.9
12.5
7.6
a.<
1 BELOM
STANOARuS
HVCIRdCAKBUNS GH/HI
* ARITHMETIC «
HC
»3
54
70
45
25
38
36
67
50
54
43
60
87
57
56
60
51
* CO
47
47
67
32
25
38
38
33
50
36
26
50
74
30
11
50
40
BOTH*
37
39
63
23
0
21
21
33
0
27
1*>
40
65
26
11
35
29
MEAN
2. 2C
2.43
2.16
2.69
2.75
2.82
2.87
2.29
2.25
2.34
2.47
2.38
2.11
2.08
2.12
2.31
2.46
SO *
0. 73
0.86
0.68
1.06
0.4d
2.87
3.01
0.81
0.53
1.25
0.99
1.49
2.25
0.49
0.33
1.24
1.71
btUNETRlC *
MEAN
<.G9
f.ii
2.06
2.50
2.47
2.50
2.18
2.22
2.12
2.29
2.14
1.63
2.03
2.10
2.09
2.22
SO
1.41
1.37
1.34
1. 40
1.20
1.50
1.52
1.39
1.27
1.54
1.50
1.55
1.86
1.28
1.18
1.5)
1.50
C»R80N HON.WIDE GM/HI
ARITHMETIC
MEAN
29.1
35.4
30.0
37. a
47.3
32.0
32.3
33.9
16.6
33.5
35.0
32.6
22.1
17.3
40.8
21.3
32.5
SO
20.7
27.5
27.9
27.6
24.4
18.5
18.0
26.6
10.3
20.0
18.9
21.4
19.7
19.1*
13.2
8.6
20.8
GEOMETRIC
MEAN *
22.6
26.0
21.5
27.9
40.8
26.9
27.5
25.1
14.9
28.1
30.9
25.4
16.7
32.7
38.6
19.5
26.6
SO
1.14
2.10
2.20
2.48
1.85
1.84
1.80
2.43
1.95
1.86
1.65
2.10
2.12
1.71
1.45
1.58
1.94
NOX GM/MI i
4
ARITHMETIC
MEAN *
>.ie
6.10
6.11
6.65
5.96
5.02
5.09
4.25
4.73
4.72
4.50
4.17
4.56
5.63
6.78
4.18
5.04
SO
1.38
1.90
1.88
2.17
1.23
1.56
1.60
0.29
2.32
1.65
1.59
1.37
1.78
1.43
1.08
1.18
1.74
GeiMFTRIC
MEAN *
5.00
5.96
5.76
6.27
5.85
4.74
4.81
4.24
4.*4
4.19
4.22
1.90
4.10
5.4>
6.70
1.95
».71
sn '
1.31
1.43
1.47
1.45
1.24
1.41
1.4)
1.07
1.67
1.50
1.45
1.50
1.71
1.30
1.18
1.43
1.48
» APPLICABLE STANOAMDS:HC-2.2 CM/MI CO-23. G-N/NI
TABLE 10
EMISSION LEVELS FCH 1970 MODEL VEHICLES
DtTROIT
MANU-
FACIURER N
*
MAKE *
AHC 20
CHRY CORP 60
PL»H 28
OUIIGE 21
CHRV 11
FORO MO CO.. 81
FORD 66
MERC 13
I INC 2
CM 171
CHEV 90
PONT 28
OLOS 23
BU1CK 23
CAOI ,-r T
MILKS ( NISC 20
TOTAL 352
MtAN
HUES
IKI
V.2
9.7
9.5
9.0
11.8
9.5
9.!
9.1
13.7
13.7
IS. 7
10.7
11.1
10.4
10. T
7.9
11.5
* 1 BELOk * HVDROCARBONS CM/MI CARBUN MONOXIDE 6M/MI NOI GM/MI
* STANDARCS * « . .
ARITHMETIC GEOMETRIC ARITHMETIC GEOMETRIC * ARITHMETIC (.EIWFTRIC
HC CO «BDTH» MEAN SO « HEAN SO HtAN SO MEAN SO MEAN SO MEAN » SO
50 50 45 2.S5 0.75 2.45 .- 1.33 26.1 IT. 7 21.6 1.87 4.86 1.62 4.60 1.42
IT 2S 13 3.05 0.93 2.92 1.35 49.5 30.4 35.1 2.19 5.47 .64 5.19 .42
25 39 21 2.72 0.77 2.61 1.35 35.9 29.7 26.2 2.29 4.81 .51 4. S3 .46
14 24 10 3.05 0.68 2.98 1.25 48.3 27.4 40.5 1.91 6.41 .63 6.17 .34
090 3.86 1.24 3.70 1.36 64.5 30.1 56.1 1.07 5.35 .09 5.25 .22
14 27 10 3.44 2.10 3.16 1.44 33.2 16.3 29.5 1.64 4.33 .48 4.09 .42
14 ,30 11 3.43 2.24 3.13. 1.45 30.7 14.4 2T.6 1.59 4.39 .58 4.11 .46
15 IS 8 3.55 1.4(1 3.30 1.46 43.0 21.6 37.8 1.73 4.02 .88 3.92 .27
000 3.09 0.06 3.09 1.02 49.B 1.5 49.S 1.03 4.44 .04 4.38 .27
29 19 9 3.08 1.31 2.64 1.48 43.6 23.0 37.5 1.78 3.85 .48 3.56 1.91
19 20 T 3.36 1.45 3.0« 1.51 44.2 23.1 3T.8 1.83 3.70 .48 3.38 l.ST
29 36 21 2.88 1.35 2.63 1.53 31.7 16.3 27.9 1.70 3.75 .25 3.55 1. 4O
48 17 13 2.3S 0.62 2.28 1.30 42.2 22.6 37.1 1.68 1.44 .26 3.22 1.45
17 4 0 3.19 0.93 9.07 1.33 57.5 25.1 52.1 1.61 4.41 .49 4.20 1.37
43 0 0 2.26 0.33 2.24 1.16 42.8 1S.7 40.6 1.42 9.78 .29 5.64 1.28
60 70 55 -2.27 0.81 2.16 1.37 23.a 12.6 21.1 1.63 J.43 1.21 3.23 1.45
24 27 14 1.08 1.46 2.86 1.45 39.4 23.3 32.9 1.86 4.27 1.62 3.45 1.51
** APPLICABLE STANOAROSlHC-2.2 GM/MI CO-23. CM/Ml
33
-------
TABLE 11
EMISSION LtVELS FOR It70 MODEL VEHICLES
DENVER
MANU-
FACTURE*
MAKE
>NC
CMRY CORP
PCYM
OCCCE
CHRV
FCXC MO CO.
FORD
NE*C
UNC
GN
CHtV
PONT
OLDS
8U1CK
CADI
VOLKS 4 MISC
TOTAL
»
N
*
20
60
32
20
a
it
5a
21
2
IT)
40
28
22
21
1
20
352
MEAN « 8ELOH
MILES * STANDARDS
IK I * * ARITHMETIC
HC » CO BOTH* MEAN SO
HYDROCARBONS GM/NI
CARBON MONOXIDE SM/NI
4.6
4.1
4.1
4.2
8.4
4.4
10.1
4.4
8.}
10.1
10.5
10. 1
12.4
10.2
12.2
*. 5
10.1
0
2
1
0
0
2
3
0
0
a
9
11
9
0
0
10
5
15
0
0
0
0
5
7
0
0
2
1
7
0
0
0
0
3
0
a
a
0
0
a
0
0
0
i
0
4
0
0
0
0
0
3.76
4.85
3.11
6.53
4.74
4.15
4.25
3.90
4.08
4.24
4.66
4.07
3.34
3.82
3.63
3.34
4.25
1.04
4.54
1.42
7.57
0.62
1.41
1.55
0.98
0.68
1.73
1.88
1.46
0.77
1.30
0.44
1.09
'.39
3,63
4.24
3,63
5.17
4.76
3.44
3.49
3.79
4.05
3.93
4.28
3.73
3.10
3.63
3.60
3.23
3.42
OtOHETRi:
MIAN SO
3,6) 1.3J
4.24 1.52
3,63 l.3»
5.17
4.7*
1.9*
3.49
3.79
4.05
3.93
.28
3.7}
3.10
1.43
J. 60
.74
.14
.33
.42
.29
.18
.48
.51
.91
.28
.38
.15
3.23 1.37
3.92 1.45
AHITHMETIC
MEAN *
46.6
74.6
71.*
81.4
106.9
42.6
SI. 5
55.2
55.1
71.0
6*. a
60.3
75.8
86.4
111.2
41.3
«5. 1
S3
16.3
32. 4
30.0
30.0
35.)
21.3
24.1
11.9
5.6
24.4
25.7
24.4
16. )
30.7
39.4
4.0
24,3
GEOMETRIC
MEAN SO
43.6 1.44
73.4 1.47
66.2 1.48
77.6
101.8
48.5
46.5
54,0
54.4
65.0
S4.4
53.6
71.9
82.1
.34
.40
.53
.58
.23
.11
.94
.50
.65
.27
.41
114.0 1.36
SO. 3 1.25
54.1 1.56
ARITHMETIC
MEAN * SO *
4.14 1.57
3.43 1.62
3.44 1.9«
3.47 1.87
3. 58 1.34
4.21
4.17
4.26
4.62
3.93
3.55
4.14
2.47
3.48
.74
.41
.54
.26
.28
.06
.44
.08
.15
2. 48 0. 80
4.16 1.25
3.83 1.51
GEOMETRIC
MEAN
3.42
3.54
3.66
3.48
3.32
3.82
3.75
3.47
4.62
3.30
3.35
3.74
2.40
3.27
2.»4
4.00
J.53
SD
1.45
1.63
1.57
1.78
1.54
1.58
1.62
1.50
1.06
1.47
1.49
1.54
1.41
1.45
1.30
1.32
1.52
* APPLICABLE STANOAROSlHC-2.2 CM/MI CO-23. GN/NI *
TABLE U
EMISSION LtVELS FOB 1970 HJOEL iftHlCLES
ASHINGTCJN
MANU-
FACTURER
MAKt
AMC
CHRV CORP
PLYM
DODGE
CMRY
FORD MO CO.
FORD
MERC
LIHC
CM
CHtV
PONT
OLDS
BU1CK
CAOI
VOLKS I MISC
TOTAL
1*
*
30
66
26
24
4
75
65
2
191
76
24
20
24
7
21
343
MEAN
MILES
IKI
10.8
10. T
10.6
10.5
11.9
11.1
11.3
11. 1
12.*
11.1
10.6*
12.4
11.5
11. 0
11.4
12.1
11.1
S
BELO. *
HYDROCARBONS GH/M1
CARBON MONOXIDE GN/MI * HiJX GX/NI
STANDARDS
HC
37
14
18
14
0
28
28
13
ICO
34
24
46
60
29
43
11
29
» CO »BOTN»
71 37
26 12
25 14
34 14
0 0
51 24
51 25
50 13
50 50
33 17
22 1
71 31
40 15
29 13
14 14
76 29
42 20
RIIHMET
C CEQMtTRIC *
MEAN SO MIAN * SU
2.57 0.9* 2.42 1.42
3.15
2.96 (
2.99 (
.01 1.01 1.35
J.89 2.84 1.11
.80 2.89 1.31
4.23 1.36 4.04 1.17
3.17
2.94
5.46
1.50
2.95
1.22
1.10
.76 2.79 1.53
.10 2.76 1.41
.89 1.49 2.25.
.07 l.iO 1.05
.71 2.61 1.S8
.72 2.91 1.55
.54 2.60 1.71
2.25 1.08 2.01 1.62
2.51 0.48 2.42 1.31
3.07
1.99 2.68 1.70
2.70 U.94 2.54 1.43
2.99 1.12 2.71 1.51
ARITHMETIC
MEAN so
21*0 17.1
19.0 IB. 5
37.0 18.0
16.0 18.0
55.4 15.0
24.9 15.1
24.2 12.9
25.5 21.1
42.8 48.1
31. 16.7
34. 18.0
21. 11.6
28. 15.3
34. 15.5
29. 11.0
14. 2 7. 2
24.0 17.4
GEOMETRIC
MEAN « SU
16.0 2.04
33.4 1.74
32.1 1.79
31.0 1.82
53.5 1.33
20. 1.85
20. 1.74
19. 2.15
25. 4.67
27. 1.74
30. 1. 74
IB. 1.77
24. 1.67
31. 1. 51
28..0 1.38
18.0 1.43
25.0 1.8*
ARITHMETIC
MEAN * SO
4. 77 1. 51
5.U2 1.78
6.00 2.00
5.60 .71
5.46 .27
4. no .84
4.84 .85
4.19 .83
1.14 .75
4.53 .54
4.17 .22
4.22 .07
4.00 0.84
5.73 1.81
6.43 2.45
3.86 0.75
4.82 1.70
GEOMF T
MEAN «
4.50
5.55
5 .b4
5.34
5.H3
4.44
4.51
4.32
2.44
4.29
3.47
4.10
3.41
5.45
6.48
3.81
».51
R|C
sn
1.42
1.37
1.40
1.38
1.26
1.52
1.46
1.91
1.79
1.40
1.39
1.29
1.24
1.18
1.52
1.22
1.44
APPLICABLE STANOAROSlHC-2.2 GM/MI CO-21. OH/MI
34
-------
Table 13
SUMMARY OF EMISSIONS (ARITHMETIC MEANS)
FOR 1970 MODEL VEHICLES
KANSAS CITY
HOUSTON
LOS ANGELES
DETROIT
WASHINGTON
DENVER
TOTAL EXCLUDING DENVER
HYDROCARBONS
gms/mi
2.66
3.00
2.46
3.08
2.99
4.25
2.83
CARBON MONOXIDE
gms/mi
33.9
40.5
32.5
39.4
29.8
65.1
35.3
N°XC
gms/mi
4.63
4.69
5.04
4.27
4.82
3.83
4.69
70-71 STANDARDS: HC-2.2 gm/mi, CO-23 gm/mi
35
-------
TABLE 1*
EMISSION LEVELS HM 19T1 MODEL VEHICLES
HOUSTON-PHASE 1
NANU-
FACTURER N
MEAN
MILES
X BELOk
STANCARDS *
HYDROCARBONS GN/MI
ARITHMETIC >
MAKE <
ANC
CHKV CORP
PLYM
OOOCE
CHAT
FORD HO CO.
FORO
MERC
LINC
CM
CHEV
PONT
CLDS
SUICK
CAC1
VOLKS I HISC
TOTAL
>
9
10
4
6
14
14
0
0
23
13
0
5
4
1
9
65
12.7
19.0
27.0
13.7
25.9
25.9
0.0
0.0
T.I
7.1
0.0
9.8
4.8
0. 0
11.8
14.5
* HC
78
80
100
67
79
79
0
100
100
0
100
100
0
88
CO
22
10
25
0
50
50
O
74
62
0
100
75
0
56
49
BOTH*
22
10
25
0
50
50
0
74
62
0
100
75
0
44
48
MEAN *
1.89
1.95
1.71
2.12
1.86
1.86
0.0
1.14
1.3k
0.0
0.78
0.94
0.0
1.55
1.58
SO
0.63
0.59
0.39
0.68
rt 0
0.32
0.32
0.0
0.39
0.33
0.0
0.17
0.21
0.0
0.12
0.55
GEOMETRIC
MEAN
1.81
i.ba
1.67
2.01
0*0
1.83
1.8}
0.0
i.oa
1.34
0.0
0.76
0.93
0.0
1.52
1.48
S3
1.38
1.14
1.28
1.38
rt_ Q
1.19
1.19
0.0
0.0
1.42
1.29
0.0
1.27
1.26
0.0
1.21
1.45
CARBON MONWIDE GM/N 1
ARITHMETIC
MEAN *
J4.5
38.2
36.4
39.4
0* 0
26.5
26.5
0.0
0.0
19.8
23.1
0.0
12.4
18.9
27.0
27.1
SD
15.1
11.1
11.0
10.8
0.3
11.0
11.0
0.0
11.8
14.1
0.0
4.0
7.8
14.5
13.8
GEOMETRIC *
MEAN
30.2
36.5
34.1
3B.1
0.0
24.1
24.3
17.3
20.0
0.0
11.7
17.9
21.9
23.6
SO
1.87
1.39
1.51
1.34
0.0
1.57
1.97
1.67
1.72
0.0
1.93
1.43
1.67
1.73
NOX G4/MI <
1
ARITHMETIC
MEAN >
3. 3«
4.19
3.06
4.93
0.0
4.24
4.24
2.&0
2.68
0.0
2.58
2.52
2.95
3. J5
SO
1.22
1.80
l.H
1.76
0 .0
1.56
1.56
0.50
0.62
0.0-
0.09
0.43
0.82
1.35
GEOMETRIC <
MEAN
1.15
3.78
2.87
4.55
0.0
3.99
3.99
0.0
2.56
2.62
0.0
2.59
2.44
0.0
2.1)4
1.12
SO '
1.52
1.65
1.52
1.64
0.0
1.43
1.43
0.0
1.21
1.26
0.0
1.03
1.17
0*0
1.34
1.46
*« AMLICABLE STANDARDSinc-2.2 CM/MI co-23. CM/MI *»
TABLE 15
EMISSION LEVaS FOR 1971 MODEL VEHICLES
LOS ANGELES-PHASE 1
MANU-
FACTURER
NAKt
AHC
CHRT CORP
PLVM
DOUGE
CH«r
FORO NO CO.
FORD
MERC
LINC
CM
CHEV
PONT
OLDS
BUCK
CADI
VOLKS 1 HISC
TOTAL
N *
'
t
23
12
6
5
30
24
6
0
40
22
6
4
6
2
11
112
MEAN
MILES
29.4
29.3
27.3
35.2
27.2
40.*
44.8
23.3
0.0
35.3
36.0
35.8
31.8
34. T
35.0
36.2
35.1
* I BELOk HYDROCARBONS GM/MI
* STANDARDS
ARIIHNET 1C GEOMETRIC
HC
100
91
92
100
80
93
92
100
0
95
91
100
100
100
100
73
92
CO
10D
30
17
67
20
87
83
100
0
78
68
67
100
100
100
55
70
BOTH* MEAN *
100 1.06
30
17
67
20
87
83
100
0
75
64
67
too
.74
.»u
.51
.61
.24
.23
.27
.0
.01
.14
.15
.69
100 O.TT
100 0.51
36 1.67
67 1.29
SO
0.35
1.41
0.43
0.64
2.96
0.43
0.46
0.30
0.0
0.47
0.55
0.26
0.20
0.17
0.11
0.56
0.80
MEAN
l.Ot
1.49
1.44
1.37
1.79
1.17
1.16
1.24
0.0
0.93
1.06
1.11
0.67
0.76
0.50
1.59
1.16
SD
1.42
1.66
1.34
1.66
2.43
1.41
1.45
1.21
0.0
1.47
1.45
1.26
1.33
1.26
1.25
1.41
1.53
CARBON MONO* IDt GN/MI
*
ARITHMETIC GEOMETRIC
MEAN
12.6
34.4
42.6
22.9
28.6
14.0
14.3
12.4
0.0
15.8
18.0
21.8
6.2
10.4
10.2
24.1
19.7
SO
5.3
19.2
19.9
19.2
5.8
7.3
8.2
1.8
0.0
9.5
8.5
14.6
2.6
2.1
3.0
B.5
13. »
MEAN
11. »
28.6
18.5
16.0
28.0
12.6
12.6
12.1
0.0
13.3
16.0
17.1
5.8
10.2
10.0
22.8
16.0
SD
1.57
2.00
1.61
2.66
1.24
1.57
1.66
1.16
0.0
1.83
1.66
2.10
1.61
1.21
1.14
1.41
1.91
ND« CM/HI
ARITHMETIC * GEOMETRIC <
MEAN »
2.36
3.49
3.40
3.55
3.65
3.33
3.21
3.82
0.0
2.89
1.03
2.96
2.94
2.43
2.36
2.69
1.07
SD *
0.69
1.13
1.23
1.08
1.16
0.67
0.54
0.96
0.0
0.65
0.75
0.43
0.38
0.50
0.02
u.85
0.86
MEAN
2.27
3.30
3.16
3.40
3.53
3.26
1.16
3.71
0.0
2.82
2.94
2.94
2.43
2.39
2.36
2.58
2.94
SD
1.15
1.44
1.54
1.40
1.32
1.23
1.20
1.32
0.0
1.26
1.29
1.16
1.13
1.25
1.O1
1.13
1.33
APPLICABLE STANOARDSlHC-2.2 GN/HI CO-23. GN/MI **
36
-------
TABLE 16
EMISSION LEVELS FOR 1971 MODEL VEHICLES
DETROIT-PHASE 1
MANU-
FACTURER
* *
N
MEAN
MILES
I BELOh
STANDARDS
HYDROCARBONS GH/M I
* « ARITHMET 1C
HAKE
AHC
CHRY CORP
PLYH
OOUGb
CHRY
FOMO HO CO.
FORO
MERC
LINC
GH
CHEV
PONT
OLDS
BUICK
CA01
* *
0
9
4
3
2
15
15
0
0
13
2
3
6
0
0.0
13.2
7.0
12.7
26.5
19.1
19.1
0.0
o.o
18. 6
17.2
12.5
25.0
21.0
0.0
HC
0
78
75
67
100
33
33
0
0
96
92
100
100
100
0
« CO
0
11
0
0
50
47
47
0
0
75
85
0
100
67
0
BOTH*
0
11
0
0
50
33
33
0
0
75
85
0
100
67
0
MEAN »
u.o
1.92
2.17
2.00
1.30
2.36
2.36
0.0
0.0
1.34
1.46
1.60
o.ei
1.26
0.0
SD
0.0
0.42
0.15
0.40
0.21
0.46
0.46
0.0
0.0
0. 54
0.61
0.74
0.32
0.22
0.0
*
GEOMETRIC
MEAN
0.0
1.67
2.16
1.97
1.30
2.32
2.32
0.0
0.0
I .25
1.36
1.52
0.77
1.25
0.0
so
o.u
1.21
1. 07
1.21
1.17
1.22
1.22
0.0
0.0
1 .45
1.44
1.62
1.46
1. 18
0.0
CARBON MONOXIDE GVM
*
ARITHMETIC *
HEAN
U. U
43.2
50. 8
45.4
25.0
31. 0
31.0
0.0
0.0
20.6
45.5
13.0
19.3
0.0
SO
0. 0
14.7
9.0
16.2
7.5
16.8
16.8
0.0
0.0
9.2
31.2
4.9
10.7
0.0
SEOMET-IIC
MEAN «
0.0
40.8
50.1
43.5
?4.5
26.5
26.5
0.0
0.0
19.0
39. 8
12.5
17.1
0.0
SO «
0.0
1.46
1.21
1.41
1.35
1.84
1.84
0.0
0.0
1.53
2.12
1.43
1 .68
O.u
NO X GM/MJ
«a| TIME TIC «
MEAN *
0.0
3.37
2.=<;
4.01
3.28
3. 1=6
3.86
0.0
0.0
2.67
1.71
2.92
2.15
0.0
SO
0.0
0.95
0.67
1.33
0.63
1 .53
1.53
0.0
0.0
0.70
0.59
0.08
0.34
0.0
GEOMfTRIC
ME»N «
0.0
3.26
2.37
1. 36
3.25
3.57
3.57
0.0
0.0
2.58
1.66
2.92
2.U
0.0
Su
0.0
1. 31
1.27
1.39
1. 21
1.52
1.52
3.3
0.0
1.31
1.42
1. 03
1.17
0.0
VOLKS i. MISC
1.77 0.67
*» APPLICABLE STANDARDS:HC-2.l GH/ M 1 CO-23. l»M/Ml »»
IA6Lt 17
EMISSION LbVtLS t-C« H71 HJOfcU VEHICLES
OENVER-CHASE 1
MANU-
FACTURER
MAHE
AMC
CHRY CORP
PLYM
OCDGE
CHRY
FORO MC CO.
FORO
MERC
LINC
GH
CHEY
PCNT
ULOS
BUICK
CADI
VOLKS C HISC
TOTAL
*
* N «
* *
20
Ib
14
3
1
26
24
1
1
57
34
5
6
9
3
23
144
MEAN
MILES
19. >
17.6
19.6
9.7
0.0
26.2
26.4
0.0
0.0
19.5
18.7
26.4
14.3
22.0
19.3
23.3
21. u
* 1
; HELO*
*
HYDI
STANDAkUS *
HC
30
0
7
u
0
4
4
0
0
63
50
100
67
78
100
35
36
*
- co »ecTH«
10
0
0
0
0
0
a
0
0
0
0
0
0
0
0
4
2
10
0
0
Cl
U
0
0
0
0
0
0
0
0
0
0
0
1
AKI THHI
MEAN *
3. 12
3.22
2.S9
4. -3
0.0
3.C7
3.12
0.0
0.0
2.16
2.37
1.82
2.13
2.08
0.67
2.43
2.63
MYOMJCAKHUNS CM/M 1
1C
SD
I. 00
0.72
1.69
0.0
1. 16
1.19
0.0
0.0
0.74
0.55
0.17
0.91
0.99
0. 16
0.64
1.03
GEDMETKI; «
MEAN * SU
3 .09
2 .91
4.06
0.0
2.96
1.00
0.0
0.0
2.03
2.31
1.62
2.00
1.92
0.65
2.35
2.46
1.31
1.26
1. 53
0.0
1.23
1.29
0.0
0.0
1.43
1.25
1. 10
1.47
1.50
1.31
1.30
1.46
ARITHMETIC
HtftN bj
HI. 4
77.0
90.2
0.0
51.2
51.6
0.0
0.0
68.6
76.5
45.9
66.4
60.5
46.9
44.3
64.0
27.3
27.9
1 9. 9
0.0
12.4
12.7
0.0
0.0
25.9
25.4
15.3
16.9
29.6
11.6
13.1
27.6
GEOMETRIC »
MEAN « SO
76.9
72.4
148.8
0.0
49.8
50.3
0.0
0.0
64.1
72.6
44.2
64.5
54.9
46.0
42.3
58.5
1.42
1.44
1.25
0.0
1.27
1.28
0.0
0.0
1.45
1.39
1.35
1.32
1.58
1.28
1.39
1.54
ARITHMETIC
HfAN * SO *
2.97
2.98
3.24
0.0
* ,-VS
4.45
0.0
0.0
2.25
2.14
2 * 5b
2.72
2.16
2. 34
7 .<.<;
f. 86
l.St
1.51
0.74
0.0
1.67
1.74
0.0
0.0
0. 79
0.82
0.46
0.68
0.86
0.76
U.77
1.42
GC ,-FTaic *
ME AM S L)
2
2*
3
0
i.
-------
FABLE 18
EMISSION LEVELS FUR 1971 MOObL VEHICLES
HOUSTON-PHASE 2
MANU-
FACTURER
MAKE
ANC
CHRV CORP
PLYN
UOUGE
CHR»
FOOD KG CO.
(OHO
MEKC
LIMC
CM
CHEV
PONT
OLDS
BUICK
CAUI
VilLKS (. M1SC
TOTAL
*
N
*
* *
9
10
4
6
0
14
14
0
0
23
13
0
i
4
1
9
65
MEAN
MILES
(M
5.7
5.3
5.3
5.3
0.0
5.3
3.0
0.0
0.0
5.2
i. 1
0.0
>. 3
6.0
0.0
». »
5.3
* I BELOW
* STANDARDS
HT ORO CAK BUNS CM /Ml
ARITHMETIC
HC
33
10
0
IT
0
14
14
0
0
61
46
0
BO
75
0
67
40
CO
33
20
0
33
0
14
14
0
0
61
46
0
100
50
0
44
38
BOTH"
11
0
0
0
0
7
T
0
0
43
23
0
eo
50
0
33
23
HE AN >
2.27
2. 97
3.09
2.84
0.0
3.20
3.20
0.0
0.0
1.87
2.22
0.0
1.32
1.65
2.10
2.41
SO
0.53
0.73
0.53
0.88
0.0
1.15
1.15
0.0
0.0
0.89
0.85
0.0
0.51
1.06
0.48
1.00
*
»
GEOMETRIC *
MEAN
2.22
2.68
3.05
2.77
0.0
3.04
3.04
0.0
0.0
l.t>7
2.05
0.0
1.25
1.43
2.05
2.21
SO
1.27
1.32
1.20
1.3?
0.0
1.38
1.38
0.0
O.C
1.64
1.56
0.0
1.39
1.81
1.27
1.55
CARBON HON. 1X106 CM/HI
AKITHMEtIC *
MEAN
21.0
51.1
73.8
36.0
0.0
40.1
40.1
0.0
0.0
24.9
28.8
0.0
12.3
30.8
31.6
33. 1
SO
11. '
25.0
22.7
11.)
0.0
22.7
22.7
0.0
0.0
16.9
17.5
0.0
4.6
20.5
19.1
21.0
GEOMETRIC *
MEAN *
26.4
45.9
71.3
)4.3
0.0
35.2
35.2
0.0
0.0
20.8
24.8
0.0
11.6
26.7
26. »
29.1
SO *
1.65
1.63
1.35
1.42
0.0
1.71
1.71
0.0
0.0
1.83
1.77
0.0
1.48
1.80
0* 0
1.87
1.86
NOX CM/ HI '
<
ARITHMETIC *
MbAN
9.05
5.43
4.20
6.24
0.0
4.85
4.85
0.0
0.0
3.55
3.81
0.0
3.68
2.68
OB 0
J.09
4.27
SO »
1.66
2.1 1
1. 11
2.2H
0.0
2.24
2.24
0.0
0.0
1.11
1.08
0.0
1.51
0.18
0* 0
1.14
1.83
GEOMETRIC .
MEAN
4.80
5.09
4. 10
5.96
0.0
4.21
4.21
0.0
0.0
3.39
3.63
0.0
3.49
2.67
0*0
£.87
3.88
SJ '
1.41
1.46
1.29
1.49
0.0
1.85
1.85
0.0
0.0
1.37
1.41
0.0
1.42
1.07
0*0
1.53
1.58
* APPLICABLE STANOARDS:HC-2.4 G-N/MI CO-23. GN/NI «
TABLE 19
EMISSION LEVELS FCR 1971 MODEL VEHICLES
LUS ANGELES-PHASE 2
MANU-
FACTURER
NAKt
AMC
CMRY CORP
PLTM
DOOGE
CHR«
FORO MO CO.
FORD
MERC
LINC
CM
CHEV
PONT
OLDS
BUICK
CACI
VOLKS I NISC
TOTAL
*
K
* *
* »
a
23
12
b
5
30
24
6
0
40
22
6
4
6
i
11
112
Mt»N
MILES
[ K)
9.4
5.4
5.4
5.9
4.8
5.6
5.6
5.7
0.0
5.2
5.2
5.5
5.4
5.2
4.8
5.1
5.4
f. BELOU
STANDARDS
*
HC
100
7a
75
100
60
87
83
100
0
85
82
83
100
63
100
64
83
HYDROCARBONS GM/MI
ARITHMETIC
CO
38
43
33
67
4U
60
54
S3
0
83
82
67
100
83
100
9
58
BOTH*
36
43
33
67
40
57
50
83
0
75
73
67
100
67
100
9
54
MEAN *
1.04
1.75
1.82
1.31
2.13
1.47
1.57
1.10
o.o
1.74
2.07
1.67
0.98
1.45
0.72
1.93
1.68
SO
0.39
0.73
0.77
0.44
0.77
0.60
0.63
O.ls
0.0
1.74
2.21
0.55
0.77
1.02
0.19
u.76
1.16
GEOMETRIC
MEAN *
1.60
1.63
1.70
1.24
2.01
1.36
1.43
1.08
0.0
1.38
1.61
1.59
0.79
1.24
0.71
I.B2
L.48
SO
1.27
1.48
1.43
1.45
1.47
1.53
1.57
1.18
0.0
1.86
1.87
1.40
2.06
1.77
1.31
1.44
1.69
CARBON HONOXIDt GM/NI *
ARITHMETIC
MEAN
27.2
39.1
41.3
24.5
51.4
22.4
24.1
15.9
0.0
17.1
17.6
23.1
6.0
17.6
13.6
32.2
25. l
SO
14.1
27.5
29.0
15. 9
31.9
13.4
14.4
5.2
0.0
14.5
13.6
26.0
2.1
6.7
4.6
9.6
19.1
GEOMETRIC «
MEAN
22.9
30.3
31.9
20.4
43.0
19.1
20.3
15.1
0.0
13.1
14.0
14.1
5.7
16.6
13.2
30.9
19.9
SO *
2.05
2.13
2. 19
1.97
1.99
1.77
1.83
1.45
0.0
2.04
1.95
3.04
1.43
1.46
1.41
1.46
2.09
NOX GH/N1
ARITHMETIC *
MEAN *
2.94
3.78
3.77
4.28
3.19
3.58
3.59
3.56
0.0
3.44
3.62
2.98
4.33
2.69
3.34
^.37
3.41
sn «
1.03
1.57
1.59
2.14
0.25
1.15
1.10
1.46
0.0
0.99
0. 9B
0.71
1.21
0.67
0.37
0.94
1.22
*
GEOMETRIC
MEAN *
2.77
3.49
3.43
3.90
3.18
3.42
3.45
3.32
0.0
3.31
3.49
2.90
4.22
2.61
3.33
i.23
1.21
SO *
1. 46
1.52
1.61
1.59
1.08
1.35
1.32
1.51
0.0
1.33
1.32
I. 28
1.29
1.31
1.12
1.42
1.42
* APPLICABLE STANDARDS:HC-2.2 GM/M1 CO-23. GN/MI «
38
-------
TABLfc 20
EMISSION LEVELS FOR 1971 MODEL VEHICLES
DETROIT-PHASE 2
MANU-
FACTl/HER
HAKE
AMC
CHKY COOP
PLYN
DODGE
CHRV
FORD NU CO.
FORD
MERC
LINC
GN
CMtV
PONT
OLDS
8U1CK
CADI
VOL US (. XI SC
TOTAL
« t
H «
* *
0
9
4
3
2
IS
15
0
0
24
13
2
3
6
0
0
46
HEAN
MILES
CPU
O.O
6.0
6.3
6.2
4.9
5.2
3.2
a. a
0.0
5.6
5.3
6.3
6.6
5.4
0.0
0.0
5.6
» BELOh
STANCAKCS
HYDROCARBONS GM/MI
ARITHMETIC »
« HC
0
33
25
67
0
7
7
0
0
42
38
50
33
50
0
0
29
CO
0
22
25
33
0
47
47
0
0
46
46
50
67
33
0
0
42
BOTM«
0
22
25
33
0
7
7
0
0
29
31
50
33
17
0
0
21
MEAN
0.0
2.51
1.12
2.C8
2.72
3.13
3.13
0.0
0.0
2.64
2.89
2.03
3.05
2.08
0.0
O.U
2.77
SO *
0.0
O.T6
0.69
1.06
0.24
0.78
0.78
0.0
0.0
1.57
1.72
1.00
2.43
0.98
0.0
0.0
1.25
*
*
GEOMETRIC
MEAN
0.0
2. 39
2.65
1.92
2.71
3.0)
3.03
0.0
0.0
2.28
2.52
1.90
2.42
1.89
0.0
0.0
2.51
SD
0.0
1.40
1.31
1.61
1.09
1.29
1.29
0.0
0.0
1.72
1.71
1.61)
2.34
1.62
0.0
u.u
1.56
CARBON MONOXIDE GMAMI
ARITHMETIC
MEAN
u .u
40.9
41.6
28.7
58.1
30.1
30.1
0.0
0.0
28.7
29.8
52.3
19.7
23.1
0.0
0.0
31.5
SO
0.0
20.2
24.0
U.5
10. 9
14.2
14.2
0.0
0.0
21.3
20.1
55.4
16.1
13.4
a. a
0.0
19.6
GEOMETRIC «
MEAN
0.0
J5.2
33.8
26.8
57.6
27.0
27.0
0.0
0.0
22.5
24.9
34.6
15. 5
19.8
a.o
0. 0
25.9
so
0.0
1.90
2.33
1.56
1.21
1.64
1.64
0.0
0.0
.2.06
1.84
3.94
2.37
2.18
0.0
0.0
1.92
NOX GM/MI «
ARITHMETIC
MEAN
O.C
4.17
4.08
4.S8
3.2<5
4.91
4.91
0.0
0.0
2.98
3.00
1.99
3.15
2.93
0.0
u.
-------
Table 22
SUMMARY OF EMISSIONS (ARITHMETIC MEANS)
FOR 1971 MODEL VEHICLES
LOW MILEAGE ENGINES - PHASE I
HOUSTON
LOS ANGELES
DETROIT
DENVER
STABILIZED ENGINES - PHASE 2
HOUSTON
LOS ANGELES
DETROIT
DENVER
HYDROCARBONS
gms/mi
1.58
1.29
1.77
2.63
2.41
1.68
2.77
3.16'
CARBON MONOXIDE
gms/mi
27.1
19.7
28.5
64.0
33.7
25.2
31.5
45.2
NOX
xc
gms/mi
3.35
3.07
3.08
2.86
4.27
3.41
3.80
3.19
Table 23
SUMMARY OF SEVEN-MODE EMISSIONS TESTS 1968-1971 MODEL VEHICLES
MANU-
FACTURER
68 TOTAL
69 TOTAL
70 EXCLUDING
DENVER
70 DENVER
71 PHASE 1
EXCLUSING
DENVER
71 PHASE 1
DENVER
71 PHASE 2
EXCLUDING
DENVER
71 PHASE 2
DENVER
N
967
982
1829
352
225
144
225
144
MEAN
MILES
(K)
34.3
23.2
11.0
10.1
0.25
0.21
5.4
5.2
% BELOW
STANDARDS
HC
46
53
32
5
87
36
59
29
co (BOTH
38 23
42 30
35 19
3 .3
60 58
2 1
49 38
10 6
HYDROCARBONS gm/mi
ARITHMETIC
MEAN SO
4.13 3.16
3.66 1.99
2.83 1.53
4.25 2.39
1.47 0.74
2.63 1.03
2.12 1.22
3.16 4.01
GEOMETRIC
MEAN SD
3.66 1.56
3.38 1.46
2.60 1.48
3.92 1.45
1.34 1.55
2.46 1.46
1.86 1.68
2.71 1.56
CARBON VONOXIDEgm/mi
ARITHMETIC
MEAN SD
49.0 29.2
43.3 23.0
35.3 21.9
65.1 29.3
23.7 15.0
64.0 27.6
29.0 20.0
45.2 22.8
GEOMETRIC
MEAN SD
41.1 1.82
37.9 1.69
29.2 1.89
59.1 1.56.
19.5 1.89
58.5 1.54
23.0 2.02
39.8 1.71
NOy gm/mi
ARITHMETIC
MEAN SO
4.60 1.94
5.33 1.83
4.69 1.73
3.83 1.51
3.16 1.10
2.86 1.42
3.74 1.52
3.19 1.41
GEOMETRIC
MEAN SD
4.11 1.70
4.95 1.53
4.36 1.49
3.53 1.52
2.99 1.39
2.54 1.63
3.46 1.49
2.88 1.61
I '
40
-------
TABLE 24
TABULATION OF EMISSION LEVELS OF 1968 MODEL VEHICLES
BY ENGINE CIO
MANU-
FACTURER
232
CHRY CORP
225
318
363
440
FORD MO CO.
200
302
340
429
GN
230
290
307
327
390
396
430
455
472
VCLKS
91
97
N
*
41
36
36
40
38
50
37
36
38
43
43
48
80
143
37
37
31
31
40
49
MEAN
MILES
IK)
29.9
35.8
41.7
38.6
36.8
32.5
37.6
38.4
32.9
33.8
29.8
32.3
32.5
32.7
33.6
35.4
36.1
40.4
30.2
12.2
* » BELOM
STANDARDS
HC
20
50
36
38
29
62
24
93
45
40
84
4-
13
55
38
61
68
87
26
71
CC
32
39
33
23
3
76
13
53
11
28
63
36
6
19
22
5
29
58
40
82
*
BOTH*
12
25
19
13
0
56
19
39
5
26
58
2
1
14
16
6
29
55
18
65
HYDROCARBONS PPM *
ARITHMETIC *
MEAN *
347.
334.
369.
361.
346.
293.
355.
274.
323.
297.
2*0.
485.
440.
284.
534.
263.
277.
211.
669.
422.
SO
134.
227.
397.
313.
152.
153.
119.
42.
166.
81.
139.
187.
213.
100.
810.
52.
151.
Ilk.
579.
373.
GEOMETRIC
MEAN *
330.
294.
326.
316.
322.
276.
339.
242.
299.
266.
219.
456.
404.
272.
364.
258.
252.
194.
542.
350.
SO »
1.36
1.58
1.60
1.54
1.44
1.34
1.35
1.35
1.42
1.32
1.49
1.42
1.48
1.34
1.99
1.21
1.52
1.4.
1.31
1.71
CARBON MONOXIDE *
ARITHMETIC *
MEAN
£.12
1.96
1.82
2.24
2.97
1.64
1.27
1.53
2.64
1.92
1.45
1.74
3.26
2.23
-2.47
2.30
2.32
1.50
2.74
1.68
SO *
1.13
1.20
0.85
1.01
1.61
1.53
0.52
0.78
1.00
0.82
0.70
0.58
1.57
0.8B
1.34
0.56
1.40
0.66
1.44
1.25
*
GEOMETRIC
MEAN
1.83
1.71
1.64
2.03
2.68
1.28
1.17
1.32
2.44
1.73
1.29
1.65
2.92
2.05
2.17
2.23
1.93
1.45
2.47
1.33
SO *
1.79
1.66
1.64
1.6(1
1.55
1.89
1.57
1.82
1.53
1.63
1.64
1.39
1.60
1.52
1.70
1.29
1.91
1.50
1.55
1.97
NOX PPM *
1
ARITHMETIC *
MEAN *
1384.
1677.
1457.
1524.
961.
1912.
1647.
1489.
1003.
1279.
1323.
1572.
90S.
1166.
931.
1206.
996.
1373.
1299.
1065.
SO *
563.
623.
457.
493.
486.
721.
478.
487.
339.
392.
407.
375.
462.
409.
465.
374.
364.
430.
436.
483.
GEOMETR 1C i
MEAN *
1255.
1495.
1359.
1443.
804.
1654.
1563.
1402.
930.
1216.
1262.
1523.
768.
1098.
821.
1160.
911.
1300.
1 188.
942.
SD '
1.64
1.81
1.54
1.41
2.03
1.97
1.42
1.44
1.55
1.39
1.37
1.30
1.92
1.43
1.69
1.32
1.60
1.42
1.64
1.73
APPLICABLE STANDARDS:HC-275PPM CO-l.St, U10PPM.2.3S FOR VOLKS I
TABLE 25
TABULATION OF tMISSION LEVELS OF 1969 MODEL VEHICLES
BY ENGINE C1D
MANU-
FACTURER
232
CHRV CORP
225
316
303
44O
FORO MO CO.
200
302
351
390
429
GM
230
250
307
350
3S6
428
430
455
472
*
N
*
*
32
39
44
37
35
26
42
74
44
32
31
35
33
163
28
65
35
36
40
MEAN
MILES
(K)
19.1
19.7
23.9
26.9
20.2
27.2
23.7
21.8
26.4
21.4
27.3
25.2
21.5
22.8
19.5
27.6
22.6
25.7
16.1
* t BELOH *
> STANDARDS *
*
HC
25"
59
41
46
51
62
7
31
SO
50
58
89
42
50
39
28
77
86
100
* CO
50
41
32
46
26
73
64
59
52
25
39
63
21
21
14
17
26
50
80
BOTH*
9
31
23
32
23
58
5
26
41
19
35
57
18
IT
11
6
26
50
ao
HYDROCARBONS PPM
ARI THME
MEAN *
350.
296.
304.
306.
349.
293.
465.
328.
295.
217.
324.
208.
309.
282.
331.
331.
243.
216.
159.
TIC »
SO
135.
106.
77.
121.
294.
112.
42ll
1124
130.
111.
244.
73.
79.
75.
163.
102.
33.
.47.
30.
liEOMET
MEAN *
332.
283.
295.
286.
300.
275.
393.
312.
279.
27(1.
276.
197.
300.
272.
304.
317.
23B.
209.
157.
*
KIC
SO *
1.37
1.32
1.28
1.44
1.60
1.42
1.66
1.36
1.35
1.40
1.64
1.39
1.28
1.33
1.45
1.34
1.22
1.29
1.19
CARBON MONOXIDE X *
*
ARITHHE
MEAN *
1.42
1.97
2.00
1.87
2.34
1.31
1.45
1.48
1.60
2.27
1.88
1.45
2.06
2.17
2.19
2.37
1.92
1.97
1.26
TIC
S3 »
0.58
0.97
0.89
1.20
1.18
0.49
0.58
0.68
Q..83
1.13
1.02
0.54
0.77
0.87
0.71
0.96
0.6&
1.59
0.38
GEOHET
MEAN *
1.29
1.76
1.79
1.53
2.04
1.23
1.35
1.34
1.41
1.98
1.66
1.35
1.95
2.00
2.08
2.21
1.80
1.63
1.20
RIC «
SO
1.63
1.63
1.65
1.77
1.76
1.44
1.50
1.57
1.69
1.76
1.65
1.50
1.40
1.51
1.41
1.45
1.47
1.74
1.36
ARITHHE
MEAN *
1805.
1849.
1536.
1457.
1154.
1867.
1880.
1572.
1503.
1006.
1075.
1129.
1420.
1444.
1091.
1265.
1450.
1350.
1441.
NOX PPN
TIC *
SO *
422.
505.
508.
484.
494.
523.
434.
476.
510.
353.
365.
386.
325.
440.
349.
405.
402.
484.
373.
GECMET
MEAN *
1755.
1759.
1436.
1363.
1048.
1783.
1826.
1460.
14O4.
926.
1009.
1061.
1378.
1363.
1032.
1185.
1386.
1177.
1388.
*
*
RIC *
SO *
1.28
1.43
1.49
Ii49
1.58
1.39
1.29
1.46
1.49
1.57
1.46
1.44
1.30
1.46
1.42
1.61
1.38
1.96
1.33
VOLKS
23.7 58 78 53 469. 316. 409. 1.64 2.01 0.99 1.38 1.39
*» APPLICABLE STANDARDS3HC-273PPM CO-l.St I410PPM.2.3I FOR VOLKS I «
41
-------
TABLE 26
TABULATION OF EMISSION LEVELS OF 1970 MODEL VEHICLES
BY ENGINE CIO«EXCLUOING DENVER
MANU-
FACTURER
CIO
AMC
212
CHRY CORP
225
lit
383
FORD NO CO.
200
102
151
190
429
CM
210
290
JOT
150
400
455
472
VOLKS L HISC
97
N
*
59
0
1O4
72
120
11
101
*l
37
11
H
Ml
1T1
92
lit
15
100
MEAN
MILES
IKI
10.0
a. 9
10.1
10.4
14.9
9.3
9.9
10.2
10.4
13.1
12.2
11.2
10.6
11.0
12.6
11.4
9.4
* S BELOh *
* STANDARDS *
« HC
42
26
45
10
25
32
1
29
10
71
95
16
27
2*
5*
49
48
* CO
51
16
60
14
49
44
29
31
22
24
25
25
21
34
40
9
71
BOTH*
36
11
19
4
21
17
5
20
16
19
19
10
12
IT
14
9
41
HYDROCARBONS GN/NI
ARITNNE
MEAN
2.42
2.88
2.39
3.53
2.T3
3.24
3.14
2.81
2.73
2.11
2.55
3.46
2.91
2.98
2.20
2.18
2.53
TIC
SO *
0.82
0.86
0.65
1.77
0.79
3.78
2.05
1.09
0.91
0.70
2.04
1.86
1.19
1.69
0.98
0.96
1.15
CEOMET
MEAN *
2.10
2.76
2.12
3.28
2.64
2.68
3.09
2.67
2.60
2.01
2.07
1.16
2.71
2.71
2.02
2.27
2.33
*
*
RIC
SO *
1.37
1.34
1.29
1.43
1.29
1.61
1.40
1.44
1.36
1.37
1.85
1.49
1.45
1.4>
1.52
1.32
1.49
CARBON MONOXIDE OH/11 *
*
ARITHME
HERN
28.1
44.4
24.5
56.0
27.7
27.3
36.4
16.7
16.0
31.2
2«.6
16.2
43.8
37.4
11.8
17.2
20. a
TIC *
SO *
18.5
22.6
14.7
28.9
18.4
13.8
20.3
20.5
16.8
14.2
11.0
17.5
22.8
Z1.7
21.6
11.2
9.3
GEONE T
MEAN *
22.3
19.0
20.7
48.2
22.5
24.0
31.2
31.2
32.1
28.0
26.2
11.5
37.8
30.6
26.1
14.7.
19.1
RIC
SO *
2.05
1.70
1.84
1.81
1.94
1.69
1.78
1.80
1.65
1.64
1.58
1.77
1.77
1.95
1.90
1.49
1.50
ARITHME
MEAN *
5.10
7.19
5.61
5.4*
5.07
3.34
5.25
S.ll
4.29
1.14
1.61
3.81
4.29
4.12
5.02
6.45
3.91
NOX GM/M1 »
TIC "
SD *
1.40
1.8S
1.39
1.47
1.93
1.07
1.36
1.37
1.18
0.89
i.4i
1.16
1.39
1.18
1.74
1.69
1.04
GEONET
MEAN *
4.89
6.92
' 5.45
5.20
4.62
3.17
5.07
5.13
4.10
3.24
3.36
3.62
4.05
3.91
4.68
6.21
3.76
RIC «
SO
1.16
1.14
1.10
1.43
1.59
1.19
1.31
1.33
1.36
1.291
1.46
1.39
1.41
1.42
1.48
1.14
1.12
*** APPLICABLE STANDARDS:HC-2.2 GM/MI CO-21. SM/MI **
TABLE 27
TABULATION OF EMISSION LEVELS OF 1970 MODEL VEHICLES
BY ENGINE CID**CENVER ONLY
MANU-
FACTURER
CIO
ANC
232
CHRY CORP
225
318
383
FORD MO CO.
30Z
351
429
GN
250
307
350
400
45S
YOLKS t MISC
97
N
* *
*
11
13
20
19
19
20
11
20
17
70
21
25
20
MEAN
MILES
m
10.3
.0
.2
.1
.5
.8
11.7
9.6
9.3
10.8
10.0
10.5
B.»
* « BELOh *
* STANDARDS *
» HC
0
0
5
0
0
0
9
45
0
1
5
8
10
* CO
27
0
0
0
0
5
0
0
0
1
5
0
0
BOTH*
0
0
0
0
0
0
0
0
0
1
0
0
0
HYDROCARBONS GM/MI
ARITHME
MAN *
3.42
4.67
3.62
4.29
4.03
4.18
3.81
3.19
5.36
4.77
3.95
3.16
1.39
TIC *
SD
0.80
1.92
1.07
0.85
1.23
1.49
1.50
2.01
1.99
1.33
2.00
0.89
1.09
GEOMET
MEAN *
3.34
4.41
1.47
4.22
3.88
3.98
3.58
2.77
5.08
4.59
3.64
3.05
3.23
RIC
SD
1.25
1.39
1.34
1.21
1.31
1.36
1.46
1.67
1.39
1.12
1.47
1.19
1.37
CARBON MONOXIDE Gt/tl »
ARITHME
MEAN
44.5
81.5
66.1
77.7
46.8
56.3
59.0
44.4
60.4
79.1
62.3
72.4
41.3
TIC *
SD *
18.9
20.5
31.9
28.2
12.8
23.7
19.7
15.1
13.3
26.2
28.5
20.3
9.0
GEOMET
MEAN
40.4
79.3
60.8
72.6
45.4
51.8
55.9
42.0
59.0
74.4
56.5
69.4
40.3
RIC *
SD *
1.62
1.28
1.49
1.48
1.29
1.54
1.43
1.40
1.24
1.45
1.58
1.37
1.25
ARITHME
MEAN
4.36
5.37
3.58
4.17
3.66
4.10
3.82
3.29
3.60
3.46
4.02
3.82
4.16
NOX. GM/MI «
<
TIC
SD *
1.76
1.60
1.37
1.39
1.39
1.27
1.61.
1.05
0.76
1.25
2.06
0.96
1.25
GEOMET
MEAN *
4.03
5.09
1.32
1.90
3.44
3.84
3.49
3.15
3.51
3.22
3.39
3.71
4.00
RIC <
SD «
1.52
1.44
1.51
1.49
1.46
1.52
1.58
1.35
1.24'
1.48
1.61
1.29
1.32
APPLICABLE STANOARDSJHC-2.2 GM/MI CO-21. GM/MI *
42
-------
Table 28
MANUFACTURER/MAKE CID GROUPINGS
MANUFACTURER
MAKE
AMC
CHRYSLER CORP.
FORD MOTOR CORP
CHEVROLET
PONTIAC
OLDSMOBILE
BUICK
CADILLAC
VOLKSWAGEN
1968-1969
232
225
318
383
440
220
302
351
390
429
230
250
307
327
350
396
350
400
428
250
350
455
350
430
472
91
97
1970
199
232
304
360
390
198
225
318
340
383
449
170
200
240
250
302
351
360
390
428
429
460
230
250
292
307
350
400
402
454
250
350
400
455
250
350
455
250
350
455
472
500
97
103
1971
232
258
304
360
198
225
318
340
360
449
98
122
170
200
240
250
302
351
360
390
400
460
140
250
292
307
350
400
402
454
350
400
455
350
455
350
455
472
500
97
103
43
-------
TABLE 29
MILfcAGE EFFFCTS FOR l<368 AND 1969 MOOtL VEHICLES
*
*
*
*
*
HC
GM/MI
CO
CM/ HI
NOX
GM/MI
YEAH
TOTAL
TOTAL
TOTAL
TOTAL
TOTAL
TOTAL
68
69
68
69
68
69
*
*
* N
*
*
313
318
313
318
313
318
*
* SIGNI-
*FICANCE
* LtyEL
* *
95
95
*
95
*
*
*
* INTERCEPT
*
GM/MI
* MEAN * STANDARD
* *
3.03
2.94
43.6
37.7
5.11
5.60
FRROK
0.558
0. ?29
5.81
2.36
0.338
0.247
SLflPF
MEAN
0.031
0.030
0.241
0.215
-0.016
-0.013
GM/MI /1000 MI
* STANDARD
* ERROR
0.015
0.013
0.180
0.094
0.010
0.010
* CHANCE
1000 MI
0.98
0.98
-
0.56
-
-
*
*
*
*
* SLOPE NOT SIGNIFICANUV DIFFERENT FROM ZERO AT AT LEAST 95? LEVEL
1968 MEAN MILEAGE = 34.6 K
1969 MEAN MILEACE - 24.0 K
HC
GM/MI
CO
CM/HI
NOX
GM/MI
* *
* *
* CITY *
* *
* *
HOUSTON
LOS ANGELES
DETROIT
DENVER
TOTAL
EXCLUDE DENVER
MOUSrCN
LOS ANGELES
DETROIT
DENVER
TOTAL
EXCLUDE DENVER
HOUSTON
LOS ANGELES
DETROIT
DENVER
TOTAL
N
65
112
48
144
225
65
112
48
144
225
64
112
48
144
224
*
* SIGNI-
FICANCE
* LEVEL
* *
99
99
99
*
99
99
99
*
99
99
99
99
99
99
99
INTEI
MEAN
1.58
1.29
1.76
2.63
1.47
27.1
19.7
28.5
64.0
23.7
3.37
3.07
3.08
2.86
3.16
TABLE 30
MILEAGE EFFECTS BY CITY IS71 MODEL VEHICLES
CEPT GM/MI * SLOP? GM/
*
* STANDARD * MEAN *
* ERROR * *
0.069
0.076
0.096
0.066
0.049
1.71
1.31
2.40
2.31
I. 00
0.169
0.169
0.173
0.119
0.073
0.158
0.079
0.191
0.106
0.125
1.25
1.13
0.63
-3.75
1.06
0.182
0.182
0.134
0.068
0.113
/1000MI
ANDARO
RROR
0.020
0.019
0.033
0.075
0.014
0.452
0.347
0.572
0.392
0.248
0.030
0.030
0.024
0.016
0.014
( CHANGE
1000 MI
7.14
4.90
7.54
-
6.35
3.18
4.66
-
-7.65
3.78
4.45
4.45
3.71
2.18
3.13
*
*
*
*
EXCLUDE DENVER
* SLOPE NOT SIGNIFICANTLY DIFFERENT FROM ZERO AT AT I EAST 95% LEVEL
44
-------
TABLE 31
MILEAGE EFFtCTS FUR 1S71 MQOtL VEHICLES
EXCLUDING DENVER
*
*
*
*
*
HC
GM/MI
CO
GM/MI
NOX
GM/MI
MAKE
AMC
PLYM
FORO
CHEV
BUICK
CADI
VOLKS
FORD
VOLKS
AMC
PLYM
DODGE
FORO
CHEV
OLOS
auiCK
CADI
r
*
* N
*
*
17
20
53
48
Id
3
20
53
20
17
20
15
53
47
12
16
3
*
* i IGNl-
*FICANCE
* LtVEL
* *
99
99
99
99
95
99
95
99
99
99
95
99
99
99
95
99
95
*
*
* INTFRCEPT GM/MI *
*
* MFAN
*
1.50
1.67
1.71
1.29
1.00
0.56
1.61
22.2
25.4
2.90
3.24
4.19
3.66
2. 84
2.78
2.35
2.43
*
* STANDARD *
* ERROR *
0. 161
0. 101
0.088
0.075
O.C73
0.071
0.1C3
1.91
2. 54
0.269
0.251
0.383
0.17?
0. 104
O.C80
0.108
0.113
SLOPF GM/MI/1000MI
MEAN *
*
0.-091
0.115
0.13-)
0.196
0.154
0.046
O.C74
1 .35
1.29
0.222
0.114
0.196
0.129
0.127
0.207
0.076
0.245
STANDARD
FRPOR
0.029
0.037
0.024
0.039
0.055
0.004
0.028
0.425
0.405
0.063
0.049
0.064
0.027
0.027
0.072
0.019
0.055
*
* t CHANGE
*
* 1030 MI
*
4.19
5.38
6.12
9.46
9.56
6.18
3.^7
4.88
4.22
5.87
3.39
3.93
3.03
3.80
5.74
2.57
7.18
*
*
*
*
*
HC
GM/HI
CO
GM/MI
NOX
OH/MI
MAKE
NONE
AMC
PLYM
FORO
CHEV
PONT
OLOS
BUICK
VOLKS
CHEV
OLOS
TABLE 32
MILEAGE EFFECTS FOR 1S71 MODEL VEHICLES
DENVER ONLY
* *
* * SIGNl-
* N *FICANCE
* * LEVEL
* * 3!
INTERCEPT GM/MI
SLOPE CM/MI/1000MI
MbAN
+ STANCAPO * MEAN * STANDARD
* FHRC1R * * ERROR
* CHANGE »
1030 MI *
20
20
24
34
5
6
9
23
34
6
95
95
99
99
99
99
95
99
99
99
74.2
77.0
51.6
76. 5
46.0
66.5
60. 6
44.4
2. 13
2.72
1 8.82
' 7.4S
2.59
4.36
6. 84
c.92
9.
-------
Table 33
SIX-CITY SURVEILLANCE PROGRAM-1957-1971 MODEL VEHICLES
OBJECTIVES:
BASIC TEST PROCEDURES:
INSTRUMENTATION
CITIES:
TOTAL VEHICLES TESTED:
DETERMINE CONTRIBUTION TO AIR POLLUTION BY VEHICLE
POPULATION RESPONSIBLE FOR 95% OF VEHICLE EXHAUST
EMISSION (CIRCA 1971); COLLECT EMISSIONS DATA DURING
STEADY-STATE MODES, ACCEL/DECEL DRIVING CYCLE;
EVALUATE CITY EFFECTS; COLLECT EVAPORATIVE
EMISSIONS DATA.
1972 FTP (CVS-C) AND 1975 FTP (CVS-CH). IN ACCORDANCE WITH
FEDERAL REGISTER, VOL. 35, NO. 219, NOV. 10. 1970; VOL. 36.
NO. 55. MARCH 20. 1971; VOL. 36. NO. 128, JULY 2, 1971. EACH
VEHICLE TESTED ONCE (EXCEPT FOR SUBSET OF REPLICATES).
EVAPORATIVE TESTS BY SHED TECHNIQUE (SAE J71). VEHICLES
TESTED IN "AS-RECEIVED" CONDITION AND ANY MEASUREMENT
OF IDLE SPEED TIMING, ETC. MADE ONLY AFTER VEHICLE
TESTED.
HC - FLAME IONIZATION ANALYZER
CO, CO2 - NON-DISPERSIVE INFRARED (NDIR)
NO, NOX - CHEMILUMINESCENT ANALYZER
HOUSTON. LOS ANGELES. DENVER, CHICAGO, ST. LOUIS, AND
WASHINGTON. D. C.
1020
46
-------
Table 34
COMPOSITE EMISSION LEVELS AS DETERMINED BY 1972 TEST PROCEDURES (EXCLUDING DENVER)
YEAR
PRE-68
67
58
59
60
61
62
63
64
65
66'
67'
TOTAL PRE-68'
CALIFORNIA
66
67
CONTROLLED
1968
1969
1970
1971
TOTAL 68-71
N
20
20
21
16
23
38
55
64
80
67
54
458
16
17
B4
89
86
101
360
MEAN
MILES
(K)
80.0
86.7
79.7
65.4
67.6
72.7
77.3
72.1
62.9
61.8
54.6
685
65.7
56.4
46.4
39.5
28.7
15.6
31.9
% BE LOW
LEVEL 1
HC
0
5
0
13
17
5
0
0
1
2
0
2
19
12
20
12
30
61
32
CO
10
5
19
13
17
8
2
3
4
3
2
5
19
6
25
10
33
46
29
NOy
*c
40
50
33
50
61
40
47
42
53
45
46
46
38
53
35
11
9
13
17
HYDROCARBONS gm/mi
ARITHfl
MEAN
7.07
10.74
11.37
9.53
6.49
9.57
10.26
8.14
12.17
9.20
8.19
9.56
8.72
6.22
6.25
5.96
4.41
3.44
4.96
AETIC
SD
2.32
10.65
8.70
8.48
3.64
6.87
8.11
3.70
13.29
6.18
3.61
8.06
8.S4
3.52
7.10
4.72
2.18
1.38
4.50
GEOMI
MEAN
6.72
8.58
9.19
7.40
5.69
7.83
8.85
7.49
9.24
8.07
7.51
8.02
6.62
5.52
5.13
5.19
4.06
3.21
4.27
TRIC
SD
1.39
1.81
1.89
2.01
1.67
1.88
1.62
1.49
1.89
1.61
1.51
1.70
2.00
1.63
1.70
1.59
1.47
1.45
1.62
CARBON MONOXIDE gm/mi
ARITHI
MEAN
80.3
85.2
81.6
86.3
84.6
85.5
105.8
91.0
97.9
103.3
103.5
95.2
78.1
81.4
78.6
73.8
56.S
47.5
63.5
IIIETIC
SD
33.2
38.6
45.0
44.8
59.8
38.8
41.8
41.4
40.7
43.2
47.7
43.5
38.3
38.0
59.0
36.2
27.2
27.2
40.9
GEOM
MEAN
73.5
76.5
70.7
75.0
66.5
77.5
98.1
82.3
90.0
94.3
94.5
85.6
70.2
74.6
63.8
66.0
50.2
40.4
53.5
ETRIC
SD
1.57
1.65
1.74
1.80
2.06
1.60
1.49
1.60
1.52
1.56
1.54
1.61
1.61
1.52
1.86
1.62
1.66
1.80
1.80
NOy gm/mi
*c
ARITH
MEAN
3.89
3.70
4.47
3.94
3.12
3.34
3.63
3.61
3.27
3.44
3.26
3.51
3.23
3.30
4.22
5.32
4.93
4.75
4.81
METIC
SD
2.15
2.44
2.39
2.37
2.25
1.57
2.15
1.71
1.66
1.71
1.45
1.87
1.44
1.45
1.88
2.01
1.64
1.73
1.85
GEOM
MEAN
3.24
2.92
3.71
3.26
Z53
2.53
3.07
3.19
£87
3.01
2.90
3.01
£91
2.98
3.75
4.93
4.66
4.42
4.43
ETHIC
SD
1.98
2.14
1.99
1.34
.97
.75
.81
.70
.70
.75
.71
1.79
1.64
1,61
1.69
1.50
1.41
1.48
1.54
EXCLUDING CALIFORNIA 66-67
'HC
CO
N°XC
3.4 gm/mi
39.0 gm/mi
3.0 gm/mi
Table 35
COMPOSITE EMISSION LEVELS AS DETERMINED BY 1972 TEST PROCEDURE (DENVER ONLY)
YEAR
TOTAL PRE-68
1968
1969
1970
1971
TOTAL 68-71
N
97
18
17
17
20
72
MEAN
MILES
IK)
65.1
42.1
38.9
26.0
15.1
30.1
% BELOW
LEVEL*
HC
0
0
12
6
0
4
CO
1
0
6
0
5
3
NOy
*C
83
83
65
59
55
65
HYDROCARBONS gm/mi
ARITHMETIC
MEAN
11.31
8.74
7.74
7.85
6.80
7.73
SD
6.13
4.08
4.89
4.23
2.08
3.89
GEOMETRIC
MEAN
10.27
8.00
6.49
6.91
6.52
6.93
SD
1.51
1.53
1.89
1.70
1.35
1.62
CARBON MONOXIDE gm/mi
ARITHMETIC
MEAN
136.8
122.9
92.6
111.2
102.7
106.4
SD
55.5
66.1
57.7
39.8
40.6
52.0
GEOMETRIC
MEAN
125.8
109.9
79.7
103.4
94.5
95.6
SD
1.53
1.60
1.72
1.45
1.54
1.59
NOv gm/mi
Ac
ARITHMETIC
MEAN
1.93
2.38
2.52
2.72
3.06
2.68
SD
1.11
1.11
1.21
1.13
1.56
1.28
GEOMETRIC
MEAN
1.66
2.19
2.20
£48
2.75
2.41
SD
1.76
1.50
1.78
1.59
1.59
1.62
HC - 3.4 gm/mi
CO 39.0 gm/mi
NOXC ~ 3-° Bm/mi
47
-------
Table 36
COMPOSITE EMISSION LEVELS AS DETERMINED BY 1975 TEST PROCEDURES (EXCLUDING DENVER)
YEAR
pRE-ea
57
58
69
60
61
62
63
64
65
66'
67'
TOTAL PRE-68'
CALIFORNIA
66
67
CONTROLLED
1968
1969
1970
1971
TOTAL 68-71
N
20
20
21
16
23
38
55
64
80
67
54
458
16
17
84
89
86
101
360
MEAN
MILES
IK)
80.0
86.7
79.7
65.4
67.6
72.7
77.3
72.1
62.9
61.8
54.6
68.5
65.7
56.4
46.4
39.5
28.7
15.6
31.9
% BELOW
LEVEL t
HC
0
5
0
13
26
5
0
2
0
5
2
4
19
24
31
20
41
70
42
CO
6
10
19
19
17
8
2
13
3
10
6
8
25
12
34
21
44
60
40
NOX
*e
40
45
38
56
61
49
S3
39
49
42
48
46
31
41
26
10
n
15
15
HYDROCARBONS gm/mi
ARITHft
MEAN
6.63
10.03
10.80
8.79
5.94
8.87
9.43
. 7.28
11.18
8.26
7.38
8.74
7.84
5.33
5.54
5.19
3.90
3.06
4.37
AETIC
SD
2.32
10.20
8.24
8.09
3.27
6.84
7.76
3.29
12.62
5.48
3.28
7.63
8.34
3.52
7.07
4.26
1.95
1.26
4.30
GEOM
MEAN
6.25
7.93
8.70
6.82
5.20
7.09
8.04
6.68
8.36
7.21
6.75
7.26
5.81
4.60
4.45
4.53
3.60
2.85
3.75
ETRIC
SD
1.42
1.84
1.88
1.98
1.68
1.94
1.65
1.50
1.92
1.63
1.52
1.73
2.03
1.70
1.73
1.56
1.46
1.44
1.61
CARBON MONOXIDE gm/mi
ARITH
MEAN
81.4
78.2
77.3
81.6
79.7
78.0
96.5
81.7
87.9
91.0
93.6
86.5
65.2
67.2
67.8
61.7
48.2
40.1
53.9
METIC
SD
30.7
36.8
45.9
43.4
56.8
36.6
38.4
38.B
36.B
38.9
44.9
40.3
36.6
37.0
57.5
31.0
24.7
24.5
37.9
GEOM
MEAN
75.6
69.7
65.8
70.0
61.8
69.6
89.2
73.1
80.8
82.3
86.1
77.4
56.B
59.7
E2.7
55.0
42.0
33.5
44.4
ETRIC
SD
1.60
1.67
1.79
1.83
2.12
1.67
1.50
1.64
1.61
1.60
1.55
1.63
1.72
1.63
1.96
1.63
1.72
1.86
1.86
NOy gm/mi
*c
ARITH
MEAN
3.83
3.62
4.49
3.94
3.06
3.33
3.64
3.66
3.37
3.57
3.28
3.54
3.40
3.42
4.34
5.45
5.05
4.81
4.92
METIC
SD
2.19
2.31
2.55
2.48
2.3S
1.67
2.16
1.77
1.66
1.82
1.46
1.91
1.54
1.50
1.92
2.02
1.67
1.78
1.88
GEOM
MEAN
3.17
2.88
3.6*
3.18
2.43
2.91
3.08
3.24
2.97
3.13
2.93
3.04
3.04
3.08
3.85
5.06
4.78
4.47
4.52
ETRIC
SD
1.99
2.12
2.08
2.02
2.04
1.76
1.82
168
1.70
1.72
1.68
1.79
1.68
1.65
1.71
1.49
1.40
1.49
1.54
EXCLUDING CALIFORNIA 66-67
HC
CO
3.4 gm/mi
39.0 gm/mi
3.0 gm/mi
Table 37
COMPOSITE EMISSION LEVELS AS DETERMINED BY 1975 TEST PROCEDURE (DENVER ONLY)
YEAR
TOTAL PRE-68
1968
1969
1970
1971
TOTAL 68-71
N
97
18
17
17
20
72
' MEAN
MILES
(K)
66.1
42.1
38.9
26.0
15.1
30.1
% BELOW
LEVEL*
HC
0
0
18
12
5
8
CO
0
0
12
0
5
4
NO*,
Ac
83
83
65
53
50
63
HYDROCARBONS gm/mi
ARITHMETIC
MEAN
10.16
7.34
6.31
6.71
5.68
6.46
SD
5.59
2.73
3.47
3.85
1.45
2.97
GEOMETRIC
MEAN
9.24
6.87
5.43
5.93
5.51
5.89
SD
1.49
1.46
1.84
1.66
1.29
1.67
CARBON MONOXIDE gm/mi
ARITHMETIC
MEAN
126.9
109.2
76.4
94.8
91.1
92.2
SD
48.5
52.6
47.7
33.8
37.6
43.7
GEOMETRIC
MEAN
117.4
99.7
65.6
89.3
83.1
83.0
SD
1.61
1.53
1.74
1.43
1.68
1.60
NOv Bm/mi
*c
ARITHMETIC
MEAN
1.39
2.20
2.59
2.78
3.08
2.67
SD
1.12
0.80
1.24
1.11
1.60
1.25
GEOMETRIC
MEAN
1.61
2.07
2.27
2.55
2.75
2.40
SD
1.79
1.43
1.76
1.67
1.62
1.61
HC
co
NOv
3.4 gm/mi
39.0 gm/mi
3.0 gm/mi
48
-------
Table 38
EMISSION LEVELS AS AFFECTED BY GEOGRAPHIC LOCATION
1972 AND 1975 CVS TEST PROCEDURES
YEAR
PRE CONTROL
EXCLUDING DENVER
PRE CONTROL
DENVER
CONTROLLED121
EXCLUDING DENVER
CONTROLLED
DENVER
N
4S8
97
393
72
MEAN
MILES
IK)
68.5
65.1
34.3
30.1
HYDROCARBONS gm/mi
MEAN<1> SO
9.S6 B.06
11.31 6.13
5.16 4.50
7.73 3.88
76*
MEAN SD
8.74 7.63
10.16 5.59
4.54 4.30
6.46 2.97
CARBON MONOXIDE gm/mi
72
MEAN SD
95.2 45.4
136.8 55.5
64.8 41.0
106.4 52.0
75
MEAN SD
86.5 40.3
126.9 48.6
54.9 38.0
92.2 417
N0>e am/mi
72
MEAN SD
3.61 1.87
1.93 1.11
4.77 1.88
2.68 1.28
75
MEAN SD
3.54 1.91
1.89 1.12
4.79 1-90
2.67 1.25
(1) ARITHMATIC MEAN & SD
(2) INCLUDES 1966, 1967, LOS ANGELES
" 72 TEST PROCEDURE
t 75 TEST PROCEDURE
49
-------
Table 39
EMISSION LEVELS BY VEHICLE WEIGHT AS DETERMINED BY 1975
TEST PROCEDURE EXCLUDING DENVER AND DENVER ONLY
EXCLUDING DENVER
VEHICLE WEIGHT (1b)
1501 - 2000
2001 - 2500
2501 - 3000
3001 - 3500
3501 - 4000
4001 - 4500
4501 - 5000
5001 - 5500
N
54
33
132
192
263
141
31
5
HC-gm/mi
ARITHM
MEAN
6.70
3.67
6.35
6.88
7.91
5.96
6.39
5.69
IETIC
S.D.
5.2
2.6
7.4
6.9
7.7
3.7
8.5
3.4
CO-gm/mi
ARITHIV
MEAN
52.53
42.31
62.51
73.70
80.33
74.48
84.81
64.40
IETIC
S.D.
23.30
28.70
36.80
40.6
47.1
40.1
48.8
37.6
NO gm/mi
ARITHIV
MEAN
2.53
3.76
3.89
4.00
4.09
4.99
4.85
5.98
IETIC
S.D.
1.5
2.0
1.8
1.9
2.0
2.0
2-4
2.1
DENVER
VEHICLE WEIGHT (1b>
1501 2000
2001 - 2500
2501 - 3000
3001 3500
3501-4000
4001-4500
4501-5000
5001 - 5500
N
12
5
23
48
56
19
4
2
HC-gm/mi
ARITHM
MEAN
6.16
5.11
7.67
8.57
9.89
9.32
5.34
5.04
ETIC
S.D.
1.9
3.6
3.3
3.7
6.7
4.9
1.5
1.9
CO-gm/mi
ARITHIV
MEAN
78.12
48.15
106.52
103.00
124.24
140.61
108.16
155.20
ETIC
S.D.
16.9
19.3
37.7
42.9
52.2
60.3
45.9
30.8
NOxcgm/mi
ARITHIV
MEAN
1.19
2.14
1.94
2.31
2.37
2.42
2.67'
2.32
IETIC
S.D.
0.4
1.3
1.1
1.3
1.3
1.1
0.5
1.2
50
-------
Table 40
ACCELERATION/DECELERATION MODES AND DRIVING SEQUENCE
\ . MODE
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
TYPE
ACCEL
DECEL
ACCEL
ACCEL
ACCEL
DECEL
ACCEL
DECEL
ACCEL
DECEL
ACCEL
DECEL
ACCEL
DECEL
DECEL
DECEL
ACCEL
DECEL
ACCEL
DECEL
ACCEL
DECEL
ACCEL
ACCEL
DECEL
DECEL
ACCEL
DECEL
ACCEL
ACCEL
DECEL
DECEL
SPEED RANGE
(mph)
0-30
30-0
0-15
15-30
30-45
45-30
30-60
60-45
45-60
60-15
15-60
60-0
0-60
60-30
30-15
15-0
045
45-15
1545
45-0
0-60
60-0
0-30
30-60
60-30
30-0
0-60
60-0
0-30
30-60
60-30
30-0
TIME IN MODE
(sec)
12
16
8
11
13
12
17
12
14
30
26
21
32
23
9
8
22
16
18
19
25
28
15
25
18
10
38
35
18
21
14
13
AVERAGE SPEED
(mph)
18.05
16.66
9.04
23.07
37.65
38.05
45.80
53.01
52.54
40.40
43.42
33.83
38.24
46.86
23.18
7.81
28.85
31.33
30.55
24.72
38.28
33.88
17.73
45.14
47.23
15.99
38.01
33.87
17.73
45.27
46.63
16.40
AVERAGE ACCEL-
ERATION RATE
(mph/sec)
2.50
-1.88
1.88
1.36
1.15
-1.25
1.76
-1.25
1.07
1.50
1.73
-2.86
1.88
-1.30
-1.67
-1.88
2.05
-1.88
1.67
-2.37
2.40
-2.14
2.00
1.20
-1.67
-3.00
1.58
-1.71
1.67
1.43
2.14
2.31
DISTANCE
(miles)
0.0602
0.0741
0.0201
0.0705
0.1360
0.1268
0.2163
0.1716
0.2043
0.3367
0.3136
0.1973
0.3313
0.2994
0.0579
0.0173
0.1759
0.1392
0.1528
0.1304
0.2654
0.2634
0.0737
0.3134
0.2362
0.0444
0.4009
0.3293
0.0886
0.2599
0.1813
0.0592
-------
Table 41
MEAN EMISSIONS DATA FOR 32 ACCELERATION/DECELERATION MODES
(FIVE CITIES EXCLUDING DENVER - 851 TESTS)
(PPM)
CONCENTRATION
MODE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
HC
757.
353.
536.
555.
650.
399.
981.
542.
847.
610.
952.
600.
925.
616.
375.
344.
806.
449.
741.
459.
1073.
629.
709.
795.
584.
382.
830.
612.
643.
850.
546.
403.
CO
3773.
1390.
2997.
2917.
2973.
1500.
7148.
2140.
4126.
1953.
6098.
1912.
5335.
2082.
1522.
1902.
4649.
1775.
4334.
1798.
7789.
1995.
3691.
4510.
2041.
1646.
4625.
2062.
3504.
5851.
2006.
1768.
CO2
19118.
6111.
10928.
17471.
24090.
8440.
33971.
12572.
32846.
9234.
31886.
8590.
29772.
10019.
6338.
5241.
23429.
7317.
23069.
6915.
31467.
8257.
17115.
29170.
9910.
6142.
27380.
8208.
15556.
30179.
9896.
5871.
NOX*
212.
41.
62.
183.
328.
95.
426.
171.
493.
106.
445.
103.
436.
115.
40.
21.
298.
67.
292.
59.
402.
86.
182.
431.
118.
39.
409.
78.
155.
410.
129.
37.
(GRAMS PER MILE)
MASS
HC
11.8
5.6
16.2
6.6
4.7
2.8
6.0
2.9
4.5
4.2
6.1
4.9
6.9
3.6
4.3
11.7
7.8
3.9
6.7
5.0
7.9
5.1
11.1
4.9
3.4
6.4
6.1
5.0
10.0
5.3
3.2
6.6
CO
121.28
46.75
190.38
72.53
45.18
22.18
90.48
23.69
45.27
27.44
81.23
32.04
82.74
25.29
36.85
138.53
93.16
32.10
81.70
41.18
118.07
33.48
120.15
57.59
24.52
58.05
70.34
34.65
113.70
75.95
24.42
60.83
C02
955.38
310.77
1060.35
674.88
573.36
192.40
669.26
216.33
564.19
198.57
662.80
219.98
721.16
187.35
231.41
562.57
730.27
200.71
677.77
239.54
742.47
211.49
861.83
582.50
183.55
324.64
649.38
209.93
779.45
610.55
185.34
300.49
NOX*
11.3
2.3
6.4
7.5
8.3
2.4
8.9
3.1
9.0
2.5
9.8
2.9
11.2
2.3
1.6
2.4
9.9
2.0
9.1
2.2
10.1
2.4
9.8
9.1
2.4
2.2
10.3
2.2
8.3
8.8
2.6
2.1
UNCORRECTED FOR HUMIDITY
52
-------
Table 42
MEAN EMISSIONS DATA FOR 32 ACCELERATION/DECELERATION MODES
(DENVER-169 TESTS)
MODE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
(PPM)
CONCENTRATION
HC
1003
446
660
740
919
508
1498
786
1274
835
1572
833
1535
895
511
451
1265
604
1126
630
1730
850
1045
1246
745
502
1346
850
894
1468
760
551
CO
10450
2456
5252
7279
10626
3041
22506
4984
16440
4093
23397
4085
19753
4857
2826
2982
16247
3524
14863
3491
25496
4555
10461
17567
5084
2813
17466
4514
8897
20384
4978
2866
C02
14172
4902
7915
12806
18105
6759
23961
10286
24223
7619
29522
7314
20438
8680
5082
3866
15999
5945
15766
5530
21238
7479
12261
19730
8719
4857
18799
7061
11108
20620
8493
4501
NOX*
112
37
33
106
212
95
230
188
341
124
199
123
198
139
44
17
133
69
128
59
156
102
94
203
146
42
191
99
93
188
150
37
(GRAMS PER MILE)
MASS
HC
12.7
6.0
16.6
7.3
5.6
3.0
7.5
3.5
5.6
4.7
8.3
5.6
9.5
4.4
5.0
13.0
10.1
4.4
8.4
5.8
10.4
5.7
13.5
6.3
3.6
7.1
8.1
5.7
11.5
7.6
3.7
7.6
CO
271.66
69.07
272.49
147.87
132.00
37.49
230.14
45.41
146.63
47.48
252.75
56.54
248.87
48.60
56.95
179.61
264.70
52.77
228.05
66.23
312.80
63.09
277.35
182.49
50.38
82.31
215.80
62.49
235.36
214.38
49.99
81.72
CO2
564.54
200.92
616.48
397.75
346.72
124.06
379.42
142.23
334.63
132.40
376.35
151.35
397.62
130.85
149.68
332.14
400.57
131.49
371.59
154.21
402.57
155.03
496.14
316.22
130.23
206.87
358.05
145.87
446.97
334.98
128.43
185.66
NOX*
4.6
1.5
2.5
3.4
4.2
1.8
3.8
2.8
4.9
2.3
3.5
2.7
4.0
2.2
1.3
1.3
3.5
1.6
3.1
1.7
3.1
2.2
3.9
3.4
2.3
1.8
3.8
2.2
3.9
3.2
2.4
1.5
"UNCORRECTED FOR HUMIDITY
53
-------
Table 43
MEAN EMISSIONS DATA FOR FIVE STEADY STATE MODES1
CITY
LOS ANGELES
DENVER
CHICAGO
HOUSTON
ST. LOUIS
WASHINGTON
ALL EXCEPT
DENVER
POLLUTANT
HC
CO
co.
NO;
HC
CO
CO,
NO*
HC
CO
CO,
NOx
HC
CO
co.
NO*
HC
CO
CO,
N0x
HC
CO
CO,
NO*
HC
CO
CO,
N02X
IDLE
1.45
17.02
72.32
0.08
1.47
17.20
60.54
0.12
1.30
15.49
63.85
0.10
1.50
18.74
76.13
0.14
1.33
16.43
64.28
0.08
1.12
13.25
65.04
0.13
1.34
16.19
68.35
0.11
ISmph
5.28
69.11
409.53
0.51
6.06
86.49
345.92
0.83
5.47
66.57
358.70
0.87
5.68
77.29
391.06
0.83
4.90
67.68
334.24
0.53
4.23
56.10
380.44
1.04
5.11
67.36
374.23
0.75
30mph
3.09
29.50
333.12
1.69
3.74
52.13
297.92
1.53
3.48
29.77
318.61
2.36
3.02
35.11
318.01
1.49
2.76
29.54
301.64
1.57
2.59
26.19
343.54
2.88
2.99
30.02
323.03
2.00
45 mph
2.91
24.60
357.90
3.78
3.94
56.43
330.09
3.07
3.59
30.53
357.65
4.37
2.83
32.38
356.13
3.47
2.58
26.13
315.26
3.38
2.57
25.40
390.83
6.04
2.90
27.79
355.55
4.21
60 mph
2.60
25.51
372.47
5.51
3.87
71.13
368.91
4.52
3.62
32.63
417.11
6.26
2.47
31.71
404.79
5.70
2.67
26.97
362.26
5.43
2.88
25.79
452.50
8.90
2.85
28.50
401.60
6.35
1. IDLE RESULTS IN GRAMS PER MINUTE.
NOX NOT CORRECTED FOR HUMIDITY.
STEADY STATE RESULTS IN GRAMS PER MILE.
-------
Table 44
FUEL EVAPORATIVE EMISSIONS USING THE ENCLOSURE TECHNIQUE
MODEL
YEAR
N
DIURNAL MEAN*
HC-GMS
- LOS ANGELES -
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1957-1971
2
4
4
2
4
5
7
9
11
13
10
11
13
11
20
126
- DENVER -
1971
20
33.04
14.10
40.49
31.33
11.94
15.42
27.37
29.43
43.36
22.83
25.56
15.18
26.34
18.20
15.22
23.67
47.20
S.D.
12.31
4.46
48.79
10.63
1.80
6.37
27.75
13.02
31.56
10.51
20.40
3.51
11.62
12.19
10.74
18.79
36.08
HOT SOAK
MEAN
HC-GMS
22.04
11.21
11.03
13.32
13.86
9.86
18.33
15.76
15.04
15.94
15.22
12.54
15.10
10.49
11.15
13.75
34.77
S.D.
8.14
9.61
5.66
2.76
4.86
2.92
15.19
9.47
6.48
8.43
8.04
6.55
5.66
5.46
6.86
7.68
24.25
COMBINED MEAN*
HC-GMS
55.08
25.32
51.52
44.64
25.80
25.28
45.71
45.14
58.39
30.77
41.78
27.72
41.44
28.68
26.37
37.41
81.97 '
S.D.
20.45
6.56
52.99
13.40
5.73
6.35
31.42
15.50
30.20
17.19
23.93
7.92
11.61
16.47
15.90
21.69
42.49
ARITHMETIC MEAN
-------
1968
KANSAS CITY
(498 CARS)
HOUSTON
(469 CARS)
1969
PERCENT BELOW STANDARDS
0 10 20 30 40 50 60 70 80 90 100
BOTH HC & CO
HC
CO
BOTH HC & CO
HC
CO
'//////////A
0 10 20 30 40 50 60 70 80 90 100
PERCENT BELOW STANDARDS
0 10 20 30 40 50 60 70 80 90 100
KANSAS CITY
(494 CARS)
HOUSTON
(488 CARS)
BOTH HC & CO
HC
CO
BOTH HC & CO
HC
CO
1
////t^XXXM
1
Y///////////A
i i i.i i i i i i
0 10 20 30 40 50 60 70 80 90 100
Figure 1 CONFORMANCE OF 1968 AND 1969 MODEL TEST FLEETS WITH EMISSION
STANDARDS (COLD START SEVEN MODE, SEVEN CYCLE FEDERAL TEST
PROCEDURE)
57
-------
KANSAS CITY
(395 CARS)
HOUSTON
(387 CARS)
LOS ANGELES
(352 CARS)
DETROIT
(352 CARS)
DENVER
(352 CARS)
WASHINGTON, D.C.
(343 CARS)
PERCENT BELOW STANDARDS
0 1.0 20 30 40 50 60 70 80 90 100
BOTH HC & CO
HC
CO
BOTH HC & CO
HC
CO
BOTH HC & CO
HC
CO
BOTH HC & CO
HC
CO
BOTH HC&CO
HC
CO
BOTH HC&CO
HC
CO
Y///////A
i
yJ^/////A
1
y//zm\
(.3%)
W//////////A
0 10 20 30 40 50 60 70 80 90 100
Figure 2 CONFORMANCE OF 1970 MODEL TEST FLEET WITH EMISSION STANDARDS
(COLD START SEVEN MODE, SEVEN CYCLE FEDERAL TEST PROCEDURE)
58
-------
30
20
i
a?
10
1.20
3.20 5.20
HC gm/mi
30
20
-j
i-
ae
10
7.20
1.25 2.50
LOG (HC)
3.75
30 i-
20
i
10
30
20
10
37.5 75.0 112.5
CO gm/mi
2.50 5.00 7.50
LOG (CO)
Figure 3 HISTOGRAMS OF HC AND CO EMISSIONS 1970 MODEL VEHICLES
EXCLUDING DENVER
59
-------
30 |-
20
O
10
30 |-
20
10
"
1.20 3.20 5.20 7.20
HC gm/mi
1.25 2.50 3.75
LOG (HC)
30 i-
20
P
3*
10
601-
40
20
ttb
37.5 75.0 112.5
CO gm/mi
2.50 5.00 7.50
LOG (CO)
Figure 4 HISTOGRAMS OF HC AND CO EMISSIONS 1970 MODEL VEHICLES
DENVER ONLY
60
-------
HOUSTON
(65 CARS)
BOTH HC & CO
HC
CO
BOTH HC & CO
LOS ANGELES
(112 CARS)
HC
CO
DETROIT
(48 CARS)
BOTH HC & CO
HC
CO
DENVER
(144 CARS)
BO IH HC & CO
HC
CO
PERCENT BELOW STANDARDS
0 10 20 30 40 50 60 70 80 90 100
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
LOW Ml
STABIL
i i i i i i i i
... '.I
1
1
1
$$$M$$M?8$M$£8!:88888Mg&A
m^%£m?ffifi@mt?pmmmfa
1
VWfiXfiS&iSgWi
83£88888$$a
|
I
1
i i i i i i i i i
0 10 20 30 40 50 60 70 80 90 100
Figure 5 CONFORMANCE OF 1971 MODEL TEST FLEET WITH EMISSION STANDARDS
(COLD START SEVEN MODE, SEVEN CYCLE TEST PROCEDURE)
61
-------
30
20
10
HC gm/mi
p
60
40
20
-.375
CO gm/mi
as
30
20
10
-15.0
gm/mi
.175 0.0 +.175
SLOPE/1000 mi
+.375
-7.0 0.0 7.0
SLOPE/1000 mi
15.0
-.375 -.175 0.0 +.175 +.375
SLOPE/1000 mi
Figure 6 MILEAGE EFFECTS (HISTOGRAM OF SLOPES) 1968 MODEL VEHICLE
62
-------
30
20
10
HC gm/mi
I
60
40
20
-.375
CO gm/mi
30
20
10
-15.0
gm/mi
.175 0.0 +.175 +.375
SLOPE/1000 mi
-7.0 0.0 7.0
SLOPE/1000 mi
15.0
-.375 -.175 0.0 +.175 +.375
SLOPE/1000 mi
Figure 7 MILEAGE EFFECTS (HISTOGRAM OF SLOPES) 1969 MODEL VEHICLES
63
-------
g
30
20
10
30
20
10
I
30
20
10
HC gm/mi
-.375 -.175 0.0 +.175 +.375
SLOPE/1000 mi
CO gm/mi
-15.0
NOX gm/mi
-.375
-7.0 0.0 7.0
SLOPE/1000 mi
M-I-H.
15.0
>Lrd"l
-.175 0.0 +.175 +.375
SLOPE/1000 mi
Figure 8
MILEAGE EFFECTS (HISTOGRAM OF SLOPES) 1971 MODEL VEHICLES
EXCLUDING DENVER
64
-------
30
20
10
HC gm/mi
r-n
o
30
20
10
-.375
CO gm/mi
.175 0.0 +.175 +.375
SLOPE/1000 mi
30 i-
20
10
15.0
gm/mi
-7.0 0.0 7.0 15.0
SLOPE/1000 mi
fttm
_n
.375 -.175 0.0 +.175 +.375
SLOPE/1000 mi
Figure 9 MILEAGE EFFECTS (HISTOGRAM OF SLOPES) 1971 MODEL VEHICLES
DENVER
65
-------
0 20 40 60 80 100
VEHICLES - CUMULATIVE %
200
E
I
I
o
o
1
cc.
100
0 20 40 60 80 100
VEHICLES - CUMULATIVE %
0 20 40 60 80 100
VEHICLES - CUMULATIVE %
Figure 10 1975 CVS EMISSION LEVELS VS. CUMULATIVE PERCENTAGE OF TEST
VEHICLES FOR PRE-CONTROL, 68-69 AND 70-71 MODEL YEARS
(ALL CITIES EXCLUDING DENVER)
66
-------
Appendix I
EMISSION REDUCTION GOALCALCULATIONS
The table on page 2 which shows the percent of the emission reduc-
tion goal that has been achieved is based on the Study of Light-Duty Vehicles
in the Six Cities1957-1971 Model-Year Survey. As explained in the text, a
significant difference exists between Denver and the other five cities (Los
Angeles, St. Louis, Houston, Chicago and Washington, D.C.). Vehicles tested
in Denver, as well as 1966 and 1967 model year vehicles tested in Los Angeles
which had been subject to California standards, were excluded from the calcu-
lations on which the table is based.
The table is based on 818 vehicle exhaust tests. The percent of
goal achieved was determined as follows:
Pre-ContHr Minus ActualHr
Percent of Goal Achieved^ = Pre.Con^c Minus St
-------
Appendix II
THE LOG-NORMAL DISTRIBUTION AS A STATISTICAL
MODEL FOR EXHAUST EMISSIONS
Histograms compiled from automobile exhaust emission surveillance
data suggest that these emissions conform to a log-normal distribution. In
the following discussion, this distribution and its properties are examined
and assumptions which might give rise to the distribution in an emissions
context are explored.
1. PROPERTIES OF THE LOG-NORMAL DISTRIBUTION
A random variable X is said to have a log-normal distribution if
log X has a normal distribution. Let Y = log X. Then Y has the probability
density function
f(y)=
where m and a are, respectively, the mean and standard deviation of the
random variable Y. Note that, with regard to the initial random variable X,
m is the mean value of log X and a is the standard deviation of log X.
Suppose, now, that one performs the inverse logarithmic transforma-
tion on m. It can be readily shown that antilog m is the geometric mean of
the random variable X. Note, also, that the distribution of Y is symmetric.
Therefore, m is not only the mean value of Y but also its median- -that is,
50% of the values of Y are larger than m and 50% are smaller than m. It
follows, therefore, that the antilog of m is the median for the distribution
of X as well as being its geometric mean.
Interpretation of .the standard deviation a and its anti logarithm
is somewhat more subtle. As long as one deals with Y, the log-transformed
variable, interpretation of its distribution follows all the conventions
normally employed in drawing inferences on the basis of the normal distribu-
tion. For example, if one adds the standard deviation to the mean, the result
M + a is a value of Y below which approximately 84% of all the values of Y
fall--that is, m + a is approximately the 84th percentile of the distribution.
It follows, therefore, that antilog (m + a ) is approximately the 84th per-
centile for the distribution of the original variable X. Similarly, if one
adds two standard deviations to obtain m + 20 , then IT + 2(7 constitutes
approximately the 95th percentile for the random variable Y and antilog
(m + 2a) constitutes approximately the 95th percentile for the random vari-
able X.
68
-------
Consider, now, the inverse logarithm of a . This quantity, anti-
log a , retains the interpretational properties of a provided it is used
in a multiplicative rather than an additive sense. If antilog m (the median
of X) is multiplied by antilog a , the result is clearly antilog Cm + a);
therefore, the multiplication yields directly the 84th percentile of the ran-
dom variable X. Similarly, if one multiplies the median of X by a2 , the
result is clearly antilog (m +2a) , or the 95th percentile of X. It is for
this reason that antilog a is sometimes referred to as the "ratio standard
deviation" in dealing with the log-normal distribution. In this report it is
called the geometric standard deviation to connote its relation to the geometric
mean.
2. A HEURISTIC JUSTIFICATION OF THE LOG-NORMAL DISTRIBUTION OF
EMISSIONS
If emission measurements tend to follow a log-normal distribution,
it is of interest to ask why this relationship holds.
A statistical process may tend toward producing a log-normal dis-
tribution if the following conditions are met:
(a) Several sources of error or variability combine in
generating the random variable under study.
(ti) These sources of error combine multiplicatively rather
than additively.
Within reasonable limits, the form of the statistical distribution of the
source variables is irrelevant.
To justify the above hypothesis, at least heuristically, consider the
central limit theorem of mathematical statistics. Though this theorem can be
expressed in various ways, the following statement suffices for present pur-
poses. If an arbitrary population distribution has mean m and finite vari-
ance a2 , then the distribution of the sample mean approaches the normal
distribution with mean m and variance ((72/n as the sample size n increases.
Sometimes it is said that the distribution of the sample means tends asymtoti-
cally to a normal distribution. How large n has to be in order to approach
closely to a normal distribution depends on the form of the distribution of
the variables being combined. As has been shown in experience with statistical
quality control concepts, normality can be closely approached if n is as small
as 3 or 4, even if the distributions of the original variables depart markedly
from a normal distribution.
Now consider the case of n sources of errorthat is, consider n
random variables X , X ,..., X . Perform the logarithmic transformation
69
-------
log X = log X. + log X_ + ...+ log X
or Y = Y + Y + ...+ Y
where Y. = log X. .
Then, if the central limit theorem applies, Y will tend asymptotically toward
a normal distribution, and this fact implies that in antilog space X will tend
toward a log-normal distribution.
There is reason to believe that the sources of variability in auto-
motive exhaust emissions combine multiplicatively. Consider, first of all,
errors of measurement in determining the concentration of a particular pollu-
tant in the exhaust stream. Let the error involved in measuring the concen-
tration C be measured by its standard deviation a . There is reason to
believe that a is proportional to C--that is, that the relative standard
deviation or coefficient of variation O/C is constant for a. particular
effluent and a particular measurement situation.
Now, consider a particular engine and let its emissions be measured
repeatedly by the process under consideration. The actual concentration C.
will vary from time to time, due to differences in ambient environmental con-
ditions, condition of linkages and engine adjustments, type of fuel, operator-
induced variations and the like. It will follow, therefore, that a will also
vary proportionally to C and that the net result will be analogous to the
multiplication of two sources of error.
Let us further consider the relationship which obtains between
repeated tests of the same automobile and variation among many automobiles in
a fleet or in a production output. It might, perhaps, be argued that if a
particular automobile has a high mean level of emissions, that same automobile
may tend to have highly-variable emissions. Conversely, an automobile having
a relatively low mean level of emissions may tend to have relatively constant
emissions. A moment's reflection will show that these statements must, to
some extent at least, be true. Since emissions are bounded on the low side
of the scale (they can not be less than zero) . it is impossible for individual
emission measurements to range upward beyond a certain level and yet maintain
a low mean, unless the high measurements are offset by a large number of
measurements close to zero. This bounding process will tend to limit the
magnitude of the standard deviation and, at the same time, may tend to produce
a skewed distribution which, itself, may resemble a log-normal distribution.
i
Now consider the combination of vehicle-to-vehicle variability, time-
to-time variability within a given vehicle, and test-to-test variability at a
given time. If test-to-test variability is proportional to the mean level of
the particular automobile at a given time, and time-to-time variability is
70
-------
proportional to the mean level for the particular vehicle, then the three
sources of variability would combine in a multiplicatively way. The result,
by virtue of the central limit theorem, would be a log-normal distribution.
71
-------
Appendix
DISCRIMINANT FUNCTION ANALYSIS OF AUTOMOBILE
EXHAUST EMISSIONS
To provide a greater insight with regard to apparent differences
between emissions for different cities or different makes of vehicles, multiple
discriminant function analysis was applied to the emission data. The procedure
is illustrated in Figure III-l^ which presents a hypothetical plot of emission-
measurements from vehicles in two cities designated simply as City A and City
B. In the left-hand part of this figure, the emissions are plotted in three-
dimensional space. Both cities exhibit a considerable amount of scatter in
the plotted points and, although the two sets of points tend to cluster about
a different center of gravity, there is a considerable amount of interpenetra-
tion of the two clouds of points. Consequently, it is impossible to construct
a plane in the space in such a way that all City A points fall on one side and
all City B points on the other.
/
CO
/ /
\ o ° o
\° o °
x \ o o °
X * 0\*
^ %yv*^^
X X X \
x X \_
X CITY A
O CITY B
Figure 111-1 DISCRIMINANT FUNCTION ANALYSIS OF AUTOMOBILE EXHAUST EMISSIONS
Now consider the right-hand side of Figure III-l. In this two-
dimensional plot, two linear combinations of the three emittants are computed:
72
-------
Fl =
* a2(CO) + a3(NOx)
F2 =
b2(CO)
The quantities FI and ₯2 are called discriminant functions and the quantities
al> a2> a3 anc^ bj, b2, £3 are weighting factors selected in such a way as to
produce minimum scatter within groups and maximum separation between groups.
Though it is still not possible to separate the groups completely, the two
sets of points show much less overlap than was previously the case. If such
a result can be achieved when comparing emissions from two or more actual
cities, then it is clear that a real difference between them is involved.
In the computation of discriminant functions, one of the functions
may be much more discriminating than the other. In fact, it may be that sub-
stantially all of the discrimination is realized along either the FI or $2
If such is the case, the weaker of the two discriminants can be ignored, and
the problem reduces to a one-dimensional problem in which classification of
points can be achieved by observing whether a particular point lies to the
left or to the right of some critical value. This value is selected (see
Figure III-2) in such a way that the two frequency distributions overlap to
the least possible degree. The principle involved is that of maximum likeli-
hood, each point being assigned to the group to which it is "most likely" to
belong.
CITY A
CITY B
Figure 111-2 MAXIMUM LIKELIHOOD DISCRIMINATION
Several points must be made here. First, it is very rare that two
sets of data will be found which will show no overlap or intermingling of points.
Second, the impression received from a discriminant function plot may be quite
different from that obtained if only the mean values of the emittants, or of
73
-------
discriminant functions derived from them, are considered. As has been previously
pointed out, a very minimal difference between the two means yUA and >UB may-be
declared statistically significant, provided the number of observations in the
two groups is sufficiently large, yet this difference may be of little practical
importance. The implications of this fact can be further appreciated from the
discriminant function plots, wherein it is seen that even though the centroids
of the two clouds of points may be distinct, the actual amount of overlap of
the two distributions precludes strong generalizations concerning a preference
for one or the other of the two distributions.
The rationale described above was applied in assessing differences
between cities and between vehicle makes. Figures III-3 through III-6 illus-
trate the analytic process as applied to low mileage 1971 model Chevrolets
operating in the four cities Houston, Los Angeles, Detroit and Denver. The
first three figures show plots of HC vs CO, HC vs NO , and CO vs NO . In these
X "
plots, each point represents a particular vehicle. The last figure in this
group is a plot of the two strongest discriminant functions. Units in this
figure are arbitrary but, being weighted combinations of the three effluents,
are proportional to grams per mile. Whether one considers the. raw data or
discriminant-function plots, it is difficult to discriminant one city from
another, with the exception of Denver and the possible exception of Los Angeles.
Both on the basis of discriminant function plots and the magnitude of differences
among the means for the several cities, it was concluded that there is little
justification for treating separately any of the cities except Denver.
As is evident in Figure III-6, Denver data is separated from the data
from the other cities primarily in the direction of the Factor 1 axis. This
fact makes it possible to simplify the comparison by considering only this
factor and plotting Denver and non-Denver data as histograms, as shown in Fig-
ure III-7. Though there is some overlap of the two histograms, the distinction
between the two categories is clear. The plot is particularly meaningful,
because it reflects the distinctiveness of Denver not in terms of a single
effluent but in terms of all effluents combined.
74
-------
CM-
CD
-g
O
a.
in
in .
OJ
0
Qk
E
o f^xs
i e§
s
a
o
o
i
0,80
T
I
1. GO 2.-*0 3.20
HC gm/mi
O -HOUSTON
Q - LOS ANGELES
X -DETROIT
A -DENVER
I !
4.00 4. BO
Figure 111-3 TWO-WAY PLOT OF EMISSIONS FOR FOUR CITIES 1971 PHASE 1 CHEVROLETS
75
-------
o
to
o
oa
O
o
V .
o-
E
t
-it-
?<.
o
o
tf
« ,
OJ
o
to
0
n
m
BO
O
©
e
z
o
s
I
D.BO
1.60 2.40 3.20
HC gm/mi
O -HOUSTON
Q -LOS ANGELES
X -DETROIT
A. -DENVER
4.00 4.80
Figure 111-4 TWO-WAY PLOT OF EMISSIONS FOR FOUR CITIES 1971 PHASE 1 CHEVROLETS
76
-------
CD B
0 S
4 I
O> ,i
X """ J
on
s
co
0
CD
§
m
O - HOUSTON
Q - LOS ANGELES
X - DETROIT
A -DENVER
I
25
i
50
75
i
100
CO gm/mi
I
125
1SJ
Figure 111-5 TWO-WAY PLOT OF EMISSIONS FOR FOUR CITIES 1971 PHASE 1 CHEVROLETS
77
-------
rw _
o
o.
OJ
G3
CD
EJ
O
O
CS
E * ~
ffi
1 1
0.5 a. a 0,5"
0
a
z
A
1 1
i..Q L.S
- HOUSTON
- LOS ANGELES
- DETROIT
- DENVER
1 !
2.0 2.
FflC2
Figure 111-6 DISCRIMINANT PLANE PLOT OF EMISSIONS FOR FOUR CITIES
1971 PHASE 1 CHEVROLETS
78
-------
28 r-
ALL CITIES
EXCLUDING DENVER
4 6 8 10 12
MAGNITUDE FACTOR # 1
14 16
Figure 111-7 DISTRIBUTION OF PRINCIPAL FACTOR (1971 PHASE 1 QHEVROLETS)
DENVER VS REST
79
-------
BIBLIOGRAPHIC DATA ' ttcport No. 2.
SHEET APTD-1544
4. 1 ulc anJ Subtitle
Automobile Exhaust Emission Surveillance
A Summary
7. Authorfs^
9. Performing Organization Name and Address
CALSPAN Corporation
12. Sponsoring Organization Name and Address
ENVIRONMENTAL PROTECTION AGENCY
Office of Mobile Source Air Pollution Control
Certification and Surveillance Division
Ann Arbor, Michigan 48105
5- Report Date
May 1973
6.
8. Performing Organization Kept.
No.
10. Project/Task/ffork Unit No.
1 1. Contract/Grant No.
68-01-0435
13. Type of Report & Period
Covered
14.
15. Supplementary Notes
u. Abstracts Tne purpose of the report is to summarize information on emissions from light-
duty vehicles. The report contains the findings and results of three exhaust emission
surveillance programs conducted by the EPA: (T) the Great Plains (Two-City) Surveillance
Program - 1968-1969 Model Year Survey; (2) the National Surveillance Pr9gram - 1970 (Six
City) and 1971 (Four City) Model Year Surveys; and (3) a study of emissions from Light-
Duty Vehicles in Six Cities - 1957-1971 Model Year Survey. The first two programs employ
ed the Federal Seven-Mode Test Procedure whereas the third program utilized the 1972 and
1975 Constant Volume Sampling (CVS) Federal Test Procedure. Hydrocarbon and carbon monox
ide emissions for the vehicles tested were assessed by comparing their mean emission
levels with applicable Federal standards. In an effort to assess the extent to which lo-
cal climate, terrain, driving practices and other geographically differentiated factors
affect emissions, vehicles were sampled in cities selected to span the range of such fac
tors. The effect of mileage accumulation on vehicle functioning, as reflected by measur-
able changes in emission levels, was explored by making two or more emission measurement:
on the same vehicle at different points in its mileage-accumulation history.
17. Key Words and'Documenc Analysis. 17a. Descriptors
Air pollution
Exhaust emissions Urban areas
Vehicles
Surveillance
Tests
Sampling
Monitoring
Hydrocarbons
Carbon monoxide
Nitrogen oxides
17b. Identificrs/Open-Ended Terms
Light-duty vehicles
Constant Volume Sampling Procedure (CVS)
Federal Test Procedures
17e. COSATI Fie Id /Group
13B
18. Availability
FORM N TlS-JS l«
Srjtemcnt
Unlimited
W.
20.
ev- 3'72' THIS FORM MAY BE REPRODUCED
Kcrur:
V\'(
R>gcr
V <- ..>*.-
' CLiss
I.AS^II
'Thi.s
H-'O
('111 is
I!-!-)
21- \
22. 1
o. t-:' t'.i-x-s
87
r i c L-
-------
INSTRUCTIONS FOR COMPLETING FORM NTIS-35 (10-70) (Bibliographic Data Sheet based on COSATI
Guidelines to Format Standards for Scientific and Technical Reports Prepared by or for the Federal Government,
PB-180 600).
1. Report Number. Each individually bound report shall carry a unique alphanumeric designation selected by the performing
organization or provided by the sponsoring organization. Use uppercase letters and Arabic numerals only. Examples
FASEB-NS-87 and FAA-RD-68-09.
i. Leave blank.
3. Recipient's Accession Number. . Reserved for use by each report recipient.
4. Title and Subtitle. Title should indicate clearly and briefly the subject coverage of the report, and be displayed promi-
nently. Set subtitle, ii used, in smaller type or otherwise subordinate it to main title. When a report is prepared in more
than one volume, repeat the primary title, add volume number and include subtitle for the specific volume.
5. Report Dote. Eaca report shall carry a date indicating at least month and year. Indicate the basis on which it was selected
(e.g., date of issue, date of approval, date of preparation.
6* Performing Organization Cade. Leave blank.
7. Authorfa). Give name(s) in conventional order ic.g., John R. Doc, or J.Robert Doe). List author's affiliation if it differs
from the performing organization.
8. Performing Organization Report Number. Insert if performing organization wishes to assign this number.
9. Performing Organization Name and Address. Give name, street, city, state, and zip code. List no more than two levels of
an organizational hierarchy. Display the name of the organization exactly as it should appear in Government indexes such
as USGRDR-I.
10. Project/Tosk/toork Unit Number. Use the project, task and work unit numbers under which the report was prepared.
11. Contract/Grant Number. Insert contract or grant numbei under which report was prepared.
12, Sponsoring Agency Nome and Address. Include zip code.
13. Type of Report and Period Covered. Indicate interim, final, etc., and, if applicable, dates covered.
14. Sponsoring Agency Cade. Leave blank.
15. Supplementary Notes. Enter information not included elsewhere but useful, such as: Prepared in cooperation with . . .
Translation of ... Presented at conference of ... To be published in ... Supersedes . . . Supplements '. .
16. Abstract. Include a brief (200 words or less) factual summary of the most significant information contained in the report.
If the reporr contains a significant bibliography or literature survey, mention it here.
17. Key Words and Document Analysis, (a). Descriptor*. Select from the Thesaurus of Engineering and Scientific Terms the
proper authorized terras that identify the major concept of the research and zre sufficiently specific and precise to be used
as index entries for cataloging.
(b). Identifier* and Open-Ended Terms. Use identifiers for project names, code names, equipment designators, etc.. Use
open-ended terms written in descriptor form for those subjects for which no descriptor exists.
(c). COSATI Field/Group. Field and Group assignments are to be taken from the 1965 COSATI Subject Category List.
Since the majority of documents are multidisciplinary in nature, the primary Field/Group assignment(s) wilt be the specific
discipline, area of human endeavor, or type of physical object. The application(s) will be cross-referenced with secondary
Field/Group assignments that will follow the primary posting(s).
IB. Distribution Statement. Denote releasability to the public or limitation for reasons other than security for example "Re-
lease unlimited". Cite any availability to the public, mth address and price.
19 & 20. Security Classification. Do not submit classified reports to the National Technical
21. Number of Page*. Insert the total number of pages, including this one and unnumbered pages, but excluding distribution
list, if any.
22. Price. Insert the price sec by the National Technical Information Service or the Government Printing Office, if known.
FORM NTIS-JB IBEV. S?72) USCOMM-I5C 14952-P7Z
------- |