ANALYSIS OF AIR RESOURCES BOARD DATA
FOR CALIFORNIA EMISSION FACTORS
TSC-PD-A231-4
August 1980
Technology Service Corporation
-------
Technology Service Corporation
2950 31st STREET SANTA MONICA, CALIFORNIA 90405 (213)450-9755
ANALYSIS OF AIR RESOURCES BOARD DATA
FOR CALIFORNIA EMISSION FACTORS
TSC-PD-A231-4
August 1980
John Gins
Nancy Chang
Michael Thiele
George Rakuljic
Draft Final Report
Contract No. 68-03-2836, Task No, 2
Submitted to:
U.S. ENVIRONMENTAL PROTECTION AGENCY
2565 Plymouth Road
Ann Arbor, Michigan 48105
Task Officer: Mr. Jay Wallace
-------
TABLE OF CONTENTS
Section Page
1. INTRODUCTION 1
1.1 BACKGROUND 1
1.2 PURPOSE 1
1.3 REPORT ORGANIZATION 2
1.4 CONCLUSIONS 2
2. MATCHING CALIFORNIA MODEL-YEAR GROUPS WITH FEDERAL
MODEL-YEAR GROUPS 4
2.1 DEFINITION OF MODEL-YEAR GROUPINGS 4
2.2 MATCHING MODEL-YEAR GROUPS 5
3. HOMOGENEITY OF THE SIX ARB DATA SETS 21
3.1 VEHICLE SELECTION AND CVS-75 EMISSION TEST 21
3.2 COMBINED DATA SET 22
4. DATA VERIFICATION FOR CARB DATA SET 24
4.1 CARB REPORT NO. 1 24
4.2 CARB REPORT NO. 4 24
4.3 CARB REPORT NO. 7 26
4.4 CARB REPORT NO. 8 27
4.5 CARB REPORT NO. 11 27
4.6 CARB REPORT NO. 12 28
5. ANALYSIS . . . 31
5.1 DATA BASE 31
5.2 DESCRIPTIVE STATISTICS OF THREE EMISSIONS UNDER
EACH MODEL-YEAR GROUP 31
5.3 REGRESSION ANALYSIS BETWEEN EMISSIONS AND ODOMETER
READINGS 35
-------
LIST OF TABLES
Table Page
2.1 Definition of Model-Year Group for Light-Duty Gasoline
Vehicles (LDGV) for HC 6
2.2 Definition of Model-Year Group for Light-Duty Gasoline
Vehicles (LDGV) for CO ........ .... 7
2.3 Definition of Model-Year Group for Light-Duty Gasoline
Vehicles (LDGV) for NOV . . 8
A
2.4 Definition of Model-Year Group for Light-Duty Gasoline-
Powered Trucks (LDGT1) for HC ......... 9
2.5 Definition of Model-Year Group for Light-Duty Gasoline-
Powered Trucks (LDGT1) for CO 10
2,6 Definition of Model-Year Group for Light-Duty Gasoline-
Powered Trucks (LDGT1) for NOV .......... 11
A
2.7 Definition of Model-Year Group for Medium-Duty Gasoline-
Powered Trucks (LDGT2) for HC 12
2.8 Definition of Model-Year Group for Medium-Duty Gasoline-
Powered Trucks (LDGT2) for CO . 13
2.9 Definition of Model-Year Group for Medium-Duty Gasoline-
Powered Trucks (LDGT2) for NOY ...... 14
A
2.10 Definition of Model-Year Group for Heavy-Duty Gasoline-
Powered Vehicles (HDGT) 15
2.11 Definition of Model-Year Group for Light-Duty Diesel-
(Powered Vehicles (LDDV) 16
*
2.12 Definition of M6del-Year Group for Light-Duty Diesel-
Powered Trucks (LDDT1) 17
2.13 Definition of Model-Year Group for Light-Duty Diesel-
Powered Trucks (LDDT2) 18
2.14 Definition of Model-Year Group for Heavy-Duty Diesel-
Powered Vehicles . ...... 19
-------
LIST OF TABLES (Cont'd)
Table Page
2.15 Definition of Model-Year Group for Motorcycles 20
4.1 Summary of Changes Made During Data Verification 25
5.1 Breakdown of Model-Year and Emission Control Device for
Light-Duty Vehicles . 32
5.2 Breakdown of Model-Year and Emission Control Device for
Light-Duty Trucks ' 33
5.3 Breakdown by Model-Year for Medium-Duty Trucks . 34
5.4 Descriptive Statistics of Hydrocarbon Emissions for California
Light-Duty Gasoline Vehicles in Each Hydrocarbon Model -
Year Group 37
5.5 Descriptive Statistics of Mileage for California Light-Duty
Gasoline Vehicles in Each Hydrocarbon Model-Year Group ... 38
5.6 Descriptive Statistics of Carbon Monoxide Emissions for
California Light-Duty Gasoline Vehicles in Each Carbon
Monoxide Model-Year Group 39
5.7 Descriptive Statistics of Mileage for California Light-Duty
Gasoline Vehicles in Each Carbon Monoxide Model-Year
Group 40
5.8 Descriptive Statistics of Oxides of Nitrogen Emissions for
California Light-Duty Vehicles in Each NO Model-Year
Group 41
5.9 Descriptive Statistics of Mileage for California Light-Duty
Vehicles in Each NO Model-Year Group 42
/\
5.10 Descriptive Statistics for California Light-Duty Trucks for
Each Hydrocarbon Model-Year Group 43
5.11 Descriptive Statistics for California Light-Duty Trucks
in Each Carbon Monoxide Model-Year Group 44
5.12 Descriptive Statistics for California Light-Duty Trucks
in Each Nitrogen Oxides Model-Year Group 45
-------
LIST OF TABLES (Cont'd)
Table Page
5.13 Descriptive Statistics of Hydrocarbon for California
Medium-Duty Trucks in Each Model-Year Group 46
5.14 Descriptive Statistics of Carbon Monoxide for California
Medium-Duty Trucks in Each Model-Year Group . . 47
5.15 Descriptive Statistics of Oxides of Nitrogen for California
Medium-Duty Trucks in Each Model-Year Group 48
5.16 Descriptive Odometer Statistics for California Medium-
Duty Trucks in Each Model-Year Group 49
5.17 Results of Regression Analysis on Hydrocarbon Emissions and
Odometer Readings for California Light-Duty Gasoline
Vehicles 50
5.18 Results of Regression Analysis on Carbon Monoxide Emissions
and Odometer Readings for California Light-Duty Gasoline
Vehicles 51
5.19 Results of Regression Analysis on Nitrogen Oxides Emissions
for California Light-Duty Gasoline Vehicles 52
5.20 Results for Regression Analysis for California Light-Duty
Gasoline-Powered Trucks .... 53
5.21 Regression Analysis Between Hydrocarbon and Odometer
Readings for California Medium-Duty Gasoline-Powered
Trucks ' 54
5.22 Regression Analysis Between Carbon Monoxide and Odometer
Readings for California Medium-Duty Gasoline-Powered
Trucks . 55
5.23 Regression Analysis Between Nitrogen Oxides and Odometer .
Readings for California Medium-Duty Gasoline-Powered
Trucks 56
-------
1. INTRODUCTION
This report summarizes the results of Technology Service Corporation's
(TSC) review of emissions data gathered by the California Air Resources Board
(CARB) and an analysis of California model-year groups and Federal (other 49
states) model-year groups.
1.1 BACKGROUND
The Office of Mobile Source Air Pollution Control within the Environ-
mental Protection Agency has the responsibility for monitoring emissions from
in-use passenger vehicles, to assess the impact of these emissions on air
quality throughout the country. The mobile-source emission estimation model,
MOBILE2 (an updated version of MOBILE1), has, as a major component, emission
factors which are ultimately used to assess the impact of mobile emissions
on air qua-lity. MOBILE2 can be used to assist regions with the State Imple-
mentation Plans, the constructing of environmental impact statements, and the
setting-up of transportation control measures. Since California has different
vehicle emission standards from the other 49 states, California is analyzed
and reported on separately in MOBILE2.
1.2 PURPOSE
The objective of this task order is to analyze emissions data gathered
by the CARB and to determine deterioration factors relating the increase in
emissions to mileage. This analysis is based on model-year groups that have
been defined by TSC according to emission control technology (i.e., air pumps,
-------
EGR, oxidizing catalysts) and California emission standards. TSC has also
related, by emission control technology, the California groups with the
Federal model-year groups already defined for MOBILE2. EPA will use this
relationship to compare California and Federal emissions for vehicles with
similar technologies.
1.3 REPORT ORGANIZATION
Chapter 2 presents the model-year groups for the California vehicles
and pairs them with Federal model-year groups. Chapter 3 analyzes the
comparability among the data bases given in six CARB reports. Chapter 4
reports the changes that needed to be made to the data because of keypunching
or coding errors. Chapter 5 presents emissions and odometer readings (mileage)
for each model-year grouping that appears in the data and tabulates their
relationship. Descriptive statistics are presented, as well as a regression
analysis between emissions and mileage.
1.4 CONCLUSIONS
There is not a perfect match between California and Federal model -
year groups, owing to differences in emission standards and control technology.
The match is sufficient that comparisons can be made, however.
The six data sets derived from the CARB reports can be combined after
some duplicates are deleted. But, while some of the individual reports might
be representative of the California vehicle fleet, the combination would not
be representative without a stratification or weighting procedure. There are
no diesels in the data base, and so an analysis of diesel vehicles was not
made.
-------
The data analysis by model-year group indicates that, for some model-
year groups, the samples are just too small. The 1975 and 1976 model-year
light-duty gasoline-powered vehicles are the best represented. The regression
analysis indicates that use of a linear model to relate emissions to mileage--
to determine a deterioration factoris unreasonable for this data because
either the percent of variation explained by using the model is too small or
the coefficients are not significantly different from zero, or because the
sample was of a size insufficient for regression analysis to be performed.
-------
2. MATCHING CALIFORNIA MODEL-YEAR GROUPS WITH
FEDERAL MODEL-YEAR GROUPS
For the difference in deterioration factors between California vehicles
and Federal vehicles to be determined, it is necessary to know that the
vehicles under comparison have the same types of emission control devices and
are subject to a similar emission standard.
2.1 DEFINITION OF MODEL-YEAR GROUPINGS
The definitions of model-year groups were made by emissions and vehicle
classes. The emissions under consideration are total hydrocarbons (HC),
carbon monoxide (CO), and nitrogen oxides (NO ). The vehicle classes are
/\
light-duty vehicles (LDGV and LDDV); light-duty trucks (LDGT1 and LDDT1), with
gross vehicle weight less than or equal to 6000 pounds; light-duty trucks
(LDGT2 and LDDT2), with gross vehicle weight greater than 6000 pounds but
less than or equal to 8500 pounds (these are called medium-duty trucks in
California); heavy-duty trucks (HDGT and HDDT), with gross vehicle weight
greater than 8500 pounds; and motorcycles. Tables 2.1 through 2.10 give the
definition of model-year groups for gasoline-powered vehicles, and Tables 2.11
through 2.14 give the definition of model-year groups for diesel-fueled
vehicles. Table 2.15 gives the definition for motorcycles.
The definition of the Federal groupings was determined by the EPA and
reflects a knowledge of the change in emission standards and emission control
technology. The definition of the California groupings was derived by TSC
from our own knowledge of emission-standard changes and changes in control
technology. This knowledge was supplemented by a literature review and
-------
conversations with EPA personnel (particularly with the task officer) and
with engineers in the automotive industry. The interaction between NO and
/\
HC controls was also taken into account in deriving the model-year groups for
California vehicles.
2.2 MATCHING MODEL-YEAR GROUPS
The main purpose of this subtask was to find a Federal model-year group
that would correspond to each California model-year group. Tables 2.1 through
2.15 indicate which Federal group corresponds to each California model-year
group for each emission and vehicle class and gives reasons for the match.
The data matching of Federal and California model-year groups was made
by matching similar emission standards and emission control technologies
between the Federal and California model-year groups. Different California
model-year groups can be coupled with one Federal model-year group. The
requirement that each California model-year group match a Federal model-year
group necessitated this duplication. Two California groups matched to one
Federal group should not be lumped together.
-------
TABLE 2.1 DEFINITION OF MODEL-YEAR GROUP FOR LIGHT-DUTY GASOLINE VEHICLES
(LDGV) FOR HC
Model -Year
California
pre-66
66-67
68-69
70-71
72-73
74
75-76
77-79
80+
Group
Federal
pre-68
68-69
68-69
70-71
72-74
72-74
75-79
80
81
Major Emission Control Devices
None
Air Pump
Air Pump
Air Pump (California 71 has EGR)
Air Pump, EGR
Air Pump, EGR, with stringent NO
standard for California
Air Pump, EGR, oxidizing catalyst
Air Pump, EGR, oxidizing catalyst
3-Way catalyst
Emission Standard
None
275 ppm
50-100 CID 410
101-140 CID 350
Over 140 CID 275
2.2 gm/mile
3.2 gm/mile
(3.4 gm/mile)b
3.2 gm/mile
(3.4 gm/mile)
0.9 gm/mile
(1.5 gm/mile)
0.41 gm/mile
(1.5 gm/mile)
0.39 gm/mile
(0.41 gm/mile)
a
ppm
ppm
ppm
cr>
Test procedure prior to 72 is 7-mode; between 72-74, CVS-72; and for 75+, CVS-75.
Federal standard is value in parentheses.
-------
TABLE 2.2 DEFINITION OF MODEL-YEAR GROUP FOR LIGHT-DUTY GASOLINE VEHICLES
(LDGV) FOR CO
Model -Year
California
pre-66
66-67
68-69
70-71
72-74
75-76
77-79
80
Group
Federal
pre-68
68-69
68-69
70-71
72-74
75-79
80
81 +
Major Emission Control Devices
None
Air Pump
Air Pump
Air Pump (California 71 has EGR)
Air Pump, EGR
Air Pump, EGR, oxidizing catalyst
Air Pump, EGR, oxidizing catalyst
3-Way catalyst
Emission Standard3
None
1.5%
50-100 CID 2.3%
101-140 CID 2.0%
Over 140 CID 1.5%
39 gm/mile
39 gm/mile
9.0 gm/mile .
(15.0 gm/mile)D
9.0 gm/mile
(7.0 gm/mile)
9.0 gm/mile
(7.0 gm/mile)
Test procedure prior to 72 is 7-mode; between 72-74, CVS-72; and for 75+, CVS-75.
DFederal standard is value in parentheses.
-------
TABLE 2.3 DEFINITION OF MODEL-YEAR GROUP FOR LIGHT-DUTY GASOLINE VEHICLES
(LDGV) FOR NOV
Model -Year
California
pre-66
66-70
71
72-73
74
75-76
77-79
SOH-
Group
Federal
pre-68
68-72
73-74
73-74
73-74
75-76
77-79
80
81 +
Major Emission
None
Air Pump
Air Pump, EGR
Air Pump, EGR
Control Devices
Air Pump, EGR, with stringent NO
standard for California
Air Pump, EGR,
Air Pump, EGR,
Air Pump, EGR,
3-Way catalyst
oxidizing catalyst
oxidizing catalyst
oxidizing catalyst
Emission Standard3
None
None
4.0 gm/mile .
(3.0 gm/mile)D
3.0 gm/mile
2.0 gm/mile
(3.0 gm/mile)
3.1 gm/mile
2.0 gm/mile
1 . 5 gm/mi 1 e
(2.0 gm/mile)
1.0 gm/mile
Test procedure prior to 72 is 7-mode; between 72-74, CVS-72; and for 75+, CVS-75.
"Federal standard is value in parentheses.
00
-------
TABLE 2.4 DEFINITION OF MODEL-YEAR GROUP FOR
LIGHT-DUTY GASOLINE-POWERED TRUCKS
(LDGT1) FOR HC (GVW < 6000 Ibs)
Model -Year
California
pre-66
66-67
68-69
70-71
72-74
75
76-78
79-80
81 +
Group
Federal
pre-68
68-69
68-69
70-71
72-74
75-78
75-78
79-80
81
82
85+
Major Emission
Control Devices
None
Air Pump
Air Pump
Air Pump
(California 71
has EGR) '
Air Pump, EGR
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
3 -way Catalyst
Emission Standard
None
275 ppm
50-100 CID 410
101-140 CID 350
over 140 CID 275
2.2 gm/mile
3.2 gm/mile h
(3,4 gm/mile)
2.0 gm/mile
0.9 gm/mile
(2,0 gm/mile}
ppm
ppm
ppm
0,41 ^0,39 gm/mile
(2.0 gm/mile)
1 .7 gm/mi;l e "
1 .7 gm/mile
0.39
(0.8 gm/mile)
Test procedure given prior to 72 is 7-mode; between 72-74, CVS-72; and
CVS-75 for 75 and later.
Federal standard is value in parentheses.
-------
10
TABLE 2.5 DEFINITION OF-MODEL-YEAR GROUP FOR
LIGHT-DUTY GASOLINE-POWERED TRUCKS
(LDGTl) FOR CO (GVW £6000 Ibs)
Model -Year
California
pre-66
66-67
68-69
70-71
72-74
75
76-78
79-80
---
81 +
Group
Federal
pre-68
6S-.69
68-69
70-71
72-74
75-78
75-78
79-80
81
82
85+
Major Emission
Control Devices
None
Air Pump
Air Pump
Air Pump (California
71 has EGR)
Air Pump, EGR
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
3
-------
11
TABLE 2.6 DEFINITION OF MODEL-YEAR GROUP FOR
LIGHT-DUTY GASOLINErPOWERED TRUCKS
(LDGT1) FOR NOV (GVW <_ 6000 Ibs)
X
Model -Year
California
pre-66
66-70
71
72-73
74
75
76-78
79-80
---
81 +
Group
Federal
pre-68
68-72
73,74
73-74
73-74
75-78
75-78
79-82
83-84
85+
Major Emission
Control Devices
None
Air Pump
Air Pump, EGR
Air Pump, EGR
Air Pump, EGR
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
Air Pump, EGR,
Oxidizing Catalyst
3^-way Catalyst
Emission Standard3
None
None
4,0 gm/mile b
(.3.0 gm/mile)
3.0 gm/mile
2.0 gm/mile
(3.0 gm/mile)
2.0 gm/mile
(3.1 gm/mile)
2.0 gm/mile
(3.1 gm/mile)
1 . 5 gm/mi 1 e
(2.3 gm/mile)
2.3 gm/mile
1 ,0 gm/mile
(.1 , 4 gm/mi 1 e )
Test procedure given prior to 72 is 7-mode; between 72-74, CVS-72; and
CVS-75 for 75 and later.
Federal standard is value in parentheses.
-------
12
TABLE 2.7 DEFINITION OF MODEL-YEAR GROUP FOR MEDIUM-
DUTY GASOLINE-POWERED TRUCKS (LDGT2) FOR HC
Model -Year
California
pre-69
69-71
72
73-74
75-76
77
78-80
___
81 +
Group
Federal3
pre-70
70-73
70,73
74*78
74-78
74.78
79-80
81
82
85+
Major Emission
Control Devices
None
Air Pump, Engine
Modification
Air Pump, Engine
Modification
EGR
EGR
EGR
EGR, Oxidizing
Catalyst
EGR, Oxidizing
Catalyst
EGR, Oxidizing
Catalyst
3-way Catalyst
Emission Standard
None
275 ppm
180 ppm h
(275 ppm)D
None
None
1,0 gm/BHP-hr
(None)
0.9 gm/mile
(1.7 gm/mile)
(1 ,7 gm/mile)
(1.7 gm/mile)
0.39 gm/mile
(0.8 gm/mile)
aTest procedure given prior to 72 is 7-mode; between 72-74, CVS-72; and
CVS-75 for 75 and later.
Federal standard is value in parentheses.
-------
13
TABLE 2.8 DEFINITION OF MODEL-YEAR GROUP FOR
MEDIUM-DUTY GASOLINE-POWERED TRUCKS
(LDGT2) FOR CO
Model -Year
California
pre-69
69-71
72
73-74
75-76
77
78-80
81-82
83+
Group
Federal9
pre-70
70-73
70-73
74-78
74-78
74-78
79-80
81
82
83-84
85+
85+
Major Emission
Control Devices
None
Air Pump, Engine
Modification
Air Pump, Engine
Modification
EGR
EGR
EGR
EGR, Oxidizing
Catalyst
EGR, Oxidizing
Catalyst
EGR, Oxidizing
Catalyst
EGR, Oxidizing
Catalyst
3<-way Catalyst
3-?way Catalyst
Emission Standard
None
1.5%
1.0% .
0,5%}b
40 gm/BHP-hr
30 gm/BHP-hr
(40 gm/BHP-hr)
25 gm/BHP-hr
(40 gm/BHP-hr)
17 gm/mile
(18 gm/mile)
(18 gm/mile)
(18 gm/mile)
(18 gm/mile)
9 gm/mile
(.10 gm/mile)
7 gm/mile
(10 gm/mile)
Test procedure given prior to 72 is 7-mode; between 72-74, CVS-72; and
CVS-75 for 75 and later.
Federal standard is value in parentheses.
-------
14
TABLE 2.9 DEFINITION OF MODEL-YEAR GROUP FOR MEDIUM-
DUTY GASOLINE-POWERED TRUCKS, (LDGT2) FOR NO
, Model -Year
California
pre-69
69-71
72
73-74
75-76
77
78-80
---
81-82
83+
Group
Federal3
pre-70
70-73
70-73
74-78
74-78
74-78
79-82
83-84
85+
85+
Major Emission
Control Devices
None
Air Pump, Engine
Modification
Air Pump, Engine
Modification
EGR
EGR
EGR
EGR, Oxidizing
Catalyst
EGR, Oxidizing
Catalyst
3-way Catalyst
3-way Catalyst
Emission Standard
None
None
None
16 gm/BHP-hr
10 gm/BHP-hr h
(16 gm/BHP-hr)
5 gm/BHP.hr
(16 gm/BHP-hr)
2.3 gm/mile
(2,3 gm/mile)
1 . 5 gm/mi 1 e
(0.9 gm/mile)
1 . 0 gm/mi 1 e
(0.9 gm/mile)
Test procedure, given prior to 72 fs 7-mode; between 72-74, CVS-72; and
CVS-75 for 75 and later,
Federal standard is value in parentheses.
-------
15
TABLE 2.10 DEFINITION OF MODEL-YEAR GROUP FOR
HEAVY-DUTY GASOLINE-POWERED VEHICLES(HDGT)
Model -Year Group
California. Federal
Hydrocarbons
pre-69 pre-70
69-71 70-73
72 70-73
73-74 74-78
75-76 74-78 .
77-79 79-83
80-82 79-83
83+ 79-83
Carbon Monoxide
pre-69 pre-70
69-71 70-73
72 70-73
73-74 74-78
75.76 74-78
77-79 79-83
80-82 79-83
83* 79-83
Oxides of Nitrogen
pre-69 pre-70
69-71 70-73
72 70-73
73-74 74-78
75-76 74-78
77-79 79-83
80-82 79-83
83+ 79-83
Major Emission
Control Devices
None
Air Pump
Air Pump
Air Pump, EGR
Air Pump, EGR
Air Pump, EGR
None
Air Pump
Air Pump
Air Pump, EGR
Air Pump, EGR
Air Pump, EGR
None
Air Pump
Air Pump
Air Pump, EGR
Air Pump , EGR
Air Pump , EGR
Emission Standard
None
275 pom
180 ppm (275 ppm)
None
None
1.5 qm/BHP-hr
1.0 gm/BHP-hr
(1.5 gm/BHP-hr)
0.5 gm/BHP-hr
(1.5 gm/BHP-hr]
None
1.5%
1.0% (1.5%) '
40 gm/BHP-hr
30 gm/BHP-hr(40 gm/BHP-hr)
25 gm/BHP-hr
25 gm/BHP-hr
25 gm/BHP-hr
tin., HC + MOX
i, 1 1 . -* -
None None
None None
None None
None 16
None 10 (16)
7. 5 (None) 5 (5-10)
None 5^6 (5-10)
None 4.5 (5-10)
Federal standard is value in parentheses,
-------
16
TABLE 2.11 DEFINITION OF MODEL-YEAR GROUP FOR LIGHT-
DUTY DIESEL-POWERED VEHICLES (LDDV)
Model -Year
California
Hydrocarbons
pre-75
75-76
77
78
79
80
81
82+
Carbon Monoxide
pre-75
75-76
77
78
79
80
81
82+
Oxides of Nitrogen
pre-75
75-76
77
78
79
80
81
82+
Group
Federal
pre-75
75-76
77
78
79
80+
80+
80+
pre-75
75-76
77
78
79
80+
80+
80+
pre-75
75-76
77
78
79 or 80
81-82
81-82
83+
Emission Standard
None
1.5 gm/mile
1 .5 gm/mile
1 . 5 gm/mi 1 e
1 . 5 gm/mi 1 e
0.39 gm/mile (0.41 gm/mile) a
0.41 gm/mile
0.39 gm/mile (0.41 gm/mile)
None
15 gm/mile
15 gm/mile
15 gm/mile
15 gm/mile
9.0 gm/mile (7.0 gm/mile)
3.4 gm/mile (7.0 gm/mile)
7.0 gm/mile
None
3.1 gm/mile
2.0 gm/mile
2.0 gm/mile
2.0 gm/mile
1.5 gm/mile (1.0 gm/mile)
1.5 gm/mile (1.0 gm/mile)
1 .0 gm/mile
Federal standard is value in parentheses,
-------
17
TABLE 2.12 DEFINITION OF MODEL-YEAR GROUP3 FOR LIGHT-
DUTY DIESEUPOWERED TRUCKS (LDDT1) (GVW < 6000 Ibs)
"Model-Year
California
Inertia!
Weight
0-3999 Ibs
pre-78
78
79+
Inertia!
Weight
4000-6000 Ibs
pre-78
78-82
83+
Groupa
Federal
pre-79
79
80-82
83-84
85+
pre-79
79
80-82
83-84
85+
i
Emissions
HC
None h
(None,2.0)
0.7)
(1 7)
\ < ' i
0.9
(0.8)
0.41
(0.8)
None
(None, 2.0)
(1.7)
(1.7)
0.5
(0.8)
0.5
(0.8)
Standard (.in
CO
None ,
(None,20)D
(18.0)
(18 0}
^ i o. u;
17.0
(10.0)
9.0
(10.0)
None
(None, 20.0)
(18.0)
(18.0)
9.0
(10.0)
9.0
(10.0)
gm/mile)
NOX
None h
CNone,3.1)D
(2.3)
(2 3)
\£. . J J
2.0
(2.3)
1.5
(0.9)
None
(None, 3,1)
(2.3)
(2.3)
2.0
(2.3)
1.5
(0.9)
Model-year groups are the same for all three pollutants.
Federal standard is value in parentheses.
-------
18
TABLE 2.13 DEFINITION OF MODEL-YEAR GROUP0 FOR LIGHT-
DUTY DIESEL-POWERED TRUCKS (LDDT2)
(6001 < GVW < 8500)
Model -Year
California
Inertia!' Weight
0-3999 Ibs
pre-78
78-80
81 +
Inertia! Weight
4000-6000 Tbs
pre-78
78-80
81-82
83+
Inertia! Weight
6001-8500 Ibs
pre-78
78-80
81-82
83+
Group3
Federal
pre-78
83-84
85+
pre-78
83-84
83-84
85+
pre-78
83-84
83-84
85+
Emission
HC
None
0.9
(0.8)b
0.39
(0.8)
None
0.9
(0.8)
0.39
(0.8)
0.41
(0.8)
None
0.9
(0.8)
0.5
(0.8)
0.5
(0.8)
Standard
CO
None
17.0
(10.0)b
9.0
(10.0)
None
17.0
(10.0)
9.0
(10.0)
7.0
(10.0)
None
17.0
(10.0)
9.0
(10.0)
9.0
(10.0)
(gm/mile)
NOX
None
2.3
1.0
(0.9)
None
2.3
1.5
(2.3)
1.0
(0.9)
None
2.3
2.0
(2.3)
1.5
(0.9)
Model-year groups are the same for all three pollutants.
""Federal standard is value in parentheses.
-------
19
TABLE 2.14 DEFINITION OF MODEL-YEAR GROUP FOR HEAVY-DUTY
DIESEL-POWERED VEHICLES
~Model *Year-Groups
California Federal
Emission Standard
Hydrocarbons
pre-77
pre-85
77-79
80-82
83+
Carbon Monoxide
pre-77
77-82
83+
Oxides of Nitrogen
pre-85
pre-85
85+
pre-85
pre-85
85+
r-«.
r^.
i
pre-77 pre-85
77-79
80-82
83+
pre-85
pre-85
85+
California
None
1.0 gm/BHP-hr
1.0 gm/BHP-hr
0.5 gm/BHP-hr
pre-73 = None
73-74 = 40 gm/ g
BHP-hr '
75-76 = 30 gm/ £
BHP-hr
25 gm/BHP-hr
25 gm/BHP-hr
Federal
! pre-78 = None
79-83-= 1.5 gm/BHP-hr
84 = 1.3 gm/BHP-hr
Same as above
Same as above
1.3 gm/BHP-hr
pre-73 = None
73 = 1.5%
74-78 = 40 gm/BHP-hr
79-83 = 25 gm/BHP-hr
84 = 15.5 gm/BHP-hr
Same as above
15.5 gm/BHP-hr
None
7.5 gm/BHP-hr
None
None
;co
I pre-84 = None
84 = 10.7 gm/BHP-hr
Same as above
Same as above
1.7 gm/BHP-hr
-------
20
TABLE 2.15 DEFINITION OF MODEL-YEAR GROUP FOR MOTORCYCLES
Model -Year Group
California
Hydrocarbon
pre-78
78-79
80-81
82+
Carbon Monoxide
pre-78
78-79
80-81
82+
Oxides of Nitrogen
pre-78
78-79
80-81
82+
Federal
pre-78
78-79
80+
80+
pre-78
78-79
80+
80f
pre-78
78-79
80+
80+
Emission Standard
None
50-169 cc
170-749 cc
Over 749 cc
5.0 gm/km
1 . 0 gm/ km
None
17 gm/km
12 gm/km
12 gm/km
None
None
None
None
5.0 gm/km
5.0 + 0.0155
(D-170) gm/km
14 gm/km
j
(5.0 gm/km)a
aFederal standard is number in parentheses.
-------
21
3. HOMOGENEITY OF THE SIX ARB DATA SETS
The Air Resources Board in El Monte, California, has performed a series
of Surveillance Test Programs for several types of California vehicles. Data
on vehicle characteristics have been collected for each of these vehicles,
and constant-volume sampling (CVS-75) emissions tests were performed. Under
this program there were thirteen reports published by the ARB. TSC examined
data contained in the following six reports:
1. 1975-1976 Model Year Surveillance Test Program Report (CARB Report No. 1),
2. Final Report of the High Mileage Catalyst Vehicle Surveillance
Test Program, 1st Series, CVS (CARB Report No. 4).
3. Medium Duty Vehicle Surveillance Test Program Series II, CVS
(CARB Report No. 7).
4. Twenty Additional Datsun Motor Vehicle Surveillance Test
Program Report (CARB Report No. 8).
5. Limited Surveillance Program to Determine Whether Early
Catalyst Failures are Occurring in 1975 Model Year Vehicles
(CARB Report No. 11).
6. Final Report of the Light^Duty Vehicle Surveillance Test Program,
2nd Series, CVS (LDVSP-ir) (CARB Report No. 12).
The objective of this chapter is to examine the six ARB reports for
information relating to vehicles and to CVS-75 emissions testing, and to
determine whether the data from the different reports can be combined.
3.1 VEHICLE SELECTION AND CVS-75 EMISSION TEST
According to the ARB reports, each data set was selected for a
different purpose. For Report No. 1, the vehicles were selected randomly
by a private-vehicle procurement contractor, to guarantee"representativeness
-------
22
of 1975 and 1976 model-year vehicles. Report No, 4 examined high-mileage
catalyst vehicles. The samples are from 1975 and 1976 model-year vehicles
having more than 50,000 miles. The data set for Report No. 7 was
representative of the California medium-duty truck population, which was
selected by Systems Controls, Inc. Report No. 8 had 20 additional vehicles
that were randomly selected to be representative of 1973 through 1977 model-
year Datsun vehicles. In Report No. 11, most of the vehicles were obtained
from car dealers or rental agencies; they were not selected to represent
sales. As for Report No. 12, the samples were designed to represent the
California State population.
Although the sampling purposes were different for each report, the
testing procedure of CVST75 was exactly the same, Each vehicle had a CVS-
75 test "as received," which required a 12-hour cold soak prior to testing.
Then, bag samples were taken during the cold-start, cold-transient, and
hot-start cycles. The emission concentrations of hydrocarbons (HC), carbon
monoxide (CO), and nitrogen oxides (NO ) as determined from each of these bag
A
samples were used to calculate the average grams per mile over the total test
distance.
3,2 COMBINED DATA SET
Since the test procedure for each report was the same, all the data
can be included in the data bank. The combined data set cannot be assumed
to be representative of the California State fleet population, however,
because the sampling plan for each report was different. The combined data
-------
23
set is valid for the study of each model-year group because it contains
both high-mileage and low-mileage vehicles in each group.
Twenty-one Datsun vehicles are replicated in Report Nos. 1 and 4.
When data sets were combined, the replicates were deleted from Report No. 4's
data set.
In the combined data set, there are 781 light-duty vehicles, 96
light-duty trucks, and 75 medium-duty trucks. The total number of vehicles
is 952.
-------
24
4. DATA VERIFICATION FOR CARB DATA SET
Before the data could be analyzed, it was necessary to perform data
verification on the data set received. This was done by scrutinizing the
values in the data set and then comparing them to those given in the ARB
reports. When a value was found to be missing or incorrect, it was changed
to reflect the information given in the reports. Table 4.1 summarizes the
changes that were made in the data set. The following sections describe in
detail the changes that were made in the data set and state the reason for
changes. Each section describes one ARB report.
4.1 CARB REPORT NO. 1
In this report, 11 of the vehicles that were American makes were
actually manufactured by Japanese companies. Three Chrysler cars had been
manufactured by Mitsubishi, four GM cars had been manufactured by Isuzu, and
four Ford cars had been manufactured by Mazda. The values of the manufacturer
variables were changed to reflect this fact. A valud had already been assigned
for Mazda; however, there were no values for Mitsubishi or Isuzu. Therefore,
the following values were assigned to them: 36 for Mitsubishi, and 37 for
Isuzu. The four Ford's were changed from a 6 (Ford) to an 11 (Mazda). The
only other change was that one car had "EG" as one of its emission control
devices. This was changed to "EGR" and was most likely a keypunch error.
4.2 CARB REPORT NO. 4
All of the vehicles tested for this report were light-duty vehicles;
however, all the values for the vehicle-type variable had been set to zero
(no code). These values were changed to 1 (light-duty vehicles) for all 66
-------
25
TABLE 4.1 SUMMARY OF CHANGES MADE DURING DATA VERIFICATION
CARB Report
Number Variable
1 MFG
MFG
MFG
EM Dev
4 Veh Type
P or F
P or F
P or F
MYR
MYR
EM Dev
EM Dev
7 Veh Type
MFG
MFG
MFG
MFG
MFG
Make
MFG
Make
8 CA/Fed
11 Veh Type
CID
CID
CID
EM Dev
EM Dev
12 .MFG
MFG
MFG
MFG
MFG
MFG
Make
EM Dev
EM Oev
EM Oev
EM Oev
EM Dev
EM Dev
EM Dev
EM Dev
Veh Type
CA/Fed
Odom
Iner Wt
Cyl
Trans
Change Number
of
From To Vehicles Vehicle Numbers
9
1
6
EG-...
0
10000
10000
20000
75
76
(Blank)
...-CO
0
0
0
0
0
0
99
0
31
0
0
2300 (cc)
2800 (cc)
1400(cc)
(Blank)
(Blank)
6
6
6
9
1
. 1
0
CVCC
*
REGR
...-EG
...-E
...-EFE-
..,-E
...-EGR-
2
8
5841
5003
1
5
36
37
n
EGR-...
1
1
2
3
76
75
AIR-EGR-OC
...-OC
3
6
21
1
9
38
38
98
98
1
1
140(cu.in.)
I7l(cu.1n.)
85(cu.in.)
AI-EGR-OC
EGR-OC
14
33
39
36
37
17
1
EM
EM
EGR
. . . -EGR
...-EFI
...-EFE-OC
. . . -EFE
... -EGR-OC
1
1
35841
5500
8
1
3
4
4
1
66
8
6
52
.
2
5
1
2
75
38
4
15
13
1
1
4
4
20
182
9
1
2
130
43
25
6
2
6
10
1
1
1
2
1
1
1
1
1
1
1
2
1
1
1
1
109,150,172
11,27,180,182
92,107,110,133
84
1-64 (All)
3,14,15(-1),18,24,39,50,64,
11, 15(-2), 17,40,43,60
1,2,4-10,12,13,16,19-23, 25(-l),
25(-2), 26-38, 41 ,42, 44-49,
51-59,61-63
7,57
8,9,32,56,61
16
26,64
All
All Ford trucks
77,63,50,12
47,58,15,17,35,62,10,14,29,
57,60,4,56,36,22
9,37,7,13,42,64,65,27,61,54,
80,44,52
70
70
31,81,53,48
31,81,53,48
1-20 (All)
All
20001,20004,20006,20023,
20041 ,20073,21077,21079,21081
21078
23001 ,23008
All AMC, Chrysler and GM
(except Suick); 6 Imports
All Ford (exceot for the 9 with
2300 cc CID) and GM (Buick); 2 Imports
1001,1004,1006,1035,4066-4082,
4085,4088,4089,4091
4170,4131 ,4334,4344,4376,4335
4027,4093
4323 ,4329 >4342 ,4373 ,4397 ,4398
4056,4175,4215,4260,4261 ,4262, .
4339,4386,4391,4392
4332
4332
1020
4018,4039
4296
4260
4330
4378
4382
4383
4237
4213,4214
4213
4213
4213
4213
-------
26
cases. The values of the pass/fail variables for all the cases had been
set to 10000 for pass or pass marginally, and 20000 for fail. Eight of the
values were changed to 1 for pass; six were changed to 2 for pass marginally;
and 52 were changed to 3 for fail. The changes to pass or pass marginally
were based on the data listings in the report. The model-year in the data
set for seven of the vehicles did not agree with the report. Two were changed
from 75 to 76, and five were changed from 76 to 75. One vehicle was missing
the emission control device listed for it in the report, so the code for the
device was added. Finally, two vehicles had CO listed as one of their
emission control devices and they were changed to OC. The disagreement
between the data set and the report for the model-year and emission-control-
device variables were most likely due to keypunch errors.
4.3 GARB REPORT NO. 7
All of the vehicles tested for this report were medium-duty trucks;
however, all the values for the vehicle-type variable had been set to zero
(no code). These values were changed to 3 for all 75 cases. Secondly, while
values were present for the make variable, the values for the manufacturer
variable had all been set to zero. These were changed according to the tables
in the report. Thirty-eight were Ford trucks, four were GMC, 15 were GM,
and 13 were Chrysler. One vehicle was a Jeep, which had no value assigned to
it. We assigned Jeep the value 38 and changed both the manufacturer and make
(which had been 99 for "other") to this value. Four of the vehicles were
International Harvester. The value 31 is a Renault if the vehicle is a car
or International Harvester if the vehicle is a truck. To avoid confusion, we
-------
27
decided to leave Renault with the value 31 and reassign International
Harvester to the value of 98. Therefore, for the four International Harvester
vehicles, we changed the values of the manufacturer variable from 0 to 98,
and the make variable from 31 to 98.
4.4 CARB REPORT NO. 8
For this report, all 20 vehicles had a value of zero for the California/
Federal variable. These values were changed to 1 to indicate California
emission controls.
4.5 CARB REPORT NO. 11
All of the vehicles tested for this report were light-duty vehicles,
but all the values for the vehicle-type variable had been set to zero. These
values were changed to 1 for all 182 cases. Twelve of the vehicles had values
for their cubic-inch displacement (CID) in cubic centimeters (cc),so these
values were converted into cubic inches. Nine Fords were changed from 2300 cc
to 140 cu. in.; one Ford was changed from 2800 cc to 171 cu. in.; and two
Datsuns were changed from 1800 cc to 85 cu. in. Finally, no emission control
-\
devices were listed for the vehicles. To determine most likely devices
for these vehicles, we searched through the other reports to see which devices
similar vehicles had, and then assigned those same devices to the vehicles in
Report No. 11. Vehicles for which no information was found, the nine Fords
which had 2300 cc for their cu. in. displacement, were left blank. The
emission control device was changed to AI-E6R-OC for 130 cases, including
all the AMC, Chrysler, and GM (except for Buick) vehicles and six imported
-------
28
cars. The emission control device was changed to EGR-OC for 43 cases,
including all the Fords (except for the nine vehicles left blank) and
Buick-GM vehicles and two imported cars.
Another problem which was noticed, but for which, no change was made,
concerned the use of the same vehicle number for more than one vehicle. In
the report, each vehicle was identified by a test number consisting of a
series of letters and numbers, which was converted to the vehicle number
in the data set by taking the number portion of the test number and adding
it to 1000 times the page number on which the data appeared in the report.
Thus, test number CAL 74 on page 21 became vehicle number 21074. For most
of the pages, the letter portions of the test number would remain the same
and the number portions would change. But for the imported cars on page
23, some of the test numbers had the same number portion and differed only
in the letter portion. Thus, the conversion resulted in assigning the same
number to more than one vehicle. The vehicles with the same numbers had
different manufacturers and makes, so they can be distinguished from one
another by those variables in th.e data set, but one cannot refer to them by
vehicle number alone.
4.6 CARB REPORT NO. 12
A number of the vehicles of American make in this report were actually
manufactured by companies in other countries. The manufacturer variables
were changed to the correct code for foreign manufacturers, according to the
information given in the report: the values for the 25 Fords manufactured by
Nissan were changed to 14 for Datsun; values for the 6 Fords that were Couriers
-------
29
were changed to 33 (Courier); the values for the 1 Ford that was a Capri
and the other that was a Cortina, both being manufactured in Europe, were
changed to 39 to represent European-made Ford passenger vehicles. Six
Chryslers were manufactured by Mitsubishi, so their values were changed to
36; and 10 GM vehicles were manufactured by Isuzu, so their values were
changed to 37. These values, 36 and 37, were assigned to these manufacturers
in Section 4.1 for vehicles in CARB Report No. 1. One vehicle was listed as
GM-Opel; so its manufacturer variable was changed from 1 to 17 for Opel, and
.its make variable was changed from 0 to 1 for Buick. . All the vehicles in
this report had a value of zero for the make variable; the Opel was the
only vehicle for which that value was changed.
Four of the vehicles had what appeared to be keypunch errors in
their lists of emission control devices. One was changed from CVCC to EM;
two were changed from "*" to EM; and one was changed from REGR to EGR. Five
other vehicles had errors in their lists of emission control devices that
could have been due to a format error in the program generating the tape.
For all five vehicles, the list of devices had been chopped off after the
first 11 characters, so the missing letters were added. The changes were
from EM-OC-FI-EG to EM-OC-FI-EGR; from AI-EGR-OC-E to AI-EGR-OC-EFI; from
AI-EGR-EFE to AI-EGR-EFE-OC; from OC-EGR-AI-E to OC-EGR-AI-EFE; and from
EM-EFI-EGR to EM-EFI-EGR-OC.
The rest of the errors that were found appear to be keypunch errors.
One vehicle was incorrectly listed as a light-duty truck, so the value for
the vehicle type was changed from 2 to 1. Two vehicles had values of 8 for the
California/Federal emissions-control variable; these were changed to 1. One
-------
30
of these two vehicles also had errors in the odometer, inertia! weight,
cylinders, and transmission variables, The changes were from 5841 to 35841
for the odometer, from 5003 to 5500 for the inertial weight, from 1 to 8
for the cylinders, and from 5 to 1 for the transmission.
-------
31
5. ANALYSIS
5.1 DATA BASE
In the ARB data set, 427 cases are coded as California vehicles, 22 as
Federal vehicles, and 503 have no code. For the purpose of this task, only
California vehicles will be considered in the analysis; the 22 Federal vehicles
were deleted from the data base. As for the remaining 503 uncoded vehicles,
*
we know that there is a 95% chance they are California vehicles, and have
included them in the data base to get a good sample size for each model-year
group. For the non-Federal vehicles, 763 are light-duty gasoline-powered
vehicles (LDGV), 92 are light-duty gasoline-powered trucks (LDGT1), and 75
are medium-duty gasoline-powered trucks (LDGT2). The breakdown of emission
control devices and model-year of the vehicles for each vehicle type are
shown in Table 5.1 and Table 5.2. In Table 5.1, we note that there are three
cars with emission control devices coded as "CARTER" or "CARTER RST." Since
the meanings of these codes are unclear, we dropped those cars from the data
base. The emission control devices were not coded for the medium-duty trucks,
which were in ARB Report No. 7. The breakdown by model-year is shown in
Table 5.3.
5.2 DESCRIPTIVE STATISTICS OF THREE EMISSIONS UNDER EACH MODEL-YEAR GROUP
For each vehicle type (LDGV, LDGT1 and LDGT2), the vehicles are
grouped by the definitions discussed in Chapter 2. Under each model-year
*
K
According to the conversation with the staff at ARB, El Monte,
California.
-------
32
TABLE 5.1 BREAKDOWN OF MODEL-YEAR AND EMISSION CONTROL DEVICE FOR
LIGHT-DUTY VEHICLES
Emission
Control
Device
None (or
no code)
FE
EM
EGR
EGR-FE
EGR-EM-EFE
EGR-TR
EGR-OC
EGR-OC-EFE '
EGR-OC-FE
EGR-OC-EM-EFE
EGR-OC-EM-FE
AI
AI-TR
AI-OC
AI-OC-FE
AI-OC-EM
AI-EGR
AI-EGR-EFE
AI-EGR-TR
AI-EGR-TR-EM
AI-EGR-OC
AI-EGR-OC-EFE
AI-EGR-OC-FE
AI-EGR-OC-EM
AI-EGR-OC-EM-FE
CARTER
CARTER RST
NOX or RNOX
Total
Model -Year
68 69 70 71 72 73
17 12 22 16 18 10
21 1
111222
2 2 1 11
1
221471
1
2 9
2
2
3 3
25 24 26 24 30 35
74 75
8
5
10
1 1
1
56
6
4
1
2
1 1
3
4
20
1
4
220
4
1
8
34 328
76
7
17
8
1
1
6
1
3
2
1
5
1
107
5
3
5
1
174
77 Total
1 104
4
21
26
3
1
1 1
73
14
3 8
1 2
1 2
19
3
9
1
7
33
2
2 11
1
52 379
2 11
4
13
1
2
2
6
63 763
-------
33
TABLE 5.2 BREAKDOWN OF MODEL-YEAR AND EMISSION CONTROL
DEVICE FOR LIGHT-DUTY TRUCKS
Emission
Control
Device
None (or
no code)
EGR
EGR-EM-EFE
EGR-OC
EGR-OC-EFE
EGR-OC-EM-EFE
AI
AI-OC
AI-EGR
AI-EGR-TR
AI -EGR-OC
AI -EGR-OC-EFE
AI- EGR-OC- EM
Total
Model-Year
72 73 74 75 76
1
1 1
2
2 8
1
1
1 4
1
13
2
6 22
1
2
1 1 1 31 35
77 Total
1
2
2
4 14
1 2
1
5
1
13
2
18 46
1
2
23 92
-------
34
TABLE 5.3 BREAKDOWN BY MODEL-YEAR FOR MEDIUM-DUTY TRUCKS
MODEL-YEAR
67
68
69
70
71
72
73
74
75
76
77
78
Total
# OF CASES
3
1
2
3
2
3
3
6
14
16
15
7
75
-------
35
group, the descriptive statistics for both emissions level and mileage
are calculated. The results,for light-duty vehicles are shown in Tables 5.4
through 5.9. We note that there are two declines among the model-year
groups: a small one between 1969 and 1970 and a large one between 1974 and
1975. This phenomenon appears for all three emissions.
For light-duty trucks the descriptive statistics for the three emis-
sions are shown in Tables 5.10 through 5.12. There is also a decreasing
trend for model-year 1972-74; however, the sample size is too small for us
to make any strong conclusions. The situation is similar for medium-duty
trucks, the results for which are given in Tables 5.13 through 5.16.
5.3 REGRESSION ANALYSIS BETWEEN EMISSIONS AND ODOMETER READINGS
The usual method of determining the deterioration factor (the increase
in emissions with respect to mileage) is to find the linear relationship
between emissions and odometer readings. The linear equation is written
in the form E = b + dM, where E is the emissions, b is the new-car emissions
(intercept), d is the deterioration factor, and M is the mileage in thousands
of miles. The coefficients are estimated, using regressions, by the following
model: E = b + dM + e, where e is the residual error.
Tables 5.17 through 5.23 give the regression results, derived using
a linear model for hydrocarbon, carbon monoxide, and nitrogen oxides emis-
sions for the three types of gasoline-powered vehicles, LDGV, LDGT1, and
LDGT2.
In general, the percent variation explained (PVE) by the model
p
(100-r ) is small (less than 20%). This PVE indicates that for this data
-------
36
the linear model is not appropriate (in the statistical sense) for describing
the change in emissions with respect to mileage.
If the residual error e is normally distributed with mean zero and
2
variance a , we can use the F-value for the analysis of variance for the
regression, to test if the coefficient d (deterioration factor) is signi-
ficantly different from zero. Even if e is not normally distributed, but
the residual is independent with respect to the emission levels, a large
F-value could indicate that d was significantly different from zero at the
95% level. This large F-value occurred in only six cases:
1. LD6V, hydrocarbon, 75-76, b = 0.51, d = 0.021, F = 39.41 with
(1,500) degrees of freedom. 7.3% of the variation was explained
(estimated over the range of 12 miles to 88,860 miles) (Table 5.17).
2. LDGV, hydrocarbon, 77-79, b = 0.32, d = 0.013, F = 9.16 with (1,61)
degrees of freedom. 13.1% of the variation was explained (estimated
over the range of 1,319 miles to 30,037 miles) (Table 5.17).
3. LDGV, carbon monoxide, 75-76, b = 7.76, d = 0.16, F = 23.05 with
(1,500) degrees of freedom. 4.4% of the variation was explained
(estimated over the range of 12 miles to 88,860 miles) (Table 5.18).
4. LDGV, nitrogen oxides, 75-76, b = 1.99, d = 0.0079, F = 9.63 with
(1,500) degrees of freedom. 1.9% of the variation was explained
(estimated over the range of 12 miles to 88,860 miles) (Table 5.19).
5. LDGT1, hydrocarbon, 75, b = -0.727, d = 0.137, F = 6.62 with (1,29)
degrees of freedom. 19% of the variation was explained (estimated
over the range of 1,857 miles to 84,866 miles) (Table 5.20).
6. LDGT2, nitrogen oxides, 78-80, b = 2.794, d = -0.068, F = 14.97
with (1,5) degrees of freedom. 76% of the variation was explained
(estimated over the range of 2,054 miles to 20,585 miles) (Table 5.23).
Case 6 should be rejected because of sample size (only 7 vehicles). The
negative intercept in Case 5 would indicate that the linear model should not be
extended outside of the mileage range. The percent variation explained by the
other four cases is noise level. Thus, it could be said that a linear model to
describe emissions using odometer mileage readings is not appropriate for this data,
-------
37
TABLE 5.4 DESCRIPTIVE STATISTICS OF HYDROCARBON EMISSIONS FOR CALIFORNIA
LIGHT-DUTY GASOLINE VEHICLES IN EACH HYDROCARBON MODEL-YEAR
GROUP (in gin/mile)
Model-Year Group
pre-66
66-67
68-69 '
70-71
72-73
74
75-76
77-79
80+
Sample Size
None
None
46
50
65
34
502
63
None
Means
6.62
4.41
4.79
4.54
0.97
0.50
Standard
Deviation
5.63
2.42
5.96
5.50
1.63
0.23
Range
24.1
11.81
34.55
29.40
24.80
1.23
Minimum
1.72
1.08
1.18
1.23
0.13
0.14
Maximum
25.82
12.89
35.73
30.63
24.93
1.37
Total 760 1.99 3,44 35.60 0.13 35.73
-------
38
TABLE 5.5 DESCRIPTIVE STATISTICS OF MILEAGE FOR CALIFORNIA LIGHT-DUTY
GASOLINE VEHICLES IN EACH HYDROCARBON MODEL-YEAR GROUP (in miles)
Model -Year
Group
pre-66
66-67
68-69
70-71
72-73
74
75-76
77-79
80+
Sample
Size
None
None
46
50
65
34
502
63
None
Means
79,476
84,519
63,750
47,776
21,561
13,707
Standard
Deviation .
26,270
26,309
21,523
15,274
20,743
6,715
Range
121,757
127,734
102,519
55,201
88,848
28,718
Minimum
27,006
24,407
14,661
24,389
12
1,319
Maximum
148,763
152,141
117,180
79,590
88,860
30,037
Total
760
33,338
30,600
152,129
12
152,141
-------
39
TABLE 5.6 DESCRIPTIVE STATISTICS OF CARBON MONOXIDE EMISSIONS FOR CALIFORNIA
LIGHT-DUTY GASOLINE VEHICLES IN EACH CARBON MONOXIDE MODEL-YEAR
GROUP (in gm/mile)
Model -Year Group
pre-66
66-67
68-69
70-71
72-74
75-76
77-79
80+
Total
Sample
Size
None
None
46
50
99
502
63
None
760
Means
76.55
62.76
52.71
11.37
6.06
23.64
Standard
Deviation
49.44
38.17
37.57
16.54
5.81
33.32
Range
185.22
159.36
211.16
133.51
25.27
217.65
Minimum .
9.02
13.63
6.80
0.31
1.38
0.31
Maximum
194.24
172.99
217.96
133.82
26.65
217.96
-------
40
TABLE 5.7 DESCRIPTIVE STATISTICS OF MILEAGE FOR CALIFORNIA LIGHT-DUTY
GASOLINE VEHICLES IN EACH CARBON MONOXIDE MODEL-YEAR GROUP
(in miles)
Model -Year Group
pre-66
66-67
68-69
70-71
72-74
75-76
77-79
80+
Sample
Size
None
None
46
50
99
502
63
None
Means
79,476
84,519
58,264
21,561
13,707
Standard
Deviation
26,270
26,309
20,957
20,743
6,715
Range
121,757
249,491
102,519
88,848
28,718
Minimum
27,006
24\407
14,661
12
1,319
Maximum
148,763
152,141
117,180
88,860
30,037
Total 760 33,338 30,600 152,129 12 152,141
-------
41
TABLE 5.8 DESCRIPTIVE STATISTICS OF OXIDES OF NITROGEN EMISSIONS FOR
CALIFORNIA LIGHT-DUTY VEHICLES IN EACH NOX MODEL-YEAR GROUP
(in gm/mile)
Model -Year
Group
pre-66
66-70
71
72-73
74
75-76
77-79
80+
Sample
Size
None
72
24
65
34
502
63
None
Mean
3.19
2.74
2.85
2.89
2.15
1.84
Standard
Deviation
1.36
1.23
1.14
1.60
1.19
1.13
Range
6.08
4/79
6.26
6.77
8.43
4.99
Minimum
0.50
0.47
0.53
0.41
0.41
0.72
Maximum
6.58
5.26
6.79
7.18
8.84
5.71
Total 760 2.34 1.27 8.43 0.41 8.84
-------
42
TABLE 5.9 DESCRIPTIVE STATISTICS OF MILEAGE FOR CALIFORNIA LIGHT-DUTY
VEHICLES IN EACH NOV MODEL-YEAR GROUP (in miles)
Model -Year
Group
pre-66
66-70
71
72-73
74
75-76
77-79
80+
Sample
Size
None
72
24
65
34
502
63
None
Mean
83,628
77,528
63,750
47,777
21,561
13,708
Standard
Deviation
25,246
29,233
2:1,523
15,274
20,743
6,715
Range
121,757
127,734
102,519
55,201
88,848
28,718
Minimum .
27,006
24,407
14,661
24,389
12
1,319
Maximum
148,763
152,141
117,180
79,590
88,860
30,037
Total
760
33,338 30,600 152,129
12
152,141
-------
43
TABLE 5.10 DESCRIPTIVE STATISTICS FOR CALIFORNIA LIGHT-DUTY
TRUCKS FOR EACH HYDROCARBON MODEL-YEAR GROUP
Model-Year
Group
Hydrocarbon
(in gm/mile)
72-74
75
76-78
Total
Odometer
(in miles)
72-74
75
76-78
Total
Sample
Size
3
31
58
92
3
31
58
92
Mean
5.08
3.22
0.80
1.76
51,728
28,825
16,413
21 ,747
Standard
Deviation
2.88
5.10
0.71
3.28
12,922
16,027
9,651
14,544
Range
5.35
26.85
3.32
27.44
25,576
83,009
49S543
83,092
Minimum
3.02
0.73
0.14
0.14
37,867
1,857
1,774
1,774
Maximum
8.37
27.58
3.46
27.58
63,443
84,866
51,317
84,866
-------
44
TABLE 5.11 DESCRIPTIVE STATISTICS FOR CALIFORNIA LIGHT-DUTY
TRUCKS IN EACH CARBON MONOXIDE MODEL-YEAR GROUP
Model- Year
Group
Sample
' Size
Mean
Standard
Deviation
Range
Minimum
Maximum
Carbon Monoxide
(in gm/mile)
72-74
75
76-78
Total
3
31
58
92
47.57
21.96
8.71
14.44
42.06
16.13
9.36
16.00
84.05
59.64
49.37
89.32
6.50
7.11
1.23
1.23
90.55
66.75
50.60
90.55
Odometer
(in miles)
72-74
75
76-78
Total
3
31
58
92
51 ,728
28,825
16,413
21,747
12,922
16,027
9,651
14,544
25,576
83,009
49,543
83,092
37,867
1,857
1,774
1,774
63,443
84,866
51,317
84,866
-------
45
TABLE 5.12 DESCRIPTIVE STATISTICS FOR CALIFORNIA LIGHT-DUTY
TRUCKS IN EACH NITROGEN OXIDES MODEL-YEAR GROUP
Model -Year
Group
Nitrogen Oxides
(in gm/mile)
72-73
74
75
76-78
Total
Odometer
(in miles)
72-73
74
75
76-78
Total
Sample
Size
2
1
31
58
92
2
1
31
58
92
Mean
1.82
5.00
2.14
1.72
1.90
50,655
53,875
28,825
16,413
21,747
Standard
Deviation
0.13
1.04
0.44
0.79
18,085
16,027
9,651
14,544
Range
0.19
0.00
3.74
2.65
4.23
25,576
0
83,009
49,543
83,092
Minimum
1.72
0.77
0.93
0.77
37,867
1,857
1,774
1,774
Maximum
1.91
5.00
4.51
3.58
5.00
63,443
53,875
84,866
51,317
84,866
-------
46
TABLE 5.13 DESCRIPTIVE STATISTICS OF HYDROCARBON FOR CALIFORNIA
MEDIUM-DUTY TRUCKS IN EACH MODEL-YEAR GROUP (in gm/mile)
Model -Year
Group
pre-69
69-71
72
73-74
75-76
77
78-80
Sample
Size
4
7
3
9
30
15
7
Mean
9.67
7.64
7.60
5.95
6.23
2.94
1.33
Standard
Deviation
6.34
1.76
3.72
1.47
8.24
1.32
1.71
Range
13.21
5.12
7.04
4.23
43.78
4.72
4.79
Minimum
5.95
6.06
3.38
4.12
1.52
1.03
0.30
Maximum
19.16
11.18
10.42
8.35
45^30
5.75
5.09
Total
75
5.45
5.87
45.0
0.30
45.30
-------
47
TABLE 5.14 DESCRIPTIVE STATISTICS OF CARBON MONOXIDE FOR CALIFORNIA
MEDIUM-DUTY TRUCK'S IN EACH MODEL-YEAR GROUP (in gm/mile)
Model -Year
Group
pre-69
69-71
72
73-74
75-76
77.
78-80
Total
Sample
. Size
4
7
3
9
30
15
7
75
Mean
74.87
61.43
66.16
77.10
53.76
35.61
20.02
52.12
Standard
Deviation
23.17
11.90
10.52-
19.82
28.58
14.04
' 28.16
27.82
Range
55.52
30.96
21.03
65.60
98.16
51.14
78.54
108.54
Minimum
51.06
46.43
55.88
47.35
14.66
13.55
4.41
4.41
Maximum
103.58
77.39
76.91
112.95
112.82
64.69
82.95
112.95
-------
48
TABLE 5.15 DESCRIPTIVE STATISTICS OF OXIDES OF NITROGEN FOR CALIFORNIA
MEDIUM-DUTY TRUCKS IN EACH MODEL-YEAR GROUP (in gm/mile)
Model- Year
Group
pre-69
69-71
72
73-74
75-76
77
78-80
Total
Sample
Size
4
7
3
9
30
15
7
75
Mean
4.94
5.72
4.01
5.64
3.83
4.70
2.09
4.30
Standard
Deviation
1.39
2.01
0.33
2.35
2.03
1.79
0.52
2.07
Range
3.23
4.79
0.65
6.79
9.56
5.89
1.59
9.56
Minimum
3.65
3.31
3.71
1.90
0.41
1.95
1.03
0.41
Maximum
6.88
8.10
4.36
8.69
9.97
7.84
2.62
9.97
-------
49
TABLE 5.16 DESCRIPTIVE ODOMETER STATISTICS FOR CALIFORNIA MEDIUM-
DUTY TRUCKS IN EACH MODEL-YEAR GROUP (in miles)
Model -Year
Group
pre-69
69-71
72
73-74
75-76
77
78-80
Sample
Size
4
7
3
9
30
15
7
Mean
95,498
53,773
67,217
53,540
48,201
29,479
10,037
Standard
'Deviation
12,330
24,713
12,499
16,853
18,725
17,007
6,664
Range
29,104
66,649
24,361
55,248
95,379
74,729
18,531
Minimum
83',797
13,259
56,656
31,666
5,755
8,677
2,054
Maximum
112,902
79,908
81,017
86,914
101,134
83,406
20,585
Total 75 45,338 25,077 110,848 2,054 112,902
-------
TABLE 5.17 RESULTS OF REGRESSION ANALYSIS ON HYDROCARBON EMISSIONS AND
ODOMETER READINGS FOR CALIFORNIA LIGHT-DUTY GASOLINE VEHICLES
Model -Year ~
Group r
68-69
70-71
72-73
74
75-76
77-79
0.004
0.039
0.025
0.052
0.073
0.131
Mean Square
for Regression
5.93
11.12
57.47
52.11
96.83
0.44
Mean Square
for Residual
32.34
5.76
35.12
29.59
2.46
0.05
F-Val
Calculated
0.18
1.93
1.64
1.76
39.41*
9.16*
ue
95% level
4.06
4.04
3.99
4.15
3.86
4.00
(DFlfDF2)
(1,44)
(1,48)
(1,63)
(1,32)
(1,500)
(1,61)
Estimated
Intercept
(gm/mi)
7.72
2.88
1.98
8.47
0.51
0.32
Estimated Slope
(gm/mi per
1000 mi driven)
-0.014
0.018
0.044
-0.082
0.021
0.013
CJl
o
Significant at the 95% level.
-------
TABLE 5.18 RESULTS OF REGRESSION ANALYSIS ON CARBON MONOXIDE EMISSIONS AND
ODOMETER READINGS FOR CALIFORNIA LIGHT-DUTY GASOLINE VEHICLES
Model -Year ^
Group r
68-69
70-71
72-74
75-76
77-79
0.008
0.022
0.011
0.044
0.059
Mean Square
for Regression
825.46
1565.76
1585.93
6040.84
124.38
Mean Square
for Residual
2480.
1454.
1409.
262.
32.
77
37
92
12
26
/
F-Val
Calculated
0.33
1.08
1.12
23.05*
3.86
ue
95%
4
4
3
3
4
level
.06
.04
.94
.86
.00
(DFlfDF2)
(1,44)
(1,48)
(1,97)
(1,500)
(1,61)
Estimated
Intercept
(gm/mi)
63.59
44.60
41.53
7.76
3.17
Estimated Slope
(gm/mi per
1000 mi driven)
0.
0.
0.
0.
0.
16
21
19
16
21
Significant at the 95% level.
-------
TABLE 5.19 RESULTS OF REGRESSION ANALYSIS ON NITROGEN OXIDES EMISSIONS FOR
CALIFORNIA LIGHT-DUTY GASOLINE VEHICLES
Model -Year 2
Group r
66-70
71
72-73
74
75-76
77-79
0.003
0. 026
0.0004
0.041
0.019
0.054
Mean Square
for Regression
0.42
0.90
0.04
3.45
13.29
4.29
Mean Square
for Residual
1.86
1.55
.1.32
2.52
1.38
1.22
F-Value
Calculated
0.22
0.58
0.03
1.37
9.63*
3.51
95% level
3.98
4.30
3.99
4.15
3.86
4.00
(DFrDF2)
(1,70).
(1,22)
(1,63)
(1,32)
(1 ,500)
(1,61)
Estimated
Intercept
(gm/mi)
3.45
2.22
2.78
1.88
1.99
2.38
Estimated Slope
(gm/mi per
1000 mi driven)
0.003
0.0068
0.0011
0.0212
0.0079
0.0392
en
ro
Significant at the 95% level.
-------
TABLE 5.20. RESULTS FOR REGRESSION ANALYSIS FOR CALIFORNIA
LIGHT-DUTY GASOLINE-POWERED TRUCKS
Model -Year
Group
Hydrocarbon
' 72-74
75
76-78
Carbon Monoxide
72-74
75
76-78
Nitrogen Oxides
72-73
74
75
76-78
Sample
Size
3
31
58
3
31
58
2
1
31
58
r2
0.47
0.19
0.04
0.42
0.07
-.002
NA
NA
0.05
0.01
LI c- u r F-Vdl U6
M6dn ScjUflrfi M&dn SQUdt*G
for Regression for Residual Calculated 95% level
7.88 8.72 0.90 161
144.81 21.89 6.62* 4.18
1.08 0.50 2.18 4.02
1493.68 2043.99 0.73 161
516.33 251.25 2.06 4.18
8.79 89.02 0.10 4.02
1.51 1.06 1.43 4.18
0.15 0.19 0.79 4.02
(OF18DF2)
(1,1)
(1,29)
(1,56)
(1,1)
(1,29)
(1,56)
(1,29)
(1,56)
Estimated
Intercept
(gm/mile)
-2.870
-0.727
0.566
156.965
29.425
8.042
1.739
1.634
Estimated
Slope (gm/mile
per 1,000 miles
Driven)
0.154
0.137
0.014
-2.115
-0.259
0.041
0.014
0.004
en
co
Significant at the 95? level.
-------
TABLE 5.21, REGRESSION ANALYSIS BETWEEN HYDROCARBON AND ODOMETER READINGS
FOR CALIFORNIA MEDIUM-DUTY GASOLINE-POWERED TRUCKS
Model- Year
Group
pre-69
69-71
72
73-74
75-76
77
78-80
Sample
Size
4
7
3
9
30
15
7
r2
0.85
0.42
0.78
0.03
0.006
0.05
0.04
Mean Square
for Regression
102.07
7.78
21.55
0.49
12.72
1.26
0.64
Mean Square
for Residual
9.33
2.15
6.15
2.40
69.83
1.78
3.40
F-value
Calculated
10.95
3.62
3.50
0.21
0.18
0.71
0.19
95* level
18.51
6.61
161
5.59
4.20
4.67
6.61
(OF, ,DF2)
(1,2)
(1,5)
(1,1)
(1,7)
(1,28)
(1,13)
(1,5)
Estimated
Intercept
(gm/mlle)
-35.509
5.161
25.247
6.744
7.932
2.42
1.822
Estimated
Slope (gm/mlle
per 1,000 miles
Driven)
0.473
0.046
-0.263
-0.015
-0.035
0.018
-0.049
en
'Significant at the 95% level.
-------
TABLE 5.22. REGRESSION ANALYSIS BETWEEN CARBON MONOXIDE AND ODOMETER READINGS
FOR CALIFORNIA MEDIUM-DUTY GASOLINE-POWERED TRUCKS
Model- Year Sample
Group Size
pre-69
69-71
72
73-74
75-76
77
78-80
4
7
3
9
30
15
7
r2
0.66
0.13
0.43
0.04
0.05
0.05
0.04
Mean Square Mean Square
for Regression for Residual
1062
107
94
122
1103
144
202
.73
.25
.37
.01
.85
.18
.97
274.
148.
127.
431.
806.
201.
911.
06
58
09
29
30
33
10
F-value
Calculated
3.88
0.72
0.74
0.28
1.37
0.72
0.22
95% level
18.51
6.61
161
5.59
4.20
4.67
6.61
(DFrDF2)
(1,2)
(1,5)
(1,1)
(1,7)
(1,28)
(1,13)
(1,5)
Estimated
Estimated Slope (gm/mlle
Intercept per 1,000 miles
(gm/mlle) Driven)
-70.903
52.228
103.100
64.69
69.43
41.174
28.776
1 . 526
0.171
-0.550
0.232
-0.329
-0.189
-0.873
en
en
-------
TABLE 5.23, REGRESSION ANALYSIS BETWEEN NITROGEN OXIDES AND ODOMETER READINGS
FOR CALIFORNIA MEDIUM-DUTY GASOLINE-POWERED TRUCKS
F-value
Model-Year
Group
pre-69
69-71
72
73-74
75-76
77
78-80
Sample
Size
4
7
3
9
30
15
7
r2
0.17
0.006
0.19
0.0004
0.03
0.07
0.76
Mean Square
for Regression
0.99
0.15
0.04
0.02
3.95
2.99
1.22
Mean Square
for Residual
2.38
4.84
0.18
6.30
4.14
3.22
0.08
Calculated
0.42
0.03
0.23
0.003
0.95
0.93
14.97*
95% level
18.51
6.61
161
5.59
4.20
4.67
6.61
(DF, ,DF2)
(1,2)
0.5)
0.1)
0.7)
0,28)
0.13)
0,5)
Estimated
Intercept
(gm/mlle)
9.401
5.367
4.771
5.492
2.883
3.901
2.794
Estimated
Slope (gm/mlle
per 1,000 miles
Driven)
-0.047
0.006
-0.011
0.003
0.020
0.027
-0.068
Significant at the 95% level.
------- |