United States
Environmental Protection
Agency
NOT TO BETAKE
November
1990
Air
x°/EPA
Cooperative Testing
Program Draft Report
ECTD
TSS
DRAFT
-------
EPA-AA-TSS-I/M-90-X
**Draft**
v 1.5
Technical Report
Report on the EPA/Manufacturer Cooperative
I/M Testing Program
By
James A. McCargar
Lisa Mouat Snapp
September 1990
NOTICE
Technical Reports do not necessarily represent final EPA
decisions or positions. They are intended to present
technical analysis of issues using data which are currently
available. The purpose in the release of such reports is to
facilitate the exchange of technical information and to
inform the public of technical developments which may form
the basis for a final EPA decision, position or regulatory
action.
Technical Support Staff
Emission Control Technology Division
Office of Mobile Sources
Office of Air and Radiation
U.S. Environmental Protection Agency
-------
Section 1: Executive Summary 2
1.1 Program Summary 2
1. 2 Results 2
Section 2: Background and Program Summary 6
2 .1 Program Overview 6
2 . 2 Recruitment Quotas 7
2 . 3 Procurement 8
2 . 4 As-Received Testing Protocols 9
2.4.1 Introduction 9
2.4.2 Basic I/M Test Procedure 9
2.4.3 Applicability of the BITP Model 11
2.4.4 As-Received FTP Test 11
2.4.5 Tank Fuel Analysis 11
2 .5 As-Received Diagnosis 13
2.6 Selection of Vehicles for Remedial
Maintenance 13
2.7 Remedial Maintenance Protocols and Post-Repair
Testing 13
2 .8 Database Structures 15
2 . 9 Program Nonconformities 15
Section 3: Basic Vehicle Characteristics of the
CTP Fleet 17
3.1. Introduction and Overview 17
3.2. Profile by Manufacturer and Quota Group 17
3.3. Mileage Profile 19
3.4. Profile by Vehicle Type 22
3.5. Additional Comments on the Base Sample 23
Section 4: As-received Emissions Analysis 24
4.1. Introduction 24
4.2. FTP Results 24
4.2.1. Sample Description 24
4.2.2. Pass/Fail Results at Certification
Standards 24
4.2.3. Excess Emissions Analysis 25
4.2.4. Analysis Using MOBILE4 Emitter
Categories 29
4.2.5. Correlation of Excess Emissions,
Emitter Category, and Odometer 32
4.3. Michigan AET Test Results 34
4.3.1. Profile of the Base Sample 34
4.3.2. Stratifications by Manufacturer and
Quota Group 35
4.3.3. Stratification by Mileage and Vehicle
Type 37
4.4. Correlation Between the FTP and Michigan AET
Results 37
4.4.1. Introduction 37
4.4.2. Excess Emissions and Emitter
Categories of the AET Failure Types ...37
11
-------
4.5. Laboratory Short Test Results 41
4.5.1. Sample Description 41
4.5.2. Second-Chance Failure Rates of the As-
Received Base Sample 41
4.5.3. Idle-Mode Short Test Variability Due
to Fuels 45
4.5.4. Correlating the Laboratory Short Tests
with the MOBILE4 Emitter Categories ...46
4.5.5. Variability Between Adjacent Idle
Modes 49
4.6 Supplemental Analysis of the AST Errors of
Commission 53
Section 5: Emission Effects of Remedial
Maintenance 55
5 .1 Introduction and Sample Descriptions 55
5.2 Total Mass Emission Reductions from the CTP
Fleet 55
5.2.1 Net Benefit of Repairs FTP-Based 55
5.2.2 FTP versus LA4 values 58
5.2.3 Net Benefit of Repairs LA4-Based 62
5 .3 Overview of the Repairs Conducted 63
5.3.1 System and Subsystem Repair Categories ...63
5.3.2 Emission Benefits per System Repair 65
5.3.3 Emission Benefits per Subsystem Repair ...69
5.3.4 Total Benefit per Subsystem 73
5.3.5 Effect of Deteriorated Catalysts on
Emissions 75
5 . 4 Analysis of Incremental Repairs 77
5.4.1 Sample Description 77
5.4.2 Benefits of Repairing to Pass I/M 78
5.4.3 Comparison to MOBILE4 Repair Estimates ...81
5.4.4 Benefits of Repairing to pass the FTP ....83
5.4.5 Emission Benefits "Lost" through
Second-Chance 85
5.4.6 Benefits of Repairing to Different
Targets 86
5 . 5 Repairs to High Emitters 87
5.5.1 Effectiveness of Repair on Marginals
vs . Highs 87
5.5.2 Benefit of Repairing Highs only 89
5.5.3 Catalyst Repairs Performed on Highs 91
5 . 6 Difficulty of Repair 93
5.6.1 Difficulty of Repair to Passing Levels ...93
5.6.2 Effectiveness of Repair at Successive
RM Steps 97
Appendix A: Failure Rates in the Michigan AST
Program 103
Appendix B: Vehicle Identifying Information for
the CTP Base Sample 104
iii
-------
Appendix C: As Received Failure Rates for
Selected Modes of the Basic I/M Test
Procedure 105
Appendix 0: AST Errors of Commas ion in the CTP
Sample 106
Appendix E: Per-Repair Emission Reductions for
All Systems: by Quota Group 107
Appendix F: Per-Repair Emission Reductions for
Statistically Significant Systems:
by Quota Group 113
APPENDIX G: Per-Repair Emission Reductions for
All Subsystems 114
Appendix H: Per-Repair Emission Reductions for
Statistcally Significant Subsystems:
by Quota Group 116
Appendix I: Total Estimated Emission Reductions
for All Subsystems 121
IV
-------
SECTION 1: EXECUTIVE SUMMARY
1.1 Program Summary
The EPA/Manufacturer Cooperative Test Program (CTP)
recruited private-owner vehicles based on failure of the
Michigan Auto Exhaust Testing (AET) Program for testing at
laboratory facilities of the EPA and seven major motor
vehicle manufacturers. The program focused on closed-loop
light-duty vehicles and light-duty trucks from model years
1981-1986. The test protocols included the Federal Test
Procedure (FTP) and a new short test procedure with segments
simulating a variety of field test conditions. The remedial
maintenance procedure called for incremental repairs and
retests to meet both FTP and short test criteria.
Data from the program were analyzed with the following
objectives in mind:
(1) Developing advice to I/M programs on improvements to
preconditioning methods and formal I/M test
procedures.
(2) Seeking, and assessing the potential of, a limited
diagnosis and repair sequence as a remedy for a
significant portion of the in-use emissions excess.
(3) Improving methods and models for estimating I/M
effectiveness in reducing emissions.
(4) Providing feedback to the manufacturers on
particular malfunction or malmaintenance types.
1.2 Results
The following results from analysis of the CTP data are
significant :
(1) Eighty-six percent of the 239 vehicles in'' the CTP
base sample failed their HC or CO certification
standards in the as-received condition, with HC+CO
failures being the most prevalent. Of the 'failures,
one vehicle was a super emitter and 1B&. were high
emitters, by the MOBILE4 definitions. The worst 40%
of the , vehicles accounted for 90% of the fleet
excess HC and 83% of the fleet excess CO. [Sections
4 . 2 and;1 4.3]
'
(2) The mean excess HC and CO emissions of the MY1981-82
vehicle group exceeded the mean excess emissions for
-------
the MY1983-86 group by about half. Discounting the
impact of the one super emitter, the fuel-injected
and carbureted vehicles had roughly comparable
excess emissions, within the model year groups. The
percentage of high emitters varied considerably
across the manufacturers. [Section 4.2.4]
(3) Excess emissions were not well correlated with
mileage; however, successively higher MOBILE4
emitter categories showed successively higher mean
mileages. [Section 4.4.2]
(4) Of the 40% of the CTP fleet that were normal
emitters, almost 80% failed their recruitment short
test for only one pollutant; HC-only failures
outnumbered CO-only failures by two to one. The
data show no reasonable alternative outpoints for
excluding large numbers of normal emitters from AET
failure without inappropriately converting many high
emitters to AET passes. [Section 4.4.2]
(5) When tested in a fully-warmed condition with either
loaded or extended-2500rpm preconditioning, the
second-chance idle failure rate of the CTP fleet was
only about 40%. Up to an additional 20% of the
fleet failed second-chance idle tests under a
variety of conditions on warmed-up vehicles. Only
on vehicles that were idle tested immediately after
an extended soak did the second chance failure rate
exceed 60%. [Section 4.5.2]
(6) Vehicles that were combined HC+CO failures on their
AET tests were not also prone to be HC+CO failures
during second-chance testing. [Section 4.5.2]
(7) The second-chance idle tests at fully warmed
condition, with loaded or extended 2500rpm
preconditioning, reduced the error of commission
rate to zero or near zero, and passed 78% or more of
the marginal emitters as well. The failure rate of
the high emitters under the same second-chance test
conditions was 60%-65%. [Section 4.5.4]
(8) Correlation between the two idle modes of a two-mode
idle test was relatively high (R2 values of close to
90%) when the vehicles were fully warmed and
preconditioned with loaded or extended-2500rpm
operation. Poorer correlation (R2 values between
31% and 85%) was shown under less ideal test
conditions. For all test conditions, the effect of
the intervening 2500rpm mode was generally to reduce
the failure rate on the second idle.
[Section 4.5.5]
-------
(9) All but two of the 33 as-received error of
commission vehicles had elevated or failing HC
scores on the Michigan AET test; slightly over one-
third failed the AET for CO. One-quarter of the Ec
vehicles fell in one GM engine displacement. Only
one of the Ec vehicles had a repeatable short test
exceedance that was diagnosed and resolved through
repair. Most Ec' s were attributed to the response
of the engine and control system to vehicle
preconditioning. [Section 4.6]
(10) Repairs eliminated 99% of the excess HC and CO
emissions in the CTP fleet; one fifth of the overall
HC reduction and one third of the overall CO
reduction were due to oxygen sensor replacement.
Catalyst and fuel injector replacements also each
contributed more than 10% to the total HC reduction,
while no other repair type contributed more than 10%
to the CO reduction. The average vehicle had 1.8
g/mi HC and 28 g/mi CO eliminated in 2.6 repair
steps. [Sections 5.2.1 and 5.3.4]
(11) The most frequent repair types were oxygen sensor
replacements (45% of repaired vehicles), catalyst
replacements (30%), and ignition tune-ups (spark
plug/wire replacement, initial timing adjustment,
idle speed adjustment) (29%). [Appendix G]
(12) The most effective repair type on a per-vehicle
basis was fuel injector replacement. This
eliminated an average 2.4 g/mi HC and 24 g/mi CO.
Other statistically significant effective HC repairs
were to the catalyst (1.1 g/mi), carburetor
(1.0 g/mi), oxygen sensor (0.8 g/mi) and fuel meter
tune-ups (0.6 g/mi). For CO, the significant
effective repairs were to the load sensor (23 g/mi),
oxygen sensor (21 g/mi), carburetor (12 g/mi), fuel
metering system tune-ups (12 g/mi), ECU (12 g/mi),
and catalyst (7 g/mi). The average reduction per
repair step for all repair types was 0.7 g/mi HC and
11 g/mi CO. [Section 5.3.3]
(13) Vehicles required an average of 1.5 repairs to
switch from high emitter to marginal or passing
emitter status on the FTP, with emitter categories
as defined by MOBILE4. The same number of steps was
needed to change from failing to passing on an idle
I/M test performed under ideal conditions. [Section
5.6.1]
(14) Marginal emitters (as defined by MOBILE4) were not
worthy repair targets in the CTP. Their emission
-------
reductions were negligible, with high emitters
achieving reductions 15 times as large. [Section
5.5.1]
(15) Repairs targeted at the most likely malfunctioning
component resulted in the early repairs being more
effective than later repairs; the first repair on a
vehicle achieved an average reduction five times
that of the third repair. [Section 5.6.1]
(16) Second-chance I/M tests apparently reduced the
incidence of unnecessary repairs to normal emitters
without greatly reducing the benefit due to
repairing high emitters. Sixty percent of the fleet
passed a second-chance I/M test performed under
ideal conditions; this 60% would have achieved less
than 25% of the total fleet repair reduction.
[Section 5.4.5]
(17) Repair types that were consistently effective at
reducing FTP emissions were also the most reliable
at correcting I/M-failed vehicles to passes. At the
system level, these were the exhaust, fuel metering,
and 3-way control systems. Subsystems included the
catalyst, carburetor, oxygen sensor, fuel metering
system tune-ups, and fuel injectors. [Sections
5.3.2 and 5.3.3]
(18) Repairing until vehicles passed I/M reduced emission
levels by approximately 75% and eliminated over half
of the FTP excess. However, this was only about two
thirds of the reduction that could be realized with
more complete repair. [Section 5.4.6]
(19) Replacement of the catalytic converter resulted in
average g/mi reductions twice that of all other
repair types for HC and 3/4 that of all other repair
types for CO, even though catalyst repairs were
generally withheld until the last repair. Evidence
of tampering or misfueling occurred on only 20% of
the vehicles that received catalyst repairs.
[Section 5.3.5]
(20) Estimates of average emission reductions due to
repair to I/M passing status are lower in the
MOBILE4 model than those seen in the CTP,
particularly for HC on fuel injected vehicles.
[Section 5.4.3}
(21) Ninety-four percent of vehicles that were repaired
from high to normal emitter levels on a transient
test could also, at that repair stage, pass an I/M
test performed under optimum conditions (57% from
-------
I/M fail to pass, 37% already passing) [Section
5.4.4]
-------
SECTION 2: BACKGROUND AND PROGRAM SUMMARY
2.1 Program Overview
The Cooperative EPA/Manufacturer I/M Test Program (CTP)
was a joint effort by the Environmental Protection Agency and
seven of the major domestic and imported vehicle
manufacturers to recruit failed vehicles from an official
state Inspection/Maintenance (I/M) program for study in a
laboratory environment. The intent of the program was to
gather data that could be combined with the results from a
number of other studies, contributing to accomplishing
several objectives. Among these were the following:1
Develop advice to I/M programs on improvements to
preconditioning methods and formal I/M test
procedures.
Seek, and assess the potential of, a limited diagnosis
and repair sequence as a remedy for a significant
portion of the in-use emissions excess.
Improve methods and models for estimating I/M
effectiveness in reducing emissions.
Provide feedback to the manufacturers on particular
malfunction or malmaintenance types.
The study focused on 1981-86 model-year vehicles with
closed-loop engine control that failed the idle test under
Michigan's decentralized Auto Exhaust Testing (AET) program.
The EPA solicited owners just prior to their expected AET
test date by direct mail for participation in the program and
also coordinated initial recruitment efforts. Each of the
participating manufacturers Chrysler, Ford, General
Motors, Honda, Nissan, Mitsubishi, and Toyota completed
the recruitment process and performed the laboratory work on
its own vehicles, at its own facilities.
Vehicles of some additional manufacturers, which had no
appropriate testing facilities in Southeast Michigan
available during the program, were recruited to the EPA Motor
Vehicle Emissions Laboratory in Ann Arbor, Michigan for
testing and repair. This group included vehicles from
American Motors, Mazda, Subaru, and Volkswagen. Facilities
for Toyota were under construction at the outset of the
program, and testing of the first ten Toyota vehicles
consequently took place at MVEL; work on the remaining six
Toyota vehicles was completed by the manufacturer.
-------
According to the CTP program plan, recruited vehicles
first underwent a series of emissions tests, including short
cycles and the Federal Test Procedure (FTP) , in the as-
received condition. Each vehicle was diagnosed for causes of
any excess emissions; it was then repaired, and retested. At
least two features distinguished the program protocols from
those in previous in-use studies conducted by EPA and others.
First, the remedial maintenance and retest steps were
incremental; that is, mass-emission and raw-gas tests were
executed following each "significant" repair. Second, the
program's short cycles incorporated a new Basic I/M Test
Procedure (BITP) that was designed to simulate vehicle
response under a variety of field I/M test conditions.
Recruitment for the Cooperative Test Program began in
February 1987, and testing continued through May 1988. Data
from the participants was collected, quality checked and
organized in a common database through the EPA facility in
Ann Arbor. Final data submissions and major corrections to
the databases were completed in May 1989.
The remainder of this section summarizes portions of the
CTP program plan that will aid the reader in understanding
the analysis to follow.
2.2 Recruitment Quotas
The Cooperative Test Program vehicles were not a random
sample from among all those that failed the Michigan Auto
Exhaust Test. Quotas on recruitment were established
according to a number of variables: manufacturer, model-year
group, fuel-metering strategy, presence of tampering or
misfueling, and presence of a pattern failure.
The manufacturer quotas distributed the testing
obligation based on the relative percentages of each
manufacturer's fleet in a 140,000-vehicle sample of failures
from the Seattle I/M program. This program was selected
because of the availability of detailed data, its procedural
similarity to the Michigan program, and its use of a
keyoff/restart step for Ford vehicles.2
Regardless of its share in the Seattle failures, each
participant agreed to test a minimum of ten vehicles; this
minimum was applied to Mitsubishi and Honda. General Motors,
Ford, and EPA each entered the program with testing targets
of 60 vehicles. Nissan was allocated 30 vehicles, Toyota 16,
and Chrysler, 15. Thus the program plan target was 261
vehicles.
Only closed-loop vehicles from the 1981 through 1986
model years were recruited. By definition, manufacturers
producing no closed-loop vehicles in the earlier model years
-------
necessarily met their recruitment goals from the later years.
In order to further focus on technologies likely to dominate
I/M fleets in the 1990s, carbureted vehicles were targeted to
be less than half of each organization's basic quota, and at
the manufacturer's option, post-1983MY carbureted vehicles
could be excluded from recruitment.
For manufacturers with closed-loop production across the
model years, some additional considerations applied: half of
the quota was to be filled by vehicles from the 1981 or 1982
model years, the other half from the 1983 or later model
years. Some additional considerations applied, in order to
ensure a sample with current technology, as well as age-
related malperformances.
In order to preclude domination of the sample by certain
important forms of tampering, the number of vehicles with
fuel inlet tampering, Plumbtesmo test failure, or catalyst
removal was limited to the greater of one vehicle or ten
percent of a manufacturer's overall CTP quota. Finally, the
participants were also expected to limit recruitment of so-
called "pattern failures" from the sample, under the
principle that limited information of value would be obtained
if a manufacturer's sample were heavily biased towards one
(or a few) vehicle groups once a pattern of problems had been
adequately diagnosed and remedied.
Because the primary recruitment quotas were based upon
model year and fuel metering strategy, four "quota groups"
were defined for each manufacturer: fuel-injected MY1981-82,
carbureted 1981-82MY, fuel-injected MY1983-I-, and carbureted
MY1983+. The quota groups will be referred to frequently in
the remainder of this report.
2.3 Procurement
Each week over the course of the program, EPA culled a
list of potential owners from a tape provided by the Michigan
Department of State, containing registration data for owners
of vehicles due to receive notices of their AET test
requirement in that week. Decoders based on the Vehicle
Identification Number (VIN) were employed to screen out
ineligible vehicles. A mailing label was generated for
eligible owners, and a direct-mail solicitation was
conducted. The solicitation letter contained boilerplate
language on the program and incentives for the owner to
participate, as well as manufacturer-specific information.
The letter emphasized that only those owners who failed their
subsequent AET test would be eligible, and it provided
instructions for interested owners to contact the appropriate
participating CTP test site or its representative.
-------
The size of a week's mailing was determined by each
site's throughput and progress towards meeting the
recruitment quotas. Where more eligible owners existed than
CTP testing capacity for the likely respondents, the pool was
reduced by randomly selecting owners from all those in the
same quota group.
Once an owner contacted a CTP recruitment site, a
telephone screening was conducted to verify eligibility.
This screening verified the Michigan AET test failure, the
timeliness of the owner response, the absence of any owner
action to remedy the failure, and set of safety and outlier
rejection criteria.3 Conformance to the any late-breaking
changes in the manufacturer's progress towards recruitment
quotas was also taken into account. Owners not excluded
during the telephone screening were scheduled for intake to
the appropriate test site. A final safety and outlier
screening was performed at the time of vehicle intake,
including a road test to project safe dynamometer operation.
2 . 4 As-Received Testinqr Protocols
2.4.1 Introduction
The CTP program plan called for each vehicle to receive
four short-cycle sequences, collectively referred to as the
Basic I/M Testing Procedure (BITP), on tank fuel (Table I).4
The BITP was then followed by the Federal Test Procedure
(FTP) on Indolene. If the vehicle was recruited with
insufficient fuel to complete BITP testing, the program plan
called for substitution of a commercial fuel. An RVP and
lead-in-fuel analysis was performed on the tank fuel.
TABLE 1
AsReceived Emissions Testing Outline
Procedure
BITP
FTP
Fuel
Tank/Commercial
Indolene
Sequence
Cold Start
Extended Loaded
Extended Idle
Restart
N/A
Prior Base Operation
75° Soak, 1 hr minimum
LA4
continuous 20-min idle
LA4 + Restart
LA4 Prep, overnight 75° Soak,
No heat build
2.4.2 Basic I/M Test Procedure
The Basic I/M Test Procedure attempted to replicate
unloaded field I/M tests under a variety of controlled
-------
conditions, and thereby provide possible explanations for the
Michigan AET results that were the basis for CTP vehicle
recruitment. These controlled conditions included the prior
("base") operation of the vehicle and any conditioning that
occurred immediately prior to the pass/fail test modes.
Examination of other factors that might have affected the
Michigan AET results, such as AET analyzer calibration
variables or operator fraud, was not in the scope of the
program.
The four sequences that make up the BITP corresponded to
(and were named after) four different types of preceding, or
"base" operation: a cold start, extended operation under
load, extended operation at idle, and execution of an engine
keyoff/restart. The base operation was then followed by one
or more simulated Two-Speed Idle tests, each with a
controlled conditioning mode. Raw-gas emission values and
engine RPM measurements were gathered throughout the BITP,
but attention was focused on seven "core" two-speed tests,
spread through the procedure. No loaded short testing
(either transient or steady-state) was included in the BITP.
Table 2 summarizes the modes of the Basic I/M Test
Procedure. The core sampling periods, which will serve as
the basis for much of the short-test analysis that is to
follow, are shaded. Certain parameters HC, CO, C02, and
engine RPM, were sampled throughout the procedure. During
the core sampling periods, these parameters were measured at
15 and 30 seconds into each mode, and at 60, 90 and 120
seconds of the second idle-neutral mode as well. Coolant
temperature was monitored during the entire cold start
sequence, and at other points (such as the extended idles),
where some significant variation might occur.
Sampling intervals outside of the core sampling period
varied according to the purpose of the mode. For a more
detailed description of the sampling procedure, refer to the
program plan.5
The selection, order, and duration of the modes reflects
a desire to test the vehicles under both ideal and non-ideal
conditions. The purpose of the Cold Start sequence, for
example, is to characterize the emissions of each vehicle at
abnormal (low) operating temperature, and then to determine
the effectiveness of various conditioning modes at achieving
normal operating temperature, and the emissions impacts of
such conditioning
The Extended Loaded sequence was the hypothetical
"ideal" test condition in the procedure; its purpose was to
measure vehicle emissions immediately following an extended
period of loaded operation. The Extended Idle sequence
presumes that the vehicle has prior loaded operation that had
10
-------
achieved normal operating condition in the recent past, but
that a period of extended idle may have caused the vehicle's
condition (and therefore its emissions) to deviate from its
ideal levels.
Finally/ the Restart sequence examines the non-ideal
aspects of a keyoff/restart on vehicle emissions. For most
vehicles, the ideal condition was presumed to not involve a
restart, and to isolate the effects of the restart from other
conditioning variables, these vehicles were not restarted
during the Cold Start, Extended Loaded, and Extended Idle
sequences, but were restarted during the Restart sequence.
On the other hand, Ford vehicles from the 1981 model year
onward were generally designed with the assumption that a
keyoff/restart would precede any idle short test.6 Use of the
restart for Fords in the CTP was therefore the opposite of
the case for non-Fords (see Table 2).
2.4.3 Applicability of the BITP Model
It is important to note that in the Cooperative Test
Program the BITP simulates field test conditions but not a
field sample. The program recruited initial idle-test
failures only; no AET-passing vehicles underwent the BITP or
other procedures from the CTP protocol.
The seven core sampling periods in the BITP thus
represent second-chance I/M tests on failed vehicles.
Without the analogous data on the passing vehicles, care must
be taken when interpreting failure rates, excess emissions
identified, variability results, and other analyses on the
BITP data. This caution applies as well when examining parts
of the BITP that had no direct analog in the Michigan AET
test, such as 2500rpm pass/fail results.
2.4.4 As-Received FTP Test
The FTP used in the Cooperative Test Program was a
standard three-bag cycle performed on Indolene (Table 1,
above) . The CTP version of the test omitted the heat build
and the Highway Fuel Economy Test.
2.4.5 Tank Fuel Analysis
Lead-in-fuel analysis was performed using x-ray
flourescence and targeted at designating the fuel as either
above or below a 0.05 g/gal standard. Reid Vapor Pressure
testing was conducted using the ASTM D323 protocol.
11
-------
TABLE 2
Modes of the Basic T/M Test
SEQUENCE
MODE*
MODE NAME
DURATION
FUNCTION
Cold Start (CS)
0 1
0 2
03
04
0$
0 6
0:7
75° Soak
Engine Start
2500rpm
le^neutra
2500rpm
>60 min
n/a
30 $0c.l
30 sec *
120 $e$
180 sec
K&yoff/Restart
1 1
Base Operation
Base Operation
Core Sampling
Core Sampling
Core Sampling
Conditioning
Core Sampling
VeWcfe* Only
Extended Loaded (XL)
01
oa
-frSL
LA 4
1372 sec
Keyolf/Resfart
Base Operation
A y"VF.«Vrtvw* " *^%i^iW!fiaa?ss'A^v
&** tomtfte&fr
Ford Vehicles Only
^
120
sec
Extended Idle (XI)
Idle-neutral
20 min
Base Operation
SampHrig
*... .: -- H «3f V
«wett«
Conditioning
Keyoff/Restaft
Restart (RS)
01
1372 sec
Idle-neutral
120 860
Base Operation
y^ ^SvSty g^sy^o-'VA^y'^ *'3SftSKsftM:
. . Core Sampling!
on-Ford Vehlcfes
/-Core Sampling
Sampling
12
-------
2.5 AsReceived Diagnosis
The last step in characterizing the as-received
condition of the CTP vehicles was a diagnosis of the engine
and emission control systems. The primary purpose of this
inspection was to identify system and component
malperformances that might explain FTP or short-test
noncompliances of the vehicles. An established protocol was
then followed for performing remedial maintenance and retests
to measure the impacts of the repairs.
In order to permit organization of the data from the
eight participating sites in a common database, a uniform
format was devised for reporting the results of the
engine/emissions system diagnosis. The format was based
largely on the ECOMP file employed by the EPA Emission
Factors testing program, the largest mainframe EPA database
on in-use vehicles. Both the CTP and ECOMP formats divide
the engine and emissions components of the vehicle into
systems and subsystems, and then code the presence and type
of malperformance detected during the vehicle inspection.
The coding is supplemented by narrative comments of the
technicians. In the CTP, an additional data recording system
was developed for coding the repair actions that were taken
on the basis of the ECOMP diagnosis (see Section 5.3.1) .
2.6 Selection of Vehicles for Remedial Maintenance
Before being released to its owner, each CTP vehicle was
required to satisfy criteria in three categories: FTP
performance, variability of the short test scores, and short
test performance relative to the 207(b) emission standards.
Table 3 summarizes these criteria.7 Information from the as-
received characterization (test results and diagnostic data)
were used to identify all vehicles that would require
remedial maintenance steps in order to meet the criteria.
2.7 Remedial Maintenance Protocols and PostRepair Testing
The remedial maintenance philosophy in the CTP was (to
the extent feasible) to measure the emissions impacts of
individual repairs. Repairs and follow-up testing were
therefore executed in steps, with the minimum number of
repairs conducted at each step that would be expected to
generate significant emissions impacts. The priority order
of repairs was determined by the exit criteria remaining to
be satisfied at that step targeting FTP compliance first,
I/M variability second, and basic I/M conformance third. If
multiple repair options were available that could reasonably
satisfy the highest priority criterion, the repair with the
13
-------
biggest projected impact was conducted first (followed by the
appropriate test sequence, to quantify the impact of the
repair.8
TABLE 3
Allowable Exit Criteria in the Remedial
Maintenance Phase
CATEGORY
FTP
I/M
Variability
I/M Basic
MILEAGE
<50K
>50K
N/A
N/A
N/A
N/A
N/A
N/A
N/A
ALLOWABLE EXIT CRITERIA
HC & CO <1.5 * cert standard AND
HC & CO <2.0 * cert standard
Analogous sampling points between
sequences show comparable values AND
successive sampling points within a
test mode show comparalbe values; OR
Observed variability traced to
unrepairable element of design OR
Variability cannot be repeatably
triggered OR
All reasonable repair efforts
completed
Core sampling emissions in extended
loaded sequence pass 207 (b) OR
Observed 207 (b) failure traced to
unrepairable element of design OR
All reasonable repair efforts
completed
The one exception to the above guideline was catalyst
replacement. Dramatic emissions improvements would normally
be expected with installation of any "green" catalyst, even
if performance of the original catalyst was acceptable. Such
replacements could mask the importance of other important
malfunctions in the engine or other emission control devices.
On-vehicle diagnosis was also anticipated to be difficult in
some cases, except where overt signs of damage to the
container or the the substrate existed. Thus, catalyst
replacement was treated as the repair of last resort.
As a guide for decision-making when faced with multiple
repair options, the CTP program plan established a guideline
for the priority order of repair, as follows:
1. Computer control and feedback system repairs,
including most repairs indicated by onboard diagnostic
systems and repairs to electronic fuel metering
components;
14
-------
2. Primary emission controls other than those in the
feedback system, including exhaust aftertreatment,
secondary air, PCV, and EGR systems;
3. Idle mixture adjustment, on vehicles with missing
limiter devices.
4. Other basic engine components.
In general, post-repair testing was conducted after each
repair step in the CTP. The test procedures employed were
all those necessary to track the vehicle's performance
relative to each of the outstanding failed exit criteria.
Thus, a repair targeted at an FTP failure on a vehicle with
remaining short test noncompliance would be followed by I/M
testing as well as FTP testing.
As a cost-saving measure, both the FTP and short test
protocols could be shortened during post-repair testing.
Test sites could perform an LA4 cycle in place of the FTP,
until a significant improvement in emissions was exhibited
between the post-repair LA4 and the weighted bag two and
three results from the previous FTP. At that point, however,
a full FTP was to be conducted before proceeding with further
repairs. For short testing, abbreviated versions of the BITP
were available that consisted of the extended loaded
sequence, plus remaining sequences showing violations of the
variability or basic I/M criteria.
2 . 8 Database Structures
Data from all of the important aspects of the program
described in the above sections were recorded in a relational
database (MICRO) on the Michigan Terminal System (MTS) and
were also downloaded to microcomputers for analysis. The
file structure of the CTP MICRO database mimics that of the
EPA Emissions Factors database, with additional datasets and
fields for the information that is unique to the CTP. A new
repair database format was constructed to categorize repair
actions and facilitate analysis. CTP program participants
were afforded access to these data through MTS accounts and
through copies of the subset microcomputer databases.
2.9 Program Nonconformities
By and large, the participating organizations in the CTP
operated independently, under the guidelines of the CTP
program plan. There was, for example, no real-time
coordination between the participants during vehicle testing
or when short turnaround decisions on repair protocols needed
to be made. Thus, each testing organization exercised its
15
-------
own judgment in unusual or borderline cases. Some
nonconformity in the data results, which complicated some of
the analytical tasks. Relevant and significant examples will
be identified in the sections that follow.
16
-------
SECTION 3: BASIC VEHICLE CHARACTERISTICS OF THE CTP FLEET
3.1.
Introduction and Overview
The participating organizations recruited 245 vehicles
for study in the Cooperative Test Program. A detailed
breakdown of the vehicle identifying information for all of
these vehicles appears in Appendix B.
Malfunctions in two of the 245 vehicles prevented
gathering sufficient as-received or post-repair data to
justify inclusion in the base analytical sample.9 Four
additional vehicles received no initial FTP, and had no post-
repair FTP that could be reasonably substituted for the
missing as-received test.10 These four vehicles are only
included in a limited number of analyses in the sections that
follow. The bulk of this report focuses on the remaining 239
vehicles, hereafter referred to as the base CTP sample.
3.2. Profile by Manufacturer and Quota Group
By and large, the actual CTP vehicle sample met the
intent of the program plan targets described in Section 2.2.
Table 4 and Figure 1 show the breakdown of the 239-vehicle
sample by manufacturer and quota group.11 As anticipated in
TABLE 4
CTP Base Sample bv Manufacturer Share
Manufacturer
General Motors
Ford
Nissan
Toyota'
Chrysler
American Motors**
Volkswagen**
Mazda"
Subaru**
Mitsubishi
Honda
TOTAL
Vehicles
58
57
20
16
15
15
14
12
12
10
10
239
Sample %
24.3
23.8
8.4
6.7
6.3
6.3
5.9
5.0
5.0
4.2
4.2
100.0
Testing split between EPA and manufacturer
** Testing performed by EPA
17
-------
the project plan, General Motors, Ford, Nissan, Chrysler,
Mitsubishi, and Honda tested their own vehicles. Testing of
Toyota vehicles was divided between EPA and the manufacturer.
Vehicles of the remaining manufacturers were tested by EPA at
MVEL.
In general, the CTP recruitment reflects patterns one
would have expected from the actual vehicle fleet. Ford and
General Motors, which together comprise half of the CTP
sample, are each dominated in the earlier model years by
carbureted vehicles, and in the later model years, by fuel
injection. Some manufacturers produced no closed-loop
vehicles in a given quota group, and thus none were present
in the program. Examples include fuel-injected 1981-82 Mazda
and Subaru. In other cases, vehicles in the quota group were
produced, but none was recruited in the CTP. Examples here
were the fuel-injected MY1981-1982 Toyotas, and carbureted
MY1983-1986 Subarus.
FJLGURE 1
Fleet Profile bv Quota Group and Manufacturer
GM TOR) NISS TOYT CHRY AMC VW MAZ SUBA MITS HCND
Fl 83-86
Garb 83-86 D Garb 81-82 Fl 81-82
18
-------
Table 5 shows the 239-vehicle fleet distribution by
quota group. The fleet comes quite close to the planned
target of 50% each for fuel injection and carburetion.
Manufacturers who have unusual representation in a quota
group are Volkswagen, with one-quarter of the fuel-injected
1981-82 vehicles, and Subaru and Toyota, whose total of ten
carbureted 1981-82 vehicles were the only contributions by
Japanese manufacturers to that quota group. Interestingly,
three of Volkswagen's fourteen vehicles were carbureted.
Although no specific recruitment targets existed for
individual model years, the sample was well-distributed
across the 1981-86 range. Of the six model years
represented, none comprised greater than 23% of the sample,
and none was less than 11%.
TABLE 5
Fleet Breakdown bv Quota Groun
Quota Group
Fuel-Injected 1981-82
Carbureted 1981-82
Carbureted 1983-86
Fuel-Injected 1983-86
TOTALS
Vehicles
23
66
50
100
239
Fleet %
9.6
27.6
20.9
41.8
100.0
3.3. Mileage Profile
The CTP sample was also well-distributed by mileage
(Figures 2-3), again without an explicit recruitment
criterion. The median mileage was close to 50,000 miles.
Not surprisingly, few (in fact only 15) of the under-50K
vehicles were in the 1981-82 quota groups. Conversely, only
14 vehicles exceeded 100,000 miles, with only one from the
1983-86 model years.
The, fleets of six of the eleven CTP manufacturers were
composed entirely of vehicles with accumulated mileage under
100,000 (Figure 4). Of these, Volkswagen is notable for
having close to 80% of its vehicles in the range 50,000 to
100,000 miles. Subaru and Toyota both had significant
numbers of very high mileage vehicles in their samples, with
25% and 18%, respectively, over 100,000 miles. Toyota
weighed in with the highest mileage vehicle, a 1982 Toyota
Corolla with 234,000 miles.
19
-------
FIGURE 2
Mileaaft Profile hy Qun1-a flrrmp
2 3 4 5 6 7 8 9 10 11 >12
Lower Bound of 10K Mileage Interval
Fl 83-86
Carb 83-86 DCarb 81-82 Fl 81-82
FIGURE 3
Cumulative Mileage Profile bv Quota
m Total
-Fl 81-82
0-Carb 81-82
Garb 83-86
0-FI 83-86
<20 <40 <60 <80 <100 <120
Mileage Category
(1000's)
20
-------
FIGURE 4
Mileage Distrihut* i nn bv Manufacturer
100
80
60
40
20
0
IliM
i i i i i i i i i i i
l*:'v TO::^B;;;:;:-:-B :^ ::' i^;Bi^:^B : -IP$HllliPH$MH-:i;::;:!;B.
:..;.. &N₯X«^* ::;-:::: ^m-::::.::: ^&&&$ ^«xW;:>S ^--.; .-:- ^«£&&& ^BSSSXsvJ ^<&x&& - ^H;
|^^HK^H^H^^HiMl|iiMHiM^Hi£MnKillsi^Hi^nv
GM PQFD NISS TOYT CHRY AMC VW MAZ SUBA MITS HCMD ALL
<100K Ll<50K
Figure 5 shows the mean mileages by manufacturer for the
base CTP fleet. The "total" entries represent mean odometer
miles; the "annual" entries are estimates of annual mileage
accumulation, derived for each vehicle by subtracting the
model year from 1987, dividing by the odometer miles, and
then averaging across the given manufacturer.
FIGURE 5
Mean Mileage Accumulation bv Manufacturer
M
i
I
e
s
(K)
POFD NISS TOYT CHRY AMC VW MAZD SUBA MITS HCtO ALL
I Total D Annual
The fleet mean odometer was 55,000 miles, and the
calculated mean annual mileage accumulation was just under
21
-------
17,000 miles. The high-mileage Toyota helped drive that
manufacturer's mean odometer reading over 70,000 miles,
highest among the participants; Mitsubishi (with no vehicles
in the 1981-82 category) yielded the lowest mean odometer
reading, 40,000 miles. Honda vehicles displayed the highest
annual mileage accumulation (24,400 miles), almost twice that
of AMC, whose 12,300-mile annual pace was the lowest among
the participants.
3.4. Profile by Vehicle Type
Overall, eight percent of the 239-vehicle CTP sample
were light-duty trucks (Table 6). Two participating
manufacturers, General Motors and Chrysler, elected not to
recruit LDTs. Two others, Mazda and Honda, produced no
closed-loop LDTs in the model years covered by the program,
and thus had none in their CTP samples. Of the remaining
manufacturers, only American Motors had an LDT percentage of
greater than 25% (from the Jeep line). Most of the trucks in
the CTP sample fell in the later model years, probably
reflecting the slower penetration of closed-loop technology
in trucks relative to that in LDVs.
TABLE 6
Fleet Breakdown bv Vehicle Type
Category
LOV
LDT
LDT %
General Motors
Ford
Nissan
Toyota
Chrysler
American Motors
Volkswagen
Mazda
Subaru
Mitsubishi
Honda
58
48
19
15
15
1 1
13
12
1 1
8
10
0
9
1
1
0
4
1
0
1
2
0
0.0
15.8
5.0
6.3
0.0
26.7
7.1
0.0
8.3
20.0
0.0
Fuel-Injected 1981-82
Carbureted 1981-82
Carbureted 1983-86
Fuel-Injected 1983-86
23
64
40
93
0
2
10
7
0.0
3.0
20.0
7.0
SAMPLE
220
19
7.9
22
-------
3.5. Additional Comments on the Base Sample
Some manufacturers showed unusual concentrations of
selected engine displacements in their samples. For example,
fourteen of the fifteen Chrysler vehicles (or 93%) were
135CID engines. In the national fleet, this displacement
accounted for between 64% and 74% of Chrysler's production of
closed-loop vehicles in model years 1981-86. Close to three-
fifths of the 57 Ford CTP vehicles were 140CID, a
displacement that was only 19% to 28% of Ford's closed-loop
production in the model years covered by the CTP. Finally,
thirteen of the 30 fuel-injected GM vehicles (or 43%) were
151CID, compared to between 10% and 20% of GM's closed-loop
production in the given model years.
Higher I/M failure rates are one possible cause, but not
the only one, for the greater representation of these
families; AET failure rates by manufacturer and displacement
are not available from Michigan.
23
-------
SECTION 4: AS-RECEIVED EMISSIONS ANALYSIS
4.1. Introduction
This section analyzes the emission results from testing
of the CTP fleet in the as-received condition. The FTP
results are analyzed first, including an analysis based on
the "excess emissions" concept and emitter categories
employed in EPA's MOBILE4 computer model. The short test
results from both the Michigan AET test and the CTP's own
short tests are analyzed and compared to the FTP results.
4.2. FTP Results
4.2.1. Sample Description
All but two of the 239 vehicles in the base CTP sample
underwent an FTP on Indolene in the as-received condition.
For the two exceptions (vehicles 257 and 337), post-repair
FTP results were substituted for the missing as-received
tests, based on the conclusion that the repairs performed
were likely to have had negligible impact on each vehicle's
emissions. (As discussed in Section 3.1, such substitutions
were unavailable or unjustified for six additional vehicles,
accounting for the difference between the 245 vehicles
recruited for the program, and the 239 vehicles used for the
bulk of this analysis.)
4.2.2. Pass/Fail Results at Certification Standards
Of the 239 vehicles in the base sample, 206 (or 86%)
exceeded their certification standards for HC, CO, or both.
The remaining 33 vehicles thus represent errors of commission
(Ec) by the original AET short test. As Figure 6 shows,
combined HC-CO FTP failures dominated the CTP sample. Only
in the later model years were there significant numbers of
CO-only failures, and few HC-only failures occurred,
regardless of model year.12
24
-------
FIGURE 6
As-Received FTP Failure Type hv Quota Group
F\ 81-82 GARB 81-82 GARB 83-86 Fl 83-86
ALL
NO FAIL DHC-ONLY HHC&CO COONLY
4.2.3. Excess Emissions Analysis
A more illuminating view of FTP emissions in the CTP
sample arises from analyzing "excess" emissions, which is the
difference between each vehicle's FTP performance and its
emission standard, for a given pollutant. Figures 7-10
scatter plot the excess HC and CO emissions of the base
sample, stratified by quota group. Identical scales have
been used in the figures to ease comparisons between the
groups; note, however, that the resolution does not permit
display of the individual vehicles concentrated very close to
the origin.
The total HC excess emissions in the 239-vehicle base
sample was 334 g, or 1.4 g per average vehicle. The total CO
excess was 5155 g, or 21.6 g average excess. As the
scatter plots show, however, the fleet displayed a broad
spread of excess emission values for both HC and CO. The
total excess values include 57 vehicles that passed their HC
standard and 41 that passed their CO standard, and thus
contributed negative "excess" emissions. Cleanest relative
to its certification standards for both HC and CO was vehicle
249, a carbureted 1982 Ford LOT, which was 0.8 g below its
1.7 g HC standard and 12.9 g below its 18 g CO standard.
25
-------
FIGURE 7
AsReceived FTP Excess Emissions
Fuel-Tn-ipr!l-p»H 1981-82 Vehicles
00
(g/mi)
140
120
100
80
60
40
20
0
-20
.....
....
....
1
f
....
B
.....
.....
.
.
i
i
i
i
i
(18.7, 55
i
i
i
i
i
i
i
i
i
.6)
-».
- 1
3456
HC (g/mi)
8
10
FIGURE 8
As-Received FTP Excess Emissions:
Carbureted 1981-82 Vehicles
00
(g/ml)
1 4fi -
120 -
1 c\ n _
1 UU
80
6rt
u
40
2 n
u
0 -
.... J
.....
.....
4
.....
a.
i i
i
....1....C...J
1 1
1 1
1 1
1 1
1 1
1 1
T
1 ! '
i ' i
....;.&..,
_ i i
...VrfvL.-.,
-.-A,-? r
!!»* T~"T~"1
1 1
1 1 1
....
.-...
1 r-
....
.....
.....
.....
.....
.....
....
....
....
....
.....
1
1
L... J-...
1
1
1
1
1
.........
1
1
-! -
1
.....
.....
3456
HC (g/mi)
1 0
26
-------
FIGURE 9
As-Received FTP Excess Emissions:
Carbureted 1983-86
CO
(g/mi)
1 40 -.
120-
1 n n -
I U U
8n .
u
6n .
u
A n ,
2n .
0 -
.....
i
... ;
i
' V
i
i
.'
.
..L
.....
.....
r----i
....
....
1 1
i
L-...
L. ...
r----
1
i
t
.....
r----i
....
....
-1
3456
HC (g/mi)
10
FIGURE 10
As-Received FTP Excess Emissions:
Fuel-Injected 1983-86 Vehicles
140 -r
120 -
100 -
CO
(g/mi)
-20
.....
'
1
....i....
1
1
1
1
1
{-
" "*!"""
i
L... J
(.----
";-
-----
....
.. ..
r
u
....
.....
.....
.....
.....
.....
----1
....
....
....
....
-----
....
.% ..
....
.....
.....
....
-1
3456
HC (g/mi)
10
27
-------
On the opposite extreme was vehicle 202, another 1982
Ford, which was a sizable 18.7 g above its 0.41 g HC
standard. (This fuel-injected LDV was also the only super
emitter in the as-received CTP sample; see section x.x.x
below) . On its own, #202 accounted for almost 6% of the
total HC excess emissions in the 239-vehicle sample. Highest
in the ranking of excess CO emissions was vehicle 605, a
fuel-injected 1982 Nissan LDV, which was 139 g above its 3.4
g standard, or close to 3% of the total excess CO from the
sample.
When the CTP vehicles are ranked by excess FTP
emissions, and the excess is accumulated as a percent of the
total excess for a given pollutant, Figure 11 results. The
curves for HC and CO were determined independently. For each
pollutant, the dirtiest 40% of the vehicles accounts for 90%
of the excess from the sample. The leveling of the curve
above 100% and eventual decline to 100% reflects the portion
of the ranked sample that was very close to, and then below,
the certification standards.
FIGURE 11
Excess Emissions in the Ranked AsReceived Fleet
120
100
% of Total
Excess
0 10 20 30 40 50 60 70 80 90 100
% of Ranked Base Sample
As the scatter plots and judgment would indicate,
considerable overlap exists between the vehicles with large
excesses for HC and CO. The ninety vehicles with the worst
28
-------
HC excess emissions together account for 90% of the fleet
excess HC; the identical 90 vehicles account for 83% of the
fleet excess CO.
4.2.4. Analysis Using MOBILE4 Emitter Categories
The EPA MOBILE4 emissions model classifies light-duty
vehicles into four categories, from lowest to highest
emitters: passing, marginal failure, high failure and
super.emitter. The passing and marginal emitters, when taken
together as a single group, are frequently referred to as
"normal" emitters.13 Boundaries between emitter categories
are defined separately for HC and CO. The pollutant with the
highest emitter category determines the category of the
vehicle; thus, a vehicle that is a high emitter on HC and a
marginal emitter on CO is referred to as a high emitter.
Table 7 shows the upper bounds of the emitter categories
for HC and CO. Note that the boundaries between the
marginal- and high-emitter categories are technology and
model-year-group specific.14 Passing emitters have HC and CO
values that are each below their respective certification
standards. A vehicle with either an HC reading or a CO
reading exceeding the respective upper bound of the high-
emitter category is classified as a super emitter.
TABLE 7
Bounds of the MQBILE4 Emitter Categories
Technology Group
Carbureted 1981-82
Fuel-Injected 1981-82
Carbureted 1983+
Fuel-Injected 1983+
HC
Pass Marginal High
cert std 1.175 10.0
cert std 0.725 10.0
cert std 0.815 10.0
cert std 0.965 10.0
CO
Pass Marginal High
cert std 17.411 150
cert std 10.499 150
cert std 10.398 150
cert std 10.558 150
Figure 12 shows the breakdown of the base CTP sample by
the MOBILE4 approach, giving the percent of the sample in the
four quota groups (and the entire sample) that fell in each
emitter category.15 Thus, approximately 15% of the CTP fleet
were passing emitters, 25% were marginals, 60% were highs,
and less than 1% that is, one vehicle was a super
emitter. In simpler terms, 60% of the sample showed
significant HC or CO FTP noncompliance, while the remaining
40% could be considered "normal" emitters.
The figure shows that the fuel-injected 1983-86 group
was distinctive for containing a disproportionately large
percentage of marginal emitters and fewer high emitters.
29
-------
Carbureted vehicles were more likely to be high emitters than
were their fuel-injected counterparts of the same model year.
Similarly, vehicles from the earlier model year grouping were
more likely to be high emitters than their counterparts in
the later model-year grouping.
FIGURE 12
As-Rereived FTP Profile hv Quota
and Emissions Cateaorv
PI 81-82 GARB 81-82 GARB 83-86 Fl 83-86
ALL
SUPER
HIGH
D MARGINAL PASS
Table 8 shows the mean excess HC and CO emissions for
all 206 FTP failures (including the one super emitter), and
then isolates the mean excess HC and CO for the 62 marginal
emitters and the 143 high emitters. Thus, for example, the
mean excess for fuel-injected 1983-86 high-emitters was 2.21
g/mi HC and 35.4 g/mi CO. The mean excesses for the marginal
emitters are quite small, implying that most of the marginals
were in the lower part of the marginal range. The mean
excess HC for the carbureted 1983-86 group is actually
negative, indicating that the CO values for those vehicles
were driving the classification. One implication of these
low values for the marginal emitters is that the emissions
repair benefit to be derived from them is quite small.
The relatively high values in the "all fails" category
show the effects of the large number of high emitters in the
sample. Thus, for example, the 2.07 g/mi HC excess and 30.5
g/mi CO excess for the carbureted 1981-82 vehicles are
30
-------
several times greater than the standards applicable to those
vehicles. The mean excess for all failures in the fuel-
injected 1981-82 group is sensitive to the presence of the
single super emitter: the HC value would drop from 2.77 g/mi
to 1.93 g/mi if the super-emitter were to be removed.
Because of the large impact of the single super emitter
and the low mean excess displayed by the marginals, the
analysis below will in most instances focus on the high-
emitter category.
The mean excess HC values for the high emitters in the
four quota groups were comparable to the mean for all high
emitters of 2.22 g/mi, although the carbureted 1983-86 value
was somewhat lower. For mean excess CO, the fuel-injected
1981-82 category was considerably dirtier than the mean, and
the carbureted 1983-86 group was once again somewhat cleaner.
TABLE 8
Mean Excess Emissions of Failures bv Quota Group
Quota Group
Fuel-Injected 81-82
Carbureted 81-82
Carbureted 83-86
Fuel-Injected 83-86
ALL
Excess HC (g/ml)
All Marginal High
Fails
2.77 0.14 2.41
2.07 0.19 2.51
1.31 -0.01 1.67
1.24 0.07 2.21
1.66 0.09 2.22
Excess CO (g/mi)
All Marginal High
Fails
42.4 1.4 52.5
30.5 3.5 36.8
19.8 2.3 24.6
20.2 2.2 35.4
25.4 2.4 35.2
The MOBILE4 emitter categories were also used to
generate Figure 13, which shows the proportion of each
manufacturer's fleet that fell in each category. Although
the data in Figure 13 have not been adjusted for mileage,
model year, or technology factors, such factors do not
necessarily explain the differences between manufacturers.
Chrysler, for example, has the highest percentage of its
fleet (80%) in the high-emitter category, yet it was the
manufacturer with the highest percentage of its fleet under
50,000 miles (see Figure 4). The Chrysler fleet also
contained an above-average percentage of late-model vehicles,
and no fuel-injected 1981-82 vehicles (Figure 1). Toyota, on
the other hand, shows the smallest percentage of vehicles in
the high-emitter category (as well as the highest percentage
of "passes,") yet its fleet had the highest average odometer
reading of any manufacturer in the sample (Figure 5).
31
-------
FIGURE 13
PTP profile hy Manufacturer
and Emi 53 q ! rm «; Catecrarv
GM PORD NISS TOVT CHRY AMC VW MAZ SUBA MITS K>D ALL
SUPER
HIGH
MARGINAL D PASS
4.2.5. Correlation of Excess Emissions, Emitter
Category, and Odometer
Linear regressions on odometer against excess emissions
in the as-received fleet yielded essentially no correlation.
The degree to which dirty cars are dirty seems not to depend
much on their age. Figure 14, for example, shows the wide
scatter when excess HC is plotted against odometer for the
carbureted 1981-82 group; plots for the other groups and for
CO are similar. The R-squared values for the linear fits
performed on each quota group ranged from 0.2% to 5.2% for
both HC and CO, with similar poor correlation shown for the
fleet as a whole.
On the other hand, the mean mileages increase across the
MOBILE4 pass, marginal, and high categories for the
technology and model year groups with significant sample
sizes in each emitter category (Figure 15) . This is
consistent with an hypothesis that greater numbers of
vehicles move into the failing emitter categories with
increased mileage, although the actual excess emissions of
vehicles within a category may not be linearly related to
mileage.
32
-------
HC Excess
(g/mi)
FIGURE 14
Mileaye vs. Excess HC in 1-hp Carbureted
1QB1-R2 Quota ftronn
9
8
7
6
5
4
3
2
1
0
- 1
1
1
1
1
1
1
r\v
r T
hB
* B
.-v-.l"ii
1L "!.
IB
BJ
1
1
1
1
_ 1
1
t
1
"^ **
g 1
T
I 1
B .!
..!L ;'..-'..
i
f L ..!.....
i
i
i
i
i
i
i
...f .....
i
i
i
i
i
i
i
i
i
i
i
i
50
100
150
200
250
Mileage (1000's of Miles)
FIGURE 15
Mean Mileage of Model Year and FuelMetering Groups
bv MOBILE4 Emitter Catecrorv
Mean
Mileage
(K)
Fl
Garb
PASS
81-82
MARG
83-86
HIGH
ALL
33
-------
4.3. Michigan AET
4.3.1. Profile of the Base Sample
As mentioned in Section 2.1 above, vehicles were
recruited for the Cooperative Test Program on the basis of
failing their official I/M short test under the Michigan
Automobile Emission Testing (AET) program. The AET test
makes a pass/fail determination at idle, following a short
period of 2500rpm preconditioning.
A scatter plot of the HC and CO scores on the AET
screening test for the 239-vehicle sample appears in Figure
16. The major divisions in the figure represent multiples of
the 207 (b) cutpoints. Values that fall on the 10% CO
horizontal and the 2000ppm HC vertical are vehicles whose
emissions exceeded the maximum ends of the analyzer ranges.
Vehicles that passed the AET cutpoints of 1.2% CO and 220ppm
HC were not recruited for the program, which accounts for the
empty range in the lower left corner of the plot.
FIGURE 16
Results of the I/M Screening Test (AET
for the Base Sample
10.8
9.6 --
8.4 *
00
3.6 -
2.
7.2 -------*-B"-.*-----fc----j--.--fc----j-.---*-----
48i":\v--;.--
2.4 ~s"jri-ii-B»a ir'i " '
J&t ' . *
1<2 ...^i'-H--.....,'.^. ^...Jr--...^ T
220 440 660 880 1100 1320 1540 1760 1980 2200
HC (ppm)
34
-------
Noteworthy in Figure 16 is the large number of vehicles
that failed one pollutant (particularly in the first "cell"
above the outpoint), but passed the other pollutant.
Similarly, note the relatively small number of vehicles that
failed both pollutants in the area just in excess of the
outpoints. On the other hand, 30% of the base sample (74
vehicles) had either HC emissions above SOOppm or CO
emissions above 4%. Twenty-two of these vehicles with
extremely high short test scores were single-pollutant
failures, including the three 2000ppm HC vehicles at the
extreme bottom right of Figure 16.
The distribution by failed pollutant is reasonably
consistent with an hypothesis that some emission control
failures (or AET test irregularities) cause high idle HC, and
others cause high idle CO, but few cause both high HC and
high CO. The vehicles seen to have high HC and CO are
approximately accountable in terms of random simultaneous
occurrence of two failures.
4.3.2. Stratifications by Manufacturer and Quota Group
The frequent pattern of AET failure for a single
pollutant was maintained across quota groups as well as for
the entire base sample (Figure 17). For each group, the sum
of HC-only and CO-only failures exceeded the number of
combined HC/CO failures. In the two 1983-86 groups and the
fleet as a whole, the number of HC-only failures was itself
almost equivalent to the number of combined HC+CO failures.
FIGURE 17
V
e
h
e
s
Fl 81-82 CARB 81- GARB 83- Fl 83-86
82 86
ALL
35
-------
These observations on the Michigan AE.T results contrast
with the results on the as-received FTP test, presented in
Figure 6. There, HC-only failures were the rarity across all
quota groups; CO-only failures occurred at a somewhat higher
rate in the 1983-86 model year groups, and combined HC/CO
failures dominated the sample. This trend is not simply an
artifact of the cutpoints used to define FTP failure; that
is, there is not a disproportionate number of combined HC+CO
failures right above the standards, in the "marginal" failure
category. In fact, of the 144 highest emitters in the sample
(the 143 highs, plus the one super emitter), 141 were
combined HC+CO failures.
In further analyzing the types of AET failure, what was
true for the quota groups was not true for the manufacturers;
different manufacturers showed different proportions of HC-
only, CO-only and combined HC/CO failures. In Figure 18,
each manufacturer's CTP sample is divided according to the
number of vehicles that fell in each of the AET failure
types. Better than 70% of the Honda and Toyota vehicles were
HC-only failures, as were 57% of the GM vehicles. Together,
these three manufacturers accounted for almost three-fifths
of the HC-only failures in the base sample.
FIGURE 18
AET Failure Types bv Manufacturer
GM FORD MISS TOYT CHRY AMC VW MAZ SUBA MITS HOND
HC-Only DHC&CO CO-Only
36
-------
Chrysler and Ford, on the other hand, showed higher
proportions of their samples (60% and 47%, respectively) in
the combined HC/CO failure category. Subaru and Volkswagen
had close to half of their samples failing CO only, while
Honda had no CO-only failures at all.
4.3.3. Stratification by Mileage and Vehicle Type
Two other brief analyses were performed on the
relationship between the AET failure type and basic vehicle
characteristics. Analysis' of the numerical relationship
between vehicle mileage and the AET scores showed essentially
no useful correlation. Finally, no qualitative difference in
the AET failure types was noticed when the base sample was
disaggregated by vehicle type.
4.4. Correlation Between the FTP and Michigan AET Results
4.4.1. Introduction
Section 4.2.2 mentioned that 206 of the 239 vehicles
(86%) in the base CTP sample failed their as-received FTP
tests at certification standards. Because all vehicles
recruited for the CTP had failed their initial Michigan AET
test, the remaining 33 vehicles that passed their as-received
FTP are considered errors of commission (Ec) by the AET short
test. This section examines more closely the relationships
between the AET test results and the as-received FTP test
results for both the FTP failures and the Ec vehicles.
4.4.2. Excess Emissions and Emitter Categories of the
AET Failure Types
Figure 19 stratifies the sample by the type of AET
failure (HC-only, CO-only, or combined HC/CO) and the MOBILE4
emitter categories. As shown in the figure, the greatest
number of vehicles fell in the high emitter category for each
of the types of AET failure. Of the vehicles that failed
their AET tests for both pollutants, for example, 61 of 83
(or 73%) were high emitters making the HC+CO failure type
the most productive at identifying vehicles with significant
FTP nonconformity. Next-most productive were the CO-only AET
failures, where 62% were FTP high emitters. Least effective
were the HC-only failures. While these failures were the
most prevalent (38% of the sample) , it was the only AET
failure type where more normal FTP emitters were identified
than high emitters.
Figure 19 also shows a correlation between the type of
AET failure and whether or not the vehicle was an error of
commission. Of the 33 Ec vehicles in the sample, only three
37
-------
failed their AET short test for both HC and CO. Of the
remainder, Ec's among the HC-only failures outnumber those in
the CO-only failures by two to one.
FIGURE 19
Breakdown nf FTP Emitter Types bv Typp of
failure
BRASS
DMARG
SHIGH
SUPE
HC-only
BOTH
AET Failure Type
CO-Only
As shown in Table 9, vehicles in the high-emitter
category accounted for 93% of the excess HC and 96% of the
excess CO from among the 206 failures in the base sample.
Each type of AET failure contributed significantly to that
total, but the combined HC+CO AET failures accounted for
nearly half of the total excess (48.8% of the HC and 48.4% of
the CO) . This reflects both the larger number of HC+CO
failures among the high emitters, as well as a generally
larger mean excess per vehicle 2.73 g/mi for HC, and
41.6 g/mi CO.
On a per-vehicle basis, the CO-only AET failures matched
the HC+CO failures in their ability to identify excess CO
emissions among the high emitters. The CO-only failures even
identified a respectable percentage of the excess HC
emissions, with a mean excess HC of 1.7 g/mi, not far below
the 2.0 g/mi mean from the HC-only AET failures. In spite of
the high percentage of error of commission vehicles among the
HC-only failures, the group nevertheless identifies nearly
one quarter of the HC excess emissions among the failed
vehicles in the sample, as well as 14% of the excess CO.
38
-------
TABLE 9
ionship Between AET Failure Type and Excess Emrss i nn,«j
for the Hi ah Emitter
AET Failure
Type
HC-oniy
HCandCO
CO-only
All Fail Types
Mean Excess Emissions
HC (g/mi)
1.97
2.73
1.69
2.22
CO (g/mi)
17.8
41.6
43.7
35.2
% of Failed-Vehicle Excess
HC
24.3%
48.8%
19.9%
93.0%
CO
14.3%
48.4%
33.4%
96.1%
Because no vehicles were recruited into the CTP with
short test scores lower than the 207 (b) cutpoints, the data
cannot be used to evaluate the effectiveness of tighter
standards on the excess emissions identified by the AET test.
However, because so little of the excess apparently arises
from the marginal emitters (and by definition, none arises
from the passing vehicles), it is worth asking what impact
raising the outpoints might have had on the number of normal
emitters identified by the test. Reducing the number of Ec
vehicles, or perhaps even marginal emitters, while
maintaining acceptable rates for identification of high
emitters, would presumably be a desirable goal.
A scatter plot of the AET scores for the normal emitters
appears in Figure 20, and Figure 21 shows the analogous data
for the high emitters. The format is similar to that
employed earlier in Figure 16, except that each of the axes
has been truncated at four times the applicable 207 (b)
standard. Clearly, many of the normal emitters fall in the
zone between the 100% and 200% of the HC and CO standards.
Not surprisingly, however, many of the high emitters do
precisely the same thing. Together, the plots show that
there is no natural break in either the HC or the CO
distribution that would suggest short test standards for
excluding large numbers of the normal emitters from AET
failure, without inappropriately converting many of the high
emitters to AET passes.
39
-------
FIGURE 20
AETScores of the Normal Emitter1? T.vinrr
00
.6
Nssr tiie 207 cb) sta.nda.rd.g
...., ...".....
» , ," «
1
1
L.....J.......
L"
;- . "'B .:
1
0 220 440 660 880
HC (ppm)
FIGURE 21
AET Srores of the Hiah Emitters Lvina
Near the 207 (b) standards.
00
3/5
.b
2 A
.*»
1 9 .
c
m
g
*
1 1
22
i
.
.1 *
_ !
*B|
!0 44
g
~
r -
10 66
10 88
HC (ppm)
40
-------
4.5. :,ahQratory c;hnrc Tear. Reaulr.g
4.5.1. Sample Description
For a variety of reasons, not all of the 239 vehicles in
the base CTP sample completed short cycle testing with the
Basic I/M Test Procedure on their tank fuel. Six of the
vehicles were found to have insufficient tank fuel for the
3IT? after the procedure was underway; consistent with the
program plan, the procedure was completed on these vehicles
with commercial fuel.
Equipment and vehicle problems, as well as deviations
from the program plan, led to omitted core sampling periods
during parts of the BITP on seventeen vehicles. For all but
eight of the vehicles, as-received indolene BITP testing was
available for the missing tank/commercial modes, and the
indolene values have been substituted for the purposes of the
analysis below. Only one vehicle from the original 239
(vehicle 104) was missing enough data to be eliminated from
the short cycle analysis entirely.
In summary, tank fuel values were available for all but
a few vehicles, and all but a few modes. Where they were
not, commercial fuel values were used; where commercial fuel
values were unavailable, Indolene values were used. Sample
sizes that are less than 239 for particular modes result when
no substitute data (regardless of fuel type) were available.
4.5.2. Second-Chance Failure Rates of the As-Received
Base Sample
Recall from Section 2.4.2 that the Basic I/M Test
Procedure includes seven "core" sampling periods, each
consisting of a 30-second first-idle mode, a 30-second
2500rpm mode, and a 120-second second-idle mode. Thus, the
core sampling periods closely resemble the EPA Two-Speed Idle
Test. However, because all CTP vehicles had already failed
one short test (the AET idle test) prior to recruitment, the
core sampling periods in the BITP procedure represent
"second-chance" short tests.
Figure 22 analyzes the as-received failure rates for
fourteen of the idle modes found in the BITP, grouped as the
paired first- and second-idle readings of the seven core
sampling periods. The 30-second point in the modes were used
in all cases, and failure type was determined by comparing
those readings to the 207(b) HC and CO cutpoints. The.
numerical data on which the figure is based are provided in
Appendix C.
41
-------
The order of the modes in the figure duplicates the
order in which they were performed during testing. within
each pair in the figure, the only difference between the two
idles is an intervening 2500rpm mode, with the exception of
the restart (RS) pair, which had both an intervening 2500rpm
and a keyoff/restart. Between pairs, the difference between
the idles is the intervening conditioning (shown in the
bottom row of the table) that preceded the second pair.16
FIGURE 22
Failure Rates of the BITP Idle Modes
Extended
2500rpm
I Ext Idle
Each stacked column in Figure 22 gives the percent of
the base sample that failed the given mode for CO-only (solid
portion), HC-only (striped), and HC+CO (crosshatched). The
total height of the column therefore gives the overall
failure rate for the mode. The base operation and
conditioning for the modes are provided below the horizontal
axis. Thus, for example, 89% of the sample failed the first
idle mode of the BITP (CS-03) , which was an unconditioned
idle following a minimum one-hour soak. Following 30 seconds
of 2500rpm operation (CS-04), the failure rate dropped to 58%
during the second idle (CS-05) .
42
-------
Figure 22 shows that considerable variability existed
between the original Michigan Auto Exhaust Test results that
were the basis for recruitment into the CTP and the simulated
field short tests of the Basic I/M Test Procedure. With the
exception of the very first mode of the procedure (following
a "cold" start), the idle failure rates ranged between 40%
and 60%. This indicates that a second-chance short test,
almost regardless of how poorly it is performed, will reduce
the idle-test failure rate substantially.
By examining the adjacent idle modes in Figure 22, one
sees that the consistent impact of 2500rpm operation (both
30-second and 180-second) was to lower the idle failure rate.
Focusing just on the core sampling periods, the failure rate
for the second idle of each core sampling period was lower
than the rate for the first, attributable to the intervening
30-second 2500rpm mode. The magnitude of the reduction
between idles was greatest when the initial operating
condition was furthest from ideal: the 31-point drop
following the soak at the very beginning of the procedure,
and an eight-point drop following the 20-minute extended idle
in the middle of the procedure. Even in the conditions
considered more ideal, however, failure rates for the second
idle were three to four points lower than the first.
The impact of loaded operation was to reduce the failure
rate as well. In fact, extended 2500rpm and loaded operation
were apparently responsible for achieving the lowest failure
rates among the 14 modes in the procedure; these are the
values in the low 40's for XL-05, XL-10, and RS-02.17
On the other hand, the failure rate increased following
extended idles: the ten-minute idle before CS-12 generated a
ten-point rise, and the 20-minute idle at XI-02 led to an
eighteen-point rise. In each of these cases, the
accumulation of 2500rpm operation (including some extended
2500rpm modes) in succeeding modes eventually reversed the
effects of the long idles.
Not surprisingly, the idle immediately following the
initial soak period showed the highest failure rates of all.
Interestingly, only extended 2500rpm and idle operation were
then necessary to bring the failure rate down within a few
points of the rates achieved by extended loaded operation
later in the procedure.
Considering the effect of 2500rpm operation elsewhere in
the procedure, the apparent rise from the combined effect of
2500rpm and restart operation between the last two idle modes
(RS-02 and RS-05) implies that the restart alone might
increase the failure rate somewhat.
The Michigan AET test and the BITP idle modes showed
differences in the types of failure as well as in the overall
43
-------
failures rates. Recall from Section 4.3.1 that single-
pollutant failures were frequent in the AET results. As
shown in Figure 22, however, combined HC+CO failures were the
rule in the BITP idle modes, outnumbering the other two
failure types in every one of the modes. In fact, the HC+CO
failures outnumbered the sum of the HC-only and CO-only
failures in all of the idle modes except two: CS-05 and XI-
02. These two modes represented non-ideal test conditions:
one soon after the soak at the beginning of the procedure,
and the other following a 20-minute idle.
Of the three failure types, the CO-only group had the
most consistent failure rates across the different idle
modes: all fell in the range from 7% to 13%. The four
highest CO-only rates all came at the beginning of the
procedure, before the first extended 2500rpm operation had
occurred. Past that point, the CO-only failure rate never
exceeded 10% of the sample. The HC-only failure rate, on the
other hand, was apparently more sensitive to the type of
prior operation. All of the modes preceded by extended
2500rpm or loaded operation had HC-only failure rates of 7%-
9%; all the modes that immediately followed extended idle or
a soak had failure rates twice that high.
The above observations support the hypothesis that many
of the vehicles recruited for the CTP program were poorly
preconditioned in their original Michigan AET test. Roughly
one-half of the failures might have been avoided by better
preconditioning.
In each idle mode in Figure 22 where the vehicles appear
to have received adequate preconditioning (e.g., CS-10, XL-
02, XI-10), approximately 20% to 25% of the sample failed the
mode for both HC and CO. The possibility exists that these
vehicles also failed the AET test for both HC and CO, and
that they could represent a consistent set of failures across
the various short test conditions encountered in the initial
test (AET) and second chance tests. Table 10 compares the
AET failure types to the failure types on on the first idle
of the extended loaded sequence (XI-02) . The data show
considerable migration among the failure types between the
initial test and this second-chance test. Of the 83 vehicles
that failed both HC and CO on the AET test, for example, only
39 were of the same failure type on the second-chance test;
38 of the 83 changed from an HC+CO failure to a pass. The
fact that the second-chance test shows the 25% HC+CO failure
rate is due largely to the migration into that category of 17
vehicles from the AET CO-only category.
Given the earlier analysis of the AET HC-only failures,
it is perhaps less surprising that more than two-thirds of
these vehicles passed the second-chance test. Almost half of
44
-------
the AET CO-only failures passed the second-chance test as
well.
TABLE 10
rnmparison of Failure Tvpgs
the AET
and the
First Idle-Neutral of the Extended Loaded Seouenre>
Second-Chance
Failure Type
HC-Only
Both
CO-Only
Pass
Total
AET Failure Type
HC-Only
15
8
2
66
91
Both
3
39
3
38
83
CO-Only
1
1 7
16
31
65
Total
1 9
64
21
135
239
4.5.3. Idle-Mode Short Test Variability Due to Fuels
One-hundred, ninety-six of the vehicles in the base
sample underwent as-received Extended Loaded short cycles on
both Indolene and tank fuels.18 The HC and CO values for the
second idle (mode 5) of these sequences were compared to
determine the effect of fuel type. For 169 of these (or 86%)
there was no change in the 207 (b) pass/fail status for either
HC or CO between the tests on the different fuels; this is
shown by the sum of the bold-print numbers in Table 11. Of
the 27 remaining vehicles, there was an essentially
negligible trend to more frequently pass the tank fuel test
(a net increase of three failures occurred on Indolene) .
Although there were six more HC failures on tank fuel than
Indolene, most of these vehicles were already consistent CO
failures, so their overall pass/fail status did not change.
TABLE 11
Changes in Pass/Fail Status for HC and CO Between Tank and
Indolene Idle Tests
CO Status
P-P
P-F
F-P
F-F
Total
HC
P-P
1 08
5
0
1 2
125
Pass/Fail
P-F
2
3
0
2
7
Status
F-P
3
0
4
6
13
(Tank-lndolene)
F-F Total
1 1
0
2
38
51
124
8
6
58
196
45
-------
Were there to be a fuels-related impact on the idle test
scores, the volatility difference between the tank and
Indolene fuels would be the most obvious explanation. Fuel
RVP levels were available for 189 of the 196 tank fuel tests
used in the above comparison. The range in RVPs of this
sample was 5.0 to 15.6, with a mean of 11.7.19 One-third of
the RVP levels were above 12.0. The 27-vehicle sample that
showed variable pass/fail results on HC, CO or both, had
RVP's in the range 9.3'to 15.6, and also had a mean RVP of
11.7. Thus no volatility-related fuel variability was
evident in the first-idle neutral of the "ideal" short cycle
from the as-received laboratory testing.
The group of 27 vehicles that change pass/fail status on
either HC or CO between the Indolene and tank-fuel tests was
divided into three roughly equal-sized groups. Eight
vehicles had small score differences (changes in HC of less
than SOppm and in CO of less than 1.0%) that nevertheless
overlap the 207 (b) cutpoint. Eleven had extreme differences
(changes in HC of more than 400 ppm or in CO of more than
4%). The remaining eight vehicles had moderate changes.
For these groups, there was no correlation between the
magnitude and the sign of the change in scores (i.e., no
trends for one fuel being more failure-prone at idle) , or
between the magnitude of the change in scores and the RVP of
the tank-fuel test.
4.5.4. Correlating the Laboratory Short Tests with the
MOBILE4 Emitter Categories
Based on the results of the previous subsection, the
question arises whether there are patterns to the FTP
emissions of the vehicles that changed pass/fail status
between the Michigan AET test and the various short cycles in
the Basic I/M Test Procedure. Figure 19 begins this analysis
with the laboratory short test results for the vehicles that
were either passing emitters or marginal emitters on their
as-received FTP tests, as well as the passing and marginal
emitters taken together as a group. Thus, these are the
vehicles that would presumably show minimal impacts from I/M-
instigated repair.
Each collection of three bars in Figure 23 represents
the percentage of the given MOBILE4 emitter category that
passed a particular core sampling mode from the Basic I/M
Test Procedure. Again, the horizontal axis gives the
significant vehicle operation that preceded the sampling
period. In each case, the data were from the second idle
mode of the core sampling period. Based on the discussion in
Section 23, the second idle of each period showed almost
uniformly higher pass rates than did the first idle.
46
-------
Thus, for example, the middle set of bars shows the
rates for the fifth mode of the Extended Loaded sequence (XL-
05) , which was the second idle-neutral following an extended
stretch of loaded operation. From the figure, 100% of the
passing emitters passed this "second-chance" idle test. For
the marginal emitters, the comparable figure on XL-05 was
85%, and for the passing and marginal emitters taken
together, 91%.
FIGURE 23
Response of the Normal FTP Emitters to
Second-Chance Short Testing
Pass
Marginal
Pass+Marginal
XL-OS
XMO RS.C
P
a
s
s
Base
Operation:
Conditioning:
Warm
Soak
None
Warm Warm Soak Extended Extended Extended Extended
Soak Extended Loaded Idle Idle Loaded
2500rpm
Extended Extended
2500rpm Idle
None None Extended Restart
2500rpm
Figure 23 illustrates the significant reduction in
failure rates of the normal FTP emitters that would probably
have accompanied any of the sequences, had one been employed
as a second-chance test in the Michigan AET program. Five of
the seven sequences show pass rates for the normal emitters
that are above 75%. For the three fully-warmed sequences
preceded by extended loaded or extended 2500rpm base
operation, the error of commission rate would have been at or
near zero.
47
-------
The apparent success of second-chance testing at passing
normal emitters does not come without cost, however. Second-
chance testing also reduces significantly the failure rates
among the FTP high emitters, and consequently reduces the
available emissions benefit to be gained by repair efforts.
Figure 24 presents the response of the FTP high emitters in
the base sample to the various second-chance tests of the
Basic I/M Test Procedure. Each column in the chart
corresponds to the percentage of the original 239 vehicles
that failed the second idle-neutral of the indicated
sequence. For example, somewhat over 70% of the vehicles
failed their second-chance cold start test, with HC+CO
failures being the largest group.
FIGURE 24
Response of the Hioh Emitters to
SecondChance Short Testing
CO-only
HHC+CO
E3 HC-only
F
a
Base
Operation:
Conditioning:
Warm
Soak
None
Warm
Soak
Warm Soak Extended Extended
Extended Loaded Idle
2500rpm
Extended Extended
2500rpm Idle
None
None
Extended Extended
Idle Loaded
Extended Restart
2500rpm
The effect of the various types of vehicle operation on
the failure rates of the high emitters was less severe than
was seen earlier in the normal emitters, but was
directionally the same. Between 28% and 39% of the original
failures fell away, depending on the second-chance test. The
48
-------
lowest second-chance failure rates occurred after extended
loaded operation (XL-05 and RS-05) , followed closely by the
rates on fully warmed vehicles with extended 2500rpm
operation (XI-10). Thus, if one were to address the error of
commission problem through such preconditioning and second-
chance testing, between 35%.. and._..40% of _tne__hAgh emitters
captured by the initial test might be lost. ~ The" "highest
failure rate occurred on the vehicle with the worst
conditioning -- the cold start sequence which was shown
earlier in Figure 23 to exhibit the worst error of commission
rate.
4.5.5. Variability Between Adjacent Idle Modes
Table 12 uses linear regression on the HC and CO
emission values to analyze the variability between the first-
idle and second-idle short test modes in the seven core
sampling sequences of the Basic I/M Test Procedure. Recall
from Section x.x that these two modes are separated by a 30-
second 2500rpm no-load mode, making the core sampling
sequence roughly equivalent to a two-mode idle test. Once
again, the base operation and conditioning modes are provided
for each of the modes as a guide to the I/M test conditions
that each simulates.
TABLE 12
Regressions on the First and Second Idle
Modes of the Core Sampling Periods
Base
Operation
Conditioning
Sample Size
HC
Slope
Intercept
R- squared
CO
Slope
Intercept
R-squared
MODE
CS-03/05
Warm
Soak
None
237
0.69
252
25.8%
0.65
2.03
30.9%
CS-07/10
Warm
Soak
Extended
2500rpm
238
0.99
25
88.5%
0.95
0.27
85.6%
CS-012/15
Warm Soak
Extended
2500rpm
Extended
Idle
237
0.97
51
72.8%
0.89
0.57
70.4%
XL-O2/05
Extended
Loaded
None
238
0.97
19
85.5%
0.96
0.18
91.1%
XI-02/05
Extended
Idle
None
237
0.94
57
69.7%
0.89
0.60
69.2%
XI-07/10
Extended
Idle
Extended
2500rpm
237
1.00
22
90.7%
0.95
0.29
88.0%
RS-02/05
Extended
Loaded
Restart
231
1.03
5
90.0%
0.97
0.12
88.7%
49
-------
For a variety of reasons, not all vehicles were tested
with each of the sequences in the Basic I/M Test Procedure.
The sample sizes in Table 12 reflect the number of vehicles
in the 239-vehicle sample where paired first-idle and second-
idle modes were available for the given segment of the
procedure.
The regressions are least-squares fits to lines of the
form y = mx + b, where m is the slope and b is the y-
intercept. In each case, the x-values were taken to be the
first idle of the pair, and the y-values were the second
idle. If there were no variability between the first- and
second-idle modes, the relationship between the emission
scores for each pollutant would be a line with slope of one,
intercept of zero, and 100% R-squared value.
As shown in the table, four segments of the procedure
showed slopes greater than 0.95 and correlation greater than
85% for both HC and CO; these were the Cold Start modes 7/10,
Extended Loaded modes 2/5, Extended Idle modes 7/10, and the
Restart modes 2/5. In two of these segments (CS-07/10 and
XI-07/10), the short test was immediately preceded by three
minutes of 2500rpm conditioning. In the other two segments
(XL-02/05 and RS-02/05), the short test was preceded by an
LA4 prep cycle.
Not surprisingly, the worst correlation (R2 values below
30% for both HC and CO) was shown by the paired values for
the Cold Start modes 2/5, which came at the very beginning-of
the procedure, following an extended soak and no
conditioning. The fact that the slope of the HC and CO
regression lines for this comparison are well below one shows
that even the short period of 2500rpm operation between the
two idle modes was sufficient to reduce emissions
considerably on many vehicles.
The two remaining segments, which also showed poorer
correlation, were the Cold Start modes 7/10 and the Extended
Idle modes 2/5. These short tests immediately followed
periods of extended idle: 10 minutes in the case of the Cold
Start, and 20 minutes in the case of the Extended Idle.
Based on the preceding discussion, periods of extended
no-load off-idle operation and extended loaded operation can
reduce the variability in short test scores that might result
from periods of extended idle operation or from testing a
vehicle too promptly after soak periods.
Figure 25 shows the actual scatter for one of the pairs
of idles with good correlation: the first- and second-idle HC
values from the extended loaded sequence. Recall that the
base operation before the first idle mode in this sequence is
50
-------
an LA-4 prep cycle. In this case, the scatter was
responsible for reducing the correlation coefficient to
85.5%. The maximum values for each axis have been set below
the actual maximums in the sample in order to allow
examination of the region around the 207(b) HC cutpoint of
220 ppm, where a number of vehicles changed pass/fail status;
no actual data point that falls beyond the range of the graph
passed HC on either idle mode.
The plot clearly shows that a number of vehicles had
significant differences between their first-idle and second-
idle HC scores, even though the 30 seconds of 2500rpm that
separated the two idles might seem inconsequential compared
to the LA-4 that preceded the first idle. Because of their
variability, some of these changed pass/fail status for HC:
data points with the open-square symbol are vehicles that
failed the first idle, but changed to pass on the second;
points marked with an open triangle passed the first idle but
failed the second.
FIGURE 25
Scatter Between First Idle and Second Idle
Following Extended Loaded Operation
2nd
Idle
H3
(ppm) 440
u
rt
u
n .
u
0 -
0 -
0.
1
0
r i
1
1
1
1
1
.... ...L. ........
1
t i
.......... L ... .....
1
. I
A A '
i
.......... ..... . .^ .
A 2 * "
^A;i ;'..
^~ " 1 1 " ^ * ^^
!jifis*-«il * 1 D
220 4'
r i
I
.- " J
"a""0""
W 66
i
I
.........J
'
0 8f
|
...J...!.j
.........J
......... J
a
JO 11
00
1st Idle HC (ppm)
51
-------
When both HC and CO are considered, 12% of the base
sample 29 vehicles changed failure type (pass, HC-only,
CO-only, HC+CO) between the two idle modes of the Extended
Loaded sequence (the sum of the bold entries in Table 13) .
Twenty-three of the 29 also changed their overall idle test
status: 15 failed the first idle and passed the second, and
the balance did the reverse. As the table shows, however,
most of this difference came from CO-only failures that were
cleaner following the 2500rpm mode; the changes in other
failure types for the most part offset each other.
TABLE 13
Distribution of the Base Sample by Failure Type on
Idle Modes Following Extended Loaded Qpp»ra-Mnn
Second Idle
(XL-05)
Pass
HC-only
HC+CO
CO-only
All
First Idle (XL-02)
Pass
126
6
2
0
134
HC-only
7
1 1
1
0
19
HC+CO
3
1
58
2
64
CO-only
5
0
2
1 4
21
All
141
18
63
16
238
TABLE 14
Distribution of the Base Sample by Failure Type
on Idle Modes Following Extended Idle Operation
Second Idle (XI-05)
Pass
HC-only
HC+CO
CO-only
All
First Idle (XI-02)
Pass
89
4
2
2
97
HC-only
1 0
28
1
0
39
HC+CO
1 2
9
52
5
78
CO-only
4
0
4
15
23
All
115
41
59
22
237
Table 14 makes a similar comparison for the idle modes
in a sequence with poor preconditioning, the first and second
idle modes of the extended idle sequence. Here, the base
operation before the first idle test is a twenty-minute idle;
the intervening mode between the two idle tests is once again
30 seconds of 2500rpm operation. The number of vehicles that
change their failure type in this case rises to 53, or 22% of
the base sample. Note, however, that almost three quarters
of these were more serious failures on the first idle; for
52
-------
example, 26 vehicles changed from first idle failures of one
type or another to passes on the second idle. This is
consistent with the information presented in Figure X above,
in that even a brief stretch of 2500rpm operation appears to
compensate for extended periods of idle operation.
4.6 Supplemental Analysis of the AET Errors of Commission
As discussed previously in this section, 33 of the 239
vehicles in the CTP base sample were errors of commission by
the Michigan AET test. A table of basic emissions data and
vehicle identifying information for these vehicles appears in
Appendix D.
Figure 19 above showed that the Ec vehicles were not
randomly distributed by AET failure type; a disproportionate
number were HC-only AET failures. The distribution of Ec
vehicles also varied by basic vehicle characteristics. All
but six of the 33 vehicles fell in the 1983-86 model years;
the fuel-injected 1983-86 vehicles were the most heavily
represented quota group, with 19 vehicles (53%) of the total
Ec fleet. Differences were also evident between
manufacturers. Six of the sixteen Toyota vehicles were
errors of commission, while there were no Chrysler Ec' s .
Trucks were over-represented; there were eight LDT Ec's out
of the 33 total (24%), while trucks only represented eight
percent of the base sample. Five out of the eight LDT Ec' s
were Fords, all of them carbureted.
One particular engine stands out in the error of
commission fleet: the GM 151 CID fuel-injected LDV. Eight
of these vehicles appear in the list of Ec's, representing
all but two of GM's total Ec's, and almost one-quarter of
the total number of Ec's in the CTP fleet. Notably, all of
the eight vehicles failed the AET test for HC.
As shown in the data of Appendix D, almost all of the
AET errors of commission in the CTP fleet had elevated AET HC
scores; a surprising number were above 400ppm HC, and three
exceeded lOOOppm. The sample included thirteen AET errors of
commission for CO, with the highest showing a 6.7% CO score.
Appendix D also provides data from the 30-second point
of the first idle-neutral from the extended loaded sequence
(XL30HC and XL30CO), performed on the as-received vehicles.
Recall that these values represent idles following an
extended period of loaded preconditioning (an LA4) . In most
cases, the HC values during this mode of the extended loaded
sequence were much lower than the comparable values from the
AET test. However, the XL30HC values for five of the 33 were
still elevated from normal levels, and three continued to be
53
-------
EC'S. None of the vehicles that was a CO error of commission
on the AET test showed failing XL30CO values during its as-
received testing, although a handful exceeded 0.5% CO.
With only a few exceptions, no component problems were
identified during the as-received diagnosis that might have
explained elevated values for the Ec vehicles on the AET
test. Thus, only five of the 33 Ec vehicles had repair
efforts, and only one of those could be considered successful
in resolving observed emissions anomalies. This was vehicle
21, a 1984 Toyota, whose malperforming oxygen sensor was
replaced, leading to normal HC and CO levels.
For most of these Ec vehicles, no specific explanation
for the elevated AET scores was available. GM attributed
elevated AET HC values in its Ec fleet to inadequate
preconditioning during the AET test. Nevertheless, two of
GM's vehicles (338 and 347) that were HC errors of commission
by the AET test also showed failing values on the extended
loaded sequence, in spite of the LA4 preconditioning. These
and others of GM's Ec group did show elevated HC values
during other idle modes of their as-received short testing.
For one AMC vehicle (31) , the AET failure was probably
attributable to an air diversion timer tied to the elapsed
time at idle.
54
-------
SECTION 5: EMISSION EFFECTS OF REMEDIAL MAINTENANCE
5 . 1 Introduction and Sample
Prior to any repair, each vehicle underwent an as-
received characterization to aid in the decision of which
repairs were necessary, if any. This characterization
included the BITP, FTP, and a complete diagnosis of the
vehicle's engine and emission control systems. If the
results from this characterization indicated that repairs
were necessary to meet the criteria for vehicle release, a
repair sequence was designed. The first criterion was a
reduction in FTP emissions to a target based on the vehicle's
certification standards. Once this criterion had been met,
repairs targeted emission levels and variability on the I/M
test. The vehicle was released when all criteria had been
met or all reasonable repair efforts were completed.
Of the 239-vehicle sample, 184 vehicles received a total
of 479 remedial maintenance (RM) steps for which mass
emissions data were collected. Of these steps, 372 were
single, isolatable repairs; the remaining 107 RM steps
included 258 repairs, for a total of 630 repairs. Each
vehicle therefore received an average of 3.4 repairs in 2.6
RM steps.
The design of the CTP dictated that catalyst replacement
be a last resort repair so that the high conversion
efficiency of a new catalyst would not mask the necessity of
other repairs; under certain conditions, the CTP program plan
did not then require a final mass emissions test. Therefore,
much of the following analysis is focussed on pre-catalyst
repairs only. This included 413 RM steps on 175 vehicles.
Vehicles that passed the CTP standards (150% of cert
standards for mileage <=50K; 200% for mileage >50K) of the
as-received FTP as well as the ideal I/M portion of the BITP,
with little variability throughout the BITP, did not undergo
repair.
5.2 Total Mass Emission Reductions from the CTP Fleet
5.2.1 Net Benefit of Repairs FTP-Based
The reduction in emissions of the CTP due to repair was
substantial, with nearly all of the emissions in excess of
certification standards being eliminated. The net emissions
benefit achieved from the repairs on these 184 vehicles was
339 g/mi HC and 5193 g/mi CO, for an average reduction per
vehicle of 1.8 g/mi HC and 28.2 g/mi CO. (The certification
55
-------
standard for the majority of these vehicles is 0.41 g/mi HC
and 3.4 g/mi CO). On a percentage basis, HC emissions were
reduced by slightly over 80%, and CO emissions by rjver 85%
for these 184 vehicles. These reductions eliminated almost
99% of the excess HC and CO FTP emissions, relative to
individual vehicles' certification standards, of the entire
CTP fleet.20 See Tables 15 and 16 for breakdowns by
manufacturer and quota group. Note that greater than 100% of
a group's excess can be eliminated; this is caused by
vehicles that are repaired to levels cleaner than their
certification standard.
TABLE 15
FTP Emission Reductions (a/mi]
for All Repairs
bv Manufacturer
HC
GM
FORD
NISS
TOYT
CHRY
AMC
VW
MAZD
SUBA
MITS
HOND
ALL
number of emissions
vehicles reduction
46
47
17
8
15
13
10
7
8
6
7
184
86.56
105.18
25.66
4.68
24.44
25.07
17.24
9.24
20.58
10.08
10.24
338.98
%
reduction
83.6%
83.3%
77.8%
42.3%
79.3%
71.0%
71 .4%
80.6%
87.6%
84.1%
91.2%
80.3%
average
reduction
1.88
2.24
1.51
0.58
1.63
1.93
1.72
1.32
2.57
1.68
1.46
1.84
total
excess*
85.60
105.82
25.96
8.36
24.68
27.74
20.55
8.72
19.06
8.98
8.65
344.12
% excess
reduced
101.1%
99.4%
98.9%
55.9%
99.0%
90.4%
83.9%
106.0%
108.0%
112.3%
118.4%
98.5%
CO
GM
FORD
NISS
TOYT
CHRY
AMC
VW
MAZD
SUBA
MITS
HOND
ALL
* excess
number of
vehicles
46
47
17
8
15
13
10
7
8
6
7
184
is based on entire
emissions
reduction
1091.3
1374.1
763.6
103.9
395.4
384.2
327.6
220.0
356.6
113.1
62.9
5192.6
239-vehicle sample
%
reduction
86.4%
87.1%
91.4%
57.9%
85.9%
78.4%
83.1%
84.9%
87.6%
82.1%
89.2%
85.5%
average
reduction
23.7
29.2
44.9
13.0
26.4
29.6
32.8
31.4
44.6
18.9
9.0
28.2
total
excess*
1051.8
1378.5
770.1
147.8
394.9
404.8
359.5
241.5
344.5
106.4
50.8
5250.7
% excess
reduced
103.8%
99.7%
99.1%
70.3%
100.1%
94.9%
91.1%
91.1%
103.5%
106.3%
123.8%
98.9%
56
-------
TABLE 16
FTP Emission Reductions (g/mi) for All Repairs --
bv Quor.a Group
HC
FI81-82
Garb 8 1-82
Garb 83-86
F! 83-86
ALL
number of
vehicles
18
62
36
68
184
emissions
reduction
52.00
122.84
58.44
105.71
338.98
%
reduction
83.6%
78.7%
80.4%
80.5%
80.3%
average
reduction
2.89
1.98
1.62
1.55
1.84
total
excess*
55.42
130.60
56.12
101.97
344.12
% excess
reduced
93.8%
94.1%
104.1%
103.7%
98.5%
CO
FI81-82
Garb 8 1-82
Garb 83-86
Fl 83-86
ALL
* excess is
number of
vehicles
18
62
36
68
184
based on entire
emissions
reduction
830.7
1869.3
829.7
1662.8
5192.6
%
reduction
90.1%
82.6%
84.0%
87.5%
85.5%
average
reduction
46.1
30.2
23.0
24.5
28.2
total
excess*
849.0
1921.8
834.0
1645.8
5250.7
% excess
reduced
97.8%
97.3%
99.5%
101.0%
98.9%
239-vehicle sample
Because of the special treatment of catalyst repairs in
the CTP, they are here separated from the analysis. Of the
184-vehicle sample, 175 vehicles received non-catalyst
repairs; for 168 of these, both pre- and post-repair FTP data
is available. This includes tests on vehicles that received
a catalyst change at a later RM stage. The net emissions
benefit from these non-catalyst repairs was 264 g/mi HC and
4544 g/mi CO, for an average reduction per vehicle of 1.6
g/mi HC and 27.1 g/mi CO. On a percentage basis, HC
emissions were reduced by over 67%, and CO emissions by
almost 80%. These reductions eliminated at least 76% of the
excess HC and 86% of the excess CO emissions of the entire
CTP fleet, and 84% and 92% of the excess HC and CO,
respectively, of this 168-vehicle sample; more reductions may
have occurred for which both pre- and post-repair data is not
available.
Obviously, catalyst repairs had a significant impact on
the overall emission reductions even though they were usually
the final repair to be performed, occurring when emission
levels had already been significantly decreased. Their
contribution to the total emissions benefit was due to the
unusually low FTP levels after a catalyst repair rather than
to excessively high levels prior to that repair; emissions
were lower, on average, prior to a catalyst replacement than
before other types of repair.
57
-------
Catalyst replacements for which FTP data is available
accounted for 10% of the RM steps that occurred; their repair
eliminated 17% of the excess HC and 8% of the excess CO of
the entire CTP fleet. An additional 1% of the RM steps also
involved catalyst replacements, but lack of data does not
allow an assessment of their emissions impact. Final
emission levels following a catalyst replacement averaged
0.33 g/mi HC and 4.2 g/mi CO approximately half the levels
of the final non-catalyst repairs, at 0.76 g/mi HC and 7.1
g/mi CO. The average reduction for a single catalyst repair
eliminated 77% of the HC and 65% of the CO ^emissions
occurring just prior to the repair, and 44% and 32%,
respectively, of the vehicle's entire as-received HC and CO
emissions. In contrast, the average non-catalyst repair
eliminated 38% of the HC and 51% of the CO emissions
occurring just prior to the repair, and 29% and 39%,
respectively, of the vehicle's as-received HC and CO
emissions. Thus, catalyst repairs were somewhat more
productive in eliminating HC than were repairs to other
components, while CO was approximately in the same range.
5.2.2 FTP versus LA4 values
One purpose of the CTP was to obtain emission benefits
of individual repairs, as measured by the FTP. However, in
order to streamline the testing process, labs were not
required to perform an entire FTP after a repair if there was
reason to believe that there had been no effective emission
benefit. In case of uncertainty, the lab could perform an
LA4 (bags 1 and 2 of the FTP) , without the extended vehicle
preconditioning required in a complete FTP. Results of the
LA4 were to be used to decide if the FTP was needed;
significant emission decreases would mean that it was.
Unfortunately, this process was not always followed, with the
follow-up FTP sometimes eliminated. For this reason, FTP
data is not available before and after each obviously
significant repair.
Rather than eliminate these repairs from the analysis
a significant proportion for at least one manufacturer it
was judged better to retain all the data and perform the
analysis on an LA4 basis. For repairs without an LA4, FTP
data was converted to LA4 data by using results from bags 2
and 3; bag 3 has the same driving cycle as bag 1, and
approximates the warmed-up vehicle condition seen in the LA4-
only tests.
It was important to determine the effectiveness of thus
modelling FTP emission reductions with LA4-derived values,
and, if effective, the approximate shift in values one could
expect. As a first step to accomplish this, FTP values were
compared to the LA4 values derived from the FTP scores on
those same vehicles. The scatterplots in Figures 26 and 27
58
-------
show that FTP-derived LA4 values tracked FTP values quite
consistently for both HC and CO, indicating that derived LA4
values are generally an effective approximation of FTP
scores.
FIGURE 26
FTP vs FTP-Derived LA4 Emission Values -- HC
LA4HC
LA4HC = 0.96(FTPHC)-0.16 R2 = 92.8% (19.1,18.6)
4 6
FTP HC (g/mi)
10
12
FIGURE 27
FTP vs FTP-Derived LA4 Emission Values CO
LA4CO
LA4CO - 0.99(FTPCO) - 1 .95 R2 - 98.9%
160
140 -
120
i A A 1 no
LA4 iwvi
rn BO .
lnlmi\ ert
\yn"l oU
40
20-
0 t
«&?
J«4S*T^
^
tF*r
rl
pfn
a
r9D °
0 20 40 60 80 100 120 140 160
FTP CO (g/mi)
59
-------
Table 17 shows the total and average FTP as-received
emissions and emission reductions for all vehicles that
received repairs, and the LA4 values derived from those same
FT? results. It also gives the ratio of the derived LA4 to
FTP values, shown as a percentage in the last row of each
section of the table. Figures 28 and 29 show the scatter in
g/mi reduction per RM as measured by the two methods.
Derived LA4 emission values and g/mi emission reductions
generally were somewhat lower than FTP values, as illustrated
in Figure 30; derived LA4 values values are the lower of each
set of lines, with dotted lines showing HC and solid
indicating CO. On the other hand, the percent reductions as
measured by the derived LA4 were several percentage points
greater than as measured by the FTP.
The similarity of the FTP and derived-LA4 values
indicate that any conclusions that would be drawn from FTP
data would not be significantly altered by the use of
derived-LA4 values. Therefore, much of the remaining
analysis is performed using derived-LA4 and actual LA4 data
as an FTP substitute. The emission impact of thus using LA4
rather than FTP values can be illustrated by repeating some
of the information found in Section 5.2.1 above, describing
net emission reductions due to repairs; this time, LA4 rather
than FTP values are used. A comparison of the two reveals
little difference in the substance of the findings.
TABLE 17
Emission Values as Measured bv the FTP and FTP-Derived LA4
HC
FTP
LA4
LA4/FTP
number
vehicles
184
184
total
as-rcvd
422.34
382.99
90.7%
total
reduction
338.98
321.91
95.0%
average
as-rcvd
2.30
2.08
90.7%
average
reduction
1.84
1.75
95.0%
%
reduction
80.3%
84.1%
104.7%
CO
FTP
LA4
LA4/FTP
number
vehicles
184
184
total
as-rcvd
6074.3
5665.9
93.3%
total
reduction
5192.6
5063.0
97.5%
average
as-rcvd
33.0
30.8
93.3%
average
reduction
28.2
27.5
97.5%
%
reduction
85.5%
89.4%
104.5%
60
-------
FIGURE 28
FTP vs FTP-Derived LA4 Emissi Reductions --
ALA4HC
ALA4HC - 0.99(AFTPHC) - 0.05 R2-93.1%
FTP HC reduction (g/mi)
FIGURE 29
FTP vs FTP-Derived LA4 Emission Reductions -- CO
ALA4CO
ALMCO - 0.99(AFTPCO) - 0.23 R2 - 99.2%
-50
0 50 100
FTP CO reduction (g/mi)
150
61
-------
FIGURE 30
Average Emission Reductions Due to Repair: FTP
FTP-D^ri vpri LA4
as received
after final repair
5.2.3 Net Benefit of Repairs LA4-Based
The net emissions benefit achieved from the repairs on
these 184 vehicles was 322 g/mi HC and 5063 g/mi CO, for an
average reduction per vehicle of 1.7 g/mi HC and 27.5 g/mi
CO. On a percentage basis, HC emissions were reduced by
slightly over 84%, and CO emissions by almost 89% for these
184 vehicles. These reductions -- as measured by the LA4
eliminated approximately 95% of the excess HC and CO FTP
emissions, relative to certification standards, of the entire
CTP fleet.
Of the 184-vehicle sample, 175 vehicles received non-
catalyst repairs for which pre- and post-repair LA4 data is
available. This includes tests on vehicles that received a
catalyst change at a later RM stage. The net emissions
benefit from these non-catalyst repairs was 255 g/mi HC and
4575 g/mi CO, for an average reduction per vehicle of 1.5
g/mi HC and 26.1 g/mi CO. On a percentage basis, HC
emissions were reduced by over 69%, and CO emissions by
almost 83%. These reductions eliminated at least 74% of the
excess HC and 87% of the excess CO emissions of the entire
CTP fleet, and 78% and 90% of the excess HC and CO,
respectively, of this 175-vehicle sample; more reductions may
have occurred for which both pre- and post-repair data is not
available.
62
-------
5.3 Overview of the Repairs Conducted
5.3.1 System and Subsystem Repair Categories
Repairs were categorized by the testing organizations,
with some advice from EPA staff, into the following systems
and subsystems.
1. induction system
heated air door assembly
temperature sensors
air filter element
hoses
other (e.g., gaskets)
2 . fuel metering system
carburetor assembly
idle mixture adjustment limiter
idle mixture adjustment
idle speed
idle speed solenoid
fuel injection components
hoses, lines, wires
choke adjustment -- notches
choke adjustment -- vacuum break
choke adjustment limiter
fast idle speed
vacuum diaphragms
electrical controls
exhaust heat control valve assembly
hoses, lines, wires
other (e.g., fuel filter, float level)
3 . ignition system
distributor assembly
initial timing
initial timing limiter
spark plugs and wires
vacuum advance assembly
spark delay devices
spark knock detector
electronic timing module
coolant temperature sensors
hoses, lines, wires
other (e.g., points, distributor cap)
4 . EGR system
egr valve assembly
back pressure transducer
delay solenoid
vacuum amplifier
vacuum reservoir
coolant temperature sensor
63
-------
hoses, lines, wires
other (e.g., gaskets, plugged manifold)
5. air injection system
air injection assembly
bypass valve, dump valve air pump system
air diverter valve
check valve
drive belt
hoses, lines, wires
other (e.g., air filter, stuck valves)
6. PCV system
pcv valve assembly
filters
hoses and lines
other (e.g., vent tube seal)
7. exhaust system
exhaust manifold, tailpipe, muffler
catalytic converter
other (e.g., mixture set tube)
8. evaporative system
evaporative canister
canister filter
canister purge solenoid/valve
hoses, lines, wires
other (e.g., gas cap, gaskets)
9. engine assembly
engine assembly
cooling system
valve adjustment
belt tensions
hoses, lines, wires
other (e.g., battery, transmission fluid)
10. three-way catalyst system
electrical control unit
oxygen sensor
barometric pressure sensor
load sensor (throttle position, manifold vacuum)
engine speed sensor
coolant temperature sensor
crankshaft position sensor
egr position sensor
egr control solenoids
air/fuel control actuator
air bypass solenoid/valve
air diverter solenoid/valve
throttle kicker/actuator
idle speed control system
hoses, lines, wires
64
-------
diagnostic bulb check
diagnostic warning
other (e.g., switches)
In addition to these categories of systems and
subsystems repaired, the CTP database includes a code listing
the nature of repair -- replaced, adjusted, cleaned,
reconnected, restored, or rebuilt. Narrative comments,
filled out for each RM step, elaborated on the exact
components, diagnostic techniques, and other details judged
relevant by the technician but not covered by the coding
system.
5.3.2 Emission Benefits per System Repair
According to the CTP program plan, repairs were to be
done one at a time, with mass emission tests before and after
each repair, and in decreasing order of their likely impact
on emission reductions. In fact, this happened much of the
time, resulting in a substantial database of isolatable
repairs with bracketing mass emission tests. For these
cases, the emissions reductions can be simply averaged over
all of the occurrences of a particular repair. Attention
must be paid to the possibility that some repair types with
apparently low average benefits were the result of
misdiagnosis as to what needed repair.
However, in a number of cases, more than one repair
occurred prior to a post-repair emission test being
performed, resulting in a number of non-isolatable repairs.
For these, simple averages for each repair type would have
resulted in counting the entire emission reduction of the
grouped repairs for each of the repairs in the group.
To overcome this problem, multiple linear regressions
were performed, using the systems listed in Section 5.3.1 as
variables. The change in emission levels for each pollutant
was regressed across the ten systems, resulting in the
emission reduction for each pollutant to due repairs to each
system. The regression for HC reduction took the form
10
ALA4HC = 0 + Z(ALA4HC)i x (indicator for system^ repair)
i-l
A similar regression was performed for CO reduction. Note
that regressions were calculated with zero as a constant
term; that is, the results were forced through the origin, so
that if no repair occurred, the result would be no emission
reduction.
The following table lists repair results at the system
level. Included are both the simple averages of the repairs
which occurred singly, and the results of the multiple
regression for all repairs. All emission values are in grams
65
-------
per mile, as measured by the LA4 .
breakdowns by quota group.
See Appendix E for
TABLE 18
Emission Reductions per System Rppair
SYSTEM
REPAIRED
Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
All
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
14
94
63
19
77
4
60
4
10
156
501
A HC
0.33
1.07
0.19
0.11
0.41
-2.51
1.20
0.31
0.90
0.55
0.62
A CO
0.0
13.2
-1.4
1.5
6.7
-9.4
7.3
3.0
6.8
17.8
9.9
MULTIPLE LINEAR REGRESSION
ALL REPAIRS
N
28
117
78
27
93
11
68
8
21
179
-
A HC
0.27
0.98
0.23
-0.16
0.30
-1.25
1.16
0.75
0.53
0.62
-
t- ratio
0.64
5.12
0.95
-0.40
1.20
-2.01
4.51
1.07
1.13
4.03
-
A CO
1.4
12.8
-0.7
-3.6
4.5
-6.5
8.0
10.7
5.9
17.8
-
t- ratio
0.28
5.70
-0.26
-0.78
1.53
-0.89
2.65
1.31
1.08
9.78
-
A stepwise regression was then performed, successively
eliminating the system with the lowest t-ratio (correlation
coefficient relative to standard error) . This method
eliminated the systems which had the least statistically
significant repairs; that is, the systems whose repairs were
the least useful at explaining an emission reduction were
eliminated. The systems with the most significant repair
benefits were the fuel metering, exhaust, and three-way
systems. For these, the two approaches give similar
reduction estimates, as shown in the following table. See
Appendix F for breakdowns by quota group.
TABLE 19
Emission Reductions per Repair to Statistically Significant
Systems
SYSTEM
REPAIRED
Fuel Meter
Exhaust
3-Way
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N A HC A CO
94 1.07 13.2
60 1.20 7.3
156 0.55 17.8
MULTIPLE LINEAR REGRESSION
ALL REPAIRS OF THESE SYSTEMS
N A HC t-ratlo A CO t-ratlo
117 1.01 5.30 13.0 5.85
68 1.20 4.68 8.4 2.80
179 0.64 4.16 17.9 9.89
66
-------
These statistically significant systems are highlighted
in dark grey in the following graphs of average emission
reductions for isolatable repairs. The light grey bars must
be taken cautiously, since the values that created these
averages are highly variable. See Appendix E for individual
figures per quota group.
FIGURE 31
Average HC Reductions per Isolatable System Repair
1.5 T
HC
g/mi
-1.5
-3 J-
INDT FUEL IGNT EGR AIR
EXH EVAP ENG 3WAY ALL
FIGURE 32
Average CO Reductions per Isolatable System Repair
20 T
CO
g/mi
10--
-10 -L
INDT FUEL IGNT EGR AIR
EXH EVAP ENG 3WAY ALL
67
-------
In general, calculating emission repair reductions at
the system level, as just done, is not illuminating, since
the repairs varied widely within the general system category.
An examination by subsystem appears in Section 5.3.3 below.
Nevertheless, some additional information can be drawn
from the analysis at the system level. As clearly shown in
Figure 29, repairs to those systems with the least
variability in FTP repair reductions were much more effective
than others at consistently getting I/M failing vehicles to
pass the I/M test. For instance, exhaust system repairs were
effective 89% of the time, due almost entirely to catalytic
converter replacements. The excellent converter efficiency
of brand new catalysts may be responsible, and this result
should be considered cautiously. Repairs to the fuel
metering and three-way systems turned I/M fails into I/M
passes 66 and 64% of the time, respectively. Fuel system
repairs usually entailed replacing the carburetor or fuel
injectors (nearly all injector replacements were on a single
basic engine model) or tuning the system (largely on
carbureted vehicles) -- that is, adjusting idle mix, idle
speed and/or initial timing . Three-way system repairs were
mostly oxygen sensor or, less frequently, ECU replacements.
FIGURE 33
I/M Pass Rates Due to System Repair bv Quota Group
Cart 81-82 9 Carb 83-86 H Fl 81-82 D Fl 83-86
EXH
FUEL
SWAY
IGNTT
Repairs to those systems with greater variability in FTP
repair reductions were, at the same time, less consistent in
reducing the I/M failure rate. Air injection system repairs,
effective at eliminating I/M failures 52% of the time, varied
among various valve replacements and repairs to the pump
68
-------
assembly, while repairs of the ignition system were
overwhelmingly tune-ups, and succeeding in getting failing
vehicles to pass I/M only 30% of the time. The missing
systems -- induction, EGR, PCV, evap, and engine -- had too
few vehicles failing I/M at the time of the repair to include
in the analysis. In general, quota group had little impact
on the effectiveness of a certain repair. Figure 33 and the
values cited above include only isolatable repairs. Missing
data indicates small sample size, rather than 0%
effectiveness.
5.3.3 Emission Benefits per Subsystem Repair
An analysis by subsystem is essentially an analysis by
component or component group. Results at this finer level of
detail can be used to better pinpoint those specific
components that have the greatest impact on emissions. The
same technique used above -- simple averages of emission
reductions for isolatable repairs, coupled with a stepwise
multiple linear regression for all repairs was repeated,
this time using subsystems as variables.
Many subsystems were eliminated from the results due to
a low occurrence of repairs. This is presumably because
these components were not often diagnosed as emission control
problems in need of repair, either because they were, in
fact, not in need of repair, or because their malfunction was
judged to not significantly affect emissions (see Appendix G
for a count of repairs per subsystem) . Of the remaining
subsystems, a step-wise multiple regression yielded seven
with statistically significant emission reductions due to
their repair. Those with more than seven cases and a t-ratio
greater than 2.0 for one or both pollutants are considered
significant.
Table 20 lists the simple averages and the results of
this step-wise regression. As with repairs categorized by
system, the most consistently effective repair types were to
the fuel system, the exhaust system mainly the catalyst
and the electronic controls for the three-way system. Not
surprisingly, some of the most important emission control
components -- the catalyst and oxygen sensor -- are
consistently effective at cleaning up both HC and CO
emissions.
Note that, of the seven repair types, five were
effective for both HC and CO; two were consistently effective
on CO only (the ECU and load sensor) . Recall from section
4.2.3 that most vehicles that were high emitters on one
pollutant were also high on the other, so that repairs were
often targeted, effectively so, at reducing both HC and CO.
All emission values in the table are in grams per mile, as
measured by the LA4. See Appendix G for results for all
69
-------
subsystems, and Appendix H for breakdowns
statistically significant subsystems by quota group.
of the
TABLE 20
Emission Reductions
Subsystem Rf»pa ] r
SUBSYSTEM
REPAIRED
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
O2 Sensor
Load Sensor
All
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
22
30
19
43
14
69
11
372
A HC
1.03
0.61
2.35
1.11
0.40
0.80
0.61
0.70
A CO
11.9
11.6
24.2
7.0
10.7
20.7
23.2
10.7
MULTIPLE
ALL REPAIRS
N
27
43
20
56
19
82
22
-
A HC
1.02
0.63
2.22
1.20
0.67
0.94
-0.27
-
LINEAR REGRESSION
OF THESE SUBSYSTEMS
t-ratlo
2.7
2.1
5.1
4.7
1.5
4.4
-0.7
-
A CO
12.6
11.8
22.5
8.5
13.4
22.9
11.2
-
t-ratlo
2.9
3.4
4.4
2.8
2.5
9.0
2.3
-
The method of repair which most often occurred on these
seven important subsystems was replacement of the main
component carburetor, fuel injectors, catalytic converter,
oxygen sensor, load sensor (manifold air pressure or throttle
position sensor), or electronic control unit. The other
frequent subsystem repair was a fuel metering tune-up, which
included adjustments to idle speed, idle mix, and/or initial
timing.
A comparison of the simple averages with the regression
correlation coefficients reveals reasonably consistent
results, except for one case -- the load sensor. This
subsystem yields an average CO reduction of 23.2 g/mi when
repairs limited to the load sensor are averaged, in contrast
to an 11.2 g/mi reduction projected by the regression.
Further investigation reveals that when all repairs to the
load sensor are averaged, including those lumped with other
repairs, the average reduction drops to 11.2 g/mi, identical
to that predicted by the regression. Of the 22 load sensor
repairs, 11 were lumped with another repair hoses, lines,
and wires; these were all Ford vehicles undergoing a recall
procedure to clean out the line leading to the map sensor,
regardless of a diagnosis indicating its necessity. These 11
repairs actually increased CO emissions by an average of 0.9
g/mi, presumably because they were often unnecessary, while
the 11 isolatable repairs decreased CO by 23.2 g/mi, on
average. This is an unusual instance, in which two repair
types were repeatedly performed together, with consistent
emission effects. In this case, the multiple regression was
not able to separate out the effects of the single repair
type. In most other cases for which we have a reasonable
70
-------
sample size, the multiple regression tracked the averages of
isolatable repairs closely.
Figures 34 and
from the preceding
subsystems that have
that pollutant. The
just the seven major
See Appendix H for
group.
35 chart the average emission reductions
table. Dark columns indicate those
statistically significant reductions for
average reduction for all repairs -- not
ones is also included in the figures.
breakdowns of these figures by quota
FIGURE 34
Avera.de HC Reduction per Isolatable Subsystem Repair
Cart FuelMtr Fuel Catalyst ECU 02 Load
Tune Inject Sensor Sensor
ALL
FIGURE 35
Average CO Reduction per Isolatable Subsystem Repair
Caib FuelMtr Fuel Catalyst
Tune Inject
ECU 02 Load
Sensor Sensor
ALL
71
-------
Many of
effective at
illustrates
subsystems at
test.
the same seven subsystems are consistently
reducing the I/M failure rate. Figure 36
the effectiveness of repairs to certain
allowing an I/M failing vehicle to pass the I/M
FIGURE 36
I/M Pass Rates Due to Subsystem Reoair
Catalyst Carburetor
Oxygen Fuel Meter Fuel Injector Ignition Tune
Sensor Tune
Catalytic converter replacements were effective in this
task 100% of the time, while replacement of the carburetor or
fuel injectors, fuel metering system tune-ups, and oxygen
replacement sensor replacements allowed I/M fails to pass
approximately 75% of the time. On the other hand, ignition
system tune-ups (i.e., spark plug or plug wire replacements
or initial timing adjustments), which did not appear earlier
as statistically significant, were effective only 20% of the
time at turning I/M fails into I/M passes. The remaining
subsystems had repairs on I/M failing vehicles too
infrequently to be included in the analysis.
Figure 37 presents the same data split according to
quota group. Note that missing data indicates small sample
size (fewer than five cases), rather than 0% effectiveness.
Both figures and the values cited in the previous paragraph
include only isolatable repairs.
72
-------
FIGURE 37
T/M Pass Rates DIIP to Subsystem Repair hv Quota Groun
Garb 81-82 M Cart 83-86 H Fi 81-82 D Fl 83-86
Catalyst Carburetor
Oxygen
Sensor
Fuel Meter
Tune
Fuel Injector Ignition Tune
5.3.4 Total Benefit per Subsystem
Many of these same subsystem repair types not only are
consistently effective at reducing emissions on individual
vehicles, but also contribute greatly to the total emission
reduction of all repairs in the CTP. Figures 38 and 39 show
the subsystem repairs that contributed greater than 5% of the
overall CTP repair reduction. These values were derived by
multiplying the average emission reduction per quota group
for a subsystem repair type (calculated with isolatable
repairs only) by the number of times a repair occurred to
that subsystem in that quota group (all occurrences
isolatable or not). This estimate of the total contribution
of that subsystem was then divided by the total benefit
realized by all repairs, generating percent contribution per
subsystem. See Appendix I for a listing of results for all
subsystems; this table is not stratified by quota. Note that
the totals do not equal 100%, due to the combining of
isolatable averages with all repair occurrences.
The most important repair at reducing fleet emissions
was the oxygen sensor, for both HC and CO. This was not only
the most frequently repaired component, being replaced on 1/3
of the repaired vehicles, but also contributed a fairly large
reduction when replaced 0.80 g/mi HC and 20.7 g/mi CO.
The catalyst was also very important, being replaced on 30%
73
-------
of the repaired vehicles. It was even more effective than
the oxygen sensor at reducing HC per vehicle, at l.li g/mi,
but only about 1/3 as effective at reducing CO.
FIGURE 38
Contribution of Subsystems to Tntal HC Repair R»n«»fit-
bv Quota Group
Garb 81-82 9 Garb 83-86 H Fl 81-82 D Fl 83-86
O2 Catalyst Fuel
Sensor Injector
Garb Fuel Fuel
Meter System
Tune Other
Ignition Air Pump
Tune
FIGURE 39
Contribution of Subsystems to Total CO Repair Benefit
bv Quota Group
%
CO
35% -r
30% -
25%
20% .
15%-.
10%--
5% -
0% - -
Garb 81-82 H Cart 83-86 El Fl 81-82 D Fl 83-86
02
Sensor
Fuel
Injector
Fuel Meter
Tune
Load
Sensor
Catalyst Carb
Fuel
System
Other
74
-------
Fuel system repairs of many kinds, including fuel
injector replacements, catalyst replacements, tune ups, and
miscellaneous repairs to other fuel system components were
also extremely effective at reducing fleet emissions of both
HC and CO. The fuel meter tune items idle speed, idle
mix, and initial timing adjustments -- had an impact based
largely on their frequency of occurrence, being performed on
almost 25% of the repaired vehicles, but with a per-vehicles
reduction of only 0.61 g/mi HC and 11.6 g/mi CO. Fuel
injector replacements occurred somewhat less frequently, on
only about 10% of the repaired vehicles, but had extremely
high levels of reduction per repair, at 2.35 g/mi HC and 24.2
g/mi CO. Carburetors were replaced about as often as fuel
injectors, but were only about half as effective per repair.
Repairs to other fuel system components were not
particularly frequent, occurring on only 6% of the repaired
vehicles, but they had extremely high average reductions per
repair, on the order of those seen for fuel injectors. About
half of these repairs to miscellaneous fuel system components
had a negative or negligible emissions benefit, including
carburetor mixture adjustments, cleaning deposits from the
throttle body, and repairing the accelerator pump or linkage;
those repairs that were effective consisted of replacing the
air cleaner vacuum line, adjusting the float level and M/C
solenoid, and replacing the jet mixture solenoid.
The air pump and ignition tune-ups both had large
impacts on the overall HC reduction, but for opposite
reasons. The air pump was repaired -- invariably this
involved replacement on only 3% of the repaired vehicles,
but had very large HC reductions upon repair, averaging 7.9
g/mi. Its repair also resulted in very large CO reductions
of 41.1 g/mi, but not large enough to overcome the small
frequency of occurrence. Ignition tune items, on the other
hand spark plug or plug wire replacements or an initial
timing adjustment, followed by an idle speed adjustment if
needed were only marginally effective per repair, at 0.34
g/mi HC reduction, but their frequency performed on 30%
of all repaired vehicles caused them to have a large HC
impact overall.
5.3.5 Effect of Deteriorated Catalysts on Emissions
Because of the significant role catalytic converters
play in emission control and because of their susceptibility
to damage through tampering and misfueling, an analysis was
conducted focusing on their role in emission levels when
malfunctioning. In the CTP fleet, 55 catalysts were replaced
on 53 vehicles (two vehicles had both an oxidation and a
three-way catalyst replaced), accounting for 11% of the RMs
that occurred. Table 21 details the emission reductions for
catalyst replacements and for all other RM types. Catalyst
75
-------
replacements eliminated slightly less than 20% of the CTP
fleet's excess HC, and slightly less than 9% of its excess
CO. Percent reductions in emission levels per RM were
significantly higher for catalyst replacements than for the
other RM types, even though, since catalyst repairs were
generally withheld until the last repair, the pre-RM emission
levels were lower on average at the time they occurred.
TABLE 21
LA4 Emission Reductions Catalyst Replacements and All
Other Repairs
HC
catalyst
other
all
number of
RMs
55
413
479
average
pre-RM
emissions
1.38
1.82
1.74
average
reduction
per RM
1.23
0.62
0.67
percent
reduction
per RM
89.0%
33.8%
38.5%
total
reduction
67.48
254.74
321.91
% total
excess
reduced
19.6%
74.0%
93.5%
CO
catalyst
other
all
number of
RMs
55
413
479
average
pre-RM
emissions
10.9
23.0
21.3
average
reduction
per RM
8.5
11.1
10.6
percent
reduction
per RM
78.1%
48.1%
49.5%
total
reduction
468.5
4575.3
5063.0
note: 55+413=468; 10 of the missing repairs were post -catalyst replacement, excluded
due to potential masking effect of new catalyst on subsequent repair reductions; 1
catalyst diagnostic rather than repair
% total
excess
reduced
8.9%
87.1%
96.4%
was
The correctness of the diagnosis that a particular
catalyst was malfunctioning is important to this analysis.
Most vehicles -- 95% -- were released from "the test program
with normal emission levels, suggesting that those that did
not receive catalyst replacements probably did not require
them. Also, catalyst replacements were normally withheld
until all other repair options were exhausted. Therefore, it
can be assumed that catalysts repairs were generally applied
only when needed, and avoided when not.
Evidence of misfueling or tampering did not play a
significant role in identifying vehicles which required a
catalyst replacement to achieve normal emission levels. For
catalyst replacement vehicles, the average lead level in the
as-received tank fuel was 0.0037 g/gal; for all CTP vehicles,
it was 38% higher, at 0.0051 g/gal. Of the 81 vehicles with
above-average lead-in-fuel levels, only 9, or 11%, received
catalyst replacements; 84% of the catalyst changes occurred
on vehicles with below-average lead-in-fuel levels. Also, of
the ten vehicles with the highest lead-in-fuel levels (0.015
76
-------
0.05 g/gal), only two received catalyst changes. Thus,
fuel lead level is not a reliable predictor of the necessity
of a catalyst replacement.
Additionally, only two of the 53 catalyst replacement
vehicles were noted to have signs of fuel inlet restrictor
damage. Two vehicles which did not receive cat replacements
also had such damage; one was nevertheless released from the
CTP with normal emitter levels achieved via other repairs,
while the other remained a high emitter due to obvious
tampering -- a missing catalyst. This vehicle certainly
would have had drastic emission reductions had a new catalyst
been installed. Thus, evidence of fuel inlet restrictor
tampering may be a reliable sign of the necessity of catalyst
replacement, but occurs infrequently.
Two-thirds of the as-received fleet underwent the
Plumbtesmo test for tailpipe lead residues; only three
failures occurred, and none of these required catalyst
changes to achieve normal emission levels. Overall, only 20%
of the vehicles that had their catalysts replaced had
evidence of misfueling or tampering, as indicated by the
Plumbtesmo test, fuel inlet restrictor damage, or above-
average fuel lead levels.
5.4 Analysis of Incremental Repairs
5.4.1 Sample Description
The focus will now move from a discussion of the repair
types that affected emission levels to the benefits actually
achieved under different circumstances of test procedure,
vehicle emitter category and repair target.
Recall that 630 repairs were performed in 479 remedial
maintenance (RM) steps on 184 vehicles. The sequence of the
RM steps on each vehicle was based on the as-received vehicle
characterization, with the repairs judged most likely to
reduce FTP emissions performed first. Once FTP values were
sufficiently low, any vehicle that still had difficulty
consistently passing the I/M test was repaired to eliminate
that problem. Transient and I/M tests performed both before
and after each repair allow a comparison of repairs that
helped the vehicle pass I/M with those that actually cleaned
up the vehicle, as measured by the FTP or LA4.
Recall from Section 4.5.2 that the core sampling period
of the BITP which followed extended loaded preconditioning
had the lowest failure rate, with approximately 40% of the
CTP sample failing. This is considered to be the BITP
sampling period which is closest to the "ideal" I/M test
condition. Analysis of the effects of repair on the I/M test
77
-------
focuses exclusively on this sampling period. The following
sections focus exclusively on non-catalyst repairs.
5.4.2 Benefits of Repairing to Pass I/M
The purpose of I/M is to determine which vehicles have
high emission levels, so that their emissions can then be
reduced through repair; an adequate repair should not only
allow a vehicle to pass a subsequent I/M test, but also clean
up its actual in-use emissions. It is therefore important to
determine if those repairs that allow an I/M failing vehicle
to pass are also the ones that actually clean up the
emissions, as measured by a mass emissions test. This
section investigates the mass emissions benefits that are
realized by repairing vehicles to pass the I/M test.
One hundred vehicles 54% of those that received
repair were failing their ideal I/M test at the time of
first repair. Eighty-four of these vehicles received a non-
catalyst repair at some point that allowed a passing I/M
score, while 11 required a catalyst replacement to pass I/M,
three never passed the ideal I/M test (these three never
received catalyst repairs, and were high or marginal emitters
at the time of release) , and the remaining two had no post-
repair I/M test data. Overall, it took an average of 1.5
remedial maintenance steps to get a failing vehicle to pass
I/M.
Both I/M scores and mass emissions data are available
for the non-catalyst repairs on 83 of the 84 vehicles which
were initially failing I/M. The net LA4-measured emissions
benefit achieved from these RMs those that first allowed
an I/M pass without a catalyst change was 178 g/mi HC and
2651 g/mi CO. Average reductions per RM per vehicle were
therefore 2.14 g/mi HC and 31.9 g/mi CO, over twice as large
as the average reduction per RM for all repairs. This single
RM step reduced as-received HC and CO emissions by about 75%
for these 83 vehicles, eliminating approximately 80% of their
as-received excess. This high level of reduction on these
vehicles eliminated over 50% of the excess HC and CO
emissions of the entire CTP fleet and achieved close to 60%
of the entire reduction seen by that fleet, although the
repairs occurred on only 47% of the vehicles, and represented
only 20% of the non-catalyst RMs performed. Additionally,
the number of high and super emitters in this group dropped
from 83% to 25% due to this single RM. Therefore, the
repairs that worked in terms of I/M pass/fail were also
apparently well suited to reduce FTP emissions.
Table 22 gives values for these vehicles at their first
I/M pass, broken down by the emitter categories before and
after the repair that caused the passing test. The reduction
is that caused by the single RM that caused the vehicle to go
78
-------
from I/M fail to I/M pass; mean emissions are those after the
repair that is, at first I/M pass. Percent excess is of
failed vehicles in the entire CTP fleet.
TABLE 22
FTP Benefits of Repairing to Pass I/M
MOBILE4
Emitter
Category
Normal to Normal
Normal to High
High to Normal
High to High
Super to Normal
Super to High
TOTAL
Number
of
Vehicles
12
2
49
18
1
1
83
Mean LA4
Reduction
HC CO
0.03 1.4
-0.64 -15.5
2.38 47.2
1.74 12.9
15.69 62.9
15.12 57.2
2.14 31.9
% Of FTP
Excess
HC CO
0.1% 0.3%
-0.4% -0.6%
33.9% 44.1%
9.1% 4.4%
4.6% 1 .2%
4.4% 1.1%
51.7% 50.5%
Mean LA4
Emissions
HC CO
0.39 4.5
1.31 19.1
0.37 4.2
1.96 27.2
0.28 2.2
3.11 8.2
0.77 9.6
The FTP benefits of repairing to pass I/M were not
dependent on fuel meter type or model year group. As Figure
40 illustrates, about 60% of the vehicles changed from high
to normal emitters after the repair to passing I/M status,
independent of quota group.
FIGURE 40
Changes in Emitter Group Due to Repair from I/M Fail
Pass bv Quota Grouo
1-0 T/M
%0f
vehicles
High-High
H High-Norm
D Norm-High
1 Norm-Norm
FI81-82
Garb 81-82 Carb 83-86
Fl 83-86
ALL
79
-------
Manufacturers were slightly more variable in their
success, but this was to an extent due to the number of high
versus normal emitters in their original sample. Overall,
repairs that turned normal emitters into high emitters while
allowing the vehicle to pass I/M were quite rare
approximately 2%. Super emitters are included with highs in
the following two figures. Note that several manufacturers
are grouped together; this is due to their small sample size
(fewer than seven vehicles) in this subset of data.
FIGURE 41
Changes in Emitter Group Due to Repair from T/M Fail to I/M
Pass bv Manufacturer
%of
vehicles
High-High H High-Norm D Norm-High Norm-Norm
GM
FORD
NISS
AMC
VW
OTHER ALL
There is some additional FTP reduction available from
more complete repair even after vehicles pass I/M. The CTP
test sequence did not specifically address this issue, but
the program nevertheless collected data on 39 vehicles that
received additional repair and mass emission testing after
they were passing I/M. These vehicles achieved an additional
LA4 reduction of 0.33 g/mi HC and 9.5 g/mi CO, on average, as
shown in Figure 42. These reductions were achieved in an
average of two additional remedial maintenance steps per
vehicle.
80
-------
FIGURE 42
Average LA4 Emissions for Vehicles with Addi 1--j npal Repair
After Passing I/M
HC
g/mi
3.5 j
3 -
2.5
2
1.5
1
0.5
0 -I-
O-HC
-co
CO
g/mi
y 40
30
..20
-.10
as received
after first I/M pass
after final repair
5.4.3 Comparison to MOBILE4 Repair Estimates
The MOBILE4 emissions model uses emission benefits
realized from repairing failing I/M vehicles to pass the I/M
test as part of its calculations for I/M credits.21 The
values used are derived from test programs conducted by EPA
and the California Air Resources Board, in which vehicles
underwent the I/M process and failures were repaired by
either commercial garage mechanics or EPA contractors
instructed not to continue repairs past the point of passing
I/M. Table 23 below compares the MOBILE4 average repair
reductions for MY 80-86 vehicles with closed-loop control
which initially failed an idle test to CTP values for a
similar set of vehicles. (The MY 80 vehicles in the MOBILE4
dataset are California only, with technology similar to that
used on Federally certified MY 81 vehicles). All vehicles
included in these tables passed the I/M test following
repair. Reductions are calculated from as-received values,
and are those realized through the RM step which took the
vehicle from I/M failing to I/M passing status. CTP
reduction values are LA4-based; MOBILE4 values are FTP-based.
81
-------
TABLE 23
MQBIT.E4 vs C.TP Average Emissiog Redurfinni
Due to Renair to I/M Pass
Carbureted Vehicles
HC
Normal
High
Super
ALL
CTP
N
7
40
0
47
CO
Normal
High
Super
ALL
N
7
40
0
47
As-Rcvd
FTP LA4
1.33 1.11
3.50 3.30
3.18 2.97
Reduction
g/ml %
0.54 48.2%
2.46 74.6%
2.17 73.1%
As-Rcvd
FTP LA4
22.9 20.8
50.9 48.8
46.8 44.6
Reduction
g/ml %
12.1 58.4%
38.0 77.9%
34.1 76.5%
MOBILE4
N
38
53
9
100
As-Rcvd
FTP
0.76
2.86
13.81
3.05
Reduction
g/ml %
0.14 18.7%
1.46 51.1%
11.67 84.5%
1.88 61.6%
N
38
53
9
100
As-Rcvd
FTP
8.8
50.9
190.2
47.5
Reduction
g/ml %
1 .8 20.8%
29.0 57.0%
174.0 91.5%
31.7 66.9%
Fuel Injected Vehicles
HC
Normal
High
Super
ALL
CTP
N
7
27
2
36
CO
Normal
High
Super
ALL
N
7
27
2
36
As-Rcvd
FTP LA4
1.29 1.16
3.20 3.04
12.91 11.87
3.36 3.16
Reduction
g/ml %
0.69 59.9%
2.31 76.0%
10.18 85.7%
2.43 76.9%
As-Rcvd
FTP LA4
28.0 5.5
56.2 54.1
62.9 62.0
51.1 45.1
Reduction
g/ml %
1 .4 25.8%
44.3 82.0%
56.8 91.6%
36.7 81.4%
MOBILE4
N
12
24
4
40
N
12
24
4
40
As-Rcvd
FTP
0.41
2.36
6.41
2.18
As-Rcvd
FTP
5.8
47.9
184.1
49.0
Reduction
g/ml %
0.08 20.0%
1.42 60.3%
4.48 69.9%
1 .33 60.9%
Reduction
g/ml %
1 .3 23.4%
32.7 68.3%
139.0 75.5%
33.9 69.5%
When vehicles in all emitter categories are combined,
the MOBILE4 percent reduction values undershoot those seen in
the CTP by 2-40%; this gap would probably be several percent
greater if the CTP values were FTP-based. The gram per mile
reduction used by MOBILE4 is also generally lower than that
seen in the CTP, partially due to lower as-received levels in
the MOBILE4 sample. It can be argued that the MOBILE4 values
are more realistic since many of the vehicles were repaired
in commercial facilities rather than emission laboratories.
On the other hand, the CTP values can be considered the level
of reduction that could be attained given improved mechanic
training in diagnosis and repair.
82
-------
5.4.4 Benefits of Repairing to pass the FTP
One hundred thirty-two vehicles -- 72% of those
receiving repair were high or super emitters at some point
during their repair cycle. Almost 3/4 of these -- 98
vehicles -- received a non-catalyst repair that turned them
into normal emitters, while 20% required a catalyst repair to
be cleaned up, and the remaining 6% never were repaired to
normal emitter levels, never having received a catalyst
replacement.
The net LA4-measured emissions benefit achieved from
these RMs the non-catalyst repairs that cleaned up a
vehicle to normal emitter levels was 202 g/mi HC and 3722
g/mi CO. Average reductions per RM per vehicle were
therefore 2.06 g/mi HC and 38.0 g/mi CO, almost three times
as large as the average reduction per RM for all repairs.
This single RM step reduced as-received HC and CO emissions
by about 85% for these 98 vehicles, eliminating over 90% of
their as-received excess. This very high level of reduction
on these vehicles eliminated approximately 60% of the excess
HC and 70% of the excess CO emissions of the entire CTP fleet
and achieved close to 2/3 of the entire reduction seen by
that fleet, although the repairs occurred on only 40% of the
vehicles, and represented only 20% of the RMs performed.
Ideal I/M tests were performed on most of these high or
super emitters. Of those that were cleaned up to normal
emitting levels, 57% became I/M passes after the repair.
Another 37% had previously been passing I/M and continued to
pass, while 6% continued to fail I/M even though they had
achieved normal emitter levels. Therefore, a total of 94% of
the vehicles that were repaired from high to normal levels on
a transient test could also, at that repair stage, pass an
I/M test performed under optimum conditions.
Table 24 gives values for these vehicles for the repair
which took them from high (or super) to normal emitter
status, broken down by the I/M pass/fail category before and
after the repair. The reduction is that caused by the single
RM that caused the vehicle to become a normal emitter; mean
emissions are those after the repair -- that is, for the
first time at normal levels. Percent excess is of FTP-failed
vehicles .in the entire CTP fleet.
83
-------
TABLE 24
I/M Benefits of Repairing fn FTP Normal EmiM-pr
Ideal I/M
Pass/Fall
Status
Pass to Pass
Pass to Fail
Fail to Pass
Fail to Fail
Unknown
TOTAL
Number
of
Vehicles
34
0
53
6
5
98
Mean LA4
Reduction
HC CO
1.01 24.4
2.75 46.9
0.99 19.4
3.18 58.5
2.06 38.0
% Of FTP
Excess
HC CO
9.9% 15.8%
42.4% 47.3%
1.7% 2.2%
4.6% 5.6%
58.7% 70.9%
Mean LA4
Emissions
HC CO
0.45 5.4
0.39 4.2
0.57 5.0
0.57 4.3
0.43 4.7
The change in I/M pass/fail status once a vehicle was
repaired to normal emitter levels was somewhat dependent on
model year and fuel metering system, but largely as a result
of variations in the I/M pass-fail levels prior to the
repair. Overall, getting vehicles to pass I/M once their FTP
or LA4 levels were low was not difficult for any quota group,
with 94% of the vehicles passing overall, and no quota group
doing worse than 88%.
FIGURE 43
Changes in I/M Pass/Fail Status Due to Repair from High to
Normal Emitter bv Ouota
%0f
Vehicles
Fail-Fai
Fail-Pass D Pass-Fail Pass-Pass
FI81-82
Cart 81-82 Cart 83-86
Fl 83-86
ALL
84
-------
Manufacturers were significantly more variable in their
I/M status changes, but again, this was almost entirely due
to the differences in pre-repair I/M status. GM and Ford had
the most difficulty in obtaining passing I/M scores when
vehicles were cleaned up on transient tests, with 14% and 9%
failure rates, respectively.
FIGURE 44
Chances in I/M Pass/Fail Status Due fo Repair from High to
Normal Emitter bv Manufacturer
%0f
Vehicles
Fail-Fai
Fail-Pass D Pass-Fail Pass-Pass
GM
FORD NISS CHRY AMC VW OTHER ALL
5.4.5 Emission Benefits "Lost" through Second-Chance
I/M tests do not have perfect pass/fail correlation with
the FTP. Vehicles that fail I/M with passing FTP emissions
are considered false failures; their repair is unnecessary
from a clean air standpoint and undesired from a consumer
cost and inconvenience standpoint. One strategy to reduce
the number of false I/M failures is to give all failing
vehicles a second-chance I/M test. Presumably, clean
vehicles that fail I/M due to inadequate preconditioning
and/or canister purge during the idle test would 'have a good
chance of passing an immediate second-chance test if preceded
by sufficient preconditioning, and would not have to be
repaired or retested later. Conversely, vehicles that are
truly dirty under normal operating conditions should, continue
to fail. However, the second-chance test still being a short
test, some dirty vehicles would pass along with those that
are clean; it is important to determine the potential repair
85
-------
benefits from these vehicles that would be lost by applying
the second chance test.
In the CTP, 138 vehicles almost 60% of the fleet
passed an ideal I/M test in as-received condition (52 highs,
53 marginals, and 33 passes) . Recall that all CTP vehicles
failed their field I/M test, so that this lab test can be
considered second-chance. We can assume that a fairly large
proportion of these CTP second-chance passes would have also
passed a second-chance test in the field, even under non-
ideal conditions. These vehicles would then not be repaired.
It would be interesting to calculate the total emissions
benefit due to repairing these vehicles that which would
be lost if they all passed second-chance. However, we do not
have repair data on all 138 of the vehicles, since a number
were released from the CTP without repair and others did not
receive mass emissions tests after repair. Mass emissions
data for post-repair tests (including catalyst repairs) is
available on 82 vehicles that passed second-chance (48 highs,
30 marginals, 4 passes). We can calculate the average
emission reduction per vehicle (from as-received to release),
with per-vehicle averages based on the as-received MOBILE4
emitter category. Summing these averages according to the
weighting of the emitter categories of the 138 vehicles
provides an estimate of the total emissions reduction that
would not be realized: 68 g/mi HC and 1211 g/mi CO, or
between 20 and 25% of the total LA4-based emission reduction
of the entire CTP fleet. This estimate is an upper bound,
since not all 138 vehicles would have passed"second-chance
given the non-ideal conditions in the field, and since the
CTP per-vehicle repair benefits are probably higher than
those in commercial facilities.
Thus, repairs to 60% of the CTP fleet provided less than
25% of the reduction, as an upper bound estimate. Over 60%
of these CTP second-chance passes were normal emitters (pass
or marginal) much higher than the general CTP fleet at 40%
normal emitters and therefore are not desirable candidates
for repair. This supports the theory that second-chance
tests can reduce the incidence of unnecessary or detrimental
repairs to cleaner vehicles without greatly reducing the
emission benefit due to repairs to those that are dirty.
5.4.6 Benefits of Repairing to Different Targets
"Repair benefit" can be defined many different ways,
based on the target which a vehicle is being repaired to
meet. I/M programs, of course, use a passing score on a I/M
test as the target; after this point is reached, there are no
further emission control repairs. The CTP database provides
information on the extent of excess emissions eliminated via
I/M-targetted repairs, and whether a substantial portion of
86
-------
"repairable" emissions remain after an I/M test is passed.
As the following table shows, repairs to I/M targets reduce
as-received emission levels by about 75%, eliminating over
half of the FTP excess. However, this is only about 2/3 of
the total reduction that can be realized with more complete
repair.
TABLE 25
Reduction in LA4 Emissions Due to Reoair to Different Taraets
TOTAL REDUCTION
(a/ml)
HC
to first I/M pass - non-cat
to first I/M pass ~ all
all non-catalyst repairs
all repairs
CO
to first I/M pass - non-cat
to first I/M pass - all
all non-catalyst repairs
all repairs
* vehicles
83
94
175
184
83
94
175
184
as-
received
253.56
281.59
367.51
382.99
3862.2
4115.7
5538.9
5665.9
after
repairs
63.83
65.87
112.77
61.08
797.1
820.8
963.5
603.0
reduction
189.73
215.71
254.74
321.91
3065.1
3294.8
4575.3
5063.0
%
reduction
74.8%
76.6%
69.3%
84.1%
79.4%
80.1%
82.6%
89.4%
% excess
reduced
55.1%
62.7%
74.0%
93.5%
58.4%
62.8%
87.1%
96.4%
AVERAGE REDUCTION
(a/ml)
HC
to first I/M pass -- non-cat
to first I/M pass - all
all non-catalyst repairs
all repairs
CO
to first I/M pass ~ non-cat
to first I/M pass - all
all non-catalyst repairs
all repairs
tt vehicles
83
94
175
184
83
94
175
184
as-
received
3.05
3.00
2.10
2.08
46.5
43.8
31.7
30.8
after
repairs
0.77
0.70
0.64
0.33
9.6
8.7
5.5
3.3
reduction
2.29
2.29
1.46
1.75
36.9
35.1
26.1
27.5
%
reduction
74.8%
76.6%
69.3%
84.1%
79.4%
80.1%
82.6%
89.4%
% excess
reduced
55.1%
62.7%
74.0%
93.5%
58.4%
62.8%
87.1%
96.4%
5.5 Repairs to High Emitters
5.5.1 Effectiveness of Repair on Marginals vs. Highs
It is important, first, to determine the effectiveness
of repairs on marginal emitters versus high emitters. Is it
worth it to capture and repair the marginals, or would the
effort be better spent focussed entirely on the highs? Sixty
percent of the CTP sample were high emitters (143 vehicles),
while 26% were marginals (62 vehicles) . As shown in Figures
45 and 46, the emissions benefit of non-catalyst repairs to
87
-------
the marginal emitters is negligible, whereas the high
emitters have substantial LA4 reductions 1.3 g/mi HC and
33 g/mi CO. In fact, the average emission reduction on high
emitters is over 15 times as great as that seen on marginals,
for both HC and CO. These figures include non-catalyst
repairs on only those vehicles that eventually received
repair (all of the highs, and three-fourths of the marginals)
and for which we have complete mass emissions data.
FIGURE 45
Average HC Repair Benefit Maraina^L :vs_JLLah_ Emi
> marginal -O- high
HC
g/mi
2.5
2
1.5
1
0.5
0
as received
after final repair
FIGURE 46
Average CO Repair Benefit Marginal vs High Emitters
CO
g/mi
* marginal ** high
as received
after final repair
88
-------
The following table simply tallies the number' of
vehicles which were dirtier on the LA4 after all repairs than
as-received, on at least one pollutant. The breakdown into
high and marginal emitters indicates that a much higher
percentage of the marginals than highs were dirtier after
repair, although catalyst changes helped clean up both
marginal and high emitters. Thus, repairs to marginal
emitters generally result in negligible repair benefits and
are much more likely to be detrimental than repairs to high
emitters; therefore, marginal emitters are not a worthy
target for I/M programs.
TABLE 26
Vehicles Dirtier After All Repairs
EMITTER
CATEGORY
High
Marginal
NON-CATALYST REPAIRS
N cleaner
123
26
N dirtier
11
11
% dirtier
8%
30%
ALL REPAIRS
N cleaner | N dirtier
137 5
31 7
% dirtier
4%
18%
5.5.2 Benefit of Repairing Highs only
We now focus on high emitters, as both the most
prevalent portion of the CTP sample and the most important
segment relative to emission reductions. This section
supplies FTP as well as' LA4 values, to provide data that can
be more easily compared to that from other programs.
FTP emission benefits in a single non-catalyst RM step
ranged from a reduction of more than 7 g/mi HC and 125 g/mi
CO at the high end (from installation of a new oxygen sensor
on a 1985 fuel injected Oldsmobile Firenza, and replacement
of the ECU and oxygen sensor on a 1984 fuel injected
Chevrolet Cavalier), to an emissions increase of 8.6 g/mi HC
and 30 g/mi CO (from replacement of a PCV fitting on a 1982
carbureted Mercury Marquis). As Figure 47 shows, a large
reduction in one pollutant did not necessarily correlate with
a large reduction in the other, although there were
relatively few cases in which one pollutant increased while
the other decreased.
Overall, the average FTP benefit of non-catalyst repairs
to high emitters, per RM, was 0.83 g/mi HC and 15.8 g/mi CO.
This reduced HC by 41% and CO by 54% from the levels
immediately prior to the RM, and eliminated 31% of the
vehicle's as-received HC and 41% of its as-received CO, on
average. Each high emitter received an average of 2.3 non-
89
-------
catalyst RMs, and eventually had 83% and 92% of its excess HC
and CO eliminated. This reduced each high emitter's as-
received emissions by 69% for HC and 81% for CO, or a total
of 1.86 g/mi HC and 34.5 g/mi CO, from average levels of 2.70
g/mi HC and 42.4 g/mi CO.
With the inclusion of catalyst repairs, 98% of the
excess HC and 99% of the excess CO from the high emitters was
eliminated, in an average of 2.6 RMs per vehicle. This
entailed reducing each high emitter's as-received FTP
emissions by 81% for HC and 87% for CO, or 2.19 g/mi HC and
35.3 g/mi CO, from average levels of 2.69 g/mi HC and 40.6
g/mi CO.
FIGURE 47
FTP Benefit per RM of Non-Catalyst Renair to Hiah Emitters
CO
reduction
(g/mi)
HC reduction (g/mi)
The average LA4 benefit of non-catalyst repairs to high
emitters, .per RM, was 0.71 g/mi HC and 13.4 g/mi CO. This
reduced HC by 36% and CO by 50% from the levels immediately
prior to the RM, and eliminated 28% of the vehicle's as-
received HC and 34% of its as-received CO, on average. Each
high emitter received an average of 2.5 non-catalyst RMs, and
90
-------
eventually had 78% and 90% of its excess HC and CO
eliminated. This reduced each high emitter's as-received
emissions by 71% for HC and 85% for CO, or a total of 1.77
g/mi HC and 33.4 g/mi CO, from average levels of 2.50 g/mi HC
and 39.5 g/mi CO.
With the inclusion of catalyst repairs, 94% of the
excess HC and 97% of the excess CO from the high emitters was
eliminated, in an average of 2.8 RMs per vehicle. This
entailed reducing each high emitter's as-received LA4
emissions by 86% for HC and 91% for CO, or a total of 2.11
g/mi HC and 34.7 g/mi CO, from average levels of 2.46 g/mi HC
and 38.1 g/mi CO.
5.5.3 Catalyst Repairs Performed on Highs
Of the 143 vehicles that were high emitters as-received,
48 eventually received catalyst replacements; mass emissions
data is available on all but one. These 47 vehicles had
achieved, on average, an LA4-based emission reduction of 0.95
g/mi HC and 15.7 g/mi CO prior to replacement of the
catalyst, reducing HC emissions 40% and CO 58% from as-
received levels. These vehicles required an average of 2.9
non-catalyst RM steps to achieve these relatively small
reductions. A single catalyst repair, on the other hand,
allowed an additional 1.29 g/mi HC and 8.8 g/mi CO,
eliminating an additional 54% and 33%, respectively, of the
vehicle's as-received emissions. This resulted in an total
reduction of 95% of the vehicle's as-received HC and CO,
bringing levels lower than certification requirements, thus
eliminating the entire excess for the vehicle.
. Vehicles that never received a catalyst repair, however,
had total reductions of 81% HC and 91% CO, in only 2.1 RMs,
partly as a result of higher initial emission levels. This
eliminated 90% of the excess HC and 95% of the excess CO for
these vehicles. Non-catalyst repair vehicles never attained
emission levels as low as catalyst-replacement vehicles, with
final values approximately twice as high. Nevertheless, most
of these vehicles were brought to normal emitter levels even
without the benefit of a new catalyst, while over half of the
catalyst-repair vehicles were still high emitters until the
catalyst was replaced, despite almost three attempts at
repairing other components. In all, over 15% of the high
emitters could not be brought to normal levels by any means
of repair other than catalyst replacement.
The following two figures and table illustrate the
emission levels as-received, following all non-catalyst
repairs, and finally after catalyst replacements, for
vehicles that had no catalyst replacement and those that did.
Only vehicles that were high emitters as-received are shown.
Values are LA4-based.
91
-------
FIGURE 48
Average HC
of
to Hiyh
Catalvsf, vs Other Repairs
cat repair vehicles -o- non-cat repair vehicles
HC
g/mi
3
2.5
2
1.5
1
0.5
0
as-received after final non- after cat
cat repair repair
FIGURE 49
Average CO Benefit: of Repair to Hioh Emi'
Catalyst vs Other Repairs
cat repair vehicles
non-cat repair vehicles
CO
g/mi
as-received after final non- after cat
cat repair repair
92
-------
[ABLE 27
Average Benefit of Repair to High Emitters
Catalyst vs Other Repairs
HC
cat repair vehicles
non-cat rapalr vahs
CO
cat rapalr vehicles
non-cat repair vena
number of
vehicles
47
95
number of
vehicles
47
95
as-received
emissions
2.40
2.50
as-received
emissions
27.1
43.9
after final
non-cat repair
1.45
0.47
after final
non-cat repair
11.4
4.5
after
cat repair
0.16
after
cat repair
2.6
Catalyst replacement was the second most frequent repair
performed on high emitters, done on 1/3 of them. The most
common was replacement of the oxygen sensor, performed on 43%
of the high emitters. This repair was highly successful at
reducing emissions, at an average LA4 reduction for the highs
of 1.23 g/mi HC and 10.1 g/mi CO, eliminating almost 1/2 of
the excess HC and 1/4 of the excess CO for the affected
vehicles. Other repairs done frequently on high emitters
included various ignition tune items and fuel metering tune
items, and repair or replacement of carburetor assemblies,
air injection system check valves and hoses, fuel injection
components, and three-way control system components such as
load sensors, hoses, and the ECU. Repairs to the induction,
EGR, PCV, evaporative, and engine assembly systems were
relatively infrequent.
5.6 Difficulty of Repair
5.6.1 Difficulty of Repair to Passing Levels
Another aspect of vehicle repair is the difficulty of
diagnosing and performing the repair (s) that will actually
reduce emissions. In the CTP, diagnosed problems were ranked
according to their likely impact on FTP levels and performed
in that order. Therefore, the earlier RMs should have had a
greater emissions benefit than those performed later. If
this holds true, we can assume that diagnosis was generally
correct, and therefore was not a major inhibiting factor in
reducing FTP levels. The remainder of this section
investigates this issue.
The following table and figure indicate, for each repair
step, the vehicles that changed MOBILE4 emitter status at
each RM stage. The "high" grouping includes high and super
emitters, while the "norm" grouping includes passes and
marginals, as previously defined. All values are based on
93
-------
the LA4, and only repairs performed prior to catalyst
replacement are included. Some totals are not equal to 100%
due to occasionally missing data.
TABLE 28
Emitter Category Changes per RM
RM
Step
1
2
3
4
5
6
7
8
9
10
ALL
MOBILE4 Emitter Category Change
Norm-Norm
30%
45%
49%
52%
65%
63%
67%
75%
100%
-
42%
Norm-High
0%
5%
3%
0%
0%
0%
17%
0%
0%
-
2%
High-Norm
37%
18%
15%
14%
10%
25%
0%
25%
0%
-
24%
High-High
32%
30%
31%
34%
25%
13%
17%
0%
0%
-
30%
HIGH includes highs and supers
NORM includes passes and marginals
FIGURE 50
Emitter Category Chancres Due to Remedial Maintenance
High-High U High-Norm D Norm-High Norm-Norm
2345678
Remedial Maintenance Step
94
-------
In general, earlier repairs were more successful than
later repairs at reducing high emission levels to normal
levels, as anticipated. This is partly because, at later RM
steps, a greater percentage of vehicles had moved into the
normal emitter category prior to that repair, thus reducing
the number available to be cleaned from high to normal
levels. Additionally, however, those that were high emitters
at a later RM step were less likely than earlier repairs to
have substantially reduced emission levels (enough to drop
them into normal emitter status) due to that repair, dropping
from a 53% chance at the first repair to a 29% chance at the
fourth and fifth. The balance of these two effects is that
each time a car is repaired, there is a two-thirds chance it
will be at normal emitting levels after the repair.
It took approximately 1.5 RM steps, on average, to clean
up an FTP- or I/M-failing vehicle to passing levels. Figure
51 gives a breakdown of the repair step after which high or
super emitters became passing or marginal emitters, as
measured by the LA4, and the step after which I/M failures
became I/M passes.
FIGURE 51
RM Step that Cleaned Up Emissions
LA4 from High to Normal Emitter: I/M Score from Fail to Pass
%of
Vehicles
100 -r
90
80
70
60
50 -
40
30
20
10
0 4-
M 6th RM
Ei 5th RM
4th RM
3rdRM
H 2nd RM
D 1stRM
LA4
I/M
There is very little difference between the two cases;
in fact, the two bars represent many of the same vehicles
moving into passing status for both tests at the same RM
stage. There is a 57% overlap between the two bars; that is,
95
-------
of the vehicles that were either high/super emitters as-
received or were failing I/M as-received, 57% were both. Of
this overlapping group, 72% passed both test types at the
same RM step, 21% passed I/M while remaining high emitters,
and 7% became normal emitters prior to passing I/M.
Quota group had a minor impact on the number of repair
steps required to turn a vehicle from a high to normal
emitter. While the MY 81-82 vehicles had nearly identical
results regardless of fuel metering type, the MY 83-86 group
showed a marked difference between carbureted and fuel
injected models. Carbureted vehicles were repaired to normal
emitter levels in only one repair 81% of the time, whereas
only 53% of the fuel injected vehicles were successfully
repaired in a single RM step. Nevertheless, each quota group
was able to achieve a success rate of 80-90% after only two
RMs, as illustrated in Figure 52.
FIGURE 52
RM Step that Cleaned Up Emissions
LA4 from High to Normal Emitter bv Quota Grouo
%of
Vehicles
100 -r
90 -
80
70
60--
50
40
30
20 -
10
0
M 6th RM
E3 5th RM
B 4th RM
M 3rd RM
Q 2nd RM
DlStRM
FI81-82 Garb 81 -82 Garb 83-86 Fl 83-86
ALL
Manufacturers were varied in their success at reducing
high emitters to normal emitters in the first repair, with
Ford and AMC having the most trouble a success rate of
only slightly over 40% and the "other" group Toyota,
Mazda, Subaru, Mitsubishi, and Honda successful 100% of
the time. (These manufacturers were grouped to achieve a
sufficient sample size). Nevertheless, at least 85% of the
high emitters had been cleaned to normal emitting levels by
the third repair, regardless of manufacturer.
96
-------
FIGURE 53
RM Step that Cleaned rip Emissions --
LA4 from High to Normal E'm icr.gr -- bv
%0f
Vehicles
100
90
SO
70
60
50
40
30
20
10
0
D 6th RM
E9 5th RM
4th RM
3rd RM
13 2nd RM
QlstRM
GM FORD NISS CHR AMC VW OTHER ALL
5.6.2 Effectiveness of Repair at Successive RM Steps
The emission benefits of each repair also decreased, in
general, as the number of RMs increased, as shown in Figures
54 and 55. Step four is out of line with the trend,
particularly for HC. This is due to a single repair on the
super emitter, which accounts for 2/3 of the total reduction
for that repair step. When this vehicle is eliminated, the
total HC reduction plummets from 23.4 to 8.2 g/mi, and the
average drops from 0.81 g/mi to 0.29 g/mi, which is
consistent with the trend across all RM steps.
Similarly, four vehicles had a large impact on the
excessive reduction in CO seen in repair step four. These
four vehicles had an average reduction of 42.6 g/mi, while
the remaining 25 had an average reduction of 3.2 g/mi. The
repairs to these four vehicles were dissimilar (rerouting
lines to the air bypass and diverter valves, or replacing the
carburetor, fuel injectors or air pump), with no apparent
reason for such large reductions all occurring at RM step
four. Elimination of them from the analysis allows the
general trend to become more clear. The undue impact of a
few vehicles can be attributed to the reduction in sample
size at later repairs; this also affects steps five through
ten.
97
-------
FIGURE 54
Total and Average HC Redurr i nns per RM
Avg HC
AVERAGE Reduction
10
Remedial Maintenance Step
FIGURE 55
Total and Average CO Reductions per RM Step
Total CO
Reduction
(g/mi)
2500
2000
TOTAL
Avg CO
AVERAGE Reduction
4567
Remedial Maintenance Step
10
In general, the early repairs were quite successful at
reducing emissions, with the first repair generating an
98
-------
average reduction five times greater than the third repair,
despite the fact that each RM step usually included a repair
to only a single component. This is encouraging, in that it
indicates both that technicians were quite successful in
identifying the required repair, and that a single
malfunctioning component, rather than a complex set of
problems, is often the cause of high emissions.
Another way to approach the issue of repair difficulty
is to look at the number of RM steps required to reduce high
emission levels by a certain percentage. The following cable
lists the first RM step in which the total emission reduction
to that point exceeded 80%, from original levels of HC>2 or
C0>20 g/mi. The average number of repair steps to achieve
this reduction was 1.8 Quota group was not a factor in the
ability to reduce high emissions by 80%, as illustrated in
Figure 52.
Again, however, manufacturer had an effect, with the
seven Subaru and two Mitsubishi vehicles at these emission
levels reduced by 80% in a single repair, while the three
Toyota vehicles never achieved this reduction even after
multiple repairs. GM and VW were never able to reduce 40% of
their vehicles by this amount, while Ford and Chrysler were
unsuccessful about 20% of the time. Overall, approximately
75% of the vehicles with HC>2 or C0>20 g/mi eventually
received repairs which were able to reduce the high pollutant
by 80% or more.
TABLE 29
RMs Needed to Reduce LA4 Emissions by >8Q%
for Vehicles with FTP HC>2 or CQ>2Q a/mi
RM
step
1
2
3
4
5
6
7
8
9
10
ALL
Number with
>80% reduction
46
18
8
5
1
1
1
0
0
0
80
Total
Number
106
75
49
27
21
15
8
7
5
1
106
Percent with
>80% reduction
43.4%
24.0%
16.3%
18.5%
4.8%
6.7%
12.5%
0.0%
0.0%
0.0%
75.5%
99
-------
FIGURE 56
RM3
t-o
LA4 Emissions hy >8Q%
for VehielPs with FTP Hf>2 . 0 and/or CQ>20
bv Quota
50 j
40
Number 30 -
of
Vehicles 20"
10-
0--
Never
Di
FI81-82
Carb 81-82
Garb 83-86
Fl 83-86
FIGURE 57
RMs Needed to Reduce LA4 Emissions by >8Q% --
for Vehicles with FTP HC>2.Q and/or CQ>2Q g/mi
bv Manufacturer
Never I
>3
3
H2
Di
Number
of
Vehicles
GM FORD NISS TOYT CHRY AMC VW MAZ SUB A MITS MONO
The sum of these various approaches to the question of
diagnostic difficulty is that diagnosis was generally not an
impediment to emission reduction. Fewer than two RMs were
required, on average, to clean up an FTP- or I/M-failing
vehicle to passing levels and to reduce high emission levels
by 80% or more. This is despite the fact that each RM
normally included only a single repair. Additionally,
100
-------
earlier repairs were twice as likely as later repairs to turn
a high emitter into a normal emitter, and also had
substantially larger average emission reductions. Of course,
some vehicles were more difficult to diagnose and repair;
five percent of the repaired vehicles were released as high
emitters, never having achieved normal emitter status. Also,
15% of the repaired vehicles received more than four RM
steps, and 4% needed more than six. However, most of these
vehicles eventually had their catalysts replaced; the design
of the CTP to delay these repairs until all other options had
been exhausted contributed to the high number of RMs in most
of these cases.
101
-------
APPENDICES
102
-------
APPENDIX A: FAIttJR* RATZS IN TBX MICHIGAN AST PROGRAM
The following table provides initial-test failure rates
(in percent) for all valid inspections performed in the
Michigan AET program in the first quarter of 1986. Data are
provided by model year, and by the aggregate of the 1981
through 1986 model years. Available fields in the raw data
did not permit isolating light-duty vehicle and light-duty
truck failures.
Table 30: Michigan AET Failure Rates by Model Year and
Manufacturer
AMC
CHRY
FORD
GM
MONO
MITS*
NISS
TOYT
VW
1981
17.8
21.2
25.0
15.9
3.8
50.0
48.3
10.4
25.5
1982
15.3
20.6
25.1
13.1
6.0
66.7
28.6
5.7
12.5
1983
16.4
14.4
14.0
8.3
11.3
15.4
15.2
8.5
10.8
1984
9.9
13.7
9.9
9.0
10.3
10.0
16.9
5.1
4.0
1985
10.8
5.3
6.6
5.4
11.6
15.4
4.4
4.2
1.5
1986
n/a
10.5
8.5
6.1
0.0
n/a
0.0
n/a
n/a
1 981-86
14.7
14.9
14.4
10.3
9.0
26.0
17.4
7.0
11.5
Small sample size
103
-------
APPENDIX B
Vehicle Identifying Information for the CTP Base Sample
Veh
001
002
003
004
005
006
007
008
009
010
011
012
013
014
015
016
017
018
019
020
021
022
023
024
025
026
027
028
029
Mfr
SUBA
AMC
AMC
SUBA
AMC
VW
AMC
VW
SUBA
SUBA
VW
AMC
AMC
TOYT
AMC
VW
SUBA
MAZD
MAZD
MAZD
TOYT
MAZD
TOYT
TOYT
TOYT
SUBA
TOYT
TOYT
AMC
Model
DL
JEEP
ALLI
WAQO
ALLI
JETT
JEEP
JETT
BRAT
GLF
GTI
JEEP
181
TERC
ALLI
RABB
SUBA
GLC
626
GL£
VAN
GL£
CORO
TERC
STAR
GL
CORO
CORO
SPIR
F/M
CARB
CARS
TBI
CARB
TBI
PFI
CARB
PFI
TBI
CARB
TBI
CARB
PFI
CARB
TBI
PFI
PFI
CARB
CARB
CARB
PFI
CARB
CARB
CARB
PFI
CARB
CARB
CARB
CARB
MY
81
83
85
82
85
85
84
84
83
81
85
85
81
85
83
81
86
85
84
85
84
83
84
84
83
81
81
83
81
CID
109
258
85
109
85
109
150
105
109
109
109
258
100
89
85
105
109
91
120
91
122
91
97
95
79
109
108
97
258
Type
LDV
LOT
LDV
LDV
LDV
LDV
LOT
LDV
LDT
LDV
LDV
LDT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
102
16
32
151
40
27
41
66
38
54
27
29
29
23
47
64
17
23
37
13
59
49
70
27
54
60
113
53
31
AETHC
534
1335
327
839
273
88
317
268
104
387
82
190
611
176
356
2000
168
71
181
63
337
97
382
231
225
180
392
978
130
AETCC
1.7
9.5
6.3
10.0
0.5
4.7
0.0
5.7
2.9
7.2
1.6
3.7
6.0
1.6
0.6
0.5
1.4
1.5
4.7
2.4
2.3
2.8
0.0
0.5
0.3
1.7
0.2
5.1
3.9
FTPHC
1.94
4.84
1.08
8.32
0.60
1.47
0.74
0.73
1.71
2.44
0.31
1.35
4.63
0.23
3.78
6.87
0.14
0.28
2.71
0.10
0.58
0.31
0.66
0.23
0.47
2.60
3.77
0.68
2.40
FTPCC
71.6
49.0
21.9
81.3
4.3
58.2
10.8
6.3
55.5
54.2
2.9
18.4
85.1
1.4
18.9
69.3
2.1
8.9
117.0
3.1
8.4
7.7
6.0
1.7
6.0
57.7
48.7
5.1
83.5
HCstc
0.41
1.70
0.41
0.41
0.41
0.41
0.80
0.41
1.70
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstc
7.0
18.0
3.4
7.0
3.4
3.4
10.0
3.4
18.0
7.0
3.4
10.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
10.0
3.4
3.4
3.4
3.4
7.0
3.4
3.4
7.0
XL05HC
33
1096
527
738
44
320
13
581
272
389
1
167
292
6
1742
1763
1
7
87
0
190
1418
34
14
99
328
389
61
9£
XL05CO
0.0
10.0
8.6
10.0
0.1
10.0
0.0
0.6
9.7
7.7
0.0
3.2
6.7
0.0
3.4
10.0
0.0
0.0
2.1
0.0
0.4
6.9
0.2
0.0
0.1
4.7
0.2
0.0
2.4
104
-------
Veh
030
031
032
033
034
035
036
037
038
039
040
041
042
043
044
045
046
047
048
049
050
051
052
053
054
055
056
057
058
059
060
101
102
Mfr
VW
AMC
AMC
VW
VW
MAZD
AMC
SUBA
MAZD
MAZD
VW
VW
AMC
VW
MAZD
VW
AMC
VW
AMC
AMC
VW
VW
SUBA
SUBA
SUBA
MAZD
SUBA
SUBA
MAZD
MAZD
MAZD
CHRY
CHRY
Model
RABB
EAGL
SPIR
RABB
RABB
Gt£
SPIR
GL10
323
RX7
RABB
RABB
181
QUAN
GL£
RABB
ALLI
RABB
181
BJOO
VANO
RABB
GL
GL
DL
626
GL10
GL10
RX7
323
626
FIFT
ARIE
F/M
PFI
CARB
CARB
PFI
PFI
CARB
CARB
PFI
PFI
PFI
PFI
CARB
PFI
PFI
CARB
TBI
TBI
CARB
PFI
TBI
TBI
CARB
CARB
TBI
CARB
CARB
PFI
PFI
PFI
PFI
CARB
CARB
CARB
MY
82
82
82
81
84
85
82
85
86
85
81
82
81
82
83
82
83
83
81
85
84
84
82
86
82
84
85
85
84
86
83
83
81
CID
105
258
151
105
109
91
151
109
98
80
105
105
101
105
91
105
85
105
101
85
117
105
109
109
109
122
109
109
80
98
122
318
135
Type
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
44
53
88
66
69
49
49
29
37
18
84
76
36
70
92
56
72
93
58
47
81
66
82
43
108
71
39
45
84
16
55
69
79
AETHC
150
578
232
223
119
228
237
465
236
1743
95
1478
338
163
350
114
541
364
376
486
225
423
294
179
0
176
141
802
477
1047
417
244
228
AETCO
1.4
0.4
6.2
0.0
4.0
6.4
3.9
0.4
0.1
0.1
1.4
0.1
0.7
8.5
1.9
6.2
0.6
0.4
0.0
6.1
4.4
6.2
5.2
2.1
4.5
1.5
2.2
2.1
2.6
0.6
5.4
3.6
3.5
FTPHC
0.45
1.13
3.50
0.60
1.33
2.30
1.35
0.20
0.43
0.26
0.95
5.21
1.93
1.85
1.08
1.92
7.98
2.11
0.57
0.74
0.71
1.98
4.24
0.16
1.97
0.62
0.27
0.20
3.40
0.28
1.05
1.31
1.42
FTPCO
5.4
14.4
104.3
3.1
15.2
54.9
28.8
2.6
3.0
1.6
13.9
30.3
14.0
72.2
13.7
69.0
37.5
25.6
4.5
10.1
7.9
35.2
34.1
3.6
47.8
10.6
4.9
4.1
40.9
4.0
14.5
14.5
33.1
HCstd
0.41
1.70
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstd
3.4
18.0
7.0
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
10.0
3.4
7.0
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
7.0
XL05HC
130
26
196
139
231
598
266
53
143
2
150
1131
299
216
53
142
2000
520
17
518
2
98
540
21
746
37
19
35
201
111
23
55
20
XL05CO
1.2
0.0
4.8
0.0
8.6
9.5
3.8
0.3
0.2
0.0
3.8
0.2
0.7
10.0
0.1
5.0
3.3
0.9
0.0
9.8
0.0
1.6
8.4
0.1
10.0
0.0
0.2
0.1
0.8
0.2
0.0
0.8
0.0
105
-------
Veh
103
104
105
107
109
110
111
112
113
114
115
116
117
201
202
203
204
205
206
207
208
209
210
211
212
213
214
216
217
218
219
220
221
Mfr
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
Model
CHAR
OMNI
CARA
RELI
DAYT
LEBA
600
RELI
LANC
LEBA
NEW
ARIE
RED
LINC
TOVW
LINC
CAPR
MUST
F150
RANG
TEMP
MARQ
TEMP
TOPA
RANG
TOPA
TEMP
TEMP
MARQ
LTD
MUST
ZEPH
RANG
F/M
CARB
CARB
TBI
CARB
PFI
PFI
TBI
TBI
PFI
PFI
PFI
CARB
CARB
CARB
TBI
TBI
CARB
CARB
CARB
TBI
CARB
CARB
TBI
CARB
TBI
TBI
TBI
TBI
CARB
CARB
CARB
CARB
CARB
MY
85
84
85
81
84
86
84
85
85
86
85
81
81
82
82
83
81
81
86
85
84
81
85
84
85
86
86
85
82
81
81
81
84
CID
135
135
135
135
135
135
135
135
135
135
135
135
135
302
302
302
140
140
300
140
140
302
140
140
140
140
140
140
302
302
140
140
140
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LOT
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
KMile
29
44
29
64
41
41
54
32
41
36
40
34
49
59
66
68
75
150
16
61
53
67
33
68
49
13
15
58
92
70
68
117
44
AETHC
292
391
150
267
152
364
537
175
242
306
872
100
334
881
975
575
321
187
250
600
244
230
518
1382
744
292
105
1333
147
640
422
183
124
AETCO
3.8
2.6
1.5
6.3
1.3
0.7
6.8
2.8
2.2
0.3
10.0
3.6
7.8
1.9
10.0
1.5
2.3
3.7
0.0
0.7
0.2
5.2
5.9
10.0
. 0.6
4.5
1.8
8.1
1.7
0.8
3.6
3.2
1.4
FTPHC
1.49
1.14
3.63
1.19
1.17
0.51
3.00
2.10
0.51
0.65
8.73
1.19
2.81
9.33
19.07
3.31
1.04
3.57
0.62
1.79
3.02
1.04
2.68
6.17
1.18
0.65
0.80
2.15
1.81
2.53
4.90
1.66
0.54
FTPCO
31.8
6.3
63.7
8.1
14.0
4.8
37.5
61.4
5.3
7.1
86.3
35.4
51.3
65.2
62.6
24.9
7.5
66.5
0.9
10.5
47.3
23.6
56.3
58.7
5.9
17.0
23.7
30.1
34.9
39.3
64.9
28.3
4.1
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.80
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
COstd
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
7.0
7.0
7.0
7.0
3.4
3.4
3.4
10.0
10.0
3.4
3.4
3.4
3.4
10.0
3.4
3.4
3.4
7.0
3.4
3.4
3.4
10.0
XL05HC
231
#N/A
406
26
136
48
2024
34
20
165
502
44
345
165
910
75
109
461
60
380
279
38
43
300
500
58
43
1205
211
25
557
155
40
XL05CO
4.0
#N/A
7.1
0.2
1.0
0.0
7.1
0.4
0.0
0.2
7.1
0.6
6.8
2.0
8.3
0.1
0.0
7.8
0.0
0.0
0.0
0.0
0.0
2.6
0.0
0.0
0.0
7.7
4.9
0.0
3.4
0.7
0.0
106
-------
Veh
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
Mfr
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
Model
LTD
MARQ
MARQ
MUST
MARQ
MUST
SABL
TOPA
LTD
TOPA
TEMP
MARQ
TEMP
TEMP
LTD
TEMP
TEMP
F150
RANG
TEMP
OOUN
TEMP
MUST
TAUR
TEMP
MUST
TEMP
F250
RANG
SABL
EXP
TEMP
MUST
F/M
CARS
CARB
TBI
CARB
CARB
CARB
PFI
TBI
CARB
CARB
TBI
CARB
TBI
TBI
CARB
TBI
TBI
CARB
TBI
TBI
CARB
TBI
CARB
TBI
TBI
CARB
TBI
CARB
CARB
TBI
TBI
TBI
CARB
MY
81
82
86
81
82
81
86
86
81
84
85
82
85
85
82
85
86
86
86
85
81
85
81
86
85
81
85
82
83
86
84
86
81
CID
255
302
302
140
302
140
183
140
302
140
140
302
140
140
302
140
140
300
177
140
302
140
140
150
140
140
140
351
122
183
98
140
140
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LOT
LDV
LDV
LDV
LDV
KMile
80
46
28
93
54
59
47
5
41
55
25
88
58
21
33
43
16
36
27
21
73
45
76
17
42
70
25
45
70
14
46
10
65
AETHC
293
161
449
327
153
132
459
144
1400
1020
312
596
326
946
268
485
313
58
309
411
117
938
312
373
644
306
262
1126
214
126
164
651
102
AETCO
5.2
1.3
0.1
2.3
1.9
1.5
5.2
3.0
8.5
0.5
3.5
7.9
3.9
10.0
3.9
0.8
3.6
2.2
0.7
3.7
3.0
4.2
5.6
4.6
5.1
0.0
0.5
0.0
3.5
1.3
1.6
8.8
2.8
FTPHC
3.03
0.33
0.39
2.14
1.82
3.71
0.25
0.36
1.51
1.54
0.47
5.19
3.84
0.82
1.98
0.60
6.75
2.30
0.91
0.29
0.57
3.49
1.83
0.73
3.23
1.33
0.61
0.88
1.44
0.16
1.07
3.60
0.38
FTPCO
42.9
1.8
2.0
23.4
27.1
63.9
4.2
10.2
27.4
42.9
10.8
65.6
17.5
17.3
30.4
17.3
63.2
63.4
7.1
6.7
9.5
24.0
36.2
17.1
65.9
23.8
14.7
5.1
11.8
2.2
18.0
65.2
8.2
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
1.70
1.70
0.41
0.41
0.41
0.41
COstd
3.4
7.0
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
7.0
3.4
3.4
7.0
3.4
3.4
10.0
10.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
18.0
18.0
3.4
3.4
3.4
3.4
XL05HC
72
71
40
588
63
144
176
27
28
87
26
452
409
926
29
14
147
50
974
34
0
965
436
51
164
31
62
147
64
21
122
826
84
XL05CO
0.1
0.0
0.0
0.3
0.0
1.2
0.0
0.1
0.0
0.0
0.0
6.2
0.1
6.6
0.0
0.0
0.1
1.1
7.8
0.2
0.0
5.1
5.5
0.0
0.8
0.0
0.0
0.0
0.0
0.0
0.0
8.7
0.0
107
-------
Veh
257
258
259
260
301
302
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
Mfr
FORD
FORD
FORD
FORD
QA
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
Model
MUST
MUST
CAPR
RANG
RIVI
REGA
MALI
BONN
CELE
MALI
RIVI
CIER
PHOE
MALI
SKYH
FIER
SEVI
CORV
CUTL
FIER
FIRE
CENT
CELE
DEVI
CAVA
RIVI
REGA
CIER
CITA
BONN
2000
GRAN
CIER
F/M
CARB
CARB
CARB
CARB
CARB
CARB
CARB
CARB
TBI
CARB
CARB
TBI
CARB
CARB
TBI
TBI
TBI
TBI
CARB
TBI
TBI
CARB
CARB
TBI
TBI
CARB
CARB
TBI
CARB
CARB
TBI
CARB
TBI
MY
81
81
81
84
85
82
82
82
86
81
81
84
81
81
83
84
81
82
81
84
85
81
83
82
86
81
82
84
81
82
83
84
84
CID
140
140
140
122
307
231
229
231
151
229
250
151
151
229
110
151
368
350
231
151
110
231
173
250
121
250
231
151
151
231
110
305
151
Type
LDV
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
78
94
83
35
10
98
83
65
5
102
68
61
64
43
58
49
54
49
57
44
27
62
116
60
18
72
113
35
140
55
59
37
60
AETHC
366
263
233
146
570
298
263
341
225
399
805
500
184
502
357
265
1167
306
602
227
167
1639
309
264
239
249
2000
244
157
172
367
252
325
AETCO
5.4
0.0
0.1
2.6
4.4
0.4
0.4
0.3
0.4
1.2
10.0
0.5
1.8
0.6
0.5
0.3
1.4
4.4
3.2
1.7
2.5
0.2
1.7
0.0
0.9
0.9
0.3
0.6
1.6
1.5
0.2
0.4
0.3
FTPHC
3.30
0.57
2.20
0.73
0.70
4.06
2.61
3.97
0.21
0.56
0.84
0.25
1.37
0.59
0.93
1.29
3.72
3.11
1.59
0.55
7.65
4.13
1.82
0.33
0.16
1.52
4.83
0.85
3.40
2.32
0.97
2.01
0.24
FTPCO
49.2
4.6
41.4
8.0
2.1
32.3
15.9
22.5
1.8
5.7
8.4
3.3
20.9
9.8
6.5
37.3
35.3
42.0
20.9
3.8
129.2
29.4
7.3
5.2
1.8
23.4
28.2
6.4
34.8
15.7
7.4
15.1
2.9
HCstd
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstd
3.4
3.4
3.4
10.0
3.4
7.0
7.0
7.0
3.4
7.0
3.4
3.4
3.4
7.0
3.4
3.4
7.0
7.0
7.0
3.4
3.4
7.0
3.4
7.0
3.4
3.4
7.0
3.4
3.4
7.0
3.4
3.4
3.4
XL05HC
450
61
49
80
1 1
254
200
754
14
19
151
84
219
338
87
45
1125
261
57
29
643
870
102
18
9
556
1885
121
507
19
173
196
41
XL05CO
5.7
0.0
0.0
0.0
0.0
2.3
0.4
0.3
0.1
0.0
0.6
0.2
6.5
0.3
0.0
0.1
1.5
1.4
0.2
0.1
10.2
0.2
0.0
0.0
0.0
7.2
2.1
0.3
9.1
0.0
0.3
0.5
0.1
108
-------
Veh
331
332
333
334
335
336
337
338
339
340
341
342
343
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
401
402
403
404
Mfr
CM
Ovl
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
HOND
HCND
HOND
HOND
Model
MONT
CMBG
CELE
REGA
CIMA
GRAN
CITA
GVE3
REGA
SKYL
CITA
DEVI
CENT
RIVI
J200
CITA
GRAN
SUNB
SUNB
DEVI
CELE
CITA
CAVA
SKYH
CAVA
GRAN
CAVA
DEVI
CIER
CIVI
CIVI
CIVI
CIVI
F/M
CARB
CARB
CARB
CARB
CARB
CARB
CARB
TBI
CARB
CARB
CARB
TBI
TBI
CARB
CARB
TBI
TBI
TBI
TBI
TBI
TBI
PFI
TBI
TBI
TBI
TBI
TBI
TBI
TBI
PFI
PFI
CARB
CARB
MY
83
81
82
81
82
81
81
82
83
81
81
81
86
81
82
84
86
86
84
83
85
85
84
84
85
86
84
83
82
86
85
84
84
CID
229
151
173
231
112
231
173
151
231
173
151
368
151
307
112
151
151
110
110
249
151
173
121
110
121
151
121
249
151
91
91
91
91
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
46
44
82
50
78
86
72
53
84
55
84
29
37
61
91
25
17
27
47
54
52
23
51
66
42
37
8
92
56
41
58
53
71
AETHC
1261
209
248
304
174
557
904
517
134
68
193
331
452
393
192
331
223
240
237
277
284
687
177
218
395
548
315
520
1062
259
327
227
383
AETCO
6.1
1.5
4.2
0.4
5.8
7.6
0.2
0.2
1.5
1.2
4.9
0.2
0.5
2.5
4.2
0.7
0.1
0.3
0.1
0.4
0.6
0.3
2.6
2.2
3.5
0.7
1.2
0.1
1.0
1.4
0.7
0.0
0.0
FTPHC
6.42
0.51
0.70
1.34
1.56
5.92
0.64
0.28
0.98
0.25
2.12
0.57
1.10
4.94
1.81
0.21
0.40
0.28
0.38
6.45
0.34
0.64
7.64
2.96
0.43
0.35
6.09
1.44
0.29
0.35
0.93
1.46
1.19
FTPCO
93.2
14.3
10:6
14.3
16.9
93.2
5.0
2.7
5.7
8.0
72.2
9.6
5.2
68.9
28.6
2.6
1.2
3.9
8.4
7.0
4.3
6.1
136.5
55.0
7.6
3.3
38.4
5.8
4.2
3.5
8.6
4.5
6.5
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstd
3.4
3.4
7.0
7.0
7.0
7.0
7.0
7.0
3.4
7.0
3.4
7.0
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
XL05HC
774
20
379
155
405
1927
34
179
181
26
230
22
287
214
274
205
123
39
153
192
13
10
260
435
133
36
236
53
33
44
371
160
172
XL05CO
10.4
0.0
6.5
0.2
9.6
8.6
0.0
0.2
1.1
0.1
5.6
0.0
0.3
0.4
5.7
0.3
0.1
0.1
0.6
0.2
0.0
0.1
6.1
9.1
0.6
o.i
1.5
0.0
0.2
0.1
4.3
0.1
0.1
109
-------
Veh
405
406
407
408
409
410
502
503
504
505
506
507
508
509
510
511
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
Mfr
HGND
HCND
HOND
HCND
HOD
HOND
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
Model
CIVI
CIVI
CIVI
CIVI
ACCO
CIVI
COLT
RAM
COLT
COLT
RAM
COLT
COLT
COLT
COLT
COLT
STAN
200S
PULS
SENT
MAX!
MAX!
200S
200S
280Z
300Z
300Z
280Z
TRUC
PULS
280Z
SENT
280Z
F/M
CARB
CARB
CARB
PFI
CARB
PFI
CARB
CARB
CARB
CARB
CARB
TBI
TBI
TBI
TBI
TBI
PFI
PFI
CARB
CARB
PFI
PFI
PFI
PFI
PFI
PFI
PFI
PFI
TBI
PFI
PFI
CARB
PFI
MY
84
84
85
86
85
85
84
85
85
84
85
85
84
85
86
86
86
82
85
84
82
83
81
81
82
84
84
81
86
83
82
84
83
CID
91
91
91
91
112
91
86
122
90
98
122
98
98
97
98
98
120
134
98
98
146
146
119
119
171
181
181
171
146
91
168
98
168
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LDV
LDV
KMile
85
77
13
35
22
53
65
50
34
59
28
21
58
26
39
16
33
95
50
35
91
69
119
90
73
42
24
57
28
55
102
48
74
AETHC
798
565
248
396
244
224
1061
493
24
746
227
2000
108
307
203
127
269
194
279
228
357
132
122
469
299
184
237
268
61
1527
152
239
138
AETCO
0.1
0.0
0.1
3.0
1.8
0.6
0.1
7.8
2.3
0.3
2.7
4.9
1.7
6.7
1.5
2.1
7.0
1.8
0.4
0.1
2.8
2.5
2.1
0.4
0.7
1.7
1.7
4.4
2.3
9.5
1.7
0.2
1.8
FTPHC
2.14
2.05
1.73
0.43
1.75
0.68
1.00
4.77
0.21
1.18
4.44
0.22
0.39
0.24
0.23
0.20
1.15
2.61
2.20
0.36
4.52
0.66
1.20
1.60
0.74
0.57
0.46
5.02
0.61
5.09
1.69
0.16
0.86
FTPCO
4.7
7.4
4.8
4.0
34.0
6.8
8.2
53.3
9.3
6.6
54.2
4.8
6.3
1.4
4.3
3.0
45.8
91.5
48.7
4.8
142.2
9.1
16.9
6.8
12.1
4.3
5.3
125.1
98.9
43.7
54.2
3.6
11.7
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
COstd
3.4
3.4
3.4
3.4
3.4
3.4
3.4
10.0
3.4
3.4
10.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
10.0
3.4
3.4
3.4
3.4
XL05HC
123
500
348
48
291
65
250
593
35
114
600
0
28
2
42
0
276
1013
904
65
167
130
35
433
19
0
0
332
60
561
181
50
226
XL05CO
0.2
0.1
0.2
0.1
1.8
0.1
0.7
7.6
1.9
0.1
8.0
0.0
0.1
0.1
0.2
0.0
2.9
8.0
1.7
0.0
3.3
2.5
0.2
0.4
0.1
0.0
0.0
6.0
3.8
5.8
2.3
0.1
4.3
110
-------
Veh
618
619
620
701
702
703
704
705
706
707
708
709
Mfr
NISS
NISS
NISS
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
Model
PULS
MAXI
200S
TERC
TERC
OORO
CELI
CELJ
OOflO
CORO
CAMR
CELI
F/M
PFI
PFI
PFI
CARB
GARB
CARB
PFI
PFI
CARB
CARB
PFI
CARB
MY
83
83
84
86
84
81
86
83
82
82
84
83
CIO
91
146
120
91
91
108
122
144
108
108
122
144
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
59
61
28
5
77
96
32
59
234
147
32
49
AETHC
446
281
293
262
339
167
335
293
445
1473
344
2000
AETCO
6.9
0.3
7.4
0.1
1.2
1.9
0.4
0.2
0.1
0.2
0.2
0.8
FTPHC
1.88
0.68
2.01
0.24
1.14
2.12
0.26
0.44
2.26
0.51
0.16
0.19
FTPCO
44.4
7.0
72.3
1.9
3.8
47.2
1.7
3.0
49.2
8.9
1.9
3.5
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstd
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
XL05HC
547
1 7
250
7
45
128
0
141
109
1
0
0
XL05CO
5.6
0.0
6.0
0.0
0.0
0.8
0.0
0.1
1.1
0.0
0.0
0.0
111
-------
Appendix C
As Received Failure Rates for Selected Modes of the Basic I/M Test Procedure
VEHICLE
BITP
Mode
CS03
CS05
CS07
CS 10
CS 12
CS 15
XL 02
XL 05
XI 02
XI 05
XI 07
XI 10
RS02
RS05
Failure
COUNT
Type
HC-onlv HC+CO CO-only
49
39
19
20
33
35
19
18
39
41
22
20
16
21
131
68
68
61
74
65
64
63
78
59
63
58
64
64
29
31
26
25
24
20
21
16
23
22
27
23
18
19
Pass
27
98
124
131
106
117
134
141
97
115
125
136
134
128
VEHICLE PERCENT
BITP
Mode
CS03
CS05
CS07
CS 10
CS 12
CS 15
XL 02
XL 05
XI 02
XI 05
XI 07
XI 10
RS02
RS05
Failure Type
HC-only
21%
17%
8%
8%
14%
15%
8%
8%
16%
17%
9%
8%
7%
9%
HC+CO CO-only
56%
29%
29%
26%
31%
27%
27%
26%
33%
25%
27%
24%
28%
28%
12%
13%
11%
11%
10%
8%
9%
7%
10%
9%
11%
10%
8%
8%
Pass
11%
42%
52%
55%
45%
49%
56%
59%
41%
49%
53%
57%
58%
55%
112
-------
APPENDIX C
Emission Values for Calculating As Received Failure Rates for Selected Modes of the Basic I/M Test Procedure
HC values in ppm; (allure >=220 ppm
CO values in % concentration: failure >=1.2%
Veh
001
002
003
004
005
006
007
008
009
010
Oil
012
013
014
015
016
017
018
019
020
021
022
023
024
025
026
027
028
029
030
031
032
033
034
035
036
037
038
039
040
041
042
CS 03
HC CO
247 6.4
2000 8.4
715 7.4
282 8.2
1386 6.7
323 5.8
278 5.2
829 6.6
326 5.4
299 4.3
498 5.3
571 0.3
936 10.0
200 1.3
533 1.0
2000 9.3
818 7.6
963 10.3
503 8.0
647 9.2
462 1.6
530 10.0
242 2.1
256 1.3
1038 5.0
202 3.9
616 1.8
277 1.0
518 2.6
441 9.7
343 5.6
1902 10.0
700 2.7
532 10.0
396 5.5
414 4.0
1956 2.2
718 1.4
192 0.6
479 2.7
903 0.2
464 4.3
CS 05
HC CO
163 6.7
848 6.3
565 8.8
615 10.0
729 5.2
142 2.0
182 4.4
148 0.3
143 3.5
211 3.9
133 0.4
65 0.0
356 4.8
135 1.2
288 2.6
625 10.0
377 8.7
182 0.1
480 8.7
44 0.0
214 0.3
56 0.1
214 1.8
207 0.3
519 2.8
180 3.1
301 2.3
436 2.8
307 4.4
112 0.4
133 0.7
698 9.3
633 0.4
104 0.6
264 2.3
370 3.0
69 0.1
280 0.9
31 0.0
209 1.7
859 0.2
227 0.8
CS 07
HC CO
15 0.0
1255 10.0
356 5.8
565 10.0
143 0.1
197 7.4
9 0.0
29 0.5
127 3.4
545 8.0
9 0.0
169 3.2
242 2.6
59 0.1
2058 3.2
1259 10.0
151 0.1
6 0.1
140 2.6
4 0.0
49 0.0
13 0.0
142 0.7
72 0.0
150 0.1
261 3.1
218 0.2
370 3.0
139 4.4
118 0.5
16 0.0
293 8.0
131 0.0
118 3.2
1024 10.0
229 3.5
39 0.2
96 0.0
10 0.0
1 50 1 .6
2000 0.0
351 0.8
CS 10
HC CO
20 0.0
1202 10.0
471 5.8
S84 10.0
86 0.1
200 7.5
8 0.0
26 0.1
130 3.5
517 7.9
8 0.0
9 0.3
231 2.3
47 0.0
1955 3.2
1014 10.0
54 0.1
4 01
136 2.6
5 0.0
74 01
16 00
118 0.5
56 0.0
187 01
238 4.0
274 0.2
326 2.4
133 4.0
148 0.9
18 0.2
309 8.0
137 0.0
116 3.5
748 9.4
222 3.5
50 0.2
81 0.0
1 1 0.0
148 1.8
1539 0.2
280 0.8
CS 12
HC CO
53 00
1537 10.0
439 64
727 100
159 01
234 8.4
4 0.0
26 01
131 3.1
471 7.5
7 0.0
96 16
192 11
41 00
2005 3.7
2000 10.0
13 00
0 0.1
116 10
3 00
438 6.3
309 5.4
162 09
57 00
261 00
301 47
323 0.2
298 22
135 33
162 1.1
340 36
251 5.8
70 0.0
142 3.9
1270 10.0
196 23
62 0.1
36 0.0
11 0.0
142 2.1
752 0.3
133 0.7
CS 15
HC CO
63 0.0
1414 10.0
429 6.2
727 100
163 0.1
246 8.5
8 00
24 0.0
122 2.8
470 7.6
8 0.0
251 5.4
222 1.4
33 0.0
2002 3.5
2000 10.0
12 0.0
1 0.1
112 1.4
2 0.0
290 0.3
92 3.7
193 1.1
80 0.0
213 0.0
301 44
363 0.2
303 2.3
127 3.6
160 1.2
290 1.1
247 6.3
145 0.0
132 3.9
806 10.0
191 2.8
204 0.3
102 0.1
11 0.0
144 2.0
844 0.2
172 0.7
XL 02
HC CO
21 0.0
1203 10.0
453 8.5
700 100
8 00
386 10.0
15 0.0
319 0.5
277 9.7
402 7.6
3 0.0
177 2.6
239 6.9
8 00
1602 3.3
2000 10.0
1 0.0
5 0.0
99 1.6
1 0.0
226 0.7
1019 5.4
38 0.2
22 0.0
157 0.1
247 5.5
357 0.2
56 0.1
99 2.4
98 0.8
20 0.0
196 ' 4.8
138 0.0
230 8.4
669 10.0
242 4.0
105 0.3
190 0.2
3 0.0
129 3.6
1079 0.2
182 0.6
XL 05
HC CO
33 0.0
1096 10.0
527 8.6
738 10.0
44 0.1
320 10.0
13 00
581 0.6
272 9.7
389 77
1 0.0
167 3.2
292 6.7
6 0.0
1742 3.4
1763 100
1 0.0
7 0.0
87 2.1
0 0.0
190 0.4
1418 6.9
34 0.2
14 00
99 0.1
328 4.7
389 0.2
61 0.0
96 2.4
130 1.2
26 0.0
196 4.8
139 0.0
231 8.6
598 9.5
266 3.8
53 0.3
143 0.2
2 0.0
150 3.8
1131 0.2
299 0.7
XI 02
HC CO
118 0.1
1686 10.0
472 7.6
847 10.0
203 1.2
632 10.0
1 0.0
776 0.3
110 3.1
389 6.8
214 4.9
141 1.5
152 1.2
30 0.0
2000 5.6
1100 10.0
351 6.9
9 0.0
88 0.5
1 00
353 55
889 6.1
127 0.8
264 0.7
266 0.2
329 3.7
335 0.2
43 0.0
105 3.0
165 2.3
775 1.8
203 4.2
63 0.0
174 7.2
525 9.3
281 3.7
182 0.2
1 0.0
4 0.0
156 3.5
1304 0.2
165 0.7
XI 05
HC CO
56 0.0
1343 10.0
813 6.9
701 10.0
169 0.2
710 10.0
It 00
326 0.6
104 2.7
349 6.2
164 3.4
169 1.9
184 0.8
146 0.2
1821 3.8
1219 100
57 0.0
8 0.0
49 09
15 0.0
386 5.7
197 5.5
176 1.1
330 1.2
217 0.1
260 4.5
407 0.2
52 0.0
104 3.1
137 2.1
477 0.4
211 4.4
168 0.0
167 7.0
169 2.4
267 4.0
164 0.3
325 0.2
6 0.0
151 3.5
1090 0.2
162 0.6
XI 07
HC CO
23 0.0
1293 100
487 80
695 10.0
58 00
277 95
6 00
68 00
106 31
275 5.3
1 1 00
110 1.5
282 6.0
135 06
1819 37
1164 10.0
4 00
2 0.0
55 18
6 0.0
204 0.9
808 62
105 0.4
99 0.1
169 0.1
285 5.0
325 0.2
77 0.2
99 3.4
140 1.2
27 0.0
211 5.2
201 0.0
140 6.2
594 9.2
250 3.3
61 0.2
192 01
5 0.0
163 3.8
1224 0.2
203 0.7
XI 11
HC CO
28 0.0
1564 10.0
509 8.3
718 10.0
71 0.1
435 10.0
8 0.0
58 0.1
108 3.1
290 5.3
1 1 0.0
195 2.6
265 6.0
69 0.2
1753 3.5
1274 10.0
3 0.0
3 00
67 2.1
5 0.0
193 0.6
894 6.2
73 0.3
53 0.3
142 0.1
246 5.8
311 0.2
58 0.0
102 3.1
131 1.1
24 0.0
225 5.1
200 0.0
152 6.5
614 10.4
258 3.5
57 0.2
282 0.2
5 0.0
164 4.0
1136 0.2
166 0.7
RS 02
HC 00
51 0.1
1363 10.0
431 9.3
1015 10.0
14 0.0
581 10.0
3 0.0
432 2.0
109 3.1
271 5.6
0 0.0
1 0.0
254 6.6
35 0.0
1690 3.6
1794 10.0
6 00
4 0.0
89 1.5
1 0.0
185 0.6
663 6.6
27 0.1
6 0.0
133 0.1
280 3.3
443 0.2
38 0.0
103 3.2
131 2.7
137 0.3
228 5.1
112 0.0
733 9.6
653 10.2
216 3.6
57 0.3
110 0.2
3 0.0
241 9.4
1338 0.2
169 0.7
RS05
HC CO
100 0.2
1383 4.0
446 9.3
1051 10.0
13 00
428 10.0
5 00
722 0.4
99 2.9
267 5.6
2 0.0
126 0.6
273 6.2
54 0.0
1744 3.5
2000 10.0
6 0.0
2 0.0
82 2.0
0 0.0
192 0.5
444 6.7
64 0.2
33 0.0
100 0.0
322 5.2
393 0.2
37 0.0
100 3.2
153 2.4
15 0.0
220 5.4
208 0.0
660 10.0
643 10.0
246 3.5
3 0.0
165 0.2
3 0.0
246 9.1
1123 0.2
245 0.7
113
-------
Veh
043
044
045
046
047
048
049
050
051
052
053
054
055
056
057
058
059
060
101
102
103
105
107
109
110
111
112
113
114
115
116
117
201
202
203
204
205
206
207
208
209
210
211
212
213
214
216
217
218
CS 03
HC 00
554 10.0
2000 6.6
517 10.0
2000 4.7
379 1.2
444 1.7
1121 9.9
- 332 7.8
709 5.8
2000 6.8
273 0.4
167 3.8
543 8.8
419 2.6
434 2.3
680 3.2
368 2.7
377 2.9
342 7.1
185 4.2
502 7.1
220 3.3
502 7.1
105 0.7
206 0.2
270 6.2
502 7.0
154 0.3
295 0.5
243 4.7
460 5.0
502 6.9
288 4.3
507 7.6
77 2.5
167 3.2
325 7.1
956 1.5
317 1.5
1851 6.6
43 1.0
206 2.0
367 2.9
657 0.5
271 2.0
537 6.5
305 1.7
93 1.3
48 0.6
CS 05
HC CO
291 8.0
1752 0.0
202 3.8
2000 4.8
455 1.2
152 0.1
873 8.3
187 4.9
650 5.7
197 3.2
118 0.3
99 2.1
246 2.4
58 0.1
103 0.4
452 2.3
140 0.1
361 5.7
72 1.3
94 0.0
437 1.0
226 6.0
121 0.3
114 1.5
233 1.0
266 5.9
308 7.1
105 1.5
76 0.1
502 7.1
123 1.2
189 1.2
102 1.1
1131 8.3
80 0.6
128 0.3
293 6.0
163 0.0
257 0.7
387 0.1
0 0.0
155 0.5
166 2.2
562 0.5
366 4.3
213 0.3
60 0.0
270 6.2
12 0.2
CS 07
HC CO
274 9.7
62 0.0
140 40
2000 2.8
430 1.3
0 0.0
353 4.2
114 0.2
379 4.2
412 6.2
132 0.2
298 7.8
59 0.1
13 0.0
17 0.0
216 0.7
278 0.4
45 0.0
25 0.0
12 0.0
274 5.3
387 7.1
124 3.4
87 0.8
34 0.0
461 7.1
33 0.1
65 0.3
49 0.2
502 7.1
41 1.6
282 5.4
959 8.7
1298 8.3
56 0.3
142 0.0
435 7.9
32 0.0
300 0.1
290 0.0
420 6.5
26 0.1
417 4.3
386 0.0
50 0.0
41 0.0
1161 8.3
1018 3.1
77 2.3
CS 10
HC CO
274 9.9
54 0.0
147 4.2
2000 2.9
376 1.1
0 0.0
319 3.0
99 0.1
192 1.7
296 6.2
65 0.2
273 7.4
60 0.0
14 0.0
19 0.0
217 0.7
120 0.2
33 0.0
20 0.0
11 0.0
251 4.9
502 7.1
79 1.6
91 0.7
31 0.0
406 7.1
22 0.2
12 0.1
207 0.4
502 7.1
42 1.7
286 5.7
88 1.6
890 8.3
57 0.1
134 0.0
436 8.1
57 0.0
370 0.1
302 0.0
32 0.0
18 0.0
285 2.5
401 0.0
53 0.0
115 0.6
1139 8.3
201 4.7
15 0.0
CS 12
HC CO
248 10.0
67 0.0
140 4.7
2000 2.7
823 03
0 0.0
305 2.7
43 0.0
117 0.9
239 4.1
22 0.0
301 7.6
55 0.1
58 0.0
35 0.0
247 0.6
169 0.1
63 0.0
10 0.0
38 0.0
280 4.4
502 7.1
21 0.9
132 1.0
44 0.1
337 6.7
7 0.1
82 0.1
74 0.2
502 7.1
26 1.1
251 6.5
1343 8.7
1168 8.3
71 0.1
696 7.0
440 7.9
120 0.0
997 4.6
79 0.1
201 2.3
68 0.0
649 4.1
750 0.1
269 3.5
462 4.8
1887 8.3
269 4.1
81 2.4
CS 15
HC CO
248 10.0
41 0.0
154 4.8
2000 2.9
798 0.3
0 0.0
293 2.5
22 0.0
144 1.1
314 6.5
102 0.1
324 8.0
57 0.0
41 0.0
84 0.2
245 0.6
141 0.2
35 0.0
6 0.0
10 0.1
283 4.7
502 7.1
46 0.1
119 0.9
28 0.0
438 7.1
19 0.2
36 0.0
282 0.3
502 7.1
85 2.4
378 6.8
165 2.4
972 8.1
74 0.1
356 1.5
405 7.6
116 0.0
697 0.7
733 0.0
36 0.0
35 0.0
322 2.3
818 0.7
95 0.0
192 0.5
1342 8.3
228 4.2
72 0.0
XL 02
HC CO
209 9.9
60 0.3
136 5.5
1884 3.7
491 1.0
14 0.0
497 9.8
3 0.0
89 1.9
505 7.9
32 0.1
693 10.0
36 0.0
20 0.1
109 0.2
173 0.8
270 0.4
64 0.0
46 0.9
34 0.0
225 3.5
192 4.4
11 0.1
144 1.6
82 0.1
1692 6.7
61 O.S
98 0.2
67 0.1
502 7.1
37 0.8
1343 6.2
1082 8.7
1049 8.3
76 0.1
109 0.0
534 8.4
67 0.0
347 0.0
344 0.0
36 0.0
54 0.0
361 2.3
414 0.0
59 0.0
71 0.0
1754 8.3
263 5.1
57 0.6
XL 05
HC 00
216 10.0
53 01
142 50
2000 3.3
520 0.9
17 0.0
518 9.8
2 0.0
98 1.6
540 8.4
21 0.1
746 10.0
37 0.0
19 02
35 0.1
201 0.8
111 0.2
23 0.0
55 0.8
20 0.0
231 4.0
406 7.1
26 0.2
136 1.0
48 0.0
2024 7.1
34 0.4
20 0.0
165 0.2
502 7.1
44 0.6
345 6.8
165 2.0
910 8.3
75 0.1
109 0.0
461 7.8
60 0.0
380 0.0
279 0.0
38 0.0
43 0.0
300 2.6
500 0.0
58 0.0
43 0.0
1205 7.7
211 4.9
25 0.0
XI 02
HC CO
202 10.0
109 0.1
136 6.0
2000 2.7
756 08
26 0.0
369 7.0
14 0.0
98 0.9
233 51
381 3.6
390 8.8
47 0.1
272 4.0
64 0.0
221 05
158 0.2
53 0.0
29 0.5
28 0.0
260 3.6
21 0.0
26 0.7
128 0.8
195 0.3
312 6.1
0N/A »N/A
86 0.2
24 0.2
466 7.1
45 1.6
218 6.2
1348 8.7
1258 8.3
83 0.1
621 6.7
372 7.0
214 0.0
769 0.7
449 2.1
171 1.6
47 0.0
241 2.9
960 1.3
228 1.9
327 2.7
1887 8.2
267 5.1
63 1.2
XI 05
HC CO
199 9.6
52 0.0
131 4.9
2000 3.0
526 1.0
31 0.0
327 4.7
16 0.0
114 1.2
338 7.1
273 1.0
387 8.8
44 0.0
157 0.1
81 0.1
219 0.6
104 0.2
45 0.0
48 0.9
6 0.0
252 3.7
151 1.7
80 0.4
121 0.9
157 0.6
437 7.1
»N/A *N/A
58 0.1
187 0.2
502 7.1
89 2.6
475 6.9
229 2.5
1580 8.2
81 0.1
333 1.4
376 6.9
158 0.0
631 0.7
286 0.1
38 0.0
57 0.0
146 1.5
794 0.7
88 0.0
97 0.0
926 4.4
203 4.3
27 0.0
XI 07
HC 00
204 9.4
64 0.0
120 4.6
2000 3.6
276 1.4
12 0.0
270 4.9
12 0.0
98 1.0
315 6.9
87 0.2
404 9.1
44 0.1
24 0.0
46 0.2
183 0.7
62 0.2
36 0.0
38 2.0
8 0.0
194 4.1
153 3.0
179 3.1
108 1.0
30 0.
338 7
0N/A ON/A
1 1 0.
50 0.
502 7.
62 1.9
325 6.5
1111 8.7
1536 8.3
73 0.1
254 0.1
370 7.2
65 0.0
462 0.0
180 0.1
218 2.3
57 0.0
722 4.1
576 0.0
70 0.0
68 0.0
807 6.8
260 5.6
66 1.3
XI 11
HC CO
211 9.5
57 0.1
127 3.9
2000 29
362 1.4
1 1 0.0
285 5.0
12 0.0
87 0.9
303 6.9
77 0.2
396 8.8
42 0.0
21 0.0
28 0.2
178 0.7
63 0.2
54 0.0
43 1.9
7 0.0
192 4.0
173 2.9
62 0.5
112 1 .0
28 0.0
284 7.0
»N/A «N/A
8 0.1
28 0.0
502 7.1
53 1.7
356 6.7
135 2.3
1153 8.3
66 0.1
211 0.0
427 7.5
59 0.0
484 0.0
201 0.1
44 0.0
39 0.0
467 2.5
544 0.0
81 0.3
55 0.0
985 5.9
224 4.5
27 0.0
RS 02
HC CO
181 10.0
41 0.0
135 6.1
1875 4.8
486 1.3
3 0.0
1146 10.0
13 0.0
»N/A 0N/A
351 7.1
577 8.3
333 6.3
22 0.0
17 0.1
72 0.2
193 0.8
531 0.8
28 0.0
78 3.0
39 0.1
216 3.6
96 0.5
24 0.9
130 1.1
42 0.0
295 5.9
22 04
11 0.2
162 0.1
502 7.1
53 1.6
214 5.6
1115 8.7
1177 8.2
263 0.4
167 0.0
572 8.4
93 0.0
382 0.1
201 0.1
35 0.0
76 0.0
»N/A »N/A
572 0.1
47 0.0
48 0.1
703 7.2
238 4.8
54 0.5
RS 05
HC CO
189 9.4
29 0.0
134 4.6
2000 4.6
359 1.4
9 0.0
1170 10.0
14 0.0
*N/A «N/A
335 7.1
487 7.1
298 5.6
22 0.0
26 0.1
1 1 0.1
311 1.5
118 0.3
33 0.0
71 1.3
18 0.0
217 3.7
348 7.0
14 0.1
122 1.0
51 0.0
313 6.7
36 0.3
13 0.2
189 0.2
502 7.1
68 18
248 6.0
1083 8.7
1088 8.3
84 0.1
189 0.0
444 8.0
94 0.0
358 0.0
154 0.1
160 2.3
48 0.0
»N/A »N/A
421 0.0
46 0.0
41 0.0
750 5.7
274 5.5
50 05
114
-------
Veh
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
257
258
259
260
301
302
304
305
306
307
308
309
310
CS 03
HC 00
632 83
259 32
206 33
451 24
266 0.3
238 0.1
222 1.2
875 2.7
259 3.7
678 8.4
115 2.0
817 2.0
127 4.4
312 2.7
421 1.6
248 0.2
565 5.5
32 0.0
81 1.3
382 3.5
54 2.2
125 1.9
97 1.3
356 3.2
361 5.5
354 4.6
404 5.9
450 5.4
97 2.1
79 0.0
110 1.4
112 2.1
199 0.3
175 3.0
727 7.0
109 1.0
2098 2.3
90 0.4
363 7.0
265 0.3
105 0.9
477 3.1
1970 9.5
678 4.2
234 0.2
302 0.2
199 0.2
340 0.3
96 2.8
CS 05
HC CO
743 75
308 1.6
235 03
1248 7.2
239 0.0
87 02
154 0.5
191 06
208 2.4
49 0.0
SO 0.1
28 0.0
102 0.0
218 0.4
323 3.0
263 0.3
759 7.9
20 0.0
86 0.6
157 0.0
115 2.7
31 0.0
100 0.7
13 0.0
509 0.8
167 3.1
636 8.4
396 5.3
92 1.9
222 0.0
52 0.0
44 0.0
11 0.0
206 2.6
514 3.0
28 0.0
859 4.3
41 0.0
339 4.2
116 0.1
23 0.0
472 2.0
304 0.3
1266 0.1
231 0.1
232 0.7
212 0.4
207 0.6
209 3.7
CS 07
HC CO
439 33
217 28
55 0.1
56 0.4
594 5.9
8 0.0
122 1.0
219 2.7
201 1.5
50 0.0
14 0.0
433 7.3
68 0.0
59 0.0
380 5.4
344 1.9
576 5.7
107 1.8
19 0.0
48 0.0
61 1.7
100 0.0
63 0.7
2 0.0
998 4.8
326 6.2
44 0.0
715 7.0
3 0.0
67 0.0
144 0.0
120 1.2
0 0.0
452 0.2
804 8.7
28 0.0
565 6.7
49 0.0
67 0.0
46 0.0
537 5.0
312 3.5
169 0.3
627 0.4
23 0.1
39 0.0
76 0.2
156 0.2
154 3.7
CS 10
HC CO
612 3.5
243 2.8
49 0.0
138 1.8
296 0.0
8 0.0
128 1.0
185 0.0
206 1.6
78 0.0
16 0.0
38 0.0
93 0.0
33 0.0
398 6.0
355 2.1
634 6.0
30 0.0
17 0.0
39 0.0
61 1.5
82 0.0
62 0.7
3 0.0
989 5.2
334 6.2
150 0.9
577 5.7
1 0.0
75 0.0
78 0.0
132 1.6
5 00
280 1.6
781 8.7
37 0.0
555 6.8
59 0.0
57 0.0
46 0.0
4 0.0
268 1.3
168 0.4
973 0.6
5 0.0
28 0.0
27 0.0
126 0.2
157 3.1
CS 12
HC CO
528 3.2
76 00
161 18
97 0.2
555 3.9
846 1.0
144 1.1
162 1.4
367 2.6
144 0.0
154 2.5
312 6.3
116 0.0
99 0.0
476 6.0
664 5.0
875 6.9
118 13
20 0.0
719 6.8
S3 0.8
667 5.8
88 1.1
7 0.0
1583 7.2
398 6.0
375 3.8
929 8.3
3 0.0
163 0.0
308 3.1
220 2.7
1 1 0.0
321 1.1
1015 8.7
467 8.6
467 5.4
127 0.0
159 0.2
84 0.8
162 0.3
217 0.6
234 0.4
1588 1.1
22 0.1
46 0.0
78 0.1
150 0.2
176 3.5
CS 15
HC CO
610 2.8
193 2.3
110 1.1
155 11
314 0.0
41 0.0
113 0.8
70 0.0
225 1.3
164 0.0
131 1.3
40 0.0
98 0.0
79 0.0
463 6.3
1005 6.2
614 6.0
48 0.0
26 0.0
500 3.4
58 1.0
363 0.3
51 0.2
3 0.0
1061 4.7
340 5.6
91 0.0
392 5.5
5 0.0
187 0.0
113 0.0
124 1.5
19 0.0
303 1.2
1071 8.7
177 0.0
463 5.8
111 0.0
103 0.0
74 0.3
33 0.0
236 0.8
184 0.4
1111 0.5
22 0.2
37 0.0
44 00
173 0.1
171 3.6
XL 02
HC CO
759 3.5
209 18
32 0.0
52 0.0
67 0.0
43 0.0
147 0.8
156 2.6
149 1.5
170 0.0
19 00
352 7.1
114 0.0
97 0.0
413 5.6
510 4.0
462 4.9
102 1.1
17 0.0
91 0.0
43 0.8
1094 7.9
19 0.0
5 0.0
1309 5.4
479 6.0
140 1.2
341 3.5
8 0.0
59 0.0
53 0.0
64 0.1
17 0.0
110 0.0
908 8.7
86 0.0
458 6.1
74 0.0
62 0.0
72 0.0
15 0.0
236 2.4
240 0.5
6 0.1
19 0.1
20 0.0
204 1.1
95 0.1
197 5.4
XL 05
HC CO
557 34
155 0.7
40 0.0
72 0.1
71 0.0
40 0.0
588 0.3
63 0.0
144 1.2
176 0.0
27 0.1
28 0.0
87 0.0
26 0.0
452 6.2
409 0.1
926 6.6
29 0.0
14 0.0
147 0.1
SO 1.1
974 7.8
34 0.2
0 0.0
965 5.1
436 5.5
51 0.0
164 0.8
31 0.0
62 0.0
147 0.0
64 0.0
21 0.0
122 0.0
826 8.7
84 0.0
450 5.7
61 0.0
49 0.0
80 0.0
11 0.0
254 2.3
200 0.4
754 0.3
14 0.1
19 0.0
151 0.6
84 0.2
219 6.5
XI 02
HC CO
869 2.8
57 00
123 1.4
236 18
278 2.6
604 0.6
221 0.9
132 1.3
283 2.1
212 0.1
68 0.6
262 4.8
228 3.0
356 2.0
474 6.1
539 4.2
601 5.6
121 1.0
37 0.1
771 8.9
53 0.8
1068 7.9
70 0.7
4 0.0
1651 6.7
643 5.4
269 1.8
646 7.1
27 0.0
336 0.5
270 2.7
193 2.3
101 0.0
448 0.0
1128 B.7
634 7.8
403 4.3
160 0.0
171 0.2
65 0.4
278 1.4
248 1.1
204 0.5
42 0.4
9 0.0
66 0.1
172 0.3
115 0.0
184 4.7
XI 05
HC CO
767 26
118 06
93 08
149 0.7
108 00
75 0.0
133 0.7
46 0.0
139 1.0
185 0.0
25 0.0
40 0.0
107 1.9
158 0.0
419 5.2
500 1.7
1091 6.9
51 0.0
80 0.0
469 4.0
54 0.9
1004 7.8
69 04
0 0.0
832 4.6
629 5.1
91 0.0
306 06
18 0.0
79 0.1
101 0.0
117 1.2
136 0.1
267 0.7
955 7.9
330 0.0
398 5.0
66 0.0
126 0.1
99 0.2
273 1.1
218 0.6
281 0.5
48 0.2
36 0.1
64 0.1
132 0.2
106 0.0
234 6.9
XI 07
HC CO
826 32
86 06
44 0.1
474 8.0
325 4.5
37 0.0
106 0.7
101 0.4
132 1.1
140 0.0
19 0.0
433 7.4
43 0.0
30 0.0
404 5.8
361 0.0
485 5.4
109 1.3
45 0.0
124 0.1
40 0.8
1274 7.9
42 0.3
59 1.7
1731 0.3
610 5.3
60 0.0
454 5.0
8 0.0
78 0.0
166 0.0
81 0.6
103 0.0
166 0.0
784 7.5
263 00
407 5.4
62 0.0
52 0.0
62 0.0
638 5.3
236 1.6
217 0.5
29 0.0
59 0.2
20 0.0
225 1.8
141 0.1
235 7.9
XI 11
HC CO
657 33
77 03
41 0.0
94 01
103 00
87 0.0
123 0.6
49 00
132 1.1
152 00
18 0.0
54 01
36 0.6
30 00
415 5.7
329 0.1
962 7.1
49 0.0
24 00
239 1.0
44 08
851 7.9
46 0.3
3 0.0
1731 05
612 5.3
71 0.0
90 0.0
10 00
90 0.0
92 0.0
102 1.2
125 0.0
172 0.0
754 7.6
264 0.0
425 5.4
67 0.0
46 0.0
70 0.0
255 1.7
237 1.6
208 0.5
11 0.1
25 0.1
19 0.0
201 1.3
59 0.1
215 7.0
RS 02
HC CO
579 35
199 1.1
32 0.0
37 0.0
85 0.0
21 00
200 0.6
125 0.4
137 1.2
104 0.0
30 0.1
344 71
90 0.0
78 0.0
»N/A ON/A
463 4.0
418 4.3
96 1.1
14 00
607 6.4
44 0.8
1114 7.8
22 0.0
5 0.0
»N/A ffN/A
1205 55
78 0.2
440 4.0
28 00
50 00
65 0.0
62 00
95 0.0
98 0.0
895 8.7
241 0.0
463 5.8
61 0.0
59 0.0
39 0.0
184 0.2
242 3.2
196 0.3
4 0.1
18 0.1
21 0.0
101 0.2
28 0.1
297 8.7
RS 05
HC CO
671 3.1
154 0.8
39 00
21 0.0
86 0.0
46 0.0
236 03
102 0.0
130 0.9
107 0.0
15 0.0
385 72
64 0.0
25 0.0
#N/A ffN/A
387 0.7
303 3.8
90 1.0
25 0.0
458 6.7
49 0.9
1144 7.9
33 0.1
0 0.0
»N/A ffN/A
1192 4.9
59 0.0
267 3.2
20 0.0
49 0.0
73 0.0
62 0.1
97 0.0
134 0.0
869 8.7
231 0.0
424 5.5
67 0.0
50 0.0
40 0.0
12 0.0
472 35
287 0.8
43 0.1
13 0.1
247 3.4
243 1.3
57 0.2
324 9.6
115
-------
Veh
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
CS 03
HC 00
190 0.9
310 01
452 0.7
662 1.0
1913 2.9
205 1.8
236 1.1
194 3.5
94 1.7
201 1.9
345 0.2
307 0.1
166 0.4
614 0.6
162 0.3
1885 4.6
307 0.1
1116 0.1
103 0.2
331 0.3
450 5.2
236 8.1
ffN/A ffN/A
266 0.2
260 4.0
470 3.6
ffN/A ffN/A
477 0.8
245 0.2
142 0.2
279 2.6
327 0.1
206 0.2
426 0.3
248 3.4
259 0.5
503 1.5
183 0.4
311 7.2
269 0.9
238 0.5
378 0.9
636 0.1
557 9.1
305 0.7
209 0.5
841 0.1
1576 0.1
88 0.2
CS 05
HC CO
54 0.1
189 0.2
371 1.4
958 1.0
830 2.6
79 0.0
278 0.6
144 0.6
403 3.9
125 0.2
104 0.1
158 0.4
128 0.5
154 0.4
219 0.8
499 0.9
312 0.0
1101 0.1
1553 0.1
313 0.5
414 7.3
88 0.1
ffN/A ffN/A
282 0.8
260 4.1
127 0.1
ffN/A ffN/A
1037 0.1
58 0.0
67 0.1
266 3.2
78 0.1
423 0.6
506 2.8
252 2.6
704 0.8
575 0.5
125 0.5
171 1.2
413 0.4
207 0.6
94 0.3
158 0.1
412 3.8
212 0.4
462 0.6
1897 3.0
957 0.1
103 0.1
CS 07
HC CO
200 03
194 03
72 02
920 1.3
332 2.7
46 0.2
40 0.1
226 5.2
666 0.4
63 0.0
18 0.0
20 0.0
311 2.4
1870 2.1
153 0.1
353 5.6
132 0.1
235 0.4
384 0.5
107 0.1
703 10.4
19 0.0
64 0.1
176 0.4
137 0.9
1682 7.5
ffN/A ffN/A
350 0.6
162 2.4
24 0.0
186 2.7
18 0.0
331 0.3
362 1.8
303 2.4
243 0.4
268 0.1
14 0.0
69 0.4
176 0.5
36 0.1
6 0.0
686 8.3
498 4.8
131 0.4
142 0.2
368 1.3
73 0.0
53 0.0
CS 10
HC CO
167 0.4
158 0.1
83 08
879 13
106 0.2
38 0.1
98 0.2
315 8.3
699 0.3
99 0.0
21 0.0
20 0.0
309 2.3
1885 1.9
119 0.1
336 5.7
58 0.2
230 0.5
283 0.5
107 0.1
996 10.4
17 0.0
65 0.1
154 0.4
126 0.5
1927 6.7
ffN/A ffN/A
293 0.3
188 1.9
27 0.0
186 2.7
18 00
308 03
324 1.6
354 2.9
151 0.2
200 0.0
12 0.0
150 0.6
156 0.5
26 0.1
38 0.2
746 8.5
555 S.6
135 0.8
135 0.2
280 1.0
67 0.0
39 0.0
CS 12
HC CO
25 0.0
182 01
842 6.5
1965 0.4
421 1.2
195 2.5
18 0.0
315 5.3
1080 0.2
213 1.2
33 0.0
14 0.0
112 0.1
155 0.8
223 0.5
352 4.8
48 0.0
219 0.1
447 0.3
53 0.0
721 8.6
24 0.0
184 1.5
182 0.2
429 80
1927 7.9
ffN/A ffN/A
928 0.3
208 0.1
30 0.0
210 3.3
32 0.0
493 0.6
233 0.9
318 5.6
109 0.7
125 0.0
81 0.1
250 2.7
163 0.1
29 0.1
12 0.0
205 2.3
720 10.0
399 5.5
26 0.0
749 1.3
97 0.0
20 0.0
CS 15
HC CO
26 0.0
193 02
58 00
1097 1.3
313 3.9
92 0.2
25 00
597 10.2
1045 02
228 1.5
24 0.0
17 0.0
129 0.4
1755 2.1
204 0.4
371 5.2
39 0.1
199 0.4
317 0.3
68 0.1
633 8.6
21 0.0
174 1.1
164 0.3
249 3.5
1662 7.5
»N/A ffN/A
348 0.2
258 0.8
29 0.1
184 2.9
31 0.0
421 0.5
218 0.4
239 4.4
238 0.1
219 0.0
21 0.0
162 0.6
246 0.4
34 0.1
37 0.1
265 4.6
613 6.8
165 0.4
52 0.1
354 1.4
81 0.0
22 0.0
XL 02
HC CO
192 0.1
76 0.0
30 0.1
1141 1.1
167 0.2
87 0.4
103 0.1
309 7.4
1483 0.2
80 0.0
18 0.0
1 1 0.0
214 2.8
1885 2.7
110 0.3
550 9.3
15 0.0
186 0.4
235 0.4
39 0.1
1079 10.4
18 0.0
375 6.6
130 0.3
458 10.2
1927 7.8
30 0.0
512 0.6
221 0.9
29 0.0
244 6.0
18 0.0
315 0.3
586 2.7
289 6.4
265 0.5
100 0.1
3 0.0
108 1.3
139 0.1
10 0.0
66 0.2
259 6.0
527 10.0
190 0.9
43 0.1
363 1.7
60 0.0
92 0.2
XL 05
HC CO
338 0.3
87 00
45 0.1
1125 1.5
261 1.4
57 0.2
29 0.1
643 10.2
870 0.2
102 0.0
1,8 0.0
9 0.0
556 7.2
1885 2.1
121 0.3
507 9.1
19 0.0
173 0.3
196 0.5
41 0.1
774 10.4
20 0.0
379 6.5
155 0.2
405 9.6
1927 8.6
34 0.0
179 0.2
181 1.1
26 0.1
230 5.6
22 0.0
287 0.3
214 0.4
274 5.7
205 0.3
123 0.1
39 0.1
153 0.6
192 0.2
13 0.0
10 0.1
260 6.1
435 9.1
133 0.6
36 0.1
236 1.5
S3 0.0
33 0.2
XI 02
HC CO
24 00
287 04
56 04
1965 06
322 2.1
101 02
21 0.0
377 69
1332 02
258 1.4
26 00
28 00
68 03
184 0.5
216 05
303 4.0
25 00
277 0.2
641 02
100 0.1
510 7.1
30 0.0
60 1.2
172 03
320 88
1927 7.9
50 0.0
541 02
239 03
79 0.0
233 5.4
82 0.0
453 0.6
393 1.2
363 6.3
196 0.6
145 0.0
434 5.3
277 4.7
102 0.1
26 0.0
188 0.5
215 4.4
779 10.0
678 8.1
25 0.0
532 0.8
274 0.0
IS 00
XI 05
HC CO
82 0.1
291 05
32 0.1
1123 1.2
412 2.3
67 0.1
81 0.1
611 10.2
1183 0.2
265 1.4
23 0.0
37 02
131 1.5
1885 2.2
211 0.5
498 6.9
24 0.0
393 0.3
279 0.5
152 0.2
453 6.7
27 0.1
311 7.4
188 0.3
287 7.7
1927 6.6
81 0.6
816 0.2
205 0.6
36 0.1
221 5.1
35 0.0
374 0.6
201 0.2
320 5.9
262 0.5
135 0.0
168 01
270 1.3
235 0.1
28 0.1
40 0.1
198 4.7
621 8.0
317 1.2
66 0.0
335 1.1
188 0.0
20 0.0
XI 07
HC CO
142 0.4
277 0.2
50 03
804 12
103 0.7
132 1.9
117 0.2
764 10.2
821 0.3
108 0.0
20 0.0
16 0.0
546 7.6
1885 1.9
163 0.3
1885 10.1
191 0.3
241 0.3
235 0.4
114 0.1
706 10.4
23 0.0
330 5.2
137 0.6
328 8.8
1927 8.4
37 0.0
274 0.7
186 1.6
37 0.2
233 5.5
39 0.1
343 0.5
193 0.9
510 5.3
152 0.4
248 0.1
95 0.2
180 1.0
218 0.6
15 0.0
IS 0.1
214 S.6
525 8.4
161 0.7
158 0.2
179 1.4
65 0.0
58 0.3
XI 1 1
HC CO
208 0.4
230 02
45 02
547 1.2
57 0.0
132 16
47 01
689 101
715 0.2
81 0.0
20 0.0
18 0.2
618 102
1306 14
154 0.3
1885 10.1
21 0.1
198 03
240 0.5
87 0.1
656 10.4
22 0.1
342 55
113 0.4
326 8.8
1927 8.4
34 0.0
135 0.4
179 1.6
37 0.1
250 6.0
19 0.0
335 0.4
187 0.9
553 5.5
128 1.0
193 0.1
63 0.1
148 0.4
153 0.2
13 0.0
6 0.1
214 5.5
573 7.7
157 0.5
120 0.2
220 1.5
56 0.0
18 0.0
RS 02
HC CO
420 0.2
111 01
32 0.0
1149 0.9
304 1.9
66 06
1 1 0.0
388 8.5
661 03
74 0.0
20 0.0
13 00
535 6.4
1885 29
113 0.3
802 92
27 0.0
139 0.1
279 0.4
42 0.1
907 10.4
77 0.4
ffN/A #N/A
130 0.4
400 9.9
1927 8.6
29 00
454 0.3
196 1.5
22 0.0
276 7.4
17 0.0
260 0.3
485 2.4
255 5.6
176 0.4
170 0.1
13 0.1
151 2.3
109 0.1
5 0.0
44 0.2
255 6.3
594 10.0
252 1.6
43 0.2
ffN/A ffN/A
94 0.0
19 0.0
RS 05
HC CO
38 0.0
97 0.1
46 0.1
999 1.5
332 4.7
39 0.1
89 0.6
699 10.1
579 0.3
84 0.0
24 0.0
25 0.2
179 1.8
1259 1.7
117 0.3
566 8.4
17 0.0
101 0.1
268 0.5
39 0.1
558 10.4
165 39
ffN/A »N/A
165 0.4
358 9.5
1037 0.5
64 0.3
149 0.3
36 0.0
22 0.0
521 7.6
22 0.0
283 0.3
467 2.5
261 4.5
261 0.7
90 0.0
73 0.1
92 0.6
183 0.2
4 0.0
141 0.2
250 62
507 10.0
135 0.5
32 0.1
ffN/A ffN/A
96 0.1
19 0.0
1 16
-------
Veh
401
402
403
404
405
406
407
408
409
410
502
503
504
505
506
507
506
509
510
511
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
701
702
703
704
705
706
707
708
709
CS 03
HC CO
330 5.0
493 1.3
329 0.2
277 2.5
185 0.2
277 0.2
225 1.1
306 2.9
500 5.0
297 3.3
487 5.5
600 10.6
217 2.9
357 0.5
600 11.2
309 5.9
369 1.7
412 5.4
1080 0.1
322 4.7
570 4.2
739 4.1
552 2.6
429 0.5
425 4.4
315 2.9
103 0.3
324 1.4
267 3.4
337 3.2
412 3.3
269 0.6
203 2.5
505 7.9
330 3.4
254 1.0
254 2.1
804 10.0
321 3.6
95 0.3
178 0.7
117 0.9
380 9.4
176 2.2
646 0.2
352 3.9
302 5.1
246 0.9
138 1.9
CS 05
HC CO
408 3.0
363 5.0
277 0.1
213 0.3
500 0.1
500 0.1
216 0.2
237 2.0
394 5.0
232 2.1
192 0.3
600 7.0
94 2.9
199 0.1
600 8.0
158 2.4
256 3.9
347 4.6
126 0.0
276 3.9
414 3.1
548 2.2
202 0.0
211 0.2
192 2.6
136 1.3
82 0.3
178 0.3
105 0.8
177 0.5
254 2.7
246 6.3
97 1.5
384 6.8
155 2.5
113 0.1
201 3.4
628 10.0
84 0.5
36 0.1
128 0.4
100 0.6
182 3.2
103 0.6
371 0.3
123 1.6
59 0.3
152 0.4
152 0.6
CS 07
HC CO
136 1.0
70 0.1
344 0.1
180 0.1
500 0.1
500 0.1
215 0.2
63 0.0
323 3.1
73 0.0
444 1.4
600 7.7
52 2.0
53 0.0
600 9.6
9 0.1
215 1.1
282 3.1
166 0.5
10 0.0
423 4.1
1013 8.2
27 0.0
118 0.0
199 2.7
144 2.0
52 0.2
551 0.5
30 0.1
10 0.0
9 0.0
366 6.5
183 4.5
406 5.1
188 2.3
114 0.2
253 4.2
634 7.5
25 0.0
217 6.3
16 0.0
56 0.0
152 0.8
18 0.0
180 0.1
140 1.3
27 0.0
3 0.0
30 0.0
CS 10
HC CO
141 1.0
65 0.1
322 0.
203 0.
297 0.2
500 0.
212 0.
62 0.
334 3.0
93 0.1
522 2.5
600 7.8
56 2.5
48 0.0
600 9.0
5 0.0
211 1.5
269 3.1
91 1.1
13 0.0
453 4.1
1013 8.2
42 0.0
90 0.0
199 2.9
147 2.2
56 0.2
546 0.4
25 0.1
7 0.0
7 0.0
373 6.4
144 4.3
501 4.3
199 2.4
158 0.2
239 4.1
596 7.2
26 0.0
229 6.2
19 0.0
S3 0.0
147 0.8
9 0.0
279 0.1
143 1.2
34 0.0
0 0.0
17 0.0
CS 12
HC CO
153 0.2
356 2.1
500 0.2
350 0.3
292 0.2
500 01
269 0.2
263 0.4
250 0.4
130 0.1
334 O.S
581 70
38 1.5
267 0.4
600 9.4
38 0.0
234 2.2
243 1.5
313 1.2
73 0.0
459 2.8
669 1.8
563 0.0
577 0.2
195 2.7
172 1.9
132 0.2
733 0.2
31 0.1
22 0.0
43 0.0
384 5.1
151 4.4
418 3.2
190 1.6
469 0.4
280 4.3
629 6.1
73 0.0
265 6.1
39 0.0
146 0.7
130 0.5
16 0.0
333 0.0
152 0.9
102 0.0
5 0.0
38 0.1
CS 15
HC CO
135 0.8
100 0.1
314 0.1
271 0.1
500 0.2
500 0.1
231 0.2
104 01
226 0.5
153 0.1
307 04
536 6.2
48 15
276 0.5
600 8.3
48 0.0
224 2.1
265 2.7
301 3.0
20 0.1
432 3.0
665 2.2
378 0.1
472 0.1
203 2.7
166 2.1
133 0.3
575 0.6
41 01
24 0.0
60 0.0
399 6.0
155 3.8
855 8.8
206 1.9
234 0.0
262 3.0
603 6.2
108 0.1
277 5.7
96 0.0
131 0.3
137 0.8
23 0.0
352 0.0
145 1.1
98 0.0
2 0.0
43 0.1
XL 02
HC CO
79 0.1
86 01
181 0.1
149 0.1
111 02
500 0.1
211 02
48 01
341 3.1
49 01
145 02
600 8.2
38 2.4
210 03
600 8.9
0 0.0
45 0.1
0 0.0
8 0.1
0 0.0
372 2.9
1013 8.0
968 2.2
41 0.0
168 3.4
110 1.9
25 0.1
392 0.4
19 0.1
0 0.0
0 0.0
314 5.8
61 37
663 8.5
164 2.3
45 0.0
232 4.4
566 6.4
10 0.0
243 5.8
6 0.0
35 0.0
112 0.5
2 0.0
148 0.1
122 0.9
0 0.0
0 0.0
0 0.0
XL 05
HC CO
44 0.1
371 4.3
160 01
172 0.1
123 0.2
500 0.1
348 0.2
48 0.1
291 18
65 0.1
250 0.7
593 76
35 19
114 0.1
600 8.0
0 0.0
28 0.1
2 0.1
42 0.2
0 0.0
276 2.9
1013 8.0
904 1.7
65 0.0
167 3.3
130 2.5
35 0.2
433 0.4
19 0.1
0 0.0
0 0.0
332 6.0
60 3.8
561 5.8
181 2.3
50 0.1
226 4.3
547 5.6
17 0.0
250 6.0
7 0.0
45 0.0
128 0.8
0 0.0
141 0.1
109 1.1
1 0.0
0 0.0
0 0.0
XI 02
HC CO
62 01
307 08
463 02
448 01
140 02
500 0.1
254 03
239 04
276 13
228 04
520 12
600 81
43 24
240 03
600 9.2
62 0.1
57 0.0
40 00
769 06
74 00
492 3.7
598 15
1015 1.1
585 0.1
165 2.3
156 2.4
141 0.7
511 0.2
43 01
249 08
21 00
340 55
183 6.3
333 25
185 26
548 1.0
285 4.9
541 5.2
108 0.2
257 6.1
120 0.1
142 O.S
27 0.0
50 0.1
214 01
127 06
30 0.0
6 0.0
9 0.0
XI 05
HC CO
92 0.1
299 08
302 01
307 0.0
186 0.2
500 0.1
254 0.2
289 0.9
264 1.2
206 0.1
371 1.2
589 75
61 1.7
261 04
600 8.6
59 01
104 01
45 0.2
438 5.8
26 0.1
546 3.2
609 1.8
1015 1.1
551 0.1
235 2.0
ISO 2.3
113 0.3
524 0.6
67 0.2
224 03
72 0.0
497 6.0
185 6.7
497 8.4
346 2.4
394 0.2
345 4.5
SS9 5.3
143 0.1
286 5.9
152 0.1
138 0.4
92 O.S
85 0.2
197 0.0
121 0.8
33 0.0
3 0.0
13 0.0
XI 07
HC CO
56 0.1
87 0.1
240 0.1
201 0.0
163 0.2
500 0.1
251 0.2
108 01
274 1.4
121 0.1
413 1.3
534 7.1
83 3.0
261 0.5
600 92
25 0.1
67 0.2
41 0.2
211 04
0 00
402 3.2
713 47
1013 14
287 00
165 2.4
142 2.4
46 0.2
512 0.4
21 0.0
18 0.0
8 0.0
360 6.1
140 5.6
419 35
219 2.7
140 0.1
243 4.4
516 5.5
36 0.0
26 0.0
39 0.0
67 0.0
112 0.8
11 0.0
214 0.1
119 1.1
3 0.0
3 0.0
1 0.0
XI 1 1
HC CO
50 0.1
77 00
251 0.1
189 0.0
234 0.2
500 01
339 02
69 01
289 18
107 0.1
409 15
585 75
77 26
219 0.4
600 88
13 01
81 0.1
101 03
144 03
0 00
399 3.1
804 57
963 1.6
242 00
163 2.2
146 22
40 02
525 0.4
19 0.1
17 0.0
7 0.0
352 58
166 68
429 32
232 2.6
218 0.2
247 4.3
520 5.3
31 0.1
14 0.0
21 0.0
68 0.0
112 07
8 0.0
184 0.1
121 1.0
3 00
0 0.0
2 0.0
RS 02
HC CO
77 0.1
79 0.1
174 0.1
146 0.1
158 0.2
500 01
482 0.3
49 0.1
391 45
89 0.1
205 03
582 78
30 18
67 03
600 84
0 00
30 00
13 02
77 08
0 0.0
378 3.1
1013 8.4
901 1.8
43 00
160 28
136 23
26 02
381 04
32 02
135 0.1
0 0.0
340 60
102 58
379 34
181 2.8
80 02
252 52
489 5.9
a o.o
1 0.0
4 00
36 0.0
76 01
2 0.0
160 0.1
98 0.6
0 0.0
0 0.0
0 0.0
RS 05
HC CO
47 0.1
320 2.9
198 0.1
186 0.1
240 0.2
500 0.1
443 0.2
53 0.1
369 4.3
88 0.1
304 08
389 5.7
28 1.6
114 0.4
554 78
0 0.0
31 00
1 1 0.2
66 0.2
0 00
312 3.1
1013 84
832 12
95 0.0
171 3.3
123 2.2
44 03
406 04
30 0.2
119 0.2
0 0.0
340 6.0
110 6.6
426 3.4
211 3.3
93 0.2
205 4.9
454 5.6
15 0.0
6 0.0
5 0.0
41 0.0
118 09
2 00
151 0.1
105 0.8
0 0.0
0 0.0
1 00
117
-------
APPENDIX D
AET Errors of Commission in the CTP Sample
V«h
31
223
249
14
20
24
206
221
250
260
701
321
336
360
11
17
21
37
39
SO
224
251
306
309
322
330
347
348
357
509
511
704
708
MIT
82
82
82
85
85
84
86
84
83
84
86
82
82
82
85
86
84
85
85
84
86
86
86
84
86
84
84
86
86
85
86
86
84
MFR
AMC
FORD
FORD
TOYT
MAZD
TOYT
POFD
FORD
FORD
FORD
TOYT
GM
am
Out
VW
SUBA
TOYT
SUBA
MAZD
VW
FORD
FORD
am
am
am
am
am
am
am
MITS
MITS
TOYT
TOYT
CID
258
302
351
89
91
95
300
140
122
122
91
250
151
151
109
109
122
109
80
117
302
183
151
151
121
151
151
151
151
97
96
122
122
Typ«
LOT
LDV
LOT
LDV
LDV
LDV
LOT
LOT
LOT
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KlliU
53
46
45
23
13
27
16
44
70
35
5
60
53
56
27
17
59
29
18
81
28
14
5
61
18
60
25
17
37
26
16
32
32
Quota Grp
Garb 81-82
Carb 81-82
Carb 81-82
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Ft 81-82
Fl 81-82
F\ 81-82
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
HCCorl
1.70
0.41
1.70
0.41
0.41
0.41
0.80
0.80
1.70
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
FTPMC
1.13
0.33
0.88
0.23
0.10
0.23
0.62
0.54
1.44
0.73
0.24
0.33
0.28
0.29
0.31
0.14
0.58
0.20
0.26
0.71
0.39
0.16
0.21
0.25
0.16
0.24
0.21
0.40
0.35
0.24
0.20
0.26
0.16
COCert
18.0
7.0
18.0
3.4
3.4
3.4
10.0
10.0
18.0
10.0
3.4
7.0
7.0
7.0
3.4
3.4
10.0
3.4
3.4
10.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
FTPCO
14.4
1.8
5.1
1.4
3.1
1.7
0.9
4.1
11.8
8.0
1.9
5.2
2.7
4.2
2.9
2.1
8.4
2.6
1.6
7.9
2.0
2.2
1.8
3.3
1.8
2.9
2.6
1.2
3.3
1.4
3.0
1.7
1.9
AETHC
578
161
1126
176
63
231
250
124
214
146
262
264
517
1062
82
168
337
465
1743
225
449
126
225
500
239
325
331
223
548
307
127
335
344
AET CO
0.4
1.3
0.0
1.6
2.4
0.5
0.0
1.4
3.5
2.6
0.1
0.0
0.2
1.0
1.6
1.4
2.3
0.4
0.1
4.4
0.1
1.3
0.4
0.5
0.9
0.3
0.7
0.1
0.7
6.7
2.1
0.4
0.2
XL30HC
20
67
53
8
1
22
67
32
64
72
6
18
512
92
3
1
226
105
3
3
43
17
19
95
11
39
265
100
43
0
0
2
0
XL30CO
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.0
0.0
0.0
0.6
0.2
0.0
0.0
0.7
0.3
0.0
0.0
0.0
0.0
0.1
0.1
0.0
0.1
0.5
0.1
0.1
0.0
0.0
0.0
0.0
-------
APPENDIX E:
PER-REP AIR EMISSION REDUCTIONS FOR ALL
SYSTEMS: BY QUOTA GROUP
These tables summarize the HC and CO emission reductions
per repair, due to repairs to the systems on vehicles in the
quota groups listed, in g/mi, as measured by the LA4
transient cycle. Figures indicate averages derived from
isolatable repairs.
Carbureted 1981-1982
SYSTEM
REPAIRED
Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
SIMPLE AVERAGES
ISOLATABLE
N
11
46
18
11
55
3
30
3
1
42
220
A HC
0.41
0.67
0.24
0.18
0.31
-3.31
1.12
0.39
1.23
1.06
0.59
REPAIRS
A CO
-0.7
8.8
0.0
2.6
6.8
-11.0
7.7
3.9
30.9
22.3
9.0
Vehicles
MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N
18
53
24
18
70
8
35
3
6
56
291
A HC
0.72
0.60
0.11
-0.24
0.18
-1.65
1.00
0.39
-0.05
0.90
-
t- ratio
1.6
2.6
0.3
-0.6
0.8
-2.8
3.5
0.4
-0.1
4.1
-
A CO
0.6
9.1
-0.3
-5.4
4.2
-4.7
9.1
3.9
2.4
19.4
-
t-ratlo
0.1
3.4
-0.1
-1.1
1.5
-0.7
2.7
0.4
0.3
7.6
-
Carbureted 1983-1986
SYSTEM
REPAIRED
Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
SIMPLE AVERAGES
ISOLATABLE
N
3
20
12
4
8
1
10
0
2
25
85
A HC
0.02
0.87
-0.20
0.08
0.16
-0.11
1.32
-
-0.01
0.42
0.47
REPAIRS
A CO
2.4
11.1
-6.5
0.3
4.1
-4.7
6.9
-
0.3
13.2
6.8
Vehicles
MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N
5
29
17
5
9
2
13
3
4
26
113
A HC
0.45
0.92
-0.33
-0.06
0.16
-0.24
1.53
1.44
-0.12
0.42
-
t-ratlo
1.0
4.4
-1.1
-0.1
0.5
-0.3
5.1
2.3
-0.2
2.0
-
A CO
12.8
13.0
-7.3
-1.0
3.9
-7.6
8.0
24.8
-3.9
13.5
-
t-ratlo
1.4
3.1
-1.2
-0.1
0.6
-0.5
1.3
2.0
-0.4
3.2
-
119
-------
Fuel
SYSTEM
REPAIRED
Induction
Fuel Meter
ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
SIMPLE
Injected 1981-1982 Vehicles
AVERAGES
ISOLATABLE
N A
0
6 1
7 0
2 -0
4 4
0
5 2
0
2 -0
14 1
40 1
HC
-
.10
.06
.06
.85
-
.11
-
.02
.06
.29
REPAIRS
A CO
-
22.8
0.1
0.0
28.3
-
6.2
-
0.4
32.0
18.3
MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N
1
6
8
2
4
0
5
0
4
15
45
A HC
0.02
1.11
0.06
-0.06
8.27
-
2.28
-
1.04
1.21
-
t- ratio
0.0
1.1
0.1
0.0
4.9
-
1.9
-
0.8
2.0
-
A CO
-0.2
22.8
0.1
0.0
38.0
-
8.6
-
27.9
35.9
-
t-ratlo
0.0
1.7
0.0
0.0
1.6
0.5
-
1.4
4.1
-
SYSTEM
REPAIRED
Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
Fuel
SIMPLE
Injected 1983-1
AVERAGES
ISOLATABLE
N
0
22
26
2
10
0
15
1
5
75
156
A
2
0
0
-0
0
0
1
0
0
HC
-
.07
.37
.00
.65
-
.98
.08
.57
.22
.56
REPAIRS
A CO
-
21.6
-0.3
-0.1
-0.3
-
7.2
0.3
7.1
14.2
10.7
986
Vehicles
MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N
4
29
29
2
10
1
15
2
7
82
181
A
-1
1
0
0
-0
1
0
0
1
0
HC
.77
.78
.72
.00
.82
.77
.94
.08
.53
.33
-
t- ratio
-1.2
3.6
1.4
0.0
-0.9
0.5
1.4
0.0
1.5
1.1
-
A CO
-14.0
17.2
1.9
-0.1
-1.0
13.3
6.8
0.3
6.4
14.1
-
t-ratlo
-0
3
0
0
-0
0
0
0
0
4
.8
.2
.3
.0
.1
.3
.9
.0
.6
.3
120
-------
HC Benefit: per System Repair
Carbureted MY 81-82 Vehicles
Cart) 81 -82
6.00 T
4.50 4-
HC
g/rrt
INDT FUEL IGNT EGR AIR
EXH EVAP ENG 3WAY ALL
-4.50 J-
HC Benefit peg System Repair "
Carbureted MY 83-flfi Vehicles
Garb 83-86
HC
g/mi
6.00 y
4.50 -
3.00
1.50-
0.00
-1.50
-3.00 J-
-4.50 1
INDT FUEL IGNT EGR AIR PCV EXH EVAP ENG SWAY ALL
121
-------
Averaye HC Benefit per System Repair
Fuel Injected MY 81-82 Vehicles
HC
FI81-82
6.00 y
4.50
3.00
1.50
0.00
-1.50
-3.00
-4.50 -
INDT FUEL IGNT EGR AIR PCV EXH EVAP ENG SWAY ALL
Average HC Benefit per System Repair
Fuel In-iected MY 83-86 Vehicles
6.00 -P
4.50 -
3.00
1.50
-1.50
-3.00
-4.50 J-
INDT FUEL IGNT EGR
Fl 83-86
PCV EXH EVAP ENG SWAY ALL
122
-------
AveraQft CO
System Repair
Carhiir«»l-Ari MV
Cart 81-82
30 T
20 +
CO
g/mi
-10-L
INDT FUEL IGMT EGR AIR
EXH EVAP ENG 3WAY ALL
Average CO Benefit per System Repair
Carbureted MY 83-86 Vehicles
Garb 83-86
40-r
30 +
CO
g/mi
EGR AIR "PW EXH EVAP ENG SWAY ALL
-20 -L
123
-------
Average CO Benefit per System Repair
Fuel In-ieeted MY 81-82
FI81-82
40 T
CO
g/mi
0
-10-.
-20
INDT FUEL IGNT EGR AIR PCV EXH EVAP ENG 3WAY ALL
Averae CO Benefit
Sstem Reair
Fuel Injected MY 83-86 Vehicles
Fl 83-86
CO
g/mi
40 j
30
20
10
0
-10
-20--
INDT FUEL IGNT EGR AIR PCV EXH EVAP ENG 3WAY ALL
124
-------
APPENDIX F:
STATISTICALLY
PBR-RBPAIR EMISSION REDUCTIONS FOR
SIGNIFICANT SYSTEMS: BY QUOTA GROUP
This table summarizes the HC and CO emission reductions
per repair, due to repairs to the systems on vehicles in the
quota groups listed, in g/mi, as measured by the LA4
transient cycle.
SYSTEM
REPAIRED
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N A HC A CO
MULTIPLE LINEAR REGRESSION
ALL REPAIRS OF THESE SYSTEMS
N A HC t-ratlO
A CO
t-ratio
Carbureted 1981-1982 Vehicles
Fuel Meter
Exhaust
3-Way
46 0.67 8.8
30 1.12 7.7
42 1.06 22.3
53 0.61 2.7
35 1.05 3.6
56 0.91 4.2
9.2
9.6
19.2
3.5
2.9
7.7
Carbureted 1983-1986 Vehicles
Fuel Meter
Exhaust
3-Way
20 0.87 11.1
10 1.32 6.9
25 0.42 13.2
Fuel Injected 19
Fuel Meter
Exhaust
3-Way
6 1.10 22.8
5 2.11 6.2
14 1.06 32.0
Fuel Injected 19
Fuel Meter
Exhaust
3-Way
22 2.07 21.6
15 0.98 7.2
75 0.22 14.2
29 1.01 5.0
13 1.51 5.1
26 0.42 2.0
14,6
7.4
13.3
3.6
1.2
3.2
81-1982 Vehicles
6 1.11 0.9
5 2.28 1.5
15 1.28 1.7
22.8
8.6
37.8
1.7
0.5
4.4
83-1986 Vehicles
29 1.67 3.5
15 0.94 1.4
82 0.38 1.3
16.2
6.8
14.3
3.2
0.9
4.5
125
-------
APPENDIX 6:
PBR-REPAIR EMISSION REDUCTIONS FOR ALL
SUBSYSTEMS
This table summarizes the HC and CO emission reductions
per repair, due to repairs to the subsystems listed, in g/mi,
as measured by the LA4 transient cycle.
SUBSYSTEM
REPAIRED
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
INDUCTION SYSTEM
Htd Air Door
Temp Sensors
Air Filter
Hoses
Other (Indt)
1
0
3
4
1
A HC
0.06
-
-0.01
1.13
0.00
A CO
-1.6
-
2.2
-2.3
-0.2
FUEL METERING SYSTEM
Cart) Assembly
Fuel Meter Tune
Idl Spd Sole
Fuel Inj
Hoses
Other
Chk Adj Vacm
Vac Diaphrms
Other (Chk)
22
30
0
19
3
4
1
2
1
IGNITION SYSTEM
Dist Assembly
Igni Tune Items
Vac Adv Assmb
Spk Delay Dev
Elect Tim Mod
Hoses
Wir/Hms/Fuse
Other
EGR SYSTEM
Vatv Assembly
Delay Solnoid
Cool Temp Sen
Hoses
Other
3
38
2
1
0
2
0
0
8
0
1
4
2
1.03
0.61
-
2.35
-0.09
2.35
0.01
-0.06
-0.01
0.00
0.34
-0.19
0.28
-
0.01
-
-
-0.01
-
-0.02
0.05
0.62
11.9
11.6
-
24.2
0.2
29.7
0.5
-1.0
-0.4
-0.5
0.4
-1.2
10.4
-
-0.5
-
-
0.6
-
0.3
-1.4
11.3
AIR INJECTION SYSTEM
Pump Assembly
Byps/Dump Vlv
Diverter Vlv
2
0
5
7.93
-
-1.45
41.1
-
-3.4
N
2
2
15
6
3
27
43
1
20
7
11
1
5
2
11
54
5
1
1
2
2
2
12
1
1
8
5
5
3
13
MULTIPLE
ALL
A HC
-0.65
-0.40
-0.57
1.16
1.18
0.89
0.49
-0.65
2.22
0.08
0.70
0.01
0.10
1.51
-0.22
0.46
-0.15
0.28
0.56
0.01
-0.53
-0.48
-0.01
-0.89
-0.02
-0.05
0.41
5.31
-0.31
-0.67
LINEAR
REPAIRS
t- rat lo
-0.4
-0.2
-0.9
1.4
0.8
2.3
1.6
-0.3
5.2
0.1
1.1
0.0
0.1
1.0 -
-0.4
1.7
-0.1
0.1
0.3
0.0
-0.3
-0.3
0.0
-0.4
0.0
-0.1
0.4
4.7
-0.2
-1.0
REGRESSION
A CO
-6.0
3.3
-2.8
7.7
3.2
10.8
10.0
-14.1
22.2
-5.8
10.7
0.5
1.5
20.9
-3.6
0.5
-0.6
10.4
1.2
-0.6
-4.2
-5.8
-1.0
-16.3
0.3
-1.5
7.0
22.6
12.3
-1.3
t- ratio
-0.31
0.16
-0.38
0.77
0.19
2.41
2.78
-0.61
4.40
-0.66
1.47
0.02
0.14
1.13
-0.50
0.16
-0.05
0.46
0.05
-0.04
-0.23
-0.32
-0.15
-0.60
0.01
-0.18
0.60
1.70
0.71
-0.17
126
-------
SUBSYSTEM
REPAIRED
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
A HC
A CO
AIR INJECTION SYSTEM (continued)
Check Valve
Drive Belt
Hoses
Wir/Hms/Fuse
Other
PCV SYSTEM
Valv Assembly
Filters
Hoses/Lines
Other
14
1
9
1
4
3
0
1
0
EXHAUST SYSTEM
Exh Manifold
Catalyst
Other
0
43
1
0.35
-0.04
-0.24
0.02
-0.30
-0.12
-
-9.68
-
-
1.11
0.50
3.8
0.1
0.5
0.4
-3.7
-0.8
-
-35.3
-
-
7.0
12.3
EVAPORATIVE SYSTEM
Evap Canister
Canister Purg
Hoses
Other
1
2
1
0
ENGINE ASSEMBLY
Eng Assembly
Cooling Sys
Valve Adj
Belt Tension
Hoses
Eng Oil
Other
2
1
1
0
1
1
2
0.08
0.58
0.02
-
0.63
-0.06
0.11
-
5.99
-0.16
0.46
0.3
5.8
0.2
-
16.6
-1.5
1.5
-
28.7
-0.8
1.9
THREE-WAY CATALYST SYSTEM
ECU
O2 Sen
Load Sensor
Eng Spd Sen
Cool Temp Sen
EGR Postn Sen
A/F Cntri Act
Air Bypas Sen
Air Divrt Act
ISC Sys
Hoses
MAT Sen
Wir/Hms/Fuse
Other
ALL
14
69
11
1
4
1
5
1
1
1
6
3
6
1
372
0.40
0.80
0.61
0.53
0.86
0.06
2.46
2.76
0.12
0.20
0.37
-0.55
0.43
-0.02
0.70
10.7
20.7
23.2
3.4
27.4
-12.8
32.6
28.2
-0.4
2.0
-3.7
10.2
26.7
-0.7
10.7
N
31
5
27
2
7
3
4
3
1
10
56
2
1
2
3
2
3
7
4
1
1
2
3
19
82
22
2
8
1
5
3
4
1
18
3
9
2
630
MULTIPLE
ALL
A HC
0.33
-2.01
-0.06
1.39
-0.46
-0.12
0.54
-3.55
-0.34
0.42
1.09
0.77
0.08
0.58
0.56
1.04
0.66
-0.24
-0.27
3.72
5.99
-0.14
0.36
0.67
0.91
0.02
0.27
0.24
0.06
2.46
1.77
-0.33
0.20
-0.58
-0.55
0.48
0.45
-
LINEAR
REPAIRS
t- ratio
0.8
-1.8
-0.1
1.0
-0.6
-0.1
0.4
-2.9
-0.2
0.6
3.9
0.6
0.0
0.4
0.5
0.7
0.6
-0.3
-0.2
1.8
3.2
-0.1
0.3
1.5
4.2
0.0
0.2
0.4
0.0
2.9
1.3
-0.3
0.1
-1.1
-0.5
0.7
0.3
-
REGRESSION
A CO
4.3
-3.7
1.0
17.8
-6.4
-0.8
3.9
-10.8
-8.2
0.8
7.8
4.1
0.3
5.8
-0.6
19.8
14.0
-8.2
-7.4
105.1
28.7
2.9
0.2
14.0
21.9
17.1
-3.7
11.4
-12.8
32.6
21.8
-4.4
2.0
-11.9
10.1
26.2
20.5
-
t- ratio
0.88
-0.28
0.20
1.09
-0.67
-0.06
0.27
-0.73
-0.36
0.10
2.36
0.25
0.01
0.37
-0.05
1.06
1.03
-0.83
-0.57
4.28
1.28
0.16
0.02
2.59
8.60
2.98
-0.24
1.42
-0.57
3.25
1.36
-0.32
0.09
-1.86
0.78
3.27
1.25
-
127
-------
APPENDIX B: PER-REP AIR EMISSION REDUCTIONS FOR STATISTCALLY
SIGNIFICANT SUBSYSTEMS: BY QUOTA GROUP
This table summarizes the HC and CO emission reductions
per repair, due to repairs to the subsystems on vehicles in
the quota groups listed, in g/mi, as measured by the LA4
transient cycle. Figures indicate the averages derived from
isolatable repairs.
SUBSYSTEM
REPAIRED
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N A HC A CO
MULTIPLE LINEAR
REGRESSION
ALL REPAIRS OF THESE SUBSYSTEMS
N A HC t-ratlo
A CO
t-ratlo
Carbureted 1981-1982 Vehicles
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
O2 Sensor
Load Sensor
ALL
15 1.34 16.2
12 0.37 7.6
0 -
19 0.85 6.2
9 1.42 21.9
15 0.32 9.6
1 2.13 76.7
149 0.55 8.3
18 1.06 2.8
19 0.61 1.7
0 -
28 1.07 3.6
11 1.17 2.5
21 0.38 1.1
1 2.13 1.3
291
13.1
11.5
-
9.6
17.9
12.8
76.7
-
3.0
2.7
-
2.7
3.2
3.2
4.1
-
Carbureted 1983-1986 Vehicles
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
02 Sensor
Load Sensor
ALL
7 0.36 2.6
10 0.63 10.5
0 -
10 1.32 6.9
1 -0.13 0.6
12 0.31 5.2
2 1.53 73.7
75 0.57 9.1
9 0.92 2.5
12 0.93 3.0
0 -
12 1.51 4.8
1 -0.13 -0.1
13 0.29 1.0
3 1.05 1.7
113
10.9
15.7
-
7.4
0.6
5.0
48.8
-
1.6
2.6
-
1.2
0.0
0.9
4.1
-
Fuel Injected 1981-1982 Vehicles
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
O2 Sensor
Load Sensor
ALL
0 -
5 1.35 27.3
0 -
3 2.55 12.6
0 -
9 1.77 42.5
2 0.16 0.8
35 1.27 19.5
0 -
5 1.35 1.0
0 -
3 2.55 1.5
0 -
10 2.03 2.2
2 0.16 0.1
45 -
-
27.4
-
12.6
-
50.1
0.8
-
-
1.9
-
0.7
-
5.0
0.0
-
Fuel Injected 1983-1986 Vehicles
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
02 Sensor
Load Sensor
ALL
0 -
3 0.30 5.0
19 2.35 24.2
11 0.87 6.3
4 -1.75 -12.0
33 0.93 25.5
6 0.19 5.0
113 0.80 12.1
0
7 -0.30 -0.3
20 2.26 4.2
13 0.94 1.4
7 -0.05 -0.1
38 1.24 3.2
16 -0.72 -1.2
181
-
-4.4
23.3
6.8
7.3
27.9
1.3
-
-
-0.5
4.3
1.0
0.8
7.1
0.2
-
128
-------
Averaye HC Benefit per Subsystem Repair
Carbureted MY 81-82 Vehicles
Cart) 81 -82
HC
g/rri
3.00 j
2.50 -
2.00
1.50
1.00
0.50
0.00
-0.50
-1.00
-1.50 4-
-2.00
.. Garb FuelMtr Fuellnj Catalyst
Tune
ECU O2 Load
Sensor Sensor
All
Average HC Benefit per Subsystem Repair --
Carbureted MY 83-86 Vehicles
Carb 83-86
HC
g/rrt
3.00 -r
2.50
2.00
1.50
1.00
0.50
0.00
-0.50
-1.00
-1.50-.
-2.00
.. Cart)
Fuel Mtr Fuel Inj Catalyst
Tune
ECU
O2 Load
Sensor Sensor
129
-------
Average HC Benefit n*»r Subsystem Repair
Fuel Tn-iected MY Sl-fl? Vehirles
FI81-82
HC
g/mi
3.00 -r
2.50
2.00
1.50-
1.00
0.50
0.00
-0.50
-1.00
-1.50 ..
-2.00 -
.. Garb Fuel Mtr Fuel Inj Catalyst
Tune
ECU O2 Load
Sensor Sensor
All
Average HC Benefit per Subsystem Repair
Fuel In-iected MY 83-fifi Vehicles
Fl 83-86
HC
g/mi
3.00 y
2.50
2.00
1.50
1.00
0.50 -
0.00
-0.50
-1.00
-1.50
-2.00
Caib Fuel Mtr Fuel Inj Catalyst
Tune
O2 Load
Sensor Sensor
All
130
-------
Average CQ Benefit per Subgygt-.em Repair --
Carbureted MY 8l-fl2 Vehicles
Garb 81 -82
CO
g/rrt
80 j
70-
60-
50--
40 -
30
20-
10-
0
-10
-20
.. Caib Fuel Mtr Fuel Inj
Tune
Catalyst ECU
O2 Load
Sensor Sensor
Afl
Average CQ Benefit per Subsystem Repair
Carbureted MY 83-86 Vehicles
Cart) 83-86
CO
g/rrt
80 j
70-
60-
50-
40-
30-
20"
10-
0
-10
-20
.. Cart) Fuel Mtr Fuel Inj Catalyst
Tune
ECU
O2
Sensor
Load
Sensor
131
-------
Average CO Benefit: per Subsystem Repair
Fuel In-ierted MY 81-82 Vehicles
FI81-82
CO
g/mi
80 j
70 -
60
50 -
40
30 -
20
10
0
-10
-20
4. Garb FuelMtr Fuellnj Catalyst
Tune
ECU
O2
Sensor
Load
Sensor
All
Average CO Benefit per Subsystem Repair
Fuel Injected MY 81-82 Vehicles
Fl 83-86
CO
g/mi
80-r
70 -
60
50--
40 -
30 -
20
10 -
0 4-
-10
-20
.. Carb FuelMtr Fuellnj Catalyst
Tune
O2 Load
Sensor Sensor
All
132
-------
APPENDIX Z:
TOTAL ESTIMATED EMISSION REDUCTIONS FOR ALL
SUBSYSTEMS
This table summarizes estimates of the total emission
reductions, in g/mi and percent of overall CTP fleet
reduction, realized by repairs to specific subsystems, as
measured by the LA4. Average reductions per repair are
calculated from isolatable repairs only. Number of repairs
includes all repairs to that subsystem, whether or not
isolatable. Totals greater than 100% are due to the
SUBSYSTEM
REPAIRED
INDUCTION SYS1
Htd Air Door
Temp Sensors
Air Filter
Hoses
Other (Indt)
N
'EM
2
2
15
6
3
FUEL METERING SYS1
CarbAssmbly
Fuel Meter Tune
Idl Spd Sole
Fuel Inj
Hoses
Other
ChkAdjVacm
Vac Olaphrms
Other (Chk)
27
43
1
20
7
11
1
5
2
IGNITION SYSTEM
Dist Assembly
Igni Tune Items
Vac Adv Assmb
Spk Delay Dev
Elect Tim Mod
Hoses
Wir/Hms/Fuse
Other
EGR SYSTEM
Valv Assembly
Delay Solnoid
Cool Temp Sen
Hoses
Other
11
54
5
1
1
2
2
2
12
1
1
8
5
AVG REDUCTION
PER REPAIR
HC CO
0.06 -1.6
-
0.01 2.2
1.13 -2.3
0.00 -0.2
EM
1.03 11.9
0.61 11.6
-
2.35 24.2
-0.09 0.2
2.35 29.7
0.01 0.5
-0.06 -1.0
-0.01 -0.4
0.00 -0.5
0.34 0.4
-0.19 -1.2
0.28 10.4
- -
0.01 -0.5
- -
- -
-0.01 0.6
-0.02 0.3
0.05 -1.4
0.62 1 1 .3
AIR INJECTION SYSTEM
Pump Assembly
Byps/Dump Vlv
Diverter Vlv
5
3
13
7.93 41.1
-
-1.45 -3.4
ESTIMATE OF
TOTAL REDUCTION
HC CO
0.12 -3.1
-
-0.17 33.2
6.79 -13.8
0.01 -0.5
27.73 320.5
26.29 498.8
-
47.09 484.3
-0.64 1.3
25.82 327.2
0.01 0.5
-0.30 -5.0
-0.02 -0.9
-0.02 -5.2
18.59 19.8
-0.95 -6.0
0.28 10.4
0.02 -1.1
-0.11 6.7
- -
-0.02 0.3
0.37 -10.9
3.09 56.4
39.66 205.6
-
-18.87 -43.7
% OF ENTIRE
CTP REDUCTION
HC CO
0.0% -0.1%
- -
-0.1% 0.7%
2.1% -0.3%
0.0% 0.0%
8.6% 6.3%
8.2% 9.9%
-
14.6% 9.6%
-0.2% 0.0%
8.0% 6.5%
0.0% 0.0%
-0.1% -0.1%
0.0% 0.0%
0.0% -0.1%
5.8% 0.4%
-0.3% -0.1%
0.1% 0.2%
0.0% 0.0%
- -
0.0% 0.1%
-
0.0% 0.0%
0.1% -0.2%
1.0% 1.1%
12.3% 4.1%
- -
-5.9% -0.9%
133
-------
SUBSYSTEM
REPAIRED
N
AVG REDUCTION
PER REPAIR
HC CO
AIR INJECTION SYSTEM (continued)
Check Valve
Drive Belt
Hoses
Wir/Hms/Fuse
Other
PCV SYSTEM
Valv Assembly
Filters
Hoses/Lines
Other
31
5
27
2
7
3
4
3
1
EXHAUST SYSTEM
Exh Manifold
Catalyst
Other
10
56
2
0.35 3.8
-0.04 0.1
-0.24 0.5
0.02 0.4
-0.30 -3.7
-0.12 -0.8
- -
-9.68 -35.3
-
1.11 7.0
0.50 12.3
EVAPORATIVE SYSTEM
Evap Canister
Canister Purg
Hoses
Other
1
2
3
2
ENGINE ASSEMBLY
Eng Assembly
Cooling Sys
Valve Adj
Belt Tension
Hoses
Eng Oil
Other
3
7
4
1
1
2
3
0.08 0.3
0.58 5.8
0.02 0.2
-
0.63 16.6
-0.06 -1.5
0.11 1.5
- -
5.99 28.7
-0.16 -0.8
0.46 1.9
THREE-WAY CATALYST SYSTEM
ECU
02 Sen
Load Sensor
Eng Spd Sen
Cool Temp Sen
EGR Postn Sen
A/F Cntrt Act
Air Bypas Sen
Air Dtvrt Act
ISC Sys
Hoses
Other
MAT Sensor
Wir/Hrns/Fuse
ALL
19
82
22
2
8
1
5
3
4
1
18
2
3
9
630
0.40 10.7
0.80 20.7
0.61 23.2
0.53 3.4
0.86 27.4
0.06 -12.8
2.46 32.6
2.76 28.2
0.12 -0.4
0.20 2.0
0.37 -3.7
-0.02 -0.7
-0.55 10.2
0.43 26.7
0.70 10.7
ESTIMATE OF
TOTAL REDUCTION
HC CO
10.77 117.6
-0.18 0.7
-6.41 12.3
0.04 0.8
-2.08 -25.9
-0.35 -2.4
-29.05 -105.8
_
62.18 393.0
1.01 24.6
0.08 0.3
1.16 11.6
0.06 0.5
1.89 49.7
-0.41 -10.8
0.46 6.1
5.99 28.7
-0.32 -1.6
1.39 5.7
7.68 203.5
65.61 1698.0
13.35 510.6
1.07 6.8
6.91 218.8
0.06 -12.8
12.32 163.1
8.28 84.5
0.49 -1.6
0.20 2.0
6.73 -66.5
-0.04 -1.4
-1.65 30.5
3.89 240.3
345.89 5455.8
% OF ENTIRE
CTP REDUCTION
HC CO
3.3% 2.3%
-0.1% 0.0%
-2.0% 0.2%
0.0% 0.0%
-0.6% -0.5%
-0.1% 0.0%
_ _
-9.0% -2.1%
_ _
19.3% 7.8%
0.3% 0.5%
0.0% 0.0%
0.4% 0.2%
0.0% 0.0%
_ _
0.6% 1 .0%
-0.1% -0.2%
0.1% 0.1%
1 .9% 0.6%
-0.1% 0.0%
0.4% 0.1%
2.4% 4.0%
20.4% 33.5%
4.1% 10.1%
0.3% 0.1%
2.1% 4.3%
0.0% -0.3%
3.8% 3.2%
2.6% 1 .7%
0.2% 0.0%
0.1% 0.0%
2.1% -1.3%
0.0% 0.0%
-0.5% 0.6%
1 .2% 4.7%
107.5% 107.8%
134
-------
1 For a more complete description of the CTP program objectives, refer
to "Program Plan: A Cooperative EPA/Manufacturer I/M Testing
Program," U.S. EPA, Office of Mobile Sources, ECTD/TSS, January 1987.
2 At the time the CTP was being designed, the Michigan AET database did
not include fields for vehicle type (LOT or LDV), fuel metering type,
and other vehicle identifiers that were factors in the CTP
recruitment. Consequently, Michigan AET program statistics available
to EPA were not specific enough to set the program quotas.
Nevertheless, some summary information from those data are provided
in Appendix A. For more details on the recruitment quotas, refer to
the CTP program plan op. cit..
3 The safety and outlier rejection criteria were baaed upon the EPA
Emissions Factors recruitment criteria, and included off-road use,
major engine modifications, and excessive towing. Other such
criteria were evaluated at. the time of vehicle intake at the test
site.
4 Some manufacturers chose to begin their as-received testing with a
simulation of the original AET test. Such a test was not part of the
program plan, however, and data for these tests are not stored in the
common CTP database.
5 CTP Program Plan, op. cit., pp. 21-25.
6 This fact is the basis for including a restart requirement in the
testing of Ford vehicles with unloaded versions of the EPA-approved
performance warranty short tests (40 CFR 85.2201-2212)
7 The criteria in the I/M Variability category actually involve a set
of numerical comparisons between the emission results in the various
modes and sequences of the Basic I/M Test Procedure. Details may be
found in the CTP Program Plan op. cit.
8 A generalized flow diagram of decision making in the CTP remedial
maintenance phase appears in Figure 3 of the program plan, op. cit.
9 Vehicle 256, a Ford Topaz, suffered a transmission failure during as-
received testing and was removed from the program by the
manufacturer. Vehicle 344, an Buick Electra, had a substantial leak
in the catalyst as-received, and was not FTP tested for safety
reasons; post-repair FTP testing was not performed due to the lack of
an emissions baseline for the vehicle.
10 initial testing for three of the four (vehicles 215, 255, and 303)was
terminated for safety reasons due to catalyst overtemperature, traced
to malperformance of other components; the fourth had an ECM failure
traced to a disconnected coolant temperature sensor.
11 Unless otherwise specified, "AMC" encompasses AMC, Jeep, and Renault
nameplates and "VW" encompasses VW of America and VW of Germany.
Divisions of other manufacturers are grouped under the principal
manufacturer name (e.g., Chrysler, Dodge, and Plymouth are grouped
under Chrysler).
135
-------
12 Unless otherwise noted, the term "CO-only failure" refers to a
vehicle that fails the relevant procedure (FTP, short cycle) for CO
but passes HC; i.e., NOX is ignored in the failure-type
classification. Failures for "HC-only" are handled analogously. If
no suffix appears on the failure type (e.g., "CO failure"), the
classification was made blind to the pass/fail status of other
pollutants. To simplify the tables and figures in this report, the
"blind" category is rarely presented on its own, but the ordering of
the other failure types has been chosen to permit easy summing of the
HC-only or CO-only category with the HC+CO category, yielding the
failures for one pollutant that are blind to the other.
13See, for example, Glover, E. L., and Brzezinski, D. J., MQBTLE4
Exhauat Emission Factors and Inspection/Maintenance Benefits for
Paaaenyer Cars. US E.P.A technical report EPA-AA-TSS-I/M-89-3 (1989).
14The upper bounds for marginal emitters in MOBZLE4 were determined by
projecting a log-normal distribution onto the emissions of a large
sample of in-use vehicles in each of several technology categories,
applying a two-standard-deviation cutoff, and then back-calculating
the emission values that corresponded to the cutoff.
15Limited sample sizes prevented application of the four emitter
categories to light-duty trucks in the development of the MOBILE4
model. We have applied the categories to both LDVs and LOTs in the
CTP analysis, however.
16 Recall from Section 2.4 that Ford vehicles alone received a
keyoff/restart step between idles in the core sampling periods
throughout all sequences of the BITP except the Restart sequence.
Because this was the baseline condition for Fords, the effect of the
procedure design is still to isolate the impact of the intervening
engine operation, which is the significant point for the analysis to
follow.
17 Note by recalling Table2 in Section 2.4 that because the restart
step followa the RS-02 mode, XL-05 and RS-02 are procedurally
identical.
18 Almost all of the vehicles that were lost to this analysis were ones
that received no Indolene Extended Loaded sequence following the as-
received FTP. Such exclusions included all of the Chryslers and
scattered cases from the other participants. Low incoming fuel
levels in two vehicles (33 and 240) prompted substitution of
commercial fuel during the Extended Loaded sequences of the tank
BITP.
19 The anomolous RVP of 5.0 for vehicle 613 was verified with the
manufacturer (Nissan) but remains unexplained.
20 "Excess" emissions are defined as that portion of the emissions above
the certification level for the vehicle, with HC and CO treated
separately. "Total excess" is the sum of the individual excess
emissions of each vehicle. At this point and for the remainder of
this report, the excess is set to zero for clean vehicles those
whose emissions are below the certification standard. Earlier
analyses in this report treated clean vehicles as negative excess.
136
-------
21
Glover, E. L., and Brzezinslci, D. J., op. cit.
137
------- |