EPA-AA-EPSD-I/M-92-01
Technical Report
Report on the EPA/Manufacturer Cooperative
I/M Testing Program
By
James A. McCargar
Lisa Mouat Snapp
September 1992
NOTICE
Technical Reports do not necessarily represent final EPA decisions or positions.
They are intended to present technical analysis of issues using data that re
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 that may form the basis for a final EPA decision, position or
regulatory action.
Emission Planning and Strategies Division
Office of Mobile Sources
Office of Air and Radiation
U.S. Environmental Protection Agency
-------
Section 1 : Executive Summary 1
1.1 Program Summary 1 .
1.2 Results 1
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 10
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 14
2.6 Selection of Vehicles for Remedial
Maintenance 14
2.7 Remedial Maintenance Protocols and Post-Repair
Testing 14
2 . 8 Database Structures 16
2 . 9 Program Nonconformities 16
Section 3: Basic Vehicle Characteristics of the
CTP Fleet 18
3.1. Introduction and Overview 18
3.2. Profile by Manufacturer and Quota Group 18
3.3. Mileage Profile 20
3.4. Profile by Vehicle Type 23
3.5. Additional Comments on the Base Sample 24
Section 4: As-received Emissions Analysis 25
4.1. Introduction 25
4.2. FTP Results 25
4.2.1. Sample Description 25
4.2.2. Pass/Fail Results at Certification
Standards 25
4.2.3. Excess Emissions Analysis 26
4.2.4. Analysis Using MOBILE4 Emitter
Categories 30
4.2.5. Correlation of Excess Emissions,
Emitter Category, and Odometer 33
4.3.- Michigan AET Test Results 35
4.3.1. Profile of the Base Sample 35
4.3.2. Stratifications by Manufacturer and
Quota Group 36
4.3.3. Stratification by Mileage and Vehicle
Type 38
4.4. Correlation Between the FTP and Michigan AET
Results 38
4.4.1. Introduction 38
4.4.2. Excess Emissions and Emitter
Categories of the AET Failure Types ...38
-11-
-------
4.5. Laboratory Short Test Results 42
4.5.1. Sample Description 42
4.5.2. Second-Chance Failure Rates of the As-
Received Base Sample 42
4.5.3. Idle-Mode Short Test Variability Due
to Fuels 46
4.5.4. Correlating the Laboratory Short Tests
with the MOBILE4 Emitter Categories ...47
4.5.5. Variability Between Adjacent Idle
Modes 50
4.6 Supplemental Analysis of the AET Errors of
Commission 54
Section 5: Emission Effects of Remedial.
Maintenance 56
5 .1 Introduction and Sample Descriptions 56
5.2 Total Mass Emission Reductions from the CTP
Fleet 56
5.2.1 Net Benefit of Repairs FTP-Based 56
5.2.2 FTP versus LA4 values 59
5.2.3 Net Benefit of Repairs LA4-Based 63
5 . 3 Overview of the Repairs Conducted 64
5.3.1 System and Subsystem Repair Categories ...64
5.3.2 Emission Benefits per System Repair 66
5.3.3 Emission Benefits per Subsystem Repair ...70
5.3.4 Total Benefit per Subsystem 74
5.3.5 Effect of Deteriorated Catalysts on
Emissions 76
5 .4 Analysis of Incremental Repairs 78
5.4.1 Sample Description 78
5.4.2 Benefits of Repairing to Pass I/M 79
5.4.3 Comparison to MOBILE4 Repair Estimates ...83
5.4.4 Benefits of Repairing to pass the FTP ....85
5.4.5 Emission Benefits "Lost" through
Second-Chance 87
5.4.6 Benefits of Repairing to Different
Targets 88
5 . 5 Repairs to High Emitters 89
5.5.1 Effectiveness of Repair on Marginals
vs . Highs 89
5.5.2 Benefit of Repairing Highs only 91
-5.5.3 Catalyst Repairs Performed on Highs 93
5.6" Difficulty of Repair 95
5.6.1 Difficulty of Repair to Passing Levels ...95
5.6.2 Effectiveness of Repair at Successive
RM Steps 99
Appendix A: Failure Rates in the Michigan AET
Program 105
Appendix B: Vehicle Identifying Information for
the CTP Base Sample 106
-111-
-------
Appendix C: As Received Failure Rates for
Selected Modes of the Basic I/M Test
Procedure 114
Appendix D: AET Errors of Commission in the CTP
S amp 1 e 119
Appendix E: Per-Repair Emission Reductions for
All Systems: by Quota Group 120
Appendix F: Per-Repair Emission Reductions for
Statistically Significant Systems:
by Quota Group 126
Appendix G: Per-Repair Emission Reductions for
All Subsystems 127
Appendix H: Per-Repair Emission Reductions for
Statistically Significant
Subsystems: by Quota Group 129
Appendix I : Total Estimated Emission Reductions
for All Subsystems 134
Appendix J: Cooperative Test Program Plan 136
-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 70% 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 4.3]
(2) The mean excess HC and CO emissions of the MY1981-82
vehicle group exceeded the mean excess emissions for
Section 1: Executive Summary
-1-
-------
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 cutpoints 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]
Section 1: Executive Summary
-2-
-------
(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
Section 1: Executive Summary
-3-
-------
reductions were negligible, with high emitters
achieving reductions 15 times as large as marginal
emitters. [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 had
they been repaired. [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 three-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 the mean of
all other repair types for HC and 3/4 that of the
mean 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 emissions model than those seen in the CTP,
particularly for HC on fuel injected vehicles. This
is at least partially due, however, to the second-
chance screening received by CTP vehicles, which
eliminated cleaner vehicles from the repair
cycle. [Section 5.4.3]
(21) Ninety-four percent of vehicles that were repaired
from high to normal emitter levels on a transient
Section 1: Executive Summary
-4-
-------
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 I: Executive Summary
-5-
-------
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.
Section 2: Background and Program Summary
-6-
-------
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
Section 2: Background and Program Summary
-7-
-------
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+, 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.
Section 2: Background and Program Summary
-8-
-------
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 Testing 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
As-Received 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
Section 2:
Background and Program Summary
-9-
-------
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
Section 2: Background and Program Summary
-10-
-------
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
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
fluorescence and targeted at designating the fuel as either
Section 2: Background and Program Summary
-11-
-------
above or below a 0.05 g/gal standard. Reid Vapor Pressure
testing was conducted using the ASTM D323 protocol.
Section 2: Background and Program Summary
-12-
-------
TABLE 2
Modes of the Basic I/M Test Procedure
SEQUENCE
Cold Start (CS)
Extended Loaded (XL)
Extended Idle (XI)
Restart (RS)
MODE*
0 1
0 2
0 3
04
05
0 6
07
oa
09
1 0
1 1
1 2
1 3
1 4
1 5
0 1
02
03
04
0 5
0 1
02
03
04
05
0 6
07
08
09
1 0
0 1
02
oa
04
05
MODE NAME
75° Soak
Engine Start
Idle-neutral
2500rpm
idle-neutral
2500rpm
Idle-neutral
Keyoff/Restart
2500rpm
Idle-neutral
Idle-neutral
Idle-neutral
Keyoff/Restart
2500rpm ,
Idle-neutral
LA 4
Idle-neutral
Keyoff/Restart
2500rpm
Idle-neutral
Idle-neutral
Idle-neutral
Keyoff/Restart
2500rpm
Idle-neutral
2500rpm
Idle-neutral
Keyoff/Restart
2500rpm
Idle-neutral
LA4
Idle-neutral
Keyoff/Restart
2500rpm
Idle-neutral
DURATION
>60 min
n/a
30 sec
30 sec
i 120 SBC
1 80 sec
30 sec
i r\t&
30 sec
120 sec
10 min
30 sec
n/a
30 sec
120 sec
1372 sec
30 BBC
n/a v
30 sec
120 sec
20 min
30 sec
n/a
30 sec
120 sec
180 sec
30, sec
n/a
30 sec
120 '*ee
1372 sec
30 see
n/a ,
30 see
120 sec
FUNCTION
Base Operation
Base Operation
Core Sampling
Core Sampling
, Core, Sampling ^ ..
Conditioning
, Core Sampling
Ford Vehicles Only
Core Sampling
Core Sampling
Conditioning
Core Sampling
Ford Vehicles Only
Core Sampling
Core Sampling
Base Operation
^Core Sampling ,
Ford Vehicles Only
Core Sampling * ,
Core Sampiina
Base Operation
Core Sampling
Ford Vehicles Only
Core Sampling
Core Sampling
Conditioning
Care Sampling
Ford Vehicles Only
Core Sampling
Core Samplinq
Base Operation
Core Sampling
Non-Ford, Vehicles Only
Core Sampling -
Core Sampling
Section 2:
Background and Program Summary
-13-
-------
2.5 As-Received 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
Section 2: Background and Program Summary
-14-
-------
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 P_hase
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 comparable values; OR
Observed variability traced to
unrepairable element of design OR
Variability cannot be repeatably
trigaered 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;
Section 2:
Background and Program Summary
-15-
-------
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
Section 2: Background and Program Summary
-16-
-------
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.
Section 2: Background and Program Summary
-17-
-------
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
1 5
15
14
12
1 2
1 0
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
Section 3:
Basic Vehicle Characteristics of the CTP Fleet
-18-
-------
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.
FIGURE 1
Fleet Profile bv Quota Group and Manufacturer
GM FCRD NISS TOYT CHRY AMC VW MAZ SUBA MITS HC>D
Fl 83-86
Garb 83-86 D Garb 81-82 Fl 81-82
Section 3:
Basic Vehicle Characteristics of the CTP Fleet
-19-
-------
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 Group
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.
Section 3:
Basic Vehicle Characteristics of the CTP Fleet
-20-
-------
FIGURE 2
Mileage Profile bv Quota Group
1 234 5 6 7 8 9 10 11 >12
Lower Bound of 10K Mileage Interval
Fl 83-86
Garb 83-86 D Garb 81-82 Fl 81-82
FIGURE 3
Cumulative Mileage Profile bv Quota Group
<20 <40 <60 <80 <100 <120
Mileage Category
(1000*3)
H Total
Fl 81-82
0-Carb 81-82
-Garb 83-86
0-FI 83-86
Section 3: Basic Vehicle Characteristics of the CTP Fleet
-21-
-------
FIGURE 4
Mileaae Distribution bv Manufacturer
GM FORD NISS TOYT CHRY AMC VW MAZ SUBA MITS HCND ALL
<100K
<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 Mileacre Accumulation bv Manufacturer
GM FORD NISS TOYT CHRV AMC VW MAZD SUBA MITS HOM) ALL
I Total Q Annual
The fleet mean odometer was 55,000 miles, and the
calculated mean annual mileage accumulation was just under
Section 3: Basic Vehicle Characteristics of the CTP Fleet
-22-
-------
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
LDV
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
Section 3:
Basic Vehicle Characteristics of the CTP Fleet
-23-
-------
3.5. Additional Comments on the Basp 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.
Section 3: Basic Vehicle Characteristics of the CTP Fleet
-24-
-------
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
Section 4: As-Received Emissions Analysis
-25-
-------
FIGURE 6
As-Received FTP Failure Type bv Quota Groun
100 -r
90
80
70
60 -
50
40 -
30 -
20 --
10
0
Fl 81-82 GARB 81-82 CARB 83-86 Fl 83-86
ALL
NO FAIL
HC-ONLY m HC&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
applicable 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.
Section 4: As-Received Emissions Analysis
-26-
-------
FIGURE 7
As-Received FTP Excess Emissions:
Fuel-In-iected 1981-82 Vehicles
00
(g/mi)
140
120
100
80
60
40
20
0
-20
I
....
I 1
L _ . . .
r.
I I
L . - . J
. . . .
I 1
L - .". _
- ...
.....
i
- . . _ j
.....
1 1
....
....
1
L ... J
i
i
i
__..!.._.
i
i
I
i
i
(18.7, 55
1
1
....!....
i
i
i
i
i
i
i '
i
1 1
.6)
I
I
1
- 1
3456
HC (g/mi)
10
FIGURE 8
As~Received FTP Excess Emissions;
Carbureted 1981-82 Vehicles
00
(g/mi)
140.
120
1 n n .
l U U
80
6ft .
U
4n .
u
20 -
0
-OH-
.... J
...V,
1
l 1
l 1
....1....C...J
1 1
1 1
1 1
1 1
1 1
' >
1 _l 1
1 I
i ...M".
' S ~ "' '
iV»* T"-T
i i
i i
1 1 1 1
«...
8
....
i i
L... J
J
i i
....
....
1 1
....
....
....
1 1
.....
.....
1 1
....
....
....
i i
.....
r....
.....
3456
HC (g/mi)
10
Section 4: As-Received Emissions Analysis
-27-
-------
FIGURE 9
s Received FTP Excess Emissions:
140 -,
120 -
100 -
80 -
60 -
(g/mi)
40 -
20 -
0 -
o n .
- c. U 1
Carbureted 1983-86 Vehicles
m;
1
1
i . ...
7
i
i
i
..--!...-
i
. . . . i - - !.
!
.....
.... J
I
....
1
....
.....
....
101234567891
HC (g/mi)
FIGURE 10
As-Received FTP Excess Emissions:
Fuel-Iniected 1983-86 Vehicles
00
(g/mi)
140-
1 "T w
120 -
1 A n .
\ U U
80
6rt -
w
40
90.
£ U "
0 -
.. .. J
.....
. - . f,'l
'
1
1
....!....
I
I
I
I
.... ' ....
1 "
1* " ^~ I "
|
1
rS5«.i
r " I
i
L-.. J
-----
m
.... (
.....
....
i
j
....
.....
...m-
.....
.....
"V
.....
....
....
....
....
.....
. . . . .
"
.....
.....
.-...
....
.....
.....
.....
-1
3456
HC (g/mi)
10
Section 4: As-Received Emissions Analysis
-28-
-------
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 4.2.4
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 As-Received Fleet
% of Total
Excess
120
100
80
60
40
20
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
HC excess emissions together account for 90% of the fleet
Section 4:
As-Received Emissions Analysis
-29-
-------
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
Upper 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
00
Pass Marginal High
cert std 17.411 150
cert std 10.499 150
cert std 10.398 1 50
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.
Carbureted vehicles were more likely to be high emitters than
Section 4: As-Received Emissions Analysis
-30-
-------
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-Received FTP Profile by Quota Group
and Emissions Category
Fl 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
several times greater than the standards applicable to those
Section 4:
As-Received Emissions Analysis
-31-
-------
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/mi)
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).
Section 4:
As-Received Emissions Analysis
-32-
-------
FIGURE 13
As-Received FTP Profile by Manufacturer
and Emissions Category
GM FORD NISS TOYT CHRY AMC VW MAZ SUBA MITS HOND 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.
Section 4: As-Received Emissions Analysis
-33-
-------
HC Excess
(g/mi)
FIGURE 14
Mileage vs. Excess HC in the Carbureted
1981-82 Quota Group
8.
6.
5.
A ,
T .
P .
0.
..........
I i
I 1
1 1
i 1
L.......-.1-........1.........J
i i
i i
* ""* *
1 * 1
......... .......................
' '
_.*! 1
! *' *' '
"',' '
. -. !. 'I.". «l . !!
i i
i i
..........
...
50
100
150
200
250
Mileage (1000's of Miles)
FIGURE 15
Mean Mileage of Model Year and Fuel-Metering Groups
bv MQBILE4 Emitter Category
Mean
Mileage
(K)
Fl
Garb
81-82
83-86
ALL
PASS D MARG M HIGH
Section 4:
As-Received Emissions Analysis
-34-
-------
4.3. Michigan AET Test Results
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 30
second 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 Test)
for the Base Samtile
GO
(%)
1 n R .
Q fi .
8.4
7 9 .
6 -
4 8 .
(
3.6 -
2 4 .
1.2 -
0 -
I
i
' B
I B """"
- -"vjv ---%-
.....'?.."."."*".....
"J» "!
""j"!5i«' J"' !
..ii'^Xi,.* ; . _"-'. .
"* ?* "_**
ifii-Af.iii^uH
------
r - - - - n
' ,
B
i" -
- - - - -
- - - - -
B
B
- - " "
1
r "H
r
£
...
IE 1
.......
.....
------
1 *H
r-----
-----
------
----<
i 1
_ ____
m
j
i
220 440 660 880 1100 1320 1540 1760 1980 2200
HC (ppm)
Section 4: As-Received Emissions Analysis
-35-
-------
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
m HC-Only
DHC&CO
CO-Only
Fl 81-82 CARB 81- GARB 83- Fl 83-86
82 86
ALL
Section 4: As-Received Emissions Analysis
-36-
-------
These observations on the Michigan AET 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
60 T
GM FCPD NISS TOYT CHRY AMC VW MAZ SUBA MITS HOND
HC-Only CUHC&CO CO-Only
Section 4:
As-Received Emissions Analysis
-37-
-------
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 Result.^
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
Section 4: As-Received Emissions Analysis
-38-
-------
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 of FTP Emitter Types bv Type of AET Failure
m PASS
D MARG
H HIGH
SUPE
HC-only
BOTH
AET Failure Type
CO-On!y
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.
Section 4:
As-Received Emissions Analysis
-39-
-------
TABLE 9
Relationship Between AET Failure Type* and. Excess Emissions
for the High Emitter Category
AET Failure
Type
HC-only
HCarxlCO
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 cutpoints 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.
Section 4:
As-Received Emissions Analysis
-40-
-------
AET
FIGURE 20
of the Normal Emi
Lvina
Near the 207(b) Standards
00
3c
.b
2 A .
.4
10
.£. ~
n -
1
.-----------.
I
' *
I
V . ' .7
r;---'11-.
" ".
..............
220
440
HC (ppm)
660
880
FIGURE 21
AET Scores of the High Emitters Lying
Near the 207(b) Standards
00
3.6
2.4 -
1.2
.-
^
220
660
HC (ppm)
880
Section 4: As-Received Emissions Analysis
-41-
-------
4.5. Laboratory Short Test Results
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 or. their tank fuel. Six of the
vehicles were found to have insufficient tank fuel for the
BITP 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.
Section 4: As-Received Emissions Analysis
-42-
-------
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
& 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).
Section 4:
As-Received Emissions Analysis
-43-
-------
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
Section 4: As-Received Emissions Analysis
-44-
-------
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 o'f 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
Section 4: As-Received Emissions Analysis
-45-
-------
the AET CO-only failures passed the second-chance test as
well.
TABLE 10
Comparison of Failure Types Between the AET Test and the
First Idle-Neutral of the Extended Loaded Sequence
Second-Chance
Failure Type
HC-Only
Both
CO-Only
Pass
Total
AET Failure Type
HC-Only
1 5
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
Section 4:
As-Received Emissions Analysis
-46-
-------
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 400ppm 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.
Section 4: As-Received Emissions Analysis
-47-
-------
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
P
a
s
s
Pass
Marginal
Pass+Marginal
XL-05
XMO RS-05
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.
Section 4:
As-Received Emissions Analysis
-48-
-------
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 High Emitters to
Second-Chance Short Testina
F
a
CO-only m HC+CO
HC-only
Base
Operation:
Conditioning:
Warm
Soak
None
Warm Warm Soak Extended Extended
Soak 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
Section 4:
As-Received Emissions Analysis
-49-
-------
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 the high 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 2.4.2 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
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%
Section 4:
As-Received Emissions Analysis
-50-
-------
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
Section 4: As-Received Emissions Analysis
-51-
-------
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
220ppm, 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
Between First Idle and Second Idle
Following Extended Loaded Operation
1 1 UU -
Son .
ou
2nd fifin .
bl IU O O \J
Idle
H3
/ V% M MM \ ^. A 0 .
(ppm) H*»U
220 -
i
..........
A ^
..........
A
./
^^
* "?*!
,^-^fi-
L.........J
L... . . .- . J
B
. '
L _ j
1 "I
i IB %"
IS "
n
a
n
,-' " * J
./......
«r
a "°
1
...... ...j
......... j
'
.........J
*mm ".j
"
.....>...j
n
0 ,TP£--
0
+
+
220 440 660
1st Idle HC (ppm)
880
1
1100
Section 4:
As-Received Emissions Analysis
-52-
-------
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,
while half as many (eight) did the reverse. As the table
shows, however, most of this difference came from five 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 Operation
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
14
21
All
141
1 8
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
Section 4:
As-Received Emissions Analysis
-53-
-------
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 22 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 LOT 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
commrssion 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
Section 4:' As-Received Emissions Analysis
-54-
-------
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 E0 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. .."; *> . ... At *&.-.:...
"
Section 4: As-Received Emissions Analysis
-55-
-------
SECTION 5: EMISSION EFFECTS OF REMEDIAL MAINTENANCE
5 . 1 introduction and Sample Descriptions
Prior to any repair, each vehicle underwent an as-
received characterization to aid in determining 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
Section 5: Emission Effects of Remedial Maintenance
-56-
-------
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 over 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
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 is
number of
vehicles
46
47
17
8
15
'-.13
-'" 10
7
8
6
7
184
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%
Section 5:
Emission Effects of Remedial Maintenance
-57-
-------
TABLE 16
FTP Emission Reductions (a/mi)
for All Reoairs
bv Quota Group
HC
FI81-82
Carb81-82
Carb 83-86
Fl 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
Carb 8 1-82
Carb 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
239-vehicle sample
%
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%
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\
Section 5:
Emission Effects of Remedial Maintenance
-58-
-------
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 I 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 derived-LA4 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
Section 5: Emission Effects of Remedial Maintenance
-59-
-------
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»0.96(FTPHC)-0.16 R2-92.8% (19.1,18.6)
8
10
12
FTP HC (g/mi)
i AX
LA4
CO
FIGURE 27
FTP vs FTP-Derived LA4 Emission Values CO
LA4CO - 0.99(FTPCO) 1 .95 R2 - 98.9%
20 40 60 80 100
FTP CO (g/mi)
120 140
160
Section 5: Emission Effects of Remedial Maintenance
-60-
-------
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
FTP 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 FTPDerived 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%
Section 5: Emission Effects of Remedial Maintenance
-61-
-------
FIGURE 28
FTP vs FTP-Derived LA4 Emission Reductions HC
LA4
HC
reduction
(g/mi)
10
8
6
4
2
0
-2
(-8.6,-9.7)
ALA4HC o 0.99(AFTPHC) - 0.05 R2-93.1%
n
a n
a a
246
FTP HC reduction (g/mi)
10
FIGURE 29
FTP vs FTP-Derived LA4 Emission Reductions CO
150
100
LA4
CO
reduction
(g/mi)
50
-50
ALA4CO - 0.99(AFTPCO) - 0.23 R2 - 99.2%
-50
0 50 100
FTP CO reduction (g/mi)
150
Section 5: Emission Effects of Remedial Maintenance
-62-
-------
FIGURE 30
Average Emission Reductions Due to Repair: FTP and
FTP-Derived LA4
FTPHC -O LA4HC
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/'M CO, for an average reduction per vehicle of 1.5
g/mi H'6*" 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.
Section 5:
Emission Effects of Remedial Maintenance
-63-
-------
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
Section 5: Emission Effects of Remedial Maintenance
-64-
-------
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
Section 5: Emission Effects of Remedial Maintenance
-65-
-------
hoses, lines, wires
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 eaqh. 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 due to repairs to each
system. The regression for HC reduction took the form
10
ALA4HC = 0 + Z(ALA4HC)i x (indicator for systemi 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
that occurred singly, and the results of the multiple
regression for all repairs. All emission values are in grams
Section 5: Emission Effects of Remedial Maintenance
-66-
-------
per mile, as measured by the LA4.
breakdowns by quota group.
See Appendix E for
TABLE 18
Emission Reductions per System Repair
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- rat lo
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-ratlo
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 that 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
Section 5:
Emission Effects of Remedial Maintenance
-67-
-------
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
Avenaoe HC Reductions per Isolatable System Repair
HC
g/mi
1.5 -r
-1.5 -
-3 -L
INDT FUEL IGNT EGR AIR
EXH EVAP ENG 3WAY ALL
FIGURE 32
Average CO Reductions per Isolatable System Reoai
20 T
CO
g/mi
10--
-10-L
INDT FUEL IGNT EGR AIR
EXH EVAP ENG 3WAY ALL
Section 5: Emission Effects of Remedial Maintenance
-68-
-------
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 illustrated in
Figure 33. This figure and the values cited below include
only isolatable repairs; missing data in the figure indicates
small sample size rather than 0% effectiveness and, thus, the
values in the figure do not always appear to match those
cited. As clearly shown in Figure 33, 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 convertor replacements. The
excellent convertor 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
Garb 81-82 H Garb 83-86 H Fl 81 -82 D Fl 83-86
EXH
FUEL
SWAY
AIR
IGNT
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,
Section 5:
Emission Effects of Remedial Maintenance
-69-
-------
effective at eliminating I/M failures 52% of the time, varied
among various valve replacements and repairs to the pump
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.
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.
.. t'
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
Section 5: Emission Effects of Remedial Maintenance
-70-
-------
subsystems, and Appendix H for breakdowns
statistically significant subsystems by quota group.
of the
TABLE 20
Emission Reductions per Subsystem Repair
SUBSYSTEM
REPAIRED
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
02 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 that 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
Section 5:
Emission Effects of Remedial Maintenance
-71-
-------
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
Average HC Reduction per Isolatable Subsystem Repair
Cart
Fuel Mtr
Tune
Fuel Catalyst
Inject
ECU O2 Load
Sensor Sensor
ALL
FIGURE 35
Average CO Reduction per Isolatable Subsystem
Carb Fuel Mtr Fuel Catalyst ECU O2 Load
Tune Inject Sensor Sensor
ALL
. Section 5: Emission Effects of Remedial Maintenance
-72-
-------
Many of the same seven subsystems are consistently
effective at reducing the I/M failure rate. Figure 36
illustrates the effectiveness of repairs to certain
subsystems at allowing an I/M failing vehicle to pass the I/M
test.
FIGURE 36
I/M Pass Rates Due to Subsystem Repair
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.
Section 5:
Emission Effects of Remedial Maintenance
-73-
-------
FIGURE 37
T/M Pass Rates Due to Subsystem Repair bv Quota Group
Cart 81-82 H Garb 83-86 El Fl 81-82
Fl 83-86
Catalyst Carburetor
Oxygen Fuel Meter Fuel Injector Ignition Tune
Sensor 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
i.qolatable 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%
Section 5:
Emission' Effects of Remedial Maintenance
-74-
-------
of the repaired vehicles. It was even more effective than
the oxygen sensor at reducing HC per vehicle, at 1.11 g/mi,
but only about 1/3 as effective at reducing CO.
FIGURE 38
Contribution of Subsystems to Total HC Repair Benefit
bv Quota Group
35% j
30% -
25% - -
% 20% -
HC 15%--
10% -
5% -
Carb 81-82 H Carb 83-86 H Ft 81 -82 D Fl 83-86
O2 Catalyst Fuel
Sensor Injector
Carb 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% j
30%
25%
20%-
15%-
10%-
5% -
0% - -
Carb 81-82 H Carb 83-86 E3 Fl 81-82 D Fl 83-86
02
Sensor
Fuel
Injector
Fuel Meter
Tune
Load
Sensor
Catalyst Carb
Fuel
System
Other
Section 5:
Emission Effects of Remedial Maintenance
-75-
-------
Fuel system repairs of many kinds, including fuel
injector replacements, carburetor 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
mixture control 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
Section 5: Emission Effects of Remedial Maintenance
-76-
-------
catalyst replacements and for all other RM types. Catalyst
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-catalyst
replacement 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 percent
reduction reduction
per RM per RM
1.23
0.62
0.67
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
note: 55+413=468; 10 of the missing repairs
due to potential masking effect of new
catalyst diagnostic rather than repair
average percent
reduction reduction
per RM per RM
8.5
11.1
10.6
78.1%
48.1%
49.5%
total
reduction
468.5
4575.3
5063.0
were post-catalyst replacement, excluded
catalyst on subsequent repair reductions; 1
% 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 that 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
Section 5:
Emission Effects of Remedial Maintenance
-77-
-------
on vehicles with below-average lead-in-fuel levels. Also, of
the ten vehicles with the highest lead-in-fuel levels (0.015
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 that 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 that 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
Section 5: Emission Effects of Remedial Maintenance
-78-
-------
period that is closest to the "ideal" I/M test condition.
Analysis of the effects of repair on the I/M test 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 that
were initially failing I/M. The net LA4-measured emissions
benefit achieved from the RMs 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
Section 5: Emission Effects of Remedial Maintenance
-79-
-------
after the repair that caused the passing test. The reduction
is that caused by the single RM that caused the vehicle to go
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 T/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.
Section 5:
Emission Effects of Remedial Maintenance
-80-
-------
FIGURE 40
Changes in Emitter Group Due to Repair from I/M Fail to I/M
Pass by Quota Group
%0f
vehicles
High-High E~3 High-Norm D Norm-High Norm-Norm
FI81-82
Garb 81 -82 Garb 83-86 Fl 83-86
ALL
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.
Section 5:
Emission Effects of Remedial Maintenance
-81-
-------
FIGURE 41
Changes in Emitter Group Due to Repair from I/M Fail to I/M
Pass bv Manufacturer
%of
vehicles
High-High H High-Norm D Norm-High Norm-Norm
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.
Section 5:
Emission Effects of Remedial Maintenance
-82-
-------
FIGURE 42
Average LA4 Emissions for Vehicles wifch Additional Repair
After Passing I/M
HC
g/mi
3.5 -r
3 -
2.5 -
2 -
1.5 -
1 -
0.5 -
0 --
0-HC
-co
CO
g/mi
-j-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
that initially failed an idle test to CTP values for a
similar|[set of vehicles. (The MY 80 vehicles in the MOBILE4
datasets||re California only, with technology similar to that
used oni£F.'ederally 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 that took the
vehicle from I/M failing to I/M passing status. CTP
reduction values are LA4-based; MOBILE4 values are FTP-based.
Section 5:
Emission Effects of Remedial Maintenance
-83-
-------
TABLE 23
MQBILE4 vs CTP Average Emission Reductions
Due to Repair 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/mi %
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%
1 1 .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/mi %
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
As-Rcvd
FTP
0.41
2.36
6.41
2.18
Reduction
g/ml %
0.08 20.0%
1 .42 60.3%
4.48 69.9%
1 .33 60.9%
N
12
24
4
40
As-Rcvd
FTP
5.8
47.9
184.1
49.0
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 two to 40%. 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. These lower levels are probably due to differences
in the vehicle sample receiving repair. The MOBILE4 data is
composed of vehicles that failed an I/M test in the field,
while the CTP data includes only those field I/M failures
that went through the additional screening of as-received
testing and were shown to require repairs. Many clean
vehicles were eliminated by this battery of second-chance
Section 5:
Emission Effects of Remedial Maintenance
-84-
-------
tests, thus raising the average pre-repair emission levels
and subsequent emission reductions of the group that
remained. 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.
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 CC. 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
that 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.
. Section 5: Emission Effects of Remedial Maintenance
-85-
-------
TABLE 24
T/M Benefits of Repairing to FTP Normal Emitter Levels
Ideal I/M
Pass/Fail
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 Quota
%0f
Vehicles
Fail-Fai
Fail-Pass D Pass-Fail Pass-Pass
FI81-82
Cart) 81-82 Carb 83-86
Fl 83-86
ALL
Section 5:
Emission Effects of Remedial Maintenance
-86-
-------
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
Changes in I/M Pass/Fail Status Due to Repair from High to
Emitter bv Manufacturer
%0f
Vehicles
GM
Fail-Fail
Fail-Pass D Pass-Fail Pass-Pass
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
Section 5:
Emission1 Effects of Remedial Maintenance
-87-
-------
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 this 60% of the CTP fleet would have
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
Section 5: Emission Effects of Remedial Maintenance
-fl8-
-------
I/M-targetted repairs, and whether a substantial portion of
"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 Renair to Different
TOTAL REDUCTION
(g/mi)
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
» 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 Hiah 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
Section 5:
Emission Effects of Remedia'l Maintenance
-89-
-------
the marginal emitters is negligible, whereas the high
emitters have substantial LA4 reductions 1.8 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
Averaae HC Repair Benefit Marginal vs High Emitters
> marginal -D- high
HC
g/mi
2.5
2
1.5
1
0.5
0
as received
after final repair
FIGURE 46
Averaae CO Repair Benefit Marginal vs Hiah Emitters
CO
g/mi
marginal -o- high
as received after final repair
Section 5:
Emission Effects of Remedial Maintenance
-90-
-------
The following table simply tallies the number of
vehicles that 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 Reoairs
EMITTER
CATEGORY
High
Marginal
NON-CATALYST REPAIRS
N cleaner | N dirtier
123 11
26 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-
catalyst RMs, and eventually had 83% and 92% of its excess HC
Section 5: Emission Effects of Remedial Maintenance
-91-
-------
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 oer RM of Non-Catalvst Repair to High 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
eventually had 78% and 90% of its excess HC and CO
Section 5:
Emission Effects of Remedial Maintenance
-92-
-------
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.
Section 5: Emission Effects of Remedial Maintenance
-93-
-------
FIGURE 48
Average HC Benefit! of Repair to High Emitters
Catalyst vs Other Repairs
HC
g/mi
3
2.5
2
1.5
1
0.5
0
cat repair vehicles
non-cat repair vehicles
as-received after final non- after cat
cat repair repair
FIGURE 49
Average CO Benefit of Repair to High Emitters
SO-,
40-
co 3°-
B/ml". 20-
10-
0-
Catalyst vs Other Repairs
- cat repair vehicles -O- non-cat repair vehicles
\
: \
V^.
i i i
as-received after final non- after cat
cat repair repair
Section 5: Emission Effects of Remedial Maintenance
-94-
-------
TABLE 27
Average Benefit of Repair to High Emitters
Catalyst vs Other Repairs
HC
cat repair vehicles
non-cat repair vehs
CO
cat repair vehicles
non-cat repair vehs
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
Section 5: Emission Effects of Remedial Maintenance
-95-
-------
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 ner RM Step
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 Changes Due to Remedial Maintenance
Number
of
Vehicles
High-High H High-Norm D Norm-High Norm-Norm
345678
Remedial Maintenance Step
10
Section 5:
Emission Effects of Remedial Maintenance
-96-
-------
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
%0f
Vehicles
100 j
90 -
80 --
70 -
60 -
50 -
40
30
20
10
0 --
EflSthRM
Ei 5th RM
Q 4th RM
H3rdRM
H 2nd RM
D 1st RM
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,
Section 5:
Emission Effects of Remedial Maintenance
-97-
-------
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 81-62 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 Group
%0f
Vehicles
100 -r
90
80
70
60
50
40
30
20
10
0
IHethRM
ii 5th RM
fi 4th RM
H3rdRM
II 2nd RM
n 1st RM
FI81-82 Cart) 81 -82 Garb 83-86 FI 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.
Section 5:
Emission Effects of Remedial Maintenance
-98-
-------
FIGURE 53
RM Step that Cleaned Up Emissions
LA4 from High to Normal Emitter bv Manufacturer
%0f
Vehicles
100
90
80
70
60
50
40
30
20
10
0
ffllethRM
i3 5th RM
II 4th RM
H 3rd RM
El 2nd RM
D 1stRM
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.
Section 5:
Emission Effects of Remedial Maintenance
-99-
-------
FIGURE 54
TotaL and Average HC Reductions per RM Step
AvgHC
AVERAGE Reduction
(g/mi)
0.9
10
Remedial Maintenance Step
FIGURE 55
Total and Average CO Reductions per RM Step
Total CO
Reduction
(g/mi)
2500
2000
1500
1000
TOTAL
AvgCO
AVERAGE Reduction
(g/mi)
20
345678
Remedial Maintenance Step
10
In general, the early repairs were quite successful at
reducing emissions, with the first repair generating an
Section 5: Emission Effects of Remedial Maintenance
-100-
-------
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 table
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 that 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>20 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%
Section 5:
Emission Effects of Remedial Maintenance
-101-
-------
FIGURE 56
RMs Needed to Reduce LA4 Emissions by >8Q%
for Vehicles with FTP HC>2.Q and/or CQ>20 g/mi
bv Quota
50 j
40--
Number 30 - -
of
Vehicles 20''
10--
0--
Never
FI81-82
Cart 81-82
Garb 83-86
Fl 83-86
FIGURE 57
RMs Needed to Reduce LA4 Emissions by >8Q%
for Vehicles with FTP HC>2.0 and/or CQ>2Q a/mi
by Manufacturer
Never Si >3 Ha Hz D 1
Number
of
Vehicles
FORD NISS TOYT CHRY AMC VW MAZ SUBA MITS HOND
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,
Section 5:
Emission Effects of Remedial Maintenance
-102-
-------
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.
Section 5: Emission Effects of Remedial Maintenance
-103-
-------
FOOTNOTES
1 For a more complete description of the CTP program objectives, refer
to Appendix J for "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 based 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 thes 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
Footnotes
-------
manufacturer name (e.g., Chrysler, Dodge, and Plymouth are grouped
under Chrysler) .
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-t-CO category, yielding the
failures for one pollutant that are blind to the other.
13See, for example, Glover, E. L., and Brzezinski, D. J., MOBILE4
Exhaust Emission Factors and Inspection/Maintenance Benefits for
Passenger Cars. US E.P.A technical report EPA-AA-TSS-I/M-89-3 (1989).
upper bounds for marginal emitters in MOBILE4 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 LDTs 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 Table 2 in Section 2.4 that because the restart
step follows 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 anomalous RVP of 5.0 for vehicle 613 was verified with the
manufacturer (Nissan) but remains unexplained.
Footnotes
-------
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.
21 Glover, E. L., and Brzezinski, D. J., op. cat.
Footnotes
-------
APPENDICES
-104-
-------
APPENDIX A: FAILURE RATES IN THE MICHIGAN AET 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
HOND
MITS*
NISS
TOYT
VW
1 981
17.8
21.2
25.0
15.9
3.8
50.0
48.3
10.4
25.5
1 982
15.3
20.6
25.1
13.1
6.0
66.7
28.6
5.7
12.5
1 983
16.4
14.4
14.0
8.3
11.3
15.4
15.2
8.5
10.8
1 984
9.9
13.7
9.9
9.0
10.3
10.0
16.9
5.1
4.0
1 985
10.8
5.3
6.6
5.4
11.6
15.4
4.4
4.2
1.5
1 986
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
Appendix A
-105-
-------
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
GIF
GTI
JEEP
181
TERC
ALLI
RABB
SUBA
GL£
626
GL£
VAN
GLC
CCFO
TERC
STAR
GL
CCfD
CORO
SPIR
F/M
CARB
CARB
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
$3
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
LOT
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
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
AETCO
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
FTPCO
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
HCstd
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
COstd
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
96
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
Appendix B
-106-
-------
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
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
Model
RABB
EAGL
SPIR
RABB
RABB
Gl£
SPIR
GL10
323
RX7
RABB
RABB
181
QUAN
GLC
RABB
ALL!
RABB
181
BJOO
VANO
RABB
GL
GL
DL
626
GL10
GL10
RX7
323
626
FIFT
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
Mf
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
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
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
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
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
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
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
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
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
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
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
11 1
23
55
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
Appendix B
-107-
-------
Veh
102
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
Mfr
CHRY
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
Model
ARIE
CHAR
OMNI
CARA
REU
DAYT
LEBA
600
REU
LANC
LEBA
NEVW
ARIE
REU
LING
TCAMM
LINC
CAPR
MUST
F150
RANG
TEMP
MARQ
TEMP
TOPA
RANG
TOPA
TEMP
TEMP
MARQ
LTD
MUST
F/M
CARB
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
MV
81
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
CID
135
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
Type
LDV
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
KMile
79
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
AETHC
228
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
AETCO
3.5
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
FTPHC
1.42
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
FTPCO
33.1
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
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
0.80
0.80
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
COstd
7.0
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
XL05HC
20
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
XL05CO
0.0
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
Appendix B
-108-
-------
Veh
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
Mfr
FORD
FORD
FORD
RORD
PORD
FORD
PORD
FORD
PORD
FORD
FORD
FORD
FORD
FORD
FORD
RORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
RORD
Model
ZEPH
RANG
LTD
MARQ
MARQ
MUST
MAFO
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
F/M
CARB
CARB
CARB
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
MY
81
84
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
CID
140
140
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
Type
LDV
LOT
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
KMile
117
44
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
AETHC
183
124
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
AETCO
3.2
1.4
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
FTPHC
1.66
0.54
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
FTPCO
28.3
4.1
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
HCstd
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
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
1.70
1.70
0.41
COstd
3.4
10.0
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
XL05HC
155
40
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
XL05CO
0.7
0.0
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
Appendix B
-109-
-------
Veh
253
254
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
Mfr
FORD
FORD
FORD
FORD
FORD
FORD
GVI
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
TB/IP
MUST
MUST
MUST
CAPR
RANG
RIVI
RGGA
MALI
BOW
CELE
MALI
RIVI
CIER
PHOE
MALI
SKYH
FIER
SEVI
CORV
CUTL
FIER
FIRE
CENT
CELE
DEVI
CAVA
RIVI
REGA
CIER
CITA
BOM
F/M
TBI
CARB
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
MY
86
81
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
R?
CID
140
140
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
?31
Type
LDV
LDV
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
KMile
10
65
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
AETHC
651
102
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
AETCO
8.8
2.8
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
FTPHC
3.60
0.38
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
FTPCO
65.2
8.2
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
HCstd
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
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
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
XL05HC
826
84
450
61
49
80
11
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
XL05CO
8.7
0.0
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
Appendix B
-1 10-
-------
Veh
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
Mfr
a/i
a/i
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
CM
CM
Model
2000
GRAN
CIER
MONT
CMB3
CELE
REGA
CIMA
GRAN
CITA
O/B3
REGA
SKYL
CITA
DEVI
CENT
RIVI
J200
CITA
GRAN
SUNB
SUMB
DEVI
CELE
CITA
CAVA
SKYH
CAVA
GRAN
CAVA
DEVI
CIER
F/M
TBI
CARB
TBI
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
Mf
83
84
84
83
8^ '
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
CID
110
305
151
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
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
KMile
59
37
60
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
AETHC
367
252
325
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
AETCO
0.2
0.4
0.3
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
FTPHC
0.97
2.01
0.24
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
0.45
0.34
0.64
7.64
2.96
0.43
0.35
6.09
1.44
0.29
FTPCO
7.4
15.1
2.9
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
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
COstd
3.4
3.4
3.4
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
XL05HC
173
196
41
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
XL05CO
0.3
0.5
0.1
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
0.1
1.5
0.0
0.2
Appendix B
-111-
-------
Veh
401
402
403
404
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
Mfr
HQND
HOND
HCND
HGND
HCND
HOND
HOND
HGND
HGND
HGND
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
Model
CIVI
CIVI
CIVI
CIVI
CIVI
CIVI
CIVI
CIVI
ACCO
CIVI
COLT
RAM
COLT
COLT
RAM
COLT
COLT
COLT
COLT
COLT
STAN
200S
PULS
SENT
MAX)
MAXI
200S
200S
280Z
300Z
300Z
2802
F/M
PFI
PFI
CARB
CARB
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
MY
86
85
84
84
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
CID
91
91
91
91
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
Type
LDV
LDV
LDV
LDV
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
KMile
41
58
53
71
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
AETHC
259
327
227
383
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
AETCO
1.4
0.7
0.0
0.0
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
FTPHC
0.35
0.93
1.46
1.19
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
FTPCO
3.5
8.6
4.5
6.5
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
HCstd
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
0.80
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
COstd
3.4
3.4
3.4
3.4
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
XL05HC
44
371
160
172
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
XL05CO
0.1
4.3
0.1
0.1
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
Appendix B
-1 12-
-------
Veh
613
614
615
616
617
618
619
620
701
702
703
704
705
706
707
708
709
Mfr
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
Model
TRUC
PULS
280Z
SENT
280Z
PULS
MAXI
200S
TERC
TERC
OOFD
CEU
CEU
CORD
OOHO
CAMR
CEU
F/M
TBI
PFI
PFI
GARB
PFI
PFI
PFI
PFI
CARB
CARB
CARB
PFI
PFI
CARB
CARB
PFI
CARB
MY
86
83
82
84
83
-4
83
83
84
86
84
81
86
83
82
82
84
83
CID
146
91
168
98
168
91
146
120
91
91
108
122
144
108
108
122
144
Type
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
28
55
102
48
74
59
61
28
5
77
96
32
59
234
147
32
49
AETHC
61
1527
152
239
138
446
281
293
262
339
167
335
293
445
1473
344
2000
AETCO
2.3
9.5
1.7
0.2
1.8
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
0.61
5.09
1.69
0.16
0.86
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
98.9
43.7
54.2
3.6
11.7
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
0.80
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
COstd
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
3.4
3.4
3.4
XL05HC
60
561
181
50
226
547
17
250
7
45
128
0
141
109
1
0
0
XL05CO
3.8
5.8
2.3
0.1
4.3
5.6
0.0
6.0
0.0
0.0
0.8
0.0
0.1
1.1
0.0
0.0
0.0
Appendix B
-1 13-
-------
Appendix C
As-Received Failure Rates for Selected Modes of the Basic I/M Test Procedu
VEHICLE COUNT
BITP
Mode
CS03
CS05
CS07
CS10
CS12
CS15
XL02
XL05
XI02
XI05
XI07
XI10
RS02
RS05
Failure )*tvbe
HC-only
49
39
19
20
33
35
19
18
39
41
22
20
16
21
HC+CO
131
68
68
61
74
65
64
63
78
59
63
58
64
64
CO-only
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
CS10
CS 12
CS 15
XL 02
XL 05
XI 02
XI 05
XI 07
XI 10
RS02
RS05
Failure Type
HC-only HC+CO CO-only
21% 56% 12%
17% 29% 13%
8% 29% 11%
8% 26% 11%
14% 31% 10%
15% 27% 8%
8% 27% 9%
8% 26% 7%
16% 33% 10%
17% 25% 9%
9% 27% 11%
8% 24% 10%
7% 28% 8%
9% 28% 8%
Pass
11%
42%
52%
55%
45%
49%
56%
59%
41%
49%
53%
57%
58%
55%
Appendix C
-1 14-
-------
Appendix C
Valo«fl for Calculating JU-R*o*iv*d railur* Rat*a for S*l*c*t*d Mod** of th* Basic I/M T*«t Proc*dur«
HC 7
oai
ASA
V9m
040
041
04>
04a
044
041
041
047
041
041
010
Olt
012
osa
OS4
OSI
Oil
_01I_
0*93
HO 00
247 «.4
8000 1.4
711 7.4
212 1.2
IIM 1.7
323 1.1
271 1.1
I2» i.e
121 1.4
280 4.2
491 1 3
171 0.!
ai to
2M 1.3
iaa t
2000 i.a
111 7.1
ooa to 3
toa i
147 1 t
462 1.<
iao 10
242 2.1
211 I.a
toai s
202 a.9
111 1.1
277 1
»1I 2.«
441 ».7
343 i.e
1902 10
700 2.7
132 10
an i.s
414 4
1954 2.2
711 1.4
IO9 A II
1 VZ W.fl
471 2.7
Ml 0.2
414 4.a
164 10
2000 1.6
117 10
2000 4.7
171 1.2
444 1.7
1121 1.9
aa2 7.6
701 I.I
2000 I.I
271 0.4
117 a.i
I4a 1.1
411 2.6
«*« 12
OJOI
HO 00
111 1.7
I4i i.a
161 6.6
611 10
721 S.2
142 2
112 4.'
141 o.a
141 35
211 1.9
taa o 4
It C
an 4.1
13! 1.1
211 *.<
121 10
177 17
112 0 1
410 1.7
214 0.3
II O.I
214 I.I
207 0.3
111 2.1
110 1.1
301 2.3
431 2.1
307 4.4
112 0.4
133 0.7
691 9.3
633 0.4
104 0.6
264 2.3
370 3
69 0.1
210 0.9
111 A
1 t)
209 1.7
III 0.2
227 O.I
291 1
1712 0
202 3.1
2000 4.6
451 1.2
152 0.1
673 1.3
117 4.9
110 5.7
197 1.2
111 0.3
99 2.1
241 2.4
51 0.1
101 04
0667
HO CO
If 0
1261 10
351 L*
Ml 10
143 0.1
107 7.4
2* O.S
127 3.4
545 1
111 3.2
S42 1.6
66 0 1
20M 32
1269 10
1(1 0.1
«fi f
U.I
140 2.6
49 0
13 0
142 0.7
72 0
ISO O.t
261 3.1
216 0.2
370 3
139 4.4
116 0.6
16 0
293 6
131 0
III 3.2
1024 10
229 3.S
39 0.2
91 0
10 0
ISO 1.1
20M 0
3S1 0.1
274 9.7
62 0
140 4
20M 2.6
4ao 1.3
353 4.2
114 0.2
379 4.2
412 6.2
132 0.2
291 7.6
S9 O.I
13 0
" o
011*
MO 00
26 0
itoi ' to
' 471 (.1
164 10
II O.t
200 7.1
21 O.I
iao a.s
517 7.9
1 0.3
231 (.3
47 C
till 3.1
1014 10
54 0.1
«A 1
U. 1
131 2.6
74 O.t
16 0
lit 0.6
66 0
167 0.1
236 4
274 0.2
326 2.4
131 4
141 0.9
II 0.2
309 1
137 0
111 3.6
741 9.4
222 3.5
60 0.2
11 0
II 0
141 1.1
1139 0.2
210 O.I
274 9.9
14 0
147 4.2
2000 2.9
371 1.1
319 3
99 0.1
112 1.7
291 1.2
IS 0.2
273 7.4
60 0
14 0
19 0
C* 12
HO CO
63 0
1637 10
431 1.4
727 10
151 O.I
234 1.4
21 0.1
tat a.i
471 7.5
91 I.I
192 11
41 0
200S 3.7
2000 10
13 0
A 1
V. 1
tu t
431 6.3
309 5.4
112 0.9
67 0
211 0
301 4.7
323 0.2
296 2.2
135 3.3
112 t.l
340 3.6
261 5.1
70 0
142 3.9
1270 10
191 2.3
62 O.I
31 0
11 0
142 2.1
752 0.3
133 0.7
246 10
67 0
140 4.7
2000 2.7
623 0.3
305 2.7
4) 0
117 0.9
239 4.1
22 0
301 7.6
55 0.1
51 0
« 2
0115
MC CO
63 0
1414 10
429 6.2
727 10
113 0.1
246 6.6
24 0
122 2.6
470 7.6
251 6.4
222 1.4
2002 3.5
2000 10
12 0
t 0.1
112 1.4
290 0.3
92 3.7
193 1.1
10 0
213 0
301 4.4
363 02
303 2.3
127 3.6
160 1.2
290 1.1
247 6.3
145 0
132 3.9
106 10
191 2.6
204 0.3
102 O.I
144 2
644 0.2
172 0.7
241 10
41 0
IS4 4.6
2000 2.9
791 0.3
293 2.5
144 t.t
314 6.5
102 O.t
324 1
57 0
41 0
64 0.2
a. 02
MC CO
21 0
1203 10
463 I.S
700 10
6 0
366 10
1ft fl
ID V
319 0.6
277 9.7
402 7.6
177 2.6
239 6.9
1602 3.3
2000 10
1 0
99 1.6
1 C
226 0.7
1019 5.4
31 0.2
22 0
157 O.t
247 6.6
357 0.2
51 O.t
99 2.4
91 O.I
20 0
196 4.1
136 0
230 6.4
169 10
242 4
105 0.3
190 0.2
129 3.6
1079 0.2
112 O.I
209 9.9
60 0.3
136 6.5
1664 3.7
491 1
14 0
497 9.6
19 1.9
605 7.9
32 O.I
693 10
36 0
20 0.1
109 0.2
a.05
MC CO
33 0
1096 10
527 6.6
736 10
44 0.1
320 10
Of
V
561 0.6
272 9.7
369 7.7
1 0
167 3.2
292 6.7
1742 3.4
1763 10
1 0
17 2.t
0 0
190 0.4
1411 6.9
34 0.2
14 0
99 O.t
321 4.7
319 0.2
6t 0
96 2.4
130 1.2
21 0
191 4.1
139 O
231 1.6
591 9.5
266 3.6
53 0.3
14 J 0.2
ISO 3.6
1131 0.2
299 0.7
2tl 10
53 0.1
142 S
2000 3.3
520 0.9
17 0
511 9.1
91 1.6
540 6.4
21 0.1
746 10
37 0
19 0.2
3S 0.1
XI 02
HC CO
116 0.1
1616 10
472 7.6
647 10
203 1.2
632 10
)f
V
776 0.3
110 3.1
319 6.6
214 4.fi
141 1.5
152 1.2
Mf
V
2000 6.6
1100 10
351 6.9
16 O.S
1 C
353 65
619 6.1
127 O.I
264 0.7
216 0.2
329 3.7
335 0.2
43 0
105 3
165 2.3
775 1.1
203 4.2
63 0
174 7.2
525 9.3
211 3.7
162 0.2
t 0
156 3.6
1304 0.2
165 0.7
202 10
109 O.t
136 6
2000 2.7
756 0.6
26 0
369 7
14 0
96 0.9
233 51
361 3.6
390 6.6
47 0.1
272 4
Si o]
XI OS
HC CO
56 0
1343 10
613 6.9
701 10
169 0.2
710 10
11 0
326 0.6
104 2.7
349 6.2
164 3.4
169 1.9
114 O.I
* jjt ft n
HO Q.C
1621 3.6
I2t9 10
67 0
49 0.9
IS 0
316 S.7
197 5.5
176 1.1
330 1.2
217 0.1
260 4.5
407 0.2
52 0
104 31
137 2.1
477 0.4
211 4.4
tei o
167 7
169 2.4
267 4
164 0.3
325 0.2
151 3.6
1090 0.2
112 0.6
199 1.6
52 0
131 4.9
2000 3
526 1
31 0
327 4.7
16 0
114 1.2
334 7.1
273 1
367 I.I
44 0
157 0.1
11 0.1
XI 07
MC CO
23 0
1293 10
417 1
69$ 10
61 0
277 9.5
61 0
106 3.1
275 5.3
11 0
110 15
212 6
mfl A
Q.B
1619 3.7
1164 10
4 0
55 1.6
204 0.9
601 6.2
105 0.4
99 0.1
169 0.1
216 5
325 0.2
77 0.2
«9 3.4
140 1.2
27 0
211 5.2
201 0
140 6.2
594 9.2
250 3.3
11 0.2
192 0.1
163 3.6
1224 0.2
203 0.7
204 9.4
64 0
120 4.6
2000 3.6
276 1.4
19 A
J3 U
270 4.9
12 0
91 1
316 6.9
67 0.2
404 9.1
44 0.1
24 0
44 02
XI 10
HC CO
26 0
1564 10
509 6.3
711 10
71 0.1
435 10
SI 0.1
101 3.1
290 5.3
11 0
195 2.6
265 6
MA 9
O.«
1753 3.5
1274 10
3 0
67 2.1
193 0.6
894 6.2
73 0.3
S3 0.3
142 0.1
246 S.I
311 0.2
51 0
102 3.1
131 1.1
24 0
225 5.1
200 0
152 6.5
614 10.4
251 3.5
57 0.2
262 0.2
164 4
1136 0.2
166 0.7
211 9.5
S7 0.1
127 3.9
2000 2.9
362 1.4
215 5
12 0
17 0.9
303 6.9
77 0.2
396 II
42 0
21 0
21 02
RB 02
HC CO
51 0.1
1313 10
431 9.3
1015 10
14 0
S61 10
432 2
109 3.1
271 S.I
0 0
1 0
254 6.1
MA
V
1690 3.6
1794 10
6 0
89 1.5
1 0
tas o.i
663 6.6
27 O.t
6 0
133 O.t
210 33
443 0.2
38 0
103 3.2
13t 2.7
137 0.3
226 S.I
112 0
733 9.6
653 10.2
216 3.6
57 03
110 0.2
241 9.4
1331 0.2
III 0.7
111 10
41 0
135 6.1
1175 4.6
411 1.3
1141 10
13 0
N/A »NM
351 7.1
577 1.3
333 1.3
22 0
17 O.t
., '» «»
R30S
HC CO
100 0.2
1313 4
441 9.3
1051 10
13 0
421 10
722 0.4
99 2.9
267 S.6
2 0
126 0.6
273 6.2
MM
U
1744 3.5
2000 10
6 0
82 2
0 0
192 0.5
444 6.7
64 0.2
33 0
too o
322 52
393 0.2
37 0
100 32
153 2.4
15 0
220 5.4
208 0
660 10
643 10
246 3.5
3 0
165 0.2
246 9.1
1123 0.2
241 0.7
169 9.4
29 0
134 4.6
2000 4.6
359 t.4
1170 10
14 0
N/A »N/A
335 7.1
417 7.1
291 5.6
22 0
21 O.t
11 0.1
Appendix C
-1 15-
-------
V*
061
059
MO
101
lOt
101
101
107
I0»
110
HI
IIS
lit
114
IK
It*
117
201
sot
so*
804
SO*
so*
S07
so*
so*
SIO
til
SIS
SI*
SI4
SI*
t17
SI*
SI*
S20
SSI
SS2
sst
224
22*
SS«
S27
S2«
22*
S30
S3I
212
SI*
214
21*
21*
217
211
21*
S40
241
9A9
*4Z
241
244
24 S
24*
247
24*
24 »
Otll
HC OP
181 S.7
177 t.a
14S 7.1
US 4.S
60S 7.1
sso 1.1
60S 7.1
10* 0.7
SOI 0.2
270 «.2
60S 7
164 0.1
206 0.6
241 4.7
460 6
60S «.«
S«l 4.1
607 7.*
77 t.6
1*7 l.S
126 7.1
«6« 1.6
117 1.6
1(61 *.*
4* 1
20* 2
187 S.*
(67 0.6
S7I 2
117 *.6
101 1.7
01 1.1
4* 0.*
812 (.1
26* 1.2
20» 1.1
461 t.4
281 0.1
211 0.1
222 1.2
(76 2.7
26* 1.7
«7( (.4
III 2
(17 2
127 4.4
lit S.7
421 l.(
24* 0.2
666 6.6
IS 0
«t 1.1
1*S 1.6
64 S.2
1S6 1.*
(7 1.1
IftC 1 9
4aV I.Z
1(1 6.6
164 4.*
404 6.8
460 6.4
A? 9 I
W/ *. 1
79 0
110 1.4
0*0*
HC CO
46S S.1
140 0.1
1*1 6.7
72 1.1
(4 0
417 1
S2( (
121 0.1
114 1.6
211 1
2(( (.9
Ml 7.1
106 1.6
76 0.1
60S 7.1
121 l.S
1(1 1.2
102 1.1
till (.1
(0 *.(
!2( 0.1
S9* «
1*1 0
S67 0.7
1*7 0.1
0 0
166 0.6
181 S.S
6*S 0.6
M* 4.1
S11 0.1
(0 0
S70 (.2
IS 0.2
741 7.6
101 18
Sll 0.1
IS4( 7.S
SI* 0
17 0.2
164 0.6
191 0.6
S0( S.4
4* 0
60 0.1
SI 0
tot o
SI* 0.4
121 1
2(1 0.1
76* 7.9
20 0
* 0.8
167 0
116 1.7
11 0
100 0.7
60* 0.(
1(7 1.1
(1* ».4
19* 6.1
Ml D
l.V
222 0
52 0
CD 07
HC CO
SI* 0.7
271 0.4
46 0
26 0
IS 0
274 6.1
1(7 7.1
124 (.4
7 M
M **S
4*1 7.1
11 0.1
** 0.1
4* O.S
60S 7.1
41 1.8
2*2 6.4
959 t.7
129* *.*
6* 0.1
142 0
416 7.9
12 0
MO 0.1
1(0 0
4SO (.8
S* 0.1
417 4.1
It* 0
60 0
41 0
11*1 (.1
101* 1.1
77 S.I
41* 1.1
217 2.*
66 O.I
6* 0.4
6*4 *.(
* 0
122 1
SI* 2.7
SOI 1.6
60 0
14 0
411 7.1
1* 0
1* 0
110 6.4
144 1.9
671 6.7
107 1.1
1* 0
41 0
11 1.7
100 0
(1 0.7
991 4.8
328 (.2
44 0
716 7
Jrt
W
67 0
144 0
ca 10
HC CO
217 0.7
120 0.2
11 0
20 0
11 0
261 4.9
tOt 7.1
7* 1.*
"i .«,-- «»
!,,$« °
' '~*M 7.1
' t* 0.1
It 0.1
S07 0.4
60S 7.1
41 1.7
S*« 6.7
*( l.«
190 t.l
67 0.1
114 0
41* t.l
17 0
170 0.1
Mt 0
It 0
It 0
2*6 2.6
401 0
51 0
116 0.*
111* *.!
201 4.7
16 0
(12 1.6
241 2.1
41 0
111 1.*
29* 0
* 0
121 1
1*6 0
20* 1.*
7* 0
1* 0
1* 0
91 0
11 0
19* *
156 2.1
(14 (
10 0
17 0
1* 0
(1 1.6
(2 0
(2 0.7
9(9 6.2
114 (.2
150 0.9
577 6.7
75 0
71 0
CO 12
HC CO
247 0.8
1(9 0.1
83 0
10 0
11 0
2(0 4.4
60* 7.1
21 0.9
lit I
44 O.t
117 (.7
7 O.t
(t O.I
74 0.2
602 7.1
2* 1.1
S61 «.5
1*4* (.7
tie* 1.1
71 O.t
((( 7
440 7.9
ISO 0
997 4.(
79 O.t
201 S.I
81 0
(4* 4.1
750 0.1
2(9 1.5
4(2 4.*
1117 1.1
2(9 4.1
(1 2.4
52* 1.2
7* 0
1(1 l.(
97 0.2
566 1.9
14* 1
144 1.1
1(2 1.4
1*7 26
144 0
164 2.6
312 (.1
11* 0
(9 0
47* (
(14 5
(76 8.9
IK 1.3
20 0
719 (.(
51 0.1
8(7 6.(
(( 1.1
15*1 7.2
111 (
175 1.1
929 (.3
183 0
301 3 1
CS 11
HC CO
245 06
141 0.2
35 0
( 0
10 O.I
2(1 4.7
502 7.1
46 0.1
119 0.9
2* 0
43* 7.1
1* 0.2
3* 0
2*2 0.3
502 7.1
IS 2.4
17* (.*
1(5 t.4
872 (.1
74 O.I
15( 1.5
405 78
IK 0
(97 0.7
733 0
31 0
35 0
322 2.3
III 0.7
95 0
192 0.5
1342 1.3
221 4.2
72 0
110 2.*
Ill 2.3
110 l.t
155 I.I
114 0
41 0
113 O.I
70 0
225 1.3
1*4 0
131 1.3
40 0
9( 0
79 0
4(1 8.1
1005 (.2
(14 1
41 0
21 0
500 1.4
51 1
3(3 0.3
51 0.2
1081 4.7
340 S.t
91 0
392 6.5
187 0
113 0
XL 01
HC CO
171 0.6
270 0.4
(4 0
41 0.9
14 0
225 1.6
192 4.4
11 0.1
144 1.6
(2 0.1
1111 1.7
II 0.5
91 0.2
17 O.I
502 7.1
17 O.I
1141 1.2
lOtt 1.7
1041 1.1
71 0.1
101 0
514 1.4
17 0
147 0
144 0
11 0
64 0
111 2.1
414 0
51 0
71 0
1764 1.1
261 5.1
57 O.I
759 1.5
209 I.I
12 0
52 0
17 0
41 0
147 O.I
161 2.1
141 1.6
170 0
II 0
352 7.1
114 0
97 0
411 5.1
510 4
4(2 4.9
102 1.1
17 0
91 0
41 06
1094 7.9
11 0
1301 6.4
479 (
140 1.2
141 3.6
51 0
51 0
XL os
HC CO
201 01
111 0.2
23 0
55 0.6
20 0
211 4
40( 7.1
26 0.2
111 t
41 0
2024 7.1
14 0.4
20 0
115 0.2
502 7.1
44 0.6
145 (.(
165 2
910 1.1
76 O.t
10* 0
461 7.1
60 0
110 0
279 0
11 0
41 0
300 2.1
500 0
51 0
43 0
1206 7.7
211 4.9
26 0
557 3.4
166 0.7
40 0
72 0.1
71 0
40 0
Sll 0.3
13 0
144 1.2
171 0
27 O.I
21 0
17 0
21 0
452 1.2
409 0.1
921 I.I
29 0
14 0
147 O.t
50 1.1
974 7.1
14 0.2
9(5 5.1
431 6.5
51 0
164 O.I
31 0
12 0
147 0
XI 02
HC CO
221 O.S
151 0.2
51 0
21 0.5
21 0
260 !.(
21 0
28 0.7
121 06
195 0.1
112 (.1
IN/A »N/A
II 0.2
24 0.2
461 7.1
46 1.6
216 62
1141 (.7
1261 1.3
11 O.t
121 1.7
172 7
214 0
769 0.7
449 2.1
171 t.l
47 0
241 2.1
960 1.1
221 It
127 2.7
1117 1.2
2*7 6.1
11 1.2
III 2.1
57 0
121 1.4
216 I.I
271 2.1
104 O.I
221 0.9
132 1.3
213 2.1
212 O.I
II O.I
212 4.1
221 1
166 2
474 6.1
531 4.2
601 5.6
121 1
37 0.1
771 1.9
53 0.1
1061 7.9
70 0.7
1(51 (.7
(41 5.4
2(1 16
646 7.1
27 0
316 0.5
270 27
HI Of
HC CO
219 0.8
104 0.2
45 0
41 0.9
1 0
262 1.7
151 1.7
10 0.4
121 O.I
157 0.6
437 7.1
N/A »N/A
61 O.I
117 0.2
502 7.1
6* 2.1
475 1.9
229 2.5
1610 1.2
II O.I
311 1.4
171 I.I
151 0
131 0.7
2*1 0.1
31 0
57 0
141 1.5
794 0.7
II 0
97 0
121 4.4
203 4.3
27 0
767 2.6
111 0.6
91 O.I
141 0.7
101 0
75 0
133 0.7
41 0
139 1
115 0
25 0
40 0
107 1.9
151 0
411 52
600 1.7
1091 6.9
51 0
10 0
469 4
54 0.9
1004 7.1
69 0.4
832 4.1
621 5.1
91 0
306 0.6
K 0
71 O.t
101 0
1107
HC CO
113 0.7
12 0.2
31 0
31 2
1 0
104 4.1
161 1
179 3.1
101 1
10 0.1
331 7.1
N/A «N/A
11 0.1
SO 0.1
502 7.1
62 1.9
326 6.5
1111 1.7
1511 1.3
73 0.1
254 0.1
370 7.2
66 0
482 0
180 O.I
211 2.3
57 0
722 4.1
571 0
70 0
61 0
807 6.6
260 5.6
66 1.3
826 3.2
16 0.6
44 0.1
474 1
325 4.S
37 0
106 0.7
101 0.4
132 I.I
140 0
19 0
433 7.4
43 0
30 0
404 5.1
361 0
415 5.4
109 1.3
45 0
124 O.I
40 0.8
1274 7.9
42 - 0.3
50 1 7
1711 0.3
610 5.3
60 0
454 S
8 0
71 0
166 0
XI 10
HC CO
171 0.7
61 0.2
54 0
41 1.9
7 0
192 4
171 2.9
62 0.5
112 1
21 0
214 7
N/A «N/A
6 0.1
28 0
502 7.1
S3 1.7
351 6.7
135 2.3
1153 1.3
66 0.1
211 0
427 7.5
51 0
414 0
201 O.I
44 0
39 0
487 2.5
544 0
11 03
55 0
915 5.9
224 4.5
27 0
657 33
77 0.3
.41 0
94 O.I
103 0
87 0
123 0.6
41 0
132 1.1
152 0
11 0
54 0.1
36 0.6
30 0
415 5.7
329 0.1
962 7.1
49 0
24 0
239 1
44 0.6
651 7.9
46 0.3
3 0
1731 0.5
612 5.3
71 0
90 0
10 0
90 0
92 0
HBOI
193 0.6
531 0.6
21 0
76 3
39 O.t
216 3.6
96 0.5
24 0.9
130 I.I
42 0
296 5.9
22 0.4
11 0.2
162 O.t
502 7.1
S3 1.6
214 5.6
1115 1.7
1177 1.2
261 0.4
167 0
572 1.4
91 0
3B2 0.1
201 O.t
35 0
76 0
N/A »N/A
572 0.1
47 0
48 0.1
703 7.2
238 4.1
54 O.S
579 3.5
199 1.1
32 0
37 0
15 0
21 0
200 0.6
125 0.4
137 1.2
104 0
30 O.I
344 7.1
90 0
78 0
N/A »N/A
483 4
418 4.3
96 1.1
14 0
607 6.4
44 0.8
1114 7.8
22 0
5 0
N/A fN/A
1205 5.5
71 0.2
440 4
28 0
50 0
65 0
RB OS
HC CO
311 1.5
118 0.3
33 0
71 1.3
18 0
217 1.7
148 7
14 0.1
122 1
61 0
111 6.7
16 0.1
11 0.2
181 0.2
502 7.1
II 1.8
246 6
1061 8.7
1088 6.1
84 0.1
189 0
444 8
94 0
356 0
1S4 0.1
160 2.3
41 0
N/A »N/A
421 0
46 0
41 0
750 S.7
274 6.5
SO O.S
671 1.1
154 0.8
39 0
21 0
86 0
46 0
236 0.3
102 0
130 0.9
107 0
IS 0
365 7.2
64 0
25 0
N/A >N/A
387 0.7
303 3.8
90 1
25 0
458 6.7
49 0.9
1144 7.9
33 0.1
0 0
N/A »N/A
1192 4.9
59 0
267 3.2
20 0
49 0
73 0
Appendix C
1 16-
-------
V*
HO
251
252
251
254
257
251
as*
210
Ml
102
104
JOS
10*
107
101
109
ito
lit
*1t
in
»u
111
It*
117
lit
1)*
MO
Ml
ut
Ml
M4
MS
M*
127
Ml
12*
110
lit
112
111
114
111
111
117
111
119
140
»4i
141
141
Mt
141
147
141
14*
ISO
1SI
1M
Ml
154
151
1SI
357
»«
001
HC CO
1«0 0.1
175 1
727 7
101 1
2081 2.1
«0 0.4
Ml 7
215 0.1
105 0.9
477 1.1
1970 9.5
171 4.2
214 0.2
102 0.2
199 0.2
140 0.1
91 2.1
190 0.9
110 0.1
452 0.7
M2 1
till 2.9
201 I.I
211 11
194 1.5
94 1.7
201 1.9
145 0.2
107 O.I
III 0.4
114 O.I
112 0.1
1115 4.1
107 O.t
mi o.t
101 0.2
111 0.1
450 1.2
211 1.1
N/A *N/A
211 0.2
210 4
470 I.I
N/A IN/A
477 O.I
245 0.2
142 0.2
279 2.1
127 0.1
201 0.2
421 0.1
241 1.4
259 0.5
501 1.5
111 0.4
111 7.2
269 0.9
211 0.5
171 0.9
111 O.I
557 I.I
105 0.7
201 0.5
«" "«
C8 01
HC CO
11 0
201 2.1
514 1
21 0
159 4.1
41 0
119 4.2
III 0.1
21 0
472 2
104 0.1
1211 O.I
211 O.I
212 0.7
212 0.4
207 O.I
209 1.7
64 O.I
119 0.2
171 1.4
951 1
(10 2.1
79 0
271 O.I
144 O.I
401 1.9
125 0.2
104 0.1
151 0.4
121 0.5
154 0.4
219 O.I
499 0.9
112 0
1101 O.t
1551 0.1
111 0.5
414 7.1
II 0.1
N/A «N/A
212 O.I
210 4.1
127 0.1
N/A (N/A
1017 O.I
51 0
17 O.t
211 1.2
71 0.1
421 O.I
601 2.1
262 2.1
704 O.I
676 0.6
125 0.6
171 1.2
411 0.4
207 O.I
94 0.1
161 0.1
412 1.1
212 0.4
412 O.I
"" >
ca 07
HC CO
120 1.2
0 0
452 0.2
104 1.7
21 0
615 1.7
41 0
17 0
41 0
617 1
112 tJ
lit 0.1
127 0.4
21 O.I
11 0
71 0.2
151 0.2
154 1.7
200 0.1
194 0.1
72 0.2
920 1.1
112 2.7
41 0.2
40 0.1
221 1.2
III 0.4
11 0
11 0
20 0
311 2.4
1170 2.1
161 0.1
151 S.I
112 0.1
215 0.4
114 0.5
107 0.1
701 10.4
19 0
14 0.1
171 0.4
117 0.9
1112 7.6
N/A N/A
Appendix C
1 17-
-------
V«h
359
360
401
402
403
404
406
406
407
401
409
410
602
$03
S04
SOS
soe
ft. A 7
9V/
SOI
6o»
610
C * |
91 1
eoi
02
eo3
04
os
606
07
eot
00
10
11
12
13
14
IS
t«
17
It
619
20
7ft 1
701
702
703
704
70S
706
707
701
_Z09_
CSOJ
HC CO
1676 0.1
19 0.2
330 6
493 11
329 0.2
277 2.S
ItS 0.2
277 0.2
226 1.1
106 2.9
600 6
297 3.3
417 S.S
00 10.*
217 2.9
3S7 0.6
00 11.2
9AO K 0
JVV 9.1
369 1.7
412 S.4
1 010 0.1
*«n 4 y
«*« *l.f
670 4.2
739 4.1
662 2.*
429 0.6
426 4.4
316 2.9
103 0.3
324 1.4
217 3.4
337 3.2
412 3.3
2«9 O.t
203 2.5
605 7.9
330 3.4
254 t
254 2.1
04 10
321 3.«
95 0.3
17ft A 7
1/0 v.f
117 0.9
310 9.4
1 7A 9 9
1/0 C.<
646 0.2
352 3.9
302 6.1
240 0.9
lit 1.9
CSOS
HC CO
957 0.1
103 O.I
401 3
361 5
277 0.1
211 0.1
500 0.1
500 0.1
21 0.2
237 2
394 5
232 2.1
192 0.3
00 7
94 2.9
199 O.I
00 t
IKft 9 A
180 «.4
266 3.9
347 46
12* 0
97ft * fl
4ifO J.N
414 3.1
64* 2.2
202 0
211 0.2
192 26
lit 1.3
12 0.3
I7t 0.1
106 O.I
177 0.5
264 2.7
241 13
97 1.5
114 61
155 25
111 O.I
201 3.4
621 10
14 0.6
16 0.1
12$ 0.4
100 O.t
112 3.2
*n* ft A
1UJ 0.0
371 0.3
123 l.t
69 0.3
162 0.4
152 0.6
csor
HC CO
73 0
53 0
lit 1
70 O.I
144 O.I
110 O.I
600 O.I
600 O.I
216 0.1
13 t
323 3.1
73 1
444 1.4
tOO 7.7
62 2
53 0
00 9.1
A 1
U.I
216 1.1
212 3.1
tit 0.5
1 A fl
IB H
423 4.1
1013 1.2
27 0
111 0
199 2.7
144 2
52 0.2
551 0.5
10 O.I
10 0
9 0
386 t.S
113 4.5
406 S.I
111 2.1
114 0.2
253 4.2
634 7.5
25 0
217 t.l
16 0
56 0
152 O.I
16 0
110 0.1
140 1.1
27 0
10 0
CS 10
HC CO
67 0
39 0
141 1
65 0.1
322 O.I
203 O.I
SI 7 0.2
MO O.I
SIS O.I
IS O.I
334 3
** O.I
6SS S.S
MO 7.1
5t 8.6
41 0
00 1
«*
D
211 1.6
269 3.1
91 t.l
« A
1 * 0
463 4.1
1013 1.2
42 0
90 0
199 2.9
147 2.2
St 0.2
54 1 0.4
26 0.1
7 0
7 0
373 6.4
144 4.3
601 4.1
199 2.4
159 02
219 4.1
596 7.2
26 0
229 62
19 0
53 0
147 O.I
9 0
279 0.1
143 1.2
34 0
17 0
CS 12
HC CO
97 0
20 0
153 0.2
351 2.1
500 0.2
350 0.1
292 0.2
500 O.I
219 0.2
2(3 0.4
250 0.4
130 0.1
334 0.5
511 7
31 1.5
217 0.4
100 9.4
36 0
234 2.2
243 1.5
313 1.2
] A
/» V
469 2.1
Ml I.I
513 0
577 0.2
195 2.7
172 1.9
132 0.2
733 02
31 0.1
22 0
43 0
314 51
151 44
411 32
190 l.t
469 04
210 43
129 t.l
73 0
215 61
39 0
141 0.7
110 0.5
16 0
313 0
152 O.t
102 0
31 0 1
CS IS
HC CO
II 0
22 0
115 O.t
100 O.I
114 0.1
271 O.I
500 0.2
500 O.t
211 0.2
104 0.1
226 0.5
153 0.1
307 0.4
Sit 1.2
41 1.5
271 O.S
fOO 1.3
224 2.1
216 2.7
301 3
20 0.1
432 3
115 2.2
371 O.I
472 O.I
203 2.7
tit 2.1
113 0.1
575 O.t
41 O.I
399 1
155 3.1
155 l.t
201 i.i
214 0
212 1
601 62
101 0.1
277 S.7
111 0.1
117 O.I
352 0
145 I.I
II 0
43 O.I
XL02
HC CO
to o
92 0.2
79 0.1
It O.I
III O.I
149 O.I
III 0.2
500 O.I
211 0.2
41 O.I
341 3.1
49 0.1
146 0.2
100 1.2
31 2.4
210 0.3
100 1.9
45 0.1
0 0
1 0.1
»A
v
172 2.9
1011 1
966 22
41 0
III 3.4
110 1.9
25 0.1
312 0.4
19 O.I
314 S.t
61 3.7
663 15
114 2.1
45 0
232 4.4
561 1.4
10 0
243 S.t
35 0
112 05
141 O.I
122 0.9
0 0
0 0
XL 05
HC CO
63 0
33 0.2
44 O.I
371 4.1
160 0.1
172 0.1
123 0.2
500 O.I
341 0.2
41 0.1
291 I.I
ts o.i
250 0.7
593 7.1
35 1.9
114 O.I
600 1
21 0.1
2 O.I
42 0.2
On
u
271 S.I
1013 1
904 1.7
ts o
117 3.3
130 S.5
15 0.2
411 0.4
It 0.1
332 t
60 1.1
561 5.1
III 2.1
50 O.I
22t 4.3
547 5.1
17 0
250 t
45 0
121 O.I
141 O.I
109 I.I
1 0
0 0
XI 03
HC CO
274 0
15 0
62 O.I
307 O.t
413 02
441 0.1
140 0.2
500 0.1
254 0.3
239 0.4
276 1.3
221 0.4
520 1.2
too t.l
43 2.4
240 0.3
tOO 9.2
62 0.1
57 0
40 0
719 06
74 0
492 3.7
S9I 1.5
1015 1.1
565 O.I
ItS 23
151 24
141 0.7
511 0.2
43 O.I
249 O.t
21 0
340 5.5
113 t.S
331 2.5
115 26
541 1
215 4.9
541 52
101 0.2
257 t.l
120 0 1
142 0.5
27 0
50 01
214 O.I
127 0.6
10 0
9 0
XI OS
HC CO
lit 0
20 0
92 0.1
299 0.6
302 0.1
307 0
lit 0.2
500 O.I
254 02
219 0.9
214 1.2
201 O.I
371 1.2
589 7.5
I 1.7
211 0.4
too t.t
59 0.1
104 0.1
45 02
439 6.1
MA 1
V.I
641 3.2
609 1.1
1015 1.1
551 0.1
235 2
ISO 23
111 01
524 06
17 02
224 0.1
72 0
497 6
ItS 6.7
497 64
346 24
394 0.2
345 45
559 5.3
143 O.I
266 5.1
152 0 1
131 0.4
92 0.5
IS 02
197 0
121 O.t
33 0
13 0
XI 07
HC CO
65 0
51 0.3
51 O.I
17 O.I
240 0.1
201 0
113 0.2
500 O.I
251 0.2
101 0.1
274 1.4
121 O.I
413 1.1
634 7.1
11 1
211 0.5
tOO 9.2
25 0.1
17 0.2
41 0.2
211 0.4
402 3.2
713 4.7
1013 1.4
267 0
115 2.4
142 2.4
46 0.2
512 0.4
21 0
It 0
t 0
360 6.1
140 S.t
419 35
219 2.7
140 0.1
241 4.4
Sit 5.5
It 0
26 0
39 0
67 0
112 01
11 0
214 O.I
119 1.1
1 0
1 0
XI 10
HC CO
56 0
II 0
50 O.I
77 0
251 O.I
169 0
234 02
500 0.1
339 02
tt O.I
219 l.t
107 01
409 15
565 7.5
77 26
219 04
600 t.t
13 0.1
II 0.1
101 0.3
144 0.3
399 3.1
104 5.7
963 1.6
242 0
111 2.2
146 2.2
40 02
525 04
19 01
17 0
7 0
352 5.6
166 t.t
429 32
212 2.6
211 02
247 4.3
520 5.3
31 O.I
14 0
21 0
tt 0
112 0.7
6 0
184 01
121 1
3 0
0 0
2 0
RS 02
HC CO
94 0
19 0
77 0.1
79 O.t
174 O.I
141 O.I
151 02
500 0.1
462 0.3
49 0.1
191 4.5
19 O.I
205 0.1
512 71
30 1.6
67 0.3
600 64
0 0
30 0
13 0.2
77 01
371 1.1
1013 64
901 It
43 0
160 21
136 2.3
21 02
361 04
32 02
135 O.I
0 0
340 6
102 58
379 34
181 2.6
80 0.2
252 5.2
469 SO
6 0
t 0
4 0
36 0
76 01
2 0
160 O.I
98 0.6
0 0
0 0
0 0
HSOS
HC CO
96 0.1
19 0
47 O.t
320 2.9
196 0.1
tit O.I
240 0.2
500 0.1
443 02
51 0.1
319 4.1
It 0.1
304 0.8
319 6.7
21 l.t
114 0.4
554 7.1
Og
V
31 0
II 0.2
tt 0.2
0 0
312 3.1
1013 1.4
132 1.2
95 0
171 3.3
123 2.2
44 0.3
406 04
30 0.2
119 02
0 0
340 6
110 66
426 34
21 1 3.3
93 0.2
205 4.9
454 56
15 0
t 0
5 0
41 0
III 09
2 0
151 O.t
105 08
0 0
0 0
1 0
Appendix C
1 1 8-
-------
APPENDIX D
AET Errors of Commission in the CTP Sample
Veh
31
223
249
14
20
24
206
221
250
260
701
321
338
360
11
17
21
37
39
50
224
251
306
309
322
330
347
348
357
509
511
704
708
MY
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
FORD
FORD
FORD
FORD
TOYT
CM
CM
CM
VW
SUBA
TOYT
SUBA
MAZD
VW
FORD
FORD
CM
CM
CM
CM
CM
CM
CM
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
98
122
122
Type
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
KMile
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
Carb 81-82
Carb 81-82
Garb 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
Fl 81-82
Fl 81-82
Fl 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
HCCert
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
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
FTPHC
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
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
1 1
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 D
-1 1 9-
-------
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
-
Appendix E
-120-
-------
Fuel
SYSTEM
REPAIRED
Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
Injected 1981-1982 Vehicles
SIMPLE AVERAGES
ISOLATABLE REPAIRS
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
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- rat lo
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-ratio
0.0
1.7
0.0
0.0
1.6
0.5
1.4
4.1
Fuel Injected 1983-1986 Vehicles
SYSTEM
REPAIRED
Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N A HC A CO
0 -
22 2.07 21.6
26 0.37 -0.3
2 0.00 -0.1
10 -0.65 -0.3
0 -
15 0.98 7.2
1 0.08 0.3
5 1.57 7.1
75 0.22 14.2
156 0.56 10.7
MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N A HC
4 -1.77
29 1.78
29 0.72
2 0.00
10 -0.82
1 1.77
15 0.94
2 0.08
7 1.53
82 0.33
181
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-ratio
-0.8
3.2
0.3
0.0
-0.1
0.3
0.9
0.0
0.6
4.3
-
Appendix E
-121-
-------
Average HC Benefit per System Repair
Carbureted MY 81-82 Vehicles
Carb81-82
HC
g/mi
6.00 j
4.50 --
3.00 -
1.50 -
0.00
-1.50
-3.00 -
-4.50 -
INDT FUEL IGNT EGR AIR
EXH EVAP ENG 3WAY ALL
Average HC Benefit per System Repair
Carbureted MY 83-86 Vehicles
Garb 83-86
HC
g/mi
6.00 -r
4.50 -
3.00
1.50 -
0.00
-1.50-.
-3.00 --
-4.50 -
INDT FUEL IGNT EGR AIR PCV EXH EVAP ENG 3WAY ALL
Appendix E
-122-
-------
Average HC Benefit per System Repair
Fuel Injected MY 81-82 Vehicles
FI81-82
HC
g/mi
6.00 -r
4.50 -
3.00 -
1.50 -
0.00
-1.50 --
-3.00 -
-4.50 -
INDT FUEL IGNT EGR AIR PCV EXH EVAP ENG 3WAY ALL
Average HC Benefit per System Repair
Fuel In-iected MY 83-86 Vehicles
Fl 83-86
HC
g/mi
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 3WAY ALL
Appendix E
-123-
-------
Average CO Benefit per System Repair
Carbureted MY 81-82 Vehicles
Garb 81-82
CO
g/mi
30 T
20 -
10 -
-10 J-
INDT FUEL IGNT EGR AIR
EXH EVAP ENG 3WAY ALL
Average CO Benefit per System Repair
Carbureted MY 83-86 Vehicles
Garb 83:86
40 -r
30 -
20 -
CO
g/mi
. 10 -
0
-10 -
-20 -
INDT FUEL
EXH EVAP ENG SWAY ALL
Appendix E
-124-
-------
Average? CO Benefit per System Repair
Fuel Innected MY 81-82 Vehicles
FI81-82
40 -r
30 -
20--
CO
g/mi
0
-10 -I-
-20 -
1
INDT FUEL IGNT EGR AIR PCV EXH EVAP ENG SWAY ALL
Average CO Benefit per System Repair
Fuel Iniected MY 83-86 Vehicles
Fl 83-86
40 -r
30 -
20
CO "
g/mi
10--
0
-10 -
-20
INDT FUEL IGNT EGR AIR PCV EXH EVAP ENG SWAY ALL
Appendix E
-125-
-------
APPENDIX F: PER-REPAIR EMISSION REDUCTIONS FOR
STATISTICALLY 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
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
Fuel Injected 1981-1982 Vehicles
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
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
Appendix F
-126-
-------
APPENDIX G:
PER-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
Carb 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
Oist Assembly
Igni Tune Items
Vac Adv Assmb
Spk Delay Dev
Elect Tim Mod
Hoses
Wir/Hrns/Fuse
Other
EGR SYSTEM
Valv 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-ratio
-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
Appendix G
-127-
-------
SUBSYSTEM
REPAIRED
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
A HC
A CO
AIR INJECTION SYSTEM (continued)
Check Valve
Drive Belt
Hoses
Wir/Hrns/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 Cntrl Act
Air Bypas Sen
Air Divrt Act
ISC Sys
Hoses
MAT Sen
Wir/Hrns/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-ratlo
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
-
Appendix G
-128-
-------
APPENDIX H: PER-REP AIR EMISSION REDUCTIONS FOR
STATISTICALLY 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
O2 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
-
Appendix H
-129-
-------
Averae HC Benefit
Subsstem Reair
Carbureted MY 81-82 Vehicles
Carb 81-82
HC
g/mi
3.00 j
2.50 --
2.00 -
1.50 -
1.00 -
0.50
0.00
-0.50
-1.00
-1.50 -
-2.00 -
.. Carb FuelMtr Fuellnj Catalyst
Tune
ECU O2 Load All
Sensor Sensor
Averacre HC Benefit per Subsystem Repair
Carbureted MY 83-86 Vehicles
Carb 83-86
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 -
X Carb Fuel Mtr Fuel Inj Catalyst
Tune
ECU O2 Load
Sensor Sensor
All
Appendix H
-130-
-------
Average HC Benefit per Subsystem Repair
Fuel Injected MY 81-82 Vehicles
FI81-82
HC
g/mi
3.00 j
2.50 --
2.00
1.50 -
1.00 --
0.50 --
0.00
-0.50
-1.00
-1.50 4-
-2.00 --
.. Carb Fuel Mtr Fuel Inj Catalyst
Tune
ECU
O2 Load
Sensor Sensor
All
Averaye HC Benefit Per Subsystem Repair
Fuel Injected MY 83-86 Vehicles
Fl 83-86
HC
g/mi
Fuel Mtr Fuel Inj Catalyst
Tune
O2 Load
Sensor Sensor
All
Appendix H
-131-
-------
Average CO Benefit per Subsystem Repair
Carbureted MY 81-82 Vehicles
Fuel Mtr Fuel Inj Catalyst
Tune
O2 Load
Sensor Sensor
Average CO Benefit per Subsystem Repair
Carbureted MY 83-86 Vehicles
Garb 83-86
CO
g/mi
80 -r
70 -
60 -
50 -
40 -
30 -
20
10
0
-10
-20
.. Garb Fuel Mtr Fuel Inj Catalyst
Tune
ECU
O2
Sensor
Load
Sensor
All
Appendix H
-132-
-------
Average CO Benefit per Subsystem Repair
Fuel Iniected MY 81-82 Vehicles
FI81-82
CO
g/mi
80 -r
70
60
50
40
30
20 -
10 --
0
-10
-20
4. Carb Fuel Mtr Fuel Inj Catalyst
Tune
ECU
O2
Sensor
Load
Sensor
All
Average CO Benefit per Subsystem Repair
Fuel In-iected MY 81-82 Vehicles
Fl 83-86
CO
g/mi
80 j
70 -
60 -
50--
40 -
30--
20
10
0
-10
-20 -
.. Carb
Fuel Mtr Fuel Inj Catalyst
Tune
O2 Load
Sensor Sensor
All
Appendix H
-133-
-------
APPENDIX I:
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
combination of isolatable averages with all repairs.
SUBSYSTEM
REPAIRED
N
INDUCTION SYSTEM
Htd Air Door
Temp Sensors
Air Filter
Hoses
Other (Indt)
2
2
15
6
3
AVG REDUCTION
PER REPAIR
HC CO
0.06 -1.6
-
-0.01 2.2
1.13 -2.3
0.00 -0.2
FUEL METERING SYSTEM
Cart Assmbly
Fuel Meter Tune
Idl Spd Sole
Fuel Inj
Hoses
Other
Chk Adj Vacm
Vac Diaphrms
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/Hrns/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
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 11.3
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%
Appendix I
-134-
-------
SUBSYSTEM
REPAIRED
N
AVG REDUCTION
PER REPAIR
HC CO
AIR INJECTION SYSTEM (continued)
Check Valve
Drive Belt
Hoses
Wir/Hrns/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 DK/rt 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%
Appendix I
-135-
-------
APPENDIX J: COOPERATIVE TEST PROGRAM PLAN
Appendix J
-136-
-------
EPA-AA-TSS-86-99
PROGRAM PLAN
A Cooperative EPA/Manufacturer
I/M Testing Program
January 1987
NOTE: This document is a revision to the 19 August 1986
draft project proposal that was distributed to the
vehicle manufacturers by the Environmental
Protection Agency. Changes to the draft are based
upon comments received during the September 1986
workshop on the program and subsequent discussions
with the participating organizations.
Prepared by: Technical Support Staff
U.S. Environmental Protection Agency
Motor Vehicle Emission Laboratory
Ann Arbor, Michigan 48105
Appendix J
-------
Table of Contents
Section
Section 1:
Section 2:
2.1
2.2
2.3
2.4
Section 3:
3.1
3.2
3.3
3.4
3.5
3.6
Section 4:
4.1
4.2
4.3
4.4
4.5
4.6
4.7
Section 5:
-..-":*
5.1
5.2
5.3
5.4
5.5
5.6
Background and Program Summary
Vehicle Sample
Basic EPA and Manufacturer Quotas
Quotas Based on Fleet
Characteristics
Quotas Related to Catalyst
Tampering and Mis fuel ing
Quotas Related to I/M Pattern
Failures
Vehicle Procurement
Introduction
Owner Solicitation Letters
Selection of Owners for
Solicitation
Prescreening
Scheduling
Intake
As-Received Characterization
Introduction
Basic I/M Test Procedure
Tank Fuel Analysis
As-Received FTP
Flagging Vehicles for I/M
Variability
Abbreviated I/M Test Procedure
Engine/Emissions System Diagnosis
Remedial Maintenance
Introduction
Objective
Remedial Maintenance Based on the
FTP Criterion
After-Repair Testing Based on the
FTP Criterion
Remedial Maintenance Based on I/M
Variability
After-Repair Testing Based on the
Variability Criterion
Page
1
5
5
5
7
8
10
10
10
12
14
15
17
18
18
19
28
28
28
35
36
38
38
38
39
42
42
43
-------
Section
Page
Section 5: (continued)
5.7 Remedial Maintenance Based on the
Basic I/M Criterion 44
5.8 After-Repair Testing Based on the
Basic I/M Criterion 44
5.9 Other FTP Testing in the. Remedial
Maintenance Phase 45
5.10 Catalyst Replacement in the Event
of a Persistent FTP Failure 45
Section 6: Exit Tasks and Owner Compensation 47
Section 7: Documentation and Reporting 49
7.1 Introduction 49
7.2 Prescreening Data 49
7.3 Intake Data 49
7.4 Data from the As-Received
Characterizations 50
7.5 Data from the Remedial Maintenance
Phase 50
7.6 Exit Phase Data 52
Appendix A: Suspected I/M Pattern Failures in
the Seattle I/M Program A-l
Appendix B: Draft Owner Solicitation Letter B-l
-------
List of Tables
Table
1.
2.
3.
4.
5.
6.
7.
Division of Program Responsibilities
Test Matrix by Organization
Modes Performed in the Sequences of
the Basic I/M Test Procedure
Detailed Breakdown of the Basic
I/M Test Procedure
Emission Scores for a Vehicle Not
Flagged for I/M Variability
Emission Scores for a Vehicle Flagged
Due to Variability Between Sequences
Emission Scores for a Vehicle Flagged
Due to Within-Mode Variability
Page
4
6
20
22
32
33
34
-------
List of Figures
Fioure
1.
2.
3.
Timeline for Progress of a Sample
Vehicle Through the CTP
Sampling Points During the Core
Sampling Period
Generalized Flow Diagram for the
Remedial Maintenance Phase
Page
11
27
40
-------
Section 1: Background and Program Summary
Vehicle Inspection/Maintenance (I/M) programs currently
operate in 31 States, including over SO urban areas. Host of
these programs incorporate an NDIR analysis of tailpipe
emission levels for hydrocarbons (HO, carbon monoxide (CO), or
both, as an indicator of in-use emissions malperformance.
Design elements of these "traditional" I/M programs have
been under investigation by EPA for several years. Engine and
emission control technologies have evolved substantially since
I/M was first implemented, implying that the causes of
emissions malperformance and the effectiveness of existing I/M
test methods could change. In addition, EPA projects that a
number of major urban areas will fail to meet the 1987
attainment deadline for the ozone National Ambient Air Quality
Standard; one possible EPA strategy for addressing ozone
nonattainment is enhancement of traditional tailpipe I/M
programs to achieve greater reductions in mobile source HC.
Finally, studies by EPA and one manufacturer have suggested
that I/M scores of some recent vehicles are more variable than
expected; causes for the variability are suspected, but not
confirmed.
Consideration of the above issues led EPA to propose a
Cooperative Test Program (CTP) between the Agency and a number
of the motor vehicle manufacturers. The basic purpose of the
program is to examine ways to improve I/M cost-effectiveness,
focusing on 1981 and later vehicles with closed-loop fuel
metering. This will be accomplished primarily through
investigating the causes of FTP and I/M emissions failure on
individual 1981 and later light-duty vehicles (LDVs) and
light-duty trucks (LOTs), determining remedies for the
failures, and assessing the emissions performance of both the
as-received and after-repair vehicles. To date, at least six
vehicle manufacturers have agreed to participate in the program.
The specific functions of the Cooperative Test Program are
the following:
o Procure a pool of 1981 and later model-year LDVs and
LOTs that have failed an official I/M short test;
o Characterize the as-received response of each
vehicle, on commercial fuel, to a variety of I/M
test conditions, including extensive loaded pretest
operation, extended idle pretest operation, and a
cold start.
-------
- 2 -
o Refuel each vehicle with Indolene, and measure the
as-received FTP emissions.
o Conduct a complete engine and emissions system
diagnosis on each vehicle for the causes of any
observed FTP and/or I/M failures.
o Order the repairs indicated by the diagnosis
according to their anticipated FTP emissions benefit.
o Beginning with the repair of highest projected
benefit, conduct remedial maintenance where
necessary to achieve FTP HC and CO levels of at most
150% of certification standards for vehicles with
less than or equal to 50,000 miles, and 200% of
certification standards for vehicles with more than
50,000 miles, verifying emissions reductions after
each significant repair with FTP and I/M testing.
o On vehicles where acceptable FTP levels have been
achieved, conduct additional remedial maintenance as
necessary to achieve acceptable short test response,
verifying emission reductions after each significant
repair with I/M testing.
o If diagnosis indicates that the vehicle is
performing as designed, trace any remaining I/M
failure to a specific response of the vehicle to I/M
testing or pretest operation.
o Determine alternative techniques for identifying
high-emitting vehicles in the recruited sample;
consider if such procedures would have yielded more
cost-effective repairs.
EPA expects that the results of the Cooperative Test
Program, when considered with other recent and concurrent
testing programs elsewhere, will aid in accomplishing the
following objectives:
o development of advice to I/M programs on
improvements to preconditioning methods and formal
I/M test procedures;
o assessment of a limited diagnosis and repair
sequence as a remedy for a significant portion of
the in-use emissions excess;
o improvement of I/M effectiveness models;
-------
- 3 -
o feedback to the manufacturers' vehicle design groups
^0X1 particular malfunction or malmaintenance types;
o feedback to the manufacturers on the adequacy of
existing service literature for addressing I/N
failures; information to influence the preparation
of improved service and training materials.
The general approach of the Cooperative Test Program will
be to recruit I/M failures from the Michigan, Auto Exhaust
Testing (AET) program through mail solicitations, to perform
testing and repair operations at Southeast Michigan facilities
of both EPA and the manufacturers, and to accumulate the data
at the EPA Motor Vehicle Emission Laboratory (MVEL) in Ann
Arbor for generation of initial reports.
The division of responsibilities between EPA and the
vehicle manufacturers during the CTP is outlined in Table 1.
The activities are aggregated into seven phases: Recruitment,
Prescreening, Intake, As-Received Characterization, Remedial
Maintenance, Vehicle Exit Tasks, and Documentation/Reporting.
As shown in the table, the objective is to complete these
phases on between 260 and 300 vehicles.
The remainder of this program plan describes each phase of
the CTP in detail. Approval of the plan and preparation for
testing are currently in progress; recruitment will begin in
January 1987, with the first tests to take place late in
January or early February (subject to each manufacturer's
ability to intake and test at that time). Dates for completion
of testing will be dictated by the quota for a given
manufacturer, the success of the recruitment scheme, and the
test organization's own scheduling limitations. However, EPA
anticipates that testing will continue into the summer of
1987. Upon completion of testing, a data-only report will be
generated by EPA. The need and schedule for other EPA reports
or regulatory actions will be assessed once the data are
assembled. A public workshop may also be held following
release of the data-only report.
-------
- 4 -
Table 1: Division of Program Responsibilities
Phase
EPA
Manufacturers
Recruitment
Prescreening
Intake
As-Received Charac-
terizaeion
Remedial Maintenance
Vehicle Exit
Coordination with MI AET
and DOS officials
o Recruitment mailings
o Receipt of initial phone
contacts for nonpartici-
pating manufacturers
o For approximately 60 to
100 vehicles*
-intake inspection
-safety road test
-intake paperwork
o For approximately 60 to
100 vehicles:
-As-Received FTP
-As-Received Z/M
-basic emissions
systems check
o For approximately 60 to
100 vehicles:
-remedial maintenance
-after-repair testing
o For approximately 60 to
100 vehicles:
-exit inspection
-exit paperwork
-owner incentives
Coordination with EPA
on scheduling and
assessment of quotas
o Receipt of initial phone
contacts for owners of
their own makes
o For approximately 200
vehicles:
-intake inspection
-safety road test
-intake paperwork
o For approximately 200
vehicles:
-As-Received FTP
-As-Received Z/M
-basic emissions
systems check
o For approximately 200
vehicles:
-remedial maintenance
-after-repair testing
o For approximately 200
vehicles:
-exit inspection
-exit paperwork
-owner incentives
Documentation and
Reporting
Receive data in standard
format and assemble data
base
Provide data to EPA in
standard format
Generate data-only re-
port with simple summary
statistics
o Generate own report(s)
as desired
-------
- 5 -
Section 2: Vehicle Samele
2.1 Baaic EPA and Manufacturer Quotas
In order to meet the program objectives, each
participating organization agrees to an overall testing quota.
A vehicle counts towards that quota if it completes all phases
of the CTF; that is, following procurement (Recruitment,
Prescreening, and Intake Phases), the vehicle undergoes the
As-Received Characterization, Remedial Maintenance Phase, and
completion of the proper documentation. As will be evident in
later sections, a vehicle need not necessarily undergo repairs
in order to complete the Remedial Maintenance Phase; these
vehicles will count towards the test facility's quota.
For most of the participating organizations, the CTF
quotas are identical to those suggested in EPA letters to the
manufacturers inviting participation in the cooperative program
(see Table 2). The levels for Honda and Mitsubishi were
reduced slightly to reflect a lower anticipated share of total
I/H failures, based on an analysis of 140,000 vehicles in the
Seattle I/M program. The Seattle program was selected because
it is similar to Michigan in its use of an idle-neutral 'short
test with 2500rpm preconditioning, and its use of a restart
procedure for vehicles manufactured by Ford. Of course, any
manufacturer may choose to test additional vehicles.
EPA's test quota for the program is 60 vehicles, but the
Agency may test as many as 100 vehicles. The EPA share will
include the fleets of manufacturers who do not have local test
facilities or who are unable to participate in the program.
Based on the quotas from Table 2, at least 260 vehicles
will complete the Cooperative Test Program. The vehicle
manufacturers will complete approximately 200 vehicles of this
total, and EPA will test a minimum of 60 vehicles.
2.2 Quotas Based on Fleet Characteristics
In order to meet the study objectives, the CTP recruitment
must control for a variety of fleet characteristics, including
model year distribution, fuel metering technology, and vehicle
type. The following considerations will apply to all
participating manufacturers:
o All test vehicles must employ closed-loop fuel
metering.
-------
- 6 -
Table 2: Test Matrix bv Organization
Seattle Data
Organization
Chry
Ford
GM
Hond
Niss
Kits
Toyt
Subtotals
Other mfrs
EPA
Totals
LDV
Fail
141
1183
962
164
433
92
115
2975
710
N/A
3685
LDT
Fail
93
436
351
0
478
22
193
1380
1151
N/A
2531
% Tot
Fail
3.8
26.0
21.1
2.6
14.7
1.8
5.0
70.1
29.9
N/A
100.0
CTP
Test
Quota
15
60
60
10
30
10
16
201*
N/A
60
261
* Participation of one additional manufacturer, uncertain at
this writing, would increase this number.
-------
- 7 -
o Carbureted vehicles must not exceed 50% of each
organization's basic test quota.
o At the test organization's option, carbureted
1983 MY and later vehicles may be excluded from
testing.
o If higher Michigan AET failure rates in light-duty
trucks threaten to skew the sample towards a limited
number of control technologies, a manufacturer may
choose to limit recruitment of LDTs, with.
consideration for similarities in the LDV and LDT
systems and the manufacturer's fleet mix.
The above requirements imply that if a manufacturer
produced only open-loop vehicles in the 1981 and 1982 model
years, its CTF testing quota will be filled exclusively with
1983 NT and later vehicles.
Additional considerations apply to those manufacturers
that produced at least some closed-loop vehicles in either the
1981 or 1982 model years:
o Half of each manufacturer's quota will be filled by
vehicles from the 1981 and 1982 model years; the
other half will be 1983 NY and later vehicles.
o NY 1981 and 1982 vehicles will be recruited without
regard to fuel metering type until 50% of the
1981/82 slots become filled with carbureted
vehicles; at that point, the testing organization
may choose to either (1) continue randomly
recruiting both carbureted and fuel injected
vehicles until either the overall limit on
carbureted vehicles or the overall 1981/82 quota is
reached, or (2) exclude additional carbureted
1981/82 vehicles from testing.
The purpose of the model year quotas is to ensure that
data are gathered on vehicles with the latest control
technologies, as well as those with age-related malperformances.
2.3 Quotas Related to Catalyst Tampering and Misfuelinq
EPA recognizes that vehicles that have been misfueled or
had their catalysts removed may not provide useful information
to the vehicle manufacturers on the particular emissions
performance of their vehicles. Therefore, each manufacturer
may choose to limit the number of vehicles with any combination
-------
- 8 -
of fuel inlet tampering, Plumbtesmo test failure, or catalyst
removal to one vehicle or 10% of that manufacturer's overall
CTP quota.* whichever is greater. After the limit is reached,
such vehicles may be returned to their owners without being
tested or counted as part of the sample, or they may be
retained and tested, at the test facility's option.
In addition to the above limitation, the remedial
maintenance procedures for vehicles with catalyst tampering or
evidence of misfueling will differ from the procedures used on
other vehicles. This issue will be addressed in Section 5.
2.4 Quotas Related to I/M Pattern Failures
EPA defines an I/M pattern failure as a vehicle group that
fails an approved I/M short test at an unusual rate, and is
known (or strongly suspected) to fail due to a common cause.
Some pattern failures are traced to malfunctions or component
defects; others occur in vehicles that are performing as
designed, but the design conflicts with some aspect of the test
procedure in a way that frequently causes failing I/M scores.
The approach to pattern failures in the CTP is based on
the objective that test slots should not be filled with
vehicles whose failures may be predicted and explained with
reasonable accuracy in advance of testing. Before the CTP
begins, EPA will therefore provide each manufacturer with
summaries of those vehicle groups that, in the Agency's
opinion, could reasonably be excluded from the CTP recruitment
pool. Each participating organization will then have the
opportunity to amend or supplement this information. When EPA
and a manufacturer agree that a pattern failure group has been
identified and adequately explained, the vehicle group will be
excluded (to the extent possible) from CTP solicitation and
recruitment.
EPA believes that some vehicle groups with observed high
failure rates in operating I/M programs will not be understood
well enough to justify excluding them from recruitment at the
outset of the program. However, the CTP testing itself may
confirm the? cause of failure in some cases, implying that
further testing of the "newly confirmed" pattern failure in the
CTP would be unproductive. In such cases, the participating
manufacturers agree to meet with EPA to reach agreement on the
status of the vehicle groups. After the need to exclude a
group from further CTP testing is agreed upon, the manufacturer
may choose to reject the affected vehicles during the
prescreening phone call (see Section 3.4). In addition, EPA
will act as quickly as possible to remove the vehicles from
future solicitation mailings.
-------
Appendix A provides an EFA list of vehicle groups that
have shown unusual failure rates in the Seattle I/M program; in
most of these cases, a common cause of failure has not been
identified:" by the Agency. Note that the analysis currently
covers only a few model years, and that open-loop vehicles have
not been excluded. Both before and during the CTP, the
manufacturers may wish to make special efforts to determine if
any of these groups display patterns of I/M failure. Examples
of such a determination might be the following: a defective
part sold only in the group in question; a group-specific quirk
in the calibration; an assembly error that could be systematic;
an act of tampering that improves on an otherwise poor
driveability characteristic.
EFA may choose on it own to recruit and test suspected
pattern failure vehicles using an approach similar to the that
of the CTP program; however, such vehicles will not count
towards EFA's testing quota.
-------
- 10 -
Section 3: Vehicle Procurement
3.1 introduction
All vehicles will be procured for the Cooperative Test
Program with direct mail solicitations to owners who are about
to undergo testing in the Michigan Auto Exhaust Testing
Program. EPA will organize and execute the solicitation
mailings; however, each participating organization may choose
to have the letters bear its own letterhead and return
address. The letter itself will prompt the vehicle owner that
fails his or her I/H test to telephone a designated
representative of the appropriate test facility or its
contractor. During the phone call, owners will be prescreened
for various acceptance criteria and then scheduled for intake
into the testing phases of the program.
Figure 1 illustrates the timeline that a sample vehicle
might follow as it progresses through procurement and testing
in the CTP. The timeline is bounded by the endpoints of the
State of Michigan vehicle registration process, beginning with
the mailing of a registration and emission-test reminder to the
owner, and ending with the deadline for vehicle registration.
For the sample vehicle in the figure, this period spans 55
days. The legal minimum in Michigan is 45 days, with almost
all vehicles falling in the 50- to 60-day range.
The procurement phases (recruitment, prescreening, intake)
together occupy about half of the sample vehicle's CTP
"lifetime." This represents a balancing of several
constraints, including the relatively tight window afforded by
the Michigan registration process, the need to keep owner
response rates high, and the need to provide the test facility
with sufficient testing, diagnosis, and repair time. The
resulting owner deadlines and recommended durations for each
procurement phase are illustrated in Figure 1 and discussed
below.
3.2 Owner Solicitation Letters
As mentioned previously, EPA holds the responsibility for
executing the CTP owner solicitation mailings. The CTP will
begin with a weekly mailing schedule, offset from the Michigan
DOS registration reminder mailings by approximately four
working days. The offset provides time for EPA to apply the
various CTP quotas to the Michigan registration files and to
generate the solicitation package.
-------
Figure 1: Timeline for Progress pf a, SaHBle Vehicle Through the CTP
CTP Phase
Week Nuaber:
Day Muaber:
Recruitment Phase -;k
'V , ',
HI DOS generates weekly tap* «itb
registration data
EPA receives copy of HI tape
HI registration reminders Miled
CTP solicitation sailed
HI Auto Exhaust Test perforated
Prescreening call date
Intake date
CTP As-Received Characterization
and Remedial Maintenance Phase
Exit task dates *
Owner registers vehicle
Registration deadline
2
4
3
II
4
IB
5
25
6
32
7
39
8
46
9
53
10
60
11
67
12
72
XS
-X
s
X-SX
X
sx
xsx
X-
f
Xc
-X
xs=
-X
s
s
Note: S===S indicates days during which the indicated CTP Phase occurred for this saaple vehicle. , ,K
^^ XX indicates range of additional days on which the indicated CTP Phase could hive Otturred for this sample vehicle, given the
preceding values of S===S.
-------
- 12 -
A draft of the CTP solicitation letter, written for a
hypothetical company called Acme Motors, is provided as
Appendix 8; the important elements to note are the following:
o potential eligibility for an important study
program, based on I/M test failure;
o potential benefits/incentives for the owner;
o qualifiers to inform the owner that acceptance to
the program is not guaranteed.
o a ten-working-day deadline for the owner to obtain
the Michigan auto exhaust test (AET);
o a two-working-day deadline between when the AET test
is performed and when the owner contacts the test
facility;
o contact phone number at the testing organization, to
either obtain additional information or to enter the
program following test failure;
o importance of not having repairs conducted prior to
intake;
o the fact that the solicitation is not transferable
to other vehicles.
3.3 Selection of Owners for Solicitation
One objective for EPA in the solicitation mailings will be
to minimize the number of owners who receive a mailing but
whose vehicles are not eligible for CTP testing. Meeting this
objective will require screening the Michigan registration
files for vehicles that should be excluded at the outset, such
as pre-1981 or open-loop vehicles. Additional groups of
vehicles will need to be excluded as test slots are filled, and
the various quotas are met; this updating process will
necessarily require close coordination between the
participating manufacturers and EPA.
The process for culling names from the Michigan files is
under development by EPA, dependent in part on the debugging of
new VIN decoding software for the 1981-1984 model years. The
likely process will be as follows:
1. On a weekly basis, the Michigan DOS will provide EPA
with a data tape containing only those vehicles that
-------
- 13 -
are due to receive registration reminders during the
upcoming week. The tape will include vehicle
identifying information (VIN, model year, make),
owner identifying information (address label
fields), and registration information (expiration
date).
2. For 1981-1984 MY vehicles, EPA will employ a VIN
decoding program to generate engine family and
emission control system information for the vehicles
on each Michigan tape. EPA will use the output of
the VIN decoder to eliminate 1981-1984 MY vehicles
from the tape that fail to meet the required fleet
characteristics (See Section 2.2).
3. For 1985-1986 MY vehicles, EPA requests that each
manufacturer provide an algorithm sufficient to
segregate all the eligible vehicles from the
Michigan data. The algorithm may be in narrative
form. An example (consistent with the approach EPA
will take to the 1981-1984 MY vehicles) would be a
decoder that uses the VIN and model year fields- to
distinguish the correct model years, vehicle types
(LDV, LOT), fuel metering types (carbureted,
fuel-injected), and feedback strategies
(closed-loop, open-loop).
4. On an ongoing basis, updates from the manufacturers
on filled quotas will be used to cull additional
vehicle types from the registration tape.
5. Currently available information on each facility's
testing capacity and quotas, and the response
patterns of vehicle owners to the solicitations will
be used to determine a target size for each
organization for the pending mailing.
6. The vehicles remaining in the culled version of the
Michigan tape will constitute the recruitment pool
for the pending mailing.
7. Owners will be randomly selected for solicitation
from the recruitment pool, until the target mailing
size for each organization has been reached.
8. Mailing labels will be generated for all those
owners who have been* selected for solicitation, and
the mailing will be executed.
-------
- 14 -
As the CTP progresses, the weekly solicitation approach
will be reevaluated and changed as necessary.
3.4 Prescreenino;
As described in Section 3.1 above, a vehicle owner who
wishes to participate in the Cooperative Test Program must
first fail an I/M test and then telephone the contact person
listed in his or her solicitation letter. One purpose of the
phone call is to provide the testing organization with an
opportunity to prescreen the vehicle. Each manufacturer is
responsible for prescreening its own vehicle makes, using the
following steps:
o delivery of a brief introduction to the Cooperative
Test Program;
o gathering data on the vehicle and its I/M test
history through a telephone questionnaire;
o rejecting vehicles that fail to meet recruitment
quotas or other acceptance criteria ("prescreen
rejection");
o providing an explanation of incentives to owners who
have passed prescreening;
o supplying owners with information necessary for
intake of the vehicle into the testing program.
The prescreening questionnaire format will be similar to
that of the EPA Emission Factors test program. The
questionnaire has two purposes: determining the eligibility of
the vehicle for the program (information that will later be
verified in person during the Intake Phase); and gathering
background information on the vehicle, its maintenance history,
and its I/M test history that may supplement the diagnostic and
testing work of the CTP. (Clearly, the maintenance information
from the> questionnaire should be used carefully, given that
owner-supplied data may be unreliable.)
The basic prescreening factors are the following:
o owner contact initiated within specified times from
mailing of solicitation letter and following
Michigan I/M test;
o I/M failure status;
-------
- 15 -
a manufacturer and model year;
o other recruitment quotas, as applicable (see
Section 2.4);
o absence of post-failure repair;
o standard Emission Factors disqualifying factors,
including off-road use, major engine modifications,
and excessive towing.
Engine modifications that EPA considers cause for
rejection include radical carburetor changes, addition of
headers, and engine switches. Vehicles with aftermarket air
conditioning systems would be accepted for testing. Vehicles
with evidence of catalyst tampering or misfueling would be
accepted, subject to the quota on such vehicles described
earlier in Section 2.3. During the prescreening call, however,
owners will be advised that the CTP is not responsible for
repairs to tampered emission controls that are needed to remedy
the Michigan AET test failure, and the promised incentive of a
passing AET certificate on such vehicles would no longer apply
(see also Section 6).
Vehicles that have received previous tests in the Michigan
AET program (either in the previous year's inspection cycle, or
multiple tests in the current year) will be accepted into the
program, provided they have not been repaired during the
current year's inspection cycle.
A form will be provided by EPA to log incoming calls so
that the effectiveness of the recruitment system and the causes
for presreening rejection can be documented. Vehicles that
meet the prescreening criteria will receive the remainder of
the vehicle history portion of the prescreening questionnaire.
3.5 Scheduling
After completing the telephone questionnaire, owners who
have passed prescreening will be given instructions for intake
into the test program itself. Each test facility may choose
between- two scheduling options: just-in-time scheduling, or
banked scheduling. With just-in-time scheduling, the facility
accepts only those vehicles that fill the open test slots in
the immediate future; once the near-term slots are full, owners
(even those that would otherwise have been CTF-eligible) are
rejected at the prescreening call. As vehicles near completion
of testing, the facility anticipates the next .test slots to
open up, and gives the "green light" to once again accept
owners during prescreening.
-------
- 16 -
With banked scheduling, the facility first fills the open
test slots in the- immediate future; then, rather than rejecting
owners once these slots are full, the facility "banks" all
vehicles that remain eligible following prescreening until new
test slots open up. As soon as the facility can anticipate the
intaJce data for the next banked vehicle, the next banked owner
is contacted and a delivery date arranged.
If a test facility chooses the banked scheduling approach,
owners should be given a closure date, past which he or she
should assume that the vehicle will not be accepted for
testing, and other remedies for the I/M failure should be
pursued. Provisions for an incentive to the owners in such
cases are left to the discretion of the participating
facilities.
Whichever scheduling approach is adopted, the facility
must always accept the first eligible owner that is available
to fill a given test slot. This is necessary to prevent
nonrandom effects from influencing the sample selection.
The just-in-time scheduling approach assumes that the flow
of vehicles through the program will be limited by the testing
capacity of each facility, while the banked scheduling approach
providas insurance against periods when the flow of vehicles
will be limited by owner response rates. At this point', EPA
has insufficient information to predict which approach will be
the most efficient. For example, low owner response rates
might imply that not all test slots can be filled on short
notice, even if the maximum recruitment pool in a given week
receives the CTP solicitation. Each participating organization
may therefore wish to make provisions for changing their
scheduling approach during the CTP, should the need arise.
Another scheduling issue concerns the rigidity of the
vehicle registration deadline imposed on the owner by the State
of Michigan. CTP testing should normally be completed one week
ahead of this deadline, in order to give the owner time to
register the vehicle. Cases may arise where the testing
facility wishes to retain a vehicle past that one-week
cushion: the remedial maintenance of a vehicle may take longer
than expected; a vehicle may become available late in its
registration cycle, yet be ideal for a hard-to-fill test slot.
In these cases, the test facility will be responsible for
providing temporary registration of the vehicle. The State of
Michigan provides two-week temporary registrations for a
five-dollar fee; this registration need not be purchased by the
owner. Proof of insurance and proof of the vehicle
identification number (from an old registration, for example)
are required. Close coordination between EPA and the
manufacturers will be necessary to keep the number of these
vehicles to a minimum.
-------
- 17 -
3.6 Intake
Intake of procured vehicles will occur at a location
agreed upon by the owner and the testing organization
(manufacturer or EPA, as appropriate). The location will
determine to some extent the order of the following steps in
the intake phase:
o delivery of the test vehicle; (optional) exchange
for leaner vehicle;
o outline of the owner incentives (see Section 6).
o correction of omissions in the prescreening
questionnaire; verification of prescreening
acceptance criteria;
o execution of intake vehicle checks, including a
"scratch/dent" inspection;
o execution of a road test, to determine if the
vehicle is safe for dynamometer and I/M testing;
o rejecting vehicles that fail to meet acceptance
criteria ("intake rejection");
o obtaining signed releases from the owner for testing
and repair of the vehicle;
o providing preliminary information on return of the
vehicle following testing.
Standard EPA intake procedures call for the owner to be
present during the vehicle inspection to limit the test
organization's liability for future claims of vehicle damage
during the test program.
Vehicles that are released to the test organization by the
owner and that pass their safety road test will proceed
immediately into the As-Received Characterization, described in
the next section.
-------
- 18 -
Section 4. As-Received Characterization
4.1 Introduction
The As-Received Characterization consists of a sequence of
FTP and I/M testing and a complete engine/emissions system
diagnosis, but no repairs. The results of these tests are used
as the baseline for judging the effectiveness of subsequent
remedial maintenance efforts. The steps in the As-Received
Characterization, conducted in order, are as follows:
1. Conduct the Basic I/M Test Procedure (Section 4.2)
on the as-received vehicle. Use tank fuel, if the
vehicle is procured with sufficient fuel; otherwise,
use a commercial fuel of the test facility's choice.
2. Perform an RVP determination and a lead-in-fuel
analysis on the fuel used in the as-received Basic
I/M Test Procedure.
3. Drain the fuel tank and fill to 40% with Indolene.
4. Conduct a standard overnight FTP prep.
5. Conduct an As-Received FTP test.
6. Analyze the results of the Basic I/M Test Procedure
in Step 1 above to flag vehicles with variable I/M
test results, either between one sequence of the
procedure and another, or between readings at
different times in a single mode (Section 4.5).
7.- Develop an Abbreviated I/M Test Procedure that
displays the failure behavior of the particular
vehicle and which will clearly show when and if a
repair has eliminated that behavior (Section 4.6).
8. Conduct the Abbreviated I/M Test Procedure on the
as-received vehicle, using Indolene fuel (no refill
necessary following the As-Received FTP).
9. Conduct a complete engine and emissions system
diagnosis. (Section 4.7).
10. If necessary, revise the Abbreviated I/M Test
Procedure according to the results of the tests on
Indolene, for later use during remedial maintenance.
Aspects of the As-Received Characterization are described
in greater detail in the sections that follow.
-------
- 19 -
4.2 Basic I/M Test Procedure
One purpose of the Cooperative Test Program is to
determine^ the effects of pretest operation and different
sampling approaches on I/M scores. (Other factors that may
have affected the Michigan AET results of a test vehicle, such
as hardware calibration variables or operator fraud, are
difficult or impossible to assess in the CTF). For this
investigation. the CTP relies on four I/M sequences,
collectively called the Basic I/M Test Procedure. The four
sequences are called the cold start sequence, the extended
loaded sequence, the extended idle sequence, and the restart
sequence.
Table 3 lists the modes that are performed in the
sequences. As shown in the table, each sequence consists of
one or more pretest operating modes, each followed by a core
four-minute emissions sampling period. Examples of- the pretest
operating modes include a one-hour 75°F soaJc, an LA4 cycle, a
20-minute idle-neutral, and a three-minute 2500 cpm-neutral.
The core sampling period consists of three modes: a first
idle-neutral of 30 seconds, a 2500 rpm-neutral of 30 seconds,
and a second idle-neutral of 120 seconds. The sequence in
which the modes occur in the Basic I/M Test Procedure may be
duplicated by reading off the modes column by column, moving
from left to right.
The Basic I/M Test Procedure is laid out in greater detail
in Table 4. The selection of modes and their duration reflects
the following testing objectives, among others:
Cold Start Sequence;
o characterize emissions of vehicle at abnormal (low)
engine operating temperature;
o determine ability of certain operating modes to
achieve normal engine operating temperature;
o characterize emissions of vehicle when normal engine
operating temperature is achieved through certain
modes of operation.
Extended Loaded Sequence;
o use extended loaded operation to achieve ideal
operating condition;
o characterize short test emissions of vehicle
immediately after the loaded pretest operation.
-------
- 20 -
Table 3: Modes Performed in the Sequences
of the Basic I/M Test Procedure
Sequence:
Initial Mode:
Succeeding Modes
Core Sampling
1st Idle-neutral
2500 rpm-neutral
2nd Idle-neutral
2500 rpm-neutral:180 sec
Core Sampling
1st Idle-neutral
2500 rpm-neutral
2nd Idle-neutral
Idle-neutral: 10 rain
Core Sampling
1st Idle-neutral
2500 rpm-neutral
2nd Idle-neutral
Cold Extended Extenaed
Start Loaded Idle Restart
75° soak LA4
Restart
X
X
X
X
X
X
X
X
X
X
X
X
20 min
Idle-N
X
X
X
LA 4,
Restart
X
X
X
X
X
X
Note: For a detailed description of the actual sequence of
steps in the Basic I/M Test Procedure, refer to Table 4.
-------
- 21 -
Extended Idle Sequence:
o use extended loaded operation to achieve ideal
operating condition;
o characterize the short test emissions of the vehicle
during and after extended idle.operation;
o follow the extended idle operation with a period of
off-idle no-load operation, and characterize the
short test emissions of the vehicle during and after
this period.
Restart Sequence
o use extended loaded operation to achieve ideal
operating condition;
o characterize the short test emissions of the vehicle
after the restart
The Basic I/M Test Procedure is conducted with the hood in
the raised position. External cooling fans may be used, but
only during modes (such as the LA4) where the vehicle is in
motion under load on the dynamometer.
During the procedure, HC (ppm), CO (%), and CO, (%)
values are measured on nondispersive infrared (NDIR)
analyzers. EPA believes that the C0t values are useful both
as a sample dilution check and as a diagnostic tool; however, a
facility that lacks equipment to simultaneously monitor CO,,
CO. and HC should give priority to measuring CO and HC. Engine
rpm values are measured using either inductive tachometers or
an ECM datalinfc, at the test facility's option. Engine coolant
temperature, which will be used in assessing vehicle
variability, should be measured at the engine block or other
location upstream of the thermostat.
The EPA's own test setup will include constant monitoring
of the basic emissions and rpm values on multichannel strip
chart recorders, as well as real-time manual recording of
specific numerical values on data sheets. Analyzers and chart
recorders will be calibrated to permit later verification of
indivfdual values. At a minimum, the EPA calibration and
maintenance requirements applying to field I/M instruments (40
CFR Part 85, Subpart W) will be employed; EPA customarily
performs calibrations on a more frequent basis. Facilities are
also encouraged to conduct frequent through-the-probe zero air
checks, as well as electrical zero checks..
Table 4 indicates the sampling intervals for the various
parameters discussed above. Consider, for example, the first
-------
- 22 -
Table 4» Detailed Breakdown of _the Basic I/M Teat Procedure
Mode
Parameters Monitored
Sequence
Wo. Name
Duration Item
A. Cold Start 1. 75*r soak 60 min
2. start engine
3. idle-neutral 30 sec
4. 2500rpm
6. 2500rpm
30 sec
5. idle-neutral 120 sec
180 sec
7. idle-neutral 30 sec
8.* restart
9. 2500rpra 30 sec
10. idle-neutral 120 sec
11. idle-neutral . 10 min
12. idle-neutral 30 sec
none/engine off
none
HC.CO,C02.RPM
coolant (°F)
HC.CO.C02.RPM
coolant (*F)
HC.CO.C02.RPM
coolant (°F)
HC,CO.C02.BPM
coolant (*F)
HC.CO.C02.RPM
coolant (°F)
none
HC.CO,C02,RPM
coolant (*F)
HC.CO.C02.RPM
coolant (°F)
HC.CO.C02,RPM
coolant (*F)
HC,CO,C02.RPM
coolant CF)
Sample at;
sec a IS, 30, &
stability
sec a 0
sec a IS, 30, &
stability
sec 3 o
sec s 15, 30, 60, 90,
120, & stability
sec s 0
sec s IS, 30, 60 ...
180, & stability
sec 3 0
sec 3 IS, 30, &
stability
sec 3 o
sec 3 is. 30, &
stability
sec s o
sec 3 15. 30, 60. 90.
120, & stability
sec a 0
min 3 1, S, 10, &
stability
min s 0, 5
sec 3 15. 30, &
stability
sec 3 0
* Ford vehicles only.
-------
- 23 -
Sequence
B. Extended
Loaded
C. Extended
Idle
Mode
Table 4 (continued)
Parameters Monitored
No. N
A. Cold Start
(cont)
Duration Item
13.* restart none
14. 2500rpm 30 sec HC.CO.C02,RPM
coolant (°F)
15. idle-neutral 120 sec HC.CO.C02,RPM
coolant (*F)
1372 sec none
30 sec HC.CO,C02,HPM
1. LA4 prep
2. idle-neutral
3.* restart
4. 2SOOrpm
none
30 sec HC.CO.C02.RPM
5. idle-neutral 120 sec HC,CO.C02,RPM
1. idle-neutral
2. idle-neutral
20 min HC.CO.C02,HPM
coolant (*F)
30 sec HC.CO,C02,RPM
3.* restart
4. 2500rpn
none
30 sec HC.CO,C02,aPM
coolant (*F)
5. idle-neutral 120 sec HC.CO,C02,RPM
coolant CF)
Sample at;
sec a 15, 30. &
stability
sec a o
sec = 15, 30, 60, 90,
120, & stability
sec s o
sec s 15, 30,
& stability
sec = 15, 30, .
S. stability
sec a 15, 30, 60, 90,
120, & stability
rain a 1, 2...20 &
stability
min =0, 5. 10
sec 3 15, 30,
& stability
sec 3 15, 30, S.
stability
sec s 0
sec 3 15, 30, 60, 90,
120, & stability
sec a 0
* Forda only.
-------
- 24 -
Sequence
C. Extended
Idle (cont)
D. Restart
Mod*
Table 4 (continued)
_ Parameters Monitored
tip. Mama
6. 2500rpa
7. idle-neutral
Duration Item
180 sec HC.CO.C02.HPM
coolant (9F)
30 sec HC.CO.C02,RPM
3.* restart
9. 2SOOrpa
none
30 sec HC,CO,C02.RPM
coolant (°F)
10. idle-neutral 120 sec HC,CO.C02,RPM
coolant (°F)
1. LA4 1372 sec cone
2. idle-neutral 30 sec HC.CO.C02.RPM
3.0 restart
4. 2SOOrpn
none
30 sec HC.CO.C02,RPM
5. idle-neutral 120 sec
HC,CO,C02.HPM
Sample at;
sec 3 15, 30, 60...
180, S. stability
sec s 0
sec s 15, 30,
& stability
sec s 15, 30, &
stability
sec 3 0
sec s 15, 30, 60, 90,
120, & stability
sec s 0
sec 3 15, 30
& stability
sac 3 15, 30
& stability
sec 3 15, 30, 60, 90.
120, & stability
* Fords only.
3 Non-Fords only
-------
- 25 -
appearance of the core sampling period, in modes three to five
of the cold start sequence. During the first idle-neutral and
the 2500 rpn-neutral, emission scores and engine rpm values are
recorded at a minimum of two points: the midpoint of the mode
(15 seconds), and the end of the mode (30 seconds). A third
set of emission/rpm readings is also taken during the mode if
the HC and CO values stabilize simultaneously for at least five
seconds after the "transport lag" at the beginning of the
mode. The elapsed time-in-mode at the end of this five-second
stabilization period is also recorded. If more than one such
five-second period of stabilized readings occurs during the
mode, values are only recorded for the first occurance.
During the second idle-neutral mode, readings are taken at
a minimum of five points: at the 15-second point, and at four
thirty-second intervals spread through the mode. As above,
emission and elapsed time readings are also taken for the first
five-second occurance of "stabilized" HC and CO values, if it
occurs.
The initial point of a mode (t » 0) is defined by engine
rpm and the mode type. The initial point for an idle-neutral
mode is the instant when the engine rpm goes above 350 rpm
(from a restart, for example), or drops below 1600 rpm
(following a 2500 rpm-neutral, for example). The initial-point
for a 2500 rpm-neutral mode is the instant when the engine rpm
goes above 2200 rpm or drops below 2800 rpm. The initial point
of a mode must be re-initiated if the vehicle deviates from
these rpm bounds. At some facilities, time-in-mode will be
monitored manually, which may raise the issue of the accuracy
of the sampling points. EPA suggests ±2 sec as an objective
for the accuracy of time measurements during the core sampling.
Determinations of "stabilized" emissions must necessarily
be more subjective. There are really two purposes for
monitoring stability in the CTP: (l) to determine if clearcut
changes occur in emission levels continuously throughout a
mode, and (2) to determine if emissions are mostly stable
throughout a mode, but clearcut gross changes do occur. The
first purpose is addressed in the CTP by recording a
five-second period of stability within each mode, if such a
period occurs. The second purpose is addressed by comparing
the five-second period of stability to other fixed-time
sampling points in the mode. With these purposes in mind, the
choice of what constitutes a "clearcut change" in emission
values is left to the test facility. The selection of a
five-second period and limiting data to only the first such
occurance reflects the fact that EPA is currently reviewing a
similar algorithm for incorporation into EPA-recomraended
procedures for computerized NDIR analyzers.
-------
- 26 -
Figure 2 illustrates the points where the fixed-time
emission values would be taken on a hypothetical vehicle
undergoing^ the core sampling procedure in the restart
sequence. 'The rpm trace shows the final few seconds of the IJU
pretest operation, the restart, and the three modes of the core
sampling period. The HC and CO emissions traces are offset in
elapsed time from the rpm trace due to the approximately
seven-second transport time in the analysis system.
The discussion above illustrates the recording of values
during the core sampling period. The process is similar for
all the other modes of the Basic I/M Test Procedure, although
the sampling intervals may vary from mode to mode.
In addition to the basic emission values, engine rpm, and
coolant temperature values, some facilities may find it
valuable to monitor additional engine or emission control
parameters during the Basic I/M Test Procedure. For example,
oxygen sensor voltage might be monitored as an indicator of
feedback control status, or changes in secondary air routing
might be recorded. The results could be used to indicate
points at which the vehicle deviates from its ideal operating
condition.
A facility should only monitor extra parameters in the
Basic I/M Test Procedure if it is clear that the process of
taking the measurements will have no impact on the emission
scores and will not modify the status of any component on the
as-received vehicle.
A review of the above description will show that the Basic
I/M Test Procedure may be used to measure to measure I/M scores
of the test vehicle during
o the formal test sequences for non-loaded short tests
specified in the EPA performance warranty regulations
o additional formal test sequences arrived at by
changing the duration of the sampling modes.
The effects of pretest vehicle operation on these scores are
examined by looking at
o pretest operating modes that are probably occurring
in actual I/M programs (extended idle, for example);
o pretest operating modes that are probably not
occurring routinely in the field, but might be
considered as required preconditioning before the
formal test (extended off-idle no-load operation,
for example).
-------
2000
JC
fc
M
&
1000
S
o
so
no
>i
n>
D
IQ
0
30
t
60
90 !20
Total Elapsed Time (seconds)
150
1RO
-------
- 28 -
Note that not all of the conditions are presumed to be
good indicators of actual emission problems with a vehicle;
scores ducing parts of the cold start sequence, for example,
may be poor indicators of emission performance because coolant
temperature has not yet reached normal operating levels.
Facilities may wish to add short test sequences that they
consider important to characterizing the vehicle's behavior in
other I/M testing situations. If additional sequences are
indicated, however, the importance of consistency between all
the participants in the program should be taken into account.
4.3 Tank Fuel Analysis
Lead-in-fuel analysis should be conducted using a
procedure at least as accurate as x-ray flourescence, and
designed to designate the fuel as either above or below a
0.5 g/gal standard. Reid Vapor Pressure testing will be
conducted with the ASTM D 323 method or an equivalent
semiautomated method. EPA has the capacity to perform either
the D 323 method or a semiautomated approach with Herzog test
equipment. With prior arrangement, EPA will perform a limited
number of these tank-fuel tests for other facilities. If the
as-received vehicle was low on tank fuel, and the facility's
commercial fuel was employed, the lead-in-fuel analysis may be
eliminated.
4.4 As-Received FTP
The FTP testing in the As-Received Characterization is
standard three-bag CVS testing on the urban driving cycle,
without a heat build, and without the Highway Fuel Economy
Test. At the test facility's option, back-to-back LA4 cycles
(without a drain and refill) may be used in place of the single
LA4 in the FTP prep. As with the Basic I/M Test Procedure, the
test facility may measure additional parameters on the FTP, if
such measurements do not in any way alter the vehicle or
emission system operation in a way that would affect the
outcome of the FTP. The test facility may also elect to employ
"slave" tires in place of the as-received tires.
4.5 Flagging Vehicles for I/M Variability
Variable short test results may be a critical factor in
I/M programs. In the Cooperative Test Program, the objectives
related to variable vehicles are the following:
o to hypothesize the role that variability might have
played in the original Michigan AET test failure;
-------
- 29 -
o to determine whether the vehicle might show variable
test results in other existing I/M programs,
triggered by factors either in the formal test
procedure or in the pretest operation of the vehicle;
o to determine whether the vehicle might show variable
test results given certain changes to the I/M test,
either in the pretest operation of the vehicle or in
the formal test itself.
o to the extent possible, explain the observed
variability, either in terms of vehicle malfunction
or aspects of vehicle design.
Vehicles are "flagged" as variable according to the
results of the Basic I/M Test Procedure, using criteria to be
described below. The consequence of being flagged is that the
vehicle will undergo somewhat different analysis during the
Remedial Maintenance Phase of the CTP. The I/M procedures that
are repeated after each FTP and used to monitor the impacts of
repair may also be different for these vehicles. As will be
seen in Section 5, this does not necessarily imply extra repair
or retest efforts.
Two types of variability are of interest: variability
between sequences of the Basic I/M Test Procedure* and
variability between samples taken at different times in the
same mode. In the first type, the likelihood is that
differences in pretest operation between the two sequences
trigger the change in emissions. For example, following the
LA4 at the beginning of the extended loaded sequence, a vehicle
might exhibit passing readings throughout the rest of the
sequence; however, the long stretch of idle pretest operation
in the extended idle sequence might trigger changes in the
vehicle's control systems, leading to much higher readings.
On the other hand, variability between samples in the same
mode might be triggered by the type or duration of the mode,
and not by pretest operation. For example, a vehicle might
show failing readings fifteen seconds into any idle-neutral
mode, but passing readings at 30 seconds regardless of how
the vehicle was operated just prior to that mode.
The question of designating a CTP vehicle . as either
variable or not variable may be straightforward in some cases.
For example, a brief review of the chart traces may show
emission scores on one sequence that are consistently well
above the Federal 207(b) cutpoints; on another sequence the
scores may be consistently very low. On one mode of the
extended loaded sequence, the emission values may take a sudden
leap from below the 207(b) standards to well above them.
-------
- 30 -
Another vehicle might show stable, passing scores on all but
the beginning of the cold start sequence. In clear cut cases
such as these, the test facility simply flags the vehicle or
not, as appropriate.
However, the evidence of variability will not always be
clear cut for every vehicle, and determining whether or not: a
vehicle is flagged will then depend upon the definition of
variability that is, what scores are compared, and how
stringent is the comparison. Obviously, a facility may choose
in the CTP to flag every vehicle that it believes is
borderline. However, an alternative approach is for each
facility to agree at a minimum to apply specific numerical
criteria in any case where the facility has doubts about
whether or not to flag a vehicle as variable.
The numerical standard in the CTP employs the HC and CO
emission scores during the core sampling periods of the various
I/M sequences of the Basic I/M Test Procedure. The specific
values used are the eighteen HC or CO scores that are taken at
fixed times during the core sampling: two HC scores and two CO
scores in the first idle-neutral (seconds IS and 30), two HC
and two CO scores in the 2500 rpm-neutral (seconds 15 and 30.),
and five HC and five CO scores in the second idle-neutral
(seconds 15, 30, 60, 90, and 120). The specific criteria based
on these "fixed time" values are laid out below. ' The
"stabilized" emissions values in a mode are not employed
because they are somewhat more arbitrary; however, facilities
that wish to incorporate the stabilized readings into the
criteria are encouraged to do so.
The determination is then made as follows:
Determining Variability Between Sequences; Variability is
determined by comparing the results of the extended loaded
sequence in turn to each of the other three sequences
(cold start, extended idle, restart) from the Basic I/M
Test Procedure. Values taken when the engine coolant
temperature was below the specified opening temperature of
the thermostat are discarded. Each remaining "fixed-time"
core sampling- value from the sequence in question is then
compared to the corresponding value from the extended
loaded sequence. For example, the HC score taken fifteen
seconds into the first idle-neutral of the restart
sequence is compared to the HC score taken fifteen seconds
into the first idle-neutral of the extended loaded
sequence. Nine such comparisons take place for each core
sampling period contained in the sequence under analysis.
If any value exceeds its counterpart in the extended
loaded sequence by a factor of two or more, and the value
fails the 207(b) cutpoints (1.2% CO; 220 ppm HC), the
vehicle is flagged as variable.
-------
- 31 -
Determining Variability Within a Mode; Variability is
determined by separately examining the results for each of
the three core sampling modes (first idle-neutral, 2500
rpm-neutral, second idle-neutral) in the extended loaded
sequence. If any HC or CO score exceeds any other score
in the same mode by a factor of two- or more, and at least
one of those tvo scores exceeds the 207(b) outpoints, then
the vehicle is flagged for variability.
The data in Tables 5 through 7 illustrate the
determination of variability on three different vehicles, using
the numerical standards described above. The sequence names,
mode numbers, and mode names in the figure are all taken
directly from the Basic I/N Test Procedure, Table 4. Emission
scores are provided for HC (ppm) and CO (%) at the indicated
time (seconds) into each mode of core sampling.
In the first example. Table 5, a number of readings in the
cold start sequence are at least twice their counterparts in
the extended loaded sequence, and they also fail the 207(b)
outpoints. For example, the fifteen-second readings of mode
five in the cold start (underlined in the figure) are 230 ppm
HC and 1.7% CO; the corresponding values in the extended loaded
sequence are 72 ppm HC and 0.51 % CO. Note, however, that all
of the failing readings in the cold start sequence occur no
later than mode 10, and that the coolant temperature does not
exceed the specified temperature of 19S°F until mode 12. This
vehicle is therefore not variable between sequences. In
addition, there is no mode in the extended loaded sequence that
meets the criteria for flagging variability within a mode.
(Data from the remaining sequences have been deleted for
simplicity).
In the second example. Table 6, the engine coolant
temperature is assumed to be normal throughout. There are
again readings in a sequence that are at least twice their
counterparts in the extended loaded sequence, and they also
fail the 207(b) outpoints. For example, the fifteen-second
readings in mode 2 of the extended idle sequence and mode 2 of
the extended loaded sequence are 230 ppm HC. 1.3% CO, and 82
ppm HC, 0.34 % CO, respectively. A similar example for the
second idle-neutral of core sampling is also underlined in the
figure. This vehicle is flagged for variability due to the
different responses in the two sequences. Note that in this
example, the sequence in question (extended idle) actually has
two core sampling periods, but only one is variable with
respect to the extended loaded sequence.
The third example. Table 7, illustrates variability within
a mode. The coolant temperature is once again assumed to be
-------
- 32 -
Table 5: Emission Scores for a Vehicle Hot
Flaqqed for I/M Variability
vehicle *is not flagged as variable, because all of the scores thai
variability criteria in the cold start sequence were during perio
rmal coolant temperature.
Mode
Sequence No . Name
Cold St 3 IN-1
4 2500
5 IN-2
7 IN-1
9 2500
10 IN-2
12 IN-1
14 2500
IS IN-2
Ex Load 2 IN-1
4 2500
5 IN-2
HC/CO
232
1.3
250
1.3
230
1.7
230
1.2
118
1.1
230
1.2
88
0.9
33
0.4
100
0.9
82
0.34
45
0.06
72
0.51
HC/CO
(30)
285
1.7
259
1.4
230
1.4
230
1.2
110
0.9
230
1.2
87
0.9
36
0.5
105
0.9
74
0.33
43
0.09
65
0.47
HC/CO
(60)
N/A
N/A
226
1.4
N/A
N/A
223
1.1
N/A
N/A
110
0.9
N/A
N/A
45
0.82
HC/CO
(90)
N/A
N/A
233
1.5
N/A
N/A
225
1.2
N/A
N/A
110
1.0
N/A
N/A
SO
0.73
HC/CO C'lant
(120) (°F)»
N/A 078-
N/A 102*
218 108"
1.4
N/A 155"
N/A 157'
225 165»"
1.2
N/A 195"
N/A 198»
115 199"
1.0
N/A 199"
N/A 199*
99 199'
0.80
N/A a not applicable to this mode
* Thermostat rated at 195'F
-------
- 33 -
Table 6: Emission Scores for a Vehicle Flagged Due
to Variability Between Sequences
The core tabling values in Modes 1, 9 and 10 of the extended idle sequence
are passing* ' Nevertheless, this vehicle is flagged as variable because the
idle-neutral nodes (Modes 2 and S) of the extended idle sequence are variable
compared to the extended loaded sequence values.
Mode HC/CO HC/CO HC/CO EC/CO
Sequence Wo. Name (15) (30) (60) (901
Ex Load 2 IM-1 82 74 N/A N/A N/A 195'
0.34
2500 45 43 N/A N/A N/A 195*
0.06
IN-2 72 65 45 50 99 195'
0.51 0.47 0.82 0.73 0.80
Ex Idle 2 IH-1 230 220 N/A H/A N/A 195*
1.3 1.3
4 2500 48 43 N/A N/A N/A 195*
0.16 0.19
5 IN-2 214 210 208 267 280 195*
1.9 1.9 2.0 2.0 1.9
7 IN-1 38 87 N/A N/A N/A 195*
0.9 0.9
9 2500 33 36 N/A N/A N/A 195*
0.4 0.5
10 IN-2 100 105 110 110 115 195*
0.9 0.9 0.9 1.0 1.0
N/A * not applicable to this mode
* Thermostat rated at 195*
-------
- 34 -
Table 7: Emission Scores for a Vehicle Flagged Due
to Within-Mode Variability
Vehicle is flagged for variability because the second idle-neutral mode shows
variability batmen the fixed time readings after 60 seconds and the readings
before 60 second*.
Sequence
Ex Load
Mode
No.
2
4
5
Name
IH-1
2500
IN-2
HC/CO
(is)
82
0.34
45
0.06
72
0.51
HC/CO
(30)
74
0.33
43
0.09
65
0.47
HC/CO
(60)
N/A
N/A
15
0.82
HC/CO
(90)
N/A
N/A
150
1.73
HC/CO
(120)
N/A
N/A
199
1.80
C'lant
<°F)*
195»
195*
195*
N/A a not applicable to this mode
Thermostat rated at 195"F
Vehicle is flagged for variability because the second idle-neutral mod* shows
variability between the fixed time readings after 60 seconds and the readings
before 60 seconds.
-------
- 35 -
normal throughout/ and the data from sequences other than the
extended loaded have been eliminated for simplicity. Note, for
example, that within the second idle-neutral of the core
sampling, the CO values at 30 seconds and 90 seconds (0.47% and
1.73%, respectively) differ by more than a factor of two, and
the latter value exceeds the 207(b) CO outpoint of 1.2%.
4.6 Abbreviated I/M Test Procedure
The Basic I/M Test Procedure could be employed after each
FTP and/or repair step; however, it is time consuming, and not
every sequence will yield valuable information for every car.
The results of the Basic I/M Test Procedure are therefore used
to design a shortened version of the procedure called an
Abbreviated I/M Test Procedure. Such an abbreviated procedure
may be used, for example, as a replacement for the Basic I/M
Test Procedure for the as-received short testing on Indolene,
which follows the FTP in the As-Received Characterization.
The Abbreviated I/M Test Procedure for a given vehicle
consists of the extended loaded sequence, plus any other
sequence which showed variable test results when compared to
the extended loaded sequence. This rule may be applied to the
examples of the previous section (assuming that all data
omitted from the examples did not show variability). Because
the first example showed no variability, the Abbreviated I/M
Test Procedure for that vehicle would consist solely of the
extended loaded sequence. The Abbreviated I/M Test Procedure
for the vehicle in the second example would be the extended
loaded sequence, plus the extended idle sequence. In the final
example, the only extra sequence necessary would again be the
extended idle sequence.
The process of determining the sequences that will make up
an Abbreviated I/M Procedure should be distinguished from
deciding when the procedure is actually to be employed. As
will be seen in the description of the Remedial Maintenance
Phase (Section S), the performance of some CTP vehicles during
the As-Received Characterization may obviate the need to
perform an Abbreviated I/M Procedure later in the program.
Like the Basic I/M Test Procedure, the Abbreviated I/M
Procedure is performed with the hood in the raised position,
and with external cooling fans applied only during modes with
loaded operation, such as the LA4. A facility may again choose
to monitor extra parameters in the Basic I/M Test Procedure,
but only if it is clear that the process of taking the
measurements will have no impact on the emission scores.
-------
- 36 -
4.7 Engine/Emissions Systems Diagnosis
The final step in the As-Received Characterization is a
complete diagnosis of tha test vehicle's engine and emission
control systems, with tha primary purpose of identifying any
repairs thai: may be necessary to remedy FTP and I/M emissions
malperformance in the vehicle. Repairs indicated by the
diagnosis, as well as after-repair testing, are conducted
during the Remedial Maintenance Phase, described in Section 5.
In order that the data from the Cooperative Test Program
and previous studies may be easily compared, EPA recommends
that the test sites tailor their diagnostic sequences to the
standard EPA Emission Factor format (the basis for the ECOMP
datafile on the Michigan Terminal System). Mechanic's data
sheets and a sample ECOMP dictionary have previously been
provided to each of the participating organizations; a
dictionary that will be specific to the CTP is under
development by EPA. The ECOMP system provides basic vehicle
and test identifying information, together with coded data on a
comprehensive inspection of the engine and emission control
system.
While EPA would prefer that the manufacturers provide data
in the ECOMF format, some may have little experience with the
coding system employed there. EPA will therefore consider
coding such data on a limited basis, if the necessary raw data
is made available by the manufacturer in the narrative comments
section.
The intent in applying the ECOMP format is not to
prescribe an order of diagnosis or to limit the scope to
particular steps. Rather, each manufacturer should take
whatever steps it feels are necessary to completely diagnose
the problems on a given vehicle. Steps as diverse as scoping
the engine, functional tests of a specific component, probing
of the vehicle's onboard diagnostic system, or monitoring
mixture dwell may be indicated.
EPA recognizes that it may be difficult to diagnose all
necessary repairs during the As-Received Characterization,
because completing one repair may be necessary for a second
malfunction to become evident. In ;uch circumstances, new
diagnostic information may subsequently lead to changes in the
order of remedial maintenance repairs. All malfunctions
discovered during the course of repair should be recorded and
treated as if discovered during As-Received Characterization.
In some cases, the test facility may diagnose no engine or
emission control malfunctions on a given vehicle. The
discussion in Section 5 of the Remedial Maintenance Phase will
-------
- 37 -
treat this circumstance in some detail. It is worth noting,
however, that the diagnosis step in the As-Received
Characterization is an appropriate point to ask the question of
whether vehicle design, rather than vehicle malfunction, plays
a determining role in any I/M failures observed in the test
vehicle. Information gathered on the vehicle's calibration or
from review of the engine parameter data gathered during the
Basic I/M Test Procedure may provide the basis for later
decisions during remedial maintenance.
The diagnostic process described above completes the
As-Received Characterization. The vehicle then proceeds to the
Remedial Maintenance (RM) Phase for repairs, additional
investigation of short test behavior, and further emissions
testing.
-------
- 38 -
Section 5: Remedial Maintenance
5.1 Introduction
In the Remedial Maintenance (RM) Phase, the results of the
As-Received Characterization for each test vehicle are compared
to a set of criteria for FTP and I/M performance. If repairs
are indicated for a vehicle, a repair sequence is designed and
executed, with intermediate testing to determine the impacts of
the repairs on FTP emissions, I/M emissions, and any previously
observed I/M test variability. As will be seen below, cases
may also arise where no repairs are performed, and the test
facility needs only to explain certain aspects of a vehicle's
emissions response before exiting the RM Phase.
5.2 Oblective
The objective of the Remedial Maintenance Phase is to meet
the following three criteria on each vehicle in the Cooperative
Test Program:
1. FTP Criterion;
EITHER the after-repair HC and CO FTP results on the
vehicle are less than the following targets:
vehicle
Mileage Target
<.50,000 150% of vehicle's cert standards
>50,000 200% of vehicle's cert standards
OR all reasonable diagnosis and repair efforts have
been completed.
2. Variability Criterion;
EITHER the vehicle will reliably satisfy the
variability criteria of Section 4.5 on any I/M
sequences that, prior to repair, had caused the
vehicle to be flagged for I/M variability,
OR any observed variability of the vehicle's short
test emissions has been traced to a specific,
unrepairable aspect of the vehicle's design,
OR efforts to repeatably instigate an I/M test
failure in the vehicle have been unsuccessful.
-------
- 39 -
OR all reasonable diagnosis and repair efforts have
been completed.
3. I/M Criterion;
EITHER the vehicle's core sampling emissions during
the extended loaded sequence meet the 207(b)
outpoints (220 ppm HC, 1.2% CO),
OR observed anomalies of the core sampling emissions
during the extended loaded sequence have been traced
to a specific, unrepairable aspect of the vehicle's
design.
OR all reasonable diagnosis and repair efforts have
been completed.
The approach to RH in the Cooperative Test Program is to
consider these criteria, in sequence. Figure 3 provides a
generalized flow diagram of this process. After entering the
KM phase in block 1, the FTP criterion is addressed in blocks
2-17. Consideration of I/M variability occupies blocks
19-29. Finally, remaining I/M nonconformity is considered in
blocks 30 - 38. The subsections below elaborate on the RM
Phase as laid out in Figure 3. Numbers in parentheses in the
text refer to the block numbers in the figure.
5.3 Remedial Maintenance Based on the FTP Criterion
Consideration of the FTP criterion begins with a
comparison of the most recent FTP results on the vehicle to the
mileage-based targets given in Section 5.1 above (2). For a
vehicle just entering the phase, the appropriate test is the
as-received FTP on Indolene. Note that vehicles pass through
the same decision point following after-repair testing (15), a
step that will be discussed below.
If a vehicle fails to meet the FTP targets, the list of
needed repairs assembled during the As Received
Characterization is updated as necessary to include any newly
discovered problems (3). If no repairs remain to be performed
(4), -yet the FTP targets have not been satisfied, the facility
examines the available data to determine if the FTP emissions
problem can at least be isolated to either the catalyst or the
engine as a whole (16). Such an analysis might involve
comparison of feed gas levels to tailpipe levels, visual
inspection of the catalyst for. catastrophic failure, and review
of available Plumbtesmo and tank fuel data. , If the reponse at
block 16 of Figure 3 is "no," the vehicle is flagged for later
-------
Figure 3: GENERALIZED FJ.OH DjAGJi^ Fpft TH.E REMEDIAL MAINTEMAMCF PHASF
Enter from As-rec'vd 1
character
zation
2. Ust FTP HC, CO
X FTP targets?
3. Propose Any changes
to list of repairs
4. Have all repairs
been completed?
S. Can FTP benefits of
repairs be qualita-
tively projected?
6. Execute repair of
choice fro* list
7. Conduct U4 test
8. (Optional) Conduct
Abbreviated I/N
test procedure
IA4 isiproveEient
signi f icant?
10. Are projected FTP
(pacts of repairs
neolioible?
16. Can FTP failure be
traced to either the
catalyst or the enoine?
II. Execute repair with
highest projected
FTP benefi
17. Flag vehicle for
future catalyst
testing
IB. Variabilil
lity and I/H
analyses complete?
12. LA4 option appro-
priate?
13. Conduct standard
FTP prep
14. Conduct FTP test
iS. Conduct Abbreviated
I/H Test Procedure
19. Is vehicle flagged
for I/H vartabTT-
itv?
20. Can the
bility be tr
reliably?
/H varia-
ggered
21. Suourize available
data on previous vari-
able I/M test results
-------
122. Is variability ex-
plained by unrepairable
elMent of desion?
f
23. Provide explanation
24. Evaluate engine
faaily for pattern
failure behavior
n'
n
b
25. Propose any changes
to lilt of repairs
26. Have" all repairs 1
been completed?
n
27. Execute repair of
choice froa list
28. Conduct Abbreviated
I/H Test Procedure
29. Variability
resolved?
30. Any failing reading
regaining in extended
loaded cnr> t
31. I/H failure ex-
plained by unrepairable
if dfltion?
32. Propose any changes
to list of repairs
33. Have all repairs
been completed?
34. Conduct repair of
choice
35. Conduct Abbreviated
I/H Test Procedure
36. Evaluate vehicle
for new evidence of
variability
37. Provide explanation
38. Evaluate engine
faiiily for pattern
failure behavior
39. Any repairs with
possible large FTP i»-
oacts since last FTP?
40. Vehicle flagged for
catalyst testing
4). Replace catalyst as
indicated
42. Retain old parts
for later EPA and aanu-
facturer testing
I 43. Go to
Exit Phase
-------
- 41 -
analysis of its catalyst (17), and the FTP criterion is
satisfied. If the response at block 16 is "yes," however, the
vehicle is not flagged for later catalyst bench testing; the
available data implicating either the catalyst or engine are
summarized, and the FTP criterion is satisfied.
If repairs remain to be performed at block 4, the facility
must decide on the order of repair. First, an attempt is made
to qualitatively project the relative FTP impacts of each
repair (5). If the testing organization is not confident of
its ability to predict which needed repairs will have large
effects and which only small effects, the facility simply
selects a repair and proceeds (6). If the impacts of all the
remaining repairs are judged to be negligible (10), the vehicle
is flagged for future catalyst testing, and again the FTP
criterion is satisfied. If only one repair remains which is
expected to have a large impact on FTP emissions, that repair
is performed.
The remaining case is where more than one repair remains
with significant projected FTP impacts. In general, the test
facility will then choose the single repair that is predicted
to have the greatest potential for meeting the FTP criterion
(11). For example, if the vehicle has failing CO emissions on
the as-received FTP, the repair selected should be the one that
the facility predicts will have the greatest impact oh the
vehicle's FTP CO emissions.
In general, repairs that would be anticipated to have
large impacts would include (but not necessarily be limited to)
the following:
1. Computer control and feedback system repairs,
including most repairs indicated by
emissions-related fault codes from onboard
diagnostic systems, and repairs to electronic fuel
metering components;
2. Repairs to primary emission controls other than
those in the feedback system, including
malfunctioning components in the exhaust
aftertreatment, secondary air, FCV, and EGR systems;
3. Adjustment of the idle mixture on vehicles with
missing limiter devices.
Malfunctions in other, more basic engine components (for
example, a cracked and misfiring spark pi'-ig) may also yield
substantial FTP benefits once repaired.
-------
- 42 -
5.4 After Repair Testing Based on the FTP Criterion
The information from the Remedial Maintenance Phase will
be most valuable if the repairs that cause large changes in FTP
and idle emissions can be specifically identified and
quantified as to their FTP and I/M impacts. For this reason,
the Cooperative Test Program calls for both CVS and I/M testing
following each significant repair, rather than solely after all
repairs to a vehicle have been completed. Two options for the
CVS testing are provided in the CTP; one is based on an LA4,
and the other, on an FTP.
The "LA4 option" is employed if a facility wishes to test
for substantial benefits from a repair without the time
required for a standard overnight FTP prep. The vehicle first
undergoes a 505-second prep (or a manufacturer-determined
equivalent), followed by an LA4 (7). I/M testing, which is
optional at this point, consists of the Abbreviated I/M Test
Procedure (8) that was designed specifically for the vehicle in
question during the As-Received Characterization. The results
of the LA4 are compared to the bag 2 and bag 3 results from the
last FTP on the vehicle. If the improvement is significant,
the vehicle is propped (13) and then receives a regular FTP
(14). If the FTP improvement is minimal, the vehicle returns
for consideration of further repair (3). The decision of
whether the use of the LA4 option is appropriate or not (12) is
left to the discretion of the test facility, based on the
earlier bag results on the vehicle and the likely impacts of
the repair being evaluated.
If the test facility determines that the LA4 option is not
appropriate or choses to proceed directly with an FTP-based
evaluation of the repair, after-repair testing consists simply
of the standard FTP prep (13), FTP (14), and the Abbreviated
I/M Test Procedure for that test vehicle (IS). As was the case
in the As-Received Characterization, the vehicles are tested
with the hood raised, and external cooling fans applied only
during modes with loaded operation.
Upon completion of the after-repair testing, the results
are once again evaluated against the FTP targets (2) to
determine if the* FTP criterion has been satisfied. If not, the
process described above is repeated, beginningwith the
reassessment of the list of repairs (3). If the FTP targets
are met, attention turns to the I/M response of the vehicle
(18).
5.5 Remedial Maintenance Based on I/M Variability
Recall that in Section 4.5, a process was described for
flagging vehicles with variable I/M results, either between
-------
- 43 -
different I/M sequences, or at different times in the same
snort test mode. Vehicles with that flag "raised" are
considered* in blocks 19-29 of Figure 3.
Resolving a problem with I/M variability requires that the
test facility be able to reliably trigger the variable I/M
results in the vehicle (20). If the facility is unsuccessful
in this effort (e.g., tvo consecutive Abbreviated I/M Test
Procedures give conflicting results on whether prolonged idling
causes an I/H failure), the known variable results on the
vehicle are summarized (21), and the variability criterion is
satisfied.
On vehicles where the variability can be reliably
triggered at the test facility, the next question is whether
there exists an aspect of design in the vehicle that explains
the variable behavior. If so, the facility evaluates the
vehicle to determine if there is a pattern of I/M failure that
will likely arise in other similar vehicles (24). For example,
timer-based secondary air routing may be the cause of variable
I/M readings. Another example would be a vehicle that shows
clear evidence of gradual catalyst or 0( sensor cooldown
during prolonged idling.
In some cases, design elements may not explain the
variability. An example might be a vehicle that will only fail
the 207(b) outpoints if it operates open loop and diverts
secondary air to the atmosphere, whose air routing valve is
stuck diverting to atmosphere; if the vehicle operates
open-loop following a restart, but closed-loop following
extended loaded operation, it may show variable results during
the Basic I/M Test Procedure.
In such a case, a repair sequence is conducted. As in
Section* 5.3 above, the list of repairs prepared during the
As-Received Characterization is updated based on any new
information from earlier remedial maintenance steps (25). If
all reasonable repairs have been completed, the variability
criterion is satisfied (26). Otherwise, the test facility
chooses a repair from those remaining to be performed (27).
The order-of-repair issue is not as important here as it was
during.consideration of the FTP criterion, because the costs of
verifying the impacts of repair are not as high. Nevertheless,
the facility should logically order the repairs according to
their anticipated ability to resolve the observed variability.
5.6 After Repair Testing Based on the Variability Criterion
The after-repair test for measuring success in meeting the
variability criterion is the Abbreviated I/M Test Procedure
designed previously in the As-Received Characterization (28).
-------
- 44 -
Attention must once again be paid to the proper use of external
cooling fans.
If necessary, the procedure may be modified according to
the information gathered in earlier stages of the Remedial
Maintenance Phase. For example, you may have variability
during three sequences during the As-Received Characterization,
but following the first repair of RM, you may only show
variability between one sequence and the extended loaded
sequence. Therefore, the third (no longer variable) sequence
need no longer be performed in the Abbreviated I/M Test
Procedure. The results of the test are used to determine if
the I/M variability is resolved (29), and if not, if further
repair steps are necessary to meet the variability criterion.
5.7 Remedial Maintenance Based on the Basic I/M Criterion
Once the FTP and I/M variability criteria have been
satisfied, what remains is to resolve any remaining I/M test
anomaly when the vehicle is tested under "ideal" test
conditions, i.e., during the core sampling period in the
extended loaded I/M sequence. Once again, the criterion can be
satisfied if the failing I/M scores are explained by an
unrepairable element of the vehicle's design (31).... For
example, secondary air may be dumped during all unloaded
2500 rpra operation. In such a case, an explanation of the
behavior is provided by the test facility (37), and the engine
family to which the vehicle belongs is evaluated for pattern
failure behavior (38).
Where a design explanation is not available, the sequence
of remedial maintenance steps follows the same path as in
Section 5.5 above. The available list of repairs is updated
(32), and if no repairs are available, the I/M criterion is
satisfied (33). Otherwise, the facility chooses and executes a
repair (34).
Cases may arise where catalyst replacement is diagnosed as
necessary to pass, the Michigan I/M test. If no evidence of
catalyst tampering or mis fuel ing has been uncovered, the
catalyst: should be replaced in these circumstances. The
question of retaining the catalyst for bench testing by EPA
should have been resolved earlier, when the FTP criterion was
being considered.
5.8 After-Repair Testing Based on the I/M Criterion
The after-repair test used to verify compliance with the
I/M criterion is again the Abbreviated I/M Test Procedure
-------
- 45 -
appropriate for the given vehicle (35). The results of this
test are evaluated for any new evidence of I/M variability (36
and 19)» and the vehicle is once again checked for failing
scores in the core sampling period of the extended loaded
sequence. The results of this check determine if the vehicle
satisfies the I/M criterion, or if further remedial maintenance
steps must be considered.
5.9 Other FTP Testing in the Remedial Maintenance Phase
As discussed above, and illustrated in Figure 3, the CTF
Remedial Maintenance Phase accommodates vehicles where repairs
are unlikely to have a substantial impact on FTP emissions.
Nevertheless, cases may arise in the CTP where a vehicle nears
the end of the Remedial Maintenance Phase having had repairs
with possible large FTP impacts, but not actually having an FTP
test conducted to measure those impacts. One such case may
arise through use of the LA4 option when the list of possible
repairs is "running out" (e.g., blocks 7, 8, 9, 3, 4, and 16
in sequence). In another example, a test facility may diagnose
the need for a repair with possible significant FTP. impact
after consideration of the FTP criterion has been completed.
Finally, the cumulative FTP effect of a number of small repairs
conducted at different points in the RM Phase might be
significant.
In cases like these in the CTP, an FTP should be performed
before the phase is completed (39). The results of this test
may point to the need for additional repair steps (2). Further
consideration of the vehicle's I/M behavior is unnecessary,
unless new evidence of variability or I/M failure has emerged
(18). The testing organization should use its best judgment
when deciding to forego an FTP. In spite of these cautions,
EPA anticipates that vehicles with more than one or two
significant repair steps will be rare.
5.10 Catalyst Replacement in the Event of Persistent FTP Failure
All restorative maintenance efforts may be insufficient to
meet tha FTP targets on some CTP vehicles, and the manufacturer
may have no explanation for the anomaly. If the test facility
was unable to determine whether it was the catalyst or the
engine that was leading to the FTP failure, such vehicles were
"flagged" during consideration of the FTP criterion for later
bench testing of the catalyst (17). The catalyst is not
actually replaced until after the three RM criteria have, been
satisfied (41), unless the replacement was necessary to
guarantee that the vehicle would pass the Michigan AET retest.
The removed catalyst is retained for bench emission
testing by EPA, and (if necessary) destructive analysis. The
-------
- 46 -
purposes of these efforts are evaluation of the condition of
the catalyst, and determination of the reasons for any
deficiency* A manufacturer may perform agreed-upon catalyst
tests in-house prior to forwarding the parts to EPA for the
above analysis.
On vehicles that require catalyst evaluation, parts costs
will be borne by the original testing organization (EPA or
manufacturer); EPA may require assistance in obtaining
replacement components quickly for vehicles tested at MVEL.
Handling of vehicles that are flagged for catalyst testing
is the final step in the Remedial Maintenance Phase. Vehicles
that have completed the HM Phase proceed to the Exit Phase.
-------
- 47 -
Section 6: Exit Tasks and Owner Compensation
Exit tasks include obtaining an official Michigan AET
reinspection of the vehicle, conducting a follovup visual
inspection of the vehicle with the owner, and owner
c ompensat ion.
In order for the owner to register the vehicle following
participation in the Coopurative Test Program, he/she must be
able to present either a Certificate of Inspection or a
Certificate of Waiver when applying for license plates.
Manufacturers that have met the requirements for conducting
official AET tests may, of course, retest the vehicles
themselves and issue certificates. Alternatively, the
manufacturers may retest the vehicles at one of their
dealerships that is an authorized AET test facility. Waiver
certificates may only be issued to owners who have met certain
cost-of-repair criteria, and this option will probably be
unavailable for participants in this program.
The visual reinspection of the vehicle is performed with
the owner present to verify the condition of the vehicle
relative to the intake inspection, and to answer any owner
questions about the repairs or tests undergone by the vehicle.
Owners who participate in the Cooperative Test Program
will be offered the following incentives to participate:
o a fully-insured, late-model loaner vehicle with a
full tank of fuel;
o a check for $50.00, provided during the Exit Phase;
o all repairs during the CTP to be conducted free of
charge;
o owner's vehicle to be returned cleaned, and with a
full tank of fuel;
o owner to receive a passing Michigan AET inspection
certificate, unless the cause of the original test
failure is determined in the course of CTP testing
to be tampering.
Each test facility may choose to establish additional
incentives to accomodate exceptional cases, such as vehicles
that are rejected during the Intake Phase do to potential
safety problems on the dynamometer, or vehicles that must be
kept longer than anticipated. As mentioned in Section 3.5, the
test facility is responsible for obtaining a temporary
registration for any vehicle that is kept in the CTP to the
-------
- 48 -
point where the owner has less than one week before the
Michigan registration deadline in which to register the vehicle.
Note that the above discussion eliminates a number of the
incentives in the original CTP proposal. Owner response rates
and attitudes will be monitored by the calltaJcers so that the
incentive structure can be reevaluated, if necessary.
-------
- 49 -
Section 7: Documentation and Reporting
7.1 Introduction
The following subsections discuss the record keeping
procedures for CTP data to be followed by each participating
organization. Some, but not all, CTP data will be reported to
EPA. The Agency will need certain information on an ongoing
basis in order to monitor progress towards filling test quotas,
to adjust the size of recruitment mailings, and to identify
significant problems with the recruitment approach.
The bulk of the data reported to EPA will be the actual
emissions and repair data on each test vehicle, data that will
be incorporated into MICRO files on the Michigan Terminal
System (NTS) with open access to all participants. The data on
each vehicle that completes at least the As Received
Characterization will be accumulated in a packet ("CTP Vehicle
Data Packet") for submission to EPA as soon as possible after
each vehicle has completed the Exit Phase. EPA will supply
sample data forms for all data to be included in the Vehicle
Data Packet.
7.2 Prescreeninq Data
Data gathered during the Prescreening Phase will include
information on the types of incoming calls and responses to the
prescreening questionnaire. EPA will provide the master for a
form on which to record the disposition of each incoming call
(including the cause for prescreening rejection), and a
separate form for summarizing statistics on the calls. In some
cases, vehicles that are rejecting during prescreening may
already have completed part of the prescreening questionnaire.
The test facility may wish to retain these questionnaires;
however, no submission of questionnaire data to EPA is
necessary in such cases. The questionnaires for vehicles that
successfully complete the Prescreening Phase will normally
include some incomplete responses or information that need to
be verified in person. These questionnaires are completed
during Intake, and later provided to EPA.
7.3 Intake Data
As in the case of the Prescreening Phase, information on
Intake Phase rejections will be provided to EPA in order to
adjust the size of future solicitation mailings, and to monitor
the need for modifications in the solicitation letter or the
solicitation method.
-------
- 50 -
Copies of the completed prescreening questionnaire on each
vehicle that completes the Intake Phase are added to the CTP
Vehicle Data Packet. Also included should be a copy of the
original Michigan AET inspection for accepted vehicles,
including verification of the analyzer manufacturer from that
test.
7.4 Data from the As-Received Characterization
In addition to basic vehicle and test identifying
information, the following data from the As-Received
Characterization are included in the Vehicle Data Packet:
o Basic I/M Test Procedure; emission scores and engine
rpm as a function of time, for each mode; coolant
temperature for the indicated modes; (optional)
narrative description of the behavior of additional
monitored engine and emission control parameters.
o Tank-fuel Analysis; results of the lead-in-fuel and
RVP determination on the tank fuel in the
as-received vehicle, or the RVP determination on the
facility's commercial fuel.
o As-Received FTP; composite and bag results for HC,
CO and NO.; travel distances by bag.
o Variability Status; flagged for variability between
sequences; flagged for variability within a mode;
not flagged for short test variability.
o Abbreviated I/M Test Procedure; emission scores and
engine rpm as a function of time, for each mode;
coolant temperature for the indicated modes;
(optional) narrative description of the behavior of
additional monitored engine and emission control
parameters.
o Diagnosis; results of the complete as-received
engine and emissions system diagnosis, in an
ECGMP-codable format; addtional narrative diagnostic
information, as appropriate.
7.5 Data from the Remedial Maintenance Phase
Because both FTP and I/M test procedures may be conducted
more than once during the Remedial Maintenance Phase, the data
are identified according to the repair "condition" of the
vehicle: as-received (RECV); after the first RM repair (REP1);
-------
- 51 -
after the "nth" repair (REPn). In addition to basic test and
vehicle identifying information, the RM data included in the
Vehicle Data Packet are the following:
o Repair data. reported by repair condition;
classification of system and repair type;
supplemental narrative comment; where appropriate.
o Condition of the vehicle at completion of the FTP
criterion; RECV, REP1, etc.
o Method of satisfying the FTP criterion; FTP targets
satisfied; FTP targets not satisfied, but FTP
problem isolated to either the catalyst or the
engine; FTP targets not satisfied, and vehicle
flagged for catalyst bench testing.
o LA4 test data. reported by repair condition
(applies only if LA4 option exercised): Composite
and bag data for HC, CO, and NO,; travel distances
by bag.
o Narrative comments on satisfying the FTP criterion
(optional)
o Condition of the vehicle at completion of the
variability criterion; RECV, REP1, etc.
o Method of satisfying the variability criterion;
variability not reliably triggered; I/M variability
explained by unrepairable element of design;
variability resolved through repair; variability hot
resolved after completion of all reasonably relavant
repairs.
o Narrative comment on satisfying the variability
criterion; comment on "non-repeatable" I/M
variability; explanation of design-triggered
variability
o Condition of the vehicle at completion of the I/M
criterion; RECV, REP1, etc.
or~ Method of satisfying1 the I/M criterion; I/M failure
explained by unrepairable element of design; I/M
failure resolved through repair; I/M failure not
resolved after completion of all reasonably relavant
repairs.
o Narrative comment on satisfying the I/M criterion:
explanation of design-triggered variability ~~
-------
- 52 -
o FTP test data, reported by repair condition:
Composite and bag data for HC, CO, and NO.; travel
distances by bag.
o Abbreviated I/M Test Procedure data, reported by
repair condition; emission scores and engine rpra as
a function of time, for each mode; coolant
temperature for the indicated modes; (optional)
narrative description of the behavior of additional
monitored engine and emission control parameters.
7.6 Exit Phase Data
A copy of the Michigan AET retest obtained by the CTP test
facility should be included in the Vehicle Test Packet.
-------
A-l
Appendix A
Suspected I/M Pattern Failures in the
Seattle I/M Procraa
-------
A-2
The Appendix A list was obtained by analyzing
approximately 140,000 LDVs and LDTs from the 1981 through 1983
model years tested in the Seattle I/M program from 1982 through
1984. Failure rates were computed for each engine family
represented in the data by applying the 207(b) outpoints (1.2%
CO and 220ppm HC) to the actual emission scores of the
vehicles. Average failure rates at the 207(b) levels were then
computed for all vehicle of like model year and vehicle type
(LDV or LDT).
A chi-squared analysis of the failure rates for the
families and the rates for the fleet was used to rank the
engine families. An engine family was included on the
"suspect" list if there was at least a 95.0% probablility that
the difference between the family's failure rate and the fleet
failure rate was not due to chance. There are 64 engine
families falling into this category in the model years examined.
-------
A - 3
Failure Rates at 207(b) Cutpoints
Seattle Suspected Pattern Failure Analysis
Model
Year
1981
1982
1983
LDV
3.7
2.6
1.4
LPT
14.3
13.2
10.4
Example of Chi-Squared Calculation
X1 - E (0, - Et)'/E,
i
where i - the possible outcomes or test status of
each vehicle; in this case, "pass" or
"fail"
Ot = observed occurances of test status "i"
for an engine family
Et = expected occurances of test status "i"
for an engine family, based on rates for
the fleet of like model year and vehicle
type (LDV. LDT)
Example; engine family = 14E2TM (1981 LDV)
n - 4034
207(b) failure rate
(this family) = 4.3%
207(b) failure rate
(Seattle '81 LDV) =3.7*
I (0, - E,)'/E4
[(4034 X .043) - (4034 X .037)]-*- 1-
4034 X .037
[(4034 x 0.957) - (4034 X .963)!1
4034 X .963
4.076
-------
SUSPECTED i/M PATTERN FAILURES IN THE SEATTLE I/M PROGRAM
RANK VEH.TVPE UVR MFC*
ENGINE tAMlLV(S)
FAILURE RATES (I)
PROGRAM 220/1.2 100/0.6
CHI-SQ
63
49
20
3
41
44
6
4
14
*B
5S»
13
'*!
1 .
1.4
U/
5
IB
12
60
62
4)
9
2»
so
as
10
IS
27
tb
3B
45
30
24
42
60
SI
48
17
id
31
J'J
1
t>a
/
/M
/
u
'. 1
J ».
LDV
LOV
LOW
LDV
LOT
LOV
LOI
LOI
LOI
LOV
LOI
LUI
I HI
1 HI
1 IIV
LUV
1.1)1
LOI
LOI
LOV
LOV
LOT
LOT
LUI
LUV
LOV
tUV
IOV
LOV
LOV
LOV
LOV
LOV
LOV
LUV
LUV
LOV
LUV
LUV
LOV
LUV
LUv
LUV
I IIV
1 IIV
1 (IV
I l«v
1 IIV
1 IIV
1 IIV
81
ai
83
81
82
83
81
82
83
III
82
81
82
ttl
BJ
82
B3
at
82
ai
Bl
63
82
81
ai
83
ai
82
82
ai
a*
83
82
82
83
83
82
83
81
81
81
at
a*
ai
ui
HI
u i
it i
II 1
U 1
040
040
040
040
040
040
590
590
590
690
590
StIO
bou
5l>0
000
560
5liO
500
500
560
sao
380
380
380
3UO
3UO
380
380
380
380
380
030
OJO
O30
030
030
03U
030
030
OJO
030
UJO
030
OJO
OJO
OJO
OJO
UJII
uyo
n/u
I4E2TU
I2S4A8
O2G2.SV&TPC.6
iiiaiMQ*
CIGS.7T4HAC5
OlGt.BV2MEAO
BVW2.0TSAF3
BVW2.0TSAF3
DVMI.BTSCWFo
BVMI.7V6FFS37F
Cvwl .7I6APF6
BVMI.716AA737PF
BIM2.3T2AF3
BIK2.3T2AF3
OTK2.0V2HFF7
CIH2.0V2GLCS
CIK2.0T2AFF8
BIH2.OT2Aea
CIK2.0T2AFF8
alKI .SV2GCI
BNS2.BVSF03
ONS2.4I2AAFM
CNS2.2T2AAFB
BNS2.2I2ABI
BNS2.0v2Altx
ONS2.0V2AAF7
BNSI.SV2AB6
CNSI.SV2ACFX
CNSI .5V2AAF6
BNSI.2V2ABV
CHSI . 2V2AAFX
OI-M3.3vlGxi;X
< >MJ. 3VIGXMJ
CtM2.3v2Ga»6
Ut^UI .6V&HMK3
OFMI .6V2GOH6
C»MI.6V2UHC2
OFMI .6V2GOC7
S.awBPF
3 . 3(iUF
2.3AM
2 . 3AHF
.2/&.OL.CC
. 2/5.0CLC/ALC
.2/S.OGCF
. y/5.0ci:L/u.CF
2/O.OGCF
j OLCf
U( H'o . W4ll(. 1
UU . 2V/IIJ4
I4E2TM
I2S4AB
02G2.SVSTPG6
IIM2TNQZ
CIGS.7T4HHC8
OIGI .6W2NEAS
BVM2.0TSFAB
BVM2.QISFA8
OVMl .UI5CVF6
BVWI . 7V6FLB37C
Cvwl . 7T6APT3
BVHl .7I6FA6.37P
BTK2.3I2AF3
DIK2.QV2HGGX
CIH2.0V2UCCS
OTH2.0T2AHH4
BIK2.0T2AGO
CIK2.0I2AGGO
BIKI .SV2GCI
BNS2.BV5FC4
ONS2.4f9FAC6
CNS2.2T2ABC7
BHS2.2I2AC2
BNS2.0V2ACO
UNS2.0V2AAC4
BNSI .&W2AC7
CNSI.SV2AOC9
CNSI.5V9FAF8
BNSI.2V2AC3
CNSI .2V2ABCU
CFM3 . 3VIGXF9
CFM2.3V2HAF7
M Ml .6VSHMF3
OFMI .6V2GOC7
tl-MI .6V2GHC2
S.BWAXC
3 . 3GUF
2.3AM
4.2/S.OGCC
5.0CCC
4.2/S.OGCC/ACC
4.2/S.OAAC
5.0CCC
ULHh . 2V4IIC I
4034
43
441
205 1
479
133
IBI
141
62
1544
II
167
64
374
II 1
1048
209
777
231
242
7B4
72
798
997
462
168
2061
569
439
108
32
104
831
B82
21
310
1823
SO
81
3192
976
941
350
24
22
14
329
218
21
144
2.7
7.0
2.0
7.3
13.8
3.0
9.4
19.9
16.1
4.5
27.3
10.2
ia. a
13.9
I.B
3.1
26.3
10.2
14.7
3.3
3.8
12.5
6. B
6.6
I.I
i.a
4. 1
1.9
3.0
II. 1
6.3
2.9
3.S
S.3
4.8
1.9
1.4
4.0
17.3
3.7
5.5
4.6
44. a
i 4. a
36.4
28.6
47.1
6.9
0.0
3.6
4.3
11.6
6.1
is. a
18.2
4.6
54. 1
74.5
51.6
I. 1
36.4
44.3
46.9
28.9
3.6
4.5
4B.8
26. a
39.0
7.0
5. 1
20. a
27.8
23.5
5.8
6.0
8.6
9.5
8.4
22.2
12.5
4.8
5.8
7.B
9.S
2.9
3.6
6.0
24.7
6.8
7.1
6.7
66.6
12.6
SO.O
35.7
52.6
19.7
14.3
8.3
13.2
11.6
17.2
27.9
38.2
8.3
68.0
90.1
64.5
10. 0
45.5
65.9
bO.9
41.2
3.6
4.5
59.8
40.3
51.5
10.3
7.0
36. 1
45.0
45.7
10. 8
10.7
18.2
19.5
16.9
44.4
31.3
6. 7
9.6
9.3
52.4
5.8
8.5
16.0
29.6
8.6
9.0
7.2
58. 9
12.5
63.6
35- 7
S3. 8
25.2
19.0
13.2
14.076
7.532
70.572
642.768
10.452
9.259
233.953
462.430
1 12.939
50.093
5. 167
122.643
63.43B
65.052
3.892
14.939
330.725
99.066
134.202
1.396
4.313
8.367
148.462
ba.ase
5.718
VS. 753
I3B.88I
106.974
58.316
103.738
12.385
B. 709
33.602
64.657
9.981
5.053
7. 199
7 .664
100. 2S3
86.091
31 .665
10.564
4030. 168
S.2I6
195. IBS
40. 735
2207 .932
156.628
6.622
8.552
-------
FAILURE RATES (»)
AN* VEH.TVPE
33
55
1 1
54
23
26
57
J4
411
J..'
>...
l.l
., 1
LOT
LUV
LUV
LOV
LOV
LOV
LOV
LOV
LOV
LOV
LOT
till/
1 LlV
LUV
MVK
81
83
82
82
83
82
81
81
83
81
83
82
B2
Bl
MFC*
020
999
999
aea
999
aaa
399
aaa
260
999
999
»dtt
J70
auu
ENGINE FAMILV(S)
BCR3.7TIAAI
OMT2.6V2BF09
CMT2.6V2BF08
CMTI.6V2BFOO
CPE2.0V6FAB3
CPE2.0V6FAB3
I8K
32L
OtlNI .SV3ACK6
BFI2.0V5FAI
OFJI.8T2AF02
LAOI.7V6FBF7
CTVI.3V2AFF
BI-HlB3vbFC3
BCR3
DMT 2
CMT2
CMTI
CPE2
XN6
I8K
32L
OMNI
BFT2
OFJI
CAOI
CIVI
. 7T1BC5
.6V2BCAO
. 6V2BCAX
.6V2BCA2
.OV6FAA2
.5V3AUC5
.OV5FAI
.8T5FFHO
. 7V6FBC4
. 3V2ACC
OCR IB3VbFLJ
N
137
29
107
174
58
48
487
39
B27
133
288
200
150
26
PROGRAM
14.6
3.4
8.4
1.7
10.3
12. S
2.7
10.3
1 . 1
4.5
3.8
3.5
2.0
7. 7
220/1 .2
30.7
6.9
20.6
5.7
13.8
20.8
5.7
15.4
3.6
9.0
20.5
5.5
6.7
1 1 .5
100/0.5
51.8
10.3
27. 1
8.0
17.2
25.0
9.4
17.9
6.5
18.0
44.4
6.0
11.3
15.4
CHI-SQ
30.067
6.355
136.898
6.603
64.605
62.784
5.467
I4.9B3
28.997
I0.4BS
31 .528
6 642
9.957
4.439
U1
-------
B-l
Appendix B
Draft Owner Solicitation Letter
-------
[ACME MOTOR COMPANY LETTERHEAD]
Dear Vehicle Owner:
You probably noticed as you opened this letter that the
mailing label shows the model and year of an Acme vehicle that
has been registered in your name. We have learned from the
State of Michigan that you are due to* receive a license tab
renewal notice on this vehicle. Your renewal notice will also
say that this year, Michigan may require you to have an Auto
Exhaust Test (AET) performed before you can buy your new tabs.
Why is this of interest to Acme Motor Company? Acme and
the United States Environmental Protection Agency are
conducting an important scientific study about cars and trucks
that fail their AET test, and we will be studying a small
number of vehicles just like yours. You may be able to help us
significantly in this program and be rewarded for your
cooperation.
Space in the program is limited, and we are looking foic a
limited number of each vehicle type. We do not have enough
information on your vehicle to guarantee now that you will
qualify. However, if you follow the directions given below,
and your vehicle is accepted into the study, we will offer you
a number of incentives to participate:
1. We will find and repair emissions problems on your
vehicle that may have caused the AET test failure.
All repairs will be free of charge (parts and
labor), and you will be issued a Vehicle Inspection
Certificate to allow registration of your vehicle.
2. When we pick up your vehicle, (at your home or place
of work) you will receive, at no cost to you, a
clean, fueled, and fully insured late-model car for
your unlimited use and convenience. We expect that
your vehicle will be needed for approximately
fifteen (15) working days.
3. Your vehicle will be returned to you cleaned and
with a full tank of fuel.
4. If your vehicle is tested, we will provide you with
a check for $50.00 in appreciation for your
participation in this program..
-------
-2-
The testing will be conducted indoors at facilities of
Acme Motor Company in Anytown, Michigan. Your vehicle will
probably accumulate less than ipo miles under simulated driving
conditions'. No unusual operations will be performed on your
vehicle and it will be fully insured for the entire test period.
If you are interested in the program, here's what you
should do:
1. Have an emissions check performed at any licensed
Michigan AET testing station, within 10 working days
of the postmark date on this letter.
2. You should inform your AET inspector that in the
event your vehicle fails, no repairs or maintenance
of any kind should be performed.
3. Only vehicles that fail the AET test are eligible
for our program. To be considered for the study,
you must contact us within 2 working days of the
time you failed your AET test at the following
number:
Acme Motor Company
Cooperative Test Program
(313)555-2222
When you call we will find out from you the
information we need to tell if your vehicle
qualifies.
4. If you pass your AET test, please do not contact us,
as we will be unable to accept your vehicle but
we do thank you for considering participation.
The enclosed information sheet answers some questions
people often have about this program. If you have additional
questions, please feel free to call the above number, and we
will be happy to help.
If you do fail your Michigan AET test, we hope to hear
from you. And thank your for your contribution to clean air in
Michigan.
Sincerely,
Edward Q. Engineer
Acme Motors Corporation
Enclosure
TS3:McCargar:law:X428:2565PLYMOUTHRD.:0614F:10/9/86
------- |