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 Post—Repair 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 FTP—Derived  LA4
HC
FTP
LA4
LA4/FTP
number
vehicles
184
184
total
as-rcvd
422.34
382.99
90.7%
total
reduction
338.98
321.91
95.0%
average
as-rcvd
2.30
2.08
90.7%
average
reduction
1.84
1.75
95.0%
%
reduction
80.3%
84.1%
104.7%

CO
FTP
LA4
LA4/FTP
number
vehicles
184
184
total
as-rcvd
6074.3
5665.9
93.3%
total
reduction
5192.6
5063.0
97.5%
average
as-rcvd
33.0
30.8
93.3%
average
reduction
28.2
27.5
97.5%
%
reduction
85.5%
89.4%
104.5%
     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
^^   X—X 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

-------