United States
       Environmental Protection
       Agency
                     NOT TO BETAKE
                          November
                            1990
        Air
x°/EPA
Cooperative  Testing
Program Draft Report
ECTD
TSS
      DRAFT

-------
                                          EPA-AA-TSS-I/M-90-X
                          **Draft**
                            v 1.5

                      Technical Report
          Report  on the EPA/Manufacturer Cooperative
                     I/M Testing Program
                             By

                      James A.  McCargar
                       Lisa Mouat Snapp

                        September 1990
                           NOTICE

Technical  Reports  do  not necessarily represent  final  EPA
decisions  or  positions.    They  are  intended to  present
technical analysis  of  issues using data which  are currently
available.   The purpose in the  release  of  such  reports  is to
facilitate  the  exchange of technical information  and  to
inform the  public of technical developments which may  form
the basis  for  a final EPA decision,  position  or  regulatory
action.

                   Technical  Support  Staff
             Emission Control Technology Division
                   Office  of  Mobile Sources
                 Office of Air and Radiation
             U.S.  Environmental  Protection Agency

-------
Section  1:    Executive  Summary	2

     1.1  Program Summary	2
     1. 2  Results	2

Section  2:    Background  and  Program  Summary	6

     2 .1  Program Overview	6
     2 . 2  Recruitment Quotas	7
     2 . 3  Procurement	8
     2 . 4  As-Received Testing Protocols	9
          2.4.1  Introduction 	9
          2.4.2  Basic I/M Test Procedure 	9
          2.4.3  Applicability of the BITP Model 	11
          2.4.4  As-Received FTP Test 	11
          2.4.5  Tank Fuel Analysis 	11
     2 .5  As-Received Diagnosis	13
     2.6  Selection of Vehicles for Remedial
               Maintenance	13
     2.7  Remedial Maintenance Protocols  and Post-Repair
               Testing	13
     2 .8  Database Structures	15
     2 . 9  Program Nonconformities	15

Section  3:    Basic  Vehicle  Characteristics  of  the
               CTP  Fleet	17
     3.1.  Introduction and Overview	17
     3.2.  Profile by Manufacturer and Quota Group	17
     3.3.  Mileage Profile	19
     3.4.  Profile by Vehicle Type	22
     3.5.  Additional Comments on the Base Sample	23
Section  4:    As-received  Emissions  Analysis	24

     4.1.  Introduction	24
     4.2.  FTP Results	24
          4.2.1.  Sample Description 	24
          4.2.2.  Pass/Fail Results at Certification
                    Standards	24
          4.2.3.  Excess Emissions Analysis	25
          4.2.4.  Analysis Using MOBILE4  Emitter
                    Categories	29
          4.2.5.  Correlation of Excess Emissions,
                    Emitter Category, and Odometer	32
     4.3.  Michigan AET Test Results	34
          4.3.1.  Profile of the Base Sample 	34
          4.3.2.  Stratifications by Manufacturer  and
                    Quota Group	35
          4.3.3.  Stratification by Mileage and Vehicle
                    Type	37
     4.4.  Correlation Between the FTP and Michigan AET
               Results	37
          4.4.1.  Introduction 	37
          4.4.2.  Excess Emissions and Emitter
                    Categories of the AET Failure Types  ...37
                              11

-------
     4.5.  Laboratory Short Test Results	41
          4.5.1.  Sample Description 	41
          4.5.2.  Second-Chance Failure Rates  of  the  As-
                    Received Base Sample	41
          4.5.3.  Idle-Mode Short Test Variability Due
                    to Fuels	45
          4.5.4.  Correlating the Laboratory Short Tests
                    with the MOBILE4 Emitter Categories ...46
          4.5.5.  Variability Between Adjacent Idle
                    Modes	49

4.6   Supplemental Analysis of  the  AST Errors of
               Commission	53

Section  5:   Emission Effects of  Remedial
               Maintenance	55

     5 .1  Introduction and Sample Descriptions	55
     5.2  Total Mass Emission Reductions  from  the CTP
               Fleet	55
          5.2.1  Net Benefit of Repairs — FTP-Based	55
          5.2.2  FTP versus LA4 values 	58
          5.2.3  Net Benefit of Repairs — LA4-Based	62
     5 .3  Overview of the Repairs Conducted	63
          5.3.1  System and Subsystem Repair Categories ...63
          5.3.2  Emission Benefits per System Repair	65
          5.3.3  Emission Benefits per Subsystem  Repair ...69
          5.3.4  Total Benefit per Subsystem	73
          5.3.5  Effect of Deteriorated Catalysts on
                    Emissions	75
     5 . 4  Analysis of Incremental Repairs	77
          5.4.1  Sample Description 	77
          5.4.2  Benefits of Repairing to Pass I/M	78
          5.4.3  Comparison to MOBILE4 Repair  Estimates ...81
          5.4.4  Benefits of Repairing to pass the FTP  ....83
          5.4.5  Emission Benefits "Lost" through
                    Second-Chance	85
          5.4.6  Benefits of Repairing to Different
                    Targets	86
     5 . 5  Repairs to High Emitters	87
          5.5.1  Effectiveness of Repair on Marginals
                    vs . Highs	87
          5.5.2  Benefit of Repairing Highs only  	89
          5.5.3  Catalyst Repairs Performed on Highs  	91
     5 . 6  Difficulty of Repair	93
          5.6.1  Difficulty of Repair to Passing  Levels ...93
          5.6.2  Effectiveness of Repair at Successive
                    RM Steps	97

Appendix  A:   Failure  Rates  in the  Michigan  AST
               Program	103

Appendix  B:    Vehicle  Identifying  Information for
               the  CTP  Base  Sample	104
                             iii

-------
Appendix  C:   As  Received Failure  Rates  for
              Selected  Modes of  the Basic  I/M  Test
              Procedure	105

Appendix  0:   AST  Errors  of Commas ion  in  the CTP
              Sample	106

Appendix  E:   Per-Repair Emission  Reductions for
              All  Systems:   by Quota  Group	107

Appendix  F:   Per-Repair Emission  Reductions for
              Statistically Significant   Systems:
              by  Quota  Group	113

APPENDIX  G:   Per-Repair Emission  Reductions for
              All  Subsystems	114

Appendix  H:   Per-Repair Emission  Reductions for
              Statistcally   Significant  Subsystems:
              by  Quota  Group	116

Appendix  I:   Total   Estimated  Emission Reductions
              for  All  Subsystems	121
                            IV

-------
                SECTION  1:   EXECUTIVE  SUMMARY
1.1  Program Summary

     The  EPA/Manufacturer  Cooperative  Test Program  (CTP)
recruited  private-owner  vehicles  based  on failure  of the
Michigan Auto  Exhaust Testing  (AET) Program for testing at
laboratory  facilities  of  the EPA  and  seven   major motor
vehicle manufacturers.   The program  focused on closed-loop
light-duty  vehicles and light-duty trucks  from model years
1981-1986.   The  test protocols  included the  Federal Test
Procedure (FTP) and a new short test procedure  with segments
simulating a variety of field  test conditions.   The remedial
maintenance  procedure  called  for  incremental  repairs and
retests to meet both FTP and short test criteria.

     Data from  the  program  were analyzed with  the following
objectives in mind:

    (1)   Developing advice to I/M  programs on improvements to
         preconditioning   methods   and  formal  I/M  test
         procedures.

    (2)   Seeking,  and assessing the  potential of,  a  limited
         diagnosis  and  repair  sequence as  a  remedy for a
         significant portion of the in-use emissions excess.

    (3)   Improving  methods   and models  for  estimating I/M
         effectiveness in  reducing emissions.

    (4)   Providing   feedback   to  the   manufacturers  on
         particular  malfunction or malmaintenance types.
1.2  Results

     The following results from analysis of the CTP data are
significant :

    (1)   Eighty-six percent  of  the 239 vehicles  in'' the CTP
         base  sample  failed  their HC  or CO  certification
         standards in the  as-received  condition,  with HC+CO
         failures being  the most prevalent.  Of the 'failures,
         one vehicle was  a super emitter and  1B&. were high
         emitters,  by  the  MOBILE4  definitions.  The worst 40%
         of  the  , vehicles  accounted for  90%  of the  fleet
         excess HC and 83% of the  fleet  excess CO.   [Sections
         4 . 2  and;1 4.3]
                '
    (2)   The mean excess  HC  and CO  emissions of the MY1981-82
         vehicle group exceeded the mean excess emissions for

-------
     the MY1983-86 group by about half.  Discounting the
     impact of the  one  super emitter,  the fuel-injected
     and  carbureted  vehicles  had  roughly  comparable
     excess emissions,  within  the model year groups.  The
     percentage  of  high emitters  varied  considerably
     across the manufacturers.   [Section 4.2.4]

(3)   Excess  emissions  were  not  well  correlated  with
     mileage;  however,   successively  higher  MOBILE4
     emitter categories  showed  successively  higher mean
     mileages.   [Section  4.4.2]

(4)   Of  the 40%  of  the  CTP  fleet  that  were  normal
     emitters,  almost 80% failed their recruitment short
     test  for  only  one pollutant;   HC-only  failures
     outnumbered  CO-only  failures  by two  to one.   The
     data show no  reasonable alternative  outpoints  for
     excluding  large numbers of normal emitters from AET
     failure without  inappropriately converting many high
     emitters to AET passes.   [Section  4.4.2]

(5)   When tested  in a fully-warmed condition with either
     loaded  or extended-2500rpm preconditioning,  the
     second-chance  idle failure  rate of the CTP fleet was
     only about 40%.   Up  to an  additional  20%  of  the
     fleet  failed  second-chance  idle  tests  under  a
     variety of conditions on warmed-up vehicles.   Only
     on vehicles that were idle tested immediately after
     an extended soak did the second chance failure rate
     exceed  60%.   [Section 4.5.2]

(6)   Vehicles that  were  combined HC+CO failures on their
     AET tests  were not also  prone  to be  HC+CO failures
     during  second-chance testing.   [Section 4.5.2]

(7)   The  second-chance  idle  tests  at   fully   warmed
     condition,   with   loaded  or  extended  2500rpm
     preconditioning,  reduced the  error  of  commission
     rate to zero or near zero,  and passed 78% or more of
     the marginal  emitters  as  well.  The  failure  rate of
     the high emitters  under the same  second-chance test
     conditions was 60%-65%.    [Section  4.5.4]

(8)   Correlation between the two idle modes of a two-mode
     idle test  was  relatively  high (R2  values  of close to
     90%)  when the vehicles  were  fully  warmed  and
     preconditioned  with  loaded  or  extended-2500rpm
     operation.   Poorer  correlation (R2  values  between
     31%  and   85%)  was  shown  under  less  ideal test
     conditions.   For all test conditions,  the  effect of
     the intervening 2500rpm mode was generally to reduce
     the   failure   rate   on   the   second   idle.
     [Section 4.5.5]

-------
(9)   All  but  two  of  the  33  as-received  error  of
     commission  vehicles  had elevated or  failing  HC
     scores  on the Michigan AET test;  slightly over one-
     third failed the AET for CO.   One-quarter of the Ec
     vehicles fell in one  GM  engine  displacement.   Only
     one of the Ec vehicles had a  repeatable  short test
     exceedance that was diagnosed and  resolved through
     repair.   Most Ec' s were  attributed to  the response
     of  the  engine  and  control  system   to  vehicle
     preconditioning.   [Section 4.6]

(10)  Repairs  eliminated 99%  of the  excess  HC and  CO
     emissions in the CTP fleet; one fifth of the overall
     HC  reduction  and  one   third  of  the  overall  CO
     reduction were  due to  oxygen sensor  replacement.
     Catalyst and  fuel  injector replacements  also each
     contributed  more than  10%  to the total HC  reduction,
     while no other repair  type contributed more than 10%
     to the  CO reduction.   The average  vehicle  had 1.8
     g/mi HC  and  28  g/mi  CO  eliminated in  2.6  repair
     steps.   [Sections 5.2.1 and 5.3.4]

(11)  The most  frequent  repair types were oxygen sensor
     replacements  (45%  of  repaired vehicles),  catalyst
     replacements  (30%),  and  ignition  tune-ups   (spark
     plug/wire replacement,  initial timing  adjustment,
     idle speed adjustment)  (29%).   [Appendix  G]

(12)  The  most effective repair  type on a  per-vehicle
     basis   was   fuel  injector   replacement.     This
     eliminated an average 2.4 g/mi  HC  and  24 g/mi CO.
     Other statistically significant effective HC repairs
     were  to  the  catalyst  (1.1  g/mi),   carburetor
     (1.0 g/mi),  oxygen sensor (0.8  g/mi)  and fuel meter
     tune-ups  (0.6  g/mi).    For  CO,   the   significant
     effective repairs were to  the load  sensor  (23 g/mi),
     oxygen  sensor (21 g/mi),  carburetor (12  g/mi), fuel
     metering system tune-ups  (12 g/mi), ECU  (12 g/mi),
     and catalyst  (7  g/mi).    The average reduction per
     repair  step  for  all repair types was 0.7 g/mi HC and
     11 g/mi  CO.   [Section 5.3.3]

(13)  Vehicles  required an  average  of   1.5  repairs  to
     switch  from high  emitter to marginal  or  passing
     emitter status on  the FTP, with  emitter categories
     as defined by MOBILE4.  The same number of steps was
     needed  to change from  failing  to  passing on an idle
     I/M test performed  under  ideal conditions.  [Section
     5.6.1]

(14)  Marginal emitters  (as  defined by MOBILE4)  were not
     worthy  repair targets in  the  CTP.   Their emission

-------
     reductions  were  negligible,  with  high  emitters
     achieving reductions  15  times as  large.    [Section
     5.5.1]

(15)  Repairs targeted at  the  most likely malfunctioning
     component resulted in the  early  repairs being more
     effective than later repairs; the first repair on a
     vehicle  achieved  an  average reduction  five  times
     that of the  third repair.   [Section 5.6.1]

(16)  Second-chance  I/M  tests  apparently  reduced  the
     incidence of unnecessary repairs  to normal emitters
     without  greatly   reducing  the  benefit  due  to
     repairing high emitters.  Sixty percent of the fleet
     passed  a second-chance  I/M  test performed  under
     ideal conditions;  this 60%  would  have achieved less
     than  25% of  the  total  fleet  repair  reduction.
     [Section 5.4.5]

(17)  Repair  types  that  were  consistently  effective  at
     reducing FTP emissions were  also  the most reliable
     at  correcting  I/M-failed vehicles to passes.  At the
     system level,  these were the  exhaust, fuel metering,
     and 3-way control  systems.   Subsystems included the
     catalyst,  carburetor, oxygen  sensor,  fuel metering
     system  tune-ups,  and  fuel  injectors.    [Sections
     5.3.2 and 5.3.3]

(18)  Repairing until vehicles passed I/M reduced emission
     levels  by approximately 75% and eliminated over half
     of  the  FTP excess.  However,  this was only about two
     thirds  of the  reduction  that  could be realized with
     more complete  repair.  [Section 5.4.6]

(19)  Replacement  of the catalytic  converter  resulted  in
     average  g/mi  reductions twice  that  of all  other
     repair  types for HC and 3/4 that of all other repair
     types  for  CO, even  though  catalyst repairs  were
     generally withheld  until  the  last  repair.   Evidence
     of  tampering or misfueling occurred  on  only  20%  of
     the  vehicles  that   received  catalyst   repairs.
     [Section 5.3.5]

(20)  Estimates of  average  emission reductions due  to
     repair  to  I/M  passing  status are  lower  in the
     MOBILE4  model   than  those  seen  in  the   CTP,
     particularly  for   HC  on  fuel injected  vehicles.
     [Section 5.4.3}

(21)  Ninety-four  percent of vehicles that  were repaired
     from high to  normal  emitter  levels on  a  transient
     test could also,  at that repair stage,  pass  an I/M
     test performed under  optimum conditions  (57%  from

-------
I/M fail  to pass,  37%  already passing)   [Section
5.4.4]

-------
         SECTION 2:   BACKGROUND AND PROGRAM SUMMARY
2.1  Program Overview

     The Cooperative EPA/Manufacturer I/M Test Program  (CTP)
was a joint effort by the Environmental  Protection Agency and
seven   of  the   major  domestic  and  imported   vehicle
manufacturers  to recruit  failed  vehicles from  an  official
state  Inspection/Maintenance  (I/M) program  for study  in  a
laboratory  environment.   The  intent  of the program was to
gather data that could be combined with  the  results from a
number  of  other  studies,  contributing  to  accomplishing
several objectives.   Among these were the following:1

    •   Develop  advice  to  I/M programs on  improvements to
       preconditioning  methods  and   formal  I/M   test
       procedures.

    •   Seek, and assess the potential  of,  a limited diagnosis
       and repair  sequence  as a  remedy  for  a  significant
       portion of the in-use emissions excess.

    •   Improve  methods  and  models   for  estimating  I/M
       effectiveness in reducing  emissions.

    •   Provide  feedback to  the manufacturers  on particular
       malfunction or malmaintenance types.

     The study  focused  on 1981-86  model-year  vehicles  with
closed-loop engine  control  that  failed  the idle  test  under
Michigan's decentralized Auto Exhaust  Testing  (AET)  program.
The EPA  solicited owners  just prior  to their  expected AET
test date by direct  mail for  participation in the program and
also coordinated initial  recruitment  efforts.   Each  of the
participating  manufacturers  —   Chrysler,   Ford,   General
Motors,  Honda, Nissan,  Mitsubishi, and Toyota  —  completed
the recruitment process and performed  the  laboratory  work on
its own vehicles, at  its own  facilities.

     Vehicles  of some additional  manufacturers,  which had no
appropriate  testing  facilities  in  Southeast  Michigan
available during  the program, were  recruited to the EPA Motor
Vehicle  Emissions  Laboratory  in  Ann  Arbor,  Michigan  for
testing  and repair.    This group included  vehicles  from
American Motors,  Mazda, Subaru, and Volkswagen.   Facilities
for Toyota were  under  construction  at the  outset  of  the
program,  and  testing  of  the  first  ten  Toyota vehicles
consequently took place  at  MVEL;  work  on the  remaining six
Toyota vehicles was  completed by the manufacturer.

-------
     According to  the CTP program  plan,  recruited vehicles
first underwent a series of emissions tests, including short
cycles  and the  Federal Test  Procedure  (FTP) ,  in  the  as-
received condition.   Each vehicle  was  diagnosed for causes of
any excess emissions;  it was  then  repaired,  and retested.   At
least two  features  distinguished  the  program protocols from
those in previous in-use studies conducted by EPA and others.
First,  the  remedial  maintenance  and  retest  steps  were
incremental; that  is, mass-emission  and  raw-gas  tests were
executed  following  each "significant" repair.   Second,  the
program's  short  cycles  incorporated a  new Basic  I/M Test
Procedure   (BITP)  that  was  designed to  simulate  vehicle
response under  a  variety of field  I/M test conditions.

     Recruitment for  the Cooperative Test  Program began in
February 1987,  and testing continued through May 1988.  Data
from  the participants  was collected,  quality checked  and
organized  in a common database through  the EPA  facility in
Ann Arbor.  Final  data submissions  and major corrections to
the databases were  completed in May  1989.

     The remainder of this section summarizes portions of the
CTP program plan that will aid the reader  in understanding
the analysis to follow.


2.2  Recruitment  Quotas

     The Cooperative Test Program vehicles were not a random
sample  from among all  those  that  failed the  Michigan Auto
Exhaust  Test.    Quotas  on  recruitment  were  established
according to a number of variables:   manufacturer,  model-year
group,  fuel-metering strategy,  presence  of  tampering  or
misfueling, and presence of a pattern failure.

     The   manufacturer  quotas   distributed  the  testing
obligation based   on  the   relative  percentages  of  each
manufacturer's  fleet  in  a 140,000-vehicle sample of failures
from  the Seattle  I/M program.   This program was selected
because of  the availability of detailed data, its  procedural
similarity to  the  Michigan  program,  and  its   use  of  a
keyoff/restart  step for  Ford vehicles.2

     Regardless  of  its  share in  the  Seattle  failures,  each
participant agreed  to test a minimum of  ten  vehicles;  this
minimum was applied to Mitsubishi  and  Honda.  General Motors,
Ford, and  EPA  each  entered  the program with testing targets
of 60 vehicles.  Nissan was  allocated  30  vehicles,  Toyota 16,
and  Chrysler,  15.   Thus  the  program plan target  was  261
vehicles.

     Only  closed-loop vehicles from  the 1981 through  1986
model years were recruited.   By definition,  manufacturers
producing no closed-loop vehicles  in the earlier model years

-------
necessarily met their recruitment goals from the later years.
In order to  further  focus  on technologies  likely  to  dominate
I/M fleets in the 1990s, carbureted vehicles were  targeted  to
be less than half of each  organization's basic  quota,  and  at
the manufacturer's   option,  post-1983MY carbureted  vehicles
could be excluded from recruitment.

     For manufacturers with closed-loop production across the
model years,  some additional considerations applied:   half  of
the quota was to be  filled by  vehicles  from the 1981  or  1982
model  years,  the other half  from  the 1983  or later model
years.  Some  additional considerations applied, in order  to
ensure  a  sample with current  technology, as  well  as age-
related malperformances.

     In order to preclude domination of the sample  by certain
important  forms  of  tampering,  the number of vehicles with
fuel  inlet tampering, Plumbtesmo test  failure, or  catalyst
removal was  limited  to  the greater  of one  vehicle  or ten
percent of a manufacturer's overall CTP quota.  Finally, the
participants  were also  expected to limit  recruitment of so-
called  "pattern  failures"  from  the  sample,  under  the
principle  that limited information  of  value would be  obtained
if a  manufacturer's  sample were heavily  biased towards one
(or a few)  vehicle  groups once a pattern of problems  had been
adequately diagnosed  and remedied.

     Because  the primary recruitment  quotas were  based upon
model year and  fuel  metering  strategy,  four "quota  groups"
were defined  for each manufacturer:   fuel-injected  MY1981-82,
carbureted 1981-82MY, fuel-injected MY1983-I-,  and  carbureted
MY1983+.  The quota  groups  will be  referred to  frequently  in
the remainder of this report.

2.3  Procurement

     Each  week over  the  course of  the program,  EPA culled a
list of potential owners from a tape provided  by the Michigan
Department of State,  containing registration data  for owners
of  vehicles   due to receive  notices  of their  AET  test
requirement  in  that   week.   Decoders  based on the  Vehicle
Identification  Number  (VIN)   were  employed  to  screen  out
ineligible vehicles.   A  mailing   label  was generated  for
eligible   owners,    and  a  direct-mail   solicitation  was
conducted.   The solicitation  letter  contained boilerplate
language  on  the program  and  incentives  for the owner  to
participate,  as  well as manufacturer-specific  information.
The letter emphasized that  only those  owners who failed their
subsequent  AET test would be  eligible,   and  it  provided
instructions  for interested owners  to  contact  the appropriate
participating CTP test site or  its representative.

-------
     The  size  of a  week's mailing  was  determined by  each
site's   throughput   and  progress   towards  meeting   the
recruitment quotas.   Where more eligible owners existed than
CTP testing capacity for the  likely  respondents, the pool was
reduced by  randomly  selecting owners from all those  in the
same quota group.

     Once  an  owner  contacted  a  CTP recruitment  site,  a
telephone  screening  was conducted  to  verify  eligibility.
This screening  verified the Michigan AET  test  failure, the
timeliness  of  the  owner response, the absence of  any owner
action to remedy the  failure,  and set of  safety  and outlier
rejection  criteria.3   Conformance to the  any late-breaking
changes  in  the manufacturer's progress towards  recruitment
quotas  was  also taken  into account.   Owners not  excluded
during the  telephone  screening were  scheduled for  intake to
the  appropriate test site.   A  final safety and  outlier
screening  was  performed  at  the  time  of vehicle  intake,
including a road test  to  project safe dynamometer  operation.

2 . 4  As-Received Testinqr  Protocols


     2.4.1  Introduction

     The CTP program plan called  for each  vehicle to receive
four short-cycle sequences,  collectively  referred  to  as the
Basic I/M Testing Procedure  (BITP), on tank  fuel (Table  I).4
The  BITP was  then  followed by  the Federal  Test  Procedure
(FTP)  on  Indolene.    If  the vehicle  was  recruited  with
insufficient fuel to complete BITP testing, the program plan
called  for  substitution of a  commercial  fuel.  An  RVP and
lead-in-fuel analysis  was performed on the  tank fuel.
                           TABLE  1

            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
     2.4.2  Basic I/M Test  Procedure

     The  Basic I/M  Test  Procedure  attempted to  replicate
unloaded  field  I/M  tests under  a variety  of  controlled

-------
conditions, and thereby provide possible explanations  for  the
Michigan  AET  results  that were  the basis  for  CTP  vehicle
recruitment.   These  controlled  conditions  included  the  prior
 ("base")  operation  of  the vehicle and  any  conditioning that
occurred  immediately  prior  to  the  pass/fail  test  modes.
Examination  of other  factors that might  have affected  the
Michigan  AET  results,  such  as  AET   analyzer  calibration
variables  or operator  fraud, was not  in  the  scope  of  the
program.

     The four sequences that make up  the BITP corresponded to
(and were named after)  four different types of preceding, or
"base"  operation:    a  cold start, extended operation  under
load, extended operation at idle,  and execution of  an engine
keyoff/restart.  The base  operation was then  followed by  one
or  more  simulated  Two-Speed  Idle  tests,  each  with a
controlled  conditioning mode.   Raw-gas emission values  and
engine  RPM  measurements were gathered  throughout  the  BITP,
but  attention  was  focused on seven  "core" two-speed tests,
spread  through  the procedure.    No  loaded  short  testing
(either transient  or steady-state)  was included in the  BITP.

     Table  2  summarizes  the modes  of the Basic  I/M  Test
Procedure.   The  core sampling periods, which will  serve as
the  basis  for much  of the  short-test  analysis that is to
follow,  are shaded.   Certain  parameters — HC, CO,  C02,  and
engine RPM,  — were sampled throughout  the  procedure.  During
the core sampling periods, these  parameters were measured at
15  and  30 seconds  into each mode,  and  at 60,  90 and  120
seconds of  the second  idle-neutral  mode as  well.    Coolant
temperature  was  monitored  during  the  entire  cold start
sequence,  and at other  points (such  as the extended  idles),
where some significant  variation might occur.

     Sampling intervals outside  of the core sampling period
varied  according  to the  purpose of the  mode.  For  a  more
detailed description of the sampling procedure, refer to  the
program plan.5

     The selection,  order,  and duration  of  the  modes reflects
a desire to test the vehicles under both ideal and  non-ideal
conditions.   The purpose  of the  Cold Start  sequence,   for
example, is to characterize the emissions of each vehicle at
abnormal  (low) operating  temperature,  and then to  determine
the effectiveness of various  conditioning modes at  achieving
normal  operating  temperature, and the  emissions  impacts of
such conditioning

     The  Extended  Loaded  sequence  was  the  hypothetical
"ideal" test condition  in  the procedure;  its purpose  was to
measure vehicle emissions  immediately  following an extended
period  of loaded  operation.   The  Extended  Idle  sequence
presumes that the vehicle  has  prior loaded  operation that  had
                             10

-------
achieved normal operating  condition  in the recent past,  but
that a period of extended  idle may have caused the vehicle's
condition  (and therefore  its  emissions)  to deviate from its
ideal levels.

     Finally/  the  Restart  sequence  examines  the non-ideal
aspects of  a  keyoff/restart  on vehicle emissions.  For most
vehicles, the  ideal  condition  was  presumed to not involve a
restart,  and to isolate the effects of  the  restart  from other
conditioning  variables,  these  vehicles were  not restarted
during the  Cold Start,  Extended Loaded,  and Extended Idle
sequences,  but  were  restarted during  the  Restart sequence.
On  the  other hand,  Ford  vehicles  from the  1981 model year
onward were generally designed  with the  assumption  that a
keyoff/restart would  precede  any  idle short test.6  Use of the
restart  for Fords  in the  CTP  was  therefore  the opposite of
the case for non-Fords  (see Table 2).

     2.4.3  Applicability of  the  BITP Model

     It  is  important to  note  that in  the Cooperative Test
Program  the BITP  simulates field  test  conditions but  not a
field  sample.   The program  recruited   initial  idle-test
failures only; no AET-passing  vehicles underwent the BITP or
other procedures from the  CTP protocol.

     The  seven  core  sampling  periods  in   the  BITP  thus
represent  second-chance  I/M  tests   on  failed  vehicles.
Without the analogous data on the passing vehicles, care must
be  taken when interpreting  failure  rates, excess emissions
identified,  variability  results, and  other  analyses  on the
BITP data.  This caution applies as well when examining parts
of  the  BITP that had  no direct analog in the Michigan AET
test, such as 2500rpm pass/fail results.


     2.4.4  As-Received FTP Test

     The  FTP used  in the Cooperative Test Program  was a
standard  three-bag  cycle  performed on Indolene  (Table  1,
above) .   The  CTP version of  the test omitted the  heat build
and the Highway Fuel  Economy  Test.


     2.4.5  Tank Fuel Analysis

     Lead-in-fuel   analysis   was   performed  using   x-ray
flourescence and targeted  at  designating  the fuel as either
above or  below a 0.05 g/gal standard.   Reid Vapor Pressure
testing was conducted using the ASTM D323 protocol.
                             11

-------
                                 TABLE  2

                 Modes of  the Basic  T/M Test
SEQUENCE
MODE*
MODE NAME
DURATION
    FUNCTION
Cold Start  (CS)
  0 1
  0 2
  03
  04
  0$
  0 6
  0:7
  75° Soak
Engine  Start
                              2500rpm
                               le^neutra
                              2500rpm
 >60 min
   n/a
 30 $0c.l
 30 sec *
 120 $e$
 180 sec
                            K&yoff/Restart
                       1 1
  Base Operation
  Base Operation
  Core Sampling
  Core Sampling
  Core Sampling
   Conditioning
  Core Sampling
     VeWcfe* Only
Extended Loaded (XL)
  01
  oa
 -frSL
   LA 4
1372 sec
                            Keyolf/Resfart
  Base Operation
  A y"VF.•«Vrtvw* •"•• •• *^%i^iW!fiaa?ss'A^v
  &** tomtfte&fr
Ford Vehicles Only
                     ^
                      120
                                               sec
Extended Idle  (XI)
        Idle-neutral
              20 min
             Base Operation
                                                             SampHrig
                                                             ••*... .:• •• -•- H «3f V
                                                              «wett«
                                                         Conditioning
                            Keyoff/Restaft
Restart  (RS)
  01
             1372 sec
        Idle-neutral
                                           120 860
             Base Operation
             y^ ^SvSty •g^sy^o-'VA^y'^ *'3SftSKsftM:
           . . Core Sampling!
           on-Ford Vehlcfes
           /-Core Sampling
                  Sampling
                                    12

-------
2.5  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
                             13

-------
biggest projected impact was conducted first (followed by the
appropriate  test  sequence,  to  quantify the  impact of  the
repair.8
                           TABLE 3
           Allowable Exit Criteria in the  Remedial
                      Maintenance Phase
CATEGORY
FTP

I/M
Variability



I/M Basic


MILEAGE
<50K
>50K
N/A
N/A
N/A
N/A
N/A
N/A
N/A
ALLOWABLE EXIT CRITERIA
HC & CO <1.5 * cert standard AND
HC & CO <2.0 * cert standard
Analogous sampling points between
sequences show comparable values AND
successive sampling points within a
test mode show comparalbe values; OR
Observed variability traced to
unrepairable element of design OR
Variability cannot be repeatably
triggered OR
All reasonable repair efforts
completed
Core sampling emissions in extended
loaded sequence pass 207 (b) OR
Observed 207 (b) failure traced to
unrepairable element of design OR
All reasonable repair efforts
completed
     The one  exception to the  above  guideline was catalyst
replacement.  Dramatic emissions improvements  would normally
be expected with  installation  of any "green"  catalyst, even
if performance of the original  catalyst  was  acceptable.  Such
replacements  could mask  the  importance of  other important
malfunctions in the engine or  other emission  control devices.
On-vehicle diagnosis was also anticipated to be difficult  in
some  cases,  except  where overt  signs  of  damage  to  the
container  or the  the  substrate  existed.    Thus,  catalyst
replacement was  treated as  the repair of last resort.

     As a guide for  decision-making when faced with multiple
repair options,  the CTP program  plan established a guideline
for the priority order of  repair, as follows:

    1.   Computer   control  and  feedback  system  repairs,
        including most  repairs indicated by onboard diagnostic
        systems  and  repairs  to  electronic   fuel  metering
        components;
                             14

-------
    2.  Primary emission  controls other  than those  in the
       feedback system,  including exhaust  aftertreatment,
       secondary air, PCV, and EGR systems;

    3.  Idle mixture  adjustment,  on  vehicles  with missing
       limiter  devices.

    4.  Other basic engine components.

     In general, post-repair testing was  conducted  after each
repair step in  the  CTP.   The  test procedures employed were
all  those  necessary  to track  the  vehicle's  performance
relative  to  each of  the outstanding  failed  exit  criteria.
Thus, a repair  targeted  at  an  FTP failure on a vehicle with
remaining short test  noncompliance would be followed by I/M
testing as well as FTP testing.

     As a  cost-saving measure, both the  FTP  and short test
protocols  could be  shortened  during post-repair testing.
Test sites  could  perform an LA4 cycle in place  of the FTP,
until a  significant  improvement  in emissions was  exhibited
between  the post-repair  LA4 and  the  weighted bag two and
three results  from the previous FTP.   At  that  point, however,
a full FTP was to be conducted  before  proceeding with  further
repairs.   For short  testing,  abbreviated  versions of the BITP
were  available that consisted  of   the  extended  loaded
sequence, plus  remaining sequences showing violations of the
variability or basic  I/M  criteria.

2 . 8  Database  Structures

     Data  from  all  of the  important  aspects  of the program
described in the above sections were recorded  in a  relational
database  (MICRO)  on  the Michigan  Terminal  System  (MTS) and
were also downloaded to  microcomputers   for  analysis.   The
file structure  of the CTP  MICRO database mimics that of the
EPA Emissions Factors database, with additional datasets and
fields for the  information that is unique to the CTP.  A new
repair database format was  constructed to categorize repair
actions  and facilitate analysis.  CTP program participants
were afforded access  to  these  data through MTS accounts and
through copies of  the  subset  microcomputer databases.

2.9  Program Nonconformities

     By and large,  the participating organizations  in the CTP
operated  independently,  under  the  guidelines  of the CTP
program   plan.    There  was,   for example,   no  real-time
coordination between the participants  during vehicle testing
or when short  turnaround  decisions  on  repair protocols needed
to be made.   Thus,  each testing organization exercised its
                             15

-------
own  judgment  in  unusual  or  borderline  cases.    Some
nonconformity in the data results, which complicated some of
the analytical tasks.   Relevant  and  significant examples will
be identified in the sections that follow.
                             16

-------
 SECTION  3:   BASIC  VEHICLE  CHARACTERISTICS  OF THE  CTP  FLEET
3.1.
Introduction and Overview
     The  participating organizations  recruited 245 vehicles
for  study  in the  Cooperative  Test  Program.   A detailed
breakdown of  the vehicle identifying  information for all of
these vehicles appears in Appendix B.

     Malfunctions  in two  of  the  245  vehicles  prevented
gathering sufficient as-received  or post-repair data  to
justify  inclusion  in  the base  analytical  sample.9   Four
additional vehicles received no initial FTP, and  had no post-
repair  FTP  that could  be  reasonably  substituted  for  the
missing  as-received  test.10   These four  vehicles are only
included in a limited number of analyses  in  the sections that
follow.  The bulk of this report  focuses  on  the remaining 239
vehicles, hereafter referred to as the base CTP sample.

3.2.  Profile by Manufacturer and Quota Group

     By and large, the actual CTP vehicle sample met the
intent of the program plan targets described in Section  2.2.
Table 4 and Figure 1 show the breakdown of the 239-vehicle
sample by manufacturer and quota group.11  As anticipated in
                           TABLE 4

             CTP Base  Sample bv Manufacturer Share
Manufacturer
General Motors
Ford
Nissan
Toyota'
Chrysler
American Motors**
Volkswagen**
Mazda"
Subaru**
Mitsubishi
Honda
TOTAL
Vehicles
58
57
20
16
15
15
14
12
12
10
10
239
Sample %
24.3
23.8
8.4
6.7
6.3
6.3
5.9
5.0
5.0
4.2
4.2
100.0
                 Testing split between EPA and manufacturer
               ** Testing performed by EPA
                              17

-------
the  project plan,  General Motors,  Ford, Nissan,  Chrysler,
Mitsubishi, and  Honda  tested  their own vehicles.  Testing of
Toyota vehicles  was divided between EPA and the  manufacturer.
Vehicles of the  remaining manufacturers were tested by EPA at
MVEL.

     In  general, the  CTP  recruitment reflects  patterns one
would have  expected from the  actual vehicle  fleet.   Ford and
General  Motors,  which  together  comprise  half  of the CTP
sample,  are each  dominated  in  the  earlier  model  years by
carbureted  vehicles,  and in  the later model  years,  by  fuel
injection.    Some  manufacturers  produced  no  closed-loop
vehicles in a  given  quota group,  and thus none were present
in the program.  Examples include  fuel-injected  1981-82 Mazda
and Subaru.  In  other  cases, vehicles in the quota group  were
produced, but  none was recruited  in  the  CTP.   Examples  here
were the fuel-injected  MY1981-1982 Toyotas,   and carbureted
MY1983-1986 Subarus.
                           FJLGURE  1
        Fleet Profile bv Quota Group and Manufacturer
          GM TOR) NISS TOYT CHRY AMC  VW  MAZ SUBA MITS HCND
        Fl 83-86
Garb  83-86 D Garb 81-82 • Fl 81-82
                              18

-------
     Table  5  shows  the 239-vehicle  fleet distribution  by
quota group.   The  fleet comes  quite  close to  the  planned
target  of  50%  each  for  fuel  injection  and  carburetion.
Manufacturers  who have  unusual representation  in  a  quota
group are Volkswagen,  with  one-quarter of the fuel-injected
1981-82 vehicles, and  Subaru  and Toyota,  whose total of ten
carbureted  1981-82  vehicles were the  only contributions  by
Japanese manufacturers  to  that quota  group.   Interestingly,
three of Volkswagen's  fourteen  vehicles were carbureted.

     Although  no specific  recruitment targets  existed for
individual  model  years,  the  sample  was  well-distributed
across  the  1981-86  range.    Of  the   six  model  years
represented, none comprised greater than  23%  of  the sample,
and none was less than 11%.
                           TABLE  5

                Fleet  Breakdown bv Quota Groun
Quota Group
Fuel-Injected 1981-82
Carbureted 1981-82
Carbureted 1983-86
Fuel-Injected 1983-86
TOTALS
Vehicles
23
66
50
100
239
Fleet %
9.6
27.6
20.9
41.8
100.0
3.3.  Mileage Profile

     The  CTP sample  was also  well-distributed  by  mileage
(Figures  2-3),  again  without an  explicit  recruitment
criterion.   The median  mileage  was  close  to  50,000 miles.
Not  surprisingly,  few  (in  fact  only  15)  of  the under-50K
vehicles were in the  1981-82 quota groups.  Conversely, only
14  vehicles  exceeded 100,000 miles,  with  only  one  from the
1983-86 model years.

     The, fleets of  six  of the  eleven CTP manufacturers were
composed entirely of  vehicles with accumulated mileage under
100,000  (Figure 4).   Of these,  Volkswagen is  notable  for
having close  to 80% of its vehicles in  the  range 50,000 to
100,000  miles.   Subaru  and  Toyota both had  significant
numbers of very high  mileage vehicles in their samples, with
25%  and  18%,   respectively,  over  100,000 miles.    Toyota
weighed in  with the highest mileage vehicle,  a  1982 Toyota
Corolla with 234,000  miles.
                             19

-------
                     FIGURE  2

         Mileaaft  Profile hy Qun1-a flrrmp
               2   3   4   5   6   7   8   9   10  11  >12
              Lower Bound of 10K Mileage  Interval
Fl  83-86
Carb 83-86 DCarb 81-82  • Fl  81-82
                    FIGURE 3
  Cumulative Mileage Profile bv Quota
                                         m Total
                                         ••-Fl  81-82
                                         •0-Carb 81-82
                                         •••Garb 83-86
                                         •0-FI  83-86
          <20  <40  <60  <80 <100 <120
              Mileage Category
                 (1000's)
                        20

-------
                            FIGURE 4

             Mileage Distrihut* i nn bv Manufacturer
     100

      80

      60

      40

      20

       0
IliM
i   i   i   •   i   i   i   i   i   i   i   i
l*:'v •TO::^B;;;:;:-:-B :^ ::' •i^;Bi^:^B  : -•IP$HllliPH$MH-:i;::;:!;B.
• •:•.•.;.. ••&N₯X«^* ::;-:::•: ^m-::::.::: ^•&&&$ ^•«xW;:>S ^•--.; .-:- ^«£&&& ^BSSSXsvJ ^•<&x&& •• ••-•••••• ^H;
|^^HK^H^H^^HiMl|iiMHiM^Hi£MnKillsi^Hi^nv
          GM  PQFD NISS TOYT CHRY AMC  VW  MAZ SUBA MITS HCMD ALL
                              <100K  Ll<50K
     Figure  5  shows the mean mileages  by manufacturer for the
base CTP  fleet.  The  "total" entries  represent  mean odometer
miles;  the "annual"  entries are  estimates of annual mileage
accumulation,   derived  for  each  vehicle  by  subtracting the
model  year  from 1987,  dividing  by the  odometer  miles, and
then averaging across the given manufacturer.
                            FIGURE 5

           Mean Mileage Accumulation bv Manufacturer
  M
  i
  I
  e
  s


 (K)
             POFD NISS  TOYT CHRY AMC  VW MAZD SUBA MITS HCtO  ALL
                            I Total   D Annual
     The  fleet  mean  odometer  was  55,000  miles,  and  the
calculated  mean  annual mileage  accumulation  was  just  under
                               21

-------
17,000  miles.   The high-mileage  Toyota helped  drive that
manufacturer's mean  odometer  reading  over  70,000  miles,
highest among  the participants; Mitsubishi  (with no vehicles
in  the  1981-82 category)  yielded the  lowest  mean odometer
reading, 40,000 miles.  Honda vehicles  displayed the highest
annual mileage accumulation (24,400 miles),  almost  twice that
of  AMC,  whose 12,300-mile annual  pace  was  the lowest among
the participants.


3.4.  Profile by Vehicle Type

     Overall,  eight percent of  the 239-vehicle  CTP  sample
were  light-duty  trucks   (Table  6).     Two  participating
manufacturers,  General  Motors  and Chrysler, elected  not to
recruit  LDTs.   Two others,  Mazda and Honda, produced no
closed-loop LDTs in the model  years  covered by the program,
and thus had none  in their  CTP  samples.   Of  the remaining
manufacturers, only American Motors had an LDT percentage of
greater than 25% (from the Jeep line).   Most  of the trucks in
the CTP sample  fell  in   the  later model   years,  probably
reflecting the  slower  penetration  of closed-loop technology
in trucks relative to  that  in LDVs.
                           TABLE  6

               Fleet Breakdown  bv Vehicle Type
Category
LOV
LDT
LDT %

General Motors
Ford
Nissan
Toyota
Chrysler
American Motors
Volkswagen
Mazda
Subaru
Mitsubishi
Honda
58
48
19
15
15
1 1
13
12
1 1
8
10
0
9
1
1
0
4
1
0
1
2
0
0.0
15.8
5.0
6.3
0.0
26.7
7.1
0.0
8.3
20.0
0.0

Fuel-Injected 1981-82
Carbureted 1981-82
Carbureted 1983-86
Fuel-Injected 1983-86
23
64
40
93
0
2
10
7
0.0
3.0
20.0
7.0

SAMPLE
220
19
7.9
                             22

-------
3.5.   Additional Comments on the Base Sample

     Some  manufacturers  showed  unusual concentrations  of
selected engine displacements  in  their  samples.  For example,
fourteen  of the  fifteen  Chrysler vehicles  (or  93%)  were
135CID  engines.   In  the  national  fleet,  this  displacement
accounted for between 64%  and  74%  of Chrysler's production of
closed-loop vehicles in model  years 1981-86.   Close to three-
fifths   of   the  57  Ford  CTP   vehicles   were   140CID,   a
displacement that was  only 19%  to 28% of Ford's closed-loop
production in  the model years covered  by the CTP.   Finally,
thirteen  of  the 30  fuel-injected  GM  vehicles (or  43%)  were
151CID,   compared to  between  10%  and 20% of GM's closed-loop
production in the given model  years.

     Higher I/M failure rates  are one possible cause, but not
the  only  one,  for  the   greater  representation  of  these
families; AET  failure  rates by manufacturer and displacement
are not  available from Michigan.
                             23

-------
         SECTION 4:   AS-RECEIVED  EMISSIONS ANALYSIS
4.1.  Introduction

     This section analyzes the  emission  results  from  testing
of  the  CTP  fleet in  the as-received  condition.   The  FTP
results are  analyzed first,   including  an analysis based  on
the  "excess  emissions"  concept  and  emitter  categories
employed  in  EPA's MOBILE4 computer model.   The short test
results from  both the  Michigan AET test  and the CTP's  own
short tests are  analyzed and  compared to the FTP results.

4.2.  FTP  Results


     4.2.1.  Sample  Description


     All but two  of  the 239  vehicles in the base CTP  sample
underwent an  FTP on Indolene in  the as-received condition.
For the two  exceptions  (vehicles  257  and 337),  post-repair
FTP  results  were  substituted  for the  missing  as-received
tests,  based  on the  conclusion that the repairs performed
were likely to  have  had negligible impact on each vehicle's
emissions.  (As discussed in  Section 3.1,  such substitutions
were unavailable or unjustified for  six  additional vehicles,
accounting  for  the  difference  between  the 245  vehicles
recruited for the program, and the 239 vehicles used  for  the
bulk of  this  analysis.)

     4.2.2.  Pass/Fail  Results at  Certification Standards

     Of the  239 vehicles in  the  base sample, 206  (or 86%)
exceeded their certification  standards for HC,  CO,  or both.
The remaining 33 vehicles thus represent  errors of commission
(Ec) by  the  original  AET short  test.   As Figure  6   shows,
combined HC-CO  FTP failures  dominated  the CTP sample.  Only
in the  later  model  years were  there significant numbers  of
CO-only   failures,   and  few  HC-only   failures  occurred,
regardless of  model year.12
                             24

-------
                           FIGURE  6

         As-Received FTP Failure Type hv Quota  Group
           F\ 81-82   GARB 81-82 GARB 83-86   Fl 83-86
ALL
                NO FAIL  DHC-ONLY HHC&CO  •COONLY
     4.2.3.  Excess Emissions Analysis

     A more  illuminating  view  of FTP  emissions  in the  CTP
sample arises from analyzing "excess" emissions,  which  is  the
difference between each vehicle's   FTP performance and  its
emission  standard,  for a  given pollutant.   Figures 7-10
scatter  plot the  excess  HC and CO emissions  of  the base
sample,  stratified by  quota group.   Identical scales have
been  used in the  figures  to ease  comparisons  between  the
groups;  note, however,  that  the  resolution does not  permit
display of the individual vehicles concentrated very close to
the origin.

     The  total  HC  excess  emissions  in the 239-vehicle base
sample was 334 g, or 1.4 g per average  vehicle.   The total CO
excess was 5155 g,  or  21.6 g  average excess.      As  the
scatter  plots show,  however,  the  fleet  displayed a broad
spread of excess emission  values  for  both HC  and  CO.   The
total excess values include  57  vehicles  that  passed their HC
standard  and 41  that  passed their CO standard,  and thus
contributed negative  "excess" emissions.   Cleanest  relative
to its certification standards for both HC  and CO was vehicle
249,  a carbureted  1982  Ford LOT, which  was 0.8 g  below  its
1.7 g HC  standard and  12.9  g below its  18 g  CO standard.
                             25

-------
                         FIGURE 7
            As—Received  FTP  Excess  Emissions
              Fuel-Tn-ipr!l-p»H  1981-82 Vehicles
  00
(g/mi)
140
120
100
 80
 60
 40
 20
  0
-20





.....








....

•••
•



....
1

f
....

B
•



•
.....


.....
•








.
. •























•














i
i
i
i
i
(18.7, 55
i

i
i
i
i
i
i
i
i


.6)
-».




           - 1
                    3456
                      HC  (g/mi)
8
10
                         FIGURE 8
            As-Received FTP Excess  Emissions:
               Carbureted  1981-82 Vehicles
  00
(g/ml)
1 4fi -
120 -
1 c\ n _
1 UU •
80 •
6rt
u •
40 •
2 n
u •
0 -


.... J
.....
.....
4
.....
a.


i i
• i
....1....C...J
1 1
1 1
1 1
1 1
1 1
1 1
	 T 	 • 	
1 •! '
i • ' i
....;.&..•,
_ i • • i
...VrfvL.-.,
-.-A,-? r
!!•»•* T~"T~"1
1 1
1 1 1


....
	
.-...
1 r-
....


.....
.....
.....
.....


.....
	
	
....


....
....
....
.....


1
1
L... J-...
1
1
1
1
1
.........
1
1
— -•!— -
1


.....
	
—
.....

                           3456
                             HC (g/mi)
                                               1 0
                            26

-------
                         FIGURE  9
            As-Received FTP Excess Emissions:
               Carbureted 1983-86
  CO
(g/mi)
1 40 -.
120-
1 n n -
I U U
8n .
u
6n .
u •
A n ,
2n .

0 -




.....


•i
... •;
•i







' V
i
i





.'••
.••
..•L
•


.....
.....



•


r----i




....
....
•1 1
i



L-...

L. ...




r----


•1
i

•






t









.....



r----i













....
....




           -1
3456
  HC  (g/mi)
10
                        FIGURE 10
            As-Received FTP Excess Emissions:
              Fuel-Injected 1983-86 Vehicles
        140 -r

        120 -•

        100 -•
  CO
(g/mi)
       -20
.....
•
—

'
1
....i....
1
1
1
1
1
•— {-—
•" "*•!"""
• •
i
L... J
	
(.---•-
";-•
-----
....
.. ..
r —
u
....
.....
	
.....
.....
.....

.....
----1
....
....
....
....
•
•
	
	
-----
....
.% ..
....
.....
.....
	
	
....
           -1
3456
  HC  (g/mi)
10
                            27

-------
     On  the  opposite extreme  was  vehicle 202,  another  1982
Ford,  which  was  a sizable   18.7  g  above  its 0.41  g  HC
standard.   (This fuel-injected  LDV  was  also the only  super
emitter  in the  as-received CTP sample;  see section  x.x.x
below) .   On  its  own,  #202 accounted for  almost  6% of  the
total HC excess emissions in the 239-vehicle sample.   Highest
in  the ranking  of  excess CO  emissions   was  vehicle 605,  a
fuel-injected 1982 Nissan LDV, which  was  139  g  above its  3.4
g standard, or  close to 3% of the  total excess CO  from  the
sample.

     When  the   CTP  vehicles  are  ranked   by  excess   FTP
emissions,  and the excess is accumulated  as a percent of  the
total  excess  for  a  given pollutant,  Figure 11 results.   The
curves for HC and CO were determined independently.   For  each
pollutant,  the dirtiest  40% of the vehicles accounts for  90%
of  the excess from  the  sample.   The  leveling  of the  curve
above  100% and eventual  decline  to 100%  reflects the portion
of the ranked sample that was  very close  to,  and then below,
the certification standards.
                          FIGURE 11

       Excess  Emissions  in the Ranked As—Received Fleet
           120
           100 ••
  % of Total
   Excess
              0  10  20  30 40  50  60  70  80  90  100

                         % of Ranked Base Sample
     As  the  scatter  plots and  judgment  would  indicate,
considerable overlap exists  between  the vehicles with  large
excesses for HC and  CO.   The ninety vehicles with the  worst
                             28

-------
HC excess  emissions together  account  for 90%  of  the fleet
excess HC; the  identical  90  vehicles  account for 83% of the
fleet excess  CO.

     4.2.4.  Analysis Using MOBILE4 Emitter Categories

     The  EPA MOBILE4 emissions  model  classifies light-duty
vehicles   into  four  categories,  from  lowest  to  highest
emitters:    passing,  marginal  failure,   high  failure  and
super.emitter.   The passing and marginal emitters, when taken
together  as  a  single  group,  are frequently  referred to as
"normal"  emitters.13   Boundaries between  emitter categories
are defined separately for HC and CO.   The  pollutant  with the
highest  emitter  category determines  the  category  of  the
vehicle; thus,  a  vehicle  that  is a  high emitter on HC and a
marginal emitter on CO is  referred to as a high emitter.

     Table 7 shows the upper bounds  of  the  emitter categories
for  HC  and  CO.    Note  that  the  boundaries  between  the
marginal-  and  high-emitter  categories  are  technology  and
model-year-group  specific.14   Passing emitters  have HC and CO
values  that  are  each below their  respective certification
standards.   A  vehicle with  either  an HC reading  or  a CO
reading  exceeding the  respective  upper bound  of  the high-
emitter category is classified  as a super emitter.
                           TABLE 7

              Bounds  of  the MQBILE4 Emitter Categories

Technology Group
Carbureted 1981-82
Fuel-Injected 1981-82
Carbureted 1983+
Fuel-Injected 1983+
HC
Pass Marginal High
cert std 1.175 10.0
cert std 0.725 10.0
cert std 0.815 10.0
cert std 0.965 10.0
CO
Pass Marginal High
cert std 17.411 150
cert std 10.499 150
cert std 10.398 150
cert std 10.558 150
     Figure 12 shows the breakdown of the base CTP sample by
the MOBILE4 approach,  giving the percent  of  the sample in the
four quota groups  (and the entire sample)  that fell in each
emitter category.15  Thus,  approximately  15% of the CTP fleet
were passing  emitters, 25% were  marginals,  60%  were highs,
and  less  than 1%  —   that  is,  one  vehicle — was  a super
emitter.    In  simpler  terms,   60%  of  the  sample  showed
significant HC or  CO  FTP noncompliance,  while the remaining
40% could be considered "normal" emitters.

     The  figure  shows that the  fuel-injected 1983-86 group
was  distinctive  for  containing a disproportionately large
percentage  of marginal  emitters  and  fewer  high emitters.
                             29

-------
 Carbureted  vehicles  were  more  likely  to  be  high  emitters than
 were  their  fuel-injected  counterparts of the  same model year.
 Similarly,  vehicles  from  the earlier  model  year  grouping were
 more  likely to be  high emitters than  their counterparts in
 the later model-year grouping.
                          FIGURE 12
            As-Rereived FTP Profile hv Quota
                    and Emissions Cateaorv
           PI  81-82  GARB 81-82 GARB 83-86  Fl  83-86
                           ALL
                SUPER
HIGH
D MARGINAL  • PASS
     Table  8  shows the mean  excess  HC and CO  emissions for
all 206 FTP  failures (including the one  super  emitter), and
then isolates the  mean  excess HC and CO  for  the  62 marginal
emitters and  the  143 high emitters.   Thus, for example, the
mean excess for fuel-injected 1983-86  high-emitters was 2.21
g/mi HC and 35.4 g/mi CO.  The mean excesses for the marginal
emitters are quite small, implying that most of the marginals
were  in the  lower part  of  the  marginal range.   The  mean
excess  HC  for  the  carbureted  1983-86  group  is  actually
negative,  indicating that the CO  values for those vehicles
were driving  the  classification.   One implication  of  these
low values  for  the marginal  emitters  is that  the emissions
repair benefit to  be  derived from them  is  quite  small.

     The relatively  high values  in the  "all  fails" category
show the effects of  the  large number of high  emitters in the
sample.  Thus, for example, the  2.07 g/mi HC  excess and 30.5
g/mi  CO excess  for the  carbureted  1981-82  vehicles  are
                              30

-------
several times greater than the  standards applicable to those
vehicles.   The mean  excess for  all failures  in  the fuel-
injected  1981-82  group  is sensitive  to  the presence of the
single super emitter:   the HC value would drop from 2.77 g/mi
to 1.93 g/mi if the super-emitter  were to be removed.

     Because of the large  impact  of  the  single  super  emitter
and  the  low  mean excess  displayed  by  the  marginals,  the
analysis  below will  in most  instances  focus  on  the high-
emitter category.

     The  mean  excess  HC values for the high  emitters in the
four quota  groups were comparable to the  mean for all high
emitters  of 2.22  g/mi,  although the  carbureted  1983-86 value
was  somewhat  lower.   For  mean excess CO,  the  fuel-injected
1981-82 category  was  considerably  dirtier than  the mean, and
the carbureted 1983-86 group  was once  again somewhat cleaner.
                           TABLE 8

       Mean Excess Emissions  of  Failures bv Quota Group

Quota Group
Fuel-Injected 81-82
Carbureted 81-82
Carbureted 83-86
Fuel-Injected 83-86
ALL
Excess HC (g/ml)
All Marginal High
Fails
2.77 0.14 2.41
2.07 0.19 2.51
1.31 -0.01 1.67
1.24 0.07 2.21
1.66 0.09 2.22
Excess CO (g/mi)
All Marginal High
Fails
42.4 1.4 52.5
30.5 3.5 36.8
19.8 2.3 24.6
20.2 2.2 35.4
25.4 2.4 35.2
     The  MOBILE4  emitter  categories  were  also  used  to
generate  Figure  13,  which  shows  the  proportion  of  each
manufacturer's  fleet that fell  in  each category.   Although
the  data  in Figure  13 have not  been  adjusted for  mileage,
model  year,  or technology  factors,   such  factors  do  not
necessarily  explain the  differences  between  manufacturers.
Chrysler,  for  example,  has  the  highest percentage of  its
fleet  (80%)  in  the  high-emitter  category,  yet  it  was  the
manufacturer with  the  highest percentage of its  fleet  under
50,000  miles   (see  Figure  4).    The  Chrysler  fleet  also
contained an above-average percentage  of late-model vehicles,
and no fuel-injected 1981-82 vehicles  (Figure 1).   Toyota,  on
the other hand,  shows  the smallest  percentage  of  vehicles  in
the high-emitter category  (as well  as  the highest percentage
of "passes,") yet  its  fleet  had  the highest  average  odometer
reading of any manufacturer in the sample (Figure 5).
                             31

-------
                          FIGURE 13
                       PTP profile hy Manufacturer
                    and Emi 53 q •! rm «;  Catecrarv
          GM  PORD NISS TOVT CHRY AMC VW MAZ SUBA MITS K>D ALL
                 SUPER
HIGH
MARGINAL D PASS
     4.2.5.  Correlation of Excess  Emissions, Emitter
          Category, and Odometer

     Linear regressions on  odometer against  excess  emissions
in the as-received fleet  yielded essentially no  correlation.
The degree to which dirty cars  are dirty  seems not  to  depend
much on  their  age.  Figure  14,  for  example, shows the  wide
scatter  when excess  HC is  plotted against odometer for  the
carbureted 1981-82 group; plots for the other groups and for
CO are  similar.    The  R-squared values for  the  linear  fits
performed on each quota group  ranged  from 0.2%  to 5.2%  for
both HC  and CO,  with similar poor correlation shown for  the
fleet as a whole.

     On the other hand, the mean mileages  increase  across the
MOBILE4   pass,   marginal,   and high   categories  for   the
technology  and  model  year  groups with  significant   sample
sizes  in  each  emitter  category  (Figure  15)  .   This  is
consistent  with  an  hypothesis   that  greater  numbers  of
vehicles  move  into  the  failing  emitter  categories  with
increased mileage,  although the actual  excess  emissions  of
vehicles  within  a category  may not  be  linearly related  to
mileage.
                              32

-------
HC Excess
 (g/mi)
                         FIGURE  14

         Mileaye vs. Excess HC  in  1-hp Carbureted
                    1QB1-R2 Quota ftronn
 9
 8
 7
 6
 5
 4
 3
 2
 1
 0
- 1




















1
1
1
1
1


1
r\v

r T
hB
•
* • B
	 •.-v-.l"ii

	 1L "•!.

IB
	 BJ 	

1
1
1
1

_ 1
• 1
• t
1 •
"^ **
g 1
T
I 1
B •••.!•
..!L ;'..-'..
i
f L .•.•!.....

i

i
i
•
i
i

i
•
i
•
•
..•.f .....

•
i
•
i


i

•
i
i
i
•

•
i
i
i



•
i
•
i


i





















                    50
                    100
     150
     200
   250
                         Mileage (1000's of Miles)
                         FIGURE 15
   Mean  Mileage of Model  Year and Fuel—Metering  Groups
                bv MOBILE4  Emitter Catecrorv
 Mean
Mileage
 (K)
               Fl
               Garb
                          PASS
81-82
                          MARG
83-86
          HIGH
ALL
                             33

-------
4.3.  Michigan AET
     4.3.1.  Profile of the Base Sample

     As  mentioned  in  Section  2.1  above,  vehicles  were
recruited  for  the Cooperative  Test Program  on  the basis  of
failing  their  official  I/M  short  test  under  the Michigan
Automobile  Emission  Testing  (AET)  program.   The  AET test
makes  a  pass/fail determination  at idle,  following a  short
period of 2500rpm preconditioning.

     A  scatter  plot  of  the HC  and  CO scores  on  the AET
screening test  for the 239-vehicle  sample  appears in Figure
16.  The major  divisions  in the figure represent multiples  of
the  207 (b)   cutpoints.    Values   that  fall  on  the  10%   CO
horizontal  and the  2000ppm HC vertical are  vehicles  whose
emissions exceeded the maximum ends of  the analyzer ranges.
Vehicles that passed the  AET  cutpoints  of 1.2% CO and 220ppm
HC were not recruited  for  the program, which  accounts for the
empty range in the lower left corner of  the  plot.
                          FIGURE 16
         Results of the I/M Screening Test  (AET
                     for the Base Sample
       10.8
        9.6	-•-

        8.4	*
   00
        3.6 •-

        2.
7.2 -------*-B"-.••*-----fc----j--.--fc----j-.---*-----


48i—•":\v--;—.--—
2.4 • ••~s"jri-ii-B»a	••••ir'i	"•	'	
     J&t •'   •.      *
1<2 ...^i'-H--.....,—'.^. —^...Jr--...^	T
                220  440  660  880  1100  1320  1540  1760 1980 2200

                                HC  (ppm)
                              34

-------
     Noteworthy in Figure 16 is the large number of vehicles
that failed one pollutant  (particularly  in  the first "cell"
above  the  outpoint),   but  passed  the other pollutant.
Similarly,  note the relatively small number of vehicles that
failed both  pollutants  in  the area  just  in excess  of  the
outpoints.    On the other hand,  30% of  the  base  sample  (74
vehicles)  had either   HC  emissions  above  SOOppm  or  CO
emissions  above  4%.    Twenty-two  of these vehicles  with
extremely  high  short  test  scores  were   single-pollutant
failures,  including the  three 2000ppm  HC  vehicles  at  the
extreme bottom right of  Figure  16.

     The  distribution  by  failed  pollutant is  reasonably
consistent  with an  hypothesis that  some   emission  control
failures (or AET test  irregularities)  cause  high idle HC,  and
others cause  high idle  CO, but  few cause   both high HC  and
high  CO.    The vehicles  seen  to  have  high  HC and  CO  are
approximately  accountable in  terms of  random simultaneous
occurrence  of two  failures.


     4.3.2.  Stratifications by Manufacturer and Quota Group

     The  frequent  pattern  of  AET  failure  for  a  single
pollutant was  maintained across  quota  groups as well as  for
the entire base sample  (Figure 17).  For each group,  the  sum
of  HC-only  and CO-only failures  exceeded  the  number  of
combined HC/CO  failures.   In the  two 1983-86 groups  and  the
fleet as a  whole,  the  number  of  HC-only failures  was itself
almost equivalent  to the number of combined  HC+CO failures.
                          FIGURE  17
   V
   e
   h
   e
   s
          Fl 81-82 CARB 81- GARB 83- Fl 83-86
                   82      86
ALL
                             35

-------
     These  observations  on the Michigan AE.T results contrast
with the  results on  the as-received FTP  test,  presented in
Figure 6.   There, HC-only  failures were the rarity across all
quota groups;  CO-only failures occurred at a somewhat higher
rate in  the  1983-86  model year  groups,   and  combined HC/CO
failures dominated  the sample.  This trend  is  not simply an
artifact of the cutpoints  used to define FTP  failure;  that
is, there is  not  a  disproportionate  number of combined HC+CO
failures right above  the standards, in the "marginal"  failure
category.   In  fact,  of the  144  highest emitters in the sample
(the  143  highs, plus  the  one  super  emitter),  141  were
combined HC+CO failures.

     In further analyzing  the types  of  AET failure,  what was
true for the quota groups was not true for the manufacturers;
different manufacturers  showed different  proportions  of HC-
only,  CO-only and  combined HC/CO failures.   In  Figure 18,
each manufacturer's CTP sample is  divided according  to the
number  of   vehicles  that  fell in each of  the AET  failure
types.   Better than 70% of the  Honda and Toyota vehicles were
HC-only failures, as  were  57% of  the  GM vehicles.   Together,
these three manufacturers  accounted for  almost three-fifths
of the  HC-only failures in  the base sample.
                          FIGURE 18

              AET Failure Types bv Manufacturer
          GM  FORD  MISS  TOYT  CHRY  AMC   VW  MAZ  SUBA  MITS HOND
                        HC-Only  DHC&CO • CO-Only
                              36

-------
     Chrysler  and Ford,  on the  other hand,  showed higher
proportions of  their  samples (60% and 47%,  respectively)  in
the combined HC/CO  failure category.   Subaru and  Volkswagen
had close  to  half of  their samples  failing CO only,  while
Honda had no CO-only  failures at all.


     4.3.3.  Stratification by  Mileage and  Vehicle Type

     Two   other  brief  analyses  were   performed   on  the
relationship between the  AET failure type and  basic  vehicle
characteristics.   Analysis' of  the  numerical  relationship
between vehicle mileage and the AET scores showed essentially
no useful correlation.   Finally,  no qualitative  difference in
the AET  failure types was  noticed when  the base  sample  was
disaggregated by vehicle  type.

4.4.  Correlation Between  the FTP  and  Michigan AET Results


     4.4.1.  Introduction

     Section  4.2.2  mentioned  that 206 of the  239  vehicles
(86%)  in  the  base CTP sample failed their as-received  FTP
tests  at  certification  standards.   Because  all vehicles
recruited  for  the CTP  had failed  their initial Michigan  AET
test, the remaining 33 vehicles that passed their as-received
FTP are considered errors of commission (Ec)  by  the AET  short
test.  This section  examines more  closely the  relationships
between  the AET  test  results  and  the as-received FTP  test
results for both the  FTP  failures  and  the Ec  vehicles.


     4.4.2.   Excess  Emissions  and Emitter Categories of  the
          AET Failure Types

     Figure 19  stratifies  the  sample by the  type  of  AET
failure (HC-only, CO-only, or combined HC/CO) and the MOBILE4
emitter  categories.    As  shown in  the figure,  the  greatest
number of vehicles fell in the high emitter category for each
of  the  types  of  AET  failure.   Of the vehicles that failed
their AET  tests for both  pollutants,  for example, 61  of 83
(or 73%)  were high emitters  — making the  HC+CO failure type
the most productive at identifying vehicles  with  significant
FTP nonconformity.  Next-most productive  were the  CO-only  AET
failures, where  62% were  FTP high emitters.   Least  effective
were the  HC-only failures.   While these  failures were  the
most prevalent  (38%  of  the sample)  , it  was  the  only  AET
failure type  where more  normal FTP emitters were  identified
than high emitters.

     Figure 19  also  shows a correlation between the  type of
AET failure and whether  or  not the  vehicle was an error of
commission.  Of  the 33 Ec  vehicles  in the  sample,  only  three
                             37

-------
 failed  their  AET short  test  for both  HC and  CO.   Of  the
 remainder, Ec's among the HC-only failures outnumber those in
 the CO-only failures by two to one.
                      FIGURE  19

Breakdown nf FTP Emitter Types bv Typp of
                                                  failure
                                                 BRASS

                                                 DMARG

                                                 SHIGH

                                                 • SUPE
            HC-only
                      BOTH

                 AET Failure Type
CO-Only
     As  shown  in Table  9,  vehicles  in the  high-emitter
category accounted for  93%  of the excess  HC  and 96% of  the
excess CO  from among the  206 failures  in the  base  sample.
Each type  of AET failure contributed significantly to that
total, but the  combined HC+CO  AET   failures  accounted  for
nearly half of the total excess (48.8% of the  HC  and 48.4% of
the  CO) .    This reflects both the  larger number  of HC+CO
failures among the  high  emitters,  as well  as  a  generally
larger mean excess  per vehicle  —  2.73 g/mi   for HC,   and
41.6 g/mi CO.

     On a per-vehicle basis, the CO-only AET  failures matched
the  HC+CO  failures in  their ability  to  identify excess  CO
emissions among the high emitters.  The CO-only failures even
identified  a   respectable  percentage  of  the  excess   HC
emissions,  with a mean  excess  HC  of  1.7 g/mi, not  far below
the 2.0 g/mi mean from the  HC-only AET failures.   In spite of
the high percentage of error of commission vehicles  among  the
HC-only  failures,  the group  nevertheless identifies nearly
one  quarter of  the  HC  excess emissions among  the failed
vehicles  in the sample,  as well as 14%  of the excess CO.
                             38

-------
                           TABLE  9
       ionship  Between AET Failure Type and Excess  Emrss i nn,«j
                for the  Hi ah Emitter
AET Failure
Type
HC-oniy
HCandCO
CO-only
All Fail Types
Mean Excess Emissions
HC (g/mi)
1.97
2.73
1.69
2.22
CO (g/mi)
17.8
41.6
43.7
35.2
% of Failed-Vehicle Excess
HC
24.3%
48.8%
19.9%
93.0%
CO
14.3%
48.4%
33.4%
96.1%
     Because  no  vehicles were  recruited into the  CTP  with
short test scores  lower  than the 207 (b)  cutpoints,  the data
cannot  be used  to  evaluate the  effectiveness  of  tighter
standards on the  excess emissions identified by the AET test.
However,  because  so  little  of  the  excess  apparently arises
from the  marginal emitters  (and  by  definition,  none arises
from the  passing  vehicles),  it  is  worth asking  what impact
raising the outpoints might  have had on  the number of normal
emitters  identified by the test.   Reducing the number of Ec
vehicles,  or  perhaps   even   marginal  emitters,   while
maintaining  acceptable  rates  for  identification  of  high
emitters,  would presumably be a  desirable goal.

     A scatter plot of the AET scores for the normal emitters
appears in Figure 20, and Figure 21  shows the analogous data
for  the  high emitters.   The   format  is  similar  to  that
employed  earlier  in  Figure 16,  except  that  each  of  the axes
has  been  truncated  at  four  times the applicable  207 (b)
standard.   Clearly,  many  of  the normal emitters  fall  in the
zone between  the  100%  and 200% of the HC  and  CO standards.
Not  surprisingly, however,  many  of the  high emitters  do
precisely  the same  thing.   Together,  the plots show  that
there  is   no  natural  break  in  either  the  HC  or  the  CO
distribution  that would  suggest short  test  standards  for
excluding  large   numbers  of  the normal  emitters  from  AET
failure,  without  inappropriately converting many  of  the high
emitters to AET passes.
                             39

-------
                         FIGURE 20
         AETScores of the  Normal Emitter1? T.vinrr
00

.6 •
Nssr tiie 207 cb) sta.nda.rd.g


...., ...".....
»• , • ," •«
1
1

L.....J.......
L"
;-• .• "'B .:









1
0 220 440 660 880
HC (ppm)
FIGURE 21
AET Srores of the Hiah Emitters Lvina

Near the 207 (b) standards.

00
3/5
.b •
2 A
.*» •
1 9 .

c
m •
g
*
•1 1

22

•i
• .
• .1 *
_ • •! • •
•*B| •• •
!0 44



• g
• • ~
r • -
10 66



•
10 88
                             HC (ppm)
                            40

-------
4.5.  :,ahQratory c;hnrc Tear. Reaulr.g


     4.5.1.  Sample Description

     For a variety of reasons, not all of the 239 vehicles in
the  base  CTP sample  completed short cycle testing  with the
Basic  I/M Test  Procedure  on  their  tank fuel.   Six of the
vehicles  were  found to  have  insufficient tank fuel  for the
3IT? after the procedure  was underway;  consistent  with the
program plan,  the  procedure was completed on  these  vehicles
with commercial fuel.

     Equipment  and vehicle problems,  as well as  deviations
from the  program plan,  led to omitted core  sampling periods
during parts of  the BITP on seventeen  vehicles.   For all but
eight of  the vehicles,  as-received  indolene  BITP  testing was
available  for  the missing  tank/commercial modes,  and the
indolene values have been substituted for the purposes of the
analysis  below.   Only  one  vehicle  from the  original  239
(vehicle  104)  was  missing  enough data to be eliminated from
the short cycle analysis entirely.

     In summary, tank fuel  values were available  for all but
a  few  vehicles,  and  all but  a  few  modes.   Where they were
not, commercial  fuel  values were used; where  commercial fuel
values were  unavailable,  Indolene values were used.   Sample
sizes that are less than 239 for particular modes  result when
no substitute data (regardless of fuel type) were available.


     4.5.2.  Second-Chance Failure  Rates of the  As-Received
           Base Sample

     Recall  from  Section  2.4.2 that   the  Basic  I/M  Test
Procedure  includes   seven "core"   sampling  periods,   each
consisting of  a  30-second  first-idle   mode,  a  30-second
2500rpm mode,  and  a  120-second second-idle mode.   Thus,  the
core sampling periods  closely resemble  the EPA  Two-Speed Idle
Test.  However,  because all CTP vehicles had already  failed
one short  test  (the AET  idle  test) prior to  recruitment,  the
core  sampling  periods  in  the  BITP  procedure  represent
"second-chance" short  tests.

     Figure  22  analyzes the  as-received failure  rates  for
fourteen of the idle modes  found in  the  BITP, grouped as the
paired  first-  and  second-idle  readings  of  the  seven  core
sampling periods.   The 30-second point  in the modes were used
in all cases,  and failure  type  was  determined by comparing
those  readings  to the  207(b)   HC  and  CO  cutpoints.    The.
numerical data on  which  the figure  is  based are provided  in
Appendix  C.
                             41

-------
     The  order  of the  modes in  the  figure  duplicates the
order  in  which  they  were performed during  testing.   within
each pair  in  the figure, the only difference between the two
idles  is  an intervening  2500rpm  mode,  with the exception of
the restart  (RS)  pair,  which had both an intervening 2500rpm
and a  keyoff/restart.   Between pairs,  the difference between
the  idles  is the  intervening  conditioning  (shown  in the
bottom row of the table) that preceded the second pair.16
                          FIGURE 22

             Failure Rates of the BITP  Idle  Modes
                   Extended
                   2500rpm
                   I Ext Idle
     Each stacked  column in Figure  22  gives the percent  of
the base sample that failed the given mode for CO-only (solid
portion),  HC-only (striped),  and HC+CO (crosshatched).     The
total  height  of the  column  therefore  gives  the  overall
failure  rate   for   the  mode.    The  base  operation   and
conditioning for the modes are  provided  below the horizontal
axis.  Thus, for example, 89%  of  the sample  failed  the first
idle mode  of the BITP   (CS-03) ,  which  was an  unconditioned
idle following a minimum one-hour soak.   Following 30 seconds
of 2500rpm operation (CS-04),  the failure rate dropped to 58%
during the  second idle  (CS-05) .
                              42

-------
     Figure  22  shows that  considerable variability existed
between the original Michigan Auto Exhaust Test  results that
were the basis for recruitment into the CTP  and the  simulated
field short tests of the Basic  I/M Test  Procedure.   With the
exception of the very first mode  of  the procedure  (following
a "cold"  start),  the idle  failure  rates ranged between 40%
and  60%.   This  indicates  that a  second-chance short test,
almost regardless of how poorly it is  performed, will  reduce
the idle-test failure rate  substantially.

     By examining the  adjacent  idle  modes in  Figure 22, one
sees that  the consistent  impact  of  2500rpm operation  (both
30-second and 180-second)  was  to lower the idle failure rate.
Focusing just on the core  sampling periods,  the  failure  rate
for  the  second  idle of each  core sampling period was  lower
than the rate for the  first,  attributable to the intervening
30-second  2500rpm  mode.    The magnitude of  the  reduction
between  idles  was  greatest  when  the  initial  operating
condition  was  furthest  from  ideal:    the  31-point drop
following the  soak  at the  very  beginning of  the procedure,
and an eight-point drop following the 20-minute extended  idle
in  the  middle  of the  procedure.   Even in the conditions
considered more ideal, however, failure  rates  for the  second
idle were three  to four points  lower  than the first.

     The impact of loaded operation was to reduce the failure
rate as well.  In fact, extended 2500rpm and loaded  operation
were apparently responsible for achieving the  lowest  failure
rates among  the  14  modes  in  the procedure;  these  are the
values in the low 40's  for  XL-05,  XL-10, and  RS-02.17

     On the  other hand, the failure  rate  increased  following
extended idles:   the ten-minute idle  before  CS-12 generated  a
ten-point  rise,  and the  20-minute idle at  XI-02  led to an
eighteen-point  rise.     In   each  of   these  cases,  the
accumulation  of  2500rpm operation  (including  some  extended
2500rpm modes)  in succeeding modes  eventually  reversed the
effects of the long  idles.

     Not  surprisingly,  the  idle  immediately  following the
initial soak period  showed  the  highest  failure rates of  all.
Interestingly, only  extended 2500rpm and  idle  operation  were
then necessary  to bring the  failure  rate down within a few
points  of  the rates achieved by  extended  loaded operation
later in the  procedure.

     Considering the effect of 2500rpm operation elsewhere in
the procedure, the apparent rise  from  the combined effect of
2500rpm and restart  operation  between the last  two idle modes
(RS-02  and  RS-05)   implies that  the  restart  alone  might
increase the  failure rate  somewhat.

     The Michigan  AET test and the BITP  idle modes  showed
differences in the types of failure as  well  as  in the overall
                             43

-------
failures  rates.   Recall  from  Section 4.3.1  that  single-
pollutant  failures  were  frequent  in  the AET  results.    As
shown in Figure 22,  however, combined HC+CO failures were the
rule  in the  BITP  idle modes,  outnumbering the  other  two
failure types in  every  one  of  the  modes.   In  fact,  the HC+CO
failures  outnumbered  the  sum  of the HC-only and  CO-only
failures in all of the  idle  modes  except two:   CS-05  and XI-
02.   These  two  modes represented  non-ideal test  conditions:
one soon  after  the  soak at  the beginning of the  procedure,
and the other following a  20-minute idle.

     Of the  three failure types,  the  CO-only group  had the
most  consistent  failure  rates across  the  different  idle
modes:   all  fell in the  range from  7%  to  13%.   The  four
highest  CO-only  rates all  came  at  the  beginning  of  the
procedure, before the  first extended  2500rpm operation had
occurred.   Past  that point, the CO-only  failure rate  never
exceeded 10% of the  sample.  The HC-only failure rate,  on the
other hand,  was  apparently more  sensitive to  the type  of
prior operation.    All of the  modes  preceded  by extended
2500rpm or loaded operation  had  HC-only failure  rates  of 7%-
9%; all the modes that  immediately followed extended  idle or
a soak had failure rates twice  that high.

     The above observations  support  the hypothesis  that  many
of the  vehicles   recruited  for the  CTP program were  poorly
preconditioned in their original Michigan AET test.   Roughly
one-half of  the  failures  might  have been  avoided by  better
preconditioning.

     In each idle mode in Figure 22 where  the  vehicles appear
to have received  adequate  preconditioning (e.g., CS-10,  XL-
02, XI-10),  approximately  20% to 25% of the  sample  failed the
mode for both HC  and CO.   The possibility exists that these
vehicles also  failed the AET  test for both  HC  and CO,  and
that they could represent  a consistent  set  of  failures across
the various short test  conditions  encountered in the  initial
test  (AET) and  second chance  tests.   Table 10  compares  the
AET failure types to the  failure types on on the first  idle
of the  extended  loaded sequence   (XI-02) .   The  data  show
considerable migration  among the  failure  types between  the
initial  test and this second-chance test.  Of  the  83 vehicles
that failed both HC and CO on the AET test,  for  example,  only
39 were of the  same  failure type on the second-chance test;
38 of the  83  changed from an  HC+CO  failure to  a pass.   The
fact that the second-chance test shows the 25% HC+CO  failure
rate is  due largely to the migration into that category of 17
vehicles from the  AET CO-only category.

     Given the earlier  analysis  of  the AET HC-only  failures,
it is perhaps  less  surprising  that  more  than two-thirds  of
these vehicles passed the  second-chance test.  Almost  half of
                              44

-------
the AET  CO-only  failures  passed the  second-chance test  as
well.
                           TABLE  10
rnmparison of Failure Tvpgs
                                      the AET
                                                   and the
      First Idle-Neutral  of  the  Extended Loaded Seouenre>
Second-Chance
Failure Type
HC-Only
Both
CO-Only
Pass
Total
AET Failure Type
HC-Only
15
8
2
66
91
Both
3
39
3
38
83
CO-Only
1
1 7
16
31
65
Total
1 9
64
21
135
239
     4.5.3.  Idle-Mode Short Test Variability  Due to Fuels

     One-hundred,  ninety-six  of the  vehicles  in  the  base
sample underwent as-received Extended  Loaded  short  cycles  on
both Indolene and tank  fuels.18  The HC and CO values  for the
second  idle  (mode  5)   of  these sequences  were compared  to
determine the effect of fuel type.   For 169 of these (or 86%)
there was no change in  the 207 (b) pass/fail status  for either
HC or  CO between the tests  on the different  fuels;  this  is
shown by  the  sum of the bold-print numbers in  Table  11.   Of
the  27   remaining  vehicles,   there  was  an  essentially
negligible  trend to more frequently pass the  tank  fuel test
(a  net  increase of three  failures occurred on  Indolene) .
Although  there  were six more  HC failures on  tank  fuel than
Indolene, most  of  these vehicles were already consistent  CO
failures, so their overall pass/fail status did not change.
                           TABLE  11

  Changes in Pass/Fail Status  for HC and CO Between Tank and
                      Indolene Idle Tests

CO Status
P-P
P-F
F-P
F-F
Total
HC
P-P
1 08
5
0
1 2
125
Pass/Fail
P-F
2
3
0
2
7
Status
F-P
3
0
4
6
13
(Tank-lndolene)
F-F Total
1 1
0
2
38
51
124
8
6
58
196
                              45

-------
     Were there to be a fuels-related impact  on  the  idle test
scores,  the  volatility  difference  between  the  tank  and
Indolene fuels  would be the most obvious explanation.  Fuel
RVP levels were available for  189 of the  196 tank fuel tests
used  in  the  above  comparison.   The  range in  RVPs of this
sample was 5.0  to  15.6,  with  a mean of 11.7.19  One-third of
the RVP  levels  were  above 12.0.  The  27-vehicle sample that
showed variable pass/fail  results  on  HC,  CO  or  both,  had
RVP's in the  range 9.3'to  15.6,  and also had  a mean RVP of
11.7.    Thus  no  volatility-related  fuel  variability  was
evident in the  first-idle neutral of the  "ideal" short cycle
from the  as-received  laboratory testing.

     The group of 27  vehicles  that change  pass/fail  status on
either HC or  CO between the Indolene and  tank-fuel  tests was
divided  into three  roughly  equal-sized  groups.   Eight
vehicles  had  small score  differences  (changes  in  HC of less
than SOppm  and  in CO of less  than  1.0%)  that nevertheless
overlap the 207 (b) cutpoint.   Eleven had  extreme differences
(changes  in HC  of more than  400  ppm or  in CO  of more than
4%).    The remaining eight vehicles  had moderate  changes.
For  these  groups,   there was  no  correlation  between  the
magnitude and the sign  of  the  change in scores   (i.e.,  no
trends for  one  fuel  being  more failure-prone  at  idle) ,  or
between the magnitude of the change in scores and the RVP of
the tank-fuel  test.


     4.5.4.    Correlating the Laboratory Short Tests with the
          MOBILE4 Emitter Categories

     Based on the results  of  the previous  subsection,  the
question  arises  whether  there  are  patterns  to  the  FTP
emissions  of  the  vehicles  that changed pass/fail  status
between the  Michigan  AET test  and  the various short  cycles in
the Basic I/M  Test Procedure.   Figure 19 begins  this analysis
with the  laboratory short test results for the  vehicles that
were either passing  emitters  or marginal emitters  on their
as-received FTP tests,  as well as  the  passing  and marginal
emitters  taken  together  as a group.   Thus,  these are  the
vehicles  that  would presumably  show  minimal impacts  from I/M-
instigated  repair.

     Each collection  of three bars in  Figure  23  represents
the percentage  of the  given  MOBILE4 emitter  category that
passed a particular  core sampling  mode  from the  Basic  I/M
Test  Procedure.   Again,  the horizontal  axis  gives  the
significant vehicle   operation that preceded   the  sampling
period.  In  each case, the data were from the second idle
mode of the core sampling period.  Based on the  discussion in
Section  23,  the  second idle   of  each  period  showed almost
uniformly higher pass rates than did the first  idle.
                             46

-------
       Thus,  for  example,  the  middle  set  of  bars  shows  the
  rates for the  fifth  mode of the Extended Loaded sequence  (XL-
  05) ,  which was the second  idle-neutral following an extended
  stretch of  loaded operation.   From the figure,  100%  of  the
  passing emitters passed this  "second-chance"  idle test.   For
  the  marginal  emitters,  the  comparable  figure on  XL-05  was
  85%,   and  for  the   passing  and   marginal  emitters  taken
  together,  91%.
                             FIGURE 23

              Response of the Normal  FTP  Emitters  to
                    Second-Chance Short Testing
                   Pass
                     Marginal
                             Pass+Marginal
                                   XL-OS
                                        XMO    RS.C
  P
  a
  s
  s
   Base
 Operation:
Conditioning:
Warm
Soak
None
 Warm  Warm Soak  Extended  Extended  Extended  Extended
  Soak   Extended   Loaded    Idle      Idle     Loaded
        2500rpm
Extended Extended
2500rpm    Idle
None     None   Extended  Restart
              2500rpm
       Figure  23  illustrates  the  significant  reduction  in
  failure rates of the normal  FTP  emitters that would  probably
  have accompanied any of the  sequences,  had one been  employed
  as a second-chance  test  in the Michigan AET program.   Five of
  the seven  sequences show  pass rates  for  the normal  emitters
  that are  above  75%.   For the  three fully-warmed sequences
  preceded   by  extended   loaded  or  extended  2500rpm   base
  operation,  the error of  commission rate would have been  at or
  near zero.
                                47

-------
       The  apparent  success of second-chance testing at passing
  normal emitters  does  not  come without cost,  however.  Second-
  chance testing also  reduces significantly the  failure rates
  among  the FTP  high  emitters,  and  consequently  reduces  the
  available  emissions benefit  to  be gained by  repair efforts.
  Figure 24  presents the response  of the FTP high emitters in
  the  base  sample  to the  various  second-chance  tests  of  the
  Basic  I/M  Test  Procedure.    Each  column   in  the  chart
  corresponds  to the percentage of  the original  239 vehicles
  that  failed  the  second  idle-neutral  of  the  indicated
  sequence.   For  example,   somewhat over  70%  of  the  vehicles
  failed  their  second-chance  cold  start  test,  with  HC+CO
  failures being the largest group.
                             FIGURE  24

                 Response of the Hioh Emitters to
                   Second—Chance  Short  Testing
• CO-only
HHC+CO
E3 HC-only
  F
  a
  Base
 Operation:
Conditioning:
Warm
Soak
None
 Warm
 Soak
             Warm Soak Extended Extended
              Extended   Loaded    Idle
              2500rpm
Extended  Extended
2500rpm    Idle
                       None
                                           None
Extended  Extended
  Idle    Loaded
Extended  Restart
2500rpm
       The effect  of  the various types of vehicle  operation on
  the failure  rates  of the high  emitters was less  severe than
  was  seen  earlier  in  the  normal   emitters,  but   was
  directionally the same.   Between 28% and 39%  of  the original
  failures fell away, depending on  the  second-chance test.  The
                                48

-------
lowest second-chance  failure rates  occurred  after extended
loaded operation  (XL-05  and  RS-05) ,  followed  closely by the
rates  on  fully  warmed vehicles  with  extended  2500rpm
operation (XI-10).   Thus, if  one  were to  address the error of
commission problem  through  such  preconditioning and second-
chance testing,  between  35%.. and._..40%  of _tne__hAgh emitters
captured  by  the initial test  might  be  lost. ~  The" "highest
failure   rate   occurred  on  the  vehicle  with  the  worst
conditioning --  the cold start  sequence  —  which was shown
earlier in Figure 23 to exhibit  the worst  error of  commission
rate.
     4.5.5.  Variability Between Adjacent Idle Modes

     Table  12  uses  linear regression on  the  HC  and  CO
emission values to analyze the  variability  between  the  first-
idle  and second-idle  short test  modes  in  the  seven core
sampling sequences  of  the Basic  I/M Test Procedure.   Recall
from Section x.x that  these two modes are  separated by a 30-
second  2500rpm no-load mode,  making the  core  sampling
sequence roughly  equivalent to a two-mode idle  test.  Once
again, the base operation and conditioning  modes  are provided
for each of the modes  as a guide to the I/M test  conditions
that each simulates.
                          TABLE 12

           Regressions  on  the First and Second Idle
              Modes  of  the Core Sampling Periods


Base
Operation

Conditioning

Sample Size
HC
Slope
Intercept
R- squared
CO
Slope
Intercept
R-squared
MODE
CS-03/05
Warm
Soak

None

237

0.69
252
25.8%

0.65
2.03
30.9%
CS-07/10
Warm
Soak

Extended
2500rpm
238

0.99
25
88.5%

0.95
0.27
85.6%
CS-012/15
Warm Soak
Extended
2500rpm
Extended
Idle
237

0.97
51
72.8%

0.89
0.57
70.4%
XL-O2/05
Extended
Loaded

None

238

0.97
19
85.5%

0.96
0.18
91.1%
XI-02/05
Extended
Idle

None

237

0.94
57
69.7%

0.89
0.60
69.2%
XI-07/10
Extended
Idle

Extended
2500rpm
237

1.00
22
90.7%

0.95
0.29
88.0%
RS-02/05
Extended
Loaded

Restart

231

1.03
5
90.0%

0.97
0.12
88.7%
                             49

-------
     For  a  variety of reasons, not all vehicles  were  tested
with each of the sequences in the Basic  I/M  Test Procedure.
The sample  sizes in Table 12  reflect the  number  of  vehicles
in the 239-vehicle sample where paired first-idle  and second-
idle  modes  were  available  for  the given  segment  of  the
procedure.

     The  regressions  are least-squares  fits  to lines  of  the
form  y = mx  +  b,  where  m  is the  slope  and b  is  the  y-
intercept.   In  each case,  the x-values  were  taken to  be  the
first  idle  of  the pair,  and  the y-values  were  the  second
idle.    If there were  no variability between the  first-  and
second-idle  modes,  the relationship  between  the emission
scores for each  pollutant  would be a line  with  slope of one,
intercept of zero,  and 100% R-squared value.

     As  shown  in the  table,   four segments of the procedure
showed slopes greater  than 0.95 and  correlation greater than
85% for both HC and CO; these were the Cold Start  modes 7/10,
Extended  Loaded modes  2/5, Extended  Idle modes  7/10, and  the
Restart modes  2/5.   In  two  of these segments  (CS-07/10  and
XI-07/10), the  short  test  was immediately preceded by  three
minutes of 2500rpm conditioning.   In the  other two  segments
(XL-02/05 and  RS-02/05), the  short  test  was  preceded by  an
LA4 prep cycle.

     Not  surprisingly, the worst correlation  (R2 values  below
30% for  both  HC and CO) was  shown by the paired values  for
the Cold Start modes 2/5, which came  at  the very beginning-of
the  procedure,   following  an  extended  soak   and   no
conditioning.   The  fact that  the  slope  of  the  HC  and  CO
regression lines for this comparison  are well  below one  shows
that even the  short  period of 2500rpm operation  between  the
two  idle  modes   was  sufficient   to   reduce   emissions
considerably on many vehicles.

     The  two  remaining  segments,  which also showed poorer
correlation,  were the  Cold Start  modes 7/10 and the Extended
Idle modes  2/5.    These  short  tests  immediately followed
periods of extended idle:  10 minutes in the case  of the Cold
Start,  and 20 minutes  in the  case  of the Extended Idle.

     Based on  the  preceding  discussion,  periods  of extended
no-load off-idle operation and extended loaded operation  can
reduce the variability in short test  scores that might  result
from periods  of extended  idle operation  or  from testing a
vehicle too  promptly after  soak periods.

     Figure 25 shows the actual scatter for one of the pairs
of idles with good correlation: the first-  and second-idle HC
values from  the extended  loaded  sequence.   Recall that  the
base operation before the first idle  mode  in this  sequence is
                              50

-------
an  LA-4  prep  cycle.    In  this  case,  the  scatter  was
responsible  for  reducing  the  correlation  coefficient to
85.5%.  The maximum values  for each axis  have been  set below
the  actual  maximums   in   the   sample   in  order   to  allow
examination of  the region  around  the  207(b)  HC cutpoint of
220 ppm, where a number of vehicles changed pass/fail  status;
no actual data point that falls beyond the range of  the graph
passed HC on either idle mode.

     The plot  clearly  shows  that  a number  of  vehicles had
significant differences between  their first-idle and  second-
idle  HC  scores,  even though  the 30  seconds of 2500rpm  that
separated the  two  idles might seem inconsequential compared
to the  LA-4 that  preceded the first idle.  Because of their
variability, some  of  these changed pass/fail status  for HC:
data  points  with  the  open-square  symbol  are vehicles  that
failed  the  first  idle,  but changed to  pass  on the  second;
points marked with an open triangle passed the first idle but
failed the second.
                          FIGURE 25

          Scatter Between First  Idle and Second Idle
             Following Extended Loaded Operation
    2nd
    Idle
    H3
   (ppm)  440
u •
rt
u •
n .
u •
0 -
0 -
0.
1
0
	 r 	 i
1
1
1
1
1
.... ...L. ........
1
t i
.......... L ... .....
1 •
. I
A A '
i •
.......... ..... . .^ .
A 2 •*• "
^A;i ;'..•
^~ " 1 1 " ^ •* ^^
!jifis*-«il * 1 D
220 4'
r 	 i

I
• .-• " J

"a""0""

W 66
	 i
•I
.........J
' •



• 0 8f
	
|
...J...!.j
.........J

......... J
a

JO 11






00
                             1st Idle HC (ppm)
                              51

-------
     When  both HC  and CO  are considered,  12% of  the base
sample — 29 vehicles — changed failure type  (pass, HC-only,
CO-only, HC+CO)  between the  two  idle modes  of the Extended
Loaded  sequence  (the sum of  the  bold entries  in  Table 13) .
Twenty-three of  the 29 also  changed  their  overall idle test
status:  15  failed  the  first  idle  and passed the second, and
the balance  did the reverse.   As  the table  shows,  however,
most of this difference  came  from  CO-only failures that were
cleaner  following  the  2500rpm mode; the  changes  in  other
failure types for the most part offset each  other.
                           TABLE 13

      Distribution of the Base Sample  by  Failure Type on
        Idle Modes Following Extended  Loaded Qpp»ra-Mnn
Second Idle
(XL-05)
Pass
HC-only
HC+CO
CO-only
All
First Idle (XL-02)
Pass
126
6
2
0
134
HC-only
7
1 1
1
0
19
HC+CO
3
1
58
2
64
CO-only
5
0
2
1 4
21
All
141
18
63
16
238
                           TABLE  14

       Distribution of the Base Sample by Failure Type
       on Idle Modes Following Extended Idle Operation
Second Idle (XI-05)
Pass
HC-only
HC+CO
CO-only
All
First Idle (XI-02)
Pass
89
4
2
2
97
HC-only
1 0
28
1
0
39
HC+CO
1 2
9
52
5
78
CO-only
4
0
4
15
23
All
115
41
59
22
237
     Table 14 makes  a similar comparison for  the  idle modes
in a sequence with poor preconditioning,  the first and second
idle modes  of the extended idle sequence.   Here,  the  base
operation before the first idle test is a twenty-minute idle;
the intervening mode between the two idle tests is once again
30 seconds of 2500rpm operation.  The number of vehicles that
change their failure type in this case rises to 53, or 22% of
the base  sample.   Note,  however, that almost  three quarters
of these  were more serious  failures  on  the first  idle;  for
                              52

-------
example, 26 vehicles changed from first idle failures of one
type  or another  to passes  on  the  second  idle.    This is
consistent with the information presented in Figure  X above,
in that even a brief stretch of  2500rpm operation appears to
compensate for extended periods of idle operation.
4.6  Supplemental Analysis  of  the AET Errors of Commission

     As discussed  previously  in this  section,  33 of the  239
vehicles in the  CTP base sample were errors of  commission by
the Michigan AET test.   A table of basic emissions data  and
vehicle identifying information for these vehicles  appears in
Appendix D.

     Figure  19  above showed  that  the Ec  vehicles were  not
randomly distributed by AET failure type;  a disproportionate
number  were  HC-only AET  failures.  The distribution  of  Ec
vehicles also varied  by basic vehicle characteristics.   All
but six of the  33 vehicles fell in the  1983-86 model  years;
the  fuel-injected  1983-86 vehicles  were the  most heavily
represented quota  group, with  19 vehicles  (53%) of  the total
Ec   fleet.     Differences   were  also  evident  between
manufacturers.    Six  of  the   sixteen  Toyota  vehicles were
errors  of  commission,  while   there  were no  Chrysler  Ec' s .
Trucks were  over-represented;  there were eight LDT Ec's  out
of the  33  total  (24%), while  trucks  only represented eight
percent of the  base  sample.   Five  out of the eight LDT  Ec' s
were Fords,  all  of them carbureted.

     One  particular  engine   stands  out  in  the   error  of
commission fleet:   the  GM 151 CID fuel-injected LDV.  Eight
of these  vehicles appear in the list  of Ec's,  representing
all but two  of  GM's total Ec's, and   almost one-quarter  of
the total number of  Ec's  in  the CTP fleet.  Notably, all  of
the eight  vehicles failed  the  AET test for HC.

     As shown in the data of  Appendix D,  almost all of  the
AET errors of commission in the CTP fleet had  elevated  AET HC
scores; a surprising  number were  above  400ppm HC,   and three
exceeded lOOOppm.  The sample  included thirteen  AET errors  of
commission for CO, with  the highest showing a 6.7% CO score.

     Appendix D  also  provides data from the 30-second point
of the  first  idle-neutral  from the extended loaded sequence
(XL30HC and  XL30CO),  performed on  the as-received  vehicles.
Recall  that  these  values  represent   idles  following   an
extended period  of loaded preconditioning  (an LA4)  .  In most
cases,  the HC values during this mode of the extended  loaded
sequence were much lower than  the  comparable values from  the
AET test.   However, the XL30HC values  for five of the 33 were
still elevated from normal levels,  and three continued to  be
                             53

-------
EC'S.  None of the vehicles that was a CO error of commission
on the AET  test  showed failing XL30CO values during its as-
received testing,  although a handful exceeded 0.5% CO.

     With only a  few exceptions,  no component problems were
identified  during  the  as-received diagnosis that might have
explained  elevated  values for the  Ec vehicles  on the AET
test.   Thus, only five  of  the 33 Ec  vehicles  had  repair
efforts,  and only one of those could be considered successful
in resolving observed  emissions anomalies.  This  was vehicle
21,  a  1984  Toyota,  whose  malperforming oxygen  sensor was
replaced,  leading to  normal HC  and CO levels.

     For most of  these Ec vehicles,  no specific  explanation
for  the  elevated AET  scores was available.   GM attributed
elevated  AET  HC  values  in  its  Ec  fleet  to  inadequate
preconditioning during the AET test.   Nevertheless,  two of
GM's vehicles (338 and 347) that were  HC  errors  of commission
by the AET  test  also  showed  failing  values on the extended
loaded sequence,  in  spite  of the LA4 preconditioning.    These
and  others  of GM's  Ec  group  did  show  elevated  HC  values
during other idle  modes of their as-received short testing.
For  one  AMC  vehicle   (31) , the  AET  failure  was  probably
attributable to  an air diversion timer  tied  to the elapsed
time at idle.
                              54

-------
    SECTION  5:   EMISSION  EFFECTS OF REMEDIAL MAINTENANCE
5 . 1  Introduction and Sample
     Prior  to any  repair,  each vehicle  underwent  an as-
received  characterization to aid  in the  decision  of  which
repairs  were  necessary,  if any.    This  characterization
included  the BITP,  FTP,  and  a complete  diagnosis  of the
vehicle's  engine  and  emission control  systems.    If the
results  from this  characterization indicated  that  repairs
were necessary to  meet the  criteria  for  vehicle release,  a
repair  sequence  was  designed.   The  first criterion was  a
reduction in FTP  emissions to a  target  based on  the  vehicle's
certification standards.   Once  this criterion had been met,
repairs targeted emission  levels and variability on  the I/M
test.   The  vehicle was  released when  all  criteria had been
met or all reasonable  repair  efforts were completed.

     Of the 239-vehicle sample,  184  vehicles received  a  total
of  479  remedial  maintenance   (RM)  steps for which  mass
emissions  data  were  collected.  Of  these steps,  372  were
single,  isolatable  repairs;  the  remaining  107   RM  steps
included  258 repairs,  for  a  total of  630   repairs.   Each
vehicle therefore  received an average  of  3.4  repairs in 2.6
RM steps.

     The design of the CTP dictated  that catalyst replacement
be  a  last  resort  repair  so  that  the  high  conversion
efficiency of a new catalyst would  not mask the  necessity of
other repairs; under certain  conditions, the CTP  program plan
did not then require a final  mass emissions test.  Therefore,
much of  the following analysis  is  focussed  on  pre-catalyst
repairs only.  This included 413 RM  steps  on 175  vehicles.
Vehicles  that  passed  the  CTP  standards   (150%  of  cert
standards for mileage <=50K; 200%  for mileage  >50K)  of the
as-received FTP  as  well as the  ideal I/M portion  of  the  BITP,
with little variability  throughout the BITP,   did not  undergo
repair.
5.2  Total Mass Emission  Reductions from the CTP Fleet


     5.2.1  Net Benefit  of  Repairs — FTP-Based

     The reduction in emissions of the CTP due to repair was
substantial, with  nearly all of the emissions  in excess of
certification standards  being eliminated.   The net emissions
benefit achieved from the  repairs  on  these 184 vehicles was
339 g/mi HC  and 5193 g/mi CO,  for an  average reduction per
vehicle of 1.8 g/mi HC and 28.2 g/mi CO.   (The certification
                             55

-------
standard for the  majority  of  these  vehicles is 0.41 g/mi HC
and 3.4 g/mi CO).   On a percentage  basis,  HC emissions were
reduced by  slightly over 80%, and CO  emissions  by rjver 85%
for these 184  vehicles.   These reductions  eliminated almost
99%  of the  excess  HC  and CO FTP  emissions,   relative  to
individual vehicles'  certification  standards,  of the entire
CTP  fleet.20    See  Tables   15  and  16 for  breakdowns  by
manufacturer and quota group.   Note that  greater than 100% of
a  group's  excess  can  be  eliminated;  this  is   caused  by
vehicles  that  are  repaired  to  levels  cleaner  than  their
certification standard.
                          TABLE 15

      FTP Emission Reductions  (a/mi]
for All  Repairs  —
                       bv  Manufacturer
HC
GM
FORD
NISS
TOYT
CHRY
AMC
VW
MAZD
SUBA
MITS
HOND
ALL
number of emissions
vehicles reduction
46
47
17
8
15
13
10
7
8
6
7
184
86.56
105.18
25.66
4.68
24.44
25.07
17.24
9.24
20.58
10.08
10.24
338.98
%
reduction
83.6%
83.3%
77.8%
42.3%
79.3%
71.0%
71 .4%
80.6%
87.6%
84.1%
91.2%
80.3%
average
reduction
1.88
2.24
1.51
0.58
1.63
1.93
1.72
1.32
2.57
1.68
1.46
1.84
total
excess*
85.60
105.82
25.96
8.36
24.68
27.74
20.55
8.72
19.06
8.98
8.65
344.12
% excess
reduced
101.1%
99.4%
98.9%
55.9%
99.0%
90.4%
83.9%
106.0%
108.0%
112.3%
118.4%
98.5%

CO
GM
FORD
NISS
TOYT
CHRY
AMC
VW
MAZD
SUBA
MITS
HOND
ALL
* excess
number of
vehicles
46
47
17
8
15
13
10
7
8
6
7
184
is based on entire
emissions
reduction
1091.3
1374.1
763.6
103.9
395.4
384.2
327.6
220.0
356.6
113.1
62.9
5192.6
239-vehicle sample
%
reduction
86.4%
87.1%
91.4%
57.9%
85.9%
78.4%
83.1%
84.9%
87.6%
82.1%
89.2%
85.5%

average
reduction
23.7
29.2
44.9
13.0
26.4
29.6
32.8
31.4
44.6
18.9
9.0
28.2

total
excess*
1051.8
1378.5
770.1
147.8
394.9
404.8
359.5
241.5
344.5
106.4
50.8
5250.7

% excess
reduced
103.8%
99.7%
99.1%
70.3%
100.1%
94.9%
91.1%
91.1%
103.5%
106.3%
123.8%
98.9%

                             56

-------
                          TABLE 16
      FTP Emission Reductions  (g/mi)  for All Repairs --
                       bv Quor.a Group
HC
FI81-82
Garb 8 1-82
Garb 83-86
F! 83-86
ALL
number of
vehicles
18
62
36
68
184
emissions
reduction
52.00
122.84
58.44
105.71
338.98
%
reduction
83.6%
78.7%
80.4%
80.5%
80.3%
average
reduction
2.89
1.98
1.62
1.55
1.84
total
excess*
55.42
130.60
56.12
101.97
344.12
% excess
reduced
93.8%
94.1%
104.1%
103.7%
98.5%

CO
FI81-82
Garb 8 1-82
Garb 83-86
Fl 83-86
ALL
* excess is
number of
vehicles
18
62
36
68
184
based on entire
emissions
reduction
830.7
1869.3
829.7
1662.8
5192.6
%
reduction
90.1%
82.6%
84.0%
87.5%
85.5%
average
reduction
46.1
30.2
23.0
24.5
28.2
total
excess*
849.0
1921.8
834.0
1645.8
5250.7
% excess
reduced
97.8%
97.3%
99.5%
101.0%
98.9%
239-vehicle sample
     Because of the special treatment of catalyst  repairs  in
the CTP, they  are  here  separated from the analysis.  Of the
184-vehicle sample,   175  vehicles  received  non-catalyst
repairs; for 168 of these, both pre- and post-repair FTP data
is available.  This includes tests  on vehicles  that  received
a catalyst  change  at a  later  RM stage.   The net  emissions
benefit from these  non-catalyst  repairs was 264 g/mi HC and
4544 g/mi  CO,  for  an  average  reduction per  vehicle of 1.6
g/mi  HC and  27.1  g/mi  CO.    On  a percentage  basis,   HC
emissions  were reduced  by over 67%,  and CO  emissions   by
almost 80%.  These  reductions eliminated at least  76% of the
excess HC  and  86%  of the  excess  CO emissions of  the entire
CTP  fleet,  and  84%   and  92%  of  the  excess  HC  and  CO,
respectively, of this  168-vehicle sample;  more reductions may
have occurred for  which both pre- and post-repair data is not
available.

     Obviously, catalyst  repairs had a  significant  impact  on
the overall emission reductions even though they were usually
the final  repair  to  be  performed,  occurring when  emission
levels  had  already been significantly  decreased.   Their
contribution to the total emissions benefit  was due to the
unusually low FTP levels  after a catalyst  repair  rather than
to excessively  high levels prior  to that  repair;  emissions
were lower, on average, prior to a  catalyst replacement than
before other types  of  repair.
                             57

-------
     Catalyst  replacements  for which  FTP  data is available
accounted for 10% of the RM steps  that  occurred;  their  repair
eliminated  17%  of  the excess HC and 8%  of the excess CO of
the entire  CTP  fleet.  An additional 1% of the RM steps also
involved  catalyst  replacements, but  lack of  data  does not
allow  an  assessment  of  their emissions  impact.    Final
emission  levels following  a catalyst  replacement  averaged
0.33 g/mi HC and 4.2 g/mi CO — approximately half the  levels
of the  final  non-catalyst repairs, at  0.76  g/mi HC and 7.1
g/mi CO.  The average  reduction for a  single  catalyst  repair
eliminated  77%  of  the  HC  and  65%  of  the CO ^emissions
occurring   just  prior  to  the  repair,  and  44%  and 32%,
respectively,  of the  vehicle's  entire  as-received HC  and CO
emissions.   In  contrast, the  average  non-catalyst   repair
eliminated  38%  of  the  HC  and  51%  of  the CO  emissions
occurring   just  prior  to  the  repair,  and  29%  and 39%,
respectively,   of   the  vehicle's   as-received  HC  and  CO
emissions.    Thus,  catalyst  repairs  were  somewhat  more
productive  in  eliminating  HC  than were  repairs to  other
components,  while CO was  approximately in the same range.

     5.2.2  FTP versus LA4 values

     One purpose of  the  CTP  was to obtain emission benefits
of individual repairs, as measured  by  the  FTP.   However,  in
order  to streamline  the  testing  process,   labs  were  not
required to  perform an entire FTP after  a repair  if there was
reason to believe  that  there had been  no effective emission
benefit.  In  case  of uncertainty,  the  lab could perform an
LA4 (bags 1 and  2  of the FTP) ,  without the extended vehicle
preconditioning required in a  complete  FTP.   Results  of the
LA4 were to  be  used  to decide  if  the FTP  was  needed;
significant  emission  decreases  would  mean  that  it  was.
Unfortunately,  this process was  not  always  followed,  with the
follow-up FTP  sometimes eliminated.   For this  reason,  FTP
data  is  not  available   before and  after  each  obviously
significant  repair.

     Rather  than eliminate these repairs  from  the analysis —
a significant proportion for at least one manufacturer — it
was judged  better  to  retain  all  the  data and  perform the
analysis on an  LA4 basis.   For repairs  without  an  LA4,  FTP
data was converted to  LA4 data  by  using results  from bags  2
and 3;  bag 3 has the  same driving cycle  as bag 1,  and
approximates the warmed-up vehicle  condition seen in the LA4-
only tests.

     It was important to determine  the  effectiveness of thus
modelling FTP  emission reductions  with  LA4-derived  values,
and,  if effective,  the approximate  shift in values one could
expect.  As a first step to accomplish  this,  FTP values were
compared to  the LA4 values  derived  from the  FTP  scores  on
those same  vehicles.   The  scatterplots  in  Figures  26  and 27
                             58

-------
show  that FTP-derived LA4  values  tracked  FTP values  quite
consistently  for both HC  and  CO,  indicating  that  derived LA4
values  are  generally  an  effective  approximation  of  FTP
scores.
                           FIGURE 26

         FTP vs FTP-Derived LA4  Emission Values -- HC


                                 LA4HC
                      LA4HC = 0.96(FTPHC)-0.16   R2 = 92.8%  (19.1,18.6)
                           4        6

                               FTP HC (g/mi)
10
12
                           FIGURE 27

         FTP vs  FTP-Derived LA4 Emission Values — CO
LA4CO
LA4CO - 0.99(FTPCO) - 1 .95 R2 - 98.9%
160
140 -
120 •
i A A 1 no •
LA4 iwvi
rn BO .
lnlmi\ ert
\yn"l oU •
40 •
20-
0 t



















«&?
J«4S*T^
^
tF*r
rl






pfn





a






r9D °














            0    20    40    60    80    100   120   140    160

                               FTP CO (g/mi)
                               59

-------
     Table  17 shows  the  total and  average  FTP as-received
emissions  and  emission  reductions  for  all  vehicles  that
received repairs, and the LA4  values derived from those same
FT? results.   It also gives the ratio of the derived LA4 to
FTP values,  shown as a  percentage  in the last  row of each
section of the table.  Figures  28 and 29 show the scatter in
g/mi  reduction  per  RM  as  measured  by  the  two  methods.
Derived  LA4   emission  values  and  g/mi  emission  reductions
generally were somewhat  lower than  FTP  values,  as illustrated
in Figure 30; derived LA4 values values are the  lower of each
set  of  lines,  with dotted  lines  showing HC  and  solid
indicating CO.   On the other hand, the percent reductions as
measured by  the  derived  LA4 were  several  percentage points
greater than  as measured by  the  FTP.

     The  similarity  of  the  FTP  and  derived-LA4  values
indicate that  any conclusions  that would be drawn  from FTP
data  would  not   be  significantly  altered  by  the use  of
derived-LA4   values.    Therefore,  much  of  the  remaining
analysis is  performed using derived-LA4  and  actual  LA4 data
as an FTP substitute.  The emission impact of thus using LA4
rather than  FTP  values  can  be  illustrated  by repeating some
of the  information found  in Section  5.2.1  above,  describing
net emission  reductions  due  to  repairs; this time, LA4 rather
than FTP values  are  used.  A  comparison of  the two reveals
little difference in  the substance of the  findings.
                          TABLE 17

  Emission  Values as Measured bv the FTP and FTP-Derived LA4
HC
FTP
LA4
LA4/FTP
number
vehicles
184
184
total
as-rcvd
422.34
382.99
90.7%
total
reduction
338.98
321.91
95.0%
average
as-rcvd
2.30
2.08
90.7%
average
reduction
1.84
1.75
95.0%
%
reduction
80.3%
84.1%
104.7%

CO
FTP
LA4
LA4/FTP
number
vehicles
184
184
total
as-rcvd
6074.3
5665.9
93.3%
total
reduction
5192.6
5063.0
97.5%
average
as-rcvd
33.0
30.8
93.3%
average
reduction
28.2
27.5
97.5%
%
reduction
85.5%
89.4%
104.5%
                             60

-------
                      FIGURE 28
FTP  vs FTP-Derived  LA4 Emissi    Reductions  --
                             ALA4HC
                 ALA4HC - 0.99(AFTPHC) - 0.05   R2-93.1%
                        FTP HC reduction (g/mi)
                      FIGURE 29
FTP  vs FTP-Derived LA4 Emission  Reductions -- CO

                             ALA4CO
                ALMCO - 0.99(AFTPCO) - 0.23   R2 - 99.2%
        -50
0          50         100
   FTP CO reduction (g/mi)
150
                          61

-------
                          FIGURE 30
      Average Emission Reductions Due to Repair:  FTP
                       FTP-D^ri vpri LA4
                as received
after final repair
     5.2.3  Net Benefit of Repairs — LA4-Based

     The net  emissions benefit achieved from the  repairs  on
these 184 vehicles  was 322  g/mi HC and 5063 g/mi  CO,  for  an
average reduction  per vehicle of  1.7  g/mi HC and  27.5  g/mi
CO.  On  a  percentage  basis,  HC  emissions were  reduced  by
slightly over  84%,  and CO emissions by almost 89%  for these
184 vehicles.  These  reductions  -- as measured by  the  LA4  —
eliminated  approximately 95%  of  the  excess  HC  and  CO  FTP
emissions,  relative to certification standards,  of the entire
CTP fleet.

     Of the  184-vehicle sample,  175 vehicles received  non-
catalyst repairs for  which  pre- and post-repair LA4  data  is
available.   This  includes  tests on vehicles that  received a
catalyst  change at  a  later  RM  stage.    The  net  emissions
benefit from  these  non-catalyst  repairs was 255 g/mi  HC  and
4575 g/mi  CO, for  an average reduction per  vehicle  of  1.5
g/mi  HC and   26.1  g/mi  CO.    On a  percentage  basis,   HC
emissions  were reduced by over  69%,  and CO  emissions  by
almost 83%.   These  reductions eliminated at least  74%  of  the
excess HC and 87%  of  the excess  CO emissions of  the  entire
CTP  fleet,  and  78%  and  90%  of the  excess  HC  and CO,
respectively,  of this  175-vehicle sample; more  reductions  may
have occurred for which both pre- and post-repair  data is  not
available.
                              62

-------
5.3  Overview of the Repairs Conducted


     5.3.1  System and Subsystem Repair Categories

     Repairs  were  categorized by  the testing organizations,
with some  advice from EPA  staff,  into the following  systems
and subsystems.

    1.   induction system
        •    heated  air door  assembly
        •    temperature sensors
        •    air filter element
        •    hoses
        •    other (e.g.,  gaskets)

    2 .   fuel metering system
        •    carburetor assembly
        •    idle mixture adjustment limiter
        •    idle mixture adjustment
        •    idle speed
        •    idle speed solenoid
        •    fuel injection components
        •    hoses,  lines,  wires
        •    choke adjustment --  notches
        •    choke adjustment --  vacuum break
        •    choke adjustment limiter
        •    fast idle speed
        •    vacuum diaphragms
        •    electrical controls
        •    exhaust heat control  valve assembly
        •    hoses,  lines,  wires
        •    other (e.g.,  fuel filter,  float level)

    3 .   ignition system
        •    distributor assembly
        •    initial timing
        •    initial timing limiter
        •    spark plugs and  wires
        •    vacuum advance assembly
        •    spark delay devices
        •    spark knock detector
        •    electronic timing module
        •    coolant temperature  sensors
        •    hoses,  lines,  wires
        •    other (e.g.,  points,  distributor cap)

    4 .   EGR system
        •    egr valve assembly
        •    back pressure transducer
        •    delay solenoid
        •    vacuum  amplifier
        •    vacuum  reservoir
        •    coolant temperature  sensor
                              63

-------
      •   hoses, lines, wires
         other  (e.g., gaskets, plugged manifold)

 5.  air injection system
      •   air injection assembly
      •   bypass valve, dump valve — air pump system
      •   air diverter valve
      •   check valve
      •   drive belt
      •   hoses, lines, wires
      •   other  (e.g., air filter, stuck valves)

 6.  PCV system
      •   pcv valve assembly
      •   filters
      •   hoses and lines
      •   other  (e.g., vent tube seal)

 7.  exhaust system
     •   exhaust manifold,  tailpipe, muffler
     •   catalytic converter
     •   other  (e.g., mixture set tube)

 8.  evaporative system
     •   evaporative canister
     •   canister filter
     •   canister purge solenoid/valve
     •   hoses, lines, wires
     •   other  (e.g., gas cap, gaskets)

 9.  engine assembly
     •   engine assembly
     •   cooling system
     •   valve adjustment
     •   belt tensions
     •   hoses, lines, wires
     •   other  (e.g., battery, transmission fluid)

10.  three-way catalyst system
     •   electrical control unit
     •   oxygen sensor
     •   barometric pressure sensor
     •   load sensor  (throttle position,  manifold vacuum)
     •   engine speed sensor
     •   coolant temperature sensor
     •   crankshaft position sensor
     •   egr position sensor
     •   egr control solenoids
     •   air/fuel control actuator
     •   air bypass solenoid/valve
     •   air diverter solenoid/valve
     •   throttle kicker/actuator
     •   idle speed control system
     •   hoses, lines, wires
                            64

-------
        •   diagnostic bulb check
        •   diagnostic warning
        •   other  (e.g., switches)

     In  addition  to  these  categories   of   systems  and
subsystems repaired,  the CTP  database  includes a code listing
the  nature  of   repair   --  replaced,  adjusted,   cleaned,
reconnected,  restored,  or  rebuilt.    Narrative  comments,
filled  out  for  each  RM step,  elaborated  on  the  exact
components, diagnostic  techniques,  and  other  details judged
relevant  by  the  technician  but not  covered  by  the coding
system.
     5.3.2  Emission Benefits per System Repair

     According to the CTP program plan, repairs  were  to be
done one at a time, with mass emission tests before and after
each repair, and  in  decreasing  order  of their likely impact
on emission reductions.   In  fact,  this  happened much of the
time,  resulting  in   a  substantial database  of  isolatable
repairs  with bracketing mass  emission tests.    For  these
cases,  the emissions  reductions  can be  simply averaged over
all  of  the occurrences  of  a particular repair.   Attention
must be paid to  the  possibility that  some  repair types with
apparently  low  average benefits  were  the   result  of
misdiagnosis as  to what needed repair.

     However,  in a  number  of  cases,  more than  one repair
occurred   prior  to  a  post-repair  emission   test  being
performed, resulting  in  a number of non-isolatable  repairs.
For  these,  simple averages for each  repair type  would have
resulted  in  counting  the  entire emission  reduction of the
grouped repairs  for each of the  repairs in the  group.

     To overcome  this problem,  multiple  linear regressions
were performed,  using  the systems listed in Section  5.3.1 as
variables.  The change in emission levels for each pollutant
was  regressed  across  the  ten  systems,   resulting in  the
emission reduction for each pollutant  to due repairs to each
system.   The regression for HC reduction took the form
                10
   ALA4HC = 0 +  Z(ALA4HC)i x  (indicator for  system^  repair)
               i-l
A similar  regression was performed for CO  reduction.   Note
that regressions were  calculated  with  zero  as  a constant
term; that is,  the results were  forced through the origin, so
that if no  repair occurred,  the result  would  be no  emission
reduction.

     The  following table  lists  repair results at  the system
level.   Included  are  both the simple averages  of the repairs
which  occurred  singly,   and  the  results   of  the  multiple
regression for  all repairs.   All  emission values are  in grams
                             65

-------
per  mile,  as measured  by  the LA4 .
breakdowns by quota group.
See Appendix  E  for
                           TABLE 18

            Emission Reductions per  System Rppair
SYSTEM
REPAIRED
Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
All
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
14
94
63
19
77
4
60
4
10
156
501
A HC
0.33
1.07
0.19
0.11
0.41
-2.51
1.20
0.31
0.90
0.55
0.62
A CO
0.0
13.2
-1.4
1.5
6.7
-9.4
7.3
3.0
6.8
17.8
9.9
MULTIPLE LINEAR REGRESSION
ALL REPAIRS
N
28
117
78
27
93
11
68
8
21
179
-
A HC
0.27
0.98
0.23
-0.16
0.30
-1.25
1.16
0.75
0.53
0.62
-
t- ratio
0.64
5.12
0.95
-0.40
1.20
-2.01
4.51
1.07
1.13
4.03
-
A CO
1.4
12.8
-0.7
-3.6
4.5
-6.5
8.0
10.7
5.9
17.8
-
t- ratio
0.28
5.70
-0.26
-0.78
1.53
-0.89
2.65
1.31
1.08
9.78
-
     A stepwise  regression  was  then performed,  successively
eliminating the  system  with the lowest t-ratio (correlation
coefficient  relative   to   standard  error) .    This  method
eliminated  the  systems  which  had  the least  statistically
significant repairs; that is, the systems whose repairs were
the  least  useful at explaining an emission  reduction were
eliminated.   The systems  with  the most  significant  repair
benefits  were the  fuel  metering,  exhaust,  and  three-way
systems.    For  these,   the  two  approaches  give  similar
reduction estimates, as  shown in the  following table.   See
Appendix F for  breakdowns by quota group.
                          TABLE 19

 Emission Reductions per Repair to Statistically Significant
                           Systems
SYSTEM
REPAIRED
Fuel Meter
Exhaust
3-Way
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N A HC A CO
94 1.07 13.2
60 1.20 7.3
156 0.55 17.8
MULTIPLE LINEAR REGRESSION
ALL REPAIRS OF THESE SYSTEMS
N A HC t-ratlo A CO t-ratlo
117 1.01 5.30 13.0 5.85
68 1.20 4.68 8.4 2.80
179 0.64 4.16 17.9 9.89
                             66

-------
     These  statistically significant systems  are highlighted
in  dark grey  in the  following  graphs  of  average  emission
reductions  for  isolatable repairs.   The  light  grey bars must
be  taken cautiously,  since  the values  that  created  these
averages are highly variable.  See Appendix  E  for individual
figures per quota group.
                           FIGURE  31

      Average HC Reductions per Isolatable System Repair
     1.5 T
 HC
 g/mi
     -1.5 ••
      -3 J-
         INDT  FUEL  IGNT  EGR  AIR
EXH  EVAP  ENG 3WAY ALL
                           FIGURE  32

      Average CO Reductions per Isolatable System Repair
     20 T
 CO
 g/mi
     10--
     -10 -L
         INDT FUEL  IGNT  EGR   AIR
EXH EVAP  ENG  3WAY ALL
                              67

-------
     In  general,  calculating  emission  repair reductions  at
the  system level,  as  just  done,  is not  illuminating,  since
the repairs varied widely within the general system category.
An examination by subsystem appears  in Section 5.3.3 below.

     Nevertheless,  some  additional  information can  be  drawn
from the analysis  at  the system level.    As  clearly  shown  in
Figure   29,   repairs  to   those  systems  with   the   least
variability in FTP repair reductions were much more effective
than others at  consistently getting I/M  failing  vehicles  to
pass the I/M test.  For instance,  exhaust system repairs were
effective  89% of  the  time,  due almost entirely to catalytic
converter  replacements.   The excellent  converter  efficiency
of brand new  catalysts may be  responsible,  and this  result
should  be  considered  cautiously.    Repairs  to  the  fuel
metering and  three-way  systems turned   I/M  fails into  I/M
passes  66  and 64%  of  the  time, respectively.   Fuel  system
repairs  usually entailed  replacing the  carburetor or  fuel
injectors  (nearly all  injector  replacements  were on  a  single
basic  engine  model)  or   tuning   the   system  (largely  on
carbureted  vehicles)  --  that  is,   adjusting  idle mix,  idle
speed and/or initial timing .   Three-way system repairs were
mostly oxygen  sensor or,  less frequently,  ECU replacements.
                          FIGURE 33

    I/M Pass Rates Due to System Repair  —  bv Quota Group
              Cart 81-82  9 Carb 83-86 H Fl 81-82   D Fl 83-86
          EXH
FUEL
SWAY
IGNTT
     Repairs to those systems with greater variability in  FTP
repair reductions were, at the same time,  less  consistent  in
reducing the I/M failure rate.   Air injection system repairs,
effective at eliminating I/M failures  52% of the time,  varied
among  various  valve  replacements and  repairs  to  the pump
                              68

-------
assembly,   while  repairs  of  the  ignition   system  were
overwhelmingly tune-ups,  and succeeding  in  getting failing
vehicles  to pass  I/M only  30%  of the  time.    The missing
systems --  induction, EGR,  PCV,  evap,  and engine -- had too
few vehicles failing I/M at the  time of  the repair to include
in the analysis.   In general,  quota  group had little impact
on the effectiveness of a  certain repair.  Figure 33 and the
values cited above  include  only isolatable repairs.   Missing
data   indicates   small   sample  size,   rather   than  0%
effectiveness.


     5.3.3  Emission Benefits per Subsystem Repair

     An analysis by subsystem  is  essentially an analysis by
component or component group.   Results at  this  finer level of
detail  can  be  used  to  better  pinpoint  those  specific
components  that  have  the  greatest  impact on  emissions.   The
same  technique used  above --  simple  averages  of  emission
reductions  for isolatable  repairs, coupled  with a  stepwise
multiple  linear  regression for  all repairs — was  repeated,
this time using subsystems  as variables.

     Many subsystems were  eliminated from the results due to
a  low  occurrence  of repairs.    This  is  presumably because
these components were not often diagnosed as  emission control
problems  in need  of repair, either  because they  were,  in
fact, not in need of repair, or because  their malfunction was
judged to not  significantly affect emissions  (see Appendix G
for  a  count of  repairs per subsystem) .   Of  the remaining
subsystems,  a  step-wise  multiple  regression  yielded seven
with  statistically  significant  emission reductions  due  to
their repair.  Those with more  than seven cases and  a t-ratio
greater than 2.0 for one  or both  pollutants are considered
significant.

     Table  20  lists the simple averages  and the results of
this step-wise regression.  As with  repairs categorized by
system, the most consistently effective repair types were to
the fuel system,  the exhaust system — mainly the  catalyst —
and  the  electronic controls for the three-way  system.   Not
surprisingly,  some of  the most important emission control
components   --  the  catalyst   and  oxygen  sensor  --  are
consistently  effective  at cleaning   up both   HC  and  CO
emissions.

     Note  that,  of  the  seven  repair  types,  five  were
effective for both HC and CO; two were  consistently  effective
on CO  only  (the  ECU and load  sensor) .   Recall from section
4.2.3  that  most  vehicles  that  were high  emitters  on  one
pollutant were also  high  on the  other,  so that repairs were
often targeted, effectively  so, at reducing  both HC and CO.
All emission values in  the table are in  grams  per  mile,  as
measured  by the LA4.   See Appendix G  for  results  for  all
                             69

-------
subsystems,   and  Appendix  H   for   breakdowns
statistically significant subsystems by quota group.
                                          of  the
                           TABLE 20
Emission Reductions
                                  Subsystem Rf»pa ] r
SUBSYSTEM
REPAIRED
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
O2 Sensor
Load Sensor
All
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
22
30
19
43
14
69
11
372
A HC
1.03
0.61
2.35
1.11
0.40
0.80
0.61
0.70
A CO
11.9
11.6
24.2
7.0
10.7
20.7
23.2
10.7
MULTIPLE
ALL REPAIRS
N
27
43
20
56
19
82
22
-
A HC
1.02
0.63
2.22
1.20
0.67
0.94
-0.27
-
LINEAR REGRESSION
OF THESE SUBSYSTEMS
t-ratlo
2.7
2.1
5.1
4.7
1.5
4.4
-0.7
-
A CO
12.6
11.8
22.5
8.5
13.4
22.9
11.2
-
t-ratlo
2.9
3.4
4.4
2.8
2.5
9.0
2.3
-
     The method of repair which most often occurred on these
seven  important  subsystems  was  replacement  of the  main
component — carburetor,  fuel injectors,  catalytic  converter,
oxygen sensor,  load sensor (manifold air  pressure  or throttle
position  sensor),  or  electronic control  unit.    The other
frequent subsystem repair was a fuel metering  tune-up, which
included adjustments to idle speed, idle mix,  and/or  initial
timing.

     A comparison of the simple averages with  the  regression
correlation coefficients  reveals  reasonably  consistent
results,  except for  one case  --  the  load  sensor.    This
subsystem yields an  average CO reduction  of  23.2 g/mi when
repairs  limited to the load sensor are averaged, in contrast
to  an  11.2  g/mi  reduction  projected  by the  regression.
Further investigation  reveals  that when all  repairs  to the
load sensor are averaged, including those  lumped with other
repairs, the average reduction drops to  11.2 g/mi, identical
to that predicted by the  regression.   Of the 22 load sensor
repairs, 11 were lumped with another repair — hoses,  lines,
and wires;  these were  all Ford  vehicles  undergoing a recall
procedure to clean  out the  line leading to  the  map  sensor,
regardless of a diagnosis  indicating  its  necessity.  These 11
repairs  actually increased CO emissions by an average of 0.9
g/mi,  presumably because  they  were often unnecessary, while
the 11  isolatable  repairs  decreased  CO  by  23.2 g/mi,  on
average.  This  is  an unusual instance,  in which two repair
types  were  repeatedly performed together,  with consistent
emission effects.   In this case, the multiple  regression was
not able  to separate out the  effects of  the  single  repair
type.    In most other  cases  for which we  have a reasonable
                             70

-------
sample size,  the multiple  regression tracked the  averages of
isolatable repairs  closely.
     Figures  34 and
from  the  preceding
subsystems that have
that pollutant.   The
just the seven  major
See  Appendix  H  for
group.
35 chart  the  average emission  reductions
 table.    Dark  columns  indicate those
 statistically significant reductions  for
 average  reduction for all repairs --  not
 ones  —  is also included in  the  figures.
 breakdowns  of  these  figures  by quota
                           FIGURE 34

     Avera.de HC Reduction  per Isolatable Subsystem Repair
          Cart   FuelMtr   Fuel   Catalyst   ECU    02    Load
                 Tune   Inject                Sensor  Sensor
                                    ALL
                           FIGURE 35

     Average CO Reduction  per Isolatable Subsystem Repair
          Caib  FuelMtr  Fuel   Catalyst
                Tune   Inject
               ECU     02     Load
                     Sensor  Sensor
ALL
                               71

-------
     Many  of
effective  at
illustrates
subsystems at
test.
 the  same  seven  subsystems  are consistently
 reducing  the  I/M  failure  rate.   Figure  36
the  effectiveness  of   repairs  to  certain
allowing an I/M failing vehicle to pass  the I/M
                          FIGURE 36

            I/M Pass Rates Due to  Subsystem Reoair
        Catalyst   Carburetor
           Oxygen    Fuel Meter  Fuel Injector Ignition Tune
            Sensor     Tune
     Catalytic converter replacements were  effective  in  this
task 100% of the time, while replacement of the carburetor or
fuel  injectors,  fuel  metering system  tune-ups,  and  oxygen
replacement  sensor replacements  allowed  I/M  fails to  pass
approximately 75%  of  the  time.  On the  other  hand,  ignition
system tune-ups  (i.e.,  spark plug or plug  wire  replacements
or initial timing  adjustments), which did  not  appear  earlier
as statistically significant,  were effective only 20%  of the
time  at  turning I/M  fails  into I/M passes.   The  remaining
subsystems  had   repairs   on  I/M  failing  vehicles   too
infrequently to be  included  in the  analysis.

     Figure  37 presents  the  same  data split according  to
quota group.   Note that missing data indicates  small  sample
size  (fewer  than  five cases), rather than  0%  effectiveness.
Both  figures  and the  values cited in the  previous  paragraph
include only isolatable repairs.
                              72

-------
                          FIGURE 37

   T/M Pass Rates DIIP to Subsystem Repair —  hv  Quota  Groun
               Garb 81-82 M Cart 83-86 H Fi 81-82    D Fl 83-86
        Catalyst   Carburetor
Oxygen
Sensor
Fuel Meter
  Tune
Fuel Injector Ignition Tune
     5.3.4  Total Benefit per Subsystem

     Many  of  these same subsystem  repair  types not only  are
consistently  effective at  reducing emissions  on individual
vehicles,  but  also contribute greatly  to  the total emission
reduction  of all  repairs  in the  CTP.   Figures 38 and  39  show
the subsystem repairs that  contributed greater than  5%  of  the
overall CTP  repair reduction.   These  values were derived  by
multiplying  the average  emission reduction  per quota group
for  a  subsystem  repair  type   (calculated  with  isolatable
repairs  only)  by  the  number of  times a  repair occurred  to
that  subsystem in  that  quota   group  (all occurrences  —
isolatable or not).   This estimate of the  total contribution
of  that subsystem was  then divided by  the total benefit
realized by all repairs,  generating percent  contribution  per
subsystem.  See Appendix  I for  a  listing  of results  for  all
subsystems; this table is not stratified by quota.   Note  that
the  totals  do  not  equal  100%,  due  to   the   combining  of
isolatable averages with all repair occurrences.

     The most  important  repair  at reducing fleet  emissions
was the oxygen  sensor, for  both HC and CO.   This was not  only
the most frequently repaired component, being replaced  on  1/3
of the repaired vehicles,  but also contributed a fairly large
reduction  when  replaced  — 0.80 g/mi  HC   and  20.7  g/mi  CO.
The catalyst  was also very important,  being replaced  on  30%
                              73

-------
of  the  repaired  vehicles.   It was  even  more  effective than
the  oxygen  sensor  at reducing  HC per  vehicle,  at  l.li g/mi,
but  only about  1/3  as effective at reducing CO.
                             FIGURE  38
   Contribution  of  Subsystems  to Tntal  HC Repair  R»n«»fit-
                          bv  Quota Group
                 Garb 81-82  9 Garb 83-86 H Fl 81-82    D Fl 83-86
           O2   Catalyst   Fuel
         Sensor         Injector
                            Garb    Fuel     Fuel
                                   Meter   System
                                   Tune   Other
                           Ignition  Air Pump
                            Tune
                             FIGURE 39
   Contribution of  Subsystems  to Total CO  Repair Benefit
                          bv Quota Group
 %
 CO
35% -r
30% - •
25% • •
20% •.
15%-.
10%--
 5% - •
 0% - -
                 Garb 81-82  H Cart 83-86 El Fl 81-82    D Fl 83-86
            02
          Sensor
               Fuel
              Injector
Fuel Meter
  Tune
 Load
Sensor
Catalyst    Carb
 Fuel
System
 Other
                                 74

-------
     Fuel  system  repairs  of  many kinds,  including  fuel
injector replacements,  catalyst replacements,  tune ups, and
miscellaneous  repairs  to other  fuel  system components were
also extremely effective at reducing fleet  emissions of both
HC and  CO.   The  fuel  meter tune items  —  idle speed, idle
mix,  and  initial  timing adjustments --  had an impact  based
largely on their  frequency  of occurrence, being performed  on
almost 25% of  the repaired  vehicles, but with  a per-vehicles
reduction  of  only  0.61 g/mi  HC and  11.6  g/mi  CO.   Fuel
injector replacements  occurred  somewhat  less  frequently,  on
only about  10% of the  repaired vehicles,  but had  extremely
high levels of reduction per repair,  at 2.35 g/mi  HC and 24.2
g/mi CO.   Carburetors were replaced about  as  often as fuel
injectors,  but were  only about half as effective per repair.

     Repairs   to  other  fuel   system  components  were not
particularly frequent,  occurring on only 6% of the  repaired
vehicles,  but  they had  extremely high average  reductions per
repair,  on the order of those seen for  fuel  injectors.   About
half of these repairs to miscellaneous  fuel  system components
had  a  negative or  negligible  emissions  benefit,  including
carburetor mixture  adjustments,  cleaning  deposits from the
throttle body,  and repairing the accelerator pump  or linkage;
those repairs  that were  effective consisted of replacing the
air  cleaner  vacuum  line, adjusting  the  float  level and M/C
solenoid,  and replacing the  jet  mixture solenoid.

     The  air  pump   and ignition  tune-ups  both  had  large
impacts  on  the  overall  HC reduction,  but   for  opposite
reasons.   The air  pump was   repaired  --  invariably this
involved replacement —  on  only 3%  of the repaired  vehicles,
but  had very  large  HC reductions upon repair,  averaging 7.9
g/mi.   Its  repair also resulted in very large  CO  reductions
of 41.1  g/mi, but  not  large enough to overcome the  small
frequency of occurrence.   Ignition  tune  items, on  the  other
hand — spark  plug or  plug wire replacements  or an  initial
timing adjustment,  followed by  an  idle  speed adjustment  if
needed — were only  marginally  effective per repair,  at 0.34
g/mi HC reduction,  but  their frequency  —   performed on 30%
of all  repaired  vehicles — caused  them to have  a large  HC
impact overall.


     5.3.5  Effect of Deteriorated Catalysts on Emissions

     Because  of  the  significant role catalytic  converters
play in emission  control and because of  their  susceptibility
to damage through tampering and misfueling, an analysis was
conducted  focusing  on  their role  in  emission  levels when
malfunctioning.  In  the CTP  fleet,  55 catalysts were replaced
on 53  vehicles  (two vehicles  had  both  an  oxidation and a
three-way catalyst  replaced), accounting for 11%  of the RMs
that occurred.  Table 21 details the emission  reductions for
catalyst replacements and for all other  RM types.  Catalyst
                             75

-------
replacements  eliminated  slightly  less than  20% of  the  CTP
fleet's  excess  HC,  and  slightly  less than 9%  of  its excess
CO.    Percent  reductions  in  emission  levels  per   RM  were
significantly higher  for catalyst replacements  than  for  the
other  RM types,  even  though,  since  catalyst  repairs  were
generally withheld until  the last repair,  the pre-RM emission
levels were lower on average at  the  time they  occurred.
                           TABLE  21
   LA4 Emission Reductions  —  Catalyst Replacements and All
                        Other Repairs
HC
catalyst
other
all
number of
RMs
55
413
479
average
pre-RM
emissions
1.38
1.82
1.74
average
reduction
per RM
1.23
0.62
0.67
percent
reduction
per RM
89.0%
33.8%
38.5%
total
reduction
67.48
254.74
321.91
% total
excess
reduced
19.6%
74.0%
93.5%

CO
catalyst
other
all
number of
RMs
55
413
479
average
pre-RM
emissions
10.9
23.0
21.3
average
reduction
per RM
8.5
11.1
10.6
percent
reduction
per RM
78.1%
48.1%
49.5%
total
reduction
468.5
4575.3
5063.0
note: 55+413=468; 10 of the missing repairs were post -catalyst replacement, excluded
due to potential masking effect of new catalyst on subsequent repair reductions; 1
catalyst diagnostic rather than repair
% total
excess
reduced
8.9%
87.1%
96.4%
was
     The  correctness  of the  diagnosis  that  a  particular
catalyst  was  malfunctioning is  important  to this  analysis.
Most vehicles --  95%  -- were released from "the test  program
with normal emission  levels,  suggesting that those that  did
not receive  catalyst  replacements  probably did not  require
them.    Also,  catalyst  replacements  were  normally withheld
until  all other repair options  were exhausted.   Therefore,  it
can be assumed that catalysts  repairs  were  generally  applied
only when needed,  and  avoided when  not.

     Evidence  of misfueling  or  tampering  did not  play  a
significant  role in  identifying vehicles  which  required  a
catalyst replacement to achieve normal emission levels.   For
catalyst replacement vehicles,  the  average  lead level  in  the
as-received tank fuel  was 0.0037  g/gal;  for  all  CTP  vehicles,
it was 38% higher, at 0.0051 g/gal.   Of the  81  vehicles with
above-average lead-in-fuel  levels,  only 9,  or  11%, received
catalyst replacements;  84%  of the  catalyst  changes occurred
on vehicles with below-average  lead-in-fuel  levels.  Also,  of
the ten vehicles with the highest  lead-in-fuel  levels  (0.015
                             76

-------
— 0.05 g/gal),  only two received catalyst  changes.   Thus,
fuel lead level is not a reliable predictor of the necessity
of a catalyst  replacement.

     Additionally, only  two of the  53  catalyst  replacement
vehicles were  noted  to have signs of  fuel inlet restrictor
damage.  Two vehicles which did not  receive cat  replacements
also had such damage; one was nevertheless released from the
CTP with normal  emitter levels achieved  via  other repairs,
while  the  other  remained  a high  emitter  due  to  obvious
tampering  --  a missing catalyst.   This  vehicle  certainly
would have  had drastic emission  reductions  had a  new catalyst
been  installed.    Thus,  evidence of fuel  inlet restrictor
tampering may be  a reliable  sign of the  necessity of catalyst
replacement,  but  occurs  infrequently.

     Two-thirds  of  the  as-received  fleet  underwent  the
Plumbtesmo  test  for  tailpipe lead residues;  only  three
failures  occurred,  and  none  of  these  required  catalyst
changes to  achieve normal emission levels.  Overall, only 20%
of  the  vehicles  that  had their  catalysts  replaced  had
evidence of  misfueling  or  tampering,  as indicated  by  the
Plumbtesmo  test,   fuel  inlet restrictor  damage,  or  above-
average fuel  lead  levels.


5.4  Analysis  of  Incremental Repairs
     5.4.1  Sample Description

     The focus will now move from a discussion of the repair
types that affected emission levels to the benefits actually
achieved  under different  circumstances  of  test  procedure,
vehicle emitter category and repair target.

     Recall that  630  repairs  were  performed in 479 remedial
maintenance (RM)  steps on 184 vehicles.  The sequence of the
RM steps on each vehicle  was  based  on  the as-received vehicle
characterization,  with the  repairs  judged most  likely to
reduce FTP emissions  performed  first.   Once FTP values were
sufficiently  low,  any vehicle  that  still had  difficulty
consistently passing  the  I/M  test  was repaired to eliminate
that problem.   Transient and I/M tests performed both before
and  after  each repair allow a  comparison of  repairs  that
helped the vehicle pass I/M with those that actually cleaned
up the vehicle,  as measured by the FTP or LA4.

     Recall from Section 4.5.2 that the core sampling period
of the  BITP  which followed extended  loaded preconditioning
had  the  lowest  failure rate,  with approximately  40%  of the
CTP  sample failing.   This  is  considered to  be the  BITP
sampling period  which is  closest  to  the  "ideal"  I/M  test
condition.   Analysis  of the effects of  repair on the I/M test
                             77

-------
 focuses  exclusively  on this sampling period.   The  following
 sections focus exclusively on non-catalyst  repairs.

     5.4.2  Benefits of Repairing to Pass  I/M

     The purpose  of  I/M is to determine  which  vehicles  have
 high emission levels,  so  that their  emissions can then  be
 reduced  through  repair; an  adequate repair should  not  only
 allow a vehicle to pass a subsequent I/M test,  but also clean
 up its actual in-use emissions.  It is therefore important  to
 determine if  those repairs  that  allow  an  I/M failing vehicle
 to  pass  are  also  the ones  that  actually  clean  up the
 emissions,  as  measured by a  mass emissions  test.    This
 section  investigates  the  mass emissions  benefits  that are
 realized by repairing vehicles  to pass the  I/M test.

     One  hundred  vehicles —  54%  of  those  that   received
 repair —  were failing their  ideal  I/M test at the time  of
 first repair.  Eighty-four  of  these  vehicles received  a  non-
 catalyst  repair at  some point  that allowed a passing I/M
 score,  while  11 required a  catalyst  replacement to pass  I/M,
 three never passed  the ideal  I/M test  (these three never
 received catalyst repairs,  and were high or marginal  emitters
 at the time  of release) ,  and the remaining two had  no post-
 repair I/M  test  data.   Overall,  it took  an average  of 1.5
 remedial maintenance  steps  to  get  a failing vehicle to  pass
 I/M.

     Both  I/M scores and mass  emissions  data are available
 for the non-catalyst  repairs  on  83 of the 84 vehicles which
were initially failing I/M.   The net LA4-measured emissions
benefit  achieved  from  these RMs  — those   that  first allowed
an I/M pass without a  catalyst change — was 178 g/mi  HC and
2651 g/mi  CO.  Average  reductions per RM per  vehicle  were
therefore 2.14 g/mi HC and  31.9 g/mi CO, over twice  as large
as the  average reduction per RM for all  repairs.   This  single
RM step reduced as-received HC and CO emissions by about 75%
 for these 83 vehicles,  eliminating  approximately 80%  of their
as-received excess.   This  high  level  of  reduction  on these
vehicles  eliminated  over  50%  of  the  excess  HC  and  CO
emissions of  the  entire  CTP fleet  and achieved close  to 60%
of the  entire reduction seen by  that fleet,  although the
 repairs  occurred on only 47% of the vehicles, and  represented
only 20%  of the non-catalyst  RMs  performed.   Additionally,
the number  of  high and super emitters  in   this  group dropped
 from 83%  to  25%   due  to  this  single RM.  Therefore,  the
 repairs  that worked  in terms  of  I/M  pass/fail were  also
apparently well suited to reduce FTP emissions.

     Table 22 gives values  for these vehicles at their first
 I/M pass,  broken  down by the  emitter  categories  before and
after the repair that caused the  passing test.   The reduction
 is that  caused by the single RM that caused the  vehicle to  go
                             78

-------
from I/M fail to  I/M pass;  mean emissions are those after the
repair —  that  is, at  first  I/M pass.   Percent  excess is of
failed vehicles in the  entire CTP  fleet.
                           TABLE 22

            FTP  Benefits  of  Repairing to Pass I/M
MOBILE4
Emitter
Category
Normal to Normal
Normal to High
High to Normal
High to High
Super to Normal
Super to High
TOTAL
Number
of
Vehicles
12
2
49
18
1
1
83
Mean LA4
Reduction
HC CO
0.03 1.4
-0.64 -15.5
2.38 47.2
1.74 12.9
15.69 62.9
15.12 57.2
2.14 31.9
% Of FTP
Excess
HC CO
0.1% 0.3%
-0.4% -0.6%
33.9% 44.1%
9.1% 4.4%
4.6% 1 .2%
4.4% 1.1%
51.7% 50.5%
Mean LA4
Emissions
HC CO
0.39 4.5
1.31 19.1
0.37 4.2
1.96 27.2
0.28 2.2
3.11 8.2
0.77 9.6
     The  FTP  benefits  of repairing to  pass  I/M were  not
dependent on  fuel  meter type or model year  group.   As Figure
40 illustrates,  about  60%  of  the vehicles  changed from high
to normal  emitters after the repair to passing  I/M  status,
independent of quota group.
                           FIGURE  40
 Changes  in Emitter  Group  Due  to  Repair  from I/M Fail
                    Pass — bv Quota Grouo
                                 1-0 T/M
  %0f
vehicles
• High-High
H High-Norm
D Norm-High
1 Norm-Norm
             FI81-82
Garb 81-82   Carb 83-86
Fl 83-86
ALL
                              79

-------
     Manufacturers  were  slightly  more variable in  their
success, but  this  was  to an extent due to the number of high
versus  normal emitters  in their original  sample.   Overall,
repairs that  turned normal emitters into high emitters while
allowing   the  vehicle  to  pass  I/M  were  quite  rare
approximately  2%.   Super emitters are included with highs in
the  following two figures.   Note that several manufacturers
are  grouped together;  this is due to their small sample size
(fewer than seven vehicles) in this subset of data.
                          FIGURE 41

 Changes  in Emitter Group Due to Repair from T/M Fail to I/M
                   Pass — bv Manufacturer
  %of
vehicles
                 High-High  H High-Norm  D Norm-High  • Norm-Norm
             GM
FORD
NISS
AMC
VW
OTHER   ALL
     There  is  some additional  FTP reduction  available from
more complete  repair  even after vehicles pass  I/M.   The CTP
test sequence  did not  specifically address this  issue,  but
the program  nevertheless  collected data on  39 vehicles that
received  additional  repair and  mass emission  testing after
they were passing I/M.  These vehicles achieved an additional
LA4 reduction of 0.33 g/mi HC and 9.5 g/mi CO,  on average,  as
shown  in  Figure 42.   These reductions  were achieved in  an
average  of  two  additional  remedial maintenance steps  per
vehicle.
                              80

-------
                          FIGURE 42
  Average LA4 Emissions for Vehicles with Addi 1--j npal  Repair
                      After Passing I/M
     HC
    g/mi

   3.5 j

    3 -•

   2.5 ••

    2 ••

   1.5 ••

    1 ••

   0.5 ••

    0 -I-
•O-HC
••-co
                             CO
                             g/mi

                             y 40



                             ••30



                             ..20



                             -.10
           as received
after first I/M pass
after final repair
     5.4.3  Comparison to MOBILE4 Repair Estimates

     The  MOBILE4  emissions  model  uses  emission  benefits
realized from repairing  failing  I/M vehicles  to pass the I/M
test  as  part  of its  calculations  for I/M  credits.21   The
values used are  derived from test  programs conducted by EPA
and the   California Air Resources Board,  in  which vehicles
underwent  the  I/M  process  and  failures  were repaired  by
either  commercial   garage  mechanics   or  EPA contractors
instructed not to continue  repairs past the  point  of passing
I/M.   Table  23  below  compares  the  MOBILE4   average  repair
reductions — for MY  80-86  vehicles with closed-loop control
which initially  failed an idle  test  — to CTP  values  for a
similar set of vehicles.   (The MY  80  vehicles in the MOBILE4
dataset are California only,  with  technology  similar to that
used on Federally certified  MY  81 vehicles).   All vehicles
included  in  these   tables  passed the I/M  test   following
repair.  Reductions  are calculated from as-received values,
and are those realized  through  the  RM step   which  took the
vehicle  from  I/M  failing  to  I/M  passing  status.    CTP
reduction  values are  LA4-based; MOBILE4  values  are FTP-based.
                              81

-------
                           TABLE 23
          MQBIT.E4 vs C.TP Average Emissiog  Redurfinni
                  Due to Renair to I/M Pass
Carbureted Vehicles

HC
Normal
High
Super
ALL
CTP
N
7
40
0
47

CO
Normal
High
Super
ALL
N
7
40
0
47
As-Rcvd
FTP LA4
1.33 1.11
3.50 3.30
3.18 2.97
Reduction
g/ml %
0.54 48.2%
2.46 74.6%
2.17 73.1%

As-Rcvd
FTP LA4
22.9 20.8
50.9 48.8
46.8 44.6
Reduction
g/ml %
12.1 58.4%
38.0 77.9%
34.1 76.5%
MOBILE4
N
38
53
9
100
As-Rcvd
FTP
0.76
2.86
13.81
3.05
Reduction
g/ml %
0.14 18.7%
1.46 51.1%
11.67 84.5%
1.88 61.6%

N
38
53
9
100
As-Rcvd
FTP
8.8
50.9
190.2
47.5
Reduction
g/ml %
1 .8 20.8%
29.0 57.0%
174.0 91.5%
31.7 66.9%
Fuel Injected Vehicles

HC
Normal
High
Super
ALL
CTP
N
7
27
2
36

CO
Normal
High
Super
ALL
N
7
27
2
36
As-Rcvd
FTP LA4
1.29 1.16
3.20 3.04
12.91 11.87
3.36 3.16
Reduction
g/ml %
0.69 59.9%
2.31 76.0%
10.18 85.7%
2.43 76.9%

As-Rcvd
FTP LA4
28.0 5.5
56.2 54.1
62.9 62.0
51.1 45.1
Reduction
g/ml %
1 .4 25.8%
44.3 82.0%
56.8 91.6%
36.7 81.4%
MOBILE4
N
12
24
4
40

N
12
24
4
40
As-Rcvd
FTP
0.41
2.36
6.41
2.18

As-Rcvd
FTP
5.8
47.9
184.1
49.0
Reduction
g/ml %
0.08 20.0%
1.42 60.3%
4.48 69.9%
1 .33 60.9%

Reduction
g/ml %
1 .3 23.4%
32.7 68.3%
139.0 75.5%
33.9 69.5%
     When  vehicles  in all  emitter categories  are combined,
the MOBILE4 percent reduction values undershoot those seen in
the CTP by 2-40%; this gap  would probably be several percent
greater if the CTP  values were  FTP-based.   The gram per mile
reduction used by  MOBILE4  is also  generally  lower than that
seen in the CTP,  partially due to lower as-received levels in
the MOBILE4 sample.  It can be argued that the MOBILE4 values
are more  realistic  since many of the  vehicles were repaired
in commercial  facilities rather than  emission laboratories.
On the other hand,  the CTP values can be considered the level
of reduction that  could be attained given improved mechanic
training in diagnosis and repair.
                              82

-------
     5.4.4  Benefits of Repairing to pass the FTP

     One  hundred   thirty-two  vehicles  --  72%  of  those
receiving repair — were high or super emitters  at  some point
during  their  repair  cycle.    Almost  3/4  of  these  --  98
vehicles --  received a non-catalyst  repair that turned them
into normal emitters, while 20% required a  catalyst repair to
be cleaned  up,  and the remaining  6%  never were repaired to
normal  emitter  levels,  never having  received  a catalyst
replacement.

     The  net LA4-measured  emissions  benefit  achieved from
these  RMs  —  the  non-catalyst repairs  that   cleaned  up  a
vehicle to normal emitter levels — was 202 g/mi HC and 3722
g/mi  CO.    Average  reductions  per   RM per   vehicle  were
therefore 2.06  g/mi HC and 38.0 g/mi  CO, almost three times
as large  as the average  reduction per RM  for  all repairs.
This single  RM  step reduced as-received HC and CO emissions
by about  85%  for these 98 vehicles,  eliminating over 90% of
their as-received excess.  This very  high  level of reduction
on these vehicles eliminated approximately 60%  of  the excess
HC and 70% of the excess CO emissions  of the entire CTP fleet
and achieved close to  2/3  of the entire  reduction  seen by
that fleet, although  the  repairs occurred on only  40% of the
vehicles,  and represented only  20%  of  the RMs performed.

     Ideal  I/M  tests  were performed on  most of  these high or
super  emitters.  Of those that  were  cleaned  up  to normal
emitting  levels, 57% became  I/M  passes after the  repair.
Another 37% had  previously been passing I/M and continued to
pass,   while 6%  continued  to  fail I/M  even though they had
achieved normal emitter levels.  Therefore, a total of 94% of
the vehicles that were repaired from high  to normal levels on
a transient  test could also,  at  that repair  stage,  pass an
I/M test performed under optimum conditions.

     Table 24 gives  values for  these  vehicles for  the repair
which  took  them from high  (or   super)  to normal  emitter
status, broken down by  the I/M pass/fail category  before and
after the repair.  The reduction is that caused  by  the single
RM that caused  the  vehicle to become  a normal  emitter; mean
emissions  are those  after  the repair  --  that is,  for the
first time at normal levels.   Percent  excess is  of  FTP-failed
vehicles .in the entire CTP fleet.
                             83

-------
                            TABLE 24
    I/M Benefits  of Repairing fn FTP Normal EmiM-pr
Ideal I/M
Pass/Fall
Status
Pass to Pass
Pass to Fail
Fail to Pass
Fail to Fail
Unknown
TOTAL
Number
of
Vehicles
34
0
53
6
5
98
Mean LA4
Reduction
HC CO
1.01 24.4
2.75 46.9
0.99 19.4
3.18 58.5
2.06 38.0
% Of FTP
Excess
HC CO
9.9% 15.8%
42.4% 47.3%
1.7% 2.2%
4.6% 5.6%
58.7% 70.9%
Mean LA4
Emissions
HC CO
0.45 5.4
0.39 4.2
0.57 5.0
0.57 4.3
0.43 4.7
     The  change  in  I/M pass/fail status  once a  vehicle was
repaired  to normal  emitter levels  was  somewhat  dependent on
model year  and fuel metering  system,  but  largely as  a result
of  variations in  the  I/M  pass-fail  levels  prior  to  the
repair.   Overall,  getting vehicles to pass I/M once their FTP
or LA4 levels  were low was not difficult  for  any quota group,
with 94%  of the vehicles  passing overall, and  no quota group
doing worse than  88%.
                           FIGURE 43

  Changes in I/M Pass/Fail  Status  Due to Repair from High to
                  Normal Emitter — bv Ouota
  %0f
 Vehicles
                   Fail-Fai
      Fail-Pass   D Pass-Fail   • Pass-Pass
              FI81-82
Cart 81-82   Cart 83-86
Fl 83-86
ALL
                               84

-------
     Manufacturers were  significantly  more  variable in their
I/M status changes, but   again,  this  was almost entirely due
to the differences in pre-repair I/M status.  GM and Ford had
the most  difficulty  in  obtaining passing  I/M scores  when
vehicles were cleaned up  on  transient  tests,  with  14% and 9%
failure rates,  respectively.
                          FIGURE 44

  Chances in I/M Pass/Fail Status  Due  fo  Repair  from High to
              Normal Emitter — bv Manufacturer
  %0f
 Vehicles
                  Fail-Fai
          Fail-Pass   D Pass-Fail   • Pass-Pass
             GM
FORD  NISS  CHRY   AMC   VW   OTHER  ALL
     5.4.5  Emission Benefits "Lost" through Second-Chance

     I/M tests do not have perfect pass/fail correlation with
the FTP.   Vehicles  that fail  I/M  with  passing FTP emissions
are considered false failures;  their repair  is  unnecessary
from a  clean  air  standpoint and  undesired from  a consumer
cost and  inconvenience standpoint.   One strategy  to reduce
the number of false I/M  failures  is  to  give all  failing
vehicles   a  second-chance  I/M  test.    Presumably,  clean
vehicles  that  fail  I/M due to  inadequate preconditioning
and/or canister purge during the  idle test  would 'have a good
chance of passing an immediate second-chance test  if preceded
by  sufficient  preconditioning,  and would not  have  to  be
repaired or  retested later.   Conversely,  vehicles  that are
truly dirty under normal operating conditions should, continue
to fail.  However, the second-chance test still being a short
test,  some dirty vehicles would  pass along with  those that
are clean; it  is  important  to determine  the potential repair
                              85

-------
benefits  from these vehicles that would be  lost  by applying
the second chance test.

      In  the  CTP,  138 vehicles — almost 60%  of the fleet —
passed an  ideal  I/M test  in as-received condition  (52 highs,
53 marginals,  and 33 passes) .  Recall  that  all CTP vehicles
failed their field  I/M  test,  so that  this  lab test  can be
considered second-chance.   We  can assume that a fairly large
proportion of  these  CTP  second-chance  passes would have also
passed a second-chance  test in the  field,  even  under non-
ideal conditions.  These vehicles  would then  not be  repaired.

      It would be  interesting to calculate the total emissions
benefit  due  to repairing these vehicles —  that  which would
be lost if they all passed second-chance.   However, we do not
have  repair  data  on all 138 of the vehicles,  since a number
were  released  from  the  CTP  without repair  and others did not
receive  mass emissions tests  after  repair.    Mass  emissions
data  for post-repair tests  (including catalyst repairs)  is
available on 82 vehicles that passed second-chance  (48 highs,
30  marginals,  4  passes).    We can  calculate  the  average
emission reduction per vehicle (from as-received to release),
with  per-vehicle  averages based  on  the as-received MOBILE4
emitter  category.   Summing  these averages according  to  the
weighting  of  the emitter  categories  of  the  138  vehicles
provides an  estimate of  the total  emissions  reduction that
would not  be  realized:    68 g/mi HC  and  1211 g/mi CO,  or
between 20 and 25%  of  the total  LA4-based  emission reduction
of the entire  CTP fleet.   This  estimate  is an upper bound,
since not  all  138 vehicles  would have passed"second-chance
given the  non-ideal conditions in the field,  and  since  the
CTP per-vehicle  repair  benefits are  probably  higher  than
those in  commercial  facilities.

     Thus,  repairs to 60% of the CTP  fleet  provided less than
25% of the reduction,  as  an upper bound estimate.   Over  60%
of these CTP second-chance passes were  normal  emitters  (pass
or marginal)  — much higher than the  general  CTP fleet at  40%
normal emitters — and therefore are  not desirable  candidates
for repair.    This  supports  the theory that second-chance
tests can reduce  the incidence of unnecessary  or  detrimental
repairs  to  cleaner  vehicles  without  greatly  reducing  the
emission  benefit due to repairs to those that are dirty.


     5.4.6  Benefits of Repairing  to Different Targets

     "Repair  benefit"  can  be  defined  many  different  ways,
based on  the  target  which  a  vehicle  is  being  repaired  to
meet.   I/M programs, of course, use a  passing  score on  a  I/M
test as the target;  after this point  is reached, there are no
further emission control  repairs.  The  CTP database provides
information on the extent of excess emissions  eliminated  via
I/M-targetted  repairs,  and  whether  a substantial  portion of
                              86

-------
"repairable"  emissions  remain after  an  I/M test  is  passed.
As the  following  table  shows, repairs to  I/M  targets reduce
as-received  emission levels  by  about 75%,  eliminating  over
half of the  FTP  excess.   However, this  is  only  about 2/3 of
the total  reduction  that can be  realized  with more complete
repair.
                           TABLE  25

Reduction in LA4 Emissions Due to Reoair to Different Taraets
TOTAL REDUCTION
(a/ml)
HC
to first I/M pass - non-cat
to first I/M pass ~ all
all non-catalyst repairs
all repairs
CO
to first I/M pass - non-cat
to first I/M pass - all
all non-catalyst repairs
all repairs
* vehicles
83
94
175
184
83
94
175
184
as-
received
253.56
281.59
367.51
382.99
3862.2
4115.7
5538.9
5665.9
after
repairs
63.83
65.87
112.77
61.08
797.1
820.8
963.5
603.0
reduction
189.73
215.71
254.74
321.91
3065.1
3294.8
4575.3
5063.0
%
reduction
74.8%
76.6%
69.3%
84.1%
79.4%
80.1%
82.6%
89.4%
% excess
reduced
55.1%
62.7%
74.0%
93.5%
58.4%
62.8%
87.1%
96.4%

AVERAGE REDUCTION
(a/ml)
HC
to first I/M pass -- non-cat
to first I/M pass - all
all non-catalyst repairs
all repairs
CO
to first I/M pass ~ non-cat
to first I/M pass - all
all non-catalyst repairs
all repairs
tt vehicles
83
94
175
184
83
94
175
184
as-
received
3.05
3.00
2.10
2.08
46.5
43.8
31.7
30.8
after
repairs
0.77
0.70
0.64
0.33
9.6
8.7
5.5
3.3
reduction
2.29
2.29
1.46
1.75
36.9
35.1
26.1
27.5
%
reduction
74.8%
76.6%
69.3%
84.1%
79.4%
80.1%
82.6%
89.4%
% excess
reduced
55.1%
62.7%
74.0%
93.5%
58.4%
62.8%
87.1%
96.4%
5.5  Repairs to High Emitters


     5.5.1  Effectiveness of Repair on Marginals  vs.  Highs

     It  is  important,  first, to  determine  the effectiveness
of repairs on marginal  emitters  versus high emitters.   Is it
worth  it  to  capture and  repair  the marginals,  or  would the
effort be better spent focussed entirely on the highs?  Sixty
percent of the CTP  sample were  high emitters  (143 vehicles),
while  26% were marginals  (62 vehicles) .   As shown in Figures
45 and 46,  the  emissions benefit  of  non-catalyst repairs to
                              87

-------
the  marginal  emitters  is  negligible,   whereas  the  high
emitters  have substantial LA4 reductions  — 1.3  g/mi HC and
33 g/mi  CO.   In fact,  the average emission  reduction on high
emitters  is  over 15 times as  great as  that seen on marginals,
for  both  HC and  CO.    These  figures  include  non-catalyst
repairs   on  only  those  vehicles that eventually  received
repair  (all  of  the highs, and three-fourths of the marginals)
and for which we have  complete mass emissions data.
                           FIGURE 45

    Average HC Repair  Benefit — Maraina^L :vs_JLLah_ Emi
                         ••>• marginal -O- high
     HC
     g/mi
2.5

 2

1.5

 1

0.5

 0
                       as received
                        after final repair
                           FIGURE 46

    Average CO Repair Benefit  —  Marginal vs High Emitters
     CO
     g/mi
                         ••*• marginal **• high
                       as received
                        after final repair
                               88

-------
     The  following  table  simply  tallies  the  number' of
vehicles which were dirtier  on  the LA4 after all repairs than
as-received,  on at  least  one pollutant.   The breakdown into
high  and marginal  emitters indicates  that  a much  higher
percentage of  the marginals than  highs were  dirtier after
repair,  although  catalyst  changes  helped  clean  up  both
marginal  and  high emitters.    Thus,   repairs to  marginal
emitters generally  result  in negligible repair benefits and
are much more  likely  to  be  detrimental  than repairs to high
emitters;  therefore,   marginal  emitters  are  not  a  worthy
target for I/M programs.
                          TABLE 26

              Vehicles Dirtier After All Repairs
EMITTER
CATEGORY
High
Marginal
NON-CATALYST REPAIRS
N cleaner
123
26
N dirtier
11
11
% dirtier
8%
30%
ALL REPAIRS
N cleaner | N dirtier
137 5
31 7
% dirtier
4%
18%
     5.5.2  Benefit of Repairing Highs only

     We  now  focus  on  high emitters,   as  both  the  most
prevalent portion  of the CTP sample and  the  most important
segment  relative  to  emission  reductions.    This  section
supplies FTP as well as' LA4  values, to provide data that can
be more easily compared to that from other programs.

     FTP emission  benefits  in a  single  non-catalyst  RM step
ranged from a  reduction of  more than  7  g/mi  HC and 125 g/mi
CO at the high end  (from installation of a new oxygen sensor
on a 1985  fuel injected  Oldsmobile Firenza,  and replacement
of  the  ECU  and  oxygen  sensor  on a  1984  fuel  injected
Chevrolet Cavalier), to an emissions increase of 8.6 g/mi HC
and 30 g/mi CO (from replacement  of  a  PCV fitting on a 1982
carbureted  Mercury Marquis).  As  Figure  47 shows,  a  large
reduction in one pollutant  did not  necessarily correlate with
a  large  reduction  in  the other,  although  there  were
relatively  few cases  in  which one pollutant  increased while
the other decreased.

     Overall,  the average  FTP benefit of non-catalyst repairs
to high emitters,  per RM,  was 0.83 g/mi HC and 15.8 g/mi CO.
This  reduced  HC   by  41%  and  CO  by  54% from  the  levels
immediately  prior  to  the  RM,  and  eliminated  31%  of  the
vehicle's as-received  HC  and 41%  of its  as-received  CO,  on
average.   Each high  emitter  received an  average  of 2.3 non-
                             89

-------
  catalyst  RMs,  and  eventually  had  83% and  92% of its excess HC
  and CO eliminated.    This  reduced  each  high  emitter's as-
  received  emissions by 69% for HC  and  81% for  CO,  or a  total
  of  1.86 g/mi  HC  and  34.5  g/mi CO,  from average  levels of 2.70
  g/mi HC and 42.4 g/mi CO.

      With the  inclusion  of  catalyst repairs,  98%  of the
  excess HC and 99%  of  the  excess CO from the high emitters was
  eliminated,   in  an average  of 2.6  RMs  per  vehicle.   This
  entailed   reducing  each  high  emitter's as-received FTP
  emissions by  81% for HC and  87%  for CO,  or  2.19  g/mi HC and
  35.3 g/mi CO, from average  levels of  2.69 g/mi  HC and 40.6
  g/mi CO.
                           FIGURE 47
   FTP Benefit per RM of Non-Catalyst  Renair  to Hiah Emitters
  CO
reduction
 (g/mi)
                             HC reduction (g/mi)
      The average  LA4  benefit  of non-catalyst  repairs to high
 emitters, .per  RM, was 0.71 g/mi  HC and 13.4 g/mi  CO.   This
 reduced HC  by  36% and CO by  50%  from  the  levels immediately
 prior  to  the  RM, and eliminated 28%  of  the vehicle's  as-
 received HC and 34% of its  as-received CO,  on average.   Each
 high emitter received an average of 2.5 non-catalyst RMs,  and
                               90

-------
eventually  had   78%  and   90%  of  its  excess  HC  and  CO
eliminated.   This  reduced  each high  emitter's  as-received
emissions by 71%  for  HC and 85% for CO, or  a total of 1.77
g/mi HC and 33.4  g/mi  CO,  from average  levels  of 2.50 g/mi HC
and 39.5 g/mi  CO.

     With  the  inclusion  of  catalyst  repairs,  94% of the
excess HC and 97% of the excess CO  from the high emitters was
eliminated,  in an  average  of  2.8 RMs  per   vehicle.   This
entailed  reducing each  high emitter's   as-received  LA4
emissions by 86%  for  HC and 91% for CO, or  a total of 2.11
g/mi HC and 34.7  g/mi  CO,  from average  levels  of 2.46 g/mi HC
and 38.1 g/mi  CO.


     5.5.3  Catalyst Repairs Performed  on Highs

     Of the 143 vehicles that  were  high emitters as-received,
48 eventually  received  catalyst replacements; mass emissions
data  is  available  on  all  but  one.   These  47  vehicles had
achieved,  on average,  an LA4-based  emission reduction of 0.95
g/mi  HC  and   15.7 g/mi  CO  prior  to  replacement  of  the
catalyst,  reducing HC  emissions  40%  and CO 58%  from as-
received  levels.    These vehicles  required  an average of 2.9
non-catalyst  RM  steps to  achieve  these  relatively  small
reductions.   A single  catalyst repair,  on  the  other  hand,
allowed   an  additional 1.29  g/mi  HC  and  8.8  g/mi  CO,
eliminating an additional 54%  and  33%, respectively,  of the
vehicle's as-received emissions.   This resulted  in an total
reduction  of  95%  of  the vehicle's  as-received HC  and CO,
bringing  levels  lower  than  certification requirements, thus
eliminating the entire  excess  for the vehicle.

    . Vehicles that never received a catalyst  repair, however,
had total reductions of 81% HC and 91% CO,  in only 2.1 RMs,
partly as a result of  higher  initial emission levels.   This
eliminated 90% of the excess HC and 95% of the excess CO for
these vehicles.   Non-catalyst  repair vehicles never attained
emission levels as low as  catalyst-replacement vehicles, with
final values approximately twice as high.  Nevertheless, most
of these vehicles were  brought  to  normal emitter levels even
without the benefit of a new catalyst,  while  over half of the
catalyst-repair vehicles were  still  high emitters  until the
catalyst  was   replaced, despite  almost three  attempts  at
repairing other  components.   In all,  over  15% of the high
emitters could not  be  brought  to normal  levels  by  any means
of repair other than catalyst replacement.

     The  following two figures  and  table   illustrate  the
emission  levels  as-received,  following  all non-catalyst
repairs,  and  finally  after   catalyst  replacements,  for
vehicles that  had no catalyst  replacement and  those that did.
Only vehicles that were high emitters  as-received are shown.
Values are LA4-based.
                             91

-------
                           FIGURE 48
Average  HC
                  of
                                       to  Hiyh
                 Catalvsf,  vs Other  Repairs
                •• cat repair vehicles      -o- non-cat repair vehicles
  HC
 g/mi
  3

2.5

  2

1.5

  1

0.5

  0
                    as-received  after final non-  after cat
                                cat repair      repair


                           FIGURE  49

   Average  CO  Benefit: of Repair to Hioh  Emi'
                 Catalyst vs  Other Repairs
      • cat repair vehicles
                                      non-cat repair vehicles
CO
g/mi
                   as-received  after final non-  after cat
                               cat repair      repair
                                92

-------
                           [•ABLE 27
        Average Benefit of Repair to  High Emitters
                  Catalyst vs Other Repairs
HC
cat repair vehicles
non-cat rapalr vahs
CO
cat rapalr vehicles
non-cat repair vena
number of
vehicles
47
95
number of
vehicles
47
95
as-received
emissions
2.40
2.50
as-received
emissions
27.1
43.9
after final
non-cat repair
1.45
0.47
after final
non-cat repair
11.4
4.5
after
cat repair
0.16
after
cat repair
2.6
     Catalyst replacement was the second most frequent  repair
performed on  high  emitters,  done on  1/3  of  them.   The  most
common was replacement of the oxygen sensor,  performed  on 43%
of the high  emitters.   This repair was highly successful  at
reducing emissions,  at an average LA4  reduction  for  the highs
of 1.23 g/mi  HC  and 10.1 g/mi CO, eliminating almost  1/2  of
the  excess  HC and  1/4  of  the  excess  CO for  the affected
vehicles.   Other repairs  done  frequently on high emitters
included various ignition  tune items  and fuel metering  tune
items, and  repair  or  replacement of  carburetor assemblies,
air  injection  system  check  valves and hoses, fuel  injection
components,  and  three-way  control system components such  as
load sensors, hoses, and the ECU.   Repairs to the  induction,
EGR,  PCV,  evaporative,  and engine  assembly  systems  were
relatively infrequent.
5.6  Difficulty of Repair


     5.6.1  Difficulty of Repair to  Passing Levels

     Another aspect  of vehicle  repair  is  the difficulty  of
diagnosing  and performing the  repair (s)  that will  actually
reduce emissions.  In the CTP,  diagnosed problems  were  ranked
according to their likely  impact on FTP  levels and  performed
in that order.   Therefore, the  earlier RMs should have had a
greater emissions benefit than those performed  later.    If
this holds  true,  we  can assume that diagnosis was  generally
correct, and therefore  was not a major  inhibiting  factor  in
reducing  FTP   levels.     The   remainder  of  this  section
investigates this issue.

     The following table and figure  indicate,  for  each  repair
step, the  vehicles that  changed MOBILE4  emitter status  at
each RM stage.   The  "high" grouping includes high  and super
emitters,  while  the  "norm"  grouping  includes  passes  and
marginals,   as  previously defined.   All  values  are based  on
                             93

-------
the  LA4,  and  only  repairs  performed  prior  to  catalyst
replacement  are included.   Some totals  are  not equal to  100%
due to occasionally missing data.
                            TABLE 28

             Emitter  Category Changes per RM
RM
Step
1
2
3
4
5
6
7
8
9
10
ALL


MOBILE4 Emitter Category Change
Norm-Norm
30%
45%
49%
52%
65%
63%
67%
75%
100%
-
42%
Norm-High
0%
5%
3%
0%
0%
0%
17%
0%
0%
-
2%
High-Norm
37%
18%
15%
14%
10%
25%
0%
25%
0%
-
24%
High-High
32%
30%
31%
34%
25%
13%
17%
0%
0%
-
30%
HIGH includes highs and supers
NORM includes passes and marginals
                           FIGURE 50

     Emitter Category Chancres  Due to Remedial Maintenance
            High-High  U High-Norm  D Norm-High   • Norm-Norm
                     2345678

                            Remedial Maintenance Step
                               94

-------
     In general,  earlier repairs  were more successful  than
later  repairs at  reducing  high  emission  levels to  normal
levels, as anticipated.   This  is  partly because,  at  later RM
steps,  a  greater percentage of vehicles had moved  into the
normal emitter category prior  to that  repair,  thus  reducing
the  number  available  to be  cleaned  from high to  normal
levels.  Additionally, however, those that were high emitters
at a later RM step were  less  likely  than  earlier repairs to
have substantially reduced  emission  levels (enough  to  drop
them into normal emitter  status)  due to that repair,  dropping
from a 53% chance  at  the  first repair to a 29%  chance at the
fourth and fifth.   The balance of these  two effects is  that
each time a  car  is repaired, there  is a two-thirds chance it
will be at normal emitting levels  after the repair.

     It took approximately 1.5 RM steps, on average,  to clean
up an  FTP- or  I/M-failing vehicle to  passing levels.  Figure
51 gives  a breakdown  of the repair step  after  which high or
super  emitters  became  passing  or  marginal  emitters,  as
measured  by  the  LA4,   and the  step after  which  I/M  failures
became I/M passes.
                          FIGURE 51

             RM Step that Cleaned Up  Emissions  —
 LA4 from High to Normal Emitter:  I/M Score  from Fail to Pass
  %of
Vehicles
100 -r

 90 ••

 80 ••

 70 ••

 60 ••

 50 -•

 40 ••

 30 ••

 20 ••

 10 ••

  0 4-
M 6th RM

Ei 5th RM

• 4th RM

• 3rdRM

H 2nd RM

D 1stRM
                   LA4
                                I/M
     There is  very little difference between  the  two cases;
in  fact,  the two  bars  represent many  of the  same  vehicles
moving  into  passing  status for  both tests at the  same  RM
stage.  There is a 57% overlap between the two bars;  that is,
                              95

-------
of  the  vehicles that  were either  high/super  emitters as-
received or  were failing I/M as-received, 57% were both.  Of
this  overlapping group,  72% passed  both test  types  at the
same  RM step,  21% passed  I/M  while  remaining high emitters,
and 7% became normal emitters prior to passing I/M.

      Quota group had a minor impact  on the number of  repair
steps required  to  turn a  vehicle  from a  high  to   normal
emitter.   While the  MY 81-82  vehicles  had nearly identical
results  regardless of  fuel metering  type, the MY 83-86  group
showed   a  marked  difference  between  carbureted  and fuel
injected models.  Carbureted vehicles were repaired to  normal
emitter  levels  in only  one  repair 81% of  the time,  whereas
only  53% of  the fuel  injected  vehicles were  successfully
repaired in a single RM  step.  Nevertheless, each quota  group
was able to  achieve  a success rate of  80-90%  after only two
RMs,  as  illustrated in Figure 52.
                          FIGURE 52

             RM Step that Cleaned Up  Emissions  —
      LA4 from High to Normal Emitter — bv Quota Grouo
  %of
Vehicles
100 -r

 90 -•

 80 ••

 70 ••

 60--

 50 ••

 40 ••

 30 ••

 20 •-

 10 ••

 0 ••
                                        M 6th RM

                                        E3 5th RM

                                        B 4th RM

                                        M 3rd RM

                                        Q 2nd RM

                                        DlStRM
FI81-82  Garb 81 -82 Garb 83-86 Fl 83-86
                                             ALL
     Manufacturers  were  varied in their  success  at reducing
high emitters  to normal emitters  in the first  repair,  with
Ford and  AMC having  the most trouble  — a success  rate of
only slightly  over 40%  —  and the "other" group  —  Toyota,
Mazda,  Subaru,  Mitsubishi,  and  Honda  — successful  100% of
the time.    (These  manufacturers  were  grouped to  achieve a
sufficient sample  size).   Nevertheless,  at least  85%  of the
high emitters  had  been cleaned to normal emitting levels by
the third repair, regardless of manufacturer.
                              96

-------
                          FIGURE 53
             RM Step  that  Cleaned rip Emissions --
      LA4  from High to  Normal  E'm icr.gr -- bv
 %0f
Vehicles
100

 90

 SO

 70

 60

 50

 40

 30

 20

 10

  0
D 6th RM

E9 5th RM

• 4th RM

• 3rd RM

13 2nd RM

QlstRM
            GM  FORD  NISS   CHR   AMC   VW  OTHER  ALL
     5.6.2  Effectiveness of Repair at Successive  RM  Steps

     The emission benefits  of  each repair also decreased,  in
general, as the number  of  RMs  increased,  as shown in Figures
54  and  55.    Step  four  is  out  of  line with  the  trend,
particularly for HC.   This is due to  a  single repair on the
super emitter, which accounts  for  2/3 of the total reduction
for that repair  step.   When this  vehicle  is  eliminated, the
total HC  reduction  plummets from  23.4 to 8.2  g/mi,  and the
average  drops  from  0.81  g/mi  to  0.29  g/mi,  which  is
consistent with the  trend across  all  RM steps.

     Similarly,  four  vehicles  had   a  large  impact on  the
excessive  reduction  in CO  seen  in repair step four.   These
four vehicles  had an average  reduction of 42.6  g/mi,  while
the remaining 25  had an average reduction of  3.2  g/mi.   The
repairs  to these four vehicles  were dissimilar  (rerouting
lines to the air bypass and diverter valves, or replacing the
carburetor,  fuel  injectors  or air  pump),  with  no  apparent
reason  for such large  reductions all  occurring  at  RM  step
four.   Elimination  of  them  from the  analysis  allows  the
general trend to  become more  clear.   The undue  impact  of a
few vehicles  can be  attributed to  the reduction in sample
size at  later  repairs;  this also  affects  steps five through
ten.
                              97

-------
                             FIGURE 54
          Total  and Average  HC Redurr i nns  per RM
                                                           Avg HC
                                                  AVERAGE Reduction
                                                        10
                        Remedial Maintenance Step
                            FIGURE  55


          Total and  Average CO Reductions per RM Step
   Total CO
  Reduction
    (g/mi)

   2500
   2000
TOTAL
         Avg CO
AVERAGE Reduction
                          4567

                        Remedial Maintenance Step
                                         10
      In  general,  the  early repairs were  quite  successful  at
reducing  emissions,   with  the  first  repair  generating  an
                                98

-------
average reduction  five  times  greater than the third repair,
despite the fact that each RM step usually included a repair
to only a single component.  This is encouraging, in that it
indicates  both that  technicians were  quite  successful in
identifying  the  required   repair,  and   that  a  single
malfunctioning component,  rather  than  a  complex set of
problems,  is often  the cause of high emissions.

     Another way to  approach  the issue of repair difficulty
is to look at  the  number of RM steps required  to reduce  high
emission levels by  a certain  percentage.   The following cable
lists the first RM  step in which  the  total emission  reduction
to that point  exceeded  80%,  from original levels of HC>2 or
C0>20 g/mi.   The average number  of  repair steps to achieve
this reduction was  1.8   Quota group  was  not a factor in the
ability to  reduce  high emissions by  80%,  as  illustrated in
Figure 52.

     Again,  however,  manufacturer had an effect,  with the
seven Subaru  and two Mitsubishi  vehicles  at these emission
levels  reduced by  80%  in  a  single  repair,  while  the three
Toyota  vehicles never  achieved  this  reduction even  after
multiple repairs.   GM and VW  were never able  to reduce 40% of
their vehicles by  this  amount,  while Ford and Chrysler  were
unsuccessful about  20%  of  the time.   Overall, approximately
75%  of  the  vehicles  with HC>2 or  C0>20  g/mi  eventually
received repairs which were able  to  reduce the high  pollutant
by 80% or more.
                          TABLE 29

        RMs Needed to Reduce LA4  Emissions by >8Q% —
           for Vehicles  with FTP HC>2 or CQ>2Q a/mi
RM
step
1
2
3
4
5
6
7
8
9
10
ALL
Number with
>80% reduction
46
18
8
5
1
1
1
0
0
0
80
Total
Number
106
75
49
27
21
15
8
7
5
1
106
Percent with
>80% reduction
43.4%
24.0%
16.3%
18.5%
4.8%
6.7%
12.5%
0.0%
0.0%
0.0%
75.5%
                             99

-------
                           FIGURE 56
         RM3
                    t-o
                              LA4 Emissions hy >8Q% —
       for VehielPs with FTP Hf>2 . 0 and/or CQ>20
                            bv  Quota
         50 j

         40 ••

   Number 30 - •
    of
  Vehicles 20"

         10-

          0--
                    Never
                      Di
               FI81-82
Carb 81-82
Garb 83-86
Fl 83-86
                           FIGURE 57

        RMs  Needed to Reduce LA4 Emissions by >8Q% --
       for Vehicles with FTP HC>2.Q and/or CQ>2Q g/mi
                        bv Manufacturer
• Never I
• >3
• 3
H2
Di
  Number
    of
  Vehicles
             GM  FORD NISS TOYT CHRY  AMC  VW  MAZ SUB A MITS MONO
     The  sum of these  various  approaches to  the  question of
diagnostic difficulty is that diagnosis  was  generally not an
impediment to emission  reduction.   Fewer than two  RMs were
required,  on  average,   to  clean  up an  FTP- or  I/M-failing
vehicle to passing levels and to  reduce  high emission levels
by  80% or  more.   This is  despite the fact  that  each RM
normally  included  only  a  single  repair.    Additionally,
                              100

-------
earlier repairs were twice as likely as  later  repairs  to turn
a  high  emitter   into  a  normal  emitter,  and  also  had
substantially larger average emission reductions.  Of  course,
some vehicles  were more  difficult  to  diagnose  and  repair;
five percent of the repaired vehicles were released  as high
emitters,  never having achieved normal emitter status.  Also,
15%  of the  repaired  vehicles received  more than  four  RM
steps,  and 4%  needed  more than six.  However, most of these
vehicles eventually had their  catalysts replaced; the design
of the CTP to delay these repairs until  all other  options  had
been exhausted contributed  to  the  high number of  RMs  in most
of these cases.
                             101

-------
APPENDICES
    102

-------
   APPENDIX  A:   FAIttJR*  RATZS IN TBX MICHIGAN AST  PROGRAM

     The following table provides initial-test failure  rates
(in percent) for all valid inspections performed in the
Michigan AET program in the first quarter of  1986.  Data are
provided by model year, and by the aggregate  of the 1981
through 1986 model years.  Available fields in the raw data
did not permit isolating light-duty vehicle and light-duty
truck failures.
   Table  30:   Michigan AET Failure Rates  by Model Year and
                         Manufacturer

AMC
CHRY
FORD
GM
MONO
MITS*
NISS
TOYT
VW
1981
17.8
21.2
25.0
15.9
3.8
50.0
48.3
10.4
25.5
1982
15.3
20.6
25.1
13.1
6.0
66.7
28.6
5.7
12.5
1983
16.4
14.4
14.0
8.3
11.3
15.4
15.2
8.5
10.8
1984
9.9
13.7
9.9
9.0
10.3
10.0
16.9
5.1
4.0
1985
10.8
5.3
6.6
5.4
11.6
15.4
4.4
4.2
1.5
1986
n/a
10.5
8.5
6.1
0.0
n/a
0.0
n/a
n/a
1 981-86
14.7
14.9
14.4
10.3
9.0
26.0
17.4
7.0
11.5
   Small sample size
                              103

-------
                     APPENDIX B
Vehicle  Identifying Information for the CTP Base Sample
Veh
001
002
003
004
005
006
007
008
009
010
011
012
013
014
015
016
017
018
019
020
021
022
023
024
025
026
027
028
029
Mfr
SUBA
AMC
AMC
SUBA
AMC
VW
AMC
VW
SUBA
SUBA
VW
AMC
AMC
TOYT
AMC
VW
SUBA
MAZD
MAZD
MAZD
TOYT
MAZD
TOYT
TOYT
TOYT
SUBA
TOYT
TOYT
AMC
Model
DL
JEEP
ALLI
WAQO
ALLI
JETT
JEEP
JETT
BRAT
GLF
GTI
JEEP
181
TERC
ALLI
RABB
SUBA
GLC
626
GL£
VAN
GL£
CORO
TERC
STAR
GL
CORO
CORO
SPIR
F/M
CARB
CARS
TBI
CARB
TBI
PFI
CARB
PFI
TBI
CARB
TBI
CARB
PFI
CARB
TBI
PFI
PFI
CARB
CARB
CARB
PFI
CARB
CARB
CARB
PFI
CARB
CARB
CARB
CARB
MY
81
83
85
82
85
85
84
84
83
81
85
85
81
85
83
81
86
85
84
85
84
83
84
84
83
81
81
83
81
CID
109
258
85
109
85
109
150
105
109
109
109
258
100
89
85
105
109
91
120
91
122
91
97
95
79
109
108
97
258
Type
LDV
LOT
LDV
LDV
LDV
LDV
LOT
LDV
LDT
LDV
LDV
LDT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
102
16
32
151
40
27
41
66
38
54
27
29
29
23
47
64
17
23
37
13
59
49
70
27
54
60
113
53
31
AETHC
534
1335
327
839
273
88
317
268
104
387
82
190
611
176
356
2000
168
71
181
63
337
97
382
231
225
180
392
978
130
AETCC
1.7
9.5
6.3
10.0
0.5
4.7
0.0
5.7
2.9
7.2
1.6
3.7
6.0
1.6
0.6
0.5
1.4
1.5
4.7
2.4
2.3
2.8
0.0
0.5
0.3
1.7
0.2
5.1
3.9
FTPHC
1.94
4.84
1.08
8.32
0.60
1.47
0.74
0.73
1.71
2.44
0.31
1.35
4.63
0.23
3.78
6.87
0.14
0.28
2.71
0.10
0.58
0.31
0.66
0.23
0.47
2.60
3.77
0.68
2.40
FTPCC
71.6
49.0
21.9
81.3
4.3
58.2
10.8
6.3
55.5
54.2
2.9
18.4
85.1
1.4
18.9
69.3
2.1
8.9
117.0
3.1
8.4
7.7
6.0
1.7
6.0
57.7
48.7
5.1
83.5
HCstc
0.41
1.70
0.41
0.41
0.41
0.41
0.80
0.41
1.70
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstc
7.0
18.0
3.4
7.0
3.4
3.4
10.0
3.4
18.0
7.0
3.4
10.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
10.0
3.4
3.4
3.4
3.4
7.0
3.4
3.4
7.0
XL05HC
33
1096
527
738
44
320
13
581
272
389
1
167
292
6
1742
1763
1
7
87
0
190
1418
34
14
99
328
389
61
	 9£
XL05CO
0.0
10.0
8.6
10.0
0.1
10.0
0.0
0.6
9.7
7.7
0.0
3.2
6.7
0.0
3.4
10.0
0.0
0.0
2.1
0.0
0.4
6.9
0.2
0.0
0.1
4.7
0.2
0.0
	 2.4
                         104

-------
Veh
030
031
032
033
034
035
036
037
038
039
040
041
042
043
044
045
046
047
048
049
050
051
052
053
054
055
056
057
058
059
060
101
102
Mfr
VW
AMC
AMC
VW
VW
MAZD
AMC
SUBA
MAZD
MAZD
VW
VW
AMC
VW
MAZD
VW
AMC
VW
AMC
AMC
VW
VW
SUBA
SUBA
SUBA
MAZD
SUBA
SUBA
MAZD
MAZD
MAZD
CHRY
CHRY
Model
RABB
EAGL
SPIR
RABB
RABB
Gt£
SPIR
GL10
323
RX7
RABB
RABB
181
QUAN
GL£
RABB
ALLI
RABB
181
BJOO
VANO
RABB
GL
GL
DL
626
GL10
GL10
RX7
323
626
FIFT
ARIE
F/M
PFI
CARB
CARB
PFI
PFI
CARB
CARB
PFI
PFI
PFI
PFI
CARB
PFI
PFI
CARB
TBI
TBI
CARB
PFI
TBI
TBI
CARB
CARB
TBI
CARB
CARB
PFI
PFI
PFI
PFI
CARB
CARB
CARB
MY
82
82
82
81
84
85
82
85
86
85
81
82
81
82
83
82
83
83
81
85
84
84
82
86
82
84
85
85
84
86
83
83
81
CID
105
258
151
105
109
91
151
109
98
80
105
105
101
105
91
105
85
105
101
85
117
105
109
109
109
122
109
109
80
98
122
318
135
Type
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
44
53
88
66
69
49
49
29
37
18
84
76
36
70
92
56
72
93
58
47
81
66
82
43
108
71
39
45
84
16
55
69
79
AETHC
150
578
232
223
119
228
237
465
236
1743
95
1478
338
163
350
114
541
364
376
486
225
423
294
179
0
176
141
802
477
1047
417
244
228
AETCO
1.4
0.4
6.2
0.0
4.0
6.4
3.9
0.4
0.1
0.1
1.4
0.1
0.7
8.5
1.9
6.2
0.6
0.4
0.0
6.1
4.4
6.2
5.2
2.1
4.5
1.5
2.2
2.1
2.6
0.6
5.4
3.6
3.5
FTPHC
0.45
1.13
3.50
0.60
1.33
2.30
1.35
0.20
0.43
0.26
0.95
5.21
1.93
1.85
1.08
1.92
7.98
2.11
0.57
0.74
0.71
1.98
4.24
0.16
1.97
0.62
0.27
0.20
3.40
0.28
1.05
1.31
1.42
FTPCO
5.4
14.4
104.3
3.1
15.2
54.9
28.8
2.6
3.0
1.6
13.9
30.3
14.0
72.2
13.7
69.0
37.5
25.6
4.5
10.1
7.9
35.2
34.1
3.6
47.8
10.6
4.9
4.1
40.9
4.0
14.5
14.5
33.1
HCstd
0.41
1.70
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstd
3.4
18.0
7.0
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
10.0
3.4
7.0
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
7.0
XL05HC
130
26
196
139
231
598
266
53
143
2
150
1131
299
216
53
142
2000
520
17
518
2
98
540
21
746
37
19
35
201
111
23
55
20
XL05CO
1.2
0.0
4.8
0.0
8.6
9.5
3.8
0.3
0.2
0.0
3.8
0.2
0.7
10.0
0.1
5.0
3.3
0.9
0.0
9.8
0.0
1.6
8.4
0.1
10.0
0.0
0.2
0.1
0.8
0.2
0.0
0.8
0.0
105

-------
Veh
103
104
105
107
109
110
111
112
113
114
115
116
117
201
202
203
204
205
206
207
208
209
210
211
212
213
214
216
217
218
219
220
221
Mfr
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
CHRY
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
Model
CHAR
OMNI
CARA
RELI
DAYT
LEBA
600
RELI
LANC
LEBA
NEW
ARIE
RED
LINC
TOVW
LINC
CAPR
MUST
F150
RANG
TEMP
MARQ
TEMP
TOPA
RANG
TOPA
TEMP
TEMP
MARQ
LTD
MUST
ZEPH
RANG
F/M
CARB
CARB
TBI
CARB
PFI
PFI
TBI
TBI
PFI
PFI
PFI
CARB
CARB
CARB
TBI
TBI
CARB
CARB
CARB
TBI
CARB
CARB
TBI
CARB
TBI
TBI
TBI
TBI
CARB
CARB
CARB
CARB
CARB
MY
85
84
85
81
84
86
84
85
85
86
85
81
81
82
82
83
81
81
86
85
84
81
85
84
85
86
86
85
82
81
81
81
84
CID
135
135
135
135
135
135
135
135
135
135
135
135
135
302
302
302
140
140
300
140
140
302
140
140
140
140
140
140
302
302
140
140
140
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LOT
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
KMile
29
44
29
64
41
41
54
32
41
36
40
34
49
59
66
68
75
150
16
61
53
67
33
68
49
13
15
58
92
70
68
117
44
AETHC
292
391
150
267
152
364
537
175
242
306
872
100
334
881
975
575
321
187
250
600
244
230
518
1382
744
292
105
1333
147
640
422
183
124
AETCO
3.8
2.6
1.5
6.3
1.3
0.7
6.8
2.8
2.2
0.3
10.0
3.6
7.8
1.9
10.0
1.5
2.3
3.7
0.0
0.7
0.2
5.2
5.9
10.0
. 0.6
4.5
1.8
8.1
1.7
0.8
3.6
3.2
1.4
FTPHC
1.49
1.14
3.63
1.19
1.17
0.51
3.00
2.10
0.51
0.65
8.73
1.19
2.81
9.33
19.07
3.31
1.04
3.57
0.62
1.79
3.02
1.04
2.68
6.17
1.18
0.65
0.80
2.15
1.81
2.53
4.90
1.66
0.54
FTPCO
31.8
6.3
63.7
8.1
14.0
4.8
37.5
61.4
5.3
7.1
86.3
35.4
51.3
65.2
62.6
24.9
7.5
66.5
0.9
10.5
47.3
23.6
56.3
58.7
5.9
17.0
23.7
30.1
34.9
39.3
64.9
28.3
4.1
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.80
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
COstd
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
7.0
7.0
7.0
7.0
3.4
3.4
3.4
10.0
10.0
3.4
3.4
3.4
3.4
10.0
3.4
3.4
3.4
7.0
3.4
3.4
3.4
10.0
XL05HC
231
#N/A
406
26
136
48
2024
34
20
165
502
44
345
165
910
75
109
461
60
380
279
38
43
300
500
58
43
1205
211
25
557
155
40
XL05CO
4.0
#N/A
7.1
0.2
1.0
0.0
7.1
0.4
0.0
0.2
7.1
0.6
6.8
2.0
8.3
0.1
0.0
7.8
0.0
0.0
0.0
0.0
0.0
2.6
0.0
0.0
0.0
7.7
4.9
0.0
3.4
0.7
0.0
106

-------
Veh
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
Mfr
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
Model
LTD
MARQ
MARQ
MUST
MARQ
MUST
SABL
TOPA
LTD
TOPA
TEMP
MARQ
TEMP
TEMP
LTD
TEMP
TEMP
F150
RANG
TEMP
OOUN
TEMP
MUST
TAUR
TEMP
MUST
TEMP
F250
RANG
SABL
EXP
TEMP
MUST
F/M
CARS
CARB
TBI
CARB
CARB
CARB
PFI
TBI
CARB
CARB
TBI
CARB
TBI
TBI
CARB
TBI
TBI
CARB
TBI
TBI
CARB
TBI
CARB
TBI
TBI
CARB
TBI
CARB
CARB
TBI
TBI
TBI
CARB
MY
81
82
86
81
82
81
86
86
81
84
85
82
85
85
82
85
86
86
86
85
81
85
81
86
85
81
85
82
83
86
84
86
81
CID
255
302
302
140
302
140
183
140
302
140
140
302
140
140
302
140
140
300
177
140
302
140
140
150
140
140
140
351
122
183
98
140
140
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LOT
LDV
LDV
LDV
LDV
KMile
80
46
28
93
54
59
47
5
41
55
25
88
58
21
33
43
16
36
27
21
73
45
76
17
42
70
25
45
70
14
46
10
65
AETHC
293
161
449
327
153
132
459
144
1400
1020
312
596
326
946
268
485
313
58
309
411
117
938
312
373
644
306
262
1126
214
126
164
651
102
AETCO
5.2
1.3
0.1
2.3
1.9
1.5
5.2
3.0
8.5
0.5
3.5
7.9
3.9
10.0
3.9
0.8
3.6
2.2
0.7
3.7
3.0
4.2
5.6
4.6
5.1
0.0
0.5
0.0
3.5
1.3
1.6
8.8
2.8
FTPHC
3.03
0.33
0.39
2.14
1.82
3.71
0.25
0.36
1.51
1.54
0.47
5.19
3.84
0.82
1.98
0.60
6.75
2.30
0.91
0.29
0.57
3.49
1.83
0.73
3.23
1.33
0.61
0.88
1.44
0.16
1.07
3.60
0.38
FTPCO
42.9
1.8
2.0
23.4
27.1
63.9
4.2
10.2
27.4
42.9
10.8
65.6
17.5
17.3
30.4
17.3
63.2
63.4
7.1
6.7
9.5
24.0
36.2
17.1
65.9
23.8
14.7
5.1
11.8
2.2
18.0
65.2
8.2
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
1.70
1.70
0.41
0.41
0.41
0.41
COstd
3.4
7.0
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
7.0
3.4
3.4
7.0
3.4
3.4
10.0
10.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
18.0
18.0
3.4
3.4
3.4
3.4
XL05HC
72
71
40
588
63
144
176
27
28
87
26
452
409
926
29
14
147
50
974
34
0
965
436
51
164
31
62
147
64
21
122
826
84
XL05CO
0.1
0.0
0.0
0.3
0.0
1.2
0.0
0.1
0.0
0.0
0.0
6.2
0.1
6.6
0.0
0.0
0.1
1.1
7.8
0.2
0.0
5.1
5.5
0.0
0.8
0.0
0.0
0.0
0.0
0.0
0.0
8.7
0.0
107

-------
Veh
257
258
259
260
301
302
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
Mfr
FORD
FORD
FORD
FORD
QA
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
Model
MUST
MUST
CAPR
RANG
RIVI
REGA
MALI
BONN
CELE
MALI
RIVI
CIER
PHOE
MALI
SKYH
FIER
SEVI
CORV
CUTL
FIER
FIRE
CENT
CELE
DEVI
CAVA
RIVI
REGA
CIER
CITA
BONN
2000
GRAN
CIER
F/M
CARB
CARB
CARB
CARB
CARB
CARB
CARB
CARB
TBI
CARB
CARB
TBI
CARB
CARB
TBI
TBI
TBI
TBI
CARB
TBI
TBI
CARB
CARB
TBI
TBI
CARB
CARB
TBI
CARB
CARB
TBI
CARB
TBI
MY
81
81
81
84
85
82
82
82
86
81
81
84
81
81
83
84
81
82
81
84
85
81
83
82
86
81
82
84
81
82
83
84
84
CID
140
140
140
122
307
231
229
231
151
229
250
151
151
229
110
151
368
350
231
151
110
231
173
250
121
250
231
151
151
231
110
305
151
Type
LDV
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
78
94
83
35
10
98
83
65
5
102
68
61
64
43
58
49
54
49
57
44
27
62
116
60
18
72
113
35
140
55
59
37
60
AETHC
366
263
233
146
570
298
263
341
225
399
805
500
184
502
357
265
1167
306
602
227
167
1639
309
264
239
249
2000
244
157
172
367
252
325
AETCO
5.4
0.0
0.1
2.6
4.4
0.4
0.4
0.3
0.4
1.2
10.0
0.5
1.8
0.6
0.5
0.3
1.4
4.4
3.2
1.7
2.5
0.2
1.7
0.0
0.9
0.9
0.3
0.6
1.6
1.5
0.2
0.4
0.3
FTPHC
3.30
0.57
2.20
0.73
0.70
4.06
2.61
3.97
0.21
0.56
0.84
0.25
1.37
0.59
0.93
1.29
3.72
3.11
1.59
0.55
7.65
4.13
1.82
0.33
0.16
1.52
4.83
0.85
3.40
2.32
0.97
2.01
0.24
FTPCO
49.2
4.6
41.4
8.0
2.1
32.3
15.9
22.5
1.8
5.7
8.4
3.3
20.9
9.8
6.5
37.3
35.3
42.0
20.9
3.8
129.2
29.4
7.3
5.2
1.8
23.4
28.2
6.4
34.8
15.7
7.4
15.1
2.9
HCstd
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstd
3.4
3.4
3.4
10.0
3.4
7.0
7.0
7.0
3.4
7.0
3.4
3.4
3.4
7.0
3.4
3.4
7.0
7.0
7.0
3.4
3.4
7.0
3.4
7.0
3.4
3.4
7.0
3.4
3.4
7.0
3.4
3.4
3.4
XL05HC
450
61
49
80
1 1
254
200
754
14
19
151
84
219
338
87
45
1125
261
57
29
643
870
102
18
9
556
1885
121
507
19
173
196
41
XL05CO
5.7
0.0
0.0
0.0
0.0
2.3
0.4
0.3
0.1
0.0
0.6
0.2
6.5
0.3
0.0
0.1
1.5
1.4
0.2
0.1
10.2
0.2
0.0
0.0
0.0
7.2
2.1
0.3
9.1
0.0
0.3
0.5
0.1
108

-------
Veh
331
332
333
334
335
336
337
338
339
340
341
342
343
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
401
402
403
404
Mfr
CM
Ovl
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
HOND
HCND
HOND
HOND
Model
MONT
CMBG
CELE
REGA
CIMA
GRAN
CITA
GVE3
REGA
SKYL
CITA
DEVI
CENT
RIVI
J200
CITA
GRAN
SUNB
SUNB
DEVI
CELE
CITA
CAVA
SKYH
CAVA
GRAN
CAVA
DEVI
CIER
CIVI
CIVI
CIVI
CIVI
F/M
CARB
CARB
CARB
CARB
CARB
CARB
CARB
TBI
CARB
CARB
CARB
TBI
TBI
CARB
CARB
TBI
TBI
TBI
TBI
TBI
TBI
PFI
TBI
TBI
TBI
TBI
TBI
TBI
TBI
PFI
PFI
CARB
CARB
MY
83
81
82
81
82
81
81
82
83
81
81
81
86
81
82
84
86
86
84
83
85
85
84
84
85
86
84
83
82
86
85
84
84
CID
229
151
173
231
112
231
173
151
231
173
151
368
151
307
112
151
151
110
110
249
151
173
121
110
121
151
121
249
151
91
91
91
91
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
46
44
82
50
78
86
72
53
84
55
84
29
37
61
91
25
17
27
47
54
52
23
51
66
42
37
8
92
56
41
58
53
71
AETHC
1261
209
248
304
174
557
904
517
134
68
193
331
452
393
192
331
223
240
237
277
284
687
177
218
395
548
315
520
1062
259
327
227
383
AETCO
6.1
1.5
4.2
0.4
5.8
7.6
0.2
0.2
1.5
1.2
4.9
0.2
0.5
2.5
4.2
0.7
0.1
0.3
0.1
0.4
0.6
0.3
2.6
2.2
3.5
0.7
1.2
0.1
1.0
1.4
0.7
0.0
0.0
FTPHC
6.42
0.51
0.70
1.34
1.56
5.92
0.64
0.28
0.98
0.25
2.12
0.57
1.10
4.94
1.81
0.21
0.40
0.28
0.38
6.45
0.34
0.64
7.64
2.96
0.43
0.35
6.09
1.44
0.29
0.35
0.93
1.46
1.19
FTPCO
93.2
14.3
10:6
14.3
16.9
93.2
5.0
2.7
5.7
8.0
72.2
9.6
5.2
68.9
28.6
2.6
1.2
3.9
8.4
7.0
4.3
6.1
136.5
55.0
7.6
3.3
38.4
5.8
4.2
3.5
8.6
4.5
6.5
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstd
3.4
3.4
7.0
7.0
7.0
7.0
7.0
7.0
3.4
7.0
3.4
7.0
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
XL05HC
774
20
379
155
405
1927
34
179
181
26
230
22
287
214
274
205
123
39
153
192
13
10
260
435
133
36
236
53
33
44
371
160
172
XL05CO
10.4
0.0
6.5
0.2
9.6
8.6
0.0
0.2
1.1
0.1
5.6
0.0
0.3
0.4
5.7
0.3
0.1
0.1
0.6
0.2
0.0
0.1
6.1
9.1
0.6
o.i
1.5
0.0
0.2
0.1
4.3
0.1
0.1
109

-------
Veh
405
406
407
408
409
410
502
503
504
505
506
507
508
509
510
511
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
Mfr
HGND
HCND
HOND
HCND
HOD
HOND
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
MITS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
NISS
Model
CIVI
CIVI
CIVI
CIVI
ACCO
CIVI
COLT
RAM
COLT
COLT
RAM
COLT
COLT
COLT
COLT
COLT
STAN
200S
PULS
SENT
MAX!
MAX!
200S
200S
280Z
300Z
300Z
280Z
TRUC
PULS
280Z
SENT
280Z
F/M
CARB
CARB
CARB
PFI
CARB
PFI
CARB
CARB
CARB
CARB
CARB
TBI
TBI
TBI
TBI
TBI
PFI
PFI
CARB
CARB
PFI
PFI
PFI
PFI
PFI
PFI
PFI
PFI
TBI
PFI
PFI
CARB
PFI
MY
84
84
85
86
85
85
84
85
85
84
85
85
84
85
86
86
86
82
85
84
82
83
81
81
82
84
84
81
86
83
82
84
83
CID
91
91
91
91
112
91
86
122
90
98
122
98
98
97
98
98
120
134
98
98
146
146
119
119
171
181
181
171
146
91
168
98
168
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LDV
LDV
KMile
85
77
13
35
22
53
65
50
34
59
28
21
58
26
39
16
33
95
50
35
91
69
119
90
73
42
24
57
28
55
102
48
74
AETHC
798
565
248
396
244
224
1061
493
24
746
227
2000
108
307
203
127
269
194
279
228
357
132
122
469
299
184
237
268
61
1527
152
239
138
AETCO
0.1
0.0
0.1
3.0
1.8
0.6
0.1
7.8
2.3
0.3
2.7
4.9
1.7
6.7
1.5
2.1
7.0
1.8
0.4
0.1
2.8
2.5
2.1
0.4
0.7
1.7
1.7
4.4
2.3
9.5
1.7
0.2
1.8
FTPHC
2.14
2.05
1.73
0.43
1.75
0.68
1.00
4.77
0.21
1.18
4.44
0.22
0.39
0.24
0.23
0.20
1.15
2.61
2.20
0.36
4.52
0.66
1.20
1.60
0.74
0.57
0.46
5.02
0.61
5.09
1.69
0.16
0.86
FTPCO
4.7
7.4
4.8
4.0
34.0
6.8
8.2
53.3
9.3
6.6
54.2
4.8
6.3
1.4
4.3
3.0
45.8
91.5
48.7
4.8
142.2
9.1
16.9
6.8
12.1
4.3
5.3
125.1
98.9
43.7
54.2
3.6
11.7
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.41
0.41
COstd
3.4
3.4
3.4
3.4
3.4
3.4
3.4
10.0
3.4
3.4
10.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
7.0
3.4
3.4
3.4
3.4
3.4
10.0
3.4
3.4
3.4
3.4
XL05HC
123
500
348
48
291
65
250
593
35
114
600
0
28
2
42
0
276
1013
904
65
167
130
35
433
19
0
0
332
60
561
181
50
226
XL05CO
0.2
0.1
0.2
0.1
1.8
0.1
0.7
7.6
1.9
0.1
8.0
0.0
0.1
0.1
0.2
0.0
2.9
8.0
1.7
0.0
3.3
2.5
0.2
0.4
0.1
0.0
0.0
6.0
3.8
5.8
2.3
0.1
4.3
110

-------
Veh
618
619
620
701
702
703
704
705
706
707
708
709
Mfr
NISS
NISS
NISS
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
TOYT
Model
PULS
MAXI
200S
TERC
TERC
OORO
CELI
CELJ
OOflO
CORO
CAMR
CELI
F/M
PFI
PFI
PFI
CARB
GARB
CARB
PFI
PFI
CARB
CARB
PFI
CARB
MY
83
83
84
86
84
81
86
83
82
82
84
83
CIO
91
146
120
91
91
108
122
144
108
108
122
144
Type
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KMile
59
61
28
5
77
96
32
59
234
147
32
49
AETHC
446
281
293
262
339
167
335
293
445
1473
344
2000
AETCO
6.9
0.3
7.4
0.1
1.2
1.9
0.4
0.2
0.1
0.2
0.2
0.8
FTPHC
1.88
0.68
2.01
0.24
1.14
2.12
0.26
0.44
2.26
0.51
0.16
0.19
FTPCO
44.4
7.0
72.3
1.9
3.8
47.2
1.7
3.0
49.2
8.9
1.9
3.5
HCstd
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
COstd
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
XL05HC
547
1 7
250
7
45
128
0
141
109
1
0
0
XL05CO
5.6
0.0
6.0
0.0
0.0
0.8
0.0
0.1
1.1
0.0
0.0
0.0
111

-------
                                Appendix C
As Received  Failure Rates for Selected Modes of the Basic I/M Test  Procedure
VEHICLE
BITP
Mode
CS03
CS05
CS07
CS 10
CS 12
CS 15
XL 02
XL 05
XI 02
XI 05
XI 07
XI 10
RS02
RS05
Failure
COUNT
Type


HC-onlv HC+CO CO-only
49
39
19
20
33
35
19
18
39
41
22
20
16
21
131
68
68
61
74
65
64
63
78
59
63
58
64
64
29
31
26
25
24
20
21
16
23
22
27
23
18
19
Pass
27
98
124
131
106
117
134
141
97
115
125
136
134
128
VEHICLE PERCENT
BITP
Mode
CS03
CS05
CS07
CS 10
CS 12
CS 15
XL 02
XL 05
XI 02
XI 05
XI 07
XI 10
RS02
RS05
Failure Type
HC-only
21%
17%
8%
8%
14%
15%
8%
8%
16%
17%
9%
8%
7%
9%
HC+CO CO-only
56%
29%
29%
26%
31%
27%
27%
26%
33%
25%
27%
24%
28%
28%
12%
13%
11%
11%
10%
8%
9%
7%
10%
9%
11%
10%
8%
8%
Pass
11%
42%
52%
55%
45%
49%
56%
59%
41%
49%
53%
57%
58%
55%
                                     112

-------
                                               APPENDIX C
Emission Values for Calculating  As Received Failure Rates for Selected  Modes of the Basic I/M Test Procedure
HC values in ppm; (allure >=220 ppm
CO values in % concentration: failure >=1.2%
Veh
001
002
003
004
005
006
007
008
009
010
Oil
012
013
014
015
016
017
018
019
020
021
022
023
024
025
026
027
028
029
030
031
032
033
034
035
036
037
038
039
040
041
042
CS 03
HC CO
247 6.4
2000 8.4
715 7.4
282 8.2
1386 6.7
323 5.8
278 5.2
829 6.6
326 5.4
299 4.3
498 5.3
571 0.3
936 10.0
200 1.3
533 1.0
2000 9.3
818 7.6
963 10.3
503 8.0
647 9.2
462 1.6
530 10.0
242 2.1
256 1.3
1038 5.0
202 3.9
616 1.8
277 1.0
518 2.6
441 9.7
343 5.6
1902 10.0
700 2.7
532 10.0
396 5.5
414 4.0
1956 2.2
718 1.4
192 0.6
479 2.7
903 0.2
464 4.3
CS 05
HC CO
163 6.7
848 6.3
565 8.8
615 10.0
729 5.2
142 2.0
182 4.4
148 0.3
143 3.5
211 3.9
133 0.4
65 0.0
356 4.8
135 1.2
288 2.6
625 10.0
377 8.7
182 0.1
480 8.7
44 0.0
214 0.3
56 0.1
214 1.8
207 0.3
519 2.8
180 3.1
301 2.3
436 2.8
307 4.4
112 0.4
133 0.7
698 9.3
633 0.4
104 0.6
264 2.3
370 3.0
69 0.1
280 0.9
31 0.0
209 1.7
859 0.2
227 0.8
CS 07
HC CO
15 0.0
1255 10.0
356 5.8
565 10.0
143 0.1
197 7.4
9 0.0
29 0.5
127 3.4
545 8.0
9 0.0
169 3.2
242 2.6
59 0.1
2058 3.2
1259 10.0
151 0.1
6 0.1
140 2.6
4 0.0
49 0.0
13 0.0
142 0.7
72 0.0
150 0.1
261 3.1
218 0.2
370 3.0
139 4.4
118 0.5
16 0.0
293 8.0
131 0.0
118 3.2
1024 10.0
229 3.5
39 0.2
96 0.0
10 0.0
1 50 1 .6
2000 0.0
351 0.8
CS 10
HC CO
20 0.0
1202 10.0
471 5.8
S84 10.0
86 0.1
200 7.5
8 0.0
26 0.1
130 3.5
517 7.9
8 0.0
9 0.3
231 2.3
47 0.0
1955 3.2
1014 10.0
54 0.1
4 01
136 2.6
5 0.0
74 01
16 00
118 0.5
56 0.0
187 01
238 4.0
274 0.2
326 2.4
133 4.0
148 0.9
18 0.2
309 8.0
137 0.0
116 3.5
748 9.4
222 3.5
50 0.2
81 0.0
1 1 0.0
148 1.8
1539 0.2
280 0.8
CS 12
HC CO
53 00
1537 10.0
439 64
727 100
159 01
234 8.4
4 0.0
26 01
131 3.1
471 7.5
7 0.0
96 16
192 11
41 00
2005 3.7
2000 10.0
13 00
0 0.1
116 10
3 00
438 6.3
309 5.4
162 09
57 00
261 00
301 47
323 0.2
298 22
135 33
162 1.1
340 36
251 5.8
70 0.0
142 3.9
1270 10.0
196 23
62 0.1
36 0.0
11 0.0
142 2.1
752 0.3
133 0.7
CS 15
HC CO
63 0.0
1414 10.0
429 6.2
727 100
163 0.1
246 8.5
8 00
24 0.0
122 2.8
470 7.6
8 0.0
251 5.4
222 1.4
33 0.0
2002 3.5
2000 10.0
12 0.0
1 0.1
112 1.4
2 0.0
290 0.3
92 3.7
193 1.1
80 0.0
213 0.0
301 44
363 0.2
303 2.3
127 3.6
160 1.2
290 1.1
247 6.3
145 0.0
132 3.9
806 10.0
191 2.8
204 0.3
102 0.1
11 0.0
144 2.0
844 0.2
172 0.7
XL 02
HC CO
21 0.0
1203 10.0
453 8.5
700 100
8 00
386 10.0
15 0.0
319 0.5
277 9.7
402 7.6
3 0.0
177 2.6
239 6.9
8 00
1602 3.3
2000 10.0
1 0.0
5 0.0
99 1.6
1 0.0
226 0.7
1019 5.4
38 0.2
22 0.0
157 0.1
247 5.5
357 0.2
56 0.1
99 2.4
98 0.8
20 0.0
196 ' 4.8
138 0.0
230 8.4
669 10.0
242 4.0
105 0.3
190 0.2
3 0.0
129 3.6
1079 0.2
182 0.6
XL 05
HC CO
33 0.0
1096 10.0
527 8.6
738 10.0
44 0.1
320 10.0
13 00
581 0.6
272 9.7
389 77
1 0.0
167 3.2
292 6.7
6 0.0
1742 3.4
1763 100
1 0.0
7 0.0
87 2.1
0 0.0
190 0.4
1418 6.9
34 0.2
14 00
99 0.1
328 4.7
389 0.2
61 0.0
96 2.4
130 1.2
26 0.0
196 4.8
139 0.0
231 8.6
598 9.5
266 3.8
53 0.3
143 0.2
2 0.0
150 3.8
1131 0.2
299 0.7
XI 02
HC CO
118 0.1
1686 10.0
472 7.6
847 10.0
203 1.2
632 10.0
1 0.0
776 0.3
110 3.1
389 6.8
214 4.9
141 1.5
152 1.2
30 0.0
2000 5.6
1100 10.0
351 6.9
9 0.0
88 0.5
1 00
353 55
889 6.1
127 0.8
264 0.7
266 0.2
329 3.7
335 0.2
43 0.0
105 3.0
165 2.3
775 1.8
203 4.2
63 0.0
174 7.2
525 9.3
281 3.7
182 0.2
1 0.0
4 0.0
156 3.5
1304 0.2
165 0.7
XI 05
HC CO
56 0.0
1343 10.0
813 6.9
701 10.0
169 0.2
710 10.0
It 00
326 0.6
104 2.7
349 6.2
164 3.4
169 1.9
184 0.8
146 0.2
1821 3.8
1219 100
57 0.0
8 0.0
49 09
15 0.0
386 5.7
197 5.5
176 1.1
330 1.2
217 0.1
260 4.5
407 0.2
52 0.0
104 3.1
137 2.1
477 0.4
211 4.4
168 0.0
167 7.0
169 2.4
267 4.0
164 0.3
325 0.2
6 0.0
151 3.5
1090 0.2
162 0.6
XI 07
HC CO
23 0.0
1293 100
487 80
695 10.0
58 00
277 95
6 00
68 00
106 31
275 5.3
1 1 00
110 1.5
282 6.0
135 06
1819 37
1164 10.0
4 00
2 0.0
55 18
6 0.0
204 0.9
808 62
105 0.4
99 0.1
169 0.1
285 5.0
325 0.2
77 0.2
99 3.4
140 1.2
27 0.0
211 5.2
201 0.0
140 6.2
594 9.2
250 3.3
61 0.2
192 01
5 0.0
163 3.8
1224 0.2
203 0.7
XI 11
HC CO
28 0.0
1564 10.0
509 8.3
718 10.0
71 0.1
435 10.0
8 0.0
58 0.1
108 3.1
290 5.3
1 1 0.0
195 2.6
265 6.0
69 0.2
1753 3.5
1274 10.0
3 0.0
3 00
67 2.1
5 0.0
193 0.6
894 6.2
73 0.3
53 0.3
142 0.1
246 5.8
311 0.2
58 0.0
102 3.1
131 1.1
24 0.0
225 5.1
200 0.0
152 6.5
614 10.4
258 3.5
57 0.2
282 0.2
5 0.0
164 4.0
1136 0.2
166 0.7
RS 02
HC 00
51 0.1
1363 10.0
431 9.3
1015 10.0
14 0.0
581 10.0
3 0.0
432 2.0
109 3.1
271 5.6
0 0.0
1 0.0
254 6.6
35 0.0
1690 3.6
1794 10.0
6 00
4 0.0
89 1.5
1 0.0
185 0.6
663 6.6
27 0.1
6 0.0
133 0.1
280 3.3
443 0.2
38 0.0
103 3.2
131 2.7
137 0.3
228 5.1
112 0.0
733 9.6
653 10.2
216 3.6
57 0.3
110 0.2
3 0.0
241 9.4
1338 0.2
169 0.7
RS05
HC CO
100 0.2
1383 4.0
446 9.3
1051 10.0
13 00
428 10.0
5 00
722 0.4
99 2.9
267 5.6
2 0.0
126 0.6
273 6.2
54 0.0
1744 3.5
2000 10.0
6 0.0
2 0.0
82 2.0
0 0.0
192 0.5
444 6.7
64 0.2
33 0.0
100 0.0
322 5.2
393 0.2
37 0.0
100 3.2
153 2.4
15 0.0
220 5.4
208 0.0
660 10.0
643 10.0
246 3.5
3 0.0
165 0.2
3 0.0
246 9.1
1123 0.2
245 0.7
                                                    113

-------
Veh
043
044
045
046
047
048
049
050
051
052
053
054
055
056
057
058
059
060
101
102
103
105
107
109
110
111
112
113
114
115
116
117
201
202
203
204
205
206
207
208
209
210
211
212
213
214
216
217
218
CS 03
HC 00
554 10.0
2000 6.6
517 10.0
2000 4.7
379 1.2
444 1.7
1121 9.9
- 332 7.8
709 5.8
2000 6.8
273 0.4
167 3.8
543 8.8
419 2.6
434 2.3
680 3.2
368 2.7
377 2.9
342 7.1
185 4.2
502 7.1
220 3.3
502 7.1
105 0.7
206 0.2
270 6.2
502 7.0
154 0.3
295 0.5
243 4.7
460 5.0
502 6.9
288 4.3
507 7.6
77 2.5
167 3.2
325 7.1
956 1.5
317 1.5
1851 6.6
43 1.0
206 2.0
367 2.9
657 0.5
271 2.0
537 6.5
305 1.7
93 1.3
48 0.6
CS 05
HC CO
291 8.0
1752 0.0
202 3.8
2000 4.8
455 1.2
152 0.1
873 8.3
187 4.9
650 5.7
197 3.2
118 0.3
99 2.1
246 2.4
58 0.1
103 0.4
452 2.3
140 0.1
361 5.7
72 1.3
94 0.0
437 1.0
226 6.0
121 0.3
114 1.5
233 1.0
266 5.9
308 7.1
105 1.5
76 0.1
502 7.1
123 1.2
189 1.2
102 1.1
1131 8.3
80 0.6
128 0.3
293 6.0
163 0.0
257 0.7
387 0.1
0 0.0
155 0.5
166 2.2
562 0.5
366 4.3
213 0.3
60 0.0
270 6.2
12 0.2
CS 07
HC CO
274 9.7
62 0.0
140 40
2000 2.8
430 1.3
0 0.0
353 4.2
114 0.2
379 4.2
412 6.2
132 0.2
298 7.8
59 0.1
13 0.0
17 0.0
216 0.7
278 0.4
45 0.0
25 0.0
12 0.0
274 5.3
387 7.1
124 3.4
87 0.8
34 0.0
461 7.1
33 0.1
65 0.3
49 0.2
502 7.1
41 1.6
282 5.4
959 8.7
1298 8.3
56 0.3
142 0.0
435 7.9
32 0.0
300 0.1
290 0.0
420 6.5
26 0.1
417 4.3
386 0.0
50 0.0
41 0.0
1161 8.3
1018 3.1
77 2.3
CS 10
HC CO
274 9.9
54 0.0
147 4.2
2000 2.9
376 1.1
0 0.0
319 3.0
99 0.1
192 1.7
296 6.2
65 0.2
273 7.4
60 0.0
14 0.0
19 0.0
217 0.7
120 0.2
33 0.0
20 0.0
11 0.0
251 4.9
502 7.1
79 1.6
91 0.7
31 0.0
406 7.1
22 0.2
12 0.1
207 0.4
502 7.1
42 1.7
286 5.7
88 1.6
890 8.3
57 0.1
134 0.0
436 8.1
57 0.0
370 0.1
302 0.0
32 0.0
18 0.0
285 2.5
401 0.0
53 0.0
115 0.6
1139 8.3
201 4.7
15 0.0
CS 12
HC CO
248 10.0
67 0.0
140 4.7
2000 2.7
823 03
0 0.0
305 2.7
43 0.0
117 0.9
239 4.1
22 0.0
301 7.6
55 0.1
58 0.0
35 0.0
247 0.6
169 0.1
63 0.0
10 0.0
38 0.0
280 4.4
502 7.1
21 0.9
132 1.0
44 0.1
337 6.7
7 0.1
82 0.1
74 0.2
502 7.1
26 1.1
251 6.5
1343 8.7
1168 8.3
71 0.1
696 7.0
440 7.9
120 0.0
997 4.6
79 0.1
201 2.3
68 0.0
649 4.1
750 0.1
269 3.5
462 4.8
1887 8.3
269 4.1
81 2.4
CS 15
HC CO
248 10.0
41 0.0
154 4.8
2000 2.9
798 0.3
0 0.0
293 2.5
22 0.0
144 1.1
314 6.5
102 0.1
324 8.0
57 0.0
41 0.0
84 0.2
245 0.6
141 0.2
35 0.0
6 0.0
10 0.1
283 4.7
502 7.1
46 0.1
119 0.9
28 0.0
438 7.1
19 0.2
36 0.0
282 0.3
502 7.1
85 2.4
378 6.8
165 2.4
972 8.1
74 0.1
356 1.5
405 7.6
116 0.0
697 0.7
733 0.0
36 0.0
35 0.0
322 2.3
818 0.7
95 0.0
192 0.5
1342 8.3
228 4.2
72 0.0
XL 02
HC CO
209 9.9
60 0.3
136 5.5
1884 3.7
491 1.0
14 0.0
497 9.8
3 0.0
89 1.9
505 7.9
32 0.1
693 10.0
36 0.0
20 0.1
109 0.2
173 0.8
270 0.4
64 0.0
46 0.9
34 0.0
225 3.5
192 4.4
11 0.1
144 1.6
82 0.1
1692 6.7
61 O.S
98 0.2
67 0.1
502 7.1
37 0.8
1343 6.2
1082 8.7
1049 8.3
76 0.1
109 0.0
534 8.4
67 0.0
347 0.0
344 0.0
36 0.0
54 0.0
361 2.3
414 0.0
59 0.0
71 0.0
1754 8.3
263 5.1
57 0.6
XL 05
HC 00
216 10.0
53 01
142 50
2000 3.3
520 0.9
17 0.0
518 9.8
2 0.0
98 1.6
540 8.4
21 0.1
746 10.0
37 0.0
19 02
35 0.1
201 0.8
111 0.2
23 0.0
55 0.8
20 0.0
231 4.0
406 7.1
26 0.2
136 1.0
48 0.0
2024 7.1
34 0.4
20 0.0
165 0.2
502 7.1
44 0.6
345 6.8
165 2.0
910 8.3
75 0.1
109 0.0
461 7.8
60 0.0
380 0.0
279 0.0
38 0.0
43 0.0
300 2.6
500 0.0
58 0.0
43 0.0
1205 7.7
211 4.9
25 0.0
XI 02
HC CO
202 10.0
109 0.1
136 6.0
2000 2.7
756 08
26 0.0
369 7.0
14 0.0
98 0.9
233 51
381 3.6
390 8.8
47 0.1
272 4.0
64 0.0
221 05
158 0.2
53 0.0
29 0.5
28 0.0
260 3.6
21 0.0
26 0.7
128 0.8
195 0.3
312 6.1
0N/A »N/A
86 0.2
24 0.2
466 7.1
45 1.6
218 6.2
1348 8.7
1258 8.3
83 0.1
621 6.7
372 7.0
214 0.0
769 0.7
449 2.1
171 1.6
47 0.0
241 2.9
960 1.3
228 1.9
327 2.7
1887 8.2
267 5.1
63 1.2
XI 05
HC CO
199 9.6
52 0.0
131 4.9
2000 3.0
526 1.0
31 0.0
327 4.7
16 0.0
114 1.2
338 7.1
273 1.0
387 8.8
44 0.0
157 0.1
81 0.1
219 0.6
104 0.2
45 0.0
48 0.9
6 0.0
252 3.7
151 1.7
80 0.4
121 0.9
157 0.6
437 7.1
»N/A *N/A
58 0.1
187 0.2
502 7.1
89 2.6
475 6.9
229 2.5
1580 8.2
81 0.1
333 1.4
376 6.9
158 0.0
631 0.7
286 0.1
38 0.0
57 0.0
146 1.5
794 0.7
88 0.0
97 0.0
926 4.4
203 4.3
27 0.0
XI 07
HC 00
204 9.4
64 0.0
120 4.6
2000 3.6
276 1.4
12 0.0
270 4.9
12 0.0
98 1.0
315 6.9
87 0.2
404 9.1
44 0.1
24 0.0
46 0.2
183 0.7
62 0.2
36 0.0
38 2.0
8 0.0
194 4.1
153 3.0
179 3.1
108 1.0
30 0.
338 7
0N/A ON/A
1 1 0.
50 0.
502 7.
62 1.9
325 6.5
1111 8.7
1536 8.3
73 0.1
254 0.1
370 7.2
65 0.0
462 0.0
180 0.1
218 2.3
57 0.0
722 4.1
576 0.0
70 0.0
68 0.0
807 6.8
260 5.6
66 1.3
XI 11
HC CO
211 9.5
57 0.1
127 3.9
2000 29
362 1.4
1 1 0.0
285 5.0
12 0.0
87 0.9
303 6.9
77 0.2
396 8.8
42 0.0
21 0.0
28 0.2
178 0.7
63 0.2
54 0.0
43 1.9
7 0.0
192 4.0
173 2.9
62 0.5
112 1 .0
28 0.0
284 7.0
»N/A «N/A
8 0.1
28 0.0
502 7.1
53 1.7
356 6.7
135 2.3
1153 8.3
66 0.1
211 0.0
427 7.5
59 0.0
484 0.0
201 0.1
44 0.0
39 0.0
467 2.5
544 0.0
81 0.3
55 0.0
985 5.9
224 4.5
27 0.0
RS 02
HC CO
181 10.0
41 0.0
135 6.1
1875 4.8
486 1.3
3 0.0
1146 10.0
13 0.0
»N/A 0N/A
351 7.1
577 8.3
333 6.3
22 0.0
17 0.1
72 0.2
193 0.8
531 0.8
28 0.0
78 3.0
39 0.1
216 3.6
96 0.5
24 0.9
130 1.1
42 0.0
295 5.9
22 04
11 0.2
162 0.1
502 7.1
53 1.6
214 5.6
1115 8.7
1177 8.2
263 0.4
167 0.0
572 8.4
93 0.0
382 0.1
201 0.1
35 0.0
76 0.0
»N/A »N/A
572 0.1
47 0.0
48 0.1
703 7.2
238 4.8
54 0.5
RS 05
HC CO
189 9.4
29 0.0
134 4.6
2000 4.6
359 1.4
9 0.0
1170 10.0
14 0.0
*N/A «N/A
335 7.1
487 7.1
298 5.6
22 0.0
26 0.1
1 1 0.1
311 1.5
118 0.3
33 0.0
71 1.3
18 0.0
217 3.7
348 7.0
14 0.1
122 1.0
51 0.0
313 6.7
36 0.3
13 0.2
189 0.2
502 7.1
68 18
248 6.0
1083 8.7
1088 8.3
84 0.1
189 0.0
444 8.0
94 0.0
358 0.0
154 0.1
160 2.3
48 0.0
»N/A »N/A
421 0.0
46 0.0
41 0.0
750 5.7
274 5.5
50 05
114

-------
Veh
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
257
258
259
260
301
302
304
305
306
307
308
309
310
CS 03
HC 00
632 83
259 32
206 33
451 24
266 0.3
238 0.1
222 1.2
875 2.7
259 3.7
678 8.4
115 2.0
817 2.0
127 4.4
312 2.7
421 1.6
248 0.2
565 5.5
32 0.0
81 1.3
382 3.5
54 2.2
125 1.9
97 1.3
356 3.2
361 5.5
354 4.6
404 5.9
450 5.4
97 2.1
79 0.0
110 1.4
112 2.1
199 0.3
175 3.0
727 7.0
109 1.0
2098 2.3
90 0.4
363 7.0
265 0.3
105 0.9
477 3.1
1970 9.5
678 4.2
234 0.2
302 0.2
199 0.2
340 0.3
96 2.8
CS 05
HC CO
743 75
308 1.6
235 03
1248 7.2
239 0.0
87 02
154 0.5
191 06
208 2.4
49 0.0
SO 0.1
28 0.0
102 0.0
218 0.4
323 3.0
263 0.3
759 7.9
20 0.0
86 0.6
157 0.0
115 2.7
31 0.0
100 0.7
13 0.0
509 0.8
167 3.1
636 8.4
396 5.3
92 1.9
222 0.0
52 0.0
44 0.0
11 0.0
206 2.6
514 3.0
28 0.0
859 4.3
41 0.0
339 4.2
116 0.1
23 0.0
472 2.0
304 0.3
1266 0.1
231 0.1
232 0.7
212 0.4
207 0.6
209 3.7
CS 07
HC CO
439 33
217 28
55 0.1
56 0.4
594 5.9
8 0.0
122 1.0
219 2.7
201 1.5
50 0.0
14 0.0
433 7.3
68 0.0
59 0.0
380 5.4
344 1.9
576 5.7
107 1.8
19 0.0
48 0.0
61 1.7
100 0.0
63 0.7
2 0.0
998 4.8
326 6.2
44 0.0
715 7.0
3 0.0
67 0.0
144 0.0
120 1.2
0 0.0
452 0.2
804 8.7
28 0.0
565 6.7
49 0.0
67 0.0
46 0.0
537 5.0
312 3.5
169 0.3
627 0.4
23 0.1
39 0.0
76 0.2
156 0.2
154 3.7
CS 10
HC CO
612 3.5
243 2.8
49 0.0
138 1.8
296 0.0
8 0.0
128 1.0
185 0.0
206 1.6
78 0.0
16 0.0
38 0.0
93 0.0
33 0.0
398 6.0
355 2.1
634 6.0
30 0.0
17 0.0
39 0.0
61 1.5
82 0.0
62 0.7
3 0.0
989 5.2
334 6.2
150 0.9
577 5.7
1 0.0
75 0.0
78 0.0
132 1.6
5 00
280 1.6
781 8.7
37 0.0
555 6.8
59 0.0
57 0.0
46 0.0
4 0.0
268 1.3
168 0.4
973 0.6
5 0.0
28 0.0
27 0.0
126 0.2
157 3.1
CS 12
HC CO
528 3.2
76 00
161 18
97 0.2
555 3.9
846 1.0
144 1.1
162 1.4
367 2.6
144 0.0
154 2.5
312 6.3
116 0.0
99 0.0
476 6.0
664 5.0
875 6.9
118 13
20 0.0
719 6.8
S3 0.8
667 5.8
88 1.1
7 0.0
1583 7.2
398 6.0
375 3.8
929 8.3
3 0.0
163 0.0
308 3.1
220 2.7
1 1 0.0
321 1.1
1015 8.7
467 8.6
467 5.4
127 0.0
159 0.2
84 0.8
162 0.3
217 0.6
234 0.4
1588 1.1
22 0.1
46 0.0
78 0.1
150 0.2
176 3.5
CS 15
HC CO
610 2.8
193 2.3
110 1.1
155 11
314 0.0
41 0.0
113 0.8
70 0.0
225 1.3
164 0.0
131 1.3
40 0.0
98 0.0
79 0.0
463 6.3
1005 6.2
614 6.0
48 0.0
26 0.0
500 3.4
58 1.0
363 0.3
51 0.2
3 0.0
1061 4.7
340 5.6
91 0.0
392 5.5
5 0.0
187 0.0
113 0.0
124 1.5
19 0.0
303 1.2
1071 8.7
177 0.0
463 5.8
111 0.0
103 0.0
74 0.3
33 0.0
236 0.8
184 0.4
1111 0.5
22 0.2
37 0.0
44 00
173 0.1
171 3.6
XL 02
HC CO
759 3.5
209 18
32 0.0
52 0.0
67 0.0
43 0.0
147 0.8
156 2.6
149 1.5
170 0.0
19 00
352 7.1
114 0.0
97 0.0
413 5.6
510 4.0
462 4.9
102 1.1
17 0.0
91 0.0
43 0.8
1094 7.9
19 0.0
5 0.0
1309 5.4
479 6.0
140 1.2
341 3.5
8 0.0
59 0.0
53 0.0
64 0.1
17 0.0
110 0.0
908 8.7
86 0.0
458 6.1
74 0.0
62 0.0
72 0.0
15 0.0
236 2.4
240 0.5
6 0.1
19 0.1
20 0.0
204 1.1
95 0.1
197 5.4
XL 05
HC CO
557 34
155 0.7
40 0.0
72 0.1
71 0.0
40 0.0
588 0.3
63 0.0
144 1.2
176 0.0
27 0.1
28 0.0
87 0.0
26 0.0
452 6.2
409 0.1
926 6.6
29 0.0
14 0.0
147 0.1
SO 1.1
974 7.8
34 0.2
0 0.0
965 5.1
436 5.5
51 0.0
164 0.8
31 0.0
62 0.0
147 0.0
64 0.0
21 0.0
122 0.0
826 8.7
84 0.0
450 5.7
61 0.0
49 0.0
80 0.0
11 0.0
254 2.3
200 0.4
754 0.3
14 0.1
19 0.0
151 0.6
84 0.2
219 6.5
XI 02
HC CO
869 2.8
57 00
123 1.4
236 18
278 2.6
604 0.6
221 0.9
132 1.3
283 2.1
212 0.1
68 0.6
262 4.8
228 3.0
356 2.0
474 6.1
539 4.2
601 5.6
121 1.0
37 0.1
771 8.9
53 0.8
1068 7.9
70 0.7
4 0.0
1651 6.7
643 5.4
269 1.8
646 7.1
27 0.0
336 0.5
270 2.7
193 2.3
101 0.0
448 0.0
1128 B.7
634 7.8
403 4.3
160 0.0
171 0.2
65 0.4
278 1.4
248 1.1
204 0.5
42 0.4
9 0.0
66 0.1
172 0.3
115 0.0
184 4.7
XI 05
HC CO
767 26
118 06
93 08
149 0.7
108 00
75 0.0
133 0.7
46 0.0
139 1.0
185 0.0
25 0.0
40 0.0
107 1.9
158 0.0
419 5.2
500 1.7
1091 6.9
51 0.0
80 0.0
469 4.0
54 0.9
1004 7.8
69 04
0 0.0
832 4.6
629 5.1
91 0.0
306 06
18 0.0
79 0.1
101 0.0
117 1.2
136 0.1
267 0.7
955 7.9
330 0.0
398 5.0
66 0.0
126 0.1
99 0.2
273 1.1
218 0.6
281 0.5
48 0.2
36 0.1
64 0.1
132 0.2
106 0.0
234 6.9
XI 07
HC CO
826 32
86 06
44 0.1
474 8.0
325 4.5
37 0.0
106 0.7
101 0.4
132 1.1
140 0.0
19 0.0
433 7.4
43 0.0
30 0.0
404 5.8
361 0.0
485 5.4
109 1.3
45 0.0
124 0.1
40 0.8
1274 7.9
42 0.3
59 1.7
1731 0.3
610 5.3
60 0.0
454 5.0
8 0.0
78 0.0
166 0.0
81 0.6
103 0.0
166 0.0
784 7.5
263 00
407 5.4
62 0.0
52 0.0
62 0.0
638 5.3
236 1.6
217 0.5
29 0.0
59 0.2
20 0.0
225 1.8
141 0.1
235 7.9
XI 11
HC CO
657 33
77 03
41 0.0
94 01
103 00
87 0.0
123 0.6
49 00
132 1.1
152 00
18 0.0
54 01
36 0.6
30 00
415 5.7
329 0.1
962 7.1
49 0.0
24 00
239 1.0
44 08
851 7.9
46 0.3
3 0.0
1731 05
612 5.3
71 0.0
90 0.0
10 00
90 0.0
92 0.0
102 1.2
125 0.0
172 0.0
754 7.6
264 0.0
425 5.4
67 0.0
46 0.0
70 0.0
255 1.7
237 1.6
208 0.5
11 0.1
25 0.1
19 0.0
201 1.3
59 0.1
215 7.0
RS 02
HC CO
579 35
199 1.1
32 0.0
37 0.0
85 0.0
21 00
200 0.6
125 0.4
137 1.2
104 0.0
30 0.1
344 71
90 0.0
78 0.0
»N/A ON/A
463 4.0
418 4.3
96 1.1
14 00
607 6.4
44 0.8
1114 7.8
22 0.0
5 0.0
»N/A ffN/A
1205 55
78 0.2
440 4.0
28 00
50 00
65 0.0
62 00
95 0.0
98 0.0
895 8.7
241 0.0
463 5.8
61 0.0
59 0.0
39 0.0
184 0.2
242 3.2
196 0.3
4 0.1
18 0.1
21 0.0
101 0.2
28 0.1
297 8.7
RS 05
HC CO
671 3.1
154 0.8
39 00
21 0.0
86 0.0
46 0.0
236 03
102 0.0
130 0.9
107 0.0
15 0.0
385 72
64 0.0
25 0.0
#N/A ffN/A
387 0.7
303 3.8
90 1.0
25 0.0
458 6.7
49 0.9
1144 7.9
33 0.1
0 0.0
»N/A ffN/A
1192 4.9
59 0.0
267 3.2
20 0.0
49 0.0
73 0.0
62 0.1
97 0.0
134 0.0
869 8.7
231 0.0
424 5.5
67 0.0
50 0.0
40 0.0
12 0.0
472 35
287 0.8
43 0.1
13 0.1
247 3.4
243 1.3
57 0.2
324 9.6
115

-------
Veh
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
CS 03
HC 00
190 0.9
310 01
452 0.7
662 1.0
1913 2.9
205 1.8
236 1.1
194 3.5
94 1.7
201 1.9
345 0.2
307 0.1
166 0.4
614 0.6
162 0.3
1885 4.6
307 0.1
1116 0.1
103 0.2
331 0.3
450 5.2
236 8.1
ffN/A ffN/A
266 0.2
260 4.0
470 3.6
ffN/A ffN/A
477 0.8
245 0.2
142 0.2
279 2.6
327 0.1
206 0.2
426 0.3
248 3.4
259 0.5
503 1.5
183 0.4
311 7.2
269 0.9
238 0.5
378 0.9
636 0.1
557 9.1
305 0.7
209 0.5
841 0.1
1576 0.1
88 0.2
CS 05
HC CO
54 0.1
189 0.2
371 1.4
958 1.0
830 2.6
79 0.0
278 0.6
144 0.6
403 3.9
125 0.2
104 0.1
158 0.4
128 0.5
154 0.4
219 0.8
499 0.9
312 0.0
1101 0.1
1553 0.1
313 0.5
414 7.3
88 0.1
ffN/A ffN/A
282 0.8
260 4.1
127 0.1
ffN/A ffN/A
1037 0.1
58 0.0
67 0.1
266 3.2
78 0.1
423 0.6
506 2.8
252 2.6
704 0.8
575 0.5
125 0.5
171 1.2
413 0.4
207 0.6
94 0.3
158 0.1
412 3.8
212 0.4
462 0.6
1897 3.0
957 0.1
103 0.1
CS 07
HC CO
200 03
194 03
72 02
920 1.3
332 2.7
46 0.2
40 0.1
226 5.2
666 0.4
63 0.0
18 0.0
20 0.0
311 2.4
1870 2.1
153 0.1
353 5.6
132 0.1
235 0.4
384 0.5
107 0.1
703 10.4
19 0.0
64 0.1
176 0.4
137 0.9
1682 7.5
ffN/A ffN/A
350 0.6
162 2.4
24 0.0
186 2.7
18 0.0
331 0.3
362 1.8
303 2.4
243 0.4
268 0.1
14 0.0
69 0.4
176 0.5
36 0.1
6 0.0
686 8.3
498 4.8
131 0.4
142 0.2
368 1.3
73 0.0
53 0.0
CS 10
HC CO
167 0.4
158 0.1
83 08
879 13
106 0.2
38 0.1
98 0.2
315 8.3
699 0.3
99 0.0
21 0.0
20 0.0
309 2.3
1885 1.9
119 0.1
336 5.7
58 0.2
230 0.5
283 0.5
107 0.1
996 10.4
17 0.0
65 0.1
154 0.4
126 0.5
1927 6.7
ffN/A ffN/A
293 0.3
188 1.9
27 0.0
186 2.7
18 00
308 03
324 1.6
354 2.9
151 0.2
200 0.0
12 0.0
150 0.6
156 0.5
26 0.1
38 0.2
746 8.5
555 S.6
135 0.8
135 0.2
280 1.0
67 0.0
39 0.0
CS 12
HC CO
25 0.0
182 01
842 6.5
1965 0.4
421 1.2
195 2.5
18 0.0
315 5.3
1080 0.2
213 1.2
33 0.0
14 0.0
112 0.1
155 0.8
223 0.5
352 4.8
48 0.0
219 0.1
447 0.3
53 0.0
721 8.6
24 0.0
184 1.5
182 0.2
429 80
1927 7.9
ffN/A ffN/A
928 0.3
208 0.1
30 0.0
210 3.3
32 0.0
493 0.6
233 0.9
318 5.6
109 0.7
125 0.0
81 0.1
250 2.7
163 0.1
29 0.1
12 0.0
205 2.3
720 10.0
399 5.5
26 0.0
749 1.3
97 0.0
20 0.0
CS 15
HC CO
26 0.0
193 02
58 00
1097 1.3
313 3.9
92 0.2
25 00
597 10.2
1045 02
228 1.5
24 0.0
17 0.0
129 0.4
1755 2.1
204 0.4
371 5.2
39 0.1
199 0.4
317 0.3
68 0.1
633 8.6
21 0.0
174 1.1
164 0.3
249 3.5
1662 7.5
»N/A ffN/A
348 0.2
258 0.8
29 0.1
184 2.9
31 0.0
421 0.5
218 0.4
239 4.4
238 0.1
219 0.0
21 0.0
162 0.6
246 0.4
34 0.1
37 0.1
265 4.6
613 6.8
165 0.4
52 0.1
354 1.4
81 0.0
22 0.0
XL 02
HC CO
192 0.1
76 0.0
30 0.1
1141 1.1
167 0.2
87 0.4
103 0.1
309 7.4
1483 0.2
80 0.0
18 0.0
1 1 0.0
214 2.8
1885 2.7
110 0.3
550 9.3
15 0.0
186 0.4
235 0.4
39 0.1
1079 10.4
18 0.0
375 6.6
130 0.3
458 10.2
1927 7.8
30 0.0
512 0.6
221 0.9
29 0.0
244 6.0
18 0.0
315 0.3
586 2.7
289 6.4
265 0.5
100 0.1
3 0.0
108 1.3
139 0.1
10 0.0
66 0.2
259 6.0
527 10.0
190 0.9
43 0.1
363 1.7
60 0.0
92 0.2
XL 05
HC CO
338 0.3
87 00
45 0.1
1125 1.5
261 1.4
57 0.2
29 0.1
643 10.2
870 0.2
102 0.0
1,8 0.0
9 0.0
556 7.2
1885 2.1
121 0.3
507 9.1
19 0.0
173 0.3
196 0.5
41 0.1
774 10.4
20 0.0
379 6.5
155 0.2
405 9.6
1927 8.6
34 0.0
179 0.2
181 1.1
26 0.1
230 5.6
22 0.0
287 0.3
214 0.4
274 5.7
205 0.3
123 0.1
39 0.1
153 0.6
192 0.2
13 0.0
10 0.1
260 6.1
435 9.1
133 0.6
36 0.1
236 1.5
S3 0.0
33 0.2
XI 02
HC CO
24 00
287 04
56 04
1965 06
322 2.1
101 02
21 0.0
377 69
1332 02
258 1.4
26 00
28 00
68 03
184 0.5
216 05
303 4.0
25 00
277 0.2
641 02
100 0.1
510 7.1
30 0.0
60 1.2
172 03
320 88
1927 7.9
50 0.0
541 02
239 03
79 0.0
233 5.4
82 0.0
453 0.6
393 1.2
363 6.3
196 0.6
145 0.0
434 5.3
277 4.7
102 0.1
26 0.0
188 0.5
215 4.4
779 10.0
678 8.1
25 0.0
532 0.8
274 0.0
IS 00
XI 05
HC CO
82 0.1
291 05
32 0.1
1123 1.2
412 2.3
67 0.1
81 0.1
611 10.2
1183 0.2
265 1.4
23 0.0
37 02
131 1.5
1885 2.2
211 0.5
498 6.9
24 0.0
393 0.3
279 0.5
152 0.2
453 6.7
27 0.1
311 7.4
188 0.3
287 7.7
1927 6.6
81 0.6
816 0.2
205 0.6
36 0.1
221 5.1
35 0.0
374 0.6
201 0.2
320 5.9
262 0.5
135 0.0
168 01
270 1.3
235 0.1
28 0.1
40 0.1
198 4.7
621 8.0
317 1.2
66 0.0
335 1.1
188 0.0
20 0.0
XI 07
HC CO
142 0.4
277 0.2
50 03
804 12
103 0.7
132 1.9
117 0.2
764 10.2
821 0.3
108 0.0
20 0.0
16 0.0
546 7.6
1885 1.9
163 0.3
1885 10.1
191 0.3
241 0.3
235 0.4
114 0.1
706 10.4
23 0.0
330 5.2
137 0.6
328 8.8
1927 8.4
37 0.0
274 0.7
186 1.6
37 0.2
233 5.5
39 0.1
343 0.5
193 0.9
510 5.3
152 0.4
248 0.1
95 0.2
180 1.0
218 0.6
15 0.0
IS 0.1
214 S.6
525 8.4
161 0.7
158 0.2
179 1.4
65 0.0
58 0.3
XI 1 1
HC CO
208 0.4
230 02
45 02
547 1.2
57 0.0
132 16
47 01
689 101
715 0.2
81 0.0
20 0.0
18 0.2
618 102
1306 14
154 0.3
1885 10.1
21 0.1
198 03
240 0.5
87 0.1
656 10.4
22 0.1
342 55
113 0.4
326 8.8
1927 8.4
34 0.0
135 0.4
179 1.6
37 0.1
250 6.0
19 0.0
335 0.4
187 0.9
553 5.5
128 1.0
193 0.1
63 0.1
148 0.4
153 0.2
13 0.0
6 0.1
214 5.5
573 7.7
157 0.5
120 0.2
220 1.5
56 0.0
18 0.0
RS 02
HC CO
420 0.2
111 01
32 0.0
1149 0.9
304 1.9
66 06
1 1 0.0
388 8.5
661 03
74 0.0
20 0.0
13 00
535 6.4
1885 29
113 0.3
802 92
27 0.0
139 0.1
279 0.4
42 0.1
907 10.4
77 0.4
ffN/A #N/A
130 0.4
400 9.9
1927 8.6
29 00
454 0.3
196 1.5
22 0.0
276 7.4
17 0.0
260 0.3
485 2.4
255 5.6
176 0.4
170 0.1
13 0.1
151 2.3
109 0.1
5 0.0
44 0.2
255 6.3
594 10.0
252 1.6
43 0.2
ffN/A ffN/A
94 0.0
19 0.0
RS 05
HC CO
38 0.0
97 0.1
46 0.1
999 1.5
332 4.7
39 0.1
89 0.6
699 10.1
579 0.3
84 0.0
24 0.0
25 0.2
179 1.8
1259 1.7
117 0.3
566 8.4
17 0.0
101 0.1
268 0.5
39 0.1
558 10.4
165 39
ffN/A »N/A
165 0.4
358 9.5
1037 0.5
64 0.3
149 0.3
36 0.0
22 0.0
521 7.6
22 0.0
283 0.3
467 2.5
261 4.5
261 0.7
90 0.0
73 0.1
92 0.6
183 0.2
4 0.0
141 0.2
250 62
507 10.0
135 0.5
32 0.1
ffN/A ffN/A
96 0.1
19 0.0
1 16

-------
Veh
401
402
403
404
405
406
407
408
409
410
502
503
504
505
506
507
506
509
510
511
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
701
702
703
704
705
706
707
708
709
CS 03
HC CO
330 5.0
493 1.3
329 0.2
277 2.5
185 0.2
277 0.2
225 1.1
306 2.9
500 5.0
297 3.3
487 5.5
600 10.6
217 2.9
357 0.5
600 11.2
309 5.9
369 1.7
412 5.4
1080 0.1
322 4.7
570 4.2
739 4.1
552 2.6
429 0.5
425 4.4
315 2.9
103 0.3
324 1.4
267 3.4
337 3.2
412 3.3
269 0.6
203 2.5
505 7.9
330 3.4
254 1.0
254 2.1
804 10.0
321 3.6
95 0.3
178 0.7
117 0.9
380 9.4
176 2.2
646 0.2
352 3.9
302 5.1
246 0.9
138 1.9
CS 05
HC CO
408 3.0
363 5.0
277 0.1
213 0.3
500 0.1
500 0.1
216 0.2
237 2.0
394 5.0
232 2.1
192 0.3
600 7.0
94 2.9
199 0.1
600 8.0
158 2.4
256 3.9
347 4.6
126 0.0
276 3.9
414 3.1
548 2.2
202 0.0
211 0.2
192 2.6
136 1.3
82 0.3
178 0.3
105 0.8
177 0.5
254 2.7
246 6.3
97 1.5
384 6.8
155 2.5
113 0.1
201 3.4
628 10.0
84 0.5
36 0.1
128 0.4
100 0.6
182 3.2
103 0.6
371 0.3
123 1.6
59 0.3
152 0.4
152 0.6
CS 07
HC CO
136 1.0
70 0.1
344 0.1
180 0.1
500 0.1
500 0.1
215 0.2
63 0.0
323 3.1
73 0.0
444 1.4
600 7.7
52 2.0
53 0.0
600 9.6
9 0.1
215 1.1
282 3.1
166 0.5
10 0.0
423 4.1
1013 8.2
27 0.0
118 0.0
199 2.7
144 2.0
52 0.2
551 0.5
30 0.1
10 0.0
9 0.0
366 6.5
183 4.5
406 5.1
188 2.3
114 0.2
253 4.2
634 7.5
25 0.0
217 6.3
16 0.0
56 0.0
152 0.8
18 0.0
180 0.1
140 1.3
27 0.0
3 0.0
30 0.0
CS 10
HC CO
141 1.0
65 0.1
322 0.
203 0.
297 0.2
500 0.
212 0.
62 0.
334 3.0
93 0.1
522 2.5
600 7.8
56 2.5
48 0.0
600 9.0
5 0.0
211 1.5
269 3.1
91 1.1
13 0.0
453 4.1
1013 8.2
42 0.0
90 0.0
199 2.9
147 2.2
56 0.2
546 0.4
25 0.1
7 0.0
7 0.0
373 6.4
144 4.3
501 4.3
199 2.4
158 0.2
239 4.1
596 7.2
26 0.0
229 6.2
19 0.0
S3 0.0
147 0.8
9 0.0
279 0.1
143 1.2
34 0.0
0 0.0
17 0.0
CS 12
HC CO
153 0.2
356 2.1
500 0.2
350 0.3
292 0.2
500 01
269 0.2
263 0.4
250 0.4
130 0.1
334 O.S
581 70
38 1.5
267 0.4
600 9.4
38 0.0
234 2.2
243 1.5
313 1.2
73 0.0
459 2.8
669 1.8
563 0.0
577 0.2
195 2.7
172 1.9
132 0.2
733 0.2
31 0.1
22 0.0
43 0.0
384 5.1
151 4.4
418 3.2
190 1.6
469 0.4
280 4.3
629 6.1
73 0.0
265 6.1
39 0.0
146 0.7
130 0.5
16 0.0
333 0.0
152 0.9
102 0.0
5 0.0
38 0.1
CS 15
HC CO
135 0.8
100 0.1
314 0.1
271 0.1
500 0.2
500 0.1
231 0.2
104 01
226 0.5
153 0.1
307 04
536 6.2
48 15
276 0.5
600 8.3
48 0.0
224 2.1
265 2.7
301 3.0
20 0.1
432 3.0
665 2.2
378 0.1
472 0.1
203 2.7
166 2.1
133 0.3
575 0.6
41 01
24 0.0
60 0.0
399 6.0
155 3.8
855 8.8
206 1.9
234 0.0
262 3.0
603 6.2
108 0.1
277 5.7
96 0.0
131 0.3
137 0.8
23 0.0
352 0.0
145 1.1
98 0.0
2 0.0
43 0.1
XL 02
HC CO
79 0.1
86 01
181 0.1
149 0.1
111 02
500 0.1
211 02
48 01
341 3.1
49 01
145 02
600 8.2
38 2.4
210 03
600 8.9
0 0.0
45 0.1
0 0.0
8 0.1
0 0.0
372 2.9
1013 8.0
968 2.2
41 0.0
168 3.4
110 1.9
25 0.1
392 0.4
19 0.1
0 0.0
0 0.0
314 5.8
61 37
663 8.5
164 2.3
45 0.0
232 4.4
566 6.4
10 0.0
243 5.8
6 0.0
35 0.0
112 0.5
2 0.0
148 0.1
122 0.9
0 0.0
0 0.0
0 0.0
XL 05
HC CO
44 0.1
371 4.3
160 01
172 0.1
123 0.2
500 0.1
348 0.2
48 0.1
291 18
65 0.1
250 0.7
593 76
35 19
114 0.1
600 8.0
0 0.0
28 0.1
2 0.1
42 0.2
0 0.0
276 2.9
1013 8.0
904 1.7
65 0.0
167 3.3
130 2.5
35 0.2
433 0.4
19 0.1
0 0.0
0 0.0
332 6.0
60 3.8
561 5.8
181 2.3
50 0.1
226 4.3
547 5.6
17 0.0
250 6.0
7 0.0
45 0.0
128 0.8
0 0.0
141 0.1
109 1.1
1 0.0
0 0.0
0 0.0
XI 02
HC CO
62 01
307 08
463 02
448 01
140 02
500 0.1
254 03
239 04
276 13
228 04
520 12
600 81
43 24
240 03
600 9.2
62 0.1
57 0.0
40 00
769 06
74 00
492 3.7
598 15
1015 1.1
585 0.1
165 2.3
156 2.4
141 0.7
511 0.2
43 01
249 08
21 00
340 55
183 6.3
333 25
185 26
548 1.0
285 4.9
541 5.2
108 0.2
257 6.1
120 0.1
142 O.S
27 0.0
50 0.1
214 01
127 06
30 0.0
6 0.0
9 0.0
XI 05
HC CO
92 0.1
299 08
302 01
307 0.0
186 0.2
500 0.1
254 0.2
289 0.9
264 1.2
206 0.1
371 1.2
589 75
61 1.7
261 04
600 8.6
59 01
104 01
45 0.2
438 5.8
26 0.1
546 3.2
609 1.8
1015 1.1
551 0.1
235 2.0
ISO 2.3
113 0.3
524 0.6
67 0.2
224 03
72 0.0
497 6.0
185 6.7
497 8.4
346 2.4
394 0.2
345 4.5
SS9 5.3
143 0.1
286 5.9
152 0.1
138 0.4
92 O.S
85 0.2
197 0.0
121 0.8
33 0.0
3 0.0
13 0.0
XI 07
HC CO
56 0.1
87 0.1
240 0.1
201 0.0
163 0.2
500 0.1
251 0.2
108 01
274 1.4
121 0.1
413 1.3
534 7.1
83 3.0
261 0.5
600 92
25 0.1
67 0.2
41 0.2
211 04
0 00
402 3.2
713 47
1013 14
287 00
165 2.4
142 2.4
46 0.2
512 0.4
21 0.0
18 0.0
8 0.0
360 6.1
140 5.6
419 35
219 2.7
140 0.1
243 4.4
516 5.5
36 0.0
26 0.0
39 0.0
67 0.0
112 0.8
11 0.0
214 0.1
119 1.1
3 0.0
3 0.0
1 0.0
XI 1 1
HC CO
50 0.1
77 00
251 0.1
189 0.0
234 0.2
500 01
339 02
69 01
289 18
107 0.1
409 15
585 75
77 26
219 0.4
600 88
13 01
81 0.1
101 03
144 03
0 00
399 3.1
804 57
963 1.6
242 00
163 2.2
146 22
40 02
525 0.4
19 0.1
17 0.0
7 0.0
352 58
166 68
429 32
232 2.6
218 0.2
247 4.3
520 5.3
31 0.1
14 0.0
21 0.0
68 0.0
112 07
8 0.0
184 0.1
121 1.0
3 00
0 0.0
2 0.0
RS 02
HC CO
77 0.1
79 0.1
174 0.1
146 0.1
158 0.2
500 01
482 0.3
49 0.1
391 45
89 0.1
205 03
582 78
30 18
67 03
600 84
0 00
30 00
13 02
77 08
0 0.0
378 3.1
1013 8.4
901 1.8
43 00
160 28
136 23
26 02
381 04
32 02
135 0.1
0 0.0
340 60
102 58
379 34
181 2.8
80 02
252 52
489 5.9
a o.o
1 0.0
4 00
36 0.0
76 01
2 0.0
160 0.1
98 0.6
0 0.0
0 0.0
0 0.0
RS 05
HC CO
47 0.1
320 2.9
198 0.1
186 0.1
240 0.2
500 0.1
443 0.2
53 0.1
369 4.3
88 0.1
304 08
389 5.7
28 1.6
114 0.4
554 78
0 0.0
31 00
1 1 0.2
66 0.2
0 00
312 3.1
1013 84
832 12
95 0.0
171 3.3
123 2.2
44 03
406 04
30 0.2
119 0.2
0 0.0
340 6.0
110 6.6
426 3.4
211 3.3
93 0.2
205 4.9
454 5.6
15 0.0
6 0.0
5 0.0
41 0.0
118 09
2 00
151 0.1
105 0.8
0 0.0
0 0.0
1 00
117

-------
               APPENDIX D
AET Errors of Commission in the CTP Sample
V«h
31
223
249
14
20
24
206
221
250
260
701
321
336
360
11
17
21
37
39
SO
224
251
306
309
322
330
347
348
357
509
511
704
708
MIT
82
82
82
85
85
84
86
84
83
84
86
82
82
82
85
86
84
85
85
84
86
86
86
84
86
84
84
86
86
85
86
86
84
MFR
AMC
FORD
FORD
TOYT
MAZD
TOYT
POFD
FORD
FORD
FORD
TOYT
GM
am
Out
VW
SUBA
TOYT
SUBA
MAZD
VW
FORD
FORD
am
am
am
am
am
am
am
MITS
MITS
TOYT
TOYT
CID
258
302
351
89
91
95
300
140
122
122
91
250
151
151
109
109
122
109
80
117
302
183
151
151
121
151
151
151
151
97
96
122
122
Typ«
LOT
LDV
LOT
LDV
LDV
LDV
LOT
LOT
LOT
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LDV
LDV
LOT
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
KlliU
53
46
45
23
13
27
16
44
70
35
5
60
53
56
27
17
59
29
18
81
28
14
5
61
18
60
25
17
37
26
16
32
32
Quota Grp
Garb 81-82
Carb 81-82
Carb 81-82
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Carb 83-86
Ft 81-82
Fl 81-82
F\ 81-82
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
Fl 83-86
HCCorl
1.70
0.41
1.70
0.41
0.41
0.41
0.80
0.80
1.70
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.80
0.41
0.41
0.80
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
0.41
FTPMC
1.13
0.33
0.88
0.23
0.10
0.23
0.62
0.54
1.44
0.73
0.24
0.33
0.28
0.29
0.31
0.14
0.58
0.20
0.26
0.71
0.39
0.16
0.21
0.25
0.16
0.24
0.21
0.40
0.35
0.24
0.20
0.26
0.16
COCert
18.0
7.0
18.0
3.4
3.4
3.4
10.0
10.0
18.0
10.0
3.4
7.0
7.0
7.0
3.4
3.4
10.0
3.4
3.4
10.0
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
FTPCO
14.4
1.8
5.1
1.4
3.1
1.7
0.9
4.1
11.8
8.0
1.9
5.2
2.7
4.2
2.9
2.1
8.4
2.6
1.6
7.9
2.0
2.2
1.8
3.3
1.8
2.9
2.6
1.2
3.3
1.4
3.0
1.7
1.9
AETHC
578
161
1126
176
63
231
250
124
214
146
262
264
517
1062
82
168
337
465
1743
225
449
126
225
500
239
325
331
223
548
307
127
335
344
AET CO
0.4
1.3
0.0
1.6
2.4
0.5
0.0
1.4
3.5
2.6
0.1
0.0
0.2
1.0
1.6
1.4
2.3
0.4
0.1
4.4
0.1
1.3
0.4
0.5
0.9
0.3
0.7
0.1
0.7
6.7
2.1
0.4
0.2
XL30HC
20
67
53
8
1
22
67
32
64
72
6
18
512
92
3
1
226
105
3
3
43
17
19
95
11
39
265
100
43
0
0
2
0
XL30CO
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.0
0.0
0.0
0.6
0.2
0.0
0.0
0.7
0.3
0.0
0.0
0.0
0.0
0.1
0.1
0.0
0.1
0.5
0.1
0.1
0.0
0.0
0.0
0.0

-------
    APPENDIX  E:
PER-REP AIR EMISSION REDUCTIONS FOR  ALL
SYSTEMS:   BY  QUOTA  GROUP
     These tables summarize the HC and CO emission  reductions
per repair, due to repairs to the systems on vehicles in the
quota groups listed, in g/mi, as measured by the LA4
transient cycle.  Figures indicate averages derived from
isolatable repairs.
Carbureted 1981-1982
SYSTEM
REPAIRED

Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
SIMPLE AVERAGES
ISOLATABLE
N
11
46
18
11
55
3
30
3
1
42
220
A HC
0.41
0.67
0.24
0.18
0.31
-3.31
1.12
0.39
1.23
1.06
0.59
REPAIRS
A CO
-0.7
8.8
0.0
2.6
6.8
-11.0
7.7
3.9
30.9
22.3
9.0
Vehicles
MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N
18
53
24
18
70
8
35
3
6
56
291
A HC
0.72
0.60
0.11
-0.24
0.18
-1.65
1.00
0.39
-0.05
0.90
-
t- ratio
1.6
2.6
0.3
-0.6
0.8
-2.8
3.5
0.4
-0.1
4.1
-
A CO
0.6
9.1
-0.3
-5.4
4.2
-4.7
9.1
3.9
2.4
19.4
-
t-ratlo
0.1
3.4
-0.1
-1.1
1.5
-0.7
2.7
0.4
0.3
7.6
-
Carbureted 1983-1986
SYSTEM
REPAIRED

Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
SIMPLE AVERAGES
ISOLATABLE
N
3
20
12
4
8
1
10
0
2
25
85
A HC
0.02
0.87
-0.20
0.08
0.16
-0.11
1.32
-
-0.01
0.42
0.47
REPAIRS
A CO
2.4
11.1
-6.5
0.3
4.1
-4.7
6.9
-
0.3
13.2
6.8
Vehicles
MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N
5
29
17
5
9
2
13
3
4
26
113
A HC
0.45
0.92
-0.33
-0.06
0.16
-0.24
1.53
1.44
-0.12
0.42
-
t-ratlo
1.0
4.4
-1.1
-0.1
0.5
-0.3
5.1
2.3
-0.2
2.0
-
A CO
12.8
13.0
-7.3
-1.0
3.9
-7.6
8.0
24.8
-3.9
13.5
-
t-ratlo
1.4
3.1
-1.2
-0.1
0.6
-0.5
1.3
2.0
-0.4
3.2
-
                              119

-------
Fuel
SYSTEM
REPAIRED

Induction
Fuel Meter
ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
SIMPLE
Injected 1981-1982 Vehicles
AVERAGES
ISOLATABLE
N A
0
6 1
7 0
2 -0
4 4
0
5 2
0
2 -0
14 1
40 1
HC
-
.10
.06
.06
.85
-
.11
-
.02
.06
.29
REPAIRS
A CO
-
22.8
0.1
0.0
28.3
-
6.2
-
0.4
32.0
18.3

MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N
1
6
8
2
4
0
5
0
4
15
45
A HC
0.02
1.11
0.06
-0.06
8.27
-
2.28
-
1.04
1.21
-
t- ratio
0.0
1.1
0.1
0.0
4.9
-
1.9
-
0.8
2.0
-
A CO
-0.2
22.8
0.1
0.0
38.0
-
8.6
-
27.9
35.9
-
t-ratlo
0.0
1.7
0.0
0.0
1.6
—
0.5
-
1.4
4.1
-

SYSTEM
REPAIRED

Induction
Fuel Meter
Ignition
EGR
Air Injection
PCV
Exhaust
Evap
Engine
3-Way
ALL
Fuel
SIMPLE
Injected 1983-1
AVERAGES
ISOLATABLE
N
0
22
26
2
10
0
15
1
5
75
156
A

2
0
0
-0

0
0
1
0
0
HC
-
.07
.37
.00
.65
-
.98
.08
.57
.22
.56
REPAIRS
A CO
-
21.6
-0.3
-0.1
-0.3
-
7.2
0.3
7.1
14.2
10.7
986
Vehicles
MULTIPLE
LINEAR
REGRESSION
ALL REPAIRS
N
4
29
29
2
10
1
15
2
7
82
181
A
-1
1
0
0
-0
1
0
0
1
0

HC
.77
.78
.72
.00
.82
.77
.94
.08
.53
.33
-
t- ratio
-1.2
3.6
1.4
0.0
-0.9
0.5
1.4
0.0
1.5
1.1
-
A CO
-14.0
17.2
1.9
-0.1
-1.0
13.3
6.8
0.3
6.4
14.1
-
t-ratlo
-0
3
0
0
-0
0
0
0
0
4

.8
.2
.3
.0
.1
.3
.9
.0
.6
.3
—
120

-------
                  HC Benefit: per System  Repair —
               Carbureted MY 81-82  Vehicles
                                    Cart) 81 -82
     6.00 T
     4.50 4-
 HC
g/rrt
           INDT  FUEL  IGNT  EGR   AIR
                                       EXH  EVAP  ENG 3WAY ALL
     -4.50 J-
                  HC  Benefit  peg System Repair  —•"•
                Carbureted MY 83-flfi Vehicles
                                    Garb 83-86
 HC
g/mi
 6.00 y

 4.50 - •

 3.00 • •

 1.50-•

 0.00

-1.50

-3.00 J-

-4.50 1
           INDT  FUEL  IGNT  EGR   AIR  PCV   EXH EVAP ENG  SWAY  ALL
                             121

-------
         Averaye  HC Benefit per System Repair —
             Fuel  Injected MY  81-82 Vehicles
HC
                                    FI81-82
 6.00 y

 4.50 • •

 3.00 • •

 1.50 ••

 0.00

-1.50 ••

-3.00 • •

-4.50 - •
          INDT  FUEL  IGNT  EGR   AIR   PCV  EXH  EVAP  ENG  SWAY  ALL
         Average HC  Benefit per  System Repair —
             Fuel  In-iected MY  83-86 Vehicles
    6.00 -P

    4.50 -•

    3.00 • •
     1.50 ••
    -1.50 ••
    -3.00 • •
    -4.50 J-
          INDT  FUEL  IGNT  EGR
                                   Fl 83-86
                                 PCV  EXH  EVAP  ENG SWAY ALL
                             122

-------
          AveraQft CO
                                   System Repair  —
                Carhiir«»l-Ari  MV
                                  Cart 81-82
      30 T
      20 +
 CO
 g/mi
     -10-L
          INDT  FUEL IGMT  EGR  AIR
EXH  EVAP  ENG 3WAY  ALL
         Average CO Benefit per System  Repair
               Carbureted MY 83-86 Vehicles
                                  Garb 83-86
     40-r
     30 +
CO
g/mi
                         EGR   AIR  "PW  EXH  EVAP  ENG  SWAY  ALL
    -20 -L
                            123

-------
         Average  CO Benefit per System Repair —
             Fuel In-ieeted MY 81-82
                                  FI81-82
     40 T
CO
g/mi
      0

    -10-.


    -20 ••
     INDT  FUEL IGNT  EGR  AIR  PCV   EXH  EVAP  ENG  3WAY  ALL
    Averae CO Benefit
                                 Sstem Reair —
             Fuel  Injected MY 83-86 Vehicles
                                  Fl 83-86
CO
g/mi
40 j


30 ••


20 ••


10 ••


 0


-10 ••


-20--
         INDT  FUEL  IGNT  EGR  AIR  PCV   EXH EVAP ENG  3WAY  ALL
                            124

-------
       APPENDIX  F:
      STATISTICALLY
PBR-RBPAIR  EMISSION REDUCTIONS  FOR
SIGNIFICANT  SYSTEMS:  BY QUOTA GROUP
     This table  summarizes  the HC and CO emission reductions
per repair, due  to repairs  to  the systems on vehicles in the
quota groups listed,  in g/mi,  as  measured by the LA4
transient cycle.
SYSTEM
REPAIRED
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N A HC A CO
MULTIPLE LINEAR REGRESSION
ALL REPAIRS OF THESE SYSTEMS
N A HC t-ratlO
A CO
t-ratio
Carbureted 1981-1982 Vehicles
Fuel Meter
Exhaust
3-Way
46 0.67 8.8
30 1.12 7.7
42 1.06 22.3
53 0.61 2.7
35 1.05 3.6
56 0.91 4.2
9.2
9.6
19.2
3.5
2.9
7.7
Carbureted 1983-1986 Vehicles
Fuel Meter
Exhaust
3-Way
20 0.87 11.1
10 1.32 6.9
25 0.42 13.2
Fuel Injected 19
Fuel Meter
Exhaust
3-Way
6 1.10 22.8
5 2.11 6.2
14 1.06 32.0
Fuel Injected 19
Fuel Meter
Exhaust
3-Way
22 2.07 21.6
15 0.98 7.2
75 0.22 14.2
29 1.01 5.0
13 1.51 5.1
26 0.42 2.0
14,6
7.4
13.3
3.6
1.2
3.2
81-1982 Vehicles
6 1.11 0.9
5 2.28 1.5
15 1.28 1.7
22.8
8.6
37.8
1.7
0.5
4.4
83-1986 Vehicles
29 1.67 3.5
15 0.94 1.4
82 0.38 1.3
16.2
6.8
14.3
3.2
0.9
4.5
                              125

-------
    APPENDIX  6:
PBR-REPAIR EMISSION REDUCTIONS FOR ALL
       SUBSYSTEMS
     This table summarizes the HC and CO emission  reductions
per repair, due to repairs to the subsystems listed, in g/mi,
as measured by the LA4 transient cycle.
SUBSYSTEM
REPAIRED
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
INDUCTION SYSTEM
Htd Air Door
Temp Sensors
Air Filter
Hoses
Other (Indt)
1
0
3
4
1
A HC

0.06
-
-0.01
1.13
0.00
A CO

-1.6
-
2.2
-2.3
-0.2
FUEL METERING SYSTEM
Cart) Assembly
Fuel Meter Tune
Idl Spd Sole
Fuel Inj
Hoses
Other
Chk Adj Vacm
Vac Diaphrms
Other (Chk)
22
30
0
19
3
4
1
2
1
IGNITION SYSTEM
Dist Assembly
Igni Tune Items
Vac Adv Assmb
Spk Delay Dev
Elect Tim Mod
Hoses
Wir/Hms/Fuse
Other
EGR SYSTEM
Vatv Assembly
Delay Solnoid
Cool Temp Sen
Hoses
Other
3
38
2
1
0
2
0
0

8
0
1
4
2
1.03
0.61
-
2.35
-0.09
2.35
0.01
-0.06
-0.01

0.00
0.34
-0.19
0.28
-
0.01
-
-

-0.01
-
-0.02
0.05
0.62
11.9
11.6
-
24.2
0.2
29.7
0.5
-1.0
-0.4

-0.5
0.4
-1.2
10.4
-
-0.5
-
-

0.6
-
0.3
-1.4
11.3
AIR INJECTION SYSTEM
Pump Assembly
Byps/Dump Vlv
Diverter Vlv
2
0
5
7.93
-
-1.45
41.1
-
-3.4

N

2
2
15
6
3

27
43
1
20
7
11
1
5
2

11
54
5
1
1
2
2
2

12
1
1
8
5

5
3
13
MULTIPLE
ALL
A HC

-0.65
-0.40
-0.57
1.16
1.18

0.89
0.49
-0.65
2.22
0.08
0.70
0.01
0.10
1.51

-0.22
0.46
-0.15
0.28
0.56
0.01
-0.53
-0.48

-0.01
-0.89
-0.02
-0.05
0.41

5.31
-0.31
-0.67
LINEAR
REPAIRS
t- rat lo

-0.4
-0.2
-0.9
1.4
0.8

2.3
1.6
-0.3
5.2
0.1
1.1
0.0
0.1
1.0 -

-0.4
1.7
-0.1
0.1
0.3
0.0
-0.3
-0.3

0.0
-0.4
0.0
-0.1
0.4

4.7
-0.2
-1.0
REGRESSION
A CO

-6.0
3.3
-2.8
7.7
3.2

10.8
10.0
-14.1
22.2
-5.8
10.7
0.5
1.5
20.9

-3.6
0.5
-0.6
10.4
1.2
-0.6
-4.2
-5.8

-1.0
-16.3
0.3
-1.5
7.0

22.6
12.3
-1.3
t- ratio

-0.31
0.16
-0.38
0.77
0.19

2.41
2.78
-0.61
4.40
-0.66
1.47
0.02
0.14
1.13

-0.50
0.16
-0.05
0.46
0.05
-0.04
-0.23
-0.32

-0.15
-0.60
0.01
-0.18
0.60

1.70
0.71
-0.17
                              126

-------
SUBSYSTEM
REPAIRED
SIMPLE AVERAGES
ISOLATABLE REPAIRS
N
A HC
A CO
AIR INJECTION SYSTEM (continued)
Check Valve
Drive Belt
Hoses
Wir/Hms/Fuse
Other
PCV SYSTEM
Valv Assembly
Filters
Hoses/Lines
Other
14
1
9
1
4

3
0
1
0
EXHAUST SYSTEM
Exh Manifold
Catalyst
Other
0
43
1
0.35
-0.04
-0.24
0.02
-0.30

-0.12
-
-9.68
-

-
1.11
0.50
3.8
0.1
0.5
0.4
-3.7

-0.8
-
-35.3
-

-
7.0
12.3
EVAPORATIVE SYSTEM
Evap Canister
Canister Purg
Hoses
Other
1
2
1
0
ENGINE ASSEMBLY
Eng Assembly
Cooling Sys
Valve Adj
Belt Tension
Hoses
Eng Oil
Other
2
1
1
0
1
1
2
0.08
0.58
0.02
-

0.63
-0.06
0.11
-
5.99
-0.16
0.46
0.3
5.8
0.2
-

16.6
-1.5
1.5
-
28.7
-0.8
1.9
THREE-WAY CATALYST SYSTEM
ECU
O2 Sen
Load Sensor
Eng Spd Sen
Cool Temp Sen
EGR Postn Sen
A/F Cntri Act
Air Bypas Sen
Air Divrt Act
ISC Sys
Hoses
MAT Sen
Wir/Hms/Fuse
Other
ALL
14
69
11
1
4
1
5
1
1
1
6
3
6
1
372
0.40
0.80
0.61
0.53
0.86
0.06
2.46
2.76
0.12
0.20
0.37
-0.55
0.43
-0.02
0.70
10.7
20.7
23.2
3.4
27.4
-12.8
32.6
28.2
-0.4
2.0
-3.7
10.2
26.7
-0.7
10.7

N

31
5
27
2
7

3
4
3
1

10
56
2

1
2
3
2

3
7
4
1
1
2
3

19
82
22
2
8
1
5
3
4
1
18
3
9
2
630
MULTIPLE
ALL
A HC

0.33
-2.01
-0.06
1.39
-0.46

-0.12
0.54
-3.55
-0.34

0.42
1.09
0.77

0.08
0.58
0.56
1.04

0.66
-0.24
-0.27
3.72
5.99
-0.14
0.36

0.67
0.91
0.02
0.27
0.24
0.06
2.46
1.77
-0.33
0.20
-0.58
-0.55
0.48
0.45
-
LINEAR
REPAIRS
t- ratio

0.8
-1.8
-0.1
1.0
-0.6

-0.1
0.4
-2.9
-0.2

0.6
3.9
0.6

0.0
0.4
0.5
0.7

0.6
-0.3
-0.2
1.8
3.2
-0.1
0.3

1.5
4.2
0.0
0.2
0.4
0.0
2.9
1.3
-0.3
0.1
-1.1
-0.5
0.7
0.3
-
REGRESSION
A CO

4.3
-3.7
1.0
17.8
-6.4

-0.8
3.9
-10.8
-8.2

0.8
7.8
4.1

0.3
5.8
-0.6
19.8

14.0
-8.2
-7.4
105.1
28.7
2.9
0.2

14.0
21.9
17.1
-3.7
11.4
-12.8
32.6
21.8
-4.4
2.0
-11.9
10.1
26.2
20.5
-
t- ratio

0.88
-0.28
0.20
1.09
-0.67

-0.06
0.27
-0.73
-0.36

0.10
2.36
0.25

0.01
0.37
-0.05
1.06

1.03
-0.83
-0.57
4.28
1.28
0.16
0.02

2.59
8.60
2.98
-0.24
1.42
-0.57
3.25
1.36
-0.32
0.09
-1.86
0.78
3.27
1.25
-
127

-------
APPENDIX B:   PER-REP AIR  EMISSION  REDUCTIONS  FOR  STATISTCALLY
           SIGNIFICANT  SUBSYSTEMS:   BY QUOTA GROUP
     This table  summarizes  the  HC  and  CO  emission  reductions
per repair, due  to repairs  to the  subsystems on vehicles  in
the quota groups listed, in g/mi,  as measured by the LA4
transient cycle.  Figures indicate the averages derived from
isolatable repairs.
SUBSYSTEM
REPAIRED

SIMPLE AVERAGES
ISOLATABLE REPAIRS
N A HC A CO
MULTIPLE LINEAR
REGRESSION
ALL REPAIRS OF THESE SUBSYSTEMS
N A HC t-ratlo
A CO
t-ratlo
Carbureted 1981-1982 Vehicles
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
O2 Sensor
Load Sensor
ALL
15 1.34 16.2
12 0.37 7.6
0 -
19 0.85 6.2
9 1.42 21.9
15 0.32 9.6
1 2.13 76.7
149 0.55 8.3
18 1.06 2.8
19 0.61 1.7
0 -
28 1.07 3.6
11 1.17 2.5
21 0.38 1.1
1 2.13 1.3
291
13.1
11.5
-
9.6
17.9
12.8
76.7
-
3.0
2.7
-
2.7
3.2
3.2
4.1
-
Carbureted 1983-1986 Vehicles
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
02 Sensor
Load Sensor
ALL
7 0.36 2.6
10 0.63 10.5
0 -
10 1.32 6.9
1 -0.13 0.6
12 0.31 5.2
2 1.53 73.7
75 0.57 9.1
9 0.92 2.5
12 0.93 3.0
0 -
12 1.51 4.8
1 -0.13 -0.1
13 0.29 1.0
3 1.05 1.7
113
10.9
15.7
-
7.4
0.6
5.0
48.8
-
1.6
2.6
-
1.2
0.0
0.9
4.1
-
Fuel Injected 1981-1982 Vehicles
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
O2 Sensor
Load Sensor
ALL
0 -
5 1.35 27.3
0 -
3 2.55 12.6
0 -
9 1.77 42.5
2 0.16 0.8
35 1.27 19.5
0 -
5 1.35 1.0
0 -
3 2.55 1.5
0 -
10 2.03 2.2
2 0.16 0.1
45 -
-
27.4
-
12.6
-
50.1
0.8
-
-
1.9
-
0.7
-
5.0
0.0
-
Fuel Injected 1983-1986 Vehicles
Carburetor
FuelMtr Tune
Fuel Injector
Catalyst
ECU
02 Sensor
Load Sensor
ALL
0 -
3 0.30 5.0
19 2.35 24.2
11 0.87 6.3
4 -1.75 -12.0
33 0.93 25.5
6 0.19 5.0
113 0.80 12.1
0
7 -0.30 -0.3
20 2.26 4.2
13 0.94 1.4
7 -0.05 -0.1
38 1.24 3.2
16 -0.72 -1.2
181
-
-4.4
23.3
6.8
7.3
27.9
1.3
-
-
-0.5
4.3
1.0
0.8
7.1
0.2
-
                              128

-------
        Averaye HC Benefit per  Subsystem Repair  —
                Carbureted MY  81-82 Vehicles
                                    Cart) 81 -82
 HC
g/rri
3.00 j
2.50 - •
2.00 • •
1.50 ••
1.00 ••
0.50 ••
0.00
-0.50
-1.00
-1.50 4-
-2.00 • •
          .. Garb   FuelMtr  Fuellnj  Catalyst
                   Tune
                        ECU     O2    Load
                              Sensor Sensor
                       All
        Average  HC Benefit per Subsystem  Repair  --
                Carbureted MY  83-86  Vehicles
                                    Carb 83-86
 HC
g/rrt
 3.00 -r
 2.50 • •
 2.00 • •
 1.50 ••
 1.00 ••
 0.50 • •
 0.00
-0.50
-1.00
-1.50-.
-2.00 • •
          .. Cart)
Fuel Mtr  Fuel Inj  Catalyst
 Tune
ECU
  O2    Load
Sensor  Sensor
                               129

-------
         Average HC  Benefit n*»r Subsystem  Repair —
               Fuel Tn-iected  MY Sl-fl? Vehirles
                                      FI81-82
 HC
g/mi
 3.00 -r
 2.50 • •
 2.00 • •
 1.50- •
 1.00 ••
 0.50 • •
 0.00
-0.50
-1.00
-1.50 ..
-2.00 - •
          .. Garb   Fuel Mtr  Fuel Inj  Catalyst
                    Tune
ECU     O2     Load
      Sensor  Sensor
                 All
        Average  HC Benefit  per  Subsystem Repair  —
               Fuel  In-iected MY 83-fifi  Vehicles
                                      Fl 83-86
 HC
g/mi
3.00 y
2.50 • •
2.00 ••
1.50 ••
1.00 ••
0.50 • -
0.00
-0.50
-1.00
-1.50 ••
-2.00 • •
            Caib   Fuel Mtr  Fuel Inj  Catalyst
                    Tune
  O2    Load
Sensor  Sensor
                       All
                               130

-------
        Average CQ Benefit  per Subgygt-.em Repair --
               Carbureted MY 8l-fl2 Vehicles
                                  Garb 81 -82
CO
g/rrt
80 j
70-
60-
50--
40 •-
30 ••
20-
10-
 0
-10
-20
        .. Caib   Fuel Mtr  Fuel Inj
                 Tune
Catalyst   ECU
  O2    Load
Sensor  Sensor
         Afl
        Average  CQ Benefit per Subsystem Repair —
               Carbureted MY 83-86 Vehicles
                                  Cart) 83-86
CO
g/rrt
80 j
70-
60-
50-
40-
30-
20"
10-
 0
-10
-20
        .. Cart)   Fuel Mtr  Fuel Inj  Catalyst
                 Tune
        ECU
  O2
Sensor
 Load
Sensor
                             131

-------
        Average CO  Benefit:  per Subsystem Repair —
              Fuel In-ierted MY  81-82 Vehicles
                                    FI81-82
 CO
g/mi
 80 j
 70 -•
 60 ••
 50 -•
 40 ••
 30 -•
 20 ••
 10 ••
  0
-10
-20
        4.  Garb   FuelMtr  Fuellnj  Catalyst
                  Tune
ECU
 O2
Sensor
 Load
Sensor
All
        Average CO Benefit per Subsystem Repair —
              Fuel  Injected MY  81-82 Vehicles
                                   Fl 83-86
CO
g/mi
80-r
70 -•
60 ••
50--
40 -•
30 -•
20 ••
10 -
 0 4-
-10
-20
        .. Carb   FuelMtr   Fuellnj  Catalyst
                  Tune
        O2     Load
      Sensor  Sensor
                 All
                              132

-------
  APPENDIX  Z:
TOTAL ESTIMATED EMISSION REDUCTIONS FOR  ALL
          SUBSYSTEMS
     This table summarizes  estimates  of  the  total  emission
reductions, in g/mi and percent of overall CTP  fleet
reduction, realized by repairs to specific subsystems,  as
measured by the LA4.  Average reductions per repair are
calculated from isolatable  repairs only.  Number of repairs
includes all repairs to that subsystem,  whether or not
isolatable.  Totals greater than 100% are due to the
SUBSYSTEM
REPAIRED
INDUCTION SYS1
Htd Air Door
Temp Sensors
Air Filter
Hoses
Other (Indt)

N
'EM
2
2
15
6
3
FUEL METERING SYS1
CarbAssmbly
Fuel Meter Tune
Idl Spd Sole
Fuel Inj
Hoses
Other
ChkAdjVacm
Vac Olaphrms
Other (Chk)
27
43
1
20
7
11
1
5
2
IGNITION SYSTEM
Dist Assembly
Igni Tune Items
Vac Adv Assmb
Spk Delay Dev
Elect Tim Mod
Hoses
Wir/Hms/Fuse
Other
EGR SYSTEM
Valv Assembly
Delay Solnoid
Cool Temp Sen
Hoses
Other
11
54
5
1
1
2
2
2

12
1
1
8
5
AVG REDUCTION
PER REPAIR
HC CO

0.06 -1.6
-
•0.01 2.2
1.13 -2.3
0.00 -0.2
EM
1.03 11.9
0.61 11.6
-
2.35 24.2
-0.09 0.2
2.35 29.7
0.01 0.5
-0.06 -1.0
-0.01 -0.4

0.00 -0.5
0.34 0.4
-0.19 -1.2
0.28 10.4
- -
0.01 -0.5
- -
- -

-0.01 0.6
„
-0.02 0.3
0.05 -1.4
0.62 1 1 .3
AIR INJECTION SYSTEM
Pump Assembly
Byps/Dump Vlv
Diverter Vlv
5
3
13
7.93 41.1
-
-1.45 -3.4
ESTIMATE OF
TOTAL REDUCTION
HC CO

0.12 -3.1
-
-0.17 33.2
6.79 -13.8
0.01 -0.5

27.73 320.5
26.29 498.8
-
47.09 484.3
-0.64 1.3
25.82 327.2
0.01 0.5
-0.30 -5.0
-0.02 -0.9

-0.02 -5.2
18.59 19.8
-0.95 -6.0
0.28 10.4
— —
0.02 -1.1
— —
— —

-0.11 6.7
- -
-0.02 0.3
0.37 -10.9
3.09 56.4

39.66 205.6
-
-18.87 -43.7
% OF ENTIRE
CTP REDUCTION
HC CO

0.0% -0.1%
- -
-0.1% 0.7%
2.1% -0.3%
0.0% 0.0%

8.6% 6.3%
8.2% 9.9%
-
14.6% 9.6%
-0.2% 0.0%
8.0% 6.5%
0.0% 0.0%
-0.1% -0.1%
0.0% 0.0%

0.0% -0.1%
5.8% 0.4%
-0.3% -0.1%
0.1% 0.2%
— —
0.0% 0.0%
- -
— —
—
0.0% 0.1%
-
0.0% 0.0%
0.1% -0.2%
1.0% 1.1%

12.3% 4.1%
- -
-5.9% -0.9%
                              133

-------
SUBSYSTEM
REPAIRED

N
AVG REDUCTION
PER REPAIR
HC CO
AIR INJECTION SYSTEM (continued)
Check Valve
Drive Belt
Hoses
Wir/Hms/Fuse
Other
PCV SYSTEM
Valv Assembly
Filters
Hoses/Lines
Other
31
5
27
2
7

3
4
3
1
EXHAUST SYSTEM
Exh Manifold
Catalyst
Other
10
56
2
0.35 3.8
-0.04 0.1
-0.24 0.5
0.02 0.4
-0.30 -3.7

-0.12 -0.8
- -
-9.68 -35.3
— —

- —
1.11 7.0
0.50 12.3
EVAPORATIVE SYSTEM
Evap Canister
Canister Purg
Hoses
Other
1
2
3
2
ENGINE ASSEMBLY
Eng Assembly
Cooling Sys
Valve Adj
Belt Tension
Hoses
Eng Oil
Other
3
7
4
1
1
2
3
0.08 0.3
0.58 5.8
0.02 0.2
- —

0.63 16.6
-0.06 -1.5
0.11 1.5
- -
5.99 28.7
-0.16 -0.8
0.46 1.9
THREE-WAY CATALYST SYSTEM
ECU
02 Sen
Load Sensor
Eng Spd Sen
Cool Temp Sen
EGR Postn Sen
A/F Cntrt Act
Air Bypas Sen
Air Dtvrt Act
ISC Sys
Hoses
Other
MAT Sensor
Wir/Hrns/Fuse
ALL
19
82
22
2
8
1
5
3
4
1
18
2
3
9
630
0.40 10.7
0.80 20.7
0.61 23.2
0.53 3.4
0.86 27.4
0.06 -12.8
2.46 32.6
2.76 28.2
0.12 -0.4
0.20 2.0
0.37 -3.7
-0.02 -0.7
-0.55 10.2
0.43 26.7
0.70 10.7
ESTIMATE OF
TOTAL REDUCTION
HC CO

10.77 117.6
-0.18 0.7
-6.41 12.3
0.04 0.8
-2.08 -25.9

-0.35 -2.4
— —
-29.05 -105.8
— _

— —
62.18 393.0
1.01 24.6

0.08 0.3
1.16 11.6
0.06 0.5
— —

1.89 49.7
-0.41 -10.8
0.46 6.1
— —
5.99 28.7
-0.32 -1.6
1.39 5.7

7.68 203.5
65.61 1698.0
13.35 510.6
1.07 6.8
6.91 218.8
0.06 -12.8
12.32 163.1
8.28 84.5
0.49 -1.6
0.20 2.0
6.73 -66.5
-0.04 -1.4
-1.65 30.5
3.89 240.3
345.89 5455.8
% OF ENTIRE
CTP REDUCTION
HC CO

3.3% 2.3%
-0.1% 0.0%
-2.0% 0.2%
0.0% 0.0%
-0.6% -0.5%

-0.1% 0.0%
_ _
-9.0% -2.1%
_ _

— —
19.3% 7.8%
0.3% 0.5%

0.0% 0.0%
0.4% 0.2%
0.0% 0.0%
_ _

0.6% 1 .0%
-0.1% -0.2%
0.1% 0.1%
— —
1 .9% 0.6%
-0.1% 0.0%
0.4% 0.1%

2.4% 4.0%
20.4% 33.5%
4.1% 10.1%
0.3% 0.1%
2.1% 4.3%
0.0% -0.3%
3.8% 3.2%
2.6% 1 .7%
0.2% 0.0%
0.1% 0.0%
2.1% -1.3%
0.0% 0.0%
-0.5% 0.6%
1 .2% 4.7%
107.5% 107.8%
134

-------
1  For a more complete description of the CTP program objectives,  refer
   to  "Program  Plan:    A  Cooperative  EPA/Manufacturer  I/M  Testing
   Program," U.S. EPA,  Office of  Mobile Sources, ECTD/TSS, January 1987.

2  At the time the CTP  was being  designed, the Michigan AET database did
   not include fields for vehicle type  (LOT or LDV), fuel metering type,
   and  other  vehicle   identifiers  that  were  factors  in  the  CTP
   recruitment.  Consequently,  Michigan AET program statistics available
   to  EPA  were  not  specific  enough  to  set  the  program  quotas.
   Nevertheless, some summary  information from  those  data  are provided
   in Appendix A.  For more details on the recruitment quotas, refer to
   the CTP program plan op. cit..

3  The safety  and outlier  rejection  criteria  were baaed upon  the EPA
   Emissions Factors  recruitment criteria,  and included off-road use,
   major  engine  modifications,  and  excessive  towing.   Other  such
   criteria were evaluated at. the  time  of  vehicle intake at the test
   site.

4  Some manufacturers  chose to begin their as-received  testing with a
   simulation of the original AET test.   Such a test was not part of the
   program plan, however, and data for these tests are not stored in the
   common CTP database.

5  CTP Program Plan, op. cit.,  pp. 21-25.

6  This  fact  is the basis  for including a restart requirement in the
   testing of  Ford vehicles with unloaded versions  of the EPA-approved
   performance warranty short tests (40 CFR 85.2201-2212)

7  The criteria  in the  I/M Variability category actually involve a set
   of numerical  comparisons between the emission results in the various
   modes and sequences of the Basic I/M Test Procedure.   Details may be
   found in the CTP Program Plan op. cit.

8  A generalized flow diagram of decision  making in the  CTP remedial
   maintenance phase appears in Figure 3  of the program plan,  op. cit.

9  Vehicle 256, a Ford Topaz, suffered a  transmission failure during as-
   received  testing  and  was   removed  from  the  program  by  the
   manufacturer.  Vehicle 344, an Buick Electra, had a substantial leak
   in  the catalyst  as-received,  and was  not  FTP tested for safety
   reasons; post-repair FTP testing was not performed due to the lack of
   an emissions baseline for the  vehicle.

10 initial testing for three of the four  (vehicles 215, 255, and 303)was
   terminated for safety reasons  due to catalyst overtemperature, traced
   to malperformance of other components; the fourth had an ECM failure
   traced to a disconnected coolant temperature sensor.

11 Unless otherwise specified, "AMC" encompasses AMC,  Jeep,  and Renault
   nameplates  and "VW"  encompasses VW of America  and  VW of Germany.
   Divisions  of other  manufacturers are grouped under the  principal
   manufacturer  name  (e.g.,  Chrysler,  Dodge,  and Plymouth  are  grouped
   under Chrysler).
                                  135

-------
12  Unless  otherwise noted,  the term  "CO-only  failure" refers  to a
   vehicle that fails the relevant procedure (FTP, short cycle) for CO
   but  passes  HC;   i.e.,   NOX  is   ignored  in  the  failure-type
   classification.   Failures  for "HC-only"  are  handled analogously.  If
   no suffix  appears on  the failure  type (e.g., "CO  failure"), the
   classification  was  made  blind to  the pass/fail  status  of other
   pollutants.   To simplify the tables and figures in this report, the
   "blind" category is  rarely presented on  its  own, but the ordering of
   the other failure  types has been chosen to permit easy  summing of the
   HC-only or CO-only  category with the HC+CO category, yielding the
   failures for  one pollutant that are blind to the other.

13See,  for  example,  Glover,  E. L.,  and Brzezinski, D. J.,  MQBTLE4
   Exhauat  Emission  Factors and  Inspection/Maintenance Benefits for
   Paaaenyer Cars.  US E.P.A technical report EPA-AA-TSS-I/M-89-3  (1989).

14The upper bounds  for marginal emitters  in MOBZLE4 were determined by
   projecting a log-normal distribution  onto  the emissions  of a large
   sample of in-use vehicles in each of  several  technology categories,
   applying a two-standard-deviation cutoff,  and then back-calculating
   the emission  values that corresponded to the cutoff.

15Limited  sample   sizes  prevented  application  of the  four  emitter
   categories to light-duty  trucks  in the development of  the MOBILE4
   model.   We have applied the categories to both LDVs and LOTs in the
   CTP analysis, however.

16  Recall  from Section  2.4  that  Ford vehicles alone  received  a
   keyoff/restart  step between idles in  the  core  sampling  periods
   throughout all  sequences  of the BITP except  the  Restart sequence.
   Because this  was the  baseline condition  for  Fords,  the effect of the
   procedure design  is still to isolate  the  impact of the intervening
   engine operation,  which is the significant point for the analysis to
   follow.

17  Note by  recalling Table2   in Section 2.4 that  because  the  restart
   step followa the RS-02  mode,  XL-05 and  RS-02  are procedurally
   identical.

18  Almost  all of the  vehicles that were lost to this analysis were ones
   that received no Indolene  Extended Loaded sequence following the as-
   received FTP.   Such exclusions included  all of the  Chryslers and
   scattered cases from the  other  participants.   Low  incoming  fuel
   levels  in two  vehicles  (33 and  240)  prompted  substitution  of
   commercial fuel during the  Extended  Loaded sequences of  the  tank
   BITP.

19  The anomolous  RVP of  5.0 for vehicle  613  was verified  with the
   manufacturer  (Nissan) but  remains unexplained.

20  "Excess" emissions are defined as that  portion of the emissions above
   the certification level   for  the  vehicle,   with HC and  CO treated
   separately.   "Total  excess"  is  the  sum  of the individual excess
   emissions of each vehicle.   At this point  and for  the remainder of
   this report,  the  excess is  set to  zero  for clean  vehicles — those
   whose  emissions  are  below  the  certification  standard.   Earlier
   analyses in this report treated clean vehicles as negative excess.
                                  136

-------
21
   Glover, E. L., and Brzezinslci,  D. J.,  op. cit.
                                   137

-------