.   EPA-AA-IMS/81-7
                            Low-Cost Approaches to
                 Vehicle Emissions Inspection and Mai'ntena ice
                                 October 1981
                                    Notice
This Report  does  not. necessarily represent  final  EPA decisions  or  positions.
It is intended to present technical analysis of  the  issue 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.
                       Inspection and Maintenance Staff
                     Emission Control 'Technology Division
                 Office  of  Mobile  Source Air  Pollution  Control
                      Office  of Air,  Noise, and  Radiation
                     U.S. Environmental Protection Agency
                              Ann  Arbor, Michigan

-------
                              TABLE OF CONTENTS
Section                                                               Page

1.0 INTRODUCTION AND SUMMARY                                            4
    1.1 Introduction                                                    4
    1.2 Summary of Key Results                                          6
    1.3 Example Designs For Low-Cost I/M Programs                      22
    1.3.1   Example Program A:                                         22
            Areas Requiring Only CO Reductions                         24
    1.3.2   Example Program B:
            Areas Requiring HC Reductions
    1.3.3   Example Program C: Maximum Benefit Program                 27

2.0 BACKGROUND ON CONVENTIONAL I/M PROGRAMS                            31
    2.1 Inspection and Reinspection                                    31
    2.2 Types of Repairs Performed                                     32
    2.3 Cost of Repairs                                                33

3.0 REDUCING REPAIR COSTS BY REDUCING INAPPROPRIATE                    34
   . AND UNNECESSARY REPAIRS
    3.1 Improvement 1:  Mechanic Training                              34
        3.1.1   Mechanic Training Courses                              34
        3.1.2   Information Distribution                               35
        3.1.3   Problem Facility Identification                        37
    3.2 Improvement 2:  Public Awareness                               37
    3.3 Improvement 3:  Price Competition                              37
    3.4 Improvement 4:  Automated Analyzers                            37
    3.5 Improvement 5:  Challenge Stations                             38
    3.6 Impacts on Costs and Emission Reductions                       38
    3.7 Conclusions                                                    38

4.0 I/M FOR PRE-1981 VEHICLES USING ONLY IDLE CO                       39
    OUTPOINTS   :     .
    4.1 Description and Summary of Impacts                             39
    4.2 Repair Types and Costs                                         42
        4.2.1   Elimination of Idle HC-Only Failures                    42
                and Repairs
        4.2.2   Simpler Repairs for Other Failed Vehicles               43
        4.2.3   Net Effect on I/M Costs                                46
    4.3 Emission Peductions                                            46
        4.3.1   Carbon Monoxide Emissions                              47
        4.3.2   Hydrocarbon Emissions                                  48
        4.3.3   Deterioration Issues                                   49
        4.3.4   Emission Reduction Benefits Model for                   49
                Idle CO I/M Programs for Pre-1981 Vehicles
    4.4 Fuel Economy Benefits                                          50

-------
                               TABLE OF CONTENTS
                                  (continued)
Section                                                               Page

5.0 I/M FOR 1981 AND LATER VEHICLES USING ONLY IDLE                    55
    CO CUTPOINTS
    5.1 Background on 1981 and Later Vehicles                          55
    5.2 Idle CO I/M for 1981 and Later Vehicles                        56
    5.3 Emission Benefits Model for Idle'CO I/M                        57
        Programs for 1981 and Later Vehicles

6.0 OVERALL EMISSION REDUCTION BENEFITS FOR I/M PROGRAMS               60
    USING ONLY IDLE CO CUTPOINTS

7.0 METHODS TO IMPROVE THE HC AND CO EMISSION REDUCTIONS               62
    FROM IDLE CO I/M PROGRAMS
    7.1 More Effective Test Procedure for 1981 and Later               62
        Vehicles
    7.2 Higher Failure Rate for Pre-1981 Vehicles                      63
    7.3 Better Repairs                                                 63
    7.4 Re-establishment of a Loose Idle HC Cutpoint                   70
    7.5 Tampering Checks                                               72
        7.5.1   Background                                             72
        7.5.2   Air Pump Checks                                        74
        7.5.3   Evaporative Emission Control System Check              77
        7.5.4   Catalyst Removal Check                                 79
    7.6 Inspection and Maintenance for Light Duty Trucks               81
References                                                             85

-------
1.0 INTRODUCTION AND SUMMARY

1.1 Introduction

The cost of automobile  Inspection  and  Maintenance (l/M)  programs is one of the
main  topics  in  any  discussion  of I/M  as  a  strategy  for   reducing  in-use
automobile emissions of hydrocarbons  (HC)  and carbon monoxide (CO).  While I/M
is a cost-effective means  to  reduce these  pollutants in  urban areas, there are
aspects  of  the most  common —  or "conventional"  —  approaches to  I/M which
could be varied  and which would reduce  the  cost of repairs  necessary  to pass
I/M and,  in  some  cases,  the  inconvenience  which  I/M causes  vehicle  owners.
This report will  focus on possible approaches  to I/M which  inherently  can be
expected to have lower  repair  costs than  the conventional approaches to I/M or
which can be  expected  to result  in fuel savings which will  offset  some  of the
cost of  the I/M program.   The report  evaluates the  emission  reductions that
can be  obtained from  each of  these   approaches  relative to  the conventional
approach.

The low-cost  forms  of  I/M  are  basically only modifications of the conventional
form, and fall into two categories.  First,   there  are I/M programs which have
the same basic  inspection and reinspection  requirements  as  a conventional I/M
program  but  include   improvements which   aim  at  reducing  the  number  of
inappropriate  and  unnecessary  repairs,  thereby lowering  the   average  repair
cost to  vehicle  owners.   I/M programs  such  as these will achieve  lower costs
with no loss in emission reduction  effectiveness.

Second,  there  are  I/M programs which  differ  from conventional   I/M  in  that no
HC  cutpoint is  used in the inspection  and  reinspection.   This  reduces  the
number  of   vehicles which  fail  inspection  and  changes  the  type   of  repairs
required  for  failed  vehicles to  pass reinspection.  The result  is a  lower
overall repair cost.   This cost  reduction is accompanied by  a. partial  loss.in
emission reduction effectiveness  for  HC.   There  are,, however,  some  improve-
ments  and  add-on  procedures  which are helpful in  increasing   the  overall  HC
emission reductions from  this second  type of  low-cost  I/M program  and which
are discussed  in this  report.   These  improvements  can make  this type  of I/M
program  more  acceptable as a  substitute  for  a conventional  I/M  program  in
urban areas which  need HC reductions  in  order to attain  the National  Ambient
Air  Quality  Standard   for  ozone.   Some  of  these  improvements  and  add-on
procedures  can also be  used in  a  conventional I/M program to  increase  its HC
effectiveness, and  are  therefore of general interest.

Because  both  forms  of  low-cost  I/M are only variations  from  the conventional
I/M program,  the  agency responsible for planning,  implementing,  and operating
a  low-cost  I/M  program will  need to  perform  all  the   tasks   it  would in  a
conventional I/M program.  These tasks include:

    0   Development of  an  Implementation Schedule.

    0   Preparation of  draft  legislation,  if  adequate   legislative  authority
        does not yet exist or is not compatible with the  low-cost approach.

    0   Initiation  of a  Public Information Program.

-------
     0    Initial  notification  of  the  repair  industry  of  program  plans  and
         schedule.

     0    Development  and  award  of any necessary contracts.

     0    Construction of  facilities  if  necessary (including challenge  stations
         in  a  decentralized  program).

     0    Adoption of  testing procedures  and  guidelines.

     0    Selection  of geographic coverage,  vehicles  subject  to  inspection,'and
         inspection cutpoints.

     0  '  Licensing of test facilities  and  testing equipment.

     0    Development  of test facility  auditing procedures and schedules.

     0    Purchase and delivery  of necessary  equipment.

     0    Initiation of mechanic training programs.

     0    Hiring and training of inspectors.

     0    Phase-in of  mandatory  program.

These  tasks are  defined in more detail  in EPA's  I/M policy  memo  of July 17,
1978  [l]r*   In addition, the improvements which  seek  to  lower repair costs by
eliminating  inappropriate  and unnecessary  repairs  require additional involve-
ment by  the responsible  agency.

In this  report's discussion of various  approaches to I/M there will often be a
distinction made between pre-1981 model year vehicles and 1981 and later model
year'vehicles.   This distinction  is  necessary  to account  for the significant
technology  differences between these two  groups.  These technology differences
cause  differences  in  the   frequency  and  types  of  emission  control  system
maladjustments  and  malfunctions  and the  types of  repairs  needed  to correct
these  problems.  There are  also differences in warranty coverage between these
two  groups  of vehicles.  These differences  are  primarily caused -by  the  more
stringent Federal standards for CO and  NOx emissions and the  implementation of
Parameter   Adjustment[2]  .and   Emission  Performance  Warranty  regulations[3]
beginning  in the  1981  model  year.   The  effect  of  these differences will be
discussed in  greater detail later  in this report.  There is  more information
available  at  this   time about  the  costs  and  effectiveness  'of  various  I/M
strategies  on pre-1981 vehicles than on 1981  and  later  vehicles and  this  will
be reflected  in the  detail  with which the two  technology types are covered.
v^••••WI(B>i^^MfcV^W^H«M^KKa^^B^A«^BMmMKHHi^B«Hm_MMBVlv^_ .         '     '          (        . '
*. Numbers in brackets refer to references at the end of the report

-------
It is expected  that the  failure  rate of  1981  and later vehicles  will  be low
(5-10%)  and some of  the  repair costs of these vehicles will  be  covered  by the
new  warranty  regulations  [2].   There  will also  be relatively  few  1981  and
later  vehicles   in  the   fleet  in  the  early  years of  a   new  I/M  program.
Therefore as  the programs begin  the repair cost .for  1981  and  later  vehicles
will not be as  important to  the  overall cost of an I/M  program  as the  repair
cost for pre-1981 vehicles.

1,2 Summary of Key Results

The most far  reaching  conclusion in  this  report is that inspecting  cars only
for carbon monoxide  (CO) emissions will  not reduce the  CO  emission reduction
of an I/M program and will reduce the HC emission reduction  by only a moderate
amount.

Repairs will  be  much simpler  and  cost much  less  in an I/M  program  that uses
only  idle  CO  cutpoints  (the "idle  CO I/M  program").   Carburetor  adjustments
will usually  be  the only necessary  repair for  pre-1981 vehicles.   These will
cost between  $6  and $10, compared to average  repair costs  from $18 to  $30 in
conventional  I/M programs.   For  1981 and  later vehicles,  using  only a  CO
cutpoint  means   that  only   cars  which  are  suffering  a  malfunction   in  the
computer-controlled  fuel system will fail the  inspection; in the  conventional
program,  some cars  fail  for high  HC  due  to  other  types   of  malfunctions.
Repair  of  fuel  system  malfunctions   causes  a  sizable  fuel   economy  benefit,
while repair  of other malfunctions  often does  not.   Consequently,  the  same
overall  fuel  savings will  be achieved  in an  idle  CO  I/M program as  in  the
conventional  program but the  failure rate,  and consequently  the  total  repair
cost, for 1981 and later vehicles will be  less.

The bottom line  for any low-cost approach to  I/M programs  is not  only  a more
favorable cost-effectiveness  value,  but an adequate reduction in  total  HC and
CO emissions  from  automobiles.   This reflects the need of areas  requiring I/M
programs to reduce these  pollutants  to attain the National  Ambient  Air Quality
Standards for ozone  and  CO.   The tables in  this  section  will provide  informa-
tion which can help  I/M  program planners  assemble  a  low-cost  I/M program using
the  idle CO  I/M  approach  as  a  base  and adding  various   emission  or  cost
enhancement options.  Options  can  be selected from the tables in  this section
to meet an  area's  individual  needs  regarding design,  costs,  and  benefits.
Following sections of  this report provide  substantiation  and  derivation  of the
benefits presented in these tables.        .    • ..

The methods used in this  section  to  compute  the  costs and  emission  benefits
from  the various  options  are  identical  to  those used in "Update  on  the
Cost-Effectiveness of  Inspection and  Maintenance".[4]   The  estimates of  fuel
economy  benefits  are derived  using   the methods described  in  "Update  on  the
Fuel Economy  Benefits  of Inspection  and Maintenance Programs".[5]  Each table
assumes  a standard  fleet  of  one million light duty  gasoline  powered  passenger
vehicles  subject  to  the   I/M   program.   All  operating  conditions   (speed,
temperature,  etc.) are national averages.

-------
Table 1  compares  the basic idle  CO  I/M program to  a conventional  I/M program
using information from Section  6.0.   Both programs would have  the  same idle CO
cutpoints, but since the idle CO  program does not use an idle  HC  cutpoint,  all
of the  vehicles  which  would have  failed for idle HC  only  in  the  conventional
I/M  program  will   pass   in   the  idle  CO  program.   Therefore,   while  the
conventional I/M program has  stringency of  20%  for pre-1981 vehicles  the idle
CO program  has  a  stringency  of  about   13%  using  the same  idle CO  cutpoints.
Both programs use the idle test for  1981 and later vehicles.


                                         Table 1

                            Comparison  of  Basic  Idle CO and
                                Conventional I/M Programs
            Stringency
               for
            Pre-1981
            Vehicles
            (Percent)
                       Percent
                      Benefit on
                     Dec. 31, 1987
                       HC     CO
  Five Year
  Emission
  Reduction
(Thousands of
Tons Removed)
 HC      CO
  Five Year
Program Cost^
(Millions of $)
Cost-
Effectiveness
($/ton)
  HC   CO
Conventional
    I/M        20

Idle CO        13
  I/M
                      34.9   33.1   46.50   526.8l/       54.10        581   5ll/

                      24.2   33.1   31.31   526.8         40.19.       642   38
1:


2:

3:
Costs  include  inspection  fee ; and  repair  costs  less vany; fuel  economy
benefits from repairs.-  •.  .-• vV '".'••''::"3;'.•'•'• ~' •".•'.' '•''-~:?--'•?-*':#-£'•&'£•'••.'~'-zC.^-' .'-- '-""•'

Program costs are divided equally  between the two  pollutants.  :.,.*.;, ("v -    - C:

These values  have been  recalculated and 'are nearly  but, not. exactly  the
same  as  reported in  "Update  on  the Cost-EffectivenessV of .Inspection  and
Maintenance". [4]     •      .    •"."''"  •'•''.''  •',;'"•'••    " '•;'-' ' ~:'"f '• .•:-•'.. '''."'•':•.-'?'
Since  only those I/M  repairs  which effect. CO emissions will  be required  the
basic  idle CO I/M program will provide the same CO benefit as  the  conventional
I/M  program,  but with  overall repair  costs  for pre-1981  model year  vehicles
reduced by about 70%.   Since  repair costs .are only one part  of  overall, program
costs, overall  program costs  in  the  basic idle CO.program  are about .26%  less
than   for   a  conventional  I/M   program.    This  reduction-;, is  entirely   in
out-of-pocket  repair  expenses-to  vehicle  owners.   The  cost-effectiveness  of
the  basic  idle  CO program  is  better for CO  than  the conventional  approach  to
I/M  and  slightly worse for HC.   An idle CO  program,  however,  lends  itself  to

-------
several  optimizations  which will  greatly  increase  the cost-effectiveness  of
the program and  will  provide other  unique  benefits  to vehicle  owners.   These
optimizations are described  briefly  in this  section  and  example low  cost. I/M
programs using these options are described in Section 1.3.

Overall, each enhancement  option strives  to  increase  the  effectiveness  and/or
cost-effectiveness of an  idle  CO I/M  program.   There are three  basic methods
to achieve  this  goal:  (1)  achieve  greater emission  or fuel  economy  benefits
from each failing vehicle,  (2)  fail  more  of  the vehicles  which  are  inspected,
(3) inspect more vehicles in the I/M area.  Each of  the following options will
use one or more of these methods.

Better Test for 1981 and Later Vehicles

Although  the  basic  idle  CO  I/M program  provides  the same  CO  benefit  as  a
conventional  I/M program,  there is a considerable  shortfall  in HC  benefit.
One of  the  simplist ways  to increase the  HC benefit of idle CO  programs is  to
use an  inspection test for  1981  and  later vehicles which  is more effective  in
identifying vehicles with full-rich fuel  system  failures.  This  is  discussed
in Section  7.1.   Increasing the  failure  rate of  these vehicles will provide
large HC and CO emission  benefits for  each  repair  dollar  spent.   Also, studies
have shown  that repairs of  vehicles  with  full-rich  fuel system failures result
in a 15%  fuel economy  benefit.[6]   The fuel savings  from  the  repair  more than
offset  the  expected average  repair  cost  of  about  S30.   Using  the  Two-Speed
Idle or  Loaded  Test for  1981  and  later vehicles  instead  of the idle test  is
expected  to.increase  the  .identification rate  of  1981 and later  vehicles with
fuel system failures from 50% to 70%.   Because at  each inspection less than 5%
of 1981  and later vehicles are  expected  to have  a fuel system  failure, their
failure rate on the I/M test will still be only a few percent.

Table 2  presents  the additional benefits associated  with  the  use of  a better
test  for 1981  and  later vehicles.   When combined  with  the  basic  idle  CO
program, the HC benefit is  increased 2.4  percentage  points from  24.2% to 26.6%
and the  CO  benefit  increases 5.8 percentage  points  from   33.1%  to  38.9%.  The
averaged cost-effectiveness of the I/M program with  this  option  is  improved to
$527/ton  for HC  and  &30/ton  for  CO   (not  shown  in  the  table, but  can  be
calculated  from  the  information in  the  tables).   The overall HC benefit with
this enhancement alone is still  less than for a conventional  I/M program.

-------
                                       Table 2

                                   Better Test for
                               1981 and Later Vehicles2
                                in an Idle CO Program
         Additional
          Percent..
         Benefit on
         Dec. 31, 1987
           H£   C£

           2.4  5.8
Additional
 Five Year
 Emissions
  Benefit
(Thousands of
Tons Removed)
 HC     CO

 2.03   60.4
Additional
 Five Year
Program Cost^
(Millions of $)
     -5.05
Incremental
Cost-
Effectiveness
($/ton) of Option
       NA3
1:  Costs  include  additional  repair costs  less  any fuel economy benefits  from
    repairs.
2:  Using  the Two-Speed  Idle or  Loaded Test  instead  of  the  idle  test  will
    increase  the  identification rate of 1981 and  later vehicles  from 50%  to
    70%.
3:  The 15% fuel economy benefit  savings in fuel more than offset  repair costs
    for 1981  and  later vehicles with full  rich  fuel system failures.   Overall
    program costs are reduced by the amount shown.
Increased Stringency                            .              .         .

Another  way to  improve  the cost-effectiveness of an  I/M  program  is to  fail
more  of  the inspected vehicles.   This can be  done in  an  idle  CO program by
increasing  the  stringency  for  pre-1981 vehicles.   This is discussed in Section
7.2.  This  will  not  increase inspection costs but will increase overall  repair
costs.   Up  to  a point,  the increase in stringency  will  result  in  increases in
HC and CO emission benefits at a  lower incremental cost per ton than  the  basic
program, since  the inspection costs have already been paid.   •

Table 3  presents  the additional  benefits from  increasing  the  stringency  in an
idle  CO  I/M  program.    The   costs  include  only the  costs  resulting   from
additional  repairs.  The basic idle  CO program  has already  absorbed all of the
inspection  cost.  Note that increasing  the  stringency  from  30%  to  40% does not
substantially  increase   the HC  and  CO benefits   but  will  cost  as  much as
increasing  the  stringency  from 20%  to 30%.  This  is  the point  where failing
more  vehicles  will result  in  less cost-effective  emission  reductions than in
the  basic   program.   For this reason a  stringency . of  more .than  30%  is not
recommended  for  an idle  CO I/M program  unless  no other option  to  increase HC
and CO emission benefits is found acceptable.   .           +.*•..•'.,.  ••  .  . ,,. .'. ..

-------
                                       10

                                         Table 3

                                 Increasing  Failure  Rates
                                  in an Idle CO Program
Got ion
 Additional
  Percent
 Benefit on
Dec. 31, 1987
   HC   CO
Additional
 Five Year
 .Emission
. Reduction
(Thousands of
Tons Removed)
  HC     CO
                                                         Additional
                                                           Five Year
                                                        Program Cost*
                                                        (Millions of £)
                        Incremental
                        Cost-
                        Effectiveness^
                        ($/ton)
                          HC   CO
Higher Stringency for
Pre-1981 Vehicles

Increased from 13% to 20%
Increased from 20% to 30%
Increased from 30% to 40%
   2.4  3.3
   3.1  2.8
   1.1  0.8
 5.07
 7.23
 2.03
80.6
85.2
30.4
2.06
2.95
2.95
203   13
204   17
727   49
1:  Costs include  additional repair costs  less  any  fuel economy benefits from
    repairs.
2:  Program costs are divided equally between  the two pollutants.
Better Repairs for Pre-1981 Vehicles

Better repairs  for  pre-1981 vehicles deserve  special  attention primarily as a
cost-saving measure, although  they are  also an HC  and CO enhancement.  Better
repairs therefore deserve  attention from  any  area  considering an  idle  CO I/M
program, not just those that might  need  it as  an  HC or CO enhancement.  Better
repairs  for  pre-1981  vehicles  will produce  fuel . economy  savings   on  these
vehicles which will offset a large  part of  the cost of the I/M program.

EPA  studies  have shown  that  a  fuel  savings of  4%  is achieved  by carburetor
adjustments  to  pre-1981  vehicles which  have failed an I/M  test,  even  if the
adjustments  are  not  performed precisely  to  specification.   It is  necessary
that adjustments be performed  closer  to  specification than they  would be  in a
basic  idle  CO or  conventional I/M program,  and that  other  types of repairs
which  can  degrade  fuel  economy  be avoided.   Because carburetor  adjustments
will usually be  the only necessary repair for pre-1981 vehicles  in an idle CO
I/M program, the types of  repairs  which  can degrade  fuel  economy are  naturally
avoided in this  type of program.   Section 4.4  describes a number of approaches
that I/M administrators can take to improve the  quality  of  carburetor adjust-
ments,  two  of which  are  mechanic  training and   a.  tight  idle  CO reinspection
cutpoint.  The fuel savings which  will result  are well worth  the effort,  since

-------
                                      11

an  additional  4%  savings  for  each  pre-1981  vehicle  which  is  failed  and
repaired translates into a total annual savings  of about £36.  This  offsets  a
large part of  the  program costs.   Also, a  small  improvement  in the HC  and  CO
emission reductions accompanies  the fuel savings.   Section  7.3 discusses  the
HC and CO  benefits.  The overall effect is  a major further improvement  in  the
cost-effectiveness of  the idle CO I/M program.

Table 4 presents  the  additional  benefits of better repairs in  an idle  CO  I/M
program.   Once  a  stringency has   been  chosen,  Table  4 will   provide  the
additional benefit that  can  be achieved with better  repairs.   For example  in
the basic  idle  CO  program with a  13% stringency  the  HC benefit  in  1981  can  be
increased 0.6 percentage points from 24.2%  to 24.8%  and the  CO benefit  can  be
increased  2.7  percentage points from  33.1% to  35.8%.   Overall  costs  can  be
reduced $13.27  million as a  result of fuel savings  from repaired  vehicles.
The result is an overall cost-effectiveness of $411/ton for HC  and  $22/ton  for
CO (not shown  in the  table) .   However, the HC benefit  in 1987  from  the basic
idle CO program with  this option alone is still not as large as the HC benefit
of a  conventional  I/M program.   One or more  other  HC  enhancements   are also
necessary to equal the HC benefit  of a conventional program.

The  benefits  from better  repairs  for 1981  and  later  vehicles  cannot   be
estimated  at  present.   It  appears  that the requirement  for  vehicles to pass
the  reinspection  test  alone will  force  a  high  quality   repair  on  these
vehicles,   unlike  the  case   for  pre-1981  vehicles,  leaving  less  room  for
improvements via special programs  of training,  etc.   This occurs because most
repairs  to  1981  and  later  models  involve   parts   replacement   instead   of
adjustments.                                                  		   .   .


-------
                                      12
Stringency
   for
Pre-1981
Vehicles
(Percent)
    13
    20
    30
    40
                                         Table 4

                          Better Repairs for Pre-1981 Vehicles
                                  in an Idle CO Program
Additional
 Percent
Benefit on
Dec. 31, 1987
 HC     CO
Additional
 Five Year
 Emission
 Reduction
(Thousands of
Tons Removed)
 HC      CO
0.6
1.3
1.9
2.1
2.7
3.3
2.7
2.0
1.40
2.28
3.29
4.18
74.5
98.3
78.9
56.6
 Additional
 Five Year
Program Cost1
(Millions of $)
                                        -13.27
                                        -20.40
                                        -30.60
                                        -40.80
Incremental
Cost-
Effecti
(£/ton)
                                           N/A
                                           N/A
                                           N/A
                                           N/A
    Costs include additional  repair  costs less any fuel-economy  benefits  from
    repairs.
    Since there  is  a net  savings,  no incremental cost-effectiveness has  been
    calculated.  Overall program costs are reduced by the amount shown.
Tampering Checks

Table  5 presents  the  incremental  benefits  from  the  addition  of  tampering
checks  to  an. idle CO  I/M program.   These  checks are  discussed  in detail  in
Section 7.5.  The costs shown in Table 5  are  entirely  repair  costs,  since  none
of  the tampering  checks  is  expected  to significantly  increase  the  assumed
inspection costs.  The checks may, however, pose  administrative.problems which
can seem large to the I/M program managers.

As can  be  seen  from  the  table,  the air pump  and  evaporative  cannister  checks,
which  consists  of  checking for  the  presence  of  these devices,  are  extremely
cost-effective.   The incremental cost-effectiveness for  the catalyst  check  for
the presence  of the catalyst  on all model  year vehicles  is  higher than  any
cost-effectiveness value presented so  far in  this summary, but is still lower
than for many non-I/M HC control strategies.[4]   The cost  of  replacing  missing
catalysts on  older cars  is the  reason  for  the high cost-effectiveness value.
About  1.4%  of   vehicles  originally  equipped  with  catalysts  have  had   them
removed.   New  OEM  replacement  catalysts cost  from  $172 to  $320.   If   all
catalysts  removed  before  the  I/M  program  begins  in  1983  are  waived,   and
catalyst presence is checked only for 1983 and  later models from then on,  only
very  foolish  vehicle owners  would  remove  their catalysts knowing  that  they
will  be checked.   The   repair  costs  for the. catalyst  check  would  then  be
insignificant.

-------
                                       13

                                         Table 5

                                 Tampering  Check  Benefits
Dotion
Additional
 Percent
Benefit on
Dec. 31, 1987
 HC     CO
Air Pump Check    1.5    4.3

Evaporative
Cannister Check   1.3    0.0

Catalyst Check
(All Vehicles)    0.9    0.5

Catalyst Check
(1983 and Later
Model Vehicles)   0.3    0.2
Additional
 Five Year
 Emission
 Reduction
(Thousands of
Tons Retooved)
 HC      CO
                 2.56
                 1.94
        68.0
        0.0
1.35    10.7
0.20    1.6
                                                     Additional
                                                     Five Year
                                                    Program Cost*
                                                    (Millions of J>.
                      0.92
                      0.25
                                       2.00
                                       0.00
                                    Incremental
                                    Cost-
                                    Effectiveness
                                    ($/ton)
                                      HC   _CO

                                     180     7
                                     129l/  -
                                     741
                                            93
                                     NA4    NA4
1:  Costs include any additional repair costs.              ''       .   .
2:  Program costs are equally divided between  the  two pollutants.
3:  Since  there  is no  CO benefit,  all costs  have been  allocated  to  the  HC
    benefit.     .                           ...... • :.._ ....  ...  . .   -     ;;,i;.
4:  Since  there  is  no  significant  cost,  no   cost-effectiveness   has  been
    calculated. '          .     •  •'    . .'''    " :  . '''  -'••''. •'="..-'". ^ • "„'.":-...  "\:.'.:~  '.'"•-'
Loose HC Cutpoints

Table  6 presents  the  additional benefits  of  including  very  loose  idle  HC
outpoints with an  idle  CO I/M program. . This option  is  discussed in  detail  in
Section  7.4.   Data on vehicles with  very  high  idle HC  scores are  very  sparse
so that there is a very  high  degree  of uncertainty in all estimates concerning
these  vehicles.    Section  7.4  does  make  some  conclusions  concerning  these
vehicles,  however, based  on  data and  experiences  with  vehicles  exhibiting
moderately high idle HC  scores.              ...

-------
                                      14

Repairs of vehicles which fail only  for  idle HC in the Portland program ranged
in cost  from  zero to $207, averaging  $41.  These  repairs  range from cleaning
or replacing  spark plugs to  extensive  ignition  system  diagnosis  and replace-
ments.   Program  costs  would  be  increased  by  the additional  cost  of  these
repairs  for any vehicles failing a  high  idle  HC cutpoint used in  an idle CO
program.

There  is evidence,  discussed  in Section 7.4,  that  repair  of  severe  spark
ignition  misfire,  which often  causes  very  high  idle  HC  measurements,  can
result  in substantial  fuel  economy benefits  averaging  from  about   7%  to  9%.
This would save vehicle owners who  needed and  received  such  repairs about $72
in fuel costs for that  year.

Section  7.4  estimates  that,   in  the  best  possible  case, about  half of  the
vehicles which  fail a 1000 ppm idle  HC  cutpoint will  obtain a $72 fuel economy
benefit.   The  repair  cost  of  $41  and  the  average  fuel  savings  of  $36
approximately cancel  each other, so in  the best case this op.tion  has  no  net
effect  on program costs.  As  Table  6  shows, there is an  HC  benefit.  Repairs
for vehicles  failing  idle HC  cutpoints  in an  idle  CO  program are  not expected
to increase the  CO benefit of  the program.

Other  factors,  however, not considered  above,  may prevent the  best case  from
occurring and thus affect the  cost  of  this option.   There  is reason  to believe
that  unnecessary  repairs,   essentially  eliminated by  use  of only idle CO
cutpoints, can  become  a problem again  if  any  HC cutpoint  is  used.   In  the
Port-land  program  there  are,  for  example,   ignition  repairs  performed  on
vehicles  which  fail  the idle  test for only  CO.   Such  ignition repairs are  not
necessary,* but are  sold to vehicle owners  who  are willing to allow mechanics
a free hand in  performing repair  to  pass the reinspection.  An idle  CO program
without  an  HC   cutpoint is   expected  to  have  very  few instances of   such
unnecessary  repairs,  since in virtually  all  cases the only  necessary  repair
will  be  a carburetor adjustment and vehicle owners can be  readily  made  aware
of  this  fact.   Including an  idle   HC  cutpoint  makes  determination  of  which
repairs  are  necessary  and  which are  not more  unclear.    The  typical vehicle
owner  will again be left essentially on his own in determining which repairs
should  be performed.   As a  result,  the  amount of repairs performed, besides
carburetor adjustments, will  likely be  greater  than  what would be  needed to
make  cars which  failed the HC  cutpoint  pass  reinspection.  These  unnecessary
repairs  will  increase  the  program  costs  without  significantly  increasing HC
benefits.  This will degrade  the  cost-effectiveness  of this  option.  If,  for
example,  only  4%  of  the   inspected   vehicles  receive   unnecessary ignition
repairs   (the   same  number  as  do  need  ignition  repairs)  and  produce  no
additional HC benefit,  the  cost per tori of  HC for this option  can  climb  from
essentially   zero   to   over  $3000/ton.   Clearly,  unnecessary .repairs  can
drastically  reduce the  cost-effectiveness  of. any I/M  program.  Section  3.0
discusses approaches which can be used to  reduce  unnecessary repairs.'
*  This  statement  is  possibly overly  pessimistic,  since  it  is  possible  that
ignition repairs performed  by  the mechanic  which.would not have been necessary
to  pass  the  I/M reinspection may  prevent  engine problems in  the. future which
would cause an  I/M inspection failure.

-------
                                      15

                                    Table 6

                            Loose  Idle HC Outpoints
         Additional
          Percent
         Benefit on
         Dec. 31, 1987
           HC   CO

           2.0  0.0
                       Additional
                        Five Year
                        Emissions
                         Benefit
                       (Thousands of
                       Tons Removed)
                        HC     CO

                        2.59   0.0
Additional
 Five Year
Program Cost^
(Millions of $)
   0 - 8.98
Incremental
Cost-
Effectiveness
($/ton) of Option
    0 - 3467
1:
Costs include additional  repair  costs  less any fuel economy  benefits  from
repairs.
Inspection of Light-Duty Trucks

Table 7  presents the  additional  benefits from  inspection  and maintenance  of
light-duty  .trucks  (LDT)   as   well  as  passenger  vehicles.   This  option  is
discussed in Section 7.6.   Reduction in HC and CO emissions is  not required  of
LDTs under EPA policy, however any emission reductions gained  from LDTs  can  be
used towards total emission reduction goals for  passenger vehicles.   Since the
technology used  in  most LDTs  is  very  similar to  that  of passenger  vehicles,
I/M for LDTs is very similar to I/M for passenger vehicles.  A  4%  fuel economy
benefit has been  included  for  the cases of  better  repair of pre-1985  LDTs and
a  15%  fuel  economy  benefit  for  1985  and  later   LDTs.   'Starting  in  1985,
light-duty trucks will  have the  types  of emission controls used  on  1981 and
later passenger cars.)  Since use of a  better  test  such as the Two-Speed  Idle
or Loaded  Test for  1985  and  later LDTs  effects very  few  LDTs  by 1987,  the
effect of selecting that option on the  LDT benefits  is  insignificant.     "  .
                                                                                 *- niC* 1  t

-------
                                      16

                                         Table 7

                                I/M for Light-Duty Trucks
                                  in an Idle CO Program
Stringency
   for
Pre-1985
Vehicles
(Percent)
  Basic

    13
    20
    30
    40
Additional
 Percent
Benefit on
Dec. 31, 1987
 HC     CO
 4.2
 4.6
 5.1
 5.3
4.4
4.9
5.4
5.6
         Additional
          Five Year
          Emission
          Reduction
         (Thousands of
         Tons Removed)
          HC      CO
5.20
6.08
7.27
7.62
 71.7
 84.1
 98.6
103.0
                    Additional
                    Five Year
                   Program Cost^-
                   (Millions of $)
3.34
3.53
3.77
4.02
                              Incremental
                              Cost-
                              Effectiveness ^
                                               CO
38
35
33
33
                                                                       HC
642
581
518
527
With Better
  Repairs 3

    13          4.3    4.9      5.58     84.1
    20          4.9    5.6      6.61     100.0
    30          5.6    5.9      8.11     111.4
    40          5.8    6.0      8.58     113.6
                                         2.29
                                         1.86
                                         1.25
                                         0.62
                                               410
                                               281
                                               154
                                                72
                                               22
                                               15
                                                9
                                                5
1:  Costs include  additional  inspection and repair costs less any fuel economy
    benefits   from  repairs.    These  costs  have   been  estimated   from  the
    cost-effectiveness  values,  which in  turn were  assumed  to  be the  same as
    for passenger  cars.   ..._.. .  .  .           ...

2:  Program  costs  are   equally  divided  between  the  two  pollutants.   These
    values  have  been assumed to  be  equal  to  those for  light-duty vehicles.
    Once  a  stringency has  been  selected, only one  row from the Table  may be
    used,  determined  by  whether  or  not  the  program  chosen  will  utilize
    mechanic training or  tight reinspection  cutpoints.

3:  These  additional benefits   correspond   to  basic  idle  CO  programs  using
    tighter reinspection  cutpoints or with  full mechanic training programs.

-------
                                      17
Exempting Pre-1975 Vehicles
Another  way  to  improve  the  cost-effectiveness  is   to  reduce   the  overall
inspection  costs   by   inspecting  fewer  vehicles.    This   reduces  emission
benefits, but they can often  be  restored  by  increasing  the failure rate of  the
remaining  inspected  vehicles  and  adding  other  options  described  in  this
report.  There  is,  of course, a  limit  to which this approach  can be expected
to work.  Restricting an idle CO program to only  1981  and  later vehicles,  for
instance,  cannot  be  optimized to  provide the  same HC  and  CO  benefit  as a
conventional  I/M  program,  even  if  all vehicles  in need  of repair  could  be
identified and repaired.

Table 8 presents the benefits  of  an idle  CO  I/M program which exempts pre-1975
vehicles from testing.   An  area  requiring only CO  reductions could use  such a
program  with  a 30%  initial stringency  and  receive as much CO benefit  as a
conventional I/M program at a lower overall cost.   Areas  requiring reductions
in HC will  find they must  add even more  options to their idle  CO I/M program
if they  are  to  achieve   the same  HC benefit  as a conventional  program.   It  is
possible however to achieve the  same HC benefit as a conventional program with
an improved  overall cost-effectiveness  if enough  cost-effective  enhancements
are used.  Since exempting  pre-1975 vehicles will  often  reduce the effective-
ness of  the options in  reducing  overall HC and CO  emissions, Tables  9 through
12 have been calculated  to  be  used  with the  results presented in Table 8.   The
benefits for use of a better  test  for  1981  and later vehicles  in Table  2 are
not  affected  by  exempting pre-1975  vehicles  and  therefore can  be  combined
directly with the results in Table 8 if that option is desired.
                                         Table  8

                      Idle CO I/M for 1975 and Later Vehicles Only
Stringency
   for
 1975-80
Vehicles  '
(Percent)
 Percent
Benefit on
Dec. 31, 1987
 HO     CO
           Five Year
           Emission
           Reduction
          (Thousands of
          Tons Removed)
           HC      CO
                   Five ..Year1 •.:
                  Program Cost
                  (Millions of
                             Cost-       V ~.~ ."/•/,-
                             Effectiveness
                             ($/ton) of Option 2
                              HC   ' :'  CO   . -;;   .-.,
  13
  20
  30
  40
22.4
24.3
26.6
27.5
28.6
31.6
34.0
34.7
29.26
32.87
38.08
39-81
435.7.
502.9
572.5
598.7
32.43
34.00
36.24
38.48
554'
.517
476'
483
37
34
32
32
1:  Costs  include,  inspection  fee  and  repair . costs   less  any. fuel  economy
    benefits from repairs.                                                     .
2:  Program costs are divided equally between the two pollutants.   -      •       •.

-------
                                  18
                                    Table  9

                                 Better Repairs
             in an Idle CO Program for 1975 and Later Vehicles  Only

Stringency
for
1975-80
Vehicles
(Percent)

13
20
30
40


Additional
Percent
Benefit on
Dec. 31, 1987
HC CO
0.5 2.7
1.3 3.3
2.0 2.7
2.1 2.0
Additional
Five Year
Emission
Reduction
(Thousands of
Tons Removed)
HC CO
1.07 72.3
2.40 95.9
3.74 74.9
4.28 51.8


Additional
Five Year
Program Cost*-
(Millions of $)

-10.14
-15.58
-23.37
-31.16
                                                                  Incremental
                                                                  Cost-
                                                                  Effectiveness2
                                                                  ($/ton)
                                                                       NA
                                                                       NA
                                                                       NA
                                                                       NA
Fuel economy benefits from repairs.
Since there  is  a net  savings,  no incremental cost-effectiveness has  been
calculated.  Overall program costs are reduced by the amount  shown.

-------
                                       19
                                         Table 10

                                     Tampering Checks
                 in an  Idle  CO  Program for 1975 and Later Vehicles Only
Option

Air Pump Check
Additional
 Percent
Benefit on
Dec. 31, 1987
 HC     CO
 1.4
Evaporative
Cannister Check   1.2

Catalyst Check
(All Vehicles)    0.9

Catalyst Check
(1983 and later
models only)      0.3
3.9


0.0


0.5



0.2
Additional
 Five Year
 Emission
 Reduction
(Thousands of
Tons Removed)
 HC      CO

1.78    49.0
                 1.43    0.0
                 1.35     10.7
                 0.20     1.6
                                                     Additional
                                                     Five Year
                                                    Program Cost
                                                    (Millions of
0.92
                      0.25
                      2.00
                      0.00
Incremental
Cost-
Effectiveness 2
($/ton)
  HC_    C0_

 258    9
              1753
 741
                      93
                     NA4
1:  Costs include any additional repair  costs.               .  ;
2:  Program costs are equally divided  between  the  two  pollutants.
3:  Since  there  is  no  CO benefit,  all  costs  have been allocated  to  the  HC
    benefit.         ••••'-.   •  •;  .'•-.••'..    • •  " ;"   •'  ':-  -. •" ;*v.';•-"  ."•"•"
4:  Since  there  is  no  significant  . cost,  no   cost-effectiveness   has  been
    calculated.                     •'.;•,  -     .          •.-• .  ".   '•- .  .-..•'	'.

-------
                                      20

                                   Table 11

                           Loose  Idle HC Cutpoints
            in an Idle CO Program for 1975  and Later Vehicles Only


                           Additional
                            Five Year
         Additional         Emissions        Additional           Incremental
          Percent            Benefit          Five Year           Cost-
         Benefit  on        (Thousands of     Program Cost1        Effectiveness
         Dec.  31,  1987     Tons Removed)     (Millions of i)      ($/ton)  of Option
           HC    CO          HC     CO
           1.9  0.0         2.42   0.0         0 - 8.39             ' 0 - 3467
1:   Costs include additional repair costs  less any fuel economy benefits  from
    repairs.
repairs

-------
                                      21

                                        Table 12

                    Idle CO I/M for 1975 and Later Light-Duty Trucks
Stringency
   for
 1975-85
   LDTs
(Percent)
Additional
 Percent
Benefit on
Dec. 31, 1987
 HC     CO
         Additional
          Five Year
          Emission
          Reduction
         (Thousands of
         Tons Removed)
          HC      CO
                  Additional
                  Five Year
                 Program Cost^
                 (Millions of $)
                            Incremental
                            Cost-
                            Effectiveness
                            ($/ton) of Option2
                             HC      CO
    13
    20
    30
    40
 3.7
 4.1
 4.4
 4.5
3.4
3.9
4.4
4.5
4.58
5.22
6.05
6.27
52.7
62.6
74.3
78.0
2.54
2.70
2.88
3.03
554
517
476
483
37
34
32
32
With Better
  Repairs 3

    13          4.0    4.0      4.96    64.4
    20          4.4    4.5      5.79    77.6
    30          5.0    4.9      6.92    86.2
    40          5.1    4.9      7.27    87.6
                                         1.82
                                         1.51
                                         1.07
                                         0.60
                                               367
                                               261
                                               154
                                                83
                                               22
                                               15
                                               10
                                               "6
1:  Costs include additional inspection and repair costs  less  any  fuel  economy
    benefits  from  repairs.   These   costs   have  been  estimated   from   the
    cost-effectiveness values, which  in turn were  assumed  to  be  the  same  as
    for passenger cars.            ...   •    .     .       ;  •;' . •  "...

2:  Program  costs  are  equally  divided  between the  two  pollutants.   These
    values  have  been  assumed  to  be equal  to  those  for  "light-duty  vehicles.
    Once a  stringency has been selected,  only  one  row from  the. .Table may  be
    used,  determined   by  whether   or  not  the  program  chosen  will  utilize
    mechanic training  or tight reinspection cutpoints.         . . '      ,'.

3:  These  additional   benefits  correspond  to  basic  idle  CO  programs using
    tighter reinspection cutpoints or with full  mechanic training  programs.

-------
                                      22

1.3 Example Designs For Low-Cost I/M Programs

The previous section of this  report  presented  a number of options  for  reducing
the cost of an I/M  program  and  for restoring its emission reduction  effective-
ness  when  necessary.   This  section  selects  from  among  these  options   and
assembles  them  into   two   examples  of  low-cost  I/M  programs.    The   example
programs  are  tailored  to   two  situations.   Program  A is  an example  for  all
areas  which need  to   obtain  only  CO  reductions  from  I/M  because  they  are
already in attainment with  the  ozone NAAQS  or  will attain it by  1982.   Program
B  is  an  example  for areas  which  require  HC reductions from  I/M  to attain  the
ozone NAAQS.

The core  of the  two examples  is the idle CO I/M  program.   Program B for areas
needing HC  reductions  incorporates HC enhancements,  described  in Sections  1.2
and  7.0.   The  examples both  suggest improvements  from  Section  3.0  to reduce
further   the   incidence  of  unnecessary   and   inappropriate   repairs.    All
cost-effectiveness  comparisons  in  this   section  use  the  conventional   I/M
program  described  in   "Update  on  the  Cost-Effectiveness  of  Inspection   and
Maintenance" [4]  to represent a typical  conventional I/M program and  use  the
methodology presented  in  that  report.   This methodology  essentially  consists
of accounting for all of the  I/M costs  and  emission reductions  for a five year
period  from  January   1,    1983  through  December  31,  1987.    Costs   include
inspection and repair costs less any fuel savings from repairs.

I/M planners who. wish  to  construct their own example  program may easily do so
using  the information  in  Tables  1-7  or  Tables  8-12.   The  stringency of  the
program  and whether pre-1975 vehicles  will be  inspected  should  be  the first
tentative choice  made;  options  should  then be added  in  turn.   When selecting
options,  better  repairs   for  pre-1981 ' vehicles  (achieved  through   mechanic
training  or stringent  reinspection limits)  and a  better  test   for  1981   and
later  vehicles  should  be  considered first,  since they  yield  net  savings  in
program  costs.   Then  other  options   should   be  considered based  on their
incremental cost-effectiveness,  their ease of implementation  and administra-
tion,  and local  preferences.   If  a  satisfactory program  in  terms of  emission
reduction  benefits  does not  result,  another  choice  regarding  stringency  and
inspection of pre-1975 vehicles should be made.

1.3.1  Example Program A; Areas Requiring Only Ambient CO Reductions

Example Program A has the following principal design  features:

(l)   Only  1975  and  later  model  year  vehicles  are  subject   to  mandatory
inspection and repairs.

(2)  Idle  CO cutpoints  are  chosen  which  will  fail  13%  of  1975  through  1980
model year vehicles at the  initial inspection.[7]

(3) The Two-Speed  Idle  Test or  Loaded  Test  for 1981  and  later vehicles with a
1.2%  cutpoint on  both  modes instead of the  idle  test.  This  will increase  the
identification rate of  vehicles with  full  rich fuel system  failures  from  50%
to  70%.   Owners  qualify  for  the  Emission  Performance Warranty  .[3]  if other
conditions are also met.         -.'..••     .

-------
                                      23

(4) A  stringent  reinspection idle  CO outpoint  of  1.0% for  1975-80 vehicles.
This will  force  better  carburetor adjustments  resulting  in  significant fuel
economy  benefits  averaging  about  4%  as  well  as  additional  CO  emission
reductions.   (Similar  results  could  be  obtained  using  mechanic  training
programs.  However,  it is  difficult  to reach  every vehicle  through mechanic
training  because  some  commercial  mechanics  will  not  submit  to  training and
some vehicle owners will perform their own maintenance.)

(5)  Any  improvements  from  Section  3.0  which  are  -compatible  with  local
conditions.

Example  Program  A will  produce  somewhat  greater  CO emission benefit  than, a
conventional I/M  program or  the  basic idle  CO program but  with an improved
cost-effectiveness and  a  lower overall cost..  Table 13 presents  the emission
benefit calculations for Example Program A.

                                        Table 13                              '

                       Comparison of Emission Reductions and Costs
                               of Example Program A and  a
                         Conventional  20% Stringent I/M  Program

                                          Five Year
                           Percent         Emission
                           Emission       Reduction   Five  Year
                         .  Benefit  on     (Thousands  Program  Cost       Source
Program                    Dec.  31, ,1987  of Tons)     (Millions of &) (Section 1.2)
                            HC      CO     HC    CO
CONVENTIONAL I/M
PROGRAM:
(20% Stringency,; idle
test for 1981 and later
vehicles)

EXAMPLE PROGRAM A:

0 13% Stringent. Idle
  CO I/M exempting
  pre-1975 vehicles
0 More Effective Test
  for 1981 and later
  vehicles
0 Tight idle CO
  reinspection cutpoint
  for 1975-80 vehicles

Program A Totals
34.9
33.1   46.50 526.8
               54.10
               Table 1
22.4    28.6   29.26 435.7 .  ..32.43  7
 2.4
 0.5
 5.8
 2.7
2.03  60.4
1.07  72.3
25.3*   37.1   32.36 568.4
 -5.05
-10.14-
                       17.24 -V
                                    -.Table 8 '
Table 2
Table 9
* Note: This example is not equivalent to a conventional I/M
        program in regards to HC benefit.            .

-------
                                      24
From Table 13 it is  simple  to  calculate .the overall program cost-effectiveness
of Example Program A in  reducing CO emissions.   The  total five  year  program
costs are divided  by the total  five  year CO emission  reductions  resulting  in
an overall CO cost-effectiveness of  &30  per ton.   This  is about  a third the
cost per  ton for  CO  of the conventional  I/M  program.   Table  14  compares the
benefits  and cost-effectiveness  of  conventional  I/M,  the  basic   idle  CO I/M
program, and Example Program A.


                                   Table  14

                        Comparison of I/M Programs for
                      Areas Requiring Only CO Reductions


                             CO Emission Benefit on      Cost-Effectiveness*
Program                        December 31, 1987             (S/.Ton. CO)

Conventional I/M          .            33.1                       102
Basic Idle CO I/M                     33.1                        76
Example Program A                     37.1                        30
* All costs have been allocated to the CO benefits.
It may  be  of interest  to  point out  that  since the fuel  savings  from repairs
more than offset  repair and  inspection costs when tight reinspection cutpoints
(or  mechanic  training  programs)  are  incorporated in  the  idle   CO  program,
increasing the failure  rate  for pre-1981  vehicles  over the 13% used in Example
Program A would  actually reduce the  overall  cost  of the  program  and  the cost
per  ton of  CO.    For   example  if  the  stringency  in  Example  Program A were
increased to 20%, equal to the  stringency of the conventional I/M program, the
CO  emission  benefit   on  December   31,  1987  would  be  increased  to  40.7%
eliminating  659.2  tons, of  CO during  the  five year period.   The  total program
cost, including the  fuel savings,  would be  reduced to  $13.37 million resulting
in a cost-effectiveness  of $20/ton  of CO.   It has  been assumed in this example
that there are  local reasons for keeping  the failure  rate as  low as  possible
and that there is no need for CO benefits beyond the minimum required amount.

1.3.2 Example Program B: Areas  Requiring HC Reductions

In areas requiring HC  reductions,  HC enhancement options from Section 7.0 (and
described  briefly in   Section  1.2)  are  necessary  to  make  idle  CO  I/M  an
acceptable substitute  for  conventional I/M.  Since all  of the HC enhancements
combined would more  than fully compensate  for  the loss in HC reduction which
occurs  from  eliminating the HC cutpoints,  I/M  planners  can  choose among the
enhancements based on  convenience,  practicality, and local preferences.   There
will be no need for  CO  emission enhancements,  although many of the HC enhance-
ments  will  also  increase   CO   benefits.    The  centralized   or  decentralized

-------
                                       25

formats  will  make some of  the  enhancements more convenient  and practical than
others.   High failures rates are  not  desirable, for  example,  in a centralized
program  since each  failure implies a  fair degree  of inconvenience  to  owners
who must make two trips  as  a result, one to a  repair facility and another back
to the inspection station.   On the other hand,  using a tighter CO reinspection
cutpoint to force higher quality  carburetor adjustments would  be quite  simple
in  a  centralized  program.    Tampering  checks  are  also  appropriate   in  a
centralized program.   It   is  easier to  train  the  relatively  few centralized
inspectors  to  perform  tampering  checks  than  it   would  be  to  train  every
operator of  a  licensed  inspection station  in a  decentralized  program.   The
example  program  described  below  is just one  possibility.   The  reader  should
have no  difficulty in constructing other examples.   All  that  is  necessary  is
to select a combination  of  HC enhancements, which will restore HC effectiveness
back to  that  of a conventional  program.

Example  Program B has  the following principle  design features:

(1) All  model year vehicles 1968 and later  are  inspected.

(2) Idle CO cutpoints  which will  fail 30%  of  pre-1981 vehicles at  the initial
    inspection.[7]

(3) A reinspection cutpoint of 1.0% CO for pre-81 vehicles.   This  will achieve
    the  same  effect as a mechanic  training  program  that reaches  every  vehicle.

(4) The  Two-Speed Idle  Test  or the  Loaded Test  for  1981 and  later vehicles
    with a  1.2%  CO cutpoint on both modes,  increasing the identification  rate
    of high emitters  among  these  vehicles.  Owners  of 1981 and  later vehicles
    qualify for  the  Emission Performance Warranty  [3] if other  conditions  are
    also met.

(5) An air  pump  inspection.*  As described in  Section 7.5.2,  this is a  simple
    check to  perform.   Although  an air  pump  check will  require opening  the
    engine  compartment,  this check is  prefered over  other simple checks  such
    as a catalyst check  because  the low expected  cost of repair of  air  pump
    disablements.  The extra  time  needed to open  the hood will  be  unimportant
    in a decentralized program.      .  ,    ,-..•.   •_•;.•      ..>..-.-.  ->., i" '.   '-,•   :   -

(6) Improvements  from Section 3.0  which are compatible with local  conditions.
Example  Program B  is approximately  equal  in  HC  reduction  effectiveness and
exceeds  the  CO reduction  effectiveness  of  a  conventional I/M  program with a
20%  stringency  for  pre-1981  vehicles.    This \is  illustrated, in : Table 15.
Although the pre-1981 failure  rate  in Example  Program B is half again  as  large
as in the conventional I/M program,  the  much lower per-vehicle  repair  cost and
fuel savings will result  in a less  costly program overall.  This  is  illustrat-
ed in Table 16.         .                        .     :  •..              '%•'..'      -
*  Any one  of  the  three  tampering  checks   (catalyst,  air  pump,  evaporative
control) described  in Section  7.5  could be  used in  this  example.  Tampering
checks  for  air pumps  and evaporative- emission control  systems  would require
opening  the  engine  compartment hood  on all vehicles.   Program B as described
here would require  the  hood  to be opened on  1981  and  later vehicles anyway if
the Two-Speed Idle Test is used instead of the  Loaded  Test.   -;   - . .

-------
                                      26
                                     Table  15

                    Comparison of Emission Reductions and Costs
                            of Example Program B  and a
                      Conventional 20% Stringent  I/M Program
Program
CONVENTIONAL I/M
PROGRAM:
(20% Stringency, idle
test for 1981 and later
vehicles)

EXAMPLE PROGRAM B:

0 Basic Idle CO I/M
0 More Effective Test
  for 1981 and later
  vehicles
0 Increased Pre-1981  .
  Stringency
   -13% to 20%
   -20% to 30%
0 Tight idle CO
  reinspection cutpoint
  for'pre-1981 vehicles
0 Air pump disablement
  check
               Five Year
Percent      .^Emission
Emission       Reduction   Five Year
Benefit on     (Thousands  Program Cost       Source
Dec. 31, 1987  of Tons)    (Millions of t) (Section 1.2)
 HC      CO     HC    CO
 34.9    33.1   46.50 526.8
                       54.10
 24.2    33.1   31.31 526.8     40.19
  2.4     5.8    2.03  60.4     -5.05
              Table 1
                                     Table 1
                                     Table 2
2.
3.
1.
1.
4
1
9
5
3.
2.
2.
4.
3
8
7
3
5.
7.
3.
2.
07
23
29
56
80.
85.
78.
68.
6
2
9
0
2.
2.
-30.
0.
06
95
60
92
Table
Table
Table
Table
3
3
4
5
   Program B Totals
 35.5
52.0   51.49 899.9
10.47

-------
                                       27

                                    Table 16

                         Cost-Effectiveness  Comparison
                           Between Example Program B
                         and  a  Conventional I/M Program
                           With CO  and HC Cutpoints


Program                    HC  Cost-Effectiveness*     CO Cost-Effectiveness*

Conventional I/M                  $581/ton                   $51/ton
Example Program B                 $102/ton                   $ 6/ton
* Some areas which are implementing I/M  to  achieve HC reductions  are  doing so
only  for  that  reason,  i.e.,  they do  not need  CO  reductions.  In  such  cases,
since  the  costs are  equally distributed between  the two  pollutants in  this
table, the HC cost-effectiveness  number would-be doubled.


1.3.3 Example Program  C: Maximum  Benefit Program

As an exercise  to demonstrate the  maximum  potential benefits of an idle  CO
program and  as a  method  to show  how  to apply  each of  the  Tables  in  Section
1.2, Example Program  C will use all of the  emission  enhancements  discussed in
this report.                                                              -

Example Program C  has  the following  principle design features:   ' "  •    "•    -: -

(l) All model year vehicles 1968  and later are inspected.        ' •^•- • •- ~" :.- .;-.;.

(2)  The  Two-Speed I'dle  Test  or Loaded Test  for  1981 and  later  passenger
vehicles and 1985  and  later light-duty. trucks with a 1.2% CO cutpoint in  both
modes  increasing   the  identification  rate  of   high   emitters   among   these
vehicles.    Owners  of 1981  and  later vehicles   qualify— for  the  Emission
Performance Warranty [3] if other  conditions  are also met.       .   :..        .;
                                .   ' " .     • .   . •_.  '..-'.'•  •••--•• /.•  • • -••-''  3, : ,  ;' •„-••;
(3)  Idle  CO cutpoints which   fail  30%  of  pre-1981  vehicles  at  the  initial
inspection.[7]    Although  a  40%  stringency  program would   provide  greater
benefits,  30%  is probably  closer  to what most I/M programs consider desirable
as a failure rate.              .                :                ..  '. .••.".   .•-.

(4)  The  reinspection   cutpoint of  1.0%  CO  for  pre-1981  passenger  car  and
pre-1985 light-duty  trucks.   This will achieve  the  same effect as  a  mechanic
training program that  reaches every vehicle.

(5) An air pump check.             :. .

-------
                                      28

(6)  An evaporative cannister check.

(7)  A catalyst check for all inspected 1975 and later passenger cars.

(8)  An idle HC inspection cutpoint of 1000 ppm.

(9)  Inspection and maintenance for all light-duty trucks below 8500 Ibs.  GVW.

Example Program C achieves  a  45%  HC  benefit and a 58%  CO benefit, much  larger
than the conventional I/M  program with  a 20%  stringency.  This  is illustrated
in Table  17.   The cost-effectiveness of Example Program C  is  calculated  and
presented in Table 18.  These  cost-effectiveness  values are  slightly  more than
for  Example Program B, but are still well below the  conventional I/M program.

-------
                                       29

                                      Table 17

                    Comparison of Emission Reductions  and  Costs
                             of Example Program C and a
                       Conventional  20% Stringent I/M Program
Program
CONVENTIONAL I/M
PROGRAM:
(20% Stringency, idle
test for 1981 and later
vehicles)

EXAMPLE PROGRAM C:

0 Basic Idle CO I/M
0 More Effective Test
  for 1981 and later
 - vehicles
0 Increased Pre-1981
  Stringency
    13% to 20%
    20% to 30%
0 Tight idle CO
  reinspection cutpoint
0 Air pump check
0 Evaporative System
  check
0 Catalyst Check
0 Loose HC cutpoint
0 Light-Duty Truck I/M

Percent
Emission
Benefit on
Dec. 31, 1987
HC CO
Five Year
Emission
Reduction
(Thousands
of Tons)
HC CO


Five Year
Program Cost
(Millions of $)




Source
(Section 1




.2)

34.9
24.2
 2.4
33.1   46.50 526.8
               54.10
33.1   31.31 526.8
 5.8
2.03  60.4
               40.19
-5.05
              Table 1
              Table 1
Table 2
2.
3.
1.
1.
1.
0.
2.
5.
4
1
9
5
3
9 :
0
6
3
2
2
4
0
••'-..- 0
• o
.. 5
.3
.8
.7
.3
.0
.5
.0
.9
5.
7.
3.
2.
1.
1.
2.
8.
07
23
29
56
94
35
59
11
80.
85.
78.
68.
0.
10.
0.
111.
6
2
9
0
0
7
0
4
2.
2.
-30.
0.
0.
2.
0.
1.
06
95
60
92
25
00
0
25 .
Table
Table
Table
Table
Table
Table
Table
Table
3
3
4
5
5
5
6
7
  Program C Totals
45.3    58.4   65.48 1022.0    13.97

-------
                                      30

                                   Table 18

                         Cost-Effectiveness  Comparison
                           Between Example  Program C
                        and a Conventional I/M Program


Pro gram'                    HC Cost-Effectiveness*     CO Cost-Effectiveness*

Conventional I/M                  $581/ton                   $51/ton
Example Program C                 $107/ton                   $ 7/ton
* Some areas which are  implementing  I/M to achieve HC  reductions  are  doing so
only  for  that  reason,  i.e.,  they  do not  need  CO reductions.  In  such cases,
since the  costs are  equally distributed  between the  two  pollutants  in  this
table, the HC cost-effectiveness number would be doubled.

-------
                                      31

2.0 BACKGROUND ON CONVENTIONAL I/M PROGRAMS

Since  1972 when  New Jersey  added  emission analyzers  to  its safety  inspection
stations,  I/M has  existed  as  a feasable,  effective  means to  reduce  in-use
emissions  from automobiles.   Congress  added  T/M  to  the  requirements  of the
Clean  Air  Act Amendments  in 1977  as a  reasonable and  required  air  pollution
control strategy for  those urban areas with  serious  air quality problems.  In
all,  30  states  are expected  to implement  some  type  of  I/M  by  1983.  Some
states, such  as  Oregon,  Arizona, and Rhode  Island have already begun emission
inspection programs.   These current  programs, while  differing from state  to
state, do  have many features in common.   This section will describe what can
be  called  a  "conventional"  I/M  program which   best  describes  the  type  of
program now being used  in  those  states  with  operating I/M  programs.   Following
sections will describe  low-cost  approaches   to  I/M  primarily by  noting the
aspects in which they differ from conventional I/M programs.

2.1 Inspection and Reinspection

All currently operating programs[8], both centralized and decentralized, use
an emission analyzer  to  measure  the  concentrations at  idle and/or  at a cruise
mode  (usually only  at   idle)  of hydrocarbons  (HC) and  carbon monoxide (CO)
coming  from  the   tailpipe(s)   of   vehicles   subject   to   inspection.   These
measurements  are compared  to standard  levels (cutpoints)  for  each  pollutant
established  for  each  vehicle  age  and   type  by   the  state or  local  agency
responsible for  the program.  If  the measurement  of  either pollutant  exceeds
its cutpoint,  then  that  vehicle  fails the test.   The percent of vehicles which
fail the emission inspection is  refered  to as the failure  rate.   This  failure
rate is typically 20  to  30 percent.   The state or local  agency can adjust the
program  failure   rate  by   adjusting'  the cutpoints   to  fail  more  or  fewer .
vehicles.  Vehicles which  fail  the  emission inspection  must   receive  enough
repairs  to reduce  their  emissions   below  the program cutpoints  and  pass  a
reinspection.    The  reinspection  cutpoints  are   identical  to  the  original
inspection  standards.    Most  failed  vehicles pass tbe  reinspection test  on
their first try.       •                                  :        .
                             : •. '   '       • ' '        '.   •' .     •   ''.      " '•'     ••/•*
There  are  several  ways  to provide  for  inspection facilities.  Decentralized
programs utilize existing  repair   facilities which  are  licensed   to perform
emission inspections  and reinspections.    Centralized  programs  involve  use  of
facilities dedicated  to high-volume, emission  testing with  repairs performed
elsewhere.   These  centralized  facilities  can either  be  operated  directly  by
the  state,  by a single contractor,  or  by  multiple  private  companies under
licenses from the state.   States should  not  feel  constrained to use only these •
options.   New Jersey  combines  advantages  of centralized and  decentralized
approaches  in their  I/M  program.   All  vehicles  are  inspected initially  at
state-run  centralized  inspection facilities.  Those  vehicles  which  fail have
the option of either  returning  to  the centralized facility for a- reinspection
or  having  the  reinspection performed  at  a   licensed  repair  facility.   This
allows  the  reinspection to  be   performed  where  the  repairs are performed  in
some  cases eliminating  a   trip  back  to  the  centralized  inspection . facility
without loading the repair service industry with the initial inspection task.

-------
                                      32

The inspection requirement  is  enforced  either  by withholding registration from
vehicles which have  not  passed their inspection or  reinspection  or by issuing
special stickers  to  cars when they  pass  the inspection  and  issuing citations
to cars with missing or expired stickers.

2.2 Types of Repairs Performed

In the majority of cases  it has  been found  that a simple carburetor adjustment
on  pre-1981  vehicles can reduce the  idle  emissions  of HC  and  CO  from most
vehicles  sufficiently to  pass the  reinspection cutpoints  used  in operating
programs.  When a  carburetor  adjustment alone  is not enough  to  reduce both CO
and HC idle emissions to  pass  the program cutpoints then in many cases other
simple  repairs  like  new  spark plugs or a  new  air filter element  can provide
the additional  nece.ssary reductions.   Other  repairs  may also be  performed in
the same visit to  the repair  facility,  at  either the owner's or the mechanic's
initiative.  Table 19 lists the  frequency  of various types  of  repairs done on
pre-1981  vehicles  as  observed  in  the  Portland  I/M  program.   Carburetor
adjustments are by far the most common  repair.


                                   Table 19

                 Typical Repairs  Performed  on Vehicles Failing
                       the Oregon State Inspection Test


                                            Model Year          Model Year
                                         1972-74 Vehicles     1975-7? Vehicles

      Total Number of Cars                       95                • 252

Were the  following items  repaired, replaced, or adjusted?  (yes)

Spark Plugs                                    35.8%               25.4%
Spark Plug Wires                               14.7%                11.9%
Points and Condenser                           31.6%                5.6%
Distributor Cap and Rotor                      17.9%                6.3%
Spark Timing Control Devices                   13.7%                10.3%
Carburetor (Idle Mixture  Adjustment)           87.4%               89.3%
Choke                                          48.4%   '            21.8%
Intake System                         .          5.3%                4.0%
Air Filter                                     31.6%               31.0%
Engine Oil                                     21.1%                15.1%
Idle Speed '                                    48.4%               73.4%
Timing                                         -'-9.5%               '^.02
Dwell                                          41.1%               26.6%
Air Injection System                            5.3%                3.2%
EGR System                                      6.3%                4.8%
PCV System                                      6.3%       "         7.5%
Valves                                          7.4%                0.8%

-------
                                      33

2.3 Cost of Repairs

Average repair  costs  on pre-1981 vehicles  h'ave  be.in reported to  be  as low as
$18.71 per failed  vehicle  in the New Jersey program and  as  high as $35.00 per
failed vehicle  in  the  Los  Angeles program.[8]  An analysis  of repair costs in
1979 in the  Portland,  Oregon I/M program[9]  showed  an average  repair  cost of
$22.00 with half of the  repairs  costing  less  than $14.00.  (Over 44 percent of
the collective  repair  bill was due  to  the 13 percent of  the  failed  cars that
had repair  costs  more  than  $50.00.)  These  low  average  and  median  costs are
significant since  the  Portland program has the most  stringent cutpoints of all
operating programs.

The low average cost of  typical  I/M  repairs on pre-1981  model year vehicles is
a  reflection  of the  simple  adjustments  necessary  in most  cases  to  reduce  a
vehicle's  emissions sufficiently to pass the reinspection test.

-------
                                      34

3.0 REDUCING REPAIR COSTS BY REDUCING INAPPROPRIATE AND UNNECESSARY REPAIRS

Most persons  in the service industry are  sincere  in their efforts  to provide
effective, low-cost  I/M  repairs  'to  their  customers.   However,  the  service
industry  is  not  perfect.    Problems  can  occur which  cause  repair   costs  to
owners to  be  higher than they would  need  to be.  Some problems  result  from a
lack  of   knowledge  in  the  repair  industry  of  proper  diagnosis  and  repair
procedures  for  emission  related  components  and  systems.   Misdiagnosis  and
ignorance  of  proper repair  procedures  can  lead to  (1)  replacement  of  parts
which have  not  failed,  (2) unnecessary adjustments,  (3) multiple repair trips
to  finally correct  the  emission  problem,   (4)  continued   non-compliance  with
emission  inspection  standards  possibly  requiring a waiverj  and (5)  encouraging
the  vehicle owner to attempt  to acquire  a  certificate  of  compliance without
completing  necessary emission  repairs  (cheating).   In addition,  there may also
be  some  cases where (1) a  mechanic will  knowingly  perform more  repairs  than
necessary   to  pass  the  emission  inspection,   some  of  which  would  not  be
considered  necessary or useful by any standard, (2)  the vehicle  owner will be
persuaded  to  purchase  services  the  owner  would  not  have   sought  otherwise
(i.e., brake  job,  lube,  wiper blades,  tires,  etc.),  (3) higher  rates will be
charged  for the repairs than  can be justified  by  the parts  and  labor costs,
and  (4)  when  the emission  inspection is performed in the  service facility, as
in  decentralized programs,  the vehicle may  be  deliberately failed  incorrectly
to  generate repair business.

These potential  problems  increase the repair cost'vehicle  owners must pay and
reduce  the cost-effectiveness of  the  I/M  program.   It  is  not possible  to
determine exactly  how frequently these  problems may  occur,  but it is  necessary
to  consider the  possibility of these problems  in the overall  design  of an I/M
program.   If  these  problems  can be  made  less  frequent,  the  cost  of the I/M
program  can be  reduced.

3.1  Improvement  1:  Mechanic Training

The  most straightforward  way  to deal  with  problems associated  with emission
repairs  is to  assure   that  mechanics  are  able  to   diagnose  the problem and
perform  the  repairs  correctly,   and  that  they  are   aware  of  the  benefits  to
their customers  and  their  business  of correct  diagnosis  and repair of emission
related  problems.   With  this  information  conscientious  mechanics  will  be
better  able  to  avoid  any  unnecessary or  inappropriate  repair work  during
emission related repairs.

3.1.1 Mechanic Training Courses

A short,  practical course  has  been  designed  by the  National Center  for Vehicle
Emissions- Control  and Safety of  Colorado  State University  (CSU)  which provides
mechanics  with  the  basics  of diagnosis  and repair  for  vehicles  failing I/M
test cutpoints.  The course  was  designed to  be  a maximum of 16 hours  in length
and  could,  for  example, be  taught over  several  evening  sessions.   A  program to
provide  this  course  or a  similar course  to  all mechanics  who will perform
emission repairs will improve  the knowledge  of the  service industry  in vehicle
emission control.   In  addition,  EPA  is  convinced   that  mandatory  mechanic

-------
                                      35

training can result  in  increased, emissions reductions and fuel savings from an
I/M program.

3.1.2 Information Distribution

Another,  less  ambitious, method  to inform  mechanics of  proper  I/M diagnosis
and  repair  procedures  is  to include  advice to  the  mechanic directly  on the
inspection  form.   The  advice  would take  the  form  of a  list of  the  repairs
which  are most  likely  needed  based  on  the  type  of  I/M  failure (HC-only,
CO-only, or both).   Table  20 is an example of  the type of advice which can be
included  on  the  inspection  forms.   (A portion of the advice in Table  20 is
specific  to  pre-1981 vehicles, but  corresponding advice  can  be  developed for
1981 and later vehicles.)  In  this  way  the mechanics  who will actually perform
repairs will have  diagnosis and  repair information  available to them  at the
time of repair.  This information can  be  supplemented by a full-scale training
program or by distribution of more detailed I/M diagnosis  and  repair manuals.

-------
                                   Table 20

                 Diagnostic  Advice  for  Use  on Inspection  Forms

For each emission problem condition, perform the  checks  and  maintenance in the
sequence  as  listed until  the  idle  measurements   for  both  HC  and  CO  are
sufficiently   below  the   idle   emission    standards  to   assure   passing   a
reinspection.

Condition:   Fail Both Idle HC and Idle CO.

Diagnosis:   Begin  with  the  maintenance  described  to  correct   the   idle  HC
             problem.   When  the  idle HC  measurement  is  acceptable,  begin
             performing  maintenance  described to correct  idle CO  problems  as
             necessary.

Condition:   Fail Idle HC.

Diagnosis:   1.   Improper   Ignition   Timing:    Compare   with   manufacturers
             specification.

             2.  Faulty Ignition or Misfire:  Check  for arcing  or disconnected
             wiring,  fouled  or damaged  spark  plugs and   insufficient  spark
             voltage.

             3.  Vacuum  Leaks:   Check  for  damaged,   missing,  or  disconnected
             hoses  and  check  for leaks around the intake manifold  and base on
             the carburetor.

             4.  EGR System Incorrectly Operating  at Idle:  Check  for proper
             valve  operation.

             5.   Idle   Speed   Set  Too   Low:   Compare  with   manufacturers
             specification.

             6.   Air/Fuel  Mixture   Set   Too   Lean:  Check   or   set  using
             manufacturers  specified procedures.  Check  the  balance  in two and
             four barrel  carburetors.

             7.  Worn Piston Rings or Valves:  Perform  compression check.

Condition:   Fail Idle  CO.   .             -            ...

Diagnosis:   1.  Dirty Air Filter: Clean or replace if necessary.

             2.  Choke  Stuck: Check for proper operation and adjustment.

             3.  PCV Svs tern'PI ugged: <"heck  valve and hose.s for restriction.

             4.  Air Pump  and  Control  Valve Inoperative:  Check  belt, bypass
             valve, hoses,  etc.
             5.   Idle   Speed   Set   Too  Low:
             specification.
Compare   with   manufacturers
              6.   Air/Fuel  Mixture   Set  Too   Rich:
              manufacturers specified procedures.
      Check
set  using

-------
                                       37

3.1.3 Problem Facility Identification

If  those  mechanics who  are  most  in need  of  emission repair  training  can be
identified,  then  the  training effort  can be more  effectively  focused on  this
smaller number of  individuals.   These  repair facilities can be identified by a
survey of vehicle  owner  complaints, average repair  costs,  or waiver rates for
service  facilities   performing   emission  repairs.    Those  facilities  which
deviate the  most  from the norm  could  be  required or  encouraged  to send their
mechanics for training.  This approach could be used as a method to eventually
train all the service industry,  starting  with  those most in need of  training.

3.2 Improvement 2:  Public Awareness

One  simple  way  to protect  vehicle owners  from inappropriate  or unnecessary
repairs is a strong public  awareness program.   This effort  should  be aimed at
telling the  vehicle  owner  what  repairs  are  appropriate  for  his/her vehicle
given it has failed  the  I/M test  in a certain way  (HC, CO,  or both)  and what
charges are reasonable for these repairs.  This information can be included on
the  inspection  form.  Armed with  this  information,  vehicle owners  would be
better prepared to enquire  about and understand  the  type  of repairs mechanics
propose to make and to know better  to  refuse any unnecessary repairs.

3.3 Improvement 3;  Price Competition

The  repair   industry  may  take  advantage  of   the  lack of   information  about
appropriate  repairs   and their  costs   in  the  beginning  of a mandatory  I/M
program to sell repairs at inflated  prices.

An  effort  could  be made as part  of  the  I/M   program  to  encourage comparison
shopping  by  vehicle  owners  in  need of  emission repairs.   Repair facilities
could  be  encouraged  to  advertise  their  fees for  various  emission  related
repairs  and  give  accurate  estimates   of  possible  repair   costs.    This
information  could be  made  mandatory.   The  program  could  also  publish  the
average repair costs  from each facility at  frequent intervals to promote price
competition by repair facilities and comparison shopping by vehicle owners.

3.4 Improvement 4:  Automated Analyzers                '".-.   . :•.'-;»  S'l:'L:.4.-".^ : . '  -'

In  decentralized  programs,  the  possibility  of vehicles_being  deliberately
failed incorrectly to increase a service  facility's  repair business is a worry
to some vehicle owners.  Even if the instances  of such dishonest}' are  few,  the
bad  publicity  and distrust  such  instances wi 1.1  generate  should  be  avoided.
All  decentralized programs  should  have  some  method   to  respond  to  consumer
complaints about  specific  instances of fraud and should also conduct  auditing
and  surveillance  programs  to detect   other  instances.   However, "any  method
which can make fraud more difficult  from  the start is also desirable.

Computerized  emission analyzersflO]   offer  a  new  tool   for  a  decentralized
program to  reduce errors  in the  passing and  failing of vehicles  in an  I/M
program.  These sophisticated analyzers are less  prone to  improper use.   These
analyzers can make it nearly  impossible for a  service  facility  to deliberately
fail a  vehicle  incorrectly  if the vehicle owner at  least makes .certain  that

-------
                                      38

his own vehicle  is being tested by  the  analyzer.   These instruments  can also
reduce  administrative  costs  by automating  data collection  and reducing  the
amount  of  program auditing  required.   With  an  automated  analyzer,  the  audit
period  can  be   lengthened  from one to  three  months,  with  a  corresponding
reduction to the number of auditors which are required!11].

EPA  has  been  impressed  by  the  performance • and  possibilities  inherent  in
computerized emission analyzers  for  decentralized I/M  programs,  and recommends
their  use  over  the  traditional garage  analyzers.    In  addition,  to make  it
feasible for I/M programs to  acquire these  analyzers,  additional lead time for
implementation  of  a  decentralized  I/M  program  using  them may  be  granted  if
requiredfl1].

3.5 Improvement 5;  Challenge Stations

"Challenge   Stations"   are   state-operated   or  closely   supervised  emission
inspection  stations  which  can  be  used  by  vehicle  owners  to  verify  their
inspection  results received  at private  garages in  a decentralized  program.
These  challenge  stations  will allow vehicle  .owners  who  are suspicious  of the
results they have  received from their  emission inspection at  private garages
to be  retested by  the state before they invest in emission repairs.

Challenge stations increase confidence  in the fairness of the  I/M  program and
provide a means  to identify inspection stations which are  incorrectly failing
vehicles through  incorrect  testing procedures, poor equipment  maintenance,  or
deliberately.   By reducing  the number  of   incorrect failures,  total  repair
costs  to  owners  will  be  reduced.   These  stations  also provide  a  means  for
distribution  of public  awareness   material   directly  to vehicle  owners  'most
concerned about  the program.

Because of the  importance of  the challenge  function, EPA  requires  some form of
it in  any decentralized  I/M program.  In  this sense,  challenge  stations  are an
essential part  of  a  conventional I/M program  and not  an  optional  improvement.
However, an  I/M program does have  the option of placing far more  emphasis  on
the challenge function than required by EPA.

3.6 Impacts  on Costs and Emission Reductions

None  of  the improvements  outlined .in  this  section will  decrease  emission
reductions  from  the  I/M  program.   Indeed,  mechanic  training  programs  can
increase emissions reductions.   They will reduce the  costs  of  the  program  to
the motoring public,  and  thus   improve  the  overall  cost-effectiveness  of  the
program.   The  size  of  the  cost   savings  cannot  be  quantified,   hut  it  is
reasonable  to assume  it will  outweigh  the  administrative  costs  to the  I/M
program of implementing the improvements.

3.7 Conclusions

All of  the improvements outlined in  this section can be applied  in  one form or
another to any  I/M program  — including any  "conventional" I/M  program  — and
should  be considered whether  the program  is  being designed,  implemented,  or is
already operating.

-------
                                       39

4.0 I/M FOR PRE-1981 VEHICLES  USING ONLY IDLE CO OUTPOINTS

4.1 Description and Summary  of Impacts

The  improvements  described  in  Section  3.0  attempt  to  reduce  I/M  costs by
reducing  the  number  of  inappropriate,  unnecessary,  and ineffective  repairs.
Such  repairs  constitute  "fat" in  a conventional  I/M  program and  can be cut
without reducing  program effectiveness.   Further reduction in I/M costs can be
achieved  only   by eliminating  some  properly  performed  repairs   that  are
appropriate in a  conventional  I/M program and which do  produce useful emission
reductions;  this  is   cutting  "lean"  from  the  I/M  program.    Clearly,  the
objective of such cuts  should  be to eliminate  or reduce types of repairs  that
are  costly  in   relation  to   the  emission  reductions  they   produce   (less
cost-effective)  and to maintain  or even  increase  repairs which  are cheap in
relation to their  emission reductions  (more cost-effective).

The number  and  types  of repairs which  are  appropriate  in  a  conventional I/M
program  are  determined  by  the  design  of  the  program,  specifically  by the
inspection and  reinspection test procedure  and  cutpoints.   Consequently, the
only  approach  which  might  have  the  effect of  reducing  the number  of   less
cost-effective  repairs   is  to  modify the  test  procedures  or  the   cutpoints.
This  section  examines one  such  modification:  simply eliminating the idle HC
cutpoint and retaining  only the  idle  CO cutpoint,  with  no  change  in the  idle
test procedure itself.   An I/M program with  this modification will  be  referred
to as an  idle  CO I/M  program.   The goal of using  only  an idle  CO cutpoint is
to  virtually  eliminate  the  occurrence  of all  repair  types  other  than   idle
mixture and speed adjustments,  because  idle mixture  adjustments  appear to be
the most  cost-effective  type  of  repair  in a   conventional  I/M program for
pre-1981 vehicles.                                                          .

Idle mixture and  speed  adjustments can  be performed  cheaply,  because there is
no  need  for  diagnosis  beyond  the  results   of  the  idle test  itself  and no
replacement parts  are needed.   This is in contrast to most  other types of I/M
repairs  (excluding perhaps  spark  timing  adjustments),  particularly ignition
system repairs,  in which time .consuming diagnosis  is needed to  pinpoint the
exact problem and  the fix involves  installing a replacement- part such as a set
of spark plugs or wires, a distributor cap, or-a  new vacuum hose.—"'-•"•; -;--—-:~-~~
In addition  to  being relatively inexpensive,  idle  mixture  adjustments produce
very  sizable emission  reductions  for  both  HC  and  CO on catalyst-equipped
vehicles.   On   such  vehicles,  a  rich  idle mixture  maladjustment  causes  the
engine to produce higher concentrations  of  HC  and CO.  .It also depletes oxygen
from the engine's exhaust,  so the  catalyst  is less able  to  convert  the HC.and
CO  to  harmless  C07  and water  vapor.   Correcting  the  idle  mixture " therefore
has a two-fold effect in reducing HC and CO emitted at the tailpipe.*'    -
* On  cars without  catalysts,  i.e. ,. those  sold  before  1975,  an  idle  mixture
  adjustment also has a large effect on CO but its effect on HC'is much less.
                       . •   ' . . -( ..•-;...>•'-   •-'. .   -;'••••  ......... TO---   •••# »-.. — «  i - •" - •_, .-i
re

-------
                                      40

Table 21  shows  a number  of  examples  illustrating  how effective  idle  mixture
adjustments  are  for  catalyst  vehicles  which  have  failed  idle  HC and/or  CO
cutpoir.ts.   The  examples  differ  from one  another with  respect  to  how  the
samples were  selected  and the procedure  used  in the  idle  mixture adjustment;
the  specific  details  for  each  example  are  given  in  footnotes  to  the  Table.
All of the examples  support  the  general effectiveness of this type of repair.
Reductions in FTP  HC  range  from 34.1% to 55.6%.   Reductions  in FTP  CO  range
from  51.8% to  71.5%.    For  comparison,  Table   21  also  presents   the  results
achieved  by  the  typical  mix of  "conventional"  I/M repairs performed  by  field
mechanics in  the Portland I/M program.

Eliminating  repairs  other than  idle  mixture adjustments  will decrease  the  HC
emission  reductions  of an idle CO I/M program  compared to a  conventional  I/M
program.   Not  all  of   the  HC  reductions are  lost,   since   idle  mixture
adjustments account  for  some of the  HC emission reductions in  a conventional
I/M program.

The fuel economy aspects of an idle  CO I/M program  will  also  differ from those
of  a conventional  I/M program,  because of  the differences   in the  types  of
repairs  performed.   Proper  correction of a rich  idle mixture  maladjustment
should always improve  fuel economy.    If  there  are no other  repairs performed
which can decrease fuel economy, a net fuel  savings should  result from the  I/M
program.

The  following sections examine  in more  detail  the  issues of  repair  types  and
costs, emission  reductions,  and fuel  economy improvements in an  idle  CO  I/M
program.  The conclusions from these sections can be summarized as  follows:

Repair Types  and  Costs - Virtually  all  cars which  fail inspection  in  an idle
CO  I/M  program  will  be  capable  of  passing  the   reinspection after  having
received only an idle  mixture adjustment.  The  average repair cost is expected
to  be $6  to  $10,  compared  to   the  range of $18.71  to  $35  in  the  currently
operating  conventional I/M programs.   There  will  also be  a   reduction  in  the
failure  rate, since  cars  will  no longer fail for high  HC emissions.   Together
these  two changes will  have a  large impact on the overall   cost  of  the  I/M
program.   The total repair  cost for  pre-1981  cars will  be  reduced  by  about
70%,  which will  in  turn  reduce  the  overall  cost  of a typical  I/M program  by
19% during the period  from 1983  through  1987.

-------
                                       41
                                      Table 21

                                Emission  Reductions
                            From Carburetor  Adjustments
                           (1975-1980 Model Year  Vehicles)
Program

FY77
Emission Facto'r*
Program

Portland Study2
Carburetor Repairs
Only

Houston Program-^
Study

Houston Program^
Study

Restorative
Maintenance^-

FY79 EF
Program^-

Portland Task
Group
Portland Study&

Vehicles
Receiving
Adjustments
(54)
(76)


(97)

(97)

(95)

(43)

(31)




FTP
Pollutant
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

., ~ • • •
Before
Carburetor
Adjustment
(g/mi)
2.75
42.2
2.50
39.4
3.54
59.7
3.54
59.7
2.13
41.1
1.93
28.4
3.26
. 47.5 -•
.* 	 •. "... .
Conventional
After
Carburetor
Adjustment
(g/mi)
1.43
14.6
1.65
19.0
1.57
20.9
1.87
26.5
1.09
11.7
1.27
13.7
1.90
. 20.6 •': /•
V '. : . . - " . . ...


Percent
Change
48.0
65.4
34.1
51.8
55.6
65.0
47.2
55.6
48.8
71.5
34.2
51.8
41.8
:56.7 ..
•T d
I/M (for comparison)
                        (310)
CQ
 2.82   :    	1.56  ,  . ;44.7
39.63     --  .19.72 -. ..- :••'  50.2
1:  Carburetor  adjustment performed  by  contractor  mechanics  to manufacturer
    specifications.  Idle HC cutpoint:  225 ppm;  Idle CO cutpoint:  1.0%       .   '
?.:  rarburetor  repairs  performed by field mechanics.   Vehicles failing Oregon
    DEQ outpoints.                        •                .     •
3:  Carburetor  adjustment performed  by  contractor  mechanics  to manufacturer
    specifications.  Vehicles failing Houston Program cutpoints.
4:  Idle mixture  adjusted so that  idle CO  is  0.5%.   Vehicles failing Houston
    Program cutpoints.
5:  Idle mixture  adjusted so that  idle  CO  is  0.2%.  Vehicles failing Oregon
    DEQ cutpoints.
6:  Typical mix of conventional I/M program  repairs.   Vehicles failing Oregon
    DEQ I/M inspection  presented here for comparison.           ~,--....,:,-<-.    ,

-------
                                      42

Emission Reductions -  The  CO emission reductions from  pre-1981  vehicles in an
idle CO  I/M program will  be virtually the  same  as those  from  a conventional
I/M program which uses  the  same  idle  CO  cutpoints.   The HC emission reductions
will be  significantly  less, but still 75%  or  more  of what they  would  be in a
conventional I/M  program.   The  reduced HC  benefits  are due to  the  absence of
ignition system and other HC-oriented repairs in the idle CO I/M program.

Fuel Economy Improvements  - A  fuel  economy  improvement of  about  4 percent can
be expected  from  an idle CO I/M  program,  if.appropriate mechanic training is
conducted or if  the I/M program has  soirie  other mechanism  to  ensure  that idle
mixture  adjustments are performed approximately correctly.  There has  been no
conclusive  evidence that  there  is  any  fuel  economy benefit associated with
conventional I/M  programs.   This  is  primarily  due to  repairs  performed  in
conventional  I/M  programs  which  degrade  fuel  economy  offsetting  the  fuel
economy  benefits associated with carburetor  adjustments alone.

4.2 Repair Types and Costs

4.2.1 Elimination of Idle HC-Only Failures and Repairs

The  most obvious  effect  of eliminating  the   idle  HC  cutpoint   for  pre-1981
vehicles from a conventional I/M program is that cars  which  would have  failed
the  inspection  only for high  idle  HC emissions will  now  pass  inspection and
will not be repaired.   Consequently,  these cars'  contribution  to the  average
repair  cost  is eliminated.   To quantify  this  effect,  the number .of  HC-only
failures and their  average repair cost in  a conventional  I/M program  must be
known.

In  its   study  of  the  Portland  I/M  program,  EPA  found  that among  1975-1977
vehicles, 8 percent of'cars failing  the  idle test in 1978  failed  for  HC only.
This number  should  not be  considered typical  of  most other  I/M  programs, for
two  reasons.   First,  the   Portland  program has  an  unusually low (stringent)
idle CO  cutpoint, so many  cars  which  would  be  HC-only failures in other states
are  HC-and-CO  failures  in  Portland.   Second,  the  cars  in EPA's sample were
relatively  new when   they  began  I/M.   The  I/M  programs which  will  begin
operation in 1982  or  1983  will  find  that  malperformances which  cause  high HC
will  be more  common  due  to  the  greater  average   age  of  catalyst-equipped
pre-1981 vehicles.  Consequently, if  they  use an HC  cutpoint, the fraction of
all  failing  cars  which  fail  only for HC will  be  larger.  EPA  estimates that
for  conventional  I/M  programs  starting  in  1982  or  1983  the HC-only  failure
rate will be about 25%  of  the overall failure rate.

The  impact  on  average  repair  cost  of  eliminating  HC-only  failures  will  be
greater  than suggested  by  their  number.  THis  is  because  HC-only failures have
a  higher  average  repair   cost  than  the   remaining  failed  cars.   In  EPA's
Portland sample,  the  1975-1977  HC-only  failures had an average  repair  cost of
$31.05.  The average   for  the  other  failed  cars was  $23.86.   This pattern  of
costs makes  sense,  since  most  of  the  cars that  fail only  the HC  cutpoint
require  some  sort  of  ignition  system  repair.  This  involves  time  consuming
diagnosis  to pinpoint  the  specific  parts  in  need  of  replacement,  or  else
replacement  of  more  parts  than really  necessary.    Ignition system  repairs
should  be   less  frequent  among  the  other  failed   cars,   since   idle  mixture

-------
                                      A3

maladjustment  is  the only cause of  idle  test failure for  many of them.  This
explanation  is  supported  by  type-of-repair data  from  the  Portland study.
Spark  plugs  and  spark plug  wires were  replaced more  o'ften  (about  twice as
often) on the HC-only  failures than on other  failed  cars.

The figures  given above for the number  of HC-only  failures  (about  25% of all
failures)  and  for  their  relative  repair cost   ($31  versus  $24)  mean  that
eliminating  HC-only  failures  should  reduce  the  total  I/M  repair  bill  for
pre-1981 vehicles  by about 30%.

4.2.2 Simpler Repairs  for Other Failed Vehicles

Because  the  vehicles that do fail the  idle test  in an  idle  CO I/M program do
not have to  pass an  HC  cutpoint on reinspection,  in principle  they should
require simpler repairs than  in a conventional  I/M  program.   The only repairs
that  are  possibly needed  are  those  which reduce idle CO  emissions; repairs
which  would  reduce  idle  HC   emissions   are  not  necessary.    In  fact,  for
virtually all  failed cars the only  repair that  is needed  is  an idle mixture,
and possibly  idle speed,  adjustment.  Argument  and  evidence  in favor of this
contention are given in the following paragraphs.

It  is  well  known  that the engine parameter  which  has  the  largest  impact on
idle CO is the idle mixture setting,  since this  is the primary control for the
idle air/fuel ratio.   Most  catalyst-equipped  vehicles  with  a properly adjusted
idle mixture  have an  idle  CO  level  near  zero.   A rich maladjustment causes
idle  CO  to  increase  far  enough  for  the vehicle  to  fail  a  typical  idle  CO
cutpoint.  Even partial correction of  the maladjustment  will reduce idle CO to
below  che cutpoint.   There are  some  other malperformances  which can increase
idle  CO,  such  as  blockages  and  leaks  inside   the  carburetor,  defective
carburetor  floats,  severely   plugged  air filters,   crankcase  oil  which  is
heavily  contaminated  with  gasoline,  and  disabled  air  pumps.   Surveillance
studies  have  repeatedly  found  that  such  problems are infrequent  compared  to
idle mixture  maladjustments [12,13,14].   Consequently,  they  should  be needed
only rarely in either a conventional  or idle CO I/M program.

Additional  evidence  to support  the   claim that  idle  mixture  adjustments  are
almost  always  sufficient  to  allow   vehicles which  have  failed an   idle  CO
cutpoint to pass  on  reinspection comes  from several  EPA test programs.*  These
programs and  their findings are summarized in  Table 22.   All  of the programs
showed  very  high pass  rates   on  the  idle  CO   test  after  an idle  mixture
adjustment,  in the range of 96.4 to 99.0 percent.       .  .  '    .•"*'"•...';
* With few exceptions,  the  report  will  treat  1973-1980 vehicles as though they
are typical of all vehicles made before  the  1981  model year.   In fact they are
not;  there  are  important  differences  between  these  and  older  vehicles.
However by  1987 —  the date  EPA  uses  to  evaluate  the effectiveness  of I/M
programs —-  1975-1980 vehicles will  acount  for  83.7% of  the  inspections and
83.4% of the emissions of all  vehicles made  before  the 1981 model year.  Hence
their behavior in an  I/M  program will far. outshadow  the remaining older cars.
In  addition,  comparatively  little data  is  available  on  pre-1975  ve'hicles,  so
the focus on  1975-1980  vehicles  is unavoidable.   Where essential,.distinctions
will be drawn based on  the  best available information.           : . •  •   :   ;

-------
                                      44
Program

Restorative
Maintenance Evaluation

FY79 Emission
Factor Program 1

Houston Program
Study !

Houston Program
Study 2

Portland Task
Group 10 3
         Table 22

     Idle CO Pass Rate
After Carburetor Adjustment

                             Idle CO Pass Rate
              Initial      at Reinspection After
   Initial    Idle CO         Only Carburetor
   Sample     ..Failure    Adjustment Among Vehicles
     Size       Rate        Failing Initial Test

      145       50.3                97.3
       64      43.8                 96.4


     480      21.7                 99.0


     480      21.7                 96.5


     102      38.2                 97.1
1:  Carburetor  adjustment  performed  by  contractor  mechanics  to  manufacturer
    specifications.
2:  Idle mixture adjusted so that idle CO is 0.5%.
3:  Idle mixture adjusted so that idle CO is 0.2%.
As Section  3.0 suggested, it  is  not necessarily enough  to arrange  things  so
that  only  simple  I/M  repairs  are  needed.    The  lack  of  understanding  by
mechanics and owners and  the economic  interest  of  repair  facilities  can result
in unnecessary  repairs  being  performed.  This  is  illustrated  by the  types  of
repairs  which  were performed  by  commercial repair  facilities  in Portland  on
cars that had  failed only the  CO  cutpoint.   Virtually all of these  cars  could
have passed with  only ar  idle  mixture  adjustment.   Table 22 shows  the  repairs
that were  actually performed.   The  average  repair cost ' for  the vehicles  was
S19.50.  This is below the average for  the cars  that  failed the  ^C  cutpoint  or
both cutpoints,  but  it  still   reflects  charges  for  maintenance  in addition  to
idle mixture adjustments.

This additional maintenance will  be easier  to eliminate  in  an  idle  CO  I/M
program  than  in a  conventional  I/M program.   In  a  conventional  I/M  program
there  are  enough  failed  cars  which actually  do need types of  repairs  other
than an idle mixture adjustment that many owners are  not  surprised and  have  no
recourse when told their vehicles is one of them.  Some mechanics may not have
learned or may  not bother to   determine which vehicles do or  do  not.  need more

-------
                                       45

maintenance.  However,  in an idle CO program,  owners and mechanics can be told
that in virtually all cases only idle mixture  adjustments are required to pass
reinspection.   Owners  then have a  much  simpler  job  in protecting  themselves
from  repair facilities  that  recommend  more maintenance.   The  number of cars
that require  more maintenance  will  be so  small  that owners  can be advised by
the  I/M  program  to  always  seek a  second opinion  before  agreeing  to it.  In
addition,  price  competition among  repair  facilities should  be  keener  in an
idle CO  I/M program  because repairs are  more  uniform.   The  price competition
will  be  additional  pressure on repair facilities  to refrain  from  performing
unnecessary maintenance.
                                    Table 23

                                Portland Program
          Typical  Repairs  Performed on Vehicles Failing Only Idle CO
                      in the Oregon  State  Inspection Test

                                       Model Year              Model Year
                                    1972-74 Vehicles        1975-77 Vehicles

Total Number of Cars      •                80                        134

Were the following items repaired,  replaced or adjusted? (yes)

Spark Plugs                              28.8%                     14.9%
Spark Plug Wires                         11.3%                      4.5%
Points and Condenser                     23.8%                      5.2%
Distributor Cap and Rotor                16.3%                      3.7%
Spark Timing Control  Devices             11.3%                      6.0%
Carburetor                               93.8%                     92.5%
Choke                                    40.0%               -  •    11.9%
Intake System                             2.5%               .       3.0%
Air Filter                               26.3%              .       24.6%
Engine Oil            .     '        '• '     16.3%           -.'..--..•=• 16.4%
Idle Speed     '- -     -             T'    56.3% '   -   -  ;  • : •- -••;'•""''V;: :-73.9%  :.
Timing               ...:....-...-•.••..-.     •42.5% -  '   	v-;''"=-^ •"•>-*->-34.3%   -•
Dwell                   .-.-••-•       ;j  :36.3%    :  '"•    . r-: ':.':  <-:;' 20.1%
Air Injection  System              •?:'"••"••."•• 3.8% •"•  "  • -.•'• -:--^"-'' :-'=r •• 3.0% .:  :
EGR System                      " . "*•      3.8% -.'     .'         : ,:-: 3.7%
PCV System                       '       '6.3%      '     '..    -'••'.•'•• 4.5%
Valves                            .-       5.0%     '      - -.-: -'•• '   '•; :-0.0%
In  conclusion,  an idle  CO  I/M program  which  also  adopts  the  improvements
discussed in Section  3.0  will very likely succeed  in reducing repair costs for
pre-1981  vehicles  to  the amount  which  is  legitimate  for  an  idle  mixture
adjustment  only.   Data from  two sources  indicate  this cost  should be  in the
range  of  $6 to  $10  per  car.  -In  the Portland  study, 58  1975-77  cars  which
received only  idle mixture and/or speed adjustments  were  charged  an average of
$9.50  by  the repair facilities.  And  for  313  1975-79 cars which  were  taken to
commercial  repair  facilities  in various parts  of the country as  part  of EPA's

-------
                                      46

FY79 Emission  Factors Program  [14],  the  average  charge  for  an  idle  mixture
adjustment was $6.29.

It  should  be mentioned  here  that  passenger  cars  produced  by  General  Motors
Corporation  (GM)  during  the  1979  and  1980  model  years were  equipped  with
carburetor idle mixture  adjustment  screws  which  are inaccessable without first
removing the carburetor.   It  can be  expected  that a  carburetor  adjustment on
these  vehicles  will  cost  more  than  has  been estimated  for  other  vehicles.
However, since  the  carburetor idle mixture  of these  vehicles  cannot normally
be adjusted, the  occurrance  of idle mixture maladjustments will  be  lower.   In
the  Houston  Program  only  3.1%  of  the  98 1979  and   1980  GM  vehicles  tested
failed  the program's  idle  CO cutpoints.   In  all,   these  vehicles  will  account
for only 11% of all pre-1981 vehicles in January  of 1983  when  all I/M programs
are  scheduled  to  be  in operation.   It  is  not  expected,  then,  that  these
vehicles will have  any significant effects on the overall repair  costs being
estimated for idle CO programs.

4.2.3 Net Effect on I/M Costs

It was  shown in Section 4.2.1 that  by  just eliminating  HC-only  failures,  30%
of  the  total repair  bill  for  pre-1981  vehicles  can be  saved  compared to  a
conventional  I/M  program.   The  cost per  vehicle  for the remaining vehicles
will also be reduced, down to  the cost of  a carburetor adjustment,  about $6 to
$10.   Using  the  repair cost  of $23..86  from  Portland   as  typical   for  these
vehicles in a conventional I/M program and $10 for an idle mixture adjustment,
the net effect is that an idle CO I/M program  will have  total  repair  costs  for
pre-1981 vehicles that are about  70%  less  than a  conventional  I/M program with
the same CO cutpoint.

EPA  has   recently   estimated  all   the   cost   components   for  a  typical
(conventional) I/M program design  [4].   Averaged  over the five years  from 1983
through 1987, repair  costs  for  pre-1981  vehicles account  for  27%  of  the total
cost of the I/M  program.   Therefore,  an  idle  CO I/M  program  costs will  be
about 19% cheaper than a conventional I/M  program  with the same  CO cutpoint as
the  result  of  reduced  costs  for  pre-1981  vehicles  alone.   Overall  program
costs  will  be reduced  about  26%,  including  the  savings  from 1981   and  later
vehicles discussed in Section 5.2.

4.3 Emission Reductions

There  are  several  operating  programs,  such  as   the  Portland  program,  which
provide data that can be used to estimate emission reductions  in conventional
I/M programs. 'Since  there are no operating  idle  CO I/M  programs,  the emission
reductions  that we  can  expect  from such  programs  can only  be  estimated  from
evidence  gathered  from  a variety  of  sources that  are  each  only  partially
relevant.  This section  will examine  the available data  which  may provide some
indication  of the  possible effects  of  idle  CO  I/M programs  on HC  and  CO
emission reductions observed in conventional I/M programs.

In  order  to provide  a  convenient analytical  framework for analyzing emission
reductions,  a failure  group  approach  will  be  used.  That,  is,  the  three  groups
of  "failed"  vehicles will be  analyzed  separately:  vehicles  which would  have

-------
                                       47

failed  HC only  if  there was  an idle  HC outpoint  but  will  now pass  instead
(HC-only  failures),  vehicles which  would have  failed HC and CO and will now
fail only idle  CO (HC-and-CO  failures),  and vehicles which would have  failed
idle CO only (CO-only  failures).

4.3.1 Carbon Monoxide  Emissions

HC-only failures - Any CO emission  benefits from vehicles which  fail  only  idle
HC  cutpoints  in a  conventional I/M program will  be lost  in an  idle  CO I/M
program since these vehicles  will  pass their idle CO cutpoint.   However,  these
vehicles  do not  produce any CO  benefits  from repair  even  in  conventional
programs  and  therefore  dropping  these vehicles  from an  idle CO  I/M  program
will not  reduce the overall  CO benefits  of the program.   This  conclusion is
supported by data  from the  Portland Study.   In .Portland,  those  vehicles which
failed only for idle  HC had  low FTP CO emissions to begin  with  and  showed no
CO improvement  from repairs.

HC-and-CO  failures -   Vehicles  which  fail  both  the   idle  HC   and  idle  CO
cutpoints in a conventional I/M program will still  fail  for  idle CO  in  an idle
CO I/M program.   It is  expected  that the  CO emission reduction from  the  repair
of  these  vehicles  in  an  idle  CO   I/M program  will be  the same  as  the CO
emission  reduction  that  would  be  observed  from  the   same vehicles In  a
conventional I/M program.   This  is expected  since  the  same repairs  used to
reduce idle CO  emissions in the conventional  program will  still be  needed in
the idle CO program in  order to pass the idle CO  cutpoints.

Also, the additional repairs  which  reduce idle HC  emissions in  a conventional
program  do not significantly  affect  CO  emissions.  As  mentioned earlier,
vehicles  in  Portland   which  failed   only  for idle HC  and received  repairs to
reduce  their  idle  HC   emissions  showed no  improvement  in their  CO  emissions.
In an EPA test program  where misfire was  induced  on a sample of  eight pre-1981
vehicles  an  average  11% misfire drastically  increased  HC  emissions  but had
little  effect  on  CO   emissions.  A correction  of misfire  on  these vehicles
would have  resulted  in  only  a small  reduction  in  FTP  CO  emissions.   In the
Houston  program[15]   additional  repairs  other  than  carburetor  adjustments
reduced FTP  CO  emissions  only  an  additional  1.25 gpm  after  the   carburetor
adjustment on  those  vehicles  which failed  both  idle HC  and idle  CO..  These
additional repairs would only have  increased the total  CO  emission  reduction
from repairs by only 2.4%.                              ....  . . ..-.^•!   . ,'  ,-

CO-only  failures - Those  vehicles  which  would  fail  only .for   idle  CO. in  a
conventional I/M program will also  fail in an idle  CO program'.   These vehicles
will receive similar repairs and therefore  tbe  sane reductions in CO  emissions
as in a conventional program since the  retest requirements will.be the same..

The conclusion  that can be  drawn is  that  for pre-1981 vehicles, dropping the
idle HC  cutpoints  from  a  conventional I/M  program,  to  arrive  at an idle CO
program, will not  reduce the overall CO  benefits of the  I/M program, despite
the elimination of idle HC  only failures and  the additional repairs  performed
on some CO failures in a conventional I/M program to pass idle HC cutpoints.

-------
                                      48

4.3.2 Hydrocarbons/Emissions

HC-only failures  -  In an  idle  CO  I/M program all of  Che  HC emission  benefits
from vehicles  which  would have  failed  the  idle  HC cutpoints  but not  the  idle
CO  cutpoint  in a conventional  I/M program  will  be  lost  since  these  vehicles
will not  be failed  in an idle CO  program.  For  1976-77  vehicles  tested  in
Portland during 1978  and  1979 in the Portland Study,  only  11% of the  total  HC
benefit observed in the program came from the idle EC only  failures.   However,
the  idle HC  only failure  rate  and  their contribution to  the  total HC  benefit
of a conventional I/M program is expected to increase as the  pre-1981  vehicles
age.  By the time I/M programs  begin operation in 1982 or 1983 it  is  estimated
that  idle  HC  only  failures  will  account  for  26%  to 27%  of  the  total  HC
emission reductions  from  pre-1981  vehicles  in conventional  I/M  programs.   By
1987 idle HC only failures will account for about 29% of  the HC benefits  from
these vehicles.

HC-and-CO Failures -  The  HC emission reductions  from  repair of  vehicles  which
fail idle CO in  the  idle  CO I/M program but would have  also failed idle HC  in
a  conventional I/M  program  are expected  to be  less  than  the  HC  reduction
observed in  conventional  I/M programs.  Primarily this  is due to  the  loss  of
additional repairs}  such  as ignition repairs, necessary in a conventional I/M
program to  reduce- idle HC emissions which  will no  longer be required in  an
idle CO program.

While there is no precise  information on how much of  the HC emission  reduction
benefit will  be  lost,  this  loss can be  estimated  by  examining  existing  data
which compares the HC  emission reductions produced by  carburetor adjustments
and  those  produced  by more  complete  tune-ups  including  ignition  repairs  on
vehicles  which fail   both  idle HC  and  idle CO  cutpoints.   In  the  Houston
Program additional repairs after the carburetor  adjustment increased  the  total
FTP  HC  emission  benefit   an  additional  9.4% on  a  small  sample  of  vehicles
initially  failing  both   idle   HC  and  idle  CO.   Also,   additional   repairs
increased  che  total  HC emission  benefit another 4.0% in  a  small  sample  of
vehicles  failing  both  idle   HC and  idle  CO  in   the  FY79   Emission Factor
program[14]  after  a  carburetor   adjustment.    The  Restorative  Maintenance
Evaluationf12]  shows  an   additional  3.8%  HC benefit  in  further  repairs  on
vehicles  failing   both idle  HC and  idle CO  after  a  carburetor  adjustment.
Vehicles  in  the  Emission Factor  and  Restorative  Maintenance  samples   were
reasonably  new when   testing  was   done  and  were not  expected  to  have   many
problems which would  cause  high HC  emissions.   Also,  the vehicles  in  these
samples would  have  received  less  of  Che additional  repairs  in a conventional
I/M  program.   To  quantify Che  Loss  precisely would  require  new  testing since
Che  existing  data only  approximaces che  expected  effects  of an  idle  CO I/M
program.

CO-Only Failures  - Vehicles which  fail  only for idle  CO in a conventional I/M
program will also  fail in an  idle  CO program and will receive essentially the
same  repairs.  The HC emission  reductions   from  these  repairs in  an  idle  CO
program will therefore be  the same as in a conventional  program.

-------
                                       49

In  conclusion,  the  total  HC  emission benefits  possible  from an  idle CO  I/M
program will  not  be as- great  as from a conventional  I/M program  for  pre-1981
vehicles.   The  loss  in HC  effectiveness  is  expected  to be  even greater  on
pre-catalyst   technology   vehicles   built  prior   to   1975,   but  there   is
insufficient  data  to  illustrate  their  loss.   An  idle  CO  I/M"  program  will
therefore  achieve  only  partial HC  emission  reduction benefits  for  pre-1981
vehicles.  Quantification of these benefits will  be  presented in Section 4.3.4.

4.3.3 Deterioration Issues

Between  inspections the  HC  and CO.  emission-  reductions  praduced; by an  I/M
repair: will gradually  deteriorate.    The.  reasons  for  this  deterioration  may
include/   emission   component.   failures   due   to   general  gradual.   wear,
maladjustments, and tampering- occurring between-  I/M inspections.   The  rate  of
this  deterioration affects  the net HC' .and CO emission benefits   that can  be
expected from an  I/M program.   EPA has done extensive  testing to  quantify  the
deterioration rate for  conventional I/M programs,  however,  no testing  program
has yet been conducted to measure  deterioration in an idle CO program.  Such  a
testing program would  be difficult  to conduct  with any  validity  since  there.
are currently no operational idle CO I/M programs which  can be evaluated.

One possible  line of speculation is that  the  deterioration  rate will be  less
in  an idle CO  I/M program since  mechanics will  have no need  to  temporarily
adjust carburetors very- lean as a  cheap way to  pass- idle; HC  cutpoints as  is
sometimes  done  in  conventional  I/M  programs.   However,  since  a  rich   idle
mixture may be  counteracting  other, problems which may  remain unrepaired in an
idle  CO I/M program, the  rate  of  deterioration of HC and CO emissions may be
higher than has been observed  in conventional  I/M  programs  if vehicles owners
maladjust  their vehicles  between I/M  inspections.   A  survey  of vehicle owners
in  EPA's  Emission  Factor  Programs a.  week after  they  had had  their  vehicles
carburetors  adjusted  to  manufacturers   specifications  indicates  that  most
vehicle owners are  satisifed with  the  performance of their vehicles after such
repairs.[16]  Although  Emission Factor Programs  do not simulate I/M  programs
and  although  carburetor-  adjustments  to  manufacturers  specifications  is  not
exactly  the repair  at  issue,  it  does  indicate  that:  vehicle  owners  can be
satisfied  with  the  performance of  their vehicles even  after  the  carburetor is
adjusted leaner than they have become accustomed to.

Lacking any relevant data, one  can  only assume  the same deterioration  behavior
for  idle  CO I/M  programs  as  for  conventional  I/M for  the  purposes  of  this
report.

4.3.4 Emission Reduction Benefits Model for Idle CO  I/M Programs
      for  Pre-1981 Vehicles

For pre-1981 model year vehicles,  EPA's simulation  model  for conventional I/M
programs can  be modified  to approximate the effects of idle  CO I/M  programs.
First,  the HC  benefit  attributable  to   vehicles  which  fail  only   HC  in  a
conventional I/M  program  are omitted.   Secondly,  the HC  benefit  attributable
to  vehicles which  fail  both pollutants  in a  conventional program  are reduced
based on assumptions concerning the  cause  of  failure.  An estimate is  made of
the  number of  conventional  HC-and-CO  failures which  have  high idle  HC  for
reasons  other  than  idle  mixture  maladjustment.   These  vehicles  are  then
modeled  as receiving  little  HC  reduction.   The  remaining   vehicles   in  the
HC-and-CO  group receive  the  normal (conventional) HC  reduction.  Finally,  the
HC  benefit attributable  tc vehicles  which  fail  only  CO  in -3  conventional
oro?ram is retained  in its entirety.

-------
                                      50

Figures  1  and 2  present sample  results  of  this  modified  simulation.   These
figures are based  on a conventional I/M  stringency  of 20% and  an  idle  CO  I/M
stringency  of  13%.*   The  two  programs  use  the  same  CO   outpoint.    The
difference in stringency is due to  the use  of an HC  outpoint in addition to an
idle  CO  cutpoint  in  the  conventional  I/M  program.   Figure  1  presents   the
composite HC  emission  factor for pre-1981  vehicles  without I/M and with each
of the two types  of  I/M.   Figure  2 shows the percent  benefits associated with
these programs for HC.

4.4 Fuel Economy Benefits

The  Portland   Study  did  not  indicate any  significant  overall  fuel  economy
benefit  as  a  result  of I/M  repairs  [17].   This  conflicts  with  data  from
various restorative  maintenance  programs  which  show about  a 4% improvement in
combined  city/highway  fuel  economy  after  emission  related  maintenance[18].
The question  is  the  following: what  specific differences  in  the repair types
performed in Portland could cause the disparity.

One piece of  the  solution is  the  fact that  carburetor  adjustments,  even when
they  are not performed  exactly  as   specified  by  the  manufacturer,  produce
sizable  fuel  economy   benefits   on   pre-1981  vehicles.   This  makes  sense
technically since a  vehicle with  an overly  rich idle mixture is consuming more
fuel  than  it  needs   and  any  carburetor   adjustment   that  results   in   a
significantly  leaner idle  mixture  should  result in less  fuel consumption  and
better  fuel  economy.   Recent  studies  in  Houston  and Portland  indicate  that
significant fuel  economy  benefits  between 1.1% and 4.6%  can  be  obtained  if
only  carburetor  adjustments  are  performed  after failing an  idle CO cutpoint.
Table 24 presents  the results of  those studies.
* The term stringency in  this  report  will  be  used to refer to Che selection of
appropriate  idle HC  and  CO  outpoints in  a  conventional  I/M program  or Che
selection of  idle CO outpoints  in  an  idle  CO  I/M program such chat Che failure
rate  in  Che  first year of  Che I/M program will  have a  failure  rate  equal to
Che  stringency,  i.e.,  a twenty  percent   stringency means  a  twenty  percent
failure  rate  in  the  first  year.    The   outpoints  selected  are  then  used
Chroughout all years of Che program's  operation.

-------
                                        51
                                     Figure 1


                       Comparison of  HC Emission Seductions
                              For Pre-1981 Vehicles
  5.50
                                  CONVENT IONfiL
                                  » / M
'0.0
        30
                                                                  30

-------
                                   52


                               Figure 2

                        Comparison of EC Benefits
                          For Pre-1981 Vehicles
 SO

 45

 40
uu
£

-------
                                       53
                                    Table 24

                             Fuel Economy Benefits
                         from Carburetor Adjustments  to
                       Vehicles  Failing  Idle CO  Cutpoints
                         (1975-80 Model Year Vehicles)
Program

Houston Program
Study1
Portland Task
Group 102
FY79 EF
Houston Site.
Number- of
Vehicles
Adjusted
Fuel
Economy
Before
Carburetor
Adjustment.
(mi/gal)
After
Carburetor
Adjustment
(mi /gal)
Percent
Change
(85)
(31)
(12.)
City
Hwy
Comb

City
Hwy
Comb

City
Hwy
Comb
13.79
20.57
16.19

14.39
20.14
16.51

14.22
21.71
16.92
14.61
21.01
16.93

14.70
20.03
16.70

14.74
22.29
17.39
5.9
2.1
4.6

2.1
-0.6
1.1

2.9
2.6
2.8
1:  Idle mixture adjusted so that idle CO is 0.5%
2:  Idle mixture adjusted so that idle CO is 0.2%
3:  Carburetor  adjusted  by field mechanics;  those
    CO after carburetor adjustment.
                                vehicles  passing  2.5%  idle
If  carburetor  adjustments  can  produce   sizable   fuel  economy  benefits  and
carburetor  adjustments  are   the  most   common  repair   in  conventional  I/M
programs, the lack of  significant  overall improvement  in fuel economy observed
in Portland must  be the  result of  repairs  other  than  carburetor  adjustments
which degrade fuel economy and  offset  the fuel  economy improvement  provided by
the  carburetor  adjustment.   Overly  retarded  ignition  spark  timing has  been
identified as  one such  repair in the  Port land. study  [19].   Analysis  of  the
available data will continue  in order  to  identify  other  repairs  which may also
degrade fuel economy.

In  idle  CO I/M  programs  carburetor adjustments   will   be  virtually the  only
repair necessary to  pass  the  reinspection idle CO  cutpoint.   It  can therefore
be stated with confidence that an  idle CO I/M  program  will  obtain fuel  economy
benefits.  The  size of the  fuel economy  benefit will depend  on  how  close
repair mechanics will  adjust  carburetors  to manufacturers  specifications.   We
do know that if mechanics adjust the carburetor such that  the  idle  CO is 0.5%,
the  fuel economy  benefit  may  be  as  high  as the  4.6%  obtained in  the  Houston
program.   This  fuel  economy  benefit can provide  significant  savings  in  fuel
costs to vehicle owners who must receive emission  related repairs.

-------
                                      54

Mechanics will not normally  be  expected to adjust carburetors much leaner  than
the program's  idle CO cutpoint,  since  they only need a  reasonable  margin of
safety  in order  to  be  sure to  pass  the  reinspection.    In  New  Jersey,   the
operating program has an idle CO  cutpoint  of 3.0% for 1975-80 vehicles and  the
average   idle   CO  measurement   after   repair   is  0.9%.    However,   if   the
reinspection cutpoint is tighter, mechanics  will be forced to adjust leaner in
order  to keep a  reasonable margin  of  safety.   In the Portland  program most
1975-30  vehicles must  pass  a 1.0% reinspection  cutpoint.   The average idle CO
measurement after  repair in this  program is 0.2%.  this means that one way to
assure that mechanics adjust carburetors  so that the idle CO measurement after
repair is 0.5%  is  to impose an appropriately  tight reinspection cutpoint.  It
has been estimated that  a  reinspection  idle CO  cutpoint  between 1.0% and 2.0%
will  result  in an average   after-repair  idle  CO  measurement  of 0.5%  or  less
with  no  adverse  side effects.  (The  Portland  I/M program has been using a CO
cutpoint  of 1%  for many  years.)  This  may mean  having a reinspection cutpoint
which  is  different   than   the  initial  inspection  cutpoint.    (An  initial
inspection  cutpoint  of  2.0% would   result  in  a  37%  stringency  for  1975-80
vehicles, higher  than desired  by many I/M planners.)   This should  cause no
conflict, since  the  purpose  of the   two  cutpoints is  different.   The initial
cutpoint  is  set  to  identify  those  vehicles  most  in need  of  repair.    The
reinspection  cutpoint  assures  that  those repairs  were effective  in reducing
the  vehicles   CO   emissions.    The  inspection  cutpoint/,   therefore,  does   not
necessarily need to be as tight as the  reinspection cutpoint.

Another  way  to assure that  most  repair mechanics will adjust  carburetors so
that  the  idle CO after repair is  0.5% is through mechanic training.  Mechanics
who do repairs  in  the  I/M  program can either be  instructed  to adjust all cars
failing  the  idle  CO  cutpoints  so  that the  idle  CO  is  0.5%  as   a  matter of
course or they can be instructed  to  do  so  if the vehicle's idle CO measurement
exceeds  0.5%   after  adjusting  the   vehicles  idle  mixture  using   manufacturer
specified   procedures.    Additional   information  on   overall   carburetor
diagnostics and repair can also be included  in the  program.

If a  tighter  reinspection cutpoint or mechanic  training is not possible a 0.5%
idle  CO  adjustment target can be  mandated by the  program  as a requirement  for
vehicles  failing  the  idle  CO cutpoint.   This approach  will  require a periodic
statistical  review  by  the  program administrators   of   post-repair  idle  CO
measurement records  from each repair facility to enforce the requirement.   The
effectiveness of this approach  will  depend heavily on how well the requirement
can be enforced.

The   emission  benefits  crom  any  of  chess  options   which   encourage  leaner
adjustments Chan would be necessary  merely Co  pass Che iricial inspeccion idle
CO cutpoint will be discussed in  Section 7.3.

-------
                                      55

5.0 I/M FOR 1981 AND LATER VEHICLES USING ONLY  IDLE CO OUTPOINTS

5.1 Background on 1981 and Later Vehicles

There are two  significant  technology  differences between pre-1981 and  1981  and
later vehicles..   First,  the: majority of  1981 and later vehicles are  expected
to  employ microprocessor-based,  closed-loop, engine  control  systems.   These
systems rely  on a. network  of, sensors to  supply information on  the: operating.
condition of  the  engine to  an on-board  microprocessor.  .The microprocessor
analyzes  this  information- and: sends operational commands  to  various actuators
(primarily- solenoids)  which: modify the air/fuel, ratio,  control the deployment
of  secondary  air,   determine  spark-  timing,  and.  on. some  systems  allow  or
disallow  exhaust  gas  recirculation   (EGS).    These   systems   represent  a
significant departure  from earlier, mechanically controlled systems.

Second, the Parameter  Adjustment  regulations[2] provide  that  the fuel metering
systems  on  1981  and  later  vehicles  be  designed  in  such  a  way  so  as  to
discourage tampering or maladjustment of  the idle mixture and choke  systems.
As  is  widely  known, maladjustment of  the  idle mixture  system has  been  the
cause  of  a significant  share  of the  in-use  emissions  problem  for  pre-1981
vehicles.  Although  even the protected  designs  can still be maladjusted  (after.
some effort),  it  iai expected, that the- 1981  and later fleet  will  be  much  less
subject: to idle mixture  and. choke maladjustment.

Due to these  technology  differences,  the  general nature  of the expected  in-use
emission  performance of  1981 and later vehicles is different than for earlier
model year vehicles.  The  in-use  emissions  performance of earlier  model  year
vehicles  (pre-1981)  is  largely  affected  by  the  rate . of  various  forms  of
maladjustment  and tampering  (chiefly  idle mixture maladjustment)  as  well  as
phenomena such as  problems  with the  ignition system and catalyst deterioration
due to aging or misfueling.   By comparison,  while  1981  and later vehicles  will
still experience  ignition system problems  and  catalyst, deterioration  to  some
extent, .the  main  phenomenon affecting  in-use  performance  is  expected  to  be
failure of the microprocessor-based fuel  control system.  These failures often
result  in very rich modes of operation  with accompanying very  high  levels  of
EC  and  CO and poor  fuel economy.   These HC  and  CO  levels are  high  enough  so
that even if only a  small  number  of vehicles experience  a fuel system  failure,
they  can  dominate the emission inventory of the  fleet  and  cause  the  average
emissions of  the  fleet to  exceed the  standards.  The early data  on the  in-use
performance   of   microprocessor-controlled  vehicles  has   given  significant
indication that  this type of behavior can  be expected for the  1981  and later
fleet.

Rich  failures  of  the microprocessor-based  fuel  control  system primarily  affect
CO  emissions—causing  them  to rise  dramatically.   These  failures also cause  HC
emissions  to   rise,  but  to  a  somewhat  lesser  degree.   Repair of  these  rich
failures  result  in  significant  fuel   economy improvements   averaging  about
15%[6).   The  average in-use HC  emissions  of the 1981 and  later fleet are  also
largely  affected  by problems with  the  ignition system  (e.g.,  misfire,  fouled
plugs,  faulty distributor components).

-------
                                      56

Readers  interested  in  gaining  more  information  on   the   projected   in-use
performance of 1981 and  later  vehicles  should refer to "Derivation of  1981  and
Later  Light  Duty  Vehicle  Emission  Factors  for  Low  Altitude, Non-California
Areas"[20].

5.2 Idle CO I/M for 1981 and Later Vehicles

Given  the  preceeding  discussion  on  the expected  in-use  emissions performance
of the  1981  and later fleet,  it  should come as no  surprise  that the emission
reductions  attributed   to   I/M  are  expected  to  primarily  come  from   the
identification  and   repair  of  vehicles   with  a   rich   failure  of  their
microprocessor-based  fuel  control  system.   Data  from   such  vehicles  in   the
field  indicate  that   some   fraction of  them  will  readily  fail an   I/M   CO
cutpoint.  The  estimate used  in  this analysis is  that   50%  of those vehicles
with  a  failure  of  the  microprocessor-based  fuel  control  system will   be
identified in  an  idle CO I/M  program using the basic idle  test.   This  is  the
same  estimate  used for  a  conventional  I/M program.  This  means  that  the  CO
benefits  in  an  idle  CO program will  be  the  same as  in a  conventional   I/M
program.   Since   these   vehicles    also   have   high   HC   emissions,   their
identification  and repair  in  an idle  CO  I/M program  will  also  result   in
significant  HC  reductions.    Ignition  problems   (the other  main  source   of
projected HC emissions  for the 1981  and  later fleet)  are not identified by  a
CO  cutpoint,   and  therefore  an  idle  CO . I/M program   will  not  get   the   HC
reduction benefits of their.repair.

A conventional  I/M using an  idle  test program is  expected to fail about 7%  of
the 1981 ami later vehicle  fleet.  An idle CO I/M program using the  same test
will   only   fail   vehicles  with  microprocessor-based   fuel  control  system
failures, or about 1.7%  of the 1981 and  later  fleet.  The  other vehicles fail
in a  conventional  I/M program for  idle  HC  emissions  only.    Since all  of  the
fuel  economy benefits  associated  with repair of  1981  and later vehicles comes
from  these vehicles  with  fuel  control system  failures,  idle  CO  I/M programs
will  achieve  the  same  overall  fuel  economy  benefits  from  1981  and  later
vehicles as  conventional I/M  programs.   This amounts to an  average 15% fuel
economy  improvement  for  each  vehicle  experiencing  a  fuel system  control
failure which is identified and repaired.

Repair  costs  for  1981 and  later  vehicles has  been estimated  to  be  about   $30
per repair [4].  These  repairs include ignition parts replacements as  well  as
repairs  to  the microprocessor-based  fuel  control system.   For  an idle   CO
program  these  costs are assumed  to  be Che  same,  although  only  fuel  control
system  repairs will be necessary.  However,  since  Che  failure rate for an idle
CO program is only about ?5% of 'che  failure  rate  of a  conventional I/M program
for 1981 and Later vehicles, repair coses are reduced about 75%.

The  following  section  will  quantify  the emission  benefits  expected  for   the
1981  and  later fleet  in an  idle  CO  program.   It will especially  document   the
loss of HC benefit due to not identifying vehicles with ignition problems.

-------
                                      57

Readers interested  in gaining more information  on the methodology involved in
estimating  I/M  benefits  for  the  1981  and   later   fleet  should   refer  to
"Derivation  of  I/M  Benefits  for  Post-1980   Light   Duty  Vehicles   for  Low
Altitude,  Non-California Areas"[21].

5.3 Emission Benefits Model for Idle CO I/M Programs for 1981 and
    Later- Vehicles

The I/M emission, benefits attributed  to a; conventional  program-  for  1981 and
later  vehicles  come  from two  sources:   the:  identification  and repair  of
vehicles with  rich  failures of  their microprocessor  control,  system  (yielding
HC anc CO benefits)  and  the HC benefits resulting from the identification and
repair  of  vehicles  with ignition  system  problems.   As. discussed  above,  the
first source of  benefits  is  not  diminished in an  Idle CO  I/M program although
the second  source  would be  lost.   EPA's  computer model which  calculates I/M
benefits for  1981  and later  vehicles  in  a  conventional  I/M  program  can  be
easily modified to account for this partial loss of HC  benefits[20].

Figures 3  and 4- compare  the  HC  and. CO  emission reduction  benefits  from  a
conventional  I/M program and  a  idle  CO  I/M  program.    The  benefits  shown
represent those from, the 1981 and later fleet only.

-------
                                       58




                                    Figure 3
                                       Later Vehicles
   t.SO
'0.0
             81    32    63
                                    35   38    37
aa    33
                                                                30

-------
                                    59

                                 Figure 4

                         Comparison of EC Benefits
                        For 1381 and Later Vehicles
 no

 35

 30
o2Q
a

-------
                                      60

6.0 OVERALL EMISSION REDUCTION BENEFITS FOR I/M PROGRAMS
    USING ONLY IDLE CO OUTPOINTS

The  results  from  models for  benefits  of  idle CO  I/M programs  for  pre-1981
vehicles and  1981  and later vehicles as  discussed in  sections  4.3.4  and  5.3,
respectively, can be combined to determine benefits for the  entire  fleet.

Benefits for CO in an idle CO I/M program are  expected to  be the same  as  for  a
conventional I/M program if  the  same idle CO  cutpoints are used.  The  effects
for HC  are  presented in  Table  20.   The  HC  benefits  on December 31,  1987  for
conventional and  idle CO programs  starting January  1, 1983  with  no  mechanic
training and  a simple  idle  test for  1981  and  later  vehicles  are shown.  It
should  be  noted that  the idle CO  programs presented  in  Table  25 would  have
higher  CO benefits  than the  conventional programs  since  the idle  CO cutpoints
for a given  failure  rate idle CO  program are  lower than  the idle  CO cutpoints
for a conventional program of the same stringency  (failure  rate).*

It  can  be  seen in  the  table  that areas  with ozone  non-attainment  problems
using an idle CO program will  likely need to implement measures to increase HC
benefits.   Methods  for  increasing HC   benefits  are  discussed  in  the   next
section.

If the  idle CO I/M program was restricted to only 19.75 and  later vehicles,  the
composite HC benefit  for idle CO programs would  be further reduced.   Pre-1975
vehicles account  for between 7.4%  and  10.7%  of  the  HC benefit  in an  idle CO
I/M program, depending  on stringency.   The  HC benefit  in conventional  programs
would also be  reduced if all pre-1975 vehicles  were  exempted from inspection.
Table  25  also  presents  the  emission  benefits of  an  idle CO  program  which
exempts pre-1975 vehicles.
* The  term stringency  in  this  report  will  be used to refer to the selection of
appropriate  idle HC  and  CO  cutpoints  in  a conventional  I/M  program  or  the
selection of  idle CO cutpoints  in  an  idle  CO I/M program such that  the  failure
rate  in the  first year of the I/M program  will  have a  failure rate equal to
the  stringency,  i.e.,  a twenty  percent   stringency means  a  twenty  percent
failure  rate  in  the  first  year.   The   cutpoints  selected  are   then  used
throughout all years of the program's  operation.

-------
                                       61

                                   Table 25

            Comparison of EC Benefits  for Conventional and  Idle  CO
                        I/M Programs*  Starting  in  1983

Stringency for                            Percent.  Benefit
Pre-1981 Vehicles         	on-December 31,  1987	
(Failure' Rate) *"*        Conventional       Idle CO         1975-*-  Idle CO
                            HC   CO.           H£  CO       -'   HC    CO

      13Z                   32:   30           24   33            22    29
      20Z                   35   33           27   36            24    32
      30%                   38   36           30   39            27    34
      40Z                   39   38           31   40            28    35
*  Further  assumption about  the I/M  programs  are no  mechanic training  and a
simple idle test for 1981. and later vehicles.

**  At  a  given  stringency  (failure  rate) ,   the   idle  CO  I/M  program  has
numerically-  lower CO outpoints- than the  conventional program, but: has equal
stringency overall.


Using the model for  emission benefits it is possible  to  calculate. the tons of
pollutants  that are  removed  by  an  T/M  program.   The  methodology  for   this
calculation  is  presented in "Update on  the  Cost-Effectiveness of  Inspection
and Maintenance [4].  Table  26  presents the results of such  a calculation for
idle  CO  programs.   For  comparison  a   conventional  I/M  program  with  a  20%
stringency  for  pre-1981  vehicles   produces   a  46,500   ton  reduction  in  HC
emissions  and  a  526,800  ton  reduction in CO emissions  from a  base  of  one
million vehicles.
                                   Table 26

                        Five Year Emission Reductions  -

               Thousands of Tons Removed* by an Idle CO Program

Stringency for
Pre-1981 Vehicles          All >fodel Years     '       Only 1975 and Later
    (Percent)                H£       C£                HC         CO

       13                   31.31    526.8             29.26      435.7
       20                   36.38    607.4             32.87      502.9
       30                   ^3.61    692.7             38.08      572.5
       40                   45.64    723.1             39.81      598.7
* A vehicle base of one million vehicles is used.

-------
                                      62

7.0 METHODS TO IMPROVE THE HC EMISSION REDUCTION FROM
    IDLE CO I/M PROGRAMS

As described  generally  in Section 4.3.2  and  as quantified  in  Section 6.0,  it
is expected  that  a conventional  I/M  program which  simply  eliminates the  idle
HC  emission   requirements  from  the   program  will   suffer   some  loss   of
effectiveness  in  reducing HC emissions'  from automobiles when  compared to  the
conventional progam.  Table 20  showed that  there  is over a 20% percent loss  in
HC  effectiveness  even  if the  CO cutpoint  is tightened  enough  to  bring  the
overall stringency  of the program back to that of  a conventional I/M program.
This will  have an impact on areas wifh  ozone air  quality  problems  which  must
reduce the overall  emissions of HC from automobiles and which  may wish to  use
the idle CO  I/M  program approach.  For this  reason,  this  section will examine
several methods which can be  used to increase  the  effectiveness  of  an idle  CO
program in reducing HC emissions.  These methods include:

         Using a more effective  test procedure  for  1981  and  later  vehicles.
         Increasing the failure  rate of pre-1981 vehicles.
         Assuring better  repairs  on pre-1981 vehicles.
         Re-establishment of a  loose idle HC cutpoint.
         Addition of tampering  checks.
         I/M for light-duty trucks.

Many of  these improvements are  also expected  to increase  the  reduction of  CO
emissions.

Most of these  improvements can  also  be  applied to conventional I/M programs  if
additional reductions of  HC and  CO emissions are desirable.

7.1 More Effective Test Procedure  for 1981 and  Later Vehicles

One means  of  increasing the  HC reductions from an  Idle  CO  program is to use a
more  effective  test  procedure   for  1981 and  later  vehicles.   The benefits
presented  in Section 5.0  were  based  on  the  use of a basic idle-in-neutral  test
for  1981  and  later vehicles.   This test  is  estimated  to  catch 50%  of  the
vehicles   with  rich  failures   of  their  microprocessor-based   fuel  control
system.   The  addition   of other pass/fail  vehicle  operating  modes  to   the
overall  test  procedure  has   been shown  to  increase  this percentage.   For
example,  the  addition of a  2500 rpm  test  mode (Two  Speed Idle  Test)  or  the
addition  of  a 30 mph/9.0 HP loaded  test  mode  ("Loaded Test)  has  been shown  to
result in  a 70%  identification  rate  of  the  vehicles with a fuel control system
failure[21!.   This  will  increase Che failure rate  for  1981  and later vehicles
from  about L.7%  to 2.4%.  Either of  these  tests  will still  qualify  vehicle
owners  for coverage under the  Emission Performance Warranty[3j.   Since  these
additional  rich  failures increase both  HC and  CO  emissions,  yet are  readily
identifiable  using  only a CO  cutpoint,  any increase in  the number of vehicles
identified by  the  CO  cutpoint  will cause the HC  benefits  to increase.   Repair
of  rich  failures  among   1981  and  later  vehicles has  produced  an average  '15%
fuel  economy  benefit.   The  fuel  savings from  these  repairs,  therefore  more
than offset the cost of an average repair cost  of about  $30.

-------
                                       63

Figures 5 and 6 show  the  improvement in HC effectiveness which  can  be-achieved
by  using  the  Two  Speed  Idle  Test  or   the  Loaded  Test.   For  purposes  of
comparison,  the  HC  effectiveness  of  using  the  basic  idle  test  is   also
included.   Figures  5  and  6- apply only  to  1981  and  later  vehicles.    In
fleetwide  terms,  the  use  of one  of these more  effective  tests  for  1981  and
later vehicles can  increase- the- composite HC emission reduction  from I/M as  of
December 31,  1987  by 2.4- percentage points.  For example, Table.  19  showed a
20% stringent  idle  CO I/M program will yield an HC reduction  of: 27%, on  this;
date.   If-  a more  effective test is, used  for 1981  and  later  vehicles,   this,
reduction would be- improved to 29%.  This option  will  increase the  tons of  HC
removed by  an  idle CO program by 2,030 tons and  CO  by 60,400  tons in. a.  base
sample of one million vehicles.

7.2 Higher Failure  Rates for Pre-1981 Vehicles

If the  idle CO cutpoint is made  more  stringent  more vehicles  will   fail their
initial idle CO  inspection, more vehicles will  receive repairs, and  therefore
it is logical  to  expect  that there will be an overall  increase  in the HC  (and
CO)  benefits of  the  idle  CO I/M program.   Also,  if the  reinspection idle  CO
cutpoints  are  identical  to  the  initital  inspection cutpoints,  tightening  the
idle  CO  cutpoints will   have  the effect  of   forcing mechanics   to adjust
carburetors  on all  failing  cars  leaner" than  they  would  have with looser-
cutpoints.   This  effect  is discussed  in: Section  4-.4.   Since these  additional
failures  will  ail  receive  the   same-   repairs   as   the  original   failures
(carburetor adjustments),  the  increase in HC  benefit  to  the  program will not
be  as   large  as  the  increase  in  the CO  benefit.   In  addition,   there  are
diminishing  returns from  failing more vehicles   as more  and  more   relatively
clean vehicles  fail  the   stringent  initial  inspection  idle CO  cutpoints  and
produce very little HC or  CO benefit from repairs.  This  is evident in Figure
7.

Figure  7 presents  the effects of increasing stringency*  on pre-1981  vehicles
subject  to  an idle  CO   I/M  program.   Figure   7 applies only to  pre-1981
vehicles.   Table   25  in   Section 6.0  presents   the  HC  and  CO  benefits   of
increasing  stringency on  the  pre-1981  vehicles and Table  26 presents  the   tons
of pollutants removed.

7.3 Better Repairs

On pre-1981 vehicles,  mechanics will on the  average only adjust carburetors  to
just  pass   an  idle  CO  reinspection  cutpoint with  a  reasonable   margin   of
safety.   If the reinspection cutpoint is fairly loose,  vehicles after repair
* The term stringency in this report will be  used  to  refer  to the selection of
appropriate  idle  HC  and  CO  cutpoints  in  a  conventional  I/M program  or  the
selection of idle CO cutpoints in an idle CO  I/M program such that the failure
rate in  the  first  year of  the  I/M  program will  have  a failure  rate  equal to
the  stringency,  i.e.,  a  twenty percent   stringency  means  a twenty  percent
failure  rate  in  the  first year.   The   cutpoints  selected  are  then  used
throughout all years of the program's operation.

-------
                                       64.
                                   Figure 5

                   Effect of a More Effective Test Procedure
                         on Che HC Emission Seductions
                          For 1981 and Later Vehicles
 2.0
err
(U
 ,0.3
a
a.
s<
  0.0
                                  MO  I/N
                                  CONVENTlaNflL
                                  I/M CS0231
                                  IffLS CO
                                   50231
                                               Benefit from using
                                               a better cest for
                                               1981 and later vehicles.
         SO    31    32   33   34    35   6*6
                                CflL£NOflB  TEP.R
                                                  37
39
30
  1:   Idle CO  Test  (50Z Identification Sate)
  2:   Two-Speed  Idle Test or Loaded Test (702 Identification Rate)

-------
                                       65
                                   Figure 6

                   Effect of. a. More Effective Test' Procedure
                              on Che HC Benefits.
                          For 1981. and Later Vehicles
  40


  3S


  30
Ul
09
u
OS-
£lO


   5
            IOL£  cn I/M
            IDLE  Cff I/H
                                             Benefit from using
                                             a better cest for
                                             1981 and later vehicles.
       30
                   22    33
34     35    36
CflL£HOflH  TEflH
37
39
30
 1:   Idle CO Test (502 Identification Rate)
 2:   Two-Soeed Idle Test or Loaded Test (707. Identification  Hate)

-------
                                    66
                                 Figure 7

                 Effect of Increasing  Idle CO I/M Stringency
                    on HC Benefits for Pre-1981 7ehicles
  SO

  45

  40
 -20
 UJ

 Sis
 Ul
 a.
  10
            CONVENTIONAL
            I/H  C20»*
            IOLE CO  I/H
IDLE CO  I/M
(30X3*
IDLE CO  I/M
        30
       92    33
34    35    36
CflLENOflfl TEflR
37
38
30
*?rogram Failure Rata in Che inicial program year.

-------
                                    67


                                 Figure 3

                         Effect of Better Repairs*
                            on the 1C Benefits
                           for Pre-1981 Vehicles
 50


 45
to.
£30.
tt

«25,


520
iu

Sis
LU
O.
  10
           CONYENTrONflL
           I/H
           IOLS: co  I/M
           IQL£ CO  I/M
           HITH «T*
80
81
32
33
                               3U    35
                              CfiLENOflfl
                                          36
37
33    30
*Bectar  Repairs are acqiiired either by  eight reinspection idle CO outpoints  or
"hrough  aechanic training programs chat reach nearly, all repaired vehicles.

-------
                                      68

will still have idle mixture's adjusted  richer, than they would have been  if  all
carburetors  were  adjusted  to  manufacturer  specifications.   In  New  Jersey's
operating  program the  idle  CO  reinspection  cutpoint for  1975-80 vehicles  is
3.0%.   After repair  the average  idle  CO  measurement  of  failed  vehicles  is
reduced  to  0.9%.   In the Portland  program,  most 1975-80 vehicles must  pass  a
1.0% idle  CO reinspection cutpoint and the average  idle  CO measurement after
repair  is 0.2%.   The  average idle CO measurement  of  vehicles failing  the idle
CO  outpoints  in  the  Houston  program  which  adjusted   all  vehicles   using
manufacturer  specifications  had an average idle CO  after adjustment of  0.15%.
This data  indicates  that there is  some room for improvement  in  the manner  in
which  mechanics  perform  carburetor   adjustments  in   I/M  programs.    Better
carburetor   adjustments   should  result  in  greater HC   (and  CO)   emission
reductions from the carburetor adjustments.

For  1981 and  later  vehicles,  however, the  logic  that improved  repair will
provide  larger  benefits does  not  apply  in  the  same way.   This  can  best  be
understood  by recalling  the types  of  failures  expected  for 1981  and  later
vehicles.   These  failures are  truly "system"  failures, where the failure  or
disconnection of one or more of  the components  which make up the  system  causes
the entire  system to  malfunction.   Proper operation  of  the system can only  be
restored once the problem is  diagnosed and  the  responsible components  either
replaced  or  reconnected.    For  1981 and  later  vehicles,  if a  vehicle has
experienced  a fuel control system failure  and fails  the I/M  test,  its  emission
will be high enough so  that  it will not be  able  to pass  the I/M test until  the
problem  has  been diagnosed  and  fixed.   The  benefits accompanying this  repair
will be  total.  It will not  be possible to  increase  the benefits  by adjusting
closer  to  manufacturer  specifications  since   no adjustments   are  involved.
Thus, it will not be possible  to  extract  more HC benefit  from 1981  and  later
vehicles by  striving  to improve repairs.   Repairs on these vehicles  are of  an
all-or-nothing nature.   Improved  repairs  on  1981  and later vehicles will have
advantages  in terms  of  lower  repair costs  (due  to   less wasted  effort  during
repair)  and  less  owner  inconvenience  (due to fewer return  trips  to the  repair
facility).

On  pre-1981  vehicles,   it  will  be  possible   to get  mechanics  to  perform
carburetor adjustments  closer to manufacturer specifications  in three ways:
         Mechanic training
         Lower  idle CO  reinspection outpoints, and
         Statistical review of idle CO reinspection results.
These approaches are described  in  Section ^.^.   One or all of these approaches
can  be  applied co  any  I/M program  to  improve  the carburetor  adjustments
performed  in   the  program.   However,   it  is   expected   that  none  of  these
approaches can  do  better than  the  results  now  being  achieved  in the Portland
program  where  most 1975-80  vehicles must pass a  1.0% idle  CO  inspection and
reinspection  cutpoint.   The  average   idle  CO  .measurement  after  repair  of
1975-77  vehicles  in Portland which fail for idle  CO  (excluding  idle-HC only
failures) is 0.3%

-------
                                       69

Better repairs  for pre-1981 vehicles  deserve  special attention primarily as  a
cost-saving measure,  although  they are also an  HC and CO enhancement.   Better
repairs  therefore  deserve attention from  any  area considering  an idle  CO  I/M
program, not just  those  that might  need  it as  an HC or CO enhancement.   Better
repairs  for  pre-1981 vehicles  will  produce   fuel economy  savings  on  these
vehicles which will offset: a large  part of the cost of the I/M  program.

EPA  studies  have  shown:  that  fuel savings; of  4Z, are achieved by  carburetor-
adjustments to  pre-1981,  vehicles which  have  failed,  an I/M-test,  even  if  the
adjustments, are- not  performed: precisely- to  specification..  It  is  necessary
that adjustments be performed  closer- to specification  than  they would be in  a ••
basic,  idle CO or- conventional I/M  program^   and  that: other types  of  repairs
which  can  degrade  fuel   economy  be avoided.    Because  carburetor  adjustments
will usually be  the only necessary repair for pre-1981 vehicles in an idle CO
I/M program, the types of repairs  which  can degrade fuel economy are  naturally
avoided in this type  of  program.   Section 4.4  describes a number of  approaches
that I/M administrators  can take  to improve the quality  of carburetor adjust-
ments,  two of which  are mechanic  training and  a tight- idle  CO reinspection
cutpoint.  The- fuel savings which- will result  are well worth the' effort,  since
an  additional  4Z  savings foe  each-  pre-1981  vehicle  which   is   failed  and
repaired translates into  a total,  annual savings of about $36.   This offsets  a
large  part of- the program, costs.-  Also,  a small, improvement: in the HC. and. CO
emission- reductions  accompanies  the fuel  savingsv  Section-  7.3 discusses the-
HC" and CO  benefits.   The overall  effect, is a  major further- improvement  in the
cost-effectiveness of the  idle CO I/M  program.

The  quality  of adjustment observed in  the Portland  study  is   expected  to  be
achieved in a conventional I/M program with mechanic training.   EPA's emission
model  for  conventional   I/M  programs  with mechanic  training  reflects this
judgment and  can therefore be modified in the same way  as  the  basic model  to
predict the maximum effect of  better- repairs  in an idle  CO  I/M program.  This
process is described in Section 4.3.4.

Figure 8 presents  Che maximum  effect of  better repairs performed in an idle CO
I/M  program  for  pre-1981  vehicles.   Figure  8  applies  only  Co  pre-1981
vehicles.  Table 27 presents the possible  overall  net  effect  in 1987 of  better
repairs on pre-1981 vehicles.

                                   Table-27

                Efface of  Better Repairs for Pre-1981 Vehicles
                           in  an  Idle  CO  I/M Program

         Stringency for            Percent Benefit  on December 31, 1987
         Pre-1981 Vehicles         No Mechanic             With Mechanic
            (Percent)               Training                 Training

                                    HC    CO                  HC     CO

               13                   24    33                  25     36
               20                   27    36                  28     40
               30                   30    39                  32     42
               40                   31    40                 '33     42

-------
                                      70

It can  be  assumed that the program  costs  associated with  better repairs  will
be  minimal  since  any  additional  time  required  to  perform  a  carburetor
adjustment more  precisely  would be insignificant  compared  to the basic  charge
for setup  and adjustment.  When fuel  savings  are  considered,  better  repairs
become  a  method  to  reduce  overall  program costs,  since  fuel  savings  from
repairs can  be  used  to  offset other program costs.  These  savings  have  been
calculated  using  figures  from  "Update  on  the   Fuel  Economy  Benefits  of
Inspection and Maintenance Programs" [5] and presented in Table  4  in  Section
1.2.   The  additional  tons  of EC  and  CO removed  through  better  repairs  have
been  calculated  using the  idle CO  emission benefit model.   These  reductions
are al? o  presented in Table  4.  Exempting  pre-1975 vehicles will  reduce  the
benefits  of   this  option.   Table  9  in  Section  1.2  presents  the   costs  and
benefits of better repairs without pre-1975 vehicles.

7.4 Re-establishment of a Loose Idle HC Cutpoint

Much  of the  loss  in  HC  benefits of  dropping the  idle  HC requirement  from a
conventional I/M program was due to  passing  vehicles in an idle CO I/M  program
which  would  have  failed  only  for  idle HC  in  a conventional  program.    These
vehicles provide a significant portion  of  the HC benefit in a conventional  I/M
program.  The effects are discussed in Section 4.3.2.

While  re-establishing an  idle HC  cutpoint  may  at first  seem to  defeat  the
basic  idea  behind an  idle CO  I/M  program,  it  is at  least  worth considering
whether a  reasonably  loose idle HC cutpoint  might identify  the  very  worst HC
emitters,   ones   for  which   repairs   to   reduce   idle  HC  would   be  very
cost-effective,  without  losing much of  the  attractive  aspects of the idle CO
program.

In the  Portland  study, the vehicles failing  only a hypothetical  500  ppm  idle
HC cutpoint  produced  77.5% of  the total HC  benefit from  all vehicles  failing
only  for  idle HC  in  the  Portland  study.   The  vehicles failing the  1000  ppm
cutpoint provided  61.3%  of the total  HC benefit  from  these  vehicles.   The  two
vehicles  failing only  the 1000 ppm idle HC cutpoint  provided  this  large a
percentage of the  total  HC benefit from the  idle  HC-only  failures  because  one
of them was  a vehicle with FTP HC  emissions  of 26.30 grams per  mile.   This is
more  than 17  times  the design  standard.  Although such  vehicles  are  rare,  they
are present  in  any fleet  and  produce very  large  HC emission  reductions when
repaired.  Because there  are  few  such  vehicles,  their repair  has  negligible
impact  on the overall  failure  rate  and  repair cost.   Also,  because very simple
problems,  such  as one  disconnected  spark plug  wire,  can  cause  very high HC
emissions, repairs can oftan be quite inexpensive.

In  an  idle  CO   I/M   program  with  a  loose  idle  'AC  cutpoint some  additional
benefit might be  obtained  from vehicles which  fail  both the  idle CO  cutpoint
and  the  loose   idle  HC  cutpoint   if   after  the   carburetor  adjustment   the
vehicle's idle HC emission still exceeds the idle HC reinspection cutpoint.

The  level  of HC benefit  observed  in  the  Portland  study for  loose  idle  HC
cutpoints may not be obtained  in an idle CO  I/M  program, because  in  many cases
all  of the  repairs   necessary  to   allow the vehicles   to  pass   the  stringent

-------
                                       71

 Portland idle HC cutpoints at reinspection will  not  be  needed in order to pass
 looser idle  HC  cutpoints.  One  possible  way  to assure that  adequate repairs
 are performed  is  to  use  a  loose  idle HC  inspection  cutpoint  (1000  ppm)  to
 identify vehicles most  in need  of  repairs  related  to  high HC  emissions,  and
 then  apply  a  more   stringent  idle  HC   cutpoint   (200-300  ppm)   in  the
 reinspection.  This  approach would  be  very  similar to the  one  described  in
 Section 4.4 for separate, idle CO inspection and reinspection cutpoints.

 Using a  special modification of  its I/M- simulation model, EPA has estimated
 that adding a 1000 ppm idle HC cutpoint to an. idle CO I/M; program may increase
 the fleetwide HC benefit  of  the  idle CO program as  of  December- 31,  1987 by as
 many- as. two  percentage   points.   For  example,   Table  25  showed  that  a  20%
 stringent idle  CO program- will achieve  a composite HC  reduction of  27%  on
 December 31, 1987.   If air HC cutpoint-  of 1000 ppm  were added,  this reduction
 would be improved to  291.  This  is  accomplished with a negligible increase  in
 the stringency of  the program, about two to  four percentage points.

 A valid concern related  to a loose  idle HC  cutpoint is the  reintroduction  of
 uncertainty in the repair  process.   In  the  idle CO  I/M program  vehicle owners
 can  be  assured  that virtually all vehicles  can  be  made   to   pass  with  a
 carburetor  idle mixture   adjustment,  possibly with  an  idle   speed  adjustment.
 also.   With the addition  of  even a.  very loose idle  HC  cutpoint,  some  vehicles
 owners will  have  to  rely on  the  repair industry   to  properly  diagnose  and
 repair their vehicle  without- knowing for  certain if the  repairs- were indeed
 necessary to  reduce  their idle  HC   emissions.   The   number of vehicle owners
 with this problem will be small, however,  since  the failure  rate  for  HC will
 be quite low.  In a centralized  idle CO I/M program, owners of  those  vehicles
 which fail a loose idle HC cutpoint  could be  given a choice of  either passing
 the retest of idle HC emissions  or  presenting a signed statement  that  a state
 certified mechanic has  performed specific  ignition   system diagnosis   and  has
 performed repairs  as  necessary.  This would be similar  to  the  waivers  provided
 in  some  I/M programs  after   a  specified "low-emission"  tune-up  by certified
 mechanics.   The suggestions outlined  in Section  3.0  will also help  reduce  the
 incidence of unnecessary  repairs.

.A loose idle HC cutpoint  may have fuel  economy advantages  since  it would often
 identify vehicles with severe ignition system misfire.  Severe (11%)  induced
 misfire on  10  vehicles  in an EPA  test program  decreased  their  overall  fuel
 economy  7.7%.   The  two   vehicles  in the  Portland  study   identified  as  idle
 HC-only  failures  by   a  1000  ppm idle  HC  cutpoint  obtained  an   overall  fuel
 economy benefit of 9.9%  from  repair.  These  instances  indicate  that there may
 be a  significant  fuel economy  benefit  in  the repair of many of  the  vehicles
 which  exhibit  extremely  high   idle HC  emissions.    The   reduction  in  fuel
 consumption for such  vehicles  can  offset: t'-e  costs  of the additional  repairs
 that  will   be  necessary   as   a  result  of  the  addition of  a  loose  idle  HC
 cutpoint.  An 8%  fuel economy  benefit,  for  instance,  provides  about   $72 per
 year in fuel  savings.  Repair costs of the idle  HC  only failures  in  Portland
 range from zero to $207,   averaging $41.  Not all  vehicles  requiring repairs  as
 a result of failing a high idle HC cutpoint are  expected to receive such large
 fuel economy  benefits,  however.  Only  about   half of  the  vehicles failing  a
 1000 ppm HC cutpoint  are  expected to receive  large  fuel economy benefits  from
 repairs.

-------
                                      72

The five  year tons  of HC  emissions  reduced have  been estimated  to be  2,590
tons in  a vehicle  base  of  1  million cars.   If pre-1975  vehicles are  exempt
from the  program the HC benefit will  be  reduced to  1.9%  on December 31,  1987
reducing HC emissions  by 2,420 tons.

Loose  idle  HC outpoints  could also  be  instituted  on a  voluntary basis.  An
idle CO  I/M  program could  easily measure  and report to  the vehicle owner  the
idle HC  levels of  each inspected  vehicle.   Vehicle owners whose vehicle's  idle
HC  levels were  extremely  high  could  be  advised  that  their  vehicle's HC
emissions  were  above  normal  and  their   vehicle   probably  was   in need of
maintenance  in  addition  to  a  carburetor  adjustment.    These  repairs  will
usually  involve  ignition  problems.   It could  be  expected  that  some vehicle
owners  would voluntarily  seek emission  related repairs  and  reduce  their HC
emissions.

Generally, the advisory HC limits  could  be considered  as  a service offered to
vehicle  owners  warning  them  when  their   vehicles   are  in  need   of repairs.
Emission  reductions  from   this  method  would  be  sporadic  and very hard to
quantify  since all repairs would be  voluntary  and  no idle  HC retest would be
required.  However,  some  vehicle  owners  would  have  repairs performed and  some
HC  benefits   would  be gained  without  significantly  increasing   the  overall
program costs.

7.5 Tampering Checks

7.5.1 Background

Another means of obtaining additional HC reductions  in an idle CO I/M program
is  to   perform  a  visual  check  of various  emission  control  systems  in
conjunction with the vehicle's  idle  CO test.   The  purpose is  to identify  those
vehicles which have  had one or more  of their emission control  systems disabled
or  removed.   Requiring that these  vehicles have their emission control systems
restored  to  proper  operation  can result  in additional  HC  (and  CO) emission
reductions.

Of  equal  importance, performing a tampering  check  as part  of the I/M process
can  discourage  new  instances  of  tampering.   This  deterrence   value   of  a
tampering check  can best  be understood  in qualitative,  "common-sense"   terms.
That is,  it  makes  sense that  vehicle  owners  will  be  less  likely  to  remove or
disable  any  of  the  emission control  systems on their  vehicles  if  they  know
Chat the  presence  of some  or  all  of those systems will  be  checked in  the I/M
program.

There  are,  however,  several drawbacks to  performing  tampering  checks  as part
of  the  I/M  test.   First,   the  time  required to properly perform  a  tampering
check slows  down the inspection process  and increases the manpower requirement
of  the I/M station.  This,  of course, adds to the cost of  the. I/M  program.

Second,  while most manufacturers equip  at least some of  their vehicles with
the emission control  systems  which  are  suggested  below  for  inclusion  in the
tampering  check,  they  do  not  design  those   systems  uniformly.   There  is
significant diversity  in what  the  various  systems  look like and where they are

-------
                                       73

placed on  the  vehicle.  This  will  require that  the  inspectors performing  the
tampering  check  receive special training.   In addition, there  will  be a  need
to determine which  vehicles were originally  equipped  with  the system  in  order
to  accurately  determine  whether   it  has  been  removed.   For  example,   many
vehicles from  the  1975-1979 model  years were  not  equipped  with air pumps..   In
order  for an  air  pump  tampering  check to. be made on  these  vehicles,  it  is
necessary  to  separate  the; vehicles  originally equipped  with  air- pumps  from
those  not  so equippedv  It should,  be  noted  that  these- considerations and  the
difficulties, they present apply especially to decentralized programs.

Third, most  of the emission: control  systems   recommended for- inclusion in  the
tampering-  check  have  been observed: in the  field to have  relatively  low  rates
of tampering (between  1-6Z)..  This  means  that the number of  cars  failed in  an
I/M program  on the basis  of the tampering check  will  likely  be  small.  This
magnifies any  added inspection cost per detected disablement.

Fourth, the replacement cost for some  of  the  components  on some systems can  be
high   (e.g.    catalyst   replacement).    This   is,   of   course,   particularly
undesirable in. an I/M program designed with the goal of minimizing costs.

In  light  of these-  considerations,  tampering, checks  should  be limited., to  an
inspection of:  those, systems with  a^ significant HC effect  (based on  both the
known  incidence  of tampering  with  the system and  the  per—vehicle  effect  of
such tampering on HC emissions).  In  addition, systems should be included only
if they are easy to inspect, relatively cheap to  repair and/or have a valuable
deterrent  effect.   The  following sections will discuss  the  systems  which best
meet  these  criteria:  air  pumps,   evaporative  emission control  systems  and
catalysts.

It  is  worth   noting  a  fifth  drawback  of a tampering  check.   This   is  the
confusion  it can cause  among vehicle  owners.   An owner  may  not understand why
it matters  if  his  older car fails  the tampering  check if the  idle  test shows
that the  car  has  low  emissions anyway.   It  can  be  difficult to  explain the
different  purposes  of  the  tampering check  and the idle  test  to  a.  layperson  in
the noise  and  hurry of  the  I/M station.  Therefore, EPA  suggests  that passing
the tampering  check be a prerequisite  to  receiving a valid  idle test.   In this
way, the owner of  a tampered vehicle  will  not receive the idle CO  score  from
his or her vehicle  if  it is below  the cutpoint as this would  only  confuse him
or  her.    There  is  no  harm in giving  the  owner  a  failing  idle  CO  score,
however,  and this may  save  him or  her some inconvenience if  the  repair of the
tampering  and  Che carburetor adjustment are  performed  in Che  trip  Co  a repair
facility.

One way  to  take advantage  of  the  deterrence value  of  tampering  inspections
without the drawbacks  associated with  inspecting  all  vehicles  is  to  develop a
method to  randomly  select a  sample  of vehicles to be  inspected  for  tampering.
Only  the  vehicles  sampled  would  be  checked for  tampering  and  all  other
vehicles would only be  given  the  emissions  test.  This would  greatly reduce
the  resources  needed  to conduct  tampering   inspections.   These few  vehicles
could  also  be  more  thoroughly  examined than  in  a  program which mandated  that
all vehicles be  inspected  for  tampering  and  therefore would be  somewhat  more
likely  to  discover any disablements.   The   expected  result  would  be  less
tampering overall Chan without any  Campering  check since  vehicle  owners may be

-------
                                      74

less likely to allow  tampering with  their vehicle's  emission controls if there
is a chance, however  small, that  they will  be detected.   Likely there would be
more tampering than  with an inspection of  every vehicle, since  at  least some
owners may prefer  to  tamper and  accept  the low  risk of  being selected for the
inspection.  If  all  owners did  this,  there would be  no deterent  effect from
the checks.

This random selection tampering  inspection has  drawbacks of its  own, however.
(1) The  program  as a whole may  be  criticized as being  "arbitrary"  and,  since
only some  vehicle owners  are  singled  out for  the  tampering  inspection,  may
criticize  the program  as  not  equitable.    (2)  If  the number  of  tampering
inspections performed, is small,  and  if  the associated  penalty for detection is
only the repair cost, many  vehicle owners may feel it  is cheaper  to take their
chances  on being  selected and  discovered  rather  than  correct  any existing
tampering or refraining  from future tampering.

For this analysis  each tampering check it  is assumed  that  once  all  instances
of  disconnected  or  removed emission  control  components  are discovered  and
repaired in the  first year of  the program (e.g., a program  starting  in 1983),
only a  few isolated  instances  of  tampering  will be  discovered in  following
years as  vehicle  owners  become  aware  that their systems  are being checked.
This does not reduce  the HC.and  CO emission benefits of  the check,  but limits
the repair costs to  the  cost of  a one-time fix of all tampered systems  at  the
start of  the  program.   Any increase in  the  inspection costs are  considered
negligible.

7.5.2 Air Pump Checks

The purpose of the air pump is  to supply  air to  the engine's  exhaust in order
to promote  the oxidation of EC and  CO  to harmless  by-products.  The  air pump
performs this  function  on  both  catalyst  and  non-catalyst  vehicles.    The  air
pump is driven by means  of  a belt which transmits power  from the  crankshaft  as
it rotates.  This  method of powering the  air pump is  the same as that  used  to
run the alternator and air conditioner  compressor.  The  air  pump  can  therefore
be  found  near or  on  the   same  plane  as   the alternator or  air  conditioning
compressor.  Its plumbing distinguishes  it.

The percentage of  vehicles  equipped  with  air  pumps  varies by  model year.   The
percentages presented in Table  28 are the  ones used  in this  analysis.
                                   Table 28

               Percent of Various Model Year Groupings Equipped
                                With Air Pumps

               Mode 1 Year Grouping        Percent Equipped With Air Pumps

                    1968-1975                           100%
                    1975-1979         •                   40%
                    1980                                100%
                    1981*                                95%

-------
                                       75

There are  three main ways the air pump can  be  tampered with.  First, the  belt
which drives  the  pump can be removed.  Second,  the  entire  unit — pump,  belt,
hoses, and  even  mounting brackets —  can be removed.   Third,  the output  hose
from  the  air  pump  can  be  disconnected  and/or  the  air routing  valve  can  be
tampered with.   This last form of tampering results in  the  air pump spinning
freely and  no  air being supplied, to  the exhaust.  AIL  three- of these forms  of
tampering can  ba  identified  by trained  inspectors  in- an I/M: lane..

The repairs^ necessary for  these  various- forms  of tampering are  self-evident.
In most cases-,, repair can. be- accomplished  by simply installing- a  new belt  or
reconnecting- a hose.  An- average  repair- cost of $20  has been assumed for  this
analysis.

The rate of air  pump tampering used  in this analysis  is that  6.6% of the air
pump-equipped  fleet  has  a tampered air pump at any one time.  This rate comes
from  an   in-use   surveillance   program  conducted   by  EPA's   Mobile  Source
Enforcement Division in  1978  [22].*

While various  surveillance  programs  have reported  various rates  of  air  pump
tampering,   the  rate  from   this  program  was  chosen  since  the  surveillance
techniques  employed  in  the  program  should,  have resulted  in. the  most  random
vehicle sample from among the. various  programs.  In addition,  the time spent
in inspecting each vehicle for tampering  was roughly what might be spent in an
operating I/M program.

The HC  and  CO  emission  increases  which  accompany air  pump tampering  for
1975-1979  model   year vehicles  were  quantified  by examining data  from. 11
vehicles  (1975-1979 model   years)  tested  with  and  without  their air  pumps
operational.   Nine  of.  these  vehicles  came-  from   the  300  car  Restorative
Maintenance  program  [12].   The  other  two   came  from  a  test  program  which
examined   regulated   and;   unregulated  exhaust   emissions   from   catalyst
vehicles[23].   These  data   indicate   that   the  average  HC  emission   level
increases  1.2  g/mi  upon air pump tampering  and  the average  CO  emission level
increases   28.0  g/mi.  COne  source  of  uncertainty  in  Che  analysis  has   to do
with  the fact  that  the  11 vehicles used  to  determine  the  emission  effects of
air pump tampering were  all  in  tuned-up condition.   The emission increases due
to air pump tampering on vehicles in  less perfect condition  may vary.)

There  is some  uncertainty as to  the  HC and  CO  effects of  air  pump  tampering
for pre-1975  model  year vehicles.  However,  these vehicles  contribute  only  a
very  small  share  of  the fleet's emissions  at  the  date of  interest  to  this
analysis (December  31,  1987^,   They  were assumed  to show  the  same percentage
affect due  to air pump  tampering  as   1975-1979  vehicles.   This assumption is
reasonable  and due to  the  small  contribution made by these vehicles,  does  not
significantly affect the analysis.
* The  rate  of 6.6%  is  an overall rate  for the air  pump  system and does  not
appear in Reference  22, which  gives  rates only  for individual components  of
the air pump  system.   The rate of 6.6%  was determined  from the original  data
base used to prepare Reference 22.

-------
                                      76

For 1981 and later model year  vehicles,  Che effects of air pump  tampering  were
quantified  by  examining the results  of EPA  laboratory programs  which took  4
vehicles representative of  1981  and later  technology  and  tested them  with  and
without  their  air  pumps  operational.   In  addition,  one  representative  1980
Ford vehicle tested  in  an  EPA  surveillance  program  in California was  found  to
have  its  air  pump disabled due to  having one  of  the  vacuum  control hoses
kinked  closed.   This vehicle  was  tested   as-received  (air pump  disabled)  as
well as after having  the  air pump  repaired (vacuum hose  unkinked).  Data  from
these five vehicles  indicate that the average EC  emission level increases  0.5
g/mi upon  air  pump tampering  for 1981  and later  vehicles and  the average  CO
emission level increases 15.0 g/mi.

No  comparable  test  data  are  available for  1980   model  year  vehicles.   1980
model year vehicles were  assumed  to  have  the same emission effects  for  air
pump tampering as  1981  and  later vehicles.  This  is because  the 1980  emission
standards  (0.41  g/mi  HC;  7.0 g/mi  CO)  are closer  to  the  1981 standards  (0.41
g/mi HC;  3.4  g/mi CO)  than to  the  1975-1979  standards (1.5  g/mi  HC;   15  g/mi
CO).

Given the  assumed  rate  of air  pump tampering (6.6%)  and given  the  emission
increases  which  accompany  air   pump  tampering  for  the  various  model   year
groupings,  fleetwide  emission  benefits and  costs  of identifying and repairing
tampered  vehicles  can  be  calculated.    It  is assumed  that all  vehicles  with
tampered air pumps are identified and repaired.

The  calculations  involved   in  determining  a fleetwide  benefit  will  not  be
presented.   Basically   the  procedure  involves   calculating  the  emissions
projected  to be  contributed by  the various  model  year groupings (i.e.  pre-75,
1975-1979,  1980,   1981  and  later)   on  December  31,  1987.   The   emission
reductions  which would  result from the repair of vehicles with  tampered  air
pumps is  then  calculated  (also  within each model  year grouping  at a  December
31,  L987  date).  The  final  step  involves  calculating  a  fleetwide   percent
benefit  by  taking  the  emission  reductions due  to  repairing  vehicles   with
tampered  air  pumps and  dividing it  by the  total  (non-I/M)  emissions  of  the
fleet.   Table  29  presents  these figures  of percent  benefit  for  both  HC  and
CO.  For  an example,  Table 25 showed  that an idle  CO I/M program with a  20%
stringency will  achieve a  composite HC reduction of  27%  on December 31, 1987.
If  an  air pump tampering  check  were  added  to   the  program  the  total HC
reduction would  be about 29%.  The  total additional five  year program cost has
been  estimated  to  be  SO.92 million  for  a fleet  of one million vehicles.
Exempting  pre-1975 vehicles  will  reduce   the  benefits.    These  benefits  are
presented  in Table  10 in Section  1.2.

-------
                                       77

                                   Table  29

           Composite Percent Benefit due to Air Pump Tampering Check

                                                   Additional  Five  Year
                                                   Emission  Reduction *
Pollutant   December- 31. 1987 Percent  Benefit    (Thousands of  Tons  Removed)

   HC                      1.5*                             2.56
   CO       :              4.33T                             68.0
  A vehicle base of one million vehicles is assumed.
7.5.3 Evaporative Emission Control System Check

The evaporative, control,  system la  intended, to capture the gasoline fumes which
are:, naturally given, off  wherever gasoline is  stored, and used*  These fumes are
made-  up  of pure  hydrocarbon-  (HC)  emissions.  Although  not. emitted  from the
exhaust  they  represent  a   significant  portion  of  a  vehicle's  total  HC
emissions.  The  evaporative  control system  captures the  fumes given  off by
both  gasoline  in  the fuel  tank and  the gasoline  present .in  the carburetor
(early systems dealt  only with evaporative losses  from  the fuel tank).  These
fumes are stored in a charcoal cannister mounted in the engine compartment and
then  routed to the  engine for  burning at appropriate  times.

Especially  for  early model year vehicles,  on which  evaporative  controls were
first  introduced,   tampering with  the system has been  observed to  be  fairly
common.   The  1978  Tampering  Study(22]   described  earlier  found  8.2%  of
1973-1974  model  year  vehicles had  tampered   evaporative  control  systems  and
1.3%  of  1975-1977  model  year  vehicles  had  been  tampered.   This  tampering can
take  the  form  of disconnected or cut  hoses,  missing cannisters  or removal of
the entire  system.   Once again, these  forms  of  tampering  are  identifiable by
trained inspectors.   An  average  repair cost of $10 has been assumed.

The calculation  of the  emission benefit  due  to  the  identification and  repair
of  vehicles with  tampered evaporative  control   systems  was  performed  in  the
sane  way  as  for  air  pump  tampering.   The   fleet  was  broken  down  into
appropriate model  year  groupings,  an  HC emission increase  due  to  evaporative
tampering and a  tampering  rate was  assigned  to each  grouping and the fleetwide
benefit  on  January   1,   1988  was   calculated.    Since  evaporative  control
technology  has  evolved along  a  different  timescale  than air  pump  or catalyst
technology, the model year  groupings used for the evaporative analysis differ:
pre-1971, 1971, 1972-1974, 1975-197"?, 1978-1980,  1981 and later.

-------
                                      78

The  emission  increases  assigned  Co  each  grouping  to  represent  a  tampered
system  come  from  an  EPA  emissions model  —  MOBILE2  —  which  models  the
emission contributions ("both  evaporative  and exhaust) from all model  years  and
classes of  mobile  vehicles  (e.g.  passenger  cars,  trucks,  motorcycles).   The
assumption  used  to determine  the  increase  in  emissions  due  to  tampering  was
that any tampering would return the  vehicle to uncontrolled  levels  (pre-1971)
of evaporative HC.

The  rates   of  tampering  used in  the  analysis  were that  8.2%  of  1971-1974
vehicles  and  1.32 of  1975  and  later  vehicles  are  tampered.    These  rates
reflect  the  data  from  the  1978   Tampering  Study[22].   A separate  study
conducted  in  California  confirmed  the  large  difference  in  tampering  rates
between pre-1975 and  1975  and later vehicles,  although  the  exact rates were
not duplicated.

Given  the  above  model year groupings,  tampering rates  and emission  increases
due  to  tampering,  a figure for  cost  and  percent benefit on  December 31, 1987
can be  calculated.   Once  again,  this calculation assumes that all  instances  of
tampering are  identified  and  repaired.  Table  30 presents the HC benefit  which
results.   (Obviously,  since  evaporative   systems  are only concerned  with   HC
emissions  there  is no CO  benefit.)   Five year  program  costs  are  expected  to
increase $0.25 million.*   Exempting  pre-1975 vehicles  from  the  program will
reduce  these benefits to  1.2%, reducing emissions of  HC by  1,430 tons.


                                   Table 30

            Composite  HC  Benefit  Due  to Evaporative  Tampering Check

                                                   Additional  Five  Year
                                                   Emission Reduction *
Pollutant   December 31, 1987  Percent Benefit   (Thousands of  Tons  Removed)

   HC                     1.3%                             1.94
* A vehicle base of one million vehicles has been made.
It  is  important  to  note that  this  benefit  refers  to the  comparable percent
reduction of  the  fleet's  exhaust HC emissions reprasented  by  the  reduction in
evaporative HC emissions.  As  described  earlier,  the total  HC  emissions from a
vehicle  come  from both  exhaust HC  and  evaporative  HC.   The  percent benefit
presented in Table 30  is  the  gram-per-mile  reduction in evaporative. HC divided
by  the  fleet's  exhaust HC  (in gram per  mile)  in  the  absence of  I/M.   While
most  discussions  of  I/M  emission  reductions  deal  only  with  exhaust  HC
emissions, it is  equally  valid to examine  reductions  in  evaporative  HC.   This
is  because  the  ultimate  goal of  the I/M program  is to reduce  total  emissions
(exhaust  and  evaporative).   Since  I/M  policy  statements  have  traditionally

-------
                                       79

been expressed  in  terms  of percent reduction  in exhaust HC, the HC  reductions
due  to  an  evaporative  tampering  check  should  be  expressed  in  terms  of  the
comparable  reduction  in exhaust HC  as  has  been  done  in  Table  29.   The  HC
benefit in  Table  30 can  therefore  be  added to  the  HC  emission  reductions  for
idle CO I/M programs in  Table  25.   For example, the 27%  HC benefit for a  20%
stringency  idle CO I/M program  can.  be  increased to about 28% with  the addition
of an evaporative emission control  check.

7.5.4- Catalyst Removal Cheek

As  is  well known,  automotive-  catalysts  lower  HC  and CO  emissions  in  the.
exhaust by  catalytically promoting- the.  oxidation  of  HC  and  CO  to harmless
by-products.  (Catalysts  on most 1981  .and later  vehicles  also  help reduce  NOx
emissions.)   Catalysts  are normally mounted  on the  underside  of  the vehicle,
along the exhaust  pipe and before  the muffler.  Some  vehicles  have catalysts
mounted inside  the  engine compartment.   If  a  catalyst  is  not  observed  by
checking underneath  a  1975 or  later model  year vehicle,  it will  be necessary
to-  open  the engine  compartment hood and either  locate the catalyst, there  or
confirm from, the emissions label put. on. every vehicle that the vehicle was  not
equipped with a catalyst  at the  factory.

Tampering with  the  catalyst takes  the  form. of. simple: removal  of  the catalyst
and  replacement with  a   straight exhaust pipe.   Since- this  is. very  easy  to
detect,  it  was  assumed that all instances  of catalyst  removal  are  identified
through an I/M tampering  check.

Repair of   this  form of  tampering  obviously  requires   installation  of  a   new
catalyst (or reinstallation of  the  old one  if it was saved).   This could be a
relatively  expensive repair.   It was  decided,  however,  to  include it  in this
report since many  states might  be  interested  in  checking  for  catalyst removal
since it is such a  flagrant  form of tampering.   In  addition,  a  catalyst check
has a valuable  deterrent effect which  costs  nothing.   Hew  catalysts  now cost
between $172  and  S320,  most of which  is  dealer  and  distributor  markup.   A
market for  lower-priced  used catalysts  may  appear,   if new  catalysts  are  not a.
requirement  of  the  program.   In any event,  lower-priced replacement catalysts
are possible if any demand is created  by  a  catalyst check.   An average cost  of
$200 per catalyst has been assumed  in this analysis.

The rate of catalyst  removal used  in  this analysis was the rate  observed  in
the 1978 Tampering survey: 1.4%[22].   This  rate  was  applied  to  the entire 1975
and later fleet.

The  HC   and  CO  emission  increases   f.;hich  accompany   catalyst  removal  were
determined  by  examining  Che  engine-out (before  the catalyst)  emissions  of a
number of vehicles  involved in several misfueling  test programs.   Before   the
vehicles were  misfueled,  they   received  both  baseline  tests  (all components
functional)  and tests  with  the  catalyst removed.  -By comparing  the results of
the  two  tests  the  percent increase  in  emissions  which  accompanies  catalyst
removal   can  be calculated.    Four  vehicles  from  the  1975-1979  model  year
grouping and  six  vehicles  representative  of  the  1981  and  later model  year
grouping were tested.  1980 model year vehicles  were assumed to  have  the same
percent  increase as  1975-1979  vehicles.  This was  done because  the  catalysts

-------
                                      80

used on 198C vehicles  are  more like Chose used  on 1975--79 vehicles  Chan  Chose
used  on  19<31   and   later  vehicles.   The  figures  of  percent  increase   were
combined  wich  the  projected  zero-mile  emission  levels  of  Che  various  model
year groupings  in  order  Co calculate  Che following  figures of  gram-per-mile
increase:  3.06 g/mi HC and 24.16  g/mi CO  for  1975-1979;  0.67 g/mi HC and  7.36
g/mi CO for 1980; 0.93 g/mi HC and  7.11 g/mi CO  for  1981  and  later.

As was  done for  Che air  pump and evaporative  control  analyses,  the rate  of
catalyst  removal,  the emission  increases due  to  catalyst  removal,  and  the
relative  contributions of  the various model  year groupings were combined  to
calculate figures of percent  benefit  at  January  1,   1988. "  Table 31  presents
Chose figures.   As  with  the benefits  presented in Tables  29  and  30,  Che  HC
benefit in Table  31 would  be added  to  the HC  benefit in Table 25,  for example,
to find Che overall  benefit  of an idle  CO I/M program chat includes  a catalyst
removal check.   Total five  year  program costs  will  be  increased  by about  $2
million.*
                                   Table 31

            Composite Percent Benefit Due Co Catalyst Removal Check

                                                 Additional  Five Year
                     Benefit on                  Emission  Reduction *
Pollutant         December 31, 1987            (Thousands of  Tons Removed)

   HC                   0.9Z                               1.35
   CO                   0.5%                             10.7
One alternative way  Co  implement  a  catalyst check which helps avoid Che  repair
cose  problem  mentioned  earlier  would  be  Co  restrict  Che check   co   Chose
vehicles sold  after  Che I/M program  begins.   That is,  vehicle  owners who had
removed  Che  caCalysC  on  Cheir  earlier  model  year  vehicles  (before  I/M
sCart-up)  would  not  be  required  Co  install  a  new  catalyst, while  owners of
newer model year  vehicles would.   This approach,  if  coupled  with an  effective
public awareness  program,  should  provide  an effective natural deterrent.   Many
vehicle  owners would be dissuaded from  removing Cheir  catalyst  under such an
approach and  those  chat  removed  their  catalyst anyway and  were subsequently
identified  in  an I/M program  would  at least  have  been forewarned.   There is
noc  expected  Co  be,  Cherefore,  any  significanC  addicional  program   cost
associated wich this program.
* A vehicle base of one million vehicles is used.

-------
                                      81

This approach does,  of course,  reduce  the  emission benefits associated with  a
catalyst check.   Table 32 presents  the emission benefits  which  result if  the
catalyst  check  is   restricted  to  1983-1988  model year  vehicles  (assumes  a
January 1, 1983 I/M  start-up).
                                   Table. 3Z

           Composite.' Percent Benefit Due to a Catalyst: Removal Check
                     Restricted to 1983 and Later Vehicles.
                                                 Additional Five Year
                       Benefit on                Emission Reduction *
Pollutant           December 31, 1987         (Thousands of Tons Removed)

   HC                     0.3%                           0.20
   CO                     0.2%                           1.6
* A vehicle base of one million vehicles is used.
7.6 Inspection and Maintenance for Light Duty Trucks

Light duty  trucks  (LDTs) such  as  pickups,  vans,  and  light  delivery vehicles
are  susceptible  to  the  same  types  of  maladjustments  and  emission component
failures as most  pre-1981 passenger  vehicles.   The  HC and CO  emissions  from
LDTs are a  significant fraction  of  emissions from all  mobile  sources in urban
areas.  While EPA policy does  not  require inspection of LDTs  in I/M programs,
any emission reductions from repairs of  LDTs  in an I/M program will contribute
towards attainment  of air  quality  standards,  just  as do  emission reductions
among  passenger  vehicles   (LDVs).   Therefore,  an  idle  CO  I/M   program  can
include LDTs and use  the HC  emission  reductions from those additional vehicles
to improve the program's  overall HC emission reduction benefits.

The  most  obvious  advantage to  this  approach   of enhancing  an  idle  CO  I/M
program is  that  it  will require  fewer  additional  design  and  administrative
complications than some  of  the other  HC  benefit enhancements  in this section.
In addition,  the  added HC  benefits  will generally  be  larger than  any  of  the
other HC enhancements  since  Che additional  HC  and CO  emission  reductions  are
derived by  testing more  vehicles rather  than  by squeezing  more  benefits  fron
the same vehicles.

LDTs  can  be  broken  down into two major subgroups  which  will be considered
separately.  LDTs with gross  vehicle  weights  (GW)  below  6000  pounds  will  be
refered to  in this  report  as LDT1.   All LDTs  between 6000  pounds and  8500
pounds  GVW  will be  LDT2.   In addition,  in order  to  discuss  the  effects  of
better  tests   and   better   repairs,   the  technology   level  of   each   LOT
corresponding  to  an  equivalent  LDV   (passenger   car)   technology  will   be
determined.   Table  33 presents  the  technology levels  and the  corresponding
model vears of LDVs,  LDTls,  and LDT2s  which correspond to those levels.

-------
                                      82

                                   Table 33

                       LOT Technology Level Equivalents
Technology                  '          Model Years Included
  Level                      LDV               LDT1              LPT2

Pre-Catalyst.               Pre-1975          Pre-1975          Pre-1979
Oxidation Catalyst         1975-1980         1975-1984       "  1979-1984
Three-Way Catalyst .        1981 and Later    1985 and Later    1985 and Later
LDV:  Light duty gasoline vehicles
LDT1: Light duty trucks below 6000 pounds GVW
LDT2: Light duty trucks between 6000 and 8500 pounds GVW


The  additional  HC  and CO  emission benefits  in tons  of pollutants  chat are
possible  by  including LDTs in  an idle CO  I/M program depend  on the emission
reductions per  LDT,  the  number of  LDTs  in  the  I/M area,  and  the  number of
miles  they drive.   Table  34  presents  the  HC and  CO  emission benefits  in an
idle  CO  program  among  the  two  groups  of  LDTs  with  and  without  other
enhancements.

-------
                                      83
                                   Table 34

                     HC Reductions Among Light Duty Trucks
                           in an Idle CO I/M Program
LDT1 (0-6000 pounds GVW):
                         HC
Stringency for
Pre-1985 Vehicles
(Failure Rate)
      13
      20
      30
      40
LPT2 (600Q-85QO pounds GVW):
       Reduction on December 31,
Idle CO Test         With
   Alone         Better Test *
         CO
HC.
CO
29.6
32.8
36.8
38.1
35.4
40.1
44.5
45.7
30.0
33.2
37.2
38.5
36.2
41.0
45.4
46.6
Stringency for
Pre-1985 Vehicles
(Failure Rate)
      13
      20
      30
      40
       Reduction on December 31,
Idle CO Test         With
   Alone         Better Test *
                         HC
         CO
HC
CO
22.1
24.2
27.3
28.2
32.1
34.9
37.8 -
38.5
23.2
25.3
- 28.4
29.3
34.6
37.4
40.3
41.0
1987
      With
  Be'tter Repairs **
     HC     CO

     31.2   40.1
     35.3   45.7
     40.6   49.1
     42.3   49.5
1987
      With
  Better Repairs **
     HC     CO
                                      23.3
                                      25.8
                                      29.7
                                      30.8
                           34.9
                           38.0
                           40.3
                           40.8
* Better Test: Two-Speed Idle Test or Loaded Test for 1985 and later LDTs.
**  Better Repairs:  Full  mechanic  training  or more   stringent  reinspection
cutpoints for Pre-Catalyst and Oxidation Catalyst LDTs.
The  emission  reductions   from   LDTs   in   Table  34  can  be  converted  into
passenger-vehicle-equivalent  HC  and  CO  emission  reductions  and   applied  to
overall  passenger vehicle  emission reduction  goals.   This  is done  by first
calculating the gram  per mile  reduction in the emission  factors  for HC and CO
from  I/M on  December 31,  1987.   Next  these reductions  are weighted  by  the
vehicle  miles  traveled of  LDTs  versus  passenger  cars  in  Chat  year.   Finally
these  weighted  reductions  are divided  by  the  appropriate  passenger vehicle
emission  factors  for  the  case  without I/M  giving the  additional  HC  and  CO
benefits  from  I/M   for   LDTs.    Table  35   presents   the  results  of  such
calculations using  standard national  LOT densities  and  mileage accumulations.
Using  the Two-Speed  Idle  or Loaded Test for  1985  and  later LDTs  provides only
an additional 0.1% benefit  for both HC and CO emissions.

-------
                                      84

                                   Table 35

                         Passenger-Vehicle-Equivalent
                   Emission Benefits from Light-Duty Trucks
 Stringency for            Additional Percent Benefit on December 31,  1987
Pre-1985 Vehicles                for Passenger Vehicle Idle CO I/M
    (Percent)                 Idle CO Test             With Better Repairs
                              HC        CO                 HCT       CO

       13                     4.2       4.4                4.3       4.9
       20                 .    4.6       4.9                4.9       5.6
       30                     5.1       5.4                5.6       5.9
       40                     5.3       5.6                5.8       6.0
For example,  a  basic idle CO I/M  program from Table 25  in  Section 6.0 with a
13%  stringency  can increase  the  program  HC  benefits  by  4.2%  and  the  CO
benefits by 4.4% for passenger  vehicles  by  including  light-duty trucks in  the
I/M program.   This  would  increase the  benefits  of a  13% stringency idle CO
program  from  24.22  to  37.5% for  CO.   The  five  year  program costs  and tons
removed  for  this option have been estimated in  Section 1.2  and  presented in
Table 7.  Exempting  pre-1975 LDTs  from  the program will reduce both the HC  and
CO benefits.  Table 12  in Section  1.2  presents  the benefits  and  costs of  LOT
I/M exempting pre-1975 vehicles.

-------
                                      85

                                  References

1.  "Inspection/Maintenance   Policy."    David  G.  Hawkins   to  EPA   Regional
    Administrators, EPA memo.  July  17, 1978.

2.  Federal  Register,  44-  FR 2960,  January  12,  1979.   (Parameter Adjustment
    Regulations).

3.  Federal  Register,  45 FR  34802,  May  22,  1980.   "Motor  Vehicles;  Emission
    Control  System  Performance- Warranty Short Tests  and Warranty Regulations;
    Final Rules."

4.  "Update  on the  Cost-Effectiveness of Inspection  and  Maintenance."   Tom
    Darlington, EPA I/M Staff.  April 1981.  EPA-AA-1MS/81-9.

5.  "Update  on  the  Fuel  Economy  Benefits  of  Inspection  and  Maintenance
    Programs."    R.   Bruce    Michael,    EPA   I/M    Staff.     April   1981.
    EPA-AA-IMS/81-10.

6.  "Inspection  and  Maintenance  for   1981   and  Later  Model  Year  Passenger
    Vehicles".   David W-  Hughes,   EPA  I/M  Staff.    June  1981.   Society  of
    Automotive Engineers, 810830.

7.  "Recommendations Regarding  the  Selection of  Idle Emission  Cutpoints  for
    Inspection  and  Maintenance   Programs  Requiring   Only  Carbon   Monoxide
    Emission   Reductions."   Susan   Vintilla,   EPA   I/M  Staff..   May  1981.
    EPA-AA-IMS/81-13.

8.  "A Survey  of Operating  Inspection/Maintenance  Program."   R.F.  Klausmeier
    and O.K.  Kirk,  Radian Corporation,  Austin,  Texas.   April 1980.    No.  DCN
    80-230-146-09.

9.  "Update  on EPA's  Study   of  the Oregon  Inspection/Maintenance  Program."
    James A. Rutherford and Rebecca  L.  Waring,  EPA I/M  Staff,  June  1980.   Air
    Pollution Control Association,  No.  APCA 80-1.2.

10. "Analysis of the Emission Inspection  Analyzer".   William  B.  Clemmens,  EPA
    ECTD, September 1980.  EPA-AA-IMS/80-5-A.

11. "Implementation  Issues  Regarding  EPA  Recommended  I/M  Emission  Analyzer
    Specification."  David  G.  Hawkins  to EPA  Regional  Administrators,   EPA
    memo.  September 24,  1980.

12. "An Evaluation  of  Restorative  Maintenance on Exhaust Emissions of  1975-75
    Model  Year  In-Use Automobiles."  Jeffrey C.  Bernard and  Jane F.  Pratt,
    Calspan  Corp.,  Buffalo,  New  York.    December  1977.   Three  sites,  four
    volumes.  EPA-460/3-77-021.

13. "A  Study  of  Emissions  from   Passenger  Cars   in   Six  Cities."   FY77.
    Automotive  Testing Laboratories,  Aurora, Colorado.   January  1979.   Two
    Volumes.  EPA-460/3-78-011.

-------
                                      86

14.  "FY79  Study  of  Emissions  From  Passenger  Cars  in  Six  Cities."   Two
    Volumes.    Automotive  Testing   Laboratories,   Inc.,   September   1980.
    EPA-460/3-80-020.

15*  "Testing Support for  an Evaluation of a Houston  I/M  Program."   Automotive
    Testing   Laboratories,   Inc.,   Aurora,   Colorado.     September   1980.
    EPA-460/3-80-021.

16.  "Final Results of the One-Week Follow-Up With  Participants  During the FT79
    and  FY80  Emission  Factor Testing  Programs."  EPA  Technology  Evaluation
    Branch, 1981.  EPA-AA-TEB/81-27.

17.  "Effects of  Inspection Maintenance Programs on Fuel  Economy."   I/M Staff
    Technical Report.   Revised June 1979.   IMS-001/FE-1.

18.  "Restorative Maintenance  Fuel Economy Analysis."   Jim Rutherford  to Janet
    Becker, EPA I/M Staff memo.  February 14,  1980.

19.  "Portland Study Fuel  Economy Analysis."  Jim  Rutherford  to Charles Gray,
    EPA I/M Staff memo.   January 28,  1980.

20.  "Derivation of 1981 and Later  Light Duty Vehicle Emission  Factors  for  Low
    Altitude,  Non-California  Areas."   EPA   I/M  Staff  Technical   Report.
    November 1980.  EPA-AA-IMS/80-8.

21.  "Derivation  of I/M  Benefits for  Post-1980 Light  Duty  Vehicles  for  Low
    Altitude, Non-California  Areas."   David  Hughes,  EPA  I/M  Staff.   January
    1981.  EPA-AA-IMS/81-2.

22.  "1978  Motor  Vehicle  Tampering  Survey."   EPA  Office   of  Enforcement.
    November 1978.

23.  "Regulated   and  Unregulated   Exhaust   Emissions   from   Malfunctioning
    Non-Catalyst  and  Oxidation  Catalyst  Gasoline  Automobiles."   EPA  ECTD,
    1980.  EPA-460/3-80-003.

-------