-------
                         Table of Contents
                                                               Page

List of Appendices                                              iii
List of Tables                                                    v
1.0  INTRODUCTION                                                 1
2.0  GLOSSARY OF KEY TERMINOLOGY                                  3
3.0  PROPOSED I/M PERFORMANCE STANDARDS                           5
    3.1  Enhanced I/M Performance Standard                        5
        3.1.1   Low  Option                                        5
        3.1.2   Medium  Option                                      6
        3.1.3   High Option                                        6
    3.2  Recommended Enhanced I/M Program Design                  6
    3.4  Basic  I/M Performance Standard                           7
4.0  EMISSION REDUCTIONS  FROM I/M PROGRAMS                        8
    4.1  Recent I/M Test  Programs (Maryland and Indiana)          8
    4.2  FTP HC/CO Correlation Comparison Between the IM240 and
         the Second-chance 2500 rpm/Idle Test                    12
        4.2.1   I/M  Test  Assessment  Criteria Overview             15
        4.2.2   Detailed  Discussion  of  Correlation and Test
                Assessment                                       19
        4.2.3   Avoiding  Errors  of Commission                     20
    4.3  Approval of Alternative Tests                           27
    4.4  Transient Testing Fast-Pass/Fast-Fail Strategies        29
    4.5  Estimating I/M Testing Credits for MOBILE4.1            31
        4.5.1   Tech4.1 Background and  Assumptions                31
        4.5.2   Evaporative  and  Running Loss Modeling, and the
                Effectiveness  of Purge/Pressure  Testing           35
        4.5.3   Benefits  of  IM240 NOx Inspections                 36
5.0  REGULATORY  IMPACT  ANALYSIS  -  ESTIMATING  COST  AND  COST
EFFECTIVENESS                                                    45


Draft                         -i-                       2/26/92

-------
    5.1  Cost of Conventional I/M Testing                        45
        5.1.1   Inspection and Administration  Costs               45
    5.2  Estimated Cost of High-Tech I/M Testing                 47
        5.2.1   General Methodology
                                                                 AQ
        5.2.2   Equipment  Needs  and  Costs                         *°
        5.2.3   Cost  to Upgrade  Centralized Networks              50
        5.2.4   Cost  to Upgrade  Decentralized  Programs            56
    5.3  Repair Costs                                            58
        5.3.1   HC and CO  Exhaust Repair Costs and Methodology   58
        5.3.2   NOx Repair Costs and Methodology                  60
        5.3.3   Evaporative System Repair  Costs and Methodology  62
    5.4  Fuel Economy  Benefits                                   64
        5.4.1   Fuel  Economy Benefits  of Evaporative  System
                Repairs                                          64
        5.4.2   Fuel  Economy Benefits  of IM240 Repairs            65
        5.4.3   Fuel  Economy Benefit for the 2500  rpm/Idle Test  66
    5.5  Recurring Failure and Repair Rates                      68
    5.6  Method for Estimating Cost Effectiveness of I/M  Programs?!
        5.6.1   Inspection Costs                                 72
        5.6.2   Repair Costs                                     72
        5.6.3   Fuel  Economy Cost Benefits                        73
6.0  REGULATORY IMPACT ANALYSIS  -  COSTS  AND  BENEFITS  OF ENHANCED
I/M OPTIONS                                                      75
    6.1  Emission Reduction Benefits                             75
    6.2  Cost Effectiveness Estimates                            77
        6.2.1   Assumptions and  Inputs  for  the Program  Options   77
        6.2.3   Cost-Effectiveness Calculations                  78
        6.2.4   National Cost  of Choosing  Low  Option  I/M         80
    6.3  National Costs and Benefits                             81
        6.3.1   Emission Reductions                               81
        6.3.2   Economic Costs to Motorists                      83
    6.4  Motorist Inconvenience Costs                            85
7.0  REGULATORY FLEXIBILITY ANALYSIS                             86

Draft                        -ii-                      2/26/92

-------
    7.1  Regulatory Flexibility Act Requirements                 86
        7.1.1   The Universe  of Affected Entities                 87
    7.2  Types of Economic Impacts of Concern                    88
    7.3  Changes in Repair Activity                              89
        7.3.2   Repair Activity in  Future  I/M Programs            90
    7.4  Changes in Emission Testing Activity in I/M Areas       91
        7.4.1   The Existing  Market in Centralized and
                Decentralized Programs                            91
        7.4.2   Future Market in  Enhanced  I/M Programs            98
        7.4.3   Centralized Programs                              98
        7.4.4   Decentralized Programs                            98
        7.4.5   Impact on  Jobs in Decentralized Programs         103
        7.4.6   National Impact on  Jobs                          107
    7.5  Mitigating the Impact of Enhanced I/M on Existing
         Stations                                               109
8.0  ONBOARD DIAGNOSTICS AND ON-ROAD TESTING                    112
    8.1  Onboard Diagnostics, Interim Provisions                112
    8.2  On-road Testing,  Interim Provisions                    112
                        List of  Appendices
Appendix
 A  Cost Effectiveness Model Version 4.1  Source Code Listing
 B  Tech4 Model Version 4.1 Source Code  Listing
 C  Evaporative Emissions  and Running Loss Emission  Factor
    Derivation
 D  Purge and Pressure Test Effectiveness  Figures and Spreadsheet
 E  Regression Analyses and Scatter Plots  for Fuel  Injected 1983
    and Later Vehicles
 F  MOBILE4.1 Technology Distribution and Emission  Group Rates and
    Emission Levels
 G  Exhaust Short Test Accuracy
 H  Data Analyses for Appendix G
 I  Evaporative System Purge and Pressure Diagrams

Draft                        -iii-                      2/26/92

-------
 J Evaporative  System Failures  and Repairs



 K MOBILE4.1  Performance  Standard Analyses, By Option



 L Identifying  Excess Emitters  with  a Remote  Sensing Device




 M Model  Year Failure Rates  by  Test  Type
Draft                        ~iv-                      2/26/92

-------
                     List of Tables
                                                               Page

4-1  IM240 Selection Standards for Stratified FTP Recruitment     9
4-2  Weighting Factors for Correcting Recruitment Biases         10
4-3  FTP, Lane IM240, and "Tank Fuel" IM240 Results              21
4-4  Vehicle 1674 NOx, FE Results Versus Dynamometer Settings    22
4-5  Two-Ways-To-Pass Criteria Applied to Lane IM240 TBI Error-of-
     Commission Vehicles                                         24
4-6  Variables Affecting CO Emissions                            26
4-7  IM240 Bag-1 Fast-Pass/Fast-Fail Analysis                    30
4-8  Short Test Identification Rates                             34
4-9  Short Test Repair Effectiveness                             35
4-10 Lane IM240 Based Emission Factor Levels with IM240 NOx
     Cutpoints                                                   39
4-11 Side Effects of I/M on NOx Emissions                        41
4-12 Lane IM240 Based Emission Factors with IM240 Cutpoints      43
5-1  I/M Program Inspection Fees                                 46
5-2  Equipment Costs for New Tests                               49
5-3  Expendables for New Tests                                   49
5-4  Peak Period Throughput Rates in Independent I/M Programs    50
5-5  Current Program Costs                                       53
5-6  Costs to Add Pressure Testing to Centralized Programs       54
5-7  Costs to Add Proposed Tests to Centralized Programs         55
5-8  Inspection Volumes in Licensed Inspection Stations          56
5-9  Costs to Conduct Pressure, Purge, and Transient Testing in
     Decentralized Programs                                      58
5-10 Average Cost of Repairing Emission Control Components       60
5-11 NOx Repair Costs                                            62
5-12 Average Repair Costs and Fuel Economy Benefits              63
5-13 Zero Improvement Vehicle Sample Size Adjustments            66
5-14 Adjusted Zero FE Benefit Vehicle Sample Size                68
5-15 Exhaust Test Failure Rates                                  69
5-16 Default Inspection Costs in CEM4.1                          72
5-17 Default Repair Cost in CEM4.1                               73
5-18 Fuel Economy Benefits in CEM4.1                             74
6-1  MOBILE4.1 Inputs for Enhanced I/M Performance Standard
     Options                                                     76
6-2  Benefits of I/M Programs Options                            77
6-3  Total Annual Program Cost                                   78
6-4  Cost per Ton Allocating All Costs to VOC                    79
6-5  VOC Cost per Ton Accounting for NOx and CO Benefit          80
6-6  Total Cost and Benefits of I/M Options                      81
6-7  Excess Cost of Choosing Low Option I/M                      81
6-8  National Benefits of I/M                                    83
6-9  Program Costs and Economic Benefits                         85
6-10 Costs of the Biennial High Option including Inconvenience   86
7-1  Affected Businesses                                         88
7-2  Repair Expenses in Enhanced I/M Programs                    90
7-3  Number of Inspection Stations by State                      92
7-4  Inspection Stations by Category                             93
7-5  Inspection Station Volumes and Incomes                      94
Draft                        -v-                       2/26/92

-------
7-6  Average Inspection Station Revenues, Costs, and Profits     96
7-7  Inspection Volumes in California                            97
7-8  Station Revenues and Profits by Volume                      98
7-9  Assumed Station Distributions                              10~
7-10 Revenues and Profits for Low and Medium Volume Stations    102
7-11 Numbers of Inspectors per Station by State                 104
7-12 Estimated Inspection FTE                                   105
                                                                1 (] *7
7-13 Summary of FTE Gains and Losses                            •"•" '
7-14 Impact on Jobs of I/M Proposal                             108
Draft                        -vi-                      2/26/92

-------
1.0  INTRODUCTION

     Despite having the best vehicle control  program  in the world,
many areas  in  the United States  continue  to measure unhealthful
levels  of  air  pollution,  approximately  half of  which  can be
attributed to motor vehicles.   As  a result, in  addition to tighter
standards  on new  vehicles  and their fuels,  the Clean  Air Act
Amendments  of  1990  (Act) require the implementation  of vehicle
Inspection and Maintenance  (I/M)  programs in areas that have been
designated as nonattainment  for ozone or  carbon monoxide.  A total
of  162  such areas  currently exist  in  the United States,  52 of
which do  not presently operate I/M programs.   Depending upon the
severity  of the nonattainment problem,   these  areas  will have to
implement  either  a  basic  I/M program  (required in  areas  with
moderate ozone nonattainment, and  in marginal  areas  with existing
I/M  programs)  or  an  enhanced  I/M  program  (required in  most
serious,  severe,  and extreme  ozone areas,  as well as most carbon
monoxide  (CO) areas registering levels greater  than 12.7 parts per
million  (ppm)).   Seventy-four of  the  162  nonattainment  areas
currently  designated  will  require the  implementation  of  an
enhanced I/M program.

     The  Environmental  Protection  Agency (EPA) has had oversight
and policy  development responsibility for  I/M programs since the
passage  of  the  Clean Air Act in  1970,   which  included  I/M as an
option for  improving air  quality.   The  first such I/M program in
the United States was begun  in New Jersey in  1974,  and the program
elements which made up this  program's design  (i.e., a centralized,
annual,  idle test of  all light-duty gasoline vehicles,  with no
waivers   or  tampering checks)  still  constitute  those  design
features  upon which  the  basic I/M performance standard is based.
However, many advances have been made in vehicle  technology since
the time of that first I/M program, and while the  idle test in use
in  many  current  programs  works   well  enough when  it  comes to
detecting emission  problems  in  older,  low-tech vehicles,  its
effectiveness as a testing  strategy  rapidly drops off as we begin
testing  newer,  more sophisticated, computer-controlled vehicles.
High-tech vehicles  need high-tech testing  which  more  closely
simulates  real-world driving conditions  and the  sort  of test to
which  vehicles  are  originally certified  -  a loaded,  transient
test,  which requires  driving the  vehicle through  a  prescribed
pattern of accelerations  and decelerations on a dynamometer.

     Much  has  also been  learned  since   1974 about the  many ways
vehicles contribute  to the  problem  of air pollution.   Previously,
it  was  thought that the majority  of  the  air pollution problem
attributable to   mobile sources  was   the  result  of  exhaust
emissions;  it  is  now  understood   that emissions  in the  form of
evaporative and running  losses are also  major contributors.   The
gasoline  evaporating in  the tank   of a vehicle and escaping into
the environment is  as  much  a source of   volatile  organic compound
(VOC)  emissions  as  are the exhaust   gases  emitted   from  the
Draft                        -1-                       2/26/92

-------
tailpipe.  Vapor  recovery  and recirculation mechanisms have been
installed on   vehicles  since  1971,  but  these  systems  can
deteriorate with time and  are often  rendered useless as a result
of wear, tampering,  and design defects.  Cost effective tests have
been developed to detect evaporative  system  failures  of this sort,
including the  evaporative  system purge test and the evaporative
system pressure test.

     Under the terms of the Clean Air Act Amendments of 1990, EPA
is  required  to establish  minimum  performance standards  for I/M
programs.    The Act  further  specifies  that  the  standard  for
enhanced I/M shall be based upon a program that employs an annual
cycle of  automated emissions analysis, performed at  a centralized
site, and enforced through the  denial  of registration.   EPA is
currently  in  the  process  of developing  and  formalizing  these
standards in  the form  of a  rulemaking.

     In  the  past,  the model program used  to  establish  the
performance  standard  assumed  a basic program along  the  lines -of
the  original  New Jersey  program  -  a standard  which  remains
essentially  unchanged  for  basic I/M programs.    For  the enhanced
I/M  performance standard,   however,  EPA is  evaluating potential
low, medium,  and high  options,  the last  of which would be based on
loaded,  transient testing,  in  conjunction with evaporative system
purge and pressure tests  (see section 3.0  for  a more  detailed
discussion of  these potential  options).   Using EPA's  MOBILE4.1
computer model, a high-tech I/M program, such as  the one proposed
by EPA's  high  option,  is  expected  to achieve emission reductions
from mobile  sources on  the order of  approximately  31% for ozone-
forming hydrocarbons (HC)  and 34% for CO (compared  to a 5% HC and
16% CO emission reductions  from the basic I/M performance standard
program design).

     Given the potentially significant  economic  impact  of  this
decision,  it is necessary to assess the  costs  and benefits of
enhanced  I/M performance  standards.    This  report  provides  the
technical background information supporting EPA's cost and benefit
projections.

     In assessing the costs and benefits of enhanced I/M,  we will
detail  the  findings of recent research and development  on  test
procedures  and vehicle emissions,   the basis for  the  computer
models  used to establish emission  benefits  and  program  cost
effectiveness, the   differences in cost   effectiveness  among
programs based upon network and test  types, as well as projections
of the  average per vehicle cost for  inspection  and  repairs,  and
the  cost  offset of the  fuel  economy benefit achieved  by making
such repairs.   The  full program listings  for the computer models
used  to assess  the  costs and benefits  of  I/M  program  design
options, as  well  as  other graphic and  tabular support data,  are
attached to this report  as  appendices.
Draft                        -2-                      2/26/92

-------
2.0  GLOSSARY OF KEY TERMINOLOGY

     Throughout this  report several key  terms  will be used  with
which the reader may not be immediately familiar.   To facilitate a
better  understanding  of   the  issues  involved,   the  following
glossary is provided.

"Concentration"  Versus  "Mass  Emissions"  Tests;   Mass emissions
tests provide a much  better indication  of vehicle  emission  levels
than concentration  tests.   A concentration  reading of 200 parts
per million  (ppm)  HC from  a subcompact  car  and the same 200  ppm
reading  from  a  large truck (which  is  entirely possible) suggest
that  the  two  vehicles  pollute  equally.    However,  this  is
incorrect.  The truck will have  a much higher volume of exhaust.
So, over a given one-mile  drive, the subcompact car  may only  emit
50  cubic feet of exhaust  gases,  whereas  the  truck may emit  500
cubic feet.  With both vehicles  emitting 200  ppm HC over the mile,
the  total  amount  of HC emitted  by the  truck will  be  10 times
greater  than  the  amount   emitted  by  the  small   car.   A  mass
emissions test allows the total  emissions per mile  to be measured;
a  concentration test  does  not.   All currently approved I/M tests
are concentration tests.  The Federal Test Procedure and the IM240
test, however, are mass emissions tests.

Error-of-Commission  (Ec):   On the basis  of an emissions test,  the
false failure of a  vehicle  as "dirty"  (i.e.,  emitting high  enough
that repair and a retest are required)  when  the vehicle, in fact,
meets EPA new car standards, based upon the Federal Test Procedure
 (see  definition below) .    Usually,  HC  and  CO EC'S are defined
without regard to NOX emissions, and vice  versa.

Error-of-Omission;   To falsely  pass as clean a vehicle which,  in
fact, exceeds EPA new car standards,  based upon the results  of the
Federal Test Procedure.

Federal Test Procedure  (FTP):  The Federal Test Procedure  (FTP)  is
a  mass   emissions  test  created to  determine  whether  prototype
vehicles  comply  with  EPA standards,  thus  allowing production
vehicles to be  certified for sale  in the United States.  The FTP
has become the  '"golden  standard" for determining vehicle emission
levels,  so  it is also  used to determine the emission levels  of
win-use" vehicles.   The FTP is  too costly to use  for I/M because
vehicles must be maintained in a  closely controlled environment
for over 13 hours.   The FTP is  based on  a 20 minute trip,  driven
once when the engine is cold, and again when  it is hot.

High-Tech  Vehicles:   Vehicles  with computerized  control   of  the
engine  and emission control  system,   especially   1983  or newer
vehicles employing  fuel injection  (either port fuel  injection  or
throttle-body injection)  as opposed  to  carburetion as  a  fuel
metering methodology.
Draft                        -3-                      2/26/92

-------
Idle  Test:   A  concentration-type emission  test to  measure the
percentage of CO and ppm  HC  in  the exhaust stream of a gasoline-
powered vehicle  operating at idle.   The  nondispersive infrared
detector  (NDIR)  equipment normally  used  gives  a  less accurate
measure  of HC  than does the  flame ionization detector   (FID)
equipment used in the FTP  and IM240 tests.

IM240   Exhaust   Test!    A mass  emissions   (as  opposed  to
concentration),  transient  short  test  run on an inertial  and power-
absorbing  dynamometer  using  a  240 second driving  cycle  loosely
based  upon the  LA-4  cycle  used in the  FTP.   EPA  has  usually
divided the  driving cycle into  2  parts or  "bags"  with separate
emissions  determinations, but   other  treatments   are  possible.
Unlike  the idle test  which  is   conducted  at a  single  speed and
expresses emissions in terms  of  percentages and ppm, the ^IM240 is
conducted  at  a  range  of accelerations   and  decelerations and
provides emissions  measurements  in terms of grams per mile  (gpm).
The  IM240  has  proved  particularly   effective  in  accurately
identifying high emitting, newer technology vehicles.

Multiple Independent Supplier Test-Only  Network;  A  hybrid program
design  with  elements  of both  centralized  and  decentralized
networks in  which  multiple participants are  licensed to  perform
I/M  testing  (as  opposed  to  a single contractor).   To establish
equivalency with traditional  centralized programs and to avoid the
decentralized discount  incorporated  in  EPA's  MOBILE  model,
participants must operate  test-only facilities and are barred from
making  repairs,  selling replacement  parts,  making  referrals,  or
otherwise engaging in activities that would violate the intention
of the  test-only requirement  (i.e., the avoidance of conflict-of-
interest) .

Pressure   Test:   A  test  whereby  inert gas  is  injected  into  a
vehicle's  evaporative  system to establish  the system's integrity
by  indicating the  presence  of  a  leak(s)  or  by confirming the
system's ability to  hold pressure.

Purge Test:   A test to determine whether  a vehicle's evaporative
emissions  system recycles the  gasoline vapors  adsorbed  on the
charcoal in the canister (i.e.,  whether  or  not the canister purges
vapors  to the engine to be combusted).   To provide  representative
operation  and  opportunity  for  the  purge  control  system  to
demonstrate its proper working order, the  purge test is conducted
on a dynamometer using  the same  240-second  transient driving  cycle
as the IM240 exhaust gas test.

250. 0 rpm/Idle Test,;  A two-speed,  steady-state concentration-type
test in which emissions are sampled at both idle  and 2500 rpm.  TO
be considered  a  pass,  a  vehicle must  pass at both speeds.   The
two-speed test has a better identification rate  for high emitting
vehicles than does the  standard  idle test.
Draft                        -4-                      2/26/92

-------
3.0  PROPOSED I/M PERFORMANCE STANDARDS


3 .1  Enhanced I/M Performance Standard

     Under the  Act,  EPA  is  required to establish  a performance
standard  for enhanced I/M  programs  including,  at a  minimum,
centralized,  annual,  automated emission  testing  of  light-duty
vehicles  and trucks,  including a  tampering check  for  emission
control devices, a misfueling check, and provisions  for including
on-road  emission testing and  inspection  of onboard  diagnostic
devices  (OBD).  The  performance standard is  defined  by completely
specifying the design of  a model or benchmark I/M program.  While
enhanced  I/M programs  need not match  the  performance standard's
model program element  by  element,  such programs must be designed
and implemented to meet or exceed the minimum emission reductions
achieved  by  the performance standard.   Any deviations  from the
performance  standard's  program design that  may  lead to emission
reduction losses must be made up by strengthening  other aspects of
the  program.    For  example,   while  the  Act  constrains  the
performance  standard for  enhanced  I/M  programs  to be based on an
annual program,  it  is  clear  that  a biennial program is more cost
effective and  results in  relatively  minor emission  reduction
losses over  those achieved  by  an annual program.   The emission
reduction losses  resulting  from  a  decision  to  test  vehicles
biennially as opposed to  annually can be made up, for example, by
extending transient exhaust testing  and purge testing  to cover
earlier  model  years  than  those  specified in  the performance
standard.  This  specific  example will be discussed  in more detail
in  section  3.2  of this report,  "Recommended Enhanced I/M Program
Design."

     In  its draft  guidance  document  issued in  April  1991,  EPA
proposed  four potential  enhanced I/M performance  standards for
comment.  More recently, in discussion with interested parties and
for  purposes  of this  document,  these four  options have  been
reduced  to  three:  a low,  a medium, and  a  high option.   Although
EPA's  proposed performance  standard most  closely  resembles the
high  option proposed  in EPA's April  1991 draft  guidance,  the
Agency is still soliciting public  comment  on  the low and medium
options.   The low,   medium,  and high  options take an incremental
approach to  advanced  technology  testing.    These options  are
presented below.    Computer  modeling runs  for  these options are
included in  Appendix  K,  while  a detailed listing of  program
elements  is provided in Table 6-1.

3.1.1     Low Option

     The  Low Option is similar to  the better  programs currently
operating, and, by the year 2000,  under conditions of temperature,
fleet composition, etc. that are  reasonably  typical  of an  average
nonattainment  area,  is projected  to yield a 10%  reduction in


Draft                        -5-                      2/26/92

-------
highway vehicle  VOCs and  a  25% reduction  in CO  over a non-I/M
scenario.   This  option  includes annual,  centralized idle testing
of model  year 1968  and  later  light-duty vehicles  and trucks  to
8,500  Gross  Vehicle Weight Rating  (GVWR),  including  a visual
inspection for the catalyst and fuel inlet  restrictor  on  1981  and
newer  vehicles.   The option  further assumes  a  pre-1981 failure
rate of 20%,  a waiver rate of  1%,  and an overall compliance rate
of 98%.

3.1.2     Medium Option

     The Medium  Option  is  identical to  the above Low  Option with
the exception that it includes  pressure  testing of  the  evaporative
system on  1971  and newer vehicles.   This  option is estimated  to
yield  a 20%  reduction in VOCs  and a 25% reduction  in CO over  a
non-I/M scenario (note  that pressure  testing is  a VOC  strategy
that  yields  no  CO benefit, making the  low and  medium options
identical  for CO purposes).

3.1.3     High Option

     The  High Option  includes  a  transient  IM240  exhaust  test
incorporating NOX cutpoints, and purge testing of the  evaporative
control system of 1986 and later vehicles  (using cutpoints of  0.8
gpm HC, 15 gpm  CO,  and 2.0 gpm NOX  for  emissions  during the  the
entire 240-second test,  but with an additional  opportunity to pass
for HC and CO by demonstrating  emissions less than 0.5 gpm and 12
gpm,  respectively,  during the  final 147  seconds).   Two-speed
testing is to be performed on 1981-1985  model year vehicles  (using
cutpoints of  1.2%  CO, 6%  CO2,  and  220 ppm HC)  while idle testing
is to be used on pre-1981 vehicles.   Idle test  cutpoints  for older
vehicles must yield  a 20%  failure rate.  The performance  standard
also  includes visual inspection of the catalyst  and  fuel inlet
restrictor on all  1984  and later vehicles  (as opposed to 1981  in
the  low  option)  and  evaporative   system  integrity  (pressure)
testing of 1983  and  later vehicles.  The High  Option is  estimated
to yield a 28% reduction in VOCs, a  31% reduction in CO,  and a 9%
reduction in NOX over a  non-I/M  scenario.


3 .2  Recommended Enhanced I/M Program Design

     The  Act requires  EPA to  establish a  performance  standard
based  on  an  annual  test  program.    States, however,  are free  to
perform  a  biennial program   under such  a  requirement  if  a
demonstration can be made that  such a program (in combination with
other features)  would be equally effective.   This demonstration is
made  using  EPA's  mobile  source emission  model  which  includes
biennial  and annual program  credits.    For example,  using  the
current version  of the  model,   MOBILE4.1,  a biennial  high option
can achieve  the same VOC  reduction achieved  by the  annual high
option performance  standard by  doing  transient/purge  testing  on
Draft                        -6-                       2/26/92

-------
1984 and  later vehicles  and  pressure testing on  1971 and  later
vehicles.   Given the added convenience and cost effectiveness  of a
biennial  program,  EPA recommends that  states  adopt the biennial
high option since  it  can  achieve reductions equal  to that of the
proposed  annual  model  program  which defines  the  performance
standard.   In addition, EPA recommends that initial  testing of new
vehicles  be delayed until the vehicle is  two or three years  old,
as the  percentage  of high emitting  vehicles  among  newer cars is
relatively small.

3.4  Basic I/M Performance Standard

     The  basic  I/M performance standard is  based upon  the program
design of the  original New Jersey program and remains  essentially
unchanged as  a result of EPA's  proposed action.   The basic I/M
performance  standard  is  estimated  to yield  a  5%  reduction in
mobile  source  VOC emissions  and a  16%  reduction  in CO.   The
performance standard  includes annual,  centralized idle testing of
model  year  1968  and later  light duty vehicles.    The  pre-1981
failure  rate  is  assumed  to  be  20%,  with  0%  waivers  and  100%
compliance.   Unlike the low option for enhanced I/M, the basic I/M
performance  standard does  not  include   testing   of  light-duty
trucks; neither does it include visual inspections of any emission
control components.
Draft                        -7-                      2/26/92

-------
 4.0   EMISSION  REDUCTIONS FROM I/M PROGRAMS


 4.1   Recent  I/M  Test Proci;rajns (Maryland and. Indiana)

      The  data  used by  EPA to assess the benefits of  high-tech I/M
 testing concepts,  including the  IM2401'  and  evaporative  system
 purge and pressure testing,  have been obtained as  a  result  of two
 special testing programs  performed under  contract  to  EPA.   The
'first testing  program  - an IM240 transient test pilot study - was
 conducted as  part of  a cooperative  project  with  the  State  of
 Maryland  in  1989,  and utilized one of the state's I/M stations for
 testing and recruiting vehicles.   This was  the  first attempt  to
 perform transient  emissions tests on consumer  vehicles in  a  high
 throughput system.  A  more extensive program  is  currently  being
 run  in Indiana.   The Maryland pilot study began testing in  August
 1989,  and continued through December of that  year,  testing a total
 of approximately 600 vehicles  for an average of approximately 120
 vehicles   per  month.    The larger-scale  Indiana  program  began
 testing in February 1990.   As of  November  1, 1991,  approximately
 8,300 vehicles  had  been tested as  part  of  the  Indiana program,
 with an average  of approximately 120 vehicles per week.  As  such,
 the  database produced  by this test program  is  the  largest  of its
 kind ever  assembled to assess I/M testing.

      The  Indiana testing contracts include  two test  facilities,  a
 laboratory in  New  Carlisle (a few miles west of  South Bend) ,  and
 an I/M station in  Hammond.   The  laboratory is owned  by Automotive
 Testing Laboratories, Inc.  (ATL), a contractor to EPA, and the I/M
 station is  owned  by  the  Indiana  Vocational-Technical  College,
 which operates the I/M  program for the  State of Indiana.  The I/M
 station includes four lanes, with ATL running one of the  four.

      EPA  has  three  separate testing contracts  in  Indiana  that
 utilize the  two  facilities: Emission factor  (EF),  I/M, and running
 loss  testing.  Reformulated fuels testing is  being  performed under
 the  EF contract.    The three   contracts  use  vehicles that are
 selected at  the  I/M station.  The selection criteria  for follow-up
 laboratory testing include  model year,  fuel metering  type, and
 results from the following tests: The  IM240,  canister purge  flow
measurement, and evaporative control  system pressure tests.

      The  goal  at the I/M station originally  was to test a  random
 sample  of 1976 and newer light  duty vehicles.   On May 15,   1991
the  recruitment  goal changed  to randomly  sample  1983  and newer
vehicles,  to  increase the  number  of  fuel-injected  vehicles
represented  in the database.  This  change was made  to reflect the
fact  that  fuel injection is rapidly replacing  carburetion as the


1   Pidgeon, W.  and Dobie, N.,  «The  IM240 Transient I/M Dynamometer Driving
    Schedule and The  Composite I/M Test  Procedure," U.S. EPA Technical Report
    Number EPA-AA-TSS-91-1, January 1991.


Draft                        -8-                       2/26/92

-------
preferred  fuel-metering  method  for  new  vehicles,   and  the
percentage of carbureted vehicles  in  the  in-use fleet will become
insignificant in the future.

     Choosing cars for further laboratory testing is driven by the
overriding  importance of testing  and assessing emissions  from -
and the  impact of  repair  on - dirty  in-use vehicles.   A random
sample of vehicles visiting the  I/M  station would result  in the
contractor recruiting mostly  clean  vehicles,  given  that  the
majority  of  excess  emissions  comes  from  a  relatively  small
percentage of vehicles known as high  to  super emitters.   To avoid
the problem and cost  of  evaluating  a majority of vehicles  that
will  ultimately be  assessed as  clean,  a stratified recruitment
plan is employed to deliberately over-recruit dirty cars,  based on
the results of IM240, purge and  pressure  tests.   Actually,  two
recruitment and lab testing programs operate  simultaneously.   In
one, a nominally 50/50 mix of IM240-clean and IM240-dirty vehicles
is  recruited  for FTP exhaust testing.   In  actual  practice,  more
clean  cars  than dirty have  been  recruited rather  than allow lab
testing  slots  to  be idle while  waiting  for a dirty car to  be
recruited.   The  Hammond I/M  lane vehicles  were   categorized  as
clean  or  dirty  using  the IM240  standards  listed in  Table  4-1.   In
the  other  lab-testing recruitment effort,  a  sample even  more
heavily  weighted  toward  purge  and  pressure test  failures  is
recruited for evaporative  and running  loss emissions testing.

                             Table  4-1

      IM240  Selection  Standards for Stratified FTP Recruitment
                             Selection  Standards
                               (grams per mile)

          Model Years           HC           CO

          1986+  *             >1.10        >15.0

          1983-85              >1.20        >16.0

   * The 1986+ standards were set to be more stringent than 1983-
     1985 standards to improve  recruitment of high emitters and
     to balance the failure rates between model year groups.

     The FTP database  that  results  from EPA's  recruitment targets
must be  corrected to  represent  the clean/dirty vehicle  ratio in
the in-use  fleet  to   correctly  determine  excess   emission
identification  rates  (IDR),  error-of-commission  rates  (Ec)  and
failure  rates  (all  important  criteria for assessing  the overall
effectiveness  of  I/M  testing  strategies) .   The  database  was
corrected using the weighting factors presented in Table 4-2.
Draft                        -9-                      2/26/92

-------
                            Table 4-2
13.
Fuel
Metering
System
PFI

g=-^Vf-t-tl--L-Llv
Lane
IM240
Results
Clean

Total
a cj.^i-^j-1-
Lane
Count
1505
97
1602
' J-^JJ- 	 W +
Lab
Sample
Count
55
19
74
Weighting
Factor
27.36
5.11

# Lab Veh
Passing
FTP
23
1
24
# Normals
Failing
FTP
24
2
26
Clean
Dirty
1555
166
73
35
21.30
4.74
25
4
32
6
   TBI
           Total    1721     108                29          38

     Weighting  factors are  used as  follows:    If the  19 dirty
vehicles that received FTP tests  in  the Port Fuel Injected  (PFI)
vehicle sample had excess HC emissions which totaled 100 gpm,  the
database would be corrected in  this case by multiplying 100 by  the
5.11 weighting  factor,  resulting in  a  corrected excess emission
rate of 511 gpm for the dirty vehicles (excess  emissions  are those
FTP-measured  emissions that exceed  the certification  emission
standards  for the  vehicle  under consideration;  an  I/M test's
identification  rate  for excess  emissions represents  one  of  the
important criteria for assessing  an  I/M test's effectiveness,  as
detailed  in section  4.2  of this report) .   In  comparison,   the
excess emissions of the IM240 clean vehicles  have  to be multiplied
by 27.36 to make their excess emissions  representative.   The total
simulated excess  emissions are the  sum of  the  simulated excess
emissions  from the  clean and dirty  vehicles  in the  I/M lane
sample.  The number of vehicles tested was similarly adjusted with
the factors  for the purpose of calculating failure  rates.    The
large sample of 55 clean  cars  in this sample provides confidence
in conclusions about a test's  relative tendency to avoid  failing
clean cars.

     Appendix H provides additional information on adjustments to
make the FTP database  representative  of the Hammond lane  fleet's
ratio  of  clean and  dirty vehicles.    Appendix  H  also  includes
tables that  allow a comparison  of  cutpoint effects  on  IDR,  I/M
failure rates,  EC rates,   and  I/M failure  rates  for FTP-passing
vehicles.

     At  the Hammond  I/M  station,  in  addition  to   the  IM240
technicians perform the official  Indiana  I/M test  (2500  rpm/ldle)
and an additional second-chance 2500 rpm/ldle test1 for those that
fail the first chance test.  Vehicles  that require a second-chance
test first receive 3 minutes of preconditioning.   The  combination
Draft                       -10-                      2/26/92

-------
of this ^enhanced" steady-state testing, along with  the  IM240  and
purge/pressure  tests  allows  for  direct  comparison  of these
alternative I/M procedures.   section 4.2 of  this  report provides  a
more detailed discussion of the results of  comparing  the  degree to
which the IM240 and the second-chance 2500  rpm/Idle test  correlate
with the FTP.

     In addition  to  assessing the IM240 for correlation with  the
FTP, several  other issues  are addressed as part  of the  Hammond
study.   Since  dirty  vehicles are repaired at the  lab, the repair
effectiveness can be  evaluated.   The  running loss  tests  allow  EPA
to  characterize  the  air  quality  impact  of  vehicles  failing
pressure and purge tests  and the  effectiveness of  repairing these
vehicles.   The  transient short  test developed by  the Colorado
Department  of  Health  (CDH-226)  as  well as a  variety of  steady-
state  tests are  performed  at  the  lab  and can be  evaluated as
potential I/M tests.   Additionally,  one of the  IM240s performed at
the lab was restricted  to inertia weight settings  of 2,500 pounds
or 3,500 pounds.  This restriction allowed EPA  to evaluate  the  FTP
correlation effect  of a  more  economical dynamometer  (with fewer
inertia weight  settings).   We found that an inertia  weight range
of  2,000  to 5,500 pounds using  four  inertia  wheels   (500,  1,000,
and 2,000 pounds  with a  fixed  wheel of  2,000 pounds)   is  worth  the
moderate additional cost.

     The  evidence  displayed  in  section  4.2  (see  below)   and
Appendices E and G of this  report graphically and quantitatively
shows the advantage of  the  high-tech  IM240  test for  the  sample of
vehicles  tested in  Indiana  in  1990  and 1991.    The  actual
calculations of the  exhaust emission reductions  of the  several
short tests are more  detailed  in  order  to best reflect the actual
characteristics of the  fleet as it ages  and changes  in technology
mix.    A  computer  model  called Tech4.1   is  used  to calculate
technology- and age-specific adjustment factors that  represent  the
effect  of I/M  programs of  different types (the  so-called "I/M
credit"),  and  these  factors  are built into  the mobile  source
emissions model MOBILE4.1.   Section 4.3  of  this document  contains
details on  the Tech4.1 model, while  Appendix  B  lists the  source
code for the model.

     Finally,  the Indiana testing program  has  revealed the true
seriousness of evaporative  emission  control  system  malfunctions
that develop  during  real world operation.   Previous EPA  testing
programs  (i.e.,  those conducted during  the last 10  years or  so)
that did  not  make  use  of  an operating I/M  lane to screen  and
recruit vehicles for more thorough laboratory testing have  focused
mostly  on vehicles that  were about  5  years  old  or younger,  in
order to most  quickly obtain information on the latest generation
of  new  technology vehicles.   When special efforts  were made to
recruit high mileage vehicles,  they tended to be vehicles that  had
accumulated  unusually  high mileage  for their age,  for  example
vehicles from owners with long commutes or who  used their vehicles
for business during  the day.   EPA  staff have  been  concerned  for
Draft                        -11-                      2/26/92

-------
some time that testing such vehicles was not giving a true picture
of  evaporative emission  problems,  which  may develop  more as  a
function  of  passing time  than  of miles  driven;  for  example,
deterioration of rubber and plastic  components would be  more time-
than mileage-based.   Also,  the recruitment practices in  the  test
programs prior to the  Indiana I/M  lane program  relied  on  owner
response to letters and phone  calls.  There has been concern  that
this resulted in a different sample  of vehicles, probably  a sample
biased towards better maintenance condition than would be  found if
owners could be solicited face-to-face,  as  they are in the Hammond
study   (where the  level  of  motorist  participation  has been
sufficiently  high   to   ameliorate  these   concerns)  .    These
differences in study design explain  why the  Indiana program  has
produced results very different  from  previous  estimates of in-use
evaporative  emissions.     EPA's  interest   in   the  high-tech
evaporative purge and pressure tests has been  in response  to these
findings.

     Because  of the  extensive  detail  of the evaporative emissions
findings from Indiana, the results of the testing  are presented in
Appendix C, rather than illustrated with figures  and tables here.
Briefly  stated,  the  Indiana  program showed that  by 13  years  of
age,  nearly   one-half  of  all  vehicles will   experience   an
evaporative  system  failure  that   renders   the   control  system
virtually   ineffective,   causing evaporative and  running  loss
emissions to  increase by factors of  up to 10 times.  Nearly all of
these failures can be detected by the combination of the  pressure
and purge  tests.   Use of only one  of these tests  finds  at  least
some of  the problem vehicles.   The  problems can  be  repaired,  and
vehicles will  then pass a  re-inspection using  the pressure and/or
purge test.   Appropriate  repairs reduce emissions  back to normal
levels.   Of course,  the purge and pressure tests  cannot  overcome
the  limited  control capacity designed into vehicles  by their
manufacturers, so under certain conditions  of  temperature  and fuel
volatility, both passing  and repaired vehicles will  fail  to  meet
the certification emission standard.

4.2  FTP HC/CQ Correlation Comparison  Between the IM24Q  and  the

     Second-chance 2500  rpm/Idle Test

     This section focuses on the comparison of the IM240 transient
test (using cutpoints of 0.8 gpm HC  and 15  gpm CO for the results
over the full 240 seconds, with  a  provision  that  a  vehicle  also
may pass by having emissions  during  the last 147  seconds  of  the
test less than or equal to  0.5 gpm  HC and  12  gpm  CO - see section
4.2.3 for  a  more  detailed  explanation of  the   two-ways-to-pass
criteria)  to  EPA's  currently  recommended   second-chance  2500
rpm/Idle test procedure2,  and details the evaluation criteria  upon
2   Tierney, E., Herzog,  E.  and  Snapp, L.  "Recommended  I/M Short Test
    Procedures For the  1990s:   Six Alternatives", U.S. EPA Technical  Report
    Number EPA-AA-TSS-90-3, January 1991.


Draft                       -12-                      2/26/92

-------
which the  comparison is based.  This  comparison  shows how an  I/M
program  based  on  one  of  the  better  currently  used   (non-
dynamometer)   I/M tests   (second-chance  2500  rpm/Idle)   can  be
improved upon by  changing to  the IM240  test,  which has  a much
better classical correlation with the FTP than the  idle  or 2500
rpm/Idle test  for matched pollutants  (see the regression analyses
including  R-squared  values and scatter plots in Appendix E for an
illustration of this better correlation).

     For  the  sake  of  the  correlation analysis  illustrated  in
Appendix  E,  only  1983  and  newer  vehicles equipped with  fuel
injection  were considered3.   The vehicles  in this sample received
both the  second-chance  2500  rpm/Idle test  and the  IM240  at the
Hammond  test  site.   At  this  time,  most I/M  programs have not
adopted second-chance testing and the test algorithms recommended
in  EPA's  Alternative Test Procedure  report,  which  calls  for  an
immediate  second-chance test  for vehicles  that  initially fail the
emission standards.  Under the recommended  procedures,  vehicles
are preconditioned in a non-loaded state for three minutes at 2500
rpm prior  to the second  test.   Second-chance testing was devised
to  reduce,  to  the extent possible, the problem of falsely failing
vehicles.    For the purposes  of  this comparison  and to  enable
analyses of  the  effectiveness of more  stringent standards, second-
chance  tests  were  performed  on  1983  and  newer  fuel-injected
vehicles  if their  emissions exceeded 100 ppm HC  or 0.5%  CO  on
their initial  2500  rpm/Idle tests.  Note that these standards are
substantially  tighter than the  standards of 220 ppm HC and 1.2% CO
used in nearly all I/M programs on 1981 and later vehicles.

     One of the  central  concerns  in  developing  a new  I/M short
test was to  devise a test that  would pass  vehicles that would pass
the FTP  and  fail those  that  would fail  the FTP.    With  that  in
mind,  the  IM240  was  devised by  truncating, splicing,  and otherwise
augmenting the first two hills of  the FTP driving  cycle.   One of
the goals  of the pilot  program was to assess  how  well  the IM240
correlates with the  FTP.   Since performing the  FTP in the Indiana
lane was  not  a  practical alternative, both  IM240s  and  FTPs were
conducted  in  the lab after the vehicles were recruited in the I/M
lane.   The lab results  of the  IM240  and the FTP  showed excellent
correlation.    One  can  conclude  that the IM240  is  an  excellent
measurement  of the true  emissions  of  the  vehicle  at  the time and
place it is performed,  given the fuel being used at  the time.

     Comparing lab FTP  and lane  IM240 results  is  problematic for
several reasons,  but still shows good correlation.   Since the lab
tests  are  performed at  a different  time from the  lane  IM240s,
intervening  factors, such as intermittent problems  or changes in
   The emission reduction benefits presented in Section 6, however, do reflect
   the  application  of  the  IM240 to carbureted vehicles as well as fuel-
   injected vehicles; the comparisons  of IDR,  EC rate,  and failure  rate for
   the  various I/M tests presented  in Appendices  G and H also  address
   carbureted vehicles.
Draft                        -13-                      2/26/92

-------
the vehicle, may affect the results.  For example, exhaust systems
are repaired, when needed, prior to the  lab tests.   Another major
problem making lab and lane comparisons difficult is the fact that
FTP tests are all done on Indolene  fuel4 while lane tests are done
on the  fuel in the  tank  of  the vehicle  as received.   Also,  the
equipment  used in the lanes  measured a  lower maximum emission
value than the lab equipment;  for example, a car would have pegged
the lane instrument for hydrocarbons at 13 gpm while in the lab it
actually measured 25 gpm.  Temperature  and preconditioning at the
lane  were  also  often  different  than  at the  lab.   For  these
reasons,   lab/lane  comparisons   say   less  about  the   actual
performance  of  the test and more  about the influence  real world
differences make on vehicle emissions.   Nevertheless, both sets of
comparisons are presented  in Appendix E of this  report.

     One  of the conclusions  evident from  the  data  collected as
part of  the Hammond study is  that for fuel injected vehicles in
particular, the high-tech IM240 test has a better correlation with
the FTP  than the conventional  idle or  2500 rpm/Idle test.   This
section and Appendix E present some illustrations  of this better
correlation.

     For  example,   one  indication  of  better  correlation  is
demonstrated  by  higher   R-squared  values  from   least-squares
regressions with FTP emissions as the dependent  variable and short
test emissions as the  independent  variable.   Statistics for these
regressions are  given  in  the regression  analyses  tables  in
Appendix E.

     The  better  correlation of the IM240  test also can  be  seen
visually  in the  scatter plots of  emissions  results  from vehicles
which received all four tests  (Appendix E).  Separate plots of FTP
versus  short test  results are included  for  each  type  of  fuel
injection  (whether PFI   or  Throttle  Body   Injection  (TBI)),
pollutant  (HC, CO, and  NOX) , and each  short test  type  (except for
idle and  2500  rpm/Idle for NOXf since representative  in-use NOX
emissions cannot be measured on these tests).  Because of the wide
range of the data, the  graphs  showing  all the data contain a lump
of  points  near  the  origin.     To  allow examination  of  the
correlation for vehicles emitting in this range, an enlargement of
the data in this range  is  also provided  for each  of the graphs in
Appendix E.

     The  above  two  indications   (R-squared values and  scatter
plots)  of better correlation do not directly enter the calculation
of the  emission reduction  advantages of the  IM240.    In  an I/M
program,  predicting  the absolute  level  of   a  vehicle's  FTP
4   Indolene is a special test fuel whose properties are held constant.   This
   is necessary because the normal changes in fuel properties of  commercial
   fuel can  change a car's emissions results even if all of the  other test
   procedure variables and vehicle variables did not change between tests.


Draft                        -14-                      2/26/92

-------
emissions is not  as  important as identifying a large majority  of
the vehicles whose emissions are likely to be high enough to merit
repair  (which  are,  themselves, a minority of the overall  in-use
fleet) .   Also,  the short test  should  pass vehicles that are not
malfunctioning,  in order to  avoid impacting  owners of  vehicles
which have emissions low enough to not merit  repair.  The figures
in Appendix G,  which  are  discussed  in  the  following  sections,
graphically demonstrate the differences between the  second-chance
2500 rpm/Idle test and  IM240 in regard to  these objectives.

     The  I/M  test must  also  do  a  good  job  of  ensuring that
vehicles that  have shown emission reductions from  repairs  large
enough to pass  re-inspection  on the short test have  also achieved
sizeable FTP  reductions.   Better  performance of one short test
versus another in identifying vehicles  as  generally clean or dirty
will also ensure that fewer vehicles  can pass reinspection without
achieving real  FTP reductions.  Therefore,  it  is clear  that the
IM240 test will be the better enforcer of good repairs.   Analysis
of  data from  vehicles  in  Indiana  that were  repaired  at the
laboratory  and  retested  on  both  the FTP and IM240  shows that
reductions measured by the  two tests  are  highly  correlated, even
better  than  the  correlation discussed  above.   Figures and
statistics to illustrate  this  are also included in Appendix E.

4.2.1     I/M Test Assessment  Criteria  Overview

     In  assessing the  overall effectiveness  of  an I/M testing
procedure,  it  is important  to determine  the test's  effectiveness
in measuring and determining  a variety of factors,  including the
IDR, the failure rate,  the error-of-commission rate, the  failure
rate among vehicles that pass  FTP standards,  and  the failure rate
for  so-called  "normal  emitters," which may  fail  an FTP  standard
but are clean enough to make it an  issue whether  they will benefit
much from normal  repair  procedures.   Each of these  is  discussed,
in turn, below.  Section 4.2.2 provides a more detailed  discussion
of the same topics.

4.2.1.1   Excess Emission Identification Rate  (IDR)

     EPA commonly uses the rate of excess  emissions  identified
during an I/M  test to  objectively  and quantitatively compare I/M
test procedures.  As mentioned earlier,  excess emissions are those
FTP-measured  emissions  that   exceed  the  certification  emission
standards for  the vehicle  under consideration.    For example,   a
vehicle  certified to the  0.41 gpm HC standard  that  failed the
second-chance  2500  rpm/Idle I/M test with an FTP result of 2.00
gpm, would have  excess emissions equalling 1.59 gpm  (i.e.,  2.00  -
0.41 = 1.59) .

     The excess emissions identification rate (IDR)  equals the sum
of the  excess  emissions  for  the  vehicles failing  the I/M test
divided  by  the total  excess  emissions  (because  of   imperfect
Draft                       -15-                      2/26/92

-------
correlation between  I/M  tests  and  the  FTP,  some I/M  passing
vehicles also have excess emissions which are used for calculating
the  total excess  emissions) .    Thus,  assuming  an I/M  area that
tests  1000 vehicles, 100  of which are  emitting 1.59  gpm excess
emissions  each,  while the  I/M  test  fails  (identifies)  80 of the
excess  emitting  vehicles,  the  excess  emission identification rate
can be  calculated as follows:

      80 failing vehicles * 1.59 gpm excess per vehicle ^    _
         100 vehicles *  1.59 gpm excess per vehicle

As can  be  seen  in Figures  1 and  4 in Appendix G,  the IM240 using
two-mode criteria has been shown to identify more excess emissions
among the  cars  tested at  the Indiana lane  than the second-chance
2500 rpm/Idle test with  current  I/M program  cutpoints.

4.2.1.2    Failure Rate

     As the IDR increases, the opportunity to  identify vehicles
for emission repairs also increases.   However, this measure is not
sufficient  for  determining which is  the more  efficient  and cost-
effective  I/M test.   Other  criteria must  also be addressed before
such an assessment can be made.   One such criterion is the failure
rate,  which  is calculated by  dividing the number of  failing
vehicles by the number of vehicles tested.   For example:

            50 vehicles failed I/M  „. , nn
            -rrrr	7—:	7—,. '   * 100 = 5%  I/M failure rate
             1000 vehicles tested

     The  ideal  I/M test  is one  that fails  all of  the dirtiest
vehicles while  passing  those below the  FTP standard or close to
it, but still above it.  The potential  emission reduction benefit
decreases as emission levels from a vehicle approach the standard,
because the prospect for  effective repair  diminishes.    Thus,
achieving a high IDR  in conjunction with  a  low failure  rate (as a
result  of identifying  fewer vehicles  passing or  close to  the
standard)  efficiently utilizes  resources.   As  the figures  in
Appendix  G show,  tightening  the  cutpoints on  the  idle  test  to
achieve IDRs comparable to the  IM240's results  in  increasing the
failure rate well beyond that of the  IM240.   For example, for 1983
and newer,  PFI  vehicles,   the  failure rate  rose from 12%  to  38%
when second-chance, two-speed cutpoints were  tightened  to 100 ppm
for HC  and 0.5% for CO,  even though the two-speed test's IDRs for
HC and  CO were  only  77%  and  82% respectively  (compared  to  the
IM240's 82% and 85% IDRs  for HC and CO,  and its 14% failure rate).
The  remaining   figures   in Appendix  G  illustrate  a  similar
relationship between  IDR  and failure  rate  for  tighter  two-speed
cutpoints  for  both  TBI and carbureted vehicles.   For a  more
specific,  model year breakdown  of failure rates among the vehicles
in the Hammond lane  sample,   by test  type,  see  the  Appendix
M,."Model Year Failure Rates by  Test Type."
Draft                        -16-                      2/26/92

-------
4.2.1.3   Error-of-Commission (Ec)  Rate

     Properly  functioning  vehicles which  pass FTP  standards
sometimes fail the  2500 rpm/Idle  test;  these are referred  to  as
false failures or as  errors of commission  (Ecs).  When  error-of-
commission vehicles are sent to repair  shops,  no emission control
system malfunctions exist.   Often, the  repair shop finds  that  the
vehicle  now  passes the test without  any  changes.   These  false
failures waste resources,  annoy vehicle  owners,  and may  lead  to
emissions  increases  as  a  result of  unnecessary  and  possibly
detrimental "repairs."  Motor vehicle manufacturers  see this as a
significant   problem,  since  it  can   contribute  to   customer
dissatisfaction  and  increased  warranty  costs.   An I/M  program
seeking larger emission reductions through more stringent  emission
test standards may actually increase  the number of false failures.
The error-of-commission rate is,  therefore,  an  important measure
for evaluating the accuracy of I/M  tests.

     To see how  an  error-of-commission  rate is calculated,  assume
an I/M area which tests 1000 vehicles,  of which  100 fail the  I/M
test, although only 50  of those 100 failing  vehicles also exceed
their  FTP  standard for HC or CO.   The error-of-commission rate
equals the number of vehicles that fail  the I/M test  while passing
the FTP  for  HC  and CO, divided by  the total number of  vehicles
which were I/M tested:

     50 vehicles failed  I/M but passed FTP HC  and CO ^    _       *
                 1000 vehicles tested                 ~

*Error-of-commission

     As  the. error-of-commission  rate  decreases, vehicle  owner
satisfaction and acceptance of  the I/M program  increases.   Thus,
while it is relatively  easy to  improve  the IDR by making the  I/M
test  standards  more  stringent,  this "improvement"  comes at  the
cost of potential increases in the  error-of-commission rate.

4.2.1.4   Failure Rate Among FTP-Passing Vehicles

     The risk of failing an I/M test with a  clean vehicle  is  not
expressed  very  clearly,  however,  by   stating  fleet  error-of-
commission rates.  Fleet rates tend to be very low, but  the impact
on  any  individual motorist can  be  very  significant.  A more
informative statistic  than error-of-commission rate is the failure
rate  among all  inspected  vehicles  which  still pass  their  FTP
standard.  This  indicates  the risk to the  owner  of having a clean
vehicle  failed.    For the  IM240  using  the  two-ways-to-pass
criteria, the failure  rate for vehicles  which pass the FTP is zero
for the  sample tested  in  Indiana  (Appendix G).   While the  false
failure  rate  for the second-chance two-speed test  is  initially
comparable to the  IM240 using the two-speed  cutpoints  in current
use, tightening  these cutpoints  to improve IDR has  the  effect  of
Draft                        -17-                      2/26/92

-------
increasing  the false  failure  for  the  steady-state test.   For
example, as illustrated in Figures I through 3 of Appendix G, _for
1983 and newer PFI vehicles,  tightening the  steady-state  cutpoints
from 220 ppm  EC and 1.2%  CO (the cutpoints  most commonly used in
current I/M programs) to 100 ppm HC and 0.5% CO has  the  effect of
increasing the test's  false failure rate from 0%  to 13% - this,
even though  the  two-speed test's IDRs for  both HC  and CO still
fall appreciably below that of the IM240.  For 1983  and  newer TBI
vehicles,  the same tightening of cutpoints achieves HC and CO  IDRs
for the steady-state test that actually exceed those of  the IM240
by  a  percentage  point or  two,  but this  at the cost of a false
failure rate  of  20% compared to no false failures  for  the IM240
(Figures 4-6, Appendix G).

     Even when the  two-ways-to-pass criteria are not used for the
IM240, the false failure rate for the  vehicles in EPA's  sample was
only  0.8%,  representing a  total of  5 EC vehicles  -  still much
lower than the false failure rate for the steady-state  test with
comparable IDRs.   Since even this number of  full-test failures was
unexpected,  given  the IM240's  similarity  to  the  FTP  and the
looseness of the 0.8/15 cutpoints compared to the 0.41/3.4 new car
standards,  section  4.2.3  is  included to   discuss  these  false
failures in depth.

4.2.1.5   "Normal Emitter"  Failure Rate

     The  IM240 failure  rate for  normal emitters   will  also be
lower.  ""Normal"  emitters  are defined,  for  the  purposes of  this
discussion,  as those vehicles that emit less  than twice  the FTP HC
standard and  less  than three times the  FTP  CO  standard.  Normal
emitters include  those vehicles that  pass  the FTP.   Repairs on
such vehicles usually do not produce large emission reductions  (at
least short of catalyst replacement, which EPA generally avoids in
its emission  repair evaluations due to  cost  and because testing
after a new  catalyst is  installed would not necessarily  indicate
what  emissions will be after  the catalyst  "wears  in"),  their
emissions  are sometimes  increased  by  inept  repairs,   and they
account for little of the total  excess  emissions (9% of  excess HC
for 1983 and newer PFI vehicles  [page  8,  Appendix H], 6% excess HC
for 1983  and  newer TBI vehicles  [page  21,  Appendix  H]  and 3%
excess  HC  for 1981 and  newer  carbureted vehicles   [page 31,
Appendix H] ) .   Therefore,  normal emitters  are  not  the most  cost
effective  to identify  for  repairs.   These vehicles  often  lack
overt defects.   Those that  fall above one  of  the  FTP  standards
obviously have some  problem,  but may  only have suffered catalyst
deterioration  (which is  difficult to  diagnose)  or  may  have  been
either poorly  designed or built  in the first place.    Thus, the
marginal  costs of  identifying  and effectively repairing these
vehicles may not always be worth the marginal benefits  that  could
be expected.
Draft                       -18-                      2/26/92

-------
4.2.2     Detailed Discussion of Correlation and Test  Assessment

     The  following  analysis shows that  the IM240 test  using the
two-ways-to-pass criteria is considerably more  powerful  as an I/M
test than the second-chance  2500  rpm/Idle test  for all technology
type vehicles, but  especially  newer  tech,  fuel-injected vehicles.
The  analysis  presumes that  the IM240 is  implemented to  achieve
higher IDRs.  Given that  rationale,  IM240 standards of 0.8 gpm HC
and  15 gpm CO for the full  test and 0.5 gpm HC  and 12 gpm CO for
the  last 147 seconds were selected for this  analysis.   These IM240
standards achieve IDRs that  are  significantly higher  than  for the
present second-chance 2500 rpm/Idle standards,  while maintaining a
false failure rate of zero.

     This discussion  is  limited to  PFI  vehicles,  as  this  is the
most commonly used fuel metering system on new vehicles.   Throttle
body injection,  which is less sophisticated, may also  be used on a
significant proportion of the  future fleet, though less than for
PFI.    Therefore,  although  analogous  figures  and  tables  are
included  in  Appendices  G   and  H  for both  TBI  and carbureted
vehicles,  they are not formally discussed.

     Figure 1 in  Appendix G provides a comparison of the  present
second-chance 2500  rpm/Idle  test  using current  standards  (220 ppm
HC and 1.2% CO) to  the more  effective, high-tech IM240 test using
the  two-ways-to-pass criteria.   Note  the following:

   •   The  FTP excess emissions  identification rates are 17% higher
      for  HC and 12%  higher  for CO with the IM240 as compared to
      the  second-chance 2500 rpm/Idle test  using the  1.2%/220 ppm
      standards.

   •   Neither test failed FTP passing vehicles.   (Without the two-
      ways-to-pass criteria,  however,  5 FTP-passing  fuel-injected
      vehicles would  have failed the IM240  yielding  an  error-of-
      commission  rate   of   0.8%  using  the  weighting   factors
      (discussed in section  2.4.1) to correct  the lab  sample to
      simulate the  lane  sample.   These vehicles are  discussed in
      section  4.2.3.)

   •   The  IM240  increases the failure rate to 14% from 12%  for the
      second-chance  2500  rpm/Idle test.

     Figure  2  in Appendix  G illustrates the power  of the IM240
test compared to the  2500 rpm/Idle test using more stringent idle
standards,  currently  in  use in California.   I/M programs might
consider California idle  standards because  the  emission  reduction
from the  program  can  be  increased and the  cost  of implementation
is relatively small.

     California uses  standards  of 1.0% CO and 100 ppm HC  for the
idle mode, while using 1.2%  CO and 220 ppm HC  for the 2500 mode.
In Figure 2,  only  the stringency of the  2500  rpm/Idle  test is


Draft                        -19-                      2/26/92

-------
increased, while the IM240 standards are the same as those used in
Figure 1  (see Appendix G for both  figures).  Note the following:

   •   The IDRs are  still  6% higher for HC and 3% higher for CO with
      the IM240 as  compared to the second-chance 2500 rpm/Idle test
      with more stringent standards.

   •   The  second-chance 2500  rpm/Idle  test  failure  rate  using
      California standards  is  34%  compared to  only  14% for  the
      IM240.   So even with the IM240's  higher  IDRs,  significantly
      fewer vehicles will need to be repaired.

   •   Thirteen percent  of the FTP-passing vehicles fail the second-
      chance  2500  rpm/Idle  test,  while  none  fail  the  IM240.
      Sending this many  cars  for  unnecessary repairs, while  also
      identifying less  excess emissions, wastes  resources.

   •   The normal emitter failure  rate  is only  2.3% for the IM240
      versus  22% for the  second-chance 2500 rpm/Idle test.   This
      means that the vehicles identified for repairs by the IM240
      are more likely to achieve significant emission reductions.

      In  Appendix G, Figure  3  compares  the same  IM240  standard to
the more stringent standards  of 0.5%  CO  and 100 ppm HC  for  both
modes of the second-chance 2500  rpm/Idle  test for PFI vehicles,
while Figures 4,   5,  and 6 present data analogous  to the  first
three figures,  but this  time  for  TBI vehicles,  and Figures 7-9
present  this  information  for  1981 and newer carbureted vehicles.
However,  second chance  testing  was only performed  on  1983  and
newer vehicles, so Figures 7, 8,  and 9  only include  second  chance
results  for  1983  and newer  vehicles,  not  for  1981  and  1982
vehicles.

4.2.3     Avoiding Errors of Commission

      I/M test procedures and  standards that  cause low  emitting
vehicles to fail I/M tests are obviously undesirable.   Using full-
test  IM240  standards  of 0.8  gpm  HC  and 15  gpm  CO   (without
incorporating  0.5/12  cutpoints  for  the last 147 seconds)  results
in  five  of the  182  vehicles that  received FTP  tests (1983  and
newer with fuel injection)  failing  when they should have  passed.
EPA staff were curious to find that  even this small number of  FTP-
passing vehicles failed the IM240  0.8/15 standards  since  the IM240
driving  schedule is taken from the  FTP and is a  hot  start  test at
the  Indiana  lane.     This  section will  investigate  why these
vehicles  falsely  failed  the IM240 test,  and  in  so  doing  will
explain  how and why two-mode criteria  were  developed.  Given  that
attempting  to  repair  properly   functioning  vehicles   wastes
resources and  adversely  affects vehicle owners,  I/M programs  and
vehicle manufacturers,  EPA  regards this as an important issue.

     The  causes  of IM240 false failures  can  be categorized  into
those that may have occurred due  to errors  in  performing  the  test


Draft                       -20-                      2/26/92

-------
and those  that occurred  because of the  interactions of  vehicle
design,  fuel properties  and test  conditions.

     In addressing these errors of commission,  we will concentrate
on  the  vehicles  actually  tested  rather  than the reweighted
database discussed elsewhere.   Of the 74  PFI  equipped 1983  and
newer vehicles  receiving FTPs,  24 passed the  FTP,  while one  of
these 24  failed the IM240  standards of 0.8/15.  Of the  108  TBI
1983 and newer vehicles receiving FTPs,  29 passed the FTP  and 4 of
these failed  the IM240 0.8/15 gpm standards.   Other  identifying
data and test results for  these vehicles are listed in Table 4-3.

                             Table  4-3

          FTP, Lane  IM240. and "Tank Fuel" IM24Q Results
                         (grams per mile)
Veh
t
1674
685
1607
1624
1663
Fuel
Sys
PFI
TBI
TBI
TBI
TBI
Model
Year
1989
1989
1988
1983
1986
Mfr.
Suzuki
Chry.
GM
GM
GM
Carbon Monoxide
Tank
Lane Fuel
IM240 IM240 FTP
15.3
16.3
16.1
19.1
23.3
1.7
2.7
1.2
11.0
3.4
3.1
2.4
2.5
2.2
0.5
Hydrocarbons
Tank
Lane Fuel
IM240 IM240 FTP
0.73
0.4
0.78
1.37
0.76
0.1 0.37
0.06 0.12
0.14 0.24
0.64 0.38
0.15 0.1
Oxides
Lane
IM240
0.8
1.3
0.9
1.4
0.7
of Nitrogen
Tank
Fuel
IM240 FTP
0.2
1.1
1.1
1.6
0.6
0.2
0.8
0.6
1.4
0.4
     The  first thing to  note is that  the lane  IM240  HC and  CO
emissions are considerably higher than for the two tests  performed
at the  laboratory.   The "tank fuel"  IM240 is the first  test the
vehicles  receive  at the laboratory,  usually  using the same  fuel
that was  in  the tank for the  lane  test.   Since the vehicles are
stored  outside prior to  testing, each  is preconditioned before
receiving the tank  fuel IM240.   The preconditioning consists  of a
3 mile  drive at 50  mph  (on  the road)  immediately  before  the  test.
The comparison of the lane and "tank fuel" tests strongly suggests
that  it  was  some  factor  at  the  lane  that  resulted  in  these
vehicles  being  errors   of  commission,  not   some  inherent
unsuitability of the IM240 test itself.

     The  following analysis  of   the  test results  suggests  that
vehicle  1674,  the  only  PFI  error-of-commission vehicle,  probably
failed  as  a  result  of testing errors at  the  lane.   The  four TBI
error-of-commission vehicles are  likely  the result of  interactions
of vehicle design,  fuel  properties and test conditions.
4.2.3.1
Vehicle 1674
Draft
                  -21-
2/26/92

-------
     One  possible explanation for vehicle  1674's results  on  the
lane IM240  test  is  that it was run at too high  an  inertia weight
 (IW) setting  on the dynamometer.  This  is important because  the
fuel  metering  system  enrichens  under  heavy loads,  causing  CO
emissions to increase, which could explain the inconsistently high
CO result at the  I/M lane.  Also,  NOX emissions typically increase
in  conjunction  with  engine  load,  while  fuel  economy  (FE)
decreases,  suggesting  one method for  determining  whether  the
dynamometer settings  were,  in  fact, inconsistent.    Table  4-4
provides  fuel economy  and NOX results for vehicle  1674  for  the
various tests conducted.

                            Table  4-4

     Vehicle  1674 NOy, FE Results Versus  Dynamometer Settings
       NOX (gpm)
Lane
IM240

0.83
   Fuel Economy  (mpg)    29.61
   IW Setting (Ibs.)

     Road Load (hp)
2125
Tank Fuel
  IM;

  0.24

  31.19

  2125
indolene
 IM24Q

  0.12

 33.90

  2125

  6.5
IW=2500#
 IM24Q

  0.18

 32.43

  2500

  6.5
     Note  that  the lane IM240 NOX is  about  4 times greater  than
any  of  the other NOX results, including  the  2,500 pound  inertia
weight  test  that  was  performed  at  the lab to  simulate  an
economical dynamometer  with  only  two  inertia weight  settings
 (i.e.,  2,500  or  3,500  pounds).   The  lane IM240 shows poorer  fuel
economy than for  any of  the other  tests.    Another indication
suggesting that vehicle 1674  represents a unique situation  is  the
fact that the other 4 vehicles which  were  errors of commission  had
fairly  consistent NOX emissions,  (see Table 4-3) even  though their
HC and CO emissions deviated  considerably from lane to  lab.

     Another  important consideration is the fact that the  vehicle
tested  at the  lane just  before vehicle 1674 used  dynamometer
settings  of  3,125  pounds  and 7  horsepower   (hp) .  Vehicle  1674
should  have  been  tested  at   2,125   pounds  and 7  hp.     The
comparatively high  NOX  emissions and fuel consumption results  in
conjunction  with  the  consistent NOX results  for the  other  4
vehicles  suggest that  1674 was,  in  fact,  tested  at the previous
vehicle's inertia weight setting of  3,125 pounds,   47% higher  than
the 2,125 pound  setting  required.

     The settings at the  Hammond lane are selected manually,  and
there is  no  feedback  circuit to document the setting.   The  only
record of the setting is a handwritten  data sheet  completed by  the
Draft
    -22-
                     2/26/92

-------
lane technician,  who may  have recorded  the correct  dynamometer
setting,  while  failing  to  change the  actual  setting  on  the
dynamometer.   Given the importance  of using correct  dynamometer
inertia weight settings,  automatic selection of  the IW settings is
essential to  ensure that vehicles  are  tested correctly.  EPA is
developing  quality  control  procedures  to insure correct settings
are chosen in I/M program lanes.

4.2.3.2   Interactions  of Vehicle  Design,   Fuel  Propertiesf  and

          Test Conditions

     Three  techniques were  considered  for preventing  IM240  false
failures  caused  by  the  interaction  of  vehicle  design,   fuel
properties and test conditions.

   1.  Reduce the  stringency of the pass/fail standard.

   2.  Allow second-chance testing for  failing  vehicles that  are
      close to the standard  and precondition vehicles  prior to the
      second-chance  IM240.

   3.  Since the IM240 is  a 240-second test,  use  the latter part of
      the test as  a  separate,  second chance to pass, if the vehicle
      fails over  the entire  test.   For  example,  consider a vehicle
      with a  composite result of  0.7 gpm HC and 16 gpm CO,  with
      Bag-2 results of 0.4  and  11 gpm.   Because it  only slightly
      exceeds the 15 gpm  CO  standard,  the test software could be
      designed to also  consider the emissions  in the  147-second
      mode of the test (i.e., Bag 2).   Vehicles pass  if they pass
      either  the full-test   outpoints  or  the  Bag-2  outpoints.
      Assuming Bag-2 outpoints of 0.5 gpm HC and 12  gpm CO,  which
      are  significantly  more  stringent   than  the  composite
      standards,  the vehicle in this example would  pass because it
      does not fail both sets of  cutpoints.   The  logic supporting
      the suitability of these pass/fail criteria will be discussed
      later.

Each of  the three strategies for avoiding false failures will be
further discussed below.

     The  first  alternative,   reducing  the stringency of  the
pass/fail  standards, is  unacceptable.   Although this strategy
decreases errors  of  commission, it  also unacceptably  reduces IDR,
as  shown in  the  tables  in  Appendix H.  For  example,  page  H-24
lists a variety of  potential cutpoints  of varying  stringency,  the
least stringent of which is  1.6/20  gpm.   Three  of  the 4 error-of-
commission TBI vehicles  were found to pass at this  standard,  but a
number of vehicles which were high emitters  did  also,  lowering the
HC IDR from 89% to  61% and the  CO  IDR  from  80%  to  58%.   These IDR
reductions would compromise the purpose of this  high-tech test and
reduce the benefits  from  an  I/M program.
Draft                        -23-                      2/26/92

-------
     Second-chance testing  and  the two-ways-to-pass criteria  are
better ways of avoiding false failures.   The  theory  behind this is
as follows.  Assuming that the test was  correctly performed  in  the
first place,  the  most likely reason  that  a  properly functioning
vehicle would fail an IM240 is  that the evaporative canister  was
highly  loaded with  fuel  vapors and  that  the vapors  were being
purged  into  the  engine  during  the test.    This  has been  a
significant cause of false  failures in  existing I/M programs  and
it has been shown that highly loaded canisters  can cause  both high
HC and CO emissions,  even though the feedback fuel metering  system
is functioning properly.

     Since the canister is being purged  during  the IM240, the fuel
vapor concentration from the canister continually decreases  during
IM240 operation.  The  decreasing fuel vapor  concentration results
in decreasing HC  and CO emissions.   So, Bag-2 results  should be
lower than  the  composite results,  on a  gram per mile basis.    On
the other  hand,  if the vehicle is  actually  malfunctioning, Bag-2
emissions  should remain  high.   For  this  reason,  second  chance
tests after preconditioning,  as shown  for current  2500 rpm/Idle
test,  should be  less  influenced by  canister purge.

     Adding  Bag-2 results  to the  pass/fail criteria completely
solves  the error-of-commission problem  for  the vehicles  in  our
test sample.  As  shown in  Table 4-5, all  4  of the TBI  error-of-
commission vehicles  will  pass the  lane  IM240 if the two-ways-to-
pass  criteria  are  applied such  that  a vehicle  fails  only  by
exceeding  both  0.8/15 gpm for the  composite  IM240  and Bag-2
cutpoints of 0.5/12 gpm.

                            Table  4-5

   Two—Ways—To—Pass Criteria Applied to Lane IM240  TBI  Error—of-
                       Commission Vehicles

                         (grams  per mile)
Vehicle
Number
685
1607
1624
1663
Lane IM240
Bag-1
;
0
2
3
1
H£
.73
.01
.60
.94
£Q
29.
44.
53.
55.

1
4
6
0

0
0
0
0
Lane IM240
Bag-2
HC.
.28
.30
.47
.28
Lane IM240
Composite
£P_ HC.
11
4.
5.
10
.2
9
0
.5
0
0
1
0
.40
.78
.37
.76
m
16.31
16.09
19.08
23.28
     Using  two-ways-to-pass  criteria has  the added  benefit  of
reducing  the  normal emitter  failure rate  for  TBI vehicles  from
5.2% to  4.1%.   At  the same  time,  the overall  TBI failure  rate
Draft
-24-
2/26/92

-------
decreases from  22%  to 21% without a perceptible  reduction  in the
IDRs  which  remain  at 89%  and 80%  for HC  and CO  respectively.
Analogous benefits  were  observed for the  PFI  sample, where the
normal emitter  failure  rate for PFI vehicles decreases  from 2.7%
to 2.3%,  and  the overall PFI  failure  rate decreases  from  15% to
14% without a perceptible  reduction in  the IDRs which  remain at
82% and 85%  for HC and CO respectively.

     All of the TBI error-of-commission vehicles passed 0.8/15 gpm
standards on  the tank  fuel IM240,  suggesting  that performing  a
second-chance test  after preconditioning would eliminate  errors-
of-commission.   Second-chance  testing would increase  the  cost of
the testing process, however, so it may be best  reserved as a last
resort.  Given  the  efficacy of the two-ways-to-pass criteria, the
error-of-commission problem may be fully resolved.   Nevertheless,
second-chance tests  on  select vehicles  during high  temperature
days  in  spring or fall  would  serve to insure no  false  failures.
Since such a procedure could be used on only the small fraction of
vehicles that are slightly above one or both of  the standards, the
impact on test costs should be  minimal.

      The following  discusses how  vehicle design,  fuel properties,
and  test  conditions can   interact  to cause   false failures.
Considering these variables can  facilitate designing algorithms
that  determine when  second-chance testing  should be  performed.
The success of utilizing the  two-ways-to-pass  criteria  indicates
that  second chance  testing will  not  often be  necessary.    This
discussion  shows  that these 4  TBI EC  vehicles  were  tested under
rather unique conditions and  explains why second  chance  testing
may be useful  for some cars.

      HC emissions and especially CO emissions have both been shown
to vary  with evaporative canister  loading,  so EPA  investigated
variables that affect canister loading such as gasoline Reid Vapor
Pressure  (RVP),  ambient  temperature, and the exhaust  interaction
caused by canister purge  flow (see Table  4-6).

     While  the  specific  fuel  RVP  for  the lane  IM240  is  unknown,
the test dates do provide some information regarding RVP.   The RVP
at service stations in the  Hammond area  was  regulated  by  EPA to  a
maximum  10.5  psi  between June 1  and September  15.   Three  of the
error-of-commission vehicles were tested outside of  this  period.
ASTM  guidelines allow RVP up to 13.5 psi in April  and September,
and up to  15  psi in  December.   Since these are  only  guidelines,
they can be exceeded.  Unseasonably warm test conditions can cause
high  RVP fuels to  vaporize at  an excessive rate, causing  high
canister  loadings.    If  the RVP  of the  fuel in these  error-of-
commission vehicles was  13.5 psi  or higher, it is  plausible that
these three vehicles had highly loaded canisters and/or high vapor
generation rates in their fuel  tanks.
Draft                        -25-                      2/26/92

-------
                            Table 4-6

                 Variables Affectincr  CO Emissions
       Vehicle
        Number
         685

         1607

         1624

         1663
Canister
Purge
Volume
(liters)
35.7
45.7
79.7
84.9
Engine
Displacement
(liters)
2.3
4.5
5.7
4.3
I/M Lane
Temperature
(° F)
100
71
57
87
                          Lane
                          Test
                         9/6/90

                        10/11/90

                        12/10/90

                         4/6/91
     Canister loading  and fuel  tank vapor generation,  in turn,
generally increase with  increasing  ambient temperature.   Vehicle
685  was  the  only error-of-commission vehicle  that  was tested
during  the  RVP-control-season,  and  the ambient  temperature
recorded at the time of  testing  was  100°F, considerably above the
typical FTP test temperature of  75°F.  The lane temperatures were
also unseasonably  high  for vehicles 1624 and 1663.

     If the canister is highly  loaded,  CO  is likely  to  increase  as
a function of increasing purge flow.  The median purge volume for
all  vehicles tested  at the  lane was  approximately  30  liters.
Vehicle 1663 had  an unusually  high  purge  volume of 85 liters and
was  tested at an  unusually  high temperature  (87°F)  in April when
RVP  is not regulated.   Similarly, vehicle 1624 had  a purge volume
of  80  liters and was tested  at a  lane  temperature  of  57°F  in
December.  If the outdoor temperature was in the thirties and the
fuel RVP was 15  psi,  the warming of  the fuel tank that  would occur
while in the I/M station is likely to have led to a high  canister
loading.   High canister  loading  combined with the high purge rate
probably caused  the high emissions.   Table 4-5 shows that  the Bag-
1  emissions  were  more  than  5 times  higher than  the  Bag-2
emissions, further  indicating  that a  highly  loaded canister was
sufficiently purged during  the  first  mode  to achieve relatively
clean Bag-2 results.

     The evidence suggests that the  fuel RVP and the conditions  in
the  Hammond  station caused  high CO  emissions.   Indiana is unique
in that  existing  buildings  were adapted for I/M testing  purposes
and, as  in the  case of  the Hammond station, tend  to be  closed-
architecture  structures like  warehouses  rather than  open,  airy
test stations.  The likelihood of the  kind of false failures  seen
here should be much less in actual I/M programs for two  reasons:
fuel RVP  will be  regulated to  9.0  psi during the  summer months,
and  the  factors  contributing to  canister loading  in the Hammond
lane are  better controlled  in  typical  I/M programs.   The  Hammond
IM240 lane residence  time is  10-20  minutes  for the research and
Draft
-26-
2/26/92

-------
development  type  testing;  also,  there  is  considerably  less
ventilation  (thus higher  temperatures  on sunny  days)  than  in
typical  I/M  facilities.   In  designing  future facilities  for
enhanced  I/M testing  lanes  it will  be  important to  keep those
considerations in mind, as  is  already  the  case  in most  centralized
programs.   Further,  the number of vehicles receiving  second-chance
tests  can be  minimized by developing algorithms that consider
canister purge rate, higher than normal ambient  temperatures,  and
probable  fuel  RVP  in  determining whether  to  allow  second-chance
testing.

4.2.3.3   Error—of—Commission  Summary

     EPA  views false  failures as a  significant problem  and  is
committed  to investigating and utilizing strategies  to prevent
their  occurrence.     The  IM240  can  help,  because  it  allows
prevention strategies,  such as two-ways-to-pass criteria, that are
not  relevant  to steady-state I/M  procedures.    The  following
summarizes  how  errors  of  commission can be prevented  in  I/M
programs by using the IM240.

  •  There is strong  evidence indicating that no PFI Ecs would have
     occurred if the  correct dynamometer inertia weight  setting had
     been  used.   Vehicle 1674's dynamometer inertia weight setting
     was  probably too  high.   EPA believes  that automation  and
     quality  control procedures will prevent such mistakes.

  •  All  of the  false  failures in EPA's  sample could have been
     prevented by implementing pass/fail criteria that consider
     Bag-2  results  in  addition  to composite results.   Under this
     strategy,  vehicles  fail  only if their composite emissions
     exceed 0.8/15 and their Bag-2  emissions exceed 0.5/12 gpm.

  •  Second-chance testing  after preconditioning on  marginally
     failing  vehicles,  especially those showing  high purge rates
     and tested at  higher than normal temperatures,  could be used
     as a  fail-safe  mechanism to  insure  that  no  errors-of-
     commission occur.

     In conclusion,  the data  indicate that these  false failures
were  caused by  unique  circumstances  rather  than inherent test
procedure flaws or  standards  that  are too stringent.  Experience
with  these  vehicles  suggests  that   the  problem  can be  solved
through automation,  quality  control,  two-ways-to-pass criteria,
and preconditioned second-chance tests on marginal failures.

4.3  Approval of Alternative Tests

     Although the IM240, purge, and pressure tests represent EPA's
current trio  of  recommended high-tech tests,  we do  not rule  out
the  possibility of  future,  valid alternatives  to  these tests,
including  fast-pass and fast-fail  transient  testing strategies
Draft                       -27-                     2/26/92

-------
(see  section  4.4,  "Transient  Testing   Fast-Pass/Fast-Fail
Strategies").    States  may  seek  approval   of  such  strategies,
contingent upon  the state's demonstrating to  EPA's  satisfaction
that  such  strategies  are  at  least  as   effective  as  EPA's
recommended  tests  at   identifying  excess   emissions   while
maintaining a  comparably  low error-of-commission  rate.  As  the
sheer number of analyses contained  in this report can attest,  EPA
does not promulgate new testing  strategies  capriciously.   Before
proposing the  IM240,  purge, and pressure  tests,  EPA  amassed  a
compelling body of  data on  each  through pilot  programs conducted
in  Maryland  and Indiana  (see  section   4.1,  "Recent  I/M  Test
Programs  (Maryland  and Indiana)" for  further discussion of  these
pilot studies) .   Rigorous  evaluations of  each were  conducted to
determine their effectiveness  at  identifying excess  emissions
while  maintaining  low  error-of-commission  rates.    Economic
analyses were  also  conducted to  assess the  cost  effectiveness of
the  tests,  as no degree of  technical excellence will  justify  a
testing  strategy  that  is  exorbitant  in   its overall cost.    For
example,  the  FTP   is  the  hallmark   against which  I/M  testing
strategies are measured, but  cannot itself be used as an I/M test,
given its cost.

     Of the potential alternatives  to  EPA's recommended tests,  the
one  which has garnered the most attention  is  the suggestion by
some that steady-state  loaded testing using  a simple non-inertial
dynamometer  be used to  perform the  purge   check.   EPA  pursued
transient  testing  instead  of  steady-state  because  our  best
engineering and  technical  judgement   suggested that  steady-state
testing as a mechanism for  conducting the purge  check would lead
to  decreased  emission  reduction  benefits,  higher  errors-of-
commission,  and,  ironically,  higher overall  costs  per ton  of
emission reductions  produced.  The  rationale behind the assumption
that higher  errors-of-commission rates would result is  the fact
that  purge  strategies  vary from  vehicle  to  vehicle,  and  the
possibility of developing  a steady-state  test that  successfully
addresses this variety  is  small to  none.    The result of failing to
address the  full  range of purge strategies  is easy  to predict:
Cars that should  pass  will fail,  leading to unnecessary expense
and hardship for motorists, with no  environmental benefit.

     It has also been suggested by  some that a loaded steady-state
test is  an  adequate alternative to  transient  emission testing.
One of the problems  with this proposal arises from the requirement
under section  182(c)(3) of the Act that programs  in  enhanced I/M
areas  achieve NOX  benefits.   EPA has  found  that  NOX  emission
testing  (as opposed to visual  inspection   of  emission  control
devices)  is essential for NO>: emission reductions.   Further  EPA
believes that a reliable steady-state  test for  NOx does not  exist
and  is  not  likely  to  be possible.   Nevertheless,   as  indicated
earlier,  EPA  is open to  demonstrations  by  states  or  their
representatives that proposed alternative testing  strategies  are
equal or superior to EPA's proposed tests  in terms of identifying
excess emissions  and keeping false failures to a minimum.
Draft                       -28-
                                                     2/26/92

-------
4 . 4  Transient Testing Fast—Pass/Fast-Fail Strategies

     Among   the  alternative  testing  strategies   that  make
environmental and economic  sense, the potential  for fast-pass  and
fast-fail transient testing  ranks  the  highest.   EPA  is in  the
process   of  looking  at   potential  fast-pass  and  fast-fail
strategies,   and preliminary results suggest  that roughly 33%  of
the vehicles tested  could be fast  passed or  failed based upon
analysis of data gathered during  the first  93  seconds  of the  IM240
 (i.e., Bag 1) using separate fast-pass and fast-fail cutpoints.

     In evaluating  potential  fast-fail  criteria, EPA looked at  a
sample  of 4,158  1983  and  newer  vehicles  tested  at  the Hammond
IM240  lane  described in  section  4.1,  1,033  (or 24.8%) of  which
failed  the   IM240.   298  (or  28.8%) of  the  1,033  vehicles that
failed would have  failed  within  the first  93  seconds of the test
if Bag 1  cutpoints  of 2.5  gpm HC, 50 gpm CO,  and 5.0  gpm NOX were
used;  there were no errors-of-commission.   Although  stricter  Bag  I
cutpoints could be  used to increase  the  percentage of fast-failed
vehicles, the  error-of-commission  (Ec)  rate would also rise.   In
turn,   when  fast-pass Bag  1 cutpoints  of 0.41/3.4/1.0 were  used,
1,074  (or 34.4%) of the 3,125 vehicles that passed overall passed
within the  first  93 seconds of  the test.  Seven  additional  false
passes were  also  recorded,  resulting in  an  error-of-omission rate
of  0.7%.    Tightening  the fast-pass  cutpoints  to   0.25/1.5/1.0
eliminates the false passes but  also reduces the fast-pass rate to
13.2%.  Table  4-7 provides further  details on the Bag 1  cutpoints
looked at in this analysis.  While more development  of  fast-pass
and fast-fail criteria is needed,  it is  reasonable  to conclude
that  criteria  can  be developed  to accurately  pass and  fail  about
one third of all vehicles tested  after  only 93 seconds rather than
the full  240 seconds.   Further,  EPA will  soon begin  collecting
second-by-second  IM240 data.  This  will allow the development of
algorithms  that will permit  especially clean  cars to pass well
before  93 second,  and others to pass  after 93 seconds, but well
before  240   seconds.   Once the  algorithms are  developed,  only
vehicles that are close to the cutpoints are  expected to  continue
for the full  240  seconds  to  ensure that  they  are  not falsely
failed.
Draft                       -29-                      2/26/92

-------
                            Table 4-7
             IM24Q Bag-1 Fast -Pa
                                       -Fail  Analysis
Fail
IM240
Total
1033
1033
1033
Fail
Fast-Fail
Total
1297
450
298

Fail
Both
902
445
298
Fail
Fast-Fail
Only
395
5
0
Fast-
Fail
^D rate
87.3%
43.1%
28.8%


EG Rate
30.5%
1.1%
0.0%
Pass
IM240
Total
3125
3152
3125
Pass
Fast-Pass
Total
2861
1081
413
Pass
Both
2730
1074
413
Pass
Fast-Pass
Only
131
7
0
Fast-
Pass
ID Rate
87.4%
34.4%
13.2%
False-
Pass
Rate
12.7%
0.7%
0-0%
   Fast Fail
  0.8/15/2.5
  2.0/40/4.0
  2.5/50/5.0
   Fast Pass
  0.8/15/2.5
 0.41/3.4/1.0
 0.25/1.5/1.0

     Another  area  that EPA  is investigating  is the  possibility
that the overall test time may be reduced.   The IM240 is itself an
FTP-like short  test  based upon  a  modified and  condensed  driving
cycle that takes as its reference the LA-4  cycle used in the FTP.
EPA  is  currently  investigating  the  possibility  of  further
abbreviating  the test  by comparing how  well  data from  either of
the two hills of the  IM240  driving cycle (i.e., Bag  1  and Bag 2)
taken  separately  correlate  with  the  current two-mode  IM240.
Preliminary  results  based upon  a  sample  of  188  1983  and  newer
fuel-injected vehicles which  were  recruited  at the  Indiana  I/M
lane  and  subsequently  retested   under lab  conditions   (which
included each vehicle  receiving  an FTP)  suggest that  analysis of
Bag 2  (i.e.,  emissions sampled during the second hill of the IM240
driving cycle)  may be  about  as good  as the  full  IM240  when it
comes to identifying vehicles  that  would  pass  or fail on the basis
of the full test.  Using Bag-2 cutpoints of 0.60/12  for HC and CO
respectively,  and looking at  Bag-2  results  only,  90%  of the excess
HC emissions  and 84%  of the  excess CO emissions were identified,
with an EC rate  of  0.7%,  as  compared to the full IM240 using the
0.8/15 cutpoints only  (i.e.,  no Bag-2 cutpoints), which identified
82% and 85% of  the  excess HC and CO emission,  respectively,  with
an EC rate of 0.8%.   These findings come  with the caveat that they
are based upon a Bag 2  sample  which followed the Bag 1 portion of
the driving cycle,  meaning that Bag 2's high degree  of correlation
with  the  IM240  may be  the  result of  preconditioning  occuring
during the Bag  1 phase.   Even if such is,   in  fact,  the case,  the
prospect of  a shorter  overall test time  still seems  good since
adequate preconditioning  for  Bag 2 could probably be  obtained in
less than 93 seconds by modifying Bag 1 to  use a higher speed over
less time.

     To determine whether or  not preconditioning is  a factor,  EPA
plans  to  begin testing  a sample  of vehicles  using what  is,  in
effect, a three bag test, beginning with  the second  hill of the
Draft
-30-
2/26/92

-------
IM240 driving cycle up  front  (hence  no  possibility for  "Bag 1"
preconditioning)  followed  by a regular IM240.   This study should
help EPA determine  (1) whether or not preconditioning is a factor
in Bag 2's  high degree of correlation with the  full test  and (2)
whether preconditioning  would improve the correlation between Bag
1 and the  full test.  In  addition,  as mentioned  above,  EPA will
soon begin  collecting second-by-second data, which  will  allow us
to determine  whether  or not there is  some  point in the  testing
cycle by which  time  if vehicle X  is  emitting  at  a rate  Y,  it will
clearly pass or fail.

4.5  Estimating I/M Testing Credits for MQBILE4 1

     As  stated earlier,  the data from  the  Indiana  program were
analyzed and  re-assembled in  a manner which allows a comparison of
I/M  program  designs  over   a wide  range  of  time  frames  and
conditions, rather than  just  for the particular sample of vehicles
tested in  Indiana.   This method  for  estimating the  effect  of I/M
program  options on  exhaust  emissions  (i.e.,  the I/M  credit)  is
fairly simple.   Using the emission factor database,  the fraction
of total vehicle FTP emissions which is identified by a particular
short test  is determined for  each of four strata of vehicles based
on FTP emission level.  Using a  subsample of  vehicles  which have
been repaired,  the  emission  reductions attributable  to  these I/M-
triggered repairs is estimated for each strata.  The Tech4.1 model
is used to  calculate the emissions impact of a given short test by
reducing  the  total  FTP  emissions identified  at  each age  by the
estimated  emission  reductions  resulting  from I/M repairs.   When
the  fleet  average emission rates  are  recalculated by considering
the  strata, the  difference  between the  I/M  and non-I/M case is
stored as an  I/M credit for use in MOBILE4.1.

4.5.1     Tech4.1 Background  and Assumptions

     The  Tech4.1 model  divides  the  1981 and  newer  light-duty
gasoline vehicle  (LDGV)  sample into  several  groups.   The 1981 and
1982 model  years are kept separate from the 1983+ model years.  In
each model  year group, the vehicles are divided by technology type
into closed-loop  port  fuel injection  (PFI),  closed-loop throttle-
body fuel  injection (TBI), closed-loop carbureted  (Garb)  and all
 (carbureted and fuel injected) open-loop  (Oplp).   Further,  each of
these groups  are divided  into emission levels for  Normal,  High,
Very High  and Super emitters.   Table F-l in Appendix  F provides
details   on  national   fleet  averages  for  passenger  vehicle
distributions by model year and technology type;  Tables F-2 and F-
3 provide  data on emitter groups  by  model  year  group,  technology
type, emission levels and rates,  and mileage accumulation.

     The  model allows  a  separate IDE  and  repair  effectiveness
estimate for  each of these divisions of the data by I/M test type,
as  illustrated in Table 4-8.  It should be noted  that  the IDRs
listed in Table 4-8  for  the  traditional I/M tests (i.e.,  the idle
Draft                        -31-                     2/26/92

-------
and 2500 rpm/Idle tests)  are based upon historical emission factor
data gathered  at  EPA's Motor Vehicle  Emission Lab  (MVEL)  in Ann
Arbor,  Michigan,  as well  as  elsewhere, and  not  at the  Hammond,
Indiana test lane.   The IDRs mentioned elsewhere in  this report
(Appendices  G  and H,  for  example)  were derived  as part  of the
Hammond study,  and are not  divided by  emitter  group, as  is the
case in Table 4-8.

     In practice,  because  of small  sample  sizes, several  of the
divisions  represented  in  Table  4-8  share information.   _ In
particular,  the  small  amount of steady-state  Loaded/Idle testing
required that  all vehicles without  Loaded/Idle testing be assumed
to have the same short test result for Loaded/Idle testing as they
had for the  2500 rpm/Idle  test  for  the purpose of determining the
IDR for the Loaded/Idle test.

     For Super emitters  (vehicles  over 10  gpm HC or 150  gpm CO) ,
the  IDR is  the  same for  all  technologies,  but  is separate for
1981-82 and 1983+ vehicles.  Most 1981-82 vehicles are  carbureted.
Most 1983+ vehicles  are  fuel injected.  There are no  Super open-
loop vehicles in the sample.

     The two fuel injection groups in the 1981-82  grouping use the
same IDRs  for Very  High emitters  (vehicles  over 1.64 gpm  HC or
13.6 gpm CO),  High emitters (vehicles over  0.82 gpm HC  or 10.2 gpm
CO) and Normals.   In some cases,  such as  the High  emitters, the
1983+ open-loop and carbureted technologies were combined.

     Repair effectiveness  (Table 4-9) was determined by dividing  the
repaired sample  by technology  into  PFI, TBI  and  Carb.   Model year
grouping was not used.  To be eligible for  the repair  effectiveness
analysis,  a repaired  vehicle must  first  fail  the short test  of
interest before repairs,  and then after  repairs,   must  pass the same
short test.  Thus,  different samples of  repaired  vehicles were used
for each short test.  The sample was  then  ranked by  before  repair
emission  level  and divided into  four equal-sized  subgroups  of
increasingly more  severe emissions  failure.   The before  and after
repair emission levels  of each subgroup were then  determined.

     When plotted, before  repair  emission level versus after repair
emission  level, these four  emission  failure points represent  a
technology specific function used to determine repair effectiveness.
Generally,  the vehicles  with higher before repair  emission  levels
get  larger absolute emission  reductions from repairs,  but  do  not
reach as clean a level after repairs as  vehicles  which began with a
milder degree  of emission failure.  Before  repair  emission levels of
High, Very High  and Super  emitters in many cases will fall between
the calculated points,  and  so had their after repair emission levels
determined by  interpolation.  Before  repair emission  levels lower
than the lowest point were  interpolated between  the  low  point  and
zero.   Before  repair emission levels  above the  highest  point were
assumed to be the same as the highest point.
Draft                        -32-                      2/26/92

-------
     Since  few of the  repaired vehicles  had Loaded/Idle  or IM240
testing  data,  it  was  assumed  that  vehicles  repaired  using  a
Loaded/Idle test  and the IM240  test  would use the  same before and
after  repair   curve  as the  2500  rpm/Idle testing.   EPA  is being
conservative in assuming  that  vehicles failing the Loaded/Idle test
or the  IM240   test,  after repair,  will  have the same  after repair
emission level as we estimate for the  2500  rpm/Idle test vehicles.
However, since the  failure  rates  of  vehicles in the  high emitter
groups are  larger for the Loaded/Idle  test  and  the IM240 transient
test than  for  the 2500 rpm/Idle test,  the total emission reduction
due to repairs will be larger.

     As  an example,   the  zero  mile HC  emission  level  of Very High
emitters  for   1983+  PFI vehicles  is  2.019  gpm  and their  slope is
taken to be the same slope as the Normals (i.e.,  0.0115 gpm/10,000
miles)  (see Table F-2) .   At  5  years  old,  the  average  mileage of
these vehicles will  be  60,829  miles.   The non-I/M emission level is
therefore:

                     2.019 + .0115*6.0829 = 2.089 gpm

     Assuming  a 2500 rpm/Idle  test is done, the HC  IDR  (see Table
4-8) for this  group  is  0.6187,  or  nearly 62% of  the total emissions
from these  vehicles  is  identified by failing vehicles using the 2500
rpm/Idle  test.   Table  4-9  shows the  results  of  a data  analysis
indicating the  predicted average after  repair  levels given  the
before repair  emission  level.  The series of points in the table are
used  to  predict  the after repair emission  levels for  all emitter
groupings,  only  dependent on  the average  before repair  emission
level  for  that  group.    The  before  repair  emission   level  falls
between the two emission  levels 1.9846 and 3.9314.  The after repair
levels  for these emissions are 0.59231  and 1.0271  respectively.
Interpolating, the  after repair  level  for the  2.089  gpm  before
repair emission level is:

       0.59231+((2.089-1.9846)/(3.9314-1.9846))*(1.0271-0.59231)=0.6153

     Therefore the after repair HC  emission level for  5 year old,
1983+ PFI vehicles tested on the 2500  rpm/Idle test  is:
                %
               0.6187*0.6153 +  (1-0.6187)*2.089 - 1.1772  gpm

     Comparing the I/M  and non-I/M cases indicates the "I/M benefit"
among Very High emitters.

                       (2.089-1.1772)72.089 = 43.6%
Draft                        -33-                      2/26/92

-------
                                Table  4-8
SuDer Emitters
Test

Idle Test
Idle Test
2500/Idle
2500/Idle
Load/ Idle
Load/ Idle
IM240
IM240

Idle Test
Idle Test
2500/Idle
2500/Idle
Load/ Idle
Load/Idle
IM240
IM240
Model
Years
81-82
83+
81-82
83+
81-82
83+
81-82
83+

81-82
83+
81-82
83+
81-82
83+
81-82
83+
PFI

0.
0.
0.
0.
0.
0.
1.
1.

0.
0.
0.
0.
0.
0.
0.
0.
fld
6048
8978
6523
8978
6523
8978
0000
0000

2736
5676
2736
6187
2736
6187
8920
8800
PFI
CQ
0.6968
0.9656
0.8577
0.9656
0.8577
0.9656
1.0000
1.0000
Very
0.3231
0.6129
0.3231
0.7465
0.3231
0.7465
0.9460
0.9400
TBI
HC.
0.6048
0.8978
0.6523
0.8978
0.6523
0.8978
1.0000
1.0000
TBI

0.
0.
0.
0-
0.
0-
1.
1.
£Q
6968
9656
8577
9656
8577
9656
0000
0000
Garb
HG
0.6048
0.8978
0.6523
0.8978
0.6523
0.8978
1.0000
1.0000
Garb
£0.
0.6968
0.9656
0.8577
0.9656
0.8577
0.9656
1.0000
1.0000
Oplp
HC
0.0000
0.0000
0-0000
0.0000
0.0000
0.0000
0.0000
0.0000
Oplp
£Q
0.0000
0.0000
0.0000
0-0000
0-0000
0.0000
0.0000
0.0000
Hicrh Emitters
0.2736
0.2651
0.2736
0.3616
0.2736
0.3904
0.8770
0.8600
0.
0.
0.
0.
0.
0.
0.
0.
3231
2695
3231
4206
3231
4337
8750
8600
0.3858
0.3640
0.4789
0.5684
0.5476
0.5684
0.8760
0.9400
0.4108
0.3180
0.5331
0.6832
0.6037
0.6832
0.8680
0.8300
0.4568
0.3640
0.6197
0.5684
0.6197
0.5684
0.8760
0.9400
0.5194
0.3180
0.6162
0.6832
0.6162
0.6832
0.8680
0.8300
High Emitters
Idle Test
Idle Test
2500/Idle
2500/Idle
Load/ Idle
Load/ Idle
IM240
IM240

Idle Test
Idle Test
2500/Idle
2500/Idle
Load/Idle
Load/ Idle
IM240
IM240
81-82
83+
81-82
83+
81-82
83+
81-82
83+

81-82
83+
81-82
83+
81-82
83+
81-82
83+
0.
0.
0.
0.
0.
0.
0.
0.

0.
0.
0.
0.
0.
0.
0.
0.
0506
2507
0506
3436
0506
3866
0930
1300

0556
0360
0556
0575
0556
0907
0450
0500
0.1135
0.2208
0.1135
0.3501
0.1135
0.3937
0-0600
0.0800
Nor
0.0774
0.0414
0.0774
0,0694
0.0774
0.1023
0.0560
0.0600
0.0506
0.0336
0.0506
0.1924
0.0506
0.1924
0.5080
0.5100
znaJL EmJL
0.0139
0.0425
0.0139
0.0476
0.0139
0.0712
0.0970
0.1000
0.
0.
0.
0.
0.
0.
0.
0.
1135
0613
1135
1532
1135
1532
4190
4200
0.0563
0-0694
0.0898
0.0694
0.0910
0.0694
0.1820
0.1800
0.0492
0.0415
0.0834
0.0415
0.0896
0-0415
0.2060
0.2200
0.2274
0.0694
0.2274
0.0694
0.2274
0-0694
0.1820
0.1800
0.1522
0.0415
0.1522
0.0415
0.1522
0.0415
0.2060
0.2200
tters
0.
0-
0.
0.
0.
0.
0.
0.
0139
0436
0139
0514
0139
0739
0750
0800
0.0188
0.0023
0.0371
0-0140
0.0371
0.0140
0.1340
0.2400
0-0204
0.0078
0.0427
0.0156
0-0427
0.0156
0.1200
0.2100
0.0093
0.0023
0-0201
0.0065
0.0201
0-0231
0.1340
0.2400
0.0131
0.0078
0.0317
0.0208
0.0317
0.0403
0.1200
0.2100
   Identification  Rate  (IDR)  is the fraction of the total  sample  emissions
   from vehicles failing the short test.
Draft
-34-
2/26/92

-------
                             Table  4-9

                  Short Test  Repair Effectiveness

                                PFI/TBI            Carb/Oplp
                          Before     After     Before     After
                          Repair     Repair    Repair     Repair
      Idle Test      HC
      Idle Test      HC
      Idle Test      HC
      Idle Test      HC
      Idle Test
      Idle Test
      Idle Test
      Idle Test
    2500 rpm/Idle*   HC?
    2500 rpm/Idle*   HC
    2500 rpm/Idle*   HC
    2500 rpm/Idle*   HC

    2500 rpm/Idle*   CO
    2500 rpm/Idle*   CO
    2500 rpm/Idle*   CO
    2500 rpm/Idle*   CO
 0.7400
 1.9223
 3.9023
14.2820
 0.4108
 0.6062
 1.0769
 1.3808
0.9677
2.0226
3.1063
8.5543
0.6224
1.1894
1.3254
1.5286
CO
CO
CO
CO
9.2708
28.0310
90.0380
190.6600
4.9900
9.4669
12.1480
20.6200
10.4870
29.5500
53.5200
134.7500
9.8624
12.9690
17.4340
18.2810
 0.8267
 1.9846
 3.9314
14.2820

10.3340
35.5180
104.5000
190.6600
0.4075
0.5923
1.0271 ,
1.3808 ''
0.9303
1.9431
2.9862
8.2523
 4.8950  i  10.6220
 9.8631    29.0530
11.9250   54.2820
20.6200  136.9700
		J
           0.5764
           1.0349
           1.1413
           1.4141

           9.2808
          12.4890
          13.1900
          13.5960
    Also used for Loaded/Idle and IM240 repair effects.
     In  the   Tech4.1  model,  the   technologies  and  emission
categories  are combined before  an  average  I/M benefit  for the
model year is calculated.

4.5.2   Evaporative  and   Running   Loss  Modeling,   and   the
        Effectiveness  of Purge/Pressure Testing

     A large  part of the additional  emission reduction available
through  the  use  of  high-tech I/M  tests  is the  result  of the
evaporative and  running loss emission reductions  achieved by the
repair of vehicles  which fail  the new evaporative system pressure
and purge tests.  The effectiveness of evaporative system pressure
and  purge  checks  in  reducing the   rate  of pressure and purge
problems was  calculated assuming that programs  with  these checks
would detect  100% of  all problems detected by the  EPA checks run
in the  Hammond I/M program.   This assumes that  the  program will
use methods similar to  the procedures used in  Indiana.  Although
all of the pressure and purge problems are assumed to be detected,
since some  problems will re-occur with time, the  average rate of
problems over the inspection cycle will not  be zero.
Draft
   -35-
                  2/26/92

-------
     For purposes  of  determination of program effectiveness,  the
combined evaporative system pressure and purge failure  rates  from
over 2,400 vehicles  tested in Indiana were  used.   The  resulting
effectiveness estimates were then used for  application of pressure
checks,  purge checks and combined pressure  and purge checks in the
MOBILE4.1 model.

     The average reduction in the rate of failure is calculated by
determining  the  rate of  failure  at the  midpoint  between  two
vehicle  ages.    The effect  of inspection  can  be  visualized  by
plotting the non-program rate over age with  the  calculated before
and after repairs  failure  rate estimates assuming  inspection  (see
figure in Appendix  D) .   At each age, vehicles due  for  inspection
are checked and necessary  repairs made.  Between inspections,  the
rate  of  failures  increases  until  the  vehicles  are  due   for
inspection again.   The  slope  of this failure  rate line  between
inspections is assumed to be equal to the slope of the non-program
line  for that vehicle  age.   This  creates  a  rising and  falling
pattern of rates resembling a saw blade.   The average  reduction in
rates is then  the average  value  of the "saw teeth" compared  with
the non-program case.

     With an  inspection program, at  age  zero,  when the  calendar
year equals the model  year,  no  vehicles  are yet one year  old and
due for  inspection;  therefore,  no reductions are made.   Assuming
an  annual  inspection,  at  age  one,  25% of the model  year  is  one
year old or older.  Therefore, the rate at one year is  reduced by
25% to reflect repairs on the vehicles due  for inspection.   By the
second year,  all  vehicles are inspected each year and  the after
repair rate  is always  zero.   The failure rate  after a check  is
always zero,  since the  detection rate is  100%.   Therefore,  the
midpoint failure rate is half the number  of failures that occur in
that year, once inspections begin.  In the  biennial  case,  vehicles
are  inspected  every  other  year  and  the  rate  of  failures
accumulates in the years between  inspections.

     This method  was  used in a computer spreadsheet to  calculate
the reduction in  failures from  evaporative  system pressure  and
purge checks used  in the MOBILE4.1  model.    The  spreadsheet  is
shown in Appendix D with  and without  formulas.   The  spreadsheet
originally contained errors which resulted  in the benefits  used in
the MOBILE4.1  model to  be  smaller than the  estimates  reported in
this  document  (which are  based  upon the corrected spreadsheet)
The version of  MOBILE4.1 released  to the  public does not  yet
reflect these changes, although they will be incorporated into the
next MOBILE release.

4.5.3     Benefits of IM24Q NOx Inspections

     None  of the  existing I/M  program  models  or  the  MOBILE4 1
model itself  are  designed  to estimate the effect of NOX  emission
inspection as part of an I/M program.  Therefore, to  estimate the
Draft                       -36-
                                                      2/26/92

-------
effect  of  an  IM240-based  NOX  inspection,  a  simple  model  was
developed.

     A  sample  of over 3,200  1983  and newer model year  vehicles,
tested  in  Hammond, Indiana  using the IM240  test procedure,  was
analyzed.   The sample was divided into  three technology  groups:
multi-point fuel injection vehicles,  throttle-body fuel  injection
vehicles and  carbureted vehicles.   Two NOX  cutpoint cases were
examined for each  technology, one  with a 10%  failure  rate  and  one
with a 20% failure  rate.

     Using  an  emission correlation  mapping  between  IM240  NOX
measurements  and  NOX measured on the  FTP, an FTP  NOX emission
level  was  estimated  for  each vehicle in  the sample.   A  linear
least-square regression  was  run for  estimated  FTP  NOX  emissions
versus mileage for each technology for two  model year  groups: 1983
through  1985  model year vehicles  and 1986 and newer model year
vehicles.   The  regressions were then  run again excluding vehicles
which  fail  the IM240 NOX inspection  first  using  the 10%  failure
rate  cutpoints  and  then the  20%  failure  rate  cutpoints.    The
exclusion  of  the  higher NOX emitters was  intended  to  represent
their deletion from the fleet through  repairs.

     Using  the  technology  mix used in MOBILE4.1,  the regressions
were weighted together to produce emission  factor  zero mile levels
and  deterioration  rates  for each model  year from  1983  through
1992.   The difference  in  the emission levels  between  the  cases
with and without NOX  failures removed is assumed to be the  benefit
from  the  IM240 NOX  emission test with only  NOx  related  repairs
performed.   Results are shown in Table 4-10.

     Since  it  is expected  that  most NOX emission  testing  will be
done along  with testing  for  HC  and CO emissions,  the side effect
of HC  and CO repairs  on NOX emissions should  also  be  accounted
for.   This effect is ignored  in the standard MOBILE4.1  model.
Typically  NOX  emissions  will increase,  on  average,  when HC  an CO
emission repairs are  performed.   The extent of this NOX emission
disbenefit  was  determined  by  calculating average  NOX emission
levels  corresponding  to  the Normal,  High,  Very High  and  Super
HC/CO emitter categories  used in the MOBILE4.1 Tech4  model.

     Using  the  post-repair emission  levels of  the  same vehicles
used to calculate  the after  repair emission levels  for HC  and CO
emissions,  the NOX emission levels of these vehicles  after  repairs
were determined.   These NO>:  emission  levels are not the  result of
NOX  related  repairs, but  a by-product  of  HC   and  CO emission
repairs.   Using the  standard HC/CO 2500 rpm/Idle test  IDRs  along
with  the  repair effects on  NOX  and the  NOX emission  rates by
emitter group  in the Tech4 model,  the effect of  NOX disbenefits
was determined for  each age of each model year.  See  Table 4-11.
Draft                        -37-                      2/26/92

-------
      The  NOX disbenefits,  as  a percent change,  are applied to the
 emission  levels estimated  from  the regression equations  at each
 age.   The  resulting NOX  emission  levels by  age are  regressed
 versus  mileage  for  each model  year to  give  the  final  emission
 factor equation  for NOX.  Comparing the emission factor results of
 the  baseline case with the  cases  with 10% or  20% NOX  emission
 testing failure  rates was  done to estimate the  benefits,  in tons,
 of the  IM240 NOX emission  test.   Results  are  shown in Table 4-12.
 For  example,  at age  5 and mean mileage of 60,829  miles,  the W20%
 fail" IM240 NOX  cutpoints  will reduce 1992  model year NOX from
 0.887 to  0.710 gpm, a reduction of 20%.

      The  final  emission factors  were  used as alternate input  to
 the  MOBILE4.1 model  and,  in  combination  with  the CEM4.1  model,
 used to calculate the  tons of NOX  emission  benefit  from  use  of
 IM240 NOX  cutpoints.   These benefits  were used in  applying  the
 cost  credit.   It should be noted that since  both  the  cases with
 and  without the  IM240 NOX  inspection cutpoints  should include  the
 disbenefits of  EC/CO repairs, the  disbenefits  do  not  effect  the
 calculation of  incremental NOX  reduction from IM240  cutpoints.
 For  simplicity  and consistency,  therefore, the disbenefits were
 not  applied to the I/M scenarios  involving  only HC/CO cutpoints.
Draft                       -38-                      2/26/92

-------
                                  Table 4-10




 Lane IM24Q  Based Emission  Factor Levels with  IM24Q  NOX Outpoints







Age    0123456789     10     11      12




Miles    0    1.3118  2.6058  3.8298  4.9876 6.0829  7.119  8.0991  9.0262  9.9031  10.7326  11.5172  12.2594




Year                                     Base
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992

1983
1984
1985
1986
1987
1988
1989
1990
1991
1992

1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1.146
1.048
0.983
0.608
0.593
0.561
0.570
0.550
0.547
0.547

1.078
1.036
0.970
0.600
0.591
0.559
0.556
0.529
0.525
0.523

0.985
0.955
0.911
0.591
0.582
0.551
0.550
0.525
0.521
0.520
1.197
1.115
1.049
0.683
0.669
0.633
0.639
0.614
0.610
0.610

1.092
1.051
0.984
0.650
0.640
0.603
0.600
0.569
0.564
0.562

0.992
0.962
0.917
0.629
0.619
0.586
0.585
0.558
0.553
0.552
1.248
1.181
1.114
0.758
0.744
0.705
0.707
0.677
0.672
0.671

1.106
1.066
0.998
0.699
0.688
0.647
0.643
0.608
0.602
0.600

0.999
0.968
0.922
0.666
0.656
0.621
0.619
0.590
0.585
0.584
1.296
1.243
1.175
0.828
0.815
0.773
0.772
0.737
0.731
0.729

1.120
1.080
1.011
0.746
0.734
0.688
0.684
0.645
0.639
0.636

1.006
0.974
0.927
0.702
0.691
0.654
0.652
0.621
0.616
0.614
1.341
1.303
1.233
0.895
0.882
0.837
0.833
0.794
0.787
0.784

1.132
1.093
1.023
0.790
0.777
0.727
0.723
0.681
0.673
0.670

1.012
0.979
0.932
0.735
0.724
0.685
0.683
0.650
0.644
0.642
1.384
1.359
1.288
0.958
0.946
0.897
0.890
0.847
0.840
0.837

1.144
1.105
1.034
0.831
0.817
0.764
0.759
0.714
0.706
0.703

1.018
0.984
0.937
0.766
0.755
0.715
0.712
0.677
0.671
0.669
1.425
1.412
1.340
1.017
1.006
0.954
0.945
0.898
0.889
0.886
10% Fail
1.155
1.117
1.045
0.871
0.856
0.800
0.794
0.745
0.737
0.733
20% Fail
1.023
0.989
0.941
0.796
0.785
0.743
0.739
0.703
0.697
0.695
1.463
1.462
1.389
1.074
1.063
1.009
0.997
0.946
0.936
0.932

1.166
1.128
1.056
0.908
0.893
0.833
0.827
0.775
0.766
0.762

1.028
0.994
0.945
0.825
0.813
0.769
0.765
0.728
0.721
0.719
1.499
1.509
1.436
1.127
1.117
1.060
1.046
0.991
0.981
0.977

1.176
1.139
1.065
0.943
0.927
0.864
0.858
0.803
0.793
0.790

1.033
0.999
0.949
0.851
0.839
0.794
0.790
0.751
0.744
0.741
1.533
1.554
1.480
1.177
1.168
1.108
1.092
1.034
1.023
1.018

1.185
1.149
1.074
0.977
0.960
0.894
0.887
0.830
0.820
0.816

1.038
1.003
0.953
0.877
0.864
0.818
0.813
0.773
0.766
0.763
1.566
1.596
1.521
1.225
1.216
1.154
1.136
1.075
1.063
1.058

1.194
1.158
1.083
1.008
0.991
0.922
0.914
0.855
0.844
0.840

1.042
1.007
0.956
0.901
0.888
0.840
0.835
0.793
0.786
0.783
1.597
1.636
1.561
1.270
1.262
1.198
1.177
1.113
1.101
1.095

1.203
1.167
1.091
1.038
1.020
0.948
0.941
0.879
0.868
0.864

1.046
1.011
0.959
0.923
0.910
0.861
0.856
0.813
0.805
0.803
1.626
1.674
1.598
1.313
1.305
1.239
1.216
1.149
1.136
1.130

1.211
1.176
1.099
1.067
1.047
0.973
0.965
0.901
0.890
0.885

1.050
1.014
0.962
0.945
0.932
0.881
0.875
0.832
0.824
0.821
Draft
-39-
2/26/92

-------
                                            Table  4-10




                                          -  continued  -









 13     14    15     16    17     18     19     20    21     22    23     24    25    Age    Regression



12.9615  13.6257   14.254   14.8483   15.4104  15.9421   16.4451   16.9209   17.3712  17.7969   18.1997  18.5806  18.941    Miles  ZML   DET




                                                     Base
1.653
1.710
1.633
1.353
1.345
1.278
1.253
1.183
1.170
1.164
1.679
1.744
1.666
1.392
1.384
1.314
1.288
1.216
1.202
1.195
1.704
1.776
1.698
1.428
1.420
1.349
1.321
1.247
1.232
1.225
1.727
1.807
1.728
1.462
1.455
1.382
1.353
1.276
1.261
1.254
1.749
1.835
1.756
1.494
1.488
1.413
1.382
1.303
1.288
1.280
1.770
1.863
1.782
1.525
1.519
1.442
1.410
1.329
1.313
1.306
1.790
1.888
1.808
1.554
1.548
1.470
1.437
1.354
1.338
1.329
1.808
1.913
1.831
1.581
1.575
1.497
1.462
1.377
1.360
1.352
1.826
1.936
1.854
1.607
1.602
1.521
1.486
1.399
1.382
1.374
1.842
1.957
1.875
1.632
1.626
1.545
1.508
1.420
1.402
1.394
1.858
1.978
1.896
1.655
1.650
1.567
1.530
1.439
1.422
1.413
1.873
1.998
1.915
1.677
1.672
1.588
1.550
1.458
1.440
1.431
1.887
2.016
1.933
1.698
1.693
1.608
1.569
1.476
1.457
1.448
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1.146
1.048
0.983
0.608
0.593
0.561
0.570
0.550
0.547
0.547
0.0391
0.0511
0.0501
0.0575
0.0581
0.0553
0.0527
0.0489
0.0481
0.0476
                                                     10% Fail
1.218
1.184
1.107
1.093
1.074
0.997
0.989
0.923
0.911
0.906
1.226
1.191
1.114
1.119
1.098
1.020
1.011
0.943
0.931
0.926
1.232
1.198
1.120
1.143
1.122
1.041
1.032
0.962
0.949
0.944
1.239
1.205
1.126
1.165
1.144
1.061
1.052
0.980
0.967
0.962
1.245
1.211
1.132
1.187
1.165
1.080
1.071
0.997
0.984
0.979
1.251
1.218
1.138
1.207
1.185
1.098
1.088
1.013
1.000
0.994
1.256
1.223
1.143
1.226
1.203
1.115
1.105
1.028
1.015
1.009
1.261
1.229
1.148
1.244
1.221
1.131
1.121
1.043
1.029
1.023
1.266
1.234
1.153
1.261
1.238
1.146
1.136
1.056
1.042
1.037
1.271
1.239
1.157
1.277
1.254
1.161
1.150
1.069
1.055
1.049
1.275
1.243
1.162
1.293
1.269
1.174
1.164
1.081
1.067
1.061
1.279
1.248
1.166
1.307
1.283
1.187
1.176
1.093
1.078
1.072
1.283
1.252
1.169
1.321
1.296
1.199
1.188
1.104
1.089
1.083
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1.078
1.036
0.970
0.600
0.591
0.559
0.556
0.529
0.525
0.523
0.0108
0.0114
0.0105
0.0381
0.0372
0.0338
0.0334
0.0303
0.0298
0.0296
                                                     20% Fail
1.054
1.018
0.965
0.965
0.952
0.900
0.894
0.849
0.841
0.838
1.058
1.021
0.968
0.984
0.971
0.918
0.912
0.866
0.858
0.854
1.061
1.024
0.971
1.002
0.988
0.935
0.928
0.882
0.873
0.870
1.064
1.027
0.973
1.019
1.005
0.951
0.944
0.896
0.888
0.885
1.067
1.029
0.976
1.035
1.021
0.966
0.959
0.910
0.902
0.898
1.070
1.032
0.978
1.051
1.037
0.980
0.973
0.924
0.915
0.911
1.073
1.034
0.980
1.065
1.051
0.994
0.987
0.936
0.927
0.924
1.075
1.037
0.982
1.079
1.065
1.006
0.999
0.948
0.939
0.935
1.077
1.039
0.984
1.092
1.077
1.019
1.011
0.959
0.950
0.947
1.080
1.041
0.986
1.104
1.090
1.030
1.022
0.970
0.961
0.957
1.082
1.043
0.987
1.116
1.101
1.041
1.033
0.980
0.971
0.967
1.084
1.045
0.989
1.127
1.112
1.051
1.043
0.990
0.980
0.976
1.086
1.046
0.990
1.137
1.122
1.061
1.053
0.999
0.989
0.985
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
0.985
0.955
0.911
0.591
0.582
0.551
0.550
0.525
0.521
0.520
0.0053
0.0048
0.0042
0.0288
0.0285
0.0269
0.0265
0.0250
0.0247
0.0246
   Draft
-40-
                                                                                  2/26/92

-------
                                  Table 4-11




                   Side  Effects of  I/M  on NOx Emissions




                       (Disbenefit of  HC/CO  repairs)






 Age    012     3     4      56789     10     11     12



Miles   0   1.3118  2.6058  3.8298 4.9876  6.0829  7.119  8.0991  9.0262 9.9031  10.7326  11.5172  12.2594
 Year
      Base
1983 0.61
1984 0.63
1985 0.63
1986 0.51
1987 0.50
1988 0.47
1989 0.47
1990 0.46
1991 0.45
1992 0.45

1983 0.61
1984 0.63
1985 0.63
1986 0.51
1987 0.50
1988 0.47
1989 0.47
1990 0.46
1991 0.45
1992 0.45
0.67
0.69
0.69
0.56
0.55
0.52
0.52
0.50
0.50
0.50

0.69
0.71
0.70
0.58
0.56
0.53
0.54
0.52
0.52
0.52
0.74
0.75
0.74
0.62
0.60
0.56
0.57
0.55
0.55
0.55

0.76
0.77
0.76
0.64
0.61
0.58
0.59
0.57
0.57
0.57
0.80
0.81
0.79
0.67
0.64
0.60
0.62
0.59
0.59
0.59

0.82
0.83
0.81
0.69
0.66
0.63
0.64
0.62
0.62
0.62
0.86
0.87
0.84
0.72
0.69
0.64
0.66
0.64
0.63
0.63

0.88
0.88
0.86
0.74
0.71
0.67
0.69
0.66
0.66
0.66
0.91
0.92
0.89
0.76
0.73
0.68
0.70
0.68
0.67
0.67

0.93
0.93
0.90
0.78
0.75
0.71
0.73
0.70
0.70
0.70
0.98
0.99
0.96
0.81
0.77
0.72
0.74
0.71
0.71
0.71
Idle
0.99
1.00
0.97
0.83
0.80
0.75
0.77
0.74
0.74
0.74
1.04
1.05
1.02
0.85
0.81
0.75
0.78
0.75
0.75
0.75

1.05
1.06
1.04
0.87
0.84
0.78
0.81
0.78
0.78
0.78
1.10
1.11
1.09
0.89
0.84
0.78
0.81
0.78
0.78
0.78

1.10
1.12
1.09
0.91
0.87
0.82
0.85
0.82
0.82
0.82
1.15
1.17
1.14
0.92
0.88
0.81
0.85
0.82
0.81
0.82

1.15
1.17
1.15
0.95
0.91
0.85
0.88
0.85
0.85
0.85
1.20
1.22
1.20
0.96
0.91
0.84
0.88
0.85
0.85
0.85

1.20
1.22
1.20
0.98
0.94
0.88
0.91
0.88
0.88
0.88
1.25
1.27
1.25
0.99
0.94
0.87
0.91
0.88
0.87
0.88

1.25
1.27
1.25
1.02
0.97
0.90
0.94
0.91
0.91
0.91
1.30
1.32
1.30
1.02
0.97
0.89
0.94
0.90
0.90
0.90

1.29
1.31
1.30
1.05
1.00
0.93
0.97
0.94
0.94
0.94
Two Speed
1983 0.61
1984 0.63
1985 0.63
1986 0.51
1987 0.50
1988 0.47
1989 0.47
1990 0.46
1991 0.45
1992 0.45
0.69
0.71
0.71
0.58
0.57
0.54
0.55
0.53
0.53
0.53
0.76
0.78
0.76
0.64
0.62
0.59
0.60
0.58
0.58
0.58
0.83
0.83
0.82
0.70
0.67
0.63
0.65
0.63
0.63
0.63
0.88
0.88
0.86
0.75
0.72
0.68
0.70
0.67
0.67
0.67
0.93
0.93
0.91
0.79
0.76
0.72
0.74
0.72
0.71
0.72
1.00
1.00
0.97
0.84
0.81
0.76
0.78
0.76
0.76
0.76
1.05
1.06
1.04
0.88
0.85
0.80
0.82
0.80
0.80
0.80
1.10
1.12
1.10
0.92
0.89
0.83
0.86
0.84
0.83
0.84
1.15
1.17
1.15
0.96
0.92
0.87
0.90
0.87
0.87
0.87
1.20
1.22
1.20
1.00
0.96
0.90
0.93
0.90
0.90
0.90
1.25
1.27
1.25
1.03
0.99
0.93
0.96
0.93
0.93
0.93
1.29
1.31
1.30
1.06
1.02
0.95
0.99
0.96
0.96
0.96
Draft
-41-
2/26/92

-------
                                             Table  4-11




                                           - continued  -









 13     14     15    16     17     18     19     20    21     22    23     24     25   Model    Regression



12.9615  13.6257  14.254   14.8483  15.4104   15.9421   16.4451  16.9209   17.3712  17.7969   18.1997   18.5806  18.941   Year    ZML   DBF




                                               Base
1.34
1.36
1.35
1.05
1.00
0.92
0.96
0.93
0.93
0.93

1.33
1.36
1.34
1.07
1.02
0.95
0.99
0.96
0.96
0.96
1.38
1.40
1.39
1.08
1.02
0.94
0.99
0.95
0.95
0.95

1.37
1.39
1.38
1.10
1.05
0.97
1.02
0.99
0.98
0.99
1.41
1.44
1.41
1.10
1.04
0.96
1.01
0.98
0.97
0.98

1.40
1.43
1.41
1.13
1.07
0.99
1.04
1.01
1.01
1.01
1.44
1.46
1.44
1.13
1.07
0.98
1.03
1.00
1.00
1.00

1.42
1.45
1.42
1.15
1.09
1.01
1.06
1.03
1.03
1.03
1.47
1.49
1.45
1.15
1.09
1.00
1.05
1.02
1.02
1.02

1.44
1.47
1.44
1.17
1.11
1.03
1.08
1.05
1.04
1.05
1.49
1.51
1.47
1.17
1.11
1.02
1.07
1.04
1.03
1.04

1.46
1.49
1.46
1.19
1.13
1.05
1.10
1.06
1.06
1.07
1.51
1.53
1.48
1.19
1.13
1.04
1.09
1.05
1.05
1.06

1.48
1.50
1.47
1.21
1.15
1.07
1.12
1.08
1.08
1.08
1.53
1.54
1.50
1.21
1.15
1.05
1.11
1.07
1.07
1.07
Idle
1.50
1.52
1.48
1.23
1.17
1.08
1.13
1.10
1.09
1.10
1.55
1.56
1.51
1.23
1.16
1.07
1.13
1.09
1.09
1.09

1.52
1.53
1.49
1.25
1.18
1.10
1.15
1.11
1.11
1.11
1.56
1.58
1.53
1.25
1.18
1.08
1.14
1.10
1.10
1.11

1.53
1.55
1.50
1.26
1.20
1.11
1.16
1.12
1.12
1.13
1.58
1.59
1.54
1.26
1.19
1.10
1.16
1.12
1.12
1.12

1.54
1.56
1.51
1.28
1.21
1.12
1.17
1.14
1.14
1.14
1.60
1.60
1.55
1.28
1.21
1.11
1.17
1.13
1.13
1.13

1.56
1.57
1.52
1.29
1.22
1.13
1.19
1.15
1.15
1.15
1.61
1.62
1.56
1.29
1.22
1.12
1.18
1.14
1.14
1.15

1.57
1.58
1.53
1.30
1.24
1.14
1.20
1.16
1.16
1.16
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992

1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
0.601 0.0550
0.617 0.0549
0.614 0.0527
0.511 0.0415
0.497 0.0383
0.470 0.0345
0.475 0.0375
0.455 0.0364
0.453 0.0365
0.452 0.0367

0.628 0.0519
0.640 0.0524
0.637 0.0505
0.530 0.0414
0.516 0.0385
0.491 0.0350
0.497 0.0377
0.479 0.0367
0.477 0.0367
0.476 0.0369
Two Speed
1.33
1.35
1.34
1.09
1.04
0.98
1.01
0.98
0.98
0.98
1.36
1.39
1.38
1.12
1.07
1.00
1.04
1.01
1.00
1.01
1.39
1.42
1.40
1.14
1.09
1.02
1.06
1.03
1.03
1.03
1.42
1.45
1.42
1.17
1.11
1.04
1.08
1.05
1.05
1.05
1.44
1.46
1.44
1.19
1.13
1.06
1.10
1.07
1.07
1.07
1.46
1.48
1.45
1.21
1.15
1.08
1.12
1.09
1.08
1.09
1.47
1.50
1.46
1.23
1.17
1.09
1.14
1.10
1.10
1.10
1.49
1.51
1.48
1.24
1.19
1.11
1.15
1.12
1.11
1.12
1.51
1.53
1.49
1.26
1.20
1.12
1.17
1.13
1.13
1.13
1.52
1.54
1.50
1.28
1.22
1.13
1.18
1.14
1.14
1.15
1.53
1.55
1.51
1.29
1.23
1.14
1.19
1.16
1.15
1.16
1.55
1.56
1.52
1.30
1.24
1.16
1.21
1.17
1.17
1.17
1.56
1.57
1.53
1.32
1.25
1.17
1.22
1.18
1.18
1.18
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
0.637 0.0509
0.645 0.0516
0.642 0.0499
0.535 0.0421
0.521 0.0395
0.497 0.0362
0.503 0.0385
0.487 0.0374
0.485 0.0374
0.485 0.0376
    Draft
-42-
2/26/92

-------
                                  Table 4-12




         Lane IM24Q Based. Emission  Factors with.  IM24Q  Outpoints




               with Disbenefits of  HC/CQ Repairs Included









Age   0     1     2     3     4     5     6     7     8     9     10     11      12



Miles   0   1.3118 2.6058  3.8298  4.9876 6.0829  7.119   8.0991  9.0262  9.9031   10.7326  11.5172  12.2594
Year
                                       Base
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992

1983
1984
1985
1986
1987
1988
1989
1990
1991
1992

1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1.146
1.048
0.983
0.608
0.593
0.561
0.570
0.550
0.547
0.547

1.078
1.036
0.970
0.600
0.591
0.559
0.556
0.529
0.525
0.523

0.985
0.955
0.911
0.591
0.582
0.551
0.550
0.525
0.521
0.520
1.235
1.145
1.081
0.707
0.695
0.662
0.667
0.646
0.642
0.641

1.126
1.080
1.014
0.673
0.664
0.630
0.627
0.599
0.593
0.591

1.023
0.988
0.945
0.651
0.643
0.613
0.611
0.587
0.582
0.580
1.288
1.214
1.147
0.787
0.775
0.740
0.742
0.714
0.710
0.709

1.142
1.095
1.027
0.726
0.717
0.679
0.675
0.642
0.636
0.634

1.032
0.995
0.950
0.692
0.684
0.652
0.650
0.623
0.618
0.617
1.336
1.271
1.208
0.861
0.851
0.814
0.812
0.781
0.775
0.774

1.154
1.104
1.039
0.776
0.766
0.725
0.720
0.684
0.676
0.675

1.037
0.995
0.953
0.730
0.721
0.689
0.686
0.658
0.652
0.651
1.374
1.325
1.260
0.930
0.922
0.884
0.877
0.841
0.835
0.833

1.160
1.112
1.045
0.821
0.812
0.768
0.761
0.721
0.715
0.712

1.037
0.996
0.952
0.764
0.757
0.724
0.719
0.689
0.684
0.682
1.413
1.378
1.313
0.995
0.989
0.949
0.939
0.899
0.891
0.887

1.168
1.121
1.054
0.864
0.854
0.808
0.801
0.757
0.749
0.746

1.039
0.998
0.955
0.797
0.790
0.756
0.750
0.718
0.712
0.710
1.448
1.427
1.361
1.059
1.053
1.013
0.999
0.956
0.947
0.943
10% Fail
1.174
1.130
1.062
0.907
0.896
0.849
0.839
0.793
0.784
0.781
20% Fail
1.040
1.001
0.956
0.829
0.822
0.788
0.781
0.749
0.742
0.739
1.480
1.473
1.406
1.119
1.114
1.072
1.054
1.006
0.998
0.993

1.179
1.137
1.068
0.947
0.936
0.885
0.874
0.825
0.816
0.812

1.040
1.002
0.956
0.860
0.852
0.817
0.809
0.774
0.768
0.766
1.509
1.517
1.448
1.175
1.172
1.128
1.106
1.056
1.045
1.040

1.183
1.145
1.074
0.984
0.973
0.919
0.907
0.855
0.845
0.841

1.040
1.004
0.957
0.888
0.881
0.845
0.835
0.800
0.792
0.790
1.536
1.558
1.486
1.227
1.226
1.181
1.155
1.101
1.091
1.084

1.187
1.152
1.079
1.018
1.008
0.952
0.938
0.884
0.874
0.869

1.040
1.005
0.957
0.914
0.908
0.871
0.860
0.823
0.816
0.813
1.565
1.595
1.526
1.275
1.278
1.230
1.200
1.143
1.131
1.125

1.193
1.157
1.087
1.050
1.041
0.982
0.967
0.909
0.898
0.894

1.041
1.006
0.959
0.937
0.933
0.895
0.883
0.844
0.836
0.833
1.590
1.631
1.561
1.323
1.325
1.275
1.243
1.183
1.170
1.164

1.198
1.164
1.091
1.081
1.071
1.009
0.993
0.934
0.922
0.918

1.042
1.007
0.959
0.961
0.956
0.916
0.904
0.864
0.856
0.853
1.614
1.666
1.594
1.365
1.368
1.319
1.283
1.219
1.206
1.199

1.202
1.169
1.097
1.108
1.098
1.036
1.018
0.956
0.944
0.939

1.043
1.009
0.960
0.982
0.977
0.938
0.923
0.882
0.874
0.871
 Draft
-43-
2/26/92

-------
                                           Table  4-12




                                         -  continued  -









 13    14     15     16     17     18     19     20    21     22    23     24    25    Model   Regression



12.9615  13.6257   14.254   14.8483  154104  15.9421   164451   16.9209  17.3712   17.7969  18.1997   18.5806   18.941   Year  ZML    DBF




                                              Base
1.637 1.660
1.699 1.728
1.626 1.655
1.405 1.443
1.409 1.448
1.357 1.395
1.320 1.354
1.252 1.285
1.240 1.269
1.231 1.262
1.680
1.757
1.685
1.478
1.484
1.429
1.387
1.314
1.299
1.290
1.698
1.783
1.711
1.510
1.519
1.462
1.417
1.342
1.327
1.318
1.716
1.807
1.735
1.541
1.549
1.493
1.445
1.367
1.351
1.342
1.733
1.832
1.761
1.571
1.579
1.522
1.471
1.392
1.376
1.366
1.748
1.854
1.783
1.597
1.607
1.547
1.496
1.414
1.396
1.387
1.763
1.875
1.803
1.623
1.632
1.570
1.519
1.435
1.416
1.406
1.779
1.895
1.823
1.645
1.655
1.594
1.539
1.454
1.435
1.425
1.791
1.913
1.842
1.667
1.679
1.616
1.560
1.473
1.453
1.444
1.803
1.931
1.860
1.689
1.699
1.634
1.578
1.488
1.470
1.460
1.816
1.948
1.878
1.707
1.719
1.653
1.596
1.506
1.486
1.475
1.825
1.962
1.893
1.726
1.738
1.671
1.612
1.520
1.501
1.490
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1.196
1.087
1.022
0.638
0.623
0.594
0.605
0.589
0.587
0.587
0.0338
0.0468
0.0463
0.0584
0.0599
0.0580
0.0543
0.0503
0.0494
0.0488
10% Fail
1.207 1.211
1.176 1.180
1.102 1.106
1.135 1.160
1.124 1.149
1.059 1.082
1.041 1.062
0.976 0.996
0.965 0.982
0.959 0.977
1.215
1.185
1.111
1.183
1.172
1.102
1.083
1.014
1.001
0.995
1.218
1.189
1.115
1.203
1.194
1.122
1.102
1.031
1.017
1.011
1.221
1.193
1.119
1.224
1.213
1.141
1.119
1.046
1.032
1.026
1.225
1.197
1.124
1.243
1.232
1.158
1.135
1.061
1.047
1.040
1.227
1.201
1.128
1.260
1.249
1.173
1.151
1.074
1.059
1.053
1.230
1.205
1.131
1.277
1.265
1.187
1.164
1.086
1.071
1.064
1.234
1.208
1.134
1.291
1.279
1.201
1.176
1.098
1.082
1.076
1.235
1.210
1.137
1.305
1.294
1.214
1.189
1.109
1.093
1.087
1.237
1.214
1.140
1.319
1.307
1.225
1.201
1.118
1.103
1.096
1.240
1.216
1.143
1.331
1.319
1.235
1.211
1.129
1.113
1.106
1.241
1.218
1.145
1.343
1.331
1.246
1.221
1.137
1.121
1.114
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1.118
1.067
1.001
0.628
0.619
0.590
0.588
0.564
0.560
0.558
0.0068
0.0082
0.0077
0.0385
0.0384
0.0355
0.0343
0.0311
0.0305
0.0302
20% Fail
1.044 1.045
1.011 1.011
0.961 0.962
1.002 1.021
0.997 1.015
0.956 0.974
0.942 0.958
0.899 0.915
0.891 0.905
0.887 0.902
1.046
1.012
0.963
1.038
1.033
0.990
0.974
0.929
0.921
0.916
1.046
1.013
0.964
1.053
1.050
1.006
0.989
0.943
0.934
0.930
1.047
1.014
0.964
1.068
1.064
1.021
1.003
0.955
0.946
0.942
1.048
1.015
0.966
1.082
1.078
1.034
1.015
0.967
0.958
0.954
1.048
1.015
0.967
1.095
1.091
1.045
1.027
0.978
0.968
0.964
1.048
1.017
0.967
1.107
1.103
1.056
1.038
0.988
0.978
0.973
1.050
1.017
0.967
1.118
1.114
1.067
1.047
0.997
0.987
0.982
1.049
1.017
0.968
1.128
1.125
1.078
1.057
1.006
0.996
0.992
1.050
1.018
0.969
1.139
1.134
1.085
1.066
1.013
1.004
0.999
1.051
1.019
0.970
1.147
1.143
1.094
1.074
1.022
1.011
1.006
1.050
1.019
0.970
1.157
1.153
1.102
1.082
1.028
1.018
1.013
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1.021
0.982
0.939
0.617
0.608
0.581
0.581
0.559
0.555
0.554
0.0018
0.0021
0.0017
0.0291
0.0294
0.0283
0.0272
0.0256
0.0252
0.0250
  Draft
-44-
2/26/92

-------
5.0  REGULATORY   IMPACT  ANALYSIS  - ESTIMATING  COST AND  COST

EFFECTIVENESS


5.1  Cost of Conventional I/M Testing

     EPA has  collected  and analyzed cost data from all operating
I/M programs that could provide the information.   EPA  has  analyzed
per vehicle costs in  I/M programs  based  upon  four  basic pieces of
information:  The I/M program agency  budget, number of  initial
tests,  the  fee  for  each test,  and the portion  returned to the
state or local government.   This discussion will deal with  three
aspects of I/M cost:  Inspection costs,  oversight  costs, and repair
costs.   In  addition,  it  should  be noted that where  programs are
referred  to  as  either  "centralized"  or  "decentralized,"  this
indicates  only  whether  or  not  the  functions  of  testing and
repairing are separated and performed by different  entities  (i.e.,
in centralized networks  - which can include multiple  independent
supplier test-only  networks - the  test  and repair functions are
separate, whereas  in  decentralized networks,  these functions are
typically performed at a single facility.

5.1.1     Inspection  and Administration  Costs

      Inspection  fees  are set  in one  of three  ways: By  a bid
process for a   contract  to  supply  inspection  services,  by
legislation or regulation establishing a maximum  fee,  or by market
forces.  Ideally the  fee is  scaled to cover the  cost  of providing
the  inspection,   cover  the  fee  to  the  state for  oversight and
management,  and  to provide  a reasonable profit  to the  operator
 (except in government-run programs).

      This ideal is not always met in actual  I/M programs.  In some
programs  the inspection fee  does not  include  a  share  for the
state's oversight costs, so these must be derived from the general
fund,   with  the  result  that  oversight   efforts   are  often
significantly underfunded.   In many decentralized programs the
maximum fee  is set below the actual cost  (with profit)  for the
test, so providers  must make up for that cost by  providing  other
goods and services.

      The economies  of scale  and efficiency of operation  in high
volume  test-only inspection  networks enable  motorists  in  these
programs to enjoy lower average inspection fees than in low volume
decentralized programs.   Based  upon  1989  I/M  audits (which
collected  information  from all  I/M programs),  and  taking into
account both inspection  and  oversight  costs,  decentralized
programs  using  computerized  analyzers  have  the  highest costs,
averaging  about   $17.70  per  vehicle;  centralized  contractor-run
programs average $8.42 per  vehicle  (recently gathered 1990 data
show  slightly different  numbers, although these have  not greatly
Draft                        -45-                      2/26/92

-------
 affected the  overall  averages).    Table  5-1  shows  the estimated
 cost  of the   I/M program  on a  per  vehicle  basis,  including
 inspection  and oversight costs.

                              Table  5-1

                     I/M Program Inspection Fees

        Decentralized Programs                 Centralized Programs

      Program      Cost Per Vehicle        Program      Cost Per Vehicle

     Anchorage            $32.00           Arizona             $6.00
     Fairbanks            $29.00          Connecticut          $10.00
    California           $48.39           Florida            $10.00
     Colorado            $11.20           Illinois             $8.07
      Georgia            $10.68          Louisville            $6.00
   Massachusetts          $17.18          Maryland             $8.53
     Michigan            $10.87          Minnesota            $8.00
     Missouri            $9.00           Nashville            $6.00
   North Carolina         $15.40          Washington            $9.00
    New Mexico           $16.00          Wisconsin            $8.73
      Nevada             $21.26
     New York            $19.92
   Pennsylvania           $9.01
      Dallas             $17.25
      El Paso            $17.25
   Davis County           $9.00
    Utah County           $9.71
   Salt Lake City         $11.49
     Virginia            $13.50

      In a  centralized,  contractor-run  program,  the  contractor
 bears the  cost  of acquiring land,  constructing  and  equipping
 inspection  facilities,  hiring and training  staff,  collecting and
 processing  data,  and doing  the routine  testing work.   The state's
 role  in  this  case is  to  make  sure  the  contractor  meets  its
 obligations and to study the  outcomes of the program to make sure
 it is meeting  the  goal.

      In a  decentralized  program,  individual  firms  and  small
 businesses  are licensed to  perform  the  inspection.   In  this case,
 the state takes primary responsibility  for many of  the  day to day
 functions,   such  as   data   collection   and processing,   public
 information, enforcement,  inspector training, and  so on.  The fact
 that  inspections are performed by many  business  entities  instead
 of one,  and that there are  more inspection sites  means  that state
 oversight   and program  evaluation  activities  need to  be  more
 intensive   in   this  type   of program  and,  as  a   result   are
 Significantly more costly.                                    '

     Costs  of  quality assurance  measures vary among programs
 depending  upon  the comprehensiveness  of  the quality  assurance
program  and are not well  documented in most state programs   EPA
Draft                        -46-                       2/26/92

-------
has recommended that the portion of the  inspection  fee  returned to
the state be dedicated to program oversight,  and  that it be  scaled
to cover the  cost  of carrying out an effective  quality assurance
program.    EPA audits  have  found that  decentralized  programs
typically have inadequate quality assurance programs, often  due to
lack of  funds, but  also  due to  insufficient  legal  authority or
willingness.

5.2  Estimated Cost of High-Tech I/M Testing


5.2.1     General Methodology

     EPA's estimates of the costs of high-tech  test procedures are
driven by  a  number  of assumptions.  The first is that  the new
tests would be performed in currently existing  inspection networks
 (i.e., centralized- and decentralized-based programs).  Therefore,
the  starting point  for  the  cost  estimates  are  the  current
inspection costs in I/M programs  as they are reported to EPA.

     The  second assumption is  that  adding the new  tests will
increase  inspection costs  in  programs  that are now  efficiently
designed and operated.    In  programs  that  are not now well
designed, current costs are likely to  be higher than necessary and
the  cost  increase  less  if efficiency  improvements are made
simultaneously.    In order  to  perform  the  high-tech tests new
equipment will have  to be acquired and additional inspector time
will be required for some test procedures.  The amount  of the cost
increase  will be  determined to  a  large  degree by the  costs of
acquiring  new  equipment  and the impact of the  longer  test on
throughput  in a high  volume operation.   Average  test volume in
decentralized  programs  is  low  enough  to   easily  absorb  the
additional  test time involved  (although at a cost  in labor  time).
Equipment  costs are analyzed in  terms  of the  additional  cost to
equip  each  inspection  site   (i.e.,  each  inspection  lane  in
centralized  inspection  networks,  and  each licensed  inspection
station in decentralized networks).

     By  focusing  on the  inspection lane  or  station  as the basic
unit  of  analysis,  the  resulting  cost  estimates  are  equally
applicable  in  large programs,  with many  subject vehicles and
inspection sites,  or small programs,  with few subject vehicles and
inspection  sites.  Previous  EPA  analyses  of  costs  in I/M programs
have  found that the  major determinants  of  inspection costs are
test  volume  and  the level  of  sophistication  of  the  inspection
equipment.    Costs  of  operating programs were  not  found  to be
measurably  affected  by  the size  of the  program  (for  further
information  the reader may  refer to  EPA's  report entitled,  "I/M
Network   Type:  Effects   on  Emission  Reductions,   Cost,  and
Convenience").   Figures  on  inspection volumes  at  inspection
stations and  lanes  are  available from I/M program  operating data.
This  information  enables the equipment  cost per vehicle  and the
Draft                        -47-                      2/26/92

-------
additional staff cost per  vehicle  to be calculated for each  test
procedure.

     The equipment cost figures presented in this paper are based
on the  costs of the  equipment EPA  believes is  best  suited  for
high-tech  testing.     They  are   current  prices   quoted   by
manufacturers, and do not  reflect  what the per unit prices might
be if this  equipment  were purchased in  volume.   Staff costs  are
based on  prevailing wage  rates  for  inspectors in  both types  of
programs  as  reported  in  conversations with  state  I/M program
personnel.   For  centralized networks,  site costs and management
overhead are  back  calculated from  current  inspection costs.   For
decentralized networks,  it is assumed that  longer  test  times could
be absorbed with no increase in sites.   The current  average volume
in decentralized stations  is 1,025 vehicles per  year (between  3
and 4 vehicles per  day,  depending upon  the  number  of days per  year
the station is open).   Consequently,  increasing the length of  the
test, to the degree that the new procedures  would, is not expected
to impact the number of  inspections that can be performed.

5.2.2     Equipment Needs and Costs

     A  pressure  metering  system,  composed  of  a  cylinder  of
nitrogen gas with a regulator, and hoses connecting the tank  to  a
pressure meter, and to the vehicle's evaporative  system is needed
to  perform evaporative  system pressure testing.    Hardware  to
interface the metering system with  a  computerized  analyzer is  also
needed and is included in the cost  estimate.  Purge  testing can be
performed by  adding a flow  sensor with a  computer interface,  a
dynamometer, and a Video Driver's Aid.  With the  further addition
of a Constant Volume Sampler (CVS)  and a flame  ionization detector
 (FID)  for  HC  analysis,  two  non-dispersive  infrared  (NDIR)
analyzers for CO and carbon monoxide  (CO2) ,  and  a  chemiluminescent
 (CI) analyzer for NOX, transient testing can be  performed.

     The  analyzers used  for the  transient test  are laboratory
grade  equipment.   They  are  designed  to  higher  accuracy   and
repeatability specifications  than  the NDIR  analyzers  used  to
perform the current I/M tests.   Table 5-2 shows  the  estimated  cost
of  equipment  for  conducting high-tech  tests.    This  quality  of
technology is essential for accurate  instantaneous measurements of
low concentration mass  emission levels.
Draft                       -48-                      2/26/92

-------
Test

Pressure

Purge



Transient
              Table 5-2

    Equipment Costs for  New  Tests


             Equipment                    price

           Metering System                 $600

           Flow Sensor                     $500
           Dynamometer                  $45,000
           Video Drivers Aid            $3,000

           CVS & Analyzers              $95f OOP

           TOTAL                      $144,100
     The  figures  in  Table  5-2  do  not  include  the  costs  of
expendable materials.   Nitrogen gas is used up  in  performing the
pressure  test.    Additionally,  the  FID  burns hydrogen  fuel.
Calibration gases are needed for each of the analyzers used in the
transient test.  Because  the  analyzers  used in  the  transient test
are designed  to  more stringent  specifications than the  analyzers
currently used in  the  field,  bi-blends,  gaseous  mixtures composed
of one  interest  gas in a diluent  (usually nitrogen)  are  used to
calibrate them.  Multi-blend  gases, such  as are  typically  used to
calibrate  current  I/M  equipment, are not  suitable.    Current
estimates for expendables are shown in Table 5-3.  The replacement
intervals are  estimated based on the usage rates observed in the
EPA  Indiana  pilot  program  and typical  inspection  volumes  as
presented  later in  this section.   Calculations of  per  vehicle
equipment  costs  presented  throughout this  report  include  per
vehicle costs of these expendables  as well.

                             Table  5-3

                     ExDendables for New Tests
      Test

   Pressure
  Material

N2 Gas
          Replacement Interval
Cost  Centralized   Decentralized
$30
250 tests
250 tests
   Transient
H2 Fuel       $60

HC Cal Gas    $60

CO Cal Gas    $60

C02 Cal Gas   $60
        2 months

        2 months

        2 months

        2 months
                I year

                1 year

                1 year

                1 year
     Staff costs  have  been found to vary between  centralized and
decentralized programs, as does the  effect  on  the  number of sites
Draft
              -49-
                          2/26/92

-------
in the network infrastructure.   Therefore,  the following sections
are devoted to separate cost  analyses  for each  network type.

5.2.3     Cost, to Upgrade Centralized Networks


5.2.3.1   Basic Assumptions

     The  starting  point in this  analysis  is  the  current  average
per vehicle inspection  cost in centralized  programs.   A figure of
$8.50  was used based upon data  from operating  programs.   This
figure  includes the cost  of one  or more retests  and  network
oversight costs.  The key  variables to consider  in estimating the
costs in  centralized networks are throughput,  equipment, and staff
costs.    Data on  these variables  were  obtained  by  contacting
program managers in  a number of  these programs, and  by surveying
program contracts  and Requests for Proposal.

     Throughput refers to the number of vehicles per hour that can
be tested in  a lane.   The  higher  the  throughput  rate,  the  greater
the number of vehicles  over which costs are spread,  and the lower
the  per  vehicle  cost.   EPA  contacted   program managers  and
consulted the contracts in  a number of  centralized programs  to
determine peak period throughput  rates in  the different  systems.
Rates were as reported in Table 5-4.

                            Table 5-4

      Peak Period Throughput Rates in Independent  I/M Programs


   Program                       Vehicles Tested  per Hour


   Arizona                                     20

   Connecticut                                25-30
   Illinois                                     25

   Maryland                                   25-35
   Wisconsin                                  25-30

     On the basis  of this information, 25  vehicles per hour  was
assumed to  represent the typical peak period  throughput  rate  or
design  capacity in  centralized   I/M  programs.   During off-peak
hours and days, throughput is lower since there  is  not  a  constant
stream of arriving vehicles.  Conversations  with  individuals  in
the  centralized   inspection  service  industry  indicate  that
inspectors start at  minimum  wage  or slightly higher, that  by the
end of the  first  year  they  earn  $5.50 to  $6  per  hour, and  that
they generally stay with the  job  for one to  three years.  Thus,  $6
per hour was used to  estimate the  average inspector's hourly wage
Draft                       -50-                      2/26/92

-------
     Estimates  of the  costs of  adding  pressure testing,  purge
testing,  and transient tailpipe testing were derived by taking the
current  costs  for the  new  equipment  to perform the new  tests,
dividing it by the number  of inspections  expected to be performed
in the lane over  a five year period and  adding  it  to the current
$8.50 per vehicle cost,  with a further adjustment  for the  impact
of test  time  on throughput,  and  thus  on  the number  of  sites and
site costs.   The  same  is done to  estimate additional  personnel
costs  associated with  adding the  new tests.   When  independent
programs  were  surveyed to   determine  the  length  of a typical
contract, it was  discovered  that  Illinois,  Florida,  and Minnesota
all have  five  year contracts, Arizona has  a seven  year contract,
and the  program in the  State of  Washington is  operating under  a
three  year  contract,  resulting in  an  average contract  length  of
five years  among the  five  programs  surveyed.    Five years  was
therefore chosen as the typical  contract length.

     The  number  of inspections expected to be performed  over the
five year contract period was derived by  calculating the  total
number of hours  of lane operation,  estimating the  average  number
of vehicles per  lane  and multiplying the two.   A lane is assumed
to operate for 60 hours a week (lane operation times were found to
vary from 54 to 64 hours per  week),  52 weeks a year for five years
for a  total of 15,600  hours.  Lanes are assumed to have  a  peak
throughput  capacity of  25 vehicles  per hour.  Modern centralized
inspection networks are designed so that they can accommodate peak
demand periods  with  all lanes operating  at this throughput  rate.
Networks are usually designed so that average throughput is  50-65%
of peak  capacity  or  13-15 vehicles  per hour.  When  operating for
15,600 hours over the life of a contract,  a centralized inspection
lane is  estimated to perform a  total  of 195,000  inspections,  or
about 39,000 per year.

5.2.3.2   The Effect of Changing Throughput

     The  addition of  evaporative  system pressure  testing  to  a
centralized program  would  result  in a  slight  decrease in  the
throughput capacity.   The addition of purge and transient testing,
along with pressure testing,  would result  in a further decrease.

     Assuming  the same  test  frequency (i.e., annual or biennial)
the reduced throughput  rate  means that the  number of lanes  needed
to test  a given  number  of  vehicles  would  increase accordingly,  as
would the size of the network infrastructure needed to support the
test program.  The result  is an  increase  in the  cost per vehicle.
Actual  consumer  cost  depends on the  test  frequency; EPA  would
encourage states  to  adopt biennial  programs  to  reduce  the  costs
and  imposition  of the program.    Less  frequent  testing  only
slightly  reduces the  emission reduction benefits  while  cutting
test costs almost in half.

     One  way to  estimate the cost would  be to simulate an  actual
network of  stations and  lanes in  a  given  city.   One could attempt


Draft                        -51-                      2/26/92

-------
to assess land  costs,  building  costs,  staff and equipment  costs,
costs for all necessary  support  systems,  and other cost  factors.
However, this approach would be  very  time  consuming  and  would  rely
on information  which  is  proprietary to  the  private contractors
that  operate   the  programs and  is,  therefore,   unavailable.
Instead, the cost of the  increased  number  of  lanes and stations  is
derived by analyzing current costs and subtracting  out  equipment,
direct personnel, and state agency  oversight  costs.  The remainder
is adjusted by  the change in throughput'in the new  system.  Then,
new  estimates   of  equipment, personnel and oversight  costs  are
added back in to obtain the  estimated total cost.

     As discussed previously, the typical  high volume station can
test  25 vehicles  per hour, performing   (in  most   cases)  a test
consisting of 30 seconds  of high speed  preconditioning or  testing,
followed by 30 seconds of idle testing.  In addition, a  short time
is spent  getting the vehicle into position  and  preparing it for
testing.   This leads to a two to three  minute  test  time  on
average,  depending  upon  what   short  test  is  performed.    EPA
recently issued alternative test  procedures for steady-state tests
that  reduce  various  problems  associated  with  those  tests,
especially false failures,  but   at a cost of  longer  average  per
test time.

     Current costs were estimated by contacting operating program
personnel,   equipment  vendors  and  contractors.    The  most
sophisticated  equipment  installation  (i.e.,  the   equipment  for
loaded  steady-state  testing)   was  used  to estimate  current
equipment costs.

     The  cost to acquire and install  a single curve dynamometer
and an  analyzer in  existing networks is about $40,000  or 210 per
vehicle using  the basic  test volume assumptions.    As  indicated
previously,  a  staff person is  assumed to  earn  $6.00  per hour.
When this figure is multiplied by 15,600 total contract hours and
divided by 195,000  vehicles, direct  staff costs  are estimated  at
48* per vehicle.  Existing centralized  networks typically  have two
staff per lane.   Thus,   total  staff costs work  out to  96*  per
vehicle.  In this  analysis  a figure  of $1.25 is  used to  estimate
the  amount of  the  state retainer.    This  reflects EPA's best
estimate  of  the  per vehicle expense  for  a  good  state quality
assurance program in a centralized  network.  Equipment,  staff,  and
state costs add  up to  $2.42  per vehicle.   Subtracting this  amount
from the  current average  of $8.50  leaves  $6.08 in  infrastructure
costs and other  overhead  expenses including employee benefits and
employer  taxes   as  shown  in Table  5-5.    This  amount  is then
factored by the change in the throughput  rate and the  equipment,
oversight, and staff costs for the new tests  are then added.
Draft                       -52-                     2/26/92

-------
                            Table 5-5

                      Current Program Costs


                            Per Vehicle     Total  Cost Less

     Increments                 Cost          Increments


     Current                                    $8.50

     Equipment                 $0.21            $8.29

     Staff                     $0.96            $7.33

     State Retainer            $1.25            $6.08

5.2.3.3   Evaporative System Pressure Testing

     Most  centralized programs  use  a  two position  test queue;
emission  test are  done  in one  position  while  emission control
devices are checked in the  other, along with  other  functions  such
as fee collection.  In this type of system the throughput rate  is
determined by  the  length  of time required to perform the longest
step in the sequence, not by length of the entire  test  sequence.
With computer automation  and refined pass/fail logic,  the pressure
test could be  performed by  an extra technician while  the emission
control device check and other functions  are performed.   In such a
system  the test  would  take between two  and three  minutes  to
perform.   As  a result,  adding  the  pressure  test is expected  to
reduce  peak  throughput  to  about   20  vehicles  per  hour.     To
accommodate peak  demand  periods  and maintain  short wait times,  a
design  throughput  rate  of half  of  capacity is assumed,  for  a
typical throughput rate of  10  vehicles per  hour.   Assuming the
same number of hours  of lane  operation, the total number of tests
per lane  is  estimated to be  156,000 over the five year contract
period.

     Calculations  of the  additional costs  of  adding pressure
testing are detailed in  Table 5-6.   These are  based  upon the costs
of equipment  and  expendables as shown in Tables  5-2 and 5-3 and
staff  costs  and throughput assumptions  detailed  in the previous
paragraphs.   Staff costs  are derived by  assuming  two  staff per
lane at a  pay rate of $6.00 per hour for  a total of  15,600 hours
in the  contract  period,  and  dividing by  156,000 vehicles tested
during that period.   Equipment costs per lane are assumed to  be
$40,600, also  divided by  156,000 vehicles.  This  works  out to 25*
per vehicle.   Based on the  cost and usage rates shown  in Table  5-
3, nitrogen gas for the pressure test costs an additional 13* per
vehicle.   Taken together,  the pressure test  would  be  expected  to
add about $2 to the cost  of an efficient,  high-volume test system.
Draft                        -53-                      2/26/92

-------
                             Table 5-6

        Costs to Add Pressure Testing to Centralized Programs

                                                  Running Total
 Increments                 Per Vehicle Cost      Cost per Vehicle

 Adjust  for  Throughput        $6.08 * 12.5/10          $7.61

 Staff                       $1.20                   $8.81

 Oversight                    $1.25                  $10.05

 Equipment and Nitrogen gas   $0.38                  $10.44



 5.2.3.4  The Transient/Purge Short Test

      The transient/purge  test  is  a longer test procedure than the
 ones  currently used in most I/M programs and is the longest single
 procedure  in  the whole  inspection  process.    Thus,  it is  the
 determining factor in lane throughput and will therefore influence
 the number  of test sites required.

      The transient  test  takes  a maximum  of  four  minutes  to
 perform.  An  additional minute is  assumed to  prepare  the vehicle
 for testing,  for  a maximum total of  five minutes.  EPA is in the
 process  of  looking   at  potential   fast-pass  and   fast-fail
 strategies, and preliminary results  suggest  that roughly  33%  of
 the  vehicles  tested  could be  fast passed or failed  based  upon
 analysis of data gathered during the first 93 seconds  of the IM240
 (i.e.,  Bag  1)  using separate  fast-pass and fast-fail  cutpoints.
 Hence,  EPA  estimates that  the average total  test time  could  be
 shortened to  at  least  four minutes per vehicle.   This  translates
 into  a throughput  capacity  of  15  vehicles  per  hour.    To
 accommodate peak  demand periods and maintain  short wait times,  a
 design  throughput rate  of half  of  capacity  is  assumed,  for  a
 typical throughput  rate of 7.5 vehicles  per  hour.  Assuming the
 same  number of  hours of  lane  operation as previously,  the  total
 number  of tests  per lane in a  transient  lane  is  estimated to  be
 117,000 over the five  year contract  period.

      State  quality  assurance program costs  would increase  given
 the complexity and diversity of the test system;  an estimate of an
 additional  50*  is  used here but  the  amount  could vary depending
 upon  the  intensity  of the oversight  function  the state chooses
 Staff costs per vehicle are calculated  using the  same  assumptions
 for wages and hours of  operation as shown in  Table 5-5;  however
 the cost is  spread  over 117,000  tests  over  the life of  the
 contract rather than  195,000.   The result is  staff costs of 80*
 per staff per  vehicle.   Four staff per lane are assumed (it may be
Draft                       -54-
                                                      2/26/92

-------
that, in a  well  configured lane, three would  suffice)  to perform
the  tests.    The  additional  tasks performed by  inspectors  in
conducting  the  new tests  - i.e.,  disconnecting  vapor lines  and
connecting them to  analytical  equipment  for  the evaporative tests
and  driving the  vehicle through the transient  driving  cycle  - do
not  require that inspectors have higher  levels of  skill than  they
do presently.  Rather,  these tasks  can be performed by comparably
skilled individuals trained to  these specific  tasks.   Total staff
costs work  out  to  $3.20  per vehicle.   Equipment  costs  for  each
test  procedure  are  derived by  taking  the  equipment  costs  from
Table  5-2  and  calculating the costs  of  five  years  worth  of
expendables using  the  figures  in Table  5-3 and  dividing  by
117,000.   The  resulting costs  estimates  are shown in  Table  5-7.
Table 5-7 shows the  result  of  factoring  the  figure  of $6.08,  from
Table 5-5, by the change  in  the throughput rate and adding in the
equipment,  staff,  and  state  costs associated with the  new  test
procedures.  The figure of  $6.08 is multiplied by  12.5/7.5,  i.e.,
the  ratio of the design throughput  rate  in the current  program to
the  design  throughput  rate  in  a program  conducting  pressure purge
and  transient testing.

                             Table 5-7

        Costs to  Add Proposed Tests  to Centralized Programs
 Increments

 Adjust for Throughput

 Staff

 Oversight

 Pressure Test

 Purge Test

 Transient Test
Per Vehicle Cost

 $6.08  *  12.5/7.5

 $3.20

 $1.75

 $0.13

 $0.41

 $0.87
 Running Total
Cost per Vehicle

     $10.14

     $13.34

     $15.09

     $15.22

     $15.64

     $16.50
     Thus,  the  cost  of  adding the  new  tests  to  centralized
networks  is  found to  be about  double  the  current average  cost.
The cost of centralized test systems has been dropping in the past
few  years as  a  result  of competitive  pressures  and  efficiency
improvements.   These  factors  may drive down the costs  of  the new
tests  as well,  especially  as  they  relate to  equipment  costs.
Given  that  conservative  assumptions  were  made  regarding  two
additional personnel,  equipment costs  of  $144,000 per  lane,  and
low  throughput rates,  the cost estimate  presented  here can  be
fairly viewed  as  a worst case assumption.    As  discussed earlier,
the important issue is the quality of the test,  not the frequency,
Draft
 -55-
      2/26/92

-------
so  doing  these  tests  on  a  biennial basis would  offset  the
increased per test cost.

5.2.4     Cost to Upgrade Decentralized Programs


5.2.4.1   Basic Assumptions

     A  methodology  similar  to  that  described  above  is  used to
estimate the costs of upgrading  decentralized programs.  Equipment
and  labor costs  are key  variables as  they were  in determining
costs for centralized programs.  However, total test  volume  rather
than throughput  and test time are  the  critical factors affecting
cost.   Licensed inspection  stations  at present  only perform, on
the  average,  about  1,025 inspections per year,  as  shown in  Table
5-8  (note that  this number is a station-weighted average).    This
analysis is  conducted based on  the assumption  that the number of
inspections per station per year on average will not  change  (i.e.,
time and  data available did not permit construction of a  supply
curve for testing services).

                             Table 5-8

        Inspection Volumes in Licensed Inspection Stations

Program              Vehicles per Year    Vehicles  per  Station

California               6,180,093                799
Colorado                 1,655,897               1,104
Dallas/Ft. Worth         1,948,333               1,624
El Paso                    278,540                1,161
Georgia                  1,118,448               1,729
Houston                  1,482,349               1,348
Louisiana                  145,175                1,037
Massachusetts            3,700,000               1,321
Nevada                     523,098                1,260
New Hampshire              137,137                 564
New York                 4,605,158               1,071
Pennsylvania             3,202,450                834
Rhode Island               650,000                 684
Virginia                   481,305                1,301
Weighted Average                                 1,025

     Two conclusions can be drawn  from this.   The first  is  that
adoption of the new tests is not likely to affect the total  number
of inspections  the  average  station can perform.   The  second is
that costs will  be  spread  over  a smaller number  of vehicles  than
in the case  of high  volume  centralized  stations.   The incremental
cost per  vehicle for each  new  test will  therefore  be  larger in
this type of inspection network.
Draft                        -56-
                                                      2/26/92

-------
     The   current  average  cost   of   vehicle   inspection   in
decentralized programs  is  about  $17.70   (again, the derivation  of
this figure can be found in EPA's  technical  information  document,
"I/M  Network Type:  Effects  on  Emission  Reductions,  Cost, and
Convenience").     Note  that this  figure   may  substantially
underestimate actual costs  since most states  limit  the  inspection
fee that a station can  charge.  In  many  cases, this fee  is  likely
to be below cost;  stations presumably obtain  sufficient  revenue  to
stay  in  business  by providing other services, which may include
engine repair.  The  costs  for adding high-tech tests are derived
by adding the incremental costs for each test type to  this amount.
Incremental costs are again  estimated by dividing the cost  of the
equipment  and  the  staff   costs  by  the  number of  inspections
expected to be performed in a year,  or 1,025 tests.

     Equipment  costs  are   spread  over  the   useful  life of the
equipment.   While a  piece  of equipment's useful  life  can vary
considerably  in actual practice,  a five  year equipment life  is
assumed.

     While large  businesses,  such  as dealerships, are often able
to  afford  to purchase  current analyzer equipment  outright, the
smaller  gas  stations  and  garages  often  have to  finance  these
purchases.   The higher  cost of  the equipment needed to perform
purge  and  transient  testing  ($144,000  for the dynamometer, CVS,
analyzers,  etc.,  as opposed to  $12,000 to  $15,000  for the most
sophisticated of  the  current NDIR-based analyzers)  makes it even
more likely that these purchases  will have  to be  financed for most
inspection  stations.   The  costs  of the  CVS and  analyzers for
transient  testing and  the  dynamometer  are   amortized  over five
years  at 12%  interest  in the analysis  in this report.   The  costs
of  the  other,  less  expensive  pieces  of   equipment  are  not
amortized, but  are assumed to be purchased outright and the cost
spread out over five  years  worth of  inspections, or 5,125 tests.

     Program personnel in decentralized  programs  were  contacted  to
determine  inspector  wage  rates.   In many cases,  inspectors are
professional mechanics earning about $25 per  hour.  However, most
states do not require inspectors to  be mechanics, and inspections
may be performed by less skilled  individuals  who  typically earn  $6
or  $7  per hour.   The  prevalence  of different  wage  rates  among
inspectors is unknown.  Therefore,  EPA assumed an average wage  of
$15  per  hour  for this  analysis.    An   overhead  rate of  40%   is
assumed,  for a total  labor  cost of $21 an hour.

5.2.4.2   The Pressure Test

     The pressure test can  be  performed  while  the inspector  is
conducting  the  anti-tampering inspection.    Thus,  virtually  no
extra labor time is added.   The incremental  cost  for this test  is
the cost of the equipment including the  nitrogen  gas,  which,  using
the costs  from  Tables 5-2  and 5-3  and  spreading them over  5,125
tests, yields a  cost  of  23$  per  vehicle.   Adding  this  to the
current average  cost  of  $17.70 yields a total  cost of $17.93.

Draft                      "  -57-                      2/26/92

-------
5.2.4.3   The Transient/Purge Short Test

     The transient/purge  test  is assumed to  require  five minutes
of the  inspector's time.   With labor  costs  at $21 per  hour, as
described above, this  works out to $1.75 per vehicle.   Equipment
costs are taken from Table 5-2, except that the costs for the  CVS,
analyzers,  and dynamometer are amortized over the five year period
as described  previously.   This  brings  the total costs  for these
pieces of equipment over  the five  year  period to $126,793 for the
CVS and  analyzer system,  and $60,060 for the dynamometer.   These
costs are divided by 5,125  inspections.   The  costs  of expendables
from Table  5-3  are  added in according  to the usage rates assumed
for decentralized programs.

     The incremental  costs  for these  test  procedures,  and  the
total cost  with pressure testing included, are  detailed  in Table
5-9.

                             Table  5-9

     Costs to  Conduct Pressurer  Purge,,  and Transient  Testing  in
                      Decentralized Procrrams
Test Procedure         Incremental Cost               Total

Pressure                     $0.23                    $17.93

Labor                        $1.75                    $19.68

Purge (Equipment only)       $12.40                    $32.08

Transient (Equipment only)   $24.97                    $57.05



5.3  Repair Costs


5.3.1     HC and CO Exhaust Repair Costs and Methodology

     The  repair  costs  for HC and CO  exhaust  emission  repairs are
split into  two elements.    One  addresses the repair costs due to
failure of a tailpipe test, such as the 2500 rpm/Idle idle test or
the loaded transient test.  The other element addresses the repair
costs of correcting tampering identified as a result of the visual
inspection  for the presence and connection  of emission control
components such as the  catalyst  (also  known as "ATP  failures")
Draft                        -58-                      2/26/92

-------
5.3.1.1   Tailpipe Emission Test Failures

     Based on  current  information from I/M programs  which collect
repair  cost  information,  the  average cost  to repair  a 1981  or
newer vehicle  failing the  2500  rpm/Idle test  is approximately $75,
including parts and labor.  For example,  1989 repair  data from the
Louisville, Kentucky I/M program shows the average cost  to  be $54
for  all  model  year  vehicles  if  only   commercial  repairs  are
included.   The  overall average cost drops  to  $42  per  repaired
vehicle  if the  cost of self  repairs  (repairs  performed by  the
individual vehicle owner) are also included5.   In addition, the $75
average  repair cost figure  is further supported by the  findings
from an  I/M  repair study conducted in California which  showed the
average  repair  cost to be  $72 for  1980 and later  model  year
vehicles6.  In this study, 500  vehicles that  failed the  California
I/M  test were recruited,  tested,  and  repaired at independent
commercial garages to  pass I/M.  Finally, a  study of repair costs
conducted by the  Oregon I/M program  in  1985  and  1986 found  the
average repair cost to be about $50 per failure.7

     The  average cost  to  repair  a  vehicle  which  fails  both  the
IM240 and the  2500  rpm/Idle  test is also assumed to  be  $75.   This
figure is based  on the fact  that these cars  are  likely  to receive
on average the same types of  repairs  as  are received by  vehicles
failing only the  2500  rpm/Idle test.  For the  vehicles  which fail
only the  IM240 emission test,  the  average repair cost  is assumed
to be  $150,  or twice  as much.  This  higher  repair  cost  accounts
for the  additional and more  thorough diagnosis needed to  identify
the causes of  the IM240 failures.   In addition, it allows for  the
possibility  of more costly engine  parts  being required to  repair
the IM240 failures.  Therefore,  blending  the  $75  cost of repairing
combined  IM240 and 2500 rpm/Idle  failures with  the  $150 cost  of
repairing IM240-only failures,  and  assuming  (based on observations
in  Indiana)  that there are  slightly more   2500  rpm/Idle/IM240
failures  than  IM240-only failures,  yields an average cost of $120.

5.3.1.2   Emission Control Inspection Failures

     The  average cost  (separated  by model year group)  to  repair
emission control components  identified  as  needing  repair  or
replacement by a visual  inspection are shown in Table 5-10.

     These  costs  were  estimated  several  years  ago,  based  on
average  retail parts and  labor costs.   For  example, the average
air pump  repair  cost reflects  the  cost  of replacing a  broken  air
5   "1989 Annual Report Vehicle  Exhaust Testing Program  Jefferson  County,
    Kentucky", April, 1990
6   "I/M Evaluation Program  Series II",  Summary from the California Air
    Resources Board's I/M Evaluation Program,  October 25, 1991.
7   Jasper,  W.  P.  "A Discussion of Reported Maintenance and Repair Expenses
    in an I/M Program",  SAE Paper 861547



Draft                        -59-                      2/26/92

-------
pump belt  or reconnecting an air  or vacuum line.   This cost  was
based  on  the   assumption  that  most  air  pump  tampering  ^or
malmaintenance will  focus on disabling the  unit by disconnecting
the belt or line rather than removing the entire unit.   If  this is
the case, then the repairs will be relatively  simple.   The  average
catalyst  replacement  cost was  based on  the  retail   cost  of  an
aftermarket   converter.    The  misfueled   catalyst  replacement
reflected the cost  of the converter plus an additional amount  to
replace the poisoned oxygen sensor.  The evaporative system repair
is the average cost of reconnecting a vapor  or vacuum  line  after a
visual inspection of the  system.   The  PCV  and gas cap  repairs  are
the average cost of replacing these components.

                            Table 5-10

       Averacre Cost of Reoairina  Emission Control Components
        Component                       Pre-81     1981+

        Air Pump                          $15       $15

        Catalyst Replacement              $150       $165

        Misfueled Catalyst  Replace        $175       $190

        Evaporative System  Repair          $5         $5

        PCV System Repair                 $5         $5

        Gas Cap Replacement               $5         $5

     Repair of  intentional  tampering failures  will  contribute
relatively little  to  the overall cost  of repairing I/M-failed
vehicles  in the  1990s,  due  to decreasing  tampering  rates.   The
estimated costs per vehicle,  therefore,  were not revisited.

5.3.2      NO3E Repair Coats and Methodology

     Repair costs for  NOX reduction, and the  supporting analysis
are discussed  separately from the HC and CO repair cost analysis
because repairs  targeted to  reduce HC and CO emissions often have
no effect on NOX emissions.  Moreover,  the Indiana data  showed that
the HC/CO  failures  and  the NOX failures were essentially separate
sets of vehicles8.  For  example, many  vehicles requiring repairs to
correct high HC  or  CO  emissions  frequently have  fairly low NOX
emissions,  and  consequently  do   not  require  NOX  repairs.
Furthermore, for those  vehicles which  are  high  NOX emitters, the
    November 1991,  EPA memorandum from E.  Glover to C.  Harvey,   "Average
    Repair Costs and Benefits from Repairing High NOX Emitters."
Draft                        -60-                      2/26/92

-------
most common repair is to the EGR system, and this often has little
impact on HC  or CO emissions.   In  other  words,  the vehicles with
excessive  HC and  CO emissions  usually need  different  types  of
repairs  than those  with  excessive NOX  emissions.    Thus,  their
repair costs were analyzed separately.

     The data used to calculate  the average cost  and benefit  of
performing  vehicle NOX repairs was  collected in the on-going EPA
Emission Factor test program at the EPA's  Motor Vehicle Emission
Laboratory  (MVEL)  in Ann  Arbor,  Michigan,  as  well  as  at the ATL
facility in South  Bend,  Indiana.    In  this  program,  large numbers
of in-use vehicles were recruited for testing and repair to better
characterize the  emissions  of  the  fleet.    However,  for  the
analysis of NOX costs, the overall database was restricted to 1983
and later model year vehicles  which had received an FTP test both
before  and after  repair, and had  been  tested  in  the  last  few
years.  As  a  result,  data from 169  1983+ model  year vehicles with
repair data were obtained.

     Most of  the 169 vehicles  were  high emitters of HC or CO,  and
had  repairs aimed at  those  pollutants,  since  EPA  had  given  the
testing  contractor instructions  to  focus on HC  and CO emissions.
In order to more accurately characterize the cost of effective NOX
repairs,  criteria  were  used  to further  select vehicles  which
clearly  had high  NOX  emissions  before  repair,  but had  achieved
lower NOX  emissions  as the result of  the repair.   These criteria
were: Before  repair  FTP emissions had to exceed 2.0 gpm NOX,  and
after repair  FTP emissions could  not exceed 1.25 gpm.  As a result
of these  criteria, 10 cars out of  169 were selected,  and  9  were
used in the final  cost analysis.   Examining the individual vehicle
repairs  of these  9  vehicles   (see Table  5-11)  shows that  all  of
them needed EGR repairs to lower the NOX emissions to levels which
could meet the  criteria.   On 6  of these  vehicles, the EGR  was
replaced, while on the other 3 the EGR passage was cleaned,  or the
delay valve was replaced.

     The tenth  car  (683), a Chevrolet  Chevette,  was removed from
the cost analysis  because the  repair it received was not targeted
toward NOX  reduction.   Instead, NOX emissions  decreased primarily
due to an  ineffective  EC/CO  repair,  which caused the engine to  go
to  a rich air/fuel mixture  as  evidenced by  a  very large  CO
emission increase  (10 to 30 gpm).

     The repair costs  of  the  9 individual vehicles  as  well  as the
overall averages are  shown in  Table 5-11.   For  example, the price
of  the  repair  parts  averaged  $44,   using  Mitchell's   Summer
Collision Estimating Guide.  The labor cost averaged $34, based on
0.68 hours  at  $50 per hour.   These labor hours were determined
using Mitchell's  1991 Mechanic's   Labor  Estimating	Gjiisie..   In
addition, each  car was  assumed to require 0.5  hours of diagnostic
time at  the labor rate  of $50/hour for  an average cost of  $25.
Summing these costs puts the  total average cost of an effective NOX
Draft                        -61-                      2/26/92

-------
repair  at  $103.    For input  into  subsequent  cost effectiveness
models this overall cost was rounded to  $100.
                             Table 5-11

                              Repair Costs

Veh
861
94

803

1095

23
545
1131

1657

41
683

MX
87
86

86

89

87
84
86

83

85
85

Make
MERC
FORD

CHRY

PONT

FORD
CADI
DOD6

CHEV

CHEV
CHEV

Model
COUGAR
THUNDERBIRD

NEW YORKER

GRAND PRIX

TAUR0S
SEVILLE
W150

CELEBRITY

S-10
CHEVETTE
NOx Repair
Description
Replaced EGR Valve
Install EGR Vacuum
Line
Replaced EGR Valve
Clean EGR Passage
Replaced EGR Valve
Assembly
Clean EGR Passage
Clean EGR Passage
Replaced EGR Delay
Valve
Replaced EGR Valve
Clean EGR Passage
Replaced EGR Valve
O2 Sensor, Coolant
Labor
Hours
0
0

0

0

1
0
0

0

0
4
.80
.30

.80

.80

.00
.70
.30

.70

.70
.60
Labor
Cost
$40
$15

$40

$40

$50
$35
$15

$35

$35
$230
Parts
Cost
Retail
$42
$0

$41

$161

$0
$0
$22

$65

$67
$39
Diag-
nostic
Cost
$25
$25

$25

$25

$25
$25
$25

$25

$25
$25
Total
Cost
$107
$40

$106

$226

$75
$60
$62

$125

$127
$294
                     Temperature Sensor,
                     Rebuilt Carburetor
AVERAGE with #683


AVERAGE without #683
                           1.07


                           0.68
$53

$34
$43

$44
$25

$25
$122

$103
 5.3.3
Evaporative System Repair Costs and Methodology
      The repair and  cost  data  used  to  calculate  the  average
evaporative  system  repair  costs  and  subsequent  fuel  economy
improvements  were  collected  during  an  EPA  running  loss  test
program conducted  at  ATL  during  the  Spring  of  1991  in  which
failing vehicles were  repaired and retested.  All comparisons were
done  with data  obtained from running  loss tests at  95°  F  using a
9.0  RVP emission test  fuel,  and 3  consecutive  LA-4 test  cycles
 (the  first LA-4  being a  cold start).

      The cost-benefit calculation was  based upon  a  sample  of 25
vehicles which  failed  either  the I/M  purge or  pressure  test in
this  test program,  and  for  which evaporative  system repair cost
information  was  available.9   Only  24  vehicles (vehicle 1667 was not
available)  were  used  to calculate  the average fuel economy cost
    July 26,  1991, EPA memorandum from E.  Glover to C. Harvey,   "Average
    Repair Costs and Benefits from Repairing Purge and Pressure Failures "
Draft
                   -62-
          2/26/92

-------


Test
Pressure
Purge
M

Total Fuel
Economy
Improvement
7.87 gpm 6.1%
8.26 gpm 5.7%
Total
Fuel
Economy
Savings
gal /mi
0.0034
0.0035


Average
Parts
Cost
$15.03
$21.89


Average
Labor
Hour
0.45
0.96
savings resulting from evaporative system repair.  The results are
shown in Table 5-12.

                            Table 5-12

          Average Repair Costs and Fuel Economv Benefits
                                                         Average
                                                          Total
                                                          Cost*

                                                          $37.76

                                                          $70.10

* Labor costs are computed from labor time using a labor rate (including
California) of $50 per hour

     The  evaporative  repair costs, excluding gas  caps,  are based
on parts  costs as invoiced by ATL.  If  the  cost  of a repair part
for a particular  vehicle  was  not  available,  then  the average cost
from the  other vehicles which also received  that  repair was used.
For example,  in  the analysis, the  value  of  $29.46,  obtained from
vehicle   (1563)   was   used as an  estimate of   purge solenoid
replacement  cost  on  two  other vehicles (1525  and 1552)  which
received that repair,  but did not have invoiced repair costs.  The
ATL invoiced gas cap  replacement  cost was available on only two
vehicles  (1532 and 1542) .  For the other vehicles which required
this  repair, the  gas cap  cost  was based on auto  dealer  retail
prices for an OEM part.   Typically, the  gas  cap  OEM retail price
was around  $7.   In  addition,  repair parts  such  as  evaporative
hoses, or inexpensive in-line tees were  assumed  to cost nothing,
except as overhead in the  labor cost of fixing them.

     The  time   of repair  is  generally  based  on   individual
diagnostic and repair durations provided by  ATL.   Typically, they
include  both the time to  diagnose the problem  and  replace  or
reattach the parts.   For example,  vehicle 1548 required 6 hours of
diagnosis to discover the cause of the  purge problem and  replace
the defective part.   Most of  the time  was spent  in  diagnosis,
though this  length was unusual  since  most diagnoses  and  repairs
were completed in a half an hour or less.

     In  some cases,  actual  labor  times were not available  to
diagnose  or  replace a particular part.   In  these cases estimates
were  made regarding  the  duration of  a typical  repair.    For
example,  gas cap  replacement  (including diagnosis)  duration was
not usually  itemized and,  therefore,  was  estimated  to   be  15
minutes.   In other cases,  repair  times from similar  repairs on
other cars were used.  However,  for the  sake of clarity, both the
Draft                        -63-                      2/26/92

-------
 parts  and labor cost basis of  each vehicle's repair are  noted in
 Table  2  of  reference 8.

 5.4  Fuel Economy Benefits


 5.4.1     Fuel Economy Benefits of Evaporative System Repairs

     The analysis of  the data shows  a substantial  fuel economy
 benefit  under  the  95°  F  test  conditions  as  the   result  of
 evaporative  system  repair.    This   fuel  economy  benefit  is
 attributed  to  two factors.  The first is increased performance and
 efficiency  of  the vehicle's engine  following  an evaporative  repair
 such as  reconnection  of  a  vacuum  line  or  a  TVS repair.    This
 increase in efficiency was directly  measured  by the CVS  equipment,
 and  it was  found that  the measured fuel economy  increased by  an
 average  3.2%  for vehicles failing the  pressure  test,  and  an
 average  2.8%  for vehicles  failing the purge test.    The  fuel
 economy  improvement is calculated by dividing the fuel consumption
 reduction by the total  fuel consumption,  as illustrated in the
 following calculations:

     4.13 grams fuel/mile/128.3 grams fuel/mile =3.2%  Pressure failures

     4.05 grams fuel/mile/143.75 grams fuel/mile =2.8% Purge  failures

     The second  factor  involved  in   the  fuel  economy  benefit
 calculation is the  utilization of the captured  HC vapor  which
 would  have  otherwise been lost as  running  loss emissions.   In a
 properly   designed   closed-loop   vehicle   the   engine   should
 effectively substitute  these  vapors  for  liquid  tank  fuel,  and
 reduce the   vehicle's  real fuel  consumption.   These vapor fuel
 flows  from the  engine  and  the  evaporative  canister  are  not
 measured during the running loss test.

     Since  actual fuel flow data were not measured, it was assumed
 that  100%  of the  captured  running loss  emissions (i.e.,  the
 difference   between  before  and   after repair  levels)  can  be
 effectively utilized as  fuel.   This assumption  may be slightly
 high given  the fact that on average  exhaust CO emissions increased
 somewhat as the  result  of evaporative repairs,  indicating that
 some of  the extra fuel was not  fully combusted.   However,  such an
 error  (i.e.,  using  an 'R' factor  of 1.0)  is probably  small,  and
 its effect  should not  be large considering that the  running loss
 reductions  are not  large  in  comparison  to total  vehicle  fuel
 consumption.

     The  running loss vapors from pressure  failures were converted
 to_liquid fuel, using an  R Factor  of 1.0, the standard density of
 Emission  Test  Fuel,  and  a  carbon  weight factor  of  0.83  for the
 fuel.

     3.74 gpm running loss  CH2.33 * R Factor =3.74 gpm liquid fuel  (CHi 85)
Draft                       -64-
                                                      2/26/92

-------
     3.74 g C/mi * (1cm3/ 0.745 g Fuel) * (1 g Fuel / 0.83 g C)  *
     (1.0 liter/ 1000 cm3) * (1.0 gal/3.79 liter)  = 0.0016 gal/mi

     The analogous running loss  vapors from  purge failures  were
converted  to  liquid  fuel,   and  the  percentage  fuel  economy
improvement was calculated in a similar manner.

     4.21 gpm running loss CH2.33 * R Factor =4.21 gpm liquid fuel (CH1.85)

     4.21 g C/mi * (1 cm3/ 0.745 g Fuel) * (1 g Fuel / 0.83 g C) *
           (1.0 liter /  1000 cm3)  * (1.0 gal/3.79  liter) = 0.0018 gal/mi

     The measured  fuel  economy  improvements  from  better engine
operation were combined with  the measured running loss reductions
to produce  evaporative  repair fuel economy benefits  of  6.1%  for
pressure failures  and 5.7%  for purge failures.   Averaging these
together produced  an  overall  fuel  economy benefit from  evaporative
repair of 5.9%.

5.4.2     Fuel Economy Benefits of IM240 Repairs

     The  fuel  economy  benefit for  repairing  a vehicle  that  has
been identified  as failing the 0.8  gpm  HC  outpoint or the 15  gpm
CO outpoint on the IM240  test  has  been estimated as  an  increase of
12.6% in overall fuel economy,  after repairs.  This  compares to an
8.0% fuel  economy  benefit  realized by  identifying and repairing
vehicles  using  the  2500  rpm/Idle  test as a yardstick.   These
percentages  are derived from data  gathered from  the  IM240  test
site in Hammond, Indiana, and are  based  upon an  average difference
in fuel economy before and after repairs.

     The 12.6%  fuel  economy  benefit assessed  for identifying  and
repairing vehicles on the basis of the IM240 test lane results is
based upon two  groups of 1983 and newer vehicles recruited at  the
Hammond test site.   The  first group included  those vehicles  that
failed the  emissions  cutpoints of  0.8 gpm  for HC and/or 15.0  gpm
for CO, which  were subsequently FTP-tested, repaired and retested
at  the  ATL  facility  in  Indiana   (a  total  of  42 vehicles) .    The
second group consisted of those vehicles that  failed the emissions
cutpoints, were  FTP-tested,  but were not  repaired   (a total of 10
vehicles).   Unrepaired vehicles were  assumed  to represent a  fuel
economy benefit of zero,  with the  net effect that the  overall  fuel
economy benefit calculation is conservative.

     The  10  IM240-failed  vehicles  mentioned above  were   not
repaired and retested because the  original design  of the testing
program sought  to  conserve  testing slots  by  applying a criteria
that only  vehicles with  an  FTP  result twice  the certification
standards  for  the  vehicle would  receive repairs and be retested.
These  unrepaired   vehicles  were  included in the analysis  to
represent that  fraction of vehicles  (i.e.,  19%,  or 10 out of 52)
expected to  fail  the IM240  (in  a future  I/M program)  but which
have only a  marginal  emissions  problem  and presumably  only  a
Draft                        -65-                      2/26/92

-------
marginal  fuel economy  loss  (if  any),  thus  requiring only minimal
repairs  which  will  not  result  in improved fuel  economy.   The
averaged  fuel economy benefit represents a harmonic average of_the
FTP  fuel  economy  before  and after repairs for  the  52  vehicle
sample group.

     Table 5-13 shows that 17 vehicles  which failed at  the Hammond
lane  were  not  repaired,  which  raises  the  issue  of why  only  10
vehicles  were  used  to  represent the  wno improvement"  vehicles.
The  logic  and  assumptions  were  as  follows.   Of the 72  vehicles
that were repaired,  only  42  (58.3%) had all  the necessary data to
do  the calculations.   Assuming  the same  attrition  rate  (duetto
incomplete data)  for the vehicles  that  did not  receive  repairs
(i.e.,  58.3% of the  17 unrepaired  vehicles) yields  a total of 10
vehicles.   The net  effect  of  assuming this  fraction  of  "zero
improvement  vehicles"  is a  lower  fuel economy benefit  for  the
IM240  (12.6%  instead  of a potential 15.7%).

                              Table 5-13

          Zero Improvement Vehicle Sample Size Adjustments

 Original #                                                  Remaining
  of Vehs    Description of Data  Used and Removed                Vehicles

   98       1983+ Failed lane 0.8 & 15 &  received FTPs              98

   17       were less than twice standard & not repaired            81

    9       were greater than twice the  standard,  but  were not     72
            repaired  due  to test  schedule  or cost (engine
            rebuild or  catalyst)

    4       had  no as-received  IM240s at ATL (IM240-based fuel     68
            economy benefits were initially  evaluated, so this
            test was required.  In  retrospect,  they should have
            been added  back  into the database for the FTP-based
            FE improvement)

    1       had  no after-repair IM240 f!643  (to verify repair     67
            success)

   25       Failed after-repair  IM240 (incomplete ATL repairs)       42

            % of 72 repaired that can be included in analysis =     58.3%

            58%  of 17  <2 x standard  & not repaired included as
            zero improvement =                                   10

5.4.3      Fuel  Economy Benefit for  the  2500 rpm/Idle Te^-fr

     The  8.0%  fuel economy  benefit assessed for  identifying  and
repairing vehicles  on the basis  of  the  2500 rpm/Idle  test is based
upon  two  groups  of  1983  and  newer   vehicles  recruited  at  the
Draft                         -66-                       2/26/92

-------
Hammond test site.   The first group consisted of  6  vehicles that
failed the  220  ppm HC  and/or  1.2% CO cutpoints on  their initial
2500  rpm/Idle  I/M  test,   received  an  IM240  before  and  after
commercial  repairs, and  received passing  scores  on the  retest.
The second group consisted  of  those  vehicles  that  returned to the
Hammond lane after  commercial repairs, but again  failed  the 2500
rpm/Idle test.   These  latter 6 retest failures  are  considered to
be the result  of incomplete repairs, which would  be corrected in
an enhanced I/M program.

     The before  and  after  IM240  fuel economy  data  was "corrected"
to reflect  FTP fuel economy  by  employing a correction factor of
1.0925, reflecting the fact  that,  on average,  FTP  fuel  economy
varies from  IM240  measured fuel economy by 9.25%.   This  variance
reflects the fact that the  IM240 and FTP are,  after all,  different
tests, using different  driving cycles, etc.   Still,  the two tests
show  a high  degree of correlation,  and, in  the  area  of  fuel
economy,  the  variance between the two  tests  is  a relatively
constant difference  of 9.25%.  Therefore, multiplying IM240 fuel
economy readings by  1.0925  yields  a  reliable  estimate of  FTP fuel
economy.

     After successful  repairs  (i.e.,  those resulting in a passing
retest),  some marginal  vehicles will fail  to  realize a noticeable
fuel  economy improvement.   Using a database of 48  cars,  it  was
determined  that  4  of the  vehicles that  failed  the  2500  rpm/Idle
I/M test were not repaired  because their FTP emissions scores were
less  than twice their certification standards,  leading to  the
conclusion  that, had  these  vehicles been repaired, their  fuel
economy benefit would be zero.  These 4 vehicles  represent 8.3% of
the  48 database vehicles  for which all  the  necessary  data  was
available.   Assuming that 8.3% of the 6 vehicles  that were still
failing I/M would not  get a  fuel economy benefit  after  repairs
yields a figure  of 0.48 vehicles that will show no noticeable fuel
economy benefit.  Given that  half  of a vehicle cannot be  added to
the  database,  each  of  the other  6  vehicles  that  did pass  after
repairs were duplicated yielding  twelve  vehicles,  and 1  vehicle
was  added to  represent the "no  fuel  economy improvement"  case.
Adding  the single  "zero  improvement vehicle"  lowered  the  fuel
economy benefit of  the  2500  rpm/Idle test  from  8.6%  to  8.0%.
Table 5-14 further details how these  numbers were arrived at.
Draft                        -67-                      2/26/92

-------
                            Table 5-14

           Adjusted Zero FE Benefit Vehicle Sample Size
   Original # of                                           Remaining
     Vehicles    Description of Data Used and Removed          Vehicles

        312      1983+ Failed IN I/M at lane                     312

        256      not recruited to lab                           56

         8       missing data                                 48

        44      dirty enough to expect an FE benefit             4

                % that failed IN I/M but too clean for a FE    8.3%
                benefit (4 of 48)

                8%  of 6  commercially  repaired included as    0.48
                zero improvement

     While  6  vehicles may  seem  like a slim database,  we did  not
want to assume too  low a  fuel  economy benefit  for the  conventional
2500 rpm/Idle test  and risk overestimating  the incremental  benefit
of the  IM240 test.   A mid-1980s  study with  actual  or simulated
commercial  repairs  of  older  technology  1981-83 vehicles showed
only a 3.5% improvement.   This has not been shown to be  applicable
to newer  technology vehicles.   We also did not  want  to claim  too
much benefit.  We did not rely on the ATL-performed repairs  (as we
did  for the  fuel  economy benefit  when using IM240 outpoints)
because  the ATL  mechanics were instructed to  repair  all  known
malfunctions that  would  likely affect FTP  HC  and  CO.   Therefore,
the emissions  and fuel  economy  benefits would likely exceed what
would actually occur with real world repairs  that  stop as  soon as
the 2500 rpm/Idle  test cutpoints are met.   In contrast, we  judged
that because  the  IM240  is  a  mass emissions test  that correlates
well with the FTP,  real world  repairs  aimed at  making vehicles
pass  the  fairly  stringent   IM240  cutpoints  would  not  be   so
different from those made by  the ATL mechanics.   The  fact  that 25
of the  67 ATL-repaired  vehicles still failed the  IM240  suggests
that ATL mechanics in general did not go too far.

5.5  Recurring  Failure  and Repair  Rates  and Fraction  of  Fleet
     Affected by Fuel Economy Benefits

     The  rates  at which  vehicles  recurrently  fail tailpipe  tests
and emission control inspections in  an ongoing  I/M program  (i.e.,
the percentage of  failing vehicles in a  program that  has been
established  for   a few  years)   are  used   within  the  Cost
Effectiveness  Model  (CEM)  for  determining repair costs.   Fuel
economy  credits  for repairs  resulting from  tailpipe  tests  are
based on  the hypothetical  failure  rates that would  occur in  the


Draft                        -68-                      2/26/92

-------
first cycle  of  the I/M program if  it  were just starting.   These
hypothetical rates in effect represent  vehicles  that  have  been  and
remain  affected  by  the  I/M  program  that  has  in  fact  been
operating.

     The exhaust  test  failure rates  for calculation  of repair
costs in CEM are in the  form of a  zero-mile  failure rate and  a
deterioration rate, such that the fraction of failing vehicles  for
a given test type is calculated by multiplying the  deterioration
rate by the average mileage  and adding  that result to the  zero-
mile  failure  rate.    Table  5-15   shows   the   zero-mile   and
deterioration  rates found in the  BLOCK DATA section of  the  CEM
program listing in Appendix A.

                            Table 5-15

                    Exhaust Test Failure  Rates
                             (fraction)
       Test


       Idle

       2-Speed

       Loaded

       IM240

       NOx
Zero—Mile


 0.00

 0.00

 0.0252

 0.00

 0.032936
 Deterioration
(per  10K miles)

  0.01

  0.01

  0.01190

  0.0373

  0.0084805
    Type


 (recurring)

 (recurring)

 (recurring)

(first-cycle)

 (recurring)
     These numbers are based  on  regressions of  emission test  data
from the  IM240 lane  in  Indiana.   In  1990  and  1991,  Indiana had
just  revitalized  its  moribund  I/M  program   and  hence  can be
considered to  represent a hypothetical  I/M  program in its first
cycle  of  inspections) .    For  the  IM240  the  first-cycle HC/CO
failure rate per  10,000  miles was 0.0373 at an average of  50,000
miles  observed among 3,436  model year  1983  and  newer  cars in
Indiana.   The  above recurring  rates  include  adjustment  of the
first-cycle  rates   by   a  factor  of   1/1.87  (e.g.,  0.01
0.0187/1.87).   This  adjustment  factor  is  the  recurring  initial
failure ratio  for  idle  testing,  derived by comparing the  Indiana
failure rates with failure rates from other operating  I/M programs
with longer histories.

     The recurring zero-mile  rate  used by the model for the IM240
is  half  of  the  first-cycle  deterioration   rate  (0.0373/2   =
0.01995).  The  recurring  deterioration rate used by the model for
the IM240  is  half of 1/1.87  times the  first-cycle failure rate.
This method  represents  a  50-50  compromise between the following
two assumptions,  either of which  would be reasonably  plausible:
Draft
           -69-
                         2/26/92

-------
(a) The  IM240 test  will require  vehicle  repairs  sufficient  to
return   the  emission control  systems  to like-new condition  thus
yielding a constant  failure  rate  equal to the rate found  for the
first 10,000  miles  of operation  (0.0373),  and (b)  IM240  repairs
will deteriorate similarly to idle and 2-speed test  repairs,  which
would yield a deterioration rate of 0.0373/1.87 = 0.01995).

     These failure rates  assume  cutpoints of 1.2% CO and  220 ppm
HC  for the  idle  and 2-speed tests,  and 0.8/15 gpm for  the  IM240
test.   For NOx, separate  cutpoints of  1.69  for PFI, 2.50 for TBI,
and 3.99 gpm  for  carbureted vehicles are used  resulting in  an
overall nominal failure  rate  of about 10%  on the IM240.

     In the case of ATP emission control component inspections CEM
calculates recurring repair rates for  the  first year  a  vehicle  is
inspected  from  the  difference  in   tampering   rates   given  by
MOBILE4.1 for the no-program  case and the with-ATP case.   There  is
also a small residual repair  rate assumed for latter years,  with a
very minor cost impact.

     In  the  case of purge  and pressure  test failures  MOBILE4.1
uses a lookup table which has different malfunction  rates for each
vehicle  age  up to  13 years,  and  older vehicles  are  assigned the
rates of the 13  year old vehicles.   The malfunction rates  range
from roughly 4% to 33%  for purge or  pressure  malfunctions,  and  8%
to  50%  for the  combination  of purge  and pressure malfunctions.
This  lookup  table  can be  found  as   the  EFFECT  array  at the
beginning  of the  FAIL  function  in the  CEM program listing  in
Appendix A.   After appropriately weighting together these  purge
and pressure failure rates, MOBILE4.1 uses them in its calculation
of evaporative and running loss emission factors  in  the  absence  of
an evaporative I/M program.  These malfunction rates  would become
the first-cycle  failure rates  for a new I/M program rather  than
recurring failure rates.

     CEM assumes these  same  initial failure  rates  in determining
fuel economy benefits of purge and pressure tests,  since the  fuel
economy  effect of  an I/M program in a given year depends  on the
difference between the  number  of failures  that  would exist  in a
no-program  case  and the  near-zero  number present with the  I/M
program in operation.   The fuel economy benefit  calculation using
these failure rates  is described in section 5.4.

     To  determine  purge and pressure  repair  costs,  CEM requires
recurring failure  rates corresponding  to an  ongoing I/M  program
wherein the  failure  rate  would be lower than the initial  failure
rate observed  in Indiana's first  cycle and used  in MOBILE4.1 and
in  the  fuel  economy benefit  calculation  to represent the  no-
program case.  The recurring  purge and pressure failure  rates used
for this  purpose  are:
Draft                       -70-                      2/26/92

-------
     Recurring Purge test failure  rate:   3.0%

     Recurring Pressure test failure  rate:     2.5%

     Recurring total Purge/Pressure failure  rate:   5.0%

The exact use  of  these rates can be seen in  the FAIL  function  of
the CEM program listing in Appendix A.

     These recurring purge and pressure test failure rates  were
derived from the  initial rates  of MOBILE4.1.   As  an example, the
5% total failure  rate  is based on roughly a  50% failure  rate for
ten year  old vehicles  indicating that  roughly  5% went bad  each
year  on  average.   For  an  analysis   that  did  not   treat age
explicitly this was an assumption that could be used for all ages,
and would  definitely not  underestimate  costs,  since much  of the
rise to the 50% failure rate happens  at higher mileages  when there
are fewer cars still in use.

5.6  Method for Estimating Cost  Effectiveness  of I/M Programs

     The  cost of  an  I/M  program is  determined  by  summing the
estimated  inspection fee  costs,  the  estimated repair costs, and
the negative  cost  of estimated fuel economy  benefits  (gallons  of
fuel saved *  $/gallon).   The emission benefits of an I/M  program
are determined by  subtracting  the estimated emissions with the
program from  the  emissions with  no  I/M program.    CEM does the
emissions  calculation by  making  multiple  runs  of MOBILE4.1 and
manipulating  the  results  of  the  various runs.   Since MOBILE4.1
does  not  include  the  necessary cost  components,  CEM  itself
calculates costs by combining the previously discussed  information
on per vehicle costs and fuel economy  benefits with  the  estimates
of failure rates.

     Since MOBILE4.1  calculates  the  emission  levels,  tampering
rates,  and misfueling rates for January 1st  of each calendar year,
CEM performs  two two  consecutive  sets of  MOBILE4.1  runs and
interpolates between them to get  an annual average emission  rate
which is then  converted  into a  ton per  year value  using  the fleet
vehicle miles  travelled  (VMT)  data  contained in  MOBILE4.1.   In
order to separate  out  costs and benefits associated with  various
portions of  an I/M program, two  intermediate MOBILE4.1 runs are
done between  the   full  program and no-program  runs.   Therefore,
each CEM run  performs a total of eight MOBILE4.1 runs as follows.
Draft                        -71-                      2/26/92

-------
     1)   Full I/M & ATP program (as requested)
     2)   Run 1 minus any ATP and evap testing
     3)   Run  2  minus  any  tailpipe  I/M,  but  with  tampering
          deterrence effect of I/M
     4)   Baseline, no program benefits at all)
     5)   Run 1 for next calendar year
     6)   Run 2 for next calendar year
     7)   Run 3 for next calendar year
     8)   Run 4 for next calendar year

5.6.1  Inspection Costs

     Inspection  costs  are  determined  by multiplying  user-input
inspection costs by the number of vehicles adjusted for compliance
rate   (percentage  of   vehicles that  fail  to   get  inspected) .
Separate  costs  are  input  for tailpipe emission  tests,  emission
control checks,  purge test,  and pressure test.  If a program calls
for biennial rather  than annual  inspections,  the inspection costs
per year are divided in half.   All  default  costs are  found in the
SETUP  routine of  the GEM program listing in  Appendix A.   Default
inspection costs are  shown  in Table 5-16.  Note  that  the cost of
performing the  purge test overlaps  many of  the  costs  associated
with transient testing, including the cost of a dynamometer, video
driver's aid  (VDA),  and the  throughput adjustment associated with
the  longer  test  time.     If  purge   testing  is assumed,  the
incremental  cost  of  including the  transient test is  relatively
minor, including the  cost  of a constant  volume  sampler (CVS)  and
the analyzers necessary to perform mass emissions  testing.

                            Table 5-16

                Default Inspection Costs in CEM4.1
           Test              Cost             Comments
Steady-state Tailpipe Test j    $10
Emission Control Checks
25C-1.75
Pressure Test             I    69C
Purge Test

Transient Emission Test
5.6.2     Repair Costs
 $6.53
$12 if biennial

Depends on checks done
Includes dyno,  adjusted thruput
  67C   |Increment  over purge cost
     Calculating total repair costs  is  performed similarly to the
inspection costs,  except that the costs  are only applied to the
percentage of vehicles estimated to  fail  a  given I/M test.  it is
Draft                        -72-                      2/26/92

-------
further adjusted  for the percentage  of  vehicles that do  not  get
repaired  because  they  require  repairs costing more  than  the
applicable  cost  waiver  limit.   Default  repair  costs  are  as
follows.

                            Table 5-17

                   Default Repair  Cost in CEM4.1


      Failure Triggering Repair     Pre—81       81+

      Idle or 2500 rpm/Idle Test      $50        $75

      Transient Test  (IM240)          N/A        $150

      Air Pump                        $15        $15

      Catalyst                        $150        $165

      Misfueled Catalyst Cost         $175        $190

      Evaporative  System              $5          $5

      PCV System                      $5          $5

      Gas Cap                         $5          $5

      Purge Test                      $70        $70

      Pressure Test                   $38        $38

      N0x                          Estimated     $10°

     In the  case  of  transient exhaust  testing,  the fraction  of
failing vehicles  that would have  failed a  2500 rpm/Idle  test  is
assigned the  repair  cost  for  the 2500  rpm/Idle test,  while  the
remainder is assigned the higher transient test repair cost.

5.6.3     Fuel Economy Cost Benefits

     Fuel economy benefits are based on cumulative repairs  made to
vehicles  that fail  an I/M  tailpipe  test  and/or an evaporative
system pressure  test.   As  described in section 5.5, the repair
rate  used  is the  first-cycle  failure rate  corresponding  to
inspection of vehicles that have not  previously been subject to an
I/M program.   The percentage improvement in fuel economy  depends
on the type of test  that was failed.   The  following  benefits  are
from the BLOCK DATA section of  the CEM program listing.
Draft                        -73-                      2/26/92

-------
                             Table  5-18

                  Fuel Economy Benefits in CEM4.1


               Test                        FE  Benefit

               2500 rpm/Idle (pre-81)           0.0%

               2500 rpm/Idle (81+)              8.0%

               IM240 (83+)                     12.6%

               Purge/Pressure                  5.9%

The model  converts  these percent MPG benefits  into  dollar benefits
using  the  VMT information from  MOBILE4.1,  fleet  average  fuel
economies  for appropriate  model  years from CEM and a  user-input
gasoline cost from CEM, which defaults  to $1.25 per gallon.
Draft                        -74-
                                                       2/26/92

-------
6.0  REGULATORY IMPACT ANALYSIS  -  COSTS  AND BENEFITS OF ENHANCED

I/M OPTIONS


6.1  Emission Reduction Benefits

     Grams  per  mile  emission  factors  were  calculated  using
MOBILE4.1   for  low,   medium,   and  high   option  enhanced  I/M
performance  standards.   The  test procedure  variables  and the
constant inputs are detailed in  Table 6-1.   All options are based
on centralized testing of  1968 and later light duty vehicles and
light duty  trucks  with annual frequency,  as  required by section
182 (c) (3) (B) of the Clean  Air  Act  Amendments  of 1990.  The major
differences in the benefits of the options are due to the type of
testing utilized,  specifically, the adoption of evaporative system
pressure  testing  in  the  medium  option  and  of transient  and
evaporative system purge  testing in the high option.   Other inputs
reflect national default  values assumed in MOBILE4.1.
Draft                        -75-                      2/26/92

-------
                             Table 6-1
>r Enhanced I/M
Low
20%
1968+
none
none •
none
none
1%
98%
central
annual
LDV/LDT1/LDT2
1981+
Yes
Yes
No
No
No
No
No
500 feet
11.5
8.7
1992
72°F
92°F
87.5°F
20.6/27.3/20.6
no
no
19.6 mph
MOB4.1 default
Performance Staj
Medium
20%
1968 +
none
1971+
none
none
1%
98%
central
annual
LDV/LDT1/LDT2
1981+
Yes
Yes
No
No
No
No
No
500 feet
11.5
8.7
1992
72°F
92°F
87.5°F
20.6/27.3/20.6
no
no
19.6 mph
MOB4.1 default
rr^r'l options
High
20%
1968-1980
1981-1985
1983+
1986+
1986+
1%
98%
central
annual
LDV/ LDT1/LDT2
1984+
Yes
Yes
No
No
No
No
No
500 feet
11.5
8.7
1992
72°F
92°F
87.5°F
20.6/27.3/20.6
no
no
19.6 mph
MOB4.1 default
        Inputs

Pre-1981 Stringency
Idle
2500 rpm/Idle
Pressure
Purge
Transient
tWaiver Rate
tCompliance Rate
*Network Type
*Test Frequency
*Vehicle Coverage
ATP MY coverage
Catalyst
Fuel Inlet
Air Pump
Tailpipe Lead Test
Evap Disablement
PCV Disablement
Gas Cap
Altitude
Period 1 RVP
Period 2 RVP
Period 2 Start Year
Minimum Temperature
Maximum Temperature
Ambient Temperature
Operating Mode
Onboard Controls
Stage II Control
Vehicle Speeds
VMT Mix

t  These percentages may not be realistic  for some programs, in which case the
   program will have to be "over designed" to make up the performance loss.

*  Clean  Air Act Amendments require  these inputs as  elements  of  the
   performance standard.

     The  gram per mile emission factors  for  each option and  the
emission  reduction benefit  as  a percentage of the no-I/M case  in
the calendar year  2000 are shown in Table 6-2.  The  no-I/M factors
were  calculated assuming  the same  RVP,  ambient  temperatures
maximum and  minimum  temperatures,  operating  modes,  altitude,
vehicle speeds,  and VMT mix variables  as  assumed  for  the  enhanced
I/M options.   Stage II and  on-board vapor recovery  system effects
were  not modeled  in  either the  no-I/M  or  enhanced I/M options
case.
Draft
-76-
                                                        2/26/92

-------
     Emission  benefits  from  basic  I/M  (the  current  performance
standard)  and from the  biennial high  option program  (which EPA
recommends) are  also  shown.   Note that  the  proposed enhanced I/M
performance  standard  listed  below  in  Table 6-2  is  an  annual
program,  as  required  by  the  Act.   Note  further  that  emission
reductions are  expressed as a percentage  of total highway mobile
source emissions.  Many  other mobile  source  programs are described
based  on light-duty  vehicles; doing so here would show  a much
higher percent benefit.

                             Table  6-2

                 Benefits of I/M Programs Options*

                        VOC Emission  Effects    CO Emission Effects
       Scenario


    Base - No I/M

      Basic I/M

      Low Option

    Medium Option

 Biennial High Option

  Proposed Enhanced
 Performance  Standard
              Emission
               Factor
                (gpm)

                2.084

                1.971

                1.870

                1.661

                1.495

                1.503
 Percent   Emission
Reduction   Factor
             (gpm)
          Percent
         Reduction
  5.4%

  10.3%

  20.3%

  28.3%

  27.9%
11.874

10.021

 8.927

 8.927

 8.223

 8.230
15.6%

24.8%

24.8%

30.7%

30.7%
*  Total Highway Mobile Source Emissions in 2000

     The results shown in Table 6-2  are our best estimates at this
time but our test programs and data  analyses are continuing and we
anticipate refining the numbers as time goes  on.


6.2  Cost Effectiveness Estimates
6.2.1
AssumDtions and Inputs for the Program  Potions
     EPA's  estimates  of  the  cost-effectiveness  of  the  three
scenarios are  based upon modeling with MOBILE4.1  and CEM4.1 with
assessments done for calendar  year  2000,   These are compared with
a modeling scenario in which no I/M  program is  assumed.

     The  low,  medium,  and  high options  are the  same  as  those
described in Table 6-1.   The assumed cost  for an I/M inspection,
including a  visual check of emission control devices,  is $8.50.
The  incremental cost  of adding  the  evaporative  system pressure
Draft
                  -77-
                   2/26/92

-------
test is  $1.94.    The  incremental  costs  of adding  the purge  and
transient tests are $5.19  and $0.87,  respectively.   As  indicated
in section 5.6.1,  the cost of the  purge test  includes  the cost of
a dynamometer and VDA,  and also reflects a throughput ^adjustment
to accommodate  the  longer test; adding  transient testing to  the
purge  test  requires  the  addition of  a CVS and  the necessary
emissions analyzers.   In  addition,  gasoline is  assumed to  cost
$1.25 per gallon.   The average repair costs  shown  in Table  5-17
were assumed.   It should  be further noted  that the  incremental
costs of  adding purge  and transient testing to a decentralized
network  ($12.40 and $24.97,  respectively)  are  larger than ^in  a
centralized network  because  of the assumption  these additional
costs will be spread out  over a smaller  test volume  (i.e., it is
assumed that the average number of vehicles tested per station in
a decentralized network will  not change).

6.2.3     Cost-Effectiveness Calculations

     Total  annual  program  costs  per  million  vehicles,   as
calculated  by  CEM4.1, are  presented  in  Table 6-3,  including
inspection costs,  repair cost and  fuel economy benefits,  shown on
an annual basis.  Note  that the total  cost  of  I/M increases  as  the
option becomes more  stringent;  also note  that  the total cost  (on a
per million vehicle basis) of a biennial high option is less  than
both the  annual  low  option  and  the basic  I/M  program.   These
results  make  it  clear that biennial  testing  should be  a  top
priority.

                             Table  6-3

                    Total  Annual Program Cost

                Scenario                    Cost

            Basic I/M                    $6,412,000

            Low Option                  $9,885,000

            Medium Option              $11,628,000

            High  Option                $11,390,000

            Biennial High  Option         $5,429,000

     The next step  is  to  calculate cost-effectiveness ratios,  or
the annual cost per  ton of  emission  reductions.   For areas  that
are required  to do  enhanced I/M due  to  ozone nonattainment  (the
majority of enhanced I/M areas),  the ratios  could be  calculated by
dividing the  annual program costs, from Table  6-3,  and  dividing
them by the annual  tons of  hydrocarbon  reductions.    The  results
are shown in. Table 6-4.  Unlike the total costs  in Table  6-3,  the
cost per ton decreases  with program stringency.   This  is because a
Draft                       -78-                      2/26/92

-------
major part  of the cost  is  the inspection and  the small marginal
cost of  doing a more  effective  test is overwhelmed  by the large
marginal benefit.  This  is  a critical factor to keep in mind when
choosing among various different  ozone control strategies.

                             Table 6-4

              Cost per  Ton Allocating All Costs to VOC

              Scenario                  Costs per Ton

              Basic I/M                     $5,410

              Low Option                    $4,404

              Medium Option                 $2,621

              High Option (Annual)          $1,694

              Biennial High Option           $879

     Since  the  I/M  program  yields  CO  benefits  as  well  as  VOC
benefits and  some areas  need reductions in both, it makes sense to
split the  cost among  pollutants.   High-tech I/M  can also obtain
significant NOX benefits and many ozone areas may need NOX control
as  well  to  bring ozone levels into compliance with EPA standards.
To  estimate the cost of only  the VOC portion of the I/M benefit,
one can  assess  what  the  cost would have been to obtain the CO and
NOX reductions by other  strategies.   If all the program costs were
allocated to  NOX reductions  (which only occur  in  the high option
program) , then  the  cost per ton  for  the  annual high-tech program
would be $6,298 per  ton and  for the biennial  high-tech program
$3,267  per  ton  of  NOX benefit.    Alternative  costs  for  NOX
reductions  are  estimated using  cost per ton  figures  to  obtain
stationary  source NOX  reductions through the use of more efficient
burners, estimated at  $300 per ton.   Allocating all of the program
costs to CO yields  a cost per ton  of about  $143 for the biennial
high option.   Costs  for  other control programs  range from roughly
$100-225  (without  fuel economy benefits)  for cold temperature CO
standards.   Oxygenated  fuels  programs range from about $200-400
per ton.  A conservative,  alternative cost per  ton figure of $125
was chosen for  this  analysis.   These  alternative cost  per  ton
figures  are   then  multiplied  by   the   annual  ton  reductions
attributable  to  the  various program scenarios.   Other assumptions
about the  cost of alternate  CO  or NOX programs would change the
cost remaining  to allocate  to VOC.   Higher costs would leave less
to  assign to VOC and vice-versa.

     Since  CO reductions  are not needed  in all  areas,  and only
about 44% of  the vehicles that will be subject to enhanced I/M are
in  CO areas,  costs are not assigned in all areas.  This is done by
Draft                        -79-                     2/26/92

-------
reducing the tons of emission reduction to 44% of full benefit and
using that result to calculate the alternative  cost per ton.

     The  results are shown in  Table  6-5.   As  expected  the costs
are  lower in all cases,  and  the biennial high  option  program is
about $461 per ton.

                            Table 6-5

        VOC  Cost per Ton Accounting for NOS and CO Benefit


                  Scenario             Cost Per Ton

             Basic I/M                    $4,518

             Low Option                   $3,655

             Medium Option                $2,242

             High Option (Annual)          $1,271

             Biennial High Option          $461

6.2.4     National Cost of Choosing Low Option  I/M

     The  Clean  Air Act  requires  nonattainment areas  to  meet
specific milestones of  15% reduction in VOC  emissions by  1996 and
a 3% reduction per year thereafter.  There are  two ways  for states
to achieve these goals: impose additional controls  on  stationary
sources   (i.e.,  those  beyond RACT  requirements)  or additional
controls on mobile sources.  The question  is:   What is the cost of
doing  a  less  stringent  I/M  program  and  getting  additional
reductions from stationary sources instead?

     A low performance standard  for I/M  means fewer tons  of VOC
reductions than a high option  program,  as  shown in Table  6-6.   The
low option system, even when implemented in a centralized network,
costs more per  ton  than the high-tech approach.  Thus,  if states
choose to implement a low option program there  is a direct cost to
the  nation  because  of  the higher  expense.    In  addition  to  the
direct cost,  there is  also an indirect cost.   As more and  more
controls are imposed on stationary sources,  the law of diminishing
returns would predict  that the  cost  per  ton  will  rise.   It  is
estimated that  the cost  of  these marginal  controls will  likely
exceed $5,000 per ton.
Draft                       -80-
                                                      2/26/92

-------
                            Table 6-6

              Total Cost and Benefits of  I/M Options


         Per Million Vehicles       Tons     Total  Cost

             High Option           6f724     $8,544,000

        Centralized Low  Option      2,245     $8,204,000

       Decentralized Low Option     2,245     $17,062,000

     To  estimate  the  total cost  of implementing a  low option
program it was assumed  that of the 56 million vehicles subject to
enhanced  I/M  42  million vehicles  would be  in a decentralized
system  and  14  million  would be  centralized.    This  reflects the
current  mix  of  programs in  the affected  areas.    It  was  also
assumed that each  ton not obtained  from  I/M would  be gotten from
stationary source controls at  $5000 per ton.   The  results are
shown in Table 6-7.  The extra  direct cost of the low option would
be  about  $353   million while  the  indirect  cost  of  the  more
expensive  stationary  source controls  amounts  to about $1,254
million, for a total  of  about $1.6 billion in excess cost.

                            Table 6-7

              Excess Cost of Choosing Low Option I/M


Low Option
Vehicles
millions
High Option 56
Centralized 14
Low Option Decentralized 42
Total

Stationary Cost
Total
6.3 National Costs
Low Option 56
High - Low
@ $5000/ton
Excess Cost
and Benefits
Benefits
tons
376,529
31,426
94,279
125,705
250,824


Cost
millions
$479
$115
$717
$832
$353
$1,254
$1,607
6.3.1 Emission Reductions
     Estimates of the total costs and  emission  reduction  benefits
of current  and future  I/M  programs were  obtained using CEM4.1.
Because average costs  and effectiveness vary between  centralized


Draft                       -81-                      2/26/92

-------
and  decentralized programs10 the costs  and  reductions were modeled
differently  for each program type.   The MOBILE4.1  output showing
the  scenarios  used are in Appendix K.  Vehicle  population figures
are  needed  in order  to  calculate  total   costs  and  emission
reductions.    Because figures  obtained  from the  states vary  in
reliability,  estimates were derived based upon  Census  data  for
each area.

     As  shown  in Table  6-8 below,  current  I/M  programs  obtain
estimated  total annual  emission reductions of 116,000  tons  of VOC
and  1,566,000  tons  of  CO.    Implementation  of the high  option
requirements  of  this proposed action  on a  biennial basis  would
yield  estimated annual  emission reductions of 384,000  tons  of VOC
and  2,345,000  tons of  CO from enhanced  I/M programs,  and  36,000
tons of VOC  and 500,000 tons of CO from basic programs.   Enhanced
I/M  programs would also reduce NOX emissions.   The  transient test
with NOX  cutpoints  designed to fail 10% to  20%  of the  vehicles
would  yield  estimated  NOX  reductions  of  9%  relative to  emission
levels with no  program in place.
10Tierney,  E.,J.  "I/M Network Type:   Effects on Emission Reductions   Cost
   and Convenience," U.S. EPA Technical Information Document,  number EPA-AA-
   TSS-I/M-89-2, January  1991



Draft                        -82-                      2/26/92

-------
                            Table 6-8

                     National  Benefits  of  I/M


               (tons of emissions reduced  annually)

                                  VQC          £Q

       Reductions from Continuing I/M Unchanged

            Centralized Areas     55,540       775,228

          Decentralized Areas     60f476       791f167

               Current Total    116,016     1,566,395

       Expected Reductions from Proposal

       Enhanced Areas           384,130     2,345,278


       Basic Areas

                  Centralized     23,289       326,290

               Decentralized     12,996       174,186

                  Basic Total     36.285       5QQ.476

       Total Future Benefits    420.415     2,845,754

Thus,  enhanced  I/M and  improvements  to  existing and  new  I/M
programs  will result in national emission reductions substantially
greater than current I/M programs.


6.3.2     Economic Costs to Motorists

     EPA has developed estimates of inspection and repair costs in
a "high-tech" I/M program.   The derivation of these estimates is
detailed  in section 5.0.   A  conventional  steady-state  I/M  test
including ATP currently costs about $8.50 per vehicle  on average
in a centralized  program,  and $17.70 per vehicle on average  in a
decentralized  program.   A  complete  high-tech  test,   including
transient,  purge,  and pressure  testing,  is  expected  to  cost
approximately $17 per  vehicle  in  an efficiently run  high-volume
centralized program.   In  a program where 1984 and  later vehicles
received  the high-tech test,  and older vehicles received a steady-
state test  and ATP,  and the  inspection were performed  biennially,
the estimated annual per vehicle cost would be about $9.  The cost
is sensitive to whether test equipment and personnel face a steady
stream of vehicles or have idle periods.  Therefore the cost would
be somewhat higher  in a test-only  multi-participant system if the


Draft                       -83-                     2/26/92

-------
inspection  network  had  more  excess  capacity  than  a  typical
centralized  program.   Test-only  stations  may  also not  be  as
proficient in  testing each  vehicle  quickly,  adding somewhat  to
costs.

     The  overall  average repair  cost for transient failures  is
estimated to  be $120.  Average repair  costs for pressure and purge
test failures  are  estimated  to  be  $38  and  $70,   respectively.
Repairs for NOX failures are estimated to cost  approximately $100
per vehicle.   Data from the Hammond test program  indicate that  it
would be  very  rare  for  one vehicle  to  need all  three of  these
repair  costs.

     These  repairs  have  been found to  produce   fuel  economy
benefits that will at least partially offset the  cost of  repairs.
Fuel economy improvements of  6.1%  for pressure test  failures and
5.7% for purge test  failures were observed.   Vehicles that  failed
the transient short  test  at  the proposed cutpoints were  found  to
enjoy a fuel economy improvement of 12.6% as a  result of  repairs.
Fuel economy  improvements persist beyond the  year of the  test.

     Currently,  there are an estimated 63,550,000  vehicles subject
to  I/M nationwide.    Of these,  23,574,000  are in centralized
programs  and  39,976,000  are  in  decentralized programs  (see
Appendix  K) .   Inspection fees  currently total  an estimated  $747
million annually, $182 million in centralized programs, and $565
million in decentralized programs.  Repair costs  are  estimated  at
$392 million,  $140  million  in centralized  programs,  and  $252
million in decentralized programs.  Current fuel  economy  benefits
are  estimated  at  $245  million,  $92  million  in  centralized
programs,  and $153 million in decentralized programs.

     As  shown  in Table  6-9  below,   estimates  using EPA's  cost
effectiveness model  show that  total  inspection costs in the  year
2000 in enhanced I/M programs accounting  for  growth in the size  of
the vehicle  fleet are expected to be $451 million,  with repairs
totaling  $710 million  assuming that  programs are biennial.    Fuel
economy benefits  are expected  to  total $825  million,  with  $617
million  attributable  to the  tailpipe emissions test and  $208
million due to  the functional evaporative tests.

     In basic I/M programs,  total  annual inspection costs  in the
year 2000 are estimated  at $162 million,  and  repair  costs are
expected to be  approximately  $113 million.

     Thus, despite significant increases  in repair expenditures  as
a result  of  the  program, the  switch  to  biennial testing and the
improved  fuel  economy benefits from  programs  will  result in  a
lower national  annual cost of the inspection program.
Draft                       -84-                      2/26/92

-------
                             Table 6-9

                Program Costs and Economic Benefits

                       (millions of dollars)
                                     Emission
                    Emission           Test      Evap
                      Test    Evap     Fuel      Fuel
             Test    Repair  Repair  Economy  Economy   Net
             Cost     Cost    Cost   Savings  Savings   Cost*

Costs and. Economic Benefits of Continuin  I/M Unchaned
Central
Decentral
Total
$182
$565
$747
Expected Costs and
$140
$252
$392
Economic
na
na

Benefits
($92)
($153)
($245)
na
na

$230
$664
$894
From Proposal
Enhanced     $451     $489    $221    ($617)    ($208)    $336

Basic

  Central    $67      $60      na     ($39)      na      $88

Decentral    $95      $53      na.     ($31)      na      $117

    Total    $162     $113            ($70)             $205

    Grand    $613     $602    $221    ($687)    ($208)    $541
    Total
* Net cost  is derived by adding inspection and  repair  costs and subtracting
  fuel economy benefits .


6 • 4  Motorist Inconvenience Costs

     There  is  an additional cost  factor associated with  I/M,  the
cost of  the time  spent by vehicle owners  in  complying  with  the
inspection  requirement.   This  cost was  estimated by assuming that
motorists'  leisure  time is worth  about  $20  per hour.  The amount
of time spent  getting an inspection can vary considerably as well
and  very  little data  on this  subject is  available.   For  the
purpose of  this  analysis,  it  was assumed  that motorists  typically
spend  roughly 45  minutes travelling to  the test  site,  getting
tested, and returning in an  efficiently designed high  volume test
program.

     EPA calculated the  cost  effectiveness  of  the biennial high
option with this additional cost included.   Table  6-10  below shows
Draft                        -85-                      2/26/92

-------
the estimated  total program cost  per  million vehicles,  the cost
per  ton  with  all  costs  allocated to VOC  reduction,  and  the
adjusted cost per ton of VOC with costs allocated among pollutants
as discussed previously.

                            Table 6-10

     Costs of the Biennial  High Option inr.lndina Inconvenience


          Total Cost               $12,254,000


          Cost per Ton

          All costs to VOC           $1,983


          Cost per Ton

          Adjusted VOC Cost          $1,566

     Comparing these figures with those in  Tables  6-4  and 6-5 show
that the  biennial high option,  with motorist inconvenience  costs
included,  is still more  cost  effective than the  low and medium
options without those costs considered.


7 . 0  REGULATORY FLEXIBILITY ANALYSIS


7.1  Regulatory Flexibility Act Requirements

     The  Regulatory Flexibility Act  recognizes  three  kinds  of
small entities  and defines  them as  follows:

   • Small  business  - any business  which  is independently  owned
     and  operated and not dominant in  its field  as defined  by
     Small Business Administration regulations under  section 3  of
     the Small  Business Act.

   • Small  organization  - any not-for-profit  enterprise that  is
     independently  owned  and  operated  and not dominant in  its
     field  (e.g., private hospitals and educational institutions).

   • Small  governmental  jurisdiction  -   any  government  of  a
     district with a population of  less than 50,000.

     Small  governmental  jurisdictions,  as  defined  above,  are
exempted from the  requirements of  this regulation.    There are  no
private non-profit organizations  involved in the operation of I/M
programs.    Consequently this analysis will  be  limited to  the
affects   on certain  small  businesses,   namely  providers  of
inspection and  repair services and  of inspection equipment.
Draft                       -86-                      2/26/92

-------
     There is a significant impact on  small  entities  whenever  the
following criteria are  satisfied:

   • Annual  compliance  costs  (annualized  capital,  operating,
     reporting,  etc.)  increase total  costs of production  for  small
     entities for the  relevant process  or product by more than 5%

   • Compliance costs  as a percent of sales for small  entities  are
     at  least  10% higher  than compliance costs as  a percent  of
     sales for large entities

   • Capital costs  of  compliance  represent a  significant portion
     of  capital available  to  small entities,  considering  internal
     cash flow plus external financing  capabilities

   • The  requirements  of the   regulation  are  likely  to result  in
     closures of small entities

     The  enhanced  I/M  performance  standard  contained in  the
proposed action includes new "high-tech"  test procedures  for  newer
vehicles and  enables  states to  obtain  significantly  higher
emission reductions  from their  I/M programs than they  have
previously.   This performance  standard will affect  different  types
of businesses differently.  Test providers will need  to invest  in
new equipment.   Repair providers  will be  repairing more vehicles
for more  types  of inspection  failures.  The enhanced performance
standard  will also affect different  types of  inspection networks
differently.

7.1.1     The Universe of Affected Entities

     The  Regulatory   Flexibility  Act's  definition of  "small
business" is  based on the Small  Business Administration's  (SBA)
definitions.   These are  listed in  13 CFR Part 121  by Standard
Industrial Code  (SIC)  categories.   The  types  of businesses that
have  either  been  licensed to  perform inspections or have been
involved  in I/M  in some other way, such as by  selling inspection
equipment, and their SIC categories are listed in Table 7-1,  along
with  the size cutoffs used by SBA to define  small  business  for
each.   Size  cutoffs  are  defined either  in terms of number  of
employees or  gross annual   revenue,  expressed  in   millions   of
dollars.
Draft                       -87-                      2/26/92

-------
                            Table 7-1

                       Affected Businesses

        Sic                  Description                 Cutoff

        5013    Automotive Part  and Supply Wholesalers     100
                (i.e.,  auto  engine  testing equipment,  employees
               electrical)

        5511    Motor Vehicle Dealers (New and  Used)       $11.5  M

        5521    Motor Vehicle Dealers (Used)               $11.5  M

        5531    Auto and Home Supply Stores               $3.5 M

        5541    Gasoline Service Stations                  $4.5 M

        7531    Top and Body  Repair Shops                  $3.5 M

        7534    Tire Retreading and Repair Shops           $7.0 M

        7535    Paint Shops                               $3.5 M

        7538    General Automotive Repair  Shops           $3.5 M

        7539    Auto Repair,  Not Elsewhere Classified,   $3.5 M
                (e.g.,  radiator  shops muffler  shops,
               transmission  shops,  etc.)

        7549    Automotive Services,  Except Repair and   $3.5 M
               Car  Washes   (e.g.,  diagnostic  centers,
               inspection centers,  towing etc.)

     Note that although all  analyzer manufacturers are "affected,"
the size  cutoff of 100 employees prevents  them from meeting  the
definition of "small business."

7.2  Types of Economic Impacts  of Concern

     This analysis  looks at the types of impacts  that  inspection
and repair  providers in existing programs will  experience as  a
result  of the  requirements of  the  proposed  action.   Since  the
requirements  for  basic I/M programs  will remain  essentially  the
same as the  current I/M requirements,  significant  impacts  are  not
expected  in  these programs.  Hence,  this analysis will focus on
existing  I/M programs that will have to  become  enhanced.   This
analysis  assumes  that  the  enhanced program  implemented will
reflect the  high  option  on  the basis that this would represent  a
"worst  case"  scenario (i.e.,  that  with the  greatest economic
impact potential).
Draft                       -88-                      2/26/92

-------
7.3  Changes in Repair Activity

     The repair industry in enhanced areas that currently have I/M
programs will  enjoy  a significant  increase  in repair  revenues.
The repair  industry  consists of motor vehicle  dealers  (SICs 5511
and 5521),  general  automotive repair shops  (SIC  7538)  and some
gasoline service stations (SIC 5541).

7.3.1     Repair Activity in Current I/M Programs

     Reliable data do  not exist on the number of repair facilities
in  I/M  program  areas  that  do  I/M repairs.    However,  repair
revenues that  accrue to the industry as  a whole can be estimated
using vehicle  population data.   EPA estimates that  there  are 64
million vehicles in  current I/M program areas, 24 million of which
are in areas with centralized programs.   Of these,  an estimated 15
million  are in  areas that will  become  enhanced.   There  are an
estimated  40  million  vehicles  in  decentralized programs.   Of
these, about 33  million are in areas that must implement enhanced
I/M.

     Repair cost information  is generally not collected  by  the
states except  when  a motorist applies for a  waiver.   However, as
described in  section 5.6,  estimates of total  repair costs  can be
made  using CEM4.1.    EPA  estimates  that $392  million  worth of
repair business  would be generated  by current I/M  programs  in the
year  2000  if  these programs continued unchanged,  $302  million in
areas that  will  go  enhanced.   Of this latter  figure, an estimated
$89  million would  be performed  in  areas that currently operate
centralized programs and $213 million in areas with decentralized
programs.
 Draft                        -89-                      2/26/92

-------
7.3.2     Repair Activity in Future I/M Programs

     The transient  test, with  its superior  ability to  identify
excess emissions,  is  expected to  generate  more repairs  than  the
steady-state tests,  while the purge and pressure tests will enable
I/M  programs to  identify excess  evaporative  emissions  for  the
first time.   Estimates using CEM4.1  indicate that  an  additional
$100 million in annual repair business  will be  generated in areas
that  currently  operate  centralized programs,  and an  additional
$212  million  in  areas  that  currently  operate  decentralized
programs as a result of  the  requirements proposed  in this action.
The  additional emission  repairs  identified by the  transient  test
are  expected to generate an additional $41 million  in  areas  that
currently have centralized programs and $79 million in  areas  that
currently have decentralized programs.  The addition  of purge  and
pressure testing is expected to  generate an additional $59 million
in  areas  that  currently have  centralized  programs,  and  $132
million in areas that currently  have decentralized  programs.   Thus
the  repair industry in  these areas  is estimated  to  receive  an
additional $312  million,  and a total of $613 million annually  as a
result of the proposed  action, as summarized in Table 7-2.

                             Table  7-2

             Repair Expenses in  Enhanced I/M Programs

                       (millions of dollars)

                     Centralized   Decentralized  All Programs

       Current             $89           $213           $302

Additional

   Transient Repairs       $41           $79           $120

 Evaporative Repairs       $59           $132           $191

      Total New            $1QQ           $211           $311

       Total              $189           $424           $613

     The $311 million in extra repair expenditures  is estimated to
comprise about 40% parts cost and the remainder  for labor, profit,
and overhead.  The automotive parts industry estimates that 20,000
jobs are created for every $1 billion  spent on  parts.   Hence,  the
additional  parts  demand  ($125 million)  will create 750  jobs  in
parts manufacturing as  well  as  additional business  for retailers
and distributors,  and is likely  to  create  more jobs for clerks and
delivery employees.   The remaining  60% is  estimated to  comprise
about 50%  profit  and overhead at  the  repair  shop  and  50% labor.
Hence, mechanics  will  earn an  additional  $93  million  over  all
program  areas.   At  an  average  pay rate  of  $25  per  hour   this


Draft                        -90-                      2/26/92

-------
translates into 1,800 full time equivalents (FTE)  over all program
areas.

     Firms that  pursue this repair  business  may need  to upgrade
repair  technician  skills and  obtain additional  diagnostic  and
other equipment  to  perform  effective repairs  on new  technology
vehicles.  Inspection  stations  in  decentralized programs,  as well
as many repair  shops  in centralized  programs, possess  emission
analyzers.  These  will be useful in testing  those vehicles  still
subject to steady-state tests and may be used to diagnose vehicles
failing  the transient  test and  to  assess repair success.   BAR90
analyzers, in  particular,  are  designed to function as  a platform
for  a variety  of engine diagnostic functions and to  download OBD
fault codes.

7.4  Changes in Emission Testing Activity  in I/M Areas

7.4.1     The  Existing Market  in  Centralized  and Decentralized
Programs

     A  number of  different types of  entities are  involved  in
providing inspections.   The  centralized programs in the States  of
New  Jersey,  Delaware,  Oregon,   and  Indiana  are  operated by  the
state,  those in  the cities of Memphis,  Tennessee,  and Washington,
D.C. are  operated  by the local  government.   These programs  cover
approximately  6  million vehicles.    All  of these programs  except
Oregon   and   Memphis  will  be  subject   to  the  enhanced  I/M
requirement.   Therefore, 5 million vehicles in government operated
programs  will  be covered by this  requirement.   The  remaining  18
million vehicles are in programs operated by private contractors
 (SIC 7549), of which 10 million vehicles are in areas  covered  by
the  enhanced  I/M requirement.    Both the  government agencies,  and
the private contractors exceed the  cutoffs for small entities.

     Inspection  providers in decentralized programs fall into all
SIC  categories  in  Table 7-1 except  5013 -  Automotive  Part  and
Supply  Wholesalers.   However,   the  prevalence of  the  different
categories among licensed inspection stations varies.   The  total
number  of inspection  stations  in  decentralized areas  covered  by
the enhanced I/M requirement  are listed in Table 7-3.
Draft                        -91-                      2/26/92

-------
                            Table 7-3

              Number of Inspection Stations by  State.



            State                        Stations

          California                       8,752

           Colorado                        1,500

           Georgia                          647

           Houston                         1,100

          Louisiana                         140

        Massachusetts                      2,800

            Nevada                          415

        New Hampshire                       243

           New York                        4,300

         Pennsylvania                      3,838

         Rhode Island                       950

           Virginia                         370

            Total                         25,055

     Data on  the  distribution  of inspection  stations among  the
different categories are not collected by most  states,  neither is
data on  the  number of  stations that fall below  the  cutoffs  for
small  entities  listed  in  Table  7-1.    However,   listings  of
inspection stations were obtained from California  and Pennsylvania
and  stations  were broken  down  into  the following  categories:
Service Stations,  gas  stations  that also perform  repairs  (5541);
Dealerships  (5511 and 5521); Independent  Repair Shops (7538);  Non-
Engine  Repair   Shops,   such  as  tire   shops,  body  shops,   or
transmission shops  (7531,  7534,  7535,  and 7539); Retailers  (5531);
and Test Only  Stations (7549).   The California  data is based on an
analysis of the  entire station  population.   The Pennsylvania data
is  based on  an  analysis  of  a  10%  random sample  of licensed
stations.
Draft                       -92-                      2/26/92

-------
                            Table 7-4
                 Inspection Stations  bv Cateaorv

Station Type
Service Stations
Dealerships
Independent Repair Shops
Non-Engine Repair Shops
Retailers
Test Only Stations
Total

Station Type
Service Stations
Dealerships
Independent Repair Shops
Non-Engine Repair Shops
Retailers
Test Only Stations
Total
California
Number
2,183
1,361
3,272
734
276
131
7978
Pennsylvania
Number
124
95
67
46
16
0
348
• T-ryrr' J
Percentage
27
17
41
9
3
2


Percentage
36
27
19
13
5
0

     Information  on  number  of  subject  vehicles  in  each  I/M
program,  and the  inspection  fee  and  the  portion of  the  fee
returned to the state in each  program is  readily  available.   EPA
also  gathers  data  on  the  number  of  licensed  stations   in
decentralized programs.  With  this information, inspection station
revenue  in  decentralized  programs  can  be  estimated.    These
estimates  for  programs  in  enhanced  I/M  areas  are  presented  in
Table 7-5.
Draft
-93-
2/26/92

-------
                                 Table 7-5
Inspection
Program
California X1
Colorado
Georgia
Houston13
Louisiana13
Massachusetts
Nevada
New Hampshire
New Yorkt
Pennsylvania
Rhode Island
Virginia
lojtai
Averages weighted
by # of stations
* Simple averages
Stations
8,752
1,500
647
1,100
140
2,800
415
243
4,300
3,838
950
220.
25055
2,088*
(i.e., non-
Station Volumes and Incomes
Vehicles Vehicles
per Year /Station
6,426,636
1,655,897
1,118,448
1,482,349
145,175
3,700,000
523,098
137,137
4,605,158
3,202,450
650,000
481.305
24r127P653
2,010,638*
•weighted)
734
1,104
1,729
1,348
1,037
1,321
1,260
564
1,071
834
684
1.301
963
Fee
$48. 3912
$9.00
$10.00
$11.25
$10.00
$15.00
$16.00
$14.00
$17.00
$8.48
$12.00
$12.50
$15.39
State
Share
$6.00
$1.50
$0.50
$3.50
$5.25
$2.50
$3.00
$1.25
$1.25
$0.48
-0-
$1.10
$3.35
Net
Revenue
$31,127
$8,279
$16,422
$10,444
$4,926
$16,518
$16,386
$7,195
$16,868
$6,675
$8,211
$14r829
$18,914
11 BAR 90 analyzers are used in these programs.   All others currently use BAR
   84 except Houston,  Louisiana, and  Rhode Island.

12 This  figure  was supplied to  EPA by the  State  in October  of 1991  and
   represents an  estimate  based upon data from  calendar  year 1990.   In its
   Third Report to the Legislature  (December 1991),  the I/M Review Committee
   reported an average cost  per inspection  of  $36.23.  This  number is based
   upon a. survey  conducted  in  September 1991,  and includes only  the  cost of
   the inspection (not the $6 fee for the certificate). The resulting figure
   of $42.23 suggests  that,  at least during September 1991,  the  average fee
   charged to motorists may  have dipped  slightly.
13
   Current  I/M inspection  is anti-tampering  only.    Station,  vehicle,  and
   income data may change  with the addition of tailpipe emissions testing'.
Draft
-94-
                                                               2/26/92

-------
     The costs  incurred by  inspection stations  are  driven by  a
number of  factors.   Labor  (i.e., the  amount  of time required to
perform the inspection and the inspector's hourly wage)  appears to
be the largest component of cost.   The  cost of the analyzer  is the
second largest  component.   PC-based  (BAR  90)  analyzers are the
latest generation of  analyzers  used  in  decentralized  programs.
Their  cost can vary  from $13,000  to  $20,000.   The  most  common
price  appears to  be approximately $15,000  each.   A  number of
service  station based  programs  in areas  required  to  implement
enhanced I/M  are currently  using  BAR  84  analyzers.   These cost
approximately  $5,000 each.   Many  stations in the older BAR 84
programs have paid off  the  cost of  their  analyzers, which in turn
decreases  their  annual inspection  expenses.   Analyzer service
contracts  and calibration  gas add lesser increments to  the total
cost.

     Estimates  were made  of  the typical   costs  incurred  by
inspection  stations,  net profits were estimated and the results
presented  in  Table 7-6.   While  large  businesses  may  be able to
afford to  purchase  current  analyzer equipment outright,  the
smaller  entities,  with  which this  analysis  is  concerned,  often
have  to  finance  these  purchases.    Analyzers are assumed  to be
purchased  and paid off over a five  year  period  at a 12% rate of
interest.   Conversations with program  personnel in decentralized
programs indicated that inspectors  are paid  about $15 per hour.
Overhead  (employers  taxes,  benefits, etc.)  is assumed to be 40%,
for a total labor cost of $21 per hour.

     Some  cost factors are subject to regional variability.  Local
data,   as  reported  by state  program officials  and  EPA Regional
offices, is used  for  such parameters  as  number  of  vehicles per
station  per year,  average length  of  test,  and  cost  of service
contracts.  Labor and equipment  costs  are  estimated as  described
previously.  In programs where the equipment specification is more
than  five  years old, the  analyzers are  assumed  to  be paid off.
This,   in turn,  increases the stations' profits.   The results are
listed in Table 7-6.
Draft                       -95-                      2/26/92

-------
                             Table 7-6

      Averaae Inspection Station_Revenues. Costs, and Profits
State
California11
Colorado
Georgia
Houston13
Louisiana13
Massachusetts14
Nevada
New Hampshire
New York11
Pennsylvania14
Rhode Island14
Virginia
Average
Average W/Q CA
Average w/o CA & NY
Vehicles
/Station
734
1,104
1,729
1,348
1,037
1,321
1,260
564
1,071
834
684
If301
963
1,086
1,091
Fee
$48.39
$9.00
$10-00
$11.25
$10.00
$15.00
$16.00
$14.00
$17.00
$8.50
$12.00
$13.50
$15.39
$12.39
$11.93
Net
Revenue
$31,127
$8,279
$16,422
$10,444
$4,926
$16,518
$16,386
$7,195
$16,868
$6,675
$8,211
$14,829
$18,914
$12r357
$10,741
Annual
Cost
$11,899
$5,202
$9,320
$7,075
$5,444
$13,498
$7,681
$4,257
$20,268
$2,811
$2,653
$5P546
$10,818
$10,238
$6,645
Net Profit
$19,228
$3,078
$7,102
$3,369
($518)
$3,020
$8,705
$2,938
($3,400)
$3,864
$5,557
$9r283
$8.096
$2,120
$4,097
     This  analysis revealed  anomalies in the  California and  New
York  programs  relative to  the others.   California  has  a much
higher average  fee than the other programs,  and estimated  average
profit  is  nearly  twice that  of the  next  highest  program.    The
estimate  for New  York reflects an  unusually  long test duration
 (see  Table  7-11)  and  shows  the average  station  operating at  a
loss; this estimate is  supported by  reports that station operators
have  sued  the  state  to  be  allowed to  charge  a  higher fee.
Therefore, average revenues  and profits were also  calculated with
data from those states omitted.

     These  figures,  based on  the  average inspection volumes  for
each state,  show that  inspection services,  by  themselves, do  not
14
   Due to the age of  the state analyzer specification, analyzer  costs are
   assumed to be paid off in stations in these programs.
Draft
-96-
                                                       2/26/92

-------
yield significant profit to the average inspection station.   While
the average  profit  is low, the  amount of revenue and  profit  can
vary  a  great  deal  among inspection stations  since  inspection
volumes vary considerably as  well.    The  best available data  on
station volumes  was  obtained  from the  California program.   The
data covers a three month time  period  and is  shown  in Table 7-7.

                            Table  7-7

                 Inspection Volumes in California
      Tests

        0

      1-100

     101-200

     201-300

     301-400

     401-500

       501+

      Total

   Total Active
Stations

 1,958

 1,156

 1,676

 1,178

   754

   469

 1.571

 8.752

 6,794
% Total

   22

   13

   19

   13

   9

   5

   18.
% Active Stations

       NA

       17

       25

       17

       11

        7

       23.
     EPA  analyzed  revenues  and profits  for inspection stations at
 different volumes;   the  results  are  presented  in  Table  7-8.
 Revenues, costs and profits are calculated as in Tables 7-5 and 7-
 6.   California  has a market based inspection fee,  (i.e.,  stations
 charge  what  the  market  will bear,  since  the state  does  not
 regulate the  fee) .   Conversations  with  California   program
 officials indicate that higher volume  stations  charge lower fees
 than the  average.  The  fees assumed for 1,200 and 2,000 inspection
 per year cases are based on figures suggested by  the state.
Draft
           -97-
                     2/26/92

-------
                            Table 7-8

              Station Revenues and Profits bv  Volume
Veh/Qtr
0
100
300
500
Veh/Year
0
400
1200
2000
Fee
$48.39
$48.39
$42.00
$32.00
Net Revenue
$0
$16,956
$43,200
$52,000
Annual Cost
$5,474
$8,974
$15,974
$22,974
Net Profit
($5,474)
$7,982
$27,226
$29,026
     These figures indicate that inspections can be profitable  if
volume  is  high,  however,   relatively  few stations  have high
inspection volumes.   Based  on the data  in  Table  7-7, 22%  of the
licensed stations perform no  inspections and therefore  are  losing
money  invested  in  equipment,  licensing,  and  training   (only
equipment  costs  are  estimated here).    An  additional 32% perform
800  inspections  per  year or  less,  and therefore appear  to  be
earning only  a modest level of profit.  22%  perform from  800  to
1,600  inspections  per year,  and  an  additional 23%  perform more
than 1,600 inspections per year.   Profitability  is  higher  in these
latter two categories.

7.4.2     Future Market in Enhanced I/M  Programs

     Test  providers will  be required to invest in new equipment
for  that  portion of  the subject vehicle fleet  that  will undergo
transient, purge,  and pressure testing.   The  total  cost to re-
equip  an  existing inspection site  to  perform  the new tests  is
estimated at  about  $144,000.    EPA  based  this  estimate   on
conversations with equipment  manufacturers over  the past  year;
more recent information indicates that a lower figure  is likely.

7.4.3     Centralized Programs

     As indicated in section  5.0,  throughput rates would  be  lower
in  centralized  lanes performing  transient, purge,  and  pressure
testing  than in  inspection  lanes  performing  the  current test
procedures.  Since programs will be able to switch from an  annual
inspection frequency to biennial  at the same time they implement
the  high-tech tests,  EPA does not anticipate  that a  significant
number  of  new  inspection   lanes  will  need  to  be  built   in
centralized programs  in order to satisfy the proposed  requirements
and maintain  waiting times at  minimal levels.

7.4.4     Decentralized Programs

     Enhanced areas  that currently  have  decentralized  programs
will have two options in meeting the requirements of  the  proposed
Draft                       -98-                      2/26/92

-------
action: they  can institute  either a multi-participant test-only
network,  or a single operator centralized system.

     If a program were to switch to a multi-participant, test-only
system, stations that currently participate in the test and repair
network would have  a  choice  between concentrating on inspections,
and  becoming test-only stations,  or  concentrating on  repairs.
That  choice  would likely  be  driven  by the  station's  current
inspection volume  and the degree to which  its  prospective income
is expected to be derived from inspection as opposed to repair and
other services.   This analysis utilizes the simplifying assumption
that stations that perform a large volume of inspections,  and that
currently  derive more income from inspection than  from repair or
other services,  would be  likely  to become test-only stations.   By
the  same reasoning, stations  that  are  more oriented toward repair
would  focus  on the  additional repair  business generated  by  the
inspections conducted elsewhere.

     Data  correlating average  inspection volume with station type
are  not  available.   However,  survey  data of  motorists  in  I/M
programs point  to  the fact that stations  that  currently  focus on
repair work  and  that  do  a steady  volume of  repairs are  often
unable  to  make  facilities  available  to  provide inspections
promptly on request15.  27% of motorists in decentralized programs
reported  being asked to bring  their  vehicles  back for  testing
another time.   20%  reported  having to  take their vehicles to more
than one station to obtain a test.  Nearly one out of three had to
leave their vehicles  for inspection.  On the average, the vehicles
had  to be  left  for  five hours.  These data suggest that a focus on
repair leads  to reduced opportunities  to  perform inspections  and
probably to lower inspection volumes as  a result.

     The  converse  appears  also to  be true.   Stations  that  are
readily  able  to provide  inspections  are often  either  unable,  or
simply  have  not  chosen  to  perform  repairs.    53% of motorists
reported taking their vehicle to another  station,  other  than  the
one  where the inspection was performed,  for repairs.

     Based on  the data  from  Pennsylvania and California,  the
following  distribution of  station  types  is   assumed for  this
analysis:
15   "Attitudes  and  Opinions Regarding Vehicle Emission Testing,
   Research.  September,  1991


Draft                  '      -99-                     2/26/92

-------
                            Table 7-9

                  Assumed Station Distributions

               Station Type               Percentage

       Service Stations                       32

       Dealerships                            22

       Independent Repair Shops               30

       Non-Engine Repair Shops                11

       Retailers                              4

       Test Only Stations                     1

     Some stations,  such as  dealerships and independent repair
shops,  would be likely to  concentrate  on I/M repairs since  their
business  already has a decided orientation toward engine  repairs.
Together,  these constitute 52% of the assumed station population.
Because of their focus  on repair,  it  is  likely that  these  stations
tend to  have  lower  inspection  volumes,  as  discussed above,  and
some of  them are  likely to  be  among  the  22%  of  stations  that
report no testing activity.   For  the  purposes of  this  analysis,  it
is assumed that  half of the  inactive  inspection stations are  in
this repair-oriented group.

     These repair oriented stations will  likely  get the majority,
though not all,  of  the  additional repair business  estimated
previously at $211  million  among all decentralized programs.   If
these stations ultimately get 85% of this business  (allowing for
15% of the repair  stations  to come from  other categories, mainly
service stations)  it will  amount to annual  revenues of  roughly
$13,000 per year.  This  would offset inspection  losses of $10,000
to $12,000 per year (Table  7-6).

     The  stations that  have higher inspection volumes  than average
are likely to be  deriving  a substantial portion of their current
profit from the inspection business and relatively  little or none
from repair.   Based on  the  California data,  it is assumed  that the
23% of the stations  that have inspection volumes of approximately
200% of  the  program average or more would be  likely  to opt  to
become  test-only  stations.    Test-only   stations,  in  those
decentralized programs where  they exist, would,  of course,  be in
this group.

     Some  stations  in  this  high volume group may  be   repair-
oriented  stations, such  as  dealerships,  independent  repair shops,
and  some  service  stations,  and may prefer to opt out of the
inspection business for more profitable repair business.   This
would create  opportunities  for other  businesses to enter the test-


Draft                       -100-                     2/26/92

-------
 only market,  including  stations  whose  current  inspection volume is
 somewhat lower.

     Current repair  revenues in decentralized  enhanced programs
 are  estimated at  $213 million.    If  this   23%  segment  of the
 stations  had  been getting 23% of  this  business  (based  on the
 foregoing  discussion,  they have probably been getting less), then
 they are giving up current annual revenues  of  $8,500 each in  order
 to pursue  the  inspection market.

     The remaining 25%  that do  not have  a  clear  orientation toward
 engine  repair,  and  that  do  not  perform  a  high  volume  of
 inspections,  are  a mix of service stations,  whose  business  is a
 mix of gasoline sales  and,  in some cases,  engine repairs including
 I/M repairs on some portion of the vehicles they test; non-engine
 repair shops,  such as  tire shops,  muffler  shops,  transmission
•shops,  etc.;  and retailers.  Members of this  group are assumed to
 make   up  the  other  half  of  the  22%  of  stations  that  do  no
 inspections.    These  stations would not be  adversely  affected by
 this  rulemaking since they  are  currently  deriving no  income from
 the inspection business.

     This   leaves  14%  of  the population  of  licensed inspection
 stations that  do  not  have  a clear orientation  toward engine repair
 and derive some income from inspections.  Since they are not high
 volume stations,  stations  in this group  do  not derive high profits
 from  inspections on the average.  Table 7-10  shows the projected
 current revenues  and  profits for these stations assuming that they
 are  evenly  distributed among  the  four low  to medium  groups  in
 Table  7-7  (those  doing  1 to 400  inspections per quarter), assuming
 that  all  stations charge  the average  fee of $48.39.    Note also
 that  the  numbers  of inspections in each  category  represent the
 mid-points of the ranges presented  in Table  7-7.    The  column
 entitled  "%  Avg  Profit"   shows the  estimated profit for  each
 category  as  a percentage of  the  program  average  profit for
 California in  Table 7-6.

     Given that the average profit in California is almost double
 that for the next most profitable program, the profits calculated
 based   on  California  data  were adjusted to reflect  projected
 national profits  for  stations with inspection  volumes  ranging from
 about  25%  to  200%  of  the  average for the  program.   The national
 average profits are based  on the figure of $4,097  obtained as the
 average net profit without  data from California and New York.
 Draft                        -101-                     2/26/92

-------
                              Table 7-10

        Revenues and Profits for Lowand Medium Volume  Stations
Veh/Qtr
50
150
250
350
% Avg
Vol.
27
82
136
191
% Total
Stations
3.36
4.90
3.36
2.38
Net
Revenue
$8,478
$25,434
$42,390
$59,346
Net
Profit
$1,254
$14,710
$28,166
$41,622
Avg. Prof it
6.5
76.5
146.0
216.0
Profit Based
on Nat ' 1 Avg
$266
$3,134
$5,982
$8,849
      The  first two categories, representing 8% of the total number
 of  stations,  appear  to earn 77% of the  program  average profit or
 less.   The two higher volume  categories,  representing  roughly 6%
 of  the total station population, derive  substantial  profits from
 the inspection business  (these  estimates are based  on data from
 California which  has  the most  profitable  inspection  program;
 profits  in other states probably do not  increase  with  increasing
 test  volume  as steeply as this analysis  suggests,  while revenues,
 on  the other hand,  do  increase  in direct proportion to volume).
 Data  on  the  relative contribution  of  inspection  revenue, compared
 to  other  types  of  business  are not  available.   Some of  these
 stations  may  be  service stations  that  are currently doing  a
 profitable business  in engine repairs,  and would continue  to do
 so.   Others,  such as  the  2.38%  earning  an estimated 216%  of the
 average profit might still opt into the test-only business where a
 high  volume  station  has opted out,  as  discussed previously.
 Others, such as the  non-engine repair shops and the retailers have
 primary lines of business unrelated to  I/M.

      However,  it may be that some of  those stations  earning 200%
 or  more of the average revenue would be unable to recoup this loss
 any other  way, and would  be  forced to  close.   The  average revenue
 loss  for  these stations would be $37,828  nationally, and $21,482
 outside California and New York.  It may  also be that some  of the
 stations   in  the  lower  profit  categories   are  so  marginally
 profitable that  loss  of inspection  business  would  result  in
 closure as well.   If 10% of this group of stations without clear
 I/M related alternatives  (14% of the  total) were  to close it would
 amount to a total of  roughly  350  stations  nationwide.

      If a single contractor centralized program were instituted in
 an  area where a  decentralized  program  is  currently operating  the
 option to pursue the test-only business would not  be  available to
 the 23% of the  station population  that would be likely to  pursue
 it.    Based on the  foregoing  analysis,  these stations  have current
 inspection volumes  of 200% or more  of  the program average,  and may
 have  average profits  of  roughly 220%  or more  of  the program
Draft
                            -102-
                                                      2/26/92

-------
average.   Members of this  group without profitable  alternatives
would also face the risk of  closure.

     The likelihood  of  closure would depend upon the  fraction  of
income derived  from  inspections.   Data on this is  not available.
Since many of these stations have other lines of business,  such  as
gasoline  sales,  auto  parts sales,  or various  types of  vehicle
repair and servicing,  the  loss  of business will not  necessarily
mean closure.  The fraction of these stations that would be unable
to  recoup  this  loss and face closure is  difficult  to estimate
given the  paucity of data.   However, if,  as before,  10% of these
stations  were  to close  as a  result  of a  switch to a  single-
contractor centralized  system,  as  well as 10%  of   the  14%  of
stations identified previously as being at risk,  then  927 stations
might close  nationwide  if all decentralized programs  in enhanced
I/M areas  switched to  centralized,  single-contractor  systems.   If
the areas  containing half of  the current  inspection stations  were
to  switch  to  single-contractor,  centralized  systems,   then
potential closures would number about 464.

     The most  severely  impacted would be the test-only  stations,
which in California  comprise 2%  of the  test  stations.   Given  that
they have no other lines of business to compensate for the  loss  of
inspection revenue, these stations would almost  certainly close  if
the  area  were  to  switch  to  a  centralized,  single-contractor
system, unless  these stations  were able to win  the  contract (some
of  these businesses  have indicated to  EPA they they would  try  to
do so) .

7.4.5      Impact  on  Jobs in Decentralized  Programs

     Table  7-11 shows the  number  of inspectors in each program,
and  the  average number  of  inspectors per  station for  all
decentralized  enhanced  programs  except Rhode  Island,  for which
data  on  the number  of  inspectors is  unavailable.   The national
weighted  average  number  of inspectors per  station excludes the
highest  and  lowest  averages  in  the  set,   those  from  New  York
 (program  officials  in  this state have indicated that the total
number of  licensed inspectors  is  likely to  include  individuals  no
longer working as inspectors)  and Massachusetts.
Draft                       -103-                     2/26/92

-------
                              Table  7-11
1* mitkyc J. JJ 	 l^jj
Stations
8,752
1,500
647
1,100
140
2,800
415
243
4,300
3,838
370
ted Averac
1- J.J.H-'y^^ ^^r*.^ 	 ff
Inspectors
18,000
2,930
2,845
2,645
513
1,208
1,249
933
21,640
19,221
1*114
re
Average
2.06
1.95
4.40
2.40
3.66
0.43
3.01
3.84
5.03
5.01
3. 01
2.05
Time per Test
25
5
10
15
15
25
10
5
40
3
5
20
California

Colorado

Georgia

Houston

Louisiana

Massachusetts

Nevada

New Hampshire

New York

Pennsylvania

yj.rcfi.nia
     Average station volumes are low  (Tables  7-5  and  7-6)  - about
four  per  day.   Given  that  there  are,  on  the average,  two
inspectors  per station,  and that  the average inspection  takes
twenty minutes to perform,  it  follows that the average  inspector
spends 40 minutes per day performing  inspections.   This  works out
to 0.08  of  an FTE  (i.e., inspections take about three  hours and
twenty minutes out of a  forty-hour work week).   Hence,  inspectors
are generally  individuals employed primarily for other  jobs (in
most cases as mechanics)  who spend a small amount  of their time on
inspections.   Communications  with  program  officials  in  these
states and  EPA's experience  in auditing  these programs  support
this conclusion.   Table  7-12  shows the estimated total  number of
FTE devoted to inspections  in the  different station  categories
developed in this analysis,  using the volume assumptions developed
previously.
Draft
-104-
                                                      2/26/92

-------
                              Table 7-12

                       Estimated Inspection  FTE

    Station Type          %_       Number  Tests/Day      FTE

Repair Oriented          52%      13,029      3         1,612

Inspection Oriented      23%       5,763      8         1,902

No Inspections           11%       2,756      0             0

Remainder                14%       3f 508      4.           579
                                          Total       4,093

     In most cases, the time  spent  on  inspections  could be  easily
re-oriented  toward other  tasks if inspection business  were to
cease, however,  some  stations might experience some  contractions
as a  result  of losing inspection business, and some  might  close,
as estimated previously.   For  the sake  of  analysis,  all FTEs
currently  devoted to inspections  in  decentralized  enhanced
programs, as shown in  Table 7-12, are  counted  as lost.  Estimates
are also made  of  additional  FTEs   lost as  a  result  of potential
station closures.

     If  a multiple independent supplier program were instituted,
it was estimated that 10% of  the 14%  of  stations  that have some
inspection  business,   and  are not  clearly positioned to  pursue
either the inspection  or repair  markets, might potentially  close.
Assuming that  these  stations  have  two  FTEs  in  addition  to
inspector FTEs, total job losses would amount  to  an additional  700
FTEs.

     In  the  event  of  a switch to a single-contractor centralized
system,  10%  of  the 23%  of  stations  that would  otherwise have
pursued  the  test-only option  would also  be  at  risk of  closing.
Potential  closures are  estimated  to  total  927.    Average non-
inspection FTE  per station in this  case is  assumed  to be 2.5 since
some larger stations would be included in  the  risk  group.  In this
case,  losses could total an additional  2,318 FTEs.

     New jobs  would be created by   the test-only program,  and  the
increased repair business  that would offset these potential  losses
to the small business  community and  to  labor.

     EPA  estimates  that   in  a  high  volume  enhanced  I/M   lane,
testing an average of  7.5  vehicles  per hour,  3-4 inspectors would
be needed  per  lane  instead  of  the  1-2  typically  employed in
current  high volume  systems.   Using an industry estimate  of  267
FTE per  million  vehicles,  and assuming a  20% retest  rate,   5,340
FTEs  are  required to  test the 33  million vehicles  in currently
decentralized programs on  a biennial basis  (this  estimate  is based
Draft                       -105-                     2/26/92

-------
on the  assumptions and methodology  developed in  section 5.2  of
this report,  "Estimated Cost of High-Tech I/M Testing").

     In a multiple independent supplier system volume would likely
be lower.   This  analysis estimates  that 4,200  inspections  per
year, or  about 16 per day would  be  likely.  Therefore,  two  or
three inspectors  per  lane would be adequate.   If two  inspectors
per  lane  were  employed,   11,525  FTEs  would  be  created  if  all
current  decentralized areas  adopted  a multiple  independent
supplier system.

     Additional jobs  that would be created  in the repair  sector
were estimated previously in  this  analysis.   Approximately  1,217
mechanic FTEs, and 506 FTEs  in auto  parts manufacturing would  be
created,  in  addition to  clerical,  delivery and  other  support
personnel.   The results are summarized in Table 7-13.

     Some new inspection  facilities  would be constructed  whether
programs adopted multiple  independent supplier networks  or  single
contractor networks,  also  creating jobs.  FTE estimates  are  based
on an industry estimate that  construction  of  an inspection  station
requires  4.79 man years   of  construction and  5.1 man  years  of
subcontracting.  An average station is assumed to have  2.4  lanes.
The number of lanes required to inspect the fleet is based  on  the
assumptions of biennial inspections  and a 20% retest  rate.   FTE
calculations are based on  the assumption that total effort,  i.e.,
modification  of  existing  structures  in those  areas  adopting
multiple  independent  supplier programs  and  construction  of  new
facilities in those areas  adopting single-contractor programs,  is
equal to that  needed  to construct  lanes  for  half of the vehicles
in decentralized  enhanced areas.   The  results  are summarized  in
Table 7-13.
Draft                       -106-
                                                      2/26/92

-------
                              Table 7-13

                    Summary of FTE  Gains  and.  Losses

     (in currently decentralized areas required to do enhanced I/M)
        Losses
              Gains
Current Inspection FTE    4,093

 Station Closures

   Multiple Independent    700

             Contractor   2,318
      New Inspector FTE

       Multiple  Independent

                 Contractor

      New Repair FTE
                                               Mechanic

                                      Parts Manufacture

                                           Construction
   11,252

    5,340




    1,217

     506

     587
Net Gain

   Multiple Independent

             Contractor

7.4.6     National Impact on Jobs
                              8,769

                              1,239
     EPA has  estimated the total FTE  in  current  I/M programs and
the  projected  changes  in FTE  nationwide  as  a  result of  the
proposed changes.   These  are  summarized in Table  7-14.   Note that
Table  7-14  includes  areas which  will   be  starting enhanced or
basic programs  from scratch,  while earlier  tallies included only
areas already operating I/M programs.
Draft
-107-
2/26/92

-------
                             Table 7-14



                   Impact  on Jobs of I/M Proposal




                 Current Test and Repair Jobs








Inspector Jobs                                     FTE




                   Decentralized Programs          6,600



                     Centralized Programs          2,500



Repair Jobs




                   Decentralized Programs            800



                     Centralized Programs          1,500




Total Current Jobs                               11,400



                 Future Test and Repair Jobs
Enhanced I/M Programs




     Inspector Jobs




        Multiple Independent Supplier




        Single Contractor




     Inspector Job Subtotal




     Repair Jobs




Basic I/M Programs




     Inspector Jobs




     Repair Jobs




Total Future Inspection and Repair Jobs




                       Other Job Gains
                     10,500




                      2.700




                 2,700 -  10,500




                      5,500








                      2,700




                        700




                11,600 - 19,400
     Parts Manufacturing
                      1,034
Draft
-108-
                                                       2/26/92

-------
     Construction                                1,800

     Small Business Services                       800

Total Net Gain in Jobs                      3,800 - 11,600

     Small  Business  Services  are  estimated  by  assuming  15
additional FTEs per urbanized area.   The 800 FTEs presented in the
table represent the  jobs generated  in the  52  urbanized  areas  that
do not have I/M  programs  now,  but will be  implementing them  as  a
result of the  proposed action.

     Whether  programs adopt  a  multiple  independent   supplier
network  or a single-contractor  centralized  one there  will  be
shifts in  job opportunities with some net  gain in either case.
Hence, the shift  to high-tech enhanced I/M may  cause significant
shifts in both business and job opportunities.   Small  businesses
that  currently  do both inspections and repairs in  decentralized
I/M programs will have to choose between the two.  Significant new
opportunities will  exist  in these  areas  for small businesses  to
continue to participate.   EPA believes there are ways  states can
help test stations make the transition to an enhanced I/M program.

7.5  Mitigating the Impact of Enhanced  I/M  on Existing Stations

     Three potential  approaches to  helping  test  stations  make the
transition are presented  here.   The first  approach would provide
direct assistance to  stations that  might be adversely affected by
the  transition to  a high-tech system.   The  second  would be  to
design the enhanced  program to  include  transitional  mechanisms to
soften the  impacts  of the new system.   The  third  would be  for
states to  establish  programs  to assist  stations and  inspectors
through  retraining  and retooling  programs.  The previous section
discussed various  strategies to assist repair technicians  in the
retest process,  including free  retests  and  priority  access  to
retest lanes,  as  well as diagnostic  and  repair assistance.

     In  some  states that are currently decentralized and will  have
to implement  enhanced I/M, analyzers have been in use for  10 years
or more  and are fully amortized.  In states that upgraded  to BAR90
equipment  (California and New York), the equipment was purchased
since 1990, and has  years  of useful life left.   A number  of other
states upgraded  their equipment  to  BAR84 in the period from  1987
to 1990.   Stations in these areas  are  likely to still be paying
for their equipment  (see the footnote to Table 7-6).   One  means by
which the  state  could provide  direct  assistance to current  test
stations would be to  set  up some  type  of state-supported  analyzer
buy-back program  for  stations  that   were no  longer  going  to
participate in either  the  test  or repair business,  possibly using
funds obtained from  inspection  fees.   BAR90  analyzers  would be
needed in the repair  business both  for  diagnostic  and repair  work
Draft                       -109-                     2/26/92

-------
as well  as  to check whether  repairs on  old technology vehicles
were effective.   BAR90 analyzers could also be used to  test  older
technology vehicles  in test-only  stations.   This  concept  would
allow stations that  were planning  to  leave the  I/M  business to
recover all  or part of  their capital investment for equipment that
could not be  used for diagnostics  and repair.    Such a buy-back
program might  allow a fairer transition to test-only  status.

     A related strategy would  be for EPA,  the states,  and industry
to support  the development  of new  and  improved uses  for  BAR90
analyzers so  that  current  as  well  as  future  analyzer owners can
use this technology more  effectively in  the repair process.  In
particular,  it was California's  intent  in developing  the  BAR90
specification for  the  computer in the analyzer,  which  is  an IBM
386 DOS-based system,  to become a platform for  vehicle diagnosis
and  repair.    EPA, the  states, and  industry could  potentially
provide technical  and  financial support  to speed the development
of such  software.   This would not  only  make  better use  of the
equipment in the field but  would serve as an excellent mechanism
for providing  critical technical assistance and  training  to the
repair community.

     A  second strategy  to mitigate  the impacts  is  to  design
transitional features  into  the program.  One approach would  be to
allow test and repair shops to  continue to do testing on vehicles
not  subject  to the  transient/purge test for some  transitional
period (note that  EPA's recommended enhanced program would require
biennial, transient/purge  tests on  1984 and  later  model  year
vehicles, and biennial  steady-state  tests  on older vehicles).  EPA
is proposing to permit a phase-out  of the  decentralized test-and-
repair portion  of the program such that all vehicles  would be
inspected in  test-only stations starting  January 1,  1996.   This
would allow  these decentralized  stations to continue  to  obtain
revenue to  recover the investment made in testing  equipment and
would allow additional time to plan other strategies to replace
the income to  be lost from testing.

     A related  approach is  to  allow  vehicles  that  have  failed
initial  inspections  in  test-only stations  to  be   retested in
existing  test  and  repair   stations  using conventional  test
techniques during  the  first inspection cycle.   This  would  allow
those stations to  attract  customers,  conduct testing  and perform
repairs,  with the  added benefit  of  sparing  the customer   from
returning to the test-only station for the retest.

     A third strategy  would be to provide targeted assistance to
stations to  assure they  were  able  to provide  high-tech  repair
services.   This  would require pre-program start-up  training to
bring repair  technicians in  these stations up  to speed  on the
high-tech tests,  vehicle diagnosis,  and  engine  repair.   It  might
mean tuition grants or  other financial  assistance.  This dovetails
with  stronger repair technician  training programs  which EPA
envisions as being part of  future I/M requirements, but  differs in
Draft                       -110-                      2/26/92

-------
terras of funding,  timing,  and intensity.   This approach might also
include financial assistance  to stations can for the  purchase  of
equipment to  perform  sophisticated diagnosis  and  repair on  new
technology vehicles  or to  upgrade tools  and equipment  for  more
sophisticated diagnosis and repair.

     EPA  does not  consider  this  to  be  an exhaustive  list  of
possible strategies.   The Agency  expects  to  receive comments from
the states and affected industries  on  additional  ways  to  mitigate
the impacts  of  upgrading to  an enhanced program on current  test
and repair stations  in decentralized programs.
 Draft                       -111-                     2/26/92

-------
8.0  ONBOARD DIAGNOSTICS AND ON-ROAD TESTING

8 .1  Onboard Diagnosticsr Interim Provisions

     EPA is required to issue onboard diagnostic (OBD) regulations
by  May 15, 1992,  while I/M  programs  will begin  OBD checks  two
years  after the regulation has been issued.   OBD checks  are  not
currently  a part  of EPA's  performance  standard and  no  credit  has
been assessed for such checks in the MOBILE4.1 model/ such will be
determined after formal  issuance  of  OBD regulations.    For  the
purpose of this cost-benefit  analysis,  the impact of OBD  has  not
been addressed.  The  impact  of OBD will  be  relatively minor up
until the  attainment deadline for serious areas,  in November 1999.
EPA  will   certainly  revisit the  issue once  OBD  regulations  are
final and  as their  implementation clarifies the potential  of this
strategy in an I/M setting.

8 . 2  On—road Testing,. Interim Provisions

     Section 182(c)(3)(B)(i) of the Act  requires EPA to  establish
a  performance  standard for  enhanced  I/M  "including  on-road
emission testing."   The Act does not specify how programs or  EPA
are to  address  the "on-road testing" requirement,  and neither is
on-road testing defined within the  Act itself.  While potentially
a  fruitful supplemental testing  strategy, it is  clear from  the
legislative history of the 1990  Amendments  that on-road  testing
was not viewed  as a potential replacement for  I/M programs, as  has
been suggested by some.  Under the section addressing enhanced  I/M
programs,  the  legislative history  notes:

     On-road emission testing is to  be  a  part  of the  emission
     testing  system,  but is  to be a  complement  to  testing
     otherwise required since on-road testing  is not  intended
     to replace such testing.   On-road emission testing may
     not be practical  in every season  or  for every  vehicle,
     and is not required.   However, it should play some  role
     in the state  program.   It is  the Committee's intention
     that  states should take  into consideration  that the
     results of on-road emission testing,  when used,  have not
     been  shown  to  be consistent with  Federal   emission
     testing procedures. [Emphasis added]

     In its current proposal  for  implementing this  requirement,
EPA  has  specified  that  on-road  testing  be defined  as "the
measurement of  HC,  CO, NOX, and/or C02  emissions  on any  road or
roadside in the nonattainment  area  or  the I/M program, " and that
it be  required  in enhanced programs and  an  option  for  basic  I/M
areas.   Minimally,  the on-road testing  effort must  evaluate  the
emission performance  of at least 0.5% of  the subject fleet each
year.   EPA believes that the  on-road  testing requirement  can be
fulfilled by a range of approaches,  including,  but  not limited  to:
Draft                       -112-
                                                      2/26/92

-------
remote sensing  devices  (RSD) ,  random road-side  pull-overs using
tailpipe tests  and emission  control device checks,  or road-side
pull-overs of vehicles with high RSD readings,  as well as through
the use  of  portable analyzers  that  can be placed  on the vehicle
prior to on-road driving.

     Of  the above  approaches,  RSD  has  gained  the most  public
attention and has generated considerable  interest.   The objective
of RSD is to remotely  measure the concentration of emissions from
vehicles as .they are operated  on public  roads,  and  in this aim,
RSD fully meets the definition of an  on-road testing strategy.  In
its current  version, RSD  works by  focusing  a  beam, or,  in some
cases, multiple  beams,  of infrared  light across  the roadway into
an infrared detector.   The concentration  of certain pollutants in
the exhaust stream  are  then determined by measuring the amount of
infrared light  absorbed  at  specific  wavelengths  as  it  passes
through  the  exhaust in  much  the same way  that astronomers study
stellar  atmospheres by  analyzing specific  portions of  a  star's
spectrum.  The analysis  is tied to a vehicle through the use of a
video  camera which records  the  vehicle's license  plate  as  it
passes through the beam(s).

     Given   its  non-intrusive  nature  and  potentially  high
throughput capabilities, RSD warranted further investigation.  EPA
has  conducted a preliminary  analysis of  RSD  (see Appendix  L,
"Identifying  Excess  Emitters  with  a  Remote  Sensing  Device:  A
Preliminary Analysis")  that investigated  the comparability of the
results  obtained to those in  the 2500 rpm/Idle  test.   EPA found
that,  under  controlled  conditions and  using  stringent  cutpoints,
RSD's  performance in measuring CO emissions was comparable to the
2500  rpm/Idle  test.  Since then, other researchers, such  as the
California  Air  Resources Board  (CARB),  have  found that  the
accuracy of the  device  for  measuring HC  emissions,  while less
accurate than  for  CO,  is within  a  practical  range  for  roadside
monitoring.   For example, CARB researchers recently  reported to
the  CARB I/M  Review  Committee16 that the  device, under  highly
controlled operating conditions,  yielded  results  that compared to
calibrated  on-board measurements  as follows:   the  remote sensors
accurately  measured CO within  ± 5% and HC  within ±  15%  of the
instrumented vehicle measurements,  respectively.   EPA,  however,
knows  of no current RSD methodology for detecting and measuring
NOX emissions,  although developmental work is  being  done in this
area.   EPA  encourages the states to be  innovative in  fulfilling
the on-road testing requirement.

     There have been and continue to  be a number of efforts in the
area  of RSD  evaluation,  including  those  at  the  University  of
16 D. Lawson,  J. Gunderson, "In-Use Emission Study and High Emitter Phase,"
   Presentation to  I/M Review Committee, Sacramento, California,  January  29,
   1992.



Draft                       -113-                     2/26/92

-------
Denver,  where  the  first RSD testing strategies  were  developed.
The bibliography17 of  research in this area  continues to  grow.

     Currently,  it is  difficult  for EPA  to  project  a  standard
"emission  credit"  for  on-road  testing  for  the  purpose  of
performance standard modeling.   Hence,  for  the  purpose  of  this
cost-benefit  analysis,  the  impact  of  on-road  testing  is  not
addressed.    Nonetheless,   emission reduction  credits  will  be
assessed for on-road  testing efforts  once additional experience is
gained in  the  actual  use of various on-road  testing  strategies,
including RSD  technology.   Under  EPA's  current proposal,  on-road
testing programs required  by  the  Act  "shall  provide  information
about  the  emission performance  of in-use  vehicles,  by  measuring
on-road emissions  through  the use  of  remote  sensing devices or
roadside  pullovers  including  tailpipe  emission  testing.    The
program shall collect,  analyze and report on-road  testing data" as
part  of  the state's  annual  report  to EPA.    EPA  shall use  this
data,  in  conjunction with  data  gathered as  part  of the Agency's
on-going  investigation of  these  testing  strategies,  to  develop
testing protocols and guidance.
 17 In addition to the sources  referenced in Appendix L,  the  following works
   have contribute to the body of information concerning RSD.

   1. D.R.  Lawson,  P.J.  Groblicki,  et.  al.,  "Emissions  for  In-use  Motor
   Vehicles in Los Angeles:  A Pilot Study of Remote Sensing and the Inspection
   and Maintenance  Program," Journal of the Air Waste Management Association,
   40(8): 1096 (1990)

   2. R.D.  Stevens and  S.H.  Cadle,  "Remote Sensing  of  Carbon Monoxide
   Emissions,"  Journal  of  the Air Waste Management  Association, 40(1):39
   (1990)

   3. G.A. Bishop,  D.H.  Stedman, et.  al.,  "IR Long-Path  Photometry, A Remote
   Sensing Tool for  Automobile  Emissions," Analytical Chemistry, 61, 671A-677A
   (1989)

   4. D.H. Stedman  and G.A.  Bishop, "Evaluation of a Remote Sensor for Mobile
   Sources CO Emissions," Report to the Environmental Protection Agency, EPA-
   600-S4-90-032.

   5. D.H. Stedman,  G.A. Bishop, et. al., On-Road  CO Remote  Sensing in the Los
   Angeles Basin.  Final  Report on Contract No. A932-189,  California Resources
   Board, Research  Division, Sacramento,  1991.

   6. D.H. Stedman  and G.A.  Bishop.  An Analysis  of On-Road Remote Sensing as
   a Tool for Automobile Emissions Control, ILENR/RE-AQ-90/05, Final Report to
   Illinois Department of Energy and Natural Resources, Springfield, IL,  1990.



 Draft    '                   -114-                       2/26/92

-------
Appendix A


Appendix B

Appendix C


Appendix D


Appendix E


Appendix F


Appendix G


Appendix H


Appendix I

Appendix J

Appendix K

Appendix L
Cost Effectiveness Model Version 4.I Source Code
Listing

Tech4 Model Version 4.1 Source Code Listing

Evaporative Emissions and Running Loss Emission Factor
Derivation

Purge and Pressure Test Effectiveness Figures and
Spreadsheet

Regression Analyses and Scatter Plots for Fuel
Injected 1983 and Later Vehicles

MOBILE4.1 Technology Distribution and Emission Group
Rates and Emission Levels

Exhaust Short Test Accuracy: IM240 vs. Second-chance
2500 rpm/Idle Test

Data Analyses for Appendix G: 1983 and Newer PFI, TBI,
and Carbureted Vehicles

Evaporative System Purge and Pressure Diagrams

Evaporative System Failures and Repairs

MOBILE4.1 Performance Standard Analyses, By Option

Identifying Excess Emitters with a Remote Sensing
Device: A Preliminary Analysis
Appendix M   Model Year Failure Rates by Test Type
Draft
                -115-
2/26/92

-------
                       APPENDIX A
COST EFFECTIVENESS MODEL VERSION 4.1 SOURCE CODE LISTING

-------
                                             Appendix A

c
C                #$*$*$*$*   CEM4.1   #$#$#$#$#
C
C           The Coat-Effectiveness Version of MOBILE4.1
C
C
C
C  Re-Revise repair coat adjustment for recurring IM240.   Dec 10, 1991
C  Fix problem in PCLEFT Tech 1 & 2 credits.
C
C  OEM version 4.IE                                        Dec 4, 1991
C  Trial change to CPFUEL, adding IMFLAG to  coverage checks.
C  Trial init of CSIZE in RESET.
C  Changed Frg/Pra repair EF benefit from 7.4% to 5.9%
C  Install new PPEFF array in FAIL routine for revised
C    detection rates.
C  Change block FAIL to FAILRT to avoid confusion with
C    Function FAIL().
C  CEM version now based on November 1991 MOBILE4.1,
C
C  CEM version 4.ID                                        November 19, 1991
C  Revise repair cost adjustment for recurring IM240
C    from .0199/10K miles to .0373 constant.
C  Block Data from .0199 to  .0373, and remove corresponding FE
C    1.87 adjustment.
C  Still needs revisions for November M4.1:
C    multiple runs and speed correction factors, etc.
C
C  CEM version 4.1C                                         November 6, 1991
C  Re-installed 1.87 FE benefit adjustment factor for non-111240
C   (as it has been for IM240)
C
C  CEM version 4.IB                                         October 30, 1991
C  Revised FAIL to skip over program reductions when doing
C  recurring case for repair costs.
C  Still has excess code for multiple evaluation years
C  but can not be used.
C                                                           October 18, 1991
C
C  CEM Version 4.1A, 10/17/91, is the last version prior to
C  cutting out the excess evaluation year capability (JCYDX > 2).
C  It includes much unused code and variables, but is fully M4.1
C  compatible plus a couple little fixes  (HDV handling in FAIL routine).
C  It includes input file list capability and 'interactive' I/O file
C  specification, but needs revision for November release of M4.1.
C
C
C
C                ****** Release MOBILE4.1 ******             November 4, 1991
C
C
C  Release MOBILE4.1 is derived from 102  (June 12, 1990) MOBILE4.
C
C  See Dnn.T-NOTES for a chronologically ordered summary of the coding changes
C  performed up through the current In-House Version.  See Dnn.T-UGBASE for the
C  In-House MOBILE4 User Guide Supplement for Version Dnn.  All In-House
C  Version files have the Dnn prefix.  The backup tape is *M4.1*, mounted by
C  Dnn.C-TAPE.  All files are on ST99.
C
C  Version 00 was opened 02/28/91 and closed 03/04/91.
C  Version 01 was opened 03/04/91 and closed 04/09/91.
C  Version 02 was opened 04/11/91 and closed 04/30/91.
C  Version 03 was opened 05/01/91 and closed 05/15/91.
C  Version 04 was opened 05/16/91 and closed 05/24/91.
C  Version 05 was opened 05/28/91 and closed 06/05/91.
C  Version 06 was opened 06/06/91 and closed 06/21/91.
C  Version 07 was opened 06/26/91 and closed 07/29/91.
C  Version 07 was rereleased as Corrected M41 on 08/28/91.
C  Version 07 was rereleased as FINAL M41 on 11/04/91.
C
C
C  MOBILE4 is the fourth version of the FORTRAN program implementation of the
C  Mobile Source Emission Model.  It updates and replaces MOBILES. The model
C  estimates HC, CO and NOx exhaust and HC evaporative emission factors by
C  motor vehicles.  For information on using the program and on the
C  differences between MOBILES and MOBILE4,  consult the User's Guide to
C  MOBILES,  EPA 460/3-84-002, and the User's Guide to MOBILE4,
C  EPA-AA-TEB-89-01.   Otherwise questions should be directed to:
C

-------
                                            Appendix  A

C         Office of Mobile Sources
C         Emission Control Technology Division
C         Test and Evaluation Branch
C         2565 Plymouth Road
C         Ann Arbor, Mich. 48105
C         Telephones 313-668-4462
C                    FTS 374-8462
C
C
C  Source Code Documentation
c  	
C
C  The source code itself is commented to assist the users who require a
C  more detailed understanding of the program than provided in the Guides.
C  This section provides the following information:
C
C  1) Execution summary
C  2) Source code structure
C  3) Source code comments structure
C     a) Subroutines and functions header format
C     b) BLOCK DATA subprograms header format
C     c) In-line comments
C  4) Subscript dictionary
C  5) Subroutine / function parameter dictionary
C
C
C  1) Execution summary
C
C  To run the program, attach the input file to logical I/O Unit 5 and the
C  output file to Logical I/O Unit 6.  If supplying optional alternate I/M
C  credits, attach the credits file to Logical I/O Unit 4.  Refer to the
C  User's Guide to MOBILE4 for more detail on setting up runs and check
C  your operating system documentation for the system commands for compiling
C  and executing the program.  There is discussion of running the program on
C  the IBM PC AT under DOS in the User's Guide.
C
C  For information on the source code itself,  refer to the Programmer's Guide
C  to MOBILE4, EPA-AA-TEB-89-01.
C
C
C  2) Source code structure
C
C     This section describes the layout of the code plus the numbering and
C     naming conventions.
C
C     The source code for the subprograms is grouped as follows:
C
C        program header
C        MAIM
C        MOBILE subroutine
C        QUITER subroutine
C        input subroutines
C        computation subroutines S functions
C        pointer functions
C        output subroutines
C        BLOCK DATA subprograms
C
C     Within each grouping and nesting level,  the routines usually appear in
C     calling order.
C
C     The User's Guide listing of the code has each subprogram starting at an
C     integral multiple of 1000, simplifying the construction of the Table
C     of Contents.  That listing also has the comments in mixed case, to
C     improve readability.  The MOBILE4 tape being mailed out to the user
C     community has this mixed case version in File 1.  An upper case only
C     version, with hyphens replacing underscores and blanks replacing plus
C     carriage controls, is in File 2, for use by installations not able to
C     process these features.  For example, EPCDIC PN character set is upper
C     case only.
C
C     Statement number ranges are defined for usage as follows:
C
C          1 -   9  -  not used
C         10 -  98  -  branching
C               99  -  branch to last RETURN / RETURN1 / STOP before END
C        100 - 199  -  READ formats
C        200 - 299  -  WRITE prompt and output formats
C        300 - 499  -  WRITE error and warning messages
C        500+       -  not used
C

-------
                                             Appendix A

C     Statement numbers are local.  For example, within each  subprogram,
C     branch numbers start at 10, READ formats  start at 100,  etc.
C
C     Given FORTRAN'S 6 character maximum on subprogram name  lengths, useful
C     naming conventions are hard to come by.   There is also  the limitation
C     posed by not wanting to change arbritrarily  all  the  subprogram names of
C     the predecessor program, MOBILE3.  Hence, the only conventions are:
C
C     a) Input driver sections all have the  suffix 'SEC'.
C     b) Pointer functions all begin with 'I'  (to  get  default INTEGER typing).
C     c) Pointer functions all contain/end with 'PT' or 'PTR'.
C     d) Output subroutines all have the prefix 'OUT'.
C
C     Otherwise, new subprogram names were picked  to indicate the purpose of
C     the code.
C
C  Note: some features during the course of  the model  development were
C  subsequently removed or "turned off".  The latter case  would occur when
C  the prospect of restoring the feature for in-house  MOBILE4 or for MOBILES
C  seemed likely.  Commented out code to this end  is done  with a 'CC1 .
C
C
C  3) Source code comments structure
C
C     Each subprogram contains a documentation  header, set up in a  standard
C     format as described below.  All variables and arrays used in  the code
C     are defined in the comments.  The subroutine / function headers define
C     local parameters and local variables / arrays.   The  BLOCK DATA headers
C     describe the COMMON storage areas.  Array subscripts and subprogram
C     parameters are specified in dictionaries  following this section.  The
C     intent is to allow the user to scan the source code  on-line for
C     definitions, rather than have to leaf  through hard-copy documentation
C     such as the Proramtiter'a Guide to MOBILE4.
C
C     a) Subroutines and functions header format
C
C        
C        C
C        C  
C        C
C        C  Called by 
C        C
C        C  Calls 
C        C
C        C  Input on call:
C        C
C        C    parameter list: 
C        C    common block(s):
C        C    // 
C        C
C        C          .
C        C
C        C  Output on return:
C        C
C        C    parameter list: 
C        C    common block(s):
C        C    // 
C        C          .
C        C          .
C        C
C        C  Local array subscripts:
C        c
C        C  (<#>,...)  -  (, ...)
C        C          .
C        C          .
C        C
C        C  Local variable / array dictionary.
C        C
C        C   Name   Type               Description
C        C  	  	  	
C        C            
C        C
C        C
C        C
C        C  Notes:
C        C
C        C  
C        C
C        C
C        
C

-------
                                             Appendix A

C        Notes on format:
C
C        A local variable is one not passed  in via the parameter list
C        or a common block.  It can be used  in the argument list of a
C        subprogram call made by the defining subprogram.
C
C        Also, if a section of the header is not needed, its format is
C        omitted.  For example, if no subprograms are called, then the
C        'C  Calls ' line is not included.  "Notes" are always included.
C
C        Finally, 'Output on return1 includes any variable / array whose
C        contents can be modified by the subprogram.  This usually will not
C        include data that can be changed by subprograms called by the
C        subprogram, aince the comment headers on those subprograms will
C        indicate what is being passed back up.
C
C     b) BLOCK DATA subprogram header format
C
C              BLOCK DATA
C        C
C        C  BLOCK DATA subprogram <#>
C        C
C        C  
C        C
C        C  Common block array subscripts:
C        C
C        C  (<#>,...)  -  (,...)
C        C          .
C        C          .
C        C
C        C  Common block dictionary:
C        C
C        C   Name   Type              Description
C        C  	  	  	
C        C  //:
C        C            
C        C
C        C
C        C
C        C  Local array subscripts:
C        C
C        C  (<#>,...)  -  (,...)
C        C          .
C        C          .
C        C
C        C  Local array dictionary:
C        C
C        C   Name   Type              Description
C        C  	  	  	
C        C            
C        C
C        C
C        C
C              COMMON // ...
C
C
C
C        C
C              DIMENSION (or type)  (subscript(s))
C
C
C        C
C              EQUIVALENCE (,)

C                 '.
C
C        C
C        C  //: 'a labels, purpose or data
c        c               or on the structure of the initializing DATA
C        C               statement(a)>
C        C
C              DATA /...
C
C
C
C                 .../
C        C
C        C  //: ...
C
C

-------
                                             Appendix A

c
c              END
c
C        Notes on format:
C
C        Local arrays are defined and then equivalenced to  common block  arrays
C        that are too large be intialized in 1 DATA  statement.   Too large -
0        would require greater than 19  continuation  lines,  the FORTRAN limit.
C
C        The local arrays dictionary is often compressed by providing a
C        generic definition of the conventions used  to construct and name the
C        set of local arrays to be equivalenced to a given  common array.
C
C        "//:" prefix is included only if a  large number of common
C        blocks are being initialized in a single BLOCK DATA  (see BD 23).
C
C
C     c) In-line statements
C
C        In-line comments are intended  to describe the algorithm being used
C        by the code.  Where MOBILE4 is identical to MOBILES  in  operation,
C        the MOBILES comments have usually been left intact.  Where the
C        method has been updated or the whole section / subprogram is new,
C        the in-line comments describe  the MOBILE4 methodology.
C
C        Particular attention is paid to the use of  indices and  pointers.
C        Some of the ambiguities of MOBILES,  such as the use  of  IDX vs JDX,
C        have been cleared up.  The usage of new keys and pointers is specified
C        in detail where they occur.
C        occur.
C
C
C   4) Subscript dictionary
C
C     Subscript variables used to index arrays in MOBILE4 are globally
C     defined, not in the sense that the values are  available to all
C     subprograms, but rather that the  same  label is used for the same
C     subscript function throughout the code.  IF is the 3  pollutants index
C     for  all of MOBILE4, although its  instances are usually  as  a local
C     variable.  As array subscripts, all these variables must be of type
C     INTEGER. The default INTEGER typing syntax rule is employed by having
C     all  subscripts begin with a letter from the range I through N.
C
C       Name    Range               Description

C     IAD       1-5    index into ID dimension of ATR for ADJFAC equation
C
C     IADJ     1-2    index into CSIZE adjustment vector ADJ7AC
C
C     IAY       1-2    anti-tampering program start  year flag:
C                        1 - in year range up thru start year
C                        2 - in year range after start year
C
C     IB       1-3    bag number:
C                        1 - cold start
C                        2 - hot stabilized
C                        3 - hot start
C
C     IBER     1-3    terms of user supplied new emission  rate  equation:
C                        1 - zero mile  level (intercept)
C                        2 - deterioration rate  (slope)
C                        3 - deterioration rate  (slope), LDGV 81+ HC/CO  50K+
C
C     1C       1-7    overlap effects  categories computed  direct from base
C                      tamper rates; each case is the set intersection of
C                      the indicated disablement types:
C                        1 - AorE * CATS
C                        2 - AorE        * NCKS
C                        3 - AorE               * TNKS
C                        4 - AIRS * CATS * NCKS
C                        5 - AIRS * CATS       * TNKS
C                        6 -        CATS * NCKS
C                        7 -        CATS       * TNKS
C                      AorE - AIRS if HC or  CO
C                           » EGRS if NOx
C

-------
                                             Appendix A
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
          8-12   no overlap effects categories computed from the 7
                 overlap cases:
                   8 - AIRS
                   9 - CATS
                  10 - NCKS
                  11 - INKS
                  12 - 


ICAT      1-3    catalyst technology configurations:
                   1 - ox catalyst / no catalyst
                   2-3 way catalyst
                   3-3 way + ox catalyst


ICH       1-20   character string array position index:
                   1 - 1st 4/8 characters (CHARACTER*4/8)
                   2 - 2nd 4/8 characters



                 also serves to subscript any other CHARACTER type array.


ICOEF2    1-6    coefficient number of 2nd degree polynomial
                 (Oth (intercept) to 2nd decree term coefficients)


ZCOEF5    1-6    coefficient number of 5th degree polynomial
                 (Oth (intercept) to 5th decree term coefficients)


ICOL      1-3    index into EFF's column dimension  (BD 24 defines abbrs):
                 1 - PREV CAT ;  2 - PREV ; 3 - SOBS


ICOLD     1-2    index into cold CO standard case of MYCOL fi CCSTD


ICR       1-2    Replace I/M credits inclusion vector NUDATA's array
                 identifier subscript:
                   1 - (CR12HC £ CR12CO) - Technologies 1-2
                   2 - CRED4V - Technologies 4+ LDGV


ICS       1-NT   secondary index: selects pointers to category
                 sizes to be summed up to get SUB (in C8TO11) and
                 DIV  (in A8TO11)


IC11      1-11   positive tampering effects categories 1-11; 12 is the
                 no effects case.


IC7       1-7    overlap tampering effects categories 1-7


ID        1-9    disablement types (1-5,9 for bag, 6-8 for evap):
                   1 *• AIRS " air pump disabled
                   2 - CATS - catalyst removed
                   3 - NCKS - misfueled - filler neck
                   4 - INKS - miafueled - tank (all but filler neck)
                   5 - EGRS - E6R disabled
                   6 — EVAP ™ evaporative system tampered - cannister only
                   7 - PCVS - PCV disabled
                   8 - CAPS - evaporative system tampered - cannister and
                              gas (fuel inlet) cap
                   9 - MISF - misfueled - all sources - NCKS + TNKS


IDBAG     1-5    bag exhaust disablement types (see ID)


IDPB      1-3    misfueling related disablement types:
                   1 - NCKS - misfueled - filler neck
                   2 - TNKS - misfueled - tank (all but filler neck)
                   3 - MISF - miafueled - all sources - NCKS + TNKS


IDU       1-5    index for Diurnal (DU) indicating full, 8AM-11AM,
                 10AM-3PM, 8AM-2PM, and multiple


IDUSER    1-8    ID cases £ order of cases presented to user in TAMFLG
                 (TAMZML £ TAMDR) S ATPFLG  (DISTYP) options.  IDUSER
                 differs as follows: 4 = MISF.  GETTAM derives TNKS =
                 MISF - NCKS for TAMDR £ TAMZML and stores the read and
                 computed values in ID order. ATPAER does not need TNKS
                 £ uses DISTYP in the read in IDUSER order.


IDX       1-25   model year (my) window index:
                   1 = 25+ my age group cell «> tamper effect is
                       always 0.0 => this case skipped in TAMPER  (only)
                   2 - 24th my cell
                   3 - 23rd my cell

-------
Appendix A
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c




IEG








IEK

IELST

IEQU


IERNEW



IEVEQ


IEVF





IEVX


IE1ST

IFAC


IFDS



IFG







IFL

IFUEL


IG










IGCSF




IGD


IGER

IGIDL





1-7








1-3

4-7

1-20


1-20



1-9


1-4





1-32


1-3

1-100


1-3



1-13







1-2

1-3


1-20










1-3




1-3


1-20

1-15


24 = 2nd my cell
25 - 1st my cell - calendar year (ICY)

tampering effects group:
1 - air pump disabled, air pump only equipped
2 =» air pump disabled, air pump/catalyst equipped
3 » catalyst removed
4 = miafueled catalyst
5 - EGR disabled
6 - EGR disabled / catalyst removed
7 - EGR disabled / misfueled

catalyst emission effects groups: IEK - IEG - 1

last non-zero tampering effects group size, given IFG

pointer to idle equation to be zeroed out due to entry
of new exhaust ef equation by user

pointer to location in default ef rates & model years
arrays where new ef equation's parameters are to be
stored (IERNEW m 0 -> no space for new equation)

HS and DU equation term coefficient - see definitions
for HSEQ in BD 09 and DUEQ in BD 11

index for gram evap rates EVF, GREVP and VGREVP:
1-hot soak
2=in-use diurnal
3— multiple diurnal
4—crankcaae

model year (my) window index for evaporative data input.
i.e. to the lEVXth myg 'a rate(s) for a given NO set

first non-zero tampering effects group size, given IFG

index into new (user supplied) emission rates arrays;
the user can enter up to 100 replacement cases .

fuel delivery system type:
1 » carbureted 2 — fuel injected OR
1 - CARbureted 2 = TBI 3 = FFI

flag names for prompts (listed by common block used) :
/FIAGS1/ /FLAGS2/ /FLAGS3/ /FLAGS4/
1 - TAMFIiG 4 - MYMRFG 8 - ATPFLG 12 - PRTFLG
2 - SPDFLG 5 - NEWFLG 9 - TEMFLG 13 - IDLFLG
3 - VMFLAG 6 - IMFLAG 10 - RLFLAG 14 - NMHFLG
7 - ALHFLG 11 - OUTFMT 15 - HCFLAG
The 14th flag (or 0 - PROMPT) is read without a prompt.

myg bounds: 1 - First year 2 - Last year

fuel type: 1 - Gasoline, 2 - Ether Blend, 3 - Alcohol
Blend

model year group (my?) pointer:
 - null string,C,CAT,D,E,F,L,LM,S,T,U,W,l,2,3,4
IG links a data array to its corresponding myg index
array. These data arrays all have my as a subscript
plus have cases of contiguous subsets of years with
the same data point. Core space is then saved by
storing each point only once per my subset and then
constructing an index array to enable the program to
find the location IG of the data point for a given
my.

model year group index in the array CSF1ST:
1 - pre-1981 model years
2 = 1981-1983 model years
3 - 1984+ model years

MYG index into 3rd dimension of TAMZML & TAMDR
1 - pre-1981 2 - 1981-83 3 - 1984+ for LDGV/T
1 - 1981+ for HDGV
myg pointer into basic emission rate arrays from IERPTR

myg pointer into idle emission rate arrays from IDLPTR


-------
                     Appendix  A
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
IG50 l-i:


IGTL 1-4

IH 1-3




IHG 1-2



IHDSAL 1-2



IBRD 1-6







ILU 1-2




IM 1-2



IMDL 1-2



INO 1-2


IF 1-3




IFAR 1-5





IPARTL 1-3


IPG 1-2



IPIM 1-2



IP1 1-3


IP2 1-3


IF50 1-2


IQG 1-6







myg pointer into the 50K+ deterioration rate arrays
ERB50K S ERU50K (the arrays differ in lookup algorithm)

running loss HC: emission rate MYG pointer

non-exhaust HC emission case:
  1 =• hot soak evaporative
  2 = diurnal evaporative
  3 ** crankcaae

non-exhaust HC technology / tampering case:
  1 - evaporative
  2 » crankcase

HDGV sales fraction case
  1 - HDGV weighing < 14000 Ibs
  2 - HDGV weighing >- 14000 Ibs

evaporative temperature types:
  1 « hot soak
  2 - running loss
  3 - diurnal, 8 AM
  4 - diurnal, 10-11 AM
  5 - diurnal, 2 FM
  6 — resting loss

bounds of a value range (ex.: tags of nonzero
category sizes):
  1 ** lower bound
  2 - upper bound

I/M program flag:
  1 - I/M program not in effect for MY x IV x ICY case
  1 - I/M program is in effect for MY x XV x ICY case

coefficients for trips per day model equation:
  1 - intercept
  2 - slope

index into NOYES character string vector:
  1 - No     2 - Yes

pollutants:
  1 = HC  = hydrocarbon
  2 «• CO  " carbon monoxide
  3 - NOx — oxides of nitrogen

parameters identifying where and when to apply user
supplied new emission rates:
  1 ™ region
  2 «• vehicle type     4 « first my of range covered
  3 - pollutant        5 - last my of range covered

index for Fartial Diurnal (PARD) indicating
8AM-11AM, 10AM-3FM and 8AM-2PM

pollutants grouped for tampering effects & SK2 scfs:
  1 - HC and CO
  2 - NOX

pollutants affected by I/M programs:
  1 - HC
  2 - CO

lower bound of output IP loop  (first pollutant to
have results printed)

upper bound of output IF loop  (last pollutant to
have results printed)

pollutant types affected by 1981+ LDGV 50K+ change in
deterioration rate

emission control equipment group; 2 more groups
computed from the stored 6 groups data:
  1 *• air pump only
  2 = air pump / catalyst
  3 =• catalyst only
  4 = EGR only
  5 = EGR / 3-way catalyst
  6 - no EGR / 3-way catalyst

-------
                                             Appendix A

c
C     IR        1-2    region  code:
C                      IR haa  same value range aa IREJN,  but is not necessarily
C                      the  scenario  field's value
C                        1  - low altitude
C                        2  - high altitude
C
C     IREC      1-999   input  record number,  on a given READ
C
C     IRON      1-15   index into EFF's row dimension - the 8 ID cases split
C                      into 15 subcases (see BD 24 definition of EFF)
C
C     IRT         1-4    temperature cases in RVP cf calculations,  the values
C                        referenced  depending on the (sub)array.  See /RVFEX1/.
C
C     IRQ1        1-3    RVP cf equation parameters set 1 (RVP > 9.0 adj.):
C                        RADJCF-(bl+b2*RVP)/b3
C
C     IRQ2        1-2    RVP cf equation parameters set 2 (T interpolation)
C                           1 - initial temperature (T range Ib)
C                           2 - temperature difference (interval length)
G
C     IRQ3        1-3    RVP & high  T combined cf formula coefficients:
C                        TCF (IB)-EXP (al* (RVP-9) +a2* (T-75)+a3* (RVP-9) * (T-75))
C
C     IRVP        1-2/4  types of non-certification RVP:  1 » Base,  2 - In-use
C                                              3 =" Uncontrolled refueling loss
C                                              4 - Hot soak (weathered)
C
C     IS        1-4    overlap tampering effects category sizes 8-11
C                       ( IS -  1C - 7 ); used in A8TO11.
C
C     ISC       1-5    underscore string repetition counter - for example,
C                      ISC  - 3 -> repeat the string 3 times
C
C     ISP       1-21   my group pointer into speed correction equation array;
C                      found by ISPPTR; indexes SKI & SCUNA1.
C
C     IS2a      1-2/3  number  of coefficients in a new scf equation:
C                           IS2C - 2  (A/X+B);   IS2N - 3 (A+B*X+C*X**2)
C
C     IT        1-9    pointer to temperature correction  equation
C                      coefficients; found by ITCPTR;  indexes TTA/4/7COF
C                      arrays.
C
C     ITECH  1-15   emission control technology class  wrt I/M impact;
C                      f(MY,IVLD) aa determined by IMPTR; indexes I/M credits
C                      arrays  CR12HC/CO £ CRD4VA/B. ITECHA/B/C are in BD8S34.
C
C     ITEM      1-2    NAMTEM  index: 1 - 'mini', 2 - 'maxi'  prefixes
C
C     ITER      1-4    unspecified string iterations counter;  same usage as
C                      ISC, only with any string
C
C     ITL       1-13   running loss  HC: defined trip lengths for 3 cycle types:
C                           1-6 - HTCC      7-12 - IA-4      13 - HFET
C
C     ITLR      1-4    running loss  HC: RVPs at which measured rl HC available
C
C     ITLT      1-4    running loss  HC: temperatures at which measured rl HC
C                                        emission rates are available
C
C     ITY       1-3    type tampering rates wrt I/MxIGD table heading substring
C                      pointer
C
C     IUDI      1-5    index into Uncontrolled Diurnal rate UDI:
C                        1  - standard FTP conditions
C                        2  - in-use  conditions
C                        3  - 8AM-11AM conditions
C                        4  - 10AM-3PM conditions
C                        5  m 8AM-2PM conditions
C
C     IV        1-8    vehicle types evaluated by MOBILE4:
C                        1  - LDGV - light duty (Id)  gasoline fueled vehicle
C                        2  - LDGT1 - Id gasoline fueled truck,  <- 6000 Iba.
C                        3  = LD6T2 » Id gasoline fueled truck,  6001-8500 Ibs.
C                        4  =• HDGV => heavy duty gasoline  fueled vehicle
C                        5  » LDDV - light duty diesel fueled vehicle
C                        6  - LDDT - light duty diesel fueled truck
C                        7  - HDDV - heavy duty diesel fueled vehicle
C                        8  - MC   «• motorcycle

-------
                                            Appendix A
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
          1-6
       vehicle type subgroups; Index correction factor arrays;
        - LM, TEMP,  A, B or LDG, defined as:
IVEFF





IVF1






IVF2








IVF3



IVIM





IVHD





IVP





IVREV





IVTAM



TV9






12ND


JCS



JDX
1-2




1-6






1-9







1-10



1-4




1-2




1-3




1-4




1-4



1-9






1-4


1-7



25-1
IVLM
1 - LDGV
2 - LDGT1
3 - LDGT2
4 - LDDV
5 - LDDT
o *• MC
indexes :
MYGSP1
BFRCOF
IVTEMP
1
2
3
4
5


=
m
at
9
=.


LDGV
LDGT1
LDGT2
HDGV
MC


MYGTCF



IVA
1 - LDGV
2 - LDGT1
3 - LDGT2
4 - MC



SCADJ1
ISP GBP
IVB
1
2
3
4



- HDGV
- LDDV
- LDDT
- HDDV



GPBSCO


IVLDG
1
2
3




- LDGV
- LDGT1
= LDGT2




TTFC


JG
          1-20
                          ALHRET
See subroutines BIGCFX £ BIGALH and function ISPPTR.


vehicle class groupings for Stage II vrs percent
efficency in reducing HC refueling losses:
  1 - LDGV, LDGT1 £ LDGT2    2 = HDGV


vehicle class types for MOBILE1 numeric output
format (OUTFMT - 1):
  1 - LDV      3 - LDT2     5 • HDD
  2 - LDT1     4 - HDG      6 - MC


vehicle class types for MOBILES numeric output
format (OUTFMT - 2):
  1 - LDGV     4 - LDGT     7 - LDDT
  2 - LDGT1    5 - HDGV     8 - HDDV
  3 - LDGT2    6 - LDDV     9 - MC


vehicle class types for MOBILE4 descriptive output
format (OUTFMT - 3): same as IVF2 plus 10 = All Vehicles


vehicle class types that can be included in MOBILE4
inspection / maintenance (I/M) programs:
  1 - LDGV    2 - LDGT1    3 = LDGT2    4 - HDGV


vehicle class types for heavy duty only (used for
units conversion of user entered hd zml s dr):
  1 - HDGV     2 - HDDV


vehicle types whose ef's can be modified by correction
factors for extra load, trailer towing S/or CO offset:
  1 - LDGV     2 - LDGT1    3 - LDGT2


vehicle types covered by the revised idle EF algorithm:
  1 - LDGV     2 - LDGT1    3 - LDGT2    4 - HDGV  (HCSCO)
IVI - IVREV, except 3 & 4 -> 3 (HDGV uses LDGT2 data)


vehicle types impacted by tampering:
  1 - LDGV     2 - LDGT1    3 - LDGT2    4 - HDGV


vehicle type categories for output arrays EFEVAP,
EFIDLE, S EFFTP-  First 8 are identical to TV'a.
  9 - LDGT
    • LDGT1 + LDGT2, each weighted by its vmt share


secondary index  (pointer into a pointer vector)


pointer to category size to be summed into DIV
(in A8TO11) or  (* TGSUSE) into EGS  (in SETEGS)


model year window index - age of vehicle relative to
ICY + 1.  In detail:
 25 » 25+ my age group cell •> tamper effect is
      always 0.0 -> this case skipped in TAMPER  (only)
 24 - 24th my cell
 23 - 23rd my cell
         2 » 2nd my cell
         1 - 1st my cell = calendar year (ICY)


       model year group pointer - IG or 21 - IG, depending on
       whether looking for match or unused cell, respectively,
       in MYGERB.  Also used (as is KG) to distinguish  (from IG)
       myg dimension size differences in BD comments.
                                                  10

-------
                                             Appendix A
c
c
c
G
C
C
C
C
C
C
C
C
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
C
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
      JOU
   JPGD
   JSTD
1-3






1,5



1-6
JO
JV
KSTD
KDX
LAY
LIM
1-9
1-2
1-2
1-21
1-2
1-2
output unit key:
  call I is assigned IOUREP's value
  cell 2 is assigned lOUERR's value
  cell 3 is assigned lOUASK's value


pointer to base tamper rate to be used in
computing CSIZE(1-3,IAY).  Depends on IPG.


emissions standard model year group for gas fueled IV.
The group key is the same as ZSTD, but JSTD is set
internally, instead of being passed in via EVMAIN.
(The ISTD definition is in the Parameter Dictionary.)


input / output  (i/o) device unit numbers


vapor pressure calculation loop S vectors index  (CALUDI)


mapping of JSTD used to index R2ALT:
    KSTD - 1, unless JSTD = 4 -> KSTD = 2


model year window index - ICY - 1979, unless result is
< 1 (-> - 1) or > 21 ( -> - 21); used by HDDMYM


same as IAY, but set for DISCAL call only.


- IM, except - 1 when there is an I/M kink, but it is
  after LDXSY & the PREV segment is being evaluated:
                    ZM   IMKINK
                                  LKINK
                                          LIM
                                                Comments
1
2
2
2
1
1
2
2
1
1
1
2
I
2
1
2
P S S: no I/M
P S Si all I/M
P: no I/M; S: kinked I/M
P: kinked I/M; S: all I/M
   LIMIT



   LINES


   LJDX



   LKINK


   NCKIAY
1-2



1-9


25-1



1-2


1-3
index for range check on a variable's value assignment:
1 " lower bound    2 •» upper bound


number of lines in user entered data set title


model year relative to LAPSY - JDX - LDXSY -I- 1;
used with CUMMIL to get mileage of "previous" segment


same as IMKINK, except IMDXSY must be within (JDX,LDXSY)


index into ECK - ICOL index
5) Subroutine / function parameter dictionary


   Variables /arrays occurring in 1 or more of the subprogram parameter
   lists are defined in this section, unless the variable is a subscript,
   in which case look it up in 4) 's subscript dictionary.  The motivation
   for defining all the parameter lists in one place is to have only one
   instance of a definition in the code plus to encourage only one use
   for a mnemonic in the code plus the combined lists contain only 18
   names after subscripts have been removed.  Whether scanning on-line or
   off a listing,  the lookup will be quick.


   Parameter array subscripts


   CSF0SE(7)         -  CSFOSE ( IC7 )
   CSF1ST(7,2,3)     -  CSF1ST ( IC7, IPG, IGCSF )
   S2LEFT(4)         -  S2LEFT ( IVTAM )
    Name
           Type              Description
   CSFOSE   R    category size factors used * AIR to calculate CSIZE
   CSF1ST   R    category size factor for 1st calculation of CSIZE
   FLGERR   I    number of flag values found to be out of range
   IBEFSW    0-2    switch identifying routine calling BEF S result returned
                      0 =• BEFIDL => always return uncorrected basic non-idle ef
                      1 - BEFIDL m return uncorrected basic ef
                      2 « EFCALX or HCCALX - return basic ef corrected for
                          operating mode, temperature, RVP (if high temp),
                          I/M, oxy fuel and tampering
   ICASE    I    flags the my groups array to be searched.
   ICY      I    calendar year for which emission factors are calculated
   INERR    I    cumulative count of data entry errors
   IOUOUT   I    unit number on which title is printed
                                                  11

-------
                        Appendix  A
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
IOUSUM
IVALUE
IY
IY1
IY2
MESSAG
MY
HERR
NREC
RVALUE
S2LEFT
VMLDGT
VMTAGE

Evap Study

Dame
EVRATE






ISTD




























ITRSTL




IVGAS



IVTL

KEYEQ




KEYEVP


RIVAL





I sum of acceptable changed i/o unit numbers
I entered integer value causing error / warning message
I year to test and adjust to 4-digit year
I lower 4-digit bound for IY
X upper 4-digit bound for IY
I message code sent by calling subprogram
I model year
I number of error if IY is out of bounds
X number of records to be skipped on next READ
R entered real value causing error / warning message
R See local dictionary definition in REFUEL header.
R sum of vmt for LDGT1 and LDGT2 (- X23 in MOBILE2)
R vehicle cumulative mileage

XI Parameter List Dictionary for CCEVRT block of routines:

Range Description
R dependent variable value from evaporative component
equation (s) calculation:

KEYEVP EVRATE
1 — > hot soak rate
2 -> diurnal rate

1-7 emissions standard model year group for gas fueled IV:
IVGAS value myg
1 1 pre-1971
2 1971
3 1972-1977
4 1978-1980
5 1981+ carbureted (KEYEQ-1 or 2)
6 1981+ TBI KEYEQ-1 or 2
7 1981+ PFI KEYEQ-1 or 2
2 1 pre-1971
2 1971
3 1972-1977
4 1978-1980
5 1981+ carbureted (KEYEQ-1 or 2)
6 1981+ TBI KEYEQ-1 or 2
7 1981+ PFI KEYEQ-1 or 2
3 1 pre-1979
2 1979-1980
3 1981+ carbureted (KEYEQ-1 or 2)
4 1981+ TBI KEYEQ-1 or 2
5 1981+ PFI KEYEQ-1 or 2
4 1 pre-1985
2 1985+ (KEYEQ-1 or 2)
5 1 pre-1978
2 1978-1979
3 1980-81
4 1982-84
5 1985+

1-2 technology type for resting losses calculation
1— fuel injected with open bottom canister
2-fuel injected with closed canister
(currently carbureted-fuel injected, open)

1-5 gasoline fueled vehicle types (— IVTEMP) :
1 - LDGV 2 - LDGT1 3 - LDGT2 4 - HDGV 5 - MC
IVGAS - IV, except MC (5 vs 8); IVEVAP is an alt name.

1-4 running loss HC: vehicle classes with rl HC emissions

1-3 HS/DU equation group selection key:
1 — passing
2 — failed purge
3 — failed pressure

1-2 Evaporative component calculation selection key
1 = hot soak 2 - diurnal

R independent variable value in evaporative component
equation (a) calculation:
KEYEVP RIVAL
1 -> hot soak -> fuel RVP
2 — > diurnal — > diurnal Index

***********************************************************
                              12

-------
                                            Appendix A

c
C     1C, IDX, IFDS, IG, IM, IP, IP1, IP2, IR, IV and JSTD are parameters
C     defined in the subscript dictionary.
C
C  Notes:
C
C  **** HOT IN USE **** - the subroutine/function is not  called by the current
C  version of MOBILE4.   Release MOBILE4 (final version),  for example,  does not
C  use WTLROL or OUTTAM.
C
C
                                                  13

-------
                                            Appendix A

c
c  MAIN
c
C  Dummy main program for MOBILE4.1.
C    Required to allow proper RUN entry to MOBILE subroutine.
C
C  Called by:  none
C
C  Calls MOBILE
C
C  The following variables are passed in argument lists or common blocks
C
C    argument lists: none
C    common blocks:
C    /SAVE01/ JCALL
C
C  Output on return:
C
C    parameter list: none
C
C  Local variable dictionary:  none
C
C  Reset JCALL (initialized to 1 by BD12)  to indicate stand-alone
C
C     COMMON /SAVE01/ JCALL,VA001(15,3,8,2),VA002(15,3,8,2)
C     JCALL-0
C     CALL MOBILE(INERR, *99)
C     STOP
C  99 STOP 1
C     END
C
C *****************************************************************
C
                                                  14

-------
                                             Appendix A

      SUBROUTINE MOBILE(INERR,  *)
C
C  Called by user's  program (example: Air Quality Analysis System).
C
C  Calls BDSAVE, CONSEC,  EFCALX,  ONESEC, OUTDT5, OUTNEW, OUTPUT,
C        PARSEC, QUITER,  REINIT,  RESTOR and REGMOD.
C
C  The following variables are passed in argument lists or common blocks  and
C  in MOBILE:
C
C    argument  lists: ICY,INERR
C    common blocks:
C    /FLAQS3/  OUTFMT
C    /IOUCOM/  IOUERR
C    /REGION/  IREJN
C    /SAVE01/  JCALL
C
C  Output on return:
C
C    parameter list: INERR
C
C  Local variable  dictionary:
C
C   Name   Type              Description

C  ICYOLD   I    calendar year from previous pass of scenario loop
C  IROLD    I    region from previous pass of scenario loop
C  ISAV93   I    flag for 1993 disclaimer message  (QUITER #130)
C                       0 « message has not yet been given
C                       1 — message has been given: not again!
C  LEAST1   I    flag:  0 - OUTPUT has not yet been called.
C                       1 - OUTPUT has been called at least 1 time.
C
C  Notes:
C
C  MAIN was converted to subroutine MOBILE in Version 01 of development MOBILE4.1
C  The /BYMYCd/ CBs  were added for Version 02.
C
C
      CHARACTER*26 BYVTN
      CHARACTER*!? YFM
      CHARACTER* 8  NAMFLG, NAMMMR
      CHARACTER* 8  USNAME
      CHARACTER*?  PN,U7
      CHARACTER* 8  NAMRVP
      CHARACTER* 4  NAMTEM
      CHARACTER* 4  NO YES, COMMA, PERIOD
      CHARACTER* 4  IMNAME, PRO JID , SCNAME
      CHARACTER*3  PSEC
      CHARACTER*!  COLON, VB
      INTEGER  PROMPT, TAMFLG, SPDFLG, VMFLAG, OXYFLG, DSFLAG
      INTEGER  ALHFLG
      INTEGER  ATPFLG, TPDFLG, RLFLAG, TEMFLG, OUTFMT, OUTPRF
      INTEGER  PRTFLG, HCFLAG, COLDFG
      INTEGER  ATPPGM,ATPFQT,DISTYP
      INTEGER  ADDCSW
      REAL JULMYR,JANMYR
      REAL LHDVM, MHDVM, HHDVM, LHDRG, MHDRG, HHDRG
C
      COMMON /ALURAC/ ACEQIP(8,3,2),ACCF(2,3,2,3,2)
      COMMON /ALUHIN/ AC,XLOAD (3) , TRAILR(3) ,ABSHUM, DB,HB
      COMMON /ALUHLT/ XLCF(3,6,3,2),TMCF(3,6,3,2)
      COMMON /ALUOUT/ ALHRET(25,4,3)
      COMMON /ALUPAC/ MYGACE (8, 3, 2) ,MYGACC (2, 3, 2)
      COMMON /ALUPLT/ MYGXLCJ6,3,2),MYGTWC(6,3,2)
      COMMON /ATPAR1/ LAPSY,LAPlST,LAPLST,LVTFLG(4)
      COMMON /ATPAR2/ ATPPGM,ATPFQT, CRATP, DISTYP (8)
      COMMON /ATPAR3/ AIR(2) ,CAT (2) ,ECK (3) , INK (2) , EGR(2)
      COMMON /ATPAR4/ CAN(2),PCV(2),CAP(2),EFF(2, 3,15)
      COMMON /BASEQl/ ERBZML(23,3,8,2),ERBDR<23,3,8,2),ERB50K(12,2, 3, 2)
      COMMON /BASEQ2 / MYGERB (23,3,8,2), MAXERB
      COMMON /BASEQ3/ NEWVEH, NEMMYF, NEWMYL, ZMLNEW, DRNEW, KINKS
      COMMON /BASEQ4/NEWPAR(5,100) ,BERNEW (3,100) ,NEWFIT (100) ,NEWCT,MAXCT
      COMMON /BASEQ5/ ERUZML(12,3,8,2),ERUDR(12,3,8,2),ERU50K(12, 2, 3, 2)
      COMMON /BASEQ6/ MYGERU (12, 2, 3, 8, 2) , MAXERU, NUMERU (3, 8, 2) ,KEYER, IGER
      COMMON /BASEQ7/ IPMOD(3)
      COMMON /BASEQ8/ IHDVMT(23,2),HDVZML(23,2) ,HDVBDR(23, 2)
      COMMON /BYMYC1/ IVPICK (8) , IMPICK, IDPICK,BYCUM (25, 8)
      COMMON /BYMYC2/ BYBEF4(3,25,8),BYTAM(3,25,4),BYFER(3,25,8)
      COMMON /BYMYC3/BYEVAP(25,8),BYRUNL(25,4),BYREFL(25,4),BYRSTL(25, 8)
      COMMON /BYMYC4/ BYVTN (8) , YFM,PN (10) , U7,PSEC (3) , VB


                                                   15

-------
                                        Appendix A

COMMON /BYMYC5/ BYIMCR(2, 25, 4) ,BYIMRE (2, 25, 4) ,REDSUM<2, 4)
COMMON /CEVCOM/ USRTPD(8),USRMPD(8)
COMMON /CEVBMY/ BMYMPD (25, 8) ,BMYTPD (25, 8) ,BMTPDF (25, 8)
COMMON /CITCIN/ UDI(5),IUDI,PFRATE(3)
COMMON /CITPAR/ SCNAME(4)
COMMON /CITRV1/ RVPBAS,RVPIUS,RVPICY, I0SESY,RVPUWX
COMMON /CITRV2/ RVPHS,RVPX,RVP090,RVP115
COMMON /CRACOM/ MAXCRA, CRAVAL(3, 8, 2) ,MYGCRA<3, 8, 2)
COMMON /CUMCOM/ CUMMIL(25,8)
COMMON /EGSCAL/ AER(11,11,2,2),TGS(11,6,4),EGS(7,2)
COMMON /EMECAL/ EM3W(3,3, 3) ,EMOX (3,3, 3) ,EMEGR<4, 3,3) ,EMI (7, 3, 3)
COMMON /EVADC1/ DUEQ(3,2,3) ,EFDU(3,2,5,5)
COMMON /EVADC2/ HIDO(3,2),R2D0(2,2)
COMMON /EVAHIA/ HIHS(3,2),HIADJ
COMMON /EVAHS1/ HSEQ<3,2,9),EFHS(3,2,5,5)
COMMON /EVAPAR/ IFDS, I3TDC, ISTDT,ISTDP,FINJ(2)
COMMON /EVAPGR/ EVP (4) , GREVP (4, 9) ,VGREVP (4) ,PARD (3) ,GRPD (3, 9)
COMMON /EVAEHD/ HDSAL, HDWGT (2) , IWMAP (5)
COMMON /EVASTD/ MYGSTD(5,S),MAXSTD,KSTD81(4,3)
COMMON /EVPDAT/ EVPAERJ3,2),EVPTGS(3,2,4),CCEMI(7,4)
COMMON /FID/  MAXFID,MYFID(16,5),TOGCF(16,5,3),VOCCF(16,5,3)
COMMON /FLAGS1/ PROMPT, TAMFLG, SPDFLG,VMFLAG,OXYFLG,DSFLAG
COMMON /FLAGS2/ MYMRFG,NEWFLG, IMFLAG, ALHFLG
COMMON /FLAGS3/ ATPFLG, TPDFLG,RLFLAG,LOCFLG, TEMFLG, OOTFMT
COMMON /FLAGS4/ PRTFLG, IDLFLG, NMHFLG,HCFLAG,COLDFG
COMMON /GSFCOM/ MAXGSF,GSFRAC(22,8,2),MYGSF(22,8,2)
COMMON /HDCCOM/ HDCFAC (37,2), LOWYR, MAXCIG
COMMON /HDDMAR/ C2BDVM (25) , LHDVM (25) , MHDVM (25) , HHDVM (25)
COMMON /HDDREG/ C2BDRG(21),LHDRG(21),MHDRG(21),HHDRG(21)
COMMON /IDLEQ1/ ZMLIDL(15,3,8,2),DR1DL(15,3,8,2)
COMMON /IDLEQ2/ MYGIDL(15,3,8,2),MAXIDL
COMMON /IDIBQ3/ UTIDOF(14,3,2)
COMMON /ICR4VA/ CRD4VA(24, 2,12,3)
COMMON /ICR4VB/ CRD4VB(24,2,12,3)
COMMON /IMPAR1/ ICYIM, ISTRIN,MODYR1,MODYR2, WAIVER(2) ,CRIM
COMMON /IMPAR2/ ILDT (4) , ITEST, NUDATA (2) , NLIM, IMNAME (20, 9)
COMMON /IMPAR3/ IMDX1, IMDX2, IMCASE, IMVEH, IM
COMMON /IMPAR5/ MYGIM2 (3,3) , LBIM4P,CRHDGV(2, 2) ,DISCNT (3)
COMMON /IMPAR6/ IFREQ,INTYP
COMMON /IM12HC/ CR12HC(19,20,5, 2)
COMMON /IM12CO/ CR12CO(19,20,5,2)
COMMON /IM240P/ DSIZE (25, 4) , IM24YR, IPRGYR, IPRSYR
COMMON /INJECT/ TBI (13, 4) ,PFI (13, 4) ,MAXFIY,MINFIY
COMMON /IOUCOM/ IOUIMD, IOUGEN, IOUREP, IOUERR, IOUASK
COMMON /IVPCOM/ IVPTRL(8) , IVPTRT (8) , IVPTRA(8) , IVPTRB(8)
COMMON /LOOKUP/ IVTAM, IQG, IPG, JPGD, IHG, IGCSF
COMMON /MAXIMA/ MAXVEH, MAXLTW, MAXPOL, MAXREG,MAXYRS
COMMON /MPDCOM/ MAXMPD , VALMPD (2,8), MYGMPD (2,8)
COMMON /MYCODE/ MY,IDX, JDX,LDXSY,LMYRVT, XAY, IMDXSY, IMKINK
COMMON /MYRCAL/ XMYM(25,8),JANMYR(2S,8),TF(25,8),TFMYM(25,8)
COMMON /MYRSAV/ AMAR (25, 8) , JULMYR (25, 8) , NEWCUM
COMMON /MYOB1/  MYTGS (11, 6, 4) ,MYEGR(4, 4) ,MYVTS (3, 2, 4)
COMMON /MYUB2/  MYCCEI(7,4)
COMMON /NAMESl/ NAMFLG (19) , NAMMMR (2, 3, 8) , NAMRVP (4) , NAMTEM (2)
COMMON /NMETH1/ ZDMTH1(2,3,2, 7,3,2),ZDMTH2(2,7,2,2),ZDMTH3(7,3,2)
COMMON /NMETH2/ MYGMTH(7,8,2),MTHMYG
COMMON /OFFSET/ OFFCO(25,3),OFFMTH(25,8)
COMMON /OMTCOM/ OMTCF(25,3,8),OMTTAM(25,3,4),OMTCFF(25,3,8)
COMMON /OPMOD1/ BFRCOF(8,8,3,6),BFSTEP(8,12,3,3),BFR50K(4,12,3)
COMMON /OPMOD2/ MY3TEP,MAXBFR,MYGBFR(8, 3, 6)
COMMON /OPMOD3/ BFMILE,KINKBF,BFGT50
COMMON /OXY1/ SHRMKT(3),OXYCNt(2),IGASHW
COMMON /OXY2/ NFUEL, MXOXMY, MAXOXY, OXYVAL (25,12), OXYBF (25,12) , OXYCF
COMMON /PROJEC/ PROJID(20)
COMMON /QUITXQ/ N1QUIT
COMMON /REGION/ FEET(2),IREJN,ALT,INITPR
COMMON /REGISF/ GSFVCT(8),USRGSF(25,2)
COMMON /RESTJL1/ EFFTP (3, 9) ,EFEXH (9) ,EFEVAP (9) ,EFREFL (9) ,EFRUNL (9)
COMMON /RESUL2/ RLGGAL (9) ,RSTGPH (9) , EFRSTL (9) ,EFIDLE (3, 9)
COMMON /RESUL3/ VFTP (3) , VEXH, VEVAP, VLOSS, VRUNLS, VRSTLS, VIDLE (3)
COMMON /RLCOMl/ ROADFE(32,4),DISPL,SPILL,OBED,OBES,OBDF(4)
COMMON /RLCOM2/ IOBMY,IVOB(4),IS2SY,NPHASE,S2EFF(4)
COMMON /RLCOM3/ RLRATE(25,4)
COMMON /RONLS1/ TLEMI (6, 3, 4, 4, 4, 4) , TLVMT (6) , TLVMTU (6) ,MAXTL
COMMON /RONLS2/ TLFAIL(6,4,4)
COMMON /RONLS3/ MYGRUL (4, 4) ,RULRVP (4) ,ROLTEM (4) ,RULSFD (3) ,MAXROL
COMMON /RESTLS/ PTGRST(13,5,2),RESTA(2),RESTS
COMMON /RVPEX1/ R9G1(3,2),RG23(3,3,3,2),ILL(2)
COMMON /RVPEX2/ MYGRVP(4,4),RVPCF(25,3,4)
COMMON /RVPNAT/ RVPLIM(2),RVPDIU(2),PFUL
COMMON /SCENE1/ SPD(8),PCCN,PCHC,PCCC
                                              16

-------
                                             Appendix A

      COMMON /SCENE2/ FCC,FCN,FHC,FHN
      COMMON /SPEED1/ SKI(6,18,3),SK2C(2,14,3,3),SK4C(2,14,3,3)
      COMMON /SPEED2/ MYGSP1 (18,2,3) ,MYGSMC (18, 2,3) ,MYGSP2 (3,3)
      COMMON /SPEEDS/ SADJ,SCUNA1(18,3,8),SCADJl(3,4),ISPGRP(3,4)
      COMMON /SPEED4/ MAXSP1,LB1STS(3),LBLAST,IVB,IVA,SCUNAO(18,3, 8,2)
      COMMON /SPEEDS/ MYHSPC(2,3),MAXHSP(3)
      COMMON /SPEEDS/ SALHCF(25,3,8),SCFIDL(25,4),SCIADJ(25,3,8)
      COMMON /SPEED7/ GPBSCO(3,3,4)
      COMMON /SIZCAL/ ATR(5,2),TGSUSE,CSIZE(12,2),CSAE(12)
      COMMON /SIZPAR/ NTERMS(4),INDXCS(5,4)
      COMMON /STRING/ NOYES(2),COMMA,PERIOD, COLON
      COMMON /TAMEQ1/ TAMZML(9, 4,3, 2) ,TAMDR<9, 4,3,2) ,TAMA50 (9, 4,3, 2)
      COMMON /TAMEQ2/ MYGTAM(3,4),FSOK
      COMMON /TAMEQ3/ COMCAP,CUMALL,caMNIM,CUMIM,QT50KA,GT50KN,GT50KI
      COMMON /TAMEQ4/ BTR(9,2)
      COMMON /TAMID1/ EMI3W(3,3,3),EMIOX(3,3,3),EMIDL(3,4,3)
      COMMON /TAMID2/ MYIDLE(3,3),IEK
      COMMON /TAMID3/ OFFIDL(3,25,4)
      COMMON /TAMOUT/ TAMBAG (3,3, 25, 4), THS (2,25, 4) , TDU (25, 4), TCC (25, 4)
      COMMON /TAMOO2/ TOB(25,4),GCONLY(25,4)
      COMMON /TAMPB1/ LPOD,PBBTR(3,2,25, 4)
      COMMON /TEMPC1/TTFC (2, 3, 9, 3) , TTFD (3, 3, 3), TT4 (2, 6,3) , TT8 (2,3, 4,3)
      COMMON /TEMPC2/ MAXTCF,MYGTCF(9,3,5),MYGTFD(3),ISHIFT
      COMMON /TEMPC3/ MDLOHI (3) , TDIFF (3) ,CHFRAC (3) , TCF (3)
      COMMON /TEMPC4/ MYGCOO(3) ,LOWCO,COOTFC(3) ,COOTFD (3)
      COMMON /TEMPC5/ NCC,MYCOL(2),CCSTD(3,2)
      COMMON /TEMPC6/ CSUB (3,2), CM0L(3, 2) ,PIVPCT (3,3, 2)
      COMMON /TEMPC7/ ICOLD,ADDCSW,MULCSW
      COMMON /TEMPS/ AMBT, TEMMIN, TEMMAX, TEMEXH (3) , TEMEVP (6)
      COMMON /TPDCOM/ TPDCO(2,8)
      COMMON /USDATA/ USNAME(4,4),NUSD,IUSD(4)
      COMMON /VMXCOM/ REGMIX(8),TFNORM(8),VMTMIX(8), VCOUNT(39,8)
      COMMON /YEARS4/ IY1941,IY1960,IY2020
C
      COMMON /SAVE01/ JCALL,VA001(15,3,8,2),VA002(15,3, 8, 2)
      COMMON /SAVE02/ VA003(24,2,12,3),VA004(24,2,12,3)
      COMMON /SAVE03/ VA005 (19, 20,5,2) ,VA006 (19,20,5,2)
      COMMON /SAVE04/ VA007(25,8),VA008(25, 8)
      COMMON /SAVE05/ VA009(9,4,3,2),VA010(9, 4,3,2)
      COMMON /SAVE06/ VA011(8)
C
C
C  The GEM Post-Processor  COMMON BLOCKS
C
      COMMON /COST02/ IFEVYR, ILEVYR
      COMMON /COST07/ CSTIM,GASCST,IMSTRT,ISCEN,MOVIM,MOVATP,REPIM(4)
      COMMON /COST27/ IGRFLG, IATPLP, IDIFF,OUTPRF
C  Add JPRGYR  & JPRSYR, 5/29/91
      COMMON /COST28/ JATP, JIM, ISTRT, JCYDX, JPRGYR, OPRSYR
C
      COMMON /COSTXX/ JMYR1,JMYR2
C
      COMMON /ZONAME/ M4IN,M4OUT,M4ZMC,C4IN
C
C      CHARACTER*20 M4IN,M40UT
C
C      CALL GETCL(M4IN)
C
      ISAV93-0
      IF(JCALL.EQ.O) GOTO  10
C
C  Subroutine  call, aetup  run.
C
      CALL REINIT
C
      IF(JCALL.EQ.l) CALL  BDSAVE
C
      IF(JCALL.EQ.2) CALL  RESTOR
      JCALL-2
C
C...  Set for pc file i/o
C
C      OPEN(ONIT-5,FILE-'M4INPOT',STATUS-'OLD')
CC     OPEN (UNIT-5, FILE-M4IN, STATUS-' OLD ')
C      OPEN (UNIT-6, FILE-'M4OUTPUT' , STATUS- 'NEW ,
C     *  CARRIAGE CONTROL-"FORTRAN")
C
C  Initialize non-COMMON variables (INERR t local variables).
C
   10 INERR-0
      ICYOLD—1


                                                   17

-------
                                             Appendix A

      IROLD=»-1
      PRECY=0
      LEAST1=0
C
C  Control and One-Time Data Sections:
C
C  Get the execution control flags  and  parameters,  and then use them to read
C  in data, if any, that  is read  in once  and applied to all scenarios.  Note:
C  alternate RETURN1 is the fatal read  error exit:  execution stops.
C
      CALL CONSEC(*99,*999)
C
C  Set up for an ATP only run.
C
      IATPLP - 1
      IF(IMFLAG.EQ.l .AND. ATPFLG.EQ.2) IATPLP  - 2
C
      CALL ONESEC(INERR,*99)
C
C  Store original input settings
C   (ISTRT=-IMSTRT set in ONESEC)
C
      ISTRT- IMSTRT
      JIM  - IMFLAG
      JATP - AIPFLG
      JERGYR - IPRGYR
      JPRSYR - IPRSYR
      JMYR1 - MODYR1
      JMYR2 - MODYR2
C
      GOTO 15
C
C  Enter scenario loop.
C
      ENTRY MOBNXT(INERR,  *)
C
   15 CALL PPINIT
C
C
C  Note that if INERR > 0 coming  out of CONSEC/ONESEC,  then the scenarios will
C  be read and processed,  but EFCALX/OUTPOT  will not be called.  The user will
C  only get error diagnostics out of the  run.
C
C  20 CALL PARSEC (ICY, INERR, *98)
   20 CALL PARSEC (ICY, INERR, *999)
C
C  Display status on screen...
C
      WRITE (0,100) ICY,PROJID
  100 FORMATC Start CEM4.1 Run for Evaluation  Year:  ',I4,/,
     *  1X,20A4 )
C
      IF(ICY.LT.ICYIM .AND. ICY.LT.LAPSY) GOTO  20
C
C  Removed old CALL CHKSEN...
C
      IFEVYR " ICY
C
C  later add direct input of f of consec  yrs (IDIFF)  to ilevyr
C
      ILEVYR - IFEVYR + IDIFF - 1
C     IDIFF - ILEVYR - IFEVYR + 1
C
C... later can change 1 to some com input number of consecutive yrs if want.
C
   22 JCYDX - ICY - IFEVYR + 1
C     WRITE (8, 102) JCYDX, ICY, IFEVYR, ILEVYR
  102 FORMAT(IX,'jcydx,icy,ifevyr,ilevyr:',416)
C
C  Enter program scenario loop.
C
      DO 25 JSCEN-1,4
C
         ISCEN - 4 - JSCEN + 1
C
C  If ATP-Only case, skip I/M MOBILE4 runs
C
         IF(IATPLP.EQ.2 .AND. JSCEN.GE.3) GOTO  26
C
         CALL RESET
C


                                                  18

-------
                                             Appendix A

C  50+ errors is  (arbritrary)  run limit  •»> atop execution.   Otherwise,
C  one or more input processing errors -> abort current scenario S then do
C  input processing for error  diagnostics on the remaining scenarios.  INERR
C  is never reset to 0, but  instead accumulates until the run ends or the
C  cutoff value  (50) is reached.
C
      IF(INERR.GT.50) GOTO 30
C     IF(INERR.GT.O) GOTO 20
      IF(INERR.GT.O) GOTO 30
C
      IF(ICYOLD.NE.ICY.OR.IROLD.NE.IREJN) CALL REGMOD (ICY, INERR)
      IF(INERR.GT.50) GOTO 30
C     IF(INERR.GT.O) GOTO 20
      IF(INERR.GT.O) GOTO 30
      ICYOLD-ICY
      IROLD-IREJN
C
C   Warning 130  - for ICY'a  1993+ MOBILE4.1's disclaimer
C
      IF(ISAV93.EQ.O.AND.ICY.GE.1993)  CALL QUITER(0.,0,120,INERR)
      IF(ICY.GE.1993) ISAV93-1
C
      CALL EFCALX (ICY, INERR)
C. . .
C     IF(INERR.GT.50) GOTO 30
C     IF(INERR.GT.O) GOTO 20
      IF(INERR.GT.O) GOTO 30
C. . .
      CALL EFSAVE(ISCEN, JCYDX)
      CALL SAVER (ICY)
      IF(OUTPRF.EQ.l) GOTO 25
      IF(OUTPRF.EQ.2  .AND. JCYDX. LE. IDIFF .AND.
      *  JSCEN.EQ.l) CALL OUTPUT (ICY)
      IF(OUTPRF.EQ.3  .AND. JCYDX.LE. IDIFF .AND.
      *  (JSCEN.EQ.l.OR.JSCEH.EQ.4))  CALL OUTPUT(ICY)
      IFJOUTPRF.EQ.4  .AND. JCYDX. LE. IDIFF) CALL OUTPUT(ICY)
      IF(OUTPRF.EQ.S) CALL OUTPUT(ICY)
C
    25 CONTINUE
C
      LEAST1-1
C     GOTO 20
      GOTO 28
    26 CALL ATPSAV(ICY)
C
    28 IF(JCYDX.LE.IDIFF) THEN
        ICY - ICY+1
        GOTO 22
      ENDIF
      GOTO 99
C
C     GOTO 20
C
    30 CALL QUITER(0.0,0,28,INERR)
C
C   If user is supplying replacement basic FTP e£ parameter records,  always
C   print the records contents  and disposition table,  even when input errors
C   have prevented the normal output routines from being called.   The only
C   exception is  when a  READ  error / end-of-file occurs in CONSEC or ONESEC.
C
C   Thia section  can be  expanded to print other optional input.
C
    98 IF(LEAST1.EQ.O.AND.NEWFLG.EQ.2)  CALL OUTNEW
C
    99 RETURN
C
C   End-of-file on 1st read of  call -> input for call  missing entirely.
C   Calling program can  branch  to attach  a new input file or terminate
C   calls to MOBILE.
C
   999 RETURN 1
C   End MOBILE
      END
                                                   19

-------
                                             Appendix A

      SUBROUTINE TAMPER(ICY)
C
C  TAMPER computes tampering effects  in  grama/mile on bag 1,  2 & 3 HC, CO,
C  & NOX and on hotsoak, diurnal  S  crankcase HC.
C
C  Called by EFCALX.
C
C  Calls BAGEME, DISATP, EFFGRP,  EMIRAT,  EVPEME,  FINDPB,  and IMCHEK.
C
C  Input on call:
C
C    parameter list: ICY
C    common blocks:
C    /ATPAR1/ LAPSY,LAP1ST,LAPLST,LVTFLG
C    /FLAGS4/ PRTFLG
C    /IMPAR1/ ICYIM
C    /IMPAR3/ IM
C    /MAXIMA/ MAXYRS
C    /TAMPB1/ LPOD
C
C  Output on return:
C
C    common blocks:
C    /LOOKUP/ IVTAM
C    /MYCODE/ MY, IDX, JDX, LDXSY, LMYRVT, IMDXSY, IMKINK
C
C  Local variable / array dictionary:
C
C   Name   Type              Description
C  ™~~™™»  —«»«  V —>««««*««MMBWM — «>«_~~~M.—««••««.•_« ~~ •• — ••••••.™™™™.»™™»™™™.«™
C  LDX1ST   I    JDX order pointer  to first  model year covered by ATP
C  LDXLST   I    JDX order pointer  to last model  year covered by ATP
C  LMYR     I    flag relating model  year to ATP  parameters
C                  1 •• model year not covered by  an ATP
C                  2 = JDX is covered by an  ATP  (but  IVTAM may not be,
C                      hence, check for  inclusion to  set  LMYRVT)
C
C  Notes:
C
C  TAMPER was changed for MOBILE4.1v6 to call FINDPB.
C
C...  ATPFLG.EQ.l check added, and  ATPFLG declared aa INT.
C. . .
      INTEGER PRTFLG, HCFLAG, COLDFG, ATPFLG
C
      COMMON /ATPAR1/ LAPSY, LAP1ST, LAPLST, LVTFLG (4)
C. . .
      COMMON /FLAGS3/ ATPFLG, TPDFLG,RLFLAG, LOCFLG, TEMFLG, OUTFMT
C
      COMMON /FLAGS4 / PRTFLG, IDLFLG, NMHFLG, HCFLAG, COLDFG
      COMMON /IMPAR1/ ICYIM, ISTRIN,MODYR1,MODYR2, WAIVER (2) ,CRIM
      COMMON /IMPAR3/ IMDX1, IMDX2, INCASE, IMVEH, IM
      COMMON /LOOKUP/ IVTAM,IQG,IPG,JPGD,IHG,IGCSF
      COMMON /MAXIMA/ MAXVEH,MAXLTW,MAXPOL,MAXREG, MAXYRS
      COMMON /MYCODE/ MY, IDX, JDX, LDXSY, LMYRVT, IAY, IMDXSY, IMKINK
      COMMON /TAMPB1/ LPOD,PBBTR(3,2,25,4)
C. . .
      COMMON /COST07/ CSTIM, GASCST, IMSTRT, ISCEN,MOVIM,MOVATP,REPIM(4)
C
C  Relate start year of anti-tampering program  (ATP)  and  1st  and last model
C  years covered to calendar year.  Put  them in JDX order.  For example:
C
C      ICY - LAPSY -> LDXSY <* 1
C      ICY - LAPSY + 1 -> LDXSY - 2
C
C
C      ICY - LAPSY + 24 -> LDXSY  -  25
C
C  JDX order is used because these  values are used to access  the JDX ordered
C  CUMMIL.  Note that LDXSY <- 0  => ICY  < LAPSY => no ATP was in effect during
C  any part of the calendar year's  25 model  year  window.
C
C  The ATP has to be in effect at least  1 year prior  to ICY for it to affect
C  ICY's emissions.  The mileage  accumulated through  LAPSY all goes to the no
C  program case.
C
C  Only calculate these values once per  scenario  => compute them here before
C  the tamper loops begin.
C
      LDXSY-ICY-LAPSY+1
      LDXIST-ICY-LAPIST+I


                                                   20

-------
                                             Appendix A

      LDXLST-ICY-LAPLST+l
C
C  Decide whether to use "Hon-I/M"  (IM-1)  or "I/M"  (IM-2)  TAMZML and TAMDR:
C
      CALL IMCHEK(1,ICY,IVTAM,IDX)
C
C  Relate I/M start year to calendar year,  using JDX ordering.   This pointer
C  will be used within IDX loop to  set  kinked tampering rate curves flag.
C
      IMDXSY-ICY- ICYIM+1
C
C  For each tampering vehicle type  and  each model year in  the 24 year window
C  from ICY-23 thru ICY  (25+ year case  is  always zeroed),  compute the bag  and
C  evaporative tampering offsets.
C
      DO 40 IVTAM-1,4
C
      CALL IMCHEK(2,ICY, IVTAM,IDX)
C
      DO 30 IDX-2,MAXYRS
C
C  MY pointer info using IDX & ICY  may  be  done here,  before  sub  calls.
C
C  Identify this loop's model year  for  use in ITAMPT lookups.
C
      MY-ICY- 1,  increases with age) my index for use  in
C  apportioning mileage weights.
C
      JDX-MAXYRS-IDX+1
C
C  Reset  (turn off) the ATP my range switch.   Turn  on if there is an ATP
C  program, it starts before the calendar  year,  the model  year JDX is in the
C  affected range and the vehicle type  has been selected by  the  user for
C  coverage.
C
C      LMYRVT - 1 -> switch is off.  ATP not a factor in my  JDX'a tampering
C      LMYRVT - 2 =»> switch is ON.  ATP is a factor in my  JDX's  tampering
C
C  Note: ATPFLG - 1 -> LAPSY - 2020 =>  LDXSY < 1 => 1st 2  criteria can be
C        checked using LDXSY.    (...Except for GEM,  which  changes ATPFLG
C        without changing LAPSY or  LDXSY accordingly.   So  we need to include
C        the ATPFLG check for use with  CEM.)
C
      LMYR-1
      LMYRVT-1
C
      IF (LDXSY.GT.l.AND.LDX1ST.GE. JDX.AND. JDX.GE.LDXLST)   LMYR-2
C
C     IF (LDXSY.GT.l.AND.ATPFLG.GT.l.AND.LDX1ST.GE. JDX.AND. JDX.GE.LDXLST)
C    *  LMYR-2
C
C  either add next line or use revised  long IF line above  for CEM...
C. . .
      IF (ATPFLG. EQ. 1) LMYR-1
C
      IF(LMYR.EQ.2.AND.LVTFLG(IVTAM) .EQ.2)  LMYRVT-2
C
      CALL IMCHEK(3,ICY,IVTAM, IDX)
C
C  Set kinked disablement rate curves flag.
C
      IMKINK-1
      IF (IM. EQ. 2. AND. JDX. GT. IMDXSY) IMKINK-2
C
C  Figure the base tampering rates  for  both exhaust and non-exhauat emissions.
C
      CALL DISATP
C
C  If ICY is beyond LPOD, use the tampering rates as of LPOD for the misfueling
C  categories (ID-3,4,9) for vehicles in ICY's fleet produced thru LPOD.
C
      IF(ICY.GT.LPOD) CALL FINDPB(ICY)
C
C  The next 3 subroutines compute the tampering bag emission additives.
C
      CALL EFFGRP
      CALL EMIRAT
      CALL BAGEME
C


                                                  21

-------
                                            Appendix A

C  The non-exhaust HC emission additives  calculation procedure  is much
C  simpler because there are no overlaps.   EVPEME  does  all  the  work.
C
      IF(PRTFLG.EQ.1.0R.PRTFLG.EQ.4)  CALL EVPEME
C
   30 CONTINUE
   40 CONTINUE
C
      RETURN
C End TAMPER
      END
                                                 22

-------
                                             Appendix A
      SUBROUTINE EFFGRP
c
C EFFGRP builds the tampering effects group
C rates from DISATP are modified to account
C types, overlap factors, anti-tampering pro
C technology group sizes. The output rates
C corresponding to the significant emission
C
C Called by TAMPER.
C
C Calls JVDJ12, ATPEFF, A8TO11, C1TO7, C8TO12
C
C Input on call:
C
C common blocks:
C /EGSCAL/ TGS
C /LOOKUP/ IVTAM
C /MYCODE/ MY,tJDX,LDXSY,LMYRVT
C /SIZCAL/ CSIZE
C
C Output on return:
C
C common blocks:
C /EGSCAL/ EGS
C /LOOKUP/ IQG, IPG, JPGD, IGCSF
"C /MYCODE/ IAY
C /SIZCAL/ TGSUSE,CSAE
C
C Local array subscripts:
C
C CSF1ST(7,2,3) - CSF1ST ( IC7, IPG, IGCSF
C
C Local variable / array dictionary:
C
C Name Type Description
C ,_™_~_-____-___~_~~______________
C CSF1ST R category size factor for 1st
C as  -  =• AorE for
C
C Notes :
C
C EFFGRP was modified for MOBILE4.1 v5 to ex
C its rates. Tampering effects are stored i
C
C

sizes array. The tampering
for technology vs tampering
gram effectiveness factors and
are divided into effects groups
impact tampering groups .



, ITAMPT, SETEGS and TECHO.



















)




calculation of CSIZE, expressed
.c. £req> / .
IC7-1-3, CATS for IC7-4-7



pand CSF1ST to 3 MYGs fi updates
n array BSIZE for CEM4.1.


C
c
c
c

c
c
c
   DIMENSION CSF1ST(7,2,3)

   COMMON /EGSCAL/ AER (11, 11, 2, 2) , TGS (11, 6, 4) , EGS (7, 2)
   COMMON /LOOKUP/ IVTAM,IQG,IPG,JPGD,IHG,IGCSF
   COMMON /MYCODE/ MY, IDX, JDX, LDXSY, LMYRVT, IAY, IMDXSY, IMKINK
   COMMON /SIZCAL/ ATR(5, 2) , TGSUSE, CSIZE (12, 2) ,CSAE (12)

   COMMON /COST32/ BSIZE(4,25,2,6,17,2),TRAT(9)

Continuation code is  IPG  - H (HC £ CO)  £ N (NOx)
MYG 1 applies to all  MY for HD6V;  the indicated ranges are for LDGV/T.
   DATA CSF1ST/
                           CSIZE
                             4
       123
MYG 1  (pre-1981)
  H  .0708,  .1420,  .1538,  .5529,  .0000,  .1513,
  N  .0948,  .1209,  .0033,  .3059,  .0039,  .3843,
MYG 2  (1981-83)
  H  .1421,  .0519,  .0364,  .2840,  .0406,  .1917,  .1469,
  N  .0698,  .0233,  .0116,  .2727,  .0260,  .2338,
MYG 3  (1984+)
  H  .1739,  .0750,  .0938,  .1011,  .0000,  .0470,
  N  .1212,  .0000,  .0000,  .0000,  .0000,  .0588,
                                                   .2500,
                                                   .0118,
                                                   .0130,

                                                   .0556,
                                                   .OOOO/

   Zero out tampering effects group  sizes  (EGS)  array.

      DO 20 IPG-1,2
      DO 10 IEG-1,7
      EGS(IEG,IPG)-0.0
   10 CONTINUE
   20 CONTINUE
                                                   23

-------
                                             Appendix A

C  Add in each equipment  (technology)  group's  before and after anti-tampering
C  program  (ATP) atart year  contribution to each EGS category for each
C  pollutant group.
C
      DO 80 IPG-1,2
C
C  Set ATR pointer for 1st 3 sizes:  HC/CO -> uae AIRS,  NOX -> use EGRS.
C
      IP(IPG.EQ.l) JPGD-1
      IF(IPG.EQ.2) OPGD-5
C
      DO 70 IQG-1,6
C
C  Lookup technology group size to be  used on  this pass.
C
      IGl-ITAMPT(l)
      TGSUSE-TGS (IG1, IQG, IVTAM)
C
C  Check how technology type £ size  affect tampering rates.
C
      CALL TECHO(*70)
C
C  HC S CO effects category  sizes are  identical  -> only 2 pollutant passes.
C
      DO 65 IAY-1,2
C
C  Skip to end of loop if no mileage accumulated (hence no tampering)
C  for the pre/post ATP year case being checked.
C
      IF(IAY.EQ.1.AND.LMYRVT.EQ.2.AND.LDXSY.GT.JDX)  GOTO 64
      IF(IAY.EQ.2.AND.LMYRVT.EQ.l) GOTO 64
C
      IGCSF-ITAMPT(6)
      CALL C1TO7(CSF1ST(1,IPG,IGCSF))
      CALL C8TO12
C
C  The first calculation pass can produce negative sizes for the no
C  overlap  (8-11) and the no effects (12)  categories.   Correct these
C  by calling A8TO11 and ADJ12, respectively.
C
      DO 30 IC=8,11
      IF(CSIZE(IC, IAY) .LT.O)  CALL A8TO11 (1C)
   30 CONTINUE
      IF(CSIZE(12,IAY).LT.O)  CALL ADJ12 (CSF1ST)
C
C  "Subsequent"  (to ATP start year)  pass + have  split mileage case =>
C  "subsequent" category size = overall size - "previous" size
C
C  I.e., in the split mileage case,  the program  computes the previous to
C  ATP start year and overall values up to this  point.   After this next
C  loop assigns the subsequent values  as residuals,  the program works
C  directly with the previous and subsequent cases in the calls to ATPEFF
C  and SETEGS.
C
      IF(IAY.EQ.1.0R.LMYRVT.EQ.1.OR.LDXSY.GT..TOX)  GOTO  50
      DO 40 IC-1,12
      CSIZE (1C, 2) -CSIZE (1C, 2) -CSIZE (1C, 1)
   40 CONTINUE
C
C  If ATP covers this MY, apply effectiveness  factors to the category sizes.
C
   50 IF(LMYRVT.EQ.2) CALL ATPEFF(*60)
C
C  Otherwise just shift lAY's sizes  to CSAE without modification.
C
      DO 55 IC-1,12
      CSAE (1C) -CSIZE (1C, IAY)
   55 CONTINUE
C
C  Set the effects group sizes.
C
   60 CALL SETEGS
C. . .
   64 DO 61 IC-1,12
        BSIZE (IVTAM, JDX,IPG,IQG, 1C, IAY)  - CSAE (1C)
   61 CONTINUE
   65 CONTINUE
   70 CONTINUE
   80 CONTINUE
      RETURN
      END


                                                   24

-------
                                             Appendix A

      SUBROUTINE EVPEME
C
C  EVPEME serve a the same functions  for evaporative (evap - hot soak + diurnal)
C  and crankcase  (cc) HC as EFFGRP,  EMIRAT  and BAGEME do  for exhaust IF,
C  producing ha S du HC tampering  ratea £ cc HC offset.   The task is simplified
C  into one subroutine because there is no  overlap categorization or
C  weighting.  The appropriate numbers  are  looked up and  multiplied
C  together with only anti-tampering effectiveness reductions (if any)  applied.
C
C  Called by TAMPER.
C
C  Calls ITAMPTrOTCALC.
C
C  Input on call:
C
C    common blocks:
C    /EVPDAT/ EVPAER,EVPTGS,CCEMI
C    /LOOKUP/ IVTAM
C    /MYCODE/ IDX, JDX,LDXSY,LMYRVT
C    /TAMEQ4/ BTR
C
C  Output on return:
C
C    common blocks:
C    /IM240P/ DSIZE
C    /LOOKUP/ IHG
C    /TAMOUT/ THS,TDU,TCC
C    /TAMOU2/ TOB,GCONLY
C
C  Local variable dictionary:
C
C   Name   Type              Description
u
C
C
C
C
C
C
C
C
C
C
C
OTR
OTRCAR
OTRINJ
PCTEQP


Notes :

EVPEME
R
R
R
R




overall (previous
carbureted hot
fuel injected
percentage of
given IG3



modified for MOBILE4.
tampering


rates in DSIZE.

+ subsequent my ranges) tampering rate
soak affecting overall
hot
soak affecting
IVTAM fleet




1 to






store


tampering
overall
equipped with




diurnal






and






rate

tampering rate
IHG technology








,




crankcaae








      INTEGER ATPFLG, TPDFLG, RLFLAG, TEMFLG, OUTFMT
C
      COMMON /EVPDAT/ EVPAER (3, 2) , EVPTGS (3, 2, 4) , CCEMI (7, 4)
      COMMON /FLAGS3/ ATPFLG, TPDFLG, RLFLAG, LOCFLG, TEMFLG, OUTFMT
      COMMON /IM240P/ DSIZE (25, 4) , IM24YR, IPRGYR, IPRSYR
      COMMON /LOOKUP/ IVTAM, IQG, IPG, JPGD, IHG, IGCSF
      COMMON /MYCODE/ MY, IDX, JDX,LDXSY, LMYRVT, IAY,IMDXSY, IMKINK
      COMMON /TAMEQ4/ BTR(9,2)
      COMMON /TAMOUT/ TAMBAG(3, 3, 25, 4) , THS (2, 25, 4) ,TDU (25, 4) , TCC (25, 4)
      COMMON /TAMOU2/ TOB (25, 4) , GCONLY (25, 4)
      COMMON /COST32/ BSIZE (4,25,2, 6,17,2) ,TRAT (9)
C
C  For MOBILE4,  crankcase (IH » 3)  processing does not change.
C  Diurnal processing only changes  in its selection of the new evaporative
C  cannister + gas  cap tampering (ID  - 8)  instead of the previously used
C  evaporative cannister only (ID = 6) .   Since aside from data point selection,
C  the procedures for calculating the diurnal and crankcase offsets are the
C  same, a loop  structure remains the convenient way to program the procedure
C  Hot soak, on  the other hand,  has become more complicated and has therefore
C  been pulled from the old IH loop and treated separately,  after the diurnal
C  and crankcaae offsets have been  computed.
C
C  MOBILE4 defers HS  and DU tampering offset  calculation to CCEVRT,
C  returning only the % equipped *  overall tampering rates instead.
C
C  The Diurnal and  Crankcase  tampering rates  are stored in BSIZE for CEM4.1.
C
C  Store the non-ATP  rates.
C
      DSIZE (IDX, IVTAM) -BTR (8, 1) +BTR(8, 2)
C
      DO 10 IH-2,3
C
                                                   25

-------
                                             Appendix A

C  There are only two evaporative  cases,  not  three,  when indexing the
C  anti-tampering effectiveness rates,  evaporative technology group sizes,
C  and the model year groups  for the  group  sizes  and the evaporative emission
C  impact rates:  for these arrays, the hot soak  and diurnal numbers are
C  identical.  For convenience we  use IHG instead of IH in these arrays.
C
      IHG-IH-1
C
C  Assign the tampering index (see paragraph  preceding DO statement).
C
      ID-9-ZH6
C
C  Get the overall tampering  rate.  As  in DISATP,  there are 3 cases:
C
C     (1) No ATP or its start  year  is  after  ICY  or IDX is not in affected my
C        range: use the "previous" base rate  without applying the
C        effectiveness rates.
C
      IF(LMYRVT.EQ.l) OTR-BTR(ID, 1)
C
C     (2) There is an ATP and  its start  year is  before IDX and IDX is in the
C        affected my range: use the "subsequent"  rate * the effectiveness
C        rate.
C
      IF (LMYRVT. EQ. 2. AND. LDXSY. GT. JDX. AND. ID. EQ. 7 )
      *  OTR-OTCALC(ID,2,IHG,2)
      IF(LMYRVT.EQ.2.AND.LDXSY.GT.JDX.AND.ID.EQ.8)
      *  OTR-OTCALC(ID-2,2,IHG,2)+OTCALC(ID,2,3,2)-OTCALC(ID-2,2,3,2)
C
C     (3) There is an ATP and  its start  year is  at or after IDX but before ICY
C        and IDX is in the affected my  range: sum the rates * their
C        corresponding effectiveness  rates.   Note that in this case, BTR holds
C        the "previous" and overall rates ->  subtract 1st from 2nd to  get the
C        "subsequent" rates.
C
      IF(LMYRVT.EQ.2.AND.LDXSY.LE.JDX.AND.ID.EQ.7)
      *  OTR-OTCALC (ID, 1, IHG, 1) +OTCALC (ID, 2, IHG, 2) -OTCALC (ID, 1, IHG, 2)
      IF (LMYRVT.EQ. 2. AND. LDXSY. LE. JDX. AND. ID.EQ. 8)
      *  OTR-OTCALC(ID-2,1,IHG,1)+OTCALC(ID,1,3,1)-OTCALC(ID-2,1,3,1)
      *     -K>TCALC(ID-2,2,IHG,2)+OTCALC (ID, 2, 3,2)-OTCALC(ID-2,2,3,2)
      *    -(OTCALC(ID-2,1,IHG,2)+OTCALC(ID,1,3,2)-OTCALC(ID-2,1,3,2))
C. . .
C  Save for CEM via EBSIZE...
C
      CALL EBSIZE(ID,OTR)
C
C  Lookup model year group pointers.
C
      IG3-ITAMPT(3)
C
      IF(IH.EQ.3) IG4-ITAMPT(4)
C
C  Lookup and apply the percentage equipped and emission impact factors.
C
      PCTEQP-EVPTGS(IG3,IHG,IVTAM)
      IF(IH.EQ.2) TDO(IDX,IVTAM)-OTR*PCTEQP
      IF(IH.EQ.3) TCC(IDX,IVTAM)-OTR*PCTEQP*CCEMI(IG4,IVTAM)
   10 CONTINUE
C
C  Now calculate the hot soak offset.   This case  is  complicated by both the
C  usage of 2 tampering types (evaporative  cannister only and evaporative
C  cannister plus gas caps) and the weighting together of fuel injected and
C  carbureted emission impact rates.
C
      IH-1
      IHG-IH
C
C  Assign the carbureted tampering index  •  evaporative canniater only,  since
C  carbureted hot soak HC is  not affected by  gas  cap tampering.   The  fuel
C  injected index is ID+2 - 8 - cannister + gaa caps.
C
      ID-6
C
C  Get the overall tampering  rates.   As with  diurnal and crankcase,  there are
C  3 cases,  but now 2 rates are computed.
C
C     (1) No ATP or ita start  year  is  after  ICY  or IDX is not in affected my
C        range:  use the "previous" base rate  without applying the
C        effectiveness ratea.
C
      IF(LMYRVT.NE.l)  GOTO 20


                                                   26

-------
                                             Appendix  A

      OTRCAR-BTR (ID, 1)
      OTRINJ-BTR(ID+2,1)
      GOTO 40
C
C    (2) There is an ATP  and its start year is before IDX and IDX is in the
C        affected my range:  use the "subsequent" rate * the effectiveness
C        rate.
C
   20 IF(LDXSY.LE.JDX)  GOTO  30
      OTRCAR-OTCALC (ID, 2, IHG, 2)
      OTRINJ-OTCALC(ID,2,IHG,2)+OTCALC(ID+2,2,3,2)-OTCALC(ID,2,3,2)
      GOTO 40
C
C    (3) There is an ATP  and its start year is at or after IDX but before ICY
C        and IDX is in  the affected my range:  sum the rates * their
C        corresponding  effectiveness rates.  Note that in this case, BTR holds
C        the "previous" and  overall rates -> subtract 1st from 2nd to get the
C        "subsequent" rates.
C
   30 OTRCAR-OTCALC(ID,1,IHG,1)+OTCALC(ID,2,IHG,2)-OTCALC(ID,1,IHG,2)
      OTRINJ-OTCALC (ID, 1, IHG, 1) +OTCALC (ID+2, 1, 3,1) -OTCALC (ID, 1, 3,1)
     *      +OTCALC (ID, 2, IHG, 2) +OTCALC (ID+2,2,3, 2) -OTCALC (ID, 2,3,2)
     *     -(OTCALC(ID,1,IHG,2)+OTCALC(ID+2,1,3,2)-OTCALC(ID,1,3,2))
C
C  Save the unweighted  (by PCTEQP)  Evaporative Cannister (EC) Only tampering
C  rate for figuring the  On  Board (OB) EC tampering impact on refueling losses.
C  Lookup model year group pointer for the percentage equipped table.
C
   40 TOB(IDX,IVTAM)-OTRCAR
      IG3-ITAMPT(3)
C
C  Lookup the percentage  equipped with canisters factor and apply it to
C  the  otr's to  get the tampered percentage of carbureted and fuel injected
C  cases for hot soak.
C
      PCTEQP=EVPTGS (IG3, IHG, IVTAM)
      THS (1, IDX, IVTAM) -PCTEQP*OTRCAR
      THS (2, IDX, IVTAM) =PCTEQP*OTRINJ
C
C  Store the gas cap only tampering values for running loss weighting.
C
      GCONLY (IDX, IVTAM) =OTRINJ-OTRCAR
      IF (GCONLY (IDX, IVTAM) .LT.0.0)  GCONLY (IDX, IVTAM)-0. 0
C
      RETURN
C  End  EVPEME
      END
                                                   27

-------
                                             Appendix A

      FUNCTION FAIL (MYR, KDX, IV, IK)
C
C  FAIL handles the effect  of Purge and Pressure failures as well as
C  evap system tampering, returning a fraction for purge/pressure.
C  Revised for CEM to  save  P/P failure rates for an in-place
C  inspection program  (i.e.,  fail rates lower than for M4.1).
C
C  Called by HCCALX.
C
C  Calls ENFORC, ITAMPT.
C
C  Input on call:
C
C    parameter list: MYR,KDX,IV,IK,IG3
C    common blocks:
C    /EVPDAT/ PCTEQP
C    /FLAGS3/ ATPFLG
C    /IMPAR1/ GRIM
C    /IMPAR2/ ILDT
C    /IMPAR6/ INTYP
C    /IM240P/ DSIZE,IPRGYR,IPRSYR
C    /LOOKUP/ IVTAM
C    /MAXIMA/ MAXYRS
C    /MYCODE/ MY
C    /TAMOUT/ TDU
C
C  Output on return:
C
C    function: FAIL
C
C  Notes.
C
C  FAIL was added for  MOBILE4.1.
C
C... for CEM...

C  PPFAIL  Block    Recurring (lower) repair rates for Purge, Press,
C                   Purge-fPress for each MY S IV for Repair Costs.
C  PPFIXR  Block    Purge/Pressure Fix rates = Full non-program fail
C                   rate  minus w/program failure rate for FE benefits.
C  ISRECR  Local    Flag  "Is Recurring" to loop again with lower
C                   P/P fail rates.
C  	
C
C
       INTEGER ATPFLG, TPDFLG, RLFLAG, TEMFLG, OUTFMT
C
       COMMON /EVPDAT/  EVPAER(3, 2) ,EVPTGS (3, 2, 4) ,CCEMI (7, 4)
       COMMON /FLAGS3/  ATPFLG, TPDFLG, RLFLAG, LOCFLG, TEMFLG, OUTFMT
       COMMON /IMPAR1/  ICYIM, ISTRIN,MODYR1,MODYR2, WAIVER(2) ,GRIM
       COMMON /IMPAR2/  ILDT (4) , ITEST,NUDATA(2) ,NLIM, IMNAME (20, 9)
       COMMON /IMPAR6/  IFREQ, INTYP
       COMMON /IM240P/  DSIZE (25, 4) , IM24YR,IPRGYR, IPRSYR
       COMMON /LOOKUP/  IVTAM, IQG, IPS, JPGD, IHG, IGCSF
       COMMON /MAXIMA/  MAXVEH,MAXLTW, MAXPOL, MAXREG,MAXYRS
       COMMON /MYCODE/  MY, IDX, JDX,LDXSY,LMYRVT, IAY, IMDXSY, IMKINK
       COMMON /TAMOUT/  TAMBAG(3,3,25,4),THS(2,25,4),TDU(25,4),TCC(25,4)
C. . .
       COMMON /COST07/  CSTIM,GASCST,IMSTRT,ISCEN,MOVIM,MOVATP,REPIM(4)
       COMMON /COST28/  JATP, JIM, ISTRT, JCYDX, JPRGYR, JPRSYR
       COMMON /COST47/  PPFAIL(25,3,10,4),PPFIXR(25,10,4)
C
       DIMENSION RATE(3,13),OVLP(5),OVLR(5),PPEFF(25,2)
C
C  Revised Purge/Pressure Failure Rate calc'a, 3/26/91
       DATA RATE   /  .043,  .043,   .080,
                     .043,  .043,   .080,
                     .043,  .043,   .080,
                     .060,  .053,   .096,
                     .060,  .053,   .103,
                     .069,  .053,   .120,
                     .094,  .053,   .150,
                     .094,  .106,   .188,
                     .142,  .152,   .258,
     *               .214,  .171,   .323,
     *               .224,  .236,   .389,
     *               .229,  .336,   .442,
     *               .229,  .336,   .451   /
C
       DATA FACTOR,DEFF /  0.50, 0.50 /
C


                                                   28

-------
                                             Appendix  A

c
C  Purge/Pressure Check Effectiveness
C
C  Purge/Pressure Check Effectiveness,  Annual,  Biennial
C
      DATA PPEFF /
     *  .000,.238,.853,.910,.836,.789,.765,.732,.798,.837,
     *  . 903, . 967, . 992, . 981, . 967, . 958, . 922, . 928, . 927, . 941,
     *  .972,.992,.981,.967,.958,
     *  .000,.238,.853,.910,.779,.789,.612,.732,.582,.837,
     *  . 733, . 967, . 945, . 973, . 947, . 925, . 849, . 850, . 701, . 853,
     *  .705,.972,.945,.973,.947 /
C
C  Initialize variables
C
      FAIL=0.0
      PRGR-0.0
      PRSR-0.0
      PPRT-0.0
      DO 10  1=1,5
        OVLP(I)=0.0
        OVLR(I)-0.0
   10 CONTINUE
C. . .
      ISRECR - 1
C
C  For  diesel S motorcycles,  return 0.0 for FAIL
C
      IF(IV.GT.4) RETURN
C
C  Determine percent  of vehicles equipped with evap systems
C  (Use KDX  and MYR to avoid problems with common block MYCODE)
C
      MY-MYR
      IDX-KDX
C
      IHG-1
      IVTAM-IV
      IG3-ITAMPT(3)
      PCTEQP-EVPTGS(IG3,1,IV)
C
C  Determine overall  Pressure and Purge failure rates
C
      JDX= (MAXYRS+1) -IDX
      JDXX — JDX
      IF(JDXX.GT.13)  JDXX=13
      PRGR-RATE (1, JDXX) *PCTEQP
      PRSR-RATE (2, JDXX) *PCTEQP
      PPRT-RATE(3,JDXX)*PCTEQP
C
      TRAT-DSIZE (IDX, IV) *PCTEQP
C. . .
C  This is where we jump in for the recurring rate loop...
C
   15 SUM-PRGR+PRSR
      IF(PPRT.GT.SUM)  PPRT-SUM
C
C  Determine  proportion of overall in each overlap
C
        OVLP ( 4) -TRAT*FACTOR
        OVLP (5) -TRAT* (1. 0-FACTOR)
        OVLP (1) =PPRT-PRGR-OVLP (4)
      IF(OVLP(1) .LT.0.0) OVLP(1)=0.0
        OVLP (2) -PRSR-OVLP (1) -OVLP (4) -OVLP (5)
      IF(OVLP (2) .LT.0.0) THEN
        OVLP(2)-0.0
        OVLP (5) -PRGR- (PPRT-PRSR)
        OVLP (4) -OVLP (4) +TRAT* (1. 0-FACTOR) -OVLP (5)
      ENDIF
        OVLP (3) -PRGR-OVLP (2) -OVLP (5)
      IF(OVLP (3) .LT.0.0) OVLP(3)-0.0
C
C  Initialize rates in overlaps
C
      DO 20 1-1,5
        OVLR(I)-OVLP(I)
   20 CONTINUE
C
C...  10/30/91 change  to  skip over  program reductions with
C    recurring repairs for  ongoing program (already part of
C    recurring RATEs)


                                                   29

-------
                                              Appendix A

c
      IF(ISRECR.EQ.2)  GOTO 40
G
C  Adjust rates  to reflect programs
C
      EADJ-1.0 - ENFORC(CRIM,1)*PPEFF(JDX,IFREQ)
      ADJ-l.O-EADJ
C
C Purge Check Alone
C
      IF(MY.GE.IPRGTTR.AND.ILDT(IV) .EQ.2) THEN
        IF(INTYP.EQ.l)  THEN
          OVLR(2)-EADJ*OVLP(2)
          OVLR(3)-EADJ*OVLP(3)
          OVLR (5) -EADJ*OVLP (5)
        ENDIF
        IF(INTYP.GT.l)  THEN
          OVLR(2)-DEFF*ADJ*OVLP(2)+EADJ*OVLP(2)
          OVLR (3) -DEFF*ADJ*OVLP (3) +EADJ*OVLP (3)
          OVLR(S)=DEFF*ADJ*OVLP(5)+EADJ*OVLP(5)
        ENDIF
          OVLR(1)-OVLP (1)+OVLP (2) -OVLR(2)
          OVLR(4)-OVLP (4)+OVLP (5) -OVLR(5)
      ENDIF
C
C Pressure Check Alone
C
      IF(Mr.GE.IFRSYR.AND.ILDT(IV) .EQ.2) THEN
        IF -OVLP (3) +OVLP (2) -OVLR (2) H-OVLP < S) -OVLR (5)
      ENDIF
C
C Combined Purge/Pressure Checks
C
      IF(INTTP.NE.3.
     *    AND.MY.GE.IPRSYR.
     *    AND.MY.GE.IPRGYR.
     *    AND.ILDT(IV).EQ.2)  THEN
        IF(INTYP.EQ.l)  THEN
          OVLR(1)-EADJ*OVLP(1)
          OVLR (2) -EADJ*OVLP (2)
          OVLR (3)-EADJ*OVLP (3)
          OVLR ( 4)-EAD J*OVLP (4)
          OVLR(5)-EADJ*OVLP (5)
        ENDIF
        IF(INTYP.GT.l)  THEN
          OVLR (1) -DEFF*ADJ*OVLP (1) +EADJ*OVLP (1)
          OVLR(2)-DEFF*ADO*OVLP(2)+EADJ*OVLP(2)
          OVLR (3 ) -DEFF*AD J*OVLP ( 3 ) +EAD J*OVLP (3)
          OVLR (4) -DEFF*AD J*OVLP (4 ) +EAD J*OVLP ( 4)
          OVLR (5) -DEFF*AD J*OVLP (5 ) +EAD J*OVLP ( 5)
        ENDIF
      ENDIF
C
C Anti-tampering program effects
C
      IF(ATPFLG.EQ.2.AND.ILDT(IV) .EQ.2)  THEN
        IF(MY.LT.IPRSYR.OR.ILDT(IV) .NE.2) THEN
          OVLR (4) -TDO (IDX, IV) *FACTOR
          OVLR (5) -TDO (IDX, IV) * (1. 0-FACTOR)
        ENDIF
      ENDIF
C
C  Return adjusted overall Pressure/Purge rate
C
C  IK — 1 : Purge only effects
C       2 : Pressure  effects
C
      IF(IK.EQ.l)  FAIL-OVLR<3)
      IF (IK. EQ.2)  FAIL-OVLR(1)+OVLR(2)+OVLR(4)+OVLR(5)
                                                    30

-------
                                             Appendix A

c...
C  Only need CEM info  for  full ATP  run...
C
      IF(ISCEN.NE.4) GOTO  99
C
C
C  Save difference between no pgm S with pgm total failure rate
C  for FE benefit calcs  (PPFEB)  in  PPFIXR...
C
      IF(ISCEN.EQ.4) PPFIXR(IDX, JCYDX, IV)  -
     *   OVLP(1)+OVLP(2)+OVLP(3)+OVLP(4)+OVLP(5)
     *  -OVLR(1)-OVLR(2)-OVLR(3)-OVLR(4)-OVLR(5)
C
C  Force lower  "recurring" failure  rates for in-place program...
C
      PRGR - 0.03
      PRSR - 0.025
      PPRT - 0.05
      ISRECR -  2
C
C  Go back and  recalc  overlaps  for recurring failures...
C
      GOTO 15
C
C  Adjusted overall  Pressure/Purge rate
C
C     Store Purge fi  Pressure Repair rates for later repair cost calcs
C
C     OVLR(l) " Press  fail only (no tampering)
C     OVLR(2) - Press  fi  Purge  fail  (no tampering)
C     OVLRJ3) - Purge  fail only (no tampering)
C     OVLRJ4) - Press  fail due  to tampering
C     OVLR(5) - Purge  £  Press  fail  due to tampering
C
C  Save recurring Purge/Press  failure £ fix rates  from loop with ATP..
C
C   PPFAIL(x,lrx,x)  -  Entire PURGE failure rate
C   PPFAILJx,2,x,x)  =  Entire PRESSURE failure rate
C   PPFAILJx,3,x,x)  -  Total PURGE fi PRESSURE failure rate
C   i.e., overlap would  need to be computed by:
C   PPFAIL(l) + PPFAIL(2)  - PPFAIL(3)
C
   40 IF(ISCEN.EQ.4  .AND.  IV.LE.4)  THEN
        PPFAIL(IDX,1, JCYDX, IV)  - OVLP (2)  + OVLP (3)
     *                          + OVLP(5)
        PPFAIL(IDX,2,JCYDX,IV)  - OVLP(l)  + OVLP (2)
     *                          + OVLP (4)  + OVLP (5)
        PPFAIL(IDX,3,JCYDX,IV)  - OVLP(l)  + OVLP (2)
     *                          + OVLP (3)  + OVLP (4)
     *                          + OVLP(5)
      ENDIF
C
   99 RETURN
C  End FAIL
      END
                                                   31

-------
                                             Appendix A

      FUNCTION PCLEFT(MY,ICY,IP,IV)
C
C  PCLEFT determines the basic emission  factor multiplicative adjustment
C   ( - percentage left of BEF - PCLEFT )  for the effects  of an inspection /
C  maintenance  (I/M) program, after  checking whether or not I/M applies to the
C  factor being computed by the calling  function.
C
C  Called by BEF.
C
C  Calls IMPTR.
C
C  Input on call:
C
C    parameter list: MY,ICY,IP,IV
C    common blocks:
C    /FLAGS2/ IMFLAG
C    /ICR4VA/ CRD4VA
C    /ICR4VB/ CRD4VB
C    /IMPAR1/ ICYIM, ISTRIN,MODYR1,MODYR2,WAIVER,CRIM
C    /IMPAR2/ ILDT,ITEST
C    /IMPAR5/ CRHDGV, DISCNT
C    /IMPAR6/ IFREQ,INTYP
C    /IM12HC/ CR12HC
C    /IM12CO/ CR12CO
C    /IM240P/ IM24YR
C
C  Output on return:
C
C    function: PCLEFT
C
C  Local variable / array dictionary:
C
C   Name   Type              Description
C  	  	  	
C  AGE1ST   I    age of the vehicle  at first inspection
C  BY       I    benefit year for technology 1 or 2  vehicle
C  IBY      I    benefit year for technology 4 plus  vehicle
C  IREM     I    remainder •> stringency  - greatest multiple of 10  < stringency
C  ISTRN    I    stringency index into technology 1  or  2  credits array
C  ITESTL   I    local version of ITEST  used for IM240  check
C  REM      R    IREM converted to REAL  value
C  WAIV     R    waiver rate used on this call,  depends on MY
C
C  Notes:
C
C  Non-TECH4+ vehicles use 19th value for 20-24 benefit year.
C  TECH4+ vehicles have arrays expanded  to MAXYRS-1.
C  PCLEFT was modified to include the IM240  check.
C
C
      CHARACTER* 4 IMNAME
      INTEGER ALHFLG
C
      COMMON /FLAGS2/ MYMRFG, NEWFLG, IMFLAG, ALHFLG
      COMMON /IM12HC/ CR12HC(19,20,5,2)
      COMMON /IM12CO/ CR12CO(19,20,5,2)
      COMMON /IM240P/ DSIZE (25, 4) ,IM24YR,IPRGYR,IPRSYR
      COMMON /ICR4VA/ CRD4VA(24,2,12,3)
      COMMON /ICR4VB/ CRD4VB (24,2,12,3)
      COMMON /IMPAR1/ ICYIM, ISTRIN,MODYR1,MODYR2, WAIVER (2) , CRIM
      COMMON /IMPAR2/ ILDT (4) , ITEST,NUDATA(2) , NLIM, IMNAME (20, 9)
      COMMON /IMPAR5/ MYGIM2 (3, 3) , LBIM4P, CRHDGV(2, 2) , DISCNT (3)
      COMMON /IMPAR6/ IFREQ,INTYP
      COMMON /MAXIMA/ MAXVEH,MAXLTW,MAXPOL,MAXREG,MAXYRS
C
      COMMON /COST07/ CSTIM,GASCST, IMSTRT, ISCEN,MOVIM,MOVATP,REPIM (4)
C
      INTEGER BY,AGE1ST
C
C  Initialize PCLEFT to "no reduction" value - 1.0.
C
      PCLEFT-1.0
C
C  If factor being calculated by BEF is  not  covered  by  I/M,  then RETURN.
C
      IF(IMFLAG.EQ.1.OR.IP.EQ.3.OR.IV.GT.4)  GOTO 99
C. . .
      IF(ILDT(IV).EQ.l.OR.
     *   ICY.EQ.MY.OR.ICY.LE.IMSTRT.OR.
     *   MY.LT.MODYR1.OR.MY.GT.MODYR2) GOTO  99
                                                   32

-------
                                             Appendix  A

C  Assign waiver rates
C  Control WAIV  (waver  rates)  for INTYP (I/M program type). INTYP 2-
C  I/M Decentralized/Computerized,  £ 3- Decentralized/Manual, if INTYP is
G   Hot a 1- Centralized I/M and the WAIV is LE 50%,  WAIV is set to zero.
C
      IF(MY.LE.1980) WAIV»WAIVER(1)
      IF(MY.GT.1980) WAIV-WAIVER(2)
      IF(WAIV.LE.0.50.AND.INTYP.GT.1) WAIV-0.0
C
      IF(IV.EQ.4) GOTO  50
C
C  Selecting I/M reduction for LDGV or LDGT.  Several parameters must be set.
C
C  Determine technology by model year and vehicle type.
C
      ITECH-IMPTR(MY,IV)
C
C  Find the benefit year:
C  Limit value according to technology type.
C
C     BY-ICY-ICYIM
C...  IF(MY.GT.ICYIM) BY-ICY-MY
      BY-ICY-IMSTRT
      IF(MY.GT.IMSTRT)  BY-ICY-MY
      IF(ITECH.LE.3.AND.BY.GT.19) BY-19
      IF(ITECH.GE.4.AND.BY.GT.MAXYRS-1) BY-MAXYRS-1
C
C  Find the age  of the  vehicle at first inspection.
C  Limit value according to technology type.
C
      AGE1ST-1
C. . .
      IF(MY.LT.IMSTRT)  AGE1ST-IMSTRT-MY+1
C
      IF(ITECH.LE.3)  THEN
        IF(AGE1ST.GT.19)  AGE1ST-19
        IF(BY+AGE1ST.GT.19) BY-20-AGE1ST
      ENDIF
C
C  For now, continue  MOBILES policy of using technology 2 credits for ITECH-3.
C
      IF(ITECH.EQ.3)  ITECH*>2
C
C  Branch on technology type:  TECH 1 & 2 form 1 group,  TECH 4+ another.
C
      IF(ITECH.GE.4)  GOTO 40
C
C  Select correct I/M credits for TECH 1 fi 2.   The same credits array is used
C  for LDGV and LDGT, but the I TECH mygs differ, so that the same MY may yield
C  a different credit for LDGV than for LDGT.
C
C  Interpolate between  10,20,30,40 £ 30% stringency.
C
      IKEM-ISTRIH-(ISTRIN/10)*10
      REM-IREM*.1
      ISTRH-(ISTRIN-IREM)/10
C
      IF (IFREQ. EQ. 1)  THEN
        IF(IP.EQ.l) THEN
          PCRED-CR12HC (BY,AGE1ST, ISTRN, ITECH)
          IF(ISTRN.LT.S.AND.IREM.GT.O)  PCRED-
     *       (CR12HC (BY,AGE1ST, ISTRN+1, ITECH) -PCKED) *REM+PCRED
        ELSE IF(IP.EQ.2)  THEN
          PCRED-CR12CO(BY,AGE1ST,ISTRS,ITECH)
          IF(ISTRN.LT.S.AND.IREM.GT.O)  PCRED-
     *       (CR12CO (BY, AGE1ST, ISTRN+1, ITECH) -PCRED) *REM+PCRED
        ENDIF
C
      ELSE IF(IFREQ.EQ.2)  THEN
        IF(IP.EQ.l) THEN
          PCRED-CR12HC (20-BY, 21-AGE1ST, ISTRN, ITECH)
          IF(ISTRN.LT.5.AND.IREM.GT.O)  PCRED-
     *       (CR12HC (20-BY, 21-AGE1ST, ISTRN+1, ITECH) -PCRED) *REM+PCRED
        ELSE IF(IP.EQ.2)  THEN
          PCRED-CR12CO(20-BY,21-AGE1ST,ISTRN,ITECH)
          IF(ISTRN.LT.5.AND.IREM.GT.O)  PCRED-
     *       (CR12CO (20-BY, 21-AGE1ST, ISTRN+1, ITECH) -PCRED) *REM+PCRED
        ENDIF
C
      ENDIF
C


                                                   33

-------
                                             Appendix A

      GOTO 60
c
C  Select I/M credits for TECH 4+ vehiclea.  For TECH  4+, there are  separate
C  credits arrays, one for annual and one for biennial inspection programs.
C  Use the Loaded/Idle Test for IM240 check.
C
   40 IBY-BY+AGE1ST-1
C
      ITESTL-ITEST
      IF(M¥.GE.IM24YR.AND.MY.GE.1981) ITESTL-3
C
      IF (IFREQ. EQ. 1) PCRED-CRD4VA (IBY, IP , ITECH-3 , ITESTL)
      IF(IFREQ.EQ.2) PCRED-CRD4VB(IBY,IP,ITECH-3,ITESTL)
      GOTO 60
C
C  Assign HDGV I/M credit.
C
   50 PCRED-CHHDGV(IP,IFREQ)
C
C  Set I/M benefit adjusted for waivers and enforcement and discount
C
   60 PCLEFT-1. 0- (PCRED* (1. 0-WAIV) *ENFORC (GRIM, 1) *DISCNT (INTYP) )
C
   99 RETURN
C  End PCLEFT
      END
                                                  34

-------
                                       Appendix  A
c
c
c.
c
c
c.
c
c.
c
G
C
c
c
c
c
c
c
c
c
c
c
c
c
c
MAIN



.Main Driver for
Model .





modified MOBILE4/Cost Effectiveness


.Calls M4MAIN, SETUP, INTWDW, FLEET, TBASE, LOOP, SUMWDW and PRINT

.Variable

Name

IATPLP

INERR
ISIZE1


ISIZE2


PPFAIL





Dictionary:

Source

Block

Par am
Par am


Far am


Block




Description

Flag indicating whether an ATP only scenario
is to be run.
MOBILE 4 error counter in subroutine QUITER
Variable containing either the size of the
floating model year window for the I/M coverage
or the earliest model year inspected in I/M.
Variable containing either the size of the
floating model year window for the ATP coverage
or the earliest model year inspected in an ATP.
Failure rates for Purge, Press, Purge+Press for
each model year and each IVGAS


REAL    JTJLMYR
INTEGER PROMPT
CHARACTER*20 M4IN,M4OUT
CHARACTER*8 USNAME
CHARACTER* 4 NOTES , COMMA, PERIOD
CHARACTER*! COLON

COMMON /COST01/ ATPFEE
COMMON /COST02/ IFEVYR, ILEVYR
COMMON /COST03/ STRNGY,RMAXF,REPEAT,FFACT
COMMON /COST04/ TAMPER(25,6,4,10,4)
COMMON /COST05/ REPLCE(25,6,10,4)
COMMON /COST06/ STOAER
COMMON /COST07/ CSTIM,GASCST, IMSTRT,ISCEN,MOVTM,MOVATP,REPIM(4)
COMMON /COST08/ FAILRT(2,25,39,4)
COMMON /COST09/ DRSTRN(5),ZMSTRN(5)
COMMON /COST10/ FEBNFT(25,10,4),FECNMr(20,4)
COMMON /COST11/ PCNTSVJ4)
COMMON /COST12/ IMFXFG, IAFXFG, IMCSFG, IAFEFG, IALDV
COMMON /COST13/ CIMFEE(25,10,4),CIMREP(25,10,4)
COMMON /COST14/ TONRED(25,2,10,4),COST(25,10,4)
COMMON /COST15/ TOTCST(2),TOTRED(2),TCOST
COMMON /COST16/ DATPFE,DIMFEE,DIMREP,0ATPRE,AATPFE,AIMFEE
COMMON /COST17/ DETER(25,2,10,4),TDETER(25,2,10, 4)
COMMON /COST18/ AIMREP,AATPKE, CFACTR,CATPFE (25, 10, 4)
COMMON /COST19/ ICOV1,ICOV2,INDVL1,INDVL2
COMMON /COST20/ TD0CT(25,2,3,10,4),TEVAP(25,10,4)
COMMON /COST21/ CATPRE(25,10,4),FOLBEN(25,20,4)
COMMON /COST22/ ATCOST(2,8)
COMMON /COST23/ VCNT(39,8),GSDSCT(8,10),HDDCnM(25,10)
COMMON /COST24/ ATEQP(20,5,4)
COMMON /COST25/ FRLOSS(25,4,4,10),FREF(25,4,4,10)
COMMON /COST26/ FRSTLS(25,8,4,10),YRSTLS(10),BRSTLS
COMMON /COST27/ IGRFLG, IATPLP, IDIFF,OUTPRF
COMMON /COST28/ JATP,JIM,ISTRT,JCYDX,JPRGYR,JPRSYR
COMMON /COST30/ EVAP (25, 8, 4,10) ,FER(3,25, 8, 4,10)
COMMON /COST31/ 3AER(11,11,2,2),SEVP(2,2),ZAER(11,11, 2,2)
COMMON /COST32/ BSIZE(4,25,2,6,17,2),TRAT(9)
COMMON /COST33/ EMCAL4(7,3,3),EVPD1(3,2,4),EVPD2(7, 4)
COMMON /COST34/ TCOONT,BYEAR,VGRORT
COMMON /COST35/ TCST,TBEN(2),ADFBEN,ADIM(2),ADAT(2),ADET(2)
COMMON /COST36/ YTONS(2,10),BTONS(2),BIDTON(2) ,AEVAP
COMMON /COST37/ YRCOST(10),YRBEN<10,2),YCOUNT(39),ACOUNT
COMMON /COST38/ STOCK(12,8),CLDIST(8)
COMMON /COST40/ YEVTON (10) ,BEVTON, YEREF (10) ,BEREF
COMMON /COST41/ YERLOS(10),BERLOS
COMMON /COST42/ ECAL02(10,6,4),ECAL03(1,2),ECAL04(20, 4)
COMMON /COST43/ EMCAL1(3,3,3),EMCAL2(3,3,3),EMCAL3(4, 3, 3)
COMMON /COST44/ LOOK1 (5) ,MYCDl (8), TAM3Q1 (6)
COMMON /COST45/ TAMD1(8,3,3),TAMD2(8,3,3),TAMD3(8,4,3)
COMMON /COST46/ REDIM(25,2,10,4),REDATP(25,2,10,4)
COMMON /COST47/ PPFAIL(25,3,10,4),PPFIXR(25,10,4)
COMMON /COST48/ PRGFEE,PRSFEE,CEVFEE(25,10,4),TRNFEE
COMMON /COST49/ PPFEB(25,10,4),APFBEN

COMMON /COSTXX/ JMYR1,JMYR2
                                             35

-------
                                             Appendix A
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
      COMMON /IONAME/ M4IN,M4OUT,M4IMC,C4IN
C
c
c
c
c
	 c
t
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
COMMON
Common block with names other than COST?? originate in
:he MOBILE4 section of this program.
/ALUHIN/ AC,XLOAD{3) ,TRAILR<3) , ABSHUM, DB, WB
/ATOAR1/ LAPSY,LAP1ST,LAPLST,LVTFLG(4)
/ATPAR2/ ATPPGM,ATPFQT,CRATP,DISTYP (8)
/CITPAR/ SCNAME(4)
/CITRV1/ RVPBAS, RVPIUS, RVEICY, IUSESY, RVPUWX
/CUMCOM/ CUMMIL(25,8)
/FLAGS1/ PROMPT, TAMFLG, SPDFLG, VMFLAG, OXYFLG, DSFLAG
/FLAGS2/ MTMRFG, NEWFLG, IMFLAG, ALHFLG
/FLAGS3/ ATPFLG, TPDFLG, RLFLAG, LOCFLG, TEMFLG, OUTFMT
/IMPARl/ ICYIM,ISTRIN,MODYR1,MODYR2,WAIVER(2) ,CRIM
/IMPAR2/ ILDT(4),ITEST,NUDATA(2),NLIM, IMNAME(20,9)
/IMPAR6/ IFREQ,INTYP
/IM240P/ DSZZE (25, 4) , IM24YR, IPRGYR, IPRSYR
/IOUCOM/ IOUIMD, IOCGEN, IOOREP , IOUERR, IOUASK
/LOOKUP/ IVTAM, IQG, IPG, OPGD , IHG, IGCSF
/MAXIMA/ MAXVEH,MAXLTW,MAXPOL,MAXREG,MAXYRS
/MYCODE/ MY, IDX, JDX, LDXSY, LMYRVT, I AY, IMDXSY, IMKINK
/MYRSAV/ AMAR(25,8),JULMYR(25,8),NEWCOM
/PROJEC/ PROJID<20)
/REGION/ FEET (2) , IRE JN, ALT, INITPR
/SAVE01/ JCALL,VA001 (15, 3, 8, 2) , VA002 (15, 3, 8, 2)
/SCENE1/ SPD(8),PCCN,PCHC,PCCC
/STRING/ NOYES (2) , COMMA, PERIOD, COLON
/TAMEQ4/ BTR(9,2)
/TEMPS/ AMBT, TEMMIN, TEMMAX, TEMEXH (3) ,TEMEVP(6)
/USDATA/ USNAME (4,4), NUSD , IUSD ( 4 )
      COMMON /VMXCOM/ REGMIX (8) , TFNORM(8) , VMTMIX (8) ,VCOUNT (39, 8)
      COMMON /YEARS4/ IY1941,IY1960,IY2020
 DRIVER  (as of 08/07/91):


 DRIVER demonstrates how to call the subroutine version  of
 MOBILE4.1


 Calls MOBILE.


 Local variable / array  dictionary:


  Name   Type              Description
 INTER
 M4CALL
C
I
flag for interactive mode ('Y') vs. batch mode  ('N')
loop index for MOBILE4.1 call loop
    PROGRAM DRIVER
    IFPRMT IsFilePrompt:
      1 - M4INPUT...C4OOTPOT
      2 - Prompt for each of 4 I/O files for each run
      3 - All I/O  files for all runs listed in one given  filena


    LOGICAL ISFILE,ISFLST
    CHARACTER*20 C4IN, C4OOT, M4IMC, TMPFIL, M4OOT1, C4OUT1
    CHARACTER*! INTER


    IOUGEN-5
    IOUGEN«0
    IOUREP-6
    IOUASK-0
    IOUERR-6
    M4IMC - 'M4IMC'
    M4OUT1 = '  '
    C40UT1 = '  '


    WRITE(IOUASK,300)
300 FORMAT(' ****** CEM4.1 DRIVER for MOBILE4.1 *****'/)


    CALL GETCL(M4IN)
                                                   36

-------
                                             Appendix A

      IF(M4IN.EQ.'  ') THEN
        IFPRMT - 1
        M4IN  - 'M4INPUT'
        M4OUT - 'M4OUTPUT'
        C4IN  - 'C4INPUT'
        C40UT - 'C4OUTPUT'
      ELSE IF(M4IN.EQ.' /'.OR.M4IN.EQ.'/'
     *    .OR.M4IN.EQ.'?'.OR.M4IN.EQ. '  ?')  THEN
C  Filename request mode. . .
        IFPRMT - 2
      ELSE
C  File list mode...
        INQUIRE(FILE - M4IN,EXIST  - ISFLST)
        IF(ISFLST) THEN
          IFPRMT - 3
          OPEN(UNIT-3, FILE-M4IN,  STATUS-'OLD')
        ELSE
          WRITE(0,200) M4IN
  200     FORMAT(' File list  not found:  ',A20)
        ENDIF
      ENDIF
C
      M4CALL-0
C
C  ******************************************************
C  BEGIN Driver LOOP To run each input/output file aet...
C  ******************************************************
C
   50 IF(IFPRMT.EQ.3) THEN
        READ(3,220,ERR-98,END-99)  M4IN,M4OUT,C4IN,C4OUT,M4IMC
  220   FORMAT (5 (A20/) )
      ELSE IF(IFPRMT.EQ.2) THEN
  230   FORMAT(A20)
        WRITE(IOUASK,240)
  240   FORMAT('  ','Pleaae enter the MOBILE4.1 Input filename',
     *    '  (Return Quits):')
        READ(IOUASK,230,END-99) M4IN
        IF(M4IN.EQ.'  ') GOTO  99
        WRITE(IOUASK,250)
  250   FORMAT('S1,'Pleaae enter the MOBILE4.1 Output filename:')
        READ (IOUASK, 230,END-99) M4OUT
        WRITE(IOUASK,260)
  260   FORMAT('fi1,'Pleaae enter the CEM4.1 Input filename:')
        READ(IOUASK,230,END-99) C4IN
        WRITE(IOUASK,270)
  270   FORMAT('S','Pleas* enter the CEM4.1 Output filename:')
        READ (IOUASK, 230, END-99) C4OUT
        WRITE(IOUASK,280)
  280   FORMAT('S','Enter the MOBILE4.1 Credita filename  (if any):')
        READ(IOUASK,230,END-99) M4IMC
      ENDIF
      IF(M4IMC.EQ.'  ') M4IMC  - 'M4IMC'
      WRITE(0,390) M4IN,M4OUT,C4IN,C4OUT,M4IMC
  390 FORMAT(' I/O Filea:  ',5(/«  ',A20)/ )
C
C  Now check exiatence of chosen M41 fi  CEM41 input filea...
C
      TMPFIL - M4IN
      INQUIRE (FILE - TMPFIL, EXIST  - ISFILE)
      IF(ISFILE) THEN
        TMPFIL » C4IN
        INQUIRE(FILE - TMPFIL,EXIST - ISFILE)
C       IF(ISFILE) THEN
C         TMPFIL - M4IMC
C         INQUIRE(FILE - TMPFIL,EXIST - ISFILE)
C       ENDIF
      ENDIF
      IF(ISFILE) THEN
CC
        OPEN(UNIT-IOUGEN, FILE-M4IN,  STATUS-'OLD')
        OPEN(UNIT-7, FILE-C4IN, STATUS-'OLD')
C
C  Append if different output  files are not given
C
        OPEN (UNIT-IOUREP , FILE=M4OUT, STATUS-' UNKNOWN' ,
     *      CARRIAGE CONTROL-"FORTRAN",  POSITION-"APPEND")
        OPEN(UNIT-8, FILE-C4OUT, STATUS-'UNKNOWN',
     *      CARRIAGE CONTROL-"FORTRAN",  POSITION="APPEND")
                                                   37

-------
                                             Appendix A

      ELSE
        WRITE(0,290) TMEFIL
  290   FORMAT(' File not found:  ',A20)
        GOTO 99
      ENDIF
C
      M4CALL-M4CALL+1
C
C ***********************************
C
C  Get Coat Effective Inputs
C
      CALL GETCEI
C       	 M4MAIN ia the modified MOBILE4 driver.   It automatically
C             runs MOBILE4 four times for each evaluation date.
C             The runs are:
C
C              1. An I/M and ATE  run
C              2. An I/M run
C              3. An I/M Oeterence run
C              4. A no I/M and no ATP program run  (baseline)
C
      ICALL - 1
   10 JCALL - 1
      IF(ICALL.EQ.l) CALL MOBILE (INERR, * 98)
CC
      IF(ICALL.GT.l .AND. ICALL.LT.SO) CALL MOBNXT (INERR, * 98)
      IF (ICALL. GE. 50 .OR. INERR. GT.O) GOTO 98
      ICALL - ICALL + 1
C     CALL MOBILE(INERR,*99)
C
      WRITE(0,310) M4CALL,INERR
  310 FORMAT(/'  ','Run #',13,'  INERR -',I2/>
C
C
C  Calculate or  assign the default coat and benefit assumptions...
C
        CALL SETUP(ISIZE1,ISIZE2,INERR)
C
C  Initialize all cost and benefit arrays to zero...
C
        CALL INTWDW
C
C
C  Normalize the vehicle fleet to the first evaluation year  requested
C  by the user.  The default fleet is a total nationwide fleet baaed on
C  in-use vehicle counts done by  MVMA...
C
        CALL FLEET
C
C  Calculate the emissions in tons for a given fleet with no
C  control program in place...
C
        CALL TBASE
C
C
C  Compute fuel economy benefits  for scenarios with I/M...
C
        IF(IATPLP.NE.2) CALL CFFUEL
C
C
C  Calculate costa and benefits for each model year and vehicle
C  type in each calendar year...
C
C
        CALL LOOP(ISIZE1,ISIZE2)
                                                  38

-------
                                            Appendix  A

c
C       	 SUMWDW auma the coat and benefits over all calendar
C             yoara.
C
        CALL SUMWDW
C
C       	 Output the programs resulta.
C
        CALL PRINT(ISIZE1,ISIZE2)
C
      GOTO 10
C
C  Close any no longer needed filea...
C
   98   CLOSE(5)
        CLOSE (6)
        CLOSE(7)
        CLOSE(8)
C       CLOSE(4)
C
      IF(IFPRMT.GT.l) GOTO 50
C
   99 WRITE(0,320)
  320 FORMAT('  ','DRIVER calla completed.')
C
      STOP
C  End Main Driver
      END
                                                   39

-------
                                             Appendix A

      SUBROUTINE PPINIT
c
C  Initializes PPFIXR and PPFAIL arrays to zero, needed for MacFortran
C
C  Called by MOBILE.
C
C  Calls none.
C
C  Input on call:
C
C    common blocks:
C
C    /BYMTC2/ BYBEF4,BYTAM,BYFER
C    /BYMrC3/BYEVAP,BYRUNL,BYREFL,BYRSTL
C
C
C  Output on return:
C
C    common blocks:
C
C    /COST47/ PPFAIL(25,3,10,4),PPFIXR(25,10,4)
C
C  Notes:
C
C  PPINIT was added for CEM4.1 9/27/91 for better MacFortran use
C
C
      COMMON /COST47/ PPFAIL(25,3,10, 4) ,PPFIXR(25,10, 4)
C
C  Init PPFIXR, PPFAIL
C
      DO 50 13-1,4
        DO 50 12-1,10
          DO 50 11-1,25
            PPFIXR(II,12,13) - 0.0
            DO 50 14-1,3
              PPFAIL(II,14,12,13) - 0.0
   50 CONTINUE
C
   99 RETURN
C  End PPINIT
      END
                                                  40

-------
                                              Appendix A

      SUBROUTINE EFSAVE(ISCEN,JCYDX)
C
C  EFSAVE stores the emission factors for each CE aanario  and
C  evaluation year.
C
C  Called by MOBILE.
C
C  Calls none.
C
C  Input on call:
C
C    parameter  liat:  ISCEN,JCYDX
C    common blocks:
C
C    /BYMYC2/ BYBEF4,BYTAM,BYFER
C    /BYMYC3/BYEVAP,BYRUNL,BYREFL,BYRSTL
C
C
C  Output on  return:
C
C    common blocks:
C
C    /COST25/ FRLOSS,FREF
C    /COST26/ FRSTLS
C    /COST30/ EVAP, FER
C
C  Notes:
C
C  EFSAVE was added for CEM4.1 VERSION 2
C
C
C
      COMMON  /BYMYC2/ BYBEF4 (3, 25, 8) ,BYTAM (3,25, 4) ,BYFER<3,25, 8)
      COMMON  /BYMYC3/BYEVAP (25, 8) ,BYRUNL(25, 4) ,BYREFL(25, 4) ,BYRSTL (25, 8)
      COMMON  /COST25/ FRLOSS(25,4,4,10),FREF(25,4,4,10)
      COMMON  /COST26/ FRSTLS(25,8,4,10),YRSTLS(10),BRSTLS
      COMMON  /COST30/ EVAP(25,8,4,10),FER(3,25,8,4,10)
      COMMON  /MAXIMA/ MAXVEH,MAXLTW,MAXPOL,MAXREG,MAXYRS
      COMMON  /SFEED6/ SALHCF(25,3,8),SCFIDL(25,4),SCIADJ(25,3,8)
C
C
      DO 10 IDX-1, MAXYRS
      DO 10 IV=1, MAXVEH
        DO 20 IP-1,MAXPOL
          JDX - MAXYRS + i - IDX
C
          FER (IP, IDX, IV, ISCEN, JCYDX) - BYBEF4 (IP, JDX, IV)
          IF(IV.LE.4)
     *      FER(IP, IDX, IV, ISCEN, JCYDX)»FER(IP, IDX, IV, ISCEN, JCYDX)  +
     *      BYTAM(IP, JDX,IV)
          FER (IP , IDX, IV, ISCEN, JCYDX) -
     *      FER (IP, IDX, IV, ISCEN, JCYDX) *SALHCF (IDX, IP , IV)
C
   20   CONTINUE
C
        EVAP (IDX, IV, ISCEN, JCYDX) - BYEVAP (JDX, IV)
        FRSTLS (IDX, IV, ISCEN, JCYDX) - BYRSTL (JDX, IV)
C
   10 CONTINUE
C
C
      DO 30 IDX-1,MAXYRS
      DO 30 IV-1,4
        JDX - MAXYRS + 1 - IDX
C
        FRLOSS (IDX, IV, ISCEN, JCYDX) - BYRUNL (JDX, IV)
        FREF(IDX,IV, ISCEN, JCYDX) - BYREFL (JDX, IV)
C
  30  CONTINUE
      RETURN
C  End EFSAVE
      END
                                                    41

-------
                                              Appendix A
      SUBROUTINE SAVER (ICY)
c

C. . SAVER saves the Emission Factors and Tampering Rates generated
C . . from each of the
C. .variables.
C
C. .Called by M4MAIN
C
C. .Calls None
C
C. .Array Subscript a:
C
C. .BSIZE<4,25,2, 6,15
C. .TAMPER (25, 6, 4, 10,
C. .FER(3,25,8)
C. .TRAT(7)
C
four (or two) MOBILE 4 runs in the appropriate








,2) - BSIZE(IV,IDX,IPG,IQG,IC,IAY)
4) - TAMPER (MYDX, IP, IISCEN, ICYDX, IV)
- FER (IP, IDX, IV)
- TRAT(IEG)

C. .Variable Dictionary:
C
C. . Name Source
C BSIZE Block
C
C EFALL Block
C FER Block
C
C ICY Par am
C ICYDX Local
C IDX Local
C IF Local
C ISCEN Par am
C IV Local
C JDX Local
C
C MY Local
C MYDX Local
C TAMPER Block
C TRAT Block
C
C
C

Description
Tampering rates in each overlap section from
MOBILE4
Emission factors from MOBILE4 .
MOBILE4 array variable which originally contains
the emission factors.
Calendar year.
Absolute calendar year 1975-1.
Forward model year index (ascending model years) .
Pollutant type index (1:HC 2: CO 3:NOx).
Control program index.
Vehicle type index.
Backward model year index (descending model
years) .
Model Year.
Absolute model year index (1950-1) .
Tampering rate factors from MOBILE 4.
Intermediate array which contains the tampering
rates .


C
c
c
c
c
c
c
c
      COMMON  /COST04/ TAMPER(25,6,4,10,4)
      COMMON  /COST07/ CSTIM, GASCST, IMSTRT,ISCEN,MOVIM,MOVATP,REPIM(4)
      COMMON  /COST28/ JATP, JIM,ISTRT, JCYDX, JPRGYR, JPRSYR
      COMMON  /COST32/ BSIZE(4,25,2,6,17,2),TRAT(9)
      COMMON  /COST40/ YEVTON (10) ,BEVTON, YEREF (10) ,BEREF
      COMMON  /COST41/ YERLOS (10) , BERLOS
      COMMON  /MAXIMA/ MAXVEH,MAXLTW,MAXPOL,MAXREG, MAXYRS
DO 40 IDX - 1,MAXYRS
  JDX   - MAXYRS +  1  -  IDX
  ICYDX - ICY - 1974
  ICYDX - ICY - 1981
  MYDX  - MAXYRS +  ICYDX - JDX
  MY    - MYDX + 1950
      DO 30 IV-1,4

      TRAT (1) -BSIZE (IV, JDX, 1,1
     *        + BSIZE (IV, JDX, 1
     *        + BSIZE (IV, JDX, 1
     *        + BSIZE (IV, JDX, 1
     *        + BSIZE (IV, JDX, 1
     *        + BSIZE (IV, JDX, 1
      IF(TRAT(1) .EQ.0.0)
     * TRAT (1)-BSIZE (IV, JDX, 1,
     *         + BSIZE (IV, JDX,
     *         + BSIZE (IV, JDX,
     *         + BSIZE (IV, JDX,
     *         + BSIZE (IV, JDX,
     *         + BSIZE (IV, JDX,
      TRAT (2) -BSIZE (IV, JDX, 1, 3
     *        + BSIZE (IV, JDX, 1
      TRAT (3) -BSIZE (IV, JDX, 1, 3
     *        + BSIZE (IV, JDX, 1
     *        + BSIZE (IV, JDX, 1
     *        + BSIZE (IV, JDX, 1
                         ,1,1)+BSIZE(IV,JDX,1,1,1,2)
                         ,1,2,1)  + BSIZE(IV,JDX,1,1,2,2)
                         ,1,3,1)  + BSIZE(IV,JDX,1,1,3,2)
                         ,1,4,1)  + BSIZE (IV, JDX, 1,1, 4, 2)
                         ,1,5,1)  + BSIZE(IV,JDX,1,1,5,2)
                         ,1,8,1)  + BSIZE (IV, JDX, 1,1,8,2)

                         2,1,1)+BSIZE(IV,JDX,1,2,1,2)
                         1,2,2,1)  + BSIZE (IV, JDX, 1,2, 2, 2)
                         1,2,3,1)  + BSIZE (IV, JDX, 1,2, 3, 2)
                         1,2,4,1)  + BSIZE(IV,JDX,1,2,4,2)
                         1,2,5,1)  + BSIZE (IV, JDX, 1,2,5,2)
                         1,2,8,1)  + BSIZE (IV, JDX, 1,2,8,2)
                         ,1,1)+BSIZE(IV, JDX, 1,3,1,2)
                         ,3,9,1)  4- BSIZE (IV, JDX, 1,3,9,2)
                         , 2,1)+BSIZE(IV, JDX, 1,3,2, 2)
                         ,3,4,1)  + BSIZE (IV, JDX, 1,3, 4, 2)
                         ,3,6,1)  + BSIZE(IV, JDX,1,3,6,2)
                         ,3,10,1)  +BSIZE(IV, JDX,1,3,10,2)
                                                    42

-------
C
C
C
C
C
C
C
      TRAT(4)-BSIZE(IV,JDX,1,
     *        + BSIZE(IV,JDX,
     *        + BSIZB (IV, JDX,
     "        + BSIZE(IV,JDX,
      TRAT(5)-BSIZE(IV,JDX,2,
     *        + BSIZE (IV, JDX,
     *        + BSIZE(IV,JDX,
     »        + BSIZE(IV,JDX,
     *        + BSIZE(IV,JDX,
     *        + BSIZE (IV, JDX,
      TRAT(6) - BSIZE(IV,JDX,
      TRAT (7) • BSIZE (IV, JDX,
      TRAT(8) - BSIZE(IV,JDX,
                                          Appendix A

                           3,3,1)+BSIZE(IV,JDX,1,3,3,2)
                           1,3,5,1)  + BSIZE(IV,JDX,1,3,5,2)
                           1,3,7,1)  + BSIZE(IV,JDX,1,3,7,2)
                           1,3,11,1)  +BSIZE(IV,JDX,1,3,11,2)
                           4,1,1)+BSIZE(IV,JDX,2,4,1,2)
                           2,4,2,1)  + BSIZE (IV, JDX, 2, 4,2,2)
                           2,4,3,1)  + BSIZE(IV,JDX,2,4,3,2)
                           2,4,4,1)  + BSIZE(IV, JDX, 2,4,4,2)
                           2,4,5,1)  + BSIZE(IV,JDX,2,4,5,2)
                           2,4,8,1)  + BSIZE (IV, JDX, 2, 4,8,2)
                           1,1,13,1)
                           1,1,14,1)
                           1,1,15,1)
C

C
Store £l«at variables  for current IM program caae
TAMPER/2nd subscript:  1-air,  2-cat miaaing/no misfueling,
                       3-mi»fueled   4-pcv, 5»evap, 6=gas cap
   TAMEER(IDX,1,ISCEN,JCYDX,IV)-THAT(1)
   TAMPER(IDX,2,ISCEN,JCYDX,TV)=TRAT(2)
   TAMPER(IDX, 3, ISCEM, JCYDX, IV) -THAT (3) +TRAT (4)
   TAMPER (IDX, 4 , ISCEN, JCYDX, IV) -TRAT (7 )
   TAMPER(IDX,5,ISCEN,JCYDX,IV)-TRAT(6)
   TAMPER (IDX, 6, ISCEN, JCYDX, IV) -TRAT (8)

30 CONTINUE
40 CONTINUE

99 RETURN
End SAVER
   END
                                                    43

-------
                                             Appendix  A
c
c.
c
c
c
c
c.
c
c.
c
c.
c
c.
c
c.
c
c.
c
c
c
G
c
c
c
c
c
c
G
c
c
c
c
c
c
c
c
c
c
c
c
c
c
    SUBROUTINE ATPSAV(ICY)

.ATPSAV ia called if an ATP only program is to be modeled.   It
 allows M4MAIN to be run only twice and atorea the EFALL  and
 TAMPER array elementa for the no control program situation and
 the ATP only program.

.Called by M4MAIN

.Calls None
.Array Subscript:

.TAMPER(25,6,4,10,4)

.Varialbe Dictionary:'
- TAMPER (IDX, IVTAM, IISCEN, JCYDX, IV)
  Name
          Source   Deacription
 ICY      Param    Calendar year.
 ICYDX    Local    Absolute calendar year index 1975—1.
 IDX      Local    Forward model year index  (ascending model
                   yeara) .
 IISCEN   Local    Mobile3 run index(1:No program, 2:Deterrence,
                   3:I/M only, 4:I/M;ATP).
 IP       Local    Pollutant type index  (1:HC 2:CO,3:NOx).
 IV       Local    Vehicle type index.
 IVTAM    Local    Emission control device tampering type index
                   (1:AIR,2:CAT,3:MISF,4:EVAP 5:PCV).
 JDX      Local    Backward model year index  (descending model
                   yeara).
 MY       Local    Model year
 MYDX     Local    Absolute model year index  (1950=1).
 TAMPER   Block    Tampering rate factora from MOBILE4.


    COMMON /COST04/ TAMPER(25,6, 4,10,4)
    COMMON /COST28/ JATP, JIM, ISTRT, JCYDX, JPRGYR, JPRSYR
    COMMON /COST30/ EVAP(25,8,4,10),FER(3,25,8,4,10)
    COMMON /COST25/ FRLOSS(25,4,4,10),FREF(25,4,4,10)
    COMMON /COST26/ FRSTLS(25,8,4,10),YRSTLS(10),BRSTLS
    COMMON /COST40/ YEVTON(10),BEVTON,YEREF(10),BEREF
    COMMON /COST41/ YERLOS (10) , BERLOS
    COMMON /MAXIMA/ MAXVEH,MAXLTW,MAXPOL,MAXREG, MAXYRS

    DO 40 IDX - 1,MAXYRS
      JDX    = MAXYRS + 1 - IDX
      ICYDX  - ICY - 1974
      ICYDX  - ICY - 1981
      MYDX   - MAXYRS + ICYDX - JDX
      MY     - MYDX + 1950
      DO 40 IV - 1,8
        IF(TV .GT. 4) GOTO 20
          DO 10 IVTAM - 1,5
          DO 10 IISCEN "1,2
            TAMPER (IDX, IVTAM, IISCEN, JCYDX, IV) -
   *           TAMPER (IDX, IVTAM, 3, JCYDX, IV)
 10       CONTINUE

 20     DO 30 IISCEN -1,2
          DO 25 IP - 1,2
            FER (IP, IDX, IV, IISCEN, JCYDX) -FER (IP, IDX, IV, 3, JCYDX)
 25       CONTINUE
          EVAP (IDX, IV, IISCEN, JCYDX) -EVAP (IDX, IV, 3, JCYDX)
          FREF (IDX, IV, IISCEN, JCYDX) -FREF (IDX, IV, 3, JCYDX)
          FRLOSS (IDX, IV, IISCEN, JCYDX) -FRLOSS (IDX, IV, 3, JCYDX)
          FRSTLS (IDX, IV, IISCEN, JCYDX) -FRSTLS (IDX, IV, 3, JCYDX)
 30     CONTINUE

 40 CONTINUE

 99 RETURN
 End ATPSAV
    END
                                                   44

-------
                                             Appendix A

      SUBROUTINE RESET
C
C..RESET resets the IMFLAG  and ATPFLG for each MOBILE4 call in M4MAIN.
C..NOW also resets IPRGYR & IPRSYR
C
C..Called by M4MAIN
C
C..Calls None
C
C..Array Subscripta:
C
C. .AER(11,11,2,2)              -  AER(I4,I3,I2,I1)
C..BSIZE(4,25,2,6,15,2)        -  BSIZE (11,12,13,14,15,16)
C. .SAER(11,11,2,2)             -  SAER(I4,I3,I2,I1)
C..SEVP(2,2)                   -  SEVP(2,2)
C. .VA023(11,11,2,2)            -  VA023(I4,I3,I2,I1)
C..VAO24(2,2)                  -  VAO24(I2,I1)
C
C..Variable Dictionary:
C
C.. Name    Source    Description
C   	    	    	
C  AER      Block     The ATP effectiveness rates  for non-evaporative
C                     exhaust emission tampering.
C  ATPFLG   Block     The anti-tampering flag  
-------
                                             Appendix A

      MODYRl = JMYRl
      MODYR2 - JMYR2
C  Now set per scenario
      IF(ISCEN.EQ.4) THEN
        ATPFLG = JATP
        IF (JIM.EQ. 2) IMSTRT - ISTRT
        IPRGYR - JPRGYR
        IPRSYR - JPRSYR
      ELSE IF(ISCEN.EQ.3) THEN
        IF(JIM.EQ.2) IMSTRT = ISTRT
CC
C      ELSE IF(ISCEN.EQ.2) THEN       I No changesB needed.
C       MODYRl - 2020
C       MODYR2 - 2020
CC
      ELSE IF(ISCEN.EQ.l) THEN
C       IMFLAG - 1
        MODVR1 » 2020
        MODYR2 - 2020
      ENDIF
C
   99 RETURN
C  End RESET
      END
                                                   46

-------
                                             Appendix A
      SUBROUTINE SETUP(ISIZE1,ISIZE2,INERR)
C
C..SETUP defines moat of the default coat and benefit assumptions
C..used in the model.
C
C..Called by MAIN
C
C..Calls QUITER
C
C..Variable Dictionary
C
C..  Name    Source   Description
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
ATCOST  Block    Cost of Replacing emission control devices as a
                 result of failing a tampering inspection. The
                 default assumptions are:  ATCOST(TECH,TYPE)
                 TYPE =• 1 : Air Pump System
                      - 2 : Catalyst Removal
                      • 3 : Misfueled Catalyst
                      » 4 : Evaporative Control System
                      = 5 : FCV System
                      — 6 : Gas Cap
ATPFEE  Block    Anti-tampering inspection fee
BYEAR   Block    Fleet size and growth base year default - 1986
CSTIH   Block    Inspection Maintenance I/M inspection fee
FFACT   Block    Fraction of annual fuel economy benefit achieved
                 by biennial I/M  (based on HC emission reduction)
GASCST  Block    Cost of one gallon of unleaded gasoline
IAFEFG  Block    Flag indicating whether user is inputting
                 optional ATP fees and costs or using the defaults
                 1 = defaults   2 = user inputs
IAFXFG  Block    Flag indicating whether the ATP has a fixed or
                 floating model year coverage
lAIiDV   Local    Variable indicating which vehicle classes are
                 to be inspected by ATP. Must be the same as
                 those inspected by I/M.
IATPLP  Block    Flag indicating whether an ATP only scenario ia
                 to be run
IFREQ   Block    Flag indicating whether the I/M program ia
                 annual or biennial
IGRFLG  Block    Flag indicating whether user ia inputting the
                 vehicle fleet distribution and the fleet growth
                 rate or is using the built-in defaults
ILDT    Block    Parameter data from MOBILE4 indicating the
                 number of vehicle classes inspected by I/M
IMCSFG  Block    Flag indicating whether user ia inputting
                 optional I/M repair coat and fees or ia using
                 the defaults
IMFXFG  Block    Flag indicating whether the I/M program has a
                 fixed or floating model year coverage
INERR   Param    MOBILE4 error counter in subroutine QUITER
ISIZE1  Block    Variable containing either the aize of the
                 floating model year window for I/M coverage or
                 the earliest model year inspected in an I/M
                 program
ISIZE2  Block    Variable containing either the size of the
                 floating modely year window for I/M coverage or
                 the earliest model year inspected in an I/M
                 program
DISTYP  Block    Array of flags indicating which emission
                 control components are to be inspected by the ATP
TVTYP   Local    Vehicle class index for all eight vehicle classes
ATPFQT   Block    Flag indicating whether ATP program is annual
                 or biennial
LAP1ST  Block    Variable containing the oldest model year
                 covered by an ATP if a fixed model year program
                 is run
MODYR1  Block    Variable containing the oldest model year
                 covered by I/M if a fixed model year program ia
                 run
MOVATP  Block    Number of model years covered in an ATP if a
                 floating model year program is run
MOVIM   Block    Number of model years covered in I/M if a.
                 floating model year program is run
REPEAT  Block    Fraction of I/M failures which fail in the
                 subsequent year also
REPIM   Block    Array containing the I/M repair costs
                  REPIM(l) - Old Tech  ;  REPIM(2) - New Tech
                  REPIM(3) - IM240 ; REPIM(4) = NOx (IM240)
RMAXF   Block    Maximum I/M failure rate that 1981 and later
                 vehicles are allowed
                                                  47

-------
                                             Appendix  A

C   TCOUNT  Block    Vehicle fleet size in base year
C   VGRORT  Block    Array containing the growth rates for each
C                    vehicle class
C
      INTEGER BYEAR,ATPFLG,PROMPT,ATPFQT,DISTYP
C
C     	 COMMON BLOCK DEFINITIONS
C
      COMMON /COST01/ ATPFEE
      COMMON /COST02/ IFEVYR, ILEVYR
      COMMON /COST03/ STRNGY,RMAXF,REPEAT, FFACT
      COMMON /COST07/ CSTIM,GASCST,IMSTRT, ISCEN,MOVIM,MOVATP,REPIM(4)
      COMMON /COST12/ IMFXFG, IAFXFG, IMCSFG, IAFEFG, IALDV
      COMMON /COST22/ ATCOST(2,8)
      COMMON /COST27/ IGRFLG, IATPLP, IDIFF,OUTPRF
      COMMON /COST34/ TCOUNT,BYEAR, VGRORT
      COMMON /COST48/ PRGFEE,PRSFEE,CEVFEE(25,10,4),TRNFEE
C
      COMMON /COSTXX/ JMYR1,JMYR2
C
      COMMON /ATPAR1/ LAPSY,LAP1ST,LAPLST,LVTFLG(4)
      COMMON /FLAGSl/ PROMPT, TAMFLG, SPDFLG, VMFLAG, OXYFLG, DSFLAG
      COMMON /FLAGS2/ MYMRFG,NEMFLG,IMFLAG,ALHFLG
      COMMON /ATPAR2/ ATPPGM, ATPFQT, CRATP, DISTYP (8)
      COMMON /IMPAR1/ ICYIM, ISTRIN,MODYR1,MODYR2,WAIVER<2) ,CRIM
      COMMON /IMPAR6/ IFREQ,INTYP
C
CC
      MODYR1 - JMYR1
      MODYR2 - JMYR2
CC
C  I/M program defaults
C
      RMAXF  - .25
      REPEAT - .20
C
C.. FFACX was 0.70 in CEM4,  changed 8/14/91
C
      FFACT  - .90
C
C  Default I/M coat assumptions:
C
      IF(IMCSFG.EQ.2) GOTO 10
      IF(IFREQ.EQ.l)  CSTIM - 10.00
      IF(IFREQ.EQ.2)  CSTIM - 12.00
          GASCST   -   1.25
          REPIM(l) -  50.00
          REPIM(2) -  75.00
          REPIM(3) - 150.00
          REPIM(4) - 100.00
C
C  Default Purge/Pressure Inspection costs...
C
          PRGFEE -  6.53
          PRSFEE -  0.69
C   TRNFEE is just increment  for IM240 over Purge insp fee.   6/24/91
          TRNFEE -  0.67
C
C  Default ATP cost assumptions:
C
   10 IF 
-------
                                          Appendix A

ATPFEE ia the inspection  coat  for the ATP .
The default assumes  any underhood inapection will cost
$1.00 regrardleaa  of the  number of items inspected,  a
lead deposit teat  will coat  an additional $0.50,  and an exterior
vehicle inapection of either catalyst,  fuel inlet reatrictor,
or gaa cap will  coat an additional $0.25. Max cost of $1.75.
20 IF(IAFEFG.EQ.2)
                                GOTO 30
                                ATPFEE -0.0
IF(
*
*
*
*
IF(
*
*
*
IF(
DISTYP(l)
DISTYP(S)
DISTYPJ6)
DISTYP(7)

DISTYP(2)
DISTYP(3)
DISTYP (8)

DISTYP(4)
.EQ.
.EQ.
.EQ.
.EQ.

.EQ.
.EQ.
.EQ.

.EQ.
2
2
2
2 )

2
2
2 )

2 )
.OR.
.OR.
.OR.


.OR.
-OR.



C
C
C
C
C
C
C
C
C
C
C
                                ATPFEE - 1.00



                                ATPFEE - ATPFEE + 0.25
                                ATPFEE = ATPFEE + 0.50

 The  default  fleet distribution and growth rates by vehicle type.

 30 IF(IGRFL6.EQ.2)  GOTO 45

. . default  growth rate used to be 2.0

   VGRORT  -  0.0
   BYEAR
   TCOONT
             1986
             1000000.
 This block determines whether a fixed or floating
 program ia to be run and transfers the model year
 window aize or program start date.
         IMFXFG =• 1
         IAFXFG - 1
         IMFXFG - 2
         IAFXFG - 2
                       Moving Model Year Coverage

                       Fixed Model Year Coverage
 45 IF (IMFXFG. EQ.l)  ISIZE1 = MOVIM
    IF ( IAFXFG. EQ.l)  ISIZE2 - MOVATP
    IF (IMFXFG. EQ.2)  ISIZE1 - MODYR1
    IF (IAFXFG. EQ.2)  ISIZE2 - LAP1ST
 99 RETURN
 End SETUP
    END
                                                49

-------
                                             Appendix  A
      SUBROUTINE INTWDW
C
C.
C
C.
C
C.
C
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C
C
C
C.
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
.INTWDW initializes window scenario array elements to  zero

.Called by MAIN
C
C
C
C
.Array Subscripta

.CATPFE(25,10, 4)
.CATPRE (25,10, 4)
.CIMFEE(25,10,4)
.CIMREP (25,10, 4)
.COST(25,10,4)
.DETER(25,2,10,4)
.FULBEN(25,10,4)
.REDATP(25,2,10,4)
.REDIM(25,2,10,4)
.TDETER<25,2,10,4)
.TONRED(25,2,10,4)
.TOTCST(2)
.TOTRED (2)
.Name   Type
    CATPFE (IDX, JCYDX, IV)
    CATPRE (IDX, JCYDX, IV)
    CIMFEE (IDX, JCYDX, IV)
    CIMREP (IDX, JCYDX, IV)
    COST(IDX,JCYDX,IV)
    DETER (IDX, IP, JCYDX, IV)
    FULBEN (IDX, JCYDX, IV)
    REDATP (IDX, IP, JCYDX, IV)
    REDIM (IDX, IP, JCYDX, IV)
    TDETER (IDX, IP , JCYDX, IV)
    TONRED (IDX, IP, JCYDX, IV)
    TOTCST(IP)
    TOTRED(IP)
Dictionary
 CATPFE Block  Array which holds and sums the ATP fees for each
               model year, calendar year and vehicle type
 CATPRE Block  Array which holds and sums the ATP repair costs
               for each model year, calendar year and vehicle type
 CIMFEE Block  Array which holds and sums the I/M fees for each
               model year, calendar year and vehicle type
 CIMREP Block  Array which holds and sums the I/M repair coats
               for each model year, calendar year and vehicle type
 COST   Block  Array which holds and sums the total cost
               for each model year, calendar year and vehicle type
 DETER  Block  Array which holds and sums the deterence benefit
               for each model year, calendar year and vehicle type
 FULBEN Block  Array which holda and sums the fuel economy benefit
               for each model year, calendar year and vehicle type
 ICYDX  Local  Calendar year index    1 - 1975
 IP     Local  Pollutant index  1 - HC   2 - CO
 IV     Local  Vehicle class index
 MYDX   Local  Model year index       1 - 1950
 REDATP Block  Array which holds and sums the ATP benefit for each
               model year, calendar year and vehicle type
 REDIM  Block  Array which holds and sums the I/M benefit for each
               model year, calendar year and vehicle type
 TCOST  Block  Array which passes the total coat of control programs
 TDETER Block  Array which passes the total deterence benefit
 TONRED Block  Array which passes the total benefit
 TOTCST Block  Array which passes the total cost of control programs


    COMMON /COST10/ FEBNFT(25,10,4),FECNMY(20,4)
    COMMON /COST13/ CIMFEE(25,10,4),CIMREP(25,10,4)
    COMMON /COST14/ TONRED(25,2,10,4),COST(25,10,4)
    COMMON /COST15/ TOTCST(2),TOTRED(2),TCOST
    COMMON /COST16/ DATPFE, DIMFEE, DIMREP, DATPRE, AATPFE, AIMFEE
    COMMON /COST17/ DETER(25,2,10,4),TDETER(25,2,10,4)
    COMMON /COST18 / AIMREP , AATPRE , CFACTR, CATPFE (25,10,4)
    COMMON /COST21/ CATPRE(25,10,4),FULBEN(25,20,4)
    COMMON /COST20/ TDUCT(25,2,3,10,4),TEVAP(25,10,4)
    COMMON /COST46/ REDIM(25,2,10,4),REDATP(25,2,10,4)
    COMMON /COST47/ PPFAIL(25,3,10,4),PPFIXR(25,10,4)
    COMMON /COST49/ PPFEB(25,10,4),APFBEN

    COMMON /COST50/ TRLOSS(25,4,3),TREVAP(25,4,3)
    COMMON /COST50/ TRLOSS(25,4,3),TREVAP(25,4,3),TRSTLS(25,4,3)

    COMMON /MAXIMA/ MAXVEH, MAXLTW, MAXPOL, MAXREG, MAXYRS
      TCOST-0.0

      DO 10 IP=1,2
       TOTCST(IP)=0.0
       TOTRED(IP)-0.0
   10 CONTINUE
                                                   50

-------
                                             Appendix A
      DO  20  IV-1,4
       DO 20 JCYDX-1,10
          DO 20 IDX - 1,MAXYRS
            COST(IDX,JCTDX,IV)-0.0
            CIMFEE(IDX,JCYDX,IV)-0.0
            CATPFE(IDX,JCYDX,IV)-0.0
            CIMREP(IDX,JCYDX,IV)-0.0
            CATPRE(IDX,JCYDX,IV)-0.0
            POLBEN(IDX,JCYDX,IV)-0.0
            FEBNFT (IDX, JCYDX, IV) -0 . 0
            PPFEB(IDX,JCYDX,IV)-0.0
            TEVAP(IDX,JCYDX,IV)-0.0
            DO 20 IP-1,2
              TOMBED(IDX,IP,JCYDX,IV)-0.0
              BEDIM(IDX,IP,JCYDX,IV) -0.0
              REDATP(IDX,IP,JCYDX,IV)-0.0
              DETER(IDX,IP,JCYDX,IV) -0.0
              TDETER(IDX,IP,JCYDX,IV)-0.0
              DO 20 IMC-1,3
                TDOCT(IDX,IP,IMC,JCYDX,IV)-0.0
C               TRSTLS(IDX,IV,IMC)-0.0
                TREVAP(IDX,IV,IMC)-0.0
                TRLOS8(IDX,IV,IMC)-0.0
   20 CONTINUE
C
      RETURN
C  End INTHDW
      END
                                                   51

-------
                                             Appendix A
c
c
c.
c.
c.
c
c.
c
c.
c
c.
c.
c.
c
c.
c
c.
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c

c
c
      SUBROUTINE FLEET
.FLEET determines the number of vehicles in each vehicle
.class in each calendar year.  FLEET also sums the total
.number of vehicles in each calendar year.


.Called by MAIN
.Array Subscripts


.STOCK(12,8)
.VCNT (46, 8)
.VGRORT(8)


.Variable Dictionary
               - STOCK(JBY,IV)
               - VCNT(IY,IV)
               - VGRORT(IV)
  Name
          Source   Description
C
C
C


C
C
C
C
C
  ACOUNT  Faram
  BCOUNT
  BYEAR
  CLDIST
  ICYDX
  rv
  JEXP
  NCOUNT
  TCOUNT
  VCNT


  VGRORT


  YCOUNT
Local
Par am
Block
Local
Local
Local
Block
Block
Block


Block


Block
The average number of vehicles over  all  the
calendar years selected in the scenario  records.
Sum of MOBILE VCOUNT for 1st eval year
The fleet size base year
Vehicle claas registration distribution
The absolute calendar year index 1 - 1975
Vehicle class index
Exponent index
Desired vehicle fleet size in 1st evaluation year
Vehicle fleet size in base year
Number of vehicles of each type in each
calendar year
Array containing the fleet growth rates  for each
vehicle claas
The total number of vehicles in a given  calendar
year
    INTEGER BYEAR
    COMMON /COST02/ IFEVYR, ILEVYR
    COMMON /COST23/ VCNT(39,8),GSDSCT(8,10),HDDCUM(25,10)
    COMMON /COST34/ TCOUNT,BYEAR,VGRORT
    COMMON /COST37/ YRCOST(10),YRBEN(10,2),YCOUNT<39),ACOUNT

    COMMON /REGISF/ GSFVCT(8),USRGSF(25,2)
    COMMON /VMXCOM/ REGMIX (8) , TFNORM (8) , VMTMIX (8) , VCOUNT (39, 8)


    DIMENSION OLDCNT (39)


 Determine base year index


    IBY - BYEAR - 1981


 Adjust YCOUNT from given base year to current IFEVYR...
 Leave TCOUNT as the base year count
 and use YCOUNT as the new count for current eval year


    DO 20 ICYDX - 1,39
      YCOUNT(ICYDX) - TCOUNT
      IF(VGRORT.NE.0.0 .AND. ICYDX.NE.IBY) THEN
        IF(ICYDX.LT.IBY)  THEN
          JEXP - IBY - ICYDX
          YCOUNT (ICYDX) - TCOUNT / ((1.0+ (VGRORT/100 .) ) ** JEXP)
        ELSE IF (ICYDX.GT.IBY)  THEN
          JEXP = ICYDX - IBY
          YCOUNT(ICYDX) - TCOUNT * ((1.0+(VGRORT/100.))**JEXP)
        ENDIF
      ENDIF
      OLDCNT(ICYDX) - 0.0
      DO 10 IV-1,8
        VCNT(ICYDX,IV)  - VCOUNT(ICYDX,IV) * GSFVCT(IV)
        OLDCNT (ICYDX) - OLDCNT (ICYDX)  + VCNT (ICYDX, IV)
 10   CONTINUE
                                                   52

-------
                                            Appendix  A

G  normalize VCNT by Desired YCOUNT/(Old VCOUNT from M.41)
C
      DO 50 ICYDX - 1,39
        ADJCNT - YCOONT(ICYDX)/OLDCHT(ICYDX)
        DO 40 IV - 1,8
          VCNT (ICYDX, IV)  - VCHT (ICYDX, TV) *ADJCNT
   40   CONTINUE
   50 CONTINUE
C
C  Sum up YCOUNT and ACOUNT...
C
  105 IFICY - IFEVYR - 1981
      ILICY = ILEVYR - 1981
C
      ACOUNT-0.0
      DO 120 ICYDX=IFICY,ILICY
        ACOUNT - ACOUNT + YCOUNT(ICYDX)
  120 CONTINUE
C
      ACOUNT = ACOUNT /  ( ILEVYR - IFEVYR 4- 1 )
C
  999 RETURN
C  End FLEET
      END
                                                  53

-------
                                             Appendix  A

      SUBROUTINE TEASE
c

C..TBASE calculate the non-program emission levels for each
C . . calendar year .
C
C. .Called by MAIN
C
C . . Calla CAHIL
C
C. .Array Subscripts :
C
C. .YTONS(2,10)
C. . JULMYR(25,8)
C. .BTONS(2)
C. .VCNT(39,8)
C








- YTONS (IP, JCYDX)
- JULMYR (JDX, IV)
- BTONS (IP)
- VCNT( ICYDX, IV)

C. .Variable Dictionary:
C
C. . Name Source
C ACCMIL Param
C
C BTONS Block
C
C EFALL Block
C ICYDX Local
C IDX Local
C IFEVYR Local
C IFICY Local
C ILEVYR Local
C ILICY Local
C IP Local
C IV Param
C JDX Param
C
C JULMYR Block
C
C MY Local
C MYDX Local
C NUMCY Local
C VCHT Block
C
C VMT Param
C YTONS Block
C
C

Description
The average accumulated mileage in a one year
period.
Average benefits over all evaluated calendar
years .
Emission factors from MOBILE4.
Absolute Calendar year index (1975-1) .
Forward model year index (ascending model years)
First calendar year evaluated.
First evaluated calendar year index (1975-1) .
Last calendar year evaluated.
Last evaluated calendar year index (1975 -1) .
Pollutant type index (1:HC, 2: CO, 3:NOx) .
Vehicle type index.
Backward model year index (descending model
year) .
Fraction of vehicles in each model year from
MOBILE4 .
Model year.
Absolute model year index (1955—1) .
The number o£ calendar years being evaluated.
Total number of vehicles in a vehilce class in a
calendar year.
Total vehicle miles traveled in a calendar year.
Total benefits in each evaluted calendar year.


      REAL JULMYR
      COMMON  /COST02/  IFEVYR,ILEVYR
      COMMON  /COST04/  TAMPER(25,6,4,10,4)
      COMMON  /COST23/  VCNT'(39, 8) ,SSDSCT (8,10) ,HDDCUM(25,10)
      COMMON  /COST25/  FRLOSS(25,4,4,10),FREF(25,4,4,10)
C. . .
      COMMON  /COST26/  FRSTLS(25,8,4,10),YRSTLS(10),BRSTLS
      COMMON  /COST27/  IGRFLG,IATPLP, IDIFF, OUTPRF
C
      COMMON  /COST28/  JATP, JIM, ISTRT, JCYDX, CPR6YR, JPRSYR
C   Add AEVAP 6/3/91
      COMMON  /COST36/  YTONS(2,10),BTONS(2),BIDTON(2),AEVAP
      COMMON  /COST30/  EVAP(25,8,4,10),FER(3,25,8,4,10)
      COMMON  /COST40/  YEVTON(10),BEVTON,YEREF(10),BEREF
      COMMON  /COST41/  YERLOS(10),BERLOS

      COMMON  /CUMCOM/  CUMMIL(25,8)

      COMMON  /MYRSAV/  AMAR(25, 8) , JULMYR(25, 8) ,NEWCUM
      COMMON  /MAXIMA/  MAXVEH,MAXLTW,MAXPOL,MAXREGrMAXYRS
      COMMON  /REGISF/  GSFVCT (8) , USRGSF (25, 2)
C
      DIMENSION YHCFTP(10)
C
C     IFICY = IFEVYR - 1974
      IFICY - IFEVYR - 1981
C     IDIFF - ILEVYR - IFEVYR +  1
C
C     	 Compute the number of  calendar years analyzed.
C
C
      DO 5 IPP - 1,2
        BTORS(IPP) =0.0
        BIDTON(IPP) =0.0
    5 CONTINUE
                                                   54

-------
                                              Appendix A

        BEVTON  - o.o
        BEREF   =0.0
        BERIiOS  -0.0
C. . .
        BRSTLS  =0.0
        BFTP    — 0.0
C
      DO 30 JCYDX - 1,IDIFF
C
C       ICYDX - (IFEVYR-1974) + JCYDX - 1
        ICYDX - (IFEVYR-1981) + JCYDX - 1
C
        YTONS(1,JCYDX)  - 0.0
        YTONS(2,JCYDX)  - 0.0
        YEVTON (JCYDX)  - 0.0
        YEREF (JCYDX)    - 0.0
        YERLOS(JCYDX)  =0.0
        YRSTLS(JCYDX)  - 0.0
C... add 8/12/91...
        YHCFTP(JCYDX)  =0.0
C
        DO 20 IV  = 1,8
C. . .
          DO 10 IDX =• 1,MAXYRS
            JDX = MAXYRS + 1 - IDX
C           MYDX  = MAXYRS + ICYDX - JDX
C           MY      MYDX + 1950
C
C      	 Determine the vehicle miles traveled (VMT)
C    VCNT now already has GSFVCT incorporated into it...
C
            CALL  CALMIL(ODX,IV,ACCMIL,VMT)
C
            VADJ  = VCNT (ICYDX, IV) *JOLMYR(JDX, IV) *1.1025E-6*.001*VMT
C
          TMP-VADJ* (EVAP (IDX, IV, 1, JCYDX) +EVAP (IDX, IV, 1, JCYDX+1) ) /2. 0
              YEVTON (JCYDX) = YEVTON (JCYDX)  + IMP
C    *           (EVAP (IDX, IV, 1, JCYDX) +EVAP (IDX, IV, 1, JCYDX+1) ) /2 *
C    *                           VADJ
              BEVTON        = BEVTON + IMP
C    *           (EVAP (IDX, IV, 1, JCYDX) +EVAP (IDX, IV, 1, JCYDX+1) ) /2 *
C    *                           VADJ
C. . .
          TMP=VADJ* (FRSTLS (IDX, IV, 1, JCYDX) +FRSTLS (IDX, IV, 1, JCYDX+1) ) /2 . 0
              YRSTLS (JCYDX) - YRSTLS (JCYDX)  + IMP
C
              BRSTLS        = BRSTLS + TMP
C
          TMP-VADJ* (FER (1, IDX, IV, 1, JCYDX) +FER(1, IDX, IV, 1, JCYDX+1) ) /2 . 0
              YHCFTP (JCYDX) = YHCFTP (JCYDX)  + TMP
C
              BFTP          = BFTP + TMP
C
          TMP-VADJ*(FER(2,IDX,IV,1,JCYDX)+FER(2,IDX,IV,1,JCYDX+1))/2.0
              YTONS(2, JCYDX)- YTONS(2,JCYDX)  + TMP
C
              BTONS(2)       - BTONS(2) + TMP
C
        IF(IV.LE.4) THEN
          TMP-VADJ* (FREF (IDX, IV, 1, JCYDX) +FREF (IDX, IV, 1, JCYDX+1)) /2 . 0
              YEREF(JCYDX)  - YEREF (JCYDX) + TMP
C
              BEREF         - BEREF + TMP
C
          TMP-VADJ* (FRLOSS (IDX, IV, 1, JCYDX) +FRLOSS (IDX, IV, 1, JCYDX+1) ) /2 . 0
                YERLOS (JCYDX) - YERLOS (JCYDX)  + TMP
C
                BERLOS        = BERLOS + TMP
C
        ENDIF
C
   10 CONTINUE
C
   20 CONTINUE
C. . .
              YTONS(1,JCYDX) - YHCFTP(JCYDX)
     *                         + YEVTON (JCYDX)  + YEREF (JCYDX)
     *                         + YERLOS(JCYDX)  + YRSTLS(JCYDX)
C
   30 CONTINUE
C


                                                    55

-------
                                            Appendix  A


            BTONS(l)  = BFTP + BEVTON + BEREF + BERLOS + BRSTLS

            DO 40 1=1,2
              BTONS(I) - (BTONS(I) / IDIFF)
   40       CONTINUE

            BEVTON = BEVTON / IDIPF
            BEREF  = BEREF  / XDIFF
            BERLOS = BERLOS / IDIFF

            BRSTLS - BRSTLS / IDIFF
            BFTP   - BFTP   / IDIFF
C
C
   99 RETURN
C  End TBASE
      END
                                                  56

-------
                                             Appendix A
c
c.,
c.,
c.,
c.,
c.
c.
c.
c.
c
c.
c.
c.
c.
c.
c
c.
c
c.
c
c.
c
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c
c.
c
c.
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
   SUBROUTINE CPFUEL


CPFUEL computes  the  fuel  economy benefits due to car
repairs.  This routine  is called only if an I/M
program is  selected.  If this ia the case, this routine
calculates  the fuel  savings accumulated from I/M
repairs over the life of  the I/M program. It is assumed
that a given vehicle receives only one fuel economy
benefit over its entire life and that once received
this benefit ia  permanent.


Because this subroutine may have the largest calendar
year coverage  (the ICYPX  looping begins at the I/M
start year  not the first  scenario year (IFEVYR)),  the
I/M failure rate calculation ia performed in this
, subroutine.


Called by MAIN


.Calls CALMIL
-Array Subscripts:


 DRSTRN(3)
,FAILRT(25,10,4)
,FEBNFT(25,10,4)
.FECNMY(20,4)
.JULMYR(25,8)
,NEWSTR(4,100)
,PCNTSV(4)
,PPFEB(25,10,4)
.WAIVER (2)

, ZMSTRN(3)


.Variable Dictionary
- DRSTRN (ITEST)
 - FAILRT(IDX,JCYDX,IV)
  - FEBNFT(IDX,ICYDX,IV)
  - FECNMY (IDX, IV)
  - JULMYR(IDX,IV)
  - NEWSTR(4, IM)
  - PCNTSV(ITECH+IM24+P/P)
  - PPFEB(IDX,ICYDX,IV)
  - WAIVER (ITECH)
- ZMSTRN(ITEST)
  Name
          Source    Description
  ACCMIL  Param     Accumulated mileage at a given age for each
                    vehicle class
  DRSTRN  Block     Growth in the I/M failure as a function of
                    mileage for new tech vehicles
  FAIL    Block     Calculated I/M failure rate (1-HC/CO,  2-NOx)
  FAIL1   Local     I/M failure rate for next year
  FAIL2   Local     I/M failure rate for current year among cars
                    which were OK after last year's repairs
  FEBNFT  Block     Final fuel economy benefit for exhaust repairs
  FECARS  Local     Number of cars for which fuel economy credits
                    should be claimed
  FECNMY  Block     Fuel economy of each vehicle type for each
                    model year in miles per gallon
  FFACT   Block     Fraction of annual fuel economy benefit achieved
                    by biennial I/M (based on HC emission reduction)
  GASCST  Block     Cost of one gallon of unleaded gasoline
  ICUTS   Block     I/M outpoint index from MOBILE4
  ICYDX   Local     Calendar year index  1 - 1975
  ICYDX2  Local     Subsequent calendar year index ICYDX2-ICYDX + 1
  ICY     Local     Calendar year
  IDX     Local     Model year index (forward counter 1 - 25)
  IFEVYR  Block     First calendar year evaluated
  IFICY   Local     First calendar year index
  IFREQ   Block     Flag indicating whether the I/M program is
                    annual or biennial
  ILEVYR  Block     Last calendar year evaluated
  ILICY   Local     Last calendar year index
  IMICY   Local     Calendar year index in which I/M starts.
                    Only used if I/M atart precedes first calendar
                    evaluated.
  ISTRIN  Block     Number of input stringency groups
  ISTRT   Block     I/M start year
  ITECH   Block     Technology group  1 = old tech ,  2 - new tech
                     Old - myr 1951-1980 LDGV ,  myr 1951-1983  LDT's
                     New - myr 1981+     LDGV ,  myr 1984+     LDT's
  ITEST   Block     I/M test type from MOBILE4
  IV      Local     Vehicle class index
  JDX     Local     Model year index (backwards counter 20 - 1)
  MY      Local     Model year
  MYDX    Local     Model year index
  NEWSTR  Block     Array containing information read from the
                    stringency records.
                        NEWSTR(1,IM)   1 = Vehicle class
                                                   57

-------
                                             Appendix A
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c

c
c
c
                      NEWSTR(2,IM)  2 - First model  year in group
                      NEWSTR(3,IM)  3 - Last model year in group
                      NEWSTR(4,IM)  4 - Group's  stringency
NMYF    Local     Firat model year in group  (from NEWSTR(2,IM)
NMYLL   Local     Last model year in group  (from NEWSTR(3,IM)
NRPEAT  Local     Cars which are not part of the REPEAT failure
                  category
PCNTSV  Block     Average fuel savings by technology group
PPCARS  Local     Number of cars for which Purge/Pressure repair
                  fuel economy credits should be claimed
PPFEB   Block     Fuel economy benefit for Purge/Press repairs
PREVCY  Block     Values used in calculating the previous
                  calendar year's fuel economy cars.   Used to
                  calculate next year's cars
PREVNF  Block     Cars repaired in previous years which have not
                  yet failed again
REPEAT  Block     Fraction of I/M failures which fail in the
                  subsequent year also.
RMAXF   Block     Maximum failure rate that 1981 and later
                  vehicles are allowed
VMT     Param     Vehicle miles travelled in a given year
WAIVER  Block     Array containing the waiver rates  for old
                  and new technology vehicles
ZMSTRN  Block     Failure rate of new technology vehicles at
                  zero miles.

NOTES: EVPFEB added 7/11/91 could be done away with  and the
       results simply added in with FEBNFT if we didn't want
       to see it separately in the printout.
  9/8/91  Add NOx to FAIL array FAILRT(IPOL,ODX,ICYDX,IV),  where
  IPOL: HC,CO-1, NOx-2


  COMMON /COST01/ ATPFEE
  COMMON /COST02/ IFEVYR,ILEVYR
  COMMON /COST03/ STRNGY,RMAXF,REPEAT,FFACT
  COMMON /COST05/ REPLCE (25, 6,10, 4)
  COMMON /COST07/ CSTIM, GASCST, IMSTRT, ISCEN, MOVTM,MOVATP, REPIM (4)

  COMMON /COST08/ FAILRT(2,25,39,4)
  COMMON /COSTO 9/ DRSTRN(5 ) ,ZMSTRN(5)
  COMMON /COST10/ FEBNFT(25,10,4),FECNMY(20,4)
  COMMON /COST11/ PCNTSV(4)
  COMMON /COST28/ JATP, JIM, ISTRT, JCYDX, OPRGYR, JFRSYR
  COMMON /COST47/ PPFAIL(25,3,10,4),PPFIXR(25,10,4)
  COMMON /COST49/ PPFEB(25,10,4),APFBEN

  COMMON /IMPAR1/ ICYIM, ISTRIN,MODYR1,MODYR2, WAIVER(2) ,CRIM
  COMMON /IMPAR2/ ILDT (4) , ITEST,NUDATA(2) ,NLIM, IMNAME (20, 9)
  COMMON /IMPAR6/ IFREQ,INTYP
  COMMON /IM240P/ DSIZE(25,4),IM24YR,IPRGYR,IPRSYR
  COMMON /MAXIMA/ MAXVEH, MAXLTW, MAXPOL,MAXREG,MAXYRS

  IFICY - IFEVYR - 1981
  ILICY - ILEVYR - 1981
  IMICY - ISTRT  - 1981
        The IMICY allows this subroutine to loop for calendar
        earlier than the first calendar evaluated  (IFEVYR.) .
      ISICY - IMICY
      IF((ILICY-IMICY).GT.19) ISICY
                                  ILICY - 19
      DO 52 IV-1,4
        IF(ILDT(IV)  .NE. 2) GOTO 52
        DO 51 ICYDX-ISICY, ILICY
          JCYDX-ICYDX-IFICY+1
          ICYDX2-ICYDX+1
          ICY - 1981 + ICYDX
          DO 50 IDX = 1,MAXYRS
            JDX = MAXYRS + 1 - IDX
            MYDX - MAXYRS + ICYDX - JDX
            MY  - MYDX + 1957
        Get vehicle miles traveled  (VMT) that year

        CALL CALMIL(JDX,IV,ACCMIL,VMT)

        Determine technology group

        ITECH-1
                                                   58

-------
                                             Appendix  A

            IF(IV.EQ.LAND.MY.GE.1981)    ITECH=2
            IF(IV.GT.LAND.MY.GE.1984)    ITECH-2
C
C     	Compute  IM failure rate
C
            IF(ITECH .EQ.  2)  GOTO 20
C
            IF(MY.GE.MODYR1.AND.MY.LE.MODYR2.AND.IHFLAG.GT.1)
     *        FAILRT(1,IDX, ICYDX, IV) - ISTRIN/100.0
C
C     Add separate IM240  failure rates, 6/6/91, S adjusted for time
C     by factor of 0.01/0.0187...  0.0199 put into DRSTRN(4) 9/9/91
C       0.4/15  t  0.0750 ... 0.0401
C       0.8/15  :  0.0373 ... 0.0199
C   Revise 12/10/91, to use half of flat zero mile
C   and half of sloped DR.
C       0.8/30  :  0.0324 ... 0.0173
C       1.6/30  t  0.0148 ... 0.0079
C
   20       IF(ITECH .EQ.  2)  THEN
              IF (MY  .GE.  IM24YR)  THEN
C               FAILRT (1, IDX, ICYDX, IV)  - ZMSTRN(4)
C    *            +  DRSTRN (4) *ACCMIL*0. 0001
C  12/10/91...
                FAILRT(1,IDX,ICYDX,IV)  -
     *            DRSTRtt(4)/2.0 + (DRSTRN(4) / (2.0*1. 87))*ACCMIL*0.0001

                FAILRT(2,IDX, ICYDX, IV)  - ZMSTRN(S)
     *            +  DRSTRN (5) *ACCMIL*0. 0001
              ELSE
                FAILRT(1,IDX,ICYDX, IV)  -
     *            ZMSTRN (ITEST) +(DRSTRN(ITEST) * (ACCMIL*. 0001) )
                FAILRT (2, IDX, ICYDX, IV)  - 0.0
              ENDIF
              IF (FAILRT (1, IDX, ICYDX, IV) . GT .RMAXF)
     *          FAILRT (1, IDX, ICYDX, IV) -RMAXF
              FAILRT(2,IDX,ICYDX,IV)  - 0.0
            ENDIF
C
C    Add check  for negative failure rate, 6/6/91 ...
C
            IF (FAILRT (1, IDX, ICYDX, IV) .LT. 0.0)  FAILRT (1, IDX, ICYDX, IV) =0.0
C
C    FAIL1  is used only for FE benefit calcs
C
            IF(ITECH.EQ.2 .AND. MY.GE.IM24YR)  THEN
              FAIL1 - ZMSTRN(4) + DRSTRN (4) *ACCMIL*0 . 0001
            ELSE
              FAIL1  - FAILRT(1,IDX,ICYDX,IV)*1.87
            ENDIF
C
C  Removed  all  code regarding NRPEAT, FAIL2, PREVCY £ PREVNF since
C  we've been using different FE calc method since 6/19/91, 12/10/91...
C
C  Calculate the number of cars to get fuel economy benefit...
C
            FECARS - FAIL1*ENFORC(CRIM,1) * (1. 0-WAIVER (ITECH) )
            IF (FECARS. GT. 1. )  FECARS-1. 0
C
C       	 Calculate annual fuel economy benefits for a vehicle in
C            a  given model year.
C
            IDXP1-MY-1967
            IF(IDXPl.LE.O)   IDXP1=1
            IF(IDXP1.GE.20)  IDXP1=20
C
C    	  Calculate fuel economy benefit in each calendar year.
C
            IF(ICYDX.GE.IFICY)  THEN
              IF(ITECH.EQ.2 .AND.  MY.GE.IM24YR)  THEN
                 FEBNFT (IDX, JCYDX, IV)  =
     *           (PCNTSV (3) *GASCST*VMT/FECNMY (IDXP1, IV) ) *FECARS
              ELSE
                 FEBNFT(IDX,JCYDX,IV)  =
     *           (PCNTSV (ITECH) *GASCST*VMT/FECNMY (IDXP1, IV) ) *FECARS
              ENDIF
C           ENDIF
C
                                                   59

-------
                                             Appendix A

C     Add FE benefit calc for Purge/Pressure Repairs 7/11/91...
C     PPFIXR already includea ENFORC(CRIM,1), but not WAIVER
C     since no waivers are expected for P/P  repairs.
C
              IF(MY.GE. JPRSYR .OR. MY. GE. JPRGYR) THEN
C               PECARS - PPFIXR (IDX, JCYDX, IV)
                PFFEB(IDX,JCYDX,IV) -
     *            PCNTSV(4)*GASCST*VMT/FECNMY(IDXP1,IV)*
     *            PSFIXR (IDX, JCYDX, IV)
              EHDIF
            ENDIF
C
C     	 Adjust annual FE benefit if the inspection is biennial.
C
            IF
-------
                                             Appendix A

      SUBROUTINE CALMIL (JDX, XV, ACCMIL, VMT)
C
C..CALMIL calculates the mileage statistics for each model year in
C..each vehicle class  in each calendar  year.
C
C..Called by CMPUTE,CTFUEL,SMCOST,TBASE
C
C..Calls None
C
C..Array Subscripts:
C
C..CUMMIL(25,8)         -  CUMMIL(JDX,IV)
C
C..Variable Dictionary:
C
C.. Name    Source  Description
C   ~—— ••    •»»—™~  _WB«.»«._~«.MWWa»_~.u__.H>»~_«»~H.MM _.».»«__ —— _~—««
C  ACCMIL   Param    The average accumulated mileage in a one year
C                    period.
C  CUMMIL   Block    The average accumulated mileage at a given
C                    vehicle  age.
C  IV      Param    Vehicle  type index.
C  JDX      Param    Backward model year index (decending model
C-                   years).
C  VMT      Param    Total vehicle miles traveled in a calendar
C                    year.
C
C
      COMMON  /COST23/  VCNT(39,8),GSDSCT(8,10),HDDCUM(2S, 10)
      COMMON  /COST28/  JATP, JIM, ISTRT, JCYDX, JPRGYR, JPRSYR
      COMMON  /CUMCOM/  CUMMIL(25,8)
      COMMON  /MAXIMA/  MAXVEH,MAXLTW, MAXPOL,MAXREG, MAXYRS
C
      IF(JDX.NE.25)  ACCMIL-  (CUMMIL (JDX, IV) +CUMMIL( JDX+1, IV) ) /2. 0
      IF (JDX.EQ. 25)  ACCMIL =  CUMMIL (JDX, IV)
C
      IF(JDX.NE.25)  VMT -  CUMMIL (JDX+1, IV)  - CUMMIL( JDX, IV)
      IFJJDX.EQ.25)  VMT -  CUMMIL(JDX,IV)    - CUMMIL(JDX-1,IV)
C
C
      RETURN
C End CALMIL
      END
                                                    61

-------
                                             Appendix  A
      SUBROUTINE LOOP(ISIZE1,ISIZE2)
c
c.
c.
c.
c.
c
c.
c
c.
c
c.
c
c.
c
c.
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c

c
c


.LOOP does the vehicle type, calendar year, model year and
. pol lut ant 1 ooping
Inside theae loopa the subroutines
.COVER, CMPUTE, CMPREP, POLL, REDUC, and SMCOST are
.called.

.Called by MAIN





.Calls COVER, CMPUTE, POLL, CMPREP and SMCOST

.Array Subacriptea




.Variable Dictionary

. Name Source
IATPLP Block

ICOV1 Local

ICOV2 Local

ICY Local
ICYDX Local
IDX Local
IFEVYR Block
IFICY Local
ILEVYR Block
ILICY Local
INDVLl Local

INDVL2 Local

IP Local
ISIZE1 Par am



ISIZE2 Par am


ISTRT Block
IV Local
JDX Local
LAPSY Block
m Local
MYDX Local


INTEGER PROMPT



Description
Flag indicating whether an ATP only scenario is
to be run
Variable indicating whether a given calendar year
is covered by I/M
Variable indicating whether a given calendar year
is covered by an ATP
Calendar year
Calendar year index
Model year index (forward 1-25)
First calendar year evaluated
Index of first calendar year evaluated
Last calendar year evaluated
Index of last calendar year evaluated
Variable indicating whether a given model year
is covered by I/M
Variable indicating whether a given model year
is covered by an ATP
Pollutant index
Variable containing either the size of the
floating model year window for I/M coverage or
the earliest model year inspected in an I/M
program.
Variable containing either the size of the
floating model year window for ATP coverage
or the earliest model year inspected in an ATP .
I/M program start year
Vehicle class index
Model year index (backwards counter 20 - 1)
ATP start year
Model year
Model year index





C
c
c
c
c
      COMMON /COST02/  IFEVYR, ILEVYR
      COMMON /COST07/  CSTIM,<3ASCST, IMSTRT, ISCEN,MOVIM,MOVATP,REPIM(4)
      COMMON /IMPAR1/  ICYIM, ISTRIN,MODYR1,MODYR2, WAIVER(2) , CRIM
      COMMON /COST12/  IMFXFG, IAFXFG, IMCSFG, IAFEFG, IALDV
      COMMON /COST19/  ICOV1, ICOV2, INDVLl, INDVL2
      COMMON /COST27/  IGRFLG, IATPLP, IDIFF,OUTPRF
      COMMON /COST28/  JATP, JIM, ISTRT, JCYDX, JPRGYR, JPRSYR

      COMMON /ATPARl/  LAPSY,LAP1ST,LAPLST,LVTFLG(4)
      COMMON /FLAGS1/  PROMPT, TAMFLG, SPDFLG, VMFLAG, OXYFLG, DSFLAG
      COMMON /MAXIMA/  MAXVEH, MAXLTW, MAXPOL, MAXREG, MAXYRS
IDIFF - ILEVYR - IFEVYR + 1
IFICY - IFEVYR - 1974
ILICY - ILEVYR - 1974
IFICY - IFEVYR - 1981
ILICY - ILEVYR - 1981
      DO 22 IV - 1,4
        DO 21 JCYDX - 1,IDIFF
C         ICYDX =  (IFEVYR-1974)  +  JCYDX - 1
C         ICY - ICYDX +  1974
          ICYDX -  (IFEVYR-1981)  +  JCYDX - 1
          ICY - ICYDX +  1981
          ICOV1 -  0
          ICOV2 =  0
          IF(IATPLP .EQ.  2) ISTRT  - 2020
                                                   62

-------
                                             Appendix A

c
          IF(ICY.LT.ISTRT  .AMD.  ICY.LT.LAPSY)  GOTO 21
          IF(ICY.GE.ISTRT)    ICOV1  - 1
          IF(ICY.GE.LAPSY)    ICOV2  - 1
C
          DO 20 IDX - 1,MAXYRS
            JDX - MRXYRS +  1  -  IDX
            MYDX - MAXYRS + ICYDX -JDX
C           MY - MYDX + 1950
            MY - MYDX + 1957
C
            IHDVL1 - 0
            IHDVL2 - 0
C
C      	 Determine the vehicles  covered by an inspection program.
C
            CALL COVER (MYDX, ICY, IV, ISIZE1, ISIZE2)
C
C      	 Compute emission  reduction  benefits and
C           failure rates.
C
            DO 10  IP -  1,2
              CALL CMEUTE (IV, ICYDX, MYDX, IP, IDX, JDX, JCYDX)
              CALL POLL (MYDX, IP, IV, IDX, JCYDX)
    10       CONTINUE
C
C      	 Compute overall anti-tampering failure/replacement
C           rates.
C
            CALL CMPKEP (IV, ICY, MYDX, IDX, JDX, JCYDX)
C
            CALL SMCOST (MYDX, ICYDX, IV, JDX, IDX)
C
    20      CONTINUE
    21    CONTINUE
    22  CONTINUE
C
    99  RETURN
C   End LOOP
       END
                                                   63

-------
                                             Appendix  A
c
c
c
c
c
c
c
c
      SUBROUTINE COVER (MYDX, ICY, IV, ISIZE1, ISIZE2)
C
c.
c.
c
c.
c
c.
c
c.
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
.COVER determines
. is covered by an


whether a given model year or calendar year
I/M or ATP program.

.Called by Loop

.Variable

. Name
IAFXFG

IATPLP

ICOV1

ICOV2

ICY
ICYDX
IMFXFG

INDVL1

INDVL2

ISIZE1



ISIZE2


ISTRT
LAPSY
MY
MYDX



Dictionary

Source
Block

Block

Local

Local

Local
Local
Block

Local

Local

Par am



Far am


Block
Block
Local
Local


Deacription
Flag determining whether the ATP has a fixed or
floating model year coverage
Flag indicating whether an ATP only scenario is
to be run
Variable indicating whether a given calendar year
is covered by I/M
Variable indicating whether a given calendar year
is covered by an ATP
Calendar year
Calendar year index
Flag indicating whether the I/M program has a
fixed or floating model year coverage
Variable indicating whether a given model year
is covered by I/M
Variable indicating whether a given model year
is covered by an ATP
Variable containing either the size of the
floating model year window for I/M coverage or
the earliest model year inspected in an I/M
program.
Variable containing either the size of the
floating model year window for ATP coverage
or the earliest model year inspected in an ATP .
I/M program start year
ATP start year
Model year
Model year index

      INTEGER PROMPT
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
       /COST05/
       /COST07/
       /COST12/
       /COST15/
       /COST19/
       /COST27/
       /COST28/
       /ATPAR1/
       /FLAGSl/
       /IMPAR2/
REPLCE(25,6,10,4)
CSTIM, GASCST, IMSTRT, ISCEN, MOVIM, MOVATP , REPIM (4)
IMFXFG, IAFXFG, IMCSFG, IAFEFG, IALDV
TOTCST(2),TOTRED(2),TCOST
ICOV1, ICOV2, INDVL1, INDVL2
IGRFLG, IATPLP, IDIFF, OUTPRF
JATP, JIM, ISTRT, JCYDX, JPRGYR, JPRSYR
LAPSY,LAP1ST,LAPLST,LVTFLG(4)
PROMPT, TAMFLG, SPDFLG, VMFLAG, OXYFLG, DSFLAG
ILDT (4) , ITEST, NUDATA (2) , NLIM, IMNAME (20, 9)
MY - MYDX + 19S7

IF(IATPLP.EQ.2)
IF ((IMFXFG.EQ.l)
IF((IMFXFG.EQ.2)
GOTO 30
                                            GOTO 30
                        .AND.  (ICOV1.EQ.1))  GOTO 10
                        .AND.  (ICOVl.EQ.l))  GOTO 20
      Seta coverage for  floating I/M programs.
   10 IF (MY.GT.ICY)                                INDVL1 = 1
      IF((MY.LE.ICY)  .AND.  (MY.GT.(ICY-ISIZE1)))   INDVL1=2
      IF(MY.EQ.(ICY-ISIZEl))                       INDVL1=3
      IF«MY.EQ. (ICY-ISIZE1-1) )  .AND.  ( (ICY-ISTRT) .GT. 1) )
     *                                             INDVL1 = 4
      IF((MY.EQ.(ICY-ISIZE1-2))  .AND.  ((ICY-ISTRT).GT.2))
     *                                             INDVL1 = 5
      IF(MY .LE.  (ICY-ISIZE1-3) )                   INDVL1 - 6
      GOTO 30
      Sets coverage for fixed  I/M programs.
   20 IF(MY.GT.ICY)
      IF(MY.LE.ICY  .AND.
                    (MY.GE.ISIZE1))
   30 IF(IArPLP.EQ.2)
      IF(IAFXFG.EQ.l  .AND.  ICOV2.EQ.1)  GOTO 40
      IF(IAFXFG.EQ.2  .AND.  ICOV2.EQ.1)  GOTO SO
      GOTO 98
                             INDVL1
                             INDVL1
                                             INDVL1
                                                   64

-------
                                             Appendix A

c
C     	 Sets coverage  for floating ATPa.
C
   40 IF (MT. GT. ICY)                                INDVL2 - 1
      IF((MY.LE.ICY)  .AND.  (MY.GT.(ICY-ISIZE2)))   INDVL2-2
      IF(MT.EQ.(ICY-ISIZE2))                       INDVL2=3
      IF((MY.EQ. (ICY-ISIZE2-1) )  .AND.  ( (ICY-LAPSY) . GT. 1) )
     *                                             INDVL2 = 4
      IF((MY.EQ. (ICY-ISIZE2-2))  .AND.  <(ICY-LAESY).GT.2))
     *                                             INDVL2 - 5
      IF(MY  .LE.  (ICY-ISIZE2-3))                   INDVL2 - 6
      GOTO 98
C
C     	 Seta coverage  for fixed ATPa.
C
   50 IF (MY.GT. ICY)                                INDVL2 - 1
      IF (MY. LE. ICY .AND.  (MY.GE. ISIZE2) )           INDVL2 - 2
C
   98 IF(ILDT(IV)   .EQ.l)  INDVL1-1
      IF(LVTFLG(IV).EQ.l)  INDVL2-1
C
   99 RETURN
C  End  COVER
      END
                                                    65

-------
                                             Appendix A
      SUBROUTINE CMPUTE(IV,ICYDX,MYDX,IP,IDX,JDX,JCYDX)
c
c.
c.
c.
c
c.
c
c.
c
c.
c.
c.
c.
c
c.
c
c.
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c

. CMPUTE computes

the overall emission reduction


.benefits by computing the differences in the MOBILE 4
. EFALL arraya.

.Called by LOOP

.Array Subscripts

. JULMYR(25,8)
.TDUCT (25, 2, 3, 10,
.VCNT (46, 8)
.WAIVER (2)







- JULMYR (JDX, IV)
4) - TDUCT (25, 2, 3, 10, 4)
- VCNT (ICYDX, IV)
- WAIVER (ITECH)












.Variable dictionary

. Name Source
ICY Local
ICYDX Local
IETYP Local
IMC Local
IMCAS1 Local
IMCAS2 Local
IF Local
I TECH Local
IV Local
JDX Local
JULMYR Block

tfl Local
MYDX Local

Description
Calendar year
Calendar year index
Control program reduction type index
Control program reduction type index
Control program reduction type counter







Second control program reduction type counter
Pollutant index 1 - HC 2 = CO
Technology type index 1 - old tech 2 - new
Vehicle class index
Model year index (backwards 20 - 1)
Distribution of vehicle fleet by model year,
One distribution per vehicle class
Model year
Model year index

tech






by calendar year by vehicle class as a result of

RDUCT1 Local

RDUCT2 Local
RDUCT Local

TDUCT Block

VCNT Block

VMT Par am
WAIVER Block



a control program
Emissions of no control program or lesser
control program
Emissions of greater control program
Difference in emissions between RDUCT1 and
RDUCT2 RDUCT = RDUCT1 - RDUCT2
Array containing final emission factors.
VMT weighting and fleet weighting have been
Number of vehicles of each type in each
calendar year
Vehicle miles travelled in a given calendar
Array containing waiver rates










done


year




C
c
c
c
c
c
c
c
c
c
c
REAL JULMYR

COMMON /COST01/ ATPFEE
COMMON /COST04/ TAMPER(25,6,4,10,4)
COMMON /COST05/ REPLCE(25,6,10,4)
COMMON /COST07/ CSTIM, GASCST, IMSTRT,ISCEN,MOVIM,MOVATP,REPIM (4)
COMMON /COST20/ TDUCT(25,2,3,10,4),TEVAP(25,10,4)
COMMON /COST23/ VCNT(39,8),GSDSCT(8,10),HDDCUM(25,10)

COMMON /COST25/ FRLOSS(25,4,4,10),FREF(25,4,4,10)
COMMON /COST26/ FRSTLS(25,8,4,10),YRSTLS(10),BRSTLS

COMMON /COST50/ TRLOSS(25,4,3),TREVAP(25,4,3)
COMMON /COST50/ TRLOSS(25,4, 3),TREVAP(25,4,3),TRSTLS(25,4,3)

COMMON /COST30/ EVAP(25,8,4,10),FER(3,25,8,4,10)
COMMON /COST40/ YEVTON(10),BEVTON,YEREF(10),BEREF
COMMON /COST41/ YERLOS(10),BERLOS
COMMON /IMPAR1/ ICYIM, ISTRIN, MODYR1, MODYR2, WAIVER (2) , CRIM
COMMON /MYRCAL/ XMYM(25,8),JANMYR(25,8),IF(25,8),TFMYM(25,8)
COMMON /MYRSAV/ AMAR(25, 8) , JULMYR (25, 8) ,NEWCUM
COMMON /REGISF/ GSFVCT(8),USRGSF(25,2)
      JCYDX2 = JCYDX  +  1
      ICY    - 1981 4- ICYDX
      MY - MYDX +  1957
      There are three cases  (IMC)  for which emission reductions
      are calculated:
            IMC
                  1  :  I/M tampering deterrence effect.
                                                   66

-------
                                             Appendix A

C           IMC - 2  :  The effect o£ I/M repairs.
C           IMC = 3  :  The effect of ATP repairs.
C
      DO 30 IMC-1,3
        IMCAS1-IMC
        IMCAS2-IMC+1
C
        IF(IDX.HE.1)  GOTO 10
C
C     	 Calculate basic emission reductions without losses.
C           for last  model year index in given window,
C           set mid-year emissions equal to Jan 1 emissions,
C           otherwise interpolate between the Jan 1st emissions
C           from one  calendar year to the next for the given
C           model year.
C
        RDUCT1-FER (IP, IDX, IV, IMCAS1, JCYDX)
        RDUCT2-FER (IP, IDX, IV, IMCAS2, JCYDX)
C
        REVAPl-EVAP (IDX, IV, IMCAS1, JCYDX)
        REVAP2-EVAP (IDX, IV, IMCAS2, JCYDX)
C
        RLOSS1-FRLOSS (IDX, IV, IMCASl, JCYDX)
        RLOSS2-FRLOSS (IDX, IV, IMCAS2, JCYDX)
C. . .
C       RSTLS1-FRSTLS (IDX, IV, IMCASl, JCYDX)
C       RSTLS2-FRSTLS (IDX, IV, IMCAS2, JCYDX)
C
        GOTO  20
C
C     	 The January 1st EFALL value for a given calendar
C           year is  averaged with the subsequent year's EFALL
C           value to  arrive at a July 1st emission value.
C
    10   RDUCT1 -  (FER(IP,IDX,IV,IMCAS1, JCYDX)  +
     *            FER(IP,IDX,IV,IMCAS1,JCYDX2))/2.0
        RDOCT2 -  (FER(IP,IDX,IV,IMCAS2,JCYDX)  +
     *            FER(IP,IDX,IV,IMCAS2,JCYDX2))/2.0
C
        REVAP1 -  (EVAP (IDX, IV, IMCASl, JCYDX) +
     *            EVAP (IDX, IV, IMCASl, JCYDX2))/2.0
        REVAP2 -  (EVAP (IDX, IV, IMCAS2, JCYDX) +
     *            EVAP(IDX,IV,IMCAS2,JCYDX2))/2.0
C
        RLOSS1 -  (FRLOSS(IDX,IV, IMCASl, JCYDX)  +
     *            FRLOSS (IDX, IV, IMCASl, JCYDX2))/2.0
        RLOSS2 -  (FRLOSS (IDX, IV, IMCAS2, JCYDX)  +
     *            FRLOSS (IDX, IV, IMCAS2, JCYDX2))/2.0
C
C. . .
C       RSTLS1 -  (FRSTLS(IDX,IV, IMCASl, JCYDX)  +
C    *            FRSTLS (IDX, IV, IMCASl, JCYDX2))/2.0
C       RSTLS2 -  (FRSTLS (IDX, IV,IMCAS2, JCYDX)  +
C    *            FRSTLS (IDX, IV, IMCAS2, JCYDX2) )/2.0
C
    20   RDUCT - RDOCT1 - RDUCT2
C
C. . .
        REVAP -  (REVAP1-REVAP2) +  (RLOSS1-RLOSS2)
C       REVAP -  (REVAP1-REVAP2) +  (RLOSS1-RLOSS2) + (RSTLS1-RSTLS2)
C
C     	 Get the  vehicle miles traveled  (VMT)
C
        CALL  CALMIL(JDX,IV,ACCMIL,VMT)
C
C       	  Change  emission reductions to WEIGHT (grams) reductions.
C             On a per vehicle basis, not a fleet basis.
C
C
C           TDUCT(IDX,IP,IMC, JCYDX, IV)
C           represents the I/M or ATP emission's benefit  (tonnage) per
C           car per  year and multiplies it by the appropriate fleet
C           size.  The units are then converted from tons to grams
C           when calculating TDUCT.
C
C   Save these 2 separately for debugging...
C
        VADJ  - VCNT (ICYDX, IV)  * JULMYR (JDX, IV) * 1.1025E-6 *
     *        VMT  !*  GSFVCT(IV) iGSDSCT (IV, JCYDX)
C
        TREVAP(IDX, IV, IMC) = (REVAP1-REVAP2) * VADJ


                                                   67

-------
                                             Appendix A

        TRLOSS(IDX,IV,IMC) =  (RLOSS1-RLOSS2) * VADJ
C
        IF(IP.EQ.l)  TEVAP(IDX, JCYDX, IV) - REVAP * VADJ
C
        IF(IP.EQ.2) REVAP = 0.0
        TDOCT(IDX,IP,IMC,JCYDX,IV) - (RDOCT+REVAP) * VADJ
C
   30 CONTINUE
C
   99 RETURN
C  End CMPUTE
      END
                                                  68

-------
                                              Appendix A
      SUBROUTINE POLL (MYDX, IP, IV, IDX, JCYDX)
c
c.
c.
c.
c.
c
c.
c
c.
c
c.
c
c.
c.
c.
c.
c.
c.
c.
c
c.
c
c.
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c

.POLL trana'fera the
.which are specific

contents from the array TDUCT into other arrays
to I/M, ATP or DETERRENCE. These new arrays are
.subsequently summed by model year, calendar year, pollutant and
.vehicle class in subroutine SUMWDW.

.Called by LOOP

•Calls None

.Array Subscripts:

.CRDTAT(3,2)
.CRDTIM(3,2)
.DETER (25, 2, 10, 4)
.REDATP (25, 2, 10, 4)
.REDIM(25,2,10,4)
.TDUCT(25,2,3,10,4)
. TONRED(25,2,10, 4)








- CRDTAT (IAYRDX, IP)
- CRDTIM(IMYRDX,IP)
- DETER (IDX, IP, JCYDX, IV)
- REDATP (IDX, IP, JCYDX, IV)
- REDIM (IDX, IP, JCYDX, IV)
- TDUCT (IDX, IP, 3, JCYDX, IV)
- TONRED(IDX, IP, JCYDX, IV)

.Variable Dictionary:

. Dame Source
CRDTAT Block

CRDTIM Block

DETER Block


IAYRDX Local

IMYRDX Local

INDVL1 Local


INDVL2 Local


IP Par am
IV Par am
MY Local
MYDX Par am
REDATP Block


REDIM Block


TDUCT Block

TOHRED Block


TONSAT Block

TONSDE Local

TONSIM Local




Description
Fraction of ATP effectiveness for vehicles which
are no longer inspected.
Fraction of I/M effectiveness for vehicles which
no longer inspected.
Array which contains the pollutant reductions for
each model year, calendar year and vehicle type
due to tampering deterence.
Index used in CRDTAT array. Relates INDVL2 to
the model years which are no longer inspected.
Index used in CRDTIM array. Relates INDVL1 to the
model years which are no longer inspected.
Flag determining whether a given vehicle class
with a given model year in a calendar year is
covered by I/M.
Flag determining whether a given vehicle class
with a given model year in a calendar year is
covered by ATP •
Pollutant type index (ascending model years) .
Vehicle type index.
Model year .
Abolute model year index (1950=1) .
Array which contains the pollutant reduction for
each model year, calendar year and vehicle type
due to ATP.
Array which contains the pollutant reductions
for each model year, calendar year and vehicle
type due to ATP.
The emission factor after it has been multiplied
by the fleet size and converted to units of tons.
Array which contains the pollutant reductions for
each model year calendar year and vehicle type
due to all control programs.
Intermediate variable holding the pollutant
reductions due to the ATP.
Intermediate variable holding the pollutant
redutions due to the deterence.
Intermediate variable holding the pollutant
reductions due to the I/M program.


C
c
DIMENSION CRDTAT (3,2), CRDTIM (3,2)

COMMON /COST01/ ATPFEE
COMMON /COST14/ TONRED(25,2,10,4),COST(25,10, 4)
COMMON /COST16/ DATPFE, DIMFEE, DIMREP, DATPRE, AATPFE, AIMFEE
COMMON /COST18/ AIMREP, AATPRE, CFACTR, CATPFE (25,10, 4)
COMMON /COST13/ CIMFEE(25,10,4),CIMREP(25,10, 4)
COMMON /COST21/ CATPRE(25,10,4),FULBEN(25,20,4)
COMMON /COST46/ REDIM(25,2,10,4),REDATP(25, 2,10, 4)
COMMON /COST17/ DETER(25,2,10,4),TDETER<25,2,10, 4)
COMMON /COST19/ ICOV1,ICOV2,INDVL1,INDVL2
COMMON /COST20/ TDUCT(25,2,3,10,4),TEVAP(25,10,4)

COMMON /COST50/ TRLOSS(25,4,3),TREVAP(25, 4, 3)
COMMON /COST50/ TRLOSS(25,4,3),TREVAP(25,4,3),TRSTLS(25,4,3)
                                                   69

-------
                                             Appendix A

      COMMON /ATPARl/ LAPSY,LAP1ST,LAPLST,LVTFLG(4)
      COMMON /ATPAR2/ ATPPGM,ATPFQT,CRATP, DISTYP (8)
C
      DATA CRDTIM/.50,.25,.125,  .50,.25,.125/
      DATA CRDTAT/.75,.50,.25,   .75,.50,.25 /
C
C     MY  - MYDX + 1950
      MY - MYDX + 1957
C     ICY - ICYDX + 1974
C
      TONSDE >0.0
      TONSIM -0.0
      TONSAT -0.0
C
C     	 Aaaign benefita  for  vehicle covered by * program.
C
      IF(INDVL1.EQ.2) TONSDE - TDUCT(IDX,IP,1,JCYDX,IV)
      IF(INDVL1.EQ.2> TONSIM - TDUCT(IDX, IP, 2, JCYDX, XV)
      IF(INDVL2.EQ.2) TONSAT - TDUCT(IDX,IP,3,JCYDX,IV)
C
C
      IMYRDX - INDVL1 - 2
      IAYRDX - INDVL2 - 2
C
      IF(INDVLl.GE.3.AND.INDVL1.LT.6)
     *  TONSDE-TDUCT (IDX, I?, 1, JCYDX, IV) *CRDTIM(IMYRDX, IP)
      IF (INDVL1. GE. 3 .AND. INDVLl. LT. 6)
     *  TONSIM-TDUCT(IDX,IP,2,JCYDX,IV)*CRDTIM(IMYRDX,IP)
      IF (INDVL2 . GE. 3 . AND. INDVL2 . LT. 6)
     *  TONSAT-TDUCT (IDX, IP, 3, JCYDX, IV) *CRDTAT (IAYRDX, IP)
C
      IF (INDVLl. GE. 3. AND.INDVL1.LT.6)  THEN
        TEVAP(IDX,JCYDX,IV)  - TEVAP(IDX,JCYDX,IV)*CRDTAT(IMYRDX,IP)
        TREVAP (IDX, JCYDX, IV)  - TREVAP (IDX, JCYDX, TV) *CRDTAT (IMYRDX, IP)
        TRLOSS (IDX, JCYDX, IV)  - TRLOSS (IDX, JCYDX, IV) *CRDTAT (IMYRDX, IP)
      ENDIF
      IF (INDVLl. EQ.l  .OR.  INDVLl. GE. 6)  THEN
        TEVAP (IDX, JCYDX, IV)  - 0.0
        IF(IV.LE.4) THEN
          DO 20 IMC-1,3
            TREVAP(IDX,IV,IMC) -0.0
            TRLOSS (IDX, IV, IMC) -0.0
   20     CONTINUE
        ENDIF
      ENDIF
C
C     	 Add IM tona to ATP tons  to gut  total.
C
      TONRED (IDX, IP, JCYDX, IV) -TONSIM+TONSAT+TONSDE
      REDIM (IDX, IP, JCYDX, IV) -TONSIM
      REDATP (IDX, IP, JCYDX, IV) -TONSAT
      DETER (IDX, IP, JCYDX, IV) -TONSDE
C
      RETURN
C  End POLL
      END
                                                   70

-------
                                             Appendix A
      SUBROUTINE CMPREP (IV, ICY, MYDX, IDX, JDX, JCYDX)
c
c.
c.
c.
c.
c
c
c
c.
c
c.
c.
c
c.
c
c.
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c


c









.CMPREP uses the differences in MOBILE 4 I/M,ATP, and no
. control program tampering rates and compute a ATP
.failure/replacements for five different tampering
. situations .

Called by LOOP

.Array Subscript

.REPLCE (25, 5, 10, 4) - REPLCE (IDX,IT, JCYDX,IV)
.TAMPER (25, 6, 4, 10, 4) - TAMPER(MYDX, IT, 3, ICYDX,IV)

.Variable dictionary:

. Name Source Description

ICY Local Calendar year
ICYDX Local Calendar year index



















ICYDX2 Local Subsequent calendar year index ICYDX2 = ICYDX+1
IT Local Emission control tampering component index
IT2 Local Tampering component index for DISTYP
DISTYP Block Array of flags indicating which emission
control components are to be inspected by
ATP
IV Local Vehicle class index
LAPSY Block ATP start year
MY Local Model year
MYDX Local Model year index
PBFAC Local Plumbtesmo efficiency factor
RETAIL Local Percentage of vehicles which refail the



the







tampering inspection the following year after
being fixed the previous year

REPLCE Block The difference in tampering rates as a result
of a control program
STOAER Block Condition determining if the Plumbtesmo
efficiency is applied
TAMPER Block The MOBILE4 tampering rates

INTEGER DISTYP
DIMENSION REFAIL (6)

COMMON /COST01/ ATPFEE
COMMON /COST04/ TAMPER(25, 6, 4,10, 4)
COMMON /COST05/ REPLCE (25, 6,10, 4)
COMMON /COST06/ STOAER
COMMON /COST40/ YEVTON (10) ,BEVTON, YEREF (10) ,BEREF
COMMON /COST41/ YERLOS (10) , BERLOS
COMMON /ATPAR1/ LAPSY, LAP1ST,LAPLST,LVTFLG (4)
COMMON /ATPAR2/ ATPP6M,ATPFQT,CRATP, DISTYP (8)
















C
c
c
c
c

c
c
c
c
   DATA PBFAC  / 1.061  /
   DATA REFAIL / O.OS,  0.05,  0.10,  0.30,  0.30,  0.30 /


   JCYDX2 - JCYDX + 1
   IDX2   - IDX - 1
   MY - MYDX + 1957


   DO 30 IT-1,6


Compute initial year ATP
failure/replacement by  substracting non-ATP (IM only)
from IM+ATP.


     IF (ICY .GT. LAPSY)   GOTO 10


NOTE: the tampering rate  for  the last model year in each
calendar year is always zero.


     REPLCE (IDX, IT, JCYDX, IV)  « 0
     IF(IDX .EQ. 2) REPLCE(IDX,IT,JCYDX,IV) «
  *          (TAMPER(IDX, IT, 3, JCYDX, IV)
  *        -  TAMPER(IDX,IT,4, JCYDX,IV) ) *ENFORC (CRATP, 2)
     IF (IDX .GT. 2) REPLCE (IDX, IT, JCYDX, IV) =
  *          (TAMPER (IDX, IT, 3, JCYDX, IV)
  *        -  TAMPER(IDX2, IT, 4, JCYDX2, IV) ) *ENFORC (CRATP, 2)

     GOTO 20
                                                   71

-------
                                             Appendix A

C  Calculate the number of replacements  in  subsequent years and
C  adjust replacements to account  for enforcement  problems.
C
   10   REPLCE(IDX,IT,JCYDX,IV) -
     *   TAMPER(IDX, IT, 3, JCYDX, IV) *REFAIL(IT) *ENFORC (CRATP, 2)
C
C  If a given emission control component is not  inspected,
C  set the replacement rate to zero.
C
   20   IT2 - IT
        IF(IT .GT. 3) IT2 = IT + 2
        IF(DISTYP(IT2) .NE. 2) REPLCE (IDX, IT, JCYDX, IV)  -  0.0
C
C  Replacement rates cannot be less than zero
C
        IF(REPLCE(IDX,IT,JCYDX,IV).LT.0.)
     *    REPLCE(IDX,IT,JCYDX,IV)  =0,0
C
C  30 CONTINUE
C
C  Adjust misfueling replacements  to account  for lead
C  deposit testing failures.
C
      IF(STOAER  .LT. 1.0 .AND. IT  .EQ. 3)
     *  REPLCE (IDX, 3, JCYDX, IV) -REPLCE (IDX, 3, JCYDX, IV) *PBFAC
C
   30 CONTINUE
C
   99 RETURN
C  End CMPREP
      END
                                                  72

-------
Appendix A
SUBROUTINE SMCOST (MYDX, ICYDX, IV, JDX, IDX>
C

C. . Calculates and sums the I/M and ATP inspection and repair coats
C
C. .Called by LOOP
C
C. .Calls CAMIL
C
C. .Array Subscripts
C
C. .ATCOST (2, 6)
C. .ATEQP (25, 4, 4)
C. .CATPFE(25,10,4)
C. .CATPRE (25, 10, 4)
C. .CIMFEE (25, 10, 4)
C. .CIMREP (25, 10, 4)
C.. COST (25, 10, 4)
C. .FAILRT(25,20,4)
C. .FEBNFT (25, 20, 4)
C. .FULBEN (25, 20, 4)
C. . JULMYR(25,8)
C. .REPLCE(25,5,10,4)
C. .REFIH(3)
C. .VCNT(46,8)
C







- ATCOST (ITECH, ITAMS)
- ATEQP (MYDX,ITAM4, IV)
- CATPFE (MYDX, ICYDX, IV)
- CATPRE (IDX, JCYDX, IV)
- CIMFEE (IDX, JCYDX, IV)
- CIMREP (IDX, JCYDX, IV)
- COST (IDX, JCYDX, IV)
- FAILRT (IDX, JCYDX, IV)
- FEBNFT (IDX, ICYDX, IV)
- FULBEN (IDX, JCYDX, IV)
- JULMYR(JDX,IV)
- REPLCE (IDX, ITAMS , JCYDX, IV)
- REPIM (ITECH)
- VCNT (ICYDX, IV)

C. .Variable Dictionary
C
C Name Source
C ACCMIL Faram
C
C ATCOST Block
C
C
C
C
C
C
C
C
C
C
C ATEQP Block
C
C
C ATPFEE Block
C CATPFE Block
C
C CATPRE Block
C
C
C CIMFEE Block
C
C CIMREP Block
C
C
C CNSPEC Local
C
C COST Block
C
C CREPAT Local
C CREPIM Local
C CSTIM Block
C FAIL Block
C FAIL2S Local
C FEBNFT Block
C FULBEN Block
C
C
C ICY Local
C ICYDX Local
C INDVL1 Local
C
C INDVL2 Local
C
C ITAM4 Block
C ITAM5 Block
C ITECH Block
C
C
C IV Local

Description
The average accumulated mileage in a one year
period
Cost of replacing emission control devices as a
result of failing a tampering inspection. The
default assumptions are: ATCOST (TECH, TYPE)
TYPE - 1 Air Pump System
- 2 Catalyst Removal
= 3 Misfueled Catalyst
* 4 Evaporative Control System
- 5 PCV System
» 6 Gas Cap
• 7 Purge Test
•* 8 Pressure Test
Array containing the fraction of each model year
in each vehicle class equipped with each
tampering component .
Anti-tampering inspection fee
Array which holds and sums the ATP fees for each
model year, calendar year and vehicle type
Array which holds and sums the ATP repair coats
for each model year, calendar year and vehicle
type
Array which holds and sums the I/M fees for each
model year, calendar year and vehicle type
Array which holds and sums the I/M repair costs
for each model year, calendar year and vehicle
type
Variable which sums and holds the I/M and the
ATP fees
Array which holds and sums the total coat for
each model year, calendar year and vehicle type
Holds the ATP repair coat
Holds the I/M repair coat
I/M inspection fee
Calculated I/M failure rate
Failure rate for 2 speed teat only
Final fuel economy benefit (before summing)
Array which holds and sums the fuel economy
benefit for each model year, calendar year and
vehicle type
Calendar year
Calendar year index
Variable indicating whether a given model year is
covered by I/M
Variable indicating whether a given model year is
covered by ATP
Emission control tampering component index
Emission control tampering component index
Technology group 1 - old tech , 2 = new tech
Old - myr 1951-1980 LDGV , myr 1951-1983 LDT's
New - myr 1981+ LDGV , myr 1984+ LDT's
Vehicle class index
    73

-------
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
                                          Appendix A

 JDX     Local    Model year index  (backwards 20-1)
 JULMYR  Block    Fraction of vehicles  in each model year from
                  MOBILE4
 MY      Local    Model year
 MYDX    Local    Model year index
 REPLCE  Block    The difference  in tampering rates as a result of
                  control programs.   See  ATCOST above for elements
 REPIM   Block    Array containing the  I/M repair costs
                   REPIM(1) - Old Tech   ;   REPIM(2)  - New Tech
                   REPIM(3) - IM240 ; REPIM(4)  - NOx (IM240)
 VCNT    Block    Array of the total  number of vehicles in given
                  calendar years

   REAL    JULMYR
   INTEGER BYEAR,ATPFLG,PROMPT,ATPFQT,DISTYP
      COMMON  /COST01/
      COMMON  /COST03/
      COMMON  /COSTOS/
      COMMON  /COST07/
      COMMON  /COST08/
      COMMON  /COST09/
      COMMON  /COST10/
      COMMON  /COST12/
      COMMON  /COST13/
      COMMON  /COST14/
      COMMON  /COST15/
      COMMON  /COST16/
      COMMON  /COST18/
      COMMON  /COST19/
      COMMON  /COST21/
      COMMON  /COST22/
      COMMON  /COST23/
      COMMON  /COST23/
      COMMON  /COST24/
      COMMON  /COST28/
   add PPFAIL 6/19/91,
      COMMON  /COST47/
   add Purge/Pressure
      COMMON  /COST48/
      COMMON  /COST49/
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
      COMMON
          /ATPAR2/
          /IMPAR1/
          /IMPAR2/
          /IMPAR6/
          /MYRSAV/
          /IM240P/
          /REGISF/
          /YEARS4/
ATPFEE
STRNGY , RMAXF , REPEAT , FFACT
REPLCE(25,6,10,4)
CSTIM, GASCST, IMSTRT, ISCEN, MOVTM, MOVATP, REPIM (4)
FAILRT(2,2S,39,4)
DRSTRN (5) , ZMSTRN (5)
FEBNFT (25,10, 4), FECNMY (20, 4)
IMFXFG,IAFXFG,IMCSFG,IAFEFG,IALDV
CIMFEE(25,10,4),CIMREP(25,10,4)
TONRED(25,2,10,4>,COST(25,10,4)
TOTCST(2),TOTRED(2),TCOST
DATPFE , DIMFEE , DIMREP , DATPRE, AATPFE , AIMFEE
AIMREP, AATPRE,CFACTR,CATPFE (25,10, 4)
ICOV1,ICOV2,INDVL1,INDVL2
CATPRE(25,10,4),FULBEN(25,20,4)
ATCOST(2,8)
VCNT(46,8)
VCNT(39,8),GSDSCT(8,10),HDDCUM(25,10)
ATEQP(20,5,4)
JATP, JIM, ISTRT, JCYDX, JPRGYR, JPRSYR
 and PPFEB,PPFIXR,APFBEN 7/11/91  ...
PPFAIL(25,3,10,4),PPFIXR(25,10,4)
inspection fees, 6/21/91 ...
PRGFEE,PRSFEE,CEVFEE(25,10,4),TRNFEE
PPFEB(25,10, 4) ,APFBEN

ATPPGM,ATPFQT, CRATP, DISTYP (8)
ICYIM, ISTRIN, MODYR1,MODYR2, WAIVER (2) , CRIM
ILDT (4) , ITEST, NUDATA(2) , NLIM, IMRAME (20, 9)
IFREQ,INTYP
AMAR(25,8),JULMYR(25,8),NEWCUM
DSIZE (25, 4) , IM24YR, IPRGYR, IPRSYR
GSFVCT < 8 ) ,USRGSF(25,2)
IY1941,IY1960,IY2020
   MY - MYDX + 1957
   ICY - ICYDX + 1974
   ICY - ICYDX + 1981

   CNSPEC -0.0
   CREPIM - 0.0
   CREPAT - 0.0
to allow addition of Purge/Pressure inspection costs,  6/21/91...

   CATPRE(IDX, JCYDX, IV) - 0.0
   CEVFEE (IDX, JCYDX, IV) =0.0
         Determine vehicle technology group.
   ITECH-1
   IF(IV. EQ. LAND. MY. GE. 1981) ITECH-2
   IF(IV.GT.LAND.MY.GE.1984) ITECH-2

   	 Get average miles accumulated during  the  calendar
         year, for the given model, and  aet  I50K index.

   I50K-1
   CALL CALMIL(JDX, IV, ACCMIL, VMT)
   IF(ACCMIL.GT.50000.0) I50K-2
         Skip cost calculation if vehicle  is  not  inspected.
      IF(INDVL1.NE.2  .AND.  INDVL2. NE. 2  .AND.  MY.LT.JPRGYR .AND.
     *   MY.LT.JPRSYR) GOTO 30
                                                   74

-------
                                              Appendix A

      IF(INDVL1.NE.2.AND.INDVL2.NE.2) GOTO 5
C
C     	 Inspection coat is the sum of the  I/M fee plus the ATP
C           fee.
C
      FEE - CST1M
G
C  TRNFEE is  incremental transient inspection  cost over purge
C  inspection coat,  PRGFEE
C. . .
      IF(MY.GE.IM24YR) THEN
        FEE - FEE + TRNFEE
        IF (IM24YR.LT. JPRGYR .AMD. MY. LT. JPRGYR) FEE - FEE + FRGFEE
      ENDIF
      IF(IFREQ.EQ.2)  FEE-FEE/2.0
      IF(INDVL1.EQ.2)
     *  CTMFEE (IDX, JCYDX, IV) =
     *    VCNT (ICYDX, IV)* JOLMYR( JDX, IV) *FEE
C
C     FEE=ATPFEE
C     IF(ATPFQT.EQ.2)  FEE-FEE/2.0
      IF(INDVL2.EQ.2)  THEN
        FEE - ATPFEE
        IF(ATPFQT.EQ.2) FEE - FEE/2.0
        CATPFE (IDX, JCYDX, IV) -
     *    VCNT (ICYDX, IV) *JULMYR (JDX, IV) *FEE
      ENDIF
C
C   Add Purge S Preaaure Inapection coat calcs, 6/21/91 ...
C   uaing PRGFEE, PRSFEE, and CEVFEE
C
   5  IF(IV.LE.4  .AND.  (MY.GE. JPRGYR .OR. MY.GE. JPRSYR) )  THEN
C       IF(IV.LE.4)  THEN
          IF (MY. GE. JPRGYR) THEN
            FEE - PRGFEE
            IF(IFREQ.EQ.2)  FEE - FEE/2.0
            CEVFEE (IDX, JCYDX, IV) - CEVFEE (IDX, JCYDX, IV) +
     *         VCNT (ICYDX, IV)* JULMYR( JDX, IV) *FEE
          ENDIF
          IF (MY. GE. JPRSYR) THEN
            FEE = PRSFEE
            IF(IFREQ.EQ.2) FEE - FEE/2.0
            CEVFEE (IDX, JCYDX, IV) = CEVFEE (IDX, JCYDX, IV) +
     *         VCNT (ICYDX, IV) *JOLMYR( JDX, IV) *FEE
          ENDIF
C       ENDIF
      ENDIF
C     CNSPEC  - CIMFEE (IDX, JCYDX, IV) + CATPFE (IDX, JCYDX, TV)
      CNSPEC  - CIMFEE (IDX, JCYDX, IV) + CATPFE (IDX, JCYDX, IV)
     *  -1- CEVFEE (IDX, JCYDX, IV)
C
   8  IF(INDVL1.NE.2)  GOTO 10
C
C      	 IM repair coat.
C     For IM240 failures, uae 2-apd repair coat for that  fraction of
C     IM240 failures that would have failed the 2-speed teat.
C
        IF(MY.GE.IM24YR) THEN
          FAIL2S  - ZMSTRN(2) + (DRSTRN(2)*ACCMIX.*0.0001)
          IF(FAIL2S.GT.RMAXF) FAIL2S - RMAXF
          IF(FAIL2S.LT.O.O) FAIL2S - 0.0
          IF (FAIL2S . LE. FAILRT (1, IDX, ICYDX, IV) ) THEN
            CREPIM - FAIL2S*REPIM(2) +
     *         ((FAILRT(1,IDX,ICYDX,IV) - FAIL2S)  * REPIM(3))
          ELSE
            CREPIM - FAIL2S*REPIM(2)
          ENDIF
C
C  Add  NOx repair costs...
C
          CREPIM  - CREPIM + FAILRT (2, IDX, ICYDX, IV) *REPIM (4)
C
        ELSE
          IF((MY.LE.1980) .OR.   (MY.LE.1983 .AND.  IV.GE.2))
     *      CREPIM-FAILRT (1, IDX, ICYDX, IV) *REPIM (1)
          IF((MY.GE.1984) .OR.   (MY.GE.1981 .AND.  IV.EQ.l))
     *      CREPIM-FAILRT (1, IDX, ICYDX, IV) *REPIM(2)
        ENDIF
C
      CREPIM  - CREPIM* (GRIM/100) * VCNT (ICYDX, IV) *JULMYR( JDX, IV)
      CIMREP (IDX, JCYDX, IV)  -CREPIM


                                                    75

-------
                                             Appendix A

c
   10 IP(INDVL2.HE.2>  GOTO 21
C
C     	AT? repair coat.
C
      DO 20 ITAM5-1,6
        ITAM4=ITAM5
        IF(ITAM4.GT.2)  ITAM4-ITAM4-1
C
        IDXP2 - MY - 1966
        IF(IDXP2.LE.O)   IDXP2-1
        IF(IDXP2.GE.20)  IDXP2-20
          CREPAT - CREPAT+(ATEQP(IDXP2,ITAM4,IV)*
     *    REPLCE (IDX, ITAMS, JCYDX, IV) *ATCOST (ITECH, ITAM5) *
     *     (CRATP/100.))
C
   20 CONTINUE
C
   21 IF ( (MY. IiT. JPRGYR .AND.  MY.LT. JPRSYR) .OR. IV.GT.4) GOTO  25
C
C     Fix Purge, Press repair costs including separate Prg+Prs cost
C     in new block PPFAIL,  6/21/91
C
      IF (MY.GE. JPRGYR)  CREPAT - CREPAT4- (PPFAIL (IDX, 1, JCYDX, IV) *
     *   (ATCOST (ITECH, 7))* (CRATP/100.))
      IF (MY. GE. JPRSYR)  CREPAT - CREPAT+ (PPFAIL (IDX, 2, JCYDX, TV) *
     *   (ATCOST(ITECH,8))*(CRATP/100.))
C
   25 CREPAT - CREPAT*VCNT(ICYDX,IV)*JULMYR(JDX,IV)
      CATPRE (IDX, JCYDX, IV)  -  CREPAT
C
C  Sum inspection, I/M and ATP costs to get total.
C  Note: If fuel economy benefits are larger than the program
C        costs, then a negative cost is assumed for that model year.
C
   30 COST(IDX,JCYDX,IV)  - CNSPEC + CREPIM + CREPAT
      IF (COST (IDX, JCYDX, IV) . GT. 0. 0 . AND.  INDVL1. EQ. 2)
     *  COST (IDX, JCYDX, IV)  -
     *  COST (IDX, JCYDX, IV)  -  ( FEBNFT (IDX, JCYDX, IV) *
     *  VCNT(ICYDX,IV)*JULMYR(JDX,IV) )
C
      IF(INDVL1.EQ.2)
     *  FOLBEN (IDX, JCYDX, IV)  - FEBNFT (IDX, JCYDX, IV) *
     *    VCNT(ICYDX,IV)*JULMYR(JDX,IV)
C
C  Add Purge/Press FE  benefit summation...
C
      IF((MY.GE. JPRGYR .OR. MY. GE. JPRSYR)  .AND.  IV.LE.4) THEN
        PPFEB (IDX, JCYDX, IV) - PPFEB(IDX, JCYDX, IV) *
     *    VCNT(ICYDX, IV)*JULMYR(JDX,IV)
C
        IF (COST (IDX, JCYDX, IV) . GT. 0 . 0 . AND. INDVL1. EQ. 2 )
     *    COST(IDX,JCYDX,IV)  -
     *    COST (IDX, JCYDX, IV)  - PPFEB (IDX, JCYDX, XV)
      ENDIF
C
C
   99 RETURN
C  End SMCOST
      END
                                                   76

-------
                                 Appendix A
SUBROUTINE SUMWDW
c

C.. SUMWDW auma emission reductions and coats across model years
C . . for each calendar
C
C. .Array Subscripts
C
C. .ADAT(2)
C. .ADET (2)
C. .ADIM(2)
C. .CATPFE (25, 10, 4)
C. .CATPRE (25, 10, 4)
C. .CIMFEE (25, 10, 4)
C. .CIMREP (25, 10, 4)
C. .COST (25, 10, 4)
C. .DETER(25,2,10,4)
C. .FULBEN(25,20,4)
C. .REDATP (25, 2, 10, 4)
C. .REDIM(25,2,10,4)
C. .TBEN(2)
C. .TONRED (25, 2, 10, 4)
C. .YRBEN(10,2)
C. .YRCOST(IO)
C
year, and store in appropriate variable



- ADAT (IP)
- ADET (IP)
- ADIM (IP)
- CATPFE (MYDX, JCYDX, IV)
- CATPRE (MYDX, JCYDX, IV)
- CIMFEE (IDX, JCYDX, IV)
- CIMREP (IDX, JCYDX, IV)
- COST (IDX, JCYDX, IV)
- DETER (IDX, IP, JCYDX, IV)
- FULBEN (MYDX, JCYDX, IV)
- REDATP (IDX, IP, JCYDX, IV)
- REDIM(IDX,IP, JCYDX, IV)
- TBEN(IP)
- TONRED (MYDX, IP, JCYDX, IV)
- YRBEN( JCYDX, IP)
- YRCOST (JCYDX)

C. .Variable Dictionary
C
C Name Source
C AATFFE Block
C AATPRE Block
C ADAT Block
C
C ADET Block
C
C ADFBEN Block
C
C ADIM Block
C AEVAP Block
C AIMFEE Block
C AIMREP Block
C APFBEN Block
C CATPFE Block
C
C CATPRE Block
C
C
C CIMFEE Block
C
C CIMREP Block
C
C
C COST Block
C
C DETER Block
C
C
C FULBEN Block
C
C
C ICY Local
C JCYDX Local
C IDX Local
C IFEVYR Block
C IFICY Local
C ILEVYR Block
C ILICY Local
C IP Local
C IV Local
C JDX Local
C MY Local
C MYDX Local
C NUMCY Local
C
C REDATP Block
C
C REDIM Block
C
C THEN Block
C
C TCST Block
C TONRED Block

Description
Final sum of the ATP inspection fees
Final sum of the ATP repair costs
Array containing the final sum of the ATP
benefits
Array containing the final sum of the I/M
deterence benefits
Containa the final sum of the fuel
economy benefits
Array containing the final sum of the I/M benefit
Final sum of the Evap+Runloas benefits in tons.
Final sum of the I/M inspection fees
Final sum of the I/M repair coats
Final sum of Purge/Pressure Fuel economy benefits
Array which holda and auma the ATP fees for each
model year, calendar year and vehicle type
Array which holda and auma the ATP repair costs
for each model year, calendar year and vehicle
type
Array which holda and auma the I/M fees for each
model year, calendar year and vehicle type
Array which holda and sums the I/M repair coats
for each model year, calendar year and vehicle
type
Array which holda and sums the total coat for
each model year, calendar year and vehicle type
Array which containa the pollutant reductions for
each model year, calendar year and vehicle type
due to tamering deterence.
Array which holds and auma the fuel economy
benefit for each model year, calendar year and
vehicle type
Calendar year
Calendar year index
Model year index (forward 1-25)
First calendar year evaluated
Index of firat calendar year evaluated
Laat calendar year evaluated
Index of last calendar year evaluated
Pollutant index (1 - HC ; 2 = CO)
Vehicle class index
Model year index (backwards 20-1)
Model year
Model year index
Number of calendar years which are being
evaluated (number of scenario records - 1)
Array which holds and sums the ATP benefit for
each model year, calendar year and vehicle type
Array which holda and sums the I/M benefit for
each model year, calendar year and vehicle type
Array containing the final total benefit from
I/M deterence, I/M and ATP
Final total cost of all control programs
Array which passea the total benefit
                                      77

-------
                                             Appendix A
c
c
c
c
c

c
















c

c

c
c

c
YRBEN Block Array which contains the individual calendar
year benefits
YRCOST Block Array which contains the individual calendar
year costs

REAL JULMYR

COMMON /COST02/
COMMON /COST14/
COMMON /COST16/
COMMON /COST17/
COMMON /COST18/
COMMON /COST13/
COMMON /COST21/
COMMON /COST20/
COMMON /COST27/
COMMON /COST35/
COMMON /COST36/
COMMON /COST37/
COMMON /COST46/
COMMON /COST47/
COMMON /COST48/
COMMON /COST49/

COMMON /COST50/
COMMON /COS-ISO/
COMMON /COST51/


COMMON /MAXIMA/




IFEVYR, ILEVYR
TOURED (25, 2, 10, 4) , COST (25, 10, 4)
DATPFE, DIMFEE, DIMREP, DATPRE, AATPFE, AIMFEE
DETER(25,2,10,4),TDETER(25,2,10,4)
AIMREP, AATPRE, CFACTR, CATPFE (25, 10, 4)
CIMFEE (25, 10, 4) ,CIMREP (25,10, 4)
CATPRE (25, 10, 4) ,FOLBEN (25, 20, 4)
TDOCT (25, 2, 3, 10, 4) , TEVAP (25, 10, 4)
IGRFLG, IATPLP , IDIFF, OUTPRF
TCST, THEN (2) , ADFBEN, ADIM (2) ,ADAT(2) ,ADET(2)
YTONS (2, 10) ,BTONS (2) , BIDTON (2) , AEVAP
YRCOST (10) , YRBEN (10, 2) , YCOUNT (39) , ACOUNT
REDIM(25,2,10,4),REDATP(25,2,10,4)
PPFAIL(25,3,10,4),PPFIXR(25,10,4)
PR6FEE,PRSFEE,CEVFEE(25,10,4) , TRNFEE
PPFEB (25, 10, 4) ,APFBEN

TRLOSS (25, 4, 3) , TREVAP (25, 4, 3)
TRLOSS (25, 4, 3) , TREVAP (25, 4, 3) , TRSTLS (25, 4, 3)
ARLOSS , AREVAP , ARS TLS


MAXVEH, MAXLTW, MAXPOL, MAXRE6, MAXYRS

IFICY - IFEVYR - 1981
ILICY - ILEVYR - 1981
C
C
c










c






c
c
c

Initialize summing

TCST - 0.0
AIMFEE - 0.0
AIMREP - 0.0
AATPRE - 0.0
AATPFE - 0.0
ADFBEN =0.0
AFFBEN - 0.0
AEVAF - 0.0
ARLOSS - 0.0
AREVAP -0.0

DO 10 IP - 1,2
TBEN(IP) - 0.0
ADIM(IP) - 0.0
ADAT(IP) - 0.0
ADET(XP) -0.0
10 CONTINUE




variables



















over vehicle type, model year and calendar v

C
c
c
c
   DO 60 JCYDX-1, IDIFF
     YRCOST(JCYDX) -  0.0
     DO 12 IP - 1,2
       YRBEN(JCYDX,IP)  -  0.0
12   CONTINUE
     DO 50 17-1,4
       ICYDX -  (IFEVYR-1981)  •»• JCYDX - 1
       ICY - 1981 + ICYDX
       DO 40 IDX - 1,MAXYRS
         JDX - MAXYRS + 1 - IDX
         MYDX - MAXYRS  +  ICYDX -  JDX
         MY - MYDX +  1957

         TCST   - TCST    + COST(IDX,JCYDX,IV)
         AIMFEE = AIMFEE+CIMFEE(IDX, JCYDX, IV) +CEVFEE (IDX, JCYDX, IV)
         AIMREP » AIMREP  + CIMREP (IDX, JCYDX, IV)
         AATPRE m AATPRE  + CATPRE (IDX, JCYDX, IV)
         AATPFE » AATPFE  + CATPFE (IDX, JCYDX, IV)
         ADFBEN - ADFBEN  + FULBEN (IDX, JCYDX, IV)
         APFBEN - APFBEN  + PPFEB (IDX, JCYDX, IV)

Sum cost over vehicle class and model  year.
As of 10/7/91 this includes PP FE benefit
                                                   78

-------
                                             Appendix  A
c
c
c
c
c
         YRCOST (JCYDX)  - YRCOST (JCYDX)  + COST (IDX, JCYDX, IV)

Sum benefits over vehicle claaa, model year and
calendar year for each pollutant.  Alao sum over
vehicle claaa and model year only.
         DO 20 IP-1,2
           TBEN(IP)  -  TBEN(IP)
           ADIM(IP)  -  ADIM(IP)
           AD AT (IP)  =  ADAT (IP)
           ADET(IP)  =  ADET(IP)
                                   + TONRED(IDX,IP,JCYDX,IV)
                                   + REDIM(IDX,IP,JCYDX,IV)
                                   + REDATP (IDX, IP, JCYDX, IV)
                                     DETER(IDX,IP,JCYDX,IV)
   20
              YRBEN(JCYDX,IP)  - YRBEN(JCYDX,IP)+TONRED(IDX,IP,JCYDX,IV)
            CONTINUE
C
C
C
Sum evaporative emission benefita

         AEVAP  - AEVAP + TEVAP (IDX, JCYDX, IV)
   15
            DO 15 IMC=1,3
               AREVAP - AREVAP + TREVAP (IDX, IV, IMC)
               ARLOSS - ARLOSS + TRLOSS (IDX, IV, IMC)
            CONTINUE
 C
 C
 C
 C
 40      CONTINUE
 50   CONTINUE
 60 CONTINUE

 Compute annual averages changing to thousands of tons and
 thousands of dollars.
       TBEN(l)
       TBEN(2)
       TCST

       AATPFE
       AIMFEE
       AIMREP
       AATPRE
       ADFBEN
       APFBEN

       AEVAP
       ADIM(l)
       ADIM(2)
       ADAT(l)
       ADAT(2)
       ADET(l)
       ADET(2)

       ARLOSS >
       AREVAP *
              TBEN(l)  * .001 / IDIFF
              TBEN(2)  * .001 / IDIFF
              TCST    * .001 / IDIFF
AATPFE *
AIMFEE *
AIMREP *
AATPRE *
ADFBEN *
APFBEN *
AEVAP *
ADIM(l)
ADIM(2)
ADAT (1)
ADAT (2)
ADET(l)
ADET (2)
.001 / IDIFF
.001 / IDIFF
.001 / IDIFF
.001 / IDIFF
.001 / IDIFF
.001 / IDIFF
* .001 / IDIFF
.001 / IDIFF
.001 / IDIFF
.001 / IDIFF
.001 / IDIFF
.001 / IDIFF
.001 / IDIFF
             ARLOSS * .001 / IDIFF
             AREVAP * .001 / IDIFF
 C

 C
    DO 30 JCYDX - 1,IDIFF
      YRCOST (JCYDX)  - YRCOST (JCYDX)  *  .001
      YRBEN(JCYDX,!) - YRBEN(JCYDX,1) *  .001
      YRBEN(JCYDX, 2) - YRBEN (JCYDX, 2) *  .001
 30 CONTINUE

    RETURN
 End SUMHDW
    END
                                                    79

-------
                                             Appendix A
      BLOCK DATA
C
C
C
C
C
C

C
C
C
G
C
C
C

C
C
C
C
   COMMON /COST09/ DRSTRN(5),ZMSTRN(5>
   COMMON /COST10/ FEBNFT(25,10,4),FECNMY(20,4)
   COMMON /COSTll/ PCNTSV(4)
   COMMON /COST24/ ATEQP(20,5,4)
   COMMON /COST38/ STOCK(12,8),CLDIST(8)


ZMSTRN(ITEST,ICUT) is the zero-mile intercept used to calculate  the
failure rates for 1981 and newer vehicles.  Idle, 2-Speed, Loaded,
IM240(.8/15),NOx(10% fail-1.69 MPFI & 2.50 TBI fi 3.99 Garb)

   DATA ZMSTRK / 0.0,0.0,0.0252,0.0,0.032936/

DRSTRN(ITEST,ICUT) is the rate per 10k miles used to calculate the
failure rates for 1981 and newer vehicles.  Idle, 2-speed, Loaded,
IM240,NOx

Revise IM240 back to initial rate 11/18/91

   DATA DRSTRN / 0.01,0.01,0.01190,0.0373,0.0084805/

FECNMY is the miles-per-gallon average for each model year
by vehicle class.  (MY 1968 - 1987+)

   DATA FECNMY/
  1 13.9, 13.9, 13.2, 13.1, 12.9, 12.6, 13.9, 14.9, 15.6, 16.7,
  1 18.5, 19.6, 21.8, 23.3, 24.6, 26.0, 27.4, 28.8, 30.2, 31.6,
  2 10.6, 10.6, 10.4, 10.2,  9.9,  9.6, 12.0, 12.6, 13.8, 14.3,
  2 15.2, 16.3, 18.1, 18.4, 18.9, 19.5, 20.2, 21.1, 22.0, 22.9,




C
C
C
C

C
C
C
C
C
C
C
C
C
C
C

C









C






C
3 7.9, 7.9, 7.7, 7.4, 7.0, 6.9, 8.8, 9.7, 9.4,
3 9.8, 11.5, 13.3, 13.8, 14.3, 14.9, 15.4, 16.0, 16.6,
4 7.9, 7.9, 7.7, 7.4, 7.0, 6.9, 8.8, 9.7, 9.4,
4 9.8, 11.5, 13.3, 13.8, 14.3, 14.9, 15.4, 16.0, 16.6,

FCNTSV is the average fuel savings by technology group.
Pre-81 2spd/idle, 81+ 2spd/idle, 83+ IM240, Purge/Press

DATA PCNTSV/ 0.0, 0.080, 0.126, 0.059 /

ATEQP determines the fraction of each model year in each
class equipped with each tampering component .

ATEQP CODE:
1-air
2-cat
3"pcv
4-evap
5»gas cap

DATA ATEQP/
LDGV
1 .00, .05, .05, .05, .10, .30, .30, .45, .40, .30,
1 .30, .30, .65, .85, .70, .60, .60, .40, .40, .30,
2 .00, .00, .00, .00, .00, .00, .00, .80, .85, .85,
2 .90, .90, .95,1.00,1.00,1.00,1.00,1.00,1.00,1.00,
3 .00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,
3 1.00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,
4 .00, .00, .00, .00,1.00,1.00,1.00,1.00,1.00,1.00,
4 1.00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,1.00,
5 20*1.00,
LDGT1
1 .00, .05, .05, .05, .10, .30, .30, .40, .40, .30,
1 .30, .50, .50, .50, .50, .50, .50, .50, .50, .50,
2 .00, .00, .00, .00, .00, .00, .00, .70, .80, .75,
2 .75, .80, .80,1.00,1.00,1.00,1.00,1.00,1.00,1.00,
3 .00, 19*1.00,
4 .00, 19*1.00, 20*1.00,
9.6,
17.2,
9.6,
17. 2/


Repair



vehicle



























LDGT2 IVTAM4-1 IVTAM4-2 IVTAM4-3 IVTAM4-4
     * 11*.00,9*.50, 11*.00,9*1.(
     * 20*1.00,
C  HDGV   IVTAM4=1
                       IVTAM4=2
     * 11*.00,9*.50, 11*.00,9*1.
     * 20*1.OO/

      END
                                   IVTAM4=3
                             00, .00,19*1.00,
3*.00,8*.05,9*1.00,

     IVTAM4=4
3*.00,8*.05,9*1.00,
                                                   80

-------
                                    Appendix  A
SUBROUTINE PRINT(1SIZE1,ISIZE2)
c
c.
c
c
c.
c
c.
c
c.
c
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c.
c
c.
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c

.PRINT lists the

overall emission benefits and costs far
the indicated calendar years combined. No discounting'"
.Called by IAIN



.Calls PRNIM,PRNATP,PRNCST,PRNBEN

.Array Sub script a

.ADAT (2)
.ADET(2)
.ADIM(2)
.BTONS(2)
.IVPRT(4)
.KCOV(2,2>
.KCOVST(2,2)
.KFREQ<2,2)
.KTYPE(4,2)
.LVTFLG(4)
.»OYES(2)
.TBEN(2)
.TTYPE(4,3)
.V6RORT(8)
.VNAME(4)
.WAIVER (2)
.YCOUNT(46)
.YRBEN(10,2)
.YRCOST(IO)
.YTONS(2,10)




- ADAT (IP)
- ADET(IP)
- ADIM(IP)
- BTONS (IP)
- IVPRT(NCOVER),IVPRT(I2ND)
- KCOV(ICH,IAFXFG)
- KCOVST(ICH,IAFXFG)
- KFREQ (ICH, IFREQ)
- KTYPE(ICH,ATPPGM)
- LVTFLO(IVTAM)
- NOYES (IMTFLS)
- TBEN(IP)
- TTYPE(ICH,ATPPGM)
- VGRORT(KI)
- VNAME (XVPTR (NCOVER) )
- WAIVER (ITECH)
- YCOUNT(IY)
- YRBEN(IY,IP)
- YRCOST(IY)
- YTONS(2,IY)

•Variable dictionary

Name Source
AATPFE Block
AATPRE Block
ACODNT Block
ADAT Block

AOET Block

ADFBEN Block

ADIH Block
AIMFEE Block
AIMREP Block
Al'FFEE Block
BTONS Block

BYEAR Block
CLDIST Block
COMMA Local
CSTIM Block
CUTCO Local
CUTHC Local
GASCST Block
SHORT BLock

IAFXFG Block

IATPLP Block

ICH Local
ICO Local
ICUTS Block
ICYIM Block
ICY2 Block
IPEVYR Block
IFREQ Block

IFICY Local
ILEVYR Block
ILICY Local
IMFXFG Block

IMTFIiG Block
IOOREP Block
ISIZE1 Block


Description
Final sum of the ATP inspection fees
Final sum of the ATP repair costs
Total vehicle count over all scenario years
Array containing the final sum of the ATP
benefits
Array containing the final sum of the I/M
deterence benefits
Contains the final sum of the fuel
economy benefits
Array containing the final sum of the I/M benefit
Final sum of the I/M inspection fees
Final sum of the I/M repair costs
Anti-tampering inspection fee
Array containing the number of tons of pollutant
without a control program
The fleet size base year
Vehicle class registration distribution
Variable holding the character ' , '
I/M inspection fee
Variable holding character data pertaining to CO
Variable holding character data pertaining to HC
The cost of one gallon of unleaded gasoline
The vehicle fleet growth rate by vehicle class
converted into percent for printing
Flag indicating whether ATP has fixed or floating
model year coverage
Flag indicating whether an ATP only scenario is
to be run
DO loop counter
DO loop counter
I/M outpoint index from MOBILE 4
Tampering deterence start year
Calendar year
First calendar year evaluated
Flag indicating whether the I/M program is
annual or biennial
Index of first calendar year evaluated
Last calendar year evaluated
Index of last calendar year evaluated
Flag indicating whether the I/M program has a
fixed or floating model year coverage
Mechanic Training flag
Output Device number
Variable containing either the size of the
floating model year window for I/M coverage or
                                         81

-------
C
c
c
c
c
c.
c
c
c
c
c
                                            Appendix A

                     the earliest model year inspected in an I/M
                     program
                     Variable containing either the size of the
                     floating model year window for I/M coverage or
                     the earliest model year inspected in an I/M
                     program
                     Number of input stringency groups
                     I/M start year
                     DO loop counter calling subroutine ATPOUT
                     Array of flags indicating which emission
                     control components are to be inspected by the
                     ATP
                     I/M test type from MOBILE4
                     I/M program type       1 = Centralized
                                            2 «• Decentralized
                     DO loop index
                      Flag indicating whether ATP program is annual
                     or biennial
                     Variable holding the character data :
                      'floating' and 'fixed'
                     Array holding a string of character data
                     Array holding the character data  :
                      'annual' and 'biennial'
                     DO loop index
                     Array holding the character data  :
                      'centralized' and 'decentralized'
                     ATP start year
                     Counter variable  1-4
                     Mechanic training program character string
                     Variable holding the project ID character
                     string
                     Array containing the final total benefit from
                     I/M deterence, I/M and ATP
                     Vehicle fleet size in base year
                     Final total cost of all control programs
                     Variable holding the character data 'idle'
                     Array containing the growth rates for each
                     vehicle class
                     Array holding character data
                     Array containing the waiver rates for old
                     and new technology vehicles
                     Variable which contains the old tech waiver rate
                     Variable which contains the new tech waiver rate
                     The fleet count in a particular calendar year.
                     Total benefits in each evaluted calendar year.
                     Array which contains the individual calendar
                     year benefits
                     Array which contains the individual calendar
                     year costs
      INTEGER ALHFLG,ATPFLG,TPDFLG,RLFLAG,OUTFMT

      COMMON /COST28/ JATP, JIM, ISTRT, JCYDX, JPRGYR, JPRSYR
      COMMON /PROJEC/ PROJID (20)

      DIMENSION IDATE(3),ITIME(3)

      just use hh:mm:sec of time w/o hundreths of second

      CHARACTER* 4 PROJID
      CHARACTER*8 CDATE,CTIME

      IREP - 3

   MTS date/time...

      CALL ANSITM(ITIME)
      WRITE (IREP, 100) PROJID, IDATE (2) , IDATE (3) , IDATE (1) ,
     *                       ITIME(1),ITIME(2),ITIME(3)
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c


ISIZE2



ISTRIN
ISTRT
IT
DISTYP


ITEST
INTYP

IVTAM
ATPFQT

KCOV

KCOVST
KFREQ

KII
KTYPE

LAPSY
NCOVER
NOTES
PROJID

THEN

TCOUNT
TCST
TTYPE
VGRORT

VNAME
WAIVER

MAIV1
MAIV2
YCOUNT
YTONS
YRBEN

YRCOST





Block



Block
Block
Local
Block


Block
Block

Local
Block

Local

Local
Local

Local
Local

Block
Local
Block
Block

Block

Block
Block
Local
Block

Local
Block

Local
Local
Block
Block
Block

Block



C 100 FORMAT('ICost Effect. Model  (CEM4.1)
      *        IX,' 
-------
                                             Appendix A

C  MacFortran date/time...
C
C     CALL DATE(IDATE(2),IDATE(3),IDATE(1))
C     CALL TIME(ISECS)
C     ITIME(l) - ISECS/3600
C     ISECS - ISECS  -  (ITIME(1)*3600)
C     ITIME<2) - ISECS/60
C     ITIME(3) - ISECS  - ITIME(2)*60
C     WRITE(IREF,100) EROJID,IDATE(2),IDATE(3),IDATE<1),
C    *                        ITIME(1),ITIME(2),ITIME(3)
C 100 FORMAT('ICoBt  Effect.  Model (CEM4.1):  ',20A4,
C     *        1X,'
-------
                                       Appendix A
SUBROUTINE PRNIM(IREP, ISIZE1)
C
C. .PRNIM lists the
C
C. .Called by PRINT
C
C. .Array Subscript a
C
C. .ILDT(4)
C. .IMSTR(4,20)
C. .IVPTR(4)
C. .KCOV(2,2)
C. .KCOVST(2,2)
C. .KFREQ(2,2)
C. .KTYPE(4,2)
C. .NEWSTR(4,100)
C. .TTYPE(4,3)
C. .VNAME(4)
C. .WAIVER (2)
C

I/M program parameters





- ILDT(IVTAM)
- IMSTR (IV, NMY)
- IVPTR (NCOVER) , IVPTR (I2ND)
- KCOV (ICH, IAFXFG)
- KCOVST (ICH, IAFXFG)
- KFREQ (ICH, IFREQ)
- KTYPE (ICH,ATPPGM)
- NEWSTR (IT, ISTRN)
- TTYPE (ICH, ITEST)
- VNAME (IVPTR (NCOVER))
- WAIVER (ITECH)

C. .Variable dictionary
C
C Name Source
C COMMA Local
C GRIM Block
C I Local
C ICH Local
C ICYIM Block
C IFREQ Block
C
C IMFXFG Block
C
C IMSTR Local
C INTYP Block
C
C IREP Local
C ISIZE1 Block
C
C
C
C ISTRIN Block
C ISTRT Block
C ITEST Block
C IV Local
C
C
C
C KCOV Local
C
C KCOVST Local
C KFREQ Local
C
C KTYPE Local
C
C NEWSTR Block
C
C
C
C
C NCOVER Local
C TTYPE Local
C VNAME Local
C WAIVER Block
C
C WAIV1 Local
C WAIV2 Local
C

Description
Variable holding the character ' , '
I/M program compliance rate
DO loop counter
DO loop counter
Tampering deterence start year
Flag indicating whether the I/M program is
annual or biennial
Flag indicating whether the I/M program has a
fixed or floating model year coverage
Stringency by model year and vehicle type
I/M program type 1 = Centralized
2 « Decentralized
Output device indicator
Variable containing either the size of the
floating model year window for I/M coverage or
the earliest model year inspected in an I/M
program
Number of input stringency groups
I/M start year
I/M test type from MOBILE4
Vehicle type 1 - LDGV
2 - LDGT1
3 - LDGT2
4 - HDGV
Variable holding the character data :
'floating1 and 'fixed1
Array holding a string of character data
Array holding the character data :
1 annual ' and ' biennial '
Array holding the character data :
'centralized' and 'decentralized1
Contains stringency information
1 : Contains vehicle class code
2 : Contains first model year covered
3 : Contains last model year covered
4 : Contains stringency value
Counter variable 1-4
Variable holding the I/M teat type character data
Array holding vehicle class character data
Array containing the waiver rates for old
and new technology vehicles
Variable which contains the old tech waiver rate
Variable which contains the new tech waiver rate

CHARACTER*8 VNAME (4)
CHARACTER*4 KTYPE (4, 2) ,KCOV(2,2) ,KCOVST (8, 2)
CHARACTER*4 TTYPE(4,3),KFREQ(2,2)
CHARACTER* 4 NOYES, COMMA, PERIOD
CHARACTER*! COLON

COMMON /COST12/  IMFXFG, IAFXFG, IMCSFG, IAFEFG, IALDV
COMMON /COST28/  JATP, JIM, ISTRT, JCYDX, JPRGYR, JPRSYR
COMMON /IM240P/  DSIZE(25,4),IM24YR,IPRGYR,IPRSYR
COMMON /IMPAR1/  ICYIM, ISTRIN, MODYR1, MODYR2, WAIVER (2) ,CRIM
COMMON /IMPAR2/  ILDT (4) , ITEST,NODATA(2) ,NLIM, IMNAME (20, 9)
COMMON /IMPAR6/  IFREQ, INTYP
COMMON /STRING/  NOYES(2),COMMA,PERIOD,COLON
                                             84

-------
                                             Appendix A

c
      DIMENSION IVPTR (4)
      DIMENSION IMSTR(4,20)
C
      DATA KCOV /
     *   'Floa','ting',
     *   'Fixe','d    '  /
      DATA KCOVST  /
     *   'Numb','er o','f mo','del ','year','a in',' win','dow:',
     *   'Olde','at m','odel1,'  yea','r co-,'vere','d:   ','     ' /
      DATA KFREQ /
     *  'Annu','al   ',
     *  "Bien1,'nial'  /
      DATA KTYPE /
     *  'Cent','rali','zed ','     ',
     *  'Dece','ntra','lize','d   '  /
      DATA TTYPE/'Idle','    ','     ','     ',
     2            '2500',' rpm','  / I','die ',
     3            'Load','ed /','  Idl','e   '/
      DATA VNAME/'LDGV    ','LDGT1   ' , 'LDGT2    •,'HDGV    ' /
C
C
      I1-MODYR1-1900
C     WRITE(IREP,100)  ISTRT,(1,1-11,83,1)
      WRITE(IREP,100)  ISTRT,ISTRIN
  100 FORMAT('0',55X,'I/M Program Selected:'/
     *   130('_'),/,
     *    '     I/M  start  year (January 1)  :      ',I4,25X,
     *    'Stringency:',IX,12 )
C
      I2-83-I1+1
      DO 20 IV-1,4
      DO 20 NMY-1,12
         IMY-MODYR1+NMY-1
         IMSTR(IV,NMY)-0
         IF (IMY. GE. MODYR1. AND. IMY. LE. MODYR2)
     *    IMSTR(IV,NMY)-ISTRIlf
    10    CONTINUE
    20 CONTINUE
      I3-80-I1+1
      WRITE(IREP,101)  ICYIM
  101 FORMAT('  ','   Tampering deterrence start:        ',I4,30X)
C
      IF(IMFXFQ.EQ.l)  WRITE (IREP, 102)  (KCOVST (ICH, IMFXFG) , ICH-1, 8) ,
     *                 ISIZE1,
     *                 (KCOV(ICH,IMFXFG),ICH-1, 2),
     *                 (IMSTR(2,I),1-1,12,1)
  102 FORMAT(4X,8A4,IX,14,2X,2A4,20X,'LDGT1 ',16(12,IX))
      IF (IMFXFG.EQ. 2)  WRITE (IREP, 103)  (KCOVST (ICH, IMFXFG) , ICH-1, 8) ,
     *                 ISIZE1,MODYR2,
     *                 (KCOV(ICH, IMFXFG) , ICH-1, 2)
C    *                 (IMSTR(2,I), 1-1,12,1)
  103 FORMAT(4X,8A4,IX,14,'-',14,2X,2A4,1SX)
C
C
      WRITE(IREP,104)  (KFREQ(ICH,IFREQ),ICH-1,2)
C    *                 (IMSTR(3,1), 1-1,12,1)
  104 FORMAT('  ','   I/M inspection frequency:          ',2A4,26X)
C    *           'LDGT2 ',16(12,IX))
C
C
      WRITE(IREP,105)  (KTYPE(ICH,INTYP),ICH-1,4)
  105 FORMAT('  ','   I/M program type:                  ',4A4,8X)
C
C
      WRITE(IREP,106) (TTYPE(ICH,ITEST),ICH-1,4) ,  IM24YR
  106 FORMAT('  ','   1981 fi later MYR teat type:        ',4A4,
     *  T75,'IM240  test model year coverage:      ',14,'+')
C
C
      NCOVER-0
      DO 30 IV-1,4
         IF(ILDT(IV).EQ.l) GOTO 30
         NCOVER-MCOVER+1
         IVPTR (NCOVER) -IV
    30 CONTINUE
C
      IF (NCOVER.EQ.l)  WRITE (IREP, 107)  VNAME (IVPTR (1) ), JPRGYR
  107 FORMAT('  ','   Vehicle types covered:            ',A4,
     *  T75,'Purge  check  model  year coverage:      ',14,'+')
      NCOMMA-NCOVER-1


                                                   85

-------
                                             Appendix A

      IF(NCOVER.EQ.2)
     *  WRITE(IREP,108)  VNAME (IVPTR(l)) , COMMA,
     *                   VNAME (IVPTR (HCOVER) ) , JPRGYR
  108 FORMAT('  ','   Vehicle types covered:     ',7X,A4,A2,A5,
     * T75,'Purge check model year coverage:     ',14,' + ')
      IF (NCOVER.EQ. 3)
     *  WRITE (IREP, 109)  (VNAME (IVPTR(I) ) ,COMMA, I-1,HCOMMA) ,
     *                   VNAME (IVPTR (NCOVER) ) , JPRGYR
  109 FORMAT('  ','   Vehicle types covered:     ',7X,A4,A2,A5,A2,A5,
     * T75,'Purge check model year coverage:     ',14,'+')
      IF(NCOVER.EQ.4)
     *  WRITE (IREP, 110)  (VNAME(IVPTR(I)),COMMA,I-1,NCOMMA),
     *                   VNAME (IVPTR (NCOVER) ) , JPRGYR
  110 FORMAT('  ','   Vehicle types covered:     ',7X,A4,A2,2(A5,A2),A5,
     * T7S,'Purge check model year coverage:     ',14,'+')
C
      WAIV1 - WAIVER (1) * 100.
      WAIV2 = WAIVER(2) * 100.
      WRITE (IREP, 120) WAIV1, JPRSYR,WAIV2, GRIM
  120 FORMAT(  '    I/M waiver rate  (pre-1981):       ',F5.!,'%',
     * T75,'Pressure check model year coverage:  ',14,'+',/,
     *         '    I/M waiver rate  (post-1981):      ',F5.1,'%',/,
     *         '    I/M compliance rate:              ',F5.1,'%')
C
  999 RETURN
      END
                                                   86

-------
c
c
                                           Appendix  A

    SUBROUTINE PRNATP (IRJEP, ISIZE2)

    INTEGER ATPPGM,ATPFQT,DISTYP

    CHARACTER*8 VNAME(4)
    CHARACTER*4 KTYPE(4, 2) ,KCOV(2, 2) ,KCOVST (8,2)
    CHARACTER*4 TINSP(3,8)
    CHARACTER*4 TTYPE (4,3) ,KFREQ (2,2)
    CHARACTER*4 NOYES,COMMA,PERIOD
    CHARACTER*! COLON

    COMMON /COST12/ IMFXFG,IAFXFG,IMCSFG,IAFEFG,IALDV
    COMMON /ATPAR1/ LAPSY,LAP1ST,LAPLST, LVTFLG (4)
    COMMON /ATPAR2/ ATPPGM,ATPFQT,CRATP,DISTYP (8)
    COMMON /STRING/ NOTES (2) , COMMA,PERIOD, COLON

    DIMENSION IVPTR(4) ,ITPTR(8)

    DATA KCOV /
   *  'Floa','ting',
   *  'Fixe','d    ' /
    DATA KCOVST /
   *  'Numb','er o','f mo
   *  'Olde','at m1,'odel','
    DATA KFREQ /
   * 'Annu','al  ',
   * 'Bien','nial' /
    DATA KTYPE /
   * 'Cent','rail','zed  ', '
   * 'Decs','ntra','lize1, 'd
                               del ','ye-ax1,'• in',1 win','dow: ',
                                yea','r oo','vere1,'d:   ','     '  /









C
c
DATA
*
*
*
*
*
*
*
DATA


TINSP/'Air '
'Cata1
'Fuel'
'Plum'
'EGR '
'Evap1
'PCV '
'Gaa '
VNAME/ 'LDGV


'Pump*
'lyat'
' Inl1
'btea'
'Syst'
• sya'
1 Syat '
•Cap '
', 'LD




et
mo
em
tern
em
/
GT1 '


                                        1LDGT2
                                                    'HDGV
 C
 C
C
C
C
C
C
C
    WRITE (IREP, 101) LAPSY,CRATP
101 FORMAT('0', SOX,'Anti-Tampering Program Selected:',/,
   *  130C '),/,
   *       T    ATP atart  year (January 1):       ',I4,25X,
   *        '   ATP  compliance rate:                ',F5.1,'%')


    WRITE (IREP, 102)  (KFREQ (ICH,ATPFQT) ,ICH-1,2) ,
   *                   (KTYPE(ICH,ATPPGM),ICH-1,4)
102 FORMAT('  ','   ATP  inapection frequency:         ',2A4,21X,
   *        '  ','   ATP  program type:                    ',4A4)


    IF (IAFXFG.EQ.l) WRITE (IREP, 103) (KCOVST(ICH,IAFXFG),ICH-1,8),
   *                                 1SIZE2
103 FORMAT('  ','    ',8A4,4X,I4,2X,2A4)
    IF(IAFXFG.EQ.2) WRITE(IREP,104) (KCOVST(ICH,IAFXFG),ICH-1,8),
   *                                 ISIZE2,LAPLST,
   *                                 (KCOV (ICH, IAFXFG) , ICH-1, 2)
104 FORMAT('  ','    ',8A4,IX,14,'-',14,2X,2A4)


    NCOVER-0
    DO 10 TV-1,4
      IP(LVTFLG(IV) .EQ.l)  GOTO 10
      NCOVER-NCOVER+1
      IVPTR(NCOVER)-IV
 10 CONTINUE

    IF (NCOVER. EQ.l) WRITE (IREP, 105) VNAME (IVPTR (1) )
105 FORMAT('  ','   Vehicle types  covered:           ',A4,A4)
    NCOMMA=NCOVER-1
    IF(NCOVER.GT.l)
   *  WRITE (IREP, 106)  (VNAME (IVPTR (I) ), COMMA, 1=1, NCOMMA) ,
   *                    VNAME (IVPTR (NCOVER) )
106 FORMAT('  ','   Vehicle typea  covered:    ',7X,3(A5,A2),A5)
      NCOVER-0
                                                   87

-------
                                             Appendix A

      DO 20 IT=1,8
        IF(DISTYP(IT).EQ.l) GOTO 20
        NCOVER=NCOVER+1
        ITPTR (NCOVER) =IT
   20 CONTINUE
C
      IF(NCOVER.EQ.l) WRITE(IREP,107)  (TINSE 
-------
                                             Appendix  A

      SUBROUTINE PRNCST(IREP)
C
      INTEGER BYEAR
      INTEGER PROMPT, TAMFLG, SPDFLG,VMFLAG
C
      COMMON /VMXCOM/  REGMIX(8) , TFNORM(8) ,VMTMIX(8) ,VCOUNT (39, 8)
C
      COMMON /COST01/  ATPFEE
      COMMON /COST07/  CSTIM,GASCST,IMSTRT,ISCEN,MOVIM,MOVATP,REPIM(4)
      COMMON /COST22/  ATCOST(2,8)
      COMMON /COST27/  IGRFLG,1ATPLP,IDIFF,OUTPRF
      COMMON /COST28/  JATP, JIM,ISTRT, JCYDX, JPRGYR, JPRSYR
      COMMON /COST34/  TCOUNT, BYEAR, VGRORT
      COMMON /COST38/  STOCK(12,8),CLDIST(8)
      COMMON /FLAGS1/  PROMPT,TAMFLG,SPDFLG,VMFLAG,OXYFLG,DSFLAG
      COMMON /COST48/  PRGFEE,PRSFEE,CEVFEE(25,10,4),TRNFEE
C
      DIMENSION IVPTR(4),DIST(8)
C
      DO 10 1-1,8
        DIST(I) -  REGMIX(I)*100.
   10 CONTINUE
C
      IF (JIM. EQ. 2  .AND.  JATP.EQ.2) THEN
C
      WRITE(IREP,101)  CSTIM,  (DIST (KJ) ,KJ-1, 8),
     *                 ATPFEE, VGRORT
  101 FORMAT (
     *  '0',11X,'Cost Assumptions',31X,'Flaet Assumptions',5X,
     *        'LDGV  LDT1  LDT2  HDGV  LDDV  LDDT  HDDV   MC   ',/,
     *   43('  '),12X,75('_'),/,
     *  '   I/M inspection cost:  (per inap) $',F6.2,15X,
     *        '  Vehicle  percentages:  ',8(FS.1,1X>,/
     *  '   ATP inapection coat:  (per insp) $',F6.2,15X,
     *        '  Growth rate:  ',F4.1,' % per year'
     *   )
C
      ELSE IF(JIM.EQ.l .AND.  JATP.EQ.2)  THEN
C
C     WRITE (IREP, 104)  ATPFEE,(VGRORT(KI),KI-1,8),
C    *                 (DIST (KJ), KJ-1,8)
      WRITE (IREP, 104)  ATPFEE, (DIST (KJ) ,KJ-1, 8) , VGRORT
  104 FORMAT(
     *  '0',11X,'Coat Assumptions',31X,'Fleet Assumptions', 5X,
     *        ' LDGV  LDT1  LDT2  HDGV  LDDV  LDDT  HDDV   MC' , /,
     *      43('  '),12X,75('_'),/,
     *  '   ATP Tnapection coat:  (per insp) $',F6.2,15X,
     *        '  Vehicle  percentages:  ',8(F5.1,IX),/,
     *  '                                     ',6X,15X,
     *        '  Growth rates  ',F4.1,' % per year' )
C
      ELSE IF(JIM.EQ.2 .AND. JATP.EQ.l)  THEN
C
      WRITE (IREP, 105)  CSTIM, (DIST (KJ) ,KJ-1, 8) , VGRORT
  105 FORMAT(
     *  '0',11X,'Coat Assumptions',31X,'Fleet Assumptions', 5X,
     *        'LDGV  LDT1  LDT2  HDGV  LDDV  LDDT  HDDV   MC   ',/,
     *     43('  -),12X,75(' '),/,
     *  '   I/M Tnapection coat:  (per inap) $',F6.2,15X,
     *        '  Vehicle  percentagea:  ',8(F5.1,1X),/,
     *  '                                     ',6X,15X,
     *        '  Growth rates  ',F4.1,' % per year' )
C
      END IF
C
      IF (JPRGYR.NE.2020  .OR. JPRSYR.NE.2020)
     *  WRITE(IREP,110)  PRGFEE,PRSFEE,TRNFEE
  110 FORMAT('   Purge inapect cost  (per insp)  : $',F6.2,/,
     *        '   Pressure inap cost  (per insp)  : $',F6.2,/,
     *        '   IM240 insp incremnt over Purge: $',F6.2)
C
      IF(IATPLP.EQ.l)  WRITE(IREP,102) GASCST,TCOONT,BYEAR
  102 FORMAT(   '    Avg.  Gasoline Coat(per gallon): $',F6.2,1SX,
     *    '  Fleet  aize is ',F9.0,' vehicles in base year  ',14)
C
C
      IF(IATPLP.EQ.2)  WRITE (IREP, 103) TCOUNT,BYEAR
  103 FORMAT(57X,
     *    '  Fleet  aize is ',F9.0,' vehicles in base year  ',14)
C
C


                                                   89

-------
                                             Appendix A

      IF ( JIM.EQ.2 .OR. JATP.EQ.2  .OR. JPRGYR.LT.2020  .OR.
     *  JPRSYR.LT.2020 ) WRITE(IREP,106)
  106 FORMAT(
     *       '0',5X,'  Repair Coata',12X,'Pre-81',4X,'81+',/,
     *           46('_')
     *       )
C
C
      IF (JIM.EQ.2) WRITE(IREP,107)(REPIM(KI),KI-1,4)
  107 FORMATC  Avg. I/M repair coat      : $ ',F4.0,4X,F4.0,
     *        '  ( IM240 repair - $ ',F4.0,' NOx  repair- $',F4.0,'  )'
     *      )
C
C
      IF (JATP.EQ.2) WRITE (IREP, 108) ( (ATCOST (KI,KJ) ,KI=1,2) ,KJ-1, 6)
  108 FORMAT (
                Air Pump repair coat      : $ ',F4.0,4X,F4.0,/,
                Catalyat replacement coat: $  ',F4.0,4X,F4.0,/,
                Miafueled catalyat coat   : $  ',F4.0,4X,F4.0,/,
                Evap. ayatem repair coat  : $  ',F4.0,4X,F4.0,/,
                PCV ayatem repair coat    : $  ',F4.0,4X,F4.0,/,
                Gas cap repair coat       : $  ',F4.0,4X,F4.
     *
C
      IF  (JPRGYR.LT.2020 .OR. JPRSYR.LT.2020)
     *   WRITE(IREP,109)((ATCOST(KI,KJ),KI-1,2),KJ-7,8)
  109 FORMAT(
     *       '  Purge repair coat         : $ ',F4.0,4X,F4.0,/,
     *       '  Preaaure repair coat      : $ ',F4.0,4X,F4.0
C    *         '   ( Purge+Preaa repair ™ Purge repair )'
     *       )
C
C
  999 RETURN
      END
                                                  90

-------
                                             Appendix  A

      SUBROUTINE PRNBEN(IREP)
C
      INTEGER BYEAR, ATPFLG, PROMPT, ATPFQT, ATPPGM, DISTYP
      REAL EVDOL,NETCST,NETBEN
C
      CHARACTER*8 VNAME(4)
      CHARACTERS KTYPE(4,2),KCOV(2,2),KCOVST(8,2)
      CHARACTER*4 TTYPE (4, 3) ,KFREQ (2, 2)
C
      COMMON /COST02/  IFEVYR,ILEVYR
      COMMON /COST16/  DATPFE, DIMFEE, DIMREP,DATPRE, AATPFE, AIMFEE
      COMMON /COST18/  AIMREP,AATPRE,CFACTR,CATPFE(25,10, 4)
C.. Add resting loaaea to output, 8/9/91
      COMMON /COST26/  FRSTLS (25, 8, 4, 10) , YRSTLS (10) ,BRSTLS
      COMMON /COST27/  IGRFLG,IATPLP, IDIFF, OUTPRF
      COMMON /COST28/  JATP,JIM, ISTRT, JCYDX, OPRGYR, JPRSYR
      COMMON /COST35/  TCST,TBEN(2) ,ADFBEN,ADIM(2) ,ADAT (2) ,ADET (2)
C   Add AEVAF  6/3/91
      COMMON /COST36/  YTONS(2,10),BTONS(2),BIDTON(2),AEVAP
      COMMON /COST37/  YRCOST(10),YRBEN(10,2),YCOUNT(39),ACOUST
      COMMON /COST40/  YEVTON (10) ,BEVTON, YEREF (10) ,BEREF
      COMMON /COST41/  YERLOS(10),BERLOS
      COMMON /COST49/  PPFEB(25,10,4),APFBEN
C
      COMMON /COST51/  ARLOSS , AREVAP , ARSTLS
C
C  Write heading
C
      WRITE (IREE, 101)  IFEVYR, ILEVYR
   101 FORMAT('  ',/,
     * 8X,'Benefita  (1000 tona/yr)  and Coats (1000 $/yr) ',
     * 'are averaged  over the calendar yeara ',14,' thru ',14,'.',/,
     *     130('_')  )
C
C       IFICY  = IFEVYR - 1974
C       ILICY  - ILEVYR - 1974
        IFICY  - IFEVYR - 1981
        ILICY  - ILEVYR - 1981
C       IDIFF  - ILEVYR - IFEVYR + 1
C
C...Separate out exhaust HC from total voc
C
      BEXTON - BTONS (1) -BEVTON-BEREF-BERLOS-BRSTLS
C
C  Write more  headings
C
      WRITE(IREP,103)
   103 FORMAT(
     * 35X,'Gal.',4X,' Program Benefits and Coata ',
     * 4X,'  Fleet   ',9X,'Emissions Without Program Elements  ',/,
     *   43X,28('  '),17X,50('  •),/,
     * 35X,'Year'74X, '  VOC       CO       Coat   ', 4X, '  Size   ',
     * 4X, ' ExhVOC      CO    Evap HC   Refuel  RnLoaa  Rat Loo  ',/,
     *   35X,4('  '),4X,2(7(' '),2X),10('_'),4X,9('_'),
     * 2X,6(2X,7T'_')) )
C
C  Write benefita  for each calendar year
C
      IYY  - IFICY  - 1
      DO 10 IY - 1,IDIFF
        IYY -  IYY + 1
        ICY2 - IY + IFEVYR - 1
        YEXTON - YTONS (1, IY)-YEVTON (IY)-YEREF 
-------
                                             Appendix A
c
c
c
      IF(JATP.EQ.l) THEN
        ATPBEN = ADAT(l)
        NETBEN - TBEN(l) 4- AEVAP
      ENDIF
      NETCST = TCST - APFBEH
      EXHBEN - TBEN(l) - AEVAP

   Write total benefit breakdown and total benefits

      WRITE (IREP,102) ATPBEN, ADAT (2) , AATPFE,
  102 FORMAT(
                      AEVAP, APFBEN,
                      ADIM(l) ,ADIM(2) ,AIMFEE,
                      AATPRE , AIMREP , ADFBEN ,
                      ADET(l) ,ADET(2) ,
                      NETBEN,TEEN(2),TCST,ACOUNT,
                      BEXTON,BTONS (2) , BEVTON,BEREF, BERLOS,BRSTLS
         43X,2(7('_'),2X),10('_'),/,
         Average Annual ATS Benefit/Fee
         Evap/RunLoaa HC s MEG Benefit
         Average Annual I/M Benefit/Fee
         Avg Ann. ATF + P/P Repair Cost
         Average Annual I/M Repair Coat
         Average Annual Fuel Savings
         Tampering Deterrence
                                         ,12X,2(F7.3,2X),F10.0,/,
                                         ,12X,F7.3,10X,'(',F10.0,')',/
                                         ,12X,F7.3,1X,F8.3,2X,F10.0,/,
                                         ,12X,18X,F10.0,/,
                                         ,12X,18X,F10.0,/,
                                         ,12X,17X,'(',F10.0,')',/,
                                         ,12X,2(F7.3,2X),/,
           43X,2(7('_'),2X),10('_'),4X,9('  '),2X,6(2X,7('_')),/,
         Evap    RnLoaa   ExhHC',5X,'Annual Average
     *
     *
     * F7.3,IX,F8.3,2X,F10.0,4X,F9.0,2X,2X,F7.3,1X,F8.3,4<2X,F7.3)
     * )
C
C  Calculate and write total benefits in kilograms per day.
C
      KBEHHC -  <(TBE»(1)/.365)*2000.)/2.2046
      KBENCO -  ((TBEN(2)/.365)*2000.)/2.2046
C
      WRITE(IREP,105) AREVAP,ARLOSS,EXHBEN,KBENHC,KBENCO,BTONS(1)
  105 FORMAT(3(F7.3,1X),17X,'(  ',17,19,'  kilograms per day  )',6X,
     *  '(  ',F7.3,' Total VOC )  ')
C
  999 RETURN
      END
                                                   92

-------
                                             Appendix  A

      SUBROUTINE EBSIZE (ID,OTR)
C
C
C     Calla Function  OTCALC
C
C     The Evap, Gas Cap  and Crankcaae tampering rates are stored
C     in BSIZE.
C
C        NHG-13  -  Evap Canister
C        NHG-14  -  Crankcaae
C        NHG-15  -  Gas  Cap
C
      COMMON  /LOOKUP/ IVTAM, IQG, IPG, JPGD,IHG,IGCSF
      COMMON  /MYCODE/ MY,IDX, JDX,LDXSY,LMYRVT,IAY, IMDXSY, IMKINK
      COMMON  /TAMEQ4/ BTR(9,2)
      COMMON  /COST32/ BSIZE(4,25,2,6,17,2),TRAT(9)
C
C  For MOBILE4, crankcaae (IH - 3)  processing does not change.
C     PCV Case
C
      IF(ID.EQ.7) THEN
        BSIZE(IVTAM,JDX,1,1,14,1) - OTR
C
C     EVAP  Canister & GAS Cap Cases
C
C     NOTE: OTCALC(i,j,k,l)
C
C        i  -  8 —  EVAP  + GAS Cap tampering rate
C               6 —  EVAP  only tampering rate
C        j  -  1 -  A Vehicle's tampering rate previous to ATP start
C               2 "  Previous + Subsequent  (Overall) tampering rate
C        k  -  3 -  ATP effectiveness code for GAS Cap
C               1 -  ATP effectiveness code for EVAP
C        1  -  1 •»  Previous tampering effectiveness code
C               2 -  Subsequent tampering effectiveness code
C
      ELSE  IF(ID.EQ.S)  THEN
C
C        No ATP Case
C
        IF(LMYRVT.EQ.l)  THEN
          BSIZE(IVTAM,JDX,1,1,13,1) - BTR(6,1)
          BSIZE (IVTAM,JDX, 1,1, 15,1) - BTR(8,1) - BTR(6,1)
C
C        ATP  Starts before the vehicle is built
C
        ELSE  IF(LMYRVT.EQ.2  .AND. LDXSY.GT. JDX) THEN
          BSIZE 
-------
                                             Appendix  A
      SUBROUTINE PRNLAP(IREP)
c
c
c
c
G
C
C
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c


PRNLAP writes scenario parameters


Called by PRINT.


common blocks :
/ALUHIN/
/CITRV1/
/FLAGS2/
/FLAGS3/
/FLAGS 4 /
/MAXIMA/
/REGION/
/RESUL1/
/RESUL2/
/RESUL3/
/SCENE!/
/STRING/
/TEMPS/
/USDATA/
/VMXCOM/

AC,XLOAD,TRAILR,ABSHUM,DB,WB
RVPICY
IMFLAG, ALHFLG
ATPFLG
PRTFLG, IDLFLG, NMHFLG, HCFLAG
MAXVEH
IRE JN, ALT
EFFTP , EFEXH, EFEVAP , EFLOSS , EFRUNL
EFIDLE
VFTP , VEXH , VEVAP , VLOS S , VRUNLS , VTDLE
SPD, PCCN, PCHC, PCCC
NOTES , COMMA, PERIOD
TEMEVP , TEMEXH
USNAME, NUSD, IUSD
VMTMIX

Local array subacripta:

NAMEVP (4)
NAMNMH (2)
NAMPOL(3)
NAMREG (3)
NAMVEH (10)


- NAMEVP ( IEVB )
- NAMNMH ( NMHFLG )
- NAMPOL ( IP )
- NAMREG ( IREON )
- NAMVEH ( IVP3 )

Local variable / array dictionary:


Name Type Description
IP1 I
IP2 I
see parameter dictionary
see parameter dictionary
NAMNMH C*8 prefix indicating type of HC (total vs non-methane)
NAMPOL C*4 pollutant names
NAMREG C*4 region names
NAMVEH C*8 vehicle class names
NWITHC I
number of substrings to be written with a comma appended
USC8 C*8 8 character underscore


C

c

c
 INTEGER ALHFLG
 INTEGER ATPFLG, TPDFLG, RLFLAG , TEMFLG , OUTFMT
 INTEGER PRTFLG,HCFLAG
 CHARACTER* 8 USNAME
 CHARACTER* 4 NOTES, COMMA, PERIOD
 CHARACTER*! COLON
 CHARACTER*! ASTMCL, ACLASS
 CHARACTER* 4 SCNAME

 COMMON /ALUHIN/ AC, XLOAD (3) , TRAILR (3) , ABSHUM, DB, WB
 COMMON /CITPAR/ SCNAME(4)
 COMMON /CITRV1/ RVPBAS,RVPIUS,RVPICY, IUSESY,RVPUWX
 COMMON /FLAGS2/ MYMRFG,NEWFLG, IMFLAG, ALHFLG
 COMMON /FLAGS3/ ATPFLG,TPDFLG,RLFLAG,LOCFLG,TEMFLG,OUTFMT
 COMMON /MAXIMA/ MAXVEH, MAXLTW, MAXPOL, MAXREG, MAXYRS
 COMMON /REGION/ FEET(2),IREJN,ALT,INITPR
 COMMON /SCENE!/ SPD(8),PCCN,PCHC,PCCC
 COMMON /STRING/ NOYES(2),COMMA,PERIOD,COLON
 COMMON /TEMPS/  AMBT, TEMMIN, TEMMAX, TEMEXH (3) , TEMEVP (6)
 COMMON /USDATA/ USNAME(4,4),NUSD,IUSD(4)

 CHARACTER*9 NAMVEH (8) , NAMNMH (2) ,USC8, NAMEVP (4)

 CHARACTER* 4 NAMREG (3 ) , NAMPOL (3)

 DATA NAMVEH/
         LDGV'
        LDGT1'
        LDGT2'
         HDGV1
         LDDV
         LDDT'
*        HDDV
*         MC ' /

 DATA NAMREG/'Low  ','High','Mid '/
                                                   94

-------
                                             Appendix  A

      DATA uses/'  	'/
c
      DATA NAMPOL/1 HC:',1 CO:','HOX:'/
C
      DATA NAMNMH/'Total    ','Non-Math'/
C
      DATA NAMEVP/'Hot  Soak  ','Diurnal  ','Multiple ','Crankcaae'/
C
      WRITE(IREP,100)
  100 FORMAT(SIX,'Local Parameters Selected: ',/,130('  ') )
C                                                      ~
      IF(NUSD.EQ.O) GOTO  10
      IF(NUSD.EQ.l) WRITE (IREP, 200) (USNAME (ICH, IUSD(1)) ,ICH-1, 4).
     *                                PERIOD
      NWITHC-NUSD-1
      IF(NWITHC.GT.O) WRITE(IREP,200)
     *   ( (USNAME (ICH, IUSD (ICT) ) , ICH-1, 4) , COMMA, ICT-1, NWITHC) ,
     *   (USNAME (ICH, IUSD (NUSD) ) ,ICH-1, 4) ,PERIOD
  200 FORMAT('  ','User  supplied',4A8,T46,A1,4A8,T78,A1,4A8,T110,A1/
     *        '  ',T1S,4A8,T46,A1,/)
C
    10 WRITE (IREP,215) SCNAME,TEMEXH,NAMREG(IREJN)
  215 FORMAT(
     *    T3,4A4,
     *    T38,'Ambient  Temp:',F5.1,2(' /',F5.1),' (F) ',
     *    185,'Region:  ',A4)
C
      WRITE (IREP, 220) PCCN,PCHC,PCCC,ALT
  220 FORMAT)
     *    T36,'Operating Mode:',F5.1,2(' /',F5.1),
     *    T83,'Altitude:',F6.0,'  Ft.')
C
      WRITE(IREP,300)
     *   RVPBAS,TEMMIN,RVPIUS,IUSESY,TEMMAX
  300 FORMAT('0',
     *            T5,  '    Base RVP: ',F4.1,
     *            T79, 'Minimum Temp:   ',F4.0,' (F)1,/,
     *            T5  ,'  In-uae RVP: ',F4.1,
     *            T35,'In-use Start Tr:  ',14,
     *            T79,'Maximum T«mp:   ',F4.0,' (F)')
C
      IF(ALHFLG.GT.l) WRITE(IREP,225)
     *    ABSHUM,
     *    AC,DB,WB
  225 FORMAT(
     *'  ',T19,'Absolute Humidity:',F6.2,
     *    T48,'AC (DB / WB):',F5.1,'  (',F5.1,' /',F5.1,')')
C
      WRITE(IREP,230)
     *    NAMVEH,(USC8,ITER-1,8),
     *    SPD
  230 FORMAT(
     *'0','  Vehicle  Type:',7X,8(2X,A8),/,
     *     '                ',8X,8(2X,A8),/,
     *'  ','    Veh.  Speeds:',7X,8F10.1)
C
      IF(ALHFLG.ST.l) WRITE(IREP,235)
     *    XLOAD,
     *    TRAILR
  235 FORMAT(
     *'  ',' Extra Load:',F11.3,2F10.3/
     *'  ','Trlr in Tow:',F11.3,2F10.3)
C
      RETURN
      END
                                                   95

-------
                                             Appendix A

      SUBROUTINE GETCEI
c
C  GETCEI reads in the Coat  Effective Inputs particular
C  to the Coat Effective model CEM4.1.
C
C  Called by MAIN.
C
C  Calls TRUEST.
C
C  Input on call:
C
C
C  Output on return:
C
C    common blocks:
C
C    /COST01/ ATPFEE
C    /COST07/ CSTIM,GASCST,IMSTRT,ISCEN,MOVIM,MOVATP,REPIM(4)
C    /COST12/ IMFXFG, IAFXFG, IMCSFG, IAFEFG, IALDV
C    /COST22/ ATCOST(2,8)
C    /COST27/ IGRFLG,IATPLP,IDIFF,OUTPRF
C    /COST34/ TCOUNT,BYEAR,VGRORT,NCOUNT
C    /COST38/ STOCK(12,8),CLDIST(8)
C    /COST48/ PRGFEE,PRSFEE,CEVFEE(25,10,4),TRNFEE
C    /YEARS4/ IY1941,IY1960,IY2020
C
C  Notes:
C
C  GETCEI was added for  CEM4.1 VERSION 2
C
C
      INTEGER BYEAR, OUTPRF
C
      ..COMMON /COST01/ ATPFEE
      COMMON /COST07/ CSTIM,GASCST, IMSTRT, ISCEN,MOVIM,MOVATP,REPIM(4)
      COMMON /COST12/ IMFXFG, IAFXFG, IMCSFG, IAFEFG, IALDV
      COMMON /COST22/ ATCOST(2,8)
      COMMON /COST27/ IGRFLG, IATPLP, IDIFF, OUTPRF
      COMMON /COST34/ TCOUNT,BYEAR,VGRORT
      COMMON /COST48/ PRGFEE,PRSFEE,CEVFEE(25,10,4),TRNFEE
      COMMON /YEARS4/ IY1941, IY1960, IY2020
C
C
      READ(7,*)
      READ (7, 700) IMFXFG,MOVIM, IMSTRT, IMCSFG, IDIFF, OUTPRF
      IDIFF - 1
      IF(OUTPRF.LT.1.0R.OUTPRF.GT.5)  OUTPRF - 3
      IF (IMCSFG.EQ. 2) THEN
        READ (7,701) CSTIM,REPIM(1) ,REPIM(2) ,GASCST,REPIM(3) ,PRSFEE,
     *              PRGFEE,TRNFEE,REPIM(4)
      ENDIF
C
      READ (7, 702) IAFXFG, MOVATP, IAFEFG
      IF (IAFEFG.EQ. 2) THEN
        READ<7,703) ATPFEE, ((ATCOST(I,J),1-1,2), J-l,8)
      ENDIF
C
      READ(7,704) IGRFLG
      IF (IGRFLG.EQ. 2) THEN
        READ (7, 705) VGRORT
        READ<7,707) TCOUNT,BYEAR
      ENDIF
C
  700 FORMAT(II,2(IX,12),3(IX,II))
  701 FORMAT(FS.2,2(1X,F6.2) , 1X,F4. 2,1X,F6.2, 3 (1X,F5.2) ,1X,F6.2)
  702 FORMAT(II,IX,12,IX,II)
  703 FORMAT(F5.2,IX,16(F4.0,1X))
  704 FORMAT(II)
  705 FORMAT(F4.1)
  706 FORMAT(8(F5.3,1X»
  707 FORMAT(F9.0,IX,14)
C
C  Call YRTEST and change 2  digit  years to 4 digit
C
      CALL YRTEST (IMSTRT, 32, IY1960, IY2020, INERR)
C
      RETURN
C
      END
                                                   96

-------
                 APPENDIX B
TECH4 MODEL VERSION 4.1 SOURCE CODE LISTING

-------
                                   B-l

                 Appendix B  : Tech 4.1 Model  Source Code
CC
CC..MOBILE4.1 I/M Credit Model for 1981 and newer LDGV
CC          With  Oxygenated Fuel Effects
CC                  (May 14,  1991)
CC                With Bag Fractions
CC                   (July 1991)
CC          With  Calculated Repair Effects
CC                  (July 20,  1991)
CC
CC..Program Main
CC
CC..COMMON Blocks and DIMENSION Statements
CC
      COMMON  /DAT01/  MYR,ISTD,ITECH,IBAG,IP,IAGE, ICUT, ITST
      COMMON  /DAT02/  AMIL(25) ,ODOM(25),TMILE(25),WGT(25)
      COMMON  /DAT03/  ENOX(4,2,2)
      COMMON  /DAT04/  FAIL(25,4,2)
      COMMON  /DAT05/  OXYE(2,3,4,3),OXYGON
      COMMON  /DAT06/  FRAC(4,12)
      COMMON  /DAT07/  ESO(2,4,4,2),EHO(2,4,4, 2)
      COMMON  /DAT08/  DN(3,4,4,2)
      COMMON  /DAT09/  ZMIL(2,4,4,2),CWO(2,4,25,4,2),CIMW(2, 4,25, 4, 2, 3)
      COMMON  /DAT10/  EWO(3,4,25,12),EIMW(2,4,25,12,3),EZM(2, 4,12)
      COMMON  /DAT11/  CREDIT(2,25,12,3, 4)
      COMMON  /DAT12/  ZML(3,4,12),ZML1(3,4,12),ZML2(3,4,12)
      COMMON  /DAT13/  BFZML1(3,4,12),BFDET1(3,4,12)
      COMMON  /DAT14/  DET(3,4,12),DET1(3,4,12),DET2(3, 4,12)
      COMMON  /DAT15/  CWOA(2,25,4,2),EWOA(3,25,12),EZMA(2,12)
      COMMON  /DAT16/  XSIDR(2,4,2,3),XHIDR(2, 4,2, 3)
      COMMON  /DAT17/  RSUP (2, 4, 2, 3) , RHIG (2, 4, 2, 3)
      COMMON  /DAT18/  SIZE(25,4,2,4)
      COMMON  /DAT19/  GV(4,2),GH(4,2),GS(4,2), BH(4, 2) ,BV(4, 2)
      COMMON  /DAT20/  EMO(2,4,4,2),ENO(3,4, 4, 2)
      COMMON  /DAT21/  BFDET2 (3,4,12),BFZML2(3,4,12)
      COMMON  /DAT22/  RMAR(2,4,2,3),RNOR(2, 4, 2, 3)
      COMMON  /DAT24/  XMIDR(2,4,2,3),XNIDR(2,4,2, 3)
      COMMON  /DAT25/  EXPO (2,4,3,2),DOTS(4,2,4, 2,2)
CC
C     OPEN(1,FILE-'M5SIZE')
C     OPEN(2,FILE-•M5TEST')
C     OPEN(3,FILE-'M5MYRf)
C     OPEN(4,FILE-'M5REG1)
      OPEN(13,FILE-'T5V7OXY.CSV')
      OPEN(14,FILE-'AAAAA')
CC
CC
CC..Calculate the mileage accumulated in each one year interval.
CC
      AMIL(l) - ODOM(l)
CC
      DO  10 IAGE-2,25
        AMIL(IAGE) -  ODOM(IAGE)  - ODOM(IAGE-l)
   10 CONTINUE
CC
CC..Model Year  Grouping
CC..ISTD  - 1  :  1981,1982 Model Year.s
CC..ISTD  - 2  :  1983 and newer Model Years
CC
      DO  600  ISTD=1,2
CC

-------
                                   B-2

                 Appendix B :  Tech 4.1 Model Source Code
CC..ITECH indicates the technology type used  in  the vehicles.
CC
CC..ITECH - 1 : Closed-Loop, Multi-Point Fuel Injected
CC..ITECH - 2 : Closed-Loop, Throttle-Body Fuel-Injected
CC..ITECH - 3 : Closed-Loop, Carbureted
CC..ITECH - 4 : Open-Loop, Any Carbureted or  Fuel  Injected
CC
      DO 600 ITECH-1,4
CC
CC..Vehicle age in years
CC
      DO 600 IAGE-1,25
CC
      CALL EGROUP
CC
CC..FTP Bag ( 1-FTP; 2-BAG1; 3-BAG2; 4-BAG3 )
CC
      DO 600 IBAG-1,4
CC
CC..Pollutant ( 1:HC, 2:CO )
CC
      DO 600 IP-1,2
CC
      CALL EMIT
      CALL IMEMIT
CC
  600 CONTINUE
CC
      CALL MYRSUB
      CALL REGR
      CALL JAN1
CC
      OPEN(7,FILE-'T5V7EF.DAT')
      OPEN(8,FILE-'T5V7ANN.IMC')
      OPEN(9,FILE-'T5V7BI.IMC')
      OPEN(10,FILE-'T5V7BF.DAT')
CC
      CALL OUTPUT
CC
C     CLOSE(1)
C     CLOSE(2)
C     CLOSE(3)
C     CLOSE(4)
      CLOSE(7)
      CLOSE(8)
      CLOSE(9)
C     CLOSE(10)
      CLOSE(13)
      CLOSE(14)
CC
      STOP
      END

-------
                                   B-3

                 Appendix B : Tech 4.1 Model  Source Code
      SUBROUTINE EGROUP
CC
CC..This routine predicts the number of vehicles in each emission
CC..level category  by technology and age.
CC
      COMMON /DAT01/  MYR,ISTD,ITECH,IBAG,IP,IAGE,ICUT, ITST
      COMMON /DAT02/  AMIL(25),ODOM(25),TMILE(25),WGT(25)
      COMMON /DAT04/  FAIL(25,4,2)
      COMMON /DAT18/  SIZE (25,4,2,4)
      COMMON /DAT19/  GV(4,2),GH(4,2),GS(4,2),BH(4, 2) ,BV(4, 2)
CC
CC..Calculate the number of "HIGH" emitting vehicles
CC
      SIZE(IAGE,ITECH,ISTD,2)  - GH(ITECH,ISTD)  * ODOM(IAGE)
CC
CC.."BH" is the change in the rate of occurrance of "HIGH"
CC..emitting vehicles assumed to occur at 50,000 miles.
CC
      IF(IAGE.GT.l  .AND. ODOM(IAGE-l).GT.5.0)
     * SIZE(IAGE,ITECH,ISTD,2)  - SIZE(IAGE-1,ITECH,ISTD,2)
     *                   + BH(ITECH,ISTD)*GH(ITECH,ISTD)*AMIL(IAGE)
      IF(SIZE(IAGE,ITECH,ISTD,2).GT.1.00) SIZE(IAGE,ITECH,ISTD,2) - 1.00
      IF(SIZE(IAGE,ITECH,ISTD,2).LT.0.00) SIZE(IAGE,ITECH,ISTD,2) - 0.00
CC
CC..Calculate the number of "VERY HIGH" emitting vehicles
CC
      SIZE(IAGE,ITECH,ISTD,3)  - GV(ITECH,ISTD)  * ODOM(IAGE)
CC
CC.."BV" is the change in the rate of occurrance of "VERY HIGH"
CC..emitting vehicles assumed to occur at 50,000 miles.
CC
      IF(IAGE.GT.l  .AND. ODOM(IAGE-l).GT.5.0)
     * SIZE(IAGE,ITECH,ISTD,3)  =• SIZE(IAGE-1,ITECH,ISTD,3)
     *                   + BV(ITECH,ISTD)*GV(ITECH, ISTD)*AMIL(IAGE)
      IF(SIZE(IAGE,ITECH,ISTD,3).GT.1.00) SIZE(IAGE,ITECH,ISTD,3) - 1.00
      IF(SIZE(IAGE,ITECH, ISTD, 3) .LT.O.O(X) SIZE (IAGE, ITECH, ISTD, 3) = 0.00
CC
CC..Calculates the  number of "SUPER" emitting vehicles
CC
      SIZE(IAGE,ITECH,ISTD,4)  - GS(ITECH,ISTD)  * ODOM(IAGE)
      IF(SIZE(IAGE,ITECH,ISTD,4).GT.1.00) SIZE(IAGE,ITECH,ISTD, 4) - 1.00
      IF(SIZE(IAGE,ITECH,ISTD,4).LT.0.00) SIZE(IAGE,ITECH,ISTD, 4) = 0.00
CC
CC..Check for logical sizes
CC
      CHECK - SIZE(IAGE,ITECH,ISTD,3)
     *      + SIZE(IAGE,ITECH,ISTD,4)
      IF(CHECK.GT.l.O)
     *    SIZE(IAGE,ITECH,ISTD,3)  - 1.0 - SIZE(IAGE,ITECH,ISTD,4)
CC
      CHECK - SIZE(IAGE,ITECH,ISTD,2)
     *      + SIZE(IAGE,ITECH,ISTD,3)
     *      + SIZE(IAGE,ITECH,ISTD,4)
      IF(CHECK.GT.1.0)
     *    SIZE(IAGE,ITECH,ISTD,2)  - 1.0 - SIZE(IAGE,ITECH,ISTD, 3)
     *                                   - SIZE(IAGE,ITECH,ISTD,4)
CC
CC..Calculate the number of "NORMAL" emitting vehicles
CC
      SIZE(IAGE,ITECH,ISTD,1)  -1.0

-------
                                   B-4

                 Appendix B :  Tech 4.1 Model Source Code


     *                         -  SIZE(IAGB,ITECH,ISTD,2)
     *                         -  SIZE(IAGE,ITECH,ISTD,3)
     *                         -  SIZE(IAGE,ITECH,ISTD,4)
C     WRITE(1,701) IAGE,ITECH,ISTD,(SIZE(IAGE,ITECH,ISTD,J),J-1,4)
C 701 FORMAT(313,4F8.4)
CC
  999 RETURN
      END

-------
                                   B-5

                 Appendix B : Tech 4.1 Model  Source Code


      SUBROUTINE EMIT
CC
CC..This routine combines  the emission levels of each emission
CC..category based on  the  predicted category size.
CC
      COMMON /DAT01/ MYR, ISTD, ITECH,IBAG,IP,IAGE,ICUT,ITST
      COMMON /DAT02/ AMIL(25),ODOM(25),TMILE(25) ,WGT(25)
      COMMON /DAT07/ ESO(2,4,4,2),EHO(2, 4, 4, 2)
      COMMON /DAT08/ DN<3,4,4,2)
      COMMON /DAT09/ ZMIL(2,4,4,2),CWO<2,4,25,4, 2) ,CIMW(2, 4,25, 4,2, 3)
      COMMON /DAT15/ CWOA(2,25,4,2),EWOA(3,25,12),EZMA(2,12)
      COMMON /DAT18/ SIZE(25,4,2,4)
      COMMON /DAT19/ GV(4,2),GH(4,2),GS(4,2),BH(4,2) ,BV(4, 2)
      COMMON /DAT20/ EMO (2,4,4,2),ENO(3,4, 4,2)
CC
      IF(IAGE.GT.l) GOTO 10
CC
CC..Emission levels at zero  mileage point
CC
      ZMIL(IP, IBAG,ITECH,ISTD)  - END(IP,IBAG,ITECH,ISTD)
CC
CC..Emission levels by age
CC
   10 ES - ESO(IP,IBAG,ITECH,ISTD)
      EH - EHO(IP,IBAG,ITECH,ISTD)
     *   +  ( DN(IP,IBAG,ITECH,ISTD)*ODOM(IAGE)  )
      EM - EMO(IP,IBAG,ITECH,ISTD)
     *   +  ( DN(IP,IBAG,ITECH,ISTD)*ODOM(IAGE)  )
      EN - ENO(IP,IBAG,ITECH,ISTD)
     *   +  ( DN(IP,IBAG,ITECH,ISTD)*ODOM(IAGE)  )
CC
CC  Adjust emissions for effects of Oxygenated Fuels
CC
      ENA=EN*OXY(IP,ITECH,EN)
      EMA-EM*OXY(IP,ITECH,EM)
      EHA-EH*OXY(IP,ITECH, EH)
      ESA-ES*OXY(IP,ITECH, ES)
CC
CC..Calculate  the  base (without I/M)  composite emission levels by age
CC
      CWO(IP,IBAG,IAGE,ITECH,ISTD)  -
     *   SIZE(IAGE,ITECH,ISTD,1)  *  EN
     * + SIZE(IAGE,ITECH,ISTD,2)  *  EM
     * + SIZE(IAGE,ITECH,ISTD,3)  *  EH
     * + SIZE(IAGE,ITECH,ISTD,4)  *  ES
      IF(IBAG.EQ.l)
     * CWOA(IP,IAGE,ITECH,ISTD)  -
     *   SIZE(IAGE,ITECH,ISTD,1)  *  ENA
     * + SIZE(IAGE,ITECH,ISTD,2)  *  EMA
     * + SIZE(IAGE,ITECH,ISTD,3)  *  EHA
     * + SIZE(IAGE,ITECH,ISTD,4)  *  ESA
CC
C     IF(IP.EQ.LAND.IBAG.EQ.l)
C    * WRITE(2,772) IAGE,ITECH,ISTD,EN,EM, EH,ES
C 772 FORMAT(314,4F7.3)
CC
  999 RETURN
      END

-------
                                   B-6

                 Appendix B :  Tech 4.1 Model Source  Code
      SUBROUTINE IMEMIT
CC
CC..This routine combines  the emission levels of each emission
CC..category based on the  predicted catagory size.
CC
      COMMON /DAT01/ MYR,ISTD,ITECH,IBAG,IP,IAGE,ICUT,ITST
      COMMON /DAT02/ AMIL(25),ODOM(25),TMILE(25),WGT(25)
      COMMON /DAT07/ ESO(2,4,4,2),EHO(2,4,4,2)
      COMMON /DAT08/ DN(3,4,4,2)
      COMMON /DAT09/ ZMIL(2,4,4,2),CWO(2,4,25, 4, 2) ,CIMW(2, 4, 25,4,2, 3)
      COMMON /DAT16/ XSIDR(2,4,2,3),XHIDR(2,4,2,3)
      COMMON /DAT17/ RSUP(2, 4, 2, 3), RHIG(2, 4, 2, 3)
      COMMON /DAT18/ SIZE(25,4,2,4)
      COMMON /DAT20/ EMO(2,4,4,2),END(3,4,4,2)
      COMMON /DAT22/ RMAR(2, 4, 2, 3) ,RNOR(2, 4, 2, 3)
      COMMON /DAT24/ XMIDR(2,4,2,3),XNIDR(2,4,2,3)
      COMMON /DAT25/ EXPO(2,4,3,2),DOTS(4,2,4,2,2)
CC
      DIMENSION STD(2)
      DATA STD / .41, 3.4  /
CC
CC..Non-I/M emission levels
CC
      ES2 - ESO(IP,IBAG,ITECH,ISTD)
      EH2 - EHO(IP,IBAG,ITECH,ISTD)
     *   +  ( DN(IP, IBAG, ITECH, ISTD) *ODOM(IAGE)  )
      EM2 - EMO(IP,IBAG,ITECH,ISTD)
     *   +  ( DN(IP,IBAG,ITECH,ISTD)*ODOM(IAGE)  )
      EN2 - ENO(IP,IBAG,ITECH,ISTD)
     *   +  ( DN(IP,IBAG,ITECH,ISTD)*ODOM(IAGE)  )
CC
CC..For each test type   ITEST - 1 :  Idle Test
CC..                     ITEST - 2 :  IM240  with 1.6/30 HC/CO Cutpoints
CC..                     ITEST - 3 :  IM240  with 0.8/15 HC/CO Cutpoints
CC
      DO 10 ITEST-1,3
CC
CC..The emissions of vehicles passing the short test are combined
CC..with the estimated emission levels of vehicles which are repaired.
CC
CC  DOTS(ITECH,A/B,IDOT,IP,ITST)
CC
      ITST-ITEST
      IF(ITEST.EQ.3) ITST-2
CC
C     IF(ES2.GT.O) REPS -  EXPO(IP,ITECH,ITEST,1)
C    *     * EXP(-EXPO(IP,ITECH,ITEST,2)/ES2)
      CALL CONECT(ES2,REPS)
      IF(REPS.LT.STD(IP)  .AND.  ES2.GT.STD(IP))  REPS-STD(IP)
      IF(REPS.GT.ES2) REPS-ES2
      IF(ES2.LT.STD(IP)) REPS-ES2
      EIMS =• (XSIDR(IP, ITECH, ISTD, ITEST) *REPS)   +
     *             ((1 - XSIDR(IP,ITECH,ISTD,ITEST))*ES2)
CC
C     REPH - EXPO(IP,ITECH,ITEST,1)
C    *     * EXP(-EXPO(IP,ITECH,ITEST, 2)/EH2)
      CALL CONECT(EH2,REPH)
      IF(REPH.LT.STD(IP)  .AND.  EH2.GT.STD(IP))  REPH-STD(IP)
      IF(REPH.GT.EH2) REPH-EH2
      IF(EH2.LT.STD(IP)) REPH-EH2

-------
                                   B-7

                 Appendix B : Tech 4.1 Model  Source Code
      EIMH -  (XHIDR(IP,ITECH,ISTD,ITEST)*REPH)  +
     *             ((1  - XHIDR(IP,ITECH,ISTD,ITEST))*EH2)
CC
C     REPM - EXPO(IP,ITECH,ITEST,1)
C    *     * EXP(-EXPO(IP,ITECH,ITEST,2)/EM2)
      CALL CONECT(EM2,REPM)
      IF(REPM.LT.STD(IP)  .AND.  EM2.GT.STD(IP)) REPM-STD(IP)
      IF(REPM.GT.EM2)  REPM=EM2
      IF(EM2.LT.STD(IP))  REPM-EM2
      EIMM =  (XMIDR(IP,ITECH,ISTD,ITEST)*REPM)  +
     *             (d  - XMIDR(IP, ITECH,ISTD, ITEST) )*EM2)
CC
C     REPN - EXPO(IP,ITECH,ITEST,1)
C    *     * EXP(-EXPO(IP,ITECH,ITEST,2)/EN2)
      CALL CONECT(EN2,REPN)
      IF(REPN.LT.STD(IP)  .AND.  EN2.GT.STD(IP)) REPN-STD(IP)
      IF(REPN.GT.EN2)  REPN-EN2
      IF(EN2.LT.STD(IP))  REPN-EN2
      EIMN -  (XNIDR(IP,ITECH,ISTD,ITEST)*REPN)  +
     *             ((1  - XNIDR(IP,ITECH,ISTD,ITEST))*EN2)
CC
      PS=-XSIDR(IP, ITECH, ISTD, ITEST)
      PH-XHIDR(IP,ITECH,ISTD,ITEST)
      PM-XMIDR(IP,ITECH,ISTD,ITEST)
      PN-XNIDR(IP,ITECH,ISTD,ITEST)
      IF(IBAG.EQ.l  .AND.  IAGE.EQ.1  .AND. IP.EQ.l)
     * WRITE(14,881) ISTD,ITECH,IP,ITEST,IAGE,REPS,REPH,REPM,REPN,
     * ES2,EH2,EM2,EN2,EIMS,EIMH,EIMM,EIMN,PS,PH,PM,PN
  881 FORMAT(5I3,12F7.2,4F7.3)
CC
CC..Emission levels by age and by test
CC..Calculate the  base (with I/M)  composite emission levels by age
CC
      CIMW(IP,IBAG,IAGE,ITECH,ISTD,ITEST) -
     *   SIZE(IAGE,ITECH,ISTD,1)  *  EIMN
     * + SIZE(IAGE,ITECH,ISTD,2)  *  EIMM
     * + SIZE(IAGE,ITECH,ISTD,3)  *  EIMH
     * + SIZE(IAGE,ITECH,ISTD,4)  *  EIMS
CC
C     IF(ITEST.EQ.3) THEN
C       CHK1-CIMW(IP,IBAG,IAGE,ITECH,ISTD,ITEST)
C       CHK2-CIMW(IP,IBAG,IAGE,ITECH,ISTD,ITEST-1)
C       IF(CHK1.GT.CHK2)  THEN
C         CIMW(IP,IBAG,IAGE,ITECH,ISTD,ITEST)-CHK2
C     WRITE (14, 882) ISTD, ITECH, IP, ITEST, IAGE, REPS, REPH, REPM, REPN,
C    *           ES2,EH2,EM2,EN2,EIMS,EIMH,EIMM,EIMN,PS,PH,PM,PN
C 882 FORMAT(5I3,12F7.3,4F7.3)
C       ENDIF
C     ENDIF
CC
CC
    10 CONTINUE
CC
  999 RETURN
      END

-------
                                   B-8

                 Appendix B : Tech 4.1 Model Source Code
      SUBROUTINE MYRSUB
CC
CC..This section combines  the technologies into
CC..model year emission  levels.
CC
      COMMON /DAT01/ MYR,ISTD,ITECH,IBAG,IP,IAGE,ICUT,ITST
      COMMON /DAT02/ AMIL(25),ODOM(25),TMILE(25),WGT(25)
      COMMON /DAT06/ FRAC(4,12)
      COMMON /DAT08/ DN(3,4,4,2)
      COMMON /DAT09/ ZMIL(2,4,4,2),CWO(2,4,25,4,2),CIMW<2,4,25,4,2,3)
      COMMON /DAT10/ EWO(3,4,25,12),EIMW<2,4,25,12,3),EZM<2,4,12)
      COMMON /DAT12/ ZML (3, 4,12), ZML1 (3, 4,12) , ZML2 (3, 4,12)
      COMMON /DAT15/ CWOA(2,25,4, 2),EWOA(3,25,12),EZMA(2,12)
      COMMON /DAT14/ DET(3,4,12),DET1(3,4,12),DET2(3,4,12)
      COMMON /DAT20/ EMO(2,4,4,2),END(3,4,4,2)
CC
      DIMENSION EFFECT(12,25)
CC
      ZERO=0.0
CC
CC..Loop by MYR, CO standard,  technology,  age, bag, &  pollutant
CC
CC..The ITEST loops only for the I/M composite emission arrays
CC
CC
      DO 300 MYR-1,12
       ISTD-1
       IF(MYR.GE.3) ISTD-2
      DO 300 IP-1,2
      DO 300 IBAG-1,4
      DO 300 ITECH-1,4
CC
CC..Zero mile emission levels by model  year
CC
      EZM(IP,IBAG,MYR) - EZM(IP,IBAG,MYR)
     * + FRAC(ITECH,MYR) * ZMIL(IP,IBAG,ITECH,ISTD)
CC
CC  Oxygenated Fuel Effect
CC
      IF(IBAG.EQ.l) EZMA(IP,MYR)-EZMA(IP,MYR)
     * + FRAG(ITECH,MYR)*ZMIL(IP,IBAG,ITECH,ISTD)
     * * OXY(IP, ITECH, ZMIL(IP,IBAG,ITECH,ISTD))
CC
      DO 300 IAGE-1,25
CC
CC..Calculates the emission levels  for
CC..January 1st dates from the emission levels by age.
CC..Since model year introduction is  on October 1st,  this
CC..requires a 75%/25% staggering.
CC
      EWO (IP, IBAG, IAGE,MYR)  »
     *   EWO(IP,IBAG,IAGE,MYR)
     * + FRAC(ITECH,MYR) * CWO(IP,IBAG,IAGE,ITECH,ISTD)
CC
CC  Oxygenated Fuel Effect
CC
      IF(IBAG.EQ.l)
     * EWOA(IP,IAGE,MYR) =• EWOA(IP, IAGE,MYR)
     * + FRAC(ITECH,MYR) * CWOA(IP,IAGE,ITECH,ISTD)
CC

-------
                                   B-9

                 Appendix B  : Tech 4.1 Model  Source Code
      DO 200 ITEST-1,3
CC
      EIMW(IP,IBAG,IAGE,MYR, ITEST)  =
     *    EIMW(IP,IBAG,IAGE,MYR,ITEST)
     *  + FRAC(ITECH,MYR)  * CIMW(IP,IBAG,IAGE,ITECH,ISTD, ITEST)
CC
CC
  200 CONTINUE
  300 CONTINUE
CC
CC  Write out the  Oxygenated Fuels  Effect
CC
      DO 774 MYR-1,12
       ISTD=1
       IF(MYR.GE.3)  ISTD-2
       IAGE-0
CC
C      WRITE(13,773) MYR, IAGE,ZERO,
C    *  E2M(1,1,MYR),EZMA(1,MYR),
C    *  EZM(2,1,MYR),EZMA(2,MYR)
CCCCCC      *  EZM(3,1,MYR),EZMA(3,MYR)
CC
      DO 774 IAGE-1,25
      DO 772 ITECH-1,4
       EMIT-ENO(3,1,ITECH,ISTD)+DN(3,1,ITECH,ISTD)*ODOM(IAGE)
       EWO(3,1,IAGE,MYR)-EWO(3,1,IAGE,MYR)
     *  + EMIT*FRAC(ITECH,MYR)
       EWOA(3,IAGE,MYR)-EWOA(3,IAGE,MYR)
     *  + EMIT*FRAC(ITECH,MYR)*OXY(3,ITECH, EMIT)
  772 CONTINUE
CC
       EFFECT(MYR,IAGE)  -
     *  (EWO(2,1,IAGE,MYR)-EWOA(2,IAGE,MYR))
     * / EWO(2,1,IAGE,MYR)
CC
C      WRITE(13,773) MYR,IAGE,ODOM(IAGE) ,
C    *  EWO (1, 1, IAGE,MYR) , EWOA (1, IAGE,MYR) ,
C    *  EWO (2,1, IAGE,MYR) , EWOA (2, IAGE,MYR) ,
C    *  EWO (3,1, IAGE,MYR) , EWOA (3, IAGE, MYR)
C 773 FORMAT(2(I4,','),F9.4,',',6(F8.3, ', '))
CC
  774 CONTINUE
CC
      DO 10 MYR-1,12
       WRITE(13,775)  (EWO(2,1,IAGE,MYR),IAGE-1, 25)
   10 CONTINUE
      DO 20 MYR-1,12
       WRITE(13,776)  (EFFECT(MYR,IAGE),IAGE=1,25)
   20 CONTINUE
  775 FORMAT(5X,f*',25(F7.2, ', '))
  776 FORMAT(5X,'*',25(F4.3, ', '))
CC
  999 RETURN
      END

-------
                                   B-10

                 Appendix B :  Tech 4.1 Model Source Code
      SUBROUTINE REGR
CC
CC..This subroutine uses a weighted regression equation to
CC..linearize the emission level  results  for each model year.
CC
      COMMON /DAT01/ MYR,ISTD,ITECH,IBAG,IP,IAGE,ICUT,ITST
      COMMON /DAT02/ AMIL(25),ODOM(25),TMILE(25),WGT(25)
      COMMON /DAT03/ ENOX(4,2,2)
      COMMON /DAT06/ FRAC(4,12)
      COMMON /DAT07/ ESO(2,4,4,2),EHO(2,4,4,2)
      COMMON /DAT08/ DN(3,4,4,2)
      COMMON /DAT10/ EWO(3f4,25,12),EIMW(2,4,25,12,3),EZM(2,4,12)
      COMMON /DAT12/ ZML (3, 4,12) , ZML1 (3, 4,12) , ZML2 (3, 4,12)
      COMMON /DAT14/ DET(3,4,12),DET1(3,4,12),DET2(3,4,12)
      COMMON /DAT20/ EMO(2,4,4,2),ENO(3,4,4,2)
CC
      DO 40 MYR-1,12
      DO 40 IBAG=1,4
      DO 40 IP-1,2
CC
      SUMX  - 0.0
      SUMY  - 0.0
      SUMXY - 0-0
      SUMXX - 0.0
CC
      N - 5
CC
      DO 10 IAGE-1,N
CC
       IF(IAGE.EQ.l) EM - EZM(IP, IBAG,MYR)
       IF(IAGE.GT.l) EM - EWO(IP,IBAG,IAGE-1,MYR)
CC
       IF(IAGE.EQ.l) XM - 0.0
       IF(IAGE.GT.l) XM - ODOM(IAGE-l)
CC
       SUMX  - SUMX  +  XM
       SUMY  - SUMY  +  EM
       SUMXY - SUMXY +   (XM*EM)
       SUMXX - SUMXX +   (XM**2)
CC
C     IF(IP.EQ.LAND.IBAG.EQ.l)
C    * WRITE(4,774) IAGE,MYR,EM,XM,SUMX,SUMY,SUMXY,SUMXX
C 774 FORMAT(214,6F8.4)
CC
   10 CONTINUE
CC
      SUM1 - N * SUMXY - SUMX * SUMY
      SUM2 - N * SUMXX - SUMX**2
      Dl -  SUM1 / SUM2
      Zl -  (SUMY/N) - Dl *  (SUMX/N)
CC
CC..Store the regression results
CC
      ZML1(IP,IBAG,MYR) - Zl
      DET1(IP,IBAG,MYR) - Dl
      ZML2(IP,IBAG,MYR) - ZML1(IP,IBAG,MYR)  +  DET1(IP,IBAG,MYR)*5.0
CC
      IF(ZML1(IP,IBAG,MYR).GE.0.0)  GO  TO  30
CC
CC..If the emission level at  zero miles is less than zero,

-------
                                   B-ll

                 Appendix B :  Tech 4.1 Model Source Code
CC..then the regression  is  altered to intercept at zero
CC
      ZML1(IP,IBAG,MYR)  - 0.0
      DET1(IP,IBAG,MYR)  - SUMXY /  SUMXX
      ZML2(IP,IBAG,MYR)  - ZML1 (IP, IBAG,MYR)  + DET1 (IP, IBAG,MYR) *5. 0
CC
   30 SUMX  =0.0
      SUMY  =0.0
      SUMXY =0-0
      SUMXX =0.0
CC
      Ml = 25 - N
      M2 = N + 1
CC
      DO 20 IAGE=M2,25
CC
       IF(IAGE.EQ.l)  EM  - EZM(IP,IBAG,MYR)
       IF(IAGE.GT.l)  EM  = EWO(IP,IBAG,IAGE-1,MYR)
CC
       IF(IAGE.EQ.l)  XM  - 0.0
       IF(IAGE.GT.l)  XM  = ODOM(IAGE-l)
CC
       SUMX  = SUMX   +   XM
       SUMY  = SUMY   +   EM
       SUMXY = SUMXY  +   (XM*EM)
       SUMXX = SUMXX  +   (XM**2)
CC
  20  CONTINUE
CC
      SUM1 - Ml *  SUMXY  - SUMX  * SUMY
      SUM2 - Ml *  SUMXX  - SUMX**2
      Dl =  SUM1 / SUM2
      Zl -   (SUMY/MI) -  Dl  * (SUMX/M1)
CC
CC..Store the regression results
CC
      DET2(IP,IBAG,MYR)  - Dl
CC
CC..Single Linear  Regression
CC
      SUMX  =0.0
      SUMY  =0.0
      SUMXY =0.0
      SUMXX =0.0
C     SUMW  =0.0
CC
      DO 60 IAGE-1,25
CC
       IF(IAGE.EQ.l)  EM  = EZM(IP,IBAG,MYR)
       IF(IAGE.GT.l)  EM  = EWO(IP,IBAG,IAGE-1,MYR)
CC
       IF(IAGE.EQ.l)  XM  = 0.0
       IF(IAGE.GT.l)  XM  = ODOM(IAGE-l)
CC
       SUMX  = SUMX   +  ( WGT(IAGE)  * XM )
       SUMY  = SUMY   +  ( WGT(IAGE)  ,* EM )
       SUMXY = SUMXY  +  ( WGT(IAGE)  * (XM*EM))
       SUMXX = SUMXX  +  ( WGT(IAGE)  * (XM**2))
CC
  60  CONTINUE

-------
                                   B-12

                 Appendix B :  Tech 4.1 Model Source Code
CC
      SUM1 - SUMXY - SUMX * SUMY
      SUM2 - SUMXX - SUMX**2
      Dl -  SUM1 / SUM2
      Zl -  SUMY - Dl * SUMX
CC
CC..Store the regression results
CC
      ZML (IP, IBAG,MYR) = 211
      DET(IP,IBAG,MYR) - Dl
CC
      IF(ZML(IP,IBAG,MYR).GE.0.0) GO TO 40
CC
CC..If the emission level at  zero miles is  less  than zero,
CC..then the regression is altered to  intercept  at zero.
CC
      ZML(IP,IBAG,MYR) - 0.0
      DET (IP, IBAG,MYR) - SUMXY  / SUMXX
CC
   40 CONTINUE
CC
CC..Since the NOx emissions are not combined from emission level
CC..groups, the NOx emission  factors can be calculated directly
CC..from the regressions.
CC
      IP-3
CC
      DO 50 MYR-1,12
       ISTD-1
       IF(MYR.GE.3) ISTD-2
      DO 50 IBAG-1,4
      DO 50 ITECH-1,4
CC
      ZML(IP,IBAG,MYR) - ZML (IP, IBAG,MYR) +
     *  ENO(IP,IBAG,ITECH,ISTD)*FRAC(ITECH,MYR)
CC
      DET(IP,IBAG,MYR) - DET(IP,IBAG,MYR) +
     *  DN(IP,IBAG,ITECH,ISTD)*FRAC(ITECH,MYR)
CC
      ZML1
-------
                                  B-13

                Appendix B  : Tech 4.1 Model  Source Code
        DET2(IP,1,MYR) =DET1(IP,1,MYR)
   56 CONTINUE
CC
CC
      CALL BAGF
CC
  999 RETURN
      END

-------
                                   B-14

                 Appendix B :  Tech 4.1 Model Source Code
      SUBROUTINE BAGF
CC
CC..This routine calculates the  bag fractions for hot/cold starts
CC
CC..Last Updated : November 15,  1988
CC
      COMMON /DAT01/ MYR,ISTD,ITECH,IBAG,IP,IAGE,ICUT,ITST
      COMMON /DAT12/ ZML (3, 4, 12), ZML1 (3, 4,12), ZML2 (3, 4, 12)
      COMMON /DAT13/ BFZML1(3,4,12),BFDET1(3,4,12)
      COMMON /DAT14/ DET(3,4,12),DET1(3,4,12),DET2(3,4,12)
      COMMON /DAT21/ BFDET2(3,4,12),BFZML2(3,4,12)
CC
      DIMENSION BFRAC(4)
CC
      DATA BFRAC  / 1.000, 0.206,  0-521,  0.273  /
CC
      DO 20 IP-1,3
      DO 20 MYR-1,12
CC
       Z2 - 0.0
       Z3 = 0.0
       D2 - 0.0
       D3 - 0.0
CC
CC..Sum up the bag regression coeffs weighted by  the FTP bag fractions
CC
      DO 10 IBAG-2,4
CC
       Z2 - Z2 + ZML1(IP,IBAG,MYR)   *  BFRAC(IBAG)
       Z3 » Z3 + ZML2(IP,IBAG,MYR)  * BFRAC(IBAG)
       D2 =- D2 + DET1(IP,IBAG,MYR)   *  BFRAC (IBAG)
       D3 - D3 + DET2(IP,IBAG,MYR)  * BFRAC(IBAG)
CC
   10 CONTINUE
CC
CC..Set the combined FTP bag  fraction  to 1.00
CC
      BFZML1(IP,1,MYR)  - Z2  / Z2
      BFZML2(IP,1,MYR) - Z3 / Z2
      BFDET1(IP,1,MYR)  - D2  / Z2
      BFDET2(IP,1,MYR) - D3 / Z2
CC
CC..Divide each bag regression coeff by  the  weighted sum
CC
      DO 20 IBAG-2,4
CC
       BFZML1(IP,IBAG,MYR) -  ZML1(IP,IBAG,MYR)  /  Z2
       BFZML2(IP,IBAG,MYR) -  ZML2(IP,IBAG,MYR)  /  Z2
       BFDET1(IP,IBAG,MYR) -  DET1(IP,IBAG,MYR)  /  Z2
       BFDET2(IP,IBAG,MYR) -  DET2(IP,IBAG,MYR)  /  Z2
CC
   20 CONTINUE
CC
      RETURN
      END

-------
                                   B-15

                 Appendix B : Tech 4.1 Model Source  Code
      SUBROUTINE JAN1
CC
CC..This subroutine  calculates the average emissions of each
CC..model year on  January first.   It creates the I/M credits
CC..and passes them  to  OUTPUT.
CC
      COMMON /DAT02/ AMIL(25),ODOM(25),TMILE(25) ,WGT(25)
      COMMON /DAT10/ EWO(3,4,25,12),EIMW(2,4,25,12,3) ,EZM<2, 4,12)
      COMMON /DAT11/ CREDIT(2,25,12,3, 4)
      COMMON /DAT12/ ZML(3, 4, 12),ZML1(3,4,12),ZML2(3,4,12)
      COMMON /DAT14/ DET(3,4,12),DET1(3,4,12),DET2(3,4,12)
CC
      DIMENSION ANSWNO(2,25,12),ANSWIM(2,25,12, 3, 3)
      DIMENSION EPRED(2,4,25,12,3),PRED(2,4,25, 12, 3)
      DIMENSION SLOPE(2,12,26),ZERO(2,12, 26)
CC
CC
      IBAG = 1
      DO 100 MYR   -  1,12
      DO 100 IP    -  1,2
CC
CC..The deteriorations  before and after the "KINK" are transferred
CC..to the array SLOPE  for each vehicle age.
CC
      DO 95 I =- 1,25
CC
       IF(I.LE.4)  SLOPE(IP,MYR,I)  - DET1(IP,IBAG,MYR)
       IF(I.GE.5)  SLOPE(IP,MYR,I)  - DET2(IP,IBAG,MYR)
CC
       IF(I.LE.4)  ZERO(IP,MYR, I)  - ZML1(IP,IBAG,MYR)
       IF(I.GE.S)  ZERO(IP,MYR, I)  - ZML2(IP,IBAG,MYR)  -
     *                               DET2(IP,IBAG,MYR)*5.0
CC
    95 CONTINUE
CC
CC..Computes the NON-I/M emission level by age,  by pollutant,
CC..by bag, and by myr.
CC
      DO 100 IAGE  -  1,24
CC
      IF(IAGE  .GT. 1) GOTO 33
CC
CC..Vehicle age is one.
CC
      ANSWNO(IP,IAGE,MYR)  -
     *.75*(ZERO(IP,MYR,IAGE)  + SLOPE(IP,MYR,IAGE)*.625*ODOM(IAGE) )  4
     *.25* (ZERO(IP,MYR,IAGE+1) +  SLOPE(IP,MYR,IAGE+1)*
     *         (.125*(ODOM(IAGE+1)-ODOM(IAGE))+ODOM(IAGE)) )
      GOTO 34
CC
CC..Vehicle age is greater than one.
CC
    33 ANSWNO(IP,IAGE,MYR)  -
     *.75*(ZERO(IP,MYR,IAGE)  + SLOPE(IP,MYR,IAGE)  *
     *         (,625*(ODOM(IAGE)-ODOM(IAGE-1))+ODOM(IAGE-1))  )   +
     *.25*(ZERO(IP,MYR,IAGE+1) +  SLOPE(IP,MYR,IAGE+1)  *
     *         (.125*(ODOM(IAGE+1)-ODOM(IAGE))+ODOM(IAGE)) )
CC
CC..Compute the I/M  emission level by age, by pollutant,
CC..by bag, by myr and  by test for IAGE -  1.  The predicted emission

-------
                                   B-16

                 Appendix B :  Tech 4.1 Model Source Code


CO..level is from the regression equation and the actual model
CC..emission level points.
CC
    34 DO 100 ITEST-1,3
CC
      EPRED(IP,IBAG,IAGE,MYRrITEST)  -
     * 1 -   ((EWO(IP,IBAG,IAGE,MYR)  - EIMW(IP,IBAG,IAGE,MYR,ITEST)) /
     *                   EWO(IP ,IBAG,IAGE,MYR))
CC
      PRED(IP,IBAG,IAGE,MYR, ITEST) -
     *     (ZERO (IP,MYR,IAGE)  + SLOPE(IP,MYR,IAGE)*ODOM(IAGE)) *
     *     EPRED(IP,IBAG,IAGE,MYR,ITEST)
CC
CC..Determine I/M credits  for each inspection frequency.
CC
CC      ITYP => 1   Annual
CC             2   Biennial 1 -  3 -  5 - etc
CC             3   Biennial 2 -  4 -  6 - etc
CC
      DO 110 ITYP -1,3
CC
       IF(ITYP.GE.2) GOTO  60
CC
CC..Annual I/M Credits
CC
      IF(IAGE .GT. 1) GOTO 50
CC
      ANSWIM(IP,IAGE,MYR,ITEST,ITYP)  -
     *  (.75*(.625*SLOPE(IP,MYR,IAGE)*ODOM(IAGE)  + ZERO(IP,MYR,IAGE)))
     * + .25*(PRED(IP,IBAG,IAGE,MYR,ITEST) + SLOPE(IP,MYR,IAGE+1)*
     *        (.125*(ODOM(IAGE+1)-ODOM(IAGE)))  )
      GOTO 60
CC
    50 ANSWIM(IP,IAGE,MYR,ITEST,ITYP)  -
     *.75*(PRED(IP,IBAG,IAGE-1,MYR,ITEST)  +
     *  SLOPE(IP,MYR,IAGE) *  (.625*(ODOM(IAGE)-ODOM(IAGE-1))) )  +
     *.25*(PRED(IP,IBAG,IAGE,MYR,ITEST) +
     *  SLOPE(IP, MYR,IAGE+1)  *  (.125*(ODOM(IAGE+1)-ODOM(IAGE)))  )
CC
      GOTO 90
CC
CC..Biennial I/M Credits
CC
CC    IMODE  -  1  :  Odd year
CC             2  :  Even year
CC
    60 IMODE  - MOD(IAGE,2)
CC
CC..Biennial 1 - 3 - 5  - etc   1st Year Exception
CC. .          same as no I/M  first year
CC
      IF(ITYP .EQ. 2 .AND. IAGE  .EQ.  1)
     *   ANSWIM(IP,IAGE,MYR,ITEST,ITYP) =
     *  (.75*(.625*SLOPE(IP,MYR,IAGE)*ODOM(IAGE)  + ZERO(IP, MYR, IAGE)))
     * + .25*(PRED(IP,IBAG,IAGE,MYR,ITEST)  + SLOPE(IP,MYR,IAGE+1)*
     *        (.125*(ODOM(IAGE+1)-OPOM(IAGE)))  )
CC
CC..Biennial 2 - 4 - 6  - etc   1st Year Exception
CC. .          same as no I/M  first year
CC

-------
                                   B-17

                 Appendix B : Tech 4.1 Model  Source Code


      IF(ITYP .EQ. 3  .AND.  IAGE .EQ.  1)
     *  ANSWIM (IP,IAGE,MYR,ITEST,ITYP)  -
     *   .75*(ZERO(IP,MIR,IAGE)  + SLOPE(IP, MYR,IAGE)*.625*ODOM(IAGE)) +
     *   .25*(ZERO(IP,MYR,IAGE)  + SLOPE(IP,MYR,IAGE+1)*
     *  (.125*(ODOM(IAGE+1)-ODOM(IAGE))+ODOM(IAGE)) )
CC
CC..Biennial  2 -  4 -  6 -  etc    2nd Year Exception
CC
      IF(ITYP .EQ. 3  .AND.  IAGE .EQ.  2)
     *         ANSWIM(IP,IAGE,MYR,ITEST,ITYP) -
     *     .75*(ZERO(IP,MYR,IAGE)  +
     *         SLOPE(IP,MYR,IAGE)*(ODOM(IAGE-l))  +
     *         SLOPE (IP, MYR, IAGE)* (.625* (ODOM(IAGE)-ODOM(IAGE-l) ) ) ) +
     *     .25*(PRED(IP,IBAG,IAGE,MYR,ITEST) +
     *         SLOPE (IP,MYR, IAGE+1)*. 125* (ODOM(IAGE+1) -ODOM(IAGE) ) )
CC
      IFdAGE.EQ.l  .OR.  (ITYP.EQ.3 .AND.  IAGE.EQ.2))  GOTO 90
CC
CC..The Principle Biennial Cases 1-3-5-etc and 2-4-6-etc
CC
CC..An Even Year  for  the  1-3-5 or An Odd Year for the 2-4-6
CC          There is  no I/M inspection that year
CC
      IF((IMODE.EQ.l  .AND.  ITYP.EQ.3)  .OR.  (IMODE.EQ.O.AND.ITYP.EQ.2))
     *        ANSWIM(IP, IAGE,MYR, ITEST, ITYP)  -
     *   .75*(PRED(IP,IBAG,IAGE-1,MYR,ITEST)  +
     *        SLOPE(IP,MYR,IAGE)*(.625*(ODOM(IAGE)-ODOM(IAGE-1)))) +
     *   .25*(PRED(IP,IBAG,IAGE-1,MYR,ITEST)  +
     *        SLOPE(IP,MYR,IAGE)*(ODOM(IAGE)-ODOM(IAGE-l))  +
     *        SLOPE(IP,MYR,IAGE+1)*.125*(ODOM(IAGE+1)-ODOM(IAGE)))
CC
CC..An Odd Year  for the 1-3-5  or An Even Year for the 2-4-6
CC..          There is an I/M  inspection that year
CC
      IF((IMODE.EQ.O  .AND.  ITYP.EQ.3)  .OR.  (IMODE.EQ.LAND.ITYP.EQ.2))
     *        ANSWIM (IP, I AGE, MYR, ITEST, ITYP)  -
     *     .75*(PRED(IP,IBAG,IAGE-2,MYR, ITEST) +
     *         SLOPE(IP,MYR,IAGE-1)*(ODOM(IAGE-1)-ODOM(IAGE-2))  +
     *         SLOPE(IP,MYR,IAGE)*(.625*(ODOM(IAGE)-ODOM(IAGE-l)))) +
     *     .25*(PRED(IP,IBAG,IAGE,MYR,ITEST)  +
     *         SLOPE(IP,MYR, IAGE+1)*(.125*(ODOM(IAGE+1)-ODOM(IAGE))))
CC
CC..Combined  1-3-5 and 2-4-6 biennial cases
CC
   90 CREDIT(IP,IAGE,MYR,ITEST,ITYP)  -
     *  (ANSWNO (IP, IAGE,MYR) -ANSWIM (IP, IAGE,MYR, ITEST, ITYP) )
     *           / (ANSWNO(IP,IAGE,MYR))
CC
  110 CONTINUE
CC
CC..Store  resulting I/M credits
CC
        CREDIT(IP,IAGE,MYR,ITEST,4)  -
     *  (CREDIT(IP,IAGE,MYR, ITEST,2)  + CREDIT(IP,IAGE,MYR,ITEST, 3))/2
CC
  100 CONTINUE
CC
      RETURN
      END

-------
                                   B-18

                 Appendix B :  Tech 4.1 Model Source Code
      SUBROUTINE OUTPUT
CC
CC..Outputs results for emission factors,  bag fractions and
CC..I/M credits.
CC
      COMMON /DATO 5/ OXYE(2,3,4,3),OXYGON
      COMMON /DAT11/ CREDIT(2, 25,12, 3, 4)
      COMMON /DAT12/ ZML (3, 4, 12) , ZML1 (3, 4,12) , ZML2 (3, 4,12)
      COMMON /DAT13/ BFZML1(3,4,12),BFDET1(3,4,12)
      COMMON /DAT14/ DET(3,4,12),DET1(3,4,12),DET2(3,4,12)
      COMMON /DAT21/ BFDET2(3,4,12),BF2ML2(3,4,12)
CC
      INTEGER IP,IBAG
      CHARACTER*4 LABI(3) /'   HC1,'   CO','  NOX1/
      CHARACTER*4 LAB2(5)/'FTP  ',
     *                   'BAG1',
     *                   'BAG2',
     *                   'BAG3',
     *                   'BAGI'/
CC
      NP = 3
      Nl - 1
      N3 - 3
CC
CC..Write out Emission Factors on Device #7
CC
      WRITE(7,100) Nl
      WRITE(7,600)
      DO 20 IP=1,NP
      WRITE(7, 500)
      DO 20 MYR-1,12
      NYR-1980+MYR
      IBAG=1
      T50 - ZML1(IP,IBAG,MYR) +  5.0*DET1(IP,IBAG,MYR)
      T100 - T50 + 5.0*DET2(IP,IBAG,MYR)
      WRITE(7,300) NYR,LAB2(IBAG),LABI(IP),
     *  ZML1(IP,IBAG,MYR),DET1(IP,IBAG,MYR), DET2(IP,IBAG,MYR),
     *  T50,T100,ZML(IP,IBAG,MYR),DET(IP,IBAG,MYR)
   20 CONTINUE
CC
CC..Write out Annual I/M credits  on  Device #8
CC
      WRITE(8,102) N3
      DO 10 ITEST-1,3
      DO 10 MYR-1,12
        NYR-1980+MYR
      DO 10 IP-1,2
        WRITE(8,200)  (CREDIT(IP,IAGE,MYR,ITEST,1),IAGE-1,24),
     *     NYR,LAB1(IP)
   10 CONTINUE
CC
CC..Write out Biennial I/M Credits on Device #9
CC
      WRITE(9,103) N3
      DO 130 ITEST-1,3
      DO 130 MYR-1,12
        NYR-1980+MYR
      DO 130 IP-1,2
        WRITE(9,200)  (CREDIT(IP,IAGE,MYR,ITEST,4),IAGE-1,24),
     *     NYR,LAB1(IP)

-------
                                   B-19

                 Appendix B :  Tech 4.1 Model Source Code


  130 CONTINUE
CC
CC..Write out Bag Fractions on  Device #10
CC
      WRITE(10,101) Nl
      DO 30 IP-1,3
      WRITE(10, 500)
      DO 30 MYR=1,12
      NYR-1980+MYR
      WRITE(10,400) NYR,LAB1(IP),
     * (BFZML1 (IP, IBAG,MYR),BFDET1 (IP, IBAG,MYR) ,BFZML2 (IP, IBAG,MYR),
     *                     BFDET2(IP,IBAG,MYR),IBAG-2,4),
     * BFZML1 (IP, 1,MYR),BFDET1 (IP, 1,MYR), BFZML2 (IP, 1,MYR) ,
     *                  BFDET2(IP,I,MYR)
   30 CONTINUE
CC
  100 FORMAT(II,/,' **',/,
     *' ** MOBILE4.1  LDGV Emission Factors1,/,1 **')
  101 FORMAT(II,/,/,
     *' ** MOBILE4.1  LDGV Bag Fractions  **',/,/,
     *23X,'Bag l',20X,'Bag 2',25X,'Bag 3',25X,'FTP',/,
     *9X, 4 ('   	'), /,
     *9X,4('   ZML1    DET1  ZML2    DET2'),/,
     *9X,4('   	    	  	    	'))
  102 FORMAT(II,/,' **',/,
     *' ** MOBILE4.1  Annual I/M Credits',
     */,'  **')
  103 FORMAT(II,/,' **',/,
     *' ** MOBILE4.1  Biennial I/M  Credits',
     */,'  **')
  200 FORMAT(24F4.3,5X,I4,A4)
  600 FORMAT (18X, '  ZML ', 3X, '  DET1  ', 3X, '  DET2 ' , 3X,
     *  '      Q 50k1,'     @100k',3X,'   ZML ',3X,'   DET ')
  300 FORMAT(
     *  IX,14,'  :  ',2A4,'-',
     *  F6.3,' +  ',F6.3,3X,
     *  F6.3,3X,2F10.3,2(3X,F6.3))
  400 FORMAT(1X,I4,A4,16F7.4)
   500 FORMAT('-')
CC
      RETURN
      END

-------
                                B-20

              Appendix B :  Tech 4.1 Model Source Code


   FUNCTION OXY(IP,ITECH,EMIT)

   COMMON /DAT05/ OXYE(2,3,4,3),OXYGON

   OXY-1- 0

   IEG=0
   IF(EMIT.LT.OXYE(1,1,ITECH,IP)) THEN
     OXY-1.0-OXYCON*OXYE(2,1,ITECH,IP)/100.
     GOTO 99
   ENDIF
   IF(EMIT.GE.OXYE(1,3,ITECH, IP)) THEN
     OXY-1.0-OXYCON*OXYE(2,3,ITECH,IP)/100.
     GOTO 99
   ENDIF
   IF(EMIT.LT.OXYE(1,2,ITECH,IP) .AND.
  *   EMIT.GE.OXYE(1,1,ITECH,IP)) IEG-2
   IF(EMIT.LT.OXYE(1,3,ITECH,IP) .AND.
  *   EMIT.GE.OXYE(1,2,ITECH, IP)) IEG-3

   Xl-OXYE(1,IEG-1,ITECH,IP)
   X2-OXYE(1,IEG,ITECH,IP)
   Yl-OXYE(2,IEG-1,ITECH,IP)
   Y2-OXYE(2,IEG,ITECH, IP)

   ODET-(Y2-Y1)/(X2-X1)
   OZML-Y1-X1*ODET

   OXY-1-(OZML+ODET*EMIT)*OXYCON/100.

99 RETURN
   END

-------
                                   B-21

                 Appendix B  : Tech 4.1 Model Source Code
      SUBROUTINE  CONECT(BIN,OUT)
CC
CC  DOTS(ITECH,A/BfIDOT,IP/ITEST)
CC
      COMMON  /DAT01/  MYR,ISTD,ITECH,IBAG,IP,IAGE,ICUT,ITST
      COMMON  /DAT25/  EXPO(2,4,3,2),DOTS(4,2,4,2,2)
CC
      OUT-EIN
CC
      IF(EIN.LE.DOTS(ITECH,1,1,IP,ITST)) THEN
        SLP = DOTS(ITECH,2,1,IP,ITST)
     *      / DOTS(ITECH,1,1,IP,ITST)
        OUT - SLP * BIN
      ENDIF
CC
      IF(BIN.GT.DOTS(ITECH,1,1,IP,ITST) .AND.
     *   EIN.LE.DOTS(ITECH,1,2,IP,ITST)) THEN
        SLP = ( DOTS(ITECH,2,2,IP,ITST) -
     *          DOTS(ITECH,2,1,IP,ITST) )
     *      / ( DOTS(ITECH,1,2,IP,ITST) -
     *          DOTS(ITECH,1,1,IP,ITST) )
        CPT = DOTS(ITECH,2,1,IP,ITST) -
     *        DOTS(ITECH,1,1,IP,ITST)*SLP
        OUT - CPT + SLP*EIN
      ENDIF
CC
      IF(BIN.GT.DOTS(ITECH,1,2,IP,ITST) .AND.
     *   EIN.LE.DOTS(ITECH,1,3,IP,ITST)) THEN
        SLP - ( DOTS (ITECH, 2, 3, IP, ITST) -
     *          DOTS(ITECH,2,2,IP,ITST) )
     *      / ( DOTS(ITECH,1,3,IP,ITST) -
     *          DOTS(ITECH,1,2,IP,ITST) )
        CPT - DOTS(ITECH,2,2,IP,ITST) -
     *        DOTS(ITECH,1,2,IP,ITST)*SLP
        OUT - CPT + SLP*EIN
      ENDIF
CC
      IF(BIN.GT.DOTS(ITECH,1,3,IP,ITST) .AND.
     *   EIN.LE.DOTS(ITECH,1,4,IP,ITST)) THEN
        SLP = ( DOTS(ITECH,2,4,IP,ITST) -
     *          DOTS (ITECH, 2, 3, IP, ITST) )
     *      / ( DOTS(ITECH,1,4,IP,ITST) -
     *          DOTS(ITECH,1,3,IP,ITST) )
        CPT - DOTS(ITECH,2,3,IP,ITST) -
     *        DOTS(ITECH,1,3,IP,ITST)*SLP
        OUT = CPT + SLP*EIN
      ENDIF
CC
      IF(BIN.GT.DOTS(ITECH,1,4,IP,ITST)) THEN
        OUT - DOTS(ITECH,2,4,IP,ITST)
      ENDIF
CC
      RETURN
      END

-------
                                   B-22

                 Appendix B :  Tech 4.1 Model Source Code
CC
CC.
CC
CC
CC
CC
CC
CC
   BLOCK DATA BD01

.This block data is used to initialize data arrays

   COMMON /DAT10/ EWO<3,4,25,12),EIMW(2,4,25,12,3),EZM(2,4,12)
   COMMON /DAT12/ ZML (3, 4,12) , ZML1 (3, 4,12), ZML2 (3, 4,12)
   COMMON /DAT14/ DET(3,4,12),DET1(3,4,12),DET2(3,4,12)
   COMMON /DAT15/ CWOA<2,25,4,2) ,EWOA(3, 25, 12) ,EZMA(2,12)
   DATA EWO
   DATA EIMW
   DATA EZM

   DATA ZML
   DATA ZML1
   DATA ZML2

   DATA DET
   DATA DET1
   DATA DET2

   DATA CWOA
   DATA EWOA
   DATA EZMA

   END
                  3600*0.0
                  7200*0.0
                    96*0.0
144*0,
144*0.
144*0,

144*0.
144*0,
144*0.
                  400*0.0
                  900*0.0
                   24*0.0

-------
                                   B-23

                 Appendix B :  Tech 4.1 Model Source Code


      BLOCK DATA BD02
CC
CC..Emission Level Data  Block
CC
      COMMON /DAT07/ ESO(2,4,4,2),EHO(2, 4, 4, 2)
      COMMON /DAT19/ GV(4,2),GH(4,2),GS(4,2),BH<4,2),BV(4,2)
CC
CC
CC..Change in the rate of  increase in the number of HIGH emitters
CC..                   BH(ITECH,ISTD)
CC                 MPFI    TBI     Carb   Oplp
      DATA BH /13.8353,13.8353,0.9643,1.0000,
     *          4.4709,15.2076,2.3200,1.0000 /
CC
CC..Change in the rate of  increase in the number of VERY HIGH emitters
CC..                   BV(ITECH,ISTD)
CC                 MPFI    TBI     Carb   Oplp
      DATA BV / 1.0000,1.0000,1-0000,1.0000,
     *          1.0000,1.0000,1.0000,1.0000  /
CC
CC..Growth in the number of HIGHS per 10,000 miles
CC..                     GH(ITECH,ISTD)
CC                MPPI      TBI       Carb     Oplp
      DATA GH /   .0065,    .0065,    .0270,     .02704,
     *            .0090,    .0040,    .0140,     .01298  /
CC
CC..Growth in the number of VERY HIGHS per 10,000 miles
CC..                     GV(ITECH,ISTD)
CC                MPFI      TBI  .     Carb     Oplp
      DATA GV /   .01933,   .01198,   .03762,    .03530,
     *            .00840,   .02012,   .01487,    .00433  /
CC
CC..Growth in the number of SUPERS per 10,000 miles
CC..                     GS(ITECH,ISTD)
CC                MPFI     TBI      Carb    Oplp
      DATA GS  /  .00257,  .00257,  .00257,  .00000,
     *            .00194,  .00194,  .00194,  .00000  /
CC
CC..Average emissions of SUPERS (from 37 EF  & IM vehicles)
CC..             ESO(IP,IBAG,ITECH,ISTD)
CC
      DATA ESO /
CC..1981,1982 model year vehicles
     1  15.259,173.84, 18.376,154.87,16.344,193.62,10.813,150.52,
     2  15.259,173.84, 18.376,154.87,16.344,193.62,10.813,150.52,
     3  15.259,173.84, 18.376,154.87,16.344,193.62,10.813,150.52,
     4   0.000,  0.00,   0.000,   0.00,  0.000,  0.00,  0.000,  0-00,
CC..1983 and newer model year vehicles
     1  11.123,189.00, 7.6414,163.53,12.874,204.06,10.460,179.79,
     2  18.083,183.59, 13.582,177.29,23.081,208.03,12.076,142.32,
     3  12.870,246.85, 13.435,266.05,13.935,265.45,10.435,197.65,
     4   0.000,  0.00,   0.000,   0.00,  0.000,  0.00,  0.000,  0.00  /
CC
CC..Emission Levels of VERY HIGH Emitters at zero miles
CC..       EHO(IP,IBAG,ITECH,ISTD)
CC
      DATA EHO/
CC..1981,1982 model year vehicles
     * 0.606,  18.049, 1.489,   25.589, 0.392,  15.029, 0.349,   18.063,
     * 1.163,  22.059, 2.896,   64.029, 0.583,    7.181, 0.954,   18.611,

-------
                                   B-24

                Appendix B  :  Tech 4.1 Model Source Code


     * 1.950,  33.396, 3.604,  47.486, 1.632,  32.110,  1.301,   25.174,
     * 1.589,  27.650, 2.583,  45.178, 1,225,  21.845,  1.524,   25-403,
CC..1983 and newer vehicles                                      - -._
     * 2 019,  22.301, 5.041,  37.189, 1.742,  22.043,  1.369,   16.248,
     * 2'.242   44 416  2.900   56.086, 2.267,  46.292,  1.705,   32.087,
     * 2.002   36.130  4.320,  55.834, 1.464,  30.566,  1.272,   31.744
     * 1.352,  34.021, 1-441,  40.120, 1.386,  34.668,  1.217,   28-081 /
CC
      END

-------
                                   B-25

                 Appendix B : Tech 4.1 Model Source  Code
CC
CC.
CC
CC
CC
CC.
CC.
CC

CC.
CC.
CC
CC
CC.
CC.
CC

CC.
CC.
CC
CC.
CC.
CC

CC.
CC.
   BLOCK DATA BD03

.Emission Level Data Block

   COMMON /DAT03/ ENOX(4,2,2)
   COMMON /DAT08/ DN(3,4,4,2)
   COMMON /DAT20/ EMO(2,4,4,2),ENO(3,4,4,2)
.Emission level of HIGHs at zero mileage
           EMO(IP,IBAG,ITECH,ISTD)
   DATA EMO/
,1981,1982 model year
  * 0.374,  3.879,  0
  * 0.758,  6.843,  1
  * 0.781,  8.496,  2
  * 0-757,  8.142,  1
.1983 and newer model
  * 0.997,  7.602,  1
  * 0.762,  8.820,  1
  * 0.703,  8.577,  1,
  * 0.840,  8.386,  1
vehicles
954,   5.778,
540,  17.730,
027,  22.919,
494,  17.342,
year vehicles
933,  12.231,
288,  18.793,
740,  17.608,
461,  17.215,
                                        0.272,
                                        0.566,
                                        0.404,
                                        0.474,
                                        0.712,
                                        0.611,
                                        0.376,
                                        0.536,
4.591,  0.128,  0.992,
5.088,  0.530,  2.013,
4.073,  0.551,  5.963,
4.430,  0-735,  8.196,

6.344,  0.834,  6.538,
5.238,  0.651,  8.072,
4.252,  0.543,  9.925,
3.871,  0.951, 10.367 /
.Emission level of NORMAL emitting vehicles at  zero mileage
           END(IP,IBAG,ITECH,ISTD)
    DATA SNO /
,1981,1982 model year vehicles
  *.188, 1.548, .380,.722,5.515,.821,.
  *.279, 3.422, .545,.789,9.081,.902,.
  *.288, 3.067, .678,.752,9.350,1.203,
  *.306, 3.368, .649,.632,8.501,.983,.
,1983 and newer model year vehicles
  *.269, 2.598, .617,.831,7.073,.966,.
  *.242, 2.953, .529,.619,7.074,.902,.
  *.222, 2.209, .635,.586,7.006,1.021,
  *.334, 4.093,.524,.594,10.555,.640,.
                034,.466,.218,.
                138,2.155,.451,
                .127,.911,.491,
                185,1.275,.512,

                080,1.052,.451,
                118,1.337,.395,
                .108,.468,.446,
                207,1.262,.478,
         071,.578,.348,
         .160,1.578,.460,
         .241,2.394,.635,
         .287,3.447,.655,

         .202,2.163,.667,
         .193,2.912,.504,
         .164,1.890,.699,
         .377,4.589,.5217
.Emission deterioration of NORMAL vehicles per  10,000 miles
               DN(IP,IBAG,ITECH,ISTD)
    DATA DN /
.1981,1982 model year
  *.0450,.5807,.1280,
  *.0556,.9654,.1345,
  *.0161,.2520,.1414,
  *.0156,.4582,.1497, .
  *.0253,.3281,.0696,
  *.0535,.8313,.0561,.
  *.0230,.2877,.0440,
  *.0654,1.0489,-0281,
,1983 and newer model
  *.0115,.1554,.0163,
  *.0137,.1382,.0248,.
  *.0134,.1572,.0419,
  *.0188,.1931,.0469,.
  *.0199,.2282,.0593,
  *.0429, .7472,.0555,.
vehicles

0395,.4866,.1112,.0487,.4784,.1557,

0119,.0323,.1264,.0245,.5045,.1626,

0181,.1893,.0657,.0183,.2174,.0876,

.0106,.0813,.0444,-0151,.1088,.0556,
year vehicles

0122,.1838,.0106,.0080,.1105,.0218,

0125,.1606,.0328,.0109,.1213,.0549,

0107, .0732,.0587,.0194,.1298, .0631,

-------
                                   B-26

                 Appendix B :  Tech 4.1 Model Source Code
     *.0248,.2093,.0540,
     *.0623,.4080,.0731,.0164,.1454,.0441, .0128, .1799, .0585/
CC
CC
CC..1983-85/1986+ split  for NOx emissions
CC        ENOX(ITECH,MYR SPLIT,ZML/DR)
CC
      DATA ENOX/
CC..Zero Mile Levels
CC      MPFI    TBI   Carb   OPLP
     * 0.689, 0.585, 0-692, 0.524,
     * 0.539, 0.316, 0.478, 0.524,
CC..Deterioration Rates
CC       MPFI     TBI    Carb    OPLP
     * 0.0185, 0.0470, 0.0558, 0.0540,
     * 0.0185, 0.0470, 0.0558, 0-0540 /
CC
      END

-------
                                   B-27

                 Appendix B : Tech 4.1 Model Source  Code
CC
CC.
CC
CC
CC
CC.
CC.
CC.
CC.
CC.
CC.
CC.
CC.
CC
CC.
CC

CC.
CC.
CC.
CC
CC.
CC

CC.
CC.
CC.
CC
CC.
CC

CC.
CC.
CC.
CC
CC
CC.
CC
   BLOCK DATA EDO4

.I/M Identification Rate  (IDR) effects block  data

   COMMON /DAT16/ XSIDR(2,4,2,3),XHIDR(2, 4,2, 3)
   COMMON /DAT24/ XMIDR(2,4,2,3),XNIDR(2,4, 2, 3)
.The fraction of excess emissions identified by the  short
.test for each emission level group.
.Modified Idle, 2spd, & Loaded  (1.2/220) for total emissions
.rather than Excess emissions for Mobile 4.1, 7/19/91.
.Format:    MPFI         TBI         GARB      OpLoop
           HC   CO     HC    CO     HC   CO    HC  CO
.  81-81
.   83+

                 XSIDR(IP,ITECH,ISTD,ITEST)

    DATA XSIDR/
.Idle test identification rate for SUPERS
  1       .6048,.6968,.6048,.6968,.6048,.6968,.0,    .0,
  2       .8978,.9656,.8978,.9656,.8978,.9656,-0,    .0,
.2500/Idle test identification rate for SUPERS
  1       .6523,.8577,.6523,.8577,.6523,.8577,.0,    .0,
  2       .8978,.9656,.8978,.9656,.8978,.9656,.0,    .0,
.Loaded/Idle test identification rate for SUPERS
  1       .6523,.8577,.6523,.8577,.6523,.8577,.0,    .0,
  2       .8978,.9656,.8978,.9656,.8978,.9656, .0,    .O/

                 XHIDR(IP,ITECH,ISTD,ITEST)

    DATA XHIDR/
.Idle test identification rate for VERY HIGHs
  1       .2736,.3231,.2736,.3231,.3858,.4108,.4568, .5194,
  2       .5676,-6129,.2651,.2695,.3640,.3180, .3640, .3180,
.2500/Idle test identification rate for VERY HIGHs
  1       .2736,.3231,.2736,.3231,.4789,.5331,-6197, .6162,
  2       .6187,.7465,.3616,.4206,.5684,.6832,.5684,.6832,
.Loaded/Idle test identification rate for VERY HIGHs
  1       ,2736,.3231,.2736,.3231,.5476,.6037,.6197, .6162,
  2       .6187,.7465,.3904,.4337,.5684,.6832,.5684,-6832 /

                 XMIDR(IP,ITECH,ISTD,ITEST)

    DATA XMIDR/
.Idle test identification rate for HIGHs
  1       .0506,.1135,.0506,.1135,.0563,.0492,.2274,.1522,
  2       .2507,.2208,.0336,.0613,.0694,.0415, .0694, .0415,
.2500/Idle test identification rate for HIGHs
  1       .0506,.1135,.0506,.1135,.0898,.0834,.2274,.1522,
  2       .3436,.3501,.1924,.1532,.0694,.0415, .0694, .0415,
.Loaded/Idle test identification rate for HIGHs
  1       .0506,.1135,.0506,.1135,.0910,.0896,.2274,.1522,
  2       .3866,.3937,.1924,.1532,.0694,.0415,.0694, .0415 /
                 XNIDR(IP,ITECH,ISTD,ITEST)
       DATA XNIDR/

-------
                                   B-28

                 Appendix B :  Tech 4.1 Model Source Code
CC..Idle test identification rate for NORMALS
     1       .0556, .0774, .0139,.0139,.0188,.0204,.0093,.0131,
     2       .0360,.0414,.0425,.0436,.0023,.0078,.0023,.0078,
CC..2500/Idle test identification rate for  NORMALS
     1       .0556,.0774,.0139,.0139,.0371,.0427,.0201,.0317,
     2       .0575,.0694,.0476,.0514,.0140,.0156,.0065,-0208,
CC..Loaded/Idle test identification rate for NORMALS
     1       .0556,.0774, .0139, .0139,.0371,.0427,.0201,.0317,
     2       .0907,.1023,.0712,.0739,.0140,.0156,.0231,.0403  /
CC
      END

-------
                                   B-29
                 Appendix B : Tech 4.1 Model  Source  Code
CC
CC.
CC
CC
CC
CC

CC
CC
CC
CC
CC
CC
CC
CC
CC.
CC.
CC
CC.
CC

CC.
CC.
CC.
CC
CC.
CC
   BLOCK DATA BD05

. I/M Repair Effects block data

   COMMON /DAT17/ RSUP(2,4,2,3),RHIG(2, 4, 2, 3)
   COMMON /DAT22/ RMAR(2,4,2,3),RNOR(2, 4,2, 3)
   COMMON /DAT25/ EXPO<2,4,3,2),DOTS(4,2,4, 2, 2)

 DOTS(ITECH,A/B,IDOT,IP,ITST)

   DATA DOTS /

  *0.7400,0.7400,.96765,.96765, .41083,.41083,.62235,.62235,
  *1.9223,1.9223,2-0226,2.0226, .60615,.60615,1.1894,1.1894,
  *3.9023,3.9023,3.1063,3.1063, 1.0769,1.0769,1.3254,1.3254,
  *14.282,14.282,8.5543,8.5543, 1.3808,1.3808,1.5286,1.5286,

  *9.2708,9.2708,10-487,10. 487, 4.9900,4.9900,9.8624,9.8624,
  *28.031,28.031,29.550,29.550, 9.4669,9.4669,12.969,12.969,
  *90.038,90.038,53.520,53.520, 12.148,12.148,17.434,17.434,
  *190-66,190.66,134.75,134.75, 20.620,20.620,18.281,18.281,

  *.82667, .82667,.93026,.93026, .40750,.40750,.57641,.57641,
  *1.9846,1.9846,1.9431,1.9431, .59231,.59231,1.0349,1.0349,
  *3.9314,3.9314,2.9862,2.9862, 1.0271,1.0271,1.1413,1.1413,
  *14.282,14.282,8.2523,8.2523, 1.3808,1.3808,1.4141,1.4141,

  *10.334,10.334,10.622,10.622, 4.8950,4.8950,9.2808,9.2808,
  *35.518,35.518,29.053,29.053, 9.8631,9.8631,12.489,12.489,
  *104.50,104.50,54.282,54.282, 11.925,11.925,13.190,13.190,
  *190.66,190.66,136.97,136.97, 20.620,20.620,13.596,13.596 /

 EXPO(IP,ITECH,ITEST,A/B)   Aexp(-B/x)-y

   DATA EXPO /
  * 0.6682, 6.1319, 0-7476, 8.4157, 1.3629, 13.4934,1.1487, 9.9503,
  * 0.6736, 6.8360, 0.6839, 7.9328, 1.1585, 10.7263,1.2079, 9.1495,
       0.6736,  6.8360,
       0.2676,  3.4348,
                    0.6839, 7.9328, 1.1585, 10.7263,1.2079,  9.1495,
                    0.3419, 3.0945, 0.7034, 5.4373, 0.5682,  2,1526,
  * 0.2911, 6.3200, 0.3262, 2.8702, 0.6309, 4.1224, 0.4449, 1.7352,
  * 0.2911, 6.3200, 0.3262, 2.8702, 0.6309, 4.1224, 0.4449, 1.7352 /

.Emission level after repairs expressed as a fraction of
.the emission level before repairs.

        RSUP(IP,ITECH,ISTD,ITEST) 4/11/91
    DATA RSUP/
.2500/Idle Test
  1       .958,
  2       .958,
.2500/Idle Test
  1       .958,
  2       .958,
.2500/Idle Test
  1       .958,
  2       .958,
emission effect
.975, .968,  .952
.975, .968,  .952
emission effect
.975, .968,  .952
.975, .968,  .952
emission effect
.975, .968,  .952
.975, .968,  .952
from repairs for SUPERS
,  .864, .930, .000, .000,
,  .864, .930, .000, .000,
from repairs for SUPERS
,  .864, .930, .000, .000,
,  .864, .930, .000, .000,
from repairs for SUPERS
,  .864, .930, .000, .000,
   864, .930, -000, .000 /
                 RHIG(IP,ITECH,ISTD,ITEST)

-------
                                   B-30

                 Appendix B :  Tech 4.1 Model Source  Code
CC.


CC.


CC.


CC
CC.
CC

CC.


CC.


CC.
CC
CC.
CC

CC.
CC.
CC.
    DATA RHIG/
.2500/Idle Test
  1        .609,
  2        .609,
.2500/Idle Test
  1        .609,
  2        .609,
.2500/Idle Test
  1        .609,
  2        .609,
    DATA RMAR/
.2500/Idle Test
  1       .609,
  2       .609,
.2500/Idle Test
  1       .609,
  2       .609,
.2500/Idle Test
  1       .609,
  2       .609,
    DATA RNOR/
.2500/Idle Test
  1       .375,
  2       .375,
.2500/Idle Test
  1       .375,
  2       .375,
.2500/Idle Test
  1       .375,
  2       .375,
emission effect of
.772, .879,  .909, ,
.772, .879,  .909, ,
emission effect of
.772, .879,  .909, ,
.772, .879,  .909, ,
emission effect of
.772, .879,  .909, .
.772, .879,  .909, ,
 repairs for
 668,  .760,
 668,  .760,
 repairs for
 668,  .760,
 668,  .760,
 repairs for
 668,  .760,
 668,  .760,
 VERY
.627,
.627,
 VERY
.627,
.627,
 VERY
.627,
.627,
HIGHS
.741,
.741,
HIGHS
.741,
.741,
HIGHS
.741,
.741 /
                 RMAR(IP,ITECH,ISTD,ITEST)
                 (Same as for Very Highs)
emission effect of
.772, .879, .909,
.772, .879, .909,
emission effect of
.772, .879, .909,
.772, .879, .909,
emission effect of
.772, .879, .909,
.772, .879, .909,
 repairs for
.668,  .760,
,668,  .760,  .
 repairs for
,668,  .760,  .
,668,  .760,  ,
 repairs for
,668,  .760,  .
,668,  .760,  .
                 RNOR(IP,ITECH,ISTD,ITEST)
emission effect of
.508, .352, .390, ,
.508, .352, .390, ,
emission effect of
.508, .352, .390, ,
.508, .352, .390, ,
emission effect of
.508, .352, .390, ,
.508, .352, .390, ,
 repairs  for
 263,  .280,  ,
 263,  .280,  ,
 repairs  for
 263,  .280,  .
 263,  .280,  .
 repairs  for
 263,  .280,  ,
 263,  .280,  ,
 HIGHS
 627,  .741,
 627,  .741,
 HIGHS
 627,  .741,
 627,  .741,
 HIGHS
 627,  .741,
 627,  .741  /
NORMALS
420,  .603,
420,  .603,
NORMALS
420,  .603,
420,  .603,
NORMALS
420,  .603,
420,  -603  /
      END

-------
                                   B-31

                 Appendix B :  Tech 4.1 Model Source Code
CC
CC.
CC
CC
CC.
CC.
CC

CC
CC.

CC.

CC.

CC.

CC.

CC.

CC.

CC.

CC.

CC.

CC.

CC.

CC
CC.
CC
CC
CC.
CC
CC
c
c
c
   BLOCK DATA BD06

.Fleet Description Block

   COMMON /DAT02/ AMIL(25),ODOM(25),TMILE(25),WGT(25)
   COMMON /DAT06/ FRAG(4,12)

.Technology Sales Fractions Projections
         FRAC(ITECH,MYR)

   DATA FRAC /
    MPFI  TBI   Carb  Oplp
.1981 Model Year
  * .061, .029, .629,  .281,
.1982 Model Year
  * .062, .106, .506,  .325,
.1983 Model Year
  * .088, .183, .486,  .244,
.1984 Model Year
  * .110, .282, .551,  .058,
.1985 Model Year
  * .307, .208, .408,  .077,
.1986 Model Year
  * .400, .270, .325,  .005,
.1987 Model Year
  * .374, .365, .250,  .010,
.1988 Model Year
  * .492 ,-407, .101,  .000,
.1989 Model Year
  * .597, .275, .125,  .003,
.1990 Model Year
  * .761, .219, .018,  .001,
.1991 Model Year
  * .795, .202, .003,  .000,
.1992 and Newer Model Years
  * .815, .185, .000,  .000 /

.Fleet January 1st VMT weighting  factors

   DATA WGT / 0-030, 0.120, 0.111,  0-099, 0-088,
  *           0.078, 0-068, 0.060,  0.054, 0.048,
  *           0-043, 0.038, 0.033,  0.028, 0.024,
  *           0.020, 0.017, 0.013,  0.010, 0.007,
  *           0.004, 0.002, 0.002,  0.001, 0.003  /
.Cumulative January 1st mileage accumulation by age

   DATA ODOM/ 1.3118, 2.6058,  3.8298,  4.9876,  6.0829,
  *           7.1190, 8.0991,  9.0262,  9.9031,10.7326,
  *          11.5172,12.2594,12.9615,13.6257,14.2540,
  *          14.8483,15.4104,15.9421,16.4451,16.9209,
  *          17.3712,17.7969,18.1997,18.5806,18.9410 /

   END
   BLOCK DATA BD07
Oxygenated Fuels Effects Block

   COMMON /DATO 5/ OXYE(2,3,4,3),OXYGON

-------
                                   B-32

                 Appendix B :  Tech 4.1 Model Source Code
      DATA OXYGON / 1.0 /
C
C  OXYE(XY,IEMIT,ITECH,IP)
C
      DATA OXYE /
C HC
     * 24*0.00,
C CO
     * 3.50,3.46, 15.26,9.87, 95-79,11.44,
     * 3.72,4.93, 15.26,9.87, 95.79,11.44,
     * 4.18,6.77, 15.26,9.87, 95.79,11.44,
     * 3.50,9.97, 15.26,9.97, 95.79, 9.97,
C NOx
     * 24*0.00 /
C
      END

-------
              APPENDIX C









EVAPORATIVE EMISSIONS AND RUNNING LOSS




      EMISSION FACTOR DERIVATION

-------
                      EVAPORATIVE HC EMISSIONS


1.0  Introduction

     There have been  more vehicles tested by EPA's Emission Factors
Program  (EFP)  since  the issuance  of MOBILE4.   From the  EFP test
facility  located  at  Motor  Vehicle  Emissions Laboratory  (MVEL)  in
Ann  Arbor,  Michigan  (EFP-MVEL),  EPA continues   recruiting   in-use
vehicles  and  testing  them with  the certification test procedure —
a  diurnal test,  followed by  exhaust emissions  test on  a Federal
Testing  Procedure  (FTP)  cycle,   and  then  a  hot   soak   test.   A
certification test  fuel  with fuel volatility level of  9.0 psi Reid
Vapor Pressure  (RVP)  is  used.   Two additional  fuels with 10.4 and
11.7 psi  RVPs   were   also  used  during  FY84   through   FY89  to
characterize the  evaporative emissions effect due to different fuel
volatility  levels  that  were  commercially  available  for   in-use
vehicles.   Since  FY90,   the  9.0 psi  RVP fuel  has  been   the only
gasoline fuel used in the EFP-MVEL testing.

     EPA  added  "Hammond Program"  (EFP-Hammond)   to  the  so-called
MVEL  "traditional"  EFP  during  FY90.  A  brief description  of this
new program is  given  in section  1.1.  Due  to the addition  of this
new  test  program,  a  different  modeling  approach  is  used  for
MOBILE4.1's evaporative diurnal and hot soak emissions.

     For  example,  in  the MOBILE4 evaporative diurnal and  hot soak
emissions   model,  there  were   seven   elements  of   evaporative
emissions:  standard  level,  emissions due to  insufficient canister
purge,  emissions   due   to   excess  fuel   volatility,  problem-free
emission  levels,  emissions  due  to malmaintenance   and/or  defect,
non-tampered  emission  levels,  and  emissions  from tampering.   The
problem-free  emission  rates  were  the  sum  of  the  first  three
elements.   The  non-tampered  emissions   were  the  emissions  from
problem-free vehicles plus the  emissions  from malmaintenance and/or
defect.   Finally,  the  in-use  evaporative   diurnal  and  hot  soak
emissions were estimated by  accounting  for emissions  from  both the
non-tampered and tampered fractions of the vehicle fleet.

     However,  in  the MOBILE4.1  evaporative emissions model,  there
are three  different types of  evaporative HC emissions:   emissions
from  pass  vehicles,  emissions  from  vehicles  that failed  pressure
test, and  emissions from vehicles  that   failed  purge test.   These
three types of  emissions are defined for both diurnal and hot soak
emissions in MOBILE4.1.
                              -1-

-------
1.1  Hammond Program

     EPA's  Hammond  program was  initiated  in FY90.   This  program
utilizes  the   testing   facilities  for   the   State  of   Indiana's
Inspection  and  Maintenance   (I/M)  programs   located   at   Hammond,
Indiana.  While  waiting  for the required  I/M  tests, in-use  vehicle
owners  were  asked to participate  additional tests defined  by  EPA's
new  EFP.    One  of  the  tests  those  I/M  vehicles  were  asked  to
participate  was  a  transient  cycle  called  "IM240,"   which  is   a
shortened   version   of    the   certification   FTP   cycle.    All
participating  vehicles  were  also  tested  for   their   evaporative
emissions  control system  to see  if there  exists  either   pressure
and/or  purge failure.

     The pressure  test  is  performed by using a gauge to  measure the
pressure  between  a  vehicle's   fuel   tank   rollover   valve  and
evaporative  canister,  by  introducing  nitrogen  at 14"  of  water
pressure.  Two minutes  after  the  initiation of  this test,  if the
evaporative  system holds   less  than 8"  pressure,  the  vehicle is
considered  to  have  pressure problem  in  its  evaporative emissions
control system, thus failing the pressure test.

     The purge test  is  performed  by  using a  flow meter measuring
the  purge  air  between  a  vehicle's  canister   and  engine.   The
cumulative purge  air over  the  entire  IM240 cycle is measured.  If
the  flow  rates  are  less   than  1.0  cubic  liter,  the  vehicle's
evaporative  emissions   control   system   is   considered  to  have
insufficient purge, thus failing the purge test.

     As of  January  30,  1991,  2,497 I/M  vehicles were  tested for
potential pressure  and  purge  failures under  the  Hammond  program.
The model years  of  these in-use vehicles ranged from 1976  to 1991.
Overall failure rates were:   12.7 percent  for the  pressure test,
11.5 percent  for  the purge  test,  and  21.1 percent  for  either.
These  failure  rates  were analyzed  and  then characterized  in terms
of vehicle's age, and are summarized in Table 1.

     Some of the failed vehicles were also  tested  for their diurnal
and  hot soak  emission  levels.  The results  seem to  suggest  that
emissions from vehicles  that  failed pressure  test are  similar to
the  emissions  from vehicles  with  no evaporative  emissions control
system, and  are somewhat higher  than  the  emissions from  vehicles
that failed purge test.


1.2  Modeling Approach

     As  mentioned   in  section 1.0,   the  MOBILE4.1   evaporative
emissions  model  include:   emissions from pass  vehicles, emissions
from  vehicles   that  failed  pressure  test,   and  emissions  from
vehicles that failed purge test.
                              -2-

-------
     Test results obtained from EFP-MVEL were used to:

     a)   characterize the fuel volatility effect on emissions,
     b)   derive emission levels for pass vehicles, and,
     c)   describe  the  emissions  effects  due  to  other parameters,
          such as fuel delivery system, vehicle type, etc.

     Results obtained from EFP-Hammond program were used to:

     a)   define pass versus pressure and/or purge failure vehicles,
     b)   estimate  the emission  levels  due to pressure and/or purge
          failures, and,
     c)   see if  the pressure/purge test  results  are applicable to
          and/or  are consistent  with  mechanics'  visual  diagnosis
          check in EPA's EFP at MVEL.

     In addition to test  data,  a theoretical vapor generating model
(based on  Wade  equation) was  also used for  the diurnal  emissions
model to:

     a)   determine the emissions  effect from a combination of fuel
          volatility,  temperature  rise,   fuel  tank  capacity,  and
          tank fill level,

     b)   describe  the  emissions  from  vehicles that  failed  either
          pressure or purge test, and,

     c)   define the emission levels from vehicles that have  two or
          more consecutive no-driving days.

     The  "pass"  vehicles  are  defined  as  vehicles  that  are  not
tampered, nor failing pressure/purge tests.  Therefore, in  terms of
the  MOBILE4  evaporative emissions  categories, the  pass  vehicles
include vehicles  that have  no  visible nor  measurable problems  in
their evaporative  emissions control system  components, or  problems
that do  not appear to directly  relate to pressure  leaks or  purge
deficiency.

     Table 2  provides a   list  of  the  criteria used  in  defining
pressure/purge  failures   as well  as  tampered  vehicles   from  the
EFP-MVEL data  base.  As  can be  seen,  "tampered"  in  the  MOBILE4.1
model is now a subsample of the failure vehicles by  definition.   In
general,  pressure  failures  indicate  the  potential  existence of  a
leakage   on  vehicle's   evaporative   emissions   control   system
components,  which  include:  gas  cap,  filler  neck,  sending  unit,
rollover valve,  and vent  hoses.   Purge  failures are  usually  caused
by   an   in-operative   canister    purge   solenoid   or  valve,   or
disconnected, missing, or damaged purge hoses.

-------
2.0  Diurnal Emissions


     As in  MOBILE4,  the diurnal emissions  (in unit of  grams  per  one
hour  test)   are  described  as  a  function  of Uncontrolled  Diurnal
Index  (UDI).   This  UDI is  the uncontrolled  diurnal emissions  (in
grams), calculated  from a  combination of weathered fuel volatility
level  (in  psi  RVP),  fuel  tank temperature  rise (in °F), fuel tank
capacity (assumed a sales-weighted 16.2  gallon  tank for LDGVs),  and
percent  tank  fill  (set  as  54.57%),  normalized at  FTP conditions
(which include using 9.0 psi RVP fuel,  60  to 84°F temperature  rise,
and  40% tank  fill).   The  calculations  of  uncontrolled   diurnal
emissions are based upon a theoretical  vapor generating model  (Wade
equation).   For  example,   with   9.0 psi   RVP  fuel,  72  to  96°F
temperature  rise  and  40%  tank fill,  the  UDI  is  calculated  to  be
1.7448.   The  followings  are  some  examples  of  the  calculated UDI
values, all  at  60  to 84°F temperature rise  and  40% tank fill,  with
different fuel volatility levels:

              Fuel Volatilities in psi RVP
             9.0      10.5    11.5     11.7
     UDI     1.0     1.4567  1.9581   2.0677

The UDIs  are used to  describe diurnal  emissions  for pass  vehicles
as well as for vehicles that failed pressure/purge tests.


2.1  1981+ Light-Duty Gasoline-Powered Vehicles and Trucks

     The MOBILE4.1 diurnal emissions  for model  years 1981 and later
light-duty  gasoline-powered  vehicles  and trucks  (LDGVs  and  LDGTs)
are  based  upon  EPA's  EFP  test  results.    Table  3  summarizes  the
emissions data obtained from EFP-MVEL in terms of  sample sizes and
average  emissions   for  various   categories   (such   as  pass/fail
vehicles,  different fuel delivery  systems and vehicle types).

     Note that due to  a lack of class 2 truck data  in  the  EFP-MVEL
sample, the  two classes of light-duty trucks (classes  1  and  2) are
combined as  one  LDGT  group.   Also,   because of  insufficient data,
two  types   of  fuel   injection   fuel  delivery   systems   (ported
fuel-injection  [PFI]  and  throttle  body fuel-injection  [TBI])  are
also being combined as one fuel-injected (FINJ)  group for the LDGTs.


2.1.1  Pass Vehicles

     Diurnal emissions  for  pass  vehicles  were  derived  from EPA's
EFP-MVEL data.  The  derived equational  form for model  years 1981+
pass light-duty vehicles and trucks  is:
                              -4-

-------
              3 + b * UDI + C  * UDI1
                                  (2.1)
where:
     Vehicle Type
     /Fuel System

     LDGV/CARB
     LDGV/PFI
     LDGV/TBI

     LDGT/CARB
     LDGT/FINJ
    Regression Coefficient
-4.3304
•0.0879
-0.8233

•7.6803
•3.8537
5.9904
0.0
0.0

9.7703
4.9437
0.0
1.4179
2.0433

0.0
0.0
The  coefficients  from  equation  (2.1)  were  derived  from EFP-MVEL
data, where  the  three UDI  values  corresponding to  9.0,  10.4, and
11.7 psi RVP  fuel volatility  levels  were 1.0,  1.4567,  and 2.0677,
respectively.   The constant  terms of  equation (2.1)  (denoted as "a"
above) were adjusted  so that when UDI is  1.0  the predicted diurnal
emissions are equal  to the average diurnal emissions at 9.0 psi RVP
from the EFP-MVEL sample.

     For  carbureted   vehicles,   the  equation  (2.1)   is  to  be
extrapolated when the UDI values are  greater than 2.0677.  However,
when  the UDI  values  are  less  than  1.0,  new  sets  of  coefficients
listed  below  are  used.   These  coefficients  were  derived from  a
two-step analysis:  first, all carbureted  vehicle data  were used to
fit an exponential function,  then,  the predicted emissions for UDIs
at 1.0 and lower were used to  fit a linear regression line.  Again,
the  constant   terms   were adjusted  so  that  when  UDI  is 1.0 the
predicted  diurnal  emissions  are  equal  to  the  average  diurnal
emissions at  9.0  psi  RVP  from the EFP-MVEL sample.   In addition,  a
lower bound of the diurnal emissions is set to  equal  to the resting
loss emissions.
     Vehicle Type
     /Fuel System

     LDGV/CARB
     LDGT/CARB
    Regression Coefficient
•0.2888
 0.1412
1.9488
1.9488
0.0
0.0
     For   fuel-injected   vehicles,   equation (2.1)   is   to   be
extrapolated  when the  UDI  values  are  beyond  the  existing  data
(i.e.,  UDI  values  are  either  less  than  1.0  or  greater  than
2.0677).  At  low UDI values  the  calculated  diurnal emissions  may
become negative.  A  lower bound of the diurnal emissions  is set to
equal to the resting loss emissions.

     Graphic presentations of  1981+ LDGV/LDGT diurnal  emissions as
a function of UDI for pass vehicles are shown in Figures 1 and 2.
                              -5-

-------
2.1.2  Failed Vehicles

     To  estimate  the  diurnal  emission  levels  for  vehicles  that
failed either  pressure or  purge test,  two  steps of  analyses  were
used:  1)  the  theoretical vapor generating model, and,  2) data  from
EFP-Hammond.

     The  theoretical vapor  generating  model  was used  to describe
the  relationships  between UDI  and  the  emission  levels  from  failed
vehicles  (see Attachment).   It  is assumed that fuel delivery  system
is  not  a significant  parameter on  failed  diurnal  emission levels,
although  the  emissions simulation  was  based  on a  late model-year
fuel-injection  system.   Uncontrolled  diurnal emissions  (in  grams)
were  calculated at  various  fuel  volatility levels  (from 7.5 to
11.0 psi  RVP)   for  the  temperature   rise  of  72  to 96°F.   These
calculated  uncontrolled  diurnal   emissions  were   used   for  the
pressure   failure   levels.    The purge  failure  levels  were  the
calculated  uncontrolled   emissions   subtracted  by  a  backpurge
effect.    The   g/test  diurnal  emissions   for   failed   LDGVs  are
expressed as linear functions of UDI:

     Dlraiied  Pressure - 2.2002 +  7.5626 * UDI

     Dlpalied  Purge - -4.2897 -r 7.4563 * UDI

At UDI -  1.7448 (72 to 96°F  temperature  rise,  9.0 psi  RVP  fuel,  and
40% tank  fill),  the diurnal emissions estimated  from the  above  two
equations  are  15.40  and   8.72 grams   for   vehicles  that   failed
pressure  and purge test, respectively.

     From  EFP-Hammond data  base,  30  LDGVs  (as  listed  in Table 4)
were also  tested  for diurnal  emissions  under the conditions  of 72
to  96°F  temperature rise, 9 psi RVP  fuel,  and 40%  tank  fill (i.e.,
also with UDI « 1.7448).  These vehicles were found  to  have failed
either  pressure or  purge  test  at  the Hammond  I/M Lane.   Their
average diurnal emissions were:

                                     G/Test
     Category            N           Diurnal

     Failed Pressure    12           27.73
     Failed Purge       18           15.58

     The  average  diurnal  emission  levels  from  EFP-Hammond  were
compared  with  the  estimated emissions  from  the theoretical  vapor
generating  model.    The  data  versus   model   ratios  of   1.801185
(  = 27.73/15.40    for    the   pressure    failure)    and   1.786687
(  = 15.58/8.72   for  the  purge  failure)  were  the  multiplicative
factors  used  to  derive  new diurnal  emission equations  for  LDGVs
that failed either pressure or purge test:
                              -6-

-------
     Dlpailed Pressure - 1.801185 *  (2.2002 + 7.5626 * UDI)
                       = 3.9630 + 13.6216  * UDI                 (2.2)

     Dlraiied Purge = 1.786687 * (-4.2897  + 7.4563 * UDI)
                    - -7.6643 + 13.3221  *  UDI                   (2.3)

     Note that  equations (2.2) and  (2.3) were  derived based  on a
sales-weighted  LDGV fuel tank  capacity of  16.2  gallons.   The fuel
tank capacities  of LDGTs are,  in general,  larger  than  the LDGVs.
Many of the  LDGTs  are also equipped with  an  auxiliary fuel tank of
a  similiar   capacity  as   the   primary  tank.   Based   on  EPA's
Certification  data  base,  the  sales-weighted  primary  fuel  tank
capacity  for  model  year   1988  LDGTls  was  19.0  gallons.   When
including the  auxiliary fuel tank,  the sales-weighted capacity for
LDGTls  became  20.8 gallons.   Therefore,   an   average  fuel  tank
capacity  ratio  of  1.17  (19.0  gallon/16.2  gallon)   is  used  to
estimate the diurnal emissions of failed LDGTs:
LDOTs
                    = 1.17 * DI F a i I e 4 LDGVs           (2.4)
     This approach  implies that  the  diurnal emissions  from  failed
LDGTs  are  117  percent  higher  than  those from failed  LDGVs.   This
assumption is  well-supported from EFP-MVEL  test  data.   As can  be
seen from Table 3,  the average diurnal  emissions  from failed  LDGTs
are 125 percent higher at  11.7 psi RVP  fuel,  and  138 percent  higher
at 9.0 psi RVP fuel, than  the average emissions from failed LDGVs.

     Under FTP  conditions, the diurnal  emissions  for  vehicles  that
failed either pressure or  purge test are calculated to be:

                             Diurnal (G/Test)
        Vehicle              Failed      Failed
         Type               Pressure     Purge

        LDGVs               17.58         5.66
        LDGTs               20.57         6.62

     In  general,  diurnal  emission  levels for vehicles  that  failed
either pressure  or  purge  test  at any  other UDI  values  for  1981+
LDGVs  and LDGTs can  be  calculated  from equations  (2.2)  through
(2.4).   Note that the pressure  failure  equation can  be extrapolated
down to  UDI « 0.0 without getting  negative emission  values.   When
UDI is  at  0.58  or  lower,  however,  the  purge  failure  equation  is
predicting zero  or negative  diurnal emissions.   A lower  bound  of
the failed  purge diurnal  emissions  is  also set  to  equal to  the
resting loss emissions.

     Graphic presentations of  1981+ LDGV/LDGT diurnal emissions  as
a function of UDI for vehicles failed either pressure  or purge  test
are shown in Figures 3 and 4.
                              -7-

-------
2.1.3  Pre-1981 LDGV Diurnal Emission Rates

     Pre-1981   model   year   LDGVs  are   predominately   carbureted
vehicles.   In  this  section,   the   diurnal  emissions   from   pass
vehicles  and  from vehicles  that failed  either pressure or  purge
test for pre-1981 LDGVs are determined.  The following is a  list  of
methodologies used for deriving these emissions:

     1)   For  pre-1971  no  evaporative  emission  standard vehicles,
all  three  emissions  are  set  to  be  the  same  as  the   MOBILE4
uncontrolled  emission  rates.   However,  if  the MOBILE4 uncontrolled
emission  rates  are lower  than the.  estimated emission  levels   from
vehicles  that  failed   pressure   test,   the  new  failed pressure
emission levels are used.

     2)  Emissions are calculated from  equations derived for  1981+
LDGV  group   (e.g.,  for   diurnal  emissions,  use  equations  (2.1)
through (2.4) defined in sections 2.1.1 and 2.1.2).

     3)  As the MOBILE4 emission rates came  from the  average values
of test data,  it  is  assumed that data from various model years were
results from  vehicles  tested  at  about  the  same  age  by  EPA's  EFP.
The average mileages of the current 1981+ sample were:

     VTYP     CARS      PFI          TBI         FIKJ

     LDGV    55,057    42,115      48,890
     LDGT    52,794                            44,759

These  average  mileages   fell  into  4-5  year  old  category.   The
fractions of  the  4 year  old LDGV  fleet  that failed either pressure
or purge  test  are (from  Table 1)  0.053  and 0.060,  respectively.
With the  emission levels from vehicles  that failed either pressure
or purge  test  being  already  defined,  the  following  relationships
are established:

     E » 0.887 * Epass  +  [0.053 * EFaile«l Pressure]
         + [0.06 * Efailed Purge]

Therefore,  the emission   levels   for  pass  vehicles  can  be  back
calculated from the above equation.

     4)   Assume the same  MOBILE4  uncontrolled emission rates.

     5)   Use the emission ratio from:

          (MOBILE4MYRi  /  MOBILE4HYR2)
        = (MOBILE4.1MYRi  / MOBILE4.1MYR2)

     6)   Use the emission ratio from:

          (Failed  PressureMYRi  /  Failed  Pressure,,YRZ)
        » (Failed  PurgenYRi / Failed
                              -8-

-------
     The MOBILE4 FTP diurnal emissions  in  g/test for pre-1981 LDGVs
at two different RVP fuels were:
           MYR Group

           Pre-1971
             1971
           1972-77
           1978-80
M4 Diurnal (G/Testl
9.0 RVP    11.5 RVP
26.08
16.28
 8.98
 5.16
47.99
38.58
23.53
14.47
The derived  MOBILE4.1  LDGV FTP diurnal  emissions for  the  two RVP
fuels are:

                      FTP  Diurnal Emissions (G/Test)
MYR Group

Pre-1971
 1971
1972-77
1978-80
1981+ Garb

Note  that  the 1981+  carbureted diurnal emissions  were calculated
from  equations (2.1)  through  (2.3).   They were  listed  above  for
comparison  purposes   only,   since  all   pre-1981   LDGVs  are   of
carbureted fuel delivery  system.   As  in MOBILE4,  the  pre-1981  FTP
LDGV  diurnal   emissions   in  g/test for any  other  RVP  fuels  are
calculated  by  interpolating  between  these  two  given  emission
levels.

     Using the assumption that in-use  vehicles were tested at  the
age  of  four,   the  MOBILE4.1  FTP  diurnal  emissions  in  g/test  for
LDGVs at two different RVP fuels are:
at 9
.0 RVP
Failed

26
15
8
4
1
Pass
.08l
.033
.27s
.393
.662
Purge
26
26
9
5
5
.081
.081
.33a
.66*
.66*
Failed
Pressure
26
26
20
17
17
.08l
.08'
.39s
.582
.582
47
37
22
13
7
Pass
.991
.383
.80s
.243
.402
at
11.5
Failed
Purge
47
47
23
18
18
.991
.99l
.S33
.422
.422
RVP
Fai

led
Pressure
47.
47.
35.
30.
30.
99'
99l
454
642
642
           MYR Group

           Pre-1971
             1971
           1972-77
           1978-80
           1981+ Garb
M4.1 Diurnal (G/Test)
9.0 RVP   11.5 RVP
26.08
16.28
 8.98
 5.16
 2.74
 47.99
 38.58
 23.53
 14.47
  9.29
                              -9-

-------
2.1.4  Pre-1981 LDGT1 Diurnal Emission Rates

     In  MOBILE4,  as  well as  its earlier  versions,  pre-1981  LDGV
emission  rates  were used  directly  for  pre-1981 LDGTls, since  there
were  insufficient  test   data  from  LDGTls  to  derive   their  own
emission  rates  and  LDGTls  have  the  same   evaporative emission
standards  as the  LDGVs.   The  MOBILE4  FTP diurnal  emissions  in
g/test for pre-1981 LDGTls at two different RVP fuels were:

                        M4 Diurnal (G/Test)
           MYR Group    9.0 RVP     11.5 RVP
           Pre-1971
             1971
           1972-77
           1978-80
26.08
16.28
 8.98
 5.16
47.99
38.58
23.53
14.47
     However,  as  discussed in section  2.1.2,  LDGTls  in general are
equipped with  larger capacity of  fuel  tank  than the  LDGVs.   From
model  years  1981+  EF  data,  where  the  majority of truck  data were
available,   the average  LDGT1  diurnal  emissions  are  found  to  be
significantly  higher  than their  LDGV  counterpart  (see  Table 3).
Thus, it is reasonable for LDGTl's  to have higher diurnal emissions
than  the  LDGVs,  whether  for  pass  vehicles,  or  for  vehicles that
failed either pressure or  purge test.

     The MOBILE4.1  LDGT1  diurnal emission rates  were  estimated  by
multiplying  a  factor  of 1.17 (the  ratio  of the average  LDGT1 and
LDGV fuel tank capacities, as discussed in section 2.1.2).

                      FTP  Diurnal Emissions (G/Test)

Pass
at 9.0 RVP
Failed Failed
Purge Pressure

Pass
at 11.5 RVP
Failed Failed
Purge Pressure
30.51
17.59
9.68
5.14
2.092
30.51
30.51
10.92
6.62
6.62
30.51
30.51
23.86
20.57
20.57
56.15
43.73
26.68
15.49
11. 452
56.15
56.15
27.88
21.55
21.55
56.15
56.15
41.48
35.85
35.85
MYR Group

Pre-1971
  1971
1972-77
1978-80
1981+ Carb

Note  that  the  1981+  carbureted diurnal  emissions were  calculated
from  equations  (2.1)  through  (2.4).   They  were  listed  above  for
comparison   purpose   only,  since   all   pre-1981  LDGTls   are   of
carbureted fuel  delivery system.  As  in  MOBILE4, the  pre-1981  FTP
LDGT1  diurnal  emissions  in  g/test  for  any  other  RVP  fuels  are
calculated by interpolating between these two given emission levels.
                              -10-

-------
     Using the  assumption that  in-use vehicles were  tested at the
age  of  four,  the  MOBILE4.1  FTP diurnal  emissions  in  a/test for
LDGTls at two different RVP  fuels are:

                        M4.1 Diurnal  (G/Testl
           MYR Group    9.0  RVP     11.5 RVP

           Pre-1971     30.51       56.15
             1971       19.05       45.13
           1972-77      10.51       27.54
           1978-80       6.05       16.93
           1981+ Garb    3.34       13.35

As can  be seen,  the  new LDGT1  diurnal  emissions  for MOBILE4.1 are
slightly higher than the diurnal  rates assumed for MOBILE4.


2.1.5  Pre-1981 LDGT2 Diurnal Emission Rates

     In MOBILE4,  the  same  post-1979  LDGT1 diurnal  emissions  were
assumed for  LDGT2s,  based  on  their  similar  evaporative  emissions
standards  and  their  similar   fuel  tank  capacities.   The  diurnal
emissions for  pre-1979 LDGT2s  were  assumed to be  the same  as  the
pre-controlled  HDGV  diurnal  emission  levels,  as  there  were  no
separate vehicle  category  of LDGT2 before  the model  year  of 1979.
The  MOBILE4   FTP diurnal  emissions  in g/test  for  LDGT2s   at  two
different RVP fuels were:

                        M4 Diurnal (G/Test)
           MYR Group    9.0 RVP     11.5 RVP

           Pre-1979     42.33       77.89
           1979-80       5.16       14.47

     Using the  same methodology  as  in LDGTls (section  2.1.4),  the
MOBILE4.1 FTP  g/test diurnal emissions  for LDGT2s  at  two  different
RVP fuels are:

               	FTP  Diurnal Emissions (g/test)	
                    at 9.0 RVP            	at  11.5 RVP
                      Failed   Failed            Failed   Failed
MYR Group      Pass    Purge  Pressure   Pass     Purge  Pressure

Pre-1979      42.33    42.33   42.33    77.89     77.89   77.89
1979-80        5.14     6.62   20.57    15.49     21.55   35.85
1981+ Garb     2.09*    6.62   20.57    11.452    21.55   35.85
                              -11-

-------
Note  that the  1981+ carbureted  diurnal  emissions  were  calculated
from  equations  (2.1)  through  (2.4).   They were  listed  above  for
comparison purpose  only,  since all 1979-80  LDGT2s are of  carbureted
fuel  delivery  system.    As  in  MOBILE4,   the   pre-1981   FTP  LDGT2
diurnal  emissions  in g/test  for  any  other RVP fuels  are  calculated
by interpolating between  these  two given emission levels.

     Using the assumption  that  in-use vehicles were  tested at  the
age  of  four,  the  MOBILE4.1  FTP diurnal  emissions  in  g/test  for
LDGT2s at two different RVP fuels are:

                        M4.1  Diurnal  (G/Test)
           MYR Group    9.0 RVP    11.5 RVP

           Pre-1979     42.33      77.89
           1979-80        6.05      16.93
           1981+ Garb     3.34      13.35


2.1.6  HDGV Diurnal Emission  Rates

     The MOBILE4 FTP diurnal emissions  in g/test  for  HDGVs  at two
different RVP fuels were:

                        M4 Diurnal (G/Test)
           MYR Group    9.0 RVP    11.5 RVP

           Pre-1985     42.33      77.89
           1985+          4.68      17.94

     The  same  algorithm  as MOBILE4 HOGV evaporative emissions were
used to  estimate model  year 1985+ HDGV diurnal  emission  levels for
pass  vehicles,  as  well  as  HDGVs  that  failed  either pressure  or
purge  test.   The  bases  used  in the calculations  were   the  1981+
LDGT1  diurnal  emission  rates   (i.e.,  2.09/11.45 g/test  for  pass
vehicles,  20.57/35.85  g/test  for  vehicles  that  failed  pressure
test, and 6.62/21.55  g/test for vehicles that failed purge test,  at
9.0  and  11.5 psi  RVP   fuels,   respectively).    In  general,   the
following equation is used to estimate HDGV diurnal emission rates:

                       *  1.5  * 0.875]
                        i  * 2.0 * 0.125]

where, the constant values of  1.5  and  2.0  were the  ratios  of the
evaporative  emission  standards  between  HDGVs  and  LDGTls.   The
constant values of 0.875  and  0.125 were  the  estimated  market shares
of HDGVs under each evaporative emissions standards.
                              -12-

-------
     The MOBILE4.1  FTP  g/test diurnal  emissions for  HDGVs  at two
different RVP fuels were:

                      FTP Diurnal Emissions  (g/test)
MYR Group

Pre-1985
1985+

     As  in  MOBILE4, the  FTP HDGV diurnal  emissions in  g/test for
any  other  RVP fuels  are  calculated by  interpolating  between  these
two  given  emission levels.   Assuming  that  in-use vehicles  were
tested  at  the age  of  four, the  MOBILE4.1  FTP diurnal emissions in
g/test for HDGVs at two different RVP  fuels  are:

                        M4.1 Diurnal (G/Test)
           MYR Group    9.0  RVP      11.5 RVP
at
Pass
42.33
3.27
9.0 RVP
Failed Failed
Purge Pressure
42.33 42.33
10.34 32.14

Pass
77.89
17.89
at 11.5
Failed
Purge
77.89
33.67
RVP
Failed
Pressure
77.89
56.02
           Pre-1985
             1985+
42.33
 5.22
77.89
20.86
2.1.7  Motorcycle Diurnal Emission Rates

     There were  no  new data available  for the motorcylce diurnal
emissions  since the  release of  MOBILE4 model.   Therefore,  the
MOBILE4.1  diurnal  emissions  for  motorcycles  are  the  same  as
those used for MOBILE4.
                              -13-

-------
2.2  Other Types of Diurnal Emissions

     Three   other   types   of   diurnal   emissions   used   in   the
MOBILE4/MOBILE4.1 models  are:   in-use full diurnal emissions  (FDI),
partial  diurnal emissions  (PDI),  and  multiple   diurnal  emissions
(MDI).   The  g/test  diurnal   emissions  discussed  in  the previous
sections  (2.1,  2.1.1  through  2.1.7) were all derived  from one  hour
heat build test results.

     Under  in-use  conditions,  however,  vehicles  may   experience
diurnal  emissions  from a much slower temperature  rise for  a  much
longer  time  frame,  for example, for  the entire period of six hours
when  the ambient  temperatures  are  increasing  from the  minimum to
the maximum.   Sometimes one full diurnal may be interrupted (or  cut
short) by the start of a  trip  during  the ambient  temperature  rising
period  so that  vehicles may  only experience a   partial diurnal.
Vehicles  may have  multiple diurnal  emissions  when  they have   not
being driven  for  two  or  more consecutive days.   Some vehicles  may
have  short  but frequent  trips  or  one long  trip  that  they  have no
diurnal  emissions  at  all during one  particular day.   The fractions
of  occurrences of  these various   types  of  diurnal   emissions   are
discussed  in  section  5.0.   The levels  of  these  various types of
diurnal emissions are discussed in the following sections.


2.2.1  In-Use G/Day Full Diurnal (FDI) Emissions

     In  MOBILE4,  the  in-use  (6-hour) g/day  full  diurnal  emissions
are the  FTP  one-hour  emission  rates  (g/test) adjusted for  121   and
124 percent at 9.0 and 11.5  psi RVP fuels,  respectively.

     There has  been very limited data available in this  area  since
the release of  MOBILE4 model.  The  adjustment  factors of 1.21   and
1.24  were derived  from  testing of  one  late model year  carbureted
vehicle with  its canister disconnected.   It  is unclear whether   the
differences  in emissions between  a slower  (six hours)  temperature
rise  and faster  (one  hour)   heat  build  should  be  classified  as
diurnal  or  resting loss  emissions,  as  the  levels  seem to  be more
similar to the  latter.  For these  reasons,  it  is   assumed that   the
21 (or 24) percent  emission increases due to the  slower temperature
rise  are  the  resting  loss  emissions  and the in-use   six-hour full
diurnal  emission  rates are the same as the one-hour FTP  diurnal
emissions.  Consequently,  the  MOBILE4  adjustment   factors of  1.21
and 1.24  used for  in-use full  diurnal emission calculations  are no
longer applicable in MOBILE4.1.
                              -14-

-------
2.2.2  Partial Diurnal (PDI) Emissions

     Many  of  the  MOBILE4  assumptions  used  for  partial  diurnals
remained unchanged for the MOBILE4.1 model, with one exception.

     By definition,  partial diurnals  occur  only when the vehicle's
engine has been  turned-off  for at least three  hours,  but less than
six hours, between the daily  minimum  and  maximum temperatures when
the  ambient  temperatures  are  increasing  (i.e.,  between  7 AM  and
3 PM).  The two  temperatures  used to  calculate the  partial diurnal
emissions are:

     1)      the ambient  temperature at  two  hours (rather  than  the
             one hour  assumed in MOBILE4)  after the  end of a trip,
             and,

     2)      the ambient  temperature at the time when  a  new trip
             starts.

The  three  types  of partial  diurnals and  their  characteristics  are
described below:

                        	Temperature	    Percent of
      Description

       8 AM - 11 AM
      10 AM -  3 PM
       8 AM -  2 PM

The  temperatures  at   10AM,  11AM,  and  2PM  are  calculated by  the
following equations:

      Tio AM  - Tmin + 0.53 * (Tmax - Tmin)              (2.5)

      Tii AM  = Tmin + 0.71 * (Tmax - Tmin)              (2.6)

      T2 PM - Tmin + 0.94 * (Tmax - Tmin)              (2.7)

where:
          Tmin - the minimum temperature of the day,  in °F
          Tmax - the maximum temperature of the day,  in °F.

The  ambient   temperature  at   3PM   is   just  the   maximum  ambient
temperature of the day.
Minimum
Tl o AM
Ti 2 PM
Tl o AM
Maximum
T 1 1 AM
Ta PM
Tj PM
Occurrence
32.36
7.09
3.69
                              -15-

-------
2.2.3  Multiple Diurnal  (MDI) Emissions

     In MOBILE4,  due to a lack of data, the  adjusted  in-use  diurnal
emissions  from tampered  vehicles were  used for  multiple  diurnal
emission  rates.   It  has been suggested that  this  assumption  may not
be  technically correct.   With  vehicles  parked  for  two  or  more
consecutive  days,  their diurnal  emission  levels are dependent  upon
many  factors:   ambient  temperature  rise,  backpurge   during   the
cooling   of   the    ambient   temperature,   canister   capacity   and
efficiency,  fuel  weathering, and  tank fill  level.   When all these
parameters  are  being considered,  the multiple diurnal emissions  may
be  lower  (on a  grams-per-day basis)  than  the  diurnal emissions
under uncontrolled or tampered conditions.

     Similar   to   the   FTP   one-hour  (or   in-use  full)   diurnal
emissions,  the MOBILE4.1 multiple   diurnal  emissions   model   also
includes  three elements:  emissions from  pass  vehicles, emissions
from  vehicles  that  failed  pressure   test,  and  emissions   from
vehicles  that failed purge  test.   Multiple  diurnal  emissions   for
vehicles  that  failed either  pressure  or  purge test  are  assumed to
be  the  same as  their   FTP  one-hour  diurnal counterparts.   These
emissions were based on both  the  theoretical  vapor generating model
and test data from EFP-Hammond (see discussions in section 2.1.2).

     Multiple  diurnal  emission rates  for  pass vehicles  were based
on  the  theoretical vapor generating  model.  Overall grams  per   day
emissions  modeled  from  late  model  year   fuel-injected  technology
vehicles  were simulated at  various  fuel  volatility  levels  for   one
to  seven  consecutive no-driving days.   The  temperature  rise of 72
to  96°F  and 40% tank fill level were assumed.  Also,  it  is   assumed
that backpurge  occures  in 50% of  the vehicles experiencing multiple
diurnals.

     To summarize  results, Table  2  from  Attachment has  been  copied
and presented at  the top  portion of Table 5.   Ratios  of emissions
from any  day (day  2  through  day 7)  to emissions  at day  1  for each
fuel  volatility  level   were calculated  and  listed  at  the lower
portion of Table 5.  These ratios, then, were weighted  according to
the  re-normalized  fractions  of occurrence for  days  2  through 7.
Finally,  a  nonlinear  regression   line  was  fitted  through   the
weighted  ratios.   The  equational  form for  the ratio  as  a function
of fuel volatility is:

     CFMOI.P.SS -  EXP (4.8491 - 0.33424 * RVP)                (2.8)

Both  the weighted  and  predicted  ratios  at   each fuel  volatility
level are given in  Table 5.   The  multiple  diurnal emissions   for
pass vehicles are  calculated  by:

     MDIpass - DIPass *  CFMDi.Pasi                         (2.9)
                             -16-

-------
     Using  equations   (2.8)  and   (2.9),   the   multiple  diurnal
emissions for  pass vehicles  are  sometimes calculated  to  be higher
than the emissions  from vehicles  that failed  purge,  especially for
pre-2.0  SHED   evaporative  emissions   standard   vehicles  (e.g.,
pre-1981  LDGVs).   One  of   the   reasons   is   that  the  emissions
calculated to  derive equation  (2.8)  were based  on late model year
fuel-injected   vehicle   performance,   while   all  pre-2.0   SHED
evaporative   emissions   standard   vehicles    are  of   carbureted
technology.    Also,  for  the  pre-2.0  SHED  evaporative  emissions
standard era,  the  emission levels for pass vehicles were similar to
the emission levels  for vehicles  that failed  purge.   Therefore,  it
is  reasonable  to  set  the  maximum  levels  of  multiple  diurnal
emissions for  pass vehicles  equal to the emissions  from vehicles
that failed purge, i.e.,

     MDIpass =MDIFalled Purge                           (2.10)

     To summarize, multiple  diurnal  emissions  for pass vehicles are
calculated by the following two methods:

     1)      use equations (2.8) and  (2.9), or,

     2)      use equation  (2.10),  if  the  calculated emission levels
             from  equations   (2.8)  and (2.9)   are greater than  the
             multiple  diurnal emissions  from  vehicles  that  failed
             purge.

These  two methods  are used for all LDGVs  and  LDGTs,  pre-1981  model
years  as well  as  1981+  model  years.   As  there is  no  multiple
diurnals  assumed  for  HDGVs  and  MCs, these  two  methods  are  not
applicable to these two vehicle types.
                              -17-

-------
3.0  Hot Soak Emissions

     As  in  MOBILE4,  the hot  soak emissions  (in  unit of  grams per
one  hour  test)  are  described  as  two  step  functions  of  fuel
volatility in psi RVP  and  ambient temperature in  °F.   The hot soak
emissions are measured at FTP  test conditions,  i.e.,  with 9.0 psi
RVP  fuel,  40% tank  fill,  and  at the  ambient temperature  of   82°F.
In  the  model,   hot  soak  emissions  are  first  adjusted  for  fuel
volatility  effect,  then corrected for  the  temperature  effect  at
ambient temperatures 41°F and higher.


3.1  1981+ LDGVS and LDGTS

     As  in  the  MOBILE4.1  diurnal  emissions  model,  the  g/test hot
soak  emissions   are  derived for pass vehicles,  for vehicles  that
failed  pressure test,  and  for vehicles  that failed  purge  test.
Table 3  also  summarizes  the average  hot soak  emissions  from the
current EFP-MVEL data base.
3.1.1  Pass Vehicles

     Hot soak  emissions  for pass  vehicles  were derived  from EPA's
EFP-MVEL data.  For model  years  1981+ pass  light-duty  vehicles  and
trucks, the  derived equational forms for g/test hot  soak emissions
are:
a + b * RVP
or,

     HS p a s s

where:

Vehicle Type
/Fuel System

   LDGV/CARB
   LDGV/PFI
   LDGV/TBI

   LDGT/CARB
   LDGT/FINJ
     d * RVP + e * RVP
                                                          (3.1)
                                                          (3.2)
                              Regression Coefficient
   0.14593   0.13823
   0.46673   0.10297
   0.198327  0.041297
0.16407
0.078327
0.13823
0.041297
-1.436657
18.437
-4.669710

-5.196600
-4.789710
                                     0.69740
                                     0.58219
                                                    0.0     0.034897
                                                   -4.2538  0.25072
                                                   0.58219  0.0
                                                            0.0
                                                            0.0
                             -18-

-------
     Equation  (3.1)   is  used  when  fuel  volatility  is   less  than
9.0 psi RVP,  while equation  (3.2)  is used  when  fuel volatility is
equal to or  high  than 9.0 psi RVP.   The  coefficients  from equation
(3.1)  were  regression  results  from  data  where  the  fuel  RVP  is
9.0 psi and  from  limited  emissions data  at lower  fuel  volatility
levels  of  6.5  and  8.0  psi  RVP.   The  coefficients  from equation
(3.2)  were  derived  from  EFP-MVEL  data,  where  the  three  fuel
volatility levels were 9.0, 10.5, and 11.7 psi RVP.

     The constant  terms  of equations  (3.1) and  (3.2)  (denoted as
"a"  and "c"  above)  were  adjusted  so  that  at  9.0 psi  RVP  the
predicted  hot soak  emissions are  equal  to  the  average  values  at
9.0 psi RVP  of  the EFP-MVEL sample.  Note  that the  lower  bound of
fuel  volatility for  equation (3.1)  is  5.0  psi RVP,  and  the lower
bound of hot soak emissions  is  set  to  equal  to  the resting  loss
emissions.

     Graphic  presentations of  1981+ LDGV/LDGT hot  soak  emissions
for  pass vehicles  as  a  function  of  fuel volatility  are  shown  in
Figures 5  and 6.   There  has  been  no  new data  available  to develop
new  hot   soak  temperature  correction   factors.    Therefore,   the
equations  used  to  correct  for temperature effect  from MOBILE4  will
be used again in MOBILE4.1  for pass vehicles.


3.1.2  Failed Vehicles

     During  FY91,  five vehicles were tested for hot  soak emissions
with  canister  connected  and  disconnected  at  two  levels  of  fuel
volatilities  (9.0  and 10.5 psi RVP) and  at  three  levels  of ambient
temperatures  (70,  82, and 95°F).   These five vehicles were:

         Model Year       Make           Fuel System

            1979          Cutlass        Carbureted
            1983          Cutlass        Carbureted
            1983          Skylark           TBI
            1990          Taurus            PFI
            1990          Lumina            PFI

     Overall  average  hot  soak emissions with canister  disconnected
(or   uncontrolled   hot    soak  emissions)   at   various   ambient
temperatures and fuel volatilities are summarized below:

          Temperature           Hot Soak (G/Test)
            in °F            9.0 RVP        10.5 RVP

              70.0            5.62             10.71
              82.0            7.93             16.11
              95.0           20.08             39.27
                              -19-

-------
     Regression   equation  has   been  derived   to   describe   the
relationships  of  the  uncontrolled hot  soak emissions,  versus  the
fuel volatility, and ambient temperature:
     HSU
        ncontrolled
   EXP  [2.2307 + 0.4443 *  (RVP  -  9.0)
   + 0.05114 * (TEMP - 82.0)]
where:
         HS = uncontrolled hot soak emissions in g/test,
        EXP » exponential function,
        RVP = fuel volatility in psi, and
       TEMP = ambient temperatures in °F.

Due  to  limited  available  data,  it  is  assumed that  the estimated
uncontrolled  hot  soak  emissions   are   the   emission  levels  for
vehicles that failed pressure test.

     From EFP-Hammond  data  base,  18  of  the 30 LDGVs  (as listed in
Table 4) were  also tested for  hot soak  emissions  at  92°F  ambient
temperature, and  with  9.0 psi RVP fuel.  These vehicles were found
to have failed purge test at the Hammond  I/M Lane.   The average hot
soak  emissions  were  10.56  g/test.   Therefore,  for  vehicles  that
failed purge test, the hot soak emissions can be  estimated from the
above failed pressure  hot soak emission  equation  but  normalized at
the average emissions from Hammond data.

     The hot soak emissions equations for failed vehicles are:

     HSFaile
-------
     Note  that  these  equations  can  be  extrapolated  beyond  the
ambient temperature  ranges of  70.0  and 95.0°F  and fuel volatility
levels of 9.0  and  10.5 psi RVP.  TO be consistent with the hot soak
emissions calculations for pass vehicles,  lower  limits are also set
at 5.0 psi RVP  fuel  volatility and 41°F ambient temperature.  There
are also  a  lower bound  of hot  soak emissions  being equal  to  the
resting  loss  emissions  and  an upper bound of 40.00  g/test  of  hot
soak emissions for LDGVs.

     Equations  (3.3)  and (3.4)  are also used  to calculate  hot soak
emissions for  all  other vehicle types  (such  as  LDGTls,  LDGT2s,  and
HDGVs) that failed pressure or purge test.

     Graphic presentations of 1981+ LDGV/LDGT  hot  soak emissions as
a  function  of  fuel  volatility at  82°F  for  vehicles  that  failed
either pressure or purge test are shown in Figure 7.
3.1.3  Pre-1981 LDGV Hot Soak Emission Rates

     The  same   list  of  methodologies  used   to   derive  diurnal
emissions, as described in section 2.1.3, is also used  for hot soak
emissions  calculations.   The  MOBILE4  FTP  hot  soak emissions  in
g/test for pre-1981 LDGVs at two different RVP fuels were:
           MYR Group

           Pre-1971
             1971
           1972-77
           1978-80
M4 Hot Soak (G/Test)
9.0 RVP      11.5 RVP
14.67
10.91
 8.27
 2.46
22.45
16.15
12.32
 4.30
The derived MOBILE4.1  LDGV FTP hot soak  emissions  for the  two  RVP
fuels are:

                       FTP Hot Soak Emissions (g/test)
MYR Group

Pass
at 9.0
Failed
Purge
RVP
Failed
Pressure

Pass
at 11.5
Failed
Purge
RVP
Failed
Pressure
Pre-1971
1971
1972-77
1978-80
1981+ Garb
14.671
10. 433
7.773
1.793
1.392
                      14.671
                      14.671
                       9.98s
                       6.332
                       6.332
14. 671
14.671
14. 67"
9.312
9.312
28.261
14. 613
11.10s
3.89s
3.18Z
28.261
28.261
19. 232
19. 232
19. 232
28. 261
28. 261
28.26'
28. 262
28. 262
                              -21-

-------
Note  that  the 1981+ carbureted hot soak  rates  were calculated  from
equations   (3.2)   through   (3.4).    They  were  listed  above   for
comparison  purpose only, as  all pre-1981  LDGVs are  of carbureted
fuel  delivery system.   As   in  MOBILE4,   the  pre-1981 FTP  LDGV  hot
soak  emissions  in g/test for any other RVP  fuels are calculated  by
interpolating between these  two given emission  levels.

      Using  the  assumption that  in-use  vehicles  were  tested at  the
age  of  four, the  MOBILE4.1  FTP  hot  soak  emissions  in  g/test  for
LDGVs at two different RVP fuels are:
           MYR Group

           Pre-1971
             1971
           1972-77
           1978-80
           1981+ Garb
           M4.1 Hot Soak (G/Test)
           9.0 RVP      11.5 RVP
           14.67
           10.91
            8.27
            2.46
            1.91
               28.26
               16.15
               12.50
                6.10
                5-47
3.1.4  Pre-1981 LDGT1 Hot Soak Emission Rates

     As  in  MOBILE4,  the  pre-1981  LDGV hot  soak  emission rates are
used also for  pre-1981  LDGTls because of their  similar evaporative
emission standards.   The MOBILE4  FTP  hot soak emissions  in g/test
for pre-1981 LDGTls at two different RVP fuels were:
           MYR Group

           Pre-1971
             1971
           1972-77
           1978-80
           M4 Hot Soak (G/Test)
           9.0 RVP    11.5 RVP
           14.67
           10.91
            8.27
            2.46
               22-45
               16.15
               12.32
                4.30
The derived MOBILE4.1  LDGT1  FTP hot soak emissions for  the  two  RVP
fuels are:

                       FTP Hot Soak Emissions (g/test)
MYR Group

Pre-1971
  1971
1972-77
1978-80
1981+ Garb
at 9.0 RVP
Pass
Failed
Purge
Failed
Pressure
at 11.5 RVP
Pass
Failed
Purge
Failed
Pressure
14.67
10.43
 7.77
 1.79
 1.08
14.67
14.67
 9.98
 6.33
 6.33
14.67
14.67
14.67
 9.31
 9.31
28.26
14.61
11.10
 3.89
 2.82
28.26
28.26
19.23
19.23
19.23
28.26
28.26
28.26
28.26
28.26
                             -22-

-------
Note that  the  1981+ carbureted hot  soak  rates  were calculated  from
equations   (3.2)   through   (3.4).    They  were  listed  above   for
comparison  purpose  only, as  all  pre-1981 LDGTls  are of carbureted
fuel delivery  system.   As   in  MOBILE4,  the pre-1981  FTP LDGT1  hot
soak emissions  in  g/test for any other RVP  fuels  are calculated by
interpolating between these  two given emission  levels.

     Assuming  that  in-use vehicles  were  tested at  the age of four,
the MOBILE4.1  FTP hot  soak  emissions in  g/test for LDGTls  at   two
different RVP fuels are:
              MYR Group

              Pre-1971
                1971
              1972-77
              1978-80
              1981+ Carb
            M4.1 Hot Soak (G/Test)
            9.0 RVP     11.5 RVP
            14.67
            10.91
             8.27
             2.46
             1.83
        28.26
        16.15
        12.50
         6.10
         5.15
3.1.5  Pre-1981 LDGT2 Hot Soak Emission Rates

     As  in MOBILE4,  the  same  post-1979  LDGT1  hot  soak  emission
rates  are  assumed for  LDGT2s,  based  on  their  similar  evaporative
emission  standards.   The  pre-1979  LDGT2  hot  soak emission  rates
were the same  as  the pre-controlled HDGV  hot  soak rates,  as  there
were no  separate  vehicle category  of  LDGT2s before the model year
of 1979.  The MOBILE4 FTP hot soak  emissions in  g/test  for  pre-1981
LDGT2s at two different RVP fuels were:
              MYR Group

              Pre-1979
              1979-80
            M4 Hot Soak (G/Test)
            9.0 RVP     11.5 RVP
            18.08
             2.46
        27.66
         4.30
The derived MOBILE4.1  LDGT2  FTP hot soak emissions for  the  two RVP
fuels are:

                      FTP Hot Soak Emissions (q/test)
MYR Group

Pre-1979
1979-80
1981+ Carb
at 9 . 0 RVP
Pass
Failed
Purge
Failed
Pressure
at 11.5
Failed
Pass Purge
RVP
Failed
Pressure
18.08  18.08
 1.79   6.33
 1.08   6.33
18.08
 9.31
 9.31
44.16
 3.89
 2.82
44.16
19.23
19.23
44.16
28.26
28.26
                             -23-

-------
As  in  MOBILE4,  the pre-1981 FTP LDGT2  hot  soak emissions in  g/test
for  any other  RVP fuels  are  calculated  by  interpolating  between
these two given emission levels.

     Assuming that  in-use vehicles were tested at the age of  four,
the  MOBILE4.1  FTP hot  soak  emissions  in g/test for  LDGT2s  at  two
different RVP fuels are:

                           M4.1 Hot Soak (g/test)
              MYR Group    9.0 RVP      11.5 RVP

              Pre-1979     18.08        44.16
              1979-80       2.46        6.10
              1981+ Garb    1.83        5.15


3.1.6  HDGV Hot Soak Emission Rates

     The MOBILE4  FTP  hot  soak emissions in g/test  for HDGVs  at two
different RVP fuels were:

                           M4 Hot Soak  (G/Test)
              MYR Group    9.0 RVP      11.5 RVP

              Pre-1985     18.08       27.66
              1985+         2.12        4.77

     As  in  the  diurnal emissions  calculations,  the  same  algorithm
from MOBILE4  HDGV evaporative emissions are used  to estimate  1985+
hot  soak emission rates for pass  vehicles, as  well as HDGVs  that
failed  either  pressure or  purge  test.   The  bases  used  in  the
calculations  were   the  1981+  LDGT1   hot  soak   emission   rates
(1.08/2.82 g/test for  pass vehicles,  9.31/28.26 g/test for vehicles
that failed pressure  test, and 6.33/19.23  g/test  for  vehicles that
failed purge test, at 9.0  and 11.5 psi RVP fuels, respectively).

     The new  MOBILE4.1 FTP g/test  hot  soak emissions  for HDGVs at
two different RVP fuels are:

               	FTP   Hot  Soak Emissions (g/test)
                	at

   MYR Group
   Pre-1985
   1985 +
at 9.0 RVP
Failed Failed
Pass Purge Pressure
18.08 18.08 18.08
1.69 9.89 14.55
at 11.5 RVP
Failed Failed
Pass Purge Pressure
44.16 44.16 44.16
4.41 30.05 44.16
                             -24-

-------
     As in MOBILE4,  the  FTP HDGV hot  soak emissions in  g/test  for
any other  RVP  fuels are  calculated by interpolating between these
two given emission levels.

     Assuming that  in-use  vehicles  were tested at the age  of four,
the MOBILE4.1  FTP  hot  soak  emissions  in  g/test  for HDGVs  at  two
different RVP fuels are:

                           Hot Soak (G/Test)
              MYR Group    9.0 RVP    11.5 RVP

              Pre-1985     18.08      44.16
                1985+       2.86       8.06


3.1.7  Motorcycle Hot Soak Emission Rates

     There  were no  new  data available  for  motorcylces  hot  soak
emission rates  since  the release of MOBILE4 model.   Therefore,  the
MOBILE4.1  hot  soak emissions for motorcycles  are  the same  as those
used for MOBILE4.
                              -25-

-------
4.0  High Altitude
     In  MOBILE4  model,  a set  of altitude  adjustment  factors was
used  to  calculate  the evaporative  diurnal and  hot soak  emissions
for  vehicles  operating  in  high altitude  region.   These  altitude
adjustment  factors  were  based  on either  actual  test  data or  from
their  emissions  standard  ratio.   For  example,  the  MOBILE4  FTP
diurnal  and hot  soak emissions  in g/test for high altitude LDGVs  at
two different RVP fuels were:
           MYR Group

           Pre-1971
            1971
           1972-76
            1977
           1978-80
     Diurnal (G/Test)
     9.0 RVP  11.5 RVP
                    Hot Soak  (G/Test)
                    9.0 RVP   11.5 RVP
     33.90
     21.16
     17.15
      8.98
     13.36
          62.39
          50.15
          44.93
          23.53
          36.85
            19.07
            14.18
            17.15
             8.27
             6.37
           29.18
           20.99
           20.96
           12.32
           11.15
Note  that  the  emission  rates  for  model years  1972-76  were actual
test results from EPA's EFP test  facility  in Denver,  Colorado.   The
emissions  for  model  year  1977  were  the  same as  the  low altitude
rates (because  of  their  same  evaporative  emissions  standard).   The
emission   rates  for  model   years   1978-80   were  estimated   by
multiplying a factor  of  2.59  to  the  low  altitude  rates.   Finally,
the emission  rates for all other model  year groups were calculated
by multiplying an adjustment factor of 1.30.

     The same set  of  altitude  adjustment  factors is used  again in
MOBILE4.1  and  are  summarized  in Table  6.  The  only exception is
that,  for  model years  1978-80,  the  MOBILE4  adjustment factor  of
2.59  has  been  changed  to 1.30.   These  high altitude  adjustment
factors   are  used for both  diurnal  and  hot  soak emissions  of  pass
vehicles,  as well  as  vehicles that failed  either pressure  or  purge
test.    For  example,   the  derived  MOBILE4.1  high  altitude  LDGV
diurnal  emissions at two RVP fuel levels are:

                   	LDGV  Diurnal  Emissions (G/Test)	
                       at 9.0 RVP
                               at 11.5 RVP
   MYR Group

   Pre-1971
     1971
   1972-76
     1977
   1978-80
   1981-83 Garb
   1984+ Garb
Pass
Failed
 Purge
33.90
19.54
10.16
8.27
5.71
2.16
1.66
33.90
33.90
13.82
9.33
7.36
7.36
5.66
 Failed
Pressure

  33.90
  33.90
  27.46
  20.39
  22.85
  22.85
  17.58
     Failed
Pass  Purge
 Failed
Pressure
62.39
48.59
33.13
22.80
17.21
9.62
7.40
62.39
62.39
45.06
23.83
23.95
23.95
18.42
62.39
62.39
56.15
35.45
39.83
39.83
30.64
                             -26-

-------
   Similarly,  the  derived  MOBILE4.1  high  altitude  LDGV
emissions at two RVP fuel levels are:

                        LDGV Hot Soak  Emissions  (G/Test)
              hot  soak
at 9.
0 RVP
Failed
MYR Group
Pre-1971
1971
1972-76
1977
1978-80
1981-83 Garb
1984+ Carb
Pass
19.
13.
10.
7.
2.
1.
1.
07
56
03
77
33
81
39
Pur
19
19
14
9
8
8
6
Fai
led
ge Pressure
.07
.07
.52
.98
.23
.23
.33
19
19
19
14
12
12
9
.07
.07
.07
.67
.10
.10
.31
at 11
.5 RVP
Failed
Pass
36.
18.
18.
11.
5.
4.
3.
74
99
87
10
06
13
18
Pur
36
36
27
19
25
25
19
ge
.74
.74
.30
.23
.00
.00
.23
Failed
Pressure
36.74
36.74
36.74
28.26
36.74
36.74
28.26
     Assuming that  in-use  vehicles were tested at the  age  of four,
the MOBILE4.1 FTP  diurnal  and hot soak emissions in g/test for high
altitude LDGVs at two different RVP fuels are:
                          Diurnal (G/Test)
MYR Group
Pre-1971
1971
1972-76
1977
1978-80
1981-83 Carb
1984+ Carb
9.0 RVP
33.90
21.16
11.30
8.98
6.72
3.57
2.74
11.5 RVP
62.39
50.15
35.07
23.53
18.81
12.08
9.29
Hot Soak (G/Test)
9.0 RVP  11.5 RVP
                                              19.07
                                              14.18
                                              10.91
                                               8.27
                                               3.20
                                               2.74
                                               2.11
           36.74
           20.99
           20.32
           12.50
            7.94
            7.11
            5.47
                              -27-

-------
5.0  Trip Related Estimates

     Most of  the MOBILE4 trip-related  estimates,  such as trips per
day (TPD), miles  per day  (MPD),  fractions of  occurrences for FDI,
MDI, and  PDI  are also  used in MOBILE4.1,  with some modifications.
There is no new data available  to  derive new estimates.   Therefore,
the data  base where these  estimates were  derived  from is the same
1979 GM-NPD survey data used in MOBILE4.


5.1  Trips Per Day and Miles Per Day

     The  trips  per  day and miles  per  day  equations  used for LDGVs
and LDGTs have the following forms:

             TPD - 4.7187 - 0.058508 * AGE               (5.1)
             MPD - JAMAR / 365.0                         (5.2)

where:   TPD = estimated number of trips per day,
        AGE - vehicle age in years,
        MPD = estimated miles driven per day,
      JAMAR » January 1 annual mileage accumulation rate, which is
              also a function of vehicle age.

These  equations  were  used  in  MOBILE4  to  calculate  the  estimated
trips per day and miles per day for vehicles at ages  1 through 20.
In MOBILE4.1, equations  (5.1)  and  (5.2) are used to  obtain  TPD and
MPD estimates for LDGVs and LDGTs for ages 1 through 25.

     The TPD  and  MPD estimates for  HDGVs  and MCs  remain  unchanged
in MOBILE4.1:

                     HDGVs    MCs
             TPD      6.88    1.35
             MPD     33.97   10.02
                             -28-

-------
5.2  Other Trip-Related Estimates — LDGVS and LDGTS

     Percents  of  other  trip-related  estimates for  LDGVs  and LDGTs
used in MOBILE4 were:

                                                           Total
        Description                            Percent     Percent

     Driving Days                              76.15       76.15

         Single Full Diurnal Days              22.37
         Partial Diurnal Days                  43.14
               i) 8 AM to 11 AM      32.36%
              ii) 10 AM to 3 PM       7.09%
             iii) 8 AM to 2 PM        3.69%
         3+ multiple Diurnal Days               3.89
         No Diurnal Days                        6.75

     No Driving Days                           23.85       23.85

         Single Full Diurnal Days              11.59
         2+ Multiple Diurnal Days              12.26

     In MOBILE4.1,  all the above  estimates  are expressed  in terms
of  vehicle's  age.   For example, the  proportions  of vehicles  having
at  least one  trip  on  any  given day  for  age 1 vehicles are  higher
than  those  for age  2  vehicles,   etc.   The  equational  forms  for
various types of driving days (in percent) are:

     Driving Days - 87.07 - 1.72 * AGE                   (5.3)
     Single Full Diurnal = 25.58 - 0.5 * AGE             (5.4)

     Partial Diurnal Days:

     8AM-11AM Partial Diurnal - 37.0 - 0.73 * AGE        (5.5)
     10AM-3PM Partial Diurnal = 8.11 - 0.16 * AGE        (5.6)
     8AM-2PM Partial Diurnal - 4.22 - 0.08 * AGE         (5.7)

     3+ Multiple Diurnal = 4.45 - 0.09 * AGE             (5.8)
     No Diurnal Days - 7.72 - 0.15 * AGE

The equational forms for various types of  driving  days  (in  percent)
are:

     No Driving Days - 12.93 + 1.72 * AGE
     Single Full Diurnal Days - 6.31 + 0.84 * AGE        (5.9)
     2+ Multiple Diurnal Days - 6.62 + 0.88 * AGE       (5.10)

Table 7 summarizes  the results.
                              -29-

-------
5.5  Other Trip-Related Estimates for HDGVs

     Percents  of  all  trip-related  estimates  for  HDGVs  used  in
MOBILE4 were:

                                                           Total
        Description                            Percent     Percent

     Driving Days                              27.10       27.10

     No Driving Days                           72.90       72.90

         Single Full Diurnal Days              49.80
         2 Multiple Diurnal Days               13.70
         3+ Multiple Diurnal Days               9.40

     Due to  a  lack of data, no equations as  functions  of  vehicle's
age were derived.  These estimates are used in MOBILE4.1 again.
                             -30-

-------
                         Table 1

              Pressure/Purge  Failure Rates
                     by Vehicle Age
             Source:   EPA's EFP-Hammond Data
 Age of          	LDGV Failure  Rates	
Vehicle           Purge          Pressure          Either

   01            0.043              0.043           0.080
   02            0.043              0.043           0.080
   03            0.043              0.043           0.080
   04            0.060              0.053           0.096
   05            0.060              0.053           0.103

   06            0.069              0.053           0.120
   07            0.094              0.053           0.150
   08            0.094              0.106           0.188
   09            0.142              0.152           0.258
   10            0.214              0.171           0.323

   11            0.224              0.236           0.389
   12            0.229              0.336           0-442
   13+           0.229              0.336           0.451
                            -31-

-------
                          Table 2

    Criteria  Used to  Define Pressure and Purge Failures
                 Emission Factors Programs
Evaporative System Component          Diagnosis	

                      Purge  Failures

Canister Purge Solenoid/Valve     Missing
                                  Disconnected or Bypassed
                                  Leaks Vacuum
                                  Sticking
                                  Inoperative

Vacuum or Vent Lines              Disconnected or Missing
                                  Plugged or Damaged
                                  Misrouted

Purge Hose                        disconnected or Missing
                                  Split or Not sealed

Canister Purge TVS                Stuck
ECM signal to purge solenoid      None detected

                    Pressure Failures

Gas Cap                           Non OEM or Missing
                                  Leaking

Sending Unit Gasket               Leaking
Fuel Tank Rollover                Valve leaking or Stuck

Fuel Tank Filler Neck             Leaking

            Both  Pressure and Purge  Failures

Canister                          Missing
Carburetor Bowl Vent Line         Disconnected or Leaks
Fuel Line to Canister             Disconnected
EFE Control Switch                Missing
Bowl Vent Solenoid                Disconnected
                        -32-

-------
                             Table  3
                                               Date:   April  26,  1991
               Statistics on Evaporative  Emissions
              (Average G/Test Hot  Soak  and  Diurnal)
                     Source:  EPA's  EFP-MVEL
 Fuel
 Delivery
 System
All
 Carbureted
 PFI
 TBI

Failed Purge
 Carbureted
 PFI
 TBI
    9.0 RVP
               10.4 RVP
                           11.7 RVP
 N
HS
DI
N
HS
DI
N
HS
DI
               Light-Duty Gasoline-Powered Vehicles
386  1.99
347  0.70
315  0.82
 Failed Pressure
  Carbureted      37
  PFI             12
  TBI              9

 (FPurge,FPressure)
  All            115
 Tampered
  Carbureted
  PFI
  TBI

 Pass
  Carbureted
  PFI
  TBI

 Others*
  Carbureted
  PFI
  TBI

 (Pass,Others)
  Carbureted
  PFI
  TBI
     2.65
     1.90
     1.75
     158
      69
      95
    3-43
    2.09
    1.42
     5.52
     4.80
     3.90
 31  5.32  5.11
 13  4.27  6.93
 13  3.26  7.19
     3.15  7.44
     3.22 10.84
     2.77  9.21
            17  7.14  8.76
             2 16.21  7.70
             2  3.27 13.52
            17  6.14  9.76
             3  9.16  9.36
             2  1.25 11.77
     3.85  7.22   43  6.86  9.51
  7  7.77 10.28
  1  0.29 13.79
  2 11.49 10.52
258  1.21
299  0.41
233  0.47
     1.38
     1.03
     0.93
             2 16.00 22.70
             fl-
             o-
      gs  1.89  3.74
      61  1.36  3.66
      79  0.85  3.32
53
22
58
2
1
1
.27
.19
.01
3
5
2
.02
.42
.38
27
3
12
3
0
4
.88
.51
.91
5
21
4
.80
.57
.77
311
321
291
1.39
0.46
0.57
1.66
1.33
1.22
122
64
91
2.33
1.32
1.39
4.20
4.50
3.51
       380
       254
       274
    4.17  9.59
    3.84  6.85
    2.99  8.91
                         30 10.25 15.19
                          9 21.07 12.63
                         10 13.69 21.71
                         37  5.10 15.93
                          7 10.96 25.27
                          7 13.92 24.42
                              100  9.97 17.24
                          7  9.05 19.36
                          1  0.20 21.44
                          1 36.43 28.28
                  253  3.25  7.25
                  219  2.97  5.64
                  206  1.77  7.68
                                     53  3.79  11.85
                                     18  3.25  10.78
                                     50  3.71   8.86
                                    306  3.34   8.05
                                    237  2.99   6.03
                                    256  2.15   7.91
*A11 vehicles that had some non-zero ECOMP codes but were not
 defined as either purge, failure, pressure failure, or tampered.
                              -33-

-------
                        Table 3  (Continued)
                                             Date:  April  26,  1991

                Statistics on Evaporative Emissions
               (Average G/Test Hot Soak and Diurnal)
                      Source:  EPA's EFP-MVEL

  Fuel
  Delivery          9.Q RVP            10.4 RVP	11.7 RVP
  System         N    HS   PI      N    HS    PI    N    HS    PI

                 Light-Puty Gasoline-Powered Trucks

 All
  Carbureted    170  1.53  2.93     6  5.35 10.01  199  4.13 14.02
  Fuel-Injected  80  0.58  1.90     0   -     -     42  1.32   8.64

 Failed Purge
  Carbureted      7  5.06  9.44     1 21.68 25.47   12 10.20 17.36
  Fuel-Injected   1  0.19  0-62     0   -     -      0   -

 Failed Pressure
  Carbureted     10  2.89  8.71     0         -     11  7.23 23.65
  Fuel-Injected   2  0.40 22.42     0                1  6.81 54.35

 (FPurge,FPres)
  All            20  3.27  9.93     1 21.68 25.47   24  8.70 21.79

 Tampered
  Carbureted      4  8.94  8.41     0   -     -      6 19.47 32.06
  Fuel-Injected   1 11.10 23.32     0                1 26.94 53.62

 Pass
  Carbureted    129  1.00  1.69     4  2.31  3.63  128  2.19 10.54
  Fuel-Injected  69  0.43  0.95     0               37  0.36  6.75

 Others*
  Carbureted     20  1.59  4.68     1  1.17 20.07   42  5.31 18.58
  Fuel-Injected   7  0.62  2.50     0   -  '   -      3  2.83  1.70

 (Pass,Others)
  Carbureted    149  1.08  2.09     5  2.08  6.92  170  2.96 12.52
  Fuel-Injected  76  0.45  1.09     0   -     -     40  0.55  6.37
*A11 vehicles that had some non-zero ECOMP codes but were not
 defined as either purge,  failure,  pressure failure, or tampered.
                             -34-

-------
                 Table 4

   Failed Pressure/Purge Test Results
      92°F Hot Soak,  9 psi  RVP Fuel
72-96°F Diurnal Heat  Build,  40% Tank Fill
        Source: EPA's EFP-Hammond

Vehicle #
576
640
660
665
724
443
736
1541
1635
1640
1530
1552
722
1542
740
1524
673
1526
1612
1532
743
1537
1544
1529
659
1533
1548
1555
730
1525



Fuel
Failed
Model Year System Hot Soak
1981
1981
1981
1981
1982
1983
1983
1983
1984
1984
1984
1984
1985
1985
1985
1985
1985
1985
1985
1986
1986
1986
1986
1987
1988
1988
1988
1988
1989
1989


Garb
PFI
Garb
Garb
Garb
TBI
Garb
TBI
PFI
PFI
TBI
TBI
PFI
PFI
Garb
Garb
TBI
TBI
TBI
PFI
TBI
TBI
TBI
PFI
PFI
PFI
PFI
TBI
PFI
PFI
N:
Average:
1.95

15.96


6.17


12.05
1.62

3.70



26.08
1.73
9.26
5.54
33.30


25.37
0.96
2.77
0.52
20.97
7.36

14.71
18
10.557
Purge Failed Pressure
Diurnal
13.32

11.93


2.68


13.36
20.08

21.92



25.55
19.30
22.85
13.01
22.22


14.40
2.80
21.34
0.65
34.61
15.44

4.90
18
15.576
Diurnal

76.22

77.92
24.76

17.55
10.85


17.01

16.38
11.38
14.47





5.41
33.45






27.37

12
27.727
                 -35-

-------
                              Table 5

                     Multiple Diurnal  Emissions
                           Pass Vehicles
                   Source:   Table  2  of Attachment
    Fractions
        of
Day Occurrence
1
2
3
4
5
6
7
0.341
0.078
0.037
0.021
0.014
0.008
0.003
 2
 3
 4
 5
 6
 7
0.484
0.230
0.130
0.087
0.050
0.019
7.5
RVP
8.0 8.5 8.7 9.0
RVP RVP RVP RVP
10.0
RVP
11.0
RVP
Calculated Emissions (in Grams)
0.2
1.6
2.6
3.3
3.8
4.3
4.7
0.4 0.7 0.8 1.1
2.4 3.4 3.8 4.7
3.6 5.0 5.7 6.8
4.5 6.2 6.9 8.2
5.3 7.0 7.9 9.2
5.8 7.7 8.6 10.0
6.3 8.2 9.0 10.5
2.5
8.7
11.9
13.8
15.0
15.8
16.4
5.3
15.5
20.0
22.3
23.6
24.3
24.7
Calculated Emission Ratios
(Relative to Day 1 Emissions)
8.0
13.0
16.5
19.0
21.5
23.5
6.0 4.9 4.8 4.3
9.0 7.1 7.1 6.2
11.3 8.9 8.6 7.5
13.3 10.0 9.9 8.4
14.5 11.0 10.8 9.1
15.8 11.7 11.3 9.5
3.5
4.8
5.5
6.0
6.3
6.6
2.9
3.8
4.2
4.5
4.6
4.7
                     Occurrence Weighted Ratio

                  12.2     8.6    6.8    6.7    5.8
                                                 4.5
              3.5
                  10.4
                   Predicted Ratio*

                     8.8    7.4    7.0
6.3
4.5
3.2
*Predicted ratios are calculated from equation:
         Ratio - EXP (4.8491 - 0.33424 * RVP)
                             -36-

-------
                             Table  6

                 High Altitude Adjustment Factors
                      Evaporative Emissions
           Vehicle
            Type

           LDGVs
            LDGTlS
Model Year
  Group

Pre-1972
1972-76
1977
1978-81
1982-83
1984 +

Pre-1972
1972-76
1977
1978-81
1982+
Adjustment
  Factor

    1.30
     *
    1.00
    1.30
    1.30
    1.00

    1.30
     *
    1.00
    1.30
    1.30
            LDGT2S

            HDGVs

            MCS
All

All

All
    1.30

    1.30

    1.30
* Based on actual test data.
                             -37,-

-------
              TABLE 7

Trip-Related Estimates (in Percent)
          LDGVs and LDGTs
  Source:   1979 GM-NPD  Survey  Data
Trip Days

Veh
Age
01
02
03
04
05
06
07
OS
09
10
11
12
13
14
15
16
17
IS
19
20
21
22
23
24
25 +

Trip
Days
85.35
83.63
81.91
80.19
78.47
76.75
75.03
73.31
71.59
69.87
69.87
68.15
66.43
64.71
62.99
61.27
59.55
57.83
56.11
54.39
50.95
49.23
47.51
45.79
44.07
Single
Full
DI
Days
25.08
24.58
24.08
23.58
23.08
22.58
22.08
21.58
21.08
20.58
20.08
19.58
19.08
18.58
18.08
17.58
17.08
16.58
16.08
15.58
15.08
14.58
14.08
13.58
13.08
Partial
Diurnal Days
SAM-
HAM
36.27
35.54
34.81
34.08
33.35
32.62
31.89
31.16
30.43
29.70
28.97
28.24
27.51
26.78
26.05
25.32
24.59
23.86
23.13
22.40
21.67
20.94
20.21
19.48
18.75
10AM
8AM
-3PM -2PM
7.95
7.79
7.63
7.47
7.31
7.15
6.99
6.83
6.67
6.51
6.35
6.19
6.03
5.87
5.71
5.55
5.39
5.23
5.07
4.91
4.75
4.59
4.43
4.27
4.11
4.14
4.06
3.98
3.90
3.82
3.74
3.66
3.58
3.50
3.42
3.34
3.26
3.18
3.10
3.02
2.94
2.86
2.78
2.70
2.62
2.54
2.46
2.38
2.30
2.22
3 +
MDI
Days
4.36
4.27
4.18
4.09
4.00
3.91
3.82
3.73
3.64
3.55
3.46
3.37
3.28
3.19
3.10
3.01
2.92
2.83
2.74
2.65
3.46
3.37
3.28
3.19
3.10
No
DI
Days
7.57
7.42
7.27
7.12
6.97
6.82
6.67
6.52
6.37
6.22
6.07
5.92
5.77
5.62
5.47
5.32
5.17
5.02
4.87
4.72
6.07
5.92
5.77
5.62
5.47
No
No
Trip
Da
14
16
18
19
21
23
24
26
28
30
31
33
35
37
38
40
42
43
45
47
31
33
35
37
38
ys
.65
.37
.09
.81
.53
.25
.97
.69
.41
.13
.85
.57
.29
.01
.73
.45
.17
.89
.61
.33
.85
.57
.29
.01
.73
Trip^ Days
Single
Full
DI
Days
7.15
7.99
8.83
9.67
10.51
11.35
12.19
13.03
13.87
14.71
15.55
16.39
17.23
18.07
18.91
19.75
20.59
21-43
22.27
23.11
15.55
16.39
17.23
18.07
18.91
2 +
MDI
Days
7.50
8.38
9.26
10.14
11.02
11.90
12.78
13.66
14.54
15.42
16.30
17.18
18.06
18.94
19.82
20.70
21.58
22.46
23.34
24.22
16.30
17.18
18.06
18.94
19.82
             -38-

-------
ATTACHMENT
 -39-

-------
                                                      Figure  1
                              1981+ Passing  Vehicles LDGV  Diurnal  Emissions  (g/test)
16 -r
         i    i"n—r~  i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i
   0.3  0.4  0.5  0.6  0.7  0.8  0.9   1    1.1  1.2   1.3  1.4   1.5  1.6   1.7  1.8  1.9   2   2.1  2.2  2.3  2.4  2.5  2.6  2.7
                                                         UDI
                                                10:34 AM 7/25/91

-------
                                                          Figure 2
                                1981+  Passing  Vehicles LDGT Diurnal  Emissions  (g/test)
20  -r
    0.3   0.4  0.5  0.6  0.7  0.8  0.9   1
1.1   1.2  1.3  1.4  1.5   1.6  1.7  1.8  1.9
                   UDI
2   2.1   2.2   2.3  2.4  2.5  2.6  2.7
                                                   10:38 AM 7/25/91

-------
                                                                       Figure  3
                                               1981+ Failed Vehicles LDGV Diurnal Emissions  (g/test)
              45 T
to
                   03  0.4  0.5  0.6  0.7  0.8  0.9   1    1.1  12  1-3  1.4   1.5  1.6  1.7  1.8  1.9   2   2.1   2.2  2.3  2.4  2.5  2.6  2.7
                                                                  10:31 AM 7/25/91

-------
                                                                       Figure 4
                                               1981+ Failed Vehicles  LDGT  Diurnal  Emissions  (g/test)
               50 T
GO
                                                  Failed
                                                 Pressure
                                                                                                      Failed
                                                                                                      Purge
                  0.3  0.4  0.5  0.6  0.7  0.8  0.9   1   1.1  1.2   1.3  1.4  1.5   1.6  1.7  1.8   1.9
2.1   2.2  2.3  2.4  2.5  2.6  2.7
                                                                8:38  AM  6/25/91

-------
                                                       Figure  5
                           1981+ Passing Vehicles at 82°F LDGV Hot Soak Emissions (g/test)
3.5 T
           5.5
6.5
7.5      8      8.5      9      9.5
   Fuel Volatility  (in psi  RVP)
10     10.5      11      11.5     12
                                                  9:11 AM 6/25/91

-------
                                                                    Figure 6
                                        1981+ Passing Vehicles at  82°F LDGT Hot  Soak Emissions  (g/test)
                3 T
en
            H
            o  2.5
            t
S
o
a
k

E
m
i
s
s
i
o
n
s
                2  -
               1.5
               0.5  -'
                0
                                                                FINJ
                                                 CAR
                          5.5
                              6.5
7.5      8      8.5      9      9.5
  Fuel  Volatility (in  psi  RVP)
10
10.5
11
11.5
12
                                                              10:10 AM 6/25/91

-------
                                          Figure 7
           1981+ Failed Vehicles at 82°F LDGV & LDGT Hot Soak Emissions  (g/test)
5.5
6.5
7.5     8      8.5      9      9.5
   Fuel Volatility (in  psi  RVP)
10
10.5
11
11.5
12
                                     9:14 AM 7/3/91

-------
                       RUNNING LOSS EMISSIONS


1.0  Introduction

     Running  loss emissions  test  program  has  been  performed and
data  collected  by  Automotive  Testing  Laboratories,  Inc.   (ATL)
located at South  Bend,  Indiana under EPA's Emission Factors Program
contract.

     The  test  program  was  designed  to  test  in-use  vehicles  with
three different driving cycles:

     1)    a  low  speed cycle  (known as  the New York City  Cycle
           [NYCC]) with an average speed of 7.1 mph,

     2)    a  Federal Testing  Procedure  (FTP)  LA-4   driving  cycle
           with an average speed of 19.6 mph, and,

     3)    a high  speed cycle (or  Highway Fuel Economy Test [HFETl)
           with an average speed of 47.9 mph.

The duration of the running loss test is  approximately one  hour for
each driving  cycle.   Therefore, the  NYC  driving  cycle  is  repeated
six times (6 bags), the two portions of the  LA-4  cycle are  repeated
three  times  (6 bags),  and the  HFET driving cycle  is  repeated five
times  (5  bags).   The  distances,  average  speeds,  and  durations  of
these three driving cycles are summarized in the following:

                          Distance   Speed     Duration
    Driving                  in       in          in
     Cycle       Bags      Miles      mph       Minutes
     NYCC        1 to 6     1.18       7.1        10.0
     LA-4        1,3,5      3.59      25.6         8.4
                 2,4,6      3.89      16.2        14.4

     HFET        1 to 5    10.20      47.9        12.8

Under  this  test  program,  many vehicles  were tested  only on  LA-4
cycles.  Fewer vehicles were tested on HFET cycle.

     The  running  loss  emissions  test  program  was  designed  to
collect  data  at  four  levels of  fuel  volatility  (7.0,  9.0,  10.4,
11.7 psi in  Reid Vapor Pressure  [RVP]) and three  levels of  ambient
temperature (80, 95,  and 105°F).


                                -47-

-------
     Not all vehicles were tested for  all  combinations of fuel RVPs
and ambient  temperatures,  however.   There  is  usually no testing at
extreme conditions,  such  as  the combinations  of high  RVP  fuel and
high  ambient  (11.7  psi/105°F),  and  low RVP  fuel  and  low ambient
(7.0 psi/80°F),  because  of  their   less  likely  occurrences  in the
real world.  Also,  if  from a test vehicle the running loss emission
results  are  low  (less  than  0.5  grams)   at   certain  fuel  and
temperature  combination  (for example,  9.0 psi/95°F),  it is assumed
that at  the combinations  of lower   fuel volatilities  and/or   lower
ambient   temperatures   (i.e.,   7.0 psi/95°F,   9.0 psi/80°F,   and,
7.0 psi/80°F),  this vehicle would have emissions  at a similarly low
level.   To  save  resources,  the  vehicle  is  not  tested  for the
combinations   of   lower   fuel   volatilities   and    lower   ambient
temperatures.  Further, there  has  been  no testings on 11.7 psi RVP
fuel shortly after the issuance of MOBILE4 in 1989.
                               -48-

-------
2.0  True Vapor Pressure
     In  MOBILE4  model, when the  test  data  is  not  available at
certain  combinations  of  fuel volatility  and  ambient  temperature,
the  g/mi running  loss  emissions  were estimated  from  a  variable
called "True  Vapor Pressure  (TVP)."   in MOBILE4.1  model,  this TVP
is used  to correlate with  the  running loss  emissions  from failed
vehicles.   These  TVPs  by  bag  are expressed  as functions  of   fuel
volatility  and fuel  tank  temperature, and  are calculated  by the
following five steps of equations:
U7

V7
     1.0223*RVP
     66.401 - 12.718*U7
     0.0018407*U7**4
                       [0.0357*RVP / (1.0 - 0.0368*RVP)]

                               1.3067*U7**2 - 0 . 077934*U7**3
     TB i  = Tmin + DTB
                                         for i=l,2,...,6
     CD7Bi = 262.0 / [(V7/6.0 + 560.0) - 0.0133] * (100.0 - TBi
           + V7
TVP
   Bi -
        14.697 - 0.5308*CD7Bi
        0.000055631*CD7Bi**3 + 1.769*10
                                     0 . 0077215*CD7B i **2
                                              *CD7B1**4
                                                  for i =
                                                   -1,2.
where:
       RVP
       TBi
      Tmin
      DTBi
                                              in °F,
     TVP
        B i  =
        fuel volatility in psi,
        cumulative tank temperature for bag i
        initial tank temperature in °F,
        cumulative tank temperature rise for bag i
        and,
        calculated bag i TVP value..
                                                   in °F,
The  TVP  values  were  calculated  for  all  combinations  of  fuel
volatilities   (7.0,   9.0,   10.4,   and  11.7 psi   RVP)   and   tank
temperature profiles  (with  the  initial  tank temperatures at 80,  87,
95, and 105°F).

     The tank temperature  rises  for  each bag were estimated  from  a
44-car tank  temperature  data collected  at  Laredo, Texas  during  the
time frames of  September  11 through  October 23, 1989, by ATL  under
EPA's contract.   In  this program, vehicles'  tank  temperatures  were
measured and  recorded while  being driven  under  the  LA-4  driving
cycle repeated  for three  times  (thus,   a total  of six  bags).   The
initial  tank  temperatures were  adjusted  to  be  within  +  2°F  of  the
ambient   temperatures  at   the   start   of   the   test.    An  ambient
temperature of  95°F  was  used.   The  fuel tanks  were  filled at  40%
level with  9.0 psi RVP fuel.
                               -49-

-------
     The  initial and  final  tank  temperatures  for  each  bag from
these  44  vehicles  were  examined and  an average  tank temperature
rise  for  each  bag  was  calculated.    The   resulting   average tank
temperature rises  for  the  six  bags are:  5,  9,  4, 5,  2, and 3°F,
respectively.   The  cumulative  tank temperature  rises (5,  14,  18,
23, 25, and 28°F),  then,  were  used in  the  above  five equations to
calculate  the corresponding TVP values.   Therefore,  the initial and
final tank temperatures for the four temperature profiles were:


 Initial Tank      	Final Tank Temperature in °F	
  Temp in  °F      Bag 1   Bag 2   Bag 3   Bag 4

   80.0
   87.0
   95.0
  105.0

Table  1  summarizes  the  calculated TVP  values for  each  bag  under
each   fuel  volatility  level  and  each  initial   ambient  (tank)
temperature.
85.0
92.0
100.0
110.0
94.0
101.0
109.0
119.0
98.0
105.0
113.0
123.0
103.0
110.0
118.0
128.0
105.0
112.0
120.0
130.0
108.0
115.0
123.0
133.0
                               -50-

-------
3.0  Running Loss Emissions Data

     Similar  to  EPA's  Hammond  program,  all  vehicles  under  the
running  loss  emissions  test  program  since  FY90 were  also checked
for pressure  and/or purge  failures  of their  evaporative emissions
control  system.   Descriptions  of  these  pressure/purge  tests were
given in the  Evaporative HC Emissions Chapter.  Failure  rates from
either pressure  and/or purge test  were described as  a function of
vehicle's age.  These  failure rates are the same as those  used for
the diurnal/hot soak emissions model.

     Therefore, the  MOBILE4.1 running  loss  emissions  model include
three  portions:    emissions   for  pass  vehicles,  emissions  for
vehicles that  failed pressure test,  and emissions for  vehicles that
failed purge test.

     For vehicles tested  prior  to  the Hammond program,  diagnoses
and comments  from mechanics  and  their running loss emissions test
results  at  9.0 psi/95°F  were examined to  see which  category  the
vehicles  should  be  grouped  into.   For example,  if diagnoses  and
comments from mechanics indicated  that there exists  a  leakage  on
one of the  evaporative emissions  control  system components  (such as
gas cap, filler neck, sending unit,  rollover valve,  or  vent hoses),
the  vehicle  is  categorized  as  "failed pressure  test".    If  there
exists  an  in-operative  canister   purge  solenoid  or   valve,   or
disconnected,  missing,  or  damaged  purge   hoses,  the  vehicle  is
grouped  as  "failed  purge   test."    If  there  were   no  problems
detected, the vehicle is categorized as "pass vehicle."

     Among  those  pre-Hammond  vehicles  that had problems detected in
their  evaporative  emissions  control  system  components   by  the
mechanics'   diagnoses  and comments,  some of  them had relatively low
running loss  emissions.   Among  those pre-Hammond  vehicles  that  had
no visible  problems  detected  in their evaporative  emissions control
system  components,   some  of  them  had  very  high   running   loss
emissions.    Due  to  this  lack of consistency,  it  was  decided that
all  LDGVs  tested  prior  to  Hammond  program  be  excluded   from
MOBILE4.1 model.   For  LDGTs,  however, the  sample  sizes were very
limited, their running  loss  emissions  test   results were  relatively
consistent,  with  no visible  problems  detected.  For these  reasons,
all LDGT data were used as pass vehicles for  the LDGT fleet.
                               -51-

-------
3.1  Exception Vehicles

     Among  the data  used  for  MOBILE4.1  analysis,  there  were  two
vehicles  with  "unusual"  test  results.   One  of   them  is  a  1986
Chevrolet  Camaro,  identified  as vehicle  #1532, which is  equipped
with  a  5.0L engine, ported fuel injection fuel delivery system,  by
engine  family  G1G5.0V8NTA8, and with  evaporative emissions  control
system  6BO-1A.  This  vehicle,  when  tested  at Hammond  I/M  lane,
passed  pressure check but failed  purge  flow  check.  The  other
vehicle  is  a  1986  Pontiac Sunbird,  identified as  vehicle  #1578,
which is  equipped  with a 1.6L engine,  throttle body fuel  injection
fuel  delivery  system,  by engine  family  G2G1.8V5TDG2,   and with
evaporative  emissions  control  system  6AO-2B.  This  vehicle was
found  to  have a  leaking  gas  cap.    It  failed pressure  check but
passed purge flow check.

     Two  things  were observed  on these  two  vehicles:   their tank
temperature  rises  were  extremely high,  and  their  emission  levels
were  equally  high  under  as-received  as  well  as   after   repair
conditions.  As  discussed  above  (in  section 2.0),  the average tank
temperature rise from an initial ambient of  95°F is  28°F.   The tank
temperature  rises  for these two vehicles, however, were  more than
40°F.   The cumulative  temperature  rises  from  the  six bags  were:
12,  22,  33, 40,  44,  47°F  for  the  Camaro,  and 9,  26, 32,  37, 41,
43°F for the Sunbird.  The  following table shows their running loss
emissions  in total  grams  (the  sum of  6 bags)  tested at 9.0 psi/95°F
from LA-4 driving cycle:

     Vehicle                                      Total Emissions
     Number      	Status	                (Grams)

     1532        as-received (failed purge)             315.5
                 after reconnect purge line             330.8
                 after replace gas cap                  319.9

     1578        as-received (failed pressure)          263.2
                 after repair                           325.5

     Note that these repair/maintenance  steps  of reconnecting purge
line  and  replacing  gas cap usually  will  reduce a major  portion of
the  evaporative  running  loss   emissions  for   most  vehicles.   The
above table  has  illustrated that for  these two vehicles,  however,
the high  levels of  emissions  may not  be caused by  either  pressure
or  purge  problems  in  their evaporative emissions   control  system.
The repair work failed  to decrease  their  emissions, even  increased
the   total   grams   slightly.     Thus,   their   extremely  high  tank
temperature rises could  be  the  main  reason why these  two  vehicles
had  such  high running  loss emissions.  For  these   reasons,  it was
decided that these  two vehicles be treated  as exceptions.   Their
average  emissions   from   as-received   and  after  repair  are  to  be
included in both the pass and failed vehicle categories.
                               -52-

-------
     Market  shares of  these  two engine  families  for  model  years
1983 through 1990  were  obtained from  EPA's Certification  files.
The  "Camaros"  included  all  models  of  Camaro  and Firebird  with
either 5.0L or 5.7L engines.  Their sales ranged  from a  low 0.46 to
a  high   2.02 percent,   with   an  average   of  1.06 percent.    The
"Sunbirds" were  models  under  the  names  of  Sunbird,  Skyhawk,  and
Firenza with either 1.8L or  2.0L engines.   Their sales  ranged from
0.48  to   2.01 percent,   with  an  average   of   1.28 percent.    An
alternative  was   to  derive  an  estimate  based  on  the  sample.   As
there  were  a total  of  60  LDGVs  in  the   MOBILE4.1  running  loss
emissions  test  program  sample,  weighting  factors  of 0.01667  each
are  used  to   represent  the  Camaros  and the  Sunbirds  in  the  fleet
(i.e.,  a   weighting  factor  of  0.96666  is  used for the  "regular"
vehicles).

     Table  2 summarizes  the  calculated  TVP values for  each  bag
under   each   fuel  volatility  level  and   each   initial  ambient
temperature  for  these two  exception  vehicles.   Table  3  shows  the
running loss emissions  (in  cumulative  grams)  of  these  two vehicles
to  be  used  for  failed  vehicles.   Table  6  shows  the running  loss
emissions  (in grams per mile) of these two  vehicles to  be used for
pass vehicles.
                                -53-

-------
3.2  Pass Vehicles

     Among the  "regular"  vehicles,  those  that  passed both  pressure
and  purge  checks  are categorized  as  the  "pass"  vehicles.   For
example,  there  are  32   model  year   1981+  pass   LDGVs   tested  at
9.0 psi/95°F for the LA-4 driving cycle.

     The  running  loss emissions for  pass  vehicles are expressed  in
unit of grams per mile (g/mi) for each bag (bags 1 through  6)  under
each  driving  cycle  (NYCC,  LA-4,  and HFET).   Note  that  there  are
only five bag  testing results from the HFET driving  cycle.   As  the
emission  levels  from five  bags  under the  combinations  of  95°F  and
fuel volatilities  of 9.0 and 11.7 psi  RVP are  relatively stable,
the  emissions   from the  sixth  bag  of the HFET driving  cycle  are
assumed  to  be the  same  as  the  emissions   from the  fifth  bag.
Therefore,  there  are  a  total  of  18  (6  bags  *  3  speeds)  emission
values  for  the combination  of  four  fuel  volatility  levels   (7.0,
9.0, 10.4,  11.7 psi RVP)  and four ambient temperature profiles  (80,
87, 95, 105°F).

     The  g/mi  running loss  emissions are  derived by  dividing the
total  grams  by  the  cumulative  trip  distances  in  miles.   For
example, for bags 1 and 2:

     RLsi,g/m\  — RLei , g / DBi ,m i

     RLe z ,gxm i
        = (RLB1)g + RLB2,g)  / (DBi,mi  + DBZ,mi)

where:

     D = trip distance in miles.

The  average  g/mi  running  loss emissions  for  each bag under each
driving cycle  for  1981+  "regular" pass  LDGVs  and  1981+ pass  LDGTs
are given  in  Tables 7 and  9.  Note  that  the  sample sizes  for the
combinations of  fuel volatility and  ambient  temperature  vary,  due
to the  limitation  of  the  test data discussed previously in sections
1.0  and 3.0.    When there  is no  actual  test  data  available,  the
sample  size is denoted as  zero  ("0"), and the  emission  values are
estimated   by    linear   interpolation/extrapolation.     Table   8
summarizes the  g/mi emissions used  for  MOBILE4.1  model 1981+ pass
LDGVs,   including  the  "regular"  pass vehicles  and  two  exception
vehicles.
                                -54-

-------
3.3  Failed Vehicles

     Due to  the limited  sample  sizes, there  is  no differentiation
made between  vehicles that  failed  pressure test  and  vehicles that
failed  purge  test.    The  criteria  used  to  define  vehicles  that
failed either pressure or purge test are the  same  as those used for
hot soak/diurnal emissions model.

     The   running   loss  emissions  for   failed   vehicles  in  the
MOBILE4.1 model  are  expressed in terms  of cumulative  grams  by bag
for   the   combinations    of   fuel   volatilities   and   ambient
temperatures.  Based on their  similar  emission levels  in cumulative
grams,  results  from  the  NYC  and LA-4  cycles were combined.   The
emissions  are   also   used  for   HFET  cycle.   The  emissions  in
cumulative grams  are  to be converted  onto grams per mile unit (and
adjusted  for  speed)  in the  MOBILE4.1 model,  as  to   be  discussed
later in section  4.0.   Emissions  for failed LDGVs are used also for
LDGTs, due to a lack of LDGT  data.   Further,  failed emission levels
are also used for vehicles made during the pre-control  era.

     Table  4   summarizes   the  running   loss  emissions  used  for
"regular"  failed  vehicles.   The  actual test  data  by  each  bag  were
used  to fit  equations  as a  function of True  Vapor   Pressure (as
discussed  in  section 2.0).  The  predicted emissions  in  cumulative
grams are  used in MOBILE4.1  model  for the  failed emission levels.
Table 5  shows  failed  emissions used for LDGVs and  LDGTs,  including
the "regular" failed vehicles  and two exception vehicles.
                                -55-

-------
4.0  Driving Characteristics

     In MOBILE4,  a VMT  weighted urban type driving characteristics
was  built  into  the  running   loss  emissions  model   so   that   the
emission  rates  were independent of the  vehicle speed.   This set  of
urban VMT  weighting factors was derived  from tabulated frequencies
by  trip distance  in miles  and average  speed in  mph  matching  the
three  driving  cycle  characteristics  used  in  the   running   loss
emissions  test program.   The  data sources  were  the  1979 GM-NPD
Survey Data,  and  an Operational Characteristics  Study  sponsored  by
EPA in Columbus, Ohio.

     In MOBILE4.1,  however, a  VMT  weighted driving characteristics
in terms  of  trip  duration  alone is used  in  the model  so  that the
emission  rates  can  be adjusted  by any  given  vehicle  speed.   The
1979 GM-NPD  Survey Data  again  are  used.   The  frequency  table was
based  on   trip  durations in  10-minute intervals,  and  the deriving
VMT weighting factors are:

     Trip  Duration     Corresponding  Percent
       in  minutes         Bag #         VMT

     10 and less            1         33.227
     11-20                2         32.883
     21-30                3         14.871
     31-40                4          7.886
     41-50                5          3.645
     52 and up              6          7.488
                           Total:    100.0

This new  set  of percent VMT weighting factors  is used  to  represent
each bag of the three driving cycles.


4.1  Pass  Vehicles

     For  pass  vehicles  in  the MOBILE4.1  model,  the  running  loss
emissions  are given for all six bags of the three driving  cycles in
grams  per  mile unit,  for  each of  the  four  fuel  volatilities  and
four ambient temperatures.  The  speed  adjustment  algorithm  for  pass
vehicles are described in the following.

     At any  given average vehicle  speed  between 7.1 and  47.9  mph,
the emissions  are  calculated  by linear  interpolation   between  the
emission  rates  at  the  two  adjacent   default  vehicle  speeds.   For
example,  emissions  at   10.0 mph are  to  be   calculated  based  on
emissions  at  7.1  mph (NYCC) and  emissions  at  19.6 mph  (LA-4),  and
emissions  at 30.0  mph are based on emissions  at  19.6 mph (LA-4)  and
emissions  at 47.9  mph (HFET):
                               -56-

-------
     ADJ = (SPD - SPD1) /  (SPD2 - SPD1)                         (4.1)

where:                   ---- ----

           ADJ = calculated speed adjustment factor,
           SPD = user input speed in mph, which is between  7.0  and
                 47.9 mph,
          SPD1 = average speed in mph for the default driving
                 cycle 1 (either NYCC or LA-4),
          SPD2 = average speed in mph for the default driving
                 cycle 2 (either LA-4 or HFET) .

Then, with the  calculated  speed adjustment factor, the g/mi running
loss  emissions  are  estimated  by  subtracting  the  resting   loss
emissions, and weighted by the VMT weighting factors:

     RESTL8/mi - RESTLg/hr /  (6.0 * SPDm , , h P )             (4.2)

     RL B i , S P D
where:
                  i.spDi + ADJ *  (RLBi,SpD3 - RLBi,spDi)
           - RESTLg/mi] * VMTBi   for i=l,2,...,6           (4.3)
         RESTL = resting loss emissions,
      RLBi,spD = calculated g/mi bag i running loss emissions at
                 speed SPD,
     RLBi,spDi = g/mi bag i running loss emissions for the default
                 driving cycle 1 (either NYCC or LA-4),
     RLBi,spD2 = g/mi bag i running loss emissions for the default
                 driving cycle 2 (either LA-4 or HFET),
         VMTB i = bag i VMT weighting factor.

     For any  average  speed below 7.1 mph but greater  than 2.5 mph,
the  emissions are calculated  from  the emissions  at  7.1 mph and an
adjustment  factor  calculated  from  the ratio of the two  speeds  (the
input speed divided  by 7.1 mph).  This same algorithm is applicable
to  where  the speed  is  above  47.9  mph  but  less  than  65.0 mph.
Therefore,  eguations (4.1) and (4.3) above are now reformulated as:

     ADJ -  SPD / SPD1

     RL Si , S P D
            =  [RLBi,spDi + ADJ * RLBi,spDi - RESTLg/mi]
              * VMTB i           ~ for i = l,2, ... ,6
where:
      SPD = user input speed in mph, which is either between 2.5
            and 7.1 mph, or between 47.9 and 65.0 mph,
                                -57-

-------
     SPD1 = average speed of 7.1 mph  from  NYC  driving cycle if SPD
            is less than --2..JD mph, or  average speed  of 47.9 mph from
            HFET driving cycle  if SPD is greater  than 47.9 mph,
     RLflt.spDi = g/mi bag i running loss emissions  from NYC
                 driving cycle  if SPD is below 7.1  mph,  or those
                 from HFET driving cycle if SPD is  above 47.9  mph.

Finally,  the  running  loss  emissions  are the total  of  the  g/mi
emissions over the six bags:
     RL = RL B 1 , S P D + RL B Z , S P D + RL B 3 , S P D + RL B 4 , S P D

        + RLBS,SPD '+ RLB6,SPD                               (4.4)
4.2  Failed Vehicles

     The  running   loss   emissions  for  failed   vehicles   in  the
MOBILE4.1  model  are  expressed   in  cumulative  grams.   For  each
combination of  fuel volatility  and  ambient temperature,  there are
only  six  emission  values,   each  representing  the  cumulative bag
emissions  for  all  three driving cycles.   The  speed  adjustment
algorithm  for  failed  vehicles  are  based  on  the  following two
assumptions:

     1)    the  durations  for  each bag  of  each  driving cycle are
           approximately the same, (i.e., about 10 minutes),

     2)    the  cumulative  emissions   in grams  by  bag under  each
           driving cycle from failed vehicles are the same.

     For  failed vehicles, the  speed  and  VMT adjustment  algorithm
includes  converting  the  emissions from  cumulative  grams onto  grams
per mile  unit by  the  given  speed, as  well as subtracting  off the
resting loss emissions and multiplying VMT weighting factor:

     RESTLg/io min= RESTLg/nr / 6.0io mln/hr

     RL8i,g/mi = [(RLBi,g - RESTL) * VMTBl]  / DB1,mi
               = (RLBi,g  - RESTL) * VMTB ! * 6 / SPD

     RLB2>g/mi = [(RLB2>g - RESTL) * VMTB2]
                 / (DB 1 ,mi + DB2 ,mi )
               = (RLB2>g  - RESTL) * VMTB2 / (2 *  Dai.mi)
               = (RLB2)g  - RESTL) * VMTB2 * 6 / (2  * SPD)
then,
     RLBi>g/mi  = (RLBi(g - RESTL) * VMTBi  * 6 / (i * SPD)
                                                  for i = 1, 2, .. . , 6
                               -58-

-------
and,

     RL SPD = RI» B 1 , g / m 1 + RL B 2 , g X m 1  +  RL B 3
            •f RLB4,gXmi + RLflS.gXmi + RLB6,gXmi

where:

         RESTL -  resting loss  emissions,
     RLBi,gxini =  calculated  g/mi bag  i running  loss  emissions  at
                  speed SPD,
           i,g =  bag  i cumulative  grams  running  loss emissions,
              i =  bag  i VMT weighting  factor,
        DBi,mi =  trip distance in  miles  for bag  i, and,
           SPD =  user input  speed  in  mph.
                                -59-

-------
5.0  Other Model Years and Other Vehicle Types
     The running  loss  emissions data described  in  Tables 5, 7, and
9  are  used  in MOBILE4.1  for  other model  year  groups  and other
vehicle types according to the following mapping:
      Data
    1981+ Pass LDGV
    1981+ Pass LDGT1
Pass/
Fail

Pass
Pass
Pass

Pass
Pass
Pass
                                       Vehicle Type
 LDGV

1972-77
1978-80
1981+
LDGT1
LDGT2
HDGV
                  1985+
          1972-77
          1978-80   1979-80
          1981+     1981+
    1981+ Failed LDGV All     Pre-1972  Pre-1972  Pre-1979 Pre-1985
                      Failed  1972+     1972+     1979+    1985+
                               -60-

-------
 Fuel   Initial
 RVP    Ambient
in psi  Temp °F
 7.0
 9.0
10.4
11.7
 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0
                           Table 1

                     True Vapor Pressure
                      Regular  Vehicles
                        True Vapor Pressure
Bag 1
5.58
6.42
7.50
9.04
7.28
8.32
9.64
11.54
8.63
9.82
11.34
13.50
9.76
11.07
12.74
15.11
Bag 2
6.67
7.64
8.88
10.65
8.63
9.82
11.34
13.50
10.19
11.55
13.27
15.72
11.47
12.96
14.86
17.54
Bag 3
7.21
8.24
9.56
11.44
9.30
10.56
12.17
14.45
10.95
12.39
14.21
16.80
12.31
13.89
15.89
18.71
Bag 4
7.94
9.04
10.46
12.48
10.19
11.54
13.27
15.71
11.96
13.50
15.46
18.22
13.42
15.11
17.26
20.27
Bag 5
8.24
9.38
10.84
12.92
10.56
11.95
13.73
16.24
12.39
13.97
15.98
18.82
13.89
15.63
17.83
20.92
Bag 6
8.72
9.91
11.44
13.60
11.14
12.60
14.45
17.06
13.05
14.70
16.80
19.74
14.61
16.43
18.71
21.93
                             -61-

-------
                           Table 2
                     True Vapor Pressure
                      Exception Vehicles
 Fuel   Initial
 RVP    Ambient
in psi  Temp °F
 7.0
 9.0
10.4
11.7
 7.0
 9.0
10.4
11.7
 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0
 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0
                        True Vapor Pressure
Bag 1

6.42
7.35
8.55
10.28
8.32
9.47
10.94
13.04
9.82
11.14
12.82
15.21
11.07
12.52
14.37
16.97

6.05
6.94
8.09
9.73
7.86
8.96
10.37
12.38
9.30
10.56
12.17
14 .46
10.49
11.88
13.65
16.16
Bag 2
Camaro
7.79
8.88
10.28
12.27
10.00
11.34
13.04
15.45
11.75
13.27
15.21
17.93
13.19
14.86
16.97
19.95
Sunbird
8.40
9.56
11.04
13.14
10.75
12.17
13.96
16.51
12.60
14.21
16.25
19.12
14. 13
15.89
18.12
21.25
Bag 3

9.56
10.84
12.48
14.80
12.17
13.73
15.71
18.50
14.21
15.98
18.22
21.37
15.89
17.83
20.27
23.69

9.38
10.65
12.27
14.56
11.95
13.50
15.45
18.21
13.97
15.72
17.93
21.03
15.63
17.54
19.95
23.33
Bag 4

10.84
12.27
14.07
16.63
13.73
15.45
17.62
20.68
15.98
17.93
20.38
23.81
17.83
19.95
22.62
26.35

10.28
11.64
13.37
15.82
13.04
14.69
16.78
19.72
15.21
17.07
19 .43
22.74
16.97
19.02-
21.59
25.18
Bag 5

11.64
13.14
15.05
17.75
14.69
16.51
18.80
22.02
17.07
19.12
21.70
25.31
19.02
21.25
24.06
27.97

11.04
12.48
14.31
16.90
13.96
15.71
17.91
21.01
16.25
18.22
20.71
24.18
18.12
20.27
22.97
26.75
Bag 6

12.27
13.84
15.82
18.63
15.45
17.34
19.72
23.07
17.93
20.06
22.74
26.48
19.95
22.27
25.18
29.24

11.44
12.92
14.80
17.46
14.45
16.24
18.50
21.68
16.80
18.82
21.37
24.93
18.71
20.92
23.69
27.56
                            -62-

-------
                              Table 3
              Cumulative Grams Running Loss Emissions
                        Exception Vehicles
    Fuel
    RVP
  in psi
    7.0
    9.0
   10.4
   11.7
    7.0
    9.0
   10.4
   11.7
Ambient
Temp °F
  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0
  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0
Running Loss Emissions (Cumulative Grams)
Bag 1
Bag 2
Bag 3
Bag 4
Bag 5
Bag 6
Camaro
1.83
1.96
2.15
2.47
2.11
2.31
2.61
3.11
2.38
2.65
3.05
3.71
2.64
2.98
3.47
4.28

0.71
1.40
2.42
4.10
1.52
2.97
5.10
8.66
3.45
5-41
8.26
13.01
5.29
7.72
11.24
17.07
2.27
3.99
7.00
12.07
4.12
8.62
15.16
25.98
10.12
16.13
24.80
39.04
15.78
23.17
33.78
51.11
Sunbi
2.00
4.66
8.48
22.72
6.73
15.79
28.93
50.59
18.83
30.89
48.24
76.64
30.19
44.99
66.18
100.64
3.83
7.98
16.62
30.78
5.07
19.69
40.76
75.27
24.59
43.92
71.58
116.55
42.82
66.43
100.05
154.42
rd
4.24
5.43
12.92
34.93
17.02
21.95
51.91
139.24
26.83
58.18
128.15
283.75
55.95
113.31
221.83
441.95
8.06
11.20
32.01
82.27
33.30
45.99
130.51
378.61
59.71
148.49
347.42
789.42
142.28
305.45
614.02
818.00*

6.68
17.12
43.89
87.78
18.73
48.02
123.67
247.17
65.67
135.05
234.04
394.43
131.16
215.68
335.71
529.16
12.55
16.73
47.79
120.61
50.25
67.15
192.70
565.10
87.34
219.75
518.46
818.00*
210.54
455.53
818.00*
818.00*

10.05
25.76
66.05
132.10
28.42
72.88
185.66
369.14
99.29
202 .69
349 .77
587.29
196.95
322.60
500.51
719.06*
36.78
49.04
122.60
240.39
100.62
134.16
336.79
664.77
181.88
367.54
630.51
818.00*
357.42
582.20
818.00*
818.00*

16.20
41.53
94.39
178.09
41.96
107.60
241.92
460.06
139.12
262.23
437.11
719.06
255.44
404.86
616.16
719.06*
*An upper  limit based on actual  test  results
                               -63-

-------
  7.0
  9.0
 10.4
 11.7
                            Table 4

            Cumulative Grams Running Loss Emissions
                        Failed Vehicles
  Fuel
  RVP   Ambient
in psi  Temp °F
 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0
             Running Loss Emissions (Cumulative Grams)
Bag 1
0.50
1.02
1.52
3.25
1.23
2.52
3.76
4.79
2.85
3.90
4.72
4.90
3.85
4.62
4.97
6.66
Bag 2
1.44
2.48
7.27
14.13
6.32
10.92
16.79
25.13
12.34
17.59
24.28
33.74
17.30
23.08
30.42
40.78
Bag 3
3.26
6.80
11.34
26.24
9.28
19.29
32.04
50.13
22.37
33.79
48.28
68.78
33.15
45.69
61.58
83.99
Bag 4
4.33
10.07
17.37
44.20
13.69
31.71
54.63
87.09
37.28
57.80
83.80
120.49
56.67
79.17
107.64
147.69
Bag 5
4.91
12.27
21.52
58.84
16.41
41.49
73.37
118.47
49.25
77.78
113.91
164.85
76.21
107.49
147.02
202.59
Bag 6
5.83
15.34
27.40
77.40
20.58
54.20
96.87
157.15
64.60
102.78
151.10
219.11
100.70
142.53
195.33
269.48
                             -64-

-------
  7.0
  9.0
 10.4
 11.7
                            Table 5

            Cumulative Grams Running Loss Emissions
                  1981+ Failed LDGVs  and  LDGTs
  Fuel
  RVP   Ambient
in psi  Temp °F
 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0
             Running Loss Emissions (Cumulative Grams)
Bag 1
0.53
1.04
1.55
3.25
1.25
2.52
3.76
4.83
2.85
3.90
4.75
5.02
3.85
4.64
5.05
6.79
Bag 2
1.46
2.54
7.29
14.24
6.29
10.96
16.97
25.57
12.41
17.79
24.69
34.54
17.49
23.45
31.07
41.95
Bag 3
3.29
6.80
11.45
26.46
9.34
19.34
32.52
52.04
22.48
34.37
50.00
73.16
33.69
47.16
64.89
91.13
Bag 4
4.43
10.21
18.06
45.56
14.10
32.22
57.05
94.62
38.13
60.60
90.70
136.21
59.34
85.22
119.88
165.22
Bag 5
5.12
12.57
22.70
61.09
17.17
42.44
77 .23
130.10
50.72
82.23
124.59
182.78
80.46
116.88
164.10
222.58
Bag 6
6.52
16.34
30.10
81.80
22.27
56.42
103.29
170.66
67.80
109.85
163.86
237.42
107.56
154.23
212.73
287.77
                             -65-

-------
                            Table 6
  7.0
  9.0
 10.4
 11.7
  7.0
  9.0
 10.4
 11.7
             Grams per Mile Running Loss Emissions
                  Exception Vehicle:   Camaro
  Fuel
  RVP   Ambient
in psi  Temp °F
 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0
 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0

 80.0
 87.0
 95.0
105.0
                Running Loss Emissions  (Grams/Mile)
Bag 1
New
1.55
1.66
1.82
2.09
1.79
1.96
2.21
2.64
2.02
2.25
2.58
3.14
2.24
2.53
2.94
3.63
Bag 2
Bag 3
Bag 4
Bag 5
Bag 6
York City Cycle
0.96
1.69
2.97
5.11
1.75
3.65
6.42
11.01
4.29
6.83
10.51
16.54
6.69
9.82
14.31
21.66
LA-4 Driving
0.51
0.55
0.60
0.69
0.59
0.64
0.73
0.87
0.66
0.74
0.85
1.03
0.74
0.83
0.97
1.19
0.30
0.53
0.93
1.61
0.55
1.15
2.02
3.46
1.35
2.15
3.31
5.21
2.10
3.09
4.50
6.81
1.08
2.25
4.69
8.69
1.43
5.56
11.51
21.26
6.95
12.41
20.22
32.92
12.10
18.77
28.26
43.62
Cycle
0.35
0.72
1.50
2.78
0.46
1.78
3.68
6.79
2.22
3.96
6.45
10.51
3.86
5.99
9 .02
13.92
1.71
2.37
6.78
17.43
7.06
9.74
27.65
80.21
12.65
31.46
73.61
167.25
30.14
64.71
130.09
173.31

0.54
0.75
2.13
5.48
2.22
3.07
8.70
25.24
3.98
9.90
23.16
52.63
9.49
20.63
40.93
54.53
2.13
2.84
8.10
20.44
8.52
11.38
32.66
95.78
14.80
37.25
87.87
138.64*
35.68
77.21
138.64*
138.64*

0.68
0.90
2.57
6.49
2.70
3.61
10.37
30.40
4.70
11.82
27.89
44 .00*
11.33
24.50
44 .00*
44 .00*
5.19
6.93
17.32
33.95
14.21
18.95
47.57
93.89
25.69
51.91
89.06
115.54*
50.48
82.23
115.54*
115.54*

1.63
2.18
5.45
10.68
4.47
5.96
14.97
29.55
8.08
16.34
28.02
36.36*
15.89
25.88
36.36*
36.36*
                             -66-

-------
                      Table 6 (Continued)
             Grams per Mile Running Loss Emissions
                  Exception Vehicle:  Sunbird
  Fuel
  RVP
in psi
  7.0
  9.0
 10.4
 11.7
  7.0
  9.0
 10.4
 11.7
Ambient
Temp °F
  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0
  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0

  80.0
  87.0
  95.0
 105.0
Running Loss Emissions (Grams/Mile)
Bag 1
New
0.60
1.19
2.05
3.47
1.29
2.52
4.32
7.34
2.92
4.58
7.00
11.03
4.48
6.54
9.53
14.47
Bag 2
Bag 3
Bag 4
Bag 5
Bag 6
York City Cycle
0.85
1.97
3.59
9.63
2.85
6.69
12.26
21.44
7.98
13.09
20.44
32.47
12.79
19.06
28.04
1.20
1.53
3.65
9.87
4.81
6.20
14.66
39.33
7.58
16.44
36.20
80.16
15.81
32.01
62.66
42.64 124.84
LA-4 Driving
0.20
0.39
0.67
1.14
0.42
0.83
1.42
2.41
0.96
1.51
2.30
3.62
1.47
2.15
3.13
4.75
0.27
0.62
1.13
3.03
0.90
2.11
3.86
6.75
2.51
4.12
6.43
10.22
4.03
6.00
8.82
13.42
Cycle
0.38
0.49
1.17
3.15
1.53
1.98
4.68
12.56
2.42
5.25
11.56
25.59
5.05
10.22
20.00
39.85
1.42
3.63
9.30
18.60
3.97
10.17
26.20
52.37
13.91
28.61
49.58
83.57
27.79
45.69
71.13
112.11

0.45
1.14
2.93
5.85
1.25
3.20
8.24
16.48
4.38
9.00
15.60
26.30
8.74
14.38
22.38
35.28
1.70
4.37
11.19
22.39
4.82
12.35
31.47
62.57
16.83
34.35
59.28
99.54
33.38
54.68
84.83
133.28

0.54
1.39
3.55
7.11
1.53
3.92
9.99
19.86
5.34
10.90
18.81
31.59
10.59
17.35
26.92
42.30
2.29
5.87
13.33
25.15
5.93
15.20
34.17
64.98
19.65
37.04
61.74
101.56*
36.08
57.18
87.03
101.56*

0.72
1.85
4.20
7.92
1.86
4.78
10.75
20.45
6.18
11.65
19.43
31.96*
11.35
17.99
27.38
31.96*
                             -67-

-------
               Table 7
Grams per Mile Running Loss Emissions
        Regular Pass Vehicles
Fuel
RVP
psi
7.0



9.0



10.4



11.7



7.0



9.0



10.4



11.7




Ambient
Temp °F
80.0
87.0
95.0
105.0
80.0
87.0
95.0
105.0
80.0
87.0
95.0
105.0
80.0
87.0
95.0
105.0
80.0
87.0
95.0
105.0
80.0
87.0
95.0
105.0
80.0
87.0
95.0
105.0
80.0
87.0
95.0
105.0

Sample
Size
0
0
12
12
0
0
8
14
0
0
0
0
0
0
0
0
0
0
12
12
0
0
32
14
4
0
0
0
0
0
0
0
               Running Loss Emissions (Grams/Mile)
Bag 1
New York
0.28
0.28
0.28
0.29
0.28
0.29
0.33
0.67
0.29
0.36
0.63
1.37
0.35
0.59
1.10
1.95
Bag 2
City
0.30
0.30
0.30
0.31
0.30
0.31
0.35
0.91
0.31
0.40
0.85
2.25
0.38
0.77
1.73
3.35
Bag 3
Cycle
0.31
0.31
0.31
0.32
0.31
0.32
0.37
1.72
0.32
0.50
1.58
6.09
0.45
1.39
4.40
9.65
Bag 4

0.31
0.31
0.31
0.33
0.31
0.32
0.38
2.80
0.32
0.61
2.55
13.71
0.53
2.20
9.53
22.60
Bag 5

0.30
0.30
0.30
0.32
0.30
0.31
0.53
3.90
0.31
0.85
3.55
22.93
0.74
3.08
15.63
38.43
Bag 6

0.30
0.30
0.30
0.32
0.30
0.31
0.63
4.87
0.31
1.04
4.45
29.25
0.73
3.85
19.88
49.17
LA-4 Driving Cycle
0.07
0.07
0.07
0.09
0.07
0.08
0.12
0.12
0.08
0.12
0.12
0.25
0.12
0.12
0.20
0.36
0.09
0.09
0.09
0.11
0.09
0.10
0.17
0.22
0.10
0.17
0.22
0.42
0.17
0.21
0.34
0.58
0.09
0.09
0.09
0.18
0.09
0.10
0.18
0.29
0.10
0.19
0.28
1.32
0.19
0.27
0.92
2. 16
0.09
0.09
0.09
0.11
0.09
0.10
0.21
0.34
0.10
0.22
0.33
4.23
0.22
0.31
2.74
7.40
0.09
0.09
0.09
0.11
0.09
0.10
0.23
0.65
0.10
0.27
0.61
6.28
0.26
0.55
4.12
10.86
0.09
0.09
0.09
0.11
0.09
0.10
0.30
1.00
0.10
0.36
0.92
7.58
0.34
0.82
5.05
12.96
                -68-

-------
         Table 7 (Continued)

Grams per Mile Running Loss Emissions
        Regular Pass Vehicles
Fuel
RVP
psi

7.0



9.0



10.4



11.7



Ambient
Temp

80.
87.
95.
105.
80.
87.
95.
105.
80.
87.
95.
105.
80.
87.
95.
105.
op

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Sample
Size

0
0
0
0
0
0
7
0
0
0
0
0
0
0
7
0
Bag
Highway
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
Running Loss
1
Fuel
02
02
02
06
02
02
02
09
02
02
08
09
02
05
09
09
Bag 2
Bag
Emissions
3
Bag 4
(Grams /Mile)
Bag 5
Bag 6
Economy Test
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.05
0.05
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
02
02
02
02
02
02
02
02
02
02
02
03
02
03
04
04
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
                 -69-

-------
                            Table 8

             Grams per Mile Running Loss Emissions
                        1981+ Pass  LDGVs

  Fuel
  RVP   Ambient    	Running Loss Emissions (Grams/Mile)
in psi  Temp °F
  7.0     80.0
          87.0
          95.0
         105.0

  9.0     80.0
          87.0
          95.0
         105.0

 10.4     80.0
          87.0
          95.0
         105.0

 11.7     80.0
          87.0
          95.0
         105.0
  7.0     80.0
          87.0
          95.0
         105.0

  9.0     80.0
          87.0
          95.0
         105.0

 10.4     80.0
          87.0
          95.0
         105.0

 11.7     80.0
          87.0
          95.0
         105.0
Bag 1
New
0.31
0.32
0.34
0.37
0.32
0.35
0.43
0.81
0.36
0.46
0.77
1.56
0.45
0.72
1.27
2.18
Bag 2
York City
0.32
0.35
0.40
0.55
0.37
0.47
0.65
1.42
0.50
0.72
1.34
2.99
0.70
1.23
2.38
Bag 3
Cycle
0.34
0.36
0.44
0.62
0.40
0.50
0.79
2.67
0.55
0.96
2.47
7.78
0.90
2.19
5.77
4.31 12.14
LA-4 Driving
0.08
0.08
0.09
0.12
0.08
0.10
0.15
0.17
0.11
0.15
0.17
0.32
0.15
0.17
0.26
0.44
0.10
0.11
0.12
0.18
0.11
0.15
0.26
0.38
0.17
0.27
0.37
0.66
0.27
0.35
0.55
0.90
Cycle
0.10
0.11
0.13
0.21
0.12
0.16
0.31
0.60
0.18
0.34
0.57
1.88
0.33
0.53
1.37
2.98
Bag 4

0.35
0.40
0.57
0.92
0.48
0.64
1.27
4.92
0.76
1.59
4.52
17.42
1.48
3.97
12.57
26.60

0.10
0.12
0.17
0.30
0.14
0.20
0.49
1.02
0 .24
0.53
0.96
5.40
0.51
0.88
3.70
8.65
Bag 5

0.35
0.41
0.61
1.02
0.51
0.70
1.58
6.41
0.83
2.02
5.89
26.14
1.87
5.18
18.83
41.68

0.11
0.13
0.19
0.33
0.16
0.22
0.56
1.47
0.27
0.64
1.37
7.33
0.61
1.23
5.17
11.94
Bag 6

0.41
0.50
0.80
1.29
0.63
0.87
1.97
7.36
1.06
2.49
6.82
31.89
2.15
6.05
22.59
55.39

0.13
0.15
0.25
0.42
0.19
0.28
0.72
1.80
0.34
0.82
1.68
8.47
0.79
1.53
5.94
13.74
                             -70-

-------
                      Table 8 (Continued)

             Grams per Mile Running Loss Emissions
                        1981+ Pass LDGVs

  Fuel
  RVP   Ambient    	Running Loss Emissions (Grams/Mile)
in psi  Temp °F
  7.0     80.0
          87.0
          95.0
         105.0

  9.0     80.0
          87.0
          95.0
         105.0

 10.4     80.0
          87.0
          95.0
         105.0

 11.7     80.0
          87.0
          95.0
         105.0
Bag 1
Highway
0.02
0.02
0.02
0.06
0.02
0.02
0.02
0.09
0.02
0.02
0.08
0.09
0.02
0.05
0.09
0.09
Bag 2
Fuel
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.05
0.05
Bag 3
Economy
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.04
0.04
Bag 4
Test
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
Bag 5

0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
Bag 6

0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
                             -71-

-------
                            Table 9

             Grams per Mile Running Loss Emissions
                        1981+ Pass  LDGTs

  Fuel
  RVP   Ambient    	Running Loss Emissions (Grams/Mile)
in psi  Temp °F
  7.0     80.0
          87.0
          95.0
         105.0

  9.0     80.0
          87.0
          95.0
         105.0

 10.4     80.0
          87.0
          95.0
         105.0

 11.7     80.0
          87.0
          95.0
         105.0
  7.0     80.0
          87.0
          95.0
         105.0

  9.0     80.0
          87.0
          95.0
         105.0

 10.4     80.0
          87.0
          95.0
         105.0

 11.7     80.0
          87.0
          95.0
         105.0
Bag 1
New
0.45
0.49
0.55
0.37
0.61
0.60
0.56
0.81
0.62
0.71
0.80
1.56
0.66
0.77
0.89
2.18
Bag 2
York City
0.41
0.45
0.50
0.55
0.61
0.54
0.46
1.42
0.56
0.63
0.71
2.99
0.57
0.74
0.90
4.31
LA-4 Driving
0.12
0.13
0.09
0.23
0.12
0.10
0.10
0.25
0.15
0.16
0.18
0.18
0.17
0.16
0.16
0.23
0.14
0.14
0.12
0.25
0.12
0.10
0.10
0.18
0.15
0.15
0.16
0.16
0.12
0.16
0.16
0.16
Bag 3
Cycle
0.34
0.41
0.48
0.62
0.77
0.62
0.53
2.67
0.66
0.86
1.14
7.78
0.72
1.23
1.74
12.14
Cycle
0.11
0.11
0.13
0.22
0.11
0.09
0.09
0.15
0.13
0.14
0.16
0.16
0.10
0.19
0.23
0.21
Bag 4

0.29
0.37
0.45
0.92
0.85
0.67
0.61
4.92
0.74
1.07
1.57
17.42
0.82
1.96
3.09
26.60

0.07
0.08
0.17
0.23
0.12
0.09
0.09
0.28
0.14
0.18
0.24
0.26
0.09
0.41
0.53
0.58
Bag 5

0.29
0.39
0.43
1.02
0.99
0.78
0.71
6.41
0.88
1.34
2.08
26.14
1.11
2.78
4.40
41.68

0.06
0.07
0.19
0.26
0.11
0.09
0.09
0.35
0.14
0.20
0.28
0.31
0.08
0.57
0.74
0.91
Bag 6

0.32
0.44
0.42
1.29
1.21
0.93
0.78
7.36
1.05
1.66
2.67
31.89
1.66
3.80
5.93
55.39

0.05
0.07
0.25
0.32
0.11
0.11
0.11
0.51
0.16
0.24
0.39
0.43
0.11
0.95
1.07
1.72
                             -72-

-------
                      Table 9 (Continued)

             Grams per Mile Running Loss Emissions
                        1981+  Pass  LDGTs

  Fuel
  RVP   Ambient    	Running Loss Emissions (Grams/Mile)
in psi  Temp °F
  7.0     80.0
          87.0
          95.0
         105.0

  9.0     80.0
          87.0
          95.0
         105.0

 10.4     80.0
          87.0
          95.0
         105.0

 11.7     80.0
          87.0
          95.0
         105.0
Bag 1
Highway
0.02
0.02
0.02
0.06
0.02
0.02
0.02
0.09
0.02
0.02
0.08
0.09
0.02
0.05
0.09
0.09
Bag 2
Fuel
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.05
0.05
Bag 3
Economy
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.04
0.04
Bag 4
Test
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
Bag 5

0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
Bag 6

0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.03
0.03
0.03
                             -73-

-------
                                                                  MOBILE4.1
                                           Technology Distribution for Gasoline Light-Duty Trucks

Model
lea
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991

Model
feK
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
Model
Year
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
CLLP
3way
Curb
0
57339
74447
109748
255529
167151
260447
260231
174645
7985
21900
CLLP
Sway
Cafe
0.0%
3.9%
3.8%
3.7%
82%
4.6%
7.4%
66%
37%
02%
0.4%
CLLP
Carb
1.8%
3.9%
13.1%
25.0%
25.8%
13.1%
12.8%
92%
7.7%
20%
1.5%
CLLP
Sway
MPFI
0
0
3438
65296
206437
861000
716266
1099548
1544412
1948880
2177487
CLLP
Sway
MEQ
0.0%
0.0%
02%
22%
6.6%
23.8%
20.4%
280%
33.1%
41.3%
43.5%
CLLP
MPFI
0.0%
0.0%
0.2%
22%
6.6%
23.8%
32.7%
41.6%
54.0%
60.1%
57.7%
CLLP
Sway
IBI
0
0
0
0
144973
328016
780032
1455516
1435056
1533017
1815964
CLLP
Sway
IBI
0.0%
0.0%
0.0%
00%
4.7%
9.1%
22.3%
37.0%
30.7%
32.5%
362%
CLLP
IBI
0.0%
0.0%
0.0%
0.0%
4.7%
13.6%
28.0%
442%
36.9%
36.7%
40.7%
Oplp
Sway
Cab
13589
0
0
20757
15454
17780
48058
15860
10998
9998
6000
Oplp
Sway
CMfe
0.9%
0.0%
0.0%
0.7%
05%
05%
1.4%
04%
02%
02%
0.1%
Oplp
Any
982%
96.1%
86.8%
72.8%
62.8%
495%
26.4%
5.0%
1.4%
12%
0.1%
Oplp
Sway
MPFI
0
0
0
0
0
0
0
0
0
0
0
Oplp
Sway
MEQ
0.0%
0.0%
0.0%
0.0%
00%
0.0%
0.0%
0.0%
0.0%
0.0%
00%













Oplp
Sway
HI
0
0
0
0
0
0
0
38829
0
0
0
Oplp
Sway
HI
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
1.0%
0.0%
00%
0.0%
Model
Xsss
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
CLLP
OxSway
Carb
26593
906
163194
631304
546263
308664
188653
101129
185103
87597
53055
CLLP
OxSway
Cert.
1.8%
0.1%
93%
213%
17.6%
85%
5.4%
2.6%
4.0%
1.9%
1.1%
Any
Carte
100.0%
100.0%
99.8%
97.8%
88.7%
62.7%
39.3%
132%
9.1%
32%
1.6%
CLLP
OxSway
MPFI
0
0
0
0
0
0
429429
534243
974850
884251
712782
CLLP
OxSway
MPFI
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
123%
13.6%
20.9%
18.7%
142%
Any
MPFI
0.0%
0.0%
02%
22%
6.6%
23.8%
32.7%
416%
54.0%
60.1%
57.7%
CLLP
OxSway
HI
0
0
0
0
368
163878
202385
283427
288728
200673
223736
CLLP
OxSway
HI
0.0%
0.0%
0.0%
0.0%
0.0%
45%
5.8%
72%
62%
43%
4.5%
Any
HI
0.0%
0.0%
0.0%
0.0%
4.7%
13.6%
28.0%
452%
36.9%
36.7%
40.7%
Oplp
OxSway
Carb
0
0
0
51116
66837
0
64044
69315
0
0
0
Oplp
OxSway
Carb
0.0%
0.0%
0.0%
1.7%
22%
0.0%
1.8%
1.8%
0.0%
0.0%
0.0%













Oplp
OxSway
MEQ
0
0
0
0
0
0
0
0
0
0
0
Oplp
OxSway
MEQ
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Model
feat
1981
1962
1983
1984
1985
1986
1987
1988
1989
1990
1991
Oplp
OxSway
HI
0
0
0
0
0
0
0
0
0
0
0
Oplp
OxSway
HI
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Any
Sway
0.9%
3.9%
4.0%
6.6%
20.0%
37.9%
515%
73.0%
67.8%
742%
803%
Oplp

Qxjd
1429888
1429253
1658258
2081626
1868467
1776740
814666
72566
53200
40500
0
Oplp

Oxid
943%
96.1%
84.1%
70.3%
602%
49.0%
232%
1.8%
1.1%
0.9%
0.0%
Any
OxSway
1.8%
0.1%
9.3%
23.1%
19.8%
13.0%
252%
25.1%
31.0%
24.9%
19.7%
Oplp

None
46007
0
51277
0
0
0
0
0
0
4969
0
Oplp

None
3.0%
0.0%
2.6%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.1%
0.0%
Any
fixjd
943%
96.1%
84.1%
703%
602%
49.0%
232%
1.8%
1.1%
0.9%
0.0%
AH vehicle counts taken from CAFE estimates, except 1989 and newer model years, which are from General Label predictions.
                                                                                  7/2/91

-------
                          APPENDIX D
PURGE AND PRESSURE TEST EFFECTIVENESS FIGURES AND  SPREADSHEET

-------
o
a:
   0.500



   0.450



   0.400



   0.350



   0.300



Sf  0.250
3


°  0.200



   0.150



   0.100



   0.050



   0.000
                                            Purge-Pressure Detection Rates

                                                         Annual
                   A— W/Program



                   ~D— No-Program
	D	 D-   D   D   D    D   D

-------
   0.500


   0.450


   0.400


   0.350


-g 0.300
ce:

 gf 0.250
 13

"L° 0.200


   0.150


   0.100


   0.050 1


   0.000
                                             Purge-Pressure Detection Rates
                                                          Biennial
      -A— W/Program


      -D— No-Program
D	
                                                   a
                                                           D—a
10
15
                                                                                            20
                                                                                                    25
                                                         Age Index

-------
Revised Purge/Pressure Failure Rates for MOBILE4.1
ANNUAL
In- lane
Age No-program
Index fail rate
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25


0.080
0.080
0.080
0.096
0.103
0.120
0.150
0.188
0.258
0.323
0.389
0.442
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451


New
Detect
Rate
0.0800
0.0800
0.0600
0.0160
0.0070
0.0170
0.0300
0.0380
0.0700
0.0650
0.0660
0.0530
0.0250
0.0070
0.0170
0.0300
0.0380
0.0700
0.0650
0.0660
0.0530
0.0250
0.0070
0.0170
0.0300
0.0380
0.0412
Annual
After
0.0800
0.0600
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000


Annual
average
0.0800
0.0600
0.0080
0.0035
0.0085
0.0150
0.0190
0.0350
0.0325
0.0330
0.0265
0.0125
0.0035
0.0085
0.0150
0.0190
0.0350
0.0325
0.0330
0.0265
0.0125
0.0035
0.0085
0.0150
0.0190

Averaqe
Annual
% benefit
0.000
0.238
0.853
0.910
0.836
0.789
0.765
0.732
0.798
0.837
0.903
0.967
0.992
0.981
0.967
0.958
0.922
0.928
0.927
0.941
0.972
0.992
0.981
0.967
0.958

0.908
BIENNIAL
No-program
Age in-lane
Index fail rate
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25


0.080
0.080
0.080
0.096
0.103
0.120
0.150
0.188
0.258
0.323
0.389
0.442
0.451
0.451
0.451
0.451
0.451
0.451
-0.451
0.451
0.451
0.451
0.451
0.451
0.451


New Detect
1
0
Rate
0.0800
0.0800
0.0600
0.0160
0.0070
0.0240
0.0300
0.0680
0.0700
0.1350
0.0660
0.1330
0.0250
0.0250
0.0240
0.0240
0.0680
0.0680
0.1350
0.1350
0.1330
0.1330
0.0250
0.0250
0.0240
0.0240
0.0630
Bien.
After
0.0800
0.0600
0.0000
0.0000
0.0070
0.0000
0.0300
0.0000
0.0700
0.0000
0.0660
0.0000
0.0250
0.0000
0.0240
0.0000
0.0680
0.0000
0.1350
0.0000
0.1330
0.0000
0.0250
0.0000
0.0240


Bien.
averaqe
0.0800
0.0600
0.0080
0.0035
0.0155
0.0150
0.0490
0.0350
0.1025
0.0330
0.0995
0.0125
0.0250
0.0120
0.0240
0.0340
0.0680
0.0675
0.1350
0.0665
0.1330
0.0125
0.0250
0.0120
0.0240

Averaqe
Bien.
% benefit
0.000
0.238
0.853
0.910
0.779
0.789
0.612
0.732
0.582
0.837
0.733
0.967
0.945
0.973
0.947
0.925
0.849
0.850
0.701
0.853
0.705
0.972
0.945
0.973
0.947

0.843

-------
Revised Purge/Pressure Failure Rates for MOBILE4.1
ANNUAL
Age
Index
1
-14A5
-14A6
-14A7
-14A8
-14A9
-14A10
-14A11
-14A12
-14A13
-14A14
-14A15
-14A16
-14A17
-14A1B
-14A19
— 1 4ATI
— iTAiV
-14A21
-14A22
-14A23
-14A24
-14A25
-14A26
-14A27
-14A28


In-lane
No-program
fail rate
0.08
0.08
0.08
0.096
0.103
0.12
0.15
0.188
0.258
0.323
0.389
0.442
0.451
0.451
0.451
0.451
O»KI
• 431
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451


Hew
Detect
Rate
-B5
-B5
-N6
-B8-B7
-B9-B8
-B10-B9
-B11-B10
-B12-B11
-B13-B12
-B14-B13
-B15-B144(B6-B5)
-B16-B154 (B7-B6)
-B17-B164(B8-B7)
-V9
-V10
-Vll
— tr| 9
"v±<
-VI 3
-VI 4
-V15
-VI 6
-VI 7
-VI 8
-VI 9
-V20
-V21
-AVERAGE (VS:V30)
Annual
After
-V5
-0.75*V6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0


Annual
average
-{W54V6)/2
-(H64V7)/2
- ** 1 f)
" \ H*. J. T V«i *. 1 / £
-(H224V23I/2
-(H234V24I/2
-(H244V25I/2
-(H254V26I/2
-(H264V27I/2
-(W274V28I/2
-(H264V29)/2
-(H294V301/2

Average
Annual
% benefit

-(5L6-X6|/((SL645L7| .21
-(SL7-X7|/((SL74$L8| '21
- (SL8-XB) / ( (SL84JL9I /2)
- (SL9-XSI / ( (SL94JLKM /Cl
- (SL10-X10) / ( (SL1045L11 ) i2
-(SL11-X11)/ ((SL1HSL12) .'2
-(SL12-X12) / ( (SL124SL13) 12
-(SL13-X13)/ (I5L134SL14I '2
-(SL14-X14)/ (1SL144SL15) /2
-(SL15-X1S)/ ((SL154SL16) '2
-(SL16-X16) / ( ($L164SL17) /2
-(5U7-X17) / ( (5L174$L1B) !2
-(SL18-X18)/ ((SL184SL1!-! i2
-(SL19-X19) / ( (SLlfuJLIfi) '2
- (5L20-X20) / ( (SL22
- (SL29-X29) / ( (SL294SL30) /2

-AVERAGE(V7:V29I
BIENNIAL
Age
Index
1
-14A59
-14A60
-14A61
-14A62
-14A63
-14A64
-14A65
-14A66
-14AS7
-14A68
-14A69
-14A70
-14A71
-1+A72
-14A73
-14A74
-14A75
-14A76
-14A77
-14A78
-14A79
-14A80
-14A81
-14A82


No-program
in-lane
fail rate
0.08
0.08
o.oa
0.096
0.103
0.12
0.15
0.168
0.258
0.323
0.389
0.442
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451
0.451


New Detect
1
-1-V56
Rate
-B59
-B59
-H60
-B62-B61
-B63-B62
-V634(B64-B63I
-B65-B64
-V654(B66-B65|
-B67-B66
-V674 (B68-B67) 4 (B60-BS9)
- (B69-B68) 4 (B60-B59)
-V64(B70-B69)4(B61-B60)
-(B71-B70)4(B62-B61)
-W714(V63-H63)
-H724(V64-H64)
-H734(V65-H65)
-H744(V66-H66)
-H75+(V«7-H67)
-H764(V68-H68)
-W774 (V69-H69I
-H784 (V70-H70)
-H794(V71-H71)
-H804(V72-H72)
-H814(V73-W73)
-H824 (\TJ4-W74)
-H834(V75-N75)
-AVERAGE (V59:V84I
Bien.
After
-V59
-0 . 25*V$57«V604 (0 . 75*V«0)
-V61«V557
-H614(V$57« (V62-H61) )
-V63
-W624(V$57*(V64-H62)>
-VS5
-H644(V$57«(H824V83)/2
-(H834V84I/2

Average
Bien.
% benefit
-(SL59-X59)/(($Lf-<'4SLCii) /;
- ($L60-X60) / ( ($L6(i4$LCl) /2
-(SL61-X61)/((SL6145L62) /2
-(SL62-X62) / ( ($L624$L63) /2
- (SL63-X63) / ( ($L634$L64 ) /2
- ($L64-X64) / ( ($L644$L£5) 12
- (SL65-X65) / ( ($L654$ie£) 12
- (SL66-X66I / ( ($L664$L67) /2
- (SI.67-X67) / ( ($L674$16B) /2
- (5L68-X68) / ( ($L684$LC9) 12
- (SL69-X69) / ( ($L«94$L70) /2
- (SL70-X70) / ( (5L704SL71 ) /2
-(SL71-X71) /( (SL714$L72) /2
- (SL72-X72) / ( (SL724SL73) /2
- ($173-X73( / ( ($L734$L74) /2
- (SL74-X74) / ( ($L744$L75) /2
- ($L75-X75| / ( ($L754$L76) /2
- (5L76-X76) / ( (SL764SL77) /2
- (5L77-X77) / ( (5L774SL78) /2
- ($L78-X78| / ( ($17845179) /2
- {$L79-X79) / ( ($L794$L80) /2
- (SL80-X80) / ( (SL804SLB1 ) /2
-(SL81-X81)/((SI.814$l82)/2
- (5182-X82) / ( (SL824$183) /2
- (SLB3-X83) / ( (SL634SL84 ) /2

-AVERAGE (V61:Y83)

-------
                       APPENDIX E









REGRESSION ANALYSES AND SCATTER PLOTS FOR FUEL-INJECTED




                1983 AND LATER VEHICLES

-------
                      Table  of  Contents
                          Appendix E
E-l     IM240 Vehicles Tested to Date               11/14/91

PFI Vehicles

E-2     Selecting One Score From the 2500/Idle Test

E-3     Regression Analyses of FTP HC versus HC Emissions Over
        IM240, Idle Test, and 2500/Idle Test for PFI Vehicles

E-4     HC Emissions - FTP vs Lane IM240

E-6     HC Emissions - FTP vs Lane 2500/Idle

E-8     HC Emissions - FTP vs Lane Idle

E-10    Regression Analyses of FTP CO versus CO Emissions
        Over: IM240, Idle Test, and 2500/Idle Test for PFI
        Vehicles

E-ll    CO Emissions - FTP vs Lane IM240

E-13    CO Emissions - FTP vs Lane 2500/Idle

E-15    CO Emissions - FTP vs Lane Idle

E-17    Regression Analyses of FTP NOx versus IM240 NOx
        Emissions

E-18    NOx Emissions - FTP vs Lane IM240

E-20    Vehicles Tested by ATL at Southbend, Indiana

E-21    HC Emissions - FTP vs Indolene-IM240

E-23    CO Emissions - FTP vs Indolene-IM240

E-25    NOx Emissions - FTP vs Indolene-IM240

E-27    EF Vehicles Tested at EPA's MVEL

E-28    HC Emissions - FTP vs Indolene-IM240

E-30    CO Emissions - FTP vs Indolene-IM240

E-32    NOx Emissions - FTP vs Indolene-IM240

E-34    Description of Repair Regressions Analysis

-------
E-35    Statistical Summaries for Variables Used in PFI Before
        and After Regression Analyses

E-36    Regression Analyses for the Change in Emission
        Constituents Before and After Repairs of the PFI
        Vehicles for the FTP vs IM240-A

E-37    EC Emissions - AFTP vs AIM240-A

E-39    CO Emissions - AFTP vs Alndolene-IM240

E-41    NOx Emissions - AFTP vs AIM240-A

E-43    Fuel Economy - AFTP vs AIM240-A

TBI Vehicles

E-45    Selecting One Score From the 2500/Idle Tests

E-46    Regression Analyses of FTP HC versus HC Emission Over:
        IM240f  Idle Test, and 2500/Idle Test for TBI Vehicles

E-47    HC Emissions - FTP vs Lane IM240

E-49    HC Emissions - FTP vs Lane 2500/Idle

E-51    HC Emissions - FTP vs Lane Idle

E-53    Regression Analyses of FTP CO versus CO Emissions
        Over: IM240, Idle Test, and 2500/Idle Test for TBI
        Vehicles

E-54    CO Emissions - FTP vs Lane IM240

E-56    CO Emissions - FTP vs Lane 2500/Idle

E-58    CO Emissions - FTP vs Lane Idle

E-60    Regression Analyses of FTP NOx versus IM240 NOx
        Emissions

E-61    NOx Emissions - FTP vs Lane IM240

E-63    Vehicles Tested by ATL at Southbend, Indiana

E-64    HC Emissions - FTP vs Indolene-IM240

E-66    CO Emissions - FTP vs Indolene-IM240

E-68    NOx Emissions - FTP vs Indolene-IM240

E-70    EF Vehicles Tested at EPA's MVEL
                           11

-------
E-77    Description of Repair Regression Analysis

E-78    Statistical Summaries for Variables Used in TBI
        Before and After Regression Analyses

E-79    Regression Analyses for the Change in Emission
        Constituents Before and After Repairs of the TBI
        Vehicles for the FTP vs IM240-A

E-80    HC Emissions - AFTP vs IM240-A

E-82    CO Emissions - AFTP - AIM240-A

E-84    NOx Emissions - AFTP vs AIM240-A

E-86    Fuel Economy - AFTP vs AIM240-A
                             iii

-------
                                                IM240 VEHICLES TESTED TO DATE
                                                                                     12/13/91
                        Test
                        FTP
  As Received
   In Micro
Indiana     Ann Arbor
              RM#1
              In Micro
          Indiana     Ann Arbor
  453
314
213
75
    RM#2
    In Micro
Indiana     Ann Arbor

  35          9
                                                   RM#3
                                                   In Micro
                                              Indiana      Ann Arbor

                                                 3           0
  LaneIM240
Tank Pud IM240
Diagnostic IM240
Non CUP IM240
  IM 240 Performed in Lab. Independent
  of all other tests.

Indolent CITP-A
IM240Lane
Change of Owner
MISC Repeat
Tank Fuel IM240
After Exhust Repair
7313
6
14
435
15
NA
NA
NA
0
0
73
0
0
0
0
NA
NA
NA
0
0
10
0
0
0
0
NA
NA
NA
0
0
2
0
0
0
0
NA
NA
NA
0
0
Indolene CITP-B
                                        NA
              NA
              267
            14
                        70
IM240-A (A)
CDH226-A (B)
Steady State- A
IM240-B (D)
CDH226-B (C)
Steady State-B
378
369
0
365
366
0
30
31
0
31
31
0
188
180
0
180
179
0
15
15
0
15
15
0
29
29
0
29
29
0
1
1
0
1
1
0
3
3
0
3
3
0
0
0
0
0
0
0

-------
Regression Analyses - Port Fuel-Injected  Vehicle Sample:

       As part of EPA's regression analysis, FTP results for HC and CO emissions were
each regressed against the corresponding emission results on the Lane IM240, the second
chance 2500/Idle Test, and the Idle Test. Regressions were also performed comparing
NOx results for the FTP and Lane IM240 (NOx emissions are not measured as part of
either Idle Test and are therefore not included in this comparison). Lasdy, FTP results for
HC, CO, and NOx were each regressed against corresponding emissions from Lab IM240s
run on Indolene fuel at the ATL lab in South Bend and at MVEL in Ann Arbor. This
section provides the results of these regressions performed on a sample of port fuel injected
(PFI) vehicles recruited as part of the Hammond study.

       EPA removed thirteen vehicles from its sample of PFI vehicles.  Three were
removed because they received repairs for exhaust leaks during the interim between
receiving an I/M test at the Hammond lane and being FTP tested at the lab. The ten
remaining PFI vehicles were excluded due to inconsistent dynamometer settings for the
Lane IM240 and the FTP. Vehicles were only removed if the horsepower or inertia weight
settings were at least 15% different between the lab and the lane.

       These differences in dynamometer settings arose from the use of different
dynamometer look-up tables at die lane and lab. Traditional certification test car lists used
for emission factor testing are too detailed for efficient use in the I/M lane. After becoming
aware of the differences in the two tables, EPA opted to use the look-up table that had been
simplified for lane use at both testing sites to avoid future inconsistencies. The resulting
sample consisted of 74  '83 and newer PFI vehicles.


Selecting  One  Score From the Second Chance  2500/Idle Test:

       To illustrate the correlation of FTP HC and CO results to the  corresponding results
on the second chance 2500/Idle Test, we had to select an HC score and CO score from the
first chance or second chance test, and from the 2500 rpm mode or the idle mode.
Concerning the first chance versus second chance results, we chose the better results
relative to die HC/CO outpoint being evaluated.

For the 2500 rpm versus idle modes, we chose the larger of the  corresponding emissions
measured.  This approach is based on the assumption mat the same outpoints would be
used on each of die modes. Hence, a vehicle would pass die two mode test if and only if
the higher of its HC scores (whedier Idle or 2500 rpm) met me HC cutpoint (with the same
also following for die CO results).


Graphing the Data:

       On each of die graphs comparing HC emissions, the three horizontal dotted lines
are positioned at the boundaries separating:

   •   die normal emitters from the high emitters (0.82 g/mi).
       the high emitters from the very high emitters (1.64 g/mi), and
   •   the very high emitters from the super emitters (10.00 g/mi).
                                     E-2A

-------
       Similarly, on each of the graphs comparing CO emissions, the three horizontal
dotted lines are positioned at the boundaries separating:

   *   the normal emitters from the high emitters (10.2 g/mi),
   •   the high emitters from the very high emitters (13.6 g/mi), and
   •   the very high emitters from the super emitters (150.0 g/mi).
                                     E-2B

-------
Regression  Analyses of FTP HC versus  HC Emissions Over:
          IM240,  Idle  Test, and 2500/Idle  Test
                    For  PFI Vehicles
Dependent variable is: HCFTP
RA2 = 80.3%
s = 1.016 with
Source
Regression
Residual
Variable
Constant
HCIM24
74 - 2 » 72 degrees of freedom
Sum of Squares df
302.637 1
74.3928 72
Coefficient s.e. of Coeff
0.198655 0.1372
1.13037 0.066

Mean Square
303
1.03323
t-ratio
1.45
17.1

F-ratlo
293



Dependent variable is:
RA2 = 39.3%
s = 1.784
Source
Regression
Residual
Variable
Constant
Higherddle
HCFTP


with 74 - 2 = 72 degrees of freedom
Sum of Squares
148.004
229.026
Coefficient
0.692609
HC... 0.003966
df
1
72
s.e. of Coeff
0.2313
0.0006
Mean Square
148
3.18092
t-ratlo
2.99
6.82
F-ratlo
46.5



Dependent variable is: HCFTP
RA2 = 44.3%
s = 1.709 with 74 - 2 a 72 degrees of freedom
Source
Regression
Residual
Variable
Constant
Idle.HC
Sum of Squares
166.842
210.188
Coefficient
0.706016
0.004541
df
1
72
s.e. of Coeff
0.2184
0.0006
Mean Square
167
2.91927
t-ratlo
3.23
7.56
F-ratlo
57.2

                           E-3

-------
HC Emissions - FTP vs Lane IM240
 (74 1983 & Newer PFI  Vehicles)
                                     Regression Line
                                      RA2 = 80.3%
                 6              8
       IM240  HC Emissions (g/ml)
10
12
               E-4

-------
FTPHC
(g/ml)

 O ...I.........
                         HC Emissions -- FTP vs Lane IM240
                           (74 1983  & Newer PFI Vehicles)
                        Enlarged for Better  Detail near Origin
2.5 -J-
 2 -r
1.5 T
 1 -r
                                                      Regresslon Line
                                                       RA2 = 80.3%
0.5 -
   **»•
 0
               0.5
1           1.5           2

    IM240 HC Emissions (g/ml)
2.5
                                      E-5

-------
                             HC Emissions -  FTP vs Lane 2500/ldle
                                (74  1983 &  Newer PFI Vehicles)
FTPHC
(g/ml)
 12
 10 —
  8 --
  6 -r
  4 --
  2 --
  0 ^
                                                              Regression Line
                                                                R*2 a 39.3%
     0       200      400      600      800     1000     1200     1400     1600     1800     2000
                                     2500/ldle HC Emissions (ppm)
                                               E-6

-------
FTPHC
(9/ml)
  3  ~
 2.5  -
  2  -r
 1.5  T
                              HC Emissions -- FTP  vs Lane 2500/ldle
                                  (74 1983  &  Newer PFI  Vehicles)
                              Enlarged for Better Detail  near Origin
                                                                    Regression Line
                                                                      R*2 = 39.3%
                   I
50
 100               150
2500/ldle HC Emissions (ppm)
200
                                                                                             250
                                                E-7

-------
FTPHC
(9/ml)
 12 -r
                                HC Emissions - FTP vs Lane Idle
                                (74 1983  & Newer PFI Vehicles)
 10 -L-
  8 -
  6 -
                                                                  Regression Line
                                                                    R*2 = 44.3%
  0 •

    0
200      400     600
800     1000     1200
 Idle HC Emissions (ppm)
1400     1600     1800    2000
                                               E-8

-------
FTPHC
(g/ml)
  3 -
                                 HC Emissions - FTP vs Lane Idle
                                 (74 1983 &  Newer PFI Vehicles)
                              Enlarged  for  Better Detail near Origin
 1.5 -
   1  -
 2 5 - -                                 •                         Regression Line
                         •                                          RA2 = 44.3%
  2 -
 0.5 t_
      k  m        .        .

                       •
                      50               100               150               200               250
                                        Idle HC Emissions (ppm)
                                               E-9

-------
Regression Analyses of  FTP CO versus CO Emissions Over:
          IM240,  Idle  Test,  and  2500/Idle  Test
                     For PFI  Vehicles
Dependent variable is: COFTP
RA2 m 90.4%
s = 17.10 with 74 - 2 = 72 degrees of freedom
Source
Regression
Residual
Variable
Constant
COIM24
Sum of Squares
199166
21052.3
Coefficient
0.446807
1.05072
df
1
72
s.e. of Coeff
2.217
0.0403
Mean Square
200000
292.392
t-ratio
0.202
26.1
F-ratio
681

Dependent variable is: COFTP
RA2 = 76.3%
s » 26.91 with 74 - 2 s 72 degrees of freedom
Source
Regression
Residual
Variable
Constant
Hlgherddle
Sum of Squares
168093
52125.7
Coefficient
4.01842
CO... 23.6992
df
1
72
s.e. of Coeff
3.446
1.555
Mean Square
200000
723.968
t-ratlo
1.17
15.2
F-ratlo
232

Dependent variable is:
RA2 = 61.8%
s m 34.19 with
Source
Regression
Residual
Variable
Constant
Idle.CO
74 - 2 = 72 degi
Sum of Squares
136071
84147.5
Coefficient
8.59165
27.9077
COFTP
ees of freedom
df
1
72
s.e. of Coeff
4.291
2.586


Mean Square
100000
1168.71
t-ratlo
2
10.8


F-ratlo
116



                           E-10

-------
FTP CO
(g/ml)

250  -
                               CO Emissions - FTP vs Lane IM240
                                (74 1983  & Newer PFI Vehicles)
200  —
150  -T--
 100  -1-
  50 -
                                      Regression Line
                                        R*2 s 90.4%
                      50
100              150
IM240 CO Emissions (g/ml)
200
250
                                             E-11

-------
                               CO Emissions - FTP vs Lane IM240
                                (74 1983 & Newer PFI  Vehicles)
Frpco                        Enlarged  for Better Detail  near Origin
(g/ml)
 50 —                                            Regression Line
                                    •              R*2 B 90.4%
 40 —
 30 ^
 20
 10
                    • •         •
             _,	:	1	.	j	i
             5        10       15       20      25       30       35      40       45       50
                                      IM240 CO Emissions (g/ml)
                                              E-12

-------
FTP CO
(g/ml)
250 -
                              CO Emissions  - FTP vs  Lane  2500/ldle
                                  (74 1983  & Newer PFI Vehicles)
                                                        Regression Line
                                                         R*2 = 76.3%
200 --
150 -»-•
100 —
 50  --
                                     2500/ldle CO Emissions (percent)
                                               E-13

-------
FTP CO
(g/ml)
50
                              CO Emissions  -- FTP vs  Lane 2500/ldle
                                  (74 1983 &  Newer PFI  Vehicles)
                              Enlarged for Better Detail  near Origin
40 -1-
30 --
20 -
                                                                                   Regression Line
                                                                                     R*2 = 76.3% '
10 --
0 -f

  0
            0.2
0.4
0.6       0.8       1        1.2
      2500/ldle CO Emissions (percent)
1.4
1.6
1.8
                                                E-14

-------
FTP CO
(g/ml)
250 -
200 --
150 -1-
100 T
 50 -
                                 CO Emissions - FTP vs  Lane Idle
                                  (74 1983  & Newer  PFI Vehicles)
                                                Regression Line
                                                  RA2 = 61.8%
                                             4         5
                                        Idle CO Emissions (percent)
                                               E-15

-------
FTP CO
(g/ml)
50 -
                               CO Emissions - FTP vs Lane Idle
                               (74 1983  & Newer PFI Vehicles)
                            Enlarged for Better  Detail near Origin
40 T
30 --
                                                             Regression Line
                                                               R*2 = 61.6%
20 4-
10
     K1    -"
.Oir
                                                                   —i—

                                                                   1.5
            0.25
0.5
0.75         1          1.25

  Idle CO Emissions (percent)
1.75
—i

 2
                                            E-16

-------
Regression Analyses of FTP NOx versus IM240 NOx Emissions
        (Note:  NOx is not Measured for the Idle Test)
Dependent variable is:
RA2 = 60.3%
s = 0.3277
with 74 - 2 =
Source Sum of Squares
Regressior 11.758
Residual 7.73394
Variable
Constant
NOIM24
Coefficient
0.259283
0.427093
NOFTP
72 degrees of
df
1
72
s.e. of Coeff
0.0587
0.0408

freedom
Mean Square
11.76
0.107416
t-ratlo
4.41
10.5


F-ratlo
109



                            E-17

-------
FTPNOx
 (g/ml)

  3 -r
                                NOx Emissions - FTP  vs Lane  IM240
                                  (74 1983 & Newer  PFI Vehicles)
 2.5  --
  2  --
 1.5  T
  1  -r
 0.5  --
                                                                   Regression Line
                                                                     R*2 s 60.3%
  o  -
             0.5
1       1.5
  2        2.5        3
IM240 NOx Emissions (g/ml)
3.5
4.5
                                                E-18

-------
FTPNOx
 (g/mi)

  1.5 -
                                NOx Emissions - FTP vs Lane IM240
                                  (74 1983 &  Newer PFI Vehicles)
                               Enlarged for Better Detail near  Origin
 1.25 --
   1 --
 0.75 -
  0.5 -
 0.25 -"
                                                                                Regression Line
                                                                                 R*2 8 60.3%
                                                                                              H

                                                                                               2
0.25
0.5
0.75         1         1.25
  IM240 NOx Emissions (g/ml)
1.5
1.75
                                               E-19

-------
Vehicles Tested  by ATL at Southbend,  Indiana
  Regressions of FTP vs lndolene-IM240
     for 63 1983 & Newer PFI Vehicles
Dependent variable is HCFTP
RA2 • 89.2%
S = 0.8037 with
Source
Regression
Residual
Variable
Constant
HC.IM24.lndo...
63 - 2 a 61 degrees
Sum of Squares
325.428
39.4065
Coefficient
0.43699
1.1596
of freedom
df
1
61
s.e. of Coeff
0.1132
0.0517

Mean Square
325
0.646008
t-ratio
3.86
22.4

F-ratio
504



Dependent variable is
RA2 . 90.3%
S = 18.57 with 63 -
Source
Regression
Residual
Variable
Constant
CO.IM24.lndo...
COFTP
2 s 61 degrees of
Sum of Squares
195708
21030.6
Coefficient
5.6588
0.991509
freedom
df
1
61
s.e. of Coeff
2.571
0.0416

Mean Square
200000
344.764
t-ratlo
2.2
23.8

F-ratlo
568

Dependent variable is
RA2»89.1%
s = 0.1780 with 63
Source
Regression
Residual
Variable
Constant
NOx.IM24.lnd...
NOX.FTP
- 2 = 61 degrees
Sum of Squares
15.8477
1.93337
Coefficient
0.145473
0.757394
of freedom
df
1
61
s.e. of Coeff
0.0356
0.0339

Mean Square
15.85
0.031695
t-ratlo
4.09
22.4

F-ratio
500

                              E-20

-------
 FTPHC
 (g/ml)
 15  -
12.5 --
 10 --
 7.5
  5  -
 2.5 -J-
   0  1

     0
                            HC Emissions  -  FTP  vs lndolene-IM240
                                 (63 1983  &  Newer  PFI Vehicles)
                               Regression Line
                                R*2 = 89.2%
-*"•
     2.5
7.5
10
                                       IM240 HC Emissions (g/ml)
                                              E-21

-------
FTPHC
(g/ml)
 3  -
2.5  ->-
 2  --
  1  -
0.5  J-.
V"*-/
                             HC Emissions  -  FTP vs lndolene-IM240
                                 (63 1983  &  Newer PFI Vehicles)
                              Enlarged for  Better Detail  near Origin
                                                     Regression Line
 0  -1	i	;	i	;	••	1	i

    0              0.5             1              1.5             2              2.5             3

                                       IM240 HC Emissions (g/ml)
                                              E-22

-------
 FTP CO
 (g/ml)
250 -
200
150 --
100 —
 50 -
                           CO Emissions  -  FTP vs lndolene-IM240
                                (63 1983 & Newer PFI Vehicles)
                     50
Regression Une
  R*2 s 90.3%
100              150
IM240 CO Emissions (g/ml)
200
250
                                             E-23

-------
                            CO Emissions  -  FTP vs lndolene-IM240
                                (63  1983 & Newer PFI Vehicles)
                              Enlarged for Better Detail  near Origin
FTP CO
(g/ml)
50 --


45 --


40 --


35 --


30 --
                                         . • •'                         Regression Line
25 —                                 .--"'                             R*2s90.3%

                               .-''                         •
20 --                       .--


15--              ..--'•

.10 --


 5
 0 -=	~	i	r
                     10       15       20       25      30      35       40       45       50
                                      IM240 CO Emissions (g/ml)
                                              E-24

-------
 FTPNOx
 (g/ml)
4.00 —
                             NOx Emissions  -   FTP  vs lndolene-IM240
                                  (63 1983  & Newer PFI Vehicles)
3.00 —
2.00 -
1.00 -T-
                                                     Regression Line
                                                       R*2 s 89.1%
                                     •H      •
0.00
    0.00
1.00
         2.00

IM240 NOx Emissions (g/ml)
3.00
4.00
                                               E-25

-------
 FTPNOx
 (g/ml)

2.00 -
1.75 --


1.50 --


1.25 --


1.00


0.75 --


0.50 --


0.25 -
o.oo -I
                             NOx Emissions  -  FTP vs lndolene-IM240
                                  (63 1983 &  Newer PFI Vehicles)
                               Enlarged for Better Detail near  Origin
                                     Regression Line
                                      R*2 = 89.1%
                      II
    0.00
0.25
0.50
0.75       1.00        1.25
  IM240 NOx Emissions (g/ml)
1.50
1.75
2.00
                                               E-26

-------
EF Vehicles Tested at EPA's MVEL
 Regressions of FTP vs  Indolene IM240
      for  143 1985-90 PFI Vehicles
Dependent variable is:
R*2 = 95.7%
HC.FTP


s = 0.5416 with 143 - 2 = 141 degrees of freedom
Source
Regression
Residual
Variable
Constant
HCIM24
Sum of Squares
927.841
41.3557
Coefficient
0.088877
1.91326
df
1
141
s.e. of Coeff
0.0472
0.034
Mean Square
928
0.293303
t-ratlo
1.88
56.2
F-ratlo
3163



Dependent variable is: CO.FTP
RA2 = 95.4%
s • 5.335 with 143 - 2 = 141 degrees of freedom
Source
Regression
Residual
Variable
Constant
COIM24
Sum of Squares
83635
4012.44
Coefficient
1.68729
1.14873
df
1
141
s.e. of Coeff
0.4594
0.0212
Mean Square
83635
28.457
t-ratlo
3.67
54.2
F-ratlo
2939

Dependent variable is: NOx.FTP
RA2 = 90.5%
s ^ 0.2012 with 143 - 2 » 141 degrees of freedom
Source
Regression
Residual
Variable
Constant
NOIM24
Sum of Squares
54.6898
5.70772
Coefficient
0.10242
0.81782
df
1
141
s.e. of Coeff
0.0237
0.0222
Mean Square
54.69
0.04048
t-ratlo
4.33
36.8
F-ratio
1351

             E-27

-------
FTPHC
(9/ml)

30.00 --
20.00  --
10.00  --
                             HC Emissions  -  FTP vs lndolene-IM240
                                   (143  1985-90  PFI  Vehicles)
                                 (Tested at EPA's Ann Arbor Lab)
                                  Regression Line
                                   B*2 = 95.7%
 0.00  *
     0.00
5.00                         10.00
      IM240 HC Emissions (g/ml)
15.00
                                              E-28

-------
FTPHC
(g/ml)
3.00 -
                             HC Emissions  -   FTP vs lndolene-IM240
                                    (143 1985-90  PFI   Vehicles)
                                  (Tested at  EPA's Ann Arbor Lab)
                               Enlarged for  Better Detail near  Origin
2.00 T
 1.00 --
 0.00
                                                Regression Line
                                                  B*2 a 95.7%
    0.00
0.50
1.00            1.50           2.00
      IM240 HC Emissions (g/ml)
2.50
                                                                          3.00
                                               E-29

-------
FTP CO
(g/ml)
300.00 -
250.00 --
200.00 -
150.00
100.00 -
 50.00 -r
                             CO Emissions  -  FTP vs lndolene-IM240
                                    (143  1985-90  PFI  Vehicles)
                                  (Tested at EPA's Ann Arbor Lab)
                      Regression Line
                       R*2 = 95.4%
  0.00 •

      0.00
50.00
100.00            150.00
 IM240 CO Emissions (g/ml)
200.00
250.00
                                               E-30

-------
FTP CO
(a/ml)
50.00 -
40.00 -
                             CO Emissions  --   FTP vs lndolene-IM240
                                    (143 1985-90  PFI  Vehicles)
                                  (Tested at EPA's Ann Arbor Lab)
                               Enlarged for Better Detail near  Origin
                                      Regression Une
                                       R*2 = 95.4%
30.00  -
20.00 -
10.00
 0.00 ^

     0.00
10.00
20.00             30.00
IM240 CO Emissions (g/ml)
40.00
50.00
                                              E-31

-------
FTPNOx
(g/ml)
                             NOx Emissions  -  FTP vs  lndolene-IM240
                                    (143  1985-90  PFI  Vehicles)
                                  (Tested at EPA's Ann Arbor Lab)
4.00 -1-
                               Regression Line
                                R*2 s 90.5%
3.00 --
 2.00 -
 1.00 4
0.00 -
    0.00
1.00
2.00              3.00
IM240 NOx Emissions (g/ml)
4.00
5.00
                                               E-32

-------
                            NOx  Emissions   -  FTP vs lndolene-IM240
                                    (143 1985-90 PFI Vehicles)
                                  (Tested at EPA's Ann Arbor Lab)
                               Enlarged for Better Detail near Origin
FTP NOx
(g/ml)
2.00 -


                                                     Regression Line
                                                       RA2 = 90.5%

1.50 -

 0.50 -
 0.00 -
    0.00                  0.50                  1.00                   1.50                   2.00

                                       IM240 NOx Emissions (g/ml)
                                              E-33

-------
                 Description  of Repair  Regression Analysis

Regression Analyses

      The difference in emission levels before and after repairs for the FTP's and Lab
IM240's run on Indolene fuel was calculated such that a positive difference would
correspond to a decrease in emissions or an increase in fuel economy. The AFTP values
(AHC, AGO, ANOx and AMPG) were regressed against the corresponding Indolene
AIM240 values. The results and graphs of these regressions are included.

Data Description

      The data used in this analysis consisted of FTP and Indolene IM240 scores for tests
performed before and after repairs. These results consisted of HC, CO, NOx, and fuel
economy values. The database was restricted to 1983 and newer PFI and TBI vehicles
which received repairs and whose as received FTP scores exceeded twice the FTP standard
for either HC or CO.  Also excluded from the analysis were vehicles which received repairs
before the as received FTP and vehicles which had differences in dynamometer settings of
greater than 15%. The resulting data set contained 41 TBI vehicles and 23 PFI vehicles.
Summary Statistics for the variables used in the PFI analyses follow.
                                    E-34

-------
                Statistical Summaries for Variables Used in
                PFI Before and After Regression  Analyses
Summary statistics for: AHC.IM240
NumNumeric = 23
Mean =1.7848
Median = 0.59000
Midrange = 3.7950
Standard Deviation = 2.3719
Range = 8.2500
Minimum = -0.33000
Maximum = 7.9200
Summary statistics for  AHC.FTP
NumNumeric = 23
Mean =2.4678
Median =1.9700
Midrange = 4.2600
Standard Deviation = 2.9115
Range = 9.0800
Minimum = -0.28000
Maximum = 8.8000
Summary statistics for: ACO.IM240
NumNumeric = 23
Mean = 53.061
Median = 9.6200
Midrange = 119.73
Standard Deviation = 75.023
Range = 248.85
Minimum = -4.6900
Maximum = 244.16
Summary statistics for ACO.FTP
NumNumeric = 23
Mean = 60.184
Median =14.580
Midrange = 115.89
Standard Deviation = 74.656
Range = 239.46
Minimum = -3.8400
Maximum = 235.62
Summary statistics for: ANOX.IM240
NumNumeric = 23
Mean = -0.36174
Median = -0.23000
Midrange = -0.90000
Standard Deviation = 0.52710
Range = 2.3000
Minimum = -2.0500
Maximum = 0.25000
Summary statistics for ANOX.FTP
NumNumeric = 23
Mean = -0.29522
Median =-0.15000
Midrange = -0.78000
Standard Deviation = 0.48197
Range = 2.2800
Minimum = -1.9200
Maximum = 0.36000
 Summary statistics for: AMPG.IM240
 NumNumeric = 23
 Mean = 2.5904
 Median =1.5500
 Midrange = 3.2800
 Standard Deviation = 2.9593
 Range = 10.780
 Minimum = -2.1100
 Maximum = 8.6700
Summary statistics for AMPG.FTP
NumNumeric = 23
Mean =2.5183
Median =1.7100
Midrange = 5.3200
Standard Deviation = 3.5426
Range = 17.720
Minimum = -3.5400
Maximum = 14.180
                                  E-35

-------
    Regression  Analyses for the Change in Emission Constituents
Before and After Repairs of the PR Vehicles for the FTP vs IM240-A
Dependent variable is:
RA2 = 90.8%
s = 0.9031
Source
Regression
Residual
Variable
Constant
AHC.IM240

with 23 - 2 = 21
Sum of Squares
169.361
17.1275
Coefficient
0.380093
1.16974
AHC.FTP





degrees of freedom
df
1
21
s.e. of Coeff
0.2376
0.0812
Mean Square
169
0.815595
t-ratio
1.6
14.4
F-ratlo
208




Dependent variable is: ACO.FTP
RA2 = 88.4%
s = 26.01 with 23 - 2 a 21 degrees of freedom
Source Sum of Squares df Mean Square
Regression 108409 1 100000
Residual 14207.5 21 676.547
Variable Coefficient s.e. of Coeff t-ratio
Constant 10.5355 6.693 1.57
ACO.IM240 0.93568 0.0739 12.7
F-ratlo
160
Dependent variable is:
RA2 = 71.2%
s = 0.2649
Source
Regression
Residual
Variable
Constant
ANOX.IM240
ANOX.FTP
with 23 - 2 s 21 degrees of freedom
Sum of Squares df Mean Square
363679 1 3.6368
1.47378 21 0.07018
Coefficient
-0.016187
0.771357
s.e. of Coeff t-ratio
0.0675 -0.24
0.1072 7.2

F-ratio
51.8



Dependent variable is:
RA2 = 36.7%

AMPG.FTP





s = 2.885 with 23 - 2 = 21 degrees of freedom
Source
Regression
Residual
Variable
Constant
AMPG.IM240
Sum of Squares
101.306
174.8
Coefficient
0.639853
0.725132
df
1
21
s.e. of Coeff
0.8074
0.2079
Mean Square
101
8.32381
t-ratlo
0.793
3.49
F-ratlo
12.2




                                E-36

-------
                            HC Emissions - AFTP vs Alndolene-IM240
               (23  1983 & Newer PFI Vehicles with Repairs and FTP > 2*(FTP.Stnd))
 A FTP
HC (g/ml)
10 -r
           8 --
           6 --
           4 -r
           n	
                                                                        Regression Line
                                                                         R*2 = 90.8%
1 	 • *'
1 (
. 9 -
I
1
1 i
234
1 1
5 6
7 £
                                          A IM240 HC (g/ml)
                                               E-37

-------
  A FTP

HC (g/nti)
                HC Emissions - AFTP vs  Alndolene-IM240


 (23  1983  & Newer  PFI Vehicles  with Repairs and  FTP > 2*(FTP.Stnd))







 4  -






3.5  --

                                                                         s*
                                                                       ^
                                                                      ^
                                                                    ^
                                                                   ^

 3  --






2.5  -                                             ^



         »                 •                 ^^    •


 2  -               .                  ^

                                      s*
                                    x'
                                   X*



1.5  -                        /""

                          ^ ^                                   Regression Line

                      ^'                                       R*2 a 90.8%
              0.5 --
   -0.5
             -0.5
                 0.5
1.5
                                             A IM240 HC (g/ml)
2.5
                                                                     Enlarged for Better Detail near Origin
                                                   E-3B

-------
  A FTP
CO (g/mi)
 -50
                        CO Emissions  -  AFTP  vs Alndolene-IM240
            (23 1983  & Newer PFI  Vehicles with Repairs and FTP > 2*FTP.Stnd)
250 —
            200 -
            150 -
            100 -
             50 -
    0

     j
-50 -
50
                                                                       Regression Line
                                                                        R*2 = 88.4%
100
150
200
250
                                      A IM240 CO (g/ml)
                                           E-39

-------
                         CO Emissions  -  AFTP vs Alndolene-IM240
             (23 1983 & Newer PFI Vehicles with Repairs and FTP > 2*FTP.Stnd)
  A FTP
CO (g/ml)
100 -
             80  -
             60  -
             40  -
             20  -r
                                                              Regression Line
                                                               R*2 = 88.4%
1
20



	 mo — _ 	
6
i
i
i
-20 -
	 1 	 • 	 r~~ 	 —
20 40 60 80



	 1
1C



                                         A IM240 CO (g/ml)
                                                                Enlarged for Better Detail near Origin
                                              E-40

-------
                      NOx Emissions   -  AFTP vs Alndolene-IM240
          (23 1983  & Newer PFI Vehicles with Repairs  and FTP > 2*FTP.Stnd)
 A IM240
NOx (g/mi)
                                                            A FTP
                                                          NOx (g/ml)
                                                                      0.5 -T-
-2.5
             -2
    Regression Line
     R*2 = 71.2%
-1.5
-1
    •   ••  •
-0.5 I    _^
                                                                     -0.5 -r
                                                                      -1 —
                                                                     -1.5
0.5
                                                                      -2 —
                                          E-41

-------
                         NOx  Emissions  -  AFTP vs Alndolene-IM240
             (23 1983 & Newer PFI  Vehicles with Repairs and FTP > 2*FTP.Stnd)
                                                         0.4 -
                                                         0.2 --
    A  IM240
   NOx (g/ml)
                                                                  A FTP
                                                                NOx (g/ml)
-0.5
-0.4
-0.3i
-0.2     •   -0,4-'
0.1
0.2
0.3
  Regression Line
    RA2 = 71.2%
                                                        -0.2 —
                                                        -0.4 -r
                                                        -0.6 —
                                                        -0.8 •+•
                                                                Enlarged for Better Detail near Origin
                                               E-42

-------
                        Fuel Economy  -  AFTP vs Alndolene-IM240
            (23 1983 & Newer  PFI  Vehicles with  Repairs and FTP > 2*FTP.Stnd)
          AFTPUPG
           (ml/gal)
          15 -
                       10 —
                                                        Regression Line
                                                          B*2 = 36.7%
                        5  ~~
                                « JHT
-4
'-2
8
10
                       -5 -
                                                         A IM240 MPG (ml/gal)
                                           E-43

-------
           Fuel Economy  -  AFTP vs Alndolene-IM240
(23 1983 & Newer PFI Vehicles with Repairs and FTP > 2*FTP.Stnd)
A FTP MPG
(ml/gal) 5
4 -
3 -
2 -
1 -
! — "~ A
,-'"" .
3 -2^-" -1 (
^-^ -1 -
"" '
-3 -
-4 -
-5 -
Regression Line
R*2 = 36.7% • ^^-'
^--^
• jn"
-""" ^1 *
— —
'" • *
i I 1 i <
i 1 2345
"
A IM240 MPG (mi/gal)

"
Enlarged for Better Detail near Origin
                             E-44

-------
Regression Analyses - Throttle-Body Injected Vehicle  Sample:

       As part of EPA's regression analysis, FTP results for HC and CO emissions were
each regressed against the corresponding emission results on the Lane IM240, the second
chance 2500/Idle Test, and the second chance Idle Test. Regressions were also performed
comparing NOx results for the FTP and Lane IM240 (NOx emissions are not measured as
part of either idle test and are therefore not included in this comparison). Lastly, FTP
results for HC, CO, and NOx were each regressed against corresponding emissions from
Lab IM240s run on Indolene fuel at the ATL lab in South Bend and at MVEL in Ann
Arbor. This section provides the results of these regressions performed on a sample of
throttle-body injected (TBI) vehicles recruited as part of the Hammond study.

       EPA removed a total of twenty-two vehicles from its sample of TBI vehicles.  Sk
were removed because of repairs performed after their I/M test but before their FTPs, all
but two of which were exhaust leak repairs. Of the remaining two pre-FTP repaired
vehicles, one received a new fuel pump and the other received an alternator and battery
replacement.  In addition to these six pre-FTP repair exclusions, sixteen TBI vehicles were
removed from the analyses due to differences of 15% or greater between lane and lab
dynamometer settings. The resulting database consisted of 108 TBI vehicles.
Selecting One Score From the Second Chance 2500/Idle Test:

       To illustrate the correlation of FTP HC and CO results to the corresponding results
on the second chance 2500/Idle Test, we had to select an HC score and CO score from the
first chance or second chance test, and from the 2500 rpm mode or the idle mode.
Concerning the first chance versus second chance results, we chose the better results
relative to the HC/CO cutpoint being evaluated.

For the 2500 rpm versus idle modes, we chose the larger of the corresponding emissions
measured. This approach is based on the assumption mat the same outpoints would be
used on each of the modes. Hence, a vehicle would pass the two mode test if and only if
die higher of its HC scores (whether Idle or 2500 rpm) met the HC cutpoint (with the same
also following for the CO results).


Graphing the Data:

       On each of the graphs comparing HC emissions, the three horizontal dotted lines
are positioned at the boundaries separating:

   •   the normal emitters from die high emitters (0.82 g/mi),
   •   die high emitters from the very high emitters (1.64 g/mi), and
   •   the very high emitters from the super emitters (10.00 g/mi).

Similarly, on each of die graphs comparing CO emissions, the three horizontal  dotted lines
are positioned at the boundaries separating:

       the normal emitters from the high emitters (10.2 g/mi).
   •   the high emitters from the very high emitters (13.6 g/mi), and
   •   the very high emitters from the super emitters (150.0 g/mi).
                                     E-45

-------
 Regression  Analyses of FTP  HC versus HC  Emissions Over:
            IM240, Idle  Test, and 2500/Idle  Test
                       For  TBI  Vehicles
Dependent variable is:
RA2 m 64.4%
s = 2.397 with 108 - 2 . 106
Source
Regression
Residual
Variable
Constant
HCIM24
Sum of Squares
1102.31
609.049
Coefficient
-0.149912
1.45609
HCFTP
degrees of freedom
df
1
106
s.e. of Coeff
0.261
0.1051
Mean Square
1102
5.74575
t-ratlo
-0.574
13.9
F-ratio
192

Dependent variable is:        HCFTP
RA2 = 8.4%
s a 3.845 with  108 - 2 = 106  degrees of freedom
Source      Sum of Squares     df
Regression       143.863          1
Residual         1567.5         106
Variable      Coefficient   s.e. of Coeff
Constant        0.89518       0.4242
Hlqherddle  ...   0.003197	0.001
Mean Square  F-ratio
    144       9.73
  14.7877

  t-ratlo
    2.11
    3.12
Dependent variable is:
RA2 = 14.0%
s = 3.727 with 108 - 2 = 106
Source
Regression
Residual
Variable
Constant
Idle.HC
Sum of Squares
238.906
1472.46
Coefficient
0.651022
0.005218
HCFTP
degrees of freedom
df
1
106
s.e. of Coeff
0.418
0.0013
Mean Square
239
13.8911
t-ratlo
1.56
4.15
F-ratlo
17.2

                              E-46

-------
 FTPHC
 (9/ml)


35 -
HC Emissions  -  FTP vs Lane  IM240
  (108  1983 &  Newer TBI Vehicles)
30 --
25 4-
                                            Regresslon Line
                                             R*2 = 64.4%
20
15 -r
10 --
 5 -1-
                                         6            8
                                      IM240 HC Emissions (g/ml)
                                    10
12
14
                                             E-47

-------
FTPHC
(g/ml)
 5  -T


4.5


 4  --


3.5  --


 3


2.5  --


 2  -


1.5  --


 1  --


0.5  --


 0
                               HC Emissions  -  FTP vs Lane IM240
                                 (108 1983 &  Newer TBI Vehicles)
                               Enlarged for  Better  Detail near Origin
                  m	•.
                                                                                Regression  Line
                                                                                     = 64.4%
             0.5
                                      1.5          2          2.5
                                       IM240 HC Emissions (g/ml)
3.5
                                              E 48

-------
  FTPHC
  (g/ml)

35 -
30 -
25 -
20 —
                           HC Emissions  -  FTP vs Lane  2500/ldle
                               (108 1983  & Newer TBI  Vehicles)
                                                                  Regression Line
                                                                    R*2 = 8.2%
15 -
10	
    0
500
         1000
2500/ldle HC Emissions (ppm)
1500
2000
                                             E-49

-------
  FTPHC
  (g/ml)

  5  --


4.5  --


  4  --


3.5  --


  3  --


2.5  --


  2  --


1.5  --


  1  -?_-


0.5


  0
      HC Emissions  -   FTP vs Lane 2500/ldle
         (108 1983 & Newer  TBI Vehicles)
       Enlarged  for Better Detail near Origin
                                                   Regression I
—r

 50
100
                                 150       200       250       300
                                     2500/ldle  HC  Emissions (ppm)
350
400
450
                       E-50

-------
  FTPHC
  (g/ml)
35 -
                               HC Emissions  -  FTP vs Lane Idle
                               (108  1983 & Newer TBI  Vehicles)
30 -J-
25 -1-
20 -1-
15 -
                     Regression Line
                       R*2 = 14.0%
10
 5 -
                         500
       1000
Idle HC Emissions (ppm)
1500
2000
                                              E-51

-------
  FTPHC
  (g/ml)

  5 -
4.5 --


 4 --


3.5 --


 3 --


2.5 --

 rt _.„.,


1.5 --
  1
0.5
                               HC Emissions  -  FTP vs Lane Idle
                               (108 1983  & Newer TBI Vehicles)
                             Enlarged for Better Detail near Origin
                                                        Regression Line
                                                         RA2 * 14.0%
        f-Vf*1.
             50
                      100
150       200       250

       Idle HC Emissions (ppm)
300
350
400
450
                                             E-52

-------
Regression Analyses of FTP CO versus  CO Emissions Over:
         IM240, Idle Test,  and 2500/Idle  Test
                    For TBI Vehicles
Dependent variable is:
RA2 = 79.9%
s = 20.97 with 108 - 2 = 106
Source
Regression
Residual
Variable
Constant
COIM24
Sum of Squares
185395
46625.6
Coefficient
-3.46603
1.41279
COFTP
degrees of freedom
df
1
106
s.e. of Coeff
2.331
0.0688
Mean Square
200000
439.864
t-ratio
-1.49
20.5
F-ratio
421


Dependent
RA2 = 46.9%
s = 34.10
Source
Regression
Residual
Variable
Constant
Hlgherddle
variable is:
with 108 - 2 » 106
Sum of Squares
108761
123260
Coefficient
4.43526
15.0538
COFTP
degrees of freedom
df
1
106
s.e. of Coeff
3.677
1.557
Mean Square
100000
1162.83
t-ratlo
1.21
9.67
F-ratlo
93.5


Dependent
s= 32.73
Source
Regression
Residual
Variable
Constant
Idle.CO
variable is:
with 108 - 2 • 106
Sum of Squares
118502
113519
Coefficient
5.15188
17.2888
COFTP


degrees of freedom
df
1
106
s.e. of Coeff
3.47
1.644
Mean Square
100000
1070.93
t-ratlo
1.48
10.5
F-ratio
111

                          E-53

-------
   FTP CO
   (g/ml)
300 -r
250 --
200 --
150	
100 --
 50 --
    CO Emissions  -  FTP vs Lane IM240
      (108 1983 &  Newer TBI Vehicles)
                                                      Regression Line
                                                        R*2 = 79.9%
            i     •
               25
50
75        100        125
  IM240 CO Emissions (g/ml)
150
175
200
                                             E-54

-------
                             CO Emissions -  FTP vs Lane IM240
                               (108 1983  & Newer TBI Vehicles)
  (g/ml)                      Enlarged for Better Detail near Origin


50 ~                                            Regression Une
45 --

40 --

35 --

30 --

25 --

20 --

15

10

 5 -

 0
                                                  ft*2 = 70.9%
   0          5          10         15         20         25         30         35         40
                                     IM240 CO Emissions (g/ml)
                                             E-55

-------
   FTP CO
   (g/ml)
300 -
250 --
200 T
150
100 —
 50 -1-
CO Emissions  -   FTP vs  Lane 2500/ldle
    (108 1983  & Newer TBI Vehicles)
                                                                    Regression Line
                                                                      RA2 = 46.9%
             *  V
     0
           456
        2500/ldle CO Emissions (percent)
10
                                              E-56

-------
  FTP CO
  (g/ml)
45 --


40 --


35 --


30 --


25 --


20 --


15 --


10
CO Emissions  -  FTP  vs Lane  2500/ldle
    (108 1983  &  Newer  TBI Vehicles)
 Enlarged (or Better Detail near Origin
                                                   Regression Line
                                                     R*2 = 46.9%
            0.2      0.4       0.6       0.8       1        1.2
                                     2500/ldle CO  Emissions (percent)
                                      1.4
1.6
1.8
                                               E-57

-------
  FTP CO
  (g/mi)
300 T
                                CO Emissions  -  FTP vs Lane Idle
                                 (108  1983  & Newer TBI Vehicles)
250 --
200 -1-
Regresslon Line
  R*2 a 51.1%
150
100 --
 50 --
                                        456

                                       Idle CO Emissions (percent)
                                                 10
                                              E-58

-------
  FTP CO
  (g/ml)
 50 -
                                CO Emissions  -  FTP vs Lane  Idle
                                 (108 1983  &  Newer TBI Vehicles)
                               Enlarged for Better Detail near Origin
                                                   Regression Line
 45 4-                                                RA2sS1.1%


 40 --


 35


 30


 25 -


 20


 15


.10


  5


  0 ^
•	

	•
             0.2       0.4       0.6       0.8        1        1.2       1.4       1.6       1.8        2
                                       Idle CO Emissions (percent)
                                               E-59

-------
Regression  Analyses  of  FTP NOx  versus IM240 NOx Emissions
        (Note:   NOx is not Measured for the Idle Test)
Dependent variable is: NOFTP.TBI
RA2 = 80.7%
s = 0.4677 with 108 - 2 = 106 degrees of freedom
Source
Regression
Residual
Variable
Constant
NOIM24.TBI
Sum of Squares
96.941
23.1857
Coefficient
-0.069973
0.734578
df
1
106
s.e. of Coeff
0.071
0.0349
Mean Square
96.94
0.218733
t-ratlo
-0.985
21.1
F-ratlo
443

                             E-60

-------
 FTPNOx
  (g/ml)

7 -r
6 --
5 -
4 --
3 -
2 --
o -
NOx Emissions -  FTP vs Lane IM240
  (108 1983 & Newer TBI Vehicles)
                                                 Regression Line
                                                   R*2 = 80.7%
                                     345
                                      IM240 NOx Emissions (g/ml)
                                              E-61

-------
   FTPNOx
    (g/ml)

   2  -
1.75 --


 1.5 --


1.25 --


   1  --


0.75 --


 0.5 \


0.25 --
    NOx Emissions  -  FTP vs Lane IM240
      (108  1983 & Newer TBI  Vehicles)
    Enlarged for  Better Detail  near Origin
                                                                  Regression Une
                                                                   B*2 = 80.7%
               0.25
0.5
0.75         1         1.25
  IM240 NOx Emissions (g/ml)
1.5
1.75
                                               E-62

-------
Vehicles Tested by ATL at  Southbend,  Indiana
   Regressions of FTP vs lndolene-IM240
     for 91 1983 & Newer TBI Vehicles
Dependent variable is HCFTP
RA2 = 90.1%
s = 1.354 with
Source
Regression
Residual
Variable
Constant
HC.lM24.lndo...

91 - 2 » 89 degrees of
Sum of Squares
1491.43
163.188
Coefficient
0.002929
1.66576

freedom
df
1
89
s.e. of Coeff
0.1537
0.0584


Mean Square
1491
1.83357
t-ratio
0.019
28.5


F-ratio
813




Dependent variable is
RA2 m 90.3%
s = 15.12 with 91 -
Source
Regression
Residual
Variable
Constant
CO.IM24.lndo...
COFTP
2 = 89 degrees of
Sum of Squares
190004
20343.9
Coefficient
-0.174774
1 .34497
freedom
df
1
89
s.e. of Coeff
1.758
0.0467

Mean Square
200000
228.583
t-ratlo
-0.099
28.8

F-ratlo
831

Dependent variable is NOx.FTP
RA2 = 92.6%
s = 0.2989 with
Source
Regression
Residual
Variable
Constant
NOx.IM24.lnd...
91 - 2 = 89 degrees
Sum of Squares
99.1838
7.9495
Coefficient
0.073869
0.845099
of freedom
df
1
89
s.e. of Coeff
0.0443
0.0254

Mean Square
99.18
0.08932
t-ratio
1.67
33.3

F-ratio
1110



                               E-63

-------
FTPHC
(g/ml)
30 -
                           HC Emissions  -  FTP vs lndolene-IM240
                                (91  1983 & Newer TBI Vehicles)
25 -
Regression Line
  R*2 = 90.1%
20 -
15
10
 5 -
 0
                                               10
                                      IM240 HC Emissions (g/ml)
                                       15
20
                                             E-64

-------
FTPHC
(g/mi)
 3  -
                            HC Emissions  -  FTP vs lndolene-IM240
                                 (91 1983  &  Newer TBI Vehicles)
                              Enlarged for  Better Detail  near Origin
2.5  -
 2  -
1.5 -
                                                                       Regression Line
                                                                         R*2 = 90.1%
  1  -
                             I    •
0.5 -a
              VI
                  0.5
         1.5
IM240 HC Emissions (g/ml)
2.5
                                              E-65

-------
 FTP CO
 (g/ml)
300 -
                           CO Emissions  -  FTP vs lndolene-IM240
                                (91  1983 & Newer TBI Vehicles)
250 -
                      Regression Line
                        B*2 = 90.3%
200 -
150 -
100 --
 50 -J-
o -i
  0
             20
40       60
 80      100     120
IM240 CO Emissions (g/ml)
140      160      180
200
                                             E-66

-------
FTP CO
(g/ml)
50 -
                            CO  Emissions  -  FTP vs lndolene-IM240
                                (91 1983  & Newer TBI Vehicles)
                             Enlarged  for  Better Detail near Origin
40 -
30 -
20 -
                                                                           Regression Line
                                                                            RA2 = 90.3%
10 -1-
 0 -
                     10
20                30

IM240 CO Emissions (g/ml)
40
50
                                             E-67

-------
 FTPNOx
 (g/ml)

6.00 --
5.00 --
4.00 -*-
3.00 -
2.00 --
                             NOx Emissions   -  FTP vs  lndolene-IM240
                                  (91 1983 & Newer TBI Vehicles)
                                           Regression Line
                                             RA2 s 92.6%
1.00 --
0.00 -r
    0.00
1.00        2.00
3.00       4.00        5.00
  IM240 NOx Emissions (g/ml)
6.00       7.00
8.00
                                                E-68

-------
 FTPNOx
 (a/ml)
2.00
1.75  -1-
                            NOx Emissions  -   FTP  vs lndolene-IM240
                                 (91 1983 & Newer TBI Vehicles)
                              Enlarged for Better Detail  near Origin
                           Regression Line
                             R*2 = 92.6%
1.50
1.25


1.00 --


0.75 --


0.50 --


0.25 --
             V •  "
        •>^    • 1
          "l  "•  •
o.oo -
    0.00
0.25
0.50
0.75        1.00        1.25
  IM240 NOx Emissions (g/ml)
1.50
1.75
2.00
                                              E-69

-------
EF Vehicles Tested at  EPA's MVEL
 Regressions of FTP  vs Indolene IM240
      for 62 1985-88 TBI Vehicles
Dependent variable is: HC.FTP
RA2 = 87.7%
s = 1 .703 with 62 - 2 = 60 degrees of freedom
Source
Regression
Residual
Variable
Constant
HCIM24
Sum of Squares
1241.84
173.984
Coefficient
0.146231
1.71372
df
1
60
s.e. of Coeff
0.2279
0.0828
Mean Square
1242
2.89973
t-ratlo
0.642
20.7
F-ratio
428

Dependent variable is: CO.FTP
RA2 a 92.5%
s a 1 1 .59 with 62 - 2 = 60 degrees of freedom
Source
Regression
Residual
Variable
Constant
COIM24
Sum of Squares
99522.9
8056.07
Coefficient
2.56957
1.10181
df
1
60
s.e. of Coeff
1.584
0.0405
Mean Square
99523
134.268
t-ratlo
1.62
27.2
F-ratlo
741

Dependent variable is: NOx.FTP
RA2 m 90.9%
s - 0.1594
Source
Regression
Residual
Variable
Constant
NOIM24

with 62 - 2 = 60 degrees
Sum of Squares
15.2107
1.52398
Coefficient s.e
0.067854
0.780974

of freedom
df
1
60
. of Coeff
0.0323
0.0319


Mean Square
15.21
0.0254
t-ratlo
2.1
24.5


F-ratio
599




             E-70

-------
FTPHC
(g/ml)
30.00 --
20.00 --
10.00 -"-
 0.00 -fl

     0.00
                             HC Emissions  -   FTP  vs lndolene-IM240
                                    (62  1985-88 TBI Vehicles)
                                  (Tested at EPA's Ann Arbor Lab)
5.00
             Regression Line
              R*2 = 87.7%
        10.00
IM240 HC Emissions (g/ml)
15.00
20.00
                                               E-71

-------
                          HC Emissions  -  FTP vs lndolene-IM240
                                 (62  1985-88 TBI Vehicles)
                              (Tested at EPA's Ann Arbor Lab)
                            Enlarged  for Better Detail near Origin
FTPHC
 .00  T
2.00 +
1.00  +
0.00
                                                            Regression Line
                                                             R*2 s 87.7%
              .  *•  ••   •
           B *•%
     k#-    -
    0.00                 0.50                 1.00                 1.50                 2.00
                                    IM240 HC Emissions (g/ml)
                                          E-72

-------
FTP CO
(g/ml)
200.00 --
150.00 --
100.00 --
 50.00 --
                             CO Emissions  --  FTP vs lndolene-IM240
                                    (62  1985-88 TBI Vehicles)
                                  (Tested at EPA's Ann Arbor Lab)
  0.00 *

      0.00
50.00
                                       Regression Line
                                         R*2 = 92.5%
        100.00
IM240 CO Emissions (g/ml)
150.00
200.00
                                               E-73

-------
FTP CO
(g/ml)
50.00 -
                             CO Emissions  -   FTP vs lndolene-IM240
                                    (62  1985-88 TBI  Vehicles)
                                  (Tested at  EPA's Ann Arbor Lab)
                               Enlarged for  Better Detail near  Origin
40.00 T
30.00 -
20.00 -
                                              Regression Line
                                                RA2 = 92.5%
10.00 T
 0.00 -
     0.00
10.00
20.00             30.00
IM240  CO Emissions (g/ml)
40.00
50.00
                                               E-74

-------
FTP NOx
                             NOx Emissions  -  FTP vs lndolene-IM240
                                     (62 1985-88  TBI  Vehicles)
                                  (Tested at EPA's Ann Arbor Lab)
3.00 -1-
2.50
2.00 —
 1.50 -
 1.00 -
 0.50 -
 0.00 -r
                                   Regresslon Line
                                     RA2 = 90.9%
    0.00
0.50
1.00
1.50        2.00        2.50
  IM240 NOx Emissions (g/ml)
3.00
3.50
4.00
                                               E-75

-------
FTPNOx
 1.50 --
 1.00
0.50 --
o.oo
                            NOx Emissions  -  FTP  vs lndolene-IM240
                                    (62  1985-88  TBI Vehicles)
                                 (Tested at EPA's Ann  Arbor Lab)
                              Enlarged for Better Detail  near Origin
                                                    Regression Line
                                                      B*2 s 90.9%
                                      •  •
    0.00
0.50
         1.00
IM240 NOx Emissions (g/ml)
1.50
2.00
                                               E-76

-------
                 Description  of Repair  Regression Analysis

Regression  Analyses

      The difference in emission levels before and after repairs for the FTP'S and Lab
IM240's run on Indolene fuel was calculated such that a positive difference would
correspond to a decrease in emissions or an increase in fuel economy. The AFTP values
(AHC, AGO, ANOx and AMPG) were regressed against the corresponding Indolene
AIM240 values. The results and graphs of these regressions are included.

Data Description

      The data used in this analysis consisted of FTP and Indolene IM240 scores for tests
performed before and after repairs. These results consisted of HC, CO, NOx, and fuel
economy values. The database was restricted to 1983 and newer PFI and TBI vehicles
which received repairs and whose as received FTP scores exceeded twice the FTP standard
for either HC or CO. Also excluded from the analysis were vehicles which received repairs
before the as received FTP and vehicles which had differences in dynamometer settings of
greater than 15%. The resulting data set contained 41 TBI vehicles and 23 PFI vehicles.
Summary Statistics for the variables used in the TBI analyses follow.
                                     E-77

-------
                Statistical Summaries for Variables  Used in
                TBI Before and After Regression Analyses
Summary statistics for: AHCIM24
NumNumeric = 41
Mean =1.3563
Median » 0.2200
Midrange = 7.8500
Standard Deviation = 3.4020
Range = 17.660
Minimum = -0.98000
Maximum = 16.680
Summary statistics for: AHC.FTP
NumNumeric = 41
Mean = 2.4634
Median = 0.3700
Midrange = 14.595
Standard Deviation = 6.0860
Range = 30.590
Minimum = -0.70000
Maximum = 29.890
Summary statistics for ACO.IM24
NumNumeric = 41
Mean =24.092
Median = 4.8800
Midrange = 84.310
Standard Deviation = 47.490
Range = 186.72
Minimum = -9.0500
Maximum = 177.67

Summary statistics for: ANOx.IM24
NumNumeric = 41
Mean = 0.04683
Median = -0.07000
Midrange = 2.2850
Standard Deviation = 1.2167
Range = 8.3700
Minimum = -1.9000
Maximum = 6.4700

Summary statistics for: AMPG.IM24
NumNumeric = 41
Mean = 2.4132
Median =1.4300
Midrange = 5.2600
Standard Deviation = 3.4721
Range = 15.420
Minimum = -2.4500
Maximum = 12.970
Summary statistics for  ACO.FTP
NumNumeric = 41
Mean = 38.661
Median = 5.4500
Midrange = 128.11
Standard Deviation = 69.2007
Range = 265.65
Minimum = -4.7200
Maximum = 260.93

Summary statistics for  ANOx.FTP
NumNumeric = 41
Mean = -0.06439
Median = -0.06000
Midrange =1.5800
Standard Deviation = 1.0174
Range = 6.8800
Minimum = -1.8600
Maximum = 5.0200

Summary statistics for:  AMPG.FTP
NumNumeric = 41
Mean = 2.0820
Median = 0.97000
Midrange = 3.6850
Standard Deviation = 3.7642
Range = 21.810
Minimum - -7.2200
Maximum = 14.590
                                  E-78

-------
    Regression Analyses for the Change in Emission Constituents
Before and After Repairs of the TBI Vehicles for the FTP vs IM240-A
Dependent variable is: AHC.FTP
R*2 m 87.2%
s = 2.209 with 41 - 2 = 39 degrees of freedom
Source Sum of Squares df Mean Square
Regression 1291.3 1 1291
Residual 190.271 39 4.87875
Variable Coefficient s.e. of Coeff t-ratlo
Constant 0.198146 0.372 0.533
AHC.IM24 1.67013 0.1027 16.3
F-ratlo
265
Dependent variable is: ACO.FTP
RA2=82.1%
s = 29.57 with 41 - 2 » 39 degrees of freedom



Source Sum of Squares df Mean Square
Regression 1 56368 1
Residual 34110.7 39
Variable Coefficient s.e. of Coeff
Constant 6.94316 5.192
ACO.IM240 1.31656 0.0985
200000
874.633
t-ratio
1.34
13.4



F-ratlo
179




Dependent variable is:
RA2 = 84.9%
s m 0.4008
Source
Regression
Residual
Variable
Constant
ANOX.IM240

with 41 - 2 = 39
ANOX.FTP

degrees of freedom



Sum of Squares df Mean Square
35.1435
6.26452
Coefficient
-0.100466
0.770364
1
39 0.
s.e. of Coeff t
0.0626
0.0521
35.14
160629
-ratio
-1.6
14.8



F-ratio
219




Dependent variable is:
RA2 m 58.3%
S = 2.489 with
Source
Regression
Residual
Variable
Constant
AMPG.IM240
AMPG.FTP
41 - 2 = 39 degrees
Sum of Squares
325.213
241.558
Coefficient
0.100224
0.821213
s.e.
0
0
of freedom
df Mean Square
1 325
39 6.19379
of Coeff
.4753
.1133
t-ratlo
0.158
7.25
F-ratio
52.5


                                 E-79

-------
 A FTP
HC (g/ml)
                         HC Emissions  -  AFTP vs Alndolene-IM240
              (41  1983 & Newer TBI Vehicles with Repairs and FTP > 2*FTP.Stnd)
        35 T
        30 +
        25 +
        20
        15  +
        10  +
         5
                                                                 Regression Line
                                                                   B*2 s 87.2%

J^
                                              T"

                                              8
                                               10
12
14
        - 5
                                        A IU240 HC (g/ml)
16
18
                                             E-80

-------
            HC Emissions   -  AFTP vs Alndolene-IM240
(41 1983  & Newer TBI  Vehicles with Repairs  and FTP > 2*FTP.Stnd)
               AFTP HC
               (g/ml)
            o
                                                              Regression Line
                                                                R*2 = 87.2%
                                                              A IM240 HC (g/ml)
                   Expanded for Better Detail near Origin
                                 E-81

-------
 A FTP
CO (g/ml)

      300 —
     250 J-
     200 —
      150 —
      100 —
      50 -L
                        CO  Emissions  -  AFTP vs Alndolene-IM240
            (41  1983 &  Newer TBI Vehicles with Repairs and FTP Z 2*FTP.Stnd)
                                               Regression Line
                                                 RA2 a 82.1%
-2O
I
      -50 —
20       40      60       80      100      120      140      160      180


                    A IM240 CO (g/ml)
                                           E-82

-------
  A FTP
CO (g/ml)
                         CO Emissions  -   AFTP vs Alndolene-IM240
             (41  1983 & Newer TBI Vehicles with  Repairs and  FTP > 2*FTP.Stnd)
50 -




40 --




30 --




20 --

   •

10 --
                                Regression Line
                                  H*2 = 82.1%
              x
               X
                                               X
                                                X
                                                  X
                                           X
                              X
                               10
20
                                                30
40
            -10

    Expanded for Better Detail near Origin
                                          A IM240 CO (g/ml)
50
                                             E-83

-------
-2
                       NOx  Emissions  -  AFTP vs Alndolene-IM240
             (41 1983  & Newer TBI Vehicles with Repairs and FTP > FTP.stnd)
       A FTP
     NOx (g/ml)
                  6 -
                  5 -
                  4 -
                  3 -
                  2 -
                  •4 	
Regression Line
  RA2 a 84.9%
                 -J *

                 -2 -
                                      A IM240 NOx (g/ml)
                                            E-84

-------
                        NOx Emissions   --  AFTP  vs Alndolene-IM240
             (41  1983 &  Newer TBI Vehicles  with  Repairs and FTP > FTP.stnd)
                                A FTP NOx
                                  (g/ml)
                                           1.5  -
                                             1  —
                                           0.5  —
                                                            Regression Line
                                                              R*2 = 84.9%
   A IM240
  NOx fa/nth
-2
-1.5
-1
-0.5
0.5
1.5
                                          -0.5  -r-
                                          -1.5  -
                                            -2  —

                                 Expanded for Better Detail near Origin
                                             E-85

-------
                        Fuel  Economy  -  AFTP vs Alndolene-IM240
            (41  1983 &  Newer TBI Vehicles with Repairs and FTP > 2*FTP.Stnd)
                 15 -
      A FTP
   MPG (ml/gal)
                 10 -1-
                  5 -
                                                 Regression Line
                                                   R*2 = 57.4%
•A
8        10        12        14
                 -5 -
                -10 -
                                     A IM240 MPG (ml/gal)
                                           E-86

-------
                         Fuel Economy  -  AFTP vs Alndolene-IM240
            (41  1983 &  Newer  TBI Vehicles with Repairs  and FTP > 2*FTP.Stnd)
                   AFTPMPG
                   (ml/gal)   5
                             3 -


                             2 --
                                                  Regression Line
                                                   R*2 a 57.4%
-3
-2
                            -2 --


                            -3 --


                            -4 --


                            -5 --

                   Expanded for Better Detail near Origin
                                                                      A IM240 MPG
                                                                          (ml/gal)
                                             E-87

-------
                  APPENDIX F









MOBILE4.1 TECHNOLOGY DISTRIBUTION AND  EMISSION




        GROUP RATES AND EMISSION LEVELS

-------
                                                                   MOBILE4.1
                                            Technology Distribution for Gasoline Passenger Cars

HP [1,1
XCK
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991

Mo [|a|
XcK
1981
1982
1983
1984
1985
1986
1987
1986
1989
1990
1991
Modal
Star
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
CLLP
3iray
Stttt
1571360
1040443
151414
761159
966873
594368
415441
277445
613700
10800
9000
CLLP
Stray
fin*
23.0%
15.9%
22%
8.6%
12.9%
75%
6.4%
4.1%
72%
0.1%
0.1%
CLLP
Cob
62.9%
50.6%
48.6%
55.1%
40.8%
31.9%
24.9%
10.1%
125%
1.8%
03%
CLLP
Stray
UPFI
413972
403436
591266
938060
2182617
2607047
2239841
2998983
4652439
5177046
6419242
CLLP
Stray
lies
6.0%
62%
8.5%
10.5%
292%
32.9%
34.7%
44.4%
54.6%
71 .8%
77.4%
CLLP
MEB
6.1%
62%
8.8%
11.0%
30.7%
392%
372%
492%
59.7%
76.1%
79.5%
CLLP
3MfBy
IBI
0
435887
794934
1414498
432479
1058186
1397261
2206819
2034135
1395418
1570776
CLLP
Stray
IBI
0.0%
6.7%
115%
15.9%
5.8%
133%
21.7%
32.7%
23.9%
19.4%
18.9%
CLLP
IBI
2.9%
10.6%
183%
282%
20.8%
265%
36.3%
40.7%
275%
21.9%
202%
Oplp
3WNBy
Cmtb
48896
54160
0
25877
25476
50
16821
0
0
0
0
Oplp
9my
£Mfe
0.7%
03%
0.0%
03%
0.3%
0.0%
03%
0.0%
0.0%
0.0%
0.0%
Oplp
Any
28.1%
325%
24.4%
5.8%
7.7%
2.4%
1.7%
0.0%
03%
0.1%
0.0%
Oplp
Sway
MEB
0
0
0
0
0
0
64
65
90
0
0
Oplp
wWwy
MEB
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%













Oplp
Stray
IBI
0
0
0
0
0
150047
41918
0
0
0
0
Oplp
Stray
IBI
0.0%
0.0%
0.0%
0.0%
0.0%
1.9%
0.7%
0.0%
0.0%
0.0%
0.0%
Model
XtK
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
CLLP
OxStmy
Garb
2731528
2265172
3211332
4140952
2085078
1936299
1188253
404916
449631
122142
15690
CLLP
Ox3w*y
Cob
393%
34.7%
46.4%
465%
27.9%
24.4%
18.4%
6.0%
5.3%
1.7%
02%
Any
Cafe
91.0%
832%
72.9%
60.9%
484%
32.4%
25.9%
10.1%
12.8%
1.9%
0.3%
CLLP
OxSwwy
MEB
1434
0
15517
40669
116557
504156
156340
322283
431553
311612
176552
CLLP
Ox3w»y
MEB
0.0%
0.0%
02%
05%
1.6%
6.4%
2.4%
4.8%
5.1%
43%
2.1%
Any
MEB
6.1%
62%
8.8%
11.0%
30.7%
392%
372%
492%
59.7%
76.1%
795%
CLLP
OxSway
IBI
201346
258479
471542
1091249
1125292
1041046
942058
542380
308598
186117
103342
CLLP
OxStmy
IBI
2.9%
4.0%
6.8%
123%
15.0%
13.1%
14.6%
8.0%
3.6%
2.6%
12%
Any
IBI
2.9%
10.6%
183%
282%
20.8%
28.4%
36.9%
40.7%
275%
21.9%
202%
Oplp
OxSvray
CMfe
932653
1181258
870343
488429
547055
40209
46809
0
25000
7000
0
Oplp
OxSvmy
CjSfe
13.6%
18.1%
12.6%
55%
73%
05%
0.8%
0.0%
03%
0.1%
0.0%













Oplp
OxStray
MEB
284
0
409
0
0
0
650
0
0
0
0
Oplp
OxStray
MEB
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Model
IMC
1981
1982
1963
1984
1985
1986
1987
1988
1989
1990
1991
Oplp
OxStmy
IBI
0
0
0
0
0
0
0
0
0
0
0
Oplp
OxStmy
IBI
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Any
Stray
29.7%
29.6%
222%
353%
482%
55.6%
63.8%
812%
85.7%
913%
96.4%
Oplp

OxM
943390
889494
817624
0
0
0
0
0
0
0
0
Oplp

OxM
13.8%
13.6%
11.8%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Any
OxSiray
565%
56.8%
66.0%
64.7%
51.8%
44.4%
362%
18.8%
143%
8.7%
3.6%
Oplp

fton*
38
10
0
0
0
0
0
0
0
0
0
Oplp

fton»
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Any
Ox Id
133%
13.6%
11.8%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%


AJ1
6844901
6528339
6924581
8900893
7481429
7931408
6447456
6752891
8515146
7210135
8294602





























GMFI
133369
636022
831142
1444036
1644484
2087185
1673716
2420729
3018191
2522963
3639046





























%QMFI
1.9%
9.7%
12.0%
162%
22.0%
263%
26.0%
35.8%
35.4%
35.0%
43.9%



























All vehicle counts taken from CAFE estimates, except 1989 and newer model years, which are from General Label predictions.
                                                                                  7/2/91

-------
                                                                                 MOBILE4.1
                                                                        Emitter Group Emission Levels*

Modsl
Ysar
Group
19B1-B2
196142
196142
1981-82
108345
1983-85
1983-85
1983*5
1986+
1986+
1986+
1986*
1981-82
1981-82
1961-82
1981-62
1983+
1983+
1983+
1983+
198142
198142
198142
196142
1983+
1983+
1983+
1983+
198142
196142
196142
196142
1963+
1983+
1983+
1983+

Emission
UM|
Group
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Noifisw
Normal
High
High
High
High
High
High
High
High
Vary High
Vary High
Very High
Vsry High
Vary High
Vsry High
Vsry High
Vsry High
Supsr
Supsr
Supsr
Supsr
Supsr
Supsr
Supsr
Supsr


Technology
Group
MPR
TO
Cart.
Oplp
MPH
TO
Can
Oplp
MPH
TO
Can
Oplp
MPH
TO
Can
Oplp
MPH
TO
Can
Oplp
MPH
TO
Can
Oplp
MPH
TO
Can
Oplp
MPH
TO
Cart.
Oplp
MPH
TO
Can
Oplp


HC
an.
0.188
0.279
0.288
0.306
0.269
0.242
0.222
0.334
0.269
0242
0222
0.334
0.374
0.758
0.781
0.757
0.997
0.762
0.703
0.840
0.606
1.163
1.950
1389
2.019
2.242
2.002
1.352
15.259
15.259
15259
0.000
11.123
16.083
12.870
0.000


HC
BEL
0.0450
0.0161
0.0253
0.0230
0.0115
0.0134
0.0199
0.0248
0.0115
0.0134
0.0199
0.0248
0.0450
0.0161
0.0253
0.0230
0.0115
0.0134
0.0199
0.0248
0.0450
0.0161
0.0253
0.0230
0.0115
0.0134
0.0199
0.0248
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
HC
Emission*
at 50.000
MJjsj
0.413
0.360
0.415
0.421
0.327
0.309
0.322
0.458
0.327
0.309
0.322
0.458
0.599
0.839
0.908
0.872
1.055
0.829
0.803
0.964
0.831
1244
2.077
1.704
2.077
2.309
2.102
1.476
15.259
15259
15.259
0.000
11.123
18.083
12.870
0.000
HC
Emissions
at 100,000
Mllss .
0.638
0.440
0541
0.536
0.384
0.376
0.421
0.582
0.384
0.376
0.421
0.582
0.824
0.919
1.034
0.987
1.112
0.896
0.902
1.088
1.056
1.324
2203
1.819
2.134
2.376
2.201
1.600
15.259
15.259
15.259
0.000
11.123
18.083
12.870
0.000


CO
ZUL
1.548
3.422
3.067
3.368
2.S98
2.953
2209
4.093
2.598
2.953
2.209
4.093
3.879
6.843
8.496
8.142
7.602
8.820
8.577
8.386
18.049
22.059
33.396
27.650
22.301
44.416
36.130
34.021
173.840
173.840
173.840
0.000
180.000
183.590
246.850
0.000


CO
EH
0.5807
0.2520
0.3281
02877
0.1554
0.1572
0.2282
0.2093
0.1554
0.1572
02282
02093
05807
0.2520
0.3281
02877
0.1554
0.1572
02282
02093
0.5807
02520
0.3281
0.2877
0.1554
0.1572
02282
02093
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
CO
Emissions
at 50,000
MJtet
4.452
4.682
4.708
4.807
3.375
3.739
3.350
5.140
3.375
3.739
3.350
5.140
6.783
8.103
10.137
9.581
8.379
9.606
9.718
9.433
20.953
23.319
35.037
29.089
23.078
45.202
37.271
35.068
173.840
173.840
173.840
0.000
189.000
183.590
246.850
0.000
CO
Emissions
at 100,000 NOx
Ifflsj OIL
7.355 0.380
5.942 0.545
6.348 0.676
6245 0.649
4.152 0.689
4.525 0.585
4.491 0.692
6.186 0.524
4.152 0.539
4525 0.316
4.491 0.478
6.186 0.524
9.686
9.363
11.777
11.019
9.156
10.392
10.859
10.479
23.856
24.579
36.677
30.527
23.855
45.988
38.412
36.114
173.840
173.840
173.840
0.000
189.000
183.590
246.850
0.000
NOx NOx
Emissions Emissions
NOx at 50.000 at 100,000
PET Mllas Mllss
0.1280 1.020 1.660
0.1414 1.252 1.959
0.0696 1.026 1.374
0.0440 0.869 1.089
0.0185 0.782 0.874
0.0470 0.820 1.055
0.0558 0.971 1.250
0.0540 0.794 1.064
0.0185 0.632 0.724
0.0470 0.551 0.786
0.0558 0.757 1.036
0.0540 0.794 1.064
























        ' All emission (avals in grains per mis. DET in grams per mile per 10,000 miles.

BLOCK.XLS
                                                                                                                                                       11/8/91

-------
                                 MOBILE4.1
                              Emitter Group Rates

Model
Year
Group
1981-82
1981-82
1981-82
1981-82
1983+
1983+
1983+
1983+
1981-82
1981-82
1981-82
1981-82
1983+
1983+
1983+
1983+
1981-82
1981-82
1981-82
1981-82
1983+
1983+
1983+
1983+
1981-82
1981-82
1981-82
1981-82
1983+
1983+
1983+
1983+

Emission
Level
Group
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
High
High
High
High
High
High
High
High
Very High
Very High
Very High
Very High
Very High
Very High
Very High
Very High
Super
Super
Super
Super
Super
Super
Super
Super


Technology
Group
MPFI
TBI
Garb
Oplp
MPFI
TBI
Garb
Oplp
MPFI
TBI
Garb
Oplp
MPFI
TBI
Garb
Oplp
MPFI
TBI
Garb
Oplp
MPFI
TBI
Garb
Oplp
MPFI
TBI
Garb
Oplp
MPFI
TBI
Garb
Oplp
Growth
Rate
per 10,000
Miles
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.0065
0.0065
0.0270
0.0270
0.0090
0.0040
0.0140
0.0130
0.01933
0.01198
0.03762
0.03530
0.00840
0.02012
0.01487
0.00433
0.00257
0.00257
0.00257
0.00000
0.00194
0.00194
0.00194
0.00000
Rate
Increase
Beyond
50.000 Miles
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
13.8353
13.8353
0.9643
1.0000
4.4709
15.2076
2.3200
1.0000
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Percent
at
50,000
Miles
85.8%
89.5%
66.4%
68.8%
90.3%
87.0%
84.6%
91.3%
3.3%
3.3%
13.5%
13.5%
4.5%
2.0%
7.0%
6.5%
9.7%
6.0%
18.8%
17.7%
4.2%
10.1%
7.4%
2.2%
1.3%
1.3%
1.3%
0.0%
1.0%
1.0%
1.0%
0.0%
Percent
at
100,000
Miles
29.9%
37.2%
33.3%
37.7%
65.0%
45.5%
60.0%
82.7%
48.2%
48.2%
26.5%
27.0%
24.6%
32.4%
23.2%
13.0%
19.3%
12.0%
37.6%
35.3%
8.4%
20.1%
14.9%
4.3%
2.6%
2.6%
2.6%
0.0%
1.9%
1.9%
1.9%
0.0%
BLOCK.XLS
11/8/91

-------
                     APPENDIX G









EXHAUST SHORT TEST ACCURACY: IM240 VS. SECOND-CHANCE




                 2500 RPM/IDLE TEST









                     PFI       p. G-2




                     TBI       p. G-5




                     CARB      p. G-8

-------
Special Note On Appendix G:

       Appendix G shows the accuracy of the IM240 test when a vehicle is failed only if its
composite emissions exceed 0.8 g/mi HC or 15 g/mi CO and its Bag 2 emissions alone exceed 0.5
g/mi HC or 12 g/mi CO. In other words, a vehicle can pass by having low emissions in Bag 2
even if Bag 1 emissions were very high.  EPA has found that this prevents inappropriately failing
clean cars with no sacrifice in identifyimg dirty cars. For this reason, EPA is proposing this
approach to IM240 cutpoints.

       Please note that second chance Idle/2500 tests were performed on 1983 and newer vehicles
only, and therefore were not available for the 1981  and 1982 carbureted vehicles used to construct
these charts. In all cases, when a second chance  Idle/2500 test was available, the better test was
used.
                                          G-l

-------
                                                  Figure 1
                               Comparison of IM240 to Second Chance 2500/Idle for
                          1983 & Newer Port Fuel Injected Vehicles from Hammond Indiana
                         IM240 Cutpoints = 0.8 & 15 g/mi Composite and 0.5 &12 g/mi Bag 2
                                Second Chance 2500/Idle Clpts = 220 ppm & 1.2%
90%  —
80%   -t-
70%
60%
50%
40%
30%
20%   -
10%
 0%
                                    85%
                  82%
                             73%
          65%
                                                       14%
                                                                 0.0%    0.0%
                                                                                    1.7%
                                                                                           2.3%
            HCIDR
COIDR
 I/M Fail Rate     I/M Fail Rate for FTP   I/M Fail Rate for
	Passing Vehs	Normal Emitters
 D 2-Spd (220ppm & 1.2%)
                       IM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                G-2

-------
                                                  Figure 2
                               Comparison of IM240 to Second Chance 2500/Idle for
                                1983 & Newer PFI Vehicles from Hammond Indiana
                          IM240 Cutpoints = 0.8 & 15 g/mi Composite; 0.5 & 12 g/mi Bag 2
                   Second Chance 2500/ldle Ctpts = 220 ppm & 1.2% @2500; 100 ppm & 1.0% @Idle
90%
80%
70%
60%
50%
40%
30%  —
20%
10%  -
                                    85%
                                                                                    22%
                                                                  13%
                                                                         0.0%
                                                                                           2.3%
            HCIDR
COIDR
I/M Fail Rate
I/M Fail Rate for FTP  I/M Fail Rate for
   Passing Vehs	Normal Emitters
 D 2-Spd (220/lOOppm & 1.2/1.0%)
                       IM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                G-3

-------
                                                  Figure 3
                               Comparison of EM240 to Second Chance 2500/Idle for
                          1983 & Newer Port Fuel Injected Vehicles from Hammond Indiana
                          IM240 Cutpoints = 0.8 & 15 g/mi Composite; 0.5 &12 g/mi Bag 2
                           Second Chance 2500/Idle Ctpts = 100 ppm & 0.5% both modes
90%
80%
70%  --
60%
50%
40%  -t-
30%   --
20%
10%
 0%
                                    85%
                                                                                    26%
                                                                 13%
                                                                        0.0%
                                                                                           2.3%
            HCIDR
COIDR
 I/M Fail Rate     I/M Fail Rate for FTP   I/M Fail Rate for
	Passing Vehs	Normal Emitters
 O 2-Spd (lOOppm & 0.5%)
                       EM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                G-4

-------
                                                  Figure 4
                               Comparison of IM240 to Second Chance 2500/Idle for
                        1983 & Newer Throttle Body Injected Vehicles from Hammond Indiana
                          IM240 Cutpoints = 0.8 & 15 g/mi Composite; 0.5 &12 g/mi Bag 2
                                Second Chance 2500/Idle Ctpts = 220 ppm & 1.2%
90%  -r
                  89%
10%
0%
                                                                  3.S
                                                                                            4.1%
                                                                         0.0%
            HCEDR
CO DDR
I/M Fail Rate
I/M Fail Rate for FTP   I/M Fail Rate for
   Passing Vehs	Normal Emitters
 02-Spd(220ppm&1.2%)
                       IM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                G-5

-------
                                                  Figure 5
                               Comparison of IM240 to Second Chance 2500/Idle for
                                1983 & Newer TBI Vehicles from Hammond Indiana
                          IM240 Cutpoints = 0.8 & 15 g/mi Composite; 0.5 & 12 g/mi Bag 2
                   Second Chance 2500/ldIe Ctpts = 220 ppm & 1.2% @2500; 100 ppm & 1.0% @Idle
100%
 90%
 80%
 70%   —
 60%  --
 50%
 40%
 30%
 20%
 10%  -
           90%
                   89%
                                                                                    16%
                                                                  12%
                                                                         0.0%
                                                                                           4.1%
             HCIDR
COIDR
 I/M Fail Rate    I/M Fail Rate for FTP   I/M Fail Rate for
	Passing Vehs	Normal Emitters
 D 2-Spd (220/lOOppm & 1.2/1.0%)
                      IM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                G-6

-------
                                                  Figure 6
                               Comparison of IM240 (o Second Chance 2500/Idle for
                        1983 & Newer Throttle Body Injected Vehicles from Hammond Indiana
                          IM240 Cutpoints = 0.8 & 15 g/mi Composite: 0.5 & 12 g/mi Bag 2
                            Second Chance 2500/Idle Clpts = 100 ppm & 0.5% both modes
100%
 40%
 20%
 10%
 0%
                                                                  20%
                                                                                     22%
                                                                         0.0%
                                                                                            4.1%
             HCIDR
COEDR
I/M Fail Rate    I/M Fail Rate for FTP  I/M Fail Rate for
                  Passing Vehs	Nonnal Emitters
 U 2-Spd (lOOppm & 0.5%)
                      IM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                G-7

-------
                                                  Figure 7
                               Comparison of IM240 to Second Chance 2500/Idle for
                             1981 & Newer Carbureted Vehicles from Hammond Indiana
                          IM240 Cutpoints = 0.8 & 15 g/mi Composite; 0.5 & 12 g/mi Bag 2
                                Second Chance 2500/Idle Ctpls = 220 ppm & 1.2%
90%
                                    84%
80%
70%
60%  —
50%
40%
30%
20%
10%  --
 0%
                                                                                           5.7%
                                                                                   4.0%
                                                                 0.0%   0.0%
            HCIDR
COEDR
 VM Fail Rate     I/M Fail Rate for FTP   I/M Fail Rate for
	Passing Vehs	Normal Emitters
    2-Spd (220ppm & 1.2%)
                       IM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                G-8

-------
                                                  Figure 8
                               Comparison of IM240 to Second Chance 2500/Idle for
                             1981 & Newer Carbureted Vehicles from Hammond Indiana
                          IM240 Cutpoints = 0.8 & 15 g/mi Composite; 0.5 & 12 g/mi Bag 2
                   Second Chance 2500/Idle Ctpts = 220 ppm & 1.2% @2500; 100 ppm & 1.0% @Idle
                                     84%
80%
70%
60%
50%  -
40%
30%  H
20%
10%  —
0%
                                                                                     6.9%
                                                                  0.0%   0.0%
                                                                                            5.7%
            HCIDR
COIDR
 I/M Fail Rate    I/M Fail Rate for FTP   I/M Fail Rate for
	Passing Vehs	Normal Emitters
 G 2-Spd (220/100ppm & 1.2/1.0%)
                       IM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                 G-9

-------
                                                  Figure 9
                               Comparison of IM240 to Second Chance 2500/Tdle for
                             1981 & Newer Carbureted Vehicles from Hammond Indiana
                          IM240 Cutpoints = 0.8 & 15 g/mi Composite: 0.5 & 12 g/mi Bag 2
                           Second Chance 2500/ldle Ctpts = 100 ppm & 0.5% both modes
90%
           83%    83%
                                    84%
80%
70%
60%
50%
40%
30%
20%
10%  -1
                             81%
                49%
            HCIDR
COIDR
I/M Fail Rate
I/M Fail Rate for FTP   I/M Fail Rate for
   Passing Vehs _ Normal Emitters
    2-Spd(100ppm&0.5%)
                       IM240 (0.8 & 15 g/mi composite; 0.5 & 12 g/mi Bag 2)
                                                G-10

-------
                 APPENDIX H









DATA ANALYSES FOR APPENDIX G: 1983 AND NEWER




     PFI,  TBI,  AND  CARBURETED VEHICLES








                 PFI       p. H-2




                 TBI       p. H-15




                 CARB      p. H-28

-------
Database  Description

       The data used to calculate these tables consisted of IM240 Lane, second chance
2500/IdIe, and FTP test scores for 1983 and newer port fuel injected and throttle body fuel
injected vehicles tested at ATL's laboratory in South Bend, Indiana. Carburetor equipped
1981 and newer vehicles are also included, but because no second chance 2500/Idle tests
were performed on the 1981 and 1982 vehicles, we only used the first chance results for
1981-82 carbureted cars.  For 1983 and later vehicles, we always used the better results
from the first or second chance tests for all fuel metering types.

       Only vehicles receiving each of these tests were used in these analyses. Vehicles
receiving repairs in the interim between the lane tests and the FTP, and vehicles exhibiting
at least a 15% variation between lane dynamometer settings and laboratory dynamometer
settings were excluded from the  database.

       These differences in dynamometer settings arose from the use of different
dynamometer look-up tables at the lane and lab. Traditional certification tables used for
emission factor testing are too detailed for efficient use in the 1/M lane so simpler tables
were developed for use at the lane. The tables developed for the I/M lane are probably
more representative than the traditional tables because they are derived from more vehicles
with various optional equipment. After becoming aware of the dynamometer setting
differences in the two tables, EPA opted to use the simplified table at both testing sites to
avoid future inconsistencies. The resulting database consisted of 74 1983 and newer PFI
vehicles and 108 1983 and newer TBI vehicles and was used to construct Appendices E,
G, and H .  Regressions were not performed for the 122 1981 and newer carbureted
vehicles.
Weighting Factor Explanation

       FTP's were performed only on vehicles recruited to the lab for additional testing.
Because the recruiting criteria for vehicles at Hammond lane was based on equal quotas for
low IM240 emitters and high IM240 emitters it did not ensure a sample representative of
the in-use fleet. Weighting factors were used to better simulate a representative sample.
The weighting factors were calculated as the ratio of '83 and newer PFI vehicles tested at
Hammond I/M lane to PFI vehicles tested at the South Bend lab. The weighting factors for
the TBI and carbureted vehicles were calculated similarly. The simulated fleet was then
created by multiplying the lab sample results by the weighting factors.


IM240 Cutpoint  Explanation

       Appendix H shows the characteristics of the IM240 test when a vehicle is failed
only if its composite emissions exceed 0.8 g/mi HC or 15 g/mi CO and its Bag 2 emissions
alone exceed 0.5 g/mi HC or 12 g/mi CO. In other words, a vehicle can pass by having
low emissions in Bag 2 even if Bag 1 emissions were very high. EPA has found that this
prevents inappropriately failing clean cars with no sacrifice in identification of dirty cars,
and therefore, is proposing this approach  to IM240 cutpoints. Tables calculated using this
approach are labeled 'Two Ways to Pass Criteria." Tables calculated using composite
IM240 cutpoints alone are also included for comparison. These Tables are labeled "Full
IM240 Test Criteria". Both tables help illustrate the method of analysis and underlying
sample sizes as well as  showing how the characteristics of the IM240 and Idle/2500 tests
vary with more or less stringent cutpoints.
                                       H-l

-------
1983+ Simulated Lane Sample: "Two
PF1 Vehicles
Excess
LaneIM240
Cut-Points
Comp+Baa 2
0.6/10 + 0.5/12
0.6/15+0.5/12
0.6/20 + 0.5/12
0.7/10 + 0.5/12
0.7/15+0.5/12
0.7/20 + 0.5/12
0.75/10+0.5/12
0.75/15+0.5/12
0.75/20 + 0.5/12
0.8/10 + 0.5/12
0.3/15 +04/12
0.8/20 + 0.5/12
0.85/10 + 0.5/12
0.85/15+0.5/12
0.85/20 + 0.5/12
0.9/10 + 0.5/12
0.9/15+0.5/12
0.9/20 + 0.5/12
1.0/10+0.5/12
1.0/15+0.5/12
1.0/20 + 0.5/12
1.1/10 + 0.5/12
1.1/15+0.5/12
1.1/20 + 0.5/12
1.2/10 + 0.5/12
1.2/15+0.5/12
1.2/16 + 0.5/12
1.2/20 + 0.5/12
1.6/10 + 0.5/12
1.6/15+0.5/12
1.6/20+0.5/12
Emissions
Identified
HC
554.8
554.8
553.6
554.8
554.8
553.6
554.8
554.8
553.6
519.7
Slfc?
518.6
519.7
519.7
518.6
490.4
482.0
480.8
472.7
413.6
412.4
411.4
352.3
351.1
411.4
352.3
352.3
351.1
411.4
352.3
339.4
Ways to Pass Criteria"
Excess
Emissions

CO
9889.4
9889.4
9798.5
9889.4
9889.4
9798.5
9889.4
9889.4
9798.5
9715.6
9715.0
9624.8
9715.6
9715.6
9624.8
9639.0
9468.0
9377.1
9592.5
8890.6
8799.8
9268.2
8566.3
8475.5
9268.2
8566.3
8566.3
8475.5
9268.2
8566.3
8243.5
Identified
HC
87%
87%
87%
87%
87%
87%
87%
87%
87%
82%
82%
81%
82%
82%
81%
77%
76%
76%
74%
65%
65%
65%
55%
55%
65%
55%
55%
55%
65%
55%
53%

CO
86%
86%
86%
86%
86%
86%
86%
86%
86%
85%
85%
84%
85%
85%
84%
84%
83%
82%
84%
78%
77%
81%
75%
74%
81%
75%
75%
74%
81%
75%
72%
Number of
Simuiated
Failures
256
256
251
256
256
251
256
256
251
229
229
224
229
229
224
201
174
169
174
119
114
147
92
87
147
92
92
87
147
92
82

Failure
Rate
16%
16%
16%
16%
16%
16%
16%
16%
16%
14%
14%
14%
14%
14%
14%
13%
11%
11%
11%
7%
7%
9%
6%
5%
9%
6%
6%
5%
9%
6%
5%
Number of
Simulated
EC'S
0
0
0
0
0
0
0
0
0
0
O/
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

EC
Rate
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehlcieg
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
H-2

-------
1983-85 Simulated Lane Sample: "Two Ways to Pass Criteria"
PFI Vehicles
Excess Excess
LaneIM240
Cut-Points
Comp+Baf 2
0.6/10 + 0.5/12
0.6/15+0.5/12
0.6/20 + 0.5/12
OJ/10 + 0.5/12
0.7/15+0.5/12
0.7/20 + 0.5/12
0.75/10 + 0.5/12
0.75/15+0.5/12
0.75/20 + 0.5/12
0.8/10 + 0.5/12
0.8/1S +0,5/12.
0.8/20 + 0.5/12
0.85/10 + 0.5/12
0.85/15+0.5/12
0.85/20 + 0.5/12
0.9/10 + 0.5/12
0.9/15+0.5/12
0.9/20 + 0.5/12
1.0/10 + 0.5/12
1.0/15 + 0.5/12
1.0/20 + 0.5/12
1.1/10 + 0.5/12
1.1/15+0.5/12
1.1/20 + 0.5/12
1.2/10 + 0.5/12
1.2/15+0.5/12
1.2/16 + 0.5/12
1.2/20+0.5/12
1.6/10 + 0.5/12
1.6/15 + 0.5/12
1.6/20+0.5/12
Emissions
Identlfled
HC
193.2
193.2
192.3
193.2
193.2
192.3
193.2
193.2
192.3
193.2
193.2
192.3
193.2
193.2
192.3
193.2
187.6
186.7
193.2
187.6
186.7
193.2
187.6
186.7
193.2
187.6
187.6
186.7
193.2
187.6
177.6
Jbf missions

CO
4637.1
4637.1
4566.0
4637.1
4637.1
4566.0
4637.1
4637.1
4566.0
4637.1
4637.1
4566.0
4637.1
4637.1
4566.0
4637.1
4524.6
4453.5
4637.1
4524.6
4453.5
4637.1
4524.6
4453.5
4637.1
4524.6
4524.6
4453.5
4637.1
4524.6
4271.7
Identlfled
H£
93%
93%
92%
93%
93%
92%
93%
93%
92%
93%
93%
92%
93%
93%
92%
93%
90%
90%
93%
90%
90%
93%
90%
90%
93%
90%
90%
90%
93%
90%
85%

CQ
95%
95%
94%
95%
95%
94%
95%
95%
94%
95%
95%
94%
95%
95%
94%
95%
93%
91%
95%
93%
91%
95%
93%
91%
95%
93%
93%
91%
95%
93%
88%
Number of
Simulated
Failures
62
62
58
62
62
58
62
62
58
62
62
58
62
62
58
62
44
40
62
44
40
62
44
40
62
44
44
40
62
44
36

Failure
Rate
36%
36%
34%
36%
36%
34%
36%
36%
34%
36%
36%
34%
36%
36%
34%
36%
26%
24%
36%
26%
24%
36%
26%
24%
36%
26%
26%
24%
36%
26%
21%
Number of
Simulated
Esi
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

EC
Rate
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehicles
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
- 0.0%
0.0%
0.0%
0.0%
0.0%
H-3

-------
1986+ Simulated Lane Sample: "Two Ways to Pass Criteria"
PFI Vehicles
Excess Excess
Lane IM240
Cut-Points
Com^>+Ba0 2
0.6/10 + 0.5/12
0.6/15+0.5/12
0.6/20 + 0.5/12
0.7/10 + 0.5/12
0.7/15+0.5/12
0.7/20 + 0.5/12
0.75/10 + 0.5/12
0.75/15+0.5/12
0.75/20 + 0.5/12
0.8/10 + 0.5/12
&8/1S+&S/I2
0.8/20 + 0.5/12
0.85/10 + 0.5/12
0.85/15+0.5/12
0.85/20 + 0.5/12
0.9/10+0.5/12
0.9/15+0.5/12
0.9/20 + 0.5/12
1.0/10 + 0.5/12
1.0/15+0.5/12
1.0/20 + 0.5/12
1.1/10 + 0.5/12
1.1/15+0.5/12
1.1/20 + 0.5/12
1.2/10 + 0.5/12
1.2/15+0.5/12
1.2/16 + 0.5/12
1.2/20 + 0.5/12
1.6/10 + 0.5/12
1.6/15+0.5/12
1.6/20+0.5/12
Emissions
Identified
H£
350.0
350.0
350.0
350.0
350.0
350.0
350.0
350.0
350.0
313.3
313.3
313.3
313.3
313.3
313.3
282.5
282.5
282.5
263.8
210.7
210.7
199.5
146.3
146.3
199.5
146.3
146.3
146.3
199.5
146.3
146.3
Emissions

CQ
4831.9
4831.9
4831.9
4831.9
4831.9
4831.9
4831.9
4831.9
4831.9
4649.5
4649.5
4649.5
4649.5
4649.5
4649.5
4569.0
4569.0
4569.0
4520.2
3962.9
3962.9
4179.8
3622.4
3622.4
4179.8
3622.4
3622.4
3622.4
4179.8
3622.4
3622.4
Identified
HC
85%
85%
85%
85%
85%
85%
85%
85%
85%
76%
76% "
76%
76%
76%
76%
69%
69%
69%
64%
51%
51%
48%
36%
36%
48%
36%
36%
36%
48%
36%
36%

CQ
79%
79%
79%
79%
79%
79%
79%
79%
79%
76%
76%
76%
76%
76%
76%
75%
75%
75%
74%
65%
65%
69%
60%
60%
69%
60%
60%
60%
69%
60%
60%
Number of
Simulated
Failures
190
190
190
190
190
190
190
190
190
161
i«i\
161
161
161
161
133
133
133
104
75
75
75
46
46
75
46
46
46
75
46
46

Failure
Rate
13%
13%
13%
13%
13%
13%
13%
13%
13%
11%
11*
11%
11%
11%
11%
9%
9%
9%
7%
5%
5%
5%
3%
3%
5%
3%
3%
3%
5%
3%
3%
Number of
Simulated
EC'S
0
0
0
0
0
0
0
0
0
0
a
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

EC
Rate
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehicles
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
H-4

-------
1983+ Simulated Lane Sample
PF1 Vehicles


Idle/2500
Cut-Points
220112
100/1.0-220/1.2
lOO/OJ
Excess
Emissions
Identified
HC
414.3
482.8
487.4
Excess
Emissions
Identified
CQ HC CO
^fcJfc ••«• dbdHb
8316.3 65% 73%
9371.0 76% 82%
9419.7 77% 82%

Number of
Simulated
Failures
188
549
603


Failure
Rate
•MUB
12%
34%
38%

Number of
Simulated EC
EC'S Rate
0 0.0%
82 5.1%
82 5.1%
Failure Rate
for FTP
Passing
Vehicles
0.0%
12.9%
12.9%
1983-1985 Simulated Lane Sample
PFI Vehicles
Excess Excess

Idle/2500
Cut-Points
220/1.2
100/1.0-220/1.2
100/0.5
Emissions
Identified
HC
178.5
181.9
181.9
Emissions
Identified
CO HC CO
4114.2 86% 84%
4189.3 87% 86%
4221.3 87% 87%
Number of
Simulated
Failures
32
68
86

Failure
Rate
19%
40%
51%
Number of
Simulated
EC'S
0
0
0

EC
Rate
0%
0%
0%
Failure Rate
for FTP
Passing
Vehicles
0%
0%
0%
1986+ Simulated Lane Sample
PFI Vehicles
Excess
Excess

ffi ffl^ 3Si Oil^ lls missions
Idle/2500
Cut-Points
220/1.2
100/1.0-220/1.2
100/0.5
Identified
HC
211.6
284.5
289.4

CQ
3714.3
4819.7
4819.7
Identified
HC
51%
69%
70%

CQ
61%
79%
79%

Number of
Simulated
Failures
157
480
508


Failure
Rate
11%
33%
35%

Number of
Simulated
EC'S
0
86
86


EC
Rate
0%
6%
6%
Failure Rate
for FTP
Passing
Vehicles
0%
13%
13%
H-5

-------
Characterize Hammond Indiana 1983+ Test Fleet as of 9/25
Number of




TfichnvivxT Laos CVI244
PR Low
High
Total
Vehicles
Actually
Tested at the
Hammond
LJU&
1505
97
1602

Number of
Vehicles
in the
Lab Sflmole,
55
19
74

Weighting
Factors
for the
Ijib Samnle
27.36
5.11



Number of
Lab Vehicles
Pflssinc FTP
23
1
24

Number of
Normal Lab
Vehicles
Filling FTP
24
2
26
Characterize Hammond Indiana 1983-1985 Test Fleet as of 9/25
Number of
Vehicles
Actually
Tested at the
Hammond
Tf f hllttlVgY I .a fie JM24Q I-flrn*
PFI Low 126
High 44
Total 170

Number of
Vehicles
in the
Lab SftmDle
7
11
18

Weighting
Factors
for the
Lalj §am.p|e,
18.00
4.00



Number of
Lab Vehicles
Passing FTP
1
0
i

Number of
Normal Lab
Vehicles
Failing FTP
5
1
6
Characterize Hammond Indiana 1986+ Test Fleet as of 9/25
Number of
Vehicles
Actually
Tested at the
Hammond
T* ChlWiVKY Lane IM240 Lane
PFI Low 1379
High 53
Total 1432

Number of
Vehicles
in the
Lab Samnle
48
8
56

Weighting
Factors
for the
Lab Samnle
28.73
6.63



Number of
Lab Vehicles
Passim FTP
22
1
23

Number of
Normal Lab
Vehicles
Failing FTP
19
1
20
H-6

-------
Characterize Simulated 1983+ In-Use Fleet
Calculated From the Lab sample and Weighting Factors





Technnlovv
PFI
Number of
Vehicles
Actually
Tested at the
Hammond
Lane
1602

Number of
Simulated
Vehicles Total Excess Emissions
Passing In Simulated Sample
FTPs HC CO
634 636.57 11441.42
Characterize Simulated 1983-1985 In-Use Fleet
Calculated





From the Lab sample and
Number of
Vehicles
Actually
Tested at the
Hammond
Technology LjlUfi
PFI
170
Weighting Factors

Number of
Simulated
Vehicles
Passing
FTPs
18




Total Excess Emissions
In Simulated Sample
H£
208.16






CQ
4874.20
Characterize Simulated 1986+ In-Use Fleet
Calculated From the Lab sample and Weighting Factors





Technology
PFI
Number of
Vehicles
Actually
Tested at the
Hammond
Laos
1432

Number of
Simulated
Vehicles
Passing
FTPs
639



Total Excess Emissions
In Simulated Sample
H£ CO
412.09 6083.07
H-7

-------
          Note: FTP Category Definitions
  Normal Emitters: FTP HC<0.82 and FTP CO<10.2
    High Emitters: FTP HC>0.82 or FTP CO>10.2
Very High Emitters: FTP HCSi.64 or FTP CO>13.6
    Super Emitters: FTP HC^IO.O or FTP CO>150.0
  Distribution of Total Excess Emissions by FTP Emitter Categories
1983+ PFI Sample
Simulated Lane Sample
Emissions
HC (g/mi)
CO (g/mi)
Total
Excess from
Normals
9%
9%
Total.
Excess from
Hiyhs
13%
7%
Total
Excess from
VM-T Hlphs
49%
40%
Total
Excess from
Supers
30%
43%
Total Excess
100%
100%
1983- 1985 PFI Sample
Simulated Lane Sample
Total
Excess from
Emissions Normals
HC(g/mi) 6%
CO (g/mi) 5%

Total Total Total
Excess from Excess from Excess from
Highs YffY Hl?hs Supers Total Excess
4% 43% 47% 100%
3% 45% 48% 100%

1986+ PFI Sample
Simulated Lane Sample
Total
Excess from
RmtwIniM Normals
HC(g/mi) 10%
CO (g/mi) 12%

Total Total Total
Excess from Excess from Excess from
Hlffhs VM-T Highs Supers Total Excess
17% 53% 20% 100%
10% 34% 43% 100%
                             H-8

-------
Distribution of Emissions Identified by FTP Emitter Categories
IM240 1983+ PFI Simulated Lane Sample: "Two Ways to Pass Criteria"
Cut-Point
Contp 4. Rag J
1.6/20 + 0.5/12

1.2/16 + 0.5/12

1.0/15+0.5/12

0.8/15+0.5/12

0.6/10 + 0.5/12


Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
0%
0%
0%
0%
0%
0%
2%
2%
2%
2%

Hl»h«
MAvMtf
0%
0%
0%
0%
0%
0%
9%
1%
8%
1%

V«T Highs
44%
40%
46%
42%
54%
44%
53%
46%
56%
47%

Supers
56%
60%
54%
58%
46%
56%
37%
51%
34%
50%
IM240 1983-1985 PFI Simulated Lane Sample: "Two Ways to Pass Criteria"
Cut-Point
C. OIDD + Bflf 2
1.6/20 + 0.5/12

1.2/16 + 0.5/12

1.0/15 + 0.5/12

0.8/15+0.5/12

0.6/10 + 0.5/12


Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
0%
0%
0%
0%
0%
0%
3%
3%
3%
3%

Highs
mff^ffm
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%

Very Highs
44%
46%
47%
49%
47%
49%
46%
48%
46%
48%

Supers
•Mll|UU>B
56%
54%
53%
51%
53%
51%
51%
50%
51%
50%
IM240 1986+ PFI Simulated Lane Sample: "Two Ways to Pass Criteria"
Cut-Point
Comp *. Bag 2
1.6/20 + 0.5/12

1.2/16 + 0.5/12

1.0/15 + 0.5/12

0.8/15 + 0.5/12

0.6/10 + 0.5/12


Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%

Highs
0%
0%
0%
0%
0%
0%
16%
3%
14%
3%

Very Highs
43%
28%
43%
28%
61%
34%
58%
41%
62%
43%

Supers
57%
72%
57%
72%
39%
66%
26%
56%
24%
54% •
                         H-9

-------
Distribution of Emissions Identified by FTP Emitter Categories
Idle/2500
Cut-point
HC/CO
220/1.2

100/1.0-220/1.2

100/0.5

1983+ PFI Simulated Lane Sample

Emission , Normals Hlyh.i
HC 0% 9%
CO 0% 4%
HC 5% 11%
CO 6% 4%
HC 6% 11%
CO 6% 4%


Very Highs
45%
36%
44%
37%
44%
37%


Supers
46%
60%
39%
53%
39%
53%
Idle/2500 1983-1985 PFI Simulated Lane Sample
Cut-point
HC/CO
220/1.2

100/1.0-220/1.2

100/0.5


Emission
HC
CO
HC
CO
HC
CO

Normals
0%
0%
2%
2%
2%
3%

Hlyhs
0%
0%
0%
0%
0%
0%

Very Hiahs
45%
44%
44%
43%
44%
43%

Supers
55%
56%
54%
55%
54%
55%
Idle/2500 1986+ PFI Simulated Lane Sample
Cut-point
HC/CO
220/1.2

100/1.0-220/1.2

100/0.5


Rmlssfnn
HC
CO
HC
CO
HC
CO

Normals
0%
1%
7%
10%
9%
10%

Highs.
18%
9%
20%
8%
20%
8%

Very High*
42%
20%
43%
28%
43%
28%

Supers
39%
70%
29%
54%
29%
54%
                          H-10

-------
1983+ Simulated Lane Sample: "Full IM240 Test Criteria"
PFI Vehicles
Excess Excess

LaneIM240
rut-Polnts
0.6/10
0.6/15
0.6/20
0.7/10
0.7/15
0.7/20
0.75/10
0.75/15
0.75/20
0.8/10
0.8/15
0.8/20
0.85/10
0.85/15
0.85/20
0.9/10
0.9/15
0.9/20
1.0/10
1.0/15
1.0/20
1.1/10
1.1/15
1.1/20
1.2/10
1.2/15
1.2/16
1.2/20
1.6/10
1.6/15
1.600
Emissions
Identified
H£
568.2
568.2
567.0
568.2
568.2
567.0
560.0
560.0
558.8
519.7
sm? "
518.6
519.7
519.7
518.6
490.4
482.0
480.8
472.7
413.6
412.4
411.4
352.3
351.1
411.4
352.3
352.3
351.1
411.4
352.3
339.4
Emissions

CQ
10032.7
9984.0
9893.2
10032.7
9984.0
9893.2
9961.9
9913.2
9822.3
9764.3
9715.0
9624.8
9764.3
9715.6
9624.8
9687.7
9468.0
9377.1
9641.2
8890.6
8799.8
9316.9
8566.3
8475.5
9316.9
8566.3
8566.3
8475.5
9316.9
8566.3
8243.5
Identified
HC
89%
89%
89%
89%
89%
89%
88%
88%
88%
82%
" $m
81%
82%
82%
81%
77%
76%
76%
74%
65%
65%
65%
55%
55%
65%
55%
55%
55%
65%
55%
53%

£Q
88%
87%
86%
88%
87%
86%
87%
87%
86%
85%
85%'
84%
85%
85%
84%
85%
83%
82%
84%
78%
77%
81%
75%
74%
81%
75%
75%
74%
81%
75%
72%
Number of
Simulated
Failure!!
398
371
366
398
371
366
343
316
306
261
234-
224
261
234
224
234
179
169
206
124
114
179
97
87
179
97
92
87
179
97
82

Failure
Rate
25%
23%
23%
25%
23%
23%
21%
20%
19%
16%
x 15%- ,
14%
16%
15%
14%
15%
11%
11%
13%
8%
7%
11%
6%
5%
11%
6%
6%
5%
11%
6%
5%
Number of
Simulated
EC'S
32
32
32
32
32
32
*
32
32
27
5
5 ,
0
5
5
0
5
5
0
5
5
0
5
5
0
5
5
0
0
5
5
0

EC
Rate
2.0%
2.0%
2.0%
2.0%
2.0%
2.0%
2.0%
2.0%
1.7%
0.3%
0,3%
0.0%
0.3%
0.3%
0.0%
0.3%
0.3%
0.0%
0.3%
0.3%
0.0%
0.3%
0.3%
0.0%
0.3%
0.3%
0.0%
0.0%
0.3%
0.3%
0.0%
Failure Rate
for FTP
Passing
Vehicles
5.1%
5.1%
5.1%
5.1%
5.1%
5.1%
5.1%
5.1%
4.3%
0.8%
0,8%
0.0%
0.8%
0.8%
0.0%
0.8%
0.8%
0.0%
0.8%
0.8%
0.0%
0.8%
0.8%
0.0%
0.8%
0.8%
0.0%
0.0%
0.8%
0.8%
0.0%
H-11

-------
1983.1985 Simulated Lane Sample: "Full IM240 Test Criteria"
PFI Vehicles
Excess Excess
Emissions Emissions Number of
LaneIM240 Identified Identified Simulated
Cut-Points
0.6/10
0.6/15
0.6/20
0.7/10
0.7/15
0.7/20
0.75/10
0.75/15
0.75/20
0.8/10
0.8/15
0.8/20
0.85/10
0.85/15
0.85/20
0.9/10
0.9/15
0.9/20
1.0/10
1.0/15
1.0/20
1.1/10
1.1/15
1.1/20
1.2/10
1.2/15
1.2/16
1.2/20
1.6/10
1.6/15
1.6/20
H£
196.6
196.6
195.7
196.6
196.6
195.7
*
196.6
196.6
195.7
193.2
193.2
192.3
193.2
193.2
192.3
193.2
187.6
186.7
193.2
187.6
186.7
193.2
187.6
186.7
193.2
187.6
187.6
186.7
193.2
187.6
177.6
CO
4684.8
4652.8
4581.6
4684.8
4652.8
4581.6
4684.8
4652.8
4581.6
4669.2
4637-1.
4566.0
4669.2
4637.1
4566.0
4669.2
4524.6
4453.5
4669.2
4524.6
4453.5
4669.2
4524.6
4453.5
4669.2
4524.6
4524.6
4453.5
4669.2
4524.6
4271.7
ac
94%
94%
94%
94%
94%
94%
94%
94%
94%
93%
93%
92%
93%
93%
92%
93%
90%
90%
93%
90%
90%
93%
90%
90%
93%
90%
90%
90%
93%
90%
85%
co
96%
95%
94%
96%
95%
94%
96%
95%
94%
96%
95%
94%
96%
95%
94%
96%
93%
91%
96%
93%
91%
96%
93%
91%
96%
93%
93%
91%
96%
93%
88%
Failures
98
80
76
98
80
76
98
80
76
80
62
58
80
62
58
80
44
40
80
44
40
80
44
40
80
44
44
40
80
44
36
Failure
Rate
58%
47%
45%
58%
47%
45%
58%
47%
45%
47%
36%
34%
47%
36%
34%
47%
26%
24%
47%
26%
24%
47%
26%
24%
47%
26%
26%
24%
47%
26%
21%
Number of
Simulated EC
Ecja
0
0
0
0
0
0
0
0
0
0
tf
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Rate
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0,0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Vehicles
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
H-12

-------
1986+ Simulated Lane Sample: "Full EVI240 Test Criteria"
PFI Vehicles
Excess Excess

LaneIM240
::nt- Points
0.6/10
0.6/15
0.600
0.7/10
0.7/15
0.7/20
0.75/10
0.75/15
0.75/20
0.8/10
0.8/15
0.8/20
0.85/10
0.85/15
0.85/20
0.9/10
0.9/15
0.9/20
1.0/10
1.0/15
1.0/20
1.1/10
1.1/15
1.1/20
1.2/10
1.2/15
1.2/16
1.2/20
1.6/10
1.6/15
1.6/20
Emissions
Identified
HE
358.7
358.7
358.7
358.7
358.7
358.7
350.0
350.0
350.0
313.3
313.3
313.3
313.3
313.3
313.3
282.5
282.5
282.5
263.8
210.7
210.7
199.5
146.3
146.3
199.5
146.3
146.3
146.3
199.5
146.3
146.3
Emissions

£2
4906.3
4906.3
4906.3
4906.3
4906.3
4906.3
4831.9
4831.9
4831.9
4649.5
4649,5
4649.5
4649.5
4649.5
4649.5
4569.0
4569.0
4569.0
4520.2
3962.9
3962.9
4179.8
3622.4
3622.4
4179.8
3622.4
3622.4
3622.4
4179.8
3622.4
3622.4
Identified
HE
87%
87%
87%
87%
87%
87%
85%
85%
85%
76%
76%
76%
76%
76%
76%
69%
69%
69%
64%
51%
51%
48%
36%
36%
48%
36%
36%
36%
48%
36%
36%

£tt
81%
81%
81%
81%
81%
81%
79%
79%
79%
76%
76%
76%
76%
76%
76%
75%
75%
75%
74%
65%
65%
69%
60%
60%
69%
60%
60%
60%
69%
60%
60%
Number of
Simulated
Failures
283
283
283
283
283
283
225
225
219
168
tm
161
168
168
161
139
139
133
110
82
75
82
53
46
82
53
46
46
82
53
46

Failure
Rate
20%
20%
20%
20%
20%
20%
16%
16%
15%
12%
12&
11%
12%
12%
11%
10%
10%
9%
8%
6%
5%
6%
4%
3%
6%
4%
3%
3%
6%
4%
3%
Number of
Simulated
EC'S
35
35
35
35
35
35
35
35
29
7
7
0
7
7
0
7
7
0
7
7
0
7
7
0
7
7
0
0
7
7
0

EC
Rate
2.5%
2.5%
2.5%
2.5%
2.5%
2.5%
2.5%
2.5%
2.0%
0.5%
0.5%
0.0%
0.5%
0.5%
0.0%
0.5%
0.5%
0.0%
0.5%
0.5%
0.0%
0.5%
0.5%
0.0%
0.5%
0.5%
0.0%
0.0%
0.5%
0.5%
0.0%
Failure Rate
for FTP
Passing
Vehicles
5.5%
5.5%
5.5%
5.5%
5.5%
5.5%
*
5.5%
5.5%
4.5%
1.0%
1.0%
0.0%
1.0%
1.0%
0.0%
1.0%
1.0%
0.0%
1.0%
1.0%
0.0%
1.0%
1.0%
0.0%
1.0%
1.0%
0.0%
0.0%
1.0%
1.0%
0.0%
H-13

-------
Distribution of Emissions Identified by FTP Emitter Categories
IM240
Cut-Point
HC/CO
1.6/20

1.2/16

1.0/15

0.8/15

0.6/10

1983+ PFI Simulated Lane Sample: "Full IM240 Test Criteria"

Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
0%
0%
0%
0%
0%
0%
2%
2%
4%
•3%

Highs
0%
0%
0%
0%
0%
0%
9%
1%
8%
1%

Very Highs
44%
40%
46%
42%
54%
44%
53%
46%
54%
46%

Supers
56%
60%
54%
58%
46%
56%
37%
51%
33%
49%
IM240
Cut-Point
HC/CO
1.6/20

1.2/16

1.0/15

0.8/15

0.6/10

1983-1985 PFI Simulated Lane Sample: "Full IM240 Test Criteria"

Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
0%
0%
0%
0%
0%
0%
3%
3%
5%
4%

Highs
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%

VM-T Highs
44%
46%
47%
49%
47%
49%
46%
48%
45%.
47%

Supers
56%
54%
53%
51%
53%
51%
51%
50%
50%
49%
1M240
Cut-Point
HC/CO
1.6/20

1.2/16

1.0/15

0.8/15

0.6/10

1986+ PFI Simulated

Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO
Lane Sample:

"Full IM240 Test Criteria"



NjjrjUflis Hjghs Vwv Highs
0%
0%
0%
0%
0%
0%
0%
0%
2%
2%
0%
0%
0%
0%
0%
0%
16%
3%
14%
3%
43%
28%
43%
28%
61%
34%
58%
41%
61%
43%


Supers
57%
72%
57%
72%
39%
66%
26%
56%
23%
53%
                          H-14

-------
1983+ Simulated Lane Sample: "Two Ways to Pass Criteria"
TBI Vehicles
Excess Excess
LaneIM240 Emissions Emissions Number of
Cut-Points
Comn+Bav 2
0.6/10 + 0.5/12
0.6/15+0.5/12
0.6/20 + 0.5/12
0.7/10 + 0.5/12
0.7/15+0.5/12
0.7/20 + 0.5/12
0.75/10 + 0.5/12
0.75/15+0.5/12
0.75/20 + 0.5/12
0.8/10 + 0.5/12
0.8/15+0.5/12
0.8/20 + 0.5/12
0.85/10 + 0.5/12
0.85/15+0.5/12
0.85/20+0.5/12
0.9/10 + 0.5/12
0.9/15+0.5/12
0.9/20 + 0.5/12
1.0/10+0.5/12
1.0/15+0.5/12
1.0/20 + 0.5/12
1.1/10 + 0.5/12
1.1/15+0.5/12
1.1/20 + 0.5/12
1.2/10 + 0.5/12
1.2/15+0.5/12
1.2/16 + 0.5/12
1.2/20 + 0.5/12
1.6/10+0.5/12
1.6/15+0.5/12
1.600+0.5/12
Identified
HC
761.7
754.0
753.5
754.8
747.2
723.8
754.8
747.2
723.8
754.8
740.4,
717.0
754.8
740.4
717.0
750.1
735.7
712.3
722.9
697.1
673.8
698.4
578.5
547.3
698.4
578.5
564.8
547.3
684.7
539.5
501.0
Identified
CO
10334.1
9949.6
9939.4
10282.4
9897.9
9077.9
10282.4
9897.9
9077.9
10282.4
97«&5
8943.4
10282.4
9763.5
8943.4
10273.0
9754.1
8934.1
10042.5
9366.4
8546.4
9938.1
8451.9
7482.2
9938.1
8451.9
7959.7
7482.2
9825.6
8188.8
7108.0
HC
•AAfc
92%
91%
91%
91%
90%
87%
91%
90%
87%
91%
89%
87%
91%
89%
87%
91%
89%
86%
87%
84%
81%
84%
70%
66%
84%
70%
68%
66%
83%
65%
61%

84%
81%
81%
84%
81%
74%
84%
81%
74%
84%
80%
73%
84%
80%
73%
84%
80%
73%
82%
77%
70%
81%
69%
61%
81%
69%
65%
61%
80%
67%
58%
Simulated
Failures
445
403
398
424
381
346
424
381
346
424
360
325
424
360
325
403
339
303
381
296
261
360
168
128
360
168
147
128
355
145
100
Failure
Rate
26%
23%
23%
25%
22%
20%
25%
22%
20%
25%
21%
19%
25%
21%
19%
23%
20%
18%
22%
17%
15%
21%
10%
7%
21%
10%
9%
7%
21%
8%
6%
Number of
Simulated
EC'S
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

EC
Rate
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehicles
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
H-15

-------
1983-85 Simulated Lane Sample: "Two Ways to Pass Criteria"
TBI Vehicles
Lane IM240
Cut-Points
Comp+Baff 2
0.6/10 + 0.5/12
0.6/15+0.5/12
0.6/20 + 0.5/12
0.7/10 + 0.5/12
0.7/15+0.5/12
0.7/20 + 0.5/12
0.75/10 + 0.5/12
0.75/15 + 0.5/12
0.75/20 + 0.5/12
0.8/10 + 0.5/12

OS/13 +OS/J2.
0.8/20 + 0.5/12
0.85/10 + 0.5/12
0.85/15+0.5/12
0.85/20 + 0.5/12
0.9/10 + 0.5/12
0.9/15+0.5/12
0.9/20 + 0.5/12
1.0/10 + 0.5/12
1.0/15+0.5/12
1.0/20 + 0.5/12
1.1/10 + 0.5/12
1.1/15+0.5/12
1.1/20 + 0.5/12
1.2/10 + 0.5/12
1.2/15+0.5/12
1.2/16 + 0.5/12
1.2/20 + 0.5/12
1.6/10 + 0.5/12
1.6/15+0.5/12
1.6/20 + 0.5/12
Admissions
Identified
HC
359.9
359.7
359.7
355.1
355.0
338.3
355.1
355.0
338.3
355.1

3S&3
333.6
355.1
350.3
333.6
351.9
347.0
330.4
351.9
339.2
322.5
351.9
286.4
269.8
351.9
286.4
277.0
269.8
335.6
239.9
223.2
Emissions
Identified
CO
5740.5
5661.0
5661.0
5704.6
5625.1
4958.1
5704.6
5625.1
4958.1
5704.6

5SBL9
4864.8
5704.6
5531.9
4864.8
5698.1
5525.4
4858.3
5698.1
5416.4
4749.3
5698.1
4972.7
4305.6
5698.1
4972.7
4631.3
4305.6
5563.7
4658.4
3991.3
HC
90%
90%
90%
89%
89%
84%
89%
89%
84%
89%

87%,
83%
89%
87%
83%
88%
87%
82%
88%
85%
80%
88%
71%
67%
88%
71%
69%
67%
84%
60%
56%
CO
87%
86%
86%
87%
86%
75%
87%
86%
75%
87%

84%
74%
87%
84%
74%
87%
84%
74%
87%
82%
72%
87%
76%
65%
87%
76%
70%
65%
85%
71%
61%
Number of
Simulated
Failures
244
229
229
229
215
194
229
215
194
229

200
179
229
200
179
215
185
165
215
170
150
215
111
91
215
111
96
91
209
83
62

Failure
Rate
44%
41%
41%
41%
38%
35%
*
41%
38%
35%
41%
•• ••
36%
32%
41%
36%
32%
38%
33%
29%
38%
30%
27%
38%
20%
16%
38%
20%
17%
16%
37%
15%
11%
Number of
Simulated
EC'S
0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

EC
Rafa
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%

0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehicles
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%

00%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
H-16

-------
198«+ Simulated Lane Sample: "Two Ways to Pass Criteria"
TB1 Vehicles
Excess Excess
LaneIM240
Cut-Points
Comp+Baf 2
0.6/10 + 0.5/12
0.6/15+0.5/12
0.6/20 + 0.5/12
0.7/10 + 0.5/12
0.7/15+0.5/12
0.7/20 + 0.5/12
0.75/10 + 0.5/12
0.75/15+0.5/12
0.75/20 + 0.5/12
0.8/10 + 0.5/12
0.8/15+0.3/12.
0.8/20 + 0.5/12
0.85/10 + 0.5/12
0.85/15+0.5/12
0.85/20 + 0.5/12
0.9/10 + 0.5/12
0.9/15+0.5/12
0.9/20 + 0.5/12
1.0/10 + 0.5/12
1.0/15 +0.5/12
1.0/20 + 0.5/12
1.1/10 + 0.5/12
1.1/15+0.5/12
1.1/20 + 0.5/12
1.2/10 + 0.5/12
1.2/15+0.5/12
1.2/16+0.5/12
1.2/20 + 0.5/12
1.6/10 + 0.5/12
1.6/15+0.5/12
1.6/20 + 0.5/12
Emissions
Identified
HC
359.1
350.0
349.6
359.1
350.0
347.1
359.1
350.0
347.1
359.1
'- 33Q&,
347.1
359.1
350.0
347.1
359.1
350.0
347.1
325.7
316.6
313.6
295.7
264.3
255.2
295.7
264.3
264.3
255.2
295.7
264.3
249.4
Emissions
Identified
CO
dKdb
4191.7
3860.8
3852.7
4191.7
3860.8
3817.0
4191.7
3860.8
3817.0
4191.7
"awe*-
3817.0
4191.7
3860.8
3817.0
4191.7
3860.8
3817.0
3909.1
3578.2
3534.4
3781.1
3241.3
3078.6
3781.1
3241.3
3241.3
3078.6
3781.1
3241.3
2990.4
HC
••Afe
98%
95%
95%
98%
95%
94%
98%
95%
94%
98%
95%
94%
98%
95%
94%
98%
95%
94%
89%
86%
85%
80%
72%
69%
80%
72%
72%
69%
80%
72%
68%
CQ
JUfc
83%
76%
76%
83%
76%
76%
83%
76%
76%
83%
, 76% <
76%
83%
76%
76%
83%
76%
76%
77%
71%
70%
75%
64%
61%
75%
64%
64%
61%
75%
64%
59%
Number of
Simulated
Failures
^Jj^f|mg£
157
131
127
157
131
124
157
131
124
157
111
124
157
131
124
157
131
124
131
105
97
105
53
41
105
53
53
41
105
53
38

Failure
Rate
AJHJUI
14%
11%
11%
14%
11%
11%
14%
11%
11%
14%
- 11*- '
11%
14%
11%
11%
14%
11%
11%
11%
9%
8%
9%
5%
4%
9%
5%
5%
4%
9%
5%
3%
Number of
Simulated
geis
0
0
0
0
0
0
0
0
0
0
f>>- '
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

EC
Rate
AHUk
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
oo*.
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehicles
^AHUUMt
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
H-17

-------
1983+ Simulated Lane Sample
TBI Vehicles
Excess Excess

Idle/2500
Cut-Points
220/1.2
100/1.0+220/1.2
100/0.5
Emissions Emissions
Identified Identified
HC CO HC CQ
591.6 8196.5 71% 67%
745.5 9947.1 90% 81%
746.7 10039.1 90% 82%
Number of
Simulated
Failures
249
625
715

Failure
Rate
14%
36%
42%
Number of
Simulated
EC'S
21
64
111

EC
Rate
1.2%
3.7%
6.5%
Failure Rate
for FTP
Passing
Vehicles
3.9%
11.6%
20.2%
1983-1985 Simulated Lane Sample
TBI Vehicles
Excess

Idle/2500
Cut-Points
220/1.2
100/1.0+220/1.2
100/0.5
Emissions
Identified
HC
289.7
375.4
376.3
Excess

Emissions

01
5007.8
6038.5
6070.0
Identified
HC
72%
94%
94%

SSL
76%
92%
92%

Nnml>er of
Simulated
Failures
138
347
362


Failure
Rate
25%
62%
65%

Number of
Simulated
EC'S
0
0
0


EC
Rate
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehicles
0.0%
0.0%
0.0%
1986+ Simulated Lane Sample
TBI Vehicles
Excess

Idle/2500
Cut-Points
220/1.2
100/1.04-220/1.2
100/0.5
Emissions
Identified
HC CO
272.9 3051.9
328.5 3543.5
328.5 3600.7
Excess
Emissions
Identified
HC CO
74% 60%
89% 70%
89% 71%

Number of
Simulated
Failures
105
209
291


Failure
Rate
9%
18%
25%

Number of
Simulated
EC'S
26
78
134


EC
Rate
2.2%
6.7%
11.6%
Failure Rate
for FTP
Passing
Vehicles
4.7%
14.0%
24.0%
H-18

-------
Characterize Hammond Indiana 1983+ Test Fleet as of 9/25
Number of


T«ehnology Lane IM240
TBI Low
High
Total
• Vehicles
Actually
Tested at the
Hammond
Lans
1555
166
1721

Number of
Vehicles
in the
73
35
108

Weighting
Factors
for the
Lab Sample
21.30
4.74

Number of
Lab Vehicles
Pawlny FTP
25
4
29

Number of
Normal Lab
Vehicles
Failing FT^
32
6
38
Characterize Hammond Indiana 1983-1985 Test Fleet as of 9/25
Number of
Vehicles
Actually
Tested at the
Hammond
TtcfaovlvgY Lant IM24Q i^ane
TBI Low 458
High 102
Total 560

Number of
Vehicles
in the
Lab Sample
31
18
49

Weighting
Factors
for the
Lab SfimDle
14.77
5.67



Number of
Lab Vehicles
Passing FTP
4
1
5

Number of
Normal Lab
Vehicles
Falling FTP
16
0
16
Characterize Hammond Indiana 1986+ Test Fleet as of 9/25
Number of
Vehicles
Actually
Tested at the
Hammond
Ttchnnjoyy Lane IM240 t.ane
TBI Low 1097
High 64
Total 1161

Number of
Vehicles
In the
Lab Sample
42
17
59

Weighting
Factors
for the
Lab Sample
26.12
3.76



Number of
Lab Vehicles
Passing FTP
21
3
24

Number of
Normal Lab
Vehicles
Falling FTP
16
6
22
H-19

-------
Characterize Simulated 1983+ In-Use Fleet
Calculated From the Lab sample and Weighting Factors





Technology
TBI
Number of
Vehicles
Actually
Tested at the
Hammond
Ljujg
1721

Number of
Simulated
Vehicles
Passing
FTPs
552



Total Excess Emissions
In Simulated Sample
B£
827.69





CO,
12239.31
Characterize Simulated 1983-1985 In-Use Fleet
Calculated From the Lab sample and Weighting Factors





Technology
TBI
Number of
Vehicles
Actually
Tested at the
Hammond
LJUB.
560

Number of
Simulated
Vehicles Total Excess Emissions
Passing In Simulated Sample
FTPs HC CO
65 400.94 6575.68
Characterize Simulated 1986+ In-Use Fleet
Calculated From the Lab sample and Weighting Factors
   Technology
      TBI
 Number of
  Vehicles
  Actually
Tested at the
 Hammond
   Lao*

   1161
                                    Number of
                                    Simulated
                                     Vehicles
                                      FTPs
560
          Total Excess Emissions
          In Simulated Sample
                HC             CQ
367.49
5051.31
                                           H-20

-------
          Note: FTP Category Definitions
  Normal Emitters: FTP HC<0.82 and FTP CO<10.2
     High Emitters: FTP HC»0.82 or FTP COS10.2
Very High Emitters: FTP HC>1.64 or FTP COS13.6
    Super Emitters: FTP HC^IO.O or FTP COS150.0
  Distribution of Total Excess Emissions by FTP Emitter Categories
1983+ TBI Sample
Simulated Lane Sample



Total
Excess from
F.mlgrfons
HC(g/mi)
C0(g/mi)
Nqrrfla's
6%
11%


Total
Excess from
Highs
18%
15%


Total
Excess from
YflT Hfghg
31%
34%


Total
Excess from
SiiDers
45%
39%




Total Eix.ce.ss
100%
100%

1983-1985 TBI Sample
Simulated Lane Sample



Total
Excess from
EjoissioQs,
HC(g/mi)
CO(g/mi)
Normals
8%
8%


Total
Excess from
Mff
19%
16%


Total
Excess from
VHT HlrfM
49%
46%


Total
Excess from
Supers
24%
31%




To(a|| Excel"1
100%
100%
f
I98t> TBI Sample
Simulated Lane Sample



Total
Excess from
Ejoissisuu
HC(g/mi)
CO (g/mi)
Normals
2%
15%


Total
Excess from
Hlchs
16%
11%


Total
Excess from
Very Highs
20%
25%


Total
Excess from
Supers
62%
49%




Total Excess
100%
100%
                           H-21

-------
Distribution of Emissions Identified by FTP Emitter Categories
1M 240 1983+ TBI Simulated Lane Sample: "Two Ways to Pass Criteria"
Cut-Point
Como + Bay 2
1.6/20 + 0.5/12

1.2/16 + 0.5/12

1.0/15+0.5/12

0.8/15+0.5/12

0.6/10+0.5/12


Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
1%
1%
1%
1%
1%
2%
2%
2%
4%
5%

Highs
1%
1%
3%
2%
14%
7%
14%
9%
14%
8%

Very Highs
24%
30%
31%
36%
32%
39%
34%
40%
34%
40%

Supers
74%
68%
66%
60%
53%
51%
50%
49%
49%
47%
IM240 1983-1985 TBI Simulated Lane Sample: "Two Ways to Pass Criteria"
Cut-Point
Comp + Bag 2
1.6/20 + 0.5/12

1.2/16 + 0.5/12

1.0/15 + 0.5/12

0.8/15+0.5/12

0.6/10 + 0.5/12


Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
0%
0%
0%
0%
1%
1%
2%
1%
5%
5%

Hlyhs
4%
2%
6%
4%
12%
6%
14%
8%
14%
8%

Very Highs
52%
47%
58%
53%
58%
56%
56%
55%
55%
53%

Supers
44%
50%
35%
43%
29%
37%
28%
36%
27%
35%
IM240 198«+ TBI Simulated Lane Sample: "Two Ways to Pass Criteria"
Cut-Point
Oomn 4- Ran 2
1.6/20 + 0.5/12

1.2/16 + 0.5/12

1.0/15+0.5/12

0.8/15 + 0.5/12

0.6/10 + 0.5/12


Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
1%
2%
1%
2%
1%
2%
1%
2%
1%
2%

Highs
0%
0%
1%
1%
17%
10%
16%
10%
15%
9%

Very Highs
7%
15%
11%
20%
9%
18%
18%
24%
20%
30%

Supers
92%
83%
87%
77%
72%
69%
65%
64%
64%
59%
                          H-22

-------
Distribution of Emissions Identified by FTP Emitter Categories
Idle/2500 1983+ TBI Simulated Lane Sample
Cut-point
HC/CO Emission Normals
220/1.2

100/1.0-220/1.2

100/0.5

HC
CO
HC
CO
HC
CO
1%
2%
4%
5%
4%
6%
Highs
11%
5%
16%
11%
16%
11%
Very Highs
30%
43%
35%
42%
34%
42%
Supers
57%
50%
45%
41%
45%
41%
Idle/2500 1983-1985 TBI Simulated Lane Sample
Cut-point
HC/CO
220/1.2

100/1.0-220/1.2

100/0.5


Emission
HC
CO
HC
CO
HC
CO

Normals
2%
2%
5%
6%
6%
6%

Hlohi
11%
5%
16%
11%
16%
11%

Very Hlgha
53%
53%
52%
50%
52%
50%

Supers
34%
40%
26%
33%
26%
33%
Idle/2500 1986+ TBI Simulated Lane Sample
Cut-point
HC/CO Emission Normals
220/1.2
100/1.0-220/1.2
100/0.5
HC
CO
HC
CO
HC
CO
0%
1%
0%
1%
0%
2%
Highs
11%
4%
16%
10%
16%
9%
Vtrv Hlgha
14%
32%
22%
36%
22%
35%
Supers
75%
63%
62%
54%
62%
53%
                           H-23

-------
1983+ Simulated Lane Sample: "Full 1M240 Test Criteria"
TBI Vehicles
Excess Excess

LaneIM240
Cut-Points
0.6/10
0.6/15
0.6/20
0.7/10
0.7/15
0.7/20
0.75/10
0.75/15
0.75/20
0.8/10
O.S/15
0.8/20
0.85/10
0.85/15
0.85/20
0.9/10
0.9/15
0.9/20
1.0/10
1.0/15
1.0/20
1.1/10
1.1/15
1.1/20
1.2/10
1.2/15
1.2/16
1.2/20
1.6/10
1.6/15
1.6/20
Emissions
Identified
HC
774.9
767.2
766.7
754.8
747.2
723.8
754.8
747.2
723.8
754.8
740,4
717.0
754.8
740.4
717.0
750.1
735.7
712.3
722.9
697.1
673.8
698.4
578.5
547.3
698.4
578.5
564.8
547.3
684.7
539.5
501.0
Emissions
Identified
CQ
10550.8
10103.6
10093.4
10345.0
9897.9
9077.9
10345.0
9897.9
9077.9
10345.0
9763,5,
8943.4
10345.0
9763.5
8943.4
10335.6
9754.1
8934.1
10105.1
9366.4
8546.4
10000.7
8451.9
7482.2
10000.7
8451.9
7959.7
7482.2
9888.2
8188.8
7108.0
HC
94%
93%
93%
91%
90%
87%
91%
90%
87%
91%
89%"
87%
91%
89%
87%
91%
89%
86%
87%
84%
81%
84%
70%
66%
84%
70%
68%
66%
83%
65%
61%
CQ
86%
83%
82%
85%
81%
74%
85%
81%
74%
85%
80%
73%
85%
80%
73%
84%
80%
73%
83%
77%
70%
82%
69%
61%
82%
69%
65%
61%
81%
67%
58%
Number of
Simulated
Failures
528
464
455
464
400
360
464
400
360
464
379
334
464
379
334
443
358
313
422
315
270
400
187
138
400
187
166
138
396
164
104

Failure
Rate
31%
27%
26%
27%
23%
21%
27%
23%
21%
27%
22%
19%
27%
22%
19%
26%
21%
18%
24%
18%
16%
23%
11%
8%
23%
11%
10%
8%
23%
10%
6%
Number of
Simulated
EC'S
57
40
36
19
19
14
19
19
14
19
, ***
9
19
19
9
19
19
9
19
19
9
19
19
9
19
19
19
9
19
19
5

EC
Rate
3.3%
2.3%
2.1%
1.1%
1.1%
0.8%
1.1%
1.1%
0.8%
1.1%
14%
0.6%
1.1%
1.1%
0.6%
1.1%
1.1%
0.6%
1.1%
1.1%
0.6%
1.1%
1.1%
0.6%
1.1%
1.1%
1.1%
0.6%
1.1%
1.1%
0.3%
Failure Rate
for FTP
Passing
Vehicles
10.3%
7.3%
6.4%
3.4%
3.4%
2.6%
3.4%
3.4%
2.6%
3.4%
3A%
1.7%
3.4%
3.4%
1.7%
3.4%
3.4%
1.7%
3.4%
3.4%
1.7%
3.4%
3.4%
1.7%
3.4%
3.4%
3.4%
1.7%
3.4%
3.4%
0.9%
H-24

-------
1983-1985 Simulated Lane Sample: "Full IM240 Test Criteria"
TBI Vehicles
Excess Excess

LaneIM240
::ut. Points
0.6/10
0.6/15
0.6/20
0.7/10
0.7/15
0.7/20
0.75/10
0.75/15
0.75/20
0.8/10
0.8/15
0.8/20
0.85/10
0.85/15
0.85/20
0.9/10
0.9/15
0.9/20
1.0/10
1.0/15
1.0/20
1.1/10
1.1/15
1.1/20
1.2/10
1.2/15
1.2/16
1.2/20
1.6/10
1.6/15
1.6/20
Emissions
Identified
HC
369.0
368.9
368.9
355.1
355.0
338.3
355.1
355.0
338.3
355.1
330.3^
333.6
355.1
350.3
333.6
351.9
347.0
330.4
351.9
339.2
322.5
351.9
286.4
269.8
351.9
286.4
277.0
269.8
335.6
239.9
223.2
Emissions

£Q
5847.3
5767.8
5767.8
5704.6
5625.1
4958.1
5704.6
5625.1
4958.1
5704.6
5S3L9
4864.8
5704.6
5531.9
4864.8
5698.1
5525.4
4858.3
5698.1
5416.4
4749.3
5698.1
4972.7
4305.6
5698.1
4972.7
4631.3
4305.6
5563.7
4658.4
3991.3
Identified
HC
92%
92%
92%
89%
89%
84%
89%
89%
84%
89%
87%
83%
89%
87%
83%
88%
87%
82%
88%
85%
80%
88%
71%
67%
88%
71%
69%
67%
84%
60%
56%

£Q
89%
88%
88%
87%
86%
75%
87%
86%
75%
87%
84%
74%
87%
84%
74%
87%
84%
74%
87%
82%
72%
87%
76%
65%
87%
76%
70%
65%
85%
71%
61%
Number of
Simulated
Failures
279
265
265
235
220
200
235
220
200
235
205
185
235
205
185
220
191
170
220
176
155
220
117
96
220
117
102
96
215
88
62

Failure
Rate
50%
47%
47%
42%
39%
36%
42%
39%
36%
42%
37%
33%
42%
37%
33%
39%
34%
30%
39%
31%
28%
39%
21%
17%
39%
21%
18%
17%
38%
16%
11%
Number of
Simulated
EC'S
20
20
20
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
0

EC
Rate
3.7%
3.7%
3.7%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
0.0%
Failure Rate
for FTP "
Passing
Vehicles
31.6%
31.6%
31.6%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
8.7%
0.0%
H-25

-------
1986+ Simulated Lane Sample: "Full IM240 Test Criteria"
TBI Vehicles
Excess Excess

LaneIM240
Cut-Polnts
0.6/10
0.6/15
0.6/20
0.7/10
0.7/15
0.7/20
0.75/10
0.75/15
0.75/20
0.8/10
(WJ3
0.8/20
0.85/10
0.85/15
0.85/20
0.9/10
0.9/15
0.9/20
1.0/10
1.0/15
1.0/20
1.1/10
1.1/15
1.1/20
1.2/10
1.2/15
1.2/16
1.2/20
1.6/10
1.6/15
1.6/20
Emissions
Identified
HC
359.1
350.0
349.6
359.1
350.0
347.1
359.1
350.0
347.1
359.1
33ftO
347.1
359.1
350.0
347.1
359.1
350.0
347.1
325.7
316.6
313.6
295.7
264.3
255.2
295.7
264.3
264.3
255.2
295.7
264.3
249.4
Emissions
Identified
CO
4268.5
3860.8
3852.7
4268.5
3860.8
3817.0
4268.5
3860.8
3817.0
4268.5
3860.8
3817.0
4268.5
3860.8
3817.0
4268.5
3860.8
3817.0
3985.9
3578.2
3534.4
3857.9
3241.3
3078.6
3857.9
3241.3
3241.3
3078.6
3857.9
3241.3
2990.4
HC
98%
95%
95%
98%
95%
94%
98%
95%
94%
98%
95%
94%
98%
95%
94%
98%
95%
94%
89%
86%
85%
80%
72%
69%
80%
72%
72%
69%
80%
72%
68%
CO
85%
76%
76%
85%
76%
76%
85%
76%
76%
85%
76%
76%
85%
76%
76%
85%
76%
76%
79%
71%
70%
76%
64%
61%
76%
64%
64%
61%
76%
64%
59%
Number of
Simulated
Failures
195
142
135
195
142
131
195
142
131
195
142
127
195
142
127
195
142
127
168
116
101
142
64
45
142
64
64
45
142
64
41

Failure
Rate
17%
12%
12%
17%
12%
11%
17%
12%
11%
17%
22%-
11%
17%
12%
11%
17%
12%
11%
15%
10%
9%
12%
6%
4%
12%
6%
6%
4%
12%
6%
4%
Number of
Simulated
EC'S
11
11
8
11
11
8
11
11
8
11
11
4
11
11
4
11
11
4
11
11
4
11
11
4
11
11
11
4
11
11
4

EC
Rate
1.0%
1.0%
0.6%
1.0%
1.0%
0.6%
1.0%
1.0%
0.6%
1.0%
1.0%
0.3%
1.0%
1.0%
0.3%
1.0%
1.0%
0.3%
1.0%
1.0%
0.3%
1.0%
1.0%
0.3%
1.0%
1.0%
1.0%
0.3%
1.0%
1.0%
0.3%
Failure Rate
for FTP
Passing
Vehicles
2.0%
2.0%
1.3%
2.0%
2.0%
1.3%
2.0%
2.0%
1.3%
2.0%
um
0.7%
2.0%
2.0%
0.7%
2.0%
2.0%
0.7%
2.0%
2.0%
0.7%
2.0%
2.0%
0.7%
2.0%
2.0%
2.0%
0.7%
2.0%
2.0%
0.7%
H-26

-------
Distribution of Emissions Identified by FTP Emitter Categories
IM240 1983+ TBI Simulated Lane Sample: "Full IM240 Test Criteria"
Cut-Point
HC/CO
1.6/20

1.2/16

1.0/15

0.8/15

0.6/10

Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO
Normals
1%
1%
1%
1%
1%
2%
2%
2%
4%
5%
Highs
1%
1%
3%
2%
14%
7%
14%
9%
15%
10%
Very Highs
24%
30%
31%
36%
32%
39%
34%
40%
33%
40%
Supers
74%
68%
66%
60%
53%
51%
50%
49%
48%
46%
IM 240 1983-1985 TBI Simulated Lane Sample: "Full EV1240 Test Criteria"
Cut-Point
HC/CO
1.6/20

1.2/16

1.0/15

0.8/15

0.6/10


Emission
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
0%
0%
0%
0%
1%
1%
2%
1%
5%
5%

High*.
4%
2%
6%
4%
12%
6%
14%
8%
16%
9%

Verr Hlyhs
52%
47%
58%
53%
58%
56%
56%
55%
53%
52%

Supers
44%
50%
35%
43%
29%
37%
28%
36%
27%
34%
IM240 1986+ TBI Simulated Lane Sample: "Full IM240 Test Criteria"
Cut-Point
HC/CO
1.6/20

1.2/16

1.0/15

0.8/15

0.6/10


EflUSSlOD
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
1%
2%
1%
2%
1%
2%
1%
2%
1%
4%

ottos
0%
0%
1%
1%
17%
10%
16%
10%
15%
9%

Vt-rv Hlchs
7%
15%
11%
20%
9%
18%
18%
24%
20%
30%

Supers
92%
83%
87%
77%
72%
69%
65%
64%
64%
58%
                          H-27

-------
1981+ Simulated Lane Sample: "Two Ways to Pass Criteria"
CARfi Vehicles
Excess Excess
Lane IM240
Cut-Points
Comp+Bait 2
0.6/10 + 0.5/12
0.5/15+ 0.5/12
0.6/20 + 0.5/12
0.7/10 + 0.5/12
0.7/15 + 0.5/12
0.7/20 + 0.5/12
0.75/10 + 0.5/12
0.75/15 + 0.5/12
0.75/20 + 0.5/12
0.8/10 + 0.5/12
0.8/15 -MJ.Syt2
0.8/20 + 0.5/12
0.85/10+0.5/12
0.85/15+0.5/12
0.85/20 + 0.5/12
0.9/10 + 0.5/12
0.9/15+0.5/12
0.9/20 + 0.5/12
1.0/10 + 0.5/12
1.0/15+0.5/12
1.0/20 + 0.5/12
1.1/10 + 0.5/12
1.1/15+0.5/12
1.1/20 + 0.5/12
1.2/10 + 0.5/12
1.2/15+0.5/12
1.2/16 + 0.5/12
1.2/20 + 0.5/12
1.6/10 + 0.5/12
1.6/15+0.5/12
1.6/20 + 0.5/12
Emissions
Identified
H£
3122.6
3090.9
3090.9
3034.3
2964.3
2953.0
2902.9
2832.9
2810.3
2850.7
2733:2
2730.6
2832.4
2734.9
2712.3
2832.4
2734.9
2687.8
2832.4
2734.9
2666.5
2832.4
2734.9
2666.5
2832.4
2734.9
2717.1
2666.5
2803.3
2705.8
2604.5
Emissions

£Q
45640.9
44022.2
44022.2
45082.8
42387.3
42129.4
44554.8
41859.4
41373.6
44286.3
41220,2
40734.4
43998.2
40932.1
40446.3
43998.2
40932.1
40026.2
43998.2
40932.1
39681.5
43998.2
40932.1
39681.5
43998.2
40932.1
40626.5
39681.5
43825.2
40759.1
39280.9
Identified
H£
94%
93%
93%
92%
89%
89%
88%
86%
85%
86%
*m
82%
85%
83%
82%
85%
83%
81%
85%
83%
80%
85%
83%
80%
85%
83%
82%
80%
85%
82%
79%

£Q
93%
90%
90%
92%
87%
86%
91%
86%
85%
91%
84%
83%
90%
84%
83%
90%
84%
82%
90%
84%
81%
90%
84%
81%
90%
84%
83%
81%
90%
83%
80%
Number of
Simulated
Failures
1354
1219
1219
1219
1004
992
1165
950
926
1112
842
819
1085
815
792
1085
815
768
1085
815
730
1085
815
730
1085
815
788
730
1061
792
683

Failure
Rate
60%
54%
54%
54%
45%
44%
52%
42%
41%
50%
38%
37%
48%
36%
35%
48%
36%
34%
48%
36%
33%
48%
36%
33%
48%
36%
35%
33%
47%
35%
30%
Number of
Simulated
EC'S
0
0
0
0
0
0
0
0
0
0
a
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

EC
Rate
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehicles
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
H-28

-------
1981+ Simulated Lane Sample
Carbureted Vehicles
Excess Excess

Idle/2500
Cut-Points
220/1.2
100/1.0+220/1.2
100/0.5
Emissions
Identified
HC
2246.1
2626.8
2741.1
Emissions
Identified
CQ HC CO
31472.1 68% 64%
37064.9 79% 76%
39716.4 83% 81%
Number of
Simulated
Failures
599
874
1095

Failure
Rate
27%
39%
49%
Number of
Simulated
EC'S
0
0
54

Cc
Rate
0%
0%
2%
Failure Rate
for FTP
Passing
Vehicles
0%
0%
29%
H-29

-------
Characterize Hammond Indiana 1981+ Test Fleet as of 9/25
Number of




Tec hnvtoiY LJUK JM2W
Carbureted Low
High
Total
Vehicles
Actually
Tested at the
Hammond
Lane
1427
812
2239

Number of
Vehicles
in the
Lab Sample
53
69
122

Weighting
Factors
for the
lifBb Sample
26.92
11.77



Number of
Lab Vehicles
Passim FT?
7
0
7

Number of
Normal Lab
Vehicles
Falling FTP
25
4
29
H-30

-------
Characterize Simulated In-Use 81+ CARB Fleet
Calculated From the Lab sample and Weighting Factors





Technology
Carbureted
Number of
Vehicles
Actually
Tested at the
Hammond
Lau
2239

Number of
Simulated
Vehicles
Passing
FTPs,
188



Total Excess Emissions
In Simulated Sample
H£ £Q
3312.86 48915.00
Distrubutlon of Total Excess Emission by Emitter Group
1981+ CARB Sample
Simulated Lane Sample


Emissions
HC(g/mi)
C0(g/mi)
Total
Excess from
Normals
3%
2%
Total
Excess from
Jlighj
7%
4%
Total
Excess from
V
-------
1981+ Simulated Lane Sample: "Full IM240 Test Criteria"
Carbureted Vehicles
Excess Excess
Emissions Emissions Number of
Lane IM240
Cut-Points
0.6/10
0.6/15
0.6/20
0.7/10
0.7/15
0.7/20
0.75/10
0.75/15
0.75/20
0.8/10
0.8/15
0.8/20
0.85/10
0.85/15
0.85/20
0.9/10
0.9/15
0.9/20
1.0/10
1.0/15
1.0/20
1.1/10
1.1/15
1.1/20
1.2/10
1.2/15
1.2/16
1.2/20
1.6/10
1.6/15
1.6/20
Identified
HC
3131.7
3099.9
3099.9
3043.3
2973.1
2960.1
2912.0
2841.7
2817.4
2859.7
2m»
2737.7
2841.4
2743.7
2719.4
2838.2
2740.4
2691.7
2838.2
2740.4
2670.4
2838.2
2740.4
2670.4
2838.2
2740.4
2721.0
2670.4
2809.1
2707.5
2604.5
Identified
CO
45839.4
44220.7
44220.7
45281.3
42541.2
42201.6
44753.3
42013.2
41445.8
44484.8
41374.3
40806.6
44196.7
41085.9
40518.5
44196.7
41085.9
40098.4
44196.7
41085.9
39753.8
44196.7
41085.9
39753.8
44196.7
41085.9
40698.7
39753.8
44023.7
40840.6
39280.9
HC
95%
94%
94%
92%
90%
89%
88%
86%
85%
86%
83%
83%
86%
83%
82%
86%
83%
81%
86%
83%
81%
86%
83%
81%
86%
83%
82%
81%
85%
82%
79%
CO
94%
90%
90%
93%
87%
86%
91%
86%
85%
91%
85%
83%
90%
84%
83%
90%
84%
82%
90%
84%
81%
90%
84%
81%
90%
84%
83%
81%
90%
"83%
80%
Simulated
Failures
1485
1350
1350
1350
1108
1085
1297
1054
1019
1243
947
911
1216
920
884
1162
866
807
1135
839
741
1135
839
741
1135
839
800
741
1112
804
683
Failure
Rate
66%
60%
60%
60%
49%
48%
58%
47%
46%
56%
42%
41%
54%
41%
39%
52%
39%
36%
51%
37%
33%
51%
37%
33%
51%
37%
36%
33%
50%
36%
30%
Number of
Simulated
EC'S
54
54
54
54
54
54
54
54
54
54
54
54
54
54
54
27
27
27
0
0
0
0
0
0
0
0
0
0
0
0
0

EC
Rate
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
2.4%
1.2%
1.2%
1.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Failure Rate
for FTP
Passing
Vehicles
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
28.6%
14.3%
14.3%
14.3%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
H-32

-------
Dtstrnbution of Excess Emissions Identified by Emitter Group
IM240
Cut-Point
HC/CO
1.6/20

1.2/16

1.0/15

0.8/15

0.6/10

1981+ C ARB Simulated Lane Sample: "Full IM240 Test Criteria"

Emission.
HC
CO
HC
CO
HC
CO
HC
CO
HC
CO

Normals
0%
0%
0%
1%
0%
1%
1%
1%
1%
1%

Hlyhs
1%
1%
1%
1%
1%
1%
1%
1%
6%
4%

Very Hlyhs
59%
68%
61%
69%
61%
69%
61%
69%
60%
69%

Supers
39%
30%
38%
29%
37%
29%
37%
29%
33%
26%
                            H-33

-------
                  APPENDIX I
EVAPORATIVE SYSTEM PURGE AND PRESSURE DIAGRAMS

-------
                      P re a aura
      The evaporative pressure  test  is  used to determine the integrity
of a vehicle's  evaporative system, and fuel tank.   It is conducted by
introducing nitrogen  pressure into the fuel tank through the tank-to-
canister vapor vent line near the canister.   The  Nitrogen is introduced
into the system until the  pressure in the fuel tank stabilizes at about
14  inches  of water (0.5  PSI) .   Fuel  tank pressurization is  done by
continually  modulating the  Nitrogen  flow into the  fuel system by
successive opening and  closing  of the  control valve  by the operator.
Modulating the  Nitrogen flow into the  system  allows  a higher pressure
Nitrogen  flow to  be  safely  used to pressurize  the  system.   Without
modulating the  flow,  the  Nitrogen pressure would have to be low,  thus
considerably lengthing the test.  If too high a pressure  is  used,  a  weak
vapor hose might bulge and rupture.

      After  the vehicle's evaporative  system is  pressurized,  it is
allowed to stand  for  up to two minutes  to determine if it can continue
to hold pressure.  A vehicle is recorded as  a failure  if  the fuel  system
pressure drops  to  less  than 8 inches of water  within the 2 minute  time
frame.
                       Pressure  Teat   Equipment

      Figure  2  shows a schematic  of the  Pressure test set-up which is
 used.   The required equipment includes an air or nitrogen gas bottle, a
 standard  regulator,  and  a magnehelic  to provide  finer  control while
 pressurizing  a  vehicle's  evaporative system. Other pieces of equipment
 include clamps  to close off vapor  lines and other assorted fittings.
                               PRESSURE TEST
                   FILLER NECK
                         NITROGEN CVUNO EH

-------
                        Purere  Teat.  Pror-oHn-ro
      The evaporative purge test is  conducted during  an  IM240  transient
dynamometer  test to detect  vehicles whose  evaporative  canister  purge
system are  inoperative.  The test  procedure includes disconnecting  the
test vehicle's vapor purge line running from the canister to the engine,
and installing a gas flowmeter in  the line.  On most  vehicles  this is a
relatively  simple  and  quick procedure.    However,  in  some cases  the
canister and its purge lines are difficult to find in  the allotted time,
and as a  result, the vehicles cannot be tested.

      After  installing the  flow  meter in the evaporative purge system,
the  vehicle  is  operated  over  the IM240  transient  cycle,   and  the
cumulative  vapor purge  flow in units  of  liters  are recorded.     The
vehicle is  recorded as  a failure if its cumulative vapor purge is less
than 1.0 liter.
                                Teafc  Euiment
      Figure 1 shows a schematic of the Purge test set-up which is  used.
The  required equipment includes a  transient dynamometer  (not  shown  in
Figure  1)  on  which to conduct  the IM240,  and a  gas  flowmeter  which
measures the  instantaneous  and cumulative vapor purge flow between  the
evaporative   canister and  the  engine  during  the IM240  test.    The
dynamometer  which is used is  a  standard Clayton ECE-50 twin  roll with
125 pound inertia weights and  Road  Load  power  control.     The  flowmeter
which  is used is  a Sierra  Total flow meter  series  730  capable  of
measuring flows from 0 to 50 liters per minute.
                                 PURGE TEST
                   FILLER NECK
                          ROLLOVER VALVE
                                  EVAPORATIVE
                                   CANISTER
            FUEL TAN*
                                                 FLOW MCTEF

-------
               APPENDIX J
EVAPORATIVE SYSTEM FAILURES AND REPAIRS

-------
Running Loss Emission Levels of Cars Recruited for Repair
at 95 F and 9.0 RVP Fuel
(3 LA-4 Cycles)

Manufacturer


General Motors
FORD
Chrysler
OTHER

ALL
Average Running Losses
Purae Failures,
g/mi

8.46
6.51
5.16
3.94

7.06
Pressure Failures
g/ffij

8.29
7.03
4.20
3.88

7.99
Repair Reduction
a/ml "

5.66
6.26
5.12
3.38

5,67
Alter Repair
a/mi "

3.22
0.55
0.07
0.23

1.76
Notes
* A total of 47 cars were tested ttiat tailed eilhor the purge test, the pressure test, or both tests. The Purge and Pressure Failure Columns
are based on 47 cars.
" A total of 40 cars were repaired. Average Repair Reductions and Alter Repair running losses in Hie above Table are based on 40 cars.

Manufactuer

General Motors
FORD
Chrysler
OTHER

ALL
Total
Failures

22
16
5
4

47
Number Of Failures hv Tvpe

Purge
Failures

11
12
4
2

29
Solenoid /
Electrical

6
6
1
1

14
Purge
LJLBLS

4
6
3
1

13
Unknown

1
0
0
0

1
Pressure
Failures

1 1
4
1
2

18

Vent
Line

1
3
0
0

5
Gas Can

9
0
0
1

10
Tank/
Sending
Unit

1
1
1
1

4
As  of 11/12/91

-------


VEH

1455
1455

1456
1456

1459
1459

1532
1532
1532

1580
1580

1712
1712

1461
1461

1530
1530

1578
1578

1457
1457

1524
1524

1533

1542
1542

1544
1544

1584
1584

1672
1672



MYR

88
88

90
90

90
90

86
86
86

90
90

87
87

87
87

84
84

86
86

85
85

85
85

88

85
85

86
86

86
86

86
86



MAKE

CHEVY
CHEVY

CHEVY
CHEVY

CHEVY
CHEVY

CHEVY
CHEVY
CHEVY

CHEVY
CHEVY

CHEVY
CHEVY

PONTIAC
PONT1AC

PONTIAC
PONTIAC

PONTIAC
PONTIAC

BUICK
BUICK

BUICK
BUICK

BUICK

BUICK
BUICK

BUICK
BUICK

BUICK
BUICK

QUICK
BUICK



MODEL

CORS
CORS

LUMI
LUMI

.UMI
.UMI

228
Z28
Z28

CAVA
CAVA

CAVA
CAVA
-
GRNAM
GRNAM

6000
6000

SUNB
SUNB

CENT
CENT

LESA
LESA

REGAL

>ARK
'ARK

CENT
CENT

CENT
CENT

LESA
LESA
'


ENG.FAM

2.0LTBI
2.0L TBI

L1G3.1V8XGZ5
L1G31V8XGZ5

3.1LPFI
3.1LPFI

G1G50V8NTAB
G1G50V8NTA8
G1G50V8NTA8

L1G22V5JFG2
L1G2.2V5JFG2

2 OL TBI
2.0LTBI

2 5L TBI
2.5L TBI

2 51 TBI
2 51 TBI

G2G18V5TDG2
G2G1.8V5TDG2

2.5L TBI
2.5LTBI

5.0L
5.0L

2.8L PR

F4G3.8VBXEB3
F4G3.8V8XEB3

G2G25V5TPG9
32G25V5TPG9

32G25V5TPG9
G2G2.5V5TPG9

S4G3.8V8XEB4
34G38V8XEB4



Purfl

PASS
PASS

FAIL
PASS

FAIL
PASS

FAIL
PASS
PASS

PASS
PASS

PASS
PASS

PASS
PASS

PASS
PASS

PASS
PASS

PASS
PASS

FAIL
PASS

FAIL

PASS
PASS

FAIL
PASS

FAIL
PASS

FAIL
PASS


Asoll
Pres

FAIL
PASS

PASS
PASS

PASS
PASS

PASS
FAIL
PASS

FAIL
PASS

FAIL
PASS

FAIL
PASS

FAIL
PASS

FAIL
PASS

FAIL
PASS

PASS
PASS

PASS

FAIL
PASS

PASS
PASS

PASS
PASS

PASS
PASS


1/12/9
TEST

RECV
RM1

RECV
RM1

RECV
RM1

RECV
RM1
RM2

RECV
RM1

RECV
RMI

RECV
RMI

RECV
RMI

RECV
RMI

RECV
RMI

RECV
RMI

RECV

RECV
RMI

RECV
RMI

RECV
RMI

RECV
RMI



TOTAL RL
a/ml
588
306

11.86
1.88

2700
13.85

14.02
14.70
14.22

563
0.34

848
7.78

1153
0.10

773
0.11

11.70
14.45
-
465
0.11

331
0.11

0.07

2.86
0.44

10.43
1.05

7.54
014

7.80
0.87



RL Reducl
a/ml

2.81


998


1315



-020


530


069


11.44


7.63


-275


4.54


320




2.42


9.38


7.40


6.93



Hot Soak

















045

































Olumal

















043

































COMMENTS

Gas Cap Leaks


Bad Purge Solenoid


Purge Solenoid Inoperative


Purge Line from Canister Disconnected
Gas Cap Leaking
Leak Possibly at Sending Unit Gasket

Gas Cap Leaks


Vent Line Leaks at connection with Steel Lin

,x,:,,,: ;...,:. «,.:.;>.:. :::.:,,;::.:;.5:i<:::.:;.;j:::-:,
Gas Cap Leaks, Filler Neck rusted at Cap


Gas Cap Does not Seal


Gas Cap Leaks


Crack in Seam ol Gas Tank and Filler Neck


TVS Broken


Unknown Problems

Gas Cap Does not Seal


Vac Line from Manifold to Purg Valve Broker


Purge Una Weathered. Collaspes when hot


Purge Solenoid Wiring Disconnected



-------


VEH

158!
MM




1661
1661

1720

1526

1552
1552

1560
1560

MYH

87
87

85
85

88
-
85

84
84

86
86

MAKE

OLDS
OLDS



OLOS
OLDS

OLDS

CADILLAC


CADILLAC
CADILLAC

CADIL
CADIL
.AC
.AC

MODEL

[)E
DE

-T
T

CUTL
CUTL

CUTL
, ,, ,
SEVI

SEVI
SEVI

ELDO
ELDO

ENG.FAM

H2G3.8V8XEB7
H2G3.8V8XEB7

F2G25V5TPG8
F2G25V5TPG8

2 8L PFI

4.1LTBI

E6G4.1V5NKA7
E6G4.1V5NKA7

G6G4.
G6G4.
V5NKA9
V5NKA9

Purn

FAIL
PASS

PASS
PASS

PASS

FAIL

FAIL
PASS

FAIL
PASS
Asol 11 112/9
Pm»

PASS
PASS

FAIL
PASS

FAIL

PASS

PASS
PASS

PASS
PASS
TEST

RECV
RM1

RECV
RM1

RECV
'
RECV

RECV
RM1

RECV
RM1

TOTAL RL
a/ml
553
2.14

880
022

9.20

5.48

711
020

6.88
0.16

RL Reducl
fl/ml

339


858


-



6.91


671

Hot Soak







076









Diurnal







1.76









COMMENTS

Purae Line Disconnected


Gas Cap Leaks


Gas Cap has a holed Drilled in it

Canister Purge Solenoid Inoperative

Canister Purge Solenoid Bad


Purge Vapor Line on top of Can. Disconnect!


-------

Vah

1460
1460

1462
1462

1525
1525

1563
1563

1574
1574

1575
1575

1647
1647

1676
1676

1450
1450

1537
1537

1561
1561

1689
1689

1802
1802
1802

* s
fex f.vVtff*
1548
1548

1639

1713
1713
1713


MYR

85
85

88
88

89
89

85
85

87
87

89
89

86
86

84
84
'
88
88

86
86

85
85

85
as

83
83
83

-.•>, ,••'
88
88

81

86
86
86


MAKE

FORD
FORD

FORD
FORD

FORD
FORD

FORD
FORD

FORD
FORD

FORD
FORD

FORD
FORD

FORD
FORD
*
MERCURY
MERCURY

MERCURY
MERCURY

MERCURY
MERCURY

MERCURY
MERCURY

MERCURY
MERCURY
MERCURY

f S -f
LINCOLN
LINCOLN

LINCOLN

LINCOLN
LINCOLN
LINCOLN


MODEL

CROWN
CROWN

MUST
MUST

MUST
MUST

TEMP
TEMP

ESCO
ESCO

TAUR
TAUR

MUST
MUST

CROWN
CROWN
'
TRAC
TRAC

COUG
COUG

TOPA
TOPA

MARQ
MARO

MARQ
MARQ
MARQ

ff x-:--.
MARK
MARK

TOWN

TOWN
TOWN
TOWN


ENG.FAM

5.0L TBI
5 OL TBI

2.3L PFI
2 3L PFI

5.0L PFI
5.0LPFI

FFM2.3V5HCF4
FFM2.3V5HCF4

HFM1.9V5FFF1
HFMI.9V5FFFI

KFM2.5V5HCF1
KFM2.5V5HCF1

GFM50V5HBF9
GFM5.0VSHBF9

EFM50V5HBF7
EFM50V5HBF7
* ,,> *
JFM16V5FZKO
JFM1.6V5FZKO

3.BL TBI
3.8LTBI

FFM2.3V5HCF4
FFM2.3V5HCF4

SOL TBI
50 L TBI

5.0L TBI
5 OL TBI
50LTBI

* ,„,.>, ,«
JFM5.0V5HBF3
JFM50V5HBF3

5.0CCF

50LPFI
5 OL PFI
5.0L PFI


Purg

FAIL
PASS

PASS
PASS

FAIL
PASS

FAIL
PASS

PASS
PASS

FAIL
PASS

PASS
PASS

FAIL
PASS

FAIL
PASS

PASS
PASS

FAIL
PASS

FAIL
PASS

FAIL
FAIL
PASS


FAIL
PASS

FAIL

FAIL
PASS
PASS

AsoM
Pres

PASS
PASS

FAIL
PASS

PASS
PASS

PASS
PASS

FAIL
PASS

PASS
PASS

FAIL
PASS

PASS
PASS
-
PASS
PASS

FAIL
PASS

PASS
PASS

PASS
PASS

FAIL
FAIL
PASS


PASS
PASS

PASS

PASS
PASS
PASS

1/12/9
TEST

RECV
RM1

RECV
RM1

RECV
RM1

RECV
RMt

RECV
RM1

RECV
RM1

RECV
RM1

RECV
RM1

RECV
RM1

RECV
RM1

rRECV
RM1

RECV
RM1

RECV
RM1
RM2


RECV
RM1

RECV

RECV
RM1
REP


TOTAL RL
Q/ml
715
025

917
239

11.63
082

526
024

3.58
Oil

1041
008

860
036

4.43
181

1.22
017

676
005

3.88
0.10

349
025

628
634
0.12

-
12.26
1.40

409

806
009
012

| :
RL Reduct
fl/ml

690


678


10.81


503


3.48


1033


8.24


262


106


6.71


3.78


324



6 16



1086





7.94


Hot Soak


































068

















Diurnal
























'









0.74

















COMMENTS

Electronic Puroe Solenoid Inoperative


Plastic Vent Line Irom Canister Missing


Connector at Can. does not make conneclioi


Purge Solenoid Inoperative


Sending Unit Gasket does not seal


Purae Vac Line burned through


Fuel Tank Vent line disconnected at Caniste


Purge Vac Line Misrouted


Broken Purge Line Fittings


Vapor Line Disconnected at Rollover Valve


Purge Solenoid Inoperative


TVS is stuck in the Closed Position


Recv - Tank Vent line plugged. Battery Acid
destroyed Canister.
RM1 - Repaired Purge Lines
RM2 - New Normalized Canister

Electrical Connection at Purge Solenoid Bad


Electrical Grounding Problem near ECM

Purge Hose Disconnected




-------

VEHicLe
1530
is?? 	
1S37
1S41
1542
1574
1578
ISIO
1647
I6«l
1667


Average






VEHICLE
1524
1S2S
1544
1548
1552
1655
1580
1561
1563
1575
15S4
15(5
1567
1640


Average
f^HESSOBE FAILURES 	 1
DIAGNOSIS oH REPAIRS 	
«EW OEM CMS CAP
IEPUCEGASCAP
^CONNECTED FUEL TANK VAPOR LNE TO ROLLOVER VALVE
1EPUCED OVEFFLOW HOSE WHICH WAS SPLIT
REPLACED FILLER NECK AND GAS CAP
5ASKET AROUND SENDNG UNIT DOES NOT SEAL PROPERLY
GAS CAP LEAKS
GAS CAP LEAKS
RECONNECT FUEL TANK VENT LINE TO CANISTER
GAS CAP LEAKS
GAS CAP LEAKS







PURGE PAiLURES

DIAGNOSIS OR REPAIRS
REPLACE THERMAL-VACUUM SWITCH
REPLACE MALE END OF CONNECTOR TO CANISTER PURGE SOEI
VAC UNE RECONNECTED FROM M ANIFOUD TO PURGE VALVE
PURGE SOUBNOD REPLACED
REPLACED FAULTY CANISTER PURGE SOLENOD
TEE FITTNG REPLACED
PURGE LWE RECONNECTED
PURSE SOLENOD REPLACED
PURSE SOLENOD REPLACED
PURGE VACUUM LME TO THE ENGINE HAS A HOLE BURNED THF
PURGE LME IS WEATHERED AND COLLAPSES WHEN HOT
PURGE LNE IS DISCONNECTED
PURGE LNE B BROKEN NEAR ENGINE
THERMAL VACUUM SWITCH REPLACED




g/t.l
171 60
-4 40
151 00
5650
5440
7640
•61 90
119 10
165 40
19300
115 30


86.04





HCRcd
g/lsl
71 90
24320
211 00
244 30
15550
123 ID
15099
64 9C
113 1C
23246
16660
7630
13000
56 60


147.14
THREE BAGS
g/rnM.
s«e
004
4S4
299
1 54
266
296
457
429
656
502


3.75



HNST
1HREEIAO3
HCRod
glifUt
1 37
735
721
643
542
361
563
320
416
851
403
402
4 14
1 40


4.75

elmllo
17. 35
6.23
1.63
•1.33
•24 18
26 II
911
1566
-4.73
4.45



5. IB



Fu«l
Consumption
HC hiel Rod
glmllo
-3 52
-13(7
1 51
766
•2 35
469
14 76
oot
•5 63
18 10
-9 (0
41 00
-0.63
41 02


6.65
	 T«l«l
FiMfeRL MC
(/mile
2323
8.27
6.17
166
-2334
2699
12.09
20.43
•0.44
11.01



•.81



Tolol
FuoURLHC
Sevlnge
fiRilll
•215
••.32
• 72
1411
307
630
2041
326
•1 67
2661
-5.77
4502
351
42.42


11.40
Part*
BASIS
Etl
Acl
Ell
Ell
Act
Ell
E»l
E>l
EM
E»l
Ell







Paris
Cost
BASIS
Ell
Ell
Ell
Act
Ell
Ell
Acl
Acl
Acl
Ell
Ell
EM
Ell
Acl




Porto
$700
$362
(000
(000
(3340
(9311
(700
(700
(000
(700
$720


$15.03





Cost
Puts
(2041
(2946
(000
(3853
$2946
(100
(000
(3704
(13015
(000
(000
(000
(000
(2041


$21.89
Lobor
BASIS
EH
Act
Ell
Ell
Act
Ell
Ell
Ell
Ell
Ell
EH







Lobor
Tim*
BASIS
Ell
Ell
Ell
Acl
Ell
Acl
Acl
Acl
Acl
EH
Ell
Ell
EM
Est




TIME
0.25
025
0.5C
OSC
IOC
1 0(
025
0.25
OS(
025
0.25


045





Labor
TIME
050
OSC
OSC
600
0.5C
OSC
OSC
OSC
1.50
05C
050
050
050
0.50


0.96
•*)> SSO/MR
1 	 r-^t
Uboi
(12 SO
(1250
(2500
(2500
(5000
(5000
(1250
(1250
(25.00
(1250
(12 SO


$22.73




'« HO/HR
Cost
Labor
(2500
(2500
(2500
(30000
(2500
(2500
(2500
(2500
(7500
(2500
(25.00
(2500
(2500
(25.00


$48.21
•@SSOHH
1 fco»t
	 fotli
$1950
(1612
(25.00
(2500
(8340
(14311
(1950
(1950
(2500
(1950
$1970


$37.76




•®$50/HR
Cosl
Total
(45.41
SS446
(2500
(338 53
(5446
(26.00
(2500
(6204
(20515
(2500
(25.00
(2500
(25.00
$45.41


$70.10
PARTS
•/gfmlle
(030
$044
(000
(000
((143
(321
(056
$034
$0.00
$0.64
•...


$1.71




PARTS
Cool / Rod
f/g/mllt
((948]
((466]
(000
(273
$961
(0.12
$000
$1129
($7784)
$0.00
(000
$000
$000
$048


$1.92
•(§> *5CyHH
LABOfl
(fg/mlle
(0.54
(151
(4.05
(1504
((2.14
$173
(1.03
$0.61
($57.34
$1.14
•-..


$2.56



'@(50/HF
LABOR
Coal / Red
S/g/milo
((11 62
((396
(2.87
(2126
(815
$301
$122
$762
($44 (61
$094
($433)
JO 56
(7.12
$059


$4.23
FUEL
eOCNGMV
gal/mllo
00091
0.0034
0.002!
o.ooo;
•00091
0.0121
0.0050
0.0084
-0.0002
0.0045
•...


0.0036



FUEI
ECONOMY
SAVINGS
gal/milt
-0.0009
•0.0026
0.0031
0.0056
00013
0.0034
0.0084
0 0014
•0.0007
0.0110
-0.0024
0.0186
0.0015
0.0175


0.0047
•^SSO/Mfi
TOTM
1 coal / h«
syg/mii*
(064
$195
(405
(1504
($3.57
(494
$161
$095
($5734
$1.77



$4.29



•e$so*»
TOTAL
Coil /Red
$/g/mlle
((21 10
($862
$287
$2399
(1776
$3.13
$1.22
$1691
($12270:
$094
($433)
$056
$712
$1 07


$6.15

-------


VEH

555
555
.- -
1458
1458

1587
1587

1635

1559

'


VEH

1541
1541

1568

1640
1640
•."/- ^
1667
1667


MYR

88
88

87
87

86
86

84

89

-. * s "


MYR

83
83

86

84
84
...:


MAKE

DHRYSLEF
CHRYSLEF
-
DODGE
DODGE

DODGE
DODGE


DODGE

PLYMOU1>

4 ': ': \ ',' ''


MAKE

RENAULT
RENAULT


MAZDA

TOYOTA
TOYOTA
%.•'• f ff<'-
eRENAU
IRENAU
• •
.


MODEL

L
L
•BA
EBA
1 »
LANC
LANC

DAYT
DAYT



DAYT


^
ACCL

t , ,,?


MODEL

ALL!
ALL!
-
HX-7
-
CELI
CELI
,,/z
AL
AL
.1
.1


ENG.FAM

JCR25V5FBE6
JCR2.5V5FBE6
-,,, " ,
22LTBI
22LTBI

GCR2.2V5FAAX
GCR2.2V5FAAX

ECR2.2V5FAA8
, , -',-. «
KCR2.5V5FBD6

\^4?l;


ENG.FAM

DAM1.4V5FFD3
DAM1.4V5FFD3
-."• fsff, vA J •-
GTK1.3V5HFD8
"
ETY28V5FBB5
ETY2.8V5FBB5
','... '...-.,.
FAM1.4V5FFA2
FAM1.4V5FFA2


Purg

FAIL
PASS

PASS
PASS

FAIL
PASS

FAIL
-
FAIL

-


Purg

PASS
PASS

FAIL

FAIL
PASS
\ f •.
PASS
PASS



As ofll/12/91
Pres

PASS
PASS

FAIL
PASS

PASS
PASS

PASS

PASS

-

TEST

RECV
RM1
'
RECV
RM1

RECV
RM1

RECV
*
RECV

^

TOTAL RL
a/ml
552
0.05
,- -
4.20
008

586
0.08

5.43

381

V*
': s --

As of 11/12/91
Pres

FAIL
PASS
-
PASS

PASS
PASS
-
FAIL
PASS
TEST

RECV
RM1
»$o
RECV

RECV
RM1

RECV
RM1
TOTAL RL
g/ml
2.57
0.06

4.79

3.08
0.57
- ,J
5.19
0.07


RL Reduct
a/ml

5.47


412


5.78





" ,


RL Reducl
g/ml

2.51




2.51


5.12


Hoi Soak














<


Hoi Soak













Diurnal














-'\ ( .
';


Diurnal



:




1




COMMENTS

Tee fining in purge line broken at Throttle Bo


Vehicle has a Broken Rollover Valve


Purge Line Broken near Engine


Canister Purge Line is Burned in Hall

Purge Valve Stuck

%^


COMMENTS

Fuel Sending Unit Gasket

-
Purges thru Oil Filler. Oil cap does not seal

TVS is Broken

-
Gas Cap Leaks


-------
                     APPENDIX K
MOBILE4.1  PERFORMANCE STANDARD ANALYSES, BY OPTION

-------
IRun »8p Mod. High Option,  Performance standard Options (annual|
 MOBILE4.K4NOV91)
0
-M120 Warningi
4.              MOBILE*.1 does not model moat 1993 and later clean Ale Act
               requirements* Emission Factors Cor CY 1993 or later are affected.
OI/M program selected:
     Start year (January 1):
     Pre-1981 MYR stringency ratet
     First model year covered:
     Last model year covered!
     Waiver rate (pre-19811s
     waiver rate (1981 and newer):
     Compliance Ratet
     Inspection type!
     Inspection frequency
     Vehicle types covered!
     1981 t later MYR test type!
 1983
  20%
 1968
 2020
  1.%
  1.%
 98.%
 Centralized
 Annual
 LDOV - Yes
U5GT1 - Yes
LDCT2 - Yes
 HDGV - No
 2SOO rpra / I
     MOBIU4.1 IM240 Transient Annual  1/M Credits  .8/15  (8/5/91)
Oldest model year covered by:
IH240 Transient Test 1986
Purge System Check 1986
Fuel System Pressure Check 1983
OAnti-tamperlng program selected:
0 Start year (January 11 : 1983
First model year covered: 1984
Last model year covered: 2020
Vohlcle types covered: LDGV , LDGT1, LDCT2
" Type: Centralized
Frequency: Annual
Compliance Rate: 98.0%
0 Air pump system disablements: No
Catalyst removals : Yes
Fuel inlet restrlctor disablements: Yes
Tailpipe lead deposit test: No
ECU disablement: No
Evaporative system disablements: No
PCV system disablements: No
Missing gas caps: No
0 Minimum Temp: 72. (F) Maximum Temp: 92.
Period 1 RVPi 11.5 Period 2 RVPt 8.7 Period 2 Start Yri 1992
OVOC HC emission factors include evaporative HC emission factors.
0
OCal. Year: 2000 I/M
Anti-tarn.
0 Veh. Type: LDGV
+
Veh. Speeds: 137?
VMT Mix:
OComposlte Emission
VOC HC:
Exhaust HC:
Eveporat HC:
Refuel L HC:
Runlng L I1C:
Rstlng L HC:
Exhaust CO:
Exhaust NOXI
0.584
rectors
1.16
0.48
0.16
0.19
0.23
0.11
6.17
0.72
Program: Yes Anblent Temp: 87.5 / 87.5 / 87.5 (F) Region: Low
Program: Yea operating Mode: 20.6 / 27.3 / 20.6 Altitude: 500
LDST1 LDGT2 LDGT HDGV LDDV LOOT HDOV
13. S
0.199
(Cm/Nile)
1.50
0.81
0.17
0.25
0.19
0.10
7.39
1.03
19.S
0.080

1.69
0.90
0.24
0.25
0.20
0.10
8.58
1.13



1.56
0.83
0.19
0.23
0.19
0.10
7.73
1.07
19. t
0.035

4.54
2.07
1.44
0.40
0.51
0.12
36.27
4.51
13. i
0.002

0.6S
0.65




1.60
1.37
19.6
0.001

0.84
0.84




1.74
1.50
19.6
0.093

2.15
2.15




11.21
9.30
(F)
. Ft.
MC
15. S
0.007

4.96
1.89
2.52


0.54
24.78
0.77

All Veh



1.503
0.796
0.213
0.192
0.204
0.098
9.230
1.753

-------
Mobile  output  from CEM 4.1    1/24/92
Run »2 : Basic I/M
MOBILE4.1(4Nov91)
Ml20 Warning:
                           MOBILE4.1 does not modal moat 1993 and later Clean Air Act
              requirements; Emission Factors for CY 1993 or later are affected.
I/M program selected:

    Start year (January 1):
    Pre-1981 MYR stringency rate:
    First imdel year covered:
    Last model year covered:
    Waiver rate (pre-1981):
    Waiver rate (1981 and newer):
    Compliance Rate:
    Inspection type:
    Inspection frequency
    Vehicle types covered:
    1961 t later MYR test type:

Oldest model year covered by:
    IM240 Transient Test
    Purge System Check
    Fuel System Pressure Check
 1983
  20%
 1968
 2020
  0.%
  0.%
100.%
 Centralized
 Annual
 LDGV - Yes
LDGT1 - No
LDGT2 - No
 HDGV - No
 Idle
 2020
 2020
 2020
                      Period 1 RVP: 11.5
                                              Minimum Temp:
                                              Period 2 RVP:
                      72.
                      8.7
                                                                 (F)
     Maximum Temp:   92.  (F)
Period 2 Start Yr:  1992
VOC HC emission factors Include evaporative MC emission factors.
Cal. Year: 2000
  '/eh. Type:
                       I/M Program: Yes
                 Anti—tarn. Program: No
       Ambient Temp:  87.5  /  97.5  /  87.5  (F) Region: Low
     Operating Mode:  20.6  /  27.3  /  20.6   Altitude:  SOO. Ft.
                           LDGT1
                                     LDST2
                                                LOST
                                                                    L0DV
                                                                              LDDT
                                                                                                        All Veh
Veh. Speeds:
VMT Mix:
19.?
O.S84
Composite Emission Factors
VOC HC:
Exhaust HC:
Evaporat HC:
Refuel L HC:
Runing L HC:
Rating L HC:
Exhaust CO:
Exhaust NOX:
Cal. Year: 2000

Veh. Type:
Veh. Speeds:
VMT Mix:
1.S8
0.52
0.36
0.19
0.42
0.11
e.se
0.73
I/M
Anti— tarn.
LDGV
19.6
0.584
Composite Emission Factors
VOC HC:
Exhaust HC:
Evaporat HC:
Refuel L HC:
Runing L HC:
Rstlng L HC:
Exhaust CO:
Exhaust NOX:
1.77
0.71
0.36
0.19
0.42
0.11
10.05
0.75
19.? 	
0.199
(Cm/Mile)
2.25
1.20
0.35
0.25
0.36
0.10
11.96
1.1S
Program:
Program:
IDGT1
19.?
0.199
(Gm/Hlle)
2.25
1.20
0.35
0.25
0.36
0.10
11.96
1.15
" 19.?
0.080

2.59
1.37
0.46
0.25
0.40
0.10
14.40
1.29
No
No
LDGT2
" 19.?
0.080

2.59
1.37
0.46
0.25
0.40
0.10
14.40
1.25



2.35
1.25
0.38
0.25
0.37
0.10
12.66
1.18
Ambient
Operating
LDGT



2.35
1.25
0.38
0.25
0.37
0.10
12.66
1.18
19
0

4
2
1
0
0
0
36
4
Temp:
Model
.?
.035

.54
.07
.44
.40
.51
.12
.27
.51
87.5 /
20.6 /
HDGV
19
0

4
2
1
0
0
0
36
4

.035

.54
.07
.44
.40
.51
.12
.27
.51
19.?
0.002

0.65
0.65




1.60
1.37
87.5 /
27.3 /
LDDV
19.?
0.002

0.65
0.65




1.60
1.37
19.?
0.001

0.34
0.84




1.74
1.50
87.5 (F)
19TS
0.093

2.15
2.15




11.21
9.30
Region: Low
20.6 Altitude: 500.
LDDT
19.?
0.001

0.84
0.84




1.74
1.50
HD0V
"""""" 19.6
0.093

2.15
2.15




11.21
9.30
19.6
0.007

4.96
1.89
2.52


0.54
24.78
0.77

Ft.
MC
19.6
0.007

4.96
1.89
2.52


0.54
24.78
0.77



1.971
0.937
0.381
0.192
0.364
0.098
10.021
1.783


All Veh



2.084
1.049
0.382
0.192
0.364
0.098
11.874
1.799

-------
CEM4.1 : Run *2  : Basic  I/M                                                                01/24/92  13:41:59
                                                   Local Parameters Selected;


Base RVP: 11.5
In-use RVP: 8.7
Vehicle Type: LDGV
Veh. Speeds: I9.fi

I/M start year (January 1) :
Tampering deterrence start:
Oldest model year covered:
I/M inspection frequency:
I/M program type:
1981 t later MXR test type:
Vehicle types covered:
I/M waiver rate (pre-1981) :
I/M waiver rate (post-1981) :
I/M compliance rate:
Cost Assumptions

Operating Mode: 20.6

In-use Start Yr: 1992
LDGT1 LDGT2
157? 157?

1553
1983
1968-2020 Fixed
Annual
Centralized
Idle
LDGV
0.0%
0.0%
100.0%
1 87.5 1 87.5 (Ft 	 Region; Low 	
/ 27.3 / 20.6 Altitude: 500. Ft.
Minimum Temp: 72. (F)
Maximum Temp: 92. (F)
HDGV LDDV LDDT HDDV MC
—&7S -T57S —157? — 157? ^575
I/M Program Selected:
stringency: 20



IM240 test model year coverage: 2020+
Purge check model year coverage: 2020+
Pressure check model year coverage: 2020+


Fleet Assumptions LDGV LDT1 LDT2 HDGV LDDV LDDT HDDV MC
   I/M inspection cost:  (per insp) $  8.00            Vehicle percentages:  63.2  1577    371    376    o75    072    3.3
                                                      Growth rate:   0.0 % per year
   Avg. Gasoline Cost (per gallon): $  1.25                 Fleet size is  1000000.  vehicles in base  year  1986

       Repair Costs            Pre-91    81+
  Avg.  I/M repair cost     i  5  50.     75.  ( IM240 repair « $ 150. NOx repair* 5   0.  )

        Benefits (1000 tons/yr) and Costs  (1000 S/yr) are averaged over the calendar years  2000 thru  2000.

 Calendar Year 2000              TotVOC     55      ExhHC    Evap    Refuel   RnLoss   RstLos

 Baseline No-Program Emissions:   21.955  123.217   10.929    4.024    2.067    3.897    1.038

                                  Program Benefits and Costs

VOC
Average Annual ATP Benefit, Fee 6.666
Evap/RunLoss Benefit, Fee
Evap/RunLoss MFC Benefit
Avg Annual I/M Benefit, Fee
Avg Annual ATP Repair Cost
Avg Ann. Prg/Prass Repair Cost
Avg Ann. IM Repair Cost,FailRt
Average Annual IM Fuel Savings
Exhaust . Tampering Deterrence
0.006

1.033




0.142
Evap RnLoss ExhHC Ann Avg 1.181
0.005 0.001 1.175 (
2936
CO
0.000


17.762




1.646
19.408
48237
Cost
0.
0.
( o.
5074.
0.
0.
3536.
( 2198.

£412.
kilograms



)




)


P
                                                                 0.0483

-------
CEM4.1 :  Run HO  High Option  (Biennial)
                                                                                            01/24/92   13:43:19
                                                   Local Parameters Selected:
                                    Ambient Temp:  87.5  /  87.5  /  87.5  (F)
                                  Operating Mode  20.6  I  27.3  /  20.6
                                                                                    Region:  Low
                                                                                 Altitude:   500.  Ft.
      Base RVP: 11.5
    In-use RVP;  9.7

  Vehicle Type:

   Veh. Speeds:
                                 In-use Start Yr: 1992

                            LDGV      LDGT1      LDGT2      HDOV       LDDV       LDDT

                            TSTJ      137?       157?      177?       T575       157?

                                                        I/M Program Selected:
                                                                             Minimum Temp:   72.
                                                                             Maximum Temp:   92.
                                                                       HDDV

                                                                       19.4
                                                                                (F)
                                                                                (F)
                                                                           MC

                                                                          19.«
   I/M stact year  (January 1):
   Tampering deterrence start:
   Oldest model year covered:
   I/M inspection  £requency:
   I/M program type:
   1991 t later MYR test type:
   Vehicle types covered:
   I/M waiver rate  (pre-1981):
   I/M waiver rate  (post-1991):
   I/M compliance  rate:
                                     1983
                                     1983
                                     1968-2020  Fixed
                                     Biennial
                                     Centralized
                                     2500  rpm / Idle
                                     LDGV,  LDCT1,  LDGT2
                                       1.0%
                                       1.0%
                                      98.0%
                                                Stringency:20
                                                        IM240 test model year coverage:
                                                        Purge check model year coverage:
                                                                                    1984+
                                                                                    1984+
                                                        Pressure check model year coverage: 1971+
                                                  Anti-Tamperlng Program Selected:
ATP inspection frequency:
Oldest model year covered*
Vehicle types covered:
Parameters covered by ATP:
Biennial
1975-2020 Fixed
LDGV , LDGT1, tDGT2
Air Punp / Catalyst
Evap System / PCV System
ATP compliance rate:
ATP program type:
/ Fuel Inlet / Plumbtesroo /
/ Gas Cap /
98.0%
Centralized
           Cost Assumptions
                                                     Fleet Assumptions
                                                                           LDGV  LDT1  LDT2  HDGV  LDDV  LDDT  HDDV   MC
  I/a Inspection cost:  (per insp) $   8.00
  ATP inspection cost:  (per insp) $   0.50
  Purge inspect cost  (per insp)  : 5   6.53
  Pressure insp cost  (per insp)  : S   0.69
  IM240 insp Incremnt over Purge: $   0.87
  Avg. Gasoline Coat(per gallon): $   1.25
       Repair Costs
                               Pre-81
                                          81+
Avg. I/M repair cost
Air Pump repair cost
Catalyst replacement cost
Mlsfueled catalyst cost
Evap* system repair cost
PCV system repair cost
Gas cap repair cost
Purge repair cost
Pressure repair cost
S 50.
5 15.
$ 150.
$ 175.
S 5.
5 5.
S 5.
$ 70.
$ 38.
75. (
15.
165.
190.
5.
5.
5.
70.
38.
                                                     Vehicle percentages:63.219.7
                                                     Growth rate:   0.0 % per year
                                                          Fleet size Is  1000000. vehicles In base year 1986
                                              IM240 repair - $ 120. NOx repair- $  0. )
       Benefits  (1000 tons/yr) and Costs  (1000 $/yr) are averaged over the calendar years 2000 thru 2000.
Calendar Year 2000
Fleet Size 1000000
Baseline No-Program Emissions:
TotVOC
21.955
Program
CO
123.217
Benefits
ExhHC
10.929
and Costs
Evap
4.024
Refuel
2.067
RnLoss
3.897
RstLos
1.
038
Average Annual ATP Benefit,Fee    0.496    4.648
Evap/Runloss Benefit, Fee         3.583
Evap/RunLoss MPG Benefit
Avg Annual I/M Benefit, Fee       1.925   30.973
Avg Annual ATP Repair Cost
Avg Ann. Prg/Press Repair Cost
Avg Ann. IM Repair Cost,FallRt
Average Annual IM Fuel Savings
Exhaust Tampering Deterrence      0.210    2.315
Evap
 1.873
        RnLoss
         1.710
ExhHC
2.631
Ann Avg  STZTT   37.937
     (   15443
                                                        228.
                                                       3057.
                                                       3363.)
                                                       4006.
                                                        888.
                                                       3577.
                                                       7015.
                                                       9979.)
            5429.
94290  kilograms per day )

-------
CEM4.1 : Run »3  :  Low Option
                                                   Local Parameters Selected:
                                                                                            01/24/92  13:42:09


Base RVP: 11.5
In-use RVP: 8.7
Vehicle Type: LDGV
Veh. Speedsi 1575

Tampering deterrence start:
Oldest model year covered:
I/M Inspection frequency:
I/M program type:
1981 4 later MiR test type:
Vehicle types covered:
I/M waiver rate (pre-1981) :
I/M waiver rate (post-1981):
I/M compliance rate:

ATP start year (January 1) :
ATP inspection frequency:
Ambient Temp: 87.5
Operating Mode: 20.6

In-use Start Xr: 1992
LDGT1 LDGT2
1575 T575

1983
1968-2020 Fixed
Annual
Centralized
Idle
LDGV, LDGT1, LDGT2
1.0%
1.0%
98.0%
/ 87.5 / 87.5 (F) Region: Low

/ 27.3 / 20.6 Altitude: 500. Ft.
Minimum Temp: 72.
Maximum Temp: 92.
HDGV LDDV LDDT HDDV
T575 1575 T575 T575
I/M Program Selected:
Stringency: 20



IM240 test model year
Purge check model year
(F)
(F)
MC
19. i





coverage: 2020+
coverage: 2020+
Pressure check model year coverage: 2020+


Anti-Tampering Program Selected:
1553
Annual
ATP compliance rate:
ATP program type:



93.01
Centralized
   Oldest model  year  covered:
   Vehicle types covered:
   Parameters  covered by ATP;

           Cost  Assumptions
                                    1981-2020  Fixed
                                    LDGV , LDCT1,  LDGT2
                                    Catalyst     /  Fuel Inlet
                                                      Fleet Assumptions
                                                                            LDGV  LDT1  LDT2   HDGV  LDDV   LDDT   HDDV    MC
   I/M  Inspection  cost:(per InspJS8.00
   ATP  inspection  cost:  (per insp)  $  0.50
   Avg. Gasoline Cost(per gallon):  $  1.25
                                                     Vehicle percentages:  63.2  1977   3.1   3.6   0.2   0.2   3.3
                                                     Growth rate:   0.0 % per year
                                                          Fleet size is  1000000. vehicles in base year 1966
                                                                                                                      T78
       Repair Costs
                                Pre-81
                                          81+
                                             (  IM240 repair - S 150. ROx repair- S  0. )
 Avg.  I/M  repair cost     :  $  537     TSl
 Air Pump  repair cost     :  $  IS.     IS.
 Catalyst  replacement cost:  $ ISO.    165.
 Misfueled catalyst  cost  :  $ 175.    190.
 Bvap. system repair cost :  $   5.      5.
 PCV system  repair cost   :  5   S.      5.
 Gas cap repair cost      :  $   5.      5.

       Benefits (1000 tons/yr)  and Costs (1000 $/yr)  are averaged over the calendar  years  2000 thru  2000.
Calendar Year 2000
Fleet Size    1000000
Baseline No-Program Emissions:
                                                             Evap
                                  21.9S5  123.217   10.929

                                  Program Benefits and Costs
                                                                       2.067
                                                                              RnLoss

                                                                                3.897
                                                                                         1.038

Average Annual ATP Benefit, Fee
Evap/RunLoss Benefit, Fee
Evap/RunLoss MPO Benefit
Avg Annual I/M Benefit, Fee
Avg Annual ATP Repair Cost
Avg Ann. Prg/Press Repair Cost
Avg Ann. IM Repair Cost,FallRt
Average Annual IM Fuel Savings
Exhaust Tampering Deterrence
VOC
0.591
0.034

1.697




0.210
Evap RnLoss ExhHC Ann Avg 2.229
0.032 0.003 2.194 (
5539
CO
T7554"


26.634




2.315
30. S3i
76136
Cost
439.
0.
( 0.)
7280.
384.
0.
5192.
( 3609.)

988S.
kilograms per







0.0738



day )

-------
CEM4.1 : Run »5  Medium Option
                                                   Local  Parameters  Selected:
                                                                                            01/24/92  13:42:28

Ambient Temp: 97.5
Operating Mode: 20.6
Base RVP: 11.5
In-use RVP: 8.7 In-use Start Irs 1992
Vehicle Type: LDGV
Veh. Speeds: 157?

I/M start year (January 1) :
Tampering deterrence start:
Oldest model year covered:
I/M inspection frequency:
I/M program type:
1981 t later MIR test type:
Vehicle types covered:
I/M waiver rate (pre-1981) :
I/M waiver rate (post-1981) :
I/M compliance rate:
LDGT1 LDGT2
157? T5T?

1983
1983
1968-2020 Fixed
Annual
Centralized
Idle
LDGV, LDGT1, LDGT2
1.0%
1.0%
98.0%
/ 87.5 / 87.5 (F) Region: Low
/ 27.3 / 20.6 Altitude: 500. Ft.
Minimum Temp: 72. (F)
Maximum Temp: 92. (F)
HDGV LDDV LDDT HDDV MC
157? — rSTS — 1575" 157? T5T?
I/M Program Selected:
stringency: 20
IM240 test model year coverage:
Purge check model year coverage:
Pressure check model year coverage:





2020+
2020+
1971+
Anti-Tamperlng Program Selected:
ATP inspection frequency:
Annual
ATP compliance rate: 98.0%
ATP program type: Centralized
   Oldest model year covered!
   Vehicle types covered:
   Parameters covered by ATP:

           Cost Assumptions
                                    1961-2020  Fixed
                                    LDGV , LDGT1, LDGT2
                                    Catalyst    / Fuel Inlet  /
                                                      Fleet Assumptions
                                                                           LDGV  LDT1  LDT2  HDGV  LDDV  LDDT  HDDV
  I/M Inspection cost:  (per  Insp)  S   8.00
  ATP Inspection cost:  (per  Insp}  $   0.50
  Purge inspect cost  (per insp)  :  $   6.53
  Pressure insp cost  (per insp)  :  $   1.94
  IM240 insp Incrennt over Purge:  $   0.87
  Avg. Gasoline Cost(per gallon):  S   1.25
                                                     Vehicle percentagest  6772iST/   971   376072072373   179
                                                     Growth rate:   0.0 % per year
                                                          Fleet size is  1000000. vehicles in base year 1986
       Repair Costs
                                  -81
                                             T  IM240  repair -  $ 150. NOx repair- S  0. )
 Avg. I/H repair cost      :  S   SO.     75.
 Air Pump repair cost      :  $   IS.     15.
 Catalyst replacement cost;  $  ISO.     165.
 MisCueled catalyst cost   :  S  175.     190.
 Evap. system repair cost  i  $    5.       5.
 PCV system repair cost    :  S    5.       5.
 Gas cap repair cost       :  $    5.       5.
 Purge repair cost         :  9   70.     70.
 Pressure repair cost      s  $   38.     38.

       Benefits  (1000 tons/yr)  and Costs  (1000 $/yr) are averaged over the calendar years 2000 thru 2000.
Calendar Year 2000

Baseline No-Program Emissions:
                                TotVOC
                                           c
                                           ExhHC

                                           10.929
                                 21.955  123.217

                                 Program Benefits and Costs
                                                            Evap    ReCuel

                                                             4.024
                                                                                      RstLos

                                                                               3.897    1.038
                                  voc
                                            CO
                                                     cost
Average Annual ATP Benefit, Fee    0.287    1.684        HTT
Evap/RunLoss Benefit, Fee         2.262                1765.
Evap/RunLoas MFC Benefit                          (     1766.)
Avg Annual I/M Benefit, Fee       1.697   26.634       7280.
Avg Annual ATP Repair Cost                              594.
Avg Ann. Prg/Pres* Repair Cost                         1743.
Avg Ann. IM Repair Cost,FailRt                         5192.
Average Annual IM Fuel Savings                    (     3609.)
Exhaust Tampering Deterrence      0.210    2.315
                                                                0.0738
Evap
 1.383
RnLoss
 0.878
                 ExhHC
                 2.194
                         Ann Avg  4.456   30.633      11628.
                               (   11075    76136  kilograms per day )

-------
Mobile,  output  from GEM 4.1    1/24/92

Run MO  High Option  (Biennial)
MOBIlE4.1(4Nov91)
M120 Warning:
                           MOBILE4.1 does not model most 1993 and later clean Air Act
              requirements; Emission factors for CY 1993 or later are affected.

I/M program selected:
    Start year (January 1):
    Pre-1981 MYR stringency rate:
    First model year covered!
    Last model year covered:
    Waiver rate (pre-1981):
    Waiver rate (1981 and newer):
    Conpliance Rate:
    Inspection type:
    Inspection frequency
    Vehicle types covered:
    1981 4 later MYR test type:
 1983
  20%
 1968
 2020
  1.%
  1.%
 98.%
 Centralized
 Biennial
 LDGV - Yes
LDGT1 - Yes
LDGT2 - Yes
 HDSV - No
 2500 rpm / I
    MOBILE*.1 IM240 Transient Biennial I/M Credits .8/15 (8/5/91)
Oldest model year covered by:
    IU240 Transient Test                1984
    Purge System Check                  1984
    Fuel System Pressure Check          1971

Anti-tampering program selected:

    Start year (January 1):
    First model year covered:
    Last model year covered:
    Vehicle types covered:

    Type:
    Frequency:
    Compliance Rate:

    Air pump system disablements t
    Catalyst removals:
    Fuel inlet restrlctor disablements:
    Tailpipe lead deposit test:
    EGR disablement:
    Evaporative system disablements:
    PCV system disablements:
    Missing gas caps:
 1983
 1975
 2020
 LDGV ,  LDGT1,  LDGT2

 Centralized
 Biennial
 98.0%
 Yes
 Yi
 Yi
 Y.
 Ho
 Ye
 Ye
 Yi
       Minimum Temp:  72.  (F)
       Period  2 RVP!  8.7
                      Period 1 RVP: 11.5

VOC HC emission factors include evaporative HC emission factors.
                                                                            Maximum Temp:  92. (F)
                                                                       Period 2 Start Yr: 1992
Cal. Year:
2000
I/M
Anti-tarn.
Program:
Program:
Yes
Yes
Ambient
Operating
Temp: 87.5
Model 20.6
/ 87.5
/ 27.3
/ 87.5
/ 20.6
(F) Region:
Altitude:
Low
500.
Ft.
  Veh. Type:
                                     LDGT2
                                                                                                        All  Veh
Veh. Speeds:
VMT Mix:
19.?
0.584
Composite Emission Factors
VOC HC:
Exhaust HC :
Evaporat HC:
Refuel L HC:
Runing L HC:
Rsting L HC:
Exhaust CO:
Exhaust NOX:
Cal. Year: 2000

Veh. Type:
Veh. Speeds:
VMT Mix:
1.16
0.48
0.16
0.19
0.23
0.11
6.32
0.72
I/M
Anti-tarn.
LDGV
19. g
0.584
Composite Emission Factors
voc HC:
Exhaust HC:
Evaporat HC :
Refuel L HC:
Runing L HC:
Rsting L HC:
Exhaust CO:
Exhaust NOX:
1.77
0.71
0.36
0.19
0.42
0.11
10.05
0.75
19.?
0.199
(Gin/Mile)
1.47
0.80
0.15
0.25
0.18
0.10
7.06
1.05
Program:
Program:
LDGT1
19.?
0.199
(Gm/Mlle)
2.25
1.20
0.35
0.25
0.36
0.10
11.96
1.15
" 19.?
0.080

1.67
0.89
0.23
0.25
0.20
0.10
8.20
1.13
No
No
LDGT2
' 19.?
0.080

2.59
1.37
0.46
0.25
0.40
0.10
14.40
1.25



1.53
0.82
0.18
0.25
0.19
0.10
7.38
1.07
Ambient
Operating
LDGT



2.35
1.25
0.38
0.25
0.37
0.10
12.66
1.18
19. S
0.035

4.54
2.07
1.44
0.40
0.51
0.12
36.27
4.51
Tempi 87.5
Mode: 20.6
HDGV
19.?
0.035

4.54
1.07
1.44
0.40
0.51
0.12
36.27
4.51
19.?
0.002

0.65
0.65




1.60
1.37
/ 87.5 /
/ 27.3 /
L0DV
~ 19.?
n.002

0.65
0.65




1.60
1.37
19
0

0
0




1
1
87.5
10.6
.?
.001

.84
.84




.74
.50
197? 	
0.093

2.15
2.15




11.21
9.30
19.6
0.007

4.96
1.89
2.52


0.54
24.78
0.77



1.495
0.796
0.206
0.192
0.203
0.098
8.223
1.750
(F) Region: Low
Altitude: 500.
LDDT
19
0

0
0




1
1

.001

.84
.94




.74
.50
HDDV
19.6
0.0?3

2.15
1.15




11.21
9.30
rt.
Mr-
lS. 6
n.oo7

4.96
l.?9
2.52


0.54
24.78
0.77

Ml Vsh



S.084
1.049
11.382
0.192
0.364
0.098
11.874
1.799

-------
Mobile  output  from GEM 4.1.   1/24/92

Run *5  Medium Option
MOBILE4.l(4Nov91>
M120 Warning:
                           MOBILE4.1 does not model most 1993 and latar Clean Air Act
              requirements; Emission Factors for CY 1993 or later are affected.

I/M program selected:
    Start year (January 1):
    Pre-1981 MYR stringency rate:
    First model year covered:
    Last model year covered:
    Halvar rate (pre-1981):
    Waiver rate (1981 and newer);
    Compliance Rate:
    Inspection type:
    Inspection frequency
    Vehicle types covered:
    1981 i later MXR test type:

Oldest model year covered by:
    IH240 Transient Test
    Purge System Check
    Fuel System Pressure Check

Anti-tampering program selected:

    start year (January 1):
    First model year covered:
    Last model year covered:
    Vehicle types covered:

    Type:
    Frequency:
    Compliance Rate:
                                        1983
                                         20%
                                        1968
                                        2020
                                         1.%
                                         1.%
                                        98,%
                                        Centralized
                                        Annual
                                        LDGV - las
                                       LDGT1 - Yes
                                       LDGT2 - Yes
                                        HDGV - No
                                        Idle
                                        2020
                                        2020
                                        1971
                                        1983
                                        1981
                                        2020
                                        LDOV ,  LDGT1,  LDCT2

                                        Centralized
                                        Annual
                                        98.0%
    Air pimp system disablements:       No
    Catalyst removals:                  Yes
    Fuel inlet restrictor disablements: Yes
    Tailpipe lead deposit test:         No
    EGR disablement:                    No
    Evaporative system disablements:    No
    PCV system disablements:            No
    Missing gas caps:                   No
                      Period 1 RVP; 11.5
                                              Minimum Temp:
                                              Period 2 RVP:
                                                            72.
                                                            8.7
                                                                 (F)
     Maximum Temp:   92.  (F)
Period 2 Start Yr:  1992
VOC HC emission factors include evaporative HC emission factors.
Cal. Year: 2000

Veh. Type:
Veh. Speeds:
VMT Mix:
I/M
Anti-tarn.
LDGV
19.?
0.584
Composite Emission Factors
voc HC:
Exhaust HC:
Evaporat HC:
Refuel L HC:
Rnnlng L HC:
Rstlng L HC:
Exhaust CO:
Exhaust NOX:
Cal. Year: 2000

Veh. Type:
Veh. Speeds:
VMT Mix:
1.35
0.52
0.21
0.19
0.32
0.11
7.01
0.72
I/M
Antl-tam.
LDGV
19.6
0.584
Composite Emission Factors
VOC HC:
Exhaust HC:
Evaporat HC:
Refuel L HC:
Runlng L HC:
Rstlng L HC:
Exhaust CO:
Exhaust NOX:
1.77
0.71
0.36
0.19
0.42
0.11
10.05
0.75
Program:
Program:
LDGT1
19.?
0.199
(Gm/Hlle)
1.66
0.86
0.20
0.2S
0.26
0.10
8.07
1.05
Program:
Program:
LDGT1
19.?
0.199
(Gm/Mlle)
2.25
1.20
0.3S
0.25
0.36
0.10
11.96
1.15
Yes
Yes
LDGT2
" 19.?
0.080

1.90
0.97
0.29
0.25
0.29
0.10
9.48
1.13
No
No
LDGT2
' 19.?
0.080

2.59
1.37
0.46
0.25
0.40
0.10
14.40
1.15
Ambient
Operating
LDGT



1.73
0.89
0.22
0.2S
0.27
0.10
8.47
1.07
Ambient
Operating
LDGT



2.35
1.25
0.38
0.25
0.37
0.10
12. ««
1.18
Tempi
Mode:
87.5 /
20.6 /
HDGV
19
0

4
2
1
0
0
0
36
4
Temp:
Mode:
.6
.035

.54
.07
.44
.40
.51
.12
.27
.51
87.5 /
20.6 /
HDGV
19
0

4
2
1
0
0
0
36
4
.6
.035

.54
.07
.44
.40
.51
.12
.27
.51
87.5 / 87.
27.3 / 20.
LDOV
19.?
0.002

0.65
0.65




1.60
1.37
87.5 / 87.
27.3 / 20.
LDDV
19.?
0.002

0.65
0.65




1.60
1.37
5 (F) Region:
6 Altitude:
LOW
500.
LDDT HDDV
19.? 191
0.001 0.

0.84 2.
0.84 2.




1.74 11.
1.50 9.
5 (F) Region:
6 Altitude:
?
093

15
15




21
30
Low
500.
LDDT HDDV
19.6 l?~
0.001 0.

0.34 2.
0.84 :.




1.74 11.
1.30 5.
?
093

15
15




21
JO

Ft.
MC
19.6
0.007

4.96
1.89
2.52


0.54
24.78
0.77

Ft.
MC
19.6
0.007

4.96
1.99
2.52


0.54
24.78
n.77


All Veh



1.661
0.838
0.252
0.192
0.281
0.098
3.927
1.752


All Veh



2.084
1.049
0.3»2
0.192
0.364
o.o«e
11.874
1.799

-------
Mobile  output  from GEM 4.1    1/24/92
Run *3 : low Option
MOBILE4.1UNOV91)
M120 Warning i
                           MOBILE4.1 does not model most 1993 and later Clean Air Act
              requirements; Emission Factors for CY 1993 or later are affected.
I/M program selected:

    Start year (January 1):
    Pre-1981 MYR stringency  rate:
    First model year covered:
    Last model year covered:
    Waiver rate (pre-1981):
    Waiver rate (1981 and newer):
    Compliance Rate:
    Inspection type:
    Inspection frequency
    Vehicle types covered:
    1981 C later MYR test type:

Oldest model year covered by:
    IM240 Transient Test
    Purge System Cheek
    Fuel System Pressure Check

Anti—tampering program selected:

    Start year (January 1):
    First model year covered:
    Last model year covered:
    Vehicle types covered:

    Types
    Frequency:
    Compliance Rate:
                                        1983
                                         20%
                                        1968
                                        2020
                                         1.%
                                         1.%
                                        98.%
                                        Centralized
                                        Annual
                                        LDGV - Yes
                                       LDGT1 - yes
                                       LDGT2 - Yes
                                        HDGV - No
                                        Idle
                                        2020
                                        2020
                                        2020
                                        1983
                                        1981
                                        2020
                                        LDGV ,  LDGT1,  LDGT2

                                        Centralized
                                        Annual
                                        98.0%
    Air punp system disablements:       No
    Catalyst removals:                  Yes
    Fuel inlet restrlctor disablements: Yes
    Tailpipe lead deposit test:         No
    EGB disablement:                    No
    Evaporative system disablements:    No
    PCV system disablements:            No
    Missing gas caps:                   No
                                              Minimum Temp:  72.  (F)
                                              Period 2 RVP:  8.7
                      Period 1 RVP: 11.5

voc HC emission factors include evaporative HC emission  factors.
                                                                            Maximum Temp:   92.  (F)
                                                                       Period 2 Start Yr:  1992
Cal. Year: 2000
Veh. Type:
Veh. Speeds:
VMT Mix:
I/M
Anti-tarn.
LDGV
":!M
Composite Emission Factors
VOC HC:
Exhaust HCi
Evaporat HCi
Refuel L RC:
Runing L HC:
Rsting L HC:
Exhaust CO:
Exhaust NOXt
Cal. Year: 2000

Veh. Type:
Veh. Speeds:
VMT Mix:
1.58
0.52
0.36
0.19
0.41
0.11
7.01
0.72
I/M
Anti-tarn.
LDGV
19.?
0.584
Composite Emission Factors
VOC HC:
Exhaust HC:
Evaporat HC:
Refuel L HC:
Runing L HC:
Rsting L HC:
Exhaust CO:
Exhaust NOX:
1.77
0.71
0.36
0.19
0.42
0.11
10.05
0.75
Program:
Program:
LDGT1
19.?
0.199
(Gm/Mlle)
1.90
0.86
0.34
0.25
0.36
0.10
8.07
1.05
Program:
Program:
LDGT1
19.? 	
0.199
(Gm/Mlle)
2.25
1.20
0.35
0.25
0.36
0.10
11.96
1.15
Yes
Yes
LDOT2
" 19.?
0.080

2.18
0.97
0.45
0.25
0.40
0.10
9.48
1.13
No
No
LDOT2
' 19.?
0.080

2.59
1.37
0.46
0.25
0.40
0.10
14.40
1.25
Ambient
Operating
LDGT


1.98
0.89
0.37
0.2S
0.37
0.10
8.47
1.07
Ambient
Operating
LDOT



2.35
1.25
0.38
0.25
0.37
0.10
12.66
1.18
Tempi 87.5
Mode: 20.6
HDGV
19.?
0.035

4.54
2.07
1.44
0.40
0.51
0.12
36.27
4.51
Temp: 87.5
Mode: 20.6
HDGV
19.?
0.035

4.54
2.07
1.44
0.40
0.51
0.12
36.27
4.51
/ 87.5 /
/ 27.3 /
LDDV
~ 19.?
0.002

0.65
0.65




1.60
1.37
/ 87.5 /
/ 27.3 /
LDDV
~~19.? 	
0.002

0.65
0.65




1.60
1.37
87.5
20.6
(F) Region: Low
Altitude: 500.
LDDT
19
0

0
0




1
1
87.5
20.6
.001

.84
.84




.74
.50
HDDV
o!o93

2.15
2.15




11.21
9.30
Ft.
MC
19.6
0.007

4.96
1.89
2.52


0.54
24.78
0.77

All Veh


1.870
0.838
0.379
0.192
0.363
0.098
8.927
1.752
(F) Region: Low
Altitude: 500.
LDDT
19
0

0
0




1
1
.3 	
.001

.94
.94




.74
.50
HDDV
197? 	
0.093

2.15
2 .13




11.21
9.30
Ft.
MC
19.6
0.007

4.96
1.89
2.52


0.54
24.78
0.77

All Veh



2.084
1.049
0.382
0.192
0.364
0.09S
11.874
1.799

-------
                        APPENDIX L









IDENTIFYING EXCESS EMITTERS  WITH A REMOTE SENSING DEVICE:




                  A PRELIMINARY ANALYSIS

-------
                                                                              911672
        Identifying Excess Emitters with a  Remote
            Sensing  Device:  A Preliminary  Analysis
                                               Edward L. Glover and William B. Clemmens
                                                            U.S. Environmental Protection Agency
ABSTRACT                                   systems  work by focusing a  beam,  or  in  some
                                             cases  multiple  beams,  of infrared light  across
   There  has   been  considerable  interest  in  the  roadway  into  an  infrared  detector.   The
applying  remote  measuring  methods to sample  instrument  determines   the  concentration  of
in-use  vehicle  emissions,  and  to  characterize  the  pollutant in  the path of  the beam based  on
fleet emission  behavior.    A  Remote  Sensing  the  amount  of  infrared  light  absorbed  by the
Device   (RSD)  was used  to  measure  on-road  detector  at  specified  wavelengths,  and  the
carbon   monoxide  (CO)   emissions   from  theoretical relationship  of carbon monoxide  to
approximately  350  in-use  vehicles  that  had  carbon  dioxide   in  auto  exhaust.    Early
undergone  transient mass  emission  testing at a  equipment  had  only  carbon  monoxide  (CO)
centralized  I/M  lane.    On-road  hydrocarbon  capability,  while  later  equipment   may  have
(HC) emissions  were also  measured  by  the RSD  the   potential   to  measure  other  emissions
on about 50  of these vehicles.   Analysis of  the  species  (primarily  hydrocarbons  - HC).
data indicates   that  the  RSD  identified  a
comparable number of  the high  CO emitters as     Two  potential  uses  of  remote  sensing
the two  speed  I/M  test only  when  an  RSD  devices  (RSD)  have been proposed   by  various
cutpoint  much   more   stringent than  current  sources.     They  are:  (1)    the   on-road
practice  was  used.   Both  RSD and  I/M  had  identification  of  gross  polluting   vehicles
significant errors  of  omission  in  identifying  (commonly   called   High   Emitters)   for
High CO Emitters  based  on the mass  emission  subsequent repair,  and   (2)   the  monitoring  of
test.  The test data were  also  used  to study  the  on-road  vehicle  emissions in  a   specified area
ability  of the  RSD  to  characterize  fleet  CO  over a  period  of time  for  program evaluation
emissions.                                    purposes.  Such  potential  uses can only  come  to
                                             fruition,   however,  if  the   remote  sensing
INTRODUCTION                              techniques are  shown to  provide accurate and
                                             reliable  results  that reflect   the   vehicle's true
   Researchers  at  the   University  of  Denver  emission  levels  over  a  variety  of  normal
and  elsewhere   [1]*,  [2],  [3], [4] are in  the  operating  conditions (e.g.,  acceleration, cruise,
process  of  developing   systems  to   remotely  hot/cold  operation,  etc.).
measure  the  concentration  emissions  from
vehicles  operating on the public  roads.   These     Several  test  programs wuh  remote  sensing
                                             devices  have  been  conducted by a  variety  of
   	——	                  organizations  including;  the   University  of
* Numbers in  brackets denote references listed at  the  Denver   (by  Professor   Donald   Stedman),
end of the paper.                                General   Motors  [4],  the  EPA   Environmental

-------
                                                                                        911672
Monitoring and  Support  Laboratory  (EMSL) in  emissions   from   a   given   vehicle  would
Las Vegas [2], and  the California Air Resources  categorically   match  IM240   mass  emissions
Board (CARS)  [5].   Many of these  studies have  from  the same vehicle.   For example, it  was of
involved  measuring  the  RSD  concentration  interest to  know  if  a  vehicle  with  a high  RSD
emissions  from  thousands of vehicles  [1],  [3].  concentration   would  also  have  high  IM240
Others  were   concerned  with   verifying  the  mass  emissions,   and   vice   versa.     The
accuracy   of   the   RSD   measurements   by  effectiveness  of the RSD  system in  identifying
directing  small  samples  of  exhaust  with  a  gross  emitters and  excess emissions  was  also
known  concentration into the  RSD beam  and  evaluated  relative  to  the  capability   of  the
comparing  the   results  to  more  traditional  standard  Indiana  I/M  test using  the  IM240  as
analyzers  [2].    Another  involved   a  roadside  the yardstick  for  excess  emissions.   Finally, the
pull-over  program  conducted by  CARB   which  effect  of  external  variables,  such  as  vehicle
compared   RSD  results  to   I/M  tailpipe  and  operating  mode,  owner  response,  weather
tampering  checks   on   in-use   vehicles   [5].  conditions,   and   siting  factors  were  also
However,  none  of  these programs   provided  a  investigated  as  to  whether  they   affect   the
quantitative  comparison  between  RSD  results  ability of the   RSD  system to  correctly identify
and   corresponding  transient   dynamometer  vehicles  with   high  emissions.
mass  emission  results on the  same set of  cars.
                                                   All RSD and IM240 testing  was conducted in
   The  goal  of this  study was  to fill this  gap  Hammond,   Indiana,   by   Automotive  Testing
and  to  provide a  comparison   between  RSD  Laboratories Inc.   (ATL)  under  contract  to  the
results  and  dynamometer  results  on  the same  U.S.  Environmental  Protection   Agency.
cars.  The driving  schedule  which  was utilized
for the  comparison was  the IM240.  It  is a  new  TEST EQUIPMENT
transient  driving  schedule  developed  by  the
EPA  [8] that consists of  the  first  two "hills" of     Figure  1   shows a  schematic of the RSD
the   new   car   certification   procedure,   the  system  in  a  roadside  setting.    The  principal
Federal   Test  Procedure (FTP),  with   some  parts  of  the system include the  IR detector and
modifications   (further   details   are  discussed  source;  a video  camera  to  record  the  license
later   in   the   paper).      Although   linear  plates  of passing  vehicles;  a   modified  police
correlations  between  the   RSD  and   IM240  radar  gun;  a  personal  computer equipped  with
emissions  were   performed,  most  of   the  an A/D  board; and special  software  developed
investigation was focused on whether the  RSD  by the University  of  Denver researchers.   Also
                                                                 included  in  the  system   are
                            Fieurel                              l^e    special    calibration

           REMOTE SENSING  DEVICE SCHEMATIC         S±T  ""  """'  '"
                            IR SOURCE
  VIDEO
  1.0.
  CAMERA
               COMPUTER
   The  RSD  system  operates
by   continuously  monitoring
the   intensity  of   the  IR
source.     When   a  vehicle
breaks  the   beam   path,  the
reference   voltage   drops  to
zero   which   signals   the
presence   of   the   vehicle.
Span   voltages    collected
before the  beam  is  blocked
and  zero  voltages  during the
blockage  are recorded.   As
the  vehicle  exits  the  beam,
samples  are   taken  over  one
second at 125 Hertz.   The C02
spectral  region  is  isolated by
filtering  at 4.3 |im and the CO
spectral   region  is  filtered  at

-------
911672

 4.6 u,m.  The instantaneous  CO  and CO2  values
 are then  regressed  using  a  linear  least  squares
 procedure.   From  the  slope  of this  regression
 the  average  CO/CO2  molar  ratio  (Q)  and  the
 average  HC/CO2 molar ratio  (Q1)  are  obtained
 and  reported.    The HC  ratio is  calibrated and
 reported  in  terms  of  propane.

    The average  Q  and Q' molar ratios  over  the
 one  second  sample  time  are  the  only  emission
 measurements  that,  are  recorded  during  an RSD
 test.   Only  these  are  recorded because  the size,
 position  and distribution  of  the exhaust  plume
  is not  known.   Therefore, to  help  safeguard  the
  quality    of   the    subsequent   emission
  calculations  which   are   based  on   these
  theoretical  relationships,  the  system  employs a
  built  in  feature  to evaluate  the  reasonableness
  of  the   observed   molar   ratio.      If  an
  unreasonable molar  ratio  is  observed,  the RSD
  reports  a  'non-linearity  error1.

     The  majority  of testing  was  conducted  using
  the   original  RSD  design  developed  by  Dr.
  Stedman.    This  unit  (designated  in the  paper as
  RSD  #1) had a  liquid  nitrogen  cooled  detector,
  and measured only  carbon  monoxide  (CO).   It
  used   two   indium   antimonide   photovoltaic
  detectors.   A second unit was used in  the later
  stages of   testing.     This   second  RSD  unit
  (designated  as RSD  #2) was air cooled,  and had
  both  HC and CO  measurement  capability.   The
  two   systems   also   had   slightly   different
  versions   of  software  controlling   the  data
  acquisition  and  processing.

  DESCRIPTION OF TESTING

     Two  different  RSD  testing  formats  were
  employed  during  this  study.    These   formats
  included  track testing at the  ATL  (Bendix) test
  facility  in   New Carlisle, Indiana, and  on-road
  testing   of   in-use   vehicles  in  Hammond,
  Indiana.     The   RSD  track  testing   was  very
  limited,   lasting   only   three   days   at   the
  beginning  of the  test  program,  and   employed
  the  first RSD unit (RSD #1).   The  purpose of the
  track   testing  was  to  compare   RSD  results
  collected   under   controlled  track  conditions
  with  FTP results  on the  same  set of vehicles.   A
  second  purpose    was    to  evaluate   the
  repeatability  of  the RSD  emissions by  replicate
  testing of   the  same  vehicle  under  controlled
  conditions.   A total of ten (10)  cars were  tested
  at  two test-track  sites.    At  the  first  test-track
  site,  five cars received multiple  RSD tests  on a
level roadway  at  speeds of 5. 10, 20, 30,  and 40
MPH.   At the second track  site, five  other  cars
were   tested  several  times   on  an  inclined
roadway  with  a  3  percent  grade  at  the same
speeds.    All   ten   of  the   vehicles   which
participated  in  this  part  of  the  RSD  test
program  had   been  recruited  to  the  ATL
laboratory    for   other   emission   testing
programs  and  were simply selected  for  the RSD
testing  based  on availability.    As  a  result of
participating  in  the  other   programs,  these
vehicles   underwent  FTP testing  and  repairs.
However,  none of the vehicles  received an RSD
test  after  repair.

   RSD   testing   at  roadside   sites  of   in-use
vehicles,   which  had   received  a   transient
dynamometer test, was  the  major effort of  this
study.   This RSD testing was  conducted  at  two
expressway  ramp  sites,  and  a  secondary  street
with two-way  traffic.  The  majority of this  on-
road testing  was conducted  at the expressway
sites,  and  employed  the  first  design  RSD  unit
(RSD  #1).  The  second design unit (RSD #2)  was
used  at  the second  test site,  a fairly  lightly
traveled,  two-lane  expressway  service  drive
with two-way  traffic.   Both  RSD test sites were
located  less  than one mile  from  the  specially-
modified  centralized  I/M  facility  in Hammond,
Indiana,  which  administered  the   transient
IM240  dynamometer  test.

   The IM240  is a  new  two bag  mass emission
test  conducted  on  a   dynamometer  over  a
transient   driving   schedule.     The   driving
schedule  for  the  IM240 test  consists of the  first
two "hills" of the new car  certification test,  the
Federal  Test  Procedure   (FTP),   with  some
modifications    to   include   more  transient
operation.   However, the driving  schedule does
not  include  a   vehicle  engine-start.     The
dynamometer  inertia  weights   and  horsepower
settings  were  selected  based on a  consolidated
list of the new  car certification settings for the
vehicle  model.   The  emissions  were  measured
using  a  CVS-CFV  system with  laboratory  grade
emissions  analyzers.    For  further  information
on  the  IM240  consult  EPA   technical   report
number EPA-AA-TSS-91-1 [8].

   The vehicles  which participated  in  the RSD
testing were  a subset  of vehicles  selected  for a
larger   program  designed   to  monitor  the
emission  performance  of in-use vehicles  using
the IM240  dynamometer  test.    These  vehicles
were selected  for IM240  testing  as  they  entered

-------
4                                                                                         911672

the  Hammond  I/M   facility   based   on  the      On  the  third day  of  testing at this site, the
availability  of   the   dynamometer,   and  the  problem of  vehicles  slowing  down  was  solved
owner's  willingness  to  participate.     If  the  by  moving  to  a second  site on  the  ramp about
vehicle  owner  declined to  participate,   that car  25  yards from  the  ramp exit.  The road grade at
went  through   the   normal   Indiana   I/M  this point  was  also  2.75  degrees.   Because the
procedure  (idle  and  2500 RPM no load), and the  vehicles were   about  to  enter  the  freeway,  it
next  car   in  line  was  chosen  to  receive  an  was more  difficult for the the  vehicle  owners
IM240.    Only  after the vehicle had  completed  to  slow down  to  look at  the  equipment while
the IM240 test was the owner  approached about  passing through the RSD  beam.  As a result, the
participating  in the  RSD drive-by test.    Drivers  average  vehicle speed  was  more  than  25  MPH
were  asked to  drive by the  RSD site  and were  and vehicle  operation was believed  to  be fairly
given  a  monetary  incentive  if  they  indicated  representative   of   ordinary  expressway   ramp
that  they   would   participate  (cars  were  not  driving.
stopped  at  the  RSD  test  site  itself).    This
additional  recruitment  procedure  for  the  RSD      After receiving  the  new  RSD  unit (RSD  #2),
testing  was  needed only   for  the  expressway  it  was decided  to move  to  a  test site  on  a
sites,  in   that   the  secondary   street  site  was  secondary   road with two-way   traffic,   about
located  so that owner  participation  was blind.    one-quarter  of  mile  from  the I/M  lane.    The
                                                 site was a straight, flat road with a posted speed
    After receiving the  IM240 test,  the  car  was  limit of 35  MPH.    It was  situated  such  that
inspected    for    catalyst   tampering    and  somewhat  more  than half  of the cars  exiting
misfueling.   It then received a  pressure  test of  the  I/M  lane  would  pass  through  the   test
its  evaporative emission control  system.   This  section.   Thus, it  was  possible  to  conduct the
procedure   took bnly a  few  minutes.   However,  RSD tests  at   this site  without  involving  the
during  this  period  the  vehicles  were  turned-  vehicle owner.   The  RSD device was set up so
off  to  conduct  the evaporative  pressure test,  that the beam  crossed both  lanes of  traffic.
The  effect  of  this   vehicle  shutdown  on
subsequent  RSD   emissions   is  unknown.  OPERATIONAL EXPERIENCE
However,   since  the  shutdown  was brief,  it is
expected to be minimal.                              The original  goal  of this  test  program  was
                                                 to  conduct RSD tests on  all the vehicles which
    The two  expressway  test sites  were  on the  received IM240 tests.    However,  for  various
northeast  cloverleaf  of the  Cline   Avenue  and  reasons not  all  the  test  vehicles  received  the
1-94  Interchange  in Hammond,  Indiana.    The  RSD test.   As shown  in Tables  1 and 2,  described
cloverleaf  was  an  entrance  ramp  leading onto  more  fully  in  the   following  sections,  there
Westbound  1-94  from   Northbound   Cline  were   several   reasons    for   vehicles   not
Avenue.     The   geometry  of  the  ramp  was  receiving  the  RSD  test,  including  recruitment
circular   with   a   slightly   upward   grade  problems,   RSD   equipment   problems,   and
throughout.                                     inclement   weather.

    Two  different test  sites  on   the  expressway      VEHICLE  RECRUITMENT  -  As noted  earlier,
ramp were  used.   The first site  was  about 50  the  vehicle  recruitment  procedures  differed
yards from the ramp entrance.   The road grade  between test sites.   At  the expressway  site it
at  this point  was  about  2.75   degrees.   Also,  was necessary  to inform  the owner of  the  RSD
because  the ground  surrounding  the   roadway  test  and   offer   a   monetary   incentive   for
was relatively  level,  this  site  was convenient  participation.    In  some  cases,  the  owners
for  setting  up  the  equipment  and parking  a  declined to  participate,  or  in  some  instances,
supporting  van.     However,  it  proved   to be  agreed  to   participate  (accepted  the monetary
unsatisfactory   because  it   allowed   curious  incentive),  but  never showed  up  for the  RSD
vehicles  owners   to  slow  down while passing  test.   In other  cases,  they  showed  up,  but  were
through the RSD  beam.   Where  it  was  estimated  missed,  because   of  communication  problems
that typical  average speeds  on  that section of  from  the  IM240  lane   to  the  roadside  RSD
the ramp  were around  15 to 25  MPH,  most of  operator,   alerting   the   operator  that   a   test
the test vehicles  were  found to  be  travelling at  vehicle was on  its  way.
speeds of 5 to 15  MPH.

-------
911672
  Sii£

  X-way
  2-lane

  Combined
        Table 1
   Vehicle Recruitment

        Owner Declined
IM240   /Vehicle Missed
 407
 673
71
12Q

191
Actual RSD
Participation

 336 (83%)
 146fSS%t

 482 (72%)
     Most  of  these  RSD  refusals  to  participate
  were  due  to  the.  vehicle  owner  citing  time
  constraints  or  other reasons.    Many  of  these
  refusals occurred  at  the  end  of  the  day  when
  the  I/M  lane  was  backed-up.    Unfortunately,  it
  was  noticed  that   a  few  of  the  owners  of
  vehicles  with  very  high  IM240  emissions,  or
  with signs  of  tampering  (fuel  inlet  restrictor
  disabled) also  tended  to  decline.    In addition,
  several  vehicle  owners  declined   because  they
  did  not want to drive on an  expressway.

     As  a result of  owner  refusals and  operator
  misses,  only 336 vehicles  out  of 407  vehicles
  actually  drove  by  the expressway  site  where
  an RSD  measurement  was  attempted (see Table
  1).

     The second  site  was  partially  chosen  so  as
  to avoid the need  to inform  the  vehicle  owners
  about  the  RSD  test.  Thus,  direct  recruitment
  was    not   necessary,  and    vehicle   owner
  involvement   was   minimized.     This   helped
  insure  that vehicle  owners  would not  change
  their driving patterns  in  the  RSD  test  section,
  and  avoided  the  issue  of some  owners  with
  high emitting  vehicles declining  to  participate
  in the  RSD test.  The  other  reason  for choosing
  a second site  was to obtain results at a location
  that  would  be  expected  to  have  different
  vehicle  operation.

     Unfortunately, as  shown  in Table  1,  a high
  RSD  test  rate  was  not  achieved  at  this  site
  either.    This  was  due   to  the  fact  that the
  location  of  the  second  site  allowed  IM240
  vehicles  to  exit  the  I/M  test  lane  in two
  directions.   Only one  of  these  directions  took
  the  vehicle by  the  RSD  site.    Therefore,  as
  indicated in Table  1,  only 146 vehicles  out  of
  266 (55%)  participated  in the  RSD testing.  The
  other   120  vehicles   were  not  involved  because
  they exited  from   I/M  lane  in   the  opposite
  direction  from the RSD test site, or they missed
  the  RSD test  because  of lack of  communication
  between the lane and the  RSD  operator.
    RSD  EQUIPMENT OPERATION  -  Operation
of  both  RSD systems was fairly straightforward
with   the   set-up   and   operating  procedures
being  nearly  identical  for both  systems.   The
only   differences   were   due   to   software
upgrades,   and   the   absence   of   cryogenic
cooling  for  the  RSD   #2  (two-lane  testing)
system.  Most of the basic  set-up and  operating
problems  were  addressed during the  first few
days   at  the  first   on-road  site   (the  initial
expressway   location).      These   problems
involved   the   typical   learning   curve,
experienced   when   operating   unfamiliar
equipment.      Beyond   that,   testing   was
reasonably    routine,   interrupted   only   by
occasional  equipment  problems,  and  inclement
weather.

    Severe   equipment   problems   with  either
RSD  units   were infrequent.    However,  one
major problem did occur  with the RSD #1.   The
problem  was  the  failure  of a  computer  board
which   controlled  the  video  acquisition  and
storage, and  it resulted in  a few days  of  down
time.    Once  diagnosed,  the board  was replaced
with a new  one  (provided  by  the  supplier  of
the RSD equipment).    The second equipment
problem  which  resulted  in  a  couple  days
downtime was the  theft  of the  roadside power
generating  equipment.    This occurred at  the
two-lane site  while  RSD  #2  was operating.   It
illustrates  a  potential  practical  problem  with
this type  of  emission   testing.    Other  more
minor    equipment   problems   which   were
resolved by   the  equipment  operator  included:
(1) the  heating  element  used   to  produce  the
collimated RSD beam burned out several times,
(2) rain entered  the  detector  and  had  to  be
dried   out  using a  heat  gun,  (3)   high  winds
potentially    causing    the   instrument    to
erroneously   make  a  reading were present  at
times.

    The number  of vehicles  lost  to   RSD testing
due  to equipment  problems  with  each of  the
RSD  units   is  shown  in  Table 2.   Overall,
approximately  16 percent  of the  RSD tests were
missed.   However,  many  of  these  problems
were  related  to the  emerging  nature  of this
technology,  and  as  it  matures,  the number  of
RSD tests lost to equipment  problems  would be
expected to   decline  drastically.

    WEATHER  EFFECTS  - One  of  the objectives
of this  study  was  to  determine if the RSD  could
operate in  inclement  weather such as  rain  or

-------
                                                                                            911672
snow.   Track  testing  was
excluded    from    this
requirement.   However,              Actual RSD
the  On-road  RSD  testing    §itg      Participation  Malfunction
was   attempted    when    x            336
inclement   weather   was    2-lane        145
present     during    the
expressway   testing   in
August,   November,   and
                               Table 2
                         Vehicle Capture Rates
Combined     482
                      Equip
      28
      21

      52
             Weather
                                 27
     27
          True RSD  Percent of   w/o Equip
           Sample
281
122

403
Participation Malfunct.

   (84%)      (91%)
   (84%)
(89%)
December.   Testing in  inclement  weather  led  to
some  missed RSD tests.   At the  expressway site,
27  out  of  the  336  participating  vehicles  (8
percent)    were  clearly  missed due  to  rain  or
snow  (see  Table 2).   Also, as indicated  in  the
instrument  description,  the  RSD  unit  has   a
built-in  quality   control   algorithm  that  flags
illogical  results  as  'invalid  readings'.    Some  of
the RSD  results discussed  in  the next  section,
and  labelled 'invalid'  may  have  been  tested  on
a  rainy day.   However,  it was  not  possible  to
segregate  the  recorded data  in  a  manner that
would  allow  an   answer  to  that   question.
Therefore,  the  actual  loss  due  to  weather
conditions   could   have   been   higher  than
indicated.

   Testing  of  the  RSD #2 unit  at  the two-lane
site was  done  in late  March.   During this period
no  inclement  weather  testing  was done due  to
the  generally  fair  weather  conditions  on   all
but two  days.    During  these two  days, RSD
testing   was    not    attempted   because   of
hailstorms,  and   the   vehicles  which   received
IM240 tests on  those  days are not  included  in
Table  2  under weather  losses.

   In  addition  to  lost   test  time,   inclement
weather   also   resulted   in   many   of  the
equipment  problems  which  were  mentioned
above.   Nevertheless,  it  was found  that  testing
could  occur when  very  light  rain was present,
although  the  number  of  non-linear   CO/CO2
errors  increased  if  the pavement  was wet  and  a
large   water "rooster  tail"  was  created  behind
the car that interfered  with the  IR  beam.   In
addition,    the   intermittent   rain   became   a
nuisance  since  all  the   equipment  had  to   be
covered  and  uncovered.
   RSD  SAMPLE -  The true  RSD
sample,  shown  in  Table  3,   was
computed   by   removing   the
number of  vehicles  that were not
tested  because  of  recruitment
problems,  equipment   problems,
or inclement  weather.    This   true
                    sample  (403  vehicles  overall) contains  all  the
                    vehicles  which  were attempted to  be  tested  by
                    the RSD under,  as best  we  could  determine,
                    reasonable  conditions.    The table includes both
                    the valid and  'invalid'  readings  based  on  these
                    measurement  attempts.   Invalid   reading  being
                    those   flagged  by  the  unit's internal  quality
                    control   algorithm.

                       Subsequent  analysis  of  the RSD's capability
                    to  identify  High  Emitters  used  only the  valid
                    readings  because  traditionally, only  valid  I/M
                    and IM240  tests  are used  for  such analyses.  If
                    necessary,  invalid  I/M  or  IM240 tests can  be
                    repeated.   However, unlike  I/M  or  IM240 tests
                    which  can  be  easily repeated,  RSD  tests  might
                    not be  conveniently  repeated.    A RSD  second
                    chance  test  would require  the vehicle  to  drive
                    past  the instrument  again   under as  nearly  as
                    identical  conditions  as  possible.    Since  repeat
                    RSD testing was  not in  the program  plan, it  was
                    not done.   Therefore,  in  assessing  the  RSD's
                    capability   to  identify  High   Emitters,   an
                    argument could be  made that the entire  True
                    RSD Sample' (in Table 3) should   be  used,  since
                    the 'invalid'  sample  may   include  some   High
                    Emitters.

                       Based on   consistency  with  past practices,
                    only  the valid  RSD readings  in   Table 3  were
                    used  for subsequent  analyses.   However,  this
                    decision  should  be  reviewed  in  'any   future
                    programs  relative  to  the   type  of  vehicle
                    sampling  planned,  since  the  effect of  using
                    the True RSD  Sample'  (as opposed to only 'valid
                    RSD  tests')  would have  the effect  of reducing
                    the   effectiveness   of  the   RSD   unit   in
                    identifying   High  Emitters.
        Site

        X-way
        2-lane

        Combined
True RSD
 Sample

   281
   122

   403
      Table 3
    RSD Sample

 Valid    Invalid
CO RSD   CO RSD

  257      24(9%)
  22      23(23%1
                                                             356
          47(12%)
        Valid
       HCRSD


          52.

          53
           Invalid
              RSD
           69(57%)

             69

-------
911672                                                                                           7

     To  clarify,  an invalid  test result as tabulated   TEST RESULTS
  in  Table  3,  could have  resulted for two reasons.
  First,  if  the  standard deviation of  the  CO/CO2      A description  of  each  vehicle tested along
  ratio,   which  is  measured   more   than  one   with its  test scores from  the IM240,  RSD,  and
  hundred times  per second, was greater  than 20   the  Indiana   I/M  tests  can   be   found  in
  percent, a  flag  would  be set.   This  prevents   reference   [9].
  rapidly  changing  CO/CO2  ratios  from  being
  recorded  as  valid.   Second,  if  the  instrument   HIGH  EMITTER  IDENTIFICATION
  could not detect a CO plume,  a  flag would  be set.
  This   prevents   pedestrians,   multiple    axle      The  primary goal  of  this  project  was  to
  trucks,  and  other  accidental   beam  blockages   evaluate  the  ability  of the  RSD  concept  to
  from being  recorded as  valid.    Wet  conditions   consistently  identify  vehicles  with  excessi-ve
  could   cause  both  these  RSD   errors  to  occur,   CO  emissions   as  determined  by   transient
  however,  as  indicated,  this potential  source  for   dynamometer  testing.    In  other  words,  could
  'invalid results'  could  not  be separated  from   the RSD properly  categorize  vehicles  (i.e., pass
  other  causes  in the  recorded data.                or fail)  based on their IM240  mass emission  test
                                                    scores.   RSD   hydrocarbon measurements  were
     The number  and percentage  of  valid  and   added later in  the  program,  and were  similarly,
  invalid RSD  CO results  for each  site suggest   but less  extensively,  analyzed.
  different   levels  of  performance  between  the
  two instruments.   For example  in  Table 3, RSD      To  achieve  this goal,  the   analysis  focused
  #1  shows  a  fairly  low  rate  of invalid readings,   on  several   areas.    The  first  addresses a  brief
  Whereas,  RSD  #2  at the  two-lane site shows  a   comparison  and correlation  of  the IM240  with
  high  rate  of  invalid   readings  (approximately   the  Federal  Test  Procedure  (FTP)  results.
  25  percent).     This  higher   rate  of  invalid   Logically,  the   next  would be  to address  the
  readings suggests that RSD #1   is  more  efficient   ability   of   the  RSD  to  categorically  identify
  and  sensitive  to low   CO  concentrations  than   vehicles  as  gross  or   high   emitters  versus
  RSD #2, possibly  due  to  the  cryogenic cooling   normal   emitters.      However,   before   the
  of  RSD  #1.    However,  the longer  RSD beam   categorical   analysis  can   take  place,   it  is
  length  at the  two-lane  road  (greater than  30   necessary  to define the  ground rules   for High
  feet) versus  the  expressway  site  (19  feet)  may   Emitters, and  from  this,  identify the  number  of
  also have  contributed  to  the higher  rate of   High Emitters in the  valid RSD sample.   Only
  invalid readings.                                 then can  the  ability  of the  RSD  to properly
                                                    categorize  vehicles   be  assessed.
      The rate  of RSD HC  invalid  readings made
  by  RSD   #2  (56   percent  in  Table   3)  is      IM240  VERSUS  FTP  -  In order  to  better
  unacceptably  high,   and   suggests   that  the  HC   understand  the  usefulness  of the  IM240 as  a
  capability  of  the RSD  in  its  current  state of   vehicle  to   evaluate  the RSD  results,  a brief
  evolution    needs     much    improvement,   comparison  of   the  IM240  to the  Federal  Test
  Potentially  one of the  problems  with the RSD   Procedure  (FTP)  is appropriate.   All  of  the
  HC  channel   is  the   very   tight   instrument   testing  to  compare  the  IM240  to  the   FTP was
  sensitivity    required  to   measure   relatively   conducted  in a separate  test   program,  and  a
  dilute  HC concentrations.   According  to  the   more  extensive  evaluation  of  those  results will
  manufacturer  of  the  RSD  unit, a difference of   be the  subject  of  some  future report.    However,
  100 ppm  HC produces a difference of only  1  mV   for the  purposes of this RSD  report,  a  simple
  by  the detector.   In  addition, the  fairly  long   regression   of   the  data  available  should   be
  RSD beam  length at  the  two-way  road site made   sufficient to  make  the point.
  the already  weak  signal  even   weaker,  and the
  detection   more   difficult.     Based  on  the      A regression of the available  IM240 and FTP
  experience  in  this program  with  an  admittedly   data  from the 300  vehicles  in Figure  2 indicates
  early   design   unit,  the  performance  and  the   reasonably   good  correlation,  as   demonstrated
  results  from  the  HC  RSD  channel  should be   by  the  correlation  coefficient   (r2 =  0.76)  and
  viewed with  caution.                             the slope  (b  = 0.99).     More importantly,  a
                                                    review   of   the  specific  vehicle data  indicates
                                                    that  the   IM240   is  excellent   at  identifying
                                                    vehicles  with high  excess  FTP emissions.   In

-------
8
                                          911672
                      Figure 2
               FTP Versus IM240
       0   20  40   60   80  100 120 140 160 180 200
                    IM240 CO (g/mlle)

fact,  for  virtually  all  vehicles  tested, where  the
model  year  ranged  between  1976  and 1989,
there were zero  errors  of  commission  by  the
IM240  relative to-  the  new  car  standards on  the
FTP  (i.e.,  all cars  that failed the  IM240 at  the
new car standards for HC,  CO, or NOx, also failed
the  FTP  for  the  same pollutant).    The almost
perfect   categorical   relationship  between   the
IM240  and  the  FTP  was  a  little surprising
because  the  IM240  does  not  include  the  cold
start  operation  found  in  the  FTP.     Only
warmed-up  operation  occurs  on  the  IM240,   and
the  results do  not  reflect  the  contribution  of
any  engine-start  emissions  (cold or  hot).   The
lack  of cold  operation  on  the IM240  should,
however,  not pose  a  significant problem,  since
all  RSD  measurements and  all  I/M  test results
in  this  study were  made on  warmed up  cars.
Therefore,   based    on   the    good   linear
correlation  of the   IM240  to the  FTP,  and  its
success  at  identifying  gross  CO  and HC  emitters,
it  can  be  viewed  as  a  representative  substitute
by  which  to judge  the RSD  results and  base  the
effectiveness of the RSD  systems.

    DEFINITION  OF   A  HIGH EMITTER   - A
necessary   part  of  the   analysis  needed   to
identify  High Emitters  was  to  determine   the
appropriate  level  with  which  to  ascertain  that
a vehicle  was a  High  Emitter.   Historically,  the
specific  levels have  been  related  to  a  vehicle's
FTP  emission   levels,   and  EPA's  modeling
programs  for  in-use fleets  (MOBILE4) [10]  have
used  a  factor of the  mean  plus  two times  the
standard deviation  of  the  mean fleet  emissions
as  the  High  Emitter  level.   To determine  High
Emitter  cutpoints   for the  IM240,  this  same
 procedure  was  applied  to  135 vehicles  which
 received  before  and  after  repair  IM240  tests
 in  another  EPA  test program.   The  resulting
 cutpoints  are   shown   in  Table   4.     As  in
 previous  analyses  of FTP  results,  this  testing
 showed  that  many  vehicles  whose  before-
 repair  emissions   exceeded   these  cutpoints
 usually  achieved  substantial  reductions  in
 emissions   from   repair.     Vehicles   whose
 emissions  were  less  than   the  cutpoints  in
 Table  4, often had no reductions or  a  very
 small  percentage  reduction  in  emissions  from
 repair.   In short,  our practical experience  has
 been  that  vehicles  with  mass emissions  above
 such  cutpoints  can  usually  be  repaired  to  a
 value  under  the   cutpoint  quite  effectively,
 while  those  below the  cutpoint  have   been
 very  difficult   to  repair  in  a cost  effective
 manner.

   Our focus in  this  analysis  was  to  evaluate
the  ability  of the   RSD  to  find High  Emitters
(HE) which  could  realistically be  expected  to
yield  substantial   emission   reductions  from
repair.   A vehicle is defined as  a HE if  its IM240
emissions exceed the cutpoints  shown  in  Table
4.    The  potential  emission reductions  from  the
repair  of  these  vehicles  was  defined  as  the
emissions  in  excess  of  me  Table  4  standards
(i.e.  greater than  10  g/mile CO  for  1983+ Model
Year  vehicles),   and  such   reductions  were
termed Repairable Excess  Emissions (RPEE).

   HIGH  EMITTERS  IN  THE SAMPLE  - The
cutpoints  in Table  4 were  used to  identify High
Emitters  (HE's) in the sample of valid RSD tests.
For  instance, at  the two  sites  there  were a  total
of 356  valid RSD  CO tests conducted.    Of that
total, 97  vehicles,  or 27 percent,  exceeded  the
IM240 cutpoints  in  Table 4.    Whereas,  only 53
valid HC  test were conducted  at  the two-lane
site.   The analyses  in the  following sections  use
the number  of valid  RSD  test in Table 5  as the
base  value   to    compute    certain    RSD
effectiveness  rates.

                     Table 4
         IM240 High Emitter* Cutpoints
 Model Year Grp       QQ            HC

 1976 -1980           30 g/mile      2.0 g/mile
 1981-1982           20 g/mile      1.0 g/mile
 1983 and Later         10 g/mile      1.0 g/mile

 *  Includes vehicles termed both "HIGH" and"SUPER"
 Emitters in MOBILE4

-------
911672
                      Tables
                    High Emitters

           Valid RSD IM240 CO
           CO Sample   Highs
                              Valid RSD IM240 HC
                             HC Sample   Highs
  X-way
  2-lane

  Comb
            257
            22
72 (28%)
25(25%)
            356     97(27%)
52.

53
                       S

                       8
    The model year  distribution  of the vehicles
  in the  'combined sample'  of valid CO tests  in
  Table  5  is  shown   in  Figure  3.    This  figure
  provides  a   finer  breakdown  of the  overall
  distribution  and  of  the  distribution of High  CO
  Emitters (HE).   It also  shows several  interesting
  points.    First,  the  1982  through 1985  model
  years   comprise  almost   50  percent  of   the
  sample,  and  they  also  include a  large portion
  of the HE vehicles.   Second, the HEs  make up  a
  discernible   portion  of   most   model  year
  samples,  even  some late  model  years.    For
  example,  for  the  1978   through   1985  model
  years,  typically 25  to  35  percent  of the  model
  year  sample  are  HE vehicles.    For  the  1989
  model  year,  the  percentage is between  5 and 10
  percent.
    To  further  simplify   subsequent  analyses,
  the  three model  year  groups  in  Table 4  were
  combined into two   groups -- a  1976  to  1982
  group,  and   a  1983  and  later  group.  However,
  the  cutpoints  in Table  4  were   still  used to
  define   High   Emitters   within  even   the
  consolidated  groups.

    By  grouping  the  97  HE   CO   vehicles
  (combined  sample  in Table   5)  into  these  two
  model   year  groups,  it  can  be shown  that  the

                        Figures
             Model Year Distribution
   20
   18

£ 16
£ 14
             Total Sample
             IM240 High Emitters
        76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
                       Model Year
older 1976-82  sample  contributes  less  than  half
(45 percent) of the total HEs  for CO.  This older
vehicle  group  also  only  contributes 46  percent
of  the  vehicles   to  the  overall  sample.    The
closeness  of the  fraction of HE's to that  for  the
overall  sample  distribution  indicates  that   the
older   vehicles   did    not    contribute   a
disproportionate  share   of the  HE  vehicles   in
this   program.

   RSD IDENTIFICATION  OF  HIGH EMITTERS
- The analysis  of the  RSD's ability to  properly
identify High  Emitters  focused  primarily on  CO
emissions,  since  a combined  total of  356 valid
RSD CO tests were conducted  at the two sites.   A
lesser  emphasis  was  placed  on  HC  emissions,
because only  53  valid HC  measurements were
obtained at  the one  site. These analyses  in most
instances  evaluated  the CO  sample  in the two
distinct  model  year  groups  (i.e.,  1976  - 1982
model  years,  and 1983 and  later  model  years).
However, due to the  small  size of the HC sample,
all  model  years  were  grouped  together.

   The  primary  data   set   used  in  the  analysis
was  formed  by  combining the  expressway (X-
way)  and   two-way  street  (2-lane)  data  sets
together.    This combination  was done  in  order
to  produce  a   larger,  and  potentially  more
representative  database.     However,  limited
separate   analyzes  were  conducted  on   the
results from  the  two principal  roadside  sites  to
illustrate   any   potential   differences,   and   to
determine   if  combining  the  data   masked any
important   results.

   RAW DATA - The first analysis was to scatter
  plot the  RSD  concentration data  versus  the
  IM240 mass  emission data.   Figures  4(a) and
  4(b) show the  combined sites  CO results  for
  each  model  year  group.   Figure  4(c)  shows
  the  HC  results  for all  model  years.    In
  viewing  these  scatter plots,  there  appears  to
  be  more  of  a  relation  between  the  RSD CO
  data and the IM240  values  for the older 1976
  -  1982  model  year  group  than  for  the 1983
  and later group.  In  any case,  the  regression
  correlation  coefficient   is  marginal  in  the
  case of  CO,  and extremely  poor  in the case  of
  HC.   Scatter plots, of course, provide  only  an
  overall  relationship,  and  Figure  4  does not
  provide   any  information  on the  ability   of
  the RSD to properly  categorize a  vehicle as  a
  High  Emitter  on the  IM240, or as  a  Normal
  Emitter.

-------
10
                                          911672
                       Figure 4
              RSD - IM240 Scatter Plots
                    (Combined Sites)
       to.o
(a)
        , 0 41983 and Lmt«r Vehicle* J
              20  40  60  M  100 120 140 160 180 200
                      IM240 CO (a/mlto)
       10.0
(b)
              20  40 60  80  100  120 140 160 180 200
                      IM240 CO (o/mito)
 (c)
                 2  3
                      4  S  6 7  8  9 10 11 12
                      IM240 HC (g/mito)
                           Figura 5
                     RSD Identification Rate
               (e)
   RSD HIGH EMITTER IDENTIFICATION RATE - The
ability of the RSD  to correctly identify  vehicles
with  high   IM240   mass   emissions  as  High
Emitters  is  a central  issue  in the evaluation of
the  performance  of  the  RSD  concept.    The
ability of the RSD  to  identify such  vehicles is
shown in Figure  S  as  a  function of RSD  CO
level.   A detailed table  of these values can be
found  in  reference  [9].   The  identification  rate
in Figure  S  is  defined  as   the  ratio  of  the
number  of   IM240  High  Emitting  vehicles
identified  by  the RSD- to  the  total   number of
High  Emitting   vehicles   identified  by   the
IM240.   If  the  RSD  were  to  identify all of the
High Emitters  identified  by  the  IM240,  the
identification  rate would be  100  percent.

   The  CO High Emitter rates  are separated  by
site  to  show any  differences  that  might   be
apparent  between  the  results.     Differences
were  possible   because   a   different  RSD
instrument was used at  each site,  and both  sites
had   different   traffic  characteristics,   (e.g.,
vehicle    speeds    and    accelerations).
Nevertheless,  a comparison  of  the  RSD CO High
Emitter  identification  curves  from   both  sites
(Figures  Sa and  Sb)  illustrates  that  roughly  the
same results  at  both  sites  were obtained over a
complete  range   of  RSD  CO  levels.    This is
especially  true in  the  case  of  the  1983 and  later
vehicles,     where     the    difference     in
identification  rates  between  sites  is typically
only   5  or  10  percent  (Figure Sa).    The   site
effect in Figure 5(b)  for the  1976 through 1982
      model   years  shows  a  slightly   larger
      difference   between   sites.     However,
      because  the results  from  both   sites  were
      rather  consistent   for  each model  year
      group, the RSD  data  from  both sites were
      combined   for  most   of  the*  analyses
      throughout  the   remainder of the  paper.

         Combining  the  RSD  results  by   site
      produced a  larger data set for analysis.   A
      further  combining  of the database  across
      the two model  year groups  would  produce
      a  still   larger data  set.    Based  on  the
      proportional    distribution   of    High
      Emitters by  model  year  in Figure  3, such
      combination   would  seem  appropriate,
      except possibly for some late model years.
      However, Figure  5(c) shows that the  RSD,
      at all  CO  levels, systematically  identifies a
      higher  proportion  of 1976  though  1982
      High Emitters than  1983  and   later  High
      Emitters.    Typically,  the  difference  is

-------
911672
                                                                                                 11
  around  10 or  20 percent.   This is evidence that
  although  the  distribution  of  High  Emitters  is
  similar  for  most  individual  model  years,  the
  ability of the RSD  to  find a  High Emitter is  not
  the  same, and  is a function of  model  year,  or
  model  year  group.    Thus,  combining  the  two
  model  year  groups for  subsequent  analysis  is
  probably not   appropriate.

     Unlike   the   CO   results,   no    sample
  stratification  was   done  for  the  HC  results
  because  the  RSD .HC sample  was relatively  small
  (only  53 valid  readings),   and  because  there
  were  very  few  High  HC  Emitting  vehicles  in
  the  1976 through 1982  model  year  range.    For
  example, only  one  vehicle in  the  1976  through
  1982 model  year group  was found to be a High
  Emitter.       Instead,    the   High   Emitter
  identification  rate as a function of HC  level was
  computed using  an  'all'  model  year  sample.
  This information is shown  in Figure 5(d).

     RSD OUTPOINT ANALYSIS - The RSD CO and HC
  levels shown in  Figure 5  can also  be used as a
  standard  to pass  or fail  vehicles.   The  exact  CO
  or  HC  values  (called   cutpoints)  which   are
  chosen,   would  reflect  the  desired   severity   of
  the  test, and  would  determine  the  number  of
  High  Emitting  vehicles  which  would fail.   The
  selected  values  are  usually chosen  based on  the
  results  of  a   cutpoint  analysis  designed   to
  maximize the  number  of High Emitters  found,
  and  minimize  the  number  of Normal  Emitters
  (i.e., not High Emitters)  which  might be  falsely
  failed.   For this cutpoint  analysis,  the  RSD  CO
  cutpoints range  from the very  tight 0.05% RSD
  CO cutpoint  to the  loose 7.50%  RSD  CO cutpoint.
  For RSD HC,  the  cutpoints ranged  from 50 ppm
  HC to 2500  ppm HC (Hexane equivalent).

     The   first   part of  this  cutpoint  analysis
  simply  compares the  number of  IM240  High
  Emitting vehicles identified  by  the  RSD  to  the
  total number of  vehicles  with an RSD  CO level
  above a given cutpoint.   As seen in  Figure  6,
  the  RSD in  all  cases  identifies  more  cars than
  are  truly High  Emitters  on the IM240.

     At cutpoints above  3%  CO in Figure 6(a)  for
  late model  cars,  there  appears  to be  reasonably
  close  agreement  in  identifying  High   Emitters
  between the  RSD and the IM240.   However,  at
  the    higher   cutpoints   fewer   cars   were
  identified.   As  the CO cutpoints are  tightened,
  the  the  RSD and  IM240 curves  begin  to diverge
  dramatically,    indicating    that    a   great
percentage of  these  new  vehicles  captured  by
the lower RSD cutpoints  are, in fact, low IM240
emitters   which  should  not  be  failed.    For
example,  at the extremely  tight 0.10%  RSD CO
cutpoint   the  number  of  1983  and  later  RSD
failures  was  more  than  140  vehicles, while the
number   of truly   High   Emitting  vehicles   was
only about 35.

    Similar comments  can  be made  for  the  1976
-  1982   model  year  group,  except  that  the
divergence between the  RSD  identification  and
the IM240 occurs  at a  higher  RSD cutpoint, in
this case 5%  CO (see Figure 6b).

                     Figure 6
        Failures versus High Emitters
              1883 + Vehicle* (Combined Sites)
                            RSD cars with
                            IM240 > lOg/mi
          0.00 1.00 2.00 3.00  4.00 5.00 6.00 7.00 8.00
                      CO Cut Point (%)
       100
     1
76 .'82 Vehicle* (Combined Sites))
                             RSD cars with
                             IM240 > lOg/mi  ,
          0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
                      CO Cut Point (%)
                  All Vehicles (Two-Lane Site)
                             RSD can with
                             IM240 > lOg/mi  ,
 500
                    1000  1500  2000  2500  3000
                      HC Cut Point (%)

-------
12
                                                           911672
   The ability of  the  RSD to  identify  HC High  vehicles  that  should   be  identified  would  be
Emitters  is   extremely  poor.    The  curves  in  missed.
Figure  6(c)   do  not   have similar  shapes,  and
tend  to  diverge  even  at high   HC   cutpoints.     The effectiveness of the RSD test in  terms  of
Furthermore,  the  number  of  High  HC  Emitter  High  Emitter   Identification   (RSD).   False
Identified  is  extremely   low  at   all  cutpoints  Failures  (Ec), and  False Passes  (Eo) is shown  in
above  1000  ppm.                                  Figure  7  for  both  model  year  groups.    A
                                                   detailed  table  of combined site  results  for  each
   As  seen   in  Figure 6,  the  RSD  can  falsely  model year  group can be  found   in  reference
identify  vehicles  that are  not  High  Emitters.  [9].
Not  shown  in Figure  6  is the opposite issue  of
High Emitting  vehicles  NOT  identified  by  the     As  indicated  in Figure  7(a) relatively  few
RSD.    Such errors,  generally  referred  to  as  1983  and later  High Emitters were  identified by
errors   of  commission (Ec)  or  False   Failures,  the RSD  system  (less   than  30  percent)  at  high
and  as  errors of omission  (Eo) or False Passes,  to moderate  cutpoints (7.5% RSD CO to 2.0%  RSD
are  inherent  in  the  selection  of  a  cutpoint  to  CO),  while  the  False   Pass  rate was extremely
maximize the number of High Emitters  found,  high  (greater  than 70  percent).    However, on
while   minimizing   the  number  of  Normal  the  plus  side,   the  False  Failure  rate  was
Emitters falsely  failed.                            generally  quite  low, with the exception of the
                                                   high rate  around 6.0%  RSD CO.   This exception
   False   Failures   were    determined   by  occurred  because  the   sample  of  vehicles  with
subtracting   the   number   of   IM240   High  very high RSD  CO scores  was so small  that one
Emitting  vehicles  that  were  identified  by  the  False  Failure   dominated  the  results.      A
RSD  from   the  total   number   of   vehicles  probably   more   representative   False  Failure
identified by the  RSD,  and  then  the  sum  was  rate   pattern  can  be   seen  in  Figure   7(c),
divided by  the  total  number
                                                                Figure?
                                                      RSD Identification Accuracy
                                 IPO.O
                                 10.0
of  vehicles   identified  by
the RSD.    In essence, this
calculation  is  a  measure  of
the    fraction    of   cars
identified  by  the RSD that
are NOT High Emitters.   A
large   value would  indicate
that   many  of the  vehicles
identified   are   not   High
Emitters,  and  should  not  y
have  been  identified.        *• *•••
   T?  i      n                5 40.0
   False    Passes    were
determined   subtracting the     20.0
number   of   IM240   High
Emitting vehicles  that  were
identified by  the RSD  from
the total  number  of  High
Emitters  identified  by the
IM240,  and  then  the   sum
was   divided  by  the  total
number  of  High  Emitters
identified   by   the  IM240.  g
This    calculation   is   a  S
measure  of  the  fraction  of  oc
High    Emitters     NOT
identified by   the  RSD.   A
large    value,    in    this
instance,   would  indicate
that   many  High  Emitting
                                                   Eo * False High Emitter PASS Rate by RSD
                                                   Ec « False High Emitter FAIL Rate by RSD
                                                   RSD * RSD Vehicle Identification Rate
                                     19834- Model Years (Combined Sites)
                                  0.0
                                 100.0 -
                                     19834- Model Ye
                                 00.0 • •••
                                 M.,
                                                                  100.0
                                                                   10.0
                                                                   iO.O
                                  1976 • 1982 Model Years (Combined Sites)
                                                                   40.0-
                                                                   20.0
                                   0.00 1.00 2.00 3.00 4.00 5.00 8.00 7.00 8.00
                                         RSD CO Cut Point (%)
               (X-way Site)
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
      RSD CO Cut Point (%)
            (b)
  All Model Years (Two-Lao* Site)
                                 40.0
                                 20.0
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
      RSD CO Cut Point (%)
            (c)
                                                                          500  1000 1500  2000  2500 3000
                                                                           RSD HC Cut Point (%)

                                                                                  (d)

-------
911672

 constructed only  from  the  RSD  results  at  the
 express way site.   It shows  that the  false failure
 rate  for   High  Emitting  vehicles  is  virtually
 zero  at  the higher  cutpoints.

     Despite the  exceptions,  the  overall  results
 from  the  1983  and  later vehicles  suggest that
 the RSD   will  not  identify  the majority  of  the
 High  Emitting  late  model  vehicles.    However,
 in  the  instances when  a  vehicle  is  identified  as
 a   High  Emitter  by  the  RSD  (at  the  higher
 cutpoints),  it  is  usually   correctly  identified.
 For example,  at the  frequently cited  cutpoint  of
 4.5%  RSD  CO, less  than 15 percent  of  the 1983
 and later  High  Emitters  were identified.  This
 means  that 85 percent  of  the  High  Emitters
 were  missed by the RSD at 4.5% RSD CO causing
  a  False Pass  rate  of 85  percent.   However,  the
 False  Failure  rate  was  only  10 percent  of  the
 RSD   failure  rate,  suggesting  that  virtually  all
  the late model RSD  failures  at 4.5%  will be High
  Emitters.

     The  1976  through  1982  model  years  show
  results  in  Figure  7(b) generally  similar  to  the
  1983   and  later   vehicles.    However,  they
  typically  had  higher  identification  rates   of
  High Emitters,  lower  rates  of False  Passes,  but
  higher  rates   of  False   Failures   at   identical
  cutpoints   (to  the  1983  and  later  model  years).
  These  somewhat  different  results   are   likely
  due  to the  inherently higher emission  levels
  of  the  older  vehicles  (because  they  were
  certified  to higher  new  car  standards),  and  a
  longer period  of  normal  deterioration  due  to
  age.   They suggest that in order to be fair to all
  model  year vehicles,  RSD  cutpoints should be
  based  on model   year  or   technology  type,
  instead of one single  value  such as  4.5% RSD
  oa

     Like  the  HC  RSD  results  shown in Figures
  5(d)  and  7(d),  the  High Emitter  Identification,
  and  the   False Failure  and   False  Pass  rates
  shown  in Figure   7(d)   were  quite  negative.
  They  indicate  that  at present,  the RSD  HC
  channel  cannot  be  used  to   accurately  and
  repeatedly  identify  High  HC Emitters  without
  simultaneously  falsely identifying  many  low
  emitting  vehicles.    For  example, even  at  the
  extremely  tight cutpoint of 25 ppm HC  (Hexane
  equivalent)  the  RSD   identified   about  80
  percent  of the  HC  High Emitters.   At higher
  and  more  reasonable  cutpoints,  the HC High
  Emitter Identification  rate  dropped  to  around
                                              13

10  percent.    Despite  this  low  rate,  the  False
Failure  rate  was  extremely  high.

   The  results also show  that  at  virtually  all
reasonable HC cutpoints  the  false  failure rate
exceeded  70  percent of  the  total  failure  rate.
In fact,  no cutpoint was  identified  which  could
reduce  the  false  failure  rate  below 70 percent
and  still  produce  a High  Emitter  identification
rate  of  more  than  20  percent.    In  addition,
during the test program  one vehicle had  IM240
HC emissions  exceeding  11 g/mile,  but recorded
an RSD HC score  of 0 ppm  HC.  Therefore, based
on  these  findings  in conjunction with the fact
that   more   than   half    of   the   RSD  HC
measurements  were  invalid,   strongly  suggests
that  at  the present  time,  the  RSD  HC system
cannot   make    repeatable   or    accurate
measurements  of  in-use   vehicle  hydrocarbon
emissions.

   Viewing the cutpoint  analysis for CO  on  an
overall   basis   would   indicate    that   the
percentage of  High Emitters  identified  versus
the overall number  of  failures is fairly high  at
cutpoints  greater  than  3.0%   RSD  CO.   This
suggests  that  typically  if a vehicle  tests   above
3.0%  RSD* CO it  is  likely  to be  a  High Emitter.
Overall  this is a  positive  sign, since it suggests
that  a moderate RSD CO cutpoint can  be  found
that  will  identify  at least  some  High Emitters
without  falsely  failing   a  large  percentage  of
the fleet.   On  the  negative  side,  however such  a
cutpoint  would  also  likely  miss  a  large portion
of the true  High  Emitters.  More High Emitting
vehicles   could be   identified  by   using   lower
cutpoints, but  that  would  increase  the numbers
of  False  Failures  to  possibly  unacceptable
levels.    Such  trade-offs  could  seriously   affect
the use of the  RSD  as a  screening  tool for High
Emitters.  One can only speculate, but  it may  be
that   requiring  a  high   RSD  reading   to  have
been  observed  on  several  different  days,   or  at
several  different  sites,  before  a   vehicle  was
classified  as  a   failure  would  improve the
balance    between   proper   and    improper
failures.

   REPAIRABLE EXCESS EMISSIONS IDENTIFIED BY
RSD  -   The   previous   section   on   cutpoint
analysis   described   the  effectiveness  of  the
RSD  system  in terms of  the  number  of  High
Emitting  vehicles  identified  by   the   RSD at
particular CO  or HC  cutpoints.    An  equally
important  consideration  is  the  amount  of
excess  emissions  represented  by  the  vehicles

-------
14
                                                                                             911672
identified  by   the   RSD  (i.e.,  which   High
Emitters  are  failed).    As indicated,  this paper
addresses  excess emissions as  emissions  above
the levels defined in  Table 4.  Defined here as
Repairable Excess  Emissions,  or  RPEE,  these
excess   emissions   represent  the   potential
emission   reductions   from   those  vehicles
identified  by  the RSD,  if they  were  properly
repaired.

   The RPEE  was calculated  for  each cutpoint
as  the  sum  of the  excess   IM240  emissions
from  all  High  Emitting vehicles  identified  by
the RSD  (i.e., those  above the  levels in Table
4), divided by  the  sum of  the  excess IM240
emissions  from  all   of the   High  Emitting
vehicles.   In  short,  it  is the fraction  of excess
emissions  from  all  cars  with mass  emissions
above  the  levels   in  Table  4  that   were
identified  by the RSD.

   The effectiveness  of the RSD  was analyzed
in  terms  of   RPEE  because   the  RPEE
identification   rate    reflects   the   actual
emission  level  of each  High Emitter, whereas
the High  Emitter  identification  rate accounts
only   for  the  number  of  High  Emitters
without   regard for   emission  level.    The
difference  between   the   two   effectiveness
parameters  arises because of the  effect that
vehicles  with  very   high mass emissions  can
have  on the total emissions of  all the vehicles
in a  given sample.   Typically,  these vehicles
(called  Super Emitters  in MOBILE4) comprise
a  large share  of the  excess  emissions in  a test
sample  (or fleet),  although   they  are  usually
few  in number.   Further,  such  vehicles  can
be  readily  repaired  to  moderate  emission
levels.     For  these  reasons,  their  correct
identification  is of  prime  importance  in the
control  of repairable  excess  emissions,  and  in
the evaluation  of the  RSD  system's  ability  to
identify   vehicles   with  repairable  excess
emissions.

   The first order  of  analysis  was  to evaluate
the RPEE  identification  rate of the  RSD.  The
difference  in  model  year groups  can be  seen
in  Figure  8.    The   results  show  that  at
cutpoints  greater than  4.0%  CO,  the  model
year  differences  are   pronounced,   with   a
greater  amount   of  RPEE  identified  for  1976
through  1982  vehicles than   for   1983   and
later  vehicles.   The  trend  of  identifying  more
from  the  old  cars  than  the newer  cars was
previously  noted  in  the  rate  comparisons  of
High  Emitter  identification  shown  in  Figure
5'(c).   However,  this  trend occurred  at  all CO
cutpoints  relative  to  the  number of  vehicles
identified.    This  previous analysis  suggested
that  the  RSD was more  effective  at  finding
1976  through  1982  High  Emitters  than  1983
and  later ones.   In  contrast to  this  previous
analysis.  Figure  8  shows  similar RPEE  rates
for   both   groups   at   tighter   cutpoints,
suggesting  that  if appropriate  cutpoints  are
chosen,  the  RSD can be  equally effective  at
identifying  the  excess  emissions  from  new
cars  as  well as old cars.

   To  better  understand  the  true relationship
between  the   number   of  High   Emitters
identified  by  the  RSD,   and  the  Repairable
Excess  Emissions   from  those  vehicles,  the
data   were  plotted  in Figure  9.    A  careful
examination  shows  that  for the  1983 and later
vehicles, at cutpoints  between  1% to 4% CO,
the RSD  identified less than half of the  High
Emitters.   However,  those few  vehicles  which
were  identified   contribute   significantly  to
the total  repairable  excess CO emissions.   At
relatively  tight  cutpoints,  the  RSD  seems  to
find   the  worst  of  the  1983  and  later  High
Emitters,  although,  not  all  of  them,  since
even  at extremely  tight  cutpoints,  the  RPEE
rate  was  only  around 80  percent.

   For   the  1976  through   1982   vehicles
(Figure  9b), the  story is a little  different,  and
possibly a  little more  positive.   For this  model
year  group,  the  RPEE   and  High  Emitter
identification  rates  appear  to   proportionally
track  each  other  better.    In  addition,  for
these  vehicles, the RPEE  rate  seems to be  a
more    linear  function   of  CO   cutpoint,
suggesting  a  stronger  relationship  between
number   of  vehicles   identified   by   RSD
emissions  and  repairable  excess  emissions.

                         Figure 8
              Rapalrabl* Excass CO Emissions
                    (Modal Year Effect)
         0.0
          0.00  1.00  2.00  3.00  4.00  S.OO C.OO  7.OO  t.OO
                      RSD CO Cut Point

-------
911672
                                                                                                      15
                                ftgun*
                     Vehicle vs Excess CO Emissions
                   D
                   RSD Vchlek Identification Rate
            1M3 Mid Later Model YUM
                                            1*7«. US2Model Yean
• xoe 4M ua ut ua
 RSDCOCMPotat
                                         ue tm ua 4.o» ue ua 7.00
                                              RSDCOCutPolm
                     missed   by  the  RSD.    Also,   as
                     mentioned   previously,   the  False
                     Failure  rate  (Ec)   in   identifying
                     the  number of  High Emitters  was
                     around    10   percent   of   the
                     identification  rate,  and  the  False
                     Pass  rate  (Eo)  was  more  than  85
                     percent  of  all  of the  High Emitters
                     in  the   model  year  group.    This
                     means that  10  percent of the  RSD
                     failures    identified    contribute
                     nothing  to the RPEE.   However,  85
                     percent   of  the   High  Emitters
                     which do   contribute  something  to
                     the RPEE  were not  identified.
     Like  the  High  Emitter  identification  rate
  shown in Figure 7, the RPEE  rates also need to
  be compared  with  the False  Failure  and False
  Pass  rates.   This  is  done to  see  if  a potential
  balance  between  high  RPEE  rates   and  low
  False Failure rates exists  for RSD.    To make
  the  comparison  easier,  the  False Failure  and
  False Pass  rates from Figures  7  were  plotted
  along with  the  RPEE  rate as  a  function  of CO
  and  HC cutpoint.    The  results of
  this   effort   are  shown  in  Figure
  10,   stratified  by   model   year
  group,  site,  and  pollutant  in  an
  analogous   arrangement  to  Figure
  7.
                                                       An optimal  balance  between  a
                                        high  RPEE  and  a  low  False  Failure  rate  is
                                        more  difficult  to  find  in  the  1976  through
                                        1982  sample.  Here,  both the  RPEE  rate and
                                        False  Failure rate seem  to  be  linear  functions
                                        of RSD CO  cutpoint.  However, if the goal is  to
                                        hold  False  Failures  to  less  than   20 percent
                                        then a 3.0% RSD CO  cutpoint  which  identifies
                                        60% of the  RPEE  seems  reasonable.   If False
                                        Failures need to  be  almost  zero, then a 6.0%

                                                        Figure 10
                                         Repairable Excess Emissions Identification
                                       |Eo = F«J
                                       EcsFal
                                       ""BE*
          = Fabe High Emitter PASS Rite by RSD
           Falae High Emitter FAIL Rate by RSD
             RSD Identification Rate of Repairable Excess Emissions
     For  1983   and  later  vehicles
  (Figure   10 a),   the  RPEE   rate
  (indicated  as  RPEE in  the  figure)
  ranged  from   10  percent  at  very
  loose  cutpoints to  70  percent  at
  about  2.0%  RSD   CO.    Further
  tightening  of  the  RSD  cutpoint
  produced   little additional  RPEE,
  but  dramatically   increased  the
  False   Failure  rate  (Ec).   This
  increased  False   Failure    rate
  suggests  that  the  2.0%  RSD  CO
  cutpoint   may   be   the   lowest
  practical    RSD   cutpoint    that
  should   be   considered   for  the
  later   model   year   vehicles
  (assuming that  the False  Failure
  rate   of  20   to  25   percent   is
  acceptable  at  the  2%   cutpoint).
  In contrast,  the  widely  reported
  4.5% RSD CO cutpoint had a  RPEE
  rate    of   about   25    percent,
  meaning  that   at  this  cutpoint,  75
  percent of the  RPEE  would  be
                            iM.a
                                1983+ Model Yean (Combined Slta)
                                1976-1982 Model Yean (Combined Sites)
                                                            M.a
                                                            21.1
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 (.00
      RSD CO Cut Point (%)
                            1M.I
  1983+ Model Year* (X-waj Site)
                                                              0.00 1.00 2.00 3.00 4.00 5.00 (.00 7.00 8.00
                                                                    RSD CO Cut Point (%)
                                                                          (b)
                                                                All Model Years (Two-Lane Site)
                               0.00 1.00 2.00 3.00 4.00 5.00 8.00 7.00 1.00
                                    RSD CO Cut Point (%)

                                          (e)
                                   500 1000 1500  2000  2500  3000
                                     RSD HC Cut Point (%)

-------
16
                                           911672
RSD  CO  cutpoint  is  more  appropriate.   For
comparison,  the  4.5%  RSD  CO   cutpoint
identified about  40 percent of  the  RPEE, and
surprisingly had  a False  Failure rate of  about
20  percent.

   Like  the  HC  High   Emitter  identification
rates, the HC RPEE  rates were extremely  poor.
In fact  they seemed to be  worse in the  sense
that  the  maximum  HC  RPEE rate was less  than
50  percent  of  the total  HC RPEE.   However,
this   was  principally  due  to  the   11  g/mile
IM240 HC  emitter which  had an RSD  HC test
result of 0 ppm.  Nevertheless,  the  HC RPEE
identification    was   very   low   for   most
cutpoints,  yet  the False  Failures  were nearly
80  percent  of  the  total  failures.     In  all,
coupling these  results with  all  the  other low
High Emitter   identification  rates,  and   high
False Failure and False Pass rates,  leads to the
conclusion  that  the  HC RSD  in  its  current
state  will  not   accurately  measure  the  HC
concentration  in  a   vehicle's  tailpipe   with
sufficient accuracy to  be able  to  determine  if
a vehicle is truly a High Emitter  and   in   need
of  repair.

RSD  and   I/M

   As indicated,  the  primary goal  of  this study
was   to  evaluate  the  ability  of  the  RSD  to
properly categorize  High Emitters.     However,
because  the IM240  vehicles had  also received
an   official  Indiana  I/M   test,    the   unique
opportunity  to  compare  the  RSD   results  with
I/M  scores  on the same  vehicles was  exercised.

   The  Indiana  I/M   test   is  a  centralized
biennial  contractor run  I/M  program  [11].   A
two-speed  idle  test (2500 RPM  and idle) is used
for   1981   and   later  vehicles.     The  207(b)
cutpoint  of 1.2%  CO  is used  for the idle  portion
of the test, and a  1%  CO value is  used for 2500
RPM.   The   207(b)  HC  standard of 220  ppm is
employed at both  speeds.   Only the idle  test is
used  for 1976  through 1980  vehicles.  The 1980
vehicles  are tested  against  a 2%  CO  -cutpoint,
while the  1976 through  1979 model  years  use a
3.5%  CO cutpoint.

   Because the  Hammond,  Indiana I/M site  was
also  used  to evaluate  second  chance   tests  and
preconditioning  cycles  as pan of  a  previously
completed   program,   all  of  the  vehicles were
tested consistent  with  the  recently  published
EPA  guidelines for short test  procedures  (12).
   RAW  DATA  -  Similar  to  the evaluation of
High  Emitter  identification by  the  RSD,  the
first  analysis  here  was  to scatter  plot  the  data.
Figures  ll(a)  and ll(b) address the I/M  CO  test
for the  late model cars,  while  Figures  ll(d)  and
ll(e)  provide  an  HC comparison  on these same
vehicles.  The idle  test for the older cars is in
Figure  ll(c).   About the best  that can  be  said
about these  data is  that the   relation   between
RSD  and  I/M  scores   is  poor   for  CO,  and
nonexistent  for  HC.    However,  it  must  be
remembered   that   the   relationship   of  short
tests  to  the certification  test is not  based on  a
relationship   of  scores,  but  a  relationship  of
pass/fail  categories.
                       FIGURE 11
                 RSD - I/M Scatter Plot
             linommaLju
   <«>
               o.a o.«  o.« o.m i.o  1.2 1.4  i.» i.e  a.o
                  I/M aOOO RPM T*at - HC (ppm x 1OOO)
             All Mod* Ynn
                  I/M Ml* TMt . HC (ppm « 1

-------
911672
             17
   RSD   VERSUS  I/M -  This
analysis     consisted     of
comparing    the    performance
parameters  previously  used  to
evaluate  the ability of  the RSD
to  identify  High  Emitters  and
excess   emissions  with  those
same  parameters  calculated  for
the I/M  tests.    For simplicity,
RSD  results are shown for only
three   cutpoints,  which include
the    frequently   referenced
4.5%   RSD  CO  cutpoint,   a
moderate  3.5% RSD CO cutpoint,
and   a   more  stringent   2.0%
cutpoint.  RSD  HC results faired
so  poorly  in   the  scatter  plots
that no   attempt  was  made  to  compare  them
with the  HC  I/M  results.

    The effectiveness of the RSD  and I/M tests
in  identifying  High  CO Emitters is  shown  in
Figure 12 for both  the  1983  and later vehicles
and for the 1976  -  1982 vehicles.   Examination
of Figure  12(a) for late model cars  shows that
the RSD  at moderate  and  high CO  cutpoints  is
generally  no  better  than   the  two-speed  I/M
test  (i.e.,  'Both'  in Figure 12a)   in  identifying
High  Emitters.  In  the  extreme  case, the RSD
found  less than  15  percent  of  the  High
Emitters  at  the   highly  reported   4.5%  CO
cutpoint,   while   the   two-speed   I/M   test
identified  almost  25  percent  of  the  High
Emitters.    However,  when the  more stringent
RSD  cutpoint of  2% CO  was used,  about  35
percent of  the  High  Emitters were  found.

    For  the  1976  through  1982  vehicles   in
Figure  12(b),  the  I/M  identification of  High
Emitters  was  marginally  better  than  the  RSD
identification  rate (at 4.5% CO).   Relative  to
the I/M  identification,  the RSD identification
improved  substantially   at   tighter   RSD
cutpoints.   For  example, at the 4.5% RSD CO
cutpoint,  the  identification  rate   of  1976
through  1982 High Emitters  was  only  around
25  percent,  slightly  less  than   I/M  rate,
however,  at the 2.0% RSD cutpoint, almost  60
percent  of  the  High  Emitters  were  found.
(Note  that  the  2500  RPM  test  results  are not
shown for  the  1981 and  1982 model years  in
the   figure   for  the   sake  of  consistency,
although,  they  would  boost  the  High  Emitter
identification  rate of  the   I/M  test  somewhat
for the  1976 -  1982 model  year group.)
                                                                     FIGURE 12
                                                            High Emitter Identification Rate
                                                                  (CO, Combined Sk*>
                                                        Mathematically,  the  RSD  and   I/M  tests
                                                    were   combined   to   determine   if   the
                                                    combination  of tests  would  identify  additional
                                                    High Emitters.   As  seen in Figure  12,  there is
                                                    an   increase   in   the   identification   rate,
                                                    suggesting  that   the   two  tests   identify
                                                    different  cars.   This  would seem   reasonable
                                                    since   the   tests    are    conducted   under
                                                    significantly   different   vehicle   operating
                                                    conditions,   and would be  expected to  find
                                                    problems   which   are  particular   to   each
                                                    operating  mode.   From the  analysis  shown  in
                                                    Figure 13, it is clear  that  a subset of different
                                                    vehicles  are identified by  each  test.   For the
                                                    1983  and  later  group,  the additional  RSD
                                                    identification is very  marginal   until the  2%
                                                    RSD cutpoint  is  applied.    However,  for the
                                                    older  model  year  group,   the  additional  RSD
                                                    identification is substantial at   all  cutpoints.
                                                    In  fact,   adding  the  I/M   and  4.5%  CO  tests
                                                    together  nearly  doubles   the   identification

                                                                       Figure 13
                                                         High  Emitting Vehicle Identification Split
                                                                        (CO only)
                                                                1983 and Lat*r Mr*
                                                                (Ule/ZSNI/MTett)
 197*-1982 MVs
  (Mhl/MTert)
                                                                 I/M
                                                                        RSD
I/M
RSD
                                                    4.5% RSD
                                                    3.5% RSD
                                                    2.0% RSD
                                                             TIYTAIf


                                                              13
                                                              13
                                                              13
                                                                         •mrua
                                                                                   TOTAl.il
         TOTAIJI

           10
                                                                                                 15
                                                                                                 25

-------
18
                                                                                             911672
rate  of  High  Emitters,  further  attesting  that
different older vehicles  fail  each  of the tests.

   It should  be pointed out.  though,  that  the
RSD  test  for the  newer  cars was  compared
against  the two-speed  idle test  in Figure  13,
whereas  the   older  cars  were  compared   to
only   the  single   speed   idle  test.     The
differences  in  I/M  test  type  could  possibly
explain   the   improved   identification    of
additional older cars  by the RSD.   In  addition,
the Indiana I/M CO cutpoint of  3.5%  CO  for
the older cars (1976-1979  models) may not  be
as stringent  as it  should  be.   Tightening  this
I/M  cutpoint  for  older  cars  would  increase
the  number  of High  Emitters  identified   by
I/M    such    that     the    identification
improvements  exhibited  by  the  RSD  in  Figure
12(b)  could  possibly  vanish.

   Another   important    factor    in    any
comparison  of I/M  with RSD are False Failure
rates.   False  Failure  rates  were  computed  as a
percentage   of  the  failure  rate,   and  they
indicate  the  percentage  of  improper  failures.
Combined with the  High  Emitter  identification
rate,  they are a  measure  of a test's  ability to
find  High  Emitters,  yet  avoid  failing
low  and  marginal  Emitters.
                                                   described 'RSD  CUTPOINT ANALYSIS.'  With this
                                                   method,  a  1983  and later  vehicle  failing the
                                                   RSD or I/M test, would only need an IM240 test
                                                   score   less  than  10  g/mile,   rather  than  3.4
                                                   g/mile, to be considered  a  False  Failure.    Also,
                                                   with  this  method,  cars  with  IM240  scores
                                                   between  3.4  and   10 g/mile  (labeled  Marginal
                                                   Emitters) would be  considered  False Failures.

                                                      Both  methods  are shown  in  Figure  14, the
                                                   first  in (a)  and (b),and  the  second  in  (c)  and
                                                   (d).   The False Failure rates  calculated in terms
                                                   of  only the High  Emitters  (second  method) are
                                                   substantially  higher  for  both  RSD  and   I/M
                                                   relative  to  the  False  Failure rates calculated by
                                                   method  1.    However,  the  False  Failure   rate
                                                   percentages  for the  two methods  were  based
                                                   on  different sample  sizes.   Therefore, a  direct
                                                   mathematical  comparison  can  not be  made.

                                                      Even  so,  a qualitative  comparison  can be
                                                   made.    In  those  cases  where  the  the  False
                                                   Failure rates for method  1  (Figures  14a  and b)
                                                   are  zero, it  is  probably  safe  to assume  that all
                                                   of  the False Failures determined  by  method  2
                                                   (Figures  14c  and  d)  were  Marginal  Emitters
                                                   with  IM240  scores  between  3.4 and  10.0  g/mile.
                                                                     HgurtU
                                                                 Falsa Failure Rate*
   For   the   comparison   of  False
Failures  between  RSD  and  I/M,  two
methods  were  used.    The  difference
was  in the level  of IM240 score below
which  a failed car  would  be  defined  as
a  False Failure.  The first method  used
new    car   certification   standards
applied  to the  IM240  results  as  the
boundaries for  False  Failures.    For
example,  if a  1983  and  later  vehicle
failed  either the  RSD or  the I/M  test,
and  had  an IM240 test   score  of  less
than   3.4   g/mile   (1983   new  car
standard),  it  was  considered  a  False
Failure.    This  method  of determining
False Failures is similar  to that  used  in
MOBILE4, except that MOBILE4 uses FTP
test  scores (as  opposed  to  IM240  test
scores).

   The second method used  to  compute
the  False  Failure  rate  utilized  the
IM240  High  Emitter  cutpoints  shown
in  Table  4 as the  boundaries for  False
Failures.   This method  is identical  to
the  method  used  in  the   previously
                                                               [tttrpintland mgli Emm»i« ]

                                              1*3 .nd Utw Mod*! Yon          _  l»7«throu«hl



Hp H.







n H H in i i
                                                                 { High Emltl»f« Only )
                                              lM3indL.tn-Mod.IY.
                                                                           1976 Ihrouih 19M Modd VMM

-------
911672

  Given  this assumption, it is  likely that
  a  majority   of  the  False  Failures
  identified  by  method   2,   where  the
  method 1  False  Failure  rate  is  greater
  than   zero,   were   also   Marginal  $'
  Emitters.     As  implied  previously,  ^
  Marginal  Emitters  are  more  difficult  *
  to repair,  and as a  result, less likely to  I
  contribute   to    overall    emission  |
  reductions.                              I

     Interestingly,  the  combination  of
  the  RSD  and  I/M   tests  shown  in
  Figure  14(a)  shows  a  lower  False
  Failure rate  than the  RSD  test alone.
  This apparent  anomaly  is  a result  of
  the  sample   size   increasing,   when
  combining the  RSD   and  I/M  samples, faster
  than   the  number  of   failures.     In   this
  particular  case,  the  I/M-only  False  Failure
  rate  was  zero,  therefore,  the  combined  rates
  were the  same as that for the RSD.   Thus,  the
  fraction  of  False  Failures   was  lower  for  the
  combined   sample   than  for   the   individual
  samples.

     Another  possible  concern  is the  similarity
  in   magnitude    of   the    High    Emitter
  identification  rates  in  Figures  12(a)  and  12(b)
  to the False Failure rates in Figures  14(c) and
   14(d).   The similarity  might suggest  that both
  tests  tend to  falsely  identify  a  vehicle  as  a
  High  Emitter almost  as  often  as they  correctly
  identify a  High Emitter.   However,  it  should be
  remembered  that  the  calculations  for  the  High
  Emitter identification  rate  were  based  on  the
  total  number  of vehicles   identified   as  High
  Emitters  by  the  IM240,  whereas   the False
  Failure  rate  was  based  on  the   number  of
  vehicles  failing  the  RSD  or I/M  test.    Because
  the number  of  IM240 High Emitters  and  the
  number   of   RSD    or  I/M   failures   were
  significantly    different,    any   numerical
  similarity  of  identification  rates  and False
  Failure rates  is  merely  coincidental.

     A  characteristic  that  must also be  evaluated
  when   comparing  RSD  to  I/M  are  the False
  Passes  or missed vehicles.   The  False  Pass  rate
  is the  fraction of High  Emitters not  identified
  by the RSD  or  I/M  test  relative to  the  total
  number  of  High  Emitters  identified   by   the
  IM240  (also   defined  near  Figure  7).    An
  alternative  calculation, in  this  case,  would  be
  to subtract  the  identification  rates  in Figure
  12 from  unity.
                                               19
                 FifunlS
              Falsa Pass Rates
              ( High Emitter. Only)
                                  (b)
    False  Pass  rates  are  shown  in  Figure  15,
and  as would  be expected, more stringent tests
(i.e.,  two-speed  idle  test),   or  tighter  RSD
cutpoints  reduce  the   number  of  vehicles
missed.   In  examining  both model  year groups
in  this  figure, it  is evident  that  the  Indiana
I/M  test missed fewer vehicles than the RSD  at
a  4.5%  cutpoint.    Only  at the  most stringent
cutpoints  did the RSD  substantially  reduce  the
number  of High  Emitters  missed by the  I/M
test.

   Finally, as  in  the  analysis   of  the  RSD's
ability  to   identify   High    Emitters,   the
comparison of  RSD  to I/M  needs to look  at  the
amount of repairable excess  emissions (RPEE)
identified  by  both  tests.    The  RSD  values  in
Figure 16  are  the same as  the  individual  RPEE
values in Figure 10.   The  I/M RPEE values were
calculated  in  a similar  fashion,  and  are  shown
in Figure  16.   Clearly,  the  Indiana I/M test  for
the  newer  vehicles  identified  more  repairable
excess emissions than the RSD  test at a 4.5% CO
cutpoint.     However,   at   the   most   stringent
cutpoint, the  RSD did better.

   For the  older  cars,  the  RSD  consistently
identified  more excess  emissions than  the  I/M
test.   In  this  case,  though,  it should  be noted
that  the  I/M test  for these cars consists of only
a single  speed  idle  test, and it is likely that  the
identification  rate   of  the  I/M  test   would
increase if a two-speed  idle test  were  used.

   In  reviewing the  comparisons between RSD
and  I/M in Figures  12  though  16, it  is  difficult
to  see any  substantial  improvement  over  the
I/M  test  with  the  RSD.   At  the  4.5% RSD
cutpoint, the  the I/M  test  for  the new  cars  (i.e.,

-------
20
                                          911672
                    Repairable Excess Emissions
the  two-speed  test)  identified   more   High
Emitters,  had  lower  False  Failures,  had  lower
False Passes,  and   identified  substantially  more
excess  emissions.    At the tightest  cutpoint,  the
RSD   did  identify  more  cars   and  excess
emissions, with  a  correspondingly  lower  False
Pass rate,  than  the  I/M test, however,  it did  so
at the  expense  of  an increase  in  False Failures.
Lowering  the  cutpoint   on  the   I/M  test   to
achieve  the  same  False   Failure rate  would   be
expected  to  produce  similar  increases  in  the
identification  of  High   Emitters  and  excess
emissions with  the  I/M test.

   For  the  older  vehicle  group, the RSD  faired
a  little  better.   At  the 4.5% cutpoint, the  single
speed  I/M   test  and  the  RSD  were  roughly
comparable  in  number of cars identified,  false
passes,  and excess  emissions.    At this  cutpoint,
the  RSD  did  have  a  measurably  lower  False
Failure  rate.   At the  lower  cutpoints, the RSD
performed  better  than the  single  speed  idle
test  on  the older  cars  with  comparable  False
Failures.

   In  reviewing  the  fictitious combination   of
the  I/M  and  RSD  tests   in these  figures,  some
interesting observations can  be  made.   For late
model   cars,   the   combination   increased
identification  somewhat over  the   I/M test,  but
with  the expected   increase  in  False  Failures.
However,   for  the   older   model  years,   the
combination   drastically    increased    the
identification  (number of vehicles,  and excess
emissions)   over   the   I/M   test   without   a
substantial increase  in False Failures.   In fact,
the combined tests  at  the highest  RSD cutpoint
(4.5%   CO)  increased  the  identification  over
both the  I/M and RSD tests while  reducing  the
False Failure rate  relative  to  the I/M test.
                Clearly,     the    fictitious
            combination  of I/M  and RSD  test
            did  better  with  the   older  cars
            than  than  the  newer cars,  and
            was better than either test  alone.
            However,    even    with    the
            combination,   only   30   to   40
            percent  of the  late  model  High
            Emitters and  60 to  70 percent  of
            the older ones  were found.  This,
            suggests  that  neither   test  is  as
            effective  as  it  should  be   in
            finding  High  Emitters.   In  the
            case  of  I/M,  there  are  several
            possible methods  to  improve  the
capture  of High  Emitters.     As  previously
indicated,  tighter  I/M  cutpoints  could   be
used,  or,  a  single  speed  idle test could  be
replaced  with  a  two-speed test - a two-speed
test  could  be  replaced  with a  loaded test, etc.
The  ultimate  replacement,  of  course,  would
be   the  IM240.    However,  for   RSD,  the
improvements   in  the  identification of  High
Emitters  are  less  obvious,  and seem  to   be
limited  to  decreasing  the  cutpoint,   with   a
near  exponential  increase  in   False  Failures
(see  Figures  6   and  7),  or  to  reducing
measurement  variability  (discussed next).

REPEATABILITY  ANALYSIS

   A  key  requirement  of  any  measurement
system  is  that  it  provide  consistent  results  to
the same stimuli.   In  the case  of RSD, Stephens
and  Cadle  [4],  and Lawson  et.al.  [5], have shown
that   under  relatively   normal  conditions,   the
RSD  can measure  with reasonable  accuracy,  CO
from a  specially  controlled  vehicle.   However,
other    elements   in    the    measurement-
consistency   equation   include   the   vehicle
characteristics,   its   operation,  and   weather
conditions.

   A previous  analysis of RSD data collected  in
Chicago   [6]  suggested  that  there  could   be
considerable  difference  between   RSD   reading
on the same car on  different days  Cat  about  the
same time  of day).   To  further  explore  this
area, several of these  other  elements  that could
cause  variability  in  the RSD   results,  and  the
effect  that these  factors  could   have   on   the
ability of  the  RSD to  accurately  identify  gross
emitters  were  analyzed.     The   track  testing
primarily   addressed   the  vehicle   and    its
operation.   The  analysis  of  the  on-road  data
addresses  vehicle  operation,  weather   effects,

-------
                        Figure 17
                  RSD Reproducibility
                 (Lmi RoW, M MPH, ATL TtH-Tnck)
 6.00
 4.00 • ••
 3.00 •
 0.00
-1.00


f FTPCO
[ 18.6 a/ml

+
•rtMMU.vwu.WAW.1v'
1
#738

"f FTPCO
I 3.20/lrt
•*
J 	 * 	 t
« \
4


#742



FTPCO Y
1.2 g/ml J


1
• 681





FTPCO
3.19/ml

{
L
«
T FTPCO 	
[ 20.5 8/ml
4

^ 4
) 	 ' 	


teao
                  TEST VEHICLE NUMBER
911672                                                                                          21

  and a  possible  combination of the  two on
  the  exhaust plume behind  a vehicle.

     TRACK  TESTING  -  The  track testing
  was conducted  over  a three day  period  at
  the  Bendix  test  track  in  New  Carlisle,   jf
  Indiana    by    Automotive    Testing   ~c
  Laboratories (ATL).   The track testing was   1
  the   first  opportunity   for    the   ATL   I
  personnel  to operate  the  RSD without  the   |
  assistance  of  the  developer.                 J  2.00 4.
                                               o
     The track   testing consisted  of driving   o  i.oo
  several  vehicles   by  the  RSD  multiple
  times  at the same  speed to obtain replicate
  RSD readings.   Several speeds  were used at
  each of two sites  -  one  a level  road,  and
  the  other  a 3  percent uphill grade.   A  total
  of 10  cars were tested, five  at each site.           percent CO  all the way  to 3% CO.   This large
                                                    variation  in RSD  readings  could  be  a  potential
     The  speeds  used   during   the  RSD  tests   problem  if a pass/fail  cutpoint were  to   be
  included 5,  10,  20,  30,  and 40 MPH.   The  data   selected that happened  to  be  in  the  upper pan
  were  spotty at many  of the  these  test  speeds   Of this  range.
  with numerous  low  voltage  errors  (i.e.,  no RSD
  readings).   The  lack  of  readings due  to  the      Conversely, vehicle  #680,  a  very high  FTP
  automatic  quality  control  feature  was  assumed   vehicle, had a similar  spread  in  RSD  readings.
  to be  partly due to  the  lack of  experience with   but  the   lower   range   of  #680's   readings
  this  equipment  by  the  operator.   As a  result,   overlapped   the  high  end  of #742's   readings.
  only at 20  MPH was sufficient data collected  to   Based  only on  this observation, it  would   be
  allow     reasonable     analysis    of    the   difficult to  discriminate  the dirty  car from  the
  reproducibility  of the  RSD  measurements.        ciean   car   under  these  operating   conditions
                                                    with an RSD cutpoint in the 2% to 4% CO range.
     Figures  17  and  18  were plotted  using  the   if  a typical  vehicle  showed the  kind  of RSD
  RSD  data  collected at  20  MPH and  the  as-   emissions  distribution  evident  in  this  example,
  received  FTP  from  each  vehicle.    Figure  17   the outcome  of  an  RSD   test could  be  based
  shows  the  repeatability of the RSD of tests done   considerably  on   chance,  or   be   a   strong
  on  the level  road,  and  Figure 18  shows those   function of specific  test  conditions,  instead  of
  from  the  3 percent  uphill  grade.   Both
  figures  provide  a visual  indication of  the
  variance among RSD scores on   the  same
  vehicle   at   a  constant   speed.     The
  accompanying  as-received  FTP   result
  provides    a   convenient   qualitative
  comparison  of  the  FTP  mass  emissions
  with RSD  concentration  emissions.

     Examination   of  Figure  17  shows
  extremely  high  variability  between  RSD
  measurements   made  under very  similar
  conditions  on  the  level  track.    Based  on
  these RSD results, it  is likely that for  some
  vehicles, the range  of RSD  results  will fall
  on both sides  of a  likely  cutpoint.   For
  example,  the   RSD  emissions  of vehicle
  #742,  which had a  low  FTP  score, varied
  from  a  low  reading  of  essentially  zero
                       Figure 18
                 RSD Reproducibility
                (UpWll R«W, 29 MPH, ATL Tot-Trad
6.00
                  TEST VEHICLE NUMBER

-------
22                                                                                           911672

being  based on  the vehicle's  FTP  emissions or      An  obvious  third  reason  is   that  different
level  of  repair.                                    cars  were  used  at  each site,  and the  observed
                                                   variability  was  simply a  matter of  coincidence.
   Examination  of  the  uphill  RSD  results in   This   possibility   is   somewhat  discounted
Figure  18,  show that they  contrast  very much   because  of   the  consistency  of  the   results
with   those  from  the  level  road  testing  in   within  each   road   site.     Furthermore,   the
Figure  17.    In terms  of  test  variability  the   difference  in  the False  Failure  rates  in  Figures
uphill  results   show  a  fairly   high  level  of   7(a)  and  7(c)  might also  imply  that  the   site
repeatability.   For  instance,  the individual  test   selection can  affect  the  variability of the RSD
results  usually  differed by  less than  0.5%  CO,   reading.   Figure  7(a)  included  both  level road
with  many  showing  much   better  repeatability   operation    and   uphill    operation   (with
than  that.   Whereas,  the  RSD readings from the   acceleration),   whereas   Figure  7(c)  included
level  site  frequently  varied  by  as  much 1.0 to   only  uphill  operation.   As  previously  indicated,
3.0%  CO,  none of uphill vehicles had  RSD scores   the  False Failure rate of IM240  High  Emitters
which  varied  to the  extent  that an individual   for  the  combined  site  (Figure  7a)  was  higher
vehicle's  pass/fail  status  would  likely  change   than  that for  the uphill  site.   It is possible  that
between  tests.                                     some  vehicles at the  two-lane site  exhibited  the
                                                   behavior  demonstrated  by  vehicle  #742,  and
   The  reasons   for  the  inconsistency  between   were  identified  as  High Emitters  by the RSD,
the uphill and  the  level  sites is not  completely   when  in fact  they  were not.   If  this  scenario
known.   One  possibility  is  that  since  the  level   truly occurred, it would  suggest that only sites
road  RSD testing was done  on  the  first day of   that  impose some  moderate  load on the  engine
testing  in this  study, the  lack of  repeatability   should  be  selected  for future testing.
may  have  been  due  to  the  operator's  lack of
experience  or   some   initial   non-recurring      ON-ROAD DATA - RSD  variability was  also
start-up  problem.   If this  were  the  case,  then   investigated  using  the   on-road vehicle  speed
these  results  would  not  be  representative  of   data  collected  at  the  expressway  ramp site.
normal   use.        However,    subsequent   Figure  19  shows  a  plot   of  the  RSD   CO
conversations  with  the   operators,  have  ruled-   concentration  emissions  versus vehicle  speed
out equipment  problems a possible  cause.   Lack   at  this  site.     No  speed  measurements  were
of  experience  could  still have  been  a factor,   recorded  at  the  two-lane site.
though.
                                                      The  vehicle  speeds  shown  in  this  figure
   A  second  reason  for  the  inconsistency may   were  measured  manually  by  the  RSD  operator
have  been  that  cars  can  vary  more  on  a  level   holding a radar  gun.  They  reflect  as closely  as
road.    This  could  easily  be  the  case,  since the   possible  a  vehicle's speed at  the  instant  of  the
load  on  the  engines  is  very light  under these   RSD  test.    The plot  shows  that  the  vehicle
conditions.    Any,   even  imperceptible  change,   speeds  ranged  from 5 to  33 MPH with a mean  of
in operating conditions during the  split second   23  MPH.  The lower speeds  (5 to  15  MPH) were
that  the  RSD  test  was  done could  be a  large   generally   recorded   at  the  first  site  on   the
percentage  of  the  required   load  for the  level   expressway ramp, and the higher  speeds  (20  to
road,   possibly  resulting  in   a large  change in   35  MPH) at  the second  site  on the expressway
CO concentration level.   If  this were truly the   ramp.
case,   then  the  results between  tests, even  on
the same car,  could be  very  different.   On the      The  scatter  and   the    linear   regression
other  hand,  the  tests  done  at the  uphill  site   results  in  Figure   19  indicate  no  relationship
placed  a  much heavier load on the  engine.   The   between  speed and  RSD CO  concentration.  The
higher  load  possibly  led to  less variability in   linear  regression  line is  almost horizontal   and
the  load  demanded  by  the  driver,  since  any   the  correlation coefficient is 0.0025.   However,
small  changes   in  load which  did  not  cause  a   the  lack  of   correlation  between   vehicle speed
noticeable  change   from  the   specified   test   and RSD results  is  not surprizing.   The more
speed  would  be  a  considerably  smaller  portion   critical  variables,  as  mentioned   above,   are
of  the higher  uphill  load.    The  smaller  load   likely  to  be  vehicle  acceleration   and  load.
variability  may  have   resulted   in  more   However,  because   of  problems   with   the
repeatable  RSD  data.                              interface  between  the   RSD  system  and   the

-------
911672
                                            23
                            Figure 19
                   On-Road Speed Effects
       ' "(jf»0.016X*0.4»  r2.0.0025
       0      5      10     15     20     25
                        VahlclaSpMd (MPH)

  radar gun,  acceleration  could not  be measured
  in this  study.
        interference   could  cause   erroneous
        results if the separation  time between
        successive  cars  was   short,   and  the
        difference  in emission  levels  of  the
        successive  plumes  was  large.    The
        potential    problem    of    plume
        interference   during   the    RSD
        measurements was  anticipated  by  its
        developers.    To  minimize  the  effects  of
        this   problem,    the   researchers
        developed  a criteria  to  detect  and
        eliminate  rapidly   changing   CO/C02
        ratios  when  two  exhaust plumes  mix
        together.  Under  such  test conditions,
        the  RSD discards  readings   when  the
        standard  deviation  of the  linear  fit  of
        CO and C02  exceeds 20  percent.
           Despite    these     precautions,
        Stephens and Cadle  [4]  found  potential
        RSD   measurement  problems   if   the
remnants  of   a  previous  vehicle's  exhaust
plume  were  present  during  the  current  RSD
measurement.      From   their   work   they
concluded  that   (1)  a
   Considering  these  results,  along  with those
from the test  track, vehicle speed (at least over                                identifies  plume
this  range)  does not  appear to  be  useful  as  a  ^'..     .*">"  lr*1  *•'"**»«   iusuui«.o  pium*
site  sellction  criteria.    As  suggested, other  "onlmeanty,    (3)  the  residual  plume  effect
criteria  are  probably  more useful  but  possibly  seemks to, last on  aT*gM  *T  ^ '"p"   ^
m(W  Mr*™* ,« Jl«,,~  «r  «htain              maybe  longer,  and    (4)  plume  nonlmeamy
                                                only  occurs   when  sequential  plumes  are  of
                                                greatly   different  concentrations.
   more difficult  to  measure or obtain.

      PLUME   INTERFERENCE   * A  potential
   problem  exists   where  the  residual   exhaust
   plume  from  a leading  vehicle  could  possibly
   create an interference in the RSD reading of  a
   closely   following  second   vehicle.     Such

                            Figure 20
                 Vehicle Separation Effect
               23456789
               Time Between Measurements (Seconds)
   In   this   study,   a   plume   interference
analysis,  similar to one  performed by  Stephens
and Cadle  was conducted using a  large sample
of  21,000  vehicles.    These  vehicles  were
         coincidentally  measured  by  the  RSD
         at the expressway  site,  in the process
         of measuring the  257 vehicles which
         received the both the IM240  test, and
         the RSD  test.  The  results were similar
         to those  obtained  by   Stephens  and
         Cadle.    For   example,  Figure  20
         compares  the  median  RSD  emissions
         stratified   into  three  levels  based  on
         the  previous  vehicle's  RSD  emissions.
         The  figure shows a definite  likelihood
         of measuring  a higher RSD  value  if
         the  time  separation  is  two  seconds  or
         less  and  the previous  RSD  CO  value
         was  more than  5  percent.   Based on
         the  results  in Figure 20,  the  RSD  CO
         emissions  of  a  typical vehicle will be
         about 0.5% CO  higher if  it is  following
         (within  one or  two seconds)  a  High
         RSD  Emitter (greater  than  5%),  than

-------
24
                                         911672
if  it  is  following  a  low  RSD High  (less than
1.0%).  For separation times of three  seconds or
more,  the  CO  medians show  less  relationship
with  the  previous  vehicle's  CO,  indicating the
previous  vehicle's  plume has  dispersed.

IN-USE  FLEET  ANALYSIS

   A second goal of  this project was to evaluate
the ability  of  the RSD  units to predict average
in-use emissions  in  a  specified  area.  This
evaluation  consisted   of   three   pans:     (1)
comparison of  estimated average RSD  CO  mass
emissions  with  IM240  mass   emissions,   (2)
comparison  of   the  median  RSD  and  IM240
emissions   collected  in  Indiana  with  RSD
emissions collected  in other states,  and  (3) use
of  the   Indiana  RSD  emissions   from  the
expressway  site  to simulate  the  collection  of
RSD  emissions  at a  number of independent
sites.
fuel  economy estimate.   Two  estimates for  an
individual   vehicle's   fuel   economy   were
available.    The   first  was  the  actual   fuel
economy  measured during the  IM240 test.   It
could be  viewed as the  best estimate available,
since  instantaneous   fuel   economies   at   the
precise  RSD  test  conditions   would  generally
not  be  available.  The second  was  an  average
by-model-year   and   by-manufacturer   fuel
economy estimate based on CAFE numbers  [4].

   ESTIMATED  MASS  EMISSIONS  •  RSD  mass
emission  estimates  in  grams  per  mile  were
calculated  using   both  of  the   above   fuel
economy estimation methods  so  that  they  could
be   compared   directly   with   IM240   mass
emission  results.   The  individual  model   year
results,  and  overall  average results  from  the
combined  sample  are  shown  in  Table  6  along
with the  IM240 mass  emissions.    Also, shown
for  comparison  are the  average by-model-year
RSD  and  I/M  concentration  measurements.
   All  three  pans  of this  analysis  required      A         .       ,   .            „„...
converting  the RSD  CO  concentrations  to  CO     .A  comparison  of  the  overall  RSD  mass
mass  emission estimates.   This  was  done  in  a  estimates at  the  bottom  of  Table  6 with  the
two  step process  following the method  outlined  IM24°  mass  Co  emissions  shows  reasonably
by  Stephens  and  Cadle   [4].   First,  the  CO  i°°d  agreement.    The  difference  between
concentrations  were  converted  to  grams  per  ^240 based  fuel economy RSD value of  15.51
gallon units  using molar  ratios from idealized  &™  1S   °,nlv 4  P?rcenVbel°w  'heu measu/ed
fuel   combustion  equations,  the  EPA  fuel  IM24°  valuc- ,  When  the  CA*E  basfd  fuuel
economy equation, and other  constants.   The  economy   values   were   substituted    the
second  step involved converting the  grams  per  difference  between  the  RSD  estimate  and  the
          *                   *P»      «^      »    T»jl1j^ft «*%A*AA«I*S4  *M  1 £.  •*&*«*•«*   A  1*lr&Ivf «A*t«n«
gallon  units  to grams  per  mile  units  using  a  *M24° in,creased  '°  16  Perce.m-   A  hke2v
                                                for the  larger  difference when  using the CAFE
                                                                  values  is that  the IM240 fuel
                              Table 6
                 Average Emissions by Model Year
                            (Combined Site)
                 IM240 Tai Remit*
Model
3m
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1976-82
1983*
AU.
Svnpto
2J2£
9
7
19
19
18
15
30
22
37
31
31
31
27
32
20
1
117
239
356
CO
lif/mimi
42.19
11.73
4032
2196
24.77
3326
21.17
27.99
1040
9.51
10.77
733
5.81
444
3.13
SJt
29.52
943
16.17
HC

la/mM\
2JZ
440
349
441
3J7
2.15
245
ISO
1.79
149
133
147
OJ9
049
041
Q.2S
3 JO
147
104
CO'
ttniD
4049
36.74
37.77
1925
26.79
1718
25.17
24j01
1634
532
1144
535
4J4
6J6
542
US
2744
947
15J1
CO"
ln/ml«>
3436
3136
3345
20J02
19.77
14J1
21J7
11.13
14.55
4.24
1O05
5.15
4.49
6.11
5.51
Ifi
24.13
137
13 JS
                                              CO
                                               1.71
                                               2.02
                                               1.90
                                               1.26
                                               1.40
                                               1.10
                                               1.66
                                               1J6
                                               1.14
                                               0.32
                                               0.81
                                               0.42
                                               0.39
                                               0.4*
                                               0.44
                                               fljfi

                                               1.54
                                               0.66
                                               0.95
                                                    Indi.ni I/M Rauta
                                                    Ida CO  gSOOCO
     2JS
     1.36
     1.61
     2.11
     0.43
     0.91
     0.75
     1.23
     OJO
     0.27
     0.77
     0.11
     0.16
     0.09
     0.11
     nm
     1.24
     0.36
     0.65
1.22
1.49
1.69
0.66
0.18
0.67
OJ1
0.13
0.12
0.13
0.00
0.4S
0.45
        during the IM240 cycta.
                                  gnrai»mte<»^
economy  is  based   on  the
actual   performance   of  the
in-use  vehicles,  whereas  the
CAFE is simply  a rough model
year    and    manufacturer
average  fuel  economy  based
on  new  vehicles  tested  over
the FTP.   In  any case, both
estimating   methods   resulted
in an average RSD  value  less
than the IM240 value

   In     examining     the
individual     model    year
results,    however,    the
relationship   between   the
estimated  RSD  emissions  and
the IM240 emissions  was not
as  consistent,  even  though
both   measurement   methods
showed     a     generally

-------
911672
                                                                                              25
decreasing pattern of  emissions  from older  to   emissions  pattern  for  the   older   vehicles.
newer  model  vehicles.   For  example, the RSD   Suggesting  that the  medians  are  similar,  and
estimates  for  the  1984  model  year with  37   that  the  same  outliers  and  small sample  sizes
ivnfn    v6re   siSmficantly   higher  than   the   effect the RSD and  the  IM240  medians equally.
1M24U,  while the  1985 RSD estimates  with 38   This is  a positive  result,  and  suggests a  good
cars were significantly lower  than  the  IM240   relationship  between  average RSD and  IM240
value.    This  dramatic  switch in RSD estimates,   results.   However,  total agreement between the
approximately  a  70  percent  decrease,  occurred   two  is not present,  and the magnitude of both
while the  IM240 values changed  only about 10   curves  suggests that the RSD  tends  to  under
percent  between  the   1984  and  1985  model   predict the  IM240  over  all model  years.
years.    Although this example  represents  the
most  extreme  case in  Table 6,  the  change  in      The RSD  data in Figure  21(a)  were  compared
RSD  estimates  from  year  to  year,  in  general,   to  the median  data  from the  previous analysis
seemed  to  vary  more  than  the IM240 average   [6]  in Figure  2l(b).    This previous  analysis
emissions.                                       included  test  data  collected  in  Chicago, Illinois
                                                  [1], Denver,  Colorado, and Ute  Pass Colorado  by
    Relative  to  two  methods of  computing  the   Dr.  Stedman.   Of the  three sites, the vehicle
RSD  estimates,  the similar results indicate  that   operation  and   test  site   topography  at  the
substituting  the  CAFE fuel  economy  estimates   Denver location was  probably  the most similar
seems  to  be a  reasonable approach.   This  is   to  the Hammond,  Indiana expressway  on-ramp.
fortunate,   since   actually   measured    fuel   The Ute  Pass location at 8000  feet with a steep
economy  values  for  individual  in-use  vehicles   uphill  grade  and  typically  strong  headwinds
would  probably   never  be  available  in  any   was  probably  the  least  similar.   The Chicago
                                                  express  way  ramp  was  marginally  similar  in
                                                  that  it likely  included  vehicle  acceleration,  but
                                                  was  a  straight entrance  ramp  instead  of  a
                                                  curvy one like  Hammond and  Denver.

                                                     Given  the  site  differences,  there   is   a
  application of  RSD to  on-road  measurements.

     OTHER  RSD  TEST  PROGRAMS  -  The CO
  concentration   results   from    other   RSD
  programs,  which   did   not   measure   mass
  emissions,  had  previously  been  analyzed  by
  the  author  [6]  relative to  their estimated  fleet   general similarity in the model year pattern  at
  mass  emissions.  To put these  other estimates in
  perspective,   they   were   compared   to   the
  Indiana  results.
     Since  the  other results  were  analyzed  in
  terms of the median values, the RSD and IM240
  median  values  from the express  way site  were
  computed,   and   plotted   in   Figure  21 (a).
  Excellent    directional   agreement   between
  medians  can be observed.   For  example,  both
  curves  seem  to  follow the  same inconsistent
                                                  all  sites  except  Ute  Pass.    The  difference
                                                  between  the  general  trend  and  a   specific
                                                  model  year  comparison  may  be   due  to  the
                                                  difference  in  sample  size.   The   Denver  and
                                                  Illinois  data  included  thousands  of  vehicles,
                                                  where   the   Hammond   data   included   only
                                                  hundreds  of  vehicles.     Nevertheless,   the
                                                  Hammond  data is  in the  same  ballpark  as the
                                                  Denver and  Chicago data.
                                 FIGURE 21
                     Median By-Model-Year CO Emissions
      7677787980ai8an84858687UM  78 77 ?• 7« 80 81 82 83 84 85 86 87 M M
                MODEL YEAR                       MODEL YEAR

                   (a)                               0)
                                                                      MULTIPLE      FLEET
                                                                   SIMULATION    -  A serious
                                                                   shortcoming  of previous  RSD
                                                                   testing  is  that  they  did  not
                                                                   include   any  mass   emission
                                                                   data   on  matched   vehicles.
                                                                   This   test  program  addressed
                                                                   that  need  on  an  individual
                                                                   vehicle    basis,   but    in
                                                                   evaluating  fleet  emissions,  it
                                                                   is  only  one  program   with
                                                                   only   one    fleet   average
                                                                   emission value  (see  Table  6).
                                                                   To  determine  how   well  the
                                                                   RSD  fleet  average in Table  6
                                                                   might    apply    to   on-road

-------
                                                                                           911672
 -casuremcnts,  many  programs  similar to  this
 ae would  need to be run.

   In  order  to obtain  a qualitative  assessment
 ithout   running  many   test   programs,  a
 omputer  program  was  developed  to  simulate
 le effect of  many sites.   Additionally, because
 le analysis in  this paper seems  to suggest that
 .SD  may  track broad  trends  in  fleet  averages
 jasonably  well,  but  does  not  do  so  well on
 idividual  vehicles,  it  would  be  useful  to
 imulate  many  sites  with  different  cars, but
 'ith   identical    test   conditions   for   each
 idividual  vehicle.

   The  correlated variables in  this  simulation
program   were  FTP/IM240  mass   emissions.
IM240/RSD  estimated   mass  emissions,  and
[M240/RSD vehicle  speed.   Only  expressway
data  was  used  for this  simulation.   It  was
•.xpected  that   the  FTP/IM240  simulation  would
 how  good  correlation  as  in Figure  2  earlier,
 nd that  the IM240/RSD speed  would  show  little
 orrelation.

   The  first  step  in  the FTP/IM240  simulation
 vas to randomly  partition  the  data  in Figure  2
 nto  two  equal sized  samples, arbitrarily  called
 .ample 'A' and sample  'B'.   Next, the average
 TP  and  IM240  CO  mass emissions  for  each
 sample group   were calculated.   The  difference
rctween the average FTP  emissions of sample  A
ind sample B  was then  determined.   Likewise
 he difference  in   the average  IM240 emissions
>f   sample  A   and B  was calculated.     This
calculation  process  created  a   unique   FTP
emissions  difference between  sample  A and B,
ind   a   unique   IM240  emissions   difference
jetween  the  samples.    This unique data   could
.hen be plotted  on an x - y plot as a point.
                                              The  process  of  randomly  partitioning  the
                                          data  in  Figure 2, and  calculating the  FTP  and
                                          1M240  emission  differences  was  repeated  1000
                                          times,  generating  the   1000  paired  FTP  and
                                          IM240 data points in  Figure  22(a).  In  this way,
                                          the  mix  of vehicles  in  each  sample  could  be
                                          different,   representing   different   fleets,  but
                                          each vehicle  would  be  tested  under  identical
                                          conditions.    Therefore,  only  the  effect   on
                                          different fleet  mixes  would  be  evaluated.

                                              An  analogous  procedure  was  repeated  with
                                          the  256 vehicles  tested  at  the expressway  site
                                          which received  both  the  RSD and  the  IM240
                                          test.    These  CO  differences  were  plotted  in
                                          Figure 22(b).   The same process  was then  used
                                          with  IM240 CO  data  and the RSD  speed data
                                          resulting  in Figure 22(c).

                                              The  best   agreement,  as expected,  of  the
                                          three comparisons  in  Figure  22  is  between  the
                                          FTP and the IM240.  This is noted by  the  high
                                          regression  correlation  coefficient,   and   the
                                          relative  absence  of data  points  in  the  upper
                                          left  quadrant  and the  lower  right  quadrant.
                                          Points  in  these  quadrants would  occur  if  the
                                          difference  between the FTP  means of sample A
                                          and  sample   B   were   negative,  and   the
                                          difference  between the IM240  means  of sample
                                          A and sample  B  were positive,  or  vice versa.   If
                                          this  had  occurred,   it  would  have  indicated
                                          there was  little  agreement  between  the  paired
                                          FTP  and   IM240  results  for that  particular
                                          simulation.

                                              In  contrast to  the  FTP/IM240  simulation is
                                          the  IM240 versus RSD  speed simulation.   As
                                          indicated,  this  simulation  was  only  conducted
                                          as  a  control  parameter  of poor  correlation,
                                          because  theoretically,  their  should  have  been
                                                Figun22
                                            Reel Simulations
                                 (Emlotoo or Speed Differences between Randomly £
[Y-OJ3X-OJ11 rlmOM  ]
                                         [Y.O.TIX-O.OM  12=055 3
[Y.O.OOSX*0.032 r2.0.0004
   •20
          .10  4  0  S  10
             FTP-CO(gftnl)
                                    •10  4   OS  10
                                      IM240-COO/ml)
    10  4   0  S  10
      IM240*CO(g/mi)

-------
911672                                                                                          27

  no  relationship  between  these variables  since   Emitters.    At  the  most  commonly  reported
  •A       came     from    non-overlapping,   cutpoint of 4.5  percent CO, the  RSD identified
  independent  tests.     The  generally  circular   less  than  15  percent  of the  late  model  High
  pattern  shown  in  Figure  22(c) shows that  this   Emitters,  which  accounted  for  only about  25
  was  the  case.     Note   that  the  RSD  speed   percent  of the  repairable  excess  emissions, and
  parameter  is  the  speed  difference  between   only  identified   about  20 percent of the  older
  sample  A  and  B,  not  the  average  speed in   model   High   Emitters.     Approximately  85
  61 c-f  ,?roup-                                     percent   of  the  late  model   High  Emitters
     Finally, comparison of the RSD  and IM240   accounting for  75  percent of  the  repairable
  mass   emission   simulation   showed   fairly   excess  emissions were  missed  by  the  RSD   in
  positive results,  although  not  as  positive as the   this  test program.
  FTP and IM240 results.   As in the  other figure,
  these  positive  results were  evidenced  by the     2.   Of  the vehicles  identified  by  the  RSD   as
  reasonable   correlation    and   the   relative   High CO  Emitters,  a moderate  portion  of  them
  absence of the points  in the two  quadrants.  In   were, in  fact,   not  High  Emitters   (i.e.,   False
  general,  these   results  generally  show  that if   Failures).   For  cutpoints  above  2  percent CO,
  the  RSD  and the  IM240  were tested at many   RSD  errors of  commission  ranged from  zero   to
  different    test   sites   under  IDENTICAL   as  high as 35  percent  (combined site)  for late
  conditions,  their   mean   emissions  would   model   year    cars,   and  from   zero   to
  typically be  similar to  one another.               approximately 25  percent for older cars.   These
                                                    ranges   of  False  Failure  rates,  however,  did
     As  a  final  point,  despite  the  good  overall   confirm  that  at least 65  to 75   percent  of the
  average   and   median   emission    agreement   vehicles identified as High  Emitters  by the  RSD
  between IM240 and  RSD,  care  must  be  taken as   truly  are High  Emitters.
  to  not  conclude   that   RSD  estimates   will
  consistently  represent  individual  vehicle  mass     3.   The  RSD was able  to operate in mildly
  emissions  over a complete driving cycle, or in-   inclement  weather   conditions  including   light
  use operation over  a  wide  range  of  conditions,   rain.    Only about  10  percent  of the CO  tests
  This fact can  be  seen in  the scatter  plots of   were  lost due to weather.   Of this reduced  total.
  Figure  4, the repeatability  plots of  Figures 16   about  another   10   percent  was lost  due  to
  and  17,  and  in  the  individual  model   year   improper  RSD   measurements  detected  by  the
  emission  results in   Table 6.    These  results   RSD's   internal  quality   control   algorithm.
  indicate   that  RSD  measurements   can  vary   Neglecting  equipment  malfunctions,  slightly
  considerably,  for   a  number  of   different   more than 15  percent of the  sampled fleet was
  reasons   which   probably   can   never   be   not  measured.
  completely  accounted  for  in  a one  second  test.
  Therefore,  the  utility  of  the  RSD   results in     4.   The  RSD  CO performance looked   more
  terms  of fleet predictions  is best if limited  to  a   like  a  traditional   I/M  test  than  a  loaded,
  specific consistent  test condition,  and  based on   transient,  mass  emission  test.    At  the   most
  a large RSD  sample.                               reported cutpoint of 4.5 percent  CO, the  RSD
                                                    was not nearly   as  effective in  identifying  High
  CONCLUSIONS                                   Emitters or  repairable  excess emissions as the
                                                    two-speed  idle   I/M test  for late  model   cars.
     The  primary  goal  of  this  test program  was   Lowering  the   RSD cutpoint   to   2   percent,
  to  evaluate  the  capability  of  the  RSD to   increased  the   performance  of  the   RSD   to  a
  properly  identify and  categorize High  Emitters   level  marginally  better  than   the  two-speed
  as  determined  by  the   IM240.    A  spin-off   test,  but  at  the expense  of higher errors  of
  analysis evaluated the  performance of the  RSD   commission (False Failures)  relative  to  the the
  relative  to  a  typical I/M  test  procedure, and the   I/M  test.   Although the  RSD  identified  fewer
  ability  of  the  RSD  to   identify  average  fleet   older  model  year  cars than  the  single  speed
  emission  levels  was  investigated.     Specific   idle  I/M test,  it  identified  slightly  more excess
  observations  and   conclusions  from   this  test   emissions   with  a  measurably  lower  False
  program are  as  follows.                           Failure  rate.

    1.  The RSD did   not,  even  at  the  most     5.   The RSD and the I/M CO  tests identified
  stringent cutpoints,  identify  all of the  High CO   different  population  groups.  The   RSD   only

-------
28                                                                                        911672

portion was  generally  smaller  for  the  newer  which  is largely  uncontrolled.    Because of the
model  group  than  for  the older  model  year  likely   uncontrolled  vehicle  operation,   RSD
group.                                          identification  improvements  might   only  be
                                                 achieved  by  lowering  cutpoints,  whereas  I/M
   6.   Combining  the  RSD CO test and  the  I/M  programs  can  lower  cutpoints,  add   program
CO  test  improved  identification  of  late  model  features  (e.g.,  functional  checks),  or change
High  Emitters,  but  at the expense of increased  test  type.
False Failures.   For older cars,  the combination
substantially    increased    High    Emitter     10.  The RSD  HC performance  was abysmally
identification   with  little  change   in  False  poor.  However,  the unit used  was  one of the
Failures  relative  to   the  I/M   test.     In  first  prototypes   developed,   and  future   units
combination  with  I/M,  reasonable  increases  in  would  be expected to improve.
identification  of   older   vehicles  were  even
observed at  the 4.5  percent  RSD  cutpoint with a     11.  Reasonable  agreement  seemed  to   exist
lower  False  Failure  rate  than  with  I/M alone.    between    the   IM240   overall   fleet-average
                                                 emission levels  and those  estimated  from  the
   7.   Selecting  an  RSD  cutpoint  below   2  RSD   measurements.    However,   substantial
percent CO  did  not  substantially  improve   the  variations   between   the   two   measurement
identification  of repairable excess CO emissions  methods    for  individual  model   years   was
from   the  late model  vehicles.    Selecting  a  observed.      Further  analysis   simulating
cutpoint above  4  percent  resulted in around  80  multiple  fleets,  tends to imply  that  there  is  a
percent of  all High  Emitting  vehicles  falsely  rough  relationship  between  the  RSD   fleet
passing the  RSD  test.    A  recommended  CO  estimates  and  the  IM240  when  identical  test
cutpoint range  for  future  test  programs  would  conditions  and  very  large samples are  used.
be  from  2%  to  3.5%  CO, depending  on   the
program  objectives,  and   the  acceptable  False     12.  The  RSD   plume  interference  analysis
Failure rate.                                     indicated that  there  is  a definite likelihood  of
                                                 measuring  a  higher  RSD  value  if  the   time
   8.   Repeatability of  the RSD CO  emission  separation  between  vehicles   is two seconds  or
measurements   on  track-tested  vehicles  was  less, and the  previous  RSD  CO  value  is  more
found  to  be  extremely  poor  under steady-state  than 5 percent.    However,   even if these  two
conditions  on  a  level  road.    However,  much  conditions  are  met,  the median  effect  is  less
improved    repeatability   in    RSD    CO  than 0.5% CO.
measurements  were   observed  on   vehicles
operating  on  a 3%  uphill  grade.   It is  thought     As  a final  note, the inability  of  the  RSD  to
that  the  reasons  for   these  variable   results  identify  even  a  significant  fraction   of  the
were  due to  individual  vehicle  differences,  and  High  Emitters   was  the  most   striking
its  precise   operating   characteristics  at  the  observation  of  this  study.   The  cause  for this
time   of   the  RSD   test,   rather  than  the  inability  to  identify  a  large  fraction  of  the
instrument   itself.     The  reasons  for  the  High  Emitters  is  probably more  a reflection  of
inconsistent   measurement  results   are  less  the  fact that  an  instantaneous  RSD  emission
important  than  the  fact  that  they  existed,  and  measurement  is  not  likely  to  be  completely
that  they  cannot  currently be  accounted   for  representative  of   an   overall   driving  cycle,
by  the RSD   system,  since the  current  system  rather   than  an   inability   of  the  RSD  to
measures    neither    vehicle   speed   nor  accurately   measure   the   instantaneous   CO
acceleration   accurately.    With  the  current  emissions  in an  exhaust plume.   Compounding
state  of RSD  development, it  is  recommended  this  problem is  the current  lack of ability  of
that  future   site  selections lean toward  those  the RSD system  to measure  the  exact  operating
sites  with  light  to  moderate   acceleration  or  mode of the vehicle  during the test.  Therefore,
uphill   loads.                                     not  only is the  RSD  system  not   capable  of
                                                 completely  measuring  the   emissions   from
   9.   Improving   the    CO   identification  overall driving, the  emission  results which are
performance   of  the   RSD  may  be  difficult  measured   are   diminished   in   usefulness
because  reducing  RSD   variability   and  because there is  no exact way  to  relate  them to
increasing    identification    performance  a  vehicle  driving mode.
appears to  be  influenced  by  vehicle  operation,

-------
911672                                                                                      29

     Even  so,  this may not be  a fatal handicap,   and Maintenance  Program",  Journal  of the Air
  since  the RSD  seems  to have the  ability  to   &  Waste  Management  Association,    August,
  identify   a  portion  of the High Emitters  in  a   1990.
  roadside   environment   quickly   and   non-
  obtrusively.   As such, this is probably the RSD     6.   October 9,  1990,  Memorandum from  E.
  system's  principle  asset,  and  may be  best used   Glover  to P.  Lorang,   "Analysis  of  Existing
  on  a  random  basis  in  conjunction with   a   Remote Sensing  Data"
  traditional   I/M  program.      For  example,
  identifying  and  repairing even a  portion   of     7.   "An  Evaluation  of Remote  Sensing for
  the  High Emitters in a CO 'hot spot' (outside  of   the   Measurement   of   Vehicle   Emissions",
  their periodic I/M  cycle) may  be  sufficient  to   Report   Prepared  for   the   California   Air
  lower the ambient levels in  the hot  spot area.    Resources  Board  by  Sierra  Research,   Report
                                                  No. SR90-08-02,  August, 1990.
  ACKNOWLEDGEMENTS
                                                    8.   Pidgeon,  W.,  and  Dobie,  N.  "IM240
     Recognition is given to Mr.  Joe  Patterson  of   Transient I/M Dynamometer Driving  Schedule
  Automotive   Test  Laboratories  (ATL) for  the   and the Composite I/M Test  Procedure",  Report
  many hours  he  spent alongside  the  expressway   No. EPA-AA-TSS-I/M-91-01.
  during  the  cold  winter  days  in  Hammond,
  Indiana.    Recognition is  also given to  Mr.     9.   June  20,   1991,  Memorandum  from N.
  Pradeep  Tripathi  of ATL   for  his  roadside   Brown  to  W.  Clemmens,    "Compilation of
  efforts,  and  especially  for  the hours he  spent   Remote Sensing Data  Collected  in Indiana".
  processing  RSD data and reading license plates
  from  video  tapes.    Special  acknowledgement     10.  Glover,  E., and Brzezinski, D.   "Exhaust
  also  goes to Dr. Donald  Stedman and Dr. Gary   Emission   Factors and  Inspection/Maintenance
  Bishop   for   all  their   valuable   advice  and   Benefits  for  Passenger Cars",   Report  No. EPA-
  support.   Finally,  thanks  goes  to  Mr.  Dennis   AA-TSS-I/M-89-2.
  McClement of ATL for his patience, and helpful
  suggestions  throughout this  process.               11.  Inspection/Maintenance     Program
                                                  Implementation Summary,   EPA Mobile  Source
  REFERENCES                                  Document,  July,  1990.

    1.  Stedman,  D.H., and Bishop,  G.A.,   "An     12.  EPA  Technical  Report,  "Recommended
  Analysis of On-road Remote  Sensing as  a Tool   I/M Short Test Procedures  For the  1990's:  Six
  for  Automobile  Emission   Control",  Report   Alternatives,  "   Report No.  EPA-AA-TSS-I/M-90-
  Prepared for the Illinois  Department of Energy   3.
  and  Natural Resources",   ILENR/RE-AQ-90/05.

    2.  Stedman, D.H.,  and  Bishop, G.A.,  "Remote
  Sensing   for  Mobile  Source  CO  Emission
  Reduction",   Final   Report  under   EPA
  Cooperative   Agreement,     CR-815778-01-0,
  August,   1990.

    3.  Bishop,  G.A.,   and   Stedman,   D.H.,
  "Oxygenated  Fuels,   A   Remote   Sensing
  Evaluation";   SAE Paper No.   891116.

    4.  Stephens.  R.D., and Cadle. S.H.,  "Remote
  Sensing  Measurements  of  Carbon  Monoxide
  Emissions   from  On-road   Vehicles",     GM
  Research Publication   GMR-7030, May, 1990.

    5.  Lawson, D.R., et al,   "Emissions from  In-
  use  Motor Vehicles  in Los  Angeles:   A Pilot
  Study  of Remote  Sensing   and the  Inspection

-------
              APPENDIX M
MODEL YEAR FAILURE RATES BY TEST TYPE

-------
               HC 2500:
               CO 2500:
                HC Idle:
         HAMMOND LANE I/M FAILURE RATES

220 ppm    220 ppm
         1.2 percent
           220 ppm
HC Idle:   220 ppm   220 ppm   220 ppm    250 ppm   350 ppm
CO Idle:  1.2 percent 1.2 percent 1.2  percent 2.0 percent 3.5 percent
IM240I
   HC:
   CO:
0.8
15
g/mi
g/mi
Model
Year
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
Lane
Sample
102
255
355
431
347
353
392
548
804
819
792
698
755
791
784
89
1

65
202
252
266
140
57
86
70
106
66
52
24
15
8
2
0
0

N.A.
N.A.
NA
N.A.
N.A.
66
§9 ;
76
J21:I€
74
56
32
18
Iffln
•• .H-;;:f:-;;
i^'f'$
mMM§

NA
N.A.
N.A.
N.A.
N.A.
98
141
105
156
105
85
39
22
12
8
0
0

57
187
226
230
102
50
67
53
86
56
39
19
12
6
2
0
0


'^M^M


78
37
47
35
49
35
24
13
4
3
1
0
0
Model
Year
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91

Lane
Sample
121
281
385
456
351
366
403
467
677
658
643
548
632
632
604
36


104
255
338
354
239
219
243
199
263
194
121
59
37
27
14
0

Note:  Shaded  area indicates official  Indiana  I/M test
                                                     Page 1

-------
                                            HAMMOND LANE I/M FAILURE RATES
               HC 2500:     —      220 ppm
               CO 2500:
                HC Idle:  220 ppm   220 ppm    22
                CO Idle: 1.2 percent 1.2  percent  1.2
                     220 ppm
                     .2 percent
220 ppm    220 ppm   220 ppm   250 ppm    350 ppm
                        percent 2.0 percent 3.5 percent
Model
Year
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
63.7%
79.2%
71.0%
61.7%
40.3%
16.1%
21.9%
12.8%
13.2%
8.1%
6.6%
3.4%
2.0%
1.0%
0.3%
0.0%
0.0%
N.A.
N.A.
N.A.
N.A.
N.A.
;^::i8:7%i-;
ISifB I?
;:H:i3.9%l..:.
IfSISS
|8$N|B
SSifitJ
4,6%
.2,4%
II;illlilI;i
:3m4%':
0,0%
0-0%
N.A.
N.A.
N.A.
N.A.
N.A.
27.8%
36.0%
19.2%
19.4%
12.8%
10.7%
5.6%
2.9%
1.5%
1.0%
0.0%
0.0%
55.9%
73.3%
63.7%
53.4%

14.2%
17.1%
9.7%
10.7%
6.8%
4.9%
2.7%
1.6%
0.8%
0.3%
0.0%
0.0%
•• : 27.5% ,
IBf&iSp
lltiliil
HS18i^
22.5%
10.5%
12.0%
6.4%
6.1%
4.3%
3.0%
1.9%
0.5%
0.4%
0.1%
0.0%
0.0%
IM240I
   HC:  0,8
   CO:  15
g/mi
g/mi
Model
Year
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
86.0%
90.7%
87.8%
77.6%
68.1%
59.8%
60.3%
42.6%
38.8%
29.5%
18.8%
10.8%
5.9%
4.3%
2.3%
0.0%
Note:  Shaded  area indicates official  Indiana  I/M test
                                                     Page 2

-------