A Study  01
       Mandatory  Engine  Maintenance
 for Reducing Vehicle  Exhaust Emissions
    Volume II. Mandatory Inspection/Maintenance
                     Systems Study
                     FINAL REPORT
                        July 1973
        In Support of:
APRAC Project Number CAPE-13-68

            for

 Coordinating Research Council, Inc.
     Thirty Rockefeller Plaza
   New York, New York  10020
    TRWk
 TRANSPORTATIONS,
'ENVIRONMENTAL
OPERATIONS
    OHE SPACE PARK • REDOKDO 8EACH CALIfORN'A 90?f8
                               and
                     Environmental Protection Agency
                       Air Pollution Control Office
                          5600 Fishers Lane
                       Rockville, Maryland 20852
SCOTT RESEARCH LABORATORIES, INC
P. O. BOX
SAN BERNARDINO. CALIFORNIA

-------
                     A Study of
       Mandatory  Engine Maintenance
  for  Reducing  Vehicle Exhaust  Emissions
     Volume II. Mandatory Inspection/Maintenance
                     Systems Study
                     FINAL REPORT
                        July 1973
        In Support of:
APRAC Project Number CAPE-13-68

            for

 Coordinating Research Council, Inc.
     Thirty Rockefeller Plaza
   New York, New York  10020
                              and
                     Environmental Protection Agency
                       Air Pollution Control Office
                         5600 Fishers Lane
                       Rockville, Maryland 20852
    TRW/
 TRANSPORTATION*
ENVIRONMENTAL
Off RATIONS
SCOTT RESEARCH LABORATORIES. INC
f. O. BOX 14I«
• AM •CNNAKDINO. CALIFORNIA

-------
                                 PREFACE

     This report, "A Study of Mandatory Engine Maintenance for Reducing
Vehicle Exhaust Emissions," consists of eight volumes.  The following are
the subtitles given for each volume:
          ®  Executive Summary, Volume I, Final Report, July 1973
          t  Mandatory Inspection/Maintenance Systems Study, Volume II,
             Final Report, July 1973
          e  A Documentation Handbook for the Economic Effectiveness
             Model, Volume III, Final Report, July 1972
          e  Experimental Characterization of Vehicle Emissions and
             Maintenance Studies, Volume IV, Year End Report, July 1972
          e  Experimental Characterization of Service Organization
             Maintenance Performance, Volume V, Final Report, July 1972
          «  A Comparison of Oxides of Nitrogen Measurements Made with
             Chemiluminescent and Non-Dispersive Radiation Analyzers,
             Volume VI, Final Report, July 1972
          •  A User's Manual and Guide to the Economic Effectiveness
             Computer Program, Volume VII, Final Report, July 1973
          t  Experimental Characterization of Vehicle Emissions and
             Maintenance States, Volume VIII, Final Report, July 1973
     The first volume summarizes the general objectives, approach and
 results of the study.  The second volume presents the results of the
 mandatory inspection/maintenance system study conducted with a computer-
 ized system model which is described in Volume III.   The experimental
 programs conducted to develop input data for the model are described in
 Volume IV (Interim Report of 1971-72 Test Effort) and V, VI, VIII.
 Volume VII is a user's manual for the computer code  and Volume VIII
 reports the experimental program and data obtained in the final test
 phase of the investigation.
     The work presented herein is the product of a joint effort by TRW
 Systems Group and its subcontractor, Scott Research  Laboratories.  TRW,
 as the prime contractor, was responsible for overall program management,
 experimental design, data management and analysis, and the economic
 effectiveness study.  Scott acquired and tested all  of the study vehicles,
 Scott also provided technical assistance in selecting emission test pro-
 cedures and in evaluating the test results.
                                     i

-------
            A STUDY OF MANDATORY ENGINE MAINTENANCE PROCEDURES
                  FOR REDUCING VEHICLE EXHAUST EMISSIONS

                             TABLE OF CONTENTS

                                                                      Page

1.0   INTRODUCTION AND SUMMARY	     1

2.0   SYSTEM STUDY APPROACH	     4

3.0   EVALUATION OF ALTERNATIVE PROGRAMS 	    12

      3.1  Mandatory Inspection and Maintenance Programs 	    12

           3.1.1  Comparison Between Year-End and Final  Results.  .  .    12

           3.1.2  Vehicle Exhaust Emissions	    19

           3.1.3  Optimal Engine Inspection and Mode Emissions
                  Inspection Criteria	    22

      3.2  Mandatory Maintenance Programs	    24

      3.3  Hybrid Inspection/Maintenance Programs	    27

4.0   SYSTEM SENSITIVITY ANALYSIS	    31

      4.1  The Impact of Vehicle Deterioration Rate on Program
           Performance	    31

      4.2  The Impact of the Inspection and Maintenance Process
           on Vehicle Rejection Rate	    32

      4.3  The Impact of Inspection Frequency on Program
           Performance	    38

5.0   STUDY CONCLUSIONS	    4!

      5.1  Inspection and Maintenance Procedures 	  ,  .    41

      5.2  System Sensitivity Analysis	    42

      5.3  Regional Implications of Vehicle Inspection and
           Maintenance .	    43
                                     ii

-------
                          LIST OF FIGURES
FIGURE                        TITLE                           PAGE

 3-1        Emission Time Histories for an Extensive
            Maintenance Program                                21

 3-2        Emission Time Histories for Mandatory
            Maintenance Programs                               26

 4-1        Influence of Vehicle Deterioration on Program
            Performance                                        33

 4-2        Mode Emission Rejection Histories                  34

 4-3        Vehicle Population Rejection Rates                 36

 4-4        Influence of Inspection Frequency on
            Procedure Effectiveness                            39

-------
                         LIST OF TABLES
TABLE                         TITLE                          PAGE


2-1      Inspection and Maintenance Procedures Studied         5

3-1      Comparison of Year-End and Final Results for
         Basic Inspection/Maintenance Procedures              13

3-2      Comparison of Optimal Results for Basic
         Inspection/Maintenance Procedures                    15

3-3      Average Abolute Mass Emission Reductions for
         Selected Optimal Procedures                          18

3-4      Vehicle Population Characteristics                   20

3-5      Optimal Pass/Fail  Criteria for Los Angeles           23

3-6      Comparison of Evaluated Programs                     30

4-1      Comparison of Program Performance for
         Various Vehicle Rejection Rates                      38

-------
                       1.0  INTRODUCTION AND SUMMARY

     This report summarizes a study of the economic effectiveness of
mandatory vehicle inspection and maintenance which was conducted as part
of the Emission Test Program.  The primary objectives of this study were
twofold:
     1)  To analyze the cost-effectiveness of mandatory vehicle inspection
         and maintenance as a means for reducing automobile exhaust
         emissions.
     2)  To evaluate the regional implications of various inspection
         and maintenance procedures and to identify the most attractive
         procedures for a given class of air quality problems.
The main thrust of the analysis was to determine the costs and emission
reductions achieved using a number of candidate procedures.  The informa-
tion presented herein reflects the latest available experimental data,
including the results from the emission deterioration study, as well as
the most recent refinements to the economic effectiveness model.
     The study was conducted using a model which is capable of evaluating
the performance characteristics and costs associated with a wide range of
inspection and maintenance procedures.  To support the model, vehicular
test data were obtained which defined the general maintenance state of
vehicle populations, the sensitivity of exhaust emissions to engine mal-
functions, the ability of service organizations to perform specific engine
maintenance, and the rate with which exhaust emissions and engine adjust-
ments deteriorate with vehicle use.
     In keeping with the growing interest in vehicle inspection and mainte-
nance a number of different procedures were examined in some detail.
These included:
          •  Mandatory inspection and maintenance
                  Engine inspection
                  Emission inspection
          •  Mandatory maintenance only
          •  A hybrid inspection and maintenance procedure which involves
             a combination of the two previous processes

-------
This evaluation provides a good comparison of these strategies and should

be of direct interest to those states and regions that have proposed or

are currently contemplating a program of inspection and maintenance.

Within, each of these alternatives, however, the basic inspection procedures

remain the same, i.e.:

          •  Direct evaluation of engine maladjustments and malfunctions
             using conventional or more sophisticated garage-type
             equipment.

          •  Indirect diagnosis of engine maladjustments and malfunctions
             using measurements of exhaust emission levels at different
             engine loading conditions

     The emission reductions predicted during this study are based on the

1972 Federal Test Procedure and are in general agreement with those reported

in the year-end report (July 1972).  The main differences in predicted

emission reductions reflect the lower rate of engine misfire and slightly

higher rate of idle fuel to air ratio enrichment with mileage determined •

in the deterioration experiment.  The net result is the prediction of

slightly lower HC reductions (7% vs 10%) and a slightly higher CO reduc-
tion (9% vs 6%) than previously reported.

     Subject to limitations in the model and supporting data base, which

are described in subsequent volumes of this report, the following specific

conclusions were drawn:

     1)  For regions characterized by a photochemical smog problem, the
         most cost-effective procedure involves a state operated emission
         inspection made with the engine operating under load followed,.
         as necessary, by an idle adjustment and ignition tune-up.

     2)  The most comprehensive photochemical smog oriented inspection
         and maintenance procedure studied yielded HC, CO and NO
         reductions of 10, 6 and -1 percent, respectively.

     3)  For regions characterized by a carbon monoxide problem, the
         most cost effective procedure involves a franchised garage idle
         inspection and adjustment program along with the mandatory
         replacement of air cleaner and PCV valve.

     4)  The most comprehensive CO oriented inspection and maintenance
         procedure studied yielded HC, CO and NO  reductions of 6, 11 and
         -2 percent, respectively.

-------
     5)  The most cost effective inspection interval  is yearly regardless
         of the inspection and maintenance procedure selected.
     6)  Vehicle rejection rates on the order of 40-60 percent are
         required to achieve reasonable emission reductions.
     7)  The figure of merit (discounted total  program cost divided
         by the weighted five year reductions of emissions) of inspection
         and maintenance procedures is sensitive to the following factors
         (given in rank order of importance):
         •  Regional characteristics of air pollution problem
         t  Garage maintenance effectiveness
         •  Rate and extent of emission and engine adjustment deterioration
         •  Vehicle population characteristics
         •  Extent of voluntary maintenance
         •  Frequency of inspection and repair
         •  Costs of unnecessary repair
         •  Inspection accuracy (errors of omission and commission)
     The results of this study indicate the necessity of evaluating the
impact of new emission control technology on the performance of inspection
and maintenance.  The present data base contains emission and maintenance
cost data for vehicles through the 1971 model year.  For this study, it
was assumed that all post-1971 vehicles had the same basic characteristics
as the 1971 vehicles.  New data should be acquired and inspection and
maintenance procedures reevaluated as major changes in emission control
technology are introduced into automobiles.
     As noted above, one objective of this report is to provide general
guidance to those states and regions contemplating inspection and mainte-
nance programs as part of their air quality implementation planning.  Pilot
studies conducted on a small scale should be performed to verify that the
estimated performance of a selected inspection and maintenance procedure
can be achieved before large scale program implementation.

-------
                         2.0  SYSTEM STUDY APPROACH

     This section presents a discussion of the inspection and maintenance
procedures examined in this study and a description of the methodology
used to assess their feasibility as a means for reducing automobile exhaust
emissions.
     The revised economic-effectiveness model  and the recently augmented
data base was used to evaluate the performance of selected inspection and
maintenance procedures.  Since each basic approach to vehicle inspection
and maintenance is subject to a large number of possible variations, speci-
fic candidate procedures were selected in advance of the actual analysis.
The results of previous studies were used to guide this selection.   Table
2-1 is a summary of the inspection and maintenance alternatives examined.
     These alternatives reflect the importance of inspecting and maintain-
ing three categories of engine maladjustments  and malfunctions:
     1)  Idle adjustments (idle speed, idle fuel  to air ratio and basic
         timing)
     2)  Ignition components affecting misfire or spark advance (NO  control)
                                                                   A
     3)  Induction system components affecting fuel to air ratio.
The idle adjustments involve labor only, are economical to repair and can
be selectively manipulated to influence all three exhaust emission species.
Ignition related components primarily influence HC emissions, although NO
                                                                         A
emissions are also strongly affected by spark timing.  The cost of repairing
these items is distributed about equally between labor and parts.  Induction
system component failures (PCV valve, air cleaner, air pump and choke system)
primarily affect CO emissions and to a lesser degree HC and NO  emissions.
                                                              A
Repair costs for these items tend to be largely determined by the cost of
replacement parts, although labor costs for air pump and choke maintenance
can be appreciable.
     Clearly, the maintenance of any combination of these three categories
of malfunctions may be desirable depending on regional air quality require-
ments.  In order to evaluate inspection and maintenance alternatives system-
atically these three categories of malfunctions have been used to define

-------
       TABLE 2-1   INSPECTION AND MAINTENANCE PROCEDURES STUDIED
•  Mandatory Inspection and Maintenance

   Engine Inspection

   1)  Idle (Franchised Garage)

   2)  Extensive A (Franchised
       Garage with Dynamometer)

   3)  Extensive B (Franchised
       Garage with Dynamometer)



   Emission Inspection

   1)  Idle (State Lane)

   2)  Extensive A (State Lane
       Emission Under Load)

   3)  Extensive B (State Lane
       Emission Under Load)


•  Mandatory Maintenance

   1)  Recommended Manufacturer's
       Maintenance
       (Franchised Garage)

t  Hybrid Inspection and Maintenance

   1)  Idle Inspection with Mandatory
       Replacement
       (Franchised Garage)
   2)  Random Idle Inspection
       (State Lane)
Inspection and Maintenance Procedure

Inspect and adjust ICO,  idle rpm,
and timing

Inspect and adjust ICO,  idle rpm,
timing, repair misfire and NO
control

Inspect and adjust ICO,  idle rpm,
timing, repair misfire,  air pump,
PCV valve, air cleaner,  NO  control
and choke system
                  adjust ICO,  idle

                  LHC; adjust  ICO,
                  repair misfire
Measure ICO, IHC;
rpm and timing
Measure ICO, IHC,
idle rpm, timing.
and NO  control
Measure ICO, IHC,  LHC, LCO; adjust
ICO, idle rpm, timing, repair mis-
fire, PCV valve and air cleaner
Undertake specific maintenance
treatment as prescribed by manu-
facturer
Inspect and adjust ICO, idle rpm,
timing, replace air cleaner and
PCV value
Inspect subset of vehicle popula-
tion using idle emission test and
perform required maintenance
ICO - Idle CO emission measurement
IHC - Idle HC emission measurement
LCO - Loaded CO emission measurement
LHC - Loaded HC emission measurement

-------
the following three maintenance options:  1) idle adjustments which are
made primarily to reduce CO emissions, 2) extensive "A" maintenance con-
sisting of idle adjustments and the repair of the ignition component fail-
ures which is performed to reduce both HC and NO  emissions, and 3) exten-
                                                A
sive "B" maintenance which adds the repair of induction related malfunctions
to the extensive "A" maintenance treatment for further CO control.
     The engine component malfunctions within these maintenance options
can be diagnosed using conventional, commercially available equipment and
procedures or inferred from exhaust mode emission measurements.  Emission
measurements must be made in appropriate engine operating modes when the
latter approach is used.  Selected emission inspection modes are summarized
in Table 2-1.  The selection of these emission inspection modes is  reported
in Volume IV.
     It was generally found that multiple emission modes did not identify
significantly more engine malfunctions than could be detected with  single
emission modes.  Exceptional cases were the diagnosis of air cleaner and
PCV valve malfunctions where vehicle rejections based upon high emission in
two engine operating modes, i.e., CO and HC under load, yielded slightly
improved failure diagnosis (5-10 percent).  To simplify both the computer
model and the implementation of inspection procedures, only three emission
inspection modes which identified a large number of component failures were
selected for study.
     The other two basic approaches—mandatory maintenance only and hybrid
inspection and maintenance—represent the two most logical alternatives to
the basic procedures that have been previously studied in depth. A program
of mandatory maintenance simply requires that all vehicles undergo  periodic
engine repair based on procedures set forth by the vehicle manufacturer.
This approach has the advantage of eliminating the need for inspection
(except for tests required to diagnose specific engine malfunctions).  Its
chief disadvantage is that it tends to be more expensive and that the adjust-
ment of all vehicles to manufacturer's specification may yield only marginal
emission reduction results (as opposed to repairing only those vehicles with
maladjustments that result in directly lowering vehicle emissions).

-------
     A hybrid program of vehicle inspection and maintenance attempts to
combine the key features of a mandatory inspection and maintenance program
with those of a mandatory engine maintenance program.  The most practical,
and perhaps most cost-effective program, involves an idle emission inspection
and tune-up followed by the mandatory replacement of two key emission-
oriented engine components--PCV valve and air cleaner.  (This procedure
is especially attractive for areas characterized with a CO problem.)
Another program which offers some promise involves a random idle inspection
followed by an extensive maintenance program for those vehicles failing the
test.  This approach is primarily oriented toward that subset of the vehicle
population that is poorly maintained and/or contains inherently high emitters,
Both basic alternatives will be examined in some detail in Section 3.0
     This study was performed in three parts as described in the following
sections:
          •  An evaluation of the inspection and maintenance alternatives
             described in Table 2-1 with regard to the specific components
             within each subsystem to be maintained and their associated
             inspection pass/fail criteria.
          •  An assessment of the basic study ground rules and assumptions
             to determine their impact on procedure effectiveness.
          •  A preliminary analysis of the regional implications of vehicle
             inspection and maintenance.
     A five-year time period was studied to determine if predicted emission
levels stabilized at the end of that period.  Projections of the performance
of mandatory vehicle inspection and maintenance programs were limited to no
more than five years because of:
          •  Uncertainty in estimating the effects of major engine mal-
             functions which may occur beyond that period
          •  Lack of experimental emission time history data beyond 50,000
             miles (i.e., approximately five years)
          •  Inability to project the impact of new control technology
             and other emission control programs (e.g., vehicle retrofit
             programs)
     The economic effectiveness model has been updated to include the
following design improvements and refinements:

-------
     1)   Characterization of the  vehicle population by three emission
         control  classes:   uncontrolled vehicles  (pre-1966), controlled
         vehicles (1966-1970)  and post-1970  vehicles (NO  control).
                                                        A

     2)   Incorporation  of a vehicle  attrition  algorithm for estimating
         changes  in  vehicle population  distributions over  time  for the
         three emission control classifications.   The algorithm presently
         characterizes  all  new vehicles entering  the population by the
         post-1970 emissions data.

     3)   Introduction of the influence  coefficients for adjustment inter-
         actions  and the addition of three new engine parameters (NOX
         control  device, choke system)  bring the  total  parameter set  to
         ten.   The influence coefficients for  interactions  relate changes
         in emission species to the  adjustment and/or repair of two engine
         components.

     4)   The use  of  tabular engine adjustment  and mode emission distribu-
         tions in lieu  of the analytic  descriptions used previously.
         This  provides  for a more accurate,  although more  complex, treat-
         ment  of  these  basic data.

     5)   Incorporation  of measured garage performance into  the  maintenance
         effectiveness  analysis.

     6)   Addition of franchised garage  inspection costs, the cost of un-
         necessary maintenance, and  state lane inspection  training and
         processing  costs into the economic  analysis.

     7)   Statistical analysis of  emission reduction estimates using  both
         empirical as well  as analytical techniques.

     8)   Determination  of optimal pass/fail  inspection criteria for both
         engine components  and mode  emissions.

     9)   The deterioration of engine adjustments  and exhaust emissions
         with  vehicle usage.

     In  addition  to  these refinements,  the economic effectiveness model
now includes the  following data sets for all three controlled fleets:

     1)   Mass  and key mode primary and  interaction influences coefficients
         (orthogonal experiments).

     2)   Emission and engine deterioration rates  (deterioration experiments),

     3)   Engine adjustment and mode  emission distributions  (Los Angeles and
         Detroit  survey).

     4)   Exhaust emission levels  for a  random  sample of vehicles (Test 1A,
         deterioration  experiment).
                                     8

-------
     5)  Franchisee! garage maintenance effectiveness (garage performance
         experiment).

     6)  Revised operating and capital cost data (garage performance
         experiment).

     7)  Engine component failure rates and mean values of adjustments for
         the failed and passed vehicle populations when various mode
         emission inspection criteria are applied (cutpoint analysis of
         deterioration experiment data).

     These model refinements and data additions, when taken collectively,
constitute the 1973 Revised Economic Effectiveness Model.  Before examin-
ing results derived using this model, it might be appropriate to review

several of the basic ground rules used in constructing the model:

         •  All mass and mode emission levels and emission reduction per-
            centages are reported as average values for the entire
            vehicle population.

         t  Pass/fail criteria for each inspection and maintenance strategy
            are optimized for the first inspection interval (year zero)
            and remain invariant with time.

         t  Maintenance of engine malfunctions and maladjustments which is
            not included in the mandatory maintenance program is assumed to
            be performed voluntarily by the owner.

         •  Estimated mean emission reductions are based on measurements
            made with the 1972 Federal Test Procedure.

         t  Engine deterioration rates, influence coefficients and the
            extent of voluntary maintenance is assumed to be the same
            throughout the country.

         •  The criteria used in passing or failing vehicles during
            inspection involves the state of engine adjustments and/or
            mode emission levels and not composite mass emission levels.

     The basic performance measure used in ranking the various procedures
is the program figure of merit.  It is calculated from aggregated values

of emission reductions and system cost as shown in Equation 3-1.


(3_1)        FOM  =   annularized, discounted total program cost
                             
                     i=l

-------
     where:
         FOM,  figure of merit for a  specific procedure,  $/ton
         W-,   normalized weighting factors  for each emission species
         E .,  integrated emission level  for each  emission species for a
              baseline (voluntary maintenance) population
         E .,  integrated emission level  for each  emission species from a
              population subjected to the specified mandatory inspection
              and maintenance procedure

     Weighting factors selected for  the  Los Angeles Basin were:
         WHC   -  0.6

         WCO   =  °'1
         W    -  0.3
            A
     As can be seen, the weighting factors  establish the degree to which
reduction in  emission components impact  the program design.   Having fixed
the weighting  of emission reductions, the model  can then determine the
optimal pass/fail criteria and system configuration.  In actuality, the
weighting of  emission reductions should  reflect the specific air pollution
problems of a  given region, i.e., if the concern  is with CO, then CO should
be assigned a  larger weighting factor than  HC.
     The figure of merit is computed using  several  different measures of
emission reduction which depend primarily on the  time period of interest.
The following  three figures of merit are ones most  often reported and used
in the analysis:
         •  The integrated average emission reduction relative to the
            baseline population (control group) achieved over a five year
            period.
         •  Emission reduction achieved  relative to the baseline population
            at the end of the fifth  year.
         •  Statistically significant emission reduction achieved relative
            to the baseline population at the end of the fifth year.
     The utility of these three figures  of  merit is that they permit the
ranking of the candidate procedures  based on either average performance
                                     10

-------
results, or the results obtained at the end of the program.   One or more
of these criteria may be employed, depending on the particular characteris-
tics of the air pollution problem encountered.
     The reader is encouraged to review Volume III for a more detailed
discussion of the basic modeling approach and the actual and implied assump-
tions which are a basic part of the Economic Effectiveness Model.
                                     11

-------
                   3.0  EVALUATION OF ALTERNATIVE PROGRAMS

     The revised computer model  was first used to evaluate the economic
effectiveness of the basic inspection and maintenance procedures previously
described.   The Los Angeles basin was selected for this evaluation because:
1) the experimental data used in the model  were developed in the Los Angeles
area, 2) the region suffers from chronic air pollution in the form of photo-
chemical smog to which the automobile contributes.

3.1  MANDATORY INSPECTION AND MAINTENANCE PROGRAMS
3.1.1  Comparison Between Year-End and Final Report Result^
     Table 3-1 presents a comparison between the results derived from this
analysis and those presented in  the year-end report.  In this comparison
the pass/fail criterion, inspection system configuration, inspection frequency,
etc. were kept the same to explore the impact resulting from differences in
the data base due to introducing information from the deterioration experi-
ment as well as from differences in the way data are handled in the economic
effectiveness model.  A cursory  examination of Table 3-1 reveals that the
program figure of merit for the  year-end analysis tends to be more attrac-
tive, i.e., lower than for the final analysis (with the exception of Case
4).  This results from the substantial reduction in the rate of ignition
misfire used in the final evaluation as well as the relative higher cost
required to properly diagnose and repair vehicles with low misfire rates.
This situation is especially true for the engine parameter inspection.
Comparing the results for Case 2 shows over a 4 percent decline (10.8 vs
6.5) in hydrocarbon reduction when the new data are used.  One encouraging
note, however, is the improvement in CO emission reductions projected in
the final results.  This can be  traced to the higher CO emission deteriora-
tion rates observed across all three control classes of vehicles (primarily
due to greater deterioration in  idle fuel to air ratio adjustments).
     It should be noted that the analytical methods used in developing the
emission distributions has been  improved thereby tending to confound an
interpretation of these results.
     The results of the year-end and final  analyses appear more consistent
for the case of the exhaust emission inspection.  This can be attributed
                                     12

-------
                       TABLE 3-1   COMPARISON OF YEAR-END  AND  FINAL RESULTS
                                   FOR BASIC INSPECTION/MAINTENANCE PROCEDURES
INSPECTION/MAINTENANCE
PROCEDURES
ENGINE PARAMETER INSPECTION
1. IDLE (FRANCHISED)
.2. EXTENSIVE A+ (FRANCHISED)
3. EXTENSIVE B++ (FRANCHISED)
EXHAUST EMISSION INSPECTION
4. IDLE (STATE LANE)
5. EXTENSIVE A+ (STATE LANE)
6. EXTENSIVE B++ (STATE LANE)
FIGURE OF MERIT
$/TONH-H-
Y F

950 1005
1780 3125
1975 2935

730 585
1305 1405
1475 1665
COST PER
VEHICLE
Y F

4.00 4.40
11.30 17.80
24.80 29.30

3.20 3.35
10.10 9.70
12.35 13.45
EMISSION REDUCTIONS, PERCENT j
HC CO NOX |
Y F

2.4 3.5
10.8 6.5
11.6 8.1

1.5 2.6
7.9 5.6
8.1 6.0
Y F

2.6 5.0
2.6 5.0
5.1 8.7

2.9 6.5
2.9 6.5
3.4 8.0
Y F
t
-0.2 -0.1
-0.2 -0.1
-1.1 -2.0

0. 0.
0. 0.
-0.3 -0.5
  + IDLE PLUS IGNITION TUNEUP, WITH DYNAMOMETER
  -+ IDLE PLUS IGNITION PLUS INDUCTION TUNEUP WITH DYNAMOMETER
    L. A.  BASIN WEIGHTING FUNCTION
Y - YEAR END RESULTS (JULY  1972)
F - FINAL RESULTS  (JULY 1973)

-------
primarily to the less severe pass/fail  criteria which results with an
emission inspection.   (The rejection rates for an engine adjustment inspec-
tion are above 70 percent while for an  emission inspection they are closer
to 50 percent.)  Reducing the rejection fraction (i.e., failing fewer cars)
tends to minimize the impact of other system variables and consequently the
results of the two analyses tend to be  similar.
     A sample set of results from the final  analysis is given in Table 3-2.
The engine inspection strategy is seen  to be less cost effective across
all maintenance treatments than is the  emission inspection procedure.  This
occurs even though the engine inspection approach normally produces larger
emission reductions.
     Because the magnitude and sign of the influence coefficients varied
between the three classes of vehicles studied and because the costs of some
repairs were larger than others, a slightly different approach was used in
this study to define optimal procedures.  In previous analyses, all engine
adjustments included in a particular maintenance procedure were inspected
in all vehicles and were maintained on  those cars which did not satisfy the
specified pass/fail criteria.  While this approach is conceptually correct,
it does not necessarily lead to the most cost effective program.  For example,
the incorporation of choke heat riser valve repair has a relative small effect
on emission reduction (due, in part,to  the low reliability of maintenance)
while the costs of its inspection and maintenance are relatively high.
Greater emission reduction can be achieved at lower cost by merely making
the pass/fail criteria for air cleaner more restrictive.  This approach
eliminates the inspection and maintenance of some engine components and
adjustments for each of the fleets in order to achieve a program which is
optimal in a cost effective sense.  This new approach has not, however,
required modification of the basic structure of the economic effectiveness
model.  Consequently, the implications  of including all engine adjustments
in the analysis can still be studied easily.
     The most attractive procedure (i.e., the one having the lowest average
figure of merit of 575) is an idle CO emission inspection (#4) followed by
the adjustment of the carburetor idle fuel to air ratio.  Timing and idle
rpm were also to be adjusted on the 1966-70 emission controlled vehicles,
                                     14

-------
TABLE 3-2 COMPARISON OF OPTIMAL RESULTS FOR BASIC
          INSPECTION/MAINTENANCE PROCEDURES
INSPECTION/MAINTENANCE
PROCEDURES
ENGINE PARAMETER INSPECTION
1. IDLE (FRANCHISED)
2. EXTENSIVE A+ (FRANCHISED)
3. EXTENSIVE B++ (FRANCHISED)
EXHAUST EMISSION INSPECTION
4. IDLE (STATE LANE)
5. EXTENSIVE A+ (STATE LANE)
6. EXTENSIVE B++ (STATE LANE)
FIGURE OF MERIT
$/TON+++
ABC

785 320 540
2335 975 1585
2865 1165 1515

575 235 335
1340 558 915
1590 650 945
COST PER
VEHICLE
D E

4.55 5.20
16.85 23.10
32.10 34.95

3.20 4.10
8.90 16.75
11.60 20.70
EMISSION REDUCTIONS, PERCENT
HC

3.3
6.3
7.8

1.7
4.7
4.9
CO

6.1
6.1
10.4

6.8
6.8
7.3
NO

-0.1
-0.1
-1.6

-0.2
-0.2
-0.5
IDLE PLUS IGNITION TUNEUP, WITH DYNAMOMETER                 A.
IDLE PLUS IGNITION PLUS INDUCTION TUNEUP WITH DYNAMOMETER   B.
L.A. BASIN WEIGHTING FUNCTION                               C.
                                                            D.
                                                            E.
                                        Computed  over  five year period.
                                        Computed  at  end of last interval.
                                        Based  on  statistically significant
                                        emissions  reductions for last  interval.
                                        Based  on  total vehicle population.
                                        Based  on  maintained portion of vehicle
                                        population.

-------
but were found not to be cost-effective to maintain on the other two vehicle
classes.  This results from the fact that retarding basic timing substan-
tially increases CO emissions on precontrolled and post-1970 vehicles, thus
cancelling the beneficial  effects on the figure of merit of reduced NO  and
                                                                      /\
HC emissions.  (The influence coefficient for CO with respect to timing is
nearly five times as large for the post-1970 vehicles as for the 1966-70
emission controlled vehicles.)  It should be noted that this procedure does
not greatly reduce exhaust emissions and, therefore, may be of little prac-
tical interest.
     The next two most cost-effective procedures, an inspection of the
idle adjustments followed by their maintenance (#1) and an emission inspec-
tion followed by extensive "A" maintenance (#5), have figures of merit of
785 and 1340, respectively.  Because of the favorable figure of merit and
the relative large HC emission reductions achieved (5 percent average and
10 percent in the fifth year), the latter procedure would be attractive for
areas with severe photochemical smog.
     The smaller incremental emission reductions effected with the emission
inspection procedure followed by the more extensive "B" maintenance, result
because of the relatively poor ability of this inspection procedure to
detect component failures which affect induction system performance.  Indeed,
virtually all of the procedures investigated achieved only modest CO reduc-
tions and had no influence on NO .  The largest CO reduction is achieved
                                /\
with a direct engine inspection procedure and "B" maintenance.  Average and
fifth year CO reductions are 10.4 and 15.2, respectively.  These results
represent an improvement over those reported in the year-end report.  Again
this can be attributed to the higher CO emission deterioration rates deter-
mined from the emission deterioration experiment.
     Another point of interest is that all procedures presented in Table
3-2 yielded statistically significant emission reductions.  This is in
contrast to the case for the year-end report where neigher idle adjustment
program provided statistically significant CO reductions.  This improvement
in procedure effectiveness can be directly related to the increase in CO
emission deterioration.
                                     16

-------
     Two cost figures are presented in Table 3-2, one based upon prorating
the cost of inspection and maintenance over the total vehicle population
and the other upon distributing inspection costs over the entire vehicle
population while prorating maintenance costs to the rejected portion of the
population only.  The overall higher program costs can be attributed to the
following:
         •  Costs of unnecessary repair
         •  Increased information processing costs
         •  Training costs
These three cost elements when taken collectively tend to exceed the costs
for performing the basic inspection and maintenance procedure.  However,
these cost projections appear consistent with estimates made on a number
of ongoing experimental and operational programs.
     Absolute values for emission reductions instead of percent emission
reductions can also be used as the criteria for evaluating the performance
of alternate inspection and maintenance procedures.  Table 3-3 presents
the emission reductions in tons per day predicted by the computer model for
the six basic inspection and maintenance procedures.   These are the average
emission reductions achieved over a five year program time span during which
the automobile population continues to grow.
     As would be expected, the largest absolute value of emission reduction
is achieved for CO.  It is noted on the other hand that CO percentage reduc-
tions are nearly comparable with the HC percent emission reductions.  Part
of this is due to the larger numerical value for CO emissions.  If absolute
emission reductions are used with the Los Angeles basin weighting factors,
then the weighted HC, CO and NO  emission reductions are 51.6, 139 and 3.9,
                               X
respectively for an engine inspection and an extensive "B" maintenance treat-
ment.   Even when the Los Angeles weighting factors are applied, the weighted
CO emission reduction is considerably larger than that of the other weighted
species and as a consequence will  strongly influence procedure selection.
Clearly, the prediction of absolute values for emission reductions is also
of direct importance when considering atmospheric loading and ambient air
pollution concentrations.
                                     17

-------
00
                              TABLE 3-3  AVERAGE ABSOLUTE MASS EMISSION
                                        REDUCTIONS FOR SELECTED OPTIMAL PROCEDURES
INSPECTION/MAINTENANCE
PROCEDURE
ENGINE PARAMETER INSPECTION
1. IDLE (FRANCHISED)
2. EXTENSIVE A + (FRANCHISED)
3. EXTENSIVE B++ (FRANCHISED)
EXHAUST EMISSION INSPECTION
4. IDLE (STATE LANE)
5. EXTENSIVE A+ (STATE LANE)
6. EXTENSIVE B++ (STATE LANE)
AVERAGE REDUCTIONS +++
HC


37
69
85

18
51
53
CO
TONS /DAY

820
820
1394

908
908
978
NQ


-1
-1
-13

-2
-2
-6
                    +   IDLE PLUS IGNITION WITH DYNAMOMETER.
                    ++  IDLE PLUS IGNITION AND INDUCTION TUNEUP WITH DYNAMOMETER,
                    +++NEGATIVE VALUES INDICATE AN ACTUAL INCREASE OVER THE BASELINE,

-------
3J.2  Vehicle jxhaust  Emisslons
     Typical levels of  vehicular exhaust emission for each of the control
groups are presented in Table 3-4.  Also shown are the distributional
composition for each control type, average odometer reading and total
                                                o
vehicle population for  the Los Angeles Air Basin  as a function of time.
The ratio of CO to HC across all control types is greater than 10 to one
while the ratio of CO to NOV approaches twenty to one.  It is this relative
                           A
imbalance between the magnitude of the three emissions species that tends
to distort the impact of a mandatory program oriented towards HC or NOV
                                                                      X
control.
     While the average  odometer readings for the total population remain
nearly invariant over time, the percentage composition for the new car
category increases dramatically (approximately ten percent per year).
These population characteristics reflect a vehicle population growth from
5.7 million in 1972 to  nearly 6.3 million by 1977.  The substantially
lower mileage accumulation rates for the pre 1966 vehicles along with
their smaller percentage of the total population in later years indicates
the marginal utility of including these vehicles in an inspection and
maintenance program.

     The population weighted, average emission time histories (over a five
year period) for the two basic inspection approaches are shown in Figure
3-1 for an extensive "B" maintenance treatment.  These profiles reveal  de-
creasing emission levels with time relative to the base line for HC (Panel
A) and CO (Panel  B) and nearly constant levels for NOV (Panel C).  Two
                                                     )\
opposing effects are shaping these profiles—emission increases due to
engine deterioration and emission decreases due to engine maintenance and
vehicle attrition and replacement, i.e. newer cars entering the population
with lower emission levels.   As previously noted, all new vehicles entering
the population are assumed to have emission and engine characteristics iden-
tical  to the most advanced control systems studied in this program (1971
3.   The Los Angeles (South Coast) Air Basin as defined by the State of
    California "consists of all Orange and Ventura Counties and portions
    of Los Angeles, Riverside, San Bernardino and Santa Barbara Counties."
                                     19

-------
                                           TABLE 3-4 VEHICLE POPULATION CHARACTERISTICS
    YEAR
VEHICLE  DISTRIBUTICN   INCREMENTAL  ODCMETER    VEHICLE EHMISSION LEVELS   VEHICLE
ro
o
PERCENT

JAN
JAN

JAN

JAN
JAN

JAN
PRECCNTROLLED
CONTROLLED (1966-1970)
POST 1970
1972 TOTALS
PRECCNTRDLLED
CONTROLLED (1966-19701
POST 1970
1973 TOTALS
PRECCNTROLLED
CONTROULED (1966-1970)
POST 1970
1974 TOTALS
PRECCNTROLLED
CONTROLLED ( 1966-1970)
POST 1970
1975 TOTALS
PRECCNTROLLED
CONTROLLED (1966-1970)
POST 1970
1976 TOTALS
PRECCNTROLLED
CONTROLLED (1966-197C)
POST 1S70
1977 TOTALS
32.
46.
21.
100.
24.
43.
32.
100.
18.
39.
42.
100.
13.
34.
52.
100.
9.
28.
61.
100.
7.
22.
70.
100.
5
2
3
0
5
5
0
0
1
7
3
0
2
6
2
0
7
7
6
0
1
7
2
0
READING
MILES
4157
8418
14014
8225
3908
7262
13022
8281
3735
6265
12192
8312
3562
5409
11471
8328
3403
4743
10783
8335
3268
4441
10108
8336

11
7
4
8
12
7
5
7
12
7
5
7
11
7
6
7
11
6
6
7
11
6
6
7
HC
.94
.26
.43
.18
.21
.37
.18
.86
.04
.24
.62
.43
.84
.10
.01
.16
.61
.95
.34
.02
.35
.79
.62
.00
CO NO POPULATION
GRAMS/MILE
129.
86.
63.
95.
131.
93.
74.
96.
128.
93.
75.
92.
124.
92.
75.
87.
121.
90.
74,
83.
116.
87.
72.
78.
80
60
40
70
72
66
39
83
50
85
24
25
92
57
17
76
06
14
11
26
99
18
04
66
3.
6.
6.
5.
3.
6.
5.
5.
3.
5.
5.
5.
3,
5.
5.
5.
3.
5.
5.
5.
3.
5.
4,
4.
97
34
00
50
92
10
84
48
89
84
6C
39
85
62
37
25
82
41
14
09
79
21
91
90

5712000
5826240

5940480

6054720
6168960

6283200

-------
oo

LU

LU
to
U
X
10

 8


 6


 4


 2
                                                     PANEL A
                                         IIIIIIIIBIIII
                                                   Illlllllllllll
                                                      PANEL B
                                    BASELINE FLEET
                                    EXTENSIVE B (EMISSION)
                                    EXTENSIVE B (PARAMETER)
                             2          3

                            TIME, YEARS
       FIGURE 3-1   EMISSION TIME  HISTORIES  FOR AN
                    EXTENSIVE MAINTENANCE  PROGRAM
                              21

-------
California NOV Controlled Vehicles).  Consequently, the lowest possible
             X
emission values occur when the entire vehicle population is replaced with
these vehicles and they are efficiently maintained.  At the end of the
fifth year approximately 70 percent of the vehicle population is represented
by the post 1970 vehicles in varying states of deterioration.
     Both extensive "B" programs have their largest impact on CO and their
smallest impact on NO .  These procedures offer the greatest emission reduc-
                     A
tion potential consistent with existing garage effectiveness and reasonable
cost limitations.  The larger CO emission reductions are attributable dir-
ectly to the larger CO emission deterioration rates determined during the
experimental program.

3.1.3  Optimal Engine Inspection and Mode Emissions Inspection Criteria
     The engine components and adjustments as well as mode emissions sel-
ected as most cost effective for use in a mandatory program are identified
in Table 3-5.  Also shown are the inspection failure criteria for each para-
meter and mode.  The engine component pass/fail criteria are given relative
to manufacturer's specification and have been rounded off to reflect inspec-
tion and repair accuracy.  For example, results from the analysis indicated
on optimal pass/fail criteria of 0.9 percent for idle CO.   Experience gained
from several of the experimental programs indicated, however, that it was
difficult to inspect and adjust idle CO within 0.1 percent of the desired.
setting.  Consequently, the one percent value was selected as optimal for
a mandatory program.
     The engine parameters selected by the model include idle CO, misfire,
PCV and air cleaner for all three control groups and vacuum choke kick for
the emission controlled and post 1970 fleets.  For emission controlled
vehicles prior to 1970, the adjustment of idle rpm and timing is also cost-
effective.  It should be emphasized that this selection is based upon mini-
mizing a figure of merit which incorporates specific weighting factors on
the emission reductions.  The application of different weighting factors
may completely alter the engine components and adjustments selected and the
values of the pass/fail criteria.
                                    22

-------
                          TABLE 3-5   OPTIMAL  PASS/FAIL CRITERIA FOR LOS ANGELES


PARAMETERS/MODE EMISSIONS

o IDLE
o IDLE CO (55)
o RPM
o TIMING (DEG)
o IGNITION
o MISFIRE (55)
o NOX
o INDUCTION
o AIR PUMP
o PCV (CFPM)


o IDLE CO (%)
0 IDLE HC (PPM)

> 0 CRUISE 45 HC (PPM)



o AIR CLEANER (DEG)
o CHOKE VACUUM
KICK (IN.)'
o CHOKE HEAT RISER


o CRUISE 45 CO (55)


OPTIMAL SYSTEM PASS/FAIL CRITERIA*
PRECONTROLLED
ENGINE
PARAMETER

1.

2.5

- -
-1.
10.

— ,--
MODE
EMISSION

1.

650.



4.


CONTROLLED
ENGINE
PARAMETER

1.
-10.
1.

2.5

.---
-0.1
10.
-0.05
---
MODE
EMISSION

1,
300,

550.



2,


POST 1970
PARAMETER

1,

2.5

m^.«^w
-0.1
10.
-0.05
	
MODE
EMISSION

1.

450.



2.


Key:
Vehicle not equipped
      	  Parameter not Inspected
      ---   Mode emission not inspected
                                                                                    * As measured from manufacturers  spec.

-------
     Generally, a specific engine adjustment will not be included in the
criteria set for one or several of the following reasons:
         •  Its repair results in an increase of one of the emissions
         •  Its inspection and/or repair costs are large relative to the
            emission reduction effected
         •  The engine adjustment or its influence is unique only to a
            small subset of the vehicle population.

     In the first category are adjustments  and repairs related  to ignition
spark timing which tend to have counter productive effects.   Upon retarding
the ignition timing the CO emissions increase several times  faster than the
HC and NO  emissions decrease.  This effect has to be viewed,  however,  in
terms of the weights assigned to each emission.  In  situations  where HC is
of primary concern the retarding of timing  may be very cost effective.

     The choke system (vacuum kick and heat riser valve) fall  generally
into the second category of low leveraged repairs.  Although these repairs
have large influence coefficients for CO emissions and are frequently
failed,their repair costs are high and maintenance effectiveness low.   The
importance of maintaining these components  in a mandatory program of inspec-
tion and maintenance is also reduced when a low weighting factor is  assigned
to CO emission reductions.
     The NO  control device and air reactor systems  are typical of the
           /\
third category in that only a small fraction of the  vehicle population is'
equipped with these control systems.
     For each engine adjustment and component selected, the analytical
procedure tended to place the pass/fail criterion near the mode of its popu-
lation distribution or near manufacturer's  specification (e.g., 50 percent
rejection rates were often typical).  Only  in this way could maximum cost
effectiveness be achieved together with the stated program emission  reduction
goals.

3.2  MANDATORY MAINTENANCE PROGRAMS
     A mandatory program of engine maintenance represents another basic
strategy for reducing vehicular exhaust emissions.  This approach eliminates

                                    24

-------
the need for inspection (except for the tests required to diagnose specific
engine malfunctions).  The mandatory program examined in the study consisted
of the repair of all vehicles whose engine state was outside the limits
of manufacture specification.  For the idle engine adjustments, e.g. idle
RPM, the procedure required making only those corrections which, on the
average, yielded an emission reduction.  To illustrate this point, consider
the case for basic timing.  If timing were detected at a setting below
manufacturer's specification, then timing was not adjusted since such repair
would tend to produce higher HC emissions.  Consequently, only those vehicles
found with timing advanced beyond specification were maintained.
     The economic effectiveness model was used to simulate the impact of
two different mandatory maintenance scenarios.  The first one, based on
the procedure described above, assumed that the effectiveness of the mainte-
nance industry was consistent with that observed in experimental programs.
The second alternative assumed that perfect garage maintenance could be
achieved.
     The program figures-of-merit developed for these two alternatives were
5030 and 2045, respectively.  Obviously, either of these two programs com-
pare very favorably with the operational procedures outlined in the pre-
vious section.  They are expensive, costing nearly $40 per car per year.
On the other hand these scenarios yielded greater HC reductions (8.1 per-
cent for case #1 and 12.7 percent for case #2) than the optimized inspection
and maintenance procedures.  Additionally, for the case of perfect mainte-
nance, CO emission reductions were substantially larger (nearly 18 percent).
Figure 3-2 depicts the impact of these procedures on emission levels as a
function of time.  The greatest change relative to the baseline occurs for
CO using an improved maintenance program, i.e. perfect maintenance.
     A mandatory maintenance program has some complicated administrative
and logistical details.  Among the more serious ones are:
         •  Lack of sufficient certified repair facilities to maintain
            the total vehicle population
         •  High probability of excessive maintenance and high costs when
            procedures are not well prescribed
         •  Highly complicated data recording and management system
            required to administer program
                                     25

-------
                                                 PAMELA
to

UJ
z
O
to
    125
    100
     75
£ ?  50
O
u
     25



      0



      8
 to
 oo
 U
      0
                                                  PANEL B
                T'""""..1
                            in..
                                 """• "-MO,
                                                      "MJ
                                                  PANEL C

^^^"i^rW^^F



^^W^ffH^ff

minium

"mmf^f

•HtlTHH^ff

"ifl^flfflff
BASELINE FLEET
MANDATORY MAINTENANCE
IMPROVED MANDATORY MAINTENANCE
       0
                           2         3

                           TIME, YEARS
          FIGURE 3-2  EMISSION TIME HISTORIES FOR

                     MANDATORY MAINTENANCE PROGRAMS

                           26

-------
         •  Little direct assurance of program effectiveness without sophis-
            ticated surveillance program
         •  Standard vehicle "performance and driveability" tune-up tends
            to increase instead of decrease exhaust emissions
     On the positive side, however, a direct mandatory program could take
advantage of current ownership habits with respect to engine maintenance.
Recent evidence from the deterioration experiments seems to indicate that
there is a high degree of voluntary maintenance occurring within the gen-
eral population.  Unfortunately, the two problems identified above-
repair reliability and "performance" tuning—tend to dilute the effective-
ness of voluntary maintenance.

3.3  HYBRID INSPECTION/MAINTENANCE PROGRAMS
     A hybrid program of vehicle inspection and maintenance attempts to
combine the key features of mandatory vehicle inspection and maintenance
with those of a mandatory maintenance program.  This study focused primarily
on two different hybrid procedures:
         •  Idle inspection and maintenance along with the mandatory
            replacement of PCV value and air cleaner
         •  Random vehicle inspection followed by extensive engine
            maintenance
An idle inspection and maintenance program with mandatory parts replace-
ment is designed primarily to reduce CO since it mainly inspects and repairs
induction oriented engine components.  Previous studies as well as the cur-
rent one have always found idle programs to be cost effective.  The major
concern with the idle program previously analyzed is that they generally
yield rather modest HC and CO emission reductions.  The basic approach
here involves augmenting the idle adjustments with the mandatory replace-
ment of the PCV valve and air cleaner on all vehicles every year.  PCV
valve and air cleaner replacement offers good emission reduction potential
for fairly reasonable costs (about $3.00 for PCV valve and $5.00 for air
cleaner).   The surveys of engine maintenance states (Volume IV) have
revealed that a rather high percentage of the vehicle population requires
the replacement of either or both of these components.  The deterioration
                                     27

-------
data (Volume VIII) indicates, furthermore,  that replaced components tend
to return to their previous state after approximately one year in service.
     The other procedure examined involves  a random idle emission inspection
of the vehicle population coupled with extensive maintenance for those
vehicles failing the test.   This program is based on the premise that main-
taining vehicles with high  emissions will yield very cost effective results.
The program further provides a flexible method of inspection (through the
random selection process) and thereby permits a state or region to gradually
increase the segment of the population that participates in the program.
This feature has several key advantages:
         •  It allows sufficient time to improve and expand current
            repair facilities
         t  It permits a state or region to gain experience with the
            operation of an inspection and  maintenance system
         •  It provides an  opportunity to acquaint the motoring public
            with the control program before implementing a complete
            system
The main drawback of this approach rests with the relative low emission
reductions which can be expected from inspecting and maintaining only a
small segment of the vehicle population.  For example, if 50 percent of the
vehicle population were randomly inspected  and 50 percent failed to pass,
then only 25 percent of the vehicle population would undergo repair.  Our
analysis has shown that if  less than 20 percent of the population are inspec-
ted and maintained, then the resultant emission reductions on a total popula-
tion basis are not statistically significant.
     Both of these hybrid procedures yield  more attractive results (with
respect to the figure of merit) than programs of mandatory maintenance only.
These procedures tend to be substantially  less expensive but yield, on the
average, somewhat lower emission reductions than mandatory maintenance.
For example, the hybrid emission inspection with replacement program pro-
duced only one-half the HC  emission reduction predicted for mandatory mainte-
nance.  For CO, however, the predicted emission reductions for the hybrid
program were nearly 50 percent greater.  NO  emission reductions for both
                                           /\
procedures were equal.

                                     28

-------
     Table 3-6 summarizes the results for the three different inspection
and maintenance strategies examined.  As can be seen, the two original
mandatory inspection and maintenance  procedures  (#1 and  #2) still appear
to yield the most cost-effective results in terms of the figure of merit.
     The two hybrid programs, however, certainly provide competitive alter-
natives.  Their figures of merit at 1425 and 1750 compare favorably with
the 1340 value for the loaded mode emission inspection procedure.  Both
hybrid procedures yielded greater emission reductions than the idle engine
inspection and maintenance program.  The mandatory maintenance programs,
on the other hand, showed less attractive performance.  Even with improved
maintenance (100 percent efficiency) the cost-effectiveness performance of
these programs appear marginal.
     The candidate procedures identified above were all designed to
reduce HC emissions with the possible exception of Case #5.  (All figures
of merit were computed using the Los Angeles emission weighting factors.)
In situations where CO is of prime importance, the relative attractiveness
of these procedures may dramatically change.
                                     29

-------
TABLE  3-6  COMPARISON OF  EVALUATED  PROGRAMS
CANDIDATE PROGRAMS
MANDATORY INSPECTION MAINTENANCE
1) ENGINE INSPECTION (IDLE)
2) EMISSION INSPECTION (EXTENSIVE A)
MANDATORY MAINTENANCE
3) EXISTING MAINTENANCE
4) IMPROVED MAINTENANCE
HYBRID INSPECTION/MAINTENANCE
5) EMISSION INSPECTION WITH REPLACEMENT
6) RANDOM INSPECTION (50%)
FIGURE OF MERIT
S/TQN+++
ABC

785 320 540
1340 560 915

5030 2200 3510
2045 870 1030

1425 1735 1010
1750 825 1280
COST PER
VEHICLE
D E

4.50 5.25
9.00 16.75

39.60 39.70
39.60 39.70

12.50 13.25
10.25 12.05
EMISSIONS REDUCTIONS, PERCENT
HC CO NO

3.3 6.1 -0.1
4.7 6.8 -0.2

8.1 5.8 -0.2
12.6 17.6 0.4

3.6 9.4 -0.2
3.8 5.1 -1.0
                           A.  COMPUTED OVER FIVE YEAR PERIOD.
                           B.  COMPUTED AT END OF LAST INTERVAL,
                           C.  BASED ON STATISTICALLY SIGNIFICANT EMISSIONS
                              REDUCTIONS FOR LAST INTERVAL.
                           D.  BASED ON TOTAL VEHICLE POPULATION.
                           E.  BASED ON MAINTAINED PORTION OF VEHICLE
                              POPULATION.

-------
                     4.0  S.YSTEW SENSITIVITY ANALYSIS

     The economic effectiveness model was used to determine the sensitivity
of model predictions to changes in the basic data set.  The primary system
elements considered in the sensitivity study were:
     •  The Impact of Vehicle Deterioration Rates on Program Performance
     •  The Impact of the Engine Deterioration on Exhaust Mode Emissions
     •  The Impact of the Inspection and Maintenance Process on Vehicle
        Rejection Rates
     t  The Impact of Inspection Frequency on Program Performance
The reader is encouraged to review the year end report to gain insight into
the impact of other system variables on program performance (e.g., levels
of voluntary maintenance and emission weighing factors).
4J  THE IMPACT OF VEHICLE DETERIORATION RATE ON PROGRAM PERFORMANCE
     The rate of engine deterioration and corresponding emission deteriora-
tion is crucial to the overall inspection and maintenance process.  Vehicle
emission deterioration can be divided into two basic segments:
     •  That accounted for by the deterioration of one or more of the
        basic ten emission oriented engine parameters.
     •  That accounted for by the deterioration of engine parameters
        other than the basic ten.
     The basic governing relationship between emissions and engine deterio-
ration is given in equation 4-1.
(4-1)
       Total emission    =   V   (Influence)      X   (Parameter)
       deterioration         ^   (coefficients)       (deterioration)
       for each                                        (rates)
       etrns!f"lon          +     E  (Emission deterioration not
       species                  o accounte(j fOF by the ten engine
                                  parameters)
Estimates of E  for each species can be made by simply subtracting the
summed product of the influence coefficients and engine adjustment deterio-
ration rates from the total emission deterioration values.  Obviously, the
more emission deterioration accounted for by the ten parameter system the
smaller the value for EQ.  Computed values for EQ averaged approximately
                                     31

-------
ten to fifteen percent of total emission deterioration for all three
emission species.
     To determine the sensitivity of program performance to adjustment and
emission deterioration rates several parametric cases were run using the
model.  The first case involved a 20 percent reduction in overall  emission
deterioration rates while the second case was characterized by a 20 percent
increase in exhaust emission deterioration rates.  The results of these
cases are presented in Figure 4-1.  Shown are the program figure of merit
(Panel A), the HC percentage reductions (Panel B), CO percentage reductions
(Panel C) and NO  percentage reductions (Panel D) for the three deterioration
                X
rates.  As expected, the performance of the loaded mode emission inspection
program improved as the rate of deterioration increases.  For example, pre-
dictions of HC reductions increase from approximately 4 percent for the low
rate of deterioration to nearly 9 percent using the higher rate of deterio-
ration.  These results reflect the highly complex and nonlinear effects of
deterioration on program performance.  Clearly, engine and emission deterio-
ration will play a crucial role in determining the optimal inspection and
maintenance procedure for a given application.

4.2  THE IMPACT OF THE INSPECTION AND MAINTENANCE PROCESS ON VEHICLE
     REJECTION RATES
     Vehicle rejection rates are affected primarily by program pass/fail
criteria and the distributional form of the particular engine parameter
and/or mode emission.  Additionally, the availability of certified repair
facilities acts as an upper limit on the number of vehicles that can be
properly maintained.
     For example, if there existed adequate service facilities to main-
tain only 40 percent of the vehicle population over the course of the year
then this would limit the number of vehicles to be maintained even though
the "optimal" rejection rate might be 60 percent.  Each of these factors
will be explored in turn.
     Figure 4-2 presents the population weighted fraction of vehicles re-
jected in an emission inspection by each separate idle or loaded mode
measurement of HC and CO.  These rejection rates are relatively constant
                                    32

-------
CO

OJ
                LOADED MODE EMISSION INSPECTION WITH EXTENSIVE B MAINTENANCE
                      FIGURE OF MERIT
                                         2000
                                                   PANEL A
                                          1000
                                                  I
                                                 LOW      HIGH

                                                   NOMINAL
                                                                     KEY

                                                                     LOW:    -20% OF NOMINAL

                                                                     HIGH:    +20% OF NOMINAL
                        PANEL B
             Z
             o
             o
             UJ
             u
             X
                10
                 0
                       I
I
I
                     LOW     HIGH

                       NOMINAL
               20
             Z
             O
             I—
             u


             2  10
O
u
                        PANEL C
                                                  _L
                           j_
                           I
                     LOW     HIGH

                       NOMINAL
                                   Q

                                   u

                                   Q
                                   UJ
o
Z
                                                                       0
                                           -5
                                              PANELD
I
                                           LOW     HIGH

                                              NOMINAL
                   FIGURE 4-1  INFLUENCE OF VEHICLE DETERIORATION  RATE ON  PROGRAM PERFORMANCE

-------
                              LOADED MODE EMISSION  INSPECTION WITH EXTENSIVE B MAINTENANCE
                     PANEL A
OJ
            Q
            LU
            I—
            U
            LU
            LU

            U

            LU
            Q_
            Q
            LU

            U
            LU
            LU
            U

            LU
            Q_
60



40



20


 0
                                               ICO
                                                         PANEL B
                         1
               1
1
1
                         12345

                       PROGRAM PERIOD -YEARS
                     PANEL C
60



40



20


 0
                                              HC45
                         12345

                        PROGRAM PERIOD ~ YEARS
Q
LU
U
LU
LU
1—
LU
s
0.


o
LU
D
LU
"^
LU
LU
U
Q_


60

40
20
0
- in

•" "
- * 	 — ^
1 1 1 1 1
012345
PROGRAM PERIOD ~ YEARS
PANEL D
60
40
20

0
CO45
••"
- "" —

1 1 11 i
D 1 2 3 45
PROGRAM PERIOD - YEARS
                                 FIGURE  4-2  MODE EMISSION REJECTION HISTORIES

-------
over the program time period and also can be shown to be insensitive to
                               4
various maintenance procedures.
     Inspections which use both idle and loaded mode HC emissions result in
a vehicle rejection level that decreases more rapidly with program duration
than does the vehicle rejection level based upon either the idle or loaded
mode CO emission inspection.  This is because the first few inspections
detect most cases of ignition misfire in the population.  The other adjust-
ments and component replacements which have a large impact on HC emissions
(NO  control and timing) tend to have low deterioration rates and/or their
   /\
failure is not adequately diagnosed using a mode emission inspection.
Indeed, the effectiveness of the emission inspection strategy can largely
be attributed to the adequate diagnosis of idle fuel-to-air ratio maladjust-
ment, misfire and perhaps air cleaner and PCV valve restrictions.
     Combining each of the separate mode rejection fractions yields an
estimate of the total vehicle rejection rate for a given period of time.
This synthesis is accomplished using probability distribution algorithmn
which is discussed in Volume III.  An inspection using more than one
emission mode, say two, results in a vehicle rejection rate which exceeds
either of the individual mode rejection fractions.  Consequently, it
should not be surprising to observe vehicle rejection rates in the neigh-
borhood of 40-60 percent even though the individual emission mode rejection
fractions are only one-half that value.
     Computed values of vehicle rejection rates relative to the total  popu-
lation are presented in Panels A and B of Figure 4-3 for the two basic
inspection approaches.  The rejection rates for an engine inspection proce-
dure are slightly higher than reported in the year end report.  This is
due primarily to the incorporation of several other engine adjustments
(e.g., vacuum choke kick) into the basic inspection process.
  It should be noted that the optimal pass/fail criteria developed for the
  first period was applied to all subsequent periods.
  This same logic also applies to the inspection of engine malfunctions.
  In this case, moreover, the anticipated vehicle rejection rates should
  be even higher since there are more basic parameters (up to 10) involved
  in the inspection process.
                                    35

-------
                 ENGINE INSPECTION STRATEGY
O_


I

to
z
o

                                                  PANEL A
   100
80
   60
   40
   20
                                               KEY:

                                     minium  REJECTED

                                               MAINTAINED
                EMISSION INSPECTION STRATEGY
                                                   PANEL B
                1          2          3

                       TIME ~ YEARS
     FIGURE 4-3   VEHICLE POPULATION REJECTION RATES
                          36

-------
     Two plots are given for the emission inspection strategy (Panel  B).
The first presents the vehicle rejection rate time history while the other
indicates the actual level of vehicle maintenance.  These rates are in
good agreement with those developed in the year end report.  The difference
between the two curves indicates the level of "commission errors" for the
given procedure.  In this case, these errors ran approximately 15-20 per-
cent.  Estimates of "omission errors" are more difficult to develop since
they required a direct examination of all vehicles in the population.
Evidence from the experimental programs indicates that "omission errors"
account for approximately 10 percent of the population undergoing a loaded
mode emission test.  That is, approximately 10 percent of the vehicle
population with basic engine malfunctions will not be detected with a mode
emission inspection.
     In general, both strategies yield rejection rates substantially higher
than would be feasible based on available service organization facilities
and public acceptance.  One approach to this dilemma is to undertake a
constrained optimization in which the vehicle rejection rate cannot exceed
a preassigned value.  The economic effectiveness model was utilized to
evaluate the impact of two different restricted rejection rates on overall
program performance for an engine inspection with extensive B maintenance.
The rejection rates selected for this exercise were 30 and 50 percent,
respectively.  Table 4-1 summarizes the results of these simulations.  As
can be seen, limiting the rejection rate to 50 percent results in a slightly
improved figure of merit relative to the unrestricted case, but yields
somewhat lower emission reductions.   For the 30 percent case, however,
the figure of merit is nearly 50 percent higher than for the other two.
This can be attributed to the substantially smaller CO emission reduction
achieved without an equivalent decrease in program costs.
  The optimization algorithm used in the economic effectiveness model
  attempts to maximize emission reductions subject to a series of cost
  and logical constraints.  It should not be surprising, therefore, to
  obtain an improved figure of merit (emission reductions/costs) by
  limiting the number of vehicles undergoing maintenance.
                                    37

-------
              TABLE 4-1   COMPARISON OF PROGRAM PERFORMANCE
                          FOR VARIOUS VEHICLE REJECTION RATES
Procedure
1 . Extensive B
2. Extensive B+ (50
Percent Rejection)
3. Extensive B+ {30
Percent Rejection)
Figure of Merit
$/Ton
2865
2770
4535
Costs
32.10
21.70
18.40
Emission Reductions, Percent
HC
7.8
5.1
4.2
CO
10.4
7.9
3.5
NOV
-1.6
-1.2
-0.8
+ Idle plus ignition plus induction  tuneup with  dynamometer


 4.3  THE  IMPACT OF  INSPECTION FREQUENCY ON PROGRAM PERFORMANCE
     Frequency of inspection is an important variable in terms of its
 impact on  procedure effectiveness.  Shorter inspection intervals result in
 more frequent vehicle repair for a larger segment of the population and
 therefore produce larger emission reductions at correspondingly higher
 costs.  Longer intervals, on the other hand, yield smaller emission reduc-
 tions for  relatively modest costs.  The basic tradeoff is to determine an
 optimal inspection interval which results in acceptable emission reductions
 for reasonable program costs.  Fortunately, there are only two periods --
 six and twelve months that are consistent with current state practices
 relating to vehicle safety inspection and registration (the  two most
 probable mechanisms for recording and administering a mandatory program).
     Figure 4-4 illustrates the impact of inspection frequency on procedure
 effectiveness for a loaded mode emission inspection with extensive B main-
 tenance.  Panel A reveals that an annual inspection yields the most cost
 effective results in terms of the figure of merit.  Both HC  and CO emission
 reductions, Panels B and C, increase as the time between inspections is
 shortened.  Correspondingly, the emission reduction achieveable for these
 species as well as for NOX (Panel D) tends towards zero as the frequency
of inspection approaches 18 months.   At this point, the voluntary
                                    38

-------
                              INFLUENCE OF INSPECTION PERIOD
                         LOADED EMISSION INSPECTION WITH EXTENSIVE B MAINTENANCE
                                               PANEL A
CO
vo
                                    20001-
9 iooo
LU
Qfi
O
LL,

PANEL B
z20r
0 O
1— — 1—
(J U
Q Q
£ 10 ~ *^^ £
c£ ^^^ £
u ^^° O
n 1 ' ' ' ' u
6 12 18
INSPECTION PERIOD -MONTHS
0
—
i i I i i
6 12 18

INSPECTION PERIOD - MONTHS
PANEL C PANEL D
20
10

0
z 5
O
\Q
T
o
' ' 1 1 1 z -,

T-~

I I i I i
6 12 18 " 6 12 H
INSPECTION PERIOD ~ MONTHS
NSPECTION PERIOD-MONTI
                FIGURE 4-4  INFLUENCE OF INSPECTION FREQUENCY ON PROCEDURE EFFECTIVENESS

-------
maintenance program becomes "equivalent" to the mandatory program and the
net emission reduction (voluntary less mandatory) becomes zero.
                                    40

-------
                          5.0  STUDY CONCLUSIONS


     The conclusions derived from this study reflect the final experimental

data and latest version of the economic effectiveness model.  This analysis

differs from the previous one (year end) in the following regards:

         o  The final data from the experimental study of engine adjustment
            and exhaust emission deterioration was used throughout the
            analysis.

         o  The economic effectiveness model was revised to reflect the
            latest experimental  data and forecasting methodologies.

         o  A wider range of inspection and maintenance alternatives were
            examined.

     A summary of the final study is presented by major areas of investi-
gation.


5.1  INSPECTION AND MAINTENANCE PROCEDURES

     The results from evaluating the three different types of programs

are discussed below.  The criteria used to evaluate the relative merit of

each of these programs was a system figure of merit defined as the total

discounted program costs per weighted ton of emission reductions.  Emis-

sion weighting factors of 0.6, 0.1, and 0.3 were applied for HC, CO and NO ,
                                                                          /\
respectively.

     1)  The results obtained using the latest experimental data and
         reconfigured economic effectiveness model indicated degradation
         in program performance when contrasted with the year end report
         results.  This can be attributed primarily to the lower rate of
         misfire detected in the vehicle population.

     2)  A hybrid program involving an idle emission inspection and mainte-
         nance program as well as the mandatory replacement of air cleaners
         and PCV valve, emerged as the most cost effective procedure for
         regions with a CO problem.

     3)  A mandatory maintenance program does not appear very cost-effec-
         tive particularly in view of the relatively low repair effectiveness
         of maintenance organizations.

-------
     4)   The largest emission reductions  were  obtained with a  direct
         engine inspection  followed by an extensive "B" maintenance
         treatment.   This  program is eight times  more costly than  the most
         cost effective procedure; however, the emission reductions achieved
         are nearly twice  as large.

     5)   A loaded mode emission inspection followed by the  repair  of the
         idle and ignition  systems emerged as  the most cost-effective
         procedure for regions having photochemical smog.

     6)   Emission inspections performed in state  lanes are  more cost
         effective than inspections performed  in  licensed or franchised
         garages.

     7)   The most cost effective parameters to maintain are:
         a)  idle CO,  b)   timing,  c)  ignition  misfire,
         d)  PCV valve   e)  air cleaner.
5.2  SYSTEM SENSITIVITY ANALYSIS

     Sensitivity analyses were performed to measure the influence of data
uncertainties and model assumptions  on  the effectiveness of various  inspec-
tion and maintenance procedures.  The candidate procedures  were  found to
be most sensitive to the following factors which are presented in order of
their importance:


     1)  Regional Air Pollution -  The uniqueness of regional  air pollution
         problems requires the development of a region  specific  program.
         The characteristics of a  given air quality problem,  as  character-
         ized by the emission weighting factor, significantly alter  the
         relative merit of candidate procedures.

     2)  Unauthorized Maintenance  -  During the deterioration experiment
         a large percent of the vehicle owners had their cars maintained
         even when specifically requested not to do so.  These data  suggest
         that owners may take deliberate actions to improve their vehicle
         performance and driveability at the expense of increased emissions.
         Certainly the cooperation of the service industry  will  be required
         if vehicle readjustments  which undermine the effectiveness  of man-
         datory vehicle inspection and  maintenance are  to be minimized.

     3)  Service Organization Repair Effectiveness - The effectiveness with
         which repair is made by a service organization has a major  impact
         on resultant emission reductions.  For example, if repair relia-
         bility were improved, larger HC and particularly larger CO  emis-
         sion reductions could be  achieved.
                                    42

-------
     4)  Emission and Engine Deterioration - Engine deterioration and
         resultant emission deterioration can have a substantial  effect
         on procedure effectiveness.   Higher rates of deterioration tend
         to improve the effectiveness of all procedures in general  and
         the more extensive procedures in particular.

     5)  Vehicle Population Characteristics - The analysis has revealed
         that shifts in the makeup of the vehicle population, i.e., the
         replacement of old cars with newer cars, will have a significant
         influence on the cost-effectiveness of the various procedures.
         As a result, it is concluded that precontrolled vehicles
         (especially for the State of California) are not an important
         factor in the long term selection of inspection and maintenance
         procedures.

     6)  Extent of Voluntary Maintenance - The extent of voluntary mainte-
         nance, as it is currently modeled, has its largest impact on CO.
         Obviously, as vehicles are better maintained voluntarily, the
         less the need for a mandatory program.

     7)  Frequency of Inspection and Repair - The frequency of inspection
         and maintenance affects program performance.  Our studies have
         continuously shown that a yearly program is most cost effective.

     8)  Costs of Unnecessary Repair - The cost of unnecessary repair
         account for a major segment of overall program cost.  In general,
         these costs increase disproportionally with the extent of repair.
         Consequently, they introduce an additional bias in favor of idle
         oriented programs.

     9)  Inspection Accuracy - Variability in inspection accuracy can have
         an important impact on the effectiveness of various inspection
         and maintenance alternatives.  This is particularly the  case for
         programs with low rejection rates or with complicated diagnostic
         procedures.


5.3  REGIONAL IMPLICATIONS OF VEHICLE INSPECTION AND MAINTENANCE

     As .presented in the year end report, two hypothetical regions were

modeled having different air quality problems and vehicle characteristics.

The economic-effectiveness of mandatory inspection and maintenance is most

influenced by the emission weighting factors assigned in various regions.

Weighting factors which stress CO emission reduction generally result in

a program that does little to control HC and NO emissions.  However, large

absolute reductions of CO occur, thereby resulting in extremely good fig-

ures of merit.  Those procedures which involve adjusting the idle fuel-to-

air ratio to specification are most effective for controlling CO emissions.
                                    43

-------
CO emission reductions can be approximately doubled by maintaining compo-
nents of the induction system but at four times the cost.   Even larger
improvements can be achieved if induction system maintenance effectiveness
is improved.
     In regions where photochemical  smog is a problem and  HC and NO  emis-
                                                                   A
sion reductions are heavily weighted, timing adjustments and ignition
system repair become important.  Little reduction of CO emissions is de-
sired since NO  emissions are adversely affected.
              A
     These regional results clearly indicate the requirement for a "cus-
tomized" program designed to meet a specific air pollution problem.   The
arbitrary application of a program developed for one region to a region
with differing air quality and demographic characteristics could yield
counterproductive results.
     The following specific conclusions emerge with respect to the basic
types of pollution problems:
Carbon Monoxide Air Quality
         o  For regions characterized by a CO problem the  most cost
            effective procedure is an idle inspection and  repair program
            along with the mandatory replacement of air cleaner and PCV
            valve.
         o  An idle program designed exclusively to reduce CO may in-
            crease HC emissions.
         o  The most extensive CO oriented procedure examined yielded
            emission reductions of 6, 11 and -2 percent for HC, CO and
            NO  respectively.
              A
Photochemical Smog Air Quality
         o  For regions characterized by a photochemical problem the
            most cost effective procedure is based on a loaded mode emission
            inspection followed by the repair of the idle  and ignition
            systems.
         o  The most comprehensive smog oriented inspection and mainte-
            nance procedure studied produced HC, CO and NO  reductions of
            10, 6 and -1, respectively.
                                    44

-------
Oxides of Nitrogen  Air  Quality

     The development  of a  maintenance  procedure  designed  to  reduce NOX
emission poses  a  serious problem which was  not examined directly during
the course of this  study.   In general, procedures  designed  to  decrease
NOX emissions will  tend to increase  HC and  CO emissions.  Consequently,
care must be exercised  in  developing such a program (if required)  in order
that one class  of air pollution  problem is  not substituted  for another.
                                     45

-------