THE ECONOMIC EFFECTIVENESS OF MANDATORY
ENGINE MAINTENANCE FOR REDUCING
VEHICLE EXHAUST EMISSIONS
VOLUME I
EXECUTIVE SUMMARY
t
c,
AUGUST 9, 1971
I N SUPPORT OF:
APRAC PROJECT NUMBER CAPE-13-68
FOR
COORDI NATI NG RESEARCH COUNCIL, I NC.
THI RTY ROCKEFELLER PLAZA
NEW YORK, NEW YORK 10020
AND
ENVI RONMENTAL PROTECTION AGENCY
AIR POLLUTI ON CONTROL OFFICE
5600 FISHERS LANE
ROCKVILLE, MARYLAND 20852
TRW
SYSTEMS GROUP
ON£ SPACE PARK • flfOONDO BEACH. CALIFORNIA 90278
-------
THE ECONOMIC EFFECTIVENESS OF MANDATORY
ENGINE MAINTENANCE FOR REDUCING
VEHICLE EXHAUST EM IS SIONS
VOLUME I
EXECUTIVE SUMMARY
AUGUST 9, 1971
I N SUPPORT OF:
APRAC PROJECT NUMBER CAPE-13-68
FOR
COORDI NATI NG RESEARCH COUNCIL, INC.
THI RTY ROCKEFELLER PLAZA
NEW YORK, NEW YORK 10020
AND
ENVI RONMENTAL PROTECTION AGENCY
AIR POLLUTION CONTROL OFFI CE
5600 FI SHERS LANE
ROCKVI LLE, MARYLAND 20852
APPROVED BY %(~tJNI ~@~~
RI CHARD R. KOPPANGt
PROJECT ENGINEER
/~~~
NEAL A. RICHARDSON
PROJ ECT MANAG ER
-
'.
TRW
6YSUIU GIlOU~
ONE SPACE PARK. REDONDO BEACH. CALIFORNIA 90278
,-
-------
PREFACE
This report consists of three volumes entitled: liThe Economic
Effectiveness of Mandatory Engine Maintenance for Reducing Vehicle
Exhaust Emissions". The following are the titles given for each
volume:
. Executive Summary, Volume I
. Modeling of Inspection/Maintenance Systems, Volume II
.
Inspection and Maintenance Procedures Development, Volume III
The first volume summarizes the general objectives, approach and
results of the study. The second volume presents the analytical modeling
of a mandatory inspection/maintenance system and simulation results
obtained using that system model. The experimental programs conducted to
develop input data for the model are described in Volume III.
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 conducted the emission instrument evaluation
and acquired and tested all of the study vehicles. Scott also provided
technical assistance in selecting emission test procedures and in evalu-
ating the test results.
i
-------
I
I -
TABLE OF CONTENTS
VOLUME I
l.
2.
INTRODUCTORY SUMMARY. .
. . . . . .
. . . . . . . . . . .
SYSTEM CONSIDERATIONS. . . . . . . . . . . . . . . . . .
2.1 Engineering Design. . . . . . . . . . . . . . . . .
2.2 Economic Factors. . . . . . . . . . . . . . ... . . .
2.3 Cons tra i nts . . . . . . . . . . . . . . . . . . . . .
2.4 Program Effectiveness. . . . . . . . . . . . . . . .
3.
PROCEDURES EVALUATION. . . . . . . . . . . . . . . . . .
3.1 Data Acquisition. . . . . . . . . . . . . . . . . .
3.1.1 Frequency and Extent of Malfunction. . . . . .
3.1.2 Emission Sensitivity to Malfunction. . . . . .
3.1.3 Emissi.on Signatures. . . . . . . . . . . . . .
3.1.4 Inspection and ~1aintenance Costs. . . . . . .
3.1.5 Deterioration of Maintenance. . . . . . . . .
3.2 Economic-Effectiveness Model. . . . . . . . . . . .
3.2.1 Emission Predictor Model. . . . . . . . . . .
3.2.2 Cost Estimator Model. . . . . . . . . . . . .
3.2.3 Implicit Cost and Operations Model. . . . . .
3.2.4 Figure of Merit. . . . . . . . . . . . . . . .
4.
ECONOMIC-EFFECTIVENESS STUDY RESULTS. . . . . . . . . . .
4.1 Engine Parameter Diagnostics. . . . . . . . . . . . .
4.2 Mode Emission Signature Analysis. . . . . . . . . . .
4.3 Conclusions and Recommendations. . . . . . . . . . .
REFERENCES. . .
. . . . . . . . . .
. . . .
... . . . . .
i i
Page
1
4
4
10
10
10
11
11
11
14
24
26
30
32
35
36
36
36
40
43
46
50
52
-------
1.
INTRODUCTORY SUMMARY
The effects of mandatory vehicle inspection and maintenance were in-
vestigated to assess the most cost-effective procedures for periodically
diagnosing and restoring to specification those engine components having
significant effects on vehicle exhaust emissions. Two classes of inspec-
tion procedures were evaluated:
. Direct measurement of engine parameter adjustments
(malfunctions) using conventional or more sophis-
ticated garage-type equipment
.
Inference of engine parameter maladjustments or
malfunctions from the measurement of engine exhaust
emission levels (signatures) under varying loading
conditions. .
The former inspection procedure would most likely be performed in a
franchised or certified, privately owned garage. The effectiveness of
combinations of direct parameter measurements which resulted in shor~
(approximately 3 minutes) and long (approximately 30 minutes). inspections
were evaluated. Because direct parameter measurements using existing
diagnosti~ equipment are time consuming, remote sensing instruments felt to
be technically feasible were hypothesized and evaluated for use in direct
. diagnosis in state inspection lanes.
The emission signature inspection procedure requires more sophisticated
equipment and instrumentation. Emission signatures for hydrocarbons,
carbon monoxide and carbon dioxide would be measured under several engine
loads. The use of emission signature inspection results in a higher vehicle
inspection throughput but at the expense of a greater investment in
capital equipment.
The general
which affect the
procedures in an
framework of the study focuses on those system elements
ranking,of the alternative inspection and maintenance
economic effectiveness sense. The tasks include:
. Systems definition
. Statistical characterization of vehicle maintenance
states in an urban region
1
-------
.
Development of procedures for inspecting and main-
taining those malfunctioning engine parameters
which significantly influence emissions
. Development of a computerized system model for
evaluating the candidate procedures within a
systematic application framework.
A summary of these tasks is presented in the following sections of this
volume.
Conclusions derived from this study are:
. The six most effective engine parameters to maintain are
the three idle adjustments (air~to-fuel ratio, rpm and
basic timing), elements of the ignition system when
causing misfire, as well as the positive crankcase ventil-
ation valve and air cleaner of the induction system. The
air injection system should be inspected and maintained on
cars equipped with this type of air pollution control
equipment.
.
Inspection and maintenance of the idle adjustments
was found to be a very cost effective procedure for
controlling carbon monoxide emissions. Typical aver-
age emission reductions over a four year period are
between 2 to 3 percent for hydrocarbons and 10 to 15
percent for carbon monoxide. Oxides of nitrogen emis-
sions are increased by 4 to 7 percent.
. Control of both hydrocarbon and carbon monoxide emis-
sions requires inspection and maintenance of the igni-
tion and induction systems in addition to the idle
parameters. Optimum inspection/maintenance procedures
yield a typical average emission reduction over a four
year period of 15-22 percent for hydrocarbons and 20-33
percent for carbon monoxide. Oxides of nitrogen emis-
sions are increased from 3 to 5 percent by this treat-
ment.
. The most cost effective inspection frequency is once
yearly.
. State inspection lanes are usually more cost effective
than franchised garages.
. Nondispersive, infrared emission measurement instruments
are preferred for state-lane applications.
2
-------
The above conclusions were based on currently available data, some
of which were either of limited reliability or based upon small test
populations. In addition, the effectiveness of maintenance was inferred
from a limited set of representative power trains (i.e., 1966-1970
vehicles). Considerable emission reductions may be possible when the
precontrolled vehicle population is included in the inspection/maintenance
program.
Additional experiments are recommended to determine the effect on
inspection/maintenance procedures of:
Precontrolled and NOx controlled vehicles in an
urban population
. The reliability with which commercial repair
agencies can diagnose and repair vehicle exhaust
. cont ro 1 and re 1 a ted sys tems
.
. Differing urban regional air quality require-
ments, vehicle population composition, and
general vehitle maintenance states.
3
-------
2.
SYSTEM CONSIDERATIONS
The relationship between system design and economic factors must be
clearly understood so that inspection/maintenance procedures are selected
in combinations to yield optimal cost and performance. Three fundamental
elements were considered in the development of the vehicle emission in-
spection/maintenance system model (Figure 2-1):
. Engineering design
. Economic factors
. Constraints
System design and the associated economic factors are dependent upon
regional demography, socioeconomic considerations and available or fore-
seeable technology.
2.1
ENGINEERING DESIGN
Engineering design factors fall into two categories: inspection
and maintenance procedures and facilities configuration. Significant
factors to be determined when selecting optimum procedures are: the
extent and complexity of inspection and maintenance procedures, the
rejection criteria, and the frequency of inspection. Configuration
factors to be considered are: .
. Number and type of inspection stations for
typical demographic regions
.
Degree of automation of procedures
. Type and complexity of sensing instr~ments
. The information system to support adminis-
trative and enforcement decisions.
Of particular interest is the definition of optimally combined
inspection/maintenance procedures for a state vehicle inspection system.
4
-------
ENGINEERING DESIGN.
I ECONOMIC FACTORS I
CONSTRAINTS
: PROGRAM EFFECTIVENESS
. PROCEDURES
.
COSTS
. ENVIRONMENT
.
TOTAL SYSTD,' COST
"TJ
......
1.0
C
,
m
N
I
-...I
c;,
QI
VI
......
c.TI n
V')
~
VI
c-+
m
3
"TJ
,
QI
3
m
:e:
o
,
7':"
INSPECTION TYPE
PASS/FAIL CRITERIA
MAINTENANCE TYPE
INSPECTION INTERVAL
:~$:t~A~:f((tY::
CAPITAL EQUIPMENT
LAND
USER INCONVENIENCE
OPERATIONS
MATERIAL
:t:aAiNiNtt
TIME
DEMOG RAPH IC
:ttlMATib
AHf~q4.:klr:t
. CONFIGURATION
.
. TECHNOLOGY
.
RESULTANT E~.1)SSICN
LEVEL
INSPECTION FACILITY - NUMBER,
LOCATION, SIZE, DESIGN
MAINTENANCE FACILITY. NUMBE R,
LOCATION
INFORMATION STORAGE/RETRIEVAL.
tti$G.~:ON.Atj':A:~ti1#:*iNG.:
USE R TIME
INSPECTION TIME
MAINTENANCE TIME
STATE.OF-THE-ART
:t:Pf:tEt::A$T$
. BENEFITS . SOCIAL
IIIIII~~II~'IIII~~II ::¥~~:!~~~:~~NN~~~~~G
MODEl
-------
Inspection/Maintenance Procedures
This study considers the program impact on the vehicle owner as
well as the extent to which exhaust emissions can be controlled. To
reduce the impact of a mandatory inspection/maintenance program on the
vehicle owner, minimum cost and minimum time maintenance procedures were
considered. From a control and enforcement point of view, this is also
desirable, since it:
. Protects the vehicle owner by minimizing his
financial obligations
. Specifies easily implemented procedures com-
patible with existing commercial practice
. Provides greater assurance that a vehicle's
emissions will be reduced, since only those
maintenance treatments shown to be effective
in reducing emissions are considered.
The approaches studied for the inspection procedures were:
. Direct diagnosis of engine parameters using
.conventional or more sophisticated garage-type
equipment. This approach is well suited to
inspection by franchised-garages.
Inferences of malfunction from measurement of
engine exhaust emission signatures. This approach
is best suited for use in state inspection lanes.
Maintenance strategies of two types were defined. Thesestrategies
involve either IIpredeterminedli maintenance or lIadaptiveli maintenance.
.
With predetermined maintenance, those components and subsystems which
influence emissions substantially were first identified. The best inspec-
tion procedures for identifying failures or maladjustments of these
components and subsystems were then selected. Vehicles failing an inspec-
tion based upon a rejection criteria for each maladjustment are then sent
to maintenance where only the specified components are corrected regardless
of the state of the remainder of the engine and emission control systems.
This approach is most effective where there is a high probability that the
inspection procedure correctly identifies the malfunction and where the
cost of inspection is approximately equal to that of the specified main-
tenance activity. As an example, an idle-emission measurement screening
6
-------
test might be coupled with specified maintenance procedures for restoring
failed vehicles to manufacturers' specifications with respect to idle rpm,
idle air-to-fuel ratio and idle timing. Some of the rejected vehicles
will receive an idle parameter adjustment even though an engine parameter
outside of those considered in the maintenance program will have caused
the vehicle to be rejected.
An adaptive maintenance program is based on diagnostic data acquired
during the initial inspection supplemented by those obtained during re-
inspection at the maintenance station. All of these data are then used to
identify components which require maintenance. Thus, the probability of
making a correct diagnosis is high, but at the expense of higher inspec-
tion costs. The diagnosis of maladjustments then controls the extent of
maintenance performed although a limit is placed upon the engine sub-
systems which are treated.
The study objective is to determine the extent to which it is economic-
ally feasible to transfer detailed diagnostic activities previously associated
with maintenance to the inspection process. In Table 2-1, generalized
procedures are shown in which various levels of diagnosis are transferred
from the maintenance to the inspection activity. Vehicle power train
screening tests require obtaining a minimum amount of diagnostic informa-
tion and implies either further and complete diagnosis at the garage or a
predetermined maintenance policy. Vehicle subsystem diagnosis and detailed
component diagnosis (possibly preceded by a screening test) place increased
emphasis upon obtaining diagnostic information during the inspection
process.
Vehicle power train exhaust emission screening tests are typified by
the idle and New Jersey ACID cycle (Reference 1) tests. In these tests
only minimal diagnostic data on actual failures of the emission control
system are obtained with the vehicle being rejected for maintenance based
solely upon abnormally high emissions. Diagnosis of actual maladjustments
and failures must be done at the maintenance station where the mechanic's
skill is important to the effectiveness of the inspection/maintenance pro-
gram in reducing emissions.
7
-------
Table 2-1.
Generalized Inspection/Maintenance Procedures
I NSPECT ION
MAINTENANCE
TYPICAL INSPECTION PROCEDURE
I OfT AILED . CAPE-14 SHORT EMISSION
- PASS1 SCRET~~~NG I . N.J. ACID CYCLE
FAIL DIAGNOSIS INDICATED . CHU I DLE MODE
. SUBSYSTEM MAl NTE NANCE
. COMPONENTS
CP . CAPE-14 LONG EMISSION
COMPLETE . CLAYTON KEY MODE
- PASS- SUBSYSTEM FAIL DIAGNOSIS OF SUBSYSTEM . CAPE-15 3 MIN. ENGINE
DIAGNOSIS SUBSYSTEM(S) MAINTENANCE PARAMETER
FAILE D
SCREENING ~ FAIL
TEST
DETAILED
COMPONENT FAIL
DIAG NO SI S
INDICATED
MAINTENANCE
. CAPE-15 30 MIN. ENGINE
PARAMETER
. CHU IDLE MODE +
MULTIMODE AND/OR ENGINE
PARAMETER
. CARB DIAGNOSTIC
PASS
PASS
I
I
I
INTERFACE
-------
With subsystem inspections an attempt is made to group components or
adjustments into logical functional combinations such that a malfunction
of any of the components of a subsystem can be determined by measurement
at a single sensing point. The Clayton key mode procedure (Reference 2),
as well as the short and long emission inspection procedures which are to
be defined under this study fall into this classification. Only those
subsystems which have failed undergo detailed diagnosis during the main-
tenance procedure.
A detailed diagnosis during the inspection process completely elim-
inates diagnostic decisions during maintenance by providing a specific
repair list. Because complex and possibly automated instrument systems
are required to perform the detailed diagnosis in a cost effective manner,
it may be advantageous to perform a screening test to limit the number of
cases sent to detailed diagnosis as is shown in Table 2-1.
As the inspection process is changed from screening to detailed
component diagnosis, inspection equipment and labor skill become in-
creasingly important. The probabi1i.ty of a correct diagnosis and repair
is increased at the expense of significantly higher inspection costs and
user inconvenience. Inspection/maintenance processes which fall into the
three generalized procedures categories just described were evaluated in
this study.
Inspection Station Siting
To estimate the capital equipment and facilities requirements, a
number of demographic characteristics are needed for the area in which an
inspection/maintenance system is to be implemented. It is expected that
the preponderance of program cost will be incurred in the metropolitan
areas of regions under consideration, and that the design of equipment
and facilities will be dictated by the requirements of these urban centers.
Because the study's primary objective is to select the best procedures
from those available, the relative effectiveness of procedures is of
primary concern. Therefore, it is felt to be important to evaluate pro-
cedures within a consistent demographic framework rather than to explore
all ramifications of regional demography. For these reasons we selected
the demographic features of the Los Angeles Basic for the study.
9
-------
2.2 ECONOMIC FACTORS
Benefits in reduced emissions obtained from an inspection/maintenance
program must be compared with cost to arrive at the program's overall
economic effectiveness. The relevant costs are the explicit and implicit
costs of such a program. Explicit costs involve resources expendi-
tures to construct facilities as well as the cost of inspection/maintenance
operations. Implicit costs, for example, reflect a cost assignable to the
time a user spends in inspection and maintenance-related activities.
While the design of an economically effective program must consider both
implicit and explicit costs, the costs which are quoted for various pro-
gram alternatives are based on explicit costs, since only these result in
a direct monetary outlay. Station location and configuration design how-
ever were determined by' including both costs.
2.3 CONSTRAINTS
Several restrictions were placed on systems to be considered in the
study. For example, this study considers exhaust emission inspection
procedures which are oriented toward malfunction diagnosis rather than to
emission measurements which correlate well with those obtained using the
Federal emission test procedure. Procedures were to be defined to diagnose
maladjustments and component failures which would result in both high HC & CO
composite emissions. The effect of these policies on NOx emissions was
estimated after the fact rather than consider~d in selecting procedures be-
cause the 1966-1970 vehicles studied were not equipped with NOx control
devices. A further consideration imposed during the selection of inspection/
maintenance procedures was that a substantial reduction be effected on at
least one of the major pollutants (i.e., 15% or greater) regardless of the
cost effectiveness of a particular procedure or strategy.
2.4 PROGRAM EFFECTIVENESS
Measures of system performance included actual emission reductions
achieved, user time lost in traveling and waiting during inspection
and maintenance, and the effectiveness of the inspection process (prob-
ability of a correct diagnosis).. These performance factors were combined
to yield an estimate of the emission reduction over a four year period as
well as system cost. A figure of system merit was formulated from
these elements and optimized by selection of the system design variables.
10
-------
3.
PROCEDURES EVALUATION
The following approach was used to evaluate the effectiveness of
mandatory maintenance of vehicle emission control systems:
. A data base was acquired from which basic emission
and cost models were developed
. Systems tradeoff studies were performed using an
analytical computer model.
3. 1
DATA ACQUISITION
Data pertinent to the evaluation of inspection/maintenance procedures
were sought to describe the following characteristics quantiatively:
. Frequency and extent of malfunctions and maladjust-
ments as they occur in vehicles in a general user
population
. Vehicle exhaust emission sensitivity to selected
engine malfunctions and maladjustments
. Emission diagnostic mode sensitivity to malfunctions
and maladjustments '
. Inspection and maintenance costs
. Deterioration rates of emission sensitive
engine and control device variables.
Experimental data.were developed as indicated in the flow diagram
of Fi gure 3-1.
3. 1 . 1
Frequency and Extent of Malfunction
Malfunction and maladjustment data from commercial diagnostic
centers were obtained and evaluated. These data were found to be
unsatisfactory for study purposes for two reasons:
. Quantitative measurements were generally not provided
(i.e., a component or adjustment was either characterized
as acceptable or as unacceptable)
. Failure criteria were not provided.
11
-------
Literature Survey
Pa ramefer- Se 1 ect ion Cri te ri a
26 Enqine parameters, P.,
identified 1
GM Idle Adjustment
-- Da ta Set
228-Vehicle test set
Mode emissions, em
Engine parameters, Pi
P.
1
Engine Parameter
Field Survey
Malfunction Frequency and Extent
244-Vehicle test set
.26 Enaine rarameters/vehic1e
P ( Pi), em
P(Pi)
Scree~ing Experiments.
Orthogonal Tests
Emission Res~nse to Malfunctions
11 Vehicles
13 Malfunctioned parameters/vehicle
33 Diagnostic modes, em
32 rmission tests/vehicle
ae.
~
ap.
Economic-effectiveness
computer model
Definitive Experiments
OrthoQonal Tests
Emission Resj)onse to Malfunctions
11 Vehicles
5 Malfunctioned parameters/vehicle
tests/vehicle
ae.
em' ~
. op.
1
Figure 3-1.
Experimental Program Flow Diagram
12
-------
Although these data provided quantitative insight into the frequency of
failures in a user vehicle population, they were inadequate for quanti-
tative analyses. Therefore, an experimental study was undertaken on a
randomly selected group of 227 vehicles to measure the frequency and
extent of malfunctions and maladjustments of components known or suspected
to have a significant effect on emissions. Table 3-1 summarizes data
obtained in this study. A more detailed description of the test program
is contained in Volume III.
Table 3-1. Summary of Extent and Frequency of Typical
Engine Parameters Measured in the Field
227 Vehicles From Parameter Survey Experiment
Estimate Estimate of
Parameter of mean Standard % Un-
- Deviation satisfactory
X S
~ Timing, deg relative to
specification 0.9 5.8 76
Vacuum Advance, deg. 19.4 6.0 33
% Available Voltage coil avail. , 303 133 12
plug reqd.
Misfire Rate at 30 mph, % 12.0 0.0 2.7
45 mph, % 15.3 5.4 4.4
60 mph, % 17.6 8.4 9.7
ICO, % 3.72 1.90 60
l:!. I RPM (rpm), relative to
specification -9 94 70
~ Float Level (in.) relative
to specification 0.02 0.22 34
AIC Restriction, deg. 40 50 23
PCV Pr:ssure, in. H20 -0.3 0.5 12
Air Pump Pressure, psi 4.1 2.2 11
Odometer Reading, mi. 33528 18643
13
-------
Diagnostic equipment and procedures typical of a well equipped diagnos-
tic center were used to characterize the maintenance states of these in-use
vehicles. Of primary interest to an emission inspection/maintenance program
are those parameters which deteriorate or malfunction frequently and which
result in emissions increases. The mean deviations from the manufacturer's
specification or a zero reference point as well as the standard error about
.
this mean are shown in Table 3-1 for the ten most emission-sensitive and most
frequently maladjusted engine parameters. Idle fuel-to-air ratio, timing,
idle rpm, secondary ignition misfire, air pump failure, air cleaner restriction,
positive crankcase ventilation valve restriction, float level, and percent
available voltage were selected as the most important parameters for further
study.
3.1.2
Emission Sensitivity to Malfunction
The influence of the parameters identified in Table 3-1 upon exhaust
emissions was first estimated from published data. Again it was found
that quantitative data describing the effects of different levels of mal-
function and maladjustment on vehicle emissions were inadequately described
in the literature.Where available, the data generally described the effect
upon emissions of combinations of maintenance such as replacement of
plugs, points and condensor (References 4, 5 and 6). Single mal-
function effects were difficult to extract. Where single mal-
functions were studied, data were available for a limited number of
power trains and for only a few variables (References 7, 8 and
9). Therefore a test program involving two statistically designed
subexperiments was conceived and conducted to obtain data describing
exhaust emission responses to malfunctions. . These subexperiments,
.. ..'" . .. .,... .
described in Figure 3-1, were:
14
-------
. Screening experiment of parameters found
frequently to be out of adjustment to determine
those which significantly affect emissions
. Definitive experiment to determine quantita-
tively with a high confidence level the
relationships between emissions and mal-
functions for those malfunctions indicated by
the screening experiment to be most important.
Screening Experiment
The ten parameters listed in Table 3-1 plus manifold vacuum leakage
(typified by disconnected or punctured vacuum hoses and associated
accessories) were evaluated in a screening experiment. The objectives of
this experiment were to:
. Identify the most important of these malfunctions
. Validate the effectiveness of statistically
designed experiments in defining emissions
response to engine parameters
. Identify the most discriminating diagnostic
emission modes.
A statistically designed, orthogonal experiment was performed to develop
linear mathematical relationships (response surfaces) between exhaust
emissions and engine parameters. The statistical design selected was a
1/64 fractional factorial experiment performed at two settings for each
variable. This design allowed for the clear derivations of all main
effects as well as an indication of interactions effects.
Both mass and composite emissions were measured simultaneously using
a modified, variable dilution mass sampling technique. In addition,
emissions were measured for 33 trial diagnostic modes encompassing known
and proposed exhaust emission inspection procedures. HC, CO, C02' NO and
N02 measurements were made for all modes/cycles.
The test sequence was as follows:
. An emissions test was performed on the as-received
vehicle
. A major engine tune up was done and the emission
tests performed.
15
\""._,-,,-- ",,-
-------
. The 11 malfunctions/maladjustments were set
following the experiment design and emission
tests were performed after each set of adjust-
ments.
An emission test as used in the above test description is defined as com-
bined mass, composite and diagnostic mode emission measurements. Only hot
composite and mass measurements were made during the statistically designed
experiment.
Eleven basic power trains were tested which were representative of
1966-1969 California vehicles having engine modification and air injection
emission control systems. The power trains were selected to reflect the
U.S. population with regard to engine type and manufacture.
Exhaust emissions sensitivity to malfunction was found to be similar
between power trains and the coefficients determined in this study were
highly correlated with those found in the literature for those adjustment
effects which were statistically significant. When combined with the para-
meter survey data, these results were used to develop an index of overall
effectiveness. This index represents the emission reductions which would
be expected when a specific parameter (Pi) is restored to nominal or
specification value in all of the vehicles in a general population.
de.
E1ij = (Pi- Pspec) ~
where
E1.. = average emissions reduction of emission j due to
1J maintenance of parameter i.
Pi- Pspec = ~verage deviation of parameter i measured
1n the parameter survey.
dej/dPi = response of emission j to parameter i as
measured in the screening experiment.
Composite effectiveness indices (E1) developed by weighting the 11 power
trains for the 11 parameters evaluated in the screening experiment are
shown in Table 3-2 with the significant parameters shaded. Those para-
meters shown to have potentially significant influences when kept in
adjustment through a mandatory vehicle inspection and maintenance program
16
-------
Emission
HC,ppm
co, ~~
-....J
NO,ppm
Table 3-2.
Most. Probable Emission Reduction (Combined Data
from Engine Parameter and Screening Experiments)
Vacuum
Advance
2.26
-6.3t
*
E1
= (P. - P ) 8ejap.
1 spec 1
*
Composite value obtained by a population weighting of 11 basic
power trains.
**
PCV, positive crankcase vent system.
T
INegative values indicate an emissions increase upon maintenance.
t
-0.03'
7.7
lRPi1
i 1 J. 5
! -0.02
I
I
I
i
I
9.J
-------
are the idle adjustments (rpm, timing, fue1-to-air ratio), misfire and
induction system components (positive crankcase ventilation system, air
injection system and air cleaner restrictions). These malfunctions and
adjustments, with the exception of misfire, were selected for definitive
study within a more statistically powerful experiment. Misfire was
omitted as it had no indicated interaction with the other parameters and
was adequately characterized. Although float level and vacuum advance
have strong effects on CO and NO emissions, their malfunction frequencies
were such that they would not be expected to have significant impact on
total emissions to the atmosphere.
Thirty-three diagnostic modes were evaluated in the screening orth-
ogonal experiment. Exhaust emission responses to changes in the selected
engine parameters were obtained for each of the diagnostic modes. Modes
were sought which were highly diagnostic of either one class of ma1fu~ction
(engine subsystem) or of specific components or adjustments.
Typical emissions responses to the range of engine malfunctions simu-
lated in the ~xperiment are shown in Tables 3-3 and 3-4. Modes are grouped
by the level of engine loading (closed throttle, low, medium and high load).
Certain malfunctions such as the restriction of the reactor air pump, PCV
valve and air cleaner were set at their extremes (plugged). In the field
they will be in varying states of deterioration. Therefore, emission
responses to these parameters generally would be expected to be less than
those shown. The following conclusions were drawn from these results:
. Exhaust emissions measured in modes which load
the engine at similar levels tend to have
similar response to malfunctions.
. CO emissions measured using modes involving low
to moderate engine load levels are sensitive to
induction system type malfunctions, such as float
setting, air leak, air cleaner and PCV. Basic
timing however can sometimes confuse the diagnosis.
. For air reactor controlled vehicles, a malfunction
of this device can confuse a CO emission diagnosis
of either the idle parameter maladjustments in the
closed throttle modes or induction system mal-
functions at load.
18
-------
Table 3-3. Typical Vehicle Malfunction Sensitivity to
Operating Mode--HC Emissions
Screening Experiment, 90% Confidence Level
~e, Change in Emissions in Going From the Low to High Level of a Parameter, ppm
Mode Misfire2.5% T"' +70 RPM+50 IC02.0% +1/411 A/CRestricted pcvP1ug
o lmlng-10o -100 -0.5 F1oat-1/811 Spec Spec
* 298 154
Idle F
, **
!Idle (11) 392 237 -145 168
I
IId1e (27) 324 128 -56 79
'30- 15 F 386 389 -409
I
150-20 F 311 363 -780 228 294
130F 313 71 -70 40
1.0 124C (28) 308 64 -72 35 36
I
130A (14) 335 100 -53 46 -75 86
/+2 mph/sec (12-14) 379 148 -88 52 67
40C (29) 321 46 -47 51
10-25 F 309 110 -36
h 5-30 F 301 94 -48 27
I
10-30A (18) 369 138 -91 81 -66 92
160s (30) 359 -119 -30 37
/60P (25) 328 77 33 -44 70
/+3mPh/sec (6-8) 332 95 -27 33 28
*
1968 Federal 7 mode cycle.
**
See Vol. III, p 3-29 for corresponding mode number location.
-------
Table 3-4. Typical Vehicle Malfunction Sensitivity to
Operating Mode - CO Emissions
Screening Experiment, 90% Confidence Level
6.e, Change in Emissions in Going From the Low to High Level of a Parameter, %
* Misfire~.5% T.. +70 RPM+50 !C02. 0% +1/411 A/CRestricted PCllug
Mode lmlng-l00 -100 -0.5 Float-l/811 Spec Spec
Idle F 2.2958 0.5525
Idle 11 2.8879
Idle 27 0.3837 2.7545 0.5717
30- 15 1 .5941 0.5348 -0.4275
~0-20 1 . 7744 0.7517
BOF -0.6160
N 24C -0.9366 -0.9947 0.6846
o
30A -0.6352
"'2 mph/sec -1.3778 -0.5147 0.6572
40C -2.7870 1.4717
0-25 -0.8153 -0.9442 0.5869
15-30 -0.8910 -0.9873 0.5547
0-30A -0.7985
DOS 0.7380
DOP 0.7287 1 . 1620 0.9888
3 mph/sec -0.9795 -0.7812 0.8499
*
See comment on Table 3-3.
-------
. The fast idle modes (1500 and 2500 rpm) are not
substantially different from the closed throttle
modes in providing diagnostic information.
. Closed throttle HC emissions will generally diag-
nose maladjusted idle rpm and timing, although an
air reactor system malfunction on vehicles so
equipped will confuse the diagnosis.
. Basic timing dramatically affects NO emissions
under any engine load condition.
. The effect of misfire rates greater than 2.5% upon
HC emissions in all modes predominates over the
effect of most other malfunctions simulated.
Definitive Emissions Response Study
The six engine parameters having the largest effectiveness
indices were evaluated in a more powerful statistical experiment
(see Figure 3-1) having the objectives to:
. Develop a generalized data bank from which engine
parameters and emission signature inspection/main-
tenance models can be synthesized for the economic-
effectiveness study.
. Determine the important interactions between engine
parameters and their effect on engine exhaust emissions.
. Characterize those parameters known to have non-
linear emissions response surfaces.
In addition to the main-line experiment, the following subexperiments were
performed:
. An air cleaner deterioration test was performed
wherein the effect of three levels of restriction
was evaluated for vehicles otherwise adjusted to
manufacturer's specifications.
.
Vehicle stability tests were performed to assess base
line emission changes over the test period ,for a vehicle
set to manufacturer's specification.
A statistically designed, orthogonal experiment was performed to charac-
terize the emissions response of the six engine parameters, Table 3-2, for 11
basic power trains. The experiment was a one-half fractional factorial design
run at two levels for four parameters and at three levels for idle CO. Three
21
-------
levels of idle CO were selected because both HC and CO composite emissions
had been previously shown to have nonlinear responses to this maladjustment
(References 5 and 6).
Population weighted response surfaces, aej/aPi' for the 11 power trains
evaluated are shown in Table 3-5 for HC, CO and NO emissions. These com-
posite values reflect the average population emission sensitivity of engine
modification and air injection system controlled vehicles. The following
conclusions are indicated:
. Both engine modification and air reactor controlled
vehicles generally respond to the same degree to
engine parameter changes.
. Air reactor controlled vehicles are slightly less
sensitive than engine modification vehicles to in-
duction system malfunctions.
. Air reactor controlled vehicles, on balance, are
as sensitive as engine modification vehicles to idle
adjustments.
. All six of the engine parameters investigated signif-
icantlyaffect emissions and were selected for evalua-
tion in the economic-effectiveness study.
The experimental design also permitted the determination of parameter
interaction effects on emissions. For example, the simultaneous adjustment
of any two engine parameters may not result in an emission change which is
the sum of the effects (taken separately). An interaction may be hypothes-
ized to have occurred under this situation which may increase or decrease
the emission change. Timing and idle rpm, for example, are shown to have
interactions which affect CO emissions for air reactor controlled vehicles.
Air Cleaner Experiment
Air cleaner experiments were performed for those power trains shown in
the preceding experiment to have an emission response to air cleaner restric-
tion. The objective of these tests was to determine the nonlinear response of
emissions to air cleaner restrictions. Clogged cleaners were simulated by
taping closed the air cleaner element and then cutting vertical slits at
five equal spaces around the circumference. The remaining engine
22
-------
Table 3-5. Summary of Composite; 7-Mode Emissions Responses to
Engine Parameters - Definitive Statistical Experiment
Parameter Response, ae/aPi
Composite + * *
Vehicle Emission Timi ng Idle CO Misfire Air * Air
Control j degree Idle RPM % % Cleaner PVC Pump
~ir Injec- HC ppm 9.94 -0.525 -4 116 1.1 28 154
tion System -7xlO-4
CO % 0.009 0.196 0.82 0.35 1.54
NO ppm 30.5 0.164 -1.07 292 0 -53
N Engine Mod- HC ppm 7.25 -0.892 -18 100 29 58
w ification
CO % -0.008 -6.6xlO-4 0.176 - 0.85 0.51
NO ppm 38.3 0.063 -8.2 -114 208
*
Emission change is based on going from the nominal to a restricted state.
+Misfire results are obtained from the screening experiment.
-------
parameters were set to the manufacturer's specification.
air cleaner restriction, Figure 3-2, was measured both as
CO at 50 mph road load and with an AC air cleaner tester.
The extent of
an increase in
The exhaust emission response to air cleaner restriction as determined
from the orthogonal test is also plotted on Figure 3-2. Except for Vehicle
No. 602, the orthogonal test data points compare favorably with those
obtained during the air cleaner experiment.
Several vehicles (602, 604 and 610) exhibited nonlinear emissions
response to an air cleaner restriction. Several vehicles had responses to
air cleaner restriction which were highly unstable as indicated by data
scatter as well as by the difficulties in setting up the simulated mal-
function. Because vehicles with nonlinear air cleaner response curves also
exhibited substantial ranges of linearity, only the linear responses derived
from the definitive test were used in the economic-effectiveness study.
3.1.3 Exhaust Emission Signatures
Diagnostic modes were again evaluated using data from the definitive
orthogonal tests to determine which were most selective in identifying mal-
functions at either the subsystem or component/adjustment level. The re-
sults of this evaluation are shown in Volume III (Sections 3 and 4) for
selected modes.
The general conclusions based upon data from the screening experiment
were still found to apply. Conclusions of a more specific nature were:
. The 33 mph Clayton key-mode cycle is equivalent
to the more heavily loaded modes of the Federal
cycle in diagnostic content.
. The 50 mph Clayton key mode contains no additional
diagnostic information than was already present in
the lower loaded modes for the engine parameters
evaluated.
. The 1500 rpm fast idle mode contains no diagnostic
information not already present in the idle modes.
There may be some utility in the 50 mph Clayton mode in revealing induction
system power circuit malfunctions and verifying the results of the 33 mph
mode. It is evident that all special emission signature modes (eg, Clayton
key-mode) can be approximated using modes of the seven-mode cycle.
24
-------
I
I ,--~ "
II I, i ' I I
, ,-/1: I d..."'.Ioo ,,' .<.,)(,
... , I ,- I ~ "~,..
..... ' I I ',,- 00
... I 'Hr..If....................."" J "'6
"'" ' ""'~"""'''~'''~ , I -it- u,
I ... ........ .,',""""'-'-"""':'''''' , _.-(I , I J
.. ..........:..... . '; ':-:,~;"'":'" ..:-:, - ... - :- - -~ -~ ~ i °
-'~ : . I' ,
r ~ ' ,
#I/I!.'-_.... '. ~ ---t-- .----
, '" i
I' !, I , 609
120
0. AIR CLEANER JESTS
. ORTHOGONAL TESTS
I
I
I
I
I
b.
100
I
I
I
I
if
I
.I
80
...
...
...
....",,""1
I
I
I
I
I
~
"-
lit
C>
1
w
Q
X
o
Z
o
::E
Z
o
co
""
-------
Exclusive use of the modes of the seven-mode Federal cycle for the
emission signature study permitted the use of data from the General Motors
(GM) idle adjustment study. These data are compared with the parameter
survey data in Table 3-6. The two sets of data agree fairly well consider-
ing the acquisition source and relatively small sample. Vehicles in the
parameter survey would be expected to be further out of specification be-
cause of their higher average mileage.
The GM data were used to develop relationships between exhaust emis-
sion rejection cut points and the mean values of the engine idle parameter
settings found for vehicles in the rejected population. The parameter
survey data were used to develop equivalent relationships for inspections
based upon direct inspections of engine adjustments. It would have been
more desirable to have used a single data set for analysis of both emis-
sion signature and engine parameter inspections. However, the high correla-
tion between the GM data and that obtained during our parameter survey
eliminates the possibility of large errors.
Mean values for the idle adjustments referenced to manufacturers'
specifications for the rejected fraction of vehicles were plotted against
varying rejection levels or cut point values for single as well as multiple
emission signature measurements (see Volume II). The following results
were obtained:
.
Idle CO emissions are the most powerful single
inspection signature (i.e., identifies the
most idle maladjustments).
. The addition of an idle HC emission inspection
identifies, on the average, larger idle rpm, and
timing adjustment deviations.
. Mean values of the idle parameters for those
vehicles rejected vary approximately linearly
with the idle CO emission rejection level.
Inspection and Maintenance Costs
3.1.4
The cost and direct labor of inspection and the associated maintenance
were established from three sources. These sources, depending on whether
conventional or advanced equipment was used, were:
26
-------
Table 3-6. Comparison of Parameter Survey and General
Motors Idle Adjustment Program
Air Injection and Engine Modification Controls
Statistic
*
GM Idle Program Parameter Survey
Size 228 227
Model, Year 1966-68 1966-69
Mileage 20,000 33,500
6Rpm. -21 -9
6Timing, deg 0.3 0.9
Idle CO, % 3.37 3.72
Air Pump 6% 10%
Idle Misfire 2.2% 2.2%
Sample:
Idle Adjustments:
"1a1functions:
N
'-J
fir
See Reference 4.
-------
. Chilton's Labor Guide and Parts Manual
. Engine parameter inspection labor times derived
from the orthogonal test experiments
. Coarse operations analyses for hypothetical
advanced equipment such as remote sensing devices.
The Chilton data was used to estimate labor costs for diagnoses
using conventional garage equipment. Judgment was required in separ-
ating the diagnosis times from actual maintenance times, since the
procedures investigated in this study are usually not specifically
evaluated in the flat rate manuals.
All labor costs for garage inspection/maintenance are charged at a rate
of $10 per hour. This is a burdened rate which includes overhead factors
and profit. Labor rates for state inspection stations consist of direct
labor at $3.50 per hour and overhead factor of 50% to account for fringe
benefits and general administration. An additional 50 cents per vehicle
is charged for program enforcement (i.e., the information system required
for recording, processing, storing and disseminating inspection/maintenance
data ).
Additional ancillary data, such as demographic constants, user incon-
venience costs and average trip speed are based upon the Los Angeles region
which was the reference point for the economic-effectiveness study. All of
the above data are described in greater detail in Volumes II and III.
Investment Costs
Regional land costs were considered in the estimation of investment
costs associated with station siting. National land cost data acquired
by TRW on a program of vehicle safety compliance performed for the
Bureau of Highways was used to evaluate land costs for rural, urban and
metropolitan regions (Reference 10). Equipment costs for diagnostic and
emissions testing were acquired from a number of vendor sources. In addi-
tion, cost and performance data for prospective, prototype emissions instru-
ments and automated data acquisition systems were developed.
A coarse operations analysis of procedures and work
tions was conducted to support the costing of inspection
layout of an inspection station is shown in Figure 3-3.
station configura-
lanes. A typical
28
-------
N
1.0
l,,~-
,
I
rY/
,...~~.
. ..'~
"-':, )-
--
ADMINSTRATIVE OFFICE
.-
,. ~
,:; L) -~.
.'. .. .,. .....J
...---JI ~ - ---
~-..
\-
Figure 3-3.
Typical Station Configuration for State Inspection
-------
An evaluation of prototype models of advanced emissions instruments
was undertaken to select those generic types most suitable to state in-
spection lanes. Precision, accuracy and cost data pointed toward non-
dispersive infrared devices as those best suited to our requirements.
Scope of Voluntary Maintenance
It is important to establish the extent and frequency of voluntary
tuneup maintenance to assess its current cost and effectiveness. An
equally important consideration is that enforced maintenance will probably
only require performing a portion (idle adjustments) of the total required
vehicle maintenance. Under these circumstances, it seems reasonable to
assume that the remainder of the required mafntenance will be performed
voluntarily at a frequency equal to that of current voluntary maintenance.
Data on voluntary maintenance are scant, usually conflicting, and vague.
For example, the often-quoted 1965 Look magazine survey indicated only "34%
of the vehicle owners had tune up work performed within a l2-month interval.
A 1966 poll of the service industry indicated perhaps 88% of the nation's
vehicles were tuned during a 12-month interval (Reference 11). Recent TRW
data based on an employee poll suggest that 79% of 1966-1969 vehicles under-
go an annual tuneup.
Data from a recent surveillance program on 1966-1969 vehicles conducted'
for Standard Oil involving a random selection of California vehicles
(Reference 12) indicate a tuneup maintenance frequency of 12 to 13 months.
A 12-month frequency was selected for the study to simplify calculational
procedures and because it represented a reasonable estimate of maintenance
frequency for a newer vehicle population in which many cars are still under
warranty.
3.1.5 Deterioration of Maintenance
Estimates of the rate at which tuneup adjustments deteriorate are also
essential to evaluate alternate maintenance treatments. Analysis of the
American Automobile Associaton (AAA) data (Reference 13) suggests that toa
first approximation, degradation models must be developed for the follow-
ing three subsystems:
. Idle adjustments of fuel-to-air ratio, idle rpm,
and timing which affect all emissions (HC, CO, and
NO) .
30
-------
. Induction subsystem. (PCV, air cleaner, and air pump)
which predominantly influences CO and NO emissions.
. Secondary ignition subsystem (misfire) which affects
only HC emissions.
The deterioration of most components and adjustments except timing generally
results in reducing NO emissions. For example, rich carburetion and re-
x
tarded timing result in lower peak combustion temperatures and, therefore,
lower NOx emissions.
NO emission deterioration rates were derived based on the high correla-
x
tion which was found between CO and NOx composite emissions using the
previously discussed air cleaner experiment data and the estimated rate
of deterioration of basic timing.
Deterioration data for CO and HC were developed for each of the major
engine subsystems-induction (PCV, air pump, carburation, manifold leak),
ignition (misfire), and idle adjustment (rpm, timing, and CO) using a combin-
ation of AAA and California Air Resources Board (ARB) data.
The rate of deterioration of engine, parameters was combined with
current, voluntary maintenance program data (frequency and effectiveness),
to reconstruct the emissions degradation shown by the California ARB surveil-
lance data. This was done assuming that average voluntary maintenance occurs
every 12,000 miles on 1966 - 1969 vehicles. The effectiveness of voluntary
maintenance was obtained from Reference 4. This set of data was adjusted
through iteration until the ARB surveillance data for composite HC and CO
were matched. The following deterioration rates expressed as a function of
mileage were derived.
31
-------
Table 3-7.
Subsystem Deterioration Rates
Deteri orati on Ra te t 8 e . /8 M t /.2
gms ml
J
Power Trai n Subsys tern
HC CO NO
7 x 10-6 -5 -5.7x10-6
Air Injection Induction 8.7 x 10
Ignition 2.3 x 10-6 0 0
9.3 x 10-6 -4 10-5
Idle 3.5 x 10
Engine Modifica- Induction 8 x 10-6 1.3 x 10-4 -8.6 x 10-6
tion
Ignition 2.3 x 10-5 0 0
1.2x10-5 -4 10-5
Idle 5.6 x 10
3.2 ECONOMIC-EFFECTIVENESS MODEL
An economic-effectiveness model was then developed to provide a method
for systematically examining the impact of various strategic (inspection-
lane or franchised-garage) and tactical (inspection/maintenance procedures)
program policies on system costs and resultant emission levels. The model
developed is capable of analyzing the effectiveness of various emission
inspection test techniques including the four alternatives to be studied on
the project (i.e.t engine parametert idle emissions and key mode). The
model calculates the relative effectiveness and cost of state-lane or fran-
chised-garage inspection operations. In additiont it can treat cases where
the re~ired maintenance treatment is either adaptive or predeterminedt that
iSt cases where either the maintenance treatments are selected on the basis
of a reinspection or specified in advance for all rejected vehicles. Volume
II presents a detailed discussion of the economic-effectiveness model.
32
-------
The general ground rules for the economic effectiveness analyses
were as follows:
. The economic-effectiveness of a particular inspection/
maintenance approach to vehicle emission control is
related to specified emission reduction goals and
economic constraints. For this study a reduction of
at least 15% for any of the emission species was
required of each maintenance strategy with an attendant
program cost not to exceed $6 per car per inspection/
maintenance interval.
. The two basic emission control approaches (air reactor
and engine modification) were treated in the model as
separate populations.
. The specific vehicle population evaluated was that of
Los Angeles with an estimated population of 4 million
emission controlled. vehicles.
. The effects of new vehicles entering and leaving the
population due to attrition and new production was
ignored.
. A 4-year performance period was assumed.
. Mean values of the emissions for both the engine para-
meter and emission signature populations were allowed to
vary with time and with the degree of maintenance, how-
ever, the variances about these means were assumed to be
time invariant.
. Emissions changes due to engine parameter deteriora-
tion were considered at the subsystem level (i.e.,
induction, ignition and idle adjustment subsystems).
. Basic maintenance was assumed to be performed voluntarily
by the vehicle owner for those subsystems not treated by
the enforced maintenance procedure.
The economic effectiveness model (Figure 3-4) consists of three major elements
. Emissions predictor model
o Cost estimator model
. Operations model
The emissions prediction model describes in detail the process involved in pre
dicting emission time histories and total integrated average emissions
based on the fraction of vehicles rejected to maintenance, emission
33
-------
INITIAL
CON DITIONS
AND
BASIC DATA
EMISSIONS PREDICTOR MODel r------:----'
I I COST ESTIMATOR MO Del
. REJECTION CRITERIA 10PERATIONS MODEl~
. AVERAGE EMISSION lEVELS I. FACILITIES . I . CAPITAL COSTS
. DECREMENT AND DECAY I.QUEUING I . OPERATING COSTS
.INTEGRATED AVERAGES I. TRAVel TIME ~ . USER TIME COSTS
~ I
I
L.._---------J
BASE
EMISSION
CASE
EB
ET FIGURE OF MERIT $
-
$
l:W. M:'
1 1
Figure 3-4.
Systems Model Flow Diagram
34
-------
reduction effected by maintenance and the maintenance deterioration with
vehicle use. The effectiveness of candidate inspection/maintenance pro-
cedures was measured by comparing the resultant emission levels to some
reference or base line case. For the present model, the base line case
was the situation where vehicles are maintained voluntarily by the owner.
Cost to implement a specific inspection/maintenance program includes: (1)
capital costs, (2) operating costs and (3) user inconvenience cost. A
figure of merit for each alternative inspection/maintenance program was
determined using the emission reduction between the test and base line
case as well as the total program cost.
3.2.1
Emission Predictor Model (Figure 3-5t
The model is configured to predict average hydrocarbon, carbon mon-
oxide and oxides of nitrogen emission levels for time-invariant base line
and test fleet populations. Figure 3-5 shows the procedures utilized by
the model for determining test population emissions over the life of the
program. The model first determines the average emission levels for each
emission (HC, CO and NO) using experimental data. Next, an inspection
procedure divides the test population into two parts: (1) a subpopulation
of accepted vehicles with low emission levels; and (2) a subpopulation of
rejected vehicles with high emission levels. The model then computes
emission levels for the two populations at the time of the next inspection.
For the accepted subpopulation, the emission levels are assumed to decay at
their preinspection rates. For the rejected subpopulation, maintenance is
performed which results in a measurable, average change i~ emissions.
Using these new emission levels as a base, the model then decays the exhaust
emission levels for this rejected subpopulation using a predetermined rate.
Once the decayed emission levels have been obtained for the two subpopulations,
the model combines them to compute the total average emission level for each
emission component.
35
-------
W
0\
REJECTION
FRACTION, r
INSPECTION
Figure 3-5.
ACCEPTED -
AVERAGE (E) A
MAINTENANCE
AE
DECAY TO
I NTERV AL
END
Emission Predictor Model Flow Diagram
TOTAL TEST
POPULATION
EMISSIONS
WEIGHTED
POPULATION
AVERAGE E
DECAY TO
INTERVAL
END
-------
,-
3.2.2 Cost Estimator Model (Figure 3-6~
Inspection, maintenance and user-time costs comprise the main elements
of the cost estimator model. The model estimates capital and direct
operating costs for both the inspection and maintenance processes. All
capital costs are discounted over a specified period and are added to
the projected direct operating costs to obtain total annual operating costs.
Total program costs over the time period considered are computed as the
sum of these annual operating costs, discounted to present value. These
costs are quoted exclusive of the user inconvenience costs which are included
in the evaluation of alternative inspection/maintenance processes.
3.2.3 Operations Model
The operations model provides the linkage between the emission pre-
dictor model and the cost estimator model. It determines size, location,
and number of facilities required to maintain the program. In making these
determinations it trades off user-inconvenience costs with the capital costs
for facilities. Thus, given any system configuration, the model compu~es
the average waiting time using a queueing algorithm. Driving times to and
from the station locations are calculated from demographic considerations
(i.e., number of vehicles per square mile). This information is used to
determine the inconvenience cost the public must bear. The operations
model analyzes the implications of a labor-dominated versus an equipment-
dominated operation. Through these considerations, it then selects the
optimal degree of automation.
3.2.4 Figure of Merit
A description has been given in the manner in which the model predicts
vehicle population emission levels and relevant system costs based on a
specified inspection/maintenance program. To measure the cost/effective-
ness of the various candidate programs, a weighted figure of merit is
applied.
Payoff, $/ton = Program Cost
EW. liE .
1 Cl
where W. = weighting factor for the "i II emission
1
lIEci = emission difference between base and inspection/maintenance
test population over four-year period.
37
-------
w
00
FOR
EACH
YEAR
Figure 3-6.
INSPECTION
TOTAL
CAP. COSTS
TOTAL
ANNUAL
COSTS
TOTAL
OP. COSTS
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
,..---------~ I
I I
I I -------...1
II NCONVEN IENCEI----- -------- ------------ ------------
I I
L..------- --.I
MAINTENANCE
CAP:
OP:
CAPIT Al
OPERATING
Cost Estimator Model Flow Diagram
-------
l.\ E . = f 48 ,000 e. dM
Cl 0 1
where
e. = "i II emission, grams/mile
1
"M = accumulated vehicle mileage, miles
The weighting factors~ Wi' were selected to reflect the fractional
emission reductions which will be achieved upon meeting the 1975
Federal vehicle emissions standards (Reference 14). These weight-
ing factors are 0.1,0.6 and 0.3 for CO, HC and NO, respectively.
Integration limits are from program initiation to 48,000 average
accumulated vehicle miles, thus reflecting a four-year program time
frame.
39
-------
4.
ECONOMIC EFFECTIVENESS STUDY RESULTS
Using the systems model and the developed emissions data, an assess-
ment was made of the relative economic effectiveness of the several pro-
posed inspection/maintenance strategies. Two general strategies were
evaluated within the framework of the systems model: (l) an engine para-
meter inspection, followed, as necessary, by specified parameter mainten-
ance; and (2) a mode emission signature analysis leading to further diag-
nosis and corrective maintenance. Each of these two approaches contains a
number of substrategies and tactics. To evaluate effectively the impact
of the more significant substrategies on program effectiveness, a computer
matrix simulation approach was used. The technique permits evaluation of
the degree to which the figure of merit (F/M) is sensitive to the design
variables.
Shown in Table 4-1 for both the engine parameter diagnostics and emis-
sion signature analysis strategies are the figure of merit, cost per in-
spection/maintenance per car and optimized emission reduction for each of
the substrategies evaluated. The figure of merit used throughout this study
is defined as the present net worth cost of the program divided by the four-
year emission reduction effected. An assessment of the "bestll alternative
within a given strategy can easily be made by comparing the corresponding
figures of merit and selecting the one with the smallest numerical value.
Annual program cost per vehicle and the attained emission reduction per-
centages have also been listed so that a convenient comparison can be made
with results from other studies.
Referring to Table 4-1, the average cost of $1 to $3 per vehicle for
the idle adjustment program is not inconsistent with the $3 per car esti-
mated during the GM adjustment program and the $6 per car for the New Jersey
inspection program (Reference 4). The corresponding idle adjustment emis-
sion reduction for hydrocarbons and carbon monoxide approximates the results
from the GM experiment (Reference 4) when their results are adjusted for
maintenance deterioration effects.
The simulation indicated that a short duration parameter inspection
performed in a state inspection lane provides the most cost-effective system
with a figure of merit of 320. The state-lane inspection procedure modeled
40
-------
Table 4-1.
Summary of the More Cost Effective
Inspection/Maintenance Procedure
~
Cost Per Emission Reduction***
Figure of Merit Vehicle HC CO NO-
Procedure ($/Ton) ($) (%) (%) (%)
i ' I '
Engine Parameter Diagnosis I I I
I
, ! I ,
1. Idle (State Lane) 320 1.50 0 15 -7
2. Idle (Franchised) 370 2.50 3 13 -3
*
3. Extensive A (Franchised) 460 6.00 18 14 0
**
4. Extensive B (Franchised) 540 12.00 22 33 -5
Emission Signature Analysis
5. Idle (State Lane) 430 2.50 2 12 -4
*
6. Extensive A (State Lane) 360 4.00 11 16 -4
**
7. Extensive B (State Lane) 410 6.00 15 20 -3
*
Idle Plus Ignition Subsystem Inspection.
'1.*
Idle Plus Ignition Plus Induction Subsystem.
***Average emission reduction over a four-year period.
-------
consisted of remotely sensing idle carbon monoxide concentration and idle
rpm. This inspection/maintenance procedure resulted in an average 15% re-
duction of carbon monoxide emissions over a four-year period following
initiation of a mandatory program. Hydrocarbon emissions were unchanged
while oxides of nitrogen emissions increased 7%. The prorated cost of this
inspection/maintenance procedure was estimated to be $1.50 per vehicle with
the cost of program administration being about one third of the total.
A short duration idle parameter inspection procedure also was evaluated
for use in franchised garages with conventional diagnostic equipment. This
procedure produced emissions changes which were similar to those resulting
from the short duration state inspection lane. procedure just described. The
cost of short duration inspection in franchised garages however was higher
than that of the similar' state-lane inspection and therefore the franchised
garage inspection procedure was found to be less economically effective with
a figure of merit of 370.
The simulation demonstrated that idle adjustment inspection/maintenance
was more cost effective than either the mixed idle plus ignition procedure
or the complete idle, ignition and induction maintenance which had figures
of merit of 460 and 540, respectively. The study indicated that the more
comprehensive inspection/maintenance procedures will result in emission
reductions of as much as 33%, however, at significantly higher program costs
(average owner annual cost of $12.00). Sections 4.1 and 4.2 present more
detailed results on the two major strategies studies.
Conclusions regarding the cost effectiveness of various inspection/
maintenance strategies are affected by the weighting factors assig~ed to
individual emissions. In the figures of merit just presented a reduction
of hydrocarbon emissions was weighted six times more heavily than an
equivalent reduction of carbon monoxide emission. Nevertheless the selected
optimum strategy is really only effective in reducing carbon monoxide
emissions. In areas having photochemical "smog" problems the figure of
merit should probably be reweighted to further increase the importance of
HC and NOx reductions. Clearly maintenance of the ignition subsystem and
even the induction subsystem will become important if hydrocarbon emissions
are to be controlled effectively.
42
-------
The equipment required for state-lane inspections always includes HC
and CO emission instruments and a simple dynamometer when loaded mode
emission signatures are required. Inspection in franchised garages requires
conventional diagnostic equipment and optionally, a simple dynamometer. A
more detailed list of equipment required for the procedures evaluated can
be found in Table 4-2.
4. 1
Engine Parameter Inspection
Of the two strategies evaluated, the engine parameter inspection
approach minimizes the number of diagnostic errors because maladjustments
and malfunctions are directly identified and maintenance is performed only
on the engine parameters shown to effect emissions significantly. Therefore
for the identical vehicle rejection fractions, this procedure would be
expected to effect larger emission reductions than an emission signature
inspection.
Emission time histories for the most cost-effective program, an idle
adjustment program, using 1966-1970 vehicles are presented in Figure 4-1.
The base line fleet case reflects voluntary maintenance. Only the idle
adjustment engine parameters are inspected annually and maintained in the
inspected fleet. Depicted in panels A, Band C of Figure 4-1 are model-
generated, time history plots of hydrocarbons (HC), carbon monoxide (CO),
and oxides of nitrogen (NOx) emissions. This idle adjustment program has
its largest impact on carbon monoxide emission levels and its least, on
oxides of nitrogen emission levels. The data are based on a one-year in-
spection/maintenance cycle over a four-year program life.
In addition to an idle adjustment program, other parameter inspection/
maintenance strategies were assessed (see Table 4-1). In these cases, the
overall figures of merit were found to be considerably poorer than that of
the idle adjustment program. Although these more elaborate procedures
substantially reduce HC emission levels below those produced by the idle
adjustment program, these reductions are more than offset by the increased
program operational costs. This higher cost is primarily associated with
the more difficult and lengthy inspections which must be made to find igni-
tion and induction system malfunctions. For example, 15 minutes of inspection
time on 100% of the vehicles are required to find those 3 to 4% which have
engines that misfire under load. This increased cost may be worth the
improved air quality in regions of chronic air pollution and large vehicle
populations.
43
-------
Table 4-2.
Equipment and Procedures Required for Diagnosing
Engine Parameter Malfunctions
~
~
Engine Parameter Emission Signature
Subsystem
Equipment Procedure Equi pment Procedure
Idle
- Rpm Tachometer Idle Rpm NDIR HC Analyzer Idle HC
- Timing Timi ng 1 i ght Basic timing NDIR HC. Analyzer Idle HC
- Fuel-to-Air NDIR ca Analyzer Idle ca NDIR ca Analyzer Idle ca
Igni t"j on-Mi sfi re Engine electronic Misfire at NDIR HC Analyzer/ 45-Mph HC
Analyzer/Dynamo- 45-Mph road- dynamometer
meter load
Induction
- PCV Pressure gage Idle crankcase NDIR ca Analyzer/ 45-Mph ca
pressure dynamometer
- Air cleaner AC air cleaner Pressure drop NDIR ca Analyzer/ 45-Mph ca
tester across element dynamometer
- Air reactor NDIR ca/ca2 Idl e dil uti on NDIR ca Analyzer/ ca Connected -
Analyzer correction dynamometer ca Disconnected
-------
5
UJ 4
-J
~
~ 3
<'
Z
0 2
VI
VI
~
W
u
:r:
0
40
UJ
....1
~
~ 30
<'
,
Z 20
o
VI
VI
~ 10
w
0
U 0
w 8
-J
~
~ 6
<'
Z 4
o
VI
VI
~ 2
o
Z
n
ANNUAL IDLE INSPECTION - ADJUSTMENT PROGRAM
PANEL A
--- ~----
~ ---
~
PANEL B
~ ~
-----. ~-----..
,.,-- -----....
PANEL C
----
- -----. -------
-----.
--.
2
T IME ~YEARS
3
co
NO
4
INSPECTION REJECTION CRITERIA
RPM~-75
TIMING ~ 7 DEGREES
ICO > 4%
HC
BASE LINE FLEET
---- INSPECTED FLEET
Figure 4-1. Emission Time History for Engine Parameter
Inspection - Idle Adjustment Maintenance
45
-------
4.2 Mode Emission Signature Analysis
The mode emission signature strategy generally results in low operat-
ing costs and relative ease of program implementation. On the other hand,
the re-occuring problem of vehicle inspection errors (omission and com-
mission) reduces the effectiveness of this procedure. That is, some
vehicles will be rejected with failures which are not part of the mandatory
maintenance program (errors of commission). In other
instances, vehicles will not be identified although they had repairable
malfunctions (errors of omission). To 'minimize the effect of this de-
ficiency and yet provide low program costs, emission inspection modes
must be selected judiciously for their ability to identify specific engine
parameter maladjustments. For this study, four emission modes--idle CO
and idle HC as well as HC and CO emissions under engine load--were selected
for use in an emission screening inspection.
The idle CO emission inspection/maintenance strategy is not as cost
effective as the idle parameter inspection/maintenance approach (figure
of merit of 430 as compared to 320 for the idle parameter inspection).
The negligible HC emissions reduction along with the larger increase in NO
emissions for the idle CO emission inspection results from the fact that the
emission inspection does not diagnose deviate ti~.ing as accurately as does
the engine parameter inspection. For the same reason, NO emissions are not
changed even when an idle HC emissions inspection mode is added. The
addition of this mode, however, does result in rejecting vehicles with
larger average deviations in idle rpm, and thus, produces a substantial
reductio~ in HC emissions. Smaller CO emission reductions result because
vehicles are being rejected with lower average values of idle CO.
The combined idle modes and the loaded HC mode emission inspection
is extremely cost effective and offers substantial emission reductions
(see Figure 4-2); 11% and 16% for HC and CO, respectively. The low figure
of merit results from the high weighting factor (0.6) on HC reduction and
significantly shorter time for inspection (one additional minute) than the
conventional scope diagnosis. The lower emissions reductions relative to
engine parameter inspections are due to the large number of errors of
omission. For example, approximately 50% of the air reactor and air
46
-------
5
~
"
0 4
l
In
~
w 3
>
w
~
Z
0 2
In
In
~
W
U
J:
0
~ 40
o
l
In 30
~
w
>
w
~ 20
Z
o
In
In 10
~
UJ
o
U
~. 8
tY
l
In 6
...J
UJ
>
UJ
~ 4
Z
0
In 2
In
~
UJ
0
Z
PANEL A HC
~--------- .------- ~-------
..... -
-.
.
BASE II NE
-
------IDlE PLUS IGNITION
---IDLE PLUS IGNITION PLUS INDUCTION
o
PANel B CO
~ --
!!...------- .----.-
~ - . ---.....
-
PANEL C NO
-----~-- ---.---- --_:._---
o
o
2 3
TIME, YEARS
4
5
Figure 4-2.
Emission Time Histories - Annual Emission Signature
Inspection Using Loaded Modes
47
-------
cleaner malfunctions go undetected. Also, idle HCemission is a poor
discriminator of timing deviation. Approximately 50% of the advanced
timing deviations remaining undiscovered.
The emission time history profiles for the two most cost effective
emission inspection/maintenance procedures are presented in panels A, B
and C of Figure 4-2. The general trend of these plots is much the same
as those exhibited in Figure 4-1. The lower hydrocarbon levels achieved
using the emission signature inspection approach is attributed to the fact
that substantial reductions are possible when misfiring ignition sUbsystems
are repaired. This failure was not cost-effective to repair in the state-
lane idle parameter inspection strategy.
48
-------
4.3 Conclusions and Recommendations
The previous discussion indicates that the cost effectiveness of the
evaluated procedures tends to be strongly influenced by the inspection pro-
cess, whereas the magnitude of the possible emissions reductions are most
sensitive to the extent of imposed maintenance. Lowest figures of merit
are estimated for the engine parameter inspection with idle adjustment
maintenance and for the emission signature approach with idle adjustment
maintenance plus ignition subsystem repair. Based on the stated assumptions
and data of this study, the following inspection/maintenance procedures are
most attractive:
. Engine idle parameter inspection and maintenance per-
formed by franchised garages.
. Emission signature inspection in a state inspection
lane followed by maintenance of the idle and ignition
subsystems within a franchised garage.
These two cases provide sufficient latitude to satisfy the needs of
varying state air quality programs.. A state with a chronic photochemical
"smog" problem may choose to maximize the reduction of all emissions; a
state which is primarily concerned with CO emissions might be satisfied
with only an idle adjustment program. Also, less populated states may not
wish to impose a state-lane inspection with its attendant startup and user
inconvenience costs.
These conclusions are based upon using the 1968 Federal test procedure
as the best means of predicting vehicle exhaust emissions to the atmosphere.
The weighting factors used for each of the emission species were selected
on the basis of a region with a photochemical "smog" air pollution problem and
having a large population of controlled vehicles such as the Los Angeles
Basin.
49
-------
Due to the limitations imposed by the above assumptions, by the rapidly
changing control system technology, and by the change in Federal emissions
measurement procedure from a volume to mass emissions basis, it is recom-
mended that the basic emissions related data in the economic-effectiveness
model be improved and expanded.
The objectives of future work would be to strengthen deficiencies in
the economic-effectiveness model by acquiring and introducing more defini-
tive experimental data to:
. Describe more accurately' the frequency and extent of
engine and control device malfunctions and their asso-
ciated emission signatures for vehicles in several
regional populations. '
. Predict the effects of maintenance on emission reduc-
tions for precontro11ed and California 1971 NO con-
trolled vehicles. x
. Describe the degradation of engine parameters with
time and the associated change in mass emissions as
measured with the 1972 Federal procedure.
. Characterize the effectiveness with which specified
maintenance can be performed by a commercial service
organization.
The upgraded model would then be available to determine the effectiveness of
inspection/maintenance as a strategy which can be used to meet regional
ambient air quality.
50
-------
9.
I'
I
I
10.
13.
oJ
I' - 14.
I
I;
, '
REFERENCES
1.
John N. Pattison, Clark Fegraus, A. J. Andreatch, and John C. Elston,
"New Jersey's Rapid Inspection Procedures for Vehicular Emissions,"
SAE Paper No. 680111.
2.
C. L. Cline and L. Tinkham, "A Realistic Vehicle Emission Inspection
System," APCA Paper No. 68-152.
TRW Proposal to the CRC, Sales No. 11608.000, "Vehicle Emissions Pro-
gram, II September 1968.
3.
4.
G. W. Dickinson, H. M. Ildvad, and R. J. Bergin, "Tune-Up Inspection
a Continuing Emission Control," Proving Ground Section General Motors
Corporation, SAE Paper No. 690141, January 13-17, 1969.
5.
M. P. Sweeney and M. L. Brubacher, "Exhaust Hydrocarbon Measurement
for Tune-Up Diagnosis," SAE Paper No. 660105, January 10-14, 1966.
G. C. Hass, D. R. Olson, John N. Pattison, M. P. Sweeney, and M
Miles Brubacher, et.al., "Preliminary Report on the Effect of Commer-
cial Tune-Ups on Automotive Exhaust Emissions."
6.
7.
M. M. Roensch, "Exhaust Emission Control - Maintenance Versus Inspec-
tion," No. 68-150, General Motors Corp., 61st Annual Meeting APCA,
June 23-27, 1968.
8.
"Variation in Automotive Exhaust Emission Versus Engine Adjustment
Variables, Three 1969 Model Automobiles," California Air Resources
Board, Project M-190, February 1969.
"Variation in Automotive Exhaust Emissions Versus Engine Adjustment
Variables for Air Injection and Engine Modification Controls,"
California Air Resources Laboratory, Project Report M-166,
November 12, 1967.
11.
"Automated Diagnostic Systems - Vehicle Inspection," TRW Report No.
09793-6002-ROOO, Final Report - Phase I on Contract FH-11-6538 for
the National Highway Safety Bureau.
Ernst and Ernst, "A Study of Selected Hydrocarbon Emission Controls,"
for NEW, July 1969.
12.
"Effectiveness of F-310 in Reducing Vehicle Exhaust Emissions,"
August 18, 1970, Technical Brief presented at Los Angeles by Chevron
Research Company.
L. J. Bintz, "Automotive Fleet Emission Program, II Automobile Club of
Southern California Report, June 15, 1968.
D. S. Barth,et a1., "Federal Motor Vehicle Emission Goals for CO, HC,
and NOx Based on Des i red Air Qual i ty Levels. II
51
------- |