Automobile Emission Control -
Technological Approaches Toward Improving
In-Use Vehicle Emissions Performance
Emission Control Technology Division
Ann Arbor, Michigan
September, 1975

Section 1
1.1 Background
The Environmental Protection Agency has been monitoring exhaust
emission levels from in-use vehicles since 1971 through the Emission
Factors Program (EFP).
While the EFP has been designed to provide input of exhaust emis-
sion levels for air quality modeling, the program has indicated that a
potential problem exists with in-use vehicle emissions performance when
these in-use vehicles are tested in the as-received state of tune-up and,
maintenance condition. These levels have been in excess of the Federal/
Emission standards. Other EPA Surveillance Programs have shown that in-
use vehicles when properly maintained and tuned to manufacturer's speci-
fications are in substantial compliance with the Federal Emission Standards
which they were designed to meet. The purpose of this study was to
ascertain from currently available data sources the cause(s) for such
substantial in-use differences in emission-related state-of-tune and to
explore and evaluate potential technical solutions.
Data from the 1975 EPA Emission Factor Program have indicated that
a substantial number of the 1975 model year low mileage vehicles were
failing to meet the 1975 standards of 1.5 HC, 15 CO, 3.1 NOx for one or
more pollutants.
A summary of the 1975 Emissions Factors data available when this
report was prepared is shown in figures 1-1 to 1-3 for the low mileage
1975 vehicles.
The technological approaches to improving in-use performance of
vehicles discussed in this report are divided into two general classes,
those that involve technological changes to the vehicle, and those that
involve technological changes in areas other than to the vehicle, for
example maintenance practices.
EPA continues to look at the entire spectrum of approaches toward
improving in-use vehicle emission performance, including inspection and
maintenance (I/M) programs, short test programs, assembly line testing,
in-use compliance (recall) programs, retrofit programs, and technical
assistance to the various states and Air Quality regions where auto-
mobile emissions are a serious problem. However, these subjects are no
treated in this report. This report deals only with technological

approaches toward improving in-use vehicle emission performance, and
therefore should be considered as only part of the treatment on the in-
use vehicle emission performance issue. The scope of this report has
been limited in order to focus in more depth on technical approaches
that the automobile manufacturers could investigate in order to improve
the emission performance of in-use vehicles. Because of the complexity
of the problems, and the paucity of data on causes and responsibilities
for vehicle malperformance, this report has relied extensively on engi-
neering judgement and assumptions.
1.2	Purpose of this Report
This report intended to serve as a basis of discussion concerning
the in-use vehicle emissions performance issue, and ways to make that
performance better.
1.3	Data Sources and Nomenclature
The primary sources of data used for the preparation of this report
c&me from reports of EPA-sponsored work that involved testing of in-use
automobiles. These and other data sources are referenced as they
appear in the discussion in the text.
The nomenclature used for emissions in this report is a triplet
abbreviation, in which the dimensions of grams per vehicle are assumed
arid the emission test procedure is the 1975 Federal Test Procedure
(FTP). Thus, 0.6 HC, 7.0 CO, 1.2 NOx means 0.6 grams per mile hydro-
carbons (HC), 7.0 grams per mile carbon monoxide (CO) and 1.2 grams per
mile oxides of nitrogen (NOx), all measured on the 1975 FTP.
Fuel Economy is described in terms of miles per gallon (MPG). The
urban or "city" fuel economy is represented by MPG^, the non-urban or
"highway" fuel economy is represented by MPG„, and the composite fuel
economy, determined by a 55/45 urban, non-urban mileage weighting, is
represented by MPG .

Composite Mean HC Emissions for 1975 Model Year Vehicles
Tested in Five* Low Altitude Sites of the FY74
Emission Factor Program
5.0 —
4.0 -
1975 FTP
3.0 -
i	2.0
1975 Emission Levels
1.0 _
US Mfgs.
Foreign Mfgs.
*The mean emissions represent data from Phoenix, Chicago, Houston^ St. Louis, and
Washington, D.C.
Figure 1-1

Composite Mean CO Emissions for 1975 Model Year Vehicles
Tested in Five* Low Altitude Sites of the FY74
|	Emission Factor Program
50.0 -
1975 FTP
1975 Emission Levels
Ford	Chrysler	AMC	US Mfgs.	Foreign Mfgs.
*The mean emissions represent data from Phoenix, Chicago, Houston, St. Louis, and
Washington, D. C.
Figure 1-2

Composite Mean NOx Emissions for 1975 Model Year Vehicles
Tested in Five* Low Altitude Sites of the FY74
Emission Factor Program
5.0 _
1975 FTP
1975 Emission Levels
Mfg.	GM	Ford	Chrysler	AMC	US Mfgs.	Foreign Mfgs.
*The mean emissions represent data from Phoenix, Chicago, Houston, St. Louis and
Washington, D.C.	'
Figure 1-3

Section 2
Discussion of The Problem
2.1	What is the Extent of the Problem?
Data from the EPA Fiscal Year 1974 Emission Factor Program (FY74
EFP) indicated that a substantial number of 1975 model year vehicles
were failing to meet the 1975 standards of 1.5 HC, 15 CO, 3.1 NOx for
one or more pollutants as illustrated in Figures 2-1 and 2-2. This is
the definition of "The Problem" as Tables 2-1 and 2-2 indicate, 66% of
all domestic vehicles tested failed one or more of the pollutant levels.
Additionally, 52% of all vehicles tested failed because of high CO
levels only or in combination with other high pollutant levels. Tables
2-3 and 2-4,which detail the results of the FY74 EFP for Los Angeles,
shows the trend for CO to be somewhat lower in Los Angeles partly at-
tributable to lower emission standards, different technology, and pos-
sibly the effect of more stringent enforcement of state air pollution
control laws.
Other data taken.both by EPA and independent sources, substantiate
the fact that 1975 model year vehicles may be exhibiting emission levels
higher than the standards. While these data are generated principally
by tests other than the FTP, usually idle HC and CO measurement, they do
again indicate that the CO levels at idle are high. Table 2-5 gives
some indication of these levels identified by idle test methods. If all
manufacturers had specifications for idle CO, then the idle CO measurements
made on in-use vehicles could be used to compare to the manufacturer's
specifications in crder to determine if the vehicles are set to manu-
facturer's specifications. Unfortunately, most manufacturers do not
have idle CO specifications, the major exception being Chrysler, with a
specification of 0.3% CO in the exhaust upstream of the catalyst at
idle. In order to estimate the degree of idle maladjustment, it has
been assumed that a value of idle CO <0.6% would correspond to properly
adjusted model year 1975 vehicles. The data in Table 2-5 would seem to
indicate that average in-use vehicles have idle CO levels 2 to 3 times
higher than the assumed specification.
2.2	What are Some of the Possible Causes of the Problem?
2.2.1 Idle CO Maladjustment
It has been well established that CO failure under FTP conditions
is primarily a direct result of the air-fuel ratio the vehicle operates
at during the test cycle. While the idle portion of the Federal Driving
Cycle is less than 20% of the total cycle, the vehicle operation during
the idle portion has a great influence on whether it passes the FTP or
not. A rich idle air-fuel ratio enhances the possibility of the vehicle
not passing the FTP. The idle mixture screws are a convenient means of
enrichening the idle mixture.

Figure 2-1
Hydrocarbon Emissions of In-Use Vehicles at 50,000 miles

Figure 2-2
Carbon Monoxide Emissions of In-Use Vehicles at 50,000 miles

Table 2-1
Composite Mean Emissions for the 1975
Model Year Vehicles Tested in Five*
Low Altitude Sites of the FY74
Emission Factor Program
1975 FTP (gm/ml)
Number	HC	CO	NOx
Manufacturer	Tested	Mean	Mean	Mean
77 -
US Mfg.
Foreign Mfg.
Overall Mfg.
*The mean emissions represent data
St. Louis, and Washington, D.C.
from Phoenix
, Chicago,

Table 2-2
Number of 1975 Vehicles That Fail the Federal Standards, by Standard,
for the Cities of Chicago, Houston, St. Louis, and Washington,
and Phoenix
Number of Vehicles That Fail

Total Number

All Three
Percent Failure
U.S. Mfg
Foreign Mfg
Overall Mfg 587	371
15 125 58 112	11	21	29

Table 2-3

Mean 1975 FTP Emissions in Gm/mi
for the 1975 Vehicles Tested
in the Los Angeles FY74 EFP






US Mfg

Foreign Mfg.

Overall Mfg.


Table 2-4
The Percent of 1975 Vehicles that Failed the Federal

in Los Angeles FY74 EFP


Percent Failing


HC or CO
HC, CO, or NOx
US Mfg
Foreign Mfg
Overall Mfg

Table 2-5
Idle Emission Levels
for 1975 Model Year Vehicles*
Data Source
Gulf Oil Corp.
Champion Spark
Plug Co.
EPA-EFP(4 cities)
Sample Size
Mean Idle Mean Idle
CO Level(%) HC Level(ppm)
*Vehicles which failed any one or a combination of pollutants.
**MSED-Mobile Source Enforcement Division, EPA.

In 1967, the Ford Motor Company introduced idle limiter caps and by
1971 practically all vehicles were equipped with these caps with the
intention of limiting the amount of adjustment of the idle mixture
screws. As evidenced by Tables 2-6 and 2-7, the number of idle limiter
caps removed for 1975 model year vehicles ranges from 13 to 34 percent
depending on location. Therefore, idle limiter caps are probably not
effective. Data from EPA's Mobile Source Enforcement Division, (MSED)
also indicates that as vehicles get older the percentage of vehicles
without limiter caps increase; see Table 2-8.
Unfortunately, data does not exist which identifies who removes the
limiter caps; why they were removed (though it is hypothesized that it
is in response to a owner driveability complaint), or if they were
installed originally. Even the latter question cannot be answered with
certainty, but caps are generally installed by the carburetor manu-
facturers after the carburetor is flow checked and before it is sent to
the vehicle manufacturer. The Ford Motor Company has told CARB that
limiter caps are functionally checked 100% during the vehicle manu-
facturing process*.
2.2.2 Other Maladjusted Parameters
The causes of high emission levels of 18 Chrysler vehicles in the
EFP which had emission levels above the standards and idle CO below 0.5%
were investigated in detail. CO remained the primary emission problem
with these 18 vehicles. Because the Bag 1 mean emission levels remained
high in relation to correctly operating vehicles, this suggested that
the reason for high CO emissions could be either the malfunction of the
electric choke, choke vacuum brake, retarded spark timing, or any com-
bination of these in conjunction with some minor idle mixture malad-
justment. While analysis of Bag 1 results do not prove choke malad-
justment, choke malfunction or maladjustments are still considered to be
the most likely as the consumer would have little incentive to greatly
retard ignition timing.
Though it would be tempting to ascribe the poor emission perfor-
mance of in-use vehicles solely to idle maladjustment, other emission
system components cannot be overlooked. NAS in their June 1973 report**,
identified 61 emission control system components which could have a
deleterious effect on exhaust emissions. Informal discussions with
representatives of both the Clayton Manufacturing Corporation and Champion
Spark Plug Company surfaced other components which could have adverse
~Letter from D. A. Jensen, Ford Motor Company to William H. Lewis,
CARB; June 4, 1976.
**Feasibility of Meeting the 1975-1976 Exhaust Emission Standards in
Actual Use, National Academy of Sciences, June 1973, pp. 59-61.

Table 2-6
Mobile Source Enforcement Division-EPA
Tampering Study Limited to Domestic Vehicles Only
Where Conducted Clnn. Chic.	N.J.
When conducted 3-4/75 9-10/75	11-12/75
Vehicle year '75 only '75 only	'75 '76
Number in sample 319 152	181 37
Percent limiter caps removed 15.05 13.8	26.5 12.8
Mean idle CO level w/caps (%) .56 .60	.7 .2
Mean idle CO level w/o caps (%) 1.4 1.9	1.7 3.2
Mean idle HC level w/caps (ppm) 66 76	96 47
Mean idle HC level w/o caps (ppm) 135 140	107 322

Table 2-7
Frequency of Idle CO Maladjustment (Percent) for 1975 Vehicles
Data Source
Champion Spark
Plug Company
Emission Factors
Sampling Size
364 (Chicago, Houston,
St. Louis, Washington,DC)
115 (LA & Phoenix)
Frequency of idle
Maladjustment from Spec.
cannot be determined
Idle Limiter caps
removed (percent)
Not reported
Not reported
34 (Phoenix only)
Specification was idle CO greater than 2.00%
Specification was idle CO greater than 1.50%

Table 2-8
Results of Mobile Source Enforcement
Dlvlslon-EPA Tampering Study*
Conducted in NJ for 1974-75
Percentage of No Limiter Caps
Model Year	1974 (%)	1975 (%)
1970	76.4	76.1
1971	46.9	67.9
1972	37.0	70.2
1973	27.2	59.2
1974	5.9	45.5
1975	-	26.5
1976	-	12.8
Avg.	46.6	56.6
*The Incidence of Tampering on Cars in New Jersey During 1975, MSED,
June 22, 1976.

effects on emissions. Champion stated that possibly up to 25% of the
vacuum spark advance diaphragms in the field could be faulty. Clayton
confirmed approximately the same levels cf vacuum advance failures, plus
vacuum hose deterioration and EGR valve sticking; but ascribed no
percentages to the latter two failures. Clayton (a manufacturer and
distributor of vehicle diagnostic data) also indicated that much of the
test equipment used for tune-up, diagnostics, and repair in the field
was faulty or miscalibrated.
Data showing the frequency of failure for system components other
than idle adjustments is more limited than idle setting data. Two
pieces of data exists which could indicate the magnitude of the problems
with other systems. These data are found in Tables 2-7 and 2-8.
Table 2-9
Frequency of Failure for Selected Components
Source Champion Spark Plug Company
All Model Years
System/Component	Sample Size	Freq. of Failure(%)
Points 3877	21.0
Cap & Rotor 4625	5.1
Ignition leads 4626	18.7
Spark Plugs 4626	31.7
PCV 4624	4.6
EGR 1247	4.0
Belts (Replaced) 4609	14.0
Hoses 4528	10.8
Air Filter 4369	34.5
The MSED in their New Jersey Tampering Study also	developed a
frequency of tampering rate for the above systems.
Table 2-10
New Jersey Tampering Study
, Frequency of Tampering(%)
Sample size-1935,(all) Sample size-230, 175—
Sys tem/Component		model years		model years	
Limiter Caps	56.6	23
PCV	0.7
Vacuum Spark Retard Mechanism	2.2	1.1
AIR	0.2	-
EGR	0.4	0.5

While Tables 2-9 and 2-10 provide only circumstantial evidence of
their system frequency of failure rates, these system do have a marked
effect on vehicle exhaust emission when they fail.
The effects of the maladjustment of other parameters and component
failures is a complicated subject. The misadjustment or failure of
nearly every emission control component on a vehicle can affect its
These effects can obscure the identification of any one single
parameter as the cause of high in-use vehicle emissions. The reason for
not being able to identify the problem from the available data is that
the status of all of the components on the vehicle is usually not
determined accurately.
2.2.3 Driveability
Driveability is a highly subjective indicator of vehicle performance.
The definitions associated with driveability (hard starting, hesitation,
stall, stumble, sag, rough idle, etc.) usually are not uniform in percep-
tion from driver to driver. The automobile manufacturers on the other
hand, have trained evaluators who drive rate the vehicles during develop-
ment of the emission control systems. For some systems driveability may
be traded off against exhaust emission levels and/or fuel economy.
The only extensive owner driveability data developed by other than
an automotive manufacturer was done by Champion and is given below.
Table 2-11
Owner Performance Report for All Model Year Vehicles
Hard starting
Rough Idle
Misfire at Hwy
% of

% of

Driver Satisfaction
Well Satisfied
Could be Better
Total Reports
1618	34.93%
2507	54.21%
426	9.21%
4551	98.40%

It can be seen that most drivers, 63% would like their vehicles to
drive better. Also note particularly, 43.7% of the drivers complained
of rough idle for their vehicles. Rough idle is a potential reason for
maladjustment of the idle circuit.
2.2.4 Inadequacy of the Service Industry
As early as 1973, NAS wrote, "The service industry at the present
time is not adequate to service the 1975-1976 cars from an emission-
control standpoint." There Is no reason to challenge or to believe that
the situation has changed drastically since the NAS report was written.
The size of the service industry may provide some indication of the
problems faced by the industry. Mr. Herb Fuhrman of the National
Institute for Automotive Service Excellence (NIASE) provided the follow-
ing breakdown of personnel. A recent government census estimated that
there are 805,000 people who were classified as or called themselves
automotive mechanics. Subtracted from this number are 265,000 people
who are working for governmental bodies such as Federal, State, and
local levels and other public bodies such as Public Works, Police, Fire
Departments. Also subtracted are 90,000 managers, instructors, and
supervisors who do not actually work on the vehicles. There are ap-
proximately 450,000 individuals who actually service vehicles or ap-
proximately one mechanic per 266 in-use vehicles.
Table 2-12 shows the problem mechanics face with servicing 266
vehicles each. This table indicates the number of times maintenance is
to be performed over 50,000 miles, as instructed in the owners mannual
furnished by the domestic manufacturers.
The fact that mechanics seemingly have a formidable task to per-
form, may be part of the explanation of poor in-use emission performance.
While the role of the mechanic is indeed a critical one in assuring good
vehicle emissions, driveability and performance, the subject of mechanic
performance has not been addressed in this report.

Table 2-12
Operation	AMC	Chry.
Replace Air Filter
Lubicate Choke Linkage
Replace Ignition Wires
Check Distributor Advance Mechanisms
Replace Rotor and Cap
Replace Points and Condensor
Adjust Ignition Timing
Check Belt Tension
Replace Fuel Filter
Check Heat Riser Valve
Adjust Idle Speed and Mixture
Repl. PCV Valve and Check Hoses
Replace Spark Plugs
Insp. Thermostatic Air Filter
Check Vapor Recovery System
Check EGR Valve
Check OSAC Valve
Adjust Valves
Replace Coolant
Check Decel Valve Operation
Change Engine Oil	10	10
Change Engine Oil Filter	10	5
a,b= Ford Motor Co. uses two maintenance schedules. Each vehicle has a permanent
sticker affixed to the inside of the glove compartment door stating which
schedule to follow.
Labor rate figured at	Parts
$ll/hr. Figures are	Labor
5 year/50,000 mile	Total
* = six cylinder engines only.
$135	$210
200	390
335	500
$145	$135
270 280
$67/yr $120/yr $87/yr $112/yr $83/yr $85/yr

Section 3
What Are Some Possible Approaches
Toward Improving In-Use Vehicle Emission Performance?
This section deals with technological approaches toward improving
in-use vehicle emission performance. It was not the intention of this
study to examine and propose regulatory or other actions to achieve
implementation of the technological approaches. In some instances the
following text suggests consideration of a regulatory approach, such as
consideration of a performance specification and test, in order to
follow the technological approach to one logical conclusion. It should
not be interpreted by the reader to be e. recommendation for a particular
regulatory action. The important regulatory issues have not been
examined, including applicability of the Clean Air Act, impact on cur-
rent regulations, cost-effectiveness, air quality impact and considera-
tion of alternate approaches. For example, in the case of performance
specifications and tests it is questionable whether the Clean Air Act
currently provides the Administrator with authority to impose such
The technological approaches and means of implementation are
discussed in this report with the intention of focusing the attention of
industry and government on the real problem of in-use vehicle emissions
performance. Hopefully, these and other yet-to-be-determined tech-
nological approaches will provide valuable means of improving the in-use
vehicle performance.
3.1 Vehicle Technology Improvements
There are three types of technological approaches discussed below.
The three approaches are: a) uprating current technology, b) redesign
of current approaches to reduce potential for maladjustment, and c)
advanced technology that would be designed and developed with improved
in-use emission performance as a high priority item.
Uprating Current Technology
This category basically involves improving the performance/reliability
characteristics of current components. To suggest that there may be
components on current vehicles that could use some improvement may be
considered by the industry to be an unwarranted attack on their current

products. However, a distinction must be made here. A component could
provide performance considered adequate to the customer, but be less
than adequate as far as emissions are considered. For example, an
ignition system that produced intermittent misfire that was either
undetected by the driver or considered to be acceptable to the driver,
could result in large HC penalties.
The above example incorporates a principle that must be considered
when studying the in-use vehicle emission problem. The principle is
that a vehicle can currently operate over a wide range of calibrations
and that the emissions can vary widely over these calibrations. For
most, current systems the range of calibrations over which the emissions
are acceptable is narrower than the range over which the vehicle can
operate. This is true for the calibrating of spark timing, air/fuel
ratio, EGR rate, choke vacuum break and choke heat-up rate, and for the
control parameters like air injection rate and catalyst efficiency.
The fact that vehicles can operate in a manner that would be accept-
abel to the public in general, while having high emissions indicates
that components and systems may have to have better than just perceptibly
acceptable performance in order to keep the emissions low.
It is from the above perspective that the uprating is considered.
Factors that can influence emissions may be so subtle, in terms of
driver perception that new levels of performance for components may be
necessary, so that in addition to just "running o.k." the vehicle also
produces acceptable levels of emissions.
The uprated ignition system components considered as having potential
for emission performance gains are improved spark advance diaphragms,
improved ignition wires, improved spark plug to ignition wire connectors,
and improved spark plugs.
It may be that up to 20% of the vacuum advance diaphragms in the
field are faulty. It is rot known if these failures are due to faulty
designs, or if they are due to tampering (i.e. changing from a ported
vacuum signal to full manifold vacuum to the distributor all the time -
possibly overloading the diaphragm).
If this really is a problem in the field (it would take a large
program to prove it as is the case with nearly every in-use issue) then
EPA or industry could consider establishing a specific performance
specification and test for advance diaphragms. An accelerated vacuum
cycling test coupled with aging tests may suffice. If the degree of
failure of these diaphragms can be significantly reduced, then some
benefits could be inferred. This is especially true if a failed diaphragm
is the initiating cause for vehicle tampering in an attempt to restore
performance and fuel economy. One way to tamper could be to advance the
entire distributor to restore some of the ignition timing advance. This
would result in potentially negative emissions consequences and possibly
other problems like detonation.

The Champion survey data indicate that 63% of all the vehicles
tested had malfunctioning spark plug wires. If this trend continues
for the 1975 and 1976 models serious effects on HC emissions and catalyst
durability may result. In order to correct this potential problem, com-
ponent materials with superior high temperature tolerance could be used.
If the problems are in the carbon conductor material itself a change
could be made to matallic wire conductors, although this would probably
require a shielded cable for RF supression.
The ignition cable/spark plug connectors are also suspect. It is
entirely possible (and has happened) that when spark plugs are changed
ignition performance is not improved, but degraded. This happens because
mechanics yank on the wire (often the only practical means to remove), not
the connector, when they detach the wires to change plugs. This can
destroy the electrical integrity of the junction. A connector of the
bayonet type which would provide a positive connection and would be
difficult to destroy by yanking on the wire could solve this problem,
which may be widespread. A performance standard for connector integrity
as a consequence of pulling off the vrire from the plug for example, 50
or 100 lbs pull, could help solve this problem.
Spark plugs themselves could be improved to yield longer life, and
reduce the frequency of replacement. Replacing the plugs as discussed
above can sometimes result in worse ignition system performance. With
the types of ignition systems on today's vehicles, coupled with unleaded
fuel, spark plug change intervals have extended. GM for example has
spark plug change intervals of 22,500 miles on most of their vehicles.
Longer spark plug life is not impossible. One manufacturer ran one
engine family for 1975 certification for 50,000 miles without a plug
change. EPA could possibly consider setting a performance standard for
plug life to ensure that better plugs are used. This suggestion is
expected to result in adverse impact on spark plug suppliers, since the
replacement market represents the bulk of the sales and any reduction in
replacement frequency might result in a large reduction in total sales
The uprated hoses are the next category of uprated technology.
Current vehicles have a large number of hoses that provide vacuum control
signals and provide for gas flow. The performance of some of these
hoses is suspect. It has been indicated, for example, that some vehicles
sold in California had EGR hoses of such poor quality that a large
percentage of new vehicles had split hoses. Besides the integrity of
the hose material itself, the way in which the hoses are attached to
fittings has been mentioned by some to be inadequate. Many of the hoses
are just pushed on to a straight fitting with no clamp or other positive
method cf attachment. EPA could consider setting performance standards
for hose material i.e. ozone resistance, bursting strength, fatigue
resistance, etc. Additionally, a performance standard could be con-
sidered for pull-off force at connecting points to ensure that positive
connections are used.

Uprated air injection systems are next. Vane-type air pumps generally
tend to wear in such a way that the air delivered at a given KPM decreases
with time. A performance specification on maximum decrease in delivered
flow might be considered, which could be met by redesign or using an
inlet air filter which some have indicated is needed.
Uprated modulating devices are the next category. On today's
vehicles, there are many devices that modulate the emission control
system. The performance of some of these devices is suspect.
As an example, in some instances it has been found that the thermal
vacuum switch (TVS) of the type used by GM and others is performing
poorly in the field. Since failure of these modulating devices more
often than not results in increased emissions, EPA could consider setting
specific durability tests for any modulating device component.
Improving the performance of the catalysts used on the vehicles
could be considered. A 100% functional check of light-off time and
stabilized conversion efficiency on the vehicle could be contemplated,
with replacement of the unit if certain efficiency performance values
were not met.
PCV valve designs could be uprated to be more resistant to con-
tamination, with a goal of no replacement ever needed, possibly encouraged
by specific durability performance tests.
Last but not least, the fuel metering system could be improved.
This could be done in two ways. First, specific durability requirements
could be set for vacuum break mechanisms, choke coils, and internal
parts of the carburetor that are subject to wear. The second approach
could be to improve the fuel. It appears that detergents in fuels may
help improve carburetor performance. EPA could consider setting a
specific performance test for in-use fuels that might result in improved
detergent activity. If this would cause gasoline producers to switch
from current additives, or to add additives that they do not add now,
resistance might be expected. Of course, additives required by EPA (if
EPA can require them be added to make emission control performance
better, as EPA could require lead to be reduced for the same reason)
would have to be examined carefully to ensure that combustion products
of the additives are benign from a health effects point of view.
Re-Design of Current Approaches to Reduce Maladjustment Potential
This approach appears feasible but it is not known if it will be
effective. If the vehicles in the field have poor performance to the
extent that encourages maladjustment to provide customer satisfaction,
then the public response to this approach could be the evolution of even
more sophisticated and costly means of tampering or a substantial public
backlash. However, it is assumed here that the vehicles perform adequately
now, and that the approaches suggested here will be effective in reducing
maladjustment with concommitant emissions benefits.

Limited Adjustability
If engine and emission control system parameters can be maladjusted
they probably will be, to some extent. If reduced adjustability could
be designed into the system maladjustments could be significantly reduced.
Basic Timing Adjustment
Many vehicles (48% according to Champion) have timing out of specifi-
cations, defined as more than + 1 degree from the specified value.
The basic timing of an engine can be adjusted by using a simple
open-end wrench by nearly anyone. In just a few minutes, basic timing
can be altered drastically. Generally, timing is advanced when tampering
is done. To prevent this, it is recommended that metal stops be incor-
porated into the the distributor mounting that would prevent the dis-
tributor from being rotated more than the minimum essential travel
required to adjust for production tolerances. These tolerances could be
quite tight for today's systems, since for pointless ignitions the
position of the cam follower cannot change.
Idle-Air Fuel Mixture Adjustment
A significant part of the in-use emissions performance problem may
be the result of incorrect carburetor adjustments. Several of which can
be made in the field. These adjustments, while potentially correcting
problems with driveability and performance, have an adverse effect on
emissions and/or fuel economy. On most of today's carburetors, many of
these can usually be made by the turn of a screwdriver or the slight
bending of a linkage rod. The vehicle owner or any mechanic can easily
make the adjustments to attempt to improve the performance of the vehicle.
Two typical carburetor maladjustments are enriched idle mixture and
modified choke operation.
The air-fuel ratio at idle is controlled by the idle mixture screws
usually located on the lower portion of the carburetor body near the
throttle plate. Backing the screws out richens the mixture by allowing
more fuel to be drawn through the idle diischarge port. There are either
one or two adjustment screws depending on the number of Venturis (barrels).
One-barrel carburetors generally have cne screw while 2- and ^—barrels
generally have two screws. On a 4-barrel carburetor, the two idle
mixture screws enter only into the primary Venturis.
In recent years, carburetor manufacturers have been installing
plastic idle limiter caps over the heads cf the mixture screws to discourage

maladjustment with the idle mixture. Unfortunately, these caps are
easily removed or defeated, which makes idle mixture adjustment almost
as simple as if there were no limiter caps at all. When limiter caps,
are removed they are destroyed, thus indicating that the idle mixture
may have been readjusted.
Attempts to improve the driveability usually involve richening the
idle mixture from the initial calibration setting. Excessive richening
will cause poor performance indicated by sluggishness and black exhaust
smoke. This richening of the mixture obviously has a definite effect on
exhaust emissions. Since excess gasoline is used, and all the air is
utilized in the burning process, unburned gasoline will remain in the
exhaust. A reasonably good indicator of the air/fuel ratio at idle is
the idle carbon monoxide concentration. In general, an increase in idle
carbon monoxide concentration will produce an increase in unburned
hydrocarbons (HC) and carbon monoxide (CO), and a decrease in nitrogen
oxides (NOx) emissions. Since all vehicles are different, the result of
richening the mixture may not be exactly the same for all vehicles, but
generally speaking there will be an increase in at least one of the
pollutants. Fuel economy is also affected when the mixture is adjusted
in that richening the mixture causes more fuel to be burned up for a
given amount of air, which can tend to decrease fuel economy.
Choke Adjustment
The choke system can also be maladjusted in the field. There are
several ways that the operation of the choke system can be altered which
would affect emissions. The choke coil rod, choke unloader, automatic
choke coil, vacuum break and electric-assist choke mechanisms are all
components of the choke system which can be adjusted.
It is suspected that choke adjustments are usually made to overcome
cold starting and cold start up and driveaway driveability problems.
Most choke systems are calibrated to a certain air-fuel ratio under cold
starting conditions. However, the "cold start" for which the choke
calibrations are the most critical for emissions is the 68°F to 86 F
start on the emission test. It is speculated that in the fall and
winter choke calibrations designed for the emission test result in less
than satisfactory starting and driveaway. The variation of driver
preferences of driveability may enhance the chances that adjustments are
made to the choke system. Many of the adjustments will richen the air-
fuel mixture delivered to the engine during choked operation and/or
increasing the time during which the choke is en.
With the electric choke mechanism, an electric heating coil is used
to shorten the choke-on time. A shorter choke-on time will cause the
mixture to lean out faster which helps to decrease emissions from a cold
engine. As indicated above, this accelerated leaning of the mixture may
result in some driveability problems under cold ambient conditions.
Disconnecting the electric choke mechanism will cause the choke valve to
open as a function of engine temperature only. The longer the choke
system is in operation, the higher the carbon monoxide (CO) emissions
because the rich warm-up mixture exists for a longer period of time.
There is also a fuel economy reduction due to the extended choke-on
t ime.

Other Carburetion Factors
There are ether factors which may contribute to poor in-use emission
performance of the carburetion system. These are the installation of
incorrect carburetors on the engine, faulty carburetor manufacturing,
and incorrect carburetor rebuilding. Each engine and vehicle combina-
tion requires a specific carburetor and calibration to meet the emission
standards. If the incorrect carburetor is installed on the engine, the
emissions as well as driveability may be affected, which may result in
adjustment of the carburetor to cure an apparent problem. There have
been instances in which incorrect carburetors have been discovered on
in-use vehicles.
Faulty or improper manufacturing of carburetors includes not only
the manufacturing done by the carburetor companies, but also the rebuilding
or remanufacturing of carburetors done by independent companies and
service stations. Missing parts and incorrect main jets, are some of the
problems that have been observed. These discrepancies could arise as a
result of possible cost savings (e.g. to standardize and reduce number
of rebuilts), insufficient stock of certain parts, or poor production
As mentioned previously, the carburetor is a major target for
controlling exhaust emissions. A desirable carburetor from the emission
control standpoint would be one which would always deliver the correct
air-fuel ratios for low emission operation and would be unable to be
maladjusted. Since a major maladjusted item is the idle mixture ad-
justment, a partial solution to the maladjustment problem would be a
carburetor with a non-adjustable idle circuit. In an attempt to develop
a non-adjustable idle circuit, several factors must be considered prior
to an initial design. These factors or variables are basically the
reasons for an adjustable idle. Production variabilities, vehicle
calibration differences and carburetor gumming are some of the major
considerations involved in the design of a non-adjustable carburetor.
Production variability is said to be an important reason for an
adjustable idle circuit. Present production technology does eliminate
most of the part-to-part difference but a small variation still exists.
A variation of a few thousandths of an inch in the diameter of a metering
rod or orifice produces a significant change in the amount of fuel
allowed through the orifice. An adjustable idle circuit compensates for
the carburetor-to-carburetor differences by allowing identical air-fuel
ratios to be obtained at idle. Eliminating these differences would
reduce the need for an idle mixture adjustment. This would mean an
improvement in production technology which will require a certain
amount of lead time and may result in increased production costs.
Since all vehicles are not identical in terms of required carburetor
calibrations, there must be a certain amount of adjustability present in
the idle circuit. Of all the different vehicle/engine/transmission

combinations available, each require calibration settings suitable for
that combination and its intended use. The same basic carburetor may be
installed on vehicles with different displacement engines, but the idle
mixture setting may be different for each combination. Besides the
vehicle differences, there are also individual driving patterns and
preferences which may justify the need for an adjustable idle circuit.
Carburetor gumming is also said to be a reason for having an idle
mixture adjustment. Gum, a product of slow oxidation, may cause operating
difficulties, such as clogging of carburetor metering jets. Also possibly
contributing toward the problem of clogged carburetor jets may be dirt
or other solid impurities in the fuel or air that pass by the filters
into the carburetor.
Freshly manufactured fuels have a relatively insignificant gum
content, but with age, varying amounts of gum may be formed. The gum
content increases with rise in temperature, increased concentrations of
oxygen, exposure to sunlight and also on contact with metals. Current
manufacturing processes have eliminated most of the gum from fuels and
many of the gasolines now contain detergent additives to prevent forma-
tion of deposits and to remove existing deposits. Even though it does
not seem to be a problem with freshly manufactured fuels, there may
exist a problem with fuels that are stored for long periods of time.
More investigation is required to determine whether or not currently
available fuel additives car reduce idle circuit gumming to a level that
would make non-adjustable idle circuits more feasible.
Limited Idle Adjustments
With present technology, some degree of adjustability of either
an idle air bleed screw or an idle fuel metering screw appears to be
required. In an attempt to approach a non-adjustable idle circuit,
limiting the amount of adjustability appears to be a worthwhile course.
This limited idle adjustability features allows for the trimming of the
idle mixture to individual engine requirements for satisfactory idle
while assuring that the exhaust emission limits will not be exceeded.
Two methods for limiting the mixture adjustment are shown in Figure 3-1.*
*Cook, F. W. "Antismog Carburetor Hardware and Test Equipment." SAE paper

In Figure 3-1 by the use of a limiting screw and a modified ad-
justment screw, the screw can be locked after the appropriate mixture
ratio has been obtained on the flow stand. The limiting screw may be
covered with a lead shot to make removal more difficult. The mixture
would not be able to be set beyond an allowable limit. In Figure 3-1
the passage into the throttle bore, below the idle mixture screw, is
sized during the flow operation to give the desired maximum air-fuel
ratio before the mixture screw is inserted. This arrangement would only
allow a ]ean adjustment at idle. Another method to limit the mixture
adjustment would be to reduce the sensitivity of the adjustment. This
can be done as shown in Figure 3-2.
By changing the shape of the idle adjusting needle, the sensitivity
would be reduced. A smaller included angle and/or more threads per
inch, would allow for finer adjustment, thus producing better control of
the idle mixture within close limits.
Limiting the adjustment and/or reducing the sensitivity to a point
where the mixture is essentially "non-adjustable" appears to be feasible,
although extensive in-use testing of candidate designs may be required.
The Separate Idle Venturi Approach
As discussed earlier, some carburetors use multiple Venturis to
perform separate functions. The "4 barrel" carburetor is an example of
different Venturis (primary and secondary) performing different jobs.
Most carburetors have the idle circuits they have because the
primary Venturis are too large in diameter to provide a good signal for
the low flows at idle. The multiple venturi approach leads to the
postulate that a separate venturi be provided for idle and near-idle
conditions only. This would be a tiny (by today's practice) venturi
with its own metering circuit. The conventional idle circuit would be
eliminated. It is not known if a separate small throttle would also be
Figure 3-2 - Fine Idle Adjustment

required, in order to handle off-idle transitions, or if acceptable
transitions could be handled by other means. If the fuel passages could
be sized close to the dimensions of current primary circuits gumming
would probably not be a severe problem. New main body carburetor castings
would probably be required, with possible lead time implications. Ethyl
developed a similar approach, which was a 3-venturi carburetor, several
years ago. However this carburetor had a idle adjustment screw, possibly
because the idle venturi could flow enough air for above 30 mph road
load conditions. In order to provide for no idle circuit, the maximum
road load flow capability would probably need to be somewhat less than
Electronic Fuel Metering System
The increased use of electronics in automobiles may allow for the
control of the fuel flow electronically in the future. Several manufacturers
are currently developing electronic fuel metering systems for use in
future vehicles. As an example, the system being developed by Chrysler
will be used.
The Chrysler Electronic Fuel Metering (EFM) System controls fuel
flow based on an air flow signal. The electrical signal that is received
from the air flow meter causes an electronic fuel control to adjust fuel
flow to a predetermined air/fuel ratio schedule. To insure that the
pump is responding accurately to the signal, a fuel flow censor measures
the flow of fuel going into the throttle body. The signal from the flow
sensor is compared to the air flow signal by the electronics and if the
proper ratio is not present, the electronics energize the pump to provide
more or less fuel until the correct ratio is obtained. Programmed ratio
changes and ambient corrections such as cold, load and acceleration
enrichments are incorporated in the electronics and are said to be
The nucleus of Chrysler's system is the precise controller that
compares the air and fuel flow and corrects it for any change in engine
requirements. For each small quantity of air measured as it enters the
air cleaner chamber, fuel is delivered to the throttle body until the
correct air-fuel ratio is obtained. Therefore, for all engine operating
conditions, fuel flow is metered depending on e.ir flow.
This system, though still in the developmental and experimental
stages, appears to show promise in terms of emission control. From the
work that has been done en electronic fuel metering, it can be concluded
that this system is potentially more accurate than a conventional carburetor
and may have improved reliability. Because of the electronics in the
system, it also may have better unit-to-unit reproducibility in produc-
tion than the carburetor. This system may not be a solutidn to the
problem of carburetor maladjustment but it may eliminate some of the
adjustable mechanical controls.

Closed Loop System of Fuel Metering
A closed loop (feedback) system Is one which the command (input)
variable depends in some way on the value of the controlled (output)
variable. This means, for example, that the air-fuel ratio delivered
by the carburetor can be a function of the composition of the exhaust
gas. Important properties of a closed loop system as compared to an
open loop control are increased accuracy in obtaining desired values of
the output variable and reduced sensitivity to internal and external
disturbances. The closed loop system may allow retention of the system
calibration over long periods of time. A schematic of the feedback
system currently in development by Holley Carburetor is shown in Figure
The basic components of the closed loop system are the exhaust gas
sensor, electronic control unit (ECU), vacuum control valve and the
feedback carburetor.
The exhaust gas sensor, a zirconium dioxide (ZrO„) sensor, is
located in the exhaust stream between the engine and the catalytic
converter. The sensor produces a voltage signal dependent on the
*'The Feedback Carburetor-Closed Loop Control of Fuel Metering," Holley
Carburetor, Submission to EPA, March 1976.
Figure 3-3 - Holley Feedback System

partial pressure of oxygen in the exhaust. This signal is the main
control input to the electronic control unit. The control unit produces
a square wave output signal of constant frequency and variable band
width to the vacuum control valve which regulates the vacuum applied to
the carburetor. The vacuum signal operates the metering systems in the
carburetor, Figure 3-4 shows a schematic of the feedback carburetor
employed in Holley's system.*
Figure 3-4 - Feedback Carburetor
Control of the main system is accomplished by varying the size of
the fuel orifice in parallel with the main metering jet. The idle
system is controlled by a variable air bleed in parallel with a fixed
air bleed. When a high vacuum signal is applied to the carburetor, it
will meter lean and as the vacuum decreases, the metering will richen.
Under certain conditions such as cold starts and wide open throttle,
sensors are used to indicate to the control unit that these conditions
exist and the exhaust gas sensor is overridden by one of these auxiliary
The feedback system development may not completely solve the carburetor
maladjustment problem, but it does offer several features. The capability
to meet the fuel metering accuracy requirements of advanced emission
*"The Feedback Carburetor-Closed Loop Control of Fuel Metering,"
Holley Carburetor, Submission to EPA, March 1976.

control systems and a simple design which uses vacuum operated metering
elements makes this system a potential component of future emission
controlled vehicles. The automatic metering of fuel may reduce the
possibility of maladjustment.
A non adjustable idle circuit, if it were a requirement, may
generate much discussion. Idle adjustment is currently provided, (ac-
cording to those who maintain the nonadjustable approach may not work)
for the following reasons: (1) to compensate for gumming, (2) to allow
the carburetor to be tailored to different engines, (3) to allow for
production tolerances, and (4) to permit adjustment after engine "break-
The degree to which the above reasons are technically valid, and
cannot be overcome by design improvements, cannot be definitively deter-
mined at this point in time. However, some information on this issue
can be found from the 1976 EPA certification durability results. In
this testing, no GM vehicle required idle mixture adjustment for 50,000
miles. The idle mixture was not adjusted to compensate for break-in, an
adjustment of idle RPM was sufficient. According to the maintenance
instructions, idle mixture adjustments are not required for the vehicles
in the field.
This promising demonstration must be weighted against the potential
requirements for some adjustability should something go wrong, and
adjustments and/or repairs beyond scheduled maintenance be required.
This possibility that adjustment might be needed for extraordinary
conditions, might mitigate against eliminating the adjustability feature
altogether, as might be inferred from the GM results.
EPA could approach the adjustability issue by following the approach
said to be under consideration by Canada and possibly by Austria. Such
an approach would be to set an idle CO limit that could not be exceeded.
For example, either three times the specification or the specification
plus 0.5% CO whichever is lower, could be used. For example on the
Chryslers, the two approaches would yield 0.9% and 0.8% idle CO for a
specification of 0.3%. A test procedure that could be used would
generate discussion. EPA mechanics could try to maladjust the vehicles.
Those vehicles that could be adjusted to idle CO values in excess of the
performance standards would fail and not be certified. Since the govern-
ment maladjusters would have nearly a free hand, this might encourage
the design and introduction of non-adjustable idle circuits.
Some choke adjustments are relatively easy to alter. On the ex-
ternal bi-metal choke like Ford uses, it takes less than one minute to
adjust the choke extremely rich. These adjustments could be eliminated
with internal posts that prevented turning the cap, or restricted the
motion to + 2 notches, which may be acceptable for a wide variety of

ambients. The ease of removing the electrical wire for the electrical
choke heater could also be reduced via performance standards of the same
nature as the idle adjustment maladjustment described above.
Other Redesigns
Current vacuum routings could be improved because simple recon-
nection of vacuum hoses can have a negative emissions impact. For
example, re-routing vacuum hoses to provide full vacuum advance to the
distributor at all times may be a common in-use practice. The systems
could be redesigned to make hoses different sizes, or make hoses with
different fittings on the ends. This approach may rot solve the pinch-
ing off, or blockage of, for example, EGR hoses. EGR hoses could be
made of metal but these could still be pinched. Making the metal hose
not easy to be replaced, (welded on) would reduce EGR line blockage.
Another approach would be to redesign the intake and exhaust manifolds
to make the EGR valve and plumbing completely hidden and out of sight.
This might involve manifold casting changes.
Some air pumps are tampered with by cutting the drive belt. The
drive could be modified so f.hat the belt that drives the air pump also
drives an accessory considered more necessary (like the alternator or
the water pump). The air pump lines could also be cast into the block
and head and the pump be put under a cover, so that it was relatively
inaccessible. New castings might also be required for this approach.
In order to prevent maladjustment by replacing centrifugal advance
springs and/or weights, the distributor mechanism could be sealed at the
factory and be replaceable only as a unit.
Air cleaners could be redesigned so that the top cover cannot be
inverted. This appears to be a common in-use "fix" which many believe
helps, when all it really does is eliminate some inlet air filtration
and render ineffective some evaporative control systems.
The degree cf altering or removing the catalyst itself is unknown
but if removal of catalyst pellets or substituting a straight pipe for a
monolith catalyst are shown to be common, then the catalyst could be put
in the same casting as the exhaust manifold. This is expected to be a
significant design challenge.
A recent survey of owners voluntarily participating in EPA emis-
sions factors program indicated a percentage (9%) of owners had used
leaded fuel regularly in their catalyst-equipped vehicles. While
subsequent interviews with these owners suggested misunderstanding or

confusion over the survey question, it is not difficult to envision
defeat of the fuel filler inlet restriction through purchase of a plastic
funnel or similar adaptor (one such device is legally sold and provided
by some dealers in Canada) in the hardware section of numerous stores.
If there is shown to be a significant catalyst poisoning problem, one
obvious solution may be in a change in the price structure of gasoline
so that leaded fuels are equal or greater in price than unleaded.
Another re-design that falls under the general reduced maladjustment
potential category deals with ambient conditions, especially pressure.
Vehicles certified for high altitude are usually designed in a way that
could make them run poorly at low altitude. To eliminate any need to
have one's car adjusted for altitude or have a high altitude car ad-
justed for low altitude, air-fuel ratio and spark advance could be fully
altitude compensated. This could prevent a possible loss of mobility of
that part of the population that buys vehicles at high altitude and
finds that those vehicles are not suitable for low altitude operation
without costly repairs.
The same approach could be used for temperature variations expected
in the field, but the technical solutions to make a car meet the emission
standards at 0°F or 90 F, for example., are not as immediately apparent
as they are for altitude.
3.2 Maintenance Technology Improvements
This section deals with improving maintenance capabilities through
technology. Other approachs that might help the capability of the
service industry are not considered here.
Make Vehicles Easier to Maintain
The complexity of today's vehicles, caused to some extent by the
types of emission control systems used by the manufacturers have tended
to make some maintenance more complicated than was the case ten years
The situation could be improved by setting a performance standard
that would involve determining an acceptable time to perform emissions-
related maintenance. This might encourage designs that were more main-
tainable and help to eliminate maintenance difficulties like jacking up
of an engine off its mounts to change a spark plug.
Making vehicles easier to maintain could also involve eliminating
the need for special tools which Ł>re sometimes only practically avail-
able from the dealer. Special generally unavailable wrenches could be
eliminated, for example, and maintenance that required tools more
specialized than those generally deemed to be available could not be

Minimum tolerances around specific components could be set so that
access is easier. Making apparently simple changes like standardized,
easy-to-see timing marks that could be seen easily and only indicated
the correct value of timing could be effective.
Better Information Dissemination
There can be literally hundreds of running changes made during a
model year to a large manufacturer's fleet. Many of these running
changes involve calibrations that must be known by the mechanics if the
vehicles are to be maintained properly. If the detailed data are not
available to the mechanic, he cannot be expected to do it properly.
Asssuming that the tune-up specifications on the required sticker are
not enough, two approaches could be taken. First, the information on
the specific vehicle could be required to be a part of the owner's
manual. Second, there could be more use of today's advances in com-
munications. It is not inconceiveable to consider that a toll-free
telephone number be set up for anyone (dealer mechanic, independent
mechanic, or owner) to call. Upon identifying the VIN all the main-
tenance information would be transmitted to the mechanic and provisions
could be made to answer specific questions.
Standardized Practices
There are different ways to tune up different vehicles. Consider
idle adjustment; some use propane addition, some use the "lean drop" and
some use smooth idle to set the idle mixture. This, coupled with dif-
ferent idle CO requirements for different vehicles, could be leading to
massive confusion by the mechanics in the field. One single method
could be recommended for use on all vehicles. Such a method would have
to be universally adaptable. The lean drop appears to be somewhat more
implementable than propane addition, although it would not overcome the
tendency of some mechanics to limit enleanment (i.e. RPM drop) in favor
of smooth idle.
Similar procedures could be used for setting timing. A standardized
procedure could involve having the same hoses connected or disconnected,
for example.
Better Diagnostic Technology
The first step here would be to improve diagnostic equipment of the
type already in the field. Performance of dwell tachometer could be
improved so that they were more accurate and consistent for example. As
far as emissions measuring equipment goes the current performance is not
good. Current instruments, idle HC/CO meters in particular, are not
adequate, if California's experience is any indication. A study done in
California found that 80 percent of the meters did not work properly.
Assuming that the nationwide performance is like that in California, it
is apparent that improvements are needed in these instruments.

Section 4
Conclusions and Recommendations
The conclusions listed herein apply to the timeframe, Summer 1976,
when this report was being prepared and they reflect the interpretations
of the limited amounts of data available.
1.	There exists a relationship between idle CO levels and the
ability of the vehicle to pass either the 1975 FTP or various state
inspection tests.
The EPA Emission Factors Program has shown that there is a good
probability that a vehicle with an idle CO level higher than specifica-
tion, or greater than 0.5 idle CO in the absence of a specification,
will fail an FTP test. Maladjusted idle air/fuel mixture is the most
common cause of high idle CO levels. The frequency of idle maladjust-
ment, the reasons for it or who is responsible for the maladjustment
cannot be established with certainty. Isolated studies have indicated
idle limiter caps may have been removed from 25 to 30+% of the 1975
vehicles investigated, but none of the studies were able to identify
why or by whom the idle caps were removed. This report speculates that
the caps were removed to enrichen the idle fuel mixture for vehicle
driveability improvements.
2.	Maladjustment of the carburetor idle circuit, which results
in high idle CO levels, needs to be prevented.
Some technically feasible approaches to a nonadjustable idle cir-
cuit are proposed and discussed in this report. However, there exist
some technical questions which need to be resolved concerning the non-
adjustable idle circuit among which are: gumming, production tolerance,
and the necessity of adjustment due to engine break in.
3.	There exists a probability, possibly as high as 37%, that a
vehicle with low (less than 0.5% idle CO) could also fail the FTP.
This could be an indication that maladjustment of carburetor
chokes may also be a cause for failure on the FTP. However, more data
are required to determine that choke maladjustments are as important as
maladjusted idle CO. Currently, maladjusted chokes appear to be a
lesser problem. Idle rpm and ignition timing show little relationship
to FTP failure rates.

4. There is no general agreement on the cause(s) of poor in-use
vehicle emissions performance.
Two probable reasons for poor emissions performance are due to a
lack of service industry capability or poor vehicle driveability, but
the entire in-use performance problem is a multifaceted relationship
which may make impossible the identification of a single source that is
immediately amenable to corrective actions.
The question remains as to why so many emission control components
are maladjusted. The appearance is given that "someone" has been tinker-
ing with these controls, but the motivations behind the tinkering remain
obscure. Is it poor instructions in the field? The business relation-
ship between the vehicle owner and service industry representative?
Ignorance of emission controls? Apathy? Mis-manufacture? Vehicle
5. More information is needed to more properly assess the in-u^e
problem, before an attempt can be made to demonstrate solutions to tafie
problem.	v
EPA has initiated a program, titled Restorative Maintenance (RM)
within its Emission Factors Program, aimed at providing information con-
cerning the magnitude, frequency, and types of emission control com-
ponents maladjustment and/or failure. Cooperation has been elicited and
received from the automotive industry. Data from the currently in
process Restorative Maintenance Program was not available for this
Considering the complexity of the in-use vehicles emission per-
formance problems outlined in this report it is not surprising that
there could be a wide range of recommendations generated by discussing
these problems. While it is realized that no single answer to this
problem exists, it is equally evident that additional work must be
conducted and aimed at both understanding this problem and validating
the proposed solutions. In this context, the following recommendations
are proposed.
1. Nonadjustable carburetor idle circuit
Because of the accessibility and apparent frequence of maladjust-
ment of the idle circuit, it is recommended that a nonadjustable idle
circuit be considered seriously. While the current adjustability idle
circuit may be considered to be necessary by the some due to corrections
of production tolerance stack up, the in-use results indicate that the
idle circuit is regarded as a convenient "correction" to in-use vehicles.

The desirability of a nonadjustable idle circuit is clear, but the
degree of nonadjustability remains open for discussion. This discussion
centers around the unanswered questions of the effects of production
tolerance stack up, fuel gumming, the carburetor industry's capability,
and most importantly the necessity for adjustment during and after
vehicle break-in and extended operation. However, potential solutions
to the idle circuit maladjustment problem exist.
Also to be resolved are the alternatives of increased idle circuit
complexity versus complete nonadjustability after carburetor installa-
tion. A sealed idle circuit, would be beneficial, but those who indicate
that idle adjustments are a necessity suggest that a more complex idle
circuit to deter "tinkering" with the idle could be the desired approach.
Both approaches deserve further study and demonstration.
2.	No adjustment of idle during certification durability testing
Attention should be given to the certification durability test
procedure, especially those maintenance portions involving the carburetor
idle adjustment. Records compiled by an EPA* report indicate that of
the 162 engine families reported on during the 1974 Certification cycle
fully one third of them did. not require, for whatever reason, an Idle
mixture adjustment. The remaining engine families were adjusted, with
the result that on the average there were 1.4 idle adjustments per
engine family during 50,000 durability miles. It is Interesting to note
that this same report indicates that the owners manual suggests that the
same engine families perform idle adjustments on the average of 3.5
times during the same mileage period. There would appear to be a
disparity between (1) certification and (2) in-use vehicle instructions
and (3) potentially what Is actually needed.
In view of apparent widespread maladjustment of idle circuits,
strong consideration should be given to not allowing the manufacturers
to adjust idle, other than possibly idle rpm, during the 50,000 mile
durability portion of certification testing. This restriction could be
used as interim feature to (1) be a forcing function for the adoption of
a nonadjustable idle circuit in production carburetors, (2) reduce
number of recommended idle adjustments required in the owner's manual,
(3) be an example to the industry of the desirability, worth, and benefits
for nonadjustment over 50,000 miles.
3.	Inclusion of a Drlveabllity test in the Certification Procedure
The fact that drlveabllity is a subjective factor does not lessen
its importance when studying in-use vehicle problems. Elsewhere in this
report it is argued that drlveabllity may be a forcing function that
*Actual and Recommended Maintenance Practices for Light-Duty Vehicles
for 1975 and Later Model Years, EPA-460/3-75-009, August, 1975.

causes or is an indirect cause of in-use vehicle idle circuit malad-
justment. Since this type of maladjustment can cause gross increases in
emissions, especially for CO, it should be mentioned that the approach
of assuring that driveability is acceptable is to make engines less
adjustable. This approach would restrict engine designers from picking
engine specifications for emission testing that the designer know (or
should know) will cause customer complaints, with assurance that the
complaints will be easily remedied in the field by the making of simple
Consideration should be given to the introduction of a standardized
driveability test to be performed in conjunction with the Certification
program. The driveability ratings could be published as a consumer
guide and/or possibly posted on the vehicle as a label in a voluntary
industry program such as was the case for fuel economy in 1975 and 1976.
Introduction of a standardized driveability test will not be easy.
Potentially, the CRC driveability test could be incorporated into thee
Certification procedure as a common basis for driveability judgement^.'
The CRC driveability test provides a test procedure which judges cold
start driveability in a manner comparable to in-use vehicle operations.
It should be emphasized that it is suspected that much of the c
driveability problems encountered by the consumer become manifested at
ambient temperatures lower than the ambient temperatures at which' the
certification test procedure is conducted. Therefore, the inclusion of
a driveability test in the certification process will require amb'ient
temperature ranges reflecting ambient temperatures closely representing
those experienced by in-use vehicles. It should also be noted tlfat
EPA has not conducted studies quantifying the temperature ranges ,neces-
sary or the detailed procedures required for lower ambient temperature
testing for driveability. Approximately one year will be necesssfary
to develop such procedures.
A driveability test could be included in the certification verifica-
tion testing done by the EPA Ann Arbor laboratory. However, the-t inevitably
subjective nature of this test will require increased education and
training of personnel to conduct the CRC driveability test, as ,well as
some increase in staffing. Expanded facilities, possibly to the extent
of needing a test track may be required. Preliminary estimates indicate
that at a minimum for a test track facility it will be necessary to allow
nine to twelve months for the development of a suitable driveability
test plus the training of drive raters. Also the addition of eight
people to the present laboratory complement and 600,000 dollars 
Initially, a minimum drive number could be established by EPA for
acceptable drive performance to serve as a guideline for industry and
consumer. This minimum guideline might tend to reduce any tradeoffs of
driveability for lower emissions levels which apparently fosters the
widespread maladjustments of the emission control systems.
4.	Driveability Standards May be Required
If a voluntary approach such as outlined in Recommendation 3 above
does not prove workable en approach toward reducing the problem in this
area might be Federally mandated driveability standards. Of course, a
thorough study of the regulatory impacts of such standards is required.
5.	Service Industry/Owner Interfacing
It would seem appropriate to determine the extent of vehicle owner
insistance for "fixing it to run right" which is responded to by the
service industry by adjusting engines other than to specifications. A
limited report done on willful tampering* conducted by EPA suggests that
the service industry may in many cases respond to owner's wishes for
"corrections". If this phenomenon is as widespread as is suspected,
then the manufacturers may need to be required to develop relatively
nonadjustable emission control systems for their vehicles.
6.	Use of "Available" Maintenance During the Certification Process
It is suggested that EPA consider using EPA mechanics (which could
be contractor personnel) to factor some approximation of in-use main-
tenance into the certification procedure. These average or typical
mechanics could be given the tools presently available to mechanics in
the typical service industry facility. Implementing this recommendation
may foster development of maintenance free emission control systems more
quickly than might be otherwise expected. It is realized that a thorough
study./of the regulatory issues involved, including cost of compliance,
lead /time and cost-effectiveness would be required to determine the
viability of this approach.
Standardize Tune-up Adjustment Methods
Qne potential approach toward reducing the problem could be the
standardization of adjustment levels between all vehicle manufacturers.
For example, there presently are three methods of idle adjustment
specified by the major automobile manufacturers; idle CO measurement
before, the catalyst, the lean drop method, and propane addition. As can
be seen;by these differences, idle adjustment procedures have increased.
Some procedures are so complex that the do-it-yourselfer or small maintenance
shop may not be able to perform his own adjustment without large capital
outlay for test instruments. A standardized idle adjustment may help
alleviate this problem. Additionally, if a standardized idle adjustment
was employed, field instructions might be more easily simplified and
better understood.
*A Study of Fuel Economy Changes Resulting from Tampering with Emission
Controls, TAEB Report 74-21 DWP.