EPA-AA-IMS-81-19
Technical Report
Effect of Low Cost Repairs on I/M Failed Vehicles
by
R. Bruce Michael
September, 1981
Inspection and Maintenance Staff
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air, Noise, and Radiation
U.S. Environmental Protection Agency
-------
Table of Contents
Section Heading
1.0 INTRODUCTION
2.0 SUMMARY AND CONCLUSIONS 4
3.0 DESCRIPTION OF TEST PROGRAM 6
3.1 Test Vehicles 6
3.2 Stage 1 Maintenance 6
3.3 Stage 2 Maintenance 6
3.4 Tests Performed 7
4.0 TEST RESULTS 8
4.1 Types of Repairs Needed to Pass the I/M Test 8
4.2 Federal Test Procedure Emissions, Idle Emissions,
and Fuel Economy Results 9
4.3 Excess Emissions Reduced by Repairs 13
4.4 Comparison of Results With Three Other Programs 15
4.5 Driveability Evaluation 16
-------
1.0 INTRODUCTION
Mechanics who repair vehicles which have failed an Inspection and Maintenance
(I/M). test are usually advised to adjust these vehicles to manufacturer
specifications, particularly for the idle mixture adjustment. This method is
known to be effective in reducing emissions while maintaining or improving
driveability and fuel economy. Most pre-1981 model year vehicles have
adjustable idle mixtures. Manufacturer specifications for idle mixture have
the disadvantage of being diverse among vehicle makes and model years,
complex, and often difficult to locate in commonly available service
publications. For this reason, it is quite probable that many if not most
mechanics repairing I/M failed vehicles ignore manufacturer specifications for
idle mixtures. Instead, it is likely that many mechanics simply adjust the
idle mixture screws, while watching the idle CO level on an emissions
analyzer, until the CO level is low enough to be sure the vehicle will pass
the cutpoints during the reinspection. This study was designed to determine
whether this simple carburetor adjustment approach is effective, whether a
universal target idle CO level .could be used, and what additional repairs are
necessary to pass the I/M test.
This study* measured the effect on emissions of specific quick and low cost
repairs which differed from manufacturer specifications. Repairs centered on
a simple carburetor adjustment to a universal target idle CO level of 0.2% for
1976 and 1978 model year vehicles which initially failed a state I/M test. No
manufacturer specifies a procedure this simple for any recent model year
vehicle. Repairs were performed in two stages, when appropriate: Stage 1,
where only the simplest repairs, mainly the carburetor adjustment to 0.2% idle
CO were made; and Stage 2, where the correction of other problems were also
made if Stage 1 repairs did not result in low idle emissions. The study
period was October 1980 through January, 1981.
The vehicles used in this program were 1976 and 1978 model year passenger -
vehicles•from Vancouver, Washington. They had been in a previous study (Test
Group No. 9) and had failed the State I/M test in Portland. The vehicles were
not screened in any way, such as for tampering. Selected vehicles were not
restricted to those failing for idle CO, although all repaired vehicles did
fail for idle CO; nearly all vehicles failed the state idle test for both HC
and CO. After their initial tests, these I/M failed vehicles were given
repairs by the contractor and subsequent retests. Since Vancouver does not
have an I/M program, these vehicles are representative of ones which fail the
I/M test in a newly instituted I/M program.
The two most important potential uses of the study results are: (1) to
recommend to states specific low cost I/M repair procedures as a possible
substitute for manufacturers specifications; and (2) to modify the mechanic
training course for emission repairs developed by Colorado State University
for EPA.
* Contractually referred to as Test Group No. 10 of EPA Contract No.
68-03-2829 with Hamilton Test Systems in Portland, Oregon.
-------
2.0 SUMMARY AND CONCLUSIONS
1. Substantial emission reductions were achieved by'simple specific repairs.
Stage 1 maintenance, mainly a carburetor adjustment to a universal idle CO
target, was all that was needed to pass the State idle test for 29 of the
35 vehicles (83%) which were repaired. Stage 1 maintenance reduced
average FTP HC emissions of all 35 vehicles by 40% and FTP CO emissions by
58%. Stage 2 maintenance, only performed on six vehicles, increased the
average reduction for the 35 vehicles to 47% for HC and 71% for CO.
2. Study results indicate that the average reductions in FTP HC and CO
emissions from specific idle CO adjustments can be as large as idle
adjustments using manufacturer specifications. A recent EPA study in
Houston also agrees with this (see Section 4.4). In most cases,
adjustments using a target idle CO would be faster and easier for
mechanics, resulting in cheaper repairs.
3. One might be concerned that a specific idle CO adjustment approach would
not leave the vehicles with as good driveability as a manufacturer
specifications approach. However, we found only two minor driveability
problems resulting from the idle adjustment, one each on two of the 35
vehicles. One vehicle reportedly had an engine surge after the adjust-
ments, and another vehicle experienced hard starting. The former vehicle
needed further repairs and received a carburetor overhaul which then
improved its drivea'bility. It is uncertain whether adjustments on either
car using the manufacturer specifications would have eliminated
driveability problems. These were the only problems reported for all 35
vehicles on seven driveability measures. Therefore, it is reasonable to
say that these adjustments did not significantly affect driveability.
4. An average city fuel economy improvement of 2.5% was gained from the Stage
1 repairs. Highway fuel economy was reduced by 0.4%. After Stage 2
repairs, the average fleet fuel economy improvement was 3.0% in the city
and 0.9% on the highway.
The six vehicles needing more than just Stage 1 maintenance benefited
greatly in fuel economy from receiving both stages of repair. Stage 1
repairs resulted in fuel economy improvements of 3.2% in the city and 7.5%
on the highway compared to their as-received levels. After Stage 2
repairs, these six vehicles received an average 8.3% fuel economy
improvement in the city and 9.2% improvement on the highway, compared to
their as-received levels. • . •
5. The city fuel economy improvement from the idle adjustments of Stage 1 is
not as large as observed from similar adjustments in two other EPA
programs; the two other programs showed improvements of over 5%. No
obvious reason exists for the difference. A possible reason would be that
vehicles in Vancouver may be better maintained with respect to fuel
economy, which if true would be a result of local maintenance habits.
Sampling variations may also be a reason for the difference.
-------
The fuel economy improvements of vehicles receiving both repair stages are
larger than those seen in a recent study from vehicles receiving
adjustments plus carburetor overhauling or replacing (see Section 4.2.2).
That study showed that 17 vehicles receiving contractor repairs had fuel
economy- improvements of 3.3% in the city and 5.5% in the highway compared
to their as-received levels. A comparison between that study and the
present study is complicated by the fact, though, that the former study
vehicles had received repairs by commerical facilities prior to contractor
repairs and also received more extensive repairs by the contractor.
Because of th'is discrepancy in repairs between the two studies and the
fact of the small number of vehicles (six) receiving both repairs in the
present study, the reader should not assume that the large fuel economy
improvements observed after completion of both repair stages will always
be achieved.
-------
3.0 DESCRIPTION OF TEST PROGRAM
3.1 Test Vehicles
Test vehicles consisted of 35 light-duty passenger vehicles from Vancouver,
Washington which failed the Portland, Oregon State Inspection Test in their
as-received condition. Model years were 1976 and 1978. These vehicles were
initially tested in a previous test program (see footnote in 'Section 4.2.1.)
which compared emissions from similar Portland and Vancouver vehicles.
Therefore, the as-received levels of the vehicles in this program were the
same as in the prior test program. All repairs were performed by contractor
mechanics.
3.2 Stage 1 Maintenance
All vehicles went through this stage. The main portions of the repair
sequence are listed below.
a. If the idle CO is greater than 0.2%, adjust the fuel metering system to
0.2% or as close as possible.
b. Adjust idle speed only if the engine is running roughly; if that is the
case, adjust the idle speed for a smooth idle.
c. If idle HC is less than 225 ppm, do not perform any more adjustments.
If idle HC is greater than 225 ppm, check vacuum hoses, specified parts of
the ignition system, and basic timing adjustment, stopping repairs as soon
as idle HC is less than 225 ppm.
3.3 Stage 2 Maintenance
Vehicles only went through this stage if repairs in the first stage did not
lower idle emissions acceptably. The following items were to be repaired as
necessary.
a. PCV system.
b. Carburetor (if the 2500 rpm CO level is over 1.0% and no other causes
than the carburetor can be detected, overhaul the carburetor).
c. Heated air intake system.
d. Dilution of the oil with gasoline (repair is to change the oil).
e. Ignition system.
f. EGR system (check to make sure the valve is not open at idle).
-------
3.4.Tests Performed
The following tests were performed on the vehicles at each repair stage.
1. Federal Test Procedure
2. 50 mph Cruise Test
3. Highway Fuel Economy Test
4. Four-Mode Idle Test
5. Loaded Two-Mode
6. Diagnostic Inspection
-------
4.0. TEST RESULTS
4.1 Types of Repairs Needed to Pass the I/M Test
Stage 1 maintenance is designed to cover simple maintenance items only.
Repairs stop as soon as the idle levels are acceptably low. Generally, the
only maintenance which needs to be performed in this stage is the simple
carburetor adjustment to reach the idle CO target of 0.2%.. This was the
expectation when t.he study was designed. Stage 2 maintenance is only begun if
idle emissions can not be reduced to 0.2% CO in the first stage or if
emissions at 2500 rpm exceed 1.0% CO.
Table 1 shows the frequency of repairs in each stage. The reason for the "Not
Applicable" column is that after the carburetor adjustment in Stage 1 no other
diagnoses or repairs were performed if idle emissions were acceptable; this
was the case for 29 vehicles. At the end of the second stage, no vehicles
still had high idle emissions.
Table 1
Repairs Performed
Stage 1
Oil diluted with gasoline
Choke Operation
Carburetor
Vacuum hoses
Spark plugs
Spark wires
Distributor cap or rotor
Basic Timing
Idle Speed
Stage 2
Oil Change
Carburetor
Heated Air Intake System
EGR valve
Repair or
Adjust Replace
35 (adjust)
3 (rebuild)
1
OK
35
35
6
3
6
4
4
4
6
1
5
6
Not
Applicable
29
29
29
29
29
29
-------
4.2 Federal Test Procedure Emissions, Idle Emissions, and Fuel Economy Results
4.2.1 All Vehicles Receiving the First Repair Sequence
Results show that the simple Stage 1 repairs had a large FTP emissions benefit
on the 29 vehicles which only needed simple adjustments. Only a small benefit
was observed from Stage 1 repairs for the six vehicles needing more complex
maintenance (mainly carburetor overhauling), however. The amount of improve-
ments for the two groups is reversed when it comes to fuel economy changes,
though. Stage 1 repairs resulted in a small fuel economy improvement for the
29 vehicle group, but a more substantial one for the six vehicle group.
Average results are shown in Table 2 for all vehicles receiving Stage 1
maintenance. Vehicles which received only the Stage 1 repairs and vehicles
which later received the Stage 2 repairs are shown separately in parts A and
B. Idle emissions were measured with a commercial garage quality instrument.
Fuel economy is shown for both the Federal Test Procedure (FTP) and Highway
Fuel Economy Test (HFET).
The study which tested the as-received emission and fuel economy levels of 200
Vancouver and Portland vehicles reported that 1978 model year Vancouver
vehicles had atypical results.* Emissions were lower than have been measured
from 1978 models in other non-I/M cities. In order to see if this had an
effect on the results of the repair stages, vehicles were separated by model
year. Table 3 presents the FTP and fuel economy results of the Stage 1
repairs by model year. Results show that the first repair stage had nearly an
equal effect on each model year; therefore model year is not an important
factor for the repairs. The remainder of the report results are not separated
by model year.
* "Emissions Reductions from Inspection and Maintenance: Vancouver Versus
Portland Snapshot," EPA-AA-IMS-81-18, R. Bruce Michael, August, 1981.
-------
10
Table 2
A. Vehicles Receiving
Only The First
Repair Sequence
Before First Repair
After First Repair
Percent change
B. Vehicles Later
Receiving The
.Second Repair
Sequence
Before First Repair
After First Repair
Percent change
C. All Vehicles
Before First Repair
After First Repair
Percent change
29
29
35
35
Effect of Stage 1 Maintenance on
Emissions and Fuel Economy
N Odometer
Federal Test
Procedure
Emissions (g/mi)
HC CO NOx
49,795 3.31 44.0 2.85
1.84 14.7 3.12
-44% -67% +9%
62,498 2.91 55.7 2.43
2.46 40.3 2.77
-15% -28% +14%
51,973 3.24 46.0 2.78
1.95 19.1 3.06
-40% -58%
Idle Emissions
HC (ppm) (CO (%)
Fuel Economy
(miles per
gallon)
FTP HFET
365
177
-52%
250
110
-56%
345
165
2.93 14.77 20.77
0.22 15.05 20.59
-92% +1.9% -0.9%
1.73 13.23 18.03
0.38 13.88 18.31
-78% +4.9% +1.6%
2.72 14.48 20.24
0.25 14.84 20.16
-52%
-91%
+2.5%
-0.4%
Federal Test
Procedure Standards
35
1.50 15.0 2.82
Portland Idle
Test Standards
35
231
1.09
-------
11
Table 3
Effect of Stage 1 Maintenance on
FTP Emissions and Fuel Economy
by Model Year
FTP
Emissions (g/mi) Fuel Economy
.N HC CO FTP HFET
1976 Model Year
Before Repair 26 3.61 51.2 13.81 19.23
After Repair . 26 2.16 20.9 14.14 19.14
Percent Change -40% -59% +2.4% -0.5%
1978 Model Year
Before Repair 9 2.17 31.2 16.86 23.85
After Repair 9 1.33 13.7 17.28 23.84
Percent Change -39% -56% +2.5% .0.0%
4.2.2 Vehicles Receiving Stage 2 Maintenance
Six vehicles received both the first and second repair stages. Results at
each stage are shown in Table 4. Five of the six vehicles mainly received
either a carburetor rebuild or replacement at the second repair stage. As can
be seen, this type of repair greatly improved fuel economy. It is interesting
to note that the five vehicles needing carburetor work all had high mileages:
the range was 43,000 to 85,000 miles.
The improvements in fuel economy are larger than the improvements seen in an
earlier study in Portland which analyzed the effects of carburetor rebuilds
and replacements for 17 vehicles. These latter vehicles had improvements of
3.3% on the FTP and 5.6% on the HFET relative to their before repair levels.*
Many other repairs were included in this latter study, however, which hinders
an exact comparison.
* A study of the Effectiveness of Mechanic Training For Vehicle Emissions
Inspection and Maintenance Programs, EPA-AA-IMS/81-11, R. Bruce Michael,
April, 1981, Table 12.
-------
12
Table 4
Vehicles Receiving
Two Repair Sequences
Before Repair
After First Repair
After Second Repair
Federal Test
Procedure
Emissions (g/mi)
N Odometer H£ CO NOx
6 62,498 2.91 55.7 2.43
6 2.46 40.3 2.77
6 1.11 7.9 3.81
Idle Emissions
HC (ppm) (CO (%)
250
110
82
1.73
0.38
0.00
Fuel Economy
(miles per
gallon)
FTP HFET
13.23 18.03
13.88 18.31
14.33 19.68
Percent Change
Stage 1 Maintenance
vs. As-Received
-15% -28% + 14%
-56% -78% +4.9% + 1.6%
Percent Change
Stage 2 Maintenance
vs. Stage 1 Maintenance
Percent Change
Stage 2 Maintenance
vs. As-Received
-55% -80% +38%
-62% -86% +57%
-25% -100% +3.2% +7.5%
-67% -100% +8.3% +9.2%
4.2.3 Emissions and Fuel Economy of All Vehicles at Each Repair Stage
Results of all vehicles at the first and the final repair stages are shown
in Table 5. The "final" repair stage means the first repair for 29
vehicles and the second repair for 6 vehicles. The final repair gave
modest further fleet reduction in HC and CO and slight fuel economy
benefits.
-------
13
Before Repair
After First Repair
After Final Repair
Table 5
Emissions and Fuel Economy of
All Vehicles After Final Repair
Federal Test
Procedure
Emissions (g/mi)
N Odometer HC CO NOx
35 51,973 3.24 46.0 2.78
35 1.95 19.1 3.06
35 1.71 13.5 3.24
Idle Emissions
HC (ppm) (CO (%)
345
165
161
2.72
0.25
0.18
Fuel Economy
(miles per
gallon)
FTP HFET
14.48 20.24
14.84 20.16
14.92 20.43
Percent Change
Stage 1 Maintenance
vs. As-Received
Percent Change
Final Repair
vs. As-Received
-40% -58% +10Z
-47% -71% *17Z
-52%
-53%
-91%
-93%
1-3.0%
-0.4%
-0.9%
4.3 Excess Emissions Reduced by Repairs
Excess emissions are defined as FTP emissions above the Federal
standards. For example, if a vehicle has FTP CO emissions of 20 grams per
mile (g/mi) and the Federal standard for that vehicle is 15, then it has 5
g/mi excess CO emissions.
The analysis of excess emissions before and after repair provides valuable
information on the repairs. Table 6 presents the excess HC and CO
emissions before and after the first repair stage. 'For comparison, excess
emissions before and after commercial I/M repa'ir are also shown from the
EPA Portland study, which evaluated the effectiveness of the Portland I/M
program.* The vehicles analyzed from the Portland study were 1975-77
model years. The vehicles which are compared in Table 6 all had the same
Federal standards for HC and CO.
* "Portland Study Element II - I/M Effectiveness Study", EPA-460/3-79-003,
May, 1979; "Portland Study Element III - Post I/M Deterioration Study",
EPA-460/3-79-009, July, 1979; and "Update on EPA's Study of the Oregon
Inspection/Maintenance Program", by Rutherford and Waring, APCA 80-1.2,
presented at Air Pollution Control Association 73rd Annual Meeting, June
24, 1980.
-------
14
Results show substantial reductions in excess emissions from the simple
repairs of Stage 1. The reductions also compare favorably with the I/M
repairs from the Portland study. As expected, the HC reductions from the
simple repairs of the first repair stage are not quite as large as from the
Portland study (64% vs. 74%), because the simple repairs focused primarily on
idle CO. The CO reductions from the simple repairs are larger than from the
Portland study, however (74% vs. 68%). The second repair stage further
increased the emission reductions, equalling the Portland study for HC and
further bettering the Portland study results for CO.
The cost-effectiveness of these simple repairs would be high, because they
approximately equal the emission reductions of the more common repairs and
should be cheaper to perform. EPA does not currently have data to estimate
what the simple repairs performed by the contractor mechanics in this study
would cost if performed by a commercial repair facility, however, so an
accurate comparison of cost-effectiveness cannot be made.
Table 6
Excess Emissions Reduced by Repairs
Comparison of Specific Repairs and Common I/M Repairs
Average Excess Emissions
Above Federal Standards
(grams per mile)
Odometer
Low Cost Repairs
(Present Study, N=35)
As-Received
After Stage 1
After Stage 2
51,973
HC
1.77
0.65
0.47
CO
31.2
8.1
3.5
Percent Change Stage 1 vs. As-Received
Percent Change Stage 2 vs. As-Received
-64%
-74%
-74%
-89%
Commercial I/M Repairs
(Portland Study, N=233)
As-Received
After Repair
30,929
1.29
0.34
25.6
8.1
Percent Change
-74%
-------
15
4.4 Comparison of Results With Three Other Programs
Three earlier EPA studies examined the effect of repairs similar to the ones
discussed in this study. Two different EPA Restorative Maintenance (RM)
studies te-sted 1975-80 model year vehicles.* One of the repair stages
contained just the adjustment of idle speed and mixture, which is nearly
identical to the adjustments performed in the first repair stage for most
vehicles in the present study except that adjustments in the -RM studies were
performed according to manufacturer specifications.
The third of the earlier studies was part of an I/M test program in Houston,
in which many vehicles received only an idle CO adjustment and then were
retested.** This is exactly the same type of adjustment as made in the
present study except that a slightly different idle CO target (0.5% instead of
0.2%) was used. The four studies (including the present one) all had repairs
performed by contractor personnel and tested vehicles which were similar in
terms of model years and mix of makes and models.
The four programs selected vehicles for repair differently, however. In
Houston vehicles above 3.0% CO received an adjustment to a. target of 0.5% CO.
In the two RM studies, vehicles failing the Federal Test Procedure standards
received carburetor adjustments to the specifications of the manufacturers.
In the present study vehicles failing the State Inspection Test (SIT) for idle
emissions received an idle CO adjustment to a target of 0.2% CO; all 35
vehicles initially failed the SIT for CO, the CO limit being 1.0% for nearly
all vehicles. Therefore, in order to compare the results most effectively,
only vehicles from the present study and the RM programs which would have
initially failed the Houston program cutpoints of 3.0% idle CO were used.
Also, vehicles in Vancouver which had any parts replaced during the Stage 1
repairs were not included, thus leaving 16 cars which received only idle speed
and mixture adjustments.
Table 7 -presents the emissions and fuel economy results from the different
programs. Results from the two RM programs are combined. Emission reductions
are very similar for the three programs. This similarity in emission
reductions supports the conclusion that simple low cost carburetor adjustments
are as effective as more complicated adjustments to manufacturer specifica-
tions. The fuel economy improvements are not the same in the three studies,
however. Both the Houston and RM programs show larger improvements in FTP
fuel economy due to repair than the present study shows. No reason for this
is apparent, although it is 'possible that different maintenance habits
contribute to the discrepancy.
*1. "An Evaluation of Restorative Maintenance on Exhaust Emissions of 1975-76
Model Year In-Use Automobiles", EPA-460/3-77-021, December, 1977.
2. "FY79 study of Emissions From Passenger Cars in Six Cities", EPA-460/3-
80-020, October, 1980.
** "Testing Support for an Evaluation of a Houston I/M Program", EPA-460/3-
80-021, October, 1980.
-------
16
Table 7
Comparison of Results With Other Programs
All Vehicles Had Initial Idle CO Greater Than 3.
Federal Test
Procedure Emissions .Idle Emissions
HC CO HC (ppm) CO (%)
Vancouver Before Repair 3.35 49.2
(N-16) After Repair* 1.65 12.9
Percent Change -51% -74%
382
158
-59%
3.94
0.23
-94%
Fuel Economy
(miles per gallon)
FTP HFET
14.79
15.15
21.13
21.06
+2.4% -0.3%
Houston Before Repair 3.14 59.1
(N-85) After Repair 1.31 23.3
Percent Change -58% -61%
290
130
-55%
4.35
0.52
-88%
13.79
14.61
20.57
21.01
+5.9% +2.1%
Restorative Before Repair
Maintenance After Repair
(N-66)
Percent Change
2.48
1.09
48.1
10.6
-56% -78%
375
105
-72%
5.17
0.29
-94%
14.25
15.08
20.71
21.07
+5.8% +1.7%
4.5 Driveability Evaluation
Driveability did not significantly change as a result of either of the repair
stages. In their as-received condition all 35 vehicles reportedly had
satisfactory driveability on seven measures: engine start, surge, stumble,
backfire, stretchiness, miss and diesel (run-on). After the first repair
stage, two of the vehicles each had one driveability problem. One vehicle had
hard starting and another had an engine surging condition. The latter vehicle
needed a carburetor overhaul, though, which gave it satisfactory driveability
again. None of the six vehicles which received the- second repair had problems
after repair.
* Only carburetor adjustments.
replaced.
No vehicles were included which had any parts
------- |