United States       Air and Radiation     EPA420-P-99-019
           Environmental Protection              May 1999
           Agency
vvEPA     Determining Repair
           Effects on IM240 Cold
           Start Emissions for
           1981  and later Light Duty
           Vehicles

           DRAFT
                               > Printed on Recycled Paper

-------
                                                                             EPA420-P-99-019
                                                                                    May 1999
                                    on                                             for
                                  Janet C. Kremer

                          Assessment and Modeling Division
                              Office of Mobile Sources
                        U.S. Environmental Protection Agency
                                     NOTICE

    This technical report does not necessarily represent final EPA decisions or positions.
It is intended to present technical analysis of issues using data which are currently available.
         The purpose in the release of such reports is to facilitate the exchange of
      technical information and to inform the public of technical developments which
        may form the basis for a final EPA decision, position, or regulatory action.

-------
                                        -Draft-
               Determining Repair Effects on IM240 Cold Start Emissions
                         for 1981  and later Light Duty Vehicles

                              Report Number M6.IM.002
                                      May 1999
                                   Janet C. Kremer
                      U.S.EPA Assessment and Modeling Division
Introduction
When the MOBILE model calculates credits from Inspection and Maintenance (I/M) programs,
one important input is the difference between emissions before and after successful I/M repairs.
These "repair effects" may vary with vehicle type and model year. They also may vary with test
type and test cutpoints, and may be different for running and start emissions.

This document examines the repair effects on start emissions. It describes how the database for
this analysis was developed. It also describes how to calculate cold start emissions from the
Federal Test Procedure (FTP) bag results. The analysis in this report covers Light Duty Vehicles
(LDVs) and Light Duty Trucks (LDTs) with model years ranging from 1983 through 1991.
Repair effects for running emissions are found in EPA report number M6.IM.001.

Background

The current version of the MOBILE model is MOBILES. MOBILES calculates I/M credits using
a module called TECHS. TECHS is an external FORTRAN program which calculates and
exports the I/M credit values to MOBILES as built-in code or as an external file.

For MOBILE6, the new I/M credit methodology will be built into the model's code. This new
methodology will operate automatically every time an I/M program is called by the model. The
MOBILE6 I/M credits will have several new features. They will incorporate new basic emission
rates and I/M credits will be calculated for start and running emissions separately. Also
MOBILE6 has new classifications for defining emitter categories. In MOBILES, there were four
emitter classifications: normal, high, very high, and super. For MOBILE6, the emitter categories
are handled by two classifications, normal and high. Normal emitters are low emitting vehicles
with emission control systems that generally function properly. High emitters have average
emission levels which are considerably higher than the overall mean emission levels.

MOBILE6 will apply separate benefits to start and running emissions to generate I/M credits for
1981 and later model year vehicles. MOBILE6 will apply the  same benefits to start and running
emissions for pre-1981 model year vehicles.

More information on the new methodology for I/M credits can be found in the report "MOBILE6
Inspection / Maintenance Benefits Methodology for 1981 through 1993 Model Year Light
Vehicles," report number M6.EVI.001.

-------
Data

A lack of "real world" data on repair effects was a constraint on this analysis. In order to be
selected for this analysis a vehicle had to meet the following set of criteria:

•      Model year 1981 and later (I/M credits for pre-81 will not be updated in MOBILE6)
•      Gasoline vehicles only
       LDV or LOT only
•      Not tested in California or Colorado (because of different standards and effects of higher
       altitudes respectively)
•      Failed its "as received" test and passed its after-repair test
•      Vehicle has both I/M240 and corresponding FTP results ( FTP results are necessary to
       evaluate cold start emissions, i.e., emissions after 12+ hour soak)

The only available data meeting these criteria were in EPA's Emissions Factor Database. After
eliminating vehicles that did not meet the above criteria, the remaining vehicles, with their FTP
and EVI240 results, created the database used in this analysis. This database contained 369 LDVs
and 41 LDTs with model years ranging from 1983 through 1993.

Analysis

Calculating Start and Running Emissions

In order to examine repair effects on start emissions, it is necessary to determine what start
emissions are. The FTP results were split into start and running fractions using the methodology
explained in detail in report number M6.STE.002 and report number M6.STE.003.

Determining Emitter Categories

All the vehicles included in this analysis were classified as meeting "Tier 0" certification
standards and were defined as either normal or high emitters. Normal emitters  were defined as
vehicles with FTP emission rates less than or equal to two times the standard for hydrocarbons
(HC) and oxides of nitrogen (NOx) and FTP emission rates less than or equal to three times the
standard for carbonmonoxide (CO). High emitters were defined as vehicles with FTP emission
rates greater than two times the standard for hydrocarbons or oxides of nitrogen or FTP emission
rates greater than three times the standard for carbon monoxide. The  cutpoints used in defining
an emitter category can be found in Table 1.

-------
Table 1: Cutpoints used in determining emitter category
Vehicle
Type
LDV
LOT
LDT1*
LDT2
Model
Year
81 -93
81 -87
88-93
2 x HC Stand.
(g/mi)
.82
1.6
1.6
3 x CO Stand.
(g/mi)
10.2
30
30
2 x NOx Stand.
(g/mi)
2.0
4.6
2.4
3.4
*Light Trucks have different NOx standards based on vehicle weight

For the purpose of this analysis, additional criteria have been used for determining emitter-class
designation. In cases where a vehicle exceeds the emission cutpoint for either HC or CO, it is
considered an HC/CO high emitter. In this report, this emitter category will be known as "Emitter
Class A."  Vehicles which exceed the NOx cutpoint are considered NOx high emitters. This
emitter category will be referred to in this report as "Emitter Class B." It is then possible for a
vehicle to be a normal emitter for Emitter Class A and a high emitter for Emitter Class B or vice
versa.

Determining Pass or Fail with Cutpoint Combinations

Determining whether a vehicle passed or failed its as-received or after-repair tests depends on the
cutpoint for the test. The EPA developed six different sets of cutpoint combinations for this
analysis that reflected the range of cutpoints currently being used in I/M lanes. These
combinations are listed in Table 2. Different cutpoint combinations were analyzed to determine
whether cutpoint selection would greatly affect the amount of emission reduction.(Of course, as
described  in M6.EVI.001, cutpoints do affect identification rates and thus I/M credits) These
cutpoints were applied to both LDVs and LDTs for all model years. As in actual I/M tests, a
vehicle had two chances to pass for HC and CO pollutants. As set by EVI240 standards, "if the
composite emission level is equal to or below the composite standard or if the Phase 2 gram per
mile emission level is equal to or below the applicable Phase 2 standard, then the vehicle shall
pass the test for the exhaust component."(EPA Report Number: EPA-AA-EPSD-EVI-93-1) Phase
2 in this context means the average  emissions in the last 129 seconds of the EVI240 test. An
example of this pass/fail technique would be if a vehicle's EVI240 composite HC emissions were
1.0 grams per mile and its Phase 2 emission were 0.5  grams per mile and the cutpoint
combination being used was  combination #2 in Table 2, the vehicle would fail the composite
standard but would pass the Phase 2 standard therefore passing the test. This same vehicle would
fail the test if it were examined using cutpoint combination #1. Each vehicle in the data set was
given a pass or fail for each cutpoint combination.

-------
Table 2: Cutpoint Combinations Used to determine Pass or Fail for Composite (C) and
Phase 2 (P2) test data.

HC
CO
NOx
Comb. #1
C
0.6
10
P2
0.4
8
1.5
Comb. #2
C
0.8
15
P2
0.5
12
2.0
Comb. #3
C
1.0
20
P2
.63
16
2.5
Comb. #4
C
1.2
30
P2
.75
24
3.0
Comb. #5
C
2.0
60
P2
1.25
48
4.0
Comb. #6
C
2.4
78
P2
1.5
60
6.0
For this analysis, the EPA examined those vehicles that failed their as-received test and passed
their after-repair test based on HC, CO and NOx cutpoints and both emitter Class A and Class B.
Each vehicle's percent reduction for start and running emissions were calculated.

Calculating Percent Emission Reduction

Percent emission reduction can be calculated using a vehicle's before- and after-repair emission
rates. Emission reductions for each pollutant were calculated using the following equation.

Percent Emission Reduction = EBR - EAR X  100
                              EBR
Where EBR = Emissions Before Repair and EAR = Emissions After Repair
Results

Table 3 and 4 contain the results of this evaluation. Vehicles were divided into categories based
on their vehicle type and emitter category. Results showed that when repairing vehicles that were
classified as HC/CO high emitters, there is a noticeable increase in NOx emissions. This would
seem to indicate that I/M programs that test and repair vehicles based only on HC and CO
failures (i.e., HC/CO high emitters) would on average increase the fleet NOx emissions. This
phenomenon has been called NOx disbenefit and is discussed in more detail in EPA report
number M6.IM.001.
ANOVA

After examining the start emission results from this table, it appears that on average, no matter
which cutpoint  combination was chosen, the same average after-repair emission levels were
achieved. Based on this hypothesis, an analysis of variance (ANOVA) was performed on the
data, examining model year, vehicle type and cutpoint combination as independent variables.
Interest in model year and vehicle type significance comes from the fact that we want to be able
to split the start emission rates in MOBILE6 by model year and vehicle type. Table 5 contains
the results of this analysis.

-------
TABLE 5: Analysis of Variance Results
Analysis of Variance Results (ANOVA p-values)*
HC/CO High Emitters
Variables
Model Year
Vehicle Type
Model Year x Vehicle Type
Cutpoint Combination
HC
.0188
.0001
.0504
.4640
CO
.0001
.0014
.0184
.6959
NOx
.0001
.6968
.9311
.7630
NOx High Emitters**
Model Year
Vehicle Type
Cutpoint Combination
.0087
.6155
.9887
.0001
.0116
.9049
.0001
.0006
.9740
*in log space
** model year x vehicle type not applicable here because of lack of data

The statistical analysis showed that repair effects vary significantly by model year and also vary
by vehicle type for all but NOx emissions of HC/CO high emitters and HC emissions of NOx
high emitters. This analysis also verified that cutpoint combination was not a significant variable
for all three pollutants.

Conclusions

After reviewing the results of this analysis EPA has chosen to use one cutpoint combination
when determining the repair effects on start emissions. Cutpoint combination #2 at .8 g/mi. HC /
15 g/mi. CO / 2 g/mi. NOx was chosen because it is felt to be the most common cutpoint
combination used by I/M programs because it is the cutpoint combination EPA suggests in the
I/M test procedure guidance, EPA Report Number EPA-AA-EPSD-IM-93-1. Based on the
results of the ANOVA analysis, the data was stratified by model year group based on EPA
standards. Using cutpoint combination #2, the mean cold start after-repair emission levels were
determined for LDVs and LDTs for Emitter Class A & B. These cold start after-repair emission
levels, found in Tables 6 and 7, will be used to develop repair effects that will be  applied to all
starts in MOBILE6. It is these newly developed repair effects that will be used to compute I/M
credits for start emissions. The methodology used to compute repair effects on starts  and to
calculate I/M credits can be found in EPA report number M6.IM.001.

-------