United States       Air and Radiation      EPA420-P-99-009
           Environmental Protection              March 1999
           Agency                    M6.EXH.007
vvEPA     Determination of
           NOx and HC Basic
           Emission Rates,
           OBD and  I/M Effects for
           Tier 1 and Later
           LDVs and LDTs

           DRAFT
                              > Printed on Recycled Paper

-------
                                                                       EPA420-P-99-009
                                                                             March 1999
                      of               HC
                 and  I/M             for       1  and
                                    and

                              M6.EXH.007

                                John W. Koupal
                               Edward L. Glover

                        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.

-------
1      Introduction

       This report discusses proposed MOBILE6 NOx and exhaustHC basic emission rates (BERs)
for Tier 1 and later light-duty vehicles (LDVs) and light-duty trucks (LDTs), and the effects of On-
Board  Diagnostic (OBD) systems and Inspection/Maintenance (I/M) on these vehicles.  The
methodology discussed in this paper can be applied to generate BERs across all vehicle classes (LDV
and LDT1 through 4) for all Tier 1 and later standards, including the TLEV, LEV, and ULEV
standards under the California LEV I and Federal NLEV program, and LEV n standards recently
adopted by California. For brevity, however, the results presented here focus on the Tier 1, LEV and
ULEV standards for each vehicle class.

       Sufficient in-use data on LDVs or  LDTs  complying with Tier 1 or later standards were
generally not available at the time of the analysis. Thus, the methodology used in the development
of Tier 1 and later BERs is based on the differences in certification standards across standard level
and vehicle class. For NOx, Tier 1 and later BERs were based on a sample of 186 LDVs certified
to a 50,000 mile 0.4 gram/mile NOx standard (the Federal Tier 1 standard). For HC, Tier 1 and later
emission rates were based on BERs developed for 1988 through 1993 Ported Fuel  Injection (PFI)
LDVs.   Using certification standard as the base for Tier 1  and later BERs has two implicit
ramifications.  First, the BERs reflect the fuel which a vehicle is certified on the meet the  standard:
Indolene for Tier 1 standards, and California Phase IIRFGfor LEV and later standards. Second, the
HC BERs are  expressed as non-methane hydrocarbon (NMHC) for  Tier 1 LDV/LDTs,  and  non-
methane organic gas (NMOG) for LEV and later LDVs/LDTs.

       On-Board Diagnostics systems were required on all LDVs and LDTs sold outside California
beginning in 1996. Tier 1 vehicles began entering the fleet in 1994, and for two years (1994 and
1995) were not equipped with OBD. For MOBILE6, it will be assumed that all 1996 and later LDVs
and LDTs are equipped with OBD systems,  which  are  designed to  detect emission system
malfunctions resulting in emissions at or above 1.5 times the applicable emission standard.1 If this
criteria is  met, a light on the vehicle's dashboard (the malfunction indicator light, or MIL) is
illuminated to alert the driver that an emissions system repair is required. Thus, the rate of emission
deterioration for Tier 1 and later vehicles must take into account the impact OBD systems will have
overall in-use  emissions, including a) the  effectiveness of these systems in detecting  emission
malfunctions, b) the owner response rate to illuminated MILs, and c) the effectiveness of repair in
addressing the detected problem.

       Beginning in 2001, all Inspection/Maintenance programs will require an OBD system check
for OBD-equipped vehicles.  In I/M areas, this will greatly increase  the rate at which illuminated
MILs are addressed, hence further improving the average rate of in-use deterioration for Tier 1 and
       lrThe "1.5 times the standard" criteria was initially required by ARB, while EPA adopted a different
malfunction threshold approach.  However, manufacturers were allowed to meet the federal program through
compliance with ARB's requirement, and most chose this option.  EPA's requirement has recently been amended
to harmonize with ARB by requiring the "1.5 times the standard" criteria for vehicle sold federally. For MOBILE6,
it will be assumed that all vehicle equipped with OBD since 1996 comply with the "1.5 times"  malfunction criteria.

DRAFT M6.EXH.007                           1                                  March 1999

-------
later vehicles.

       Most Tier 1 and later vehicles will be equipped with an OBD system and, if in an I/M area,
subject to OBD-based I/M rather than traditional exhaust I/M. However, some Tier 1 vehicles will
not be equipped with OBD (model years 1994 and 1995), and some OBD-equipped vehicles will be
subject to an exhaust I/M program (prior to calendar year 2001).

       To model emissions under these scenarios, a methodology for generating basic emission rates
was developed for the following cases:

       No OBD/No EVI (Base) applies to pre-OBD Tier 1 vehicles (1994 and 1995 model years).
       It is also used as  a basis for the computation of BERs under the OBD-only and OBD/EVI
       cases.

       OBD-Onlv (OBD/ No EVT) applies to 1996 and later OBD-equipped vehicles where an I/M
       program is not present.

       OBD/IM applies to 1996 and later OBD-equipped vehicles where an I/M program which
       conducts  OBD checks is present.  An exhaust test may or may not be performed;  it is not
       differentiated from an I/M program with both  OBD checks  and exhaust testing, since
       additional I/M reductions are not given for exhaust testing if OBD checks are performed.

       Exhaust I/M represents a situation in which only  an exhaust test is conducted in an I/M
       program (an EVI240, ASM, or Idle test).  This will apply to 1994 and 1995 model year Tier
       1 vehicles in all calender years, and all vehicles in  calender years prior to 2001.

This report gives an overview of the basic approach for generating BERs, then describes the specific
details of BER development for NOx and exhaust HC for each case.

2      Basic Emission Rate Derivation Concept

       The basic concept underlying the generation  of Tier 1  and later BERs is similar to the
approach used to develop the I/M credits for 1981 through 1993 vehicles.2  For the No OBD/No EVI
case, this concept segregates in-use vehicles into "normal" and "high" emitters.  High emitters are
those vehicles which have emission control systems which are malfunctioning in some way, and are
producing average emission levels which are considerably higher than the overall mean emission
levels; the threshold for defining a high emitter for NOx and HC is 2.0 times the intermediate life
(50,000 mile) certification emission  standard.   The remainder of the fleet are considered to be
properly functioning, and are considered normal emitters; by definition, these vehicles are below 2.0
       Glover, E., and Brzezinksi, D., "MOBILE6 Inspection/Maintenance Benefits Methodology for 1981
through 1993 Model Year Light Vehicles", Draft MOBILE6 Report No. M6.IM.001, March 1999. Hereafter
referred to as "Tier 0 I/M Report"

DRAFT M6.EXH.007                          2                                 March 1999

-------
times the intermediate life certification standard.  It is important to note that both pollutants are
considered independently when determining whether a vehicle is a high emitter. Thus, a vehicle
could be a high NOx emitter, but a normal HC emitter.

       Although the segregation of vehicles into the "normal" and "high" categories (and their
thresholds) is a somewhat arbitrary modeling method, the concept that average in-use emissions are
driven by a group of vehicles emitting well above the applicable standard is supported by data from
many years of EPA vehicle test and repair programs.  This phenomena is contributed to by the
"go/no go" nature of emission control technology components such as EGR valves and air pumps,
and the high sensitivity of emissions to degradations in performance of other critical components
such as the catalyst, oxygen sensor(s) and fuel injectors.   Two important assumptions in the
development of BERs for Tier 1 and later vehicles are a) the rate at which vehicles malfunction and
become high emitters is independent of the certification standard level, and b) the average emission
levels for high emitters becomes higher relative to the standard&s the certification standard becomes
lower.

       Under this methodology, average in-use emissions are computed as a weighted average of
high and normal emitters.  Figure  lisa  general graphical view  of the  concept with the y-axis
representing emissions in grams per mile, (or grams for start emissions) and the x-axis representing
mileage.3 Three lines are shown in Figure 1 which show a) the average or basic emission rate, b) the
normal emitter emission rate, and c) the high emitter emission level.
                                        Figure 1
               GENERAL EMISSION FACTOR SCHEMATIC
        EMISSIONS
                                         MILEAGE
       MOBILE6 uses vehicle mileage as a surrogate for vehicle age. Age and mileage are used interchangeably
throughout this document.
DRAFT M6.EXH.007
                                                                           March 1999

-------
       The basic emission rate is shown as Line A.  This line represents the average emissions of
the fleet as a function of both normal emitters and high emitters.

       Line B represents the average emissions of normal-emitting vehicles. These are the vehicles
which have emission control systems which are generally performing as designed. The line is shown
as a linear function of mileage to reflect the gradual deterioration that normal vehicles experience,
primarily due to catalyst degradation over the life  of the vehicle. Normal emitter emissions are
generally expressed by a least squares linear regression of emissions versus mileage.

       Line C represents the average emissions of high-emitting vehicles.   This line is a flat
horizontal line because emissions from these vehicles do not appear to be a strong function of
mileage, based on previous analysis of Tier 0 data4 and born out by analyses of Tier 1 NOx data
presented in the following section.   The underlying phenomena expressed here is that emission
control malfunction will drive high emissions regardless of vehicle mileage; as discussed in
subsequent sections, what changes as the vehicle ages is the probability of malfunction, rather than
the emission levels resulting from a malfunction.

       Line A represents the weighted average of lines B and C, based on appropriate weighting
factors for normal and high emitters.  On a fleet-wide basis, this weighting factor represents the
fraction of high emitters in the  fleet, as a function of vehicle age; on a per-vehicle  basis, this
weighting factor can be considered to be the probability the vehicle will be a high emitter at a given
age.   This weighting factor can  be derived at any given vehicle age A (represented by vehicle
mileage) by transforming Equations 1 and 2 into Equations 3 and 4 below.
       4Tier 0 I/M report

DRAFT M6.EXH.007                           4                                  March 1999

-------
       Highs + Normals = 1                                               Eqn 1

and

       AVE = High_ave * Highs + Norm_ave * Normals                      Eqn 2

Solving for the variables Highs and Normals produces:

       Highs = (AVE - Norm_ave) / (High_ave - Norm_ave)                  Eqn 3

       Normals = 1  - Highs                                               Eqn 4

Where:

Highs = fraction of High emitters, age = A
Normals = fraction of Normal emitters, age =A
AVE is the average emission rate, age = A
High_ave is the high emitter emission average (independent of age)
Norm_ave is the normal emitter emission average, age = A

3      NOx BERs and Emitter Fractions: No IM/No OBD Case

3.1    TierlLDVs

       The No EVI / No OBD case was developed first because it did not require accounting for high
emitters which underwent repair due to OBD MIL-on; hence, the methodology closely followed the
basic emission rate derivation concept outlined in the previous section. Tier 1 LDVs served as the
basis for developing BERs across vehicle class (LOT 1 through 4), and standard level (LEV, ULEV)
for both NOx and NMHC. Thus,  the derivation of these BERs is the first step in the generation of
all BERs for all OBD/EVI cases, vehicles classes and standards.  Tier 1  and later BERs were first
generated in FTP space and subsequently split into running and start components, as discussed in
Section 7;  this deviates from the  approach used 1981 through 1993 vehicles, for which start and
running BERs were developed independently.

       The data used to generate No OBD/No EVI BERs for Tier 1 LDVs were 186 LDVs and LDTs
tested by the California Air Resources Board (ARE) as part of Surveillance Programs  12 through
14; these data were provided to EPA by ARB specifically for this analysis. The model years of these
vehicles ranged from 1989 through 1996, and all were certified to a 50,000 mile intermediate useful
life NOx standard of 0.4 g/mi, the same as the Federal Tier 1 LDV/LDT1 standard. Figure  2
graphically presents the ARB data, in terms of FTP emissions versus mileage (in ten thousand mile
units).
DRAFT M6.EXH.007                          5                                March 1999

-------
                   Figure 2 - ARE Surveillance Data FTP NOx Results
ARB Surveillance Data - FTP Results
0.4 NOx LDV/LDT
2.2
2.0 -
1.8
1.6 -
= 1.4 -
1 1.2
5
o 1-° •
0.6
0.4 -
0.2
On
*
•
•
•
High Emitters
•
«
«
•
* * *
• ** * 2,0 x Standard
• » « * * *
* * . *vv * * . • .*
»i »„*** %»»IX<.V» .-••:. * *
* * ^i* * «*i ^A **^ * * * Normal Emitters
* * JM»* *^ ^ *. » *^ ** * * .^ A. * *
* ^***»* *^ ^ ^ • **

0 2 4 6 8 10 12 14 16 18 20
Mileage/10000
       Using these data as a starting point, emissions for normal emitters, high emitters and the
fraction of high emitters were derived, as outlined in the following sections.

3.1.1   Normal and High Emitter Emission Level

       The average FTP normal emitter emission level was obtained by separating the normal
emitters from the high emitters in the ARB sample according to the "2.0 times the standard" criteria
(i.e. all vehicles in the sample above 0.8 g/mi were defined as high emitters, as shown in Figure 2).
Using this cutpoint, 19 vehicles were defined as high emitters, and the remainder (167) were defined
as normal emitters.  The BER for normal emitters was developed by fitting a linear regression of
FTP emissions versus mileage;  the result of this regression is shown in Equation 5 (the variable
'odom' is in units  often thousand miles).
       Norm_Ave(g/mi)
0.265 + 0.0147 * odom
Eqn 5
       The high emitter NOX emission level was obtained by calculating the mean emission level
of the 19 vehicles defined as High Emitters:
       High_Ave(g/mi)
1.278
Eqn 6
DRAFT M6.EXH.007
                                                                         March 1999

-------
3.1.2  High Emitter Fractions

       As discussed in Section 2, high emitter fractions allow the computation of average in-use
emissions based on normal and high emitting BERs. These fractions increase with vehicle age, and
for OBD-equipped vehicles will be used as the basis for computing OBD and/or I/M benefits.  It is
at this stage of the NOx BER computation that key assumptions regarding the representativeness of
recruitment-based5 in-use emission testing program become relevant.  In general, there is concern
that recruitment-based in-use emission testing programs are less likely to procure dirtier vehicles,
and hence do not gather data over the full range of the in-use fleet.  These programs typically have
low  response rates (less than 25 percent), and in-use data from programs with higher rates of
participation (e.g. I/M programs) show higher emissions than voluntary recruitment programs. As
such, for the purpose of developing MOBILE6 basic emission rates for 1981 through 1993 vehicles,
a "high emitter adjustment factor" was added to voluntary recruitment data to account the potential
bias  in recruitment.6 Since the ARE surveillance program is also a recruitment-based program, a
similar methodology was applied for Tier 1 and later vehicles

       In addition to a high-emitter correction factor, a second adjustment was required to "remove"
the impact of California's I/M program, which  he vehicles included in the ARB sample were
subjected to.  This was necessary in order to develop basic emission factors which reflected no I/M
program.

       Both the high-emitter and "no I/M" adjustments are relevant at this stage of the calculation
because they  only affected the weighting of high  and normal emitters  - in other words, they were
used to increase only the number of high emitters estimated in the fleet.  As discussed in Section 2
(and shown in Equations (1) through(4)), the high/normal weighting factor is a function of normal
and high emitter emission levels and their combined average in-use emission level. Emissions for
normal and high emitters  as computed  in Equations (5)  and (6) were  not affected  by these
adjustments.  The high-emitter and "no I/M" adjustments were instead applied directly to the in-use
average emission levels; based on normal and high emitter emissions computed from Equations (5)
and (6), the high emitter fractions were then derived using Equations (1) through (4).  This had the
effect of increasing the weighting of high emitters in the fleet relative to the pre-adjustment level.
This methodology is detailed in the following steps:

(1)    The unadjusted average in-use NOx emission level as a function of mileage was computed
       from the ARB dataset. This FTP NOx emission average was obtained by linear regression
       5"Recruitment-based" in-use emission testing programs are defined here as programs in which vehicles are
procured for testing from the general population in exchange for money and/or other incentives. Participants are
initially contacted through mail or phone solicitation based on registration mailing lists, and participation is strictly
voluntary.

       6Enns, P. et al, "Determination of Running Emissions as a Function of Mileage for 1981 through  1993
Model Year Light-Duty Cars and Trucks", MOBILE6 Report No. M6.EXH.OO. Referred to as "Tier 0 Running
BER Report"

DRAFT M6.EXH.007                            7                                  March 1999

-------
       of the raw ARB data (all 186 vehicles) versus mileage. The intercept of the regression was
       set equal to the normal emitter intercept from Equation (5) (0.265  g/mi).  The resulting
       regression equation is as shown:

       NOx FTP (g/mi)     =      0.265 + 0.031 * Odom                    Eqn 7
(2)    The second step was to generate an additive high emitter correction factor to account for the
       potential effects of recruitment bias on the ARB sample. The Tier 0 high emitter correction
       factor proposed for 1988 through 1993 PFILDVs (obtained from comparison of EVI240 data
       and data collected through voluntary recruitment programs)7 was calculated in FTP space,
       and fit with a linear regression as a function of mileage. The Tier 0 adjustment was reduced
       by 25 percent to account for reductions in high emitter emissions expected between Tier 0
       and Tier 1 vehicles, and to prevent unduly large  high emitter fractions. The resulting high
       emitter correction factor  is shown in Equation 8. It is in units of FTP grams per mile, and
       is a function of mileage.  This adjustment is added to the average NOx emission function
       from Step 1 above.
       HECF  =     0.00795 * Odom                                       Eqn 8a

       Where:

       HECF is the high emitter correction factor.
       Odom is the mileage, in ten thousands

       At zero miles, this adjustment adds 3 percent to the average in-use emission level; at 100,000
       miles it adds 14 percent.  These percent increases are comparable to those calculated from
       the data for 1988 through 1993 PFI LDVs with and without a high-emitter adjustment.

(3)    The third step was to generate a multiplicative "No I/M" correction factor to account for the
       I/M effect in the ARB sample. Most vehicles under California's Smog Check II program
       will undergo a basic idle I/M test,8 adjustment factors by vehicle age were developed by
       using MOBILES program with and without an idle I/M program (using by-model year output
       for a run in the year 2010).  At each vehicle age, a correction factor based on the ratio of the
       "without I/M" and "with I/M" runs was computed; these factors were regressed with mileage
       to create a correction factor as a function of mileage.  The resulting equations are shown as
       follows:
       7Tier 0 Running BER Report

       8"Smog Check Program Fact Sheet: A History of Inspection and Maintenance Programs in California",
Smog Check Web Site (http://smogcheck.ca.gov), July 1997

DRAFT M6.EXH.007                            8                                March 1999

-------
      NOIMC      =     1.057- 0.0038 * Odom                         Eqn 8b
      (NOIMC      >     1.00)

      NOIMC      =     1.00                                         EqnSc
      (NOIMC      <     1.00)

      NOIMC      =     0.00                                         EqnSd
      (Odom =      0)

(4)    The corrected in-use average NOx FTP results (C_NOXFTP) were obtained by applying the
      high emitter and No I/M correction factors from Equations 8a through 8d to the raw NOx
      FTP value from Equation 7, as shown in Equation 9.
       C_NOXFTP   =     (NOx FTP + HECF) * NOIMC                  Eqn 9
(5)     The fraction of high emitters in the fleet under NO I/M and NO OBD conditions (labeled
       HighBASE) is calculated by inserting the value of C_NOXFTP, Norm_ave, and High_ave
       (from Equations (5) and (6)) into Equation (3).  Mathematically, this is shown in Equation
       10.

       HignBASE = (C_NOXFTP - Norm_ave) / (High_ave - Norm_ave)         Eqn 10

       The resulting high emitter fractions for the No OBD/No EVI case for ages one through 25 are
       shown in Appendix A, Table A-l (mileage levels as a function of age are shown in Appendix
       A, Table A-3).

       The average in-use FTP-based NOx emission level for Tier 1 LDVs can be calculated at any
vehicle age using Equation (2), based on the terms "High_ave" (Equation 6), "Norm_ave" (Equation
5), and "Highs" (Equation 10).  According to Equation (4), "Normals" is simply 1 - Highs.
3.2    All Other Standard Levels and Vehicle Classes

      NOx  BERs for No OBD / No EVI conditions are required for LDVs under post-Tier  1
standards, and all Tier 1 and later LDTs (LDT1 through 4).  At the time of this analysis, EPA was
not aware of any dataset which provided an adequate sample of in-use data for these combinations
of vehicles class and standard level. BERs for these classes were derived from the Tier 1 LDVBERs
developed above, using a set of specific assumptions about how average emissions for normal and
high emitters, and high emitter fraction, would apply across standard level and class.
DRAFT M6.EXH.007                          9                               March 1999

-------
3.2.1   Normal Emitters

       It was assumed that for post-Tier 1 LDVs and Tier and later LDTs, normal emitter NOx
emissions will on average maintain the same performance relative to the applicable 50,000 mile
standard as Tier 1 LDVs.  Thus, normal emitter BERs for all post-Tier 1 LDVs and Tier 1 and later
LDTs were developed by calculating the ratio of the applicable standard level ("std") to the Tier 1
LDV standard, and applying this ratio to the Tier 1 BER (zero-mile level and deterioration rate), as
follows:

Norm_ave ZML(std) = Norm_ave ZML(tierlldv) * (Cert Std(std) / Cert Std(tierl Idv))              Eqn 1 la

Norm_ave DR(std)  =  Norm_ave DR(tierl Idv) * (Cert Std(std) / Cert Std(tierl Idv))                 Eqn 1 Ib

For example, normal-emitting LDV LEV BERs were generated by multiplying the normal-emitting
Tier 1  ZML and DR from Equation (5) by 0.5 (0.2 g/mi divided by 0.4 g/mi, the  50,000 miles
standards).

       The rationale behind this approach is that basic emission levels for properly operating
vehicles should receive the full benefit of reduced standards, including lower deterioration rates for
lower standard levels.  This approach presumes that normal emitters for all standards and vehicle
classes will on average achieve the same compliance margin ("headroom") with the 50,000 mile
certification standard as the normal emitters observed in the ARB data. With regards to trucks, this
approach acknowledges that LOT emission performance relative to the standard is expected to be
similar to LDVs because of increased similarities in a) emission control technology, b) manufacturer
design practices, and c) driving and usage patterns.

3.2.2   High Emitters

       High Emitter BERs are meant to estimate emissions from vehicles that significantly exceed
their certification standards  due to malfunctioning emission control systems. A key assumption in
the development of high-emitter BERs is that, as emission standards are lowered  (or "raised" for
truck BERs), emission levels for high emitters  will not be changed in proportion to the standard
change. Because emission  control equipment on these vehicles will  be degraded or completely
malfunctioning, high emitter emissions would be less dependent of certification standard level.  To
illustrate,  compliance with  LEV NOx standards  are expected to result from improved catalyst
performance rather than reductions in engine-out  emissions; as such, engine-out  emissions from
LEVs and Tier Is are likely more similar than the difference in standards, on a relative basis. A
degraded  catalyst  on a LEV would thus result in emissions more similar to a Tier 1 with a
comparably degraded catalyst.

       BERs for high emitters were developed for post-Tier 1 LDVs and Tier 1 and later LDTs by
taking the average of the Tier 1 LDV high emitter NOx BER (1.278 g/mi) and the BER that would
result if the ratio of 50,000 mile standards were applied to the Tier 1 BER, according to Equation
12:

DRAFT M6.EXH.007                           10                                 March 1999

-------
High_ave(std) = average[High_ave(tierlldv) * (Cert Std(std) / Cert Std(tierl Idv), High_ave(tierlldv)]    Eqn 12

       The result of this average is a high emitter BER which is 50 percent proportional to the
change in standard, reflecting that malfunctioning vehicles will derive some benefit on average from
lower emission standards, but not the full benefit as afforded to normal emitters.   High emitter
emissions are therefore tied closely to the Tier 1 LDV high emitter level, and the change in  high
emitter emissions is "muted" relative to changes in standard.

       Normal and high emitter BERs for LDVs and LDTs complying with Tier 1, LEV and ULEV
standards are presented in Appendix B.

3.2.3   High Emitter Fractions

       The rate at which vehicles become high emitters under the No OBD / No EVI scenario was
assumed constant for all vehicles and standard classes.  Thus, the age-based high emitter fractions
developed in Equation 10 and presented in Appendix A were applied to Tier 1 and later BERs for
all classes.  The rate of emission control  technology malfunction was assumed the same between
LDVs and LDTs, given that their emission technology and usage patterns are increasingly similar.
Reduced certification standards are also not expected to influence the rate at which emission control
technology malfunctions, because a) manufacturer's design and durability practices are not expected
to differ between Tier 1 and later  standards, and b)  many cases of emission control degradation
and/or malfunction are owner-induced, and hence outside the manufacturer's liability for in-use
emission performance.  It should be noted that the high-emitter fractions in Appendix A are shown
to vary by class, due to differences in accumulated mileage at a given age. At the same mileage, the
high emitter fractions are the same across all classes.

       The No OBD / NO I/M average in-use NOx emission rate for vehicle/standard = (V,S) can
be calculated at any vehicle age using Equation (2), based on a) the Tier 1 LDV "High_ave" and
"Norm_ave" terms from Equations (5) and (6) adjusted  as described above based on the (V,S)
standard level, and b) the base (No OBD / No EVI)  high emitter fractions from Appendix A, Table
A-l.
4      NMHC/NMOG BERs and Emitter Fractions: No IM/No OBD Case

       The development of NMHC/NMOG BERs shared many of the methodological assumptions
outlined for NOx in Section 3. As with NOx, NMHC BERs for Tier 1 and NMOG BERs for LEV
and later LDVs and LDTs were developed off of "base" LDV BERs; the primary difference between
the methodologies for the two pollutants was the source of the base data.  At the time of this analysis,
sufficient in-use data on vehicles complying with EPA's Tierl NMHC standards (for any vehicle
class) were not available.  The ARB surveillance dataset used for the NOx analysis included 58
LDVs certified to a 0.25 g/mi 50,000 mile NMHC standard, the same as Federal Tier 1. However,
these vehicles did not start entering the market until 1993; hence, most of the vehicles were of age
three years or less at the time of testing (the average mileage of this sample was roughly 38,000

DRAFT M6.EXH.007                           11                                March 1999

-------
miles, versus 58,000 for the 0.40 NOx sample). As such, these data were judged to be inadequate
for assessing overall in-use emission performance of Tier 1 LDVs.

       Tier 1 and later HC BERs were instead based on proposed MOBILE6 BERs for model year
1988 through 1993 Tier 0 LDVs with ported fuel inj ection (PFI). These BERs were developed based
on several thousand vehicles tested by auto manufacturers, EPA, and through I/M programs.  The
Tier 0 emission rates were considered a good starting point for developing Tier 1 and later BERs
because emission control technology used on later Tier 0 vehicles (e.g., 3-way catalysts and ported
fuel injection) are generally similar to those used on Tier 1 and later vehicles. A comprehensive
treatment of Tier 0 BERs and the datasets used to derive them are contained in the reports under
Reference 7, and thus are not presented here.  However, for this  analysis a simplifying step was
performed to generate a linear form of Tier 0 normal-emitter BERs, since as proposed for MOBILE6
these are expressed as nonlinear functions.  The resulting normal and high emitting Tier 0 BERs
(expressed as Total Hydrocarbon, or THC) used as a basis for Tier 1 and later BERs are shown in
Equations (13) and (14).

       Norm_Ave(g/mi)    =     0.16 +  0.0186 * odom                    Eqn 13

       High_Ave(g/mi)     =     2.076                                   Eqn 14

"odom" represents mileage in units often thousand miles.

       Using these values for normal and high emitters as a starting point, normal and high emitter
NMHC/NMOG BERs for all Tier 1 and later LDVs and LDTs were developed using the identical
methodology as for NOx (described in Sections 3.2.1 and 3.2.2) based on the ratio of the applicable
50,000 mile standard level to the Tier 0 level of 0.41  g/mi. Since Tier 1 standards are expressed as
NMHC and LEV and later standards are expressed as NMOG, the shift from THC to these pollutants
is accounted for in the standard ratios.

       The high emitter fractions developed for 1988 through 1993 Tier 0 PFI LDVs were used as
the No OBD/No EVI emitter fractions for all Tier 1 and later vehicles,  because of the expected
similarity in vehicle malfunction rates across standard level and vehicle class as discussed in Section
3.2.3.  These fractions are show in Appendix A, Table A-2.

       The No OBD / No EVI average in-use NMHC/NMOG emission rate for vehicle/standard =
(V,S) can be calculated at any vehicle age using Equation (2), based on a) the Tier 0 "High_ave" and
"Norm_ave" terms from Equations  (13) and (14) adjusted based on the (V,S) standard level, and b)
the base (No OBD / No EVI) high emitter fractions from Appendix A, Table A-2.

5      Effects of OBD and OBD-based  I/M for NOx and  HC

       Separate BERs were developed for all standard and vehicle classes to account for the effects
of OBD and OBD-based I/M programs. The methodology used to account for these programs were
identical for NOx and HC, based on reducing the fraction of high emitters in the fleet from the No

DRAFT M6.EXH.007                           12                               March 1999

-------
OBD/No IM case. Thus, emission levels for normal and high emitters were not changed under these
programs, only the fraction of highs in the fleet.  This methodology introduces a new category of
vehicle into the fleet: "Repaired" emitters. These vehicles are high emitters that are flagged by an
OBD system and undergo successful repair. For the OBD-only and OBD/EVI cases, these vehicles
are treated distinctly from normal and high emitters.

5.1    OBD Effectiveness

       OBD effectiveness is defined by three parameters: a) the probability the OBD system will
detect a failure (MIL-on Rate), b) the probability an  owner will respond to a MIL-on (Response
Rate), and c) the average after-repair emission level for responding vehicles (Repair Level). At the
time of this analysis, sufficient in-use data were not available to determine in-use patterns for these
parameters.  Thus, estimates were developed for the projected likelihood of malfunction detection,
owner response and repair level; stakeholder comments on these estimates are encouraged.  These
are presented in Sections 5.1.1 through 5.1.3 for both  the OBD-only and OBD/EVI cases.

5.1.1   MIL-on Rate

       For all vehicle  classes and standards, it is assumed that OBD  will  detect (i.e., set  an
appropriate code and illuminate the MIL) 85 percent  of high emitters. Because high emitters are
defined independently for HC and NOx, this response rate is assumed to apply equally to both
pollutants. The remaining 15 percent of fleet will not be identified, and thus will remain in the fleet
as high  emitters.  No deterioration in the ability of the  OBD system to correctly identify high
emitters is assumed. Because this parameter is solely dependent on the vehicle's OBD system, it is
the same for I/M and non-I/M areas.

5.1.2  Response Rate

       In order to obtain emission reductions from a vehicle equipped with an OBD system, not only
must the system correctly identify the vehicle, but the motorist must also respond to the MIL and
take corrective action in a timely manner. A key assumption for this analysis is that the response rate
in OBD-based I/M areas (where repair is required) will be much higher than in non-I/M areas. In
OBD-based I/M areas, the MIL-on response rate is assumed to be 90 percent over the entire life of
the vehicle. Response was not set at 100 percent to account for waivers from, or evasion of, the I/M
program.

       For non-IM areas, OBD response is assumed to  be a function of vehicle  warranty. It is
assumed that an owner is much more likely to respond to a MIL-on when repairs will be paid for by
the manufacturer. Three mileage bins were therefore developed: 1) 0 through 36,000, the standard
bumper-to-bumper warranty period; 2) 36,001 to 80,000, for  which federal law mandates that
catalysts and electronic  control modules (ECMs) remain under warranty; and 3) above 80,000, for
which no warranty is in effect (extended warranties are not accounted for in this methodology).

       Under 36,000 miles, it is assumed that 90 percent of MIL-on vehicles will be repaired. This

DRAFT M6.EXH.007                           13                                March 1999

-------
is based on the judgment that for new vehicles still under warranty, owners will have little hesitation
in addressing a MIL-on. The 10 percent loss accounts for a small percentage of owners who will not
respond to a MIL-on even with the warranty incentive.

       Between 36,000 miles and 80,000 miles, it is assumed that 10 percent of MIL-on vehicles
will be repaired. This response rate is greatly reduced from the pre-36,000 mile level to account for
the discontinuation of warranty coverage on several emission-related components (e.g. secondary
air, EGR, oxygen sensors, fuel injectors),  and reduced willingness of owners to make emission-
related repairs on an aging vehicle in the absence of an I/M program.

       Above 80,000 miles, it is assumed that no MIL-on vehicles will be repaired. This assumption
reflects the end of warranties, the lower economic value of the vehicle, and the (further) reduced
willingness of owners to make emission-related repairs in the absence of an I/M program.

5.1.3   Repair Level

       Implementation of OBD requires a third emitter category - "repaired" emitters. Repaired
emitters are assumed to have constant emissions at the after-repaired emission level,  although a
subset of these vehicles "migrate" back to the high emitter category. The emission level after an
OBD-induced repair is assumed to be 1.5 times the applicable 50,000 mile certification standard.
For Tier 1 LDVs and LDTs, this repair level is the required threshold for illuminating the MIL up
to 50,000 miles; the repair level is lower than the OBD threshold above 50,000 miles, since the 1.5
times the standard criteria applies to the full useful life standard above this mileage. For LEVs, OBD
detection is required at 1.75 times the full useful life standard at all mileages; hence, the estimated
repair level is below the OBD threshold at all mileages. Because repaired emitters are assumed not
to deteriorate for the remainder of their life (unless they migrate into the high emitter category), the
benefit of repair increases over the life of the vehicle.   Repaired emitter emissions levels are
generally lower than average normal emitter  emissions at higher mileages. The repaired emitter
"BERs" are presented across standard and vehicle class in Appendix B.
5.2    High Emitter Fractions for OBD and OBD-based I/M

       Equations 15 through 17 were used to calculate the high emitter growth rate under the OBD
and OBD-based I/M scenarios (HighOBD). Overall, the high emitter fraction in a given year is a
function of a) the number of high emitters in the previous year, b) the base high emitter "growth rate"
in the absence of OBD or I/M, and c) the OBD effectiveness assumptions outlined in Section 5.1.
The subscript T is the vehicle age. High(O) is assumed to be zero. MOBILE6 will assign a value
of 'odom' for each age 'i'.

       Nonhigh(i)          =      1.0 - HighBASE(i)                                Eqn 15

       Delta_High(i)        =      HighBASE(i)-HighBASE(i-l)                        Eqn 16


DRAFT M6.EXH.007                            14                                March 1999

-------
       Growth_High(i)      =     Delta_High / Nonhigh(i)                         Eqn 17
HighOBD(i)  = HighOBD(i-l) + [(l-OBD)*MIL*Growth_High(i)*(l-HighOBD(i-l))] +
             [(l-MIL)*Growth_High(i)*(l-HighOBD(i-l)]                     Eqn 18

Where:

HighOBD(0)   =     0.0
MIL         =     0.85
Nonhigh     =     the fraction of normal and repaired vehicles
Growth_High =     the growth rate of high emitters (or, the rate at which "nonhighs" migrate into
                    the high emitter category)

'OBD' is the OBD response rate; 0.90 for OBD-based I/M, and 0.90/0.10/0.0 for mileage bins (0 -
36K), (36K - 80K), and (80K+).

An elaboration on Equations 15 through 18 is as follows: for a given vehicle age, the fraction of high
emitters is a) the number of highs from the year before, plus b) the number of MIL-on highs added
in that year due to OBD nonresponse (a function of "nonresponse" rate, MIL-on rate, and the high
emitter growth rate applied to the available pool of normal and repaired vehicles), plus c) the number
of highs added in that year that the OBD system did not detect (a function of MIL-"off'  rate and the
high emitter growth rate applied to the available pool of "non-highs"). The high emitter growth rate
for a given year is the absolute increase in high emitters under the No OBD / No EVI case from the
previous year divided by the fraction on nonhighs - i.e., the available pool of vehicles which can
become high emitters.

       Once the high emitter fraction is calculated for the OBD or OBD/EVI cases, the fraction of
repaired emitters can be calculated as the difference between the fraction of high emitters that would
occur without OBD or I/M (HighBASE,  from Equation (10)) and the  fraction of high emitters with
OBD  and/or I/M from Equation 18.  In equation form,

Repaired(i)   =     HighBASE (i) - HighOBD (i)                               Eqn 19

The rate of normal emitters remains constant  between the No OBD / No I/M, OBD-only and
OBD/EVI case; only the number of high emitters decrease, directly replaced by repaired emitters. The
emitter fractions for normal, high and repaired emitters for the OBD Only and OBD/EVI cases are
shown in Appendix A (Tables A-l and A-2) for NOx and HC, by vehicle class.

5.3    BER Calculation for OBD and OBD-based I/M

       Calculation of average in-use FTP-based NOx and NMHC emission rates at a given vehicle
age for the OBD-Only and OBD/EVI cases are similar to the methodology for No OBD / No I/M
vehicles (Equation (2)); the primary  differences are a) use of HighOBD rather than HighBASE emitter

DRAFT  M6.EXH.007                           15                                March 1999

-------
fractions, and b) addition of a term to account for repaired emitters. As mentioned, the normal and
high emissions rates are unchanged from the No OBD / No I/M case.  This computation is as
follows:

AVE  =     HighOBD * High_ave + Normal *Norm_ave + Repaired * Rep_ave        Eqn 20
Rep_ave is 1.5 times the applicable 50,000 mile certification standard.

5.4    Results

       FTP-based BERs for the No OBD / No IM, OBD only and OBD/IM cases for Tier 1, LEV
and ULEV standards across all vehicle classes are shown in Appendix C for NOx and Appendix D
for NMHC/NMOG. The full range of standards that will be handled by MOBILE6 (primarily TLEV)
are not shown, but can be derived using the methodology  outlined in this paper. These charts
highlight a key outcome of the "50 percent proportional" treatment of high emitters discussed in
Section 3.2.2. BERs for standards below the base standard (Tier 1 LDV) are higher relative to the
applicable standard, but have more OBD and/or I/M benefit (on a percent basis) relative to Tier 1
LDVs. BERs for standards above the base standard (Tier 1 LDV) are lower relative to the applicable
standard, but have less OBD and/or I/M benefit (on a percent basis) relative to Tier 1 LDVs. This
is rooted in the assumption that average high emitter emissions are not reduced in proportion to the
standard change, thus increasing the disparity between normal and high emitter for lower standards.
As standards are lowered, high emitters will have a more  pronounced effect on average in-use
emissions, and repair of these high emitters will result in greater relative emission benefit.

6      NOx and HC BERs for Exhaust-Only I/M

       Since an OBD check is currently an unproven concept in an I/M program, some I/M credit
scenario must be developed  for those areas that use traditional  exhaust I/M test procedures.  This
scenario will be likely used frequently until  calendar year 2001.  By that time, it is assumed that I/M
test procedures utilizing OBD checks on vehicles equipped with OBD will be in place.

6.1    No OBD with Exhaust I/M

       The "No OBD / Exhaust I/M" emission levels for this scenario are calculated using the
methodology described in  draft MOBILE6  document M6.EVI.001 ("MOBILE6 Inspection /
Maintenance Benefits Methodology for 1981 through 1993 Model Year Light Vehicles").  This
methodology utilizes I/M exhaust test identification rates and after repair effectiveness levels based
on data collected from the Arizona I/M program.  The "No I/M"  and the "With Exhaust I/M"
emission rates are used to calculate the I/M emission level and I/M credits for situations where
exhaust-only I/M tests are being performed on Tier 1 vehicles without OBD. The only vehicles
which will fall in this category are the 1994 and 1995 model years certified to Tierl standards.  In
this case, the structure of the I/M credits is identical to the Tier 0 I/M credits with the exception that
the Norm_ave, High_ave,  and fraction of highs in the fleet (HighBASE) are different from analogous

DRAFT M6.EXH.007                           16                               March 1999

-------
Tier 0 parameters. Equation 21 defines this case mathematically.

 AVE =  Norm_ave*(l-HighBASE) + High_ave*HighBASE*(l-IDR)  + HighBASE*IDR*W*High_ave*RW +
              Norm_ave*R*HighBASE*IDR*FIX + High_ave*HighBASE*IDR*NC                Eqn21

IDR is the identification rate of high emitters using an exhaust emission test.
R is the after repair emission level of vehicles repaired to pass an exhaust I/M test.
Fix is the fraction of vehicles which are repaired.
NC is the fraction of vehicles which are in non-compliance following their I/M test
W is the fraction of vehicles which receive a cost or other type of waiver.
RW is the after repair level of the vehicles which get waived. It is shown as a fraction of the high emitter level.

(see report M6.IM.001 for a full explanation of these terms).

6.2    OBD and Exhaust I/M

       In  this scenario, the vehicles in the fleet are  OBD compliant, but the state continues to
conduct an exhaust I/M test; this is  most likely scenario prior to calendar year 2001.  For this
scenario, the same I/M equations and assumptions used to model the 1981-93 Tier 0 vehicles are
used.  The primary difference is the fraction of high emitters is reduced somewhat due to the effects
of OBD program (i.e., HighOBD is  substituted for a higher rate of high emitters used in TierO.

AVE   =      Norm_ave*(l-HighBASE)     +   Rep_ave* (Repaired)  +  High_ave*HighOBD*(l-IDR) +
              HighoBD*IDR*W*High_ave*RW +
              Norm_ave*R*HighOBD*IDR*FIX + High_ave*HighoBD*IDR*NC          Eqn22
7      Derivation of Running and Start BERs for NOx and HC

       MOBILE6 requires a split between running and start emissions; because Tier 1  and later
BERs were developed on an FTP-basis, factors will be required allow derivation of running and start
BERs from FTP-based BERs.  Running fractions were first derived by applying the methodology
outlined in MOBILE6 report number M6. STE.002 ("The Determination of Hot Running Emissions
From FTP Bag Emissions") to separate Tier 1 and LEV datasets. For Tier Is, a sample of 31 LDVs
and LDTs vehicles tested by auto manufacturers as part of the SFTP rulemaking were used.9 For
LEVs, a sample of 34 LDVs and LDTs tested by auto manufacturers, CRC and  API and used for the
development of proposed MOBILE6 fuel sulfur effects.10  Based on an analysis of separate factors
for LDVs and LDTs, a  single factor which applied to both LDVs and LDTs was judged to be
adequate.  For NOx, the  resulting Tier 1 and LEV factors were found to be identical; the following
equation was therefore used to generate running BERs for all standards and classes:
       9SFTP Data CD
       10Rao, V., "Fuel Sulfur Effects On Exhaust Emissions", EPA Draft MOBILE6 Report No. M6.FUL.001

DRAFT M6.EXH.007                            17                                 March 1999

-------
       Running NOx BER (g/mi)          =      0.60 * FTP NOx BER       Eqn 23

       For NMHC/NMOG, separate factors were developed for Tier Is and LEVs:

       Tier 1 Running NMHC BER (g/mi)  =      0.16  *FTPNMHCBER          Eqn 24
       LEV Running NMOG BER (g/mi)  =      0.30  *FTPNMOGBER          Eqn 25

In MOBILE6, start BERs are related to FTP and Running BERs as shown in Equation (26):11

FTP BER = (Running BER*7.5 +Start BER*0.43+Start BER*0.57*0.16) / 7.5          Eqn 26

Where:

Running BERs =    the results of Equations (23) through (25)
7.5            =     total miles of the LA4
0.43/0.57     =     Bag I/Bag 3 weighting across total FTP
0.16         =     the ratio of Bag 3 emissions to Bag 1 emissions, based on HC 1988-1993 PFI
                   BERs proposed for MOBILE6 (applied to both HC and NOx)12
Using this equation and the running and FTP BERs developed above, start factors were derived
according to Equation (27):

       Tier I/LEV Start NOx BERs (grams)       =     5.75 * FTP BER           Eqn 27
       Tier 1 Start NMHC BER (grams)          =     12.1* FTP BER           Eqn 27
       LEV Start NMOG BER (grams)           =     10.1* FTP BER           Eqn 27

These factors were applied equally to each emitter class: normal, high and repaired. The resulting
BERs for Tier 1, LEV and ULEV across all  classes are shown in Appendix B.
       nGlover, E., and Carey, P., "Determination of Start Emissions as a Function of Mileage and Soak Time for
1981-1993 Model Year Light-Duty Vehicles", MOBILE6 report number M6.STE.003

       12This factor varies somewhat across pollutant and vehicle grouping, but was held constant for the
derivation of start factors.

DRAFT M6.EXH.007                          18                               March 1999

-------
                     APPENDIX A:
a) Emitter Fractions: No OBD/No IM, OBD Only, OBD/IM Cases
          b) Vehicle Mileage As A Function of Age

-------
Table A-l: NOx Emitter Fractions
Age
(Years)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
LDV
All
Normal
1.000
0.946
0.909
0.872
0.836
0.801
0.767
0.734
0.701
0.670
0.639
0.609
0.580
0.552
0.524
0.494
0.465
0.436
0.409
0.382
0.356
0.330
0.305
0.281
0.258
0.236
Base
High
0.000
0.054
0.091
0.128
0.164
0.199
0.233
0.266
0.299
0.330
0.361
0.391
0.420
0.448
0.476
0.506
0.535
0.564
0.591
0.618
0.644
0.670
0.695
0.719
0.742
0.764
OBD Only
High
0.000
0.013
0.022
0.058
0.094
0.128
0.162
0.199
0.234
0.268
0.302
0.334
0.366
0.397
0.428
0.461
0.492
0.523
0.553
0.583
0.611
0.639
0.666
0.693
0.718
0.743
Repair
0.000
0.041
0.069
0.070
0.070
0.071
0.071
0.068
0.065
0.062
0.059
0.056
0.054
0.051
0.048
0.046
0.043
0.040
0.038
0.035
0.033
0.031
0.028
0.026
0.024
0.022
OBD/IM
High
0.000
0.013
0.022
0.031
0.041
0.050
0.059
0.069
0.079
0.088
0.098
0.108
0.118
0.128
0.139
0.150
0.162
0.174
0.186
0.199
0.212
0.225
0.239
0.253
0.267
0.282
Repair
0.000
0.041
0.069
0.097
0.123
0.149
0.174
0.197
0.220
0.242
0.263
0.283
0.302
0.320
0.338
0.356
0.373
0.390
0.405
0.419
0.433
0.445
0.456
0.466
0.475
0.482
LOT 1/2
All
Normal
1.000
0.934
0.885
0.837
0.790
0.744
0.700
0.657
0.617
0.577
0.540
0.500
0.462
0.426
0.391
0.359
0.328
0.300
0.273
0.249
0.227
0.206
0.188
0.171
0.156
0.142
Base
High
0.000
0.066
0.115
0.163
0.210
0.256
0.300
0.343
0.383
0.423
0.460
0.500
0.538
0.574
0.609
0.641
0.672
0.700
0.727
0.751
0.773
0.794
0.812
0.829
0.844
0.858
OBD Only
High
0.000
0.015
0.063
0.110
0.156
0.204
0.251
0.297
0.341
0.383
0.423
0.465
0.506
0.545
0.581
0.616
0.649
0.679
0.708
0.734
0.758
0.779
0.799
0.817
0.833
0.848
Repair
0.000
0.050
0.052
0.054
0.055
0.052
0.049
0.046
0.043
0.040
0.037
0.035
0.032
0.030
0.027
0.025
0.023
0.021
0.019
0.017
0.016
0.014
0.013
0.012
0.011
0.010
OBD/IM
High
0.000
0.015
0.028
0.040
0.053
0.066
0.079
0.092
0.105
0.118
0.132
0.147
0.162
0.178
0.193
0.209
0.225
0.241
0.256
0.272
0.287
0.302
0.317
0.331
0.345
0.359
Repair
0.000
0.050
0.088
0.123
0.158
0.190
0.221
0.251
0.278
0.304
0.328
0.353
0.376
0.397
0.415
0.432
0.447
0.459
0.470
0.479
0.486
0.491
0.495
0.498
0.499
0.499
LDT3/4
All
Normal
1.000
0.929
0.876
0.824
0.773
0.725
0.678
0.633
0.590
0.548
0.505
0.461
0.419
0.378
0.339
0.302
0.265
0.231
0.197
0.165
0.135
0.106
0.079
0.052
0.027
0.004
Base
High
0.000
0.071
0.124
0.176
0.227
0.275
0.322
0.367
0.410
0.452
0.495
0.539
0.581
0.622
0.661
0.698
0.735
0.769
0.803
0.835
0.865
0.894
0.921
0.948
0.973
0.996
OBD Only
High
0.000
0.017
0.069
0.119
0.168
0.221
0.271
0.319
0.366
0.411
0.457
0.504
0.549
0.593
0.635
0.676
0.715
0.752
0.788
0.822
0.855
0.886
0.916
0.944
0.971
0.996
Repair
0.000
0.054
0.056
0.057
0.058
0.055
0.051
0.048
0.044
0.041
0.038
0.035
0.032
0.028
0.026
0.023
0.020
0.017
0.015
0.012
0.010
0.008
0.006
0.004
0.002
0.000
OBD/IM
High
0.000
0.017
0.030
0.043
0.057
0.071
0.085
0.099
0.114
0.129
0.145
0.162
0.180
0.199
0.218
0.239
0.260
0.283
0.307
0.334
0.362
0.394
0.431
0.476
0.535
0.629
Repair
0.000
0.054
0.094
0.133
0.169
0.204
0.237
0.267
0.296
0.323
0.351
0.377
0.401
0.423
0.442
0.460
0.474
0.486
0.495
0.501
0.503
0.499
0.490
0.472
0.438
0.367

-------
Table A-2: NMHC/NMOG Emitter Fractions
Age
(Years)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
LDV
All
Normal
0.982
0.975
0.958
0.934
0.910
0.887
0.865
0.843
0.822
0.801
0.782
0.763
0.744
0.727
0.710
0.693
0.677
0.662
0.647
0.633
0.620
0.606
0.594
0.582
0.570
0.559
Base
High
0.018
0.025
0.042
0.066
0.090
0.113
0.135
0.157
0.178
0.199
0.218
0.237
0.256
0.273
0.290
0.307
0.323
0.338
0.353
0.367
0.380
0.394
0.406
0.418
0.430
0.441
OBD Only
High
0.004
0.006
0.010
0.033
0.056
0.077
0.099
0.121
0.143
0.164
0.185
0.205
0.224
0.242
0.260
0.277
0.294
0.310
0.325
0.340
0.354
0.368
0.381
0.393
0.405
0.417
Repair
0.014
0.019
0.032
0.033
0.035
0.036
0.037
0.036
0.035
0.034
0.033
0.032
0.032
0.031
0.030
0.029
0.029
0.028
0.028
0.027
0.026
0.026
0.025
0.025
0.024
0.024
OBD/IM
High
0.004
0.006
0.010
0.016
0.022
0.028
0.033
0.039
0.045
0.050
0.056
0.061
0.066
0.072
0.077
0.082
0.087
0.092
0.096
0.101
0.106
0.110
0.114
0.118
0.123
0.127
Repair
0.014
0.019
0.032
0.050
0.068
0.086
0.102
0.118
0.133
0.148
0.163
0.176
0.189
0.202
0.214
0.225
0.236
0.246
0.256
0.266
0.275
0.284
0.292
0.300
0.307
0.314
LOT 1/2
All
Normal
0.982
0.969
0.942
0.910
0.879
0.850
0.821
0.794
0.767
0.743
0.719
0.697
0.676
0.656
0.638
0.621
0.606
0.591
0.578
0.566
0.555
0.545
0.536
0.528
0.521
0.515
Base
High
0.018
0.031
0.058
0.090
0.121
0.150
0.179
0.206
0.233
0.257
0.281
0.303
0.324
0.344
0.362
0.379
0.394
0.409
0.422
0.434
0.445
0.455
0.464
0.472
0.479
0.485
OBD Only
High
0.004
0.007
0.033
0.063
0.092
0.123
0.152
0.180
0.207
0.233
0.257
0.280
0.302
0.322
0.341
0.358
0.375
0.390
0.403
0.416
0.427
0.437
0.446
0.455
0.462
0.469
Repair
0.014
0.023
0.025
0.027
0.029
0.028
0.027
0.026
0.025
0.024
0.024
0.023
0.022
0.021
0.021
0.020
0.020
0.019
0.019
0.019
0.018
0.018
0.018
0.017
0.017
0.017
OBD/IM
High
0.004
0.007
0.014
0.022
0.029
0.037
0.045
0.052
0.060
0.067
0.074
0.080
0.087
0.093
0.099
0.105
0.110
0.115
0.120
0.124
0.128
0.132
0.135
0.138
0.141
0.143
Repair
0.014
0.023
0.044
0.068
0.091
0.113
0.134
0.154
0.173
0.191
0.207
0.223
0.237
0.251
0.263
0.274
0.284
0.294
0.302
0.310
0.317
0.323
0.329
0.334
0.338
0.342
LDT3/4
All
Normal
0.982
0.967
0.936
0.902
0.869
0.837
0.807
0.778
0.750
0.724
0.699
0.675
0.653
0.631
0.611
0.592
0.574
0.557
0.541
0.526
0.511
0.498
0.485
0.473
0.462
0.452
Base
High
0.018
0.033
0.064
0.098
0.131
0.163
0.193
0.222
0.250
0.276
0.301
0.325
0.347
0.369
0.389
0.408
0.426
0.443
0.459
0.474
0.489
0.502
0.515
0.527
0.538
0.548
OBD Only
High
0.004
0.008
0.037
0.069
0.100
0.133
0.165
0.195
0.223
0.250
0.276
0.301
0.324
0.346
0.367
0.387
0.406
0.423
0.440
0.456
0.471
0.485
0.498
0.510
0.522
0.532
Repair
0.014
0.025
0.027
0.029
0.031
0.030
0.029
0.028
0.027
0.026
0.025
0.024
0.023
0.022
0.022
0.021
0.020
0.020
0.019
0.019
0.018
0.018
0.017
0.017
0.016
0.016
OBD/IM
High
0.004
0.008
0.015
0.024
0.032
0.040
0.049
0.057
0.064
0.072
0.080
0.087
0.094
0.101
0.108
0.115
0.121
0.127
0.133
0.139
0.144
0.149
0.155
0.159
0.164
0.168
Repair
0.014
0.025
0.049
0.074
0.099
0.122
0.145
0.165
0.185
0.204
0.221
0.238
0.253
0.267
0.281
0.294
0.305
0.316
0.326
0.336
0.345
0.353
0.360
0.367
0.374
0.380

-------
Table A-3: Draft MOBILE6 Cumulative Mileages (10,000 miles)
Age (Years)
1
2
3
4
5
6
7
8
9
10
11
12
13
LDV
1.491
2.908
4.256
5.537
6.755
7.912
9.013
10.059
11.054
12.000
12.899
13.753
14.566
LDT1/2
1.950
3.788
5.519
7.146
8.672
10.100
11.436
12.681
13.839
14.914
15.910
16.829
17.676
LDT3/4
2.133
4.120
5.970
7.692
9.297
10.791
12.183
13.478
14.685
15.809
16.856
17.830
18.738
Age (Years)
14
15
16
17
18
19
20
21
22
23
24
25

LDV
15.338
16.072
16.770
17.434
18.064
18.664
19.234
19.776
20.291
20.781
21.247
21.690

LDT1/2
18.453
19.165
19.815
20.406
20.941
21.425
21.861
22.252
22.602
22.914
23.191
23.438

LDT3/4
19.583
20.371
21.104
21.786
22.422
23.014
23.566
24.079
24.557
25.003
25.418
25.804


-------
              APPENDIX B:
TIER 1, LEV and ULEV BERs By Emitter Category

-------
Table B-l: NOx Basic Emission Rates
Vehicle
Class
LDV/T1
LDT2/3
LDT4
Standard
Class
Tierl
LEV/ULEV
Tierl
LEV/ULEV
Tierl
LEV/ULEV
50K
Standard
(g/mi)
0.4
0.2
0.7
0.4
1.1
0.6
Mode
FTP
Running
Start (grams)
FTP
Running
Start (grams)
FTP
Running
Start (grams)
"Normal" BER
(g/mi)
ZML
0.265
0.159
1.524
0.133
0.080
0.765
0.464
0.278
2.668
DR
0.0147
0.0088
0.0845
0.0074
0.0044
0.0426
0.0257
0.0154
0.1478
"High"
BER
(g/mi)
1.28
0.77
7.35
0.96
0.58
5.51
1.76
1.05
10.10
"Repaired"
BER
(g/mi)
0.600
0.360
3.450
0.300
0.180
1.725
1.050
0.630
6.038
SAME AS LDV/T1 TIER 1
FTP
Running
Start (grams)
FTP
Running
Start (grams')
0.730
0.438
4.198
0.398
0.239
2.289
0.0404
0.0242
0.2323
0.0220
0.0132
0.1265
2.40
1.44
13.78
1.60
0.96
9.200
1.650
0.990
9.488
0.900
0.540
5.175

-------
Table B-2: NMHC/NMOG
Vehicle
Class




LDV/T1








LDT2







LDT3







LDT4





Standard
Class

Tierl


LEV


ULEV


Tierl


LEV


ULEV


Tierl


LEV

ULEV

Tierl



LEV


ULEV

50K
Standard
(g/mi)

0.25


0.075


0.04


0.32


0.10


0.05


0.32


0.16

0.10

0.39



0.195


0.117

Mode
FTP
Running
Start (grams)
FTP
Running
Start (grams)
FTP
Running
Start (grams)
FTP
Running
Start (grams)
FTP
Running
Start (grams)
FTP
Running
Start (grams)
FTP
Running
Start (grams)
FTP
Running
Start (grams)
Basic Emission Rates
"Normal" BER
(g/mi)
ZML
0.098
0.016
1.183
0.029
0.009
0.293
0.016
0.005
0.162
0.125
0.020
1.509
0.039
0.012
0.394
0.020
0.006
0.197
0.125
0.020
1.509
0.063
0.019
0.636
DR
0.0113
0.0018
0.1364
0.0034
0.0010
0.0343
0.0018
0.0005
0.0182
0.0145
0.0023
0.1750
0.0045
0.0013
0.0455
0.0023
0.0007
0.0229
0.0145
0.0023
0.1750
0.0073
0.0022
0.0737
SAME AS LDT2
FTP
Running
Start (grams)
FTP

Running
Start (grams)
FTP
Running
Start f crams')
0.152
0.024
1.835
0.076

0.023
0.770
0.046
0.014
0.462
0.0177
0.0029
0.2136
0.0088

0.0026
0.0889
0.0053
0.0016
0.0535
"High"
BER
(g/mi)
1.67
0.27
20.17
1.23
0.37
12.40
1.14
0.34
11.51
1.85
0.30
22.31
1.29
0.38
13.04
1.17
0.35
11.77
1.85
0.30
22.31
1.44
0.43
14.54
LEV
2.03
0.33
24.44
1.53

0.46
15.45
1.33
0.40
13.43
"Repaired"
BER
(g/mi)
0.375
0.060
4.526
0.113
0.034
1.141
0.060
0.018
0.606
0.480
0.077
5.794
0.150
0.045
1.515
0.075
0.022
0.758
0.480
0.077
5.794
0.240
0.072
2.424

0.585
0.094
7.061
0.293

0.087
2.954
0.176
0.052
1.773

-------
                 APPENDIX C:
NOx FTP-Based Composite BERs for Tier 1, LEV and ULEV

-------
                                 LDV/LDT1 Tier 1 NOx
o.o
   0.0
2.5
5.0
7.5        10.0
      Mileage/10000
12.5        15.0       17.5        20.0
                     •NoOBD/ No
                                  •OBD/No
                                           - - - OBD/

-------
                             LDV/LDT1 LEV/ULEV NOx
o.o
  0.0
2.5
5.0
7.5        10.0
      Mileage/10000
12.5
15.0
17.5
20.0
                     •NoOBD/ No
                                 •OBD/No
                                         - - - OBD/

-------
                                     LDT2/LDT3 Tier 1 NOx
   1.6



   1.4



   1.2
«  1.0
E
"55
|  0.8
O)
o.
t  0.6
   0.4
   0.2
   0.0
     0.0
2.5
5.0
7.5         10.0

      Mileage/10000
12.5
15.0
17.5
20.0
                         •NoOBD/ No
                                    •OBD/No
                                             - -  - OBD/

-------
                             LDT2/LDT3 LEV/ULEV NOx
o.o
  0.0
2.5
5.0
7.5        10.0
      Mileage/10000
12.5       15.0       17.5       20.0
                     •NoOBD/ No
                                 •OBD/No
                                         - - - OBD/

-------
                                     LDT4 Tier 1 NOx
o.o
   0.0         2.5         5.0         7.5
    10.0        12.5
Mileage/10000
15.0        17.5        20.0
                      •NoOBD/ No
       •OBD/No
 - - - OBD/

-------
                                     LDT4 LEV/ULEV NOx
   1.4
   1.2
   1.0
I  0.8
to
CO
   0.4
   0.2
   0.0
     0.0
2.5
5.0
7.5         10.0
      Mileage/10000
12.5
15.0
17.5
20.0
                         •NoOBD/ No
                                   •OBD/No
                                            - - - OBD/

-------
                APPENDIX D (Corrected):
NMHC/NMOG FTP-Based Composite BERs for Tier 1, LEV and
                       ULEV

-------
                               LDV/LDT1 Tierl NMHC
o.o
  0.0
2.5
5.0
7.5        10.0
      Mileage/10000
12.5
15.0
17.5
20.0
                     •NoOBD/ NolM
                                  •OBD/No
                                          - - - OBD/

-------
                                 LDV/LDT1  LEV NMOG
o.o
   0.0
2.5
5.0
7.5        10.0       12.5
      Mileage/10000
15.0        17.5
20.0
                      •NoOBD/ NolM
                                   •OBD/No
                                           - -  - OBD/

-------
                               LDV/LDT1 ULEV NMOG
o.o
   0.0
2.5
5.0
7.5        10.0        12.5
      Mileage/10000
15.0        17.5
20.0
                     •NoOBD/ NolM
                                  •OBD/No
                                           - - - OBD/

-------
                               LDT2/LDT3 Tier 1 NMHC
o.o
  0.0
2.5
5.0
7.5        10.0        12.5
      Mileage/10000
15.0       17.5
20.0
                     •NoOBD/ NolM
                                  •OBD/No
                                          - - - OBD/

-------
                                LDT2/LDT3 LEV NMOG
o.o
   0.0
2.5
5.0
7.5        10.0        12.5
      Mileage/10000
15.0
17.5
20.0
                      •NoOBD/ NolM
                                   •OBD/No
                                           - - - OBD/

-------
                               LDT2/LDT3 ULEV NMOG
o.o
  0.0
2.5
5.0
7.5        10.0        12.5
      Mileage/10000
15.0        17.5
20.0
                     •NoOBD/ NolM
                                  •OBD/No
                                          - -  - OBD/

-------
                                    LDT4 Tier 1 NMHC
o.o
   0.0
2.5
5.0
7.5         10.0        12.5
      Mileage/10000
15.0        17.5
20.0
                       •NoOBD/ NolM
                                    •OBD/No
                                             -  - - OBD/

-------
                                     LDT4 LEV NMOG
o.o
   0.0
2.5
5.0
7.5         10.0
      Mileage/10000
12.5
15.0
17.5
20.0
                      •NoOBD/ NolM
                                    •OBD/No
                                             -  - - OBD/

-------
                                   LDT4 ULEV NMOG
o.o
   0.0
2.5
5.0
7.5         10.0       12.5
      Mileage/10000
15.0
17.5
20.0
                      •NoOBD/ NolM
                                   •OBD/No
                                            - - - OBD/

-------