Brake and Tire Wear Emissions from
On-road Vehicles in MOVES2014
oEPA
United States
Environmental Proloclion
Agency
-------
Brake and Tire Wear Emissions from
On-road Vehicles in MOVES2014
Assessment and Standards Division
Office of Transportation and Air Quality
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
that 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.
United States
Environm«nl.»l ProlecliQn
Agency
EPA-420-R-14-013
December 2014
-------
Table of Contents
1 Introduction 2
2 Brakewear 2
2.1 Literature Review 2
2.2 Developing Rates for MOVES 5
2.2.1 Emissions during braking 5
2.2.2 Activity 8
2.2.3 Emission Rate for Light Duty vehicles 11
2.2.4 PMio/PM2.5 Brake Wear Ratio 12
2.2.5 Brake Wear Emissions for Heavy-Duty Vehicles and Other Vehicle Types 13
3 Tirewear 17
3.1 Introduction 17
3.2 Methodology 20
3.2.1 PMio/PM2.5 Tire Wear Ratio 24
4 Next Steps 24
Appendix A Deceleration from PERE 26
Appendix B Brake and Tire Wear Emission Rates 29
Appendix C Literature Review 35
Appendix D Responses to Peer-Review Comments 37
References 53
-------
1 Introduction
The mobile source paniculate matter inventory includes exhaust emissions and non-exhaust
emissions. Exhaust emissions include particulate matter attributable to engine related processes
such as fuel combustion, burnt oil, and other particles that exit the tailpipe. Non-exhaust
processes include brake wear, tire wear, suspension or resuspension of road dust, and other
sources. Particulate matter from brakes and tires is defined as the airborne portion of the "wear"
that can be created by abrasion, corrosion, and turbulence. These wear processes can result in
particles being suspended in the atmosphere. The size, chemical composition, and emission rate
of particles arising from such sources contributes to atmospheric particle concentrations.
However, these particles are composed of different species and size than exhaust particulate
matter.l
The literature review for the development of the brake and tire wear emission models was
conducted in 2006 and 2007, the models were developed for MOVES2010, and this report was
written in 2008. However this documentation was not revised until the peer review complete in
2014 and no revisions to the model were made. As of 2007, the references in this report were
recent, yet there were likely a few publications on particulate matter from brake and tire wear
which were not included in the original literature review. In the sections below we present the
studies from the literature conducted at the time, as well as the models that were developed based
on the best data presented in the papers cited. A more recent literature search and a potential
model update will be conducted in the future.
2 Brakewear
2.1 Literature Review
There are two main types of brakes used in conventional (or non-hybrid electric) vehicles: disc
brakes and drum brakes. In a drum brake the components are housed in a round drum that
rotates with the wheel. Inside the drum are shoes that, when the brake pedal is pressed, force the
shoes against the drum and slow the wheel. By contrast, disc brakes use an external rotor and
caliper to halt wheel movement. Within the caliper are brake pads on each side of the rotor that
clamp together when the brake pedal is pressed.2
The definition of wear versus airborne PM seems to have slightly different definitions in the
literature. In this paper it is generally the mass of material lost, whether in the brake pads or the
tires. A fraction of that wear is airborne PM. Some studies look at both wear and airborne PM,
others look at one or the other. In brakes, the composition of the brakeliner has an influence on
the quantity and makeup of the released particles. Disc brakes are lined with brake pads while
drum brakes use brake-shoes or friction linings. These materials differ in their rate of wear, their
portion of wear particles that become airborne, and the size as well as composition of those
particles. Both types of brakes use frictional processes to resist inertial vehicle motion. The
action of braking results in wear and consequent release of a wide variety of materials
(elemental, organic and inorganic compounds) into the environment.
The overall size or mass of the brake pads also varies with vehicle type. Typically trucks use
larger brakes than passenger vehicles because the mass of vehicle that requires slowing down or
stopping is greater. In 2004, most light duty vehicles used disc brakes in the front and drum
-------
brakes in the rear. Disc brakes tend to have improved braking performance compared to drum
brakes and have correspondingly higher cost. Disc brakes are sometimes used on rear wheels as
well for higher performance (sportier) vehicles.
As a complicating issue, the particulate matter from brakes is dependent on the geometry of the
brakes, wheels and rims. The air flow through the rims to cool the brakes and rotors play a key
role in determining the wear characteristics. The emissions are also sensitive to driver activity
patterns, where more aggressive stop and go driving will naturally cause greater wear and
emissions.
There are only a very limited number of publications on brake wear PM emissions. There are
even fewer publications discussing size distributions and speciation, and none quantifying
emissions modally on which to directly base a model. This section summarizes the limited
literature as of 2006. More details of the literature on brake and tire wear can be found in
Appendix D. One of the earliest studies on brake wear emissions was done in 1983.3 Particulate
emissions from asbestos-based brakes from automobiles were measured under conditions
simulating downtown city driving. The report presented a systematic approach to simulating
brake applications and defining parti culate emissions, and was used in the development of the
EPA PARTS model.4 For PARTS, EPA calculated PMio emission factors for light-duty gasoline
vehicles of 12.5 mg/mi for brake wear. Since 1985, the asbestos in brakes has been replaced by
other materials, and newer studies have been conducted. These factors suggest the need for this
update of the emission factors applicable to more modern vehicles.
Garg et al. (2000) conducted a study in which a brake dynamometer was used to generate wear
particles under four wear conditions (much of the background information provided in the
previous paragraphs are from this paper).5 The study was performed using seven brake pad
formulations that were in high volume use in 1998. Measurements were taken on both front disc
as well as rear drum brakes. The study measured mass, size distribution, elemental composition,
as well as fiber concentration at four temperature intervals. The report also estimated PM2.5 and
PMio emissions for light-duty vehicles of 3.4 and 4.6 mg/mile, respectively for small vehicles,
and PM2.s and PMio emissions of 8.9 and 12.1 mg/mile, respectively for pickup trucks.
Sanders et al (2003)6 looked at three more current (as of-2003) classes of lining materials: low
metallic, semi-metallic and non-asbestos organic (NAO) representing about 90% of automotive
brakes at that time. Three kinds of tests were conducted: a dynamometer test, a wind tunnel test
and a track test at the Ford Dearborn proving grounds. Three sets of brake conditions were used:
(a) the first set of tests evaluated all three materials on a brake dynamometer under mild and
aggressive driving conditions, the urban driving program (UDP) with a set of 24 stops and a -7.9
m/s2 deceleration called the Auto Motor und Sport magazine (AMS) test; (b) a series of high
speed 1.8 m/s stops of a mid-size sedan with low metallic brakes were conducted in a wind
tunnel; and c) measurements of the same vehicle on a test track where collected where
decelerations were made from 60 mph at 0.15, 0.25 and 0.35 g-forces, the latter corresponding to
the AMS test to compare to the brake dynamometer. The latter test included low metallic as well
as NAO materials. The authors found that the mean particle size and the shape of the mass
distribution are very similar for each of the three linings, however they found that the low
metallic linings generate 2-3 times the number of wear particles compared to semi-metallic and
NAO linings. They also found that wear (and portion of wear that is airborne PM emissions)
increased non-linearly with higher levels of deceleration. Wear debris composition was found to
-------
have the most abundant elements consisting of Fe, Cu, Si, Ba, K and Ti, although the relative
composition varied significantly by brake type. The authors further found that 50-70% of the
total wear material was released in the form of airborne particles.
Table 2-1 contains the emission rates derived from the literature review conducted in support of
MOVES2009. While there are emission rates presented from other papers, this paper largely
relies on the Sanders et al. paper as it includes the widest array of materials currently in use,
measurement techniques, and deceleration ranges in a scientifically designed study. It is the only
paper from which modal rates can be derived. It is also the most recent of the papers listed and
improves on the measurement methods introduced in its predecessors. The other papers results
are provided as a source of comparison. Note that the range of rates from Sanders et al. (2003)
largely covers the range presented in the other papers as well. When determining the rates below,
the values from Garg et al. (2000), are also used.
Table 2-1 Non-Exhaust PM Emissions (per vehicle) from Mobile Sources Literature Values of emission
factors from brake lining wear (largely cited in Luhana et al. (2004)'s literature review
Literature Source
Luhana etal. (2004)
Sanders et al. (2003)
Abu-Allabanetal.(2003)
Westurland(2001)
Garg et al(2000)
Rauterberg-Wulff (1999)
Carbotech(1999)
Cha et al.(1983) used in PARTS
Vehicle Type
Light Duty
Heavy Duty
Light Duty
Light Duty
Heavy Duty
Light Duty
Heavy Duty
Passenger Cars*
Large Pickup
Trucks
Passenger Cars
Heavy Duty
Vehicles
Light Duty
Heavy Duty
Cars and Trucks
PM25
[mg/km]
0-5
0-15
3.4
8.9
PM10
[mg/km]
0-79
0-610
1.5 -7.0
0-80
0-610
6.9
41.2
4.6
12.1
1.0
24.5
1.8-4.9
3.5
7.8
* In this table, "passenger cars" are equivalent to light duty cars. "Light Duty" on their own includes all
Light-duty vehicles, including trucks though the studies are not all equivalent in their definitions.
-------
2.2 Developing Rates for MOVES
2.2.1 Emissions during braking
The MOVES2009 braking emission rate is based on the average of:
(1) Composition of brake pad
(2) Number (and type) of brakes
(3) Front vs rear braking
(4) Airborne fraction
and explicitly accounts for:
(1) Particle mass size distribution (PM2.5 vs PMio)
(2) Braking intensity
(3) Vehicle class: Light-Duty vs Heavy-Duty
As discussed in Sanders et al. (2003), most brake pads (at the time of the publication of that
paper) are either low-metallic, semi-metallic (full-truck), or non-asbestos organic (full-size car).
Using the results from this study, we make the following assumptions which are consistent with
those used in the paper.
equal mix of the three brake types,
four brakes per light duty vehicle, including 2 front disc brakes, and 2 rear drum brakes
- 2/3 of braking power (and thus emissions) in front brakes (1/3 rear)3
- the fraction of total PM below 2.5um is ~ 10% (+/-5%)b
- 60% of brake wear is airborne PM (+/- 10%).
We also do not compensate for the different average weights of the vehicles (though the MOVES
VSP bins scale emissions with mass). We assume there is an equal mix of the three brake types
because the market share penetration is not known.
For each test cycle from Sanders et al. (2003) and Garg et al. (2000), the following figures show
how we went from the measured results to emission rates of g/hour (for deceleration times only)
at various deceleration speeds. Sanders et al. (2003) used three measurement techniques, a filter,
an Electrical Low Pressure Impactor (ELPI), and a Micro-Orifice Uniform Deposition Impactor
(MOUDI). While all three measurement techniques produced similar results, we show all here.
Test results are shown for the UDP and wind tunnel tests from Sanders et al. (2003), as well as
the Garg et al. (2000) analysis. The latter paper adds another deceleration point for comparison.
The AMS results are not presented in the Sanders paper, however, the authors provided the data
for the purposes of this study.
a Based on discussions with author of paper Matti Mariq at Ford Motor Company and consistent with the Garg et al.
(2000) paper, who used 70%. Some of the other assumptions in this list is also from these discussions
b More will be discussed below.
-------
Table 2-2 - UDP results0
Test brake lining
PMi n emiss.
(mg/stop/brake)
UDP
low metallic
semi-metallic
Non-asbestos
Average/stop/brake
Avg. /veh
deceleration =
avg. brake time in sees =
avg. emissions in mg/stop =
emission rate for the UDP test =
filter
6.9d
1.7
1.1
3.2
9.7
ELPI
7.0
1.7
1.5
3.4
10.2
0.0012 km/s2
13.5 sees
9.95 Mg/stop
2.65 g/hr
Table 2-3 - Wind Tunnel results
Test brake lining
deceleration
Initial Velocity V(0) =
avg. brake time in sec =V(0)/dec
avg. emissions in mg/stop =
emision rate for the wind tunnel test
emss.
(mg/stop/brake)
Tunnel
low metallic
filter*
44
ELPI
45
MOUDI
40
0.0018
0.0267
14.8
129.0
31.4
in km/s2
in km/s
sees
mg/stop
g/hr
0 As these are intermediate values, the number of significant digits may exceed the precision known, however they
are kept in this presentation, and rounded for the final results. The UDP decelerations are the average decelerations
from those measured in the Sanders paper. The average brake times were determined with the assistance of one of
the original authors of the paper (Matti Mariq) who supplied the second by second trace. The filter PM10 were
determined by multiplying the total PM reported in Table 5 of the paper with the PM10 to total PM ratio determined
from the ELPI measurement.
d Sanders et al, reports the total filter PM to be 8.2 mg/brake/stop. In order to get PM10 equivalent, we applied the
ELPI ratio from table 5 in the reference. So 6.9 = 8.2* (7/8.3). The other numbers were calculated in a similar
fashion. Also, the avg per vehicle emissions is the avg stop/veh/brake emissions multiplied by 3. This is based on
the assumption made earlier that 2/3 of braking comes from the front brakes and 1/3 from the rear brakes.
-------
Table 2-4 - AMS results
Test brake lining
PM10 emiss.
(mg/stop/brake)
AMS
low metallic
semi-metallic
Non-asbestos
Average=
Avg/veh rate =
filter
800
510
550
620
1116
ELPI
70
63
92
75
135
deceleration =
Initial Velocity V(0) =
avg. break time in sec =V(0)/dec
avg. emissions in mg/stop for PM 10=
emision rate for PMi0 for the AMS test=
avg. emissions in mg/stop for PM2 5=
emision rate for PM2 5 for the AMS test=
0.0079
0.0278
3.5
1116
1143
135.0
138.2
in km/s2
inkm/s
sees
mg/stop
g/hr
mg/stop
g/hr
Table 2-5 - Garg et al. (2000) results
Test brake lining
PM10 emiss.* PM2S** (mg/stop/brake)
avg. over all
temp.
semi -metallic #1
semi-metallic #5
NAOS #2
NAOS #3
NAOS#7
Grand Avg. =
1.85
0.82
2.14
0.89
1.41
1.42
1.35
0.60
1.57
0.66
1.03
1.04
mg/stop
deceleration =
Initial Velocity V(0) =
avg. break time in sec =V(0)/dec
avg. emissions in mg/stop for PM10 =
emision rate for PM10 for the GM test=
avg. emissions in mg/stop for PM2 5 =
emision rate for PM2 5 for the test=
0.00294
0.0139
4.7
1.42
1.08
1.04
0.79
in km/s2
inkm/s
sees
mg/stop
g/hr
mg/stop
g/hr
We used these four data points to fit a power function to determine the emission rate at different
deceleration levels shown in the following figure. The AMS test, at higher decelerations, clearly
has a significant influence on results of the curve fit. Additional high speed tests could be used
for future refinement of this data.
-------
Figure 2-1- Brake wear PM2.5 emission rates in units of grams per hour for light duty vehicles as a function of
deceleration rate based on Sanders et al. (2003) and Garg et al. (2000)
160.0
140.0
2.0
4.0 6.0 8.0
deceleration in m/s2
10.0
2.2.2 Activity
In the previous section, we determined the rate of particulate matter emissions during braking in
units of grams per hour (per vehicle) as a function of deceleration level for a light-duty vehicle.
However, for MOVES, we also need to determine the frequency of different levels of braking.
The MOVES vehicle specific power (VSP) bins are relatively coarse for braking.6'7 There is a
large braking bin (operating mode 0) that contains a large fraction of driving activity, however
there are also a number of "coasting" bins that also contain braking events in each speed
category (Table 2-6). Each of these deceleration operating modes include some braking as well
as cruise and coasting operation (where the throttle is closed or nearly closed, but the brakes are
not applied). Therefore, the emission rate assigned to these bins need to contain the appropriate
average rates including the mix of driving and deceleration frequencies, and including
decelerations that do not include braking.
e While this document does not provide a detailed discussion of vehicle specific power, the light
duty emission rate report have an extensive discussion
-------
Table 2-6. VSP Operating Mode Bins by VSP and speed. Operating mode 0 and 1 (not listed) are braking and
idle respectively
VSP Class
(kW/ tonne)
30+
27-30
24-27
21-24
18-21
15-18
12-15
9-12
6-9
3-6
0-3
<0
Speed Class
(MPH)
1-25 25-50
r r r
30
r r
29
r r
16
28
r r
27
r r r
15 25
r r
14 24
r r r
13 23
r r
12 22
r r
11 21
50+
40
39
38
37
35
33
Operating mode where braking is assumed
We estimated the fraction of activity that is braking within each of the "coasting" bins by first
determining the coast down curve, then combining that with the activity fraction as seen in real-
world driving surveys.
o
The coastdown curves were generated using the Physical Emission Rate Estimator (PERE).
This was done by using the coastdown equations from PERE, and calculating the deceleration at
each speed when the forward tractive power is zero. We assumed all activity below coastdown is
braking and all activity above the curve is low throttle deceleration. Figure 2-2 shows coastdown
curves for cars of a variety of weights (and coastdown coefficients). The dotted curve is a
typical coast down curve for this class of vehicle, where 1,497 kg is the typical mass of a light
duty vehicle. The average weight for passenger cars used in MOVES is 1,497 kg.
-------
Figure 2-2- Modeled Coastdown curves using the PERE model for a variety of light-duty vehicles masses
900
1200
— 1497
— 1800
— Poly. (1497)
O)
CO
re
o
O
-1
-1.2
-1.4
-2
0.0001454x2 +0.0000622X -0.1758
R2=1.0
Speed (mph)
The deceleration activity was determined from two real world instrumented vehicle studies: one
from Kansas City and the other in Los Angeles. The Kansas City study was conducted by EPA
and Eastern Research Group (ERG) in 2005 to study real world driving activity and fuel
economy on conventional as well as hybrid electric vehicles.9 Over 200 vehicles were recruited,
though for the current analysis, only the activity data from the conventional, or non-hybrid,
population were examined. The Los Angeles activity data was conducted by Sierra Research for
the California Department of Transportation with both instrumented vehicles as well as chase car
data10'11'12. The deceleration data was analyzed for both of these studies.
Table 2-7 shows the distribution of braking activity across deceleration levels from both of these
studies. As expected, the vast majority of braking occurs during mild decelerations rather than
full (high decel) stops. More information about the PERE coastdown calculation process is
described in Appendix A.
10
-------
Table 2-7 - Activity Distribution of braking activity in the LA and Kansas City studies for each deceleration
bin.
Decel (mph/s)
1
2
o
6
4
5
6
7
8
9
10
11
12
13
14
LA
urban
37.1%
26.3%
17.9%
10.2%
5.6%
1.6%
0.64%
0.28%
0.17%
0.10%
0.05%
0.03%
0.01%
0%
LA
rural
27.1%
27.9%
20.2%
12.2%
8.2%
2.4%
0.98%
0.41%
0.26%
0.13%
0.09%
0.05%
0.01%
0.01%
KC
54.5%
26.3%
12.8%
4.6%
1.3%
0.30%
0.07%
0.02%
0.02%
0.01%
0.01%
0%
0%
0%
AVG
39.5%
26.9%
17.0%
9.0%
5.0%
1.4%
0.6%
0.2%
0.2%
0.08%
0.05%
0.03%
0.01%
0%
2.2.3 Emission Rate for Light Duty vehicles
The emission rate curve from Figure 2-1 was combined with the average activity in Table 2-7
discussed above (using a sum of the product) to calculate MOVES rates for light duty vehicles.
This gives an average PM2.5 braking emission rate of 0.557 g/hr.
However, as mentioned earlier, MOVES has brake emissions in not only VSP op-mode bin 0
(defined as the braking bin), but also in modes 1,11,21,33. Idle (zero speed, op-mode bin 0)
braking occurs in the transition (deceleration) from non-zero speed to zero speed which is a
small amount of activity in this bin. Bins 12 and 22 also contain a very small amount of braking,
which are ignored - i.e, the rates in these bins are set to zero. The brake emission rate in the
other bins were reduced by the amount of braking activity in each bin. These braking fractions
were derived by combining the amount of average activity from Kansas City and LA above and
the coast down curves from PERE discussed earlier. The results are shown in Table 2-8 below.
f For example, the PM2 5 emission rate in VSP bin 11 for light-duty vehicles is 0.557 * 0.978 = 0.546 g/hr
11
-------
Table 2-8 - Vehicle Specification (top) and Fraction of Activity in VSP bin that is braking (last 5 rows) for a
variety of vehicle types (motorcycle and bus activity fractions were copied from Light-duty and heavy-duty
trucks respectively).
wgt(kg)
CrO (rolling resistance)
Cd (drag coeff)
A (frontal area mA2)
vsp bin
0
1
11
21
33
Mid-size
car(LDV)
1497
0.008
0.32
2.25
1
0.0437
0.975
0.641
0.115
SUV
(LOT)
1800
0.008
0.36
2.5
1
0.0437
0.975
0.661
0.122
LHDT
(<=14k)
5602
0.008
0.37
2.75
1
0.0316
0.913
0.743
0.126
LHDT
(>14k)
9333
0.008
0.44
6.7
1
0.0316
0.906
0.685
0.116
MHDT
13517
0.01
0.44
6.7
1
0.0316
0.91
0.725
0.121
HHDT
22680
0.01
0.44
8.64
1
0.016
1
0.641
0.068
2.2.4 PM1(/PM25 Brake Wear Ratio
MOVES stores PM2.s brake wear emission rates by operating mode bin, then estimates
emission rates by applying a PMio/PM2.s ratio. The PMio/PM2.s ratio is based on the assumptions
that the mass fraction of particles below PMio is 0.8, and the mass fraction of particles below
PM2.s is 0.1. More specifically, Sanders et al. (2003), report PM "fractions and cutoffs of 0.8 at
10 |im, 0.6 at 7 |im, 0.35 at 4.7 |im, 0.02 at 1.1 |im, and <0.01 at 0.43 |im for the UDP stops
typical of urban driving". These assumptions result in a PMio/PM2.s ratio of 8. Where no PM2.5
values were reported, we calculated PM2.5 from PMio emission rates using this fraction. This
estimate widely varies in the literature. Abu- Allaban et al. (2003) reports that only 5-17% of
PMio is PM2.5, which is consistent with Sanders. Garg et al. (2000), report 72% of PMio is
PM2.5, which is disputed by Sanders et al. (2003). The current study does use the PM2.5
measurement reported by Garg et al. (2000), however in reality, this single value has little impact
on the curve fit in Figure 2-1, which is dominated by the more recent data from Sanders et al.
(2003).
The emission rates in g/hr PM2.s and PMio by operating mode and regulatory class are included
in Appendix B. The rates are calculated per the methodology described above and is
independent of model year and environmental conditions. The average PM2.5 and PMio brake
wear emission rates for passenger cars and trucks from three urban county inventories, using
MOVES2014 are displayed in Table 2-9. MOVES brake wear emission rates by source type will
vary according to the inputs of average speed, and VMT by road type, which impacts the
distribution of operating modes within each source type in MOVES.
12
-------
Table 2-9 Average PM2.5 and PM10 brake wear emission rates (mg/mile) for passenger cars and trucks from
3 urban county inventories using MOVES2014
Passenger Cars (21)
Passenger Trucks (3 1)
PM25
3.7
6.2
PM10
29.8
49.8
The average passenger car MOVES PMio emission rates of 29.8 mg/mi (output from the model)
is compared to the previous studies (in the literature) in Table 2-1. Carbotech (1999), Sanders et
al. (2003), Garg et al. (2000), are all laboratory measurements and have significantly smaller
reported emission rates than the present study. On the other hand Luhana et al. (2004), Abu-
Allaban et al. (2003), Westurland (2001), and Rauteberg-Wulff (1999) are roadside measurement
or tunnel measurements. These studies generally have higher emissions than laboratory
measurements. The MOVES rates largely generated from Sanders et al. (2003), are also
considerably larger than the publication cites. This is largely due to the fact that Sanders et al.
(2003), cites results primarily from the UDP braking events which are significantly milder than
the AMS decelerations. Through the modeling described in this paper, the AMS deceleration
rates are weighted in to the milder deceleration emission rates to give higher rates comparable
now to some of the results achieved from the tunnel and roadside studies. The light duty rates
are thus calibrated to laboratory measurements adjusted to real-world factors, and "validated" to
be within the range of roadside and tunnel measurements.
2.2.5 Brake Wear Emissions for Heavy-Duty Vehicles and Other Vehicle
Types
There is very little literature on direct heavy-duty brake emissions measurements. To decelerate,
heavy-duty vehicles employ technologies such as disc and drum as well as other braking
methods including downshifting and engine (or "jake") braking. A scientific study comparing
the emissions and relative activity of each of these methods of braking is beyond the scope of
this report. In order to estimate brake wear emission factors for heavy-duty vehicles an
engineering analysis was combined with results from a top-down study performed by Mahmoud
Abu-Allaban et al. (2003).13 The authors collected particulate matter on filters near roadways and
apportioned them to sources utilizing Chemical Mass Balance, CMB, receptor modeling along
with Scanning Electron Microscopy. The study was performed at roadside locations in Reno,
Nevada and Durham, North Carolina where intensive mass and chemical measurements were
taken. The authors of the paper attempted to collect and differentiate between PM measurements
from tailpipe, tire, road dust, and brake from light- and heavy-duty vehicle types. Compared to
the other papers described in the previous section (on light-duty braking) that include heavy-duty
rates, the Abu-Allaban paper is one of the most recent studies of its kind performed at the time of
the writing of this paper. The results are consistent with the heavy-duty rates measured from
Luhana et al. (2004) as well as Westurland (2001), but is the only paper to measure PM2.5. The
paper's light-duty rates are also aligned with the rates determined above.
In this study, PM2.5 brake wear emission rates for heavy duty vehicles ranged from 0 to 15
mg/km (0 to 24 mg/mi). For this analysis we have assumed the emission rate was the midpoint
of the range of emission factors, or 12 mg/mi. For the purposes of populating MOVES rates, we
13
-------
do not employ the measured emission rate directly from this study due to the extreme uncertainty
and variability of measurement and locations selected. Rather, we rely on the paper's
comparison of light-duty to heavy-duty emission factors. On table 5 of the paper, the emission
rates for the exit ramps are reproduced below. Only the exit lanes were included of the many
roads where measurements were collected. The remainder of the roads are represented by the
average and the (min to max) range reported in the table.
Table 2-10 Brake Wear Emission Rates reproduced from Abu-Allaban et al. (2003)
Location
J. Motley Exit
Moana Lane Exit
Average over all
roads
Range (min to max)
of measurements on
all roads
Vehicle Type
Heavy-Duty
Light-Duty
Heavy-Duty
Light-Duty
Heavy-Duty
Light-Duty
Heavy-Duty
Light-Duty
PM10 (mg/km)
610 ±170
79 ±23
120 ± 33
10 ±3
124 ±71
12 ±8
0 to 610
OtoSO
PM25 (mg/km)
0±0
0±0
0±0
0±0
2±2
1±0
Otol5
Oto5
Due to the difficulty of differentiating a small brake emissions signal from the much larger signal
coming from tailpipe, tire wear and road dust combined, there is much uncertainty in these
measurements - yet another reason why adjusted laboratory measurements were favored above.
Clearly PM2.5 was difficult to measure from most sites. Interestingly, the heavy-duty
measurements were highest on the exit lanes for PMio, however (rather inexplicably), the other
road types had higher emissions than the exit lanes for PM2.5. For these reasons, we rely more on
averages to determine our ratio of heavy-duty to light-duty brake emission factors. From these
measurements, we can determine that the average ratio of FID to LD brake emissions is 10 and 2
for PMio and PM2.5 respectively.8 On average, based on Table 2-10, the ratio is 7.6 for PMIO.
The following table compares the ratio for the remaining studies for comparison.
Table 2-11- Ratio of Heavy-Duty to Light-Duty PM from the literature.
Study
Luhana et al. (2004)
Abu-Allaban etal. (2003)
Westurland(2001)
Rauterburg-Wulff (1999)
Carbotech(1999)
PM2.5
o
5
PMIO
7.7
7.6
6.0
24.5
0.7
For the purposes of MOVES, a simpler model requiring a single ratio of HD to LD brake
emissions and another ratio of PMio to PM2.s brake emissions is attractive - particularly since the
data to populate the model is sparse. Also the broad range of uncertainties in the literature can
support such simplification. Based on the range in the table, above, the value of the ratio chosen
is 7.5, very close to the ratio as measured by Abu-Alaban et al. (2003), and consistent with the
range of studies.
8 Though it is not shown in the table here, according to Abu-Alaban, based on the highest sampling sites (maximum
measurements from the table), the ratio of HD to LD brake emissions is 41 and 16 for PM10 and PM25 respectively.
14
-------
The estimated emission factors for all other categories of vehicles (between light and heavy-
duty) were derived by linearly interpolating the rates between light-duty and combination heavy-
duty vehicle classes by their respective weights as shown in the figure below. This is based on a
rather simple engineering (and unproven in this study) hypothesis that the relative brake
emissions is proportional to the weight of the vehicle classes relative to (and bounded by) light
and heavy-duty vehicles. The hypothesis is based on the assumption that relative mass of the
vehicles is proportional to the relative energy required to stop the vehicles. Figure 2-3 below
shows the relative mass of light- and heavy-duty vehicles. The corresponding emission rates are
in Table 2-12.
Figure 2-3 - Interpolated Brake PM2.5 Emission Rates by Regulatory Class Weight. Passenger Cars and
Combination Heavy duty Trucks Define the Slope.
4.5
I 4
^ 3.5
I 2.5
l/l
E
LU
0)
2
ro 1.5
in 1
oi
0.5
0 •'
0
10000 20000 30000 40000 50000
Regulatory Class Weight
60000
70000
Table 2-12contains average brakewear PM emission rates from three MOVES runs for urban
counties using MOVES2014, for each source types. As mentioned earlier, average emission rates
by source type will vary for local users according to inputs that impact the operating mode
distribution of vehicle operation
Table 2-12. Scaling Emission Rates to their vehicle class, regclass id is the MOVES identifier for regulatory
class.
MC
LDGV
LOT
HD<=14k
HD>14k
MHDD
HHDD
Urban Bus
regclasswt in Ibs.
830
8000
13,474
12,358
20,575
29,808
59,369
30,000
regclassid
10
20
30
41
42
46
47
48
g/hr
0.056
0.56
0.94
0.87
1.4
2.1
4.2
2.1
15
-------
Table 2-13 Average PM2.5 and PM10 brake wear PM emission rates (mg/veh-mile) for the MOVES
regulatory classes from three urban county inventories using MOVES2014
sourceTypelD
11
21
31
32
41
42
43
51
52
53
54
61
62
Source Type
Motorcycle
Passenger Car
Passenger Truck
Light Commercial Truck
Intercity Bus
Transit Bus
School Bus
Refuse Truck
Single Unit Short-haul Truck
Single Unit Long-haul Truck
Motor Home
Combination Short-haul Truck
Combination Long-haul Truck
PM25
0.3
3.7
6.2
6.2
23.3
12.6
13.2
25.6
13.6
16.6
11.5
24.3
26.6
PM10
2.6
29.8
49.8
49.8
186.4
100.9
105.3
204.4
109.2
132.4
92.2
194.0
213.0
16
-------
3 Tirewear
3.1 Introduction
Tires are an essential part of any vehicle and the number and size of tires increase with the size
of the vehicle. Contact between tires and the road surface causes the tires to wear, with the rate
dependent on a variety of factors.
EPA's previous estimates of tire wear are contained in the PARTS model, and are emission rates
of 0.002 grams per mile per wheel. Two LDV studies from the 1970s are the basis for these
emission rates. The PARTS emissions factors are based on tests of older bias-ply tires rather
than more modern radial tire technologies The National Resource Council report on the
MOBILE model, suggested that the PARTS rates may be out of date.14
Tire wear occurs through frictional contact between the tire and the road surface. Friction causes
small and larger particles to wear from tire, which are then either released as airborne
particulates, deposited onto the road surface or retained in the wheel hub temporarily or
permanently until washed off. The road surface causes friction and abrasion and therefore the
roughness of the surface affects the wear rate by a factor of 2-3.15
In addition to road surface roughness, tires wear is dependent upon a combination of activity
factors such as route and style of driving, and seasonal influences. Heavy braking and
accelerating (including turning and road grade) especially increases tire wear. The route and
style of driving determine the amount of acceleration. Highway geometry is a key factor with
rise and fall in roads also resulting in increased tread wear. The acceleration of the vehicle
determines the forces applied to the tire, and includes turning. Tire wear due to tire/road interface
is determined by and is directly proportional to these forces.16 The season results in temperature,
humidity and water contact variations. Wear rates are lower in wet compared to dry conditions.
Finally vehicle characteristics also influence tire wear. Key factors are the weight, suspension,
steering geometry, and tire material and design. Axle geometry changes result in uneven wear
across the tire width. The type of tire influences the wear significantly. In particular, the
physical characteristics like the shape of the tire (determined by stiffness), the rubber volume
(tread pattern), and the characteristic of the tire (rubber type etc.). As a consequence of different
manufacturing specifications, different brands of tires wear at different rates. Retreads are also
considered to wear more than new tires. Wear rate studies on tire fleets reported in Bennett &
Greenwood (2001) also indicated that retreads had only about 75% of the tire tread volume that
new tires had. Cenek et al. (1993) reported that 20% of New Zealand passenger tire sales were
retreads and that retreads made up 75% of the tire tread in a sample of buses in the New Zealand
fleet.17 However, modeling emissions from retreads was deemed beyond the scope of the report
According to the literature, the most straightforward method for determining tire wear is the
periodic measurement of tread depth. However, variations in the extent of wear across the tire
and irregularities in tire shape could lead to inaccurate measurements. Determining tire weight
loss is a more sensitive approach than the measurement of tire depth, though care must be taken
to avoid errors due to damage to tires as a result of their removal from the vehicle and hubs, and
material embedded in the tire. To minimize damage to the tire, Lowne (1970) weighed both the
wheel and tire simultaneously after the wheel was brushed and stones embedded in the tire were
17
-------
1 &
removed. Table 3-1 shows a summary of the literature search conducted as of 2006 on the mass
of tire wear.
Wear rates for tires have typically been calculated based on tire lifetime (in kilometers traveled),
initial weight and tread surface depth. Tire wear occurs constantly for moving vehicles, but may
be significantly higher for cars which tend to brake suddenly or accelerate rapidly. Tire wear
rates have been found to vary significantly between a wide range of studies.l
Speed variation is an important factor as well. Carpenter & Cenek (1999) have shown that the
effect of speed variation is highest at low speeds as a result of inertial effects and effective
90
mass. They also examined lateral force effects on tires and assessed tire wear on routes of
different amounts of horizontal curvature and found that there was little variation.
Tire abrasion is difficult to simulate in the laboratory, since the varied nature of the road and
driving conditions influence wear rates in urban environments. Hildemann et al. (1991)
determined the chemical composition of tire wear particles using a rolling resistance testing
machine at a tire testing laboratory over a period of several days.21 Rauterberg-Wulff (1999)
determined particle emission factors for tire wear using modeling in combination with
measurements conducted in the Berlin-Tegel tunnel.22
Tire wear rates have been measured and estimated for a range of vehicles from passenger cars to
light and heavy duty trucks with results reported either as emission per tire or per vehicle. Most
of the studies report only wear, not airborne PM. The wear rates found in the literature are
summarized in Table 3-1 below and are converted to a per vehicle rate (units are in per vehicle
kilometer). A range of light-duty tire wear rates from 64-360 mg/vehicle/km has been reported in
the literature. Much of the variability in these wear rates can probably be explained by the factors
mentioned above. These studies made no distinction between front and rear tires, even though
they can wear at different rates.23
18
-------
Table 3-1 - Tire wear rates found in the literature. Rates are per vehicle. Estimated number of tires is
described later.
Source
Kupiainen,K.J. etal(2005)24
Luhana et al (2003)
Councell,T.B. etal(2004)
U.S. Geological Survey25
Warner et al. (2002)26
Kolioussis and Pouftis (2000)27
EMPA (2000)28
SENCO (Sustainable Environment
Consultants Ltd.) (1999)29
Legret and Fagotto (1999a)
Baumann(1997)30
Garben(1997)31
Gebbe (1997)32
Leeetal(1997)33
Sakai,H (1995)
Baekken(1993)34
CARB (1993)
Muschack (1990)
Schuring and Clark (1988)35
Pierce,R.N. (1984)
Malmqvist (1983)36
Gottle (1979)37
Cadleetal. (1978)38
Dannis (1974)39
Remarks
Measured tire wear rate
Measured tire wear rate
Calculated rate based on literature
Average tire wear for a vehicle
Average estimated tire wear
Light duty vehicle tire wear rate
Heavy duty vehicle tire wear rate
Light duty vehicle tire wear rate
Wear rate for trucks
Estimated rate for light duty vehicles
Estimated rate for heavy vehicles (>3.5t)
Passenger car tire wear rate
Heavy duty vehicle tire wear rate
Articulated lorry tire wear rate
Bus tire wear rate
Passenger car tire wear rate
Light duty vehicle tire wear rate
Heavy duty vehicle tire wear rate
Motorbike tire wear rate
Passenger car tire wear rate
Light duty vehicle tire wear rate
Heavy duty vehicle tire wear rate
Motorbike tire wear rate
Estimated tire wear rate
Measured tire wear rate
Estimated tire wear rate
Passenger car tire wear rate
Estimated tire wear rate
Estimated tire wear rate
Estimated tire wear rate
Estimated tire wear rate
Estimated tire wear rate
Measured tire wear rate
rate in mg/vkm
9mg/km-PM10
2 mg/km -PM25
74
200
97
40
53
798
53
1403
68
136
80
189
234
192
64
112
768
32
53
110
539
26.4
64
184
200
120
120
240-360
120
120
120
4
90
While there is significant literature on tear wear, there is relatively little published on airborne
particulate matter from tires. In this report, a model for tire wear rates are first determined, and
then a discussion of the modeling of airborne PM2.5 and PMio follows building off the wear
model.
19
-------
3.2 Methodology
This report begins by estimating the tire wear from light-duty vehicles, then based on the per tire
wear, extrapolates to other vehicle types. Then the emission rates are derived from the wear
rates. The method primarily depends on the data from work published by Luhana et al. (2004)
wherein wear loss rates for tires have been determined gravimetrically for in-service cars.40 At
the time of this analysis, this paper was both a recent and comprehensive study. The authors
weighed car tires at two-month intervals, and asked drivers to note the details of each trip
undertaken. Five test vehicles (labeled A-E) were selected for the tests. Of these vehicles A
(1998 Audi A3), B (1994 Ford Mondeo), C (1990 Peugeot 205) and E (1992 Vauxhall Cavalier)
were front-wheel drive vehicles (FWD). According to the driver surveys, the predominant road
type used by vehicles A and B were motorways, for vehicle D (1990 Ford Sierra) it was rural
roads and motorways for vehicle C it was suburban roads, and for vehicle E, it was rural roads.
Vehicle D was excluded from this study since it was a rear-wheel drive (RWD) vehicle. RWD
vehicles are relatively uncommon amongst passenger vehicles in the United States, and the wear
from this particular vehicle was more than double the other FWD vehicles. It is uncertain
whether the discrepancy from this vehicle was because it was a rear-wheel drive or for some
other reason. The selection of vehicles was based primarily on driving conditions, as defined by
the main type of road used by the owner and annual distance driven.
Results from the Luhana et al. (2004) study indicated that the lowest tire wear rates (56 mg/vkm
and 67 mg/vkm respectively ) were for vehicles A and B that were driven predominantly on
motorways. Vehicles C and E had very similar wear rates (around 85 mg/vkm) although these
vehicles tended to be driven on different roads. Based on the wear rates from the four front-
wheel drive cars alone, the study concluded that the average wear rate is around 74 mg/vkm.
This value seems to lean towards the lower end of the range of wear rates reported in the
literature.
The data presented in Table 3-2 includes calculations for the distances completed by each vehicle
between successive tests, the estimated average trip speeds and predominant road types for the
equivalent periods. It was assumed that the weight of the wheels remained constant during the
tests, and any weight loss was due solely to the loss of tire rubber during driving.
h vkm is "vehicle kilometer" and assumes four times a per tire rate for light-duty vehicles.
20
-------
Table 3-2: Data from Luhana et al. (2004) with measurements of tire wear for a variety of trips
vehicle
tests
testl-A
test2-A
tests -A
test4-A
testl-B
test2-B
tests -B
test4-B
testl-C
test2-C
tests -C
test4-C
tests -E
test4-E
Avg. trip
speed
km/hr
90.3
90.6
93.9
92.7
65.4
71.9
74.4
70.2
44.5
42.9
48.8
50.4
61.3
65.8
Tire Wt. Loss (per axle)
Front mean (g/km)
0.0202
0.0209
-
0.0172
0.0298
0.0262
0.019
0.0297
0.0312
0.0331
0.0284
0.0532
0.037
0.0265
Rear Mean
(g/km)
0.0092
0.0126
0.0069
0.0086
0.0087
0.0091
0.004
0.007
0.0047
0.0132
0.0064
0.0045
0.0104
0.0109
total wt. loss (per
vehicle)
g/km
0.0589
0.0669
-
0.0516
0.077
0.0705
0.0461
0.0735
0.0718
0.0925
0.0697
0.1153
0.0948
0.0749
total wt. loss
(per vehicle)
g/mi
0.0947
0.1076
-
0.083
0.1239
0.1135
0.0742
0.1183
0.1155
0.1489
0.1121
0.1855
0.1525
0.1205
avg. speed
mi/hr
56.1
56.3
58.4
57.6
40.6
44.7
46.2
43.6
27.7
26.7
30.3
31.3
38.1
40.9
Note: Vehicles A and B were driven mainly on motorways (freeways)
Vehicle C was driven on Suburban Roads and
Vehicle E was driven mostly on Rural roads
Using the above data on average speed and total weight loss an exponential regression curve was
fitted which was characterized by an R2 value of 0.43. The actual and predicted values are
presented in Figure 3-1.
A weak negative correlation is shown between tire wear and average trip speed, with wear being
around 50% higher at an average speed of 40 km/h (dominated by urban driving) than at an
average speed of 90 km/h (dominated by motorway driving).
21
-------
Figure 3-1 Relationship between light-duty tire weight loss (per vehicle) and mean trip speed between tests
Tire weight loss vs mean trip speed (actual)
0.25
0.00
10.0
20.0
30.0 40.0
mean trip speed in mi/hr
50.0
60.0
70.0
The shape of the curve in Figure 3-1 deserves some discussion. It can be seen from the curve
that the wear is maximum at zero speed and goes down as the speed goes up. This is based on the
extrapolation of the fitted curve. It may seem counter-intuitive that emissions is highest when
speed equals zero, however, it is important to note that the relationship does not take
accelerations (and turning) into account. Much of the tirewear occurs when the magnitude of a
vehicle's acceleration/deceleration is at its greatest, e.g. at low speeds when the vehicle is
accelerating from rest, or when the vehicle is braking hard to stop. A more improved
relationship would be by VSP bin, however there is insufficient data to characterize tire wear on
a second-by-second basis to enable binning by operating mode bins. The model has been
simplified to be based on speed at this time. However, for MOVES, the emission rate at zero
speed is set to zero to avoid anomalous results in project level analyses where increased idling
would result in an over prediction of tire emissions.
The predicted values as determined above are for passenger cars (LDVs). To determine tire wear
loss rates for other regulatory classes it was assumed that total tire wear per vehicle is dependent
upon the number of tires on the vehicle which in turn is a function of the number of axles per
vehicle by vehicle class. The latter data were found to be available in the Vehicle Inventory and
Use Survey (VIUS 2002) data base. This data enabled the calculation of tires per vehicle for each
of the six truck classes and thereby tire-wear losses for the different truck categories (regulatory
classes) were determined. The average number of tires per truck is given in Table 3-3 below.
22
-------
Table 3-3 - Average Number of Tires per Truck - Calculated from 2002 VIUS Survey of axle count.
RegClassID
10
20
30
41
42
46
47
48
RegClass name
MC
LDV
LOT
LHD<=14K
LHD45
MHDD
HHDD
Urban Bus
Average Tires Per
Vehicle
2.0
4.0
4.0
5.5
6.0
7.0
14.9
8.0
* Note: Tires per vehicle for LDT is the same as that for LDV
In a future study, another literature search should be conducted to search for differences in (per
tire) wear and emission rates from heavy-duty tires compared to those from the light-duty
market. There is another assumption made for the sake of simplicity, which is to keep the
emission rates of the tractive wheels identical to those of the wheels disconnected from the
drivetrain axles. A more recent literature search may also help determine whether another
approach is warranted.
Now that the average tire wear is quantified, it is critical to determine the fraction of that wear
that becomes airborne PM. The literature indicates that probably less than 10% of car tire wear
is emitted as PMio under 'typical' driving conditions but the proportion could be as high as 30%
(Boulter2005a). According to Luhana et al. (2004), PMio appears to be released from (all 4) tires
at a rate of between 4 and 6 mg/vkm for passenger cars. This suggests that generally between
around 1% and 15% by mass of passenger car tire wear material is emitted as PMio (though
much higher proportions have been reported in some studies). For this study, it is assumed that
8% of tire wear is emitted as PMio (average of 1% and 16%. According to Kupiainen et al
(2005), PM2.5 fractions were on average 15% of PMio.24 Based on this study, it is assumed that
1.2% of the total tire wear is emitted as PM2.s to develop our brakeweare emission rate. The 1.2%
is derived from assuming that 8% of tire wear to be emitted as PMio and 15% of PMio is PM2.5.
We then convert the g/vehicle/mile brakewear emission rates to g/hr by multiplying by the
average speed of each MOVES speed bin. The g/hour brakewear emission rate for all regulatory
classes used in MOVES can be found in Appendix B. MOVES applies the same brake wear
emission rate for all vehicle fuel types (gasoline, diesel, flex-fuel, and CNG) within a MOVES
regulatory class. The average PM2.s tire wear emission rates in (mg/mile) for each regulatory
class, from three urban county inventories in MOVES2014 is shown in Table 3-4.
23
-------
Table 3-4 Average PM2.5 and PMi0 tire wear PM emission rates (mg/veh-mile) for the MOVES regulatory
classes from three urban county inventories using MOVES2014
sour ceTy pel D
11
21
31
32
41
42
43
51
52
53
54
61
62
sourcetypename
Motorcycle
Passenger Car
Passenger Truck
Light Commercial Truck
Intercity Bus
Transit Bus
School Bus
Refuse Truck
Single Unit Short-haul Truck
Single Unit Long-haul Truck
Motor Home
Combination Short-haul Truck
Combination Long-haul Truck
PM2.5
0.7
1.5
1.5
1.5
4.4
2.9
2.7
5.1
2.7
3.1
2.4
4.7
5.2
PM10
4.9
9.8
10.0
10.2
29.3
19.7
17.8
34.3
17.7
20.6
15.8
31.6
34.9
3.2.1 PMlt/PM25 Tire Wear Ratio
MOVES stores PM2.s tire wear emission rates by operating mode bin (in this case, speed bins),
then estimates PMio emission rates by applying a PMio/PM2.s ratio. Thus MOVES applies a
PMio/PM2.s ratio of 6.667, which is based on the particle size distribution of tire wear measured
by Kupianen et al. (2005)1. The average PMio emission rates from three urban county
inventories using MOVES2014 are displayed in Table 3-4.
4 Next Steps
As mentioned in the earlier section, this report underwent revisions since the previous version,
but these changes were largely editorial in nature in response to the peer review. There were no
changes made to the model or the rates since MOVES2010. There are a number of updates that
can be made to both this report and the model.
As a number of years have passed, it is possible that there are more publications in the literature
or airborne brake and tire emissions from mobile sources. These papers may shed light on
emission rates, size distributions, activity or speciation of PM. There is especially little
information in the literature on the latter. These newer papers can either be used to modify the
model, or validate the current rates.
The MOVES model has undergone changes since MOVES2010b. MOVES2014 includes some
changes to the vehicle specifications described in this report. For example, the default
1 The PM10/PM2.5 ratio is derived from dividing the PM10 fraction of total PM, by the PM25 fraction of total PM, :
.087.012 = 6.667 from values reported by Kupianen et al. (2005).
24
-------
assumptions regarding axle count (and thus number of wheels per vehicle), average weights,
aerodynamics, and rolling resistance, of certain regulatory classes have changed. The weights
will have a more significant impact on the brake rates (in particular) than the latter coefficients.
For brakes, the analysis from this study also only looked at front wheel drive brakes and
primarily from vehicles equipped with disc brakes in the front and drum brakes in the rear (the
most common light duty configuration). It was beyond the scope of this study to modify the
rates the fraction of vehicles with four disc brakes, or to update the speciation profile for brake
emissions, or to capture more advanced technology vehicles with electric regenerative braking.
Vehicles with four disc brakes should presumably have higher, while hybrids and electric
vehicles should have lower brake emissions. Moreover, the incident rate of other forms of
decelerating a truck such as downshifting and engine (or Jake) braking are also not considered in
this study due to a lack of data.
Since the writing of this report, the only change that was made to the brake wear model in
MOVES2014 was that for project level analysis, the emission rates in the idle bin was set to
zero. This was done to avoid results where users may get increasing brake emissions in
particular cases where idle rates are high. As mentioned above, the idle operating mode bin does
contain a small amount of deceleration when a vehicle transitions from motion to non-motion
(stop). However, if a user is increasing idle rates based on local knowledge compared to the
MOVES default, it is logical to assume that they should not get higher brake emissions.
Therefore, for inventory mode, the emission rates were maintained as described above in the idle
bin and the change was made only to project level analysis.
The idling tire wear emission rate is set to zero in the default emission rate table (Appendix B).
Thus, for idling tire wear emission rate is zero for both project level and inventory mode.
For tire emissions, it was beyond the scope of this study to quantify the differences in emissions
(per tire) between light duty and heavy duty tires (and everything in between). It was also
beyond the scope of this study to look at how trends in rolling resistance improvement may
increase or decrease tire wear emissions. Finally a more complete model including speciation of
tire and break PM, was beyond the scope of this study. Some of the references employed did
include some of these measurements, however brake material has been known to evolve over
time. These are all subjects for future study.
25
-------
Appendix A Deceleration from PERE
This appendix briefly describes some of analytical methods used to determine the deceleration
point at which coasting becomes braking. A full description of the PERE model is provided in a
separate EPA report as cited earlier. This section, provides additional information beyond what
can be found in the PERE documentation.
The basis for the tractive load equations in there PERE model are found in the A, B, C
coastdown coefficients described in the report. The author of this report conducted coastdown
testing on a -2001 Nissan Altima on relatively "flat" roads in Southeast Michigan. The A, B, C
coefficients for this vehicle can be found in the EPA database. The A,B,C tractive load
equations in PERE were converted to a coastdown curve and plotted compared to the data below.
The area above the curve is throttle and the area below the curve is braking. The curve itself is
"coasting" on neutral gear.
Coast Down - Modeled and Measured (altima on 1-94 and service dr; geanneutral)
80
100 120
Time (seconds)
140
160
180
200
Figure A-l Coast Down- Modeled and Measured (Altima on 1-94 and Service Drive; Gear: neutral)
Based on these coastdown equations, a series of coastdown curves are generated as a function of
vehicle mass. As in the previous plot, the area under the curve is braking and the area above the
curve is throttling.
26
-------
a.
"oi
8
T3
O)
^
(A
8
u
900
1200
1497
1800
— Poly. (1497
-2
y = -0.0001454X2 + 0.0000622x - 0.1758
R2 = 1.0
\
Speed (mph)
Figure A-2. Coast down Curves as a Function of Vehicle Mass
A PERE simulation is run on the FTP cycle and the braking episodes are flagged in the figure
below (for a typical 1497kg LDV).
27
-------
100 -i
90
70
60
50
40
30
20
10
7
- o
6600
6650
6700
6750
6800
time
6850
6900
6950
7000
Figure A-3 Braking Episodes over the FTP cycle
28
-------
Appendix B Brake and Tire Wear Emission Rates
This Appendix includes the brake and tire emission rates as a function of regulatory class and
operating mode which are stored in the MOVES emissionrate table.
Table B-l PM2.5 Brake Emission Rates by Regulatory Class and Operating Mode (g/hr)
regclassID
10
10
10
10
10
20
20
20
20
20
30
30
30
30
30
40
40
40
40
40
41
41
41
41
41
42
42
42
42
regClassName
MC
MC
MC
MC
MC
LDV
LDV
LDV
LDV
LDV
LOT
LOT
LOT
LOT
LOT
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD45
LHD45
LHD45
LHD45
opModelD
0
1
11
21
33
0
1
11
21
33
0
1
11
21
33
0
1
11
21
33
0
1
11
21
33
0
1
11
21
opModeName
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
MeanBaseRate
(g/hr)
0.055664
0.0024472
0.054488
0.03584
0.0056
0.55846
0.024472
0.546
0.35896
0.0644
0.940406
0.0412091
0.919425
0.623323
0.115046
0.865
0.027334
0.789745
0.642695
0.10899
0.865
0.027334
0.789745
0.642695
0.10899
1.44
0.045504
1.31472
1.06848
29
-------
42
46
46
46
46
46
47
47
47
47
47
48
48
48
48
48
LHD45
MHD67
MHD67
MHD67
MHD67
MHD67
HHD8
HHD8
HHD8
HHD8
HHD8
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
33
0
1
11
21
33
0
1
11
21
33
0
1
11
21
33
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Moderate Speed Coasting; VSP< 0;
25<=Speed<50
Cruise/Acceleration; VSP< 6; 50<=Speed
0.18576
2.09
0.066044
1.9019
1.51525
0.25289
4.16
0.06656
4.16
2.66656
0.28288
2.1
0.0336
2.1
1.3461
0.1428
30
-------
Table B-2 PM2.5 Tire Emission Rates by Regulatory Class and Operating Mode (g/hr)
regclassID
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
30
30
30
30
regClassName
MC
MC
MC
MC
MC
MC
MC
MC
MC
MC
MC
MC
MC
MC
MC
MC
MC
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LDV
LOT
LOT
LOT
LOT
opModelD
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
400
401
402
403
opModeName
idle
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
idle
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
idle
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
MeanBaseRate
(g/hr)
0
0.0031775
0.00601
0.01116
0.015525
0.01922
0.0223
0.02484
0.026915
0.02852
0.02979
0.03075
0.031405
0.0318
0.03198
0.03192
0.0318
0
0.006355
0.01202
0.02231
0.031065
0.03844
0.0446
0.04968
0.053795
0.05708
0.05958
0.06145
0.062755
0.06354
0.063895
0.06391
0.063525
0
0.006355
0.01202
0.02231
31
-------
30
30
30
30
30
30
30
30
30
30
30
30
30
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
41
41
41
41
41
41
41
41
41
41
41
LOT
LOT
LOT
LOT
LOT
LOT
LOT
LOT
LOT
LOT
LOT
LOT
LOT
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 10k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
404
405
406
407
408
409
410
411
412
413
414
415
416
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
400
401
402
403
404
405
406
407
408
409
410
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
idle
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
idle
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
0.031065
0.03844
0.0446
0.04968
0.053795
0.05708
0.05958
0.06145
0.062755
0.06354
0.063895
0.06391
0.063525
0
0.0087725
0.016595
0.0308
0.042885
0.05308
0.061575
0.06861
0.07427
0.0788
0.082305
0.08485
0.086625
0.08772
0.088205
0.0882
0.087675
0
0.0087725
0.016595
0.0308
0.042885
0.05308
0.061575
0.06861
0.07427
0.0788
0.082305
32
-------
41
41
41
41
41
41
42
42
42
42
42
42
42
42
42
42
42
42
42
42
42
42
42
46
46
46
46
46
46
46
46
46
46
46
46
46
46
46
46
46
47
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD <= 14k
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
LHD45
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
MHD67
HHD8
411
412
413
414
415
416
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
400
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
idle
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
idle
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
idle
0.08485
0.086625
0.08772
0.088205
0.0882
0.087675
0
0.0095
0.017965
0.03335
0.04644
0.05748
0.066675
0.07428
0.08043
0.08532
0.0891
0.0919
0.09383
0.09498
0.09555
0.09548
0.09495
0
0.011045
0.02089
0.03878
0.054
0.06682
0.077525
0.08637
0.09352
0.0992
0.10359
0.10685
0.109065
0.11046
0.111085
0.11102
0.1104
0
33
-------
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
HHD8
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
Urban Bus
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
idle
speed < 2.5mph
2.5mph <= speed < 7.5mph
7.5mph <= speed < 12.5mph
12.5mph <= speed < 17.5mph
17.5mph <= speed <22.5mph
22.5mph <= speed < 27.5mph
27.5mph <= speed < 32.5mph
32.5mph <= speed < 37.5mph
37.5mph <= speed < 42.5mph
42.5mph <= speed < 47.5mph
47.5mph <= speed < 52.5mph
52.5mph <= speed < 57.5mph
57.5mph <= speed < 62.5mph
62.5mph <= speed < 67.5mph
67.5mph <= speed < 72.5mph
72.5mph <= speed
0.023655
0.04474
0.08305
0.115635
0.14312
0.16605
0.18495
0.200305
0.21248
0.22185
0.2288
0.23364
0.23658
0.2379
0.23779
0.236475
0
0.01271
0.024035
0.04462
0.06213
0.0769
0.089225
0.09936
0.107625
0.11416
0.119205
0.12295
0.12551
0.12708
0.12779
0.12775
0.12705
34
-------
Appendix C Literature Review
Table C-l Brief review of literature on brake and tire wear
Luhana,L.;Sokhi,R.;Warner,L.;Mao,H;
Boulter,P;McCrae,I.S.;Wright,J and Osborn,D,"Non-
exhaust paniculate measurements:results," Deliverable
8 of the European Commission DG TrEn, 5th
Framework PARTICULATES project, Contract No.
2000 -RD. 11091, Version 2.0 , October 2004.
2004
Non-exhaust particle research was conducted in
the Hatfield road tunnel. Combined tire and break
wear emissions for PM10 from LDVs and HDVs in
the tunnel were found to be 6.9mg/vkm and
49.7mg/vkm respectively. These emission factors
from the Hatfield Tunnel Study appears to be at
the lower end of the range of values reported
elsewhere. The report also includes a literature
review which examines the state of the art in the
field. Tire wear and brake wear rates are listed
below.
Sanders, Paul G.;Xu, Ning ;Dalka, TomM; and
Maricq, M. Matti, "Airborne Brake Wear Debris: Size
Distributions, Composition, and a Comparison of
Dynamometer and Vehicle Tests".Environ. Sci.
Technol., 37,4060-4069,2003
2003
A brake wear study was performed using seven
brake pad formulations that were in high volume
use in 1998. Included were low-metallic,semi-
metallic and non-asbestos organic (NAO)
brakes.The quantity of airborne PM generated by
automotive disk brakes was measured on a brake
dynamometer that simulated : urban driving (low
velocity,low g) and the Auto Motor und Sport
(AMS,high velocity, highg). Airborne fractions
from the low-metallic and semi-matallic linings
were 5 and 1.5 times higher than the NAO lining.
L.R.Warner; R.S. Sokhi;
L.Luhana ; P.O. Boulter; and I. McCrae,"Non-exhaust
particle Emisions from Road Transport", Proceedings
of the 11th International Symposium on Transport and
Air Pollution, Graz, 2002.
2002
The paper presents preliminary results of
gravimetric determination of tire and brake wear
for cars, and chemical analysis of ambient particle
samples for source identification using Inductively
Coupled Plasma (TCP) spectrometry. Results
suggest that the average loss rates of tire and brake
material are 97 and 9 mg/vkm respectively. The
ICP analysis shows a high relative abundance of
Ba,Sb,Zr and Sr for brake and Zn for tire material.
The chemical analysis also suggests that for tire
wear it is much more difficult to use metal
concentrations as tracers.
Abu-Allaban, M.;Gillies, J.A.;Gertler,A.W.;Clayton
,R.; and Proffitt,D., "Tailpipe, re-suspended road dust,
and brake wear emission factors from on-road
vehicles," Atmospheric Environment, 37(1),5283-
5293,2002.
2002
Intensive mass and chemical measurements were
performed at roadside locations to derive brake-
wear emission factors from in-use vehicles. PM10
emission rates for LDSI vehicles ranged from 0 to
80 mg/vkm and for HDVs from 0 to 610 mg/vkm.
The PM2 5 emissions ranged from 0 to 5mg/vkm
for LDSI vehicles and from 0 to 15mg/vkm for
HDVs. Emissions from brake wear were highest
near motorway exits.
Lukewille,A.;Bertok,I.;Amann, M.,
Cofala,J.;Gyarfas,F.;Heyes,C.;Karvosenoja,N.;Klimont
35
-------
Z.; and Schopp, W., " A framework to estimate the
potential and costs for the control of fine paniculate
emissions in Europe",/Z4&4 Interim Report IR-01-
023,Laxenburg, Austria,2001.
Westerlund ,K.G.," Metal emissions from Stockholm
traffic -wear of brake linings ",The Stockholm
Environment and Health Protection Administration,
100,64,Stockholm,Sweden,2001.
2001
Westerlund estimated the amount of material lost
due to brake wear from passenger cars and heavy
goods vehicles. The PM10 emission factors were
determined to be 6.9 and 41.2mg/vkm for LDVs
and HDVs respectively.
Garg, B.D.; Cadle, S.H.; Mulawa,P.A.; Groblicki,
P.J.;Laroo,C.; and Parr,G.A., "Brake wear paniculate
matter emissions", Environmental Science &
Technology, 34(21),4463,2000b.
2000
A brake wear study was performed using seven
brake pad formulations (non-asbestos) that were in
high volume use in 1998. Brakes were tested on a
brake dynamometer under four wear conditions.
The brake application was designed to simulate
real world events by braking from 50km/h to
Okm/h at a deceleration of 2.94 m/s2. The
estimated range of PM emission rates for small
vehicles to large pickup trucks are 2.9 -7.5
mg/vkm and 2.1 -5.5 mg/vkm for PMi0 and
PM25 respectively.
Annette Rauterberg-Wulff, "Determination of
emission factors for tire wear particles up to lOum by
tunnel measurements", Proceedings of 8th
International Symposium on Transport and Air
Pollution, Graz, 1999.
1999
PM10 emission factors were determined for tire
and brake wear using receptor modeling in
combination with measurements conducted in the
Berlin-Tegel tunnel. Tire wear emission factors
for LDVs and HGVs in the tunnel was calculated
to be 6.1 mg/vkm and 31 mg/vkm. For brake wear
it was 1.0 and 24.5 mg/vkm respectively.
Carbotech, "PM10 Emissionsfaktoren:Mechanischer
"'^irbeitsunterlage, , 17,1999
1999
Cited in Lukewille et al.(2001). The PM10 brake
wear emission factor for LDVs was determined to
be 1.8 mg/km and for HDVs it was 3.5 mg/vkm.
Cha,S.; Carter,?.; andBradow, R.L., "Simulation of
automobile brake wear dynamics and estimation of
emissions,"£lE Transactions Paper, 831036, Society
of Automotive Engineers, Warrendale,
Pennsylvania, 1983
1983
Paniculate emissions from asbestos-based brakes
from automobiles were measured under conditions
simulating downtown city driving. The report
presents a systematic approach to simulating brake
applications and defining paniculate emissions.
Based on the 1.6:1.1 wear ratio between disc and
drum brakes,the estimated airborne paniculate
(PMio) emission rate was estimated to be
12.8mg/vmi or 7.9 mg/vkm.
36
-------
Appendix D Responses to Peer-Review Comments
This section provides a verbatim list of peer reviewer comments submitted in response to the
charge questions for the Brake and Tire Wear Emissions/Temperature Effects Report.
D. 1 Adequacy of Selected Data Sources
Does the presentation give a description of selected data sources sufficient to allow the reader to
form a general view of the quantity, quality and representativeness of data used in the
development of emission rates? Are you able to recommend alternate data sources might better
allow the model to estimate national or regional default values?
D. 1.1 Dr. Chris Frey
Table 2-1 of the report is a helpful summary of data sources that were reviewed by EPA as a
possible basis for estimating brake wear emission rates. EPA chose to base the brake wear
emission rates for light duty vehicles mostly on a study by Sanders et al. (2003). The brake wear
emission rates for heavy duty vehicles are based mostly on a study by Moahmoud Abu-Allabah
et al. (2003). The tire wear emission rates are based mostly on a study by Lahuna et al. (2004).
There is a need for more critical discussion of the representativeness of the data from these
studies for the U.S. onroad fleet. Since all three of these key studies are approximately a decade
old, a question naturally emerges in the reader's mind as to whether the brake lining, brake shoe,
or tire materials that were the basis of these studies are representative of materials currently in
use. Furthermore, are the vehicles measured representative of vehicles currently in use in terms
of the most relevant characteristics, such as vehicle weight, and factors such as the ratio of brake
pad or brake shoe area to vehicle weight, and tire tread dimensions to vehicle weight, and so on.
If these questions are not answerable, then explain why they cannot be answered. However, it is
important to indicate that these issues were at least considered, even if there is not a quantitative
basis upon which to make a judgment.
RESPONSE: An explanation was added to the paper, which describes that the paper was
originally drafted in 2006, and the literature review was current at the time. The
MOVES team had not the resources to update this paper in timeforMOVES2014 due to
the many other updates required. A Next Steps section was also added to the report
describing what a future study could update. As to whether the papers are
representative, this is a difficult question to answer, as these are the only papers that
could be found at the time, there simply is not a large amount of research conducted on
airborne tire and brake wear par ticulate matter emissions. The author of the report
attempted to analytically adjust the data to be as representative as practicable.
37
-------
D.I.2 Dr. JoeZietsman
The literature used as the basis for this work is sufficient. I am not aware of any other literature
that has been overlooked. The only concern is that the literature is quite old (newest study is
from 2004 for brake wear and 2005 for tire wear).
On Table 2-1, the vehicle type classification for the Warner et al study refers to both "passenger
cars" as well as "light duty". I am not sure what the distinction is or whether it is a typo. In the
Sanders work, UDS, UDP, and AMS should be defined. On looking at the source reference looks
like UDS may be a typo. It seems as though the description of the Sanders study on page 3 needs
to be corrected - for example, -7.9m/s should be -7.9m/s2; specify what g is in the context of the
decelerations, etc.
RESPONSE: These issues have been addressed in the report. The table has also been
shortened to only include papers with measurements; the papers with brake emissions
estimates based on models and literature reviews have been omitted.
On page 7 - first sentence should refer to accelerations if referring to Figure 2.5.
RESPONSE: This has been clarified in the report.
On page 15, effect of horizontal curvature is discussed. It is assumed vertical curvature or grade
could have an effect on tire wear, but it was not mentioned in this work.
RESPONSE: Agreed: thank you for the suggestion, this has been added to the report.
In Table 3-1 - there is no mention of the Luhana study
RESPONSE: We added the Luhana et al. (2003) study to Table 3-1.
In Table 3-3 and related text- it should be better clarified how total weight loss is calculated.
RESPONSE: the table description now includes the information that the tires were
weighed.
On Page 20, Table 5 is referred to as "above", and it is actually below; Table 6 is referred to, but
it doesn't exist.
RESPONSE: this has been corrected.
D. 2 Clarity of Analytical Meth ods and Procedures
Is the description of analytic methods and procedures clear and detailed enough to allow the
reader to develop an adequate understanding of the steps taken and assumptions made by EPA
38
-------
to develop the model inputs? Are examples selected for tables and figures well chosen and
designed to assist the reader in understanding approaches and methods?
D. 2.1 Dr. Chris Frey
No.
Consider Figure 2-1. This is apparently an excel worksheet that was copied into the report. The
figure is labeled as "UDP results." What is UDP? Spell it out. Results for what, exactly? The
table reports "PM10 emiss." for "filter" and "ELPI" (again, always make tables and figures self-
documenting - what is ELPI?). I looked in Sanders et al. (2003) to try to figure out where these
reported numbers come from. For UDP, low metallic, I find in Table 5 that "filter" results are
reported for "total mass" in mg/stop/brake. However, the numbers in the draft report are not the
same as from Sanders et al. (2003). For example, the low metallic filter PM10 emiss. is reported
as 6.9 (are the units mg/stop-brake? - not very clear given that units are not given for each
column of data). Let's assume that this is 6.9 mg/stop-brake. In Table 5 of Sanders et al. (2003),
I find that the individual tests range from 6.2 mg/stop-brake to 11.7 mg/stop-brake, with an
average of 8.3 mg/stop-brake. Thus, as a reader, I cannot figure out either why did EPA choose
one test from among the multiple reported by Sanders et al. (2003), or what adjustment did EPA
make from the average of 8.3 mg/stop-brake to arrive at 6.9 mg/stop-brake? On the other hand,
for the ELPI results, EPA seems to be reporting the same values as shown by Sanders et al.
(2003) for the average of all reported tests for each of the three types of brake pad linings.
RESPONSE: All of these have been clarified in the section 2.1. For example, the
acronyms have been spelled out and the units are more clear. Describing what exactly
these instruments do is beyond the scope of this report. We also removed UDP from the
heading of Figure 2-1. As for the reason why the paper's UDP number doesn 't match
Sanders et al.: We calculated 6.9 mg/stop/brake from the 8.2 mg/stop/brake. The filter
number reported in Sanders is TOTAL PMMass. In order to get the PM10 mass, the
ELPI ratios were used, thus 8.2 * (7/8.3). Likewise, 1.7 was determined from 1.7 *
(1.7/2) etc. A footnote was added with an example calculation.
Also unclear: how does EPA go from average/stop/brake (what is this... isn't this average
emission rate in mg/stop-brake) to "Avg./veh" (again, what is this - always report units). It
would help to show an example calculation. The ratio of avg./veh to 'average/stop/brake" is
approximately 3, which may be based on an assumption that the non-drive wheels have !/2 the
brake wear of the drive wheels. This could be more clear.
RESPONSE: The avgper vehicle emissions is the avg stop/veh/brake emissions
multiplied by 3. This is based on the assumption made earlier that 2/3 of braking comes
from the front brakes and 1/3 from the rear brakes. Footnote has been added.
39
-------
As far as the deceleration in Figure 2-1, where does this come from? Is this an average of all
decelerations from all stops during the UDP? This could be more clear. Is the value of 0.0012
km/s2 reported by Sanders et al. (2003) or was this inferred by EPA. If the latter,
how?Similarly, what is the basis for the "average brake time in sees" that is reported in Figure 2-
1?
Similarly, what is the basis for the "avg . emissions in mg/ stop" that is reported in Figure 2-1?
RESPONSE: The following footnote was added below the table (f). The UDP
decelerations are the average decelerations from those measured in the Sanders paper.
The average brake times were determined with the assistance of one of the original
authors of the paper (Matti Mariq) who supplied the second by second trace.
The bottom of Figure 2-1 includes "emission rate for the UDP test" of 2.65 gms/hr. Since the
previous lines indicate that time is for braking or per stop, clarification is needed as to the time
basis for 2.65 g/hr. Is this based only on braking time? Is this based on total travel or trip time?
RESPONSE: Yes, the emission rate is only during braking events and times as clarified
in the text near the figure.
Similar comments apply to Figures 2-2, 2-3, and 2-4.
Without clear documentation of how the emission rates were estimated, it is not possible to
comment on whether the curve fit in Figure 2-5 is reasonable. The fit shown is reasonable given
the numbers used in fitting the curve, but the basis for, and meaning of, the numbers is unclear.
Additional comments regarding needs for improved communication of the data and methods are
given in detail in a section on "Specific Comments"
D.2.2 Dr. JoeZietsman
The methods and approaches are adequate. It is not clear in this report how the exact
measurement (of the PM emissions) was conducted in the source studies, as well as the basis for
assumptions/measurements regarding apportionment (of what gets emitted into the air as PM10)
and what percentage of PM10 is PM2.5. For example, on Page 20 (last sentence) - 8% of tire
wear as PM10 assumption is not referenced.
RESPONSE: these have been addressed throughout the document. For example the 8%
text is made more clear on page 22.
D.3 Appropriateness of Technical Approach
Are the methods and procedures employed technically appropriate and reasonable, with respect
to the relevant disciplines, including physics, chemistry, engineering, mathematics and statistics?
40
-------
Are you able to suggest or recommend alternate approaches that might better achieve the goal of
developing accurate and representative model inputs? In making recommendations please
distinguish between cases involving reasonable disagreement in adoption of methods as opposed
to cases where you conclude that current methods involve specific technical errors.
D. 3.1 Dr. Chris Frey
EPA is making use of data collected by others and reported in the literature. The selected
references generally seem to be reasonable.
The methods and procedures employed by EPA include judgments about the representativeness
and appropriateness of the selected data sets for use as a basis for developing MOVES inputs,
and regarding the statistical analyses conducted based on the selected data. Given lack of
sufficient data from which to develop more detailed models, EPA has developed relatively
simply models. The general approach is reasonable. However, the communication of what was
done could and should be more clear and complete. Ideally, sufficient information should be
communicated regarding the underlying data and inference approaches such that an independent
investigator can reproduce the results and obtain the same answer. Many of the detailed
comments given below under "specific comments" are aimed at this objective.
RESPONSE: These comments have been addressed throughout the report as described in
greater detail in the more detailed comments.
D.3.2 Dr. JoeZietsman
No response.
D. 4 Appropriaten ess of Assumptions
In areas where EPA has concluded that applicable data is meager or unavailable, and
consequently has made assumptions to frame approaches and arrive at solutions, do you agree
that the assumptions made are appropriate and reasonable? If not, and you are so able, please
suggest alternative sets of assumptions that might lead to more reasonable or accurate model
inputs while allowing a reasonable margin of environmental protection.
D. 4.1 Dr. Chris Frey
In general, I agree that EPA has done a reasonable job with a very limited data set to make
inferences and develop data and fitted models for use with MOVES.
D.4.2 Dr. JoeZietsman
No response.
D. 5 Consistency with Existing Body of Data and Literature
Are the resulting model inputs appropriate, and to the best of your knowledge and experience,
reasonably consistent with physical and chemical processes involved in exhaust emissions
41
-------
formation and control? Are the resulting model inputs empirically consistent with the body of
data and literature that has come to your attention?
D.5.1 Dr. Chris Frey
Yes.
D.5.2 Dr. JoeZietsman
With regard to the above three questions, based on the available data the methods are appropriate
and reasonable. I concur with what is noted in the report regarding improvements with more
data, for example allocating tire wear by VSP bin, etc. I am not aware of any current datasets or
methods that can enhance this work.
D. 6 Tire and Brakewear PM2.5 and PM10 Emission Rates and
Speciation
MOVES2014 estimates total PM2.5 andPMlO emission rates from brake and tire wear.
Additional PM2.5 speciation capabilities have been incorporated in MOVES2014 for exhaust
emission processes. What recommendations do you have for EPA for incorporating the PM2.5
andPMlO speciation of tire andbrakewear for future versions of MOVES?
D. 6.1 Dr. Chris Frey
Clearly, it will be desirable to include speciation of PM10 and PM2.5 into future versions of
MOVES for brake and tire where when sufficient data are available to support such estimates. In
the current report, EPA could provide at least a paragraph summarizing what is known about the
PM composition of brake wear debris, to expand upon some text on page 3. To the extent that
there is or isn't information on PM composition of tire wear, EPA should add at least a paragraph
to discuss this and, if applicable, summarize available information. Of particular interest is what
are the key indicator species or components of brake wear and of tire wear, and are these
sensitive to the materials used, or is there insufficient information to address these points? There
is no information offered in the section on tire wear regarding the chemical composition of tear
wear debris or emissions. Is this because no information is available?
RESPONSE: The literature had very limited discussion of PM speciation. There were
some measurements of this, however, it was beyond the scope of this current modeling
exercise. We added discussion in Section 4 Next Steps, that we would like to do more
work on PM speciation of brake and tire wear.
D.6.2 Dr. JoeZietsman
While the additional PM2.5 speciation capabilities in MOVES 2013/2014 (in terms of added
species, as relevant) will be useful, a first step for brake and tire wear emissions would be to
better estimate and justify the following factors: a) PM-10 to PM2.5 ratio; b) percentage of brake
and tire wear that is actually emitted as PM-10 and PM2.5.
42
-------
RESPONSE: We hope to improve on these emission factors as we become aware of new
literature and data. We have added a statement that we would like to improve the
information on size distributions in the Next Steps Section.
D. 7 General/Catch-All Review
Please provide any additional thoughts or review of the material you feel important to note that
is not captured by the preceding questions.
D. 7.1 Dr. Chris Frey
This is a significant report that documents an important part of the MOVES emission factor
model, which is used nationally for a wide variety of regulatory and other analyses. As such, it
is critically important that the report be well written and very clear. While the current draft of
the report is good in many respects, it comes across as a draft and is not in final form in terms of
the critical thinking needed to make sure that it clearly communicates information to the reader.
For each of the major sections on brake and tire wear, it will help the reader to have clearly
labeled sections that deal with light duty vehicles and with all other vehicle source categories. It
will also help to clearly define and consistently use terms and concepts. For example, it is
critically important to have a clear quantitative definition of break wear and of tire wear, and to
be clear as to whether these are rates per wheel, per vehicle, or other (specify).
RESPONSE: We added Light- Duty to Section 2.23, and Heavy-duty and Other Vehicles
to Section 2.2.5.
Since this is a formal technical report, the use of first person should be avoided. Statements
regarding how an analysis was done, or regarding judgments that were made, can be made
without self-reference.
In general, be careful about significant figures. It is pretty rare in this type of work that data are
known with more than 3 significant figures. However, in various places, numbers are reported
with 5 or 6 significant figures, and often with 4. Even if the original data might be known with
many significant figures, its adoption for use in representing a national fleet introduces
uncertainty, since the original data may not represent the U.S. national fleet as it exists today.
RESPONSE: Significant figures for intermediate numbers were largely maintained. The
final numbers were appropriately adjusted. Some of the figures/tables have reduced sig
figs now as well.
Many specific comments are given below that elaborate on responses given above in response to
the charge questions.
43
-------
Specific Comments (numbers refer to page/paragraph/line ... e.g., 3/3/1 means page 3, paragraph
3, line 1)
RESPONSE: Unless otherwise noted, all minor comments have been addressed in the
report in the relevant sections. Responses are provided to the major comments.
Given that this report will be finalized perhaps in 2013 or 2014, explanation is needed for the
reader as why the literature review starts off with information from 2004. Is this because more
recent data are not available?
Page 2, 3rd paragraph in Section 2.1, next to last line "is acceptable" does not fit here.
Page 2, 4th paragraph in Section 2.1: while these statements seem intuitively reasonable, they
are stated as if they are known facts. However, no references are cited. How do the authors
know that these are accepted facts? Or are these the author's hypotheses or opinions? If factual,
then cite reference(s). If these are hypotheses, then say so.
RESPONSE: we believe that these statements in this particular paragraph are obvious to
anyone with a basic knowledge of vehicle brakes. While not all readers of this paper may
have knowledge of brakes, some of the text in this paper are meant to be include basic
instructive material as well. As such, we do not believe that relatively "basic " statements
like this require references. However, the subsequent statements in the following 2
paragraphs are more specific. We 've added a parenthetical when the Garg paper is first
mentioned that much of the basic PM information comes from this paper. This paper is
very important in the list of this study's references.
Please carefully define what is meant by "wear" and then use the definition consistently. Does
"wear" refer to a mass rate of emissions or loss per tire per braking episode, or is it a time-based
rate for a tire (or a vehicle), or a mileage-based rate? If this is not defined, then readers will
make their own assumptions as to what this means. Does "greater wear" mean greater rate of
wear, or more accumulative lost mass regardless of time period?
RESPONSE: the definition of wear versus airborne PMhas been differentiated in various
points in the paper. Wear means slightly different things in the literature, but in this
paper it is generally, the mass of material lost, whether in the brake pads or the tires. A
fraction of that wear is airborne PM. Some studies look at both wear and airborne PM,
others look at one or the other. We added discussion providing a definition of brake
wear, and airborne PMin Section 2.1.
3/2/1: Page 3, 2nd paragraph, a 2000 study is not "recent" in 2013... delete "In recent studies,"
3/2/6: "ranged from 3.4 mg/mile to 4.6 mg/mile" is the correct way to write the range at the end
of the same paragraph.
3/3/1: "currently used" - refers to 2003 or to now?
44
-------
3/3/7: this list is hard to read because apparently not all brake linings were measured in all types
of tests, but the reader has to reread this a few times to really figure this out.
3/3/9: -7.9 m/s is not an acceleration. Is this supposed to be-7.9 m/s2?
3/3 - near end of paragraph... it would help to give a summary of how the PM composition
varies by brake type.
RESPONSE: We added as much information as we could, but we do not have perfect
information.
End of page 3 - why is there no discussion of other references, particular ones published shortly
before or after Sanders et al. (2003), such as Warner et al. (2002), Abu-Allaban et al. (2002), and
others. From Table 2-1, the reader infers that some of these could also be useful. If EPA judged
that they are not useful, a rationale should be given.
RESPONSE: a paragraph was added to the paper here.
Related to page 3 - is the current market share of each brake lining type known? If so, please
summarize. If not, then say so.
RESPONSE: We added a sentence saying that this is not known in this section.
4/1/1: "is based on the average of the" is better than "averages"
4/3: why are only results from Sanders et al. (2003) used here... give an explanation.
5/1/1: how sensitive are results to the assumption regarding equal mix of brake types?
RESPONSE: this is answered indirectly earlier in the report where there is a summary
provided from the paper regarding the differing results from the varying materials.
5/1/2: is there some basis for the assumption that 2/3 of braking power is in the front brakes?
Actually, the assumption made here is that the rate of brake lining wear is twice that for front
drive wheels than for rear nondrive wheels. Is the assumption really based on "power"?
RESPONSE: this information was provided'by'Matti Matiq from Ford who is a co-author
on the Sanders paper and helped us with the data and information like this in this report.
This is a "rule of thumb " in the industry. It is also consistent with what is written in
Garg and that is now added to the footnote.
5/1/4: what is meant by "total PM"? is this total suspended particulate matter?
Page 8 - include definitions of opmode bins 0 and 1.
8/1: to reader it is unclear as to why 1,497 kg is used. Explain that this is a typical weight of a
sedan passenger car. Is this the only selected weight? Why not others? What about larger
vehicles?
45
-------
RESPONSE: this is the average weight of passenger cars in MOVES. Other vehicle type
weights are provided elsewhere in the report. We added this to the report.
Page 9, figure 2-7: "coastdown curves" for what, based on what data sources? Figure and table
captions generally need to be more specific throughout the report. Also, the number of
significant figures given for the equation borders on the absurd.
RESPONSE: PERE is a vehicle simulation model, that has undergone separate peer
review. A complete description of how these plots were developed would be lengthy and
distract from the focus of the report. We do not believe that it is required for what is a
minor part of the quantification of the braking emissions. However, more detail was
added with three figures on the PERE coast down and braking estimates in Appendix A.
9/1/3: "Eastern" not "Easter"
9/2/2: it is very unclear as to how the distribution of braking activity across speed and
deceleration can be determined from the numbers given in Table 2-7.
9/2/2: "vast majority of braking" - how is this known? What is a "minor slowdown"? (give
quantitative criteria for this).
RESPONSE: The following table quantitatively shows deceleration frequencies drop as
braking becomes more severe. We disagree that an actual quantitative definition of mild
and full decelerations is required here. This is not a distinction that is made in the
MOVES model.
11: Table 2-8: in text or footnote, show the equations that define or use the constants given such
as "wgt", "CrO", "Cd", "A"... and preferably use mathematical nomenclature. Also, do not
report "mA2". This should be "m2" Please use source bin terminology and please define what
the columns are in this table... i.e. the last 5 columns should have a superheader of "Vehicle
Type." Does MOVES have source ids for "Compact," "Mid-size," "SUV," "mddt," and
"tractor"?. If these are assumptions that are meant to apply to source bin ids in MOVES, then
use proper MOVES terminology to avoid confusion.
RESPONSE: MOVES does not have source types for compact, mid-size or SUV. These
source type definitions are typical for vehicle simulation models like PERE. These
models are capable of greater precision than MOVES, and an attempt is made here to
model greater detail first, then aggregate the results up to MOVES source bins.
Table 2-8 has some definition problems, in that the first four rows of numbers are not fractions
and thus should not be defined as such, and yet the last 7 rows are in a sense undefined, because
there is no header in the table that defines what they are. There is inconsistency in the number of
decimal places given. Also, the basis for these numbers is unclear - where do they come from
and how exactly where they derived?
46
-------
RESPONSE: The table has been cleaned up. Also more information has been included in
the text near this table as well as appendix.
11/1: it would help to have an example calculation showing how these emission rates were
estimated.
13/1: please give some rationale as to why this study was used and/or why others were not. Was
this the only relevant study? The most recent study? The best study? Just stating that it was
used is not sufficient to explain to the reader why it was used. Also, please indicate what type of
instruments were used and at least some summary of how the sampling was done, and how it was
possible to associate the measurements with a per vehicle emission rate specific to brake wear.
RESPONSE: a lengthy description has been added on page 13 and follow ing.
13/1/7: "by ratioing"... ratio is not a verb, "by taking the ratio of..." More importantly, please
explain the empirical basis for these ratios. Upon what measurements are they based, and how
was it possible to distinguish among the vehicle categories - e.g., how was it possible to
apportion measurements to vehicle types?
Table 2-2: the caption "scaling to other vehicle class" may mean something to the author, but
lacks sufficient detail and specificity to mean much to the reader. What exactly is contained in
this table and what is the source or basis of the information? Tables and figures should be self-
documented to the extent possible, to make very clear as to what is the content. Please define or
explain specialized terms - e.g., "regclasswt" and "regclassid" are variables used in MOVES
(define them). What is the basis of the weight ratio? Is this based on weights given in MOVES
(for clarity, report the weights so that the basis of the weight ratios is more clear). What is the
source of the "mg/mi" and why is there no description of what this is in the column header?
Similarly, "gms/hr**" has no explanatory title nor is there any explanation of whether "**" is
meant to be a footnote. This type of sloppy and incomplete documentation of tables leads to
long-term confusion as to the basis of data contained in MOVES, and causes a lot of problems
for MOVES users for many years to come. Thus, it is essential that these reports be well
documented.
RESPONSE: A paragraph has been added before Figure 2-4 in addition to a Figure 2-4
and additional tables (Tables 2-12, 2-13) to help clarify.
13/2/2: "a PM10/PM2.5 ratio" - is one ratio used regardless of opmode bin? Source category?
Etc.? Needs to be more clear.
RESPONSE: this has been clarified in the report that there is one ratio used for
simplification.
13/2: The PM10/PM2.5 ratio is "based on the assumption that the mass fraction of particles
below PM10 is 0.80, and that the mass fraction of particles below PM2.5 is 0.1" D this is
suggested revised text. I do not think that the authors really mean to refer to the fraction of
47
-------
particle number, but rather mean to refer to fraction of particle mass, yet the text implies that
these ratios are by particle number. There is a big difference. Also, explain the basis of these
assumptions. Are these mass fractions of 0.8 (why 0.80 - 2 sig figs) and 0.1 (why only 1 sig fig)
based on measurements, a wild guess?
RESPONSE: Additional clarification on the size distribution used to derive the
PM10/PM2.5 fraction from Sanders et al. (2003) is provided in Section 2.2 A.
13/2/6: what are examples of cases either in which the PM2.5 values are known, or are not
known, so that the reader has some idea of how extensive is the reliance on the assumption?
13/3: could delete "Tires are an essential part of any vehicle and"
RESPONSE: this is a matter of style. We disagree on this deletion.
14/2: this paragraph ("Tire wear occurs through...") needs references. Some of the points here
are repetitive of previous text.
14/4 ("The key influences...") this text comes across as repetitive although it does more
specifically refer to vehicle characteristics, whereas other lists given previously seem not be
specifically about vehicle characteristics. However, in general, this material could be better and
more clearly organized with less repetition and with citation to literature to support statements of
apparent fact that presumably have an empirical basis.
14/5 "Retreads are considered..." "considered" is extremely vague, "estimated" would be
better. However, the text that follows does not support this claim. It merely indicates that
retreads have less tire volume than the original tread. Why does this imply more wear? And
what does more wear really mean - a higher rate of wear in terms of mass per stop, mass per
time, mass per mile?
RESPONSE: this is another issue that is commonly known. The references include some
other interesting facts, however, the report now lists that modeling emissions from
retreads is beyond the scope of the report in Section 3.1 .
14/6: any statement that starts with "According to the literature" must end with one or more
cited references.
RESPONSE the subsequent paragraphs and table have many references on the technique
that is introduced in this sentence.
Page 15: Table 3-1 appears but I cannot find anywhere in the text where it is mentioned or
discussed. If it is not mentioned or discussed, why is it here, or if it is here, why isn't it
mentioned or discussed? Also, clarify specifically what is meant by the tire wear rates - are
these rates per vehicle and, if so, based on how many tires per vehicle? The mass reported here
is for what PM size range? Why isn't Luhana et al. (2004) summarized in this table, since it is
the only reference really used?
48
-------
15/1/1-2: delete "that have been carried out"
15/2/2: "m9ass" should be "mass"
15/3/2: delete "That being said" (too colloquial for a formal technical report).
15/4/2: change to "with results reported either as"
16/1: there is a disconnect between stating that data are primarily from one reference and yet
that reference is not included in the summary table just above.
16/1/middle - there are syntax problems here.
Table 3-2: caption is too vague. What type(s) of vehicles? What type(s) of tires? What is the
source of these data? The table is not very clear either - e.g., mean tire weight loss "mg/km" -
does this refer to per tire or per vehicle? The numbers reported in the table have inconsistent
numbers of decimal places and/or significant figures.
17/1: does mg/vkm refer to all tires on a vehicle, or one tire? Clarify throughout.
17/2/1: "the tables below" - cite the specific table(s).
Table 3-3. The caption "Data used for the analysis" takes the prize for vagueness. I don't think
there could be a more meaningless caption. The caption should specifically indicate the
content of the table. The lack of thought that went into this table is frustrating for the reader.
These are some examples: "Front-wheel drive vehicles only" is not a valid column header, it
should be part of the caption; "units" does not describe what is in the first column, yet why
not use the same labels for both Tables 3-2 and 3-3 to denote the same tests (i.e. be
consistent); the average trip speed of the first data row of 90.3 km/hr is not the same as the
corresponding number in Table 3-2; is the tire weight loss per tire or the sum of both wheels
on the axles? i.e. are the units g/km-axle or g/km-tire?; total weight loss for all tires (could
be more clear) and could have one superheader over both of the columns that contain this
information. Why is average speed given in the 2nd and 7th columns, rather than in adjacent
columns if the only difference is a units conversion?; where do these data come from (what
reference), and why not (in the footnote) specifically state the year, make, and model of each
vehicle?
RESPONSE: This and other tables throughout the report have more descriptive captions
now. The table has also been reformatted and redone so that the information is more
clear to the reader.
Figure 3-1 and 3-2: the caption should indicate the vehicle type and whether the g/mi is per tire
or per vehicle. If per vehicle, based on the assumption of how many tires/vehicle? Also, are
mean trips speeds based on a particular time period of travel? One could divide a trip arbitrarily
into short segments and obtain a wider range of mean trip speeds. However, if the averaging
time is something like 60 seconds, 600 seconds, or 3000 seconds, that would affect the
appropriate use of these rates.
49
-------
RESPONSE: The paper does not specify the trip times.
Figure 3-2 - this figure seems unnecessary. The fitted curve is in Figure 3-2. If the purpose is to
show the curve extrapolated beyond the range of the observed data, this could be done in figure
3-1 by using a dashed line for the fitted curve when it is outside the range of mean trip speeds.
RESPONSE: this figure has been removed.
19/1/1: rewrite as "Based on extrapolation of the fitted curve, wear is highest at zero speed and
decreases as the speed increases." Why is this counter-intuitive? Lower mean trip speed might
imply more variability in speed and, hence, more acceleration/deceleration(braking) that could
increase tire wear. Thus, to me, this curve is intuitive. Delete "it is important to note" which is a
passive phrase that has no content. The statement that "the relationship" does not take
accelerations (and turning) into account is probably false. Aren't these data based on real-world
driving? If so, then the observed wear rates implicitly take into account these factors, and the
curve is fitted to these data - thus, the curve implicitly accounts for these.
RESPONSE: more text is included in this section to help clarify why the authors believe
why some readers may believe this to be counterintuitive.
20/1/2: how is it known that the vehicle is "braking hard"?
RESPONSE: this is a descriptive term, not scientific. We do not feel that change to the
text is required.
20/1/2-4: text here is a bit awkward... could say that there is insufficient data to characterize tire
wear on a second by second basis to enable binning by operating mode bins.
20/2/1: this first sentence needs to be stated up front - i.e. that this section focuses on LDVs.
Tire wear rates for other than LDVs should be in a new section. The assumption that tire wear is
just based on the number of tires also presumes that tire wear for larger tires is the same as for
LDV tires, and that tire wear is the same regardless of the ratio of weight/number of tires on a
vehicle. These assumptions should be more clearly enumerated and discussed. Might it be the
case that all else being equal (e.g., weight, acceleration, speed, road surface, etc.), tires with
larger tread surface in contact with the ground would have less wear, but that more weight per
tire would increase wear?
RESPONSE: We added a sentence at the beginning of Section 3.2 stating that the
analysis starts with light-duty, and then is extended to other vehicle types. A number of
engineering assumptions were required here. There is clearly a lack of information in
the literature (as of when this study was completed) that would give emission rates for a
variety of regulatory classes. A paragraph was added on future work.
20/2/4: Should say "The latter data were found in the ..." s
Table 3-4: does the 2002 VIUS Survey report the number of tires per truck? Or the number of
axles? If the latter, then it is not valid to imply that the number of tires is from the survey.
50
-------
However, in either case, the survey can be cited with appropriate explanatory text in a footnote.
The report is inconsistent in how vehicle categories are defined and described. If "regclass" is
actually used here, which doesn't' seem quite right (aren't these source id?), then also give the
MOVES code associated with each for complete clarity. There is no way that there are 519
billion LDVs in the U.S. - something is wrong with all of the numbers in the 3rd and 4th
columns of this table. Some explanation is needed of the "Tires Per Vehicle" - this must be
"Average Tires Per Vehicle," since no vehicle can have 5.5 or 14.9 tires, and 7 tires would also
be quite unusual. Some explanation of the basis for the number of tires per vehicle for each
vehicle type is needed. The assumption that the number of tires for LDT is the same as for LDV
seems reasonable, but should be explained - i.e. LDT includes SUV, minivan, and pickup trucks
which are typically 4 wheels. Larger pickup trucks that might have 6 wheels are not considered
or do they fall into another category?
RESPONSE: We added clarification in the Table heading, that the tires were calculated
from the 2002 VIUS Survey axle count. We also removed the survey weighting factors
that were not helpful from for the reader. We also, specify that the Tires are calculated as
the averages in the Table headings, which can yield the non-integer numbers. As
discussed in the MOVES2014 Population and Activity Report, trucks with 6 wheels are
classified in RegClassID 41, not as LDT.
20/3/1: text refers to "Table 5" - should this be Table 3-5? (but Table 3-5 seems to be on a
different topic).
20/4: This paragraph is unclear. It starts by stating that probably less than 10% of car tire wear
is emitted as PM10 based on Boulter (2005a) and then goes on to say that results of Kupiainen et
al. (2005) were used by Boulter. If Kupiainen is the original source of the data/information, then
this could be more clear and conveyed consistently in the paragraph. Also, here and in general
throughout the report, clarify if the mg/vkm values are per tire or per vehicle.
RESPONSE: this sentence was removed. Kupiainen and Boulter are research
collaborators and co-authors, judging from their series of publications.
20/5: what is the basis for assuming 8% of total tire wear is PM10 and that 15% of PM10 is
PM2.5 For clarity, what happens to the other 92% of tire wear? Is it assumed that none of it
ever re-emitted to the atmosphere?
RESPONSE: References and description of how the 8%, and 15% mass fractions were
selected are added in Section 3.2. We also added a footnote (h) describing how the
PM10/PM2.5 ratio is derived from the PM mass fractions in Section 3.2.1.
Table 3-5: please clarify if the g/mi tire wear is per vehicle or per tire. What is "avgbinspeed"
and "RegclassID" - should define/explain in footnotes. The last 8 columns need a superheader
of "Regulatory Class" and it would help if each of the individual columns included descriptive
51
-------
text in the header. Avebinspeed should have units. SpeedBinID and OpModelD also need to be
defined/explained.
Response: The previous table is removed, and is replaced with Table 3-4, which includes
output byMOVES2014 by sourcetypelD. We also specify that the tire wear emission
rates are per/vehicle-mile. We also included information on the exact rates used in
MOVES (g/hour), in Appendix B.
21/1: for clarity, insert information that 15% of the mass of PM10 is estimated to be emitted as
PM2.5.
For the references, the reference format does not seem to follow any standard convention.
Response: We have provided information such that readers can find the sources.
D. 7.2 Dr. Joe Zietsman
Overall, this report on brake and tire wear emissions in MOVES 2014 is well documented and
satisfactory. Most of my comments are related to clarifications that are needed rather than
methodological issues.
52
-------
References
1 Harrison, R.M., A. M. Jones, J. Gietl, J. Yin, D. C. Green, "Estimation of the Contributions of
Brake Dust, Tire Wear, and Resuspension to Nonexhaust Traffic Particles Derived from
Atmospheric Measurements," Environmental Science & Technology, 2012.
2 Edmunds.com, http://www.edmunds.com/car-technology/brakes-drum-vs-disc.html
3 Cha, S., P. Carter, R. Bradow., "Simulation of Automobile Brake Wear Dynamics and
Estimation of Emissions", SAE Paper 831036. Society of Automotive Engineers, Warrendale,
PA, 1983.
4 EPA, PARTS Documentation
5 Garg, B. D., S. H. Cadle, P. A. Mulawa, P. J. Groblicki, C. Laroo, G. A. Parr, "Brake Wear
Particulate Matter Emissions", Environmental Science and Technology, 34(21), 4463-4469,
2000.
6 Sanders, P.G., N. Xu,T. M. Dalka, M. M. Maricq, "Airborne Brake Wear Debris: Size
Distributions, Composition, and a Comparison of Dynamometer and Vehicle Tests",
Environmental Science & Technology,37 (18), 4060-4069, 2003.Note that 7 is missing. Maybe
this will be cleaned up after all track changes are accepted.
7 EPA, Development of Emission Rates for Light-Duty Vehicles in the Motor Vehicle Emissions
Simulator (MOVES2010),198 pp, EPA-420-R-11-011, August 2011)
8 Nam, E.K., Giannelli, R., "Fuel Consumption Modeling of Conventional and Advanced
Technology Vehicles in the Physical Emission Rate Estimator (PERE) ", EPA document number
EPA420-P-05-001, 2004
9 EPA report, "Kansas City PM Characterization Study". EPA420-R-08-009. 2005.
10 Sierra Report No. SR02-07-04, "Task Order No. 2 SCF Improvement - Field Data
Collection," July 2002
11 Sierra Report No. SR02-07-03, "Task Order No. 3 SCF Improvement - Vehicle
Instrumentation and Instrumented Vehicles," July, 2002
12 Sierra Report No. SR02-07-05, "Task Order No. 7 SCF Improvement - Driving Data
Collection, South Coast Air Basin," July, 2002
13 Abu-Allaban, M., Gillies, J.A.,Gertler,A.W., Clayton ,R., Proffitt,D., "Tailpipe, re-suspended
road dust, and brake wear emission factors from on-road vehicles," Atmospheric Environment,
37(l),5283-5293, 2002.
14 National Research Council" Modeling Mobile-Source Emissions Committee to Review EPA's
Mobile Source Emissions Factor (MOBILE) Model, Board on Environmental Studies and
Toxicology", Transportation Research Board, National Research Council 200.
15 Dunlop South Pacific tires report to the Ministry of Transport, Government of New Zealand
53
-------
16 Maitre, O., Siissner, M., Zarak, C., "Evaluation of Tire Wear Performance", SAE Technical
Paper 980256, doi: 10.4271/980256, 1998.
17 Cenek, "Tyre Wear Modleing for HDM4", 1993.
18 Lowne, R. W., "The Effect of Road Surface Texture on Tire Wear," Vol. 15, pp. 57-70, 1970.
19 Bennett, C.R., Greenwood, ID., "Modeling road use and environmental effects in HDM -4".
HDM-4 Ref. Vol. 7. World Road Association, Paris, 2001.
90
Carpenter, P., P. Cenek, "Tyre Wear Modeling for HDM4," Opus International Consultants,
Limited, New Zealand, 1999.
91
Hildemann L. M., Markowski G. R., Cass G. R., "Chemical composition of emissions from
urban sources of fine organic aerosol", Environmental Science & Technology 25, 744-759,
1991.
99
Rauterberg-Wulff A., "Determination of emission factors for tyre wear particles up to lOjim
by tunnel Measurements", Proceedings of 8th International Symposium 'Transport and Air
Pollution', 1999.
23 Luhana L., Sokhi R., Warner L., Mao H., Boulter P., McCrae I, Wright I, Osborn D.,
"Measurement of Non-exhaust Particulate Matter", Deliverable 8 of Particulates-project.
European Commission, DG TrEn, 5th Framework Programme. 96 p, 2004.
24 Kupiainen, K.J., Tervahattu, H., Raisanen, M., Makela, T., Aurela, M., Hillamo, R., Size and
composition of airborne particles from pavement wear, tyres, and traction sanding.
Environmental Science & Technology 39, 699e706, 2005.
9S
Councell, T.B., Duckenfield, K.U., Landa, E.R., Callender, E.,. Tire-wear particles as a source
of zinc to the environment. Environmental Science & Technology 38, 4206-4214, 2004.
r\r
Warner L., R.S. Sokhi, L.Luhana, P.G. Boulter, I. McCrae, Dept. of Environmental Sciences
Univ. of Hertfordshire,UK, Non-exhaust particle emissions from road transport, 11th
International Conference "Transport and air pollution", 2004.
97
Kolioussis, M., Pouftis, C., "Calculation of tyre mass loss and total waste material from road
transport", Diploma Thesis, Laboratory of Applied Thermodynamics, Report No. 0010,
Thessaloniki, Greece 2000.
28 EMPA "Anteil des Strassenverkehrs an den PM10 und PM2.5 Imissionen", NFP41, Verkehr
und Umwelt, Dubendorf, Switzerland, 2000.
SENCO (Sustainable Environment Consultants Ltd.) "Collation of information on parti culate
pollution from tyres, brakes, and road surfaces", 23 March, 1999, Colchester, Essex, UK 2000.
30 Exemplarische Erfassung der Umweltexposition Ausgewaehlter Kauschukderivate bei der
bestimmungsdemaessen Verwendung in Reifen uind deren Entsirgung. UBA-FB 98-003
31 Garben et al. Emissionkataster Kraftfahrzeugverkehr Berlin 1993, IVU GmbH Berlin,
Gutachten im Auftrag der Senatsverwaltung fur Stadtenwicklung, Umweltschutz und
Technologie, Berlin, unveroeffentlich, 1997.
54
-------
32 Gebbe et al. Quantifizierung des Reifenabriebs von Kraftfahrzeugen in Berlin, ISS-
Fahrzeugtechnik, TU Berlin, i.A. der Senatsverwaltung fur Stadtenwicklung, Umweltschutz und
Technologie, Berlin, 1997.
33 Lee, P.K., Touray, J.C., Baillif, P., Ildefonce J.P., "Heavy metal contamination of settling
particles in a retention pond along the A-71 motorway in Sologne, France", The Science of the
Total Environment, 201, 1-15, 1997.
34 Baekken, T. , "Environmental effects of asphalt and tyre wear by road traffic", Nordisk
Seminar-og Arbejdsrapporter 1992:628 Copenhagen, Denmark 1993.
35 Schilling, D. I, Clark, J. D., Rubber Chemistry and Technology, 61, 669-687, 1988.
36 Malmqvist, P. A. "Urban storm water pollutant sources", Chalmers University, Gothenberg,
Sweden, 1983.
37 Gottle, A.,Ursachen und Mechanismen der Regenwasserverschmutzung - Ein Beitrag zur
Modellierung der Abflussbeschaffenheit in st dt. Gebieten. Berichte aus Wassergutewirtschaft
und Gesundheitsingenieurwesen, TU Munchen H.23, 1979.
TO
Cadle, S. H., Williams, R. L., "Gas and particle emissions from automobile tyres in laboratory
and field studies", Rubber Chemistry and Technology, 52(1), 146-158, 1978.
39 Dannis, M. L., "Rubber dust from the normal wear of tyres", Rubber Chemistry and
Technology, 47, 1011-1037, 1974.
40 Luhana, L., Sokhi, R., Warner, L., Mao, H , Boulter, P., McCrae, IS., Wright, J. Osborn, D,,
"Non-exhaust particulate measurements:results," Deliverable 8 of the European Commission DG
TrEn, 5th Framework PARTICULATESproject, Contract No. 2000 -RD. 11091, Version 2.0 ,
October 2004.
55
------- |