Final
Report
CRC
Project
E­
24­
1
Measurement
of
Exhaust
Particulate
Matter
Emissions
from
In­
Use
Light­
Duty
Motor
Vehicles
in
the
Denver,
Colorado
Area
03/
16/
98
Prepared
for:
Timothy
C.
Belian,
Deputy
Director
Coordinating
Research
Council
219
Perimeter
Center
Parkway,
Suite
400
Atlanta,
Georgia
30346­
1301
Prepared
by:

Steven
H.
Cadle,
Patricia
Mulawa
and
Eric
C.
Hunsanger
General
Motors
R&
D
Center
MD
480­
106­
269
Warren,
MI
48090­
9055
Ken
Nelson,
Ronald
A.
Ragazzi,
Richard
Barrett,
and
Gerald
L.
Gallagher
Colorado
Department
of
Public
Health
and
Environment
Air
Pollution
Control
Division
15608
E.
18th
Ave.
Aurora,
CO
80011
Douglas
R.
Lawson
Cooperative
Institute
for
Research
in
the
Atmosphere
Colorado
State
University
Fort
Collins,
CO
80523
Kenneth
T.
Knapp
U.
S.
Environmental
Protection
Agency
MD­
46
Research
Triangle
Park,
NC
27711
Richard
Snow
Clean
Air
Vehicle
Technology
Center,
Inc
Table
of
Contents
Executive
Summary
....................................................................................................................
i
List
of
Abbreviations
and
Acronyms
Used
in
this
Report
........................................................
vi
1.
Introduction
..........................................................................................................................
1
2.
Experimental
.........................................................................................................................
5
Vehicle
recruitment
.................................................................................................................
5
Vehicle
inspection
...................................................................................................................
7
Vehicle
testing.........................................................................................................................
7
Gaseous
emissions
..................................................................................................................
8
PM
emissions........................................................................................................................
10
Sample
analysis
....................................................................................................................
11
Fuel
......................................................................................................................................
13
Smoking
vehicle
population...................................................................................................
13
3.
Quality
Assurance/
Quality
Control
....................................................................................
15
Site
correlations
....................................................................................................................
15
Emissions
variability.............................................................................................................
17
Filters
...................................................................................................................................
21
Carbon
measurements
...........................................................................................................
23
Teflon­
Pallflex
comparisons
..................................................................................................
26
4.
Emission
Rates
of
PM
and
Regulated
Gaseous
Pollutants..................................................
30
Vehicles
................................................................................................................................
30
FTP
emission
rates
of
regulated
pollutants.............................................................................
31
IM240
emission
rates
of
regulated
pollutants
.........................................................................
32
Idle
test
emission
rates
and
concentration...............................................................................
32
PM
FTP
emission
rates
.........................................................................................................
33
Comparison
of
PM
and
HC
FTP
emission
rates
.....................................................................
36
IM240
PM
emission
rates......................................................................................................
37
5.
Particle
Number
and
Size
...................................................................................................
55
Particle
number
.....................................................................................................................
55
Particle
size
distributions.......................................................................................................
58
6.
Chemical
Composition
of
the
Particulate
Matter................................................................
69
Carbon..................................................................................................................................
69
Sulfate
and
nitrate
emission
rates...........................................................................................
71
Fuel
sulfur
............................................................................................................................
77
XRF
elements
.......................................................................................................................
79
Comparison
of
outdoor
and
indoor
composition
.....................................................................
84
Polynuclear
aromatic
hydrocarbons
.......................................................................................
85
PAH
hazardous
air
pollutants................................................................................................
99
Steranes/
hopanes.................................................................................................................
100
7.
Emissions
Inventory..........................................................................................................
136
8.
Conclusions..........................................................................................................
.............
139
Acknowledgments/
References................................................................................................
140
Appendix
A.
Vehicle
Recruitment
Letters
and
Vehicle
Inspection
Form..............................
A.
1
Appendix
B.
Vehicle
Descriptions..........................................................................................
B.
1
Appendix
C.
Correction
of
EPA
Winter
Regulated
Pollutant
Data
.....................................
C.
1
Appendix
D.
Visual
Observation
of
the
Frequency
of
Smoking
Vehicles
.............................
D.
1
Appendix
E.
University
of
Denver
Remote
Sensing
Observation
of
Smoking
Vehicles
.......
E.
1
Appendix
F.
Emission
Rate
Statistics
...................................................................................
F.
1
Appendix
G.
Idle
Test
Data
..................................................................................................
G.
1
i
Executive
Summary
A
vehicle
emissions
study
was
conducted
in
the
Denver,
Colorado
area
in
July­
August
of
1996
and
in
January­
February
of
1997.
The
goals
of
the
program
were
to
measure
exhaust
particulate
matter
(
PM)
emission
rates
from
in­
use
light­
duty
gasoline
and
diesel
vehicles,
to
obtain
representative
particle
size
distributions,
to
chemically
characterize
the
PM,
and
to
compare
the
results
to
those
from
PART5,
the
current
Federal
mobile
source
PM
emission
factor
model.
The
EPA
anticipates
using
these
results
and
those
from
two
companion
CRC
studies
to
assist
in
a
planned
revision
of
the
PART5
model.
In
addition,
results
of
the
chemical
characterization
of
the
collected
PM
are
being
provided
to
the
Northern
Front
Range
Air
Quality
Study
(
NFRAQS)
for
use
in
source
apportionment.
Study
participants
were
the
Colorado
Department
of
Public
Health
and
Environment
(
CDPHE),
the
U.
S.
Environmental
Protection
Agency,
the
General
Motors
R&
D
Center,
and
Colorado
State
University.
Chemical
analyses
were
performed
under
subcontract
by
the
Desert
Research
Institute
(
DRI).
The
study
was
funded
by
the
CRC
and
the
DOE,
through
the
National
Renewable
Energy
Laboratory
and
by
the
NFRAQS.
It
also
benefited
from
in­
kind
contributions
from
all
the
participants.
Total
Petroleum
provided
free
gasoline
coupons
for
use
as
an
incentive
in
vehicle
recruitment.

Light­
duty
vehicles
were
recruited
in
six
categories:
1991­
96
gasoline,
1986­
90
gasoline,
1981­
85
gasoline,
1971­
80
gasoline,
1971
or
newer
smoking
gasoline,
and
1971
or
newer
diesels.
Vehicles
were
soaked
overnight
before
testing
on
a
chassis
dynamometer
using
the
Urban
Dynamometer
Driving
Schedule
of
the
Federal
Test
Procedure
(
FTP).
Testing
was
done
with
the
vehicles
in
as­
received
condition,
i.
e.
without
preconditioning
or
fuel
changes.
The
mandatory
oxygenated
fuel
program
was
in
effect
during
the
winter
portion
of
the
study.
The
FTP
test
was
followed,
in
sequence,
by
an
IM240
test
and
a
no­
load
idle
test.
Regulated
emissions
and
PM­
10
samples
(
particulate
matter
smaller
than
10
µ
m
in
diameter)
were
collected
for
all
three
FTP
phases
and
the
IM240.
PM­
10
samples
were
not
collected
during
the
idle
test.
During
the
summer
portion
of
study,
111
vehicles
were
tested
once.
During
the
winter
portion
of
the
study,
84
different
vehicles
were
tested
twice,
once
on
the
CDPHE
dynamometer
located
indoors
at
a
temperature
of
60
°
F,
and
once
outdoors
on
the
EPA
transportable
dynamometer
at
the
prevailing
ambient
temperature.
Winter
outdoor
testing
was
done
on
vehicles
soaked
overnight
outdoors.

Quality
assurance/
quality
control
(
QA/
QC)
procedures
included
comparisons
of
calibration
gases
for
the
two
test
sites,
testing
of
a
correlation
car
at
the
two
test
sites,
repeat
of
the
FTP
for
every
10th
vehicle,
repeat
of
every
10th
IM240,
collection
of
duplicate
filter
samples,
analysis
of
filter
samples
for
deposit
homogeneity,
and
collection
of
backup
filters
to
assess
the
magnitude
of
adsorption
of
organic
carbon
on
filters.

Average
emission
rates
for
the
regulated
emissions
and
exhaust
PM­
10
are
summarized
in
Table
1.
FTP
PM
emission
rates
for
the
1991­
97
gasoline
vehicles
were
very
low,
averaging
2.8
mg/
mi
in
the
summer
and
3.7
mg/
mi
in
the
winter
indoors.
The
PM
emission
rate
increased
with
vehicle
age
to
an
average
of
95.5
mg/
mi
in
the
summer
and
54.2
mg/
mi
in
the
winter
indoors
for
the
1971­
80
vehicles.
FTP
PM
emission
rates
for
ii
vehicles
tested
outdoors
in
the
winter
were
higher
than
those
indoors.
All
of
the
increase
occurred
during
the
cold
start
portion
of
the
test.

Table
1.
Average
FTP
Regulated
and
PM­
10
Emission
Rates
Period
Category
Number
HC
CO
NOx
PM
g/
mi
mg/
mi
Summer
1991­
96
20
0.24
2.35
0.42
2.82
Summer
1986­
90
22
1.03
14.7
1.32
44.4
Summer
1981­
85
25
2.7
35.9
1.54
47.4
Summer
1971­
80
25
6.08
62.6
2.46
95.5
Summer
Smoker
9
3.59
38.8
1.88
225
Summer
Diesel
10
ND
3.99
4.57
811
Winter­
indoor
1991­
96
10
0.56
6.74
0.51
3.51
Winter­
indoor
1986­
90
14
0.63
9.23
1.44
11.8
Winter­
indoor
1981­
85
16
2.57
35.7
2.76
35.9
Winter­
indoor
1971­
80
16
2.78
44.3
2.11
54.2
Winter­
indoor
Smoker
15
4.96
52.6
2.28
395
Winter­
indoor
Diesel
12
0.82
1.94
1.73
460
Winter­
outdoor
1991­
96
9
1.02
11.6
0.53
24.9
Winter­
outdoor
1986­
90
14
1.06
15.5
1.61
28.5
Winter­
outdoor
1981­
85
16
3.39
38.7
2.31
48.2
Winter­
outdoor
1971­
80
16
4.49
52.1
2.01
82.6
Winter­
outdoor
Smoker
15
8.71
65.6
2.29
434
Winter­
outdoor
Diesel
12
1.1
1.76
1.55
503
The
average
PM­
10
emission
rates
for
smoking
gasoline
vehicles
were
225
mg/
mi
in
the
summer
and
395
mg/
mi
in
the
indoor
winter
tests.
The
difference
is
primarily
attributable
to
better
identification
and
recruiting
of
smoking
vehicles
in
the
winter.
Diesel
vehicles
averaged
811
mg/
mi
in
the
summer
and
460
mg/
mi
in
the
indoor
winter
tests.
There
were
significant
differences
in
vehicle
age
and
class
between
the
summer
and
winter
diesel
vehicles.

FTP
HC
and
PM­
10
emission
rates
were
compared
since
other
studies
have
used
HC
emission
rates
as
a
surrogate
for
PM
emission
rates.
Correlations
between
HC
and
PM­
10
for
individual
vehicles
were
weak,
with
R2<
0.25.
The
correlation
was
much
better
(
R2=
0.89)
when
the
averages
for
each
gasoline
vehicle
category
during
the
summer
and
the
winter
were
compared.

FTP
and
IM240
emission
rates
of
HC,
CO,
and
NOx
were
well
correlated
with
R2
values
ranging
from
0.72
to
0.95.
Correlations
between
FTP
and
IM240
PM­
10
emission
rates
ranged
from
0.97
for
all
vehicles
in
the
summer
study
to
0.53
for
only
the
non­
smoking
gasoline
vehicles
in
the
summer
study.
These
results
suggest
that
the
IM240
is
effective
in
identifying
high
PM­
10
emission
rate
vehicles.

Particle
size
distributions
were
obtained
from
33
vehicles
using
impactors.
Results
are
predominantly
from
the
high
PM­
10
emitting
vehicles,
since
there
is
insufficient
mass
from
iii
low
emitting
PM­
10
vehicles
for
accurate
impactor
measurements.
The
mass
median
diameter
(
MMD)
for
non­
smoking
gasoline
vehicles
was
0.13
µ
m.
On
average,
91%
of
the
PM
mass
was
present
in
particles
2.5
µ
m
or
smaller.
Both
the
smoking
vehicles
and
the
diesel
vehicles
had
a
MMD
of
0.18
µ
m,
with
97%
of
the
mass
being
emitted
in
particles
2.5
µ
m
or
smaller.

The
number
of
particles
larger
than
0.01
µ
m
in
diameter
was
measured
continuously
during
FTP
tests
of
92
gasoline
vehicles
and
12
diesel
vehicles.
The
data
showed
that
PM
emissions
are
very
dynamic,
tending
to
occur
during
changes
in
throttle,
much
like
the
gas
phase
emissions.
This
has
implications
for
tunnel
and
roadside
studies
of
in­
use
PM
emissions.
Integrated
particle
counts
were
moderately
well
correlated
with
PM
mass
for
the
summer
study
(
R2
=
0.70),
but
were
not
as
well
correlated
for
the
winter
study
(
R2
=
0.39).

Chemical
analysis
was
performed
on
a
representative
subset
of
filter
samples.
Only
outdoor
samples
were
analyzed
for
the
winter
since
these
are
most
representative
of
actual
fleet
emissions.
"
Organic"
carbon
(
OC)
and
"
elemental"
carbon
(
EC)
were
analyzed
by
the
thermal­
optical
reflectance
method.
Soluble
sulfate
and
nitrate
were
determined
by
ion
chromatography.
X­
ray
fluorescence
(
XRF)
was
used
for
elemental
analysis.
Combined
filter
and
PUF/
XAD
cartridge
samples
were
solvent
extracted.
The
PUF/
XAD
cartridge
is
used
behind
the
particle
filter
to
collect
volatile
and
semi­
volatile
organic
compounds.
Extracts
were
analyzed
by
gas­
chromatography/
mass
spectrometry
for
polynuclear
aromatic
hydrocarbons
(
PAH)
and
for
steranes
and
hopanes.

On
average,
90%
of
the
PM
emitted
from
the
vehicles
is
carbonaceous
material,
where
carbonaceous
material
is
defined
as
the
sum
of
the
OC
(
multiplied
by
1.2
to
account
for
other
elements
such
as
hydrogen
and
oxygen)
and
the
EC.
The
OC
and
EC
fractions
of
the
carbonaceous
material
are
highly
variable
from
vehicle­
to­
vehicle.
The
OC
fraction
tends
to
increase
with
increasing
PM
emission
rate
for
gasoline
vehicles,
reaching
more
than
90%
of
the
carbonaceous
material
for
some
smoking
vehicles.
The
OC
fraction
is
also
higher
on
average
during
FTP
phase
two
than
phase
one
or
three.
Diesel
vehicles
average
approximately
50%
OC.

Sulfate
emission
rates
were
low,
with
vehicle
category
averages
for
the
gasoline
vehicles
ranging
from
0.13
to
1.7
mg/
mi.
Diesel
sulfate
emissions
were
higher,
averaging
5.6
mg/
mi
in
the
summer
and
2.7
mg/
mi
in
the
winter.
Fuel
sulfur
content
in
the
Denver
area
was
0.036
and
0.034
wt.
%,
respectively,
for
regular
grade
gasoline
and
#
2
diesel
fuel
in
the
summer
and
0.023
and
0.034
wt.
%,
respectively,
for
regular
grade
gasoline
and
#
2
diesel
fuel
in
the
winter.
A
comparison
of
the
sulfur
emitted
as
sulfate
(
measured
by
ion
chromatography)
to
the
total
sulfur
(
measured
by
XRF)
indicated
that
approximately
40%
of
the
sulfur
present
in
the
PM
was
in
the
form
of
water­
soluble
sulfate.

Forty
elements
were
measured
by
XRF,
12
of
which
were
commonly
detected.
These
were
magnesium,
aluminum,
silicon,
phosphorus,
sulfur,
chlorine,
calcium,
iron,
copper,
zinc,
bromine
and
lead.
Several
of
these
compounds
are
present
due
to
the
consumption
iv
of
engine
oil.
The
average
emission
rates
for
the
sum
of
the
elements
ranged
from
0.27
to
3.6
mg/
mi
for
the
gasoline
vehicles.
Their
maximum
percent
contribution
to
the
average
PM
emission
rate
for
the
vehicle
categories
was
nine
percent
for
the
1991­
96
gasoline
summer
vehicles.
This
elemental
data
will
be
used
as
part
of
the
vehicle
emissions
profile
for
source
apportionment.
Several
of
the
metals
are
also
of
interest
as
air
toxics.
In
this
regard,
some
of
the
measurements
showing
less­
than­
detectable
quantities
of
metals
are
as
important
as
the
measurements
on
more
abundant
species.
Finally,
the
emission
rates
of
transition
metals
are
of
interest
because
they
have
been
postulated
to
play
a
role
in
PM
health
effects.

Sixty­
five
PAH
compounds
were
measured.
The
primary
purpose
was
to
contribute
to
the
mobile
source
profile
for
source
apportionment
studies.
However,
it
is
recognized
that
PAHs
are
also
of
interest
as
hazardous
air
pollutants
(
HAPs).
Essentially
all
of
the
measured
compounds
were
emitted
by
all
vehicles
tested.
Considerable
variability
was
found
in
the
relative
emission
rates
of
the
individual
compounds
from
vehicle­
to­
vehicle,
between
FTP
phases,
and
between
some
vehicle
categories.
The
largest
difference
was
between
the
diesels
and
gasoline
vehicles,
with
the
diesels
emitting
relatively
lower
naphthalene
and
methyl­
naphthalenes
than
the
gasoline
vehicles.
Unfortunately,
these
compounds
are
present
in
the
gas
phase
and
do
not
appear
to
correlate
well
with
the
PM.
A
rigorous
analysis
of
the
variability
of
the
emission
rates
and
their
potential
usefulness
as
source­
receptor
fitting
compounds
is
beyond
the
scope
of
this
study.
Their
utility
will
be
judged
as
part
of
the
NFRAQS
study.

Hopanes
and
steranes
are
two
classes
of
hydrocarbons
found
in
motor
oil.
It
has
been
suggested
by
others
that
these
compounds
can
serve
as
tracers
for
motor
vehicle
PM
in
the
atmosphere,
with
the
possibility
that
they
can
be
used
to
distinguish
between
diesel
and
gasoline
PM.
Eighteen
hopane
and
sterane
compounds
were
measured.
Emission
rates
were
generally
very
low,
frequently
in
the
microgram
per
mile
range.
The
compounds
were
emitted
by
all
of
the
tested
vehicles,
although
some
compounds
were
frequently
below
detection
limits
for
the
lower
PM
emitting
vehicles.
There
was
little
difference
in
the
relative
emission
rates
of
these
compounds
between
diesel,
smokers,
or
gasoline
vehicles.
It
is
concluded
that
they
may
be
excellent
tracers
for
mobile
source
PM
if
they
are
not
emitted
by
other
sources,
but
they
can
not
be
used
to
distinguish
between
diesel
and
gasoline
PM
emissions.

A
simple
light­
duty
PM­
10
emissions
inventory
estimate
was
prepared
for
the
Colorado
Front
Range
area
using
the
emission
rates
determined
in
this
study
and
daily
vehicle
miles
traveled
data
supplied
by
the
State
of
Colorado.
Summer
and
winter
PM­
10
emissions
are
estimated
at
1.5
and
1.7
tons/
day,
for
the
summer
and
winter,
respectively.
Fleet
average
PM­
10
emission
rates
are
33.9
and
38.4
mg/
mi,
respectively,
for
the
summer
and
winter.
Non­
smoking
light­
duty
gasoline
vehicles
are
the
major
source
of
the
particles.
Major
uncertainties
remain
in
the
number
of
smoking
vehicles
on­
road,
and
in
the
low
temperature
PM
emission
rate
for
late­
model
gasoline
vehicles.
v
Results
of
the
comparison
of
data
obtained
in
this
program
to
PART5
model
predictions
are
not
available
at
this
time.
The
CRC
and
the
DOE
together
with
the
State
of
Texas
and
the
South
Coast
Air
Quality
Management
District
are
sponsoring
companion
studies
of
inuse
vehicle
PM
emissions
at
the
Southwest
Research
Institute
and
the
University
of
California
at
Riverside's
College
of
Engineering­
Center
for
Environmental
Research
and
Technology
(
CE­
CERT).
A
report
comparing
the
results
of
all
three
studies
will
be
prepared
after
the
final
reports
from
each
study
are
available.
vi
List
of
Abbreviations
and
Acronyms
Used
in
this
Report
A/
O
AQIRP
Auto/
Oil
Air
Quality
Improvement
Research
Program
AAMA
American
Automobile
Manufacturers
Association
AIR
Secondary
air
injection
CARB
California
Air
Resources
Board
CAWRSS
Clark
and
Washoe
Remote
Sensing
Study
CDPHE
Colorado
Department
of
Public
Health
and
Environment
CO
Carbon
monoxide
CO2
Carbon
dioxide
CRC
Coordinating
Research
Council
D
Diesel
vehicle
DOE
U.
S.
Department
of
Energy
DRI
Desert
Research
Institute
EAA
Electrical
aerosol
analyzer
EC
"
Elemental"
carbon
as
measured
by
thermal/
optical
reflectance
method
EMFAC7G
Version
7G
of
California's
mobile
source
emission
model
EMFAC
EPA
U.
S.
Environmental
Protection
Agency
FTP
Federal
Test
Procedure
G
Gasoline
vehicle
g/
mi
Grams
per
mile
GC/
MS
Gas
chromatograph/
mass
spectrometer
GM
General
Motors
H
High
particle
emitting
vehicle
HAP
Hazardous
air
pollutant
HC
Hydrocarbons
I/
M
Motor
vehicle
inspection/
maintenance
program
IARC
International
Agency
for
Research
on
Cancer
IC
Ion
chromatography
IM240
Emissions
test
for
motor
vehicles
lasting
240
seconds
L
Low
particle
emitting
vehicle
LD
Light
duty
LDD
Light­
duty
diesel
vehicle
LDG
Light­
duty
gasoline
vehicle
Lpm
Liters
per
minute
M
Medium
particle
emitting
vehicle
mg/
mi
Milligrams
per
mile
ML
Medium­
low
particle
emitting
vehicle
MMD
Mass
median
diameter
MOUDI
Micro
orifice
uniform
deposit
impactor
ND
Not
detected
NFRAQS
Northern
Front
Range
Air
Quality
Study
NIST
National
Institute
of
Standards
and
Technology
vii
NOx
Oxides
of
nitrogen
OAQPS
EPA
Office
of
Air
Quality
Planning
and
Standards
OC
"
Organic"
carbon
as
measured
by
thermal/
optical
reflectance
method
PAH
Polynuclear
aromatic
hydrocarbon
compounds
PART5
EPA's
mobile
source
model
for
particle
emissions
PM
Particulate
matter
PM­
10
Particulate
matter
having
an
aerodynamic
diameter
less
than
10
µ
m
PM­
2.5
Particulate
matter
having
an
aerodynamic
diameter
less
than
2.5
µ
m
POM
Polycyclic
organic
matter
Ppm
Parts
per
million
PpmC
Parts
per
million
carbon
PUF
Polyurethane
foam
used
for
measuring
semi­
volatile
organic
compounds
QA/
QC
Quality
assurance/
quality
control
RTP
Research
Triangle
Park,
North
Carolina
S
Smoking
vehicle
SO2
Sulfur
dioxide
Std.
Dev.
Standard
deviation
TC
Total
carbon
TWC
Three­
way
catalyst
TOR
Thermal/
optical
reflectance
method
UDDS
Urban
dynamometer
driving
schedule
VMT
Vehicle
miles
traveled
XAD
A
polymer
resin
used
for
measuring
semi­
volatile
organic
compounds
XRF
X­
ray
fluorescence
1
1.
Introduction
Atmospheric
particulate
matter
has
long
been
a
concern
because
it
reduces
visibility,
can
cause
damage
to
exposed
surface
materials,
soils
monuments
and
other
architectural
structures,
and
can
cause
health
effects.
Health
effects
concerns
have
increased
recently
due
to
epidemiological
studies
that
have
correlated
increases
in
human
morbidity
and
mortality
in
several
cities
throughout
the
country
to
PM­
2.5,
particulate
matter
with
a
diameter
of
2.5
micrometers
or
less
(
Dockery
and
Pope,
1994;
Health
Effects
Institute,
1995;
Kaiser,
1997).
In
response,
the
EPA
has
retained
the
PM­
10
standard
while
setting
new
national
ambient
air
quality
standards
for
PM­
2.5
of
65
µ
g/
m3
for
24
hr
and
15
µ
g/
m3
as
an
annual
average.
The
new
standards
will
increase
the
emphasis
on
reducing
combustion­
related
particles
since
they
are
predominately
smaller
than
2.5
µ
m.

Vehicles
contribute
to
the
atmospheric
particulate
burden
via
three
processes:
1)
direct
exhaust,
brake
wear,
and
tire
wear
emissions,
2)
reentrainment
of
dust,
and
3)
through
the
secondary
formation
of
particulate
matter
from
the
atmospheric
chemistry
of
HC,
NOx,
and
SO2
emissions.
Nationally,
all
highway
vehicles
account
for
a
very
small
proportion
of
the
total
PM­
10.
However,
in
urban
areas,
where
population
exposures
are
the
highest,
the
impact
of
all
three
processes
makes
vehicles
a
major
PM­
10
source.
Vehicle
exhaust
PM
emissions
and
tire
wear
PM
emissions
are
mostly
carbonaceous
matter.
The
state­
ofknowledge
regarding
vehicle
emissions
contributions
to
the
carbonaceous
portion
of
the
ambient
PM­
2.5
in
the
South
Coast
Air
Basin
(
Cass,
1997;
Rykowski
and
Hrebenyk,
1997)
has
recently
been
reviewed.
Diesels
were
found
to
be
the
largest
source,
followed
by
meat
cooking,
wood
smoke,
gasoline
vehicle
emissions,
and
road
dust.
The
relative
order
of
importance
of
the
sources
varies
somewhat
depending
on
the
analysis
method
used.

There
is
little
information
regarding
the
primary
particulate
emission
rates
from
current
inuse
light­
duty
vehicles.
Measurements
of
tire
and
brake
wear
emission
rates
were
last
reported
in
the
1970'
s.
A
few
measurements
of
exhaust
particulate
emission
rates
have
been
reported
recently.
Sagebiel
et
al.
(
1997)
measured
PM­
10
emission
rates
from
23
inuse
vehicles
that
were
recruited
as
high
HC
or
CO
exhaust
emitters
during
the
1994
Clark
and
Washoe
Remote
Sensing
Study
(
CAWRSS)
conducted
in
Nevada.
These
vehicles
were
emission
tested
at
roadside
on
the
IM240
test
cycle
with
the
EPA
transportable
chassis
dynamometer.
Particulate
emission
rates
were
readily
determined.
The
average
rate
for
non­
smoking
vehicles
was
0.051
grams­
per­
mile
(
g/
mi),
while
the
rate
for
those
identified
as
having
visible
smoke
was
0.558
g/
mi.
Given
the
success
of
this
study,
additional
measurements
were
made
during
1995
in
Orange
County,
CA.
(
Cadle
et
al.,
1997;
Lawson
et
al.,
1996).
Again,
vehicles
were
recruited
on
the
basis
of
having
high
remote
sensing
readings
of
CO
or
HC,
and
were
tested
at
roadside
using
the
EPA
transportable
chassis
dynamometer
and
the
IM240
test
cycle.
Eighty­
six
non­
smoking
vehicles
had
an
average
PM­
10
emission
rate
of
0.094
g/
mi,
while
17
smokers
had
an
average
emission
rate
of
0.40
g/
mi.
In
both
of
these
studies
correlation
between
PM­
10
and
HC
emissions
was
poor.
Chemical
characterization
of
the
collected
PM­
10
was
also
conducted.
2
A
third
study
(
Mulawa
et
al.,
1997)
was
conducted
in
Alaska
during
the
1994/
95
winter.
Ten
properly
functioning
light­
duty
vehicles
were
recruited
and
tested
on
the
EPA
transportable
dynamometer
using
the
FTP
cycle.
Tests
were
conducted
at
three
temperatures,
with
two
fuels.
Three
of
the
vehicles
were
retested
at
the
EPA­
RTP
facility.
A
large
fraction
of
the
Alaska
data
was
invalidated
due
to
problems
operating
at
very
low
temperatures
and
the
inexperience
of
the
local
personnel
who
operated
the
equipment.
Therefore,
most
of
the
conclusions
were
based
on
the
results
from
the
three
vehicles
tested
at
EPA.
It
was
found
that
total
PM
emissions
increase
with
decreasing
temperature,
that
most
of
the
particle
mass
was
emitted
during
the
cold­
start
portion
of
the
test,
that
there
was
a
good
correlation
between
PM­
10
emission
rates
and
the
HC
emission
rate,
and
that
use
of
an
oxygenated
fuel
reduced
PM­
10
emission
rates.

The
limited
data
available
from
FTP
tests
of
properly
functioning
vehicles
indicate
that
modern,
low
mileage
vehicles
have
low
PM
emission
rates.
The
three
aforementioned
vehicles
tested
in
the
Alaska
study
averaged
0.003
g/
mi
PM
at
room
temperature.
Twelve
1986­
1990
light­
duty
vehicles
measured
by
Mulawa
and
Dasch
(
1995)
averaged
0.007
g/
mi
total
PM.
Most
recently
the
Environmental
Research
Consortium
tested
13
low
mileage
1994­
1998
model
year
vehicles
and
6
high
mileage
1995­
1998
vehicles
on
the
FTP
using
California
phase
II
fuel
(
Chase
et
al.,
1998).
The
high
mileage
vehicles
had
average
odometer
readings
of
94,000
miles.
The
low
mileage
cars
and
trucks
had
average
total
PM
emission
rates
of
0.0006
and
0.0012
g/
mi,
respectively.
The
three
high
mileage
cars
and
3
high
mileage
trucks
had
average
total
PM
emission
rates
of
0.0008
and
0.0016
g/
mi,
respectively.

In
contrast,
Durbin
et
al.
(
1998)
studied
smoking
vehicles
in
the
South
Coast
Air
Basin.
The
study
did
an
intensive
visual
survey
of
on­
road
vehicles
and
concluded
that
between
1.1
and
1.8%
of
the
light­
duty
fleet
emits
some
visible
smoke,
and
that
these
are
composed
mainly
of
both
gasoline
and
diesel
vehicles
that
are
8
to
18
years
old.
Twentythree
smoking
vehicles
that
were
tested
on
the
FTP
had
an
average
emission
rate
of
0.399
g/
mi.

The
contribution
of
exhaust,
tire,
and
brake
primary
particulate
emissions
from
the
onroad
fleet
are
estimated
using
the
EPA
PART5
model,
or
the
CARB
EMFAC7G
model.
The
EMFAC7G
model
made
major
revisions
in
the
light­
duty
vehicle
PM
emission
estimates
compared
to
earlier
versions
of
the
model,
largely
incorporating
methodologies
and
emission
rates
used
in
PART5.
The
algorithms
used
in
these
models
are
much
simpler
than
those
used
for
exhaust
gaseous
emissions
(
Rykowski
et
al,
1996).
They
do
not
account
for
temperature,
altitude,
start
emissions
vs.
running
emissions,
fuel
composition
(
except
sulfur),
mileage,
emitter
category,
etc.
This
is
due
to
a
lack
of
data
on
the
effects
of
these
parameters.
Light­
duty
gasoline
carbonaceous
PM
emission
rates
are
provided
as
a
function
of
four
vehicle
types;
non­
catalyst
vehicles,
pre­
1981
catalyst
vehicles
with
and
without
AIR,
and
post­
1980
catalyst
vehicles.
The
light­
duty
gasoline
fleet
average
post­
1980
catalyst
vehicle
emission
rate
of
carbonaceous
material
is
0.004
g/
mi.
The
models
also
calculate
sulfate
emission
rates
for
light­
duty
gasoline
vehicles
as
a
function
of
3
catalyst,
AIR
(
secondary
air
injection),
and
fuel
sulfur
content.
The
SO2
conversion
factors
appear
to
date
from
the
late
1970'
s.
A
typical
sulfate
emission
rate
for
a
catalyst
vehicle
without
AIR
is
0.005
g/
mi,
and
with
AIR
it
is
0.016
g/
mi.
The
PART5
default
light­
duty
gasoline
fleet
average
total
PM
exhaust
emission
rates
for
1995
and
2010
are
0.014
and
0.012
g/
mi,
respectively.
For
the
draft
EMFAC7G,
the
corresponding
rates
are
0.014
and
0.010
g/
mi,
respectively.

The
accuracy
of
the
PART5
and
EMFAC7G
models
is
unknown.
The
failure
to
include
high
emitters
will
result
in
a
significant
underestimation
of
the
light­
duty
fleet
average
PM­
10
emission
rate.
For
example,
at
the
two
sites
studied
in
Orange
County,
the
recruited
light­
duty
gasoline
vehicles
had
an
average
PM
emission
rate
of
0.14
g/
mi,
and
represented
approximately
9%
of
the
fleet.
Almost
all
of
the
emissions
were
carbonaceous
material
(
Cadle
et
al.,
1997).
Thus,
these
vehicles
alone
would
contribute
0.013
g/
mi
of
carbonaceous
material
to
the
light­
duty
gasoline
fleet
average
PM
emission
rate
at
that
location.
In
addition,
the
models
do
not
separate
cold
start
emissions
from
the
total
emission
rate
and
do
not
account
for
the
effect
of
ambient
temperature.
Properly
functioning
vehicles
emit
most
of
the
exhaust
particulate
mass
during
the
cold
start,
and
the
cold
start
rate
is
a
function
of
ambient
temperature.
Recent
activity
studies
in
California
suggest
that
the
frequency
of
cold
starts
may
be
significantly
underpredicted.
Thus,
the
models
may
be
underestimating
PM
emissions
from
properly
functioning
vehicles,
especially
at
low
temperatures.
On
the
other
hand,
the
models
may
overpredict
sulfate
emissions,
since
recent
data
suggest
little
tendency
for
current
catalysts
to
produce
sulfate.
In
addition,
the
models
do
not
give
credit
for
reductions
in
emissions
from
vehicles
with
improved
emission
control
technology
in
future
years.
Thus,
future
year
projections
may
well
overestimate
PM­
10
exhaust
emissions.

Two
approaches
can
be
taken
to
improve
our
knowledge
regarding
the
impact
of
lightduty
vehicle
exhaust
PM
emissions.
The
first
is
the
bottom­
up
approach
in
which
large
numbers
of
vehicles
are
randomly
recruited
and
emission
tested
under
a
variety
of
conditions.
Given
the
highly
skewed
distribution
of
PM
emission
rates
from
in­
use
vehicles,
it
would
be
necessary
to
recruit
a
very
large
number
of
vehicles
to
obtain
an
accurate
distribution.
This
is
not
practical
within
the
scope
of
this
study.
However,
given
the
dearth
of
information
on
emission
rates
from
in­
use
light­
duty
vehicles,
the
testing
of
even
a
modest
number
of
vehicles
will
provide
a
significant
improvement
in
our
predictive
capabilities,
and
may
even
be
sufficient
to
conclude
that
the
emissions
are
not
significant
compared
to
other
sources.
The
second
approach
is
the
top­
down
source
apportionment
method.
This
methodology
utilizes
chemical
source
signatures
to
separate
the
various
sources
of
the
observed
atmospheric
PM­
10.
It
may
be
possible
to
separate
the
light
and
heavy­
duty
contributions
to
atmospheric
PM
with
this
approach.
Source
profiles
utilizing
a
combination
of
trace
species,
including
some
of
the
organic
particulate
compounds,
are
available.
However,
these
profiles
are
based
on
limited
numbers
of
samples
and
have
not
been
independently
verified.
To
some
extent,
this
top­
down
approach
suffers
from
the
same
problem
as
the
bottom­
up
approach,
in
that
source
samples
must
be
collected
from
a
large
enough
set
of
vehicles
to
ensure
that
the
source
signature
is
robust.
The
top­
down
approach
also
suffers
from
the
inability
to
distinguish
between
on­
road
and
off­
road
4
combustion
sources.
The
ideal
situation
will
be
when
both
the
top­
down
and
bottom­
up
methods
agree
on
the
relative
importance
of
the
emissions.

This
program
was
designed
to
provide
information
for
both
the
bottom­
up
and
top­
down
approaches
to
assessing
the
impact
of
exhaust
emissions
of
light­
duty
vehicles
on
PM
air
quality.
A
total
of
173
in­
use
light­
duty
gasoline
(
LDG)
and
22
light­
duty
diesel
(
LDD)
vehicles
were
tested
in
as­
received
condition
on
the
Urban
Dynamometer
Driving
Schedule
(
UDDS)
of
the
FTP
and
the
IM240
driving
cycle.
PM­
10
emission
rates
were
measured
for
each
of
the
three
phases
of
the
FTP.
A
subset
of
the
FTP
tests
included
impactor
particle
size
distributions,
continuous
particle
counts,
and
the
determination
of
the
chemical
composition
of
the
PM.
Eighty­
four
of
the
vehicles
were
tested
during
the
winter
at
two
temperatures.
Results
are
compared
to
current
emission
factor
model
predictions,
and
will
be
available
for
use
in
future
PM
emissions
model
updates.
It
should
be
noted
that
the
Coordinating
Research
Council
together
with
the
Department
of
Energy,
the
South
Coast
Air
Quality
Management
District,
and
the
State
of
Texas
is
funding
two
related
studies
that
are
making
similar
measurements
on
light­
duty
vehicles.
CE­
CERT
(
University
of
California's
College
of
Engineering­
Center
for
Environmental
Research
and
Technology)
will
test
at
least
100
LDG
and
19
LDD
vehicles
recruited
in
the
South
Coast
Air
Basin
and
the
Southwest
Research
Institute
will
test
60
LDG
and
8
LDD
vehicles
recruited
in
the
San
Antonio,
TX
area.
Altogether
a
total
of
at
least
333
LDG
and
49
LDD
vehicles
will
be
tested.

This
program
is
also
part
of
the
State
of
Colorado's
Northern
Front
Range
Air
Quality
Study
(
NFRAQS),
which
was
conducted
during
the
winter
of
1996,
the
summer
of
1996
and
the
fall­
winter
of
1996­
1997.
The
overall
goals
of
the
study
are:
1)
to
determine
the
sources
of
existing
air
pollution
in
the
Denver
urban
region,
attributing
to
each
identified
source
or
source
category
an
estimate
of
its
emissions
of
particles,
particulate
precursors,
and
other
substances;
and
2)
to
collect
data
necessary
to
support
informed
decisionmaking
for
the
future,
including
measurement
of
existing
exposure
levels
and
the
assessment
of
how
best
to
meet
state
and
visibility
goals
and
federal
air
quality
standards.
The
first
policy­
relevant
objective
of
the
study
is
to
do
a
source
apportionment
of
carbonaceous
particles
in
the
NFRAQS
study
region.
The
Desert
Research
Institute,
which
is
the
prime
NFRAQS
contractor,
will
use
the
chemical
compositions
determined
in
this
study
to
create
source
profiles
for
both
LDG
and
LDD
vehicles.
Separate
NFRAQSsponsored
studies
are
collecting
samples
of
heavy­
duty
diesel
vehicle
emissions,
woodburning
emissions,
and
meat
cooking
emissions.
Profiles
will
be
generated
from
the
chemical
analysis
results
from
each
of
these
source
types.
It
is
hoped
that
the
profiles
will
be
sufficiently
unique
to
enable
a
source
apportionment
of
the
carbonaceous
PM
in
the
NFRAQS
region.
Descriptions
of
the
profiles
and
the
results
of
the
source
apportionment
study
will
be
given
in
the
NFRAQS
final
report.
5
2.
Experimental
The
study
was
conducted
at
the
Colorado
Department
of
Public
Health
and
Environment
Aurora
Emissions
Technical
Center,
in
the
eastern
Denver
metropolitan
area.
The
summer
portion
of
the
study
was
conducted
in
July
and
August
of
1996,
while
the
winter
portion
was
conducted
in
January
and
February
of
1997.
Vehicles
were
tested
once
in
as­
received
condition
during
the
summer
portion
of
the
study.
During
the
winter,
vehicles
were
tested
twice
in
as­
received
condition,
once
at
60
°
F
and
once
at
the
prevailing
ambient
temperature.
Details
of
the
vehicle
recruitment,
inspection,
testing,
and
sampling
follow.

Vehicle
recruitment.
A
goal
of
the
study
was
to
obtain,
as
best
as
possible,
an
unbiased
set
of
vehicles
representing
the
in­
use,
on­
road,
light­
duty
vehicle
fleet
in
the
Northern
Front
Range
area
of
Colorado.
It
was
anticipated
that
the
fleet
PM
emissions
would
be
dominated
by
three
categories
of
vehicles,
smoking
gasoline
vehicles,
diesel
vehicles,
and
poorly
maintained
or
malfunctioning
vehicles.
The
latter
are
expected
to
occur
most
frequently
in
the
older
vehicle
population.
Thus,
to
ensure
that
a
reasonable
number
of
the
higher
emitting
vehicles
were
acquired
for
testing,
the
total
number
of
vehicles
to
be
tested
was
allocated
to
six
categories.
These
categories
were
1991­
1997
gasoline,
1986­
1990
gasoline,
1981­
1985
gasoline,
1971­
1980
gasoline,
1971
or
newer
smoking
gasoline,
and
1971
or
newer
model
year
diesel­
powered
vehicles.
The
1971
cutpoint
was
taken
because
current
emission
factor
models
generally
consider
emissions
from
the
last
25
years
of
vehicles.
It
is
recognized
that
the
1971­
1980
gasoline
category
includes
both
catalyst
and
non­
catalyst
equipped
vehicles.
The
1981
model
year
is
when
three­
way
catalyst
technology
was
first
introduced.
In
1986,
significant
improvements
were
made
in
vehicle
emissions
control
systems.
Smoking
gasoline
vehicles,
which
are
vehicles
with
visible
smoke
while
driven
on
the
road,
and
diesels
are
a
very
small
percentage
of
the
in­
use
fleet,
but
are
expected
to
contribute
substantially
to
the
fleet
PM
emissions.
A
random
recruitment
of
the
195
vehicles
tested
in
this
program
might
not
have
captured
any
vehicles
from
these
two
categories.
Therefore,
special
efforts
were
made
to
obtain
these
vehicles.
As
a
secondary
goal,
a
target
of
70%
light­
duty
(
LD)
cars
and
30%
LD
trucks
was
set
for
each
category.
These
percentages
of
LD
cars
and
trucks
are
representative
of
the
Denver
area
motor
vehicle
fleet.

Initial
plans
were
to
recruit
vehicles
by
having
the
State
Police
stop
on­
road
vehicles
in
the
Denver
area.
The
drivers
would
then
be
solicited
at
roadside
to
volunteer
their
vehicle
for
the
testing
program.
Unfortunately,
this
approach
was
prevented
by
last­
minute
legal
questions.
Therefore,
vehicles
were
recruited
from
customers
at
two
local
merchants,
Cub
Foods
(
a
large
discount
supermarket)
and
K­
Mart.
Recruitment
was
accomplished
by
setting
up
tables
with
large
signs
requesting
customer
participation.
Drivers
were
offered
an
incentive
of
$
100,
a
$
25
coupon
from
Total
Petroleum
redeemable
for
goods
or
fuel,
and
a
free
rental
car
during
the
time
of
testing
or
its
cash
equivalent.
This
approach
worked
well.
Because
of
insufficient
staffing,
it
was
not
possible
to
keep
records
of
the
number
of
acceptances/
rejections,
but
it
is
estimated
that
40%
of
the
drivers
who
inquired
about
the
program
accepted
the
offer.
Of
those
who
accepted
the
offer
to
participate
in
the
program
at
the
two
locations,
about
30
percent
were
rejected
on
the
spot
because
the
6
vehicles
were
not
mechanically
suitable
for
dynamometer
testing.
The
most
common
reasons
for
rejection
were
exhaust
system
leaks
and
bald
tires.
Once
the
vehicles
arrived
at
the
Aurora
Emissions
Technical
Center,
CDPHE
personnel
gave
them
a
thorough
inspection
that
resulted
in
the
rejection
of
20
percent
of
those
vehicles.
While
it
is
logical
to
assume
that
vehicles
rejected
from
the
program
due
to
a
poor
state
of
repair
may
have
higher
average
emissions
than
the
accepted
vehicles,
we
are
not
aware
of
any
study
that
has
quantified
that
effect.

The
one
difficulty
with
this
approach
was
that
suitable
numbers
of
the
oldest
vehicles
were
not
recruited
in
a
timely
manner.
It
had
been
decided
to
test
all
the
oldest
vehicles
towards
the
end
of
the
program
in
order
to
minimize
contamination
of
the
exhaust
sampling
system
by
high
PM
emitters.
Unfortunately,
it
turned
out
that
these
were
the
most
difficult
vehicles
to
recruit,
because
so
many
of
them
had
mechanical
problems
and
there
are
relatively
few
of
them
in
the
vehicle
population.
Therefore,
an
advertisement
soliciting
volunteers
for
the
program
was
run
for
one
issue
in
the
Aurora
Sentinel,
a
local
paper.
At
approximately
the
same
time,
television
news
coverage
of
the
program
was
broadcast.
These
two
events
resulted
in
a
large
number
of
calls
from
interested
volunteers.
Between
30
and
50
vehicles
were
recruited
from
these
calls.

It
was
planned
to
recruit
smoking
vehicles
through
the
City
and
County
of
Denver's
smoking
vehicle
program.
In
this
program,
trained
personnel
from
the
City
and
County
of
Denver
record
license
plate
numbers
of
in­
use
smoking
vehicles.
The
registered
owners
of
those
vehicles
are
notified
by
certified
mail
that
they
must
have
their
vehicle
repaired
within
a
specified
period
of
time
and
show
evidence
to
inspectors
that
the
necessary
repairs
have
been
made,
or
they
will
be
required
to
appear
in
court.
Owners
of
vehicles
that
had
been
identified
within
the
program
were
sent
a
letter
asking
them
to
volunteer
their
vehicle
for
testing.
This
approach
met
with
very
limited
success
(
six
recruited
vehicles)
during
the
summer
portion
of
the
study.
Therefore,
the
aforementioned
advertisement
was
also
used
to
solicit
smoking
vehicles.
In
all,
only
9
of
the
planned
20
smoking
vehicles
were
recruited
during
the
summer,
and
one
of
these
was
through
a
personal
contact.
For
the
winter
portion
of
the
study,
the
City
and
County
of
Denver
approach
was
even
less
successful.
One
vehicle
was
recruited
by
the
City
and
County
of
Denver,
five
vehicles
were
recruited
as
a
result
of
letters
sent
to
Basic
I/
M
program
emissions
inspection
stations,
and
the
remaining
nine
vehicles
were
recruited
either
by
word
of
mouth
or
from
a
list
of
vehicles
failing
the
smoke
portion
of
the
enhanced
I/
M
inspection
(
a
copy
of
the
letter
sent
to
owners
of
Basic
I/
M
stations
is
included
in
Appendix
A).

No
diesel
vehicles
were
recruited
from
motorists
shopping
at
the
local
merchants.
For
the
summer
study,
diesels
were
recruited
by
CDPHE
personnel
contacting
local
repair
shops,
used
vehicle
dealers,
and
diesel
service
businesses.
For
the
winter,
a
cash
incentive
was
provided
to
the
owners/
managers
of
diesel
inspection
stations
to
identify
and
recruit
diesel
vehicles
from
the
local
diesel
I/
M
inspection
lanes.
A
copy
of
the
letter
sent
to
the
diesel
inspection
stations
is
included
in
Appendix
A.
Most
of
the
diesels
in
the
winter
program
came
from
these
referrals,
but
one
came
from
an
owner
who
had
participated
in
the
7
summer
program
with
another
diesel
vehicle
and
heard
of
the
winter
program
through
word
of
mouth.

Table
2.1
gives
the
summer
and
winter
recruitment
targets
for
each
vehicle
category
and
the
actual
number
of
vehicles
tested.
The
number
of
vehicles
recruited
for
the
winter
was
lower
than
that
for
the
summer
since
these
vehicles
were
tested
twice.
A
description
of
every
vehicle
tested
is
given
in
Appendices
B­
1
through
B­
4.

Table
2.1.
Number
of
Vehicles
Recruited
Category
Summer
Winter
Planned
Recruited
Cars
Trucks
Planned
Recruited
Cars
Trucks
1971­
1980
25
25
18
7
15
17
11
6
1981­
1985
25
25
19
6
15
16
12
4
1986­
1990
20
22
14
8
10
14
8
6
1991­
1996
20
20
14
6
10
10
7
3
Smoking
20
9
8
1
15
15
11
4
Diesel
10
10
3
7
10
12
9
3
Vehicle
inspection.
Recruited
vehicles
were
driven
by
their
owners
to
the
Aurora
Emission
Test
Facility,
where
they
received
a
thorough
inspection.
The
inspection
accomplished
three
purposes.
First,
it
was
determined
if
the
vehicle
was
in
good
enough
condition
to
be
tested
on
the
chassis
dynamometers.
Vehicles
were
rejected
if
they
had
significant
fluid
leaks,
if
the
exhaust
system
had
leaks,
if
the
tires
were
dangerously
worn,
or
if
the
vehicle
was
judged
to
have
such
major
mechanical
problems
that
it
could
mechanically
fail
during
testing.
Second,
vehicle
information
such
as
odometer
reading,
fuel
grade,
engine,
transmission
and
emission
control
system
was
ascertained
and
recorded.
Much
of
this
information
is
presented
in
Appendix
B.
Third,
body
condition
was
recorded
in
the
event
damage
occurred
during
vehicle
testing.
Appendix
A
contains
a
copy
of
one
of
the
vehicle
inspection
sheets.

In
seven
cases
during
the
winter
portion
of
the
study,
a
vehicle
that
was
needed
to
complete
a
category,
but
had
been
rejected
during
inspection,
was
repaired
so
it
could
be
tested.
In
each
case,
a
portion
of
the
exhaust
system
was
repaired
or
replaced.
The
vehicles
were
then
prepped
on
the
dynamometer
to
ensure
that
their
exhaust
system
had
been
conditioned
and
any
forming
oil
present
on
new
parts
was
removed.
These
are
the
only
instances
where
a
vehicle
received
dynamometer
preparation.
The
seven
repaired
vehicles
had
the
following
WNFRAQS
numbers:
312,
315,
332,
370,
375,
381
and
388,
the
latter
two
being
diesels.
Repairs
were
not
expected
to
affect
the
vehicles
emission.
The
repair
costs
ranged
from
$
35
to
$
275,
with
an
average
cost
of
about
$
130.

Vehicle
testing.
Vehicles
were
tested
as
received,
including
the
use
of
the
tank
fuel.
Only
in
instances
where
insufficient
fuel
was
present
to
complete
the
testing,
was
fuel
added.
The
fuel
added
was
a
local
commercial
fuel.
Immediately
following
acceptance
for
the
study,
each
vehicle
received
an
incoming
two­
speed
idle
test
for
HC
and
CO.
This
was
8
followed
by
a
soak
period
to
bring
all
vehicle
components
to
an
equilibrium
temperature.

During
the
summer,
all
vehicles
were
soaked
indoors
overnight
at
72
°
F.
Half
the
gasoline
vehicles
were
tested
on
the
CDPHE
chassis
dynamometer,
while
the
other
half
were
tested
on
the
EPA
transportable
chassis
dynamometer.
All
diesels
in
the
summer
were
tested
on
the
CDPHE
dynamometer.
Both
dynamometers
used
8.6"
diameter
dual
rolls.
The
EPA
dynamometer
was
located
outdoors
under
a
canopy.
Vehicles
tested
outdoors
were
pushed
from
the
soak
area
onto
the
dynamometer
immediately
before
testing.

During
the
winter,
all
vehicles
were
tested
twice,
once
on
each
dynamometer.
Half
the
vehicles
were
tested
indoors
first,
and
half
outdoors
first.
The
indoor
testing
was
done
after
a
soak
indoors
at
60
°
F.
The
temperature
was
lowered
from
the
standard
FTP
test
temperature
range
to
better
reflect
the
maximum
winter
temperature
in
Denver
and
to
minimize
the
influence
evaporative
emissions
from
high
volatility
winter
fuels
might
have
on
the
exhaust
emissions.
The
outdoor
testing
was
done
after
an
overnight
outdoor
soak
at
ambient
temperatures.
The
crankcase
oil
temperature
of
the
vehicles
was
measured
with
a
thermocouple
immediately
before
testing.
It
was
anticipated
that
oil
temperature
would
be
a
better
indicator
of
the
vehicle
condition
at
the
time
of
testing
than
the
ambient
temperature.
However,
ambient
temperature
was
also
recorded.

The
vehicles
were
driven
on
the
UDDS
of
the
FTP.
This
is
the
standard
exhaust
emissions
test
which
has
been
used
for
both
certification
and
emissions
inventory
development
since
1975.
It
represents
city
driving
in
Los
Angeles.
There
are
three
phases
(
also
referred
to
as
bags)
to
the
test
that
represent
cold
start,
hot
stabilized,
and
hot
start
operation.
These
phases
are
referred
to
as
bags
1­
3.
After
completion
of
the
FTP,
the
vehicles
were
turned
off
for
a
few
minutes,
then
restarted
and
an
IM240
test
was
driven.
The
IM240
test
cycle
is
used
in
the
Denver
area's
enhanced
inspection
and
maintenance
program.
It
is
started
from
idle
with
the
vehicle
in
hot
stabilized
condition
and
lasts
240
seconds.
It
was
designed
to
mimic
the
FTP.
The
purpose
of
conducting
this
test
was
to
determine
if
it
could
be
used
as
a
short
test
to
characterize
PM
emissions
from
light­
duty
vehicles.
Following
the
IM240,
regulated
emissions
were
measured
at
idle
using
a
bagged
idle
test
or
its
equivalent.
The
test
was
conducted
at
normal
"
as
received"
curb
idle.
For
standard
transmission
vehicles
the
test
was
done
in
neutral.
For
automatic
transmission
vehicles,
the
test
was
conducted
in
drive.
This
was
done
for
intersection
modeling
purposes
by
the
CDPHE.
No
PM
emissions
were
collected
during
the
idle
test.
This
report
will
not
analyze
the
idle
emissions
data.

Gaseous
emissions.
Regulated
emissions
(
HC,
CO
and
NOx)
and
CO2
were
measured
during
all
FTP
and
IM240
tests.
Diesel
vehicles
require
the
use
of
a
heated
FID,
which
did
not
operate
properly
during
the
summer
portion
of
the
study.
Therefore,
no
HC
data
are
reported
for
the
diesels
tested
during
the
summer.
For
the
winter,
the
CDPHE
heated
FID
worked
well,
and
a
heated
FID
was
added
to
the
EPA
test
site.
The
CDPHE
measured
gas
phase
emissions
using
the
standard
bag
collection
system.
Their
analyzer
bench
was
replaced
between
the
two
studies.
The
EPA
used
modal
dilute
exhaust
measurement,
rather
than
bag
collection.
Their
analyzers
were
operated
on
fixed
ranges,
9
which
limits
their
accuracy
somewhat
at
very
high
and
very
low
concentrations.
During
the
winter,
the
dilution
air
for
the
EPA
exhaust
tunnel
was
preheated
to
150
°
F
and
the
tunnel
was
wrapped
with
insulation
to
ensure
that
water
condensation
was
not
a
problem.
In
addition,
a
heated
sampling
line
was
used
between
the
dilution
tunnel
and
the
trailer
that
housed
the
analytical
instruments
and
data
acquisition
computer.

Near
the
end
of
the
winter
study,
it
was
discovered
that
the
pump
on
the
EPA
sample
line
had
developed
a
leak
approximately
one­
third
of
the
way
through
testing.
A
small
leak
in
the
pump
was
exacerbated
by
partial
plugging
of
the
filter
protecting
the
pump.
The
filter
plugging,
which
was
probably
caused
by
the
testing
of
three
smokers
back­
to­
back,
caused
a
large
increase
in
the
vacuum
at
the
pump.
A
secondary
effect
of
this
problem
was
a
decrease
in
response
time.
The
timing
and
magnitude
of
the
leak
was
corroborated
by
examining
CO2
emission
rates
determined
for
bags
2
and
3
of
the
FTP
and
the
IM240
at
the
CDPHE
test
site
and
the
EPA
test
site
on
the
same
vehicles.
If
everything
was
operating
properly,
the
emissions
should
be
very
similar.
Bag
1
of
the
FTP
was
not
included
in
the
analysis
since
the
cold
start
at
different
temperatures
would
be
expected
to
influence
CO2
emissions.
The
slope
of
the
linear
correlation
for
CO2
was
0.95
before
the
smokers
were
run,
and
0.72
after
the
smokers
were
run.
No
trends
in
the
leak
rate
with
time
could
be
observed.
Based
on
this
analysis,
and
flow
tests
of
the
leaking
pump,
all
affected
data
from
the
EPA
test
site
were
reprocessed
to
account
for
the
change
in
response
time
(
a
3­
5%
increase)
and
a
leak
of
25%.
HC
data
for
the
diesels
were
not
affected
by
this
correction
since
the
heated
FID
was
operated
with
a
separate
pump.
Also,
PM
samples
were
collected
with
a
separate
sampling
system,
and
thus
were
not
affected
by
this
problem.
More
details
on
this
correction
are
provided
in
Appendix
C­
1.

Figure
2.1
shows
the
difference
in
FTP
composite
CO2
emission
rates
between
the
CDPHE
and
EPA
sites
after
the
corrections
were
made.
No
corrections
were
made
to
runs
0­
89
and
255­
267,
the
latter
being
after
the
pump
was
replaced.
On
average,
it
is
expected
that
the
CO2
emission
rate
will
be
higher
for
the
EPA
site
than
the
CDPHE
site,
due
to
the
temperature
difference.
Figure
2.1
indicates
that
the
differences
are
random
up
to
run
182
and
after
the
pump
replacement
before
run
249.
However,
there
is
a
consistent
bias
towards
higher
CO2
emission
rates
for
runs
183­
248.
These
can
not
be
explained
by
the
differences
in
ambient
temperature.
Since
the
CO2
rates
are
high,
but
within
the
variability
of
the
other
measurements,
possible
explanations
include
a
decrease
in
the
leak
rate,
an
independent
measurement
problem
with
the
CO2,
or
that
the
data
are
correct.
The
differences
in
the
HC,
CO
and
NOx
emission
rates
for
the
same
tests
were
examined
for
each
FTP
phase
to
determine
if
they
showed
the
same
pattern.
Definite
conclusions
could
not
be
drawn
based
on
non­
parametric
Wilcoxon
tests,
regressions,
or
t­
tests,
in
part
because
variability
is
generally
higher
for
these
species
than
for
CO2.
Therefore,
it
was
not
possible
to
determine
if
there
was
an
independent
measurement
problem
with
the
CO2.
If
however,
all
gaseous
emission
rates
for
runs
183­
248
(
except
diesel
HC)
are
further
adjusted
by
a
16%
increase,
the
ratios
of
positive
to
negative
differences
between
CDPHE
and
EPA
measurements
become
more
consistent
over
all
tests
for
HC,
CO
and
NOx
as
well
as
for
CO2.
We
can
not
therefore
discount
the
possibility
that
there
are
errors
on
the
order
of
16%
with
the
gaseous
emissions
rates
for
runs
183­
248,
excluding
the
diesel
HC
10
measurements.
Such
errors
have
no
impact
on
the
conclusions
of
this
report.
A
more
detailed
examination
of
the
differences
is
presented
in
Appendix
C­
2.

PM
emissions.
PM
samples
were
collected
through
two
isokinetic
probes
inserted
into
each
of
the
dilution
tunnels.
These
probes
were
connected
to
University
Research
Glassware
PM­
10
cyclones,
which
operated
at
a
flow
rate
of
28.3
lpm.
Flow
rates
were
held
constant
with
mass
flow
controllers
and
were
checked
periodically
with
a
dry
gas
meter.
Flow
controller
readouts
were
logged
on
data
acquisition
systems,
with
the
exception
of
the
EPA
site
in
the
summer,
which
relied
on
visual
checks
of
the
readout.
Each
cyclone
was
attached
to
a
26.5­
cm
long
straight
tube
for
flow
straightening.
The
tube,
in
turn,
was
connected
to
a
Y
fitting
which
accommodated
two
filter
holders.
Solenoid
valves
activated
by
the
test
computer
systems
were
used
to
switch
between
filters.
Switching
occurred
at
the
end
of
phase
1
of
the
FTP
to
the
phase
2
filter
and
at
the
end
of
phase
3
to
the
IM240
filter
(
filter
holders
were
changed
during
the
soak
period
between
phases
2
and
3
of
the
FTP).

PM­
10
samples
were
collected
simultaneously
on
both
37­
mm
diameter,
2.0­
µ
m
pore
size
Gelman
Teflo
and
37­
mm
diameter
Pallflex
Tissue
Quartz
2500
QAT­
UP
filters.
Quartz
filters
were
prefired
at
900
°
C
for
three
hours
to
remove
any
carbon.
Filter
plugging
problems
were
encountered
during
the
summer
for
a
few
high­
emitting
vehicles.
Therefore,
the
37­
mm
Teflo
filters
were
replaced
with
either
47­
mm
Teflo
filters
or
37­
mm
or
47­
mm
T60A20
Pallflex
filters
for
the
high­
emitting
gasoline
vehicles
and
the
diesels
during
the
winter.
A
subset
of
tests
was
run
with
a
vapor­
phase
trap
for
polynuclear
aromatic
hydrocarbons
(
PAH)
placed
behind
the
quartz
filter.
The
trap
consisted
of
5
g
of
XAD­
4
resin
(
polystyrene/
divinylbenzene
polymer)
sandwiched
between
two
polyurethane
foam
(
PUF)
plugs.
In
addition,
some
tests
were
performed
with
back­
to­
back
quartz
filters
and
back­
to­
back
Teflo/
quartz
filter
pairs
to
test
for
organic
carbon
adsorption.
Both
media
blanks
and
field
blanks
of
all
filters
and
PUF/
XAD
lots
were
retained
and
analyzed.

All
Teflon
and
Pallflex
filters
were
weighed
on
a
microbalance
after
equilibration
at
constant
temperature
and
humidity.
Filters
were
stored
in
plastic
Petri
dishes
in
a
refrigerator,
and
transported
between
laboratories
using
a
cooler
with
blue
ice
packs.
PUF/
XAD
samples
were
stored
in
glass
jars
with
a
sheet
of
aluminum
foil
placed
between
the
opening
and
the
lid.
These
samples
were
also
stored
and
transported
cold.

Particle
size
distributions
were
determined
with
MSP
micro
orifice
uniform
deposit
impactors
(
MOUDI).
The
impactors
were
run
on
the
CDPHE
site
only,
since
the
physical
arrangement
of
the
EPA
dilution
tunnel
made
collection
of
impactor
samples
very
difficult.
The
summer
study
used
an
eight­
stage
impactor
with
a
backup
filter.
The
winter
study
used
a
ten­
stage
impactor
with
a
backup
filter.
The
impactors
were
operated
at
an
actual
flow
rate
of
30.3
lpm.
The
particle
size
ranges
for
the
individual
stages
were
corrected
for
the
reduced
pressure
at
Denver's
altitude.
Uncoated
aluminum
foils
were
used
as
collection
media.
Particle
bounce
can
be
a
problem
with
impactor
samples
using
uncoated
collection
substrates
(
Lawson,
1980).
We
are
not
aware
of
any
studies
of
this
potential
11
problem
for
exhaust
PM
with
MOUDI
impactors.
Foils
were
stored
and
weighed
in
the
same
manner
as
the
filter
samples.
No
chemical
analysis
was
done
on
these
samples.

Continuous
particle
counts
were
made
during
most
emission
tests
performed
on
the
CDPHE
site
with
a
TSI
Electrical
Aerosol
Analyzer
(
EAA).
The
analyzer
was
operated
continuously
at
a
setting
that
measures
all
particles
greater
than
0.01
µ
m
aerodynamic
diameter.
This
analyzer,
which
was
designed
for
atmospheric
studies,
would
be
saturated
by
the
high
particle
concentrations
present
at
times
in
most
of
the
mobile
source
tests.
Therefore
it
was
necessary
to
further
dilute
the
sample
drawn
from
the
dilution
tunnel.
This
was
accomplished
by
using
one
or
more
diluters
in
series.
The
diluters
were
cartridge
filters
with
different
size
"
pinholes".
The
dilution
ratio
of
each
diluter
was
determined
by
placing
it
in
the
output
stream
of
an
aerosol
generator
and
using
the
EAA
to
determine
the
change
in
particle
number.
Exhaust
measurements
were
made
at
dilution
ratios
of
22
to
785
above
that
already
generated
by
the
dilution
tunnel.

Particle
counts
made
with
the
EAA
in
this
manner
are
complicated
by
the
fact
that
the
EAA
response
is
not
identical
for
all
particle
sizes.
Under
normal
operation,
the
analyzer
is
scanned
through
its
size
range,
and
the
applicable
response
factors
are
used
to
determine
the
total
number
of
particles.
When
operated
at
a
single
size
setting,
as
in
this
study,
an
integrated
response
factor
must
be
used
that
reflects
the
size
distribution
of
the
particles
being
measured.
For
calibration
purposes,
the
EAA
was
used
to
measure
particle
size
distributions
from
two
vehicles
operated
under
cruise
conditions.
It
is
not
possible
to
obtain
size
distributions
during
cyclic
operation
since
the
scanning
time
is
approximately
two
minutes.
The
response
factor
was
determined
by
comparing
the
"
step
3"
voltage
(
0.01
µ
m
setting)
to
the
total
particle
number
calculated
using
TSI
Model
390039
EAA
software
(
Twomey/
Kapadia
algorithm).
An
integrated
response
factor
of
1.25
x
105
particles/
cc/
volt
determined
from
these
tests
was
applied
to
all
of
the
emission
tests.
It
must
be
recognized
that
if
the
size
distribution
changes,
the
response
factor
can
also
change.
Therefore,
the
particle
numbers
should
be
considered
valid
to
no
better
than
a
factor
of
two.
This
high
uncertainty
has
no
effect
on
other
measurements
in
the
program
since
the
particle
number
data
is
not
used
in
any
calculations.

Sample
analysis.
A
representative
subset
of
the
samples
was
selected
for
chemical
analysis.
All
analyses
discussed
in
this
report
were
performed
at
DRI.
Samples
were
also
provided
to
NIST
for
the
determination
of
14C
content.
These
analyses
were
performed
as
part
of
the
CRC/
NFRAQS­
funded
Project
A­
20,
"
Carbon
Isotopic
Analysis
of
the
Northern
Front
Range
Air
Quality
Study's
Summer
and
Winter
1996­
1997
Program."
Results
will
be
given
in
the
final
report
for
that
project.

Sulfate
and
nitrate
were
determined
by
ion
chromatography
(
IC).
Teflo
or
Pallflex
filters
were
extracted
in
15
ml
of
40:
60
water/
isopropyl
alcohol
in
an
ultrasonic
bath
for
60
min.,
then
in
a
mechanical
shaker
for
60
min.
This
is
the
standard
extraction
solution
for
mobile
source
samples.
It
has
been
shown
to
be
more
efficient
in
extracting
sulfate
from
mobile
source
samples
than
water
alone
because
the
alcohol
acts
as
a
wetting
agent
to
help
penetrate
the
hydrophobic
samples.
The
ultrasonic
bath
temperature
was
maintained
12
below
27
°
C.
Ammonium
nitrate
is
easily
lost
from
filters
during
sampling
as
nitric
acid.
Thus,
nitrate
values
may
be
low.
A
more
accurate
sampling
procedure
would
utilize
a
nitric
acid
denuder
and
a
nylon
backup
filter.
Such
methods
have
been
used
in
the
past,
and
very
little
nitrate
or
nitric
acid
has
been
found.
Since
nitrate
and
nitric
acid
have
never
been
found
to
be
emitted
in
significant
quantities
from
automobiles,
it
was
not
the
focus
of
this
study.

"
Organic"
carbon
(
OC)
and
"
elemental"
carbon
(
EC)
were
determined
on
0.512
cm2
punches
removed
from
the
quartz
filters
using
the
DRI
thermal
optical
reflectance
(
TOR)
technique
(
Chow,
1993).
With
this
method,
OC
is
defined
as
carbon
removed
by
pyrolysis
at
550
°
C
in
a
helium
atmosphere,
while
EC
is
the
carbon
removed
after
pyrolysis
by
the
addition
of
2%
oxygen.
OC
and
EC
results
are
corrected
for
charring
by
the
optical
monitoring
of
the
reflectance.
Teflo
and
Pallflex
filters
were
analyzed
for
38
elements
by
energy
dispersive
x­
ray
fluorescence
(
XRF)
analysis
(
Chow
and
Watson,
1994).
PAHs
are
determined
from
the
quartz
filter/
PUF/
XAD
samples.
The
PUF
plugs
are
Soxhlet
extracted
with
acetone,
while
the
filters
and
XAD
resin
are
Soxhlet
extracted
with
dichloromethane.
The
extracts
were
combined
and
reduced
to
1.0
ml
by
rotary
evaporation
and
analyzed
for
approximately
70
PAHs
by
GC/
MS
in
selected
ion
monitoring
mode.
Samples
have
deuterated
internal
standards
added
before
extraction
to
measure
the
recovery
efficiency
of
the
process.
Hopanes
and
steranes
are
analyzed
from
these
extracts
by
GC/
MS
as
well.

The
primary
purpose
of
measuring
the
PAHs
was
to
create
source
profiles
that
can
be
used
to
determine
the
relative
contributions
of
LDG
and
LDD
exhaust
PM
emissions
to
the
atmospheric
particle
burden.
Some
of
the
PAHs
have
also
been
designated
by
the
EPA
as
air
toxics.
Samples
for
PAH
analysis
require
extensive
preparation,
making
the
overall
sample
analysis
cost
high.
To
limit
the
number
of
PAH
analyses,
FTP
runs
from
multiple
vehicles
were
pooled
on
a
bag
basis.
Since
pooling
of
samples
with
large
differences
in
emission
rate
would
only
result
in
a
composite
sample
dominated
by
the
high
emitter,
pooling
was
done
on
the
basis
of
emission
rate,
rather
than
by
vehicle
age.
Insufficient
PUF/
XAD
samples
were
collected
to
permit
pooling
by
both
vehicle
age
and
emission
rate.
For
the
summer,
a
total
of
29
FTP
filter
sets
were
pooled
into
15
3­
bag
samples.
These
included
four
diesels
pooled
into
three
samples;
six
smokers
into
three
samples;
seven
low
emitters
into
three
samples,
six
medium
emitters
into
three
samples,
and
six
high
emitters
into
three
samples.
Filter
samples
from
all
of
the
FTPs
that
were
pooled
for
PAH
analysis
were
also
analyzed
individually
by
IC,
XRF,
and
TOR.
This
permitted
the
determination
of
the
emission
rates
of
the
sulfate,
nitrate,
trace
elements,
OC
and
EC
on
a
vehicle
model
year
category
basis,
while
still
allowing
the
calculation
of
the
pooled
sample
emission
rates
for
source
profiles.

For
the
winter,
42
FTP
filter
sets
were
pooled
into
17
3­
bag
samples.
Eleven
diesels
were
pooled
into
five
samples,
eight
smokers
into
three
samples,
six
low­
emitters
into
two
samples,
five
mid­
low
emitters
into
two
samples,
eight
medium
emitters
into
three
samples,
and
four
high
emitters
into
two
samples.
As
before,
filter
samples
from
all
of
the
FTPs
that
were
pooled
for
PAH
analysis
were
also
analyzed
individually
by
IC,
XRF,
and
13
TOR.
All
of
the
winter
PUF/
XAD
samples
were
collected
from
vehicles
operated
outdoors
on
the
EPA
dynamometer,
since
that
was
the
most
realistic
condition
for
winter
vehicle
profile
development.
To
determine
if
there
is
a
difference
in
the
composition
of
PM
emitted
at
the
indoor
temperature
as
opposed
to
that
outdoors,
18
indoor
bag
1
samples
were
analyzed
by
IC
and
XRF.
Samples
from
bags
2
and
3
of
the
FTP
were
not
analyzed
since
it
was
unlikely
that
there
would
be
significant
differences
in
composition
during
hot
stabilized
or
hot
start
operation.

Fuel.
Vehicles
were
tested
using
the
as­
received,
on­
board
fuel.
Vehicle
fuel
samples
were
not
collected.
However,
fuel
sulfur
content
is
of
interest
since
it
is
used
by
the
PM
emission
factor
models
to
estimate
sulfate
emissions.
Sixteen
fuel
samples
were
collected
from
service
stations
during
the
summer
and
20
samples
were
collected
in
the
winter.
The
samples
were
sent
to
Core
Laboratories
(
Houston,
Texas)
for
analysis
of
fuel
sulfur.
The
AAMA
also
collects
and
analyzes
fuel
samples
during
the
summer
and
winter
in
Denver.
The
gasoline
samples
collected
in
this
study
put
more
emphasis
on
mid­
and
premium
grade
gasolines
than
the
AAMA
program,
so
results
would
be
complementary.
Fuels
tested
are
presented
later,
in
Table
6.6.
Average
sulfur
content
is
also
given
later,
in
Table
6.7.
The
AAMA
samples
(
AAMA,
1996/
7)
are
analyzed
for
the
complete
set
of
standard
fuel
properties,
including
oxygenates.
Readers
interested
in
the
fuel
composition
are
referred
to
the
AAMA
report.

Smoking
vehicle
population.
This
test
program
specifically
recruited
smoking
vehicles
because
of
their
obvious
high
PM
emission
rates.
For
the
emission
data
to
be
useful,
the
in­
use
frequency
of
smoking
gasoline
vehicles
in
the
on­
road
population
must
be
estimated.
Two
programs
were
run
to
provide
an
estimate
of
this
population.
The
first
was
a
visual
roadside
survey
of
smoking
vehicles
in
the
Denver
area.
This
effort,
which
is
described
in
detail
in
Appendix
D,
found
that
16
of
11,899
vehicles
emitted
visible
smoke,
a
percentage
of
0.13.
The
second
approach
was
an
analysis
of
the
frequency
of
smoking
vehicles
observed
by
remote
sensing
at
the
westbound
6th
Avenue
off­
ramp
from
northbound
I­
25
site
in
Denver
conducted
by
the
University
of
Denver.
Remote
sensing
has
not
been
optimized
for
opacity
measurements,
and
has
not
been
calibrated
or
undergone
independent
validation
studies.
Thus,
interpretation
of
the
results
is
uncertain.
It
was
found
that
0.81
±
0.24%
of
the
gasoline
vehicles
had
significantly
increased
exhaust
plume
opacity.
14
15
3.
Quality
Assurance/
Quality
Control
Several
quality
assurance
procedures
were
used
to
validate
the
testing
and
sampling
results.
First,
dynamometer
coastdowns
(
55
to
45
miles
per
hour)
were
performed
on
both
dynamometers
using
the
same
vehicle.
The
results
were
13.38
sec
for
the
CDPHE
site
and
13.58
sec
for
the
EPA
site,
indicating
that
both
dynamometers
were
providing
similar
loads
to
the
vehicles.
Next,
cylinder
calibration
gases
were
exchanged.
Results
are
given
in
Table
3.1.

Table
3.1.
Cylinder
Gas
Standards
Comparison
Gas
CDPHE
EPA
Scott1
Propane,
ppmC
27.6
27.9
27.6
Propane,
ppmC
165
167
165
CO,
ppm
97.1
115
99.2
CO,
ppm
971
993
963
NOx,
ppm
69.8
72.0
70.6
NOx,
ppm
88.0
89.3
90.2
CO2,
%
2.62
2.61
2.68
CO2,
%
3.50
3.53
3.56
1)
Scott
was
the
cylinder
gas
supplier
Agreement
was
within
three
percent
for
all
samples
except
the
low
CO
concentration,
which
differed
by
18%.
Since
EPA
uses
modal
analysis
without
range
changing,
their
analyzer
must
cover
a
very
broad
range
of
CO
concentrations,
with
the
low
end
being
the
least
important
for
integrated
values.
Thus,
this
low
CO
point
was
at
the
low
end
of
their
calibration
and
consequently
suffered
in
accuracy.
The
CDPHE
measures
emissions
in
bags,
and
must
have
accurate
low
CO
readings.
Overall,
the
agreement
was
judged
to
be
excellent.

Site
correlations.
During
the
summer,
results
from
the
two
sites
could
be
directly
compared
by
running
the
same
vehicle.
This
was
not
possible
during
the
winter
due
to
the
temperature
differences.
A
1991,
3.1L
Corsica
was
chosen
as
the
correlation
vehicle
for
the
summer
study.
It
was
operated
on
each
site
once
a
week.
Results
are
given
in
Table
3.2.

The
high
PM
result
on
the
first
EPA
test
was
discounted
since
this
was
the
first
run
after
the
equipment
was
setup,
and
the
tunnel
had
not
been
conditioned.
Considerable
variability
between
emission
tests
is
routinely
observed.
These
can
be
caused
by
changes
in
the
vehicle's
operation
and
in
changes
in
how
the
vehicle
is
driven.
Few
data
are
available
about
the
reproducibility
of
PM
emissions
from
light­
duty
vehicles.
For
the
regulated
gaseous
species,
the
CO
emission
rate
is
the
most
susceptible
to
variations
in
driving,
with
some
vehicles
experiencing
changes
of
a
factor
of
two
in
emission
rate
with
different
drivers,
even
though
the
tests
are
driven
within
the
allowed
test
cycle
speed
16
deviation.
The
last
test
on
the
EPA
site
was
driven
by
a
different
driver
than
the
first
two
tests.
For
comparison,
the
Auto/
Oil
Air
Quality
Improvement
Research
Program
(
A/
O
AQIRP)
used
the
ratios
of
emission
rates
from
two
tests
to
determine
if
a
third
test
should
be
run.
Acceptance
ratios
were
1.33,
1.70,
and
1.29
for
HC,
CO
and
NOx,
respectively.
The
HC
and
CO
emission
rates
fall
outside
these
criteria.
This
could
still
be
due
to
driver
differences
rather
than
test
site
differences.

Table
3.2.
Correlation
Car
FTP
Results
Site
Date
PM,
mg/
mi
HC,
g/
mi
CO,
g/
mi
NOx,
g/
mi
CO2,
g/
mi
EPA1
7/
30/
96
9.2
0.47
3.7
0.56
432
EPA
8/
5/
96
2.6
0.54
4.2
0.71
438
EPA2
8/
12/
96
2.7
0.34
1.7
0.51
419
CDPHE
8/
6/
96
0.9
0.21
1.4
0.56
396
CDPHE
8/
13/
96
0.9
0.23
2.0
0.53
397
1)
First
test
run
after
equipment
was
set
up
2)
EPA
driver
change
Emission
results
for
the
two
test
sites
can
be
compared
for
all
vehicles
during
the
winter.
The
primary
difference
between
the
sites
was
the
test
temperature,
the
effect
of
which
should
be
greatest
on
bag
1
emissions.
Therefore,
for
QA
purposes,
the
emission
rates
are
compared
for
bags
2
and
3
only.
Table
3.3
gives
the
average
PM,
HC,
CO,
and
NOx
emission
rates
for
all
non­
smoking
vehicles,
the
smoking
vehicles,
and
the
diesels.
Averages
are
for
paired
data
only.
Results
for
CO2
are
not
given
since
the
EPA
data
used
the
CO2
comparison
between
the
two
sites
to
correct
for
a
leak
that
occurred
during
some
of
the
testing
(
see
Appendix
C).
Considering
the
differences
in
dynamometers,
drivers,
emission
measurement
method
(
bag
versus
modal),
temperature,
and
vehicle
conditioning
between
subsequent
tests,
the
agreement
in
average
values
is
very
good.
Figures
3.1
and
3.2
show
comparisons
in
HC
and
CO
bag
2
emission
rates
for
the
gasoline
vehicles.
While
there
is
considerable
scatter
in
the
data,
the
regressions
and
regression
coefficients
indicate
good
overall
agreement.
The
PM
data
have
much
more
scatter,
with
the
result
that
the
regressions
were
very
poor
for
the
gasoline
vehicles,
R2
of
0.23
and
0.08
for
bags
2
and
3.
PM
correlations
were
better
for
smokers
and
diesels,
which
when
compared
together,
had
R2
values
of
0.74
and
0.68
for
bags
2
and
3,
respectively.

Results
from
IM240
tests
conducted
at
both
sites
were
examined
with
all
vehicles
pooled
together.
The
average
masses
emitted
at
the
CDPHE
and
EPA
sites
were
135
and
148
mg/
mi,
respectively.
The
R2
value
for
a
linear
correlation
was
only
0.40,
although
this
correlation
was
greatly
influenced
by
one
point.
17
Table
3.3.
Average
FTP
Bag
2
and
3
Winter
Emission
Rates
Vehicles
Site
Bag
2
Bag
3
PM
mg/
mi
HC
CO
NOx
PM
mg/
mi
HC
CO
NOx
­
g/
mi­
­
g/
mi­

Gasoline
CDPHE
16.2
1.35
19.2
1.52
23.3
1.52
22.7
2.09
Gasoline
EPA
19.0
1.58
17.2
1.63
21.7
1.47
17.0
1.79
Smokers
CDPHE
298
4.37
44.6
1.96
315
4.28
44.0
2.68
Smokers
EPA
233
5.56
47.0
2.19
268
4.62
41.5
2.53
Diesels
CDPHE
380
0.80
1.86
1.87
495
0.64
1.68
1.56
Diesels
EPA
370
1.09
1.73
1.63
425
0.79
1.50
1.29
Emissions
variability.
Vehicle
emissions
variability
can
be
examined
by
comparing
the
results
of
the
replicate
FTPs
that
were
run
on
every
10th
vehicle.
The
first
test
on
each
vehicle
was
performed
in
the
as­
received
condition,
whereas
the
second
test
had
the
first
FTP
as
a
preconditioning
step.
Thus,
variability
between
tests
is
expected
to
be
greater
than
that
for
normal
vehicle
testing
which
utilizes
preconditioning
to
improve
reproducibility.
Variability
is
also
expected
to
be
higher
for
vehicles
in
a
poor
state
of
maintenance
than
for
well­
maintained,
properly
functioning
vehicles.
Table
3.4
gives
the
emission
rates
for
replicate
FTP
tests.
The
average
of
the
ratios
of
the
first
to
second
test
and
the
standard
deviation
of
the
ratios
are
given
in
the
last
two
rows
of
the
table.
The
average
agreement
is
excellent,
although
it
is
clear
that
large
differences
can
occur
for
individual
tests.
The
standard
deviation
is
by
far
the
highest
for
the
PM
measurement,
80%
vs.
the
next
highest
standard
deviation
of
29%
for
CO.

Replicate
IM240
tests
were
run
back­
to­
back,
rather
than
on
the
successive
days
required
by
the
FTP
test
because
of
the
cold
soak
period.
Since
an
FTP
was
run
immediately
before
the
first
IM240,
some
preconditioning
had
been
performed.
Emission
rates
for
the
summer
IM240
replicates
are
given
in
Table
3.5.
The
average
ratio
of
the
first
to
second
test
and
the
standard
deviation
of
the
ratios
are
given
in
the
last
two
rows
of
the
table.
While
the
average
ratio
is
highest
for
PM,
this
is
influenced
greatly
by
the
last
test.
Deleting
that
test
results
in
an
average
ratio
of
0.92
with
a
standard
deviation
of
0.34.
Table
3.6
gives
the
results
from
the
back­
to­
back
IM240
tests
for
the
winter.
The
average
ratios
and
standard
deviations
have
outliers
removed
(
division
by
0,
and
one
test
pair
with
a
NOx
ratio
of
12).

Overall,
PM
emissions
variability
is
higher
for
the
FTPs
than
the
IM240s.
As
noted
above,
two
factors
may
explain
this
behavior.
First
is
the
lack
of
any
vehicle
preconditioning
before
the
first
FTP.
In
contrast,
the
first
IM240
has
an
FTP
as
a
preconditioning
step.
Second,
the
IM240s
are
short
tests
run
back­
to­
back,
whereas
the
FTPs
are
much
longer
tests
and
are
run
on
subsequent
days.
Thus,
the
FTP
is
more
likely
to
experience
the
vagaries
of
vehicle
performance.
We
are
not
aware
of
any
other
studies
of
in­
use
light
18
duty
exhaust
PM
emissions
that
can
be
used
to
compare
to
the
variability
found
in
this
study.

Table
3.4.
FTP
Emission
Rates
From
Replicate
Tests
Vehicle1
Run
Period
Test
Temp,
°
F
PM,
mg/
mi
HC,
g/
mi
CO,
g/
mi
NOx,
g/
mi
CO2,
g/
mi
32
4472
summer
72
18
0.62
4.67
1.12
309
32
4485
summer
72
5.6
0.52
3.32
1.21
306
116
4580
summer
72
15
1.98
60.3
1.47
451
116
4586
summer
72
14
1.95
61.7
1.42
461
56
4520
summer
72
5.9
1.84
20.6
1.10
385
56
4533
summer
72
6.0
1.72
21.2
1.02
380
65
96074
summer
72
43
4.00
21.9
1.79
279
65
96090
summer
72
33
4.33
23.8
1.91
300
118
96152
summer
72
31
4.20
48.3
2.59
376
118
96158
summer
72
62
4.63
25.2
3.33
396
31
96037
summer
72
142
6.56
86.7
2.98
467
31
96050
summer
72
147
6.33
86.6
2.72
460
302
4680
winter
60
2.0
0.32
6.22
0.90
532
302
4693
winter
60
4.5
0.28
6.00
0.96
543
336
4752
winter
60
1.6
0.52
13.0
0.56
413
336
4761
winter
60
2.2
0.40
9.01
0.63
418
358
4816
winter
60
174
1.29
15.3
1.09
395
358
4847
winter
60
161
1.55
16.3
1.15
396
381
4888
winter
60
101
0.43
1.43
0.84
334
381
4891
winter
60
56.7
0.49
1.33
0.84
335
305
97035
winter
37
6.0
0.34
7.78
1.30
443
305
97038
winter
32
16.6
0.32
7.21
1.18
431
335
97099
winter
55
44.0
2.42
22.3
2.42
466
335
97106
winter
55
37.2
2.27
25.2
2.35
427
356
97183
winter
36
20.2
1.42
15.7
1.46
311
356
97192
winter
34
22.3
1.69
15.6
1.38
273
380
97233
winter
30
371
0.54
1.25
0.81
257
380
97236
winter
31
293
0.54
1.35
0.82
259
Ave.
Ratio
of
1st
to
2nd
Test
1.15
1.02
1.11
0.99
1.01
Std.
Dev.
of
Ratios
0.80
0.14
0.29
0.09
0.05
1)
Runs
for
the
same
vehicle
listed
in
order
of
test
date,
starting
with
the
first
test
19
Table
3.5.
Results
from
Summer
Back­
to­
Back
IM240
Emission
Tests
Vehicle
Run
PM,
mg/
mi
HC,
g/
mi
CO,
g/
mi
NOx,
g/
mi
CO2
g/
mi
6
4445
26.34
0.47
3.16
0.51
316
6
4446
25.42
0.39
3.27
0.51
315
4
4476
0.19
0.64
5.46
2.09
280
4
4477
1.10
0.43
7.84
1.48
286
63
4507
61.64
4.22
18.58
2.17
563
63
4508
50.33
3.30
19.98
2.09
563
88
4539
58.65
4.11
88.64
3.85
527
88
4540
50.44
3.79
93.35
3.45
502
119
4571
36.96
1.30
33.32
2.80
831
119
4572
52.36
1.21
33.14
2.77
840
2
96004
6.83
3.44
87.76
0.63
284
2
96005
7.25
2.22
91.00
0.62
271
25
96041
6.90
0.59
6.70
1.43
440
25
96042
6.60
0.42
3.70
1.40
485
65
96075
21.84
0.87
26.89
1.53
244
65
96076
31.13
0.97
39.90
1.20
240
71
96106
121.53
13.14
107.43
4.07
415
71
96107
132.56
14.10
106.13
3.70
409
101
96140
3.48
1.07
4.31
1.61
309
101
96141
4.41
1.13
5.54
1.51
292
126
96171
55.83
0.30
18.76
1.45
670
126
96172
37.73
0.46
28.40
1.35
650
152
96202
410.33
1.70
5.40
2.57
305
152
96203
88.77
1.00
4.53
1.86
284
Ave.
Ratio
1.23
1.18
0.97
1.12
1.01
Std.
Dev.
1.12
0.31
0.31
0.15
0.04
20
Table
3.6.
Results
from
Winter
Back­
to­
Back
IM240
Emission
Tests
Vehicle
Run
PM,
mg/
mi
HC,
g/
mi
CO,
g/
mi
NOx,
g/
mi
CO2,
g/
mi
310
4684
6.73
0.32
9.38
0.64
260.1
310
4685
3.83
0.23
5.42
0.58
271.6
317
4715
2.71
0.36
3.70
16.55
293.7
317
4716
2.69
0.29
5.34
1.38
293.0
332
4749
84.1
2.15
24.27
4.17
237.1
332
4750
97.6
2.17
26.93
4.03
240.5
349
4779
0.00
0.20
1.71
1.00
311.2
349
4780
0.00
0.08
0.90
0.77
313.3
337
4810
425
6.31
120.74
1.94
418.7
337
4811
468
6.97
126.31
1.88
415.5
362
4841
21.2
2.88
48.67
2.28
361.6
362
4842
14.7
2.96
39.70
2.53
362.0
377
4864
2.45
ND
ND
ND
ND
377
4865
3.14
0.05
0.44
0.68
361.1
388
4904
679
0.30
2.05
2.62
445.3
388
4905
667
0.29
2.03
2.60
440.3
309
97032
2.48
0.11
0.96
1.44
355.7
309
97033
7.09
0.17
1.19
1.59
351.7
320
97069
24.5
5.46
26.58
2.70
441.1
320
97070
22.1
5.36
30.14
2.67
429.6
335
97100
37.6
1.99
23.77
2.75
439.5
335
97101
35.3
1.30
13.88
2.54
424.9
340
97131
1.83
0.00
0.84
0.42
242.0
340
97132
6.44
0.13
1.92
0.48
253.4
362
97162
9.55
1.95
42.96
1.40
294.8
362
97163
14.4
2.05
45.74
1.42
303.2
376
97196
1.03
0.01
0.70
0.17
226.7
376
97197
1.66
0.00
1.18
0.12
229.4
371
97227
7.88
0.69
7.22
1.19
237.5
371
97228
9.35
0.80
6.79
1.26
232.0
391
97258
339
0.42
0.90
0.48
236.2
391
97259
467
1.26
1.55
0.52
251.7
Ave.
Ratio
0.90
1.03
1.03
1.04
0.99
Std.
Dev.
0.38
0.58
0.44
0.15
0.03
21
Filters.
Additional
QA/
QC
measures
were
taken
to
ensure
the
quality
of
the
filter
data.
Both
calibration
weights
and
two
control
filters
were
weighed
at
least
once
during
each
weighing
period.
They
were
re­
weighed
after
every
20th
filter,
or
if
balance
drift
was
encountered.
The
standard
deviation
on
the
control
filters
was
±
6
µ
g.
The
control
filters
were
stored
in
closed
Petri
dishes
until
the
time
of
weighing.
It
was
observed
that
they
required
much
longer
to
reach
a
constant
weight
than
sample
filters
that
had
been
equilibrated
with
the
room
air
before
weighing.
Thus,
it
is
likely
that
the
observed
variability
in
the
control
filters
represents
a
worst
case
situation.

Tunnel
blanks
are
filter
samples
collected
from
the
dilution
tunnel
operating
at
full
flow
with
no
vehicle
exhaust.
PM
collected
on
a
tunnel
blank
filter
comes
from
the
reentrainment
of
PM
deposited
in
the
tunnel
during
preceding
vehicle
testing
and
any
ambient
PM
that
passed
through
the
tunnel
air
inlet
filter.
Tunnel
blanks
were
collected
periodically
throughout
the
study.
Results
are
given
in
Table
3.7,
both
as
the
mass
collected
and
the
equivalent
emission
rate
for
an
FTP
or
an
IM240
test.
Tunnel
blanks
were
lower
during
the
summer
than
the
winter
at
both
sites.
During
the
summer
study,
an
effort
was
made
to
recruit
and
test
the
newest
vehicles
first,
followed
by
the
older
vehicles
Table
3.7.
Tunnel
Blanks
Site
Season
Mass
Collected,
µ
g
FTP
equivalent
emission
rate,
mg/
mi
IM240
equivalent
emission
rate,
mg/
mi
CDPHE
summer
15
0.076
0.049
CDPHE
summer
13
0.066
0.042
CDPHE
summer
10
0.051
0.033
CDPHE
summer
10
0.17
0.112
EPA
summer
7
0.10
0.07
EPA
summer
12
0.16
0.11
EPA2
summer
25
0.36
0.23
CDPHE
winter
31
0.54
0.35
CDPHE2
winter
9
0.27
0.18
EPA
winter
48
0.70
0.46
EPA
winter
23
0.33
0.22
EPA
winter
54
0.79
0.51
EPA
winter
85
1.20
0.78
EPA1
winter
111
1.39
0.90
EPA1
winter
77
1.06
0.68
EPA2
winter
176
1.18
0.77
1)
Collected
immediately
after
a
smoking
vehicle
2)
Collected
immediately
after
a
diesel
vehicle
22
and
finally
the
smokers
in
order
to
minimize
tunnel
contamination
by
the
higher
emitting
vehicles.
Unfortunately,
the
older
vehicles
and
smokers
were
the
hardest
to
find
and
recruit.
Given
the
time
constraints
in
the
field,
older
vehicles
and
smokers
were
tested
throughout
the
winter
testing
period.
Diesels
were
tested
at
the
end
of
the
sampling
period
during
both
the
summer
and
winter
portions
of
the
study.

Routine
emission
tests
simultaneously
collected
samples
on
both
Teflon
and
quartz
filters.
A
few
FTP
tests
were
conducted
with
Teflon
filters
in
both
filter
holders
for
comparison
purposes.
Results
are
given
in
Table
3.8.
Samples
collected
at
the
CDPHE
site
during
the
winter
are
not
shown.
Examination
of
the
data
indicated
a
consistent
difference
between
the
two
sample
trains
during
the
winter.
It
was
subsequently
determined
that
the
body
of
the
cyclone
used
for
the
quartz
sample
filter
was
leaking.
Therefore,
these
samples
were
not
used.
Agreement
between
the
duplicate
filters
was
generally
good.
Eight
IM240
tests
were
also
conducted
during
the
summer
with
duplicate
Teflon
filters.
Results
are
shown
in
Table
3.9.
Again,
comparisons
were
judged
to
be
acceptable.

Table
3.8.
Duplicate
Filter
Samples
for
FTP
Tests
Site
Season
Test/
Filter
Bag
1,
mg/
mi
Bag
2,
mg/
mi
Bag
3,
mg/
mi
FTP,
mg/
mi
CDPHE
summer
1A
87.4
20.9
23.2
35.3
CDPHE
summer
1B
114
27.0
27.0
44.9
CDPHE
summer
2A
8.8
3.7
6.1
5.5
CDPHE
summer
2B
10.3
4.1
5.6
5.8
CDPHE
summer
3A
3.4
2.1
2.5
2.5
CDPHE
summer
3B
2.4
1.4
2.4
1.9
EPA
summer
4A
15.3
4.1
13.1
8.9
EPA
summer
4B
15.9
4.1
13.9
9.2
EPA
summer
5A
6.7
1.3
2.0
2.6
EPA
summer
5B
5.2
1.3
1.7
2.2
EPA
summer
6A
4.3
0.0
0.0
0.9
EPA
summer
6B
5.4
0.0
0.3
1.2
EPA
winter
9A
76.1
33.7
22.0
40.1
EPA
winter
9B
73.8
30.1
20.5
37.8
EPA
winter
10A
39.1
32.9
15.8
30.2
EPA
winter
10B
38.7
30.4
13.6
28.1
Finally,
tests
were
conducted
with
Teflon
and
Pallflex
filters
run
simultaneously.
Some
of
the
tests
were
performed
at
the
CDPHE
site.
However,
the
aforementioned
leak
in
the
cyclone
invalidated
these
results.
Table
3.10
gives
the
results
for
the
EPA
site.
Large
percentage
differences
were
observed
for
the
first
two
runs.
We
have
no
explanation
for
this
apparent
problem.
Subsequent
runs
had
better
agreement.
Note
that
the
last
filter
pair
was
a
comparison
between
two
Pallflex
filters.
Similar
testing
conducted
at
SwRI
and
CE­
CERT
leads
to
the
conclusion
that
Pallflex
was
suitable
for
vehicle
emission
PM
measurements
(
private
communication).
23
Table
3.9.
Duplicate
Filter
Samples
for
Summer
IM240
Tests
Site
Test
#
Filter
A,
mg/
mile
Filter
B,
mg/
mi
A­
B,
mg/
mi
CDPHE
1
26.4
22.1
4.3
CDPHE
2
25.4
24.8
0.6
CDPHE
3
2.2
3.6
1.4
CDPHE
4
1.1
5.4
4.3
EPA
5
6.8
6.9
0.1
EPA
6
7.3
6.6
0.7
EPA
7
0.3
0.8
0.5
EPA
8
0.0
0.0
0.0
Table
3.10.
Comparison
of
Teflon
and
Pallflex
Filter
Samples
Run
Teflon
PM,
mg/
mi
Pallflex
PM,
mg/
mi
Difference
97172
1.95
7.37
5.4
97175
9.95
18.2
8.2
97178
59.8
65.0
5.2
97181
91.3
97.6
6.3
97184
12.7
13.4
0.7
971871
14.5
17.3
2.8
1)
Comparison
of
two
Pallflex
filters
Carbon
measurements.
As
will
be
discussed
later
in
this
report,
it
was
found
that
the
total
carbon
measured
on
some
quartz
filters
during
the
summer
portion
of
the
study
exceeded
the
total
mass
measured
on
the
Teflon
filters.
Since
elements
such
as
hydrogen,
oxygen,
sulfur,
and
nitrogen
are
associated
with
organic
particulate
matter,
the
total
carbon
should
be
consistently
less
than
the
measured
mass.
Similar
discrepancies
were
observed
in
the
CAWRSS
study
as
well
(
Sagebiel
et
al.,
1997).
At
that
time,
no
cause
of
the
difference
could
be
proven,
but
DRI,
who
performed
the
analyses,
indicated
that
inhomogeneity
in
the
filter
deposits
was
the
most
likely
explanation
since
the
carbon
analysis
is
performed
on
a
small
punch
removed
from
the
filter.
Errors
should
be
random
unless
the
punches
are
not
taken
randomly.
Inhomogeneities
were
visibly
present
on
a
few
filter
samples
collected
during
this
study.
However,
these
were
excluded
from
the
filter
set
selected
for
carbon
analysis.
Visual
inspection,
however,
is
not
adequate
to
screen
for
possible
filter
deposit
inhomogeneities.
Therefore,
multiple
punches
for
carbon
analysis
were
taken
from
three
filters
during
both
the
summer
and
winter
portions
of
the
study.
Results
are
given
in
Table
3.11.
With
the
exception
of
the
center
punch
from
the
second
filter,
results
indicate
reasonable
uniformity.
24
Table
3.11.
Carbon
Results
From
Multiple
Punches
Period
Filter
ID
OC*
EC*
TC*
Summer
DYTQ052C1
12.8
24.2
37.1
DYTQ052M2
14.6
26.5
41.0
DYTQ052E3
9.6
26.5
36.0
Average
12.3
25.7
38.0
Std.
dev.
2.1
1.1
2.1
Summer
DYTQ053C4
18.1
6.9
25.1
DYTQ053M
7.7
0.3
7.8
DYTQ053E
7.3
0.0
7.3
Average
7.5
0.2
7.6
Std.
dev.
0.2
0.2
0.3
Summer
DYTQ222
585.2
614.5
1199.7
DYTQ222C
584.8
676.0
1260.8
DYTQ222M
597.1
685.1
1282.0
DYTQ222E
578.5
624.7
1203.4
Average
586.8
661.9
1248.7
Std.
dev.
7.7
26.6
33.2
Winter
DYDQ1436C
34.7
169.4
203.1
DYDQ1436M
33.8
184.0
216.8
DYDQ1436E
32.3
170.5
201.5
Average
33.6
174.6
207.1
Std.
dev.
1.21
8.13
8.41
Winter
DYDQ0700C
102.9
6.20
108.5
DYDQO700M
101.4
5.50
106.4
DYDQ0700E
95.8
6.20
101.4
Average
100.0
5.97
105.4
Std.
dev.
3.74
0.40
3.65
Winter
DYDQ1051C
28.6
18.9
46.9
DYDQ1051M
30.1
18.3
47.8
DYDQ1051E
26.5
17.6
43.5
Average
28.4
18.27
46.1
Std
dev.
1.81
0.65
2.27
*
Units
in
µ
g/
filter
for
the
summer
and
mg/
mi
for
the
winter
1)
Punch
from
center
of
filter
2)
Punch
from
midway
between
center
and
edge
of
filter
3)
Punch
from
edge
of
filter
4)
Rejected
as
outlier
based
on
abnormal
carbon
evolution
temperature
profile
25
It
has
long
been
recognized
that
filter
sampling
of
organic
particulate
matter
can
have
errors
due
to
both
the
desorption
of
organic
material
from
the
particles
deposited
on
the
filter
and
from
the
adsorption
of
organic
material
onto
both
the
particles
and
the
filter
material
itself
(
Cadle
et
al.,
1983).
One
way
of
estimating
the
impact
of
adsorption
onto
the
filter
is
to
collect
samples
with
a
dual
filter
stack
of
quartz
filters.
This
was
done
for
four
FTP
tests
during
the
summer.
The
filter
analysis
results
are
given
in
Table
3.12.
The
data
in
this
table
give
the
carbon
in
mg
emitted
during
the
indicated
phase
of
the
FTP
test.
The
last
four
columns
give
the
percent
of
the
total
OC
and
EC
from
both
filters
that
were
measured
on
the
front
and
back
filters.
For
the
phase
1
filters,
an
average
of
7.5%
of
the
collected
OC
was
on
the
backup
filter.
However,
for
the
phase
2
filters
24.7%
of
the
OC
was
found
on
the
backup
filters.
The
amount
of
OC
collected
on
the
backup
filter
was
relatively
constant,
suggesting
an
adsorption
process
is
occurring.
The
interpretation
is
complicated,
however,
by
the
presence
of
EC
on
the
backup
filter.
Any
EC
on
the
backup
filter
should
be
due
to
particle
penetration
(
unless
the
EC
measured
by
TOR
was
really
OC).
If
that
is
the
case,
then
the
OC
associated
with
those
particles
must
be
present
on
the
backup
filter
as
well.
The
reported
EC,
however,
is
close
to
the
uncertainty
in
the
measurement,
and
may
not
represent
particle
penetration.
The
amount
of
OC
on
the
backup
filter
can
not
be
used
to
correct
the
front
filter
since
it
is
unknown
how
much
of
the
OC
was
adsorbed
from
gas
phase
HC
that
passed
through
the
front
filter
as
compared
to
the
amount
that
was
revolatilized
from
the
front
filter
and
redeposited
on
the
backup
filter.

Additional
dual
filter
experiments
were
performed
in
the
winter.
Dual
quartz
filters
were
collected
simultaneously
with
dual
Teflon/
quartz
filter
packs
during
eight
IM240
tests.
Results
are
shown
in
Table
3.13.
The
average
OC
collected
on
the
backup
filter
for
the
dual
quartz
filter
pack
was
equivalent
to
an
emission
rate
of
6.1
mg/
mi
OC.
A
relatively
constant
amount
of
OC
was
observed
on
the
backup
filter,
supporting
the
earlier
suggestion
that
this
is
due
to
adsorption
of
OC.
For
lightly
loaded
filters,
such
as
run
97073,
the
backup
filter
OC
emission
rate
is
equal
to
34%
of
that
based
on
the
front
filter
alone.
For
heavily
loaded
samples,
such
as
run
97088,
the
adsorbed
OC
on
the
backup
filter
is
less
than
one
percent
of
the
emission
rate
based
on
the
front
filter.
With
the
exception
of
the
first
run
listed
in
Table
3.13,
the
Teflon/
quartz
backup
filter
collected
more
OC
than
the
quartz/
quartz
backup
filter.
The
increase
is
presumed
to
be
due
to
OC
that
that
is
adsorbed
on
the
front
quartz
filter,
but
passes
through
or
is
released
from
deposits
on
the
Teflon
filter.
Table
3.13
also
shows
that
a
small
amount
of
EC
was
reported
on
the
backup
quartz
filters.
Similar
amounts
were
seen
for
both
the
Teflon
and
quartz
backup
filters.
As
before,
these
levels
are
close
to
the
uncertainty
of
the
measurement
method
and
probably
represent
analytical
error.
The
total
carbon
collected
on
the
Teflon/
quartz
backup
filter
averaged
13.9%
of
the
total
carbon
collected
on
the
front
quartz
filter.
The
additional
TC
collected
behind
the
Teflon
filter
averaged
6.2%
of
the
total
carbon
on
the
quartz
filter.
No
speciation
of
the
backup
filter
organic
carbon
was
performed.
Overall,
it
is
concluded
that
quartz
filters
can
capture
substantially
more
OC
than
Teflon
filters.
It
is
not
possible
to
determine
if
all
of
the
adsorbed
OC
was
initially
present
in
the
gas
phase,
or
if
some
of
it
was
released
from
particles
collected
on
the
front
filter.
However,
it
is
likely
that
the
OC
determined
from
quartz
filters
is
an
overestimate
26
of
the
actual
particulate
matter
OC.

Table
3.12.
Carbon
Measured
on
Front
and
Back
Summer
Quartz
Filters
Front
Filter
Back
Filter
%
of
OC
on
%
of
EC
on
Run
Phase
OC
EC
TC
OC
EC
TC
Front
Back
Front
Back
96047
1
115
66.0
181
9.0
0.9
9.9
92.8
7.2
98.6
1.4
96078
1
128
38.7
167
12.0
2.0
13.9
91.4
8.6
95.1
4.9
96081
1
25.2
28.0
53.2
3.7
1.1
4.8
87.2
12.8
96.2
3.8
96090
1
236
249
485
3.2
0.0
3.1
98.7
1.3
100.0
0.0
MEAN
92.5
7.5
97.5
2.5
STD
DEV
4.1
4.1
1.9
1.9
%
of
OC
on
%
of
EC
on
Run
Phase
OC
EC
TC
OC
EC
TC
Front
Back
Front
Back
96047
2
25.7
16.9
42.7
7.9
1.2
9.1
76.5
23.5
93.2
6.8
96078
2
38.5
3.1
41.6
12.3
3.1
15.5
75.8
24.2
49.7
50.3
96081
2
11.1
1.0
12.0
4.8
0.6
5.4
69.8
30.2
62.5
37.5
96090
2
12.4
3.7
16.1
3.3
0.7
3.9
79.2
20.8
84.4
15.6
MEAN
75.3
24.7
72.4
27.6
STD
DEV
3.5
3.5
17.3
17.3
%
of
OC
on
%
of
EC
on
Run
Phase
OC
EC
TC
OC
EC
TC
Front
Back
Front
Back
96047
3
24.3
25.9
50.2
7.6
1.6
9.2
76.1
23.9
94.0
6.0
96078
3
41.0
9.3
50.2
7.5
0.9
8.4
84.5
15.5
91.5
8.5
96081
3
11.9
3.2
15.1
2.7
1.8
4.5
81.4
18.6
64.0
36.0
96090
3
13.7
14.1
27.8
0.0
0.2
0.0
100.0
0.0
98.9
1.1
MEAN
85.5
14.5
87.1
12.9
STD
DEV
8.9
8.9
13.6
13.6
Carbon
units
are
total
mg
emitted
per
phase,
front
filters
are
carbon
tunnel
blank
corrected
Teflon
­
Pallflex
Comparisons.
During
the
summer,
Teflon
filter
samples
from
high
emission
vehicles
were
observed
to
lose
some
flow
due
to
plugging.
Therefore,
in
the
winter
Pallflex
T60A20
filters
were
used
in
place
of
Teflon
membrane
filters
for
these
vehicles.
Three
sets
of
simultaneously
collected
Teflon
membrane
and
Pallflex
T60A20
filters
were
collected
and
analyzed
by
IC
and
XRF
for
comparison
purposes.
One
of
the
filters
was
lost
during
analysis.
All
T60A20
filters
had
high
nitrate
backgrounds
and
higher
backgrounds
of
most
of
the
elements
determined
by
XRF.
Therefore,
the
only
valid
comparison
for
these
filters
was
for
sulfate.
The
sulfate
emission
rate
determined
from
the
two
Teflon
membrane
filters
was
0.098
and
0.181
mg/
mi
while
it
was
0.086
and
0.174,
respectively,
from
the
Pallflex
filters.

Table
3.13.
Carbon
Measured
on
Front
and
Back
Winter
Quartz
Filters
27
Run
Cycle
OC,
mg/
mi
EC,
mg/
mi
TC,
mg/
mi
Front
Qback1
Tback2
Front
Qback
Tback
Front
Qback
Tback
97054
IM240
44.7
15.6
8.7
35.5
0.9
0.3
79.7
15.9
8.5
97063
IM240
71.2
3.4
4.3
61.8
0.8
1.3
132
3.7
5
97066
IM240
163
7.1
14.4
19.6
1.4
1.6
182
7.9
15.5
97069
IM240
16.7
5.7
10.8
3.2
0.2
0.3
19.2
5.3
10.5
97073
IM240
8.2
0
2.3
1.8
0
0
9.4
0
1.7
97082
IM240
89.1
4.6
9.1
26.8
0.1
0.3
115
4.1
8.8
97088
IM240
666
6.5
11.8
169
0.5
0.4
835
6.4
11.6
97091
IM240
145
5.7
12.5
42.5
0.4
0.5
187
5.5
12.4
1)
Qback
is
the
quartz
backup
filter
behind
a
quartz
front
filter
2)
Tback
is
the
quartz
backup
filter
behind
a
Teflon
front
filter
28
29
30
4.
Emission
Rates
of
PM
and
Regulated
Gaseous
Pollutants
This
section
presents
the
average
emission
rates
of
PM
and
the
regulated
pollutants
as
well
as
the
relationships
between
them.
Results
will
focus
on
the
average
emission
rates
for
the
six
categories
of
vehicles
tested:
1971­
80,
1981­
85,
1986­
90,
1991­
96,
smokers,
and
diesels.

Vehicles
Table
4.1
gives
the
number
of
vehicles
in
each
category
as
well
as
their
average
age
and
odometer
reading.
1996
model
year
vehicles
were
defined
as
being
0.5
years
old
during
the
summer
and
1.0
year
old
during
the
winter.
Not
all
drivers
were
asked
if
their
odometer
had
rolled
over
(
exceeded
100,000
miles
on
an
odometer
that
reads
a
maximum
of
99,999
miles),
and
some
drivers
did
not
know
if
the
odometer
had
rolled
over.
Thus
the
odometer
readings
on
old
vehicles
are
highly
uncertain.
The
average
odometer
readings
in
Table
4.1
assumed
that
any
vehicle
older
than
1986
with
an
odometer
reading
less
than
50,000
miles
had
rolled
over
once.
Uncorrected
values
are
given
in
the
footnote
to
the
Table.
These
odometer
readings
should
be
viewed
as
estimates
since
errors
can
be
produced
by
factors
other
than
rollover.
The
biggest
differences
between
the
summer
and
winter
vehicle
fleets
are
for
the
smokers
and
the
diesels.
Winter
smokers
averaged
significantly
older
than
those
from
the
summer,
although
the
reported
average
odometer
readings
are
essentially
the
same.
The
range
in
smoker
model
years
was
large,
1969­
1990
in
the
winter
and
1966­
1989
in
the
summer.
The
average
age
of
diesels
in
the
summer
was
7.3
years,
while
it
was
15.3
in
the
winter.
This
was
the
result
of
selecting
vehicles
during
the
winter
that
could
be
accommodated
on
the
EPA
chassis
dynamometer.
Most
newer
vehicles
are
in
relatively
large
trucks
which
are
too
heavy
for
the
EPA
dynamometer.
Thus,
the
winter
diesel
vehicles
were
dominated
by
older
passenger
cars
that
ranged
in
model
year
from
1979
to
1989.
The
summer
diesels
ranged
in
model
year
from
1980
to
1994.
Appendices
B­
1
and
B­
3
list
all
of
the
individual
vehicles.

Table
4.1.
Vehicle
Age
and
Odometer
Vehicle
Category
Number
Average
Age
Average
Odometer1
Number
Average
Age
Average
Odometersummer
winter
1971­
80
25
19.8
117,000
17
19.3
149,000
1981­
85
25
13.2
142,000
16
13.4
113,000
1986­
90
22
8.7
109,000
14
8.5
102,000
1991­
96
20
3.0
40,000
10
3.2
65,000
Smoker
9
13.8
110,000
15
17.4
119,000
Diesel
10
7.3
89,000
12
15.2
161,000
1)
Uncorrected
odometer
averages:
Summer
1971­
80,
85,000;
summer
1981­
85,
134,000;
winter
1971­
80,
130,000;
winter
1981­
85,
95,000;
winter
diesel,
151,000.

FTP
emission
rates
of
regulated
pollutants
Table
4.2
gives
the
average
emission
rates
for
each
vehicle
category
tested
during
the
summer
and
the
winter.
Winter­
CDPHE
was
the
31
indoor
site
while
winter­
EPA
was
the
outdoor
site.
In
all
cases,
the
average
emission
rates
increase
with
increasing
age.
Figure
4.1
shows
the
relationship
between
vehicle
age
and
FTP
HC
emission
rate
for
the
individual
summer
gasoline
vehicles,
excluding
recruited
smokers.
The
increase
in
emissions
with
age
is
obvious,
although
it
is
also
seen
that
some
older
vehicles
have
relatively
low
emissions.
All
vehicles
less
than
six
years
old
had
low
HC
emissions.
Figure
4.2
shows
the
same
relationship
for
NOx.
Again
emissions
increase
with
age,
but
the
range
is
considerably
lower
than
for
HC.
As
expected,
the
diesels
have
relatively
low
HC
and
CO
emissions
but
high
NOx.
The
standard
deviation,
minimum
value
and
maximum
value
for
each
of
the
average
emission
rates
are
given
in
Appendix
F1.
The
highest
individual
HC
and
CO
emission
rates
for
diesel
vehicles
were
2.89,
and
9.08
g/
mi.,
respectively,
while
those
for
gasoline
vehicles
were
45
and
347
g/
mi,
respectively.

Table
4.2.
Average
FTP
Regulated
Emission
Rates
Period
Category
Number
HC
CO
NOx
­­
g/
mi
­­
Summer
1991­
96
20
0.24
2.35
0.42
Summer
1986­
90
22
1.03
14.7
1.32
Summer
1981­
85
25
2.7
35.9
1.54
Summer
1971­
80
25
6.08
62.6
2.46
Summer
Smoker
9
3.59
38.8
1.88
Summer
Diesel
10
ND
3.99
4.57
Winter­
CDPHE
1991­
96
10
0.56
6.74
0.51
Winter­
CDPHE
1986­
90
14
0.63
9.23
1.44
Winter­
CDPHE
1981­
85
16
2.57
35.7
2.76
Winter­
CDPHE
1971­
80
16
2.78
44.3
2.11
Winter­
CDPHE
Smoker
15
4.96
52.6
2.28
Winter­
CDPHE
Diesel
12
0.82
1.94
1.73
Winter­
EPA
1991­
96
9
1.02
11.6
0.53
Winter­
EPA
1986­
90
14
1.06
15.5
1.61
Winter­
EPA
1981­
85
16
3.39
38.7
2.31
Winter­
EPA
1971­
80
16
4.49
52.1
2.01
Winter­
EPA
Smoker
15
8.71
65.6
2.29
Winter­
EPA
Diesel
12
1.1
1.76
1.55
Winter
indoor
and
outdoor
measurements
on
the
same
vehicles
can
also
be
compared.
In
section
3,
it
was
shown
that
FTP
bag
2
and
3
emission
rates
were
in
good
agreement
between
the
indoor
and
outdoor
sites.
Table
4.2,
on
the
other
hand,
shows
that
FTP
emission
rates
consistently
averaged
higher
outdoors
than
indoors,
as
would
be
expected
for
the
lower
temperatures.
Clearly,
all
of
this
difference
is
attributable
to
the
bag
1
emissions,
which
are
higher
due
to
the
longer
warm­
up
times
for
the
emission
control
system.
Average
vehicle
oil
temperature
and
ambient
temperature
at
the
start
of
the
emissions
tests
are
given
in
Table
4.3.
In
most
cases
the
oil
temperature
lagged
the
ambient
temperature
since
testing
was
done
after
an
overnight
soak.
Average
test
32
temperatures
for
each
vehicle
category
were
from
22
to
34
°
F
colder
outdoors
than
indoors.

Table
4.3.
Winter
Outdoor
Average
Vehicle
Test
Temperature
Vehicle
Category
Oil
Temperature
Ambient
Temperature
°
F
1971­
80
28.2
30.9
1981­
85
36.6
40.7
1986­
90
31.9
38.4
1991­
96
35.0
38.0
Smoker
39.9
44.4
Diesel
39.5
40.1
IM240
emission
rates
of
regulated
pollutants.
IM240
average
emission
rates
are
given
in
Table
4.4
for
both
the
summer
and
winter
periods.
Appendix
F2
gives
the
standard
deviation,
the
minimum
and
the
maximum
values
for
each
category
as
well.
Since
the
IM240
is
a
hot
start
test,
little
effect
of
temperature
is
expected.
Figure
4.3
shows
the
relationship
between
the
indoor
and
outdoor
sites
for
HC
on
the
same
vehicles.
The
linear
regression
slope
of
1.03
and
the
R2
of
0.82
are
good
considering
the
differences
in
preconditioning
and
other
factors
between
the
test
sites.
The
slopes
and
regression
coefficients
for
CO
were
0.74
and
0.73,
respectively,
while
those
for
NOx
were
0.56
and
0.79,
respectively.
The
NOx
regression
had
a
significant
non­
zero
intercept.

IM240
emission
rates
of
regulated
species
were
compared
to
those
measured
on
the
FTP
for
both
the
summer
and
winter.
Figure
4.4
shows
the
relationship
for
CO
during
the
summer,
while
Figure
4.5
shows
that
for
NOx
at
the
EPA
site
during
the
winter.
All
vehicles,
including
the
smokers
and
diesels
were
included
in
these
comparisons.
Correlation
coefficients
are
good,
0.82
and
0.90
for
CO
and
NOx,
respectively.
Table
4.5
gives
the
slopes
and
regression
coefficients
for
all
of
the
regulated
pollutants.
In
general,
regression
coefficients
indicate
a
strong
relationship
between
the
IM240
and
the
FTP
emission
rates.
Agreement
is
closest
for
NOx.
In
all
cases,
CO
and
HC
linear
regression
slopes
indicate
that
emissions
are
lower
on
the
IM240
than
the
FTP.
This
was
especially
true
for
the
EPA
site
in
the
winter
due
to
the
elevated
FTP
emission
rates
at
the
low
cold
start
temperature.

Idle
test
emission
rates
and
concentration.
Vehicles
received
both
an
incoming
two
speed
idle
test
and
a
post
IM240
idle
test.
These
tests
were
added
to
the
program
for
other
purposes
and
won't
be
analyzed
for
this
report.
The
results,
however,
are
given
in
Appendix
G.

Table
4.4.
IM240
Average
Emission
Rates
Period
Category
N
HC
CO
NOx
33
­­
g/
mi
­­
Summer
1971­
80
22
4.74
59.7
2.90
Summer
1981­
85
24
1.72
25.4
1.76
Summer
1986­
90
20
0.77
14.6
1.47
Summer
1991­
96
17
0.13
2.18
0.43
Summer
Smoker
7
2.23
33.0
2.19
Summer
Diesel
10
ND
3.10
4.14
Winter­
CDPHE
1971­
80
16
2.24
40.6
2.67
Winter­
CDPHE
1981­
85
15
1.76
29.7
3.59
Winter­
CDPHE
1986­
90
14
0.57
7.64
2.93
Winter­
CDPHE
1991­
96
9
0.51
5.04
0.63
Winter­
CDPHE
Smoker
15
3.84
42.5
3.05
Winter­
CDPHE
Diesel
12
0.63
1.68
1.49
Winter­
EPA
1971­
80
15
2.15
27.4
1.81
Winter­
EPA
1981­
85
16
1.27
20.9
2.18
Winter­
EPA
1986­
90
14
0.39
7.62
1.65
Winter­
EPA
1991­
96
9
0.35
2.94
0.45
Winter­
EPA
Smoker
15
3.81
40.0
2.36
Winter­
EPA
Diesel
12
0.64
1.45
1.15
Table
4.5.
Slope
and
Regression
Coefficients
for
Linear
Regressions
Between
the
FTP
and
IM240
Emission
Rates
Period
Species
Slope
R2
Summer
HC
0.73
0.95
Summer
CO
0.75
0.82
Summer
NOx
1.14
0.84
Winter­
CDPHE
HC
0.64
0.87
Winter­
CDPHE
CO
0.77
0.82
Winter­
CDPHE
NOx
1.10
0.87
Winter­
EPA
HC
0.46
0.72
Winter­
EPA
CO
0.53
0.85
Winter­
EPA
NOx
0.88
0.90
PM
FTP
emission
rates.
Table
4.6
gives
the
average
PM­
10
emission
rates
by
vehicle
grouping
for
both
the
summer
and
winter
portions
of
the
study.
Appendix
F3
gives
the
standard
deviation,
the
minimum
value,
and
the
maximum
values
for
each
vehicle
category
as
well.
The
effect
of
vehicle
age
is
pronounced,
with
the
summer
1971­
80
vehicles
emitting
an
average
of
95.4
mg/
mi
compared
to
average
emissions
of
2.8
mg/
mi
from
the
summer
1991­
96
vehicles.
Similar
results
were
observed
for
the
winter
measurements.
Figure
4.6
shows
the
FTP
PM
emission
rates
as
a
function
of
vehicle
age
for
the
summer
non­
smoking
gasoline
vehicles.
It
is
clear
that
most
new
vehicles
have
low
PM
emission
rates,
but
that
after
10
years
there
is
a
wide
range
of
emissions.
The
winter
CDPHE
average
for
the
1971­
80
and
1991­
96
vehicles
in
Table
4.6
are
given
both
for
all
tested
34
vehicles,
and
for
only
those
that
match
those
tested
at
the
EPA
site
outdoors,
a
difference
of
one
vehicle
in
each
case.
This
permits
a
valid
direct
comparison
between
the
indoor
and
outdoor
data.
This
comparison
is
shown
graphically
in
Figure
4.7.
Emission
rates
decrease
with
age
for
all
three
groups
(
summer,
winter
indoor,
and
winter
outdoor).
The
indoor
winter
emission
rates
average
lower
than
those
for
comparable
vehicles
tested
during
the
summer
for
the
1971­
80,
1981­
85,
and
1986­
90
vehicles.
Emission
rates
were
similar
for
the
1991­
96
vehicles
(
2.82
in
the
summer,
3,74
in
the
winter).
It
is
possible
that
the
decrease
between
the
two
seasons
is
due
to
fuel
changes.
Oxygenated
fuels
are
mandated
in
the
winter
in
the
Denver
area,
but
not
the
summer.
It
was
found
in
a
study
of
PM
emission
rates
from
vehicles
using
Alaska
fuel
(
Mulawa
et
al.,
1997),
that
adding
ethanol
reduces
PM
formation
at
temperatures
from
72
to
 
20
°
F.
However,
other
changes
in
fuel
composition
such
as
sulfur
or
aromatic
content
could
impact
PM
emission
rates
as
well.
Fuel
composition
is
examined
later
in
this
report.
As
expected,
average
PM
emission
rates
in
the
outdoor
tests
exceeded
those
of
the
indoor
tests.
The
winter
outdoor
PM
emission
rates
were
lower
than
those
from
the
summer
for
the
1971­
80
and
1986­
90
vehicles.
1981­
85
vehicles
had
slightly
higher
average
emissions
in
the
winter
than
summer.
The
major
difference
is
for
the
1991­
96
vehicles,
which
had
a
much
higher
winter
outdoor
average.
The
winter­
EPA
PM
average
drops
to
12.6
with
the
removal
of
one
vehicle
with
a
PM
emission
rate
of
124
mg/
mi.
Thus,
the
emission
rate
for
this
category
remain
uncertain.

The
smoking
vehicle
and
diesel
vehicle
populations
between
the
summer
and
winter
were
not
well
matched.
Thus,
these
emissions
should
not
be
compared.
It
is
appropriate,
however,
to
compare
the
indoor
and
outdoor
tests.
For
the
smoking
vehicles,
the
emissions
averaged
9.8%
higher
outdoors.
The
diesels
also
averaged
9.3%
higher
outdoors.
This
is
a
much
lower
impact
than
seen
for
the
randomly
recruited
gasoline
vehicles.
For
example,
the
difference
in
the
1981­
85
vehicles
was
34%.

At
least
some
of
the
difference
in
temperature
sensitivity
between
the
non­
smoking
gasoline
vehicles
and
the
diesel
and
smoking
gasoline
vehicles
is
due
to
the
relative
importance
of
emissions
in
the
three
phases
of
the
FTP.
As
noted
earlier,
only
phase
1
is
affected
significantly
by
the
ambient
temperature.
Table
4.6
gives
the
fraction
of
the
total
mass
emitted
over
the
entire
FTP
that
was
emitted
in
each
phase.
Since
the
FTP
weighs
the
three
phases,
this
is
not
the
same
as
the
fractional
contribution
to
the
FTP.
Emission
rate
by
phase
are
given
in
Table
4.7.
For
the
non­
smoking
gasoline
vehicles,
the
majority
of
the
mass
is
emitted
in
phase
1.
Due
to
the
temperature
effect,
the
fraction
in
phase
1
is
even
greater
for
the
vehicles
in
the
winter
outdoor
tests
than
the
indoor
tests.
Figure
4.8
plots
the
vehicle
oil
temperature
and
the
ambient
temperature
immediately
before
the
FTP
test,
as
well
as
the
FTP
phase
1
total
mass
emission
rates
outdoors
(
EPA)
and
indoors
(
CDPHE)
for
each
paired
vehicle
in
the
1981­
85
vehicle
category.
The
PM
emissions
are
higher
outdoors
for
13
of
the
15
vehicles.
Phase
2
results
are
shown
for
the
same
vehicles
in
Figure
4.9.
The
emissions
are
much
lower,
and
the
differences
appear
to
be
random
and
within
the
test
variability.
Figure
4.10
shows
the
effect
on
diesel
phase
1
FTP
PM
emissions.
Ten
out
of
12
vehicles
had
higher
emissions
at
the
reduced
temperature.
Phase
2
FTP
diesel
emissions
are
shown
in
Figure
4.11.
Only
4
of
the
12
vehicles
had
higher
emissions
at
the
lower
temperature.
In
addition,
the
emissions,
while
lower
than
35
during
phase
1,
are
still
substantial.
Thus,
the
overall
impact
of
temperature
was
less
for
the
diesel
than
the
non­
smoking
gasoline
vehicles.

Table
4.6.
FTP
PM
Emission
Rates
and
Bag
Fractions
Period
Category
N
Fraction
Bag
1
Fraction
Bag
2
Fraction
Bag
3
PM,
mg/
mi
Summer
1971­
80
25
0.64
0.18
0.19
95.5
Summer
1981­
85
25
0.66
0.17
0.17
47.4
Summer
1986­
90
22
0.61
0.21
0.18
44.4
Summer
1991­
96
20
0.58
0.19
0.23
2.82
Summer
Smoker
9
0.49
0.24
0.27
225
Summer
Diesel
10
0.54
0.16
0.29
811
Winter­
CDPHE
1971­
80
16
0.62
0.16
0.22
54.2
Winter­
CDPHE
(
match
vehicles
to
EPA)
1971­
80
15
0.63
0.15
0.22
56.6
Winter­
CDPHE
1981­
85
16
0.59
0.2
0.21
35.9
Winter­
CDPHE
1986­
90
14
0.71
0.13
0.16
11.8
Winter­
CDPHE
1991­
96
10
0.62
0.16
0.21
3.51
Winter­
CDPHE
(
match
vehicles
to
EPA)
1991­
96
9
0.62
0.18
0.20
3.74
Winter­
CDPHE
Smoker
15
0.45
0.28
0.26
395
Winter­
CDPHE
Diesel
12
0.41
0.27
0.32
460
Winter­
EPA
1971­
80
15
0.77
0.11
0.13
82.6
Winter­
EPA
1981­
85
16
0.75
0.12
0.13
48.2
Winter­
EPA
1986­
90
14
0.68
0.16
0.15
28.5
Winter­
EPA
1991­
96
9
0.83
0.11
0.07
24.9
Winter­
EPA
Smoker
15
0.59
0.22
0.2
434
Winter­
EPA
Diesel
12
0.50
0.24
0.25
503
Table
4.7.
FTP
PM
emission
rates
by
phase.

Period
Category
N
Phase
1
mg/
mi
Phase
2
mg/
mi
Phase
3
mg/
mi
FTP
PM,
mg/
mi
Summer
1971­
80
25
235
54.2
68.4
95.5
Summer
1981­
85
25
98.9
32.1
38.5
47.4
Summer
1986­
90
22
91.7
33.6
28.8
44.4
Summer
1991­
96
20
5.7
1.8
2.4
2.82
Summer
Smoker
9
253
231
192
225
Summer
Diesel
10
1646
398
964
811
36
Winter­
CDPHE
1971­
80
16
139
27.4
40.3
54.2
Winter­
CDPHE
(
match
vehicles
to
EPA)
1971­
80
15
147
28.2
42.1
56.6
Winter­
CDPHE
1981­
85
16
82.6
20.3
30.2
35.9
Winter­
CDPHE
1986­
90
14
27.9
7.3
8.2
11.8
Winter­
CDPHE
1991­
96
10
8.9
1.7
2.8
3.51
Winter­
CDPHE
(
match
vehicles
to
EPA)
1991­
96
9
9.4
1.9
2.9
3.74
Winter­
CDPHE
Smoker
15
742
298
315
395
Winter­
CDPHE
Diesel
12
699
377
494
460
Winter­
EPA
1971­
80
15
290
26.1
37.9
82.6
Winter­
EPA
1981­
85
16
159
19.0
21.4
48.2
Winter­
EPA
1986­
90
14
70.6
16.9
18.7
28.5
Winter­
EPA
1991­
96
9
81.3
13.4
6.3
24.9
Winter­
EPA
Smoker
15
1179
231
272
434
Winter­
EPA
Diesel
12
875
384
448
503
Comparison
of
PM
and
HC
FTP
emission
rates.
In
previous
studies,
the
lack
of
PM
emission
measurements
on
light­
duty
gasoline
vehicles
has
forced
the
use
of
HC
as
a
surrogate
for
PM.
In
the
EPA
Motor
Vehicle­
Related
Air
Toxics
Study
(
EPA,
1994),
it
was
estimated
that
PM
is
1.1%
of
the
mass
of
total
HC
emitted
on
an
FTP.
In
the
CAWRSS
and
Orange
County
studies,
the
relationship
between
IM240
PM
and
HC
was
examined
and
found
to
be
very
weak.
In
the
Alaska
study
on
FTP
PM
emissions
(
Mulawa
et
al.,
1997)
a
good
correlation
was
found.
The
relationship
between
PM
and
HC
emission
rates
was
reexamined
for
this
study.
Figure
4.12
plots
the
PM
and
HC
emission
rates
for
the
individual
non­
smoking
gasoline
vehicles
tested
during
the
summer.
The
correlation
for
a
linear
regression
is
poor
(
R2=
0.23).
The
linear
correlations
for
the
winter
indoor
and
outdoor
data
were
similar
with
R2
of
0.16
and
0.24,
respectively.
Thus,
it
is
clear
that
HC
is
not
a
good
surrogate
for
gasoline
PM
emissions
on
an
individual
basis.
This
is
not
unexpected
since
the
PM
emissions
can
arise
from
at
least
two
separate
phenomena,
rich
air/
fuel
operation
and
oil
consumption.
On
the
other
hand,
examination
of
the
average
emissions
rates
in
Tables
4.2
and
4.6
indicate
that
both
average
PM
and
HC
do
increase
with
vehicle
age.
Therefore,
the
average
PM
and
HC
emission
rates
for
the
non­
smoking
gasoline
vehicles
from
the
summer
and
the
indoor
and
outdoor
winter
portions
of
the
study
were
compared.
The
result,
shown
in
Figure
4.13,
shows
that
average
values
based
on
vehicle
age
have
an
R2
of
0.89.
The
slope
of
the
linear
regression
indicates
PM
increases
15
mg/
mi
for
each
g/
mi
of
HC.
This
is
somewhat
higher
than
the
1.1%
value
used
by
the
EPA.

IM240
PM
emission
rates.
Table
4.8
gives
the
average
IM240
PM
emission
rates
by
vehicle
category
and
study
period.
The
standard
deviation,
minimum
value,
and
maximum
value
for
each
category
and
study
period
are
given
in
Appendix
F4.
These
averages
are
not
matched
with
those
for
the
FTP
in
Table
4.6
since
the
number
of
vehicles
in
some
categories
is
less.
The
main
reason
for
measuring
IM240
PM
emission
rates
was
to
compare
the
results
to
those
of
the
FTP.
If
results
are
in
good
agreement,
then
the
37
IM240
data
obtained
in
previous
studies
could
be
used
with
higher
confidence
as
a
representation
of
in­
use
emissions.
Also,
further
consideration
could
be
given
to
making
IM240
measurements
of
PM.
The
comparison
of
IM240
and
FTP
PM
emission
rates
for
the
summer
is
given
in
Figure
4.14.
This
regression
includes
all
vehicles,
including
diesels.
One
high
emitting
diesel
was
omitted
from
the
comparison
since
it
was
so
far
off
scale.
Including
this
point
changed
the
R2
to
0.97
and
the
slope
to
0.86.
Including
the
diesel
and
smoking
vehicles
in
the
comparison
greatly
enhances
the
correlation.
Figure
4.15
shows
the
comparison
for
non­
smoking
gasoline
vehicles
only.
There
is
much
more
scatter
in
the
data
and
the
R2
drops
to
0.53.
The
comparisons
for
the
winter
indoor
and
outdoor
data
sets
are
shown
in
Figures
4.16
and
4.17,
with
all
vehicles
included.
The
regressions
are
good
(
R2
of
0.88
and
0.84),
and
the
slopes
are
surprisingly
close
to
unity
(
1.02
and
0.91).
The
good
agreement
is
somewhat
surprising
given
the
impact
of
cold
start
emissions
on
the
FTP.
Presumably,
the
regressions
are
driven
by
the
higher
emitters,
which
have
less
of
a
cold
start
impact
than
the
low
emitters.

Table
4.8.
IM240
PM
Emission
Rates
Period
Category
N
Average
PM,
mg/
mi
Summer
1971­
80
22
82.1
Summer
1981­
85
24
21.9
Summer
1986­
90
20
21.6
Summer
1991­
96
17
2.18
Summer
Diesel
8
762
Summer
Smoker
7
192
Winter­
CDPHE
1971­
80
16
43.0
Winter­
CDPHE
1981­
85
16
16.38
Winter­
CDPHE
1986­
90
14
14.4
Winter­
CDPHE
1991­
96
9
1.57
Winter­
CDPHE
Diesel
12
403
Winter­
CDPHE
Smoker
15
350
Winter­
EPA
1971­
80
16
27.6
Winter­
EPA
1981­
85
16
31.2
Winter­
EPA
1986­
90
14
172
Winter­
EPA
1991­
96
9
4.24
Winter­
EPA
Diesel
12
394
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
5.
Particle
Number
and
Size
Particle
number.
The
number
of
particles
larger
than
0.01
µ
m
in
diameter
was
monitored
continuously
using
an
Electrical
Aerosol
Analyzer
(
EAA).
Figures
5.1
through
5.3
plot
the
EAA
trace
for
the
three
phases
of
the
FTP.
Results
are
given
for
three
of
the
summer
emission
tests.
These
were
selected
to
represent
vehicles
with
high,
medium,
and
low
PM­
10
emission
rates.
The
dark
solid
line
in
the
figures
is
the
FTP
speed­
time
trace.
Both
the
speed­
time
trace
and
the
EAA
traces
are
plotted
in
actual
time,
i.
e.
the
EAA
data
were
not
corrected
for
any
lags
due
to
transfer
lines.
Therefore,
the
EAA
data
are
probably
shifted
a
few
seconds
to
longer
times
compared
to
the
FTP
trace.
Note
also
that
the
relative
particle
count
scale
(
vertical
scale)
is
expanded
for
the
phase
2
and
3
plots
compared
to
phase
1.

Several
conclusions
can
be
drawn
from
these
data.
First,
the
particle
count
from
the
high
PM
mass
emitter
is
consistently
higher
than
the
count
from
the
medium
mass
emitter,
which
in
turn
is
higher
than
that
for
the
low
emitter,
for
phase
1
of
the
FTP.
This
suggests
that
the
integrated
particle
count
may
be
roughly
proportional
to
particle
mass.
Note
that
the
medium
emitter
emits
higher
numbers
of
particles
than
the
high
emitter
in
phases
2
and
3
of
the
FTP.
This
was
also
reflected
in
the
measured
mass
emissions.
Second,
the
number
of
particles
emitted
is
highly
variable,
even
for
the
high
emitter.
Some
of
the
variation
in
particle
number
is
due
to
the
change
in
exhaust
dilution
ratio
in
the
dilution
tunnel,
which
is
proportional
to
exhaust
flow.
It
is
apparent,
however,
that
spikes
in
particle
number
emissions
are
associated
with
driving
events,
such
as
acceleration.
This
has
been
observed
in
other
tests
as
well
(
Maricq
et
al.,
1998)
and
is
also
commonly
observed
for
the
emissions
of
gaseous
regulated
pollutants.
This
suggests
that
PM
emission
rates
are
cycle
dependent,
and
that
real­
world
PM
emission
studies
conducted
under
steady­
state
operation
will
tend
to
underestimate
the
number
of
particles
emitted
and
probably
the
particle
mass.
Third,
in
many
instances,
all
three
vehicles
show
peaks
in
particle
number
emissions
at
the
same
point
in
the
cycle,
indicating
a
consistency
in
the
particle
formation
events
for
different
vehicles.

The
areas
under
the
EAA
curves
were
integrated
to
determine
the
total
number
of
particles
emitted
on
each
phase
of
the
FTP.
Twenty­
five
FTP
runs
from
the
summer
were
judged
to
have
valid
data.
All
of
these
runs
were
for
gasoline
vehicles.
Results
are
given
in
Table
5.1
along
with
the
corresponding
measured
FTP
mass
emission
rates.
For
the
winter,
there
were
84
FTP
runs
with
EAA
data,
twelve
of
which
were
for
diesel
vehicles.
All
EAA
data
were
collected
indoors
on
the
CDPHE
site.
Table
5.2
gives
the
EAA
particle
emission
rates
for
the
gasoline
vehicles
in
the
winter,
while
Table
5.3
gives
the
rates
for
the
diesel
vehicles.
As
before,
the
corresponding
PM­
10
mass
emission
rates
are
also
given.
Linear
regressions
were
run
between
the
EAA
particle
number
and
the
PM­
10
mass
emission
rates
for
both
the
summer
and
the
winter.
Results
were
best
for
the
summer
data,
which
are
shown
in
Figure
5.4.
This
figure
includes
data
from
all
three
phases
of
each
FTP.
While
there
is
clearly
considerable
scatter
in
the
data,
the
R2
of
0.70
is
good.
Correlation
coefficients
were
similar
for
each
FTP
phase
considered
separately
(
R2
of
0.71,
0.64,
and
0.77)
and
for
the
composite
FTP
(
R2
of
0.69).
There
was
more
scatter
in
the
winter
data,
as
shown
in
Figure
5.5,
which
plots
all
the
FTP
phases
from
the
55
gasoline
vehicles.
The
R2
is
poor,
only
0.29.
One
difference
in
the
two
data
sets
is
that
the
winter
includes
more
high
emitters.
It
is
possible
that
a
few
large
particles
were
collected
on
some
samples.
This
could
significantly
affect
the
particle
number
to
mass
ratio.
Therefore,
we
examined
only
the
data
with
mass
emission
rates
of
300
mg/
mi
or
less.
The
correlation
improved
slightly
to
an
R2
of
0.39.
It
does
not
appear
that
a
few
vehicles
are
dominating
the
correlation.
The
R2
values
for
correlations
between
PM­
10
mass
emission
rate
and
the
particle
number
emission
rate
for
each
FTP
phase
were
0.36,
0.16,
and
0.25
for
the
winter.
Correlations
were
slightly
better
for
the
diesel
vehicles,
with
an
overall
R2
of
0.35
with
all
phases
considered,
and
individual
FTP
phase
1,
2,
and
3
correlations
of
0.48,
0.26,
and
0.27,
respectively.

PM­
10
and
total
particle
number
need
not
be
well
correlated.
A
large
number
of
very
small
particles
contribute
little
to
the
total
mass.
For
example,
it
takes
one
billion
0.01
µ
m
particles
to
equal
the
mass
of
one
10
µ
m
particle,
assuming
they
all
are
spheres
with
equal
density.
Thus,
the
total
particle
emission
rate
could
undergo
large
variations
with
little
effect
on
the
emitted
PM­
10
mass
concentration.
The
observation
that
PM­
10
and
particle
number
correlate
as
well
as
they
do
suggests
that
the
particle
size
distributions
are
fairly
constant.
56
Table
5.1.
Summer
EAA
Particle
Emission
Rates
and
PM­
10
Mass
Emission
Rates
Run
#
Phase
1
Phase
1
Phase
2
Phase
2
Phase
3
Phase
3
FTP
Composite
particle/
mi
mg/
mi
particle/
mi
mg/
mi
particle/
mi
mg/
mi
Particle/
mi
mg/
mi
4444
3.16E+
14
113.6
1.50E+
14
27.0
1.45E+
14
27.0
1.83E+
14
44.9
4448
4.92E+
13
10.3
1.86E+
13
4.1
3.78E+
13
5.6
3.02E+
13
5.8
4457
1.01E+
14
5.0
7.65E+
12
2.9
1.42E+
13
2.6
2.88E+
13
3.3
4460
1.59E+
13
3.7
2.53E+
13
2.4
9.58E+
12
3.2
1.90E+
13
2.9
4463
9.40E+
12
2.4
2.16E+
11
1.4
8.39E+
11
2.4
2.29E+
12
1.9
4466
1.00E+
14
10.4
5.31E+
13
2.7
3.33E+
13
2.1
5.75E+
13
4.1
4469
2.54E+
13
15.9
8.39E+
11
4.0
1.13E+
13
6.2
8.80E+
12
7.0
4472
4.31E+
14
57.2
1.23E+
14
7.0
1.09E+
14
10.1
1.83E+
14
18.2
4475
1.12E+
14
5.6
3.31E+
13
5.4
1.43E+
13
0.0
4.42E+
13
3.9
4482
2.29E+
14
28.7
3.85E+
13
11.9
1.66E+
14
15.8
1.13E+
14
16.4
4490
3.77E+
14
33.3
1.52E+
14
9.2
6.99E+
13
5.7
1.76E+
14
13.2
4496
4.49E+
13
2.1
1.48E+
13
0.5
1.93E+
12
0.3
1.75E+
13
0.8
4499
1.77E+
14
5.4
7.74E+
13
2.0
7.79E+
13
6.0
9.82E+
13
3.8
4502
3.15E+
15
379.1
2.27E+
15
253.9
3.03E+
15
246.4
2.66E+
15
277.5
4506
3.33E+
15
105.7
2.04E+
15
28.9
2.13E+
15
40.3
2.33E+
15
47.9
4510
1.86E+
15
151.0
1.11E+
15
127.9
7.63E+
14
48.5
1.17E+
15
110.9
4513
5.02E+
13
4.0
2.89E+
14
7.7
7.25E+
13
1.6
1.80E+
14
5.2
4535
3.75E+
14
13.1
2.01E+
14
9.0
3.19E+
14
13.7
2.69E+
14
11.1
4538
2.80E+
15
182.1
4.92E+
14
17.6
1.04E+
15
33.7
1.12E+
15
56.1
4548
2.03E+
13
3.8
1.23E+
12
2.3
5.46E+
11
2.8
5.00E+
12
2.7
4554
1.15E+
14
14.2
2.65E+
14
6.7
7.34E+
13
3.2
1.81E+
14
7.3
4557
8.99E+
13
7.2
3.18E+
13
0.5
1.20E+
13
1.6
3.84E+
13
2.2
4566
5.29E+
13
2.6
4.20E+
13
1.5
4.23E+
12
1.4
3.39E+
13
1.7
4580
3.45E+
14
32.5
1.49E+
14
10.4
1.94E+
14
10.2
2.02E+
14
14.9
4583
8.45E+
14
48.5
5.28E+
14
29.6
5.44E+
14
25.0
5.98E+
14
32.2
Table
5.2.
Winter
EAA
Particle
Emission
Rates
and
PM­
10
Mass
Emission
Rates
from
Gasoline
Vehicles
Run
#
Phase
1
Phase
1
Phase
2
Phase
2
Phase
3
Phase
3
FTP
Composite
particle/
mi
mg/
mi
particle/
mi
mg/
mi
Particle/
mi
mg/
mi
particle/
mi
mg/
mi
4656
4.05E+
14
168.3
6.66E+
13
10.9
6.4E+
13
13.8
1.36E+
14
44.3
4659
1.17E+
14
23.9
2.95E+
12
1.4
6.5E+
12
1.6
2.75E+
13
6.1
4662
2.3E+
14
51.2
7.45E+
12
0.9
9.84E+
12
1.4
5.43E+
13
11.4
4665
1.31E+
15
283.6
6.73E+
13
8.9
2.37E+
13
16.0
3.09E+
14
67.0
4668
9.94E+
13
26.7
8.52E+
12
0.9
3.87E+
12
0.1
2.61E+
13
6.0
4671
3.37E+
13
7.0
9.95E+
12
1.8
3.44E+
12
0.9
1.31E+
13
2.7
4674
3.95E+
14
87.0
1.5E+
14
79.2
2.58E+
14
77.3
2.3E+
14
80.3
4677
3.32E+
13
30.4
5.59E+
12
4.2
6.29E+
12
12.2
1.15E+
13
11.8
4680
7.15E+
13
6.4
3.86E+
13
0.6
1.86E+
13
1.2
4E+
13
2.0
4683
1.08E+
14
21.2
2.87E+
13
6.2
1.68E+
13
5.7
4.18E+
13
9.2
4687
1.78E+
14
55.3
9.56E+
13
42.2
8.97E+
13
40.9
1.11E+
14
44.5
4690
5.54E+
13
8.9
7.01E+
12
0.1
8.58E+
12
1.9
1.75E+
13
2.4
4693
1.21E+
14
7.2
6.34E+
13
4.6
2.29E+
13
2.2
6.42E+
13
4.5
4695
1.52E+
15
231.7
9.78E+
14
163.7
1.13E+
15
221.0
1.13E+
15
193.6
4698
3.99E+
14
29.5
1.8E+
14
8.2
2.01E+
14
8.8
2.31E+
14
12.8
4704
2.87E+
14
12.9
8.18E+
14
29.0
4.63E+
14
20.6
6.1E+
14
23.4
4707
4.97E+
14
186.1
3.71E+
14
47.6
8.04E+
14
57.6
5.16E+
14
79.1
4711
4.44E+
14
97.6
1.43E+
14
11.9
2.81E+
14
18.6
2.42E+
14
31.3
57
4714
3.73E+
13
6.5
3.24E+
12
1.8
1.27E+
12
2.1
9.76E+
12
2.8
4718
5.14E+
14
263.3
5.7E+
14
32.5
5.53E+
14
59.8
5.54E+
14
88.0
4721
2.35E+
14
84.6
4.49E+
13
18.1
6.92E+
13
25.4
9.12E+
13
33.9
4725
3.84E+
14
659.3
2.4E+
14
270.7
3.3E+
14
414.0
2.94E+
14
390.0
4731
7.79E+
14
1362.7
3.56E+
14
1134.9
3.85E+
14
849.4
4.52E+
14
1103.7
4735
1.75E+
14
50.4
9.36E+
12
4.5
1.75E+
13
5.5
4.6E+
13
14.3
4739
6.58E+
14
711.0
3.31E+
14
494.3
4.13E+
14
428.0
4.21E+
14
520.9
4742
6.04E+
14
458.7
3.16E+
14
181.1
2.93E+
12
268.1
2.9E+
14
262.5
4745
5.23E+
13
20.3
8.18E+
13
20.2
4.4E+
13
14.8
6.54E+
13
18.8
4748
3.64E+
14
302.5
1.99E+
14
126.9
2.23E+
14
149.0
2.4E+
14
169.4
4752
1.57E+
13
5.2
8.9E+
11
0.8
1.06E+
12
0.2
4E+
12
1.6
4755
3.04E+
14
85.1
4.78E+
13
12.6
6.7E+
13
25.7
1.05E+
14
30.7
4758
1.14E+
14
126.4
1.34E+
14
244.3
9.35E+
13
134.4
1.19E+
14
189.8
4761
2.56E+
13
6.6
1.73E+
12
0.8
4.5E+
11
1.5
6.31E+
12
2.2
4763
1.91E+
14
56.6
1.32E+
14
35.9
8.75E+
13
39.1
1.32E+
14
41.1
4766
4.18E+
14
24.0
1.06E+
14
3.7
8.58E+
13
8.3
1.65E+
14
9.1
4769
1.07E+
13
6.2
3.27E+
12
1.2
5.57E+
12
5.2
5.45E+
12
3.3
4772
1.09E+
14
9.6
1.13E+
11
0.0
2.22E+
11
4.6
2.28E+
13
3.3
4775
1.86E+
13
6.3
9.31E+
12
1.4
2.63E+
12
2.6
9.39E+
12
2.7
4778
3.13E+
12
2.2
7.88E+
10
0.0
6.25E+
10
1.0
7.07E+
11
0.7
4782
2.42E+
13
17.4
1.43E+
13
3.1
2.18E+
13
4.4
1.84E+
13
6.4
4785
5.18E+
14
270.6
4.61E+
14
580.2
3.38E+
14
270.3
4.39E+
14
431.1
4788
8.58E+
14
110.3
3.18E+
14
20.5
2.01E+
14
27.2
3.98E+
14
40.9
4791
2.38E+
13
3.5
1.29E+
13
5.2
4.53E+
12
1.1
1.29E+
13
3.7
4794
5.6E+
14
95.6
9.53E+
13
18.9
3.45E+
13
14.4
1.75E+
14
33.5
4797
2.54E+
14
57.4
5.38E+
13
5.3
7.8E+
13
28.9
1.02E+
14
22.5
4800
4.55E+
13
14.9
2.89E+
13
8.8
3.88E+
13
10.6
3.5E+
13
10.6
4809
9.25E+
14
1177.1
3.24E+
14
109.0
5.54E+
14
738.0
5.11E+
14
501.2
4813
2.5E+
14
58.2
8.83E+
12
3.6
1.44E+
14
36.4
9.59E+
13
23.9
4816
5.86E+
14
325.1
2.23E+
14
141.8
2.6E+
14
119.0
3.09E+
14
173.5
4819
1.19E+
14
44.5
2.09E+
14
132.8
1.73E+
14
107.1
1.8E+
14
107.5
4822
7.73E+
13
30.9
4.96E+
13
6.6
3.62E+
13
6.3
5.16E+
13
11.6
4825
2.46E+
13
4.1
7.27E+
11
0.1
1.39E+
12
1.9
5.84E+
12
1.4
4828
3.75E+
13
10.0
1.42E+
13
0.6
1.46E+
12
0.2
1.55E+
13
2.4
4831
3E+
14
125.4
2.28E+
14
120.0
2.42E+
14
167.2
2.47E+
14
134.1
4834
1.76E+
14
278.9
2.38E+
14
653.4
3.14E+
14
514.4
2.46E+
14
537.7
4837
2.6E+
14
70.0
8.52E+
13
44.4
1.4E+
14
36.2
1.36E+
14
47.5
4840
4.55E+
14
190.8
1.38E+
14
32.1
2.1E+
14
41.3
2.23E+
14
67.4
4844
3.83E+
14
174.9
1.49E+
14
30.1
2.13E+
14
43.0
2.15E+
14
63.6
4847
6.97E+
14
385.2
2.84E+
14
156.2
2.8E+
14
2.1
3.69E+
14
161.4
4851
7.71E+
12
1.9
7.24E+
12
1.2
2.85E+
12
1.4
6.14E+
12
1.4
4854
9.86E+
12
4.0
4.79E+
12
1.3
7.65E+
11
1.9
4.74E+
12
2.0
4857
1.06E+
14
50.9
1.03E+
14
29.7
2.97E+
13
21.9
8.37E+
13
31.9
4860
8.23E+
13
29.7
5.74E+
13
14.9
5.76E+
13
13.0
6.26E+
13
17.5
4863
1.03E+
13
4.7
2.72E+
11
1.4
1.76E+
11
0.2
2.33E+
12
1.8
4876
1.81E+
13
23.7
2.02E+
12
1.9
3.93E+
11
5.3
4.92E+
12
7.4
4879
4.21E+
14
235.6
5.94E+
13
9.6
1.85E+
14
15.8
1.68E+
14
57.9
4882
1.61E+
14
53.9
5.15E+
14
117.2
1.92E+
14
35.3
3.53E+
14
81.7
4885
9.92E+
13
16.4
1.46E+
14
23.0
1.18E+
13
6.6
9.96E+
13
17.1
Table
5.3.
Winter
EAA
Particle
Emission
Rates
and
PM­
10
Mass
Emission
Rates
from
58
Diesel
Vehicles
Run
#
Phase
1
Phase
1
Phase
2
Phase
2
Phase
3
Phase
3
FTP
Composite
particle/
mi
mg/
mi
particle/
mi
mg/
mi
particle/
mi
mg/
mi
particle/
mi
mg/
mi
4891
4.22E+
14
99.8
3.37E+
14
69.0
2.37E+
14
ND
3.28E+
14
ND
4894
3.83E+
14
525.5
4.36E+
14
543.9
3.68E+
14
530.4
4.06E+
14
536.4
4897
3.7E+
14
602.9
3.06E+
14
229.3
2.95E+
14
354.9
3.16E+
14
341.2
4900
9.31E+
14
940.4
6.75E+
14
383.4
7.93E+
14
646.5
7.6E+
14
570.1
4903
1.02E+
15
866.6
8.4E+
14
407.3
7.94E+
14
642.7
8.65E+
14
566.7
4907
1.03E+
15
1129.1
1.09E+
15
789.5
7.13E+
14
1045.6
9.71E+
14
930.0
4910
6.45E+
14
940.7
4.79E+
14
724.7
4.06E+
14
830.8
4.93E+
14
798.5
4913
7.15E+
14
1120.7
2.4E+
14
416.7
4.1E+
14
528.4
3.85E+
14
593.1
4916
4.24E+
14
474.7
5.72E+
14
310.7
4.4E+
14
329.7
5.05E+
14
349.9
4919
4.82E+
14
388.9
5.94E+
14
136.8
4.77E+
14
281.4
5.39E+
14
228.5
4926
4.6E+
14
240.4
3.53E+
14
121.8
4.22E+
14
162.4
3.94E+
14
157.3
4929
5.31E+
14
1056.3
3.41E+
14
373.4
3.89E+
14
498.2
3.93E+
14
546.8
59
Particle
size
distributions.
Particle
size
distributions
were
taken
with
MOUDI
impactors.
During
the
summer
an
eight­
stage
impactor
with
a
final
stage
cutpoint
of
0.19
µ
m
was
used.
Considerable
difficulty
was
experienced
with
this
measurement.
High
emission
vehicles
tended
to
plug
the
impactor
while
low
emission
vehicles
provided
insufficient
mass
for
weighing.
The
plugging
was
determined
to
primarily
be
caused
by
the
Teflo
final
filter.
We
assume
that
the
very
small
particles
reaching
this
filter
penetrated
into
the
pores
and
plugged
them.
All
but
seven
size
distributions
were
rejected.
During
the
winter
a
ten­
stage
impactor
with
a
final
stage
cutpoint
of
0.068
µ
m
was
used.
This
impactor
has
finer
particle
size
resolution
than
the
one
used
in
the
summer.
Thus,
a
lower
percentage
of
the
mass
was
deposited
on
the
final
filter.
In
addition,
the
final
filter
was
switched
from
a
2
µ
m
pore
size
Teflo
filter
to
a
Pallflex
T60A20
filter.
Results
were
much
better.

The
impactors
were
operated
continuously
over
the
entire
FTP,
except
for
a
few
runs
in
the
summer
when
the
run
was
terminated
early
due
to
plugging.
Examination
of
size
distributions
for
the
low
PM
emitters
indicated
some
variability
in
the
impactor
stage
blank
values,
with
a
tendency
towards
decreasing
blanks
throughout
the
winter
portion
of
the
study.
Therefore,
subjective
judgment
was
necessary
in
choosing
the
blank
correction
value
for
each
sample.
In
all
cases,
a
blank
was
used
that
resulted
in
no
impactor
stages
with
a
negative
mass
change.
Uncertainty
in
the
blank
adds
uncertainty
to
the
overall
particle
size
distributions,
especially
for
the
low
emission
vehicles.
This
should
be
kept
in
mind
when
examining
the
following
data.

Figure
5.6
shows
two
typical
particle
size
distributions
from
diesel
vehicles
and
two
from
smoking
gasoline
vehicles
from
the
winter
study.
In
all
cases
the
maximum
collected
mass
is
on
the
0.22
µ
m
cutpoint
stage.
Figure
5.7
shows
distributions
from
four
gasoline
winter
vehicles.
The
maximum
collected
mass
is
on
either
the
0.12
or
the
0.22
cutpoint
stage.
There
is
some
indication
of
bimodal
distributions
for
the
gasoline
vehicles,
with
the
second
mass
mode
at
larger
particle
diameters.
Most
of
the
gasoline
vehicle
size
distributions
are
for
older
vehicles,
since
there
was
insufficient
mass
collected
from
the
newer
vehicles
to
provide
accurate
impactor
size
distributions.
It
must
be
remembered
that
because
the
impaction
foils
were
not
coated
with
grease,
particle
bounce
and
particle
loss
in
the
impactor
were
probably
significant
(
Lawson,
1980).
Coated
impaction
foils
were
not
used
since
the
chance
of
losing
mass
during
handling
foils
is
high,
and
collected
sample
mass
was
low.

Tables
5.4
and
5.5
give
the
impactor
results
for
the
summer
and
winter,
respectively.
The
mass
collected
by
the
impactor
is
listed
in
column
3.
The
corresponding
PM­
10
mass
collected
on
Teflo
filters
is
given
for
comparison.
The
impactor
samples
were
drawn
directly
from
the
dilution
tunnel;
i.
e.,
a
PM­
10
cyclone
was
not
used.
Therefore,
it
is
possible
for
the
impactors
to
collect
more
mass
than
the
filter
samples
that
were
collected
after
the
PM­
10
cyclone.
To
facilitate
the
mass
comparison,
any
mass
collected
on
the
first
stage,
which
has
a
cutoff
of
21.9
µ
m,
was
deleted
from
the
total
mass
reported
in
the
Tables.
On
the
other
hand,
some
losses
of
particles
within
impactors
is
considered
unavoidable.
Also,
since
the
last
five
stages
operate
at
reduced
pressure,
some
semivolatile
material
may
be
lost.
Thus,
it
is
common
to
have
the
sum
of
the
mass
collected
by
impactors
to
be
less
than
the
mass
collected
on
filters.
The
average
ratio
of
impactor
mass
60
to
filter
mass
for
the
gasoline
vehicles
was
0.82
in
the
summer
and
0.89
in
the
winter.
The
average
ratio
was
higher
for
the
winter
smoking
vehicles,
0.99,
but
lower
for
the
diesels
0.80.
Overall,
the
agreement
between
the
impactor
and
filter
mass
is
considered
acceptable.

Plots
of
cumulative
mass
percentage
versus
the
log
of
the
particle
size
were
prepared
for
each
sample.
Interpolation
between
points
was
used
to
determine
the
mass
median
diameter
(
MMD)
of
each
sample
and
the
percent
of
mass
with
a
particle
size
less
than
2.5
µ
m.
These
are
listed
in
the
last
two
columns
of
Tables
5.4
and
5.5.
Gasoline
vehicles
had
the
lowest
average
MMD,
0.12
µ
m
in
the
winter
and
0.15
µ
m
in
the
summer.
The
summer
value
drops
to
0.12
µ
m
if
the
distribution
with
a
MMD
of
0.32
is
eliminated.
Smoking
vehicles
and
diesels
had
the
same
average
MMD,
0.18
µ
m.
The
diesel
MMD
is
similar
to
that
reported
by
others
using
EAA
measurements
(
Groblicki
and
Begeman,
1979),
but
lower
than
the
0.3
µ
m
diameter
reported
by
Chan
and
Lawson
(
1981),
who
used
coated
impaction
stages
in
cascade
impactors.
Smaller
MMDs
have
been
reported
for
catalyst
equipped
vehicles.
For
example,
Groblicki
(
1976)
reported
MMDs
of
0.02
to
0.03
µ
m
for
oxidation
catalyst
vehicles
equipped
with
air
pumps.
The
PM
from
these
vehicles
was
dominated
by
sulfate.
That
was
not
the
case
for
the
vehicles
in
this
study.

The
fraction
of
the
mass
contained
in
particles
smaller
than
2.5
µ
m
is
relevant
to
the
new
PM
NAAQS.
Tables
5.4
and
5.5
show
that
on
average
92
and
91
percent
of
the
mass
from
gasoline
vehicles
was
smaller
than
2.5
µ
m.
For
smoking
and
diesel
vehicles
the
percent
less
than
2.5
µ
m
increased
to
an
average
of
97
and
98%,
respectively.
These
results
can
be
compared
to
those
reviewed
by
Energy
and
Environmental
Analysis,
Inc.
(
1985).
They
reported
that
the
PM
mass
exhaust
emissions
from
light­
duty
vehicles
operated
on
unleaded
fuel
had
an
average
of
89%
of
the
mass
smaller
than
2.0
µ
m,
while
diesels
had
90%
of
the
mass
smaller
than
2.0
µ
m.
Durbin
et
al.
(
1998)
measured
particle
size
distributions
from
smoking
gasoline
and
diesel
vehicles.
They
found
that,
on
average,
91.8%
of
the
mass
was
smaller
than
2.5
µ
m.

Table
5.4.
Summer
Impactor
Results
Run
No.
Vehicle
Type
Impactor
Total
Mass,
mg
Filter
Total
Mass,
mg
Mass
Ratio
Impactor/
Filter
MMD,
µ
m
Fraction
<
2.5
µ
m
4545
G
6.23
8.89
0.70
0.18
0.97
4576
G
3.19
NA
NA
0.13
0.98
4560
G
2.40
2.77
0.87
0.13
0.97
4527
G
2.35
3.04
0.77
0.12
0.93
4542
G
0.31
0.26
1.19
0.08
0.91
4533/
35
G
0.27
0.47
0.58
0.12
0.86
4523
G
0.16
NA
NA
0.32
0.82
average
0.82
0.15
0.92
61
Table
5.5.
Winter
Impactor
Results
Run
No.
Vehicle
Type
Impactor
Total
Mass,
mg
Filter
Total
Mass,
mg
Mass
Ratio
Impactor/
Filter
MMD,
µ
m
Fraction
<
2.5
µ
m
4907
D
8.89
13.75
0.65
0.21
0.99
4929
D
6.34
6.23
1.02
0.16
0.99
4894
D
6.08
7.32
0.83
0.21
0.99
4910
D
5.56
7.22
0.77
0.21
0.99
4897
D
3.92
5.40
0.73
0.21
0.90
4926
D
1.87
2.39
0.78
0.11
0.98
4891
D
1.03
NA
NA
0.14
0.99
average
0.80
0.18
0.98
4809
S
10.64
9.80
1.09
0.20
0.99
4695
S
5.71
5.08
1.12
0.17
0.89
4847
S
5.07
NA
NA
0.17
0.99
4748
S
4.47
4.64
0.96
0.18
0.97
4816
S
4.09
4.76
0.86
0.17
0.99
4758
S
3.79
4.20
0.90
0.21
0.99
4819
S
2.33
2.35
0.99
0.18
0.98
average
0.99
0.18
0.97
4873
G
2.18
2.67
0.82
0.13
0.93
4879
G
1.80
2.10
0.86
0.17
0.92
4882
G
1.22
1.75
0.70
0.10
0.97
4711
G
1.10
1.01
1.09
0.11
0.87
4837
G
1.02
1.22
0.84
0.12
0.96
4763
G
0.97
1.06
0.91
0.12
0.90
4701
G
0.96
0.89
1.08
0.11
0.77
4794
G
0.92
1.03
0.89
0.12
0.90
4860
G
0.53
0.46
1.15
0.07
0.92
4797
G
0.50
0.73
0.69
0.12
0.97
4885
G
0.27
0.37
0.72
0.07
0.95
4863
G
0.17
0.04
4.64
0.17
0.91
average
0.891
0.12
0.91
1)
Run
4863
omitted
from
average
62
63
64
65
66
67
68
69
6.
Chemical
Composition
of
the
Particulate
Matter
Carbon.
The
bulk
of
the
PM
mass
is
expected
to
be
carbonaceous
material.
The
TOR
carbon
analysis
separates
that
material
into
"
organic"
(
OC)
and
"
elemental"
carbon
(
EC).
This
separation
is
important
for
multiple
reasons.
First,
EC
absorbs
light
efficiently,
and
thus
has
a
much
greater
impact
on
visibility
reduction
in
the
atmosphere
than
organic
carbon.
Second,
tracer
compounds
such
as
the
PAHs
discussed
later
are
associated
with
the
organic
carbon
fraction.
It
may
be
instructive
to
examine
their
composition
in
terms
of
their
fraction
of
the
organic
matter
rather
than
the
total
PM.
Third,
these
same
measurements
are
made
on
atmospheric
PM
samples.
Therefore,
the
OC/
EC
split
may
provide
some
information
for
source
apportionment.
Finally,
the
fraction
of
OC
may
qualitatively
provide
information
as
to
the
impact
of
oil
consumption
on
vehicular
PM
emissions.
It
may
be
possible
to
refine
the
estimate
regarding
oil
consumption
further
by
examining
the
temperature
at
which
the
OC
evolves
in
the
analytical
procedure.

For
the
summer
portion
of
the
study,
OC
and
EC
were
measured
for
each
FTP
phase
from
38
vehicles,
six
1971­
80,
seven
1981­
85,
eleven
1986­
90,
four
1991­
96,
six
smokers,
and
four
diesels.
The
first
step
in
analyzing
the
data
was
to
determine
what
fraction
of
the
total
mass
is
accounted
for
by
the
carbon.
Figure
6.1
shows
the
comparison
between
total
carbon
and
mass
for
all
vehicles.
The
plot
includes
points
representing
each
phase
of
the
FTP.
The
regression
line
has
a
slope
of
0.97,
and
an
R2
of
0.71.
The
near
equivalence
between
total
carbon
and
mass
indicates
a
problem
in
the
data.
Carbon
can
not
constitute
all
the
mass,
since
carbonaceous
material
includes
other
elements
such
as
hydrogen,
oxygen,
nitrogen,
and
sulfur,
and
some
non­
carbonaceous
material
is
present
as
well.
In
fact,
it
is
generally
assumed
that
the
organic
carbon
measurement
for
a
source
sample
should
be
multiplied
by
a
factor
of
1.2
to
account
for
the
other.
Inspection
of
Figure
6.1
shows
a
set
of
points
that
have
much
higher
carbon
mass
than
total
mass
which
are
significantly
influencing
the
regression.
Since
most
of
these
points
have
a
relatively
high
mass,
it
was
decided
to
examine
the
mass
vs.
total
carbon
relationship
separately
for
the
non­
smoking
gasoline
vehicles,
the
smokers,
and
the
diesels.
The
results
are
shown
in
Figures
6.2­
6.4.
The
linear
regression
for
the
non­
smoking
gasoline
vehicles
(
Figure
6.2
­
one
very
high
mass
emission
rate
point
removed)
has
a
slope
of
0.77
and
an
R2
of
0.87.
It
is
shown
below
that
approximately
80%
of
the
total
carbon
from
the
summer
samples
is
organic
material.
Multiplying
this
fraction
of
the
total
carbon
by
the
1.2
factor,
shows
that
carbonaceous
material
is
89%
of
the
total
mass.
This
relationship
was
examined
for
each
phase
of
the
FTP
for
the
non­
smoking
vehicles
as
well.
The
slopes
of
the
linear
regressions
for
phases
1­
3,
respectively,
were
0.79,
0.66,
and
0.80,
while
the
R2
values
were
0.84,
0.87
and
0.87.
Figure
6.3
shows
the
results
for
the
smoking
vehicles.
The
average
values
are
much
higher
than
for
the
non­
smoking
vehicles,
and
there
is
a
suggestion
that
there
are
two
sets
of
points,
one
set
with
a
relatively
good
agreement
between
carbon
and
mass,
and
one
where
the
total
carbon
is
much
higher
than
the
total
mass.
The
diesels,
Figure
6.4,
have
consistently
higher
total
carbon
than
mass.

Figure
6.5
shows
the
relationship
between
total
carbon
and
PM­
10
mass
for
the
winter
non­
smoking
gasoline
vehicles.
The
slope
of
0.76
is
similar
to
that
obtained
for
the
summer
samples
while
the
R2
of
0.995
is
better
(
the
R2
was
0.81
before
one
outlier
was
70
removed).
The
same
relationship
for
the
smoking
vehicles
had
a
slope
of
0.78
and
an
R2
of
0.99
with
the
two
highest
emission
rate
points
removed.
Inclusion
of
the
two
points
gave
a
slope
of
0.53
with
an
R2
of
0.98.
The
diesel
vehicles
had
a
slope
of
0.45
and
an
R2
of
0.63.
There
was
considerable
scatter
in
the
highest
emission
rate
values.
Considering
only
those
emission
rates
below
1000
mg/
mi
resulted
in
a
slope
of
0.58
and
an
R2
of
0.89.
If
the
regression
was
forced
through
zero,
the
slope
increased
to
0.71
with
an
R2
of
0.84.
Overall,
the
winter
data
are
consistent,
indicating
that
most
of
the
PM­
10
mass
is
carbonaceous
material.

The
cause
of
the
poor
agreement
between
total
carbon
and
PM­
10
mass
for
the
summer
diesel
samples
and
some
of
the
summer
smoking
vehicle
samples
has
not
been
found.
Possible
explanations
include
inhomogeneous
filter
deposits
for
some
heavily
loaded
samples
or
undetected
leaks
for
some
heavily
loaded
PM­
10
filters.
The
former
explanation
is
not
a
concern,
since
the
main
purpose
of
the
carbon
analysis
was
to
provide
the
ratio
of
OC
and
EC.
Inhomogeneous
deposits
should
not
affect
this
measurement.
Possible
leaks
associated
with
the
measurement
of
PM­
10
mass
are
of
concern.
However,
there
is
no
other
indication
that
these
occurred.

The
split
between
OC
and
EC
for
each
sample
was
examined
as
well.
Results
are
given
in
Table
6.1
for
the
summer
samples
and
Table
6.2
for
the
winter
samples.
The
samples
have
been
grouped
by
vehicle
type
(
gasoline,
smokers
and
diesel).
For
source
apportionment
purposes,
DRI
will
group
gasoline
vehicles
together
based
on
their
PM
emission
rates,
rather
than
their
model
year.
The
emission
rate
categories
used
are
low
(
L),
medium
low
(
ML),
medium
(
M)
and
high
(
H).
The
third
column
of
Tables
6.1
and
6.2
(
ID)
indicates
the
source
apportionment
grouping,
with
the
number
referring
to
the
corresponding
pooled
sample
for
PAH.
D
and
S
refer
to
diesel
and
smoker,
respectively.
Thus,
an
ID
of
H2
indicates
that
the
vehicle
is
used
in
the
second
pooled
high
gasoline
PM
emitter
sample.
The
split
between
OC
and
EC
for
FTP
phase
1
summer
samples
is
presented
graphically
in
Figure
6.6.
The
solid
bar
is
the
percent
EC.
Non­
smoking
gasoline
vehicles
are
indicated
by
the
open
OC
bar,
the
smoking
gasoline
vehicles
with
the
horizontal
stripped
OC
bar,
and
the
diesels
with
the
diagonal
stripped
OC
bar.
Within
these
categories
the
vehicles
are
arranged
in
order
of
increasing
FTP
emissions
from
left
to
right.
There
is
clearly
a
trend
towards
a
higher
OC
fraction
with
increasing
emission
rate
for
the
gasoline
vehicles.
The
smoking
gasoline
vehicles
have
relatively
high
OC
fractions,
with
the
two
highest
emitters
having
more
than
96%
of
the
carbon
as
"
organic"
material.
The
diesels,
on
the
other
hand,
have
relatively
low
OC
fractions.
Figure
6.7
shows
similar
trends
for
the
winter
FTP
phase
1
samples.
Unlike
the
summer
samples,
two
of
the
highest
gasoline
PM­
10
emitters
have
relatively
high
EC
fractions.
Figures
6.8
and
6.9
plot
the
data
for
summer
FTP
phases
2
and
3.
The
phase
2
OC
fraction
is
higher
than
that
of
phase
1
for
many
of
the
vehicles.
Results
were
similar
for
the
winter.
The
FTP
phase
difference
in
OC
fraction
is
summarized
in
Figure
6.10,
which
gives
the
average
OC
71
fraction
for
each
of
the
vehicle
categories
by
FTP
phase.
The
1981­
85,
1986­
90,
1991­
96,
and
diesel
vehicles
clearly
have
lower
average
percent
OC
in
phase
1
than
phase
2.
This
is
not
the
case
for
the
older
gasoline
vehicles
and
the
smoking
gasoline
vehicles.

The
TOR
carbon
analysis
method
measures
the
OC
evolved
sequentially
at
four
temperatures;
120,
250,
450,
and
550
°
C.
The
percentage
of
the
total
OC
evolved
at
each
of
these
temperatures
was
determined
for
each
FTP
phase
for
all
of
the
winter
samples.
Volatilization
of
organic
carbon
under
conditions
that
have
been
shown
to
remove
motor
oil
has
been
used
to
estimate
the
contribution
of
motor
oil
to
the
PM
mass.
While
motor
oil
has
not
been
characterized
using
TOR,
differences
in
the
percent
OC
removed
at
a
given
temperature
between
vehicles
might
indicate
a
difference
in
the
motor
oil
contribution.
Consistent
differences
between
vehicle
types
could
also
be
used
to
strengthen
the
differences
in
profiles
for
source
apportionment.
Figures
6.11­
6.13
give
the
results.
Vehicles
are
arranged
in
the
same
order
as
in
Figure
6.7.
G,
S,
and
D
on
the
xaxis
stand
for
gasoline,
smoker,
and
diesel
vehicles,
respectively
(
there
are
8
smokers
and
11
diesels).
For
FTP
phase
1
(
Figure
6.11),
more
than
50%
of
the
OC
is
removed
at
250
°
C
for
all
the
vehicles.
Better
than
90%
of
the
OC
is
removed
at
250
°
C
for
the
vehicles
identified
as
smokers.
In
contrast,
approximately
80%
of
the
OC
is
removed
at
250
°
C
for
the
diesels.
For
FTP
phase
2
(
Figure
6.12),
the
gasoline
vehicles
tend
to
have
a
higher
percentage
of
the
OC
removed
by
250
°
C
.
Phase
3
(
Figure
6.13)
is
similar
to
phase
1.
The
large
vehicle­
to­
vehicle
variability
suggests
that
the
OC
removal
temperature
will
not
be
very
useful
in
distinguishing
between
these
three
source
types.
The
results
also
suggest
that
motor
oil
could
be
a
significant
contributor
to
the
OC
fraction
of
the
PM
for
all
the
vehicles.

Sulfate
and
nitrate
emission
rates.
Table
6.3
gives
individual
and
average
FTP
sulfate
and
nitrate
emission
rates
by
vehicle
category
for
the
summer
study.
Minimum
detection
limits
for
nitrate
and
sulfate
were
approximately
0.04
mg/
mi.
Only
21
of
117
filter
extracts
had
nitrate
concentrations
in
excess
of
the
measurement
uncertainty.
In
contrast,
101
of
the
117
filter
extracts
had
sulfate
concentrations
greater
than
the
measurement
uncertainty.
Only
one
vehicle,
a
1989
diesel,
had
an
FTP
nitrate
emission
rate
in
excess
of
1
mg/
mi.
The
highest
nitrate
FTP
emission
rate
for
a
gasoline
vehicle
was
0.17
mg/
mi.
Sulfate
FTP
emission
rates
were
also
generally
low,
with
only
one
gasoline
vehicle
having
an
emission
rate
in
excess
of
1
mg/
mi.
In
contrast,
all
four
diesel
vehicles
tested
had
sulfate
emission
rates
of
1
mg/
mi
or
higher.
The
1989
diesel
vehicle
with
significant
nitrate
emissions
also
had
high
sulfate
emissions
(
17.8
mg/
mi).
Sulfate
emissions
for
this
vehicle
were
high
in
both
phases
1
and
3,
whereas
nitrate
was
high
only
in
phase
1.
FTP
results
for
the
winter
study
are
given
in
Table
6.4.
No
nitrate
emission
rates
are
reported
for
the
smokers
or
the
diesels
because
the
filters
used
for
these
vehicles
had
a
high
nitrate
background.
For
the
gasoline
vehicles,
nitrate
emission
rates
were
very
low,
averaging
less
than
0.1
mg/
mi.
The
highest
nitrate
emission
rate
was
0.26
mg/
mi.
Overall,
it
is
concluded
that
nitrate
is
not
a
significant
contributor
to
PM
emissions.
The
gasoline
vehicle
winter
sulfate
emission
rate
averaged
less
than
0.6
mg/
mi
for
each
vehicle
category,
with
only
two
vehicles
having
an
emission
rate
greater
than
1.0
mg/
mi.
The
smokers
had
higher
sulfate
emission
rates
in
the
winter
than
summer,
with
5
of
the
8
vehicles
exceeding
1.0
mg/
mi.
Diesels
also
had
higher
sulfate
emission
rates
than
the
gasoline
vehicles,
averaging
2.74
72
mg/
mi.

Table
6.1.
OC
and
EC
Percentages
by
FTP
Phase
for
Each
Summer
Vehicle
Vehicle
Run
MY
ID1
Ph1
%
OC
Ph1
%
EC
Ph2
%
OC
PH2
%
EC
PH3
%
OC
Ph3
%
EC
FTP
PM,
mg/
mi
1971­
80
4510
1975
H3
88.10
11.88
96.30
3.69
83.08
16.96
110.9
1971­
80
4506
1976
H1
90.83
9.17
96.84
3.23
81.30
18.70
47.9
1971­
80
4551
1976
90.86
9.14
72.82
27.18
74.84
25.26
254.6
1971­
80
96078
1976
76.75
23.25
92.54
7.46
81.57
18.51
21.2
1971­
80
4538
1978
86.29
13.71
90.54
9.46
86.84
13.21
56.1
1971­
80
96105
1978
74.61
25.39
81.33
18.67
92.53
7.48
122.4
Mean
82.13
17.87
84.31
15.69
83.95
16.11
113.56
Std
dev
6.69
6.69
7.86
7.86
6.53
6.56
89.16
1981­
85
96047
1981
M3
63.49
36.51
60.32
39.68
48.46
51.54
21.8
1981­
85
96081
1981
47.40
52.60
91.87
8.47
78.79
20.94
3.2
1981­
85
96074
1983
H1
48.88
51.12
43.90
56.10
45.89
54.24
42.9
1981­
85
96186
1983
H2
94.41
5.59
94.86
5.20
93.44
6.56
68.9
1981­
85
96090
1983
M3
48.71
51.29
77.16
23.09
49.27
50.73
32.6
1981­
85
4479
1984
M1
92.10
7.94
90.94
9.17
81.63
18.49
28.3
1981­
85
4560
1985
H3
89.65
10.36
79.74
20.26
81.24
18.76
99.8
Mean
81.22
18.80
85.67
14.43
76.39
23.64
57.39
Std
dev
18.85
18.84
7.41
7.45
16.41
16.40
29.11
1986­
90
4482
1986
M1
38.79
61.27
89.68
10.32
48.16
51.84
16.4
1986­
90
4490
1986
M2
95.41
4.59
85.58
14.42
85.58
14.17
13.2
1986­
90
96010
1986
27.84
72.16
71.47
25.84
0.00
109.86
1.3
1986­
90
96164
1986
87.16
12.84
95.19
4.81
85.88
14.12
766.3
1986­
90
96071
1987
L2
58.82
41.03
91.42
8.46
75.01
24.73
6.4
1986­
90
96065
1987
M2
41.49
58.51
59.07
40.93
42.70
57.30
27.8
1986­
90
96068
1988
L1
57.83
41.90
93.88
6.12
48.17
51.83
2.1
1986­
90
96183
1989
H2
76.24
23.76
76.88
23.19
70.33
29.67
52.5
1986­
90
96180
1989
93.53
6.47
95.29
4.71
92.29
7.71
38.9
1986­
90
96044
1990
L3
65.98
34.02
94.32
5.46
66.66
33.34
9.7
1986­
90
96102
1990
L3
73.30
26.75
90.20
9.80
57.13
42.70
9.2
Mean
77.26
22.75
89.17
10.79
71.60
28.36
27.58
Std
dev
10.11
10.11
7.35
7.42
12.88
12.83
18.75
1991­
96
96143
1991
L1
69.78
30.41
87.54
12.28
87.69
12.13
1.8
1991­
96
4554
1992
L2
43.89
56.11
58.30
41.70
63.21
37.22
7.3
1991­
96
4487
1992
49.30
50.70
82.80
17.20
51.34
45.40
0.9
1991­
96
4524
1994
L1
69.37
30.80
69.54
30.46
90.02
11.23
2.2
Mean
58.09
42.01
74.54
25.41
73.07
26.50
3.04
Std
dev
11.65
11.56
11.47
11.52
16.36
15.10
2.51
Smoker
96062
1986
S1
97.62
2.38
99.18
0.82
97.27
2.74
442.5
Smoker
96201
1988
S1
96.85
3.15
99.53
0.47
97.29
2.71
551.5
Smoker
96189
1983
S2
92.07
7.93
60.98
39.01
89.03
10.97
180.6
Smoker
96192
1984
S2
77.07
22.93
86.09
13.90
37.73
62.27
144.3
Smoker
4502
1971
S3
93.29
6.71
92.71
7.31
95.15
4.85
94.0
Smoker
96161
1966
S3
83.53
16.48
89.11
10.85
79.44
20.58
277.4
Mean
86.49
13.51
82.22
17.77
75.34
24.67
174.10
Std
dev
6.61
6.61
12.49
12.49
22.42
22.42
67.09
Diesel
4593
1993
D1
29.67
70.33
45.18
54.80
20.65
79.37
282.4
Diesel
4605
1995
D1
33.76
66.24
66.79
33.25
55.03
44.99
143.3
Diesel
4595
1994
D2
41.04
58.96
55.96
44.04
43.62
56.38
288.7
Diesel
4589
1989
D3
67.60
32.40
66.79
33.21
57.80
42.20
3988.2
Mean
43.02
56.98
58.68
41.33
44.28
55.74
1175.63
Std
dev
14.77
14.76
8.96
8.94
14.64
14.64
1624.87
1)
L,
M,
H,
S,
D
refer
to
low,
medium,
high
PM
emitters,
smoker,
and
diesel.
The
associated
number
is
a
sample
id.

Table
6.2.
OC
and
EC
Percentages
by
FTP
Phase
for
Each
Winter
Vehicle
73
Vehicle
Run
MY
ID1
Ph1
%
OC
Ph1
%
EC
Ph2
%
OC
Ph2
%
EC
Ph3
%
OC
Ph3
%
EC
FTP
PM,
mg/
mi
1971­
80
97056
1971
H1
43.2
56.8
87.7
12.3
42.0
58.0
276.0
1971­
80
97053
1975
M1
71.7
28.3
78.0
22.0
48.6
51.4
103.0
1971­
80
97158
1978
H2
66.3
33.7
33.6
66.4
54.1
45.9
124.0
1971­
80
97161
1978
ML2
58.9
41.1
66.9
33.1
72.5
27.5
31.3
1971­
80
97013
1979
M1
82.3
17.7
96.6
3.4
97.4
2.6
61.6
1971­
80
97096
1980
M3
64.0
36.0
64.8
35.2
80.1
19.9
56.9
Mean
64.4
35.6
71.3
28.7
65.8
34.2
108.8
Std
dev
13.1
13.1
22.1
22.1
21.2
21.2
88.5
1981­
85
97165
1981
M3
69.0
31.0
93.5
6.5
93.8
6.2
41.9
1981­
85
97025
1983
H1
10.2
89.8
23.2
76.8
26.7
73.3
152.0
1981­
85
97062
1984
ML1
51.3
48.7
70.5
29.5
51.5
48.5
40.3
1981­
85
97115
1984
ML1
40.0
60.0
74.3
25.7
37.0
63.0
21.6
1981­
85
97112
1984
M2
38.5
61.5
65.3
34.8
21.3
78.7
55.1
1981­
85
97059
1985
M2
35.6
64.4
74.2
25.8
50.6
49.4
54.1
Mean
40.7
59.3
66.8
33.2
46.8
53.2
60.8
Std
dev
19.4
19.4
23.4
23.4
26.1
26.1
46.3
1986­
90
97072
1986
L1
74.0
26.0
79.0
21.0
67.1
32.9
9.3
1986­
90
97044
1987
L2
39.1
60.9
83.3
16.7
42.3
57.7
8.1
1986­
90
97016
1989
L1
69.2
30.8
92.9
7.1
75.0
25.0
9.7
1986­
90
97047
1989
M1
77.8
22.2
92.9
7.1
90.7
9.3
73.9
1986­
90
97099
1990
ML1
48.0
52.0
51.0
49.0
29.6
70.4
44.0
1986­
90
97137
1990
L2
41.3
58.7
79.6
20.4
70.0
30.0
5.2
Mean
58.2
41.8
79.8
20.2
62.5
37.5
25.0
Std
dev
17.4
17.4
15.4
15.4
22.5
22.5
28.0
1991­
96
97168
1991
L2
70.6
29.4
90.0
10.0
91.7
8.3
8.2
1991­
96
97075
1993
ML2
26.2
73.8
55.0
45.0
72.7
27.3
10.8
1991­
96
97134
1993
M3
58.0
42.0
64.8
35.2
44.6
55.4
51.3
1991­
96
97127
1993
L1
49.7
50.3
83.0
17.0
75.0
25.0
6.3
1991­
96
97109
1994
H2
92.0
8.0
97.6
2.4
76.5
23.5
123.0
Mean
59.3
40.7
78.1
21.9
72.1
27.9
39.9
Std
dev
24.4
24.4
17.7
17.7
17.1
17.1
50.0
Smoker
97093
1971
S1
55.7
44.3
67.8
32.2
70.0
30.0
1350.0
Smoker
97146
1975
S3
94.1
5.9
96.4
3.6
96.2
3.8
179.0
Smoker
97050
1976
S1
74.5
25.5
69.6
30.4
47.7
52.3
836.0
Smoker
97081
1977
S2
83.3
16.7
87.6
12.4
82.4
17.6
214.0
Smoker
97103
1981
S2
85.7
14.3
96.6
3.4
94.1
5.9
887.0
Smoker
97084
1983
S3
86.9
13.1
98.0
2.0
95.5
4.5
185.0
Smoker
97152
1985
S3
82.9
17.1
92.0
8.0
88.2
11.8
238.0
Smoker
97087
1989
S2
94.1
5.9
98.6
1.4
93.7
6.3
704.0
Mean
82.2
17.8
88.3
11.7
83.5
16.5
574.1
Std
dev
12.4
12.4
12.7
12.7
16.9
16.9
436.8
Diesel
97251
1979
D1
44.2
55.8
31.2
68.8
55.6
44.4
933.0
Diesel
97242
1979
D1
45.0
55.0
60.6
39.4
45.2
54.8
763.0
Diesel
97245
1979
D4
21.9
78.1
14.4
85.6
14.4
85.6
485.0
Diesel
97248
1980
D3
25.5
74.5
14.6
85.4
23.1
76.9
462.0
Diesel
97233
1981
D4
13.3
86.7
16.0
84.0
16.9
83.1
371.0
Diesel
97239
1981
D5
13.2
86.8
18.2
81.8
16.9
83.1
194.0
Diesel
97267
1982
D3
64.9
35.1
16.4
83.6
17.1
82.9
607.0
Diesel
97254
1982
D2
57.6
42.4
69.2
30.8
68.1
31.9
504.0
Diesel
97257
1982
D5
58.5
41.5
80.0
20.0
73.6
26.4
539.0
Diesel
97270
1983
D2
15.1
84.9
16.8
83.2
15.1
84.9
749.0
Diesel
97261
1984
D5
44.8
55.2
37.8
62.2
35.4
64.6
306.0
mean
36.7
63.3
34.1
65.9
34.7
65.3
537.5
Std
dev
19.5
19.5
24.6
24.6
22.5
22.5
215.8
1)
L,
M,
H,
S,
D
refer
to
low,
medium,
high
PM
emitters,
smoker,
and
diesel.
The
associated
number
is
a
sample
id.

Table
6.3.
Summer
Sulfate
and
Nitrate
FTP
Emission
Rates
Vehicle
RUN
#
MY
ID
NO3
SO4
mg/
mi
mg/
mi
1971­
80
4510
1975
H3
0.060
0.382
1971­
80
4506
1976
H1
0.050
0.100
1971­
80
4551
1976
0.170
8.693
1971­
80
96078
1976
0.041
0.348
1971­
80
4538
1978
0.044
0.262
1971­
80
96105
1978
0.159
0.207
74
mean
0.087
1.665
min
0.041
0.100
max
0.170
8.693
1981­
85
96047
1981
M3
0.000
0.068
1981­
85
96081
1981
0.000
0.038
1981­
85
96074
1983
H1
0.041
0.144
1981­
85
96186
1983
H2
0.000
0.121
1981­
85
96090
1983
M3
0.027
0.206
1981­
85
4479
1984
M1
0.000
0.172
1981­
85
4560
1985
H3
0.000
0.139
mean
0.010
0.127
min
0.000
0.038
max
0.041
0.206
1986­
90
4482
1986
M1
0.000
0.192
1986­
90
4490
1986
M2
0.000
0.050
1986­
90
96010
1986
0.000
0.022
1986­
90
96164
1986
0.119
0.483
1986­
90
96071
1987
L2
0.000
0.108
1986­
90
96065
1987
M2
0.000
0.041
1986­
90
96068
1988
L1
0.000
0.072
1986­
90
96183
1989
H2
0.054
0.141
1986­
90
96180
1989
0.000
0.330
1986­
90
96044
1990
L3
0.000
0.120
1986­
90
96102
1990
L3
0.000
0.095
mean
0.016
0.150
min
0.000
0.022
max
0.119
0.483
1991­
96
96143
1991
L1
0.000
0.052
1991­
96
4554
1992
L2
0.000
0.101
1991­
96
4487
1992
0.000
0.003
1991­
96
4524
1994
L1
0.000
0.046
mean
0.000
0.051
min
0.000
0.003
max
0.000
0.101
Smoker
96062
1986
S1
0.000
0.751
Smoker
96201
1988
S1
0.000
0.505
Smoker
96189
1983
S2
0.000
0.091
Smoker
96192
1984
S2
0.051
0.322
Smoker
96161
1966
S3
0.066
0.180
Smoker
4502
1971
S3
0.138
0.209
mean
0.043
0.343
min
0.000
0.091
max
0.138
0.751
Diesel
4593
1993
D1
0.084
1.057
Diesel
4605
1995
D1
0.064
1.449
Diesel
4595
1994
D2
0.597
2.124
Diesel
4589
1989
D3
4.311
17.790
mean
1.264
5.605
min
0.064
1.057
max
4.311
17.790
75
Table
6.4.
Winter
Sulfate
and
Nitrate
FTP
Emission
Rates
Vehicle
Run
#
MY
ID
NO3
mg/
mi
SO4
mg/
mi
1971­
80
97056
1971
H1
0.263
1.550
1971­
80
97053
1975
M1
0.071
0.421
1971­
80
97161
1978
ML2
0.060
0.372
1971­
80
97158
1978
H2
0.044
0.474
1971­
80
97013
1979
M1
0.051
0.110
1971­
80
97096
1980
M3
0.048
0.372
mean
0.090
0.550
min
0.044
0.110
max
0.263
1.550
1981­
85
97165
1981
M3
0.050
0.139
1981­
85
97025
1983
H1
0.188
0.048
1981­
85
97112
1984
M2
0.073
0.901
1981­
85
97115
1984
ML1
0.045
0.350
1981­
85
97062
1984
ML1
0.064
1.272
1981­
85
97059
1985
M2
0.078
0.866
mean
0.083
0.596
min
0.045
0.048
max
0.188
1.272
1986­
90
97072
1986
L1
0.062
0.196
1986­
90
97044
1987
L2
0.038
0.106
1986­
90
97047
1989
M1
0.056
0.123
1986­
90
97016
1989
L1
0.041
0.141
1986­
90
97137
1990
L2
0.029
0.157
1986­
90
97099
1990
ML1
0.033
0.933
mean
0.043
0.276
min
0.029
0.106
max
0.062
0.933
1991­
96
97168
1991
L2
0.044
0.281
1991­
96
97134
1993
M3
0.074
0.493
1991­
96
97127
1993
L1
0.018
0.055
1991­
96
97075
1993
ML2
0.032
0.205
1991­
96
97109
1994
H2
0.055
0.489
mean
0.045
0.305
min
0.018
0.055
max
0.074
0.493
Smoker
97093
1971
S1
ND
7.027
Smoker
97146
1975
S3
ND
0.443
Smoker
97050
1976
S1
ND
2.535
Smoker
97081
1977
S2
ND
0.589
Smoker
97103
1981
S2
ND
1.185
Smoker
97084
1983
S3
ND
1.038
Smoker
97152
1985
S3
ND
0.261
Smoker
97087
1989
S2
ND
1.326
mean
ND
1.801
min
ND
0.261
max
ND
7.027
Diesel
97251
1979
D1
ND
3.508
Diesel
97245
1979
D4
ND
3.550
Diesel
97242
1979
D1
ND
5.220
Diesel
97248
1980
D3
ND
4.170
Diesel
97239
1981
D5
ND
1.208
Diesel
97233
1981
D4
ND
1.985
Diesel
97267
1982
D3
ND
1.938
Diesel
97257
1982
D5
ND
1.560
Diesel
97254
1982
D2
ND
3.036
Diesel
97270
1983
D2
ND
3.238
Diesel
97261
1984
D5
ND
0.760
mean
ND
2.743
min
ND
0.760
max
ND
5.220
Since
the
number
of
cold
and
hot
starts
may
be
different
in
real
world
operation
from
the
weighting
factors
used
for
the
FTP,
it
is
useful
to
examine
the
emission
rates
by
FTP
phase.
Table
6.5
gives
the
average
sulfate
emission
rates
for
each
vehicle
category
by
FTP
76
phase.
Emissions
are
highest
in
phase
1
(
cold
start
operation)
and
lowest
in
phase
2
(
hot
stabilized
operation)
during
both
the
summer
and
winter.

It
is
commonly
assumed
that
sulfate
emitted
from
vehicles
is
present
as
sulfuric
acid.
Sulfuric
acid
is
very
hygroscopic.
It
is
also
frequently
assumed
that
each
sulfate
ion
has
seven
water
molecules
associated
with
it
under
the
standard
temperature
and
humidity
conditions
used
in
weighting
rooms.
Accounting
for
the
water
would
raise
the
mass
contribution
to
the
measured
PM
by
a
factor
of
2.3;
i.
e.,
1
mg/
mi
sulfate
would
give
rise
to
2.3
mg/
mi
PM
mass.
For
low
PM
emission
rate
vehicles,
the
sulfate
and
associated
water
could
be
a
significant
fraction
of
the
total
PM
emission.
Since
acidity
measurements
were
not
made,
it
is
not
know
how
much,
if
any,
of
the
sulfate
in
this
study
was
present
as
sulfuric
acid.
Overall,
it
does
not
appear
that
sulfate
and
associated
water
are
significant
contributors
to
the
PM
emission
rate
for
the
in­
use
gasoline
fleet.

Table
6.5.
Average
Sulfate
Emission
Rates
for
Each
FTP
Phase
Category
Period
Phase
1
mg/
mi
Phase
2
mg/
mi
Phase
3
mg/
mi
1971­
80
Summer
7.51
(
0.69)
1
0.13
0.20
1981­
85
Summer
0.27
0.08
0.11
1986­
90
Summer
0.28
0.15
0.24
1991­
96
Summer
0.08
0.05
0.03
Smokers
Summer
0.33
0.20
0.34
Diesels
Summer
9.19
(
3.76)
1
2.31
(
0.81)
1
9.16
(
1.28)
1
1971­
80
Winter
1.57
0.26
0.35
1981­
85
Winter
0.91
0.29
0.94
1986­
90
Winter
0.45
0.17
0.35
1991­
96
Winter
0.83
0.12
0.29
Smokers
Winter
6.03
(
2.24)
1
0.61
0.94
Diesels
Winter
7.51
1.37
1.76
1)
Values
in
(
)
are
the
averages
with
the
highest
emission
rate
removed
The
IC
sulfate
measurements
can
be
compared
to
the
total
sulfur
XRF
measurements.
Figure
6.14
plots
XRF
sulfur
versus
IC
sulfur
for
all
vehicles
in
the
summer,
where
the
IC
results
have
been
adjusted
to
remove
the
mass
of
the
associated
oxygen.
This
figure
plots
data
from
each
of
the
three
FTP
phases
rather
than
the
composite
FTP
data.
The
linear
regression
had
a
slope
of
0.35
and
an
R2
of
0.74.
The
average
ratio
of
IC
sulfur
to
XRF
sulfur
for
the
summer
was
0.40,
0.50,
and
0.45
for
FTP
phases
1­
3,
respectively.
For
the
winter,
the
correlation
between
IC
sulfur
and
XRF
sulfur
was
stronger,
having
a
slope
of
0.39
and
an
R2
of
0.92
for
all
vehicles
and
all
FTP
phases.
For
gasoline
and
smoking
vehicles
the
slope
was
0.34
and
the
R2
was
0.95.
The
average
ratio
of
IC
sulfur
to
XRF
sulfur
for
all
vehicles
in
the
winter
was
0.44,
0.39,
and
0.40
for
FTP
phases
1­
3,
respectively.
Removing
the
diesels
or
smokers
from
the
average
had
little
effect
on
these
ratios.
Overall,
it
is
apparent
that
while
total
sulfur
is
well
correlated
with
the
IC
sulfur,
the
majority
of
the
sulfur
is
not
emitted
as
water
soluble
sulfate.
This
is
true
for
all
three
FTP
phases.
The
non­
soluble
sulfate
sulfur
may
be
a
combination
of
organosulfur
compounds
and
zinc
dithiophosphate,
both
of
which
are
present
in
motor
oil.
Motor
oil
is
77
clearly
a
major
contributor
to
the
PM
mass
in
many
of
the
samples.
Analytical
errors
can't
be
ruled
out.
However,
the
extractions
and
analyses
were
performed
by
the
same
laboratory
that
analyzed
the
NFRAQS
ambient
samples.
Good
agreement
was
found
between
sulfate
and
XRF
for
these
samples
(
Chow
et
al.,
1998).

Fuel
sulfur.
For
well­
maintained
vehicles,
most
of
the
sulfur
emitted
in
the
exhaust
comes
from
the
fuel.
The
predominant
combustion
product
of
fuel
sulfur
is
sulfur
dioxide.
A
very
small
fraction
is
oxidized
to
sulfate,
or
is
emitted
in
its
original
form
as
unburned
fuel
organosulfur
compounds.
For
vehicles
with
high
oil
consumption,
some
of
the
sulfur
can
come
from
the
oil
as
well.
To
put
the
sulfate
emission
rates
discussed
above
into
perspective,
it
is
necessary
to
know
the
fuel
sulfur
content.
Fuel
sulfur
was
not
determined
for
individual
vehicles.
However,
fuel
samples
were
collected
at
the
pump
from
local
service
stations
during
both
the
summer
and
winter
studies.
In
addition,
the
American
Automobile
Manufacturers
Association
(
AAMA)
conducts
a
gasoline
survey
in
Denver
each
summer
and
winter
and
reports
fuel
sulfur
content
as
well
as
other
fuel
properties.
These
are
also
gasoline
service
station
samples.
The
samples
collected
in
our
program
were
chosen
to
complement
the
AAMA
data
base.

Table
6.6
gives
the
results
for
the
samples
collected
in
this
program.
Summer
samples
were
collected
September
19,
1996.
There
were
two
regular,
five
mid­
grade
and
seven
premium­
grade
gasoline
samples
and
two
#
2
diesel
fuel
samples.
The
average
weight
%
sulfur
was
0.0329,
0.0178,
0.0146,
and
0.0342,
respectively.
For
the
winter,
samples
were
collected
February
10,
1997.
There
were
three
regular,
five
mid­
grade,
and
eight
premium
gasoline
samples,
together
with
four
#
2
diesel
fuels
and
one
off­
road
diesel
fuel.
The
weight
%
sulfur
content
averaged
0.0148,
0.0193,
0.0109,
0.0358,
and
0.300,
respectively.
In
the
winter,
five
fuel
samples
were
analyzed
at
separate
laboratories.
Results
for
these
analyses
are
given
in
the
last
column
of
Table
6.4.
The
average
ratio
between
the
results
from
the
separate
laboratories
was
1.08.
The
greatest
difference
was
24%.

The
AAMA
Denver
1996
summer
survey
(
AAMA,
1996)
analyzed
15
regular­
grade,
one
intermediate­
grade,
and
two
premium­
grade
gasoline
samples,
as
well
as
nine
#
2
diesel
fuel
samples.
Each
sample
within
a
grade
category
was
from
a
different
brand
service
station.
Average
sulfur
content
of
the
samples
was
0.036,
0.017,
0.011,
and
0.034
wt.
%,
respectively.
The
winter
survey
analyzed
16
regular­
grade,
one
mid­
grade,
and
two
premium­
grade
gasoline
samples
and
10
#
2
diesel
fuel
samples.
Average
sulfur
content
was
0.024,
0.010,
0.008,
and
0.033
wt.
%,
respectively.
Note
that
these
are
not
sales
weighted
averages.
The
required
information
for
sales
weighting
was
not
available.
Results
from
both
this
study
and
the
AAMA
are
summarized
in
Table
6.7.
Included
in
Table
6.7
is
the
sulfur
content
average
for
results
from
both
AAMA
and
this
program.

Table
6.6.
Fuel
Sulfur
Analysis
Sample
No.
Brand
Brand
Name
Fuel
Grade
Octane
Rating
Pump
No.
Address
Sulfur,
wt
%
Sample
No.
Sulfur,
wt
%
Sulfur,
wt
%
Summer
Winter
78
1
Conoco
Unleaded
Plus
Mid­
grade
87
2
13690
E.
Colfax
0.0142
13
0.0166
2
Conoco
Super
Unleaded
Premium
91
4
0.0054
12
0.0127
4
Citgo
Mid­
grade
Unleaded
Mid­
grade
87
2
14490
E.
Colfax
0.0337
14
0.0353
5
Citgo
Super
Premium
Premium
91
2
0.0251
15
0.0277
6
Phillips
66
Super
Clean
Premium
Unleaded
Premium
91
1
14491
E.
Colfax
0.0077
3
0.0081
7
Phillips
66
Mid­
Octane
Unleaded
Plus
Mid­
grade
87
1
0.0073
4
0.0126
8
Phillips
66
Unleaded
Regular
Regular
85
1
0.0079
5
0.0135
9
Texaco
Super
Unleaded
Premium
91
7
14081
E.
Colfax
0.0116
7
0.0086
10
Texaco
Power
Plus
Premium
87
7
0.0246
6
0.0087
11
Sinclair
Unleaded
Regular
85
5
15291
E.
Colfax
0.0579
1
0.0278
12
Sinclair
Super
Unleaded
Premium
91
3
0.0199
2
0.0174
14
Total
Unleaded
Plus
Mid­
grade
87
3
1190
S.
Chambers
0.0120
16
0.0178
15
Total
Premium
Unleaded
Premium
91
3
0.0077
17
0.0037
Total
Regular
Unleaded
Regular
85
3
1190
S.
Chambers
20
0.0032
16
Amoco
Silver
Mid­
grade
87
1
15300
E.
Mississippi
0.0212
18
0.0142
Amoco
Ultimate
Lead
Free
Premium
Premium
91
1
15300
E.
Mississippi
19
0.0109
3
Conoco
Diesel
3
13690
E.
Colfax
0.0344
11
0.0346
0.028a
Conoco
Diesel
Premium
Premium
Diesel
14
1100
S.
Havana
9
0.0348
0.030a
13
Sinclair
Diesel
2
15291
E.
Colfax
0.0340
8
0.0352
0.032a
Total
Premium
Diesel
Premium
Diesel
Cetane
50
12796
E.
Colfax
10
0.0387
0.040a
Conoco
Off­
road
#
2
Diesel
Off­
road
21
0.2996
0.3233b
a)
analysis
by
Total
Petroleum
b)
analysis
by
Conoco
Table
6.7.
Average
Fuel
Sulfur
Fuel
Summer
Winter
CRC
AAMA
Wt.
Ave.
CRC
AAMA
Wt.
Ave.
#
%
S
#
%
S
%
S
#
%
S
#
%
S
%
S
Regular
2
0.034
15
0.036
0.036
3
0.015
16
0.024
0.023
Mid­
grade
5
0.018
1
0.017
0.018
5
0.019
1
0.010
0.018
Premium
7
0.015
2
0.011
0.014
8
0.011
2
0.008
0.010
#
2
Diesel
2
0.034
9
0.034
0.034
4
0.036
10
0.033
0.034
Off­
road
Diesel
0
ND
ND
ND
ND
1
0.300
ND
ND
0.300
A
vehicle
operating
with
a
fuel
economy
of
20
mpg
on
a
0.02
wt.
%
sulfur
fuel
consumes
approximately
31
mg/
mi
sulfur.
While
some
sulfur
is
stored
on
the
catalyst,
over
time
essentially
all
fuel
sulfur
is
expected
to
be
emitted
in
the
exhaust.
The
non­
smoking
gasoline
vehicle
sulfate
emission
rate
averaged
0.4
mg/
mi.,
which
is
an
average
sulfatesulfur
emission
rate
of
0.13
mg/
mi.
Thus,
on
average,
the
conversion
of
fuel
sulfur
to
sulfate
was
approximately
0.4%
for
the
non­
smoking
gasoline
vehicles
(
assuming
20
mpg
fuel
economy).
Of
course,
it
is
also
possible
that
some
of
the
observed
sulfate­
sulfur
79
derives
from
the
combustion
of
motor
oil.

XRF
elements.
Emission
rates
of
various
elements
are
of
interest
since
they
constitute
part
of
the
mobile
source
PM
source
signature
used
in
source/
receptor
modeling.
Trace
metals
are
of
interest
because
one
of
the
hypotheses
for
human
health
effects
of
PM
involves
reactions
catalyzed
by
transition
metals.
Also,
some
metals
are
considered
to
be
air
toxics.
Emission
rates
for
40
elements
were
obtained
by
XRF.
Of
these,
only
12
elements
had
emission
rates
that
averaged
higher
than
the
measurement
uncertainty
for
the
summer
samples
while
14
elements
averaged
higher
than
the
measurement
uncertainty
for
the
winter
samples.
Table
6.8
lists
the
40
elements
and
the
average
uncertainty
of
the
measurement
associated
with
each
element
in
the
summer
and
winter
for
the
Teflo
filters.
Uncertainties
were
higher
for
the
Pallflex
T60A20
filters.
This
table
gives
the
lower
quantifiable
limits
for
elements
that
were
not
detected.
The
average
FTP
emission
rates
by
vehicle
category
for
quantifiable
elements
are
listed
in
Table
6.9
for
the
summer
and
Table
6.10
for
the
winter.
Also
given
in
the
tables
are
the
minimum
and
maximum
emission
rates
for
each
category.
The
emission
rates
for
each
element
averaged
less
than
one
mg/
mi
for
the
non­
smoking
gasoline
vehicles,
with
the
exception
of
silicon
for
the
summer
1971­
80
vehicles,
and
iron
for
the
1971­
80
and
the
1981­
85
winter
vehicles.
The
summer
silicon
emission
rate
of
2.5
mg/
mi
is
dominated
by
one
high
emitting
vehicle.
Smoking
vehicles
have
higher
emission
rates
for
most
of
the
elements,
in
line
with
their
overall
higher
PM
rate.
Emission
rates
for
the
diesels
were
higher
than
for
gasoline,
with
silicon,
sulfur,
calcium,
iron
and
zinc
all
exceeding
an
average
of
one
mg/
mi
in
the
summer.
Zinc
and
silicon
are
not
reported
for
diesels
in
the
winter
because
of
the
high
filter
background
for
these
elements.

Table
6.11
gives
the
average
emission
rate
for
the
elements
as
the
percent
of
the
average
FTP
PM
mass
emission
rate
for
each
vehicle
category.
On
a
percent
basis,
the
diesel
emissions
do
not
appear
uniquely
high.
The
sum
of
the
12
elements
account
for
only
0.77
percent
of
the
average
PM
mass
for
the
smokers,
but
9.0
percent
of
the
PM
mass
for
the
1991­
96
vehicles
for
the
summer
vehicles.
The
latter,
however,
is
dominated
by
the
relatively
high
contribution
of
silicon.
This
was
not
seen
for
the
winter
1991­
96
vehicles
which
had
a
3.2
percent
contribution
of
the
elements
to
the
PM
mass,
and
only
a
0.28
percent
contribution
from
silicon.
Winter
smokers
and
diesels
have
a
low
percent
contribution
to
the
PM
mass
in
part
due
to
inability
to
measure
some
of
the
elements.
Overall,
it
is
seen
that
these
elements
are
a
relatively
small
contributor
to
the
PM
mass.
80
The
average
summer
FTP
emission
rates
of
the
elements
are
presented
graphically
in
Figure
6.15
for
the
gasoline
vehicles
only;
diesel
results
were
omitted
to
avoid
compressing
the
scale.
Note
that
the
silicon
emission
rate
for
1971­
80
vehicles
is
off
scale.
Figure
6.16
compares
the
winter
emission
rates
for
all
vehicles.
Rates
tend
to
decrease
with
decreasing
age
of
the
vehicles.

Correlating
the
emission
rates
of
the
elements
gives
some
information
regarding
their
source.
Zinc
and
phosphorus
most
likely
are
derived
from
motor
oil,
since
zinc
dithiophosphate
is
a
standard
anti­
wear
and
anti­
oxidant
oil
additive.
Thus,
it
is
not
surprising
that
Figure
6.17
shows
a
high
correlation
(
R2=
0.90)
between
these
species
for
the
summer
gasoline
vehicles.
The
winter
gasoline
vehicles
had
the
same
correlation
(
R2=
0.90)
but
a
slope
of
1.7
vs.
1.5
for
the
summer.
Calcium
and
magnesium
are
present
in
motor
oil
as
detergent
additives
and
copper
is
added
as
an
antioxidant.
These
elements
also
correlate
reasonably
well
with
phosphorus,
with
R2
values
of
0.77,
0.68,
and
0.63,
respectively,
for
the
summer
and
0.95,
0.86,
and
0.93,
respectively,
for
the
winter.
On
the
other
hand,
sulfur
is
expected
to
be
predominantly
from
the
gasoline,
yet
it
also
correlates
moderately
well
with
phosphorus
with
R2
=
0.63
in
the
summer
and
0.79
in
the
winter,
suggesting
that
a
significant
fraction
may
be
derived
from
the
motor
oil.
The
correlation
shown
in
Figure
6.18
between
summer
phosphorus
and
PM
mass,
R2=
0.29,
suggests
that
much
more
than
oil
consumption
is
involved
in
the
production
of
PM,
which
is
predominantly
carbonaceous
material.
The
R2
for
the
winter
correlation
between
phosphorus
and
PM
mass
was
0.52.

Silicon
is
sometimes
present
as
an
additive
to
motor
oil,
but
is
not
added
to
gasoline
since
it
can
damage
emission
control
components.
However,
the
silicon
emission
rate
did
not
correlate
with
phosphorus
(
R2=
0.03
in
the
summer
and
R2=
0.21
in
the
winter).
This
could
simply
indicate
that
the
silicon
concentration
in
motor
oil
is
highly
variable.
Silicon
could
also
be
present
from
the
ingestion
of
road
dust.
PM­
2.5
road
dust
composition
was
determined
as
part
of
the
NFRAQS
study.
The
ratio
of
silicon
to
iron,
aluminum,
and
calcium
averaged
1/
0.19/
0.32/
0.15
(
Eric
Fujita,
private
communication).
The
same
ratios
for
the
summer
and
winter
exhaust
PM­
10
(
which
is
mostly
PM­
2.5),
were
1/
0.35/
0.11/
0.80
and
1/
1.97/
0.25/
0.36,
respectively.
The
summer
1971­
80
exhaust
ratios
were
excluded
from
the
average
since
the
silicon
appeared
to
be
anomalously
high.
The
ambient
and
exhaust
ratios
indicate
that
road
dust
could
be
a
significant
contributor
to
all
of
these
species.
To
further
explore
this
possibility,
additional
correlations
were
examined.
While
iron
and
aluminum
were
weakly
correlated
(
R2=
0.44
in
the
summer
and
0.58
in
the
winter),
neither
was
correlated
with
silicon
in
the
summer
and
they
were
weakly
correlated
with
silicon
in
the
winter
(
R2=
0.56
for
iron
and
0.30
for
aluminum).
Thus,
the
relative
importance
of
wear
and
corrosion
products,
oil
consumption,
and
road
dust
ingestion
can
not
be
determined
from
the
data.
An
additional
source
of
silicon
could
be
gasket
grease,
but
it
seems
unlikely
that
this
would
contribute
significantly
to
the
emissions
from
older
vehicles.

Tables
6.12
and
6.13
give
the
average
emission
rates
as
a
function
of
FTP
phase
for
the
summer
and
winter
vehicles.
Emission
rates
tend
to
be
highest
during
the
cold
start
portion
of
the
test,
and
lowest
during
hot,
stabilized
operation.
The
same
general
pattern
81
is
observed
for
total
PM
mass
emission
rates.

Table
6.8.
Average
Uncertainties
in
the
XRF
Measurement
of
Elements
on
Teflon
Membrane
Filters
Element
Summer,
mg/
mi
Winter,
mg/
mi
Element
Summer,
mg/
mi
Winter,
mg/
mi
Sodium
0.1993
0.1116
Selenium
0.0046
0.0027
Magnesium
0.0375
0.0298
Bromine
0.0044
0.0020
Aluminum
0.0357
0.0167
Rubidium
0.0038
0.0023
Silicon
0.0200
0.0115
Strontium
0.0039
0.0022
Phosphorus
0.0143
0.0074
Yttrium
0.0050
0.0032
Sulfur
0.0120
0.0092
Zirconium
0.0057
0.0031
Chlorine
0.0287
0.0128
Molybdenum
0.0100
0.0057
Potassium
0.0204
0.0091
Palladium
0.0335
0.0186
Calcium
0.0206
0.0110
Silver
0.0369
0.0203
Titanium
0.0931
0.0524
Cadmium
0.0393
0.0222
Vanadium
0.0414
0.0223
Indium
0.0456
0.0257
Chromium
0.0137
0.0058
Tin
0.0572
0.0324
Manganese
0.0113
0.0054
Antimony
0.0680
0.0391
Iron
0.0083
0.0056
Barium
0.2520
0.1436
Cobalt
0.0134
0.0135
Lanthanum
0.3434
0.1946
Nickel
0.0071
0.0036
Gold
0.0219
0.0127
Copper
0.0070
0.0035
Mercury
0.0105
0.0061
Zinc
0.0072
0.0039
Thallium
0.0099
0.0068
Gallium
0.0074
0.0049
Lead
0.0098
0.0058
Arsenic
0.0131
0.0181
Uranium
0.0091
0.0051
Table
6.9.
Summer
FTP
Emission
Rates
for
Elements
Identified
by
XRF
Category
Stats.
PM
Mg
Al
Si
P
S
Cl
Ca
Fe
Cu
Zn
Br
Pb
Sum
­­
mg/
mi
­­
1971­
80
mean
102.2
0.029
0.018
2.516
0.094
0.512
0.230
0.071
0.113
0.012
0.099
0.012
0.043
3.75
min
21.2
0.003
0.009
0.126
0.026
0.114
0.006
0.043
0.021
0.003
0.060
0.001
0.008
max
254.6
0.090
0.036
13.503
0.198
1.829
1.218
0.092
0.243
0.030
0.179
0.061
0.095
1981­
85
mean
42.5
0.009
0.010
0.211
0.071
0.147
0.007
0.136
0.060
0.011
0.108
0.000
0.006
0.78
min
3.2
0.000
0.003
0.030
0.021
0.031
0.000
0.038
0.025
0.001
0.031
0.000
0.001
max
99.8
0.031
0.021
0.636
0.164
0.252
0.025
0.371
0.220
0.031
0.214
0.000
0.019
1986­
90
mean
85.8
0.046
0.041
0.129
0.115
0.225
0.012
0.170
0.088
0.025
0.186
0.000
0.027
1.07
min
1.3
0.000
0.000
0.024
0.001
0.014
0.000
0.000
0.016
0.001
0.003
0.000
0.001
max
766.3
0.259
0.285
0.390
0.637
1.464
0.086
0.972
0.273
0.227
1.055
0.001
0.254
1991­
96
mean
3.0
0.000
0.010
0.183
0.005
0.019
0.002
0.003
0.036
0.001
0.006
0.000
0.003
0.27
min
0.9
0.000
0.005
0.018
0.002
0.004
0.000
0.000
0.011
0.001
0.003
0.000
0.002
max
7.3
0.001
0.019
0.657
0.009
0.037
0.008
0.009
0.061
0.002
0.009
0.001
0.004
Smokers
mean
281.7
0.148
0.029
0.249
0.241
0.629
0.039
0.312
0.085
0.019
0.369
0.009
0.035
2.17
min
94.0
0.015
0.010
0.112
0.131
0.233
0.020
0.120
0.024
0.002
0.077
0.000
0.000
max
551.5
0.365
0.046
0.426
0.454
0.992
0.074
0.629
0.188
0.049
0.645
0.044
0.155
Diesel
mean
1175
0.402
0.303
3.189
0.634
4.504
0.139
1.329
3.151
0.019
1.731
0.009
0.150
15.6
min
143.3
0.000
0.060
1.096
0.036
0.840
0.050
0.074
0.202
0.002
0.129
0.002
0.010
max
3988
1.556
0.909
7.631
2.193
14.655
0.386
4.099
11.617
0.057
5.728
0.023
0.511
Table
6.10.
Winter
FTP
Emission
Rates
for
Elements
Identified
by
XRF
82
Category
Stats.
PM
Mg
Al
Si
P
S
Cl
Ca
Fe
Cu
Zn
Br
Pb
K
Cr
Sum
­­
mg/
mi­­
1971­
80
mean
108.7
0.078
0.090
0.714
0.225
0.581
0.068
0.241
1.013
0.018
0.246
0.057
0.222
0.027
0.021
3.603
1971­
80
min
31.3
0.034
0.018
0.101
0.029
0.170
0.017
0.083
0.073
0.004
0.047
0.000
0.000
0.008
0.006
0.590
1971­
80
max
276.1
0.125
0.190
2.406
0.676
1.366
0.156
0.440
2.759
0.036
0.446
0.305
0.834
0.049
0.048
9.837
1981­
85
mean
42.6
0.064
0.103
0.371
0.150
0.505
0.028
0.181
1.074
0.015
0.236
0.028
0.103
0.020
0.017
2.893
1981­
85
min
21.6
0.014
0.013
0.064
0.028
0.143
0.013
0.043
0.072
0.001
0.036
0.001
0.008
0.004
0.003
0.443
1981­
85
max
55.1
0.098
0.215
0.768
0.277
0.931
0.038
0.289
2.929
0.027
0.393
0.090
0.308
0.045
0.033
6.440
1986­
90
mean
25.0
0.051
0.098
0.132
0.079
0.217
0.041
0.126
0.236
0.009
0.149
0.002
0.019
0.015
0.010
1.183
1986­
90
min
5.2
0.015
0.014
0.056
0.024
0.062
0.007
0.041
0.057
0.001
0.028
0.000
0.000
0.006
0.004
0.315
1986­
90
max
73.9
0.128
0.401
0.333
0.291
0.736
0.092
0.368
0.596
0.041
0.639
0.006
0.071
0.026
0.017
3.747
1991­
96
mean
39.9
0.046
0.035
0.110
0.097
0.316
0.014
0.139
0.293
0.013
0.190
0.001
0.019
0.012
0.007
1.294
1991­
96
min
6.3
0.023
0.008
0.018
0.010
0.053
0.003
0.016
0.031
0.001
0.014
0.000
0.001
0.004
0.004
0.187
1991­
96
max
123.0
0.086
0.057
0.181
0.255
0.712
0.023
0.390
0.832
0.037
0.466
0.004
0.061
0.019
0.012
3.135
Smoker
mean
574.1
ND
ND
ND
ND
2.114
0.127
0.401
1.885
0.064
ND
0.036
0.282
ND
ND
4.909
Smoker
min
179.4
ND
ND
ND
ND
0.308
0.068
0.000
0.000
0.002
ND
0.000
0.003
ND
ND
0.381
Smoker
max
1349.9
ND
ND
ND
ND
6.723
0.298
1.090
12.54
0.173
ND
0.146
1.596
ND
ND
22.56
Diesel
mean
537.5
ND
ND
ND
ND
2.343
0.217
0.137
0.476
0.025
ND
ND
0.142
ND
ND
3.340
Diesel
min
193.9
ND
ND
ND
ND
0.917
0.050
0.000
0.023
0.000
ND
ND
0.000
ND
ND
0.990
Diesel
max
933.2
ND
ND
ND
ND
3.980
0.804
1.245
0.934
0.107
ND
ND
0.695
ND
ND
7.765
Table
6.11.
Percent
of
Average
FTP
Mass
Emission
Rate
Category
Season
Mg
Al
Si
P
S
Cl
Ca
Fe
Cu
Zn
Br
Pb
Sum
­­
percent­­
1971­
80
Summer
0.028
0.018
2.462
0.092
0.501
0.225
0.069
0.111
0.012
0.097
0.012
0.042
3.669
1981­
85
Summer
0.021
0.024
0.496
0.167
0.346
0.016
0.320
0.141
0.026
0.254
0.000
0.014
1.835
1986­
90
Summer
0.054
0.048
0.150
0.134
0.262
0.014
0.198
0.103
0.029
0.217
0.000
0.031
1.247
1991­
96
Summer
0.000
0.333
6.100
0.167
0.633
0.067
0.100
1.200
0.033
0.200
0.000
0.100
9.000
Smokers
Summer
0.053
0.010
0.088
0.086
0.223
0.014
0.111
0.030
0.007
0.131
0.003
0.012
0.770
Diesel
Summer
0.034
0.026
0.271
0.054
0.383
0.012
0.113
0.268
0.002
0.147
0.001
0.013
1.327
1971­
80
Winter
0.072
0.083
0.657
0.207
0.535
0.063
0.222
0.932
0.016
0.227
0.052
0.204
3.314
1981­
85
Winter
0.149
0.242
0.870
0.351
1.185
0.066
0.424
2.521
0.036
0.553
0.065
0.241
6.789
1986­
90
Winter
0.204
0.391
0.526
0.317
0.868
0.165
0.504
0.942
0.036
0.596
0.006
0.075
4.727
1991­
96
Winter
0.115
0.087
0.276
0.243
0.792
0.035
0.349
0.734
0.033
0.475
0.004
0.048
3.240
Smoker
Winter
ND
ND
ND
ND
0.368
0.022
0.070
0.328
0.011
ND
0.006
0.049
0.855
Diesel
Winter
ND
ND
ND
ND
0.436
0.040
0.026
0.089
0.005
ND
ND
0.026
0.621
83
Table
6.12.
Summer
Average
Emission
Rates
of
XRF
Elements
by
FTP
Phase
Category
Phase
Mass
Mg
Al
Si
P
S
Cl
Ca
Fe
Cu
Zn
Br
Pb
­­
mg/
mi
­­
1971­
80
1
327
0.068
0.041
10.711
0.208
1.997
1.031
0.190
0.371
0.031
0.271
0.053
0.112
1971­
80
2
43.4
0.019
0.010
0.426
0.051
0.111
0.025
0.032
0.030
0.006
0.034
0.002
0.015
1971­
80
3
44.8
0.018
0.017
0.331
0.092
0.160
0.019
0.057
0.078
0.011
0.095
0.001
0.044
1981­
85
1
108
0.031
0.017
0.326
0.162
0.393
0.017
0.264
0.129
0.030
0.260
0.001
0.016
1981­
85
2
23.9
0.003
0.006
0.201
0.026
0.069
0.004
0.073
0.030
0.003
0.037
0.000
0.002
1981­
85
3
28.4
0.005
0.011
0.141
0.088
0.111
0.006
0.162
0.066
0.011
0.129
0.000
0.006
1986­
90
1
100
0.089
0.071
0.166
0.211
0.319
0.014
0.290
0.195
0.034
0.327
0.000
0.041
1986­
90
2
13.0
0.033
0.028
0.136
0.073
0.237
0.014
0.141
0.046
0.019
0.123
0.000
0.020
1986­
90
3
81.9
0.060
0.039
0.125
0.175
0.273
0.011
0.275
0.111
0.025
0.274
0.000
0.031
1991­
96
1
6.35
0.000
0.012
0.231
0.010
0.034
0.008
0.011
0.092
0.003
0.013
0.001
0.005
1991­
96
2
2.30
0.001
0.010
0.221
0.003
0.017
0.001
0.000
0.013
0.001
0.002
0.000
0.002
1991­
96
3
1.94
0.000
0.007
0.075
0.004
0.010
0.002
0.002
0.036
0.001
0.006
0.000
0.002
Smoker
1
335
0.126
0.059
0.258
0.304
0.911
0.057
0.305
0.119
0.022
0.403
0.029
0.116
Smoker
2
225
0.178
0.022
0.225
0.175
0.472
0.037
0.181
0.031
0.022
0.279
0.004
0.010
Smoker
3
235
0.133
0.021
0.282
0.291
0.672
0.040
0.342
0.110
0.018
0.488
0.008
0.031
Diesel
1
2510
0.601
0.403
4.627
1.148
6.781
0.308
2.451
2.925
0.027
3.250
0.011
0.190
Diesel
2
437
0.197
0.103
1.234
0.248
1.736
0.070
0.438
0.614
0.005
0.594
0.005
0.048
Diesel
3
1568
0.639
0.607
5.819
0.982
8.040
0.144
2.185
8.133
0.039
2.756
0.015
0.313
Table
6.13.
Winter
Average
Emission
Rates
of
XRF
Elements
by
FTP
Phase
Vehicle
Phase
Mass
Mg
Al
Si
P
S
Cl
Ca
Fe
Cu
Zn
Br
Pb
K
Cr
­­
mg/
mi
­­
1971­
80
1
401
0.170
0.257
2.064
0.732
2.028
0.143
0.662
3.699
0.056
0.807
0.206
0.769
0.066
0.007
1971­
80
2
37
0.050
0.049
0.373
0.082
0.212
0.052
0.137
0.290
0.007
0.098
0.022
0.041
0.016
0.005
1971­
80
3
35
0.063
0.047
0.399
0.127
0.247
0.044
0.135
0.435
0.010
0.123
0.012
0.167
0.021
0.006
1981­
85
1
43
0.084
0.123
0.339
0.165
0.698
0.030
0.198
1.729
0.019
0.306
0.047
0.162
0.024
0.019
1981­
85
2
24
0.030
0.058
0.417
0.062
0.221
0.030
0.082
0.339
0.006
0.106
0.006
0.026
0.011
0.015
1981­
85
3
37
0.077
0.102
0.343
0.165
0.540
0.031
0.195
1.164
0.019
0.263
0.029
0.108
0.020
0.014
1986­
90
1
66
0.048
0.160
0.199
0.109
0.353
0.048
0.152
0.305
0.010
0.186
0.002
0.024
0.018
0.006
1986­
90
2
12
0.043
0.061
0.103
0.046
0.134
0.043
0.098
0.172
0.006
0.092
0.001
0.013
0.014
0.005
1986­
90
3
19
0.070
0.121
0.136
0.121
0.275
0.032
0.161
0.307
0.014
0.232
0.002
0.027
0.013
0.006
1991­
96
1
128
0.111
0.088
0.283
0.301
0.992
0.028
0.424
0.806
0.043
0.615
0.004
0.058
0.026
0.005
1991­
96
2
23
0.020
0.020
0.067
0.030
0.125
0.009
0.047
0.138
0.004
0.054
0.001
0.006
0.008
0.005
1991­
96
3
10
0.050
0.026
0.068
0.084
0.211
0.012
0.121
0.242
0.011
0.153
0.001
0.019
0.010
0.007
Smoker
1
1369
ND
ND
ND
ND
6.316
0.152
0.342
4.500
0.121
ND
0.113
0.514
ND
ND
Smoker
2
359
ND
ND
ND
ND
0.775
0.128
0.376
0.471
0.034
ND
0.005
0.070
ND
ND
Smoker
3
398
ND
ND
ND
ND
1.577
0.107
0.494
2.682
0.080
ND
0.038
0.519
ND
ND
Diesel
1
937
ND
ND
ND
ND
5.524
0.430
0.274
1.625
0.054
ND
ND
0.255
ND
ND
Diesel
2
410
ND
ND
ND
ND
1.411
0.159
0.117
0.176
0.015
ND
ND
0.097
ND
ND
Diesel
3
481
ND
ND
ND
ND
1.732
0.164
0.073
0.181
0.021
ND
ND
0.142
ND
ND
84
Comparison
of
outdoor
and
indoor
composition.
The
focus
of
this
program
was
to
characterize
the
composition
of
the
PM
from
the
winter
outdoor
samples,
since
those
are
the
best
characterization
of
the
PM
emitted
into
the
winter
Denver
atmosphere.
However,
a
limited
comparison
was
done
between
the
outdoor
and
indoor
samples.
Eleven
phase
1
FTP
samples
from
gasoline
vehicles
tested
indoors
were
analyzed
by
IC
and
XRF
for
comparison
to
the
outdoor
samples.
Phase
2
and
3
samples
were
not
run,
since
they
are
not
expected
to
be
significantly
affected
by
the
ambient
temperature.
Results
are
given
in
Table
6.14.
The
NFRAQS
number
identifies
the
vehicle
tested.
The
second
column
gives
the
outdoor
PM­
10
phase
1
FTP
emission
rate
for
that
vehicle.
The
third
column
gives
the
ratio
of
outdoor
to
indoor
PM
emission
rates.
Note
that
while
most
vehicles
had
higher
PM­
10
emission
rates
outdoors,
vehicles
345,
362,
and
348
had
lower
rates.
The
following
columns
give
the
relative
change
in
rate
of
the
indicated
species
in
relationship
to
the
change
in
the
PM­
10
emission
rate.
For
example,
if
the
outdoor
to
indoor
PM­
10
emission
rate
ratio
was
3.0,
and
the
sulfate
outdoor
to
indoor
ratio
was
also
3.0,
then
the
relative
change
in
rate
in
Table
6.14
would
be
1.0.
Missing
values
indicate
that
one
of
the
measured
rates
was
below
minimum
quantifiable
limit.
Results
are
highly
variable,
with
both
large
increases
in
some
emission
rates
and
large
decreases
in
others.

Phase
1
FTP
PM­
10
filter
samples
were
also
analyzed
for
three
smoking
vehicles
and
three
diesels.
The
samples
used
the
Pallflex
T60A20
filters
that
had
high
nitrate
blanks
and
relatively
high
XRF
detectable
elements.
Thus,
the
only
legitimate
comparison
that
could
be
made
was
for
sulfate.
This
comparison
is
shown
in
Table
6.15.
As
with
the
gasoline
vehicles,
the
sulfate
tends
to
increase
at
a
rate
equal
to
or
greater
than
the
increase
in
the
PM­
10
rate.

Table
6.14.
Relative
Change
in
Phase
1
FTP
Emission
Rates
Relative
to
the
Change
in
PM
Emission
Rate
NFRAQ
#
Outdoor
PM,
mg/
mi
Out/
In
PM
Ratio
Relative
change
in
rate
compared
to
change
in
PM
rate
NO3
SO4
Mg
Al
Si
P
S
Cl
Ca
Fe
Cu
Zn
Pb
317
22.2
3.43
0.32
0.52
1.80
0.17
0.36
0.36
0.19
0.16
349
13.1
5.84
0.11
0.17
0.66
0.60
0.73
0.46
0.37
8.61
323
40.7
3.17
0.87
6.93
1.74
1.79
1.44
1.99
0.92
1.56
2.03
335
72.3
1.43
0.62
9.86
1.71
9.29
5.46
0.58
7.11
18.19
16.37
336
38.3
7.37
0.36
0.70
0.32
0.63
1.01
0.57
0.60
0.80
4.04
0.63
345
71.2
0.55
1.41
0.88
3.79
1.87
0.75
3.14
1.27
1.11
1.99
0.38
3.80
3.15
1.12
362
115
0.60
1.27
1.49
2.30
0.10
2.17
1.50
0.61
1.98
5.53
3.27
2.44
1.76
348
89.7
0.94
8.89
0.67
5.69
0.75
0.38
1.75
0.72
0.50
1.55
1.60
2.31
1.78
342
204
8.51
0.39
0.24
0.50
0.49
0.51
1.06
1.28
0.35
1.14
1.40
1.27
1.61
3.11
322
1120
4.25
0.47
2.65
0.29
1.89
0.83
0.22
1.43
0.38
0.64
4.99
0.40
0.25
0.09
338
589
10.26
0.08
0.07
0.10
0.58
0.81
0.26
0.69
0.35
0.28
0.71
0.23
Mean
215.95
4.21
1.45
2.37
1.89
1.13
0.88
1.91
1.47
0.54
1.60
3.51
1.80
3.48
1.35
85
Table
6.15.
Relative
Change
in
Sulfate
Emission
Rate
for
Smoking
and
Diesel
Vehicles
NFRAQ
#
Category
Outdoor
PM
mg/
mi
Out/
In
PM
ratio
SO4
ratio
SO4
ratio/
Mass
ratio
329
S
5951
1.56
9.81
6.30
326
S
130
1.03
1.60
1.56
321
S
2074
3.15
6.10
1.94
383
D
1467
1.30
2.15
1.66
389
D
377
1.57
1.85
1.18
391
D
596
1.13
1.70
1.50
Polynuclear
aromatic
hydrocarbons
(
PAH).
The
primary
purpose
of
measuring
the
PAHs
was
to
provide
additional
compounds
for
use
in
creating
source
profiles.
PAHs
are
also
of
interest
as
one
of
the
five
air
toxics
listed
by
the
EPA
as
being
important
for
mobile
sources.
As
was
noted
earlier,
PAH
was
determined
on
composited
samples.
Composites
for
non­
smoking
gasoline
vehicles
were
formed
from
vehicles
having
similar
emission
rates,
regardless
of
the
model
year.
These
composites
are
identified
as
either
being
low
emitters
(
L),
medium
low
emitters
(
ML),
medium
emitters
(
M),
or
high
(
H)
emitters.
Thus,
the
non­
smoking
gasoline
vehicle
PAH
emissions
data
can
not
be
examined
using
the
vehicle
model
year
categories
used
in
the
rest
of
the
report.
On
the
other
hand,
the
carbon,
anion,
and
elemental
analyses
discussed
earlier
were
all
performed
on
an
individual
vehicle
basis.
Thus,
that
data
can
be
composited
to
match
the
PAH
data.

Tables
6.16
and
6.17
list
the
15
summer
and
15
winter
composite
samples.
No
data
are
listed
for
winter
L1
and
H2
samples
since
one
filter
extract
was
lost
from
each
of
these.
Samples
labeled
S
and
D
are
from
smoking
gasoline
vehicles
and
diesel
vehicles,
respectively.
The
number
of
vehicles
in
each
composite
is
also
given
in
the
Tables.
For
most
samples
it
was
two.
Also
given
is
the
average
FTP
PM
emission
rate
for
the
composite,
the
average
FTP
OC
emission
rate,
and
the
average
FTP
PAH
emission
rate.
Table
6.17
is
organized
so
that
the
PM
emission
rate
within
an
emitter
category
increases
from
top
to
bottom.
The
PAH
emission
rates
include
both
gaseous
and
particulate
compounds;
i.
e.
those
captured
on
the
filter
and
those
captured
in
the
PUF/
XAD
trap
behind
the
filter.
Since
the
extracts
from
the
filters
and
PUF/
XAD
traps
were
analyzed
together,
it
is
not
possible
to
separate
these
by
capture
media.
Naphthalene
was
the
most
abundant
PAH
present.
However,
there
were
significant
and
highly
variable
background
naphthalene
concentrations
in
the
summer.
Therefore,
summer
naphthalene
emission
rates
are
not
reported,
and
all
total
PAH
emission
rates
exclude
naphthalene,
unless
otherwise
noted.

Examination
of
Tables
6.16
and
6.17
shows
that
total
PAH
does
not
correlate
closely
with
PM
emission
rates.
The
lowest
summer
PM
emission
rate
sample,
L1,
has
PAH
emissions
almost
equal
to
PM
emissions.
Samples
L2
and
L3
have
much
higher
average
PM
emission
rates
than
L1,
but
essentially
the
same
total
PAH
emission
rate.
Similarly,
the
highest
total
PAH
rates
for
the
high
emitters,
10.4
mg/
mi,
was
for
the
sample
with
the
86
lowest
PM
emission
rate,
45.4
mg/
mi.
It
is
possible
that
the
mixing
of
gas
and
particle
phase
PAH
is
confounding
the
relationship
between
PM
and
PAH.
Tables
6.16
and
6.17
also
list
the
average
FTP
OC
emission
rate.
It
would
be
logical
to
compare
the
PAH
emission
rates
with
the
organic
fraction
of
the
PM
rather
than
the
total
PM
mass.
Unfortunately,
the
problems
in
measuring
the
diesel
and
smoker
summer
carbon
limit
the
utility
of
this
approach
for
the
smokers
and
diesels
in
the
summer
samples.
Note,
for
example,
that
the
summer
S1
and
S2
samples
have
more
OC
than
PM
mass.
A
regression
of
the
winter
OC
vs.
PAH
showed
considerable
scatter,
with
an
R2
of
0.48.

Sixty­
nine
PAH
compounds
were
determined
in
the
summer
samples
and
65
were
determined
in
the
winter
samples.
The
difference
in
number
was
due
to
the
deletion
of
four
compounds
from
the
summer
list
that
had
negligible
emission
rates
for
vehicles.
Table
6.18
lists
the
average
summer
emission
rates
of
the
individual
PAHs
for
each
emitter
category.
The
compounds
are
listed
in
the
same
order
that
they
are
eluted
from
the
GC.
Thus
the
order
corresponds,
approximately,
with
their
vapor
pressure.
Designation
of
a
compound
as
A­,
B­,
etc.
refers
to
isomers
that
have
not
been
uniquely
identified.
Of
the
compounds
listed,
all
but
eight
had
concentrations
that
were
significantly
greater
than
the
measurement
uncertainty
in
the
summer.
The
eight
with
high
uncertainty
relative
to
their
concentrations
were
1,8­
dimenaphthalene,
D­
trimethylnaphthalene,
G­
methylpyrene,
Retene,
Benzonaphthothiophene,
7­
Methylbenz[
a]
anthracene,
7­
Methylbenzo[
a]
pyrene,
and
Dibenz(
ah+
ac)
anthracene.
Figure
6.19
gives
a
graphical
presentation
of
the
average
emission
rates
for
summer
high
emitter
samples.
Note
that
the
rates
are
plotted
on
a
log
scale
and
cover
four
orders
of
magnitude.
Tables
6.18
and
6.19
also
assign
each
PAH
into
a
group.
These
groups
are
a
somewhat
arbitrary
way
of
compositing
the
PAHs
to
facilitate
comparison.
They
were
selected
after
an
examination
of
the
variability
of
compounds
between
emitter
categories
and
FTP
phases.
Compounds
whose
emission
rates
tended
to
change
in
a
similar
manner
were
grouped
together.
Generally,
isomers
were
lumped
into
the
same
group.
These
groups
can
be
viewed
primarily
as
PAHs
with
similar
volatilities.

The
PAHs
can
be
formed
during
combustion
or
can
be
unburned
components
of
the
fuel
and
motor
oil.
It
was
beyond
the
scope
of
this
program
to
test
individual
fuel
and
motor
oil
samples
for
PAH
content.
Therefore,
it
is
not
possible
to
determine
the
sources
of
observed
PAH.
Some
of
the
observed
emission
variability
is
likely
due
to
differences
in
composition
of
the
fuel
and
motor
oil,
as
well
as
differences
in
the
relative
contributions
that
the
fuel
and
motor
oil
make
to
the
PM
from
different
vehicles.

The
suitability
of
using
these
compounds
for
source
profiles
was
judged
qualitatively
by
comparing
the
emission
rates
between
samples.
The
wide
range
in
concentrations
of
the
individual
PAHs
makes
graphical
comparisons
on
a
concentration
basis
difficult.
Therefore,
the
ratios
of
individual
compounds
were
compared.
This
was
done
in
three
steps.
First,
the
consistency
between
composites
within
the
same
emitter
category
was
examined.
Second,
the
composites
in
each
category
were
averaged,
and
the
average
composites
were
compared.
These
averages
were
not
weighted
by
emission
rate.
Third,
differences
between
the
three
FTP
phases
were
examined
for
the
average
composites.
87
An
example
of
the
consistency
in
relative
emission
rates
within
an
emitter
category
is
shown
in
Figure
6.20
which
plots
the
ratios
between
summer
sample
composites
L2
and
L3.
Missing
ratios
are
for
species
whose
rates
were
below
the
detection
limit
for
at
least
one
of
the
samples.
There
is
good
overall
agreement
for
most
compounds.
Somewhat
more
variability
was
observed
for
the
ratios
of
L3
and
L2
with
L1,
probably
because
L1
was
the
lowest
emission
rate
sample.
Differences
between
samples
were
greater
for
the
summer
medium,
high
and
smoker
categories.
This
is
illustrated
in
Figure
6.21,
which
Table
6.16.
Average
Summer
FTP
Emission
Rates
for
the
Composite
PAH
Samples
Sample
No.
of
Vehicles
PM
Emission
Rate,
mg/
mi
OC
Emission
Rate,
mg/
mi
PAH
Emission
Rate,
mg/
mi
L1
3
2.03
3.68
1.98
L2
2
6.86
4.94
1.82
L3
2
9.45
6.94
2.32
Average
6.11
5.19
2.03
M1
2
22.4
8.76
3.46
M2
2
20.5
9.43
3.95
M3
2
30.4
14.8
9.19
Average
24.4
11.0
5.51
H1
2
45.4
22.2
10.4
H2
2
60.7
64.6
4.43
H3
2
105
58.8
9.28
Average
70.5
48.5
8.05
S1
2
497
529
18.6
S2
2
163
222
29.9
S3
2
186
123
45.5
Average
282
291
31.3
D1
2
213
143
34.2
D2
1
289
227
49.9
D3
1
3990
2879
53.0
Average
1497
1083
45.6
Table
6.17.
Average
Winter
FTP
Emission
Rates
for
the
Composite
PAH
Samples
Sample
No.
of
Vehicles
PM,
mg/
mi
OC,
mg/
mi
PAH,
mg/
mi
PAH
w/
o
napht,
mg/
mi
L2
3
7.2
5.7
4.45
2.98
88
ML2
2
21.2
12.5
25.30
14.72
ML1
3
35.3
13.7
10.01
5.92
average
28.3
13.1
17.7
10.3
M3
3
50.2
27.8
18.93
13.68
M2
2
54.6
19.9
16.75
10.23
M1
3
79.4
55.4
23.54
15.09
average
61.4
34.4
19.7
13.0
H1
2
212.6
64.1
75.71
38.88
S3
2
201.0
158.2
19.61
11.42
S2
3
601.6
455.1
39.26
24.32
S1
3
1091.8
431.7
170.96
104.42
average
631.5
348.3
76.6
46.7
D5
3
346.1
142.2
7.87
7.07
D4
2
427.3
47.5
12.12
9.00
D3
2
534.0
57.8
21.79
17.18
D2
2
627.8
152.6
42.55
37.48
D1
2
848.1
259.6
58.36
47.57
average
556.7
131.9
28.5
23.7
Table
6.18.
Average
Summer
PAH
FTP
Emission
Rates
ID
Group
Compound
Ave
L
Ave
M
Ave
H
Ave
S
Ave
D
­­
mg/
mi
­­
1
1
2­
menaphthalene
0.734
2.059
2.764
11.026
6.053
2
1
1­
menaphthalene
0.341
1.147
1.528
6.108
4.421
3
1
2,6+
2,7­
dimenaphthalene
0.072
0.214
0.343
1.334
1.579
4
1
1,7+
1,3+
1,6­
dimenaphthalene
0.136
0.362
0.620
2.653
3.680
5
1
2,3+
1,4+
1,5­
dimenaphthalene
0.043
0.140
0.217
0.869
1.285
6
1
1,2­
dimenaphthalene
0.018
0.062
0.095
0.476
0.523
7
1
1,8­
dimenaphthalene
0.000
0.000
0.000
0.005
0.000
8
2
Biphenyl
0.037
0.068
0.095
0.336
1.051
9
2
A­
Methylbiphenyl
0.003
0.002
0.003
0.005
0.039
10
2
2­
Methylbiphenyl
0.005
0.013
0.019
0.075
0.341
11
2
B­
Methylbiphenyl
0.007
0.016
0.018
0.070
0.222
12
2
3­
Methylbiphenyl
0.022
0.052
0.069
0.258
1.625
13
2
4­
Methylbiphenyl
0.011
0.026
0.032
0.119
0.632
14
2
C­
Methylbiphenyl
0.028
0.047
0.065
0.196
1.098
15
3
A­
Trimethylnaphthalene
0.033
0.106
0.174
0.597
1.651
16
3
1­
Ethyl­
2­
methylnaphthalene
0.010
0.033
0.048
0.183
0.415
17
3
B­
Trimethylnaphthalene
0.030
0.101
0.182
0.674
2.022
18
3
C­
Trimethylnaphthalene
0.028
0.096
0.156
0.589
1.993
19
3
2­
Ethyl­
1­
methylnaphthalene
0.002
0.014
0.020
0.052
0.120
20
3
D­
Trimethylnaphthalene
0.000
0.000
0.000
0.000
0.000
21
3
E­
Trimethylnaphthalene
0.018
0.060
0.101
0.399
1.431
22
3
F­
Trimethylnaphthalene
0.011
0.035
0.058
0.225
0.720
23
3
G­
Trimethylnaphthalene
0.015
0.035
0.065
0.286
0.935
24
3
H­
Trimethylnaphthalene
0.0029
0.0129
0.0196
0.0826
0.2009
25
3
1,2,8­
Trimethylnaphthalene
0.0061
0.0136
0.0213
0.0551
0.5416
26
4
Acenaphthylene
0.027
0.142
0.424
1.519
0.773
89
27
4
Acenaphthene
0.015
0.024
0.047
0.173
0.180
28
4
Phenanthrene
0.010
0.069
0.086
0.405
1.419
29
4
Fluorene
0.025
0.079
0.147
0.440
0.758
30
4
1­
Methylfluorene
0.012
0.030
0.037
0.138
0.551
31
4
B­
Methylfluorene
0.005
0.014
0.017
0.067
0.375
32
4
C­
Methylfluorene
0.047
0.048
0.059
0.172
0.747
33
5
A­
Methylphenanthrene
0.021
0.033
0.032
0.091
0.815
34
5
2­
Methylphenanthrene
0.026
0.037
0.037
0.104
0.902
35
5
B­
Methylphenanthrene
0.0041
0.0088
0.0116
0.0338
0.0437
36
5
C­
Methylphenanthrene
0.0117
0.0182
0.0211
0.0562
0.5553
37
5
1­
Methylphenanthrene
0.0133
0.0193
0.0192
0.0534
0.4512
38
5
3,6­
Dimethylphenanthrene
0.0061
0.0080
0.0073
0.0231
0.2347
39
5
A­
Dimethylphenanthrene
0.0083
0.0105
0.0092
0.0310
0.3099
40
5
B­
Dimethylphenanthrene
0.0042
0.0052
0.0045
0.0153
0.1462
41
5
C­
Dimethylphenanthrene
0.0118
0.0156
0.0148
0.0488
0.5358
41
5
1,7­
Dimethylphenanthrene
0.0061
0.0086
0.0079
0.0262
0.2231
43
5
D­
Dimethylphenanthrene
0.0038
0.0047
0.0036
0.0130
0.1380
44
5
E­
Dimethylphenanthrene
0.0055
0.0090
0.0059
0.0956
0.1913
45
6
Anthracene
0.0220
0.0481
0.0572
0.1466
0.5619
46
6
9­
Methylanthracene
0.00084
0.00097
0.00130
0.00354
0.02005
47
6
Fluoranthene
0.04839
0.04743
0.07047
0.17579
0.89560
48
6
Pyrene
0.05564
0.05206
0.08402
0.30144
1.05026
49
6
A­
Methylpyrene
0.00575
0.00716
0.00847
0.00926
0.06618
50
6
B­
Methylpyrene
0.00307
0.00371
0.00417
0.01893
0.06110
51
6
C­
Methylpyrene
0.00535
0.00677
0.00909
0.04189
0.11719
52
6
D­
Methylpyrene
0.00193
0.00276
0.00441
0.02112
0.03465
53
6
E­
Methylpyrene
0.00239
0.00312
0.00388
0.01806
0.09252
54
6
F­
Methylpyrene
0.00159
0.00268
0.00428
0.01875
0.07317
55
6
G­
Methylpyrene
0.00000
0.00000
0.00000
0.00000
0.00000
56
7
Retene
0.00000
0.00008
0.00000
0.00024
0.00227
57
7
Benzonaphthothiophene
0.00000
0.00009
0.00028
0.00189
0.00355
58
7
Benz(
a)
anthracene
0.00159
0.00380
0.00821
0.02872
0.07683
59
7
7­
Methylbenz[
a]
anthracene
0.00004
0.00013
0.00018
0.00059
0.00080
60
7
Chrysene
0.00126
0.00374
0.00850
0.03280
0.15281
61
7
Benzo(
b+
j+
k)
fluoranthene
0.00007
0.00294
0.01250
0.03526
0.11486
62
7
Benzo[
e]
pyrene
0.00035
0.00276
0.00978
0.02835
0.05881
63
7
Benzo[
a]
pyrene
0.00052
0.00247
0.01068
0.04167
0.03923
64
7
7­
Methylbenzo[
a]
pyrene
0.00012
0.00017
0.00023
0.00040
0.00162
65
7
Indeno[
123­
cd]
pyrene
0.00033
0.00204
0.00791
0.02076
0.04757
66
7
Dibenz(
ah+
ac)
anthracene
0.00007
0.00033
0.00087
0.00158
0.00380
67
7
Benzo(
b)
chrysene
0.00808
0.01281
0.01497
0.02729
0.12681
68
7
Benzo(
ghi)
perylene
0.00089
0.00578
0.01832
0.09187
0.08292
69
7
Coronene
0.00046
0.00265
0.00932
0.05404
0.02068
Total
PAH
(
w/
o
Naphth)
2.03
5.51
8.05
31.30
45.63
90
Table
6.19.
Average
Winter
PAH
FTP
Emission
Rates
ID
Group
Compound
L2
Ave
ML
Ave
M
H1
Ave
S
Ave
D
­­
mg/
mi
­­
1
0
Naphthalene
1.465
7.334
6.737
36.829
29.892
4.877
2
1
2­
menaphthalene
0.714
3.600
3.754
12.607
13.551
2.821
3
1
1­
menaphthalene
0.337
1.943
1.969
6.900
7.571
2.131
4
1
2,6+
2,7­
dimenaphthalene
0.067
0.399
0.462
1.722
2.524
1.413
5
1
1,7+
1,3+
1,6­
dimenaphthalene
0.085
0.594
0.679
2.644
3.560
2.113
6
1
2,3+
1,4+
1,5­
dimenaphthalene
0.028
0.190
0.213
0.843
1.242
0.888
7
1
1,2­
dimenaphthalene
0.012
0.087
0.094
0.278
0.566
0.288
8
1
1,8­
dimenaphthalene
0.0000
0.0022
0.0016
0.0087
0.0163
0.0054
9
2
Biphenyl
0.0293
0.1154
0.1509
0.5320
0.5228
0.6945
10
2
2­
Methylbiphenyl
0.0000
0.0087
0.0177
0.0985
0.0919
0.1884
11
2
3­
Methylbiphenyl
0.0176
0.0698
0.1084
0.3202
0.3769
0.9959
12
2
4­
Methylbiphenyl
0.0095
0.0358
0.0567
0.1510
0.1788
0.4248
13
3
A­
Trimethylnaphthalene
0.0243
0.1313
0.1631
0.4863
0.8941
0.9944
14
3
1­
Ethyl­
2­
methylnaphthalene
0.0076
0.0388
0.0494
0.1380
0.2549
0.2580
15
3
B­
Trimethylnaphthalene
0.0237
0.1350
0.1825
0.5087
0.9237
1.0971
16
3
C­
Trimethylnaphthalene
0.0258
0.1308
0.1804
0.4964
0.9136
1.3156
17
3
2­
Ethyl­
1­
methylnaphthalene
0.0018
0.0078
0.0098
0.0249
0.0778
0.0599
18
3
D­
Trimethylnaphthalene
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
19
3
E­
Trimethylnaphthalene
0.0153
0.0884
0.1231
0.3432
0.6214
1.0090
20
3
F­
Trimethylnaphthalene
0.0205
0.0946
0.1332
0.3449
0.6251
0.9230
21
3
G­
Trimethylnaphthalene
0.0105
0.0457
0.0723
0.1850
0.3392
0.3992
22
3
1,2,8­
Trimethylnaphthalene
0.0007
0.0016
0.0040
0.0050
0.0910
0.0297
23
4
Acenaphthylene
0.0868
0.3299
0.6858
1.6119
3.2736
0.3791
24
4
Acenaphthene
0.0441
0.0949
0.0566
0.1478
0.1747
0.0890
25
4
Phenanthrene
0.3355
0.5740
1.1630
2.0699
2.5742
1.2781
26
4
Fluorene
0.0522
0.1358
0.2222
0.4868
0.7726
0.4077
27
4
A­
Methylfluorene
0.0230
0.0657
0.1110
0.2382
0.3672
0.3912
28
4
1­
Methylfluorene
0.0125
0.0295
0.0519
0.0961
0.1713
0.2715
29
4
B­
Methylfluorene
0.0061
0.0159
0.0246
0.0433
0.0828
0.0838
30
4
C­
Methylfluorene
0.0525
0.1469
0.2648
0.5462
0.6345
0.2070
31
5
A­
Methylphenanthrene
0.0486
0.0617
0.1160
0.2303
0.2362
0.2951
32
5
2­
Methylphenanthrene
0.0603
0.0716
0.1262
0.2429
0.2569
0.3150
33
5
B­
Methylphenanthrene
0.0076
0.0207
0.0340
0.0274
0.0575
0.0249
34
5
C­
Methylphenanthrene
0.0339
0.0482
0.0808
0.1303
0.1578
0.2352
35
5
1­
Methylphenanthrene
0.0398
0.0484
0.0806
0.1400
0.1458
0.1758
36
5
3,6­
Dimethylphenanthrene
0.0151
0.0181
0.0244
0.0545
0.0442
0.0846
37
5
A­
Dimethylphenanthrene
0.0198
0.0206
0.0311
0.0683
0.0495
0.0967
38
5
B­
Dimethylphenanthrene
0.0099
0.0108
0.0156
0.0342
0.0250
0.0504
39
5
C­
Dimethylphenanthrene
0.0313
0.0360
0.0503
0.0929
0.0861
0.1833
40
5
1,7­
Dimethylphenanthrene
0.0141
0.0194
0.0247
0.0429
0.0390
0.0852
41
5
D­
Dimethylphenanthrene
0.0078
0.0093
0.0130
0.0248
0.0227
0.0534
42
5
E­
Dimethylphenanthrene
0.0136
0.0147
0.0212
0.0428
0.0339
0.0586
43
6
Anthracene
0.1238
0.1278
0.2805
0.5430
0.6093
0.2246
44
6
9­
Methylanthracene
0.0050
0.0044
0.0077
0.0198
0.0136
0.0122
45
6
Fluoranthene
0.2182
0.2668
0.4352
1.0176
0.5869
0.1810
46
6
Pyrene
0.2565
0.3561
0.5585
1.4643
0.7531
0.2089
47
6
B­
Methylpyrene
0.0048
0.0059
0.0081
0.0219
0.0150
0.0094
48
6
C­
Methylpyrene
0.0120
0.0155
0.0202
0.0557
0.0397
0.0237
49
6
D­
Methylpyrene
0.0048
0.0074
0.0091
0.0240
0.0212
0.0092
50
6
E­
Methylpyrene
0.0036
0.0058
0.0066
0.0211
0.0154
0.0167
91
51
6
F­
Methylpyrene
0.0056
0.0065
0.0080
0.0313
0.0191
0.0150
52
7
Retene
0
0
0
0
0
0
53
7
Benzonaphthothiophene
0.00000
0.00052
0.00053
0.00185
0.00360
0.00146
54
7
Benz(
a)
anthracene
0.00153
0.00437
0.00504
0.03413
0.03546
0.01654
55
7
7­
Methylbenz[
a]
anthracene
0.00000
0.00000
0.00000
0.00059
0.00072
0.00010
56
7
Chrysene
0.00153
0.00479
0.00560
0.03782
0.03208
0.02381
57
7
Benzo(
b+
j+
k)
fluoranthene
0.00000
0.00728
0.00952
0.10682
0.08899
0.04631
58
7
Benzo[
e]
pyrene
0.00020
0.00221
0.00349
0.05521
0.03757
0.01601
59
7
Benzo[
a]
pyrene
0.00073
0.00312
0.00353
0.06513
0.04550
0.01083
60
7
7­
Methylbenzo[
a]
pyrene
0.00000
0.00000
0.00000
0.00099
0.00161
0.00227
61
7
Indeno[
123­
cd]
pyrene
0.00000
0.00188
0.00257
0.04385
0.03870
0.00797
62
7
Dibenz(
ah+
ac)
anthracene
0.00000
0.00010
0.00007
0.00271
0.00343
0.00145
63
7
Benzo(
b)
chrysene
0.00000
0.00000
0.00000
0.00059
0.00230
0.00073
64
7
Benzo(
ghi)
perylene
0.00020
0.00548
0.00841
0.17844
0.12492
0.01512
65
7
Coronene
0.00047
0.00443
0.00603
0.14185
0.08425
0.00504
92
shows
the
ratios
between
PAH
emission
rates
for
H3
and
H1.
Note
that
the
ratios
show
definite
trends,
with
sample
H3
having
higher
emission
rates
for
the
higher
molecular
weight
compounds
than
sample
H1.
Also
note
that
the
variability
is
greatest
for
the
last
13
compounds.
This
was
generally
the
case.
Finally,
Figure
6.22
shows
the
one
case
where
there
were
very
large
differences
in
the
summer
between
different
samples
in
an
emission
category.
Sample
D3
was
a
single
diesel
vehicle.
It
had
much
higher
emissions
than
those
in
D2
and
D1,
and
clearly
had
a
much
different
PAH
emissions
profile
than
the
vehicles
in
D1.
On
the
other
hand,
D1
and
D2
had
very
similar
profiles.
Similar
variability
in
the
relative
emission
rates
of
individual
PAHs
was
seen
in
the
winter
samples
as
well.
Since
all
of
the
PAHs
are
present
in
all
of
the
samples,
there
are
no
distinct
marker
compounds
that
can
be
used
to
distinguish
between
PM
emissions
from
non­
smoking
gasoline
vehicles
and
those
of
smoking
gasoline
vehicles
or
diesels.
Thus,
separations
will
have
to
rely
on
differences
in
the
relative
emission
rates
of
the
compounds.
It
is
concluded
that
a
careful
analysis
of
the
emissions
variability
is
needed
if
these
compounds
are
going
to
be
used
in
source
profiles.
Such
an
analysis
is
beyond
the
scope
of
this
report.

A
comparison
between
emission
rates
of
individual
PAHs
for
the
summer
composite
medium
emitter
gasoline
and
the
composite
smoker
is
shown
in
Figure
6.23.
The
smoking
vehicles
have
more
low
and
high
molecular
weight
PAHs
than
the
medium
emission
rate
vehicles;
however,
the
variability
does
not
appear
to
be
any
greater
than
between
some
of
the
individual
samples
within
the
composites.
The
summer
smokers
and
high
emitters
were
in
closer
agreement
than
the
smokers
and
medium
emitters.
The
summer
profile
for
the
low
emitters
was
different
than
that
of
the
other
categories,
as
shown
in
Figure
6.24.
The
high
emitters
clearly
have
much
higher
emission
rates,
on
a
relative
basis,
of
the
low
and
high
molecular
weight
PAHs
than
do
the
low
emission
rate
vehicles.
Finally,
Figure
6.25
shows
the
ratios
between
the
diesel
vehicles
and
the
high
emitters.
The
profiles
are
very
different,
with
the
diesels
having
relatively
high
emission
rates
for
most
of
the
midrange
molecular
weight
compounds.
Results
were
similar
for
the
winter
samples.
The
biggest
difference
between
the
diesel
and
gasoline
vehicles
is
with
the
low
molecular
weight
PAHs.
All
of
these
species
remain
in
the
gas
phase.
Therefore,
they
may
not
be
good
indicators
of
the
PM
emission
rate.
More
work
is
needed
to
compare
gas
vs.
particle
bound
PAHs
as
the
tracers
for
these
vehicle
types.
Such
work
is
beyond
the
scope
of
this
report.

There
are
also
differences
in
PAH
emission
rate
profiles
between
the
three
FTP
phases.
The
differences
were
greatest
for
the
low
and
medium
emitting
gasoline
vehicles,
which
had
relatively
higher
emissions
of
both
the
low
and
high
molecular
weight
compounds
during
the
cold
start
(
bag
1)
than
during
bags
2
or
3.
This
difference
is
illustrated
in
Figures
6.26
and
6.27
which
show
the
ratios
between
bags
1
and
2
and
bags
1
and
3,
respectively,
for
the
summer
medium
emitting
vehicles.
For
the
high
emitting
vehicles
the
mid
molecular
weight
range
PAH
emission
rates
were
approximately
half
those
of
the
low
and
high
molecular
weight
PAHs
in
bag
1
compared
to
bag
2.
For
smoking
vehicles
the
differences
between
bags
1
and
2
were
mostly
in
the
high
molecular
weight
compounds,
which
had
rates
approximately
double
those
of
the
low
and
mid
molecular
weight
PAHs.
Diesels
had
higher
low
range
PAH
emissions
in
bag
1
than
2
and
3.
When
generating
profiles
for
source
apportionment
studies,
these
differences
should
be
considered
since
the
93
local
activity
may
be
different
from
that
assumed
in
the
FTP
weighting
of
the
three
bags.

While
Figures
6.20
 
6.27
give
a
good
visual
indication
of
differences
in
PAH
profiles,
it
is
desirable
to
provide
more
quantitative
comparisons.
That
was
done
by
examining
the
lumped
PAH
emission
rates
in
the
groups
defined
in
Tables
6.18
and
6.19.
Tables
6.20
and
6.21
give
the
average
PAH
emission
rates
for
the
groups
during
the
summer
and
winter.
Note
that
the
winter
data
includes
a
group
0,
which
consists
only
of
naphthalene.
It
is
not
present
in
the
summer
data
since
the
naphthalene
data
were
invalidated.
For
comparison
purposes,
Tables
6.22
and
6.23
give
the
percent
of
the
total
PAH
in
each
group
for
the
summer
and
winter.
Examination
of
these
tables
shows
that
variations
within
an
emitter
category
are
generally
as
large
as
differences
between
categories.
The
exception
is
the
difference
between
the
diesels
and
the
gasoline
vehicles.
As
noted
earlier,
the
diesels
have
a
lower
percentage
of
the
group
1
PAH
compounds
and
a
higher
percentage
of
the
group
2
and
3
PAH
compounds
(
see
Table
6.18
for
compound
group
designations).

Tables
6.24
and
6.25
give
the
percentage
contribution
of
each
group
of
PAH
compounds
to
the
total
PAH
as
a
function
of
FTP
phase
and
vehicle
category
for
the
summer
and
the
winter.
The
tables
show
that
group
1
compounds
are
a
lower
percentage
of
the
total
PAH
for
phase
2
than
for
phases
1
and
3.
The
difference
is
made
up
in
groups
3­
6.
The
difference
is
greatest
for
the
low
emission
gasoline
vehicles,
and
decreases
with
increasing
emission
rate.
The
diesels
show
a
similar,
although
small,
difference
between
FTP
phases.
As
noted
before,
the
diesel
vehicles
have
a
lower
group
1
PAH
percentage,
and
higher
group
2,3,5,
and
6
percentage
emissions.

PAH
emission
rates
were
also
expressed
as
their
percentage
of
the
total
PM
emission
rate.
Results
are
given
in
Table
6.26
for
the
summer
samples
and
Table
6.27
for
the
winter
samples.
These
percentages
are
confounded
by
the
fact
that
a
large
fraction
of
the
measured
PAHs
are
present
in
the
gas
phase.
Partitioning
between
the
gas
and
particle
phases
depends
on
a
number
of
factors
such
as
sample
temperature
and
gas
phase
concentration
and
is
likely
to
be
different
in
the
atmosphere
than
in
exhaust
samples.
Group
1­
2
compounds
are
most
likely
to
be
in
gas
phase;
group
3­
5
may
be
partitioned
between
phases
and
group
6­
7
compounds
are
likely
to
be
in
the
particle
phase.
The
NFRAQS
CMB
analysis
uses
all
of
the
PAHs
to
help
apportion
ambient
particulate
to
gasoline
and
diesel
vehicles.
Therefore,
it
makes
some
sense
to
examine
how
all
the
PAHs
compare
to
PM.
Expressed
in
this
manner
the
variability
between
samples
is
high,
and
distinctions
between
the
diesels
and
the
other
vehicles
is
blurred.

Table
6.20.
FTP
Summer
Emission
Rates
of
the
PAH
Compounds
Lumped
into
Groups
(
Naphthalene
not
included)

Sample
Sum
PAH
Group1
Group2
Group3
Group4
Group5
Group6
Group7
94
­­
mg/
mi
­­
L1
1.96
1.27
0.09
0.18
0.12
0.13
0.16
0.016
L2
1.82
1.23
0.10
0.12
0.12
0.10
0.13
0.010
L3
2.32
1.53
0.14
0.17
0.18
0.13
0.15
0.016
MI
3.46
2.61
0.12
0.26
0.21
0.12
0.10
0.041
M2
3.96
2.89
0.17
0.33
0.25
0.14
0.18
0.016
M3
9.12
6.45
0.37
0.94
0.76
0.28
0.25
0.063
H1
10.45
7.73
0.36
1.13
0.91
0.12
0.17
0.048
H2
4.44
2.62
0.21
0.42
0.64
0.20
0.31
0.047
H3
9.27
6.35
0.33
0.99
0.90
0.21
0.27
0.210
S1
18.54
12.15
0.88
2.87
1.45
0.64
0.47
0.095
S2
29.94
19.34
1.17
2.19
4.84
0.64
1.16
0.604
S3
45.41
35.92
1.14
4.37
2.45
0.49
0.64
0.398
D1
34.15
18.81
3.69
6.59
2.04
2.01
0.92
0.083
D2
49.80
19.49
6.05
13.86
4.41
3.54
2.21
0.246
D3
52.95
14.32
5.28
9.64
7.96
8.09
5.79
1.869
Table
6.21.
FTP
Winter
Emission
Rates
of
the
PAH
Compounds
Lumped
into
Groups
(
Naphthalene
included)

Sample
Sum
PAH
Group
0
Group
1
Group
2
Group
3
Group
4
Group
5
Group
6
Group
7
­­
mg/
mi­­
L2
4.45
1.46
1.24
0.056
0.130
0.613
0.302
0.634
0.0047
ML2
25.30
10.58
10.75
0.291
0.978
1.726
0.354
0.569
0.0539
ML1
10.01
4.09
2.88
0.169
0.370
1.059
0.405
1.024
0.0144
M3
18.93
5.25
7.22
0.312
1.132
2.590
0.829
1.578
0.0173
M2
16.75
6.52
5.60
0.237
0.349
2.543
0.340
1.080
0.0810
M1
23.54
8.45
8.69
0.452
1.273
2.606
0.685
1.343
0.0360
H1
75.71
36.83
25.00
1.102
2.532
5.240
1.131
3.199
0.6700
S3
19.61
8.19
6.97
0.345
1.147
1.905
0.419
0.607
0.0254
S2
39.26
14.94
15.35
0.627
2.500
4.228
0.658
0.886
0.0668
S1
170.96
66.54
64.77
2.539
10.575
18.020
2.386
4.726
1.4052
D5
7.87
0.81
2.04
0.575
1.754
1.462
0.716
0.506
0.0174
D4
12.12
3.12
3.91
0.681
1.869
1.416
0.479
0.491
0.1537
D3
21.79
4.60
7.59
1.519
3.985
2.198
1.162
0.593
0.1435
D2
42.55
5.07
14.77
3.768
10.776
4.558
2.783
0.702
0.1230
D1
58.36
10.79
19.98
4.975
12.046
5.903
3.151
1.211
0.3005
95
Table
6.22.
Percent
of
Total
Summer
FTP
PAH
Emitted
in
Each
Group
Sample
Group1
Group2
Group3
Group4
Group5
Group6
Group7
­­
%
of
total
PAH
without
naphthalene­­
L1
64.80
4.52
9.03
6.29
6.51
8.04
0.81
L2
67.63
5.70
6.74
6.65
5.58
7.17
0.53
L3
65.92
6.17
7.23
7.62
5.81
6.56
0.68
MI
75.52
3.61
7.43
5.94
3.39
2.91
1.19
M2
72.96
4.32
8.20
6.24
3.43
4.44
0.40
M3
70.77
4.07
10.33
8.34
3.08
2.72
0.69
H1
73.94
3.42
10.77
8.68
1.12
1.60
0.46
H2
58.99
4.80
9.43
14.39
4.45
6.88
1.06
H3
68.58
3.58
10.68
9.75
2.24
2.91
2.26
S1
65.53
4.72
15.49
7.80
3.44
2.51
0.51
S2
64.61
3.90
7.31
16.16
2.14
3.87
2.02
S3
79.10
2.50
9.62
5.40
1.09
1.41
0.88
D1
55.09
10.81
19.29
5.98
5.89
2.71
0.24
D2
39.14
12.16
27.83
8.85
7.10
4.43
0.49
D3
27.04
9.97
18.21
15.03
15.28
10.93
3.53
Table
6.23.
Percent
of
Total
Winter
PAH
Emitted
in
Each
Group
Sample
Group
1
Group
2
Group
3
Group
4
Group
5
Group
6
Group
7
­­%
of
total
PAH
(
without
naphthalene)­­
L2
41.68
1.89
4.36
20.53
10.12
21.26
0.156
ML2
73.03
1.98
6.64
11.72
2.40
3.86
0.366
ML1
48.62
2.85
6.25
17.89
6.85
17.29
0.244
M3
52.78
2.28
8.28
18.94
6.06
11.54
0.127
M2
54.75
2.31
3.41
24.86
3.32
10.56
0.792
M1
57.62
3.00
8.44
17.27
4.54
8.90
0.239
H1
64.31
2.83
6.51
13.48
2.91
8.23
1.723
S3
61.06
3.02
10.04
16.67
3.67
5.32
0.222
S2
63.13
2.58
10.28
17.38
2.71
3.64
0.275
S1
62.03
2.43
10.13
17.26
2.29
4.53
1.346
D5
28.83
8.14
24.81
20.69
10.13
7.15
0.246
D4
43.47
7.56
20.76
15.73
5.32
5.46
1.707
D3
44.14
8.84
23.19
12.79
6.76
3.45
0.835
D2
39.41
10.05
28.75
12.16
7.43
1.87
0.328
D1
42.01
10.46
25.32
12.41
6.62
2.55
0.632
96
Table
6.24.
Percent
of
Sumer
PAH
in
Each
Group
by
FTP
Phase
(
without
naphthalene)

Category
Group
1
Group
2
Group
3
Group
4
Group
5
Group
6
Group
7
­­
%
of
total
PAH
­­
Average
Phase
1
L
79.7
4.23
7.23
4.40
1.89
2.15
0.38
Average
Phase
2
L
47.0
5.97
8.90
11.34
11.72
14.20
0.91
Average
Phase
3
L
61.0
9.57
6.08
5.29
7.33
9.40
1.29
Average
Phase
1
M
78.9
3.39
8.83
5.64
1.38
1.11
0.75
Average
Phase
2
M
56.7
4.96
10.91
11.75
7.37
7.57
0.71
Average
Phase
3
M
73.3
4.87
8.73
6.05
3.04
3.34
0.63
Average
Phase
1
H
71.9
3.46
9.24
10.00
1.34
2.75
1.29
Average
Phase
2
H
65.0
3.84
11.87
11.14
2.97
3.69
1.45
Average
Phase
3
H
72.5
3.94
9.83
8.52
1.91
2.44
0.88
Average
Phase
1
S
72.2
3.80
8.95
8.14
1.75
3.14
2.06
Average
Phase
2
S
70.9
3.14
10.45
10.45
2.05
2.18
0.80
Average
Phase
3
S
73.3
3.50
10.29
8.02
1.67
2.16
1.04
Average
Phase
1
D
37.3
14.40
24.93
8.32
7.41
6.55
1.11
Average
Phase
2
D
31.7
11.71
25.09
12.49
10.86
6.69
1.49
Average
Phase
3
D
48.6
7.79
15.76
9.17
10.33
6.24
2.08
Table
6.25.
Percent
of
Winter
PAH
in
Each
Group
by
FTP
Phase
(
without
naphthalene)
Category
Group
1
Group
2
Group
3
Group
4
Group
5
Group
6
Group
7
­­%
of
total
PAH­­
Phase
1
L2
64.25
2.41
5.27
16.68
4.08
7.21
0.09
Phase
2
L2
12.92
1.11
3.68
27.61
17.18
37.28
0.22
Phase
3
L2
47.32
2.36
3.07
12.79
10.58
23.66
0.21
Average
Phase
1
ML
74.30
2.53
6.60
11.24
1.54
3.26
0.52
Average
Phase
2
ML
57.21
1.86
6.52
16.72
5.69
11.82
0.18
Average
Phase
3
ML
68.77
2.46
6.31
9.64
3.97
8.64
0.20
Average
Phase
1
M
63.86
2.86
8.37
16.92
2.31
5.26
0.42
Average
Phase
2
M
48.20
2.20
5.57
23.31
6.44
13.96
0.31
Average
Phase
3
M
57.39
3.23
9.22
14.43
5.21
10.27
0.25
Phase
1
H1
68.12
2.84
6.82
12.65
1.23
4.93
3.42
Phase
2
H1
58.94
2.75
6.25
15.60
4.46
11.43
0.58
Phase
3
H1
76.02
3.27
6.59
5.84
2.19
5.92
0.17
Average
Phase
1
S
64.27
2.66
10.40
16.16
1.28
3.26
1.96
Average
Phase
2
S
58.59
2.19
10.08
19.23
3.72
5.89
0.31
Average
Phase
3
S
66.07
2.93
9.70
14.72
2.15
3.56
0.86
Average
Phase
1
D
45.18
10.31
25.21
10.32
5.01
2.87
1.10
Average
Phase
2
D
39.17
9.42
25.98
14.36
7.87
2.83
0.37
Average
Phase
3
D
40.04
9.91
25.61
13.13
7.05
3.49
0.76
97
Table
6.26.
PAH
Expressed
as
a
Percent
of
Summer
PM
Emissions
by
Group
Sample
Group1
Group2
Group3
Group4
Group5
Group6
Group7
­­
Percent
of
PM
Emission
Rate
­­

L1
62.6
4.37
8.73
6.08
6.29
7.77
0.78
L2
17.9
1.51
1.79
1.77
1.48
1.90
0.14
L3
16.2
1.52
1.78
1.87
1.43
1.61
0.17
MI
11.7
0.56
1.15
0.92
0.52
0.45
0.18
M2
14.1
0.83
1.58
1.21
0.66
0.86
0.08
M3
21.3
1.22
3.10
2.51
0.92
0.82
0.21
H1
17.0
0.79
2.48
2.00
0.26
0.37
0.11
H2
4.3
0.35
0.69
1.05
0.33
0.50
0.08
H3
6.0
0.31
0.94
0.86
0.20
0.26
0.20
S1
2.4
0.18
0.58
0.29
0.13
0.09
0.02
S2
11.9
0.72
1.35
2.98
0.39
0.71
0.37
S3
19.3
0.61
2.35
1.32
0.27
0.35
0.21
D1
8.8
1.73
3.09
0.96
0.95
0.43
0.04
D2
6.8
2.10
4.81
1.53
1.23
0.77
0.09
D3
0.4
0.13
0.24
0.20
0.20
0.15
0.05
Table
6.27.
PAH
Expressed
as
a
Percent
of
Winter
PM
Emissions
by
Group
Sample
Group
0
Group
1
Group
2
Group
3
Group
4
Group
5
Group
6
Group
7
­­
Percent
of
PM
Emission
Rate­­
L2
20.42
17.34
0.79
1.81
8.54
4.21
8.85
0.065
ML2
49.81
50.64
1.37
4.60
8.13
1.67
2.68
0.254
ML1
11.59
8.15
0.48
1.05
3.00
1.15
2.90
0.041
M3
10.46
14.38
0.62
2.26
5.16
1.65
3.14
0.034
M2
11.94
10.27
0.43
0.64
4.66
0.62
1.98
0.149
M1
10.65
10.96
0.57
1.60
3.28
0.86
1.69
0.045
H1
17.32
11.76
0.52
1.19
2.46
0.53
1.50
0.315
S3
4.07
3.47
0.17
0.57
0.95
0.21
0.30
0.013
S2
2.48
2.55
0.10
0.42
0.70
0.11
0.15
0.011
S1
6.09
5.93
0.23
0.97
1.65
0.22
0.43
0.129
D5
0.23
0.59
0.17
0.51
0.42
0.21
0.15
0.005
D4
0.73
0.92
0.16
0.44
0.33
0.11
0.12
0.036
D3
0.86
1.42
0.28
0.75
0.41
0.22
0.11
0.027
D2
0.81
2.35
0.60
1.72
0.73
0.44
0.11
0.020
D1
1.27
2.36
0.59
1.42
0.70
0.37
0.14
0.035
PAH
hazardous
air
pollutants.
The
EPA
has
been
constructing
emissions
inventories
for
98
hazardous
air
pollutants
(
HAPs).
Polycyclic
organic
matter
(
POM)
is
on
the
Office
of
Air
Quality
Planning
and
Standards
(
OAQPS)
list
of
40
priority
HAPs.
There
are
a
large
number
of
compounds
that
qualify
as
POM.
EPA
has
recognized
that
it
is
not
feasible
to
inventory
all
of
these
compounds,
and
has
recommended
that
attention
be
focused
on
three
options.
The
first
option
is
to
inventory
the
sum
of
the
seven
PAHs
(
7­
PAH)
that
have
been
determined
by
the
International
Agency
for
Research
on
Cancer
(
IARC)
to
be
animal
carcinogens.
These
are
benz(
a)
anthracene,
benzo(
a)
pyrene,
benzo(
b)
fluoranthene,
benzo(
k)
fluoranthene,
chrysene,
dibenz(
a,
h)
anthracene,
and
indeno(
1,2,3­
cd)
pyrene.
The
second
option
is
to
expand
this
list
to
16
compounds
(
16­
PAH)
by
including
acenaphthene,
acenaphthylene,
anthracene,
benzo(
ghi)
perylene,
fluoranthene,
fluorene,
naphthalene,
phenanthrene
and
pyrene.
The
third
option
is
to
use
extractable
organic
matter
(
EOM)
as
a
surrogate
for
the
POM.
All
of
these
compounds
were
measured
in
this
study,
although
benzo(
b)
fluoranthene
was
not
separated
from
either
benzo(
k)
fluoranthene
or
benzo(
j)
fluoranthene.
Thus,
the
7­
PAH
sum
from
this
study
includes
benzo(
j)
fluoranthene.
EOM
was
not
determined
in
this
study,
however,
OC
as
measured
by
the
TOR
method
is
similar
to
the
EOM.
Table
6.28
lists
the
7­
PAH,
16­
PAH
and
OC
determined
for
the
winter
study.
Data
are
not
summarized
for
the
summer
study
due
to
the
lack
of
naphthalene
data
and
the
problems
with
some
of
the
OC
measurements.
Correlations
were
run
between
the
three
classes,
both
including
and
excluding
the
diesels,
and
including
each
phase
of
the
FTP
rather
than
the
composite
FTP
given
in
the
Table.
Not
surprisingly,
correlations
were
low.

Table
6.28.
Winter
FTP
Emission
Rates
of
HAP
Classifications
Sample
7­
PAH
mg/
mi
16­
PAH
mg/
mi
OC
mg/
mi
L2
0.0234
11.78
5.7
ML2
0.2905
44.64
12.5
ML1
0.0163
10.27
13.7
M2
0.0398
19.10
19.9
M3
0.6763
87.64
27.8
M1
0.0096
2.25
55.4
H1
0.1121
4.78
64.1
S3
0.1028
6.80
158.2
S1
0.0916
8.83
431.7
S2
0.2184
16.17
455.1
D4
0.0038
2.59
47.5
D3
0.0344
12.57
57.8
D5
0.0087
5.93
142.2
D2
0.0110
8.92
152.6
D1
0.0445
9.82
259.6
Steranes/
hopanes.
It
has
been
suggested
that
steranes
and
hopanes
make
good
marker
compounds
for
mobile
source
PM,
and
that
the
emission
rates
for
gasoline
and
diesel
99
vehicles
may
be
different
enough
that
their
PM
emissions
can
be
distinguished
from
each
other
(
Rogge
et
al.,
1993).
Steranes
and
hopanes
were
measured
in
the
same
sample
extracts
as
the
PAHs.
Thus,
their
emission
rates
are
for
the
same
composite
samples
as
the
PAHs.
Table
6.29
lists
the
individual
compounds
that
were
measured
and
gives
the
average
uncertainty
for
all
tests
for
the
composite
FTP.
Average
uncertainties
were
lower
for
the
winter
measurements
compared
to
the
summer
measurements.
The
individual
uncertainties
tended
to
increase
as
the
concentration
increased.

Table
6.30
gives
the
average
FTP
emission
rates
of
each
compound
for
the
summer.
Missing
data
indicate
that
the
emission
rate
for
the
compound
was
less
than
the
measurement
uncertainty.
Emission
rates
are
very
low.
The
average
ratio
of
the
emission
rate
to
the
uncertainty
in
the
rate
for
the
reported
measurements
ranged
from
a
high
of
5.5
for
sitostane
to
1.9
for
trisnorhopane­
1.
Several
of
the
compound's
emission
rates
averaged
only
two
to
three
times
the
uncertainty.
Thus,
it
should
be
recognized
that
significant
errors
are
probable
in
the
rates.
The
next
to
last
row
of
the
table
gives
the
sum
of
the
emission
rates
of
the
measured
compounds,
followed
by
its
percentage
of
the
PM
emission
rate.
Thus,
it
is
seen
that
the
measured
compounds
accounted
for
0.78%
of
the
PM
emissions
for
the
low
emitters,
but
only
0.086%
of
the
PM
for
the
diesels.

Table
6.31
gives
the
average
FTP
emission
rates
for
the
winter.
A
larger
portion
of
the
measurements
in
the
winter
were
below
the
measurement
uncertainty,
resulting
in
more
non
detectables
than
in
the
summer.
The
average
ratio
of
emission
rate
to
uncertainty
in
the
emission
rate
ranged
from
a
high
of
6.1
for
diasterane­
1
to
2.4
for
cholestane­
2.
The
sum
of
the
emission
rates
for
each
category
is
given
in
the
next
to
last
row
of
the
table,
followed
by
its
percentage
of
the
total
PM
emission
rate.
Winter
hopane
and
sterane
emissions
constituted
a
lower
percentage
of
the
PM
than
during
the
summer,
by
factors
of
two
to
four.
This
difference
can
not
be
attributed
entirely
to
the
greater
number
of
non
detectable
compounds
in
the
winter.
The
cause
of
the
reduction
is
not
known.

Figures
6.28
and
6.29
compare
the
summer
emission
rates
of
the
hopanes
and
steranes
for
each
category.
The
pattern
of
emissions
is
very
similar
with
the
exceptions
of
norhopane­
2,
hopane­
2
and
hopane­
3
which
were
not
detectable
for
some
emitter
categories.
Figure
6.30
shows
the
similar
comparison
for
the
winter.
The
low
and
medium
low
emitters
were
excluded
from
this
comparison
because
of
the
large
number
of
non­
detectable
compounds.
Also,
the
five
highest
molecular
weight
compounds
were
omitted
from
the
comparison
since
none
were
detected
in
the
winter.
The
highest
emission
rates
are
for
the
smokers.
Both
the
diesels
and
smokers
have
higher
relative
emission
rates
of
the
lower
molecular
weight
compounds
than
the
medium
and
high
emitters.
Overall,
it
is
concluded
that
these
compounds
may
make
excellent
tracers
for
mobile
source
particulate
emissions.
However,
there
is
no
indication
that
they
can
be
used
to
distinguish
between
diesel
and
gasoline
vehicle
PM
emissions.
It
may
be
that
the
primary
source
of
these
compounds
is
motor
oil.
However,
no
measurements
of
these
species
were
made
in
either
the
fuel
or
motor
oil.

Tables
6.32
through
6.36
give
the
average
emission
rates
of
the
hopanes
and
steranes
by
FTP
phase
for
each
emitter
category.
Only
those
rates
that
were
greater
than
the
measurement
uncertainty
are
included
in
the
averages.
As
expected,
many
of
the
100
compounds
show
a
distinct
difference
in
rates
between
phases,
indicating
that
profiles
used
for
source
apportionment
might
benefit
from
using
the
actual
cold
start/
stabilized
running/
hot
start
data
for
a
given
area
rather
than
using
the
FTP
default
weighting.

Table
6.29.
Average
Uncertainty
in
Hopane
and
Sterane
Measurements
Compound
Emission
Rate,
mg/
mi
Summer
Winter
ergostane
0.00813
0.00106
sitostane
0.00453
0.00151
diasterane­
1
0.00725
0.00119
diasterane­
2
0.00346
0.00103
cholestane­
1
0.00791
0.00193
cholestane­
2
0.00368
0.00221
cholestane­
3
0.01338
0.00225
trisnorhopane­
1
0.00660
0.00089
trisnorhopane­
2
0.00413
0.00106
norhopane­
1
0.01374
0.00286
norhopane­
2
0.00545
0.00209
hopane­
1
0.00760
0.00143
hopane­
2
0.00445
0.00155
homohopane­
1
0.00277
0.00077
homohopane­
2
0.00258
0.00077
hopane­
3
0.00234
0.00077
bishomohopane­
1
0.00299
0.00077
bishomohopane­
2
0.00273
0.00077
101
Table
6.30.
Summer
Average
FTP
Emission
Rates
for
Hopanes
and
Steranes
Compound
Ave
L
Ave
M
Ave
H
Ave
S
Ave
D
mg/
mi
ergostane
0.0031
0.0050
0.0123
0.0534
0.1168
sitostane
0.0043
0.0054
0.0082
0.0357
0.0874
diasterane­
1
0.0012
0.0047
0.0044
0.0336
0.3135
Diasterane­
2
0.0014
0.0034
0.0041
0.0257
0.0471
cholestane­
1
0.0033
0.0047
0.0089
0.0475
0.1009
cholestane­
2
0.0014
0.0042
0.0028
0.0166
0.0355
cholestane­
3
0.0027
0.0033
0.0070
0.0452
0.0915
Trisnorhopane­
1
0.0016
0.0041
0.0037
0.0173
0.0427
Trisnorhopane­
2
0.0010
0.0032
0.0033
0.0130
0.0309
norhopane­
1
0.0089
0.0087
0.0145
0.0598
0.1561
norhopane­
2
ND
0.0125
ND
0.0053
ND
hopane­
1
0.0071
0.0080
0.0100
0.0349
0.1074
hopane­
2
0.0020
0.0080
0.0036
0.0277
Homohopane­
1
0.0035
0.0036
0.0051
0.0148
0.0500
Homohopane­
2
0.0023
0.0032
0.0039
0.0106
0.0387
hopane­
3
ND
0.0028
ND
ND
ND
Bishomohopane­
1
0.0019
0.0023
0.0031
0.0079
0.0284
Bishomohopane­
2
0.0016
0.0024
0.0024
0.0054
0.0229
Sum
0.047
0.090
0.094
0.43
1.30
%
of
PM
0.78
0.37
0.13
0.15
0.086
102
Table
6.31.
Winter
Average
FTP
Emission
Rates
for
Hopanes
and
Steranes
Compound
L2
Ave
ML
Ave
M
H
Ave
S
Ave
D
mg/
mi
Ergostane
ND
0.00108
ND
ND
0.01064
0.00356
Sitostane
ND
0.00201
ND
0.00532
0.03344
0.00994
diasterane­
1
ND
ND
ND
0.00326
0.01493
0.01019
diasterane­
2
ND
ND
ND
0.00179
0.01006
0.00521
cholestane­
1
ND
0.00148
0.00214
0.00350
0.02215
0.00998
cholestane­
2
ND
ND
0.00351
0.00342
0.01343
0.00951
cholestane­
3
ND
0.00112
0.00352
0.00328
0.02029
0.00785
trisnorhopane­
1
ND
0.00135
0.00188
0.00323
0.00421
0.00395
trisnorhopane­
2
ND
ND
ND
ND
0.00486
0.00362
norhopane­
1
0.00376
0.00766
0.00939
0.01644
0.02347
0.01965
norhopane­
2
0.00641
0.00623
0.01249
0.01623
0.01105
0.01323
hopane­
1
0.00427
0.00501
0.00721
0.01203
0.01533
0.01251
hopane­
2
ND
0.00372
0.00485
0.00818
0.00719
0.00811
homohopane­
1
ND
0.00134
ND
ND
ND
ND
homohopane­
2
ND
ND
ND
ND
ND
ND
hopane­
3
ND
ND
ND
ND
ND
ND
bishomohopane­
1
ND
ND
ND
ND
ND
ND
bishomohopane­
2
ND
ND
ND
ND
ND
ND
Sum
0.014
0.031
0.045
0.077
0.19
0.12
%
of
PM
0.20
0.11
0.073
0.036
0.030
0.021
Table
6.32.
Average
Summer
Emission
Rates
of
Hopanes
and
Steranes
by
FTP
Phase
Composite
Ergostane
sitostane
diasterane­
1
diasterane­
2
Cholestane­
1
cholestane­
2
­­
mg/
mi
­­
L
Phase
1
0.00251
0.00388
0.00183
0.00159
0.00231
0.00115
L
Phase
2
0.00304
0.00347
0.00098
0.00133
0.00327
0.00112
L
Phase
3
0.00363
0.00594
0.00121
0.00152
0.00401
0.00198
M
Phase
1
0.00719
0.00947
0.00618
0.00491
0.00872
0.00344
M
Phase
2
0.00495
0.00556
0.00309
0.00325
0.00596
0.00193
M
Phase
3
0.00336
0.00386
0.00221
0.00211
0.00474
0.00156
H
Phase
1
0.01647
0.01076
0.00881
0.00735
0.01319
0.00415
H
Phase
2
0.00981
0.00665
0.00272
0.00251
0.00759
0.00224
H
Phase
3
0.01408
0.00919
0.00448
0.00481
0.00834
0.00274
S
Phase
1
0.08063
0.05678
0.05444
0.04332
0.07880
0.02625
S
Phase
2
0.04075
0.02601
0.02403
0.01812
0.03353
0.01252
S
Phase
3
0.05762
0.03859
0.03658
0.02720
0.05090
0.01709
D
Phase
1
0.08982
0.06675
0.08096
0.03816
0.07090
0.02550
D
Phase
2
0.10828
0.08241
0.33561
0.04590
0.09858
0.03317
D
Phase
3
0.15304
0.11239
0.44692
0.05602
0.12788
0.04729
103
Table
6.33.
Average
Summer
Emission
Rates
of
Hopanes
and
Steranes
by
FTP
Phase
Composite
cholestane­
3
trisnorhopane­
1
trisnorhopane­
2
norhopane­
1
Norhopane­
2
hopane­
1
­­
mg/
mi
­­
L
Phase
1
0.00297
0.00162
0.00115
0.00896
ND
0.00581
L
Phase
2
0.00177
ND
ND
0.00574
ND
0.00513
L
Phase
3
0.00398
0.00183
0.00114
0.01451
ND
0.01166
M
Phase
1
0.00622
0.00345
0.00314
0.02076
ND
0.01219
M
Phase
2
0.00496
0.00222
0.00210
0.01098
ND
0.00703
M
Phase
3
0.00229
0.00142
ND
0.00926
ND
0.00649
H
Phase
1
0.01214
0.00506
0.00338
0.01911
ND
0.01249
H
Phase
2
0.00503
0.00339
0.00203
0.01196
ND
0.00825
H
Phase
3
0.00684
0.00342
0.00219
0.01578
ND
0.01150
S
Phase
1
0.07110
0.02698
0.01968
0.09039
0.00616
0.05341
S
Phase
2
0.03339
0.01274
0.01002
0.04619
ND
0.02654
S
Phase
3
0.04860
0.01877
0.01386
0.06329
ND
0.03715
D
Phase
1
0.05213
0.03274
0.02372
0.11069
ND
0.08073
D
Phase
2
0.09200
0.03674
0.02845
0.14576
ND
0.10265
D
Phase
3
0.12008
0.06138
0.04100
0.20999
ND
0.13661
104
Table
6.34.
Average
Summer
Emission
Rates
of
Hopanes
and
Steranes
by
FTP
Phase
Composite
hopane­
2
homohopane­
1
homohopane­
2
hopane­
3
Bishomohopane­
1
bishomohopane­
2
­­
mg/
mi
­­
L
Phase
1
ND
0.00285
0.00098
ND
0.00091
ND
L
Phase
2
ND
0.00256
0.00169
ND
0.00140
0.00138
L
Phase
3
0.00293
0.00560
0.00428
ND
0.00358
0.00282
M
Phase
1
ND
0.00650
0.00522
ND
0.00372
0.00315
M
Phase
2
ND
0.00271
0.00205
ND
0.00160
0.00170
M
Phase
3
ND
0.00324
0.00239
ND
0.00113
H
Phase
1
ND
0.00619
0.00459
ND
0.00343
0.00270
H
Phase
2
ND
0.00402
0.00313
ND
0.00246
0.00190
H
Phase
3
ND
0.00619
0.00475
ND
0.00392
0.00319
S
Phase
1
0.00555
0.0217
0.0160
ND
0.01225
0.00828
S
Phase
2
0.00274
0.0111
0.0079
ND
0.00578
0.00422
S
Phase
3
0.00366
0.0169
0.0117
ND
0.00874
0.00560
D
Phase
1
ND
0.0412
0.0286
ND
0.0226
0.0195
D
Phase
2
ND
0.0474
0.0391
ND
0.0267
0.0219
D
Phase
3
ND
0.0615
0.0454
ND
0.0359
0.0274
Table
6.35.
Average
Winter
Emission
Rates
of
Steranes
by
FTP
Phase
Composite
Ergostane
sitostane
diasterane­
1
diasterane­
2
cholestane­
1
Cholestane­
2
cholestane­
3
­­
mg/
mi
­­

ML
Phase
1
ND
ND
ND
ND
ND
ND
ND
ML
Phase
2
ND
ND
ND
ND
ND
ND
ND
ML
Phase
3
0.004
0.0074
0.0030
0.0029
0.0055
0.0022
0.0041
M
Phase
1
ND
ND
ND
ND
ND
ND
ND
M
Phase
2
ND
ND
ND
ND
ND
ND
ND
M
Phase
3
ND
ND
ND
ND
0.0082
0.013
0.014
H
Phase
1
ND
0.024
0.015
0.0080
0.016
0.015
0.015
H
Phase
2
ND
ND
ND
ND
ND
ND
ND
H
Phase
3
ND
ND
ND
ND
ND
ND
ND
S
Phase
1
0.030
0.067
0.037
0.026
0.071
0.033
0.061
S
Phase
2
0.011
0.025
0.013
0.0089
0.016
0.012
0.016
S
Phase
3
0.0064
0.025
0.016
0.010
0.015
0.010
0.020
D
Phase
1
0.011
0.023
0.022
0.010
0.025
0.013
0.017
D
Phase
2
ND
0.015
0.0098
0.0054
0.0079
0.015
0.010
D
Phase
3
0.0097
0.023
0.012
0.0081
0.012
0.0098
0.014
105
Table
6.36.
Average
Winter
Emission
Rates
of
Hopanes
by
FTP
Phase
Composite
Trisnorhopane­
1
Trisnorhopane­
2
norhopane­
1
norhopane­
2
hopane­
1
hopane­
2
­­
mg/
mi
­­

ML
Phase
1
0.0030
ND
0.0076
0.0076
0.0052
0.0038
ML
Phase
2
ND
ND
0.0069
0.0066
0.0044
0.0043
ML
Phase
3
0.0027
0.0025
0.0092
0.0045
0.0060
0.0026
M
Phase
1
0.0027
ND
0.0084
0.0094
0.0062
0.0054
M
Phase
2
0.0028
0.0017
0.0087
0.012
0.0065
0.0063
M
Phase
3
0.0047
0.0025
0.011
0.016
0.0094
0.0024
H
Phase
1
0.0062
0.0038
0.025
0.022
0.018
0.0025
H
Phase
2
0.0043
ND
0.016
0.017
0.012
0.012
H
Phase
3
ND
ND
0.011
0.011
0.0077
0.0072
S
Phase
1
0.0048
0.0078
0.016
0.0082
0.011
0.0056
S
Phase
2
0.0042
0.0043
0.027
0.014
0.017
0.0094
S
Phase
3
0.0037
0.0038
0.022
0.0072
0.015
0.0041
D
Phase
1
0.0052
0.0060
0.026
0.0087
0.016
0.0036
D
Phase
2
0.0043
0.0031
0.017
0.015
0.011
0.0092
D
Phase
3
0.0039
0.0037
0.020
0.014
0.013
0.0094
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
7.
Emissions
Inventory
It
is
intended
that
the
data
presented
in
this
report
be
used
by
state
and
federal
agencies
to
improve
PM
emission
inventory
estimates.
A
complete
examination
of
the
PM
emission
inventory
is
beyond
the
scope
of
this
report.
However,
a
simple,
preliminary
estimation
of
the
light­
duty
exhaust
emissions
inventory
for
the
Colorado
Front
Range
area
was
made
using
the
following
assumptions.
First,
it
is
assumed
that
the
FTP
weighting
of
cold
starts,
hot
starts,
and
hot
stabilized
running
represents
actual
in­
use
activity
data
for
the
region.
Second,
we
assume
that
the
FTP
is
an
adequate
representation
of
driving
in
the
region,
even
though
it
is
well
known
that
it
omits
some
driving
conditions
that
may
increase
particulate
emissions.
Third,
it
is
assumed
that
the
smoking
gasoline
vehicle
VMT
is
0.1%
of
the
total.
This
is
based
on
a
limited
visual
observation
survey
of
smoking
vehicles
done
for
this
program.
Fourth,
we
assume
that
the
vehicles
tested
in
this
program
represent
an
unbiased
sample
with
an
adequate
population
to
provide
reasonable
average
emission
rates.
All
of
these
assumptions
could
result
in
low
estimates
of
real­
world
PM
emissions.

Table
7.1
gives
the
average
PM
emission
rates
for
the
summer
period
used
in
this
exercise.
No
distinction
was
made
between
truck
and
car
PM
emission
rates,
since
the
population
of
vehicles
tested
was
judged
to
be
too
small
to
subdivide
further.
The
averages
do,
however,
contain
approximately
30%
light­
duty
trucks.
Since
only
9
smoking
vehicles
were
tested
in
the
summer,
the
average
of
the
summer
and
winter
indoor
smoking
vehicle
emission
rates
(
300
mg/
mi)
was
used
for
this
category.
Also,
both
summer
and
winter
indoor
diesel
tests
were
combined
to
obtain
an
average
light­
duty
diesel
emission
rate
of
637
mg/
mi
for
1972­
80
vehicles
and
326
mg/
mi
for
1981­
85
vehicles.
No
newer
lightduty
diesels
were
tested.
Newer
diesel
vehicles
tested
in
the
summer
were
in
the
heavyduty
classification.
Therefore
an
emission
rate
of
80
and
100
mg/
mi
was
assumed
for
the
1991­
96
and
1986­
1990
diesel
vehicles,
respectively.
The
winter
PM
emission
rates
are
given
in
Table
7.2.
Since
there
was
little
difference
between
indoor
and
outdoor
diesel
PM
emission
rates,
the
same
average
rate
was
used
for
the
winter
as
the
summer.

Table
7.1.
Summer
PM
Emission
Rates
Vehicle
Type
1991­
96
1986­
90
1981­
85
1972­
1980
PM
Emission
Rate,
mg/
mi
LDGV
2.8
44.4
47.4
95.5
LDGT1
2.8
44.4
47.4
95.5
LDGT2
2.8
44.4
47.4
95.5
LDDV
80
100
579
637
LDDT
80
100
326
637
LDGV
Smoker
330
330
330
330
137
Table
7.2.
Winter
PM
Emission
Rates
Vehicle
Type
1991­
96
1986­
90
1981­
85
1972­
1980
PM
Emission
Rate,
mg/
mi
LDGV
24.9
28.5
48.2
82.6
LDGT1
24.9
28.5
48.2
82.6
LDGT2
24.9
28.5
48.2
82.6
LDDV
80
100
579
637
LDDT
80
100
326
637
LDGV
Smoker
330
330
330
330
Table
7.3
gives
the
Colorado
Front
Range
area
daily
vehicle
miles
traveled
(
VMT)
for
each
category.
With
the
exception
of
the
smoking
vehicle
VMT,
these
data
were
supplied
by
the
State
of
Colorado.
Smoking
vehicle
VMT
was
arbitrarily
assigned
to
the
two
oldest
year
categories,
since
most
smoking
vehicles
are
relatively
old.
Multiplying
the
emission
rates
and
the
daily
VMT
gives
the
tons
of
PM
per
day
in
each
of
the
categories.
Results
for
the
summer
and
winter
are
given
in
Tables
7.4
and
7.5.

Table
7.3.
Front
Range
Daily
VMT
by
Vehicle
Category.

Vehicle
Type
Daily
VMT/
1000
1991­
96
fraction
1986­
90
fraction
1981­
85
fraction
1972­
80
fraction
LDGV
29,308
0.501
0.316
0.137
0.046
LDGT1
9,883
0.424
0.323
0.162
0.091
LDGT2
4,936
0.373
0.323
0.184
0.12
LDDV
185
0.122
0.057
0.651
0.169
LDDT
114
0.131
0.114
0.638
0.117
LDGV
Smoker
400
0.0
0.0
0.5
0.5
138
Table
7.4.
Front
Range
Summer
PM
Emissions
by
Vehicle
Category.

Vehicle
type
1991­
96
1986­
90
1981­
85
1972­
80
Total
Metric
tons/
day
LDGV
0.041
0.41
0.19
0.13
0.77
LDGT1
0.012
0.14
0.076
0.086
0.32
LDGT2
0.005
0.071
0.043
0.057
0.18
LDDV
0.002
0.001
0.070
0.020
0.092
LDDT
0.001
0.001
0.024
0.008
0.035
LDGV
Smoker
0
0
0.066
0.066
0.13
Total
1.53
Table
7.5.
Front
Range
Winter
PM
Emissions
by
Vehicle
Category.

Vehicle
type
1991­
96
1986­
90
1981­
85
1972­
80
Total
Metric
tons/
day
LDGV
0.37
0.26
0.19
0.11
0.93
LDGT1
0.10
0.091
0.077
0.074
0.35
LDGT2
0.046
0.045
0.044
0.049
0.18
LDDV
0.002
0.001
0.070
0.020
0.09
LDDT
0.001
0.001
0.023
0.008
0.035
LDGV
smoker
0
0
0.066
0.066
0.132
Total
1.72
A
total
of
1.53
and
1.72
tons
per
day
of
PM­
10
is
estimated
to
be
emitted
in
this
area
in
the
summer
and
winter,
respectively.
For
the
summer
the
largest
contributing
category
(
26.8%)
is
the
1986­
90
light­
duty
gasoline
vehicles.
All
1991­
96
light­
duty
cars
and
trucks
combined
account
for
only
3.8%
of
the
total.
Smoking
vehicles,
even
at
an
estimated
0.1%
of
the
VMT
contribute
8.5%
of
the
total.
Since
this
number
is
highly
uncertain,
the
smoking
vehicle
contribution
may
be
significantly
larger.
The
small
number
of
light­
duty
diesel
cars
and
trucks
also
account
for
8.3%
of
the
total.
For
the
winter,
the
largest
single
category
is
the
1990­
96
vehicles,
which
contribute
21.2%
of
the
total
PM­
10.
This
value
is
also
highly
uncertain
since
the
average
emission
rate
for
this
category
comes
from
only
nine
vehicles,
two
of
which
had
very
high
emission
rates.
The
whole
fleet
average
PM­
10
emission
rate
for
the
summer
and
winter,
respectively,
was
33.9
and
38.4
mg/
mi
Given
the
number
of
assumptions
that
have
gone
into
these
estimates
it
is
obvious
that
significant
errors
are
possible.
Despite
that,
the
overall
estimate
is
probably
better
than
that
currently
used
since
the
database
is
greatly
improved.
In
addition,
the
exercise
is
useful
in
pointing
towards
areas
where
improved
data
are
needed.
139
8.
Conclusions
The
major
conclusions
from
this
study
are:

 
Late
model
(
1991­
96)
light­
duty
gasoline
vehicles
have
low
exhaust
PM
emission
rates
under
FTP
conditions.
 
Average
exhaust
PM
emission
rates
are
higher
for
older
model
year
vehicles
than
for
late
model
year
vehicles.
 
PM
emission
rates
are
significantly
higher
in
the
cold
start
(
phase
1)
portion
of
the
FTP
than
in
the
hot
stabilized
or
hot
start
portions
(
phases
2
and
3).
 
Exhaust
PM
emission
rates
increase
at
low
temperatures.
As
expected,
the
increase
is
for
FTP
phase
1
only.
 
An
observed
decrease
in
the
average
PM
emission
rate
under
FTP
conditions
between
the
summer
and
the
winter
is
consistent
with
previously
reported
effects
of
oxygenated
fuels.
 
Average
PM
emission
rates
are
similar
for
the
FTP
and
the
IM240,
suggesting
that
the
IM240
can
be
used
to
estimate
fleet
average
PM
emission
rates.
 
Impactor
size
distributions
show
that
more
than
91%
of
the
PM­
10
is
smaller
than
PM­
2.5.
Mass
median
diameters
ranged
from
0.12
µ
m
to
0.18
µ
m.
 
Continuous
particle
counts
show
that
particle
emission
rates
are
highly
variable
during
FTP
tests,
with
the
highest
rates
occurring
during
cold
start
and
accelerations.
 
Most
of
the
particulate
mass
is
carbonaceous
matter.
The
average
fraction
of
carbonaceous
matter
that
is
organic
is
higher
for
gasoline
vehicles
than
diesel
vehicles.
 
Sulfate
emissions
averaged
less
than
1
mg/
mi
for
the
gasoline
fleet,
but
were
higher
for
the
diesel
fleet.
 
Gasoline
and
diesel
vehicles
emit
the
same
PAH
compounds.
However,
there
are
differences
in
the
relative
emission
rates
of
some
of
the
compounds.
These
differences
may
be
useful
in
source
apportionment
studies
for
separating
diesel
from
gasoline
PM
emissions.
 
Gasoline
and
diesel
vehicles
emit
the
same
hopane
and
sterane
compounds.
There
is
little
difference
in
the
relative
emission
rates.
Therefore,
these
species
are
not
useful
in
separating
diesel
and
gasoline
PM
emissions
in
source
apportionment
studies.
 
The
fleet
average
PM
emission
rate
for
the
Colorado
Front
Range
was
estimated
at
33.9
mg/
mi
for
the
summer
and
38.4
mg/
mi
for
the
winter.
The
largest
uncertainties
are
in
the
number
of
in­
use
smoking
vehicles
and
the
winter
PM
emission
rate
for
latemodel
gasoline
vehicles.
140
Acknowledgments
This
program
would
not
have
been
possible
without
the
assistance
of
the
following
individuals
to
whom
we
are
indebted.
Deborah
Kielian
from
the
City
and
County
of
Denver
assisted
in
recruiting
smoking
vehicles.
Phil
Gee,
Bill
Miron,
Sherry
Larson,
Zina
Washington,
and
Dawn
Mirabile
from
the
CDPHE
helping
recruit
vehicles
and
provided
other
assistance.
Peter
Groblicki
from
GM
calibrated
the
electrical
aerosol
analyzer
for
use
in
this
study.
Jerrold
Faircloth,
Versal
Mason,
and
Michael
Wheeler
helped
operate
the
EPA
emission
test
site.
Judith
C.
Chow,
Clifton
L.
Frazier,
Eric
M.
Fujita,
Terry
Hays,
Steven
D.
Kohl,
John
Sagebiel,
Ewa
Uberna,
John
G.
Watson
and
Barbara
Zielinska
at
the
Desert
Research
Institute
contributed
to
the
sample
analysis
and
validation.
We
also
thank
the
Coordinating
Research
Council,
the
NFRAQS
program,
the
Department
of
Energy,
and
Total
Petroleum
for
their
financial
support
of
this
project.

References
American
Automobile
Manufacturers
Association,
International
Fuel
Survey,
Summer
1996
and
Winter
1997.

Cadle,
S.
H.,
P.
A.
Mulawa,
J.
Ball,
C.
Donase,
A.
Weibel,
J.
Sagebiel,
K.
Knapp,
and
R.
Snow,
Particulate
Emission
Rates
from
In­
Use
High
Emitting
Vehicles
Recruited
in
Orange
County,
CA.
Environ.
Sci.
Technol.,
31,
3405­
3412,
1997.

Cadle,
S.
H.,
P.
J.
Groblicki,
and
P.
A.
Mulawa,
Problems
in
the
Sampling
and
Analysis
of
Carbon
Particulate.
Atmos.
Environ.
17,
593­
600,
1983.

Cass,
G.
R.,
Contribution
of
Vehicle
Emissions
to
Ambient
Carbonaceous
Particulate
Matter
­
A
Review
and
Synthesis
of
the
Available
Data
in
the
South
Coast
Air
Basin.
Final
Report
to
the
Coordinating
Research
Council,
1997.

Chan,
T.
L.,
and
D.
R.
Lawson,
Characteristics
of
Cascade
Impactors
in
Size
Determination
of
Diesel
Particles.
Atmos.
Environ.
15,
1273­
1279,
1981.

Chase,
R.
E.,
G.
J.
Duszkiewicz,
T.
E.
Jensen,
D.
Lewis,
E.
J.
Schlaps,
A.
T.
Weibel,
H.
J.
Wimette,
S.
H.
Cadle,
and
P.
A.
Mulawa,
Particle
Mass
Emission
Rates
from
Current­
Technology,
Light­
Duty
Gasoline
Vehicles.
Proceedings
of
the
AWMA
International
Specialty
Conference,
PM2.5:
A
Fine
Particle
Standard.
Long
Beach,
CA.
(
In
Press),
1998.

Chow,
J.
C.,
J.
G.
Watson,
L.
C.
Pritchett,
W.
R.
Pierson,
C.
A.
Frazier,
and
R.
G.
Purcell,
The
DRI
Thermal/
Optical
Reflectance
Carbon
Analysis
System:
Description,
Evaluation
and
Applications
in
U.
S.
Air
Quality
Studies.
Atmos.
Environ.
27A,
1185­
1201,
1993.
141
Chow,
J.
C.,
and
J.
G.
Watson,
Guidelines
for
PM­
10
Sampling
and
Analysis
Applicable
to
Receptor
Modeling.
U.
S.
Environmental
Protection
Agency,
Office
of
Air
Quality
Planning
and
Standards;
Washington
D.
C.,
EPA­
452/
R­
94­
009,
1994.

Chow,
J.
C.,
J.
G.
Watson,
E.
Fujita,
L.
W.
Richards,
W.
Neff,
D.
Dietrich,
and
S.
Hering,
Northern
Front
Range
Air
Quality
Study
Volume
A:
Ambient
Measurements.
Final
Report
prepared
for
Colorado
State
University,
1998.

Dockery,
D.
W.
and
C.
A.
Pope
III,
Acute
Respiratory
Effects
of
Particulate
Air
Pollution.
Ann.
Rev.
Public
Health.
15:
107­
132,
1994.

Durbin,
T.
D.,
M.
R.
Smith,
J.
M.
Norbeck,
and
T.
J.
Truex,
Population
Density,
Particulate
Emission
Characterization,
and
Impact
on
the
Particulate
Inventory
of
Smoking
Vehicles
in
the
South
Coast
Air
Quality
Management
District.
J.
Air
&
Waste
Manage.
Assoc.,
In
press.
1998.

Energy
and
Environmental
Analysis,
Inc.,
Size
Specific
Total
Particulate
Emission
Factors
for
mobile
Sources.
EPA
460/
3­
85­
005,
1985.

Groblicki,
P.
J.,
General
Motors
Sulfate
Dispersion
Experiment:
Aerosol
Sizing
Measurements.
General
Motors
Research
Publication
GMR­
2127,
1976.

Groblicki,
P.
J.
and
C.
R.
Begeman,
Particle
Size
Variation
in
Diesel
Car
Exhaust.
Society
of
Automotive
Engineers
Paper
790421,
1979.

Health
Effects
Institute,
Particulate
Air
Pollution
and
Daily
Mortality:
Replication
and
Validation
of
Selected
Studies.
Health
Effects
Institute
report,
1995.

Kaiser,
J.,
Showdown
Over
Clean
Air
Science.
Science,
277,
466­
469,
1997.

Lawson,
D.
R.,
Impaction
Surface
Coating
Intercomparison
and
Measurements
with
Cascade
Impactors.
Atmos.
Environ.,
14,
195­
199,
1980.

Lawson,
D.
R.,
S.
Diaz,
E.
M.
Fujita,
S.
L.
Wardenburg,
R.
E.
Keislar,
Z.
Lu,
and
D.
E.
Schorran,
Program
for
the
Use
of
Remote
Sensing
Devices
to
Detect
High­
Emitting
Vehicles.
Final
Report
to
the
South
Coast
Air
Quality
Management
District,
1996.

Maricq,
M.
M.,
D.
Podsiadlik,
and
R.
Chase,
Beyond
Mass
Emissions:
An
Examination
of
the
Size
Resolved
and
Transient
Nature
of
Particle
Emissions
by
Motor
Vehicles.
Proceedings
of
the
AWMA
International
Specialty
Conference,
"
PM2.5:
A
Fine
Particle
Standard."
January
1998.
In
Press.

Mulawa,
P.
A.,
S.
H.
Cadle,
K.
Knapp,
R.
Zweidinger,
R.
Snow,
R.
Lucas,
and
J.
Goldbach,
Effect
of
Ambient
Temperature
and
E­
10
Fuel
on
Particulate
Matter
Emissions
from
Light­
Duty
Vehicles.
Environ.
Sci.
Technol.,
31,
1302­
1307,
1997.
142
Mulawa,
P.
A.
and
J.
M.
Dasch,
Characterization
of
Exhaust
Particulate
Matter
from
1986
Through
1990
Model
Year
Light­
Duty
Gasoline
Vehicles.
General
Motors
Research
Publication
8456,
1995.

Rogge
W.
F.,
L.
M.
Hildemann,
M.
A.
Mazurek,
G.
R.
Cass,
and
B.
R.
T.
Simoneit,
Sources
of
Fine
Organic
Aerosol.
2.
Noncatalyst
and
Catalyst­
Equipped
Automobiles
and
Heavy­
Duty
Trucks.
Environ.
Sci.
Technol.,
27,
636­
651,
1993.

Sagebiel
J.,
B.
Zielinska,
P.
Walsh,
J.
Chow,
S.
H.
Cadle,
P.
A.
Mulawa,
K.
Knapp,
R.
Zweidinger,
and
R.
Snow,
PM­
10
Dynamometer
Exhaust
Samples
Collected
from
In­
Service
Vehicles
in
Nevada.
Environ.
Sci.
Technol.,
31,
75­
83,
1997.

Rykowski,
R.
A.,
T.
L.
Darlington,
and
J.
Heuss,
Exhaust
Particulate
Emissions
from
Gasoline­
Fueled
Vehicles.
Proceedings
of
the
World
Car
Conference
`
96,
Center
for
Environmental
Research
and
Technology,
University
of
California,
Riverside,
1996.

Rykowski,
R.
A.
and
B.
Hrebenyk,
Contribution
of
Gasoline­
Fueled
Vehicles
to
Ambient
Levels
of
Fine
Carbonaceous
Particulate
Matter.
AIR
document
No.
071697.
Final
report
to
the
Coordinating
Research
Council,
July
1997.

United
Stated
Environmental
Protection
Agency,
National
Air
Quality
and
Emissions
Trends
Report,
1993.
EPA­
452/
R­
94­
009,
1994.
