image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
image
An analysis of vehicular particle number emissions based on long-term
roadside and urban background measurements
S. Klose
1
, W. Birmili
1
, B. Wehner
1
, A. Wiedensohler
1
, T. Tuch
1,2
, J. Voigtländer
1
, U. Franck
2
and M. Ketzel
3
1
IfT – Leibniz Institute for Tropospheric Research, Permoserstraße 15, 04318 Leipzig, Germany
2
UFZ – Helmholtz-Centre for Environmental Research, Permoserstraße 15, 04318 Leipzig, Germany
3
Department of Atmospheric Environment, National Environmental Research Institute, Frederiksborgvej 399, 4000 Roskilde, Denmark
Email: klose@tropos.de
Motivation
Objectives and methods
Th
l
d bi
i l d t
t (2005/2006) i l d
National Environmental
Research Institute
The analysed biennial dataset (2005/2006) includes
aerosol
particle number size distributions in a highly-trafficked street canyon
(Leipzig-Eisenbahnstraße) and in the urban background (Leipzig-IfT)
traffic-induced ultrafine (< 0.1 μm) and fine particles (< 1 μm) are known to be a
hazard to human health: because of their size they can be deeply deposited
high aerosol particle concentrations in densely built and much frequented areas
(Leipzig Eisenbahnstraße) and in the urban background (Leipzig IfT)
traffic
census using an automated video detection system
measurement
of local wind speed and direction at the roadside measurement site
high aerosol particle concentrations in densely built and much frequented areas
because of a lack of ambient ventilation and fresh air supply
to
study the vehicle-induced contribution to urban air pollution: calculation of
particles size-resolved emission factors indicating the number of particles emitted by
a vehicle per driven distance
consideration of turbulence-depending particle transport
meteorology within a street canyon is especially characterised by the creation of
analysis as a function of diurnal and meteorological parameters, seasonal effects as
well as changes in fleet composition
emission factors are furthermore required as input for urban air quality modelling
extensive vorticies
dispersion from the tailpipe to the sampling point is quantified with the dispersion
model OSPM (Operational Street Pollution Model)
Measurements
Dispersion modelling with OSPM
emission factors are furthermore required as input for urban air quality modelling
model OSPM (Operational Street Pollution Model)
regular
street canyon with
roadside site Leipzig-Eisenbahnstraße
urban background site Leipzig-IfT
twin
differential mobility particle
two lanes
10,500
vehicles/working day
twin differential mobility
sizer (3-800 nm) at a height of 16 m
1.5 km distant from Leipzig-
Eisenbahnstraße
twin
differential mobility
particle sizer (3-800 nm) at a
height of 6 m above ground
Eisenbahnstraße
10000
15000
Leipzig-Eisenbahnstraße
Leipzig-IfT
4-12 nm
gg
ultrasonic
anemometer on
the roof (24 m)
Row of houses with particle
inlet and anemometer.
0
5000
N [cm
-3
]
automated
video detection
system: counting and
classifying of vehicles
0
10000
15000
12-40 nm
N [cm
-3
]
classifying of vehicles
(Voigtländer et al., 2006)
G
d l
f th
it
Schematic illustration of the basic model principles in OSPM Concentrations are calculated as a sum of the direct plume
40000
0
5000
40-120 nm
OSPM is used to determine the dilution factor
F
(Berkowicz et al., 1997)
the model is particularly suitable for regular street canyons
Ground plan of the site
and its sourrounding area.
Schematic illustration of the basic model principles in OSPM. Concentrations are calculated as a sum of the direct plume
contribution and the recirculating pollution (Berkowicz et al., 1997).
10000
cm
-3
]
N [cm
-3
]
3000
6000
the model is particularly suitable for regular street canyons
the dilution factor
F
depends on the wind-induced turbulence and the turbulence
vehicles induce themselves, the following information is required:
100
1000
dN/dlog D
p
[c
0
traffic at Eisenbahnstraße
600
800
passenger car-like
lorry-like
M [h
-1
]
vehicles induce themselves, the following information is required:
roof level wind direction and wind speed
topography of the street canyon:
traffic speed and density
3
10
100
800
10
100
Leipzig-Eisenbahnstraße
Leipzig-IfT
0
200
400
Md
Td
Wdd Thd
Fid
Std
Sd
M
Calculation of emission factors
Results & Discussion
Average particle number size distribution, 2005/2006.
orientation, height, width, distance to next crossing streets, height of particle inlet
Weekly cycle of particle number concentration and traffic density.
3
10
100
800
D
p
[nm]
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday
To verify the results of the simulation,
1E15
Calculation of emission factors
Results & Discussion
3 10
4
4x10
4
measured
simulated
cm
-3
]
traffic
the run of the time series of the
simulated particle concentration can be
d ithth
dt ffi
1E14
1E15
m
-1
]
6
Mo
We
Fr
Su
Tu
Th
Sa
0
1x10
4
2x10
4
3x10
4
N
traffic
[c
wind
topography
traffic
estimated
emission factor
elling
compared with the measured traffic-
related increment concentration.
1E13
logD
p
[veh
-1
km
wind speed [m/s]
0
2
4
6
dilution factor
estimated
li
it it
OSPM
ward mode
If the dilution is described adequately
with respect to changing wind directions
1E12
average size distribution
fitted nucleation mode
fitted soot mode
dE/d
90
180
270
360
wind direction [°]
line source intensity
forw
with respect to changing wind directions
and wind speeds, the dilution factor
F
(s m
-2
) can be used to determine the
3
10
100
800
1E11
Measured vs. simulated particle concentration for changing inflow conditions.
Average size distribution of the emission factor of an average vehicle.
D
p
[nm]
268
270
272
274
276
278
280
0
Julian day of the year 2006
simulated
concentration
measured
concentration
verification
emission factor
E
(veh
-1
km
-1
)
:
8x10
14
1x10
15
N
dilution factor
modelling
1E16
1E17
6x10
14
E [veh
-1
km
-1
]
FM
N
E
traffic
=⋅
d
real
line source intensity
inverse m
1E14
1E15
D
p
[veh
-1
km
-1
]
2x10
14
4x10
14
E
emission factor in the range of 4-30 nm
linear regression
with
N
traffic
being the particle concentration of the traffic-related
increment (
N
roadside
–N
background
( cm
-3
)) and
M
being the traffic density
traffic
real
emission factor
1E12
1E13
dE/dlog
lorry-like
passenger car-like
-10
0
10
20
30
0
ambient temperature [°C]
linear regression
Emission factor of ultrafine particles vs. ambient temperature.
Size distributions of emission factors for different vehicle classes.
(veh s
-1
).
Rf
emission factor
410 100 800
1E11
D
p
[nm]
two
aerosol population modes in the size-resolved emission spectrum: the
nucleation mode (diameter 14 nm) and the soot mode (diameter 98 nm)
References
Berkowicz, R., Hertel, O., Larsen, S., Sørensen, N., und Nielsen, M. 1997. Modelling traffic pollution in streets. Ministry of
Environment and Energy, National Environmental
Research
Institute.
Imhof, D., Weingartner, E., Prevot, A., Ordonez, C., Kurtenbach, R., Wiesen, P., Rodler, J., Sturm, P., McCrae, I., Ekström, M., and
Baltensperger, U. (2006). Aerosol and NOx emission factors and submicron particle number size distributions in two road
tunnels with different traffic regimes. Atmos. Chem. Phys., 6:2215–2230.
Ketzel M Wåhlin P Berkowicz R and Palmgren F 2003 Particle and traces gas emission factors under urban driving
dependency
of the nucleation mode (particles between 4 and 30 nm) on the
ambient temperature resulting in a annual cycle, e.g.
1
14
1
km
E(summer)
=5.9⋅10
veh
1
14
1
km
E(winter)
=8.9 ⋅10
veh
Ketzel, M., Wåhlin, P., Berkowicz, R., and Palmgren, F. 2003. Particle and traces gas emission factors under urban driving
conditions in Copenhagen based on street and roof-level observations. Atmospheric Environment, 37:2735–2749.
Voigtländer, J., Tuch, T., Birmili, W., and Wiedensohler, A. 2006. Correlation between traffic density and particle size distribution in
a street canyon and the dependence on wind direction. Atmos. Chem. Phys., 6, 4275-4286.
an average lorry-like vehicle emits approximately 80-times as many particles as an
passenger car like vehicle does:
16
1
k
1
E(l
lik )
410
h
1
14
1
km
E(passenger car-like)
=
4.8
⋅10
veh
fairly
agreement with other comparable studies
Acknowledgements
This work was supported by UBA contract UFOPLAN No 20442204/03 and the EU Marie Curie
passenger car-like vehicle does:
1
16
1
km
E(lorry -like)
=
4
⋅10
veh
LibiI i
fT
hiR
hPhiD
A
lG
yg
p
1
14
1
km
2.8
10 veh
−1
14
1
km
1.5
10 veh
(Ketzel et al., 2003)
(Imhof et al., 2006)
This work was supported by UBA contract UFOPLAN No. 20442204/03 and the EU Marie Curie
Reintegration Grant FP6-2002-Mobility-11 contract No. 510583.
Leibniz Institute for Tropospheric Research – Physics Department, Aerosol Group