Chang’an University
Traffic emission factors of particle number measured
in a street canyon in Stockholm, Sweden
Deng Shunxi a,c Christer Johansson a,b
a Stockholm University, Sweden；
；
b Environment and Health Administration, Stockholm, Sweden；
；
c Chang’an University, Xi’an, China
Chang’an University
Chang’an University
INTRODUCTION
Road traffic is one of the major sources for particulate matter in
urban areas. Some studies have shown that traffic exhaust
particles have the adverse health effects due to their higher
number and surface area to larger particles, and to their increased
toxicity on a per mass basis (Nel, 2005).
Emission
factors (EFs), which are defined as the mass or number
of an air pollutant emitted per vehicle and distance traveled or
mass of fuel used, are widely used to characterize road traffic
particle emissions.
Chang’an University
INTRODUCTION
In
this work, a data analysis of the measurements of particle size
number and in a congested urban street canyon, Stockholm, is
conducted.
One
year measurements of particle number and traffic and
meteorological data are used to estimate average particle number
EFs by using an inverse modeling method.
The
size resolved particle number EFs for passenger cars (PC),
medium-duty (MD) diesel trucks and heavy-duty (HD) buses are
estimated by using multilinear regression.
The
results are compared to those of previously published values
determined by similar method and conditions in recent years.
Chang’an University
EXPERIMENTAL
The
first monitoring site : Hornsgatan street canyon, located in
central Stockholm, Sweden, is a four lane street 24m wide,
surrounded by 24 m high buildings on both sides.
The
second monitoring site : Rosenlundsgatan, corresponding to
the urban background site, is located on the roof (25 m above the
street level) of a building ca. 600 m southeast from Hornsgatan
street,.
The measurements were simultaneously performed at two
monitoring sites..
Chang’an University
EXPERIMENTAL
Chang’an University
EXPERIMENTAL
The
particle number concentration, at both street and background
roof sites, was measured using identical Condensation Particle
Counters (CPC TSI 3022) that has a lower cut-off size of 7 nm
(PN7).
The
particle number size distribution measurements were
conducted with differential mobility analyzer (DMA) coupled
with a condensation particle counter (CPC TSI 3010).
Distribution
consisted of 16 size bins from 20 to 400 nm. Another
size channel (Dp～10nm) was added by subtracting the integrated
number concentration obtained with the DMA, from the total
number concentration measured with the CPC 3022.
Chang’an University
EXPERIMENTAL
Meteorological
parameters including wind speed, wind direction,
temperature and relative humidity are also observed at roof level.
For
traffic, hourly mean values of categorized traffic volumes and
speed were obtained from automatic counters at the street canyon
site.
According
to the characteristic of urban vehicles in Stockholm,
PC :vehicle with a wheel pair distance smaller than 5.0 m.
MD : vehicles with 5-10 m wheel pair distance, mainly being
diesel powered trucks or vans
HD : larger 10 m distance vehicle, almost being ethanol powered
public buses
.
Chang’an University
RESULTS AND DISCUSSION
weekdays
,s
ra
c
re
g
n
e
ss
a
p
,s
ra
c
re
g
n
e
ss
a
p
,s
ra
c
re
g
n
e
ss
a
p
1
-
1
-
3000
) 2500
h
2000
h
e
v( 1500
la
to 1000
t 500
0
3000
) 2500
h
2000
h
e
v( 1500
la
to 1000
t 500
0
3000
) 2500
1
h 2000
h
e
v( 1500
la
to 1000
t 500
0
passenger car s
medi um t r ucks
t ot al
buses
0 2 4 6 8 10 12 14 16 18 20 22
l ocal t i me ( h)
Sat ur days
passenger car s
medi um t r ucks
t ot al
buses
0 2 4 6 8 10 12 14 16 18 20 22
l ocal t i me ( h)
Sundays
passenger car s
medi um t r ucks
t ot al
buses
0 2 4 6 8 10 12 14 16 18 20 22
l ocal t i me ( h)
Traffic flow rate
400
350
300
250
200
150
100
50
0
,s
kc
ur
t
m
ui
de
m
se
su
b
1
-
400
350
300
250
200
150
100
50
0
,s
kc
ur
t
m
ui
de
m
se
su
b
1
-
400
350
300
250
200
150
100
50
0
,s
kc
ur
t
m
ui
de
m
se
su
b
)
h
he
v(
)
h
he
v(
1
-
)
h
he
v(
Hourly
average of traffic
flow rate measured during
the measurement campaign.
During working days, the
average traffic volume per
day was 32202 vehicles, and
the fraction of PC, MD, and
HD buses were 87.7%, 9.0
% and 3.3 %, respectively..
Chang’an University
RESULTS AND DISCUSSION
Worki ng days
Weekends
120000
120000
)
3
m100000
c/
#
(
r 80000
e
b
m60000
u
n
e
l 40000
c
i
t
r 20000
a
P
St r eet
Roof
St r eet - Roof
3)m
c/100000
#(
re 80000
bm
un 60000
el
ci 40000
tr
aP20000
St r eet
Roof
St r eet - Roof
0
0
0
2
4
6
8
0
10 12 14 16 18 20 22
2
4
6
8 10 12 14 16 18 20 22
Local t i me ( hour )
Local t i me ( hour )
Weekends
Worki ng days
500
500
St r eet
Roof
St r eet - Roof
400
3)
m300
/
g
u
(
x200
O
N
St r eet
Roof
St r eet - Roof
400
)
3
m300
/
gu
(
x200
O
N
100
100
0
0
0
2
4
6
8
10 12
14
16 18
20
0
22
2
4
6
8
10
12 14
16
18
20 22
Local t i me ( hour )
Local t i me ( hour )
Weekends
Worki ng days
2. 5
2. 5
St r eet
Roof
St r eet - Roof
2. 0
)
3
m1. 5
/g
m
(
O1. 0
C
St r eet
Roof
St r eet - Roof
2. 0
)
3m
/1. 5
g
m
(
O1. 0
C
0. 5
0. 5
0. 0
0. 0
0. 51
0
2
4
6
8
10
12 14
Local t i me ( hour )
16
18
0. 27
0. 20
0. 46
1. 04
1. 16
20 22
Local t i me ( hour )
0. 82
0. 71
Diurnal patterns of the
concentrations for NOx, CO
and PN7
The
variation characteristic
of the traffic source is clearly
visible both on working days
for the street concentrations
and the difference street-roof
with the morning rush hour
peak and the broader
afternoon rush hour..
Chang’an University
RESULTS AND DISCUSSION
p
D
g
o
ld
/N
d
p
D
g
o
ld
/N
d
p
D
g
o
ld
/N
d
80000
)3 70000
m
c/ 60000
se 50000
lc 40000
it
ra 30000
p
( 20000
10000
0
5000
)3 4500
4000
m
/cs 3500
3000
e
lc 2500
it 2000
ra 1500
p
( 1000
500
0
St r eet st at i on
Average particle size
distribution
wor kdays
Sat ur days
Sundays
The
10
100
Di amet er ( nm)
1000
Roof st at i on
wor kdays
Sat ur days
Sundays
maximum particle
number concentrations are
found in the sizes around 20
nm.
At
10
80000
70000
)3 60000
m
c/ 50000
se 40000
lc
it 30000
ra 20000
p
( 10000
0
- 10000 10
100
Di amet er ( nm)
1000
St r eet - r oof
wor kdays
Sat ur days
Sundays
100
Di amet er ( nm)
1000
roof level, we found a
rather wide maximum shifted
to larger particle sizes in the
range 20-100nm.
Chang’an University
Averaged diurnal evolution of particle size distributions at Hornsgatan and the
roof background and the difference street- roof
Chang’an University
Estimation of PN emission factors
According
to OSPM (the Operational Street Pollution Model)
( Berkowicz et al. 1997), the average emission strength in a street
canyon can be calculated as
Qh = qh Nh =
ch,street − ch,background
Fh (meteorology,traffic)
=
ch,diff
Fh
is the emission density (pollutant m-1 s-1); qh is the EF
(pollutant vehicle-1 m-1); Nh is total traffic volume (vehicle s-1); ch,
street and ch,background are the concentrations measured in the street
canyon and in the urban background (pollutant m-3 ),
Qh
F is dilution, factors, calculated by OSPM, (s/m2), using the
meteorological parameters and traffic flow data
.
Chang’an University
Estimation of PN emission factors
For a specific hour, h, it is possible to estimate EFs for each
categorized vehicle. The total emission density can be expressed
as
Qh = ∑ N k , h × qk , h
k
.
Applying multiple regression analysis methods, the EFs for any
type k of vehicles can be determined as a solution of this linear
equation system.
Chang’an University
Estimation of PN emission factors
Emission factors for Particle number are obtained by using above equations
Particle number
Coefficients
Std. Error
t Stat
R2
1.68E+14
6.18E+12
27.13
0.59
3.29E+14
6.62E+12
49.64
0.83
9.19E+13
2.22E+12
41.33
0.77
3.85E+13
1.53E+12
25.13
0.55
1.16E+14
6.45E+13
1.79
MD
2.78E+15
5.35E+14
5.19
HD(Buses)
9.76E+15
1.76E+15
5.56
Total fleet
6.27E+14
1.20E+13
52.06
Vehicle type
Particles
veh-1 km-1
PN7-20nm
PN20-40nm
PN40-100nm
Total fleet
PN100-400nm
PC
PN>7nm
Units
0.86
0.84
Chang’an University
RESULTS AND DISCUSSION
number EF for the fleet mix : (6.27±0.12)×1014 particles veh-1
km-1,
The
The
number EF for three different vehicle categories :
Passenger cars (PC):
(1.16±0.65)×1014
particles veh-1 km-1
Medium-duty (MD) : (2.78±0.54)×1015
Heavy-duty (HD) buses : (9.76±1.76)×1015
The
ratios of the EF for PC, MD and HD to that for total vehicle
fleet are 0.2, 4.4 and 15.6 respectively, which indicates both diesel
vehicle (trucks) and ethanol buses are important sources of PN
emission in urban area of Sweden.
Chang’an University
Comparison of number EF measured in street canyon/ urban rood or
urban road tunnel
The
results of this study are comparable to the EFs measured on
the Stockholm tunnel by Kristession et al. (2004), Ketzel et al.
(2003), Jones and Harrison (2006) and Imhof et al. (2005). Figure
shows the comparisons.
km )
1. 2E+16
1. 0E+16
1
-
8. 0E+15
number EF( veh
1
-
Fl eet mi x
PC
MD
HD
6. 0E+15
4. 0E+15
2. 0E+15
0. 0E+00
Kr i st ensson
et al .
Jones &
Har r i son
Ket zel et al .
I mhof et al .
Thi s wor k
Chang’an University
RESULTS AND DISCUSSION
There may be many reasons that result in the differences, but at
least they can be explained by the different instruments with
different lower cut diameters that were used.
The relatively high particle number EF values measured in this
work may in part be a feature of the measurement instrument, but
may also be influenced by the fact that a significant proportion of
the MD vehicle fleet at Hornsgatan (9%), being diesel fueled
vehicles.
Almost all the public local buses are ethanol fueled vehicles with
much higher particle emission. The vehicle fleet emission value
for particle number is particularly sensitive to the proportion of
public buses.
So far the emission of ethanol buses is not well known..
Chang’an University
Thank you!