in-field verification of particulate emission factors of
TRANSCRIPT
In-field verification of particulate emission factors of road traffic D. Imhof1), U. Baltensperger1), E. Weingartner1), R. Gehrig2), M. Hill2)
1) Laboratory of Atmospheric Chemistry, Paul Scherrer Institut (PSI), CH-5232
Villigen-PSI, Switzerland 2) Eidgenössische Materialprüfungs- und Forschungsanstalt (EMPA), Überland-
strasse 129, CH-8600 Dübendorf, Switzerland ABSTRACT Extended field measurements of PM10 and PM1 concentrations at the kerbside of several streets with different traffic regimes were performed in Switzerland using a comprehensive set of instruments. Besides PM1 and PM10, concentrations of total number, surface area, and black carbon along with full size distributions were determined. From the differences between up- and downwind concentrations (or differences between kerbside and background concentrations for urban sites), “real-life” emission factors were derived using NOx concentrations to calculate dilution factors. INTRODUCTION The research project “Verification of PM10 emission factors of road traffic” was jointly realised by the Swiss Federal Laboratories for Materials Testing and Research (EMPA) and the Paul Scherrer Institute (PSI). The goal of this project was to characterise and quantify the road-traffic induced particle emissions for different traffic regimes and different processes (exhaust pipe emissions, emissions from abrasion and resuspension). Concentration measurements of the ambient air were performed on both sides of busy roads. During meteorological conditions with an approaching air flow perpendicular to the street, it is possible to determine the contribution of the local traffic from downwind-upwind differences. At the urban sampling sites, where this concept could not be realised, these contributions were calculated from the differences of the kerbside and nearby background sites. RESULTS Emission factors were determined by an indirect approach because the dilution rate was not known from the measurement configuration. Hourly dilution factors were calculated from the measured concentration differences of nitrogen oxides (NOx), the number of vehicles, and the NOx emission factors published for the Swiss vehicle fleet and for different traffic regimes. The emission factors for particles were then computed from the measured concentration differences, assuming that these undergo the same dilution as nitrogen oxides. Two vehicle categories were distinguished: LDV (light-duty vehicles < 6 m, i.e., gasoline and diesel passenger cars, vans, motor cycles) and HDV (heavy-duty vehicles > 6 m, i.e., lorries and coaches).
In order to distinguish between exhaust pipe emissions and emissions from abrasion and resuspension the PM10 and PM1 fractions were measured separately. PM1 was interpreted as direct exhaust pipe emissions, and PM10 as total particulate traffic emissions. The difference PM10-PM1 thus represents the emissions from abrasion and resuspension. Particle size spectra with high temporal and size resolution in the diameter range 18 nm < D < 10 µm were obtained, focussing on a detailed characterisation of fine particle emissions up to 1 µm. Particle number emission factors shown in this presentation were calculated for the particle size range from 18 to 50 nm (N0.05) representing the nuclei mode particles of exhaust gas emissions as well as for the total particle number (Ntot), i.e. number concentration of particles with D > 7 nm. The particle volume emission factor V0.3 (size range D = 18 to 300 nm, assuming that every particle has a spherical shape) is dominated by soot emissions, because condensation particles (D < 50 nm) are too small to significantly contribute to the volume concentration. Table 1: Particle emission factors calculated for two traffic regimes (Birrhard: Freeway, 4 lanes, speed limit 120 km/h; Zürich-Weststrasse: Urban main road, 2 lanes, speed limit 50 km/h, traffic lights).
N0.05 Ntot V0.3 PM10 PM1 Sampling site
#/km #/km cm3/km mg/km mg/km
LDV 4.10E+13 6.50E+14 0.02 63 16
All 1.40E+14 1.34E+15 0.03 83 33 Birrhard
HDV 8.10E+14 6.90E+15 0.16 267 193
LDV 2.50E+13 1.00E+14 0.03 49 11
All 6.70E+13 4.50E+14 0.08 104 29 Zürich-Weststrasse
HDV 7.40E+14 5.40E+15 0.94 703 342 CONCLUSIONS Direct mass measurements of PM1 and PM10 showed relatively high measuring uncertainties. The differences between downwind and upwind of these parameters were quite small, which resulted in emission factors with high standard deviations. Therefore the distinction between LDV and HDV turned out to be difficult. The measured particle size spectra have a lower measuring uncertainty. Due to the fact that the traffic contribution to particle number concentrations was much higher than the background value, particle number emission factors could be computed with smaller error bars. However, the conversion of a number into a mass concentration results in high uncertainties, because one has to assume the effective particle density, which depends on the mobility diameter and the chemical composition of the particles. The emission factors (N0.05, V0.3, PM1 and PM10) found for HDV were about 10 times higher than for LDV. Abrasion and resuspension processes represent a significant part of the total primary PM10 emissions of road traffic.
In-field verification of PM emission factors of road traffic
David Imhof, Urs Baltensperger, Ernest Weingartner (PSI) Robert Gehrig, Matz Hill (EMPA)
7th ETH Conference on Combustion Generated Nanoparticles
Motivation
Emission factors of motor vehicles are well-known from test stand measurements.
For real-life conditions, differences are expected because of maintenance and age of the vehicles aswell as other dilution factors and rates of theexhaust gas in ambient air.
Only little is known about emission factors of ultrafine particles.
In a field project real-life emission factors were determined for various traffic situations.
Concept and measured parameters
Wind direction
Sampling site 2Sampling site 1
BackgroundBackground +Traffic influence
Roadway
Instrument Parameter
SMPS (DMA + CPC, TSI 3010) Size distribution: 7 - 300 nm
Optical Particle Counter (Grimm 1.108) Size distribution: 0.3 - 20 µm
CPC (TSI 3022A) Number concentration N (D > 7 nm)
Betagauge (Eberline FH 62-I-R, Andersen) Mass concentration PM 10
Betagauge (Eberline FH 62-I-R, Andersen) Mass concentration PM 1
Aethalometer (AE10, Magee Scientific) Black Carbon
Diffusion Charger (LQ 1-DC) Particle surface area concentration
NOx analyzer (ML 8841) NO, NOx
CO analyzer (APMA 350E) CO
Sampling sites in the area of Zürich
Birrhard Zürich-Rosengartenstrasse
Zürich- Weststrasse Aathal
Humlikon
Sampling site Traffic situationAathal (only PM measurement)* Main road (50 km/h)
Birrhard Freeway (4 lanes, 120 km/h)
Humlikon Highway (2 lanes, 80 km/h)
Zürich-Rosen-gartenstrasse*
Urban main road (50 km/h, slope 8%)
Zürich-Weststrasse*
Urban main road (50 km/h, traffic lights)
*) Background station is represented by a nearby sampling site
Particle number and volume size distribution close to a motorway
0
10000
20000
30000
40000
50000
60000
1 10 100 1000Particle diameter D [nm]
Num
ber c
once
ntra
tion
dN/d
logD
p [c
m-3
]
DownwindUpwind
02000400060008000
10000120001400016000
1 10 100 1000Particle diameter D [nm]
∆dN/
dlog
Dp [c
m-3
]
Difference
02468
10121416
1 10 100 1000Particle diameter D [nm]
Volu
me
conc
entr
atio
ndV
/dlo
gDp
[µm
3 /cm
3 ]
DownwindUpwind
-1
0
1
2
3
4
5
1 10 100 1000
Particle diameter D [nm]
∆dV/
dlog
Dp [µ
m3 /c
m3 ]
Difference
Average diurnal variation (I)V0.3
-2
0
2
4
6
8
10
12
0 3 6 9 12 15 18 21 0
hour
V0.3
[µm
3 /cm
3 ]
DownwindUpwindDiffe re nc e
Traffic frequency
0
1000
2000
3000
4000
5000
0 3 6 9 12 15 18 21 0hour
vehi
cles
/h
LDVHDVTotal
N0.05
0
5000
10000
15000
20000
25000
0 3 6 9 12 15 18 21 0hour
N0.0
5 [p
artic
les/
cm3 ] Downwind
UpwindDifference
PM10
0
5
10
15
20
25
0 3 6 9 12 15 18 21 0hour
PM10
[µg/
m3 ]
DownwindUpwindDiffe re nc e
PM1
-4
-2
0
2
4
6
8
10
12
0 3 6 9 12 15 18 21 0
hour
PM1
[µg/
m3 ]
DownwindUpwindDiffe re nc e
NOx
0
10
20
30
40
50
60
70
80
0 3 6 9 12 15 18 21 0hour
NOx [
ppb]
DownwindUpwindDifference
Average diurnal variation (II)
∆dN/dlogD [cm-3]
6 * 104
104
103
102
101
SMPS: Difference downwind - upwind
Hour
Parti
cle
diam
eter
D [n
m]
00 03 06 09 12 15 18 21 2418
100
300ELPI (Downwind)
Number concentrationdN/dlogD [cm-3]
104
103
102
101
100
10-1
Hour
Parti
cle
diam
eter
D [µ
m]
00 03 06 09 12 15 18 21 240.03
0.1
1
8
Calculation of emission factors∆NOx = ————— · nLDV + ————— · nHDV
EFNOx(LDV) EFNOx(HDV)
D D
D = —————————————————EFNOx(LDV) · nLDV + EFNOx(HDV) · nHDV
∆NOx
∆Ci = ————— · nLDV + ————— · nHDV
EF(LDV) EF(HDV)
D D
∆NOx: Concentration difference of NOx [µg/m3]
EF(LDV): Emission factor of light-duty vehicles [mg/km]
EF(HDV): Emission factor of heavy-duty vehicles [mg/km]
n: Vehicle number [1/h]
D: Dilution [m2/h]
∆Ci: Concentration difference of species I [x/m3]
Concentration difference vs traffic frequency N 0.05
-20000
0
20000
40000
60000
0 500 1000 1500 2000
vehicles/h
∆N 0
.05
[#/m
3 ]
V 0.3
-10
0
10
20
30
40
0 500 1000 1500 2000
vehicles/h
∆V 0
.3 [µ
m3 /m
3 ]
PM1
-40
-20
0
20
40
60
0 500 1000 1500 2000
vehicles/h
∆PM
1 [µ
g/m
3 ]
PM10
-50
0
50
100
150
0 500 1000 1500 2000
vehicles/h
∆PM
10 [µ
g/m
3 ]
Emission factor per vehicle vs HDV fraction V 0.3
-0.2
0
0.2
0.4
0.6
0.8
1
0 0.1 0.2 0.3
Fraction HDV
EF V
0.3
[cm
3 /km
]
PM10
-500
0
500
1000
1500
2000
0 0.1 0.2 0.3
Fraction HDV
EF P
M10
[mg/
km]
PM1
-400
-200
0
200
400
600
800
1000
0 0.1 0.2 0.3
Fraction HDV
EF P
M1
[mg/
km]
N 0.05
-1E+14
0
1E+14
2E+14
3E+14
4E+14
5E+14
0 0.1 0.2 0.3
Fraction HDV
EF N
0.05
[#/k
m]
Emission factorsBIRRHARD
ZUERICH-WESTSTRASSE
N0.05
0.E+00
2.E+14
4.E+14
6.E+14
8.E+14
1.E+15
LDV All HDV
EF N
0.05
[#/k
m]
Ntot
0.E+00
2.E+15
4.E+15
6.E+15
8.E+15
LDV All HDV
EF N
tot [
#/km
]V0.3
0.00
0.20
0.40
0.60
0.80
1.00
LDV All HDV
EF V
0.3
[cm
3 /km
]
PM10
0
200
400
600
800
LDV All HDV
EF P
M10
[mg/
km]
PM1
0
100
200
300
400
LDV All HDV
EF P
M1
[mg/
km]
N0.05
0.E+00
2.E+14
4.E+14
6.E+14
8.E+14
1.E+15
LDV All HDV
EF N
0.05
[#/k
m]
Ntot
0.E+00
2.E+15
4.E+15
6.E+15
8.E+15
LDV All HDV
EF N
tot [
#/km
]
V0.3
0.00
0.20
0.40
0.60
0.80
1.00
LDV All HDV
EF V
0.3
[cm
3 /km
]
PM10
0
200
400
600
800
LDV All HDV
EF P
M10
[mg/
km]
PM1
0
100
200
300
400
LDV All HDV
EF P
M1
[mg/
km]
Vertical profiles of NOx, particle surface area concentration and particle number size distribution
05
101520253035404550
0 50 100
DC signal [µm2 cm-3]
Upwind
05
101520253035404550
0 100 200 300DC signal [µm2 cm-3]
Downwind
05
101520253035404550
0 5 10 15 20
NOx [ppb]
Heig
ht [m
]
Upwind
05
101520253035404550
0 50 100 150NOx [ppb]
Hei
ght [
m]
Downwind
NOx Surface area
12000 - 30000 5000 - 12000 2000 - 5000 800 - 2000 200 - 800 20 - 200 7 - 20 2 - 7 1 - 2 0.2 - 1
Particle number concentrationdN/d(logD) [cm-3]
Hei
ght [
m]
0.03 0.1 1 80
10
20
30
40
50Upwind
Particle diameter D [µm]
Hei
ght [
m]
0.03 0.1 1 80
10
20
30
40
50 Downwind
Correlation particles - NOxDownwind
y = 1.9782x - 0.6255R2 = 0.8933
0
5
10
15
20
25
30
35
40
0 5 10 15 20
NOx(h=5-30m)-NOx(background) [ppb]
DC s
igna
l(h=5
-30m
) -
DC s
igna
lback
grou
nd [µ
m2 c
m-3
]
0
5
10
15
20
0 1 2 3 4
DCTraffic/NOxTraffic
Hei
ght [
m]
ConclusionsPM10 and PM1 are determined by direct mass measurements which show relatively high measuring uncertainties.
SMPS data have a low measuring uncertainty, but the conversion into a mass concentration will result in high uncertainties.
ρeff = 15.175 Dmob-0.666
R2 = 0.9494
0
0.2
0.4
0.6
0.8
1
1.2
0 100 200 300 400 500 600 700 800
Mobility diameter [nm]
Effe
ctiv
e de
nsity
[g c
m-3]
Emission factors (N0.05, V0.3, PM1, PM10) found for HDV are about 10 times higher than for LDV. (Park et al., 2003)
Our results correspond well to findings in other field studies with similar traffic situations.