ambient elemental signatures of diesel and automotive particulate matter by size, time, and...
TRANSCRIPT
Ambient Elemental Signatures of Diesel and Automotive Particulate Matter by Size, Time, and Composition.
1Thomas A. Cahill, 1Earl Withycombe, Steven S. Cliff, Michael Jimenez-Cruz, Lee Portnoff, and 2Kevin D. Perry, DELTA Group, University of California, Davis, http://delta.ucdavis.edu, 1The Health Effects Task Force, American Lung Association, Sacramento-Emigrant Trails, 2Dept. of Meteorology, University of Utah
Sources of information – joint studies with the DELTA Group
Laboratory studies – NREL/U. Minnesota/DRI diesels; UCD S-XRF analysis –(Lawson, Watts, Zielenska et al.), plus DRI Lube oil (Fujita)
Prior field studies – prior ARB/UC Davis work; HEI/DRI Tuscarora Tunnel (Gertler et al 2002)
Quasi-ambient and ambient applications Interstate 5 on downtown Sacramento and Watt Ave
School – (Lung Assoc – Sacramento Emigrant Trails) Fresno FACES studies (ARB) Highway 50 in South Lake Tahoe – TRPA (Tahoe
Regional Planning Agency)
U. Minn. Dynamometer Diesel tests
1810
5.623.2
1.81
0.560.32
0.180.1
0.0560.032
0.0180.01
Coarse MOUDI Stage Diameter (micrometers) Ultra- Fine
-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
Mic
rog
ram
s/c
m2
Mass/200Sulfur
LeadZinc x 10
Calcium Phosphorus
For micrograms/m3, times 8.7DELTA Group, S-XRF, UC Davis
Diesel Particles by MOUDI Impactor and S-XRFSample Run # 4, CA Fuel; no grease
U. Minnesota Dynamometer Diesel Tests
1810
5.623.2
1.81
0.560.32
0.180.1
0.0560.032
0.0180.01
Coarse MOUDI Stage Diameter (micrometers) Ultra-Fine
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
Mic
rog
ram
s/c
m2
Mass/200 Sulfur Zinc x 10 Phosphorus
For micrograms/m3, times 8.7DELTA Group, S-XRF, UC Davis
Diesel Particles by MOUDI Impactor and S-XRFSample Run # 11, CA Fuel; no grease
Time-of-Flight Organic Speciation vs. Particle Size (Prof. Kelly, UC Davis)
DOE Gasoline/Diesel PM Split StudyParticle-Phase PAH in Lubrication Oil
Lube Oil - Diesel
0
50
100
150
200
250
300
350
400
ccs_
10
ccs_
ib
hcs
_11
hcs
_11
n
hcs
_1
3.2
hcs
_9
e
hcs
_ia
hcs
_ii
id_
i
id_
iii
ug
/g
Lube Oil - Spark Ignition
0
50
100
150
200
250
300
350
400
si_
1
si_
3
si_
5c1
si_
6c1
si_
6c3
si_
7c2
si_
8c1
si_
8c3
si_
9c2
si_
9c4
si_
10c2
si_
10c4
corone
dbanth
bghipe
incdpy
bapyrn
peryle
bepyrn
m_7bpy
bbjkfl
chry56m
baa7_12
bzantone
chrysn
m_7baa
baanth
bzcphen
Summary of laboratory tests
Most diesel mass lies below 0.25 µm (very fine/ultra fine) Sulfur correlates with emitted very fine/ultra fine mass, and is
associated with the diesel fuel Diesel tracers, some ultra-fine, comes from combustion of
lubricating oil stabilized with zinc thiophosphate and CaCO3 ; the amounts are highly variable (factor of 3)
Ultra fine diesel particles have much more massive and less volatile organic components than coarser diesel components, as required by the Kelvin effect.
Lubricating oil for spark ignition vehicles has more and heavier PAHs than lubricating oil for diesels
Diesel particles are highly light absorbing at all wavelengths, 350 nm – 820 nm
1810
5.623.2
1.81
0.560.32
0.180.1
0.0560.032
0.0180.01
Coarse MOUDI Stage Diameter (micrometers) Ultra- Fine
-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
Mic
rog
ram
s/c
m2
Mass/200Sulfur
LeadZinc x 10
Calcium Phosphorus
For micrograms/m3, times 8.7DELTA Group, S-XRF, UC Davis
Diesel Particles by MOUDI Impactor and S-XRFSample Run # 4, CA Fuel; no grease
PM 2.5
PM 0.25 ?
PM 10
Average Zn to mass, all DRI tests, 1800 1300
Prior field studies (Gertler et al 2002)
0 20 40 60 80 100
Percent Heavy Duty Vehicles ( 7 - 8 axles)
0
50
100
150
200
Em
issio
n F
acto
r (m
g/k
m)
PM10 Emission rates from Tuscarora Tunnel StudyGertler et al, 2002
y = 168x + 12; R2 = 0.86
19701972
19741976
19781980
19821984
19861988
19901992
19941996
19982000
20022004
Year
0
200
400
600
800
1000
1200
Em
issi
on fa
ctor
(mg/
km)
Heavy Duty ( 7 - 8 axle) diesels Light Duty vehicles x 10
PM2.5 Aerosol Emission Factors, Heavy Duty and Light Duty Vehicles Gertler et al, Health Effects Institute (2002)
Note: CA RFG vehicles 0.4 to 2 mg/km
Table 1 Comparison to heavy duty and light duty PM10 and PM2.5 emission rates form the Gertler at al 2002 Tuscarora Tunnel studies and other studies. Parameter Heavy duty
(mg/km) Light duty (mg/km)
Mixed (mg/km)
PM10 mass Gertler 2002 Tuscarora 181 + 13 10 + 11 87 + 54 PM2.5 mass Gertler 2002 Tuscarora 135 + 18 14 + 13 62 + 42 PM10 mass Gillies 2001 Sepulveda na Na 69 + 30 PM2.5 mass Gillies 2001 Sepulveda na Na 53 + 27 PM2.5 mass Norbeck 1998 In-use (med) 18 + 9 PM2.5 mass Norbeck 1998 In-use (high) 185 + 50 PM10 mass Sagebiel 1997 High CO, HC 346 smoke PM10 mass Sagebiel 1997 High CO, HC 32 no smoke From these results, we see that diesel is about 18 times worse than light duty vehicles for PM10 emissions and 10 times worse than light duty vehicles for PM2.5 emissions, and that the worst case smoking car is about the same as the average diesel. Incidentally, these emission values are sharply lower than occurred only a decade ago.
Parameter Diesel Trucks(mg/km)
Cars(mg/km)
Ratio Comment
PM2.5 mass 135 + 18 14 + 13 10 Mass? Truck PM10 = 181
PM2.5 OC 112 + 43 2.8 + 1.1 40 2nd biggest ratio
PM2.5 EC 185 + 66 3.3 + 1.2 55 biggest ratio
PM 0.25 Zn na na 10 < 0.25 μm
PM 0.25 Cu na na 10 < 0.25 μm
Heavy organics ??? A few out of 92
PM2.5 NH3, S < 0.8 1999; Some S in gasoline
Gasses (g/km) (g/km)
CO2 748 + 73 156 + 15 4.8 Roughly fuel mileage
CO < 0.6 1.9 + 0.7 < 0.3 NO/CO > 90 (!)
NO (as NO2) 11.9 + 1.9 0.4 + 0.07 28 3rd biggest ratio
THC 1.5 + 0.8 0.4 + 0.2 3.7
Summary for prior field tests
Diesel truck have roughly 10 (PM2.5) to 18 (PM10) times the mass emission rate of light duty vehicles
Diesels and light duty vehicles emit species seen in the laboratory tests, plus others in the PM2.5 mode.
There are a number of possible tracers that separate diesels from gasoline powered vehicles, including a few heavy PAHs
Gross emitting light duty vehicles, a few percent of all vehicles, can be as bad or worse per vehicle than the average heavy duty diesel.
Current ambient/quasi-ambient studies
Interstate 5 in downtown Sacramento urban
Watt Avenue on Arden Middle School suburban
Highway 50 in South Lake Tahoe The Fresno Super-site
suburban
Lung Assoc. Sacramento Transect Study Site Map and PM2.5
aggregated data #1, #3 - light blue = rain, yellow = “clear”, rest = fogs, wet and dry
0
5
10
15
20
25
30
35
Site
ug
/m3
Rain (Dec 17-22)
Clear (Dec 24-29)
Rain (Dec 29-31)
Clear (J an 1-6)
Fog (J an 7-10)
Fog/Rain (J an 10-12)
Fog/Rain (J an 12-15)
I-80
I-5
I-80
Hwy-50
Hwy-99
Sacramento – a highway nexus, (I-5, I-80, Hwy 50, Hwy 99) and close to violations of PM2.5 standards
121
12
23
34
45
56
66
77
88
99
1010
1111
1212
12
Month of the Year by week
0
10
20
30
40
50
60
70
80
90
Mic
rogr
ams/
m3
Fine PM2.5 Aerosols at 13th and T Street, Sacramento2002
Smoke from Oregon forest fires
EPA 24 hr standard
EPA annual standard
Interstate 5 in Sacramento Impact of I-5 on Downtown Sacramento
10,500 light duty, 1125 heavy duty (> 5 axel) vehicles/hr,
10 traffic lanes, nearest 100 m to Crocker Art Museum site
Sound wall; some trees but not a barrier Prevailing wind in daytime from southwest,
across I-80; stagnation low winds from southeast Cut section freeway; complex terrain prevents
direct line source modeling; use very fine diesel tracers
DELTA Group slotted 8 DRUM Impactor• 8 size ranges:
• Inlet ( ~ 12) to 5.0 μm• 5.0 to 2.5 μm• 2.5 to 1.15 μm• 1.15 to 0.75 μm• 0.75 to 0.56 μm• 0.56 to 0.34 μm• 0.34 to 0.26 μm• 0.26 to 0.09 μm
• 10.4 l/min, critical orifice control, ¼ hp pump
• 6.5 x 168 mm Mylar strips• For 42 day run, 4 mm/day,
time resolution = 3 hr.• Field portable
• 10 kg, 43 × 22 × 13 cm
43
cm
Very fine (0.26 > Dp > 0.09 µm) DRUM sample, 3 weeks, South Lake Tahoe; 1cm high, true color
UC Davis DELTA Group S-XRF x-ray spectrum, light sample; no blank subtraction
Co57 γ ray standard
1213
1415
1617
1819
2021
2223
2425
2627
2829
3031
12
34
56
78
910
1112
1314
1516
December, 2002 January, 2003
0102030
405060
70
Nan
ogra
ms/
m3
Sulfur Zinc x 10 Phosphorus x 10
ALASET HETF Sacramento I-5 Transect StudyDELTA DRUM very fine particles (0.26 > Dp > 0.09microns) , S-XRF analysis
Possible tracers of diesel exhaust
Predict 4.3 2.9 µg/m3 diesel mass (U. Minn.)
rain rain rain Fog and haze
Crocker Art Museum site
29 30 31 1 2 3 4 5 6 7 8 9
January, 2003
0
5
10
15
20
25M
icro
gra
ms/m
3
Crocker Art ARB 13th & T Sacramento River
"Droplet" Mode Aerosol Mass1.15 > Dp > 0.75 microns
29 30 31 1 2 3 4 5 6 7 8 9
January, 2003
0
5
10
15
20
25
Mic
rogra
ms/m
3
Crocker Art ARB 13th & T Sacramento River
Very Fine Mode Aerosol Mass0.26 > Dp > 0.09 microns
Observe 6 1 µg/m3 mass
rain
Impact of Watt Avenue on Arden Middle School
3800 light duty, 53 heavy (> 5 axel) vehicles/hr Stop light at NW corner of school grounds 9 traffic lanes, nearest 15 m to school buildings Prevailing wind in daytime from southwest,
across Watt Avenue; stagnation winds weak from southeast
No wall or vegetation barrier; line source modeling easy from Tuscarora PM2.5 data
HETF Site
Sebastian Way site
Arden Way
Wa
tt A
ve
Effect of roadway distance and configuration on downwind concentrations of lead 1.
Roadway Distance 27 m 40 m 100 m 160 m
At grade Calculated 4.0* 3.4 1.4 0.41
Pasquill C *not scaled!
At grade Measured 4.0 3.1 1.4 0.35
Depressed Measured 4.5 1.7 0.26
Elevated Measured (2 sites)
4.8 2.3 3.1 (3.5)(one site)
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 1 2 3 4
January (dotted lines - weekends) February, 2004
0
10
20
30
40
ng/m
3
Sulfur Zinc x 10 Potassium/2
Aerosols at Sebastian Way - UpwindVery fine (0.26 > Dp > 0.09 micron) aerosols
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 1 2 3 4
January (dotted lines = weekends) February, 2004
0
10
20
30
40
ng
/m3
Sulfur Potassium/2 Zinc x 10
Aerosols at Arden Middle School - DownwindVery fine (0.26 > Dp > 0.09 micron) aerosols
Application of Tuscarora data Sliding box model on roadway Upwind regional PM2.5 data (CARB) Regional meteorology (NOAA’s ARL HYSPLIT 4) PM2.5 inlet (IMPROVE); 3 stage rotating DRUM
impactor at 1.15, 0.34, and <0.15 μm diameter Line source model, neutral stability Soft beta gauge PM2.5 mass ~ 7.0 ± 1.5 µg/m3 (
~8 µg/m3 with uf correction) Calculated mass, 3.7 µg/m3 light, 1.7 µg/m3 trucks,
for 5.4 µg/m3; with literature values, 7.7 µg/m3
Summary of Watt Avenue impacts
Mass numbers calculated are in fair to good agreement with observed values, car and truck
About 1/3 of the mass is diesel, even through diesels were only about 1.5% of the average vehicle mix
Very fine aerosols at the downwind school site have many of the elements predicted from the laboratory and Tuscarora studies,
Very fine aerosols 300 m upwind (Sebastian Way) are much lower that downwind at the school, and much lower than downtown Sacramento.
Very fine aerosols characteristic of diesels/smoking cars in Fresno > 1 km from freeways
1516
1718
1920
2122
2324
2526
2728
2930
12
34
56
78
910
1112
1314
1516
1718
1920
21
November December
0
50
100
150
200
250
300
Nan
ogra
ms/
m3
Phosphorus x 10 Sulfur Potassium Zinc x 10
Aerosols at the Fresno First Street Super-siteVery fine (0.26 > Dp > 0.09 micron) elemental concentrations for FACES, CARB
S-XRF analyses via DELTA Group, UC Davis
Predicted diesel vf/uf mass 11/25 – 12/17, 9 ± 6 µg/m3
1516
1718
1920
2122
2324
2526
2728
2930
12
34
56
78
910
1112
1314
1516
1718
1920
21
November December
0
5
10
15
20
Con
cent
ratio
ns/m
3
Mass by soft betas (ug/m3) Zinc by S-XRF (ng/m3)
Very fine (0.26 > Dp > 0.09) Aerosols at Fresno, CA
1516
1718
1920
2122
2324
2526
2728
2930
12
34
56
78
910
1112
1314
1516
1718
1920
21
November December
0
5
10
15
20
25
30
Mas
s ug
/m3
Mass measured Mass predicted based on U. Minn. diesels, CA fuel
Very fine (0.26 > Dp > 0.09) Aerosols at Fresno, CASupersite, First Street, > 1 km from nearest freeway
Nov 25 – Dec. 17, measured mass = 4.8 ± 1.0, predicted mass = 6.6 ± 4.8,
estimated total very fine mass = 8.8 μg/m3
Persistence of very fine/ultra fine particles We find high concentrations of very fine particles
near freeways in periods of stagnation, We also find significant concentrations (6 µg/m3)
over large areas of downtown Sacramento well away (> 400 m) from the nearest freeway source
Even higher levels are seen in Fresno in winter, in a residential neighborhood > 1 km away from the nearest freeway.
Recent work in Los Angeles finds ultra fines falling off from freeways roughly as rapidly as diffusion-limited CO, so coagulation appears to be limited and lifetimes long (for the non H2SO4 fraction).
Study of ultrafine particles near a major highway with heavy-duty diesel traffic Zhu et al (2002)
Study of ultrafine particles near a major highway Zhu et al (2002); Lead from Cahill et al (ARB, 1974)
-240-200
-160-120
-80-40
040
80120
160200
240280
320360
400
Upwind Distance from Freeway 710 (m) Downwind
0
100
200
300
400
500
600
700
800
900
1000
Re
lativ
e v
alu
es
CO Black Carbon Number Lead (1974)
CO, Pb - cars
BC, number -diesels
Conclusions
Diesel exhaust mass lies almost entirely in the very fine mode < 0.25 µm; ultra-fine mode misses ~ ½ of the diesel mass.
Laboratory and field studies are consistent in composition and particle size, allowing calculation of diesel impacts.
Separation of diesel and smoking car exhaust is difficult, organic tracers unstable and uncertain, and costs high.
Smoking car exhaust appears to be as toxic as diesel exhaust per unit mass.
Diesel and smoking car exhaust represents most of the very fine mass in Sacramento and Fresno, even well away from freeways.
Diesel and smoking car exhaust must be measured in ambient urban and sub-urban sites where it is a significant and toxic fraction of PM2.5 mass.
Proposals for California
We propose getting automobile smoke listed as a Toxic Air Contaminant (TAC) under California Prop 65, joining diesel Diesel is presently 70% of the cancer risk of all current California TACs,
combined. We have initiated PM0.25 mass measurements in the CA Central
Valley, using High Schools and air district support. We have deployed robust, simple and inexpensive monitors to begin
generating data on possible PM0.25 – health associations We propose that these urgent but unmet needs can be addressed via a
new California PM0.25 mass standard It would include diesels, smoking cars, and other high temperature processes
(coal fired power plants, smelting, welding, ….) that generate very fine and ultra fine particles of suspected health impacts.
It avoids the difficulty and expense of routinely resolving the diesel-smoking car overlap, since it appears both are toxic and both must be controlled
It could in the future allow the possibility of US EPA re-evaluating the PM2.5 standard if it turns out that 2.5 μm – 0.25 μm mode aerosols prove less harmful to human health.
Acknowledgements American Lung Association – Sacramento Emigrant Trails
Health Effects Task Force – many volunteers, modest funding, excellent oversight
National Renewable Energy Laboratory – DRI diesel tests Health Effects Institute – Tuscarora Tunnel studies California Air Resources Board – Fresno FACES study Tahoe Regional Planning Agency and CalTrans – Tahoe
Studies Department of Energy - Lawrence Berkeley NL and the
Advanced Light Source – S-XRF capabilities National Science Foundation – ACE-Asia DRUM samplers
“…ultra-fine particles near a busy road” – (Gramotnev and Ristovski Atm. Env. 2003)
1112
1314
1516
1718
1920
2122
2324
2526
2728
2930
311
23
45
67
89
1011
1213
1415
16
January February, 2002
0
10
20
30
40
50
60
Nan
ogra
ms/
m3
Zinc * 10 Sulfur Phosphorus x 20
Very fine (0.26 to 0.09 micron) Aerosols at South Lake TahoeTracers of diesel exhaust/smoking cars
Wood Smoke at South Lake Tahoe
1011
1213
1415
1617
1819
2021
2223
2425
2627
2829
3031
12
34
56
78
910
1112
1314
1516
17
August September
0102030405060708090
Na
no
gra
ms/m
3
0.75 to 0.56 0.56 to 0.34 0.34 to 0.26 0.26 to 0.09
Aerosols at South Lake Tahoe, Summer, 2002Non-Soil Potassium, UCD DELTA DRUM, S-XRF Analysis
Oregon fire smoke