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