Ambient Elemental Signatures of Diesel and Automotive Particulate

Ambient Elemental Signatures of Diesel and Automotive Particulate Matter by Size, Time, and Composition. 1 Thomas A. Cahill, 1 Earl Withycombe, Steven S. Cliff, Michael Jimenez-Cruz, Lee Portnoff, and 2 Kevin D. Perry, DELTA Group, University of California, Davis, http: //delta. ucdavis. edu, 1 The Health Effects Task Force, American Lung Association, Sacramento-Emigrant Trails, 2 Dept. of Meteorology, University of Utah

Sources of information – joint studies with the DELTA Group n n n 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 o o o 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

U. Minnesota Dynamometer Diesel Tests

Time-of-Flight Organic Speciation vs. Particle Size (Prof. Kelly, UC Davis)

DOE Gasoline/Diesel PM Split Study Particle-Phase PAH in Lubrication Oil

Summary of laboratory tests o o o 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 Ca. CO 3 ; 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

PM 0. 25 ? PM 10 PM 2. 5 Average Zn to mass, all DRI tests, 1800 1300

Prior field studies (Gertler et al 2002)

Parameter Diesel Trucks (mg/km) Cars (mg/km) Ratio PM 2. 5 mass 135 + 18 14 + 13 10 Mass? Truck PM 10 = 181 PM 2. 5 OC 112 + 43 2. 8 + 1. 1 40 2 nd biggest ratio PM 2. 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 PM 2. 5 NH 3, S < 0. 8 1999; Some S in gasoline Roughly fuel mileage Gasses (g/km) CO 2 748 + 73 156 + 15 4. 8 CO < 0. 6 1. 9 + 0. 7 < 0. 3 NO (as NO 2) 11. 9 + 1. 9 0. 4 + 0. 07 28 THC 1. 5 + 0. 8 0. 4 + 0. 2 3. 7 Comment NO/CO > 90 (!) 3 rd biggest ratio

Summary for prior field tests o o Diesel truck have roughly 10 (PM 2. 5) to 18 (PM 10) 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 PM 2. 5 mode. There 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 o Interstate 5 in downtown Sacramento n o Watt Avenue on Arden Middle School n o o urban suburban Highway 50 in South Lake Tahoe The Fresno Super-site n suburban

Lung Assoc. Sacramento Transect Study Site Map and PM 2. 5 aggregated data #1, #3 - light blue = rain, yellow = “clear”, rest = fogs, wet and dry I-5 Hwy-50 I-80 Hwy-99

Sacramento – a highway nexus, (I-5, I-80, Hwy 50, Hwy 99) and close to violations of PM 2. 5 standards

Interstate 5 in Sacramento o Impact of I-5 on Downtown Sacramento n n n 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: • • 43 cm • • • 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

Very fine (0. 26 > Dp > 0. 09 µm) DRUM sample, 3 weeks, South Lake Tahoe; 1 cm high, true color

UC Davis DELTA Group S-XRF x-ray spectrum, light sample; no blank subtraction Co 57 γ ray standard

Crocker Art Museum site Predict 4. 3 2. 9 µg/m 3 diesel mass (U. Minn. ) rain Fog and haze

rain Observe 6 1 µg/m 3 mass

Impact of Watt Avenue on Arden Middle School n n n 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 PM 2. 5 data

Watt Ave HETF Site Arden Way Sebastian Way site

Effect of roadway distance and configuration on downwind concentrations of lead 1. Roadway At grade Distance Calculated Pasquill C 27 m 4. 0* *not 40 m 3. 4 100 m 1. 4 160 m 0. 41 0. 35 scaled! At grade Measured 4. 0 3. 1 1. 4 Depressed Measured 4. 5 1. 7 0. 26 Elevated Measured (2 sites) 4. 8 2. 3 3. 1 (3. 5) (one site)

Application of Tuscarora data o o o o Sliding box model on roadway Upwind regional PM 2. 5 data (CARB) Regional meteorology (NOAA’s ARL HYSPLIT 4) PM 2. 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 PM 2. 5 mass ~ 7. 0 ± 1. 5 µg/m 3 ( ~8 µg/m 3 with uf correction) Calculated mass, 3. 7 µg/m 3 light, 1. 7 µg/m 3 trucks, for 5. 4 µg/m 3; with literature values, 7. 7 µg/m 3

Summary of Watt Avenue impacts o o 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 Predicted diesel vf/uf mass 11/25 – 12/17, 9 ± 6 µg/m 3

Nov 25 – Dec. 17, measured mass = 4. 8 ± 1. 0, predicted mass = 6. 6 ± 4. 8, estimated total very fine mass = 8. 8 μg/m 3

Persistence of very fine/ultra fine particles o o We find high concentrations of very fine particles near freeways in periods of stagnation, We also find significant concentrations (6 µg/m 3) 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 diffusionlimited CO, so coagulation appears to be limited and lifetimes long (for the non H 2 SO 4 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) BC, number diesels CO, Pb - cars

Conclusions o o o 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 PM 2. 5 mass.

Proposals for California o We propose getting automobile smoke listed as a Toxic Air Contaminant (TAC) under California Prop 65, joining diesel n o We have initiated PM 0. 25 mass measurements in the CA Central Valley, using High Schools and air district support. n o Diesel is presently 70% of the cancer risk of all current California TACs, combined. We have deployed robust, simple and inexpensive monitors to begin generating data on possible PM 0. 25 – health associations We propose that these urgent but unmet needs can be addressed via a new California PM 0. 25 mass standard n n n 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 PM 2. 5 standard if it turns out that 2. 5 μm – 0. 25 μm mode aerosols prove less harmful to human health.

Acknowledgements o o o o 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 Cal. Trans – 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)

Wood Smoke at South Lake Tahoe Oregon fire smoke