Roadway Impacts & Mitigation: Past, Present, Future

1. Freeways and Secondary Roadways - something old, something new, something borrowed, something green A discussion for the UC Davis workshop on near roadway impacts and mitigation, January 24, 2007 Thomas A. Cahill Physics Department and the DELTA Group, University of California, Davis with the Breathe Calif./SET Health Effects Task Force

2. Why this sudden interest in secondary roadways? Exacerbation of pre-existing conditions� Increased population and roadway traffic, especially on secondary roads in residential neighborhoods Difficulty of present roadway models to predict near-roadway transport of pollutants in non-ideal conditions Difficulty of traffic models to reflect accurately actual roadway pollutant emissions New data on Health impacts � near freeway epidemiological studies very fine and ultra fine particle studies Emissions � car exhaust/diesel lube oil Transport � very fine and ultra fine particles, and Mitigation � somethings old, somethings new

5. I will give three examples of some relevant work we are doing� Example 1. Meshing the extensive 1970s work on freeways and transport with current data Roadway design has a major impact on downwind transport, Downwind transport is dispersion driven, and Urban very fine/ultra fine aerosols are dominated by secondary road emissions, not freeways For fine, US EPA MOBILE - 2/3 diesel, 1/3 cars For fine, CA ARB EMFAC 2007 - 1/3 diesel, 2/3 cars Note: Much of the early work is published in the gray literature but not available electronically. I will scan some of this material into the workshop CD.

6. Another example of some relevant work in progress� Example 2. Studies of heavily traveled secondary roads (Watt Avenue, �) in Sacramento Higher vf/uf impact by 65,000 v/day (1.5% diesel) than by freeway I-5 (170,000 v/day, 10% diesels) � potential reasons�. Closer spacing to receptors, Lack of barriers roadway to receptors, Stop and go traffic Dirtier diesels ? Local impact dominated by 3 winter months Organic matter very fine/ultra fine in size and richer in heavy PAHs that diesel exhaust

7. The last project is of some work on near-roadway mitigation (later talks) Example 3. Mitigation by vegetation � why now? highly efficient lung capture for particles < 0.1 �m Used oil from spark emission vehicles higher in PAHs than from diesel vehicles (Fujita et al, DRI) Major laboratory PAH concentrations in sizes < 0.15 �m Major near roadway ambient heavy PAH (BaP,..) concentrations < 0.1 �m Enhanced deposition velocities for particles < 0.1 �m due to increased diffusion � 15 x at 0.02 �m vs. 0.2 �m Vegetation is an attractive way to get a deposition surface Best mitigation: Get the small fraction (circa < 2%) of gross emitting cars off the road!

8. Example 1. Meshing the extensive 1970s work on freeways and transport with current data Lead was a unique, conserved, and, we now know, ultra fine tracer of car exhaust There was a major ARB/CalTrans effort 1972 -1974 to understand lead from Los Angeles freeways Our component: 5 LA freeway sites (1 flat, 2 cut, 2 raised section) 120 transects, each with 2 hr resolution, day and night, 5 size modes, Full elemental analysis > sodium Line source modeled concentrations

12. Results for the Santa Monica freeway � also used by US EPA for their model

13. Lateral transport from freeways: theory, lead from at grade and from cut (depressed) freeway configurations

14. Lateral transport of ultra fine particles � efficient transport, no coagulation!

15. Lateral transport at grade

16. What was the effect of the two upwind freeways? Assume 8 km upwind Assume �sliding box� mixed cell = 4 m Assume LA inversion = 400 m Assume no coagulation, settling, etc. Then = Concentration ~ 1% of freeway peak Even adding the San Diego freeway,, no more than a few % from freeways. Actual upwind value circa 15% of near roadway peak Thus, circa � BC and particle number from non-freeway sources Note: ARB EMFAC ~ 2/3 cars, 1/3 diesels

17. January 6, 2007

18. And more �.Check the distances: < 530 m, <1060 m, 1060 m to 1600 m, > 1600 m

19. Lateral transport � at grade, cut and fill � no trees or barriers

20. New information � vehicular emissions Size and composition of diesel aerosols, including ultra fines (U. Minn./DRI/UC Davis) Roadway studies of diesel and auto emission rates California CERC and Nevada DRI laboratory data HEI Tuscarora PA tunnel study freeway studies CA Air Resources Board studies of freeway ultra fines Breathe California (ex- Amer. Lung Assoc) studies of secondary roadways in Sacramento Toxicity of used diesel and spark emission vehicle Lubricating oils - Nevada Desert Research Inst. EPA Region IX/ASU/UC Davis organics, trucks, trains, and cars

21. U. Minn. Dynamometer Diesel tests; DRI mass and sulfates, DELTA Group S and elements

22. U. Minnesota Dynamometer Diesel Tests; same California fuel, different engine � no mention of smoke

23. New information on the toxicity of car exhaust There is evidence that spark emission car exhaust has more heavy PAHs than diesel truck/bus exhaust Theory of PAH formation makes small cylinder vehicles worse than large cylinder vehicles Temperature of formation for PAHs is low, < 600 C Higher cylinder wall to volume ratio, cars vs trucks Gertler at al 2002 had the benzo-a-pyrene emission rates roughly the same per vehicle, cars vs trucks, for the HEI Tuscarora Tunnel Study We find relatively high ultra fine mass from the lubricating oil in CNG busses, ~ � diesel busses Eric Fujita at Desert Research Institute showed used spark emission lubrication oil was 10 to 20 times higher in PAHs than used diesel oil

24. Typical daytime traffic 50 m south of sampling site

26. Cars have more PAHs in their oil than diesels

28. Example #2: Arden Middle School at Watt Avenue � 65,000 v/day, > 98% cars

29. Watt Ave Traffic - 15 m from school: No mitigation by road configuration possible

30. HETF data unexpected fine mass � mis-tuned natural gas water heater � fixed in 30 min!

31. Result � Very fine/ultra fine mass at Arden Middle School

33. Mass in the finer fractions

34. Comparison of composition: Arden Middle vs. Roseville rail yard

35. Surprise! Vanadium is most likely associated with bunker oil combustion from ships in the port of Oakland.

36. PM2.5 Sac. 13th and T, Jan - Mar, 2007

37. PM2.5 Del Paso Manor, Jan - Mar, 2007

38. Even size and time resolved aerosols track across 4 miles of Sacramento (wood smoke)

39. PM2.5 Del Paso Manor, Mar - May, 2007

40. Surprise! Upwind and downwind are essentially equivalent (TMC, ASU in progress)

42. Car exhaust dominates Watt Avenue PAH concentrations

43. Example #3: Mitigation via vegetative capture - theory Very fine (< 0.25 ?m) and ultra fine (< 0.1 ?m) diameter particles have suspected health impacts via several mechanisms including - Insoluble ultra fine particles in the lung and heart Carcinogens in the lung Very fine (< 0.25 ?m) and ultra fine (< 0.1 ?m) diameter particles have relatively high removal rates via diffusion if a surface is close Vegetation can provide such a surface

46. �To separate the effect of the planting from the freeway configuration would be a great mistake� (Cahill et al, ARB 502,1974) �The embankments of the cut section freeways were heavily planted.� The Santa Monica cut-section site had �a dense thicket of bushes ~ 20 feet high at the crest of the embankment hard against the right of way fence.� The Harbor cut section freeway site � was similarly planted, with eucalyptus and bushed extending higher than 30 feet on the downwind site� �The thickets were quite dense, and effectively cut the wind in their lee.�

47. UC Davis Mechanical Engineering 20 m wind tunnel

48. Size distribution of flare aerosols � scaled to tunnel wind velocity

49. Test in progress: Erin is watching the readout of wind velocity

50. Mitigation of very fine and ultra fine particles by vegetation (preliminary: ongoing HETF project)

51. Removal of very fine particles

52. Conclusions � wind tunnel vegetative capture studies Typically 75% of very fine particles 0.26 to 0.09 ?m are removed by 2 m of vegetation at 1 mi/hr wind velocity. Calculations indicate 95% removal of ultra fine particles in the same situation. This process becomes inefficient with wind velocities above 3 mi/hr Redwood and deodar are better than live oak.

53. Chui adding oleander branches

54. Very preliminary � branches versus empty box

55. With vegetative barriers on both sides (and ideally the median) of roadways, one benefits by - At high and medium wind velocities, turbulence mixes and lofts roadway pollutants At medium and low wind velocities, the barriers slow lateral transport and allow vehicular waste heat to loft pollutants At low wind velocities, very fine and ultra fine particles will be captured as they migrate through the semi-transparent barriers.

56. Mitigation options � we must move in parallel on all of them! Roadway source improvements, including Cleaner engines, fuel, and new artificial lubricating oils Removal of gross emitting vehicles ( ~ 3%) from roadways (worst 1% vehicles = ? 30% of vf/uf mass) Reduced traffic via transportation alternatives Roadway design options � �Complete Streets� Highway design; cut section, tunnel (cleaned!) Pollution barriers � use waste heat and vegetation to loft and trap uf particles (walls alone don�t work) Reduced Transport efficiency to residences Distance! This should be a key factor in new roads. Pollution barriers, especially vegetation Residential indoor air quality improvement Positively pressurized filtered receptors

57. Opportunities for new directions Add spark emission vehicle particulates to California's Proposition 65 Toxic Air Contamination roster? Not likely, I am told by those who know On road sensing for smoking cars and the means to remove them from the highways Worth an effort; we (BC/SET) already helped add smoke sensing to the inspection program last year. A new very fine or ultra fine mass standard Relatively easy implementation Uses existing filter � mass infrastructure Avoids routine expensive organic speciation (archived)