1 Long-term Measurement of Ultrafine Particles in the Urban and Rural EnvironmentDirk Felton, P.E. Research Scientist III Seminar: AWMA Niagara Frontier Section January 25, 2017 Division of Air Resources Bureau of Air Quality Surveillance
2 What are Ultrafine Particles?Solid or liquid droplets in a specific size range Primary Natural Sources: sea spray, volcanos, wildfires Anthropogenic Sources: Outdoor: combustion emissions Indoor: manufactured products including printer toner, carbon nanotubes and medical and skin care products Secondary Nucleation and Gas to Particle reactions Nucleation – Particles form because environmental conditions favor particles over gasses Combustion emissions - emitted gasses cool leading to particle formation Gas to Particle reactions Sulfate and photochemistry (SOA)
3 Invisible Particles Scatter LightRed Blood cell Hair Bacteria Virus DNA Oxygen 1 Meter 1 Million Microns 100 Nanometer 0.1 Microns 10 Nanometer 0.01 Microns 1 Nanometer 0.001 Microns 0.1 Nanometer 0.0001 Microns 10,000 Microns 1,000 Microns 100 Microns 10 Microns 1 Microns 100,000 Microns Rain Drops Mist/Fog Tobacco Smoke Visible Particles Invisible Particles Scatter Light Ultrafine particles Molecules UFP Size: How small are we talking about? A Micron is a Millionth of a Meter A Nanometer is a Billionth of a Meter Wavelength of Visible Light: 390 – 700 nm Each black line as you move to the right divides the size by 10 UFP are defined by size – nm 2 orders of magnitude Particles in a size that can stay suspended cover about 6 orders of magnitude UFP size range is roughly between the wavelength of light and the size of molecules Oxygen molecule 0.5 nanometers
4 Weight in a volume of air (µg/m3) Red Blood cell Hair Bacteria Virus DNA Oxygen 1 Meter 1 Million Microns 100 Nanometer 0.1 Microns 10 Nanometer 0.01 Microns 1 Nanometer 0.001 Microns 0.1 Nanometer 0.0001 Microns 10,000 Microns 1,000 Microns 100 Microns 10 Microns 1 Microns 100,000 Microns Particles measured by Weight in a volume of air (µg/m3) Particles measured by Number in a volume of air (#/cm3) Particles larger than 100 nm are large enough to have enough mass to be measured by determining the weight of particles collected on a filter. We can’t weigh ultrafine particles because of geometry PM-2.5 particles have a volume of about 8 cubic microns – 100 nm particles have a volume nearly indistinguishable from zero Volume = 4/3 Pi R^3 A 100 nm particle has % of the volume of a PM-2.5 particle
5 1971 – Total suspended particles Red Blood cell Hair Bacteria Virus DNA Oxygen 1 Meter 1 Million Microns 100 Nanometer 0.1 Microns 10 Nanometer 0.01 Microns 1 Nanometer 0.001 Microns 0.1 Nanometer 0.0001 Microns 10,000 Microns 1,000 Microns 100 Microns 10 Microns 1 Microns 100,000 Microns 45 Years: Evolution of PM NAAQS 1971 – Total suspended particles 1987 – PM10 daily 150, Annual 50 mg/m3 1997 – PM2.5 daily standard 65 mg/m3 2006 – PM2.5 daily standard 35 mg/m3 2012 – PM2.5 annual standard lowered from 15 to 12 mg/m3 20?? – PM NAAQS (currently under review) Total Suspended Particulates PM10 Ultrafine particles PM2.5 Ambient Ultrafine Particles are not regulated in the US What about regulation? 45 years of ambient air quality standards and monitoring methods have been based on particle mass NAAQS are Based on impacts to health and include 4 criteria: Indicator, Avg Time, Level, Form To regulate UFP, the indicator and monitoring methods will have to change In the EU, UFP emissions standards are being developed.
6 Size Distribution of Ambient ParticlesNucleation: secs-minutes Aitken: hours-days Accumulation: days-weeks Coarse: Particles in the environment are typically found in one of these size categories - occur as a result of a formation process or because they are a size that is relatively stable Formation (short lived): Nucleation and Coarse mode Stable: Aitken and accumulation mode Nucleation – evaporation, diffusion Aitken – diffusion, coagulation Accumulation – in between diffusion and gravitational settling (long range transport) Coarse – deposition settling
7 What Processes are Involved?Increase UFP (Due to primary combustion emissions) Direct emission of carbon or metallic particles Emitted Gases convert to Particles Nucleation and Condensation Decrease UFP (creates gradient away from source) Volatilization, Diffusion, Dispersion Coagulation and Condensation These processes are effected by environmental conditions: (temp, sunlight, humidity, wind speed) Gas to Particle reaction includes photochemistry The processes that decrease UFP occur more slowly when it is cold and less humid Condensation can move particles into the UFP range or more likely out of the UFP range
8 Health Implications UFP can easily be inhaled and pass into the lungsThe small size of UFP allows for interstitial passage from nasal and lung tissue into the bloodstream where normal particle removal processes are ineffective Biological responses from oxidative stress such as inflammation and increased blood pressure are evident even when the particles are low in toxicity and are strongest with smaller particles1 Composition of UFP including presence of transition metals may increase toxicity2 Biological responses may be due to particle size, number, composition, surface area or charge 1Pieters, Environmental Health Perspectives, March 2015. 2Romieu, European Respiratory Journal, January 2008. We do know that there are biological responses to UFP. Uncertainty over what causes biological response, size and number, composition, charge, surface area
9 Review for 2012 PM NAAQS The 2009 Integrated Science Assessment (ISA) found that the linkage between UFP and cardiovascular and respiratory short-term health effects was Suggestive not Causal CONSISTENCY of the observed association STRENGTH of the observed association SPECIFICITY of the observed association TEMPORALITY of the observed association GRADIENT clear exposure-response relationship PLAUSIBILITY credible association EXPERIMENTAL results that support the relationship COHERENCE evidence from various fields The EPA requires some level of certainty before they set a NAAQS for a pollutant These Factors describe the relationship between the pollutant and the health effect The EPA began a review of the PM NAAQS in 2015 and should complete the process in 2018.
10 NYSDEC Ambient UFP MonitoringUrban Neighborhood: Queens College Seasonality, diurnal pattern, urban background Rural: Pinnacle State Park Steuben County Seasonality, long range transport (Whiteface?) Near Road: Buffalo on I-90 Between Exit 51 and 52 Rochester on I-490 at Exit 20 Queens on LIE 495 at Exit 23 (Summer 2017) Mobile source impact, diurnal pattern, seasonality Peace Bridge Special Study: (concluded Sept, 2015) Busti Avenue close to PB Plaza PS 198 Neighborhood background (Summer) Where do we Monitor UFP? Whiteface started in June
11 Instrumentation: Ultrafine Particle NumberAPI Model 651 TSI Model 3783 Water CPC Lower size cut 7nm (0.007 microns) 1 Micron Cyclone Inlet 2nd Unit was on Loan from the Manufacturer First instrument designed for continuous ambient air quality monitoring Data units are # of particles/cm3 (minute or hour) First released in 2011 Went through revisions – updated version was released in 2013 Before this instrument was developed researchers didn’t collect long-term datasets
12 Condensation Particle CounterCPC Condensation Particle Counter UFP enter from the bottom and grow in size due to condensation UFP are much smaller than the wavelength of light so condensation is used to grow the particles into a larger size so they can be counted by pulsing a laser
13 UFP Data: Precision The two instruments demonstrate excellent precision At very high particle concentrations one instrument is slightly higher One of the inlets appears to have frozen for a day in February
14 UFP Data: Precision 1-Hr AveragesThe precision of the UFP instrument is better than any of our gas analyzers We are confident in our ability to measure UFP accurately
15 UFP Instrument Siting UFP Monitors require shelters with full HVAC (20-25 Deg C) UFP Cyclone Inlet The UFP instruments are installed with the same inlet configuration at each site The water CPC is robust and can be operated in remote field locations where an environmental shelter is available.
16 UFP Data: 1-Hr Averages (particles/cm3)At this urban location emissions are consistent: UFP are typically higher on the weekdays and data show a strong daily pattern (weekends in orange) Month of Peace Bridge Data – The UFP varies every day – low at night, high during the day but peaks vary Sources are consistent day to day The concentration of UFP at this location varies mostly due to environmental conditions Consistent with a pollutant that has a high source oriented component than transport component
17 UFP Data: 1-Hr Averages If you smooth the data you can see an underlying multi-day pattern due to meteorology (WD, WS, Temp, RH) The data show evidence of daily sources combined with multiday patterns During the Fall and Spring we often have multi day inversions that can lead to build up of locally generated pollutants Weather can decrease UFP rain and high winds increase UFP – cold, low humidity and low wind speeds
18 How does UFP Respond to Environmental Changes?We are confident that we can measure UFP How can we characterize the impact the environment has on UFP?
19 Peace Bridge UFP: Seasonal Time SeriesData are presented with Winter in the Center of the plot UFP is more stable in cold temps < Evaporation < Humidity < Particle Growth also Low B Layer Sources are similar year round but UFP is twice as high in the winter The time of year at least here in Buffalo can cause a 2 to 3 fold increase in UFP The published studies of UFP don’t show this and I realized that the early research on UFP was all done in Los Angeles
20 Peace Bridge UFP: Sorted by Temp (Deg C)At colder temps, the UFP range is higher but the low values are similar year round Local mobile source emissions are consistent year-round
21 Rolling Weekly Ave. UFP and Temp (Deg C)At colder temps, the average UFP is 2 to 3 times higher than the average during hot weather
22 Rolling Weekly Ave. UFP vs Temp (Deg C)XY Plot (UFP vs Temp ºC) shows the relationship between UFP and temp at this location
23 How does the Temperature Dependence of UFP in Buffalo Compare to other Locations?
24 Rolling Weekly Ave. UFP vs Temp (Deg C)Boise, Idaho Near Road 2012 Data Winter in Center of time series plot Boise Near Road AADT: 103,000 Boise, Idaho the dependence and the actual concentrations are very similar to Buffalo Boise was one of the first cities to have a near road monitor
25 Rolling Weekly Ave. UFP vs Temp (Deg C)Rubidoux, CA 2012/13 Data Winter in Center of time series plot UFP increases by about 28% during cooler periods But the concentrations are lower overall Central Valley east of Los Angeles
26 Rolling Weekly Ave. UFP vs Temp (Deg C)Queens, NYC Data Winter in Center of time series plot UFP is higher at low temperatures but UFP also increases at higher temps Baseline is elevated in NYC Here we see that there is a another phenomenon going on that increases UFP as temperatures increase
27 Rolling Weekly Ave. UFP vs Temp (Deg C)Long Beach, CA 2012/13 Data Winter in Center of time series plot UFP increases by 20% - 40% at low temps UFP is increased by a factor of 3 at high temps Long Beach is the 2nd busiest Port in the country and has a tremendous amount of truck traffic to move all that cargo
28 Seasonal Diurnal UFP: Long Beach, CAMATES IV Study Summer afternoon Peak is evidence of Local Secondary UFP production Elevated UFP in the Summer occurs in the afternoon, not at commuting times Local SOA > Local Primary Production → high VOC Multiple Air Toxics Exposure Study South Coast Air Basin
29 Secondary UFP Production: QueensSunset 1-Hr OC PM data are elevated as temps increase above 20-25º C Elevated OC PM2.5 matches temp range when UFP increases We don’t have information on UFP species but we do have PM-2.5 species data.
30 What size fractions increase at what temp?This plot shows the UFP size fractions sorted by temp Smaller incr more when it is colder than warmer Larger incr more when it is warmer Combined data 2009 – 2010 and TSI 3031 Particle Detection Limit: 20 nm TSI 3031 separates UFP into 6 size categories by charging the particles and by using a Differential Mobility Analyzer to determine the number of particles in each size category. This data is limited to particles larger than 20nm
31 UFP and Environmental ConditionsCold: UFP are elevated when temperatures are cold because the removal processes (evaporation and agglomeration) are reduced Hot: UFP can be elevated when temperatures are hot because production processes are increased but: Requires precursors (pollutants) and sunlight Must overcome increased removal processes Moderate: UFP typically are lowest when temperatures are (15º ± 3º C) Good for human exposure?
32 Urban, Rural and Near-road UFP Data in New York State
33 Urban UFP Data: Queens, NYCAverage UFP is 20,000/cm (5 times the rural site) Primary UFP increases during morning and evening commuting hours The mid-day peak indicates a photochemical process – likely Secondary Organics Urban UFP only approached rural levels after a 4 day Nor’easter in October 2015 Diurnal Plot Urban site UFP higher than rural site average 99% of the time
34 Rural UFP Data: Steuben CountySite is removed from sources Average UFP is 4,100/cm % of urban UFP There is no evidence of a morning rush hour The Midday peak indicates a photochemical process – likely Secondary Organics At night UFP briefly increases due to condensation and then decreases due to condensation as particles grow out of the UFP size range At rural sites we can see how UFP behaves when we are away from sources
35 Near Road Site for the Buffalo/Niagara CBSANew EPA regulations require NY to have 2 monitors in NYC 1 in Buffalo 1 in Rochester The Buffalo site is 20 m from I-90 between Exit 51 and 52 (Cheektowaga) AADT is 131,019 EPA recommends monitoring UFP and Black Carbon at these sites EPA is trying to get a better understanding of the near road gradient Sites are located between 20m and 50m from travel lane The houses in this picture are about 50-60m from the travel lane
36 Near Road Air Quality MonitoringMillions of people in this country live near roadways New Federal Air Quality Standards for NO2, CO and PM2.5 require Near Road Monitoring in cities (MSAs) with populations over 1 Million 400m 500m from road 300m 200m 100m 10 Million people live within 100m of a road with 25,000 AADT 60 Million people live within 500m of a road with 25,000 AADT This is one of the reasons why it is important to understand the gradients of mobile source pollutants The further mobile source pollutants travel away from the road, the greater the number of people who are exposed
37 Near Road, Urban and Rural UFP DataBuffalo Near Road UFP average concentrations are highest during the morning commute and higher than NYC urban average UFP concentration is a better indicator of proximity to a source than it is to the size of the source Near Road data in comparison to urban neighborhood and rural data
38 Busti Avenue Downwind, Max Impact SitePeace Bridge (Mobile Source Study) Busti Avenue Downwind, Max Impact Site The shelter was located about 40 m from the Peace Bridge Plaza and more than 200 m from I-190 Customs Clearance Lanes Customs Clearance Lanes We monitored PM2.5, UFP and Air Toxics for 1 year
39 UFP and Theoretical Distance to Mobile SourcesUFP is often used as an indicator (tracer) of mobile source emissions because it has a very steep gradient downwind from roads UFP returns to background within a few hundred meters of edge of road Wind Direction Early research - Los Angeles – Ground Zero for vehicle emissions research The gradient is steep because the smallest particles generated by mobile source emissions are short-lived The steepest part of the decay of UFP occurs between 20 and 60 m from the edge of the road Zhu et al., JAMA 2002, Atm. Env., 2002 Karner, Environmental Science & Tech, 2010
40 UFP and Actual Distance to Mobile SourcesWind Direction Buffalo Near-road site is 20 m from I-90 Peace Bridge site is 40 m from the customs clearance lanes Early research - Los Angeles – Ground Zero for vehicle emissions research The gradient is steep because the smallest particles generated by mobile source emissions are short-lived The steepest part of the decay of UFP occurs between 20 and 60 m from the edge of the road Zhu et al., JAMA 2002, Atm. Env., 2002 Karner, Environmental Science & Tech, 2010
41 Peace Bridge UFP Data The Near-road UFP are higher than the Peace Bridge max impact site The Near road site is closer to the roadway and the AADT is higher than the traffic on the Peace Bridge and I-190 The shape of the peak is important. The Busti Ave peak is later – bridge delays and traffic increase through midday The near road site is 20 m from highway traffic Busti Ave is 40 m from bridge traffic The comparison between the two sites supports the strong gradient of UFP away from the roadway
42 Seasonal UFP: Summer vs WinterAt the near road site 20m from the road the summer and winter UFP are similar At the Busti Ave site 40m from the Peace Bridge the summer UFP are ½ of the winter 20m vs 40m - Seasonal comparisons also support the strong gradient of UFP 1. The decay of UFP is not as evident in the winter 2. There is no decay within 20 m. – turbulent area This may have an implication for exposure?
43 Can we Identify Sources of UFP with Ambient Data?
44 Buffalo Near Road UFP DataBuffalo Near Road UFP Weekday-Weekend comparison show that motor vehicles are responsible for more than 50% of the UFP at this location The easiest analysis is weekday vs weekend
45 Peace Bridge UFP and Auto TrafficAt the Peace Bridge, auto traffic correlates with UFP on weekdays but, on weekends, auto traffic increases and UFP decreases Auto traffic and UFP are not proportional What about cars vs trucks UFP is higher on weekdays and lower on weekends – opposite of auto traffic
46 Peace Bridge UFP and Truck TrafficThe pattern of UFP closely matches the hourly average weekday truck traffic on the Peace Bridge The UFP and Truck traffic are lower on the weekends but the diurnal patterns are very similar The weekend offset is likely due to other sources of UFP including biomass burning The weekend UFP offset begins late on Friday and winds down late on Sunday
47 Why Do we Monitor UFP in a Rural Area?
48 Rural UFP Data: Steuben CountySite is removed from sources so natural processes that increase and decrease UFP can be evaluated Midday peak is higher in the winter In the summer, the production of UFP is higher in the evening due to gas to particle reactions and condensation when evaporation is at a minimum At rural sites we can see how UFP behaves when we are away from sources We saw this diurnal plot before but this time summer and winter are separate There is a constant tug of war between particle production and particle loss Condensation evaporation agglomeration condensation
49 Rural UFP and Volatile PM MassThe site has a 1405DF TEOM that provides the volatile fraction of PM2.5 The volatile fraction of PM2.5 is about 20% of the total at this site in the wintertime UFP are short-lived and sometimes correlate with the volatile fraction of PM-2.5
50 Rural UFP and Volatile PM MassVolatile PM2.5 correlates with UFP midday Since there are no local sources, the increase in PM confirms the presence of new particles The decrease in volatile PM2.5 in the afternoon suggests UFP are evaporating not agglomerating and not growing The decrease in volatile mass after noon suggests that the UFP are evaporating and not agglomerating
51 Rural UFP and Volatile PM MassVolatile PM2.5 is anti- correlated with UFP in the evening As the sun goes down, condensation increases In the evening, some new particles are formed but particles “grow” out of the UFP range and quickly increase volatile PM2.5 mass Once the particles are over 100 nm they are counted as UFP but they become a bigger component of PM-2.5 The increase in volatile mass in the evening suggests that the UFP are agglomerating and not evaporating
52 Daily Change in UFP in Rural AreasIncrease UFP Gas to Particle Reaction (new) Condensation (1nm) Result: New particles Decrease UFP Evaporation (gas) Agglomeration (mass) Condensation (100nm) Result: Gas, larger particles There is a constant tug of war between particle production and particle loss: temp, sunlight, availability of pollutants Pollutants can quickly change state between UFP, gas and larger particles Nitric Acid – Sulfate - Organic Aerosols When I talk about UFP to the public they are usually concerned about the health impact of particle number What we are learning is that some pollutants can quickly change from small particles to gases and we are all exposed to these pollutants no matter what state they are in.
53 Findings: Behavior of Ultrafine ParticlesUFP decreases within meters of emission source Near-road - after the turbulent zone UFP concentrations are sensitive to environmental factors UFP lowest in moderate temperatures UFP typically highest in the winter when UFP are more persistent and primary UFP travel further into neighborhoods UFP can be highest in hot weather when conditions exist for particle production Diffusion
54 Conclusions UFP Monitoring technology has greatly improved Water CPC has enabled this work Millions of people live near busy roadways and are exposed to elevated levels of UFP PM NAAQS are currently under review UFP are included in the Assessment (ISA) Rural monitoring sites provide evidence of UFP transport, production and loss mechanisms Urban sites have higher UFP due to the combination of production and primary sources Near Road sites show steep gradient for UFP
55 Next Steps: Study to Address Intra-urban UFP1-Yr study with 3 fixed sites and 4 portable monitors How does mobile source UFP move from busy roadways through urban areas in a typical community with grid shaped roadways? Where are the UFP hotspots in an urban area? What vehicle type is primarily responsible for high UFP in urban areas? How well does a central fixed monitor represent the UFP in an urban area?
56 Thank You Connect with us: Facebook: Twitter: twitter.com/NYSDEC Flickr: Dirk Felton, P.E. Research Scientist III 625 Broadway Albany, NY