SYTTY
URBAN PM10 AND PM2.5 CONCENTRATIONS AND TRAFFIC RELATED EXPOSURE TO FINE PARTICLES
Project leader: Taisto Raunemaa, University of Kuopio, Department
of Environmental Sciences, P.O.Box 1627, FIN-70211 Kuopio, Finland, tel.
+358- 17-163235, e-mail: Taisto.Raunemaa@uku.fi
| PUBLICATIONS |
| TIIVISTELMÄ SUOMEKSI |
Researchers:
Petri Tiitta, University of Kuopio, e-mail: ptiitta@uku.fi
Ari Leskinen, University of Kuopio, tel. +358-17-163 238, e-mail: Ari.Leskinen@uku.fi
Jarkko Tissari, University of Kuopio, tel. +358-17-163 235, e-mail:
Jarkko.Tissari@uku.fi
Tarja Yli-Tuomi Department of Chemical Engineering, University of Clarkson,
Tel. +1 315 268 6655, e-mail: ylituomi@clarkson.edu
Financing SYTTY organisation: Tekes
Funding from SYTTY / Total funding of project (€): 138586/193247
Person-months of work funded by SYTTY / Total person-months of work:
46/63
KEY WORDS: PM2.5, PM10, fine particles, traffic related particles,
impactor comparison
EXTENDED ABSTRACT
1 Introduction
The standards of the mass of suspended particle matter smaller than 10 µm (PM10) and 2.5 µm (PM2.5) size are revised since studies claim that breathing of particulate matter at concentrations below the current standard level are likely to cause significant health effects (E.g. Dockery and Pope, 1994). Premature death and an increase in respiratory illness has been documented (U.S.EPA Federal Register, 1997). Standardized and accurate measurement techniques and criteria for the location of monitoring are important in the assessment of ambient air quality with a view to obtain comparable information (EUR-Lex, 1999). In addition to sampler performance the criteria to sampling height and its proximity to local sources can play a significant role in the ability to assess human exposure (Chow 1997). Also the effects of gaseous emissions have to be reconsidered due to conversion to particles and adsorption on particle surfaces. The locations closest to the emission source are critical as they are within the breathing zone of people.
Large variations in concentrations between different monitor instruments were observed in recent survey in Finland (Yli-Tuomi and Raunemaa, 1997 and 1998, Raunemaa and Yli-Tuomi, 1997, Raunemaa, 1998). The present study was initiated in order to compare particle instruments for PM10 and PM2.5 monitoring. An USEPA reference method and a European reference method were chosen for basic comparisons. A new impactor sampler was included in various tests. Information on the monitor site placement was produced by performing field measurements close to the walkway of a local road. A new test procedure was developed during the project by adopting a large volume environmental chamber for long term sampler tests. Reliable results in exposure assessment are possible only if environmental conditions during monitoring and in sample analysis are also well controlled.
2 Methods
Sampler tests
Measurements were carried out in the city of Kuopio which is situated
in central Finland with population of approximately 90 000. Three different
study environments were applied: city monitoring site (Kasarmipuisto),
traffic influenced site (Savilahti) and the large volume environmental
chamber (Laboratory for Atmospheric Physics and Chemistry). In total six
methods were applied in the tests: two reference methods - EU standard
(Digitel DPM) and EPA standard (PQ-100, EPA-WINS), a new candidate instrument
(Dekati PM10/2.5 sampler), the TEOM monitor (EPA equivalent method), the
Gent PM2.5/10 sampler and Andersen virtual impactor PM2.5/10. The tests
method EN12341:99 (British Standard 1999) was applied when performing tests
with the candidate sampler.
More than 400 aerosol samples were collected during instrument comparison and site criteria experiments in 1998-2001. In most cases sampling was conducted for 23 hours, usually between 11 a.m. to 10 a.m, but shorter periods were possible in the environmental chamber experiments. Reports on the sampler tests are given by Tiitta et.al (1999) and (2000a). The 143 m3 volume environmental chamber made of 150 ?m Teflon attached on steel structures situates on the laboratory terrace. Particles are introduced into the chamber through injection pipelines. The mass concentration in the chamber could be adjusted between very low and several hundreds of µg/m3 concentration levels (variation <5%) for hours. Several instruments can be positioned on the test rig (Figure 1). A blower produces air currents (i.e. wind) on the samplers. Equal velocities to normal field situations (0-4 m/s) are possible. The testing times are largely reduced and samplers can be kept in equal environmental conditions. A long test can be performed within one week instead of days or months on the field. The test procedure may require that high particle concentrations are necessary, outdoors this might be possible only in heavy traffic situations where sampler testing is not possible.
Figure 1. Layout of the 143 m3 environmental chamber at University of Kuopio.
Traffic influenced site tests
Major research interest was in PM2.5 monitors due to their potential
to characterisize alveolar deposition. PM2.5 concentration was measured
at 10-85 m distances and 1.9-6.9 m heights to obtain data for monitor placement
on the side of traffic road. In the first series of measurements,
PM2.5 values were detected for 27 days close to the 18 000 vehicles/day
road in summer 1999 (Tiitta et al. 2000b).
In the summer 2000, three different heights (1.9, 4.9 and 6.9 m) and
two different monitoring distances (23 m and 59 m) were examined. PM2.5
concentration on opposite side of the road was also measured (Tiitta et
al. 2000c).
Identical EPA-WINS samplers were chosen for the curbside study because they are portable, battery- operated, easy to use and in a weatherproof case.The traffic flow, wind speed and direction, ambient temperature and relative humidity were registered. A Mettler MT5 microbalance (Mettler-Toledo AG) was applied for all gravimetric analysis of samples; the microbalance had an accuracy of 1 µg.
3 Results
Sampler tests
Sampler tests form a series of experiments with different sampler combinations.
Testing of the virtual impactor, applied in Swedish research, was conducted
in co-operation with the Chalmers Tekniska Högskolan researchers (Tiitta
et al.2000a). These measurements were carried out in the environmental
chamber using high particle concentrations. Good correlation between the
EPA-WINS PM2.5, Dekati PM2.5 and Virtual Impactor impactor samplers (R2>0.9)
were observed. Normalized differences between the devices ranged
between 0-15%. A poor correlation was documented for the Gent-impactor
at high particle concentrations and Gent sampler was therefore removed
from the tests. The sampler suites for background or moderate level monitoring
only.
Figure 2. a) PM10 concentrations by the candidate sampler (Dekati PM10) against the reference sampler (LVS-DPM) within limits of the acceptable envelope (+10µg/m3 when conc.<100µg/m3 and +10% when conc.>100µg/m3). b) PM2.5 concentrations by the candidate sampler (Dekati PM2.5) against the reference sampler (EPA-WINS).
With large particles higher concentration values than by the reference instrument were observed for the Dekati PM2.5 sampler. This was proposed be due to bounce effects on the impactor plates. The sampler performance was later corrected by the manufacturer.
There did not appear any significant difference in PM10 between the four instruments applied in the field tests at the city monitoring site. Average PM10 concentrations in Kasarmipuisto location were: DPM 11.7 µg/m3, EPA 11.1 µg/m3, Dekati 10.8 µg/m3, TEOM 11.1 µg/m3. Computed relationship between the candidate (Dekati PM10) and the reference sampler (LVS-DPM) by linear regression analysis, and between two candidate samplers met comparability requirement in large concentration range (4 – 400 µg/m3). Comparison with EN12341:1999 showed that the candidate sampler (Dekati PM10) meets the requirements for an EU-reference equivalence method (Tiitta, Nuutinen and Raunemaa, 2001). The results for PM2.5 in the city site indicated larger dispersion of data than by PM10 but the correlation was still good (Figure 2).
Traffic influenced site tests
Average PM2.5 concentration on the traffic influenced test site was
8.3 µg/m3 (range 1.3 - 17.6 µg/m3 ). In the1999
tests average PM2.5 concentration was found to decrease by 32% when moving
from 10m to over 60m distances (Tiitta et al. 2000b). The wind direction
affected the PM2.5 considerably. When downwind situations were examined,
the traffic related fine particle concentration varied from 28% to 44%
at 23m and from 8% to 40% at 59m distance. As expected, horizontal gradient
was strongest at the 1.9m height. At 4.9 m the concentration decline was
below 5% and no gradient was observed at 6.9 m.
The concentration and size distribution of fine particles below 1 µm were detected by UCPC and SMPS analyzers in a separate test series (Figure 2) at 10 m distance. With 20000 vehicles/day the daily average number concentration was 22 000 /cm3 composed of 20-30nm nucleation particles. The appearance of ultrafine particles below 100nm reveals that at close distance gas-to-particle conversion is effective and subsequent condensational growth of particles will follow.
Figure 3a)
Figure 3b)
Figure 3: Time series concentration of ultrafine particles and their size distribution in Savilahti site in summer 2000 from midnight (0:00) to 9:00 a.m. The effect of rush hour on particle development is obvious. Time series was measured by using a) ultrafine CPC and b) SMPS-system. In Figure 3a also average value and + standard deviation (solid lines) have been marked. These processes influence at least fine mode particles. The characteristics of particle structure and composition has been furthermore investigated in related works by the group (e.g. Ålander et al. 1999 and 2000, Dua et al. 1999).
Substantially high content of organics was typical by the site particles - as characterized by OC/EC ratio of 1.5 in PM2.5 particles (Tiitta et al. 2000c). The black carbon, which is representative of EC, was separately analyzed using an Aethalometer instrument. The BC concentration followed traffic density with peak values during morning rush hours. Organic particles are sensitive to environmental conditions and either evaporate or grow. Substantially small difference in PM2.5 concentrations observed at various distances from the road may accordingly reflect high portion of gasoline powered vehicles on the site. Gasoline exhaust particles can evaporate easily (Ålander et al. 1999).
4 Conclusions
As a conclusion on monitor performance fine particle PM10 and PM2.5 concentrations can be monitored down to 1 µg/m3 in 24 h sampling by adopting the three impactor devices analyzed in this work. Equal results of PM10 obtained in field measurements indicate that the PM10 cut-off by those instruments works properly. Problems in PM10 monitoring can thus be accounted for use of totally incorrect methods. In PM2.5 measurements larger variations between instrumental results were observed. PM2.5 is more difficult to monitor correctly than PM10 because smaller particles are more unstable and sensitive to conditions in sampling atmosphere outside and inside the sampler. Particles experience changes on the collection substrate or are incorrectly collected even if the penetration into the sampler interior is correct.
A single monitor site cannot produce an exact value on traffic related fine particle exposure as local wind pattern affects the measurement. A monitor at <5m height and < 25m distance from road can result in uncertain estimate of footwalk exposure to particles. When placed at those locations, traffic originated splashing effects should be avoided. In order to produce more relevant data and give also an estimate for regional concentration another similar monitor should be preferable placed at 80-200m distance from the road.
It is highly recommended that PM10 and PM2.5 are both monitored on the site as PM2.5 gives an approximation for fine mode particles, and thus alveolar deposition, and PM10 indicates the thoracic aerosol component. The usefulness of size fractioned concentration measurements for lung exposure estimation has been shown in the studies of combustion derived particles (Leskinen et al. 2000, Raunemaa et al. 2002 in these proceedings). The difference PM10-PM2.5 can be used to disclose the portion of coarse mode particles entering lungs. As fine and coarse mode particles have different sources and properties, these measurements can support the source term analysis. Specific results on the masses of different mode particles also help to design effective control strategies.
Appendix
App. 1. Schematic layout for particle modes and PM10 and PM2.5 sampling criteria (Wilson and Suh, 1997 ).
4 References
British standard (1999) EN12341:1999. Air quality- Determination of the PM10 fraction of suspended particulate matter- Reference method and field test procedure to demonstrate reference equivalence of measurement methods.
Chow, J. C., 1995. Measurement Methods to Determine Compliance with Ambient Air Quality Standards for Suspended Particles. Journal of Air and Waste Management Association 45, 320-382.
Dockery, D. W., Pope III, C. A., (1994). Acute respiration effects of particulate air pollution. Annual Review of Public Health 15, 107-132.
Dua S.K., Hopke P.K. and Raunemaa T.M.,(1999). Hygroscopicity of Diesel Aerosols. Water, Air and Soil Pollution 112, 247-257.
EUR-Lex, 1999. Council Directive 1999/30/EC of 22 April 1999 relating to limit values for sulphur dioxide, nitrogen dioxide and oxides of nitrogen, particulate matter and lead in ambient air. Official Journal L 163, 41-60.
Wilson, W. and Suh, H, (1997). Fine particles and coarse particles: concentration relationships relevant to epidemiological studies. J. Air & Waste Manage. Assoc. 47 (12), 1238
Yli-Tuomi T and Raunemaa, (1997a). T. PM10 Concentrations in Urban Sites in Finland. Journal of Aerosol Science, Vol. 28/S1, S233 - S234
Yli-Tuomi T. ja Raunemaa T. (1997b): Taajamien PM10-pitoisuudet Suomessa. Kuopion yliopiston ympäristötieteiden laitosten monistesarja, 24/1997
Yli-Tuomi T. and Raunemaa T. (1998). Comparison of PM Measurement Devices Used in Finland. J. Aerosol Sci., 29, Suppl. 1, S149.
U.S. EPA Federal Register, 1997. National Ambient Air Quality Standards
for Particulate Matter: 40 CFR Part 50, Federal Register 62:138, July 18.