Future Air Quality in Danish cities 6. Air Quality at Street Level6.1 Validation of OSPM Predictions The OSPM model is used to predict air pollution levels in the street of Jagtvej in Copenhagen. The hourly time-series produced by the Urban Background Model is used as input to the OSPM model together with COPERT III based emission factors, and parameters on traffic in the street, street configuration and meteorology. 6.1 Validation of OSPM PredictionsTable 6.1 shows that the OSPM model underestimates the observed levels for NOx, NO2 and CO, especially for CO. Ozone levels are overestimated as a consequence of the underestimation of NOx. Benzene levels are well predicted since benzene emission factors are determined by invert calculations with the OSPM. Table 6.1 [Look
here] Figure 6.1 [Look here] Figure 6.1 shows a general good agreement between modelled and observed concentrations although predicted levels are systematically underestimated for NOx, NO2 and CO possible due to underestimation of emission factors. 6.2 Possible Underestimation of COPERT III emissionsUnderestimation by COPERT III As discussed in the previous chapter, COPERT III emission factors may underestimate real world emissions on the road since better results were obtained with the Urban Background Model using emission factors that were about a factor 2 higher for CO for passenger cars without catalysts and about at factor 2 higher for NOx for lorries. In Figure 6.2, the possibility of underestimation is further investigated by comparison of COPERT III emission factors (new emissions) and the formerly used emission factors (old emissions) as input for OSPM calculations. The figure shows the ratio between CO and NOx for modelled and measured values in Jagtvej, Copenhagen for working days, Saturdays and Sundays. If the ratio between vehicle emissions of CO and NOx is correct then the slope of the regression lines of modelled air quality levels will be identically to the measured concentrations in the street air. It is seen that the slope of modelled air quality levels using COPERT III emission factors is very different from the measured ratio between CO and NOx in the street air. Much better results are obtained with the old emission factors. This indicates that the ratio between COPERT III emission factors for CO and NOx is incorrect since it does not comply with the ratio found in the measured street air. The emission factors in the OSPM model are adjusted according to travel speed during working days based on emission factors at 50 km/h. The same method has been applied for both new and old emission factors. For Saturdays, the travel speed is assumed to be 50 km/h, and the old emissions give almost a perfect fit in this situation between modelled and measured ratios of CO and NOx indicating that the ratio between CO and NOx is correct for the old emission factors and questionable for COPERT III, see Figure 6.2. Nevertheless, COPERT III emission factors have been applied throughout the study although predicted air quality levels become underestimated. For prediction of future concentrations in the urban background or in the street, observed levels have been applied from 1995 as a baseline for calibration, and the modelled trend as an index has been used to estimate future levels to give realistic predicted air quality levels that can be compared to air quality limit values. Figure 6.2 [Look here] 6.3 Future Air Quality at Street LevelTable 6.2 sums up the OSPM model runs for the difference scenario years for the future air quality at street level in Copenhagen. Table 6.2 [Look here] The table gives the predicted development in annual levels, and 98- and 99.8-percentiles in ppb/ppm, µg/m3/mg/m3 and as an index with the reference year equals 100. Future predicted levels are also given for Jagtvej in Copenhagen using observed levels from 1995 as a base and the development represented by the index. The calibration is required to give realistic future air quality predictions that can be compared to limit values because too low air quality levels are predicted using COPERT III emission factors without calibration. NO Becomes Limiting Factor in Forming NO2 Catalyst cars were introduced in Denmark in 1990/91 and reduce NOx emissions (NO and NO2). NO2 observed levels in Jagtvej were more or less constant during 1990-95 indicating that ozone was the limiting factor in forming NO2 in reactions between NO and ozone. From 1995 to 1998, measurements show a downward trend in NO2 levels, and this trend is also reproduced by the OSPM model. During 1995-2010/2020, 98- and 99.8-percentiles of NO2 are predicted to decrease about 50% and 35%, respectively. The predictions show that NO becomes the limiting factor in forming NO2 in reactions with ozone in the future due to the steadily decreasing NOx emissions (NO and NO2, NO constitutes about 95% of NOx vehicle emissions). Ozone Ozone levels increase because less NO emitted from vehicles in the street is available for ozone depletion. CO and Benzene CO levels are predicted to decrease by a factor of 4 and benzene levels by a factor of 10 from 1995 to 2010. The predicted downward trends of CO and benzene are also support by observed levels during 1995-1998. 6.4 Comparison With Air Quality GuidelinesAir Quality Guidelines A summary of present EU air quality limit values, WHO guidelines and Danish EPA criteria for the modelled pollutants is presented in Table 6.3. The Danish EPA air quality criteria were set up to minimize of adverse health effects. The air quality criteria are not administrative limit values but should be regarded as desired long-term objectives (Larsen et al. 1997). New EU limit values have to be met in 2010. A margin of tolerance has been defined to secure that limit values will be met in 2010. The margin of tolerance given as a percentage in the table refers to the year the directive entries into force. The margin of tolerance is equally stepped down each year to reach 0% in 2010. Member states have to take local action if the margin of tolerance is exceeded. Table 6.3
1 Not to be exceeded on more than 20 days per calendar year averaged over three years NO2 The EU limit value for NO2 for long-term exposure was exceeded in 1995 and the limit value for short-term exposure is tangent. However, the margin of tolerance of 50% in 1999 is not exceeded. The predicted NO2 levels in 2010 at Jagtvej are about half of the EU limit value in 2010. The Danish EPA criteria for short-term and long-term exposure is exceeded for all scenario years until 2015-2020. CO The EU limit value for CO will be between the 98- and 99.8-percentile. The EU limit value for CO was not exceeded in 1995, and the margin of tolerance of 50% will not be exceeded in the expected year of entry into force of the directive (2000). In 2010 the predicted CO levels will be 10-20% of the EU limit value in 2010. The EU limit value and WHO guidelines are identically for CO. The Danish EPA has not suggested criteria for CO. Benzene The EU limit value for benzene was exceeded in 1995. The margin of tolerance of 100% will not be exceeded based on modelled levels in 2000, the expected year of entry into force of the proposed directive. The predicted levels in 2010 will be about half of the EU limit value. WHO guidelines and Danish EPA criteria are exceeded for all scenario years. Ozone at Street Level The average ozone levels in the street will increase due to a decrease in NO vehicle emissions in the street leaving less NO for depletion of ozone in forming NO2, see Table 6.2. However, the sum of NO2 and O3 will decrease. The highest levels calculated as a 8 hour running maximum will slightly decrease over the years because the highest ozone levels in the regional background are predicted to decrease. The proposed EU limit value for ozone is 120 m g/m3 as a 8 hour running maximum not to be exceeded on more than 20 days per calendar year averaged over three years. This short-term limit value was not exceeded in 1995 nor is it predicted to be exceeded in 2010 and the following years despite an increase in average ozone levels in the street. Ozone in Urban Background In Table 6.4 exceedances of the ozone threshold of 120 m g/m3 are given for the urban background. The urban background is a better indicator for ozone exposure of the population than levels in the streets since ozone levels are influenced by NO emissions. Since the number of exceedances are less than 20, the EU limit value is not violated in the urban background. The number of exceedances of the threshold value of 120 m g/m3 increases over the years. This is due to the general increase in ozone levels in the urban background that will cause more peak values to exceed the 120 m g/m3 threshold. However, the model overestimates ozone levels as was seen in the previous Chapter 5, Table 5.5, and the presented exceedances in Table 6.4 are based on modelled ozone data that have not been adjusted to the observed level in 1995. Furthermore, since several modelled values are close to the threshold value 120 m g/m3 and the model overestimates ozone levels, it is likely that there will be few exceedances of this threshold in future observed ozone levels in the urban background. Table 6.4
6.5 Future Air Quality in 103 Copenhagen StreetsOSPM Calculation for 103 Copenhagen Streets Based on OSPM calculations for 103 different streets in the Copenhagen Area, an empirical relation between traffic density and street air quality for NO2 and benzene was established in 2000 and 2010, see figure 6.3-5. The streets represent a wide range of traffic loads and street configurations however with a little less detailed information about traffic and street configuration data compared to data available for Jagtvej. Traffic density is here defined as average daily traffic divided by the width of the street. The modelled emission reductions and predicted urban background levels in 2000 and 2010 by the present study has been applied. This relation can be applied for crude assessment of the air quality in a street just knowing the traffic density as defined. Since urban background data for Copenhagen was used, the street levels will be overestimated in other Danish cities where the urban background concentrations are lower. Since the relation was established for urban streets in built-up areas, air quality levels will be overestimated if applied for rural roads where dispersion characteristics are different. It is seen that annual levels of NO2 and benzene in 2000 are exceeding the limit value for 2010. Figure 6.3 [Look here] Figure 6.4 [Look here] Figure 6.5 [Look here] Air Quality Levels in 2010 NO2 and benzene in 2010 The predicted development in future air quality levels in 2010 for NO2 and benzene for the 103 Copenhagen streets is given in Figure 6.6-8. Similar to the scenario 2010, it is assumed that traffic loads are constant in the streets considered while an increase on main roads of 17% is assumed corresponding to a general traffic increase in the road network considered of 10% 1995-2010. Air quality levels in 2010 are predicted to decrease for NO2 and benzene, and none of the considered Copenhagen streets will violate the limit values of NO2 and benzene for 2010. Figure 6.6 [Look here] Figure 6.7 [Look here] Figure 6.8 [Look here] 6.6 Preliminary Assessment of Particulate Air Pollution at Street LevelIntroduction In this section, a preliminary assessment of the particle levels in selected streets in Denmark is carried out and levels are related to the new EU limit values for PM10. The impacts of future particle emission reductions are also briefly discussed. The assessment is based on measurements since air quality models for particles are not fully developed. Health Effects It is recognised that particles in urban air are responsible for serious health effects, i.e. long-term effects like cancer, and cadio-vascular decease and acute effects like allergy or irritation of eyes, nose and throat (Larsen et al. 1997). Particles are often characterised by the mass determined as PM10 or PM2.5, particulate matter less than 10 m m and 2.5 m m, respectively. New EU Limit Values The regulation from the Danish Ministry of Environment no. 836 dated 10.12.1986 on air quality includes limit values for TSP (Total Suspended Particles), i.e. 300 µg/m3 as 24 hour average and 150 µg/m3 as annual average. A new EU directive "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" gives limit values for particulate matter (PM10). The Member States have to comply with the 24 hour limit value 50 µg/m3 - not to be exceeded more than 35 times per year and 7 times per year - before 2005 and 2010, respectively. For annual averages the limit values are 40 µg/m3 and 20 µg/m3 for 2005 and 2010, respectively. WHO has not recommended a limit value for PM, because knowledge is missing and no lower observed effect level has been identified. Consequently, the EU Commission has also realised that our knowledge about adverse health effect and the sources and chemical/physical characteristics of particles is too limited; therefore it has been decided to revise the limit values for particles within a few years when more information is available. The directive also includes obligations for the Member States to collect data on smaller particles PM2.5. However, investigations have shown that the correlation between particle concentration and health effect increases with decreasing particle diameter. It is therefore important to determine the concentration given as number of particles in many size intervals. Characteristics of Particles The particle size distribution is an important factor that needs to be addressed whenever the PM pollution is concerned. A major contribution to particulate pollution in urban areas is believed to be from traffic, especially diesel powered vehicles. Particles emitted from car engines, petrol as well as diesel engines, are formed at high temperatures in the engine, in the exhaust pipe or immediately after emission to the atmosphere. These particles are in the so-called nucleation mode and the diameter of the particles is < 0.2 m m, ultrafine particles. Other particle modes are accumulation mode (fine particles), > 0.2 m m - 2 m m, which typically are formed by chemical reactions of (e.g. SO2 and NOx to form sulphate and nitrate), coagulation, condensation of gases on particles or other relatively slow processes. The last mode is the coarse particles > 2 m m, which typically are formed mechanically by traffic turbulence, wind erosion etc. These larger particles may also cause health effect. The size distribution and the main characteristics of urban particles are shown schematically in Figure 6.. Figure 6.9 [Look here] Trends and Levels of TSP and PM Pollution in Denmark The total suspended particulate matter (TSP) is determined in the National Air Quality Monitoring Programme (LMP) (Kemp and Palmgren, 1999) by weighing of the aerosol filters. The samplers collect particles up to an aerodynamical diameter of around 25 µm, but this cut-off varies from about 10 to 50 µm depending on the wind speed (Kemp 1993). The particles are a mixture from the different source types, but the coarse particles (> 2.5 µm) of windblown dust of local origin are expected to dominate. The fine particle fraction includes contributions of long range transported soil dust and particles from combustion processes, e.g. sulphate and nitrate particles. TSP was measured in 1998 as 24 hour average values at street stations in the major Danish cities: Copenhagen, Odense and Aalborg and at the regional background station of Lille Valby about 40 km outside Copenhagen. The measurements at Lille Valby started in the beginning of 1995. Statistics from 1998 are shown in Table 6.5 (Kemp and Palmgren, 1999). The old limit values were not exceeded. Table 6.5.
Trends The trends of TSP are shown in Figure 6.10. The general trend has been a decrease of about 30-50% during 1988-1998 for the street stations. A major part of the mass of the particles (coarse particles in Figure 6.) is windblown dust and may be considered to be either of "natural" origin, constructions or re-suspended particles from the roads. The particles from combustion processes are in the fine particle fraction, and it is expected to decrease in the future due to emission reductions. The observed trend in TSP may be a result of i.e. better cleaning of emissions from power plants, obligatory three way catalysts (TWC) on petrol cars, restrictions on the diesel exhaust, and more green agricultural fields during winter (less soil dust). TSP has been measured at the rural station Lille Valby for almost 4 years. The levels are between on third to half of the levels at the urban street stations. Figure 6.10 [Look
here] PM10 Continuous measurement of PM10 was started in July 1998. Sampling in 24 hour intervals is performed using an OPSIS SM200 sampler at Jagtvej, Copenhagen. The particles are collected on membrane filters (Millipore type AA). The PM10 is determined both on-line with the build-in b -gauge and gravimetric, using the same procedure, as for TSP. TSP is approx. 35% higher than PM10, that is, PM10 constitutes about 74% of TSP. Figure 6.11 [Look
here] In Table 6.6, the PM10 level in selected streets in Denmark have been estimated based on the above relation between TSP and PM10. Tabel 6.6
1 PM10 equals 74% of TSP It is seen that the estimated PM10 levels in 1998-99 are below the new limit value for 2005 but exceed the limit value for 2010. Denmark has a national objective to reduce particle vehicle emission by 50% in urban areas 1988-2010, and further reductions after 2010. The increase in penetration of catalyst converters reduce particle emissions for petrol powered vehicles due to unleaded petrol. Catalysts become mandatory in 1990. New stringent particulate emission standards for especially diesel powered vehicle will reduce particle emissions. The conversion to diesel with a low content of sulphur will also reduce particulate emissions. Previous assessments indicate that the total particulate emissions (as mass) from vehicle within the EU will decrease by about 70% 1995-2010 including expected increases in traffic (Iversen 1999). Based on a few number of European studies, WHO has estimated that the particulate emission from vehicles in urban areas contributes about 40-60% of PM10 (WHO 1999). Due to the above mentioned vehicle particulate emissions regulation it is likely that the PM10 will decrease in the future but it is difficult to predict how much based on existing knowledge and to predict if the limit value of 2010 will be met. The above figures indicate that it might be a problem. Fine and Ultrafine Particles The fine and especially the ultrafine particles emitted directly from the diesel and petrol fuelled vehicles contribute only a little to the particle mass TSP and PM10. It is therefore necessary to use other measurement techniques to measure these particles. In addition, we have some indication that the number of ultrafine particles, which can penetrate into the deepest parts of the lungs, is important to assess the health risk of particulate air pollution. A precise determination of the emission of particles from the actual car fleet is necessary for analysis of the problem in urban areas, investigations of the health impacts, and recommendations of abatement measures to be taken to reduce the pollution. Particle Size Distribution In order to characterise the particle pollution emitted directly from car engines, a method to measure the ultrafine particle mode has been developed. The method uses a Differential Mobility Analyser, DMA. This method is based on particle size fraction separation by the particles’ mobility, determined by movement of charged particles in electrical fields. The DMA measures with a high time resolution which is necessary for identification of traffic air pollution in order to separate this source from other types of air pollution. Measurements have been carried in busy streets in Copenhagen (Jagtvej) and Odense (Albanigade) comprising long time-series of particle spectra in connection with the normal monitoring of air pollutants, i.e. NOx/NO2, CO, benzene, O3 and SO2. In this way it was possible to determine the contribution from local traffic in the street by subtraction of the urban background concentration from the concentration measured in the street, and by inverse model calculation by the street pollution model OSPM (Berkowicz et al. 1997). The method has been used on stable pollutants like NOx, CO and benzene (Palmgren et al. 1999). Preliminary investigations have shown that the ultrafine particles do not change size significantly during the residence time in the street, i.e. less than a few minutes (Vignati et al., 1999). The DMA method gives the size distribution in the range 0.01 – 0.7 m m. However, the distribution is not determined simultaneously, but by sweeping over the size range during a few minutes. The DMA was also applied for laboratory studies of the emissions from vehicles. Examples of results are shown in Figure 6.. It is seen that particles from diesel powered vehicles are a little smaller than particles from petrol powered vehicles. Analysis shows that diesel vehicles on average emit about 25 times as many particles as petrol vehicles. The contribution from diesel and petrol vehicles was almost the same at Jagtvej because of few diesel vehicles. The contribution of ultrafine particle from diesel vehicles at Albanigade in Odense, which is a more typical city street, was much higher than from petrol vehicles (Palmgren and Wåhlin, 1999). Figure 6.12 [Look here] The particle number spectrum at Albanigade in Odense during rush hour. [Look here] The number distribution can be translated to volume distribution (or mass distribution assuming mass density 1), see Figure 6.13. It is seen that the relatively few larger particles (Figure 6.12) contributes significantly to the mass. Traffic contributes about ¾ of the mass of ultrafine particles (PM0.2). In this case, it is also seen that the petrol powered vehicle mass contribution is comparable with the non-traffic contribution for ultrafine particles. Figure 6.13 [Look here] Further investigations needed The knowledge about the air pollution with particulate matter is still rather limited. By the new PM10/PM2.5 methods and the application of DMA for measurement of ultrafine particles from traffic, possibilities have opened to obtain valuable data. Systematic measurements, including long time-series, by these methods at representative sites will improve the possibilities for health studies substantially. However, more knowledge is needed about the chemical/physical properties of the particle, e.g. chemical composition, surface properties and morphology; for this purpose it is necessary to include other analytical techniques, e.g. SEM and micro probe analysis. The characterisation of the particles is also important for quantification of the contribution from different sources and parameterisation of the properties of the particles to be included in air quality models. This is necessary for decisions on abatement measures to be taken to reduce the health impacts of particulate air pollution and to evaluate the effects of the measures taken.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||