Aarhus University Seal

No. 156: The Particle Project 2011-2013

Nøjgaard, J. K., Massling, A., Christensen, J. H., Nordstrøm, C., & Ellermann, T. 2014. The Particle Project 2011-2013. Aarhus University, DCE – Danish Centre for Environment and Energy, 51 pp. Scientific Report from DCE – Danish Centre for Environment and Energy No. 156. http://dce2.au.dk/pub/SR156.pdf


In addition to the measurements included in the Danish Air Quality Monitoring Programme, detailed physical and chemical measurements of atmospheric particles have been conducted in a series ofresearch projects on atmospheric particles since 2001. The objectives of this fourth Particle Project was to i) continue the record of particle mass and number measurements, ii) evaluate the sources to urban PM2.5, and iii) to provide scenario calculations of PM2.5 in 2020, 2025 and 2030 to evaluate if Denmark will be able to meet future reduction targets.

PM10 and PM2.5 were measured with TEOM (Tapered Element Oscillating Microbalance) at the different stations (rural background, urban background and urban curbside) in and around Copenhagen between 2002 and 2013. For TEOM-PM2.5 the data show a slow decreasing trend over the years. In general, regional and long-range transported aerosols dominate the PM2.5 mass at the background stations, while the urban curbside is also impacted by exhaust and non-exhaust emissions from traffic. The decreasing trend in PM2.5 reflects the general situation in Europe, where both precursors to particle formation and primary particle emissions are decreasing, which lowers ambient mass levels at the background sites. A decreasing trend in TEOM-PM10 has as well been observed at the urban background, while the data series at the rural background is too short to make any conclusions.

Particle number size distributions were measured with DMPS (Differential Mobility Particle Sizer) at the different stations (rural background, urban background and urban curbside) in and around Copenhagen during the time period 2002-2013. Particle numbers show a slight decreasing trend for the rural background and urban background stations. This decrease is much more pronounced at the urban curbside, which is highly influenced by traffic exhaust emissions. This is also indicated by the small particle mean diameter compared to the other stations. The number of particles in the lowest size regime (diameter between 6 and 40 nm) is about four times higher at the urban curbside compared with the urban background. In the middle size regime (diameter between 40 and 110 nm) the values are still twice as high at the urban curbside as at the urban background, which is showing the high traffic contribution in this size class. In general, the decreasing trend in particle number is in accordance with improvements in engine exhaust technologies and past changes in fuel composition, which lead to lower particulate emissions.

Sources to urban PM2.5 were evaluated based on campaigns during the winter 2011 and summer 2012 by use of the receptor model COPREM. Particle samplers were placed in an urban street and urban background from which a traffic source profile was derived by subtraction of the measured species’ concentrations at the two sites. Apportionment of the carbonaceous fraction was accomplished by use of source specific chemical markers, e.g. radiocarbon (14C) to track anthropogenic sources of fossil carbon (combustion of gasoline, diesel, oil and coal). Secondary Organic Aerosols (SOA) was tracked by specific markers for biogenic volatile organic compounds (BVOC), which are oxidized to form aerosol particles in the atmosphere. SOA is also formed by oxidation of anthropogenic volatile organic compounds (AVOC), which are composed of fossil carbon and thus included in the sources Ship emissions, Coal or Oil. One exception is VOCs released during residential wood burning. A part of this is oxidized in the atmosphere to form SOA, which cannot be separated from BSOA. Biomass Burning (BB), of which the largest source is believed to be residential wood burning in Denmark and neighbour countries, is traced by use of four organic tracers in addition to inorganic potassium.

The applied tracers were used to successfully track their specific sources. Concentrations of elemental carbon (EC, “soot”) was highest during the winter campaign, where the activity of residential wood combustion is highest, but with a clear signal all year due to traffic emissions and other combustion processes. SOA tracers were highest during the summer campaign, though products from coniferous trees were also measured during the winter campaign. Also, most tracers for biological aerosol particles (PBAP) particles such as decomposition of leaves and other plant materials were only detected during the summer campaign. Markers for residential wood combustion were only vaguely present during the summer campaign.

PM2.5, EC and OC differed largely between the two campaigns and sites, which was particular evident for the winter 2011/summer 2012 concentrations of PM2.5. However, increased PM concentrations were measured at all Danish sites in 2011. Absolute winter concentrations of PM2.5, EC and OC were all higher than summer concentrations. The fraction of fossil carbon, on the other hand, showed only 4% variation with season. In contrast, a large difference was observed between urban background (30%) and urban street (50%).

In the winter campaign, Polycyclic Aromatic Hydrocarbons (PAH) were measured in the urban background and urban street in the PM2.5 and PM10 fractions during the winter campaign. On average, two thirds of benzo[a]pyrene were found in the PM2.5 fraction. Furthermore, urban curbside concentrations of benzo[a]pyrene were generally higher than urban background concentrations, evident of traffic being a source of urban PAH.

Source apportionment of urban background PM2.5 identified Secondary Inorganic Aerosols (SIA) to be the most abundant aerosol fraction averaging 4.3 µg/m3 over the two campaigns (36%). SOA accounted for 1.2 µg/m3, which was less than expected and could be a consequence of the relatively few applied tracers for SOA, which do not encompass all SOA sources and thus possibly underestimates this source. Primary emissions included marine emissions “sea salt” (1.0 µg/m3; 8%) and the minor crustal and primary biological Aerosol Particles. Ship emissions accounted for 0.62 µg/m3 (5%). Other fossil sources were Coal (0.81 µg/m3; 7%), Oil (0.03 µg/m3; 0.3%) and Vehicular Traffic (0.26 µg/m3; 2%).

Sources of particles from BB are believed to be mainly residential wood burning. However, contributions other than national sources are likely also to play a role. According to the source apportionment analysis BB accounts for 0.88 µg/m3 of PM2.5 averaged over the winter and summer campaigns. However, the source contribution is not evenly distributed over the year, for which reason the mitigation potential is higher during the heating season. On average, this source contributed by 1.5 µg/m3 of PM2.5 during the winter campaign, but only 0.3 µg/m3 during the summer campaign.

Dominating sources to EC are BB (36%) and Traffic (44%). Ship emissions, Oil and Coal, are responsible for 12%, 1% and 7% of EC, respectively.

Organic carbon mass was attributed to SOA (44%), BB (22%) and Traffic emissions (9%). Furthermore, Ship emissions accounted for another 22% of organic carbon. Minor contributions were attributed to biological sources (PBAP) such as microorganisms and decomposition of leaves. These findings agree with other European studies of organic carbon in PM2.5.

The major uncertainty related to the present source apportionment analysis is a result of the fact that annually averaged source contributions are estimated from only 8 weeks of measurements campaigns, representing only 15% of the year. Due to the high cost of chemical measurements, however, it is very expensive to carry out longer campaigns using conventional laboratory analysis of chemical markers. A more promising approach is Aerosol Mass Spectrometry (AMS), from which continuous measurement series of organic and inorganic compounds can serve as input for source detailed apportionment analysis. AMS instruments have been purchased (with funding from the Villum Foundation) and installed temporarily at Risø in 2014 - 2015.

A national exposure reduction target of 2.1 µg/m3 was derived for Denmark based on an average exposure indicator of 14 µg/m3 for the reference year 2010. This reduction target is based on the measurements in urban background in Copenhagen, Aarhus and Aalborg.

The model calculated reduction in PM2.5 from 2010 to 2020 will most probably be slightly too low, due to limitations in the model applied. First, a likely decrease in SOA cannot be accounted for. Second. a reduction in particle-bound water cannot be addressed by the model. However, a simple estimate based on the results from the measurements has been used to take the reductions of water into account. This leads to a slightly larger reduction in PM2.5.

When particulate bound water is taken into account, the Gothenburg 2020 scenario will lead to a reduction of PM2.5 of 1.6 µg/m3. The EC baseline scenarios for 2025 and 2030 will lead to a reduction of 2.0 µg/m3 and the EC scenario 2025A will lead to a reduction of 2.9 µg/m3. Therefore, it is only EC scenario 2025A, which will lead to a reduction exceeding the national emission reduction target for Denmark of 2.1 µg/m3. The EC baseline scenarios for 2025 and 2030 leads to a reduction just below the target, while the Gothenburg scenario is well below the target.

The proposal from the EU Commission for a new directive on national emission ceilings suggest more stringent emission commitments from the EU member states for 2025 and 2030 than the base line emissions scenarios applied in this project (Amann, 2013). The emission scenarios are estimated to lead to reductions in PM2.5 of 1.6, 2.3 and 2.9 µg/m3 for 2020, 2025 and 2030, respectively. Hence, it is estimated that the national exposure reduction target can be met in 2025 and may be somewhere between scenario 2020 and scenario 2025 since the calculated reductions in PM2.5 do not account for reductions in the secondary organic particles.