Aarhus University Seal / Aarhus Universitets segl

No. 254: Mapping the health and environmental impact of air pollution in the Capital Region of Denmark

Jensen, S.S., Brandt, J., Christensen, J.H., Geels, C., Ketzel, M., Plejdrup, M. S., Nielsen, O.-K. (2018): Kortlægning af luftforureningens helbreds- og miljøeffekter i Region Hovedstaden, DCE – Nationalt Center for Miljø og Energi, 127 s. - Videnskabelig rapport fra DCE - Nationalt Center for Miljø og Energi nr. 254. http://dce2.au.dk/pub/SR254.pdf

Summary

Background and Purpose

Air pollution has significant negative effects on human health and well-being with significant socio-economic consequences, as it has negative effects on the natural environment. 

This report seeks to answer the following questions:
How is the air quality in the Capital Region of Denmark today and in the future? 
How is the air quality in relation to limit values for air quality as well as the World Health Organization's guidelines for air quality? 
What sources contribute to air quality, and how much originates from the Capital Region and outside the region? 
What are the health effects of air pollution and the associated external costs?  What are the environmental effects of air pollution?
The main objective is to identify the health and environmental effects in the Capital Region in 2014 and 2025. This is done through the following activities:

  • An air quality assessment, which describes the concentration distribution of background concentrations, as well as street concentrations and compare this with current limit values for air quality as well as the World Health Organization's guidelines for air quality
  • A source apportionment which describes the distribution of total emissions on different sources, and how they are distributed geographically. In addition, a source attribution that estimates the source contributions to the urban background concentrations, and a source attribution for 98 streets in Copenhagen
  • Estimation of the health effects and external costs related to air pollution in the Capital Region. The external costs are the social costs.
  • Description of the environmental effects of ozone, deposition of nitrogen and sulfur as well as levels of heavy metals, and compare this with critical loads and levels.

The study

Air Quality Assessment

An air quality assessment is carried out that describes the spatial distribution of background concentrations with a resolution of 1 km x 1 km, as well as street concentrations at address level in the Capital Region. This description is based on data from a national data set, which is called Air Quality at Your Street (http://luftenpaadinvej.au.dk). Furthermore, a summary of the results from the fixed measuring stations in the Capital Region is carried out and compared with EU limit values for air quality and WHO air quality guidelines.

Source apportionment

A source apportionment is carried out for the Capital Region. It includes an emission inventory which describes total emissions and the distribution of emissions on different sources, and how they are distributed geographically. 

In addition, a source attribution that estimates the source contribution to the urban background concentrations thereby providing an overview of how many micro per cubic meter the different sources contribute to urban background concentrations seen as an average over all 1 km x 1 km grid cells in the Capital Region.

Furthermore, a source attribution for 98 streets in Copenhagen is also given providing information about how much different vehicle categories contribute to street concentrations.

Health impacts and related external costs

Health effects and related external costs are calculated for the total air pollution in the Capital Region. The total air pollution includes all sources from the Capital Region, and all other sources in Denmark and abroad. This also describes how much of total air pollution originates from local sources and how much is from sources outside the Capital Region. Moreover, calculations are carried out for each type of emission source in the Capital Region to quantify the contribution of the different sources. In principle, the Capital Region is able to regulate these sources. 

The calculations are carried out with the integrated assessment model system EVA (Economic Valuation of Air pollution (Brandt et al., 2011a,b; 2013a,b), which is based on the so-called impact-pathway method. The EVA-system calculates the health impacts and related external costs based on information about the sources of pollution and their location, the dispersion of air pollution as well as exposure of the population, the dose-response relationship between exposure and health effects, and the valuation of health effects, also referred to as external costs related to health effects from air pollution. 

Modelling of air quality is based on the regional air pollution model DEHM and the urban background model UBM resulting in calculations performed on a 1 km x 1 km grid resolution. Urban background concentrations are the general air pollution in the city and reflect the concentrations in a park, a backyard or at the roof of buildings. Urban background concentrations differ from street concentrations, which represent the concentrations in the height of 2 m at the facade of buildings. Street concentrations are calculated using air quality model OSPM.

Calculations are carried out for 2014, which is the latest year for which there are updated emissions for Denmark on 1 km x 1 km resolution. Furthermore, calculations are done for 2025 based on the expected emission development. 

The EVA system includes population data with a spatial resolution of 1 km x 1 km. A new dataset based on the Central Person Registry (CPR) from 2017 has been obtained and projected to 2025 based on information from Statistics Denmark for the age groups that are part of the EVA-system.

Environmental effects

Potential environmental effects related to air pollution in the Capital Region are described indirectly: eutrophication as a result of nitrogen deposition; acidification as a result of mainly sulphur deposition; crop damage due to ozone exposure; as well as bioaccumulation and effects of heavy metals. Dep-osition and levels are compared with the critical loads and levels.

Main Findings

Air quality assessment for urban background concentrations

For urban background concentrations of NO2, the Capital Region has relatively high concentrations and in large contiguous areas as compared to distribution of urban background concentrations in other larger cities of Denmark. The highest concentrations are in Copenhagen, but throughout the Greater Copenhagen Area there are relatively high concentrations. The major transport corridors have also elevated concentrations. Road traffic is a major local source of NO2, and the regional contribution is modest. Ship traffic in Øresund also contributes. The Capital Region is located in an area with some of the highest urban background concentrations of PM2.5 in Denmark. PM2.5 is dominated by long-range transport with a clear gradient up through the country from south to north, due to sources south of Denmark, but local sources also play a role.

For urban background concentrations of PM10, the Capital Region is in a middle position compared to the rest of Denmark. PM2.5 is part of PM10, but there is also a significant contribution from sea salt from sea spray, as is evident on the north shore of the Capital Region. 

Model calculations for 2025 show a reduction in concentrations from 2014 to 2025. Urban background concentrations are expected to decrease 27% for NO2, 13% for PM2.5 and 17% for PM10 due to expected emission reductions in Europe.

Ozone concentrations are expected to increase by about 4%. Concentrations of ozone in Denmark are rising as a result of the reduction of NOx emissions in Denmark leading to less NO to consume ozone in formation of NO2.

Air quality assessment for street concentrations

The limit value of annual mean concentrations is 40 µg/m3 for NO2. Calculations performed as part of Air Quality at Your Street for 2012 may give an indication of whether or not the limit value is exceeded. Therefore, calculated exceedances are called indicative exceedances. The official announcement of exceedances of limit values is carried out as part of the annual reporting under the National Air Quality Monitoring Program, which is based on measurements from the Danish fixed monitoring stations (Ellermann et al., 2016).

As can be seen from the comparison between model results and measurements, there is some uncertainty on model results (see Table 5.1 of the report). Hence, there will also be considerable uncertainty on the number of exceedances. The vast majority of all calculated indicative exceedances for NO2 in Denmark is in the Capital Region, and these are located in Copenhagen and surroundings. There are a total of 1,066 indicative exceedances in the region. 0.2% of all 454,259 addresses in the Capital Region exceed the limit value.  

Of the 1,066 exceedances in Copenhagen and its surroundings there are 88 sites exceeding 50 µg/m3 which only occur in Copenhagen, and 6 of these exceed 60 µg/m3. The highest concentrations occur typically at very busy streets, at low travel speeds as well as in street canyons. High share of heavy-duty vehicles would also imply higher concentrations but in Air Quality at Your Street a standard vehicle distribution is assumed for all city streets. In the National Air Quality Monitoring Program air quality modelling is carried out each year for 98 selected streets in Copenhagen based on best available traffic data (traffic counts whereas Air Quality at Your Street is based on a national traffic model). Since 2012, there has been a decreasing trend in the number of exceedances of the NO2 limit value from 19 to 9 in 2015 (Ellermann et al., 2016). 

There are no calculated exceedances of limit values for PM2.5 and PM10.

Air quality assessment based on monitoring

In the national air quality monitoring program the developments in air quality is monitored at a number of permanent measurement stations. In the Capital Region, there are the following stations: two street stations: H.C. Andersens Boulevard and Jagtvej in Copenhagen, an urban background station H.C. Ørsted Institute in Copenhagen, an urban background/suburban station in Hvidovre, as well as a regional background station in Lille Valby-Risø. The Danish Environmental Protection Agency has the responsibility to ensure compliance the limit values. An air quality plan must be prepared if the limit values are exceeded, and outline how and when the exceedances are brought to an end.

There is a downward trend in NO2 concentrations for street stations, urban background stations and regional stations from 2012 onwards. In particular the downward trend is due to the ongoing replacement of the car fleet, which results in lower NOx emissions. Lower Danish and foreign emissions also contribute to the downward trend of the regional background stations. It is only the measurements at H.C. Andersens Boulevard, which exceeds the limit value of 40 µg/m3. There is also a decreasing trend for PM2.5 and PM10 concentrations, and the limit values are not exceeded. The limit value for PM2.5 is 25 µg/m3 and 40 µg/m3 for PM10 both as annual means.

The number of particles is also measured at selected stations, although there is no limit value for number of particles. When particles are counted it is proxy for ultrafine particles (PM0.1 i.e. particles with a diameter below 0.1 micrometers). There is a decreasing trend in concentrations for the street station, urban background stations, and regional station. Concentrations are approximately halved over the measurement period for street and urban background concentration from 2002 to 2015. In particular, the downward trend is due to the ongoing replacement of the car fleet with more vehicles with particle filters. The introduction of the Low Emission Zone in 2008 for heavy-duty vehicles has also contributed.

Comparison with limit values and the WHO guidelines

The EU limit values are implemented into Danish legislation. The World Health Organization (WHO) has prepared air quality guidelines. These guidelines are not legally binding. WHO guidelines are half of the EU limit values for PM2.5 (i.e. 10 µg/m3) and PM10 (i.e. 20 µg/m3) while they are the same for NO2 (40 µg/m3). 

WHO guidelines for PM2.5 are exceeded at the measurement stations in streets and touched in urban background areas but not exceeded in rural areas in 2016 in the Capital Region. By reduction of emission of PM2.5 from traffic it would be possible to comply with the WHO guidelines for PM2.5. Particulate emissions from the exhaust pipe can be reduced to technological measures, but reduction of non-exhaust (road wear, tire wear and brake wear) needs reduction in traffic.

WHO guidelines for PM10 are only exceeded in the streets in 2016. If it was possible to remove all particulate matter from traffic in the streets it would be possible to comply with the WHO guidelines for PM10. Concerning reduction of PM10 emissions the same apply as written above for PM2.5.

Reduction of NO2 will contribute to compliance with the EU limit value and WHO guidelines in the busy streets, since road traffic is also one of the emission sectors that will reduce NOx emissions in the future, and thus also reduce street concentrations.

Model calculations for 2025 indicate that the background pollution of PM2.5, PM10 and NO2 will be reduced which will further help in reaching compliance with WHO guidelines. 

Emission inventory and source apportionment

The largest source of NOx emissions is road transport, while for particles it is wood stoves and wood pellet boilers, etc. This applies for both 2014 and 2025. The findings are based on the national emission inventory which is distributed into a 1 km x 1 km grid based on various geographic variables. The total emissions are expected to be reduced respectively for NOx, PM10 and PM2.5 with 33%, 12% and 18%.

International shipping within 25 km of the Capital Region is a significant source of NOx emissions, as it is equal to about 2/3 of all NOx emissions in the Capital Region in 2014, and the same order of magnitude in 2025. However, for particles international shipping has a smaller share compared to emissions in the Capital Region in 2014 and 2025. The contribution to the urban back-ground concentration in the Capital Region is, however, not as large as may be expected due to the dominant southwest wind direction which blows the pollution away from the region, and due to the distance from the ship routes to the Capital Region. For international shipping NOx emissions are expected to increase slightly (2%), while PM10 and PM2.5 are projected to decline by 31% from 2014 to 2025.

Source attribution for urban background concentrations

The contribution of different emission sources in the Capital Region to the urban background concentration has been calculated. It shows how many mi-crograms per cubic meter the individual emission sources contribute to the average urban background concentration.

Overall, all sources in the Capital Region contributes with 37% and neighboring municipalities with 6% for NO2. That is, local sources contribute about 43%, while the other half is the regional contributions and international ship-ping within 25 km.

The two largest local emission sources in the Capital Region are road transport and wood stoves. Just taking into account the local emission sources within the Capital Region, road transport contributes by about 59% for NO2, and 19% and 17% for PM10 and PM2.5, respectively. Wood stoves contribute with about 4% for NO2, and 49% and 63% for PM10 and PM2.5, respectively. Thus, road transport contributes mainly to NO2 and wood stoves mainly to particles.

Urban background pollution of PM10 and PM2.5 are dominated by the regional concentration contribution. The regional contribution is determined by sources from Denmark and Europe. If we look at the contribution that blows into the outer border of the Capital Region it includes the modelled DEHM contribution, neighboring municipalities and international shipping within 25 km. These contributions constitute about 90% of PM10 and 91% of PM2.5 of urban background pollution in the Capital Region. On the contrary, the contribution from emission sources in the Capital Region is 10% and 9%, respectively.

Wood stoves are the largest local contributor to particulate pollution with 0.7 µg/m3 corresponding to 5% and 6% of urban background concentrations for PM10 and PM2.5, respectively.

Road transport is the second largest local contributor to particulate pollution with 0.25 µg/m3 and 0.18 µg/m3 corresponding to 1.8% and 1.6% of urban background concentration for PM10 and PM2.5, respectively. 

Calculations for 2025 indicate that the percentage contribution of road transport to urban background concentrations will decrease for NO2, and there is also a smaller reduction for PM10 and PM2.5, while the percentage al-location in 2014 and 2025 does not change much for wood stoves.

Source attribution to street concentrations

Source contribution to NO2 street concentrations has been carried out for 98 streets in Copenhagen in 2014 based on data from the National Air Quality Monitoring Program. Street concentrations include the contribution from the regional background (calculated with DEHM), a contribution from the city's emissions (calculated with UBM) and a contribution from traffic emissions in the specific streets (calculated with OSPM). The street contribution is the street concentrations minus urban background concentrations and is an indi-cator of the contribution to street concentrations arising from the traffic emis-sions in a specific street. The magnitude of the street contribution depends on traffic volume, vehicle distribution, travel speed and street geometry. The av-erage vehicle distribution for the 98 streets is 80% passenger cars, 15% vans and 5% trucks and buses. Since vehicle distribution is different from street to street, there will also be differences in the source distribution from street to street.

On average, passenger cars contribute 48% to the street contribution for NO2, vans with 20%, trucks with 15% and buses with 17%. The heavy-duty traffic (trucks and buses) thus contributes with around 33%. Despite the fact that trucks and buses make up only 5% of the traffic volume they contribute rela-tively much, since the emission factors for trucks and buses are about 10 times higher than for passenger cars and light-duty commercial vehicles.

The contribution of buses has decreased since 2014 with the retrofitting of SCRT (combined NOx catalytic converter and particulate filter) of about 300 urban buses in Copenhagen (Jensen et al., 2016). 

A detailed analysis was carried out for the street of Jagtvej in Copenhagen to assess exhaust and non-exhaust for particles. Non-exhaust includes mechanically formed particles from road wear, tire wear, brake wear and re-suspen-sion. Non-exhaust constitutes by far the largest part of the particle mass from traffic. For PM10 exhaust is about 21% and non-exhaust around 79%. For PM2.5 it is about 38% and 63%, respectively. As an example, if all the exhaust could be removed (e.g. by 100% electric cars) this would remove all exhaust, but the non-exhaust part is likely to be the same.

Premature deaths and morbidity

The total annual number of premature deaths in 2014 is approximately 1,150 in the Capital Region due to ambient air pollution levels based on both Danish and foreign emission sources. EU air quality limit values are not exceeded for particles and hence premature death occurs below the limit values. It is expected to decline by about 4% to around 1,010 in 2025, as a combination of lower PM2.5 concentrations compensating for slightly higher ozone concentrations and a larger and older population.

Premature deaths is almost exclusively due to deaths caused by long-term exposure to particulate pollution. A smaller proportion of premature deaths is due to shorter time periods with elevated concentrations (episodes) of primarily ozone.

Health effects of long-term exposure to particulate pollution accumulates throughout life from birth to death for everyone who is exposed. The long-term impact can induce cardiovascular and respiratory ailments. Therefore, premature deaths are especially among people who have been exposed for many years, e.g. elderly and people who are particularly sensitive due to prior existing diseases. Infants are also particularly sensitive, but deaths among infants represent a very small share.

The number of premature deaths is calculated based on the number of years of life lost, where one premature death corresponds to 10.6 years of life lost.

In addition to premature deaths, there are many cases of morbidity. It in-cludes chronic bronchitis and discomfort for children and adults with asthma (use of bronchodilator, cough, and respiratory symptoms), hospital admis-sions related to respiratory disorders and blood clot in the brain, cases of heart failure, lung cancer, as well as many with reduced activity (sick days).

Other diseases are also affected by air pollution, but are not included in the calculations because there is still too much uncertainty about what diseases and the precise extent of these diseases.

The contribution from foreign countries to Denmark is estimated to 2,840 premature deaths which is 76% of the total number of cases in Denmark, while the contribution from Danish emissions contributes with 890 premature deaths in Denmark (24%). 

The contribution from Danish emissions to the number of premature deaths in Europe (excl. Denmark) is estimated to about 2,280 cases/year. The “import” of air pollution related health impacts is therefore a little larger than the “export”. It is also seen that Danish emissions cause about three times the number of premature deaths in foreign countries compared to Denmark (Ellermann et al., 2017).

There will also be a contribution from indoor air pollution from indoor sources. The World Health Organization (WHO) has for high-income countries in Europe estimated 3 premature deaths per 100.000 inhabitants (WHO, 2014), e.g. about 171 for Denmark. Since the population in the Capital Region is 31.1% of Denmark about 53 premature deaths are due to indoor air pollution in the Capital Region. This is about 4% of the total premature deaths due to outdoor and indoor air pollution.

Health effects in the Capital Region distributed on local emission sources

It is examined how much the local emission sources in the Capital Region contributes to health effects in the Capital Region. The purpose of these calculations is to quantify how much the local emission sources in the Capital Region influence for health effects in the Capital Region.

122 premature deaths are attributable to emission sources in the Capital Region in 2014, and 105 in 2025. In relation to the total number of premature deaths due to all the air pollution from Danish and foreign sources, local sources in the Capital Region contribute to about 11% of all premature deaths in 2014 (122/1150) and around 10% (105/1010) in 2025. This also means that about 90% of all premature deaths in the Capital Region are caused by emissions outside the Capital Region.

The two largest local sources of premature deaths are wood burning stoves (77 in 2014 and 67 in 2025) and road transport (23 in 2014 and 18 in 2025). 

Since Danish emissions cause around the same number of premature deaths in Denmark as abroad, it is also expected that local emission sources in the Capital Region contribute with approximately the same numbers of prema-ture deaths outside the Capital Region as they do within the Capital Region.

External costs due to all air pollution

The total external costs in the Capital Region due to all air pollution from Danish and foreign emission sources are around DKK 9.5 billion by 2014, which is expected to decrease to DKK 8.2 billion in 2025. 

The external costs are almost exclusively due to particles, which here includes primarily emitted particles, secondary inorganic particles (nitrate, sulphate, ammonium), secondary formed organic particles (SOA) and sea salt. The external costs related to particulate matter are DKK 8.4 billion in 2014 and DKK 7.4 billion in 2025.

The external costs related to ozone is about DKK 0.8 billion in the Capital Region in 2014 and around DKK 0.9 billion in 2025. Ozone is not emitted directly but is formed in the atmosphere from emissions of NOx, hydrocarbons and CO. Ozone is harmful to health, and is therefore also associated with external cost. The external costs are increasing from 2014 to 2025 since ozone concentrations are increasing as a result of lower NOx emissions.

Contribution from CO to the external costs is insignificant in comparison to the other substances with around DKK 0.0031 billion in the Capital Region in 2014 and DKK 0.0029 billion in 2025. 

The major part of the external costs is due to premature deaths, since the economic valuation of premature death is relatively high in comparison to e.g. morbidity and sick days.

External costs due to local emission sources

The total external costs in the Capital Region is DKK 0.85 billion in 2014 and DKK 0.74 billion in 2025 due to local emission sources in the Capital Region. The total costs associated with health effects are due to exposure to ozone and PM2.5. The contribution of ozone is negative, as local NOx emissions reduces ozone concentrations, and the total cost is dominated by PM2.5.

The local emissions in the Capital Region contributes about 9% in 2014 and 2025 of the total external costs of all Danish and foreign sources. On the contrary, about 91% of all external costs in the Capital Region are due to emission sources outside the Capital Region.

There are also external costs outside the Capital Region associated with the local emission sources that are not included, since e.g. NOx emissions will be converted into secondary particles and cause health effects. These health effects will take place outside the model area, as the chemical transformation takes time. The model area is a square around the Capital Region and covers a minimum of 25 km from the region's border.

The main local emission sources in the Capital Region that contribute to the external costs in the Capital Region are wood stoves, accounting for approximately 66% in 2014 and 65% in 2025 followed by road traffic accounting for 15% in 2014 and 2025.

Environmental effects

The average nitrogen deposition is about 10 kg N/ha in 2014, which exceeds or is close to the critical load for certain sensitive natural habitats. Calculations of deposition in 2025 show that the lower range for the critical loads for the most sensitive habitats like lobelia lakes and raised bogs are still exceeded. But for the remaining part of the §3-natural areas calculations indicate that if emissions follow the projections, critical loads are not exceeded in 2025.

In Denmark sulphur depositionen has decreased by approx. 70% since 1989 and the level in the Capital Region in 2014 and 2025 are under the critical loads for typical natural habitats.

Calculations for 2014 and 2025 of ozone (AOT40 value) show that levels are below the target value for the protection of vegetation.

Measurements of a number of heavy metals are routinely carried out in the national air quality monitoring program. Measured levels of heavy metals are below the limit value or target values.

Uncertainties

The EVA-system is based on the impact pathway approach covering emissions from sources, the dispersion and chemical transformation in the atmosphere, exposure of the population, health effects, and economic valuation of these health effects. There are uncertainties associated to all these elements further discussed in Chapter 10. 

Important assumptions are that all particles are assumed to be equally dangerous with the mass of particulate matter (PM2.5) as indicator for health effects. 

No independent health effects of NO2 are assumed. Dose-response relationships are being implemented into the EVA-system and is expected to lead to higher but not much higher health impacts. 

The calculation method can underestimate the importance of local sources such as road traffic, which affects how much this source attributes to the health effects and external costs. 

A Swedish study has attempted to quantify the health effects in Sweden with a breakdown of the contribution from regional sources and from local sources, where local sources are road traffic and wood burning (Gustafsson et al., 2014). With the assumptions that local sources have higher relative risk than regional sources they get about 50% more premature deaths (5,300 against 3,500). With the same assumptions, one would probably get something similar for Denmark, but there is a lack of further documentation of how local sources attribute to higher relative risk before we can include this with sufficient certainty in the EVA system.

Assumptions about the economic valuation of a premature death also have significant impact on the estimation of the total external costs. The Ministry of Finance has in August 2017 announced new values for a statistical life, which is significantly higher than the values that are used in the present report. The new values are being implemented in the EVA system, and are expected to lead to about 50% higher external costs than the calculations described in the present report.