Baseline Projections of Greenhouse Gases and Air Pollutants in the European Union up to 2030

Transcription

1 ` European Consortium for the Modelling of Air Pollution and Climate Strategies EC4MACS Baseline Projections of Greenhouse Gases and Air Pollutants in the European Union up to 2030 EC4MACS Final Assessment Editor: Markus Amann International Institute for Applied Systems Analysis IIASA January 2013

2 The authors This report was compiled by authors of the partner institutions that participate in the EC4MACS project: Markus Amann, IIASA Jens Borken Kleefeld, IIASA Hannes Böttcher, IIASA Janusz Cofala, IIASA Jean Paul Hettelingh, RIVM/CCE Chris Heyes, IIASA Mike Holland, EMRC Alistair Hunt, Metroeconomica Zbigniew Klimont, IIASA Leonidas Mantzos, NTUA Leonidas Ntziachristos, LAT/AUTh Michael Obersteiner, IIASA Max Posch, RIVM/CCE Uwe Schneider, Uni Hamburg Wolfgang Schöpp, IIASA Anne Wagner, Ricardo AEA Peter Witzke, UniBonn Wilfried Winiwarter, IIASA Acknowledgements This report was produced by the EC4MACS (European Consortium for the Modelling of Air pollution and Climate Strategies) project with financial contributions of the LIFE financial instrument of the European Community. Disclaimer The views and opinions expressed in this paper do not necessarily represent the positions of IIASA or its collaborating and supporting organizations. This report documents the application of the EC4MACS model toolbox for the assessment of future air pollution and climate strategies of the European Union as of early Thus, the assessment presented in this report does not incorporate projections of economic activities and energy use that were being developed late 2012, and does not reflect new national data that has been provided to the EC4MACS team by national experts in the course of the bilateral consultations held in Thus, all quantitative results presented in this report need to be considered as provisional. In particular, after completion of this final report of the EC4MACS project, the EC4MACS toolbox has been applied for policy analyses to support the review of the Thematic Strategy on Air Pollution and the revision of the EU air legislation in Results of these policy analyses, which incorporate later information on economic and energy scenarios, as well as improved information provided by national teams, have been presented to the European Commission and stakeholders in a series of reports, and are available at MitigationofAirPollutionandGreenhousegases/TSAP review.en.html. More information on the Internet More information about the EC4MACS methodology and interactive access to models and their results is available at the Internet at

3 ` Executive Summary During the last decades, Europe has successfully eliminated the most visible and immediately harmful effects of air pollution. However, there is ample and robust scientific evidence that even at present rates Europe s emissions to the atmosphere pose a significant threat to human health, ecosystems and the global climate, though in a less visible and immediate way. Refined scientific methods reveal that, e.g., via the long term exposure to fine particulate matter, current levels of air pollution shorten statistical life expectancy of the European citizens by several months. Biodiversity and Europe s genetic resource base is under threat from the excessive release of nitrogen to the atmosphere from energy combustion and intensive agriculture. Europe s greenhouse gas emissions, currently twice as high on a per capita basis as the world average, and historically responsible for about a quarter of current concentrations in the atmosphere, make a significant contribution to global climate change. This report presents an outlook into the likely development of emissions of greenhouse gases and air pollutants and their impacts up to 2030 as can be envisaged from current expectations on economic development and the implementation of existing legislation on air pollution controls in the European Union. This report adopts as a central assumption projections of economic development developed in the year 2009, and considers national and EU wide energy, climate, agricultural and air pollution policies that have been implemented by spring The report does not include the targets on renewable energy sources and on greenhouse gas emissions from the non ETS sector that were agreed in EU's Climate and Energy Package early The analysis starts from a recent projection of population development, which suggests a 6% increase in the population of the EU 27 due to continued immigration. By 2020 total GDP in the EU 27 is assumed to be 30% higher than in 2005, and 50% in Energy and climate policies will show distinct effects on future energy consumption in Europe, and decouple the levels of economic activity (GDP) from energy consumption. Total energy consumption is assumed to remain at the 2005 level up to Although renewable energy will increase its market share to some extent, no major changes in the composition of fuel use are projected up to 2030 despite the assumed 50% increase in GDP. Increases in car ownership in the new Member States will be compensated by saturation effects in the old Member States. Further growth, however, will occur for freight transport, although the 33% increase to 2030 is lower than the assumed growth in GDP. For the agricultural sector, the baseline suggests a decline in the numbers of cattle and sheep and increases in pigs and chicken. These changes in human activity levels, together with dedicated policies to reduce emissions of greenhouse gases and air pollutants, will have distinct impacts on future pollution of the atmosphere in Europe. Most notably, the baseline projection suggests a certain decline of greenhouse gas emissions, reaching 8% in 2020 and 16% in 2030 relative to Larger reductions (between 33% and 66%) are expected for emissions of air pollutants (i.e., SO 2, NO x, PM2.5, VOC), due to the on going structural changes in the economy and the effects of new emission control legislation. These changes in baseline emissions will have distinct impacts on air pollution impacts on human health, forests, vegetation, freshwater, crops and materials. Health impacts from exposure to fine particulate matter, which is associated for the year 2000 with a shortening of statistical life expectancy of 10 months in the EU, would decline by about 50% up to The number of premature deaths that are attributable to exposure to ground level ozone (33,000 in 2000) would decline by one third in 2020 and by more than 50% in Similar positive impacts are computed for vegetation and ecosystems. For instance, ecosystems area where biodiversity is threatened by excess nitrogen deposition will shrink from 1.1 million km 2 in 2000 to below 900,000 km 2 in 2030, and acidification will remain an issue at less than 0.5 percent of the European forest area. However, despite these significant improvements, the anticipated baseline development of emissions to the atmosphere will not be sufficient to achieve sustainable environmental conditions that safeguard human health and ecosystems services. Per capita greenhouse gas emissions will still be at 9.2 tons CO 2 eq/person/yr in 2020 and at 8.3 tons/person/yr in 2030, which is significantly higher than the approximately two tons that would be available in a budget approach that allocates equal GHG emissions i

4 to all people in the world while limiting temperature increase to 2 degrees Celsius. Also for air pollution, despite the impressive reductions in precursor emissions, fine particulates in ambient air will still cause life shortening of almost five months to the European population. It is estimated that the European population would still suffer a loss of 200 million life years and experience 19,000 premature deaths because of ozone exposure. Biodiversity will remain threatened by excess nitrogen input at 900,000 km 2 of ecosystems, including 340,000 km 2 which are legally protected, inter alia as Natura2000 areas. It is unlikely that the baseline development will achieve full compliance with the air quality limit values for PM10 and NO 2 throughout Europe. In general, despite the envisaged improvements, the European Union will fail to reach the environmental targets that have been established in the EU Thematic Strategy on Air Pollution for 2020 for human health and biodiversity. There is scope for additional measures that could alleviate the remaining damage and move closer to the objectives of the Sixth Environment Action Program. Full application of readily available technical emission reduction measures in the EU could reduce health impacts from PM by 2020 by another 30% and thereby gain more than 55 million life years in the EU. It could save another 3,500 premature deaths per year because of lower ozone concentrations. Further controls of agricultural emissions could protect biodiversity at another 270,000 km 2 of ecosystems against excess nitrogen deposition, including 120,000 km 2 of Natura2000 areas and other protected zones. It could eliminate almost all likely exceedances of PM10 air quality limit values in the old Member States; in the urban areas of new Member States additional action to substitute solid fuels in the household sector with cleaner forms of energy would be required. Such Europewide emission controls would also eliminate in 2030 all likely cases of non compliance with EU air quality standards for NO 2 with the exception of a few stations for which additional local measures (e.g., traffic restrictions, low emission zones) would be necessary. Based on well established techniques, health benefits, i.e., reduced mortality and morbidity from less exposure to fine particulate matter and ground level ozone, can be expressed in monetary terms. Based on a most conservative estimates, the health benefits from the additional measures would account to 25 billion /yr in 2020; upper estimates range up to 157 billion /yr. Obviously, the additional measures that would achieve these benefits come at a certain costs, which are estimated at billion /yr. These direct emission control costs constitute on average about % of the GDP, however there are large variations across Member States, mainly due to differences in economic wealth. These investments into air pollution control equipment replace productive investments and do not produce economic utility to the consumers. Depending on the sector, these investments consume between 0.2% of the produced value added (in ferrous and non ferrous metals industries) and 4% (for electricity production). However, at the same time production of pollution control equipment will indirectly induce higher activity in other sectors. Overall, it is estimated that the emission controls of the maximum feasible reduction case would lead to % lower GDP in 2030 compared to the baseline projection, which assumes for 2030 about 50% higher GDP compared to The slightly lower GDP increase is mainly a consequence of a slower growth in household consumption as productive resources are diverted for environmental investments. However, models do not suggest net impacts on employment, although trends in different economic sectors can vary. It should be mentioned that the Maximum Technically Feasible Reduction scenario analyzed in this report portrays a hypothetical extreme case in which all technical measures that are currently available were applied to their full extent. In reality, well chosen sub sets of measures will achieve large portions of the feasible benefits at a small fraction of costs, so that such measures emerge as highly cost effective means for further improvements of air quality in Europe. The EC4MACS model system offers a practical toolbox for exploring cost effective portfolios that yield high returns of health and environmental benefits. ii

5 ` Table of contents Introduction... 2 The EC4MACS project... 3 The EC4MACS toolbox... 4 The EC4MACS consortium... 4 Drivers of future pollution... 8 Economic growth... 8 Energy prices... 9 Lifestyle changes Sectoral economic development Energy demand Modelling future energy demand Baseline energy consumption up to Transport activities Modelling future transport activities Baseline trends in the transport sector Land use and agricultural activities Modelling land use and agricultural activities Baseline trends in land use Baseline trends in agricultural activities National trends in agricultural activities Emissions of the baseline scenario Modelling future emissions of greenhouse gases and air pollutants Emission control policies considered in the baseline Carbon dioxide (CO 2 ) emissions CO 2 emissions from energy use CO 2 emissions from the land use (LULUCF) sector Methane (CH 4 ) emissions Nitrous oxides (N 2 O) emissions F gas emissions Total GHG emissions Sulphur dioxide (SO 2 ) emissions Nitrogen oxides (NO x ) emissions Fine particulate matter (PM 2.5 ) emissions Volatile organic compounds (VOC) emissions Ammonia (NH 3 ) emissions iii

6 The scope for further emission reductions The mitigation potential for SO The mitigation potential for NO x The mitigation potential for PM The mitigation potential for NH The mitigation potential for VOC Emission control costs Costs for implementation of the current legislation Costs for further emission control measures Macro economic impacts Air pollution control expenditures as shares of sectoral value added Impacts on economic growth Changes in production and employment Health impacts The EC4MACS health impact assessment methodology Premature mortality from exposure to fine particles Morbidity impacts from the exposure to fine particles Health effects from exposure to ground level ozone Vegetation and ecosystems impacts Methodology and key issues Threat to biodiversity Acidification Vegetation damage from ozone Protection of Natura2000 areas Excessive nitrogen deposition in Natura2000 areas Risks for biodiversity of all ecosystems Acidification of forest soils Acidification of surface waters Risks from ground level ozone to vegetation Risks from ground level ozone to forests Risks from ground level ozone agricultural crops Compliance with air quality limit values Compliance with PM10 air quality limit values Compliance with NO 2 limit values Monetary valuation of benefits Methodology and key issues iv

7 ` Human health benefits Quantification of health impacts Valuation of health impacts Health benefits from the emission control scenarios Non health impacts Agricultural crops Material damage Un monetised non health benefits Cost benefit analysis Comparison of costs and benefits Uncertainties Discussion Conclusions v

8 List of acronyms CAFE Clean Air For Europe Programme of the European Commission CAPRI Agricultural model developed by the University of Bonn CH 4 Methane CLRTAP Convention on Long range Transboundary Air Pollution CO 2 Carbon dioxide EC4MACS European Consortium for Modelling Air Pollution and Climate Strategies EMEP European Monitoring and Evaluation Programme ETS Emission Trading System of the European Union for CO 2 emissions EU European Union F gases Fluorinated gases GAINS Greenhouse gas Air pollution Interactions and Synergies model GW Gigawatt = 10 9 watts IIASA International Institute for Applied Systems Analysis kt Kiloton = 10 3 tons LULUCF Land use, land use change and forestry Mt Megaton = 10 6 tons NEC National Emission Ceilings NH 3 Ammonia NO x Nitrogen oxides NO 2 Nitrogen dioxide N 2 O Nitrous oxides O 3 PJ Ozone Petajoule = joule PM10 Fine particles with an aerodynamic diameter of less than 10 µm PM2.5 Fine particles with an aerodynamic diameter of less than 2.5 µm PRIMES Energy Systems Model of the National Technical University of Athens SNAP Selected Nomenclature for Air Pollutants; Sector aggregation used in the CORINAIR emission inventory system SO 2 Sulphur dioxide TSAP Thematic Strategy on Air Pollution UNFCCC United Nations Framework Convention on Climate Change VOC Volatile organic compounds vi

9 ` Introduction 1

10 Introduction Europe has successfully tackled the most visible effects of air pollution During the last decades Europe has successfully eliminated the most visible and immediate harmful effects of air pollution. Urban smog episodes like in London in the 1950s are now history, the air in industrial regions has been cleaned up since the 1960s, Scandinavian lakes are recovering from acidification in the 1970s, the large scale forest die back of the 1980s in central Europe has come to an end, and episodes of high ozone concentrations are less frequent than in the 1990s. However, there is ample and robust scientific evidence that even at present rates Europe s emissions to the atmosphere pose a significant threat to human health, ecosystems and the global climate, though in a less visible and immediate way. Refined scientific methods reveal that, e.g., via the long term exposure to fine particulate matter, current levels of air pollution shorten statistical life expectancy of the European citizens by several months. Biodiversity and Europe s genetic resource base is under threat from the excessive release of nitrogen to the atmosphere from energy combustion and intensive agriculture. Europe s greenhouse gas emissions, currently twice as high on a per capita basis as the world average, and historically responsible for about a quarter of current concentrations in the atmosphere, make a significant contribution to global climate change. However, solving the less visible problems constitutes serious policy challenges. To protect the livelihood of European citizens, the sustainability of the services provided of its ecosystems and to avoid dangerous interference with the global climate system, additional efforts are required to control the release of harmful substances to the atmosphere. In principle, a host of measures is available to further reduce emissions in the future. However, as many of the low hanging fruits have been harvested by now, further action will put higher demand on economic resources, especially at a time when resources are strained by the economic crisis. In addition, we gain increasing insight into the interactions and interdependencies of the various measures that could lead even to counterproductive outcomes of strategies if they are ignored. However, if put in context, informed decision making could develop strategies that maximize the synergies between different measures to safeguard environmental improvements for all relevant aspects while minimizing economic resources for their implementation. An outlook into the likely development of future emissions and impacts This report presents an outlook into the likely development of emissions and resulting impacts up to 2030 as can be envisaged from the current expectations on economic development and the implementation of existing legislation on air pollution controls in the European Union. It is compiled at the mid term of the EC4MACS project cycle, and provides a first holistic perspective into the next 20 years that emerges from the new integration of the participating EC4MACS models. On that basis the report provides information for subsequent considerations of enhanced policy response strategies to protect living conditions for humans and ecosystems from atmospheric pollution at the local, regional and global scale. Based on recent expectations on future economic growth The future pressure on the environment from humans and cost effective response strategies will be critically influenced by the types and quantities of economic activities. The recent economic crisis has clearly demonstrated how difficult it is to accurately predict economic development, and that any prediction is associated with deep uncertainties. However, this report adopts the most recent post economic crisis projections as a central assumption, as they incorporate the economic downturn that occurred in 2008 and Structure of the report Section 1 reviews the methodology of this analysis to support future policy decisions on climate and air pollution strategies in the European Union. Section 2 summarizes assumptions on the future development of key drivers of emissions and on policies that will influence these drivers. Section 3 presents estimates of baseline emissions of greenhouse gases and air pollutants for the Member States and economic sectors. The potential for further emission reductions is discussed in Section 4, and costs and macro economic impacts of such measures are presented in Sections 5 and 6. Sections 7 9 discuss air quality impacts on human health and vegetation, as well as compliance with air quality limit values. The monetary benefits of additional measures are presented in Section 10, and conclusions are drawn in Section

11 ` The EC4MACS project There are critical linkages, interactions and feedbacks between social, economic and environmental systems The impact of human activity on the Earth s atmosphere is critically determined by numerous linkages, interactions and feedbacks between the social, economic and environmental systems. Much, although not all, is understood about individual aspects that are involved in local air pollution and global climate change. Only an integrative perspective that brings together the relevant aspects can provide informative and accurate knowledge on the current state and the likely future development, and the scope for measures to reduce negative impacts of human activities on the atmosphere. EC4MACS provides a scientific toolbox to provide a holistic perspective on emission control strategies The European Consortium for Modelling of Air pollution and Climate Strategies (EC4MACS), a project funded under the EU LIFE programme, develops a network of well established modelling tools that enable a comprehensive integrated assessment of the policy effectiveness of emission control strategies for air pollutants and greenhouse gases (Box 1). Box 1: Objectives of the EC4MACS project Clean air and climate change are central fields of EU environmental policy. New scientific findings demonstrate important interactions and potentially large economic synergies between air pollution control and greenhouse gas mitigation. Model analyses, based on latest scientific findings and validated data, can provide valuable information for cost effective policy strategies. The key objectives of EC4MACS are: Provide scientific and economic analyses for the revision of the EU Thematic Strategy on Air Pollution and the European Climate Change Programme (ECCP) Improve existing models by including recent scientific findings Update of input data Achieve acceptance of modelling tools and input data by stakeholders Make modelling tools available to the public over the Internet EC4MACS simulates the cause effect chain to identify cost effective policy responses The EC4MACS approach assumes cause effect relationships between interacting components of social, economic, and environmental systems. These include driving forces of environmental change (e.g., industrial production), pressures on the environment (e.g., discharges of pollutants to the atmosphere), state of the environment (e.g., air quality in different regions in Europe), impacts on population, economy, ecosystems (e.g., reduced life expectancy from the exposure to air pollution), and the response of society (e.g., in form of emission control policies). This sequence of cause effect relationships is often called the DPSIR (Drivers Pressure State Impact Response) concept (Figure 1). Energy/agricultural projections Emission control options Emissions Costs Atmospheric dispersion Air pollution impacts, Basket of GHG emissions OPTIMIZATION Environmental targets Figure 1: The cause effect chain for atmospheric pollution considered in EC4MACS The EC4MACS model toolbox allows simulation of the impacts of policy actions that influence future driving forces (e.g., energy consumption, transport demand, agricultural activities), and of dedicated measures to reduce the release of emissions to the atmosphere, along their impacts on total emissions, resulting air quality, and a basket of air quality and climate impact indicators. Furthermore, through the GAINS optimization tool, the framework allows the development of cost effective response strategies that would meet environmental policy targets at least costs. 3

12 Cost effective response strategies need to consider cause effect relationships for multiple pollutants and multiple effects. Emissions to the atmosphere cause a variety of negative impacts on climate and air quality. In addition, pollution does not comprise a single chemical substance, but consists of a cocktail of many pollutants originating from a wide range of human activities and natural sources that can be controlled to different extents at different costs. Thus, cost effective response strategies need to consider cause effect relationships for multiple pollutants and multiple effects, and how they are interconnected with each other (Table 1). Table 1: The multi pollutant/multi effect frame work of EC4MACS PM (BC, SO 2 NO x VOC NH 3 CO CO 2 CH 4 N 2 O OC) Health impacts: PM (Loss in life expectancy) HFCs PFCs SF 6 O 3 (Premature mortality) Vegetation damage: O 3 (AOT40/fluxes) Acidification (Excess of critical loads) Eutrophication (Excess of critical loads) Climate impacts: Long-term (GWP100) ( ) ( ) ( ) ( ) ( ) ( ) Near-term forcing ( ) ( ) ( ) Carbon deposition to the Arctic and glaciers The EC4MACS toolbox The GAINS model, as the core integrated assessment tool of EC4MACS, represents the cause effect chains for health impacts, vegetation damage and climate change, taking into account the sources and control potentials of five air pollutants and six greenhouse gases. In particular, GAINS describes the simultaneous effects of specific control measures on the emissions these air pollutants and greenhouse gases, and the physical and chemical interactions of these emissions in the atmosphere. EC4MACS brings together a suite of wellestablished sectoral scientific models for practical policy analyses However, the GAINS model does not cover the full range of relevant driving forces that cause pollution, nor does it represent the full range of environmental and economic impacts of pollution. As these aspects cannot be ignored in the design of cost effective response strategies, the EC4MACS project has linked existing computer modelling tools to enable a holistic and coherent assessment of policy response options (Box 2, Figure 2). Global/ hemispheric boundary conditions European policy drivers Costeffectiveness Impacts Transport TREMOVE, COPERT Energy POLES PRIMES GAINS GEM-E3 Agriculture CAPRI Land use EU-FASOM, DNDC Ecosystems Atmosphere TM5 CCE-IMPACTS EMEP/CHIMERE ALPHA-2 Figure 2: The EC4MACS model suite that describes the full range of driving forces and impacts at the local, European and global scale. The EC4MACS consortium The EC4MACS consortium brings together renowned institutions that have ample experience with analyses for air pollution and climate change policies of the European Union (Box 3). The consortium is coordinated by the International Institute for Applied Systems Analysis (IIASA). 4

13 ` Box 2: Models participating in the EC4MACS framework The GAINS integrated assessment model explores cost effective multi pollutant emission control strategies that meet targets to protect human health, ecosystems and on total greenhouse gas emissions. The PRIMES energy model simulates the response of energy consumers and the energy supply systems to different pathways of economic development and exogenous constraints. The TREMOVE transport model simulates current and future transport demand, fleet structures and emissions. The CAPRI agricultural model describes the development of the agricultural sector in response to a wide range of policy actions. The EU FASOM model explores welfare maximizing total land use strategies, including greenhouses gas emission control and carbon sink strategies that meet wider environmental objectives on inter alia soil, water and biodiversity protection. The DNDC model calculates associated greenhouse gas emissions. The EMEP and CHIMERE atmospheric dispersion models address chemical processes and the transport of pollutants in the atmosphere at the regional and local scales. The CCE IMPACTS assessment module develops impact assessment methods and indicators for air pollution effects to different types of vegetation and ecosystems and collects necessary input data from all countries. The ALPHA2 model quantifies the health and environmental benefits of emission control strategies in monetary terms. The GEM E3 economic model explores the macro economic impacts of emission control strategies for all Member States of the EU by simulating the interactions between the economy, the energy system and the environment. Box 3: Partners in the EC4MACS project The International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria IIASA ( an international, non governmental research institute, has developed the GAINS integrated assessment models for air pollution and greenhouse gases. It applied the model, inter alia, for the scientific support of the CAFE programme of the European Commission ( ), for the development of the EU National Emission Ceilings Directive ( ), the EU Acidification Strategy (1997), and various protocols of the UNECE Convention on Long range Transboundary Air Pollution. Under this Convention IIASA hosts the Centre for Integrated Assessment Modelling (CIAM) of the European Monitoring and Evaluation Programme (EMEP). The Coordination Center for Effects (CCE), Bilthoven, Netherlands The CCE ( located at the Netherlands Institute for Public Health and the Environment (RIVM), collaborates as the Programme Centre of the International Cooperative Programme on Modelling and Mapping, with 26 National Focal Centres to provide operational methodologies to support impact assessment of air pollution policies. Results of CCE's work supported effect based UNECE protocols (1994, 1999), protocol review ( ) and EC policy processes including the Acidification Strategy (1997), the NEC Directive (1999) and the CAFE programme. The E3M Lab of the National Technical University (NTUA), Athens, Greece NTUA ( has developed and applied several energy and economic models for all EU member states. These include the medium term econometric engineering MIDAS model, the PRIMES energy model, and the GEM E3 general equilibrium macroeconomic model. Model applications have been carried out for the European Commission and for public authorities in many countries. The Laboratory of Applied Thermodynamics in the Aristotle University of Thessaloniki (LAT/AUTh), Thessaloniki, Greece LAT/AUTh (lat.eng.auth.gr) specializes in emissions control technology for mobile sources for impact assessment studies, emission standard development, monitoring policies, and inventorying. The institute is responsible for the development of the COPERT tool on behalf of EEA and UNECE, the TRENDS tool on behalf of DG TREN and EUROSTAT, and it is a member of the European Topic Centre on Air and Climate Change. In the area of transport scenarios, it has taken part in the Auto Oil II and the CAFE programmes. The Institute for Agricultural Policy, Market research and Economic Sociology (IAP) of the University of Bonn, and EuroCARE GmbH, Bonn, Germany IAP ( bonn.de/agpo/rsrch/capri/capri_e.htm) and EuroCARE ( bonn.de) have developed the agricultural sector model CAPRI, which is used to analyse the impact of policy changes on European agriculture. CAPRI is continuously improved under a number of research projects carried out by a network of European researchers, with funds from the European Commission and EEA. The CAPRI baseline served as input into GAINS in the context of CAFE. EMRC, AEA Technology, MetroEconomica, UK The benefit assessment team, coordinated by EMRC, a specialist in impact assessment and cost benefit assessment, includes AEA Technology, one of the UK's largest environmental consultancies, and MetroEconomica, a specialist in environmental economics. All three have participated extensively in Community funded work, for example the cost benefit analysis of the CAFE programme. 5

14 6

15 ` Drivers of pollution 7

16 Drivers of future pollution Types, volumes and locations of human activities are significant drivers of atmospheric pollution. While all these aspects are dynamically changing over time, the future development path is not unique and depends on many factors, which are genuinely uncertain. EC4MACS employs a suite of economic, energy and agricultural models to develop a coherent quantification of future human activity patterns based on recent macroeconomic forecasts of renowned institutions. Economic growth The EC4MACS macroeconomic scenario comprises numerical projections of population, GDP (volume), households income, and sectoral activity (using gross value added in volume as a proxy) for 22 sectors, in each EU Member State. After the economic crisis, the baseline scenario assumes GDP to grow by 50% until 2030 After the economic downturn in 2008/2009, the baseline scenario assumes sustained economic growth along the 2009 Ageing Report (European Economy, April 2009) and the intermediate Scenario 2 "Sluggish Recovery" presented by the European Commission in the Europe 2020 strategy (Box 4). Compared to precrisis projections, the economic recovery will compensate for some of the loss in GDP during the recent recession, but will not reach the earlier growth projections (Figure 4). All economic sectors will be affected. Continued immigration will let European population grow by six percent up to 2030, mainly in the old Member States The EC4MACS baseline scenario employs the population projections of the EUROPOP2008 convergence scenario from EUROSTAT ((Giannakouris 2010)). Thereby, by 2030 continued immigration will let population increase by 6% relative to 2005, mainly in the old Member States. This will also enlarge labour force in the EU accordingly. However, population in the new Member States is expected to decline (Figure 3). 160% 140% 120% Relative to % 80% 60% 40% 20% 0% Austria Belgium Bulgaria Cyprus Czech Rep. Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU Figure 3: Population trends assumed for the Member States Figure 4: Assumed GDP growth and comparison to a prerecession projection For instance, for 2020 an eight percent lower GDP is assumed now compared to the pre crisis projection. Assumptions on economic growth after the recovery remain unchanged, i.e., at 2.2% per year between 2016 and 2020, 1.8% per year between 2020 and 2025, and 1.65% per year between 2025 and

17 ` Box 4: Key rationales of the macro economic scenario The economic prospects for the EU are divided in three periods: the recession ( ), the recovery ( ) and the low but stable growth period (beyond 2022). The financial crisis induced a marked deterioration of global economic prospects in the final quarter of The causes of a vicious recession spiral were the loss of financial assets, the reduction in business confidence accompanied with increased uncertainty, and the resulting reduction in bank lending (credit freeze). The credit rationing practice synchronized worldwide had a detrimental effect for emerging economies through reduction in global trade (credit facilitation to trade was dramatically decreased). Thus, exports of the EU were seriously affected downwards. Credit rationing together with increased uncertainty resulted in a slowdown of private investment in all sectors and lowered households expenditures in durable goods and new houses. The rate of private savings increased, exerting further depressive effects on consumption. Altogether, drop of exports, lower consumption and investment explain recession in GDP terms for the EU economies. To alleviate the effects of the crisis, the governments put in place extraordinary measures, including reduction of basic interest rates, expansion of money supply and facilitation of credit availability. These measures removed the effects of credit rationing and reduced the shadow interest rate, and so encouraged investments and spending in durable goods and houses. The low levels of oil and commodity prices facilitate economic growth as costs of domestically produced goods fall. As these measures expand worldwide, global trade is facilitated again by increased credit availability. Thus, demand is progressively re established in the EU, concerning both exports and domestic consumption and investment. These trends characterize the recovery period, which may last until Although current differences between old and new Member States will decline over time, full convergence is not assumed until For southern countries, similar growth patterns are assumed, however with somewhat lower long term prospects than those of the new Member States. Although current differences between old and new Member States will decline over time, full convergence is not assumed until Energy prices The baseline scenario assumes a relatively high oil price environment compared with previous projections. Price trajectories for gas and coal are derived from the PROMETHEUS model based on conventional wisdom perspective of the development of the world energy system. World fossil fuel prices will remain high International fuel prices are projected to grow over the projection period with oil prices reaching 88 $/bbl (73 /bbl; in 2008 $/ ) in 2020 and 106 $/bbl (91 /bbl) in 2030 (Figure 5). Gas prices follow a trajectory similar to oil prices reaching 62 $/boe (51 /boe) in 2020 and 77 $/boe (66 /boe) in 2030, while coal prices increase during the economic recovery period and reach almost 26 $/boe (21 /boe) in 2020; afterwards they would stabilize around 29 $/boe (25 /boe). The recovery process is accompanied by efficiency and productivity gains in many sectors. As a result, growth prospects of the EU are in percentage terms somehow larger than before the crisis, albeit for a limited time period. Based on this logic, the projection displays higher growth rates for the period compared to a similar projection carried out before the crisis. Despite this, a permanent loss of GDP and welfare is encountered when considering the entire period from 2008 to Growth patterns differ across the EU. Old Member States in northern and central Europe suffer more from the recession than the others. They will recover more slowly, but stay on a significant and positive growth path over the long term. While the new Member States have undergone an important depression too, it is assumed that their recovery will be more pronounced than the EU average. Slower growth rates are assumed in the longer run as these countries approach the performance of the old Member States. Figure 5: Assumed evolution of energy prices (in constant US $ 2008) 9

18 Lifestyle changes Per capita income will grow by 2%/year Following the demographic and macro economic assumptions, per capita income will increase at an average rate of about 2% per year. However, differences in per capita income, notably between the old and the new Member States, will remain even in the long term, although the gap is slowly closing (Figure 6). consumption in Europe until In contrast, demand for milk products would decline (Figure 8). Meat consumption will increase further, but demand for milk products will fall 160% 140% 120% Relative to % 80% 60% Figure 6: Assumed development of per capita GDP The anticipated increase in personal wealth will have profound impacts on personal lifestyles, and thereby on activities that cause pollution to the atmosphere. Car use will slowly saturate, but new EU Member States will catch up fast In the past, travelled mileage and car use have steadily increased with growing income. This trend is expected to slow down in the coming decade in the old Member States. However, the anticipated economic convergence of the new Member States will let travel demand grow fast during the entire projection period (Figure 7). GDP/capita (1000 Euro/year) GDP: Old Member States GDP: New Member States Mileage: Old Member States Mileage: New Member States Figure 7: Assumed evolution of passenger travel demand in the old and new Member States (solid lines: GDP/capita; dashed lines: travel demand) As a consequence of higher incomes, dietary habits are expected to change. Together with increased population, this should lead to a 25% higher meat Annual mileage (1000 km/year) 40% 20% 0% Figure 8: Assumed development of the demand for animal products (butter, cheese and meat) Demand for cereals will increase by 25% Similar to the demand for animal products, cereal consumption in the EU 27 is expected to increase by about 25% until 2030 following the increased population and enhanced lifestyles (Figure 9). Relative to % 140% 120% 100% 80% 60% 40% 20% 0% Butter Cheese Meat Population GDP/capita Cereals consumption Population GDP/capita Figure 9: Assumed development of cereal consumption Sectoral economic development As a result of modified consumption patterns and the economic restructuring process, the sectoral composition of future economic activities is likely to change in the future. The GEM E 3 model (Box 5) has been used to develop an internally coherent picture of sectoral economic activities in the 27 Member States 10

19 ` that is consistent with the overall macro economic assumptions discussed before. Box 5: The GEM E3 model The GEM E3 model explores the macro economic impacts of emission control strategies for all Member States of the EU by simulating the interactions between the economy, the energy system and the environment. It is an applied general equilibrium model, simultaneously representing World regions and European countries, linked through endogenous bilateral trade flows and environmental flows. GEM E3 includes all simultaneously interrelated markets. It formulates separately the supply and demand behaviour of the economic agents, which are considered to optimise individually their objective while market derived prices guarantee global equilibrium. The model considers explicitly the market clearing mechanism and the related price formation in the energy, environment and economy markets: prices are computed by the model as a result of supply and demand interactions in the markets and different market clearing mechanisms, in addition to perfect competition, are allowed. The model formulates production technologies in an endogenous manner allowing for price driven derivation of all intermediate consumption and services from capital and labour. In the electricity sector, the choice of production factors can be based on the explicit modelling of technologies. For the demand side the model formulates consumer behaviour and distinguishes between durable (equipment) and consumable goods and services. The model is dynamic, recursive over time, driven by accumulation of capital and equipment. Technology progress is explicitly represented in the production function, either exogenous or endogenous, depending on R&D expenditure by private and public sector and taking into account spill over effects. 0.7% per year (Figure 11). Furthermore, restructuring will continue, shifting the production mix towards higher value added product varieties, such as special steel, special ceramics and high quality glass. Particularly high growth is assumed for the pharmaceuticals and organic chemical industry, while a slowdown is expected for the building materials industry. Iron and steel industry is projected to remain active in the EU, benefiting from the restructuring towards higher use of scrap material and the production of higher quality end products as a result of technology progress. Production of energy intensive goods will remain in the EU million Euro Services Earth and metals Construction Engineering Food, textiles, paper Other Figure 10: Development of GDP, EU 27 The GEM E3 model estimates that the service sectors, which generated in 2005 about 72% of the EU s gross value added, will increase their share to approximately 75% by 2030 (Figure 10). Non energy intensive industries are expected to maintain their current share in total value added of about 13.5%. Within this group, engineering industries producing equipment goods and pharmaceutical and cosmetics industries will grow faster than the average. The service sector and non energy intensive industries will grow faster than others Energy intensive industries (i.e., chemicals, basic metals, building materials, pulp and paper) represent a small share in total value added (3.4% in 2005). The baseline scenario assumes that the bulk of industrial activity in this sector will remain in the EU territory and grow by GDP relative to % 160% 140% 120% 100% 80% 60% 40% 20% 0% Iron and steel Chemicals Paper and pulp Textiles and leather Other Services Energy sector Non ferrous metals Non metallic minerals Food, drink, tobacco Engineering Construction Agriculture Total Figure 11: Sectoral trends in gross value added, relative to

20 Energy demand Modelling future energy demand The PRIMES energy model has been used to simulate the future energy market The PRIMES model (Box 6) has been used to quantify implications of the economic development in the various sectors on energy demand and supply in the EU 27. Based on assumptions on economic development, world market energy prices and energy policy interventions, the PRIMES model simulates the behaviour of different economic actors of the energy system and how they influence the energy market. The EC4MACS baseline considers the EU Energy and Climate Package The EC4MACS baseline scenario employs the PRIMES 2010 baseline energy projection, which considers energy policies that have been instituted by spring 2009, including the targets on renewable energy sources (RES) and on greenhouse gas emissions from the non ETS sector that are defined in the Climate and Energy policy package (Box 7). Box 6: The PRIMES model of the energy sector The PRIMES model, developed by the National Technical University Athens (Greece), simulates the response of energy consumers and the energy supply systems of the Member States of the European Union to different pathways of economic development and exogenous constraints. The model determines the equilibrium by finding the prices of each energy form such that they quantity which producers find best to supply matches the quantity consumers wish to use. The equilibrium is static (within each time period), but repeated in a time forward path under dynamic relationships. PRIMES represents energy demand and supply as well as pollution abatement technologies in great detail, and simulates the market clearing under different economic, industrial and energy policies and environmental regulations. In addition to EU wide policies, the scenario includes national policies aiming at higher energy efficiency and the support of renewable energy sources. Box 7: Energy policy measures assumed in the baseline and how they are reflected in the PRIMES model Measure How the measure is reflected in PRIMES Regulatory measures on energy efficiency Eco design implementing measures 1 Eco design Framework Directive 2005/32/EC Adaptation of modelling parameters for different product groups. As 2 Stand by regulations 2008/1275/EC requirements concern only new products, the effect will be gradual 3 Simple Set to boxes regulation 2009/107/EC (marginal in 2010; rather small in 2015 and up to full effect by 2030). The 4 Offices/street lighting regulation 2009/245/EC potential envisaged in the Eco design supporting studies and the 5 Household lighting regulation 2009/244/EC relationship between cost and efficiency improvements in the model s 6 External power supplies regulation 2009/278/EC Other energy efficiency regulations database were cross checked. 7 Labelling Directive 2003/66/EC Enhancing the price mechanism mirrored in the model 8 Cogeneration Directive 2004/8/EC National measures supporting cogeneration are reflected 9 Directive 2006/32/EC on end use energy efficiency National implementation measures are reflected and energy services 10 Buildings Directive 2002/91/EC National measures e.g. on strengthening of building codes and integration of RES are reflected 11 Energy Star Program (voluntary labelling) Enhancing the price mechanism mirrored in the model Regulator measures for energy markets and power generation 12 Completion of the internal energy market (including provisions of the 3 rd package) The model reflects the full implementation of the Second Internal market Package by 2010 and Third Internal Market Package by It simulates liberalised market regime for electricity and gas (decrease of mark ups of power generation operators; third party access; regulated tariffs for infrastructure use; producers and suppliers are considered as separate companies) with optimal use of interconnectors. 12

21 ` Baseline energy consumption up to 2030 With the assumptions on economic development, international fuel prices and energy policies that are described above, the PRIMES model estimates for 2010 total energy consumption in the EU 27 about 3% below the 2005 level. By 2015, total energy consumption would then return to the 2005 level, and shrink to the 2010 level thereafter despite the 50% increase in GDP. Despite a 50% increase in economic activity, baseline scenario indicates a stabilization of energy The decoupling between GDP growth and primary energy consumption (Figure 12) in the baseline scenario emerges as a direct consequence of the economic restructuring towards less energy intensive sectors, autonomous technological progress and dedicated energy policies that promote energy efficiency improvements. Relative to % 140% 120% 100% 80% 60% 40% 20% 0% GDP Population Total energy consumption Energy intensity of GDP Figure 12: Baseline trends in energy consumption, its key drivers (GDP and population) and energy intensities between 2000 and 2030 Energy from renewable sources is expected to triple by 2030, while coal consumption would decline by 17% While the total volume of energy consumption is suggested to remain at today s level, the structural composition of fuels and energy sources is anticipated to change (Figure 13, Table 3). Most importantly, current policies for renewable energy sources are expected to increase biomass use by two thirds in 2030 compared to 2005, and to triple energy from other renewable sources (e.g., wind, solar). In contrast, coal consumption is expected to decline by 18% by 2030, and oil consumption is calculated to be 13% lower than in Fuel consumption (1000 Petajoules/year) Figure 13: Baseline energy consumption by fuel in the EU 27 After economic recovery, energy efficiency policies will reduce energy demand in households and industry Different trends are expected for different economic sectors. Until economic recovery will be achieved, energy demand in the transport sector is expected to increase by 9% up to 2020 (relative to 2005), and by 3% for households and industry. After that time, progressive implementation of energy efficiency measures will show full effect, especially in the domestic and transport sectors where lower energy consumption is calculated for 2030 than for 2020 (Figure 14,Table 3). In contrast, fuel input to the power sector will decline between 2005 and 2010 and remain at that level for the remaining simulation period. Fuel consumption (1000 Petajoule/year) Other renewables Biomass Nuclear Gas Oil Coal GDP Non energy Transport Industry Households Power sector Conversion GDP Figure 14: Baseline energy consumption by sector in the EU GDP (trillion /yr) GDP (trillion /yr) 13

22 Table 2: Baseline energy consumption by sector in the EU 27 (1000 PJ) Conversion Power sector Households Industry Transport Non energy Total Table 3: Baseline energy consumption by fuel in the EU 27 (1000 PJ) Coal Oil Gas Nuclear Biomass Other renewables Total Although energy intensities of Member States are anticipated to converge, significant differences will prevail in 2030 The PRIMES model provides country specific projections of baseline energy consumption for all EU Member States (Table 4). As an overall feature of the EC4MACS baseline projection, energy intensity improvements of GDP will occur in all countries over time, and the large discrepancies between countries that prevail in 2005 are gradually converging. It is noteworthy that per capita consumption is expected to increase in some of the new Member States. However, even for 2030 significant differences in energy intensities are expected to remain. On a per capita basis, energy consumption levels are more similar among the EU 27 countries (Figure 15). Per capita energy use (Terajoue/person/year) AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU 27 Figure 15: Per capita energy consumption by Member State, 2005, 2020,

23 ` Table 4: Baseline energy consumption by country (Petajoules) Austria Belgium Bulgaria Cyprus Czech Rep Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU Energy intensity (Terajoue/Million /year) AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU 27 Figure 16: Energy intensities of GDP in the Member States, 2005, 2020 and

24 Transport activities Mobile sources make substantial contributions to emissions of atmospheric pollutants. While the stringent technological standards that have been introduced in the past have substantially lowered emissions from individual vehicles, their positive effect was counteracted by a general increase in the volume of transport activities. Thus, emissions in the future will be determined by the development of the overall vehicle stock, its fuel consumption, and its structural composition, as the possibilities for reducing emissions depend on vehicle types and fuel used. Modelling future transport activities The TREMOVE transport estimates future fleet composition and emissions To put all these aspects into perspective, in EC4MACS the PRIMES projection of overall trends in transport activities and fuel consumption is complemented by analyses with the TREMOVE, COPERT and FLEETS models. These models simulate the future composition of vehicle fleets in different countries and estimate emissions (Box 8). Box 8: Modelling of vehicle stock and emissions from mobile sources The TREMOVE transport model (v3.3) estimates future stock replacement rates of different vehicle categories (i.e., scrappage and new registrations) and resulting emissions, based on activity data delivered by PRIMES, detailed data on the current vehicle stock for EU27 members collected in the FLEETS project, and the latest COPERT 4 (v7.1) emission factors. The fleet turnover is modelled with TREMOVE. The stock data are based on national registration data as collected in the FLEETS project. The fleet is layered by age and thus the respective emission standard. A wide array of legislation will change the vehicle fleet in Europe The European Union has issued comprehensive legislation on energy efficiency and emission controls that will have profound impacts on the future vehicle fleet in Europe (Box 9). The baseline scenario assumes that all this legislation will be implemented according to schedule and deliver the envisaged technological improvements. Box 9: Transport policy measures assumed in the baseline and how they are reflected in the PRIMES model Regulation on CO2 from cars 2009/443/EC Limits on emissions from new cars: 135 g CO2/km in 2015, 115 in 2020, 95 in 2025 in test cycle. The 2015 target should be achieved gradually with a compliance of 65% of the fleet in 2012, 75% in 2013, 80% in 2014 and 100% in Penalties for non compliance are dependent on the number of grams until 2018; starting in 2019 the maximum penalty is charged from the first gram. Regulation EURO 5 and /715/EC Fuel Quality Directive 2009/30/EC Biofuels directive 2003/30/EC Labelling regulation for tires 2009/1222/EC Emission limits introduced for new cars and light commercial vehicles Modelling parameters reflect the Directive, taking into account the uncertainty related to the scope of the Directive addressing also parts of the energy chain outside the area of PRIMES model (e.g., oil production outside EU). Support to biofuels such as tax exemptions and obligation to blend fuels is reflected in the model The requirement of 5.75% of all transportation fuels to be replaced with biofuels by 2010 has not been imposed as the target is indicative. Support to biofuels is assumed to continue. The biofuel blend is assumed to be available on the supply side. Decrease of perceived costs by consumers for labelling (which reflects transparency and the effectiveness of price signals for consumer decisions). Regulation Euro VI for heavy duty vehiclesemissions limits introduced for new heavy duty vehicles. 2009/595/EC 16

25 ` Baseline trends in the transport sector The baseline scenario assumes that demand for passenger travel and freight transport will continue to grow in the coming decades. For passengers, it is estimated that total mobility (person kilometers) would increase by more than one third between 2005 and 2030, with a doubling of the air transport mileage (Figure 17). billion km/yr Figure 17: Total passenger mileage by transport modes Vehicle mileage will grow faster than the demand for personal mobility Higher car ownership facilitated by higher income levels will reduce the average occupancy, so that vehicle mileage will grow by 50%, i.e., even more than passenger mileage (Figure 18). Relative to % 140% 120% 100% 80% 60% 40% 20% 0% By car By rail By air Transport volume (total travel distance by people in cars) Total vehicle mileage Fuel consumption by passenger cars Figure 18: Passenger transport demand, vehicle mileage and fuel consumption by passenger cars However, the new fuel efficiency requirements would off set the impacts of the growth in vehicle mileage on energy consumption and thereby decouple total fuel consumption from travel growth. The share of diesel passenger cars is likely to remain at the 2010 level While the market share of diesel (light duty) cars increased significantly over the last years, models suggest no further growth for the future (Figure 19). Petajoule Figure 19: Diesel and gasoline consumption for light duty (passenger) vehicles Freight transport intensity of the European economy will not change significantly Despite continuing harmonization and expansion of the common European market, no significant changes in freight transport intensities are foreseen for the future, as GDP is shifting more to the service sector (Figure 20). GDP/capita (1000 Euro/yr) Gasoline Diesel GDP/capita: Old Member States GDP/capita: New Member States GDP/capita: EU 27 Transport intensity Old Member States Transport intensity New Member States Transport intensity EU Figure 20: Assumed development of GDP and freight transport intensity (freight volumes times transport distance per GDP) in the old and new Member States. Due to the high share of the service sector in the GDP of the old Member States, transport intensities (i.e., tonkm per GDP) are much lower than in the new Member States. The convergence process in the new Member States is likely to increase transport intensities in the near term, while after 2015 intensities should decline due to the shift to less material intensive economies tons.km/1000 Euro 17

26 Total freight transport volumes will continue to rise, but at a lower rate than the overall economy The pertained growth of economic activities, together with further integration, harmonization and expansion of the common European market, will lead to a steady increase in freight transport volumes (Figure 21). Road transport by trucks accounts for more than 70% of total ton kilometres. Total road fuel consumption is projected to stagnate due to more efficient vehicles In summary, new legislation on fuel efficiency should slow down fuel demand for total road transport despite the expected increases in travel distance and freight volumes (Figure 23). In particular, new vehicle technologies are expected to stabilize fuel demand for road freight transport vehicles despite the envisaged increase in truck mileage by about one third By truck By rail By inland ships 160% 140% 120% billion ton kms Relative to % 80% 60% 40% GDP Freight transport volume (distance*load) Figure 21: Freight transport volumes (tons kilometers) by transport mode There are strong differences in the structures of the transportation sectors between the old and new Member States. The baseline assumes a sharp decline in the shares of rail and waterway transport in the new Member States. As a consequence, road transport volumes in the new Member States will increase at a faster rate than GDP in the coming years; after 2015, however, growth trends (relative to GDP) should converge to those of the old Member States (Figure 22). 20% 0% Total truck mileage Truck fuel consumption Figure 23: Transport demand, vehicle mileage and fuel consumption by heavy duty vehicles On a per capita basis, fuel consumption for road transport will increase in the new Member States While new legislation is expected to reduce fuel consumption in general, economic convergence will increase it in some of the new Member States (Figure 24). 250% 200% Figure 24: Fuel consumption on a per capita basis Relative to % 100% 50% 0% GDP: Old Member States Freight transport volume: Old Member States GDP: New Member States Freight transport volume: New Member States Figure 22: Road freight transport volumes (tonkilometers) New vehicles with tighter emission standards penetrate the fleet As old vehicles retire from the fleet, new vehicles complying with the latest European exhaust emission standard will gradually penetrate the fleet (Figure 25, Figure 26). Since new vehicles have higher mileage than older vehicles, their share in activity is even higher than their share in stock. This leads to a reduction of the fleet average emission factor, with the notable exception of NO x emissions. 18

27 ` Share of total vehicle mileage 100% 90% 80% 70% 60% 50% 40% 30% Diesel Gasoline Euro 6 Euro 5 Euro 4 Euro 3 Euro 2 Euro 1 No control Share of total vehicle mileage 100% 90% 80% 70% 60% 50% 40% 30% Euro VI Euro V Euro IV Euro III Euro II Euro I No control 20% 20% 10% 10% 0% Figure 25: Penetration of Euro standards for passenger cars in the EU 27 0% Figure 26: Penetration of Euro standards for heavy duty road vehicles in the EU 27 Table 5: Fuel consumption for transport by Member State (in PJ, based on the fuel sold concept) Fossil diesel Fossil gasoline Biofuels Austria Belgium Bulgaria Cyprus Czech Rep Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU

28 Land use and agricultural activities Agricultural activities and land use are important sources of harmful emissions to the atmosphere. The production of agricultural commodities (food, feed, fibre, bio fuels) requires land area. Production capacities are limited by the amount of land required for other purposes, by the market prices of the respective products, and by biophysical constraints. Cost effective policy responses aiming at the protection of the atmosphere must consider all these factors, including the implications of initiatives in other policy areas. Modelling land use and agricultural activities In EC4MACS, agricultural and energy models are linked by the European Forest and Agricultural Sector Optimization (EUFASOM) model (Box 10). The EUFASOM and GLOBIOM models simulate the competition for available land resources between the agricultural, forest and energy sectors EUFASOM uses the results of the PRIMES and the CAPRI models and derives, under market conditions, land use and bio fuel production by crop type (forest biomass, arable crops, perennial crops such as perennial grasses and short rotation woody crops). The CAPRI model quantifies implications of changes in lifestyles and agricultural policies on agricultural activities in Member States. The CAPRI model (Box 11) quantifies the implications of the changes in lifestyles, in the macro economic development and of specific agricultural policy measures on the different types of agricultural activities. Box 11: The CAPRI model The CAPRI model simulates the impacts of policy measures on a highly differentiated set of agricultural activities. It is a partial equilibrium model for the agricultural sector developed for policy impact assessment of the Common Agricultural Policy and trade policies from global to regional scale with a focus on the EU. It builds upon an analysis of observed historical trends, on expert information, and on standard economic modelling. CAPRI employs profit maximising rationales for the economic actors, considering nutrient balances and emissions from the production system. The CAPRI market module is a spatial multi commodity model. Equilibrium ensures cleared markets for products and young animals, and the feeding needs of national herds. CAPRI includes welfare analysis considering economic and environmental indicators. CAPRI covers the entire EU agriculture through 280 regional models and 1900 farm type models. Crop shares, yields, stocking densities, and fertilizer application rates are considered at 1x1 km resolution. Box 10: The EUFASOM and GLOBIOM models The EUFASOM and GLOBIOM models simulate land use management under environmental, political, and technical change. They use as drivers the baseline projections on biofuel use from the PRIMES model and related assumptions on population growth, economic development (GDP), and technical progress rates. Data on potential yields and GHG emissions and removals for agricultural and forest management alternatives are derived from the more detailed forestry model (G4M) and the agricultural model (EPIC). For baseline and policy scenarios, the economic land use modules project domestic production and consumption, net exports and prices of wood and agricultural products and changes in land use for EU Member States and other world regions. Sector specific information from the economic modules is used by the forest and agricultural modules to project GHG emissions and removals for detailed land management options. These detailed models cover activities afforestation, reforestation, deforestation and forest management, cropland and grazing land management. The EC4MACS baseline analyzes the impacts of EU agricultural, energy and climate policies The EC4MACS baseline considers the implications of a range of recent policy initiatives that have impacts on land use and agricultural activities (Box 12). Box 12: Agricultural policies included in the baseline The Health Check of the Common Agricultural Policy (CAP) Abolition of the Set aside (regulation 73/2009) and milk quota regulations Agricultural premiums are largely decoupled from production levels The WTO December 2008 Falconer proposal The biofuel targets of the EU Energy and Climate package as modeled by PRIMES The Nitrates Directive 20

29 ` Baseline trends in land use Population growth and economic development increases the demand for built up land, and at the same time for forest and agricultural products. On top of these factors, climate policies that enhance reforestation and biofuel production will put additional pressure on European land resources. The integration of energy, agricultural and land use models in EC4MACS provides a holistic perspective on the resulting competition for land. Carbon stock (Gt C) Managed forests Afforestation/deforestation For the baseline assumptions, the various drivers are expected to cause small, but not insignificant changes in European land use patterns. Increasing demand for built up land will reduce land available for agriculture, forestry and biofuel production Overall, economic development will convert more land into built up area, so that the overall area for forests, crops and grassland in the EU 27 is estimated to shrink by three percent. Forest area and crop land are more or less stable, inter alia due to climate policies that promote carbon storage in forests and biofuel production. In contrast, a nine percent decline is anticipated for pasture (Figure 27). Million hectares Managed forests Cropland Grassland Figure 27: Land use for forests, crops and grassland Forest management will retain carbon stored in woody biomass While the forest area in the EU 27 will remain almost constant, by 2030 forest management should recover the loss in carbon storage in woody biomass that occurred between 2000 and 2005 (Figure 28) Figure 28: Carbon stored in woody biomass in the European forests Higher productivity for food crops will free up land for biofuel production Increasing agricultural productivity will provide the crop demand for food and fodder on 10% less area. The freed up land, together with gains from smaller setaside areas, will be used for biofuel production. (Figure 29). million hectares Cereals Other crops Oil seeds Sugar beets Biofuels Set aside and fallow land Figure 29: Use of crop land for different purposes The area for biofuel production in the EU 27 will increase by a factor of six up to However, different countries are expected to embark on different pathways. While perennial energy crops will become important for bio fuel production in France and Germany, the situation is different for the forest rich countries Sweden and Finland. 21

30 Baseline trends in agricultural activities Crop production recover to the 2005 levels Total volumes of staple crop production is likely to recover back to the 2005 level after the recent decline in sugar beet production (Figure 30). While sugar beets crops are expected to remain at the lower level, the loss in volume will be compensated by higher crops of wheat and other grains. Fertilizer use will rise again Fertilizer use has declined between 2000 and 2010 due to increased efficiency of fertilizer application and changes in crop types (e.g., reduced sugar beet production). However, in the long run, the baseline scenario anticipates a return to the 2000 levels due to the intensification of agricultural production, inter alia, due to increased bio fuel production Other fertilizer use Urea fertilizer use Soft wheat Barley Grain Maize Rape Potatoes Sugar Beet Other Vegetables Million tons N Million tons Figure 32: Baseline trend in fertilizer use in the EU Figure 30: Baseline trend in the production agricultural stale crops for non fodder purposes Biofuel production in the EU will increase by a factor of 10 The renewable energy targets of the EU Energy and Climate package will lead to a significant increase of total biofuel production. In the short run, biofuel production will shift from oil seeds to cereals, while in the long run new energy crops are expected to supply the majority of input (Figure 31). PJ Sugar beet Cereals Oil seeds New energy crops Cattle numbers will decrease, but there will be more pigs and chicken More significant changes are expected in the livestock sector as a consequence of the EU agricultural policy reform. Despite or because of the abolition of the milk quota regime under the Health Check, dairy cow numbers in the EU will decline up to 2015, compensated by higher productivity. This has implications on other cattle, which will decline too, but at a lower rate. Pig and poultry numbers, which are not strongly influenced by new policies, are expected to continue their increase, although their further growth might be limited by local environmental constraints (Figure 33). Livestock units (millions) Dairy cows Cattle Pigs Chicken Sheep Figure 31: Baseline trend in biofuel production Figure 33: Baseline trend in livestock numbers in the EU

31 ` National trends in agricultural activities Even larger differences are expected for livestock numbers in the Member States (Table 6, Figure 34). While, e.g., for cows and cattle, increases are expected in Belgium, Ireland and the Netherlands, decreases of up to 30% are foreseen for many new Member States as a consequence of the EU agricultural policy reform. However, the anticipated decline in livestock numbers will be compensated, at least to some extent by higher productivity in response to the abolition of the milk quota regime under the Health Check. Table 6: Livestock numbers of the EC4MACS baseline scenario, by Member State (1000 animals) Cows and cattle Pigs Chicken and poultry Austria Belgium Bulgaria Cyprus Czech R Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU Relative to % 100% 80% 60% 40% % 0% AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 34: Changes in cattle livestock numbers by Member State, relative to

32 24

33 ` Emissions 25

34 Emissions of the baseline scenario Modelling future emissions of greenhouse gases and air pollutants Within the EC4MACS framework, the GAINS (Greenhouse gas Air Pollution Interactions and Synergies) model is employed to calculate for 2020 and 2030 emissions of the various greenhouse gases and air pollutants (Box 10). Calculations consider future economic activities, country and sector specific emission factors and progressing implementation of the emission control legislation that is currently laid down in national laws. Box 13: The GAINS integrated assessment model The GAINS (Greenhouse gas Air Pollution Interactions and Synergies) model explores cost effective multi pollutant emission control strategies that meet environmental objectives on air quality impacts (on human health and ecosystems) and greenhouse gases. GAINS, developed by the International Institute for Applied Systems Analysis (IIASA) in Laxenburg (Austria), brings together data on economic development, the structure, control potential and costs of emission sources, the formation and dispersion of pollutants in the atmosphere and an assessment of environmental impacts of pollution. GAINS addresses air pollution impacts on human health from fine particulate matter and ground level ozone, vegetation damage caused by ground level ozone, the acidification of terrestrial and aquatic ecosystems and excess nitrogen deposition to soils, in addition to the mitigation of greenhouse gas emissions. GAINS describes the interrelations between these multiple effects and the pollutants (SO 2, NO x, PM, NMVOC, NH 3, CO 2, CH 4, N 2 O, F gases) that contribute to these effects at the European scale. GAINS assesses, for each of the 43 countries in Europe, more than 1000 measures to control emissions to the atmosphere. It computes the atmospheric dispersion of pollutants and analyses the costs and environmental impacts of pollution control strategies. In its optimization mode, GAINS identifies the least cost balance of emission control measures across pollutants, economic sectors and countries that meet userspecified air quality and climate targets. The GAINS model calculates emissions using activity statistics and predictions, fuel, sector and countryspecific emission factors, taking into account the degree of dedicated emission control measures at a given source at each point in time. Emission control policies considered in the baseline GAINS calculations consider a detailed inventory of national emission control legislation (including the transposition of EU wide legislation) as of 2009, and assume that these regulations will be fully implemented in all Member States according to the foreseen time schedule. For CO 2, regulations are included in the PRIMES calculations as they affect the structure and volumes of energy consumption (Box 14). For non CO 2 greenhouse gases and air pollutants, EU and Member States have issued a wide body of legislation that limits emissions from specific sources, or have indirect impacts on emissions through affecting activity rates (Box 11). Box 14: Policies and regulations affecting CO 2 emissions that are considered in the baseline scenario EU directives and regulations aiming at efficiency improvements, e.g., for energy services, buildings, labelling, lighting, boilers Regulation on new cars (involving a penalty for car manufacturers if the average new car fleet exceeds 135 g CO 2 /km in 2015, 115 in 2020, 95 in 2025 in test cycle) Strong national policies supporting use of renewable energy; however compliance with the 20% target share of renewable energy is not mandatory Co generation Directive Carbon Capture and Storage (CCS) demonstration plants Harmonisation of excise taxes on energy The Emission Trading Scheme (ETS) Directive Box 15: Legislation considered for non CO 2 GH emissions Landfill Directive Waste Directive, EU waste treatment hierarchy Nitrate Directive CAP reform, CAP health check F gas Directive Motor vehicles Directive The European Emission Trading System (ETS) Other relevant legislation: Transport emission related regulations For air pollutants, the EC4MACS baseline assumes the regulations described in Box 13. However, the analysis does not consider the impacts of other legislation for which the actual impacts on future activity levels cannot yet be quantified. This includes compliance with the air quality limit values for PM, NO 2 and ozone established by the new Air Quality Directive, which could require, inter alia, traffic restrictions in urban areas and thereby 26

35 ` modifications of the traffic volumes assumed in the baseline projections. Although some other relevant directives such as the Nitrates Directive are part of current legislation, there are some uncertainties as to how the measures can be represented in the framework of integrated assessment modelling. Although in the past EU legislation imposed increasingly stringent emission limit values for vehicles, NO x emissions from diesel cars under real world driving conditions turned out as much higher than the limit values. In fact, NO x emissions have not decreased for Euro 2 to Euro 4 limit values. The (higher) real world emission factors are used in the EC4MACS baseline. At the time of completion of COPERT IV (i.e., when latest measurement data referred to Euro 4 cars), there were no representative measurements available for emissions from Euro 5 diesel cars, let alone from Euro 6 cars. Therefore, it was assumed for the baseline scenario that real world emission factors for Euro 5 cars decrease at the same rate as the limit value in relation to Euro 4 value. Similarly, it was also assumed that the same proportional reduction in NO x emissions relative to Euro 4 applies to Euro 6. Thus, while a successful reduction rate was assumed, the resulting absolute emission level is still two times higher than the nominal level of the limit value. NOx: Large Combustion Plants Directive EURO standards, including adopted Euro 5 and Euro 6 for light duty vehicles Euro standards, including adopted Euro V and Euro VI for heavy duty vehicles EU emission standards for motorcycles and mopeds up to Euro 3 Legislation on non road mobile machinery Higher real life emissions for diesel heavy duty and light duty vehicles compared with the test cycle IPPC requirements for industrial processes as currently laid down in national legislation National legislation and national practices (if stricter) NH 3 IPPC Directive for pigs and poultry production as interpreted in national legislation National legislation including elements of EU law, i.e., Nitrates and Water Framework Directives Current practice including the Code of Good Agricultural Practice Box 16: Legislation considered for air pollutant emissions SO 2 : Large Combustion Plants Directive Directive on the sulphur content in liquid fuels Directives on quality of petrol and diesel fuels, as well as the implications of the mandatory requirements for renewable fuels/energy in the transport sector IPPC requirements for industrial processes as currently laid down in national legislation Sulphur content of gasoil used by non road mobile machinery and inland waterway vessels (reduction from 1000 ppm to 10 ppm) according to the Proposal COM(2007) 18 of the Directive of the European Parliament and of the Council to amend Directives 98/70/EC and 1999/32/EC. National legislation and national practices (if stricter) VOC: Stage I Directive (liquid fuel storage and distribution) Directive 96/69/EC (carbon canisters) Euro standards, including adopted Euro 5 and Euro 6 for light duty vehicles EU emission standards for motorcycles and mopeds up to Euro 3 Fuel Directive (RVP of fuels) Solvents Directive Products Directive (paints) National legislation, e.g., Stage II (gasoline stations) PM2.5 Large Combustion Plants Directive EURO standards, including the adopted Euro 5 and EURO 6 standards for light duty vehicles EURO standards, including adopted Euro V and Euro VI for heavy duty vehicles. Legislation on non road mobile machinery IPPC requirements for industrial processes as currently laid down in national legislation National legislation and national practices (if stricter) 27

36 Carbon dioxide (CO 2 ) emissions CO 2 emissions from energy use In 2005, power generation constituted the largest source of carbon dioxide (CO 2 ) emissions in the EU 27, contributing 36% to the total (excluding LULUCF emissions from land use and land use changes). Energyand non combustion related emissions from industrial sources accounted for 25%, transport for 21% and households for 16% Current climate policies should lead to a 16% decline in CO 2 emissions The energy and climate policies that are assumed in the baseline projection are expected to lead to a definite decoupling between economic growth and CO 2 emissions. Despite the 50% increase in GDP that is assumed for 2030 relative to 2005, CO 2 emissions would decline by 18% (Figure 35). In 2020, CO 2 would be 13% below the 2005 level. Million tons CO Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Together, the power and domestic sectors will deliver three quarters of the total CO 2 reduction Largest reductions (in absolute and relative terms) are expected in the power sector and domestic sectors, whose emissions would decline by a quarter compared to Thereby they would deliver about three quarters of the total emission reduction. The emission trading regulations for in the power sector would cut emissions before 2020, while for the domestic sector a continuous decline of emissions is anticipated. Emissions from road transport and energy use in industry are expected to decline by about nine percent (Figure 36), but non energy related CO 2 emissions in the industrial sector are likely to increase with growing production levels (Table 7). Emissions relative to % 140% 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 36: Sectoral trends in baseline emissions of CO 2 in the EU 27 Figure 35: Baseline emissions of CO 2 in the EU 27 by SNAP sector (excluding LULUCF) Table 7: Baseline emissions of CO 2 by SNAP sector (excluding emissions from the LULUCF sector) (million tons) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum

37 ` While for the whole EU 27 baseline CO 2 emissions decrease by 18% in 2030 compared to 2005, there are considerable differences in the trends of individual Member States. As a general pattern, emissions in the old Member States will decline, while many new Member States will have slightly higher baseline emissions in 2030 than in Largest reductions are calculated for the UK, Portugal and Malta, where in 2030 baseline emissions are approximately one third lower than in In contrast, a nine percent increase is anticipated for Latvia (Table 8). Table 8: Energy related baseline emissions of CO 2 by country (million tons and change relative to 2005) [Mt] [Mt] [Mt] Change [Mt] Change [Mt] Change Austria % 68 13% 64 19% Belgium % % 134 3% Bulgaria % 49 9% 45 17% Cyprus % 8 2% 8 2% Czech Rep % % % Denmark % 43 17% 41 20% Estonia % 17 1% 16 6% Finland % 52 13% 51 14% France % % % Germany % % % Greece % 103 8% 105 6% Hungary % 60 1% 57 7% Ireland % 42 11% 44 7% Italy % 445 9% % Latvia % 9 18% 9 9% Lithuania % 16 6% 15 1% Luxembourg % 12 0% 12 3% Malta % 2 36% 2 44% Netherlands % % % Poland % 332 4% 328 3% Portugal % 50 28% 47 33% Romania % 102 4% 94 11% Slovakia % 42 4% 42 2% Slovenia % 20 18% 15 11% Spain % 347 7% 360 3% Sweden % 40 24% 36 33% UK % % % EU % % % tons/person/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 37: Per capita CO 2 emissions by Member State, 2005 and

38 CO 2 emissions from the land use (LULUCF) sector Deforestation and the management of cropland constitute a source of CO 2 emissions to the atmosphere, while grasslands, afforestation and managed forests provide enhanced carbon storage in soil and woody biomass. Human induced emissions and sinks from this sector are accounted in the UNFCCC inventories under the Land use, land use change, and forestry (LULUCF) category. In 2005, emissions from land management contributed in the EU 27 another 2.5% to the energy related CO 2 emissions; about 70% were caused by cropland management. In contrast, management of forests and grassland lead to a sink of carbon corresponding to 8.3% of energy related emissions; 90% was related to forest management. The success of EU forests in absorbing carbon at these high rates in the past has been due to a variety of factors, most importantly that European forests growth rates have been higher than past harvest rates. The net carbon sink from this sector will drop by 40% until 2030 With current trends and policies, the baseline projection suggests that the net carbon sink will decline by about 40% until 2030 (Figure 38). The main factor for this fall is the shrinking sink in existing forests by more than 50%. Deforestation emissions will decrease at the same rate, but they contribute less to total emissions. Emissions and removals from cropland and grassland will stay relatively stable. The drastic decline in carbon removals from forest management is partly compensated by an increasing sink of new forests. However, in 2030 forest management will remain the dominating term in Europe s LULUCF carbon budget. million tons CO Deforestation Cropland management Management old forests Afforestation and reforestation Grazing land management Sum Figure 38: Emissions from the LULUCF sector The drivers of the decline of the forest sink are a projected increase in demand for wood for energy and material use combined with shifts in the forest structure toward older forests that lower the strength of forest carbon accumulation. Carbon sink will shrink, inter alia, due to the growing demand for energy wood The PRIMES model estimates that bio energy from fellings and traditional fuel wood will increase by 43 million m³ between 2005 and Future changes in forest management practices as a response to this increased demand will affect the capacity of EU forests to store carbon. In other words, a potential conflict exists between climate policies targeting carbon storage in forests and the increased use of forest based biomass for energy purposes. EC4MACS explored this trade off through scenario analyses. The reference scenario that was compared to the baseline includes the national renewable targets of EU member states for 2020 to meet the EU target of a share of 20% renewable energy sources in energy consumption in 2020 as well as the 20% reduction in GHG emissions compared to As a result, until 2020 the forest management sink will further decrease by 4 11% (or Mt CO 2 ) compared to the baseline (Figure 39). This effect is currently not accounted for, as the emission reduction target of 20% excludes land use emissions and removals. In the longer run regrowth effects can, however, revert the trend again. Carbon sink (million tons CO2/year) without renewable energy targets with renewable energy targets Figure 39: Development of the carbon sink in the EU 27 as a consequence of the EU renewable energy targets The development of total LULUCF emissions for each EU Member State separated by model combinations can be found in Table 9. While Czech Republic, Estonia, Germany, Ireland, Netherlands and Poland emerge as sources of CO 2 in the LULUCF sector in 2005, all other countries appear as net sinks in the early phase. Over 30

39 ` time, however, some countries change from sinks to sources (Belgium, Slovakia), while only Ireland becomes a net sink. Germany is the largest net emitter in 2005; towards 2030, Poland will assume the leading country. Largest total net removals exist in Romania, France and Italy throughout the entire simulation period. Table 9: Baseline projection of total LULUCF emissions by country, as projected by G4M for afforestation, reforestation, forest management, and the EUFASOM model for crop management and grassland management (million tons CO 2 per year). No estimates are available for Cyprus and Malta. Positive values indicate sources, negative values sinks of CO Austria Belgium Bulgaria Cyprus Czech Republic Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU

40 Methane (CH 4 ) emissions In 2005, agricultural activities (mainly livestock farming) emitted almost half of the methane (CH 4 ) emissions in the EU 27. Another one third of emissions originated from waste treatment (from solid waste disposal and wastewater treatment), and 14 percent from fuel extraction and distribution (i.e., coal mining and distribution of natural gas). Methane emissions will decline by 20% compared to 2005 due to EU regulations Over the last years, the EU has implemented comprehensive policies to reduce methane emissions in the future, which should lead to a 20% decline of CH 4 emissions 20% by 2030 compared to 2005 (Figure 40). More than half of the decline has already materialized by 2012, so that the further decline in the future is expected to emerge at a less rapid pace. Million tons CH Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 40: CH 4 emissions of the EU 27 by SNAP sector Two thirds of the emission reduction will emerge from improved waste treatment Especially large reductions occur for waste treatment, where the progressing implementation of current EU legislation on solid waste disposal and waste water management, particularly in the new Member States, will lead to a sharp decline of CH 4 emissions in the coming years (Figure 41). In 2030, EU 27 emissions from this sector should be 45% lower than in Emissions relative to % 140% 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 41: Sectoral trends in baseline emissions of CH 4 in the EU 27 The second largest contributions to emission reductions will come from improved gas distribution networks, for which losses will be cut by about 45% up to In contrast, emissions from the agricultural sector are expected to increase by four percent (Table 10). Table 10: Baseline emissions of CH 4 by SNAP sector (kilotons) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum

41 ` Large differences between old and new Member States There are large differences in the evolution of methane emission between countries. Many new Member States will reduce their CH 4 emissions by 30% up to 47%, mainly as a result of the implementation of EU waste management regulations and the on going upgrades of gas distribution networks. In contrast, emissions in most old Member States would decline less, as much of the waste management legislation has already been implemented in the past. Also, emissions from the agricultural sectors contribute a larger share to total emissions, and this sector is not expected to dramatically reduce its emissions in the future. For instance, only marginal changes are anticipated for, e.g., Belgium, Denmark and Ireland. Table 11: Baseline emissions of CH4 by country (kilotons and change relative to 2005) [Mt] [Mt] [Mt] Change [Mt] Change [Mt] Change Austria % % % Belgium % 324 1% 328 1% Bulgaria % % % Cyprus % 43 16% 47 8% Czech Rep % % % Denmark % 252 7% 267 2% Estonia % 62 29% 63 28% Finland % % % France % % % Germany % % % Greece % % % Hungary % % % Ireland % 601 3% 623 1% Italy % % % Latvia % 70 27% 79 18% Lithuania % % % Luxembourg % 26 7% 28 16% Malta % 10 12% 9 17% Netherlands % % % Poland % % % Portugal % % % Romania % % % Slovakia % % % Slovenia % 84 19% 73 29% Spain % % % Sweden % % % UK % % % EU % % % kg/person/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 42: Per capita emissions of CH 4 in 2005 and 2020 (kg/person/year) 33

42 Nitrous oxides (N 2 O) emissions Emissions from soils constituted the major source of N 2 O emissions in 2005 (about 70% of total emissions), although their exact quantification is still associated with significant uncertainties. A comparably small share of emissions was contributed by industrial activities, such as production of nitric acid and adipic acid in the chemical industry. N 2 O emissions are expected to fall by about 10% until 2030 Overall, emissions from the EU 27 are expected to fall by about 10% until 2030 (Figure 43). However, there are two counteracting mechanism that influence the overall trend: The inclusion of industrial sectors (i.e., production of nitric acid and adipic acid in the chemical industry) with cheap mitigation potentials into the EU emission trading system (ETS) will cut N 2 O emissions from these sources by about 80%. As there are considerable economic gains to be reaped, it is likely that these emission reductions will happen in the next few years. In contrast, the intensification of agriculture, inter alia to enable biofuel production, will raise N 2 O emissions from soils by about six percent. Other sectors make only small contributions to total emissions in absolute terms, so that even large relative changes do not significantly change the overall picture (Figure 44). However, it should be noted that the quantification of N 2 O emissions from soils is associated with significant uncertainties, and that new emission inventory techniques could change the conclusions. Million tons N2O/year Figure 43: N 2 O emissions of the EU 27 by SNAP sector Emissions relative to % 160% 140% 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Figure 44: Sectoral trends in baseline emissions of N 2 O in the EU 27 Table 12: Baseline emissions of N 2 O by SNAP sector (kilotons) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Statistical difference Sum

43 ` N 2 O emissions will decline in the old Member States, but will increase in the new Member States The different shares of chemical industries and different trends in future biofuel production will lead to rather inhomogeneous emission trends in the Member States. As a general pattern, N 2 O emissions in many old Member States are likely to decline, due to the extension of the ETS sector. For instance, for 2030 a 20 25% reduction of the N 2 O emissions in Germany and the Netherlands is estimated. In contrast, emissions in the new Member States, especially in those where biofuel production is likely to expand, will increase. Examples are Latvia and Estonia, with a 40 to 50% growth in N 2 O (Figure 20). Table 13: Baseline emissions of N 2 O by country (kilotons and change relative to 2005) [kt] [kt] [kt] Change [kt] Change [kt] Change Austria % 17 10% 18 14% Belgium % 26 14% 27 13% Bulgaria % 13 2% 15 11% Cyprus % 1 1% 1 5% Czech Rep % 16 14% 17 10% Denmark % 17 8% 17 10% Estonia % 3 20% 3 40% Finland % 16 15% 15 15% France % 188 7% 191 5% Germany % % % Greece % 25 7% 25 6% Hungary % 30 2% 33 12% Ireland % 23 12% 24 11% Italy % 87 23% 90 21% Latvia % 6 36% 7 56% Lithuania % 14 14% 15 6% Luxembourg % 2 15% 2 14% Malta % 0 21% 0 19% Netherlands % 37 26% 37 26% Poland % 75 1% 82 8% Portugal % 16 2% 16 1% Romania % 50 3% 52 0% Slovakia % 8 28% 8 25% Slovenia % 4 2% 4 0% Spain % 100 7% % Sweden % 20 9% 20 8% UK % 85 13% 87 10% EU % % % 7 6 kilogram/person/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 45: Per capita emissions of N 2 O in 2005 and 2020 (kg/person/year) 35

44 F gas emissions Use of refrigerants for air conditioning and cooling is the major source of F gas emissions in the EU 27. In 2005, these activities accounted for 85% of all F gas emissions. The remaining 15% originated from industrial processes, e.g., the production of semiconductors, the use of SF 6 in the electrical industry and from styrofoam production. While emission intensities of industrial production will decline, further penetration of air conditioning will cause a 25% increase in EU emissions Million tons CO2eq/yr Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture New regulations on the emissions for many of these sources will lead to declining emission intensities. In particular, emissions from industrial processes are expected to decrease by 40% up to 2030 (Figure 46). However, increasing demand for mobile and stationary air conditioning and cooling will boost these activity levels, so that baseline emissions from this sector are likely to increase by 35% compared to 2005 despite progressing implementation of the recent legislation (Figure 47). In total, F gas emissions are calculated to increase by 25% in the baseline case (Table 14) Figure 46: F gas emissions of the EU 27 by SNAP sector Emissions relative to % 200% 150% 100% 50% 0% Industrial processes Solvent use Figure 47: Sectoral trends in baseline emissions of F gases in the EU 27 Table 14: Baseline emissions of F gases in the EU 27 by SNAP sector (million tons CO2eq) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum

45 ` Most countries will increase their F gas emissions in the future Due to the differences in industrial structures, climatic conditions and personal lifestyles, trends in F gas emissions show large variations across the Member States (Figure 48). Emissions in 25 out of the 27 countries are expected to increase (by up to 76%, e.g., in Spain). Declines are calculated for Greece and Malta, although from a rather high level in Overall, baseline emissions in 2020 grow by 10% in 2005, and by 25% in 2030(Table 15) Table 15: Baseline emissions of F gases by country (million tons CO2eq and change relative to 2005) [kt] [kt] [kt] Change [kt] Change [kt] Change Austria % 2 33% 2 26% Belgium % 2 39% 2 46% Bulgaria % 0 0% 0 24% Cyprus % 0 0% 0 25% Czech Rep % 1 8% 1 25% Denmark % 1 24% 1 31% Estonia % 0 31% 0 56% Finland % 1 7% 1 22% France % 12 6% 14 10% Germany % 17 18% 17 14% Greece % 2 65% 2 59% Hungary % 1 27% 1 47% Ireland % 1 23% 1 52% Italy % 8 38% 10 60% Latvia % 0 40% 0 60% Lithuania % 0 47% 0 53% Luxembourg % 0 6% 0 13% Malta % 0 20% 0 20% Netherlands % 4 58% 4 75% Poland % 2 6% 2 32% Portugal % 1 0% 1 22% Romania % 1 13% 1 21% Slovakia % 0 29% 0 52% Slovenia % 0 29% 0 41% Spain % 8 42% 10 76% Sweden % 2 40% 2 61% UK % 12 1% 14 20% EU % 79 10% 89 24% 0.6 tons CO2eq/person/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 48: Per capita emissions of F gases in 2005 and 2020 (tons CO2eq/person/year) 37

46 Total GHG emissions For the Kyoto protocol of the UNFCCC, emissions of the six greenhouse gases are aggregated to a basket of total greenhouse gas emissions, by weighting the six gases by their global warming potentials. On this basis, CO 2 contributed about 83%, CH 4 9%, N 2 O 7%, and F gases 1% to total greenhouse gases in 2005 Total greenhouse gas emissions will decline by 17% in 2030 At the aggregated level, greenhouse gas emissions of the EU 27 will decrease in the baseline case by 13% in 2020 and by 17% in 2030 relative to 2005 (Figure 49). million tons CO2eq Figure 49: Total GHG emissions of the EU 27 by gas Despite different trends, the relative shares of gases will remain The strongest decrease is anticipated for methane, while F gas emissions are expected to grow (Figure 50). However, this will not lead to major shifts in the relative shares of the different gases in 2030 Relative to % 120% 100% 80% 60% 40% 20% 0% CO2 CH4 N2O FGAS Total Figure 50: Trends in GHG emissions by gas CO2 CH4 N2O FGAS The shares of different sectors in total GHGs will change Power generation was the largest sources of GHG emissions in 2005, contributing 30% to the total basket. Industry emitted 20%, road transport 18%, and households 15%. (Figure 51) Million tons CO2eq Figure 51: Total GHG emissions of the EU 27 by SNAP sector 44 percent of the emission reduction between 2005 and 2030 results from the power sector, 24% from households and small sources, 14% from industry, and 9% from road transport. In contrast, agricultural emissions will grow by three percent (Figure 52). Emissions relative to % 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 52: Sectoral trends in baseline emissions of total GHGs in the EU 27 This will reduce the share of power generation from 30% to 27%, and of small sources from 18% to 15%. In 2030, agriculture will then increase its contribution to 9% of the basket of greenhouse gas emissions. 38

47 ` GHG emission reductions span a wide range over countries Most countries will reduce their total GHG emissions, with largest cuts in Malta, Portugal, UK and Germany (between 30% and 41%). In contrast, emissions would increase in Cyprus, Poland and Belgium (by 1 2%), and most notably in Latvia by 10% (Table 16). On a percapita basis, Luxembourg and Ireland will remain the countries with highest emissions (Figure 53). Table 16: Baseline emissions of total greenhouse gases (million tons CO2eq and change relative to 2005) [kt] [kt] [kt] Change [kt] Change [kt] Change Austria % 80 12% 76 16% Belgium % % 151 2% Bulgaria % 60 12% 56 18% Cyprus % 9 1% 10 1% Czech Rep % % % Denmark % 55 15% 54 17% Estonia % 19 1% 18 6% Finland % 61 14% 61 15% France % % % Germany % % % Greece % % 123 9% Hungary % 76 4% 73 9% Ireland % 63 10% 66 6% Italy % % % Latvia % 13 12% 13 10% Lithuania % 23 2% 22 6% Luxembourg % 14 1% 13 1% Malta % 2 32% 2 40% Netherlands % % % Poland % 392 2% 390 2% Portugal % 64 28% 61 31% Romania % 141 6% % Slovakia % 48 1% 48 3% Slovenia % 23 14% 18 12% Spain % 418 5% 435 2% Sweden % 53 21% 49 27% UK % % % EU % % % ton CO2eqperson/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 53: Per capita emissions of total GHGs in 2005 and 2020 (tons CO2eq/person/year) 39

48 Sulphur dioxide (SO 2 ) emissions Most of sulphur dioxide (SO 2 ) emissions from human activities originate from the combustion of sulphurcontaining fuels such as coal and oil. Thus, the volumes of such fuels that are burned as well as the application and efficiency of dedicated emission control technologies are key determinants of total emissions. In the year 2005 more than two thirds of the EU 27 SO 2 emissions was generated in the power sector, and 21% by industrial sources (Table 17). SO 2 emissions will fall drastically in the future The significant changes in fuel consumption levels and patterns that are expected in the baseline projection, together with progressing implementation of emission control measures, will lead to a reduction of about 70% in SO 2 emissions in the coming decades (Figure 54). million tons SO2/year Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 54: SO 2 emissions of the EU 27 by SNAP sector Most of the SO 2 decline will emerge until 2020, for which emissions are estimated to be 68% below their 2005 levels. In the decade after 2020, emissions would then fall to 72%. The power sector will cut its emissions by almost 90% More than 80% of the drop will come from the power sector, which will reduce its emissions by almost 90% compared to This is a direct consequence of the decarbonisation in response to the EU climate targets, as well as of the progressing implementation of the IED and IPPC directives. The domestic sector will decrease its SO 2 emissions by about 50%, while industrial emissions are expected to fall by 40% (Figure 55). Emissions relative to % 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 55: Sectoral trends in baseline emissions of SO 2 in the EU 27 Table 17: Baseline emissions of SO 2 in the EU 27 by SNAP sector (kilotons) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum

49 ` All countries will reduce SO 2 emissions, with up to 90% decrease in some new Member States Legislation and changes in the energy structures will lead to declining SO 2 emissions in all Member States (Table 18). However, particularly large drops are anticipated for some of the new Member States (e.g., Bulgaria, Estonia, Romania, etc.) where emissions will fall by 80 to 90%. These countries exhibited extraordinary high per capita levels of SO 2 emissions in 2005 (Figure 56), and current legislation will bring them down close to the EU 27 average level. Table 18: Baseline emissions of SO 2 by country (kilotons and change relative to 2005) [kt] [kt] [kt] Change [kt] Change [kt] Change Austria % 18 32% 17 35% Belgium % 82 42% 76 45% Bulgaria % % 85 91% Cyprus % 4 89% 5 88% Czech Rep % % 94 53% Denmark % 11 38% 11 39% Estonia % 14 81% 9 88% Finland % 33 54% 40 45% France % % % Germany % % % Greece % % 84 84% Hungary % 59 54% 56 57% Ireland % 29 63% 26 66% Italy % % % Latvia % 5 9% 4 19% Lithuania % 15 67% 15 67% Luxembourg % 1 35% 1 39% Malta % 3 77% 1 88% Netherlands % 42 36% 40 39% Poland % % % Portugal % 62 72% 59 74% Romania % % % Slovakia % 40 56% 39 57% Slovenia % 16 61% 11 73% Spain % % % Sweden % 27 22% 25 28% UK % % % EU % % % kilograms/person/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 56: Per capita emissions of SO 2 in 2005 and 2020 (kg/person/year). The scale is cut at a value of 60, the actual value of Bulgaria in 2005 is 116 kg/person/year. 41

50 Nitrogen oxides (NO x ) emissions In 2005, mobile sources were the largest source of NO x emissions in the EU 27. Road transport contributed 44% of total emissions. Together with off road mobile sources (16%), this sector was responsible for 60% of all emissions. 22% of NO x originated in the power sector, 13% in industry, and 6% in households. NO x emissions will drop by more than 60% These emissions and the sector contributions will change significantly in the future as a consequence of recent EU legislation. Overall, NOx emissions are expected to decline by more than 60% in 2030 (Figure 57). This decline will happen gradually as a consequence of the staged introduction of more stringent emission controls to new vehicles and plants; in 2020, NO x emissions should be 48% below the 2005 levels. Million tons NOx/year Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 57: NO x emissions of the EU 27 by SNAP sector Largest reductions should occur in the transport sector Assuming successful implementation of Euro 6/VI standards, vehicle NO x emissions should drop by more than 80% in 2030 (Figure 58). However, real life emissions of new regulations are somewhat uncertain due to a number of factors (including prescribed test cycles, etc.). Emissions from power generation are expected to decrease by 60%. For the industrial and domestic sectors, smaller changes are anticipated. Emissions relative to % 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 58: Sectoral trends in baseline emissions of NO x in the EU 27 If transport emissions develop as planned, the power sector would make the largest contribution to total NO x emissions in However, since road transport constitutes currently the major source of total emissions in the EU 27, the overall decline in NO x emissions will strongly depend on the implementation success of new regulation for mobile sources (Table 19). Table 19: Baseline emissions of NO x in the EU 27 by SNAP sector (kilotons) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum

51 ` In contrast to other pollutants, national trends for NO x emissions differ less than for other pollutants. Largest reductions are expected for countries where transport has a comparably large share in total emissions, where NO x would drop by up to 77% in Smallest declines range in the order of 50% (Table 20). There are large differences in per capita emissions across countries, owing to different transport habits and, inter alia, the sales of fuels to international customers (Figure 59). Table 20: Baseline emissions of NO x (kilotons and change relative to 2005) [kt] [kt] [kt] Change [kt] Change [kt] Change Austria % 95 53% 71 65% Belgium % % % Bulgaria % 73 60% 58 68% Cyprus % 13 43% 9 60% Czech Rep % % % Denmark % 85 54% 62 66% Estonia % 22 43% 16 60% Finland % % 93 53% France % % % Germany % % % Greece % % % Hungary % 85 49% 61 64% Ireland % 65 51% 53 60% Italy % % % Latvia % 26 28% 16 56% Lithuania % 29 55% 22 66% Luxembourg % 18 64% 11 78% Malta % 4 64% 2 78% Netherlands % % % Poland % % % Portugal % % 85 68% Romania % % % Slovakia % 58 41% 47 52% Slovenia % 27 50% 16 71% Spain % % % Sweden % 97 56% 72 67% UK % % % EU % % % kg/person/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 59: Per capita emissions of NO x in 2005 and 2020 (kg/person/year). The vertical scale is cut at 50 kg, the actual value for Luxembourg in 2005 is 105 kg/person/year. 43

52 Fine particulate matter (PM 2.5 ) emissions In 2005, small sources in the domestic and service sectors contributed about 30% of total PM 2.5 emissions in the EU 27, followed by mobile sources (road transport 15%, off road sources 9%), industry (23%) and the power sector (9%). Legislation directed at other pollutants will decrease PM2.5 emissions by about 40% Current legislation that is often directed towards other pollutants will also have an impact on PM2.5 emissions. Overall, baseline PM 2.5 emissions are expected to decrease by 40% between 2005 and 2030, with a 30% cut in 2020 (Figure 60). Million tons PM2.5/year Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 60: PM2.5 emissions of the EU 27 by SNAP sector Largest cuts would come from diesel emissions As a side effect of regulations for PM10 emissions, PM2.5 from mobile sources is expected to decrease by about 75% in The power sector would cut its emissions to a similar extent, partly due to the switch away from coal. Even without dedicated PM regulations, emissions from the domestic sector would drop by 40% due to phase out of solid fuels. In contrast, industrial emissions are expected to decline by 20% only (Figure 61). Emissions relative to % 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 61: Sectoral trends in baseline emissions of PM2.5 in the EU 27 The domestic sector will remain the largest source, but industrial emissions will gain higher relative shares as other sectors implement more ambitious control measures (Table 21). Table 21: Baseline emissions of PM2.5 in the EU 27 by SNAP sector (kilotons) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum

53 ` PM2.5 emissions in Member States decline between 30 and 70% Depending on the structure of the emission sources, PM2.5 will decline to different extents in the Member States (Table 22). While cuts are larger than 30% everywhere, they reach up to 70% in countries where diesel makes large contributions to total national emissions (e.g., Malta, Cyprus, Luxembourg). Also, outliers in per capita emissions (e.g., Estonia, Portugal) would see largest improvements (Figure 62). Table 22: Baseline emissions of PM 2.5 (kilotons and change relative to 2005) [kt] [kt] [kt] Change [kt] Change [kt] Change Austria % 14 37% 12 45% Belgium % 21 25% 20 29% Bulgaria % 35 32% 30 41% Cyprus % 1 53% 1 67% Czech Rep % 28 25% 23 37% Denmark % 20 36% 16 49% Estonia % 8 58% 7 65% Finland % 22 28% 19 39% France % % % Germany % 85 31% 81 34% Greece % 35 37% 30 46% Hungary % 24 16% 20 30% Ireland % 8 41% 7 48% Italy % % 93 35% Latvia % 16 16% 12 37% Lithuania % 11 16% 9 31% Luxembourg % 2 45% 1 50% Malta % 0 59% 0 68% Netherlands % 14 44% 13 47% Poland % % 81 35% Portugal % 63 39% 59 43% Romania % % 97 37% Slovakia % 11 48% 10 50% Slovenia % 6 31% 5 50% Spain % 92 33% 85 38% Sweden % 17 39% 17 41% UK % 53 41% 49 46% EU % % % kg/person/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 62: Per capita emissions of PM2.5 in 2005 and 2020 (kg/person/year) 45

54 Volatile organic compounds (VOC) emissions In 2005, solvent use was responsible for 37% of total VOC emissions. Road vehicles and off road mobile sources contributed 23% and 9%, respectively, and industrial sources 16%. Existing legislation is expected to cut total emissions from road vehicles by 80% and from off road sources by 70%. VOC emissions will decline by 40% EU legislation will cut baseline VOC emission by about 40% up to 2030, with the largest decline before 2020 (Figure 63). Million tons VOC/year Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 63: VOC emissions of the EU 27 by SNAP sector Emission legislation for vehicles will deliver 50% of the total emission reduction Emission laws for vehicles will cut VOC emissions from this sector by more than 80% and thereby deliver 50% of the total emission reduction in the EU 27. In contrast to NO x, no major implementation failure of current emission legislation for mobile sources is observed for VOC (Table 23). Legislation for solvents should reduce emissions from this activity by about 20% up to 2030, while no significant changes are expected for emissions from industrial sources (Figure 64). As a consequence of the tight regulations on emissions from mobile sources, about 50% of the remaining emissions will come from solvent use. Emissions relative to % 160% 140% 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 64: Sectoral trends in baseline emissions of VOC in the EU 27 Table 23: Baseline emissions of VOC in the EU 27 by SNAP sector (kilotons) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum

55 ` Per capita VOC emissions will converge across countries By 2030, countries with highest per capita emissions in 2005 will enjoy the largest emission reductions (up to 60%). In other countries where per capita emissions are lower, VOC will decline by only 20% (Table 24). This will lead to convergence of per capita emissions within the EU 27 at an average that is below the lowest country level in 2005 (Figure 65). Table 24: Baseline emissions of VOC (absolute amounts [kt] and change to the year 2005) [kt] [kt] [kt] Change [kt] Change [kt] Change Austria % % % Belgium % % % Bulgaria % 86 36% 71 47% Cyprus % 5 46% 5 51% Czech Rep % % % Denmark % 75 43% 65 50% Estonia % 22 42% 17 54% Finland % 96 37% 85 44% France % % % Germany % % % Greece % % % Hungary % % 96 39% Ireland % 51 26% 50 27% Italy % % % Latvia % 49 29% 40 42% Lithuania % 54 33% 50 37% Luxembourg % 6 58% 6 61% Malta % 3 28% 3 24% Netherlands % % % Poland % % % Portugal % % % Romania % % % Slovakia % 57 17% 55 20% Slovenia % 32 26% 26 42% Spain % % % Sweden % % % UK % % % EU % % % kg VOC/person/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 65: Per capita emissions of VOC in 2005 and 2020 (kg/person/year) 47

56 Ammonia (NH 3 ) emissions Agriculture is the dominating sources of NH 3 emissions in Europe. In 2005, more than 90% of the NH 3 emissions in the EU originated from this sector. 44% of agricultural emissions were caused by cattle farming, 20% originated from pigs, and 205 from the use and production of fertilizer. In contrast to the other air pollutants, only minor changes are expected for NH 3 emissions Although NH 3 emissions are subject to targeted controls in the agricultural sector, and will be affected as a side impact of emission legislation for road transport (i.e., of improved catalytic converters), only slight changes in total emissions in the EU 27 are expected in the long run (Figure 66; Table 25). While emissions declined between 2000 and 2010, they are expected to return to the 2005 level by kilotons NH3/year Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 66: NH 3 emissions of the EU 27 by SNAP sector Emissions from cattle farming remain constant Although some legislation on NH 3 emissions has been established for the agricultural sector in the EU, they will not lead to substantial changes in total agricultural emissions. Emissions from fertilizer use are expected to increase by about 10%; emissions from pigs and chicken would recover to the 2005 levels, while only minor changes are anticipated for the largest source category, i.e., for cattle farming. In this context, the 50% decline in emissions from road transport will not have large impacts for national total emissions, as they contributed only three percent in 2005 (Figure 67). Emissions relative to % 120% 100% 80% 60% 40% 20% 0% Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvent use Road transport Non road mobile Waste treatment Agriculture Figure 67: Sectoral trends in baseline emissions of NH 3 in the EU 27 Table 25: Baseline emissions of NH 3 in the EU 27 by SNAP sector (kilotons) Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum

57 ` Implementation of advanced emission control legislation will lead to substantially lower NH 3 emissions in some Member States There are large differences in the projected evolution of NH 3 emissions across Member States; these are explained, in addition to the ongoing shifts in production, by differences in the stringency of national emission control legislation. For instance, in Denmark as a country with strict regulation, emissions are expected Table 26: Baseline emissions of NH 3 (kilotons and change relative to 2005) to decline by 23%, while increases of more than 20% are anticipated for Poland, Latvia and Lithuania (Table 26). On a per capita basis, there are large disparities across Member States There are also large differences in the per capita emissions across the EU due to different structures of the agricultural sector. For instance, per capita emissions of Ireland were 3.5 times above the EU average in 2005 (Figure 68) [kt] [kt] [kt] Change [kt] Change [kt] Change Austria % 64 5% 67 10% Belgium % 72 2% 75 1% Bulgaria % 66 2% 68 5% Cyprus % 6 10% 6 7% Czech Rep % 66 16% 66 17% Denmark % 54 25% 56 24% Estonia % 12 2% 14 15% Finland % 30 12% 29 14% France % 627 4% 636 2% Germany % 589 0% 593 0% Greece % 53 6% 53 6% Hungary % 75 3% 75 3% Ireland % 108 6% 111 3% Italy % 394 3% 409 0% Latvia % 16 26% 19 49% Lithuania % 52 17% 54 23% Luxembourg % 7 6% 7 16% Malta % 2 0% 2 3% Netherlands % 135 1% 142 6% Poland % % % Portugal % 71 2% 74 2% Romania % 153 5% 155 4% Slovakia % 23 18% 22 23% Slovenia % 19 2% 19 2% Spain % 363 0% 379 4% Sweden % 48 10% 48 11% UK % 302 5% 306 4% EU % % % kg NH3/pesron/year AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 68: Per capita emissions of NH 3 in 2005 and 2020 (kg/person/year) 49

58 50

59 ` Scope, costs and economic implications of further measures 51

60 The scope for further emission reductions The GAINS model contains a database on measures that could bring emissions down beyond the baseline projections. All these measures are technically feasible and commercially available, and the GAINS model estimates for each country to scope for application in addition to the measures that are mandated by current legislation. The Maximum Technically Feasible Reduction (MTFR) scenario explores to what extent emissions of the various substances could be further reduced beyond what is required by current legislation, through full application of the available technical measures, without changes in the energy structures and without behavioral changes of consumers. However, the MTFR scenario does not assume premature scrapping of existing capital stock; new and cleaner devices are only allowed to enter the market when old equipment is retiring. The mitigation potential for SO 2 0% MALT CYPR ESTO BULG SKRE GREE LITH HUNG ROMA PORT ITAL UNKI POLA LUXE SPAI SLOV IREL FRAN LATV NETH BELG FINL AUST DENM CZRE GERM SWED EU27 If all currently available emission control measures were fully applied (with the constraints discussed above), SO 2 emissions in the EU 27 could be 35% lower than what is expected for the implementation of current legislation (Figure 69). This would enable an 80% reduction of total SO 2 emissions relative to SO 2 : On top of current legislation, technical measures are available that could reduce EU emissions by another 35% :Agriculture 08:Non road mobile 06:Solvents 04:Industrial processes 02:Domestic 09:Waste management 07:Road transport 05:Fuel extraction 03:Industrial combustion 01:Energy sector Emission reduction potential relative to baselne 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Other SO2 sources Residential and Commercial Industry Other Combustion Refineries Oil fired Power Plants Agricultural Waste burning Industry: Other Processes Industry: Paper and Pulp Conversion Combustion Coal fired Power Plants Figure 70: Mitigation potentials for SO 2 emissions in 2030 on top of current legislation kilotons SO CLE MTFR CLE MTFR Figure 69: SO 2 emissions in 2010, 2020 and 2030, for the current legislation (CLE) and maximum technically feasible reduction (MTFR) cases, EU 27 Overall, the largest mitigation potentials are estimated for emissions from heavy industry; however, potentials in individual Member States show large variations depending on the structures of energy systems and the stringency of existing emission legislation (Figure 70). 52

61 ` The mitigation potential for NO x For NO x, there is scope for further reductions of combustion and process related emissions in some industrial sectors (e.g., glass production). Assuming that the Euro 6 emission standards deliver the envisaged reductions, and ignoring the potential for further tightening of emission standards for mobile sources, relative to 2005 NO x emissions in the EU 27 could be 15% lower in 2020, and 20% in 2030 (Figure 71). As for SO 2, there are significant structural differences across countries (Figure 72). NO x emissions could be further reduced below the baseline level by 15 20% The mitigation potential for PM2.5 Technologies that are currently on the market could cut PM2.5 emissions by up to 40% below the baseline level (Figure 73). kilotons PM :Agriculture 08:Non road mobile 06:Solvents 04:Industrial processes 02:Domestic 09:Waste management 07:Road transport 05:Fuel extraction 03:Industrial combustion 01:Energy sector kilotons NOx :Agriculture 08:Non road mobile 06:Solvents 04:Industrial processes 02:Domestic 09:Waste management 07:Road transport 05:Fuel extraction 03:Industrial combustion 01:Energy sector 0 CLE MTFR CLE MTFR Figure 73: PM2.5 emissions in 2010, 2020 and 2030, for the current legislation (CLE) and maximum technically feasible reduction (MTFR) cases, EU 27 0% ROMA PORT BULG LITH LATV SLOV HUNG ESTO GREE FINL DENM CZRE SKRE SPAI FRAN AUST POLA ITAL MALT BELG SWED GERM NETH UNKI IREL CYPR LUXE EU CLE MTFR CLE MTFR Figure 71: NO x emissions in 2010, 2020 and 2030, for the current legislation (CLE) and maximum technically feasible reduction (MTFR) cases, EU 27 Emission reduction potential relative to baselne 0% 5% 10% 15% 20% 25% 30% 35% 40% CYPR ROMA PORT SKRE HUNG SPAI IREL UNKI GREE ITAL LITH CZRE FRAN BELG MALT BULG ESTO NETH POLA FINL DENM LATV AUST SWED LUXE SLOV GERM EU27 Other NOx sources Conversion Combustion Industry: Lime Industry: Agglomeration plant sinter Power Plants Industry Other Combustion Industry: Cement Industry: Glass Industrial Boilers Figure 72: Mitigation potentials for NO x emissions in 2030 on top of current legislation Emission reduction potential relative to baselne 10% 20% 30% 40% 50% 60% 70% 80% Other PM sources Agricultural Waste burning Residential waste burning Residential and Commercial: Biomass Residential and Commercial: Coal Industry: Other Processes Industry: Aluminum Industry: Fertilizer Production Industry: Iron and Steel Industry: Cement Coal fired Power Plants Figure 74: Mitigation potentials for PM2.5 emissions in 2030 on top of current legislation Primary emissions of PM2.5 could be further reduced by up to 40% The potential differs greatly over the Member States (Figure 74), essentially due to the different relative contributions from small heating devices in the domestic sector to total PM2.5 emissions. 53

62 The mitigation potential for NH 3 Compared to other pollutants, there is less scope for reductions of NH 3 emission in Europe. However, the measures that are considered in the GAINS model could cut total European emissions by about one third (Figure 75); variations across countries are caused by different stringencies of already existing legislation (Figure 76). Although there is less scope for reducing NH 3 emissions, available measures could cut emissions in the EU by about 30% beyond current legislation kilotons NH :Agriculture 08:Non road mobile 06:Solvents 04:Industrial processes 02:Domestic 09:Waste management 07:Road transport 05:Fuel extraction 03:Industrial combustion 01:Energy sector CLE MTFR CLE MTFR Figure 75: NH 3 emissions in 2010, 2020 and 2030, for the current legislation (CLE) and maximum technically feasible reduction (MTFR) cases, EU 27 The mitigation potential for VOC The GAINS model contains practical measures that could, if fully implemented, decrease VOC emissions in Europe by another 35% below the current legislation baseline projection (Figure 77). kilotons VOC Figure 77: VOC emissions in 2010, 2020 and 2030, for the current legislation (CLE) and maximum technically feasible reduction (MTFR) cases, EU 27 About two thirds of the potential emerges from solvent use, and this potential is rather uniformly distributed across Member States (Figure 78). 0% 0 10:Agriculture 08:Non road mobile 06:Solvents 04:Industrial processes 02:Domestic 09:Waste management 07:Road transport 05:Fuel extraction 03:Industrial combustion 01:Energy sector CLE MTFR CLE MTFR LATV ROMA BULG SLOV MALT CZRE HUNG GREE IREL LITH SKRE DENM POLA GERM AUST ESTO PORT FINL SPAI UNKI FRAN ITAL CYPR SWED NETH LUXE BELG EU27 Emission reduction potential relative to baselne 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% SKRE ESTO SPAI FRAN LITH AUST CYPR PORT HUNG GERM ROMA ITAL GREE POLA SLOV LUXE IREL SWED LATV MALT CZRE UNKI FINL BULG DENM BELG NETH EU27 Other NH3 sources Urea Fertilizer Poultry Non Dairy Cattle Agricultural Waste burning Sheep and Goats Pigs Dairy Cattle Figure 76: Mitigation potentials for NH 3 emissions in 2030 on top of current legislation Emission reduction potential relative to baselne 10% 20% 30% 40% 50% 60% 70% Other VOC sources Agricultural Waste burning Sectors falling under the Solvent Directive Other Industries Other Industries: Solvent Use Industry: Chemicals Industry: Oil Production and Distribution Figure 78: Mitigation potentials for VOC emissions in 2030 on top of current legislation 54

63 ` Emission control costs The GAINS model provides for all countries and economic sectors estimates of costs for technical emission control measures that are implied by a given emission control scenario. These estimates account for societal resources that are diverted from productive investments; cost data for specific technologies are taken from the international literature, and include upfront investments, capital costs and operating expenditures. Costs for implementation of the current legislation Costs for the implementation of the emission control measures that are imposed by the current air pollution legislation increase in the EU 27 in absolute terms from 60 billion in 2010 to 98 billion in 2030 following the implementation of already agreed measures (Table 29). In relative terms, air pollution control costs increase from 0.53% of GDP in 2010 to 0.62% in 2020, and then fall up to 2030 as GDP increases. However, there are large differences across countries owing to different economic wealth. Due to their low GDP, emission control costs put significantly higher burden to the new Member States (e.g., up 3% in Bulgaria) than to the old Member States, where costs account to typically 0.5% of GDP (Figure 80). In 2010, 46% of total costs emerge for the control of emissions from road transport, and this share increases to 58% in 2030 with the penetration of Euro 6 standards. Costs in the power sector will fall in absolute and relative terms, from currently about 20% of total emission control costs to 10% in The share of the agricultural sector remains at 2 3% (Figure 79). Table 27: Air pollution emission control costs by SNAP sector, for the current legislation (in million /yr), EU Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport Off road transport Waste treatment Agriculture Sum % of GDP 0.53% 0.62% 0.58% Billion /yr Agriculture Waste management Non road mobile Road transport Solvents Fuel extraction Industrial processes Industrial combustion Domestic Energy sector Figure 79: Air pollution emission control costs for implementation of the current legislation (EU 27) 3.5% 3.0% 2.5% % of GDP 2.0% 1.5% 1.0% 0.5% 0.0% AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 80: Costs for implementation of emission control measures that are laid down in current air pollution legislation, by country, as percentage of GDP 55

64 Costs for further emission control measures The EC4MACS framework also enables estimating costs of additional measures beyond those that are laid down in current legislation. As an extreme case, this report explores costs of the maximum technically feasible emission reductions, i.e., a case in which all currently available technical emission control measures were implemented to the full extent, considering constraints imposed by the natural turnover of the capital stock. Thus, the scenario assumes application of the most efficient emission control measures to all new emission sources, but does not consider premature scrapping of existing capital stock before the end of its technical life time. While earlier analysis with the GAINS model has pointed out that such a hypothetical scenario is certainly not a cost effective step towards further environmental improvements, future implementation of all additional measures that are available but not required by current legislation would imply costs of about 40 billion /yr, i.e., around 0.25% of GDP (Table 28). These costs would occur on top of the current legislation costs of ~0.6 %/GDP. Again, these additional measures would pose rather different burdens to the economies in the difference Member States. While for most of the old Member States costs of these measures account for % of GDP, they would require up to 3% in some of the new Member States (Figure 82). Additional measures would be most expensive in the domestic sector and for solvents; about two thirds of the costs of the MTFR portfolio would occur in these two sectors. Known measures in the agricultural sector account for about 15% of the total costs (Figure 81). GAINS cost effectiveness analysis) could achieve about 80% of the environmental benefits at only 20% of the costs of the full portfolio. Table 28: Additional air pollution emission control costs by SNAP sector for the maximum technically feasible emission reduction scenario, on top of the costs for the current legislation case (in million /yr), EU Power generation Domestic sector Industrial combustion Industrial processes Fuel extraction Solvents Road transport 0 0 Off road transport Waste treatment Agriculture Sum % of GDP 0.27% 0.24% Billion /yr Agriculture Waste management Non road mobile Road transport Solvents Fuel extraction Industrial processes Industrial combustion Domestic Energy sector Figure 81: Cost for implementation of the Maximum Technically Feasible Emission Control Measures (MTFR), on top of the current legislation (EU 27) However, as it is shown in the recent policy analysis, well chosen sub sets of these measures (e.g., with the 3.5% 3.0% 2.5% % of GDP 2.0% 1.5% 1.0% 0.5% 0.0% AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SL ES SE UK EU Figure 82: Additional costs for implementation of maximum technically feasible emission control measures (on top of current air pollution legislation), by country, as percentage of GDP 56

65 ` Macro economic impacts Costs for the full implementation of all additional emission control measures put significant burdens to economic actors. The GEM E3 model has been employed to assess the macro economic implications of such a hypothetical emission control strategy. Air pollution control expenditures as shares of sectoral value added Expenditures of the MTFR scenario to reduce air pollution were estimated by the GAINS model. Additional investments and operating costs for pollution control equipment account on average for about 0.2% of the value added of the production sectors in the EU 27. Expenditures of the MTFR scenario account for 0.2% of value added in the EU 27 There are, however, large differences between the old Member States with already stringent legislation and comparably high value added, for which the expenditures account to 0.1 to 0.15% of the value added of the production sectors. In contrast, this share can reach up to 1.25% in the new Members States (e.g., Romania, Bulgaria) due to the lower value added (Table 29). For households, emission control costs account for % of the total income (in Sweden, Belgium, Germany, UK, etc.), and up to 3.6% in the new Member States. Table 29: Expenditures for air pollution control measures of the Maximum Technically Feasible scenario, as shares of sectoral value added and households income Production sectors Households Austria 0.20% 0.24% 0.24% 0.18% 0.17% 0.15% Belgium 0.11% 0.14% 0.14% 0.05% 0.05% 0.05% Bulgaria 0.79% 0.83% 0.79% 1.52% 1.26% 1.12% Cyprus 0.22% 0.22% 0.21% 0.05% 0.04% 0.03% Czech Republic 0.38% 0.36% 0.35% 0.63% 0.66% 0.58% Denmark 0.23% 0.23% 0.23% 0.18% 0.16% 0.15% Estonia 0.77% 1.07% 1.13% 2.13% 1.91% 1.65% Finland 0.26% 0.28% 0.27% 0.45% 0.43% 0.41% France 0.16% 0.16% 0.16% 0.30% 0.28% 0.26% Germany 0.16% 0.17% 0.16% 0.06% 0.07% 0.07% Greece 0.27% 0.30% 0.29% 0.20% 0.16% 0.15% Hungary 0.30% 0.32% 0.30% 0.50% 0.45% 0.42% Ireland 0.22% 0.21% 0.20% 0.11% 0.12% 0.10% Italy 0.16% 0.19% 0.19% 0.09% 0.08% 0.09% Latvia 0.94% 0.98% 0.94% 3.61% 3.31% 2.74% Lithuania 1.41% 1.43% 1.41% 1.17% 1.02% 0.96% Luxembourg 0.07% 0.08% 0.08% 0.02% 0.02% 0.01% Malta 0.50% 0.42% 0.36% 0.02% 0.02% 0.02% Netherlands 0.15% 0.17% 0.16% 0.01% 0.01% 0.02% Poland 0.53% 0.56% 0.61% 0.94% 0.85% 0.89% Portugal 0.29% 0.33% 0.31% 0.38% 0.33% 0.28% Romania 1.26% 1.19% 1.19% 2.28% 1.92% 1.74% Slovakia 0.54% 0.56% 0.54% 0.31% 0.31% 0.30% Slovenia 0.14% 0.17% 0.17% 0.21% 0.16% 0.13% Spain 0.20% 0.22% 0.21% 0.12% 0.11% 0.11% Sweden 0.12% 0.15% 0.14% 0.02% 0.02% 0.02% UK 0.11% 0.13% 0.13% 0.05% 0.04% 0.04% EU % 0.21% 0.20% 0.19% 0.17% 0.16% 57

66 Pollution control expenditures reduce productive investments and consumption, but also induce additional demand for new equipment The GEM E3 model considers the purchase of end ofpipe emission control measures by economic agents, which reduces the budget for production or consumption purposes (Figure 83). The MTFR scenario implies highest expenditure for the electricity supply sector, followed by the oil sector and agriculture. Percentage of value added 5% 4% 3% 2% 1% 0% Agriculture Coal Oil Gas Electricity supply Ferrous and non Chemical Products Other energy intensive Electric Goods Transport equipment Other Equipment Consumer Goods Construction Transport Market Services Non Market Figure 83: Air pollution control expenditures by sector as shares of value added for the MTFR scenario (EU27) These expenditures for pollution control do not add to productive investments and do not generate (economic) utility for consumers. However, at the same time, pollution control equipment will have to be produced by the rest of the economy (Figure 84), which will indirectly induce higher activity for the production of the equipment. Ferrous and non ferrous metals 2% Construction 10% Other Equipment Goods 32% Market Services 10% Chemical Products 3% Electric Goods 15% Transport equipment 28% Figure 84: Economic sectors producing air pollution abatement technology Impacts on economic growth The direct effects of emission controls on the EU economy relate to i) demand for materials and equipment for producing the abatement technologies and ii) to lower resources of agents for production and consumption, as part of their budget is spent for purchasing abatement equipment. In addition, costs for production increase. In the MTFR case, the additional demand for abatement technologies amounts to roughly 0.2% of EU s GDP. This demand drives higher activities in different sectors and induces higher investments, especially in sectors that produce emission control equipment. As GEM E3 is a general equilibrium model, it can consistently account for the chain of direct and indirect effects of these changes in the economy, including foreign trade and the impacts of re adjusted costs and prices. For example, higher costs incurring for production sectors owing to the purchasing of abatement equipment affect foreign trade. Thus, exports tend to decrease and imports tend to increase. Such effects are not present in the rest of the world regions, as it is assumed that the EU unilaterally undertakes the pollution reduction policy. The simulation assumes that the EU current account as percentage of GDP remains unchanged in the pollution reduction scenario, compared to the Reference projection; this is assumed for comparability reason. For this purpose, the model simulates re adjustment of the EU basic interest rate. The maximum emission control scenario would reduce GDP growth by 0.11% 0.14% Overall, the GEM E3 model suggests for the MTFR case GDP in the EU 27 to be between 0.11% and 0.14% lower than in baseline scenario (for comparison, GDP is assumed to increase by 50% between 2005 and While the difference is small in magnitude, the net effects of further pollution controls reduction on the EU economy are estimated negative. The impacts of cost increases and resource deviation are found higher than the effects by the demand pulled by abatement equipment. However, this analysis only quantifies the impacts on economic performance, but does not consider the health and environmental benefits of air pollution control (e.g., increased productivity of labor from lower morbidity and of agricultural production due to lower ground level ozone). 58

67 ` Changes in production and employment The main factor driving the GDP decline is the reduction in household consumption (Table 30). Table 30: Macro economic impacts of the MTFR scenario compared to the baseline case GDP 0.14% 0.12% 0.11% Investments 0.04% 0.04% 0.04% Private consumption 0.25% 0.20% 0.18% Exports +0.06% +0.06% +0.05% Imports +0.05% +0.05% +0.05% Reduced household consumption emerges as the major macro economic impact of stricter air pollution controls Households need to use part of their income for purchasing abatement equipment that does not generate utility and reduces consumption of other goods and services. Private consumption is also negatively affected by the general rise of domestic prices, as production sectors face higher costs owing to anti pollution expenditures. Again, these estimates do not consider the welfare gains from air pollution reduction on human health, e.g., on lower expenditures for health care and other feedback of reduced pollution. No major net impacts on employment are expected from the implementation of the MTFR air pollution controls In GEM E3, the labor market is modeled to be flexible, so that real wages are not readjusted upwards to recuperate lost purchasing power of income. Labor participation will increase as households seek to restore their real incomes by higher employment. The decrease in real wages offsets negative effects on employment from the lower demand for labor due to lower activity. Hence, total employment is found almost unchanged from the baseline case, mostly as a result of labor market readjustments (Table 32). Should the labor market be more rigid, for example when wages are indexed to the consumer price index, negative effects on employment would be encountered. Some sectors win, some loose Lower domestic demand and lower consumption lead to lower total productive investments compared to the baseline. However, sectors with higher demand from the production of emission abatement equipment increase their investments. The energy sector, transport, agriculture and consumer goods industries see lower activities compared to the baseline (Table 31). In contrast, sectors that produce pollution control equipment increase their production. In particular, production of other equipment goods and electrical goods grow by 0.3% and 0.4% more, which leads to positive indirect effects for the intermediate goods industries. The negative effects found for the energy sectors can be attributed to the increase in production costs, due to high abatement costs in these sectors, and the ensuing decrease in energy demand. Table 31: Changes in domestic production (% changes from reference case) in the EU 27 as a consequence of the implementation of the maximum technically feasible emission control measures Agriculture 0.11% 0.11% 0.10% Coal 0.96% 0.88% 0.89% Oil 0.17% 0.16% 0.15% Gas 0.50% 0.50% 0.52% Electricity supply 0.77% 0.73% 0.70% Metals production 0.18% 0.17% 0.16% Chemical products 0.01% 0.03% 0.03% Oth. energy intens. 0.01% 0.00% 0.00% Electric goods 0.41% 0.35% 0.32% Transport equip. 0.33% 0.33% 0.30% Other equipment 0.30% 0.29% 0.27% Consumer goods 0.04% 0.04% 0.04% Construction 0.08% 0.06% 0.06% Transport 0.01% 0.01% 0.01% Market services 0.02% 0.01% 0.01% Non market services 0.04% 0.04% 0.03% Table 32: Impacts on employment in the EU Agriculture 0.18% 0.17% 0.15% Coal 1.16% 1.04% 1.04% Oil 0.38% 0.31% 0.28% Gas 1.20% 1.11% 1.05% Electricity supply 1.20% 1.11% 1.03% Metals production 0.30% 0.28% 0.25% Chemical products 0.01% 0.04% 0.04% Oth. energy intens. 0.02% 0.02% 0.01% Electric goods 0.55% 0.45% 0.40% Transport equip. 0.51% 0.48% 0.43% Other equipment 0.50% 0.48% 0.45% Consumer goods 0.01% 0.00% 0.00% Construction 0.10% 0.08% 0.08% Transport 0.04% 0.03% 0.02% Market services 0.04% 0.02% 0.02% Non market services 0.07% 0.06% 0.05% 59

68 There are important differences across Member States At the Member State level, differences between the MTFR and the baseline scenarios are larger, essentially for two reasons: First, despite common EU legislation, current emission control legislations differ across countries. Thus, countries with currently more stringent legislation face lower additional MTFR costs (e.g., Sweden) than countries with less stringent frameworks. Further, these additional expenditures constitute different shares in relation to the overall economic wealth; e.g., in the new Member States with lower GDP, air pollution control costs have larger impacts than in the richer countries (Table 33; Figure 85). Table 33: Impacts of the implementation of MTFR emission control measures on GDP, investments and private consumption (in % difference to the baseline scenario) GDP Investments Consumption Austria Belgium Bulgaria Cyprus Czech Rep Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK % Austria Belgium Bulgaria Cyprus Czech Rep. Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK 0.0% Difference to baseline 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% % Figure 85: Difference in GDP for the maximum control case in 2020, 2025 and 2030 compared to the baseline scenario 60

69 ` Impacts of air pollution 61

70 Health impacts The preceding analyses clearly demonstrate that the dynamics of economic development combined with progressing implementation of dedicated climate and air pollution control legislation will change the current levels of harmful emissions in the coming decades. The baseline scenario suggests only moderate changes in greenhouse gas emissions, but drastic declines are anticipated for most air pollutants. These changes will have impacts on climate, air quality, human health, crops, forests, fisheries, materials and the natural environment in Europe. While climate impacts of the anticipated changes in European greenhouse gas emissions will only occur in the long run and will be strongly compounded with emission developments in other world regions, emission changes within the EU will have direct impacts on air quality and the resulting environmental effects in the near term. are considered by experts in the fields of health, ecology, materials science, etc. to be damaging even at very low concentrations. The overall approach for quantification of effects is known as the impact pathway approach (Figure 86), a straightforward logical progression from emission and pollutant dispersion to quantification of impacts and, where appropriate, monetary damage. The EC4MACS model toolbox provides a comprehensive assessment of the health and ecosystems impacts of air pollution Particular concern is focused on the effects of airborne particles, ozone and the deposition of pollutants that have acidic or nutrifying properties. The particles of concern may be emitted directly from factories, homes, vehicles, etc., or formed from other pollutants by chemical reactions in the atmosphere. These pollutants Figure 86: The impact pathway approach The EC4MACS health impact assessment methodology A wide range of scientific studies reveals significant health impacts from the exposure to fine particulate matter (Box 17). EC4MACS estimates mortality and morbidity effects in the European population that are attributable to current and future exposure to PM2.5 in ambient air. The health impact assessment methods used in EC4MACS are informed principally by a review carried out on behalf of the WHO under the EU s Clean Air For Europe (CAFE) Programme and work led by the Institute of Occupational Medicine under various studies including the ExternE research project as well as CAFE (see Hurley et al., 2005). For the EC4MACS Final Assessment, the improved methodology considers now also the contribution of secondary organic aerosols to ambient PM2.5 concentrations. Box 17: Health effects of fine particles The effects of inhaling particulate matter have been widely studied in humans and animals and include asthma, lung cancer, cardiovascular issues, and premature death. The size of the particle is a main determinant of where in the respiratory tract the particle will come to rest when inhaled. Because of the size of the particles, they can penetrate the deepest part of the lungs. Particles smaller than 2.5 micrometres, PM2.5, tend to penetrate into the gasexchange regions of the lung, and very small particles (<100 nanometres) may pass through the lungs to affect other organs. There is indication that PM2.5 leads to high plaque deposits in arteries, causing vascular inflammation and atherosclerosis a hardening of the arteries that reduces elasticity, which can lead to heart attacks and other cardiovascular problems. Researchers suggest that even short term exposure at elevated concentrations could significantly contribute to heart disease. 62

71 ` Premature mortality from exposure to fine particles For the year 2000, it is estimated that exposure to fine particles in ambient air has shortened statistical life expectancy of European citizens by 10 months on average, and by more than 13 months in industrial areas, especially in the New Member States (Table 34. The decline in precursor emissions of PM2.5, i.e., of primary particles, SO 2, NO x NH 3 and VOC, expected in the coming decades from the implementation of current legislation would reduce the loss in life expectancy to 5.2 months in 2020, and to 4.8 months in In addition, the envisaged improvements in air quality are calculated to lower infant mortality by about 350 cases/year. Implementation of the additional emission control measures of the Maximum Technically Feasible Reduction (MTFR) scenario would yield additional reductions in ambient PM2.5 levels below what is expected for the current legislation baseline. These improvements would gain another 1.3 months of life expectancy and thereby about 60 million life years for the European population. In industrial areas, these measures would increase life expectancy by up to three months. Table 34: Premature mortality from the exposure to fine particulate matter Baseline MTFR Baseline MTFR Loss in statistical life expectancy (months) Years of life lost (million years) Infant mortality (cases/year) Baseline 2030 Baseline Improvement offered by MTFR in 2030 Figure 87: Loss in statistical life expectancy attributable to exposure to PM2.5 from anthropogenic sources (in months). 63

72 Table 35: Loss in statistical life expectancy attributable to the exposure to fine particulate matter from anthropogenic sources by Member State (months) Baseline MTFR Baseline MTFR Austria Belgium Bulgaria Cyprus Czech Rep Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU Morbidity impacts from the exposure to fine particles In addition to earlier deaths, human exposure to fine particulate matter causes various morbidity effects such as bronchitis, hospital admissions, days of restricted activity and lesser symptoms ( the magnitude of impacts, with the numbers for lesser effects (e.g., use of respiratory medication) much greater than those for more severe impacts (bronchitis and hospital admissions). Table 36). For most effects, effect estimates decline by 50% between 2000 and There is a clear logic to Table 36: Morbidity impacts in the EU 27 related to PM 2.5 exposure Baseline MTFR Baseline MTFR Chronic Bronchitis (27yr +) Cases/yr 225, , , , ,190 93,952 Respiratory Hospital Cases/yr 88,745 61,852 50,238 37,037 46,836 33,817 Admissions (All ages) Cardiac Hospital Admissions Cases/yr 54,732 38,147 30,984 22,842 28,886 20,856 (All ages) Restricted Activity Days ( days/yr 486, , , , , ,538 64yr) Respiratory medication use 1000 days/yr 5,751 3,472 2,833 2,100 2,564 1,866 (children 5 14yr) Respiratory medication use 1000 days/yr 39,879 28,734 23,524 17,342 22,033 15,901 (adults 20yr +) Lower Respiratory Symptom 1000 days/yr 265, , ,340 95, ,118 84,721 days (5 14yr) LRS among adults (15yr +) 1000 days/yr 411, , , , , ,

73 ` Health effects from exposure to ground level ozone There is scientific evidence that short and long term exposure to ground level ozone causes additional morbidity and mortality effects, which are independent from the impacts caused by fine particulate matter (Box 18). Box 18: Health impacts from ground level ozone The main health effects of short term (a few hours) exposure to ozone include irritation of throat and eyes, coughing, wheezing, inflammation of lungs and difficulties in breathing. Ozone at high concentrations can cause lung inflammation (irritation) even after only a few hours of exposure. Airways respond to the exposure by covering the affected areas with fluid and by contracting the lung muscles. Breathing becomes more difficult and lung capacity decreases. The lungs will usually recover within a few days after exposure to elevated concentrations of ozone. However, if ozone exposure is experienced over a longer period of time or on a number of repeated occasions within a year, chronic damage to lung tissue may occur. This means that the lung function may be affected and the lining may lose some of its ability to serve as a protective barrier against microbes, harmful chemicals and allergens. There is little evidence from short term epidemiological studies to suggest a threshold ozone concentration at the population level. Long term studies do not indicate a threshold either. It is estimated that the reduction of ozone precursor emissions in Europe between 2000 and 2010 has avoided about 3900 cases of premature deaths per year due to short term exposure to ozone. For 2020, the EC4MACS baseline suggests a further decline by about 5000 cases/year, so that in total the number of premature deaths will be about 30% lower than in the year By 2030, the baseline results in a 37% decline relative to 2000, or 18,800 premature deaths per year in absolute terms. Deaths linked to short term (acute) exposure to ozone in Europe are shown in Table 38, for each country in 2000, 2020 and These future improvements in the EC4MACS final assessment are more optimistic than earlier calculations, mainly due to different expectations on the development of hemispheric background ozone levels. Back in the early 2000s, assumptions on future hemispheric background ozone were based on extrapolations of the increasing ozone trends observed in Mace Head at the west coast of Ireland. Based on findings of the LRTAP Task Force on Hemispheric Transport of Air Pollution (HTAP), the current calculations adopt a more optimistic perspective about the future development and assume only little changes in the hemispheric ozone background around Europe for the coming decades. This more optimistic perspective offers also an enlarged scope for further reductions of health impacts from ozone. It is estimated that further technical measures within the EU could reduce premature mortality by 45 50% below the 2000 level (Table 37). However, new research indicates potentially significant impacts on premature mortality from long term exposure to ozone. As the WHO REVIHAAP project has not yet provided advice on this issue, this effect is not yet quantified for the TSAP 2012 analyses. It is notable that results for hospital admissions and for use of respiratory medication (though not for days of restricted activity) are of a similar order of magnitude to the results for PM 2.5. The geographical distribution is shown in Figure 88. Table 37: Health impacts related to ozone exposure Baseline MTFR Baseline MTFR Premature mortality from short term exposure Cases/yr 33,065 25,541 20,636 17,329 18,994 15,260 Respiratory hospital admissions, over 65 years Cases/yr 26,565 24,706 23,835 20,075 25,870 20,856 Minor restricted activity days, working adults 1000 days/yr 84,984 67,857 53,466 45,014 47,213 38,039 Respiratory medication use (adults) 1000 days/yr 28,907 24,233 20,302 17,094 19,103 15,

74 Figure 88: The health relevant ozone indicator (SOMO35) for the year 2000 (left panel) and the baseline emissions of 2020 [ppb.days] Table 38: Premature mortality attributable to the exposure to ground level ozone [cases of premature mortality] Baseline MTFR Baseline MTFR Austria Belgium Bulgaria Cyprus Czech Rep Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU

75 ` Vegetation and ecosystems impacts Air pollution threatens the productivity and sustainability of vegetation and ecosystems in a variety of ways; however, not all damage mechanisms are fully understood, and it is often difficult to estimate damage in quantitative terms. Methodology and key issues To draw an as complete a picture as possible of the multiple undesired effects of air pollution, EC4MACS addresses a wide range of impacts to vegetation and ecosystems based on state of the art scientific understanding, to derive quantitative indicators of the benefits that result from emission control measures. Threat to biodiversity Increased content of nitrates in soil frequently leads to undesirable changes in vegetation composition, and many plant species are endangered as a result of eutrophication in terrestrial ecosystems, such as the majority of orchid species in Europe (Box 19). Ecosystems (like some meadows, forests, and bogs that are characterized by low nutrient content and speciesrich, slowly growing vegetation adapted to lower nutrient levels) are overgrown by faster growing and more competitive species poor vegetation, like tall grasses, that can take advantage of unnaturally elevated nitrogen levels and the area may be changed beyond recognition and vulnerable species may be lost. For example, species rich fens are overtaken by reed or reed grass species, and forest undergrowth affected by runoff from a nearby fertilized field can be turned into a thick nettle and bramble shrub. Box 19: Threat to biodiversity from eutrophication Chemical forms of nitrogen are most often of concern with regard to eutrophication because plants have high nitrogen requirements so that additions of nitrogen compounds stimulate plant growth (primary production). This is also the case with increased levels of phosphorus. Nitrogen is not readily available in soil because N 2, a gaseous form of nitrogen, is very stable and unavailable directly to higher plants. Terrestrial ecosystems rely on microbial nitrogen fixation to convert N 2 into other physical forms (such as nitrates). However, there is a limit to how much nitrogen can be utilized. Ecosystems receiving more nitrogen than the plants require are called nitrogen saturated. Saturated terrestrial ecosystems contribute both inorganic and organic nitrogen to freshwater, coastal, and marine eutrophication, where nitrogen is also typically a limiting nutrient. While a full quantification of the threats to biodiversity is complex, indicators can be used to characterize current and future risks of loss in biodiversity. In particular, risks of excess nutrient deposition can be represented via the effect on soil chemistry using the exceedance of computed critical loads (Box 20). Alternatively, new methods have been developed to quantify the impacts of excess deposition on plant species diversity. The CCE Environmental Impact Assessment (CCE EIA) methodology establishes protection levels against adverse effects (eutrophication, acidification and ground level ozone) of air pollution on ecosystems in European countries, which should safeguard the risk to ecosystem services. These protection levels are called critical loads and levels. Box 20: Critical loads and the quantification of the impacts of excess exposure Over the last years, such critical loads have been modelled and mapped for European ecosystems, and under EC4MACS extended to Natura2000 areas and to biodiversity indicators (Hettelingh et al. 2009). Input data to compute critical loads have been submitted by a network of collaborating National Focal Centres in 30 countries operating under the effect based programme of the Convention on Long range Transboundary Air Pollution (LRTAP). However, the quantitative understanding of the biological effects that occur once these critical loads are exceeded is limited as these vary as function of a number of elements including soil type, meteorology, land use and species occurrence. This is especially important from the point of view of risks caused by excessive deposition of nitrogen. The risk of nitrogen inputs in the environment is increasing and well documented (Galloway et al. 2008). Despite the lack of operational vegetation models on a European scale, improved knowledge is urgently required of the relationship between excessive nitrogen deposition and effects on biodiversity in Europe. Therefore, under EC4MACS, the CCE has extended the European application of computed critical loads to so called empirical critical loads of nutrient nitrogen, derived from nitrogen addition experiments both in the field and in the laboratory. 67

76 As a quantitative indicator of the impacts of excess nitrogen deposition on biodiversity in nature protection areas, the EC4MACS project developed a methodology to assess for an ecosystem the richness in species as a function of nitrogen deposition in excess of the critical loads (Box 21). Box 21: Estimating the impacts of nitrogen deposition on the species richness of ecosystems As a quantitative estimate of the change of biodiversity in future scenarios of nitrogen deposition, the species richness of (i) (semi ) natural grasslands (EUNIS class E) (ii) arctic and (sub ) alpine scrub habitats (EUNIS class F2) and (iii) under storey vegetation as defined by the Sorensen similarity index of coniferous boreal woodlands (EUNIS class G3 A C) is calculated. For these three vegetation classes, dose response curves (Bobbink and Hettelingh 2011) are applied to the European harmonized land cover map). Acidification Forest soil biology and chemistry can be seriously damaged by acid rain. Some microbes are unable to tolerate changes to low ph and are killed. The enzymes of these microbes are denatured (changed in shape so they no longer function) by the acid. The hydronium ions of acid rain also mobilize toxins such as aluminium, and leach away essential nutrients and minerals such as magnesium. Soil chemistry can be dramatically changed when base cations, such as calcium and magnesium, are leached by acid rain thereby affecting sensitive species. Box 22: Modelling threats to ecosystems from acidification As an indicator for the threat of acidification for terrestrial and aquatic ecosystems, excess deposition above site specific critical loads is estimated for current and future emission scenarios. Critical loads for acidification are computed using a steady state modelling approach. Thereby, a time independent equilibrium is assumed between acidifying and acidity buffering processes in the root zone of vegetation that is classified according to the European Information System. A critical load value is then computed by using acceptable limit values for specific soil chemistry related indicators (such as the ratio between base cations and the soil solution concentration of aluminium). More detail on the steady state modelling approach and acceptable limit indicators and values has been documented in the mapping manual ( The European database of critical loads, held at the RIVM Coordination Centre for Effects (CCE), brings together data provided by National Focal Centres of the EU Member States. For countries that did not provide data, critical loads were computed by the CCE based on its European background database. The EU 27 part of the European database covers about 2 million km 2 of ecosystem area. Both the lower ph and higher aluminium concentrations in surface water that occur as a result of acid rain can cause damage to fish and other aquatic animals. At ph lower than 5, most fish eggs will not hatch and lower ph can kill adult fish. As lakes and rivers become more acidic biodiversity is reduced. Acid rain has eliminated insect life and some fish species, including the brook trout in some lakes, streams, and creeks in geographically sensitive areas. However, the extent to which acid rain contributes directly or indirectly via runoff from the catchment to lake and river acidity (i.e., depending on characteristics of the surrounding watershed) is variable. Vegetation damage from ozone Exposure to ground level ozone causes negative effects on forest trees such as reduced photosynthesis, premature leaf shedding and growth reductions. These effects potentially have important negative consequences for carbon sequestration, biodiversity and other ecosystem services provided by forests trees such as reducing soil erosion and decreasing flooding and avalanches. Ozone sensitive forest tree species including birch, beech, Norway spruce, Sessile oak, Holm oak, and Aleppo pine are present across large areas of Europe. Effects have already been detected in ambient ozone in Europe, for example, visible injury has been detected on forest trees in ICP Forests surveys, reduced stem growth has been reported in Switzerland and leaf loss has been reported in Greece. Flux based critical levels have been derived for beech, birch and Norway spruce and are applicable for protection against loss of carbon storage and environmental protection by trees. Box 23: Modelling ozone damage to vegetation Evidence on vegetation damage from ozone in ambient air has been documented for 18 European countries, from Sweden in the north to Greece in the south (Mills et al. 2011). Most ozone sensitive crops include wheat, soybean, pulses and tomato, with potato, sugar beet, rape and maize being moderately sensitive. Grasslands (especially uplands, dry grasslands and woodland fringes), heaths and wetlands are amongst the most ozone sensitive (semi natural) habitats in Europe. A model has been developed to estimate stomatal ozone flux across Europe for a number of important species ((Emberson et al. 2000)). The model calculates ozone flux using European Monitoring and Evaluation Programme (EMEP) model ozone concentrations in combination with estimates of the atmospheric, boundary layer and stomatal resistances to ozone transfer. The model simulates the effect of phenology, irradiance, temperature, vapour pressure deficit and soil moisture deficit on stomatal conductance. Critical levels have been defined for effects of ozone on vegetation that take into account the varying effects of climatic conditions, soil moisture and phenology on the amount of ozone taken up by the stomatal pores. 68

77 ` By impacting on growth, seed production and environmental stress tolerance, ozone affects the vitality and balance of (semi )natural vegetation ecosystems and the ecosystem services they provide including carbon storage, water storage and biodiversity. The floral diversity of this vegetation type makes generalisations on effects and the establishment of critical levels applicable across Europe more difficult than for crops and forest trees. Although a whole canopy flux model has been developed, current fluxbased critical levels for this vegetation type are based on effects on individual species. The ICP Vegetation Evidence Report found that effects of ozone were most widespread across Europe for species of the genus Trifolium (clover species). Ozone exposure experiments have confirmed that these species are amongst the most sensitive to ozone, with reductions in biomass, forage quality and reproductive ability noted at ambient and near ambient concentrations in many parts of Europe. Protection of Natura2000 areas In addition to fragmentation and climate change, excess nitrogen deposition constitutes a major threat to biodiversity in these protected areas, as excess sulphur and nitrogen deposition affect plant species diversity and soil chemistry of ecosystems. Under the Natura2000 programme, the European Union declared specific nature reserve areas to maintain and restore natural habitats listed in the directive at Favourable Conservation Status (Box 24). Box 24: Natura2000 areas The Natura2000 network of nature protection areas has been established under the 1992 Habitats Directive. The aim of the network is to assure the long term survival of Europe's most valuable and threatened species and habitats. It is comprised of Special Areas of Conservation (SAC) designated by Member States under the Habitats Directive, and also incorporates Special Protection Areas (SPAs) which they designate under the 1979 Birds Directive. Natura2000 is not a system of strict nature reserves where all human activities are excluded. Whereas the network will certainly include nature reserves most of the land is likely to continue to be privately owned and the emphasis will be on ensuring that future management is sustainable, both ecologically and economically. The establishment of the network of protected areas also fulfils a Community obligation under the UN Convention on Biological Diversity. States have provided critical loads for their protected areas. Particularly wide spread impacts on biodiversity are currently estimated for the Benelux area, southern Germany, Poland and northern Italy, i.e., areas with high ammonia deposition. Overall, the anticipated reductions in baseline emissions in the EU will lead to improvements in this biodiversity indicator throughout Europe. Implementation of all available emission control measures could eliminate such risks in most areas, although biodiversity will remain threatened in northern Italy. Excessive nitrogen deposition in Natura2000 areas In 2000, most Natura2000 areas in central and southern Europe for which critical loads estimates have been provided were exposed to unsustainable level of nitrogen deposition; nitrogen deposition posed a threat to the biodiversity in almost 80% (or 430,000 km 2 ) of the protected zones in Europe. By 2030, the expected declines in NO x emissions would reduce the threatened area to 62% and would leave about 340,000 km 2 unprotected. Full application of the available measures to reduce ammonia emissions could safeguard biodiversity against excess nitrogen deposition in another 120,000 km 2 of the nature protection areas (Table 39, Figure 90). However, it should be noted that these estimates are incomplete as not all Member Figure 89: Change in biodiversity expressed through species richness in Natura2000 areas for EUNIS classes E, F2 and through abundance in G3. Dark shadings indicate locations in which that change is more than 5%. In Figure 89, dark fields indicate 0.50º 0.25º grid cells where the number of species is estimated to change by more than five percent as a result of nitrogen deposition. The size of the coloured spots is proportional to the total Natura2000 ecosystem area within the respective grid cell (based on (Bobbink and Hettelingh 2011)). 69

78 Baseline 2030 Baseline 2030 MTFR Figure 90: Percent of Natura2000 areas (i.e., protected ecosystems under the Birds and Habitats Directives) which receive nitrogen deposition in excess of their critical loads for eutrophication. For grey shaded areas, no critical loads for protected zones have been provided by National Focal Centres. Table 39: Natura2000 areas with nitrogen deposition in excess of the critical loads for eutrophication Natura2000 areas where critical loads for eutrophication are exceeded (1000 km 2 ) CLE MFR CLE MTFR Percentage of Natura2000 areas where critical loads for eutrophication are exceeded CLE MFR CLE MTFR Austria Belgium Bulgaria Cyprus % 100% 100% 100% 100% 100% Czech Rep % 86% 75% 47% 70% 37% Denmark % 100% 99% 99% 99% 95% Estonia % 42% 31% 12% 30% 9% Finland % 4% 2% 1% 2% 1% France % 80% 72% 37% 66% 31% Germany % 53% 48% 33% 46% 30% Greece % 98% 97% 93% 97% 84% Hungary % 97% 92% 68% 86% 68% Ireland % 21% 20% 5% 20% 5% Italy % 58% 46% 23% 42% 18% Latvia % 92% 85% 56% 85% 51% Lithuania % 97% 96% 86% 96% 83% Luxembourg % 100% 97% 93% 94% 86% Malta Netherlands % 89% 88% 82% 88% 82% Poland Portugal % 100% 99% 88% 99% 83% Romania % 92% 90% 74% 89% 67% Slovakia % 93% 86% 76% 84% 70% Slovenia % 66% 36% 4% 22% 1% Spain % 98% 97% 87% 97% 81% Sweden % 40% 25% 16% 23% 12% UK EU % 71% 65% 46% 62% 41% 70

79 ` Risks for biodiversity of all ecosystems Excessive deposition of nitrogen threatens biodiversity not only the areas that receive special protection under the Natura2000 program, but also at the larger scale to many European ecosystems (Table 40). about 230,000 km 2 in 2030, but will still leave more than half of European ecosystems at risk. However, additional measures, especially for ammonia emissions, could protect another 275,000 km 2. It is estimated that the decline in NO x emissions of the baseline scenario will lower the unprotected area by Table 40: Ecosystems area with nitrogen deposition in excess of the critical loads for eutrophication (absolute area and percentage) Ecosystems area where critical loads for eutrophication are exceeded (1000 km 2 ) Percentage of ecosystems area where critical loads for eutrophication are exceeded CLE 2020 MFR 2030 CLE 2030 MTFR CLE 2020 MFR 2030 CLE 2030 MTFR Austria % 66.8% 52.1% 19.3% 47.4% 14.8% Belgium % 1.1% 0.7% 0.0% 0.6% 0.0% Bulgaria % 61.0% 44.4% 24.2% 38.3% 17.9% Cyprus % 100.0% 100.0% 100.0% 100.0% 100.0% Czech Rep % 89.9% 82.5% 58.1% 79.0% 46.6% Denmark % 99.7% 99.3% 97.4% 99.3% 94.8% Estonia % 31.4% 20.9% 9.4% 20.2% 8.0% Finland % 7.8% 4.4% 1.7% 3.3% 1.2% France % 82.3% 74.5% 41.2% 69.6% 34.9% Germany % 52.9% 47.8% 34.1% 46.2% 30.8% Greece % 98.2% 97.4% 91.1% 96.2% 84.7% Hungary % 96.1% 90.0% 66.6% 84.7% 66.0% Ireland % 19.1% 17.0% 4.5% 17.7% 4.8% Italy % 60.4% 48.6% 26.6% 45.3% 21.4% Latvia % 90.7% 83.2% 53.4% 82.6% 49.8% Lithuania % 98.1% 97.3% 85.1% 97.3% 82.2% Luxembourg % 100.0% 97.9% 95.6% 96.8% 92.5% Malta % 0.0% 0.0% 0.0% 0.0% 0.0% Netherlands % 89.2% 87.9% 82.5% 87.9% 82.1% Poland % 72.5% 68.6% 50.7% 68.9% 49.1% Portugal % 99.8% 99.6% 85.2% 99.6% 78.0% Romania % 95.2% 93.3% 80.6% 92.6% 75.9% Slovakia % 95.0% 89.2% 79.8% 87.1% 73.2% Slovenia % 72.9% 43.6% 4.1% 26.2% 1.5% Spain % 97.5% 96.2% 86.7% 96.0% 82.2% Sweden % 29.7% 21.7% 11.8% 20.0% 8.4% UK % 45.3% 32.0% 11.3% 30.9% 8.0% EU % 61.5% 55.6% 40.8% 53.7% 37.1% Acidification of forest soils In the second half of the 20 th century, European forests were strongly affected by acid deposition, which exceeded over large areas the sustainable levels that have been established through site specific critical loads. However, the steep reductions in SO 2 and NO x emissions that are implied by the baseline projection will profoundly reduce deposition and bring it below critical loads in many European forests. For the year 2000 it is calculated that the critical loads for acidification have been exceeded in a forest area of almost 250, ,000 km 2, i.e., in about 1.4% of the forests within the EU 27 for which critical loads have been reported. Especially the decline in SO 2 emissions has eliminated the threat from acidification of forest soils in 130,000 km 2 up to 2010, and for the baseline scenario the situation will resolve for another 50,000 km 2 up to 2030 (Table 41). Additional measures within the EU could lead to further improvements, so that in % of forests could be protected. Full implementation of these additional measures could achieve protection also in the former black triangle (i.e., in Poland, Czech Republic and the eastern parts of Germany), while residual problems would remain in the Netherlands due to high ammonia density (Figure 91). 71

80 Table 41: Forest area at risk of acidification (absolute area and percentage of total forests) Forest area where critical loads for acidification are exceeded (1000 km 2 ) Percentage of forest area where critical loads for acidification are exceeded CLE 2020 MFR 2030 CLE 2030 MTFR CLE 2020 MFR 2030 CLE 2030 MTFR Austria % 0.0% 0.0% 0.0% 0.0% 0.0% Belgium % 0.7% 0.2% 0.0% 0.1% 0.0% Bulgaria % 0.0% 0.0% 0.0% 0.0% 0.0% Cyprus % 0.0% 0.0% 0.0% 0.0% 0.0% Czech Rep % 7.1% 5.0% 2.0% 4.2% 1.3% Denmark % 1.3% 0.1% 0.0% 0.1% 0.0% Estonia % 0.0% 0.0% 0.0% 0.0% 0.0% Finland % 0.0% 0.0% 0.0% 0.0% 0.0% France % 0.8% 0.4% 0.0% 0.3% 0.0% Germany % 1.3% 0.6% 0.1% 0.4% 0.1% Greece % 0.2% 0.1% 0.0% 0.0% 0.0% Hungary % 0.7% 0.5% 0.1% 0.5% 0.1% Ireland % 0.1% 0.0% 0.0% 0.0% 0.0% Italy % 0.0% 0.0% 0.0% 0.0% 0.0% Latvia % 0.7% 0.4% 0.0% 0.4% 0.0% Lithuania % 3.2% 3.0% 1.8% 3.0% 1.6% Luxembourg % 1.3% 1.2% 0.4% 1.2% 0.0% Malta % 0.0% 0.0% 0.0% 0.0% 0.0% Netherlands % 7.4% 7.1% 6.1% 7.0% 5.9% Poland % 3.7% 2.2% 0.8% 2.1% 0.6% Portugal % 0.1% 0.1% 0.0% 0.1% 0.0% Romania % 0.1% 0.0% 0.0% 0.0% 0.0% Slovakia % 0.6% 0.3% 0.0% 0.2% 0.0% Slovenia % 0.0% 0.0% 0.0% 0.0% 0.0% Spain % 0.0% 0.0% 0.0% 0.0% 0.0% Sweden % 0.8% 0.6% 0.5% 0.6% 0.5% UK % 0.7% 0.4% 0.1% 0.3% 0.1% EU % 0.7% 0.4% 0.2% 0.4% 0.2% Baseline 2030 Baseline 2030 MTFR Figure 91: Percentage of forest area where deposition exceeds the critical loads for acidification 72

81 ` Acidification of surface waters After the decline of SO 2 emissions before 2000, acidification of freshwater bodies remained to be problematic only in a few countries in the EU, i.e., Ireland, Sweden and the UK (Table 42). The 2005 Thematic Strategy on Air Pollution established the target to further reduce the threatened catchment area by 39% between 2000 and Table 42: Freshwater catchment area with acid deposition in excess of the critical loads for acidification (1000 km 2 ) CLE 2020 MFR 2030 CLE 2030 MTFR CLE 2020 MFR 2030 CLE 2030 MTFR Ireland % 3.6% 3.4% 0.6% 3.4% 0.6% Sweden % 27.6% 23.3% 20.3% 22.7% 20.0% UK % 3.1% 2.4% 1.3% 2.2% 0.9% EU % 23.9% 20.1% 17.4% 19.6% 17.1% Risks from ground level ozone to vegetation Risks from ground level ozone to forests Ground level ozone constitutes another risk factor that threatens the vitality of European forests. At the time of writing this report, EC4MACS employs the AOT40 indicator as a risk measure of the threat from ozone exposure. Improvements in the AOT40 indicator are shown in Figure 92. Figure 92: An indicator of ozone damage to forest trees (AOT40) calculated for 2000 (left panel) and 2020 (right panel) As an alternative and potentially more representative indicator of ozone damage to trees, Figure 93 provides changes of the ozone fluxes for deciduous trees that are calculated for the EC4MACs baseline scenario and the Maximum Feasible Reduction case. Highest excess of the critical level occurs in Mediterranean countries (Malta, Italy, Slovenia), although significant excess dose spread over large areas in central Europe. The anticipated reductions of ozone precursor emissions that are estimated in the EC4MACs baseline will reduce ozone fluxes across Europe. However, ozone fluxes are strongly determined by long term exposure during the entire vegetation period and to a lesser extent by ozone peak episodes. The importance of ozone peak level is further diminished as such episodes are often associated with dry periods, when stomata are closed and ozone uptake by plants is reduced. Thus, reductions of European precursor emissions that are most effective for the reduction of peak ozone levels have less impact on long term ozone levels and thereby on the uptake of phytotoxic ozone by plants. Thus, the current analysis suggests that ozone damage to trees will decline to a lesser degree (by about 15%) than European precursor emissions, and will be strongly influenced by the future evolution of hemispheric ozone background levels (Figure 93). 73

82 Phytotoxic ozone dose (mmol/m2) CLE 2030 MTFR 0 Austria Belgium Bulgaria Cyprus Czech Rep. Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU 27 Figure 93: Anticipated trends in ozone fluxes (phytotoxic ozone dose (POD1); mmol/m 2 leaf area) that are harmful for forest trees, for the EC4MACS baseline and MTFR scenarios. The no effect critical level (4 mm/m 2 ) is indicated by the black line Risks from ground level ozone agricultural crops A similar picture emerges for ozone damage to agricultural crops (Figure 94). Emission reductions within Europe are somewhat more effective than for trees, although a large share of the phytotoxic ozone dose is Phytotoxic ozone dose (mmol/m2) connected to long term ozone levels that are strongly determined by the evolution of hemispheric background ozone. It should be noted that these calculations (POD3) refer to non irrigated crops and assume low ozone damage in situations with water deficits. Under irrigated situations, which are wide spread in Mediterranean countries, crop losses from ozone will be much higher CLE 2030 MTFR 0 Austria Belgium Bulgaria Cyprus Czech Rep. Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK EU 27 Figure 94: Anticipated trends in ozone fluxes (phytotoxic ozone dose (POD3); mmol/m 2 leaf area) that are harmful for agricultural crops, for the EC4MACS baseline and MTFR scenarios 74

83 ` Compliance with air quality limit values As one important instrument of clean air policy, EU legislation has established air quality limit values for a range of pollutants. Member States have legal obligations to comply with these limit values everywhere, even at hot spots. While these limit values are not necessarily designed to minimize actual damage of polluted air on human health and ecosystems, they provide important benchmarks for measuring progress towards clean air in Europe. However, even at hot spots, air quality is not only influenced by nearby emission sources, but through the long range transport of pollutants in the atmosphere also from more distant sources, often in other countries. Thus, effective emission control measures must not only address local sources next to the monitoring stations and within the same city, but also emissions in many neighboring countries. Up to now it was difficult to devise a coherent assessment of the impacts of various emission sources at different spatial scales, and to combine the assessment of Europe wide emission control strategies with the scope for local measures, e.g., within urban zones. As an important new feature, EC4MACS developed a methodology to estimate future compliance with air quality limit values for AIRBASE monitoring stations that brings together emission sources from all scales (Box 25). A hybrid downscaling methodology determines for street canyon and hot spot AIRBASE stations the differences in observed concentrations to the measurements at the nearest background observation sites, and relates them to corresponding quantities that can be derived from available models. This makes it possible to modify the contributions of the different source types for future emission control scenarios. Box 25: The EC4MACS methodology for estimating compliance with air quality limit values at hot spot locations In brief, the downscaling methodology determines for street canyon and hot spot AIRBASE stations the differences in observed concentrations to the measurements at the nearest background observation sites, and relates them to corresponding quantities that can be derived from available models. This makes it possible to modify the contributions of the different source types for future emission control scenarios. The methodology is described in more detail in (Kiesewetter et al. 2013). It has been implemented for all AIRBASE stations for which sufficiently representative temporal coverage of all required components was reported and, for NO 2, where in the past annual mean concentrations exceeded 20 μg/m 3. In total, the analysis is carried out for 1843 stations for PM10 and 1174 stations for NO 2. Compliance with PM10 air quality limit values It is estimated that in the coming decades, with very few exceptions, the baseline development in emissions will lead to a far reaching elimination of current noncompliance cases in most of the old EU 15 Member States. However, due to the persistence of solid fuel use for home heating in small stoves in some of the new Member States, exceedances of the PM10 limit values are expected to prevail in urban areas in Poland, Slovakia and Bulgaria (Figure 96). Full application of available end of pipe technical emission control measures could achieve compliance at almost all stations in the old Member States already in However, for urban areas in the new Member States, these end of pipe measures will not be sufficient to eliminate exceedances without dedicated action to substitute solid fuels in the household sector with cleaner forms of energy. For 2000, likely compliance, considering the model uncertainty margin, has been estimated for about 40% of the stations for which the analysis has been carried out. Under the baseline development of emissions, more than 80% of stations would be safely in compliance by 2020, and likely non compliance is estimated for three percent of the stations (Figure 95). This situation will prevail to Stations reporting exceedences in 2009 with sufficient data Number of stations for which PM10 exceedances are modelled Modelled exceedances 2009 Baseline 2020 MTFR 2020 Baseline 2030 MTFR 2030 Figure 95: Computed compliance with the annual mean PM10 limit value at the 1843 AIRBASE stations 75

84 Implementation of the maximum technically feasible emission control measures could reduce the cases of firm non compliance to less than 100 stations in Europe (i.e., about five percent of all stations). As mentioned above, most of these stations are located in the new Member States, where local heating with solid fuels will still make large contributions to urban PM10 levels (Figure 96). Thus, implementation of the available technical measures (e.g., those that are currently discussed under the Products Standards directive) will be an important (and in most cases sufficient) step towards compliance with PM10 air quality limit values in the old Member States. However, in the new Member States, these technical measures alone are unlikely to bring air quality in compliance with the existing limit values within urban areas. Additional efforts, e.g., promotion of fuel switching policies that stimulate the substitution of coal and fuel wood by cleaner forms of energy could alleviate the situation needs further analysis. In this context, targeted use of structural funds of the European Union in economically less developed zones in the New Member States could support the accelerated introduction of cleaner heating devices and thereby make important contributions to clean air policies Baseline 2030 Baseline 2030 MTFR Figure 96: Compliance with PM10 limit values. For compliance assessment, deducts the contribution from natural sources are deducted from the measurements and model calculations. It is assumed an annual mean concentration of 30 μg/m 3 corresponds to 30 daily exceedances of the 50 μg/m 3 daily limit value. grey: <25 μg: compliance with limit values likely blue: μg: compliance uncertain red: >35 μg: compliance unlikely 76

85 ` Compliance with NO 2 limit values A similar analysis has been carried out for NO 2 limit values. As a consequence of the assumptions taken in the baseline scenario, in particular about the effectiveness of Euro 6 standards, a steep drop in total NO x emissions from the transport sector is calculated in Europe. Such a rapid decline would have profound impacts on the compliance with NO 2 limit values, as NO 2 concentrations in street canyons are dominated by the contributions from near by traffic sources. For the period , compliance has been estimated for about one third of the 1174 AIRBASE stations for which the analysis has been carried out (Figure 98. The share of stations with non compliance would then fall to about 6% in 2020 and 1% in 2030 (Figure 97). Implementation of the maximum technically feasible emission control measures (MTFR) could eliminate in 2030 almost all non compliance cases in Europe. It should be noted that the MTFR scenario assumes only Europe wide technical emission control measures, and does not consider the potential for local measures (e.g., traffic restrictions, low emission zones, etc.) Stations reporting exceedences in 2009 with sufficient data Number of stations for which NO2 exceedances are modelled Modelled exceedances 2009 Baseline 2020 MTFR 2020 Baseline 2030 MTFR 2030 Figure 97: Computed compliance with the annual mean NO 2 limit value at the 1174 AIRBASE stations Obviously, this decline in non compliance stations is a direct consequence of the anticipated sharp cut in NO x emissions from traffic sources that has been assumed in the TSAP 2012 baseline, especially from Euro 6 diesel vehicles Baseline 2030 Baseline 2030 MTFR Figure 98: Computed annual mean NO 2 concentrations at AIRBASE monitoring stations: grey dots: <35 μg: compliance with annual limit value likely; blue dots: μg: compliance uncertain; red dots: >45 μg: compliance unlikely 77

86 78

87 ` Benefits of emission controls 79

88 Monetary valuation of benefits While poor air quality compromises many aspects of human health, natural vegetation and agricultural crops, and action to reduce emissions will result in benefits to all these categories, earlier work (e.g. Holland et al, 2011) has revealed highest monetary value of benefits from emission reductions for human health. Quantifications of the monetary benefits from higher Methodology and key issues Within EC4MACS, the ALPHA and ALPHA RiskPoll methodologies have been developed to derive estimates of monetary benefits from air pollution control measures in Europe (Box 26). This methodology builds on earlier work for the original Gothenburg Protocol and EU NEC Directive, which was developed as part of the EU funded ExternE (Externalities of Energy) project during the 1990 ( For the EU s Clean Air For Europe (CAFE) programme, a thorough review of the methods was conducted, widely consulted with stakeholders, and subject to a formal, independent and international peer review. Box 26: The EC4MACS benefits assessment methodology The EC4MACS benefits assessment module quantifies the health and environmental benefits of emission control strategies in monetary terms. It employs the ALPHA and ALPHA RiskPoll tools based on the ExternE methodology, which follows a logical progression through the following stages: Quantification of emissions (in CAFE, covered by the RAINS model); Description of pollutant dispersion across Europe (in CAFE, covered by the RAINS and EMEP models) Quantification of exposure of people, environment and buildings that are affected by air pollution Quantification of the impacts of air pollution Valuation of the impacts Although a wide field of academic research and literature has emerged on the quantification of monetary benefits to human health and ecosystems, a number of issues constitute formidable challenges for a generally accepted approach, and current methods often remain controversial among stakeholders, inter alia, since it often involves value judgments that are not always shared by all stakeholders. Box 27 lists some of the most frequently contested topics in such a benefit assessment, and outlines the approach taken for the EC4MACS analysis. agricultural productivity and improved sustainable conditions resulted in lower numbers, although this is in part a function of the problem of quantifying ecosystem damage/benefits in monetary terms. The analysis presented here is focused primarily on the assessment of health impacts across Europe for the scenarios listed above. Box 27: Key issues for the valuation of benefits The willingness to pay (WTP) concept underpins any economic transaction: without a willingness to pay, the price requested for a good there will be no sale. Approach to valuation of mortality: The traditional method to valuing mortality applies the value of a statistical life (VSL) to an estimate of the number of deaths linked to a particular cause. An alternative applies the value of a life year (VOLY) to the estimated change in life expectancy aggregated across the population. EC4MACS applies the second approach. Mortality valuation estimates: Under the EC4MACS project review has been undertaken of VOLY and VSL research since the earlier work on the CAFE Programme was completed. A particular development concerns the finalisation of the ECfunded NEEDS project (Desaigues et al. 2011). However, it has been concluded that a peer review and further scrutiny of the NEEDS data is necessary before they are factored into the core analysis, bearing in mind the need for consistency in analysis carried out across the European Commission more generally in this area. On this basis the figures agreed under the earlier CAFE work ( 52,000 median/ 120,000 mean for VOLY and 0.98 million median/ 2 million mean for VSL in year 2000 prices) are applied here, though only the extremes are reported. The 40,000 VOLY estimate is also applied for sensitivity analysis. It is notable that all estimates for mortality valuation used here are considerably lower than those applied in similar work in the USA. Differences in valuation across Europe: WTP inevitably varies with income, with the result that valuation estimates will vary from country to country. Whilst this makes obvious sense for some impacts, such as the labour costs of repairing buildings, it is extremely controversial when applied to the non market aspects of health effect valuation, especially mortality. This is not problematic when considering issues that are confined within national borders, but difficulties inevitably arise when considering transboundary air pollutants. EC4MACS applies an average valuation across the EU. Gaps in knowledge of valuation: There are some significant gaps in the literature on valuation relative to air pollution impacts, especially on the concern valuation of ecosystems. Reliability of WTP estimates: A large number of biases have been identified that can affect the results of the contingent valuation surveys used to derive WTP estimates relating to health and other impacts. 80

89 ` Human health benefits Quantification of health impacts The health impacts quantified in this report are listed in Box 28, with details of the population considered for each effect. While the recent literature has provided substantial new information on the quantification of a wide range of different health impacts, the EC4MACS analysis restricts itself to impacts that are well documented and for which robust response functions are available. Thereby, the resulting benefit quantifications constitute a conservative estimate, as on many other health endpoints for which response functions are less robust are not considered. In line with advice of the World Health Organization, the EC4MACS analysis treats all particles, irrespective of source and chemical composition, as equally. Also, health effects are only estimated for particles that emerge from human activities; there is evidence of negative health effects from particles originating from natural sources (soil dust, sea salt, etc.), but these are not considered in this analysis. For the purposes of the present report, the effect of chronic exposure to PM2.5 on mortality is expressed in two ways, in terms of the loss of life expectancy (expressed as the total number of life years lost annually across the affected population) and the number of deaths brought forward (expressed as number of cases (deaths) per year). The loss of life expectancy is the preferred measure of impact on theoretical and practical grounds, though deaths brought forward is included for valuation purposes. The two estimates are not additive. Quantification of impacts only against exposure to ozone and PM2.5 does not mean that there are no effects of exposure to NO 2 and SO 2 on health. However, it is felt that separate inclusion of functions for these pollutants would incur a serious risk of double counting the effects quantified when using the functions based on PM 2.5 exposure, so it is not done. Box 28: Health impacts quantified in the EC4MACS benefit analyses Impact / population group Population Exposure metric Mortality from acute exposure All ages O 3, SOMO35 Respiratory Hospital Admissions Over 65 years O 3, SOMO35 Minor Restricted Activity Days (MRADs) 15 to 64 years O 3, SOMO35 Respiratory medication use Adults over 20 years O 3, SOMO35 Mortality from chronic exposure as life years lost or premature deaths Over 30 years PM 2.5, annual average Infant Mortality 1 month to 1 year PM 2.5, annual average Chronic Bronchitis Over 27 years PM 2.5, annual average Respiratory Hospital Admissions All ages PM 2.5, annual average Cardiac Hospital Admissions All ages PM 2.5, annual average Restricted Activity Days (RADs) 15 to 64 years PM 2.5, annual average Including lost working days 15 to 64 years PM 2.5, annual average Respiratory medication use 5 to 14 years PM 2.5, annual average Respiratory medication use Over 20 years PM 2.5, annual average Lower Respiratory Symptom days 5 to 14 years PM 2.5, annual average Lower Respiratory Symptom days Over 15 years PM 2.5, annual average Valuation of health impacts The monetary valuation of the various health benefits listed in Box 29 employs the unit values provided in Box 30. These values seek to describe the full economic effect of the impacts that they are linked with. For health impacts, for example, this will include elements associated with the costs of health care, lost productivity amongst workers and aversion to premature death or ill health. As discussed above, mortality impacts are quantified both in terms of deaths brought forward and the loss of life expectancy. Deaths are valued using a longestablished metric, the value of statistical life (VSL, also known as the value of a prevented fatality, VPF), whilst changes in life expectancy are valued using the value of a life year (VOLY). Despite major differences of more than an order of magnitude in the unit valuations, there is significant overlap in the ranges of results, as a result of the much larger number of life years lost compared to estimated premature deaths. Unlike PM mortality effects, which are related to long term exposure in our methodology and that generate a substantial change in life expectancy, the effect of ozone on mortality is presently only linked to short term exposures. There is some consensus that the shortening of life will be small, perhaps a year on average, and in many cases much less. The values presented in Box 30 are representative of willingness to pay in EU Member States. Being based in willingness to pay, they are income dependent. Here, however, only EU averaged values are used. 81

90 Box 29: Updated values for the health impact assessment (price year 2005) Impact / population group Unit cost Unit Ozone effects Mortality from acute exposure 57,700 / 138,700 /life year lost (VOLY) Respiratory Hospital Admissions 2,220 /hospital admission Minor Restricted Activity Days (MRADs) 42 /day Respiratory medication use 1 /day of medication use PM 2.5 effects Mortality from chronic exposure as: Life years lost, or Premature deaths 57,700 / 138, to 2.22 million /life year lost (VOLY) /death (VSL) Infant Mortality 1.6 to 3.3 million /case Chronic Bronchitis 208,000 /new case of chronic bronchitis Respiratory Hospital Admissions 2,220 /hospital admission Cardiac Hospital Admissions 2,220 /hospital admission Restricted Activity Days (RADs) 92 /day Respiratory medication use 1 /day of medication use Lower Respiratory Symptom days 42 /day Health benefits from the emission control scenarios emissions would be in each year all else being equal, in particular, factoring out changes in population. Applying the methodology discussed above, substantial health benefits could emerge from additional emission control measures that are not implemented in the current legislation baseline projection. There are nearly nine million life years lost per year across Europe in 2000 (nearly 6 million in the EU27) in 2000 and many more cases of hospital admissions, chronic bronchitis and various effects that may be minor at the level of the individual, but which could affect a very large number of people. These figures are roughly halved by 2030 under current legislation. Figure 99 shows the distribution of monetary damage across impact categories taking the case where mortality is valued using the median estimate of the VOLY. The example taken is for the MTFR scenario in It is clear that effects quantified against PM 2.5 exposure greatly dominate effects quantified against ozone exposure. Overall effects of chronic exposure on mortality account for two thirds of damage. For morbidity, chronic bronchitis (11%), restricted activity days (9%) and lower respiratory symptom days (6%) all make significant contributions. In contrast, infant mortality, hospital admissions and respiratory medication use make a negligible contribution to the total damage. The benefits analysis presented in this report focus mainly on the reduction in impact from moving to the MTFR scenario. However, Figure 100 shows how effects are reduced over time as a consequence of current legislation. A formal quantification of the benefits of current legislation would need to develop counterfactual scenarios that would show how large Figure 99: Proportion of damage attributable to each impact category (median VOLY applied for mortality impacts) Figure 100: Fall in health damage from air pollution under current legislation, 2000 to 2030 (blue line). Additional reduction using further measures contained in the GAINS model shown in red from

91 ` Table 43: Change in estimated annual health benefits from baseline to MFR, due to air pollution for core scenarios, EU27. *Life years lost and deaths from chronic exposure to PM 2.5 are alternate measures of the same effect Acute Mortality (All ages) Premature deaths O3 3,914 4,427 4,763 Respiratory Hospital Admissions (65yr +) Cases O3 3,760 4,468 5,014 Minor Restricted Activity Days (15 64yr) Days O3 8,451,329 9,025,509 9,173,841 Respiratory medication use (adults 20yr +) Days O3 3,208,417 3,536,629 3,709,693 Chronic Mortality (All ages) Life Years Lost * Life years lost PM 739, , ,797 Chronic Mortality (30yr +) deaths * Premature deaths PM 78,941 78,135 82,523 Infant Mortality (0 1yr) Premature deaths PM Chronic Bronchitis (27yr +) Cases PM 36,418 35,111 36,238 Respiratory Hospital Admissions (All ages) Cases PM 13,201 12,668 13,020 Cardiac Hospital Admissions (All ages) Cases PM 8,141 7,813 8,030 Restricted Activity Days (15 64yr) Days PM 67,587,352 63,135,236 63,025,253 Respiratory medication use (children 5 14yr) Days PM 733, , ,772 Respiratory medication use (adults 20yr +) Days PM 6,182,361 5,950,626 6,131,237 Lower Respiratory Symptom days (5 14yr) Days PM 34,013,907 32,051,513 32,396,962 LRS among adults (15yr +) Days PM 62,497,788 60,195,523 62,069,634 Table 44: Monetised equivalent of health impacts due to air pollution, EU27, million/year, 2005 prices. (a) Results for baseline, *Duplicate mortality estimates are for sensitivity analysis 2000 CLE 2020 MTFR 2020 CLE 2030 MTFR 2030 Acute Mortality (All ages) median VOLY O3 1,981 1,416 1,190 1,404 1,129 Acute Mortality (All ages) mean VOLY O3 4,762 3,403 2,860 3,376 2,715 Respiratory Hospital Admissions (65yr +) O Minor Restricted Activity Days (MRADs 15 64yr) O3 3,569 2,246 1,891 1,983 1,598 Respiratory medication use (adults 20yr +) O Chronic Mortality (All ages) LYL median VOLY PM 337, , , , ,017 Chronic Mortality (All ages) LYL mean VOLY PM 810, , , , ,229 Chronic Mortality (30yr +) deaths median VSL PM 544, , , , ,273 Chronic Mortality (30yr +) deaths mean VSL PM 1,108, , , , ,033 Infant Mortality (0 1yr) median VSL PM 2, Infant Mortality (0 1yr) mean VSL PM 4,577 1,585 1,135 1, Chronic Bronchitis (27yr +) PM 46,959 28,807 21,232 27,079 19,542 Respiratory Hospital Admissions (All ages) PM Cardiac Hospital Admissions (All ages) PM Restricted Activity Days (RADs 15 64yr) PM 44,713 23,605 17,386 20,752 14,953 Respiratory medication use (children 5 14yr) PM Respiratory medication use (adults 20yr +) PM LRS symptom days (children 5 14yr) PM 11,155 5,432 4,004 4,919 3,558 LRS among adults (15yr +) with chronic symptoms PM 17,271 9,988 7,363 9,369 6,762 Total, with median VOLY 467, , , , ,654 Total, with mean VOLY 944, , , , ,685 Total, with median VSL 672, , , , ,109 Total, with mean VSL 1,239, , , , ,670 In summary, health benefits that would emerge from the implementation of the additional emission reduction measures beyond the current legislation baseline would amount to billion/year in 2020, and billion/year in 2030, depending on the chosen basis for the valuation of human life (Table 45). Restricting the evaluation to end points that do not depend on the value of human life, the economic gain from reduced days with restricted activity (RAD) from the additional emission control measures amount to approximately 6 billion/year. Table 45. Health benefits of the MFR scenario relative to baseline for the EU727, [ million/year, 2005 prices] Total with median VOLY 61,597 58,205 Total with mean VOLY 121, ,561 Total with median VSL 104, ,080 Total with mean VSL 194, ,

92 Non health impacts Detailed quantification of effects of the policy scenarios on ozone damage to crops and acid damage to buildings requires additional pollutant metrics to those made that are readily available in the EC4MACS analysis. A simpler approach has therefore been taken, details of which were given in the supporting EC4MACS documentation ( It is noted that there are several limitations of this approach for quantifying non health impacts: It only permits quantification of crop and utilitarian material damage. It does not fully quantify effects on either utilitarian buildings or crops. For example, no account is taken of changes in the productivity of grassland that may impact production of livestock and associated goods, and no account is taken of the effects of particle emissions on building soiling. It only accounts for effects within the EU. It is based on emission scenarios for 2010, which may introduce significant error, particularly for ozone impacts due to their non linearities and dependence on the overall pollution climate. Damage to other non health receptors, notably ecosystems and cultural heritage, has not been quantified. The simplified method draws on past /tonne emission estimates of marginal damage to the various receptors. Agricultural crops Various air pollutants are known to affect the production and quality of agricultural and horticultural products, with most concern focused on ozone. SO 2 and nitrogen deposition also affect yield, with both positive effects via fertilisation and negative effects mediated through pests and disease. The two main effects on crop production from exposure to ozone are considered to be direct effects on yield and loss of crops through visible injury to leaves, etc. that are sold for consumption. Response functions are available for the former, and there is knowledge of the concentrations capable of causing visible injury. However, to quantify the latter through to economic damage would require more detailed analysis of pollutant concentrations, crop distributions, etc. than is possible at a pan European scale at the present time, so only direct effects on yield are quantified. The first step in the analysis is mapping of crop distributions across Europe, which has been done as part of pan European research on land use. These maps can be combined with concentration data to provide an estimate of exposure for different crops, and then response functions are applied to estimate yield loss. Response functions are available for a variety of the most important European crops, as shown in Box 31, which also provides information on sensitivity. Valuation is carried out using world market prices provided by UN FAO. Box 30: Crops considered in the damage assessment, and their relative sensitivity Sensitive (Critical Level 5 ppm.h) Cotton, Lettuce, Pulses, Soybean, Salad Onion, Tomato, Turnip, Water melon, Wheat Moderately sensitive (Critical Level of 5 10 ppm.h) Potato, Rapeseed, Sugarbeet, Tobacco Moderately resistant (Critical Level of ppm.h) Broccoli, Grape, Maize, Rice Insensitive (Critical Level >20 ppm.h) Barley, Fruit (plum & strawberry) Material damage Damage to stone buildings and monuments was, with ecological damage, one of the main drivers for the development of policy on transboundary air pollution in the 1970s, 1980s and early 1990s. The pollutants chiefly responsible for such damage were SO 2, ozone and soot. Collaborative research across Europe, coordinated through the Task Force on Materials under the UNECE Convention on Long Range Transboundary Air Pollution (CLRTAP), has given a good understanding of the types of materials affected by different air pollutants. Response functions are available that permit quantification of the loss of material (various types of stone, steel, zinc, etc.) at different levels of pollution. These functions are complex, requiring knowledge of the concentration of several pollutants and also of weather conditions (humidity, temperature and rainfall). They have, however, been successfully used to describe damage to utilitarian buildings (those without significant cultural merit) and structures across Europe. A first step in the quantification of damage to utilitarian buildings is description of the stock at risk, in terms of the types of structure affected by pollution and the materials at risk. The stock at risk used here has been 84

93 ` derived using information from surveys in the Czech Republic, Stockholm, Sarpsborg (Norway), the UK and the western Lander of Germany. Response functions are available for many materials, of which the following are most significant for the analysis limestone, mortar, sandstone, steel and zinc (used in galvanised steel). Uncertainty again arises through limited knowledge of the way that materials in use are exposed to pollution, relative to the way that the test samples used to derive response functions are exposed. Estimates of economic damage to materials within the EU are presented in Section 5. Unfortunately, it is not yet possible to quantify damage to materials used in structures of cultural significance in monetary terms. Data is lacking at several steps in the impact pathway, for example concerning the stock at risk (number of buildings, etc., number, size and quality of specific elements such as statues, and so on), repair and maintenance costs for diverse objects and structures, and information on the way that people value the cultural loss that occurs through damage. It is, however, possible to quantify degradation rates for the different materials under different pollution climates. Box 31: Damage to materials Acid rain can also cause damage to certain building materials and historical monuments. This results when the sulphuric acid in the rain chemically reacts with the calcium compounds in the stones (limestone, sandstone, marble and granite) to create gypsum, which then flakes off. This result is also commonly seen on old gravestones where the acid rain can cause the inscription to become completely illegible. Acid rain also causes an increased rate of oxidation for metals, and in particular copper and bronze. Visibility is also reduced by sulphate and nitrate aerosols and particles in the atmosphere. The rate of damage has reduced very significantly in the last 30 or 40 years as a result of reduced emissions of SO 2 and soot, especially in urban areas where of course many historic structures are to be found. Table 46: Benefits of the MTFR scenario to materials and crops compared to baseline, M/year Austria Belgium Bulgaria Czech Rep Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden UK Total Un monetised non health benefits In addition to the effects to crops and materials, there are also of course effects on ecosystems from eutrophication, ozone and acidification. As noted already, it is not currently possible to value impacts to ecosystems. Problems arise because of the diversity of benefits that good ecological health brings (existence values for biodiversity, recreational opportunity, water management, natural products such as berries and mushrooms and so on) and a current inability to link pollutant deposition to these effects beyond identification of areas at risk of critical loads exceedance and information on recovery times. Given the magnitude of health benefits, some commentators have concluded that adding in ecological benefits would make little difference to the results. However, this is not necessarily the case, for the following reasons: Early pan European action on transboundary air pollution was directed towards acid rain controls specifically because of great concern over ecological impacts. Although the area of ecosystems at risk from exceedance of the critical load for acidification has been reduced as a consequence of emission controls, ecosystems in some parts of Europe remain at significant risk. Very large areas have been identified as being at risk from exceedance of the critical load for nitrogen / eutrophication. 85

94 The effects of exceedance of the critical load for eutrophication on biodiversity at many sites throughout Europe can be considered profound. On this basis, and despite the fact that ecosystem impacts are not monetised here, it is concluded that there is a good basis for considering that inclusion of ecosystem effects could increase overall benefit results significantly. protection. It is worth considering, however, what differences there may be in policies aimed primarily at protecting health or ecosystems, relative to the extent of control of each pollutant and the areas subject to the greatest level of control. To some extent, it may be considered that policies aimed at health improvement via control of particle and ozone exposures will also yield benefits for ecological Cost benefit analysis For the purposes of cost benefit analysis, the straight comparison between the baseline and MTFR scenarios as performed here is crude. The additional measures contained in the MTFR scenario will vary greatly in costeffectiveness. Analysis elsewhere (Holland et al, 2012) shows that amongst these measures some will record a substantial net benefit, whilst others will impose a substantial net cost. Finer resolution for the analysis would permit closer identification of how far it would be appropriate, on cost benefit grounds, to go with emission reductions. However, this is unimportant for the present analysis given that it is not immediately of policy relevance. Comparison of costs and benefits The cost benefit analysis shown in Table 47, taking aggregate costs and benefits for the modelled domain and for the EU 27, demonstrates net benefits for the MTFR scenarios in all cases, accepting that limited account is taken so far of uncertainties other than those linked to mortality valuation. Table 47: Costs and benefits to health, crops and materials of the MFR scenarios compared to the baseline scenario, M/year Costs over baseline 37,625 40,722 Net benefits over baseline: Total with median VOLY 24,679 18,187 Total with mean VOLY 84,992 74,544 Total with median VSL 67,772 68,062 Total with mean VSL 157, ,586 median VOLY) the benefit cost ratio is around two for all countries and around 1.5 for the EU27. Table 48: Benefit cost ratios for the policy scenarios for 2030, M/year, accounting for effects on health, materials and crops. Top half, all countries, lower half, EU27. Net benefits, EU Total with median VOLY Total with mean VOLY Total with median VSL Total with mean VSL Uncertainties An extended analysis of uncertainties affecting the costbenefit analysis has been carried out using the TUBA framework, which has been partly developed through EC4MACS. This part of the analysis proceeds through four stages listed in Box 32. Box 32: Four steps in the uncertainty analysis 1. Quantification of the probability of benefits exceeding costs, based on description of the distribution of benefits around a given estimate. 2. Sensitivity analysis (here focused on alternatives for mortality valuation) 3. Consideration of unquantified biases 4. Development of conclusions taking account of stages 1, 2 and 3. Table 48 provides an alternative way of comparing costs and benefits using benefit cost ratios. A net cost would be shown if this ratio fell below one, but this does not apply to any of the scenarios considered. Taking the author s preferred position (mortality valued using the The analysis of uncertainties has quantified the possible consequences of variability with respect to data used for (e.g.) response functions and valuation, and sensitivity to key assumptions. It has also sought to provide a comprehensive overview of unquantified biases that affect the results, focusing on the likelihood that the 86

95 ` benefits of action will exceed costs. By accounting for these various elements, the analysis of uncertainties provides a comprehensive overview of the robustness of results. Results of the quantitative uncertainty analysis indicate that in all cases where analysis considers all European countries, there is at most a 10% probability that benefits would not exceed costs. In all cases using a VOLY higher than from (Desaigues et al. 2011) or using the VSL, the probability of costs exceeding benefits is less than 10%. Restricting analysis to the EU27 reveals a slightly different pattern. There is around a 10% probability of costs exceeding benefits when mortality is valued with the median VOLY and a 25 to 40% probability when the Desaigues et al VOLY is applied. Again, there is substantially less than a 10% probability of a net cost when applying the VSL. Overall, results consistently indicate that the move to the MFR scenario would be beneficial to society on economic grounds. [Leaving aside the non marginal nature of the scenario comparison, for the purposes of illustration.] Discussion This report demonstrates the quantification of health benefits of air pollution control policies, together with benefits through reduced damage to crops and building materials used in utilitarian applications. The basis for this quantification is effective use of the ALPHA and ALPHA Riskpoll models alongside the GAINS model, reflecting progress made via the EC4MACS project since its inception. The bias analysis indicates that there are more biases that either reduce estimates of benefits or increase estimates of cost than vice versa; in other words the biases overall appear likely to act against any package of measures passing a cost benefit test. The weighted equivalent (again, here based on the author's views, rather than, e.g. an expert panel though there is no reason that such a panel could not be convened in the future) provides a very similar outcome. Accepting the views expressed in the bias analysis as a reasonable representation of reality would suggest that the conclusion above that benefits are likely to exceed costs becomes more robust when additional uncertainties are brought into the equation. A caveat in relation to the bias analysis concerns those elements where it was concluded that the direction of bias was unclear (these concerned costs to agriculture and costs to domestic users). Here and elsewhere, the robustness of the analysis could be improved through further discussion and data collection if it was thought necessary. Beyond this point there is still, overall, a net benefit, but it is derived purely from the surplus of benefits earlier in the cost curve. The analysis presented here provides only the start and end points for the cumulative curve, giving a result very similar to that of the right hand figure. More detailed work would be needed to identify the social optimum within the range. The principal purpose of this report is to demonstrate the methods developed under the EC4MACS project, dealing with benefits assessment. It has been noted that certain parts of the analysis would be performed differently were they part of a policy exercise rather than a research study. A good example concerns the scenarios considered here: policy work would include a series of intermediate scenarios to better understand the evolution of costs and benefits between the extremes of the baseline and MTFR scenarios. This can be illustrated as follows. Figure 101 (using illustrative data only) shows the marginal costs and benefits of abating a pollutant. Marginal benefits are more or less constant throughout, but marginal costs rise as progressively more expensive options are introduced. Measures costing more than 100/tonne in this example (covering abatement beyond 70 to 80%) would not be justified on cost benefit grounds. The right hand figure uses the same data but shows cumulative costs of abatement. Cumulative benefits exceed cumulative costs throughout. At first sight, it may be thought that all measures should be included. Closer inspection reveals the fall off in net benefit (the gap between the lines) from about 70% abatement. Figure 101: Illustration of marginal and cumulative analysis of costs and benefits. 87

96 88

97 ` Conclusions 89

98 Conclusions The EC4MACS integrated model framework has been used to explore the likely development of future emissions of air pollutants and greenhouse gases in Europe and their impacts on human health and environment. The analysis shows that, while Europe has successfully eliminated the most visible and immediately harmful effects of air pollution, there is ample and robust scientific evidence that even at present rates Europe s emissions to the atmosphere pose a significant threat to human health, ecosystems and the global climate, though in a less visible and immediate way. The outlook into the likely development of emissions of greenhouse gases and air pollutants and their impacts up to 2030 adopts as a central assumption projections of economic development developed in the year 2009, and considers national and EU wide energy, climate, agricultural and air pollution policies that have been implemented by spring The report does not include the targets on renewable energy sources and on greenhouse gas emissions from the non ETS sector that were agreed in EU's Climate and Energy Package early The analysis starts from a recent projection of population development, which suggests a 6% increase in the population of the EU 27 due to continued immigration. By 2020 total GDP in the EU 27 is assumed to be 30% higher than in 2005, and 50% in Energy and climate policies will show distinct effects on future energy consumption in Europe, and decouple the levels of economic activity (GDP) from energy consumption. Total energy consumption is assumed to remain at the 2005 level up to Although renewable energy will increase its market share to some extent, no major changes in the composition of fuel use are projected up to 2030 despite the assumed 50% increase in GDP. Increases in car ownership in the new Member States will be compensated by saturation effects in the old Member States. Further growth, however, will occur for freight transport, although the 33% increase to 2030 is lower than the assumed growth in GDP. For the agricultural sector, the baseline suggests a decline in the numbers of cattle and sheep and increases in pigs and chicken. These changes in human activity levels, together with dedicated policies to reduce emissions of greenhouse gases and air pollutants, will have distinct impacts on future pollution of the atmosphere in Europe. Most notably, the baseline projection suggests a certain decline of greenhouse gas emissions, reaching 8% in 2020 and 16% in 2030 relative to Larger reductions (between 33% and 66%) are expected for emissions of air pollutants (i.e., SO 2, NO x, PM2.5, VOC), due to the on going structural changes in the economy and the effects of new emission control legislation. These changes in baseline emissions will have distinct impacts on air pollution impacts on human health, forests, vegetation, freshwater, crops and materials. Health impacts from exposure to fine particulate matter, which is associated for the year 2000 with a shortening of statistical life expectancy of 10 months in the EU, would decline by about 50% up to The number of premature deaths that are attributable to exposure to ground level ozone (33,000 in 2000) would decline by one third in 2020 and by more than 50% in Similar positive impacts are computed for vegetation and ecosystems. For instance, ecosystems area where biodiversity is threatened by excess nitrogen deposition will shrink from 1.1 million km 2 in 2000 to below 900,000 km 2 in 2030, and acidification will remain an issue at less than 0.5 percent of the European forest area. However, despite these significant improvements, the anticipated baseline development of emissions to the atmosphere will not be sufficient to achieve sustainable environmental conditions that safeguard human health and ecosystems services. Per capita greenhouse gas emissions will still be at 9.2 tons CO 2 eq/person/yr in 2020 and at 8.3 tons/person/yr in 2030, which is significantly higher than the approximately two tons that would be available in a budget approach that allocates equal GHG emissions to all people in the world while limiting temperature increase to 2 degrees Celsius. Also for air pollution, despite the impressive reductions in precursor emissions, fine particulates in ambient air will still cause life shortening of almost five months to the European population. It is estimated that the European population would still suffer a loss of 200 million life years and experience 19,000 premature deaths because of ozone exposure. Biodiversity will remain threatened by excess nitrogen input at 900,000 km 2 of ecosystems, including 340,000 km 2 which are legally protected, inter alia as Natura2000 areas. It is unlikely that the baseline development will achieve full compliance with the air quality limit values for PM10 and NO 2 throughout Europe. In general, despite the envisaged improvements, the European Union will fail to reach the environmental targets that have been established in the EU Thematic Strategy on Air Pollution for 2020 for human health and biodiversity. There is scope for additional measures that could alleviate the remaining damage and move closer to the objectives of the Sixth Environment Action Program. Full application of readily available technical emission reduction measures in the EU could reduce health impacts from PM by 2020 by another 30% and thereby gain more than 55 million life years in the EU. It could save another 3,500 premature deaths per year because of lower ozone concentrations. Further controls of agricultural emissions could protect biodiversity at another 270,000 km 2 of ecosystems against excess nitrogen deposition, 90

99 ` including 120,000 km 2 of Natura2000 areas and other protected zones. It could eliminate almost all likely exceedances of PM10 air quality limit values in the old Member States; in the urban areas of new Member States additional action to substitute solid fuels in the household sector with cleaner forms of energy would be required. Such Europe wide emission controls would also eliminate in 2030 all likely cases of non compliance with EU air quality standards for NO 2 with the exception of a few stations for which additional local measures (e.g., traffic restrictions, low emission zones) would be necessary. Based on well established techniques, health benefits, i.e., reduced mortality and morbidity from less exposure to fine particulate matter and ground level ozone, can be expressed in monetary terms. Based on a most conservative estimates, the health benefits from the additional measures would account to 25 billion /yr in 2020; upper estimates range up to 157 billion /yr. Obviously, the additional measures that would achieve these benefits come at a certain costs, which are estimated at billion /yr. These direct emission control costs constitute on average about % of the GDP, however there are large variations across Member States, mainly due to differences in economic wealth. These investments into air pollution control equipment replace productive investments and do not produce economic utility to the consumers. Depending on the sector, these investments consume between 0.2% of the produced value added (in ferrous and non ferrous metals industries) and 4% (for electricity production). However, at the same time production of pollution control equipment will indirectly induce higher activity in other sectors. Overall, it is estimated that the emission controls of the maximum feasible reduction case would lead to % lower GDP in 2030 compared to the baseline projection, which assumes for 2030 about 50% higher GDP compared to The slightly lower GDP increase is mainly a consequence of a slower growth in household consumption as productive resources are diverted for environmental investments. However, models do not suggest net impacts on employment, although trends in different economic sectors can vary. It should be mentioned that the Maximum Technically Feasible Reduction scenario analyzed in this report portrays a hypothetical extreme case in which all technical measures that are currently available were applied to their full extent. In reality, well chosen sub sets of measures will achieve large portions of the feasible benefits at a small fraction of costs, so that such measures emerge as highly cost effective means for further improvements of air quality in Europe. The EC4MACS model system offers a practical toolbox for exploring cost effective portfolios that yield high returns of health and environmental benefits.

100 References Bobbink R, Hettelingh J P (2011) Review and revision of empirical critical loads and dose response relationships. Coordination Centre for Effects, RIVM, Bilthoven, The Netherlands Desaigues B, Ami D, Bartczak A, et al. (2011) Economic valuation of air pollution mortality: A 9 country contingent valuation survey of value of a life year (VOLY). Ecological Indicators 11: doi: /j.ecolind Emberson LD, Ashmore MR, Cambridge HM, et al. (2000) Modelling stomatal ozone flux across Europe. Environmental Pollution 109: doi: /S (00) Galloway JN, Townsend AR, Erisman JW, et al. (2008) Transformation of the Nitrogen Cycle: Recent Trends, Questions, and Potential Solutions. Science 320: doi: /science Giannakouris K (2010) Regional population projections EUROPOP2008: Most EU regions face older population profile in Population and social conditions, Eurostat Statistics in Focus 1: Hettelingh J P, Posch M, Slootweg J (2009) Progress in the modelling of critical thresholds, impacts to plant species diversity and ecosystem services in Europe. Netherlands Environmental Assessment Agency, Bilthoven, The Netherlands Kiesewetter G, Borken Kleefeld J, Heyes C, et al. (2013) Modelling compliance with NO2 and PM10 air quality limit values in the GAINS model. International Institute for Applied Systems Analysis, Laxenburg, Austria Mills G, Hayes F, Simpson D, et al. (2011) Evidence of widespread effects of ozone on crops and (semi )natural vegetation in Europe ( ) in relation to AOT40 and flux based risk maps. Global Change Biology 17: doi: /j x 92

101 `

102 There are important interactions and potentially large economic synergies between climate and air quality strategies and the objectives of EU social and economic policies. Model analyses, based on latest scientific findings and validated data, can provide valuable information on the design of (cost-)effective strategies that meet multiple policy objectives. A consortium of leading scientific institutions has developed a toolbox of well established modelling tools to explore the synergies and interactions between climate change, air quality and other policy objectives. The EC4MACS toolbox is now ready for scientific and economic analyses to inform the revision of thethematic Strategy on Air Pollution in 2013 and the European Climate Change Programme on climate strategies beyond The EC4MACS toolbox informs about the costs and benefits of the various policy options to reduce reduce greenhouse gas emissions and to further improve air quality in the European Union while maximizing the benefits to EU energy, transport and agricultural policies. More information: