THE NAMEA AIR FOR BELGIUM (1990/ )

Transcription

1 Aaaaa&& Federal Planning Bureau THE NAMEA AIR FOR BELGIUM (1990/ ) Sébastien Gilis Guy Vandille February 2006

2 ACKNOWLEDGEMENTS This report has benefited from funding by the European Commission, GD Environment and GD Eurostat, by means of the Grant Agreement nr for the operation entitled "Statistics on environment accounts - NAMEA air emission accounts". Apart from these project sponsors we also wish to thank all the people who provided the data necessary to construct the present NAMEA Air, more specifically Michael Govaert of the BIM-IBGE; Marie-Rose Van den Hende, Miet D'Heer and Caroline De Bosscher of the VMM ; Isabelle Higuet of the DGRNE; and Luc Avonds, Ingrid Bracke and Koen Hendrikx of the Federal Planning Bureau. We also wish to thank the people who provided the data which enabled us to perform the calculations necessary to write up chapter V, namely Perry Francis of the UK Office for National Statistics; Sjoerd Schenau of the Central Statistical Bureau of the Netherlands; and Ingrid Bracke, Delphine Bassilière, Francis Bossier of the Federal Planning Bureau. Our gratitude extends also to Johan Wera and Luc Avonds of the Federal Planning Bureau for their aid with the calculations for chapter IV, as well as to Geert Bryon of the Federal Planning Bureau who provided the computer programs necessary for the calculations supporting this chapter, as well as for chapter VI. Finally, words of gratitude are also entitled to Bart Hertveldt and the Direction Committee of the Federal Planning Bureau for proofreading the draft version of this report.

3 TABLE OF CONTENTS Acknowledgements Table of contents Executive Summary Introduction 1 I Data compilation methodology 2 A Economic data 2 1 Output, value added and intermediate consumption at current prices 2 2 Output, value added and intermediate consumption at constant prices 3 3 Employment 3 4 Household consumption 3 B Regional Pollution data 4 1 Brussels-Capital The energy balance The greenhouse gas and the Corinair inventory The emission inventory of ozone depleting substances and greenhouse gases 5 2 Flanders The atmospheric emission inventory The energy balance The emission inventory of ozone depleting substances and greenhouse gases 8 3 Wallonia The Corinair inventory and the energy balance The emission inventory of ozone depleting substances and greenhouse gases 9 C Air pollution by transport 10 1 Road transport 11 2 Air transport 11 3 Water transport 12 D Data precision 12

4 II The results 14 A Economic data 14 1 Global economic evolution 14 2 Household expenditure 15 3 Analysis by industry 17 B Pollution data 19 1 Definitions 19 a The greenhouse effect 19 b Acidification 20 c Photochemical pollution 20 2 Global air pollution 21 3 Air pollution by industries 23 4 Air pollution by households 28 III Eco-efficiency 32 A Definitions 32 B Eco-efficiencies and ecogains for industries 33 C Eco-efficiencies and ecogains for household consumption 44 IV Input-output analysis 45 A. Allocation of emissions and energy use by domestic producers to final demand for individual products The matrix of pollution and energy use coefficients Allocation of emissions and energy use to final demand for individual products 48 B. Air pollution, energy use and international trade The matrix of pollution and energy use coefficients Allocation of emissions and energy use to final demand for individual products 54

5 V Medium term air pollution projections 57 A Methodology 58 1 The pollution coefficients 58 2 The BAT-method 60 B Results 63 1 Calculation of the pollution coefficients 63 2 The impact of expected economic growth on air emissions 64 3 Extrapolation of the evolution of the emissions 66 4 The national BAT method with constant technology in the most eco-efficient region 68 5 The national BAT method with extrapolation of the evolution from the past 70 6 The international BAT method with constant technology in the most ecoefficient region 73 7 Conclusion 75 VI Decomposition analysis of changes in CO 2 emissions in Belgium 77 A Methodology 78 1 Theory 78 2 Application to the Belgian environmental accounts 79 B Results 82 1 Total industry CO 2 emissions 82 2 CO 2 emissions by the primary, the manufacturing and the services industry 84 3 CO 2 emissions by the largest polluters 85

6 References 89 Annex 1: Industry classification, based on NACE rev. 1, and household classification, based on COICOP categories. 90 Annex 2a : Data precision (1 = highest, 5 = lowest, na = not available) 92 Annex 2b : Data precision (1 = highest, 5 = lowest, na = not available) 94 Annex 3 : Relative eco-efficiencies and eco-gains based on employment: 8 most and 8 least efficient sectors 96 Annex 4: Relative eco-efficiencies and eco-gains based on employment: 8 highest gains and losses 99 Annex 5: Weight of each industry in value added and employment in 2002 and evolution between 1990 and 2002 (in %). 102 Annex 6: Greenhouse gas emissions of type 1 ( averages and changes; in %) 105 Annex 7: Greenhouse gas emissions of type 2 ( averages and changes; in %) 108 Annex 8: Greenhouse gas emissions of type 3 ( averages and changes; in %) 111 Annex 9: Acidification ( averages and changes; in %) 114 Annex 10: Photochemical pollution ( averages and changes; in %) 117 Annex 11: Lead ( averages and changes; in %) 120 Annex 12: PM10 ( averages and changes; in %) 123

7 Executive summary The current report presents the NAMEA Air for Belgium for the years 1990/ The NAMEA Air is a satellite account of the national accounts. It contains data on air pollution as well as economic data, both allocated to industries and household consumption categories. The data on air pollution have been adjusted as to abide by the resident principle, the principle which also rules the economic data in the national accounts. Furthermore, emissions from own account transport have been allocated to the different industries, instead of allocating this type of emissions to the functional transport sector. Indicators of air pollution In order to get an overview of the mass of information contained in the environmental accounts, pollutants were regrouped into three environmental themes. The three environmental themes are the greenhouse effect, acidification, and photochemical pollution. The greenhouse effect is presented under three different forms each regrouping a different set of pollutants. The first form will be called Greenhouse effect of type 1. It only contains three traditional greenhouse gases. It is defined as follows: CO * N 2 O + 21 * CH 4 It is expressed in terms of tons of CO 2 equivalents, with the GWPs 1 of individual gases used as weights on each pollutant to reflect the potential relative stress on the environment of each substance. This weighting procedure is also used in the other two greenhouse effect indicators. The second form under which the greenhouse effect is presented, called Greenhouse effect of type 2, adds two more pollutants two those used in order to calculate Greenhouse effect of type 1. It is defined as follows: CO *N 2 O + 21*CH *HFC *SF 6 This indicator is currently probably the most relevant since it takes into account all the greenhouse gases considered in the Kyoto protocol. 2 1 Greenhouse Warming Potentials. The GWP is a measurement technique to define the relative contribution of each greenhouse gas to atmospheric warming. CO 2 has been assigned a GWP of 1, against which all other greenhouse gases are compared. 2 PFCs are also considered in the Kyoto protocol, but since these data are only available for the entire economy, and thus not disaggregated at all, the NAMEA framework can not add anything as concerns the use of PFCs. PFCs have therefore been excluded from the analysis, though they are included in the tables. i

8 The third form under which the greenhouse effect is presented, called Greenhouse effect of type 3, adds two more pollutants to the Kyoto-indicator. It is defined as follows: CO *N 2 O + 21*CH *HFC *SF *CFC *HCFC 22 The acidification indicator is a simple Potential Acid Equivalent index, which does not take into account the location of the emissions nor their transport through the atmosphere. It expresses three acid rain precursors, SO 2, NO x and NH 3, in hydrogen ions. It are these hydrogen ions that form acid rain. One hydrogen ion thus corresponds to one acidification equivalent, two of which can be released by one molecule of SO 2, while for the other two gases only one ion can be produced per molecule. The acidification equivalents of the three substances can thus be calculated by dividing the weight of each substance by its molecular mass and multiplying it with the number of hydrogen ions that can be released from one molecule of the particular substance. This leads to the following formula 3 : Acidification = * SO * NOx * NH 3 This gives a figure showing the potential acidification due to air pollution, without taking into account any particular circumstances like the wind, humidity or other climatic variables influencing the degree to which the potential for acidification is realised. Potential photochemical pollution is measured with a formula that takes into account four pollutants. 4 Photochemical pollution = 1.22 * NOx + NMVOC * CO * CH 4 This indicator should be interpreted with caution, as chemical reactions causing tropospheric ozone and photochemical pollution are very complex and influenced by a multitude of factors. Air pollution in Belgium in the period 1990/ This section analyses the evolution of global air pollution in Belgium between 1990 and Global air pollution includes emissions from both industrial activity and household consumption. Figure A presents the three greenhouse effects (GHE). Values for greenhouse effects of type 2 and 3 are of course higher than type 1 since they include more gases in their composition. However, their evolution is quite similar. From 1990 to 2002, GHE 1 has increased by 2.5%, but the main part of this growth happened between 1990 and 1999 while in the period greenhouse gas emissions decreased. If we only consider the period, emissions decreased by 4.1%. 3 source: Eurostat (2000), page Bron: de Leeuw, F. (2001), p ii

9 GHE 2 closely follows GHE 1. Over the period, it decreased by 3.5%. The level of GHE 3, which also includes CFCs and HCFCs, is significantly higher than for GHE 1 and 2, but its decrease was also higher, namely 7.2%. Figure A: Evolution of greenhouse gas emissions Greenhouse effect 1 Greenhouse effect 2 Greenhouse effect 3 Acidification and photochemical pollution have undergone a much larger decrease than GHE. As witnessed by figures B and C, this decrease already started in the first half of the nineties. Acidifying emissions decreased by almost 28% between 1990 and Photochemical pollution decreased by almost 24% in that same period. Figure B: Evolution of Acidification iii

10 Figure C: Evolution of Photochemical pollution Table A presents the shares of the main industries and the main consumption categories in total emissions. These shares are computed as the average for the available period, that is to say 1990 to 2002, except for GHE 2 and GHE 3, for which the average was calculated over the period Table A: Shares of industries and households consumption in total emissions Greenhouse effect of type 1 Greenhouse effect of type 2 Greenhouse effect of type 3 Acidification Photochemical Pollution Total Industries A-C Primary Products D Manufacturers E Energy I Transport E-Q Services not allocated Total consumption Transport Heating Other For all type of emissions, the average share of the industries was higher than the one of household consumption. It ranged from 72% for photochemical pollution to 87% for acidification, while the share of industries in greenhouse gas emissions averaged 78.5%. The shares of the manufacturing industries and the services industries were approximately equal. Only for photochemical pollution we observe a significantly larger share for the services industries. Among the latter the energy and the transport industry were the major contributors to all types of air pollution considered. The primary products iv

11 industries were equally important as the manufacturing and the services industries as concerns acidifying emissions. For the other emissions their share was a lot lower. Heating by households was responsible for an equally large share of greenhouse gas emissions as the energy industry, about 15%. Transport by the households accounted for a share in photochemical pollution equally large as that of the transport industry, approximately 20%. Table B shows the share of the main groups of industries in total industry emissions and the growth of these emissions. The shares and the growth are quite similar for each of the three greenhouse gas emission indicators. The manufacturing industries and the services industries each contributed about 43% to greenhouse gas emissions. However, emissions by the manufacturing industries decreased for all three types of greenhouse gas emissions, while emissions by the services industries only decreased for type 3. The largest decrease of greenhouse gas emissions was recorded for the primary producers. The primary products industries contributed around a third to the industry emissions of acidifying substances, which is the same share as manufacturing industries and services industries. However, emissions of primary producers and manufacturing industries have gone down by 30 to 40%, while emissions due to the services industries have decreased by only 2.6%. The smaller decrease for the services industries was to a large extent due to the energy industry, of which the emissions increased by almost 50%. An even larger increase is observed for photochemical emissions by the energy industry. However, for this type of air pollutant emissions by the services industries decreased more than the emissions by the manufacturing industries and the primary producers. This follows from the fact that the share of the energy industry in photochemical pollution is not that important. The transport industry accounted for about a quarter, and the manufacturing industries for over 40%. Table B: Average Share and evolution 5 of Air Pollution by groups of industries Industry Greenhouse effect of type 1 Greenhouse effect of type 2 v Greenhouse effect of type 3 Acidification Photochemical Pollution Share Growth Share Growth Share Growth Share Growth Share Growth TOTAL A-C primary Products D Manufacturers E Energy I Transport E-Q Services not allocated Eco-efficiency and ecogains 5 Evolution for greenhouse effect of type 2 and 3 is computed on the period Evolution of all other variables is computed on the period

12 The eco-efficiency of an industry is normally defined as the amount of an economic variable, like output or value added, generated by that industry per unit of pollution emitted by the same industry. In order to put the focus on pollution we use the inverse ratio. This means that the more polluting the industry is, the higher the indicator will be. In other words, our indicator is an inverted eco-efficiency indicator. Furthermore, as we want to be able to directly deduce whether an industry is more or less polluting than total production, we use relative shares in the economy and in pollution. All industries for which the indicator is found to be higher than one, are more polluting than the entire industry. The indicator, which for reasons of readability we will continue to simply call eco-efficiency instead of inverted relative eco-efficiency, looks as follows: Eco-efficiency of industry or household consumption category i with respect to pollutant j = relative share of industry or household consumption category i in emissions of type j relative share of industry or household consumption category i in the economy The eco-efficiency of the industries will be measured with respect to both value added and employment. The eco-efficiency of the household consumption categories is measured with respect to the share in total household expenditure. In order to compare the environmental performance of different industries the level of the eco-efficiency ratio does not reveal much, because certain industries are more polluting than others due to the particular activities they perform. A better indicator for this purpose is the change in eco-efficiency. When economic growth of an industry is larger than emissions growth we talk of an ecogain for that industry, as opposed to an ecoloss when the opposite is true. The ecogain (or -loss) of industry i with respect to pollutant j is equal to the percentage change of value added or employment in industry i minus the percentage change of emissions of pollutant j by industry i. The ecogain (or -loss) of household consumption category i with respect to pollutant j is equal to the percentage change in real expenditure for this household consumption category i minus the percentage change in emissions of pollutant j by this household consumption category i. Table C presents the average eco-efficiency with respect to value added of the Belgian production sector in the period , as well as the change in eco-efficiency during that period. As can be readily ascertained, ecogains were achieved for the three types of air pollution considered. The largest gains were achieved for acidification. Value added increased over 50% faster than emissions of acidifying substances. The gains for photochemical pollution were only slightly lower. Ecogains for greenhouse gas emissions were less than half the gains for acidification. The largest gains were obtained for greenhouse gases of type 3, the lowest for the Kyoto greenhouse gases (type 2). This implies that emissions of CFCs and HCFCs have decreased faster than the other greenhouse gases, while emissions of HFCs and SF 6 have decreased least. vi

13 Table C: Average Eco-efficiencies and ecogains based on value added. 6 Greenhouse effect of type 1 Greenhouse effect of type 2 Greenhouse effect of type 3 Acidification Photochemical Pollution Avg. Eco-Eff. Eco- Gains Avg. Eco-Eff. Eco- Gains Avg. Eco-Eff. Eco- Gains Avg. Eco-Eff. Eco- Gains Avg. Eco-Eff. Eco- Gains Total A-C Primary Products D Manufacturing E Energy I Transport E-Q Services Table C also teaches us that the services industry is, as expected, most eco-efficient, in the sense that it clearly pollutes less per unit of value added than the rest of the economy. The services industry does contain two industries which are more polluting per unit of value added than the total economy, namely the energy and the transport industry. The transport industry was the most eco-inefficient industry as concerns photochemical pollution, but it also achieved the largest ecogains. The energy industry was the most ecoinefficient industry concerning greenhouse gases. It also achieved the largest ecogains, except for greenhouse gases of type 1, for which the primary products industry obtained even larger gains. The primary products industry also achieved the largest ecogains as concerns acidification, the type of air pollution for which it is most eco-inefficient by far. The only ecolosses observed between 1990 and 2002 are on account of the energy industry, which did not perform well in the fields of acidification and photochemical pollution. The manufacturing industry was less eco-efficient than the entire economy, as expected. It achieved above average ecogains for greenhouse gas emissions of types 2 and 3 and for acidifying emissions. Table D is the equivalent of table C, but this time employment was used in order to measure eco-efficiency. We notice that with respect to employment ecogains were obtained with respect to all the types of air pollution considered, although these gains were lower than with respect to value added. This reflects the increase in labour productivity which took place between 1990 and The ecogains of the primary products industry stand in contrast to the ecolosses incurred by the energy industry, which measured against employment, was more eco-inefficient than the primary products industry, even as concerns acidifying emissions. Table D: Eco-efficiencies and ecogains based on employment. 7 6 Efficiencies shown are computed for the period or, for greenhouse gases of types 2 and 3, for the period. 7 Efficiencies shown are computed for the period or, for greenhouse gases of types 2 and 3, for the period. vii

14 Greenhouse effect of type 1 Greenhouse effect of type 2 Greenhouse effect of type 3 Acidification Photochemical Pollution Avg. Efficien. Eco- Gains Avg. Efficien. Eco- Gains Avg. Efficien. Eco- Gains Avg. Efficien. Eco- Gains Avg. Efficien. Total A-C primary Products D Manufacturers E Energy I Transport E-Q other Eco- Gains Table E presents the eco-efficiencies and the ecogains for household consumption. Important ecogains have been achieved with respect to acidification and photochemical pollution. In the period real consumption expenditure increased between 40 and 50% faster than these types of emissions. Ecogains with respect to greenhouse gas emissions were a lot lower, just like for the industries. Table E: Eco-efficiencies and ecogains for household consumption Greenhouse effect of type 1 Greenhouse effect of type 2 Greenhouse effect of type 3 Acidification Photochemical Pollution Avg. Efficien. Eco- Gains Avg. Efficien. Eco- Gains Avg. Efficien. Eco- Gains Avg. Efficien. Eco- Gains Avg. Efficien. Eco- Gains Total consumption from which : Transport Heating Other Household heating was responsible for approximately 15 times as much emissions of greenhouse gases as its share in consumption expenditure, and almost 9 times as much acidifying emissions. For this type of air pollution household heating was about twice as inefficient as household transport, and ecogains were a lot lower. The latter is true for photochemical pollution as well. Although for this type of air pollution transport was still more eco-inefficient. Ecogains with respect to greenhouse gases of types 2 and 3 were almost equal for heating and transport, while for greenhouse gases of type 1 ecogains were a lot higher for transport. This discrepancy is due to the estimation of transport expenditure prior to 1995, which engenders a sharp increase in transport expenditure between 1990 and Input-output analysis In this section we make use of the Belgian input-output tables for 2000 in order to allocate emissions of polluting substances into the air by domestic producers to final viii

15 demand for 143 different products. 8 The input-output tables make it possible to take into account the pollution generated by the production of all the intermediate goods used during the production of the final goods The objective is thus to determine for each pollutant emitted during production in Belgium, as well as for the environmental themes, to what extent each product was responsible for their emission in Emissions will also be allocated to exports and imports as to determine net exports of different types of air pollution and energy use. Table F shows for each of the 14 pollutants for which data are available in the NAMEA Air, for the 3 environmental indices and for energy use, the five products with the highest percentage shares, as well as the sum of the shares of these five products. These are the products, the final demand for which generates most pollution in Belgium, both directly and indirectly. The definition of the product codes is given in table G. Approximately one third of all three types of air pollution, greenhouse gases, acidification, and photochemical pollution, was caused by the final demand for only 5 products. As concerns the Kyoto greenhouse gas index, the same five products show up as for CO 2 and energy use, which illustrates that CO 2 is the prime greenhouse gas and that greenhouse gas emissions are closely linked to energy use. These products are basic chemical products (24A1) 10 ; iron and steel, ECCS-ferro-alloys and pipes (27A1); nonferro metals (27B1B); electricity, steam and hot water (40A1A); and general civil construction (45B1A). The fact that construction activities are among the major generators of greenhouse gases might look surprising. Direct emissions from construction activities are indeed quite modest. However, the production of the inputs used in construction activities, e.g. manufacturing of articles made of concrete, gypsum and ciment, iron profiles or metal construction works, induces a large amount of emissions. Other products like electricity, iron and steel, or basic chemicals are not surprising to find among the top generators of greenhouse gases, as their production generates a high level of emissions directly. Table F : Shares of the five most important products in air pollution. 24A1 27A1 40A1A 27B1B 45B1A Sum of shares CO A1 01A1 24D1 15A1 24F1 8 Since air pollution, especially greenhouse gas emissions, is closely linked to energy use, the latter was allocated to the 143 products in the same way is the most recent year for which coherent calculations can be made, since detailed data on final demand or input-output data are not available for a more recent year. 10 These include among others industrial gases, colouring matters, plastic and synthetic rubber. ix

16 N2O A1 15A1 92D1 92A1 91A5 CH A1B 24A1 52A1 40A1A 25B1 HFCs A1 45B1A 26D1 40A1A 51A1B SF C1 45B1A 24A1 27A1 62A1 NOx A1B 24A1 40A1A 27A1 45B1A SOx A1 15A1 15E1 55B1 15F1 NH A1 23A1B 45B1A 34A1 60C1 NMVOC A1 27B1B 27B1A 45B1A 60C1 CO A1 51A1B 27B1B 24A1 70A1A CFCs B1 25A1 45B1A 63B1 51A1B HCFCs C1 60A1 05A1 63B1 15B1 PM A1 15A1 27A1 15E1 27B1B Pb A1 27A1 40A1A 27B1B 45B1A Kyoto A1 15A1 24A1 15E1 23A1B Acidification A1 24A1 27B1B 60C1 45B1A Photochemical A1A 24A1 27A1 27B1B 45B1A Energy Source: own calculations Final demand for basic chemicals was not only responsible for the largest share of CO 2 - emissions, but also for the largest share of N 2 O-emissions. Also among the top-5 for this pollutant were two other products from the chemical industry, namely pharmaceuticals and other chemical products 11, next to agriculture and meat and meat products. All these products, except for meat and meat products, have high direct emissions coefficients. The emissions of N 2 O as a consequence of the production of meat and meat products are mainly indirect emissions. Final demand for agriculture and meat and meat products was also, not surprisingly, the most import contributor to emissions of CH 4. Once again, emissions on account of meat and meat products were mainly indirect emissions. 11 These include among others explosives, glue and etheric oils. x

17 Table G: Definition of the product codes of table 18 Code Label 01A1 Agricultural products, hunting and related services 05A1 Fishing 15A1 Meat and meat products 15B1 Fish products 15E1 Dairy products 15F1 Mill products and starch 23A1B Refined petroleum products 24A1 Basic chemicals 24D1 Pharmaceutical products 24F1 Other chemical products 25A1 Tyres and other rubber products 25B1 Plastic products 26A1 Glass products 26D1 Concrete, cement, lime, stone, and other non metallic mineral products 27A1 Iron and steel, ECCS-ferro-alloys and pipes 27B1A Cold steel products, wire, and non-eccs-ferro-alloys 27B1B Non-ferro metals 34A1 Car engines, cars and lorries 40A1A Electricity, steam and hot water 45B1A General civil construction 51A1B Wholesale trade 52A1 Repair of consumer articles 55B1 Restaurants, pubs and catering 60A1 Transport by rail 60C1 Goods transported by road, truck rental, and transport via pipelines 62A1 Air transport 63B1 Freight handling, warehousing and other transport supporting activities 70A1A Rental of houses and other buildings by the owners 91A5 Non-market services 92A1 Audiovisual services, radio and television 92D1 Sports and recreation market services The same two products were also the top contributors to acidification, mainly due to their enormous share in NH 3 emissions. Dairy products was the third food related product among the top five. Emissions on account of meat and meat products as well as of dairy products were chiefly indirect emissions. The other two top five generators of acidifying pollutants were basic chemicals and refined petroleum products. For the refined petroleum products this was due mainly to their high share in SOx emissions, while basic chemicals were among the top five emitters of both SOx and NOx. Direct emissions are more important for the refined petroleum products, while basic chemicals are important xi

18 generators of acidifying emissions due to the indirect emissions from their use of electricity, gas and refined petroleum products. The direct pollution coefficient of basic chemicals is almost three times lower as for refined petroleum products. Basic chemicals were also by far the most important contributor to emissions of NMVOCs. These emissions were both of a direct and an indirect nature, with the indirect emissions being mainly on account of their use of refined petroleum products. Their use of electricity and gas also engenders an important amount of indirect emissions of NOx and CH 4. As a consequence basic chemicals played a major role in emissions of photochemical pollutants. The most important generator of photochemical emissions were iron and steel, however. Just like for the non-ferro metals this was due to their important share in the direct emissions of CO. Other products among the top five were road transport of goods and general civil construction. Both of them can be found among the top five generators of NOx-, NMVOC- and CO emissions. General civil construction generates a large amount of photochemical pollutants indirectly as a consequence of their use of bricks, concrete, gypsum, and iron and steel profiles. Road transport of goods was also the largest contributor to emissions of fine dust particles (PM10). Not surprisingly, two other transport related products are among the top five, namely transport by rail, and freight handling, warehousing and other transport supporting activities. Fish was identified as another important generator of PM10 emissions. Both fishing and fish preparations were among the top five. Emissions on account of fish products were chiefly indirect emissions. The production of food and basic metals contributed most to emissions of lead. Almost half of all the lead emissions were emitted in order to produce agricultural products and meat and meat products. Dairy products were also among the top five. The high position of meat products and dairy products is due to their intensive use of agricultural products, which were identified as the product with the highest pollution coefficient for lead. Somewhat less than a fifth of all lead emissions could be attributed to the production of iron and steel and non-ferrous metals. Table H shows the amounts of pollution and energy use caused by exports and by the hypothetical replacement of imports by domestic products. This allows us to calculate net exports of pollution and energy use, defined here as the difference between real pollution and energy use by domestic producers of export goods minus hypothetical pollution and energy use necessary to replace imports by domestic products. Table H also compares this expression for net exports of pollution and energy use to real total pollution and energy use. Table H: Air pollution and energy use embedded in international trade flows. Emissions or Emissions or Net exports Total energy saved energy for of emissions emissions by imports exports or energy energy use or Net exports as % of total emissions or energy use xii

19 CO2 1000ton N2O ton CH4 ton HFCs ton SF6 ton NOx ton SO2 ton NH3 ton NMVOC ton CO ton PM CFCs ton HCFCs ton Pb ton Kyoto ton Acidification ton Photochemical ton Energy use TeraJoules Source: own calculations In 2000 a little bit more energy was used by the Belgian industries to produce export goods than was saved by not having to produce the imported goods. Energy use for exports was around one percent higher than energy use for imports, while net exports of energy were also equal to around one percent of total energy use by the Belgian industries. Does this imply that Belgian exports are energy intensive? Not really. Since exports were five percent higher than imports in 2000 it were in fact imports which were more energy intensive. 12 Belgian exports were nevertheless relatively greenhouse gas intensive, although to a modest degree, as exports of greenhouse gases exceeded imports by over 6%. Net exports of greenhouse gases were equal to 5% of total greenhouse gas emissions as measured by the Kyoto index. This was due to emissions of CO 2, N 2 O, HFCs and SF 6. Net exports of CH 4, however, were negative. Exports contained 15% less CH 4 than imports. This implies that imports were CH 4 intensive. This was also the case for HFCs, as pollution embedded in exports exceeded pollution embedded in imports by no more than 3%. Belgian exports in 2000 were not only greenhouse gas intensive, but also contained far more photochemical pollution than imports. Exports contained 9% more photochemical pollution than imports. As a consequence, 6% of total photochemical pollution by the Belgian industries was exported. Except for CH 4, all types of photochemical pollutants were exported. No less than 14% of total carbon monoxide emissions was exported, for NMVOCs this was 6%, and for NOx 5%. Belgian imports were intensive in acidifying emissions. Although exports exceeded imports by 5%, acidifying emissions embedded in exports were lower than acidifying emissions embedded in imports. This was solely due to NH 3 emissions, of which net imports were 12 The energy intensiveness of economic activities is a relative concept. Depending on the spectrum of comparison a certain economic activity can be labeled energy intensive or not. So it could well be the case that Belgian exports are energy intensive compared to exports of its neighbouring countries, but this was clearly not the case with respect to Belgian imports in xiii

20 equal to 16% of total NH 3 emissions by the Belgian industries. Belgian exports were NOx and SOx intensive, however. Exports contained between 7 and 8% more of both pollutants than imports. Belgium also was an exporter of CFCs in Emissions of HCFCs, PM10 and lead were saved by virtue of international trade. Imports contained a higher amount of these pollutants than exports. Net imports of PM10 equalled 14% of total emissions by the Belgian industries. For lead this was equal to 9%. Medium term air pollution projections The NAMEA method to calculate projections of air pollution consists of linking pollution coefficients of industries and household consumption categories to economic forecasts of industry output, value added or employment for the industries, and household consumption expenditure for the households. 13 If forecasts exist about the evolution of value added, output, employment or household consumption expenditure, then the evolution of pollution can simply be calculated by application of the pollution coefficients to the forecasted change in the economic variables. This results directly in emissions of different kinds of air pollution by industry and by household consumption category. In order to optimize the policy relevance of the projections we opted to calculate them on a regional basis. The regionalised version of the NAMEA Air for Belgium contains emissions of CO 2, N 2 O, CH 4, NOx, SOx, NH 3, NMVOC and CO. These emissions consist of transport emissions, as well as process emissions and emissions from combustion. End-of-pipe measures are taken into account. The eight pollutants enable us to calculate indices for greenhouse gas emissions, acidifying emissions and photochemical emissions. Forecasts for the regional economic variables were generated by the HERMES model of the Federal Planning Bureau for the period Table I shows the correspondence between the HERMES industries 15 and the NAMEA industries at the NACE two digit level. For these 15 industries 3 regional pollution coefficients were calculated for the period for each of the 8 pollutants, with value added as the denominator. 16 This results in a set of 2520 pollution coefficients, which constitute the starting point for the emission projections. Next to the projections for the eight pollutants we also calculated three indices, the traditional greenhouse gas index, the potential acidification index and the photochemical pollution index. 13 Value added is probably the best indicator to measure industry activity, since output also contains purchases of intermediate goods, which does not generate emissions by itself. The production of these intermediate goods of course does generate emissions, but as far as these goods are produced domestically their production is taken into account in the value added of other industries. Employment can only serve as a measure of activity for services industries. 14 The projections were generated before the new version of the NAMEA Air was ready. They are thus based on the previous NAMEA Air for Belgium which contained air pollution data up until Remark that in the HERMES model industries 13-14, 24, 26 and form one industry, namely the industry of the intermediary goods. Because these industries are quite important with respect to air pollution we decided to calculate separate pollution coefficients. The value added growth rate of these four industries is of course identical. 16 Pollution by households was excluded from the paper because no regional household expenditure data exist. The results are thus restricted to air pollution by the industries. xiv

21 Table I : Correspondence between HERMES industries and NAMEA industries HERMES industry NAMEA industry by NACE code Agriculture, hunting and forestry Energy 10-12, 23, Consumer goods 15-22, 25, Non-energetic ores Chemicals and chemical products 24 Other non-metallic mineral products 26 Basic metals and fabricated metal products Investment goods Construction 45 Trade, hotels and restaurants Transport, storage and communication Financial intermediation Real estate, renting and business activities; Other 70-74, community, social and personal service activities Public administration and education; private 75, 80, 95 households with employed persons Health and social work 85 Five different scenarios concerning emissions of greenhouse gases, acidifying emissions and photochemical emissions by the Belgian industries were simulated. On the basis of the NAMEA Air database for the years projections were calculated until We found that without any further changes to the pollution coefficients as of 2001, air pollution of all three types would rise by around 30% by If the evolution from the past was extrapolated into the future, greenhouse gas emissions would still increase by almost 14% and acidifying emissions by a little over 2%. The level of photochemical emissions would drop by just over 3%. For acidifying and photochemical emissions these results were better than in the scenario in which the three Belgian regions were assumed to acquire the best domestically available 2001 technology by For greenhouse gases the opposite was true. Applying the best domestically available 2001 technology in 2015 would lead to lower emissions of greenhouse gases than if ecoefficiency gains were repeated in the period. This shows that these gains were rather modest. In the case of the catch-up scenario in which the three Belgian regions are assumed to acquire the best domestically available 2015 technology, calculated by extrapolating the evolution from the past into the future, all three types of pollution would fall, greenhouse gases by somewhat more than 7%, acidifying and photochemical emissions by around 18%. For greenhouse gases and acidifying emissions this decrease is still smaller than if the Belgian industries were to acquire the best internationally available 2001 technology by 2015, where the international scope is limited to three of Belgium s neighbouring countries, the UK, the Netherlands and Germany. The best domestically available technology in 2015 would in other words not match up to the best internationally available technology in This shows that there should be room for improvement of the eco-efficiency in several industries by means of the application of foreign technologies and production processes. For photochemical emissions application xv

22 of the best internationally available 2001 technology would not lead to better results than the application of the best domestically available 2015 technology. Table J shows the results by industry for this final scenario. Table J: change in air pollution by industries in Belgium by application of the best available international technology in 2001 (in %). Greenhouse gases Acidifying emissions Photochemical emissions Agriculture, hunting and forestry Consumer goods Energy Non-energetic ores Chemicals and chemical products Other non-metallic mineral products Basic metals and fabricated metal products Investment goods Construction Trade, hotels and restaurants Transport, storage and communication Financial intermediation Real estate, renting and business activities; Other community, social and personal service activities Public administration and education; private households with employed persons Health and social work Total Decomposition analysis of changes in CO2 emissions in Belgium Recently the Belgian environmental accounts were extended with a NAMEA Energy 17, which contains energy supply and energy use data by industry. Since air pollution, and especially CO 2 emissions, are closely linked to energy use, this addition to the Belgian environmental accounts allows us to perform a decomposition analysis of the changes in air polluting emissions by the Belgian industries, which are available in the NAMEA Air for the period 1990/ By means of decomposition analysis one can investigate the factors driving the evolution of air pollution. It allows us to discern the impact on air pollution of volume changes from the impact of efficiency changes. These efficiency changes can be linked to either energy efficiency or the emission intensity of the energy mix. Concurrently, the impact of changes in the structure of the economy can be taken into account. 17 See: Gilis S., Janssen L. and G.Vandille (2006) xvi

23 The changes in Belgian CO 2 emissions were decomposed into four underlying causes. These causes are economic growth, measured by GDP, changes in the structure of the economy, measured by the share of value added of the different industries in GDP, the energy intensity of value added creation, and the emission intensity of the energy used. The starting equation looks as follows: POL = POL E E VA VA GDP GDP With: POL = physical amount of pollution in (thousands of) tons E = total energy used in TeraJoules VA = value added in millions of euros GDP = gross domestic product This equation was applied to the 34 industries for which consistent air pollution data exist in the NAMEA Air for Belgium 18, as well as to a set of 3 aggregated industries (primary goods, manufacturing industries, and services), of which the last is further disaggregated into the energy industry, the transport industry and other services. At this disaggregated level the sum of the last two explanatory variables shows the impact on pollution of economic growth of the industry. Economic growth of the industry consists of two components, namely economic growth of the entire economy (GDP), and the change of the share of the industry in total value added (VA/GDP). Summing the impacts of the latter over all industries results in the impact of the change of the economic structure on pollution. The decomposition equation is also applied to the total economy directly. In that case the third explanatory variable is always equal to one and will not play any role in the decomposition. The interpretation of the energy and the emission intensity is the same whether separate industries are considered or the entire economy. The impact of changes in emission intensity should be interpreted as following from a change of the energy mix, as long as no technology is used to capture CO 2 emissions. The latter is currently not yet the case in Belgium. Table K shows the impact of the four explanatory variables on the percentage change of emissions of CO 2 by all industries in Belgium during the period. The decomposition was made year by year, after which the yearly effects were summed. The same result is obtained when the decomposition is performed on the data for the years 1990 and 2002 directly. The table shows the results using different levels of disaggregation. The first row shows the results when the decomposition is performed on data for the entire economy directly. The second row shows the results when primary producers are discerned from manufacturing industries and services sectors, after which the effects for these three industries are summed. The final row shows the results when the decomposition is applied to the 34 industries for which consistent data exist in the NAMEA Air and Energy, after which the sum is taken over the 34 effects. 18 See table 26 in chapter IV xvii

24 Table K: Effect on total industry CO 2 emissions growth between 1990 and 2002 Emission intensity Energy intensity Economic structure Economic growth No disaggr Low disaggr NAMEA disaggr CO 2 emissions by the Belgian industries increased by 2.4% between 1990 and However, according to economic growth during that period CO 2 emissions by the Belgian industries should have grown by almost 23%. Changes in the other variables must thus have reduced CO 2 emissions. This is indeed the case. The emission intensity shows that the energy used by the Belgian industries has become around 8% less CO 2 intensive in the period under consideration. The highest impact (-8.3%) is measured at the lowest level of disaggregation, because part of the impact of changes in the economic structure is picked up by the emission intensity. As a corollary of this, the lowest impact (-7.5%) is measured at the highest level of disaggregation. The level of disaggregation seems to have the biggest impact on the effect of changes in energy intensity. According to the direct decomposition of the data for the entire economy, the decrease of the energy intensity brought down CO 2 emissions by 12%. However, when we use the NAMEA disaggregation the impact of the drop in energy intensity is reduced to only 7%. The change in the economic structure between 1990 and 2002 has reduced CO 2 emissions by almost 6%, of which somewhat less than half was due to changes between primary producers, manufacturing industries and services sectors, while the rest was due to changes within these three aggregated industries. The fact that the level of disaggregation has such an important effect on the impact of energy intensity shows that the restructuring of the Belgian economy has favoured less energy intensive industries. Table L shows the impact of the four explanatory variables on CO 2 emissions by the three aggregated industries. Two sets of results are presented, one in which the decomposition was performed on the data for the three industries directly, and a second in which the decomposition was performed on the data of the 34 NAMEA industries, after which the effects were summed in order to achieve the impact on the three aggregated industries. The impact of economic growth between 1990 and 2002 was approximately equal for the primary and manufacturing industries, while for the services industry it was somewhat higher. 19 The impact of the change in the economic structure was negative for the manufacturing industry (-7.6%) and positive for the primary (6.5%) and the services industry (1.9%). The impact of the change in the economic structure is accentuated when the data are analysed at a higher level of disaggregation. The negative impact on CO 2 emissions by the manufacturing industry is more than doubled, the positive impact on 19 Remark that economic growth stands for total GDP growth in Belgium. It does not represent economic growth of the different industries. The impact of economic growth of the different industries is the sum of the impact of total GDP growth on the one hand and the impact of the change in the industry s share in value added, or the economic structure, on the other. It is thus quite normal that the impact of the fourth explanatory variable is similar across industries. xviii