retour à l'accueil nouvelle fenêtre vers www.appa.asso.fr Pollution atmosphérique, climat, santé, société

Articles

Modeled effects of an improved building insulation scenario in Europe on air pollution, health and societal costs

Calculs des effets sur la pollution atmosphérique, la santé et l’économie d’un programme d’amélioration de l’isolation des bâtiments en Europe

Jakob Hjort Bønløkke, Gitte Juel Holst, Torben Sigsgaard, Ulrik Smith Korsholm, Bjarne Amstrup, Iratxe Gonzalez-Aparicio et Jens Havskov Sørensen

p. 1-16

[Version imprimable] [Version PDF]

Résumé

Contexte : en Europe, une partie importante de l'énergie produite est utilisée pour le chauffage domestique et pour la climatisation. La qualité de l'isolation des bâtiments a ainsi un impact significatif sur la pollution de l'air.
Objectifs : modéliser et calculer les effets d'une amélioration importante de l'isolation des bâtiments existants en Europe sur les niveaux de pollution de l'air, sur la santé et sur l'économie.
Méthodes : l'énergie utilisée dans deux scénarios différents a été comparée entre 2005 et 2020 : un scénario d’un programme de l'isolation des bâtiments existants en Europe et un scénario de statu quo. Les variations des émissions issues de ces deux scénarios ont été intégrées dans un modèle de la qualité de l'air (the Comprehensive Air-Quality Model with extensions). Les variations annuelles moyennes des principaux polluants atmosphériques ont été calculées pour chaque pays. Des données venant de l’Organisation Mondiale de la Santé (OMS) et de l'Union Européenne (UE) sur les populations et sur les impacts des polluants ont été utilisées pour déduire quels sont les effets sur la santé et l’économie. La qualité de l'air intérieur ne faisait pas partie de l’étude.
Résultats : avec le programme de l'isolation des bâtiments existants en Europe, les niveaux moyens annuels de la pollution atmosphérique particulaire fine (PM2,5) variaient de -0,008 µg/m3 (Finlande) à -0,538 µg/m3 (Belgique). Le nombre moyen d'années de vie gagné par année par 100 000 adultes était de 24,3 (intervalle de confiance 95 % de 0,9 à 54,5). Le nombre total d'années de vie gagnées chaque année variait, selon les pays, entre 31 en Finlande à 22 524 en Allemagne. Le nombre total d'années de vie gagnées était de 78 678 en Europe. Un total de 7 173 cas de bronchite chronique pourrait être évité chaque année. Plusieurs autres effets sur la santé étaient améliorés de façon similaire. Les coûts pour la société s’élevaient à 6,64 milliards d’euros par an.
Conclusions : en plus de la réduction des émissions de carbone, un programme de l'isolation des bâtiments existants en Europe aurait des avantages substantiels sur la santé grâce à l’amélioration de la pollution atmosphérique. Les effets sur la santé et sur l’économie peuvent contrebalancer de façon significative les coûts d'investissement et devraient être pris en compte lors de l'évaluation des stratégies d'atténuation du réchauffement climatique.

Abstract

Background: In Europe a substantial share of the energy supply is used for domestic heating and cooling. The quality of building insulation thus significantly impacts air pollution.
Objectives: To model the effects of an improved building insulation scenario in Europe on air pollution levels and the resulting effects on health and economy.
Methods: Projected energy savings between 2005 and 2020 were calculated for an improved building insulation scenario and a business as usual scenario. The resulting changes in emissions (e.g. from power plants) were used in the Comprehensive Air-Quality Model with extensions. Mean annual changes in the main air pollutants were derived for each country. World Health Organization (WHO) and European Union (EU) data on populations and on impacts of pollutants were used to derive health effects and costs. Effects on indoor air quality were not assessed.
Results: Projected effects on the mean annual change in PM2.5 varied from −0.008 μg/m3 (Finland) to −0.538 μg/m3 (Belgium). The mean number of life-years (LY) gained annually per 100000 adults was 24.3 LY (range 0.9 to 54.5). The total number of LY gained annually varied from 31 in Finland to 22524 in Germany, totaling 78678 LY in Europe. A total of 7173 cases of persistent chronic bronchitis could be avoided annually. Several other health outcomes improved similarly. The saved societal costs totaled 6.64 billion € annually.
Conclusions: In addition to carbon emission reductions, an improved building insulation scenario in Europe would have substantial benefits on health through improvements in air pollution. Health effects and societal cost savings may significantly counterbalance investment costs and should be taken into account when evaluating strategies for mitigation of global warming.

Entrées d'index

Mots-clés : externalités, isolation des bâtiments, morbidité, mortalité, pollution atmosphérique particulaire fine, pollution de l’air

Keywords: air pollution, building insulation, externalities, fine particulate matter, morbidity, mortality

Texte intégral

1. Background

Public health benefits and reduced societal expenses have been largely missing features of the energy consumption and greenhouse gas policies, despite the publication of several “costs of air pollution-related ill health” studies (Pervin et al., 2008; Haines and Dora, 2012). For more than three decades such studies have consistently suggested substantially improved public health and saved costs to society from reductions in air pollution (Zmirou et al., 1999) and impacts and costs have been thoroughly reviewed and estimated (Bickel and Friedrich, 2005; Holland et al., 2005; Hurley et al., 2005). More recently, studies have demonstrated the potential co-benefits to health and economy World-wide from actions to mitigate greenhouse gas emissions (Haines et al., 2009). Control of fossil-fuel particulate black carbon was suggested to be an effective means to slow global warming as well as to improve health in 2002 (Jacobson, 2002), an idea pursued in recent publications (Anenberg et al., 2012; Bond et al., 2013) although without calculation of costs.

The share of energy that is used for domestic heating and cooling is substantial. Thus, building insulation may affect air pollution and public health significantly through changed energy demands leading to changes in air pollution. This connection has rarely been investigated in contrast to studies on energy consumption, job creation, and on carbon dioxide (CO2), which are common. Levy et al. (2003) estimated the effects of insulation retrofits (to IECC 2000 insulation levels) in existing housing in the United States on ambient pollutant emissions, public health, and the corresponding saved societal costs whereas Wilkinson et al. (2009) estimated the effects in the UK of interventions to improve the energy efficiency of heating of the housing stock on indoor environment and subsequent health effects. Both studies demonstrated appreciable potential for improved public health owing to the scenarios they investigated.

We have previously described the projected changes in major air pollutants in 6 zones of Europe resulting from an improved building insulation scenario and given details on the emission estimation, modeling of air pollutants, and test of quality of predictions (Korsholm et al., 2012). Particulate matter reductions was found to vary from 1.2% in north-eastern Europe to 9% in north-western Europe and we hypothesized that in some countries the health effects of these changes would be substantial. Health effects other than those related to improved ambient air may occur as a result from improved building insulation. When building insulation is altered, indoor air quality may be altered too, both due to changed infiltration rates and due to changed behavior of dwellers. It is hard to predict the extent of such changes and the balance between positive changes (e.g. improved thermal comfort, less indoor wood smoke) and the negative ones (e.g. decreased ventilation and increased humidity and risk of mold growth). It is out of scope of this paper to model health effects due to changes in indoor air quality resulting from the insulation scenario although possible effects will be discussed.

Our primary aim was to illustrate the likely extent of improvements to public health through changes in criteria air pollutants at a regional scale from an ambitious building insulation retrofit and new building scenario – extending these from the 6 zones considered in our previous work to all countries in the region considered. A secondary aim was to calculate the range of externalities in terms of economic savings associated with the health effects. We only considered health effects for which there is broad consensus on the impact and the associated costs and did not consider damage to crops or infrastructure from air pollution. Compared with the relatively few previous studies on health effects of changes to building insulation our study differed by covering a larger region and population; by being based on an air-quality model providing details on criteria pollutant concentrations; by applying life-table analyses at the national level; and by including extensive sensitivity analyses assuming different impacts and cost.

2. Methods

The study comprised 25 European Union (EU-25) states: Finland, Sweden, Austria, Belgium, Denmark, Germany, Ireland, France, Luxembourg, Netherlands, United Kingdom, Italy, Spain, Greece, Portugal, Estonia, Latvia, Lithuania, Poland, Romania, Slovakia, Slovenia, Bulgaria, Czech Republic, and Hungary. In these states, an improved insulation scenario was compared with a business-as-usual scenario and the scenario and the methods applied have been described previously (Korsholm et al., 2012). Briefly, the building insulation scenario ran from 2005 to 2020 with an annual retrofit rate of 2% assuming ambitious insulation levels in new and retrofit buildings. Thus decreasing energy expenditure on heating and cooling was compared with a scenario assuming no changes to current insulation and retrofit practices in Europe. The improved insulation scenario considered roof, wall and floor insulation only and did not include windows, ventilation systems, etc. Thermal conductivity values came from Ecofys (ECOFYS, 2007) and were projected to decrease in all regions and by more than 50% in the regions with poorest insulation (Korsholm et al., 2012). All other variables than the retrofit rate and insulation efficiency were kept constant at the 2005 level; e.g. energy source mix for heating and construction rate of new buildings.

2.1. Air pollution data

Mean annual changes for the 15-year period in the main air pollutants particulate matter less than 10 µm (PM10) and less than 2.5 µm (PM2.5), sulphur dioxide, nitrogen oxides, carbon monoxide, ozone, and volatile organic compounds were calculated in the Comprehensive Air-Quality Model with extensions (CAMx) by modeling emissions, emission changes in the two scenarios, atmospheric chemistry and meteorology (Korsholm et al., 2012). Meteorological data from 2009 was used as this was found to be closest to the European normal. Model predictions were controlled by use of data from 8 measurements stations, finding correlations between modeled and measured data of 0.67-0.68 for ozone. Ozone was converted from µg/m3 to ppb by multiplying with 0.5097 assuming an approximative summer temperature of 25°C. In no case did changes in PM10 differ from PM2.5 as the entire change in PM was contained within the PM2.5 fraction of PM10 (Korsholm et al., 2012). Concentration changes were averaged over each country and concentration-response functions applied to the total population as if evenly distributed and exposed to air pollutant concentration changes.

2.2. Population data

Population, morbidity, and mortality data were extracted at the national level from the European Detailed Mortality Database. For infant mortality (between 1 and 12 months) the source was the European Health for All Database (WHO Regional Office for Europe, Copenhagen, DK).

Data from the most recent year were preferred as available in April 2012. Data from 2010 were available for 5, from 2009 for 10, from 2008 for 5, and from 2007 and 2005 for one country each.

2.3. Concentration-response functions (CRF)

CRF are relative risks (RR) describing changes in existing risks associated with measured or modeled concentrations of air pollutants in a population. They are not real dose- or exposure-response functions because the concentrations are based on mean outdoor levels not taking peoples exact location into account. CRF used in this study were applied to the population of each country separately and effects summed. The factors were based on CRF from the literature on mortality and selected health endpoints (morbidity) corresponding to the factors endorsed by the European Commission DG Environment Clean Air for Europe Program (CAFE) (Watkiss et al., 2005) and the more recent EU program Health and Environment Integrated Methodology and Toolbox for Scenario Development (HEIMTSA, 2011). The same CRF are used by the European Environment Agency and several of them by the OECD (OECD, 2007) and the US EPA (US-EPA, 2012).

2.4. Mortality

Mortality from all causes among adults aged 30+ years was assumed to change linearly. Based on the RR of 6% (95% CI 2;11%) per 10 μg/m3 change in mean annual PM2.5 observed in the American Cancer Society Study (Pope et al., 2002) the CRF is given by 1.006(−1/10) for a 10 μg/m3 reduction in PM2.5. None of the other pollutants investigated are considered to have separate impacts on long-term mortality (HEIMTSA, 2011).

For mortality in infancy (from age 1 to 12 months) a linear function with a RR of 0.4% (95% CI 0.2; 0.7%) per μg/m3 change in mean annual PM10 was applied as in (Woodruff et al., 1997).

For effects of ozone, the RR published by Jerrett et al. (2009) on respiratory mortality among adults aged 30+ years of 0.4% (95% CI 0.1%;0.67%) per ppb was used. As this factor applies to the annual mean of the daily 1 h maximum that was (71.62/56.68) 1.26 times higher than the annual daily mean, the modeled ozone concentration changes were multiplied by this factor in order to account for the difference.

For calculation of changes in LY the life table method described by Miller and Hurley (2003) was applied using life-tables from the Institute of Occupational Medicine (Miller, 2011). Specifically, an IOMLIFET ALL_CAUSE table was used for each country individually entering country-specific data on demography, all-cause mortality and modeled air pollution changes as described previously. For modeling of effects of changes in ozone the IOMLIFET MULTI_CAUSE table was used, entering mortality from all respiratory causes (ICD J00-J99) only. Because effects have been documented in the summer season only and ozone was modeled for entire calendar years, the CRF was halved in order to yield halved effects.

2.5. Morbidity

Morbidity effects considered in the analysis were chronic bronchitis (ICD J40-42), cardiac (ICD I00-52) and respiratory (ICD J00-99) emergency admissions, restricted activity days (RAD), use of medication for respiratory disease, and lower respiratory symptoms.

Effects on morbidity were taken from the EU program (HEIMTSA, 2011), assuming similar baseline incidences, similar employment rates, and similar linear CRFs in the 25 states assessed as in the European studies that formed the basis of the methodology.

Thus, the CRF for incidence of chronic bronchitis was calculated as a 2.2% increase per μg/m3 change in mean annual PM10, an annual incidence of 0.39% among adults and that 90% of the population did not have persistent chronic bronchitis. This corresponds to 7.7 (95%: 0.7; 14) new persistent cases annually of chronic bronchitis per μg/m3 PM10 per 100000 adults aged 18+ years.

For emergency cardiac hospital admission rates the CRF was calculated as 0.43 (95% CI: 0.22; 0.65) additional admissions per μg/m3 increase in PM10 per 100000 total population annually. For emergency respiratory hospital admissions rates the CRF was calculated as 0.56 (95% CI: 0.43; 0.62) additional admissions per μg/m3 increase in PM10 per 100000 total population annually.

The CRF for RAD was 9020 (95% CI: 7920; 10130) additional RAD per μg/m3 increase in PM2.5 per 100000 adults aged 18-64 annually.

The CRF for bronchodilator use among children with asthma was 210 (95% CI: −890; 1400) additional days of bronchodilator usage per μg/m3 increase in PM10 per 100000 children aged 5-14, per year. Among adults with asthma the CRF was 930 (95% CI: −930; 2800) additional days of bronchodilator usage per μg/m3 increase in PM10 per 100000 adults aged 20+ years annually.

The CRF for days with lower respiratory symptoms (LRS) among children was 18600 (95% CI: 9310; 27900) additional LRS including cough days per μg/m3 increase in PM10 per 100000 children aged 5-14 annually. Among adults the corresponding figure was 3 900 (95% CI: 330; 7200) additional LRS including cough days per μg/m3 increase in PM10, per 100000 adults annually.

HEIMTSA operates with additional impacts (e.g. primary care consultations, work loss days, minor restricted activity days) but as these are originally considered (Hurley et al., 2005) as secondary or sensitivity functions we did not include them.

  

2.6. Valuation

Table I. Valuation (€) of the health effects quantified in the study.
Estimation (€) des effets sur la santé dans l’étude.

Valuation (median)

Valuation (mean)

Mortality (deaths, VSL)

1018000b

2080000b

Mortality (life years lost, VOLY)

125000b

54000a

Infant mortality (deaths, VSL)

1503000a

3060000b

Chronic bronchitis (cases)

208000a

Hospital admissions

2364a

Restricted activity days in working age

97a

Respiratory medication use all ages

1a

Lower respiratory symptoms in people with chronic symptoms (all ages)

42a

a core analysis. b sensitivity analyses.

Mortality and morbidity was valued economically at 2005 values. If not stated otherwise this was done in accordance with the methodology of the CAFE program (Hurley et al., 2005) as listed in table 1. For the core analyses the mean VOLY (value of a life-year) from CAFE was applied to LY lost due to chronic exposure among adults, whereas the infant mortality valuation was conducted by use of the median VSL (value of a statistical life) times the mean marginal rate of substitution (of 1.5) as in the CAFE Program (Hurley et al., 2005).

2.7. Sensitivity analyses

As suggested in the CAFE program, both the mean and median VSL and the VOLY approaches were used for sensitivity analyses for the sake of transparency. Similarly, mean marginal rates of substitution of 1.0 and 2.0 were applied in sensitivity analyses of the value of saved lives of infants.

For sensitivity analyses on the percent change in all-cause adult mortality the proposed 75% plausibility interval of a CRF between 0.1% and 1.2% per μg/m3 increase in annual PM2.5 suggested by the expert elicitation of COMEAP (COMEAP 2009) was used. Also the central estimates on the percent change in all-cause mortality ranging from 0.7 to 1.6 % per μg/m3 increase in annual PM2.5 suggested to the EPA by a panel of experts (US-EPA, 2006) was applied.

For the remaining health end-points the 95% CI as given by HEIMTSA were used for sensitivity analyses.

3. Results

The population in the 25 EU states totaled 495.7 million of which 4.9 million were infants and 324.5 million were adults aged 30+ years - the ages in which changes in mortality from air pollution was calculated.

Figure 1. Changes (decreases) in mean air pollutant concentrations from the improved building insulation scenario in the EU-25 states.
Changements dans les niveaux des moyennes annuelles des polluants dans les 25 pays européens avec le programme d’isolation des bâtiments.

The changes in main pollutant concentrations that have previously only been published by region are visualised in Figure 1. The mean annual change in PM2.5 caused by the insulation scenario varied almost two orders of magnitude between −0.008 μg/m3 (Finland) and −0.538 μg/m3 (Belgium).

Accordingly the number of LY gained among adults 30+ years varied greatly from 0.9/100000 persons/year in Finland to 54.5/100000 persons/year in Belgium - the unweighted 25-country mean being 24.3/100000/year.

Table II. Annual PM change and gain in life-years (LY) in the EU-25 states due to the insulation scenario.
Les changements du niveau des moyennes annuelles de PM2,5 et du nombre d'années de vie gagnées dans les 25 pays européens avec le programme d’isolation des bâtiments.

PM change

Population

LY

LY/100000

adults

LY

at 0.1%/µg*m−3

LY

at 1.6%/µg*m−3

Finland

−0.008

5.34E+06

31

0.9

5

79

Sweden

−0.018

9.38E+06

98

1.6

17

249

Austria

−0.098

8.39E+06

528

9.5

90

1338

Belgium

−0.538

1.05E+07

3712

54.8

630

9363

Denmark

−0.204

5.45E+06

719

20.5

121

1801

Germany

−0.383

8.18E+07

22618

39.9

3850

57277

Ireland

−0.175

4.46E+06

374

14.7

63

941

France

−0.205

6.21E+07

8236

21.1

1397

20811

Luxembourg

−0.367

4.89E+05

94

30.4

16

239

Netherlands

−0.530

1.66E+07

5045

47.2

857

12742

United Kingdom

−0.300

6.18E+07

11752

30.3

1990

29618

Italy

−0.076

5.98E+07

2927

7.0

498

7418

Spain

−0.102

4.56E+07

2771

9.1

470

7006

Greece

−0.057

1.13E+07

395

5.1

67

999

Portugal

−0.089

1.06E+07

629

8.9

107

1593

Estonia

−0.030

1.34E+06

35

4.2

6

89

Latvia

−0.041

2.25E+06

89

6.2

15

226

Lithuania

−0.075

3.34E+06

233

11.2

39

589

Poland

−0.417

3.82E+07

12273

52.1

2083

30986

Romania

−0.126

2.14E+07

2250

16.4

379

5644

Slovakia

−0.174

5.42E+06

670

20.0

113

1686

Slovenia

−0.084

2.04E+06

117

8.6

20

297

Bulgaria

−0.068

7,62E+06

471

9.3

80

1199

Czech Republic

−0.190

1,05E+07

1404

20.4

240

3574

Hungary

−0.182

1.00E+07

1714

26.1

293

4363

All states

4.957E+08

78678

24.25a

13446

200126

PM change in µg/m3. LY gain in years. a unweighted mean

As table II shows, the total number of LY gained annually varied to an even greater extent from 31 in Finland to 22524 in Germany, totaling 78678 LY in the EU-25.

Figure 2. Annual changes (gains) in life-years (LY) following the improved building insulation scenario in the EU-25 states. In blue : totals (left axis); in red : per 100 000 adults (right axis).
Changements dans les nombres d'années de vie gagnées dans les 25 pays européens avec le programme d’isolation des bâtiments. En bleu (axe de gauche), les nombres totaux ; en rouge (axe de droite), les nombres pour 100 000 adultes.

In figure 2 this is further illustrated by including LY/100000 inhabitants/year. The value of saved LY among adults in Europe was 4.25 billion €/year (table IV).

Among infants the total number of avoided deaths was 7/year. The societal costs saved amounted to 10.4 million €/year (table IV).

A total of 7173 cases of persistent chronic bronchitis were avoided annually, varying from 3 in Finland and Estonia to 2000 in Germany. The annual net gain from this reduction was 1.49 billion €/year.

A total of 1142 emergency admissions were avoided annually, of which 1/4 would have occurred in Germany. The saved costs amounted to 2.7 million €/year.

Regarding changes in number of RAD among adults and days with LRS among children and adults, these totaled −6.6E+06/year, −94.4 E+06/year, and −152.6 E+06/year respectively.

The changes in number of days with bronchodilator use among children and adults totaled −25398/year and −840994/year respectively and resulted in total savings of 866,391 €/year.

Table III. Annual change in morbidity in the EU-25 states due to the insulation scenario.
Les changements de morbidité dans les 25 pays européens avec le programme d’isolation des bâtiments.

Chronic bronchitisa

Emergency admissions

RADb

Days on medicationc

LRSd

Finland

-3 (0,-5)

0 (0,-1)

-2.6 (-2,-3)

-0.3 (0,-1)

-2.3 (-1,-4)

Sweden

-10 (-1,-18)

-2 (-1,-2)

-9.1 (-8,-10)

-1.2 (1,-4)

-8.4 (-2,-14)

Austria

-52 (-5,-94)

-8 (-5,-10)

-47.6 (-42,-54)

-6.3 (7,-20)

-41.8 (-10,-72)

Belgium

-344 (-31,-625)

-56 (-37,-72)

-315.3 (-277,-354)

-41.7 (46,-130)

-295.9 (-76,-504)

Denmark

-66 (-6,-121)

-11 (-7,-14)

-62.5 (-55,-70)

-8.1 (9,-25)

-59.8 (-16,-101)

Germany

-2011 (-183,-3656)

-310 (-203,-397)

-1773 (-1557:-1991)

-242.9 (263,-754)

-1557 (-356,-2689)

Ireland

-45 (-4,-82)

-8 (-5,-10)

-45.2 (-40,-51)

-5.5 (6,-17)

-42.4 (-12,-72)

France

-765 (-70,-1391)

-126 (-83,-162)

-705 (-619,-792)

-92.6 (103,-291)

-676 (-177,-1147)

Luxembourg

-11 (-1,-20)

-2 (-1,-2)

-10.4 (-9,-12)

-1.3 (2,-4)

-9.6 (-3,-16)

Netherlands

-534 (-49,-972)

-87 (-57,-112)

-503.3 (-442,-565)

-64.8 (72,-203)

-466.4 (-121,-793)

United Kingdom

-1123 (-102,-2042)

-184 (-120,-225)

-1042.1 (-915,-1170)

-135.6 (150,-425)

-960 (-244,-1637)

Italy

-291 (-26,-528)

-45 (-30,-58)

-258.1 (-227,-290)

-35.1 (38,-109)

-226 (-52,-390)

Spain

-295 (-27,-536)

-46 (-30,-59)

-275.9 (-242,-310)

-35.7 (39,-111)

-231.1 (-54,-399)

Greece

-41 (-4,-74)

-6 (-4,-8)

-36.7 (-32,-41)

-4.9 (5,-15)

-31.6 (-7,-55)

Portugal

-59 (-5,-108)

-9 (-9,-12)

-54.3 (-48,-61)

-7.2 (8,-22)

-48.1 (-12-83)

Estonia

-2 (0,-5)

0 (0,-1)

-2.3 (-2,-3)

-0.3 (0,-1)

-2 (-1,-3)

Latvia

-6 (-1,-11)

-1 (-1,-1)

-5.4 (-5,-6)

-0.7 (1,-2)

-4.4 (-1,-8)

Lithuania

-15 (-1,-28)

-2 (-1,-3)

-14.5 (-13,-16)

-1.9 (2,-6)

-12.6 (-3,-22)

Poland

-988 (-90,-1796)

-157 (-103,-202)

-963.9 (-846,-1083)

-118.7 (130,-370)

-802 (-193,-1375)

Romania

-169 (-15,-307)

-27 (-17,-34)

-161.6 (-142,-181)

-20.4 (22,-64)

-135.9 (-33,-233)

Slovakia

-58 (-5,-106)

-9 (-8,-12)

-58 (-51,-65)

-7 (8,-22)

-47.6 (-12,-82)

Slovenia

-11 (-1,-20)

-2 (-1,-2)

-10.3 (-9,-12)

-1.3 (1,-4)

-8.5 (-2,-15)

Bulgaria

-33 (-3,-60)

-5 (-3,-7)

-30.8 (-27,-35)

-4 (4,-12)

-25.3 (-6,-44)

Czech Republic

-126 (-11,-229)

-20 (-13,-25)

-120.5 (-106,-135)

-15.1 (16,-47)

-96.7 (-22,-167)

Hungary

-114 (-10,-208)

-18 (-12,-23)

-106.7 (-94,-120)

-13.8 (15,-43)

-91.4 (-22,-157)

All states

-7173

(-652,-13041)

-1142

(-750,-1464)

-6615

(-5808,-7429)

-866.4

(949,-2701)

-5882

(-1433, -10081)

95% CI in brackets. a Change in annual incidence of the disease. b Restricted activity days x 1000. c x 1000; all ages included. d Days with lower respiratory symptoms; all ages included.

­

Table IV. Saved societal costs (€) in the EU-25 states due to the insulation scenario.
Coûts économisés (€) dans les 25 pays européens avec le programme d’isolation des bâtiments.

Adult life-years

Infant death

Chronic bronchitis

Emergency admissions

RADsa

LRSb

All health effectsc

Finland

1.68E+06

2.77E+03

5.76E+05

1.06E+03

2.48E+05

9.81E+04

2.6E+06

Sweden

5.28E+06

1.17E+04

2.09E+06

3.85E+03

8.78E+05

3.51E+05

8.61E+06

Austria

2.84E+07

5.49E+04

1.08E+07

1.93E+04

4.62E+06

1.75E+06

4.56E+07

Belgium

1.99E+08

5.45E+05

7.15E+07

1.32E+05

3.06E+07

1.24E+07

3.14E+08

Denmark

3.82E+07

2.19E+05

1.38E+07

2.60E+04

6.06E+06

2.51E+06

6.08E+07

Germany

1.22E+09

1.8E+06

4.18E+08

7.32E+05

1.72E+08

6.54E+07

1.87E+09

Ireland

2.00E+07

7.1E+04

9.42E+06

1.83E+04

4.39E+06

1.78E+06

3.56E+07

France

4.42E+08

1.11E+06

1.59E+08

2.98E+05

6.84E+07

2.84E+07

6.99E+08

Luxembourg

5.07E+06

9.03E+03

2.25E+06

4.20E+03

1.01E+06

4.03E+05

8.75E+06

Netherlands

2.71E+08

5.95E+05

1.11E+08

2.06E+05

4.88E+07

1.96E+07

4.51E+08

United Kingdom

6.29E+08

2.1E+06

2.34E+08

4.34E+05

1.01E+08

4.03E+07

1.01E+09

Italy

1.57E+08

2.69E+05

6.05E+07

1.06E+05

2.5E+07

9.49E+06

2.53E+08

Spain

1.49E+08

3.7E+05

6.14E+07

1.09E+05

2.68E+07

9.71E+06

2.47E+08

Greece

2.12E+07

4.52E+04

8.44E+06

1.49E+04

3.56E+06

1.33E+06

3.46E+07

Portugal

3.38E+07

7.68E+04

1.23E+07

2.21E+04

5.26E+06

2.02E+06

5.35E+07

Estonia

1.90E+06

4.42E+03

5.19E+05

9.34E+02

2.24E+05

8.24E+04

2.73E+06

Latvia

4.79E+06

1.69E+04

1.20E+06

2.14E+03

5.2E+05

1.85E+05

6.72E+06

Lithuania

1.25E+07

3.77E+04

3.21E+06

5.83E+03

1.41E+06

5.28E+05

1.77E+07

Poland

6.58E+08

1.76E+06

2.05E+08

3.72E+05

9.35E+07

3.37E+07

9.93E+08

Romania

1.20E+08

6.93E+05

3.51E+07

6.30E+04

1.57E+07

5.71E+06

1.77E+08

Slovakia

3.58E+07

1.59E+05

1.21E+07

2.20E+04

5.63E+06

2,00E+06

5.57E+07

Slovenia

6.30E+06

9.08E+03

2.29E+06

4.03E+03

1,00E+06

3.56E+05

9.96E+06

Bulgaria

2.54E+07

1.22E+05

6.91E+06

1.22E+04

2.99E+06

1.06E+06

3.65E+07

Czech Republic

7.58E+07

1.45E+05

2.62E+07

4.65E+04

1.17E+07

4.06E+06

1.18E+08

Hungary

9.26E+07

1.72E+05

2.37E+07

4.25E+04

1.04E+07

3.84E+06

1.31E+08

All states

4.25E+09

1.04E+07

1.49E+09

2.7E+06

6.42E+08

2.47E+08

6.64E+09

a Restricted activity days.
b Days with lower respiratory symptoms; all ages included.
c including days on medication.

The sum of societal costs saved by declining morbidity amounted to 2.38 billion €/year (table IV). Detailed information with 95% CI and split by nation is given in table III.

In total, summing up mortality and morbidity effects among all age groups, the total economical savings in the 25 states due to the insulation scenario was 6.64 billion €/year.

The projected changes in ground ozone were infinitesimal and the effects on life expectancy and economy similarly indiscernible from zero.

3.1. Sensitivity analyses

In the sensitivity analysis performed with median rather than mean VOLY, the saved societal cost associated with adult mortality amounted to 9.84 billion €/year, a 48% increase from the core analysis. In the analyses based on mean and median VSL the saved societal costs amounted to 13.67 and 6.69 billion €/year respectively; i.e. either a doubling or no significant change from the core analysis.

The sensitivity analyses performed with the mean rather than median value of a saved infant’s life resulted in a 2-fold increase, i.e. saved costs of 20 million €/year, increasing to 26 million €/year when using the mean marginal rate of substitution of 2.0.

The sensitivity analyses performed with the CRF ranges for adult mortality suggested by the COMEAP and EPA expert panels resulted in saved costs of 726 million €/year with a CRF of 0.1%  per μg/m3 increase in annual PM2.5 and 10807 million €/year with a CRF of 1.6%  per μg/m3 increase in annual PM2.5. These extremes correspond to 11-163% of the core analysis costs.

Table V. Sensitivity analysis on adult mortality from PM2.5 exposure. Life years gained annually due to insulation scenario.
Analyse de sensibilité sur la mortalité des adultes due aux PM2,5. Nombre d'années de vie gagnées avec le programme d’isolation des bâtiments.

Sens

Core

Sens

Sens

Sens

COMEAP lower bound suggestion

0.1%/µgm-3

0.6%/µgm-3

US EPA lower bound judgment 0.7%/µgm-3

COMEAP upper bound suggestion 1.2%/µgm-3

US EPA upper bound judgment 1.6%/µgm-3

Finland

5

31

36

60

79

Sweden

17

98

114

190

249

Austria

90

526

610

1022

1338

Belgium

630

3684

4277

7156

9363

Denmark

121

708

822

1375

1801

Germany

3850

22524

26149

43763

57277

Ireland

63

370

429

719

941

France

1397

8178

9494

15896

20811

Luxembourg

16

94

104

182

239

Netherlands

857

5014

5820

9738

12742

United Kingdom

1990

11643

13517

22627

29618

Italy

498

2913

3382

5665

7418

Spain

470

2752

3195

5351

7006

Greece

67

392

456

763

999

Portugal

107

626

726

1216

1593

Estonia

6

35

41

68

89

Latvia

15

89

103

172

226

Lithuania

39

231

268

449

589

Poland

2083

12187

14148

23767

30986

Romania

379

2217

2574

4310

5644

Slovakia

113

662

769

1288

1686

Slovenia

20

117

136

227

297

Bulgaria

80

471

547

916

1199

Czech Republic

240

1404

1630

2730

3574

Hungary

293

1714

1990

3332

4363

All states

13446

78678

91339

152983

200126

­

Table VI. Sensitivity analyses on societal costs from adult mortality from PM2.5 exposure. € saved annually with insulation scenario.
Analyse de sensibilité sur les coûts économisés (€ par an) dus au nombre d'années de vie gagnées avec le programme d’isolation des bâtiments.

Sens

Core

Sens

Sens

Sens

COMEAP lower bound suggestion 0.1%/µgm-3

0.6%/µgm-3

US EPA lower bound judgment 0.7%/µgm-3

COMEAP upper bound suggestion 1.2%/µgm-3

US EPA upper bound judgment 1.6%/µgm-3

Finland

2.86E+05

1.68E+06

1.95E+06

3.26E+06

4.27E+06

Sweden

9.02E+05

5.28E+06

6.13E+06

1.03E+07

1.34E+07

Austria

4.85E+06

2.84E+07

3.30E+07

5.52E+07

7.23E+07

Belgium

3.40E+07

1.99E+08

2.31E+08

3.86E+08

5.06E+08

Denmark

6.53E+06

3.82E+07

4.44E+07

7.43E+07

9.72E+07

Germany

2.08E+08

1.22E+09

1.41E+09

2.36E+09

3.09E+09

Ireland

3.41E+06

2.00E+07

2.32E+07

3.88E+07

5.08E+07

France

7.54E+07

4.42E+08

5.13E+08

8.58E+08

1.12E+09

Luxembourg

8.66E+05

5.07E+06

5.64E+06

9.84E+06

1.29E+07

Netherlands

4.63E+07

2.71E+08

3.14E+08

5.26E+08

6.88E+08

United Kingdom

1.07E+08

6.29E+08

7.30E+08

1.22E+09

1.60E+09

Italy

2.69E+07

1.57E+08

1.83E+08

3.06E+08

4.01E+08

Spain

2.54E+07

1.49E+08

1.73E+08

2.89E+08

3.78E+08

Greece

3.62E+06

2.12E+07

2.46E+07

4.12E+07

5.40E+07

Portugal

5.77E+06

3.38E+07

3.92E+07

6.57E+07

8.60E+07

Estonia

3.24E+05

1.90E+06

2.20E+06

3.69E+06

4.83E+06

Latvia

8.17E+05

4.79E+06

5.56E+06

9.31E+06

1.22E+07

Lithuania

2.13E+06

1.25E+07

1.45E+07

2.43E+07

3.18E+07

Poland

1.12E+08

6.58E+08

7.64E+08

1.28E+09

1.67E+09

Romania

2.04E+07

1.20E+08

1.39E+08

2.33E+08

3.05E+08

Slovakia

6.11E+06

3.58E+07

4.15E+07

6.95E+07

9.10E+07

Slovenia

1.08E+06

6.30E+06

7.32E+06

1.23E+07

1.60E+07

Bulgaria

4.34E+06

2.54E+07

2.95E+07

4.95E+07

6.48E+07

Czech Republic

1.3E+07

7.58E+07

8.8E+07

1.47E+08

1.93E+08

Hungary

1.58E+07

9.26E+07

1.07E+08

1.8E+08

2.36E+08

All states

7.26E+08

4.25E+09

4.93E+09

8.26E+09

1.08E+10

Details on the sensitivity analyses are provided in tables V and VI. The extreme ranges of the sensitivity analyses obtained by combining the smallest CRF with the lowest valuation and the biggest CRF with the highest valuation yielded a range between 1.5 and 40 billion € saved annually, i.e. 23-602 % of the core analysis valuation.

4. Discussion

Our analysis of health effects associated with an improved insulation scenario compared with a business as usual scenario in Europe from 2005 to 2020 revealed substantial benefits and particularly so regarding the number of LY lost in Central Europe. Effects, however, were discernible in all of the 25 EU states studied except Finland and Sweden. The annual health benefits within the EU-25 included 78,678 saved LY and societal cost savings of 6.64 billion €. The study provided detailed results for health effect known to be associated with air pollution on the country level as well as sensitivity analyses assuming different impacts and costs. The analyses covered a population of almost 0.5 billion and a large region, relied on suggested ranges of impacts and costs provided by CAFE/HEIMTSA (Hurley et al., 2005; HEIMTSA, 2011) and by expert elicitations for core and sensitivity analyses in accord with suggested methods (Pervin et al., 2008). In addition, we applied life-table analyses at the national level that account for population dynamics caused by historical exposure. Changes in criteria air pollutant concentrations on a per country basis was derived from an air-quality model, and were found to be in line with results from other state-of-the-art regional air-quality models (Korsholm et al., 2012). Uncertainties in relation to the insulation levels, energy sources and consumption, the scenario, and the models used are discussed in detail therein.

Our analysis is a one-year picture, assuming 2009 meteorology and 2009 populations (or as close as possible in states without 2009 data) of a sustained improved insulation scenario policy from 2005-2020. Economic valuation is expressed in 2005 value. It is conservative, including only health effects and costs agreed upon in the CAFE/HEIMTSA reports (Hurley et al., 2005; HEIMTSA, 2011). Thus it is an investigation of the health effects and associated costs that would have occurred some years from 2005 if an improved insulation scenario had been implemented in new and existing houses rather than a study of the effects in a particular real year.

Sufficient time for the health effects to change fully after exposure reduction is inherently assumed in the study and we did not include lag-times. This assumption is reasonable considering that there is “a fair amount of evidence for a good proportion of the benefits from a reduction in PM2.5 appearing in the first few years” (Walton, 2011). We did not convert the changes in LY into numbers of avoidable deaths although this is commonly used to express mortality effects. As stated by COMEAP LY gained or lost is “the most comprehensive way of capturing the full effects” and “is the most relevant index for policy analysis”. A factor of 1/10.6 can be used to convert LY lost or gained into number premature deaths as done in the CAFE reports (Watkiss et al., 2005) although greater accuracy would require country-based disease-specific mortality rates.

We also did not include interest rates, as it would require focusing on specific spans of years and because there is no commonly agreed upon interest rate for use in environmental health impact studies across the EU-25 states. Applying interest rates can change the economic consequences of projected changes significantly. If, in our case, we assume the full effect in 2009 (from 2005) and we apply a 3% interest rate, the saved costs would be 6.06 rather than 6.64 billion €. Extending this into 2020, the final year of our scenario, would reduce the saved costs substantially as would higher interest rates applied in some EU states. To what degree such an extension would be counterbalanced by an increased population of elderly particularly susceptible to the effect of air pollution is unpredictable.

A full cost-benefit analysis was out of scope of the paper as were effects of pollutants on crops and constructions. Effects on air pollution, and thus on health, from possible energy scenarios other than the improved insulation scenario were not considered. Therefore, and in contrast to Levy et al. (2003), we did not consider production of the insulation material or costs of the improved insulation scenario. Nishioka et al. (2006) investigated insulation from current practice to the levels recommended by the 2000 International Energy Conservation Codes (IEE, 2000) in new and existing housing in the U.S., considering energy reduction in the homes, energy for production of mineral wool, economic impacts for the homeowners, and interest rates and observed that “the total disease-adjusted life years saved from the fuel supply chain is four times larger than the total disease-adjusted life years added from the mineral wool supply chain”. We have no reason to believe that this would be substantially different in the region of the EU-25. Atmospheric chemistry interactions are non-linear and without running the model for each scenario it is unpredictable how concentrations of air pollutants would change given other projected changes to emission than the improved insulation scenario. An example is that the amount of secondary ammoniated sulfate and nitrate formed is dependent on the available atmospheric ammonia that changes with farming practices (Yim et al., 2013).

Considerations on how to best calculate costs associated with loss of LY and morbidity endpoints are clearly also out of scope of this paper. Such considerations are, however, important and revisions of the valuation conducted as part of the CAFE program almost a decade ago are possibly warranted. The diversity of the EU-25 economies further complicates the issue. Our cost-evaluation approach with application of similar costs per outcome across Europe was based on EU programs such as CAFE (Holland et al., 2005) and HEIMTSA (HEIMTSA, 2011) and in line with the OECD guidance on environmental cost benefit analysis (OECD, 2006).

The CAFE analysis reported that the annual societal cost of the total amount of air pollution in the EU-25 states were approx. 513 billion € in 2000 (using VOLY mean as in the present study) (Watkiss et al., 2005). Our model suggests that 1.3% could be saved with the proposed insulation program.

The CAFE quantification of health impacts and subsequent valuation was done for the European Commission DG Environment and aimed at consistency with the WHO “Systematic Review of Health Aspects of Air Quality in Europe” (Holland et al., 2005). It has formed the basis of previous European quantifications of effects of air pollution. The HEIMTSA report from 2011 was based on the CAFE results but reviewed the CRF extensively and was used for our study (HEIMTSA, 2011). However, only the CRF for chronic bronchitis and respiratory hospital admissions changed from CAFE to HEIMTSA. Valuation of mortality and morbidity was dealt with extensively in the CAFE reports (Holland et al., 2005) and included a substantial discussion of the use of the VSL versus the VOLY approach. To our knowledge no extensive work has been published on this issue in Europe since the CAFE report suggestion of comparing median and mean costs from both VSL and VOLY. These methods and estimates are widely used, e.g. by the European Environment Agency (EEA, 2011).

Several other adverse health effects than those considered in this study have been sufficiently documented for inclusion in evaluations of air pollution related health effects and costs (Bickel and Friedrich, 2005; HEIMTSA, 2011). Additional effects on, e.g. restricted activity days, lung cancer, or asthma are commonly included in similar studies, e.g. (Wong et al., 2004; Brandt et al., 2013). In addition, effects on intrauterine growth intelligence and lung development in childhood, and associations with diseases including diabetes, appendicitis, airway infections, and rheumatoid arthritis have been reported, e.g. in (Medina-Ramon et al., 2006; Gauderman et al., 2007; Brauer et al., 2008; Hart et al., 2009; Kaplan et al., 2009; Puett et al., 2011; Andersen et al., 2012; Bellinger, 2013). In addition, some of the CRF applied in our study are probably underestimating the health effects. This particularly regards the calculations on hospitalizations as these are based on studies of short term changes in air pollution. It has consistently been shown that long-term effects of PM on mortality are several times stronger than short-term effects. It is unlikely that hospitalization rates should differ in that respect. In addition PM2.5 effects on mortality are usually stronger per µg than are effects of PM10 (reflecting the fact that PM2.5 makes up part of PM10). In the case of infant mortality the CRF was available only for PM10 although in this study the entire change in PM10 was caused by changes in PM2.5 - probably underestimating the effect.

The changes in air pollutant concentrations in this study are mean changes based on a 15-year long period assuming application of the improved insulation scenario. During such a period a likely occurrence is a substantial increase of elderly people in the region under study, resulting in a larger group of vulnerable people and thus greater potential for positive effects of decreased pollution. Other changes in lifestyle and in disease prevalences are predictable. Consideration of such demographic changes was recently demonstrated to have significant impact on long-term studies of health and social cost impacts from air pollution (Flachs et al., 2013).

Despite its widespread use in other studies, the CRF used for adult mortality (of 0.6% per μg/m3 PM2.5) could also be an underestimate. Re-analyses of the Harvard Six Cities Study as well as some analyses on the American Cancer Society Study have suggested greater effects, in particular when taking into account socio-economic determinants (Krewski et al., 2000). Accordingly, a CRF higher than 0.6% has been applied in recent studies (Levy et al., 2010; Anenberg et al., 2012; Shindell et al., 2012; Yim et al., 2013). The fact that our core estimate of 6.6 billion € is in the lower sixth of the range of the sensitivity analyses based on expert elicitations partly corroborates this view (US-EPA, 2006). In a report by the UK Committee on the Medical Effects of Air Pollutants (COMEAP, 2009) a plausibility distribution based on Members’ consolidated views was developed suggesting the use of the coefficients 1% and 12% for use in sensitivity analysis. In an expert elicitation from the United States Environmental Protection Agency (US-EPA, 2006) median estimates ranged from a 0.7 to 1.6 % decrease in annual, adult, all-cause mortality per 1 μg/m3 decrease in annual average PM2.5. We considered these ranges for sensitivity analyses more appropriate than the confidence intervals provided in a single study. Yet the core change in mortality that we calculated is not negligible and it corresponds to approximately 0.1% of the total mortality in the considered states. This figure is 10 times higher than in the study by Levy et al. (2003) on retrofit of insulation in the US. Several methodological differences between the studies may explain the difference. Most importantly Levy et al. assessed an IEE 2000 insulation scenario that may be less strict than our improved insulation scenario; the U.S. energy supply differs from the European; the study only considered retrofitting existing houses; and it passed from emission changes over intake fraction to health rather than modeling atmospheric chemistry and passing from concentrations in ambient air to health effects.

Changed quality of indoor air as a result of increased insulation, of changed concentrations of pollutants penetrating from outdoors, or from less indoor emissions in homes heated with wood stoves would be likely additional effects of the improved insulation scenario. Without changes to ventilation, houses become tighter with increased insulation which results in deteriorating indoor air quality due to increased humidity but also in greater thermal comfort and less infiltration of polluted ambient air in cities. Although indoor air quality can significantly affect health (Pekkanen et al., 2007) it was out of scope to estimate these very complex effects.

A model of a household energy efficiency program in the UK, focusing on indoor air effects from increased insulation, phasing out of indoor fossil fuel combustion, and average temperature reduction revealed that more than 115 disease-adjusted life years could be saved per million population for each of the 33 megatons CO2 saved from just the insulation improvements (Wilkinson et al., 2009). However, the study also demonstrated the importance of improved ventilation, which if not ensured when insulation improves may increase indoor radon, secondhand tobacco and mold problems. On the other hand, improved insulation can also help protecting dwellers from thermal stress in a warming climate (Haines and Dora, 2012). However, if adequate ventilation is not built into energy efficient building projects and indoor pollutants increase as a result, the negative health effects may end up dominating the positive effects.

5. Conclusions

This analysis showed that an ambitious building insulation scenario of new houses and with a 2% annual retrofit ratio of existing houses in Europe could result in 78678 saved LY with the strongest effects in Central Europe and in reductions to morbidity and societal costs that would not be trivial. Sensitivity analyses indicate that the effects may be underestimated. Health effects associated with decreased carbon emissions as well as from changed indoor air quality were not considered. Our results suggest that climate mitigation costs associated with housing insulation will be partly counterbalanced by societal savings.

Abbreviations used

CAMx, Comprehensive Air-Quality Model with extensions; CAFE, Clean Air for Europe, CO2, carbon dioxide; CRF, concentration-response function; EPA, Environmental Protection Agency; EU, European Union; IEE 2000, the 2000 International Energy Conservation Codes; HEIMTSA, Health and Environment Integrated Methodology and Toolbox for Scenario Development; ICD, International classification of disease; LRS, lower respiratory symptoms; LY, life-years; OECD, Organisation for Economic Co-operation and Development; PM, particulate matter; RAD, restricted activity days; RR, relative risk; VOLY, value of a life-year; VSL, value of a statistical life; WHO, World Health Organization.

Competing financial interests declaration

The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication. None of the authors had any other perceived or actual competing financial interests. In particular, no optional future funding from any of the sponsors has been proposed during this research which could depend on the current results.

The study was sponsored by the European Insulation Manufacturers Association (EURIMA). JHB received a fee of 1000 USD from Rockwool Denmark for participating in and presenting results at a meeting.

Références

Andersen ZJ, Raaschou-Nielsen O, Ketzel M et al. (2012). Diabetes incidence and long-term exposure to air pollution: a cohort study. Diabetes Care, n° 35, p. 92-98.

Anenberg SC, Schwartz J, Shindell D et al. (2012). Global air quality and health co-benefits of mitigating near-term climate change through methane and black carbon emission controls. Environ Health Perspect., n° 120, p. 831-839.

Bellinger DC. (2013). Prenatal Exposures to Environmental Chemicals and Children's Neurodevelopment: An Update. Saf Health Work, n° 4, p. 1-11.

Bickel P, Friedrich R. (2005). ExternE Externalities of Energy Methodology 2005 update. European Commission, Directorate-General for Research Sustainable Energy Systems Brussels; 287 p.

Bond TC, Doherty SJ, Fahey DW et al. (2013). Bounding the role of black carbon in the climate system: a scientific assessment. J Geophys Res, n° 118, p. 5380-5552.

Brandt J, Silver JD, Christensen JH et al. (2013). Contribution from the ten major emission sectors in Europe and Denmark to the health-cost externalities of air pollution using the EVA model system - an integrated modelling approach. Atmos Chem Phys Discuss, n° 13, p. 5871-5922.

Brauer M, Lencar C, Tamburic L et al. (2008). A cohort study of traffic-related air pollution impacts on birth outcomes. Environ Health Perspect, n° 116, p. 680-686.

COMEAP (2009) Long-term Exposure to Air Pollution: Effect on Mortality. 196 p. Available from: http://www.comeap.org.uk/documents/reports/39-page/linking/75-long-term-exposure-to-air-pollution-effect-on-mortality2

ECOFYS. (2007). U-Values for better energy performance of buildings. 104 p. Available from: http://www.ecofys.com/en/publication/u-values-for-better-energy-performance-of-buildings/

EEA. (2011). Revealing the costs of air pollution from industrial facilities in Europe (EEA Technical report, n° 15, p. 1-74.

Flachs EM, Sørensen J, Bønløkke JH, Brønnum-Hansen H. (2013). Population dynamics and air pollution: the impact of demographics on health impact assessment of air pollution. J Environ Public Health, 760259.

Gauderman WJ, Vora H, McConnell R et al. (2007). Effect of exposure to traffic on lung development from 10 to 18 years of age: a cohort study. Lancet, n° 369, p. 571-577.

Haines A, Dora C. (2012). How the low carbon economy can improve health. BMJ, 344: e1018–8.

Haines A, McMichael AJ, Smith KR et al. (2009). Public health benefits of strategies to reduce greenhouse-gas emissions: overview and implications for policy makers. Lancet, n° 374, p. 2104-2114.

Hart JE, Laden F, Puett RC et al. (2009). Exposure to traffic pollution and increased risk of rheumatoid arthritis. Environ Health Perspect, n° 117, p. 1065-1069.

HEIMTSA. (2011). Health and Environment Integrated Methodology and Toolbox for Scenario Development. D 3.1.2/3/4 – Final report on risk functions used in the case studies, 244 p. Report n° GOCE-CT-2006-036913-2.

Holland M, Hunt A, Hurley F et al. (2005). Methodology for the Cost-Benefit analysis for CAFE: Volume 1: Overview of Methodology, 122 p.

Hurley F, Hunt A, Cowie H et al. (2005). Methodology for the Cost-Benefit analysis for CAFE: Volume 2: Health Impact Assessment, 149 p.

Jacobson MZ. (2002). Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming. J Geophys Res, p. 16-22.

Jerrett M, Burnett RT, Pope CA et al. (2009). Long-term ozone exposure and mortality. N Engl J Med, n° 360, p. 1085-1095.

Kaplan GG, Dixon E, Panaccione R et al. (2009). Effect of ambient air pollution on the incidence of appendicitis. CMAJ, n° 181, p. 591-597.

Korsholm US, Amstrup B, Boermans T et al. (2012). Influence of building insulation on outdoor concentrations of regional air-pollutants. Atmos Environ., n° 54, p. 393-399.

Krewski D, Burnett RT, Goldberg MS et al. (2000). Reanalysis of the Harvard Six Cities Study and the American Cancer Society Study of Particulate Air Pollution and Mortality. Res Rep Health Eff Inst., p.1-97.

Levy JI, Buonocore JJ, Stackelberg von K. (2010). Evaluation of the public health impacts of traffic congestion: a health risk assessment. Environ Health, 9:65.

Levy JI, Nishioka Y, Spengler JD. (2003) The public health benefits of insulation retrofits in existing housing in the United States. Environ Health, 2:4.

Medina-Ramon M, Zanobetti A, Schwartz J. (2006). The effect of ozone and PM10 on hospital admissions for pneumonia and chronic obstructive pulmonary disease: a national multicity study. Am J Epidemiol, n° 163, p. 579-588.

Miller BG. (2011). IOMLIFET version 2011 Spreadsheets for life-table calculations, 1 10.

Miller BG, Hurley JF. (2003). Life table methods for quantitative impact assessments in chronic mortality. J Epidemiol Community Health, n° 57, p. 200-206.

Nishioka Y, Levy JI, Norris GA. (2006). Integrating risk assessment and life cycle assessment: a case study of insulation. Hum Ecol Risk Assess, n° 12, p. 552-571.

OECD. (2006). Cost-Benefit Analysis and the Environment: Recent Developments; 1-318. Available from:http://www.oecd.org/env/tools-evaluation/cost-benefitanalysisandtheenvironmentrecentdevelopments.htm.

OECD. (2007). Environmental Performance Reviews DENMARK. Nov 7, 257 p.

Pekkanen J, Hyvarinen A, Haverinen-Shaughnessy U, et al. (2007). Moisture damage and childhood asthma: a population-based incident case-control study. Eur Respir J, n° 29, p. 509-515.

Pervin T, Gerdtham U-G, Lyttkens C. (2008). Societal costs of air pollution-related health hazards: A review of methods and results. Cost Eff Resour Alloc., 6:19.

Pope CA, Burnett RT, Thun MJ et al. (2002). Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA, n° 287, p. 1132-1141.

Puett RC, Hart JE, Schwartz J, et al. (2011). Are particulate matter exposures associated with risk of type 2 diabetes? Environ Health Perspect, n° 119, p. 384-389.

Shindell D, Kuylenstierna JC, Vignati E et al. (2012). Simultaneously Mitigating Near-Term Climate Change and Improving Human Health and Food Security. Science, Jan 13, 335(6065), p. 183-9.

US-EPA. (2006). Expanded expert judgment assessment of the concentration-response relationship between PM2.5 exposure and mortality. UNITED STATES ENVIRONMENTAL PROTECTION AGENCY. Final report. Available from: http://www.epa.gov/ttn/ecas/regdata/Uncertainty/pm_ee_report.pdf

US-EPA. (2012). Regulatory Impact Analysis for the Final Revisions to the National Ambient Air Quality Standards for Particulate Matter. U.S. Environmental Protection Agency. ; 1-460. Report No.: EPA-452/R-12-003.

Walton HA. (2011). Development of proposals for cessation lag(s) for use in total impact calculations,  82 p. Available from: http://www.comeap.org.uk/images/stories/Documents/Reports/supporting%20paper%20-%20walton.pdf

Watkiss P, Pye S, Holland M, CAFE CBA: Baseline Analysis 2000 to 2020. (2005), 122 p.

Wilkinson P, Smith KR, Davies M et al. (2009). Public health benefits of strategies to reduce greenhouse-gas emissions: household energy. Lancet, n° 374, p. 1917-1929.

Wong E, Gohlke J, Griffith W et al. (2004). Assessing the health benefits of air pollution reduction for children. Environ Health Perspect, n° 112, p. 226-232.

Woodruff TJ, Grillo J, Schoendorf KC. (1997). The relationship between selected causes of postneonatal infant mortality and particulate air pollution in the United States. Environ Health Perspect, n° 105, p. 608-612.

Yim SH, Stettler ME, Barrett SR. (2013). Air quality and public health impacts of UK airports. Part II: Impacts and policy assessment. Atmos Environ., n° 67, p. 184-192.

Zmirou D, Deloraine A, Balducci F et al. (1999). Health Effects Costs of Particulate Air Pollution. J Occup Environ Med, n° 41, p. 847-856.

Pour citer ce document

Référence papier : Jakob Hjort Bønløkke, Gitte Juel Holst, Torben Sigsgaard, Ulrik Smith Korsholm, Bjarne Amstrup, Iratxe Gonzalez-Aparicio et Jens Havskov Sørensen « Modeled effects of an improved building insulation scenario in Europe on air pollution, health and societal costs », Pollution atmosphérique, N° 225, 2016, p. 1-16.

Référence électronique : Jakob Hjort Bønløkke, Gitte Juel Holst, Torben Sigsgaard, Ulrik Smith Korsholm, Bjarne Amstrup, Iratxe Gonzalez-Aparicio et Jens Havskov Sørensen « Modeled effects of an improved building insulation scenario in Europe on air pollution, health and societal costs », Pollution atmosphérique [En ligne], N° 225, mis à jour le : 23/05/2017, URL : http://lodel.irevues.inist.fr/pollutionatmospherique/index.php?id=4780, https://doi.org/10.4267/pollution-atmospherique.4780

Auteur(s)

Jakob Hjort Bønløkke

Section of Environment, Work and Health, Department of Public Health, Aarhus University, Aarhus, Denmark
Corresponding author. Aarhus University, Department of Public Health, Bartholins Allé 2, 8000 Aarhus C, Denmark. jb@mil.au.dk

Gitte Juel Holst

Section of Environment, Work and Health, Department of Public Health, Aarhus University, Aarhus, Denmark. gjho@mil.au.dk

Torben Sigsgaard

Section of Environment, Work and Health, Department of Public Health, Aarhus University, Aarhus, Denmark. ts@mil.au.dk

Ulrik Smith Korsholm

Danish Meteorological Institute, Department of Research and Development, Centre for Meteorological Models, Copenhagen, Denmark. usn@dmi.dk

Bjarne Amstrup

Danish Meteorological Institute, Department of Research and Development, Centre for Meteorological Models, Copenhagen, Denmark. bja@dmi.dk

Iratxe Gonzalez-Aparicio

Danish Meteorological Institute, Department of Research and Development, Centre for Meteorological Models, Copenhagen, Denmark. European Commission, Joint Research Centre, Institute for Environment and Sustainability, Climate Risk Unit, Ispra, Italy. iratxe.gonzalez-aparicio@jrc.ec.europa.eu

Jens Havskov Sørensen

Danish Meteorological Institute, Department of Research and Development, Centre for Meteorological Models, Copenhagen, Denmark. jhs@dmi.dk