- Email: [email protected]
Health costs of automobile pollution
REVUE FRANCAISE D'ALLERGOLOGIE
ET D'IMMUNOLOGIE CLINIQUE
Health costs of automobile pollution A. RABL,
J . V. S P A D A R O
SUMMARY
RESUME
The methodology developed by the ExternE (,,External Costs of Energy,,) Project of the European Commission is used to estimate the health costs associated with air pollution due to tail pipe emissions from cars. The analysis begins with emissions data for several car types in the current fleet and for three driving sites: a trip in Paris, a trip from Paris to Lyon, and travel in the rural southwest of France. Atmospheric dispersion and chemistry is modeled, both at the local and the regional scale, including the formation of secondary pollutants (ozone as well as nitrate and sulfate aerosols). Health impacts are quantified using linear dose-response functions, based on a survey of the epidemiological literature. The economic valuation is based on the willingness-to-pay to avoid a harmful impact; of particular importance is the cost of a year of life lost (YOLL), here taken as 0.083 MEuro (derived from a ,,value of statistical life,, of 3.1 MEnro). Except for post 1997 gasoline cars whose emissions are very low, the resulting damage costs per km are not much smaller than the price of fuel, and the n u m b e r of YOLL is comparable to the n u m b e r lost by traffic accidents.
CoOt de la sant6 et pollution automobile. - La mOthodologie mise au point par le projet ExternE (External Costs of Energy) de la Commission EuropOenne est utilisOe pour estimer les coots des consOquences sur la santo de la pollution atmosphOrique due aux 5missions des pots d'Ochappement des voitures. L'analyse porte d'abord sur les 6missions des diffOrents types de voitures en service actuellement, puis sur trois sites de conduite: un dOplacement dans Paris, u n d6placement de Paris gt Lyon et u n voyage dans les zones rurales du Sud-Ouest de la France. On 6tablit la dispersion atmosph6rique et la chimie, tant ~t l'6chelle locale que r6gionale y compris la formation de polluants secondaires (ozone aussi bien que a&osols de nitrates et de sulfates). Les cons&quences sur la sant6 sont quantifi6es ~ l'aide de fonctions lin6aires doses-r6ponses, d'apr& une analyse de la litt6rature 6pid6miologique. L'estimation 6conomique est bas6e sur le consentement payer pour 6viter u n effet nocif: particuli&rement important est le coot de la perte d ' u n e ann6e de vie (YOLL = year of life cost) ici estim6 gt 0,083 millions d'Euro "(d'apr& une ,,valeur de vie statistique,, de 3,1 millions d'Euro). Sauf pour les voitures ~t essence post&ieures gt 1977, dont les 6missions sont tr&s faibles, le coot par km des dommages qui en r6suhent n'est gu6re plus basque le coot de l'essence, et le nombre de YOLL est comparable au nombre de d6c6s par accident de la circulation.
KEY-WORDS: Au" pollution. - Health mlpacts. - Damage costs. Diesel cars. - Gasoline cars
MOTS-CLI~S: Pollution atmosph6rique. - Effets sur la sant6. CoOt des dommages. - Moteurs diesels - Moteurs fi essence.
INTRODUCTION This paper presents a summary of the health d a m a g e costs d u e to air p o l l u t i o n f r o m cars i n F r a n c e , as e s t i m a t e d b y t h e E x t e r n E P r o j e c t ( , ~ E x t e r n a l Costs o f E n e r g y , , ) o f t h e E u r o p e a n C o m m i s s i o n [ E x t e r n E 1995, E x t e r n E 1 9 9 8 ] . A c c o r d i n g to E x t e r n E [1998] t h e vast m a j o r i t y
Tir6s & part: A. Rabl, Centre d'l~nerg6tique, l~cole des Mines, 60, boulevard Saint-Michel, 75272 PARISCedex 06 (France).
( o v e r 95 %) o f t h e total d a m a g e cost is d u e to h e a l t h i m p a c t s , a n d a m o n g h e a l t h costs t h e d o m i n a n t i t e m is r e d u c e d life e x p e c t a n c y . C h r o n i c b r o n c h i t i s is also i m p o r t a n t , a n d so a r e i m p a c t s f o r a s t h m a t i c s . C a n c e r s h a v e also b e e n q u a n t i f i e d , b u t t h e i r c o n t r i b u t i o n to t h e total cost is v e r y small. E x t e r n e h a s n o t i d e n t i f i e d a l l e r g y as a n e n d p o i n t o f a u t o m o b i l e air p o l l u t i o n .
RABL A., SPARADOV - Health costs of automobile pollut=on Rev. fr. Allergol., 2000, 40 (1), 55-59
56
• A. RABL ETJ. V. SPA1L4_DO/
For environmental policy one needs to know which source of pollution causes how much damage. Therefore the ExternE methodology begins at the pollution source rather than at the measured ambient concentration of a pollutant. This involves an analysis of the impact pathway for each pollutant, from source to receptors (population, crops, buildings, etc.): • specification of the technologies and emissions (e.g. kg/yr of NOz from tailpipe) ; • calculation of increased pollutant concentration in all affected regions (e.g. btg/m 3 of 0 3, using models of atmospheric dispersion and chemistry for O 3 due to N02) ; • calculation of physical impacts (e.g. n u m b e r of asthma attacks due to 0 3 using dose-response functions); • economic valuation of impacts (e.g. multiplication by cost of asthma attack). The damage is summed over all affected receptors. For details of this analysis, the reader is referred to publications by the ExternE [4, 5] Project; shorter accounts can be found in Rabl, Spadaro and McGavran [10], and Rabl and Spadaro [12].
directly. The most important cost comes from mortality due to particles, calculated on the basis of Pope et al. [8]. Another important contribution comes from chronic bronchitis due to particles [1]. In addition there may be significant direct health impacts of SO 2, but for direct impacts of NO9 the evidence is less convincing. In ExternE [5] the working hypothesis has been to use the dose-response (DR) functions for particles and for 0 3 as basis. Effects of NO 2 and SO 2 are assumed to arise indirectly from the particulate nature of nitrate and sulfate aerosols, and they are calculated by applying the particle DR functions to these aerosol concentrations. With this assumption the impacts of NO 2 and SO 2 become very large, but this is uncertain because there is insufficient evidence for the health impacts of the individual components or characteristics (acidity, solubility, ...) of particulate air pollution. In particular there is a lack of epidemiological studies of nitrate aerosols because until recently this pollutant has not been monitored by air pollution monitoring stations. All DR functions for health impacts of air pollution have been assumed linear at the population level, in view of the lack of evidence for thresholds at current ambient concentrations.
KEY ASSUMPTIONS Dispersion modeling
Monetary valuation
For most air pollutants from combustion, atmospheric dispersion is significant over hundreds to thousands of kin. Both local and regional effects are important. We have therefore used a combination of local and regional dispersion models to account for all significant damages. For modeling dispersion over the short range we have used two gaussian plume models: ISC [16] and Roadpol [15]. At the regional scale we have used two different models, the Harwell Trajectory model as implemented in the EcoSense software [7] of ExternE, and the EMEP model of the Norwegian Meteorological Service [13], the official model for the analysis of transboundary pollution in Europe.
The goal of the monetary valuation of damages is to account for all costs, market and non-market. For example, the valuation of an asthma attack should include not only the cost of the medical treatment but also the willingness to pay to avoid the suffering. If the willingness to pay for a nonmarket good has been determined correctly, it is like a price, consistent with prices paid for market goods. Economists have developed several tools for determining non-market costs; of these tools c o n t i n g e n t valuation has enjoyed increasing popularity in recent years. The results are considered sufficiently reliable. It turns out that damage costs of air pollution are dominated by mortality. The key parameter is the so-called value of statistical life VSL (really the collective willingness to pay for reducing the risk of premature death). In ExternE [5], a Europeanwide value of 3.1 MEuro ($3.6 million) was chosen for VSL, close to similar studies in the USA. Unlike previous studies which simply multiplied the n u m b e r of premature deaths by VSL, ExternE [5] bases the valuation on the years of life lost (YOLL). The value of a YOLL due to air pollution is taken as 0.083 MEuro.
Health impacts of air pollution A consensus has been emerging among public health experts that air pollution, even at current ambient levels, aggravates respiratory and cardiovascular diseases and leads to premature mortality (e.g. Wilson and Spengler [17], ERPURS [3]). There is less certainty about specific causes, b u t most recent studies have identified fine particles as a prime culprit; ozone has also been implicated
Rev. fr. Allergol.,
2000,40,
1
57
/ HEALTH COSTS OF AUTOMOBILE POLLUTION • Table I. - H e a l t h damage costs per kg of pollutant, and tailpipe emissions per k i n , both for trip from Paris to Lyon. For the primary pollutants (PM2.5 and CO) the cost per kg is about 14 t i m e s higher for travel i n P a r i s and about 7 times lower for rural travel in SW of France. ,,cat.,, = catalytic converter. 1 E u r o = 6 . 5 6 F F = $ 1 . 0 5 to 1.20.
Pollutant
Health damage cost Euro/kg
Emzsszo?Is~g/kin
Old cars (before 1997) wzthoatGas°l~necat.I
w#hGas°linecat.
New cars (since 1997)
Dzesel
Gasoline
D*esel < 0 08 ~
m
PM25
160
0.03 d
0.01 a
0.15 d
0.002 ~
SO 2
10
0.03 b
0.03 b
0.03 b
0.03 b
NO 2
15.7
3 36 b
0.79 d
0.62 a
< 0.03 a
< 0.51 a
VOC
0.7
1.62 a
0.12 a
0.11 d
< 0.38 :~
< 0.19 a
CO
0.02
13.18 d
1.29 a
0.56 a
< 2 20 a
< 1.00 a
0.03 b
a E m i s s i o n s limits i m p o s e d by d i r e c t w e 9 4 / 1 2 / C E E , M a r c h 23, 1 9 9 4 f o r n e w v e h i c l e s as o f 1 / 1 / 9 7 . T h e limits a r e s p e c i f i e d f o r C O , H C + NO2, a n d f i n e p a r t i c l e s . H C a n d N O 2 e m i s s i o n s a r e c a l c u l a t e d w i t h a split f a c t o r e q u a l to t h e r a t i o o f H C . N O 2 in t h e e n n s s i o n f a c t o r s f r o m J o u m a r d et al. [6], c a l c u l a t e d f o r o u r r e p r e s e n t a t i v e d r i v i n g cycle. b C a l c u l a t e d b y m u l t i p l y i n g f u e l c o n s u m p t i o n b y 0 0 5 % , t h e a d m i s s i b l e s u l f u r c o n t e n t as o f 1 0 / 1 / 9 6 . c b a s e d o n C O N C A V ( E [2]. a b a s e d o n J o u m a r d et al. [6], as i n t e r p r e t e d b y S p a d a r o et al. [14] f o r d r i v i n g c o n d i n o n s o f P a r i s - L y o n trip.
DAMAGE COST PER KG OF POLLUTANT
The damage costs per kg of pollutant are listed in column 2 of table I. Particles emitted by cars are PM 2~ (where PM d designates particles with diameter less than d microns) and are especially harmful because they penetrate deep into the lungs. A very important issue is the variation of the damage with emission site: health impacts d e p e n d on the population distribution in the affected region. In this regard there is a difference between primary and secondary pollutants. Primary pollutants, for instance particles, cause damage in the form in which they are emitted. Some pollutants are transformed in the atmosphere to secondary pollutants and cause damage in the latter form. For example, SO 2 is transformed into sulfate aerosols and NO 2 into nitrate aerosols; NO 2 is also a precursor of ozone. SO9 can be harmful both directly as a primary pollutant and as a precursor of sulfates (especially sulfuric acid). Damage due to primary pollutants varies strongly with local conditions, especially when emitted at g r o u n d level. The damage of secondary pollutants, by contrast, is quite insensitive to the conditions in the vicinity of the source. This is because the chemical reactions take some time and the formation of nitrate and sulfate aerosol particles occurs over distances of tens to hundreds of kin. The formation of ozone is somewhat faster and occurs over several km to tens ofkm. In this paper Rev. fr Allergol , 2000, 40, 1
we make the approximation (good to about 30 % for sites in France) that the damage of secondary pollutants does not vary with emission site. It may appear surprising that the externe estimate for CO damage is so small. This may well be due to the difficulty of correctly identifying its effects in epidemiological studies.
EMISSIONS AND DAMAGE COSTS PER KM
Columns 3 to 7 of table I shows the tailpipe emissions assumed in this study. For existing cars they are based on emissions measured byJoumard et aL [6], as interpreted by Spadaro et al. [14]. For new cars they are taken as the regulatory limits imposed in France for all new vehicles as of J a n u a r y 1, 1997 (with exceptions noted); the real emissions may turn out to be lower. There are no regulations for particle emissions from gasoline cars, an item not considered significant in the past. However, in view of the high damage cost per kg of particles even small emissions can make an appreciable contribution to the total cost. They have been measured by CONCAWE [2] for two gasoline cars of the kind being sold now with catalytic converter, and we base our PM2.5 calculations on this report. The uncertainties of this item are high, not only because it is based solely on two cars but because the emissions are close to the detection limits of the instrumentation. We interpret the numbers in
58
Q A. R A B L ETJ. E SPARADO /
I
Diesel, new
2_
]
Gasoline, new
Diesel, old
i__]
NZ
Gasoline, old, with cat. Gasoline, old, no cat. 0
10
20
30 mEuro/km
40
50
60
Fig. 1. - Health cost due to emissions from tad pipe, for diesel and gasoline cars driven from Pans to Lyon. The n u m b e r s for new cars are u p p e r limits because they correspond to emissions equal to 1997 regulations. While costs of SO 2, NO 2 and VOC do not vary m u c h with emission site, the damage due to particle and CO emissions would be about 14 times larger in Paris, and about 7 times smaller m the rural SW of France. The uncertainty is large, about a factor of 4 in either direction.
the CONCAWE report to imply a PM 25 emission of 2 m g / k m for urban driving (with uncertainty range from 1 to 4 rag/kin). This point illustrates the diffÉculty of obtaining representative data about the current vehicle fleet. There is continual evolution in technologies, tastes and driving patterns, and reliable data about the current situation are hard to get. The evolution has been particularly rapid with regard to emissions. The problems of estimating the emissions come on top of the great variability of the damages with emission site, making it doubly difficult to extract typical results suitable for policy applications. The damage costs in milli Euro per km are shown in figurel. The contribution of CO, according to the assumptions of externe, is so small that we do not even show it in the figure. We emphasize that the uncertainties are very large. By considering the uncertainty- distributions for each element of the impact pathway analysis, Rabl and Spadaro [1999] have estimated that the damage costs could be a factor of 4 larger or smaller. The main uncertainties arise from the epidemiology and from the value of a Yoll.
CONCLUSION It is instructive to put the numbers in perspective. A natural comparison for the damage cost is the cost of fuel. Taking simple r o u n d numbers, a gasoline price in France of about 0.9 Euro/1
(of which some 80 % is tax) and a consumption of 81 per 100 km, one finds a fuel cost of 72 m E u r o / k m . While the variability with car type and emission site makes a precise comparison difficult, the numbers in figure 1 are somewhat but not much smaller than the fuel cost, with the exception of new (post 1997) gasoline cars whose damage cost is only 3 m E u r o / k m . Another comparison is with the years of life lost (YOLL) due to traffic accidents; this comparison is especially interesting because it avoids the uncertainties and controversies surrounding the value of a YOLL. The YOLL due to air pollution can be extracted from the damage costs by noting that approximately 85 % of the total health damage is due to mortality which has been evaluated at 83000 Euro/YOLL. The results are shown in table II. To estimate the YOLL due to traffic accidents, about 8000 per year in France, we assume a loss of 45 YOLL per accident on average. Allocating this to the average distance driven (7800km/yr per person times 56 million persons) one finds approximately 0.8 5(oll per million km. This is of the same order of magnitude as the numbers in table II, only new (post 1997) gasoline cars being much lower. It is ironic that the diesel, which had been encouraged on the grounds of higher energy efficiency and lower emissions, would now look so poor in terms of health damage, especially when driven in large cities. Indeed, except for particle emissions the diesel is relatively clean, especially compared to gasoline cars without catalyst (requiRev Jr. Allergol , 2000, 40, 1
/ HEALTH COSTS OF AUTOMOBILE POLLUTION •
59
T a b l e II. - Years O f Life L o s t (YOLL) p e r m i l l i o n k m d u e to air p o l l u t i o n e m i t t e d b y cars. F o r corn ~arison traffic a c c i d e n t s c a u s e 0.8 Y O L L p e r million k m . N e w -- rest 1997.
•
o fT y p ~
Gasohne, old without ~at.
Gasohne, old wzth cat.
Diesel, old
Gasohne, new
Dzesel, new
Paris
0,80
0.58
5.02
0.07
1.89
Paris-Lyon
0.61
0.16
0.35
0.03
0 22
Site ~ o( travel
car
red in France only since 1993). Fortunately the particle trap for diesel cars is close to being marketed and it can reduce any remaining particle emissions by about 90 %. To sum up, using the assumptions of the ExternE Project [1998], we have f o u n d that the health cost of air pollution from cars is large large enough to merit the attention it has received. However, there has been impressive progress on the part of the automotive industry, and the
latest generation of gasoline cars entail much lower damages. It may be wise to carry out costbenefit analyses before deciding how much further the emissions should be reduced.
Acknowledgments This work has been supported m part by the ExternE Project of the Joule Programme of the European Commission, DG12. We thank our colleagues in the ExternE Project for helpful discussions.
REFERENCES 1. Abbey D.E., Lebowltz M D., Mills P.K., Petersen EE, Lawrence Beeson W., Burchette R.J. - Long-term ambient concentratmns of particulates and oxidants and development of chromc disease in a cohort of nonsmoking California residents Inhalatzon Toxzcol, 1995, Z 19-34. 2. CONCAWE. - A study of the number, saze & mass of exhaust particles emitted from european diesel and gasoline vehicles under steady state and european driving cycle conditions. Report No 98/51 CONCAWE (the off companies' European organazanon for envaronment, health and safety), Madouplein 1, B-1210 Brussels, 1998. 3. ERPURS - Analyse des liens 5~court terme entre pollunon atmos.ph&ique et sant6 (Analysis of short term correlanons between mr pollution and health). Evaluanon des Risques de la Pollution Urbame snr la Sant6. Observatoire R6gional de Sant6 d'Ile-deFrance, 21-23, rue/vlmllis, F-75015 Paris, 1997. 4. ExternE. Externe : Externalities of Energy. Vol. 1: Summary: Vol. 2: Methodology; Vol. 3' Coal and Lignite; Vol. 5. Nuclear 1SBN 92-827-5210-0, 1995. Published by the European Commission, Directorate-General XII, Science Research and Development. L-2920 Luxembourg, 1995. 5. ExternE. - Externe. Externahties of Energy. New results, to be pubhshed by Externe Program of European Commissmn Darectorate-General XII, Science Research and Development, 1998. 6. Joumard R., R. Vidon, L. Paturel, C. Pruvost, P Tassel, G. De Soete, A I. Sabra: - l~volutmn des l~missions de Polluants des Voitures Particuh~res lors du D6part Moteur Froid (Evolutmn of pollutant emissaons from cars, beginning with cold start). Report INRETS no 197. Bron, France. 1995. 7. Krewatt W., Trukenmueller A, Mayerhofer E, Fnedrich R. Ecosense - an In tegraled Tool for Environmental Impact Analysis. In: Kremers, H , Pillmann, W. (Ed.): Space and Time an Envaronmental InformaUon Systems. Umwelt-Informatik aktuell, Band 7. Metropolis-Verlag, Marburg, Germany, 1995
8, Pope C.A., Thun M.J., Namboodri M.M,, Docker)' D.W., Evans J.S., Speizer EE., Heath C W, - Particulate air pollution as a predictor of mortahty in a prospective study of US adults Amer. J Resp. C~t*cal Care Med, 1995, 151, 669-674. 9. Rabl A., Eyre N. - An Estimate of regional and global O~ Damage from Precursor N O 2 and VOC Emissions. Environment Internatmnal, 1998, 24, 835-850. 10. Rabl A, Spadaro J.'C, McGavran ED. - Health Risks of Air Pollution from Incinerators: a Perspective. Waste Management and Research, 1998, 16, 365-388 11. Rabl A., Spadaro JN, - Environmental Damages and Costs: an Analysis of Uncertainnes. Enwronment Internatzonal, 1999, 25, 2946. 12. Rabl A., SpadaroJ.V. - Les coots environnementaux de l'6nergie. Rapport final d&embre 1997 pour I'ADEME et la Commission Europ~enne. 13. Simpson D. ,,Photochemical model calculations over Europe for two extended summei periods: 1985 and 1989. Model results and comparison with observaUons,,. Atmosph~r~ Enwronment, 1993. 27A, 921-943. 14. SpadaroJ.V., Rabl A, EJourdam E., Coussy E - External Costs of Air Pollutaon: Case study and results for transport between Paris and Lyon. InternatwnaIJ VehzcleDeszuz, 1998. 20, 274-282. 15 Vossmiotis G., Arabatzis G., Asslmacopoulos D. - Description of Roadpol: A Gaussian Dispersion Model for Line Sources, program manual, Natmnal Technical Umversaty of Athens, Greece, 1996. 16. Wackter D.J., Foster J.A. - Industrml source complex ([SC) dzsper~on model user's g'u~de. 2nd edztzon. Vol. 1, EPA 450/4-88-002a. US Enwronmental Protectaon Agency, Research Triangle Park, North Carolina, 1987. 17. Wilson R., Spengler J.D., editors 1996 Partzcles zn Our A~r, Concentratzons and Health Effects. Harvard Umversity Press, Cambridge, MA.
mum Rev. f r Allergol., 2000, 40, 1
ET D'IMMUNOLOGIE CLINIQUE
Health costs of automobile pollution A. RABL,
J . V. S P A D A R O
SUMMARY
RESUME
The methodology developed by the ExternE (,,External Costs of Energy,,) Project of the European Commission is used to estimate the health costs associated with air pollution due to tail pipe emissions from cars. The analysis begins with emissions data for several car types in the current fleet and for three driving sites: a trip in Paris, a trip from Paris to Lyon, and travel in the rural southwest of France. Atmospheric dispersion and chemistry is modeled, both at the local and the regional scale, including the formation of secondary pollutants (ozone as well as nitrate and sulfate aerosols). Health impacts are quantified using linear dose-response functions, based on a survey of the epidemiological literature. The economic valuation is based on the willingness-to-pay to avoid a harmful impact; of particular importance is the cost of a year of life lost (YOLL), here taken as 0.083 MEuro (derived from a ,,value of statistical life,, of 3.1 MEnro). Except for post 1997 gasoline cars whose emissions are very low, the resulting damage costs per km are not much smaller than the price of fuel, and the n u m b e r of YOLL is comparable to the n u m b e r lost by traffic accidents.
CoOt de la sant6 et pollution automobile. - La mOthodologie mise au point par le projet ExternE (External Costs of Energy) de la Commission EuropOenne est utilisOe pour estimer les coots des consOquences sur la santo de la pollution atmosphOrique due aux 5missions des pots d'Ochappement des voitures. L'analyse porte d'abord sur les 6missions des diffOrents types de voitures en service actuellement, puis sur trois sites de conduite: un dOplacement dans Paris, u n d6placement de Paris gt Lyon et u n voyage dans les zones rurales du Sud-Ouest de la France. On 6tablit la dispersion atmosph6rique et la chimie, tant ~t l'6chelle locale que r6gionale y compris la formation de polluants secondaires (ozone aussi bien que a&osols de nitrates et de sulfates). Les cons&quences sur la sant6 sont quantifi6es ~ l'aide de fonctions lin6aires doses-r6ponses, d'apr& une analyse de la litt6rature 6pid6miologique. L'estimation 6conomique est bas6e sur le consentement payer pour 6viter u n effet nocif: particuli&rement important est le coot de la perte d ' u n e ann6e de vie (YOLL = year of life cost) ici estim6 gt 0,083 millions d'Euro "(d'apr& une ,,valeur de vie statistique,, de 3,1 millions d'Euro). Sauf pour les voitures ~t essence post&ieures gt 1977, dont les 6missions sont tr&s faibles, le coot par km des dommages qui en r6suhent n'est gu6re plus basque le coot de l'essence, et le nombre de YOLL est comparable au nombre de d6c6s par accident de la circulation.
KEY-WORDS: Au" pollution. - Health mlpacts. - Damage costs. Diesel cars. - Gasoline cars
MOTS-CLI~S: Pollution atmosph6rique. - Effets sur la sant6. CoOt des dommages. - Moteurs diesels - Moteurs fi essence.
INTRODUCTION This paper presents a summary of the health d a m a g e costs d u e to air p o l l u t i o n f r o m cars i n F r a n c e , as e s t i m a t e d b y t h e E x t e r n E P r o j e c t ( , ~ E x t e r n a l Costs o f E n e r g y , , ) o f t h e E u r o p e a n C o m m i s s i o n [ E x t e r n E 1995, E x t e r n E 1 9 9 8 ] . A c c o r d i n g to E x t e r n E [1998] t h e vast m a j o r i t y
Tir6s & part: A. Rabl, Centre d'l~nerg6tique, l~cole des Mines, 60, boulevard Saint-Michel, 75272 PARISCedex 06 (France).
( o v e r 95 %) o f t h e total d a m a g e cost is d u e to h e a l t h i m p a c t s , a n d a m o n g h e a l t h costs t h e d o m i n a n t i t e m is r e d u c e d life e x p e c t a n c y . C h r o n i c b r o n c h i t i s is also i m p o r t a n t , a n d so a r e i m p a c t s f o r a s t h m a t i c s . C a n c e r s h a v e also b e e n q u a n t i f i e d , b u t t h e i r c o n t r i b u t i o n to t h e total cost is v e r y small. E x t e r n e h a s n o t i d e n t i f i e d a l l e r g y as a n e n d p o i n t o f a u t o m o b i l e air p o l l u t i o n .
RABL A., SPARADOV - Health costs of automobile pollut=on Rev. fr. Allergol., 2000, 40 (1), 55-59
56
• A. RABL ETJ. V. SPA1L4_DO/
For environmental policy one needs to know which source of pollution causes how much damage. Therefore the ExternE methodology begins at the pollution source rather than at the measured ambient concentration of a pollutant. This involves an analysis of the impact pathway for each pollutant, from source to receptors (population, crops, buildings, etc.): • specification of the technologies and emissions (e.g. kg/yr of NOz from tailpipe) ; • calculation of increased pollutant concentration in all affected regions (e.g. btg/m 3 of 0 3, using models of atmospheric dispersion and chemistry for O 3 due to N02) ; • calculation of physical impacts (e.g. n u m b e r of asthma attacks due to 0 3 using dose-response functions); • economic valuation of impacts (e.g. multiplication by cost of asthma attack). The damage is summed over all affected receptors. For details of this analysis, the reader is referred to publications by the ExternE [4, 5] Project; shorter accounts can be found in Rabl, Spadaro and McGavran [10], and Rabl and Spadaro [12].
directly. The most important cost comes from mortality due to particles, calculated on the basis of Pope et al. [8]. Another important contribution comes from chronic bronchitis due to particles [1]. In addition there may be significant direct health impacts of SO 2, but for direct impacts of NO9 the evidence is less convincing. In ExternE [5] the working hypothesis has been to use the dose-response (DR) functions for particles and for 0 3 as basis. Effects of NO 2 and SO 2 are assumed to arise indirectly from the particulate nature of nitrate and sulfate aerosols, and they are calculated by applying the particle DR functions to these aerosol concentrations. With this assumption the impacts of NO 2 and SO 2 become very large, but this is uncertain because there is insufficient evidence for the health impacts of the individual components or characteristics (acidity, solubility, ...) of particulate air pollution. In particular there is a lack of epidemiological studies of nitrate aerosols because until recently this pollutant has not been monitored by air pollution monitoring stations. All DR functions for health impacts of air pollution have been assumed linear at the population level, in view of the lack of evidence for thresholds at current ambient concentrations.
KEY ASSUMPTIONS Dispersion modeling
Monetary valuation
For most air pollutants from combustion, atmospheric dispersion is significant over hundreds to thousands of kin. Both local and regional effects are important. We have therefore used a combination of local and regional dispersion models to account for all significant damages. For modeling dispersion over the short range we have used two gaussian plume models: ISC [16] and Roadpol [15]. At the regional scale we have used two different models, the Harwell Trajectory model as implemented in the EcoSense software [7] of ExternE, and the EMEP model of the Norwegian Meteorological Service [13], the official model for the analysis of transboundary pollution in Europe.
The goal of the monetary valuation of damages is to account for all costs, market and non-market. For example, the valuation of an asthma attack should include not only the cost of the medical treatment but also the willingness to pay to avoid the suffering. If the willingness to pay for a nonmarket good has been determined correctly, it is like a price, consistent with prices paid for market goods. Economists have developed several tools for determining non-market costs; of these tools c o n t i n g e n t valuation has enjoyed increasing popularity in recent years. The results are considered sufficiently reliable. It turns out that damage costs of air pollution are dominated by mortality. The key parameter is the so-called value of statistical life VSL (really the collective willingness to pay for reducing the risk of premature death). In ExternE [5], a Europeanwide value of 3.1 MEuro ($3.6 million) was chosen for VSL, close to similar studies in the USA. Unlike previous studies which simply multiplied the n u m b e r of premature deaths by VSL, ExternE [5] bases the valuation on the years of life lost (YOLL). The value of a YOLL due to air pollution is taken as 0.083 MEuro.
Health impacts of air pollution A consensus has been emerging among public health experts that air pollution, even at current ambient levels, aggravates respiratory and cardiovascular diseases and leads to premature mortality (e.g. Wilson and Spengler [17], ERPURS [3]). There is less certainty about specific causes, b u t most recent studies have identified fine particles as a prime culprit; ozone has also been implicated
Rev. fr. Allergol.,
2000,40,
1
57
/ HEALTH COSTS OF AUTOMOBILE POLLUTION • Table I. - H e a l t h damage costs per kg of pollutant, and tailpipe emissions per k i n , both for trip from Paris to Lyon. For the primary pollutants (PM2.5 and CO) the cost per kg is about 14 t i m e s higher for travel i n P a r i s and about 7 times lower for rural travel in SW of France. ,,cat.,, = catalytic converter. 1 E u r o = 6 . 5 6 F F = $ 1 . 0 5 to 1.20.
Pollutant
Health damage cost Euro/kg
Emzsszo?Is~g/kin
Old cars (before 1997) wzthoatGas°l~necat.I
w#hGas°linecat.
New cars (since 1997)
Dzesel
Gasoline
D*esel < 0 08 ~
m
PM25
160
0.03 d
0.01 a
0.15 d
0.002 ~
SO 2
10
0.03 b
0.03 b
0.03 b
0.03 b
NO 2
15.7
3 36 b
0.79 d
0.62 a
< 0.03 a
< 0.51 a
VOC
0.7
1.62 a
0.12 a
0.11 d
< 0.38 :~
< 0.19 a
CO
0.02
13.18 d
1.29 a
0.56 a
< 2 20 a
< 1.00 a
0.03 b
a E m i s s i o n s limits i m p o s e d by d i r e c t w e 9 4 / 1 2 / C E E , M a r c h 23, 1 9 9 4 f o r n e w v e h i c l e s as o f 1 / 1 / 9 7 . T h e limits a r e s p e c i f i e d f o r C O , H C + NO2, a n d f i n e p a r t i c l e s . H C a n d N O 2 e m i s s i o n s a r e c a l c u l a t e d w i t h a split f a c t o r e q u a l to t h e r a t i o o f H C . N O 2 in t h e e n n s s i o n f a c t o r s f r o m J o u m a r d et al. [6], c a l c u l a t e d f o r o u r r e p r e s e n t a t i v e d r i v i n g cycle. b C a l c u l a t e d b y m u l t i p l y i n g f u e l c o n s u m p t i o n b y 0 0 5 % , t h e a d m i s s i b l e s u l f u r c o n t e n t as o f 1 0 / 1 / 9 6 . c b a s e d o n C O N C A V ( E [2]. a b a s e d o n J o u m a r d et al. [6], as i n t e r p r e t e d b y S p a d a r o et al. [14] f o r d r i v i n g c o n d i n o n s o f P a r i s - L y o n trip.
DAMAGE COST PER KG OF POLLUTANT
The damage costs per kg of pollutant are listed in column 2 of table I. Particles emitted by cars are PM 2~ (where PM d designates particles with diameter less than d microns) and are especially harmful because they penetrate deep into the lungs. A very important issue is the variation of the damage with emission site: health impacts d e p e n d on the population distribution in the affected region. In this regard there is a difference between primary and secondary pollutants. Primary pollutants, for instance particles, cause damage in the form in which they are emitted. Some pollutants are transformed in the atmosphere to secondary pollutants and cause damage in the latter form. For example, SO 2 is transformed into sulfate aerosols and NO 2 into nitrate aerosols; NO 2 is also a precursor of ozone. SO9 can be harmful both directly as a primary pollutant and as a precursor of sulfates (especially sulfuric acid). Damage due to primary pollutants varies strongly with local conditions, especially when emitted at g r o u n d level. The damage of secondary pollutants, by contrast, is quite insensitive to the conditions in the vicinity of the source. This is because the chemical reactions take some time and the formation of nitrate and sulfate aerosol particles occurs over distances of tens to hundreds of kin. The formation of ozone is somewhat faster and occurs over several km to tens ofkm. In this paper Rev. fr Allergol , 2000, 40, 1
we make the approximation (good to about 30 % for sites in France) that the damage of secondary pollutants does not vary with emission site. It may appear surprising that the externe estimate for CO damage is so small. This may well be due to the difficulty of correctly identifying its effects in epidemiological studies.
EMISSIONS AND DAMAGE COSTS PER KM
Columns 3 to 7 of table I shows the tailpipe emissions assumed in this study. For existing cars they are based on emissions measured byJoumard et aL [6], as interpreted by Spadaro et al. [14]. For new cars they are taken as the regulatory limits imposed in France for all new vehicles as of J a n u a r y 1, 1997 (with exceptions noted); the real emissions may turn out to be lower. There are no regulations for particle emissions from gasoline cars, an item not considered significant in the past. However, in view of the high damage cost per kg of particles even small emissions can make an appreciable contribution to the total cost. They have been measured by CONCAWE [2] for two gasoline cars of the kind being sold now with catalytic converter, and we base our PM2.5 calculations on this report. The uncertainties of this item are high, not only because it is based solely on two cars but because the emissions are close to the detection limits of the instrumentation. We interpret the numbers in
58
Q A. R A B L ETJ. E SPARADO /
I
Diesel, new
2_
]
Gasoline, new
Diesel, old
i__]
NZ
Gasoline, old, with cat. Gasoline, old, no cat. 0
10
20
30 mEuro/km
40
50
60
Fig. 1. - Health cost due to emissions from tad pipe, for diesel and gasoline cars driven from Pans to Lyon. The n u m b e r s for new cars are u p p e r limits because they correspond to emissions equal to 1997 regulations. While costs of SO 2, NO 2 and VOC do not vary m u c h with emission site, the damage due to particle and CO emissions would be about 14 times larger in Paris, and about 7 times smaller m the rural SW of France. The uncertainty is large, about a factor of 4 in either direction.
the CONCAWE report to imply a PM 25 emission of 2 m g / k m for urban driving (with uncertainty range from 1 to 4 rag/kin). This point illustrates the diffÉculty of obtaining representative data about the current vehicle fleet. There is continual evolution in technologies, tastes and driving patterns, and reliable data about the current situation are hard to get. The evolution has been particularly rapid with regard to emissions. The problems of estimating the emissions come on top of the great variability of the damages with emission site, making it doubly difficult to extract typical results suitable for policy applications. The damage costs in milli Euro per km are shown in figurel. The contribution of CO, according to the assumptions of externe, is so small that we do not even show it in the figure. We emphasize that the uncertainties are very large. By considering the uncertainty- distributions for each element of the impact pathway analysis, Rabl and Spadaro [1999] have estimated that the damage costs could be a factor of 4 larger or smaller. The main uncertainties arise from the epidemiology and from the value of a Yoll.
CONCLUSION It is instructive to put the numbers in perspective. A natural comparison for the damage cost is the cost of fuel. Taking simple r o u n d numbers, a gasoline price in France of about 0.9 Euro/1
(of which some 80 % is tax) and a consumption of 81 per 100 km, one finds a fuel cost of 72 m E u r o / k m . While the variability with car type and emission site makes a precise comparison difficult, the numbers in figure 1 are somewhat but not much smaller than the fuel cost, with the exception of new (post 1997) gasoline cars whose damage cost is only 3 m E u r o / k m . Another comparison is with the years of life lost (YOLL) due to traffic accidents; this comparison is especially interesting because it avoids the uncertainties and controversies surrounding the value of a YOLL. The YOLL due to air pollution can be extracted from the damage costs by noting that approximately 85 % of the total health damage is due to mortality which has been evaluated at 83000 Euro/YOLL. The results are shown in table II. To estimate the YOLL due to traffic accidents, about 8000 per year in France, we assume a loss of 45 YOLL per accident on average. Allocating this to the average distance driven (7800km/yr per person times 56 million persons) one finds approximately 0.8 5(oll per million km. This is of the same order of magnitude as the numbers in table II, only new (post 1997) gasoline cars being much lower. It is ironic that the diesel, which had been encouraged on the grounds of higher energy efficiency and lower emissions, would now look so poor in terms of health damage, especially when driven in large cities. Indeed, except for particle emissions the diesel is relatively clean, especially compared to gasoline cars without catalyst (requiRev Jr. Allergol , 2000, 40, 1
/ HEALTH COSTS OF AUTOMOBILE POLLUTION •
59
T a b l e II. - Years O f Life L o s t (YOLL) p e r m i l l i o n k m d u e to air p o l l u t i o n e m i t t e d b y cars. F o r corn ~arison traffic a c c i d e n t s c a u s e 0.8 Y O L L p e r million k m . N e w -- rest 1997.
•
o fT y p ~
Gasohne, old without ~at.
Gasohne, old wzth cat.
Diesel, old
Gasohne, new
Dzesel, new
Paris
0,80
0.58
5.02
0.07
1.89
Paris-Lyon
0.61
0.16
0.35
0.03
0 22
Site ~ o( travel
car
red in France only since 1993). Fortunately the particle trap for diesel cars is close to being marketed and it can reduce any remaining particle emissions by about 90 %. To sum up, using the assumptions of the ExternE Project [1998], we have f o u n d that the health cost of air pollution from cars is large large enough to merit the attention it has received. However, there has been impressive progress on the part of the automotive industry, and the
latest generation of gasoline cars entail much lower damages. It may be wise to carry out costbenefit analyses before deciding how much further the emissions should be reduced.
Acknowledgments This work has been supported m part by the ExternE Project of the Joule Programme of the European Commission, DG12. We thank our colleagues in the ExternE Project for helpful discussions.
REFERENCES 1. Abbey D.E., Lebowltz M D., Mills P.K., Petersen EE, Lawrence Beeson W., Burchette R.J. - Long-term ambient concentratmns of particulates and oxidants and development of chromc disease in a cohort of nonsmoking California residents Inhalatzon Toxzcol, 1995, Z 19-34. 2. CONCAWE. - A study of the number, saze & mass of exhaust particles emitted from european diesel and gasoline vehicles under steady state and european driving cycle conditions. Report No 98/51 CONCAWE (the off companies' European organazanon for envaronment, health and safety), Madouplein 1, B-1210 Brussels, 1998. 3. ERPURS - Analyse des liens 5~court terme entre pollunon atmos.ph&ique et sant6 (Analysis of short term correlanons between mr pollution and health). Evaluanon des Risques de la Pollution Urbame snr la Sant6. Observatoire R6gional de Sant6 d'Ile-deFrance, 21-23, rue/vlmllis, F-75015 Paris, 1997. 4. ExternE. Externe : Externalities of Energy. Vol. 1: Summary: Vol. 2: Methodology; Vol. 3' Coal and Lignite; Vol. 5. Nuclear 1SBN 92-827-5210-0, 1995. Published by the European Commission, Directorate-General XII, Science Research and Development. L-2920 Luxembourg, 1995. 5. ExternE. - Externe. Externahties of Energy. New results, to be pubhshed by Externe Program of European Commissmn Darectorate-General XII, Science Research and Development, 1998. 6. Joumard R., R. Vidon, L. Paturel, C. Pruvost, P Tassel, G. De Soete, A I. Sabra: - l~volutmn des l~missions de Polluants des Voitures Particuh~res lors du D6part Moteur Froid (Evolutmn of pollutant emissaons from cars, beginning with cold start). Report INRETS no 197. Bron, France. 1995. 7. Krewatt W., Trukenmueller A, Mayerhofer E, Fnedrich R. Ecosense - an In tegraled Tool for Environmental Impact Analysis. In: Kremers, H , Pillmann, W. (Ed.): Space and Time an Envaronmental InformaUon Systems. Umwelt-Informatik aktuell, Band 7. Metropolis-Verlag, Marburg, Germany, 1995
8, Pope C.A., Thun M.J., Namboodri M.M,, Docker)' D.W., Evans J.S., Speizer EE., Heath C W, - Particulate air pollution as a predictor of mortahty in a prospective study of US adults Amer. J Resp. C~t*cal Care Med, 1995, 151, 669-674. 9. Rabl A., Eyre N. - An Estimate of regional and global O~ Damage from Precursor N O 2 and VOC Emissions. Environment Internatmnal, 1998, 24, 835-850. 10. Rabl A, Spadaro J.'C, McGavran ED. - Health Risks of Air Pollution from Incinerators: a Perspective. Waste Management and Research, 1998, 16, 365-388 11. Rabl A., Spadaro JN, - Environmental Damages and Costs: an Analysis of Uncertainnes. Enwronment Internatzonal, 1999, 25, 2946. 12. Rabl A., SpadaroJ.V. - Les coots environnementaux de l'6nergie. Rapport final d&embre 1997 pour I'ADEME et la Commission Europ~enne. 13. Simpson D. ,,Photochemical model calculations over Europe for two extended summei periods: 1985 and 1989. Model results and comparison with observaUons,,. Atmosph~r~ Enwronment, 1993. 27A, 921-943. 14. SpadaroJ.V., Rabl A, EJourdam E., Coussy E - External Costs of Air Pollutaon: Case study and results for transport between Paris and Lyon. InternatwnaIJ VehzcleDeszuz, 1998. 20, 274-282. 15 Vossmiotis G., Arabatzis G., Asslmacopoulos D. - Description of Roadpol: A Gaussian Dispersion Model for Line Sources, program manual, Natmnal Technical Umversaty of Athens, Greece, 1996. 16. Wackter D.J., Foster J.A. - Industrml source complex ([SC) dzsper~on model user's g'u~de. 2nd edztzon. Vol. 1, EPA 450/4-88-002a. US Enwronmental Protectaon Agency, Research Triangle Park, North Carolina, 1987. 17. Wilson R., Spengler J.D., editors 1996 Partzcles zn Our A~r, Concentratzons and Health Effects. Harvard Umversity Press, Cambridge, MA.
mum Rev. f r Allergol., 2000, 40, 1