L. De Nocker, L. Int Panis, R. Torfs 5oergm/%g 200, B-2400 ... · It is based on a detailed 'impact-pathway analysis', ... Belgium has a relatively high population density of 326

External costs of passenger transport in

Belgian cities

L. De Nocker, L. Int Panis, R. Torfs

Vito, Flemish Institute for Technological Research5oergm/%g 200, B-2400 Mo/, gg/gmmEMail: [email protected], [email protected], [email protected]

Abstract

The objective of this paper is to present intermediate results of ongoing researchprojects concerning the assessment of marginal, external costs of transport inBelgium. These studies cover all environmental impacts from different modes(road, railway and waterway) for both passenger and goods transport. Withineach transport mode, distinction is made between different options (e.g.motor/fuel technologies, traffic density, location). The presentation will focus onthe environmental impacts of airborne pollution from urban passenger transport.The assessment of these environmental impacts is based on the ExternEmethodology and it is part of a European wide project co-funded by theEuropean Commission. It is based on a detailed 'impact-pathway analysis',which quantifies impacts on human health and the environment in 4 consecutivesteps: specification of emissions, dispersion simulation, impact assessment withdose-response functions and monetary valuation. This methodology is applied toBelgian cities to demonstrate the importance of urban population density. Theresults show the importance of public health impacts as well as those ofgreenhouse gas emissions. The paper will focus on the differences betweendifferent technologies, fuels, traffic conditions and locations. Based on this datawe will discuss to which extent the benefits from stricter emissions standards arecompensated by changes in the Belgian car fleet. It shows that the benefits of theintroduction of catalysts for petrol cars are smaller than one might haveexpected. Public transport and private cars will be compared, taking into accountseat occupancy rates. Present-day diesel busses perform quite well provided theirseat occupancy rate is high enough. Older diesel busses however are likely tohave higher impacts compared to modern petrol cars.

Transactions on the Built Environment vol 41, © 1999 WIT Press, www.witpress.com, ISSN 1743-3509

502 Urban Transport and the Environment for the 21st Century

1 Introduction

1.1 Scope of the research projects

If we drive our car into the city, we have probably thought about how long thiswould take us and what it would cost us. Most people don't take into account theimpacts of their journey on public health, the historical buildings in the citycentre or on forests a 1000 kilometres downwind. These damages to man and theenvironment are called external costs, as they are not reflected in market prices.Estimates of external costs provide the basic data for analysing a myriad ofquestions related to transport and environmental policies especially in urbanenvironments.Therefore we estimate external costs for all environmental impacts from differentmodes (road, railway and waterway) for both passenger and goods transport. Ourevaluation of air pollution impacts is based on the accounting framework of theEuropean ExternE project. Earlier estimates for Belgium using ExternE datawere based on an extrapolation of case studies for neighbouring countries(Mayeres et al/) or on less up-to-date methodologies (De Nocker et al. ). Ourstudies also include estimates of externalities of other environmental impacts aswell as accidents and congestion.

1.2 Objectives of this paper

Our objective is to present a selection of intermediate results of ongoing researchon the assessment of marginal, environmental costs of transport in Belgium.Although we emphasise the impacts from the use of vehicles (mainly airpollution), other impacts are included in the study.In this paper we only report on a single item from our study: the environmentalcosts due to passenger transport on the road. We distinguish different types ofpassenger cars, fuels, cylinder capacities, technologies and traffic conditions. Wewill highlight the specific results for urban environments and compare themagainst rural conditions.In the following paragraphs, we first describe the ExternE methodology and itsimplementation for this study. Special attention is given to the calculation ofemission factors and the definition of the urban locations. In the results sectionwe first present the external costs for different passenger cars in urbanenvironments compared to rural conditions. The impact of congestion onenvironmental externalities is evaluated. We then briefly discuss howexternalities vary between different technologies. Finally, we explore theenvironmental benefits of diesel busses and electric trams for public transport incities.In addition to the results given in De Nocker et al/, data on more types ofengines and the composition of the fleets of passenger cars and city busses areincorporated in the results. Therefore we can state some conclusions more clearlyand present an estimate for the total annual air pollution cost of passenger cars inBelgium.


Urban Transport and the Environment for the 21st Century

STEPS in the analysis

Stepl

503

EMISSIONS

(e.g. g NOx/vehicle.kmfor vehicle Y / trajectory Z)

Step 2

Dispersion models

change of concentrations(e.g. fig/nf NOx in affected

regions in Europe)

Step 3

IMPACT

(e.g. Impact of NOx on human health)

Step 4

DAMAGE

EUROcent / vehicle.km)

CONCENTRATION

Exposure-responsefunction

Monetary valuation

Figure 1. The impact pathway methodology.

2 Methodology

2.1 The Impact Pathway Methodology

The evaluation of environmental impacts is based on the accounting frameworkof the European ExternE project. This framework was originally developed toaccount the externalities of electricity generation (1991-1997, EO% EC*). Since1996, it has been extended and refined to account for energy related impacts oftransport.The ExternE accounting framework is based on the 'impact pathway'methodology which represents the long way from a 'burden' to an 'impact' and



an external cost (see Figure 1. The impact pathway methodology.). Impacts onhuman health and the environment are quantified in 4 consecutive steps:determination of emission factors, dispersion simulation, impact assessment withdose-response functions, and monetary valuation. We will not discuss all thesemethodological issues in detail in this paper. Detailed information concerning themethodology can be found in EC , EC^ and De Nocker et alAThe impact pathway methodology integrates the state of the art expertise ofseveral scientific disciplines within one consistent accounting framework. Thedispersion models (step 2), dose-response functions (step 3) en prices (step 4)were selected for this project from scientific literature by a large group ofscientists who are experts in the different disciplines. The ExternE methodologyis therefore considered to be the most complete and up-to-date methodology forthe calculation of externalities of energy related impacts. Furthermore, theaccounting framework is updated and adjusted in an ongoing ExternE project(1998-1999) to make sure that every recent evolution in the relevant fields ofscience is included. The two major factors that influence the calculated damagesin our comparisons are emissions and population density. Both are discussedfurther in the following paragraph.

2.2 The implementation of the Impact Pathway Methodology

2.2.1 Calculation of emission factorsAcquiring accurate quantitative data about actual exhaust emissions for thedifferent pollutants is a complex issue due to the multitude of parametersinvolved (fuel, technology, cylinder capacity, driving pattern, etc.).For passenger cars we have identified the most representative fuels and motortechnologies. At this moment petrol and diesel cars make up more than 99% ofthe Belgian car fleet. LPG cars only take up a marginal portion (less than 1%).According to international and European emission standard legislation, we alsodistinguish between EUROO, EURO1, EURO2 and EURO3 types of vehicles.In addition we defined six drive types which relate to typical locations andassociated traffic conditions. The three distinct drive types are 'urban','highway' and 'rural drive', and for each we distinguish between peak and non-peak traffic conditions. To highlight the environmental problems that areassociated with traffic in today's cities, urban peak-traffic was defined as stop-and-go traffic whereas in the rural drive it merely reflects increased trafficdensity.A very detailed account of the methodology that was used in calculating theemission factors is given in De Nocker et al/. In this paper, we only deal withthe most important results. Because emphasis is on urban transport, the highwaydrive type is not considered here and rural results are only given as a benchmark.

2.2.2 Environmental impacts in different types of citiesEarlier results have shown that environmental externalities from transport arevery site specific. To calculate public health impacts we must accurately take thepopulation density into account, especially when applying the methodology tourban environments.


Urban Transport and the Environment for the 21st Century 505

Belgium has a relatively high population density of 326 inhabitants per knf andit has many urbanised areas, smaller towns and cities, with only a few majoragglomerations. Consequently, the 'rural' case for Belgium is still characterisedwith high population densities, compared to e.g. case studies for France. Becauseof the importance of traffic in the more populated areas in the country, the case'country average' is chosen as the reference case, as it is more representativethan a truly rural area (e.g. in the Ardennes) would be. The 'big city' case isbased on a simplification of population and geographical data for Brussels. Thecase 'small city' refers to a typical town of 80000 inhabitants and a populationdensity of 5000 inh/knr such as Aalst, Kortrijk, La Louviere or Mechelen.We have used two types of atmospheric dispersion models for respectively thelocal (up to 50 km from the source) and regional (Europe wide) scale. Whereasthe European models use large grids (50x50 km), the local dispersion is based onvery small gridcells (4x4 km). The dispersion modelling also takes into accountthe differences between rural and urban areas related to dispersion parametersand meteorological data - in particular lower ground level wind speeds in urbanareas.

3 Results

3.1 Environmental damage costs in cities versus rural areas

The large differences in the emissions of different vehicles are also reflected inimportant differences in environmental damage costs. Because it would beimpossible to present all results in this paper, we have chosen to give results foran average Belgian car. Aggregated results are presented in this way for the firsttime. This key-figure is simply a weighted average of the external costs of alldifferent types of passenger cars based on the composition of the Belgian fleet.In this way we provide an easy basis for comparison both for some types of carsthat we will highlight in the next paragraphs as well as for comparisons overtime. Some figures for the 1998 fleet are given in Table 1 (in Eurocent, 'Ect').In 1998 the external cost of an average Belgian car was 2.2 Ect/vehicle.km inrural areas. The external costs of diesel cars are higher than those of petrol cars.However, we find much higher numbers for big cities especially in congestedtraffic (Table 1). This means that the environmental cost of driving a car inBelgium is significant and of the same order of magnitude as the private fuelcosts (2.5 - 12 Ect/km). The environmental cost exceeds the private fuel costs incongested urban traffic.In these results we have included impacts for any pollutant when the necessarydata on its emission, dispersion and detrimental effect is available. Therefore theresults that are presented in the tables are in fact subtotals. Among the monetisedimpacts, the impacts on human health and global warming are the mostimportant categories. There is a growing consensus about the important healthimpacts from fine particulates, including nitrate and sulphate aerosols formedfollowing the emissions of NO% and SO2- The global warming impacts ofgreenhouse gas emissions are more controversial.



Table 1. Estimated environmental external costs of passenger carsin Belgium (in Eurocent/vkm for the 1998 fleet)

Locations -

4> vehiclesAverage carAvg. petrolAvg. diesel

rural

2.22.12.4

small city

3.52.74.9

big citynormal traffic

6.64.011.0

big citycongested

traffic

16.99.230.4

Table 2. Environmental costs for different emission control technologies(Eurocent/vehicle.km for urban traffic in big cities)

Passenger cars* technology & fuel *> 6 years old (EUROO)2-6yearsold(EURO1)< 2 years old (EURO2)

petrol6.11.71.6

diesel13.79.67.5

LPG5.41.51.2

The global warming impacts and the health impacts of nitrates, that are dominantfor petrol cars, are of a global or regional nature. The health impact ofparticipates, that is the most important category for diesel cars, is dominant in thelocal range.Compared to the impacts on public health and global warming, the damages tocrops and materials are much less important, contributing only marginally to thetotal damage.The impacts of acidification on forests and other ecosystems can be quantified interms of changes in surface area for which critical loads are exceeded. It is notyet possible to value this impact in monetary terms so it could not be included inour figures and will not be discussed here.

3.2 Evaluation of different vehicle technologies in an urban environment

We have obtained results for numerous different types of vehicles, cylindercapacities, exhaust control technologies, etc. (see Table 2 & Table 3 for somekey figures). In this paper we will only stress the striking difference betweenpetrol en diesel fuelled cars. This difference is especially important in urbanareas because of the adverse health effects that are caused by small dieselparticles. Therefore the external cost doubles when a petrol car is driven in thecity from a rural area, but quadruples for diesel cars. This effect is even morestriking for new (EURO2) cars. New LPG cars have the lowest externalities (1.2Ect/vehicle.km in big cities and 0.6 Ect/vkm in rural areas) but at present thesecars are uncommon in Belgium.



3.3 Evolution of environmental damage costs in urban areas

3.3.1 The impact of renewal of the Belgian fleet of passenger carsWhen interpreting the results from Table 1, we have to keep in mind that it is aweighted average including the external cost of cars with different cylindercapacities and technologies. Therefore the result will change from year to year,reflecting the renewal and the changing composition of the fleet. Older cars aregradually being replaced by newer cars that comply with stricter emissionstandards (Table 2) eventually resulting in the externalities shown in Table 3.Therefore the environmental external cost of Belgian cars is diminishing.When we look at the renewal and the changing composition of the Belgian fleetof passenger cars, there are two important observations that can be made. Thefirst observation is that in 1998, 50% of the Belgian car fleet consisted of 'old'EUROO vehicles. Only 29% of the fleet consisted of petrol cars equipped with athree-way-catalyst (introduced in 1993 to comply with EURO1 legislation). Atpresent the newest petrol cars (EURO2) only make up 7% of the Belgian fleet.These modern cars have external costs of about 1.6 Ect in cities (see Table 2 &3) and well below 1 Ect in rural areas. The share of diesel cars that comply atleast with the EURO1 legislation is 55%, but the external costs of EURO1 andEURO2 diesels remain relatively high because the main impact is caused byparticulates. The second observation is the clear shift from petrol to dieselfuelled cars in the composition of the Belgian fleet of passenger cars (e.g. from34% in 1996 to 37% in 1998). A main consequence of both observations madeabove is that the annual decrease of the environmental damage costs is limited.Whereas the external costs of a 1998-built petrol car are almost 4 times lowerthan those of a 1992-built car, the average external cost only diminished slightly.In 1996 the environmental cost of an average Belgian car was 7.1 Ect/vkm, in1998 it was down to 6.6 Ect/vkm. Still the average Belgian car has relativelyhigh externalities when compared to the newest models (Table 3). The effect ofimproved technology is thus only slowly beginning to show and the benefitsfrom stricter emissions standards are partly compensated by changes incomposition of the car park.

3.3.2 Do smaller cars have smaller impacts ?As a general rule, smaller cars have lower external costs than those with largerengines, the difference being about 30%. This conclusion holds true for all dieselcars that drive in rural areas, villages and small towns. However there are twoexceptions to this rule in which the external costs are only slightly lower.The first exception to be made is for small EURO1 and EURO2 petrol cars.These cars have only slightly lower external costs in rural areas because theirmain impact is caused by the long-range effect of NO* (and nitrates).The second exception is for driving in urban locations, especially when traffic iscongested. If we look at the suitability of small EURO2 cars for urban trafficfrom an environmental point of view, the relative differences with larger cars areless pronounced (Table 3). The external costs of small diesel cars (cylindercapacity < 2 1) are about 20% lower and those of small petrol cars (cylindercapacity < 1.4 1) only 10-15% lower than those caused by larger EURO2 cars.



The relative difference diminishes even further when we look at marginalexternal costs in congested city traffic, up to the point where it becomesinsignificant. This may be explained by the fact that small cars tend to producerelatively higher emissions during urban driving.However, it should be stressed that we do not have specific emission factors forsome very small new city cars. It is well possible that the environmental impactsfrom these special vehicles are lower than those of our "small petrol cars".

Table 3. Environmental costs of modern EURO2 passenger cars in Belgium(in Eurocent/vehicle.km for urban traffic in big cities)

EURO2 cars\k engine & fuel *

< 1.411.4-21>2I

petrol1.51.61.7

diesel-7.08.5

LPG-1.2-

3.4 Opportunities for public transport

For urban transport, it is interesting to compare cars with results for publictransport. In this paper the discussion is limited to large diesel busses, as theytake up the major part of the present-day fleet of public transport vehicles. Theresults for electric trams in Figure 2 are only shown for comparison. Thisestimate is based on average figures for electricity production in Belgium (DeNocker et al. ). The external costs of power generation have also been estimatedwith the ExternE methodology. The external costs per vehicle.km of dieselbusses vary from 142 Ect/vkm for EUROO buses to 58 Ect/vkm for present-daybuses and 42 Ect/vkm for future EUROS buses. These figures apply for big citiesand normal urban traffic conditions. The positive trend is due to a decrease inparticulate emissions and, to a lesser extent, lower NO% emissions. As particularsemissions are responsible for a major portion of the externalities, it is expectedthat the introduction of alternative fuel buses (CNG, LPG) would lead tosignificantly lower results. In order to compare different transport modes,external costs should be expressed in terms of passenger.kilometre instead ofvehicle.kilometre. Consequently, as shown in Figure 2, seat occupancy rates area predominant factor of public transport externalities. The straight dotted lineindicates the fixed external costs of an average passenger car with an occupancyof 1.3 persons per car. It was demonstrated before that present-day EURO2 busesperform better with respect to recent diesel cars if the occupancy rate is morethan 15%. In order to have lower externalities than recent petrol cars, occupancyrates should at least be 50%. However widely different rates (between 5% and60%) were obtained when comparing different engine technologies. Therefore nounequivocal result could be presented and the resulting data were not useful forpolicy makers. By combining data on external costs and the actual compositionof the fleets of passenger cars and city busses, we created the graph shown inFigure 2. If the occupancy rate of diesel busses is higher than 25%, theenvironmental costs of public transport are lower.



14

12

I 10o>O)

I 8

§IHI

Average Belgian bus

Average Belgian car(1.3 occupants)

o% 25% 50%

Occupancy rate

75% 100%

Figure 2.1998 comparison of public and private passenger transport in alarge Belgian city

A similar comparison in a situation with congested traffic would lead to the sameconclusions, as emissions for bus and car approximately increase with the samefactor with respect to urban non-peak driving. As occupancy rates of publictransport are usually quite high during peak hours, modern buses should performquite well in peak periods. However, at some times occupancy rates are wellbelow 25% (the average being about 15%) so that the environmental cost ofpublic transport with diesel busses in Belgian cities is higher when compared toprivate transport.At this moment we have no reason to believe that new technologies are beingintroduced faster in the fleet of city busses than in the fleet of private passengercars. Therefore we don't expect the relative position of both curves to changeover time.

3.5 Aggregated estimates of environmental damage costs for Belgium

The total annual environmental cost of private passenger transport in Belgium isestimated to be about 3 billion Euro. This figure is simply the addition of themarginal external costs for all types of passengers cars multiplied by their annualmileages under different conditions and in different locations.About half of this cost can be attributed to urban traffic although it is estimatedthat only about 30% of traffic is on urban roads. Any increase in urban traffic(especially congested traffic) will therefore cause a serious increase in the totalenvironmental external costs. Further data on the distribution of traffic ondifferent types of roads (rural, urban, highway, congested) needs to be collectedto corroborate this conclusion and allow predictions to be made.



4 Conclusions

ExternE offers a methodology to assess in a detailed way environmentalimpacts and damages from transport.Whereas impacts related to human health, global warming and the man-made environment are well accounted for, ecological impacts can only beexpressed in physical terms.Externalities of passenger cars in urban traffic are high and can exceed theprivate fuel costs.Diesel cars have much higher environmental impacts than petrol cars,especially in densely populated cities.Most types of small cars have lower externalities than large cars in ruralareas. The benefits of small cars for urban traffic are smaller than expected,especially for petrol cars and in congested traffic.Environmental externalities of diesel busses for public transport are higherthan for private transport when seat occupancy rates are under 25%.The total annual environmental cost of Belgian passenger cars is estimatedto be about 3 billion Euro.

5 Acknowledgements

This paper is based on work co-financed by the Belgian Federal Office forScientific, Technical and Cultural Affairs (OSTC) in the 'Sustainable Mobility'program and by the JOULE III program of the European Commission.

6 Bibliography

1. Mayeres, L, Ochelen, S. & Proost, S., The marginal external costs of urbantransport, Public Economics Research, 51, KULeuven-CES, 1997.

2. De Nocker, L., Vergote, S., Vinckx, L.,Wouters, G., Marginal External Costof Peak and Non-Peak Urban Transport in Belgium, Nota de lavoro 12.99,pp. 3-21, FEEM, Milan, 1999.(see website : http://www.feem.it/web/resun/_sim5.html)

3. European Commission, DGXII, Science Research and Development,JOULE, Externalities of Fuel Cycles. Report numbers EUR 16520 EN to16525 EN, 1995. (see website : http://externe.jrc.es/)

4. European Commission, DGXII, Science Research and Development,JOULE, Methodology Report, 2nd Edition. , 1998, In press.

5. European Commission, DGXII, Science, Research and Development,JOULE, External costs of Transport. Final Report., 1998, In press.

6. De Nocker, L., Torfs, R., Wouters, G., Belgian Implementation of theExternE methodology, Scientific report, VITO, Mol, 1998.


Documents

L. De Nocker, L. Int Panis, R. Torfs 5oergm/%g 200, B-2400 ... · It is based on a detailed 'impact-pathway analysis', ... Belgium has a relatively high population density of 326