dynamics of the introduction of new passenger car technologies

DYNAMICS OF THE INTRODUCTION OF NEW PASSENGER CAR TECHNOLOGIES

The IPTS Transport technologies model

Panayotis Christidis Ignacio Hidalgo Antonio Soria

June 2003

Report EUR 20762 EN

European Commission Joint Research Centre (DG JRC) Institute for Prospective Technological Studies http://www.jrc.es Legal notice Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use which might be made of the following information. Report EUR 20762 EN © European Communities, 2003 Reproduction is authorised provided the source is acknowledged. Printed in Spain

The IPTS transport technologies model

i

TABLE OF CONTENTS

TABLE OF CONTENTS..........................................................................................................................i

LIST OF FIGURES..................................................................................................................................ii

LIST OF TABLES...................................................................................................................................iii

EXECUTIVE SUMMARY.....................................................................................................................iv

1 INTRODUCTION .........................................................................................................................11

2 MODEL STRUCTURE.................................................................................................................13

3 CAR OWNERSHIP AND NEW REGISTRATIONS ................................................................15 3.1 MODELLING OF CAR FLEET AND NEW REGISTRATIONS .............................................................19 3.2 DYNAMICS OF CAR SCRAPPING AND FLEET AGEING..................................................................20 3.3 USED CARS ..............................................................................................................................28

4 MARKET SEGMENTATION .....................................................................................................30 4.1 THE WOOD ALLOCATION ALGORITHM .....................................................................................31 4.2 GRAPHICAL EXAMPLE OF WOOD ALGORITHM..........................................................................35 4.3 LONG-TERM USER COSTS .........................................................................................................36 4.4 DIESELIZATION OF THE FLEET AND THE CALIBRATION OF CONSUMER’S CHOICE. .....................38

5 TRANSPORT DEMAND..............................................................................................................43

6 FUEL CONSUMPTION AND EMISSIONS ..............................................................................47 6.1 FUEL CONSUMPTION OF DIESEL AND PETROL CARS ..................................................................47 6.2 TRENDS IN FUEL ECONOMY ......................................................................................................48

7 ADDITIONAL INDICATORS.....................................................................................................51

8 MAIN RESULTS...........................................................................................................................52

9 SCENARIO-BUILDING CAPABILITIES .................................................................................65

10 FUTURE EXTENSIONS..........................................................................................................67

11 REFERENCES..........................................................................................................................68

ANNEX I: MODEL DATABASE .........................................................................................................70


ii

LIST OF FIGURES

Figure 3-1 : Households with regular use of car, UK................................................................ 16 Figure 3-2: Driving licence holders statistics, UK .................................................................... 16 Figure 3-3: Car ownership function based on the Gompertz model.......................................... 19 Figure 3-4: New car registrations versus scrapped cars, EU-15................................................ 23 Figure 3-5: Average age of passenger cars, EU ........................................................................ 25 Figure 3-6: Average age of passenger cars and heavy duty vehicles, accession countries (AC)

........................................................................................................................................... 25 Figure 3-7: Automobile Survival Rate (US) [Transportation Energy Data Book, 21].............. 26 Figure 3-8: Automobile Scrappage Rate [Transportation Energy Data Book, 21] ................... 26 Figure 3-9: Automobile Medium Lifetime [Transportation Energy Data Book, 21] ................ 27 Figure 3-10: Life Span of Cars by country [APME, 1999] ...................................................... 27 Figure 4-1: Assumed development of car segment split according to owner categories........... 30 Figure 4-2: Light duty diesel vehicle vs. gasoline vehicle sales in the EU .............................. 38 Figure 4-3: Passenger car fleet distribution (1995 data) for EU 15. [Hickman, 1999] ............. 39 Figure 4-4: Evolution of the emission standards for diesel and gasoline engines in Europe,

[Homeister, 2001].............................................................................................................. 41 Figure 4-5: Comparison of the American, European and Japanese emission standards

[Homeister, 2001].............................................................................................................. 41 Figure 5-1: Annual mileage as a function of the passenger car age (1990 data) [Hickman, 1999]

........................................................................................................................................... 43 Figure 5-2: In-use fleet vehicle miles travelled by year in US vs EU [Martec, 2002] .............. 43 Figure 5-3: Fuel consumption per vehicle by year in US vs EU [Martec, 2002] ..................... 44 Figure 5-4: Relation between engine type/size and the annual mileage (*1000) of passenger

cars in EU 15 (1995 data) [Hickman, 1999]...................................................................... 45 Figure 6-1: US vs EU average fuel economy (per model year) [Martec, 2002]........................ 48 Figure 6-2: ACEA agreement [ACEA, 2002] ........................................................................... 49 Figure 6-3: Average engine displacement of new sold vehicles [Martec, 2002]....................... 50


iii

LIST OF TABLES

Table 3-1: Income, variable cost and fixed cost elasticities of car ownership .......................... 15 Table 3-2: New cars in EU-15 by segments and bodies............................................................ 17 Table 3-3: Average engine power of new cars .......................................................................... 18 Table 3-4: New passenger car registrations in Western Europe, Jan-Oct 2002 ........................ 18 Table 3-5: Passenger car sales by region................................................................................... 18 Table 3-6: Vehicle scrapping and recycling in the UK ............................................................. 23 Table 3-7: Estimates of number of vehicles scrapped ............................................................... 24 Table 3-8: Used car sales in various EU countries .................................................................... 28 Table 3-9: Passenger car registrations in France (‘000s)........................................................... 28 Table 3-10: First time passenger car registrations in Ireland .................................................... 28 Table 3-11: Car sales in USA.................................................................................................... 29 Table 3-12: Breakdown of the volume of used car sales by age of car, UK, 1998 ................... 29 Table 4-1: Sales of diesel cars in western Europe 1990-2001................................................... 39 Table 4-2: Fuel consumption of diesel and gasoline models of same engine size .................... 40 Table 4-3: Price of cars with diesel engines and gasoline engines............................................ 41 Table 5-1: Annual mileage, mileage distribution and representative speeds for passenger cars in

the EU15 (Reference year 1995) [Hickman, 1999] .......................................................... 46 Table 6-1: Specific fuel consumption and CO2 emissions of the ACEA newly registered

passenger cars in the EU and each EU member state [ACEA, 2002]................................ 49 Table 8-1: Car ownership levels (passenger cars per 1000 inhabitants) ................................... 54 Table 8-2: Total number of new passenger car registrations (thousands) ................................. 54 Table 8-3: Total number of passenger cars removed from the stock (thousands) ..................... 55 Table 8-4: Total number of passenger cars in circulation (millions)......................................... 55 Table 8-5: New electric car sales (thousands) ........................................................................... 56 Table 8-6: New fuel cell car sales (thousands).......................................................................... 56 Table 8-7: New large diesel car sales (thousands)..................................................................... 57 Table 8-8: New large gasoline car sales (thousands)................................................................. 57 Table 8-9: New hybrid car sales (thousands) ............................................................................ 58 Table 8-10: New light diesel car sales (thousands) ................................................................... 58 Table 8-11: New light gasoline car sales (thousands) ............................................................... 59 Table 8-12: Share of each technology in new car shares in EU-15 (%).................................... 59 Table 8-13: Car use (average km per car per year) ................................................................... 59 Table 8-14: Total car use (total car* km per year, millions) ..................................................... 60 Table 8-15: Transport intensity (total car*km per euro GDP, average EU 2000=100) ............ 60 Table 8-16: Consumption of fossil fuel for transport (passenger cars, ktoe) ............................ 61 Table 8-17: Total CO2 emissions for passenger cars (million tons) .......................................... 61 Table 8-18: Average fuel consumption of new gasoline passenger cars (l/100km).................. 62 Table 8-19: Average fuel consumption of new diesel passenger cars (l/100km)...................... 62 Table 8-20: Average fuel consumption of all gasoline passenger cars in stock (l/100km) ....... 63 Table 8-21: Average fuel consumption of all diesel passenger cars in stock (l/100km) ........... 63 Table 8-22: Average age of passenger cars in circulation......................................................... 64 Table 8-23: Average age of removal of passenger cars from circulation .................................. 64


iv

EXECUTIVE SUMMARY The IPTS transport technologies model aims to describe the dynamics of the passenger car market and the introduction of new technologies in the sector. The model simulates the impact of changes in fuel and car prices, technological development, income, and user preferences on the supply and demand of different passenger car technologies. It was originally planned as an extension of the POLES energy market model, but it may be used as a standalone model as well. The main driver for the development of the IPTS transport technologies model was the need for a tool that would help the analysis of policy measures that could lead to the reduction of fuel consumption and emissions through the acceleration of the introduction of alternative technologies. The main objective of the initial phase of development was therefore to construct a model that could explain past trends and provide a plausible outlook for the future, rather than predict the exact future values of the model variables. Although the model is considered as sufficiently accurate in its predictive capacity, its main function for the IPTS is the analysis of “what if?” scenarios describing alternative paths as regards future policy measures, technological development, socio-economic trends and other external factors that may –directly or indirectly- influence the dynamics of adoption of new transport technologies and, eventually, the impacts of transport on fuel consumption and emissions. Overview of methodology The model simulates the way that consumer choices concerning passenger cars are influenced by changes in car and fuel prices, technological development and general socio-economic trends. Data from 1965 to 1998 have been used for the calibration of the model parameters that describe the dynamics of the car market and define the weight of each variable in the model equations. Using exogenous projections for fuel prices, GDP and population growth, and the future technical and economic characteristics of each technological option, the model provides an outlook for the potential of each option and the subsequent implications for energy consumption and environmental impacts. The first step of the simulation is the estimation of the expected changes in car ownership levels for each EU member state. As in most comparable models, car ownership levels are modelled with a Gompertz function that uses the changes of GDP per capita as input. The calibration of the model on the basis of past data allows the definition of the country specific parameters that affect factors such as the saturation level of demand, the changing elasticities as income grows and the differences in the speed of the effects. The exogenous projections needed, i.e. GDP and population growth, are obtained from EUROSTAT and UN projections. The number of new registrations corresponds to the changes in the car park. These are a result of either the change in the overall car ownership level, or of the replacement of cars that are scrapped or removed from the car park (i.e. in the case of used car exports). The number of the cars removed each year from the park is modelled through country-specific survival rate curves for each cohort of cars that has entered the market since 1965. The survival curve rate can itself change over time, either as a result of technological progress or because of higher or lower incomes that accelerate or delay car scrapping. The model distinguishes between 3 types of users for each country, each corresponding to different preferences as regards car use and -as a result- different responsiveness to costs and technological characteristics. The classification of user types for each country was based on urbanisation statistics and projections, while the model parameters that define user choices are


v

endogenous, a result of the model calibration for the years 1980-1997. The model currently includes 7 technological options, conventional (internal combustion engines using gasoline or diesel, light or large) or emerging (electric, gasoline-electric hybrid, fuel cells). The technical and economic characteristics of each technological option have been defined through a number of sectoral techno-economic characterisation studies. These studies have provided a database of historic data for the conventional technologies and of projections by industry experts as regards the outlook for both conventional and emerging technologies. In addition, the model uses the projected future fuel prices of the POLES model in order to estimate the future usage cost for each option. The share of the new registrations that each option can have in the future depends therefore on the combination of the number of users of each type in each country, their responsiveness to the technical and economic characteristics of each technology, the technological progress of each option, and the development of fuel prices in the future. The model also allows the introduction of infrastructure capacity limits. Transport demand in the model is also affected by the fluctuations of fuel prices, GDP growth, the general trends of increased transport intensity, and the changes in the costs for each market segment. Combining the projected demand, the breakdown of the car park in technologies, and the expected efficiency for each technology in each generation, the model provides an outlook for fuel consumption and CO2 emissions for each country, as well as for each of the market segments covered. Figure 1: Overview of model links

Scenario analysis capabilities

POLES model • fuel price projections • GDP and population

Sectoral studies • techno-economic

characterisation

Car ownership dynamics

Fleet ageing

User shares and utilities User costs

Market shares

Transport demand

Fuel consumption & CO2 emissions

IPTS model


vi

The model currently allows four general types of scenarios to be analysed: • Technological scenarios Technological scenarios can provide the outlook for the penetration of new technologies and their impacts on the indicators measured under different assumptions than the ones currently used. Technology development is expressed in the model in terms of fuel economy and car prices for a given level of performance. The values used in the baseline scenario are derived from specific studies on the potential of the various alternative technologies. Alternative scenarios can be constructed by using more optimistic or pessimistic development paths that would change the competitive position of each alternative and influence the speed of its adoption. Technology scenarios in the model are mainly constructed through changing the exogenous variables (input) of the model. • Policy measures Policy measures are perhaps the most interesting type of scenario analysis, and the one that provides more flexibility. The policy measures that scenarios can cover include the following:

• fuel taxes: changing the level of taxation of some or all fuels influences the total demand for transport and the share of each car technology

• carbon taxes: imposing a tax that is based on the carbon content of the fuel used can favour technologies that produce less CO2 and accelerate the introduction of alternatives such as hybrids and fuel cells.

• subsidies: subsidising a specific technology changes its competitive position and increases its sales

• emission limits: imposing an emission limit (e.g. gCO2/km) favours alternative technologies and/or leads to smaller cars being used

• accelerated scrapping schemes: providing a financial stimulus to scrap cars can accelerate the renewal of the car stock and reduce total fuel consumption and emissions

• zero emission zones: imposing a zero emission limit in urban areas leads to the acceleration of the introduction of electric, hybrid and fuel cell vehicles

• combination of policies: e.g. using the carbon tax revenue to subsidise alternative technologies can significantly speed up their introduction

Policy measures in the model can be introduced by changing a number of parameters in the model that, in turn, affect the costs and prices calculated by the model. • Socio-economic trends User choices are influenced by the socio-economic profile and other country-specific issues. They can significantly affect the dynamics of the transport sector, transport demand and the potential of alternative technologies. User choices are modelled in the context of specific socio-economic trends, covering market segments, user types, the degree of urbanisation, the environmental awareness of users, etc. that are endogenous to the model. For example, possible scenarios of socio-economic trends can include different degrees of responsiveness to environmental pressure (user awareness), urban sprawl (urbanisation), changing household structures and demographics. Normally, such scenarios involve modifications of the equations describing the dynamics of the model and, as far as user choices are concerned, price elasticities for each specific user group. • External factors


vii

The main external factors that the model can analyse in terms of scenarios are fuel prices and GDP growth. In its current form, the model uses as input the projections of fuel prices from the POLES model, and the same assumptions for GDP growth that the POLES model uses. Scenarios that can be carried out include the investigation of the impact that a higher or lower price of, and/or a faster or slower economic growth would have on the dynamics of the car market, total demand, fuel consumption and CO2 emissions. Indicative results The model is a scenario analysis tool, rather than a prediction tool. It can provide a general outlook for the future developments in the sector, but this corresponds mainly to the expectations of industry experts as regards the development of the various technological options, and thus strongly depend on whether those expectations are materialised in the future. According to the model’s baseline projections, car ownership levels in the EU-15 are expected to continue to increase but, especially after 2015, will probably reach saturation at values between 600 and 650 cars per 1000 inhabitants (Figure 2). As a result, and in combination with the demographic and technological trends, annual new car registrations are expected to stabilise around 15 million for EU-15. The number of cars removed from the stock (scrapped or exported as used cars outside the EU-15) is expected to rise to almost 14 million per year, as a result of the 10-15 year lag compared with the increase of car ownership in the 1990’s. The total number of passenger cars in circulation in EU-15 will rise to almost 220 millions by year 2020, an increase of 23% compared to year 2000. Figure 2: Car ownership levels in EU-15, data and model projections

0

100

200

300

400

500

600

700

1980 1985 1990 1995 2000 2005 2010 2015 2020

year

cars

per

100

0 in

habi

tant

s

datamodel

Figure 3: New car registrations in EU-15, data and model projections


viii

8

9

10

11

12

13

14

15

16

1980 1985 1990 1995 2000 2005 2010 2015 2020

year

new

regi

stra

tions

(mill

ion

cars

)

test datatest model

Figure 4

Share in New Registrations, by technology (EU)

05

101520253035404550

2000 2005 2010 2015 2020

year

% in

new

regi

stra

tions ELECTRIC

FUELCELLHDIESELHPETROLHYBRIDLDIESELLPETROL

As regards the penetration of new vehicle technologies, the model results suggest that only hybrid vehicles have the potential for a wide scale introduction by 2010 (figure 4). Electric vehicles show a limited potential, concentrated mainly in some niche markets (urban areas in countries with cheap electricity), while fuel cells may capture a significant part of the market only by the end of the 2010’s. Another trend that can be identified is that of the shift from gasoline to diesel. The expected improvements in diesel technology could provide significant


ix

cost savings and comparable performance with gasoline technology. A gradual replacement of gasoline cars with either diesel or hybrid (mainly gasoline-electric hybrids) is therefore expected in the medium term (2010-2015). In the longer term, conventional ICEs and hybrids may gradually lose their share to fuel cells (probably using hydrogen from reformed gasoline, natural gas or methanol), depending on the progress made in the development of fuel cell technologies. Concerning the total number of kms driven in each country, both the number of cars and the average distance are expected to rise and, as a result, total car use may increase by about 33% between 2000 and 2020. Average car use is expected to stabilise around 14000 kms per car per year by 2020. However, notable differences among member states can be seen, due to the different lifestyles, geography, urbanisation and urban sprawl levels, and differences in statistics. Most of the expected changes in the driving factors can lead to increases in the average distance driven, but a certain saturation level for each country is expected to be reached in the next 15-20 years. This projection implies that the overall increase in passenger transport demand that is expected in the next 20 years is most probably going to be covered by other modes (notably air transport for long distances), since car passenger transport will have reached saturation levels. This is highlighted in the projections for transport intensity that corresponds to passenger car transport. The ratio of kms driven to GDP is expected to continue rising until around 2005, but will tend to fall afterwards, since the growth in GDP will not be accompanied by a comparable growth in car passenger transport (figure 5). Total fuel consumption and CO2 emissions are expected to follow a similar trend, reaching a maximum around year 2005 and starting to fall soon afterwards. This is the result of the expected fuel economy of passenger cars improving faster than the rate of growth of total car use. An important part of the improvement of fuel economy is expected to come from the introduction of hybrids and, later on, fuel cells but -even without these alternatives- the evolution of gasoline and diesel ICEs according to the EURO standards and the ACEA agreement should be enough to prevent CO2 emissions from rising further. Fuel economy in Europe is expected to improve, but differences will still exist between member states due to the differences in user choices. The improvement of the average fuel economy for the whole car stock is expected to be even larger, since the majority of the cars that entered into circulation in the 1980’s and 1990’s will have been replaced in the next 10 years by much more fuel efficient cars. The average age of cars in circulation is expected to rise slightly in the next 20 years, from 7.4 to 8.3 years. This is mainly the result of demographics in Europe (age distribution of car owners) and the saturation in car ownership levels (total demand). The effects of improved car technology on the length of a car’s life (either technical or economic) seem to become marginal, and the average age of car scrapping (or removal from circulation in general) is stable. However, in both average age indicators, significant differences exist among member states. These differences are mainly the result of the different socio-economic conditions, car costs, disposable income and (new and used) car market operation in each country.

Figure 5


x

Passenger transport / GDP intensity

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

AUTBLX DNK

ESPFIN

FRAGBR

GRC IRL

ITANLD PRT

RFASWE

EU-15

Member State

km/e

uro2

000 2000

2005201020152020

Figure 6

CO2 emissions from passenger road transport, EU-15

400

420

440

460

480

500

520

540

560

580

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

MtC

O2


11

1 INTRODUCTION It is widely acknowledged that the transportation sector will become one of the fastest-growing energy consumption sectors in the next decades at global level. Passenger transportation demand seems to be a crucial sector in carbon emissions in advanced economies and is likely to be responsible for large emissions increases in emerging economies under a business-as-usual projections. A salient feature in the sector is that passenger transportation becomes more and more focused on road transportation, and, from the technological point of view, relatively locked-in around a single family of technologies, i.e. the internal combustion engine (ICE). This is opposite to what happens in other major carbon-emitting sectors such as the power generation one, where the technology portfolio is becoming more diversified, offering many technology substitution patterns. There are, however, signs that the technological lock-in in transport passenger transportation sector is more and more channelled towards some alternative technological options. A significant change is taking place in the passenger car sector, in the form of the massive introduction of Diesel engines. Electric vehicles, hybrid vehicles and fuel-cell-powered vehicles are technological options that appear to be coming closer to the economic viability threshold. The model presented in this report has been developed in order to provide some prospective insight to what could be the likely evolution of the sector and the related crucial indicators (energy consumption, costs, carbon emissions, etc.). Automobile technology has evolved significantly in the past 20 to 30 years. Up to the 1970’s gasoline was the prime fuel for propulsion of passenger cars and emissions were not an issue at that time. With the oil crisis in 1973 vehicle fuel economy became a major awareness, and alternative fuels and alternative propulsion systems were getting increased attention. At the same time, the awareness for harmful exhaust gas emissions was also rising, leading to the first emission limits in vehicle certification. The spark ignited gasoline engine is still the major power source for current light duty vehicles. Regulated emissions of gasoline engines have been reduced dramatically with the introduction of the three-way catalyst, coupled to electronically controlled gasoline injection (in the throttle body). Gasoline technology has been continuously improved and optimised. With a renewed stress on fuel economy and CO2 emissions (linked to the fear for global warming), fuel saving technologies like gasoline direct injection (GDI) or variable valve actuation are gaining interest. In the past, diesel engines were mainly used for stationary purposes or for heavy duty vehicles. The real diesel engine breakthrough in passenger cars began in the 1970’s. Their improved fuel economy and better durability was the drive for growing sales of diesel vehicles. With the evolution from indirect injection to turbo diesel direct injection and the introduction of electronic diesel control (e.g. common rail), fuel economy was even improved further and emissions were lowered. Noise and performance -the main drawbacks compared to gasoline engines- have improved tremendously in the newest diesel technologies. Concerning exhaust gas after-treatment, the three-way catalyst technology cannot be used for diesel engines, and this is the main reason for the difference in emissions between gasoline and diesel engines. Current diesel engines are usually equipped with an oxidation catalyst to reduce CO and HC emissions, the other emissions are lowered by measures in engine control. However for NOx and PM emissions diesels could not reach the levels obtained by gasoline engines. This may change in the near future with the introduction of particulate traps and DeNOx devices. The automotive diesel became very popular in Europe with sales reaching half total vehicle sales in certain countries. In the United States, however, diesel passenger cars were not


12

successful. This has to do with different fuel pricing strategies, different emission legislation and the bad image of “smoking” diesel engines in the US. In the 1970’s alternative fuels got into the picture for reasons of security of energy supply. By the end of the 1980’s, the growing concern for the environmental impact of automobiles re-stimulated the interest in alternative fuels. The most popular alternative fuels at the moment are LPG, alcohols (ethanol and methanol), natural gas, and for diesel engines biodiesel. Other fuels that have not managed to reach sufficient levels of commercialisation yet are hydrogen (especially for use in fuel cells), DME and synthetic fuels. Most vehicles operating on alternative fuels are also capable of using conventional fuels such as gasoline (in the case of LPG, natural gas, alcohols) or diesel (in the case of biodiesel). The oil crisis of 1973 also initiated the believe in electric vehicles. Numerous projects were launched in the 1970s, built on the assumption that batteries could be improved rapidly. However batteries have remained a major drawback for pure electric vehicles and the EV-market kept well below expectations. By the end of the 1990’s hybrid vehicles were introduced to the market. These do not have the main drawbacks of battery electric vehicles (low range, long recharging), moreover they are environmentally friendly and reach better fuel economy than conventional vehicles. Hybrid vehicles really seem to find their way to the customers. After the year 2000 the manufacturers’ interest in fuel cell vehicles is growing.


13

2 MODEL STRUCTURE The IPTS transport technologies model aims to describe the dynamics of the passenger car market and of the adoption of new technologies in the sector. It was originally planned as an extension of the POLES energy market model , but it may be used as a standalone model as well. It is not a traditional optimization model, but rather a simulation model for the demand for passenger transport, the selection of transportation equipment (the car fleet and its relative characteristics), as well as the estimation of the potential impacts in terms of fuel consumption and emissions. Only passenger car transportation has been considered in the model in the current stage, because of its growing relative importance, and special emphasis has been put on seven car technologies:

Car technologies Internal combustion engine vehicle (ICEV):

• Gasoline (small size)

• Gasoline (large size)

• Diesel (small size)

• Diesel (large size)

Hybrid electric-internal combustion vehicle (HEV)

Battery powered electric vehicle (EV)

Fuel cell powered vehicle (FCV)

The object of the modelling exercise is to create a tools capable of capturing the process of technology substitution in the sector, in response of policy instruments as well as external boundary conditions such as the regulatory framework, the evolution of fuel prices, technology costs, etc. In this respect, the model should be able to describe past trends in the sector, such as the on-going “dieselization” of the car fleet in many countries, as well as the possible impact of accelerated technology substitution schemes (early retirement subsidies, etc.) in carbon emissions, and energy demand from the sector. The model distinguishes three customer categories (URBAN, COUNTRY, and SALESMAN), depending on the annual mileage required by them. The geographic coverage of the model is:

Countries Acronym Name Acronym Name

FRA France SWE Sweden GBR United Kingdom AUT Austria ITA Italy FIN Finland

RFA Germany IRL Ireland ESP Spain CAN Canada PRT Portugal USA United States GRC Greece MEX Mexico BLX Belgium and Luxembourg CHN China NLD The Netherlands JPN Japan DNK Denmark NDE India


14

In that respect, future extension of the model will consist in the addition of the Former Soviet Union, the Central and Eastern European countries, and the South American countries. The final goal is to reach the same geographical detail of POLES. The model runs in a normal PC using Vensim 5.0. The database is made up of two Microsoft Excel files containing information about car prices and costs, fuel costs, fleet, GDP, population, energy consumption, emissions, new registrations and technical parameters. The data sources are very diverse: EUROSTAT (Transport and Communications Yearbook), OECD (Statistical Trends in Transport, CO2 Emissions from Transport), US Department of Energy (Transportation Energy Data Book, Transportation Energy Efficiency Trends), International Road Federation (World Road Statistics), IEA (Energy Prices and Taxes) and the POLES model database. The cost data used for EV, HEV and FCV come from studies made bay the Institute of Transportation Studies, University of California, and Energy and Environmental Analysis, Inc.


15

3 CAR OWNERSHIP AND NEW REGISTRATIONS The decision to buy, substitute or scrap a passenger car is a choice made by each individual on the basis of various financial, socio-economic and technological parameters. Car ownership is an important element of modern society and its level often reflects a country’s average income and underlying consumption patterns. Although the degree may vary depending on local conditions, there is an evident correlation between car ownership and per capita income that has been consistent in the past and is expected to remain so in the future. The relevant literature generally suggests that aggregate car ownership levels increase with income until they reach a saturation level, when the income elasticity of car ownership falls to zero [Greenspan and Cohen (1996),Dargay and Gately (1999), Schafer and Victor (2000), Medlock and Soligo (2002)]. Saturation levels for Western Europe and North America are estimated to range between 600 and 700 passenger cars per 1000 inhabitants. Such levels would mean that the majority of holders of a driving license who can afford owning a car actually owns one (or more). At the level of the individual, the decision to buy a car depends on household income and utility, and can be compared to the purchasing behaviour demonstrated for typical durable goods. In general terms, the majority of households buy at least one passenger car if their finances permit it, and replace older cars for newer ones -if they can afford it- in order to increase performance, reduce maintenance costs, or just enjoy the feeling of buying a new car. The household’s income is the main determinant for the purchasing decision and the chances of buying a car increase when income increases. On the other hand, the price of the car, as well as the cost of using and maintaining it, act as inhibiting factors. An example of how these three variables, i.e. household income, fixed costs and variable costs, influence car ownership is given in Table 3-1. Country specific parameters also affect the extent of the impact of each variable. Differences in urbanisation or the quality of public transport may explain a different behaviour in different countries. People living in metropolitan areas with good public transport services do not need a car as much as people living in rural areas.

Table 3-1: Income, variable cost and fixed cost elasticity of car ownership

Norway Denmark Holland Income +0.33 +0.41 +0.15 Variable costs -1.33 -0.78 -0.41 Fixed costs -2.65 -1.29 -0.80 Source: Bjorner (1999) Demographics and lifestyles also play an important role. A household owning a second or third car is becoming a frequent case (Figure 3-1). The increased participation of women in the workforce in the last 30 years (and the additional income they bring to the household) has largely contributed to the increase in car ownership. In addition, the fact that a larger proportion of the younger generations holds driving licences, a phenomenon even more evident for women, suggests that a larger number of potential car buyers exists today compared to ten or twenty years ago (Figure 3-2). As surveys from France and the UK show, the gaps between generations, which have been important in the past, have been smoothened out for households whose head is born after the 40's, implying that the diffusion of car ownership over generations is nearing completion [Dargay, Madre et al. (2000)].


16

Figure 3-1 : Households with regular use of car, UK

Figure 3-2: Driving licence holders statistics, UK

Source: DETR (2001) The new car purchases at an aggregate level correspond to the sum of the effect of the change in car ownership levels plus the substitution of cars that are removed from the car park (scrapped or exported as used cars). This implies that the overall car ownership level of a country is determined by the average income, the extent to which car ownership levels are saturated, and the country-specific parameters related to urbanisation and transport patterns. A number of cars are also removed each year, being too old or too expensive to maintain. New car registrations represent the cars bought by individuals that did not own a car before (because they didn’t need or want one, or because they couldn’t afford one, or because they were too young to have a driving licence), and the cars bought in order to replace a car that was scrapped. The type of car purchased depends on numerous variables, not all of them quantifiable. Purchase and use costs are certainly important parameters, and probably provide a reliable indicator of market segmentation: high-price cars are bought by high-income households, low-income households are more sensitive to price differences, etc. There are, however, numerous non-economic factors that influence the type of car the individual consumer buys. Such factors may include the prestige associated with a certain brand or car size, fashion or lifestyle trends, advertising and marketing, and technological progress itself. The combination of the above factors in the last decade has resulted in a highly fragmented car market with strongly connected segments. The share of the small and lower-medium segments has increased, probably as a result of the gradual fall of car prices compared to the lower and medium income levels (Table 3-2). The fact that more young drivers and women buy cars, combined with the increasing number of second and third cars in one household, is probably the main reason. But issues such as congestion, the lack of parking space in urban areas, or the increasing environmental concerns of consumers could also be drivers for the trend toward smaller cars. The introduction of modern small cars such as Smart or Mercedes A1 can be seen as either a trend-setter or, from the opposite point of view, as the reaction of auto manufacturers to the trends in society. Consumer tastes have also changed concerning body type. Saloons dominated the market in 1995, but newer car designs are rapidly increasing their share in consumer preferences. The trend towards more Sport-Utility Vehicles (SUVs) that was impressive in the U.S.A. during the last decade seems to be taking off in Europe too. The share of another special category, 4x4 passenger cars, has doubled in the last 10 years, reaching 5% of sales in 2001 [ACEA (2002)].


17

Table 3-2: New cars in EU-15 by segments and bodies

source: ACEA (2002) Consumers demonstrate a strong tendency towards cars of higher performance (Table 3-3). Aided by the relative drop in car and fuel prices compared to average income which is accompanied by (and, to a large extent, driven by) improvements in car technology and fuel efficiency, the average power of new cars has risen dramatically during the last decade. Another important factor for this trend is the fact that a large share of new car sales is in fact a replacement of the car that the individual owned before. Consumers tend to upgrade cars; they often sell or scrap their old car in order to buy a larger, faster or more expensive one. This results in an increased cost for the consumer that cannot be explained in strict economic terms. The utility of a car with improved performance, probably better than that actually needed, cannot be measured, since it depends on each consumer’s perceived needs and choices. The type of engine and/or fuel used seems to be a secondary decision. Since only two real alternatives exist today, i.e. gasoline and diesel fuelled internal combustion engines, the individual’s choice depends largely on price, performance and use costs. Since diesel technology has made great improvements in the last 10 years in terms of performance, fuel efficiency and environmental impacts, the share of diesel fuelled sales has risen from under 14% in 1990 to over 35% in 2001. Except some limitations for diesel cars in urban areas (e.g. in Greece), and the still limited supply of a full range of diesel models (e.g. Japanese cars or smaller models), there doesn’t seem to be any other differentiating factor between gasoline and diesel cars of comparable performance, apart from cost. However, the same cannot be easily said about emerging technologies such as electric cars, fuel cells or car fuelled by natural gas or bio-fuels. Whereas gasoline and diesel are established fuels, and the internal combustion engine is a proven technology, the emerging alternatives still lag in terms of maturity, infrastructure and fuel availability, safety and public perception, and are not yet considered as alternatives by consumers. Whether they will eventually become such, when their cost falls to competitive levels, remains an open question.


18

Table 3-3: Average engine power of new cars

source: ACEA (2002)

Table 3-4: New passenger car registrations in Western Europe, Jan-Oct 2002

Table 3-5: Passenger car sales by region

Year 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Western Europe

13233 13500 13497 11450 11934 12021 13083 13138 14038 15049 14742

NAFTA 10532 9466 9449 9653 10154 9424 9390 9333 9357 10023 N/a Japan 5103 4868 4454 4200 4210 4444 4669 4492 4093 4154 4260 Asia (excl Japan)

1994 2086 2355 2853 2972 3267 3533 3599 2468 3333 1063

Eastern Europe

1995 1697 1731 1654 1560 1533 1729 1906 1820 1900 N/a

Other Markets

1730 1775 1989 2344 2617 2970 3088 3423 3012 2706 N/a

Total 34587 33392 33475 32154 33447 33659 35492 35891 34788 37165 N/a Source: Economist Intelligence Unit

Country JAN FEB MAR APR MAY JUN JUL AUG SEP OCT YTDPC-Passenger Cars

EU (15) EC (12) B 55,872 52,611 52,851 51,858 40,912 37,632 34,850 26,145 31,678 37,702 422,111DK 7,075 8,132 10,240 10,030 11,889 11,204 9,190 8,858 8,752 8,770 94,140F 192,089 173,064 198,302 207,891 186,265 200,835 199,321 121,781 153,069 188,209 1,820,826D 243,400 233,281 321,840 304,326 282,632 308,450 282,560 244,287 266,005 281,059 2,767,840GR 27,021 21,403 23,348 29,365 23,403 23,003 29,109 19,310 21,154 21,065 238,181IRL 31,075 24,173 18,814 15,482 18,456 11,724 12,468 8,742 6,348 5,075 152,357I 248,300 197,200 205,500 191,300 209,100 180,300 202,400 101,300 165,200 190,600 1,891,200L 3,368 4,713 5,278 5,276 4,068 3,779 3,782 2,191 2,798 3,565 38,818NL 70,838 42,523 47,046 40,361 44,728 46,953 39,907 36,261 40,908 43,173 452,698P 18,960 19,831 23,476 18,892 25,968 24,030 25,450 13,566 13,737 15,353 199,263E 95,946 105,271 127,211 112,851 127,698 125,734 152,945 79,416 80,812 111,783 1,119,667UK 205,476 93,515 423,727 208,976 208,669 207,330 195,637 87,245 432,661 184,145 2,247,381EC (12) 1,199,420 975,717 1,457,633 1,196,608 1,183,788 1,180,974 1,187,619 749,102 1,223,122 1,090,499 11,444,482

A 23,291 20,398 29,666 30,320 26,295 25,904 26,260 18,813 21,402 24,388 246,737SF 12,132 8,446 9,519 11,066 12,709 10,866 10,287 10,221 8,985 9,719 103,950S 16,293 18,848 22,936 24,311 24,617 22,861 17,410 18,618 22,101 23,249 211,244Efta (3) 51,716 47,692 62,121 65,697 63,621 59,631 53,957 47,652 52,488 57,356 561,931

total EU (15) 1,251,136 1,023,409 1,519,754 1,262,305 1,247,409 1,240,605 1,241,576 796,754 1,275,610 1,147,855 12,006,413 EFTA (3) IS 423 418 410 530 906 1,081 725 567 474 0 5,534

N 7,386 7,024 7,068 8,833 7,912 7,324 8,724 7,273 6,985 8,089 76,618CH 22,001 21,754 28,087 29,947 29,668 29,396 26,686 19,803 21,348 22,827 251,517

total Efta (3) 29,810 29,196 35,565 39,310 38,486 37,801 36,135 27,643 28,807 30,916 333,669

total Total 1,280,946 1,052,605 1,555,319 1,301,615 1,285,895 1,278,406 1,277,711 824,397 1,304,417 1,178,771 12,340,082

Source: Association Auxiliaire de l'Automobile


19

3.1 Modelling of car fleet and new registrations The number of cars per capita is closely related to the per capita income level. The relationship has the form of an S-shaped curve; car ownership increases slowly at the lowest income levels, and then more rapidly as income rises, finally to slow down as saturation is approached. The theoretical basis for the estimation of the overall car ownership level comes from Dargay and Gately(1999). They use the Gompertz model, where the long run level of the vehicle-population ratio can be written as:

( )*t t

GteC G eβαγ ⋅ ⋅

= ⋅ In this equation α and β are negative parameters defining the shape, or curvature, of the function and γ is the saturation level, since for β<0 *lim ( )

tt

tGC G γ

→∞= . The parameter α

determines the value for Gt =0, i.e.:

*t tG C γ → ∞ ⇒ →

( )*0 0t tG C eαγ= → = ⋅ The long run income elasticity is calculated by appropriate differentiation as:

*lnln

tGtLR t

t

d C G ed G

βη α β ⋅= = ⋅ ⋅ ⋅

Figure 3-3: Car ownership function based on the Gompertz model

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

GDP per capita (kEUR)

Vehi

cle/

Popu

latio

n

The cars per capita have are modelled by using a modified Gompertz function. The Gompertz function describes the long run relationship between vehicle ownership and per capita income. In order to account for lags in the adjustment of vehicle ownership to per capita income, a simple partial adjustment mechanism is implemented:

*

1 1( )t t t tC C C Cθ− −= + ⋅ −


20

where θ is the speed of the adjustment (0<θ<1). The lags represent the slow adjustment of car and vehicle ownership to increased income due to the necessary build-up of savings to afford ownership, the gradual changes in housing patterns and so on. Substituting Ct

* in the last equation, the result is the equation used in the model:

1(1 )Gte

t tC e Cβαγ θ θ

⋅⋅−= ⋅ ⋅ + − ⋅

where γ is the saturation level, θ is the speed adjustment factor of the curve, α and β are negative parameters defining the shape of the Gompertz function and Gt is the GDP per capita (in thousands of € ppp per capita). The total number of car registered is estimated regardless the car category, technology or customer profile. This total number of car registered is distributed among customer categories and technology types according to the rules described in the following sections. The model uses the value calculated for speed adjustment given in Dargey (1999) and the adapted saturation levels calculated by Medlock and Soligo (2002). The curve parameters were re-calibrated in order to allow for the change in GDP per capita reference year. The number of new registrations corresponds to the changes in the car park. These are a result of either the change in the overall car ownership level, or the replacement of cars that are scrapped or removed from the car park (i.e. in the case of used car exports). The number of new cars registered per country (or the car sales) is a function of the variation of the number of cars per capita, the population and the total amount of scrapped vehicles:

( )1t t t t tNCR C C POP TOTSCR−= − ⋅ + where: NCRt is the number of new car registrations at time t Ct the per capita car ownership TOTSCRt the number of cars removed from the car park during the period 3.2 Dynamics of car scrapping and fleet ageing In the model, car scrapping refers to the case of a car being removed from circulation. This can be the consequence of one or more of the following: • the car is not usable any more (e.g. because of mechanical problems or as a result of an

accident) • the car is too expensive to use or maintain: other options such as buying another car or

owning no car at all are more attractive for the owner, and no other buyer can be found • the car is sold in another country as a used car: this case has more to do with statistics

rather than with actual scrapping, but it also implies that the perceived value of the used car in the country of origin is lower than in the destination

• specific regulations or measures prohibit the use of a specific type or age group of cars, or stimulate the early retirement of older vehicles

In all of the above cases, the underlying variables are the cost of use and the remaining value of the car. A micro-economic approach modelling the decision to buy or scrap a car was used in Adda and Cooper (1997) in order to estimate the impacts of car scrapping subsidies in France. Individuals are assumed to have a certain utility from buying or scrapping a car, and a


21

change of the relative costs through government intervention greatly affects the overall car market, accelerating the speed of replacement when scrapping subsidies are given. The models used by Alberini, Harrington et al. (1998) simulate the owner's decision to keep, repair or scrap their old vehicles. This decision depends critically on the owner's perceived value of the vehicle that, in turn, depends on the mileage and condition of the car, and declines systematically with its age. Greenspan and Cohen (1996) divide scrapping into two types, engineering scrapping and cyclical scrapping. Engineering scrapping results mainly from age-dependent physical wear and tear, while cyclical scrapping is a result of business cycles. The non-engineering component of scrapping also depends on the price of gasoline, the price of new vehicles, and on the cost of repairs. An increase in their price delays the purchase of new vehicles and the scrapping decision, while a reduction in the cost of repairs encourages increased repair of vehicles and less scrapping. The price of new vehicles relative to repair costs is considered as significant in explaining total scrapping. The dynamics of the car park and, subsequently, the scrapping behaviour in each country were modelled on the basis of the historical data on new car registrations and car park figures, since only limited data were available concerning car scrapping. In the general form, the dynamics of fleet ageing are described by a set of 3 equations:

( )( )

,0

,

1b i

t car icar i

car i t i i

a e

i

CARFLEET SURPC

SURPC NCR SR

SR e− ⋅

∞

=

−

⋅

=

= ⋅

= −

∑∑

Where CARFLEETt the number of cars in circulation at time t SURPCcar,i the number of car survivors of category car per period (cohort), i.e. from those that belongs to category car and are i years old at time t. i the age of the cohort SRi the survival rate of the cohort i years old, which is computed also following a Gompertz survival model. The parameters a and b define the survival rate curve for the specific country and technology. Their values where calibrated through the process of multi-variate optimisation that determined the values that provide the best match between the car park numbers this set of equations calculates and the data from statistics between 1970 and 1998. The survival curve rate can itself change over time, either as a result of technological progress or because of higher or lower incomes leading to scrapping cars earlier or later. The model allows the incorporation of the technological progress in the determination of parameters a and b and therefore endogenise it. However, the short-term economic parameters that affect scrapping are not included, since they are outside the scope of the model. A secondary control of the calibration consists of the comparison of the calculated average age of the fleet with published data:

( )( ),0

0.5car icar i

tt

SURPC iAVGFLTAGE

CARFLEET

∞

=

⋅ +=

∑∑

An inherent problem of this method of calculation is the average age of the surviving cars that are older than 15 years old, i.e. the tail problem of the above series. For the cohorts older than


22

15 years, the average value used in the model is 17 years. Given the small number of cars that fall into that category, the final result of the calculation is influenced in a very limited way. An additional stage in the calculation involves at each time step the re-grouping of the cohorts into four age groups (SURcar,agegroup, survivor per age group). The four groups are “new” (for cars below 5 years old), “semi new” (between 5 and 10 years old), “mature” (between 10 and 14 years old) and “old” cars (for cars above 15):

( )

( )

( )

( )

4

, , , ,1

9

, , ,5

14

, , ,10

, , , , 1 , 15 16

car new t car t cart t ii

car smn t car t ii

car mat t car t ii

car old t car old t car t

SUR NCRT SURPC

SUR SURPC

SUR SURPC

SUR SUR SURPC SROT

−=

−=

−=

− −

= +

=

=

= + ⋅

∑

∑

∑

where SURPCcar,t-i is the number of survivors per period and car category in the year t-i and SROT16 is the survival rate for old cars (older than 16). The summation of these values is the fleet per category. The total fleet is the summation by car category. These two values allow computing the share of each technology in the fleet. The equation for the scrapped cars from the category "old" is:

16 , ,(1 ) car old tcar

OLDSCR SROT SUR= − ⋅∑

The total number of scrapped cars, from all categories, is:

( )15

, 1 151

t car t i t i t i tcar i

TOTSCR NCRT SRPP SRD OLDSCR SCRPP− − − − −=

= ⋅ ⋅ + −∑∑

The scrappage can be defined in terms of two components. The first one is the “engineering” scrappage and reflects physical or built-in deterioration, i.e. the physical wear and tear that increases with vehicle age or use, although by an amount that varies with model year. The second component reflects the cyclical scrappage, the prices of new vehicles, repairs and fuels. This kind of scrappage is not considered in the model.


23

Figure 3-4: New car registrations versus scrapped cars (Thousand cars/year), EU-15

0

2000

4000

6000

8000

10000

12000

14000

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

new car registrations cars removed from the fleet

source: IPTS

Table 3-6: Vehicle scrapping and recycling in the UK

1997 1998 1999 2000 Number of End of Life Vehicles

Cars

1700000

1600000

(est.) 1600000

1832431

Vans 200000 200000 200000 184706 Total 1900000 1800000 1800000 2017137 Average weight of vehicle (kgs)

1025 1030 1030 1030

Weight of vehicles for disposal

1947500 1854000 1854000 2078000

Re-use and recovery percentage

76% 74% 77% 80%

Source: SMMT


24

Table 3-7: Estimates of number of vehicles scrapped

TRENDS/RT Müller 1994deregistered scrapped

Country Sector vehicles vehicles percentageB Buses 2094 496 24%B Heavy Duty Vehicles 13856 51 0%B Light Duty Vehicles 19228 16859 88%B Motorcycles 18636 6571 35%B Passenger Cars 405394 274546 68%D Buses 9835 3795 39%D Heavy Duty Vehicles 96689 57988 60%D Light Duty Vehicles 117313 70304 60%D Motorcycles 280921 76160 27%D Passenger Cars 1968036 2147133 109% *)DK Buses 983 377 38%DK Heavy Duty Vehicles 14188 9484 67%DK Light Duty Vehicles 18320 12268 67%DK Motorcycles 10517 1230 12%DK Passenger Cars 128405 88932 69%E Buses 2633 1768 67%E Heavy Duty Vehicles 13049 8255 63%E Light Duty Vehicles 66840 42391 63%E Motorcycles 139942 14596 10%E Passenger Cars 593879 341224 57%F Buses 7718 3483 45%F Heavy Duty Vehicles 51883 45771 88%F Light Duty Vehicles 265745 235034 88%F Motorcycles 194856 85736 44%F Passenger Cars 1568162 1607248 102% *)GR Buses 487 243 50%GR Heavy Duty Vehicles 826 780 94%GR Light Duty Vehicles 3303 3121 95%GR Motorcycles 4714 1161 25%GR Passenger Cars 17784 21564 121% *)I Buses 382 3249 852% **)I Heavy Duty Vehicles 5849 29382 502% **)I Light Duty Vehicles 4026 48762 1211% **)I Motorcycles 3439 110107 3202% **)I Passenger Cars 63696 956722 1502% **)IRL Buses 4636 45 1% ***)IRL Heavy Duty Vehicles 41851 7349 18% ***)IRL Light Duty Vehicles 69315 5064 7% ***)IRL Motorcycles 295574 3113 1% ***)IRL Passenger Cars 1328675 52108 4% ***)L Buses 83 21 25%L Heavy Duty Vehicles 1752 1 0%L Light Duty Vehicles 465 127 27%L Motorcycles 1063 238 22%L Passenger Cars 18341 12937 71%NL Buses 1828 421 23%NL Heavy Duty Vehicles 63800 22340 35%NL Light Duty Vehicles 539 22340 4144% ****)NL Motorcycles 39833 6570 16%NL Passenger Cars 505493 386195 76%P Buses 287 155 54%P Heavy Duty Vehicles 1500 10117 674% *****)P Light Duty Vehicles 1668 11226 673% *****)P Motorcycles 16865 138 1% *****)P Passenger Cars 31349 11296 36%UK Buses 9864 4674 47%UK Heavy Duty Vehicles 48965 61093 125% *)UK Light Duty Vehicles 195824 244369 125% *)UK Motorcycles 67649 104135 154% *)UK Passenger Cars 1620850 1419677 88%*) Impossible figure.**) Deregistrated vehicles for Italy from TRENDS/RT possible too low by factor of 10-40.***) Figures for Ireland from TRENDS/RT too high with regard to population.****) Deregistrated LDV from Holland from TRENDS/RT too low by approx. a factor of 100.*****) Deregistrated HDV+LDV from Portugal too low, MC too high by a factor of approx. 10

Source: LAT (2002)


25

Figure 3-5: Average age of passenger cars, EU

source: EEA, Eurostat

Figure 3-6: Average age of passenger cars and heavy duty vehicles, accession countries (AC)

Source: EEA


26

Figure 3-7: Automobile Survival Rate (US) [Transportation Energy Data Book, 21]

Automobile Survival Rate

0,0

20,0

40,0

60,0

80,0

100,0

4 7 10 13 16 19 22 25 28

Vehicle age (years)

Surv

ival

rate 1970 model year

1980 model year

1990 model year

Figure 3-8: Automobile Scrappage Rate [Transportation Energy Data Book, 21]

Automobile Scrappage Rates

0,0

5,0

10,0

15,0

20,0

25,0

30,0

35,0

0 10 20 30

Vehicle age (Years)

Scra

ppag

e R

ates

1970 model year1980 model year1990 model year


27

Figure 3-9: Automobile Medium Lifetime [Transportation Energy Data Book, 21]

Automobile Median lifetime

11,5 12,5

16,1

02468

1012141618

1970 model year 1980 model year 1990 model year

Year

s

Figure 3-10: Life Span of Cars by country [APME, 1999]


28

3.3 Used cars An intermediate phase between the purchase of a new car and its eventual scrapping is the sale of the car as a used car. The same principles apply in this case for the buyer and seller of a used car. The buyer may be a first time buyer (usually young), or somebody replacing an even older car, who cannot afford (or does not consider it worth the cost) to buy a new car of similar characteristics. The fact that the various segments of the car market are strongly connected also means that some buyers prefer to buy a larger second hand car than a smaller new car. From the seller’s point of view, in the majority of cases the sale of a car is related to the purchase of a newer and/or larger car. Statistics at an EU level show that the used car market is very active, and in most cases larger than that for new cars (Table 3-8). Especially for the UK and the Netherlands, the figures suggest that the average car should be expected to change owner at least 4 times before it is finally removed from circulation. Data from the UK shows that although older cars form the larger part of the used car market, a significant number is less than 2 years old. The factors that influence the dynamics of the used car market are to a large extent cyclical. New car prices, parallel imports, interest rates, insurance costs, emission related taxes, warranty regulations, as well as the EU guarantees legislation and the ELV recycling legislation affect the price and the attractiveness of used cars. In addition, the behaviour of company cars and rental fleets influences the supply of second hand car, while the distribution structure of both new and used cars influences the market’s operation. Demographics still play a role, with car driver profiles and buying behaviour largely affected by age and income.

Table 3-8: Used car sales in various EU countries

Used car sales (m)

Used car sales per 1000 population

Used/new car sales ratio

UK 7.5 126 3.3 Germany 7.4 91 2.0 France 4.7 80 2.4 Italy 2.2 38 0.9 Netherlands 1.7 110 3.2 Spain 1.4 35 1.2 Belgium 0.7 63 1.4 Portugal 0.5 49 1.9 Denmark 0.4 67 2.2 Sweden 0.3 30 1.1 Norway 0.3 77 2.4 Source: UKCC (2000)

Table 3-9: Passenger car registrations in France (‘000s)

Year 1985 1990 1995 2000 2001 New cars 1766 2309 1931 2134 2255 Used cars 4803 4759 4129 5082 5396

Source: METLTM

Table 3-10: First time passenger car registrations in Ireland

Year 1997 1998 1999 2000 2001


29

New cars 125818 138538 170322 225269 160908 Used cars 41554 39565 36878 24003 15237

Source: CSO

Table 3-11: Car sales in USA

Year 1990 1995 1997 1998 1999 2000 New cars (000s) 13890 14730 15130 15600 16960 17410

Used cars (000s)

37530 41758 41240 40840 40740 41620

Av. price new (US$)

16350 19819 20214 20276 20534 21850

Av. price used (US$)

5830 7776 8164 8213 8674 8715

Source: CNW

Table 3-12: Breakdown of the volume of used car sales by age of car, UK, 1998

Age of used cars (years) %

0-2 13.2 3-5 22.3 6-8 19.5

9 and over 45.1 Total 100.0

Volume (m) 7.5 Source: UKCC (2000) Important the used car market may be, for most of the purposes the model presented in this paper aims at its consideration can be skipped. Indeed, in aggregate terms, the environmental effect associated to the car fleet is almost independent of the ownership patterns (in principle it seems reasonable to assume that used cars are mainly used in the same customer category they were originally purchased. The selected way adopted to model the demand for new car registrations (as an indirect function of the floating stock of cars plus an accurate description of the scrapping processes) allows us to avoid to deal in great detail with second-hand car market. This is certainly not so when international second hand car trade is concerned (an issue to take into consideration specially in accession countries, for instance). For the time being, this international used car market has not been considered.


30

4 MARKET SEGMENTATION The model allows the segmentation of the market into both technologies (CATCAR) and user type (CATCUSTOMER) and allocates new car registrations per technology for each user group. The underlying assumption is that each type of user has different preferences that can expressed in terms of utility. The utility for each category of use or customer is expressed as follows:

1 1 1 2

2, , 1

t t t t t tcust cust cust

t t t

G G P P C CG P C

cust t cust tU U eα β γ− − − −

−

− − −⋅ + ⋅ + ⋅

−= ⋅ where: • αcust is the adjustment parameter for the income per capita. • Gt is the GDP per capita. • βcust is the adjustment parameter for urban population • Pt is the percentage of urban population. • γcust is the adjustment parameter for the number of cars per capita • and Ct is the variable the number of cars per 1000 inhabitants. These utilities are used to compute the share of each customer in the total number of new car registrations per country:

,,

,

cust tcust t

cust tcust

UCUSTSH

U=

∑

The idea of the utility equation is to represent the following evolution of the customer shares as the per capita income increases.

Figure 4-1: Assumed development of car segment split according to owner categories.

Assumed development of car segment split according to owner categories

0%10%20%30%40%50%60%70%80%90%

100%

Not Developed Low Developed Developing Developed Richest

Salesman (business car)Country (main car or household)Urban (second or further car)

Having split the total new registrations per customer category, the decision to select some particular car technology depends on the utilization costs associated with each car category. Such costs depend on the fuel costs, the fixed costs (discounted during the lifetime of the vehicle) of each type of car, the expected mileage, and the stack/battery replacement costs in the case of electric, hybrid and fuel cells.


31

4.1 The Wood allocation algorithm The share of each car category per customer is computed in two stages. Initially, it is assumed that there are no capacity constraints, and each market segment is allocated its “a-priori” share following a Weibull distribution with the adjusted user cost for each technology as input:

,,

,

bcar cust

car cust bcar cust

car

ADCOSTDEMSH

ADCOST=

∑

where: DEMSHcar,cust is the share of registrations of car category car for each user group cust (demand share without capacity restrictions). ADCOSTcar,cust is the adjusted (by the user preference) cost of car category car for user cust b is the parameter of the Weibull distribution. This “a-priori” share can be viewed as a kind of absolute measure of the “atractiveness” of a given car (or technology) category within each customer category. The second step of the calculation involves the application of the Wood algorithm that allocates demand by priority to the technologies. This is done by using the Vensim function "ALLOCATION BY PRIORITY". Before explaining the particular application of the Wood algorithm to thi sparticular allocation problem, an overview of the method is given in the following. The variables used in the function are the total demand that needs to be met (TOTDEM) , the individual capacities of supply of each technology (CAPi, with 1<=i<=N), an indicator of the attractiveness of each technology (PRIi), and a parameter that defines the minimum and maximum limit of attractiveness at which the technology can be allocated a share (WIDTHi). The function "ALLOCATION BY PRIORITY" returns an allocation (ALLOCi) of the total demand (TOTDEM) to each supplying technology:

( , , , , )i i i iALLOC ALLOCATE BY PRIORITY CAP PRI N WIDTH TOTDEM=

The Wood algorithm has five desired properties:

• Committed capacities of all the technologies must sum to the total demand, under all conditions.

ii

ALLOC TOTDEM=∑

• All commitments and market shares must be positive.

0iALLOC i≥ ∀

• No technology can commit more than its capacity.

i iALLOC CAP i≤ ∀

• Under conditions of extreme excess demand, each technology should use its entire capacity.

i i ii

TOTDEM CAP ALLOC CAP i> ⇒ = ∀∑


32

• Under conditions of limited demand, uniquely unattractive technologies should use little or nothing. If there is a uniquely attractive technology with high capacity, it should win virtually the entire market, shutting out its competitors.

In order to demonstrate how the algorithm works, a graphical example would be useful. Let us assume that only three technologies have to fulfil the demand for new car registrations. These three technologies are named A, B and C. For each of these technologies, the potential allocation for each technology can be represented with a rectangle whose base length is equal to the parameter WIDTHt and whose area corresponds to the total capacity for this particular technology. The capacity in this particular case would be determined by the infrastructure i.e. refuelling stations availability, etc as well as other considerations, like the production capacity of the car manufacturing sector itself. On the horizontal axis (attractiveness), the midpoint of each rectangle corresponds to the attractiveness of the respective technology PRIi. Its boundaries lie at PRIi -/+ 0.5*WIDTHi

The operational principle of the Wood algorithm Given as exogenous parameters for each technology the total capacity CAPi and the width WIDTHi, the height of the rectangle representing each technology is given by:

ii

i

CAPHEIGHTWIDTH

=

height(B)

A

B

C

width (B)

PRI(B) PRI(C) PRI(A)

Atractiveness/Priority


33

The allocation algorithm proceeds by sweeping the capacity map from left to right, until total demand TOTDEM matches the sum of the swept areas in all rectangles (i.e. the dashed areas in Figure 1). A value of WIDTH equal to 0 would lead to “winner-takes-it-all” situations, since the most attractive technology would meet the whole demand (as long as its capacity permitted it), and no capacity rectangle would overlap. Very high values of WIDTH in relation to the attractiveness of each technology PRI would reduce the importance of the difference in attractiveness between technologies and lead to a more uniform distribution of shares among them. Back to the allocation procedure in the passenger transportation sector, the number of technologies under car is N=7 (LPETROL, HPETROL, LDIESEL, HDIESEL, ELECTRIC, FUELCELL, HYBRID). The case under consideration is a very particular one, since the allocation in the transportation sector will take place among technologies very mature (HPETROL, LPETROL, LDIESEL, HDIESEL) and totally non-mature technologies (ELECTRIC, FUELCELL, HYBRID). Mature technologies do not exhibit capacity constraints, i.e. they would be able to satisfy the entire demand for new car registration. In this respect, CAPcar,cust for these technologies should be very large (and, accordingly, also HEIGHTcar,cust). The capacity for the three emerging technologies is limited by both the production capacity constraints and the limitations imposed by the refuelling systems availability, and prescribed exogenously according to input acquired from specific studies and expert advice. In the practical implementation of the Wood algorithm in the passenger transportation sector, the capacity of supply is expressed as the share of total demand that the technology can satisfy. For conventional technologies (i.e. diesel and gasoline) it is assumed that capacity equals 100% of demand, i.e. there are no capacity constraints. For emerging technologies, and in particular for fuel cells, the capacity is variable and expressed as the potential maximum percentage that could be supplied by the corresponding technology. The parameter WIDTH controls the degree of overlap of the level of attractiveness of the various alternative technologies. In this model, we will use as a measure of priority the above-mentioned share of the technology if no capacity constraints existed, DEMSHcar,cust. All technologies under car should have the same WIDTH, since it is assumed that the relative weight of the attractiveness of each technology -and therefore their degree of overlap- is the same. The value of WIDTH used in this application is a function of the value of the attractiveness of the best alternative:

( ),cust car custcarWIDTH Max DEMSHλ= ⋅

where λ is the overlap factor ( 0λ ≥ ); in the case of this application, selecting 2λ = has resulted in a distribution of shares that fit historical data for gasoline and diesel closely. The total demand for new registrations to share out among the car categories is given a value of 1 (corresponding to 100% of new registrations), so the final formulation in the model becomes:

, ,( , , , ,1)car cust car car cust custCARSH ALLOCATE BY PRIORITY CAP DEMSH N WIDTH= The Wood algorithm is applied independently for each customer category cust.


34

Having calculated the share of each technology in the demand of each user type (CARSHcar,cust), the new registrations according to technologies and user groups can be calculated:

, ,

, , , , ,

, , ,

cust t t cust t

car cust t car cust t cust t

car t car cust tcust

NCRC NCR CUSTSHNCRCCFIN CARSH NCRC

NCRTFIN NCRCCFIN

= ⋅

= ⋅

= ∑

where: NCRCcust,t is the number of new car registrations per user type (CATCUSTOMER) NCRt is the number of new car registrations at year t (par. 3.1) CUSTSHcust,t is the share of each user type (par. 4) NCRCCFINcar,cust,t is the number of new registrations per user type and technology NCRTFINcar,t is the total number of new registrations per technology


35

4.2 Graphical example of Wood algorithm To illustrate the Wood algorithm’s application to this particular allocation problem, a graphical and numerical example would be useful. If there are no capacity constraints, the potential allocation for each technology can be represented with a rectangle whose base length is equal to the parameter WIDTHcust and whose area corresponds to the total capacity for this particular technology (i.e. 1). The area of this rectangle would correspond to 100% of capacity. For an overlap factor 2λ = , WIDTHcust is the attractiveness of the best alternative multiplied by two. On the horizontal axis (attractiveness), the midpoint of each rectangle corresponds to the attractiveness of the respective technology DEMSHcar,cust. Its boundaries lie at DEMSHcar,cust -/+ 0.5*WIDTHcust. For those technologies that do exhibit capacity constraints, the shape of the rectangle is determined in a similar way. The crucial point in the design of the allocation procedure is the estimate of this limited capacity with respect to those technologies able to fulfil the entire demand. Let us illustrate the complete procedure in the following example, one of them formulated assuming that no technology constraints exist for any technology, and another one introducing a capacity constraint in a given technology. Let us assume firstly that no capacity constraints exist. Since there are no capacity constraints, the height of each rectangular is equal to 1 divided by WIDTHcust (that, in this example, is 0.74 for all technologies, twice the maximum value of DEMSHcar,cust), i.e. is equal to 1.35. x=DEMSHcar,cus

(midpoint of rectangle’s base)

x=DEMSHcar,cus +WIDTHcust/2

(right boundary of rectangle)

Allocation with no capacity constraints.

Allocation limit x= 0.415

Allocation with reduced capacity for HYBRID to

5% of total demand.

Allocation limit at x=0.399

LPETROL 0.08 0.45 0.047 0.069 HPETROL 0.36 0.73 0.425 0.447 LDIESEL 0.03 0.4 0.001 HDIESEL 0.37 0.74 0.438 0.460 ELECTRIC 0.01 0.38 FUELCELL 0.04 0.41 0.015 HYBRID 0.11 0.48 0.088 0.006


36

The algorithm works from right to left over the priority axis x. The capacity allocated to each alternative is summed up until the total of 100% of demand is reached. This means that for a certain point on the x axis, the sum of the areas to the right of that point would sum up to 1. E.g. for x= 0.50, the only technologies that are considered are HDIESEL and HPETROL. The capacity that is allocated to them at that point equals 1.35* (0.74-0.50)= 0.324 for the former and 1.35* (0.73-0.50)= 0.311 for the latter, giving an overall of 63.5% of total demand. In the case with no capacity constraints, i.e with all capacity rectangles exhibiting the same height, the line should move to the left until the whole demand is met. This happens at x= 0.415 and the demand is covered by 4 technologies. The relative low competitive position of LPETROL and LDIESEL has been calibrated because the example corresponds to the customer category SALESMAN. The share obtained by HYBRID in this case is 0.088, which definitely too high. This is basically due to the fact that no capacity constraints have been imposed to this technology. In the case of reduced capacity for HYBRID, the area of the technology’s rectangle is reduced and –a as result- its height. So, if the capacity for HYBRID were reduced to 5% of the network, the height of its rectangle would be 0.068. The demand would be met at x= 0.399 and HYBRID would still get a certain share of the total demand (0.6%). In this case, FUELCELL would also get a small share of the demand (1.5%), greater that the one corresponding to HYBRID because no constraints are assumed for FUELCELL. This example merely aims at showing how the algorithm operates requiring to deliver meaningful results a careful calibration of the parameters WIDTHcust and CAPcar,cust 4.3 Long-term user costs Long-term user costs are an important variable in the process of allocation of new registrations to the different technologies. The general principle of the methodology is to use an indicator for the fixed and variable costs that each user group would have for each technology they can choose from. There are several steps involved in the procedure, starting with the estimation of the purchase costs and fuel economy for a standard vehicle.

0.750.500.25

HDIESEL

HPETROL

HYBRID

FUELCELL

LDIESEL

LPETROL

ELECTRIC

x, DEMSHcar,cus


37

The average car price (AVGCPR) corresponds to the value of the reference car for each technology. This is calculated by taking actual market prices, historical data, and input from specific studies that provide projections for future prices (see Annex for additional information), and estimating the development of the average purchase cost per kW for each technology, separating into the propulsion specific cost and the common parts cost.

1

1

, , , ,

,, ,

, , , 1

(1 )(1 )

((1 ) 1)

t

t

car t car t car t car t

AVSCRAGEcar t

car cust t AVSCRAGEcar cust seminew t

AVGCPR CPCST AVENGPWR PSPCST

AVGCPR r rFCCST m

KMCAR r

−

−−

= + ⋅

⋅ ⋅ += ⋅ +

⋅ + −

where: AVGCPRcar,t is the average car price for each technology CPCSTcar is the average price of the common parts of the car (i.e. not including the engine) AVENGPWRcat,t is the average engine power for each technology PSPCSTcar,t is the propulsion specific cost for each technology (expressed in €/kW) FCCSTcar,cust,t is the discounted fixed cost of the car per km r is the discount rate AVGSCRAGEt-1 intervenes as the implied economic life of vehicles (equal to the average age of removal from the car stock in the previous year) m is the maintenance cost (expressed as a percentage of the discounted fixed cost)

, , , 1car cust seminew tKMCAR − captures the expected mileage of the car type for the customer category. The selected value should be equal to the age group-averaged mileage for the same combination of technology and user type in the previous year, but for sake of simplicity the value corresponding to semi-new cars is used. The discounted fixed cost of the vehicle varies by country, technology and user. A discount rate of 8% is currently used in the model and the economic life of the vehicle is considered to be equal to the implied economic life from the ageing dynamics results. The expected mileage that converts the cost in €/km also depends on the type of user and is considered to be equal to the average mileage of a semi-new car (5-10 years old) of the same technology for that user group in the previous year. The calculation of the variable costs is based on fuel price projections derived from the POLES model and from specific studies carried out by IPTS concerning the potential development of fuel efficiency for each technology. In the general case, the variable costs are:

, , , , ,car cust t fuel t fuel car cust tFUCST CPS TOE PKFC= ⋅ ⋅ where: FUCSTcar,cust,t is the variable cost per km CPSfuel,t is the price of the fuel (per toe, in constant prices) TOEfuel is the conversion factor for the energy content of each fuel PKFCcar,cust,t is the fuel economy, expressed in litres (of the specific fuel) per 100 km In the case of the internal combustion engine vehicles, the fuel costs are the product of the fuel price (in €/toe) by the fuel consumption (in l/100 km) and the appropriate energy conversion factor. For EV, the fuel cost is computed as the product of the electricity price by the weight (in kg), the specific energy (in kWh/kg) and the charge frequency (number of charges/100 km) of the battery system, using also the conversion factor from toe to kWh.


38

The long-term car utilization costs (in €/km) are calculated with the previous values, the fixed costs, the fuel costs FUCSTcar (expressed in €/100km), the expected annual mileage EXPMILcust and the replacement costs in the case of EV and FCV:

, , , , , ,

, , , , , ,

car cust t car cust t car cust t

car cust t car cust t car cust t

CUCST FCCST FUCSTADCOST CUCST USERFREF

= +

= ⋅

The model allows for indirect costs, such as user preferences, to be taken into account, through the introduction of the USERPREFcar,cust,t parameter. For example, if using large diesel vehicles in urban areas entails indirect costs for the user (congestion, parking, etc.), USERPREFHDIESEL,URBAN,t would have a value larger than 1. 4.4 Dieselization of the fleet and the calibration of consumer’s choice. Since the introduction of alternative vehicle and fuel technologies in Europe has been minimal up to now, the calibration of the model dynamics and technology substitution patterns has been carried out on the basis of the dynamics of the dieselization of the fleet in Europe. The competition between gasoline and diesel technologies is a good example of how the various factors captured by the model can influence the penetration of each technology. The main rationale behind the model is that fuel prices, fuel efficiency and car costs influence the decision of each user type in each country in a different but predictable way. Capturing the way that these variables have influenced the balance between gasoline and diesel in the past can therefore provide a good proxy for the behaviour and the dynamics of the car market if more alternative technologies are introduced. European drivers have progressively decided to purchase more and more Diesel cars over the last 20 years. This technological shift is replicated in the model and the parameters characterising this customer decision making are calibrated according to it, assuming that the same driving factors will hold for the future market penetration of other alternative technologies. As shown in Figure 4-2, diesel demand in Europe has increased dramatically with the advent of direct injection turbo diesels with improved fuel economy and performance comparable to that of gasoline based alternatives.

Figure 4-2: Light duty diesel vehicle vs. gasoline vehicle sales in the EU

source: Martec, 2002

Figure 4-3 shows the 1995 passenger car fleet broken down by fuel type and engine capacity, for each EU 15 Member State. It is clear that the great majority of cars have gasoline engines


39

smaller than 2.0 l. Diesel cars were around 15% on average in 1995, while LPG vehicles have a significant presence only in Italy and the Netherlands. The distribution of the vehicles within the various emission categories is closely related to their age (since the various emission standards were introduced on a fixed time scale in most Member States). The average age of passenger cars is between 7 and 8 years, but there are again variations from country to country: the oldest cars are in Finland where the average age is about 11 years, while the youngest fleet is in Luxembourg, with an average age of about 4 years [Hickman, 1999].

Figure 4-3: Passenger car fleet distribution (1995 data) for EU 15. [Hickman, 1999]

Table 4-1: Sales of diesel cars in western Europe 1990-2001

Diesel (%) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001Austria 25,7 22,1 26,2 31,6 39,9 42,8 49,4 53,3 54,5 57,4 61,9 65,4 Belgium 32,7 30,8 31,8 36,9 42,4 46,8 45,7 49,8 52,2 54,3 56,3 62,6 Denmark 4,1 2,6 2,4 2,2 2,7 2,9 2,9 3 4,7 9,4 13,2 17,9 Finland 5,2 4,9 5 7,8 5,6 6,7 13,5 14,6 15,3 15,7 France 33 38,4 39 45,5 47,6 46,5 39,2 41,8 40,2 44,1 49 56,2 Germany 9,8 11,8 14,8 14,6 16,6 14,5 15 14,9 17,6 22,4 30,3 34,5 Greece 1,1 0,7 0,7 0,7 Ireland 13,6 15,5 14,2 15,6 16,7 15,9 13,3 11,3 12,8 10,9 10,1 12,9 Italy 7,3 5,7 7,6 8,4 9,1 9,9 16,5 17,5 22,5 29,4 33,6 36 Luxembourg 21,3 16,7 25,1 27,6 28,1 28,5 32,4 35,2 38,8 42 50,4 58,2 Netherlands 10,9 11 11,6 11 12 13,9 15,3 17,1 20,3 22,8 22,5 22,9 Portugal 4,9 7 7,8 11,3 11,6 10,7 12,6 16,9 18,8 20,9 24,2 27 Spain 14,6 12,8 16,6 23,2 27,5 33,6 37,5 42,2 47,8 50,6 53,1 52,5 Sweden 0,6 0,9 0,8 3,1 3,2 2,7 5,2 7,6 11 7,2 6,3 5,6 United Kingdom

6,4 8,7 12,5 19 21,7 20,2 17,8 16,1 15,3 13,8 14,1 17,8


40

European Union

23,1 22,6 22,9 22,8 25,3 29 32,8 36,4

Iceland 15,1 16,4 17,1 13,3 Norway 2,6 6,4 10,8 14,5 9,7 6 7,2 6,2 6,7 8,2 9 13,3 Switzerland 3 2,7 3,1 3,5 4,7 4,2 4,7 5,1 5,9 6,8 9,3 13,3 EFTA 5,9 4,7 5,5 5,4 6,4 7,4 9,5 13,3 West. Europe 13,8 14,6 17 19,8 22,6 22 22,3 22,3 24,8 28,4 32,1 35,8 Source : AAA (Association Auxiliaire de l’Automobile)

Table 4-2: Fuel consumption of diesel and gasoline models of same engine size

Mercedes Benz 190 (Model 1985) VW Jetta (Model 1990)

Fuel Diesel Gasoline Difference Diesel Gasoline Difference L/100 Km (Urban) 7,8 12,4 4,6 6,4 9,4 3 L/100 Km (Non-urban) 7,1 10,7 3,6 5,5 7,4 1,9 L/100 Km (Mixed) 7,6 11,2 3,6 5,9 8,7 2,8 VW Golf (Model 2002) Peugeot 206 (Model 1997) Fuel Diesel Gasoline Difference Diesel Gasoline Difference L/100 Km (Urban) 6,9 10,2 3,3 6,6 10,7 4,1 L/100 Km (Non-urban) 5,2 8,1 2,9 4,1 6,1 2 L/100 Km (Mixed) 6,2 9,4 3,2 5 7,7 2,7 Audi A4 2.4 Pack + Mercedes Classe S S320 BA5 Fuel Diesel Gasoline Difference Diesel Gasoline Difference L/100 Km (Urban) 12,5 14 1,5 11,4 17,3 5,9 L/100 Km (Non-urban) 6,3 6,8 0,5 6 8,2 2,2 L/100 Km (Mixed) 8,6 9,4 0,8 8 11,5 3,5 Fuel Energy savings Difference (%) L/100 Km (Urban) 3,73 30% L/100 Km (Non-urban) 2,18 28% L/100 Km (Mixed) 2,77 29% The relative price of diesel oil is one of the cost factors that explain the growing popularity of diesel cars. The price of diesel oil at the gasoline station has always been lower than gasoline in most western European countries. Despite rapid increases in the taxation of diesel oil it still remains cheaper than gasoline. The sample of selected cars gives an average of nearly 30 % in fuel savings for the diesel engine. These economies in fuel consumption added to a lower price for the fuel represent strong incentives to buy cars with diesel engines. But, the fuel economy of diesel cars with engines of the same engine size as a gasoline car is counterbalanced by two factors. First, the diesel car has lower performance in speed and acceleration. Second, the diesel car is more expensive. In terms of performance, a diesel car still gives less power for a given engine size than a gasoline car. To compensate for this, a bigger engine needs to be fitted to the diesel car. In recent years car manufacturers have developed more powerful diesel cars like the Volkswagen 1.9 TDI with 110 kW for an engine of 1.9 litres. The comparable gasoline engine is 1.8 litres with turbo.


41

The reason why a diesel car is more expensive than a gasoline car can be found in the fact that the engine is much heavier and more expensive because of a much higher compression ratio. Another reason is that diesel cars were marketed later (the first diesel engine car the Mercedes-Benz 260 D was produced in 1936 and was sold roughly 2000 units before the Second World War) and are sold in smaller quantities than gasoline cars. The price differences seem to persist over the last 20 years. The Volkswagen 1.9 TDI with 110 kW costs 24,200 € and the Volkswagen 1.8 GTI costs 21,900 € in Germany in 2002. Mercedes-Benz often charge the same price for a diesel car with the same performance as a gasoline car despite the fact that the diesel car has a bigger engine. In 2002 Mercedes-Benz E-Class 240 gasoline car with 178 bhp costs 28,040 GBP in Great Britain and Mercedes-Benz E-Class 270 CDI diesel car with 177 bhp costs 27,435 GBP. In Germany is the diesel car 700 € more expensive than the gasoline car.

Table 4-3: Price of cars with diesel engines and gasoline engines

Year Model Price of gasoline car

Price of diesel car

Difference in price (%)

1980 VW Rabbit 4-door Hatchback 5 890 6 415 8,9%

1980 Audi 5000 Sedan 10 600 11 400 7,5% 1981 Isuzu I-Mark 6 069 7 194 18,5% 1981 Peugeot 505 10 990 11 990 9,1% 1982 VW Jetta 4-door Sedan 8 595 9 240 7,5% 1982 Volvo 4-door Sedan 10 885 13 180 13,9% 2002 Audi A4 35 770 37 860 5,5% 2002 Mercedes 320 71 400 66 600 -7,2% 2002 Peugeot 206 (2.0 l) 16 800 18 050 6,9%

Figure 4-4: Evolution of the emission standards for diesel and gasoline engines in Europe, [Homeister, 2001]

Figure 4-5: Comparison of the American, European and Japanese emission standards [Homeister, 2001]


42


43

5 TRANSPORT DEMAND This section focuses on the modelling of the use of the transportation equipment. Assuming a given structure of the fleet (whose dynamics was described in the preceding section), the short-run module replicates the use of this equipment by the customers, depending on their characteristics and their economic restrictions (what cars are used and how much). The fuel taxation policy in Europe has resulted in a steady decline in fuel use per vehicle • High operating costs have successfully encouraged consumers to purchase desired

vehicles/technology and to adapt use patterns consistent with the stated policy objectives. • The policy also meets the EU's objectives for CO2 emissions by controlling fuel

consumption.

Figure 5-1: Annual mileage as a function of the passenger car age (1990 data) [Hickman, 1999]

Figure 5-2: In-use fleet vehicle miles travelled by year in US vs EU [Martec, 2002]


44

After declining modestly between 1975-1980 , US miles travelled per vehicle light duty) per year have increased steadily at +1.3% for US light vehicle fleet. European VMT has declined at about a - 0.44% rate over the same period as EU governments have increased taxes to maintain high vehicle operating costs.

Figure 5-3: Fuel consumption per vehicle by year in US vs EU [Martec, 2002]

The US consumption per vehicle posted its largest decline during the 1975-1980 period (Shah of Iran falls) as CAFE was being phased in. • Gains in US fuel use per vehicle stagnated from 1985-1995. • Fuel use per vehicle has been climbing since 1995, reflecting consumer demand for light

trucks and increased VMT/vehicle. • CAFE has reduced relative vehicle operating costs, encouraging consumers to burn more

fuel (recognizing CAFE freeze since 1995). In the model, the change in transport demand is estimated on the basis of average car use (kms per year). The change is assumed to be the result of the combination of dynamic factors (fuel prices FPRzon,cat, income per capita GDPPORzon and a constant trend (TRENDzon,cust) per user type per country to explain the rest of the change.

( ), , , , , , , , 1 , , ,1zon age car cust t zon age car cust t zon car cust zon custKMCAR KMCAR DYNCF TREND−= ⋅ + +

( ), , , 1

, ,, , 1 , 100

zon car t zon t zonzon car cust cust zon

zon car t zon t

FPR GDPPOP LTINELDYNCF FPREL GDPGRWFPR GDPPOP

−

−

= ⋅ ⋅ + ⋅ where: KMCARzon,age,car,cust,t is the annual usage of each car depending on the country, age, technology and user type (km/year) DYNCFzon,car,cust is the change due to the dynamic factors captured by the model (fuel price and GDP changes) TRENDzon,cust is the constant trend not explained by the dynamic factors


45

FPRzon,car,t is the fuel price for year t for the specific technology GDPPOPzon,t is the GDP per capita

( ), , , 1

, , 1 ,

zon car t zon t

zon car t zon t

FPR GDPPOPFPR GDPPOP

−

−

⋅

captures the change in fuel price in relation to the change in

GDP per capita FPRELcust is the price elasticity of transport demand for each user type GDPGRWzon is the average growth in GDP per capita LTINELzon is the long term income elasticity of transport demand The long term income elasticity of transport demand LTINELzon is modelled with a Gompertz function, in order to capture the saturation effect of transport demand (the rate of increase of transport demand drops as income increases):

( )21 2 zon zonPAR GDPPOP

zon zon zonLTINEL PAR PAR GDPPOP e ⋅= ⋅ ⋅ ⋅ where: GDPPOPzon is GDP per capita PAR1 and PAR2 are the parameters of the Gompertz function

Figure 5-4: Relation between engine type/size and the annual mileage (*1000) of passenger cars in EU 15 (1995 data) [Hickman, 1999]

To obtain typical emission figures, the number of vehicles should be combined with the annual mileage of each vehicle type, the average mileage distribution between urban, rural and highway traffic and representative speeds in each of these traffic types. Table 5-1 shows typical values for the EU15. These data have been used to adjust and calibrate the parameters in the demand equation for KMCAR zon,age,car,cust,t.


46

Table 5-1: Annual mileage, mileage distribution and representative speeds for passenger cars in the EU15 (Reference year 1995) [Hickman, 1999]

Mileage distrib. (%) Repres. Vehicle speed (km/h)

Fuel Size Emission class Annual Mileage (km) Urban Rural HWay Urban Rural HWay

Gasoline < 1,4 l Pre ECE 7495 44 42 14 21 59 91 ECE 15-00/01 6977 43 42 15 22 59 94 ECE 15-02 7060 42 43 15 24 65 98 ECE 15-03 8501 41 43 16 26 71 103 ECE 15-04 9543 38 46 16 25 70 103

Improved conventional 8709 36 38 26 34 71 105

Open loop 9011 37 39 24 36 74 105 EURO 1 11810 37 44 19 28 71 103 EURO 2 11810 37 44 19 28 71 103 EURO 3 11810 37 44 19 28 71 103 EURO 4 11810 37 44 19 28 71 103

1,4 - 2,0 l Pre ECE 7005 36 47 17 22 61 95 ECE 15-00/01 9328 31 48 21 24 60 99 ECE 15-02 10058 31 47 22 26 65 105 ECE 15-03 10030 35 44 21 29 70 112 ECE 15-04 11747 37 45 18 27 70 109

Improved conventional 10848 37 38 25 36 74 123

Open loop 10787 36 39 25 35 72 120 EURO 1 13934 37 42 21 31 72 115 EURO 2 13934 37 42 21 31 72 115 EURO 3 13934 37 42 21 31 72 115 EURO 4 13934 37 42 21 31 72 115

> 2,0 l Pre ECE 8319 34 47 19 23 61 97 ECE 15-00/01 12550 29 49 22 25 57 104 ECE 15-02 12473 31 48 21 27 63 109 ECE 15-03 12366 36 43 21 31 70 117 ECE 15-04 13727 36 44 20 28 68 113 EURO 1 17401 37 40 23 33 73 120 EURO 2 17401 37 40 23 33 73 120 EURO 3 17401 37 40 23 33 73 120 EURO 4 17401 37 40 23 33 73 120

Diesel < 2,0 l Uncontrolled 14214 38 44 18 27 69 109 EURO 1 17619 39 43 18 29 69 112 EURO 2 17619 39 43 18 29 69 112 EURO 3 17619 39 43 18 29 69 112 EURO 4 17619 39 43 18 29 69 112

> 2,0 l Uncontrolled 14873 36 43 21 27 70 112 EURO 1 18259 38 42 20 28 69 113 EURO 2 18259 38 42 20 28 69 113 EURO 3 18259 38 42 20 28 69 113 EURO 4 18259 38 42 20 28 69 113

LPG All Uncontrolled 20046 44 37 19 21 64 104 EURO 1 20696 41 38 21 22 63 104 EURO 2 20696 41 38 21 22 63 104 EURO 3 20696 41 38 21 22 63 104 EURO 4 20696 41 38 21 22 63 104

2-stroke All Uncontrolled 5150 30 60 10 30 80 100


47

6 FUEL CONSUMPTION AND EMISSIONS The simulation of the fuel consumption starts with the calculation of the average annual kilometres per car and customer for each age group. 6.1 Fuel consumption of diesel and petrol cars The annual fuel consumption per age, car and customer category is:

, , , , ,car age cust car age car age custAFC PKFC KMCAR= ⋅ where PKFCcar,age is the average specific fuel consumption, depending on the age of the car KMCARcar,age,cust is the average annual mileage depending on customer category and age (chapter 5) The total consumption per age and car category is:

( ), , , ,car age cust car age cust car agecust

TFC CUSTSH AFC SUR= ⋅ ⋅∑

where SURcar,age is the number of cars of each cohort that is still in the car park (par. 3.2). The fuel consumption per car and age, PKFCcar,age, is a function of the historical values of the fuel consumption per 100 km and the number of cars of each category:

( )

( )

( )

4

, , , ,1

, ,, ,

9

, ,5

, ,, ,

14

, ,10

, ,, ,

, ,

car t car t car t i car t ii

car new tcar new t

car t i car t ii

car smn tcar smn t

car t i car t ii

car mat tcar mat t

car old t

NCRT PKFC SURPC PKFCPKFC

SUR

SURPC PKFCPKFC

SUR

SURPC PKFCPKFC

SUR

PKFC

− −=

− −=

− −=

⋅ + ⋅=

⋅=

⋅=

∑

∑

∑

2, ,

, ,

car mat t

car smn t

PKFCPKFC

=

With these variables it is possible to calculate the average consumption, the total annual consumption per car and the total shares of gasoline and diesel cars, and the shifts from light to heavy cars. The CO2 emissions are computed by technology, country and fuel from the energy consumption using conversion factors from fuels to emissions.


48

6.2 Trends in fuel economy In the EU, there are no national limits on fuel consumption of vehicles manufactured and sold in Europe. Fuel is instead taxed at relatively high levels (compared to the US) to encourage consumers to purchase and operate vehicles with acceptable fuel economy. As a consequence, fuel economy figures were seriously better than in the USA. The fuel consumption figures were obtained from the official certification tests. These results may be seriously different from real-world fuel consumption. Also the test cycles are different in the US (FTP) and the EU (ECE). However the average load in both cycles is comparable, so the tendencies and differences expressed in the following figures should be realistic.

Figure 6-1: US vs EU average fuel economy (per model year) [Martec, 2002]

The average rated fuel economy for all vehicles sold in the USA has been drifting lower since 1987 as consumers have shifted from cars to SUVs and other light trucks. The US model year fleet average CAFE peaked in 1987 at 26.2 MPG (cars and light trucks) and has fallen ~6.5% to 23.9 MPG in 2001 The EU average fuel economy climbed steadily from 1975-2000: 1.7% CAGR. The "pause" in EU fuel economy gains shown from 1985-1990 was due to the introduction of new emissions standards, which temporarily curbed fuel economy improvement In July 1998, the European Automobile Manufacturers Association (ACEA) made a voluntary agreement with the European Commission to reduce CO2 emissions from new cars by around 25% by 2008 (to an average of no more than 140g CO2/km. Of course, this is directly linked to fuel consumption.


49

Figure 6-2: ACEA agreement [ACEA, 2002]

100

120

140

160

180

200

1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015

Model year

CO2

(g/k

m)

-11%

-25%

-36%

6,0 l/100km gasoline5,3 l/100km diesel

5,1 l/100km gasoline4,5 l/100km diesel

ACEA Target2008

EU proposal2012

Over the period 1995 and 2001, new European gasoline cars and new diesel cars reduced their average fuel consumption from 7.9 l/100km to 7.3 l/100km and 6.6 l/100km to 5.8 l/100km, respectively. The average fuel consumption for gasoline and diesel fuelled cars combined fell from 7.6 l/100km (31mpg) to 6.7 l/100km (35mpg) [ACEA, 2002].

Table 6-1: Specific fuel consumption and CO2 emissions of the ACEA newly registered passenger cars in the EU and each EU member state [ACEA, 2002]


50

Figure 6-3: Average engine displacement of new sold vehicles [Martec, 2002]


51

7 ADDITIONAL INDICATORS The survival rate per period, based on the Gompertz survival model, is given by the following equation:

( )( )

1SRPB t iSRPA e

t iSRPP e⋅ −⋅

− = −

This is the complementary of the scrappage rate, which, according to the historical data, should follow an S-shaped curve, being SRPA and SRPB parameters. The annual change of SRPPt-1 is equal to the derivative:

( )( )( ) SRPB t iSRPB t i SRPA e

t it i

SRPA SRPB eSRDSRPP

⋅ −⋅ − + ⋅

−−

⋅ ⋅=

with these values it is possible to compute the number of scrapped cars per period:

( ), 1t i car t i t i t icar

SCRPP NCRT SRPP SRD− − − − −= ⋅ ⋅∑

the number of cars surviving each period and car category in the year t-i:

, , 1car t i car t i t iSURPC NCRT SRPP− − − −= ⋅ and by age category:

( )

( )

( )

4

, ,1

9

, ,5

14

, ,10

,

t new car t i t i t ii

t smn car t i t i t ii

t mat car t i t i t ii

t old

SCR SCRPP SR SRD

SCR SCRPP SR SRD

SCR SCRPP SR SRD

SCR OLDSCR

− − −=

− − −=

− − −=

= ⋅ ⋅

= ⋅ ⋅

= ⋅ ⋅

=

∑

∑

∑


52

8 MAIN RESULTS The main results of the IPTS transport technologies model for the member states of the European Union are summarised in the following tables. These results correspond to the business-as-usual case, i.e. the baseline projections for the input variables used by the model (GDP growth, fuel prices, car prices, trends in fuel economy, etc.). Although the model is considered sufficiently reliable, it should be noted that, since its objective is mainly the comparison of alternative policy measures, these projections should be used as the basis for comparisons or trend analysis, and not as a means for prediction of the future value of the variables. Car ownership levels in the EU-15 (Table 8-1) are expected to continue increasing but, especially after 2015, will probably reach saturation at values between 600 and 650 cars per 1000 inhabitants. As a result, and in combination with the demographic and technological trends, annual new car registrations are expected to stabilise around 15 million for EU-15 (Table 8-2). The number of cars removed from the stock (scrapped or exported as used cars outside the EU-15) is expected to rise to almost 14 million per year, as a result of the 10-15 year lag compared with the increase of car ownership in the 1990’s (Table 8-3). The total number of passenger cars in circulation in EU-15 will rise to almost 220 millions by year 2020, an increase of 23% compared to year 2000. It is interesting to note that in the two countries with the largest car stock, Germany and Italy, the number is expected to remain stable after 2015 (Table 8-4). As regards the penetration of new vehicle technologies, the model results suggest that only hybrid vehicles have the potential for a wide scale introduction by 2010 (Table 8-5 to Table 8-12). Electric vehicles show a limited potential, concentrated mainly in some niche markets (urban areas in countries with cheap electricity), while fuel cells may capture a significant part of the market only by the end of the 2010’s. Another trend that can be identified is that of the shift from gasoline to diesel. The expected improvements in diesel technology could provide significant cost savings and comparable performance with gasoline technology. A gradual replacement of gasoline cars with either diesel or hybrid (mainly gasoline-electric hybrids) is therefore expected in the medium term (2010-2015). In the longer term, conventional ICEs and hybrids may gradually lose their share to fuel cells (probably using hydrogen from reformed gasoline, natural gas or methanol), depending on the progress made in the development of fuel cell technologies. Average car use is expected to stabilise around 14000 kms per car per year by 2020. However, notable differences among member states can be seen, due to the different lifestyles, geography, urbanisation and urban sprawl levels, and differences in statistics. Most of the expected changes in the driving factors can lead to increases in the average distance driven, but a certain saturation level for each country is expected to be reached in the next 15-20 years (Table 8-13). As regards the total number of kms driven in each country, both the number of cars and the average distance are expected to rise and, as a result, total car use may increase by about 33% between 2000 and 2020 (Table 8-14). This projection implies that the overall increase in passenger transport demand that is expected in the next 20 years is most probably going to be covered by other modes (notably air transport for long distances), since car passenger transport will have reached saturation levels. This is highlighted in the projections for transport intensity that corresponds to passenger car transport. The ratio of kms driven to GDP is expected to continue rising until around 2005, but will tend to fall afterwards, since the growth in GDP will not be accompanied by a comparable growth in car passenger transport (Table 8-15).


53

Total fuel consumption and CO2 emissions are expected to follow a similar trend, reaching a maximum around year 2005 and starting to fall afterwards (Table 8-16 and Table 8-17). This is the result of the expected fuel economy of passenger cars improving faster than the rate of growth of total car use. An important part of the improvement of fuel economy is expected to come from the introduction of hybrids and, later on, fuel cells but -even without these alternatives- the evolution of gasoline and diesel ICEs according to the EURO standards and the ACEA agreement should be enough to prevent CO2 emissions from rising further. Table 8-18 and Table 8-19 show the expected average fuel consumption of new gasoline and diesel passenger cars. The average for Europe is expected to improve, but differences will still exist between member states due to the differences in user choices. The improvement of the average fuel economy for the whole car stock is expected to be even larger, since the majority of the cars that entered into circulation in the 1980’s and 1990’s will have been replaced in the next 10 years by much more fuel efficient cars (Table 8-20 and Table 8-21). The average age of cars in circulation is expected to rise slightly in the next 20 years, from 7.4 to 8.3 years (Table 8-22). This is mainly the result of demographics in Europe (age distribution of car owners) and the saturation in car ownership levels (total demand). The effects of improved car technology on the length of a car’s life (either technical or economic) seem to become marginal, and the average age of car scrapping (or removal from circulation in general) is stable (Table 8-23). However, in both average age indicators, significant differences exist among member states. These differences are mainly the result of the different socio-economic conditions, car costs, disposable income and (new and used) car market operation in each country.


54

Table 8-1: Car ownership levels (passenger cars per 1000 inhabitants)

2000 2005 2010 2015 2020 Austria 508 556 587 609 623 Belgium (incl. Lux.) 485 541 578 602 619 Denmark 393 473 529 568 595 Finland 428 502 552 585 607 France 486 540 576 601 617 Germany 532 569 595 613 625 Greece 306 409 483 536 573 Ireland 373 467 529 570 597 Italy 554 581 602 616 627 Netherlands 433 506 555 587 609 Portugal 330 394 449 496 535 Spain 420 470 514 551 580 Sweden 462 524 566 595 613 United Kingdom 429 501 549 583 605 EU-15 476 528 565 592 610

Table 8-2: Total number of new passenger car registrations (thousands)

2000 2005 2010 2015 2020 Austria 340 326 332 329 323 Belgium (incl. Lux.) 569 513 566 555 563 Denmark 190 179 197 184 186 Finland 212 147 169 167 156 France 2695 2414 2659 2518 2543 Germany 3557 3647 3597 3600 3526 Greece 328 293 289 291 251 Ireland 157 121 150 141 139 Italy 2026 1905 1893 1883 1804 Netherlands 737 646 740 711 729 Portugal 268 294 312 321 327 Spain 1034 994 1033 968 914 Sweden 380 278 324 313 304 United Kingdom 2801 2792 3049 3064 3105

EU-15 15.3M 14.5 M 15.3 M 15.0 M 14.9M


55

Table 8-3: Total number of passenger cars removed from the stock (thousands)


EU-15 10.5 M 11.1 M 12.9M 13.3 M 13.7 M

Table 8-4: Total number of passenger cars in circulation (millions)

2000 2005 2010 2015 2020 Austria 4.0 4.4 4.6 4.7 4.8 Belgium (incl. Lux.) 5.1 5.7 6.1 6.3 6.5 Denmark 2.1 2.5 2.8 3.0 3.1 Finland 2.2 2.5 2.7 2.9 2.9 France 28.1 31.1 33.3 34.6 35.4 Germany 44.2 46.6 47.8 48.3 48.3 Greece 2.9 4.0 4.7 5.3 5.6 Ireland 1.4 1.7 2.0 2.1 2.2 Italy 32.3 33.2 33.8 34.0 33.9 Netherlands 6.4 7.6 8.4 8.8 9.2 Portugal 3.3 3.9 4.3 4.7 5.0 Spain 17.1 18.8 20.4 21.6 22.4 Sweden 3.8 4.2 4.6 4.8 4.8 United Kingdom 24.4 28.7 31.7 33.6 34.9

EU-15 177.3 194.8 207.2 214.7 218.9


56

Table 8-5: New electric car sales (thousands)

Time (Year) 2000 2005 2010 2015 2020 Austria n/a 0 2 2 1 Belgium (incl. Lux.) n/a 1 13 28 44 Denmark n/a 0 0 1 1 Finland n/a 0 4 12 19 France n/a 2 32 69 109 Germany n/a 3 32 70 94 Greece n/a 0 1 2 3 Ireland n/a 0 1 2 3 Italy n/a 3 26 42 58 Netherlands n/a 1 18 38 73 Portugal n/a 0 1 1 1 Spain n/a 1 9 13 11 Sweden n/a 1 22 43 64 United Kingdom n/a 4 54 125 303

EU-15 n/a 16 215 446 784

Table 8-6: New fuel cell car sales (thousands)

2000 2005 2010 2015 2020 Austria n/a 0 0 9 53 Belgium (incl. Lux.) n/a 0 0 8 74 Denmark n/a 0 0 17 71 Finland n/a 0 0 2 25 France n/a 0 1 74 462 Germany n/a 0 1 46 293 Greece n/a 0 0 0 12 Ireland n/a 0 0 4 30 Italy n/a 1 4 11 32 Netherlands n/a 0 0 12 100 Portugal n/a 0 0 2 26 Spain n/a 1 3 13 62 Sweden n/a 0 0 5 43 United Kingdom n/a 1 3 20 326

EU-15 n/a 4 13 223 1611


57

Table 8-7: New large diesel car sales (thousands)


EU-15 1790 2320 2048 1739 1375

Table 8-8: New large gasoline car sales (thousands)


EU-15 4189 3084 2736 1874 1017


58

Table 8-9: New hybrid car sales (thousands)

2000 2005 2010 2015 2020 Austria n/a 4 24 50 69 Belgium (incl. Lux.) n/a 9 55 105 144 Denmark n/a 6 38 57 53 Finland n/a 4 24 39 42 France n/a 37 243 468 610 Germany n/a 44 256 536 795 Greece n/a 5 29 60 79 Ireland n/a 2 13 23 30 Italy n/a 38 207 377 552 Netherlands n/a 14 88 161 200 Portugal n/a 6 37 75 104 Spain n/a 15 89 161 237 Sweden n/a 7 45 74 84 United Kingdom n/a 67 404 732 947

EU-15 n/a 255 1552 2918 3945

Table 8-10: New light diesel car sales (thousands)


EU-15 2208 2903 3034 2729 2117


59

Table 8-11: New light gasoline car sales (thousands)


EU-15 7106 5969 5714 5119 4021

Table 8-12: Share of each technology in new car shares in EU-15 (%)

2000 2005 2010 2015 2020 Light gasoline 46 41 37 34 27 Large gasoline 27 21 18 12 7 Light diesel 14 20 20 18 14 Large diesel 12 16 13 12 9 Electric 0 0 1 3 5 Fuel cell 0 0 0 1 11 Hybrid 0 2 10 19 27

Table 8-13: Car use (average km per car per year)

Time (Year) 2000 2005 2010 2015 2020 Austria 11,910 12,498 12,549 12,584 12,502 Belgium (incl. Lux.) 13,501 14,052 13,992 13,859 13,608 Denmark 16,592 17,661 18,000 18,388 18,690 Finland 15,889 17,384 17,944 18,204 18,239 France 13,999 14,778 14,888 14,889 14,667 Germany 12,029 12,701 12,912 13,035 13,014 Greece 14,973 16,349 16,520 16,658 16,588 Ireland 13,487 14,584 14,378 14,404 14,184 Italy 10,320 10,655 10,595 10,596 10,553 Netherlands 14,573 15,390 15,461 15,461 15,212 Portugal 16,475 17,502 17,877 18,316 18,640 Spain 10,364 11,086 11,235 11,345 11,298 Sweden 14,668 15,751 15,879 16,004 15,970 United Kingdom 15,384 16,147 16,207 16,118 15,808

EU-15 12,760 13,549 13,719 13,802 13,713


60

Table 8-14: Total car use (total car* km per year, millions)


EU-15 2.262 M 2.639 M 2.843 M 2.963 M 3.002 M

Table 8-15: Transport intensity (total car*km per euro GDP, average EU 2000=100)

Time (Year) 2000 2005 2010 2015 2020 Austria 88.8 89.9 84.3 77.6 70.8 Belgium (incl. Lux.) 93.7 97.6 94.0 87.6 80.5 Denmark 93.8 108.7 112.4 109.7 104.7 Finland 105.0 116.1 116.6 110.8 102.6 France 100.8 105.2 101.3 94.7 86.7 Germany 103.0 103.9 97.7 90.1 82.3 Greece 105.0 139.2 148.9 148.9 141.7 Ireland 70.0 79.3 78.9 74.6 67.6 Italy 98.2 96.8 90.5 83.9 77.4 Netherlands 92.5 101.4 100.5 95.0 87.7 Portugal 124.5 141.5 147.5 148.6 146.0 Spain 91.8 95.6 93.0 88.6 82.9 Sweden 103.7 112.1 110.4 104.3 96.3 United Kingdom 104.4 117.2 118.3 113.3 105.2

EU-15 100.0 105.3 102.3 96.4 89.1


61

Table 8-16: Consumption of fossil fuel for transport (passenger cars, ktoe)


EU-15 177,713 189,173 181,798 170,258 154,511

Table 8-17: Total CO2 emissions for passenger cars (million tons)


EU-15 523.5 558.7 538.1 504.0 456.9


62

Table 8-18: Average fuel consumption of new gasoline passenger cars (l/100km)


EU-15 7.4 6.9 6.4 6.1 5.8

Table 8-19: Average fuel consumption of new diesel passenger cars (l/100km)


EU-15 6.3 5.7 5.2 4.9 4.6


63

Table 8-20: Average fuel consumption of all gasoline passenger cars in stock (l/100km)


EU-15 9.0 8.6 8.1 7.6 7.2

Table 8-21: Average fuel consumption of all diesel passenger cars in stock (l/100km)


EU-15 10.0 8.1 6.8 6.0 5.6


64

Table 8-22: Average age of passenger cars in circulation


EU-15 7.4 7.6 7.9 8.1 8.3

Table 8-23: Average age of removal of passenger cars from circulation


EU-15 11.4 11.3 11.3 11.4 11.4


65

9 SCENARIO-BUILDING CAPABILITIES The main driver for the development of the IPTS transport technologies model was the need for a tool that would help the analysis of policy measures that could lead to the reduction of fuel consumption and emissions through the acceleration of the introduction of alternative technologies. The main objective of the initial phase of development was therefore to construct a model that could explain past trends and provide a plausible outlook for the future, rather than predict the exact future values of the model variables. Although the model is considered as sufficiently accurate in its predictive capacity, its main function for the IPTS is the analysis of “what if?” scenarios describing alternative paths as regards future policy measures, technological development, socio-economic trends and other external factors that may –directly or indirectly- influence the dynamics of adoption of new transport technologies and, eventually, the impacts of transport on fuel consumption and emissions. The model currently allows four general types of scenarios to be analysed: Technological scenarios Technological scenarios can provide the outlook for the penetration of new technologies and their impacts on the indicators measured under different assumptions than the ones currently used. Technology development is expressed in the model in terms of fuel economy and car prices for a given level of performance. The values used in the baseline scenario are derived from specific studies on the potential of the various alternative technologies. Alternative scenarios can be constructed by using more optimistic or pessimistic development paths that would change the competitive position of each alternative and influence the speed of its adoption. Technology scenarios in the model are mainly constructed through changing the exogenous variables (input) of the model. Policy measures Policy measures are perhaps the most interesting type of scenario analysis, and the one that provides more flexibility. The policy measures that scenarios can cover include the following:

• fuel taxes: changing the level of taxation of some or all fuels influences the total demand for transport and the share of each car technology

• carbon taxes: imposing a tax that is based on the carbon content of the fuel used can favour technologies that produce less CO2 and accelerate the introduction of alternatives such as hybrids and fuel cells.

• subsidies: subsidising a specific technology changes its competitive position and increases its sales

• emission limits: imposing an emission limit (e.g. gCO2/km) favours alternative technologies and/or leads to smaller cars being used

• accelerated scrapping schemes: providing a financial stimulus to scrap cars can accelerate the renewal of the car stock and reduce total fuel consumption and emissions

• zero emission zones: imposing a zero emission limit in urban areas leads to the acceleration of the introduction of electric, hybrid and fuel cell vehicles

• combination of policies: e.g. using the carbon tax revenue to subsidise alternative technologies can significantly speed up their introduction

Policy measures in the model can be introduced by changing a number of parameters in the model that, in turn, affect the costs and prices calculated by the model. Socio-economic trends


66

User choices are influenced by the socio-economic profile and other country-specific issues. They can significantly affect the dynamics of the transport sector, transport demand and the potential of alternative technologies. User choices are modelled in the context of specific socio-economic trends, covering market segments, user types, the degree of urbanisation, the environmental awareness of users, etc. that are endogenous to the model. For example, possible scenarios of socio-economic trends can include different degrees of responsiveness to environmental pressure (user awareness), urban sprawl (urbanisation), changing household structures and demographics. Normally, such scenarios involve modifications of the equations describing the dynamics of the model and, as far as user choices are concerned, price elasticities for each specific user group. External factors The main external factors that the model can analyse in terms of scenarios are fuel prices and GDP growth. In its current form, the model uses as input the projections of fuel prices from the POLES model, and the same assumptions for GDP growth that the POLES model uses. Scenarios that can be carried out include the investigation of the impact that a higher or lower price of, and/or a faster or slower economic growth would have on the dynamics of the car market, total demand, fuel consumption and CO2 emissions.


67

10 FUTURE EXTENSIONS In its current form, the IPTS transport technology model can be considered as being in only its first stage of development, having reached a level where it describes with sufficient accuracy the dynamics of the passenger car market in EU-15 and 6 other major countries. A number of possible extensions and improvements have been already identified, and most of them are already underway: • Extension of geographic coverage:

• Addition of candidate countries for enlargement of EU (13 candidates plus Croatia)- to be carried out during year 2003

• Gradual extension to reach global coverage • Extension of technologies covered:

• Addition of natural gas (including modification of existing ICEs to natural gas)- to be carried out during year 2003

• Addition of bio-fuels- to be carried out during year 2003 • Improvement of detail of fuel cells and hybrid vehicles (covering additional sub-

groups)- to be carried out during year 2003 • Coverage of specific market sectors

• Coverage of fleets, and especially taxis • Detailed modelling of used car market- to be carried out during year 2003

• Integration of other sectors • Connection with freight transport model- to be carried out during year 2003 • Addition of 2 wheeled vehicles and SUVs • Extension of pollutants covered- to be carried out during year 2003


68

11 REFERENCES 1. ACEA (2002): European Automobile Manufacturers Association,

http://www.acea.be/ACEA/auto_data.html. 2. Adda, J. and R. Cooper (1997): "Balladurette and Juppette: a discrete analysis of

scrapping subsidies", Cambridge, MA, National Bureau of Economic Research. 3. Alberini, A., W. Harrington, et al. (1998): "Fleet turnover and old car scrap policies",

Resources for the Future. 4. APME (1999): “Plastics, a material of choice for the automotive industry - insight into

plastics consumption and recovery in Western Europe”, Association of Plastics Manufacturers in Europe, July 1999.

5. Bjorner, T. B. (1999): “Demand for car ownership and car use in Denmark: a micro econometric model”, International Journal of Transport Economics XXVI(3): 377- 397.

6. Bouwman, M.; Development of a dynamic supply model on passenger transportation, Interim Report IR-98-097/November, IIASA, Laxenburg, Austria.

7. CNW, CNW Marketing Research, www.cnwbyweb.com/. 8. CSO, Central Statistics Office Ireland, http://www.cso.ie/principalstats/pristat8.html. 9. Dargay, J. and D. Gately (1999): “Income's effect on car and vehicle ownership,

worldwide: 1960-2015”, Transportation Research Part A 33(2): 101-138. 10. Dargay, J. M., J.-L. Madre, et al. (2000): "Car Ownership Dynamics Seen Through the

Follow-up of Cohorts: A Comparison of France and the UK", TRB Annual Conference, Washington D.C.

11. Delucchi, M.A.; Lipman, T.E.; An analysis of the retail and lifecycle cost of battery-powered electric vehicles, Transportation Research Part D 6 (2001) 371-404.

12. DETR (2001): "Transport trends: 2001 Edition", Department of the Environment, Transport and the Regions, UK.

13. EPA (2001): “Light-Duty Automotive Technology and Fuel Economy Trends 1975 Through 2001”, United States Environmental Protection Agency, September 2001.

14. EPA (1995): “Notebook Project, Profile of the Motor Vehicle Assembly Industry”, U.S. Environmental Protection Agency, September 1995.

15. Greenspan, A. and D. Cohen (1996): "Motor vehicle stocks, scrappage, and sales", Board of Governors of the Federal Reserve System (U.S.A.).

16. Hackney, J.; Neufville, R.; Life cycle model of alternative fuel vehicles: emissions, energy and cost tradeoffs, Transportation Research Part A 35 (2001) 243-266.

17. Hickman, A.J. (1999): “Methodology for calculating transport emissions and energy consumption”, Deliverable 22 for the project MEET.

18. Homeister N.L. (2001): “Vehicle Emissions Standards Around the Globe”, Ford Motor Company, 27 June 2001.

19. Landwehr, M. and C. Marie-Lilliu (2002): "Transportation projections in OECD regions", Paris, International Energy Agency.

20. LAT (2002): Transport and Environment Database System, Detailed Report 1: Road Transport, Aristotle University of Thessaloniki.

21. Martec (2002): “Fuel Economy: a critical assessment of public policy in the US vs the EU”, Martec White Paper, April 2002.

22. Medlock, K. B. and R. Soligo (2002): “Automobile Ownership and Economic Development”, Journal of Transport Economics and Policy(forthcoming).

23. METLTM, Ministère de l'Équipement, des Transports, du Logement, du Tourisme et de la Mer, http://www.equipement.gouv.fr/statistiques/chiffres/transpor/transpo_.htm.

24. Schafer, A. and D. G. Victor (2000): “The future mobility of the world population”, Transportation Research Part A 34(3): 171-205.

25. SMMT, "Motor Industry Facts 2002", UK, The Society of Motors Manufacturers and Traderds Limited.

26. Transportation Energy Data Book (2001), Oak Ridge national Laboratory, USA.


69

27. UKCC (2000): "New cars: a report on the supply of new motor cars within the UK", Competition Commission, United Kingdom.


70

ANNEX I: MODEL DATABASE A database containing techno-economic parameters defining eight current and prospective motor car technologies has been compiled, with data taken as much as possible from published sources. However, these sources tend to focus on particular vehicle types and often do not cover the timeframe required in this study. Consequently some interpolation and extrapolation has been necessary to cover the timeframe and range of vehicles under consideration, and to maintain consistency across the database. Eight types of motor car were considered distinguished by their propulsion systems and the fuels they use (Table 1). The Gasoline Fuel Cell car was added to the original list of seven vehicle types included in the work programme at the request of IPTS. For each vehicle type two car sizes were examined namely a typical urban car and a typical long distance family car. One exception was the electric vehicle for which only urban car data were collected. This is because range limitations make it unlikely that this vehicle would be used as a long distance family car. The hybrid car was assumed to be based on a gasoline fuelled engine. Table 1 Vehicle types and time-steps covered

Car type Size 2000 2010 2020 2030 Gasoline ICE Urban

Family X X

X X

X X

X X

Diesel ICE Urban Family

X X

X X

X X

X X

Hybrid Urban Family

X N/A

X X

X X

X X

Electric Urban X X X X Natural gas ICE Urban

Family N/A X

X X

X X

X X

Fuel Cell (gasoline) Urban Family

X X

X X

X X

Fuel cell (hydrogen) Urban Family

X -

X X

X X

Fuel cell (methanol) Urban Family

X X

X X

X X

Data have been assembled for 10 year time-steps from 2000 to 2030 (Table 1). Data collection for the fuel cell powered vehicles was started from 2010 because these have not yet achieved commercial scale deployment. The database contains values for the following:

- Vehicle price - Propulsion system price per vehicle - Propulsion system price per unit of capacity (kW) - Maintenance cost per year - Fuel efficiency (GJ/km) - Carbon dioxide emission (kgC/km) - Estimated distribution of sales between vehicle types in 2000

The maintenance cost covers servicing and replacement parts for the vehicles, but does not include insurance. Fuel costs are not included because this depends on assumptions on future fuel prices and vehicle utilisation.


71

The location of each data set in the work-sheets making up the database is listed in Table 2. Alternative sets of fuel efficiency assumptions are given in work-sheets “Energy” and “Energy Alt”. Table 2 Location of data sets in the database

Worksheet Data Set 2000 2010 2020 2030

Vehicle price 2000-cost 2010-cost 2020-cost 2030-cost Propulsion system price per vehicle

2000-cost 2010-cost 2020-cost 2030-cost

Propulsion system price per unit of capacity (kW)


Maintenance cost per year


Fuel efficiency (GJ/km)

Energy Energy Energy Energy

Fuel efficiency (GJ/km)

Energy-Alt Energy-Alt Energy-Alt Energy-Alt

Carbon dioxide emission (kgC/km)

Energy Energy Energy Energy

Carbon dioxide emission (kgC/km)

Energy-Alt Energy-Alt Energy-Alt Energy-Alt

Distribution of sales

2000-Use N/A N/A N/A

Data have been presented for each EU Member State and for North America in euros (2000). I.1. Data Sources The motor manufacturing industry is highly cost competitive and companies do not release much detail of their technical performance and future price targets into the public domain. Consequently the compilation of data has to draw on a broad range of research and concept studies produced from government agencies as well as the manufacturers. The information sources used to support this database are presented in the Bibliography. The sources of information generally cover one vehicle/fuel combination and a limited time frame. Therefore to build up a complete database requires interpolation and extrapolation from and between these sources. To achieve internal consistency across the database checks and comparisons were made between technologies and with current vehicle prices. EU database price assumptions Vehicle Prices: The database covers the prices of vehicles rather than their production costs. This is because most available data consider prices and it is not known what the manufacturers’ profit margins are, or how these are allocated between vehicle types and national markets. Also new vehicle types such as hybrid vehicles may be priced initially below the cost of manufacture in order to win market share. Furthermore it is prices that influences the choices


72

and decisions of end-users. The prices presented excluding purchase tax and VAT because this could well change over the timeframe of the database. Definition of Vehicle Types: Definitions of urban cars and long distance family cars were needed to guide the collection of price and fuel efficiency data. This was done by selecting a “basket” of current production models to represent each type. For the European Union the two types were characterised by the following models:

Urban Car Opel Corsa Ford Fiesta Renault Cleo Peugeot 106 VW Polo Long Distance Family Car BMW318i Audi A4 Ford Mondeo Opel Vectra VW Passat

The prices of gasoline and diesel cars in the EU vary between Member States therefore separate prices were gathered in the database for each country. Prices for gasoline vehicles were taken from DG Competition’s database of year 2000 car prices (which exclude national purchase tax and VAT). Prices were obtained for each car in the urban car and long distance family car “baskets”, and average values calculated and used for the database. Prices for diesel vehicles were calculated from these as follows. The prices of 2002 gasoline and diesel equivalents were collected from the official Internet sites of the relevant car manufacturer for the UK, Germany, Netherlands and Italy, and for each model in the car “baskets”. The average ratio of prices across the four countries was calculated for each model and used in combination with the DG Competition gasoline car price data to calculate a diesel pre-tax price equivalent for each model in 2000. The urban car and long distance family car “baskets” were then, as before, used to calculate average values for the database. The other types of car commercially available in year 2000 were hybrid and electric cars (urban only) and natural gas powered cars (large family only) (Table 1). Data on the prices of these were not available for all Member states. Therefore estimates were made as follows:

• For hybrid cars the UK price of the Honda Insight was used. Prices in other Member States were estimated assuming that this model would have the same price differential between countries as other Honda models.

• For electric cars the French price of the Peugeot 106 was used. Prices in other

Member States were estimated assuming that this model would have the same price differential between countries was the gasoline powered Peugeot 106.

• For natural gas powered cars the price was based on the purchase price differential

compared to an equivalent gasoline car. A standard differential of 3365 €(based on an UK differential of £2400 for a large familiy car, from the UK Powershift website) was applied across all Member States.

Data on the balance of propulsion system and non-propulsion system costs for gasoline, electric and natural gas vehicles were available from existing UK data. This information was


73

used to calculate the price breakdown between propulsion system and non-propulsion system elements for the gasoline, electric and natural gas cars included in the database. It was assumed that this breakdown was the same for all Member States. The propulsion system prices for diesel and hybrid vehicles were estimated assuming that the non-propulsion system prices were the same as for gasoline. Consequently propulsion system prices were obtained by subtracting non-propulsion system prices from the full price of the vehicles. Data were obtained for the maintenance charges of each of the vehicles in the urban car and family car “baskets” for the UK. Maintenance is a standardised process that on average should take the same amount of labour, equipment and spare parts in all Member States. However, the price will vary reflecting in particular the different labour charges. Therefore the price of maintenance in each Member State was estimated from the UK values using OECD Purchasing Power Parities (PPPs) for the year 2000. The maintenance charge for the hybrid urban car was assumed to be the same as for the equivalent gasoline car. Maintenance charges for the natural gas cars were scaled from the gasoline car maintenance charges by the ratio of the price of the vehicle to that of the equivalent gasoline car. Maintenance charges in each Member State for the electric cars were calculated from a UK value of £250 (includes an average of £50 servicing + £200 battery replacement per annum) using OECD PPPs as for gasoline and diesel vehicles. As regards the future prices of Gasoline and Diesel Vehicles, three assumptions underpin the data: • In line with published data, prices of the vehicles are assumed to stay level over the

period to 2030. This in part reflects the historic trend for manufacturers to absorb development costs through economies of production.

• The current price differentials between EU Member States is not sustained, and prices converge linearly such that they are level across the Union by 2030.

• Prices converge on current prices in Germany the largest EU market. Of course this does not mean that taxation of motor vehicles will also converge between Member States. This is assumed to be a scenario assumption in any modelling work in which the data is used. Price estimates for hybrid, electric and fuel cell cars have been collected from a range of sources for a UK study of low carbon options. These sources give either estimates of prices or price differentials with conventional vehicles for differing timeframes. The estimates assume economies from mass production, and are generic with no differentials between countries. These data have been used to estimate the prices given in the database by adding the price differentials or additional prices to the prices of gasoline cars. By doing this internal consistency has been maintained across the database. Price differences between EU Member States was maintained, but, in line with the data for gasoline and diesel cars, price convergence was assumed to be attained across the Union by 2030. Future propulsion system prices and maintenance charges were assumed to stay as a fixed proportion of the total vehicle price. No data have been given for fuel consumption costs because this would involve estimating future fuel prices and vehicle utilisation, both of which are scenario variables to be set for specific energy modelling studies. I. 2. EU database fuel efficiency assumptions


74

Fuel efficiencies for gasoline and diesel cars were taken from published information based on the “Combined Test Cycle”. The values reported are the averages for each “basket” of cars. Fuel efficiencies for hybrid, electric and natural gas vehicles are also reported for 2000. The hybrid data is based on the Honda Insight and the electric car data on the Peugeot 106. Data for the natural gas fuelled car was calculated using UK CO2 emission data for the 2002 Volvo S60 Bi-Fuel CNG vehicle, for CNG running relative to gasoline running emissions. The differential was used to scale data for the gasoline ‘Family Car’ to give a CNG fuelled vehicle equivalent. Future fuel efficiencies for gasoline and diesel vehicles were based on the targets for 2010 adopted by European, Japanese and Korean manufacturers under the so called ACEA voluntary agreement. Manufacturers have indicated that these targets stretch conventional propulsion systems to the limit for efficiency, therefore no further improvement should be anticipated beyond 2010. This position is represented by the data sheet “Energy”. However, some information in the literature suggests further improvement in efficiency beyond 2010. This position is represented in the data sheet “Energy Alt”. Fuel efficiencies for other vehicles were gathered from a range of sources. Much of this information expressed efficiencies as improvements relative to gasoline vehicles. Therefore the efficiencies reported have been calculated relative to the gasoline vehicle data. I. 3. North American database price assumptions Once again definitions of urban cars and long distance family cars were needed to guide the collection of price and fuel efficiency data. This was done by selecting a “basket” of current production models to represent each type. For North America the two types were characterised by the following models:

Urban Car Honda Civic Ford Escort Toyota Corolla VW Golf Dodge Neon Long Distance Family Car Ford F Series Pickup Chevrolet Silverado Pickup Ford Explorer Dodge Ram Pickup Ford Ranger Pickup

The latter selection reflects the high popularity and volume of sales of four-wheel drive light trucks (sports utility vehicles, pickups, etc) in North America. In line with North American preferences only gasoline fuelled cars were considered. The prices of gasoline cars were taken from a published price list, with no allowance made for possible inter-state variations. The prices reported are averages for the “basket” of vehicles selected and exclude purchase tax. The other types of car commercially available in year 2000 were hybrid and electric cars (urban only) and natural gas powered cars (large family only). Data on the prices of these were derived from the following:


75

• For hybrid cars the USA price of the Honda Insight was used. • For electric cars the North American price was calculated from the price of the UK

electric urban car, scaling it up relative to the price differential between the UK and US gasoline urban cars.

• For natural gas powered cars the North American price was calculated from the price

of the UK natural gas family car, scaling it up relative to the price differential between the UK and US gasoline family cars.

Data on the balance of propulsion system and non-propulsion system costs for gasoline, electric and natural gas vehicles were available from existing UK data. This information was used to calculate the price breakdown between propulsion system and non-propulsion system elements for the gasoline, electric and natural gas cars included in the database. The propulsion system prices for hybrid vehicles were estimated assuming that the non-propulsion system prices were the same as for gasoline. Consequently propulsion system prices were obtained by subtracting non-propulsion system prices from the full price of the vehicles. Data were obtained for the maintenance charges of each of the vehicles in the urban car and family car “baskets” for the USA. Average values for each of the “baskets” of cars are reported. The maintenance charge for the hybrid urban car was taken from reported values for the Honda Insight in the USA. Maintenance charges for the natural gas cars were scaled from the gasoline car maintenance charges by the ratio of the price of the vehicle to that of the equivalent gasoline car. Maintenance charges for US electric cars were calculated from the gasoline urban car equivalent scaled up by the difference in cost between UK electric vehicle maintenance relative to that of UK gasoline urban cars. As for the EU database no data have been given for fuel consumption costs. This is because it would involve estimating future fuel prices and vehicle utilisation, both of which are scenario variables to be set for specific energy modelling studies. I. 4. North American database fuel efficiency assumptions Fuel efficiencies for gasoline cars were taken from the US Automotive Fuel Economy Programme 2000 report. The urban car efficiency was calculated from the average fuel consumption of all cars and the family car efficiency was calculated from the average fuel consumption of all light trucks. Fuel efficiencies for hybrid, electric and natural gas vehicles are also reported for 2000. The hybrid data is based on the Honda Insight US EPA fuel consumption data and the electric car data on the Peugeot 106 efficiency, scaled up using the difference in efficiency of EU and US gasoline urban vehicles. Data for the natural gas fuelled car was calculated by scaling the gasoline family car efficiency to the difference in efficiency of the natural gas powered 2002 Ford F-150 pickup truck and its gasoline equivalent. Future fuel efficiencies for all vehicles were estimated by scaling them relative to the EU vehicle efficiencies to maintain consistency across the data set in terms of technological development assumptions. I. 5. Distribution of sales in the EU in year 2000


76

Data are given for the distribution (in percent) of sales between: • Gasoline Urban cars • Gasoline Long Distance Family cars • Diesel Urban cars • Diesel Long Distance Family cars for each EU Member State. These values have been estimated from ACEA data on total vehicle sales, percentage of diesel vehicles in sales and the average engine capacity of the vehicles sold in year 2000. The estimates are based on the assumptions that: • The same proportions of gasoline and family cars are sold in the urban car and long

distance family car size ranges. • The mix of vehicles in each size category have the same average engine capacity in all

Member States. From these assumptions the overall average engine capacity can be related to the engine capacities of each category through the relationship:

( ) ( )% %GU N GL T N P DU N DL T N D

Average Engine CapacityT

⋅ + ⋅ − ⋅ + ⋅ + ⋅ − ⋅ =

Where GU is the average engine capacity of gasoline urban cars GL is the average engine capacity of gasoline long distance family cars DU is the average engine capacity of diesel urban cars DL is the average engine capacity of diesel long distance family cars P% is the percentage of petrol cars in year 2000 sales D% is the percentage of diesel cars in year 2000 sales T is total sales N is sales of urban cars For these estimates urban cars were take to be made up of segments A, B and C of the standard market segmentation used by ACEA (i.e. mini, supermini and lower medium). Long distance family cars were taken to be made up of the remaining segments; D, E, F, G, H and I (upper medium, executive, luxury saloon, specialist sports, dual purpose and MPV). Average engine capacities were estimated from a “basket” of models taken from each category. Data for the division of sales between market sectors are available from SMMT for the UK. Comparison between the estimates based on the above equation and actual data showed a good match. For example the urban car share of sales was estimated to be 62.8% and the actual share was 63.1%. I. 6. Limitations to the database


77

The database supplied should be regarded as an objective assessment of prospective vehicle prices and fuel efficiencies up to 2030. It is based on a range of information sources, and also on assumptions needed to bridge gaps in the data available. Consequently it is important to noted that the data are subject to uncertainties which could benefit from assessment in modelling sensitivity analyses. Key issues are: • Vehicle prices are presented excluding taxation. This is the most practical approach

because taxation may change over the period to 2030. Indeed differential taxation is one option for encouraging the choice of vehicles and could be the subject of modelling assessments.

• Current differentials in vehicle prices are affected by a range of factors including taxation, demand, marketing strategies, etc as well as the costs of production and delivery. It is uncertain how these differentials will change in future. The view taken herein is that prices will converge as the wealth of EU Member States converge, but alternative differentials could be assessed in sensitivity studies.

• There is some contention over the potential for further fuel efficiency improvements in conventional vehicle propulsion systems. Two options have been included in the database to support sensitivity studies.

• Estimates have been provided of the mix of vehicle sales by type and fuel use for each EU Member State in year 2000. It is expected that the future mix of sales will be one aspect covered by the modelling studies. The year 2000 estimates give good agreement with UK data and look sensible for other EU countries with the exception of Italy for which the approach places all sales in the urban car category. Italy does have a large proportion of small car sales but 100% urban cars seems unlikely.

I. 7. Bibliography 1. ACEA, 2002. Automotive Key Figures. http://www.acea.be 2. AFDC, 2002. Alternative Fuels Data Centre (website), US Department of Energy:

http://www.afdc.doe.gov/afvehicles.html 3. AM, 2002. Official consumer websites of automobile manufacturers (Ford, Opel, VW,

Audi, BMW, Renault, Peugeot) for Germany, Netherlands, Spain & Italy. 4. Brown, LF (2001). “A comparative study of fuels for on-board hydrogen production for

fuel-cell-powered automobiles”. International J. of Hydrogen Energy 26 (2001) 381-397 5. CG, 2002. “Consumer Guide, Car & Truck Test – The smart shopper’s guide to a great

deal”; US magazine, June 2002 issue. 6. Chalk et al. (2000). “Challenges for fuel cells in transport applications”; Chalk, SG,

Miller, FJ, Wagner, FW; J. of Power Sources 86 (2000) 40-51. 7. Competition DG, 2000. “Car price differentials in the European Union on 1 May 2000”,

Competition DG, May 2000. 8. DTI, 2001. Digest of United Kingdom Energy Statistics. 9. DTI, 2002. “Options for a Low Carbon Future”, A report produced for the UK

Department of Trade and Industry, the Department for Environment, Food and Rural Affairs and the Performance and Innovation Unit by AEA Technology in collaboration with the Imperial College Centre for Energy Policy and Technology, February 2002.

10. Edmunds.com, 2002. True Cost to Own online database from Edmunds.com website: http://applications.edmunds.com/products/tco/TCOintroController

11. EST, 2002. UK Energy Savings Trust’s Powershift website: http://www.est-powershift.org.uk

12. ETSU (1994). “An Appraisal of UK Energy Research, Development, Demonstration and Dissemination”, report no. R83.


78

13. EUI, 2002. News article on car prices in the EU; The European Commission representation in Ireland website: http://www.euireland.ie/news/market/0202/carpricesintheEU.htm

14. EVW, 2002. EV World website: http://ww.evworld.com 15. Ford, 2002. Ford’s Th!nk electric car website: http://www.thinkmobiligy.com 16. FTEN, 1999a. FT Energy Newsletter – Energy Economist Ballard: “Leading the fuel cell

charge”. Issue 216 – 21/10/1999 17. FTEN, 1999b. FT Energy Newsletter – Energy Economist Fuel Cells: “Coming to a car

near you”. Issue 209 – 29/03/1999 18. GMC, 2001. “Well-to-Wheel Energy Use and Greenhouse Gas Emissions of Advanced

Fuel/Vehicle Systems – North American Analysis”, General Motors Corporation, Argonne National Laboratory, BP, ExxonMobil and Shell; June 2001.

19. Harding, GG (1999). “Electric vehicles in the next millennium”. J. of Power Sources 78 (1999) 193-198.

20. Hart et al. (2000). “Methanol infrastructure – will it affect the introduction of SPFC vehicles?”, Hart, D, Leach, MA, Fouquet, R, Pearson, PJ, Bauen, A; J. of Power Sources 86 (2000) 542-547

21. HEH, 1999. “Top 20 Hitlist: Best Selling Vehicles in USA; Jan 99”, Honda European Homepage website: http://www.europeanhonda.demon.nl/car_sales_stats_jan99.htm

22. Höhlein et al. (1999). “Fuel cell trains for road traffic.”, Höhlein, B, Biedermann, P, Grube, T, Menzer, R; J. of Power Sources 84 (2000) 203-213

23. KFB, 2000a. “Hydrogen or methanol in the transportation sector?”, KFB 2000 (Swedish Environment Agency).

24. KFB, 2000b. “Hybrid Electric Vehicles - An alternative for the Swedish market?”, KFB 2000 (Swedish Environment Agency).

25. KFB, 2000c. “Clean Vehicles with Electric Drive”, Final report from the Swedish research, development and demonstration programme on electric and hybrid vehicles, KFB (Swedish Environment Agency), 1993-2000

26. KFB, 2000d. “The Costs and Benefits of Electric Vehicles - Should battery, hybrid and fuel-cell vehicles be publicly supported in Sweden?”, KFB 2000 (Swedish Environment Agency)

27. Lipman, Timothy (2000). “Manufacturing and lifecycle costs of battery electric vehicles, direct-hydrogen fuel cell vehicles, and direct-methanol fuel cell vehicles”. Draft for the American Institute of Aeronautics and Astronautics, personal communication.

28. Lynch Motor Ltd, 2002. Website for electric vehicle supplier – Lynch Motor Co. Ltd, UK: http://www.lynchmotor.com

29. MIRA, 2002. Advanced Powertrain Projects, MIRA Ltd UK website: http://www.mira.co.uk/app/

30. MIT, 2000. “On the Road in 2020 – A life-cycle analysis of new automobile technologies”, Malcolm A. Weiss, John B. Heywood, Elisabeth M. Drake, Andreas Schafer, and Felix F. Au Yeung; Energy Laboratory, Massachusetts Institute of Technology; October 2000.

31. NBB, 2001. National Biodiesel Board (USA) website : http://www.Biodiesel.org/fuelfactsheet.htm

32. NHTSA, 2000. “Automotive Fuel Economy Program – Annual Update Calendar Year 2000”, National Highway Traffic Safety Administration, US Department of Transportation, 2000.

33. NRC, 2000. “Alternative Fuels for US Transportation”, January 2000, Committee on Alternative Transportation Fuels, Transportation Research Board, US National Research Council

34. Ogden et al. (1999). “A comparison of hydrogen, methanol and gasoline as fuels for fuel cell vehicles”, Ogden, JM, Steinbugler, MM, Kreutz, TG; J. of Power Sources 79 (1999) 143-168


79

35. Ogden, JM (1999). “Developing an infrastructure for hydrogen vehicles: a Southern California case study”. International J. of Hydrogen Energy 24 (1999) 709-730

36. Padro, CE, and Putsche, V (1999). “Survey of the Economics of Hydrogen Technologies”. NREL/TP-57027079, National Renewable Energy Laboratory.

37. RMI, 1999. “A Strategy for the Hydrogen Transition”, Rocky Mountains Institute, USA, 1999.

38. SMMT, 2002. Motor Industry Facts. www.smmt.co.uk 39. USCAR, 2001. USCAR Website 2001 - Next Generation Power Trains,

http://www.uscar.org/ 40. USDOE, 1997. “Batteries for Electric Drive Vehicles: Evaluation of Future

Characteristics and Costs through a Delphi Study”, Transportation Technology R & D Centre, US Department of Energy, 1997.

41. USDOE, 1999a. “A Full Fuel-Cycle Analysis of Energy and Emissions Impacts of Transportation Fuels Produced from Natural Gas”, December 1999, Transportation Technology R & D Centre, US Department of Energy

42. USDOE, 1999b. “Effects of fuel ethanol use on fuel-cycle energy and greenhouse gas emissions”, January 1999, Transportation Technology R & D Centre, US Department of Energy

43. USDOE, 2000a. “FY2000 Progress Report for Fuel Cell Power Systems”, Energy Efficiency and Renewable Energy Office of Transport Technology, US Department of Energy

44. USDOE, 2000b. “FY2000 Progress Report for Advanced Technology Development Program”, Energy Efficiency and Renewable Energy Office of Transport Technology, US Department of Energy

45. USDOE, 2001. “Future US Highway Energy Use: A Fifty Year Perspective”, May 2001 Draft, Office of Transport Technology, Energy Efficiency & Renewable Energy; US Department of Energy

46. USEPA, 2002a. US EPA’s Fuel Economy website and searchable database: http://www.fueleconomy.gov

47. USEPA, 2002b. Fuel Economy Information website, Office of Transportation and Air Quality, US EPA: http://www.epa.gov/otaq/mpg.htm

48. VTFSG, 1999. “Alternative and Future Fuels & Energy Sources for Road Vehicles”, December 1999, Vehicle Technology Fuels Sub-Group (Transport Issue Table National Climate Change Process), USA.

49. WFCC, 1999. “Market Actors - Promoting the Benefits and Applications of Fuel Cells”, presentation at Sixth Grove Fuel Cell Symposium 1999, World Fuel Cell Council

50. Which? Online, 2002. “Car Running Costs 2002”, database of UK cars for 2001 from the ‘Which? Online’ website, Motoring section (members’ area): http://www.which.net/motoring/contents.html

dynamics of the introduction of new passenger car technologies

Documents