a contribution to cleaner vehicle technologies final version for distribution

p.iii

A contribution to cleaner vehicle technologies

Jean-Marc Timmermans

May 2010

Jean-Marc

Rectangle

p.iv

Colophon Print: Silhouet, Maldegem © 2010 by Jean-Marc Timmermans, Vrije Universiteit Brussel© 2010 VUBPRESS Brussels University Press VUBPRESS is an imprint of ASP nv (Academic and Scientific Publishers nv)Ravensteingalerij 28B-1000 BrusselsTel. ++32 (0)2 289 26 50 Fax ++32 (0)2 289 26 59E-mail: [email protected] www.vubpress.be ISBN 978 90 5487 751 6 NUR 973 / 977 / 961 Legal Deposit D/2010/11.161/075 All rights reserved. No parts of this book may be reproduced or transmitted in any form or by any means,electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of theauthor or the publisher.

p.v

Ik draag dit werk graag op aan mijn dochtertjes Elise & Charlotte

Jean-Marc

Rectangle

p.vi

Jean-Marc

Rectangle

p.vii

Jury Prof. dr. ir. J. Tiberghien, Chairman, Vrije Universiteit Brussel, Belgium Prof. dr. ir. R. Pintelon, Vice-Chairman, Vrije Universiteit Brussel, Belgium Prof. dr. ir. Ph. Lataire, Advisor, Vrije Universiteit Brussel, Belgium Prof. dr. ir. J. Van Mierlo, Advisor, Vrije Universiteit Brussel, Belgium Prof. dr. ir. P. Van den Bossche, Secretary, Erasmus Hogeschool Brussel, Belgium Prof. dr. ir. J. Melkebeek, Universiteit Gent, Belgium Prof. dr. eng. J. A. Ferreira, Delft University of Technology, The Netherlands

Jury

p.viii

Jean-Marc

Rectangle

p.ix

Acknowledgements Via deze weg wens ik mijn promotor Prof. Philippe Lataire te bedanken om het promotorschap te willen aangaan voor het voorliggend onderzoekswerk. Terzelfdertijd gaat ook mijn dank uit naar mijn co‐promotor Prof. Joeri Van Mierlo die mij de kans bood om mee te werken aan een aantal interessante onderzoeksprojecten rond de milieu‐evaluatie van voertuigen die mee aan de basis liggen van het voorliggend proefschrift. Ook wens ik mijn dank te betuigen aan wijlen Prof. Maggetto. Zijn charisma en doorzettingsvermogen waarmee hij onze vakgroep en het onderzoek leidde waren steeds bijzonder aanmoedigend. Verder wens ik Philippe te bedanken voor de aangename samenwerking in het kader van de onderzoeksprojecten rond de ontwikkeling van elektrische fietsen voor postbedeling. In het bijzonder mijn dank voor de vele kilometres testritten die hij met de verschillende prototypes met heel veel zorg aflegde. Ook Francis Heymans en Jens Nietvelt ben ik in dit kader enorm veel verschuldigd. Ook wens ik Francis te bedanken voor het zorgvuldige onderhoud van onze prototypes en testmodellen elektrische fietsen en voor de uitstekende en zeer aangename technische ondersteuning. Peter Van den Bossche wens ik eveneens te bedanken voor de aangename en leerrijke samenwerking gedurende de voorbije jaren. Peter liet mij kennis maken met “de wondere wereld der normen”, zorgde voor een uitstekende gidsing tijdens gezamenlijke uitstappen naar congressen en deelde ons zijn liefde voor de elektriciteit, batterijen, de serie bekrachtigede gelijkstroommotor en uiteraard voor het batterij elektrisch voertuig en hiervoor ben ik hem zeer dankbaar. Verder zijn er de vele (ex‐)collega’s van de vakgroep ETEC en de onderzoeksgroep MOBI die ik wens te vernoemen voor de aangename werksfeer die zij allen hebben gecreëerd. In het bijzonder wens ik Julien Matheys, Nele Sergeant, Heijke Rombaut, Maarten Messagie, Fayçal Boureima, Riccardo Barrero, Bavo Verbrugge, Noshin Omar, Thierry Coosemans, Frederik Van Mulders te bedanken voor hun collegialiteit en vriendschap. Ook wens ik alle collega’s van onze vakgroep via deze weg te danken voor de fijne samenwerking. In het bijzonder wens ik Nico Smets, Daan De Wilde, Gert Weyns, Pedro Maciel, Steven Van Damme, Johan Deconinck en Gert Nelissen te vermelden voor de interessante discussies, aangename samenwerking en hun vriendschap. Further, I would like to thank all the non‐dutch speaking (ex‐) colleages of our research group and of our department for the daily nice working environment they all helped to create! Verder wens ik nog een aantal mensen (opnieuw) te danken voor het vele naleeswerk van de eerste versies van dit proefschrift: Julien Matheys, Joeri Van Mierlo, Jan Cappelle, Philippe Lataire en Nele Sergeant. Ondanks jullie overdrukke agenda kon ik toch op jullie onmisbare hulp alsook op jullie voortdurende aanmoedigingen tijdens het schrijven van dit werk blijven rekenen. Ook zijn er de vele mensen waarmee ik in het kader van de verschillende onderzoeksprojecten heb samengewerkt. Ook hen wens ik van harte te bedanken. In het bijzonder: Jean Vander Elst, Tim Panneels, Dirk Van den Berk, Johan Borgers, Tania Van Mierlo, Leen Govaerts, Tobias Denys. Particular gratitude to all persons that were involved in the NEPH project that forms the basis of a large part of the research work described in this text. Many people from different countries and companies have all contributed directly or indirectly to the work presented. Thank you all for the nice collaboration

Acknowledgements

p.x

and interesting discussions. I would like to mension the following persons in particular: Jean‐Michel Durand, Dr. Gössel, Dr. Gloëckner, Mr. Salami, Dr. Kukuk, Pierre Prenleloup, Antoine Juan, Frye Blake, Sirko Müller, Simon Smith, Marc Pouw, Nuno Bettencourt, Antonino Scribellito and Prof. Zahoransky. Dit werk was niet mogelijk geweest zonder de voortdurende steun en liefde van mijn vrouw Stefanie Hugé en van mijn twee lieve dochtertjes Elise en Charlotte! Daarom wens ik hen eveneens via deze weg te bedanken voor de onmisbare steun.

De Pinte, 27 maart 2010 Jean‐Marc Timmermans

Acknowledgements

p.xi

Summary Air pollution is an important concern for modern society as it induces important negative effects on the living environment as well as on human health. Road transportation is identified as an important contributor of this air pollution and therefore significant efforts are required to lower harmfull emissions and consequently the impacts of road vehicles. Different vehicle technologies are being used in road vehicles and each represent different characteristics with regard to emissions to the air. Different fuels are being used as the energy carrier for vehicular application. Alternative vehicle technologies are being developed such as hybrid and electric propulsion systems for road vehicles. Several environmental problems due to air pollution of road vehicles, such as global warming and air quality depletion, have been identified. When evaluating the environmental effects of road vehicles and when comparing different vehicle technologies a clear and sound methodology for environmental assessment is required. However, often incomplete or partial comparisons are used to evaluate environmental performance of road vehicles. In this work the different sources of emissions and the main environmental damages related to the use of road vehicles are described. An overview of different possible vehicle technologies and potential policy measures is provided. Further in this dissertation a new methodology for the environmental assessment of road vehicles is described. This methodology is mainly used for an environmental rating tool for road vehicles. The rating tool allows ranking different road vehicles based on their impact on the environment. In order to be able to compare road vehicles with different power system technologies or using different kinds of fuels, a well‐to‐wheel framework is used. In this way both tailpipe emissions and the emissions related to the fuel production phase are taken into account. Different aspects of the environmental impact of the vehicle such as the greenhouse effect, air quality depletion and noise pollution are integrated in the methodology. For this purpose, the methodology includes four different damage categories: global warming, human health impairing effects, harmful effects on ecosystems and noise. Further, the different impact categories are aggregated and used to calculate a single indicator, called Ecoscore. The Ecoscore is a number between 0 and 100: the higher the Ecoscore the lower the environmental impact of the vehicle. The applicability of the Ecoscore methodology is demonstrated with real vehicle data. From the analysis made with this methodology, a positive evolution of the environmental performance of vehicles through time is observed. This is mainly due to the ever more stringent European emission regulations. A low environmental impact (and therefore a high Ecoscore) is obtained for a battery electric vehicle charging its battery from the electricity grid. Also hybrid petrol‐electric vehicles and the CNG vehicles obtained a high Ecoscore. The LPG vehicle shows the best environmental performance amongst all conventional vehicles, whereas Euro 4 petrol and diesel vehicles have a similar Ecoscore. When considering the complete range of road vehicles on the roads in Belgium, the emissions of CO2 from more recent vehicles, which were directly related to the fuel consumption, were not always reduced. The positive influence of an improved engine technology is sometimes annihilated by an increase of the vehicles’ weight or an increased energy consumption caused by certain on‐board options. In recent years the newest generation of diesel vehicles has caught up its delay concerning their environmental performances on the petrol vehicles. Furthermore, the difference between the environmental performance of those vehicles and the environmental performance of the LPG vehicles has been reduced. The environmental assessment methodology was found to be robust and the environmental indicator Ecoscore is thus applicable as a policy instrument (taxation, incentives, consciousness raising campaigns, etc.) to support the use of environmentally friendly vehicles. The ambition of the Ecoscore environmental rating tool is to lead to a common system for policy measures in Belgium and possibly in other European countries, to promote the introduction and use of cleaner vehicles. The Ecoscore methodology can also be used for the ranking of heavy duty vehicles and of two‐wheelers. A calculation

Summary

p.xii

tool is developed and can be used to calculate the Ecoscore for a small set of vehicles available in the tool as an illustration. Own vehicle data can be introduced and the different parameters of the methodology can be changed by the user to investigate the influence of these parameters on the end result. As was mentioned earlier, new technologies can be a solution to lower the impact on the environment of road vehicles and of transportation in general. The well‐to‐wheel energy use and the environmental impact of electric vehicles were found to be clearly lower compared to that of conventional vehicles (e.g. diesel or petrol cars). In this context, light electric vehicles are in this context a promesing new means of transportation and more particularly an interesting alternative for goods delivery. Many European postal operators show interest for using these light electric vehicles for postal distribution. In particular electric bicycles are found attractive for lowering the physiological stress of postmen during the delivery of mail items. The introduction of automatic sorting machines in the postal business leads to an increased availability of postmen for delivery within unchanged working hours. For this reason, the delivery rounds can become longer with more mail items to to be distributed. Light electric vehicles have a role to play in softening this increased workload for the postmen. Moreover, electric bicycles allow ageing postmen to continue their mail and parcels delivery duty, hence limiting the social impact of early retirement. Postal operators are also seeking to lower the total cost of ownership (operational costs) of their delivery fleet and light electric vehicles and electric bicycles in particular are a promising option to substitute thermal mopeds for instance. But also the raising consciousness of climate and pollution issues, stimulate postal operators to continuously look for new and cleaner vehicles. However, at the beginning of this research, most electric bicycles available on the European market did not satisfy the requirements of the postal operators. In this context, a European project called NEPH was created. An extensive questionnaire was performed by the department of Electrical Engineering and Energy Technology to collect information on the requirements and wishes of the postal operators. From the analysis of the feedback of the postal operators a set of main postal mission parameters that influence a mail delivery mode were determined: total trip distance, number of stops on the delivery round, the payload, cumulative height difference, etc. A first important result of this research was the translation of the postal requirements into specifications for the sizing of the electric power system starting from typical and available postal mission parameters. For this purpose, a calculation tool that uses an approximate calculation method to estimate the required electrical energy of the power system has been developed. This method uses the postal mission parameters as input parameters for the calculation tool. The tool is illustrated by using it to calculate the required energy capacity of an electric bicycle for typical postal delivery use. From the analysis a clear gap between the required energy capacity for such a mission and the energy capacity from available (normal consumer) electric bicycles was demonstrated. The calculation tool allows determining the required energy capacity in case of each individual postal delivery round. Another way of using the calculation tool is to match or to optimize the distribution rounds to the energy capacity of the battery pack that is used. Test rides with several prototypes of electric bicycles for postal delivery have demonstrated that in case of demanding use of the bicycle, the heating of the wheel motor caused the motor temperature to reach the maximal allowable motor temperature. For this purpose, a thermal management system for the electric permanent magnet d.c. wheel motor is proposed. The thermal management modulates the maximum allowable armature current to the measurement of the motor’s temperature. This principle allows using the maximal capability of the motor without causing sudden motor shut‐down. Indeed, the maximum armature current is only lowered in case of excessive heating of the motor. Overheating is prevented as the armature current is progressively reduced with increasing motor temperature. An optimization of the energy consumption of a power‐on‐demand power system, used for the NEPH power system, is discussed. Postmen indeed tend to minimize their own contribution to the total traction power in case of the NEPH power system. The optimization principle links the maximal allowable armature current of the d.c. motor to the armature voltage. This optimization principle was implemented on new NEPH controllers and validated on the road. An important reduction of 29% and 38% was obtained. This solution allows to encourage the cyclist to deliver more biomechanical power as the contribution of the motor to the total traction power is reduced with increasing vehicle speed. However, at start the maximal traction force of the motor is still available and thus still allows lowering the physical stress on the postmen.

Summary

p.xiii

List of Symbols

Symbol Description Units Part I dij Environmental damage cost of emission spicies i and location

of emission j (ACEEE) ($c/mg)

Di,j Partial damage of pollutant j to the category i / eij Quantity of emission spicies i and location of emission j ,

averaged of the(ACEEE) (g/mi)

EC Energy content (kJ/kg) Ecoscore Ecoscore value (environmental assessement) / Ecoscorecapacity Ecoscore taking into account the passenger capacity of busses / Ecoscorei Ecoscore of vehicle i / Ecoscorefleet Average Ecoscore of the fleet / Ecoscorefleetuse Average Ecoscore of the fleet use, based on vehicle mileage / Edirect Direct emission (BIM Cleaner Drive) (g/km) EDX Environmental damage index (ACEEE) ($c/mi) Ee

j,direct Direct emission of pollutant j linked to the engine (for heavy duty vehciles)

(g/kWh)

Ei,j,direct Indirect emission value for pollutant j and damage category i (g/km) Ei,j,indirect Indirect emission value for pollutant j and damage category i (g/km) Eindirect Indirect emission (BIM Cleaner Drive) (g/km) Ej Emission value of pollutant j (g/km) Ej,direct Total direct emission of pollutant j contributing to damage

category i (g/km)

Ej,indirect Total indirect emission of pollutant j contributing to damage category i

(g/km)

Etot Total emissions (BIM Cleaner Drive) (g/km) FC Fuel consumption (L/100km) Fj Indirect emission factor for pollutant j (mg/kWh) kCO2 Emission factor for CO2 (kg/L) kS Sulphur content of the fuel (mg/kg) or (ppm) Leq(t1,t2) Equivalent sound pressure level between time instance t1 and

t2 (dB)

Lp Sound pressure Level (dB) Mi Mileage of vehicle i (km) Mtot Total mileage of all vehicles of the fleet (km) Ni Number of vehicles of model i in the vehicle fleet (#) Ntot Total number of vehicles of the vehicle fleet (#) p(t) Acoustic pressure or sound pressure (as a function of time) (Pa) p0 Reference pressure 2,105 (Pa) Qi Total damage of category I of the assessed vehicle / qi Normalised damage of category i (%) Qi,ref Total damage of category i of reference vehicle / SECi,j,rural Rural specific external cost of pollutant j to the category i (€/kg) SECi,j,urban Urban specific external cost of pollutant j to the category i (€/kg)

List of Symbols

p.xiv

Symbol Description Units TI Total impact of the assessed vehicle (%) TIi Total impact of vehicle i (%) αi Weighting factor of damage category i (%) γ Electric energy consumption (kWh/100km) δi,j,direct Direct impact factor of pollutant j to the category i / δi,j,indirect Indirect impact factor of pollutant j to the category i / εdirect Direct energy consumption (MJ/100km) εindirect Indirect energy consumption (MJ/100km) εwtw Well-to-Wheel energy consumption (MJ/100km)

ICEη Average value of the efficiency of the engine (%)

ρf Fuel density (g/L) or (kg/Nm³) σurban urban mileage distribution (%) σrural rural mileage distribution (%) χ Expended energy (MJ/MJ) ωind indirect emission weighting factor (BIM Clean Vehicles) (-) Part II A Absolute altitude (above sea level) (m) a Vehicle acceleration (m/s²) Ar Relative altitude (m) Astart Absolute altitude of the starting point of the trip (m) BAM Battery Ageing Margin (%) CR Coefficient of rolling resistance (-) CSC Capacitance of the super capacitor (F) CX Aerodynamic drag coefficient (-) d run (in the context of a slope) (m) Ddelivery Active distance: part of the round during which mail is

distributed (m)

Dmission Total distance of the delivery trip or mission (m) dw Wheel diameter (m) Eallowed allowed energy consumption of the power unit (J) Ebat Energy content of the battery pack (J) Ebat_req Required electrical energy from the battery pack (J) Ecd Required mechanical energy for continuous driving (J) Ehd Required mechanical energy to cover a height difference hcumul (J) ESC Energy content of the super capacitor (J) ESC_nom Nominal energy content of the super capacitor (J) Ess Kinetic energy required to accelerate a vehicle to vs (J) Etm Required mechanical energy allocated to the elelctric motor (J) Etot Total required mechanical energy, at the level of the wheel (J) FA Acceleration force (N) FR Rolling resistance (N) FS Slope resistance force (N) Ftm Traction force from the electric motor, at the level of the

wheel (N)

Ftot Total resistance force (N) FW Aerodynamic resistance force or aerodynamic drag (N) g Gravitational constant 9,81 (m/s²) H Heart rate (bpm) H Average heart rate of the trip (bpm) hcumul Cumulative height difference (m) Hr Relative heart rate (%) Ia d.c. motor armature current (A)

List of Symbols

p.xv

Symbol Description Units Ia_0 d.c. motor no-load current (A) Ia_lim Limit value of the controller for the d.c. motor armature

current (A)

Ia_max d.c. motor maximal allowed armature current (A) Ia_nom nominal value of the d.c. motor armature current (A) Ibat battery current (A) kT d.c. motor torque constant (Nm/A) M Total mass of vehicle (kg) Mbat Mass of the battery pack(s) (kg) Md Mass of driver (kg) Mmot Mass of the electric motor unit (kg) Mp Mass of payload (kg) Mplat Mass of the vehicle platform (excluding the power system) (kg) Mv Unladen mass of vehicle (kg) Pavg_allowed Average allowed power use of the power unit (W) Pbat Electric power delivered by the battery pack (W) Pcon Electric power delivered by the motor controller (W) Pmot Mechanical motor power (W) Pmot Mechanical power of the motor, at the level of the shaft (W) Pmot_nom Nominal power of the motor (W) PR Power to overcome rolling resistance (W) PS Power to overcome a slope (W) Ptc Mechanical traction power from the cyclist, at the level of the

wheel (W)

Ptm Mechanical traction power from the electric motor, at the level of the wheel

(W)

Ptot Total required mechanical power to overcome all resistive forces

(W)

PW Power to overcome aerodynamic drag (W) rw Wheel radius (m) S Frontal area of the vehicle (or bicycle including the driver) (m²) SOCSC State-of-Charge of the super capacitor (J) stops Number of stops on the delivery round (#) Ta Time limit of power limit integration (s) Tdriving Time of driving (exclusive stop times) (s) Tem Electromagnetic torque (Nm) Tgr Twist grip postion (%) Th Time horizon for power consumption (s) Tmission Time duration of delivery round or mission (s) Tw_tot Total torque at the wheel (Nm) Ua d.c. motor armature voltage (V) Ubat Battery voltage (V) v Speed of the vehicle (m/s) vgrade Speed on maximal grade (m/s) vs Maximal speed reached between two stops (m/s) VSC Voltage of the super capacitor (V) VSC_rated Rated voltage of the super capacitor (V) vw Headwind speed relative to ground (m/s) α Road inclination (°) δm Energy density (Wh/kg) Δh rise (in the context of a slope) (m)

conη Efficiency of the motor controller (%)

conη Average efficiency of the motor controller (%)

_ maxconη Maximum value of the efficiency of the motor controller (%)

List of Symbols

p.xvi

Symbol Description Units

motη Efficiency of the electric motor (%)

_ maxmotη maximum value of the efficiency of the electric motor (%)

ptη Efficiency of the electric power train (%)

_pt cdη Global efficiency of the power train, related to continuous driving operation

(%)

_pt hdη Global efficiency of the power train, related to hill climbing operation

(%)

_pt missionη Global efficiency of the electric power train (%)

_pt ssη Global efficiency of the power train, related to start and stop operation

(%)

tmη Efficiency of the transmission and gear system (%)

θlim Limit value of the motors’ operational temperature (°C) θmot Motors’ operational temperature (°C) θ1, θ2 Temperature boundary values for thermal motor

management (°C)

ξ Assistance factor (%)

cdξ Global assistance factor, related to continous driving operation

(%)

hdξ Global assistance factor, related to hill climbing operation (%)

missionξ Global assistance factor, related to a certain mission (%)

ssξ Global assistance factor, related to start and stop operation (%)

ρair Air density (kg/Nm³) ωp rotational speed of the pedals (rad/s)

List of Symbols

p.xvii

List of Abbreviations

Abbreviation Description Part I ACEEE American Council for Energy Efficient Economy AUL Auto-Umweltliste CARB California Air Resources Board CCGT Combined Cycle Gas Turbine CNG Compressed Natural Gas COPERT COmputer Programme to calculate Emissions from Road Transport CRT Continuous regenerating trap DALY Disability Adjusted Life Years DIN Deutsches Institut für Normung DPF Diesel Particulate Filter EC European Commission EDX Environmental Damage indeX EEC European Economic Community EEV Enhanced Environmental Vehicle FIA Fédération Internationale de l’Automobile GHG Greenhouse Gas GWP Global Warming Potential ICE Internal Combustion Engine IFEU Institut für energie-und Umweltforschung Heidelberg IPCC Intergovernmental Panel on Climate Change ISO International Organisation for Standardization LCA Life Cycle Assessment LCI Life Cycle Inventory LCIA Life Cycle Impact Assessment LHV Lower Heating Value LPG Liquefied Petroleum Gas LT Long Term MEET Methodology to Estimate Emissions from Transportation MIRA-T Milieurapport – Transport MIVB Maatschappij voor het Intercommunaal Vervoer te Brussel NEDC New European Driving Cycle NIS Nationaal Instituut voor Statistiek (Belgium) PDF Potentially Disappeared Fraction PEM Proton Exchange Membrane PEM Proton exchange membrane RME Rapeseed Methyl Ester SCR Selective Catalytic Reduction SPL Sound Pressure Level STIB Société des Transport Intercommunaux de Bruxelles


p.xviii

Abbreviation Description SUV Sports Utility Vehicle TOFP Total ozone formation potential TTW Tank-to-Wheel VAT Value Added Taxes VCD Verkehrsclub Deutschland VOC Volatile Organic Compounds VSP Vehicle Simulation Programme WHO World Health Organization WTT Well-to-Tank WTW Well-to-Wheel Part II BMS Battery Management System BAM Battery Ageing Margin CEN European Committee for Standardisation DoD Depth of Discharge EPAC Electric Power Assisted Cycle ETEC Energie Technologie en Electrotechniek (NL) E-Tour Electric Two-wheelers On Urban Roads (project acronym) HMI Human-machine-interface IFMA Internationale Fahrrad- und Motorrad-Ausstellung IGBT Insulated gate bipolar transistor LEV Light Electric Vehicle MOSFET Metal oxide semiconductor field effect transistor NEPH New Electric Postmen Helper NiCd Nickel-cadmium NiMH Nickel-metal-hydride PAS Power Assist System PTT Posterijen, Telegrafie en Telefonie (The Netherlands) SLI Starting, Lighting and Ignition (battery) SoC State-of-charge TNT Thomas Nationwide Transport Pb-acid Lead-acid Li-ion Lithium-ion

Chemical compounds and groups

Chemical symbol IUPAC name (common name) Ar Argon C3H8 Propane C4H10 Butane CFC Chlorofluorocarbon: group of organic compound containing carbon,

chlorine and fluorine CH4 Methane CO Carbon monoxide CO2 Carbon dioxide H2SO4 Sulphuric acid HC Hydrocarbons HFC Hydrofluorocarbons HNO3 Nitric acid (aqua fortis)


p.xix

Chemical symbol IUPAC name (common name) MMT Methylcyclopentadienyl manganese tricarbonyl N2O Dinitrogen monoxide (Nitrous oxide or laughing gas) NH3 Ammonia NMHC Non-Methane Hydrocarbons NO Nitrogen monoxide (nitric oxide) NO2 Nitrogen dioxide NOX Nitrogen oxides: group of binary compounds of oxygen and nitrogen O3 Trioxygen (ozone) PAH Polycyclic aromatic hydrocarbons PAN Peroxyacetylnitrate Pb Lead PFC Perfluorocarbons PM Particulate Matter PM10 Particulate Matter fraction < 10 μm PM2,5 Particulate Matter fraction < 2,5 μm POP Persistent organic pollutant SF6 Sulphur hexafluoride SO2 Sulphur dioxide SOX Sulphur oxides: group of binary compounds of oxygen and sulphur TEL Tetra ethyl lead UFP Ultra Fine Particles fraction <100nm VOC Volatile Organic Compound: group of organic chemical compounds,

vaporizing readily under normal conditions


p.xx

Jean-Marc

Rectangle

p.xxi

Table of Contents

Jury ........................................................................................................................................................... vii

Acknowledgements ................................................................................................................................... ix

Summary.................................................................................................................................................... xi

List of Symbols ......................................................................................................................................... xiii

List of Abbreviations ................................................................................................................................xvii

Table of Contents......................................................................................................................................xxi

List of Figures...........................................................................................................................................xxv

List of Tables ...........................................................................................................................................xxix 0 General introduction.............................................................................................................................1

0.1 Environmental assessment of road vehicles..................................................................1 0.2 Light electric vehicles for postal delivery.......................................................................1 0.3 Outline...........................................................................................................................2

PART I: Environmental assessment of road vehicles ...........................................3

1 Environmental damages related to the use of road vehicles ................................................................5 1.1 Environmental effects of atmospheric pollution from road vehicles.............................5 1.2 Foremost atmospheric pollutants from road vehicles...................................................8 1.3 Regulations for air quality ...........................................................................................12 1.4 Environmental problems due to air pollution from road vehicles ...............................14

2 Relation between vehicle technology and pollution ...........................................................................19 2.1 Sources of emissions ...................................................................................................19 2.2 Direct emissions ..........................................................................................................20 2.3 Indirect emissions........................................................................................................24 2.4 Noise ...........................................................................................................................25

3 What can be done? .............................................................................................................................27 3.1 Introduction ................................................................................................................27 3.2 Technological solutions ...............................................................................................27 3.3 Policy measures...........................................................................................................33

List of Contents

p.xxii

4 Environmental assessment tool ..........................................................................................................39

4.1 Environmental assessment tools for road vehicles: state of the art............................39 4.2 Well-to-Wheel framework ..........................................................................................44 4.3 Well-to-Wheel energy consumption ...........................................................................45 4.4 Methodology for environmental assessment..............................................................46 4.5 Results for the well-to-wheel energy consumption.....................................................57 4.6 Results for the environmental assessment methodology ...........................................58 4.7 Sensitivity analysis of the environmental assessment methodology...........................67 4.8 Methodology adaptations for assessment of engines for heavy-duty vehicles ...........69 4.9 Methodology adaptations for two-wheelers...............................................................72 4.10 Ecoscore calculation tool and Ecoscore website ...................................................75 4.11 Conclusions of the environmental assessment methodology ...............................78

PART II: Light Electric vehicles - Case study: The development of an electric bicycle for postal distribution ...................................................................................81

5 Light electric vehicles for postal delivery............................................................................................83 5.1 What are Light Electric Vehicles? ................................................................................83 5.2 Market of LEVs ............................................................................................................84 5.3 Different appearances of LEVs ....................................................................................84 5.4 Electric two-wheelers and electric bicycles .................................................................86 5.5 Bicycles for postal delivery: a brief history ..................................................................91 5.6 Electric bicycles: a brief history ...................................................................................93 5.7 Electric bicycles for postal delivery..............................................................................94 5.8 Other electric vehicles for postal delivery ...................................................................96

6 Postal requirements..........................................................................................................................101 6.1 Towards a better ‘last mile’.......................................................................................101 6.2 The NEPH project ......................................................................................................102 6.3 Questionnaire for postal operators ...........................................................................103 6.4 Examples of light electric vehicle experiences...........................................................107 6.5 Technical specifications: requirements or requests? ................................................109

7 Mathematical model for required mechanical power and energy ....................................................113 7.1 Goal ........................................................................................................................... 113 7.2 Required mechanical power......................................................................................113 7.3 Required mechanical Energy .....................................................................................119 7.4 Model for mechanical power and energy estimation based on postal requirements124

8 Drive technologies for electric bicycles .............................................................................................127 8.1 Introduction ..............................................................................................................127 8.2 Constituent power system components of electric bicycles......................................127 8.3 Sizing of the energy storage system ..........................................................................142 8.4 Integration of the electric power system in the bicycle ............................................147 8.5 Driving range of electric bicycles ...............................................................................152 8.6 The SAFT – Heinzmann electric power system ..........................................................154

9 Model to estimate the required energy capacity of the battery .......................................................163 9.1 Introduction ..............................................................................................................163 9.2 Features of the battery energy model.......................................................................163 9.3 Programme language ................................................................................................164 9.4 Main user interface ...................................................................................................164 9.5 Calculation methodology - Iteration algorithm .........................................................166 9.6 Programming structure .............................................................................................170 9.7 Battery energy model validation ...............................................................................170 9.8 Battery energy model results ....................................................................................172

List of Contents

p.xxiii

10 Optimization of the thermal management of a PMDC motor .........................................................179 10.1 Introduction.........................................................................................................179 10.2 Overloading an electric motor.............................................................................180 10.3 Description of the standard thermal protection..................................................180 10.4 Relating the maximum armature current to the motor’s temperature...............185 10.5 Comparative on-road tests ..................................................................................186 10.6 Conclusions on thermal management.................................................................188

11 Optimization of the energy consumption of the electric drive system ...........................................189 11.1 Introduction on the energy consumption optimization.......................................189 11.2 Description of the standard control principle......................................................190 11.3 Relating the armature current limit to the output voltage..................................191 11.4 Relating armature current limit to twist grip position and output voltage..........193 11.5 Modulating the maximum power limit to the available battery energy..............194 11.6 Comparative on-road tests ..................................................................................195 11.7 Test Results .........................................................................................................197 11.8 Conclusions on the energy consumption optimization........................................201

PART III: General conclusions and Future Work ................................................203

12 General conclusions ........................................................................................................................205 12.1 General conclusions PART I: ................................................................................205 12.2 General conclusions PART II: ...............................................................................206

13 Future work.....................................................................................................................................209 13.1 Future research in the field of environmental assessment of road vehicles .......209 13.2 Future research in the field of light electric vehicles for postal delivery .............210

Appendix 1..............................................................................................................................................213

Bibliography............................................................................................................................................217

Journal Publications................................................................................................................................227

Conference papers..................................................................................................................................229

Notes ......................................................................................................................................................233

Conversions ............................................................................................................................................238

List of Contents

p.xxiv

Jean-Marc

Rectangle

p.xxv

List of Figures Figure 1-1: Air pollution spatial and temporal resolution combinations.....................................................5 Figure 1-2: Different effects of air pollution ...............................................................................................6 Figure 1-3: Composition exhaust gases (volume percentage).....................................................................8 Figure 1-4: Evolution of the CO2 emissions of the Transport sector in Europe ...........................................9 Figure 1-5: Size range of airborne particles ..............................................................................................10 Figure 1-6: Observation of NO2 concentration (1015 molecules/cm²) in the troposphere (GOME satellite)..................................................................................................................................................................11 Figure 1-7: Emissions of NOX, NH3 and SO2 from the transport sector (based on MIRA-T 2007) ..............16 Figure 2-1: Life stages related to vehicle use ............................................................................................19 Figure 2-2: Direct and indirect emissions related to vehicle use............................................................... 20 Figure 2-3: Example of fuel production pathway: natural gas ..................................................................25 Figure 3-1: Sign for Low Emission Zone in London....................................................................................36 Figure 3-2: Picture of an air quality signalization panel in Heidelberg, Germany......................................37 Figure 4-1: Cleaner Drive-index calculation scheme ................................................................................40 Figure 4-2: Share of life cycle stages of the Life Cycle Assessment for the Environmental Certificate from Mercedes-Benz, absolute values on the right...........................................................................................43 Figure 4-3: Distribution of the emissions to the air for the different stages of a vehicle’s life cycle.........43 Figure 4-4: Classification of the pollutants to the different damage categories .......................................50 Figure 4-5: Transformation of Total Impact to Ecoscore...........................................................................56 Figure 4-6: Ecoscore Methodology Overview (step 1 & 2)........................................................................56 Figure 4-7: Ecoscore Methodology Overview (step 3, 4 & 5) ....................................................................57 Figure 4-8: Well-to-wheel energy consumption for passenger vehicles ...................................................58 Figure 4-9: Greenhouse gases for the vehicle selection – split up per pollutant and per stage ................60 Figure 4-10: Air quality depleting emissions for the vehicle selection – split up per pollutant and per stage .........................................................................................................................................................60 Figure 4-11: Air quality depleting emissions for the vehicle selection – with adapted vertical scale........61 Figure 4-12: Total impact for the vehicle selection – category split up.....................................................62 Figure 4-13: Distribution WTT and TTW environmental impact ...............................................................62 Figure 4-14: Overview range of Ecoscore for passenger vehicles .............................................................63 Figure 4-15: Total impact of different vehicle technologies and electricity production............................64 Figure 4-16: The Ecoscore of long term emissions abatement estimations for passenger cars, based on Smokers et al. (2004) ................................................................................................................................65 Figure 4-17: Ecoscore of the 20 best sold passenger vehicles and average Ecoscore ...............................66 Figure 4-18: Sensitivity of each impact contribution for the Ecoscore methodology ...............................68 Figure 4-19: Sensitivity analysis – maximum deviation on the end result.................................................69 Figure 4-20: Total impact and Ecoscore for the set of motorized two-wheelers – split up per category..74 Figure 4-21: Total impact and Ecoscore for mopeds, electric mopeds, electric motorcycles and electric bicycles .....................................................................................................................................................75 Figure 4-22: Welcome screen of the Ecoscore calculation tool ................................................................75 Figure 4-23: Overview screen of the selected vehicle’s data .................................................................... 76 Figure 4-24: Results screen of the Ecoscore calculation tool ....................................................................76 Figure 4-25: Parameter setting screen of the Ecoscore calculation tool – tab damage calculation ..........77 Figure 4-26: Ecoscore website screenshot (www.ecoscore.be)................................................................78 Figure 4-27: Ecoscore logo........................................................................................................................79 Figure 5-1: Wavecrest Tidal Force M750 at EVS20, California 2003..........................................................85

List of Figures

p.xxvi

Figure 5-2: Twike at the EVS24 in Norway, 2009 ......................................................................................85 Figure 5-3: Veloform CityCruiser I in Brussels, © Veloform ......................................................................85 Figure 5-4: Venturi Eclectic at the 2006 Paris Motor Show © Florian.......................................................86 Figure 5-5: Picture of the e-solex ©Frederic Dinh ....................................................................................87 Figure 5-6: Yamaha Passol electric scooter...............................................................................................87 Figure 5-7: Applied mounting places for the electric motor ..................................................................... 87 Figure 5-8: Power addition at the pedal axis.............................................................................................88 Figure 5-9: Power addition at the transmission ........................................................................................88 Figure 5-10: power addition at the level of the rear wheel ......................................................................88 Figure 5-11: Power addition at the level of the road ................................................................................88 Figure 5-12: “Le porteur de dépêches à Bruxelles”, 1887[148].................................................................92 Figure 5-13: Mail delivery from a penny-farthing, 1900[149] ...................................................................92 Figure 5-14: “Hen and Chickens” Pentacycle or Centre-cycle, circa 1882 [150]........................................92 Figure 5-15: Tricycle with forward basket carrier, 1934 ...........................................................................92 Figure 5-16: Group of cycling postmen ready to start their round in the early 1920s [151] .....................93 Figure 5-17: Postal delivery bicycle, Cologne 2008 ...................................................................................93 Figure 5-18: SpeedBike AL26 postal delivery bicycle ................................................................................93 Figure 5-19: Electric post bicycle in Berlin, 2007 ○c Lhoon .........................................................................95 Figure 5-20: Electric bicycle for postal delivery for La Poste [160]............................................................95 Figure 5-21: NEPH prototype Granville electric bicycle for postal delivery, 2007 ©Philippe Lataire ........96 Figure 5-22: Wattworld electric bicycle for the Belgian Post ....................................................................96 Figure 5-23: The Segway® i2 Cargo Personal Transporter at PostExpo 2006 ............................................97 Figure 5-24: Distribution trolley with auxiliary electric motor from Asoma..............................................97 Figure 5-25: E-trailer in combination with a (mechanical) postal bike at Deutshe Post............................98 Figure 5-26: Oxygen electric scooter for Belgian Post at PostExpo 2006 ..................................................98 Figure 5-27: “Free Duck” electric quadricycle evaluated by Poste Italiane [163]......................................99 Figure 5-28: Modec electric van at PostExpo 2006 ...................................................................................99 Figure 5-29: Smith electric vehicle truck at PostExpo 2006 ......................................................................99 Figure 6-1: Logo of the NEPH project......................................................................................................102 Figure 6-2: Eureka-label ..........................................................................................................................102 Figure 6-3: Cumulative number of rounds with a given length of active and non-active part of the distribution round...................................................................................................................................106 Figure 6-4: Relative altitude and calculated cumulative climbing height difference of an extreme caddy delivery trip.............................................................................................................................................106 Figure 6-5: Picture of the Roodrunner, Springtime used by TPG Post.....................................................108 Figure 6-6: Electric caddy for postal distribution from Expresso GmbH..................................................109 Figure 7-1: Hill-profile – representation of the rise and run of a slope ...................................................114 Figure 7-2: Power needed to overcome wind resistance in function of the bicycle speed v at various values for the headwind vw.....................................................................................................................116 Figure 7-3: Illustration of the term cumulative height difference...........................................................120 Figure 7-4: Decreasing payload over trip distance..................................................................................121 Figure 7-5: Post office on top of hill ........................................................................................................122 Figure 7-6: Delivery area on top of hill....................................................................................................122 Figure 7-7: Evenly spread climbing profile ..............................................................................................122 Figure 7-8: Influence of the distribution of the height difference on the required mechanical energy ..123 Figure 7-9: Analysis of the required mechanical energy as function of different trip parameters..........124 Figure 7-10: Sensitivity analysis of the model for required mechanical energy ......................................125 Figure 8-1: Ragone plot for different electrochemical energy storage systems (cell level).....................129 Figure 8-2: cycle life as function of the depth-of-discharge (example of lead-acid)................................129 Figure 8-3: Accel Pro Tri-CarGo powered with a direct methanol Fuel Cell ............................................132 Figure 8-4: The UltraCap e bike prototype of KaHo Sint-Lieven..............................................................133 Figure 8-5: Overview of battery pack and bicycle masses from commercial available electric bicycles..134 Figure 8-6: Selection of different assistance modes (left) and modulation of electrical assistance with a thumb switch (right) ...............................................................................................................................137 Figure 8-7: Display of the Sparta ION electric bicycle .............................................................................138 Figure 8-8: Pedal sensor of the Heinzmann drive system [142] ..............................................................139 Figure 8-9: Contactless pedal torque sensor [190] .................................................................................139

List of Figures

p.xxvii

Figure 8-10: Energy consumption and charging time for a set of 12 electric bicycles.............................140 Figure 8-11: Bike Tree - bike sharing and storage system ...................................................................... 141 Figure 8-12 Japanese underground bicycle parking system....................................................................141 Figure 8-13: Sanyo Solar Parking Lot for electric bicycles .......................................................................141 Figure 8-14: Influence of the energy storage technology on the required mechanical energy...............142 Figure 8-15: Electric propulsion system efficiencies ...............................................................................145 Figure 8-16: Mechanical power addition at different levels of the power train .....................................148 Figure 8-17: Velocity Dolphin with dual drive system – type B ...............................................................148 Figure 8-18: Power adding at the level of the chain (Panasonic motor unit) – type C ............................149 Figure 8-19: Pedal axis motor from BikeTec – type D, © Philippe Lataire...............................................149 Figure 8-20: Human powered series hybrid configuration......................................................................150 Figure 8-21: Battery inside the frame tube of the Sparta ION electric bicycle ........................................151 Figure 8-22: SAFT flat battery packs (pilot version) ................................................................................151 Figure 8-23: SAFT prismatic battery pack (pilot version).........................................................................151 Figure 8-24: Overview of range and energy capacity for a set of 12 electric bicycles .............................153 Figure 8-25: Saft – Heinzmann electric power system ............................................................................ 155 Figure 8-26: Heinzmann hub motor - exploded view (translated from [207]) ........................................155 Figure 8-27: Twist grip Heinzmann power unit .......................................................................................156 Figure 8-28: SAFT Smart-VH-battery module (NiMH) [178] ....................................................................157 Figure 8-29: energy discharged for pulsed discharge current at different rates .....................................158 Figure 8-30: NEPH batteries capacity margin..........................................................................................159 Figure 8-31: parallel connection of the battery packs.............................................................................159 Figure 8-32: NEPH batteries parallel or sequential discharging ..............................................................160 Figure 8-33: Schematic representation of the NEPH power system .......................................................161 Figure 8-34: NEPH bicycle prototype LUDO ............................................................................................161 Figure 8-35: NEPH tricycle prototype from MIFA A.G. ............................................................................ 162 Figure 9-1: Battery energy calculation ....................................................................................................164 Figure 9-2: The front panel of the energy calculation tool......................................................................165 Figure 9-3: Output screen of the energy calculation tool (3D plot) ........................................................166 Figure 9-4: Iterative battery energy calculation......................................................................................167 Figure 9-5: SAFT Flat battery pack combinations (Pilot and Industrial versions).....................................168 Figure 9-6: Stepwize battery weight function for the NEPH Pilot Flat batteries .....................................168 Figure 9-7: Nested loops for parameter calculation ...............................................................................170 Figure 9-8: Comparison of calculation tool with measurements for start and stop operation ...............171 Figure 9-9: Comparison of calculation tool with measurements for continuous driving operation........172 Figure 9-10: Required battery capacity as a function of the number of stops and of the total trip distance................................................................................................................................................................174 Figure 9-11: Battery pack capacity as a function of the number of stops and of the total trip distance.175 Figure 9-12: Output screen of the energy calculation tool (2-D plot of parameter 1).............................175 Figure 9-13: Depth of discharge as a function of the total trip distance and of the number of stops.....176 Figure 9-14: Battery pack weight as a function of the total trip distance and of the speed achieved between 2 stops .....................................................................................................................................177 Figure 9-15: Required battery energy as a function of the payload and of the number of stops............178 Figure 10-1: Picture of the Heinzmann PMDC wheel motor ...................................................................181 Figure 10-2: Characteristics of the PMDC motor at rated armature voltage (warmed up) .....................181 Figure 10-3: Example of an excess armature current profile ..................................................................182 Figure 10-4: The effect of an excess current on the motor temperature................................................183 Figure 10-5: Evolution of the motors’ temperature during an on-road test simulating postal distribution................................................................................................................................................................184 Figure 10-6: Picture of the test campaign in Neder-Over-Heembeek (January 2008), © Francis Heymans From Left to Right: Jean-Marc Timmermans, Prof. Philippe Lataire, Jens Nietvelt, Jean Vander Elst (Bike Events), Dr. Klaus Gössel (Heinzmann GmbH) ........................................................................................184 Figure 10-7: Armature current limit as function of the motors’ temperature ........................................185 Figure 10-8: Comparison of motors’ temperature evolution ..................................................................186 Figure 10-9: Plots of vehicle speed, altitude and heartbeat rate in case of thermal management switched off ...........................................................................................................................................................187

List of Figures

p.xxviii

Figure 10-10: Plots of vehicle speed, altitude and heartbeat rate in case of thermal management switched on ............................................................................................................................................187 Figure 11-1: Controller output characteristics when turning the pedals ................................................190 Figure 11-2: Controller output characteristics without turning the pedals.............................................191 Figure 11-3: Controller output characteristics when turning the pedals (solution 1) .............................192 Figure 11-4: Controller output characteristics without turning the pedals (solution 1)..........................192 Figure 11-5: Controller output characteristics when turning the pedals (solution 2) .............................193 Figure 11-6: controller output characteristics without turning the pedals (solution 2) ..........................193 Figure 11-7: Energy made available versus time (increment of Eallowed is stopped when driving is stopped for more than Ta )....................................................................................................................................194 Figure 11-8: Prototype used for the on-road comparative tests.............................................................196 Figure 11-9: Energy consumption versus time for the three controllers ................................................197 Figure 11-10: Histogram of power spectrum of Pbat, when using controller number 1...........................198 Figure 11-11: Histogram of power spectrum of Pbat, when using controller number 2...........................198 Figure 11-12: Histogram of power spectrum of Pbat, when using controller number 3...........................199 Figure 11-13: Relative altitude versus trip distance................................................................................200 Figure 11-14: Relative heartbeat in function of the trip distance ...........................................................200

List of Figures

p.xxix

List of Tables Table 1-1: Limit values according to the European directive 2008/50/EC ................................................13 Table 1-2: Air quality standards according to the World Health Organization..........................................13 Table 1-3: Examples of GWP (updated values between brackets), 100 years time horizon......................15 Table 2-1: Maximal sulphur content of fuels for road vehicles .................................................................23 Table 2-2: Nitrous oxide (N2O) emission values ........................................................................................23 Table 2-3: Emission factors related to the production phase of different fuels for road vehicles............25 Table 2-4: Maximum sound levels for road vehicles .................................................................................26 Table 2-5: Maximum sound pressure levels for motorized two- and three-wheelers ..............................26 Table 4-1: Overview BIM - Clean Vehicles methodology........................................................................... 42 Table 4-2: Expended energy for different fuel production options and electricity production.................46 Table 4-3: Fuel characteristics used in the Ecoscore methodology...........................................................47 Table 4-4: Indirect emission factors Fj for different fuel options ..............................................................49 Table 4-5: Impact factors for the global warming category ...................................................................... 52 Table 4-6: Specific external costs (SEC) for human health effects.............................................................53 Table 4-7: Overview of average mileage distribution for different vehicle categories (anno 2004) .........53 Table 4-8: Impact factors δ for human health effects...............................................................................53 Table 4-9: Impact factors δ for effects on ecosystems..............................................................................54 Table 4-10: Tank-to-Wheel and Well-to-Tank emissions of the reference vehicle for passenger cars and light duty vehicles .....................................................................................................................................54 Table 4-11: Summary of the parameters used for the Ecoscore methodology.........................................57 Table 4-12: Vehicle selection details for Well-to-Wheel energy consumption analysis ............................58 Table 4-13: Vehicle selection details for environmental assessment methodology analysis ....................59 Table 4-14: N2O and CH4 emission ratio’s for different fuel types ............................................................70 Table 4-15: Tank-to-Wheel and Well-to-Tank emissions of the reference situation for engines for heavy-duty vehicles .............................................................................................................................................70 Table 4-16: Summary of the parameters used for the adapted Ecoscore methodology for MIVB-STIB....71 Table 4-17: Emission limits for motorized two-wheelers..........................................................................72 Table 4-18: Vehicle selection details and COPERT categories for motorized two-wheelers....................73 Table 4-19: Tank-to-Wheel and Well-to-Tank emissions of the reference vehicle for two-wheelers .......73 Table 4-20: Details of selected motorized two-wheelers, including electric bicycles ...............................74 Table 5-1: advantages and disadvantages of the different power train topologies of electric bicycles ....89 Table 6-1: Advantages and drawbacks of conventional vehicles for postal delivery...............................104 Table 6-2: Advantages and drawbacks of currently used electric vehicles for postal delivery................104 Table 6-3: Common level of interest for light electric vehicles for postal delivery .................................107 Table 6-4: Technical requirements for NEPH electric bicycle..................................................................110 Table 6-5: Technical specifications for NEPH electric caddy ...................................................................110 Table 6-6: Technical specifications for NEPH electric 3-wheeler.............................................................111 Table 8-1: Overview of the subjective assessment of the electrical assistance ......................................154 Table 8-2: SAFT Mobility batteries characteristics ..................................................................................157 Table 9-1: Battery pack combinations SAFT Flat Pilot (36 Volts).............................................................173 Table 11-1: Energy consumption comparison.........................................................................................197

List of Tables

p.xxx

Jean-Marc

Rectangle

p.1

0 General introduction

0.1 Environmental assessment of road vehicles Environmental pollution is an issue that causes great concern, not only on a local scale, but also at international level. Tackling the important problem of air pollution forms a unique challenge for mankind. Based on global observations of important tropospheric pollutants, as described in the literature, the region formed by the north of Belgium (Flanders) and the south of the Netherlands appears to be one of the most polluted in the world, after northern Italy and the north-eastern part of China [1]. Persons and goods transportation cause an important part of the emissions of atmospheric pollutants [2]. Therefore it is essential to understand the correlation between transport and environment in order to be able to tackle transportation’s negative impacts [3]. To reduce harmful emissions due to the transport sector, efficient policy measures have to be implemented by the relevant authorities, especially in strongly urbanized regions. The introduction of ‘cleaner vehicles’ is one of the most promising potential measures policy makers have at their disposal for energy use reduction and for cutting air pollutant emissions [4]. In this context the question: “Which vehicles are the most environmentally friendly?” remains a key issue [5]. To be able to answer this question, a comprehensive and transparent methodology has been developed with the aim to compare the environmental burden caused by vehicles with different drive trains and using different fuels. The first part of this thesis describes the main environmental damages related to the use of road vehicles. Next, the different sources of emissions related to the use of road vehicles are described. The main goal of this first part of the presented research was to develop a new and pragmatic environmental rating tool for the assessment of the environmental performances of road vehicles. A methodology based on a well-to-wheel basis has been developed. This methodology finally results in a pragmatic rating tool that allows comparing road vehicles with different drive trains or fuel use. This tool is called Ecoscore and was developed between 2003 and 2005 for use by the Flemish government. Continuous efforts are made to further improve the applicability of the Ecoscore methodology for use in policy measures and to extend it to new types of fuels and new kinds of vehicles.

0.2 Light electric vehicles for postal delivery Light electric vehicles (LEV) are a promising new kind of device to extend the range of personal mobility vehicles [6]. But LEVs are also a promising alternative for vehicles goods delivery (e.g. postal service). Personal electric vehicles can offer several potential benefits to consumers and to society including lower running costs [7], reduced trip times, and a lower environmental impact. As they are potentially suitable for mail delivery, a clear interest for LEVs exists among the European postal operators. In particular electrically assisted bicycles for postal delivery seem to have a very high potential. This is also

General introduction

p.2

confirmed by the fact that several European postal operators have launched calls for tenders to purchase electrically assisted bicycles, tricycles, caddies and trolleys to implement in their postal vehicle fleet. However, at the start of this research most products (electric bicycles) available on the market did not satisfy the high requirements of the (European) postal organisations. From this observation, a European consortium was created with the goal of developing a range of power systems for use in light electric vehicles for postal delivery. The Vrije Universiteit Brussel was the academic partner of this consortium. The main goal of the research work was to translate the postal requirements into specifications for the sizing of the electric power system to be used. For this purpose, a calculation tool has been developed that uses an approximate calculation method to estimate the required electrical energy of the power system starting from typical postal operation parameters. In case of highly demanding use of the electric postal delivery bicycle, the heating of the wheel motor causes the motor temperature to exceed the maximal allowable temperature. For this purpose, a thermal management system for the electric motor was implemented allowing maximizing the vehicle performance without causing sudden system shut-down. Finally, different solutions are proposed for lowering the energy consumption in case of a “power-on-demand” electric power system.

0.3 Outline The text is organised in three parts. The first part of this dissertation deals with the environmental assessment of road vehicles. Part I contains 4 chapters:

• Chapter 1 describes the main environmental damages related to the use of road vehicles • In chapter 2 the different sources of emissions related to the use of road vehicles are

presented • In chapter 3 an overview is given of different possible solutions to reduce the environmental

damage caused by road vehicles • Chapter 4 gives the description of an environmental assessment tool for road vehicles that

has been developed at the department ETEC The second part of the dissertation deals with light electric vehicles, which are considered as a part of the solution for the problems related to the current use of road vehicles. In particular, this part of the text describes the work that has been done for the development of an electric bicycle for postal distribution. Part II contains 7 chapters:

• In chapter 5 light electric vehicles are introduced, with special attention for their application for postal delivery

• Chapter 6 presents the postal requirements in view of using light electric vehicles for delivery • In chapter 7 a mathematical model is presented that translates the postal requirements in

terms of required mechanical power and energy • Chapter 8 gives a description of the available drive technologies for use in light electric

vehicles. In particular, the relation between the required mechanical energy and the corresponding electrical energy from the electric power system is described

• Chapter 9 presents the description of a model to estimate the required energy capacity of the battery starting from the postal requirements and considering a specific drive technology

• In chapter 10 the optimisation of the thermal management of an electric power system for use in electric bicycles for postal delivery is discussed.

• In chapter 11 different solutions are presented to optimize the energy consumption of this electric power system

In the third and last part of this text the final conclusions are made:

• In chapter 12 the general conclusions of both part I and part II are expressed • In chapter 13 a proposal for future work is given

General introduction

p.3

PART I: Environmental assessment of road vehicles

Jean-Marc

Rectangle

p.4

Jean-Marc

Rectangle

p.5

1 Environmental damages related to the

use of road vehicles

1.1 Environmental effects of atmospheric pollution from road vehicles In this paragraph a number of main considerations about atmospheric pollution will be highlighted. A first element is the spatial resolution, or the scale, at which the effects caused by the pollutants take place. The effects of air pollution, caused by the exhaust from road transportation, cover the whole range of spatial scales, ranging from the effects on a local level to worldwide effects. On a local scale, pollution mainly affects public health. Secondly, on a regional scale, pollutants affect buildings and ecosystems through many chemical transformations of the pollutants (acid rain, photochemical reactions) and again human health. At a global level, pollutants contribute to climate change and depletion of the ozone layer. A second characteristic is the temporal resolution. Some pollutants increase at the incidence of short term peak concentrations, while others act over periods of many months or even years [8]. These characteristics are illustrated with some examples of combinations of spatial and temporal resolution in the figure 1‐1 below. This picture is based on reference [8].

Figure 1‐1: Air pollution spatial and temporal resolution combinations

Environmental damages related to the use of road vehicles

p.6

In this figure the following examples are illustrated:

• At a local level (or urban level) the negative effects of carbon monoxide (CO), nitrogen oxides (NOX), volatile organic compounds (VOC) and particulate matter (PM) on human health require knowledge of the emissions in short time frames (hours to days).

• At a global level the effect of emissions of greenhouse gases (carbon dioxide, methane, nitrous oxide…) as well as of substances that are damaging the stratospheric ozone layer (nitrous oxide… ) require knowledge of emission levels in larger time frames (years).

Besides these examples other combinations exist. At a local and at a regional level we can identify acidification, caused mainly by nitrogen oxides and by sulphuric oxides and we can identify photochemical pollution, related to the emissions of carbon monoxide, nitrogen oxides and volatile organic compounds from which ozone is formed on warm days. These pollutants have direct and indirect effects in a time frame of days to months. The direct and indirect effects of the different pollutants caused by transportation on the numerous receptors (humans, ecosystems…) result in a whole range of different and complex phenomena. This is illustrated in the figure 1-2 below, showing a set of relations between pollutants and receptors through different kinds of effects. This picture is based on reference [9].

Primary elements

Secondary elements:Nitrates, sulfates

Organic compounds…

Photochemical oxidantsOzone, aldehydes

Peroxy acetyl nitrate …

Nitrogen oxides

Sulfur oxides

Carbon monoxide

Volatile OrganicCompounds

Poly-aromaticHydrocarbons

Heavy Metals

Carbon dioxide

POLLUTANTS

Contamination

Corrosion

Limitation of visibility

Impact on

human health

Eco-toxicity

Acidification

Eutrophication

Climatechange

Metals

EFFECTS RECEPTORS

Humans(Psychic)

Humans(biological)

Cultures

EcosystemsIn water

Ecosystemson land

Ecosphere

Figure 1-2: Different effects of air pollution

Some pollutants, such as carbon dioxide (CO2) are chemically inert and tend to accumulate in the atmosphere, causing modification of its physical properties (concentration of greenhouse gases). Other pollutants are chemically active and can react with substances they encounter (corrosion of metals, toxicity for living organisms…) or with other pollutants within the atmosphere leading to new pollutants (secondary pollutants). The latter is true for carbon monoxide (CO), nitrogen oxides (NOX) and for volatile organic compounds (VOC) that chemically evolve in the troposphere. They are activated by sunlight and are at the basis of photochemical pollution, characterised by the production of ozone (O3) and other dangerous substances for human health and for the environment (peroxyacetyl nitrate or PAN, aldehyde, nitrate acids, hydrogen peroxide and others) [10].


p.7

Trough different processes (absorption or dilution of gases, sedimentation of particles, etc.) different exhausted compounds are migrating to the surface causing potential disturbance of the biochemical balance of the exposed environment (acidification or eutrophication1 of ecosystems, etc.). Metals and some non-biodegradable organic compounds are ending up in the food chain and are piled up in living structures. These can cause poisoning of certain organisms. From MIRA-T [2], the environmental report of Flemish Region, it appears that with unchanged concentrations of particulate matter (PM10 and PM2,5) all residents of the Flemish Region will lose one third of a healthy life year, expressed as 1/3 DALY2 (disability-adjusted life year). The fraction of particulate matter of 2,5 micrometer or less (PM2,5), that is highly concentrated in areas with dense traffic, becomes more and more important due to better knowledge of their effects on human health. This importance is related to the location of sedimentation in the respiratory system. The fraction greater than 2,5 micrometer is depositing in the upper part of the respiratory system, whereas the smaller parts (PM2,5) are penetrating to the bronchi and the alveoli of the lungs. Larger particles are therefore related to syndromes occurring in the upper part of the respiratory system (e.g. asthma), while as the smaller particles are associated to more severe effects, such as cardiovascular effects resulting in hospitalisation or even death. Also the number of particles and their composition are playing a role in the effects on human health. Toxicological research points out that the presence of carcinogenic elements and metals on fine particles, soot and ultra fine particles or UFP (particles smaller than 0,1 μm) is causing the health-related impact of particulate matter. The contribution of the sector of transportation to these compounds is considered to be important and thus is assumed to be an important source of health impairing effects. Some gases that are present in the atmosphere let through solar radiation, but are absorbing the heat reflected by the Earth in such a way that a liveable climate is being created in the lower part of the atmosphere. Without these greenhouse gases (sometimes abbreviated to GHG), life on Earth would not be possible. Due to anthropogenic activity, responsible for the exhaust of greenhouse gases, this equilibrium is being disturbed. Climate change, rise of the sea levels, expanding deserts and the disappearance of biotopes are some of the consequences. These effects can have serious implications at the long term and at a global level. The most important greenhouse gases related to human activities are: carbon dioxide (CO2), chlorofluorocarbons (CFCs), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulphur hexafluoride (SF6), methane (CH4), nitrous oxide (N2O) and ozone (O3) [11]. Acidification is caused by the binding of sulphur (S) and nitrogen (N) that were emitted into the atmosphere. Sulphur oxide in the atmosphere is transforming mainly to sulphuric acid (H2SO4) after a few hours or days, usually in the presence of a catalyst such as NO2. Nitric oxide (NO) is being converted (via oxidation) into nitrogen dioxide (NO2) and consequently converted into nitric acid (HNO3). Ammonia (NH3) is released into the atmosphere as a result of agrarian activity. The emissions of SO2, NOX and NH3 are often expressed in ‘acid equivalents’ such as SO2 - equivalents. An important health effect of road transportation is the formation of photochemical smog or summer smog. Certain volatile organic compounds (VOCs) and nitrogen oxides (NOX) react under the influence of solar radiation and high temperatures to form ozone (O3). The concentration of ozone can rise significantly in warm periods. High concentrations of ozone cause irritation of the throat, nose and eyes, respiratory difficulties can lead to accelerated ageing of the lungs. Especially young children and elderly people are sensitive to ozone.

1 Eutrophication is an increase in the concentration of chemical nutrients in an ecosystem to an extent that it increases the primary productivity of the ecosystem with subsequent negative environmental effects such as reduction in water quality. 2 DALY is a quantification for the loss of one year of “healthy” life. It is calculated as the aggregation of the years of life lost due to premature mortality in the population and the years lost due to disability for incident cases of the health condition.


p.8

The majority of air pollutants have a direct impact on human health. This negative effect on health is uttered in different forms: serious health effects (toxic compounds, effects on the respiratory system and respiration, neurologic effects) and chronic effects (carcinogenic compounds, chronical conditions of the respiratory system, effects on the cardiovascular system, development of allergies and asthma). The current state of knowledge is insufficient to provide a complete insight in the different systems of air pollutants and their different effects. However, it is possible in some cases to correlate an increased concentration of a pollutant in a certain environment with the effects on certain receptors. These relations are being quantified by the use of dose-response functions. The effects of air pollution caused by transportation are significantly higher in urban areas. This can be explained by the combination of two elements: a significant density of sources of pollutants on the one hand and high density and proximity of the receptors (humans, buildings etc.) on the other hand. Some of the studies that were incorporated into the framework of the European project ‘ExternE’, devoted to the evaluation of external costs of the energy and transportation sectors, have shown that the impacts on the local level are dominant in the complete set of damages in case of pollution caused by road vehicles [12, 13].

1.2 Foremost atmospheric pollutants from road vehicles Air pollution, related to the use of (conventional) road vehicles, is caused (mainly) by the combustion of the fuel in the internal combustion engine or ICE. For the combustion, the vehicle requires oxygen from the surrounding air. Pure air is composed of 20,947% of oxygen but mainly contains nitrogen (78,084%). Further, pure air contains 0,934% of argon (Ar) and 0,0314% of carbon dioxide (CO2) and a few other gases (0,004%). The combustion of fossil fuel results in the production of carbon dioxide. However, only about 14% of the exhaust gases of a passenger vehicle are carbon dioxide. Besides carbon dioxide, also about 70% nitrogen gas and 12% water vapour is released [14]. In figure 1-3 (based on [15]) the composition (percentages) of the exhaust gases from a petrol vehicle is shown. The numbers in this graph correspond with the earlier mentioned percentages. In the paragraphs below, the most relevant pollutants from road vehicles are shortly described.

Figure 1-3: Composition exhaust gases (volume percentage)


p.9

1.2.1 Carbon dioxide Carbon dioxide is the typical product of the combustion process of hydrocarbons. The exhaust of carbon dioxide is therefore proportional to the fuel consumption of a vehicle. Carbon dioxide is colourless and odourless. Carbon dioxide is however an important greenhouse gas that contributes to global warming (see paragraph 1.4.2 on page 14). When looking at the evolution of the CO2 – emissions of the transport sector in Europe (see figure 1-4) an important increase of the emissions, compared to the base year 1990 can be identified. Further we learn that the transport sector is responsible for 23% of the total CO2 emissions. The figures used for this graph are based on the EU energy and transport in figures - statistical pocketbook 2009 [16].

Figure 1-4: Evolution of the CO2 emissions of the Transport sector in Europe

1.2.2 Carbon monoxide Carbon monoxide (CO) is a colourless, odourless and tasteless gas, which in high concentration is highly toxic to humans and animals. Carbon monoxide causes oxygen shortage and can lead to suffocation. CO will bind with the haemoglobin in the blood to form carboxyhaemoglobin, which is ineffective for delivering oxygen to bodily tissues. The most common symptoms are headache, feeling of weakness, dizziness and vomiting. With increasing concentration of carboxyhaemoglobin in the blood, the effects are feeling sleepy, disturbance of motoric function and ultimately coma. When more than 66% of the haemoglobin has been bound, this will cause death. Carbon monoxide poisoning is the most common type of fatal air poisoning in many countries [17]. The concentration of CO in the atmosphere can vary strongly and depends mainly on the traffic situation and on the wind. Carbon monoxide concentrations in the blood of people living in the city can be twice as high as concentrations in the blood of people living in the countryside. The concentration of CO can reach very high levels in closed spaces like underground parking lots, or inside vehicles [18]. Intoxication from carbon monoxide sometimes occurs in badly ventilated residences with old or badly maintained heating systems.


p.10

1.2.3 Particulate matter A differentiation is made between fractions of particulates, based on their size. Most commonly used is the aerodynamic diameter and which is related to where in the respiratory tract the particle will deposit, if inhaled. The notation PM10 is used to denote the fraction of 10 micrometers or less. By analogy, PM2,5

represents the fraction of particles with an aerodynamic diameter of less than or equal to 2,5 micrometer. Particles below 100 nanometres are classified as ultrafine particles or UFP. An overview of the range of airborne particles and their sizes is given in figure 1-5. This figure is based on reference [19]. Particulate matter (PM10) can remain in suspension in the air for hours or even days. As these particles are very small, they can penetrate very deeply into the lungs. Because PM10 is often bound to other harmful substances, exposure and inhalation can lead to an important number of health problems such as asthma, lung cancer, cardiovascular problems and even premature death [13, 20, 21]. Because of their small dimensions the smallest fraction of PM10 can penetrate into the deepest parts of the lungs where they cause health problems [21, 22]. Even at relatively low concentrations (<50μg/m³) health impairing effects occur [23]. Furthermore, road traffic also causes particles that were deposited on the roads to be blown and carried away. These particles are not only originating from the tailpipe of the vehicles, but also from the materials used in the roads and from the vehicle’s tires and brakes. Recent research pointed out that the metals on brake wear particles cause damage to the lungs and also increase inflammatory responses [24].

Figure 1-5: Size range of airborne particles

1.2.4 Sulphur oxides Sulphur oxides (SOX) are very soluble in water and consequently can easily be absorbed through the mucous membrane of the bronchial tubes. Sulphur dioxide (SO2) is a colourless gas, which can cause irritations. People suffering from asthma are especially sensitive to sulphur oxides [21]. Fossil fuels often contain sulphur compounds and consequently their combustion generates sulphur dioxide. The sulphur content for fuels designated for road vehicles is regulated by European directive 98/70/EG. Since January 2005 the maximum was set at 50ppm. Since January 2009 this has been further lowered to 10ppm. Natural sources of SO2 are volcanoes and oceans. In the air sulphur dioxide is converted into sulphuric acid (H2SO4) and together with nitric acid it also contributes to acid rain (see paragraph 1.4.4 on page 15). Furthermore, sulphur dioxide causes acidification, which in turn damages aquatic and terrestrial ecosystems, surface water, agricultural and forestry yields and buildings [9].


p.11

1.2.5 Nitrogen oxides Nitrogen oxides (NOX) are forming a group of important pollutants mainly produced by road transportation. They are generated via combustion at high temperatures. Nitrogen dioxide acts as an indicator of the air quality and plays an important role in the formation of ozone (1,22 TOFP1) [25]. It is responsible for the formation of photochemical smog and acid rain (see paragraph 1.4.1 on page 14 and paragraph 1.4.4 on page 15 for more details). Nitrogen oxides are forming acids, when absorbed in the mucous membranes of the nose or the oral cavity, and cause irritation of the bronchial tubes, coughing, and in higher concentrations also lack of breath and even death [13]. People suffering from asthma can be very sensitive for NOX [26]. Nitrogen oxides also lead to eutrophication of soil, ground- and surface water and as a result lead to negative impacts on aquatic and terrestrial ecosystems, surface water and agricultural and forestry yields [9]. Based on global observations of tropospheric NO2, the densely populated area covering the northern part of Belgium and the southern part of The Netherlands and Rurh (Germany) appears to be one of the most polluted regions of the world, ranking third after Northern Italy and the north-eastern part of China [1]. An image of such an observation of Europe [1] can be seen in figure 1-6.

Figure 1-6: Observation of NO2 concentration (1015 molecules/cm²) in the troposphere (GOME satellite)

1.2.6 Other pollutants Besides the pollutants described earlier, some other less known pollutants related to the use of road vehicles can be mentioned:

• Volatile organic compounds (VOCs), is the name grouping a large number of chemical compounds, like toluene, xylene, benzene, etc. Some of them cause important health damaging effects such as carcinogenicity. Benzene is added to unleaded petrol as an anti-knock additive. Exposure to benzene is occurring while refilling petrol cars. Most important effects of benzene are its impact on the central nerve system, fatigue, headache, dizziness, lowered muscular strength and insomnia. Further it is considered as carcinogenic and increases the chance of leukaemia [27].

• Polycyclic aromatic hydrocarbons (PAHs) are generated by incomplete combustion of organic

compounds. PAHs are present under the form of mixtures and most PAHs are absorbed by particles. Their concentration in the air strongly depends on the location. In urban areas road

1 Total Ozone-Forming Potential


p.12

transportation is an important source of these PAHs. Six of the PAHs that are present in the exhaust gases of vehicles are classified as probably carcinogenic.

• 1,3-Butadiene is a strongly reactive component, present in the exhaust emissions of road

vehicles. The emissions of 1,3-butadiene are proportional to the emissions of the other hydrocarbons and should lower with the use of catalysts for road vehicles. It is classified as probably carcinogenic [27].

• If methanol and ethanol are to be blended with petrol, it would lead to the emission of

aldehyde and ozone precursors, through partial oxidation of these alcohols. The latter would then cause an increased impact on human health due to ozone formation.

• The use of tetra ethyl lead (TEL) in the past as an anti-knock additive caused an important

amount of lead was emitted into the environment. Epidemical research has shown that lead is highly toxic even at low concentrations. Known effects are the negative influence on the IQ of children [28] and problems with hearing [29]. Since the 1980s, campaigns for lowering lead in fuels have lead to a significant decrease of the exposure of humans to lead. Since January 2000, lead was completely banned from fuels used by road vehicles in Europe (Directive 98/70/EC) and is replaced by other octane enhancers. MMT (methylcyclopentadienyl manganese tricarbonyl) is currently used as an anti-knock additive for petrol. However possible negative effects of MMT on human health [30] have lead to new restrictions of these additives in fuels (cfr. 2009/30/EC).

Some of the previously described emissions also have indirect effects as they lead to other chemical reactions, once emitted into the atmosphere. The formation of tropospheric ozone for instance is triggered by sunlight and is an important secondary pollutant of nitrogen oxides and hydrocarbons emissions (photochemical pollution). Ozone (O3) is toxic for living organisms and human beings and causes damage to the cells of the bronchial tubes [31, 32].

1.3 Regulations for air quality In this paragraph a brief overview of the European directives related to Air Quality will be discussed. With regard to air quality, the following European directives can be mentioned:

• Directive 92/72/EEC and amendment 2002/3/EC on air pollution by ozone • Directive 80/779/EEC and amendment 89/427/EEC setting limit concentration for SO2 and

PM • Directive 82/884/EEC setting limits for the Lead concentration • Directive 85/203/EEC and amendment 85/580/EEC for N2O concentration limits • The previous three directives have been adopted by the 1999/30/EC directive related to limit

values for sulphur dioxide, nitrogen dioxide and oxides of nitrogen, particulate matter and lead in ambient air

• Directive 2000/69/EC related to limit values for carbon monoxide and benzene in the ambient air.

• Directive 2004/107/EC relating to arsenic, mercury, nickel and polycyclic aromatic hydrocarbons in ambient air

• Directives 96/61/EC and 96/62/EC concerning ambient air quality and management and concerning integrated pollution prevention and control

• The new directive 2008/50/EC on ambient air quality and cleaner air for Europe, (incorporating the directives 1999/30/EC, 2000/69/EC and 2002/3/EC)


p.13

A summary of some of the most important limit values for pollutants are given in the table 1-1 below.

Table 1-1: Limit values according to the European directive 2008/50/EC

Pollutant Period Limit value Allowed number of exceeding Entry into force SO2 1 hour 350 μg/m³ <24 January 2005 1 day 125 μg/m³ <3 January 2005 NO2 1 hour 200 μg/m³ <18 January 2010 1 year 40 μg/m³ January 2010 PM10 24 hours 50 μg/m³ <35 January 2005 1 year 40 μg/m³ January 2005 Pb 1 year 0,5 μg/m³ January 2005 Benzene 1 year 5 μg/m³ January 2010 CO 8 hour 10 mg/m³ January 2005 O3 8 hour 120 μg/m³ 25 January 2010

These directives determine the limit values and recommended values. Their goals are to protect the human health as well as the environment (animals, plants…) and are based on toxicological, eco-toxicological and epidemiological knowledge. One can distinguish limit values and recommended values. Limit values are legally binding. The recommended values are suggested but are legally not imposed. These recommended values should allow avoiding serious and permanent damage for health or for the environment. In the case of ozone, threshold values were defined. When a threshold value is exceeded, this can lead to the obligatory information of the citizens or to take actions that lead to the lowering of the emissions. A problem in this regard is that pollutants that have been emitted at a certain location at a certain point in time can lead for instance to ozone formation at another location at a later point in time (see also 1.1 on page 5). With regard to air quality, the World Health Organization (WHO) provides standards for a range of air pollutants. The primary aim of these standards is to provide a uniform basis for protecting public health from the effects of air pollution. The WHO has defined a set of recommendations that are considered to be acceptable levels of air pollution in terms of impacts on human health and the environment. An overview of these recommendations as defined by the WHO is provided in table 1-2 [19].

Table 1-2: Air quality standards according to the World Health Organization

Pollutant Recommended max. level WHO Averaging time Carbon monoxide 100 mg/m³ 15 minutes 60 mg/m³ 30 minutes 30 mg/m³ 1 hour 10 mg/m³ 8 hours Nitrogen dioxide 200 μg/m³ 1 hour 40 μg/m³ 1 year Sulphur dioxide 500 μg/m³ 10 minutes 20 μg/m³ 1 day Ozone 100 μg/m³ 8 hours PM10 20 μg/m³ 1 year 50 μg/m³ 1 day PM2,5 10 μg/m³ 1 year 25 μg/m³ 1 day


p.14

1.4 Environmental problems due to air pollution from road vehicles In paragraph 1.2, an overview is given of the primary air pollutants related to the use of road vehicles, as well as their main effects on human health. In this paragraph some important environmental issues caused by these pollutants will be highlighted.

1.4.1 Ozone formation The atmosphere can be subdivided in different layers. The first layer is the troposphere and begins at ground level and extends to a height of 7 (at the poles) to 17 kilometres (at the equator). Above the troposphere a second layer extends to a height of about 50 kilometres. This layer is called the stratosphere and is characterized by an increasing temperature with height. Most ozone in the atmosphere is contained in the stratosphere and forms the well known ozone layer which has a vital function as it protects us against harmful levels of ultraviolet radiation from the sun. In this context however tropospheric ozone is meant and is known to be the result of photochemical air pollution. Photochemical ozone, also called ‘summer smog’ is formed from nitrogen oxides and some volatile organic compounds through photolysis. Nitrogen dioxide (NO2) is transformed into nitric oxide (NO) and atomic oxygen (O). Atomic oxygen (O), strongly reactive, now reacts with molecular oxygen (O2) to form ozone (O3). Nitric oxide (NO) is strongly reactive and reacts within minutes with ozone (O3) to form molecular oxygen (O) and again nitrogen dioxide (NO2). This is a reversible reaction and will result in a dynamic equilibrium. However, this equilibrium is changed by the presence of volatile organic compounds (VOC) that will react with nitric oxide (NO) to form also NO2 without the transformation of ozone (O3) into oxygen (O2). Ozone can be seen (mainly) as a secondary pollutant as it is the result of the reaction of the exhaust of nitrogen oxides and volatile organic compounds in the air. Ozone is formed mainly in the summer, with high concentrations around midday. ‘Summer smog’ should not be confused with ‘winter smog’. The main components of winter smog are sulphur dioxide and fine particles. During long cold periods with stable weather conditions a layer of cold air becomes ‘trapped’ under a warmer layer of air at a few hundreds of meters height. The cold air that is heavier will remain low and sulphur dioxide and particulate matter will not be able to disperse.

1.4.2 Global warming Global warming is a natural mechanism in which solar radiation that reaches the Earth, is partially absorbed and partially reflected. This reflected (infrared) radiation is then absorbed and transformed into heat by, among others, carbon dioxide and water vapor. Depending on the concentration of these greenhouse gases, the temperature of the atmosphere is determined and influenced. Without the presence of these greenhouse gases, life on Earth would not have been possible. Through photosynthesis, there is an exchange of greenhouse gases between the atmosphere and the biosphere. Physical and chemical interactions between the atmosphere and the oceans are also causing an exchange of greenhouse gases. Both exchanges account for 220 Giga tons of carbon per year. Due to anthropogenic activity (mainly the use of fossil fuels and deforestation) this equilibrium has been disturbed. An increase of the CO2 level in the atmosphere has been revealed. The Intergovernmental Panel on Climate Change (IPCC) mainly relates this increase to the burning of large amounts of fossil fuels and to agricultural activities. The most important greenhouse gases that originate from human activity are carbon dioxide (CO2), chlorofluorohydrocarbons (CFCs), hydro fluorocarbons (HFCs), polyfluorhydrocarbons (PFCs) and sulphur hexafluoride (SF6), methane (CH4), nitrous oxide and ozone [33]. The impact of these gases on global warming is quite complicated, but it is known to be dependant of their concentrations but also of their potential to absorb infrared radiation of the Earth. To be able to compare the ability of the different gases to absorb infra red radiation, Global Warming Potential (GWP) has been defined and has been listed by the IPCC. Direct Global Warming Potentials are one type of


p.15

simplified index, based upon radiative properties that allow estimating the future impact of emissions. GWPs are based on a number of elements, such as the radiative efficiency and the decay rate of the gas concerned. Some examples of GWPs are given in table 1-3 based on [34] and updated values are provided between brackets based on [35]:

Table 1-3: Examples of GWP (updated values between brackets), 100 years time horizon

Greenhouse gas 100 Year GWP Carbon dioxide (CO2) 1 Methane (CH4) 23 (25) Nitrous oxide (N2O) 296 (298) HFC-134a (CH2FCF3) 1300 (1430) Sulphur hexafluoride (SF6) 22200 (22800)

It is found extremely difficult to forecast the consequences of current human activities on the greenhouse effect. Climate change, augmentation of the sea levels, the extension of deserts en the loss of biotopes are all possible effects. The consequences of a climate change are global and have an effect on a large time scale. These effects can be dissected to the natural living environment, human health and the economy. Higher temperatures can result in a more frequent occurrence of tropic cyclones. The supply of sweet water on a regional level could change significantly. The latter can cause increasing dryness as well as flash floods. Another consequence of global warming is the melting of the glaciers [36].

1.4.3 Ozone depletion The ozone layer is located in the stratosphere, at a height between 15 to 45 kilometres. This ozone layer has a crucial role for life on Earth as it protects from ultra violet radiation from the sun. Moreover, ultra violet radiation is harmful for humans, animals and plants. If humans are overexposed to ultra violet radiation, this can cause skin cancer. Ozone depletion is caused mainly by chlorofluorocarbons or CFCs. However, only 2 percent of total CFC emissions are originating from the transport sector [5]. At the South Pole, since the 1980s and on a yearly basis, starting from September and for a few months, a large hole in this ozone layer is occurring. Also above Europe, scientists have observed a small but continuously increasing thinning of the ozone layer. This thinning has negative effects on the plankton population. Plankton uses sunlight and produces oxygen trough photosynthesis. Through this process, plankton absorbs a large quantity of greenhouse gases. Ozone depletion will contribute indirectly to global warming. Ozone itself also influences global warming, however rather limited. In the stratosphere, ozone absorbs ultra violet radiation from the sun and in the troposphere it absorbs infra red radiation from the Earth. A decrease of ozone in the stratosphere will have a cooling effect, while superfluous ozone in the troposphere will cause a rise of the ambient temperature. Also the emissions of nitrogen oxides are causing a large problem. Not only will they contribute to summer smog (see paragraph 1.4.1), if they reach the stratosphere they will also, through several chemical reactions, affect the ozone layer [36].

1.4.4 Acid Rain Acid rain is caused by emissions of compounds of sulphur, nitrogen or carbon, which react with the water molecules in the atmosphere to produce acids. Sulphur dioxide is transformed within hours or days mainly to sulphuric acid, while nitrogen oxide is transformed to nitrogen dioxide and next to nitric acid and to ammonia. Ammonia is mainly released by agrarian activity. SO2 is mainly originating from the energy sector, while NOX is caused primarily by road traffic. The emissions of sulphuric acid, nitric oxides and ammonia are often expressed in acid equivalents (e.g. SO2-eq., Zeq). Most important effects on


p.16

ecosystems are caused by acidification due to atmospheric deposit of sulphur and of nitrogen containing compounds. Acidification leads to a disturbance of the composition of the surface water, the soil and air. The latter causes a disturbance of the biodiversity, deterioration of soil quality and damage to vegetation, trees and an increase of nitrates concentration in groundwater. Further, acid rain also has an impact on buildings. The main sources for these pollutants are the industry, electricity production plants, refineries, road traffic and heating of buildings. Ammonia is mainly originating from the agricultural sector, sulphur dioxide from the energy sector. Nitric oxides are however mainly emitted by road vehicles. Due to more stringent European regulations for new vehicles and for their fuels, emissions from road transport causing acidification are decreasing [37, 38]. The evolution of NOX, NH3 and SO2 emissions related to the transport sector is shown in figure 1-7. This graph is based on the results of MIRA-T 2007 [39].

Figure 1-7: Emissions of NOX, NH3 and SO2 from the transport sector (based on MIRA-T 2007)

When the different pollutant levels from figure 1-7 are weighted by their potential to form acids (NOx: 0,02174 Zeq/g; NH3: 0,05882 Zeq/g; SO2: 0,03125 Zeq/g) NOx is the most important pollutant for acidification. NOX contributes in 2005 for 94% to acidification from the transport sector.

1.4.5 Noise Noise and vibrations are disturbing some human activities such as sleeping, communication and reading. The human health is hereby indirectly influenced through fatigue, headache, insomnia, increased blood pressure etc. The risk for a heart attack is increased due to the exposure to noise. Traffic is responsible for about 60% of all disturbing noise in the Flemish Region [40, 41]. Noise pollution is an issue in many densely populated and/or industrialized areas, among which the Flemish Region. It is a threat for both the quality of life and for human health. Beside noise from road traffic and air traffic, noise from neighbours and noise from recreational activities are the most important sources of noise [40]. The actual restrictions on noise levels of passenger vehicles have resulted in a lowered noise level originating from the vehicle’s engine. However, noise from tyre rolling at high speeds is dominating total noise levels from traffic. Further, also the driving style plays a role in the noise levels associated with the use of road vehicles. The influence on humans depends on different characteristics of the noise: tone, periodicity, intensity. The sound pressure level (expressed in decibels) is considered as the most important parameter [5]. The sound pressure level (sometimes denoted with SPL) Lp can be defined as:


p.17

020.log (dB)p

pL p= Equation 1-1

In the equation 1-1 p is the effective value of the sound pressure, expressed in Pascal and p0 stands for the reference pressure (2.105 Pa). The equivalent sound level pressure Leq defined between two time instants t1 and t2 is defined as the quadratic mean of the acoustic pressure p(t):

2 2

1 2 22 1 01

1 ( )( , ) 10.log .

( )

t

eq

t

p tL t t dt

t t p t=

− Equation 1-2

The sound pressure level LeqA (8h-20h)1 is used to characterize the noise originating from traffic. For traffic, an A-weighting filter, based on the equal loudness contour, is used and is a commonly used way to emphasize frequencies around 3 to 6 kilohertz, where the human ear is most sensitive, while attenuating very high and very low frequencies. Noise and vibrations are disturbing some human activities such as sleeping, communication, reading etc. The human health is hereby indirectly influenced through fatigue, headache, insomnia, increased blood pressure etc. The risk for a heart attack is increased due to the exposure to noise. Traffic is responsible for about 60% of all disturbing noise in the Flemish Region [40, 41]. Besides the impact on humans and human health, noise pollution also impacts on ecotopes of ‘noise sensitive’ species (in particular birds). The noise emissions caused by road traffic are strongly dependent of the traffic volume, the speed of the vehicles and the technical characteristics of the vehicles and of the road. For passenger vehicles at low speed (up to 40 - 50 km/h) the noise is mainly caused by the internal combustion engine. At higher speeds the noise from the tires becomes more and more significant. In general diesel engines are generating more vibrations and noise compared to petrol engines or engines running on LPG or CNG. Electric motors however produce only very small levels of sound. Finally, the noise of vehicles also depends on the type and width of the tires and of the type of the road (concrete, asphalt, special asphalts…).

1.4.6 Other effects Besides the earlier described damages of road vehicles and their effects such as air pollution and associated noise other effects exist as well. The quality of the living environment is affected by less open space. The land-use for driving and parked cars results in less space for playgrounds, parks or other recreational locations. In addition, safety in these spaces has also lowered. Scattering is the splitting of land in smaller parts or less connected parts. The infrastructure for road traffic has a main cause in the scattering and leads to damage of the biodiversity. Roads are barriers for animals that want or need to move from one place to another (e.g. for breeding). This leads to animals getting killed by road traffic and results in local extinction of animal species. Light pollution is also an effect related to road traffic, in particular by lighting of public roads. In this way a public road with street lights can be a barrier for some animals. The presence of road infrastructure lowers the infiltration of precipitation by the ground. Also surface water is more quickly rinsed because of removal or replacement of ditches by tubes. Less ground water is the result of the construction of infrastructure. Finally road vehicles, in particular diesel vehicles, can lead to unpleasant smells.

1 The equivalent sound level pressure between 8h00 and 20h00, according to Equation 1-2


p.18

Jean-Marc

Rectangle

p.19

2 Relation between vehicle technology

and pollution

2.1 Sources of emissions The different emissions related to road traffic, causing different effects or damages as discussed in the previous chapter, originate from different sources. In this chapter we will make an inventory of all main pollutants that are related to the use of road vehicles. In this work, only air pollution is considered. Beside the gaseous emissions from vehicles also other forms of pollution related to the use of road vehicles exist: polluting liquids (oils) or waste water etc. However, if we focus now on air pollution and if we make an inventory of all emission related to the use of a road vehicle we should consider the different stages of the vehicle’s complete cycle life. Beside the vehicle itself, we also need to consider the fuel that is used to drive it. These different stages range from the extraction of raw materials for the production stage until the final dumping, landfill or recycling stage. The set of all different stages is often referred to as ‘Cradle-to-Grave’ (see figure 2-1).

Figure 2-1: Life stages related to vehicle use

However, in practice it seems very difficult to collect sufficient data for all these different stages. This is especially true, when using this inventory framework for making a comparative analysis of different vehicles, using diverse types of fuels or having different kinds of propulsion systems.

Relation between vehicle technology and pollution

p.20

From the literature [42-47] it is found that the vehicle use stage and fuel production stage are dominant for all of the considered pollutants compared to the contributions of the other stages. Furthermore it is found that only very little variation exists in the emissions related to the production phase for vehicles of different kinds and brands. For comparative analysis, including the manufacturing stage would thus have little or no added value. In the case of vehicles with alternative propulsion systems (e.g. hybrid and electric vehicles) the use of new components (such as large rechargeable battery packs) sometimes causes an increase of the pollutant emissions related to the production of these new components [48]. However, these emissions may easily be compensated throughout the vehicle use phase due to lower fuel consumption. Finally, the emissions related to the end-of-life (disposal, scrapping and recycling processes) are much less compared to the stage of vehicle usage and fuel production. More information about LCA studies of vehicles can be found in paragraph 4.1.5 on page 42. The stages related to the extraction of raw materials and to the fuel production process are in this context often referred to as the ‘Well-to-Tank’ stage. Well-to-Tank is sometimes abbreviated as WTT. The emissions related to this stage are often denoted as the indirect emissions. The stage of vehicle use is in analogy called then the ‘Tank-to-Wheel’ stage. Tank-to-Wheel is sometimes abbreviated as TTW. Emissions related to the vehicle use are also called the ‘direct emissions’. Because of the above mentioned arguments and as shown in figure 2-2, the inventory framework can be limited to these two stages of vehicle use and of fuel production, including the extraction of the required resources.

Figure 2-2: Direct and indirect emissions related to vehicle use

2.2 Direct emissions Direct emissions can be described as the emissions related to the use-phase of the vehicle. Beside the exhaust or tailpipe emissions, also other types of direct emissions can be identified: fuel tank evaporation emissions, tyre and brake abrasion. Tailpipe emissions are the emissions that are formed during the combustion process of the thermal engine used in today’s common road vehicles. For the combustion process, a vehicle requires oxygen (O2). Ambient air contains around 21% of oxygen but also contains other components. Ambient air mainly contains nitrogen (approximately 78%). The combustion of the fuel results in the production of carbon dioxide (CO2). However, only about 14% of the exhaust gases of a passenger vehicle are carbon dioxide. Besides carbon dioxide, the exhaust gases also contain about 70% nitrogen gas and 12% water vapour [14]. In what follows, the direct emissions from road vehicles will be subdivided into regulated emissions and unregulated emissions. Only tailpipe emissions are considered and will be denoted as the direct emissions.


p.21

2.2.1 Regulated direct emissions All vehicles that are sold on the European market have to undergo a homologation procedure at which the exhaust of pollutants is measured while driving a certain test cycle. In this way these pollutants are measured in controlled conditions and according to a standardised driving cycle. These emissions are not allowed to exceed the postulated limits of the directive. Different directives apply for different vehicle categories:

• Directive 70/220/EEC and its amendments for vehicles of category M1 and N1(1)

• Directive 88/77/EEC and its amendments for engines used in vehicles (categories M2, M3, N2 and N3)

• Directive 97/24/EC and its amendments for motorized two-wheelers These directives are related to the following emissions:

• CO, NOX, hydrocarbons (HC) in case of petrol vehicles • In the latest amendment of directive 70/220/EEC (2007/715/EC) limit values for PM

emissions of petrol vehicles are defined (Euro 5 and Euro 6) • CO, NOX, combined HC + NOX and PM in case of diesel vehicles • For Euro 1 and Euro 2 in case of diesel vehicles only the combined emission limits of HC + NOX

were specified Originally the emissions of passenger vehicles and light duty vehicles were limited by the directive 70/220/EEC and this directive was amended several times since then. For instance the amendments by directives 93/59/EC and 96/69/EC introduced the well known Euro 1 and Euro 2 emission limits. Later, the directive 98/69/EC introduced the Euro 3 and Euro 4 emission limits. Also more stringent terms for the quality of fuels for vehicles were dictated, in particular with regard to the sulphur content of the fuels (see paragraph 1.2.4 for more details). In case of diesel vehicles, starting from Euro 3, both NOX and combined HC + NOX emission levels are known. Consequently the level of HC emissions can be calculated by subtracting the values of both data. For older vehicles (pre Euro 3), only the combined HC + NOX emission level is known. A repartition can be made by using repartition factors proposed by IFEU [49] and also used by Cleaner Drive [50, 51]. It is important to notice that the limit values differ for diesel and petrol vehicles. Furthermore, different limits apply for passenger vehicles and for light duty vehicles. Further it is important to know that emission data originating from homologation procedure tests are not corresponding with emission levels that occur in real traffic circumstances. The emission values measured during homologation can deviate because of the following:

• Due to the ageing of the vehicle and its engine and because of possible bad maintenance of the engine and other components (e.g. the catalyst) the emissions are assumed to increase over time. Directive 98/69/EC however foresees a sustainable test after 80.000 kilometre. After this mileage the vehicle emissions are not allowed to surpass the initial stated limit values.

• The emissions related to a cold start were not taken into account before homologation date 1st January 2002.

• The average acceleration in the driving cycle used for the homologation is much lower than that occurring in real traffic situations. Especially during heavy accelerations large emission peaks have been observed (up to 30 times larger) [52].

1 See paragraph PART I: 4.4.1.1 on page 47 for more details on the vehicle categorization


p.22

2.2.2 Unregulated direct emissions Non-regulated direct emissions can be subdivided into emissions that are directly proportional to the fuel consumption of the vehicle and those emissions that do not have a direct relation with the fuel consumption. The well known exhaust gas CO2 is an example of an unregulated direct emission that is directly proportional to the vehicle’s fuel consumption. The European Directive 1999/94/EC dictates that the CO2 emission level of a vehicle is obligated to be communicated. The CO2 emissions (ECO2) are calculated from the fuel consumption trough the following relation [53].

2 210. .CO COE k FC= Equation 2-1

With:

• ECO2: emission value for CO2 expressed in grams per kilometre • FC: fuel consumption expressed in litre per 100 kilometres • kC02: emission factor for CO2 expressed in kilogram per litre

The emission factors for CO2 (kCO2) for different fuels are given in table 4-3 on page 47, where the main fuel characteristics are discussed. In case of CNG used as a fuel, the fuel consumption is expressed in Nm³ per 100 kilometre1. The homologation procedure (93/116/EC) foresees a measurement with two reference gases G20 and G25. For further analysis in a Belgian framework the reference gas G20 is considered as its composition is closest to the natural gas available in Belgium [54]. The emission factor kCO2 for compressed natural gas is now expressed in kilogram per Nm³. Another unregulated direct emission is SO2 that is also proportional to the fuel consumption and that can be calculated as follows [53].

2 2

.10. . 2. .

100f

SO SO fuel

FCE k FC S

ρ = =

Equation 2-2

With:

• ESO2: emission value for SO2 expressed in grams per kilometre • kSO2: emission factor for SO2 expressed in kilogram per litre • Sfuel: sulphur content of the fuel, expressed in gram sulphur per gram fuel • FC: fuel consumption expressed in litre per 100 kilometre • ρf: fuel density expressed in grams per litre

The emission factors for SO2 (kSO2) for different fuels are given in table 4-3 on page 47, where the main fuel characteristics are discussed. The sulphur content of fuels for vehicles is regulated by the European Directive 98/70/EG and its amendments. Since January 2005 the maximum sulphur content was set at 50ppm. Since January 2009 this has been further lowered to 10ppm. In case of compressed natural gas, the sulphur content and consequently the direct emission value for SO2 is considered as zero[54]. In case of LPG before January 2005 a sulphur content of 15mg/kg is considered, based on [55]. After this date, the same limit as for petrol fuel applies.

1 Volume of natural gas is expressed in normal cubic metre as the physical properties of gases depend on the pressure and temperature. Normal cubic metre or Nm³ refers to the volume at 0°C and 1 atmosphere pressure.


p.23

For biodiesel (RME) the maximum permitted sulphur content is 100 mg/kg (according to E DIN 51 606). An overview of the sulphur content values expressed in mg/kg for the different fuels is given in table 2-1.

Table 2-1: Maximal sulphur content of fuels for road vehicles

Date Petrol Diesel LPG CNG (G20) Biodiesel (RME) 01/01/2000 150 mg/kg 350 mg/kg 15 mg/kg 01/01/2005 50 mg/kg 50 mg/kg 15 mg/kg 01/01/2009 10 mg/kg 10 mg/kg 10 mg/kg

↑ 0 mg/kg

↓

↑ 100 mg/kg

↓ Furthermore, nitrous oxide (N2O) is also an unregulated direct emission that plays a crucial role in climate change (see paragraph 1.4.2). COPERT III [56] proposes for this pollutant global speed independent emission factors that include both warm and cold start emissions. In the framework of the Cleaner Drive project[57, 58], estimations for this emission factor have been made for different fuel uses and engine ages (based on Euro norm). For the older vehicles (pré-euro), no emission data are available, so estimations of the emissions and of the consumption were made based on the COPERT III methodology [56]. For biodiesel (RME) the same values for N2O emissions as for regular diesel fuel can be taken as an estimation [54]. In table 2-2 below these emission values are shown.

Table 2-2: Nitrous oxide (N2O) emission values

Petrol Diesel LPG CNG (G20) Biodiesel (RME)

Pré – Euro 0,005 g/km 0,027 g/km 0,015 g/km 0,015 g/km 0,027 g/km Euro 1 0,027 g/km 0,002 g/km 0,015 g/km 0,015 g/km 0,002 g/km Euro 2 0,013 g/km 0,005 g/km 0,012 g/km 0,012 g/km 0,005 g/km Euro 3 0,005 g/km 0,008 g/km 0,005 g/km 0,005 g/km 0,008 g/km Euro 4 0,005 g/km 0,008 g/km 0,005 g/km 0,005 g/km 0,008 g/km

The emission values will be used for the environmental assessment methodology for passenger vehicles and light duty vehicles. Another unregulated direct emission gas is methane (CH4) which has an important impact on global warming (see also paragraph 1.4.2 on page 14). For passenger cars and light duty vehicles, the Well-to-Wheel study from General Motors [59] provides estimations for the direct methane emissions:

• For petrol and LPG fuelled vehicles: 0,02 g/km • For diesel fuelled vehicles: 0,01 g/km • For CNG fuelled vehicles: 0,124 g/km

Emissions related to homologated hydrocarbons can be further subdivided:

• The VOC can be split into a methane fraction and a non-methane fraction • The non-methane fraction can be split into components that are responsible for negative

effects on the environment and others

The COPERT II methodology [60] offers a calculation method to determine the methane fraction from the VOC emission value. The COPERT III methodology [56] further offers updated weight percentage factors to determine the composition of the non-methane volatile organic components (NMVOC). This methodology provides calculation methods for different VOC emissions of conventional vehicles, in particular petrol, diesel and LPG vehicles. For emissions of specific organic compounds of vehicles running on compressed natural gas, estimations can be made based on the MEET project [8].


p.24

Other pollutants can not be determined from fuel consumption nor from other pollutant values, such as polycyclic aromatic hydrocarbons (PAHs) and persistent organic pollutants1 (POPs). Again the COPERT methodology proposes emission factors to calculate the emission levels for a large list of these PAHs and POPs expressed in micrograms per kilometre. These emission factors are speed independent and cover both warm and cold start emission conditions. Finally, lead (Pb) can also be mentioned in this context as its emissions are proportional to the fuel consumption. Since 2000 lead (TEL as an octane-enhancer) has been banned from fuels for road vehicles and replaced by other additives (e.g. MMT) and so TEL is not further considered in this framework.

2.3 Indirect emissions Beside the tailpipe emissions related to the use of the vehicle, as discussed in the previous paragraph 2.2 on page 20, there are emissions related to the production and distribution phase of the fuel. The following elements of this phase can be considered:

• Extraction of raw materials (crude oil extraction, natural gas exploitation…) • Transport and storage of the raw materials (bunkers, pipelines, railway transport, trucks…) • Processing of the raw materials to fuels or production of electricity (fuel refinery process,

electricity generation…) • Distribution and storage of the fuel (pipelines, road transport, electricity distribution

network…) • Refilling of fuel (leakage, vaporization, compression)

The analysis of the emissions related to the production of the fuel or electricity generation is complicated because many different production pathways are possible and exist. When considering the different raw materials (crude oil, natural gas, solar and wind energy, biomass…) a large list of different fuels can be produced that can be applied in different vehicle technologies. For instance, petrol fuel is refined from crude oil and can be used to power a petrol ICE vehicle. Another example is the production of compressed hydrogen from natural gas that can be used in a (prototype) fuel cell electric vehicle. In a study from General Motors a large number of pathways has been identified for the production of fuels, together with fifteen vehicle technologies. As an example the production of electricity from natural gas can be considered. This electricity could be used to recharge the battery pack of a fuel cell hybrid electric vehicle. However, the electricity could also be used to produce hydrogen (through electrolysis). This hydrogen could be used as a fuel for the fuel cell of the considered hybrid vehicle. On the other hand, hydrogen can also be produced directly from natural gas. Natural gas can be used as a fuel in a CNG hybrid electric vehicle. This example is illustrated in the figure 2-3 below, based on reference [10].

1 Organic compounds that are resistant to environmental degradation. For this reason these pollutants tend to accumulate and are capable of long-range transport


p.25

Figure 2-3: Example of fuel production pathway: natural gas

For the emissions related to the fuel production, the European study MEET [8] can be considered as it provides a common basis of comparison for Europe. The life cycle emissions of different fuels were calculated, taking into account the extraction and transport of crude oil (or other materials), the refinery process and distribution of the fuel. In the case of biodiesel, the influence of the agricultural process is also taken into account. The emissions of different pollutants related to the production phase of different fuels are summarized in the table 2-3 below.

Table 2-3: Emission factors related to the production phase of different fuels for road vehicles

CO NMHC NOX PM CO2 SO2 N2O CH4

FUEL TYPE mg/kWh mg/kWh mg/kWh mg/kWh mg/kWh mg/kWh mg/kWh mg/kWh

Petrol 18,4 761,4 151,9 8,6 33100 236,2 0 62,6

Diesel 16,6 315,4 129,6 3,6 24500 174,2 0 56,5

CNG (G20) 5,0 99,0 38,2 2,9 14759 60,8 0 805,3

LPG 14,8 202,7 116,3 5,4 21600 114,1 0 58,0 Biodiesel (RME)

493,1 280,4 871,9 66,6 -172786 245,5 0 0

Source: MEET (1999) [61]

2.4 Noise As for the regulated emissions of direct vehicle emissions, information on the external noise produced by a vehicle is available from homologation tests. These tests allow characterizing the sound pressure level of the vehicle that is assessed. Noise is considered in this context because it is an important kind of pollution related to road vehicles that is like emissions transmitted through the air. For homologation purposes, the measurement is performed according to the method described in the European directive 70/157/EEC. The measurement involves the noise of the vehicle in motion while the maximal sound level is recorded. More details on the test procedure are available in the directive. The European directive 70/157/EEC and its amendments have been continuously setting more stringent limits (latest amendment 2007/34/EC). A simplified overview of the evolution of the limit values over time is provided in the table below.


p.26

Table 2-4: Maximum sound levels for road vehicles

1972 1982 1988/1990 1995/1996

Passenger car 82dB(A) 80dB(A) 77dB(A) 74dB(A)1

Buses 89dB(A) 82dB(A) 80dB(A) 78dB(A)

Heavy duty 91dB(A) 88dB(A) 84dB(A) 80dB(A) The noise of motorized two- and three-wheelers is regulated by the European directive 78/1015/EC (and amendment 97/24/EC) which foresees a gradual reduction of the maximal sound pressure level. An overview of these limits is given in the table 2-5. Currently, a new test procedure that reflects the noise behavior in urban conditions more realistically is being evaluated and new limit values are being determined.

Table 2-5: Maximum sound pressure levels for motorized two- and three-wheelers

Two- and three-wheel motor vehicle 1980 1988/1989 1993/1994

Two-wheel moped

≤25 km/h 66 dB(A)

> 25 km/h 71 dB(A)

Three-wheel moped 76 dB(A)

Motorcycles

<80 cm³ 78 dB(A) 77 dB(A) 75 dB(A)

80 < … < 175cm³ 80-83 dB(A) 79 dB(A) 77 dB(A)

>175 cm³ 83-86 dB(A) 82 dB(A) 80 dB(A)

Tricycles - - 80 dB(A) Besides the vehicle also the tires that are used play an important role in the noise levels. Many different models of tires exist and can be changed easily throughout the vehicle’s life. The noise produced by tires is also regulated by means of the European directive 2001/43/EC.

1 This value is increased with 1dB(A) if the vehicle is equipped with a direct injection diesel engine


p.27

3 What can be done?

3.1 Introduction In the previous chapters an overview of the environmental issues related to the use of road vehicles has been given. Before looking at possible solutions to lower the environmental impact of road vehicles, it is important to put things in a larger perspective and to consider transportation in general. The need of people to move and the distances they have to travel can be lowerd with a good urban planning. This should allow people to travel on foot or by bicycle more easily. A good network of public transportation can lower the need to use a car. However, transportation with passenger cars can’t be avoided completely. Improvements of the environmental performances of road vehicles can be achieved at many levels. There can be enhancement by technological improvements of the vehicles (see paragraph 3.2 below). Also policy makers can play a role by using adequate measures promoting the use of cleaner vehicles or by taking actions to lower the impact of current vehicles. An overview of possible policy measures in this regard is given in paragraph 3.3 of this chapter.

3.2 Technological solutions Technological improvements or new technologies can help to lower the impact of road vehicles on the environment. In this paragraph an overview of different possible elements of technological improvements or new technologies for vehicles in this regard will be given.

3.2.1 Vehicle mass and low weight materials The mass of a vehicle influences the required mechanical energy needed for propulsion. An augmentation of the vehicle’s total mass causes not only the rolling resistive force to amplify but also increases the force needed for accelerating the vehicle. Additionally, driving up a hill will require more energy if the vehicle mass is increased. The influence of the vehicle mass on the different resistive forces and on the required mechanical energy is described in detail in chapter 7 starting at page 113. In literature the effect of the vehicle mass on fuel consumption was investigated and as expected fuel consumption increases as mass increases [62]. If the total vehicle mass can be lowered, this can have a positive effect on the energy consumption of the vehicle. Therefore continuous research and development is made to improve the weight of the vehicle. The use of light weight materials in vehicles can therefore result in improved energy consumption. On the other hand, smaller cars weigh less than large cars or SUVs and therefore generally have a lower energy consumption. To what extent the vehicle mass influences the fuel consumption and emissions is however rather complex. This will be influenced also by the vehicle’s driving cycle. If the

What can be done?

p.28

required mechanical power is changed (e.g. lowered) this will cause the vehicle’s power unit to run in a different point of operation. The point of operation in case of a thermal motor can be characterized by the motor’s rotational speed (expressed in rpm) and the motor’s output torque (expressed in Nm). The fuel consumption and the exhaust emissions are changing depending on this point of operations and are sometimes described by means of so-called engine-maps. This influence can be investigated by means of dedicated simulation software. An example of such a software is VSP (Vehicle Simulation Programme, developed at the VUB) and allows to evaluate among other things the influence of the vehicle mass making use of motor engine-maps considering a chosen driving cycle [63-65]. Research in this regard has pointed out that an increase of the vehicle mass with 200 kg, makes the fuel consumption of a small vehicle (< 1 ton) rise with 8% to 13% and with 3% to 5% for larger vehicles (1,7 tons) [66]. Based on reference [67], an increase of 6,7% was observed for a mid-sized passenger car (1500 kg) in case of an additional charge of 100 kg. A disadvantage related to the use of light weight materials for vehicles is that these materials are more difficult to recycle (low material purity and mixture of materials). Furthermore, the production of these materials is energy intensive and expensive. Moreover, the repairability is sometimes negatively influenced by the use of light weight materials [68]. The weight of a vehicle is also increased due to safety measures (e.g. airbags) but also due to more comfort equipment (e.g. airconditioning system).

3.2.2 Aerodynamic frame design Besides the vehicle mass, also the resistance related to the aerodynamic drag plays a role. This element plays an important role especially when the vehicle is driving at higher speeds (quadratic increase with speed). This is reflected in the mathematical expression for this part of the whole of resistive forces (see paragraph PART II: 7.2.3 on page 114). Besides the speed also the shape of the vehicle plays a role. The dependence of the shape can be expressed with the aerodynamic drag factor CX, multiplied with the frontal surface S. Therefore, both the size and shape of the vehicle influences the energy requirements and thus the consumption. Some manufacturers try to optimize fuel efficiency by optimizing, among other parameters, the aerodynamic drag of the vehicle. For instance, Toyota’s third generation Prius is claimed to have the lowest drag coefficient of all commercially available passenger cars (CX of only 0,25, compared to a typical value of 0,30 – 0,35 [69]). The mounting of a ski box or carrier for bicycles can also influence the aerodynamic drag (both on CX and surface S), causing the resistive forces to increase. Consequently, this also affects the fuel consumption of the vehicle with an increase of 10% in case of a ski box or 20% to 30% for a bicycle carrier [66, 67].

3.2.3 Rolling resistance The rolling resistance is also a parameter which has an influence on a vehicle’s fuel consumption. Rolling resistance is characterized by the rolling resistance coefficient CR. More details on the mathematical formulation of rolling resistive forces can be found in paragraph PART II: 7.2.1 on page 113. Typical values of rolling resistance coefficients are ranging from 0,007 to 0,014. The rolling resistance is influenced by several elements. Of course the type of tires is important, but also the road surface. Furthermore, also the tyre pressure (under-inflation), load, alignment and temperature determine the effective share of the rolling resistance. Tyre manufacturers are developing tires using materials that lower the rolling resistance with minimal effects on wear life and on safety (braking properties, especially in wet conditions)[70]. Further, under-inflation strongly influences the rolling resistance of vehicles. Under-inflation (maintenance by drivers) has a big influence. A reduction from 2 bar to 1,6 bar leads to an increase of the rolling resistance of about 10%. Large variations of the rolling resistance of tires from the same size but different brand are observed (+/- 10%). The type of underground plays an important role. Important characteristics of the road that influence rolling resistance are the texture, stiffness, temperature and longitudinal evenness of the road. Rolling resistance for the same tyre on stiff road surfaces can differ strongly, depending on the surface texture of the road (about +/- 15%). Finally, there are no conflicts between optimization of rolling resistance and tyre noise [71].

What can be done?

p.29

3.2.4 Alternative fuels In this paragraph an overview of fuel technologies that have an influence on the environmental impact of a vehicle will be given. Beside the most commonly used fuels, petrol and diesel fuel, other types of fuels can be used [72]. An overview with the main characteristics of each option is provided.

3.2.4.1 Liquefied petroleum gas The majority of the road vehicles that run on liquefied petroleum gas (LPG) are petrol vehicles that have been adapted (also called ‘retrofitted’). However, there is a tendency from the car manufacturers to bring vehicles on the market that have been developed to run on LPG fuel. The market share of these vehicles is however still very small. Liquefied Petroleum Gas (LPG) is a by-product of oil refinery and is composed mainly of propane (C3H8) and butane (C4H10). The exact proportion of the composing components is varying strongly depending upon the land of origin. In the United Kingdom, for instance, LPG is composed of 90% of propane, while in Italy LPG is containing only 20% of propane. In Belgium the propane/butane proportion is 60/40. In the past, this caused for problems when travelling through different countries but nowadays this variability is no longer a problem for the engines [73]. Expressed in litre per 100 km, the consumption of LPG vehicles is higher compared to their petrol counterparts and even more compared to diesel vehicles. This is mainly because of the lower volumetric energy density1 of the LPG fuel. The efficiency of the LPG engine is comparable to that of a petrol engine [72]. The production of LPG is however less energy demanding than that of petrol fuel (in other words lower indirect energy consumption). See also 4.5 on page 57. Taking into account both the use and production of the fuel, LPG vehicles have a higher CO2 emission per kilometre, compared to diesel vehicles, but somewhat lower than petrol vehicles. LPG vehicles have lower levels of toxic emissions (NOX, SO2, CO, HCs) compared to petrol vehicles. However, this is only achieved if maintenance is performed regularly and motor management is well fine-tuned. If not, some emissions can strongly increase (NOX and HCs) [74]. The emission reduction from retrofitted LPG vehicles is strongly variable and depends on the quality of the fitting. LPG is available all over Europe, with highest concentration of distribution points in The Netherlands, Belgium, France, Italy and Great Britain. In Belgium, LPG is available in about 550 fuel stations [5].

3.2.4.2 Compressed natural gas Like petrol engines, engines that run on compressed natural gas are spark-ignition engines. As for the vehicles that run on LPG, also a large share of the vehicles that run on CNG are adapted or retrofitted. Especially for heavy duty vehicles (e.g. buses) more and more manufacturers use specially designed engines running on CNG. In this way, the engine and the motor management system can be optimized for the use of CNG. Some vehicles can run on both petrol and natural gas and are called ‘bi-fuel’ vehicles. The typical range for a passenger vehicle running on CNG is 200 to 250 kilometres. To achieve the same range as with conventional vehicles, the fuel tank should be 5 times larger for CNG compared to petrol. Compressed natural gas is stored in the vehicle under very high pressure (up to 200 bars). Special charging infrastructure that stores natural gas at a pressure of about 250 bars allows making a quick fill and requires about ten minutes. If natural gas is not available at high pressure, a refill can be done by using a compressor that is connected directly to the vehicle’s tank. This way of refilling requires however about five hours for a complete tank. The compression of the natural gas requires energy and causes additional emissions during the fuel cycle. Natural gas is mainly composed of methane. The exact composition can differ from country to country [53].

1 Volumetric energy density is a term used to quantify the amount of energy stored in a fuel (or other energy storage system) per unit of volume – mostly expressed in mega Joules per litre (MJ/L)

What can be done?

p.30

3.2.4.3 Biodiesel Biofuels are a range of fuels that are produced from plants or from biomass in general instead of from crude oil like conventional petrol or diesel fuel. The European commission strives to have 5,75% (in volume) of biofuels in the total fuels for transportation by the end of the year (2010) [75]. The target for 2020 is set to 10%. Especially biodiesel is considered to be a viable alternative in the short term. Biodiesel can be produced from plant-derived oil, like rape seed, seed from sunflower, palm oil or soya oil. But also waste vegetable oil or fat from animals can be used to produce biodiesel. Biodiesel is consequently considered as a renewable energy source. To achieve a fuel of high quality, these oils need to be treated and this is done by chemical transformations, for example with methanol. The result is methyl ester, like rapeseed methyl ester (RME). This product is useful for use in an engine for vehicles [2] but has somewhat deviating characteristics compared to (conventional) diesel oil, in particular in terms of lower heating value and fuel density. Biodiesel can be blended with conventional diesel fuel. Similar to diesel vehicles, vehicles running on biodiesel will emit more NOX and PM compared to petrol vehicles. Compared to a conventional diesel, the use of biodiesel will lead to an increased level of NOX (about +10%), but the emissions of PM and CO and hydrocarbons would drop significantly (up to 50%) [54]. Most of the greenhouse gases associated with the use of biodiesel are related to the agriculture phase, the oil extraction and the chemical transformation processes involved. During the (energy intensive) agricultural phase important amounts of N20 are produced as well. A reduction of 40% up to 60% in greenhouse gases is considered for the use of biodiesel compared to conventional diesel fuel [76]. A differentiation can be made between first and second generation biofuels. First generation biofuels are produced from basic feedstock that competes with food production. In case of second generation biofuels, biomass is collected from residual non-food parts of crops and non-food crops. The main problem of these second generation biofuels is the limited availability of the feedstock [77]. More research related to both the indirect and direct emissions of different types of bio-fuels is needed to be able to make a comparison with fossil fuels. A recent study in this field shows big differences for the different possible production chains of different bio-fuels [78]. Also the production through gasification of second generation bio-fuels is promising from an environmental point of view [79].

3.2.4.4 Hydrogen Hydrogen fuelled engines are claimed to have several advantages, among which is the low concentration of pollutants in the exhaust gases compared to internal combustion engines using traditional fuels. Furthermore, a hydrogen fuelled engine is capable of very lean combustion, and a better efficiency is reached. A typical value for the efficiency achieved with hydrogen fuelled engines is about 30% [80, 81]. Instead of burning it, hydrogen can also be used in a fuel cell for on-board electricity production. In a hydrogen fuel cell, hydrogen is converted into water by taking oxygen from the ambient air and electrical energy is being generated. This electricity can be used to power an electric power unit of a vehicle. The Proton Exchange Membrane (PEM) fuel cell is considered the most promising for road transport applications in the short and mid term. An advantage of using hydrogen in a fuel cell for road vehicles is the absence of direct emissions[82], except H2O. Hydrogen gas is only seldom available as such in nature. However it can be produced from largely available water or from fossil fuels or even from biomass. Hydrogen thus needs to be produced in some way. Many different production pathways are possible. Often electrolysis of water using electricity is considered for hydrogen production. Electrolysis allows considering renewable electricity generation or electricity from nuclear plants. However, hydrogen can also be produced from natural gas (see Figure 2-3 on page 25) or from methanol. Each production pathway will be characterized by energy consumption and associated emissions. Current hydrogen production (for the chemical industry) is mainly produced from natural gas. New technologies for producing H2 are being developed such as production from biomass or as a by-product of other processes [4, 83]. The storage of hydrogen is a difficult issue. If hydrogen is liquefied, this requires very low temperatures (20°K) and leakages due to evaporation are occurring (2% to 4% per day). Storage under high pressure (700 bars) is accompanied by with large losses due to the energy demanding compression of the gas. An alternative solution is the storage of hydrogen in metal hydride structures or by absorption in carbon nano-tubes [82].

What can be done?

p.31

3.2.5 Alternative propulsion systems Beside alternative fuels as described in the previous paragraph, also new types or alternative vehicle propulsion systems are considered for improving the overall energy efficiency and pollution from road vehicles. Battery, hybrid and fuel cell electric vehicles can be classified as vehicles with an alternative propulsion system. A brief discussion about these alternatives will be given in this paragraph.

3.2.5.1 Battery electric vehicles Battery electric vehicles dispose of a (pure) electric propulsion system including a rechargeable battery pack as the on-board energy storage. This type of propulsion system for road vehicles exists already since the end of the 19th century but were repressed because of the success of the internal combustion engine technology [84]. More recently (1980s and 1990s) these vehicles became popular again, because of their great potential as environmentally friendly vehicles. A big advantage of battery electric vehicles is that they can be recharged with electricity from various types of energy sources: oil, gas, nuclear, wind, solar. A battery electric vehicle is driven by an electric motor (or motors) powered from the battery pack. The use of an electric motor drive allows implementing regenerative braking. The efficiency of an electric motor is also significantly larger than of an internal combustion engine: 80% up to 90% compared to 15% up to 40% for conventional thermal engines [4, 5]. Another advantage is that an electric drive consumes almost no energy when standing still. An advantage in terms of environmental performance is that they produce no direct pollutants and a low noise emission (tyre noise is dominant). When considering global environmental performance, it is important to consider also the electricity production, transport and distribution phase. As stated earlier, electricity can be produced in different ways and allows choosing for sustainable options such as wind and solar power or choosing for nuclear power. This leads to various pathways with different impact levels. An important disadvantage of battery electric vehicles however is their small range compared to conventional vehicles (80km up to 200km according to some manufacturers) and is related to the battery technology. New battery technology such as lithium-ion is promising to further increase this driving range. For recharging, a simple household electricity plug is sufficient. However larger electricity connections can allow fast charging. A full recharge takes 5 up to 8 hours. Fast charging installations would be able to reduce this charging time down 20 minutes or lower. A potential market introduction of battery electric vehicles could be via car-sharing services that show interest in this kind of vehicle.

3.2.5.2 Fuel cell electric vehicles In an electric vehicle, the rechargeable battery pack can be replaced by a fuel cell stack acting as the energy storage unit. This kind of vehicle is called a fuel cell electric vehicle. There exist many different types of fuel cells. The operating temperature is an element of distinction, as well as the electrolyte used. However hydrogen fuel cells are considered most often as an option for vehicular applications. A hydrogen fuel cell uses oxygen from the ambient air and hydrogen from a reservoir to generate electricity. Hydrogen can also be generated on-board the vehicle from methanol, petrol, ammonia or natural gas, by use of a reformer. An important advantage of fuel cell electric vehicles is that they use an electric motor as the drive unit providing a highly efficient conversion from electrical energy to mechanical energy. The proton exchange membrane fuel cell (PEM) has a solid electrolyte and is characterised by a high energy density. Further it is considered as most appropriate for automotive application as it responds relatively fast (compared to other types of fuel cells) to dynamic loads and as it operates at relatively low temperature (90°C) [63]. Regenerative braking is however not possible in a pure1 fuel cell electric vehicle, since the fuel cell stack can not be used in the inverse way (electrolysis). If a battery pack is added, this would be possible, however then we have a hybrid electric vehicle and no longer a (pure) fuel cell electric vehicle. Some main advantages and disadvantages of hydrogen are discussed in paragraph 3.2.4.4 on page 30. Further, fuel cell electric vehicles (passenger car) can be designed for a range up to 400 km, depending on the

1 pure fuel cell electric vehicle means a vehicle that is there is only one single energy storage system on-board, namely the fuel cell system.

What can be done?

p.32

size of the tank and whether hydrogen is stored in liquid or in gaseous form. In Belgium, hydrogen is distributed by use of barrels. No infrastructure is yet available to refill hydrogen at the current gas stations.

3.2.5.3 Hybrid electric vehicles Hybrid electric vehicles are a large range of vehicles with various configurations of hybrid propulsion systems. Two large categories of hybrid vehicles can be identified: serial hybrids and parallel hybrids. In serial hybrid vehicles the wheels are driven solely by the electric motor that takes his energy from more than one energy storage system. For instance, the motor could take the electrical energy from a battery pack or from a generator-engine group. An example of such a vehicle is the Renault Kangoo Elect’road. If the wheels are driven by both an electric motor and by a thermal engine (both are mechanically coupled) this is called a parallel hybrid vehicle. An example of a parallel hybrid vehicle is the Honda Civic IMA. A specific combination of both structures (serial and parallel) can lead to a more complex system such as the Toyota Hybrid System (as used in the Toyota Prius) and is called a combined hybrid. In this structure, one engine and two electrical machines (a motor and a generator) are mechanically coupled by use of a planetary gear system. Through this coupling the combined hybrid system can function as a parallel hybrid, but the planetary gear allows redirecting (part of) the power from the engine to the connected generator. Now the engine is also used to recharge the battery pack and the power flow from the engine is similar to that in a serial hybrid. For this reason, this is called a combined hybrid. Finally there exist hybrid electric vehicles that can also be recharged from the grid (as a battery electric vehicle). This kind of hybrids are called ‘plug-in hybrid electric vehicles’. Hybrid vehicles can strongly reduce the energy consumption of the vehicle (30% to 50% reduction). The emissions of hybrids depend on the type of fuel that is used for the thermal engine of the hybrid configuration. Generally, lower emission levels can be achieved as the thermal engine can operate in a more optimal point of operation. However, this strongly depends on the management system that is used and that determines the power flow of the different energy storage systems.

3.2.6 Combustion engine technologies

A detailed overview of new engine technologies or of research topics of combustion engines would be out of the scope of this thesis. However, a brief discussion about some main engine technologies related to emission reduction techniques will be given below. The ever more stringent emission standards for vehicles with an internal combustion engine have lead to the development and implementation of a series of new engine technologies. Considerable research is being performed to improve the performance and emission levels of internal combustion engines and to improve their electronic control [4]. An overview of these technologies and their relation with emission standards can be found in the literature [85]. Third generation common rail systems have been developed and are used for direct injection for both diesel and petrol engines and are used to optimize fuel injection timing and quantity. On diesel engines it features higher pressure for fuel injection systems. Together with appropriate electronic control this can improve motor noise and vibrations. Exhaust gas recirculation (EGR) is an emission reduction technique used to achieve lower NOX emissions in both petrol and diesel engines. A part of the engine exhaust is fed back to the engine cylinders and leads to a lower combustion temperature. Nitrogen oxides are formed at a faster rate at high temperatures. Catalytic converters (3-way) are used to convert toxic combustion by-products into less toxic substances through a chemical reaction. In a catalytic converter carbon monoxide (CO) and hydrocarbons (HC) are oxidized to form respectively carbon dioxide (CO2) and nitrogen (N2). In a three-way catalytic converter a reduction reaction of nitrogen oxides to nitrogen and oxygen occurs. The catalytic converter works most efficiently if the engine is running slightly above stoechiometric point (between 14,6:1 and 14,8:1 air to fuel ratio, by weight). The conversion can be very effective, however only in a very narrow band of air to fuel ratio. This requires closed loop control using one or more oxygen sensors. However, the use of catalysts has also some undesired side-effects such as increased ammonia (NH3) and nitrous oxide (N20) emission levels [86]. However, these emissions are not limited by the current emission standards. Catalytic converters also need to be warmed up before their conversion efficiency reaches a satisfying

What can be done?

p.33

point of operation. This means that at a cold start, the catalytic converter does not work yet. Only after a few kilometres driving the catalyst’s temperature will be sufficient to achieve effective emission reduction. New techniques are developed to achieve more rapidly the desired temperature of the catalytic converter. For compression ignition engines (diesel) a diesel catalytic converter can be used. The excess of oxygen in the exhaust is used to convert carbon monoxide to carbon dioxide and hydrocarbons to water vapor and carbon dioxide. In order to reduce NOX emissions of diesel engines, several techniques are applied such as the selective catalytic reduction (SCR). This converter requires ammonia (injection of urea) to covert NOX to nitrogen and water vapor. Daimlers’ ‘BlueTec’ systems for instance uses a NOX reducing system with a urea-based reductant (AdBlue). Measures are required to reduce the exhaust of excess ammonia at the tailpipe. Also other techniques exist such as ‘DeNOX catalyst’ that traps NOX by adsorption in a porous material. In order to lower the exhaust of particles from diesel engines, filter systems are used. These systems are often denoted as diesel particle filter or DPF. Many different configurations exist. A main distinction can be made between catalyzed DPF and uncatalyzed DPF. An example of the latter is the continuous regenerating trap or CRT that is composed of an oxidation catalyst and a filter. Nitrogen monoxide from the exhaust gases is converted into nitrogen dioxide in this catalyst and this nitrogen dioxide is used to react with the diesel particles in the filter of the CRT. An adverse effect however is the shift in particle size due to the use of a DPF. A larger part of small particles (20nm -50nm) are resulting from the use of these DPF [15].

3.2.7 Driving style Not only the vehicle technology, the kind of fuel or the type of propulsion system are determining to which extent environmental damage is caused, but also the way the vehicle is used. In other words the driving style and the traffic situation play an important role. The traffic situation has an influence on the speed cycle of the vehicle and thus determines the fuel consumption and associated emissions. Often infrastructure is adapted mainly to influence and improve traffic flow or safety or both. However, these adaptations or measures also affect the driving style of drivers. Well known examples of infrastructure measures are speed ramps, zone 30, phased traffic lights and roundabouts. Speed ramps are introduced to lower the speed of vehicles in areas where high vehicle speeds are dangerous for cyclists and pedestrians. Building a speed ramp is thus a measure to increase the safety of traffic. A speed ramp will force the driver to slow down. However, often drivers will accelerate again once the speed ramp is behind them. This will cause additional accelerations and thus an increase of the fuel consumption and of the emissions [66]. Zone 30 is an area where the maximal allowed speed is 30 km/h. If well designed these zones should create a better traffic flow and less speed variations. If this is the case, lower fuel consumption and lower emissions will result. Phased traffic light motivates drivers to maintain a certain speed so they will have a green light at the next traffic lights. The phased traffic lights create a more steady speed profile and will result in a lower fuel consumption and less emissions. As a conclusion, infrastructure measures can have a positive influence on traffic flow and safety, but the influence on the environmental performance is not always positive. Sometimes the effect is difficult to assess or to quantify.

3.3 Policy measures Policy makers at different levels (international, national, regional) can use measures to influence the impact on the environment due to the use of road vehicles. In this paragraph an overview of some potential measures at the disposal of the policy makers will be discussed. Some of the policy measures are on a voluntary basis while others are mandatory for the involved actors. Also the impact and time scale of the measures can vary quite strongly from one another. Nevertheless, the following categories of policy measures or instruments can be identified:

What can be done?

p.34

• Pricing Instruments • Mandates for procurement and selling (fleet quota) • Control Instruments – user advantages • Communication and dissemination instruments

Some examples for each of the above mentioned types of policy measures to promote the use of cleaner vehicles or to lower the impact of vehicles on the living environment are described in the five following paragraphs.

3.3.1 Pricing Instruments Many different kinds of pricing instruments exist and can be categorised as follows:

• Fiscal measures • Road pricing measures • Subsidies

Often policy makers are using fiscal measures to regulate or to influence markets or to adjust people’s activities. Also for the use of road vehicles and traffic fiscal measures are applied. Car taxes for different aspects of the usage of road vehicles exist: the purchase of a car, the possession of a car and the use of a car. Different kinds of fiscal measures exist and many others are possible to implement. When looking at measures dealing with the purchase of road vehicles, the following taxes are encountered: value added taxes (VAT) on the purchase of the vehicle and various taxes for the registration of the vehicle. Furthermore, the possession of a vehicle is also used to claim (often on a yearly basis) taxes. Also taxes on (obliged) insurance for vehicles can be seen as a tax on its possession. Finally, the use of a vehicle is charged with several forms op taxation. For instance taxes on fuels for road vehicles are applied in most countries. Another form of taxation on the use of a vehicle is the levy of tolls and the introduction of labels or vignettes to have access to (certain parts of) the road infrastructure. Transport related taxes are significant in the total tax income in Belgium. The fiscal income of vehicle taxation (registration and circulation tax) amounts up to 0,7 % of the GDP in Belgium, which is 1,5 % of the total fiscal income [87]. Fuel taxes or excises are a form of taxation that is related to the use of the vehicle. Fuel taxes are included in the energy related environmental taxes. The European Union imposes minimum tariffs for fuel excises. In most countries (including Belgium) petrol is submitted to higher tax levels compared to diesel. Transport fuel taxes represent 3 % of the total tax income in Belgium [87]. This variety of fiscal measures is used by the different authorities to generate revenues in general and in particular to cover expenses related to the exploitation and maintenance of road infrastructure. At the same time these fiscal measures are a very effective tool to influence the behaviour of people in the usage of road vehicles. For instance, a differentiation in fuel taxes in favour of a certain type of fuel (e.g. diesel fuel) can lead to a particular purchase behaviour and to a steered car fleet composition. Fiscal measures are ideal tools for being adapted in function of the environmental characteristics of the vehicle. In Belgium the car taxes are however still based on technical characteristics (i.e. the vehicle’s engine displacement combined with a fuel depending supplement - diesel cars pay a supplement as compensation on fuel taxes). Thus, it is not (necessarily) taking into account the environmental performance of the considered vehicle. However, more and more European countries are introducing fiscal measures based on the environmental characteristics of the vehicle, usually bases on direct CO2 emission levels. In Belgium, a reduction of the revenue tax is granted to the buyers of “clean vehicles”. Vehicles presenting low CO2 emission levels are considered “clean vehicles”. This reduction of the revenue tax is

modulated based on the CO2‐emission levels of the vehicle. This reduction on the income tax was in place from the 1st of January until the 30th of June 2007. In the meantime this measure has been

What can be done?

p.35

replaced by a direct discount on the purchase price of the vehicle. The system of income tax reduction was also applied to the purchase of new diesel vehicles equipped with a particulate filter. A major drawback of this system is that the benefit for the consumer is paid back two years after the investment in a cleaner vehicle was made. This would lower the impact of this measure on the purchase decision [87, 88]. Often, in this regard the definitions or criteria used to evaluate the environmental performance of the car are limited (e.g. only considering CO2, technology based) and/or are often discontinuous (e.g. the use of threshold values). Road pricing measures are mechanisms used to reduce vehicle use and/or to encourage shifting the vehicle use to other times or places. The most common types of road pricing are based on distance, place and/or time. An example of a road pricing mechanism is the congestion charge (cfr. the London Congestion Charge). When benefits are given to drivers of environmentally friendly vehicles (e.g. lower tolls) a higher share of clean vehicles using the route or entering the city can be expected [87]. Subsidies can be defined as direct incentives given at the moment of purchase and which can not be qualified as fiscal incentives. European regulations about subsidies are quite complicated and there is a clear shift from subsidies towards fiscal incentives for promoting environmentally friendly vehicles.

3.3.2 Mandates – procurement and selling quota A fleet quorum in this context is a policy measure that postulates a target for the composition of car fleets. Another approach of setting targets is postulating targets for the newly purchased cars by the fleet owner. This can be applied on both public and private car fleets. Also targets for the car manufacturers for the share of environmentally friendly vehicles sold could be used as a policy measure. A voluntary variant of fleet quota are the ‘convenants’ or voluntary agreements between the competent authorities and fleet holders or car manufacturers. An example of a fleet quota for car manufacturers is the Californian Law imposing the car manufacturers a minimum sales share of zero emission vehicles. On a voluntary basis and at the initiative of some private or public fleet holders (e.g. postal operators), actions to evolve towards ‘greener fleets’ are made. Both private and public companies that want to contribute to sustainable development and sustainable mobility are making initiatives in this regard and use fleet quota to postulate targets for their own vehicle fleets. Emission credits or trade is a market mechanism that allows manufacturers or other actors to exchange emission reduction credits (in case emissions are below the agreed targets). This system allows to minimize global costs for the concerned actors and society and to have more flexibility towards compliance with emission limits. Emission credits are linked to measures using quota: involved fleet owners or stakeholders can either buy appropriate (clean) vehicles to comply with the quota or have the possibility to buy emission credits from fleet holders that already surpassed their quota.

3.3.3 Control Instruments – user advantages Policy makers can use a set of initiatives to promote environmentally friendly vehicles. These initiatives are sometimes referred as user advantages. User advantages aim at offering benefits to users of environmentally friendly vehicles. An example of such a measure, is a reduced (or free) parking fee for environmentally friendly vehicles. User disadvantages aim to discourage the users of vehicles with bad environmental characteristics. An example of such a discouraging measure is prohibiting the entrance to a certain area (and/or at a certain time) of vehicles not corresponding to a minimum emission standard. This kind of area is sometimes called a “low emission zone”. A possible definition of a low emission zone is: “a certain geographic area where specific traffic rules and restrictions apply in order to lower the impact of air pollution”.

What can be done?

p.36

Low emission zones do not directly stimulate people to buy or use environmentally friendly vehicles, though are creating privileges for this kind of vehicle having a lower impact on the environment by allowing them to enter these specific zones. This kind of measure is similar to the variable levy of toll (modulated in function of the environmental performance of the vehicle). For example, since 1996 in Sweden low emission zones for heavy duty vehicles have been introduced in the large cities (Stockholm, Göteborg, Malmö and Lund). This kind of measures often has as a goal to reduce traffic intensity, to promote cleaner vehicles and to improve the local air quality. This allows lowering the impact on human health and/or buildings. Historical centres, often with limited accessibility due to very small lanes and historical pavement, are not accessible for (motorized) vehicles. Sometimes exceptions are made for zero or low emission vehicles. A picture of a sign indicating a zone allowed for low emission vehicles in London is shown in figure 3-1. For delivery services such as postal delivery, these areas are difficult to serve. The use of zero emission vehicles and in particular light electric vehicles can be a solution (see Part II starting at page 81).

© picture from Eric Hands

Figure 3-1: Sign for Low Emission Zone in London

3.3.4 Communication and dissemination instruments Informing people correctly and raising public awareness on societal problems and their possible consequences is a very important element in the large set of measures and actions that can be taken by politicians and public authorities. Scientific background on the issue helps to found the statements and proposed solutions. Information for the broad public has to be clear and easy to understand while often the problems and situation are rather complex to understand or to explain. Therefore it is important to find a good compromise between simplifications and correctness when developing tools or material for sensibility actions towards a broad public.

What can be done?

p.37

© Picture from veg_eric

Figure 3-2: Picture of an air quality signalization panel in Heidelberg, Germany

What can be done?

p.38

Jean-Marc

Rectangle

p.39

4 Environmental assessment tool

4.1 Environmental assessment tools for road vehicles: state of the art Several different environmental rating tools for vehicles have been developed over the past decades. Environmental rating tools allow assessing the environmental performance of a vehicle and most importantly, they allow making a comparison of different vehicles. In order to give an answer to the question “Which vehicles are most environmentally friendly?”, as stated in the general introduction in paragraph 0.1 on page 1, an environmental rating tool for vehicles is important for policy purposes. In what follows, an overview of existing environmental evaluation tools for vehicles will be given.

4.1.1 ACEEE’s environmental damage index and green score The American Council for Energy-Efficient Economy (ACEEE) has published a methodology for the environmental rating of vehicles (both cars and trucks) [89]. This methodology for environmental rating combines the fuel life cycle and the vehicle life cycle in their assessment. The latter also includes the manufacturing phase of the vehicle. For the manufacturing phase of the vehicle, a characterization is made based on the vehicle mass (curb weight). For electric vehicles, the methodology also considers the battery weight and the required number of battery replacements. The methodology uses both regulated and unregulated tailpipe emissions and considers airborne emissions from the fuel supply phase. Adjustment factors to emission standards are used, to reflect real-world driving. The end-of-life phase is however not considered in this methodology. The methodology defines an environmental damage index, abbreviated as EDX, and is the sum of a set of damage functions [89] (see equation 4-1):

($c/mi) ($c/g) (g/mi).ij ijEDX d e= Equation 4-1

Where i is an index over emission species and j is an index over location of emissions. dij Is an environmental damage cost and eij is the quantity of emissions averaged over a vehicle’s operational life. These damage functions are expressed in monetary terms and in this way they quantify the environmental cost related to the use of a certain vehicle. Damage cost estimates from Delucchi [89, 90] are used to calculate the damage functions. Based on the environmental damage index, a score was derived that gives a number between 0 and 100 which is easier to understand. This score is called the “Green Score”. The higher the Green Score, the better the vehicle rates. Based on this Green Score, a class ranking is performed: Superior, Above Average, Average, Below Average and Inferior [91].

Environmental assessment tool

p.40

4.1.2 Cleaner Drive-index The Cleaner Drive project was part of the European 5th Framework Programme. A European consortium has developed an environmental rating system for passenger vehicles. The cleaner drive methodology considers both indirect emissions and direct emissions1. For the indirect emissions, this methodology uses emission factors from the MEET study [8]. The following pollutants related to the fuel cycle are considered: CO2, NOX, CH4, NMHC, SO2 and PM. The cleaner drive methodology can be used to assess vehicles running on petrol, diesel, LPG, CNG or using electricity. The methodology considers regulated emissions and fuel consumption data for the direct emission levels. Two different damage categories are considered: global warming and air quality deterioration. IPCC’s global warming potentials (see also table 1-3 on page 15) are used to weight the impact of the different greenhouse gases considered (CO2, CH4 and N2O). External costs from the ExternE project [92] are used to calculate the damage related to the different air quality pollutants (CO, HC, NOX and diesel PM emissions). These external costs, expressed in €/kg reflect the overall damage to human health and to ecosystems. An overview of the Cleaner Drive-index calculation scheme can be found in figure 4-1 below. This figure is taken from reference [50].

Figure 4-1: Cleaner Drive-index calculation scheme 2

4.1.3 VCD Auto-Umweltliste Verkehrsclub Deutschland (VCD) is an organization for environmentally conscious drivers and annually publishes a car list of the 300 most environmentally friendly cars sold on the German market. This list, the Auto-Umweltliste (AUL), is meant to inform customers. The cars of this list are assessed on the basis of a set of environmental damage categories. A global score is determined by weighting the different damages. The methodology used for the calculation is developed by the Institut für Energie- und Umweltforschung Heidelberg (IFEU) [86] and has been adapted over the years.

1 In this context only tailpipe emissions are considered as the direct emissions. 2 GDI stands for Gasoline Direct Injection and AQ stands for Air Quality


p.41

The different considered environmental impact categories (2009) with their respective weighting factors are [93]:

• Global warming (60%) • Noise (20%) • Impact on nature (5%) • Effects on human health (15%)

The environmental score starts with 100 basic points, to which points from a set of evaluating categories are added or subtracted. These categories include: power, top speed, CO2-equivalent, noise, HC+NOX and particulate matter. The environmental assessment foresees a quantification of a set of impact categories based on a ‘distance to target’ methodology. The latter is represented on a scale ranging from 0 for the worst to 10 for the best score. The different environmental impact categories are then weighted based on a panel method (see weighting factors listed earlier). The maximum speed of the vehicle was used as a parameter to make an estimation of the real life emissions compared to the emissions of the standard test cycle. The maximum speed and the vehicle weight were used to estimate the emissions related to the vehicle production phase. Besides the list and the environmental score for the vehicles, also a ranking of the different vehicle manufacturers is made. This ranking takes into account several environmental considerations of the vehicle production process (recycling, used materials and resources…) [94].

4.1.4 BIM Clean Vehicles A definition for “Clean Vehicles” has been defined for the Brussels Environmental Institute (BIM-IGBE) by the Vrije Universiteit Brussel and the Université Libre de Bruxelles [51, 53]. In this framework, an environmental assessment methodology has been developed. Based on this methodology an environmental score has been defined. The methodology is based on a Well-to-Wheel framework (see paragraph 4.2 on page 44 for more information) and the total emissions are obtained by considering both direct or TTW emissions and indirect or WTT emissions. To take into account the distance between the source of pollution and the receptor, a weighting is applied in the calculation of the total emissions (see equation 4-2):

.total direct ind indirectE E Eω= + Equation 4-2

Five different damage categories are considered: global warming, human health (cancer and respiratory effects), ecosystems (acidification), damage to buildings and noise. Both regulated and unregulated emissions are considered. An overview of these emissions for the different damage categories can be found in table 4-1. For calculating the contribution to global warming, GWPs are used (see also paragraph 1.4.2 on page 14). For health effects, DALYs are applied and for damage to ecosystems the term PDF.m².year is used1. DALY stands for Disability Adjusted Life Years and PDF stands for “Potentially Disappeared Fraction”. External costs are used to express the damage to buildings. An overview of the different damage factors can in addition be found in table 4-1. All calculated damages are normalized by comparing the damage to the damage of a reference vehicle. Finally, a weighting is applied to combine the different damages into one single indicator. The methodology can also be applied for heavy duty vehicles and for motorized two-wheelers, if minor adaptations are made.

1 PDF is the quantification for the yearly surface of an area where animals will no longer appear due to pollutant deposits.


p.42

Table 4-1: Overview BIM - Clean Vehicles methodology

Classification Weighting Inventory Unit Damage factor

CO2 GWP 1 CH4 GWP 23

Global Warming 25%

N2O GWP 296 HC Daly/kg 6,46E-07 NOX Daly/kg 8,87E-05 CO Daly/kg 7,31E-07 SO2 Daly/kg 9,78-06

Human Health: Cancer and Respiratory effects

50%

PM Daly/kg 3,75E-04 NOX PDF.m².y/kg 5,713 Ecosystems:

Acidification 10%

SO2 PDF.m².y/kg 1,04 SO2 €/kg 8,3 Damage to Buildings 5% PM €/kg 259

Noise 10% 1

4.1.5 LCAs from vehicle manufacturers and other LCAs Many car manufacturers perform Life Cycle Assessments (LCAs) of the vehicles they produce. LCAs for vehicles require a large amount of detailed information and data, so often simplifications and estimations are made. Different manufacturers have different approaches in performing LCAs and therefore comparison of the results is often difficult or sometimes not even possible. A possible reason for this is that different phases of the life cycle are considered. However, most of the manufacturers do use an LCA methodology in accordance to the related ISO standards (ISO 14040:2006 and 14044:2006) to perform the assessments. The following phases of the vehicle life cycle are often (but not always) taken into account:

• Vehicle production • Fuel life cycle • Vehicle use phase (based on a certain life time or mileage) • End-of-Life options (including reuse, recycling, landfill and energy recovery)

Maintenance is an example of a phase that is often disregarded. Some examples of such LCA studies of car manufacturers are listed below:

• Volkswagen Group, the Environmental Commendation [95] • Volvo, Environmental Product Information [96] • Mercedes-Benz, Umwelt Zertifikat [97] • Toyota, Environmental & Social Report Eco-VAS [98]

In figure 4-2 an example of the results of the LCA study from Mercedes-Benz is shown [99]. On this graph, the contribution of the different life cycle phases of the new S-Class for the selected pollution parameters are shown by means of a bar diagram. An interesting result from this study is that the in-use emissions (including the fuel cycle emissions) dominate all other aspects of the vehicle life cycle. Except for the SO2 emissions, the vehicle operation phase and the fuel production phase are the most important life stages of the total life cycle. In the figure 4-2 the following abbreviations are used: abiotic depletion potential (ADP), global warming potential (GWP), photochemical ozone creation potential (POCP), eutrophication potential (EP) and acidification potential (AP). These are different midpoint categories used to assess the impact of the different emissions [99].


p.43

Figure 4-2: Share of life cycle stages of the Life Cycle Assessment for the Environmental Certificate from Mercedes-Benz, absolute values on the right

The domination of the vehicle use phase for most pollutants is also confirmed by reference [42]. In this (independent) study made by the University of Michigan, similar results have been obtained. These results are shown in figure 4-3.

Figure 4-3: Distribution of the emissions to the air for the different stages of a vehicle’s life cycle

Often these LCAs are limited to the inventory of the different emission levels and can be denoted as life cycle inventory (LCI) studies. They do not foresee an impact assessment (LCIA) to calculate the damages caused by the different emissions to different damage categories. Other LCA studies only focus on a specific part of the vehicle life cycle or a vehicle component rather than a complete vehicle. An interesting example of such an LCA study deals with the comparative assessment of different battery technologies for use in hybrid and in electric vehicles [48, 100, 101].

4.1.6 Other environmental ranking tools for vehicles The list below provides some examples of the many other environmental ranking tools that do not include a real damage calculation or impact assessment but are based (only) on emission levels. These ranking tools are in most cases developed or used by consumer organisations or by public authorities for information and public awareness purposes. In some cases public authorities use these ranking tools as a legal basis for determining the level of special subsidies or other incentives.


p.44

• The Eco-Test, developed by the FIA (Fédération Internationale de l’Automobile) [102, 103] • Environmental Performance Label by CARB (Californian Air Resources Board), U.S. [104] • Auto Umwelt Zertifikat and Umwelt ranking by Öko-Trend Institute, Germany [105] • Green vehicle guide from the Environmental Protection Agency, U.S. [106] • Automotive Environmental Index from J.D. Power and Associates, U.S. [107] • “CO2-gids van de auto” from the Belgian Federal government [108]

This list is however not exhaustive and many of the above mentioned evaluation tools are continuously updated or adapted in accordance with the state of knowledge of emissions inventory and (if applicable) of damage impact calculation. More information on these and other evaluation tools can also be found in references [47, 109, 110]

4.1.7 Findings on existing environmental assessment tools From the overview provided in the preceding paragraphs it can be concluded that a large number of environmental assessment tools and environmental ranking tools exists. While LCA is considered as the most extensive and complete methodology, it also comprises some important problems. The most important problem is the difficulty to obtain sufficient and reliable data for the inventory phase. Vehicle manufacturers dispose of the best and most complete data for this purpose. In addition, it was found that the vehicle operation and the fuel production phase compose the most important share of the total life cycle of the vehicle for most emissions to the air. Some of the assessment tools or ranking tools described earlier are focusing only on the vehicle’s use phase and on the fuel production cycle. All of the above mentioned environmental assessment tools and environmental ranking tools have shortcomings and advantages, but most of them provide a lot of useful information. These experiences and knowledge were used for the development of the environmental assessment tool for road vehicles described in the following chapters of this thesis.

4.2 Well-to-Wheel framework To be able to compare the contribution of different vehicles to air pollution, a rating methodology has been developed within the framework of this PhD. This methodology is based on a Well-to-Wheel (WTW) framework. This means that, next to the Tank-to-Wheel (TTW) or direct1 emissions, the Well-to-Tank (WTT) or indirect emissions, due to the production and distribution of the applied fuel, are taken into account as well. Moreover, an impact calculation allows assessing the health impact related to the inventoried emissions, incorporating their impact pathway. This evaluation method is more valuable for decision-making and policy, compared to evaluations based on emission levels only [111, 112]. This approach allows comparing vehicles with different fuel technologies (petrol, diesel, liquefied petroleum gas, compressed natural gas, biofuels, etc.) or different drive train technologies (internal combustion engines, hybrid electric drive trains, battery electric drive trains, fuel cell electric drive trains, etc). Emissions resulting from the vehicle assembly and from the production of its constituting components will not be taken into account. Nor are the maintenance phase and recycling phase of end-of-life vehicles. The main reason for the limitation of the methodology to a WTW framework is data availability for all road vehicles (light duty, heavy duty and two-wheelers). The available literature states that the use phase (TTW) is dominant for most considered pollutants and states that only small differences are expected when comparing different drive trains. The use of large secondary batteries in the case of hybrid and electric vehicles has a limited environmental impact thanks to the high recycling rate of this type of batteries [48, 100, 101, 113]. The environmental assessment of the vehicle is performed through a sequence of five steps: inventory, classification, characterisation, normalisation and weighting. This structure corresponds to the

1 In this context only tailpipe emissions are considered as direct emissions.


p.45

standardised LCA (Life Cycle Assessment) methodology1. Each of these five steps will be discussed in detail in the following paragraphs.

4.3 Well-to-Wheel energy consumption The WTW framework that was used to develop the environmental assessment methodology can also be used to evaluate the energy consumption of different fuel and vehicle combinations. For this purpose the WTW energy consumption can be defined as the aggregation of a WTT or indirect energy consumption and the TTW or direct energy consumption. The direct energy consumption of a vehicle can be calculated from its fuel consumption and from the characteristics of the fuel used. The direct energy consumption of a vehicle can be expressed in mega joule per 100 kilometres (MJ/100km).

6

1. . .

10direct f EC FCε ρ= Equation 4-3

Where:

• εdirect = the direct energy consumption of the vehicle expressed in mega joules of fuel per 100 kilometres

• FC = Fuel consumption of the vehicle, expressed in litre per 100 kilometres2 (in case of light duty vehicles or two-wheelers)

• ρf = Fuel density, expressed in gram per litre3 • EC = Energy content of the fuel, expressed in kilojoules per kilogram4 • the factor 1/106 in this formula is a conversion factor

The fuel characteristics (fuel density and energy content) of a range of vehicle fuels are discussed in paragraph 4.4.1.1 and shown in the table 4-3. The indirect energy consumption depends on the fuel production pathway or electricity production pathway. Each fuel can be produced from one or more resources (or feedstocks) as the source of primary energy. The combination of steps necessary to turn a resource into a fuel and to bring that fuel to the vehicle is denoted as the WTT pathway [114]. These production pathways can be characterised by the net energy consumed, expressed in mega joules per mega joule energy content of the final fuel. This energy consumption excludes the energy content of the resource or feedstock, but aggregates the extra energy consumptions of each fuel production step (extraction, on-site conditioning and storage, transportation, refinery, distribution). The energy expended will be symbolised with the Greek letter χ. The indirect energy consumption can now be calculated from the direct energy consumption as follows:

.indirect directε χ ε= Equation 4-4

Where:

• εindirect = the indirect energy consumption due to the fuel production pathway expressed in mega joules per 100 kilometres

• χ = the energy expended for the considered WTT pathway of the fuel or electricity, expressed in mega joules per mega joule energy content of the final fuel or electricity

• εdirect = the direct energy consumption of the vehicle expressed in mega Joules per 100 kilometres

1 According to the ISO 14040 series on LCA methodology 2 For vehicles running on compressed natural gas, the fuel consumption is expressed in Nm³ per 100 km 3 The fuel density of natural gas is expressed in gram per Nm³ 4 The energy content of natural gas is expressed in kg per Nm³


p.46

An important study in this regard is the WTW analysis of future automotive fuels and powertrains in the European context that was carried out by CONCAWE and EUCAR [114]. This study provides expended energy values χ for a large set of fuel production pathways as well as for electricity production pathways. Some of the values that are of interest for the analysis that will follow are shown in the table below [114]:

Table 4-2: Expended energy for different fuel production options and electricity production

Fuel pathway expended energy χ (MJ/MJ) petrol from crude oil 0,14 diesel from crude oil 0,16 LPG import from remote gas field 0,12 CNG pipeline 4000km 0,19 Biodiesel RME (min) 0,70 Biodiesel RME (max) 1,14 Electricity EU mix 1,87 Electricity wind 0,03

The aggregated WTW energy consumption is then calculated according to equation 4-5:

( ) 6

11 . . . .

10wtw indirect direct f EC FCε ε ε χ ρ= + = + Equation 4-5

Where:

• εwtw = the WTW energy consumption of the vehicle and fuel combination expressed in mega joules per 100 kilometres

Results of this methodology to calculate the well-to-wheel energy consumption are shown in paragraph 4.5 on page 57.

4.4 Methodology for environmental assessment

4.4.1 Emissions inventory In analogy with the life cycle assessment framework, this part of the assessment is denoted as the life cycle inventory analysis (LCI). As mentioned in paragraph 4.2 a WTW framework was chosen, including tailpipe emissions and emissions proportional to the fuel consumption of the vehicle. A large number of factors influence the vehicle’s tailpipe emissions and fuel consumption. The most important factors are the vehicle’s drive train technology (see also paragraph 3.2 on page 27) and the vehicle equipment (e.g. vehicular air-conditioning systems), the traffic situation and the driving behaviour [66] (see also paragraph 3.2.7 on page 33). Furthermore, aging effects of the motor can result in an increase of the emission levels of vehicles over time. Inclusion of in-use-compliance into the homologation directive (directive 98/69/EC) could ensure that emission limits are respected for longer operation times of the vehicle. These variations make it very difficult to compare vehicles with each other. Type approval emission values can present some differences as compared to real vehicle emissions, but provide a common evaluation basis for all vehicles to be assessed. Because this methodology is intended for use as a policy tool, it is important to use emission data that are available for all individual road vehicles. Since 2002, homologation data from passenger cars were collected by the Belgian federal ministry for mobility and transportation. Since 1998, the Belgian federation of the automotive industry “Febiac” owns a vehicle database called “Technicar” including vehicle fuel consumption measurements [115]. Both of these information sources were used for the development of the methodology.


p.47

4.4.1.1 Fuel characteristics The type of fuel used by the vehicle is an important parameter for the environmental assessment. Each fuel type is characterised by its energy content1 (EC) and its density (ρ). The carbon content of a fuel is related to the amount of direct CO2 emissions of the vehicle, while the sulphur content of the fuel determines the quantity of direct SO2 emissions. The fuel characteristics of a range of vehicle fuels are shown in the table 4-3 below [10, 114].

Table 4-3: Fuel characteristics used in the Ecoscore methodology

ENERGY CONTENT

EC (LHV)1 DENSITY

ρ CO2 EMISSION

FACTOR kCO2 MAX. SULPHUR CONTENT Sfuel

SO2 EMISSION

FACTOR kSO2

Petrol 42900 kJ/kg 750 g/ L 2,392 kgCO2/L 50 ppm 0,755 kgSO2/L

Diesel 43000 kJ/kg 835 g/L 2,640 kgCO2/L 50 ppm 0,835 kgSO2/L

LPG 46000 kJ/kg 550 g/L 1,662 kgCO2/L 15 ppm 0,165 kgSO2/L

CNG (G20) 45100 kJ/kg 717 g/Nm³ 1,819 kgCO2/Nm³ 0 ppm 0 kgSO2/Nm³

Biodiesel (RME) 36800 kJ/kg 890 g/L 2,497 kgCO2/L 100 ppm 1,76 kgSO2/L Source: Fuel characteristics CONCAWE/EUCAR 2006

The energy content (EC) and density of the fuel (ρ) are used to calculate the indirect emissions linked to the fuel production. This calculation will be described in more detail in the paragraph 4.4.1.3. In case of light duty vehicles (categories M1 and N1)2 and two-wheelers (categories L1-L6), the consumption of the vehicle is given in litre per 100 kilometres. The vehicles of category M1 are motor vehicles designed for the carriage of passengers and comprising no more than eight seats in addition to the driver’s seat. Vehicles of the category N1 are motor vehicles designed and constructed for the carriage of goods and having a maximum mass not exceeding 3,5 tonnes (70/156/EC). For vehicles running on compressed natural gas, the fuel consumption is expressed in Nm³ per 100 km. In case of electric vehicles the electric energy consumption is expressed in kilowatt-hour (kWh) per 100 km. For heavy duty vehicles (categories M2, M3, N2 and N3)3 the fuel consumption is characterised at the level of the engine and is given in gram of fuel per delivered kilowatt-hour (see paragraph 4.8 on page 69 for more information).

4.4.1.2 Direct Emissions Direct emissions are linked to the use phase of the vehicle. Each vehicle sold on the European market has to be compliant with the type approval test. These tests give information about the so-called regulated emissions: carbon monoxide (CO), hydrocarbons (HC), nitrogen oxides (NOX) and, in the specific case of diesel vehicles, particulate matter (PM). In the case of passenger vehicles and light-duty vehicles (M1 and N1) these emissions are expressed in grams per kilometre. For heavy-duty vehicles (M2, M3, N2 and N3) emission levels are expressed in grams per delivered kilowatt-hour. The latter emissions are evaluated on the level of the combustion engine of the vehicle. The emissions and consumption of the vehicle will thus further be dependent on the application (the same motor can be used in different types of vehicles). Besides the regulated emissions, some unregulated emissions are considered as well: carbon dioxide (CO2), sulphur dioxide (SO2), nitrous oxide (N2O) and methane (CH4). Both carbon dioxide and sulphur dioxide can be calculated starting from the fuel consumption and using the corresponding emission factors from table 4-3. The direct emissions of N2O are mainly dependent of the applied technology. Within the Cleaner Drive project [57] (see paragraph 4.1.2 on page 40) estimations of N2O emissions were performed for vehicles complying with the different “euro emission standards”. For the older vehicles (pre-euro), no emission data are available, so estimations of the emissions and of the consumption were made based on the COPERT III methodology [56]. The vehicle classification as proposed in COPERT III, was used to categorise these older vehicles within the Ecoscore methodology.

1 In this context the lower heating value (LHV) is used (most common in Europe) 2 Vehicle categories as described in European directive 70/156/EC – annex II 3 Vehicle categories as described in European directive 70/156/EC – annex II


p.48

Direct methane (CH4) emissions are technology dependent and emission values resulting from the WTW study of General Motors [59] were used.

4.4.1.3 Indirect Emissions The indirect emissions are those related to the extraction and transportation of the raw materials for the fuel production, together with the emissions linked to refining and distributing the carburant. When considering the use of electric or plug-in hybrid vehicles, emissions related to electricity generation and distribution are taken into account. Indirect emissions are directly proportional to the fuel or energy consumption of the assessed vehicle. The formula used for the indirect emission calculation of the Ecoscore methodology in the case of light duty vehicles or two-wheelers can be found hereafter.

, , 11

1. . . .

3,6.10i j indirect jE F EC FCρ= Equation 4-6

Where:

• Ei,j,indirect = indirect emission value for pollutant j and damage category i, expressed in gram per kilometre

• Fj = Indirect emission factor for pollutant j, expressed in milligram per kilowatt-hour (see table 2)

• ρ = Fuel density, expressed in gram per litre1 • EC = Energy content of the fuel, expressed in kilojoules per kilogram2 • FC = Fuel consumption of the vehicle, expressed in litre per 100 kilometres (in case of light

duty vehicles or two-wheelers) • The factor 1/3,6.1011 in this formula is a conversion factor.

In the case of electric vehicles, the electric energy consumption γ can be expressed in kilowatt-hour per 100 kilometres. The equation 4-6 can be consequently written as equation 4-7:

, , 5

1. .

10i j indirect jE F γ= Equation 4-7

Where:

• Ei,j,indirect = indirect emission value for pollutant j and damage category i, expressed in gram per kilometre

• Fj = Indirect emission factor for pollutant j, expressed in milligram per kilowatt-hour (see table 2)

• γ = the electric energy consumption of the electric vehicle expressed in kilowatt-hour per 100 kilometres

• The factor 1/105 in this formula is a conversion factor. In the case of heavy duty vehicles, the consumption data are given in gram per kilowatt-hour and so the factor ρ in equation 4-6 becomes redundant. The indirect emission factors (Fj) used in equation 4-6 can be found in table 4-4. The conventional fuels (petrol, diesel and LPG) are characterised using emission factors from the European MEET study [116]. The fuel consumption of natural gas vehicles is given in cubic meter per 100 kilometres. Because of important variations in the energy content of natural gas (poor and rich), the natural gas composition in Belgium was compared with different available gas data. The most comparable natural gas was G20, so this type was selected. Indirect emissions data related to biodiesel fuel are based on the WTW study of General Motors [59], as far as CO2 emissions are concerned, and on the MEET study [8] for the other emissions.

1 For vehicles running on compressed natural gas, the fuel consumption is expressed in Nm³ per 100 km 2 The energy content of compressed natural gas is expressed in kg per Nm³


p.49

Table 4-4: Indirect emission factors Fj for different fuel options

CO NMHC NOX PM CO2 SO2 N2O CH4

FUEL TYPE mg/kWh mg/kWh mg/kWh mg/kWh mg/kWh mg/kWh mg/kWh mg/kWh

Petrol 18,4 761,4 151,9 8,6 33120 236,2 0 62,6

Diesel 16,6 315,4 129,6 3,6 24480 174,2 0 56,5

CNG (G20) 5,0 99,0 38,2 2,9 14759 60,8 0 805,3

LPG 14,8 202,7 116,3 5,4 21600 114,1 0 58,0 Biodiesel (RME)

493,2 280,4 871,9 66,6 -172786 245,5 0 0

Elec. Renew. 0 0 0 0 0 0 0 0

Elec.CCGT95 78 129 495 0 447500 0 0 266

Elec. Belg. 2003

30 44 392 42 277683 388 1,558 3,571

Source: based on MEET 1995, VITO 2005 and Electrabel 2003

Where “Elec. Renew.” stands for electricity produced from renewable energy sources, “Elec. CCGT95” stands for electricity produced from a combined cycle gas turbine anno 1995 and “Elec. Belg. 2003” stands for the Belgian electricity mix anno 2003. These emission factors are related to the electricity production of Electrabel inclusiding transportation and distribution. However these figures exclude the emissions related to the production of the energy source used to supply the electricity plant (e.g. the emissions related to the production and supply of natural gas for a CCGT plant).

4.4.2 Classification of the damage categories In accordance to the LCA framework this part of the assessment can be seen as the first step in the life cycle impact assessment (LCIA). The classification can be described as the determination of the flows that are taken into account for the impact assessment for the different considered damage categories considered. In this way LCI results can be classified into impact or damage categories in accordance to ISO 14042. Some environmental assessments stop at the level of the inventory phase and do not perform an LCIA. Other assessments continue with a calculation of the damages. A distinction can be made between mid-point damage categories and end-point damage categories. For midpoint damage categories, the quantification of the impact is made rather early in the cause-effect chain and LCI results are grouped in common mechanisms such as global warming, ecotoxicity or respiratory effects. Endpoint damage categories try to describe the cause-effect chain up to the endpoints which are: human health, climate change, ecosystem quality. A disadvantage of the latter approach is that the quantification of endpoint damages often are characterised by high uncertainties [117]. The environmental assessment methodology proposed in this framework encompasses three main damage categories:

• Global warming • Air quality • Noise

The damage category “air quality” is further subdivided into:

• Effects on human health • Effects on ecosystems

The damage categories global warming and noise can be seen as midpoint categories, while the categories human health and ecosystems are endpoint categories. This is mainly determined by the availability and type of damage factors used for the considered impact pathways.


p.50

Each of the considered damage categories is shortly described in the remainder of this paragraph. The different damage categories will be combined to a total impact as will be further explained in paragraph 4.4.5 on page 54. The different pollutants that are taken into account for the different damage categories are schematically represented in figure 4-4.

Figure 4-4: Classification of the pollutants to the different damage categories

4.4.2.1 Global Warming On the 16th of February 2005, the Kyoto-protocol has officially come into application. It has the ambition to reduce the global emissions of greenhouse gases with at least 5 percent within the time period 2008-2012, compared to the level of 1990. This included a reduction of 8 percent for the European Union and of 5,2 percent for the Flemish region [118]. Flemish industry and agriculture have succeeded in decreasing their emissions, but this reduction was largely compensated by the continuous increase of the CO2 exhaust in the transport sector, mainly in road transportation [3]. Global warming contributes to the rise of the sea level, to an increased occurrence of extreme weather conditions. It leads to shifts and extinctions of biotopes and to potable water shortage. Possible effects on human health are among others: increased occurrence of respiratory effects and cardiovascular diseases due to heat waves and increased number of infections and diseases due to floods [119]. CO2 is the largest contributor to global warming and is mainly originating from the combustion of fossil fuels used for human activities. Nitrous oxide (N2O) and methane (CH4), both emissions exhausted among others by road vehicles, are also contributing to climate change. See also table 1-3 on page 15 for the relative contribution (GWPs) of these greenhouse gases to the global warming effect.

4.4.2.2 Air Quality Carbon monoxide causes oxygen shortage and can lead to suffocation. The concentration of CO in the atmosphere can vary strongly and depends mainly on the traffic situation and on the wind. Carbon monoxide concentrations in the blood of people living in the city can be twice as high as concentrations in the blood of people living in the countryside. The concentration of CO can reach very high levels in closed spaces like underground parking lots, or inside vehicles [18]. Nitrogen oxides (NOX) are forming a group of important pollutants mainly produced by road transportation. Nitrogen oxides are forming acids, when absorbed in the mucous membranes of the nose or the oral cavity, and cause irritation of the bronchial tubes, coughing, and in higher


p.51

concentrations also lack of breath and even death [13]. People suffering from asthma can be very sensitive to NOX. Nitrogen oxides also lead to eutrophication of soil, ground- and surface water and as a result lead to negative impacts on aquatic and terrestrial ecosystems, surface water and agricultural and forestry yields [9]. Particulate matter (PM10) can remain in suspension in the air for hours or even days. As these particles are very small, they can penetrate very deeply into the lungs. Because PM10 is often bound to other harmful substances, exposure and inhalation can lead to an important number of health problems and even to death [13]. Sulphur oxides (SOX) are very soluble in water and can consequently easily be absorbed through the mucous membrane of the bronchial tubes. People suffering from asthma are especially sensitive to sulphur oxides. Furthermore, sulphur dioxide causes acidification, which in turn damages aquatic and terrestrial ecosystems, surface water, agricultural and forestry yields and buildings [9]. Volatile organic compounds (VOC) is the name regrouping a large number of chemical compounds, like toluene, xylene, benzene, etc. Some of them present some important health damaging effects such as carcinogenicity. Some of the previously described emissions also have indirect effects as they lead to other chemical reactions, once emitted into the atmosphere. The formation of tropospheric ozone for instance is triggered by sunlight and is an important secondary pollution of nitrogen oxides and hydrocarbons emissions (photochemical pollution). Ozone is toxic for living organisms and human beings and causes damage to the cells of the bronchial tubes.

4.4.2.3 Noise Noise is one of the main annoyances related to traffic, especially for people living in city centres or close to busy traffic roads. During the last decades, an important deterioration of the ambient sound level was observed. This was accompanied by increased damages on human health due to sleep disturbance and deteriorated audibility during the daytime. The exact influence of noise on human health is hard to quantify, but long-lasting exposure to high sound levels can lead to direct effects like: quickened heartbeat rate, increased blood pressure, cardiovascular diseases, colitis, stomach ulcers, headache and dilatation of the pupils. Indirect health effects are triggered because noise leads to nervous tension and stress.

4.4.3 Characterisation of the damage effect In this part of the environmental assessment methodology, the inventoried and classified emissions for each damage category (the LCI results) will be used to determine the impact. In other words, it will make the link between the inventoried emissions and the environmental impacts. In accordance to the life cycle assessment framework, this part of the assessment is the next step of the life cycle impact assessment (LCIA) after the classification (see paragraph 4.4.2). This step of the methodology allows weighting the impact of the different emissions to a certain damage or effect considered. For instance, different greenhouse gases (e.g. CO2, CH4, N2O) contribute to the global warming effect to a different extent. For the impact assessment different methodologies exist. Midpoint characterization factors are based on equivalency principles, i.e. midpoint characterization scores are expressed in kg-equivalents of a substance compared to a reference substance (for example kgeq CO2 into air). End-point characterization factors of any substance (for example DALY) can be obtained by multiplying the midpoint characterization potentials with the damage characterization factors of the reference substances [117]. In some impact calculation methodologies an additional step is made in which the physical damage or impact is translated into costs. These costs represent external costs. External costs related to a product are those costs to society (environmental damage, congestion, health effects…) that are not included in the purchase price of the product. The emerging of external costs is based on the principle that the welfare of society can be optimized if external costs are internalized. Currently, European studies and programs are being performed regarding external costs and more specifically to external costs of transport. One of the studies trying to quantify is the earlier mentioned ExternE project [92].


p.52

Depending on the considered damage category, different impact factors were used for the characterisation of the damage due to both the indirect and direct emissions. The damage caused to human health is related to the location of the emissions, thus requiring different impact factors for indirect and direct emissions. The calculation of the partial damage D of each pollutant can be represented by the following equation:

, , , , , , ,. .i j i j indirect j indirect i j direct j directD E Eδ δ= + Equation 4-8

Where:

• Di,j = partial damage of pollutant j to the category i • δi,j = impact factor of pollutant j to the category i • Ej = total amount of emissions of pollutant j contributing to the category i

The total damage Q of each damage category can be obtained adding up the partial damages for the different damage categories:

,i i jj

Q D= Equation 4-9

Where:

• Qi = total damage of the category i • Di,j = partial damage of pollutant j to the category i

The contributions of the different greenhouse gases to global warming are calculated using the 100 years time horizon global warming potentials (GWP), as defined by the IPCC in their third assessment report [120]. As was described in paragraph 4.4.1, starting at page 46, the inventoried greenhouse gases for the assessment methodology for road vehicles are carbon dioxide, methane and nitrous oxide. The impact factors are not different for indirect and direct greenhouse gas emissions and are listed for the three considered pollutants in table 4-5.

Table 4-5: Impact factors for the global warming category

Greenhouse gas j δglobal warming, j

Carbon dioxide (CO2) 1 GWP Methane (CH4) 23 GWP Nitrous oxide (N2O) 296 GWP

External Costs were used to allocate a weighting for the contributions of the different inventoried air quality depleting emissions to quantify the effects on human health and the effects on ecosystems. For the human health damage category, external costs are used and are based on the European project ExternE [11] with updated values (baseline 2000), as described in [92, 121]. These external costs are values expressed in monetary terms per kilogram of emission of a certain pollutant, and reflect the overall damage cost of the effect of the inventoried emission on human health. The external costs used for the damage calculation are obtained by a contingent valuation method1 and are available for both urban and extra-urban (or rural) situations. An overview of these specific external costs can be found below in table 4-6:

1 Contingent valuation is an economical valuation technique based on questionnaires to stated preferences.


p.53

Table 4-6: Specific external costs (SEC) for human health effects

Air Quality pollutant j SEChuman health, j, rural SEChuman health, j, urban

Hydrocarbons (HC) 3 €/kg 3 €/kg Carbon monoxide (CO) 0,0008 €/kg 0,0032 €/kg Particulate matter (PM10) 103,49 €/kg 418,61 €/kg Nitrogen oxides (NOX) 1,152 €/kg 1,483 €/kg Sulphur dioxide (SO2) 6,267 €/kg 14,788 €/kg

A weighted average of urban and rural external costs is used for the Ecoscore assessment methodology, using the national split between urban and rural mileage as a weight factor for each category of vehicles (light duty, heavy duty and two-wheelers). The required adaptations for the assessment of heavy duty road vehicles and motorized two-wheelers can be found in paragraph 4.8 on page 69 and paragraph 0 on page 72 respectively. The average mileage distribution (anno 2004) was obtained from the Belgian National Institute of Statistics (NIS) and is given in table 4-7 below.

Table 4-7: Overview of average mileage distribution for different vehicle categories (anno 2004)

MILEAGE DISTRIBUTION (σ) RURAL URBAN Light duty (M1, N1) 75% 25% Heavy duty (M2, M3, N2, N3) 90% 10% Two-wheelers (L1-6) 60% 40%

In case of the damage categories human health and ecosystems, the impact factors δ, as used in equation 4-8, can now be calculated as the weighted average of urban and rural specific external costs (SEC) following equation 4-10 and equation 4-11.

, , , ,i j indirect i j ruralSECδ = Equation 4-10

and

, , , , , ,. .i j direct urban i j urban rural i j ruralSEC SECδ σ σ= + Equation 4-11

Where:

• δi,j = impact factor of pollutant j to the category i • σurban/rural = urban/rural mileage distribution percentage • SECi,j,urban/rural = urban/rural specific external cost of pollutant j to the category i

The specific external costs from table 4-6 in combination with the mileage distribution for light duty vehicles from table 4-7 following the equation 4-10 and equation 4-11 result in the impact factors for the damage category of human health effects described in table 4-8.

Table 4-8: Impact factors δ for human health effects

Air Quality pollutant j δhuman health, j, indirect δhuman health, j, direct

Hydrocarbons (HC) 3 €/kg 3 €/kg Carbon monoxide (CO) 0,0014 €/kg 0,0032 €/kg Particulate matter (PM10) 182,27 €/kg 418,61 €/kg Nitrogen oxides (NOX) 1,235 €/kg 1,483 €/kg Sulphur dioxide (SO2) 8,397 €/kg 14,788 €/kg

For the damage calculation of impacts on ecosystems due to acidification and eutrophication, external costs are used. Abatement costs of emission reductions for NOX and SO2, as presented by Vermoote and De Nocker [122], are used. The values for the impact factors for effects on ecosystems can be found in the table below:


p.54

Table 4-9: Impact factors δ for effects on ecosystems

Air Quality pollutant j δecosystems, j, indirect δecosystems, j, direct

Nitrogen oxides (NOX) 0,176 €/kg 0,176 €/kg Sulphur dioxide (SO2) 0,113 €/kg 0,113 €/kg

The damage calculation for noise is somewhat different compared to the calculation for damage due to emissions of pollutants. The logarithmic decibel scale is used to describe emitted sounds. The A-weighting is used, since it takes into account the sensitivity of human hearing. In this methodology, the inventoried noise level is decreased with a base value of 40 dB(A), corresponding to a non-disturbing background sound level, to obtain values proportional to the inconveniences. The calculation of noise related damage is given through equation 4-12.

40 ( )noise noise noiseQ D E dB A= = − Equation 4-12

4.4.4 Normalisation: reference vehicle To quantify the relative severity of the evaluated damages of each damage category, a normalisation step based on a specific reference value is performed. The damage associated with a theoretical vehicle was taken as the reference point.

,

ii

i ref

Qq

Q=

Equation 4-13

Where:

• qi = normalised damage on category i • Qi = total damage of the assessed vehicle on category i • Qi,ref = total damage of the reference vehicle on category i

This reference vehicle can be considered to be a vehicle with target emission values. Therefore it was chosen to take the Euro 4 emission limits for passenger cars, introduced by directive 98/69/EC as a reference for the light vehicles range. The reference vehicle is considered to be a petrol car. As far as CO2 emissions are concerned, the reference level of 120 grams per kilometre was considered. It corresponds to the target value of the European Union to reach its overall CO2 emissions reduction goal by the year 2012. The noise emission reference has been set at 70dB(A). An overview of the emission levels of the reference vehicle is provided in the table 4-10 below:

Table 4-10: Tank-to-Wheel and Well-to-Tank emissions of the reference vehicle for passenger cars and light duty vehicles

Sound CO2 N2O CH4 CO HC NOX PM10 SO2 FC [dB(A)] [g/km] [g/km] [g/km] [g/km] [g/km] [g/km] [g/km] [mg/km] [L/100km]

TTW 70 120 0,005 0,02 1 0,1 0,08 0 3,79 5,02 WTT - 14,85 0 0,028 0,0081 0,34 0,068 0,004 0,106 (0,4484)1

4.4.5 Weighting system The final step of the methodology consists in the weighting of the different damage categories, before aggregating them to obtain the total impact (TI) of the vehicle to be assessed. This total impact is a

1 This value is the energy consumption related to the TTW fuel consumption, expressed in kWh per km, and is multiplied with the different indirect emission factors Fj to calculate the WTT emissions


p.55

single indicator that represents the overall environmental damage of a road vehicle, within the boundaries of the methodology. However, this step also reduces information and represents a simplification of the assessment made. Nonetheless a single metric such as the total impact or the Ecoscore (see later) can be useful and valuable for information purposes. No scientific consensus is achieved regarding the valuation of the different impacts to provide consumers with information [42]. However, the presented methodology incorporates a weighting system that allows policy makers and all stakeholders concerned to allocate relative importance to the different damage categories. This (subjective) weighting is implemented with equation 4-14 and equation 4-15.

100. .i ii

TI qα= Equation 4-14

With

1ii

α = Equation 4-15

Where:

• TI = total impact of the assessed vehicle • αi = weighting factor of damage category i • qi = normalised damage of category i

The reference vehicle itself presents a total impact of 100. A vehicle, with higher or lower emission levels when compared to the reference vehicle, will have a total environmental impact higher, respectively lower than 100. The weighting factors (αi), as used in the Ecoscore methodology, are based on a weighting method allowing reflecting policy priorities and decision maker’s opinions. The weighting factors were determined by a stakeholder group including representatives from governmental administrations, political parties, the automotive sector, environmental NGOs, consumer organisations and others. They can be found in table 4-11. For communication purposes towards a broad public, it is important to use a score that is easy and quick to understand. That’s why the total impact (TI) is transformed into a score ranging from 0 to 100, 0 representing an infinitely polluting vehicle and 100 indicating an emission free and silent (below or equal to 40dB(A)) vehicle. The reference vehicle corresponds to an Ecoscore of 70. The transformation is based on an exponential function (see equation 4-16), so it can not deliver negative scores (see figure 4-5).

0,00357.Ecoscore = 100.e TI− Equation 4-16


p.56

0

10

20

30

40

50

60

70

80

90

100

0 50 100 150 200 250 300 350 400 450 500

Total Impact

Ecoscore

Reference Vehicle

Ecoscore = 100.exp (-0,00357. TI)

Figure 4-5: Transformation of Total Impact to Ecoscore

4.4.6 Overview of the methodology In figure 4-6 the Ecoscore methodology is shown schematically. The successive steps of the methodology, as described above, can be identified.

Ei,j, indirect

Ei,j, direct

Global Warming

Tank-to-Wheel (direct)

Regulated Emissions

• CO • NOX • HC • PM

Non-regulated Emissions • CO2 • N20 • CH4 • SO2

Well-to-Tank (indirect)

• CO • NOX • HC • PM

• CO2 • N20 • CH4 • SO2

Fuel consumption Indirect emission factors

Noise emission

Air Quality

Human Health

Ecosystems

Noise

Well-to-Wheel

Step 2: Classification Step 1: Inventory

Ej, indirect

Ej, direct

Figure 4-6: Ecoscore Methodology Overview (step 1 & 2)


p.57

Step 3: Charcterisation

Impact factors

Step 4: Normalisation

Qi/Qi,ref

Reference vehicle

Step 5: Weighting

Σαi.qi ΣDi,j

Weighting factors α

qi

Total Impact

Qi

Qi,ref αi

Partial Damage

Calculation

δi,j, indirect

δi,j, direct

Ei,j, indirect

Ei,j, direct

Di,j

Figure 4-7: Ecoscore Methodology Overview (step 3, 4 & 5)

Table 4-11 shows an overview of the parameters used for the Ecoscore methodology. The different damage categories are given, with their contribution to the end score (weighting), their different contributing pollutants (inventory) and their damage factors (characterisation).

Table 4-11: Summary of the parameters used for the Ecoscore methodology

CLASSIFICATION WEIGHTING INVENTORY UNITS CHARACTERISATION α rural urban

CO2 GWP 1 1 CH4 GWP 23 23

1) Global Warming 50%

N2O GWP 296 296 2) Air Quality (40%)

HC €/kg 3 3 CO €/kg 0,0008 0,0032 PM10 €/kg 103,49 418,61 NOX €/kg 1,152 1,483

2a) Human Health 20%

SO2 €/kg 6,267 14,788 NOX €/kg 0,176 0,176 2b) Ecosystems 20% SO2 €/kg 0,113 0,113

3) Noise 10% Sound level dB(A) x-40

4.5 Results for the well-to-wheel energy consumption As an illustration of the WTW energy consumption methodology as described in paragraph 4.3 on page 45, a comparative analysis has been made for a selection of vehicles. The details of these vehicles are shown in table 4-12. All vehicles are complying with the Euro 4 emission limits and have a cylindrical capacity of about 1600 cm³. The electric car is equiped with an air cooled 20 kW d.c. motor (Leroy Somer). The mentioned consumption data are based on the NEDC cycle, except for the battery electric vehicle. The energy consumption of the electric vehicle is based on real life use (and thus can be considered as pessimistic compared to the other consumption data).


p.58

Table 4-12: Vehicle selection details for Well-to-Wheel energy consumption analysis

EMISSION STANDARD FUEL USE BRAND – TYPE ENGINE DISPLACEMENT CONSUMPTION

[cc] [per 100km]

Euro 4 Petrol VOLKSWAGEN GOLF 1595 7,0 L

Euro 4 Diesel OPEL ASTRA 1686 4,6 L

Euro 4 LPG OPEL VECTRA 1598 9,8 L

Euro 4 CNG OPEL Astra Caravan 1600 6,42 Nm³

Euro 4 Hybrid Petrol

TOYOTA PRIUS 1497 4,3 L

Electric Electricity PEUGEOT 106 Electric - 17 kWhe

The results of the WTW energy consumption are shown in figure 4-8. From these data an important advantage for the hybrid petrol vehicle can be observed. The influence of the production pathway for biodiesel and for electricity on the WTT energy consumption is clearly visible. The vehicles that run on LPG and on CNG have a noticeable higher WTW energy consumption compared to the petrol vehicle. The diesel vehicle presents a lower energy consumption than the petrol vehicle. In this comparison, the battery electric vehicle, using electricity produced from renewable wind energy, performs the best in terms of energy consumption. The second best option seems to be the hybrid petrol vehicle.

32 19 27

118

191

2946

114

226

139

169

168

168

243241

61

2

61

0

50

100

150

200

250

300

350

400

Petrol Hybrid Petrol Diesel Biodiesel (min) Biodiesel (max) LPG CNG Battery Electric(EU mix)

Battery Electric(wind)

WT

W (

MJ

/10

0km

)

TTW energy consumption (MJ/100km) WTT energy consumption (MJ/100km)

Figure 4-8: Well-to-wheel energy consumption for passenger vehicles

4.6 Results for the environmental assessment methodology To demonstrate the applicability of the methodology, a set of 22 passenger vehicles has been selected. Vehicles using different types of fuel and complying with different emission standards were chosen. The selected cars have engine capacities which are representative for the national car fleet. This selection allows to analyse the evolution of the environmental impact of vehicles complying with different emission regulations and to analyse the differences between vehicles using different fuel types and drive train technologies. Characteristics of the selection of vehicles [10] are detailed in table 4-13.


p.59

Table 4-13: Vehicle selection details for environmental assessment methodology analysis

EMISSION STANDARD FUEL USE BRAND – TYPE ENGINE DISPLACEMENT CONSUMPTION

[cc] [per 100km]

PRE ECE Petrol No specific vehicle – based on COPERT III 1400 - 2000 14,2 L

ECE 15/00-15/01 Petrol No specific vehicle – based on COPERT III 1400 - 2000 10,1 L

ECE 15/02 Petrol No specific vehicle – based on COPERT III 1400 - 2000 9,2 L



PRE EURO Diesel No specific vehicle – based on COPERT III > 2000 7,5 L

PRE EURO LPG No specific vehicle – based on COPERT III > 2000 9,1 L

Euro 1 Petrol No specific vehicle – based on COPERT III 1400 - 2000 9,7 L

Euro 1 Diesel No specific vehicle – based on COPERT III > 2000 6,7 L

Euro 1 LPG No specific vehicle – based on COPERT III > 2000 9,3 L

Euro 2 Petrol No specific vehicle – based on COPERT III 1400 - 2000 9,7 L

Euro 2 Diesel No specific vehicle – based on COPERT III > 2000 6,7 L

Euro 2 LPG No specific vehicle – based on COPERT III > 2000 9,3 L



Euro 3 LPG TOYOTA AVENSIS 1598 9,6 L



Euro 4 LPG OPEL VECTRA 1598 9,8 L

Euro 4 CNG OPEL ASTRA Caravan 1600 6,42 Nm³

Euro 4 Hybrid Petrol

TOYOTA PRIUS 1497 4,3 L

Battery-electric Electricity PEUGEOT 106 Electric - 17 kWhe

4.6.1 Well-to-Wheel emissions inventory A first analysis that can be made is the inventory of the emission levels and the classification of these emissions following the earlier described well-to-wheel methodology. This allows evaluating the emissions of the set of vehicles on a WTW basis and classified for the different damage categories as described in paragraph 4.4.2. A first group of emissions are the GHGs contributing to the global warming damage category. In figure 4-9 these greenhouse emissions for the set of vehicles from table 4-13 are shown. The emissions are differentiated per pollutant type and also per stage. The emissions are expressed in grams per kilometre and two important sources in terms of mass of emissions are identified. The TTW CO2 emissions and the WTT CO2 emissions have the largest contribution to the total mass of GHGs. The contribution of these gases on global warming depends on the damage factors and can be largely different for the different considered emissions of GHGs.


p.60

0

50

100

150

200

250

300

350

400

Pe

tro

l PR

E E

CE

Pe

tro

l EC

E 1

5/0

0-0

1

Pe

trol

EC

E 1

5/0

2

Pe

trol

EC

E 1

5/0

3

Pe

trol

EC

E 1

5/0

4

Die

sel P

re E

uro

LP

G P

re E

uro

Pe

tro

l Eu

ro 1

D1e

sel E

uro

1

LP

G E

uro

1

Pe

tro

l Eu

ro 2

Die

sel E

uro

2

LP

G E

uro

2

Pe

tro

l Eu

ro 3

Die

sel E

uro

3

LP

G E

uro

3

Pe

tro

l Eu

ro 4

Die

sel E

uro

4

LP

G E

uro

4

CN

G E

uro

4

Hyb

rid

Pe

tro

l Eur

o 4

Ba

tter

y E

lect

ric

Gre

en

ho

us

e g

as

em

issi

on

s (

g/k

m)

WTT CH4WTT N2OWTT CO2TTW CH4TTW N2OTTW CO2

Figure 4-9: Greenhouse gases for the vehicle selection – split up per pollutant and per stage

A second group of pollutants is the group of air quality depleting emissions, which have an effect on human health and on ecosystems. In figure 4-10 these emissions for the set of vehicles from table 4-13 are shown. Again, the emissions are expressed in grams per kilometre and a dominant pollutant for the old vehicles (Pre-euro), namely the TTW emission of carbon monoxide is recognized. The direct emissions of carbon monoxide were strongly lowered and this was induced by the introduction of the European directive 93/59/EC on gaseous emissions from road vehicles, defining the Euro 1 emission limits. In figure 4-11 the same emissions are shown but with an adapted scale for the Euro 1 and later vehicles.

0

5

10

15

20

25

30

35

40

Pe

tro

l PR

E E

CE

Pe

tro

l EC

E 1

5/0

0-0

1

Pe

trol

EC

E 1

5/0

2

Pe

trol

EC

E 1

5/0

3

Pe

trol

EC

E 1

5/0

4

Die

sel P

re E

uro

LP

G P

re E

uro

Pe

tro

l Eu

ro 1

D1e

sel E

uro

1

LP

G E

uro

1

Pe

tro

l Eu

ro 2

Die

sel E

uro

2

LP

G E

uro

2

Pe

tro

l Eu

ro 3

Die

sel E

uro

3

LP

G E

uro

3

Pe

tro

l Eu

ro 4

Die

sel E

uro

4

LP

G E

uro

4

CN

G E

uro

4

Hyb

rid

Pe

tro

l Eur

o 4

Ba

tter

y E

lect

ric

Air

qu

alit

y g

as

em

iss

ion

s (g

/km

)

WTT SO2

WTT NOX

WTT PM

WTT CO

WTT HC

TTW SO2

TTW NOX

TTW PM

TTW CO

TTW HC

Figure 4-10: Air quality depleting emissions for the vehicle selection – split up per pollutant and per stage


p.61

0

1

2

3

4

5

6

7

8P

etr

ol P

RE

EC

E

Pe

tro

l EC

E 1

5/0

0-0

1

Pe

trol

EC

E 1

5/0

2

Pe

trol

EC

E 1

5/0

3

Pe

trol

EC

E 1

5/0

4

Die

sel P

re E

uro

LP

G P

re E

uro

Pe

tro

l Eu

ro 1

D1e

sel E

uro

1

LP

G E

uro

1

Pe

tro

l Eu

ro 2

Die

sel E

uro

2

LP

G E

uro

2

Pe

tro

l Eu

ro 3

Die

sel E

uro

3

LP

G E

uro

3

Pe

tro

l Eu

ro 4

Die

sel E

uro

4

LP

G E

uro

4

CN

G E

uro

4

Hyb

rid

Pe

tro

l Eur

o 4

Ba

tter

y E

lect

ric

Air

qu

alit

y g

as

em

iss

ion

s (g

/km

)

WTT SO2

WTT NOX

WTT PM

WTT CO

WTT HC

TTW SO2

TTW NOX

TTW PM

TTW CO

TTW HC

Figure 4-11: Air quality depleting emissions for the vehicle selection – with adapted vertical scale

Figure 4-12 illustrates the total impact, split up per damage category, of this selection of passenger vehicles. This figure allows the comparison of the different damage categories for each vehicle. When considering global warming, diesel vehicles present a lower impact compared to their petrol counterparts. Thanks to their efficient drive train, of all conventional vehicles, diesel vehicles contribute least to global warming. The use of electric vehicles however, is characterised by the lowest impact on global warming. Hybrid electric drive trains make it possible to lower the fuel consumption of thermal engines and thus to reduce their impact on global warming [123]. Furthermore, fuel efficiency has improved with time. Due to the high impact of particulate matter on human health, diesel vehicles perform significantly worse than all other vehicles regarding this damage category. However, this difference in impact is decreasing thanks to the latest emission regulations (Euro 4 and later) and could be improved further in the future, with more stringent emission limits, imposing the use of particle filters. The lower impact of LPG and especially CNG vehicles, compared to a petrol vehicle, is mainly due to lower indirect emissions combined with lower levels of air quality depleting emissions. Also, a positive evolution can be observed regarding the impact on ecosystems. This is due to the reduction of the NOX emissions thanks to the use of advanced injection techniques. The NOX exhaust level could be improved further by implementing DeNOX catalysts, selective catalytic reduction (SCR) systems or NOX absorbers [124]. Also the lower sulphur content of the fuels (50 ppm) has had a beneficial influence on SO2 emissions during the last decade. A further reduction to 10 ppm of the sulphur content of diesel and petrol fuels was available starting from January 2007. As far as noise pollution is concerned, the consecutive noise directives have reduced the pollution levels. Hybrid and electric vehicle technologies show an even lower level of noise pollution than the most recent conventional technologies.


p.62

37,337,1

63,264,569,5

75,8

85,3

75,4

50,0

68,4

62,9

55,0

62,5

50,3

29,9

16,327,3

23,915,3

27,6

69,4

27,0

0

100

200

300

400

500

600

Petr

ol P

RE E

CE

Petr

ol E

CE 1

5/00

-01

Petr

ol E

CE 1

5/02

Petr

ol E

CE 1

5/03

Petr

ol E

CE 1

5/04

Die

sel P

re E

uro

LPG

Pre

Eur

o

Petr

ol E

uro

1

Die

sel E

uro

1

LPG

Eur

o 1

Petr

ol E

uro

2

Die

sel E

uro

2

LPG

Eur

o 2

Petr

ol E

uro

3

Die

sel E

uro

3

LPG

Eur

o 3

Petr

ol E

uro

4

Die

sel E

uro

4

LPG

Eur

o 4

CNG

Eur

o 4

Hyb

rid

Petr

ol E

uro

4

Batt

ery

Elec

tric

Tota

l im

pact

0

10

20

30

40

50

60

70

80

90

100

Ecos

core

Noise

Ecosystems

Human Health

Global Warming

EcoScore

Figure 4-12: Total impact for the vehicle selection – category split up

Figure 4-13 shows the total environmental impact split up according to the contribution from the WTT emissions and the TTW emissions. From this analysis it can be observed that the TTW or direct impact is dominant for all kinds of vehicles except for the battery electric vehicle which is characterised by no direct emissions. In case of the petrol vehicle, the WTT or indirect contribution to the total impact is larger compared to the other kinds of vehicle.

0

50

100

150

200

250

300

350

400

450

500

550

600

Petr

ol P

re E

CE

Petr

ol E

CE 1

5/00

-01

Petr

ol E

CE 1

5/02

Petr

ol E

CE 1

5/03

Petr

ol E

CE 1

5/04

Die

sel P

re E

uro

LPG

Pre

Eur

o

Petr

ol E

uro

1

Die

sel E

uro

1

LPG

Eur

o 1

Petr

ol E

uro

2

Die

sel E

uro

2

LPG

Eur

o 2

Petr

ol E

uro

3

Die

sel E

uro

3

LPG

Eur

o 3

Petr

ol E

uro

4

Die

sel E

uro

4

LPG

Eur

o 4

CNG

Eur

o 4

Hyb

rid

- Pet

rol E

uro

4

Batt

ery

Elec

tric

Tota

l Im

pact

WTT impact

TTW impact

Figure 4-13: Distribution WTT and TTW environmental impact


p.63

4.6.2 Range of Ecoscore for vehicle database In the framework of the Ecoscore project, a large database with all registered vehicles in Belgium was available. This database was used to make the analyses of the methodology as described earlier. To have an idea about the range of the Ecoscore results of all passenger vehicles from the database, the results were classified per fuel type and per emission standard class (Euro 4, Euro 3…). The range of Ecoscores for these different groups of vehicles was presented in a chart.

Figure 4-14: Overview range of Ecoscore for passenger vehicles

On figure 4-14 a large range of Ecoscores for the different groups can be seen. Vehicles with larger energy consumption (e.g. sports utility vehicles) have a much lower score compared to vehicles with an average or low energy consumption. Consequently a Euro 4 vehicle doesn’t necessarily have a better Ecoscore than a Euro 3 vehicle.

4.6.3 Alternative fuels and drive trains To analyse the potential environmental improvements of vehicles with alternative fuels or drive trains more thoroughly, the last six cars of the selection of table 4-13 were analysed in more detail. Direct and indirect impacts were analysed separately and are shown in figure 4-15. Also the rating (Ecoscore) is indicated on the graph for each vehicle. Because the environmental impact of an electric vehicle is strongly dependent on the electricity production, three different scenarios were worked out: the Belgian electricity production mix, anno 2003 (ElecBelg03), electricity generation with combined cycle gas turbines (ElecCCGT95), and electricity generated with renewable energy sources, yielding no in-use emissions (ElecRenew).


p.64

4,87,627,8

17,04,47,8 5,6

13,3

8,9

4,621,6

9,78,6

1,24,2

10,19,4

9,8 11,1

16,47,3

9,3

9,7

10,7

9,39,3

10,011,3

11,038,1

46,6

53,845,6

55,80,71,8

3,229,6

3,2

1,02,2

7,2

20,87,2

75,8 75,4

85,3 85,0

69,5

96,7

63,264,5

0

50

100

150

200

Petro

l Euro

4

Diese

l Euro

4

LPG Euro

4

CNG Euro

4

Hybrid

Pet

rol E

uro 4

ElecB

elg03

ElecC

CGT95

ElecR

enew

To

tal

imp

act

0

10

20

30

40

50

60

70

80

90

100

Ec

os

co

re

Ecosystems direct

Human Health direct

Global Warming direct

Noise

Ecosystems indirect

Human Health indirect

Global Warming indirect

EcoScore

Figure 4-15: Total impact of different vehicle technologies and electricity production

When comparing the results for an electric vehicle charged with electricity originating from a CCGT plant1, with an electric vehicle charged using the Belgian electricity production mix, an important difference in the exhaust of greenhouse gases can be observed. It’s important to notice that the Belgian electricity production park contains nuclear power plants delivering around 54% (in 2006) of the total electricity production [125]. The potential substitution of nuclear power and coal power plants with new and highly efficient CCGT plants will alter the composition of indirect emissions related to electricity consumption. When charged with electricity originating only from CCGT plants, the impact of electric cars on human health and ecosystems is lower and is nearly totally compensating the large contribution to global warming. Battery electric cars allow the use of renewable electric energy (wind turbines, photovoltaic cells, etc.), reducing the total impact of road transportation on the environment and on human health even further. The direct impacts of LPG vehicles are comparable to those of petrol vehicles, but a substantial advantage compared to petrol can be noticed for the indirect emissions related to the LPG fuel production and transportation. Compressed Natural Gas vehicles show a larger total environmental benefit than LPG vehicles. The benefit is comparable to that of the hybrid electric vehicle. When taking a closer look however, some important differences appear between both technologies. CNG vehicles have a lower indirect impact and the damage to human health and ecosystems is significantly lower than for the hybrid car. On the other hand, the contribution of CNG vehicles to global warming is slightly higher compared to that of the hybrid car. Previous observations demonstrate that the question “which vehicles are environmentally friendly?” is indeed difficult to answer but the Ecoscore methodology presented in this work overcomes this problem and allows the evaluation of the overall environmental performance of road vehicles.

4.6.4 Long term emission abatement Another possible analysis of the environmental impact of passenger vehicles is the long term potential of emission abatement technologies. Based on the results of an assessment of the emission reduction potential for passenger cars, as described by Smokers [126], the Ecoscore of Euro 3 petrol- and diesel-

1 indirect emission data from Electrabel


p.65

engine cars (real world emissions in cold start tests with ambient temperatures of 9°C) and for the respective estimated attainable long term (LT) situation, was calculated. The figure below shows the results of the Ecoscore for these assumptions, together with the Euro 3 and Euro 4 limits and actual emissions measured during the Euro 3 type approval test. For the consumption data, two different situations were considered: vehicles with a mass of 1100 kg and vehicles with a mass of 1300 kg respectively.

Emission abatement for passenger cars

54,9 56,6 57,954,9

69,7

48,3

59,0

53,0

44,9

69,0

51,7 53,4 54,551,7

60,5

46,9

57,3

51,4

43,6

67,7

0

10

20

30

40

50

60

70

Lim

it E

uro

3

Lim

it E

uro

4

Type

Tes

t Eur

o 3

Euro

3 re

al w

orld

Long

term

Lim

it E

uro

3

Lim

it E

uro

4

Type

Tes

t Eur

o 3

Euro

3 re

al w

orld

Long

term

Ecoscorevehicle mass 1100kg

vehicle mass 1300kg

PETROL DIESEL

Figure 4-16: The Ecoscore of long term emissions abatement estimations for passenger cars, based on Smokers et al. (2004)

The Ecoscore of the Euro 3 petrol vehicle, with the emission levels measured during the certification test (Type Test Euro 3), is even better than the Ecoscore corresponding to the Euro 4 emission limits. However, the influence of driving under real world conditions, leads to an Ecoscore which is lower than the Ecoscore corresponding to type test emissions and this results in an Ecoscore (almost) equal to that of the Euro 3 emission limits. The latter results from lower real world NOX emissions, compared to the emission Euro 3 limit, that are compensating the impact of higher real world CO and HC emissions. For the Euro 3 diesel-engine vehicle however, a difference between the Ecoscore of the Euro 3 emission limits and the one corresponding to the real world emissions can be seen. A significant improvement of the environmental performance is observed for the estimated long term emission abatement technologies for both petrol- and diesel-engine vehicles. The long term estimation of the fuel consumption depends on the technology and results in two different reduction percentages [126]. For the highest reductions, leading to the highest Ecoscore, the application of a hybridisation has been considered.

4.6.5 Average Ecoscore of a vehicle fleet Often, the environmental performance of a vehicle fleet instead of individual vehicles is of interest. The environmental performance of the different vehicles constituting this fleet will determine the environmental performance of the fleet. For this reason, the Ecoscore indicator is a useful quantifier and allows the Ecoscore methodology as the framework for the assessment of the vehicle fleet. Further, the number of pieces of each vehicle model is of importance. In most vehicle fleets, certain models of vehicles are occurring multiple times. As an indicator for the environmental performance of a vehicle fleet a weighted average Ecoscore is proposed:


p.66

1Ecoscore .Ecoscorefleet i i

itot

NN

= Equation 4-17

Where:

• Ecoscorefleet = the average Ecoscore of the fleet • Ecoscorei = the Ecoscore of vehicle model i • Ni = number of vehicles of model i in the vehicle fleet • Ntot = total number of vehicles of the vehicle fleet

As an illustration of this average Ecoscore, the environmental performance of the twenty best sold passenger vehicles in Belgium in the first quarter of 2006 was calculated. The number of vehicles that was sold of each model was used for the weighting following the formulation in equation 4-17. This exercise was based on the information available in a publication of October 2006 of the Belgian consumer organisation “Test Aankoop” and based on data from Febiac1. The results of this exercise are presented in figure 4-17. The Ecoscore of each individual model (Ecoscorei) was calculated and is shown in this figure. The weighted average (Ecoscorefleet) was successively calculated using the sales numbers of each model. The fleet in this regard is the group of newly sold and registered vehicles of the first quarter of 2006. The average Ecoscore for this fleet is 64,2 and is shown in the figure by means of a green dotted line. The red line on the figure corresponds to the Ecoscore of the reference vehicle of the methodology and has a value of 70. Only two models have an Ecoscore of 70 or higher.

59 6266

70 68 66 66 66 6763 65 63 65 65

60

71

61 58 59

0

10

20

30

40

50

60

70

80

90

100

Ope

l Ast

ra

VW G

olf

Rena

ult M

egan

e

Peug

eot 2

07

Peug

eot 3

07

Rena

ult C

lio

Citr

oën

C3

Ope

l Cor

sa

Ford

Foc

us

BMW

3 re

eks

VW P

olo

Citr

oën

Xsar

a Pi

cass

o

Toyo

ta C

orol

la

Ford

Fie

sta

Aud

i A4

Citr

oën

C4

Citr

oën

Berl

ingo

Aud

i A6

VW P

assa

t

Ecos

core

0

2.000

4.000

6.000

8.000

10.000

12.000

14.000

16.000Number sold in Belgium (1st semester 2006)

Ecoscore

number of vehicles sold

Figure 4-17: Ecoscore of the 20 best sold passenger vehicles and average Ecoscore

The mileage of the individual vehicles also plays a role in the absolute impact of a vehicle fleet. If this information is available, an environmental assessment of the use of a vehicle fleet can be performed. Again, the Ecoscore can be used as the basis of such an assessment.

1Ecoscore .Ecoscorefleet use i i

itot

MM

= Equation 4-18

1 Belgian federation of the automotive industry


p.67

Where:

• Ecoscorefleet use = the average Ecoscore of the fleet use, based on mileage • Ecoscorei = the Ecoscore of vehicle i • Mi = mileage of vehicles i of the vehicle fleet • Mtot = total mileage of all vehicles of the fleet

A disadvantage of this formulation is that the total mileage of the vehicle fleet is disregarded. However, the total mileage of a vehicle fleet also influences the absolute environmental impact. A possible solution is to consider for all vehicles the Total Impact (TIi) and their mileage (Mi) and to consider the product of both as a new indicator, reflecting the absolute environmental impact of the vehicle (see equation 4-19). The latter can be made to perform assessments on a yearly basis and as a framework for green fleet management.

Environmental indicator fleet use .i ii

M TI= Equation 4-19

Where:

• TIi = the Total Impact of vehicle i • Mi = mileage of vehicles i of the vehicle fleet

4.7 Sensitivity analysis of the environmental assessment methodology A sensitivity analysis was performed to evaluate the robustness of the ranking methodology. Different analyses have been performed, to investigate the influence of the successive calculations on the results of the Ecoscore methodology. The influence of the parameter mileage distribution σ (see paragraph 4.4.3) was investigated. When the urban mileage distribution percentage was increased (and the rural mileage distribution percentage lowered accordingly), an increase of the total impact of the diesel vehicles (in particular the older ones) was observed. However the relative ranking of the different vehicles technologies was not influenced. The influence of the weighting system, parameter α (see paragraph 4.4.5) was also investigated. When changing the obtained weighting factors from table 4-11, to an equal weight for each damage category (αi=25%, for all categories i), no significant influence was observed and no influence on the ranking of the different vehicles occurred. Furthermore, the influence of the impact factors δi,j (see paragraph 4.4.3) was analysed. Therefore the individual contributions Di,j

(see equation 4-8) of each pollutant j to a corresponding damage category i, were each time multiplied by a factor 2 (100% increase of the contribution) for each pollutant and for each damage category. In this way the sensitivity of the environmental ranking methodology to the emission levels or to their damage factors can be assessed. The vehicle selection from table 4-13 was used again to demonstrate the results of the performed sensitivity analysis. The deviation of the total impact (TI), due to the multiplication with a factor 2, for each pollutant and for each damage category, was calculated for all these vehicles. The results of this analysis are shown in figure 4-18.


p.68

-25%

-15%

-5%

5%

15%

25%

35%

45%P

etro

l P

re E

CE

Pet

rol

EC

E 1

5/00

-01

Pet

rol

EC

E 1

5/02

Pet

rol

EC

E 1

5/03

Pet

rol

EC

E 1

5/04

Die

sel

Pre

Eu

ro

LP

G P

re E

uro

Pet

rol

Eu

ro 1

Die

sel

Eu

ro 1

LP

G E

uro

1

Pet

rol

Eu

ro 2

Die

sel

Eu

ro 2

LP

G E

uro

2

Pet

rol

Eu

ro 3

Die

sel

Eu

ro 3

LP

G E

uro

3

Pet

rol

Eu

ro 4

Die

sel

Eu

ro 4

LP

G E

uro

4

CN

G E

uro

4

Hyb

rid

- P

etro

lE

uro

4

Bat

tery

Ele

ctri

c

deviation %

Global Warming - CO2 Global Warming - N2O Global Warming - CH4 Ecosystems - NOx Ecosystems - SO2

-25%

-15%

-5%

5%

15%

25%

35%

45%

Pet

rol

Pre

EC

E

Pet

rol

EC

E 1

5/00

-01

Pet

rol

EC

E 1

5/02

Pet

rol

EC

E 1

5/03

Pet

rol

EC

E 1

5/04

Die

sel

Pre

Eu

ro

LP

G P

re E

uro

Pet

rol

Eu

ro 1

Die

sel

Eu

ro 1

LP

G E

uro

1

Pet

rol

Eu

ro 2

Die

sel

Eu

ro 2

LP

G E

uro

2

Pet

rol

Eu

ro 3

Die

sel

Eu

ro 3

LP

G E

uro

3

Pet

rol

Eu

ro 4

Die

sel

Eu

ro 4

LP

G E

uro

4

CN

G E

uro

4

Hyb

rid

- P

etro

lE

uro

4

Bat

tery

Ele

ctri

c

deviation %

Human Health - NMVOC Human Health - CO Human Health - PM10 Human Health - NOx Human Health - SO2

Figure 4-18: Sensitivity of each impact contribution for the Ecoscore methodology

The largest change of the total impact is obtained for diesel vehicles, when doubling the contribution of particulate matter (PM10) to the human health impact category. This means that in the case of diesel vehicles, the result of the Ecoscore methodology is significantly influenced by this pollutant, because of its important impact on the health of people. This observation is corresponding to the high external cost of particulate matter, used in the damage characterisation. To evaluate the robustness of the methodology and the relative positioning of the assessed vehicles, it is important to characterise the influence of the models’ sensitivity on the end result for the assessed vehicle. The results of the sensitivity analysis are presented in figure 4-19. The environmental indicator Ecoscore was calculated and the range of results (up to their maximum deviation), are indicated with an error bar. These analyses indicate a high robustness of the methodology for vehicles complying with the latest emission levels. Older vehicles and diesel vehicles show a larger sensitivity for errors on emission values and/or damage factors.


p.69

24 28 27 27 30

50

37

6255

37

6963

50

68 64 6370

76 7585

1615

0

10

20

30

40

50

60

70

80

90

100

Petr

ol P

re E

CE

Petr

ol E

CE 1

5/00

-01

Petr

ol E

CE 1

5/02

Petr

ol E

CE 1

5/03

Petr

ol E

CE 1

5/04

Die

sel P

re E

uro

LPG

Pre

Eur

o

Petr

ol E

uro

1

Die

sel E

uro

1

LPG

Eur

o 1

Petr

ol E

uro

2

Die

sel E

uro

2

LPG

Eur

o 2

Petr

ol E

uro

3

Die

sel E

uro

3

LPG

Eur

o 3

Petr

ol E

uro

4

Die

sel E

uro

4

LPG

Eur

o 4

CNG

Eur

o 4

Hyb

rid

- Pet

rol E

uro

4

Batt

ery

Elec

tric

Figure 4-19: Sensitivity analysis – maximum deviation on the end result

Finally, the influence of the exponential transformation (see paragraph 4.4.5) was investigated by comparing it with the use of a linear transformation. The main conclusion is that the introduction of an exponential transformation has no influence on the ranking of the different vehicles assessed with the Ecoscore methodology and has as an advantage that negative scores can be avoided, because very high total impacts will result in an Ecoscore approaching zero. A possible disadvantage is that the differentiation becomes smaller for verry high values of the total impact.

4.8 Methodology adaptations for assessment of engines for heavy-duty vehicles

Besides passenger cars and light duty cars the methodology can be applied for heavy duty vehicles if subjected to some adaptations. Heavy duty vehicles are considered in this regard as the range of the following categories of road vehicles (following directive 70/156/EC – annex III):

• Minibuses: category M2 ( > 8 passenger seats; maximum mass < 5 tonnes) • Buses: category M3 ( > 8 passenger seats; maximum mass > 5 tonnes) • Medium trucks: category N2 (carriage of goods; 3,5 tonnes < maximum mass < 12 tonnes) • Heavy trucks: category N3 (carriage of goods: maximum mass > 12 tonnes)

A first important difference between the assessments of heavy-duty vehicles compared to passenger cars is the direct emission data available for these categories of vehicles. Homologation is made at the level of the heavy-duty engines and not at the level of the vehicle. The difference lays in the units used to express the emissions and the fuel consumption. Emission data of passenger vehicles are expressed in gram per kilometre while the emission data for engines for heavy-duty vehicles are expressed in gram per kilowatt-hour mechanical energy delivered by the engine. The fuel consumption is given in grams of fuel per kilowatt-hour mechanical energy. This is the reason why emission levels from engines for heavy-duty vehicles are not directly comparable with these of passenger vehicles. For the calculation of the direct emissions of methane and nitrous oxide, another approach is used. For heavy-duty vehicles, emission factors from the ACEEE methodology [89], which are related to the CO2 emission level, are used. These emission factors can be found in the table below:


p.70

Table 4-14: N2O and CH4 emission ratio’s for different fuel types

N2O/CO2 ratio CH4/CO2 ratio [%] [%]

Petrol 0,04024 - Diesel 0,00354 0,0000354 Biodiesel (RME) 0,00354 0,0000177 CNG (G20) 0,03159 0,0002124 LPG 0,03159 -

As described in paragraph 4.4.3, a weighted average of the urban and rural external costs is used for the damage calculation. Therefore the national split between urban and rural mileage is used as a weight factor following equation 4-10 and equation 4-11 (see page 52). The average mileage distribution for heavy duty traffic was obtained from the Belgian National Institute of Statistics (NIS) and is given in table 4-7 on page 53. When using the Ecoscore methodology for the assessment of engines for heavy-duty vehicles, a different reference situation is chosen. The Euro III EEV (Enhanced Environmental Vehicle) emission limits, introduced by directive 96/96/EC are used, together with a fuel consumption of 200g/kWh and a noise emission of 76dB(A). The concept of EEV is not related to an obligation or limit but can be used for the award of an environmental label. An overview of these emission data is given in the table 4-15 below:

Table 4-15: Tank-to-Wheel and Well-to-Tank emissions of the reference situation for engines for heavy-duty vehicles

Sound CO2 N20 CH4 CO HC NOX PM10 SO2 FC [dB(A)] [g/kWh] [g/kWh] [g/kWh] [g/kWh] [g/kWh] [g/kWh] [g/kWh] [g/kWh] [g/kWh]

TTW 76 628 0,022 0,022 1,5 0,25 2 0,02 0,02 200 WTT - 58,9 0 0,136 0,040 0,76 0,31 0,009 0,42 2,40

4.8.1 Modifications to assess heavy-duty vehicles The Ecoscore methodology for heavy-duty vehicles as described above, allows an assessment of the engines. Often, the vehicle rather than the engine used in that vehicle needs to be assessed. Therefore, some further adaptations can be made to the methodology. These further adaptations are related to the direct emissions of the heavy-duty vehicle. The fuel consumption of the heavy-duty vehicle needs to be available. In that case, this fuel consumption data can be used to convert the emission data from the engine (Ee ,expressed in gram per kilowatt-hour) to emission data linked to the vehicle (E in gram per kilometre). However, to perform this conversion, information about the engine efficiency is required. For this purpose, an average value of the engine efficiency

ICEη could be used.

, ,5

1. . . . .

3,6.10e

j direct ICE j directE FC EC Eρ η=


p.71

Where: • Ej,direct = direct emission value for pollutant j, linked to the vehicle, expressed in gram per

kilometre • Ee

j,direct = direct emission value for pollutant j, linked to the engine, expressed in gram per kilowatt-hour

• ICEη = average value of the efficiency of the engine

• ρ = Fuel density, expressed in gram per litre1 • EC = Energy content of the fuel, expressed in kilojoules per kilogram2 • FC = Fuel consumption of the vehicle, Expressed in litre per 100 kilometres • The factor 1/3,6.10-5 in this formula is a conversion factor.

This approach was developed and used for the evaluation of different types of urban buses in the study for the MIVB/STIB, the Brussels public transportation company [127]. Besides this adaptation, an additional damage category dealing with damage to buildings was included. Further the weighting factors were adapted to a more urban point of view. A weight of 50% was given to the damage category “effects on human health” reflecting the importance of these effects in an urban context. The reference vehicle has also been customized for this study. For this study, an average efficiency of the internal combustion engine was estimated as follows: 35% for diesel buses and 30% for CNG buses [128]. The damage associated to a Euro IV complying standard-sized (capacity of 65 passengers) bus was taken as a reference point[128]. An overview of the modified methodology for the MIVB/STIB is shown in table 4-16.

Table 4-16: Summary of the parameters used for the adapted Ecoscore methodology for MIVB-STIB

CLASSIFICATION WEIGHTING INVENTORY UNITS CHARACTERIZATIONS α rural urban

CO2 GWP 1 1 CH4 GWP 23 23

1) Global warming 25%

N2O GWP 296 296 2) Air quality (60%)

HC €/kg 3 3 CO €/kg 0,0008 0,0032 PM10 €/kg 103,49 418,61 NOX €/kg 1,152 1,483

2a) Human health 50%

SO2 €/kg 6,267 14,788 NOX €/kg 0,176 0,176 2b) Ecosystems 10% SO2 €/kg 0,113 0,113

3) Noise 10% Sound level dB(A) x-40 x-40 SO2 €/kg 0 8,3 4) Building damage 5% PM €/kg 0 259

For the specific application for the assessment of urban buses, it can be useful to take the capacity of the buses into account in order to assess the environmental impact per transported passenger. In that case the Ecoscore can be used to calculate an environmental impact indicator per transported passenger. This was done in the framework of the MIVB/STIB study, using the following formula [128]:

Bus XCapacity Bus X

Reference

CapacityEcoscore = Ecoscore .

Capacity Equation 4-20

The capacity of the reference vehicle is used in this formula (CapacityReference) to normalise the capacity of the assessed bus and was set at 65 passengers. Finally, when evaluating a fleet of buses, it is important to take into account their mileage. The more intensively a bus is used, the higher its absolute impact on

1 For vehicles running on compressed natural gas, the fuel consumption is expressed in Nm³ per 100 km 2 The energy content of compressed natural gas is expressed in kg per Nm³


p.72

the environment will be. For this purpose in the MIVB/STIB study, the yearly mileage of each bus was combined with the Total Impact of each bus to determine the environmental impact of the bus fleet:

Environmental impact bus fleet Total Impact Bus X .yearly mileage Bus Xi ii

= Equation 4-21

4.9 Methodology adaptations for two-wheelers The assessment methodology described earlier can also be applied for motorized two- (and three-) wheelers. An overview of the emission limits of the regulated direct emissions for motorized two-wheelers is provided in table 4-17 [54]:

Table 4-17: Emission limits for motorized two-wheelers

Norm Directive Entry into force

Category CO (g/km)

HC (g/km)

NOX (g/km)

HC+NOX (g/km)

17/06/1999 L1 (≤50 cc) 6 - - 3 17/06/2002 L1 (≤50 cc) 1 - - 1,2 17/06/1999 L2(≤50 cc) 12 - - 6 17/06/2000 L2(≤50 cc) 3,5 - - 1,2 17/06/1999 L3(>50 cc, 2-stroke) 8 4 0,1 - 17/06/1999 L3(>50 cc, 4-stroke) 13 3 0,3 - 17/06/1999 L4-L5 (>50 cc, 2-stroke) 12 6 0,15

Euro 1 97/24/EC

17/06/2000 L4-L5 (>50 cc, 4-stroke) 19,5 4,5 0,45 01/07/2004 <150cc 5,5 1,2 0,3 - Euro 2

2002/51/EC

01/07/2004 ≥150cc 5,5 1,0 0,3 - 01/01/2007 <150cc 2,0 0,8 0,15 - Euro 3 2002/51/EC 01/01/2007 ≥150cc 2,0 0,3 0,15 -

For the unregulated direct emissions of nitrous oxide (N2O), estimations based on the COPERT methodology [56] can be used. An emission of 0,001 gN2O/km for mopeds (cylindrical capacity < 50m³) and an emission of 0,002 gN2O/km for motorcycles (cylindrical capacity > 50m³) is used. The fuel consumption data for motorized two-wheelers are not yet available from homologation certificates. For this reason, the COPERT methodology was used to estimate the fuel consumption of motorized two-wheelers. Different categories of two-wheelers are available within the methodology. As an illustration of the methodology for two-wheelers, the data from the COPERT methodology was used to assess the environmental performance. A differentiation is made between vehicles with 2-stroke and 4-stroke engines. A further differentiation is based on the cylindrical capacity of the engine. An overview of these vehicle types is given in table 4-18. An electric scooter from Peugeot (Scootelec) was also added to the list of mopeds.


p.73

Table 4-18: Vehicle selection details and COPERT categories for motorized two-wheelers

Type Fuel type Engine Consumption

Moped

Euro 0 – noise 1989 two-stroke - petrol <50cm³ 3,3 L/100km

97/24 stage I two-stroke - petrol <50cm³ 1,9 L/100km

97/24 stage II two-stroke - petrol <50cm³ 1,4 L/100km

Peugeot Scoot’Elec electricity 2,8 kW 8,0 kWhe/100km

Motorcycle

Euro0 noise 1989 two-stroke - petrol >50cm³ 4,1 L/100km

97/24/EC noise 1993 two-stroke - petrol >50cm³ 3,0 L/100km

Euro0 noise 1989 four-stroke - petrol 49 - 250cm³ 3,4 L/100km

Euro0 noise 1989 four-stroke - petrol 249 - 750cm³ 4,5 L/100km

Euro0 noise 1989 four-stroke - petrol >750cm³ 5,8 L/100km

97/24/EC noise 1993 four-stroke - petrol 49 - 250cm³ 3,9 L/100km

97/24/EC noise 1993 four-stroke - petrol 249 - 750cm³ 3,9 L/100km

97/24/EC noise 1993 four-stroke - petrol >750cm³ 3,9 L/100km The average mileage distribution for two-wheelers was obtained from the Belgian National Institute of Statistics (NIS) and is given in table 4-7 on page 53. For two-wheelers 40% urban traffic against 60% rural traffic is considered and will be taken into account for the damage calculation for the category air quality. To assess two-wheelers, a different reference vehicle has been selected, with emissions corresponding to the emission limits for 2006 from European directive 2002/51/EC, a fuel consumption of 3 litres per 100 kilometre and a noise emission of 76dB(A). This results in the following emission data for the reference vehicle for the methodology for two-wheelers:

Table 4-19: Tank-to-Wheel and Well-to-Tank emissions of the reference vehicle for two-wheelers

Sound CO2 N20 CH4 CO HC NOX PM10 SO2 FC [dB(A)] [g/km] [g/km] [g/km] [g/km] [g/km] [g/km] [g/km] [g/km] [L/100km]

TTW 76 66 0,002 0,2 2 0,8 0,15 0 0,002 3 WTT - 8,9 0 0,02 0,005 0,2 0,04 0,002 0,06 0,27 The results of the assessment methodology for two-wheelers, considering the vehicle data from table 4-18, are shown in figure 4-20.


p.74

Two-Wheelers (Copert III categories)

45,4

66,1

77,7

90,4

40,1

55,0

64,160,1

50,5

66,3 64,8 64,9

0

50

100

150

200

250

300

<50cm³ <50cm³ <50cm³ PeugeotScootelec

>50cm³ >50cm³ 49 -250cm³

249 -750cm³

>750cm³ 49 -250cm³

249 -750cm³

>750cm³

Euro 0 -Noise1989

97/24stage I

97/24stage II

Euro 0Noise1989

97/24/ECNoise1993

Euro 0Noise1989

Euro 0Noise1989

Euro 0Noise1989

97/24/ECNoise1993

97/24/ECNoise1993

97/24/ECNoise1993

To

tal

Imp

act

0

10

20

30

40

50

60

70

80

90

100

Noise

Ecosystems

Human Health

Global Warming

Ecoscore

MotorcycleMoped

two-stroketwo-stroke four-stroke

Figure 4-20: Total impact and Ecoscore for the set of motorized two-wheelers – split up per category

This analysis shows the positive evolution of the environmental impact of two-wheelers in time thanks to the introduction of ever more stringent emission limits. An important difference can be seen between the environmental impact of two-stroke and four-stroke two-wheelers, in particular for the category ‘impact on human health’. The engine displacement of the vehicle also affects the environmental performance of the motorized two-wheelers. The methodology used to assess the environmental impact of two-wheelers allows making a comparision between the impact of electric bicycles versus thermal mopeds and electric scooters. The energy consumption of a set of electric bicycles (see figure 8-10 on page 140) has been measured and was used for a comparative assessment. The low energy consumption of this kind of vehicles leads to a very low impact compared to thermal mopeds and even to electric scooters.

Table 4-20: Details of selected motorized two-wheelers, including electric bicycles

Type Fuel type Engine Consumption

Euro 0 – noise 1989 two-stroke - petrol <50cm³ 3,3 L/100km

97/24 stage I two-stroke - petrol <50cm³ 1,9 L/100km

97/24 stage II two-stroke - petrol <50cm³ 1,4 L/100km

Peugeot Scoot’ Elec electricity 2,8 kW 8,0 kWhe/100km

Vectrix VX-1 scooter electricity 21 kWp1 10,7 kWhe/100km

Electric Bicycle (minimum) hybrid electric <0,25kW 0,5 kWhe/100km

Electric Bicycle (maximum) hybrid electric <0,25kW 2,0 kWhe/100km

1 Peak power of the electric motor according to the manufacturer (no data available for continuous rated power) [129] website; "Vectrix Electrics," (accessed on: 17/12/2009) Available at: http://www.vectrix.com/.


p.75

45,4

66,1

77,7

90,4 88,1

99,5 98,4

0

50

100

150

200

250

<50cm³ <50cm³ <50cm³ PeugeotScootelec

Vectrix electricscooter

Electric bicycles(maximum)

Electric bicycles(minimum)

Euro 0 - Noise1989

97/24 stage I 97/24 stage II

To

tal I

mp

act

0

10

20

30

40

50

60

70

80

90

100

Eco

sco

re (

TW

)

Noise

Ecosystems

Human Health

Global Warming

Ecoscore

Figure 4-21: Total impact and Ecoscore for mopeds, electric mopeds, electric motorcycles and electric bicycles

4.10 Ecoscore calculation tool and Ecoscore website In the framework of the Ecoscore project in commission of the Flemish administry of environmental affairs (LNE1), a calculation tool has been developed. The purpose of this calculation tool was to disseminate the Ecoscore methodology in such a way that the user of the tool could calculate the Ecoscore of individual vehicles and evaluate the influence of most parameters. A short list of vehicles with different fuel or technologies use are available within the calculation tool, but the user can also introduce own vehicle data if desired. For this purpose, MS Excel® in combination with the Visual Basic framework within MS Excel® was used to develop an easy to understand user interface. The welcome screen of the interface for the calculation tool of the Ecoscore for passenger cars (M1) or light duty vehicles (N1) can be seen in figure 4-22.

Figure 4-22: Welcome screen of the Ecoscore calculation tool

1 Departement Leefmilieu, Natuur en Energie van de Vlaamse Overheid


p.76

The user of the Ecoscore calculation tool can choose between introducing own vehicle data or selecting a vehicle from a database. Once a vehicle has been selected in the database or the user has introduced all the required vehicle data, an overview of the vehicle data is shown. The user can confirm the data by pushing the yellow calculation button, or can choose to change the selected vehicle or data.(see figure 4-23).

Figure 4-23: Overview screen of the selected vehicle’s data

The last step in the calculation tool is the results screen where the Ecoscore of the selected vehicle is shown (see figure 4-24). Besides the Ecoscore, the total impact TI is presented (in percent) as well as the normalised damages qi for all four damage categories. The weighting factors αi used for the calculation of the total impact and of the Ecoscore are shown also on the results screen.

Figure 4-24: Results screen of the Ecoscore calculation tool


p.77

The user of the calculation tool can also change some of the parameters used to assess the environmental damage of the selected vehicle. The following parameters can be changed:

• The reference vehicle data (see also table 4-10 on page 54) • The weighting factors αi • The damage impact factors δi,j, rural and δi,j, urban

In order for the user to maintain a clear overview on the structure of the assessment methodology and on the parameters that are used, these parameters are arranged in three different tabs within the calculation tool (see figure 4-25).

Figure 4-25: Parameter setting screen of the Ecoscore calculation tool – tab damage calculation

A website (www.ecoscore.be) has been set up by VITO where the Ecoscore of all passenger vehicles can be consulted (see figure 4-26). The methodology used for the calculation of the Ecoscore corresponds with the assessment methodology described in this work. The complete vehicle record data from VITO (based on databases from DIV1 and Febiac2) has been used to develop this website. A logo for Ecoscore has been created and is shown in figure 4-27 on page 79.

1 Directie Inschrijving van Voertuigen (DIV); en: Vehicle Registration Service (Belgium) 2 Belgische automobiel- en tweewielerfederatie; en: Federation of automobile and two-wheeler manufacturers and importers.


p.78

Figure 4-26: Ecoscore website screenshot (www.ecoscore.be)

4.11 Conclusions of the environmental assessment methodology This chapter describes the methodology of the environmental rating tool called Ecoscore. A comprehensive and transparent overview and a summary of the environmental rating system for road vehicles, called Ecoscore, have been presented. This methodology allows the assessment of vehicles with different drive trains and using different fuels. The Ecoscore methodology is based on a well-to-wheel basis and combines different impact categories such as global warming and human health effects. The Ecoscores of real vehicles were calculated and discussed to demonstrate the applicability of this methodology. From the analysis made with the Ecoscore methodology, a positive evolution of the environmental performance of vehicles through time can be observed. This is mainly due to the ever more stringent European emission regulations. A low environmental impact (and therefore a high Ecoscore) was obtained for the battery electric vehicle (Peugeot 106 electric). Also the hybrid petrol-electric vehicle (Toyota Prius) and the CNG vehicle (Opel Astra) obtained a favorable Ecoscore. The LPG vehicles show the best environmental performance amongst all conventional vehicles, whereas Euro 4 petrol and diesel vehicles have a similar Ecoscore. On the other hand, the emissions of CO2 from more recent vehicles, which are related to the fuel consumption, are not always reduced. The positive influence of an improved engine technology is sometimes annihilated by an increase of the vehicles’ weight or an increased energy consumption caused by certain on-board options. It is noticeable that the newest generation of diesel vehicles has caught up its delay concerning their environmental performances on the petrol vehicles. Furthermore, the difference between the environmental performance of those vehicles and of the LPG vehicles has been reduced. Finally, on the basis of a sensitivity analysis, the robustness of the methodology has been evaluated. As a general conclusion, one can state that the environmental rating system is robust and thus applicable as a policy instrument (taxation, incentives, consciousness raising campaigns, etc.) to support the acquisition and use of environmentally friendly vehicles. The ambition of the Ecoscore environmental rating tool is to lead to a common system for policy measures in Belgium and possibly in other European countries, to promote the introduction and use of cleaner vehicles. In order to define implementation pathways for a consistent policy, not only the environmental impact but also barriers for purchase (technical and economical barriers, market-related barriers, legislative and regulatory and psychological barriers) and use of road vehicles should be taken into account. Fiscal measures should be based on the “polluter-pays” principle [130] and the cost-effectiveness of possible policy measures (road pricing, fiscal measures, modulated vehicle taxation, subsidies, regulatory policy…) on the overall environmental performance (quantified by the Ecoscore) of the national vehicle fleet can be analysed as a next step. The Ecoscore methodology will be used by the


p.79

three Regions in Belgium (Flanders Region, Brussels Capital Region and the Walloon Region) for future fiscal measures. Introducing the methodology on a European level might be the most effective approach. So, all vehicle manufacturers would consider this rating tool as an important factor in their marketing and development strategies. Further research should be made at different levels. An update of the indirect emission data is important to be able to assess the current fuel production chains. Research related to both the indirect and direct emissions of different types of biofuels is needed to be able to make an objective and sound comparison with fossil fuels. A recent study in this field shows large differences for the different possible production chains of different biofuels [78]. Also the production of second generation biofuels is promising from an environmental point of view [79] and should consequently be analysed in more detail.

Figure 4-27: Ecoscore logo


p.80

Jean-Marc

Rectangle

p.81

PART II: Light Electric vehicles - Case study: The development of an electric bicycle for postal

distribution

“aide toi, l’électricité t’aidera” (Denys Klein)[131]

Jean-Marc

Rectangle

p.82

Jean-Marc

Rectangle

p.83

5 Light electric vehicles for

postal delivery

5.1 What are Light Electric Vehicles? Light Electric Vehicle, sometimes referred to as LEV, is a term frequently used to designate a range of rather small vehicles equipped with an electric drive train. At first it deals with electric vehicles, but not necessarily only battery-electric vehicles. The range of vehicles that is mostly designated with this term is that of vehicles that are not homologated as passenger cars. It concerns mainly electrically driven two-wheelers such as bicycles and scooters, but also derivates such as tricycles and quadricycles. In case of electric scooters or electric mopeds, they can be homologated as a ‘moped’ or ‘motorcycle’, depending on the technical characteristics, the performances and on the legislation in the country where they are intended to be used. However, a formal definition of light electric vehicles does not (yet) exists. A proposal of such a definition will be made later on in this paragraph. When searching for a possible definition for LEV, the term ‘light’ could be linked to the vehicle’s mass, however this parameter alone would not be a good choice to delimit certain vehicle categories as used for road legislation. A possible limit could be 350 kg for the unladen mass of the vehicle (without driver or payload) and exclusive the mass of the batteries as used for the category ‘light quadricycles (L6e)’ in Directive 2002/24/EC related to type approval of two or three-wheel motor vehicles. A more appropriate approach would be to link the term ‘light’ to the performance of the vehicle. A possible limit in this regard would be to consider LEVs as electric vehicles having a limited maximum speed. A possible limit value for the maximal vehicle speed could be 45 km/h. This speed limit of 45 km/h seems to be the greatest denominator of most countries and would allow being compatible with the European driving license category AM for scooters and mopeds1. This could be adapted to 30 mph (48,3 km/h) to be compatible with the road legislation of most States of the United States of America. Additionally it would be useful to limit at the same time the power of the electric motor, in order to consider the vehicle as a ‘light electric vehicle’. A value of 4kW could be used as the maximal power of the electric motor. In this context the maximum continuous rated power of the electric motor could be used to validate the previous condition. The value of 4 kW again would allow having common conditions with existing vehicle categories used for EU vehicle type approval (see also Directive 2002/24/EC). This definition would also be compatible with the specific category of four-wheeled electric vehicles, called ‘Neighborhood Electric Vehicle’ in the United States. The latter is related to the US department of Transportation classification of ‘low speed vehicles’ which have a maximum speed of 25 mph (40,2

1 New driving licence category, valid in all EU countries, for two- and three-wheel vehicles with a maximum design speed of not more than 45km/h (Directive 2006/126/EC)

Light electric vehicles for postal delivery

p.84

km/h), but also a minimum speed of 20 mph (32,2 km/h). Several initiatives are made to increase the maximum speed of 25 mph. Others propose a new category of ‘medium speed vehicles’ with a maximum speed of 35 mph (56,3 km/h). Furthermore, in some countries, limits for vehicle mass are applicable and could possibly interfere with definitions not considering mass as a limiting factor. It is clear that all these differences in definitions and especially in road legislation are hampering the market introduction of these new ‘light electric vehicle’ concepts. As a conclusion it can be stated that there is a clear need and advantage for further standardization and harmonization in the road legislation. As a conclusion, a possible definition for light electric vehicle could be: A light electric vehicle is a vehicle, having two, three or four wheels having a maximum design speed of not more than 45 km/h and is equipped with an electric motor with a continuous rated power not exceeding 4 kW. The unladen mass of the vehicle shall not exceed 350 kg, not including the mass of the batteries.

5.2 Market of LEVs The market of light electric vehicles worldwide is experiencing a strong and continuous growth, reaching a worldwide fleet of almost 23 million vehicles in 2008. The introduction of the “Yamaha PAS” system in Japan was the real start of this success. Since 2002, the market of electric bicycles in China started booming and now dominates the worldwide sales figures. In Europe the sales numbers are also increasing with a significant increase in sales numbers in 2008 compared to 2007 (from 250 000 to 550 000), mainly due to success of electric bicycles in The Netherlands[132]. In the Flemish Region in 2005 the market was about 10 000 pieces sold by a minority of the bicycle dealers[133]. Today almost every dealer has at least one model of electric bicycle in his store. The image of electric bicycles has also evolved from ‘bicycle for old or disabled people’ to a more fashionable good looking bicycle also attracting more and more younger people.

5.3 Different appearances of LEVs As described earlier, a large range of vehicles with quite different appearances and applications can be denoted as light electric vehicles. Some vehicles have a cockpit while others are not covered. Also the position of the driver can differ and can be used to make some differentiation. The most common positions are:

• riding position • recumbent and semi-recumbent position • sitting upright • standing upright

In the riding position the body weight of the driver is spread over the handlebars, the pedals and the saddle. The riding position is the best known position as applicable to bicycles and motorcycles and also for some tricycles. This position is linked with the presence of pedals, however sometimes not used for propulsion as is the case with motorcycles for instance. An example of an LEV with riding position is the “Tidal Force M750 of Wavecrest” (U.S.A.), powered by a 750W brushless hub motor from “Matra”. Exceptional is the battery located in the front wheel of this electric bicycle and the ability to regenerative braking and cruise control. A picture of this hybrid electric bicycle is shown in figure 5-1.


p.85

Figure 5-1: Wavecrest Tidal Force M750 at EVS20, California 2003

Recumbent or laid-back reclining position is claimed to be more ergonomic compared to the riding position, especially for pedal powered vehicles. This position also allows reducing the frontal profile and in this way allows reducing the aerodynamic drag (see paragraph 7.2.3 on page 115). Also the possible lower center of mass resulting from this specific driver position is claimed to be safer compared to the classic riding position. A wide variety of configurations exist: long to short wheel base; large, small or a mix of wheel sizes; overseat, underseat or no-hands steering; and rear wheel or front wheel pedal drive. Also variants with three wheels or recumbent tricycles exist[134]. Many of these recumbents exist with an electric drive train, often using an ‘electric motor kit’ with hub motor such as the “Heinzmann” motor kit and “Bionix” motor kit[133]. An example of a recumbent tricycle with a human-electric hybrid drive is the “Twike” equipped with a 3kW electric motor [135]. However, this vehicle has a maximal speed of 85km/h and therefore doesn’t fit completely the earlier definition of light electric vehicle. Another example is the “CitiCruiser” from “Veloform” that is used as a velotaxi but meanly as an eye-catcher and as an outdoor advertising platform in cities or at large events. This semi-recumbent three-wheeler is equipped with a “Heinzmann” 250W hub motor in the front wheel. The speed of the electric assistance motor is limited to 25km/h. A ricksha from “Veloform” in Brussels is shown in figure 5-3. Instead of transporting people, goods can also be transported for delivery purposes with the “Veloform DeliveryCruiser”.

Figure 5-2: Twike at the EVS24 in Norway, 2009 Figure 5-3: Veloform CityCruiser I in Brussels, © Veloform

Other electric vehicles allow the driver to sit upright. Mostly it concerns four-wheeled vehicles such as electric golf cars or derivates. An example of such a vehicle is the “Free Duck” from “Ducati Energia” and can be seen in figure 5-27 on page 99. Another example is the urban electric car “Venturi Eclectic” which is equipped with a 4kW electric motor and has a maximum speed of 45km/h [136]. A picture of this light electric four-wheeler with sitting position for two persons can be seen in the figure 5-4 below.


p.86

Figure 5-4: Venturi Eclectic at the 2006 Paris Motor Show © Florian

Some electric vehicles allow moving while the driver is standing up. In most cases the driver can find stability by holding a handlebar or leaning against a part of the vehicle. An example of such a vehicle is the “Segway” self-balancing human-transporter and can be seen in figure 5-23 on page 97. Electric steps or sometimes called electric scooters are also an example of a light electric two-wheeler with standing up position of the driver. An example of such an electric step is the “eZip electric scooter” where the driver is standing upright on a platform. These steps are however often equipped with a saddle, allowing the driver also to sit down if he or she wants to. The range of light electric vehicles as described in this paragraph could also be divided into sub-categories, based on the number of wheels:

• Light electric two-wheelers • Light electric three-wheelers • Light electric four-wheelers • Others

Finally, the presence of pedals is an important element of distinction in the wide range of light electric vehicles. The pedal drive on light electric vehicles creates a combination of human power and electric power resulting in what could be considered a human-electric hybrid vehicle. To describe the wide range of light electric vehicles, the best approach is to use a combination of the above mentioned parameters. Even then, some particular vehicle will not be described or identified correctly, which illustrates the great diversity existing amongst light electric vehicles.

5.4 Electric two-wheelers and electric bicycles Light electric two-wheelers can now be considered as a subset of the group of light electric vehicles. However, different kinds of two-wheelers, equipped with an electric motor, exist. Most commonly the following light electric two-wheelers are encountered: electric scooters, electric mopeds and electric bicycles. Electric motorcycles, such as the “Vectrix VX-1” electric scooter (with brushless DC motor – 21kW peak) [129] is not a light electric two-wheeler because of its high performance in terms of speed (up to 100 km/h). Another example of an electric scooter is the “Oxygen” scooter with a limited speed of 45 km/h and therefore can be seen as a light electric vehicle, considering the definition proposed earlier. More details are given in paragraph 5.8 and a picture is shown in figure 5-26 on page 98. Another example is the “Yamaha Passol L” electric commuter that was presented at the Tokyo Motor Show in 2003. This scooter has a 580 Watt brushless d.c. motor and has a maximum speed of 42 km/h [137] (see figure 5-6). An example of an electric two-wheeler that could be considered as an electric moped is the “e-solex” of a French company that is a modern version of the well known French “cyclomoteur VéloSolex”, which had a petrol engine on top of the front wheel driving it through a ceramic friction roller-wheel directly on the tyre. Instead, the “e-solex” has a 400 Watt brushless hub motor in the rear wheel. The box on top of the front wheel is just a container for luggage. Two versions are available, one with a maximum speed


p.87

of 25 km/h and one with a maximum speed of 35 km/h. The “e-solex” can, if required, be driven by the pedals only [138].

Figure 5-5: Picture of the e-solex ©Frederic Dinh Figure 5-6: Yamaha Passol electric scooter

The focus of this work is however dealing with electric bicycles and will be considered in more detail hereafter.

5.4.1 Diversities amongst electric bicycles Some important properties of electric bicycles are that they are based on a bicycle frame and that they have pedals allowing the driver or cyclist to move the vehicle. Moreover, a bicycle is a pedal-powered two-wheeler. Sometimes mopeds, having pedals, resemble to electric bicycles and the difference is therefore potentially unclear. Important differences exist between different models of electric bicycles. The differences are mainly in how the electrical power system is integrated into the bicycle. An interesting classification of electric two-wheelers has been made based on the location of the electric motor [133]. A schematically representation of this classification can be seen in figure 5-7 and is based on reference [133]. Four different locations have been identified: near the saddle tube (a), in the hub of the front wheel (b), near the bottom bracket (c) and in the hub of the rear wheel (d).

(a) (b)

(c) (d)

(a) (b)

(c) (d)

Figure 5-7: Applied mounting places for the electric motor

Another approach to assess differences amongst electric bicycles is how the electric power system is integrated into the bicycles’ propulsion system. Because an electric bicycle also has a drive unit with pedals for human power input, it can be seen as a hybrid vehicle. Some manufacturers then also use the name of hybrid bike (e.g. the “Mercedes Hybrid Bike” – using the “Sanyo” hub wheel motor like in the “Sachs” electric bicycle). When looking at the power train topology of most common systems, electric bicycles can be identified as parallel hybrids of two drive units. The first unit is the classical pedal driven


p.88

system composed of pedals connected with the rear wheel trough a (chain) transmission and frequently with a gear system. The second drive unit is composed of the electric motor and the mechanical connection to one of both wheels. The power of the electric motor drive unit and the power of the pedals and cranks are at some point combined in the bicycle’s power train. Different solutions for adding or combining both mechanical powers are used. This is also closely linked with the motor’s choice or design. The mechanical power addition can be realized at different levels of the hybrid propulsion system:

• At the level of the road (type A) • At the level of the driving (rear) wheel (type B) • At the level of the transmission or chain drive (type C) • At the level of the pedal axis (type D)

Each of the different solutions is illustrated with a picture from commercial available electric bicycles shown at the IFMA 2006 Exposition in Köln, Germany (see figure 5-8 till figure 5-11). These different levels of power addition can be linked with the four power train topologies (type A – D) as proposed by the author in paragraph 8.4.1 on page 147.

Figure 5-8: Power addition at the pedal axis Figure 5-9: Power addition at the transmission

Figure 5-10: power addition at the level of the rear wheel Figure 5-11: Power addition at the level of the road

In some models the mechanical power is combined at the level of the pedals axis, before the power is transmitted to the driving wheel. In this case one single transmission (mostly a chain) and possibly a gear system is transmitting the total propulsive power to the rear wheel of the bicycle. The rear wheel is the single driving wheel. This system has the clear advantage that the gear system (rear derailleur or hub gear) of the pedal drive unit is not only available to extend the speed range for effective and comfortable pedaling but also for an optimal use of the electric motor power. At stand still or when climbing on a hill the driver can choose an appropriate gear ratio to have a maximal traction force available from the electric motor and from pedaling. An example of an electric bicycle using this configuration is the “Flyer F-series” from Biketec [139] or the “Grubber assist system” that can be built in the seat tube of any bicycle.


p.89

Some other models combine the mechanical power of the motor with the mechanical power of the pedals at the level of the transmission or the chain drive. This system has also the advantage that the gear system is available for both the pedal drive as for the motor and allows an optimal use of the electric motor power with the gear system if available. An Example of an electric bicycle using such a system is the “Panasonic” drive system used on different bicycle models such as the “Helkama Velox”. Sometimes the mechanical power addition is realized at the level of the driving wheel. In this case both drive units are connected with the drive wheel through their individual transmission (mostly chain or belt). An example of an electric bicycle with two separate transmissions connected to the rear driving wheel is the “Dolphin Express”[140], which uses a toothed belt to transmit the power of the electric motor to the rear wheel. Hub motors are often used in electric bicycles. In some models the hub motor is located in the rear wheel. In this topology the mechanical power of the electric motor is directly available at the rear wheel and is combined at the level of the rear driving wheel with the pedal power transmitted through the chain drive. An example of this kind of electric bicycle is the “Sparta Ion”[141]. In case a hub motor is installed in the front wheel, the mechanical power is added at the level of the road on which the electric bicycle is driven. This configuration is characterized with to driving wheels: pedal drive of the rear wheel and electric motor drive in the front wheel. An important advantage of such a system is the ease of mounting and maintenance. An example of electric bicycles with front hub wheel motors can be found at “Estelle” using “Heinzmann” electric drives [142]. The main advantages and distadvantages for the different power train topologies are summarized in the table below:

Table 5-1: advantages and disadvantages of the different power train topologies of electric bicycles

ADVANTAGES DISADVANTAGES type A ease of mounting and maintenance

traditional rear wheel chain drive no adaptations of the frame required

possible slipping of the front wheel

type B traditional rear wheel chain drive no adaptations of the frame required except rear fork

large rear fork width required

type C optimal use of motor power with the gear system favourable location of the centre of mass

increased wear of the chain drive increased mechanical complexity frame needs to be adapted

type D optimal use of motor power with the gear system favourable location of the centre of mass

increased wear of the chain drive increased mechanical complexity frame needs to be adapted

5.4.2 Relation between motor power and human power The modulation of the electric power from electric bicycles is also an important element that creates differences from a technical point of view. These differences can also result in differences in performances of the many available electric bicycles [133]. In some models of electric bicycles, the motor power is modulated in function of the cyclist’s effort. Cycles, equipped with pedals and an auxiliary electric motor, which cannot be propelled exclusively by means of this auxiliary electric motor is used as a definition of electric power assisted cycles (short: EPACs). However, this definition does not quantify the ratio between the motor power and the human power. For the development of the Japanese Yamaha PAS (Power Assist System) for the ‘Electro Hybrid Bicycle’ it was the purpose to make the human power account for 50% of the total energy supply in order to save on electrical energy requirement. Further it was considered preferable that human (pedal-) power would both activate and control the electric power. Also the speed of the vehicle is (indirectly)


p.90

measured and used to reduce the power assist at a certain speed and to be discontinued from a speed of 24km/h. The latter was introduced to comply with the requirements of the Japanese Transport Ministry [143]. These requirements resulted in a proportional pedal power control system. A sensing device continuously measuring the cyclist pedal force is needed to realize such a control system. This system can be described as a ‘pedal power control system’. Aside the Yamaha pedal assist system, incorporating such a sensor, other torque sensing devices or devices sensing the cyclist’s pedal force were developed to be used in electric bicycles. Other electric bicycles use a ‘power-on-demand’ system. For this kind of system a twist grip or lever is used to have an input from the driver to modulate the electric motor power. In most cases1, this is combined with a pedal speed sensor, to detect movement of the pedals. The latter is done to be able to comply with the European Directive 2002/24 (see paragraph 5.4.3). In this way the turning of the pedals can be made imperative to activate the electric motor. However, the effort of the driver can be very limited. ‘Just turning the pedals’ without really pushing can be sufficient to detect the driver’s pedals movement and to activate the electrical assistance. In this way a very large share of the total propulsive force can be delivered by the electric motor. For a power-on-demand system, the relation between motor and human power is not fixed, but is controlled by the cyclist. In addition, turning the pedals can be made obsolete for low speeds (less than 6km/h). This functionality of assistance at low speed is clarified in paragraph 5.4.3 below. In this way starting from stand still can be possible using only the electric motor. The latter is found very useful for postal applications characterized by frequent stop and go operation with important payloads. An example of such a system is the “Heinzmann” drive unit, used in various models of electric bicycles.

5.4.3 European Directive 2002/24 In order to be able to consider an electric bicycle as a “normal” bike from a legal point of view, it needs to fulfil certain conditions. However, if one of the conditions is not fulfilled (e.g. the maximal vehicle speed with electrical assistance is higher than 25km/h), insurance compensations might be denied to the victim in case of an accident. In the Belgian road legislation[144] for instance, several definitions have been incorporated, among which the following: (Article 2.15.1.) "Non motorised vehicle": each vehicle with two or more wheels, which is propelled by means of pedals or handgrips, by one or more of the users and which is not equipped with an engine, such as a bicycle, a tricycle or quadricycle. The presence of an electrical assistance motor with a nominal rated power of up to 0,25 kW, of which the traction force is gradually reduced and finally interrupted when the vehicle reaches a speed of 25 km/h, or sooner, if the cyclist stops pedalling, brings no modification in the classification as a ‘non motorized vehicle’. The above definition contains the description of a bicycle with electrical assistance motor in conformity with the definition according to the European directive 2002/24/EC. Most of the electric bicycles offered on the European market comply with this restriction, however not all of them do. With some electric bicycles the electrical assistance motor continues to provide traction force at speeds higher than 25km/h.However, in these cases the motor power generally remains below 0,25kW. The typical maximum speeds obtained with this type of electric bicycles are around 38 to 40 km/h. An example is the “Flyer F-series” of the manufacturer “Biketec” [139] that was available on the Swiss market. In Switzerland, an appropriate category of vehicle has been defined (‘light motorcycle’), which allows the user to contract a limited insurance coverage and to wear a bicycle helmet. In Belgium however, this category doesn’t exist (yet) and so the latter kind of electric bicycle would be classified as a motorcycle of category B. Consequently, full (moped) insurance is required and also wearing a motorcycle helmet (instead of a bicycle helmet) becomes mandatory. Another example of non-compliance with the Belgian road legislation is the group of electric bicycles, which provide electric traction without requiring the driver to turn the pedals. In these circumstances, the electric bicycles are functioning as electrically powered mopeds. However, the maximal speed of the vehicle, upon functioning as an ‘electric moped’, varies. Some electric bicycles can only be operated fully electrically to a very limited speed (typically 6km/h), whereupon turning the pedals becomes necessary

1 Except most electric bicycle models from the United States of America


p.91

to activate the electrical assistance at higher speeds. In Germany the category of 'electrically assisted bicycles', which is thus considered as an 'ordinary bicycle', is extended to this type of electric bicycles where pedalling at low vehicle speed is not necessary. This consideration is valid for vehicle speeds up to 6km/h. As from a speed higher than 6km/h, turning the pedals is again necessary to keep the electrical assistance active. This functionality is also denoted as ‘start-up assistance mode’

5.4.4 European standard EN15194 In January 2009 a new standard was published by European Committee for Standardisation, CEN concerning electrically power assisted cycles. Its aim is to provide a standard for the assessment of electrically powered cycles of a type which are excluded from type approval by the European Directive 2002/24/EC (see paragraph 5.4.3). This European Standard specifies safety requirements and test methods for the assessment of the design and assembly of electrically power assisted bicycles and of sub-assemblies used for EPACs. Further, EN 15194 specifies requirements and test methods for the engine power management systems of EPACs and for the electrical circuits including the charging system. The following requirements for the power management are included:

• assistance shall be provided only when the cyclist pedals forward • assistance shall be cut off when the cyclist stops pedalling forward, respecting certain cut off

distance • the output or assistance shall be progressively reduced and finally cut off as the vehicle

reaches the maximum assistance speed • the assistance shall be progressively and smoothly managed

Further, the standard contains requirements concerning start-up assistance mode, electromagnetic compatibility and descriptions of test methods for checking several technical requirements. Finally requirements and guidelines for marking and labelling are stated in this standard as well as the obligation of providing instructions for use for the EPAC, containing specific elements related to the electric system.

5.5 Bicycles for postal delivery: a brief history The bicycle has known a very surprising evolution and history. The first step in the development of the bicycle was most likely the development made by the German Baron von Drais in 1817 of the foot-propelled two-wheeled “laufmaschine” or “running machine”. This early bicycle had front wheel steering and was entirely made of wood [145]. The next step in the evolution of the bicycle was the invention of the pedaled velocipede [134] around 1866. Successive to and thanks to the development of spokes around 1870, the front wheel of the pedaled velocipede was made larger and larger. This allowed achieving greater speeds and this was the beginning of the decade of the “high wheeler” or “ordinary” bicycle. James Sterley played a prominent role in this development. When bicycles became available, they very soon were also used by postmen and postwoman1 in Europe and the US to deliver mail. They allowed increasing the speed of the mail deliveries and the distances postmen could travel. In the figure 5-12 an illustration of a postmen riding on a high wheeler in 1887 delivering mail in Brussels is shown[146]. This type of vehicle was used for an extensive period in several European countries. In the United Kingdom, this vehicle was called a “penny- farthing”2 and used for postal delivery by the Royal Mail [147]. See a picture of a Victorian postman that delivers mail from a penny farthing bicycle in 1900 in figure 5-13 below.

1 Hereafter the term ‘ postmen’ will be used to denote both men and woman delivering mail 2 The name penny-farthing comes from the British penny and farthing coins, one much larger than the other and refers to the large leading front wheel (penny) and the much smaller rear wheel (farthing) of this early bicycle.


p.92

Figure 5-12: “Le porteur de dépêches à Bruxelles”, 1887[148] Figure 5-13: Mail delivery from a penny-farthing, 1900[149]

It appeared not to be easy to drive a high-wheeler that was rather preserved for adventurous athletic men. Therefore continuous developments were made to further improve the comfort, balance and safety for the driver. In this context, aside the bicycle many other practical machines with three or four wheels were developed, often using (protected) chains for transmitting pedal power to the driving wheel(s). An interesting, however not wide spread example of such a machine, is the centre-cycle or pentacycle, designed and patented by Edward Burstow, an architect from Horsham, Sussex in 1882. Postal officials at Horsham tried out these cycles for both postal and telegraph delivery work [150] (see the picture of Figure 5-14). The continuous developments and improvements of the tricycle (and its derivates) lead to the arrangement a front-steering ‘modern tricycle’ that was very similar to the emerging form of the modern ‘safety bicycle’1 (anno 1884-1885), as we still know it today. This bicycle had direct steering and had a close-to-diamond-shape frame. In 1888 Dunlop patented the pneumatic tires that offered greater comfort and improved safety.

Figure 5-14: “Hen and Chickens” Pentacycle or Centre-cycle, circa 1882 [150]

Figure 5-15: Tricycle with forward basket carrier, 1934

Many technological developments first used for tricycles, such as chain drives and variable ratio transmission, where adapted for use on bicycles. Carrier tricycles, especially designed for transporting goods where also widely used for postal delivery for instance in the UK (see figure 5-15). Also in the Netherlands this kind of tricycle, the “bakfiets” was introduced in the 1930s and was popular at the contemporary PTT (now TNT Post BV) for distributing parcels or other large items. The regular (safety) bicycle remains however a more commonly used mean of transportation for postal delivery because of its great advantages in maneuverability and small track compared to tricycles or other vehicles.

1 Safety bicycle is a term to denote the type of bicycle as we still know it today, being safer than its predecessor, the high wheeler, because the rider's feet were within reach of the ground, making it also easier to stop.


p.93

Figure 5-16: Group of cycling postmen ready to start their round in the early 1920s [151]

Even today the bicycle is still used in many countries by different postal organizations for the daily mail delivery. These bicycles have continuously evolved adopting new bicycle technologies. For postal application, the frame has been adapted for the high payloads using oversized tubing and front and rear carriers have been added. Also the rims and spokes are enforced to withstand heavy loads and impacts from steps. To improve the stability at standstill or to easily park the bicycle, a kick stand connected below the front carrier is sometimes added. An example of such a “cargo-bike” used by the German Post can be seen in figure 5-17. Further, various designs of the payload carriers have been made, to adapt to different types of postal bags or boxes for transporting mail items. An example of postal bicycle with payload carrier for mail boxes can be found in figure 5-18 [152].

Figure 5-17: Postal delivery bicycle, Cologne 2008 Figure 5-18: SpeedBike AL26 postal delivery bicycle

5.6 Electric bicycles: a brief history The history of electric bicycles is closely related to that of the conventional bicycle, that was briefly described in the paragraph 5.5 above, and to that of the motorcycle. Very soon, steam engines and mainly fuel combustion engines were mounted on early bicycle platforms. A first interesting example is the steam bicycle from Michaux and Perraux from 1868-1869. This was a two-wheel bicycle that combined front wheel pedal-power with power coming from the steam engine. The power from the steam engine was transmitted by a pulley and belt system to the rear wheel. This can be seen as the first hybrid drive for bicycles, which combines biomechanical power with steam power [133]. Many other designs were developed afterwards, often using combustion engines, resulting in the development of motorcycles. In early versions of motorcycles, the pedals where used to start up the engine and to have a back-up in cause of engine failure, rather than intentionally to have a hybrid drive. The first use of electric motors to propel a bicycle appeared only in the late 1890s. Several patents were taken for a series of inventions of “electric bicycles”. In 1895 O. Bolton had invented an early electric bicycle with a 6-pole brushed d.c. hub motor, mounted in the rear wheel. However this vehicle did not have pedals, so was rather an early electric motorcycle instead of an electric bicycle. The first electric bicycle with a motor in the hub of the crankshaft axle (with pedals) was invented shortly afterwards in 1897 by H.W. Libbey. In 1899 J. Schnepf invented an electric bicycle with a pulley on the motors’ shaft


p.94

that is pressed against the top of the rear wheel (friction roller-wheel). This electric bicycle could also charge its battery when coasting downhill (first regenerative braking applied on electric bicycles). Because of improvements on combustion engine technology, cheap oil and the electric starter, battery electric propulsion for bicycles and for vehicles in general was somewhat repressed until the early 1970s. The oil crisis caused acceleration in the development of alternatives for petrol driven transportation and caused a new boost in electric vehicle developments. Many developments and prototypes were built all over the world (mainly in Europe, the United States and Japan) but it took until the late 1980s and early 1990s before real commercially-viable products became available. An early electric bicycle that used a torque sensing device, for modulating the motor power to the drivers’ effort, was developed from 1993 in Japan by the Yamaha Motor company (PAS system) [143]. The introduction of the Yamaha PAS is sometimes considered as the beginning of the age of the modern electric bicycle [132]. This was meanly due to the specific legislation in Japan, requiring a measurement of the pedal force. Other developments of electric bicycles however often used switches or twist grips for controlling the electric motor power [153-156]. In the late 1990s these developments lead to the currently commercialized electric bicycles and were successfully imported into Europe knowing an ever increasing sales number until today [132]. Some electric bicycles and electric scooters are being marketed today in Europe and have been evaluated by European or national initiatives such as the Swiss initiative NewRide [157]. An important project in this regard was also the E-TOUR project (Electric Two-wheelers on Urban Roads) and was funded under the Energy Program of the European Commission. The overall objective of E-TOUR was to demonstrate, evaluate and promote the use of electric two-wheelers in Europe [158]. The VUB was actively involved and contributed by polling the users of a lending service of electric bicycles and by the development of an objective performance test for “electrical power assisted bicycles” [133, 159]. The E-TOUR project showed that the performances of (normal consumer) electric bicycles offered on the European market (anno 2000) were often disappointing. This was mainly due to a lack of reliability of the offered products. The users also experienced a lack of range. In addition, the large difference between the range announced by the manufacturers and the actual range obtained in real use caused a feeling of disappointment among the users. Furthermore, the vehicle was experienced as too heavy and too expensive. However these drawbacks of the conventional commercially available electric bicycles, the feature of electrical assistance offered on the vehicle showed great opportunities and potential for professional use such as postal delivery. As the postmen, delivering mails (and parcels) by bike, need to carry along a large payload, additional traction force coming from the electrical power system, helps to perform their daily duty. Further, in the search of lowering the total cost of ownership of their delivery vehicle fleet, many postal operators are looking for alternative solutions. In addition postal companies are looking for more sustainable or greener ways of performing the delivery. Hybrid and electric vehicles are in this context a possible solution to lower the environmental impact of the mail delivery business. Also, the postal requirements (see chapter 6, page 101) correspond to very high levels for performance, reliability and service.

5.7 Electric bicycles for postal delivery The use of electric bicycles for postal application was evaluated by different postal organizations in Europe shortly after their introduction in the late 1990s. One of the first significant projects in Europe in this context was made by the German Post. In the figure 5-19 a picture of an electric bicycle for postal delivery with a Yamaha PAS electric power system from “Deutsche Post” is shown.


p.95

Figure 5-19: Electric post bicycle in Berlin, 2007 �cLhoon

Meanwhile almost every postal organization has performed some experiments using electric bicycles for their delivery mission. However, their use for mail delivery was not always successful. Often, low product quality and limited range were cause of failure. Some (early) cheap electric bicycles were not designed and intended to be used for postal delivery and showed low ergonomic quality and performance. Because of the strong interest of postal operators and the great potential market some postal organizations and bicycle manufacturers joined forces and were able to develop qualitative electric bicycles for postal delivery. The French Post for instance has several internal projects for developing and evaluating different kinds of light electric vehicles. This also included the evaluation of electric bicycles for postal delivery. Today about 600 electric bicycles are in service at La Poste, using the Panasonic pedal assist system[160]. A picture of this electric bicycle is shown in figure 5-20.

Figure 5-20: Electric bicycle for postal delivery for La Poste [160]

An important project in this regard is the European NEPH project (see paragraph 6.2 on page 102) that resulted in a common set of specifications for electric bicycles for postal delivery for European postal organizations. In the framework of this European industrial project some prototypes were built. An example of a prototype of the Belgian bicycle manufacturer “Ludo N.V.” with brand name “Granville” is shown in figure 5-21. This prototype has a “Heinzmann” drive unit and uses two flat shaped NiMH battery (two times 36V – 14,5Ah) packs mounted aside the rear wheel. These prototype batteries where developed by SAFT batteries in the framework of the NEPH project.


p.96

Figure 5-21: NEPH prototype Granville electric bicycle for postal delivery, 2007 ©Philippe Lataire

More recently in Japan, the electric bicycle market has known a significant increase after the new road traffic law coming into application in December 2008. The new law allows increasing the share of the motor power compared to the human power from 1:1 to 2:1 for speeds up to 10 km/h. Most popular applications in Japan since this new law are for rental services and parcel delivery. The Japan Post Group introduced 80 electric bicycles with trailer in Tokyo and Osaka and another 800 electric bicycles throughout the country for postal delivery. In Europe, in 2008, the Belgian Post launched a call for tenders for electric bicycles for postal delivery. Currently three different potential suppliers are withheld. One electric bicycle from “Wattworld” (figure 5-22) is currently under extensive test for further evaluation, in particular to evaluate the performance of the batteries in wintertime.

Figure 5-22: Wattworld electric bicycle for the Belgian Post

5.8 Other electric vehicles for postal delivery A first interesting example is the “Segway Personal Transporter” that is also being customized for postal applications. The “Segway i2 Cargo” (see figure 5-23) consists of a platform with two wheels with a self-balancing drive unit. The Cargo version is equipped with two customizable hard cases mounted above each wheel. To move, the driver has to lean slightly forward or backwards and by moving the leaning steer left or right he can influence the direction of motion. The manufacturer claims a maximum range of 38km for the “Segway i2” with lithium-ion batteries.


p.97

Figure 5-23: The Segway® i2 Cargo Personal Transporter at PostExpo 2006

Electric caddies or trolleys are also used for postal delivery, but are not classified as vehicles. The example of the electric caddy manufactured by the German company “Expresso” is discussed later in paragraph 6.4 and can be seen in figure 6-6. Another example is the distribution trolley of the Finnish company “Asoma”. This trolley is widely used by the Fin Post “Itella”. One of the main reasons to introduce an electric caddy is to literally and proverbially reduce the weight on the shoulder of the postmen. But also the use of an electric caddy is a means for increasing the payload and to solve the problem of higher payload requirements. Finally, it can avoid the use of relay boxes or points where additional payload is picked up.

Figure 5-24: Distribution trolley with auxiliary electric motor from Asoma

The combination of a bicycle with a trailer is a possible solution to increase the payload capacity for postal delivery. The ‘E-trailer’ of the former German company “Sun+Cycle” pushes the leading bicycle. The trailer has a loading capacity of 220 liters. A commercial leaflet of the E-trailer announces a range of up to 50km. “Sun+Cycle” was bought in 2007 by Hercules and did not continue the activities with the E-trailer. In other cases an electric bicycle is sometimes combined with a (non electric) trailer, again to increase the payload capacity (volume and/or weight). A picture of the “Sun+Cycle E-trailer” can be seen in figure 5-25.


p.98

Figure 5-25: E-trailer in combination with a (mechanical) postal bike at Deutshe Post

Also tricycles are offered for postal delivery services. An interesting example is the Tri-CarGo from Accel-Pro that is powered by a direct methanol fuel cell. More details and a picture (figure 8-3) of this heavy duty electric tricycle is given in paragraph 8.2.1.5 on page 132. In 2006 the Belgian post has bought 700 new four-stroke scooters, to replace their current fleet of environmental unfriendly two-stroke scooters. At the same time they decided to evaluate the use of electric scooters for postal delivery. The Belgian post launches as the first European postal operator to test a small fleet of 50 electric scooters from “Oxygen”[161]. This electric scooter has a 3,5 kW – 65V brushless d.c. wheel motor that can develop 250Nm. The manufacturer claims a payload capacity of 20kg in the front and 75kg at the back. The scooter can be equipped with two up to four battery packs of 12V and 100Ah. A maximum speed of 45km/h and a driving range of 120km is stated in commercial leaflets [162]. Meanwhile, the Belgian post has decided not to continue the use of the “Oxygen” electric scooter due to frequent technical problems. A picture of this scooter shown at “PostExpo 2006” in Amsterdam can be found in figure 5-26.

Figure 5-26: Oxygen electric scooter for Belgian Post at PostExpo 2006

The electric quadricycle from figure 5-27 is produced by the Italian “Ducati Energia” and was introduced from 2007 at several European operators as the “Free Duck”[163]. This vehicle is equipped with two electric motors of 2kW each, integrated into both rear wheels. Two versions are available: an electric version using 8 lead-acid battery packs (12V-42Ah). The announced performances are a range of 50km, a maximum speed of 45km/h and a payload capacity of 200kg. This quadricycle is also available with a range extender that consists of a 4 stroke 100cc petrol engine that recharges the batteries when they have reached a low state-of-charge. It is currently being evaluated by “Poste Italiane” and by the Belgian Post. However, in Belgium, “De Post/La Poste” encounters difficulties with the ergonomics and safety issues of this vehicle [164].


p.99

Figure 5-27: “Free Duck” electric quadricycle evaluated by Poste Italiane [163]

However out of scope of this chapter, it is interesting to mention that beside the light electric vehicles, also other (battery) electric vehicles are being considered for postal delivery: cars, van and even trucks. Without going into much detail on these vehicles, some examples are shown in the figure 5-28 and figure 5-29.

Figure 5-28: Modec electric van at PostExpo 2006 Figure 5-29: Smith electric vehicle truck at PostExpo 2006

The “Modec” battery electric van has a maximum speed of 50 mph (c.a. 80 km/h) and a payload capacity of 2 tons. The maximum range is 100 miles (c.a. 160 km) or 60 miles (c.a. 97 km), depending on the choice of battery [165]. “Smith Electric Vehicle” is an example of a battery electric truck with a payload capacity of 4,3 tons. The maximum range claimed is 100 miles (c.a. 160 km) and the vehicle has a maximal speed of 50 mph (c.a. 80 km/h) [166]. A picture of this truck from “TNT Post”, shown at the “PostExpo 2006”, can be seen in figure 5-29.


p.100

Jean-Marc

Rectangle

p.101

6 Postal requirements

6.1 Towards a better ‘last mile’ The postal service can be divided into different parts. The collection of mail through post offices and mail boxes can be seen as the first part of the journey that is made by post items. After collection, the mail items are transported to mail sorting centers, from where they continue their journey to the appropriate distribution center and finally the post office nearby their final destination. Transportation in these parts, sometimes over very long distances, can be made by different modes of transportation: cars and vans, large trucks, trains and even ships and airplanes. The distance covered in these first parts of the journey is much larger compared to the distance to be covered from the final post office from which it will be delivered shortly by the postman (or –woman) to the requested destination address. Therefore, the last part of the journey made by post items is sometimes called ‘the last mile’. The work of postmen, delivering mail items, is currently divided into two main parts. The first part consists of sorting the mail inside at the post office. The second part of the work consists of the postmen delivering the mail items outside on a pre-defined round [167]. The introduction of automatic sorting machines in the postal industry leads to a reduction of the working time of the inside part of the work of postmen in the near future. This will have an influence on the future organization of the work of postmen. If one wants to maintain the daily working time constant, the delivery part of the work of postmen will become longer. The latter will require changes in the delivery routes of the postmen that will become longer with more mail items to deliver. Also the mail items are changing. The volume of normal letters has decreased and the volume of non addressed mail and of parcels has increased. Additionally, not to increase the physiological stress at work and to preserve the postmen’s health over their career, one will require search for improved tools and/or vehicles for postal delivery [168]. Moreover, during the last decade in many countries a large part of the delivery rounds done by foot or by bicycle were replaced with rounds with a moped or with a car. Unfortunately, this evolution has lead to increased operational costs for the postal organization and an increased number of accidents and subsequently absenteeism. The increasing running costs are due to increasing fuel costs, insurance and maintenance. The frequent starts and stops while driving cause increased wear of parts (brakes, coupling) of cars and mopeds with combustion engines. But also the raising consciousness of climate and pollution issues, stimulate postal operators to continuously look for new and cleaner vehicles [160]. A clear interest for LEVs, suitable for postal delivery, exists among the European postal operators [169]. In particular electrically assisted bicycles for postal delivery seem to have a very high potential. This is also confirmed by the fact that several European Postal operators have launched call for tenders to purchase a large number of electrically assisted bicycles, tricycles and trolleys to implement in their postal vehicle fleet.

Postal requirements

p.102

As the postmen, delivering mails (and parcels) by bike, need to carry along a large payload, additional traction force coming from the electrical power system helps to perform his daily duty. Further, in the search of lowering the total cost of ownership of their delivery vehicle fleet, many postal operators are looking for alternative solutions for standard cars and mopeds. Further, congested cities make it increasingly expensive to organise timely postal deliveries with standard cars or delivery vans[169]. In addition postal companies are looking for more sustainable or greener ways of performing the delivery. Hybrid and electric vehicles are in this context a possible solution to lower the environmental impact and to give a ‘green’ image to the postal operators.

6.2 The NEPH project NEPH stands for New Electric Postmen Helper and is the acronym of the European industrial project that was officially launched in 2005 and formally closed in 2009. This project was granted the internationally recognised Eureka label (E!3364), that acknowledges it’s innovation and sustainability (see figure 6-2). The logo of this project can be seen in figure 6-1. The goal of Eureka is to stimulate research efforts of enterprises and to develop their innovation capacity [168]. The NEPH project was related to the study and development of a range of innovative electric power train systems designed for integration in a range of personal mobility devices, for helping postmen with their mail deliveries. This project had two clear main objectives:

• developing of a range of mobility devices to deliver mail in urban and suburban areas • contributing to the 'European Sustainable Development Policy'.

The range of NEPH power trains developed are designated for integration into the following devices: electric assisted trolleys, electric assisted bicycles and related types (three-wheelers, four-wheelers). As is symbolised by the logo of the NEPH project (see figure 6-1), the main focus of the project was however on electric two-wheelers for postal distribution.

Figure 6-1: Logo of the NEPH project Figure 6-2: Eureka-label

The industrial partner leading the NEPH project was the French battery manufacturer SAFT S.A. Two deputy leaders were appointed to practically organise and manage this project involving eleven partners. The Vrije Universiteit Brussel, department of Electrical Engineering and Energy Technology (ETEC) together with PostEurop were assigned to help managing the NEPH project. Different European postal organisations participated to this project and were gathered around the intermediate organisation PostEurop. The participating postal organisations were:

• Royal Mail Group (United Kingdom), • La Poste (France), • TNT Post (The Netherlands), • Poste Italiane (Italy) and • Itella (Finland).

The Belgian Post was involved as an observer. PostEurop and these postal organizations played a facilitating role in the specification phase of the project as well as for testing of the prototypes of the NEPH vehicles. The Vrije Universiteit Brussel was involved as a research partner and had to function,

Postal requirements

p.103

together with PostEurop as a neutral partner between the postal operators on the one hand and the industrial partners on the other hand. The NEPH industrial project is composed of different work packages on which more details can be found in publications about the NEPH project [46, 170, 171].

6.3 Questionnaire for postal operators A questionnaire was set-up to acquire insight in the postal operator’s experiences and needs for postal delivery. The goal was to collect information about the requirements from the postal companies concerning their vehicles used in their delivery fleets. For this purpose, an extensive questionnaire has been conducted among the participating postal operators of the NEPH project (see paragraph 6.2). The questionnaire consists of six parts:

• Part I: General information of the postal organisation • Part II: Actual use of postal vehicle fleet

o Information regarding the use of conventional vehicles o Information regarding the use and experiences with electric vehicles

• Part III: Interest for NEPH vehicle range • Part IV: Technical issues and standards • Part V: Service and maintenance requirements • Part VI: Cost related issues and market outlook

From this extensive questionnaire (48 pages long) a lot of information and data from the postal operators could be retrieved. An overview of most relevant elements from this information is discussed in this paragraph.

6.3.1 The postal vehicle fleet for delivery: a wide range of means of transportation

A large range of modes of transportation have been used in the long history of mail delivery. The commonly used modes of transportation used by postmen for the task of mail and parcel delivery vary depending mainly on the region or country, population density, topographic situation and weather conditions. Before the age of motorized vehicles, mail was delivered mainly by foot or by the use of horses (horseback riding) or by horse drawn coaches (mail transportation). Also other animals, such as camels were used for postal services in some parts of the world. For instance, postmen operating in mountainous areas with snow, use skis or snowshoes. Nowadays, postal organizations have a wide variety of vehicles at their disposal to organize the delivery part of the postal service. Mail delivery by foot is still broadly used, especially in densely populated areas (e.g. city centers, urban areas). Caddies are sometimes used on foot rounds to increase the payload capacity but also to reduce the ‘weight on the shoulder’ of postmen. Further, the postmen by foot sometimes use metro or trams to go from the postal office to their area of delivery. Next, bicycles are traditionally widely spread for postal delivery in urban and sub urban areas. Sometimes other kinds of human-powered vehicles (e.g. tricycles) are being used. Conventional motorized vehicles such as mopeds and motorcycles (with two and four stroke petrol engines), cars and vans (with petrol or diesel engines) are used by postal organizations all over the world. Finally, alternative vehicles technologies, such as electric and hybrid drive trains or engines with alternative fuels (e.g. compressed natural gas) are currently evaluated and used by many postal organizations. In chapter 5, an overview of the available different kinds of light electric vehicles for postal delivery are given. In order to investigate the experiences of the postal companies in regard with the use of different conventional vehicles for mail delivery, it was asked by means of the questionnaire to sum up the advantages and drawbacks for each different kind of mode of transportation. An overview of the common experiences for the different postal organizations with their conventional vehicles is given in the table 6-1 below.

Postal requirements

p.104

Table 6-1: Advantages and drawbacks of conventional vehicles for postal delivery

ADVANTAGES DRAWBACKS By foot No breakdowns, little equipment, no

problems with stairs Not protected from weather, shorter trips (need to refill), small payloads, Fatigue

Manual caddy

Higher payload than by foot, adapted for use in city centers, cheap, reduced fatigue

Lower payload than by car, not very adapted for parcels, not adapted for stairs, borders of walkways, slower than by foot

Bicycle Possibility to operate in areas with limited access, no emissions, cheap, no influence of traffic jams

Low payload capacity

Petrol Moped

Lower working load than bicycle, cheaper than car

Not possible to operate on all pedestrian or bicycle roads, expensive, technical durability, environment

Petrol or diesel car

High payload, high confort Traffic jams, expensive, environmental issues, noise

From the overview of advantages and drawbacks it can be concluded that for each delivery mode there exist shortcomings and often tradeoffs have to be made. It is clear that the trade-off will depend on the situation of the delivery round determined by its location. To cover all locations, a combination of different kinds of vehicles will be needed. There is no such thing as an ideal solution for all delivery rounds. The same question was posed in the questionnaire for their possible experiences with the electric vehicles they evaluated or used for mail delivery. An overview of the common experiences for the different postal organizations with electric vehicles is given in the table 6-2 below.

Table 6-2: Advantages and drawbacks of currently used electric vehicles for postal delivery

ADVANTAGES DRAWBACKS Electric caddy

Decrease of working load, less fatigue Enhanced payload

Cost, problems with some batteries (lead acid), size, maintenance

Electric bicycle

Decrease of working load, keep medically sensitive people at work

Quite heavy(e.g. steel frame) Limited range especially in winter conditions, maintenance, limited range

Electric 3 or 4-wheeler

Increased payload capacity, can drive on cycling roads

Heavier and wider than a conventional bicycle, expensive, maintenance

Electric moped

No emissions, silent Limited range, Cost, life-cycles of the batteries, sometimes limited torque

Electric car Silent, low energy costs, positive feedback from costumers

Limited range, high total costs, low life-cycle of the batteries, moderate payload

The same conclusions as with the table of conventional vehicles apply for the use of electric vehicles for postal delivery. However, there are also some specific remarks related to the use of an electric drive train. Early experiences with electric vehicles for postal delivery have pointed out that the range of the vehicle is an important issue. Also the weight and cost seems to be a point of attention. These issues can be related to the design of the battery pack (energy capacity) and the technical characteristics of the batteries that are used (energy density and cost). This shows the importance of better understanding the relation between required energy capacity of the batteries used and the application of postal delivery, characterised by a large diversity of situations. Results of detailed investigation of these aspects are described in chapters 7 and 9.

Postal requirements

p.105

6.3.2 Parameters influencing the choice of mail delivery mode Postal delivery requires route planning and is a typical example of what is called the Street Routing Problem in the field of Operational Research. The street routing problem is characterized by a high number of customers, a short service time for each customer, the density of the street network is high, the distance between two successive customers is short and sometimes complicated traffic regulations and restrictions are involved [172]. Postal operators use software applications to help to manage and optimize their route planning. An example of such a software application is Georoute from GIRO which is used by postal administrations in Belgium, Canada, Germany, Ireland, The Netherlands, Norway, Portugal and the United Kingdom [173, 174]. Saving on the last stage of the letter mail transport, “the last mile”, is an important way to reduce the total costs of postal organizations and therefore they constantly are seeking to optimize the routing of their postmen [175]. The postal organizations have a certain vehicle fleet available for delivery and this fleet is composed of different kinds of vehicles: mainly bicycles, mopeds and cars. The software for route planning applications can take into account customizable vehicle characteristics (payload capacity, maximum speed and traffic restrictions of the different mail delivery modes). The use of this software can lead to an optimization of the individual rounds, sometimes resulting in a reduced number of routes for a given area. Also, the software can be used to compare different modes of transportation (e.g. bicycle compared to moped) resulting in a best allocation of the available modes for mail delivery and consequently in a possible replacement strategy. In order to gain more insight in how the different postal companies of the NEPH project choose the mail delivery mode for the different delivery areas, it was asked by means of the questionnaire to sum up the parameters influencing this selection. From the feedback of the different postal organizations, the following parameters that influence the choice of a delivery mode or of a specific vehicle were identified:

• Total distance of the delivery trip or mission (Dmission, expressed in km) o Active distance: part of the round during which mail is distributed (Ddelivery,

expressed in km) o Non-active distance: part of the round during which the postman is not

distributing mail but traveling between the post office and the location of delivery

• Density of the distribution points

o Number of stops on the delivery round: denoted as stops o horizontal density (number of letterboxes per unit of distance) o vertical density (number of different letterboxes at one address or stop)

• Payload

o mass of payload: Mp, expressed in kg o volume of payload Vp: often expressed in liters or sometimes in m³

• Type of area o Topological circumstances

Slope or Road grade: grade, expressed as a percentage (see also paragraph 7.2.2 on page 114)

Cumulative height difference: hcumul, expressed in meters o Distance from the post office to the first delivery point (is related to the non-

active distance) o Town planning (accessibility of mail boxes, historical centers, street density, urban

area, suburban area, rural area) o Traffic situation (busy traffic, traffic lights, one way traffic, restricted access etc…)

Most postal operators have detailed information about the lengths of their current distribution rounds, on which they base their distribution strategy. An interesting data is the break up of the total trip length into the active part, during which mail is put into the mailboxes, and the non-active part, during which

Postal requirements

p.106

the postman is moving from the post office to the zone of distribution or back. An example of the cumulative number of rounds with a given length of the active and of the non-active part of the distribution round for non-electric bicycles and mopeds is given in figure 6-3 below. For confidentiality reasons, the source of these figures is not included.

Figure 6-3: Cumulative number of rounds with a given length of active and non-active part of the distribution round

Topological information concerning the delivery rounds is often not known and consequently not used at the high level management in any of the postal organizations. It appeared that for economic reasons the postal operators didn’t purchase the software module on topological data. At the local level this issue is often treated in a pragmatic way. However, these parameters are essential for a well-considered delivery strategy as well as to determine the required characteristics of the NEPH-power systems. As an illustration the topographic information of an extreme delivery trip by foot with assistance of a caddy is shown in figure 6-4.

Figure 6-4: Relative altitude and calculated cumulative climbing height difference of an extreme caddy delivery trip

0

10

20

30

40

50

60

70

0 200 300 750 850 950 1050 1550 1600 1700 1750 2100 2200 2850 3450 3600

distance from start point (m)

relative altitude (m)

cumulative climbing height difference (m)

height (m)

Postal requirements

p.107

We can read from the figure a total trip length of about 3,7 kilometre and a cumulative height difference of almost 60 meters for this delivery trip by foot.

6.3.3 NEPH implementation strategy As discussed earlier, the introduction of the automatic sorting machines has caused the outdoor working time to increase during the last decade. These changes have lead to a replacement of a large number “clean” delivery modes (by foot, by bicycle) with “polluting” delivery modes (mopeds and cars)[160]. However, due to high maintenance costs, risk of accidents and environmental concerns with mopeds and cars, postal organizations are looking for other solutions. For this reason, postal organizations are interested in LEVs suitable for postal delivery. The questioned postal organizations have however diverging opinions on which of the current vehicles should be replaced by which electric vehicles. However a common strong interest for electric bicycles for postal delivery was identified. The different devices proposed to the postal organizations in the framework of the NEPH project are listed in table 6-3 below altogether with the common features of the interest of the postal operators for each of these devices:

Table 6-3: Common level of interest for light electric vehicles for postal delivery

NEPH VEHICLE LEVEL OF INTEREST Electric bicycle Strong common interest Electric caddy Strong interest from most of the operators Electric 3 or 4-wheeler Moderate interest from some operators Electric moped or scooter Low interest

Further, the questionnaire has lead to the discovery of an interesting aspect of the vehicle replacement strategy: the substitution of (a part of) the conventional mopeds with electric bicycles could lead to an added value for the postal operators. This replacement strategy is based on the current situation of reliability and performance of thermal mopeds and scooters as well as on the related accidentology and noise production. An electric bicycle for postal delivery is claimed to combine the many advantages of a (conventional) bicycle with the added value of providing mechanical propulsion as a conventional moped. Furthermore, it should be a vehicle specifically developed for a high payload without allowing the high (and dangerous) top speeds as is the case with a conventional moped. The electric bicycle could consequently replace (part of) both conventional bicycles and mopeds in the fleets of the postal operators. The following requirements for future postal delivery business were also identified from the questionnaire: The delivery trips will get longer because of the increase of the use of highly effective automatic sorting machines. Many postal organizations wanted to have a reduction of the number of short-term storages or relay points as these appeared to be very expensive. In general, as European postal organizations have become commercial companies (or will be shortly), they continuously seek to increase the time efficiency of the complete delivery chain. Therefore, in their replacement strategy they do not want the overall cost to increase compared to the actual cost of the currently used delivery modes

6.4 Examples of light electric vehicle experiences Some European postal operators already had some experiences with light electric vehicles. Often these experiences are on a rather small level with internal tests (not in service) of single or small quantities of vehicles. Some larger postal organisations have larger internal projects with own developments. For instance the French Post has an internal project called VIF (Véhicules Innovatives pour Facteur). With

Postal requirements

p.108

this project the French postal organisation is, as many others, looking for alternative modes of locomotion that have added value compared to the actual situation:

• Keeping postmen longer in active service • Eliminating relay boxes • Entering urban zones, forbidden for thermal vehicles • Replacing motorised two-wheelers • Contributing to sustainable development

Further on this paragraph an overview of some experiences of European postal operators with light electric vehicles is given. In the Netherlands, TNT Post (former TPG Post, former PTT Post) had some first experiences from the year 2000 with mail delivery by using a tricycle with a large load container in the front. The Netherlands have a long tradition in using 3-wheel bikes. This tricycle with two front wheels is equipped with an electric motor, assisting the drivers’ pedal drive and allowed increasing the payload compared to a traditional bicycle. This electric tricycle was developed by Springtime and was marketed as the ‘Roodrunner’. The Roodrunner had a claimed load capacity of 150kg and 352 litre and was equipped with a 250W electrical assistance power unit. The trike can be equipped with 2 up to 6 battery packs of 12V – 30Ah. Two battery types are available: maintenance free lead acid or nickel zinc. However, TNT Post has meanwhile stopped using this tricycle because of too many technical problems.

Figure 6-5: Picture of the Roodrunner, Springtime used by TPG Post

(Picture from Philippe Lataire)

In Finland, a large part of the distribution rounds in the urban regions are done on foot using manual caddies or distribution trolleys. Moreover in most other European countries caddies are used in urban area’s. In the Netherlands, a trolley named ‘postboy’ was introduced in the early 60’s and was used for postal distribution in the cities. Using caddies for postal distribution is an effective way to increase the payload capacity and to lower the fatigue of shoulders and neck of postmen, compared to delivery on foot with a shoulder bag. Many variants of caddies exist with 2 up to 4 wheels and with a wide range of load capacity. Some manufacturers also offer caddies for postal distribution equipped with an auxiliary electric motor. An example of such an electric caddy for postal distribution is shown in the figure 6-6 below. Equipping a caddy with an electric motor allows to further increase the payload and to reduce the weight on the shoulder of the postmen.

Postal requirements

p.109

Figure 6-6: Electric caddy for postal distribution from Expresso GmbH

The example of the electric caddy is produced in Germany by the “Expresso Company” and is marketed as a mail distribution cart “MagiCexpress”. The “MagiCexpress” is fitted with sensor handles that include a hand detection system. Together with the control system, this allows the drive motors to be activated only if the operator has grasped both sensor handles firmly. Forwards/reverse movement of the caddy is determined by pushing or pulling the sensor handles. If the same force is applied to both handles, movement is along a straight line. If different forces are applied to the handles, the trolley moves in a curve. The two rear wheels contain both a permanent magnet EC-DC hub wheel motor of 200W. This caddy is capable of transporting 150kg of payload. The batteries used for this vehicle are two lead gel type batteries of 12V and 50Ah. Volume and weight for the battery pack in this vehicle is less critical as with bicycles for instance. For this reason, the lead gel batteries seem to be a good and cost effective solution.

6.5 Technical specifications: requirements or requests? An estimation of the toughest requirements had to be met by the NEPH vehicles was performed through an inquiry for delivery routes that were experienced as particularly difficult by the postmen. For instance, a cumulative height difference of 150m up to 200m during one trip with the NEPH bicycle had to be taken into account. Further, the vehicle should be able to climb slopes up to 15%. The active trip length and the horizontal density determine the number of starts and stops. It was observed that the total number of stops during for instance a bicycle delivery round is very high (up to 600 times). This leads to the fact that a significant amount of energy will be required from the electric power system and is much higher compared to the amount of energy needed for normal (non-postal) use. The maximum allowable payload mass, obviously, an important parameter and it very strongly influences the required size of the battery pack (see chapter 9) as well as the required traction force of the drive train (see paragraph 7.2 on page 113). In order to estimate the required energy and traction force needed from the electric power system, the collected data was used to calculate the energy capacity of the batteries and the torque from the motor. For this purpose a design-tool was developed and is described in detail in chapter 9 (see page 163). The torque delivered by the electric motor needs to be sufficiently large, especially if one wants to reduce the required effort on the pedals of the postmen. The force to be applied on the pedals of the bicycle was limited for reasons of ergonomics and health [168]. The requests and requirements of the postal organizations towards the performances of the electric vehicles to be developed were asked to the postal operators. This was finally converted to a set of technical specifications which were used as a reference during the development phase of the NEPH project.

Postal requirements

p.110

An overview of these technical specifications is given in the following three tables below. In table 6-4 the technical specifications for the NEPH bicycle can be found. In table 6-5 the corresponding figures for the NEPH caddy are given. Finally in table 6-6 the figures for the NEPH 3-wheeler can be found. With regard to the maximal speed of the vehicle on a flat road, two numbers are given. When looking to the figures in the different tables, a large correspondence between the figures for the NEPH bicycle and for the NEPH 3-wheeler can be found. The only difference is the payload mass which is higher in case of the 3-wheeler. This can be explained by the fact that an electric 3-wheeler is conceived as a way to further increase the payload capacity compared with an electric bicycle. A larger space for payload on a 3 –wheeler platform together with more stability due to the presence of 3 symmetrically positioned wheels should allow to increase the payload.

Table 6-4: Technical requirements for NEPH electric bicycle

TECHNICAL SPECIFICATIONS ELECTRIC BICYCLE ELECTRIC BICYCLE Symbol unit Average1 Maximal2

Payload mass Mp (kg) 50 70 Total trip length Dmission (km) 20 35 Average/Maximal speed on flat road v / vmax (km/h) 15/253 15/25 Number of Starts & Stops stops (#) 400 550 Speed attained between 2 stops vs (km/h) 10 10 Maximum road grade grade (%) 15 15 Speed on max grade vgrade (km/h) 6-7 6-7 Cumulative climbing height difference hcumul (m) 150 200

Table 6-5: Technical specifications for NEPH electric caddy

TECHNICAL SPECIFICATIONS ELECTRIC CADDY ELECTRIC CADDY Symbol unit Average Maximal

Payload mass Mp (kg) 50 150 Total trip length Dmission (km) 6 8 Maximal speed on flat road vmax (km/h) 5 5 Number of Starts & Stops stops (#) 400 600 Speed attained between 2 stops vs (km/h) 5 5 Maximum road grade grade (%) 15 15 Speed on max grade vgrade (km/h) 3 3 Cumulative climbing height difference hcumul (m) 50 100

1 The average specifications indicate performances that are feasible all at the same time 2 The maximum specifications indicate performances that are feasible, but not all at the same time 3 Two possible maximum speeds of the vehicle are feasible. However, higher speeds imply lower traction force capabilities

Postal requirements

p.111

Table 6-6: Technical specifications for NEPH electric 3-wheeler

TECHNICAL SPECIFICATIONS ELECTRIC 3-WHEELER ELECTRIC 3-WHEELER Symbol unit Average Maximal

Payload mass Mp (kg) 100 200 Total trip length Dmission (km) 20 35 Average / Maximal speed on flat road v / vmax (km/h) 15/25 15/25 Number of Starts & Stops stops (#) 400 550 Speed attained between 2 stops vs (km/h) 10 10 Maximum road grade grade (%) 15 15 Speed on max grade vgrade (km/h) 6-7 6-7 Cumulative climbing height difference hcumul (m) 150 200

In case of a bicycle , the payload is sometimes limited by internal rules of the postal organizations. Limits for individual bags or pouches are set, as well as limits for the total mass of the payload mounted on the bicycle. These limits envisage limiting the physiologic stress when handling the bags or pouches and driving the bicycle. Maximum value for total payload mass is applied for safety reasons when driving or handling the loaded bicycle. Despite these limitations, it was experienced by some postal organizations that in practice postmen often exceed these limits so they can shorten their working time by avoiding a refill. Cases where reported where postmen charge a bicycle with a total payload mass of 80 kg! The requirements of the postal operators with regard to the human-machine-interface (sometimes abbreviated as HMI) are quite limited. The basic function of an on/off switch with a clear indication by means of a LED is requested. If an ‘economic driving mode’ is present, this should be possibly activated and deactivated from the human-machine-interface. Further, the state-of-charge (abbreviated as SoC) of the battery pack is also required, and should be visible or readable while driving. As an option, the latter could be expressed in terms of remaining distance or remaining number of starts and stops. Additional, a state of health indication and an indication of when the next maintenance is required is found useful but not mandatory by the postal operators. Only as an option, a reading of the vehicle speed and a clock could be added to the human interface. In case of an electric caddy, a “dead man function” is requested for safety reasons. With reference to the battery pack of the NEPH vehicles, in trip recharging is not desired. Also battery pack exchange alongside the delivery trip is by most operators considered not acceptable because of the required logistics. So the total required battery capacity should, by preference, being carried from the start of the delivery trip onto the vehicle. For recharging the battery packs, it is important this can be done by simply connecting the charger to the battery pack and plugging in the charger. Each battery pack should have its own charger, or charging multiple packs should be automatically handled. A clear sign or signal that the charger is correctly connected should allow the user to be ensured that the charging procedure has correctly initiated. Finally, charging time should be less than 8 hours and should be possible from a standard electricity plug. It should be mentioned that the ambient temperature and conditions for charging should be consider. Moreover, charging must be possible in open air or in an unheated garage for instance. Charging at battery temperatures below 0°C can result in irreversible damage in case of Lithium-ion.

Postal requirements

p.112

Jean-Marc

Rectangle

p.113

7 Mathematical model for required

mechanical power and energy

7.1 Goal This chapter describes how the required mechanical power and energy from the electric bicycle can be predicted to satisfy the postal requirements, as described in chapter 6. Therefore it is important to impart knowledge about the different forces that act on a vehicle in general or on the bicycle in particular. In the successive paragraphs the different elements that influence the required mechanical power and energy are discussed [63, 134, 167].

7.2 Required mechanical power The power needed to propel a vehicle, along with its driver and possible payload, is determined by different elements. Some elements can be calculated using basic mechanical formulations; others can only be estimated using empirical relations. In this paragraph, an overview of the different resistive forces acting on a vehicle while driving is described. In particular the forces that play a role in the case of riding on a bicycle are discussed. From these forces, the corresponding mechanical power is calculated.

7.2.1 Rolling resistance The rolling resistance is the phenomenon of the energy losses caused by the contact of the tyres with the road surface and by tyre deformation. At very low vehicle speed and in absence of any significant wind or slope, the rolling resistance becomes dominant in case of cycling (not considering any acceleration). Rolling resistance can be divided into tyre resistance and ground resistance. Tyre resistance is caused by the deformation of the tyre to a much harder road surface. Ground resistance arises from pressing a tyre into a ‘soft’ surface. Empirically it was found that rolling resistance increased when additional load was carried. Further, parameters such as tyre pressure, temperature and vehicle speed are found to have an influence. However, empirical values for CR, representing a certain combination of circumstances, are used for vehicular power analysis.

αcos... gMCF RR = Equation 7-1

Mathematical model for required mechanical power and energy

p.114

The associated power PR, expressed in Watt, needed to overcome rolling resistance can be quantified as follows:

vgMCvFP RRR .cos.... α== Equation 7-2

Here, α is the slope of the road (see paragraph 7.2.2) and v is the speed of the vehicle, expressed in meters per second. An alternative empirical method to calculate the rolling resistance FR, taking into account the deformation of the tyre, due to total vehicle mass M, and the wheel diameter1, dw, is given by the equation 7-3.

.( /100 )

( / 26") 0, 25rc

Rw

f M kgF

d=

+ Equation 7-3

In this formulation, the influence of the road inclination is not considered. The factor frc has been determined empirically by the Heinzmann GmbH Company and is set to 10N.

7.2.2 Slope resistance Hill climbing resistance or slope resistance is the resistance that is experienced by any wheeled vehicle, for moving upwards a slope, which may occur along the vehicles’ route. The associated resistive force, in this text referred to with the ‘slope resistive force’ (FS, expressed in Newton), is determined by the total weight of the vehicle and by the slope of the hill that is concerned. The slope or grade of a road can be quantified in different ways. In the figure 7-1, an arbitrary road’s hill profile is represented:

�

Figure 7-1: Hill-profile – representation of the rise and run of a slope2

In mathematics, the slope of a line is defined by the quotient of the rise divided by the run. In the figure above, the rise is designated with Δh and the run is indicated by d, both expressed in meters. The slope of a road is often quantified by the grade, and can be defined between two different points A and B along the route:

1 Please note that the wheel diameter in this formula is expressed in inch (see conversion factor on page 238) 2 Bicycle used on this figure is from MIFA A.G.


p.115

.100% .100%h rise

graded run

Δ= = Equation 7-4

Another way to quantify the slope is to use the angle α that exists between a horizontal line through the point A and the line trough points A and B along the vehicle’s route.

arctan arctanh rise

d runα Δ= = Equation 7-5

The slope resistance force can now be expressed as follows:

αsin..gMFS = Equation 7-6

Note that this force can have either a positive or a negative value, corresponding to an upward slope or a downward slope respectively. The mechanical power that is needed for climbing a hill is the product of the slope resistance force FS and the speed of the vehicle v, expressed in meters per second:

vgMvFP sS .sin... α==

Equation 7-7

If the slope becomes negative (i.e. going downhill), the mechanical power related to the slope PS becomes negative and could lead to a negative value of the total power (see paragraph 7.2.4.) This negative total power results in an acceleration-force (see paragraph 7.2.5), could be regenerated through the electrical power system when capable to do so or dissipated in the brakes.

7.2.3 Aerodynamic drag Aerodynamic drag or wind resistance can become an important part of the total resistance, in case of high vehicle speeds (typically above 5m/s). Two main types of aerodynamic forces can be identified: one part of this force is normal to the surface of the resisted body (vehicle and driver) and the other part is tangential to the surface of the resisted body (skin friction). The formulation for the aerodynamic resistance force, FW expressed in Newton, uses a non-dimensional drag coefficient, CX. In cases where skin friction is dominant (for streamlined bodies) over the normal pressure forces, the surface area is used instead of frontal area, in combination with a “surface area drag coefficient”. In case of ‘normal’ road vehicles or bicycles in particular, the frontal surface, S expressed in m², is used in combination with a (frontal area) drag coefficient CX. The product of CX and S is multiplied with the dynamic pressure of the air flow to obtain the aerodynamic drag force:

20,5. . . .( )W X air wF C S v vρ= + Equation 7-8

Where:

S = frontal area, expressed in m2 CX = drag coefficient, non-dimensional ρair = air density expressed in kg/m3 v = vehicle or bicycle speed, expressed in m/s vw = headwind speed relative to the ground, expressed in m/s

The sum of the bicycle speed, v expressed in m/s, and the headwind speed, vw expressed in m/s, is called the relative air speed. The power that is needed to overcome aerodynamic drag, PW can now be expressed as follows:


p.116

2. 0,5. . . .( ) .W W X air wP F v C S v v vρ= = + Equation 7-9

0

50

100

150

200

250

300

350

400

0 1 2 3 4 5 6 7 8

bicycle speed (m/s)

Po

wer

nee

ded

to

ove

rco

me

win

d r

esis

tan

ce (

Wat

t)

0 m/s

3 m/s

6 m/s

12 m/s

21 m/s

value headwind speed:

Figure 7-2: Power needed to overcome wind resistance in function of the bicycle speed v at various values for the headwind vw

In figure 7-2 this power, needed to overcome wind resistance, is calculated in function of the bicycle speed. A value of 0,53 is considered for the product CX.S and 1,23kg/Nm³ for the air density ρair. The different lines correspond to different levels of head wind speed ranging from 0 km/h (no wind) up to 21 km/h.

7.2.4 Steady-state power equation The different power expressions related to the different resistive forces, described in previous chapters, can be combined to calculate the total required power Ptot to overcome all resistive forces at a constant speed condition:

( ).tot S R W S R WP P P P F F F v= + + = + + Equation 7-10

This is the power needed at the level of the driving wheel(s) of the vehicle. This equation can be rewritten to structure the steady-state power equation:

2. .(sin .cos ) 0,5. . . .( ) .tot R X air wP M g C C S v v vα α ρ = + + + Equation 7-11

7.2.5 Acceleration force If we want to increase the speed v of the vehicle, a propulsive force Fp higher than the resistive forces is needed. The change in vehicle speed can be related to an acceleration force FA. This acceleration force FA can be defined as:

aMdtdvMFA .. == Equation 7-12


p.117

It should be noted that not only the linear speed of the vehicle will increase but also the rotational speed of the rotating parts of the vehicle (wheels, pedals, motor…) will increase, with increasing vehicle speed. To take this into account, one could use a slightly higher apparent mass in the above expression instead of the total mass M. However, in this analysis, this aspect will not further be considered because of its little significance. See appendix 1 for an example of the moment of inertia of a wheel motor and the influence on the apparent mass. The acceleration force FA can now be introduced in the power equation:

( ) ( )2. . sin .cos 0,5. . . . . .tot R X air w

dvP M g C C S v v M v

dtα α ρ = + + + +

Equation 7-13

vFP tottot .=

Equation 7-14

The total force, Ftot in equation 7-14, is now the mechanical force that is required at the level of the wheels of the vehicle to overcome all drag forces and to achieve the acceleration, a, of equation 7-12. Example: Let’s consider the following situation: a postman is driving a cargo bicycle equipped with an electrical power system and is driving on the level at a constant speed of 5 m/s (18km/h). A small headwind of 1 m/s is present. The postman, fully equipped with his working clothes has a mass of 85kg and is transporting a useful charge of 45kg (mail items to be distributed). The electrical cargo bicycle has a net mass of 45kg (postal bicycle, electrical power system and empty mailbags). The total vehicle mass M in this situation is 175kg (including the driver). A rolling resistance coefficient CR of 0,0084 is corresponding to the use of large utility tires carrying a large mass over a rather rough asphalt road. A drag coefficient CX and frontal surface S of 1 and 0,53m2 respectively, are corresponding to the rather little aerodynamic shape of the cargo bicycle fully charged in combination with the position of an average postman driving his cargo bicycle. The different resistive forces can be calculated: The slope resistive force FS = 0 N The rolling resistive force FR = 0,0084 . 175 kg . 9,81 m/s2 . cos(0) = 14,4 N The aerodynamic drag force FW = 0,5 . 1 . 0,53m3 . 1,23 kg/m3. (5 m/s + 1 m/s)2 = 11,8 N In this steady state condition the mechanical power needed for the bicycle’s propulsion is: Ptot = ( FS + FR + FW ).v = ( 0 N + 14,4 N + 11,8 N ) . 5 m/s = 131 W The same conditions on a slope of only 1,5% will nearly double the required power: Now FS = 175 kg . 9,81 m/s2 . sin(arctan(0,015)) = 25,8 N And Ptot = ( FS + FR + FW ).v = ( 25,8 N + 14,4 N + 11,8 N ) . 5 m/s = 260 W From this example we can recognise the importance of the slope conditions on the required power.

7.2.6 Rotational speed and torque of the wheel The corresponding torque at the level of the wheel Tw_tot, expressed in Nm, can be derived from the total required power of the wheel:


p.118

_ .2

tot w totw tot

w

P d PT

vω= = Equation 7-15

Here, ωw is the rotational speed of the wheel, expressed in rad/s and dw is the diameter of the wheel in meters.

7.2.7 Required power related to climbing ability From the steady-state power equation from paragraph 7.2.4 we can also calculate the required mechanical power at the level of the wheel of the bicycle related to a desired climbing ability. From paragraph 6.5 we have an indication of the required (or requested) climbing ability for postal delivery with an electric bicycle. This climbing ability is expressed with a value for the maximum grade (grade in percent) and a value for the vehicle speed on the maximum grade (vgrade). The steady-state power equation 7-11 can be written in function of this climbing ability values.

( )1 2_ . . ( ).cos(tan ) 0,5. . . .( ) .req grade R X air grade w gradeP M g grade C grade C S v v vρ− = + + +

Equation 7-16

In this formulation, the grade is expressed in radians and the bicycle speed is expressed in meters per second. Example: Let’s consider the climbing ability requirements from table 6-4 and further the same conditions from the previous example: a postman is driving a cargo bicycle equipped with an electrical power system and is driving on a maximum road grade of 15% at a constant speed of 7 km/h (1,94 m/s). No headwind is present. The postman, fully equipped with his working clothes has a mass of 85kg and is transporting a useful charge of 45kg (mail items to be distributed). The electrical cargo bicycle has a net mass of 45kg (postal bicycle, electrical power system and empty mailbags). The total vehicle mass M in this situation is 175kg (including the driver). A rolling resistance coefficient CR of 0,0084 is corresponding to the use of large utility tires carrying a large mass over a rather rough asphalt road. A drag coefficient CX and frontal surface S of 1 and 0,53 m2 respectively, are corresponding to the rather little aerodynamic shape of the cargo bicycle fully charged in combination with the position of an average postman driving his cargo bicycle. In this steady state condition the mechanical power needed for the bicycle’s propulsion is: Preq_grade = ( M.g. (grade + CR) cos(tan-1grade) + 0,5.S.CX.ρair.(vgrade+vwind)2).vgrade = ( 175kg.9,81m/s.(0,15+0,0084).cos(tan-10,15) + 0,5.0,53m².1,23kg/m³.(1,94m/s+0m/s)2).1,94m/s² = 525,3 W Let’s consider an electric bicycle with a single driving wheel (rear) of 26 inch (or 66cm) diameter This situation requires a torque developed at the wheel of: Tw_tot = (0,66m / 2).(525,3W / 1,94m/s)= 87,5 Nm From this example we learn the important power and torque requirements corresponding to the demanding postal requirements


p.119

7.3 Required mechanical Energy In this paragraph a description is given of how the required mechanical energy to perform a certain journey or a postal distribution round has been estimated. An option to calculate the required energy would be to use a simulation programme, using the above power formulation to calculate the total mechanical power over time from a certain speed profile v(t). The total required mechanical energy, at the level of the wheels, would then be calculated as follows:

= missionT

tottot dttPE0

).( Equation 7-17

With Tmission the time duration of the mission or trip and Ptot(t) is the associated mechanical power profile of the trip. The total length of the mission Dmission can be calculated from the speed profile, or inversely the timeduration could be estimated from the total length of the mission.

0( ).

missionT

missionD v t dt= Equation 7-18

However, the speed profile v(t) often is not available or is difficult and expensive to measure in real life. Therefore another approach to estimate the required energy for a certain trip is proposed. A (round) trip of length Dmission by bicycle can be analysed by looking at the following elements:

• Accelerating the vehicle a number of times from standstill to a certain speed vs • Continuous driving during a certain driving time1 Tdriving on the level • Covering a certain cumulative height difference hcumul during the trip

The first two elements are often easy to calculate for postal delivery rounds. The third element sometimes is more difficult to acquire but can be determined from the route description by using a topographic map.

7.3.1 Accelerating from standstill The required mechanical energy to accelerate the vehicle to a certain speed vs can be expressed by a kinetic energy term:

2..5,0 sss vME = Equation 7-19

In this formulation, vs corresponds to the maximal speed that is reached between two stops.

7.3.2 Continuous driving The required mechanical energy for continuous driving on the level at a speed v during a certain driving time Tdriving, expressed in seconds can be calculated as follows:

1 The driving time is different from the mission time, as it exclude the time at which the bicycle speed is zero


p.120

( )2. . 0,5. . . . . .cd R X air w drivingE M g C C S v v vTρ = + +

Equation 7-20

The speed v used in this formula is an average speed value. The headwind vw is highly unpredictable and therefore difficult estimate. If the trip is a round trip, one could ease the above expression by setting vw to zero. For low values of vw it could be justified as an approximation by assuming that during the trip the wind direction and wind speed are constant and that the aerodynamic drag coefficient is symmetric for a cyclist. The driving time used in this formula can be estimated as:

missiondriving

DT

v≅ Equation 7-21

With Dmission the trip distance of the delivery trip (or mission) and v the average driving speed.

7.3.3 Covering a height difference The required mechanical energy related to covering a certain cumulative height difference hcumul, expressed in meters, during the trip can be calculated using the expression for potential energy:

cumulhd hgME ..= Equation 7-22

This quantity does not contain the energy that is needed to move. This is covered by the energy for continuous driving (see equation 7-20). The term cumulative height difference, as introduced in paragraph 6.3.2, is illustrated with a picture in figure 7-3.

Figure 7-3: Illustration of the term cumulative height difference

7.3.4 Influence of decreasing payload Typical for postal delivery is a decreasing payload while the trip is continued. This affects the power requirements and thus the energy requirement throughout the delivery trip. In particular the total vehicular mass M has to be précised for use in the different power and energy equations. The total mass can be defined as:

pdv MMMM ++= Equation 7-23


p.121

Where: • M the total mass or Gross Vehicle Mass, in kg • Mv the mass of the empty vehicle, related to tare weight or unladen weight, in kg • Md the mass of the driver or cyclist in kg • Mp the mass of the payload that is transported in kg

��

� ��

��

��

��

��

��

� ��

��

� � � � � � � � �

Figure 7-4: Decreasing payload over trip distance

A simplified but plausible evolution of the mass of the payload is represented in figure 7-4. One could define an average vehicle mass M’ and average payload Mp’ as follows:

( )epppppdv MMMMwithMMMM _0_0_ .21 −−=′′++=′ Equation 7-24

This average vehicle mass could be used in equation 7-19 when needed to calculate the required energy related to the starts and stops with decreasing payload. Moreover:

20,5. .stops

ss i si

E M v=

With Mi the total vehicle mass at start number i. When considering a steady decrease of the payload over the stopping points this formula can be rewritten as:

( ) ( )

( ) ( )

( )

2 2_

_ 0 _ 2_ 0

_ 0 _ 2_ 0

_ 0

_ 0

.0,5. ( ) .0,5.

* * .0,5.

* * .0,5.

*

stops stops

ss i s v d p i si i

stopsp p e

v d p si

stopsp p e

v d p si

p p

v d p

E M v M M M v

M Mstops M M M i v

stops

M Mstops M M M i v

stops

M Mstops M M M

= = + +

− = + + − − = + + −

−= + + −

( )_ 2.( 1)* .0,5.

2e

s

stops stopsv

stops

+


p.122

When considering a high number of starts and stops, this expression can be simplified:

( ) 2_ 0 _ 0 _

1* 1 2.( ) .0,5. 1ss v d p p p e s

stopsE stops M M M M M v if

stops

+= + + − − ≈

Or after identification with equation 7-24:

( ) 2 2(q.e.d.) * 0,5. *0,5. . ss v d p s sE stops M M M v stops M v′ ′= + + = Equation 7-25

It should be noted that the speed reached between two stops vs will fluctuate and is a function of the distance between the stopping points, but will also depend on the total vehicle mass at the time of acceleration. The mass to be considered in equation 7-22 expressing the required mechanical energy to cover a certain height difference, when taking into account a decreasing payload is more difficult to express. Two extreme situations could be imagined. In a first situation, the post office from where the delivery trip begins (and also ends), is located on top of a hill while the area for the delivery of mail (payload) is located in a valley. In the second and opposite situation, the post office is located in a valley while the area for delivery is located completely on the top of a hill. These two extreme situations are illustrated in figure 7-5 and figure 7-6 respectively.

Figure 7-5: Post office on top of hill Figure 7-6: Delivery area on top of hill

However, most likely the height difference of a delivery trip is distributed along the delivery route. For a precise calculation topographic information of the delivery trips is needed. If we assume a climbing profile, as illustrated in figure 7-7, that is evenly spread along the complete mission, we can use the average payload from equation 7-24 again:

( )[ ] cumulepppcumulhd hgMMMhgME ...21.. _0_0_ −−=′= Equation 7-26

��

� ��

��

��

��

��

��

� ��

��

� � � � � � � � ��

��

Figure 7-7: Evenly spread climbing profile


p.123

The mass M to be used in the equation 7-20 to calculate the required mechanical energy related to continuous driving part of the delivery trip, reflecting the energy to overcome rolling resistance and aerodynamic drag is again depending on the evolution of the effective payload during the delivery round. If we again consider an evenly spread decrease of the payload, the same expression for M’ (see equation 7-24) can be used:

( )2. . 0,5. . . . . .cd R X air w drivingE M g C C S v v vTρ ′= + +

Equation 7-27

7.3.5 Total required mechanical energy The total mechanical energy to complete a trip can now be attained by adding the different elements of the energy calculation as described in earlier paragraphs:

tot ss cd hdE E E E= + +

Equation 7-28

From this total mechanical energy, an estimation of the required electrical energy for the electrical power system of the vehicle can be made. But first we need to investigate the power train topology and power control strategies of this electrical power system. This will be done in chapter 8, in paragraphs 8.4.1 and 8.2.4 respectively. As an illustration, the two extreme situations (see paragraph 7.3.4) of the location of the post office and delivery area with regard to the height difference are used in a small analysis. For comparison, the situation of the height difference distributed along the delivery route is considered. For this analysis the following situation is considered: a trip with total length of 20km, a cumulative height difference of 100m, a constant payload of 45kg, 600 stops along the trip length and the speed reached between two stops is set at 12km/h.

Figure 7-8: Influence of the distribution of the height difference on the required mechanical energy

From this analysis we learn that the total required mechanical energy for the above mentioned situation is about 325 Wh. The contribution of the three different elements from equation 7-28 can be read from the figure 7-8. The contribution of the start and stop operation is the largest of all three (about 170 Wh). The contribution related to continuous driving is found to have a value of about 115 Wh. The contribution of covering the height difference (100 meters) is the smallest. The values for required


p.124

energy in case of the two extreme sitations range from 49 Wh to 37 Wh. This variation of 12 Wh represents a relative influence of about 3,7% on the total required energy for this particular situation. The situation of a distributed height difference along the delivery trip, corresponding to equation 7-27, leads to an intermediary result of 43 Wh.

7.4 Model for mechanical power and energy estimation based on postal requirements

As described earlier, a speed profile is often not available or difficult to measure. Therefore a model, based on the postal parameters, to estimate the required mechanical power and energy has been developed. Based on the postal conditions of use, described in chapter 6, an estimation of these requirements in terms of mechanical energy can be made. A Matlab™ script was developed to make an analysis of the required mechanical energy in function of the different ‘postal’ parameters.

Figure 7-9: Analysis of the required mechanical energy as function of different trip parameters


p.125

For this analysis a reference situation of a delivery trip was chosen. This reference corresponds to a trip with total length of 20km. During this reference trip a cumulative height difference of 100m is covered. A constant payload of 45kg and 600 stops were considered. The speed reached between two stops is set at 12km/h. This reference situation is represented by a purple star in the different plots from the Figure 7-9. In the different plots the calculated mechanical energy is shown, by means of lines of equal energy, for ranges of values for the relevant parameters: total trip distance, number of stops, the cumulative height difference, the speed reached between two stops and the payload. These graphs show how the required mechanical energy is evolving in function of the different parameters. If one read the different plots, it is important to keep in mind that often the different parameters are linked. For instance there is a link between the payload and the number of starts and stops or between the payload and the total trip distance. This means that not all points of the graphs can be linked to a realistic situation of use. However the area around the reference point can be considered as ‘realistic’ in case of postal delivery application. A sensitivity analysis has been made, to investigate which parameter has the largest impact on the required mechanical energy.

Figure 7-10: Sensitivity analysis of the model for required mechanical energy

For this analysis the same reference situation as earlier was chosen. Next, the different parameters are changed and the effect on the required mechanical energy was calculated. In the figure 7-10 the results of this analysis are plotted on one graph. Horizontally the variation of the parameter relative to the reference value is shown. Each line corresponds to the results for analysis of one parameter. In the legend of the graph, the different considered parameters are linked to the colours of the lines. From this analysis we can observe that the speed reached between two stops is by far the most significant parameter. The second most important parameter is the number of starts and stops that is performed during the delivery trip. This means that the kinetic energy term ESS from equation 7-19 is dominant in the total required mechanical energy for this specific practice of postal delivery.


p.126

These statements are important for the particular application of electric bicycles for delivery of mail & parcels. Indeed, it can be considered that most postmen want to minimize the time duration of their mission, hence using the maximum speed capability of their vehicle. The energy carrying capability of the bicycle being limited and the energy content being closely related to the weight and cost of the battery, limiting the maximum speed between delivery points is an important issue.


p.127

8 Drive technologies for electric bicycles

8.1 Introduction In this chapter an overview of the available drive technologies used for electric bicycles is given. For this purpose both literature and own work has been used. This chapter further contains five paragraphs:

• In paragraph 8.2 the constituent components of the electric power system used for electric bicycles are explained.

• In paragraph 8.3 the sizing of the energy storage system is made clear. The assistance factor and efficiency are defined that allow describing the relation between the required mechanical energy and the required electrical energy of the energy storage. This formulation will be used in chapter 9 to estimate the battery energy capacity.

• Paragraph 8.4 deals with the integration of the electric power system in the bicycle platform. An own classification of electric bicycles, based on the differences in the mechanical transmission of the motor power, is presented.

• In paragraph 8.5 the important paramter, driving range of electric bicycles, is discussed. Own results from a test campain whith a set of contemporary electric bicycles are presented in this paragraph.

• Finally, paragraph 8.6 descibes in more detail the electric power system that was used in the framework of the NEPH project in which the department played an important role.

8.2 Constituent power system components of electric bicycles The electric power system of electric bicycles is composed of different elements or components. For the description of the drive technology used in electric bicycles, the following elements can be distinguished:

• the energy storage system (in most cases a rechargeable battery pack) • the electric motor and motor controller • the human-machine interface and sensors used • charging and parking infrastructure

In some cases the power system is partially integrated into other elements of the bicycle or different elements of the power system are merged as one element. As an example the motor controller can be integrated into the bicycle frame or the controller can be integrated into the housing of the motor. In the paragraphs 8.2.1 till 8.2.5, these different components are discussed.

Drive technologies for electric bicycles

p.128

8.2.1 Energy storage systems for electric bicycles The on-board energy storage of an electric bicycle can be seen as a key component of the electric propulsion system as it determines its performances to a great extend. Over the past decades, significant improvements have been made in the energy storage capability of batteries. These improvements were induced by large development programs for space-, aircraft and military applications [167] but also by more specific programs for automotive traction applications (electric and hybrid electric drive trains) [132]. From these developments, new and more mature battery technologies became available for use in LEVs (particularly in electric bicycles). Batteries are not the only possible energy storage systems for electric bicycles. However commercially available electric bicycles are usually equipped with rechargeable battery packs, prototypes of electric bicycles (and other LEVs) using a fuel cell system combined with super capacitors or a battery pack as the (on-board) energy storage exist. Different battery technologies have been used for the various commercially available models of electric bicycles. The most commonly used types of batteries are:

• Lead-acid • Nickel-cadmium • Nickel-metal-hydride • Lithium-ion

Each of these types comes in a wide range of variants. For instance, lithium-ion has many types of cathodes and anodes that can be combined. Four main types of lithium-ion batteries, based on the cathode are currently mass produced: lithium-nickel-cobalt oxide, lithium-mixed oxide, lithium-manganese oxide and lithium-iron phosphate [132]. These different types of batteries are not only used for electric bicycles but also for other LEVs and for electric and hybrid electric vehicles. Beside these most common types of batteries, also other chemistries can be used (e.g. zinc air batteries, nickel zinc batteries, sodium nickel chloride and others). A noteworthy example are sodium nickel chloride (Na-NiCl4) batteries, better known as ZEBRA1 batteries. These batteries operate at high temperatures (270°C-350°C) and are used for specific applications such as hybrid busses or electric vehicle applications. The nominal capacity of a battery is generally expressed in milli Ampere-hours (mAh) at the cell level and mostly in Ampere-hours (Ah) at battery pack level. The capacity is often denoted with “CY” where Y is a number that indicates the time of the discharge period expressed in hours. For instance C10, is the battery capacity when discharged in 10 hours for a certain constant discharge current. The discharge current value corresponds to a complete discharge in exactly 10 hours. This quantity is always determined at a given value of battery temperature and at a certain discharge rate. However, information on the discharge period is often disregarded in product information. Or sometimes the discharge period considered by the battery manufacturer is not corresponding with typical discharge period of the envisaged application. The state-of-charge (or SoC) of a battery is the ratio of the remaining energy content of the battery to the nominal energy content and is expressed in percent. Depth-of-discharge (DoD) is the ratio of the amount of energy extracted from the battery to the nominal energy content.

1 Stands for Zeolite Battery Research Africa Project


p.129

Spec

ific

pow

er (W

/kg)

Specific energy (Wh/kg)

0 20 40 60 80 100 120 140 160 180 200

100000

10000

1000

100

10

1

Lead-acid

Supercapacitor

NiCdNiMH

NaNiCl

Li-PolymerLi-Ion

Figure 8-1: Ragone plot for different electrochemical energy storage systems (cell level)

Specific energy (or energy density by mass) is the ratio of the energy content of a battery to the weight of the battery and the specific power is the ratio of the maximal power1 of the battery to the mass of

the battery. Specific energy ( mδ ) is expressed in Watt-hour per kilogram while specific power is expressed in kilo Watt per kilogram. Electrochemical energy storage systems such as batteries and super capacitors can be compared by use of a bubble chart plotting specific power against specific energy. This specific type of chart is called the Ragone plot. Also other energy storage systems (e.g. fuel cell system, petrol engine system) can be represented on such a chart. The Ragone chart for a set of battery types and super capacitors is shown in figure 8-1 [113]. The volumetric energy density of a battery is the ratio of the energy content to the volume of the battery and is expressed in Watt-hour per litre. The cycle life of a battery is the number of charging and discharging cycles that can be made before the useful energy that can be retrieved has dropped below 80% of the nominal capacity. Cycle life is strongly dependent of the DoD and thus should be specified. This dependence is illustrated in figure 8-2 [176].

Figure 8-2: cycle life as function of the depth-of-discharge (example of lead-acid)

1 Maximum power of the battery is defined as the power at which the discharge current depresses the battery terminal voltage to 2/3 of the open circuit voltage (cfr. IEC 61982-3)


p.130

Usually 80% DoD is considered to express deep cycles that correspond with the use of the batteries for electric bicycle application. A realistic life span for lead batteries is approximately 200 - 300 deep cycles [177]. A realistic cycle life for nickel-metal-hydride batteries is about 300 - 500 cycles [177] and for lithium-ion batteries is about 500 - 1000 cycles [177, 178]. Beside cycle life a battery has also a limited life time in absolute time. This life time is called the calendar life and is expressed in years. Typical calendar life for nickel-metal-hydride batteries is 2 years. After the claimed calendar life, the battery pack has degraded below acceptable performances. The energy capacity will drop and the internal resistance will increase in most cases. Further, the self discharge of the battery will also increase due to aging effects of the battery. Obtaining correct information about both cycle life and calendar life is important for the potential users in order to avoid any misunderstanding or disappointment and to enable them to calculate real operational cost, including battery replacement. It is important to make the distinction between a battery pack and a battery cell. Battery cells are individual unit cells that, in most cases, have a cylindrical or prismatic shape or that have the shape of a pouch. A battery pack is a unit that is composed of a number of battery cells that are electrically connected in parallel or in series (or a combination). A battery pack might also be completed with electronics, sensors, fuses, active cooling system. These additional components are often enclosed with the strings of battery cells in a (rugged) casing. This casing provides the required mechanical robustness and ensures protection against moisture, water, dust, etc. Two different types of battery cell variants are available for most battery chemistries: optimized for energy and optimized for power. To understand the difference between both variants, the structure of the electrodes has to be considered. A technique used to improve the power performance of a cell is adding current collection points or modifying the structure of the electrode, which results in a lower internal current flow resistance [132]. Each of the above mentioned types is shortly discussed hereafter, highlighting their main characteristics.

8.2.1.1 Lead-acid batteries Lead-acid batteries (often abbreviated as Pb-acid) are probably the best known type of battery and have been used since a long time. This type is well known for powering the starter motor, the lighting system and the ignition in case of a spark ignition motor (also denoted as SLI batteries). Lead-acid batteries are also widely used for industrial electric vehicles such as forklifts. Two main different types of lead-acid batteries are available. The first type is the (classic) vented battery using a liquid electrolyte. This type requires regular maintenance (adding distilled water) to ensure good operation. A second type is the valve regulated lead-acid battery (VRLA). Two main types of VRLA batteries can be identified: absorbent glass mat in which the electrolyte is absorbed into a mat of fine glass fibres and the gell battery with a gelified electrolyte. This type of lead-acid batteries is sometimes called maintenance free lead batteries because, in contrast to the vented type, they do not require maintenance (inspection and topping up water). However, they are more expensive compared to the vented type and also require the use of special chargers to avoid overcharging and to ensure the claimed cycle life (typically 200 à 300 – but depending on the DoD) [176, 177]. For use in electric bicycles, the main disadvantage of lead-acid batteries is their low specific energy (typically 30 Watt-hour per kilogram [113, 177]) due to the use of lead plates for the electrodes. The volumetric energy density of lead-acid batteries is low, at about 70 Wh/L. Their low cost compared to all other types of batteries, is however a big advantage. For this reason this kind of battery is still used nowadays for electric bicycles. In particular the VRLA batteries are used for electric bicycles because of no need for maintenance and because of free orientation of the cells.


p.131

8.2.1.2 Nickel-cadmium batteries The specific energy of nickel-cadmium batteries (often abbreviated as NiCd1) is nearly two times larger compared to lead-acid batteries (typically around 50 Watt-hour per kilogram [113]). The volumetric energy density of nickel-cadmium batteries is much higher compared to that of lead-acid batteries: 100 up to 150 Wh/L. Nickel-cadmium batteries also have a high power density and allow fast charging. However, this type of battery emits hydrogen gases when charging and it has a rather high cost compared to lead-acid (about 5 times higher) [63]. A further disadvantage of the nickel-cadmium battery is the memory effect2. If a nickel-cadmium battery is repeatedly recharged after being only partially discharged, it will tend to lose its maximum energy capacity. The battery seems to remember the smaller energy capacity that was drawn before being recharged. This effect should not be confused with irreversible lose of capacity of batteries due to age and use. Another main disadvantage is the environmental concern about the presence of cadmium in this battery. The issue of containing cadmium and the comparison with other battery technologies was investigated in the framework of the European project SUBAT and is extensively discussed in the literature [48, 100, 113]. Meanwhile, nickel-cadmium batteries are, no longer sold in Europe because of the environmental and health impact of cadmium. Nickel-cadmium batteries are meanwhile replaced with either nickel-metal-hydride batteries or with lithium-ion batteries. However, in the past (from the early 1990s until about 2005) this type of battery was widely used for electric bicycles.

8.2.1.3 Nickel-metal-hydride batteries Nickel-metal-hydride battery (abbreviated as NiMH) is a type of battery similar to the nickel-cadmium battery. The negative electrode in this type of battery is made of a hydrogen-absorbing alloy instead of cadmium (as in the case of the nickel-cadmium battery). Nickel-metal-hydride has a somewhat higher energy density compared to nickel-cadmium [63]. Typical specific energy value for nickel-metal-hydride battery batteries is 65 à 80 Wh/kg [132, 179]. The volumetric energy density of nickel-metal-hydride battery batteries is about 140 - 170 Wh/L [179]. These values represent an advantage for the nickel-metal-hydride battery compared to the nickel-cadmium battery. This kind of battery can be sealed (< 60Ah). A disadvantage is the high cost due to the rare earth compounds used [180]. The self discharge rate of this type of batteries was somewhat higher compared (5 to 10% per day) to nickel-cadmium. New developments allowed overcoming this drawback by using a new type of separator in the cells. This type of battery is extensively used in consumer electronics. They are also suitable for high current drain applications due to their low internal resistance. The power-optimized version of nickel-metal-hydride battery batteries, are now fitted to commercialized hybrid vehicles such as the Toyota Prius [113] and Honda Insight. Also for LEV and electric bicycles this type of battery is widely used.

8.2.1.4 Lithium-ion batteries Two main variants of lithium batteries are currently in use for electric traction applications: lithium-ion and lithium-ion polymer. The latter is a variant of the lithium-ion battery in which the organic solvent electrolyte is (partially) replaced with a solid polymer composite. The main advantages of lithium-ion polymer over lithium-ion are the lower manufacturing cost and being more robust to mechanical stress. Lithium is the lightest metal with the highest electrochemical potential. Lithium-ion batteries (often abbreviated as Li-ion) are a type of rechargeable batteries in which lithium-ions move between anode and cathode. Lithium-ion batteries are not to be confused with the (disposable) primary batteries that use lithium metal or compounds as the anode. This type of battery (Li-ion) offers considerably higher specific energy performance compared to the other mentioned battery types. The specific energy value ranges from 130 up to 190 Wh/kg. The volumetric energy density for lithium batteries vary strongly from about 300 to 400 Wh/L [177, 181]. Lithium-ion batteries have better values for both weight and volume compared to nickel-metal-hydride battery batteries. The advantage in terms of weight is however more important (about half the weight) than the advantage in terms of volume.

1 NiCad is sometimes encountered but is a registered trademark of SAFT for their nickel-cadmium batteries. 2 Memory effect, also known as the lazy battery effect


p.132

Lithium-ion batteries have become common in consumer electronics, especially for portable electronics (laptops, cell phones…) because of their beneficial energy to mass ratio. Lithium-ion batteries have become more and more widely used for electric bicycle application as well. They do not suffer from the memory effect (see nickel-cadmium batteries). The main disadvantage consists of the safety issues. Lithium-ion batteries may explode when overheated or overcharged to an excessive voltage level. The safety risks can be lowered if appropriate measures are taken (thermal interruption for overcharge/discharge, vent for overpressure, shut down separator for excessive temperature). Further, lithium-ion batteries are characterised by a higher internal resistance, compared to nickel-cadmium and nickel-metal-hydride battery batteries.

8.2.1.5 Fuel cells Rechargeable batteries are not the only possible energy storage system for LEV. Also prototypes of LEV with fuel cells instead of batteries are being developed. Different types of fuel cell systems exist. The most known type is the hydrogen fuel cell that recombines hydrogen and oxygen into water while generating electrical energy (see also paragraph 3.2.5.2 on page 31 about fuel cell electric vehicles). The main obstacles for using fuel cells as energy storage for LEV are the high cost and the problems related to fuel storage, in particular when hydrogen is concerned (see also paragraph 3.2.4.4 on page 30 about hydrogen). Also the integration of the auxiliaries required is an important barrier for the use of fuel cells in LEV or electric bicycles. However, some fuel cells do not use hydrogen and are interesting options. The direct methanol fuel cell is a relatively new type of fuel cell system. The biggest advantage of this type of fuel cell is that it uses liquid methanol (CH3OH) as a fuel and thus avoids the use of hydrogen and that it does not require a reformer. In a direct methanol fuel cell, the anode functions as a catalyst to extract hydrogen from the methanol. An example of a fuel cell powered power train for LEV is the direct methanol fuel cell from the German manufacturer Clean Mobile A.G. This fuel cell is for instance integrated in the electric tricycle for postal delivery Tri-CarGo from Accell-Pro (see figure 8-3) [182]. This delivery tricycle is designed for a payload of 150kg in the front and 50kg in the rear. The fuel cell system uses a methanol cartridge of 5 litre and is combined with a 20Ah lithium-ion battery pack to power the electric power train [132, 183].

Figure 8-3: Accel Pro Tri-CarGo powered with a direct methanol Fuel Cell

8.2.1.6 Super capacitors Electric double layer capacitors, abbreviated as EDLCs, are capacitors with a very high energy density when compared to conventional capacitors. The specific energy density ranges from 5 up to 20 Wh/kg. EDLCs are characterised by a low rated voltage and use a small separator between the layers instead of a bulky electrolyte or dielectricum, resulting in an efficient packaging and high capacitance. EDLCs are better known and commercialized as super capacitors or ultra capacitors. An important advantage of EDLCs is the cycle life. Claimed cycle life for commercially available products is 1 million cycles (between


p.133

specified voltage and half rated voltage). Nominal capacitances C up to 3000F and rated voltage VSC_rated of typically 2,7 V are available today in large cylindrical cells with a volume of 0,475L [184].

2_ _ _ _

1 1. .

2 2SC nom SC rated SC rated SC ratedE CV Q V= = Equation 8-1

Where:

• ESC_nom is the nominal energy content of the super capacitor expressed in Joules • VSC_rated is the rated voltage of the super capacitor in Volt • C is the rated capacitance of the super capacitor, expressed in Farads • QSC_rated is the charge of the super capacitor corresponding with the rated voltage VSC_rated,

expressed in Coulomb (or Ampère seconds) Another important characteristic of EDLCs is that their voltage is proportional to their state-of-charge. This requires that either their operating range is limited to high state-of-charge or dedicated power electronics is used to compensate the widely varying voltage [63]. The relation between the voltage of the super capacitor VSC and its state-of-charge SOCSC is:

_ .SC SC rated SCV V SOC= with

_

SCSC

SC rated

QSOC

Q= Equation 8-2

Where:

• VSC_rated is the rated voltage of the super capacitor in Volt • SOCSC is the state-of-charge of the super capacitor at a certain value of the capacitors voltage

VSC, expressed in percent • QSC is the charge of the super capacitor corresponding with a certain value of the capacitors

voltage VSC, expressed in Coulomb (or Ampère seconds) Figure 8-4 shows an e-bike with a 250W hub motor and a 500F - 16,2V UltraCap module composed of 6 cells1 in series. The prototype uses a Heinzmann 250 Watt rated wheel motor in the front wheel. The freewheel has been removed from the hub motor by the manufacturer and allowed to use the motor to generate power for the recharging of the ultracapacitors. The range of the prototype was determined in practice and is 1,4 km of pure electric driving (no pedalling) [185].

Figure 8-4: The UltraCap e bike prototype of KaHo Sint-Lieven

1 The cells used are BCAP3000 ultracapacitors from Maxwell Technologies


p.134

8.2.1.7 Batteries from contemporary electric bicycle models A set of 12 electric bicycles, available on the Belgian and Italian market (anno 2008) was investigated by the department ETEC, in the framework of comparative tests commissioned by the Belgian consumer organisation

1. The main results obtained from this study have been described in references [186, 187].

An overview of the obtained weight data from the test samples is given in the figure 8‐5. The mass of the battery pack and the mass of the bicycle without the battery pack are shown individually in this figure. As the type of battery (chemistry used) and the nominal energy capacity affect the mass of the battery pack, this information

2 is also included in the graph. A large variation in the mass of the battery packs of commercially available electric bicycles can be observed. The lightest battery pack weighs only 1,9 kilogram while the most heavy had a mass of 13,7 kg. This large variation also causes a large variation of the total mass of the electric bicycles (including their batteries pack). The lightest electric bicycle of this set weighs 24,9 kg while the heaviest one weighs 38,7 kg. The heaviest bicycle uses a lead‐acid battery and the heaviest battery pack is of the lead‐acid type. However, its energy capacity also shows the highest value (504 Wh) of all battery packs of the set of test samples.

23,526,4

24,220,7

23,821,0

27,123,7 25,2 25,3

21,6

29,0

3,41,9 6,20

4,2

2,5 7,00

6,6 13,713,3 13,4

5,95

2,5

0

5

10

15

20

25

30

35

40

45

50

Testsample 1

Testsample 2

Testsample 3

Testsample 4

Testsample 5

Testsample 6

Testsample 7

Testsample 8

Testsample 9

Testsample 10

Testsample 11

Testsample 12

Weight (kg)

0

100

200

300

400

500

600

Energy capacity (W

h)

battery packbicycle without batteryenergy capacity

Figure 8‐5: Overview of battery pack and bicycle masses from commercial available electric bicycles

When one evaluates the batteries used with the different electric bicycles, three types of batteries can be identified: nickel‐metal‐hydride batteries, lithium‐ion batteries and lead‐acid batteries. The lead based batteries have the disadvantage to be heavy, but are very cheap. Nickel‐metal‐hydride batteries are significantly more expensive, but are used more frequently as they have a higher energy density compared to their lead‐based counterparts. Lithium‐ion batteries have the highest energy density and consequently are much lighter than the other types of batteries, for the same quantity of useful energy. Also the price of this type of batteries is much higher than for lead batteries, but becomes more and more competitive with nickel‐metal‐hydride batteries. Many of the recent models of electric bicycles are equipped with this kind of battery, which results in easier handling of the battery pack. The memory‐effect, that was prominent with the nickel‐cadmium batteries, is not an issue with the above‐mentioned battery technologies. Meanwhile, nickel‐cadmium batteries are no longer sold in Europe because of the environmental and health impact of the cadmium.

1 Test Aankoop / Test Achats – consumer organization Belgium 2 Energy capacity values based on manufacturer/supplier information


p.135

8.2.1.8 Selection of the energy storage system for electric bicycles The requirements for the energy storage systems for electric bicycles are manifold. The most critical and most argued requirements are the range requirement (or autonomy) and the attained cycle life of the battery. From the E-TOUR project (see also paragraph 5.6 on page 91), it was concluded that the autonomy was the main reason for disappointment regarding electric bicycles (models of the year 2000) [133]. The different battery technologies show different characteristics for these requirements. Further these characteristics depend on the battery temperature, discharge and charge current profiles and thus depend on the application or mission. The autonomy of a set of recent models (as from the previous paragraph) was determined in real life circumstances in the Brussels region when used for commuting. The results of these tests are shown in figure 8-24 on page 153. The energy capacity of an individual cell can be optimised (appropriate cell design and electrochemistry) but the energy capacity of a battery pack represents a different challenge. The different cells inside a pack have similar but never exactly the same characteristics due to variations in the cell manufacturing process. The cell voltage and temperature have to remain within certain limits to ensure safe operation, required capacity and acceptable cycle life. For this reason, a battery management system needs to monitor these different conditions and intervene when necessary. Voltage and temperature measurement at individual cell level is often not implemented for economic reasons; however most advanced battery packs have multiple temperature sensors and measure voltages at multiple intermediate levels. Then of course weight and volume, but also the cost, are determining factors when choosing a suitable energy storage technology. Weight and volume are restraining for electric bicycles when integrating the energy storage unit with attention to both ergonomics and vehicle stability. The presence of the battery pack should not hamper the cyclist while using the bicycle. Also, the battery pack should not restrain the possibility to load the bicycle with a useful charge. The latter is even more important in case of electric bicycles for postal distribution. Further, the position of the battery pack has to be well considered with regard to stability of the bicycle at stand still and the stability when driving. These aspects will be discussed more extensively in paragraph 8.4.2 on the integration of the battery pack on the bicycle platform.

8.2.2 Electric motors used for electric bicycles This paragraph will briefly introduce the different motor types in connotation with their use in electric bicycles. Many different types of electric motors are being used for electric bicycles. The most popular type of motor for electric bicycles however is the brushless d.c. motor [133]. Another type of motor that is regularly applied for electric bicycles is the brushed d.c. motor. However, brush wear can be a concern in case of brushed d.c. motors. Nevertheless, the brushes can in some cases be replaced throughout its lifetime as a part of maintenance of the motor or of the electric bicycle. Induction motors are on the other hand hardly ever used for small size electric vehicles [188]. A noteworthy exception is perhaps the linear induction motor integrated with the rim of a bicycle, developed by Prof. Hofer [189] The use of high-energy permanent magnets for electric motors for this application is widely spread and allows increasing the overall efficiency as they produce the motor’s magnetic field without consuming electrical energy. An overview of d.c. motors for light electrically driven vehicles is provided by dr. Gössel [188]. The family of d.c. motors can be listed as follows:

• Mechanically commutated motors o Permanent magnet excited o Electromagnetically excited

• Electronically commutated motors o Permanent magnet excited o Switched reluctance

Further, a distinction can be made between axial flux motors and radial flux motors. Axial flux design is often used in hub motors or wheel motors in combination with a disc shaped rotor (also called disc motor). Hub motors using an internal gear system as well as direct drive (no gear) exist. The use of a


p.136

gear system allows optimizing the volume and the torque range of the motor. However, disadvantages are higher noise production. Over the years, the hub motor has become widespread in the current electric bicycle product range. One of the main reasons for this evolution could be the advantage that the hub motor requires only little adaptation of the bicycle frame. On the other hand, it is quite a challenge to design a slim hub motor that can be fitted in the rear- or front fork of a bicycle.

8.2.3 Motor controllers for electric bicycles The electric power provided to the motor used in an electric bicycle has to be controlled in function of the operational parameters and of the driver’s input. This is done by the motor controller that is often customized for the motor that is used. Therefore electric motors and their controller for use in electric bicycles are in most cases seen as a fixed pair of components. Depending on the type of motor that is used (see paragraph 8.2.2), different types of motor controllers are used for electric bicycles. The main function of the controller is to regulate the current that flows to the motor and in this way controlling the torque of the electric motor. In most cases the magnetic field of the motor is available from the use of rare earth permanent magnets. Nevertheless, in some cases the motor controller also has to regulate the excitation current for the magnetic field of the motor. Most common d.c. motor controller for electric bicycles is the step down converter (also called the buck converter) with a one quadrant operation (forwards motor operation). This converter is based on a basic converter topology requiring only one power switching component. Depending on the power and voltage range, simple MOSFETs are used to control the motor armature current. In case electromagnetic excitation is used (instead of permanent magnet excitation), the motor controller contains an additional field current chopper (with transistors) to control the (low) field current. For a.c. motors and for brushless d.c. motors, the motor controller exists of a voltage source inverter fed by the d.c. voltage from the battery pack (or other energy storage system). This inverter is composed of one or multiple legs that create the desired waveform by activating the switches in the appropriate sequence and timing. The number of inverter legs depends on the number of phases of motor but in most cases three phase motors are most common. In case of power converters for brushless d.c. motors, the rotor position has to be detected. This can be done by using a sensor system (Hall sensors) or by observation of the motor voltages. Beside the basis function of controlling the motor power, other functions have to be ensured by the motor controller of an electric bicycle. Maximum speed of the vehicle at which the electric motor delivers power, dictated by legislative constraints (see also paragraph 5.4.3 on page 90) could be fulfilled by the natural characteristic of the motor or by proper selection of the gear system, however in some cases this is controlled electronically through the intermediate of the motor controller. Motor overheating prevention is often integrated into the motor controller unit. The modulation of the electric motor power can be made dependent of the driver’s input by the use of sensors. This input can be the pedal frequency and the cyclist’s effort which is measured by use of a pedal-force sensing device as for EPACs (see paragraph 5.4.2 on page 89). On some models, different assistance levels (high, normal, economy…) can be selected by the driver by pushing buttons on a control panel or a display unit (in most cases mounted on the handlebar). In other cases, the controller output is regulated by means of a twist grip. In some cases also the battery management system (BMS) is integrated in the ‘motor controller unit’.

8.2.4 Human-machine-interfaces and sensors used for electric bicycles In the context of this work, the term human-machine-interface (hereafter HMI) means the assembly of command buttons, display units, twist grip, status led’s etc. These HMI come in a wide variety of appearances and levels of functionalities. The following functionalities of the HMI for electric bicycles can be identified:


p.137

• On/off of the electric power system • State-of-charge indication • Charging status • Different modes of electric assistance • Modulation of the electric assistance • Anti-theft protection • Failure and maintenance indication • Command of the lighting system

In the most basic versions of a HMI an on/off button is foreseen to switch the electrical system of the electric bicycle on and off. The location and the reachability of this on/off button for the cyclist is however different for the different models on the market. The on/off button is sometimes combined with a key switch that disconnects the battery voltage from the electric power unit. Next to this, an indication of a low State-of-charge (hereafter SoC) of the battery pack is available on most models. The resolution of the indication of the SoC is different form one model to the other. A separate SoC-indicator is sometimes available at the level of the battery pack. In particular when the battery pack is removable, this additional SoC indication is often foreseen. This indicator on the battery pack allows checking the SoC of the battery separate from the bicycle. An indication of charging status of the battery can sometimes be found. This status indication can be found on the battery pack, on the display or on the charger. The indication of actual charging is sometimes not available, which can result in inconvenient situations. The selection of different modes of electrical assistance is available on many models. A picture of a control panel with buttons for the selection of different modes of assistance is shown in figure 8-6 (picture on the left). More advanced HMI are using a monochrome display and integrate the functionalities of a cyclocomputer1, as often used on regular (non-electric) bicycles, providing the speed, trip distance, time, average speed, maximum speed, ambient temperature etc. Some models use a twist grip or a thumb switch for the modulation of the electric assistance level. This is used for the so-called power-on-demand systems (see also paragraph 8.2.1.8). A picture of a handle bar with a thumb switch integrated is shown in the figure 8-6 (picture on the right).

Figure 8-6: Selection of different assistance modes (left) and modulation of electrical assistance with a thumb switch (right)

Also anti-theft protection is sometimes integrated in sophisticated HMI systems. An example of such a protection is a removable display that functions as an electronic key. The bicycle can only be used if the display is connected and the display has an identification code that recognizes if it is connected to the right and authorized bicycle. A key switch can also be seen as an anti-theft protection and is available on some models of electric bicycles. Failure indication and maintenance information is integrated in the HMI of some electric bicycle models. A key switch is sometimes foreseen and disconnects the battery voltage from the electrical power unit.

1 Sometimes the term speedometer is used


p.138

Finally, the lighting system of the bicycle is sometimes integrated. In this case, buttons that switch on and off the lights are added on the command or display unit. An example of an advanced display unit is shown in the figure 8-7.

Figure 8-7: Display of the Sparta ION electric bicycle

Beside the HMI, sensors are used to retrieve more input from the driver or from the bicycle. This input is in some cases required for the used control strategy. Two main power control strategies of the electrical assistance can be identified:

• Pedal power control system • Power-on-demand system

Detection of pedalling is required for complying with the European directive 2002/24/EC (see paragraph 5.4.3). Most electric bicycles thus make use of a detection device for sensing the rotation of the pedal axis (see an example in figure 8-8). In the standard EN 15194 for EPACs it is further specified that assistance shall be provided only when the cyclist pedals forward (see also paragraph 5.4.4). Depending on the control strategy and on the manufacturer’s system design, other sensors are also used. In the case of electric bicycles with measurement of the effort of the cyclist (force applied on the pedals) other type of sensors are used. This kind of system is described as a pedal power control system (see paragraph 8.2.1.8). For this type of electric bicycle the contribution to the traction force of vehicle by the electric motor (Ptm) can be made proportional to the contribution of the cyclist (Ptc). For instance, the assistance motor could be assigned to double the effort made by the cyclist. Which distribution is chosen between the motor power and the power from the cyclist can further be controlled in function of the other driving parameters (speed, different assistance modes selected by the user…). It concerns a technically quite complex system, which can be found on the more expensive bicycle models. These systems offer a great ease of use, in particular when starting from standstill as no additional handling is required to activate the electric assistance. The assistance factor ξ(t) has been defined in paragraph 8.2.1.8 and allows to quantify the relationship between the power delivered by the electric motor Ptm and the total traction power Ptot. The assistance factor for this kind of systems has been described by equation 8-7 and shows the influence of the driver’s input through the pedal torque and the pedal frequency. Pedal force sensor technology for electric bicycles is a well-patented area of technology. In many models, the pedal sensor is completely integrated within the drive unit and thus the pedal sensor is not available as a separate OEM product but is rather considered as a part of the complete drive unit. Most pedal sensor systems are based on the measurement of a mechanical strain or deformation, making use of strain gauges. Strain gauges can be applied either on a non-rotating part (e.g. on the rear wheel axis) or on the rotating parts, using means to power the strain gauge bridge and to receive the measurement signal. These means can be slip rings, rotary transformers or wireless communication. Other, more sophisticated systems make use of dynamic torque transducers based on for example the measurement of magnetic reluctance (see example in Figure 8-9). In the case no sensor is used for measuring the cyclists effort, the electric motor power can also be modulated manually by the cyclist, by means of a twist grip or with a push button (see Figure 8-6, right picture). This is the case for the so-called power-on-demand system. Again different systems are


p.139

available, each resulting in a different level of ergonomics and comfort. The assistance motor now only contributes when the pedals are turning, possibly without the cyclist needing to push them hard. The assistance factor for this kind of systems has been described by equation 8-8 and shows the influence of the driver’s input through the twist grip position (Tgr) and possibly also the pedal frequency.

Figure 8-8: Pedal sensor of the Heinzmann drive system [142] Figure 8-9: Contactless pedal torque sensor [190]

8.2.5 Chargers and infrastructure In most cases the charger for an electric bicycle is provided as a separate component. However, in some models, the charger is built-in the bicycle or the battery pack. An example of an electric bicycle with an on-board charger is the “Flyer F-series” from Biketec [139]. An important remark concerning battery chargers for electric bicycles is their protection class. Often chargers can not be used outdoors or even in humid rooms. This requires that the battery (or the bicycle) has to be brought inside a (non-humid) room to be charged. In some cases warning pictograms are put on the charger to avoid using the charger in the rain. Practical experience [186] has also pointed out that some chargers show important heating and so attention to the placement of the charger is required when using these specific models. On some chargers a pictogram is put not to place the charger in the sun, or it is mentioned in the user manual to place the charger on a hard surface and not to cover the charger. Chargers are designed to be connected to the utility grid and accept nominal input voltages of 115V/240V a.c. and a mains frequency of 50/60Hz. Typical power ratings for these chargers are 200 up to 350 Watts. The output current of these chargers are typically 1 – 4 Amps. The procedure to connect the charger can differ from model to model. In some cases a specific sequence of the connections has to be respected (e.g. first plugging in the charging plug into the battery and then connecting the charger to the utility grid). Different indicator LEDs can be found on different models of chargers for electric bicycles. In many cases there is a LED indicating if the charger is connected to the utility grid. In some cases a LED indicates that the charging of the battery is in progress. The indication of a completely charged battery is sometimes foreseen. The time needed for recharging the battery of an electric bicycle is also an important aspect of their potential use. The charging time and the energy consumption for the grid were recorded for a series of electric bicycles [186, 187, 191]. The electric bicycles were used until the battery packs were completely discharged (complete shutdown of the system) and successively the battery was recharged in the laboratory with registration of both the charging time and energy consumption. This test was repeated at least 3 times for the different models and the average values of these tests are shown in figure 8-10.


p.140

4:20

5:08

7:05

5:35

3:36

2:54

2:05

1:58

5:41

5:42

3:57

5:00

0

50

100

150

200

250

300

350

400

450

500

550

Test

sam

ple

1

Test

sam

ple

2

Test

sam

ple

3

Test

sam

ple

4

Test

sam

ple

5

Test

sam

ple

6

Test

sam

ple

7

Test

sam

ple

8

Test

sam

ple

9

Test

sam

ple

10

Test

sam

ple

11

Test

sam

ple

12

Ener

gy c

onsu

mpt

ion

(Wh/

com

plet

e ch

arge

)

0:30

1:00

1:30

2:00

2:30

3:00

3:30

4:00

4:30

5:00

5:30

6:00

6:30

7:00

7:30

8:00Charging time (hh:mm)

Energy consumption (Wh/complete charge)

Charging time (hh:mm)

379Wh

260Wh

240Wh

204Wh

260Wh

173Wh 180Wh

504Wh

370Wh504Wh

240Wh

252Wh

Figure 8-10: Energy consumption and charging time for a set of 12 electric bicycles

These test samples are corresponding to the test samples from figure 8-5 in paragraph 8.2.1.7. Quite large variations in charging times of completely discharged batteries, could be observed (from less than 2 hours up to 7 hours of charging time). This is an important element to consider if one would decide to recharge his battery pack underway (in particular if the trip is larger than the driving range). The energy taken from the electricity grid to fully recharge the batteries differs also from one model to the other, however it is related to the energy capacity of the battery pack. EnergyBus [132, 192] is a new industry standard for connecting the components of LEVs. One of the standardisation topics is the use of a plug family for chargers. This initiative could improve the exchangeability of chargers and batteries for electric bicycles. Parking infrastructure for bicycles in the cities, in particular at public places (e.g. at railway stations), has known important progress. However, still large improvements need to be made in order to further increase the modal shift towards bicycles or other light vehicles. Especially the complementarity with public transportation is key [5]. Some initiatives with rental services for bicycles in large European cities know great success (e.g. Villo! In Brussels [193], Velib in Paris [194]). Plans for large scale rental stations for electric bicycles exist and might be reality in the near future. The electric bicycle can in this way play an important role in the improvement of urban mobility through substitution of other means of transportation. Moreover, electric bicycles were found to be suitable for commuting as well as for shopping and leisure [133, 195]. Compared to non-electric bicycles, the requirements for infrastructure are largely similar (safe cycling lanes, parking, etc.). However, one important difference is the need for recharging the batteries. At most public parking facilities for bicycles, a power outlet to plug in the charger is not available. The required adaptations to enable recharging of electric bicycles, is rather limited. Preferably this would be implemented on a covered (protection from rain) and secured (to prevent theft) parking facility. The required connection is standard and the power levels involved are limited. The energy consumption is also limited, so a subscription fee system, giving access to the parking facility and the power outlets, would be preferred. An interesting example of new initiatives for bicycle storage and sharing systems is the Bike Tree project [196]. The Bike Tree is a (solar powered) bike parking system with a visual appealing (and remarkable) design and offers the possibility to recharge an electric bicycle. A picture of the Bike Tree is shown in figure 8-11. Another example of a parking station is the Japanese underground parking system with an automatic moving carrier that receives the bicycle above ground and then stores the bicycle in the underground storage space. A picture of the Tokyo underground parking system is shown in figure 8-12 [132].


p.141

Figure 8-11: Bike Tree - bike sharing and storage system

Figure 8-12 Japanese underground bicycle parking system

Also parking stations especially designed for electric bicycles exist. An example is the solar powered parking station for electric bicycles from Sanyo [197] (see Figure 8-13).

Figure 8-13: Sanyo Solar Parking Lot for electric bicycles

Battery exchange stations for electric bicycles exist and are used in Germany to service the Deutshe Post. These battery exchange stations are placed in apartment buildings and allow to pickup a fully charged battery pack along their delivery route. These exchange stations are at the same time used as charging stations and thus a (partially) discharged battery can be inserted and can be recharged. Postal operators that have large fleets of electric bicycles are interested in charging stations to manage the recharging of a large number of batteries. This recharging station could offer additional functionalities such as battery


p.142

state of health monitoring. A number of spare batteries could be made available to ensure or enhance the operability of the fleet of electric bicycles for postal delivery.

8.3 Sizing of the energy storage system When considering the sizing of the energy storage system, in particular a battery pack, starting from a set of given mission requirements (as discussed in paragraph 6.5) and the physical properties of the vehicle (as discussed in paragraph 7.2), there exists an interaction between the mass of the energy storage system, and the physical properties of the vehicle (volume and mass). The physical properties, together with the mission requirements determine the required mechanical energy and thus indirectly the required electrical energy storage. This influence is schematically represented in the figure 8‐14. The different symbols used in the scheme are commented in the remainder of this paragraph.

Figure 8‐14: Influence of the energy storage technology on the required mechanical energy

The vehicle’s physical properties in this scheme are the vehicle mass Mv, the mass of the driver, Md, the aerodynamic drag coefficient and associated surface, S.CX and the condition of the tires, determining the rolling resistance coefficient CR. In this regard some additional external factors, acting on the vehicle, should be considered as well: the air density ρ and the headwind speed vw. In this context the vehicle mass should be split up into the mass of the bicycle platform (without the electric power system), Mplat, the mass of the battery pack, Mbat, and the mass of electric motor unit, Mmot.

v plat bat motM M M M Equation 8‐3

The mission requirements are defined by the geographical specifications of the trip, the payload mass, Mp, and its variation along the mission trip length Dmission. The latter is determined by the geographical


p.143

distribution of the stops (or delivery points) along the route. The road grade or slope α along the trip will also influence the required mechanical energy. The road type(s) and the tires determine the rolling resistance coefficient CR. The driver’s input determines the contribution of the electrical motor power unit to the total required mechanical power and energy. The influence of the driver on the motor power depends on the type of system or technology used and can be controlled through the use of a twist grip, a pedal sensor or other variants (see also paragraph 8.2.4). The electric power unit technology will also determine the vehicle’s performance (e.g. its maximal acceleration). In turn, the vehicle’s hybrid propulsion system performance influences the speed cycle, v(t) or v(Dmission), of the vehicle and influences in this way also the required mechanical energy. See also equation 7‐14 and equation 7‐17 in paragraph 7.2 and paragraph 7.3 respectively. The electric power unit technology and the driver’s input will determine how much electric power is delivered and how much electric energy has to be delivered by the electrical energy storage system. The total required mechanical traction power Ptot from equation 7‐17 can be split into a contribution from the cyclist Ptc and a contribution of the electric motor Ptm. These power quantities are considered at the level of the driving wheel(s). This can be expressed by the following equation.

tot tc tmP P P Equation 8‐4

Integration of equation 8‐4 over time gives rise to the definition of the required mechanical energy allocated to the electric motor at the level of the driving wheel Etm, related to a certain driving cycle. The considered driving cycle has total time duration of Tmission (expressed in seconds).

0( ).

missionT

tm tmE P t dt Equation 8‐5

The assistance factor ξ(t) can be defined, quantifying the relationship between the power delivered by the electric motor Ptm and the total traction power Ptot . The definition of the assistance factor allows making a comparative performance analysis of electrical assisted bicycles. A dedicated testbench was developed for such analyses at the department ETEC and has been described in the literature [133, 159, 195, 198‐200].

( )( )

( )tm

tot

P tt

P t

Equation 8‐6

This assistance factor ξ(t) is a number between 0 and 1 (and is often expressed in percentage). If all traction power is coming from the cyclist, ξ equals 0 and when all traction power is coming from the electric motor, ξ equals 1. As is mentioned earlier, the value of ξ(t) will be influenced by the driver’s input and by the electric power unit technology. In other words, this parameter may change by the way the cyclist is using the electric bicycle and can be seen as a user‐dependant parameter. Two important types of main power unit technologies for electric bicycles can be distinguished in this regard. A first technology can be described as a pedal power control system such as the Yamaha PAS system (see also paragraph 5.4.2). The traction power delivered by the electric motor Ptm depends in this case on the cyclist’s effort. The cyclist’s effort is measured by the use of one or more sensors (see paragraph 8.2.4). The control strategy used in the motor power unit can be described by using the earlier mentioned assistance factor ξ(t). In case of a pedal power control system, the assistance factor can be expressed as a function of the cyclist torque exerted on the pedals Tc(t) and of the rotational speed of the pedal axis ωp(t).

( ) ( , )p ct T Equation 8‐7


p.144

A dedicated test bench for electric bicycles was developed at the Vrije Universiteit Brussel for measuring the relationship ξ(t)between the power delivered by the electric motor Ptm and the total traction power Ptot in function of the bicycle speed v(t) and the cyclist’s torque Tc(t) [133, 159]. The assistance factor ξ(t) can be made dependant on different pre-defined assistance modes (e.g. eco-mode, normal mode…) that can be chosen by the cyclist on a display unit or panel (see also paragraph 8.2.4). A second kind of technology can be described as a power-on-demand system. This type of power unit technology makes use of a twist grip or thumb switch for the input of the driver. This input is then used for modulating the electric motor power. In many cases, the rotational speed of the pedal axis ωp is also measured and used to control the motor. This can be described by expressing the assistance factor ξ(t) as a function of the rotational speed of the pedal axis ωp(t)and twist grip position Tgr(t).

( ) ( , )p grt Tξ ξ ω= Equation 8-8

Next to the instantaneous assistance factor ξ(t), a global assistance factor missionξ related to a certain

mission (or driving cycle) can be defined as the quotient of the required mechanical energy from the electric motor Etm (as defined in equation 8-5) and the total required mechanical energy Etot, at the level of the wheels. The considered driving cycle has total time duration of Tmission (in seconds).

0

0

( ). ( ).

( ).

mission

mission

T

tottmmission T

tot tot

t P t dtE

E P t dt

ξξ =

Equation 8-9

This parameter missionξ can be used to determine the required mechanical energy allocated to the

electric motor, Etm (see equation 8-5), from the required total mechanical energy Etot at the level of the wheels (see equation 7-17), related to a certain mission. The global assistance factor can be seen as the weighted arithmetic average of the (instantaneous) assistance factor, relative to the total power. In this expression, Ptot(t) has the role of the weight function.

.tm mission totE Eξ= Equation 8-10

The global assistance factor can also be linked to specific driving situations or typical driving cycles. The global assistance factor can in this regard be defined for the three driving situations as used in paragraph 8.3 for analysing the required mechnanical energy. In this context, the total required mechanical energy was divided into three elements: the energy for accelerating a number of times from standstill to a certain speed, the energy for continuous driving and the energy for hill climbing. Three corresponding global assistance factors are defined:

_tm ssss

ss

E

Eξ = Equation 8-11

_tm hdhd

hd

E

Eξ = Equation 8-12

_tm cdcd

cd

E

Eξ = Equation 8-13

In this formulation, Etm_ss, Etm_hd, Etm_cd are respectively the required mechanical energy from the electric motor for start and stop operation, for covering a height difference and for continuous driving.


p.145

In order to investigate the energy flow in the electric propulsion system, the instantaneous power quantities are considered. The instantaneous power quantities at the level of the different components of the propulsion system can be defined as in equation 8-14 till equation 8-16. Successively the mechanical motor power Pmot, the electric power delivered by the motor controller Pcon and the electric power delivered by the battery (or other energy storage system) Pbat are considered. The relationship between these different successive power quantities can be expressed by the use of instantaneous efficiencies quantifying the effectiveness of the power conversion of these different components. In this regard the efficiency of the transmission system ηtm, the efficiency of the electric motor ηmot and the efficiency of the motor controller ηcon can be distinguished.

1( ) . ( )

( )mot tmtm

P t P ttη

= Equation 8-14

1( ) . ( )

( )con motmot

P t P ttη

= Equation 8-15

1( ) . ( )

( )bat concon

P t P ttη

= Equation 8-16

1 1( ) . ( ) . ( )

( ). ( ). ( )bat tm tmcon mot tm pt

P t P t P tt t tη η η η

= = Equation 8-17

The different relations of the power quantities (equation 8-14 till equation 8-16) can be combined to a general expression that gives the relation between the required power at the level of the driving wheel from the electric motor Ptm and the required electric power from the battery pack Pbat. This expression is given in equation 8-17. The separate efficiencies can now be combined by the use of the efficiency of the electric power train ηpt. An overview of these power quantities and efficiencies is schematically represented in figure 8-15.

Figure 8-15: Electric propulsion system efficiencies

Furthermore, the required electrical energy from the battery pack Ebat_req can be defined as the integration of the required electric power from the battery pack over the mission time Tmission. This is shown by equation 8-18.

_ 0( ).

missionT

bat req batE P t dt= Equation 8-18

In analogy with the global assistance factor, a global efficiency of the electrical power train

_pt missionη

can be defined:


p.146

0_

_0

( ). ( ).

( ).

mission

mission

T

pt battmpt mission T

bat req bat

t P t dtE

E P t dt

ηη =

Equation 8-19

This parameter

_pt missionη can be used to determine the required electrical energy from the battery

pack, Ebat_req (equation 8-18) related to a certain mission, from the required mechanical energy allocated to the electric motor, Etm . The global efficiency can be seen as the weighted arithmetic average of the (instantaneous) efficiency, relative to the total power. In this expression, Ptot(t) has the role of the weight function.

__

1.bat req tm

pt mission

E Eη

=

Equation 8-20

The global efficiency can also be linked to specific driving situations or typical driving cycles, in analogy with the global assistance factor. The global efficiency can thus be defined for the three driving situations as used in paragraph 8.3 for analysing the required mechnanical energy. In this context, the total required mechanical energy was divided into three elements: the energy for accelerating a number of times from standstill to a certain speed, the energy for continuous driving and the energy for hill climbing. Three corresponding global efficiencies can be defined:

__

_ _

tm sspt ss

bat req ss

E

Eη = Equation 8-21

__

_ _

tm hdpt hd

bat req hd

E

Eη = Equation 8-22

__

_ _

tm cdpt cd

bat req cd

E

Eη = Equation 8-23

In this formulation, Ebat_req_ss, Ebat_req_hd, Ebat_req_cd are respectively the required electrical energy from the battery pack for start and stop operation, for covering a height difference and for continuous drivng. The substitution of equation 8-10 into equation 8-20 finally gives the relation between the required electrical energy from the battery pack, Ebat_req and the required total mechanical energy Etot from equation 7-17.

__

.missionbat req tot

pt mission

E Eξ

η=

Equation 8-24

This relation points out the influence of the electric power unit technology that is concerned. This was schematically shown in figure 8-14. The required electrical energy from the battery pack Ebat_req can now be calculated starting from the required mechanical energy from the electric motor Etm, by using the global efficiency of the electric

power train _pt missionη , and the global assistance factor

missionξ of the power unit, related to the mission

or driving cycle. The energy content of the battery pack Ebat will have to be equal or greater than the required electrical energy from the battery pack Ebat_req if the mission has to be fulfilled without intermediate recharge or battery exchange. This is expressed by equation 8-25.


p.147

_bat bat reqE E≥ Equation 8-25

This expression is also used later in the calculation model for the battery energy described in chapter 9. The mass of this battery pack is proportional with the energy content of the battery pack and is determined by the specific energy, m

batδ expressed in Ws/kg, of the selected battery technology. Note

that the specific energy values that are available in product data sheets or in literature is the value for the (naked) cells thus excluding the battery electronics (BMS) and excluding the casing. However, the casing can represent an important share of the total mass. This is especially true if the casing needs to be rugged. Depending on the integration of the batteries, the use of batteries for electric bicycles requires a high degree of mechanical protection.

1.bat batm

bat

M Eδ

= Equation 8-26

8.4 Integration of the electric power system in the bicycle

8.4.1 Integration of the electric power unit into the hybrid drive train From a technical point of view, one can divide the vitiating group of electric bicycles into more specific classifications. An important difference is the location of its electric motor. Mainly two possibilities are used: electric bicycles equipped with a hub motor and electric bicycles with a motor located near the bottom bracket. Furthermore, electric bicycles equipped with a hub motor can be distinguished on the basis of the hub motor being mounted in the front wheel or in the rear wheel. In exceptional cases, another location for the motor is used (e.g. behind the saddle, on top of the rear wheel…). These different kinds of integration are represented in figure 5-7 in paragraph 5.4.1 on page 87. Also different solutions for mechanical transmission of the motor power can be identified. A possible point of distinction in this regard is at what level of the hybrid power train the mechanical power of both drive units are combined. Four different power train topologies have been identified and are denoted between brackets with the letters A, B, C or D:

• At the level of the road (type A) • At the level of the driving (rear) wheel (type B) • At the level of the transmission or chain drive (type C) • At the level of the pedal axis (type D)

These four different kinds of integration are discussed and illustrated with pictures of the power units of different commercially available products (see figure 5-8 till figure 5-11 in paragraph 5.4.1 on page 87). Schematic representations of these four levels of mechanical power addition can be found in figure 8-16. The four identified levels of power addition are shown with an orange cross with the corresponding letter (A, B, C or D)


p.148

A

Transmission& Gears

Transmission& Gears

Electrical Energy Storage


Transmission & Gear system


(fro

nt)

Whe

el(f

ront

) W

heel

Bio-mechanical energy storageBio-mechanical energy storage Cyclist musclesCyclist muscles

(rea

r) W

heel

(rea

r) W

heel

Road

Road

AA

Electric power unit

Electric power unit

DD CCBB

A,B

AA

Transmission& Gears

Transmission& Gears





(fro

nt)

Whe

el(f

ront

) W

heel

Bio-mechanical energy storageBio-mechanical energy storage Cyclist musclesCyclist muscles

(rea

r) W

heel

(rea

r) W

heel

Road

Road

AA

Electric power unit

Electric power unit

DD CCBB

A,B

Figure 8-16: Mechanical power addition at different levels of the power train

Type A can be described as a double drive parallel hybrid with two driving wheels. The front wheel is driven by the electric power unit while the rear wheel is driven by the pedals through the chain. For the type A, the transmission from the scheme in figure 8-16 can be disregarded as generally a hub motor in the front wheel is applied. However, a gear system can be involved as some hub motors use an internal gear system as opposite to direct drive motor systems. Type B can be described as a double drive parallel hybrid with a single driving wheel. Type B can also involve a hub motor, installed in the rear wheel. In this case, the same comments can be made as for type A. However, this type also exists with a motor installed outside the driving wheel and using a transmission. An example of an electric bicycle with such type of propulsion system is the Velocity Dolphin [201]. A picture of the Dolphin electric bicycle and a detail of its hybrid drive train is shown in figure 8-17.

Figure 8-17: Velocity Dolphin with dual drive system – type B

Type C can be described as a single drive parallel hybrid. Type C combines the motor power and the biomechanical power at the level of the single transmission. An example of such a system is the Panasonic pedal assist system. A drawing of the mechanical lay-out of this system is shown in figure 8-18 [202]. The single chain transmission is transmitting the power of both the pedals and the motor.


p.149

Figure 8-18: Power adding at the level of the chain (Panasonic motor unit) – type C

Type D can be described as a single drive parallel hybrid like type C. The difference with type C is that the motor power is added to the pedal power at the level of the pedal axis. Many mechanical solutions are used to add both powers at the level of the pedal axis. An example is given in figure 5-8 in paragraph 5.4.1. Another example of this type of configuration is the electric motor of the BikeTec Flyer F series [139].

Figure 8-19: Pedal axis motor from BikeTec – type D, © Philippe Lataire

. The electrical assistance systems of type C and D, with the motor mounted near the pedal axis, are contributing to the propulsion of the bicycle via the single chain transmission and gear system of the bicycle. This arrangement allows employing the motor power with the bike’s gear system in a favourable way to ride uphill or to have an optimal assistance at start. Because the complete traction force of the bicycle is now transferred trough the chain transmission, the wear1 of these parts will be increased. For this type of integration the mass of the drive unit is concentrated in the middle of the bicycle (near the pedal axis). This configuration is positive for the comfort and the stability of the bicycle. In case of type A, when the motor is placed in the front wheel, this allows using a traditional rear wheel traction system (chain, gear system). On slopes or on slippery roads slipping of the motorized front wheel can occur. Putting more weight on the front wheel can reduce this potential problem. If a hub motor is used (both front- or rear-wheel mounted), it should be able to handle the complete range of operation (in terms of speed and torque) without a gear system. The motor should be able to bring the bicycle to its maximum speed within the legal framework. It should deliver sufficient traction force at low speeds for starting or for riding uphill. This is not the case for all the models of electrically assisted

1 Practical experience at the department ETEC pointed out that the chain of this type of electric bicycles needed to be replaced after about 8000km mileage.


p.150

bicycles offered on the market [186, 187, 203]. Consequently, the ones which don’t are not appropriate for demanding or challenging uphill cycling. Finally, a series hybrid configuration for an electric bicycle can be considered. This concept has been described in the literature by Fuchs and Blatter [134, 204]. Mechanical human power is converted into electrical energy by using a pedal driven generator. The output of the electric generator is connected to a converter. This converter is connected to the motor controller of an electric drive unit and to a battery pack (or other energy storage system). This concept is schematically represented in figure 8-20.



Electric motorElectric motor Transmission & Gear system


Bio-mechanical energy storageBio-mechanical energy storage Cyclist musclesCyclist muscles Transmission

& GearsTransmission

& Gears

Dri

ving

whe

elD

rivi

ng w

heel

generatorgenerator

converterconvertermotor controller

motor controller

Road

Road



Electric motorElectric motor Transmission & Gear system


Bio-mechanical energy storageBio-mechanical energy storage Cyclist musclesCyclist muscles Transmission

& GearsTransmission

& Gears

Dri

ving

whe

elD

rivi

ng w

heel

generatorgenerator

converterconvertermotor controller

motor controller

Road

Road

Figure 8-20: Human powered series hybrid configuration

8.4.2 Integration of the energy storage system in the bicycle platform Many different ways of integration of the energy storage system can be found on the currently available electric bicycle models. A first important difference is whether the battery pack is removable or not. The possibility to remove the battery can be important, as it allows recharging the battery separately from the bicycle. The battery could be recharged inside, while the bicycle is parked outside. Battery exchange (battery swapping) becomes a possibility and allows, lengthening the range, if combined for instance with a battery exchange station (see paragraph 8.2.5). The handling of a removable battery pack is also important and is strongly influenced by the weight of the battery pack. The shape and an ergonomic hand grip can further improve the handling. The accessibility of the charging plug is also an important aspect. Preferably the charging plug can be reached without removing the battery pack. This allows recharging the battery both mounted on the bicycle or not. Measures should be taken to protect the charging connector on the battery from rain and dirt. Different locations of the battery packs of electric bicycles can be found. The most common locations are: on top of or under the rear carrier, behind the seat tube, on the down tube of the bicycle frame and inside a side bag. Some models have a more particular location for the placement of the battery. The Sparta ION for instance has a battery pack built inside the curved down tube (see figure 8-21). This electric bicycle has also an optional external battery pack in a rugged side bag that can be attached to the rear carrier that can be used as a range extender. This means that the main battery (inside the down tube) will be used first and when discharged, the external battery pack will be used. In some exceptional models, such as the “Tidal Force M750” the battery pack is built in the wheel (see figure 5-1 on page 85). Another example is the concept from Ez-Wheel which integrates all the functions of a conventional power train in a wheel [205].


p.151

Figure 8-21: Battery inside the frame tube of the Sparta ION electric bicycle

Furthermore, the presence of the battery pack should not hamper the driver while cycling and preferably not restrain the payload capacity of the bicycle. The latter is especially important if the electric bicycle is intended for delivery purposes. The integration of the battery pack should also take the stability of the bicycle while driving and at standstill into account. Asymmetric weight distribution (left-right) should be avoided. In the framework of the NEPH project (see paragraph 6.2), special attention was paid to this aspect. The integration of the battery pack for electric postal delivery bicycles is closely linked to the type of bags or containers used to hold the mail. Many variants of postal bags and containers exist and are used by the different (European) postal operators. The design of the luggage carriers (rear and front) is often adapted to the specific type of bags or containers. Typical kind of side bags that are used by the Belgian post is the “twin bag”. This bag is placed over the rear carrier and has a storage compartment at each side of the bicycle next to the rear wheel. The point of gravity for this kind of bag is lower compared to bags placed on top of the rear carrier. A lower point of gravity generally results in more stability. In the framework of the NEPH project, a flat design of the battery pack was made allowing to mount the battery packs behind the post bags (see figure 8-22). This design resulted in an ergonomic solution without sacrificing payload capacity. A variant of the battery pack with a prismatic shape to be mounted on top of a rear carrier is shown in figure 8-23.

Figure 8-22: SAFT flat battery packs (pilot version) Figure 8-23: SAFT prismatic battery pack (pilot version)


p.152

8.5 Driving range of electric bicycles Forecasting the driving range of an electric vehicle or of an electric bicycle in particular, is not easy. The results strongly depend on the way the bicycle is used. The most important parameters defining the range of the bicycle are the mass, the topography (steepness of the hills), the speed profile and the effort (or power) delivered by the driver (cyclist). The first parameter, the mass M, plays an important role in the amount of energy that is required to complete a certain trip. The higher the mass, the more energy is required for the same trip. Here, the total mass is important: the mass of the bicycle, but also the mass of the cyclist as well as a possible cargo (e.g. postal items). A second parameter is the topography of the route. It is clear that more energy is required to ride up a hill compared to drive the same distance on the flat. Furthermore, the speed profile is also important; in particular the number of starts and stops is essential for the estimation of the required electrical energy. An itinerary which can be cycled at a (more or less) constant speed and without any stop, will require less energy from the electrical traction chain compared to a route with the same length but where the cyclist needs to stop and start quite often (for example in urban traffic). This is even more applicable if the electric bicycle is to be used for the distribution of mail, flyers or folders and needs to stop and start again at every doorstep along the delivery route. The efficiency of the used electric power train (ηpt) also affects the electrical energy consumption at the battery side and thus the driving range of the electric vehicle. Important differences exist in the level of the power assistance provided by the electrical motor of the electric bicycles. This can be quantified with the assistance factor ξ as discussed in paragraph 8.2.1.7. Generally, one can state that the more power the electric motor provides to the traction chain, the larger the energy consumption at the side of the battery. When the cyclist is required to provide a large part of the required effort by himself (low value of assistance factor ξ), the driving range can be significantly increased. How large this effort is, depends on the working principle of the electrical assistance that is used and differs from model to model. In general it is well described in the user manual. Some models of electric bicycles let the cyclist modulate the level of electrical assistance by means of a twist grip for example (power-on-demand system). The driving range of these electric bicycles therefore depends on the way the electric assistance is used by the user. With other models of electric bicycles the degree of the power assistance is predetermined by the associated control system of the bicycle and the power of the electric motor is modulated in function of the cyclist’s effort (pedal power control system). In this way, the share of human power and electrical power can be managed by the motor system. Some manufacturers offer different modes of electrical assistance (e.g. economic, normal, sportive…) that can be selected by the user or pre-programmed by the manufacturers and its dealers. This is done to find an optimal compromise between energy consumption (driving range) and the level of assistance depending on the application. In the framework of comparative test campaigns of electric bicycles for a consumer organisation, the driving range of a set of 12 electric bicycles available on the Belgian and Italian markets in 2008 was investigated [186, 187]. The different electrically assisted bicycles, with fully charged battery pack, were used on the test track until the battery was completely discharged and the electrical assistance of the bicycle shut down. The travelled distance was recorded and the results are shown in figure 8-24. The lowest value of the result of each bicycle corresponds to the driving range obtained by the test person who was using the available assistance at maximum extent, thus minimizing his own effort while driving the different models. The highest setting of the electrical assistance of each model was used for these tests.


p.153

0

10

20

30

40

50

60

Test

sam

ple

1

Test

sam

ple

2

Test

sam

ple

3

Test

sam

ple

4

Test

sam

ple

5

Test

sam

ple

6

Test

sam

ple

7

Test

sam

ple

8

Test

sam

ple

9

Test

sam

ple

10

Test

sam

ple

11

Test

sam

ple

12

Rang

e (k

m)

0

100

200

300

400

500

600

Ener

gy c

apac

ity

(Wh)

Driving range (km)

Energy capacity (Wh)

Figure 8-24: Overview of range and energy capacity for a set of 12 electric bicycles

From this figure we learn that quite large variations of the driving ranges occur. As described before, the driving range is expected to depend on the use and on the user (or cyclist). Some models tend to have smaller variations (about 5 km difference between lowest and highest range). The correlation between the driving range and the energy content of the battery pack is rather poor. Test sample 6 has a very large variation of the driving range (about 22 km difference). These results clearly show that it is very hard to determine and quantify the driving range of a certain model. In total, the different test persons have cycled over 4000km during these test drives. The on-road tests allowed making a general assessment about the driving behaviour during the different test samples. The importance of these different criteria depends on the type of use of the bicycle. When the bicycle is used in an urban environment with a large number of start and stops (crossroads, traffic lights...), the assistance at start becomes an important criterion. For postal application the behaviour of the assistance at start is also important as the mission is characterised with a large number of stops. If the bicycle is used in a hilly environment the behaviour of the assistance when driving uphill, is very important for the users' comfort and appreciation. Based on the findings made during the test trips, a subjective assessment of the electrical assistance was given in three different situations: assistance at start, assistance on a flat road and assistance when driving uphill. An overview of the subjective assessment of the electrical assistance can be found in table 8-1.


p.154

Table 8-1: Overview of the subjective assessment of the electrical assistance

Maximal speed Assistance at start Assistance on a flat road

Assistance on slopes

Test sample 1 25 km/h none good good Test sample 2 25 km/h very good very good average Test sample 3 25 km/h average average average Test sample 4 25 km/h none good good Test sample 5 28 km/h very good very good very good Test sample 6 20 km/h none average insufficient Test sample 7 20 km/h none good insufficient Test sample 8 28 km/h none good average Test sample 9 22 km/h none very good good Test sample 10 28 km/h none very good very good Test sample 11 30 km/h none good average Test sample 12 25 km/h very good very good very good In the second column of the table, the maximal speed at which the electrical assistance provides traction is mentioned. Four of these 12 test samples provide assistance above a vehicle speed of 25 km/h. This causes the vehicle to be considered as a motorised vehicle according to the definitions of the European directive 2002/24/EC (see paragraph 5.4.3). Furthermore, a lot of the models from the 12 test samples offer no assistance at start. The assistance on a flat road is good to very good for most models. However, the appreciation of the assistance on slopes is quite different for the different tested models electric bicycles. Test sample 5 scored “very good” for the appreciation of the electrical assistance for all three considered conditions. Test sample 6 seems to perform worst for the different criteria. It is not always easy to determine the exact available electrical energy of the battery pack installed on the bicycle. However, most of the times, the nominal voltage and the rated capacity of the battery pack are mentioned in the user manual and/or on the battery pack. Often the current capacity of the pack is mentioned and expressed in Ampere-hours, but often without outlining at which discharge period this value is true. The frequently used and standard discharge periods of 5h or of 20h can be assumed as a consequence. However, the battery pack of an electric bicycle is discharged much faster in normal usage (typically about 1 to 2 hours). Moreover, a battery will deliver less energy (Ampere-hours) if more current is drawn than the rated value (Peukert’s law). Also the battery temperature is affecting the useful energy capacity of the battery pack. Both temperature and discharge current will vary during the normal use of the electric bicycle. For example, a battery pack with a rated capacity of 10Ah (no further details available) and with a nominal voltage of 24V, results in an expected available energy content of 240Wh. In general the battery pack of an electric bicycle is used until 50 to close to 100 percent of its useful capacity (high DoD or deep cycles). The depth of discharge also has an influence on the cycle life of the battery. To improve the lifetime performance of the battery pack, the maximal depth of discharge could be limited (by the manufacturer). Furthermore, for some battery types the useful capacity may increase the first few tens of cycles compared to the initial capacity. Taking the latter into account and due to Peukert’s law, the actual useful capacity can deviate significantly from the expected value.

8.6 The SAFT – Heinzmann electric power system The SAFT – Heinzmann electric power system is the result of the collaboration of two European manufacturers: SAFT s.a., a French battery manufacturer [178] and Heinzmann GmbH., a German motor and motor drive manufacturer [206]. The NEPH project (see paragraph 6.2 on page 102) was an important substantiation of this collaboration through the development of a dedicated electric power system for use in postal delivery bicycles and tricycles. The department ETEC of the Vrije Universiteit Brussel played an important role in this development, in particular through its role of work package leader of the specification phase of the NEPH project. A schematic representation of the NEPH power system is shown in figure 8-25. From this figure we see that the power system is composed of the


p.155

battery packs that function as the on-board energy storage system. Multiple battery packs can be involved (see paragraph 8.6.2 on page 156). Further, the electric power unit is composed of a motor controller unit and a wheel motor, installed in the front wheel. This system can be identified as type A, according to the definition of paragraph 8.4.1 (see page 147). The electric (postal) bicycle equipped with this power system, can be described as a double drive parallel hybrid with two driving wheels. The front wheel is driven by the electric power unit while the rear wheel is driven by the pedals through the chain. A gear system is involved because the hub motor used has an internal gear system. In the following paragraphs the technologies used for the developments in the framework of the NEPH project are described.

Figure 8-25: Saft – Heinzmann electric power system

8.6.1 The Heinzmann power unit Heinzmann produces a various hub motors with integrated two-stage reducing gears for use in light electric vehicles. In particular they make wheel hub motors for electric bicycles or other light electric vehicle applications. Further, they produce the motor controllers for use with these electric motors. The motor concerned is a mechanically commutated permanent magnet excited d.c. motor with an axial field flux design [188]. An exploded view of this motor is shown in figure 8-26 [207].

Figure 8-26: Heinzmann hub motor - exploded view (translated from [207])

The magnetic system of the motor consists of two separate symmetrically arranged systems using high-energy magnets (type neodymium iron boron, abbreviated as NdFeB). A drum-type collector with


p.156

carbon brushes is used to conduct the power to the disc shaped iron-free armature. The motor has a nominal voltage level of 36 Volts. The continuous rated power of the motor has been set to 250 Watt. The motor can however, for short instances deliver more power. The maximal torque and current of the motor are limited by the maximal admissible torque of the gear system used. The motor used for the NEPH project has a maximal output torque of 60Nm and a maximal armature current of 33 Amperes. The wheel motor can be built in a spoked wheel of different sizes. The electric motor is controlled by a step down converter. The controller unit uses a pedal sensor to detect the forward turning of the pedals. This sensor is shown in figure 8-8 on page 139). If the pedals are not turning, the motor speed is limited, corresponding to a vehicle speed of about 6km/h (walking speed). The full range of the twist grip position is allocated to the reduced speed range (0-6km/h), in case the pedals are not turning. This functionality is also denoted as ‘start-up assistance mode’. This allows activating the electric assistance when walking aside the bicycle. As soon as the pedals are turning, the maximal motor speed is available again. A twist grip is used to regulate the desired speed of the motor. The Heinzmann power unit is thus a power-on-demand system. A picture of the twist grip from Heinzmann can be found in figure 8-27. The global efficiency

ptη , as defined in paragraph 8.3, of the Heinzmann power unit is estimated by the

manufacturer to: 60% in case of continuous driving, 40% in case of for start & stop operation and 60% à 70% for in case of hill climbing [208].

Figure 8-27: Twist grip Heinzmann power unit

8.6.2 SAFT Modular energy storage system Besides the Heinzmann electric power unit, the electric power system for the NEPH vehicles includes a SAFT battery pack. Two different nickel-metal-hydride battery packs of 36 Volts are available and where used to offer a modular range of energy storage systems to power the range of NEPH vehicles. The 36 Volts battery pack is available with two different cell sizes (smaller size D and larger size F). SAFT has meanwhile developed also a Lithium-ion battery pack of 36 Volts (based on SAFT’s Lithium-ion MP cells) that can be used for the NEPH power system. The main characteristics of these different battery packs can be found in table 8-2. A picture of the battery pack without the casing, with the cells stacked in a prismatic shape, can be seen in figure 8-28. A picture of the NEPH flat and the prismatic battery pack can be seen in figure 8-22 and figure 8-23 respectively (see page 151). The number of battery packs that are required to perform a specific postal delivery round can vary from 1 small pack to up to 3 larger packs. See also paragraph 9.8 about the battery energy model results starting on page 172.


p.157

Table 8-2: SAFT Mobility batteries characteristics

Mass Dimensions (lxbxh) Capacity Voltage Pilot Prismatic Ni-MH (D) 8,5 kg 300x199x95 mm 9,0 Ah 36 V Ni-MH (F) 11,6 kg 300x199x95 mm 14,5 Ah 36 V Pilot Flat Ni-MH (D) 9,2 kg 405x375x45 mm 9,0 Ah 36 V Ni-MH (F) 12,7 kg 405x375x45 mm 14,5 Ah 36 V Industrial Prismatic Ni-MH (D) 6,3 kg 262x185x80 mm 9,0 Ah 36 V Ni-MH (F) 9,1 kg 262x185x80 mm 14,5 Ah 36 V Li-ion (MP) 4,5 kg1 225x185x80 mm 13,0 Ah 36 V Industrial Flat Ni-MH (D) 6,8 kg1 405x375x45 mm 9,0 Ah 36 V Ni-MH (F) 9,7 kg1 405x375x45 mm 14,5 Ah 36 V Li-ion (MP) 5,1 kg1 405x375x45 mm 13,0 Ah 36 V

Figure 8-28: SAFT Smart-VH-battery module (NiMH) [178]

8.6.3 NEPH battery discharge cycle and ageing margin The use of the battery pack for the NEPH power system requires a dynamic discharge current profile. The maximal discharge current of the application is higher than the rated value and was limited to 33 Amperes2. The discharge current leads to increased heat dissipation during discharge compared to a continuous discharge at nominal rate. Further, the current rate also influences the useful energy at each discharge cycle. Two different tests were performed at the department ETEC on a SAFT Ni-MH (D) module with pulsed discharge currents of different magnitude. The discharge current was pulsed at 25 Amperes for the first test and at 30 Amperes for the second test. The pulses had a time duration of 3 seconds and were followed by a rest period of 10 seconds. The discharged energy and the number of cycles were recorded until the battery BMS shut off the battery from the load. The results of this test are shown in figure 8-29. From this figure the influence of the discharge current rate is clearly visible. The cycle life of the battery pack is limited by the calendar life and is defined at a certain operation temperature (i.e. 20°C). The cycle life further depends on various parameters such as charge rate, discharge rate, depth of discharge (DoD), period of rest between charge and discharge, storage conditions etc. Deviating use from the standard rated conditions of operation will affect the cycle life. In particular, a lower DoD will increase the expected cycle life. A higher battery temperature will lower the expected cycle life. For warranty purposes, several of these parameters are measured and logged by the battery management system of the battery pack. These logged data can be used to estimate the effective life time of the battery and to compare it with the warranty claims.

1 Estimation of the weight, no prototype yet available 2 Discharge duration at this rate is limited to 90 seconds for thermal reasons


p.158

Energy Discharged [Wh]/Number of cycles

436

619

371

445

0

100

200

300

400

500

600

700

0 2000 4000 6000 8000 10000 12000 14000Time [s]

discharged energy (pulsed @25A)number of cycles (pulsed @25A)discharged energy (pulsed @30A)number of cycles (pulsed @30A)

Figure 8-29: energy discharged for pulsed discharge current at different rates

An important parameter in this context is the DoD. The useful capacity of the battery pack will fade during its service life. A degradation of 20% of the initial capacity will occur by the end of the battery’s service life. For the NEPH project, the fading of the battery’s capacity was considered unacceptable for postal operations. To ensure a more continuous driving range performance, a battery ageing margin (BAM) of 30% of the initial capacity was agreed on by the NEPH consortium and is applied to calculate the SoC of the battery. This means that after 70% discharge of the initial full capacity, the HMI will indicate 0% SoC at this point, throughout the complete life of the battery. This principle is schematically represented in the figure 8-30. Depending on the number of discharge cycles made (or its current calendar lifetime) a certain amount of energy will still be available in the battery pack (see the red or orange dotted line). The battery pack will however not be disconnected from the application when it has reached the 30% margin zone but the information given on the human machine interface will indicate a “complete discharged battery”. The instructions to the user corresponding with this indication are to connect the battery as soon as possible to the charger and to recharge the battery. When the battery is only used outside this 30% margin zone (above the bold horizontal red line), the cycle life is expected to be extended (see the green discharge lines). The latter can be explained by the fact that the cycle life is inversely proportional to the DoD. An important advantage of this principle is that the vehicle performance will remain constant throughout the life time of the battery pack. In other words, the electric bicycle will be operable on the same postal delivery cycle during the complete cycle life of the battery.


p.159

Figure 8-30: NEPH batteries capacity margin

8.6.4 Parallel operation of the batteries The analysis of the required battery capacity for typical postal requirements has pointed out an important quantity of battery capacity is needed. The required capacity exceeds the capacity of the individual batteries from table 8-2 in case of postal delivery application. So, more than one battery pack is required. Additionally, it was found that most postal operators don’t want to have to recharge during the postal delivery mission. Battery swapping or exchange would be a possibility, however most postal operators want to avoid the use of relay boxes. Therefore, the complete required battery capacity has to be charged onto the bicycle. If two or more battery packs are mounted on the bicycle, this offers the possibility to connect both battery packs to the (Heinzmann) electric power unit. Inside the battery packs a MOSFET switch is used to enable or disable the discharging.

Figure 8-31: parallel connection of the battery packs

When different battery packs are connected in parallel, these switches allow disconnecting or connecting each battery pack individually. This feature is shown in figure 8-31. Communication between the battery packs and the motor controller has been elaborated and allows choosing between different discharge-modes. The communication protocol used is not standardised but has been developed specifically for the SAFT-Heinzmann power system. The user can choose between a parallel discharge mode and a sequential discharge mode. In the case of the parallel discharge mode, all switches (“discharge enable” on figure 8-31) are closed. Alternatively the user can select the sequential discharge


p.160

mode in which one battery pack is discharged after the other. These different discharge modes are graphically represented in figure 8-32. The sequential discharging has the advantage that different battery technologies (e.g. nickel-metal-hydride and Lithium-ion) can be mixed. Also batteries of different ages (an old battery combined with a new battery) can be combined in this application.

Figure 8-32: NEPH batteries parallel or sequential discharging

The parallel discharge mode is set as the standard discharge mode of the NEPH power system. The parallel discharge mode has the advantage that the power required by the Heinzmann power unit (Pbat) is divided between the two (or more) battery packs. Now, the heat dissipation due to the discharge current of each battery pack will be lower, compared to the situation where the total discharge current has to be delivered by one single pack. The reduced heat dissipation has an important positive influence on the cycle life of the battery pack. Measurements from SAFT have pointed out an increase of about 50% of the life time compared to one single pack in case of pulsed discharging1.

8.6.5 NEPH power system and NEPH prototypes The SAFT-Heinzmann electrical power system is composed of the following components:

• hub motor • motor controller unit • pedal sensor • twist grip • battery pack(s)

1 Discharge pulses of 33 A with duty cycle of about 25%


p.161

Figure 8-33: Schematic representation of the NEPH power system

The controller unit is communicating with the battery packs to allow an optimal, reliable and safe operation of the power system. Furthermore, the controller receives and sends information from and to the human-machine-interfaces (twist grip and pedal sensor) and receives a temperature signal from the motor. The SOC of the battery pack is calculated by the BMS of the battery pack and is communicated to the Heinzmann controller unit. The latter transfers this information to the LEDs on the twist grip (HMI). These different components and their integration on a bicycle frame is shown schematically in figure 8-33. Several platforms that are integrating the NEPH electric power system components have already been developed and evaluated. Some of these platforms can be seen in the figure 8-34 and the figure 8-35.

Figure 8-34: NEPH bicycle prototype LUDO

The prototype from figure 8-34 has been developed by Ludo n.v. for the call for tenders of the Royal Mail Group in 2007. The NEPH prototype electric tricycle for postal delivery has been developed by MIFA AG and can be seen in figure 8-35. Further, it can be mentioned that the department ETEC was strongly involved in the development of all these prototypes of the NEPH project. Close collaboration between the department and both platform builders allowed that the first prototypes were tested intensively before putting them at the disposal of the postal organisations. These internal tests resulted in useful feedback for the platform builders and modifications or improvements of the prototypes could be made. These improvements were obtained at multiple levels. Suggestions for the improvement of the bicycle’s ergonomics and an optimal integration of the components of the power system were made. The stability of the bicycle at both standstill and when riding was investigated and particular attention was paid to the minimisation of shimmy.


p.162

Figure 8-35: NEPH tricycle prototype from MIFA A.G.

. In the next chapter of this dissertation, a calculation tool that can be used to size the battery pack for the NEPH power system as described in this paragraph will be discussed. Starting from the postal requirements the corresponding required mechanical energy can be calculated. This mechanical power will now be further used to determine the required electrical energy from the on board storage system of the power system.


p.163

9 Model to estimate the required energy

capacity of the battery

9.1 Introduction This chapter describes a model that has been used to estimate the required energy of the battery, Ebat_req. This model will be refered to as the “battery energy model” in the remaining of this text. The purpose of the battery energy model is to have a tool that can be used to estimate the required energy of the battery pack used on an electric postal bicycle. The most important feature of the battery energy model is to use the postal mission parameters, as described in chapter 6, as the input for this model. In chapter 7 (see page 113), a model to calculate the required mechanical power and energy has been elaborated. In particular, an alternative methodology to calculate the total required mechanical energy, Etot, starting from basic postal requirements is given in paragraph 7.3 (see page 119). In chapter 8, an overview of contemporary drive technology used for electric bicycles was given. This chapter revealed the relation between the total required mechanical energy Etot from chapter 7 and the required battery energy Ebat_req from chapter 8 (see equation 8-24). In particular the SAFT-Heinzmann electric power system was described in paragraph 8.6 (see page 154). The battery energy model that is described in this chapter has been elaborated for this NEPH electric power system.

9.2 Features of the battery energy model The purpose of the battery energy model is to have a calculation tool that determines the required energy for the battery pack, starting from the postal requirements. The battery energy model to calculate the required battery energy can be represented with a flow chart diagram as in figure 9-1. The model can be represented by a black box. A set of postal requirements is provided to the battery energy model. The model provides the user with the calculation results, in particular the required battery energy. Besides the postal requirements, the model also uses a set of physical properties of the vehicle and some drive technology parameters.

Model to estimate the required energy capacity of the battery

p.164

Figure 9-1: Battery energy calculation

. Besides the required battery energy, the following other results are an output of the battery energy model:

• The required battery capacity (for a certain value of the battery’s nominal voltage) • The weight of the battery pack (for a given set of battery pack combinations) • The battery pack capacity (for a given set of battery pack combinations)

9.3 Programme language As the programming language for the battery energy calculation tool, National Instruments LabView™ has been used. LabView™ is a high-level programming tool that uses a graphical programming language. One of the main advantages of this programme language is that it provides a better overview of the programme structure, compared to a text based programming language (e.g. C++). The most common uses of LabView™ are data acquisition and control application. Nevertheless other applications such as simulations and calculations are also possible.

9.4 Main user interface Figure 9-2 shows the main user interface, or front panel of the battery energy calculation tool. A short description of how to use the energy calculation tool is given below.


p.165

Figure 9-2: The front panel of the energy calculation tool

Before running the calculation tool, the user can select the different parameters for input and output of the analysis. The calculation tool allows to vary two input parameters (parameter 1 and parameter 2) between a begin value and an end value. Further, the user should specify a value for the incrementation (step value input parameter 1 and 2) to vary from the begin value till the end value. The use of this incrementation value is explained in detail in paragraph 9.6. The following 7 input parameters can be selected as variables for the analysis:

• Total trip length Dmission (km) • Number of stops (#) • Cumulative height difference hcumul (m) • Payload Mp (kg) • Driver mass Md (kg) • Average speed v (km/h) • Speed attained between 2 stops vs (km/h)

The user can also choose between different output parameters:

• Required Battery capacity (Ah) • Required Battery energy (Wh) • Battery pack capacity (Ah) • Battery pack weight Mbat (kg)

Once the different parameters are choosen and set, one can run the calculation tool. Depending on the computer’s processor speed and on the choosen step sizes, the calculation tool will compute the results within a few seconds at most. The user can see the results on three different tabs of the front panel. A first tab is called “3D Plot” and is shown in figure 9-3. On this screen, the user can read the selected output parameter (e.g. the battery pack capacity) as a function of both selected parameters 1 and 2 (e.g. the number of start and stops and the total trip length). This tab contains both an intensity graph and a 3-dimensional graph. The intensity graph displays 3-dimensional data on a 2-dimensional plot by using color to display the values of the third dimension. A color scale is shown next to the graph. The intensity graph also has 2 cursors. Each


p.166

cursor displays the x, y, and z values for a specified point on the graph. The 3-dimensional graph shows the same results but with a 3-dimensional perspective.

Figure 9-3: Output screen of the energy calculation tool (3D plot)

On the other 2 tabs, a slice of the results is shown with an XY Graph. The slices can be made at different values for both parameters. In case of “2D Plot parameter 1”, the user can specify the value of parameter 2 at wich the slice should be made. In figure 9-12 on page 175 the tab “2D Plot parameter 1” is shown.

9.5 Calculation methodology - Iteration algorithm As was discussed in paragraph 8.3, the total weight of the vehicle is unknown, as the weight of the battery pack that is required is unknown at the start of the calculation. A set of battery pack combinations can be defined in the model and are considered for the calculations. The different battery pack combinations should be ordered with increasing battery energy. The calculation tool uses an iteration algorithm to determine the required battery energy. A flowchart of the used calculation methodology is shown in figure 9-4. At start, the battery energy model receives input through the user interface with values for the postal mission requirements (green block). Also the vehicle’s physical properties (rolling resistance coefficient, aerodynamic drag coefficient…) and the drive technology parameters (motor unit weight, efficiency of the electric power train…) have been initialised within the calculation tool (two upper purple blocks). If desired, the user can also change these different properties and parameters. However, the postal mission requirements (green block) are considered as the main input of the model. If the energy of the selected battery combination Ebat is not larger than or equal to the required battery energy Ebat_req, the calculation tool will select the next battery pack combination and recalculates the block “battery energy model”. The calculation iteration is stopped when the selected battery energy is larger or equal to the required battery energy (see the dicision block in the flow chart). This conditional iteration is the implementation of the expression from equation 9-5 (see page further in this paragraph on page 169). As a final step, the calculation results are exported to a spreadsheet file (tab delimited) as well as to different plots on the user interface.


p.167

Figure 9-4: Iterative battery energy calculation

9.5.1 Battery pack combinations As mentioned earlier, the calculation tool requires a set of possible battery pack combinations for the on-board energy storage. The modular approach of the NEPH energy storage system has been described in 8.6.2. The main characteristics of the SAFT Mobility batteries are shown in table 8-2. The 36 Volts battery pack is available with two different cell sizes (smaller size D and larger size F). All possible combinations of these two variants (VH-D module and VH-F module) are considered. 1, 2 or 3 battery packs are considered as possible configurations. When considering the “flat” shaped battery packs from table 8-2 two different implementations can be considered: the prototype version with a tooled aluminium casing or an industrial version with an extruded version of the casing. The latter version has a lower weight. The total mass of the different battery pack configurations (in kg), the corresponding energy capacity (in Ah) and the energy density (in Wh/kg) are shown in figure 9-5. The shift in energy density and the shift in weight due to the use of an industrial version of the casing, compared to the prototype version is indicated in this figure with a black arrow (in case of 1 VH-D module).


p.168

0

5

10

15

20

25

30

35

40

45

50

55

60

1 Pack 1 Pack 2 Packs 2 Packs 2 Packs 3 Packs 3 Packs 3 Packs 3 Packs

1 VH-D module 0 VH-D module 2 VH-D module 1 VH-D module 0 VH-D module 3 VH-D module 2 VH-D module 1 VH-D module 0 VH-D module

0 VH-F module 1 VH-F module 0 VH-F module 1 VH-F module 2 VH-F module 0 VH-F module 1 VH-F module 2 VH-F module 3 VH-F module

Total Capacity (Ah)

Total Mass Industrial (kg)

Total Mass Pilot (kg)

Energy Density Industrial (Wh/kg)

Energy Density Pilot (Wh/kg)

Figure 9-5: SAFT Flat battery pack combinations (Pilot and Industrial versions)

The different possible configurations of battery packs lead to a stepwise relation between the battery pack weight and the required energy. This stepwise function is shown in figure 9-6, in case of the prototype flat battery packs. From this figure it can be seen that the combination of three VH-D modules makes no sense compared to the combination of two VH-F modules. This combination will therefore not be considered futher in the analyses.

0

5

10

15

20

25

30

35

40

45

0 200 400 600 800 1000 1200 1400 1600 1800

Required Energy (Wh)

Bat

tery

pac

k w

eig

ht

(kg

)

0xD - 1xF

2xD - 0xF

1xD - 1xF

2xD - 1xF

1xD - 0xF

0xD - 3xF

1xD - 2xF

0xD - 2xF

3xD - 0xF

Figure 9-6: Stepwize battery weight function for the NEPH Pilot Flat batteries

9.5.2 Required battery energy calculation The calculation of the required energy capacity of the battery pack starting from the postal mission requirements is based on the required mechanical energy calculation as described in paragraph 7.3 (see page 119).


p.169

The total required mechanical energy Etot is calculated as the sum of the following three elements:

tot ss cd hdE E E E= + +

Equation 9-1

Where:

• Ess is the required mechanical energy for start and stop operation (see equation 7-25) • Ehd is the required mechanical energy related to the cumulative height difference (see

equation 7-26) • Ecd is the required mechanical energy for continuous driving (see equation 7-27)

The relation between the total required mechanical energy Etot and the mechanical energy delivered by the electric power unit Etm is quantified by using a global assistance factor related to a certain

missionmissionξ , as discussed in paragraphs 7.3 and 7.3.4.

For the battery energy model also, a global assistance factor is used. Three different global assistance

factors are implemented ξ and allocated to the three different elements of equation 9-1. This results in

the following equation:

. . .tm ss ss hd hd cd cdE E E Eξ ξ ξ= + + Equation 9-2

The relation between the the mechanical energy delivered by the electric power unit Etm and the required battery energy Ebat_req is quantified by introducing a global efficiency related to a certain mission

_pt missionη , as discussed in paragraphs 7.3 and 7.3.4.

The battery energy model uses global efficiencies of the power trainptη , linked to the different

elements of equation 9-1 to calculate the required battery energy Ebat_req:

__ _ _

. . .ss hd cdbat req ss hd cd

pt ss pt hd pt cd

E E E Eξ ξ ξ

η η η= + +

Equation 9-3

Finally, the battery ageing margin symbolised as BAM (in percent), as decribed in 8.6.3 (see page 157) is also implemented in the battery energy model:

_*_ (100% )

bat reqbat req

EE

BAM=

− Equation 9-4

Where:

• BAM is the battery ageing margin, expressed in percent • Ebat_req is the required battery energy expressed in Joules • Ebat_req* is the required battery energy taking into account the battery ageing margin,

expressed in Joules This required battery energy considering the BAM, Ebat_req* is finally compared to the different battery energy Ebat values from the battery pack combinations:

*_bat bat reqE E≥ Equation 9-5

This relation is analogous to equation 8-25 and is used in the iteration algorithm shown in figure 9-4.


p.170

9.6 Programming structure The structure of the programming of the battery energy calculation tool is based on two nested for-loops. The inner for-loop calculates the battery energy to completion of incrementing parameter 2 starting at a specified begin value until a specified end value with a specified step or increment value. The outer for-loop executes to completion of incrementing parameter 1. The selection of parameter 1 and parameter 2 can be made by the user before execution. The begin value, the end value and step (increment) value for each parameter can be introduced in the main user interface (see also paragraph 9.4). As an illustration, an example will be given, corresponding to the settings shown in figure 9-2. The program will start the calculation with both parameter 1 (e.g. total trip length) and parameter 2 (e.g. number of stops) equal to their respecitive begin values (1 km and 1 stop). After the execution of the first battery energy calculation, parameter 2 (number of stops) will be increased with the step value (e.g. 1 stop) and if still lower than or equal to the specified end value of parameter 2 (e.g. 600 stops), the program will calculate again using the new values (1km and 2 stops). The latter procedure will be repeated unit the parameter 2 (number of stops) has reached a value greater than the end value (i.e. 601). Now, the parameter 1 will be increased with the specified step value for parameter 1 (e.g. 0,1 km). If parameter 1 is still lower than or equal to the specified end value of parameter 1 (e.g. 35,0 km), the program will continue with the inner loop of parameter 2. If both parameters have reached a value greater than their end values, the calculation results are exported to the display and to a file. The results corresponding to this example can be found in figure 9-10.

Figure 9-7: Nested loops for parameter calculation

9.7 Battery energy model validation The battery energy model that has been developed was validated by measurements performed with a prototype of an electric postal delivery bicycle. The bicycle platform was constructed by MIFA A.G., partner in the NEPH project. This postal delivery bicycle was equipped with the NEPH power system (see paragraph 8.6 for more details).


p.171

Specific tests were organised and performed to measure the energy consumption of the start and go operation, which is typical for postal usage of the bicycle. Both battery current and voltage were measured while the bicycle was used to accelerate and stop 4 times successively. This test was repeated several times and different payloads were applied. The driver was asked to accelerate up to a speed of about 14 km/h. The results of the energy consumption, calculated from the measured current and voltage at the battery side, were compared to the output of the battery calculation tool. The results of these tests are shown in figure 9-8. The pink line in figure 9-8 corresponds to the calculated energy consumption, with the energy calculation tool, for a situation of 4 stops, with a speed of 14 km/h between 2 stops. The same was done for a speed of 14,5 km/h (red line) and for 12,7 km/h (orange line). The important influence of the speed reached between 2 stops is visible from this exercise. As can be seen in figure 9-8, the calculation corresponds quite well with the measured values. However, for lower values of the payload, the deviation is somewhat bigger. A possible reason is the variation of the motor efficiency between the different points of operation. In case of worst case calculations for sizing of the battery pack, the parameter value of the speed achieved between 2 stops is best slightly overestimated.

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

payload (kg)

Ener

gy (W

s)

averagecalculated 14km/hcalculated 14,5km/hcalculated 12,7km/hLi ( l l t d 14k /h)

Measured capacity for 4 starts/stopsFlat road, acceleration to about 14 km/hDistance: about 80 mDriver: 70 kgBattery: 9 kg (SAFT 36V VHF prototype)Measuring devices: about 4 kgMIFA protoype with heinzmann motor: 45 kg

Figure 9-8: Comparison of calculation tool with measurements for start and stop operation

A different mode of operation was analysed and corresponds to driving on a flat road at a continuous speed. To exclude wind and to assure a completely flat surface (excluding height differences), energy measurements were done in a gym court. The prototype was loaded with different weights of payload and a number of rounds were driven around an indoor football court. Again the obtained results were compared with the NEPH design tool. The results (expressed Wh/km), in case of a payload of 75 kg and 40 kg and can be seen in figure 9-9. A good correlation can be observed from this comparison. For lower speeds, the deviation becomes somewhat larger. This could be explained the rolling resistance being dependant of the vehicle speed.


p.172

0

2

4

6

8

10

12

14

16

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

average speed v (km/h)

Con

sum

ptio

n (W

h/km

)

calculation tool 75kg

measurements 75kg

calculation tool 40kg

measurements 40kg

Figure 9-9: Comparison of calculation tool with measurements for continuous driving operation

9.8 Battery energy model results The battery energy model has been used in the framework of the NEPH project to analyse the required battery capacity for the requirements from several European postal operators. Further, the battery energy model has also been used for analyses of the specifications of public tenders. The requirements used in these specifications for calls for tenders were however confidential. In this paragraph the results of the battery energy model that has been developed will be illustrated with a fictive, however realistic, example. Starting from a set of postal requirements, a potential supplier of electric postal delivery bicycles wants to know how much electrical energy storage is required. Let’s consider the following situation: an electric delivery bicycle equipped with the NEPH power system has the following characteristics

• the bicycle platform mass: Mplat = 33kg • diameter frontwheel (wheelmotor installed): dw = 24 inch or 61 cm • the motor power unit has a mass of Mmot = 4kg

The following physical properties are considered:

• aerodynamic drag coefficient and frontal surface: S.CX = 0,53m² • rolling resistance coefficient: CR = 0,00854 • air density: ρair = 1,225 kg/m³

The following postal mission requirements are considered for this analysis:

• the payload mass: Mp = 60kg • total trip length: Dmission = 30 km • average speed: v = 15 km/h • total number of stops = 300 • covered height difference: hcumul = 100 m • speed reached between two stops: vs = 10 km/h • the postman driving the electric delivery bicycle has a mass of: Md = 80kg


p.173

For this example, a global assistance factor ξ = 100% is used for all three elements: continuous driving,

for start & stop operation and for hill climbing. This corresponds to a situation where the driver maximally uses the electric motor system. The driver pedals (to keep the motor system active) but without really pushing on the pedals. This situation can be considered as a worst-case calculation for the estimation of the required battery energy Ebat_req. The global efficiency

ptη of the power unit is set to: 60% for continuous driving, 40% for start & stop

operation and 60% for hill climbing, as explained in paragraph 8.6.1 on page 155.

The NEPH flat pilot battery modules with a nominal voltage of 36V are considered. The masses Mbat of the different combinations are corresponding to those given in figure 9-6 and are summarized in table 9-1. A BAM of 70% is considered for the calculations (corresponding to the NEPH power system specifications described in paragraph 8.6.3)

Table 9-1: Battery pack combinations SAFT Flat Pilot (36 Volts)

Combination code

# Ni-MH (D) # Ni-MH (F) Combined mass

Combined capacity

Battery Energy Ebat

A 1 0 9,2 kg 9,0 Ah 324 Wh B 0 1 12,7 kg 14,5 Ah 522 Wh C 2 0 18,4 kg 18,0 Ah 648 Wh D 1 1 21,9 kg 23,5 Ah 846 Wh E 0 2 25,4 kg 29 Ah 1044 Wh F 2 1 31,1 kg 32,5 Ah 1170 Wh G 1 2 34,6 kg 38 Ah 1368 Wh H 0 3 38,1 kg 43,5 Ah 1566 Wh

For the analysis, different postal parameters were selected. In a first plot, the total trip distance and the number of stops where taken as parameter 1 and parameter 2 in the battery energy model. As the output parameter, the required battery capacity, expressed in Ah is taken. A 3-dimentional plot is used in figure 9-10 to represent the results of the analysis. When looking in detail to this plot, one can recognise small non linearities. These small steps are due to the stepwise relation between the battery pack mass and the required battery capacity (see figure 9-6). The number of stops is varying between 1 and 600 stops and the total trip distance is varying from 1 to 35 kilometer. The calculated required battery capacity varies from 3,8Ah and 31,8Ah. However, both parameters are not realistic or relevant over the complete range for the envisaged application (i.e. postal delivery). Especially, the upper and lower values of the range are rather improbable but are shown in this analysis to reflect the evolution of the output parameter in function of the considered input paramters.


p.174

Figure 9-10: Required battery capacity as a function of the number of stops and of the total trip distance

In a second plot, the total trip distance and the number of stops were selected as parameter 1 and parameter 2 in the battery energy model. As the output parameter, the battery pack capacity, expressed in Ah is taken. A 3-dimensional plot is used in figure 9-11 to represent the results of the analysis. The areas corresponding to the different battery pack combinations are indicated using the codes given in table 9-1. The number of stops is varying between 1 and 600 and the total trip distance is varying from 1 to 35 kilometer. For the choosen ranges of the input parameters, the calculated battery pack capacity ranges from 9Ah (combination A) to 32,5Ah (combination F). In between, the other battery pack combinations (B, C, D and E) are corresponding to areas of operation that are delimited by straight parallel lines. These parallel lines give rise to the staircase appearance of the 3-dimensional plot. The successive boundery lines correspond with the maximal performances of the matching battery pack combination in terms of parameter 1 and parameter 2. In other words, by considering the boundary line of a certain battery pack combination (e.g. the orange boundery line in the figure corresponding to combination C) and considering a certain value for parameter 2 (e.g. 300 stops) one can determine the maximal possible value for parameter 1 (about 20 km for this example). To increase the readability of the results, one can use contour plots or make slices of the 3-dimensional plot.

One of the small steps


p.175

Figure 9-11: Battery pack capacity as a function of the number of stops and of the total trip distance

The user interface that was developed, allows making slices of the 3-dimensional plot. As an illustration, a slice has been made at the value of parameter 2 equals to 300 stops. The corresponding 2-dimensional plot is shown in figure 9-12. A moveable cursor (see red cross) on this graph allows reading the boundary value of the output parameter (i.e. the battery pack capacity). The example shown in figure 9-12 corresponds to the one shown in figure 9-11.

Figure 9-12: Output screen of the energy calculation tool (2-D plot of parameter 1)

Boundery line for combination C


p.176

Depending on the distance of a certain point of operation (i.e. a combination of values for parameter 1 and parameter 2) to the boundary line of the battery pack combination that is used, the DoD of the battery will differ. The DoD of the battery, related to the full initial battery capacity Ebat, has been calculated and the results are shown in figure 9-13. The maximal value for the DoD is 70% which corresponds to a full discharge if a battery ageing margin of 30% is considered.

Figure 9-13: Depth of discharge as a function of the total trip distance and of the number of stops

The same example was also used to calculate the battery pack weight in function of the total trip distance and the speed attained between 2 stops again. A 3-dimensional plot is used in figure 9-14 to represent the results of the analysis. The areas corresponding to the different battery pack combinations are indicated using the codes given in table 9-1. The speed achieved between 2 stops is varying from 1 to 18 km/h and the total trip distance is varying from 1 to 35 kilometer. For the choosen ranges of the input parameters, the calculated battery pack capacity ranges from 9Ah (combination A) to 43,5Ah (combination H). In between, the other battery pack combinations (B, C, D, E, F and G) are corresponding to areas of operation that are delimited by curved lines. This curved line is due to the quadratic function of the speed appearing in the equations of the total required mechanical energy (see equation 7-25 till equation 7-28). The large influence of the speed reached between two stops on the required battery energy and thus on the battery pack weight can be seen in this figure.


p.177

Figure 9-14: Battery pack weight as a function of the total trip distance and of the speed achieved between 2 stops

As a final analysis, the payload mass and the number of stops were taken as parameter 1 and parameter 2 in the battery energy model. As the output parameter, the required battery energy, expressed in Wh is taken. A coloured 2- dimensional contour plot is used in figure 9-15 to represent the results of the analysis. The number of stops is again varying between 1 and 600 stops and the payload mass is varying from 0 to 100 kilogram. From this analysis it can be concluded that in case of the boundary conditions of this exercise, the required battery energy corresponding to postal usage is much larger than that compared to non postal usage. As an illustration the operation point characterised by 300 stops (rather modest for postal distribution) and a payload mass of 60 kilogram can be consider. This point corresponds to a required battery energy of about 800 Wh. This corresponds to the battery pack combination D. Compared to the battery energy from contemporary available (non-postal) electric bicycles (see figure 8-10) this result illustrates the gap between the battery packs used for (normal use) electric bicycles and the required battery pack energy corresponding to the postal requirements of this exercise, in case of the NEPH power system. Another possible application of the calculation tool is to optimize a delivery route for a given configuration of the electric power system. The length of the route could be maximized within the capability of the available battery energy taking into account the postal mission parameters (number of stops, payload, cumulative height difference…)


p.178

Figure 9-15: Required battery energy as a function of the payload and of the number of stops


p.179

10 Optimization of the thermal

management of a PMDC motor

10.1 Introduction The thermal performance of an electrical machine is very important as it strongly affects the lifetime of the machine. Moreover, overheating of an electrical machine occurs when it has not been well sized for the intended application or when a motor is overloaded for a too long time. Overheating can lead to the degradation of permanent magnets (if used), to the failure of the isolation materials or to the failure of the bearings. In the worst case it can even cause toxic smoke and fire with risk of large material damage and possible severe injuries of persons [209]. An electric machine normally is characterized by a set of rated values. Often, the rating is of the class ‘continuous running duty’ designated with the ‘S1 duty cycle’ (cfr. IEC 60034-1). The latter refers to an operation at a constant load maintained for sufficient time to allow the machine to reach thermal equilibrium (i.e. Δθ < 2K/h with θ the temperature of the motor). In case the electrical machine is used as a motor, the rated output is the mechanical power available at the shaft (expressed in Watts or sometimes in horsepower1). Unless otherwise specified, an electrical machine is designed to operate at certain site operating conditions: the altitude shall not exceed 1000m above sea level; the ambient temperature shall not exceed 40°C and the ambient temperature shall not be lower than 0°C in case of a rated output less than 600W (IEC 60034-1). If an electric motor is used for traction in an (light) electric vehicle, the operating conditions can deviate from what is assumed as normal site operation conditions. Further, for d.c. electric motors supplied from static power converters, the rated form factor2 for which the motor is designed should not be exceeded. Moreover, energy losses and temperature rise will increase if the rated form factor is surpassed [210]. As can be seen from the conditions above cited the use of an electric motor in an electric power system for (light) vehicles requires a lot of attention and consideration with respect to the application and corresponding operation conditions. Therefore it is important that measures are taken by the manufacturer to prevent overheating and hence to avoid (early) damaging of the motor. The measures to avoid overheating can be taken at the level of the motor (e.g. by means of a thermal switch) or the can be incorporated within the motors’ drive unit. In the latter case, the measures can be more advanced taking into account several operating conditions an are often referred to with the term ‘thermal management’. Implementing a thermal

1 1hp – (metric) horsepower – corresponds to 736W 2 form factor of the d.c. motor armature current is defined as the ratio of the rms-value to the average value

Optimalization of the thermal management of a PMDC motor

p.180

management into a drive unit (motor and its drive) allows using the motor at its full capabilities and improving the operational ability of its application.

10.2 Overloading an electric motor A good selection or design of an electric motor takes into account the peak torque and maximum speed requirements of the application. Based on a model of the mechanical system (see chapter 7) a load profile can be obtained and allows for instance to estimate peak torque and maximum speed requirements. In some cases, it can be interesting to allow overloading an electric motor for a short period of time. This is especially true for vehicular applications where also volume and mass of the components have to be limited. Overloading could be avoided if the motor is dimensioned for the required peak power of its application. However, in many cases this would lead to an unnecessary increased cost. In case of electric vehicles or of electric bicycles in particular, legislative limitations exist with regard to the ‘continuous rated power’ of the motor used (cfr. EC/2002/24). Overloading an electric motor means that the output of the motor (in this case the mechanical power at the shaft) surpasses the rated output value. In case of a d.c. motor used for traction, the primary function is the rated torque. In this context, overloading is considered as applying a higher mechanical torque resulting in an increased traction force of the vehicle. Increased mechanical torque is proportional to the armature current in case of separately excited d.c. machines. This can be done without risk of damage as long as the temperatures of the different parts of the motor are not exceeding their respective maximum temperatures. Depending on the thermal class of the motor this corresponds to the maximum temperature of the windings isolation, the maximal operational temperature of the bearings and in the case permanent magnets are used, the maximal temperature of these magnets. Also a big heat capacity of the armature and of the collector drum improves the overload capacity of the motor.

10.3 Description of the standard thermal protection In the forthcoming paragraphs a description of the thermal protection present in the Heinzmann electric power unit is given.

10.3.1 Heinzmann electric drive system The drive system that is considered consists of a mechanically commutated (brushed) permanent magnet direct current motor with an axial magnetic field. The wheel hub motor has a slim design and has a disc shaped rotor. The axial magnetic field is created by a symmetrically arranged magnetic system consisting of high energy magnets (NdFeB). This can be seen in the exploded view of the motor in figure 8-26 on page 155. Further, the enclosed construction of the motor, required for outdoor operation, results in a reduced dissipation of heat to the ambient air in which it operates.


p.181

Figure 10-1: Picture of the Heinzmann PMDC wheel motor

The electrical characteristics of this PMDC motor are summarised and shown in the figure 10-2. In this graphic, the speed is shown in per units (p.u.), related to the no load speed n0 [188, 208]. On this figure, the current characteristic (red line) can be read on the scale on the right handside. The power characteristic (orange line) can be read on the scale on the right handside but should be multiplied with 10W. The speed characteristic (blue line) can be read on the scale on the left handside and is expressed in per unit. Finally, the efficiency characteristic (light blue line) can be read on the scale on the left handside.

PMDC motor charateristics

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 20 40 60 80 100 120 140 160 180 200 220

Motor Torque (Nm)

0

10

20

30

40

50

60

70

80

90

100

110

120

efficiencyspeed (p.u.)current (A)n 0 = 169 rpm - corresponds with v =19,4km/h (for 24" wheel)

P mot_nom = 250W

I a_n = 9A

η mot_nom = 73%

n nom = 157 rpm

I a_max = 33A

η mot_max = 78%

63Nm 15Nm

armature current (A)power output (x10W)

987W

Figure 10-2: Characteristics of the PMDC motor at rated armature voltage (warmed up)

The continuous rated power of the motor is determined at 250 Watts. The rated voltage of the motor is 36 Volt. The maximal efficiency of this motor, ηmot_max, is about 78%. A thermal switch (Klixon) is installed into the motor which interrupts the armature current as soon as the motor temperature, near the permanent magnets has reached the value of 145°C. This thermal switch protects the motor from overheating due to current ratings higher than the continuous rated value or nominal value for a too long time. The nominal value of the armature current Ia_nom can be estimated from the rated motor power Pmot_nom, the rated armature voltage Ua_nom and the efficiency of the motor ηmot. Often only the maximal efficiency ηmot_max value is known but this value can also be used for an approximation of the nominal armature current value Ia_nom.

__

_max _9.

mot noma nom

mot a nom

PI AUη≈ = Equation 10-1


p.182

From the motor characteristics [188, 208] of figure 10-2 the motor torque constant kT can be determined and was found to be 2,01 Nm/A. The no load current Ia_0 can also be read and is 1,71 A and corresponds with a vehicle speed of 19,4 km/h in case of a 24” wheel. A nominal value for the armature current results in an electromagnetic torque at the shaft of the motor Tem of:

[ ] [ ]_ 0.( ) 2,01 / .(9 1,71) 15em T a aT k I I Nm A A Nm= − = − = Equation 10-2

In case of a wheel size of 24” as used for the NEPH prototype, this torque results in a traction force of:

. / 15 / 0,305 50tm em wF T r Nm m N= = = Equation 10-3

In order to increase the motors’ traction force, it can be exploited at an excess armature current. This excess current will cause also a higher temperature of the motor since the heating is proportional to the time and to the square of the armature current. The maximal allowable armature current Ia_max is set at 33A (see figure 10-2) for this motor. This maximum is mainly because of mechanical constraints. This armature current leads to an increased electromagnetic torque Tem and traction force related to the motor Ftm of:

[ ] [ ]_ 0.( ) 2,01 / .(33 1,71) 63em T a aT k I I Nm A A Nm= − = − = Equation 10-4

. / 63 / 0,305 207tm em wF T r Nm m N= = = Equation 10-5

This increased torque and traction force is useful for the application but the heating of the motor becomes an issue if the time that an increased armature current (Ia > 9A) is applied, becomes large. In a first approach the heating of the motor can be linked to the square of the armature current. In figure 10-3, an example of an excess current profile is illustrated. The rated current in a continuous type of duty leads to a thermal equilibrium after a sufficient long time. The temperature that is attained is considered as the limit of the motors’ operational temperature θlim. If an excess current is applied, this will lead to an increased heating of the motor. Depending on the profile of this current, this can lead to a motor temperature exceeding the earlier defined maximal temperature. An example of increased heating due to an excess current is shown in figure 10-4.

Figure 10-3: Example of an excess armature current profile


p.183

Figure 10-4: The effect of an excess current on the motor temperature

The electric motor of the Heinzmann drive systems is using this principle to achieve a maximal output (torque) using a motor that is continuous rated for a power of 0,25kW. The motor is protected using a thermal switch that cuts of the electrical power as soon as the motors’ temperature has reached the limit temperature θlim of 145°C. This allows having a maximal output torque available of about 63Nm, for short time which is very useful for the application of postal distribution. The expected load profile corresponding to postal delivery is characterised by periods of no-load alternated with periods of full and/or partial load. In most circumstances of postal delivery, the excess current up to maximal armature current Ia_max of 33A does not lead to overheating of the motor. However, it was found that for the typical Belgian situation where, the law requires the mailboxes to be placed near the sidewalk, the corresponding load profile did not longer allow the motor to reach a thermal equilibrium beneath the maximal allowable temperature of the motor. This resulted in a sudden shut down of the electrical motor system during endurance tests with stop and go operation with the different prototypes. This is illustrated in figure 10-7of paragraph 10.4. The functioning of the standard thermal protection of the Heinzmann drive system can be described as:

lim _ lim_ _ max

lim _ lim_

if

if 0mot a th a

mot a th

I I

I

θ θθ θ

< = ≥ =

Equation 10-6

10.3.2 On-road tests for investigating the motors’ thermal behaviour To investigate the motors’ thermal behaviour in real use situation, several on-road tests simulating postal distribution were performed. During these tests, the motors’ temperature was recorded by means of two NTC thermistors placed at different locations inside the motor. For these tests a prototype electric bicycle for postal distribution was charged with a payload of 48,5 kg. Together with the mass of the driver (70 kg) this lead to a total system mass of 175 kg. The selected test route was located in Neder-Over-Heembeek in the Brussels Capital Region. This area is characterised by some important hills. The round has a length of 2,2 km and there are about 150 distribution points (horizontal density of 7 stops per hectometre).


p.184

Two test rounds were performed, with a pause of about 7 minutes in between them. At the end of the second test round, the motor had already reached the maximal allowable temperature of 145°C at the level of one of the two thermistors. The evolution of the motors’ temperature can be seen in figure 10-5.

On-road measurement Motor Temperature Neder-Over-Heembeek (August 3th, 2006)

0

20

40

60

80

100

120

140

160

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

Time (minutes)

Tem

per

atu

re (

°C)

NTC yellow

NTC red

Limit

check up

round 1

round 2

Total mass: 175 kg (payload: 48,5kg - driver: 70kg )Ambient temperature: 19°C (raining)Trip: Neder Over Heembeek (2,2 km per round; 2 rounds)Driving style: simulation of postal distribution (c.a. 154 stops per round)

Figure 10-5: Evolution of the motors’ temperature during an on-road test simulating postal distribution

This test showed that in some specific situations, the load profile in combination with the excess current lead to the maximal allowable motor temperature level causing a sudden system shut down. For this reason, a thermal management has been proposed and was implemented by Heinzmann GmbH. This is discussed in the next paragraph.

Figure 10-6: Picture of the test campaign in Neder-Over-Heembeek (January 2008), © Francis Heymans From Left to Right: Jean-Marc Timmermans, Prof. Philippe Lataire, Jens Nietvelt, Jean Vander Elst (Bike Events), Dr. Klaus Gössel

(Heinzmann GmbH)


p.185

10.4 Relating the maximum armature current to the motor’s temperature

In order to find a solution to the sudden system shut down due to overheating, a suitable thermal management was to be developed. A possible solution would be to decreasing the maximum armature current level. However, this would have the disadvantage that the torque output and in consequence the traction force of the postal bicycle would decrease proportionally. Furthermore, the heating of the motor depends on the load profile and in less demanding situations (e.g. a low payload or delivery in a flat region) there will be no need to lower the armature current. Therefore a solution that would adapt itself to the load profile is proposed. In this context, the maximum armature current value would be lowered only if the load profile is causing the motor to heat up to a temperature near to the maximal allowable motor temperature. In order to develop a thermal management that takes into account the load profile, one has to measure the motor’s temperature θmot. For this measurement a thermistor is installed inside the motor. The principle of the thermal management of the wheel motor that was developed foresees an armature current limit which is equal to the maximal armature current for as long as possible. For a large part of the motors’ temperature range, the current limit is unchanged. Only if the motors’ temperature becomes high and close to the maximal temperature, the armature current limit decreases with increasing motor temperature. This principle is illustrated with the graph of figure 10-7.

0

5

10

15

20

25

30

35

0 20 40 60 80 100 120 140 160 180 200

θ_mot (°C)

I_a_

lim

_th

(A

)

I_a_lim_th = I_a_max

I_a_lim_th=f(θ_mot)

θ_1

θ_2

I_a_max

θ_lim

Figure 10-7: Armature current limit as function of the motors’ temperature

In this graph the thermal protection described in the previous chapter is represented by the blue line. An important hysteresis is present, due to the mechanical system of the thermal activated circuit breaker (Klixon) used as protection. This hysteresis is shown in the figure whith the dotted blue line. This hysteresis will cause an important time delay in the reactivation of the motor circuit after a thermal shut down. The proposed thermal management, based on the measurement of the motors’ temperature is shown with the orange line. This principle is based on different temperature ranges, bounded by the temperatures θ1 and θ2. The armature current limit Ia_lim_th is described by the following equations:


p.186

1 _ lim_ _ max

21 2 _ lim_ _ max _ max

2 1

2 _ lim_

if

( )if .

( )

if 0

mot a th a

motmot a th a a

mot a th

I I

I I I

I

θ θθ θθ θ θθ θ

θ θ

< =

− ≤ < = − − ≥ =

Equation 10-7

10.5 Comparative on-road tests The above principle of thermal management has been implemented by Heinzmann GmbH in a prototype controller. This controller was made available for internal testing at the department ETEC for evaluation. The controller with the thermal management was used on a three-wheeler for postal distribution from the German company MIFA A.G. This prototype was used on a test route between the VUB campus Etterbeek and a residential area in Koningslo-Vilvoorde. The test route had a total length of 14,8 km and was characterised by some important hills. The total accumulated height difference for the route was 113m. The effect of the thermal management was evaluated by comparing the evolution of the motors’ temperature with and without the thermal management activated. The result of an on road test with recording of the motors’ temperature can be seen in the figure 10-8.

Motor Temperature profile

0

25

50

75

100

125

150

175

200

225

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

Time (s)

Te

mp

era

ture

(°C

)

rit 2008 02 18 pm

rit 2008 02 23 am

MIFA electric 3-wheelerOn road, commuting, VUB-VilvoordeDistance: 14.8 kmAcc. Heigth difference: 113 m

thermal management switched off

thermal management switched on

Figure 10-8: Comparison of motors’ temperature evolution

From these test results it can be observed what the influence of the thermal management on the motors’ temperature is. The same test route was completed with the thermal management switched on and switched off. The same route for both test rides results in a visibly isomorphous temperature evolution. The part of the temperature curve above a temperature of 175°C is clearly different for both records. In the situation where the proposed thermal management is switched on, the temperature is flattened above 175°C compared to the situation where the thermal management was not activated. Of course this limitation has an influence on the performance of the vehicle at that point. Because the motors’ armature current was limited, the output torque was reduced proportionally. This resulted in a reduced traction force delivered by the wheel motor and was reflected in both the speed of the vehicle and the drivers effort. This was confirmed by the recording of the bicycle speeds for both situations. Aside the bicycle speed, also the altitude was recorded along the test route together with the drivers’ heartbeat rate. The results of the recordings of the vehicle speed, altitude and drivers’ heartbeat rate is shown in figure 10-9 and figure 10-10.


p.187

Figure 10-9: Plots of vehicle speed, altitude and heartbeat rate in case of thermal management switched off

Figure 10-10: Plots of vehicle speed, altitude and heartbeat rate in case of thermal management switched on

For comparing both figures, the curve corresponding with the evolution of the altitude can be used as a reference, rather than the time. As can be expected, the limitation of the armature current for avoiding overheating of the motor, caused a decrease in vehicle speed in the last large ascending of the test route (corresponding with timestamps of about 48 minutes till 55 minutes). In this example the vehicle speed lowered from 14 km/h to about 10 km/h. Also the heartbeat rate of the driver was higher, due to lowered contribution of the electrical power system. The maximal heartbeat rate was increased from less than 110 bpm to about 118 bpm in this particular situation.


p.188

10.6 Conclusions on thermal management With the proposed algorithm for thermal management, the electrical drive system was prevented from sudden shut down due to overheating. In case of excessive use of the electrical traction so that the motors’ temperature reaches almost the maximal allowable temperature, the motors’ armature current will be lowered with increasing motor temperature. This decrease in current encourages the cyclist to deliver more mechanical input power to the vehicle. A new equilibrium between the drivers’ effort and the electrical power establishes so the motors’ temperature remains within acceptable limits. This possibly leads to a lower vehicle speed, depending on the drivers’ ability (or willingness) to deliver more biomechanical power. In cases the load profile is less demanding and the motors’ temperature is not critical, the full output torque (as was also available in the standard drive system) remains at the disposal of the vehicles drive train. The tests pointed out that the values of temperatures θ1 and θ2 that were chosen give satisfactory performance and allow the driver to find a new equilibrium of electrical and biomechanical power in case of demanding (leading to high operational temperatures) situations.


p.189

11 Optimization of the energy

consumption of the electric drive system

11.1 Introduction on the energy consumption optimization This chapter discusses the optimization of the energy consumption of the Heinzmann electric drive system for use in an electric bicycle for postal distribution. An optimisation algorithm was developed to reduce the required energy capacity from the battery packs. This optimisation has been developed to avoid the need for more than one battery pack to complete a postal delivery trip. The control principle that is described below, corresponds to the control software for the Heinzmann electric drive system used on the electric bicycles for postal distribution that were developed in the framework of the Eureka NEPH project [211-213]. The purpose of the proposed control software change is to limit the energy consumption of a postal bicycle equipped with the drive system mentioned above. It was reported by the Belgian Post [214] that, according to the experience acquired during their internal tests, the postmen use the full twist grip position to move from one postal delivery point to the next one, with minimal muscular effort. The purpose of this software change is to encourage the postmen to perform more muscular effort during delivery with the postal electric bicycle hence needing less energy from the battery. It is considered that a drive system using a control based on the measurement of the cyclists' effort (using a force detecting device) allows implementing a proportional share of efforts. Using this measurement of effort, the power from the electric drive can be made proportional to the driver’s biomechanical input power. However, the proposed solution without measurement of effort allows improving the energy consumption in the short term based on software change only. In addition, the use of an expensive force or torque sensing device, often patented, can be avoided in this way. A solution with a control strategy based on effort measurement (i.e. a pedal power control system) would require more development time and leads to an increased initial vehicle cost and an increased components count. The latter can lead to a decreased availability and maintainability in postal service.

Optimalization of the energy consumption of the electric drive system

p.190

11.2 Description of the standard control principle The drive system uses a motor control unit including an electronic buck chopper to control the energy transfer from the battery to the permanent magnet DC motor. The twist grip position (Tgr) is converted directly into a duty cycle control leading to output characteristics as shown in figure 11-1 and figure 11-2. The twist grip position Tgr, or the mechanical position of rotation is described by a percentage of the twist grip’s rotation compared to the maximal possible rotation. This results in a value between 0% and 100%, where 0% corresponds to no rotation of the twist grip and 100% corresponds to the maximal possible rotation. The output characteristic shows the relation between the output current and the output voltage of the motor control unit for different positions of the twist grip (20%, 40%, 60%, 80% and 100%). From the characteristic it can be seen that the output voltage is proportional to the twist grip position but has no influence on the output current. However, the output current is limited by the controller to the limit value (i.e. 33Amps). This can be seen as a pure voltage regulation of the motor. Additionally, a current limiting algorithm is included based on a PI controller. As the motor controller imposes the d.c. motor voltage, and given that this motor has a permanent magnetic field, the bicycle speed is principally determined by the twist grip position. Pushing harder on the bicycle pedals only has a minor influence on the bicycle speed and hence the cyclist has no encouragement to deliver more muscular power at cruising speed. This leads to high energy consumption and leads to a fast battery energy depletion hampering the postal application.

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35 40

U_a (V)

I_a (A)

I_a = I_a_lim

Tgr = 20%

Tgr = 40%

Tgr = 60%

Tgr = 80%

Tgr = 100%

Figure 11-1: Controller output characteristics when turning the pedals


p.191

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35 40

U_a (V)

I_a (A)

I_a = I_a_lim

Tgr = 20%

Tgr = 40%

Tgr = 60%

Tgr = 80%

Tgr = 100%

Figure 11-2: Controller output characteristics without turning the pedals

These characteristics lead to a use for postal delivery during which the bicycle accelerates at motor current limit (Ia_lim) when moving from one postal box to the next one. In Belgium, this acceleration occurs very often during the mission, as it is mandatory that the postal boxes are reachable from the pedestrians' sidewalk. Hence, the postman moves on the sidewalk from one mailbox to the next one without getting off the bicycle and with only very limited need to park the bicycle for short time. The Heinzmann motor system develops high acceleration force so the postman feels no need to deliver more muscular power than to comfortably turn the pedals. The current limit (Ia_lim) is only influenced by the motor temperature measurement. Ia_lim_th equals Ia_max when the motor temperature is normal. When the motor temperature approaches its maximum allowable value, Ia_lim_th, is gradually lowered. See chapter 10, in particular equation 10-7 on page 186.

11.3 Relating the armature current limit to the output voltage The first solution for lowering the battery energy consumption is to relate the current limit Ia_lim to the controller's output voltage, and hence connecting the available traction force to the bicycle speed. The output characteristic in the case of turning the pedals is shown in figure 11-3. The output characteristic when the pedals are not turned is shown in figure 11-4.


p.192

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35 40

U_a (V)

I_a (A)

I_a = I_a_lim

Tgr = 20%

Tgr = 40%

Tgr = 60%

Tgr = 80%

Tgr = 100%

I_a_lim

P_max_1

Figure 11-3: Controller output characteristics when turning the pedals (solution 1)

A proposed mathematical expression for the current limit is given by equation 11-1.

)/,min( 1max_lim__lim_ athaa UPII = Equation 11-1

With this expression the current limit Ia_lim is lowered inversely proportional with the output voltage Ua. As the available motor traction force lowers with the speed of the bicycle, pushing harder on the pedals leads to higher speed and hence encourages the cyclist to deliver a significant part of the moving effort. Maximal starting force will remain available for the postmen to accelerate the bicycle from standstill to a minimal speed. An adequate value of Pmax_1 has to be found experimentally starting from a value of 170 W.

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35 40

U_a (V)

I_a (A)

I_a = I_a_lim

Tgr = 20%

Tgr = 40%

Tgr = 60%

Tgr = 80%

Tgr = 100%

I_a_lim

P_max_1

Figure 11-4: Controller output characteristics without turning the pedals (solution 1)


p.193

11.4 Relating armature current limit to twist grip position and output voltage

It is expected that in solution 1 at reasonable speed, typically 8 km/h or higher, the twist grip will give a ‘dead’ impression, as in most driving conditions the driver will operate at current limit and twist grip having no influence over a large turning range. Figure 11-5 gives an alternative, combining also a better operation when the twist grip is operated with no crankshaft rotation as shown in figure 11-6. A proposed mathematical expression for the current limit is then as given by equation 11-2.

_ lim _ lim_ max_1min(( . ), ( . / ))a a th gr gr aI I T P T U= Equation 11-2

It is considered that this control is safer for winter driving, the existing control leading easily to front wheel slipping in winter road conditions.

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35 40 45

U_a (V)

I_a (A)

I_a = I_a_lim

Tgr = 20%

Tgr = 40%

Tgr = 60%

Tgr = 80%

Figure 11-5: Controller output characteristics when turning the pedals (solution 2)

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35 40 45

U_a (V)

I_a (A)

I_a = I_a_lim

Tgr = 20%

Tgr = 40%

Tgr = 60%

Tgr = 80%

Figure 11-6: controller output characteristics without turning the pedals (solution 2)


p.194

11.5 Modulating the maximum power limit to the available battery energy

It is to be considered that in the previous solutions the full motor capability is never used and that in some exceptional conditions this could be useful or needed and therefore should be made available. In this solution, the value of Pmax_1, as used in equation 11-1 and in equation 11-2, is made dependent on the available battery energy. Therefore it is considered in this solution that the battery energy is released gradually along the mission. Here it is proposed to link the mission to its expected time duration (Tmission), the mission distance measurement being not available in the controller. If the cyclist has been using less energy than foreseen during some time, then the power limit is raised, and vice versa. The allowed energy consumption (Eallowed(t)) is as shown in figure 11-7 and corresponds to an average power use Pavg_allowed, of approximately 100 W if Tmission is set to 5 hours and a typical battery energy content Ebat (fully charged) of 500Wh (following equation 11-3 and equation 11-4). If the mission is stopped for a time larger than Ta, then the incrementing of the Eallowed(t), counting the allowed energy consumption since the start of the mission, is stopped. This is also illustrated in figure 11-7 with an example of 3 long stops of approximately 20 minutes each. Incrementing resumes as soon as the mission is continued. This is detected by energy consumption calculation and is implemented in equation 11-5. As shown by this equation, only the difference between Eallowed(t) and the actually consumed energy is implemented as a variable.

0

100

200

300

400

500

600

0 1 2 3 4 5 6 7 8

Elapsed time since start of mission (h)

Ava

ilab

le E

ner

gy

(Wh

)

Figure 11-7: Energy made available versus time (increment of Eallowed is stopped when driving is stopped for more than Ta )

_ ( )avg allowed bat missionP E T= Equation 11-3

= dtPtE allowedavgallowed . ifrunning*)( _ Equation 11-4

when pedals are turningifrunning 1

ifrunning 0 when stopped for more than Ta

= =

Equation 11-5

The difference between the allowed energy consumption and the actually consumed energy, if positive, is made available for consumption in a time horizon Th (typical value 50 sec). This leads to the power limit Pmax_3 given by equation 11-6.


p.195

max_ 3 _0

1( ) . ( ).

t

con avg allowed bath

P t P P dtT

η= − Equation 11-6

The integral of equation 11-6 has an upper bound given by the product Ub . Ia_max . Ta and has a lower bound given by 0. The factor

conη represents the average controller efficiency and has a typical value of

90%. The integration has an initial value at switching on given by Pavg_allowed. If the drive is operated constantly at current limit, then Pmax_3 tends to a value near Pavg_allowed as shown by equation 11-7.

max_ 3 _lim ( ) .con avg allowedtP t Pη

→∞= Equation 11-7

The integral upper bound is fitted to stop the time counting in case the mission is stopped for an extended time, longer than Ta (typical value of 150 sec). The estimation of the energy delivered by the battery, given by equation 11-8 as part of equation 11-6, can also be delivered by the battery management system, if available.

0( ) .

t

bat batE t P dt= Equation 11-8

The purpose of integrating the difference as shown in equation 11-6 is to obtain a robust algorithm, which is tolerant to measurement errors. The integration is reset to Pavg_allowed every time the controller is switched on. Pmax_3 is introduced in the solution 1 or 2 as:

max_1 max_ 3 _ maxmin( , . )bat aP P U I= Equation 11-9

If a measurement of Ubat is not available, an estimate of Ubat can be used. If a measurement of Ibat is not available for the calculation of Pbat, then Pbat can be calculated from the duty cycle and the motor current Ia. An approximate value of both is available at the level of the control algorithm in the cotroller.

11.6 Comparative on-road tests A comparative test was performed to investigate the influence of the new control algorithm that was implemented into two different prototype controllers of the Heinzmann electric drive system for electric bicycles. In particular the influence of the new control algorithm on the energy consumption is of interest. For this purpose on road tests were organized with a prototype electric postal bicycle (see a picture of the prototype manufactured by Ludo nv. in Figure 11-8)


p.196

Figure 11-8: Prototype used for the on-road comparative tests

Three different controllers were installed side by side on this electric postal bicycle (see Figure 11-8) and were connected one after the other:

• Controller nr. 1: the controller as used for the Belgian Post Project. 870-00-360 • Controller nr. 2: the first new controller with new control algorithm (corresponding to

solution 1 from paragraph 11.3) 870-00-360-LUDO-66 SW 111.0.05+Power_Limit+Eco-Mode+Parameter _1

• Controller nr. 3: the second new controller with new control algorithm 870-00-360-LUDO-66 SW 111.0.05+Power_Limit+Eco-Mode+Parameter _2

A test route in Neder-Over-Heembeek (Brussels Capital Region), the same as in paragraph 10.3.2, was chosen to perform this comparative on road test. Postal distribution was simulated by performing a full stop, with 3 seconds waiting, at every mail box encountered along the test route. The test route is situated in an urban area with a high horizontal density of distribution points. In addition, the area has some important slopes. The test route is characterized by the following figures:

• Total distance: 2,4 km • Number of mail boxes: 150 boxes • Total Mass: 196,4 kg

(Bicycle: 41,4 kg; Batteries: 26,4 kg; Cyclist: 85 kg; Payload: 43,6kg) • Cumulative height difference: 37m

The same test route was completed, using the three different controllers. The battery current, Ib, and the battery voltage, Ub, were recorded, using a Fluke Scopemeter (type196B). The battery current is measured, using a LEM current transducer (type LEM-100S). A sampling frequency of 12,5Hz was used, allowing a maximal recording time of 50 minutes. From this measurement data, the instant power taken from the battery, Pbattery, is calculated:

)()()( tUtItP bbBattery ⋅= Equation 11-10

Then, the energy consumption can be derived for the three different test rounds:

=t

BatteryBattery dttPtE tripofbegin

)()( Equation 11-11

Three controllers


p.197

11.7 Test Results

11.7.1 Energy delivered by the battery The same test route was completed three times, using respectively controller number 1, 2 and 3. The battery voltage and the battery current were recorded during these tests. Below, the calculated energy consumption of the drive system (here expressed in Watt-hour) at the level of the battery is plotted against time (expressed in minutes): The following colour code will be used in the graphs below: • Blue: test results corresponding with the test using controller number 1 • Red: test results corresponding with the test using controller number 2 • Green: test results corresponding with the test using controller nubmer 3

Figure 11-9: Energy consumption versus time for the three controllers

From this graph an important difference in energy consumption can be observed between the controller 1 on the one hand and the controllers 2 and 3 on the other hand. The numerical values, taken from this graph, can be read in the table below:

Table 11-1: Energy consumption comparison

TOTAL ENERGY CONSUMPTION AT

END OF THE TEST ROUTE (Wh) RELATIVE ENERGY CONSUMPTION COMPARED

WITH CONTROLLER 1 Trip with controller 1 (blue) 89,7 100% Trip with controller 2 (red) 63,5 71% Trip with controller 3 (green) 55,8 62%


p.198

The influence of the new control algorithm, with power limitation, on the power taken from the batteries can also be seen in the power spectrum of the power taken from the battery pack. In the three histograms below the influence of the power limitation can be observed. The power was calculated for every time step of 80ms. The occurrences of 0 Watt power use was removed from the data sets for reason of clarity of the plots. The difference in parameter setting between the controller number 2 and the controller number 3 can also be seen.

Figure 11-10: Histogram of power spectrum of Pbat, when using controller number 1



p.199


From the three histograms, it is clear that the new control algorithm causes a shift in the power delivered from the battery towards lower power levels and concentrates the power around a mean certain value. This mean value is different for controller number 2 and controller number 3, according to the parameter setting in the software. In particular, the parameter Pmax_1 is different for controller number 2 and controller number 3. As no significant difference in mission duration (Tmission) could be observed the significant decrease in energy consumption could be explained in an increase of the driver’s effort.

11.7.2 Effort from the cyclist During the comparative on road tests, the cyclist was equipped with a heart rate sensor (Polar, type CS600) and the evolution of the heart rate, H (beats per minute) of the cyclist was recorded for the three test rounds with the different controllers. Further, the Polar CS600 also registers the altitude, A (meter above sea level) and covered distance, D (km) during the test rounds. The relative altitude Ar is calculated as follow:

startrA A A= − Equation 11-12

The measurement of the altitude is plotted against the trip distance in the figure below and helps to have an idea of the topology of the test route and to interpret the heart rate measurements.


p.200

Figure 11-13: Relative altitude versus trip distance

Important slopes can be identified in the beginning of the test round, after about 1250 meter and after about 1900 meter. The heart beat is compared to the average heart rate of that trip and gives the relative heart beat rate:

trip theof rateheart avarage the~

with ~

HHHHr = Equation 11-13

In the figure below, the relative heart rates are plotted against the trip distance (in km):

Figure 11-14: Relative heartbeat in function of the trip distance

In this figure an increased heart rate of the cyclist can be observed in case of the use of controller number 2 and number 3 compared with the case when using controller number 1. This can be explained by a higher effort from the cyclist. The latter is especially true when driving uphill (see the first 100 meter, after 1250 meter and at the last part of the trip after 1900 meter).


p.201

11.8 Conclusions on the energy consumption optimization The purpose of the software change was to optimise (to lower) the energy consumption from the battery packs and to encourage the postman to participate whith more own muscular power to drive the vehicle. The first test results, described in the previous paragraph, show an important lowering of the energy consumption from the battery pack for a given mission. Now significant change in mission time was observed. On the hills, in case of the optimized controllers, the heartbeat from the cyclist is higher compared to the case of the standard controller. This is an indication that the cyclist had to deliver more effort and thus was encouraged to deliver a larger share of the total traction force. This adaptation in software is considered to be a good solution for lowering the energy consumption of the electrical power system used in the electric postal bicycle with the Heinzmann drive system.


p.202

Jean-Marc

Rectangle

p.203

PART III: General conclusions and Future Work

Jean-Marc

Rectangle

p.204

Jean-Marc

Rectangle

p.205

12 General conclusions

12.1 General conclusions PART I: In chapter 1 an overview was given of the most important environmental effects of atmospheric pollution related to the use of road vehicles. The primary atmospheric pollutants from road vehicles were briefly introduced and their main effects and impact on human health have been summarized. A summary of the relative European regulations for the quality of the air was provided. Finally, the most important environmental issues related to pollutants from road vehicles have been discussed. In chapter 2 an inventory of all the main sources of the pollutants that were introduced in chapter 1 is made. The different life stages of a road vehicle have been identified. Based on a literature survey, the vehicle use phase and fuel production phase were found most dominant when making a comparative analysis of the air emissions of different vehicles. A well-to-wheel (WTW) framework was identified as the most suitable approach for the inventory of polluting emissions. Further, possible data sources for the inventory of the well-to-tank (WTT) emissions were investigated as data availability is an important boundary condition for the environmental assessement. Two main categories of emissions were considered: direct or tailpipe emissions and indirect emissions related to the production of the fuel. The direct emissions can be further subdivided into regulated emissions and unregulated emissions. For the regulated direct emissions, homologation data were considered as they provide emission values that have been determined by simulating an identical and standard driving cycle. Other unregulated direct emissions were estimated from the fuel consumption of the vehicle by using emission factors or were estimated on the basis of technical characteristics of the vehicle (type of fuel, emission standard). Finally, indirect emissions were estimated from the fuel consumption by using emission factors related to the type of fuel. For electric energy consumption, in the case of electric vehicles, indirect emission factors were also used to estimate the emissions values related to the electricity production process. In chapter 3 a brief description of technical solutions and as well as possible policy measures to reduce the environmental impact of road vehicles was provided. In particular, the different possible measures were evaluated to be adapted in function of the environmental characteristics of the vehicle. The diverse possible solutions discussed in this paragraph show the importance and need for a scientifically sound yet pragmatic and clear definition of “environmental friendly vehicles”. Chapter 4 described the methodology of the environmental rating tool that has been developed in the framework of the research at hand. A state-of-the-art of environmental assessment tools was made and the main shortcomings and advantages have been discussed. The WTW framework was used as the basis for the environmental assessment tool. The WTW energy consumption of a vehicle has been defined as an additional indicator. The newly developed methodology for the environmental assessment

General conclusions

p.206

of road vehicles is further explained and detailed using five steps, similar to that used for LCA: emissions inventory, classification of the emissions into different damage categories, characterisation or damage calculation, normalisation of the different damages by considering a reference vehicle and finaly the aggregation of the different normalised damages into one single indicator (total impact) by use of a weighting system. Additionally a rescaling of the total impact was used to transform the results into an easily understandable number between 0 and 100. This number is called the Ecoscore. The higher the Ecoscore the lower the environmental impact of the vehicle. The methodology that has been developed, allows the assessment of road vehicles with different drive trains and using different fuels. Real vehicle results were calculated and discussed to demonstrate the applicability of this methodology. From the analysis made with the Ecoscore methodology, a positive evolution of the environmental performance of vehicles through time can be observed. This is mainly due to the ever more stringent European emission regulations. A low environmental impact (and therefore a high Ecoscore) was obtained for the battery electric vehicle (Peugeot 106 electric). Also the hybrid petrol-electric vehicle (Toyota Prius) and the CNG vehicle (Opel Astra) obtained a high Ecoscore. The LPG vehicle shows the best environmental performance amongst all conventional vehicles, whereas Euro 4 petrol and diesel vehicles have a similar Ecoscore. When considering the complete range of road vehicles on the road in Belgium, the emissions of CO2 from more recent vehicles, which were directly related to the fuel consumption, were not always reduced. The positive influence of an improved engine technology is sometimes annihilated by an increase of the vehicles’ weight or an increased energy consumption caused by certain on-board options. It is noticeable that the newest generation of diesel vehicles has caught up its delay concerning their environmental performances on the petrol vehicles. Furthermore, the difference between the environmental performance of those vehicles and the environmental performance of the LPG vehicles has been reduced. Finally, on the basis of a sensitivity analysis, the robustness of the methodology has been evaluated. As a general conclusion, one can state that the environmental rating system is robust and thus applicable as a policy instrument (taxation, incentives, consciousness raising campaigns, etc.) to support the use of environmentally friendly vehicles. The ambition of the Ecoscore environmental rating tool is to lead to a common system for policy measures in Belgium and possibly in other European countries, to promote the introduction and use of cleaner vehicles.

12.2 General conclusions PART II: In chapter 5 the large and variated range of light electric vehicles has been introduced. Because of the absence of a clear and uniformal definition, a proposal for a definition of light electric vehicles has been made. The electric bicycles have been identified as a specific subset of the group of light electric vehicles and the most important differences from a technical point of view have been discussed. In particular a classification based on the integration of the electric power system of the electric bicycle has been proposed. A brief history of the use of bicycles for postal distribution has been presented. A description of the recent history of the use of electric bicyles for postal delivery was given and has been complemented with an overview of the use of other types of electric vehicles for postal distribution. In chapter 6, the NEPH project, that was the basis for the research described in part II, has been presented. The questionnaire that has been set up by the author in the framework of this project and the feedback of the postal operators to this questionnaire have been discussed and analysed. Based on this feedback, a set of parameters influencing the mail delivery mode have been identified. A state-of-the-art of the use of light electric vehicles for mail delivery by the postal operators was made. This chapter provided a set of technical parameters that describe the requirements (or requests) from the postal operators. These postal mission parameters form an important basis for further analysis and have been used throughout the remainder of the work at hand. Chapter 7 gave a description of the main physics involved in driving a bicycle. In particular a mathematical model for calculating the required mechanical power and required mechanical energy has been described in this chapter. For this purpose, the different forces that act on the vehicle have been described and considered. Detailed knowledge of the speed cycle of the bicycle would be required to calculate the required mechanical energy with this formulation. However this information is often not

General conclusions

p.207

available or difficult and expensive to determine. Therefore, an alternative method has been derived to estimate the required mechanical energy to perform a certain journey or postal distribution round. This alternative method uses the postal mission parameters derived in chapter 6 as input parameters. The drive technologies used for contemporary electric bicycles have been described in chapter 8. In this chapter an overview of the available drive technologies used for electric bicycles was given. Several drive train parameters have been measured and presented. The constituent components of the electric power system used for electric bicycles were explained. Further, this chapter reveiled the relation between the total required mechanical energy Etot from chapter 7 and the required battery energy Ebat_req. The sizing of the energy storage system has been detailed. The global assistance factor and efficiency allowing the description of the relation between the required mechanical energy and the required electrical energy of the energy storage were defined. Finally, in this chapter, the electric power system that was used in the framework of the NEPH project was described. The model to estimate the required energy capacity of the battery was described in chapter 9. The purpose of the battery energy model is to have a tool that can be used to estimate the required energy of the battery pack in case of an electric postal bicycle. The most important characteristic of the battery energy model is that the postal mission parameters, as described in chapter 6, can be used as input for this model:

• Total trip length Dmission (km) • Number of stops (#) • Cumulative height difference hcumul (m) • Payload Mp (kg) • Driver mass Md (kg) • Average speed v (km/h) • Speed reached between 2 stops vs (km/h)

A software tool has been developed and illustrated by using it to calculate the required energy capacity in case of a typical set of postal mission parameters corresponding to the postal requirements resulting from the NEPH project. With this example, a clear gap was demonstrated between the required battery energy content for such a mission (about 800Wh) and the available energy content on the contemporary models of (normal consumer) electric bicycles (about 300Wh). However, in the case of a power-on-demand system, as used for the NEPH project, the required energy from the battery is strongly dependant of the input of the driver. The required energy can be strongly lowered if the cyclist is considered to deliver more energy. The calculation tool now allows forecasting the performance of a specific configuration of the modular NEPH drive train. The energy capacity can be calculated for each individual postal delivery round and will vary from region to region, depending on the mission. The modular approach of the NEPH project in combination with the design tool, will allow each postal organization to compose their postal electric bicycles, corresponding to their specific needs and will allow them matching and optimising the distribution rounds to the capacity of the NEPH vehicles. Chapter 10 dealt with the optimisation of the thermal management of the motor used in the NEPH power system. The measurement of the motors’ temperature was used to modulate the maximum allowable armature current of the permanent magnet d.c. wheel motor. The proposed principle of thermal management was implemented on a prototype electric bicycle for postal delivery and has been validated. Hereby, the thermal management of the motor was considered to be improved in view of the application as it allows the user to use the electric motor for an extended time. The limit on the armature current is only lowered in case of excessive heating of the motor. Sudden motor shut down due to overheating is prevented as the armature current is progressively reduced with increasing motor temperature. Chapter 11 described possible solutions to lower the energy consumption of the power-on-demand system used for the NEPH power system. It was confirmed that the postmen tend to minimize their own contribution to the required traction force. The power-on-demand system allows to use the electric power at nearly 100% of assistance factor (i.e. the electric motor delivers all required traction force). A first solution to reduce the electric energy consumption of the NEPH power system relates the armature

General conclusions

p.208

current limit to the output voltage of the motor controller. In a second solution the armature current limit is related to both the output voltage and the twist grip position. In a third solution, the armature current limit is linked to the available battery energy. The first solution has been implemented in two new NEPH motor controllers and was evaluated on the road with two different parameter settings. A significant decrease in energy consumption was observed for both new controllers. From the on road tests, a reduction of 29% and 38% reduction were obtained. This solution allows encouraging the cyclist to deliver more biomechanical power as the contribution of the motor to the total traction power is reduced with increasing speed. However, the traction force from the motor available at start is not compromised. This is important if the physical stress on the postmen is to be limited.

General conclusions

p.209

13 Future work

13.1 Future research in the field of environmental assessment of road vehicles

An important action for future research in the field of environmental assessment of road vehicles is the continuous actualisation of the different damage factors that integrate the current state of knowledge on the effects of the different pollutants on the environment and on human health. These actualisations of the damage factors include updated values for the global warming potentials (GWPs), for external costs and for other impact factors. Further, the continuous actualisation of the indirect emission data that reflect the current state of production technologies and pathways for different fuel types and electricity is also important to enable the assessment of the current fuel production chains. In particular, the knowledge of emission levels related to improved or alternative fuels production chains or related to electricity production were crucial for assessing new and future fuel options as well as to asses emerging vehicle technologies. Also the examination of the evolution of the fuel characteristics, such as the lower heating value, fuel density, sulphur content, is important for future evaluations. In addition, the emission levels for the production pathways of alternative fuels (hydrogen, bio-fuels) are essential for the future research work. The collection of emission data and/or fuel or energy consumption data of new and alterative vehicles should ensure up to date input for the environmental assessment tool. The use of data that is more representative for the real emission levels of road vehicles should be further investigated. This should be done keeping in mind that the input data of all different vehicles should be comparable and based on a same type of utilization of the vehicle. The use of more realistic drive cycles for homologation is key in this regard. In other words, the development of an updated test procedure for homologation should be stimulated. Quality control of emission data available in vehicle databases should be made more efficient and performant through maximal automatisation of this control. Identification of possible corrections (e.g. emission data in milligram instead of in gram per kilometer) could also be automated and could further improve data quality. Moreover, the quality of the data available in vehicle databases is crucial for fleet analyses. Fleet analyses including the evaluation of possible future scenarios form also an interesting option for future research. The extention of the existing assessment methodology from a well-to-wheel approach to a complete cradle-to-grave or LCA is an important element in the context of the evaluation of vehicles using alternative fuels or drive trains and when comparing these vehicles to conventional vehicle types. The

Future work

p.210

influence of the use of new materials (e.g. composites) or components (in particular batteries, electric machines) in vehicles on their overal environmental impact is an important subject of future research. Sensitivity analyses are an important element in the evaluation of the influence of variations in emission data and of variations in damage factors used in the methodology. Statistical analyses of emission input data with the environmental assessment methodology can provide new insights in the variation of the environmental impact of vehicles. Further extention of the research consists in the environmental assessments of vehicles to assessments of transport with other means of transportation or the assessement of multi-modal transport. Particular challenges are the identification of suitable reference situations for these assessments. A possible application of the environmental assessment methodology that has been elaborated in the framework of this research is the use of the methodology within multi-criteria analyses. In this way, the presented environmental assessment methodology can be combined with socio-economic evaluations. These multi-criteria analyses can be applied for the evaluation and motivation of different options for the renewal of vehicle fleets or for the selection between different alternatives of cleaner vehicle technologies. Finally, a possible application of the environmental assessment methodology is the use of the methodology for a definition of a cost function that is applied in the power flow management for hybrid vehicles. Currently, these cost fuctions use engine efficiency maps. This could be extended to more advanced cost functions using emission maps of different pollutants and it this way improving the environmental impact of hybrid vehicles.

13.2 Future research in the field of light electric vehicles for postal delivery

The introduction of electric bicycles on the European market can be denoted as a success. This was one of the merits of the demonstration projects E-Tour in which the department of Electrotechnical Engineering and Energy Technology (ETEC) actively participated. The main shortcomings have been identified and since then the performance of (a part of) the commercially available electric bicycles has evolved positively. Especially the driving range of electric bicycles has increased thanks to the use of new battery technologies. However, (large) differences still exist in the price, quality and performances of electric bicycles offered on the European market. The work performed in the assessment of normal consumer electric bicycles has resulted in a test program that allows identifying these differences. Continuous test programs, possibly in collaboration with consumer organizations can help to sensibilize and inform users of these differences and to provide feedback to the manufacturers or assemblers. Beside electric bicycles, the next evolution will be the introduction of electric scooters and motorcycles. In this regard, a new challenge will be to define a good test cycle to define the energy consumption and range in the framework of mandatory homologation of these motorized vehicles. Thanks to the NEPH project, the postal organizations have learned to know the potential of light electric vehicle and in particular of electric bicycles for postal distribution application. This has given rise to several initiatives from postal organizations to assess and even commercially exploit this kind of vehicles for their daily mail delivery. One of these initiatives is also the Green Fleet project from PostEurop and has as a main objective to provide a platform for best-practices with regard to the use of “green” vehicles for postal distribution. This project does not only deal with light electric vehicles but also with passenger cars, delivery vans and even small, medium and large trucks. In order to guide the different postal organizations towards the best available alternatives to convert their current delivery vehicles fleet into a more sustainable and environmental friendly fleet, multi-criteria analysis will be required. This analysis should combine environmental aspects with sound economical considerations. Nowadays, pure purchase costs are still overruling when purchase decisions are taken. Instead, total cost of ownership (TCO) in combination with externals costs should be determined for the different available

Future work

p.211

options (including electric bicycles). The definition of a correct TCO for postal delivery business and the gathering of the required input for TCO analysis is in this regard an interesting option for future work. The use of light electric vehicles for postal delivery could be further supported by creating favourable legislative and regulatory conditions. A derogation to use light electric vehicles on the sidewalk for the specific application for postal delivery corresponds with a real need. An extension of the existing exemption of type approval for electric bicycles should be considered for the specific application of postal delivery purposes. The extension could consist of an increase of the maximal allowable continuous rated power of the auxiliary electric motor installed from 0,25kW to 0,5kW. Further, a close follow-up of the use and performance of electric bicycles in postal service should be considered for future work. Preferrably, this should happen in collaboration with the involved manufacturers/assemblers and with the postal organization(s). Real-life data about range, speed, route, stop times, battery state of charge, motor temperature, etc. could provide useful feedback for further optimization. As most postal organizations use route planning software to optimize the delivery routes and the use of the vehicle fleet, light electric vehicles could become a specific type of vehicle with its own specific characteristics. The inclusion of electric bicycles (and other LEV) into route planning software is a potential and interesting field of future work. A qualitative analysis of the use and maintenance of electric bicycles for postal delivery would be useful to obtain practical feedback from the users (mailmen) and from the people responsible for maintance and repair. Such a qualitative assessment, based on questionnaires could give a better insight in the real world performances of electric bicycles for postal delivery.

Future work

p.212

Jean-Marc

Rectangle

p.213

Appendix 1 Moment of Inertia of a wheel motor

Figure A- 1: Schematic representation of an electric bicycle system with wheel motor, whith regard to acceleration power

The system of a (postal) bicycle with a wheel motor can be represented by the three different elements shown in Figure A- 1. The wheel motor installed in the hub of the (front-) wheel has a mass of m1 and a radius of r1. The wheels of the bicycle have a mass m2 and a radius of r2. The rest of the bicycle (frame, driver, payload…) has a mass of M. Considering that the bicycle moves at a speed v, the rotational speed of the wheel ωw and of the wheel motor can be determined as:

2w

v

rω = Equation A- 1

The power needed for acceleration FA can be written as the sum of the rotational power and the translational power:

. .A tot w aP F vτ ω= +

( ) ( )1 2 1 2. . . .A w w w

dvP I I M m m v

dtω ω ω= + + + +

( ) ( )1 2 1 22

. . . .A w w

v dvP I I M m m v

r dtω ω= + + + +

( )1 21 22

2

. .A

I I dv dvP v M m m v

r dt dt

+= + + +

Appendix 1

p.214

( ) ( )1 2 1 22 2

. . .A

d v v dvP I I M m m v

dt r r dt

= + + + +

1 21 22

2

.A

I I dvP M m m v

r dt

+= + + +

Equation A- 2

The moment of inertia of the bicycle wheel (without the motor) I2 can be approximated by:

22 2 2.I m r= Equation A- 3

In the approximation the mass of the wheel (rims, tyre, spokes) is considered to contribute to the moment of inertia as a cylindrical shell with radius r2 and mass m2. After substitution of Equation A- 3 in Equation A- 2 this leads to:

11 2 2

2

2. .A

I dvP M m m v

r dt

= + + +

Equation A- 4

A wheel motor with an internal gear system with gear ratio i is considered. Now, not all elements of the wheel motor have the same rotational speed. The wheelmotor can, for this analysis, be divided into three parts. One part (a) consists of the elements of the motor rotating with a speed equal to that of the wheel ωw. The second part (b) groups the elements that are rotating with a speed i.ωw. The third part (c) consists of the elements of the motor that are not rotating (the shaft). The mass of the wheel motor can consequently be subdivided according to these three parts, respectively ma, mb and mc. The moment of inertia of the wheelmotor I1 with respect to the axis of the wheel shaft can be expressed as:

21 .a bI I i I= + Equation A- 5

With:

2.

2a a

a

m rI = and

2.

2b b

b

m rI = Equation A- 6

In these approximations the masses ma and mb are considered to contribute to the moment of inertia as a mass distributed along the circumference of a massive cylindrical (with homogeneous mass density) with radius r1 and masses ma and mb respectively. After substitution of Equation A- 5 and Equation A- 6 into Equation A- 4 this leads to:

2 22

1 2 22

. .12. . . .

2 2a a b b

A

m r m r dvP M m m i v

r dt

= + + + +

2 22

1 2 2 22 2

.2. . . . .

2 2a a b b

A

r m r m dvP M m m i v

r r dt

= + + + +

Equation A- 7

or

. .A app

dvP M v

dt= Equation A- 8

Appendix 1

p.215

With:

2 22

1 2 2 22 2

.2. . . .

2 2a a b b

app

r m r mM M m m i

r r

= + + + +

2 22

2 2 22 2

.. . .

2 2a a b b

app tot

r m r mM M m i

r r

= + + +

Equation A- 9

From this expression it can be seen that the influence of the rotating parts of the motor on the moment of inertia of the system will depend on three parameters: the ratio of the rotating parts ra and rb to the wheel radius r2; the gear ratio i of the internal gear system and the mass of the rotating elements ma and mb. This apparent mass is now calculated for the Heinzmann wheel motor used in a NEPH prototype electric bicycle for postal delivery. The following data were obtained from datasheets or from own measurements:

• radius of the wheel motor: ra = r1= 7,5 cm • radius of the rotating parts at rotational speed i.ωw : rb = 5,5 cm • radius of the wheels (diameter: 24 inch) : r2 = 30,48 cm • mass of the wheel motor: m1 = 3,90 kg • mass ma of the rotating elements with a speed i.ωw = 0,20 kg • mass mb of the rotating elements with a speed ωw = 1,55 kg • mass of the wheels : m2 = 2x2,30 kg = 4,60 kg (for 2 wheels) • gear ratio of the internal gear system: i = 16,96

A total mass Mtot of 150 kg is considered. Now the apparent mass can be calculated:

( )

2 22

2 2 22 2

2 2

2

.. . .

2 2

7,5 1,55 5,5 0,20150 2.2,3 . 16,96 .

30,48 2 30,48 2

150 4,6 0,047 0,94 155,6

a a b bapp tot

r m r mM M m i

r r

cm kg cm kgkg kg

cm cm

kg kg kg kg kg

= + + +

= + + + = + + + =

This leads to an apparent mass Mapp which is about 4% higher than the total mass Mtot. The share of the electric motor in the apparent mass increase is less than 0,7% in case of a total mass of 150kg. Thanks to a good design of the wheel motor, the influence of the corresponding moment of inertia can be kept relatively low.

Appendix 1

p.216

Jean-Marc

Rectangle

p.217

Bibliography [1] Beirle, S., Platt, U., Wenig, M., and Wagner, T., “Weekly Cycle of No2 by Gome Measurements: A

Signature of Anthropogenic Sources,” Atmospheric Chemistry and Physics, vol. 3, pp. 2225-2232, 2003.

[2] de Vlieger, I., Cornelis, E., Int Panis, L., Schrooten, L., Govaerts, L., Pelkmans, L., Logghe, S., Vanhove, F., De Ceuster, G. et al., Mira (2005) Milieurapport Vlaanderen, Achtergronddocument 2005, Transport, Vlaamse Milieumaatschappij, 2005.

[3] European Commission, White Paper: European Transport Policy for 2010: Time to Decide, Brussels, 2001.

[4] Van Mierlo, J., Maggetto, G., and Lataire, P., “Which Energy Source for Road Transport in the Future? A Comparison of Battery, Hybrid and Fuel Cell Vehicles,” Energy Conversion and Management, vol. 47, no. 17, pp. 2748-2760, 2006.

[5] Van Mierlo, J., and Macharis, C., Goederen- En Personenvervoer: Vooruitzichten En Breekpunten, Antwerpen - Apeldoorn: Garant, 2005.

[6] Ulrich, K., “Estimating the Technology Frontier for Personal Electric Vehicles,” TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, vol. 13, no. 5-6, pp. 448-462, OCT-DEC 2005, 2005.

[7] Pickering, A., “Green and Clean,” Postal Technology International, vol. June 2009, pp. 2, 2009. [8] Hickman, A. J., Hassel, D., Joumard, R., Samaras, Z., and Sorenson, S., Methodology for

Calculating Transport Emissions and Energy Consumption - Deliverable 22 for the Meet Project, European Commission, Directorate General Transport, Crowthorne, United Kingdom, 1999.

[9] Goedkoop, M., and Spriensma, R., The Eco-Indicator 99 a Damage Oriented Method for Life Cycle Impact Assessment - Methodology Report, Ministerie van Volkshuisvesting, Ruimtelijke Ordening en Milieubeheer, Amersfoort, 2001.

[10] Timmermans, J.-M., Van Mierlo, J., Govaerts, L., Verlaak, J., De Keukeleere, D., Meyer, S., and Hecq, W., Taak 1: Eenduidige Methode Voor De Bepaling Van De Ecoscore Van Voertuigen, 2005.

[11] European Commission, Externalities of Energy - Externe Project, vol. 2, European Commission, Directorate-General XII, Sience, Research and Development, Brussels, 1995.

[12] Externalities of Energy - Externe Project, vol. 2, European Commission, Directorate-General XII, Sience, Research and Development, Brussels, 1995.

[13] Pilkington, A., Hurley, F., and Donnan, P., Health Effects in Externe Transport: Assessment and Expore-Response Functions, Institute of Occupational medicine, Edingburgh, 1997.

[14] Spilstijns, J., Mobilis-Studie over De Mobiliteit: Febiac, 1992. [15] Lenaers, G., "Emissies: Laag, Lager, Laagst Met Uitlaatgasnabehandeling." [16] Commission, E., Eu Energy and Transport in Figures - Statistical Pocketbook 2009, Brussels:

European Communities, 2009. [17] Omaye, S. T., “Metabolic Modulation of Carbon Monoxide Toxicity,” Toxicology, vol. 180, no. 2,

pp. 139-150, 2002. [18] Carlisle, A., and Sharp, N., “Exercise and Outdoor Ambient Air Pollution,” BRITISH JOURNAL OF

SPORTS MEDICINE, vol. 35, no. 4, pp. 214-222, AUG 2001, 2001. [19] Chan, M., and Danzon, M., Air Quality Guidelines, Global Update 2005, Copenhagen, Denmark:

World Health Organisation, 2006.

Bibliography

p.218

[20] Le Tertre, A., Quenel, P., Medina, S., Le Moullec, Y., Festy, B., Ferry, R., and Dab, W., “Short Term Modelling between Air Pollution and Health. An Example: So2 and Total Mortality Paris, 1987-1990,” Revue D Epidemiologie Et De Sante Publique, vol. 46, no. 4, pp. 316-328, 1998.

[21] Dockery, D. W., Schwartz, J., and Spengler, J. D., “Air-Pollution and Daily Mortality - Associations with Particulates and Acid Aerosols,” Environmental Research, vol. 59, no. 2, pp. 362-373, 1992.

[22] Schwartz, J., “Particulate Air-Pollution and Chronic Respiratory-Disease,” Environmental Research, vol. 62, no. 1, pp. 7-13, 1993.

[23] Dockery, D. W., and Pope, C. A., “Acute Respiratory Effects of Particulate Air-Pollution,” Annual Review of Public Health, vol. 15, pp. 107-132, 1994.

[24] Gasser, M., Riediker, M., Mueller, L., Perrenoud, A., Blank, F., Gehr, P., and Rothen-Rutishauser, B., “Toxic Effects of Brake Wear Particles on Epithelial Lung Cells in Vitro,” Particle and Fibre Toxicology, vol. 6, no. 1, pp. 30, 2009.

[25] Chang, T. Y., and Rudy, S. J., “Ozone-Forming Potential of Organic Emissions from Alternative-Fueled Vehicles,” Atmospheric Environment. Part A. General Topics, vol. 24, no. 9, pp. 2421-2430, 1990.

[26] Berglund, M., Bostrom, C. E., Bylin, G., Ewetz, L., Gustafsson, L., Moldeus, P., Norberg, S., Pershagen, G., and Victorin, K., “Health Risk-Evaluation of Nitrogen-Oxides,” Scandinavian Journal of Work Environment & Health, vol. 19, pp. 1-72, 1993.

[27] Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. Benzene, Lyon, France: I.A.R.C. - Iternational Agency for Research on Cancer, 1989.

[28] Needleman, H. L., and Gatsonis, C. A., “Low-Level Lead-Exposure and the Iq of Children - a Meta-Analysis of Modern Studies,” Jama-Journal of the American Medical Association, vol. 263, no. 5, pp. 673-678, 1990.

[29] Otto, D. A., Benignus, V. A., Muller, K. E., and Barton, C. N., “Effects of Age and Body Lead Burden on Cns Function in Young-Children 1 Slow Cortical Potentials,” Electroencephalography and Clinical Neurophysiology, vol. 52, no. 4, pp. 229-239, 1981.

[30] Zayed, J., “Use of Mmt in Canadian Gasoline: Health and Environment Issues,” American Journal of Industrial Medicine, vol. 39, no. 4, pp. 426-433, 2001.

[31] Hassett, C., Mustafa, M. G., Coulson, W. F., and Elashoff, R. M., “Murine Lung Carcinogenesis Following Exposure to Ambient Ozone Concentrations,” Journal of the National Cancer Institute, vol. 75, no. 4, pp. 771-777, 1985.

[32] Beckett, W. S., McDonnell, W. F., Horstman, D. H., and House, D. E., “Role of the Parasympathetic Nervous-System in Acute Lung Response to Ozone,” Journal of Applied Physiology, vol. 59, no. 6, pp. 1879-1885, 1985.

[33] Externalities of Fuel Cycles - Externe Project, European Commission, Directorate-General XII, Science, Research and Development, JOULE, Luxembourg, 1995.

[34] Ipcc Third Assessment Report - Climate Change 2001, pp. 873, New York, USA: University Press, 2003.

[35] Ipcc Fourth Assessment Report: Climate Change 2007: The Physical Science Basis, New York: Cambridge University Press, 2007.

[36] website; "Ozonlaag - Milieuloket," (accessed on: 25/01/2010) Available at: http://www.milieuloket.nl/.

[37] De Geest, C., Milieurapport Vlaanderen 2002, Syntheserapport, 1.5 Vervoer & Verkeer (Mira 2002), Vlaamse Milieumaatschappij, 2002.

[38] De Geest, C., De Vlieger, I., Cornelis, E., Int Panis, L., Schrooten, L., Logghe, S., Vanhove, F., De Ceuster, G., and De Geest, C., Milieurapport Vlaanderen 2004, Achtergrondocument, 1.6 Transport (Mira 2004), Vlaamse Milieumaatschappij, 2004.

[39] De Geest, C., De Vlieger, I., Cornelis, E., Int Panis, L., Schrooten, L., Govaerts, L., Pelkmans, L., Logghe, S., Vanherle, K. et al., Milieurapport Vlaanderen 2007, Achtergronddocument, Transport (Mira 2007), Vlaamse Milieumaatschappij, 2007.

[40] De Geest, C., Milieurapport Vlaanderen 2003, Syntheserapport 2.7 Lawaai (Mira 2003), Vlaamse Milieumaatschappij, 2003.

[41] De Geest, C., Milieurapport Vlaanderen 1999 (Mira 1999), Vlaamse Milieumaatschappij, 1999. [42] Batterman, S. A., Mattick, W. O., and Ranka, L. I., “Evaluating the Environmental Performance of

Passenger Vehicles,” International Journal of Life Cycle Assessment, vol. 6, no. 4, pp. 251-253, 2001.

Bibliography

p.219

[43] MacLean, H. L., and Lave, L. B., “Evaluating Automobile Fuel/Propulsion System Technologies,” Progress in Energy and Combustion Science, vol. 29, no. 1, pp. 1-69, 2003.

[44] Maclean, H. L., and Lave, L. B., “Life Cycle Assessment of Automobile/Fuel Options,” Environmental Science & Technology, vol. 37, no. 23, pp. 5445-5452, 2003.

[45] MacLean, H. L., Lave, L. B., Lankey, R., and Joshi, S., “A Life-Cycle Comparison of Alternative Automobile Fuels,” Journal of the Air & Waste Management Association, vol. 50, no. 10, pp. 1769-1779, 2000.

[46] Timmermans, J.-M., Matheys, J., Cappelle, J., Lataire, P., Van Mierlo, J., Maggetto, G., and Van den Bossche, P., “New Electric Postman Helper: From User Requirements to Technical Design Parameters,” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium, Yokohama, Japan, pp., 2006.

[47] Boureima, F.-S., Sergeant, N., Wynen, V., and Van Mierlo, J., Clean Vehicle Research: Lca and Policy Measures (Clever) - Report Task 1.2 Overview of Environmental Vehicle Assessments, Belgian Science Policy, Brussels, 2007.

[48] Matheys, J., Van Mierlo, J., Timmermans, J.-M., and Van den Bossche, P., “Life-Cycle Assessment of Batteries in the Context of the Eu Directive on End-of-Life Vehicles,” International Journal of Vehicle Design, vol. 46, no. 2, pp. 189-203, 2008, 2008.

[49] Aktualisierung Der Bewertungskriterien Der Vcd-Auto-Umweltliste, Institut für Energie und Umweltforschung Heidelberg Gmbh (IFEU), Bonn, 1997.

[50] Eisenman E.M.; Gasser, T. H., U.; Pulfer, M. ; Reutimann, F., Cleaner Drive, Hindernisse Für Die Markteinführung Von Neuen Fahrzeug-Generationen, Vereinigung Schweizerischer Verkehrsingenieure, 2004.

[51] Van Mierlo, J., Timmermans, J.-M., Maggetto, G., Van den Bossche, P., Meyer, S., Hecq, W., Govaerts, L., and Verlaak, J., “Environmental Rating of Vehicles with Different Alternative Fuels and Drive Trains: A Comparison of Two Approaches,” Transportation Research Part D - Transport and Environment, vol. 9, no. 5, pp. 387-399, 2004.

[52] Toussaint, Y., “Metropol, Un Outil D Evaluation De La Consommation Et Des Emission Des Vehicules Automobiles - Aplication Aux Vehicules Thermiques Et Electriques,” Universite de Liege, Belgium, 1999.

[53] Van Mierlo, J., Definitie Van Het Begrip Schone Voertuigen, ETEC-VUB / CEESE-ULB, Brussels, 2001.

[54] Verlaak, J., Govaerts, L., De Keukeleere, D., Timmermans, J.-M., Van Mierlo, J., Meyer, S., and Hecq, W., Ecoscore Rapport Taak 3: Verzamelen Gegevens En Integratie Emis Databank, 2005.

[55] Hendriksen, P., Evaluation of the Environmental Impact of Modern Passenger Cars on Petrol, Diesel, Automotive Lpg and Cng, 03.OR.VM.055.1/PHE, TNO, 2003.

[56] Ntziachristos, L. S., Z., Copert Iii, Computer Programme to Calculate Emissions from Road Transport, Methodology and Emission Factors, European Topic Centre on Air Emissions (LAT/AUTh), Copenhagen, Denmark, 2000.

[57] Pohl, H., “Vehicle Environmental Rating,” in Proceedings of the Fuel Cell, Hybrid and Battery Electric Vehicle Symposium, EVS20, Long Beach, CA, pp., 2003.

[58] Van Mierlo, J., Timmermans, J. M., Maggetto, G., Van den Bossche, P., Meyer, S., Hecq, W., Govaerts, L., and Verlaak, J., “Environmental Rating of Vehicles with Different Alternative Fuels and Drive Trains: A Comparison of Two Approaches,” Transportation Research Part D-Transport and Environment, vol. 9, no. 5, pp. 387-399, 2004.

[59] Choudhury, R., Wurster, R., and Weber, T., Gm Well-to-Wheel Analysis of Energy Use and Greenhouse Gas Emissions of Advanced Fuel/Vehicle Systems – a European Study, L-B- Systemtechnik GmbH, Germany, 2002.

[60] Ahlvik, P., Eggleston, S., Gorissen, N., Hassel, D., Hickman, A. J., Joumard, R., Ntziachristos, L., Rijkeboer, R., Samaras, Z. et al., Copert Ii Computer Programme to Calculate Emissions from Road Transport - Methodology and Emission Factors, European Environment Agency, Copenhagen, Denmark, 1997.

[61] Meet Methodology for Calculating Transport Emissions and Energy Consumption. [62] Tolouei, R., and Titheridge, H., “Vehicle Mass as a Determinant of Fuel Consumption and

Secondary Safety Performance,” Transportation Research Part D: Transport and Environment, vol. 14, no. 6, pp. 385-99, 2009.

Bibliography

p.220

[63] Van Mierlo, J., “Simulation Software for Comparison and Design of Electric, Hybrid Electric and Internal Combustion Vehicles with Resprect to Energy, Emissions and Performances,” Electrical Engineering and Energy Technology, Vrije Universiteit Brussel, Brussels, 2000.

[64] Van Mierlo, J., and Maggetto, G., “Vehicle Simulation Program: A Tool to Evaluate Hybrid Power Management Strategies Based on an Innovative Iteration Algorithm,” Proceedings of the Institution of Mechanical Engineers Part D-Journal of Automobile Engineering, vol. 215, no. D9, pp. 1043-1052, 2001.

[65] Van Mierlo, J., and Maggetto, G., “Innovative Iteration Algorithm for a Vehicle Simulation Program,” Ieee Transactions on Vehicular Technology, vol. 53, no. 2, pp. 401-412, 2004.

[66] Van Mierlo, J., Maggetto, G., van de Burgwal, E., and Gense, R., “Driving Style and Traffic Measures - Influence on Vehicle Emissions and Fuel Consumption,” Proceedings of the Institution of Mechanical Engineers Part D-Journal of Automobile Engineering, vol. 218, no. D1, pp. 43-50, 2004.

[67] website; "Ecodrive," (accessed on: 30/01/2010) Available at: http://www.ecodrive.org/. [68] Schexnayder, S. M., Das, S., Dhingra, R., Overly, J. G., Tonn, B. E., Peretz, J. H., Waidley, G., and

Davis, G. A., Environmental Evaluation of New Generation Vehicles and Vehicle Components, Office of Advanced Automotive Technologies, Office of Transportation Technologies, U.S. Department of Energy, 2001.

[69] Automotive Handbook, 2nd edition ed., Stuttgart, Germany: Robert Bosh GmbH, 1986. [70] Tires and Passenger Vehicle Fuel Economy, Informing Consumers, Improving Performance,

Transportation Research Board, Washington, D.C., 2006. [71] Glaeser, K.-P., "Procedures to Measure Tyre Rolling Resistance." [72] De Vlieger, I., Schrooten, L., Govaerts, L., Pelkmans, L., Van Mierlo, J., Timmermans, J.-M.,

Matheys, J., Macharis, C., Van Geirt, F. et al., Milieurapport Vlaanderen 2006, Achtergronddocument, 1.6 Transport, Vlaamse Milieumaatschappij, 2006.

[73] Murray, J., Lane, B., Lillie, K., and McCallum, J., The Report of the Alternative Fuels Group of the Cleaner Vehicle Task Force : An Assessment of the Emissions Performanceof Alternative and Conventional Fuels, Cleaner Vehicles Task Force, Great Britain, 2000.

[74] De Vlieger, I., Cornelis, E., Int Panis, L., Schrooten, L., Logghe, S., Vanhove, F., De Ceuster, G., and De Geest, C., Milieurapport Vlaanderen 2004, Achtergrondocument, 1.6 Transport, 2004.

[75] Groenboek: Op Weg Naar Een Europese Strategie Voor Een Continue Energievoorziening, European Commission, Brussels, 2000.

[76] Fulton, L., and Howes, T., Biofuels for Transport, an International Perspective, International Energy Agency, office of energy efficiency, 2004.

[77] Boureima, F.-S., Wynen, V., Sergeant, N., Messagie, M., Rombaut, H., and Van Mierlo, J., Bioses: Bio-Fuel Sustainable End Use, Task 2.1: Emission-Impact Well-to-Tank, Vrije Universiteit Brussel, Brussels, Belgium, 2009.

[78] De Ruyck, J., Lavric, D., Bram, S., Novak, A., Jossart, J. M., Remacle, M. S., Palmers, G., Dooms, G., and Hamelinck, C., Liquid Biofuels in Belgium in a Global Bio-Energy Context., Belgian Science Policy, SPSDII, 2005.

[79] Boerrigter, H., and van der Drift, B., “Biosyngas, Key-Intermediate in Production of Renewable Transportation Fuels, Chemicals, and Electricity: Optimum Scale and Economic Prospects of Fischer-Tropsch Plants,” in 14th European Biomass Conference & Exhibition, Paris, France, pp., 2005.

[80] Verhelst, S., and Wallner, T., “Hydrogen-Fueled Internal Combustion Engines,” Progress in Energy and Combustion Science, vol. 35, no. 6, pp. 490-527, 2009.

[81] Verhelst, S., and Sierens, R., “Aspects Concerning the Optimisation of a Hydrogen Fueled Engine,” International Journal of Hydrogen Energy, vol. 26, no. 9, pp. 981-985, 2001.

[82] Zegers, P., "Waterstof En Brandstofcellen." [83] Bogart, S. L., “Comparison of Investment and Related Requirements for Selected Hydrogen

Vehicle System Pathways,” Journal of Fusion Energy, vol. 21, no. 3-4, pp. 181-191, 2002. [84] Van den Bossche, P., “The Electric Vehicle: Raising the Standards,” Vakgroep Elektrotechniek en

Energietechnologie, Vrije Universiteit Brussel, Brussels, 2003. [85] Rinolfi, R., Bernard, L., and Ferrari, A., “What Powertrains for the Future?,” Fuels and

Powertrain, vol. 1808, pp. 365-376, 2003.

Bibliography

p.221

[86] Lenaers, G., “On-Board Real Life Emission Measurements on a 3 Way Catalyst Gasoline Car in Motor Way-, Rural- and City Traffic and on Two Euro-1 Diesel City Buses,” Science of the Total Environment, vol. 190, pp. 139-147, 1996.

[87] Sergeant, N., and Van Mierlo, J., Eco-Efficiency of the Brussels Mobility Based on the Ecoscore Methodology, Prospective Research for Brussels, 2009.

[88] Denys, T., and Govaerts, L., Clean Vehicle Research: Lca and Policy Measures, Report of Task 1.3 - Overview of Policy Measures, Belgian Science Policy, 2007.

[89] Decicco, J., and Kliesch, J., Rating the Environmental Impacts of Motor Vehicles: Aceee's Green Book Methodology, ACEEE, 2001.

[90] Decicco, J., and Kliesch, J., Rating the Environmental Impacts of Motor Vehicles: Aceee's Green Book Methodology, 2002 Edition, ACEEE, 2002.

[91] website, 2010; "Greenercars.Org | the Most Comprehensive and Scientific Environmental Vehicle Ratings Resource," (accessed on: 06/02/2010) Available at: http://www.greenercars.org/.

[92] Friedrich, R., and Bickel, P., Environmental External Costs of Transport, pp. 300: Springer Verlag, 2001.

[93] Egli, K., “So Sind Die Autos Bewertet,” VCS Magazin / Auto-Umweltliste 2009, vol. 1, pp. 61-63, 2009.

[94] website, 2010; "Vcs Verkehrs-Club Der Schweiz -Â Auto-Umweltliste," (accessed on: 05/02/2010) Available at: http://www.verkehrsclub.ch/de/politik-kampagnen/ratgeber/auto-umweltliste.html.

[95] website, 2010; "Volkswagen Group - Sustainability and Responsibility," (accessed on: 05/02/2010) Available at: http://www.volkswagenag.com/vwag/vwcorp/content/en/sustainability_and_responsibility.html.

[96] website, 2010; "Volvo Cars and the Environment 2008," (accessed on: 05/02/2010) Available at: http://www.volvocars.com/intl/top/about/news-events/pages/default.aspx?itemid=57.

[97] website, 2010; "Mercedes-Benz Deutschland - Feinstaubplakette - Umwelt-Zertifikat," (accessed on: 05/02/2010) Available at: http://www.mercedes-benz.de/content/germany/mpc/mpc_germany_website/de/home_mpc/passengercars/home/passengercars_world/news_and_events/particulate_matter_badge/environmental_certificate.html.

[98] website, 2010; "Toyota Environmental & Social Report 03 Development and Design," (accessed on: 05/02/2010) Available at: http://www.toyota.co.jp/en/environmental_rep/03/kaihatu07.html.

[99] Finkbeiner, M., Hoffmann, R., Ruhland, K., Liebhart, D., and Stark, B., “Application of Life Cycle Assessment for the Environmental Certificate of the Mercedes-Benz S-Class,” International Journal of Life Cycle Assessment, vol. 11, no. 4, pp. 240-246, 2006.

[100] Matheys, J., Timmermans, J.-M., Van Mierlo, J., Meyer, S., and Van den Bossche, P., “Comparison of the Environmental Impact of Five Electric Vehicle Battery Technologies Using Lca,” International Journal of Sustainable Manufacturing, vol. 1, no. 3, pp. 318-329, 2009.

[101] Matheys, J., Van Autenboer, W., Timmermans, J. M., Van Mierlo, J., Van den Bossche, P., and Maggetto, G., “Influence of Functional Unit on the Life Cycle Assessment of Traction Batteries,” International Journal of Life Cycle Assessment, vol. 12, no. 3, pp. 191-196, 2007.

[102] Schmidt, S., Ecotest, Testing and Assessment Protocol, ADAC Technik Zentrum and FIA Foundation, 2009.

[103] website, 2010; "Ecotest," (accessed on: 06/02/2010) Available at: http://www.ecotest.eu/Pages/Home.aspx.

[104] website, 2010; "Environmental Performance Label - California Air Resources Board," (accessed on: 06/02/2010) Available at: http://www.arb.ca.gov/msprog/labeling/labeling.htm.

[105] website, 2010; "Oeko-Trend Institut - Auto-Umwelt Zertifikat," (accessed on: 06/02/2010) Available at: http://www.oeko-trend.de/start/index.php?page=2&lang=en.

[106] website, 2010; "Environmental Protection Agency U.S. - Green Vehicle Guide," (accessed on: 06/02/2010) Available at: http://www.epa.gov/greenvehicles/Index.do.

[107] website, 2010; "2008 Automotive Environmental Index | J.D. Power," (accessed on: 06/02/2010) Available at: http://www.jdpower.com/autos/articles/2008-Automotive-Environmental-Index.

Bibliography

p.222

[108] website, 2010; "Portaal Belgische Overheid - Publicaties - Co2-Gids Van De Auto," (accessed on: 06/02/2010) Available at: http://www.belgium.be/nl/publicaties/publ_co2gids_2010.jsp?referer=tcm:117-29565-64.

[109] Van Mierlo, J., Meyer, S., and Hecq, W., Verslag Wp1 Definitie Van Het Begrip Schone Voertuigen, BIM-IGBE and Brussels Capital Region, Brussels, 2001.

[110] report, Environmentally Friendly Vehicle (Efv) Draft Feasibility Statement, ACEA, 2009. [111] van Wee, B., Janse, P., and Van Den Brink, R., “Comparing Energy Use and Environmental

Performance of Land Transport Modes,” Transport Reviews, vol. 25, no. 1, pp. 3-24, Jan, 2005. [112] van Wee, B., “Cost-Benefit Analysis and Health Impacts of Road Traffic Emission: An Overview

of Literature and a Categorisation of Policy Measures,” in 8th NECTAR Conference, Las Palmas G.C., pp., 2005.

[113] Van den Bossche, P., Vergels, F., Van Mierlo, J., Matheys, J., and Autenboer, W., “Subat: An Assessment of Sustainable Battery Technology,” Journal of Power Sources, vol. 162, no. 2, pp. 913-919, 2006.

[114] Edwards, R., Larive, J.-F., Mahieu, V., and Rouveirolles, P., Well-to-Wheels Analysis of Future Automotive Fuels and Powertrains in the European Context - Well-to-Tank Report, EUCAR, CONCAWE and JRC, 2007.

[115] website, 2010; "Febiac - Technicar," (accessed on: 06/02/2010) Available at: http://www.febiac.be/public/technicar.aspx?lang=NL.

[116] Joumard, R., Methods of Estimation of Atmospheric Emissions from Transport: European Scientist Network and Scientific State-of-the-Art, Institut National de Recherche sur les Transports et leur Sécurité (INRETS), France, 1999.

[117] Jolliet, O., Margni, M., Charles, R., Humbert, S., Payet, J., Rebitzer, G., and Rosenbaum, R., “Impact 2002+: A New Life Cycle Impact Assessment Methodology,” International Journal of Life Cycle Assessment, vol. 8, no. 6, pp. 324-330, 2003.

[118] Pershing, J., “Fossil Fuel Implications of Climate Change Mitigation Responses,” in Proceedings of an IPCC Expert Meeting, pp., 2000.

[119] Martens, W. J. M., “Climate Change, Thermal Stress and Mortality Changes,” Social Science & Medicine, vol. 46, no. 3, pp. 331-344, 1998.

[120] Ipcc Third Assessment Report - Climate Change 2001, pp. 873, New York, USA: University Press, 2003.

[121] Lindberg, G., “Recent Progress in the Measurement of External Costs and Implications for Transport Pricing Reforms,” European Journal of Transportation and Infrastructure Research, vol. vol.3, no. no.4, pp. pp. 387-400, 2003.

[122] Vermoote, S., and de Nocker, L., Valuation of Environmental Impacts of Acidification and Eutrophication Based on the Standard-Price Approach, DG Research, 5th framework programme, 2003.

[123] Edwards, R., Griesemann, J.-C., Larive, J.-F., and Mahieu, V., Well-to-Wheels Analysis of Future Automotive Fuels and Powertrains in the European Context, EUCAR, CONCAWE and JRC, 2004.

[124] Lenaers, G., Pelkmans, L., and Debal, P., “The Realisation of an on-Board Emission Measuring System Serving as a R&D Tool for Ultra Low Emitting Vehicles,” International Journal of Vehicle Design, vol. 31, no. 3, pp. 253-268, 2003.

[125] Eu Energy and Transport in Figures - Statistical Pocketbook 2009, Brussels: European Communities, 2009.

[126] Smokers, R., Gense, R., Rijkeboer, R., and Dijkhuizen, A., “Expert Judgement on the Long Term Possibilities of Conventional Emission Abatement Technologies for Passenger Cars,” Topics in Catalysis, vol. 30-1, no. 1-4, pp. 439-443, 2004.

[127] Borsellini, L., and Matheys, J., Autobus Propres, Coördinatie en Technische Studies MIVB/Coordination et Etudes Techniques STIB, 2006.

[128] Matheys, J., Timmermans, J.-M., Van Mierlo, J., and Maggetto, G., “Environmental Assessment of the Past, Present and Future Urban Bus Fleets, the Advantages of Battery, Trolley and Hybrid Electric Busses ” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 22), Yokohama, Japan, pp., 2006.

[129] website; "Vectrix Electrics," (accessed on: 17/12/2009) Available at: http://www.vectrix.com/. [130] De Borger, B., and Proost, S., Mobiliteit: De Juiste Prijs, pp. xviii+312, Leuven / Apeldoorn:

Garant, 1997.

Bibliography

p.223

[131] Klein, D., “La Bicyclette À Assistance Électrique,” REE - Revue de l'Electricite et de l'Electronique, no. 5, pp. 25-29, 2007.

[132] Jamerson, F. E., Electric Bikes Worldwide Reports 2009, Napels, Florida, 2009. [133] Cappelle, J., “Performance Analysis of Pedal Electric Cycles: An Objective and Subjective

Approach,” Departement of Electrical Engineering and Energy Technology, Vrije Universiteit Brussel, Brussels, Belgium, 2008.

[134] Wilson, D. G., Bicycle Science, Third Edition ed., Massachusetts: The MIT Press, 2004. [135] website; "Twike Home," (accessed on: 18/12/2009) Available at: http://www.twike.com/. [136] website; "Venturi Eclectic, Powered by Innovation," (accessed on: 19/12/2009) Available at:

http://www.venturieclectic.fr/#. [137] website; "Yamaha Motor," (accessed on: 19/12/2009) Available at: http://www.yamaha-

motor.co.jp/global/. [138] website; "E-Solex," (accessed on: 18/12/2009) Available at: http://www.e-solex.be/. [139] website; "Biketec - Willkommen. Flyer-Elektrovelos," (accessed on: 18/12/2009) Available at:

http://www.biketec.ch/. [140] website; "Dolphin Express at Extraenergy.Org," (accessed on: 18/12/2009) Available at:

www.ExtraEnergy.org. [141] website; "Sparta - Specialist in Speciale Fietsen," (accessed on: 18/12/2009) Available at:

http://www.sparta.nl/BE/. [142] website; "Estelle - Elektrorad Service," (accessed on: 18/12/2009) Available at:

http://www.elektrorad-service.de/estelle. [143] Katsuoka, T., Koyama, H., Nakamichi, S., and Ikuma, K., “Research and Development of Electro

Hybrid Bicycle with Pas (Power Assist System),” in Small Engine Technology Conference, Pisa, Italy, pp. 375-382.

[144] website, 2009; "Wegwijs in Het Belgisch Verkeersreglement," (accessed on: 28/12/2009) Available at: http://www.wegcode.be/index.php.

[145] Jamerson, F. E., Electric Bikes Worldwide Reports 2009, Napels, Florida. [146] Renoy, G., Histoire De La Poste, Brussels, Belgium: éditions Racine, 1999. [147] Lallement, P., Velocipede, U.S. patent no. 59,915, United States of America. [148] Renoy, G., Histoire De La Poste, Brussels, Belgium: éditions Racine, 1999. [149] website, 2009; "Pedal Cycles, a Victorian Postman on a Penny-Farthing Bicycle," (accessed on:

07/12/2009) Available at: http://postalheritage.org.uk/collections/museum/transport/pedalcycles.

[150] website, 2009; "The Bpma at Blists Hill, Shropshire," (accessed on: 07/12/2009) Available at: http://postalheritage.wordpress.com/2009/04/22/the-bpma-at-blists-hill-shropshire/.

[151] website, "Group of Cycling Postmen Ready to Start Their Round in the Early 1920s," line-of-cyclists-500.jpg, ed., The British Postal Museum & Archive, 1920.

[152] website; "Speedliner Mobility Gmbh," (accessed on: 15/12/2009) Available at: http://speedliner-mobility.com/bikes/.

[153] Oman, H., and Morchin, W. C., “Electric Bicycles for Commuting to Work,” Proceedings of the 27th Intersociety Energy Conversion Engineering Conference (IEEE Cat. No.92CH3164-1), pp. 3.179-83 vol.3|6 vol. 2870, 1992.

[154] Kumar, B., and Oman, H., “Power-Control for Battery-Electric Bicycles,” Proceedings of the Ieee 1993 National Aerospace and Electronics Conference, Naecon 1993, Vols 1 and 2, pp. 428-434, 1993.

[155] Oman, H., Morchin, W. C., and Jamerson, F. E., “Electric-Bicycle Propulsion Power,” Wescon 95 - Conference Record, pp. 555-560, 1995.

[156] Morchin, W. C., “Battery-Powered Electric Bicycles,” Northcon/94. Conference Record|Northcon/94. Conference Record, pp. 10.1109/NORTHC.1994.643354, 1994.

[157] Schwegler, U., “Electric Scooters: Technologies and Markets,” in EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium, Stavanger, Norway, pp.

[158] Ton, V., E-Tour - Electric Two-Wheelers on Urban Roads, Brussels, Belgium, 2003. [159] Lataire, P., Timmermans, J.-M., Maggetto, G., Van den Bossche, P., Meeusen, R., Van Gheluwe,

B., and Cappelle, J., “Electrically Assisted Bicycles: Demonstration, Characterisation, Health Benefit,” Revue E Tijdschrift, vol. 119 - 3, pp. 32-39, 2003.

[160] Poisson, P., “Fait Marquant - Le Vélo À Assistance Électrique - Un Vélo Jaune Tout Simplement,” REE - Revue de l'Electricite et de l'Electronique, no. 5, pp. 23-24, 2007.

Bibliography

p.224

[161] website; "De Post Neemt Als Eerste Europese Postoperator Elektrische Bromfietsen Voor Postbodes in Gebruik," (accessed on: 11/12/2009) Available at: http://www.post.be/site/nl/postgroup/press/releases/2006/20060901_scooters.html.

[162] Oxygen, "Technical Features Postscooter," Oxygen, 2003. [163] website; "Posteurop - Green Post Project," (accessed on: 10/12/2009) Available at:

http://www.posteurop.org/index.jsp?nodeId=22766&. [164] website, 2009; "Post Test 'Doodskist Op Wielen'," (accessed on: 15/12/2009) Available at:

http://www.standaard.be/artikel/detail.aspx?artikelid=DMF14122009_009&. [165] website, 2009; "Modec - Zero Emission Commercial Vehicle," (accessed on: 11/12/2009)

Available at: http://www.modeczev.com/index.asp. [166] website, 2009; "Smith Electric Vehicles," (accessed on: 11/12/2009) Available at:

http://www.smithelectricvehicles.com/. [167] Morchin, W. C., and Oman, H., Electric Bicycles: A Guide to Design and Use, Hoboken, New

Jersey: John Wiley & Sons, Inc., 2006. [168] Theurel, J., Offret, M., Gorgeon, C., and Lepers, R., “Physiological Stress Monitoring of Postmen

During Work,” WORK-A JOURNAL OF PREVENTION ASSESSMENT & REHABILITATION, vol. 31, no. 2, pp. 229-236, 2008, 2008.

[169] Pickering, A., “Last Mile - Pedal Power: With the Imminent Arrival of Eu Regulations to Reduce Co2 Emissions by 40 Percent by 2010, Isn't It Time to Think About Implementing Greener Delivery Methods?,” Postal Technology International, 2009.

[170] 2009; "Eureka | a Europe-Wide Network for Market Oriented R&D," (accessed on: 26th November 2009) Available at: http://www.eureka.be.

[171] Timmermans, J.-M., Nietvelt, J., Lataire, P., Van Mierlo, J., Matheys, J., and Van den Bossche, P., “New Electric Postman Helper - Sustainable Light Electric Vehicles for Postal Distribution,” in EET-2007 European Ele-Drive Conference, Brussels, pp. 7, 2007.

[172] Matis, P., “Decision Support System for Solving the Street Routing Problem,” Transport, vol. 23, no. 3, pp. 230-235, 2008.

[173] Lapalme, G., Rousseau, J. M., Chapleau, S., Cormier, M., Cossette, P., and Roy, S., “Georoute - a Geographic Information-System for Transportation Applications,” Communications of the Acm, vol. 35, no. 1, pp. 80-88, 1992.

[174] website, 2010; "Giro - Recognized Leader for Public Transit and Postal Distribution Software," (accessed on: 11/01/2010) Available at: http://www.giro.ca/en/.

[175] Irnich, S., “Solution of Real-World Postman Problems,” European Journal of Operational Research, vol. 190, no. 1, pp. 52-67, Oct, 2008.

[176] Amiri, M., Esfahanian, M., Hairi-Yazdi, M. R., and Esfahanian, V., “Minimization of Power Losses in Hybrid Electric Vehicles in View of the Prolonging of Battery Life,” Journal of Power Sources, vol. 190, no. 2, pp. 372-379, 2009.

[177] Khateeb, S. A., Farid, M. M., Selman, J. R., and Al-Hallaj, S., “Design and Simulation of a Lithium-Ion Battery with a Phase Change Material Thermal Management System for an Electric Scooter,” Journal of Power Sources, vol. 128, no. 2, pp. 292-307, 2004.

[178] website, 2010; "Saft Batteries," (accessed on: 28/02/2010) Available at: http://www.saftbatteries.com/.

[179] Taniguchi, A., Fujioka, N., Ikoma, M., and Ohta, A., “Development of Nickel/Metal-Hydride Batteries for Evs and Hevs,” Journal of Power Sources, vol. 100, no. 1-2, pp. 117-124, Nov, 2001.

[180] Morchin, W. C., and Oman, H., Electric Bicycles - a Guide to Design and Use, Hoboken, New Jersey: John Wiley & Sons, Inc., 2006.

[181] Takei, K., Ishihara, K., Kumai, K., Iwahori, T., Miyake, K., Nakatsu, T., Terada, N., and Arai, N., “Performance of Large-Scale Secondary Lithium Batteries for Electric Vehicles and Home-Use Load-Leveling Systems,” Journal of Power Sources, vol. 119, pp. 887-892, Jun, 2003.

[182] website, 2010; "Accell Pro - Dedicated B2b² Solutions," (accessed on: 25/02/2010) Available at: http://www.accell-pro.com/.

[183] website, 2010; "Clean Mobile A.G," (accessed on: 25/02/2010) Available at: http://www.4what.net/cleanmobile/.

[184] website, 2010; "Maxwell Technologies: Ultracapacitors Overview," (accessed on: 01/02/2010) Available at: http://www.maxwell.com/ultracapacitors/index.asp.

Bibliography

p.225

[185] Cappelle, J., Polfliet, N., Jacxsens, M., and Timmermans, J.-M., “Design of an E-Bike with Ultracaps as the Only Energy Source and with Regenerative Braking,” in EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium, Stavanger, Norway, pp. 7, 2009.

[186] Timmermans, J.-M., and Lataire, P., Verslag Testen Elektrische Fietsen - Test Aankoop, Vrije Universiteit Brussel, 2008.

[187] Timmermans, J.-M., Matheys, J., Lataire, P., Van Mierlo, J., and Cappelle, J., “A Comparative Study of 12 Electrically Assisted Bicycles,” in 24th International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 24), Stavanger, Norway, pp. 11, 2009.

[188] Gössel, K., Grüneberg, H., and Glöckner, M., “Contribution to Future Oriented Electric Drive Technology for Light Vehicles E.G. Bikes, Trikes, Wheel Chairs,” in 18th Battery, Hybrid and Fuel Cell Electric Vehicle Symposium, Berlin, Germany, pp., 2001.

[189] Hofer, K., “Linear Motor for Cylinders and Wheels,” ICEM 2000 Proceedings. International Conference on Electrical Machines|ICEM 2000 Proceedings. International Conference on Electrical Machines, vol. 1, no. 4, pp. 233-6 vol.1|4 vol. 2015, 2000.

[190] Lataire, P., "Electric Bicycles." [191] Timmermans, J.-M., Matheys, J., Lataire, P., Van Mierlo, J., and Cappelle, J., “A Comparative

Study of 12 Electrically Assisted Bicycles,” World Electric Vehicle Journal, eds: AVERE, published by WEVA, vol. 3, no. 1, pp. 11, 2009.

[192] website, 2010; "Energybus - Standard for Connecting Electric Components Fo Light Electric Vehicles," (accessed on: 01/02/2010) Available at: http://www.energybus.info/.

[193] website, 2010; "Villo! Bicycle Sharing System in Brussel," (accessed on: 03/03/2010) Available at: http://en.villo.be/.

[194] website, 2010; "Vélib' - Vélos En Libre-Service À Paris - Site Officiel - Paris," (accessed on: 01/02/2010) Available at: http://www.velib.paris.fr/.

[195] Cappelle, J., Timmermans, J.-M., Lataire, P., Maggetto, G., and Van den Bossche, P., “Electrically Assisted Cycling around the World ” in 20th International Electric Vehicle Symposium (EVS 20), Long Beach, California, pp., 2003.

[196] website, 2010; "Bike Tree - Secure Bicycle Schare & Storage System," (accessed on: 25/02/2010) Available at: http://www.biketree.com/.

[197] website, 2010; "Sanyo to Install “Solar Parking Lot” and Provide “Eneloop Bikes” in Setagaya, Tokyo," (accessed on: 01/02/2010) Available at: http://sanyo.com/news/2009/11/30-1.html.

[198] Lataire, P., Timmermans, J.-M., Maggetto, G., Van den Bossche, P., and Cappelle, J., “Electrically Assisted Bicycles,” International Journal of ARM, vol. 8, no. 1, pp. 6, 2007.

[199] Lataire, P., Timmermans, J.-M., Maggetto, G., Cappelle, J., and Van den Bossche, P., “Electrically Assisted Bicycles,” in 3rd COE Workshop on Human Adaptive Mechatronics (HAM), Tokyo, Japan, pp., 2006.

[200] Cappelle, J., Timmermans, J.-M., Lataire, P., and Maggetto, G., “A Personalised Testing Method for E-Pacs,” in 21st Worldwide Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 21), Monaco, pp., 2005.

[201] website, 2010; "Velocity Dolphin," (accessed on: 02/02/2010) Available at: http://www.velocity.ch/.

[202] website, 2010; "Electric Bikes from Circuitcycles," (accessed on: 02/01/2010) Available at: http://www.circuitcycles.ie/the-technology/.

[203] Timmermans, J.-M., Lataire, P., and Van Mierlo, J., Rapport Elektrisch Ondersteunde Fietsen - Nederlandse Consumentenbond, Vrije Universiteit Brussel, 2008.

[204] Fuchs, A., “Series Hybrid Drive-System: Advantages for Velomobiles,” Human Power eJournal, vol. 15, no. 5, 2008.

[205] website, 2010; "Ez-Wheel - Electric Wheel Solutions," (accessed on: 02/03/2010) Available at: http://www.ez-wheel.com/EN/ez-Wheel.html.

[206] website, 2010; "Heinzmann Gmbh & Co. Kg," (accessed on: 03/12/2009) Available at: http://www.heinzmann.de/.

[207] Glöckner, M., “Zur Motorisierung Von Kleinfahrzeugen Mit Radnabenmotoren,” Solarmobil Zeitschrift, vol. April 2007, no. 65, pp. p.17-21, 2007.

[208] Gössel, K., personal communication "Post - Epac - Bedarf an Batterieenergie Und Die Erforderliche Batteriekapazität," 12/07/2007, Schoünau, 2005,

[209] Compter, J. C., "Electric Drives for Precision Engineering Designs," SpecAmotor, 2007.

Bibliography

p.226

[210] Mohan, N., Undeland, T. M., and Robbins, W. P., Power Electronics: Converters, Applications, and Design, 2nd ed. ed., New York: Wiley, 1995.

[211] Timmermans, J.-M., Nietvelt, J., Lataire, P., Van Mierlo, J., Matheys, J., and Cappelle, J., “New Electric Postmen Helper - Development and Evaluation,” World Electric Vehicle Journal, published by Japan Automobile Research Institute (JARI), vol. 1, no. 1, 2008.

[212] Timmermans, J.-M., Nietvelt, J., Lataire, P., Van Mierlo, J., and Matheys, J., “New Electric Postmen Helper: Development and Evaluation,” in 23th International Electric Vehicle Symposium and Exposition (EVS 23), Anaheim, California, USA, pp., 2007.

[213] Timmermans, J.-M., Matheys, J., Cappelle, J., Lataire, P., Van Mierlo, J., Maggetto, G., and Van den Bossche, P., “New Electric Postman Helper: From User Requirements to Technical Design Parameters,” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 22), Yokohama, Japan, pp., 2006.

[214] Borgers, J., personal communication "Internal Tests with Prototypes Electric Bicycles for Postal Delivery at De Post / La Poste," Brussels, 2009,

Bibliography

p.227

Journal Publications [J1] Timmermans, J.-M., Matheys, J., Lataire, P., Van Mierlo, J., and Cappelle, J., “A Comparative

Study of 12 Electrically Assisted Bicycles,” World Electric Vehicle Journal, eds: AVERE, published by WEVA, vol. 3, no. 1, pp. 11, 2009.

[J2] Timmermans, J.-M., Nietvelt, J., Lataire, P., Van Mierlo, J., Matheys, J., and Cappelle, J., “New Electric Postmen Helper - Development and Evaluation,” World Electric Vehicle Journal, published by Japan Automobile Research Institute, vol. 1, no. 1, 2008.

[J3] Timmermans, J.-M., Matheys, J., Van Mierlo, J., and Lataire, P., “The Development and Applicatio of an Environmental Rating Tool to Stimulate the Use of Cleaner Road Vehicles,” WSEAS Transactions on Environment and Development, vol. 9, no. 2, pp. 7, 2006.

[J4] Timmermans, J.-M., Matheys, J., Van Mierlo, J., and Lataire, P., “Environmental Rating of Vehicles With Different Fuels and Drive Trains: a Univocal and Applicable Methodology,” European Journal of Transportation and Infrastructure Research, vol. 6, no. 4, pp. 21, 2006.

[J5] Matheys, J., Timmermans, J.-M., Van Mierlo, J., Meyer, S., and Van den Bossche, P., “Comparison of the environmental impact of five electric vehicle battery technologies using LCA,” International Journal of Sustainable Manufacturing, vol. 1, no. 3, pp. 318-329, 2009.

[J6] Matheys, J., Sergeant, N., Timmermans, J.-M., Boureima, F.-S., Wynen, V., and Van Mierlo, J., “Potential Reductions of CO2 emissions due to the landside accessibility of Brussels Airport through adapted policy measures and use of electric vehicles,” World Electric Vehicle Journal, vol. 3, pp. 14, 2009.

[J7] Matheys, J., Sergeant, N., Timmermans, J.-M., Boureima, F.-S., and Van Mierlo, J., “Improvement of the CO2 balance of the landside accessibility of Brussels Airport through implementation of electric vehicles and general policy measures,” Journal of Asian Electric Vehicles, vol. (accepted for publication), 2009.

[J8] Matheys, J., Boureima, F.-S., Sergeant, N., Timmermans, J.-M., and Van Mierlo, J., “The Influence of the "de minimis" clause of the European Emissions Tranding Scheme on CO2 emissions of Flights between Developing Countries and Belgium,” Transportation Research Part A, vol. (Reviewing), 2009.

[J9] Coosemans, T., Barrero, R., Timmermans, J.-M., Van Mulders, F., and Van Mierlo, J., “Data Acquisition System for Optimization of Series Hybrid Propulsion Systems,” World Electric Vehicle Journal, eds: AVERE, published by WEVA, vol. 3, no. 1, 2009.

[J10] Van Mulders, F., Timmermans, J.-M., McCaffree, Z., Van Mierlo, J., and Van den Bossche, P., “Supercapacitor Enhanced Battery Traction Systems - Concept Evaluation,” World Electric Vehicle Journal, vol. 2, no. 2, pp. 13, 2008.

[J11] Van den Bossche, P., Van Mulders, F., Van Mierlo, J., and Timmermans, J.-M., “The Evolving Standardization Landscape for Electrically Proppelled Vehicles,” World Electric Vehicle Journal, published by WEVA, vol. 2, 2008.

[J12] Matheys, J., Van Mierlo, J., Timmermans, J.-M., and Van den Bossche, P., “Life-cycle assessment of batteries in the context of the EU Directive on end-of-life vehicles,” International Journal of Vehicle Design, vol. 46, no. 2, pp. 189-203, 2008.

[J13] Matheys, J., Van Mierlo, J., Timmermans, J.-M., and Van den Bossche, P., “LCA of Batteries in the Context of the EU Directive on End-of-Life Vehicles,” International Design and Development, vol. 46, no. 2, pp. 14, 2008.

[J14] Van den Bossche, P., Van Mierlo, J., Timmermans, J.-M., Matheys, J., Maggetto, G., and Vergels, F., “Evolutions in Hydrogen and Fuel Cell Standardisation: The HarmonHy

Journal Publications

p.228

Experience,” World Electric Vehicle Journal, published by Japan Automobile Research Institute, vol. 1, no. 1, pp. 6, 2007.

[J15] Van Mierlo, J., Timmermans, J.-M., Maggetto, G., and Van den Bossche, P., “Peak Power Based Fuel Cell Hybrid Propulsion System,” World Electric Vehicle Journal, published by Japan Automobile Research Institute, vol. 1, no. 1, pp. 7, 2007.

[J16] Matheys, J., Van Autenboer, W., Timmermans, J. M., Van Mierlo, J., Van den Bossche, P., and Maggetto, G., “Influence of functional unit on the life cycle assessment of traction batteries,” International Journal of Life Cycle Assessment, vol. 12, no. 3, pp. 191-196, 2007.

[J17] Lataire, P., Timmermans, J.-M., Maggetto, G., Van den Bossche, P., and Cappelle, J., “Electrically assisted bicycles,” International Journal of ARM, vol. 8, no. 1, pp. 6, 2007.

[J18] Van Mierlo, J., Timmermans, J.-M., Maggetto, G., Van den Bossche, P., Meyer, S., Hecq, W., Govaerts, L., and Verlaak, J., “Environmental rating of vehicles with different alternative fuels and drive trains: a comparison of two approaches,” Transportation Research Part D - Transport and Environment, vol. 9, no. 5, pp. 387-399, 2004.

[J19] Lataire, P., Timmermans, J.-M., Maggetto, G., Van den Bossche, P., Meeusen, R., Kempenaers, F., Van Gheluwe, B., and Cappelle, J., “Electrically assisted bicycles: demonstration, characterisation, health benefit,” Revue E Tijdschrift, vol. 119e jaargang, no. 3, pp. 32-39, 2003.

Journal Publications

p.229

Conference papers [C1] Timmermans, J.-M., Matheys, J., Lataire, P., Van Mierlo, J., and Cappelle, J., “A Comparative

Study of 12 Electrically Assisted Bicycles,” in 24th International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 24), Stavanger, Norway, pp. 11, 2009.

[C2] Timmermans, J.-M., Culcu, H., Coosemans, T., Van Mierlo, J., and Lataire, P., “Test platform for hybrid electric power systems: use for design project for students,” in 13th European Conference on Power Electronics and Applications (EPE 2009), Barcelona, Spain, pp. 7, 2009.

[C3] Timmermans, J.-M., Van Mierlo, J., Lataire, P., Van Mulders, F., and Mccaffree, Z., “Test platform for hybrid electric power systems: development of a HIL Test Platform,” in 12th European Conference on Power Electronics and Applications (EPE 2007), Aalborg, Denmark, pp. 1-7, 2007.

[C4] Timmermans, J.-M., Van Mierlo, J., Lataire, P., Van den Bossche, P., and Van Mulders, F., “Test Platform for Hybrid Electric Power Systems for Use in Future Dual Use Applications: Development of a HIL Test Platform,” in 7th International All Electric Combat Vehicle Conference (AECV 2007), Stockholm, Sweden, pp., 2007.

[C5] Timmermans, J.-M., Nietvelt, J., Lataire, P., Van Mierlo, J., Matheys, J., and Van den Bossche, P., “New Electric Postman Helper - Sustainable Light Electric Vehicles for Postal Distribution,” in EET-2007 European ELE-Drive Conference, Brussels, Belgium, pp. 7, 2007.

[C6] Timmermans, J.-M., Nietvelt, J., Lataire, P., Van Mierlo, J., and Matheys, J., “New Electric Postmen Helper: Development and Evaluation,” in 23th International Electric Vehicle Symposium and Exposition (EVS 23), Anaheim, California, USA, pp., 2007.

[C7] Timmermans, J.-M., Zadora, P., Cheng, Y., Van Mierlo, J., and Lataire, P., “Modelling and design of super capacitors as peak power unit for hybrid electric vehicles,” in 2005 IEEE Vehicle Power and Propulsion Conference pp. 8, 2006.

[C8] Timmermans, J.-M., Matheys, J., Van Mierlo, J., and Lataire, P., “Ecoscore, an Environmental Rating Tool for Road Vehicles,” in 4th WSEAS International Conference on Environment, Ecosystems and Development (EED'06), Venice, Italy, pp., 2006.

[C9] Timmermans, J.-M., Matheys, J., Cappelle, J., Lataire, P., Van Mierlo, J., Maggetto, G., and Van den Bossche, P., “New Electric Postman Helper: From User Requirements to Technical Design Parameters,” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 22), Yokohama, Japan, pp., 2006.

[C10] Timmermans, J.-M., Zadora, P., Van Mierlo, J., Cheng, Y., and Lataire, P., “Modelling and Design of Super Capacitors as Peak Power Unit for Hybrid Electric Vehicles,” in IEEE Vehicle Power and Propulsion Conference (VPPC 2005), Illinois, Chicago, pp. 701-708, 2005.

[C11] Timmermans, J.-M., Van Mierlo, J., Matheys, J., and Maggetto, G., “The Development of an Environmental Rating Tool for Application in Policy Measures,” in 21st Worldwide Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 21), Monaco, pp., 2005.

[C12] Timmermans, J.-M., Cheng, Y., Van den Bossche, P., Van Mierlo, J., and Lataire, P., “Super Capacitors as Peak Power Unit for Hybrid Electric Vehicles in Future Dual-use Applications,” in 6th International All Electric Combat Vehicle Conference (AECV 2005), Bath, United Kingdom, pp., 2005.

[C13] Timmermans, J.-M., Van Mierlo, J., Maggetto, G., Meyer, S., Hecq, W., Govaerts, L., and Verlaak, J., “The comparison of two environmental rating systems: BIM-Ecoscore vs. EC-Cleaner Drive,” in EET 2004 European ELE-Drive Transportation Conference, Estoril, Portugal, pp., 2004.

Conference papers

p.230

[C14] Wynen, V., Timmermans, J.-M., Sergeant, N., Boureima, F.-S., and Van Mierlo, J., “Evaluation of the potential environmental impact reduction from passenger travel through adapted policy measures and the use of integrated sustainable Multi-modal transport modes,” in BIVEC-GIBET Transport Research Day, Brussels, pp. pp:821-835, 2009.

[C15] Sergeant, N., Boureima, F.-S., Matheys, J., Timmermans, J.-M., and Van Mierlo, J., “An Environmental analysis of FCEV and H2-ICE vehicles using the Ecoscore methodology,” in 24th International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 24), Stavanger, Norway, pp., 2009.

[C16] Rahier, H., Bertoia, C., Hubin, A., Timmermans, J.-M., Tiberghien, J., Wastiels, J., Van Mierlo, J., Vanherzeele, H., and Van Biesen, L., “Multidisciplinary technology projects in the second year of bachelor in engineering at the Vrije Universiteit Brussel,” in SEFI 37th annual conference 2009 (SEFI 2009), Rotterdam, The Netherlands, pp., 2009.

[C17] Matheys, J., Timmermans, J.-M., Sergeant, N., Rombaut, H., Boureima, F.-S., and Van Mierlo, J., “The influence of a "de minimis" clause on the CO2 emissions of developing countries" flights to or from the European Union - A Belgian case-study,” in Environment and Transport in different contexts, Ghardaïa, Algeria, pp., 2009.

[C18] Matheys, J., Sergeant, N., Timmermans, J.-M., Boureima, F.-S., Wynen, V., and Van Mierlo, J., “Potential reductions of CO2 emissions due to the landside accessibility of Brussels Airport through adapted policy measures and use of electric vehicles,” in 24th International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 24), Stavanger, Norway, pp., 2009.

[C19] Coosemans, T., Barrero, R., Timmermans, J.-M., Van Mulders, F., and Van Mierlo, J., “Data acquisition system for optimization of series hybrid propulsion systems,” in 24th International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 24), Stavanger, Norway, pp., 2009.

[C20] Cappelle, J., Polfliet, N., Jacxsens, M., and Timmermans, J.-M., “Design of an e-bike with UltraCaps as the only energy source and with regenerative braking,” in EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium, Stavanger, Norway, pp. 7, 2009.

[C21] Van Mulders, F., Timmermans, J.-M., Van Mierlo, J., and Van den Bossche, P., “A basic model for evaluating a direct battery and supercapacitor parallel connection,” in EET-2008 European ELE-Drive Conference, Geneva, pp., 2008.

[C22] Van den Bossche, P., Van Mulders, F., Van Mierlo, J., and Timmermans, J.-M., “Electric vehicle standardization: conflict, collaboration and cohesion,” in EET-2008 European ELE-Drive Conference, Geneva, pp., 2008.

[C23] Sergeant, N., Matheys, J., Timmermans, J.-M., Rombaut, H., and Van Mierlo, J., “The influence of potential policy measures on the eco-efficiency of personal vehicle mobility in Brussels,” in Urban Transport Conference Proceedings, pp., 2008.

[C24] Matheys, J., Rogolle, C., Sergeant, N., Timmermans, J.-M., Rombaut, H., and Van Mierlo, J., “Analysis and Improvement of "The Last Mile" to and from the national airport as part of the mobility policy in the Brussels urban area,” in Conference on Urban Transport, Malta, pp., 2008.

[C25] Van Mulders, F., Van Mierlo, J., Van den Bossche, P., Timmermans, J.-M., and McCaffree, Z., “Supercapacitor enhanced battery traction systems - concept evaluation,” in 23th International Electric Vehicle Symposium and Exposition (EVS 23), Anaheim, California, USA, pp., 2007.

[C26] Van Mulders, F., Timmermans, J.-M., McCaffree, Z., Van Mierlo, J., and Van den Bossche, P., “Preliminary evaluation of the use of supercapacitors in a battery powered electric go-kart,” in EET-2007 European ELE-Drive Conference, Brussels, Belgium, pp., 2007.

[C27] Van Mierlo, J., Timmermans, J.-M., and Van den Bossche, P., “A novel educative interface based on a vehicle simulation tool for hybrid propulstion system assessement,” in 23th International Electric Vehicle Symposium and Exposition (EVS 23), Anaheim, California, USA, pp., 2007.

[C28] Van den Bossche, P., Van Mulders, F., Van Mierlo, J., Cheng, Y., and Timmermans, J.-M., “Evolution of international standardization of electrically propelled vehicles,” in EET-2007 European ELE-Drive Conference, Brussels, Belgium, pp., 2007.

[C29] Sergeant, N., Van Mierlo, J., Timmermans, J.-M., and Matheys, J., “The development of an LCA tool for vehicles with conventional and alternative fuels and drive trains,” in EET-2007 European ELE-Drive Conference, Brussels, Belgium, pp., 2007.

Conference papers

p.231

[C30] Sergeant, N., Van Mierlo, J., Matheys, J., and Timmermans, J.-M., “An LCA tool for conventional and alternative vehicles,” in 23th International Electric Vehicle Symposium and Exposition (EVS 23), Anaheim, California, USA, pp., 2007.

[C31] Matheys, J., Festraets, T., Van Mierlo, J., Macharis, C., Sergeant, N., and Timmermans, J.-M., “Aviation and climate change: a comparison of the overflights of the Belgian territory and the local avaiation activities,” in 1st CEAS European Air and Space Conference, Berlin, Germany, pp., 2007.

[C32] Matheys, J., Festraets, T., Timmermans, J.-M., Sergeant, N., and Van Mierlo, J., “Alternative road vehicles, electric rail systems, short flights: an environmental comparison,” in 23th International Electric Vehicle Symposium and Exposition (EVS 23), Anaheim, California, USA, pp., 2007.

[C33] Van Mierlo, J., Cheng, Y., Timmermans, J.-M., and Van den Bossche, P., “Comparison of fuel cell hybrid propulsion topologies with super-capacitor,” in 12th International Power Electronics & Motion Control Conference, pp. 501-5, 2006.

[C34] Van Mierlo, J., Cheng, Y., Timmermans, J.-M., Maggetto, G., and Van den Bossche, P., “Peak Power Based Fuel Cell Hybrid Propulsion System,” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 22), Yokohama, Japan, pp., 2006.

[C35] Van den Bossche, P., Vreven, K., Van Mierlo, J., Timmermans, J.-M., Cheng, Y., and Maggetto, G., “Comparative modeling of traction batteries in view of electric vehicle applications,” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 22), Yokohama, Japan, pp., 2006.

[C36] Van den Bossche, P., Van Mierlo, J., Timmermans, J.-M., Matheys, J., Maggetto, G., and Vergels, F., “Evolutions in Hydrogen and Fuel Cell Standardization: the HarmonHy experience,” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 22), Yokohama, Japan, pp., 2006.

[C37] Van Mierlo, J., Cheng, Y., Timmermans, J.-M., and Van den Bossche, P., “Comparison of Fuel Cell Hybrid Propulsion Topologies with Supercapacitor,” in 12th International Power Electronic and Motion Conference (EPE-PEMC 2006), Portoroz, Slovenia, pp., 2006.

[C38] Matheys, J., Timmermans, J.-M., Van Mierlo, J., and Maggetto, G., “Environmental Assessment of the Past, Present and Future Urban Bus Fleets, the Advantages of Battery, Trolley and Hybrid Electric Busses ” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 22), Yokohama, Japan, pp., 2006.

[C39] Matheys, J., Timmermans, J.-M., Van Autenboer, W., Van Mierlo, J., Maggetto, G., Meyer, S., De Groof, A., Hecq, W., and Van den Bossche, P., “Environmental Impact of Electric Vehicle Traction Batteries: A comparison using LCA,” in 2nd Conference Environment & Transport - Transport and Air Pollution, Reims, France, pp., 2006.

[C40] Matheys, J., Timmermans, J.-M., Van Autenboer, W., Van Mierlo, J., Maggetto, G., Meyer, S., De Groof, A., Hecq, W., and Van den Bossche, P., “Comparison of the Environmental impact of 5 Electric Vehicle Battery Technologies using LCA,” in 13th CIRP International Conference on Life Cycle Engineering (LCE 2006), Leuven, Belgium, pp., 2006.

[C41] Lataire, P., Timmermans, J.-M., Maggetto, G., Cappelle, J., and Van den Bossche, P., “Electrically assisted bicycles,” in 3rd COE Workshop on Human Adaptive Mechatronics (HAM), Tokyo, Japan, pp., 2006.

[C42] Cappelle, J., Timmermans, J.-M., Lataire, P., Van Mierlo, J., and Maggetto, G., “The Pedelec Market in Flanders: The Position of the Bicycle Dealers,” in 22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 22), Yokohama, Japan, pp., 2006.

[C43] Van Mierlo, J., Timmermans, J.-M., Lataire, P., and Van den Bossche, P., “Project oriented education: build your own electric go-kart,” in 11th European Conference on Power Electronics and Applications (EPE 2005) Dresden, Germany, pp. 9, 2005.

[C44] Matheys, J., Van Autenboer, W., Van Mierlo, J., Timmermans, J.-M., Maggetto, G., Meyer, S., Arnaud, D. G., Hecq, W., and Van den Bossche, P., “Comparative Sustainability Assessment of Different Electric Vehicle Traction Batteries Using LCA,” in 21st Worldwide Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 21), Monaco, pp., 2005.

[C45] Cappelle, J., Timmermans, J.-M., Lataire, P., and Maggetto, G., “A personalised testing method for E-PACs,” in 21st Worldwide Battery, Hybrid and Fuel Cell Electric Vehicle Symposium (EVS 21), Monaco, pp., 2005.

Conference papers

p.232

[C46] Cappelle, J., Timmermans, J.-M., Lataire, P., Maggetto, G., and Van den Bossche, P., “Electrically Assisted Cycling around the World ” in 20th International Electric Vehicle Symposium (EVS 20), Long Beach, California, pp., 2003.

Conference papers

p.233

Notes

Notes

p.234 Notes

p.235 Notes

p.236 Notes

p.237

Notes

p.238

Conversions (vehicle) speed 1 mile per hour = 1,609 kilometre per hour 1 kilometre per hour = 0,621 miles per hour 1 meter per second = 3,6 kilometre per hour 1 kilometre per hour = 1 metre per second length / distance 1 inch = 2,54 centimetre 1 centimetre = 0,394 inch 1 mile = 1,609 kilometre 1 kilometre = 0,622 mile energy 1 kilowatt hour = 3600 kilojoule 1 joule = 1 wattsecond volume 1 gallon (US, liquid) = 3,785 litre 1 litre = 0,264 gallon (US, liquid) Via: http://www.onlineconversion.com

Conversions

a contribution to cleaner vehicle technologies final version for distribution

Documents