Electric Vehicles: Technology, Research and Development [1 ed.] 9781617283901, 9781607411420

Citation preview

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ELECTRIC VEHICLES: TECHNOLOGY, RESEARCH AND DEVELOPMENT

No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, or any other professional Electric Vehicles: Technology, Research andmedical Development, Nova Science Publishers,services. Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

ELECTRIC VEHICLES: TECHNOLOGY, RESEARCH AND DEVELOPMENT

GERALD B. RAINES

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

EDITOR

Nova Science Publishers, Inc. New York

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009 by Nova Science Publishers, Inc. All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Any parts of this book based on government reports are so indicated and copyright is claimed for those parts to the extent applicable to compilations of such works.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA Electric vehicles : technology, research, and development / editor, Gerald B. Raines. p. cm. Includes index. ISBN  (HERRN) 1. Electric vehicles. I. Raines, Gerald B. TL220.E4545 2009 629.22'93--dc22 2009000156

Published by Nova Science Publishers, Inc. Ô New York

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

CONTENTS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Preface

vii

Chapter 1

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles K. T. Chau

Chapter 2

Environmental Friendly Inter-City Aircraft (ENFICA-FC) and Preliminary Analysis for 2-seat Aircraft Conversion into Fuel Cells Powered Innovative System G. Romeo, G. Frulla, E. Cestino and F. Borello

1

41

Chapter 3

Clarifying the Debate on Electric Two-Wheelers in China Christopher Cherry

Chapter 4

Examination of State Estimators for Electrochemical Energy Storage Devices by Means of a Hardware-in-the-Loop System Mark Verbrugge, Damon Frisch, Trudy Weber, Arthur Bekaryan and Ping Liu

Chapter 5

On The Fuel Economy Of Hybrid-Electric Powertrains Tomaž Katrašnik

Chapter 6

Energy Efficiency Policy: Budget, Electricity Conservation andFuel Conservation Issues Fred Sissine

163

Chapter 7

Hydrogen and Fuel Cell Vehicle R&D: FreedomCAR and the President’s Hydrogen Fuel Initiative Brent D. Yacobucci

195

Chapter 8

Department Of Transportation, National Highway Trafffic Safety Administration: Federal Motor Vehicle Safety Standards; Occupant Protection In Interior Impact; Side Impact Protection; Fuel System Integrity; Electric-Powered Vehicles: Electrolyte Spillage And Electrical Shock Protection; Side Impact Phase-In Reporting Requirements U. S. Government Accountability Office

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

67

87

105

201

vi

Contents

Chapter 9

Advanced Lithium-Ion Batteries for Plug-in Hybrid-Electric Vehicles Paul Nelson and Khalil Amine

205

Chapter 10

Energy Independence and Security Act of 2007: A Summary of Major Provisions Fred Sissine

221

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Index

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

243

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

PREFACE With ever increasing concern on environmental protection and energy conservation, there is a fast growing interest in electric vehicles (EVs) from automakers, governments and customers. As electric propulsion is the core of EVs, there is a pressing need for researchers to develop advanced electric motor drives for various classes of EVs, including the battery, hybrid and fuel cell vehicles. Such issues are addressed in this book. The development and use of a fuel cell based power system for propulsion of electric aircrafts is discussed. A study done on the flight mechanics of the new aircraft, to verify the new flight performance, is also examined. Electric powered two-wheelers have risen in popularity in China over the past several years. This book investigates the growth of these electric two-wheelers in China and compares their environmental and safety impacts to those of alternative modes of transportation. Futhermore, the design and implementation of a hardware-in-the-loop system for the development, verification, and validation of algorithms used to construct state estimators for batteries and supercapacitors is addressed. There are several different kinds of devices that can be used to achieve electrochemical energy conversion. Some of these conversion technologies are reviewed, as well as their impact on the environment. The method used to control a power-train of a hybrid electric vehicle is discussed as well as how both the engine and the electric machine may achieve respective higher efficiencies after using this method. The regulated and unregulated emissions of diesel engines operating on different sulfur content fuels are also looked at. Energy efficiency issues include research and development priorities, funding for climate-related efficiency programs, implementation of equipment efficiency standards, regulation of vehicle fuel efficiency, and electricity industry ratemaking for energy efficiency profitability. Such issues are addressed in this book. Chapter 1 - With ever increasing concern on environmental protection and energy conservation, there is a fast growing interest in electric vehicles (EVs) from automakers, governments and customers. As electric propulsion is the core of EVs, it is a pressing need for researchers to develop advanced electric motor drives for various classes of EVs, including the battery, hybrid and fuel cell vehicles. This chapter firstly reviews various electric motor drives, including DC, induction, switched reluctance and permanent magnet (PM) brushless

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

viii

Gerald B. Raines

types, for application to EVs. As the development of PM brushless motor drives is being actively engaged, their motor configurations and operating principles are delineated. Secondly, two emerging PM brushless motor drives, namely the double-stator PM brushless motor drive and the hybrid-PM brushless motor drive, are discussed. Their motor configurations, design equations, analysis approaches, converter topologies and control strategies are described when necessary. Then, their operating characteristics are evaluated. Finally, the concept of integrated PM motor drive systems is revealed. It takes the definite advantages of high efficiency, good controllability, compact size and lightweight, which are particularly essential for future vehicles. Hence, three latest integrated PM brushless motor drive systems, namely the magnetic-geared PM brushless motor drive system, the memoryPM brushless integrated-starter-generator system and the PM brushless electric variable transmission system, are discussed. Future research directions of these integrated systems are particularly identified. Chapter 2 - The main objective of the ENFICA-FC project (ENvironmentally Friendly Inter City Aircraft powered by Fuel Cells), funded by European Commission, is to develop and validate the use of a fuel cell based power system for propulsion of more/all electric aircraft. The following items shall be pursued: a) A fuel cell system shall be designed, built and tested in laboratory ready to be installed on board for flying; b) A high efficiency brushless electric motors and power electronics apparatus for their control shall be designed and manufactured ready to be installed on board for flying; c) high efficiency would be obtained by an optimised aerodynamic propeller design; d) A study of the flight mechanics of the new aircraft will be carried to verify the new flight performance; e) Validation of the overall high performance of an all electric aircraft by means of flight test. The fuel cell system will be installed in a light sport aircraft which will be flight and performance tested as a proof of functionality and future applicability for inter city aircraft. A selection of most suitable aircraft for conversion is presented and the light sport aircraft Rapid 200 chosen. The high efficiency two-seat existing aircraft Rapid 200, manufactured by Jihlavan Aircraft, has been selected over more than 100 light sport aircrafts after the preliminary reported evaluation based on merit index as indicated. The aircraft will be used for the conversion from internal combustion engine to an electric one. Analysis about the COTS equipments for electric propulsion system has been performed and presented. Design indication of an optimal propeller complete the identification phase that continue with the analysis of some parameters influencing the general configuration of the converted aircraft and the mission items. Preliminary consideration about the definition of storage configuration are presented and some safety issues are considered for H2-gas management. Design indications and conversion limitations conclude the reported activity. Chapter 3 - Electric two-wheelers have entered the ever-expanding transportation market in China, fully penetrating the market in most cities and rural areas. Over the past five years, more than 50 million electric two-wheelers have been sold in China. Electric two-wheelers have surpassed bicycle ridership in many cities, moving this mode beyond filling small niches that electric vehicles have historically filled in developed nations. Electric two-wheelers are used for many types of transportation, from commuting to goods delivery to law enforcement. This chapter outlines the recent historic growth of electric two-wheelers and scooters in China, discussing several of the controversial issues surrounding them, including safety, congestion, environmental performance, and competition with public transit. Electric twowheelers provide high levels of personal mobility, matching that of the automobile. They do

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Preface

ix

so with much fewer of the environmental, congestion, and safety externalities associated with personal cars. Currently, though, most electric two-wheeler users would otherwise ride a traditional bicycle or bus, obscuring the relative impact of electric two-wheelers on China’s urban transportation systems. When balancing electric two-wheelers’ effect on mobility and access, environment, safety, and congestion in the entire transportation system, they debatably provide net benefits. This chapter clarifies those debates and frames the challenges in ways that can be quantified. Chapter 4 - We have designed and implemented a hardware-in-the-loop system for the development, verification, and validation of algorithms used to construct state estimators for batteries and supercapacitors. The intent of the work is to allow algorithm developers to test their algorithms rapidly and in the context of the actual application. Promising results are shown for a carbon based electric double layer capacitor (a supercapacitor); the modeled vehicle configuration corresponds to the General Motors Saturn Vue Green Line 36V HEV, with the NiMH battery replaced by a supercapacitor pack. While the work is pragmatic in terms of addressing vehicle applications, the theoretical underpinnings and mathematical methods are relatively new, evolving, and comprise substantial complexity. Chapter 5 - The treatise presents an extensive simulation and analytical analysis of the energy conversion phenomena in parallel and series hybrid-electric powertrains. Parameters of both hybrid powertrains are evaluated and compared to parameters of the conventionalinternal combustion engine powertrain. Simulation approach is based on an accurate and fast forward-facing simulation model that is capable of capturing dynamics of the powertrain components. Moreover, the treatise offers an analytical approach based on the energy balance equations in order to analyze and predict energy conversion efficiency in both hybrid powertrains. The analysis covers broad range of parallel and series hybrid powertrain configurations. Very good agreement between simulation and analytical results gives confidence in the accuracy of the performed analysis and confirms the validity of the analytical framework. Combined simulation and analytical analysis enables deep insight into energy conversion phenomena in hybrid powertrains. It reveals advantages and disadvantages of both hybrid powertrain concepts and their variations running under different operating conditions. The analysis thus indicates guidelines that lead to optimum fuel economy of particular powertrain concept operating according to the specified drive-test cycle. It can be concluded from the presented results that: 1.) parallel hybrid powertrain features better fuel economy than the series one for the applied test cycles, 2.) both hybrid powertrain configurations feature the best fuel economy at light duty application and 3.) electric conversion efficiency has significant influence on the fuel economy enhancement of hybridelectric powertrains. Chapter 6 - Energy efficiency issues include research and development (RandD) priorities, funding for climate-related efficiency programs, implementation of equipment efficiency standards, regulation of vehicle fuel efficiency, and electricity industry ratemaking for energy efficiency profitability. The Bush Administration has proposed an Advanced Energy Initiative (AEI) to accelerate hydrogen programs. For the Department of Energy’s (DOE’s) energy efficiency RandD programs, the Administration seeks $484.7 million, with increases for Hydrogen and Hybrid/Electric Propulsion. The request would cut $74.8 million from the Weatherization Program and eliminate controversial funding earmarks. The Housepassed version of the FY2007 Energy and Water Appropriations Bill (H.R. 5427) would fund

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

x

Gerald B. Raines

AEI and cut earmarks. The Senate Appropriations Committee has also approved AEI funding and would cut earmarks even further than the House. Energy efficiency programs have long been justified for the ability to reduce petroleum use and curb environmental impacts such as air pollution. This made it economically and administratively convenient to have them also serve as part of a low cost “no regrets” policy to reduce greenhouse gas (especially CO2) emissions. In addition to DOE funding, H.R. 5386 would provide about $100 million for the Environmental Protection Agency’s energy efficiency program, and the Senate Appropriations Committee’s version of H.R. 5522 would provide about $200 million for energy efficiency-related programs in developing countries. DOE’s implementation of equipment efficiency standards has been a subject of some congressional criticism. The Energy Policy Act of 2005 (EPACT, P.L. 109-58) directed DOE to report to Congress on actions taken to address the concern. In response, DOE issued a schedule for rulemakings on 30 products. EPACT also raised the goals for energy efficiency in federal agencies and provided modest tax incentives for efficiency in certain vehicles and buildings. Automobile fuel efficiency regulation has been one of the most controversial aspects of energy efficiency policy. The Corporate Average Fuel Economy (CAFE) program for new cars and light trucks achieved significant energy savings through 1985 but has remained relatively flat since then. Critics say that recent CAFE increases for light trucks are too small, given concerns about high gasoline prices, air emissions. Proponents counter that larger CAFE increases would pollution, and CO2 compromise safety and cause hardship for manufacturers. The National Action Plan for Energy Efficiency aims to defer the need for 20,000 megawatts of new electric power plant capacity. Its success will depend mainly on the ability of state regulators to make energy efficiency profitable for electricity companies, by addressing the link between profits and sales. Chapter 7 - FreedomCAR and the Hydrogen Fuel Initiative are two complementary government-industry research and development (R&D) policy initiatives that promote the development of hydrogen fuel and fuel cell vehicles. Coordinated by the Department of Energy (DOE), these initiatives aim to make mass-market fuel cell and hydrogen combustion vehicles available at an affordable cost within 10 to 15 years from the launch of the initiatives. However, questions have been raised about the design and goals of the initiatives. This report discusses the organization, funding, and goals of the FreedomCAR and Fuel partnerships, and issues for Congress. Chapter 8 - Pursuant to section 801(a)(2)(A) of title 5, United States Code, this is our report on a major rule promulgated by the Department of Transportation, National Highway Traffic Safety Administration, entitled “Federal Motor Vehicle Safety Standards; Occupant Protection in Interior Impact; Side Impact Protection; Fuel System Integrity; ElectricPowered Vehicles: Electrolyte Spillage and Electrical Shock Protection; Side Impact Phase-In Reporting Requirements” (RIN: 2127-AJ10). We received the rule on September 14, 2007. It was published in the Federal Register as a final rule on September 11, 2007. 72 Fed. Reg. 51,908. The final rule upgrades the Federal Motor Vehicle Safety Standard (FMVSS) No. 214, “Side impact protection.” The final rule does so by incorporating a dynamic pole test into FMVSS No. 214 requiring manufacturers to improve head and thorax protection to occupants of

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Preface

xi

vehicles that crash into poles and trees, or that are laterally struck by a higher-riding vehicle, and to reduce fatalities and injuries caused by partial ejections through side windows. Additionally, the final rule enhances the moving deformable barrier test by adding a second dummy to better represent the at-risk population in vehicle-to-vehicle side crashes. Enclosed is our assessment of the NHTSA’s compliance with the procedural steps required by section 801(a)(1)(B)(i) through (iv) of title 5 with respect to the rule. Our review indicates that NHTSA complied with the applicable requirements. Chapter 9 - In this study, electric-drive vehicles with series powertrains were configured to utilize a lithium- ion battery of very high power and achieve sport-sedan performance and excellent fuel economy. The battery electrode materials are LiMn2O4 and Li4Ti5O12, which provide a cell area-specific impedance of about 40% of that of the commonly available lithium-ion batteries. Data provided by EnerDel Corp. for this system demonstrate this low impedance and also a long cycle life at 55oC. The batteries for these vehicles were designed to deliver 100 kW of power at 90% open- circuit voltage to provide high battery efficiency (97-98%) during vehicle operation. This results in battery heating of only 1.6oC per hour of travel on the urban dynamometer driving schedule (UDDS) cycle, which essentially eliminates the need for battery cooling. Three vehicles were designed, each with series powertrains and simulation test weights between 1575 and 1633 kg: a hybrid electric vehicle (HEV) with a 45-kg battery, a plug-in HEV with a 10-mile electric range (PHEV10) with a 60-kg battery, and a PHEV20 with a 100-kg battery. Vehicle simulation tests on the Argonne National Laboratory’s simulation software, the Powertrain System Analysis Toolkit (PSAT), which was developed with MATLAB/Simulink, showed that these vehicles could accelerate to 60 mph in 6.2 to 6.3 seconds and achieve fuel economies of 50 to 54 mpg on the UDDS and highway fuel economy test (HWFET) cycles. This type of vehicle shows promise of having a moderate cost if it is mass produced, because there is no transmission, the engine and generator may be less expensive since they are designed to operate at only one speed, and the battery electrode materials are inexpensive. Chapter 10 - The Energy Independence and Security Act (P.L. 110-140, H.R. 6) is an omnibus energy policy law that consists mainly of provisions designed to increase energy efficiency and the availability of renewable energy. This report describes the key provisions of the enacted law, summarizes the legislative action on H.R. 6, and provides a summary of the provisions under each of the titles in the law. The highlights of key provisions enacted into law are as follows: Corporate Average Fuel Economy (CAFE). The law sets a target of35 miles per gallon for the combined fleet of cars and light trucks by model year 2020. Renewable Fuels Standard (RFS). The law sets a modified standard that starts at 9.0 billion gallons in 2008 and rises to 36 billion gallons by 2022. Energy Efficiency Equipment Standards. The adopted bill includes a variety of new standards for lighting and for residential and commercial appliance equipment. The equipment includes residential refrigerators, freezers, refrigerator-freezers, metal halide lamps, and commercial walk-in coolers and freezers. Repeal of Oil and Gas Tax Incentives. The enacted law includes repeal of two tax subsidies in order to offset the estimated cost to implement the CAFE provision.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

xii

Gerald B. Raines

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The two most controversial provisions of H.R. 6 that were not included in the enacted law were the proposed Renewable Energy Portfolio Standard (RPS) and most of the proposed tax provisions, which included repeal of tax subsidies for oil and gas and new incentives for energy efficiency and renewable energy.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Electric Vehicles: Technology, Research and Development ISBN 978-1-60741-142-0 Editor: Gerald B. Raines © 2009 Nova Science Publishers, Inc.

Chapter 1

ELECTRIC MOTOR DRIVES FOR BATTERY, HYBRID AND FUEL CELL VEHICLES K. T. Chau* International Research Center for Electric Vehicles, Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China

ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

With ever increasing concern on environmental protection and energy conservation, there is a fast growing interest in electric vehicles (EVs) from automakers, governments and customers. As electric propulsion is the core of EVs, it is a pressing need for researchers to develop advanced electric motor drives for various classes of EVs, including the battery, hybrid and fuel cell vehicles. This chapter firstly reviews various electric motor drives, including DC, induction, switched reluctance and permanent magnet (PM) brushless types, for application to EVs. As the development of PM brushless motor drives is being actively engaged, their motor configurations and operating principles are delineated. Secondly, two emerging PM brushless motor drives, namely the double-stator PM brushless motor drive and the hybrid-PM brushless motor drive, are discussed. Their motor configurations, design equations, analysis approaches, converter topologies and control strategies are described when necessary. Then, their operating characteristics are evaluated. Finally, the concept of integrated PM motor drive systems is revealed. It takes the definite advantages of high efficiency, good controllability, compact size and lightweight, which are particularly essential for future vehicles. Hence, three latest integrated PM brushless motor drive systems, namely the magnetic-geared PM brushless motor drive system, the memory-PM brushless integratedstarter-generator system and the PM brushless electric variable transmission system, are discussed. Future research directions of these integrated systems are particularly identified. *

Professor. Department of Electrical and Electronic Engineering, The University of Hong Kong, Pokfulam Road, Hong Kong. Phone: 852-2859-2704; Fax: 852-2559-8738; Email: [email protected]

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

K. T. Chau

2

1. INTRODUCTION Vehicles equipped with an internal combustion engine have been in existence since 1885. Although engine vehicles are continually improved by modern automotive electronics, they need a major change to improve the fuel economy and to reduce the exhaust emission. Electric vehicles (EVs), including battery, hybrid and fuel cell vehicles, have been identified to be the most viable solution to fundamentally solve the problems associated with engine vehicles (Chan and Chau, 2001; Ehsani et al., 2005). Electric motor drives are the core technology for EVs. The basic requirements of an electric motor drive for EVs are summarized below (Chau and Chan, 2007; Ehsani et al., 1997; Zhu and Howe, 2007):

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

• • • • • • • • • • • •

high torque density and high power density; wide speed range, covering low-speed creeping and high-speed cruising; high efficiency over wide torque and speed ranges; wide constant-power operating capability; high torque capability for electric launch and hill climbing; high intermittent overload capability for overtaking; high reliability and robustness for vehicular environment; low acoustic noise; reasonable cost; high-efficiency generation over a wide speed range for hybrid vehicles; good voltage regulation over wide-speed generation for hybrid vehicles; low exhaust emission for hybrid vehicles.

The purpose of this chapter is to give an overview of electric motor drives for EVs. Firstly, a review of traditional electric motor drives, namely the DC, induction, switched reluctance (SR) and permanent magnet (PM) brushless ones, will be conducted. Secondly, two emerging PM brushless motor drives, namely the double-stator PM brushless motor drive and the hybrid-PM brushless motor drive, will be presented. Thirdly, the latest integrated PM motor drive systems, namely the magnetic-geared PM brushless motor drive system, the memory-PM brushless integrated-starter-generator system and the PM brushless electric variable transmission system, will be discussed. Finally, a conclusion will be drawn.

2. REVIEW Among different types of electric motor drives, there are four types that are viable for EVs, namely the DC, induction, SR, and PM brushless motor drives. They possess fundamentally different motor topologies as illustrated in Figure 1. Basically, they are classified into two main groups, namely the brushed and brushless groups, and each group can be further classified into different subgroups as illustrated in Figure 2. It should be noted that the branches that are not viable for EVs have been pruned. Table 1 also lists the previous or latest applications of those viable motor drives to flagship EVs.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

Figure 1. Typical EV motor topologies.

Figure 2. Classification of EV motor drives.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

3

4

K. T. Chau Table 1. Applications to flagship EVs

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.1. DC Motor Drives DC motor drives used to be widely accepted for EVs. Based on the location of field excitation, they can be grouped as the self-excited DC and separately excited DC types. Based on the source of field excitation, they can also be grouped as the wound-field DC and PM DC types. As determined by the mutual interconnection between the field winding and the armature winding or the use of PM excitation, the whole family consists of the separately excited DC, series DC, shunt DC and PM DC types. Because of their technological maturity and control simplicity, various DC motor drives have ever been applied to different EVs as listed in Table 1. For the separately excited DC motor drive, the field and armature voltages can be controlled independent of each other. The torque-speed characteristic is linearly related that speed decreases as torque increases and speed regulation depends on the armature circuit resistance. For the series DC motor drive, the field current is the same as the armature current. An increase in torque is accompanied by an increase in the armature current and hence an increase in flux. The speed must drop to maintain the balance between the supply and induced voltages. The torque-speed characteristic has an inverse relationship. For the shunt dc motor drive, the field and armature are connected to a common voltage source. The corresponding characteristic is similar to that of the separately excited DC motor drive. For the PM DC motor drive, it has relatively higher power density and higher efficiency because of the spacesaving benefit by PMs and the absence of field losses. Owing to the low permeability of PMs, similar to that of air, armature reaction is usually reduced and commutation is improved. However, since the field excitation in the PM DC motor drive is uncontrollable, it cannot readily attain the operating characteristics similar to that of other wound-field DC motor drives. In general, speed control of DC motor drives can be accomplished by two methods, namely armature control and field control. When the armature voltage is reduced, the armature current and hence the motor torque decrease, causing the motor speed to decrease. In contrast, when the armature voltage is increased, the motor torque increases, causing the motor speed to increase. Since the maximum allowable armature current remains constant and the field is fixed, this armature voltage control has the advantage of retaining the maximum torque capability at all speeds. However, since the armature voltage cannot be further increased beyond the rated value, this control is used only when the DC motor drive operates below its base speed. On the other hand, when the field voltage of the DC motor is weakened

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

5

while the armature voltage is fixed, the motor induced EMF decreases. Because of low armature resistance, the armature current will increase by an amount much larger than the decrease in the field. Thus, the motor torque is increased, causing the motor speed to increase. Since the maximum allowable armature current is constant, the induced EMF remains constant for all speeds when the armature voltage is fixed. Hence, the maximum allowable motor power becomes constant so that the maximum allowable torque varies inversely with the motor speed. Therefore, in order to achieve wide-range speed control of DC motor drives for EVs, armature control has to be combined with field control. By maintaining the field constant at the rated value, armature control is employed for speeds from standstill to the base speed. Then, by keeping the armature voltage at the rated value, field control is used for speeds beyond the base speed. All DC motor drives suffer from the same problem due to the use of commutators and brushes. Commutators cause torque ripples and limit the motor speed, while brushes are responsible for friction and radio-frequency interference. Moreover, due to the wear and tear, periodic maintenance of commutators and brushes is always required. These drawbacks make them less reliable and unsuitable for maintenance-free operation, and limit them to be widely applied for modern EV propulsion.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.2. Induction Motor Drives At present, induction motor drives are the most mature technology among various brushless motor drives. There are two types of induction motors, namely the wound-rotor and squirrel-cage. Because of high cost, need of maintenance and lack of sturdiness, wound-rotor induction motors are less attractive than squirrel-cage counterparts for electric propulsion in EVs. Hence, squirrel-cage induction motors are loosely named as induction motors. Apart from those advantages of brushless motor drives, induction motor drives possess the definite advantages of low cost and ruggedness. Speed control of induction motor drives is considerably more complex than that of DC motor drives because of the nonlinearity of the dynamic model with coupling between direct and quadrature axes. A number of control strategies have been developed to allow induction motor drives to be applicable for electric propulsion. There are three state-of-the-art control strategies, namely variable-voltage variable-frequency (VVVF) control, field-oriented control (FOC) which is also called vector control or decoupling control, and pole-changing control. On top of these control strategies, sophisticated control algorithms such as adaptive control, variable-structure control and optimal control have also been employed to achieve faster response, higher efficiency and wider operating ranges. The VVVF control strategy is based on constant volts/hertz control for frequencies below the motor rated frequency, whereas variable-frequency control with constant rated voltage for frequencies beyond the rated frequency. For very low frequencies, voltage boosting is applied to compensate the difference between the applied voltage and induced EMF due to the stator resistance drop. Because of the disadvantages of air-gap flux drifting and sluggish response, the VVVF control strategy is becoming less attractive for high-performance EV induction motor drives. In order to improve the dynamic performance of induction motor drives for EV propulsion, FOC is preferred to VVVF control. By using FOC, the mathematical model of

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

K. T. Chau

6

induction motors is transformed from the stationary reference frame (d-q frame) to the general synchronously rotating frame (x-y frame). When the x-axis is purposely selected to be coincident with the rotor flux linkage vector, the reference frame (α-β frame) becomes rotating synchronously with the rotor flux. Hence, the motor torque T can be obtained as:

T=

3 M2 p i s α is β 2 Lr

where p is the number of pole pairs, M is the mutual inductance per phase, Lr is the rotor inductance per phase, and isα and isβ are the α-axis component and β-axis component of

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

stator current, respectively. This torque equation is very similar to that of separately excited DC motor drives. Therefore, by means of this FOC, the motor torque can be effectively controlled by adjusting the torque component isβ as long as the field component isα remains constant. Hence, induction motor drives can offer the desired fast transient response similar to that of separately excited DC motor drives. It is well known that a change of the number of pole pairs of induction motor drives can adjust the rotating-field synchronous speed. The squirrel-cage type takes a definite advantage over the wound-rotor type that it is able to automatically adapt the pole number of the rotor to that of the stator. In early time, this pole-changing control was implemented by using mechanical contactors, and only two or three discrete speeds were achieved. With the advancement of power electronics and control technologies, the pole-changing control can be implemented electronically. Hence, the high-speed constant-power capability can be remarkably extended, which is particularly desirable for EV cruising. Recently, new design approaches have been developed to improve the power density of EV induction motor drives by up to 30% (Wang, et al., 2005). Interdisciplinary design considerations on mechanical vibration and acoustic noise of induction motor drives have also been analyzed for EVs (Lo et al., 2000). Also, electronically pole-changing schemes have been developed for EV induction motor drives, which can significantly extend the constantpower operating region to over four times the base speed (Jiang et al., 2003).

2.3. Switched Reluctance Motor Drives SR motor drives have been recognized to have considerable potential for EVs. They have the definite advantages of simple construction, low manufacturing cost and outstanding torque-speed characteristics. The operating principle of SR motor drives is based on the ‘minimum reluctance’ rule. According to the co-energy principle, the reluctance torque produced by one phase at any rotor position is given by:

1 dL T (θ , i ) = i 2 2 dθ

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

7

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

where θ is the rotor position angle, i is the phase current and L is the phase inductance. They have two significant features. One is that the direction of torque is independent of the polarity of the phase current. Another is that the motoring torque can be produced only in the direction of rising inductance ( dL / dθ > 0 ); otherwise, a negative torque (or braking torque) is produced. So, each phase can produce a positive torque only in half a rotor pole-pitch, hence creating the torque ripple. Also, because of the heavy saturation of pole tips and the fringing effect of poles and slots, they usually exhibit acoustic noise problems. The SR motor drives have basically two operating modes. When the speed is below the base speed, the current can be limited by chopping, so-called current chopping control (CCC). In the CCC mode, the torque and thus the constant-torque characteristic can be controlled by changing the current limits. During high speed operation, however, the peak current is limited by the EMF of the phase winding. The corresponding characteristic is essentially controlled by phasing of switching instants relative to the rotor position, so-called angular position control (APC). In the APC mode, the constant-power characteristic can be achieved. In short, SR motor drives have two fundamental problems hindering their application to EVs – acoustic noise and control nonlinearity. Over the years, fuzzy sliding mode control has been developed for SR motor drives in EVs (Zhan et al., 1999). This control approach incorporates both fuzzy logic and sliding mode control in such a way that the former functions to handle the motor nonlinearities while the latter is used to reduce the control chattering. Also, the corresponding constant-power region has been extended to 3−7 times the base speed through phase advancing excitation (Rahman et al., 2000; Inderka et al., 2002). On the other hand, an active vibration cancellation technique for SR motor drives has been proposed, which induces an anti-phase vibration to cancel a specified vibration mode and hence reduces the acoustic noise (Long et al., 2005).

2.4. Permanent Magnet Brushless Motor Drives PM brushless motor drives are becoming more and more attractive for EVs, since they inherently offer high efficiency, high power density and high reliability. The key problems are their relatively high PM material cost and uncontrollable PM flux. In recent years, research and development of EV motor drives has been focused on PM brushless motor drives, aiming to fully utilize their PM material and to control the resultant air-gap flux. Various viable motor topologies have been proposed. Based on the location of PMs, they can be classified as the rotor-PM and stator-PM classes. According to the operating current and no-load EMF waveforms, they can also be classified as the PM brushless AC (BLAC) and PM brushless DC (BLDC) types (Chan et al., 1996).

2.4.1. Rotor-PM Brushless Motor Topologies The rotor-PM brushless motor topologies are most popular. According to the position of PMs in the rotor, they can further be classified as surface-mounted, surface-inset, interiorradial and interior-circumferential topologies as shown in Figure 3 (Gan et al., 2000; Zhu and Howe, 2007).

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

K. T. Chau

8

The generated torque of these rotor-PM brushless motors consists of two components, namely the PM torque and reluctance torque, which are given by:

T=

[

3 p ψ m I q − (Lq − Ld )I d I q 2

]

where p is the number of pole-pairs, ψ m is the stator winding flux linkage due to the PMs,

Ld , Lq are respectively the d-axis and q-axis stator winding inductances, and I d , I q are

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

respectively the d-axis and q-axis currents.

Figure 3. Rotor-PM brushless motor topologies.

For the surface-mounted PM brushless motor topology, the PMs are simply mounted on the rotor surface by using epoxy adhesives. Since the permeability of PMs is near to that of air, the effective air-gap is the sum of the actual air-gap length and the radial thickness of the PMs. Hence, the corresponding armature reaction field is small and the stator winding inductance is low. Also, since the d-axis and q-axis stator winding inductances are nearly the same, its reluctance torque is almost zero. For the surface-inset PM brushless motor topology, the PMs are inset or buried in the rotor surface. Thus, the q-axis inductance becomes higher than the d-axis inductance, hence producing the additional reluctance torque. Also, since the PMs are inside the rotor, it can withstand the centrifugal force at high-speed operation, hence offering good mechanical integrity.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

9

For the interior-radial PM brushless motor topology, the PMs are radially magnetized and buried inside the rotor. Similar to the surface-inset one, the PMs are mechanically protected, hence allowing for high-speed operation. Also, because of its d-q saliency, an additional reluctance torque is generated. Different from the surface-inset one, this interior-radial topology adopts linear PMs which are easier for insertion and can be easily machinable. For the interior-circumferential PM brushless motor topology, the PMs are circumferentially magnetized and buried inside the rotor. It takes the definite advantage that the air-gap flux density can be higher than the PM remanent flux density, so-called the flux focusing. Also, it holds the merits of good mechanical integrity and additional reluctance torque. However, because of significant flux leakage at the inner ends of PMs, a nonmagnetic shaft or collar is generally required.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.4.2. Stator-PM Brushless Motor Topologies The stator-PM brushless motor topologies are with PMs located in the stator, and generally with salient poles in both the stator and rotor. So, they are usually termed as doublysalient PM (DSPM) motors. Since the rotor has neither PMs nor windings, these DSPM motors are mechanically simple and robust, hence very suitable for high-speed operation. According to the shape and location of the PMs, they can be classified as the yoke-linearmagnet, yoke-curved-magnet, tooth-surface-magnet and tooth-interior-magnet motors as shown in Figure 4.

Figure 4. Stator-PM brushless motor topologies.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

10

K. T. Chau

The yoke-linear-magnet motor topology is most commonly adopted in DSPM motors (Chau et al., 2005; Cheng et al., 2001; Cheng et al., 2003). Although they are salient poles in the stator and rotor, the PM torque significantly dominates the reluctance torque, hence exhibiting low cogging torque. Since the variation of flux linkage with each coil as the rotor rotates is unipolar, it is very suitable for the BLDC operation. On the other hand, when the rotor is skewed, it can offer the BLAC operation. The major disadvantage of this topology is the relatively low torque density, as resulted from its unipolar flux linkage. The yoke-curved-magnet motor topology is very similar to the previous one, except the shape of PMs. Different from the yoke-linear-magnet one, the periphery of this topology is essentially circular. Also, since there is more space to accommodate the PMs, this DSPM motor can achieve higher air-gap flux density. Its major drawback is the difficulty in machining the curved PMs and inserting them into the stator core.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 5. Waveforms of PM brushless motor drive operations.

The tooth-surface-magnet motor topology is commonly termed as the flux-reversal PM motor, since the flux linkage with each coil reverses polarity as the rotor rotates (Deodhar et al., 1997; Zhu and Howe, 2007). Each stator tooth has a pair of PMs of different polarities mounted onto the surface. Hence, the flux linkage variation is bipolar so that the torque density is higher than that of the conventional DSPM motor. However, since the PMs are on the surface of stator teeth, they are more prone to partial demagnetization. Also, significant eddy current loss in the PMs may be resulted. The tooth-interior-magnet motor topology is commonly termed as the flux-switching PM motor (Zhu et al., 2005; Zhu and Howe, 2007). In this topology, each stator tooth consists of two adjacent laminated segments and a PM, and each of these segments is sandwiched by two circumferentially magnetized PMs. Hence, it enables flux focusing. Compared with the rotorPM topologies, this flux-switching motor has less armature reaction, hence offering higher electric loading. Since its back EMF waveform is essentially sinusoidal, this machine is more suitable for the BLAC operation.

2.4.3. Brushless AC and DC Operations As aforementioned, PM brushless motor drives have two basic operations, namely the BLAC and BLDC, as shown in Figure 5. Actually, each PM brushless motor can operate at both modes if the torque density, torque smoothness and efficiency are not of great concern. For the PM BLAC motor drives, they operate with sinusoidal current and sinusoidal air-gap

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

11

flux so that they need high-resolution position signal for closed-loop control, hence desiring a costly position encoder or resolver. In contrast, for the PM BLDC motor drives, they operate with rectangular current and trapezoidal air-gap flux so that they just need a low-cost sensor for phase-current commutation. Nevertheless, the PM BLAC motor drives allow for openloop operation, whereas the position feedback is mandatory for the PM BLDC motor drives.

2.4.4. Constant-Power Operation EV motor drives desire to operate over a very wide speed range, especially high-speed constant-power operation for cruising. Different types of PM brushless motor drives may adopt different methods to achieve constant-power operation. For the PM BLAC motor drives, constant-power operation can readily be offered by using flux-weakening control. The maximum flux-weakening capability is achieved when the motor is designed to have unity per-unit d-axis inductance (Soong and Ertugrul, 2002):

Ld I r

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ψm

=1

Figure 6. Torque-speed characteristics of PM brushless motor drives.

where ψ m is the PM flux linkage, Ld is the d-axis winding inductance and I r is the rated current. In general, the ratio of Ld I r /ψ m is less than unity. So, the higher the ratio, the higher will be the flux-weakening capability. The flux-weakening control has been comprehensively studied in various PM BLAC motor drives (Zhu et al., 2000; Uddin and Rahman, 2007). For the PM BLDC motor drives, constant-power operation is more complex. Since the operating waveforms are no longer sinusoidal, d-q transformation and hence flux-weakening control are ill-suited. Nevertheless, the corresponding constant-power operation can be offered by using advanced conduction angle control (Chan et al., 1995; Kim et al., 1997). Figure 6 shows the torque-speed characteristics of the PM brushless motor drives without control and with control (either flux-weakening control for the BLAC or advanced conduction angle control for the BLDC). It illustrates that the speed range of constant-power operation can be significantly extended. On the other hand, Figure 6 gives a comparison of the torquespeed characteristics of the PM BLAC motor drive and the PM BLDC motor drive. It can be seen that the BLAC motor drive offers higher torque and power capabilities than the BLDC motor drive employing the 2-phase 120° conduction. Nevertheless, the BLDC motor drive

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

K. T. Chau

12

employing the 3-phase 180° conduction can offer better high-speed power capability, but with the sacrifice of low-speed torque capability (Zhu and Howe, 2007). Moreover, for the PM BLDC motor drive with multiphase polygonal windings (Wang et al., 2002), the corresponding back EMF, rather than the air-gap flux, can be directly varied to enable constant-power operation. Similarly, the split-winding DSPM motor drive (Cheng et al., 2003) can perform constant-power operation by varying the effective number of armature winding turns.

3. EMERGING PM BRUSHLESS MOTOR DRIVES Because of the uncontrollable PM flux, PM brushless motor drives generally suffer from the difficulty in air-gap flux control. Thus, their constant-power operating ranges are limited, while some sophisticated operations such as on-line efficiency optimization or fluxstrengthening operation cannot be achieved. In the following, two kinds of emerging PM brushless motor drives will be discussed, which possess different configurations and employ different strategies to achieve effective air-gap flux control.

3.1. Double-Stator PM Brushless Motor Drives

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 7 shows the configuration of the double-stator PM brushless motor (DS-PMBM) drive (Niu et al., 2007; Chai et al., 2005). It consists of two concentric stators with two sets of three-phase windings and one cup-shape rotor with PMs mounted on both inner and outer surfaces. Thus, it offers the following advantages: • • • •

• •

The use of double stators can significantly increase the torque density, and can provide versatile connection modes for generation over a wide speed range. The cup-shape rotor can effectively shorten the magnetic circuit length, thus improving the torque density. Since the permeability of PMs is similar to that of air, the mutual inductance between the inner stator and the outer stator is negligible, thus improving the controllability. Since the coil span of both stator windings is one slot pitch, the flux path of each phase is independent so that the mutual inductance between phase windings is negligible, thus further improving the controllability. Since the motor adopts a fractional number of slots per pole per phase, namely the slot pitch is 11/12 pole pitch, the cogging torque can be significantly reduced. The mutipole structure can shorten the magnetic circuit length while the one-slotpitch coil-span winding arrangement can shorten end-windings, thus improving the utilization of both iron and copper materials, and hence the torque density.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

13

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 7. DS-PMBM configuration.

In order to achieve high starting torque, the double-stator windings can be arranged in such a way that the same phases are connected in series. On the other hand, in order to achieve constant generated output voltage over a wide range of speeds, the double-stator windings can adopt versatile connection modes. When there is a spatial angle of 20° between the two stators, it yields six possible ways to connect the two sets of three-phase windings. As shown in Figure 8, different composite vectors can be generated by using different winding connections. When the rotor speed ω is below its base speed ωb , the same phases (for instance, E A and Ea ) are connected in series to form E1 , so-called the Mode 1 of operation. Between ωb and 1.2ωb , the adjacent anti-phases are connected in series to form

E2 , so-called the Mode 2. Similarly, between 1.2ωb and 1.5ωb , E3 is formed under the Mode 3 of operation. Then, between 1.5ωb and 2ωb , the adjacent phases are connected in series to form E4 under the Mode 4 of operation. Similarly, between 2ωb and 3ωb , E5 is formed under the Mode 5 of operation. Finally, between 3ωb and 6ωb , the same anti-phases are connected in series to form E6 under the Mode 6 of operation. Based on the corresponding vector diagrams, it can be deduced that E1 ≈ 6 E6 ,

E2 ≈ 5E6 , E3 ≈ 4E6 , E4 ≈ 3E6 and E5 ≈ 2E6 , indicating that the generated EMF can be discretely tuned in terms of E6 . For further fine tuning or with speed beyond 6ωb , the conventional flux-weakening control can be employed to adjust the output voltage.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

14

K. T. Chau

Figure 8. Winding connections and vector diagrams.

In order to perform electronic winding connections, a matrix converter is used. As shown in Figure 9, this matrix converter is composed of a matrix of power switches, which takes the definite advantage that it can provide versatile control and regulation of power flow. Table 2 shows the switching combinations to realize the six modes of operation. For instance, the switches S X 1 , SY 2 and S Z 3 are turned on while the others are off when operating in the Mode 1 with ω < ωb . With a very wide range of speeds, all six modes of operation occur. The corresponding generated output voltage waveforms are shown in Figure 10. It can be seen that the magnitude of these output voltages can be maintained almost constant over a speed range of six times the base speed.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

15

Figure 9. Control of winding connections.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 2. Switching combinations

3.2. Hybrid-PM Brushless Motor Drives In general, hybrid-PM brushless motor drives are referred to the motors that the PM excitation is hybridized with the DC field winding excitation to produce the desired magnetic field. There are many attractive features due to the presence of hybrid-PM field: • •

• •

•

By varying the polarity and magnitude of the DC field winding current, the air-gap flux density becomes easily controllable. By realizing flux strengthening, the motor can offer the exceptionally high-torque feature, which is very essential for cold cranking engines, or providing temporary power for vehicular overtaking and hill climbing. By realizing flux weakening, the motor can offer the exceptionally wide-speed constant-power feature, which is very essential for vehicular cruising. By online tuning the air-gap flux density, the motor can maintain constant voltage output under generation or regeneration over a very wide speed range, which is very essential for battery charging. By online tuning the air-gap flux density, the motor can also offer efficiencyoptimizing control, which is highly desirable for EVs.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

16

K. T. Chau

Figure 10. Generated output voltage in different modes at different speeds.

3.2.1. Double-Stator Claw-Rotor Hybrid-PM Brushless Motor Topology Figure 11 shows the configuration of a double-stator claw-rotor hybrid-PM motor drive which is specially designed for EVs (Chan et al., 1996). It has a unique structure which comprises of an outer stator with three-phase AC winding, an inner stator with DC field winding, and a claw-type rotor with integrated PMs. Thus, the components of air-gap flux, respectively produced by PMs and the DC field winding, are magnetically shunt in nature. The advantages and special features of this motor drive are summarized as follows:

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles •

•

•

By adopting the claw-type rotor structure, the leakage flux can be minimized and the construction becomes compact. Moreover, by locating the field winding in the inner stator, the motor axial length can be shortened and the material consumption can be reduced. Due to the existence of both PMs and the field winding, the motor can be designed to achieve higher air-gap flux density and hence higher power density. Increasingly, the mounting of PMs adopts the flux-focusing arrangement, which allows the air-gap flux density even higher than the operating flux density of individual PMs. By controlling the direction and magnitude of the DC field current, the air-gap flux can be flexibly adjusted, hence the torque-speed characteristics can be easily shaped to meet the special requirements for EV propulsion. Particularly, by using field current control to weaken the air-gap flux produced by PMs, the speed range for constant-power operation can be significantly extended. By properly controlling the applied voltage and DC field current, the efficiency of the motor drive can be optimized throughout the whole operating range. Thus, the efficiency at those operating regions for EV propulsion, such as high-torque lowspeed hill-climbing and low-torque high-speed cruising, can be improved.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

•

17

Figure 11. Double-stator claw-rotor hybrid-PM motor topology.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

18

K. T. Chau

Although this hybrid-PM brushless motor drive possesses the above advantages, it does have some shortcomings. Particularly, the structure is quite complicated, which desires artful integration. Also, its motor design and optimization desire three-dimensional electromagnetic field analysis, which is lengthy and tedious.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2.2. Doubly-Salient Hybrid-PM Brushless Motor Topology Figure 12 shows another hybrid-PM brushless motor topology (Chau et al., 2003; Chau et al., 2006b). It consists of a stator with six salient poles and a rotor with four salient poles. The stator incorporates two types of windings, namely the three-phase armature winding and the DC field winding, and PM poles. The rotor has neither PMs nor windings, hence, offering high mechanical integrity for high-speed operation.

Figure 12. Doubly-salient hybrid-PM brushless motor topology.

The three-phase armature winding operates like that for a DSPM motor, whereas the DC field winding not only works as an electromagnet but also as a tool for flux weakening and efficiency optimization. Notice that flux-weakening operation is necessary for high-speed EV cruising, whereas efficiency-optimizing control is essential for long EV driving range. Also, there is an extra air bridge in shunt with each PM. If the field winding MMF reinforces the PM MMF, this extra flux path will assist the effect of flux strengthening. On the other hand, if the field winding MMF opposes the PM MMF, this extra flux path will favor the PM flux leakage, hence amplifying the effect of flux weakening. As a result, with a proper design of the air-bridge width, a wide flux-regulating range can be obtained by using a small DC field excitation. Under the assumptions that the fringing effect is negligible and the permeability of the iron core is infinite, a linear variation of flux linkage is resulted, where the maximum value occurs at the alignment between the rotor pole and the stator pole, whereas the minimum value occurs at their nonalignment. When the flux linkage is increasing, an armature current

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

19

with a positive value is applied to the phase winding, hence producing a positive torque. Similarly, when the flux linkage is decreasing, a negative current is applied to the winding so that a positive torque is also produced. Thus, two possible torque producing zones are fully utilized. Although this topology takes the definite advantages of high mechanical integrity and flexible air-gap flux control which are highly desirable for EVs, its stator is relatively bulky and the corresponding leakage flux is also significant.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2.3. Outer-Rotor Doubly-Salient Hybrid-PM Brushless Motor Topology Figure 13 shows an improved doubly-salient hybrid-PM brushless motor topology (Chau et al., 2006a). There are DC field windings in shunt with PMs in the inner stator, air bridges in shunt with PM pieces, armature windings in salient poles of the outer stator, and salient poles (without windings or PMs) in the outer rotor. It offers the following advantages:

Figure 13. Outer-rotor doubly-salient hybrid-PM brushless motor topology.

•

•

The outer-rotor topology can enable full utilization of the space of inner stator (the part beneath the armature windings) to accommodate both the PMs and DC field windings, hence improving the power density. Also, since both the PMs and DC field windings are embraced by the rotor, the problem of flux leakage can be minimized. Since the rotor does not involve any windings or PMs, it can provide high mechanical integrity which is essential to handle the high starting torque and to withstand high-speed operation.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

K. T. Chau

20 •

•

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

•

The DC field windings can provide flexible flux control, including both flux strengthening and flux weakening. When operating as a starter motor, the DC field windings are excited to produce additional flux in the same direction as the PM flux, hence strengthening the air-gap flux linkage to achieve very high starting torque. On the other hand, when operating as a generator, the air-gap flux linkage is regulated by varying the DC field current in such a way that the PM flux is weakened or strengthened by the DC field, hence realizing the desired constant output voltage over a wide speed range. The air bridges can bypass the PM flux when the two field excitations are opposite, hence amplifying the effect of flux weakening. Since the stator adopts fractional-slot concentrated windings, it can effectively reduce the cogging torque which usually occurs in PM BLDC motors. Also, it can shorten the length of end-windings, hence saving the copper material and improving the power density.

Figure 14. Magnetic equivalent circuit.

To illustrate the ability and the range of controllable flux, an equivalent magnetic circuit of the motor at no load is shown in Figure 14, where RPM is the PM reluctance, Rb is the airbridge reluctance, Rg is the air-gap reluctance, FDC is the DC field winding MMF, and FPM is the PM MMF. Hence, it yields two design equations:

⎞ FDC + ⎛⎜ Φ g + = − 1⎟ ⎜Φ g0 ⎟ FPM ⎝ ⎠

⎞ ⎛ R PM ⎜⎜ + 1⎟⎟ ⎠ ⎝ Rb

FDC − ⎛⎜ Φ g − = 1− ⎜ Φ g0 FPM ⎝

⎞ ⎛ R PM ⎜⎜ + 1⎟⎟ ⎠ ⎝ Rb

⎞ ⎟ ⎟ ⎠

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

21

where Φ g 0 is the air-gap flux at FDC = 0 , FDC − and Φ g − are under flux weakening, and

FDC + and Φ g + are under flux strengthening. Given the air-gap flux ranges of (Φ g + Φ g 0 ) = 3 and (Φ g − Φ g 0 ) = 1 / 3 , the relationship between ( RPM Rb ) , ( FDC − FPM ) and ( FDC + FPM ) can be obtained as listed in Table 3. When selecting ( RPM Rb ) = 7 , it results

( FDC − = FPM 12) and ( FDC + = FPM 4) , indicating that the DC field winding excitation

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

needs only 8.3% of PM excitation for flux weakening and 25% of PM excitation for flux strengthening to achieve 9 times change of air-gap flux. The operating principle of this motor drive is similar with that of a DSPM motor drive, except that the flux is controllable. Figure 15 shows the magnetic flux density distributions with flux weakening (–350 A-turns), with no flux control and with flux strengthening (+1000 A-turns). It can be seen that the air-gap flux can be effectively controlled. Consequently, the developed torque waveforms with and without flux strengthening are shown in Figure 16. It depicts that the developed torque under flux strengthening can be boosted up by about 3 times, which is very essential to offer high starting torque. On the other hand, the no-load rectified output voltage characteristics with and without flux control are shown in Figure 17. It depicts that the output voltage can maintain constant over the whole speed range, which is very essential for battery charging.

Figure 15. Magnetic flux density distributions with and without flux control. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

22

K. T. Chau

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 3. Selection of reluctance ratio

Figure 16. Developed torque waveforms with and without flux control.

Figure 17. Rectified output voltage characteristics with and without flux control. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

23

4. INTEGRATED PM BRUSHLESS MOTOR DRIVE SYSTEMS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The development of EV propulsion is no longer limited to design and operation of a particular motor drive. Actually, the latest research directions have been focused on system integration for propulsion. In the following, three emerging integrative technologies are identified and discussed, namely the integration of magnetic gearing and PM brushless motor drives for battery or fuel cell vehicles, the integration of PM brushless starter motors and generators for mild hybrid vehicles, and the integration of PM brushless motor drives and electric variable transmission for full hybrid vehicles.

Figure 18. Comparison of in-wheel motor drives. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

24

K. T. Chau

Figure 19. Magnetic-geared PM brushless motor drive.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4.1. Magnetic-Geared PM Brushless Motor Drive System For EVs, PM brushless motor drives are very attractive since they inherently offer high power density and high efficiency. In particular, in-wheel PM brushless motor drives can play the role of electronic differential (Chan and Chau, 2001). As the wheel speed is only about 600 rpm, the in-wheel PM brushless motor drive is either a low-speed gearless outer-rotor one or a high-speed planetary-geared inner-rotor one. Although the outer-rotor one takes the advantage of gearless operation, its low-speed operation causes bulky size and heavy weight. On the other hand, although the inner-rotor one takes the merits of reduced overall size and weight, the planetary gear inevitably involves transmission loss, acoustic noise and regular lubrication. Recently, magnetic gears are becoming attractive, since they inherently offer the merits of high efficiency, reduced acoustic noise, and maintenance free (Atallah and Howe, 2001). By artfully integrating the magnetic gear into a PM BLDC drive, the low-speed requirement for direct driving and the high-speed requirement for motor design can be achieved simultaneously (Chau et al., 2007). Figure 18 gives a schematic comparison of the existing planetary-gear inner-rotor topology and the magnetic-geared outer-rotor topology for inwheel motor drives. It can be seen that the outer-rotor topology not only offers reduced size and weight, but also eliminates all the drawbacks due to the mechanical gear. Its detailed structure is shown in Figure 19. The artfulness is the share of a common PM rotor, namely the outer rotor of a PM BLDC motor and the inner rotor of a concentrically arranged magnetic gear.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

Figure 20. Air-gap flux densities in magnetic-geared PM brushless motor drive.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

25

26

K. T. Chau

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 21. Developed torques in magnetic-geared PM brushless motor drive.

The operating principle of this magnetic-geared PM BLDC motor drive is similar to that of a high-speed planetary-geared inner-rotor motor drive, but with the difference that this one is an outer-rotor motor drive. Namely, the motoring operation is the same as the PM BLDC motor drive. Firstly, the stator is fed by three-phase voltages, which are rated at 220 Hz, to achieve the rated speed of 4400 rpm. Then, the magnetic gear steps down the rated speed to 600 rpm, which in turn boosts up the torque for direct driving. The torque transmission is based on the modulation of the air-gap flux density distributions along the radial and circumferential directions. By using the finite element method, it can be seen that the space harmonic is successfully modulated by the 25 stationary steel pole-pieces from 3 pole-pairs in the inner air-gap to 22 pole-pairs in the outer air-gap as shown in Figure 20. Hence, the developed torque in the outer rotor can be significantly amplified to about 7 times that of the inner rotor as shown in Figure 21.

4.2. Memory-PM Brushless Integrated-Starter-Generator System In conventional automobiles, the starter motor and generator are separately coupled with the engine as shown in Figure 22, hence providing high starting torque for cold cranking and generating electricity for battery charging, respectively. This arrangement takes the advantage of simplicity, but suffers from poor utilization of both machines and hence resulting in heavy weight and bulky size. In order to incorporate both functions in a single system, the development of integrated-starter-generator (ISG) system is becoming attractive. As shown in Figure 23, the ISG system for mild hybrid vehicles serves for both cold cranking and battery charging, while eliminating the use of transmission belts and flywheels.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

27

The electric machine is the core of the ISG system. Among different types of machines, the hybrid-PM brushless motor topologies are most attractive. Namely, they can offer effective air-gap flux control, hence enabling flux strengthening for cold cranking, flux weakening for high-speed constant-power operation, flux regulation for wide-speed constantvoltage generation and flux tuning for online efficiency optimization. The aforementioned types of hybrid-PM brushless motor topologies, including the double-stator claw-rotor one, the doubly-salient one and the outer-rotor doubly-salient one, are suitable for ISG application. However, they all have a key drawback – the continual excitation of DC field windings for flux control will significantly increase the copper loss, hence deteriorating the inherent merit of high efficiency in PM brushless motors. The memory-PM brushless motor is a new class of flux-controllable PM brushless motors, which has the distinct ability to change the intensity of magnetization and also memorize the flux density level in the PMs (Ostovic, 2003). As shown in Figure 24, this topology consists of aluminum nickel cobalt (AlNiCo) PMs sandwiched by iron core, which are then mechanically fixed to a nonmagnetic shaft. The online magnetization is achieved by properly applying a short DC current pulse flowing through the stator armature winding to change the magnetization level of the AlNiCo PMs in the rotor. However, it suffers from complicated control of armature current for PM magnetization, and the possibility of accidental demagnetization due to armature reaction especially during high-rate regenerative braking. By incorporating the concept of memory-PM brushless motor into the outer-rotor doublysalient hybrid-PM brushless motor, a new outer-rotor doubly-salient memory-PM brushless motor is resulted, which can offer effective and efficient air-gap flux control (Yu et al., 2008). The high effectiveness is due to its direct magnetization of PMs by magnetizing windings, whilst the high efficiency is due to the use of temporary current pulse for PM magnetization. The configuration of this machine is shown in Figure 25, which adopts a 5-phase outer-rotor double-layer-stator structure. The use of 5 phases rather than 3 phases is to enhance the torque smoothness which is desirable for vehicular operation. The use of outer-rotor design is to enable full utilization of the stator space, namely the armature windings located in the outer layer whilst both the PMs and magnetizing windings located in the inner layer, hence achieving a compact structure. Also, since the outer rotor is simply composed of salient poles with no PMs or windings, it is very robust and suitable for vehicular operation. Moreover, since the armature windings and the PMs are located in different layers of the stator, and the armature adopts fractional-slot windings with the coil span equal to the slot pitch, the PMs can be immune from accidental demagnetization by armature reaction. Similar to the conventional memory-PM brushless motor, the PM material used in this motor is AlNiCo alloy. Its demagnetization curve can offer a high remanent flux density Br to enable high air-gap flux density, and a relatively low coercive force H c to enable online magnetization. Different from the conventional memory-PM one, the PM magnetization level of this motor is tuned by applying a current pulse to the magnetizing windings. Thus, it does not need to control the d-axis armature current, which is very complicated and may even conflict with the motor control strategy.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

28

K. T. Chau

Figure 22. Separated-starter-generator for conventional vehicles.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 23. Integrated-starter-generator for mild hybrid vehicles.

As depicted in Figure 26, the level of full magnetization is denoted by the operating point P0 which is the intersection of the demagnetization curve and the load line. In order to lower the PM magnetization level, a negative current pulse is applied so that the operating point shifts from P0 to Q1 . After this current pulse, it moves along the recoil line Q1R1 and then settles at P1 . Similarly, in order to raise the PM magnetization level, a positive current pulse is applied so that the operating point shifts back to P0 . Therefore, by adjusting the magnitude and polarity of the current pulse, the PM magnetization level and hence the air-gap flux density can be flexibly controlled. By using finite element analysis, the magnetic field distributions at no load of this motor under different PM magnetization levels, namely full, half and weak, are shown in Figure 27. It can be observed that the magnetic field intensity changes in accordance with the PM magnetization levels. The corresponding air-gap flux density distributions are shown in Figure 28. It can be found that the air-gap flux density can be effectively controlled over a range of about 4 times. On the other hand, the effect of armature reaction of this motor can be assessed by applying full-load current to the armature windings while deactivating the PMs. As depicted in Figure 29, the armature reaction field concentrates on the outer stator only, thus confirming that the double-layer structure can virtually avoid accidental demagnetization of the PMs.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 24. Conventional memory-PM brushless motor topology.

Figure 25. Outer-rotor doubly-salient memory-PM brushless motor topology.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

29

30

K. T. Chau

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 26. Demagnetization curve of AlNiCo PMs.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 27. Magnetic field distributions under different PM magnetization levels.

Figure 28. Air-gap flux density distributions under different PM magnetization levels.

Figure 29. Magnetic field distribution due to armature reaction. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

31

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

32

K. T. Chau

Figure 30. No-load EMF waveforms at different speeds.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

33

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In order to assess the capability of this motor to maintain constant output voltage over a wide range of speeds, it is operated at 900 rpm (rated speed), 1800 rpm and 3600 rpm with and without using PM magnetization control. The corresponding no-load EMF waveforms are shown in Figure 30. It confirms that by online tuning the PM magnetization levels, the noload EMF can be maintained constant, which is essential to effectively charge the battery over a wide range of speeds. It is interesting to note that the new motor exhibits two unique features. On one hand, by totally demagnetizing 2 of the 6 PM pieces, the number of PM poles becomes 4, so-called the pole-dropping technique as illustrated in Figure 31. On the other hand, by fully reversely magnetizing 2 of the 6 PM pieces, the number of PM poles becomes 2, so-called the polereversing technique as illustrated in Figure 32. The corresponding effects at the rated speed of 900 rpm are shown in Figure 33, which shows that the no-load EMF is reduced by 29% when using the pole-dropping technique to achieve 4 PM poles and by 85% when using the polereversing technique to achieve 2 PM poles. Hence, the no-load EMF can be maintained constant when applying the pole-dropping technique at 1274 rpm or applying the polereversing technique at 5914 rpm. Nevertheless, in order to maintain the no-load EMF constant at other speeds, the control of magnetization levels of the PMs is still necessary.

Figure 31. Magnetic field distribution using pole-dropping technique.

Figure 32. Magnetic field distribution using pole-reversing technique.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

34

K. T. Chau

Figure 33. No-load EMF waveforms at rated speed under different numbers of PM poles.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4.3. PM Brushless Electric Variable Transmission System In 1997, Toyota developed the first electric variable transmission (EVT) system for its flagship hybrid vehicle, Prius, which is a full hybrid. The schematic configuration of this EVT is shown as Figure 34, which is mainly composed of a planetary gear, a motor and a generator. The key is the planetary gear which is depicted in Figure 35, where the sun gear is connected to the generator, the ring gear (also called the crown gear) to the transmission shaft, and the carrier (also called the cage or yoke) to the engine. Thus, the engine transfers power outwards via the planet gears (also called the pinion gears) to the ring gear which is coupled to the transmission shaft that drives the wheels, and also inwards to the sun gear which then generates electrical energy via the generator. The angular velocities of all gear shafts are governed by a constraint equation:

ω s + k gωr − (1 − k g )ωc = 0 where k g = Z r / Z s is the base gear ratio, Z s is the number of teeth of the sun gear, Z r is the number of teeth of the ring gear, and ωs , ωr , ωc are the angular velocities of the sun gear, ring gear and carrier, respectively. The motion equations can be expressed as:

J sω& s = η sTs −

J cω& c = Tc +

1 η rTr kg

kg + 1 kg

η cTr

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

35

Figure 34. Planetary-geared EVT system.

where J s is the total inertial torque of the sun gear and its connecting elements referred to the sun gear shaft, J c is the total inertial torque referred to the carrier shaft, Ts , Tr , Tc are the external torques acting on the sun gear shaft, ring gear shaft and carrier shaft, respectively, and η s , η r , η c are the coefficients of internal power losses in the sun gear shaft, ring gear

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

shaft and carrier shaft, respectively.

Figure 35. Planetary gear structure.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

K. T. Chau

36

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 36. Concentrically arranged PM brushless EVT system.

Figure 37. Control strategy of EVT system.

By controlling the power taken by the generator and then feeding back into the motor, the engine speed can be maintained constant when the transmission shaft speed is varying. Thus, a continuously variable ratio between the engine speed and the wheel speed can be achieved (Miller, 2006). Hence, this EVT system takes the following advantages: •

•

•

Because of the absence of clutches or shifting gears, it can significantly improve the transmission efficiency and reduce the overall size, hence increasing both the energy efficiency and power density. In the presence of continuously variable ratio between the engine speed and the wheel speed, the engine can always operate at its most energy-efficient operating point, hence resulting in a considerable reduction of fuel consumption. The system can fully enable the idle stop, electric launch, regenerative braking and full-throttle acceleration features, which are particularly essential for the full hybrids.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

37

However, this planetary-geared EVT system inherits the fundamental drawback of planetary gearing, namely the transmission loss, gear noise and need of regular lubrication. In recent years, active research has been conducted to eliminate this mechanical planetary gear while retaining EVT propulsion. One viable approach is the use of concentrically arranged machines to realize power splitting for the full hybrids (Eriksson and Sadarangani, 2002). Figure 36 shows the configuration of the concentrically arranged PM brushless motor drive to realize EVT without planetary gearing. The primary machine is a double-rotor PM brushless motor, while the secondary machine is an outer-rotor PM brushless motor. They are mechanically coupled by a common shaft, and electrically connected via two power converters. When installing this PM brushless EVT system in a full hybrid, it offers four modes of operation, namely the cranking, charging, launching, and continuously variable transmission (Cheng et al., 2007): • •

• •

In the cranking mode, the battery delivers the power to crank the engine via the primary machine until the engine reaches the speed for ignition. In the charging mode, the battery is either charged by the engine via the primary machine when the vehicle stops motion or by the secondary machine during regenerative braking. In the launching mode, the battery delivers the power to launch the vehicle via the secondary machine without using the engine. In the continuously variable transmission mode, the primary and secondary machines are controlled so that the optimal operating line of the engine can be achieved. As shown in Figure 37, once a specific throttle level is assigned, the optimal operating point Pe (ωe , Te ) of the engine, namely its torque and speed, can be specified. While

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ignoring the EVT power losses, the operating point Pt (ωt , Tt ) along the constant power line is the required speed and torque at the transmission shaft. The EVT control strategy is that the primary machine adopts speed control to change the speed by Δω , while the secondary machine employs torque control to change the torque by ΔT . So, the control strategy keeps the engine operating at the maximum efficiency for all driving conditions. In order to reduce the system weight and size, the two machines can be integrated into a single machine. The key is to share the outer rotor of the primary machine with the rotor of the secondary machine so that the stator is placed concentrically around the outer rotor (Hoeijmakers and Ferreira, 2006). Figure 38 shows the configuration of an integrated PM brushless EVT system. The corresponding principle of operation is the same as the two concentrically arranged PM brushless EVT system. This EVT system takes the definite advantages of highly compact and lightweight. However, it still suffers from the drawbacks of using slip rings and carbon brushes to transfer electric power between the inner rotor and the power converter. It is anticipated that the development of ‘totally brushless’ configurations for the PM brushless EVT system will be a major research direction in the field of hybrid vehicles.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

38

K. T. Chau

Figure 38. Integrated PM brushless EVT system.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5. CONCLUSION In this chapter, an overview of electric motor drives for battery, hybrid and fuel cell vehicles has been presented, with emphasis on motor topologies, drive operations and control strategies. Particularly, two emerging PM brushless motor drives, namely the double-stator PM brushless motor drive and hybrid-PM brushless motor drive, have been analyzed and discussed. Also, three integrated PM brushless motor drive systems, namely the magnetic geared PM brushless drive system for battery or fuel cell vehicles, the PM brushless ISG system for micro and mild hybrid vehicles, and the PM brushless EVT system for full hybrid vehicles, have been revealed and elaborated. It is anticipated that these three integrated systems will be the major research directions of electric propulsion for EVs.

ACKNOWLEDGMENT I would like to express heartfelt thank to Professor C. C. Chan for his invaluable comment on this chapter, as well as Mr. Chunhua Liu, Miss Shuangxia Niu and Mr. Chuang Yu for their tireless effort on this work. Especially, I must express my indebtedness to my wife, Joan Wai Yi, and my son, Aten Man Ho, for their patient and support all the way.

REFERENCES Atallah, K. & Howe, D. (2001). A novel high performance magnetic gear. IEEE Trans. Magnetics, 37, 2844-2846. Chai, F., Cui, S. & Cheng, S. (2005). Performance analysis of double-stator starter generator for the hybrid electric vehicle. IEEE Trans. Magnetics, 41, 484-487.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Electric Motor Drives for Battery, Hybrid and Fuel Cell Vehicles

39

Chan, C. C. & Chau, K. T. (2001). Modern Electric Vehicle Technology. Oxford: Oxford University Press. Chan, C. C., Chau, K. T., Jiang, J. Z., Xia, W., Zhu, M. & Zhang, R. (1996). Novel permanent magnet motor drives for electric vehicles. IEEE Transactions on Industrial Electronics, 43, 331-339. Chan, C. C., Jiang, J. Z., Xia, W. & Chau, K. T. (1995). Novel wide range speed control of permanent magnet brushless motor drives. IEEE Tran. Power Electronics, 10, 539-546. Chau, K. T. & Chan, C. C. (2007). Emerging energy-efficient technologies for hybrid electric vehicles. IEEE Proceedings, 95, 821-835. Chau, K. T., Jiang, J. Z. & Wang, Y. (2003). A novel stator doubly fed doubly salient permanent magnet brushless machine. IEEE Trans. Magnetics, 39, 3001-3003. Chau, K. T., Li, Y. B., Jiang, J. Z. & Liu, C. (2006a). Design and analysis of a stator doubly fed doubly salient permanent magnet machine for automotive engines. IEEE Trans. Magnetics, 42, 3470-3472. Chau, K. T., Li, Y. B., Jiang, J. Z. & Niu, S. (2006b). Design and control of a PM brushless hybrid generator for wind power application. IEEE Trans. Magnetics, 42, 3497-3499. Chau, K. T., Sun, Q., Fan, Y. & Cheng, M. (2005). Torque ripple minimization of doubly salient permanent magnet motors. IEEE Trans. Energy Conversion, 20, 352-358. Chau, K. T., Zhang, D., Jiang, J. Z., Liu, C. & Zhang, Y. (2007). Design of a magnetic-geared outer-rotor permanent-magnet brushless motor for electric vehicles. IEEE Trans. Magnetics, 43, 2504-2506. Cheng, M., Chau K. T. & Chan, C. C. (2001). Design and analysis of a new doubly salient permanent magnet motor. IEEE Trans. Magnetics, 37, 3012-3020. Cheng, M., Chau, K. T. & Chan, C. C. (2003). New split-winding doubly salient permanent magnet motor drive. IEEE Trans. Aerospace Electronic Systems, 39, 202-210. Cheng, Y., Cui, S., Song, L. & Chan, C. C. (2007). The study of the operation modes and control strategies of an advanced electromechanical converter for automobiles. IEEE Trans. Magnetics, 43, 430-433. Deodhar, R. P., Andersson, S., Boldea, I. & Miller, T. J. E. (1997). The flux-reversal machine: a new brushless doubly-salient permanent magnet machine. IEEE Trans. Industry Applications, 33, 925-934. Ehsani, M., Gao, Y., Gay, S.E. & Emadi, A. (2005). Modern Electric, Hybrid Electric, and Fuel Cell Vehicles: Fundamentals, Theory, and Design. Boca Raton: CRC Press. Ehsani, M., Rahman, K. M. & Toliyat, H. A. (1997). Propulsion system design of electric and hybrid vehicles. IEEE Trans. Industrial Electronics, 44, 19-27. Eriksson, S. & Sadarangani, C. (2002). A four-quadrant HEV drive system. IEEE Vehicular Technology Conference Record, 1510-1514. Gan, J., Chau, K. T., Chan, C. C. & Jiang, J. Z. (2000). A new surface-inset, permanentmagnet, brushless DC motor drive for electric vehicles. IEEE Trans. Magnetics, 36, 3810-3818. Hoeijmakers, M. J. & Ferreira, J. A. (2006). The electric variable transmission. IEEE Trans. Industry Applications, 42, 1092-1100. Inderka, R. B., Menne, M. & De Doncker, R. W. A. A. (2002). Control of switched reluctance drives for electric vehicle applications. IEEE Trans. Industrial Electronics, 49, 48-53.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

40

K. T. Chau

Jiang, S. Z., Chau, K. T. & Chan, C.C. (2003). Spectral analysis of a new six-phase polechanging induction motor drive for electric vehicles. IEEE Trans. Industrial Electronics, 50, 123-131. Kim, Y., Kook, Y. & Ko, Y. (1997). A new technique of reducing torque ripples for BDCM drives. IEEE Trans. Industrial Electronics, 44, 735-739. Lo, W. C., Chan, C. C., Zhu, Z. Q., Xu, L., Howe, D. & Chau, K. T. (2000). Acoustic noise radiated by PWM-controlled induction machine drives. IEEE Trans. Industrial Electronics, 47, 880-889. Long, S. A., Zhu, Z. Q. & Howe, D. (2005). Effectiveness of active noise and vibration cancellation for switched reluctance machines operating under alternative control strategies. IEEE Trans. Energy Conversion, 20, 792-801. Miller, J. M. (2006). Hybrid electric vehicle propulsion system architectures of the e-CVT type. IEEE Trans. Power Electronics, 21, 756-767. Niu, S., Chau, K. T., Jiang, J. Z. & Liu, C. (2007). Design and control of a new double-stator cup-rotor permanent-magnet machine for wind power generation. IEEE Trans. Magnetics, 43, 2501-2503. Ostovic, V. (2003). Memory motor. IEEE Industry Applications Magazine, 9, 52-61. Rahman, K. M., Fahimi, B., Suresh, G., Rajarathnam, A. V. & Ehsani, M. (2000). Advantages of switched reluctance motor applications to EV and HEV: design and control issues. IEEE Trans. Industry Applications, 36, 111-121. Soong, W. L. & Ertugrul, N. (2002). Field-weakening performance of interior permanentmagnet motors. IEEE Trans. Industry Applications, 38, 1251-1258. Uddin, M. N. & Rahman, M. A. (2007). High-speed control of IPMSM drives using improved fuzzy logic algorithms. IEEE Trans. Industrial Electronics, 54, 190-199. Wang, T., Zheng P. & Cheng, S. (2005). Design characteristics of the induction motor used for hybrid electric vehicle. IEEE Trans. Magnetics, 41, 505-508. Wang, Y., Chau, K. T., Chan, C. C. & Jiang, J. Z. (2002). Design and analysis of a new multiphase polygonal-winding permanent-magnet brushless dc machine. IEEE Trans. Magnetics, 38, 3258-3260. Yu, C., Chau, K. T., Liu, X. & Jiang, J. Z. (2008). A flux-mnemonic permanent magnet brushless motor for electric vehicles. AIP Journal Applied Physics, 103, 07F103:1-3. Zhan, Y. J., Chan, C. C. & Chau, K.T. (1999). A novel sliding mode observer for indirect position sensing of switched reluctance motor drives. IEEE Trans. Industrial Electronics, 46, 390-397. Zhu, Z. Q. & Howe, D. (2007). Electrical machines and drives for electric, hybrid and fuel cell vehicles. IEEE Proceedings, 95, 746-765. Zhu, Z. Q., Chen Y. S. & Howe, D. (2000). On-line optimal field weakening control of permanent magnet brushless ac drives. IEEE Trans. Industry Applications, 36, 16611668. Zhu, Z. Q., Pang, Y., Howe, D., Iwasaki, S., Deodhar, R. & Pride, A. (2005). Analysis of electromagnetic performance of flux-switching permanent magnet machines by nonlinear adaptive lumped parameter magnetic circuit model. IEEE Trans. Magnetics, 41, 4277-4287.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Electric Vehicles: Technology, Research and Development ISBN 978-1-60741-142-0 Editor: Gerald B. Raines © 2009 Nova Science Publishers, Inc.

Chapter 2

ENVIRONMENTAL FRIENDLY INTER-CITY AIRCRAFT (ENFICA-FC) AND PRELIMINARY ANALYSIS FOR 2-SEAT AIRCRAFT CONVERSION INTO FUEL CELLS POWERED INNOVATIVE SYSTEM G. Romeo∗, G. Frulla!, E. Cestino+, F. Borello+ Politecnico di Torino Dept. of Aerospace Engineering, Italy

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT The main objective of the ENFICA-FC project (ENvironmentally Friendly Inter City Aircraft powered by Fuel Cells), funded by European Commission, is to develop and validate the use of a fuel cell based power system for propulsion of more/all electric aircraft. The following items shall be pursued: a) A fuel cell system shall be designed, built and tested in laboratory ready to be installed on board for flying; b) A high efficiency brushless electric motors and power electronics apparatus for their control shall be designed and manufactured ready to be installed on board for flying; c) high efficiency would be obtained by an optimised aerodynamic propeller design; d) A study of the flight mechanics of the new aircraft will be carried to verify the new flight performance; e) Validation of the overall high performance of an all electric aircraft by means of flight test. The fuel cell system will be installed in a light sport aircraft which will be flight and performance tested as a proof of functionality and future applicability for inter city aircraft. A selection of most suitable aircraft for conversion is presented and the light sport aircraft Rapid 200 chosen. The high efficiency two-seat existing aircraft Rapid 200, manufactured by Jihlavan Aircraft, has been selected over more than 100 light ∗

Full Professor in Airplane Design and Aerospace Structures. C. Duca degli Abruzzi 24, 10129 Turin, Italy. +390115646820, [email protected] ! Associate Professor in Aerospace Structural Analysis and Design + +

Researchers in Aerospace Structural Analysis and Design Researchers in Aerospace Structural Analysis and Design

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

42

G. Romeo, G. Frulla, E. Cestino et al. sport aircrafts after the preliminary reported evaluation based on merit index as indicated. The aircraft will be used for the conversion from internal combustion engine to an electric one. Analysis about the COTS equipments for electric propulsion system has been performed and presented. Design indication of an optimal propeller complete the identification phase that continue with the analysis of some parameters influencing the general configuration of the converted aircraft and the mission items. Preliminary consideration about the definition of storage configuration are presented and some safety issues are considered for H2-gas management. Design indications and conversion limitations conclude the reported activity.

LIST OF NOTATION

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

AC, DC APU C CG COTS DOF FC H2 MTOW L P, PFC PEM Pf R

χ

=alternating or direct current Ri (Re) = internal (external) tank radius =auxiliary power unit T =propeller thrust =shaft torque V =aircraft velocity =centre of gravity Vt = Internal tank volume =component off-the-shelf W (Wt)= weight ( weight of the H2 tank) = degree of freedom h =flight altitude = fuel cell STACK mac = mean aerodynamic chord mH2= H2 mass =H2 = Hydrogen gas =max take-off weight n = load factor = cylinder length p = internal tank pressure =power, power at fuel cell ηFC = fuel cell efficiency = proton exchange membrane λ = advance ratio = performance factor (mH2/Wt) Ω =propeller angular velocity =blade radius ρ = air density = torque coefficient (or power coefficient) τ = thrust coefficient

1. INTRODUCTION Growing interest in the application to renewable energy sources has led to a rapid development of fuel cell technologies based on H2. Much has been said about hydrogen being the "fuel of the future" due to its abundance and its non-polluting combustion products. Less has been said about the fact that other forms of energy must be used to produce the hydrogen which will be used as fuel. Most hydrogen is bound up in compounds such as water or methane, and energy is required to break the hydrogen free from these compounds, then separate, purify, compress and/ or liquefy the hydrogen for storage and transportation to usage points. Widespread production, distribution and use of hydrogen will require many innovations and investments to be made in efficient and environmentally-acceptable production systems, transportation systems, storage systems and usage devices. The successful attempt to adapt H2 technologies to ground transportation opens interesting opportunities for the applications of fuel cell related technologies to aircrafts. Fuel cells could

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

43

become the main power source for small general aviation aircraft or could replace auxiliary power units (APU) on larger aircrafts, obtaining all-electric or more-electric machines. There are several potential advantages in using such a power source, ranging from environmental, and economic issues to performance and operability aspects. During the combustion of kerosene in today's engines, carbon dioxide (CO2), water (H2O) and lesser amounts of sulphur dioxide (SO2), carbon monoxide (CO), nitrogen oxide (NOx) and unused hydrocarbons (HC) are produced; the last three substances are considered to cause greenhouse effect. Large emission of NOx is produced by engine at airport flight phase. Furthermore the engine technology used in the majority of the general aviation aircraft is decades old. Although the public health dangers of alkyl lead compounds, have been greatly reduced since the ban on leaded automotive fuels, the annual use of hundred million litres of leaded aviation gasoline still presents significant hazards to individuals. Clearly the replacement of combustion engines or APU with fuel cells powered electric motors can guarantee a massive reduction of this kind of pollution, since the only emission produced by a fuel cell is water (discussion is open if water vapours emitted at high altitude can produce a long period greenhouse effect too). Air pollution isn’t the only kind of pollution that can be reduced (if not eliminated) through fuel cell driven power systems; noise is indeed considered an important form of pollution produced by aircrafts and it’s particularly important for airplanes taking-off from airports located in urban areas or during night when noise abatement regulations are even more stringent. Nowadays, the fuel cell system power density, defined as the power output per unit weight, is the greatest issue concerning fuel cell application to aerospace; it is estimated that nearly a 20-times increase in power density (20 kW/kg) is required to enable all-electric flight of a large commercial aircraft. Moreover, even if hydrogen contains an amount of energy per unit mass three times higher than kerosene, the significantly lower density of hydrogen leads to the necessity to adopt pressurized or cryogenic fuel tanks; either solution means extra weight and volume. Another key aspect is represented by airline companies independency from oil market. Even if the problem is still controversial, above all about timing, it’s generally believed that a peak in oil production has been reached or will be reached in very few decades [1], the consequence being a fast increase in oil price. This price rise could be detrimental for airline companies if oil is the only available energy source. Hydrogen has the advantage of being obtainable from a variety of sources, so that it’s less prone to market fluctuation and raises. As mentioned above, hydrogen can provide some advantages in performances too; for example an all-electric or more-electric aircraft shows an higher reliability due to replacement of mechanical systems with electric ones. Higher reliability implies also lower maintenance costs. Moreover electric propulsion is much less sensitive to altitude than combustion engines and thus greater altitudes (i.e. smaller power demand) can be reached. Different studies have been undertaken in recent years about fuel cells in transport aeronautics. Boeing is involved in feasibility investigation of innovative APU for large transport aircrafts in order to lower noise and emissions [2] and is currently developing and testing an all-electric two-seater aircraft using a PEM fuel cell as propulsion energy source [3]. A research in this field was also carried out by the Foundation for Advancing Science and Technology Education's (FASTec, USA); FASTec sponsored a research on designing, building, and testing an aircraft powered by fuel cells and advanced rechargeable batteries. Unfortunately, after a promising start, no new information can be found on the state and

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

44

G. Romeo, G. Frulla, E. Cestino et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

development of the project. Airbus is coordinating the European Commission (EC) founded project CELINA [4] whose goals cover a wide area of the topic: to define basic requirements for fuel cell power system integration and to cope with safety, certification, maintenance and installation of such an innovative system; to design an appropriate fuel cell system and related subsystems; to evaluate fuel cell performances when installed on the aircraft; to integrate fuel cells with existing systems. In this framework, European Commission has selected ENFICA-FC as one of the cofounded project in the Aeronautics and Space priority of the 6th Framework Programme. The main goals of the project is to develop a fuel cell based power system for more/all electric aircrafts belonging to different categories and to prove the feasibility of a fuel cell propelled small aircraft by converting an existing combustion engine two-seater airplane. The ENFICA-FC consortium, coordinated by prof. G.Romeo of Politecnico di Torino, consists of 9 partners covering all the needed competences: Israel Aircraft Industries, Evektor, Jihlavan Airplanes, Politecnico di Torino aerospace department, Université Libre de Bruxelles and Università di Pisa are involved in different areas of aircraft design and manufacturing as well as in safety and security issues, Intelligent Energy and Politecnico di Torino energy department are mainly involved in fuel cell power systems development, hydrogen distribution and storage is managed by Air Product while METEC is in charge of the administrative aspect of the project. As mentioned above, two are the key goals of the project: 1. A feasibility study in order to provide a preliminary definition of new and innovative power systems based on different fuel cell technologies; the first necessary step of this study is the identification of non-electrical subsystems which can be properly replaced by electric ones. The investigation concerns also the feasibility from the safety, certification and maintenance point of view. A life cycle cost will be defined too. 2. The conversion of a modern and conventional two-seat aircraft into an all-electric airplane powered entirely by fuel cell. For this application the whole power system need to be re-designed, built and installed in a properly selected commercial aircraft. The ambitious target is to fly-test the converted aircraft by the end of the three years project. Progresses in design of the new fuel cell configuration are presented in the following sections.

2. SELECTION OF THE MOST SUITABLE AIRCRAFT AND DESCRIPTION OF CURRENT PROPULSION SYSTEM AND PERFORMANCE The main objective of the ENFICA-FC project is to develop and validate the use of a fuel cell based on power system for propulsion of more/all electric aircraft. Since newly developed propulsion system is still limited in terms of available power output, only airplanes in Ultralight, VLA or LSA categories may be found suitable for conversion into an electric motor driven aircraft powered by fuel cells. A database of candidates has been established by

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

45

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

available information found on websites, including basic information such as geometrical characteristics, performances and costs [5-8]. The majority of aircraft is from the ULTRALIGHT category, supplemented by several VLA and LSA airplanes [9,10]. Airplanes highlighted in the graphs include: EV-97 Eurostar, Rapid 200 (airplanes produced by consortium members) and Super Dimona HK-36 (airplane selected by Boeing for Fuel Cell application). Limited power available from the fuel cell technology makes weight concern critical. Figure 1 clearly indicates that ultralight aircraft have lower power demands for take off and climb showing that in this case MTOW is the decisive factor. Even airplanes with high aerodynamic efficiency (and high aspect ratio) from VLA and LSA category have worse results in this area. Fig 1a indicates that ultralight aircraft will be more suitable for application where very limited power is available. A databank containing 125 commercial aircrafts was considered for selection of the most suitable one; the selection procedure identifies each aircraft through nine parameters: 1) Aspect Ratio; 2) Operative Empty Weight / Maximum Take-Off Weight ratio; 3) Wing Loading; 4) Stall speed; 5) Take-Off Length; 6) Landing Length; 7) Minimum sink speed; 8) Minimum required power for Level Flight; 9) Conversion Cost.

Figure 1a. (Continued)

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

46

G. Romeo, G. Frulla, E. Cestino et al.

Figure 1a. Selection Procedure and comparison of several parameters between selected aircrafts – Jihlavan Airplane Rapid 200.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1b. Typical Results of the multicriteria analysis.

Figure 2. Overview on Propulsion System Power Electronics.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

47

48

G. Romeo, G. Frulla, E. Cestino et al.

In order to consider all the parameters concurring in the selection procedure multi-criteria analysis, whose flow chart is reported in Figure 1a, has been used. Results for a portion of the analysed aircrafts is reported in Figure 1b. The index merit results shown in Figure 1b have shown Rapid 200 as a possible good candidate for conversion as because of the good performance as well as the advantage to be a Partner’s Aircraft (JIHLAVAN Airplanes). Weights that appear in the flow-chart of Figure 1a were chosen on the basis of different expertise judgement provided by consortium partners with experience in VLA and LSA design: for the conversion analysis, the minimum sink speed, the minimum power at level flight and the conversion cost are the most influencing parameters [11]. The propulsion system configuration of the converted aircraft is reported figure 2. It includes: a) Battery; b) Fuel cell system; c) H2 system; d) DC/DC converter; e) DC/AC converter. The power should be provided from fuel cell stack for 20 kW and an additional 20 kW from batteries. The battery pack should work for about 15 minutes, for boosting the aircraft during take-off and initial climb from 0 to 1000 m. The overall system (motor + batteries + inverter + converter) should weight not more than 120 kg. A hypothetical mission profile includes take-off, climb from the sea level to an altitude of 1000 m (with an average climbing ratio of 2,5 m/s), level flight at 1000 m at about 144 km/h, descent and landing. The total mission time is about 1 h. Considering a level flight condition at 1000m at speed between 100 km/h and 170 km/h, Rapid 200 KP-2U shows a slightly lower value of the required power. At 144 km/h, the net power required for level flight from Rapid 200 is 14 kW; while the power required at fuel cell + battery system is 18 kW. See section VI for more preliminary results about this item.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3. H2 STORAGE SYSTEM DESIGN AND SAFETY ASPECTS A. Design Guideline and Configuration Selection Fuel cell-powered air vehicles have the potential to provide a solution to air quality problems but the availability of H2 on-board, for fuel cells operations, become a key-issue in order to have an acceptable cruise range. Compressed H2 is considered a good near-term solution [12] for motor vehicles also if cryogenic liquid H2 or adsorbed element (metal hydride) can be used. The pressurized tank option is considered in this preliminary design analysis in order to detect an interesting configuration suitable for the scope of the ENFICAFC project. The maximum H2 mass is in fact limited to specific mission requirements (max 1h flight time) pointing out the advantage using pressurized gas than liquid H2. The cylindrical configuration with two closing hemispheres, is investigated due to the fact that its installation on-board seems easier than a big sphere which has higher performance factor (Pf), both for available internal volume and for available shape. The performed preliminary parametric analysis is based on the evaluation of the Pf under different conditions and deals with the identification of specific configuration for installing the tank on board. For this reason some simplifications are introduced and a qualitative derivation is used as a guideline for selection. Elementary formula are applied with the assumption of the thin wall vessels hypotheses. The compressibility parameter of the H2 gas is not included at this stage ([13]) as the calculation are just for the purpose of comparing and choose different solutions

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

49

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

and not in order to evaluate the direct performance parameter that is dependent on other important manufacturing and design items such as materials selection and tank typology (composite lay-up selection, material properties variations, configuration selection according to type 3 or type 4 etc..). The comparisons are made using a standard steel properties. A further investigation will be necessary, including the analysis of existing normative requirements and reliability analysis. Different H2 storage masses are considered referring to different flight energy requirement also if not directly connected to the ENFICA design/mission assumptions. Fixing the H2 on-board mass and the reference temperature, the required internal volume Vt is determined in function of p. As a consequence, Ri is determined from volume (L fixed), and thickness s is calculated by means of a failure criterium. As the thickness s and material density are known the Pf can be determined [14,15,16]. The Pf in function of pressure, for assumed L and fixed H2 mass content , is reported in figure 3a. Initial observations are as follows: for short cylinders (L lower than about 2.0m ) the Pf increases with internal pressure (at fixed L) but after 250atm negligible increment is determined; for long cylinders (L longer than 2.0m) the Pf reveals a maximum and for higher pressure tend to be reduced. Hence, as the installation of pressure tank on-board could be more easy if the pressure is low, it is possible to address the choice to long tank at relatively low pressure.

Figure 3a. Performance factor in function of internal pressure for a fixed H2 mass and different cylinder lengths.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

50

G. Romeo, G. Frulla, E. Cestino et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3b. Performance factor in function of internal pressure for a fixed cylinder length at different H2 masses.

The effect of the stored H2 mass is reported in figure 3b in function of pressure for fixed L. In this case the cylinder is short and the curve is increasing with pressure. The advantage in weight with increasing pressure is quite low for high pressure. The Pf tends to be reduced if H2 mass is increased. The application of a relatively long multi-cylinder tank with limited H2 stored mass could be more efficient than a big short one. Some more important design indications in order to identify the best storing configuration for ENFICA-FC scope are determined: • • •

It is not always convenient to increase the internal pressure for a reduction in volume and looking for a better Pf; As the installation of pressure tank onboard could be more easy if the pressure is low, it is possible to address the choice to long tank at relatively low pressure; a possible choice could be the application of multiple cylindrical tanks containing low H2 mass than a big one storing the total H2 mass.

B. Safety Aspects Preliminary Evaluation and Related Issues In order to successfully implement a hydrogen application, the safe production, storage, transport, handling and use of hydrogen is imperative. Actually, the normative documents, dedicated specifically to the use of hydrogen, are very few. These documents were developed

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

51

primarily for the industrial application. Only in recent years, several organizations such ISO, through its Technical Committee 197 (ISO/TC 197), or the IEC, with the Technical Committee 105 (IEC/TC 105), have intensified the work about the development of codes, standards and guidelines aimed at the project and the exercise of no-industrial application of the hydrogen, such vehicles, stations for the supply and fuel cells. There isn’t a specific standard concerning H2 tank in aeronautics, so it has been resorted to the vehicle field ( [17,18,19] and [20,21,22,23]). Hydrogen has some unique properties compared to other fuels. Some of these unique characteristics can make it safer to work with, while others make it more hazardous. Three main aspects should be considered at first : H2 is a gas, H2 is a compressed (in our case) gas, H2 is a Fuel. The application on board of H2-Gas, requires some investigation about its typical hazards [24.25.26]. Hydrogen gas is colorless, odorless and flammable. It forms flammable and explosive mixtures with air over a wide range of concentrations. Its flammability consists in contact with open flames or electrostatic discharge. H2 gas is very light and rises rapidly in air, if gas is not burning, it may collect in the upper levels of structures, creating an explosion hazard. Auto-ignition temperature is 1058 ˚F (570 ˚C). With respect to Explosion, it is sensitive to mechanical impact. For this reason impact against container must be avoided and electrical equipment in storage areas must be explosion proof. For a given pressure and hole size, hydrogen will leak approximately 2.8 times faster than natural gas and 5.1 times faster than propane on a volumetric basis. Fuel lines, un-welded connections, etc. are potential leakage or permeation sites so they should never pass through the passenger compartment, eliminating the potential for hydrogen concentration into the cabin. The amount of hydrogen present in the fuel cell or internal combustion engine is very small and low pressure when operating, and none while shut down. Fuel cell stacks develop leaks either internally (between flow paths) or externally to the ambient environment. To deal with this potential leakage, fuel cell stacks are typically enclosed and the enclosure is vented with forced air in order to prevent hydrogen accumulation. Poor electrical performance can be considered as an effect of H2 leaks. A gaseous hydrogen plumbing system that is truly leakfree is nearly impossible to build without all welded joints [25], minimizing leaks is obviously desirable condition. In addition, adequate ventilation in the vicinity of the hydrogen system is a must. For this reason a H2 application must include a leak detection system which has to be active also when the vehicle is positioned in a parking/maintenance area. The rapid formation of a combustible mixture is balanced by its faster dispersal and generally shorter duration of a flammable hazard than other fuels on an equal volume basis. Hydrogen is nontoxic, but it can cause asphyxiation in a confined area. Therefore, one of the most important ways to ensure the safe use of hydrogen is to make sure that there is adequate ventilation or specific segregation of H2 areas. Prolonged exposure effect on material behavior is possible. In some high strength steels H2 exposition can cause them to lose their strength (hydrogen embrittlement), eventually leading to failure. Material compatibility can give a great impulse to the safety of the entire system. Some indications about the mode of failure of typical H2 storage system can be derived by [25, 27]. Failure mode in normal operation or/and in collision/impact/crash/emergency can be considered. Possible failure modes in normal operation include: a) Catastrophic rupture due to manufacturing defect in tank, defect caused by abusive handling of tank, chemical etching and destruction of the epoxy resin in one area of the tank, stress rupture; b) large H2 release due to faulty pressure relief device tripping without cause, chemically induced fault in

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

52

G. Romeo, G. Frulla, E. Cestino et al.

tank wall; c) Slow H2 leak due to defect in tank and stress cracks in tank liner du to pressure cycling , faulty pressure relief device, faulty coupling from tank to feed line of first valve connected to tank. Potential collision failure modes are : Catastrophic rupture due to a collision impact, puncture of a sharp object or external fire combined with failure of pressure relief device to open. Large relief of H2 by puncture of sharp object , fire-created hole in tank wall, operation of a pressure relief device in a fire. Slow H2 relief due to fire-induced openings in fuel line connection, impact-induced openings in fuel connection line. Regarding of catastrophic rupture of storage cylinder, cited reference suggests that the fibre wrapped tank probability of rupture would be extremely low. Each tank is tested at 1.5 times its operating pressure and samples of each lot are tested till rupture. Each tank design must be qualified at 2.25 times the normal operating pressure. Each class of tank is also subjected to gunfire and must not explode but leak through the hole. Collision failure should also be unlikely: fibre wrapped tanks must be tested in an equivalent crash at 52mph to solid wall. The tank must survive and not loose pressure. Remaining scenario consists of tank explosion due to : over-pressurization (i.e in fire) and the pressure relief device fails (improbable due to the fact that the device is melting under temperature). The conversion of a two-seat aircraft into a novel all-electric one powered by fuel cell is under investigation as related to the ENFICA Project. The storage system is one part of the energy system of the aircraft and the compressed gas characteristics affect some structural components of the systems itself. The original general aviation aircraft can be included into the CS-VLA category [10]. A preliminary analysis of such a normative can give first indications for design and safety compliance. As the conventional aircraft structures have to be designed according to section C of that normative, in the case of the High pressure components of the energy system, specific articles have to be considered. Specific ultimate inertia load factors are defined under emergency landing condition. They are referred to each item of mass that could injure an occupant if it came loose. H2 tank could be considered such an item and designed accordingly. In particular not only the joints and support of the tank must be designed according to this article but also the specific valves at the top and bottom of the tank and the fittings of the circuit. Furthermore due to the fact that the H2 tank is a basic part of the aircraft, its structure can be viewed as a primary one: if a component is safety critical and for which an adequate safe-life must be evidenced, AMC 527(b).1. could be applied. Subpart D, involves also the manufacturer/supplier of the H2 tank guaranteeing that “each questionable design detail and part having an important bearing on safety in operations, must be established by tests” and so on. Specific requirement of a drop test of the entire aircraft is also indicated in order to check the landing gear system. In this case the compliance of inertia loads shall be demonstrated for the presence of the tank. Ventilation and fire protection have to be evaluated in connection to other items related to fuel valves and controls. With respect to the design and manufacturing of H2 tank installed onboard, specific normative for similar terrestrial constructions must be considered. In this case ISO-TC gives several indications. In the same sense the ASME standard suggests procedure for design pressure vessels. Standards give indications about design and manufacturing and testing pressure vessels for industrial gas and H2-gas. As a guideline for installation of such a vessel, standards impose that the manufacturer must supply a specific report for certification about any items the standard considers applicable. This includes also attachments and supports, Pressure relief devices, Certification of capacity of safety and safety-relief valves, rules governing testing and rules governing inspection. All the pressure vessels shall be provided

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

53

with protection against over-pressure. This protection shall be provided by one or a combination of methods: direct spring-loaded safety or safety-relief valves, indirectly operated safety or safety-relief valves such as pilot-operated ones, Rupture disks. Set pressures are defined according to design pressure, operating temperature and combination of devices.

4. DESIGN OF OPTIMAL PROPELLER The results of an optimized propeller for the ENFICA-FC project are presented. Two algorithms were used: the first allows for the determination of the geometric characteristics of the maximum efficiency propeller for a given operative condition and profile distribution along the blade. The output of this first analysis is the chord and twist angle distributions along the blade, together with its efficiency, torque and thrust for the operative condition. The second algorithm allows for the evaluation of the propeller efficiency, torque and thrust for a given geometry when the blade pitch and operative condition are changed. Both algorithms were developed by the Politecnico of Torino and have been validated by comparison with experimental data [28]. Two coefficients, thrust coefficient and torque coefficient, have been defined,:

τ=

T ρΩ 2 R 4

χ=

C P = since P = C ⋅ Ω 2 5 ρΩ R ρΩ3 R 5

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

(torque coefficient could also be written as function of power consumed by the propeller at the shaft) Another dimensionless parameter is introduced, the so called advance ratio:

λ=

v ΩR

The design of the optimized blade has been carried out in two different conditions: cruise (case A) and climb (case B). The two designs are then compared. In both cases, the design of the optimized blade requires the a priori knowledge of the number of propeller blades: the analysis is carried out for 2 and 3 blades. The optimization algorithm uses an aerodynamic theory based on classical results obtained from the integration of vortex theory, wing theory and momentum theory [29,30]. A compressibility correction is used introducing a semiempirical factor which corrects the lift and drag coefficients. Experimental code validation is illustrated in [31]. The evaluation of the aerodynamic performance of the propeller and its optimisation requires the detailed knowledge of the airfoil used for the blades which should provide the values of drag and lift coefficients as function of the angle of attack and Reynolds number. A ClarkY profile is selected for the first iteration of propeller design. An aerodynamic database has been created as function of the angle of attack [-10° /+25°] and Reynolds number [from 103 up to 5 x 106]. The following parameters are adopted for designing the blade in cruise condition: h=1000m;

V= 40m/s;

Ω=2000rpm;

T=340N;

R=0.8m;

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

blade=2

54

G. Romeo, G. Frulla, E. Cestino et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

a)

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

b)

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

55

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

c)

d) Figure 4. Blade optimized for the cruise conditions. A) Chord and twist angle distributions vs dimensionless radius. B) Thrust coefficient vs advance ratio C) Efficiency vs advance ratio D) Torque coefficient.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

56

G. Romeo, G. Frulla, E. Cestino et al.

The chord and twist angle distributions versus the dimensionless radius for a blade optimized for the cruise condition are shown in figure 4. In these conditions the efficiency of the propeller reaches the 90%. The optimal collective pitch angle is equal to 22.8 deg. Using the second algorithm [28], it’s possible to evaluate the behaviour of the propellers outside the design condition. The propeller is considered fixed-blade: the collective pitch angle is fixed and equal to the optimal one. This means that when the flight condition changes, the angular velocity of the propeller (and so the angular velocity of the engine) has to change. However, also different collective pitch angles have been taken in account. Efficiency versus advance ratio is also plotted in Figure 4. Results about three blade propeller in cruise condition and two and three blade propeller optimized in climb condition are not reported because differences are quite light; however, the two blades propeller seems to have a slight advantage in term of weight and it is easier to built and cheaper then a three blades one. The analysis has shown the hypothetical possibility to obtain a 90% efficiency propeller for cruise condition and about 80% for climb condition. This results are quite good if compared with the maximum theoretical efficiency that it’s possible to reach in such conditions.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5. PRELIMINARY PERFORMANCE AND ANALYSIS OF THE AIRCRAFT STABILITY Starting from the basic empty weight configuration and adding the new electric and energy system, the new aircraft Centre of Gravity (CG) position would be determined in an attempt to maintain it within the conventional value. When possible the solutions suggested by the energy system provider shall be maintained; however, since the CG & flight performances shall depend from the FCS system, it should be necessary to carry an interaction process in order to maintain or improve the best performances of the aircraft. The preliminary 3D drawing of the aircraft as well the analytical study of CG and flight performances of the aircraft have been started. A flight dynamic analysis has been performed, in the preliminary design phase, to have a first estimation of the airplane Handling Qualities and a 6DOF mathematical model has been therefore used to describe the flight dynamics of the aircraft [32,33]. The approach taken in analyzing the dynamic behaviour of the airplane was to use the solutions of the uncoupled, linearized equations of motion. To obtain the equations, a MATLAB-SIMULINK flight dynamic software was used. For each flight condition, after the trim conditions have been determined, it is possible to obtain the aircraft response to any step impulse of flight control surfaces. A maximum weight of 550 kg has been considered for a preliminary analysis of the aircraft stability. Two flight conditions have been considered: cruise and climb; a steady flight at 1000m at an average velocity of 40 m/s and a climb flight at 500m at an average velocity of 30 m/s, with a ramp angle of 4,0 deg. In both cases, the response of the aircraft at a flight control impulse has been evaluated at 1 deg impulse of elevator, for the longitudinal stability, at 1 deg impulse of aileron and of rudder for the lateral/directional stability. Eingenvalues, damping and frequencies for the longitudinal and lateral/directional stability are determined Table 1. The assumptions made in the analysis are that the effects of lateral motion on the

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

57

aerodynamic and propulsion forces and moments associated with the lift, drag and thrust forces are negligible. Of course, in the model linearization, we also assume that the motion of the vehicle is undergoing small changes in the variables ΔV (airspeed), Δα angle of attack), Δq (pitch rate) and Δθ along with small inputs in the elevator deflection control Δ e. Characteristic roots of the longitudinal model are given by the eigenvalues of the system matrix considering a steady flight at 1000 m. We can distinguish two basic modes: the phugoid mode and the short-period mode. The phugoid mode is the one that has the longest time constant and is usually lightly damped. Responses of the aircraft that are significantly affected by the phugoid mode, are the velocity and the pitch attitude. There is very little response seen in the angle of attack when the aircraft is excited in the phugoid mode. The short-period mode is displayed in the motion of the airplane Δα and Δq. It has a relatively short time constant (hence the name short-period). In most situation, this mode is relatively well damped. All the longitudinal modes (short and long period) are stable (Table1). In figure 5a is also reported the response of the aircraft evaluated at 1 deg impulse of elevator. According to the previously indicated assumptions, the motion in the lateral axis is decoupled from the longitudinal dynamics. The lateral dynamic model is described by a set of 4 linear ordinary differential equations in the variables Δβ side slip angle), Δp (rolling angular velocity) Δr (yaw angular velocity) and Δφ bank angle. There are 3 modes associated with the lateral motion analysis: The spiral mode is a slow mode that is associated with a real root depicting predominantly motion in the roll attitude. Its value is significantly affected by the damping in roll from the rolling moment due to roll rate. In cruise condition this mode may even be unstable; but since it is a very slow mode, the pilot can interact and correct satisfactorily for the spiral instability. The Dutch-roll mode is an oscillatory mode with significant components in the yaw and the roll variables. The roll mode is usually associated with a real root which is very stable. The motion is predominantly in roll rate and settles down quite quickly. In Figures 5b time responses in the lateral motion to a separately applied impulse input at the aileron control surfaces are shown. One lateral mode is not stable but it has an high period, which means that it could be easily controlled The same result was obtained for directional modes. It is possible to conclude that as expected, the aircraft has a stable behaviour in cruise condition, but it needs to be controlled by the pilot, especially in the lateral manoeuvres. In climb conditions, the aircraft is stable in its longitudinal plane, but it’s not stable in its lateral plane: however, since the period of its unstable modes is very high, it could be easily controlled. Table 1. Longitudinal and lateral dynamic: Eigenvalue analysis results Longitudinal Dynamic -1.18E-2±2.72E-001i -2.48+3.05i

Type Phugoid Short period

Lateral Dynamic 3.63E-003 -5.7E-001±3.75i -1.10E+001

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Type Spiral Mode Dutch-roll Mode Roll Mode

58

G. Romeo, G. Frulla, E. Cestino et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

a)

b) Figure 5. a) Response of the aircraft at 1 deg impulse of elevator; b)Response of the aircraft at 1 deg impulse of aileron.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

59

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

6. PARAMETRIC ANALYSIS OF ENFICA-FC DEMONSTRATION MISSION A preliminary parametric analysis has been carried out to analyze which improvements, in the ENFICA-FC mission, would be possible as efficiency and performances of few elements increases. At moment only two parameters have been take into account: H2 tank storage efficiency (the ratio between the H2 mass [kg] and the sum of H2 and tank mass [kg]) and power density of the fuel cell stack. The efficiency of the FC stack is maintained in this case fixed to 0.4. Updated evaluation has to be considered with different and improved efficiency. The needed power of each mission phase is depicted in Figure 6. This preliminary analysis interests three quantities: the volume occupied by the power plant (batteries, converter, inverter, FC stack and the electric motor), the payload and the cruise time improvement (endurance improvement). Clearly the latter two aspects can be traded-off to modify the demonstration mission profile. Quantities not involved in the parametric analysis are chosen on the basis of data obtained from already developed technologies; some COTS evaluation are reported in summary ([34,35] for more details). The whole Power Supply Drive system is envisaged to consist of: 1) A fuel cell producing about 20-22kW , 2) Batteries for take off and climbing , 3) a DC/DC converter system managing fuel cell and batteries , 4) a DC/AC inverter driving the electric motor , 5) an Aircraft Power Management Unit and Distribution. A Li-ion battery pack is considered with an energy density of 100 Wh/kg and weight less than 50 kg (expected cycle-life at least 400 cycles). Since the FC stack and the batteries have an output voltage which depends from the time, a DC/DC converter is necessary to stabilize the input characteristics in the DC/AC inverter. The total weight of the DC/DC converters should be equal or less of 15 kg and should have efficiency equal or higher than 95%. The total weight of the DC/AC inverters should be equal or less of 15 kg and should have an efficiency equal or higher than 95%. The electric brushless motor should have the following characteristics: Angular velocity between 1500 and 2500 r.pm, Efficiency higher than 95%, Maximum continuous torque less or equal 250 Nm, Weight: less or equal 30 kg. A weight of 36 kg is expected for inverter, converter and power management unit with a volume of about 0.017 m3. For a given battery voltage and stack voltage/current curve, the power flow is a function of control variable of this device (it is interfaced with the Aircraft Management Unit as a function to the pilot commands). Table 2. COTS Components data Battery energy density (including battery management unit) Battery average density Power distribution power density (converter, inverter) Power distribution system average density Engine weight Engine volume Maximum power at the FC output

100 2000 2500 2000 35 0.028 20000

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Wh/kg kg/m3 W/kg kg/m3 kg m3 W

60

G. Romeo, G. Frulla, E. Cestino et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 6. Power required at the FC stack and batteries.

A summary of the main components characteristics chosen for simulation are reported in table 2. In Figure 6, a typical mission profile as considered in the Enfica-fc project, is reported. As it’s possible to notice, since the maximum power at the FC stack is 20kW, the battery have to supply the extra required power during the take-off and the climb phase. It appears that the battery have to supply a maximum continuous power of 17 kW (during the take-off). Using this information and considering a possible 5 minutes long emergency flight entirely powered by battery, the battery pack is designed. The power distribution system (inverter, converter and power management unit) is designed, considering a peak power of the system of 40 kW. It is possible to extrapolate the power required only from the FC in the different mission phases (notice that when the power required at the power plant is greater than 20 kW, the power supplied from the FC is 20 kW, otherwise the stack supplies all the power required). Using this data and assuming a constant FC efficiency (0.4), the H2 [kg] needed is evaluated with the following equation [16]:

mH 2 =

PFC ⋅ t 32166.67 ⋅η FC

where the mass is expressed in kg, power in W, t in hours and ηFC. is the fuel cell efficiency. A 4% reserve is considered. With this assumptions the H2 mass for the basic mission results of 1.15 kg. The storage pressure [atm] has been also changed in order to evaluate the power plant volume (Figure 7). It’s observed that, excluding liquid H2, there is a negligible gain in volume for pressure above 250 atm. For a fixed storage pressure, there is a strong reduction in the power plant volume for relative low values (200 and 1000 W/kg) of the FC power density. After 1000 W/kg the slope of the curve decreases and there is not a great gain in volume.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 7. Power plant volume vs FC power density (function of storage pressure – atm).

a)

b) Figure 8. a) Payload vs FC power density (function of H2 tank storage); b) Payload against H2 tank storage efficiency for different FC power density [W/kg].

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

61

62

G. Romeo, G. Frulla, E. Cestino et al.

The aircraft payload is analysed, as function of the H2 tank storage efficiency and the fuel cell power density (Figure 8a, Figure 8b). It seems that exists a limit: for FC power density greater than 1000 W/kg and H2 storage efficiency greater than 16% the payload increase is negligible. Under the assumption to devote all the available weight to hydrogen, the cruise time is analyzed. It grows with the storage efficiency in a linear way but it has an asymptotic behaviour versus the FC power density; for FC power density greater than 1500 W/kg the endurance increase is negligible. It appears that in this case it’s more convenient to increase the H2 storage efficiency than the FC power density (Figure 9a, 9b). Finally, the dependence of the three parameters is reported in Figure 10a and Figure 10b. All of the items included into the propulsion system have some specific safety problems, different from those indicated in section.III for H2 management. Such evaluation will be included in a subsequent activity of FMECA/FTA not reported here.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

a)

b) Figure 9. a ) Cruise time vs FC power density for different storage efficiency [%] ; b) Cruise time vs storage efficiency for different FC power density [W/kg].

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

a)

b) Figure 10 a,b. Variation of payload and endurance capability for ENFICA-FC mission.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

63

64

G. Romeo, G. Frulla, E. Cestino et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CONCLUSIONS The conversion of a two-seats general aviation aircraft has been selected and analyzed for installation of a fuel cell power system. A statistical analysis has been presented from the survey of more than 100 airplanes. On the strength of the comparison between these aircrafts, Rapid 200 has been considered the aircraft with the best characteristics for the conversion into electric-motor-driven airplane powered by fuel cells. The low power required at low velocity (about 144 km/h) for a level flight has been evaluated an important parameter for the selection. A preliminary evaluation of some safety aspects and design indication about H2gas management and installation have been pointed out and design indication about the storage configuration assessed. Optimal propeller design and result has been performed and presented including design point indication and performance data. The analysis has shown hypothetical possibility to obtain a 90% efficiency propeller for cruise condition and about 80% for climb condition. A confirmation of such kind of optimal solution is expected from improved CFD analysis that is under investigation including the effect of nacelle and venting systems for a validation of this such an efficiency. A new preliminary aircraft Centre of Gravity position has been determined, starting from the basic empty weight configuration and adding the new electric and energy system, in an attempt to maintain CG within the conventional value. Since the CG & flight performances shall depend from the several FC elements position, an interaction process is in progress in order to maintain or improve the best performances of the aircraft. The preliminary 3D drawing of the aircraft as well the analytical study of CG and flight performances of the aircraft have been started. A mathematical model has been therefore developed to describe the flight dynamics of the aircraft to obtain a first estimation of the airplane Handling Qualities. As expected, the aircraft has a stable behavior in cruise condition but it needs to be controlled by the pilot, especially in the lateral maneuvers. More precise evaluation will be made when the CG position of the modified aircraft will be evaluated in more detail. The preliminary parametric sizing carried as function of FC power density and of the H2 tank storage efficiency has been shown that future developments in FC technology and storage tank could drastically improve the aircraft performances, surpassing the piston engine powered one. However, it’s necessary to consider also other parameters, such as the FC efficiency, the power plant (converter, inverter, batteries and engine) power density and the position of the various component inside the aircraft and how they influence the balance and stability. The manufacturing or acquisition of all the elements (aircraft structure, FC system, power supply drive, power management unit, etc.) are being acquired by the several partners composing the ENFICA-FC consortium aiming to the assembling in the aircraft for a first flight test in Summer 2009.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Environmental Friendly Inter-City Aircraft and Preliminary Analysis…

65

ACKNOWLEDGMENTS The authors acknowledge the important contribution of European Commission by the funding programmes ENFICA-FC, EC 6th FP – Contract No. AST5-CT-2006-030779. The authors acknowledge the contribution of Enfica partners.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[10] [11]

[12] [13] [14]

[15]

[16] [17] [18] [19]

NETL, “Peaking of World Oil Production: Recent Forecasts”, DOE/NETL2007/1263. Glover, B., “Fuel cells opportunity”, Proceedings of AIAA/AAAF Aircraft Noise and Emissions Reduction Symposium, Monterey, USA, May 2005. http://www.boeing.com/phantom/news/2007/q1/070327e_nr.html Volker H., “Airbus Fuel Cell Systems For Aeronautic Applications. A Clean Way from Kerosene to Energy”, Proceedings of 25th ICAS Congress, Hamburg, D, Sept. 2006. Taylor, J.,” Jane’s All the World Aircraft”, Jane’s, London, England, UK, 1976. Raymer, D.P., “Aircraft Design: A Conceptual Approach”, Education Series, AIAA, USA, 1989. Roskam, J., “Methods for Estimating Drag Polars of Subsonic Airplanes”, published by Author, 1971. Lan, C., and Roskam, J., “Airplane Aerodinamics and Performance”, Roskam Aviation and Engineering Corp., Ottawa, KS, 1980 LAMAC, “Design Standards for Advanced Ultra-Light Aeroplanes”, Les Cedres, Canada. 2001 CS-VLA, “EASA Certification Specification for very Light Aeroplanes”. Romeo, G., Frulla, G., Moraglio, I., Cestino, E., Borello, F., and Novarese, C., “EnficaFC Deliverable D 6/1 Internal Technical Report”, Politecnico di Torino, Torino, Italy, 2007. Thiokol Propulsion: “ High-pressure Conformable Hydrogen Storage for Fuel Cell Vehicles” Proceedings of 2000 Hydrogen Program Review – NREL/CP-570-28890. Harvey J.F. “ Theory and design of pressure vessels”. Van Nostrand Reinhold Company. NY 1980. Neel Sirosh, “Hydrogen Composite Tank Program “, Proceedings of the 2002 U.S. DOE Hydrogen Program Review NREL/CP-610-32405. QUANTUM Technologies WorldWide 17872 Cartwright Road, Irvine, CA 92614 Mitlitsky F., Weisberg A.H., Myers B.,” VEHICULAR HYDROGEN STORAGE USING LIGHTWEIGHT TANKS”. Proceedings of the 2000 U.S. DOE Hydrogen Program Review NREL/CP-570-28890. Colozza J.A., “ Hydrogen Storage for Aircraft Applications Overview”. NASA/CR – 2002-211867. EN 12245. ISO 15869. Klein, G., Zapf, B., “Hydrogen application in the aircraft sector”, Lisbon, November 2004.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

66

G. Romeo, G. Frulla, E. Cestino et al.

[20] Chaudourne, S., Tombini; C., Perrette, L., Junker, M., ”Standardization and regulation on hydrogen systems in Europe and in the world present status and future developments”. [21] Randy Dey, “ISO/TC 197 activities”, Grenoble, September 2003. [22] ISO 15916 – Basic considerations for the safety of hydrogen systems. [23] ISO 23 – Business plan of ISO/TC 197 hydrogen technologies. [24] J.L. Alcock, L.C. Shirvill, R.F. Cracknell “Compilation of Existing Safety Data on Hydrogen and Comparative Fuels”. EIHP2 - Contract N° ENK6-CT2000-00442 WP N°:5 Rep. Date: July 2001. [25] Hydrogen Fuel Cell Engines and Related Technologies. College of the Desert. Palm Desert, CA, USA. 2001 [26] “Hydrogen: The fuel of the future”, Plug-Power Inc. 2000. [27] DOE/CE/50389-502 May 97. [28] D’Angelo, S., Berardi, F., and Minisci, E., “Aerodynamic Performances of Propellers with Parametric Considerations on the Optimal Design”, The Aeronautical Journal, Vol. 106, No. 1060, June 2002. [29] Pistolesi, E., “Aerodinamica”, UTET, Torino, Italy, 1932. [30] Goldstein, “On the Vortex Theory of Screw Propellers”, Proc Royal Society, 123, 1929. [31] Romeo, G., Frulla G., Cestino, E., and Corsino, G., “Heliplat: Design, Aerodynamic, Structural Analysis of Long-Endurance Solar-Powered Stratospheric Platform”, Journal of Aircraft , Vol. 145, No. 24, 9 Dec. 1996, pp. 44-46. [32] Etkin, B., “Dynamics of Atmospheric Flight, 1st ed.”, Wiley, New York, 1972. [33] Perkin, C.D., and Hage,R.E., “Airplane Performance Stability and Control”, 1st ed., Wiley, New York, 1960. [34] G. Romeo, I. Moraglio and C. Novarese , “ ENFICA-FC: Preliminary Durvey & Design of 2-seat Aircraft powered by Fuel Cells electric propulsion “. 7th AIAA Aviation Technology, Integration and Operations Conference (ATIO) 18-20 September 2007, Belfast , Northern Ireland. (AIAA-2007-7754). [35] G. Romeo, F. Borello, E. Cestino, I. Moraglio, C. Novarese , “ ENFICA-FC: Environmental Friendly Inter-City Aircraft and 2-seat aircraft powered by Fuel Cells electric propulsion “. AIRTEC 2nd C Int. Conference – Supply on the Wings”, 24-25 October 2005, Frankfurt/Main, Germany.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Electric Vehicles: Technology, Research and Development ISBN 978-1-60741-142-0 Editor: Gerald B. Raines © 2009 Nova Science Publishers, Inc.

Chapter 3

CLARIFYING THE DEBATE ON ELECTRIC TWO-WHEELERS IN CHINA Christopher Cherry* University of Tennessee–Knoxville, Knoxville, TN 37996-2010, USA

ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Electric two-wheelers have entered the ever-expanding transportation market in China, fully penetrating the market in most cities and rural areas. Over the past five years, more than 50 million electric two-wheelers have been sold in China. Electric twowheelers have surpassed bicycle ridership in many cities, moving this mode beyond filling small niches that electric vehicles have historically filled in developed nations. Electric two-wheelers are used for many types of transportation, from commuting to goods delivery to law enforcement. This chapter outlines the recent historic growth of electric two-wheelers and scooters in China, discussing several of the controversial issues surrounding them, including safety, congestion, environmental performance, and competition with public transit. Electric two-wheelers provide high levels of personal mobility, matching that of the automobile. They do so with much fewer of the environmental, congestion, and safety externalities associated with personal cars. Currently, though, most electric two-wheeler users would otherwise ride a traditional bicycle or bus, obscuring the relative impact of electric two-wheelers on China’s urban transportation systems. When balancing electric two-wheelers’ effect on mobility and access, environment, safety, and congestion in the entire transportation system, they debatably provide net benefits. This chapter clarifies those debates and frames the challenges in ways that can be quantified.

*

Assistant Professor. Civil and Environmental Engineering, University of Tennessee-Knoxville, 223 Perkins Hall, Knoxville, TN 37996-2010. Phone: 865-974-7710; Mobile: 865-684-8106; Fax: 865-974-2669; http://web.utk.edu/~cherry

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

68

Christopher Cherry

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

INTRODUCTION Electric two-wheelers, also referred to as electric bikes (e-bikes) or light electric vehicles (LEVs), have emerged into the Chinese transportation market in a significant way in the past decade. With modest beginnings in the late 1990s, electric two-wheelers have become very popular since 2002. This rapid rise in popularity was due to a number of forces converging, including rapid urbanization and city expansion, a legacy of bicycle ridership, increased multi-worker families, and restrictions on gas motorcycle use. In addition to economic or cultural factors, electric two-wheeler technology underwent some significant advances in motor, battery and controller technology (Weinert 2007; Weinert, Ma et al. 2007). With about 20 million electric two-wheelers manufactured in 2007, it is estimated that over 40–60 million electric two-wheelers now operate on China’s roadways (Figure 1). This unprecedented growth has forced many urban areas to quickly develop policies to support or restrict electric two-wheeler use. Some cities, such as Shenzhen and Guangzhou, have banned electric twowheelers, citing safety, congestion and environmental problems, while others have fully supported them (Fairley 2005; Guangzhou Daily 2006). Others have taken a more targeted approach, restricting those that are not technically classified as bicycles; most notably, Chengdu has recently announced a licensing scheme to phase out faster electric scooters from operating in the city center. The central government attempted to classify electric twowheelers early in their development. They mandated that they would be less than 45 kg and slower than 20 km/hr, effectively making them electric bicycles (China Central Government 1999). With these design mandates, they were later given the same rights to the road as bicycles (China Central Government 2004). Most notably, they could operate in the bicycle lane and they are exempt from helmet and licensing laws, reducing barriers to use. This chapter will provide a background on electric two-wheeler growth in China and clarify some of the arguments for and against them (CHR Metals 2008). The following sections will focus on clarifying some of the largest debates currently underway in relation to the economics of travel, environmental performance, safety, congestion, and mobility and accessibility. Results of this work are largely drawn from the author’s dissertation and other subsequent work (Cherry 2007; Cherry and Cervero 2007; Weinert, Ma et al. 2007; Cherry, Weinert et al. 2008).

ELECTRIC TWO-WHEELER TECHNOLOGY DEFINED Electric two-wheelers in China are produced by hundreds of companies producing relatively low volumes of electric two-wheelers and often other products. Many of the largest electric two-wheeler companies are subsidiaries of bicycle companies or of companies that produce household electric appliances. Still others are solely electric two-wheeler manufacturers. There are some clear leaders in the industry, but these leaders still only individually produce 5–10 percent of the Chinese market. One of the key factors currently being discussed by central government policy makers is how to organize the industry to establish healthy and strong competition that encourages the production of high-quality vehicles.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Clarifying the Debate on Electric Two-Wheelers in China

69

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Electric Two-Wheeler Production.

The primary components for most electric two-wheelers in China are an in-hub motor (350–1500W) powered by a set of lead acid batteries (36–72V) with the electric current regulated by a controller. The primary reason for the large number of entrants to the electric two-wheeler production market is the open modular (OM) characteristics of the industry, discussed by Weinert et al. ( 2007). This design allows component makers to develop standalone modules that can then be used by any number of electric two-wheeler companies to assemble an electric two-wheeler. Most electric two-wheeler companies simply assemble modules from component suppliers, and the finished vehicle is branded by the company. Because of this structure, many electric two-wheelers have the same design, but are assembled and branded by different companies. This industrial structure has resulted in potentially ruinous competition, low profit margins, and low investment in research and development (Weinert, Ma et al. 2007). Electric two-wheelers primarily come in two varieties, with a spectrum of designs. They can be classified as bicycle style electric bikes (BSEB) or scooter style electric bikes (SSEB) (Jamerson and Benjamin 2007), as shown in Figure 2. It is important to note that the motive technology of these vehicles do not differ significantly. The body design varies as well as the scale of components, but the main components are the same (motor, battery, and controller). In general, SSEBs have larger batteries, higher-powered motors and, as a result, are heavier, faster, and have more range. Early-technology electric two-wheelers had a maximum speed of 20–30 km/hr and a range of 35–50 km. The most recent technology electric two-wheelers can exceed 40 km/hr with an advertised range of over 100 km. Many SSEBs have safety features that are not found in most BSEBs, including headlights, taillights, turn signals, horns, mirrors and even disk brakes. These features add to the weight of the vehicle, but also improve its safety if those features are used appropriately.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

70

Christopher Cherry

Figure 2. Bicycle Style Electric Bike (BSEB) and Scooter Style Electric Bike (SSEB).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Electric two-wheelers can provide high levels of mobility and thus accessibility in many Chinese cities, where trip lengths are increasing and residents are engaging in more complex trip-making activities. City policy makers are asking if this improved mobility is worth the perceived costs in terms of added congestion, environmental impacts, and safety. For each city, the answer to this question depends on predominant alternative mode shift, the energy mix of electricity generation, and the city’s environmental, safety, and mobility goals. Traditional bicycles have high environmental performance on almost all metrics, a good safety record, and provide high levels of mobility but with limited range. If policy can induce a shift toward bicycles, then that could be a good solution provided the city is organized to make destinations accessible by bicycle. Buses are environmentally better than electric two-wheelers on some metrics and worse on others. What environmental advantages buses have, they lack in mobility. Electric two-wheelers prove a balance of car-like personal mobility coupled with low energy use and emissions. One main advantage of electric twowheelers is in the extent to which they delay or replace car use or ownership. Even small shifts from electric two-wheelers to cars result in major impacts.

USER BEHAVIOR There have been several published surveys of electric two-wheeler users in the past several years. In 2006 electric two-wheeler and bicycle riders were surveyed in Kunming, Shanghai, and Shijiazhuang. Detailed results are presented in Cherry and Cervero (Cherry and Cervero 2007), Weinert et al. (Weinert, Ma et al. 2007), and Lin et al. (Lin, He et al. 2008). Again in 2008, electric two-wheelers were surveyed in Kunming using a similar instrument used in 2006 by Cherry and Cervero (Cherry and Cervero 2007). Electric two-wheeler users were asked a number of questions to gauge their use patterns and identify any significant changes over time. While these surveys asked many detailed questions related to use behavior, perceptions, and demographics; perhaps the most policy relevant question was to determine the “next best” transportation alternative for electric two-wheeler users to determine if electric two-wheeled vehicles are taking ridership from bicycles, buses, or personal cars. Figure 3 summarizes the results of this question from these surveys.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Clarifying the Debate on Electric Two-Wheelers in China

71

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. Alternative Mode Preference of Electric Two-Wheeler Users.

Clearly, in two of the three cities surveyed, electric two-wheeler users would otherwise use the bus for the majority of their trips, followed by bicycle. This finding is inconsistent with expectations that electric two-wheeler users are would-be bicycle riders. When asked what mode they used before buying an electric two-wheeler, a higher proportion of the respondents stated that they previously rode a bicycle, indicating that electric two-wheelers might be taking riders from bicycles and disrupting their transition into buses. Kunming and Shanghai are known for their high quality public transit services, indicating that bicyclists have already transitioned to bus, while Shijiazhuang still has very high bicycle ridership and perhaps poorer transit service. While pulling riders from transit and bicycles might be an undesirable outcome, it is worth noting that some 10–15 percent of riders would otherwise use a personal car or taxi. This car displacement provides great economic, environmental, and congestion benefits that could outweigh the dis-benefit of diverting bus and bicycle users. Vehicle ownership and future purchase plans were asked of Kunming electric twowheeler users and the results show that household electric two-wheeler and car ownership has remained about constant over the past two years, 1.2 electric two-wheelers and with 0.25 cars per household (Figure 4). Motorcycle and bicycle ownership has gone down slightly indicating that electric two-wheelers have replaced motorcycles. It is difficult to infer what effect electric two-wheelers have in redirecting the current trends of automobile ownership growth, but it is promising to see auto ownership maintain a constant level while incomes have risen among electric two-wheeler using households. None-the-less, some 20 percent of electric two-wheeler using households plan on purchasing a car in the next year, producing consistent findings as Weinert et al. (Weinert, Ma et al. 2007).

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

72

Christopher Cherry

Figure 4. Electric Two-Wheeler Household Vehicle Ownership and Purchase Plans.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ECONOMICS Electric two-wheelers provide one of the lowest-cost forms of mobility, with operating costs averaging about 0.06 RMB/km. Over half of the operating costs are in battery purchases, about 0.03 RMB/km. The total average user cost (including purchase price) ranges from 0.10–0.12 RMB/km (Jamerson and Benjamin 2004). To put this in perspective, Kunming’s bus fare is 1.5 RMB (including an average of 0.5 transfers) with an average trip length of 3.6 km. This is about 0.42 RMB/km. Shanghai’s bus trips are more expensive, about 3 RMB (including transfers), with an average trip length of 5.2 km or about 0.57 RMB/km (Cherry and Cervero 2007). These user costs are 4–5 times the user costs of electric twowheelers. This does not include subsidies to either system (operating costs, capital costs, and road allocation), which have a tendency to be very high for public transit systems. Taxi trips are even more expensive, well over 2 RMB per kilometer in most cities. Traditional bicycles, because of very low purchase and operating costs, are the most cost effective form of mobility. The low user costs of electric two-wheelers could result in personal budget reallocation increasing consumption in other areas, such as food, health care, or housing, presumably improving their overall quality of life.

ENVIRONMENT The electric two-wheeler industry is subject to larger industries in which it has little influence, such as raw material production and electricity generation. As such, many of the environmental impacts of the electric two-wheeler are unchangeable by the electric two-

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

73

Clarifying the Debate on Electric Two-Wheelers in China

wheeler industry itself. Electric two-wheeler emission rates and those of competing modes are shown in Table 1. Electric two-wheeler emission rates are subject to power plant emission rates. Most of the production impacts are a result of larger production processes, like steel and lead production. Most of the lead pollution occurs during mining, production or recycling processes, in which electric two-wheeler manufacturers have little influence. If governments exert pressure on the electric two-wheeler industry to clean up, there is little the electric twowheeler industry can do to improve some of the most difficult challenges they are facing, particularly SO2 emissions from power plants; lead (Pb) pollution from mining, smelting and disposal processes; and SO2 and PM emissions from steel and lead production processes. As the electric two-wheeler industry is growing, they are exerting more influence, for better or worse, on major industries like battery production. The use of electric two-wheelers is somewhat unique because the spatial distribution of environmental externalities is different than most motorized modes.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 1. Lifecycle Environmental Impacts per Passenger Kilometer Traveleda,b PM (g/paxkm) 0.09-0.28

CO (g/paxkm) 3.4-10.1 0.080.01-0.04 0.04-0.14 0.32e

Card

Energy Use CO2 (kWh/100 (g/pax-km) pax-km) 47-140 102-306

Bus

8.7-26.2

24.2-96.8

Motorcycle

21-42

64-128

0.04-0.08 0.20-0.40 6.3-12.5e

Bicycle

4.88

4.70

0.01

BSEB

3.8-7.6

15.6-31.2

SSEB

4.9-9.9

20.2-40.5

SO2 (g/paxkm) 0.23-0.69

0.06

Unkn 0.0070.07-0.14 0.07-0.14 0.014e 0.0090.09-0.17 0.10-0.19 0.017e

HC (g/paxkm) 0.57-1.67 0.0080.030e 1.132.25e Unkn 0.0270.053e 0.0320.064e

NOX (g/paxkm) 0.44-1.32 0.140.54e 0.080.15e Unkn 0.0100.020e 0.0140.027e

Pbc (mg/paxkm) 18-53 1-4 16-32 0 145-290 210-420

a

Assuming lifespan of 1,000,000 km, 20,000 km, 60,000 and 50,000 km for bus, bicycle, motorcycle and electric bike, respectively. b Ranges indicate assumed average load factors of 1–3 pax for car; 25-75 pax for bus; 1 pax for bicycle; 1–2 pax for motorcycle; and 1-2 pax for electric bike. Notably, multiple passengers on e-bikes are illegal in many cities and energy use requirements are raised, which is not accounted for here. c Assuming 100% recycle rate and one battery every 10,000 km for electric bikes and one battery every 3 years or 250,000 kilometers for buses, one battery every 3 years or 75,000 km for car, one battery every 3 years or 18,000 km for motorcycle (Wang, Huo et al. 2006). d Sullivan et al. 1998-LCA of Generic US Car (cautiously compare due to different methodology). e Only use phase emission rate, no production processes included. Note: some fields are Unknown (Unkn) because data are not available for the emission of these pollutants from production processes and/or power plant emissions.

Local Impacts Electric two-wheelers emit little local air pollution, to the extent that power plants are sufficiently removed from the urban area. Any shift from electric two-wheelers to alternative motorized modes will certainly result in increased local pollution in the urban area. However,

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

74

Christopher Cherry

most power plant emissions do enter urban areas, increasing the effects of electric twowheeler use. During the use phase electric two-wheelers emit 10–20 times more SO2 from power plants than buses emit from combustion of diesel fuel, on a per-passenger-kilometer basis. The public health effect of this increase is reduced (by a factor of 3 in Kunming) by the fact that power plant emissions are somewhat remote (Zhou, Levy et al. 2006). Nonetheless, public health impacts as a result of increased SO2 will likely increase. Buses emit significantly more NOX and PM than electric two-wheelers and the public health impacts of increases in these pollutants could negate any SO2 savings as a result of a shift from electric two-wheeler to bus. Electric two-wheelers emit far fewer air emissions than cars so even a small displacement of car trips results in large reductions of local air pollution.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Non-Local Impacts Unlike traditional motorized modes, most of the life cycle environmental impacts of electric two-wheelers are incurred during the production processes. Most electric twowheelers and their components are manufactured in eastern provinces and municipalities, primarily Zhejiang and Jiangsu Provinces, and Shanghai and Tianjin Municipalities and thus, most environmental externalities of the production processes are borne by populations in those provinces. Increases in electric two-wheeler use in a non-industrial city impose more environmental externalities on the populations of the industrial provinces than they impose on populations living in cities where electric two-wheelers are used. Lead (Pb) is perhaps the greatest negative environmental externality related to electric two-wheeler use. Electric two-wheelers use, and thus emit, several orders of magnitude more lead than competing modes (on a per passenger kilometer basis). However, most of the negative health effects related to lead exposure are from mining, smelting, recycling and battery manufacturing processes, impacts largely outside of the control of electric twowheeler manufacturers. High recycling rates are important to reduce solid waste but recycling will not solve the lead pollution problem from production processes or the informal recycling sector. A positive non-local impact is electric two-wheelers’ influence on national energy demand. As China industrializes, energy supplies and security are essential to robust and sustainable economic growth. Every year, the transportation sector consumes a higher portion of China’s energy demand, mostly because of rises in motorized transportation modes. Aside from bicycles, electric two-wheelers have the lowest life cycle energy impacts of any mode, consuming very little energy per kilometer. Reducing energy consumption in the transportation sector can benefit all Chinese residents by reallocating energy to the nontransportation sectors.

Global Impacts Comparable to the energy impacts, CO2 emission rates of electric two-wheelers are lower or comparable to any other motorized mode, primarily because of the energy efficiency of battery electric vehicle systems. As the effects of climate change become more tangible, the option of using renewable or low carbon based energy sources to power the transportation

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Clarifying the Debate on Electric Two-Wheelers in China

75

sector is very appealing. The international community has a stake in the transportation choices made by China and it is difficult to find a motorized mode that compares to electric two-wheelers in terms of low greenhouse gas emissions.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Policy Response Most of these environmental problems are symptoms of larger problems related to China’s reliance on coal powered electricity, poor production practices of major industries, and China’s rising demand for motorized transportation. There is some potential that electric two-wheeler motor efficiency and battery performance can produce some marginal gains in efficiency within the electric two-wheeler industry, but changes in the electric two-wheeler industry will have little impact on these larger problems. As China industrializes, electricity generation and production processes will become cleaner and, as a result, electric twowheelers will also become cleaner. Very few of the environmental impacts of electric two-wheelers are local in nature and most cities would like to have all of the benefits of electric two-wheelers while bearing few of the environmental impacts. Because of this, it seems most appropriate to enact environmental policy at a national or regional level. Specifically, the level of lead emission is unacceptable and can easily (but expensively) be mitigated. Alternative battery technologies, such as Lithium ion (Li-ion) or Nickel Metal Hydride (NiMH), that are more environmentally benign do exist and are available on more expensive electric two-wheelers, but the Chinese market has not yet adopted these because of the price (Weinert, Burke et al. 2007). Rising lead prices has caused the price gap to narrow, but even at current prices, lithium ion batteries still have about a 60–100 percent lifecycle price premium (Cherry, Weinert et al. 2008). There are a couple of ways to induce a shift to more environmentally friendly battery technologies. One way is to simply mandate a ban on lead acid batteries used in electric twowheelers. All electric two-wheeler manufacturers are regulated by standards bureaus and this could be another requirement. Judging by current lack of compliance to existing rules, this type of mandate might be difficult to enforce. Another way to induce this technology shift would be to exact a tax on lead acid batteries in the electric two-wheeler industry. This tax would act as an incentive to pull manufactures toward cleaner technologies. This tax could be high enough to promote a transition such that a portion of lead acid batteries would shift to Li-ion or NiMH so that net lead discharge compared to the dominant alternative is zero. This would be a “lead neutral” policy. High end SSEB’s could have advanced batteries, while low-end BSEB’s could maintain lead batteries. A “zero lead” policy would tax lead acid batteries so they are competitive in price with the cleaner alternative, NiMH or Li-ion. This would hopefully induce a 100 percent shift to NiMH or Li-ion, having the same effect as a ban. A tax would generate public revenue that could be invested in environmental mitigation, providing further benefits. Of course, as more electric two-wheelers switch to alternative batteries, tax revenue will decline. Any forced or tax induced shift to alternative battery technologies would raise the purchase price of an electric two-wheeler by 20–25 percent and increase the operating cost by 80–100 percent, borne by the users. This cost increase would presumably result in a marginal shift away from electric two-wheelers. None-the-less, electric two-wheelers would remain one of the cheapest forms of transportation, with user operating costs around 0.12 RMB/km.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

76

Christopher Cherry

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

SAFETY Safety is one of the biggest reservations policy makers have with electric two-wheelers. Unfortunately, electric two-wheeler’s are vulnerable road users, like bicyclists or pedestrians and when a crash occurs (particularly with an automobile), it often results in fatality or serious injury of the vulnerable road user, regardless of fault. There are two main interactions that occur that could reduce the safety of the transportation system. Electric two-wheelers operate in a shared bicycle lane, so the two modes are constantly interacting, causing potential conflict. As electric two-wheeler speed and mass increase, the differential between each mode widens causing safety challenges. On the contrary, under the current system, interactions with automobiles occur primarily at intersections. As electric two-wheelers increase in speed and mass, the differential between cars and two-wheelers narrows, which is especially important in mixed flow conditions. These two scenarios have caused a dichotomy of opinion on the development of larger, motorcycle type electric two-wheelers. Some propose that electric two-wheelers should be classified as bicycles and maintain their smaller size and speed, while others propose that they should grow to larger and faster vehicles to mix better with cars (Ni 2008). While it is perceived that electric two-wheelers are unsafe road users, since they are relatively low speed and low mass vehicles, they could be safer but more vulnerable than opponents suggest. There have been clear directives meant to improve the safety of electric two-wheelers in bicycle lanes. Some of these directives have been successful, while others have failed. Because of these failures, most cities have an abundance of electric two-wheelers on the road that do not meet safety requirements, particularly speed and weight limitations. Public officials should identify the effect of electric two-wheeler speed on safety of the transportation system and develop enforceable standards that limit speed to safe levels such that conflict between electric two-wheeler users and others in bicycle lanes is minimized. Additionally, design and operation strategies should be developed to segregate slower electric two-wheelers from car traffic. Some industry members are fighting the development of such a standard, stating that it would limit innovation and potential technology evolution to heavier “light electric vehicles” such as electric mini-cars. If the industry is allowed to move in this direction, there must be clear re-classification of electric two-wheelers so that heavier electric motorcycles require registration, driver training and licensing, and helmet use. Policy makers must seriously consider whether a faster, heavier electric motorcycle belongs in the car lane, bicycle lane, or on the roadway at all. One of the fundamental challenges associated with safety analysis is data availability and consistency. In general, detailed analysis is not possible without access to data collected by local Public Security Bureaus. Many recent studies have focused on self reported data or aggregate data compiled in statistical yearbooks (Weinert, Ma et al. 2007; Lin, He et al. 2008; Ni 2008). These data have value, but can be misleading. It is difficult to disentangle the effects of different fatality rates and attribute vulnerable road users with all of the “fault” in a crash because the casualties are borne by those vulnerable populations, regardless of fault. Moreover, it is difficult to identify the counterfactual if electric two-wheelers were not available without information on driver demographics and other data. To start, one must

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Clarifying the Debate on Electric Two-Wheelers in China

77

identify the alternative modes electric two-wheeler users shifted to another mode, then identify the safety impacts of that shift.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

MOBILITY AND ACCESSIBILITY Electric two-wheelers provide the greatest benefits in terms of the accessibility and mobility advantages they provide to users. Bicycles have good mid-range mobility benefits and very good environmental benefits, but to the extent that social forces are precluding bicycle use, then electric two-wheelers are a good alternative. Electric two-wheelers have the greatest accessibility and mobility advantage over buses. Because of the large shift away from buses in Kunming and Shanghai, an average electric two-wheeler user saves well over 100 hours of travel time per year if all of the same trips are made on alternative modes. This is consistent with survey results indicating that less than one percent of respondents stating they would not take a trip if electric two-wheelers where not available (Cherry and Cervero 2007). Using average wages, the value of this travel time savings exceeds $100 per year. As an alternative to identifying travel time savings, one can analyze the increase in access to destinations. For example, accessibility analysis in Kunming showed that within 20 minutes of travel time, electric two-wheelers proved over three times more job access than buses (Figure 5). Electric two-wheeler users overwhelmingly choose electric two-wheelers because of their mobility improvements. Travel time was a significant predictor of mode choice in all of the mode choice models and users stated that the primary reason for choosing an electric twowheeler was because it is fast (Cherry and Cervero 2007). The only mode that has the same mobility and accessibility advantages are cars, although cars have other features that make them more attractive, such as comfort and status. However, cars are unable to avoid congestion in the same ways that electric two-wheelers can. One way to induce a positive shift away from electric two-wheelers to buses (without heavy regulation that would also induce a negative shift toward cars) is to improve public transportation services in the city. Electric two-wheelers can compliment public transit use by increasing accessibility to regional rapid transit systems. Improved quality and security of bike parking facilities can encourage this behavior. Increasing speeds of buses to compete with electric two-wheelers would be a difficult task. Because of lost time accessing bus stops and relatively short trip lengths, the operating speeds of buses must be very fast to match the door to door travel time of electric two-wheelers. Consider taking a 5 km trip on a bus or an electric two-wheeler. An electric two-wheeler can make the trip in 21 minutes. To provide comparable door-to-door service, a bus would have to operate at 25 km/hr including stops, which is difficult to accomplish in an urban area, even with exclusive right-of-way (Chang 2005; Hook 2005). Because of congested conditions several sources report average operating speeds of about 10 km/hr. Rapid bus service, with exclusive lanes and/or priority, have been reported to operate at about 15–16 km/hr, on average (Kunming Urban Traffic Research Institute 2004; Hook 2005; Kunming University of Science and Technology 2005). Figure 6 shows the relative advantage of electric two-wheelers compared to buses and bicycles from a mobility perspective. The distribution of trip lengths from the previously discussed surveys are shown on the bottom left, with the 50th percentile trip length equal to 3.5 km. In this case,

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

78

Christopher Cherry

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

considering about ten minutes of access, wait and egress time, even a “fast bus” struggles to provide the same mobility as an electric two-wheeler. More dedicated lanes would improve bus performance and make them more competitive with personal modes. Complimenting improved operating speed, improvements in urban form such that origins and destinations are matched along bus corridors would reduce transfers and walk time, which would reduce the inherent disadvantage of public transportation. Ultimately, clean buses that overcome the mobility problem are the best option, but this service is difficult and expensive to provide.

Figure 5. Mode Specific Jobs Access within 20 Minutes of Kunming City Center.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Clarifying the Debate on Electric Two-Wheelers in China

79

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 6. Mobility Advantages of Electric Two-Wheelers Relative to Alternative Modes.

COST EFFECTIVENESS OF TRAVEL Making some assumptions and drawing on economic analysis in the literature, all of these costs and benefits can be monetized to some extent to develop a measure of cost effectiveness of travel by each mode. Matthews and Lave (2000) summarize studies that monetize the cost of social damage caused by different air pollutants. As expected, the variation of remediation costs is high among pollutants and it is difficult to transfer findings from this study to the case of China. In this case, some of the direct and indirect costs are monetized to reflect the cost of traveling 100 kilometers by each mode. Table 2 breaks down these costs by mode and shows that electric two-wheelers are the most cost effective mode of transportation in China, considering a subset of total costs. Most of these costs are based on assumptions, which could add some uncertainty to the values. The user costs are those borne by the user of the system. These primarily include operating and purchase costs of vehicles and fuels to travel 100 kilometers. Also included is travel time, which is represented by lost productivity that is monetized by the average wage of electric two-wheeler users in the two case study cities, Kunming and Shanghai. All monetized energy, material, and labor costs during the production process are assumed to be embedded into the purchase price of the vehicle, while all monetized costs of fuel production are

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

80

Christopher Cherry

assumed to be represented within the price of the fuel. This assumption is slightly flawed because the Central government heavily regulates fuel (gasoline, diesel and electricity) prices. The indirect costs primarily represent those externalities that are not explicitly included in the price of the product and therefore, not paid by the user or government. These include the cost to mitigate environmental impacts as well as the public health cost of traffic safety impacts.

Direct Costs

Table 2. Cost Effectiveness of Travel by Competing Modes in China

Battery Cost Electricity (Use) Gasoline (Use) Fareb

Unit Cost $34/battery $0.08/kWha

Rate of Impact 10,000km/battery 1.3kWh/100km

$0.56/La $0.19-0.38 per Trip

7.9L/100km 4.4km/trip

E-bike $0.34 $0.10

Cost per 100 pax-km Bicycle Bus Car

Purchase Pricec $0.50 $0.25 $0.17 $5.21 d Travel Time $9.70 $12.70 $20.62 $9.30 Safety $500,000 per 0.20 fatalities per $10.00 (car) Statistical Lifee million-v(p)kt Safety (bicycle) $500,000 per 0.013 fatalities per $0.65 Statistical Lifee million-v(p)kt Safety $500,000 per 0.019 fatalities per $0.95 (e-bike) Statistical Lifee million-v(p)kt SO2 $1.24/kgf lifecycle g/100km $0.02 $0.00 $0.00 $0.09 Remediation CO2 $45/tonneg lifecycle g/100km $0.12 $0.02 $0.24 $1.38 Remediation Lead Acid +$69/battery 10,000km/battery $0.69 Battery Alternativeh Partial Cost of 100 kilometers of travel $12.42 $13.62 $27.53 $30.40 a From (Weinert, Ma et al. 2007) and (Metschies 2007). b It is assumed that fare covers all operating costs in China, including fuel, maintenance and labor, but not including vehicle capital costs. c Assumes E-bike, Bicycle, Bus and Car cost $275, $50, $85,000, and $10,000, respectively and have useable lives of 50,000, 20,000, 50,000,000, and 197,000, passenger kilometers respectively (Sullivan, Williams et al. 1998; People's Daily Online 2002; Volvo 2006). d Using speeds and trip lengths derived in Cherry (Cherry 2007). e Midpoint Value of Statistical Life (VOSL) estimates from (Liu 1997; Feng 1999; Brajer and Mead 2003). f Cost effectiveness study of reducing SO2 (Li, Guttikunda et al. 2004). These improvements also have more minor co-benefits of reducing NOX and PM, not accounted for here. g Approximate value of a metric tonne of CO2 from Clean Development Mechanism (CDM) framework. h Cost of remediating all lead pollution from lead acid battery use in electric two-wheelers by shifting to alternative battery technology (NiMH or Li-Ion). Not relevant for cars or buses because of different battery technology requirements. Indirect Costs

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

$4.42 $6.50

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Clarifying the Debate on Electric Two-Wheelers in China

81

Admittedly, this table does not represent all costs, but it does show that, making some contentious assumptions, electric two-wheelers provide the most cost effective mobility of any mode in China, even compared to traditional bicycles and loaded buses. This accounting does not include public subsidy of road infrastructure, the effects of each mode on congestion, nor does in explicitly include the public health impacts of pollution.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

EFFICIENCY OF ROAD INFRASTRUCTURE ALLOCATION Public transit makes the most efficient use of surface transportation infrastructure from a people moving perspective. Like any personal mode of transportation, electric two-wheelers contribute to congestion by interacting with cars and buses, causing and bearing delay, particularly at intersections. They also occupy physical space on the road network, competing for space with bicyclists. To the extent that bicycle traffic has reduced over the past decade, electric two-wheelers are filling unused capacity in the bicycle lane, perhaps improving the people-moving efficiency of the average road link. Buses are the most efficient users of road space but have problems with mobility. Bicycles and electric two-wheelers are less efficient users of road space than buses, but more efficient users of road space than cars. Again, the overall impact on road capacity depends on the distribution of dominant alternative modes. High volumes of two-wheelers in bicycle lanes do not mix well with mixed flow lanes of larger vehicles. There are particular conflicts with right turning cars crossing through the bike lane that reduces right turning capacity, which in turn reduces auto and bus throughput. Geometric changes along with signal phase changes could reduce this conflict considerably and some novel strategies have been implemented in Kunming and Shanghai to address this problem. Developing optimum operational strategies to mitigate the conflict between bicycles and cars at intersections is an exciting area of future research. Electric two-wheelers also require more parking facilities than other efficient modes but they require much lower space for parking than a typical car parking space. Including access ways and the parking space itself, cars generally require about 28 m2/vehicle. Electric twowheelers require about 2 m2/vehicle of parking space. Bicycles require about 0.75 m2/bicycle. Bus users require no parking. This added parking demand could be a significant cost of electric two-wheeler use. However, all modes generally pay for their parking (on the order of 0.5–2 RMB per trip for two-wheelers).

OTHER EXTERNALITIES This chapter attempted to identify the largest costs and benefits and frame them in a way that could be quantified and compared by some metrics. This would hopefully inform policy on whether electric two-wheelers should or should not be banned or otherwise regulated in Chinese cities. It did not consider other externalities, some significant and important for Chinese policy makers. This study also did not include analysis of the sprawl inducing effects of electric twowheeler use. As transportation costs (monetary and time) reduce, people are able to afford and willing to live in communities that do not have sufficient mix of uses and are often not

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

82

Christopher Cherry

organized toward mass transit. To some extent electric two-wheelers are low cost, high mobility modes of personal transportation that could result in a demand for housing that is not oriented around mass transportation infrastructure, making it difficult to efficiently serve Chinese cities with public transit. Electric two-wheelers don’t have huge parking requirements, like cars, so they will not necessarily demand land uses that are separated by parking lots, but electric two-wheelers might exacerbate job and housing mismatch problems that lead to unsustainable suburbanization and sprawl. One other important omission of this study is that it did not take into account the effect of electric two-wheelers in promoting or inducing more long term auto ownership. This study looks at a snapshot in China’s motorization process and does not include the long term effects of building a culture of personal mobility in China. Based on a survey conducted in May 2008 in Kunming, 20 percent of electric two-wheeler owners plan on buying a car within the next year, within the household. It is unclear which motorization direction China will move toward. Electric two-wheelers could be a stepping stone to full automobile ownership, hastening the arrival of cars into more households. Alternatively, electric two-wheelers could fill the personal mobility needs of Chinese citizens and thus slow or reduce the transition to automobiles, particularly with uncertainty in fuel prices. The surveys conducted in Kunming and Shanghai indicate that there is a proportion of electric two-wheeler users that would use a car if electric two-wheelers were not available. A similar survey conducted in Shijiazhuang asked about future vehicle purchase behavior of bicycle and electric two-wheeler users and found that electric two-wheeler users were much more likely to buy a car in the next year than bicyclists (Weinert, Ma et al. 2007), though this could be an income effect. These results indicate that electric two-wheelers might hasten car ownership for some and deter it for others. Electric two-wheelers have much lower costs than personal cars, but provide about the same levels of access and mobility in a city. There are some factors that influence automobile purchase that are difficult to quantify but important to consider, namely status and improved comfort. An important area of future research will be to quantify some of the factors that influence car purchase decisions so that more sustainable modes can improve to provide the comfort, range, and personal mobility needs of Chinese residents. This study does not explicitly calculate net public health benefits or costs, in terms of increased mortality or morbidity. These are difficult metrics to calculate given limited studies in the Chinese context and difficulty identifying mortality rates for various pollutants and processes, particularly production processes and lead pollution. Also, the health benefits of active transport (walking and bicycling) are not included as Chinese residents who shift from bicycle to motorized modes have shown significant weight gain and incidence of obesity (Bell, Ge et al. 2002). Any shift from bus, walk, or bicycle modes to electric two-wheelers will result in less exercise and increased obesity, causing public health effects that are not accounted for in this study.

CONCLUSION This chapter took a critical look at electric two-wheelers, addressing problems that opponents have noted, such as increased pollution and poor safety. While this chapter does not intend to answer all of the questions or promote or discourage electric two-wheelers, the

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Clarifying the Debate on Electric Two-Wheelers in China

83

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

author hopes to frame the questions so that policy makers can identify the important issues and identify metrics to measure in order to answer those questions. Opponents often cite these impacts, but do not estimate the net impact (as a result of mode shift) of regulating electric two-wheeler use in a city. While they cite these problems with electric two-wheelers, they rarely cite the same problems with the automobile, especially in auto-oriented cities like Beijing and Guangzhou, which have attempted full bans of electric two-wheelers in the last two to three years. It is true that the massive amounts of lead pollution from lead acid batteries are a significant problem, but this is not a problem with electric two-wheelers per se, but rather part of a larger problem with the lead production industry. It is also a problem that can be remedied, but will cost electric two-wheeler users or the government a significant sum of money to upgrade to better-performing and more environmentally-benign batteries. The benefits of electric two-wheeler use—primarily through saved travel time, potential reduced roadway fatalities, reduced energy use, and reduced CO2 and NOX emissions, are large enough in magnitude to justify such an investment by users or subsidy by the government to push the evolution to better batteries. As power plant emission rates reduce over time and production processes clean up, electric two-wheelers (and all modes) will become cleaner to produce and operate. Table 3 gives a qualitative direction and magnitude of transportation impacts of electric two-wheelers compared to various competing modes. Each mode is identified with local and non-local impacts (over the lifecycle). Electric two-wheelers perform well against cars in almost all metrics but lead. They perform well on many metrics, but poorly on others compared to buses, and they perform poorly on nearly all metrics compared to bicycles. Overall, considering the alternative mode splits in Kunming and Shanghai, electric two-wheelers provide benefits to the transportation system, primarily because they displace a small percentage of very poorly performing automobiles and provide much more mobility to would-be bus riders Table 3. Direction and Magnitude of Electric Two-Wheeler Advantage or Disadvantage

Energy CO2 NOX SO2 PM Lead (Pb) Safety Mobility Access User Cost

Electric two-wheeler (dis)advantage compared to: Bicycle Bus Car NonNonNonLocal Local Local Local Local Local ~ + ++ ++ + + +++ ++ ++ +++ --++ + + + -------++ ++ +++ ~ ++ +++ ~ -++ +++

Overall Mode Shift Impacts in General (with Kunming and Shanghai types of mode split) + + + ~ -+ ++ ++ ++

- Electric two-wheelers perform poorly on this metric compared to alternative. + Electric two-wheelers perform well on this metric compared to alternative. Multiple + or – indicates stronger advantage or disadvantage.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

84

Christopher Cherry

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Electric two-wheelers erode transit ridership and reduce bicycle use, both modes that outperform electric two-wheelers on many environmental metrics. However, use of electric two-wheelers does not preclude use of buses or even bicycles. Buses are still a valuable form of transportation and used by electric two-wheeler users in foul weather or for longer-distance trips. Electric two-wheelers also provide high levels of mobility in local areas and can serve as very efficient feeder vehicles to a regional rapid transit system. This type of system requires secure parking at stations and perhaps vehicle sharing systems established at the terminals so that a transportation system user has access to vehicles at both ends of the trip. Some subway stations in Shanghai and Beijing have extensive bicycle and electric twowheeler parking facilities near terminal stations, but most stations lack parking infrastructure, limiting the access of those outside of walking distance. Electric two-wheelers have some clear benefits in terms of mobility with low user and public costs; it might be worth subsidizing the transition to alternative batteries. If electric two-wheeler batteries are replaced every two years, the added cost per year to shift to a Li-ion or NiMH battery is on the order of 200–300 RMB/yr ($28–43/yr). The previous sections showed that the value of travel time savings per user exceeds this value by a factor of three. Additionally, if you consider the public subsidy that could be required to meet the added travel demand on public transit, it would be cost effective to simply subsidize the industry to support electric two-wheeler use that, in the absence of lead pollution, would have clear environmental benefits. Since most electric two-wheeler impacts are non-local, national or regional policy must be developed that supports the sustainable development of electric two-wheelers in China. Currently, most regulatory policy is being made on a local level and outright bans are not the appropriate policy approach. Fixing the few problems with electric two-wheelers will make them one of the most sustainable and cost-effective transportation options available to Chinese residents.

ACKNOWLEDGEMENTS Portions of this work were funded by the UC Berkeley Center for Future Urban Transport–A Volvo Center of Excellence, The Clean Air Initiative in Asia, and The National Science Foundation. The author wishes to acknowledge his advisors, Robert Cervero and Adib Kanafani, as well as partners on this endeavor, Yang Xinmiao, Pan Haixiao, Xiong Jian, and Jonathan Weinert.

REFERENCES Bell, A. C., K. Y. Ge, et al. (2002). “The Road To Obesity Or The Path To Prevention: Motorized Transportation And Obesity In China.” Obesity Research 10(4): 277-283. Brajer, V. And R. W. Mead (2003). “Blue Skies In Beijing? Looking At The Olympic Effect.” Journal Of Environment And Development 12(2): 239-263. Chang, J. (2005). BRT Developments In China. Environment 2005 Conference Sustainable Transport And Cities: Improving Transit Systems, Abu Dhabi, UAE, 1/29/05.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Clarifying the Debate on Electric Two-Wheelers in China

85

Cherry, C. (2007). Electric Two-Wheelers In China: Analysis Of Environmental, Safety, And Mobility Impacts. Civil And Environmental Engineering, University Of CaliforniaBerkeley. Phd Dissertation. Cherry, C. And R. Cervero (2007). "Use Characteristics And Mode Choice Behavior Of Electric Bike Users In China." Transport Policy 14(3): 247-257. Cherry, C., J. X. Weinert, et al. (2008). Electric Bikes In the Peoples Republic of ChinaImpact On The Environment And Prospects For Future Growth.Clean Air Initiative-Asia, Asian Development Bank. China Central Government (1999). General Technical Standards Of E-Bike (Gb17761-1999) From National E-Bike Compelling Standards. China Central Government (2004). National Road Transportation Law. Chr Metals (2008). World Lead: Recent Developments In China's E-Bike Market. Fairley, P. (2005). “China's Cyclists Take Charge.” IEEE Spectrum. Feng, T. (1999). Controlling Air Pollution In China: Risk Valuation And The Definition Of Environmental Policy. Edward Elgar. Guangzhou Daily (2006). Guangzhou Bans Electric Bikes. Guangzhou Daily. Guangzhou, China. Hook, W. (2005). “Institutional And Regulatory Options For Bus Rapid Transit In Developing Countries: Lessons From International Experience.” Transportation Research Record: Journal Of The Transportation Research Board 1939: 184-191. Jamerson, F. E. And E. Benjamin (2004). Electric Bikes Worldwide Reports With 2005 Update-10,000,000 Light Electric Vehicles In 2004. Seventh Edition. Jamerson, F. E. And E. Benjamin (2007). Electric Bikes Worldwide Reports -20,000,000 Light Electric Vehicles In 2007. Eighth Edition. Kunming University Of Science And Technology (2005). Kunming City Bus Network Optimization. Kunming Urban Traffic Research Institute (2004). Kunming Brt System Study. G-030907042. Li, J., S. K. Guttikunda, Et Al. (2004). “Quantifying The Human Health Benefits Of Curbing Air Pollution In Shanghai.” Journal Of Environmental Management 70: 49-62. Lin, S., M. He, Et Al. (2008). “Comparison Study On Operating Speeds Of Electric-Bicycle And Bicycle: Experience From Field Investigation In Kunming.” Transportation Research Record, Journal Of The Transportation Research Board 2048: 52-59. Liu, J. T., Hammitt, J.K., Liu, J.L. (1997). “Estimated Hedonic Wage Function And Value Of Life In A Developing Country.” Economic Letters 57: 353-380. Matthews, H. S. And L. B. Lave (2000). “Applications Of Environmental Valuation For Determining Externality Costs.” Environmental Science And Technology 34(8): 13901395. Metschies, G. P. (2007). International Fuel Prices 2007. GTZ 5th Edition. Ni, J. (2008). Electric Two-Wheelers In China: Analysis Of Safety. Luyuan Electric Vehicle Company. http://www.Luyuan.Cn/Showxwdt.Asp?Art_Id=117 People's Daily Online (2002). Shanghai To Purchase 2,010 Volvo Buses. People’s Daily Online. Sullivan, J. L., R. L. Williams, Et Al. (1998). Life Cycle Inventory Of A Generic U.S. Family Sedan Overview Of Results Uscar Amp Project. Society Of Automotive Engineers 982160.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

86

Christopher Cherry

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Volvo (2006). Environment Product Declaration-Volvo 8500 Low-Entry. Wang, M., H. Huo, Et Al. (2006). Projection Of Chinese Motor Vehicle Growth, Oil Demand, And Co2 Emissions Through 2050. Argonne National Lab Anl/Esd/06-6. Weinert, J. (2007). The Rise Of Electric Two-Wheelers In China. Transportation Technology And Policy, University Of California, Davis. Phd Dissertation. Weinert, J. X., A. F. Burke, Et Al. (2007). “Lead-Acid And Lithium-Ion Batteries For The Chinese Electric Bike Market And Implications On Future Technology Advancement.” Journal Of Power Sources 172(2): 938-945. Weinert, J. X., C. T. Ma, Et Al. (2007). “The Transition To Electric Bikes In China: History And Key Reasons For Rapid Growth.” Transportation 34(3): 301-318. Weinert, J. X., C. T. Ma, Et Al. (2007). “The Transition To Electric Bikes In China: Effect On Travel Behavior, Mode Shift, And User Safety Perceptions In A Medium-Sized City.” Transportation Research Record: Journal Of The Transportation Research Board 1938: 62-68. Zhou, Y., J. I. Levy, Et Al. (2006). “The Influence Of Geographic Location On Population Exposure To Emissions From Power Plants Throughout China.” Environment International 32(3): 365-373.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Electric Vehicles: Technology, Research and Development ISBN 978-1-60741-142-0 Editor: Gerald B. Raines © 2009 Nova Science Publishers, Inc.

Chapter 4

EXAMINATION OF STATE ESTIMATORS FOR ELECTROCHEMICAL ENERGY STORAGE DEVICES BY MEANS OF A HARDWARE-IN-THE-LOOP SYSTEM Mark Verbrugge, Damon Frisch, Trudy Weber, General Motors Corporation

Arthur Bekaryan and Ping Liu HRL Laboratories

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT We have designed and implemented a hardware-in-the-loop system for the development, verification, and validation of algorithms used to construct state estimators for batteries and supercapacitors. The intent of the work is to allow algorithm developers to test their algorithms rapidly and in the context of the actual application. Promising results are shown for a carbon based electric double layer capacitor (a supercapacitor); the modeled vehicle configuration corresponds to the General Motors Saturn Vue Green Line 36V HEV, with the NiMH battery replaced by a supercapacitor pack. While the work is pragmatic in terms of addressing vehicle applications, the theoretical underpinnings and mathematical methods are relatively new, evolving, and comprise substantial complexity.

INTRODUCTION State estimators [1-9] that can adaptively characterize the performance of power subsystems provide a rational means to effectively control systems employing more than one power subsystem, as is the case for hybrid electric vehicles wherein a battery or supercapacitor is combined with an internal combustion engine to provide propulsion power. For the energy storage device, we seek the adaptive identification of the energy content (state of charge, or SOC), predicted power capability (state of power, or SOP), and performance

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

88

Mark Verbrugge, Damon Frisch, Trudy Weber et al.

relative to the new and end-of-life condition (state of health, or SOH) [10-20]. A representative HEV propulsion system is depicted in Figure 1, and the analysis we shall provide corresponds to this schematic. The SOC and SOP are used by the propulsion system controller to determine if and when the engine should be turned off, the battery should be charged or discharged, or other actions should be taken, particularly with respect to accessory-power loads. Previous analyses have shown promise for the use of weighted recursive least squares (WRLS) with exponential forgetting [11,14,16,18,20] for adaptive parameter regression when model reference adaptive systems are employed. For this approach, a model of the plant (e.g., the battery or supercapacitor) is constructed, and the parameters appearing in the model are regressed adaptively from the available measurements. The time weighting of data is damped exponentially; hence, new data has a preferential impact in determining the value of regressed parameters and thus the state of the system. In order to verify and validate the algorithms upon which the state estimators are based, it is helpful to place the energy storage device in an experimental set up that closely mimics the intended application. In Reference [21], a so-called hardware-in-the-loop (HWIL) system is described wherein an electrochemical energy storage device is placed within the set up, and the state estimator can be analyzed. It is the purpose of this work to implement the described set up on an electric double layer capacitor cell; a schematic of the set up is provided in Figure 2, which is described in more detail in the next section. Clutch Heat engine Motor/generator

Transmission

Differential

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

v(t) Supercapacitor

+

+

AC DC

Power inverter

Figure 1. Block diagram for a hybrid electric vehicle (HEV) employing a 36 V pack comprising 20 double layer capacitors of the type characterized in this work. The schematic and the vehicle model used in this work reflect the (front wheel drive) propulsion architecture of the 2006 GM Saturn Vue HEV (which employs a 36 V NiMH battery). We chose a supercapacitor cell as it is the most straightforward electrochemical energy storage device to analyze with a state estimator, and thus we can better clarify the character and utility of the HWIL set up. A plot of the open-circuit potentials for electrochemical cells of interest for HEV applications is provided in Figure 3. Because the voltage part of state estimator algorithms employs the extracted open-circuit potential to deduce the SOC, those energy storage devices that exhibit substantial and linear voltage change with SOC exhibit greater signal to noise in the determination of the SOC. Second, the actual measured voltage of the capacitor is much closer to that of the open-circuit value in the case of supercapacitors because of their low impedance (i.e., low sheet resistance, ohm-cm2). Last, supercapacitors

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

89

Examination of State Estimators for Electrochemical Energy Storage Devices …

can be modeled, for the purpose of state estimators, by a very simple RC equivalent circuit (Figure 4), which simplifies the equation system for the state estimator. Electrochemical Energy Storage Device Electrochemical Instrumentation

Post Test Data Analysis, Housekeeping

EC Cell Interface

Power Demands HWIL Controller Vehicle Model

Vehicle Configuration

Available Power

Voltage, Current, Temperature

Subsystem Interface

SOC/SOH/SOP Algorithm Electrochemical Model

Vehicle Analysis

Figure 2. Hardware-in-the-Loop system (HWIL). 4.5

Conventional lithium ion (Hard carbon anode)

4

Open-circuit voltage, V

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.5

Lithium ion (Iron phosphate cathode)

3

Lithium ion (Titanate anode)

2.5

2

Lead acid (Valve Regulated) 1.5

NiMH 1

Supercapacitor C/nonaqueous electrolyte/C

0.5

0 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Fractional state of charge Figure 3. Representative equilibrium potentials for various electrochemical energy storage devices. Hysteresis is displayed for the nickel metal hydride potential behavior, as the charge potential is greater than that of discharge (indicated by the arrows).

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

90

Mark Verbrugge, Damon Frisch, Trudy Weber et al.

C

R

+

-

+

(-

I

+

V

-

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 4. Equivalent circuit representation for the supercapacitor cell. In addition to these motivations, supercapacitors are of immediate interest for traction applications for mild hybrid architectures, and much has been written on the behavior of these devices [23-47]. If it is desired to have the engine (or fuel cell) off upon deceleration or when the vehicle is stopped, and one is willing to forego substantial downsizing of the primary power plant, then capacitors based on high-surface-area carbon electrodes and acetonitrile solvents represent a promising low-cost, high-performance option over a broad range of temperatures [43]. There is concern, however, over the abuse tolerance of acetonitrile-based systems, particularly in the context of cell overcharge, and this provides motivation for clarifying the accuracy of control algorithms at the cell level versus the pack level, as much less energy is associated with a single cell than a pack. It is also worth noting that for plug-in HEVs designed to provide zero-emission-vehicle (ZEV) range, the use of a high-energy battery coupled to a high-power capacitor is a promising architecture, as the power to energy ratio of the battery is reduced, and performance and cycle life for the tandem system can be improved. To summarize automotive interests in supercapacitors of the type investigated here, there is strong interest in these devices for mild HEVs and plug-in HEVs, but less interest in the devices for a conventional charge-sustaining strong HEV, as a single energy storage device (e.g., a lithium ion battery) would likely provide the most effective energy storage configuration. Asymmetric capacitors that utilize a robust battery electrode along with an activated carbon (capacitor) anode [49-50] may provide a means to deliver high power density and sufficient energy density for charge-sustaining strong HEVs. This paper is organized as follows. The next section overviews the HWIL system, after which a description of the state estimator is provided. The test protocol, including the driving scheduling and its implementation in the context of the HWIL system, are then presented, followed by a discussion of results, with specific emphasis on (1) the importance of random test selection for verification and validation and (2) an error analysis of the SOC and power prediction capability of the state estimator.

SYSTEM OVERVIEW The HWIL system comprises three modules (cf. Figure 2): the Electrochemical Cell Interface (ECI), the Subsystem Interface, and the Vehicle Model. The modules are linked together via the HWIL controller, which passes data and commands between the modules. The ECI module provides electrical and thermal environment for the electrochemical cell.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Examination of State Estimators for Electrochemical Energy Storage Devices …

91

Based on the measured response of the cell through this interface, the HWIL controller simulates the vehicle energy storage system. A single-channel testing system and microcontroller (Arbin Instruments) is employed. The tester (model BT-2000) is configured for large electrochemical energy storage devices and provides up to kW at potentials between 0.6 and 5 V. For increased accuracy, the linear bipolar channel is configured with two current ranges, 1000 amps and 20 amps. Rise times range from 0.5 to 1 millisecond. The tester is controlled by a microcontroller with the Motorola Coldfire processor. It includes a generic high-level command set based on the TCP/IP communications protocol, which allows for a direct hardware interface from the HWIL controller, minimizing the system response time. The electrochemical cell is housed within a thermal chamber (Model 105A Test Equity, Thousand Oaks, Ca.). This chamber has a programmable range from -40 to 130°C, controllable within °C. For all of the data reported herein, the electrochemical cell is an electric double layer capacitor acquired from Saft America [14]. The nominal capacitance is 3500 F, and the maximum voltage is 2.8 V. The Subsystem Interface module includes a State Estimator algorithm rendered in the Matlab/Simulink tool. The Subsystem Interface module supplies the SOC and available power (SOP) estimates for the capacitor deduced from the history of voltage and current measurements The vehicle model is based on the Hybrid Powertrain Simulation Program (HPSP), a modeling tool developed by GM and based on the backward-driven simulation approach [51,52]. The analysis is initiated by an instantaneous road load requirement specified by a driving cycle (vehicle speed as a function of time). Empirical quasi-steady state models using efficiency maps and system specific input parameters such as inertias and gear ratios represent the components of the powertrain. The speed and torque requirements of each component are tracked backwards from the road load through the driveline components eventually determining the engine speed and torque operating points. The current and voltage requirements for the electric drive are determined from the torque, speed, and acceleration requirements of the electrical components, which include their electrical and mechanical losses. An electric accessory load is also accounted for in the model. This approach is ideal for following a driving schedule to determine the engine operating regions under optimum controls of the powertrain or based on specialized control and energy management strategies. Within the HWIL system, HPSP performs its calculations and generates system requests at a regular time intervals (typically 0.1 to 0.25 seconds). These calculations are used to continuously update the vehicle performance parameters (fuel consumption, available supercapacitor power, etc.). HPSP subsequently records all powertrain related activity over the test profile and formats them for post processing. The HWIL controller contains software written in C++ that resides on a dedicated HWIL computer along with the HPSP vehicle model and elements of the Subsystem Interface module. The primary function of the controller is to interface the multiple application environments associated with the modules. Because these modules are intended to be flexible, the responsibility for overall abuse tolerance, robustness and performance is handled by the HWIL controller. It is also the control application for the cell tester, relaying power commands from the vehicle model, recording cell current, voltage and temperature and transferring data to the Subsystem Interface module. The HWIL controller oversees the test

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

92

Mark Verbrugge, Damon Frisch, Trudy Weber et al.

cell performance parameters, providing them to the vehicle model, and stopping or limiting a test if these parameters are exceeded.

STATE ESTIMATOR ALGORITHM The adaptive state estimator comprises voltage-based and coulomb-based models. With the exception of the power-capability expressions to be described in the next section, the algorithm is identical to that described in Reference [[14],[15]] and our description is terse albeit mathematically complete where there is overlap with the previous publication. An electriccircuit model is employed to describe the relationship between the currents and voltages observed at the terminals of the capacitor (giving rise to the voltage based SOCV) and a coulomb-accumulation model that describes the open-circuit voltage based on the history of currents seen by the capacitor (giving rise to the coulomb-counting based SOCI), which can include self discharge and current inefficiency on charge. The electrical circuit model is illustrated This equivalent circuit is used to extract a current-voltage expression, which is then employed to characterize the capacitor. In addition, the current-voltage expression forms the basis for the voltage based SOCV. While quite rudimentary, it has been shown that for the class of electric double layer capacitors we investigate here (activated carbon electrodes with an acetonitrile solvent), the equivalent circuit can be used to describe experimental data over a broad range of currents, potential, and temperatures [[14],[46]].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Voltage-based Model By summing the voltage around the circuit shown in Figure 4, we can express the capacitor voltage as ppp as ζ =t

V=

1 Q ( 0) + IR + ∫ I ζ dζ , C C ζ =0

(1)

which can be recast as the following fully recursive relation for time step N:

V

N

= (t N − t N −1 )

I | N + I | N −1 1 + I | N R + (V 2 C

N −1

− I | N −1 R ) .

(2)

Thus to determine the voltage at time N, V|N, one only needs the present value of the current and the previous time-step values for the current and voltage. The next step is to formulate an adaptive procedure for the estimation of R and C from a history of currents and voltages, which provide the basis for state estimation of the capacitor. The following definitions streamline notation:

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Examination of State Estimators for Electrochemical Energy Storage Devices …

x1 = I t − I t −Δt

x2 = ( I t + I t − Δt )

m1 = R

m2 =

93

Δt 2

1 C

y = Vmeas ,t − Vmeas ,t − Δt Equation 2 can now be recast as

y = m1 x1 + m2 x 2 , and the error ε that we seek to minimize corresponds to

ε = ∑ w j [y j − (m1 x1, j + m2 x2, j )]2 . N

j =1

Time t corresponds to index j=N in the above. Two equations for the two parameters to be extracted can be generated by setting the partials

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

m1 =

∂ε ∂ε and to zero. This leads to ∂m1 ∂m2

1 1 (V1, yV2, 2 − V2, yV1, 2 ) and m2 = (V2, yV1,1 − V1, yV2,1 ) , Det Det

(3)

where the matrix determinant Det is given by

Det = V1,1V2, 2 − V12, 2 . The variances correspond to

Vuv

N

⎛ = ⎜ s u ,v ⎜ ⎝

N

−

su

N

sv

sw

N

N

⎞ 1 ⎟ , ⎟ sw ⎠ N

(4)

in which the sums refer to N

N

N

N

j =1

j =1

j =1

j =1

s w = ∑ w j , s u = ∑ w j u j , s v = ∑ w j v j , s u ,v = ∑ w j u j v j , and u and v refer to x1, x2, or y. It should be noted that the matrix system is symmetric, V1,2=V2,1. Variances have been used in these expressions, along with the normalization

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

94

Mark Verbrugge, Damon Frisch, Trudy Weber et al.

associated with the division by sw in order to keep the resulting matrix elements nearer to unity (i.e., well scaled). Before making the summations recursive, we address the weight factor wj. There are two reasons why one would want to weight data sets differently. First, some observations may be subject to greater disturbance; here, a disturbance refers to a phenomenon not accounted for in the system model. Second, newer observations are generally more important than older observations in determining the state of the system and therefore should usually be given a larger weight factor relative to older observations. For these two reasons, we decompose the weight factor into a time-weighting factor λ and a general weight factor γ; the latter of which can be used to associate a specified weighting of selected events (e.g., discharge events over charge events). Hence,

wj = γ jλ

N− j

.

(5)

yields an exponential decay in the influence of past It can be shown that the use of λ data points on the determination of the current value of m1 and m2: N− j

λN − j = e ln λ

N− j

= e ( N − j ) ln λ ≈ e − ( N − j )(1−λ ) for λ → 1 .

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

More complete discussions on exponential forgetting can be found in section 5.6.1 of Ljnug and Söderström [[6]], section 6.2 of Anderson and Moore [[2]], and sections 5.3 and 5.4 of Kulhavý [[9]]. The summations are now made recursive with the following definitions:

sw

N

= γ N + λ (sw

su

N

= γ N u N + λ ( su

s u ,v

N

N −1

),

N −1

), sv

= γ N u N v N + λ ( s u ,v

N −1

N

= γ N v N + λ ( sv

N −1

), and

(6)

)

with su ,v = s v ,u . Initially,

sw 1 = γ 1 , su 1 = γ 1u1 , sv 1 = γ 1v1 , and

(7)

su ,v 1 = γ 1u1v1 . At this point the equations needed to regress the parameters m1 and m2 have been fully stated. At the beginning of each time step, the recursive sums 6 are calculated (Eqs. 7 for the first time step), which are then substituted in the variances 4; upon determining all variances, the parameters m1 and m2 are deduced by means of Eqs. 3, thereby providing C and R. We now address the voltage-based SOC calculation. The charge on the capacitor depicted in Figure 4 can be expressed as

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Examination of State Estimators for Electrochemical Energy Storage Devices …

95

ζ =t

Q = Q ( 0) +

∫ I ζ dζ

= C (V − IR) .

ζ =0

Q may be viewed as the magnitude of the charge on the (symmetric) capacitor electrodes. We define the voltage-based SOC, SOCV, in terms of the minimum charge on the capacitor Qmin relative to its maximum value Qmax allowed under normal operating conditions,

SOCV = = =

Q − Qmin Qmax − Qmin C (V − IR) − Qmin , CVmax I =0 − Qmin V − IR − Vmin Vmax

− Vmin I =0

(8)

I =0 I =0

where Qmin and Qmax correspond to the minimum and maximum voltages (Vmin and Vmax, respectively) under zero current conditions. In vehicle applications, the power electronics (i.e., the power inverter) used to convert the dc current associated with the energy storage device (e.g., a battery, fuel cell, or capacitor) to ac current for the electric machines require that Vmin ≥ 0.5Vmax in order that the power inverter maintains an acceptable efficiency. Thus

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

for the figures presented herein, Vmin = 0 in formulating the SOC, and we investigate SOC’s ranging from 50 to 100 percent in the power capability analyses (1.4 ≤ V0 ≤ 2.8 V); 100 percent SOC and Vmax are taken to correspond to 2.8 V, which implies that V0 = 1.4 V refers to 50% SOC.

Coulomb-based Model and Composite SOC Expression The previous section provided for the extraction of C and R from a data stream provided the determinant reflects a well-posed equation system. We now use this information along with coulomb counting to construct the SOC algorithm. We can also calculate an SOC based on coulomb counting so as to construct a currentbased SOC, SOCI,

⎛ I + I t − Δt ⎞ Δt ⎟⎟ SOC I (t ) = SOC (t − Δt ) + ⎜⎜ t − Q Q min ⎠ 2 ⎝ max I t + I t − Δt 1⎛ = SOC (t − Δt ) + ⎜ ⎜ C ⎝ Vmax I =0 − Vmin

⎞ Δt ⎟ ⎟ 2 I =0 ⎠

.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

(9)

96

Mark Verbrugge, Damon Frisch, Trudy Weber et al.

Both the voltage and current based SOC’s contain useful information, and a weighted average is thus rendered as the final composite SOC,

SOC = wSOC I + (1 − w) SOCV ,

(10)

with the weight factor w chosen to be closer to 1 for enhanced stability and closer to zero for increased responsiveness. For this work, w=0.99 was used. Note that in the formulation of Eq. 9 for SOCI, we increment from the previous value of SOC (not SOCI), thereby linking explicitly the current-based SOC to the voltage based SOC.

Power Capability The maximum discharge power can be expressed as:

Pmax, discharge = IV = IVmin . That is, when the capacitor voltage obtains its lowest acceptable value, the max discharge power results. Hence this section of the work is focused on the application of a constant voltage so as to obtain either the maximum discharge (V = Vmin) or charge (V = Vmax) power of the electric double layer capacitor. First, we consider the available instantaneous power; i.e., the power available before the charge on the capacitor is depleted significantly by the discharge event. In this case, V = V (0) + IR , where V(0) is the system voltage at zero current immediately prior to the Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

discharge [V(0)=Q(0)/C, cf. Eq.1], and

Pmax, discharge = IVmin =

[Vmin − V (0)] Vmin . R

(11)

Similarly, the instantaneous charge power for the ohmic battery is given by

Pmax, charge = IVmax =

[Vmax − V (0)] Vmax . R

(12)

The ohmic battery does not address transient effects, which are important for times that are greater than ~0.01RC. One can differentiate Eq. 1, set dV/dt = 0, and obtain the following expressions for the P(t):

[Vmin − V (0)] Vmin e −t /( RC ) R . [Vmax − V (0)] −t /( RC ) = Vmax e R

Pmax, discharge = IVmin = Pmax, charge = IVmax

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

(13)

Examination of State Estimators for Electrochemical Energy Storage Devices …

97

Thus as t → 0 , the instantaneous power expressions are obtained.

TEST PROTOCOL

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

For all of the tests reported herein, the capacitor was initially set to 2.3V (i.e., time t = 0 in Figure 5. The state estimator was provided default values for the resistance R and capacitance C; the state estimator converged to the substantially constant values for R and C after about 500 s, as noted in Figiure 6. The first 600 seconds of the test correspond to 10 minutes of the US EPA Urban Cycle (defined by the velocity versus time relationship). While this forcing of the state estimator to converge at the start of a drive event is a severe test relative to initialization methods employed in actual vehicle applications, it does allow one to assess the convergence properties of the state estimator. During the first 600 s, the HWIL systems randomly selects which test will be run once 600 s is reached and when during the next 770 s the test will be conducted. Three different tests can be selected: (1) SOC test, (2) discharge power test, or (3) charge power test. For the results presented in this work, the number of SOC, discharge power, and charge power tests correspond to 67, 70, and 63. For verification and validation of an algorithm, it is important to not introduce systematic aberrations into the analysis, thus explaining the random selection of both the test time and type. In addition, while not further addressed in this work, a myriad of different drive schedules and events must also be investigated. (The HWIL system is fully automated, allowing an extensive collection of data for analysis.) For the SOC tests, at predetermined test time, the projected SOC rendered by the state estimator is recorded, the current is set to zero for 180 seconds, after which the cell potential is recorded in order to acquire a voltage-based SOC,

SOCV =

V − Vmin Vmax

I =0

I =0

− Vmin

I =0

(14)

per Eq. 8 under equilibrium conditions. The cell is then discharged at low current (6 A) to measure the coulombic capacity and ascertain a second measure of the cell SOC per Eq. 9. For the power tests, at predetermined test time, the projected power 0.5 and 1 second into the future is rendered by the state estimator and recorded (cf. Eq. 13), the voltage is set to its minimum or maximum value for the discharge or charge power test, respectively, and the measured power after 0.5 and 1 seconds is recorded.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

98

Mark Verbrugge, Damon Frisch, Trudy Weber et al. EPA Urban Cycle 100

SOC Test

Charge Test

Discharge Test

Min.Test Start Time

Power, W

1500 1300

60

1100

40

900

20

700

0

500

-20

300

-40

100

-60

-100

-80

-300

-100 0

200

400

600

800

1000

1200

Cell power, W

Vehicle Velocity, km/hr

80

Vehicle Cycle

-500 1400

Time, s

Figure 5. Vehicle velocity, power profile of the electric double layer capacitor, and (randomly selected) times power and SOC tests.

Algorithm “converges” after ~500 s

C

R

-

+

+

(-

4000

I

+

V

-

1.4

3500

1.3

3000

1.2

2500

1.1

2000

1 0

200

400

600

800

1000

1200

Regressed capacitance, F

Regressed resistance, mohm

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1.5

1500 1400

Time, s

Figure 6. Regressed capacitance and resistance for the equivalent circuit representation over the drive cycle depicted in Figure 5. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Examination of State Estimators for Electrochemical Energy Storage Devices …

99

DISCUSSION As noted previously, the state estimator has effectively converged after 600 s (Figure 6), after which the state of the system is clarified. The results for the SOC are presented in Figure 7. We find that the voltage-based projection (Eq. 14) is in good agreement with the measured SOC and that the coulombic-based projection over-estimates the SOC. The error in the coulombic-based projection is not large, and the cause for the systematic over-prediction has not been resolved. A plot of the projected maximum (discharge and charge) power capability versus the measure maximum power is provided in Figure 8. For times less than 0.5 seconds into the future, the state estimator provides a very accurate estimate of the power capability. For longer times, the state estimator under-predicts the maximum power capability for both charge and discharge. One second is a long time relative to serial data transfer on a vehicle communications system, and the erosion of the accuracy for times greater than 0.5 seconds is not likely to cause concern in future vehicle applications, as the power capability will be updated with frequency greater than 0.5 Hz. A more complete statistical analysis of the regressed parameters R and C is provided in Figure 9. For both R and C, the standard deviation in both parameters relative to all of the tests conducted is near 10% of the average value of each parameter. Tests similar to those depicted in Figure 9 indicate that no systematic trend in error is observed in the regressed values of R and C, which bodes well for the robustness of the employed approach.

90

Projected SOC

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

95

85

Open-circuit voltage for measured SOC Coulomb titration for measured SOC

80 80

85

90

95

Measured SOC

Figure 7. Projected and measured state of charge. The (randomly selected) times at which the SOC was measured are indicated by the circular symbols in Figure 5.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

100

Mark Verbrugge, Damon Frisch, Trudy Weber et al. 800 600 400

Discharge, 0.5 s

Discharge, 1 s

Charge, 0.5 s

Charge, 1 s

Projected power, W

200

Charge

0 -200 -400 -600

Discharge

-800 -1000 -1000

-800

-600

-400

-200

0

200

400

600

800

Measured power, W

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 8. Projected and measured maximum charge and discharge power. The (randomly selected) times at which the power tests were conducted are indicated by the diamond (discharge) and triangular (charge) symbols in Figure 5.

SUMMARY AND CONCLUSION A hardware-in-the-loop (HWIL) system has been designed and implemented such that vehicles appropriately modeled with a vehicle simulation program can be examined in terms of the energy management system at the cell level. Because supercapacitors are the most straightforward electrochemical energy storage devices to quantify in terms of a state estimator, and because they are of practical interest for HEVs, we have exercised the HWIL system using a supercapacitor cell integrated into the 36V Saturn Vue Greenline Hybrid; i.e., the supercapacitor cell is part of a capacitor pack that replaces the NiMH battery in the vehicle for this exercise. The state estimator results from the combination of (1) an electrochemical model based on the underlying physical chemistry of the supercapacitor and (2) an optimal parameterestimation scheme (weighted recursive least squares). Outputs from the state estimator include the projected power capability for both charge and discharge and the state of charge. The ability of the state estimator to approximate these quantities is examined quantitatively during synthetic drive cycles. The closed loop system is fully automated, enabling one to ensure randomness in the selection of both the type and time of various tests so as to avoid

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Examination of State Estimators for Electrochemical Energy Storage Devices …

101

any systematic aberrations in the algorithm verification and validation process. The state estimator also renders the regressed electrochemical parameters used in the model upon which the algorithm is formulated. These values are useful in constructing state of health indicators. 3100

Standard deviations R : 0.012 mohm C : 20 F

Average capacitance Average resistance

Regressed capacitance (discharge)

Regressed resistance, mohm

1.15

3080

Regressed capacitance (charge)

3060

3040

3020

1.1

3000

2980

2960

Regressed capacitance, F

1.2

1.05 2940

Regressed resistance (charge) 2920

Regressed resistance (discharge) 1 600

2900

700

800

900

1000

1100

1200

1300

1400

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Time of test, s

Figure 9. Statistical analysis of test results. With reference to Figure 5, 70 discharge and 63 charge tests were conducted; the combined results for the regressed resistances and capacitances are presented in this plot. Figure 6 presents results for a continuous drive through the entire test duration, this plot provides the combined results of all the discrete tests. The ordinates are expanded in this plot relative to Figure 6.

REFERENCES [1] [2] [3] [4] [5]

Applied Optimal Estimation, edited by A. Gelb, M.I.T. Press, Cambridge, MA, 1974. B. D. O. Anderson and J. B. Moore, Optimal Filtering, Prentice-Hall, Englewood Cliffs, NJ, 1979. P. S. Maybeck, Stochastic Models, Estimation and Control, volume 141-1 of Mathematics in Science and Engineering, Academic Press, 1979. B. Widrow and S. D. Stearn, Adaptive Signal Processing, Prentice-Hall, Englewood Cliffs, NJ, 1985. W. L. Brogan, Modern Control Theory, second edition, Prentice-Hall, Englewood Cliffs, NJ, 1985.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

102 [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

[20] [21]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[22] [23] [24] [25] [26] [27] [28] [29] [30]

[31]

[32] [33] [34]

Mark Verbrugge, Damon Frisch, Trudy Weber et al. L. Ljnug and T. Söderström, Theory and Practice of Recursive Identification, MIT Press, 1986. M. G. Bellanger, Adaptive Digital Filters and Signal Analysis, Marcel Dekker, New York, NY, 1987. K. J. Åström and B. Wittenmark, Adaptive Control, Addison-Wesley, 1989. R. Kulhavý, Recursive Nonlinear Estimation. A Geometric Approach, Springer, London, 1996. S. Pillar, M. Perrin, A. Jossen, J. Power Sources, 96(2001)113. M. W. Verbrugge and E. D. Tate, J. Power Sources, 126(2004)236. G. L. Plett, J. Power Sources, 134(2004)252-261, 262-276, and 277-292. C. Pulamarasetty, P. Singh, H. Chen, X. Wang, and D. Reisner, SAE paper 2004-011584, 2004. M. W. Verbrugge, P. Liu, and S. Soukiazian, J. Power Sources, 141(2005)369. M. W. Verbrugge and D. R. Frisch, United States Patent 6947855, September 2005. M. W. Verbrugge, D. Frisch, and B. Koch, J. Electrochem. Soc., 152(2005)A333. H. Asai, H. Ashizawa, D. Yumoto and H. Nakamura, Y. Ochi, SAE paper 2005-010807, 2005. M. W. Verbrugge and B. J. Koch, J. Electrochem. Soc., 153(2006)A187. K. A. Smith, “Electrochemical Modeling, Estimation and Control of Lithium Ion Batteries,” Ph.D. Thesis, Department of Mechanical and Nuclear Engineering, The Pennsylvania State University, 2006. M. W. Verbrugge, "Adaptive, multi-parameter battery state estimator with optimized time-weighting factors," accepted by J. Appl. Electrochem., November 2006. C. Massey, A. Bekaryan, P. Liu, L. Turner, D. Frisch, T. Weber, and M. Verbrugge, SAE paper O5CV-137, 2005. D. H. Fritts, J. Electrochem. Soc., 144(1997)2233. B. E. Conway, Electrochemical supercapacitors: scientific fundamentals and technological applications, Kluwer Academic/Plenum, New York, NY, 1999. R. Kötz and M. Carlen, Electrochim. Acta, 45(2000)2483. A. Burke, J. Power Sources, 91(2000)37. S. Buller, E. Karden, D. Kok, and R. W. DeDoncker, IEEE Transactions on Industrial Applications, 38(2002)1622. A. Chu and P. Braatz, J. Power Sources, 112 (2002) 236. R. B. Wright, D. K. Jamison, and T. Q. Duong, Abstract 237, 204th Meeting, The Electrochemical Society, Inc., 2003. D. Y. Jung, Y. H. Kim, S. W. Kim, and S-H. Lee, J. Power Sources, 114 (2003) 366. M. W. Verbrugge, “Supercapacitors and Automotive Applications,” Proceedings Volume for the World Summit on Advanced Capacitors, Washington, DC, August 2003. L. Li, “Effects of Activated Carbon Surface Chemistry and Pore Structure on the Adsorption of Trace Organic Contaminants from Aqueous Solution,” Ph.D. Dissertation, North Caroline State University, Raleigh, NC, 2002. B. Pillay, “Design of Electrochemical Capacitors for Energy Storage,” Ph.D. Dissertation, University of California, Berkeley, CA, 1996. B. Pillay and J. Newman, J. Electrochem. Soc., 143(1996)1806. V. Srinivasan and J. W. Weidner, J. Electrochem. Soc., 146(1999)1650.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Examination of State Estimators for Electrochemical Energy Storage Devices … [35] [36] [37] [38] [39] [40] [41] [42] [43]

[44] [45] [46] [47] [48] [49] [50] [51]

J. Ong and J. Newman, J. Electrochem. Soc., 146(1999)4360. D. Dunn and J. Newman, J. Electrochem. Soc., 147(2000)820. C. Lin, J. A. Ritter, B. N. Popov, R. E. White, J. Electrochem. Soc., 146(1999)3168. C. Lin, B. N. Popov, H. J. Ploehn, J. Electrochem. Soc., 149(2002)A167. H. Kim and B. N. Popov, J. Electrochem. Soc., 150(2003)A1153. S. Devan, V. R. Subramanian, R. E. White, J. Electrochem. Soc., 151(2004)A905. G. Sikha, R. E. White, and B. N. Popov, J. Electrochem. Soc., 152(2005)A1682. S.R.S. Prabaharan, R. Vimala, Z. Zainal, J. Power Sources, 161(2006)730. M. W. Verbrugge and P. Liu, "On the merits of supercapacitors for vehicle propulsion systems and open questions," proceedings from the Advanced Capacitors World Summit, July 17-19, 2006, San Diego, CA USA. L. Wang, M. Fujita, M. Inagaki, Electrochim. Acta, 51(2006)4096. R. Miller, Electrochim. Acta, 52(2006)1703. M. W. Verbrugge and P. Liu, J. Electrochem. Soc., 153(2006)A1237. M. Sevilla, S. Álvarez, T.A. Centeno, A.B. Fuertes, F. Stoeckli, Electrochim. Acta, 52(2007)3207. Watanabe, H. Nakazawa, M. Itagaki, S. Suzuki, I. Shitanda, J. Power Sources, 164(2007)415. X. Wang and J. P. Zhenga, J. Electrochem. Soc., 151(2004)A1683. S. A. Kazaryan, S. N. Razumov, S. V. Litvinenko, G. G. Kharisov, and V. I. Koganb, J. Electrochem. Soc., 153(2006)A1655. Part 2 of “Well-to-Wheel Energy Use and Greenhouse Gas Emissions of Advanced Fuel/Vehicle Systems – North American Analysis, Executive Summary Report”, Argonne National Laboratory, Center of Transportation Research, 2001. (Posted at http://www.transportation.anl.gov/ publications/index.html.) T. Weber, “Vehicle System Modeling in the Automotive Industry,” ARO/ERC Engine Modeling Symposium, University of Wisconsin, Madison, June 2003.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[52]

103

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Electric Vehicles: Technology, Research and Development ISBN 978-1-60741-142-0 Editor: Gerald B. Raines © 2009 Nova Science Publishers, Inc.

Chapter 5

ON THE FUEL ECONOMY OF HYBRID-ELECTRIC POWERTRAINS

∗

Tomaž Katrašnik Faculty of Mechanical Engineering, University of Ljubljana, Aškerčeva 6, SI-1000 Ljubljana, Slovenia, tel.: +386 1 4771 310, fax.: +386 1 4771 310 e-mail: [email protected] ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The treatise presents an extensive simulation and analytical analysis of the energy conversion phenomena in parallel and series hybrid-electric powertrains. Parameters of both hybrid powertrains are evaluated and compared to parameters of the conventional-internal combustion engine powertrain. Simulation approach is based on an accurate and fast forwardfacing simulation model that is capable of capturing dynamics of the powertrain components. Moreover, the treatise offers an analytical approach based on the energy balance equations in order to analyze and predict energy conversion efficiency in both hybrid powertrains. The analysis covers broad range of parallel and series hybrid powertrain configurations. Very good agreement between simulation and analytical results gives confidence in the accuracy of the performed analysis and confirms the validity of the analytical framework. Combined simulation and analytical analysis enables deep insight into energy conversion phenomena in hybrid powertrains. It reveals advantages and disadvantages of both hybrid powertrain concepts and their variations running under different operating conditions. The analysis thus indicates guidelines that lead to optimum fuel economy of particular powertrain concept operating according to the specified drive-test cycle. It can be concluded from the presented results that: 1.) parallel hybrid powertrain features better fuel economy than the series one for the applied test cycles, 2.) both hybrid powertrain configurations feature the best fuel economy at light duty application and 3.) electric conversion efficiency has significant influence on the fuel economy enhancement of hybrid-electric powertrains.

∗

A version of this chapter was also published in Energy Conversion: New Research, edited by Wenzhong Lin published by Nova Science Publishers, Inc. It was submitted for appropriate modifications in an effort to encourage wider dissemination of research.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

106

Tomaž Katrašnik

NOMENCLATURE FR

fuel rack position (-) torque (Nm) mass (kg) ICE or EM speed (rpm)

M m n nbat

number of battery modules (-)

nc

number of engine cycles (-)

nst

number of piston strokes (-)

P peff

power (W) mean effective pressure of the ICE (MPa)

QLHV

lower fuel heating value (J)

Rbr = Wbr WICE

ratio of braking work to effective work of baseline ICE (-)

REG = WEG ,in WICE

ratio of generator input work to effective work of baseline

RV = Vdownsized Vbaseline t V W

η

ICE (-) swept volume ratio (-) time (s) swept volume of the ICE (m3) work (J) efficiency (-)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Subscripts:

b br c ch dch eff EG EL EM EM , G EM , M ES f h ICE

baseline braking engine cycle charge discharge effective electric generator electric electric motor electric motor operating in the generator mode electric motor operating in the motor mode electric storage fuel, fuel conversion hybrid internal combustion engine

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

in ind max mech min neg

( n (P

n M ICE ,b ,max ICE ,b , max

)

)

out P RB S tc

107

input indicated maximum mechanical minimum negative engine speed at maximum torque of the baseline ICE engine speed at maximum power output of the baseline ICE output parallel regenerative braking series test cycle

Abreviations: CS HF LDA

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

lhs rhs SOC TF

control strategy hybridization factor boost pressure controlled fuel limiting device (German abbreviation for “Ladedruckabhängiger Vollastanschlag”) left hand side of the equation right hand side of the equation state of charge torque factor

INTRODUCTION Global concerns on sustainable energy use and environmental protection place high demands on the energy conversion efficiency and on the exhaust pollutant generation of the ground transportation fleet. On the other hand, customers demand vehicles with improved drivability, i.e. better performances. These contradictory goals obviously call for new technologies to play the fundamental role. New vehicle technologies in the form of electric vehicles (EVs), hybrid electric vehicles (HEVs) and fuel cell vehicles (FCVs) have emerged as possible solutions [1]. Among the alternative powertrains being investigated, the hybrid electric vehicle (HEV) consisting of an internal combustion engine (ICE) and an electric motor (EM) (generally accepted definition of HEV according to Chan [2]) is considered to offer the best promise in the short to mid term due to the use of smaller battery pack and their similarities with the conventional vehicles [1,3].

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

108

Tomaž Katrašnik

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Scheme of a) parallel and b) series hybrid powertrain.

HEV can be assigned to either parallel hybrid (Fig. 1 a)), series hybrid (Fig. 1 b)), or their combination [4]. Parallel hybrids allow both the ICE and the electric motor to deliver power in parallel to drive the wheels. However, in the parallel hybrid configuration, mechanical connection between the components does not allow arbitrary optimization of the ICE operating conditions as is the case in the series hybrids. In parallel hybrids both the engine and the electric motor are generally coupled to the drive shaft of the wheels, and propulsion power may therefore be supplied by the engine alone, by the electric motor alone or by both. The electric motor can be used as a generator to charge the battery by regenerative braking or by absorbing power from the engine when its output is greater than that required to drive the wheels. The parallel hybrid therefore needs only two propulsion devices, since the electric motor can be used as a generator. Another advantage over the series HEV is that a smaller engine and a smaller electric motor can be applied to get the same top performance until the battery is depleted. Therefore, in passenger car application, the parallel hybrid configuration has been used in many HEVs that have come on the market [3]. In a series hybrid, thermal energy is first converted to mechanical energy via an ICE and then to electric energy via an electric generator (EG). The converted electric energy either charges the battery or can bypass the battery to propel the wheels via an electric motor. Although it has an added advantage of simplicity of the drive train, it needs three propulsion devices, i.e. the engine, the generator and the electric motor, which need to be sized for the maximum sustained power if the series HEV is designed to climb a long grade. The majority of presently produced hybrid electric vehicles run on spark ignition gasoline engines (gasoline engines) and the majority of the studies related to the reduction of fuel consumption due to powertrain hybridization are also made with gasoline engines as the ICE power source [1,3,5,6]. On the other hand, it is well known that compression ignition diesel engines (diesel engines) are inherently more efficient in comparison with gasoline engines [79] and the difference is even more pronounced at part loads where gasoline engines supplementary suffer from higher throttling losses. It is therefore reasonable to select diesel engine as the ICE power source to develop hybrid powertrain with high energy conversion efficiency, since it is still the most efficient energy converter of fossil fuels for vehicle

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Energy Conversion Efficiency in Hybrid-Electric Powertrains

109

propulsion [9]. Additionally, the use of diesel engines in HEVs is also favorable in comparison with the gasoline engines due to the added safety of the diesel engines, since higher flashpoint of the diesel fuel provides additional margin of safety during a collision [8]. Analyzing improvement in the fuel economy of hybrid powertrains, it can be concluded that larger improvements are observed for HEVs equipped with gasoline engines in comparison to the baseline gasoline engine vehicles than for HEVs equipped with diesel engines in comparison to baseline diesel engine vehicles. However, these conclusions arise from the fact that lower efficiency platform was selected as the baseline configuration enabling larger improvements in the energy conversion efficiency as presented in the section 2. Recently, many papers concerning simulation based analysis of the fuel economy and pollutant emissions of HEV have been published [1,3,5-8,10-14]. The majority of the simulation models rely on the map based or lookup tables based approach [3,6,7,10-14]. Governing equations are commonly solved by backward-facing models or hybrid backward/forward models that are closely related to the strictly backward-facing models [6,10,11,13,14]. Backward-facing models feature high computational speed, whereas many authors [6,10-12] agree that their drawbacks come from the assumption that the trace is met and from the use of steady-state efficiency or loss maps. Furthermore, dynamic effects are not included in the steady-state maps or in the backward-facing model’s estimate of the energy use, whereas backward-facing models do also not deal in the quantities measurable in a vehicle [6,10-12]. Backward-facing models are fundamentally empirical models [11]. It should also be noted that map based approaches have their limitations when adequate and accurate scaling of the ICEs is concerned. Moreover, dynamic behavior of the turbocharged engines could not be properly modeled due to the use of steady-state data functions. On the other hand, it is obvious that dynamic models can be included naturally in a forward-facing vehicle model, since forward-facing model deals in the quantities measurable in an actual powertrain [7,10-12]. In order to adequately model steady-state and dynamic operation of the ICE it is necessary to properly consider geometrical and mechanical characteristics of the ICE and related thermo-dynamical and fluid-mechanical processes, especially when dynamic models of the turbocharged ICEs are concerned. Mechanistic models also enable a straight forward modeling of downsized engines. According to Ref. [10-12], the main argument against the use of forward-facing approach is its large time-consumption. However, a real time forward facing model adequate for simulating dynamic operation of hybrid powertrains was introduced by the author in Refs. [15,16] and is summarized subsequently. Considering all previously discussed facts it can be concluded that it is necessary to build a fast and accurate forward-facing simulation model adequate for simulating dynamic operation of the hybrid powertrains in order to adequately represent their dynamic behavior. This treatise is intended to provide fundamental knowledge of the energy flow and of the energy loss phenomena in the parallel and series hybrid powertrains. Therefore, energy balance equations that enable more profound analysis of the energy conversion efficiency than simple comparison of the fuel consumption of different powertrain configurations solely, were derived for both hybrid powertrains. The motivation for developing hybrid vehicles is obviously striving for higher fuel economy. However, in order to achieve this goal it is necessary to have knowledge of the influences of the hybrid powertrain configuration, of the energy flows through their constituting components, their efficiencies etc. on the fuel consumption of the particular hybrid powertrain. Analytical framework that enables analysis of separate energy flows and energy losses of particular energy paths was therefore derived.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

110

Tomaž Katrašnik

Fast and accurate simulation model that is also introduced in the treatise, provides input parameters for the analytical analysis on the one hand, whereas the simulation results of the fuel consumption are also used to verify the results of the analytical analysis on the other hand. Thus, combined simulation and analytical analysis enables deep insight into energy conversion phenomena in hybrid powertrains and highlights the conditions and influences that lead to improved fuel economy of particular hybrid powertrain. Analytical background and analysis of the parallel hybrid powertrain configurations were already introduced by the author in Refs. [15,17], whereas analytical background of series hybrid powertrain and a comparative analysis were presented in Refs. [16,18]. This treatise therefore focuses on extensive and detailed analysis of the energy conversion phenomena in parallel and series hybrid powertrains with the emphasis on similarities and differences between both powertrain configurations, on the size and efficiency of their components and on the applied drive cycles.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ANALYTICAL FRAMEWORK An analytical approach to energy conversion phenomena in parallel and series hybrid powertrains is presented in this section. Analytical analysis of the energy conversion phenomena in hybrid powertrains is derived from the energy balance equations under consideration of energy production, storage and consumption, as well as component efficiencies. It is widely accepted that hybridization of the powertrain allow for improved fuel economy [1,3-5,7,8,19-21], which results in the reduction of pollutant emissions [35,7,19,21]. However, this statement is not unconditional, since powertrain configuration, i.e. parallel, series, series-parallel or complex HEV [2,4], its constituting components, and their relative power ratios and efficiencies as well as the applied driving-test cycle significantly affect the fuel economy of hybrid powertrains. Hybrid powertrains inherently posses the following mechanisms that could allow for fuel economy enhancement over the conventional powertrain systems: 1) regenerative braking, 2) operating of downsized ICE at higher efficiency and 3.) application of stop&start strategy. It is obvious that facts brought forward in all three items possess the potential to improve fuel economy of hybrid powertrains. However, it is also necessary to account for losses due to electric energy production, storage and consumption when analyzing the energy conversion efficiency of the hybrid powertrains. Analysis of the fuel consumption is commonly performed for a specific test cycle. All the quantities defined subsequently are therefore averaged values over the whole test cycle. It should be noted that production and consumption of the energy is considered by the appropriate sign and thus all values of the work are considered positive. It is obvious that it is necessary to analyze fuel consumption of hybrid powertrains with balanced SOC of the electric storage systems, i.e. SOC of all electric storage systems at the beginning of the test cycle should equal to SOC at the end of the test cycle. It can be concluded that energy conversion efficiency of the powertrain could be related to its fuel consumption during the test cycle, if SOC of the batteries is balanced.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

111

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Energy Conversion Efficiency in Hybrid-Electric Powertrains

Figure 2. Test cycle points and power outputs/inputs of a) ICE powertrain, b) parallel and c) series hybrid powertrain.

ICE powertrain Let us first consider the conventional powertrain, i.e. consisting of an ICE only; in the subsequent text conventional powertrain denotes a powertrain consisting of a baseline ICE. Energy consumed to propel the vehicle or dynamometer according to the test cycle ( Wtc ) is equal to the energy produced by the ICE ( WICE ) reduced by the energy consumed by the brakes ( Wbr ); Fig. 2.a). Thus, ttc

ttc

ttc

0

0

0

Wtc = ∫ Ptc dt = WICE − Wbr = ∫ PICE dt − ∫ Pbr dt ,

(1)

where Ptc is the instant power imposed by the test cycle, PICE is the instant power of the ICE and Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

112

Tomaž Katrašnik

Pbr (t ) = PICE ,min (t ) − Ptc (t )

(2)

is the difference between negative power required to motor the engine at particular speed ( PICE ,min ) and negative torque imposed by the test cycle. Let us define the effective efficiency of the ICE as the ratio of the energy delivered by the ICE to the energy supplied by the fuel, thus ttc

η eff

∫0 PICE dt WICE = = . m f ,tc QLHV m f ,tc QLHV

(3)

The effective efficiency is equal to product of the fuel conversion efficiency and the mechanical efficiency of the ICE

η eff = η f η mech .

(4)

The fuel conversion efficiency of the ICE is defined as the ratio of work performed to the piston during the engine cycle, i.e. ratio of indicated work (indicated work of the cycle ( Wc ) multiplied by the number of engine cycles ( nc ) during the test cycle) to energy supplied by the fuel (mass of fuel injected during the test cycle ( m f ,tc ) multiplied by the lower fuel

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

heating value ( QLHV ))

nc ∫ pcyl dV Wc nc ηf = = c . m f ,tc QLHV m f ,tc QLHV

(5)

Mechanical efficiency of the ICE is defined as the ratio of the energy delivered by the ICE to indicated work performed during the test cycle, thus t tc

∫P

ICE

η mech =

dt

0

nc ∫ p cyl dV

.

(6)

c

Combining eq. (1) and eq. (3) gives

Wtc = η eff m f ,tc QLHV − Wbr .

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

(7)

113

Energy Conversion Efficiency in Hybrid-Electric Powertrains

Parallel hybrid powertrain Considering the parallel hybrid powertrain (Fig. 1 a)), it is evident from Fig. 2 b) that energy consumed to propel the vehicle or the dynamometer according to the test cycle is produced by the ICE and by the EM, whereas electric storage systems are charged by regenerative braking and/or by running the ICE at higher torque. In the parallel hybrid powertrain electric energy is never produced and consumed simultaneously, since they apply only one electric machine that operates either as the electric motor or as the electric generator. The energy balance of parallel hybrid powertrain is therefore equal to

Wtc = WICE ,h + WEM ,out − WEG ,in − Wbr ,h ,

(8)

where

WEM ,out = η ELWEG ,in ,

(9)

since SOC of all electric storage systems at the beginning of the test cycle is equal to SOC at the end of the test cycle. The electric energy conversion efficiency is defined as

η EL = η EGη EM η ES ,

(10)

where the efficiency of the electric generator is defined as ttc

ttc

0

0

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

η EG = ∫ PEG ,out dt

∫P

EG ,in

dt ,

(11)

the efficiency of the electric motor is defined as ttc

η EM = ∫ PEM ,out dt 0

ttc

∫P

EM ,in

dt

(12)

0

and the efficiency of the electric energy storage systems, i.e. charge-discharge efficiency of electric storage systems and efficiency of electric energy transfer that also considers losses of power converters, is defined as

η ES = η ES ,chη ES ,dischη eet .

(13)

Inserting eq. (9) into eq. (8) and combining with eq. (1) gives after rearrangement

Wtc = WICE ,h − (1 − η EL )WEG ,in − Wbr ,h = WICE − Wbr .

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

(14)

114

Tomaž Katrašnik Inserting eq. (3) into eq. (14) gives

Wtc = η eff ,h m f ,tc ,h QLHV − (1 − η EL )WEG ,in − Wbr ,h = η eff m f ,tc QLHV − Wbr

(15)

and after rearrangement

m f ,tc ,h m f ,tc

=

η eff η eff ,h

⎛ (1 − η EL )WEG ,in Wbr ,h − Wbr ⎜⎜1 + + WICE WICE ⎝

=

η eff η eff ,h

⎛ Wtc + (1 − η EL )WEG ,in + Wbr ,h ⎜⎜ Wtc + Wbr ⎝

⎞ ⎟⎟ ⎠

(16)

or

m f ,tc ,h m f ,tc

⎞ ⎟⎟ . ⎠

(17)

It is evident that a hybrid powertrain utilizes fuel energy during the test cycle more efficiently than a conventional powertrain if

m f ,tc ,h m f ,tc

< 1.

(18)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

or considering eq. (16)

η eff ,h (1 − η EL )WEG ,in Wbr ,h − Wbr . > 1+ + η eff WICE WICE

(19)

It can be concluded that parallel hybrid powertrain utilizes fuel energy more efficiently than conventional powertrain if the ratio of the effective efficiency of the ICE in hybrid powertrain to the effective efficiency of the baseline ICE, i.e. lhs of eq. (19), is larger than rhs of eq. (19) that includes constant term, i.e. 1, losses due to electric energy production, storage and consumption, and the term considering positive effects due to regenerative braking. Alternative interpretation arises from eq. (17): parallel hybrid powertrain utilizes fuel energy η eff ,h > η eff and more efficiently than conventional powertrain if

((1 − η )W EL

EG ,in

+ Wbr ,h ) < Wbr , or if either of the two effects prevails over the other one.

Series hybrid powertrain Considering series hybrid powertrain (Fig. 1 b) and 2 c)), it is evident that energy supplied by the ICE is converted into electrical energy via electric generator and is then used to propel the wheels by the electric motor and/or to charge the electric storage systems. The electric storage systems could also be charged by regenerative braking via operation of the

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

115

Energy Conversion Efficiency in Hybrid-Electric Powertrains

electric motor in the generator mode. Considering Fig. 2 c) it can therefore be concluded that the electric motor operating in motor mode ( WEM ,out ) is applied to deliver power when torque of the test cycle is positive, whereas energy is consumed by regenerative braking ( WRB ) and braking ( Wbr , h ) when torque of the test cycle is negative

Wtc = W EM ,out − W RB − Wbr ,h .

(20)

Electric energy consumed by the electric motor is supplied by the electric generator and/or by the electric storage systems

WEM ,out = η EM , M (η eet , EG → EM WEG → EM + η eet , ES → EM WES → EM ) .

(21)

The electric energy transfer efficiencies (η eet ) were redefined in order to account for electric energy losses of particular energy paths; additional index indicates particular energy path. Efficiencies of electric energy transfer also consider losses of power converters. The value of the electric energy always refers to the amount of the energy generated by the source. It was already stated earlier that the series hybrid powertrain consists of two electric machines: of the electric generator and of the electric motor that can operate in the motor mode to propel the vehicle or in the generator mode to recuperate the energy via regenerative braking. The efficiency of the electric generator in the series hybrid powertrain (η EG ) is equal to

η EG of the parallel hybrid powertain (eq. (11)), and efficiency of the electric motor

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

operating in the motor mode ( η EM , M ) is equal to

η EM of the parallel hybrid powertain (eq.

(12)), whereas efficiency of the electric motor operating in the generator mode, i.e. regenerative braking, is defined by ttc

η EM ,G = ∫ PEM → ES dt 0

ttc

∫P

RB

dt .

(22)

0

Energy produced by the ICE is used to propel the electric generator

WICE = WEG ,in = WEG ,out η EG ,

(23)

whereas electric energy produced by the electric generator is consumed by electric motor and/or used to charge the batteries

WEG ,out = WEG → EM + WEG → ES .

(24)

The electric energy recuperated by the regenerative braking is used to charge the electric storage systems

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

116

Tomaž Katrašnik

WEM → ES = η EM ,GWRB .

(25)

The energy balance equation for the electric storage systems is

WES → EM = η ES (η eet , EG → ES WEG → ES + η eet , EM → ES WEM → ES ) , since SOC of all electric storage systems is balanced.

(26)

η ES is also redefined when

considering series hybrid powertrain

η ES = η ES ,chη ES ,disch .

(27)

Combining eqns. (20,21) and (23,24,25,26) gives after rearrangement

Wtc = η EGηeet,EG→EMη EM ,M WICE − η EM ,M (ηeet,EG→EM − ηeet,EG→ESη ESηeet,ES→EM )WEG→ES − (1 − η EM ,Gηeet, EM →ESη ESηeet,ES→EMη EM ,M )WRB − Wbr,h

(28) Combining (1) and (28) and inserting (3) gives

Wtc = ηEGηeet,EG→EMηEM,Mηeff ,h m f ,tc,hQLHV −ηEM,M (ηeet,EG→EM −ηeet,EG→ESηESηeet,ES→EM ) ×WEG→ES − (1 −ηEM,Gηeet,EM→ESηESηeet,ES→EMηEM,M )WRB − Wbr,h = ηeff m f ,tcQLHV − Wbr

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

(29) and after rearrangement

m f ,tc,h m f ,tc

=

ηeff ⎛ 1 ⎜⎜1 + [ηEM,M (ηeet,EG→EM −ηeet,EG→ESηESηeet,ES→EM ) ηEGηeet,EG→EMηEM,Mηeff ,h ⎝ WICE

×WEG→ES + (1 −ηEM,Gηeet,EM→ESηESηeet,ES→EMηEM,M )WRB ] +

Wbr,h − Wbr ⎞ ⎟ WICE ⎟⎠ (30)

Considering inequality (18) leads to

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

.

117

Energy Conversion Efficiency in Hybrid-Electric Powertrains

η EGη eet , EG→EM η EM ,M η eff ,h 1 [η EM ,M (η eet,EG→EM − η eet,EG→ESη ESη eet,ES →EM ) > 1+ η eff WICE × WEG→ES + (1 − η EM ,Gη eet , EM →ESη ESη eet , ES →EM η EM ,M )WRB ] +

Wbr ,h − Wbr WICE (31)

A series hybrid powertrain utilizes fuel energy more efficiently than conventional powertrain if the ratio of the energy conversion efficiency of the energy path ICE→EG→EM (η EGη eet , EG → EM η EM , M η eff ,h ) to the effective efficiency of the baseline ICE, i.e. lhs of eq. (31), is larger than rhs of eq. (31) that includes constant term, i.e. 1, electric energy losses due to energy paths EG→ES→EM and RB→ES→EM, and the term considering positive effects due to regenerative braking.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

SIMULATION MODEL AND POWERTRAIN CONFIGURATIONS The scope of the proposed treatise is to analyze the energy conversion efficiency of hybrid and conventional powertrains. The simulation model should therefore assure accurate evaluation of energy flows and energy losses. Power transfer between the components is calculated by corresponding equations presented in section 2. in differential form according to the corresponding control strategies presented in this section. A forward-facing model was applied for the analysis of both hybrid powertrain configurations. Simulation software is an extended version of author’s thermodynamic and fluid mechanics simulation code for modeling the stationary and transient operation of various configurations of ICEs [22-25]. For the purpose of this analysis simulation model was upgraded with the models of other constituting components of parallel and series hybrid powertrains as well as with the model incorporating power control strategies. The simulation code is capable of simulating various loads including vehicle dynamics as well as engine dynamometer and chassis dynamometer test cycles, whereas conventional ICE or hybrid powertains are applied as power sources. Simulation model is very fast when compared with other forward-facing dynamic simulation models, i.e. it is capable of performing real time simulations of hybrid powertrains with integration time steps 1-5×10-5s (depending on the engine speed) on a single processor PCs. It is therefore not significantly slower than the backward facing model ADVISOR with solution time on the order of 1/10 real time [10]. However, it should be noted that time steps of backward facing models that typically amount to 1s, as proposed in [11,12], do not allow simulation of the dynamics of hybrid powertrain components during transient operation.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

118

Tomaž Katrašnik

Internal combustion engine model The simulation model of the ICE, introduced in Ref. [22] and briefly described in Refs. [23-25], is based on the 0-D filling and emptying method. The differential equations for mass and energy conservation are solved in each integration time step throughout the entire engine, bypassing the need to prepare engine subsystem maps in advance as they could be indirectly influenced by the electric devices. It was shown in Ref. [26] that transient engine parameters such as ICE power output, boost pressure and turbocharger speed calculated with 0-D and with 1-D model introduced in Ref. [26] coincide very well and that results of both models agree very well with experimental data. Therefore, the use of 0-D model is preferred for the purpose of this analysis due to much higher computational speed. Table 1. Basic data of the MAN turbocharged diesel engine

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Engine Number of cylinders Bore [mm] Stroke [mm] Compression ratio Max. torque [Nm] Max. power [kW] Turbocharger

MAN D 0826 LOH 15 6 108 125 18:1 862@1400rpm 158@2400rpm HOLSET H1E-8264BF- H16WA8

The MAN D0826 LOH 15 (Table 1) turbocharged diesel engine is applied as the baseline internal combustion engine; denoted baseline ICE in the subsequent text. Much effort was devoted to the accuracy of the turbocharger simulation, as it in turn, via boost pressure and boost pressure controlled fuel limiting device (LDA), determines the quantity of fuel injected and therefore the engine power output during transient operating regime. Special care was also devoted to precise measuring of mechanical losses, since mechanical efficiency essentially influences effective efficiency of the ICE as indicated in section 2. Model of mechanical losses is therefore based on highly accurate experimental data [27]. It is furthermore necessary to scale model of mechanical losses appropriately in order to adequately represent downsized engines. Scaling factors based on performed analysis [27] agree very well with the results reported by Thiele [28]. The accuracy and applicability of the 0-D code depends strongly on the amount of experimentally determined input data available. Experimentally determined flow coefficients of inlet and exhaust valves, characteristics of the fuel delivery pump and those of the boost pressure controlled fuel limiting device, intercooler, twin entry turbine and combustion parameters based on the results of the analysis of measured in-cylinder pressure diagrams etc. were therefore applied in the simulation code. A reliable and accurate computer simulation program compatible with the experimental results in stationary and transient engine operation was thus developed [22-25]. Agreement between measured and calculated engine power, boost and exhaust pressure as well as turbocharger speed was better than 4% for different steady-state engine running conditions. Accuracy of the results of transient engine operation is of crucial importance for the purpose of the presented investigation. Fig. 3 represents simulated and measured engine dynamometer

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

119

braking force for a sudden load increase from peff =0.4 to 1.4 MPa at 1500 rpm [22-25]. Mean effective pressure is defined by

p eff =

n st πM , V

(32)

where nst =4 for 4-stroke engines and nst =2 for 2-stroke engines. Control system of the

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

applied engine dynamometer was not designed for transient measurements, therefore, oscillations of the engine speed occurred during the initial sequence of the engine transient operation. Simulated braking force of the dynamometer was calculated from the instantaneous computed engine power whereas corresponding time derivative of the measured instantaneous engine speed presented the input parameter of the simulation. Very good agreement between measured and simulated dynamometer braking force (Fig. 3) implies: 1.) agreement between measured and simulated turbocharger speed that in turn defines boost pressure and, via LDA, cyclic fuel delivery, 2.) agreement in the mass flow through the engine that, in connection with the cyclic fuel delivery, determines indicated mean effective pressure and 3.) adequacy of the model of mechanical losses that determines brake mean effective pressure.

Figure 3. Comparison between measured and simulated dynamometer braking force for

peff = 0.4→1.4

MPa.

Models of electric components An electric machine provides a fast response to torque and speed demand [29]. The average ratings of electric systems can be derived from steady-state operation and characteristics of the systems, but peak ratings estimated from the steady-state results are often inadequate [30]. It is therefore necessary to use sufficiently small calculation time steps to encounter for peak values during large load transients, such as those which occur during fast acceleration of the vehicle or dynamometer. However, according to Amrhein and Krein [30] tradeoffs are to made in terms of detail versus simulation time, since detailed behavior of

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

120

Tomaž Katrašnik

a dc-to-dc converter switching at 100 kHz of higher, for example, requires nanosecond time resolution. In this study switching power converters are therefore implemented as averaged models, as proposed by Amrhein and Krein [30]. Whereas, time steps dictated by the ICE model (1-5×10-5s) are sufficiently small to accurately model dynamics of the electric systems (other than dc-to-dc converters), in contrast to backward-facing approach models [11,12] that typically use time steps of 1s. It is discernible from analysis presented in section 2. that efficiency of electric devices strongly affects energy conversion efficiency of hybrid powertrains. Electric machines and power converters have reached a high level of maturity, whereas electric storage devices represent an intensively researched area. Analysis of the energy conversion efficiency presented in this analysis is therefore preformed for batteries and for ultracapacitors offering significantly higher charge-discharge efficiencies. Battery charging, discharging and determination of the SOC was modeled according to the model proposed by Kutluay et al [31]. The battery charge/discharge model takes into account the change of actual capacity with the charge/discharge current and temperature. Kutluay et al [31] reported that the accuracy in SOC estimation better than 3-4% has been obtained for all operating conditions. Considering batteries, two control strategies are widely applied and investigated in the literature (see also Ref. [1,5,8]): 1) charge sustaining control strategy, which ensures that a battery is never discharged below a certain limit and 2) charge sustaining constant performance strategy, where batteries are of such size that they do not limit the performance of the electric motor. In this study a third constraint was placed on the number of battery modules claiming for charge-discharge efficiency of the batteries (charging efficiency multiplied by discharging efficiency) of approximately 65%. This value is in a good agreement with the data published in Refs. [32-34]. Number of battery modules was thus selected according to the strictest of the three constraints: 1.) SOC of the battery units must be greater than 0.4, 2.) batteries are of such size that they do not limit the performance of the electric motor and 3.) charge-discharge efficiency of the batteries is approximately 65%. The first constraint prevents damage to the battery due to deep discharge and prolongs the life of the battery [8], whereas discharge does not cause major drifts in the battery open circuit voltage [35]. The simulations are performed with characteristics of new batteries. Aging of the batteries is therefore not taken into account. The batteries are not charged above SOC>0.75, since charging efficiency drops significantly for higher SOC [35-38]. The Genesis 12V, 28 Ah VRLA battery is considered as the module of the storage system. Basic storage module used in the simulation consists of 25 batteries connected in series, whereas the number of modules is selected upon the criteria outlined in the previous paragraph. The number of battery modules is discrete, therefore an appropriate number of battery modules in best agreement with proposed criteria was chosen. Section 4. also comprises results of a parametric study on the influence of the electric storage efficiency on the fuel consumption of the hybrid powertrains, where number of battery modules was varied to alter the charge-discharge efficiency. Ultracapacitors have many merits compared to the batteries, especially with respect to specific power at high rate, thermal stability, charge-discharge efficiency and life-cycle [34]. Ultracapacitors have the advantage of near-instantaneous energy delivery in contrast to batteries, which experience high internal losses if they are discharged too quickly. Energy stored in ultracapacitors can be transferred to the dc-bus at nearly any discharge rate. When

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

121

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

developed for vehicles, ultracapacitors could be expected to last as long as a car [13]. This is because it is possible to cycle ultracapacitors very quickly without having the large decrease in life cycle that most chemical batteries experience. Thus, ultracapacitors are of considerable interest for HEV applications [30]. Typically, ultracapacitors have much higher power density and lower energy density compared with batteries, however, UC models with high energy density have been produced recently as indicated below. The charge-discharge efficiency of the ultracapacitor is generally in the range from 85% to 98%, according to Ref. [32-34,39]; this value is also in good agreement with the results presented in this analysis - section 4. Therefore, the main constraint placed on the number of ultracapacitor modules is keeping the ultracapacitor SOC above certain minimum value, since power density and charge-discharge efficiency are very high. According to Mierlo et al [13], an ultracapacitor voltage drop of 50% can be allowed, which corresponds to 75% of the energy content. A model of the ultracapacitor consists of the capacitance and an equivalent series resistance, as proposed by Amrhein and Krein [30]. The NESSCAP EMHSP-0051 C0-340R0 ultracapacitor with the nominal capacitance of 51 F and rated voltage of 340 V was applied when analyzing hybrid powertrain parameters with UC as the electric energy storage device. The NESSCAP ultracapacitor features high energy density that assures the use of ultracapacitor as the unique electric energy storage device.

Figure 4. Characteristics of the base electric machine, [15].

The model of the electric motor evaluates torque output, electric energy consumption and efficiency estimation based on measured input data of the electric machine and input signals from other sub-models of the simulation model. Similarly, electric energy production, required torque input and efficiency estimation are determined in the model of the electric generator. A prototype electric motor-generator produced and tested by ISKRA Avtoelektrika d.d. was applied in the parallel hybrid powertrain and as electric motor in the series one.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

122

Tomaž Katrašnik

Performance characteristics of the electric motor-generator are presented in Fig. 4. The electric machine was scaled according to constraints given in the section 3.3. in order to ensure required torque output of the powertrain, whereas its efficiency characteristics were also modified simultaneously according to the data provided by the manufacturer. Characteristics of STAMFORD UCM 274F (max. input power 94.6kW, max. efficiency 93%) were used to simulate electric generator in the series hybrid powertrain. Switching power converters are implemented as averaged models, as proposed by Amrhein and Krein [30].

Powertrain configurations

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Let us first analyze increase of the effective efficiency of the ICE due to its downsizing. Effective efficiency is inversely proportional to brake specific fuel consumption of the ICE and could thus be related to the fuel economy of the ICE. Effective efficiency is equal to the product of fuel conversion efficiency and mechanical efficiency of the ICE as stated in eq. (4).

Figure 5. Fuel conversion efficiency of the MAN turbocharged diesel engine, [15].

Fig. 5 shows measured fuel conversion efficiency (eq. (5)) of the MAN turbocharged diesel engine. It is discernable that fuel conversion efficiency features the highest values at high engine mean effective pressure, i.e. high engine load, and medium engine speed. High fuel conversion efficiency represents the optimum between combustion parameters, heat losses to the combustion chamber walls and work consumed/produced during the working medium exchange phase. It should be noted that combustion parameters always have to be optimized in compliance with the engine emissions.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

123

Figure 6. Mechanical efficiency of the MAN turbocharged diesel engine, [15].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Analyzing measured mechanical efficiency (eq. (6)) of the MAN turbocharged diesel engine (Fig. 6) it can be concluded that mechanical efficiency increases with decreasing engine speed and with increasing engine mean effective pressure. Both facts are quite intuitive, since mechanical losses increase with increasing engine speed, whereas relative influence of the mechanical losses diminishes with increasing mean effective pressure of the engine at the same engine speed.

Figure 7. Effective efficiency of the MAN turbocharged diesel engine.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

124

Tomaž Katrašnik

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 8. Effective efficiency of the downsized ICE;

RV =0.5.

Resulting from fuel conversion efficiency and mechanical efficiency, high effective efficiency area of the MAN turbocharged diesel engine (Fig. 7) is moved towards lower engine speed compared to the high fuel conversion efficiency area. Engine operating point corresponding to the fully loaded vehicle running at a constant speed of 50 km/h in the fifth gear on a flat road is indicated in the Fig. 7. It is discernable that the engine operates at low engine mean effective pressure and therefore at low effective efficiency of the engine. This phenomenon is typical for steady-state operation of vehicle engines, except for operation near the maximum vehicle speed, whereas the phenomenon is more pronounced for the light duty vehicles. The difference between maximum engine torque and steady-state torque could be used for vehicle acceleration or hill climbing and thus influences performance and drivability of the vehicle. It is obvious that mean effective pressure should be increased at the same power output, i.e. downsizing of the ICE, in order to attain higher effective efficiency and thus lower fuel consumption at equal vehicle operation regime. Fig. 8 shows the same vehicle operating point for the downsized ICE featuring 50% of the baseline engine swept volume and equal engine configuration. It is discernable that the effective efficiency of the downsized engine increased for approximately 50%. It is therefore obvious that downsizing of the ICE represents one of the major mechanisms that could allow for fuel economy enhancement of hybrid powertrains over the conventional powertrain systems as introduced in section 2. Turbocharged MAN diesel engine was downsized considering geometrical and mechanical scaling analogy of the engine, i.e. all engine components including turbocharger were scaled in compliance with decrements in the swept volume of the ICE. Simulation model was thus run for all downsized engines reflecting adequate thermo-dynamical and fluid-mechanical processes in the downsized engines. As the result, slightly lower effective efficiency of the downsized engine could be observed for the downsized engine (Fig. 7 and 8)

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

125

Energy Conversion Efficiency in Hybrid-Electric Powertrains

due to higher mechanical losses, higher heat transfer losses, higher pressure losses and lower efficiency of the turbocharger of the downsized engine. However, considering indicated operating point it can be concluded that decrease in the effective efficiency due to downsizing is negligible compared to increase of the effective efficiency due to operation of the downsized engine at higher mean effective pressure. Applying only the downsized ICE to power the vehicle can result in inadequate transient response and poor drivability [40,41] due to reduced maximum power output of the ICE. In order to keep an acceptable level of driving comfort and performance for the user it is thus necessary to increase specific power output of the ICE or to apply the assistance of the electric motor. Specific power output of the engine could be increased by increasing engine speed or by increasing maximum mean effective pressure. Increasing of the engine speed is not favorable when high effective efficiency of the ICE is concerned due to higher mechanical losses, and thus power output should be increased by increasing the maximum mean effective pressure. However, modern turbocharged ICEs commonly feature high maximum mean effective pressures and it is not feasible to increase them significantly. Application of an electric motor in parallel hybrid powertrains or application of an adequate electric motor in series hybrid powertrains together with application of high efficiency downsized ICE could thus represent an environmental sustainable powertrain configuration. In the presented analysis, the swept volume of the turbocharged diesel engines was scaled in decrements of 10% of the baseline engine swept volume down to 50% of the baseline one that represents the smallest ICE that still complies with all analyzed test cycles and sustains battery SOC; more details are provided in the section 4. The swept volume ratio is introduced to represent the downsized engines

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

RV =

Vdownsized . Vbaseline

(33)

The ICEs with RV equal to 0.9, 0.8, 0.7, 0.6 and 0.5 corresponding to mild-, semi- and strong hybrids (classification according to Ref. [6]) were analyzed with the parallel hybrid powertrain, and the ICEs with RV equal to 0.7 and 0.5 and different control strategies were analyzed with the series hybrid powertrain. Cyclic fuel delivery of the ICE is scaled according to its swept volume in order to enable clear and comprehensive analysis. Fig. 9 shows maximum mean effective pressures of the downsized ICEs; labels denote RV . It can be seen that maximum mean effective pressure of downsized ICEs is smaller than that of the baseline one due to higher mechanical losses, higher heat transfer losses, higher pressure losses and lower efficiency of the turbocharger of the downsized engine as discussed previously, whereas the difference is more pronounced below 1600 rpm. Cyclic fuel delivery is limited by the boost pressure below 1600 rpm and thus lower efficiency of a smaller turbocharger directly influences power output of the engine via boost pressure and boost pressure controlled fuel limiting device; see Ref. [22-25] for more details.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

126

Tomaž Katrašnik

Figure 9. Mean effective pressures of the baseline and of downsized ICEs; labels denote

RV .

The components of both hybrid powertrains were sized according to the following constraints

(M

ICE , h

+ M EM ) n (M

ICE , b , max

)

= M ICE ,b n (M ICE , b , max )

(34)

for the parallel hybrid powertrain and

M EM n (M = M ICE ,b n (M ICE , b , max ) ICE , b , max )

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

for the series one, where

(

n M ICE ,b ,max

(35)

)

represents engine speed that corresponds to the

maximum torque of the baseline engine. Conventional, i.e. ICE, powertrain and both hybrid powertrains thus feature the same maximum torque at n (M ICE ,b ,max ) . The above proposed approach was adapted due to typically different torque characteristics of the EM (Fig. 4) and of the ICE (Fig. 9; torque is equivalent to peff for particular HF). Fig. 10 shows maximum torques of parallel hybrid powertrain configurations defined by eq. (34). It is discernable from Fig. 10 that it is more reasonable to compare parallel hybrid powertrains with the same maximum torque at n (M ICE ,b ,max ) than to compare parallel hybrid powertrains with the same power output at the engine speed that corresponds to the maximum power output of the baseline engine, n (PICE ,b , max ) , since torque outputs of the parallel hybrid powertrains with small ICEs, i.e. small RV s, would be significantly higher at lower ICE/EM speeds. The adequacy of the proposed approach is also confirmed when analyzing vehicle accelerations and maximum speeds of the vehicle for powertrain configurations shown in Fig. 10 that is presented in Ref. [17], since maximum speed of the vehicle and time required for acceleration from 0-100 km/h are similar for all powertrains, whereas time required for acceleration from 0-50 km/h acceleration reduces with decreasing RV . Similar conclusion could also be drawn for the series hybrid powertrain.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

127

Energy Conversion Efficiency in Hybrid-Electric Powertrains

Figure 10. Torque output of powertrains with different hybridization factors; labels denote

RV , [15].

Hybridization factor (HF’) [1,5-8] can be evaluated when parallel hybrid powertrain is considered

HF' =

PEM . PEM + PICE ,h n (P ICE , b , max )

(36)

However, considering eq. (34) and the above analysis it is more convenient to define alternative hybridization factor (HF), as proposed in [15], by

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

HF =

M EM

M EM + M ICE ,h = const. n (M

.

(37)

ICE , b , max )

Hybridization factors of particular parallel hybrid powertrains are: HF=0.101 and HF’=0.083 for RV = 0.9 , HF=0.203 and HF’=0.17 for RV = 0.8 , HF=0.304 and HF’=0.26 for RV = 0.7 , HF=0.42 and HF’=0.362 for RV = 0.6 , and HF=0.531 and HF’=0.464 for

RV = 0.5 . It can be noted that HFs of downsized ICEs do not coincide exactly with RV due to lower maximum mean effective pressure of the downsized engines as presented in Fig. 9.

Driving test cycle Engine parameters are evaluated according to the ETC (European Transient Cycle) engine dynamometer transient cycle [42]. An engine dynamometer version of the ETC was chosen rather than a vehicle one, since it enables adequate evaluation of the changes in the powertrain configuration solely, excluding the influences of the gearshift strategy, vehicle parameters, and control strategies during the vehicle stops. The number of the degrees of

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

128

Tomaž Katrašnik

freedom is therefore substantially reduced enabling clearer comparison of the results and application of less sophisticated and universally applicable control strategies. For generating the ETC test points, the engine needs to be mapped prior to the test cycle in order to determine the speed vs. torque curve, as proposed in Ref. [42]. It is discernable from the section 3.3. that all hybrid powertrains feature the same peak torque at the engine speed that corresponds to the maximum torque of the baseline engine and that all analyzed hybrid powertrains have the same minimum and maximum engine speed. It is therefore possible to drive all powertrain configurations according to the same transient test cycle, thus ensuring credibility of the comparisons of different powertrain configurations. Engine speed derivative represents the input parameter of the simulation model, whereas engine/motor and dynamometer torque are used to govern the control units of the ICE and/or the electric motor-generator. This procedure assures agreement between input and actual engine speed, whereas some discrepancies between input torque and engine output torque could arise due to transient response lag of the turbocharged engines. Therefore, it is not possible to define the representative ETC test cycle for all powertrain configurations in advance despite similar maximum torque vs. speed characteristics of different hybrid powertrain configurations (Fig. 10). In order to obtain a representative cycle, each of the analyzed hybrid powertrains was driven according to the appropriate ETC. It is obvious that engine speed traces agree perfectly, however it is necessary to determine torque trace that could be met by all powertrain configurations in order to adequately evaluate their performances. Therefore representative test cycle composed of minimum values of the time depending torque values of all analyzed hybrid powertrains was set up, including negative torques. It should be noted that such representative cycle still complies with the regulations in section 3.9.2. and 3.9.3. in Ref. [42] that regulate test run validation. According to section 3.8.1. in Ref. [42] engine speed and torque command set points shall be issued at 5 Hz or greater, which is easily handled with the proposed simulation model, however, the majority of backward-facing models do not fulfill this demand. According to section 3.7. in Ref. [42] test shall start after engine preconditioning phase, therefore, simulation is started with hot engine. Section 2.2. in Ref. [42] orders that negative torque values of the test cycle motoring points shall take on unnormalized values determined in either of the following ways: negative 40 % of the positive torque available at the associated speed point or mapping of the negative torque required to motor the engine from minimum to maximum mapping speed. Negative 40 % of the positive torque available typically features much larger absolute value in comparison to the negative torque required to motor the engine at particular engine speed and therefore represents vehicle driving conditions more realistic, although much higher negative torque values could occur during rapid decelerations. Fig. 11 shows test cycle points of the ETC where negative torque values of the test cycle take on unnormalized values equal to negative 40 % of the positive torque available at the associated speed (density of dots represents the frequency of the engine operation in the corresponding area).

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

129

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 11. Test cycle points of the European Transient Cycle.

It was shown in Ref. [17,18] that the average torque of the ETC is relatively high. Furthermore, it was found out in Ref. [10,17,18] that the drive cycle has a significant influence on the optimum combination of the component sizes and on the fuel economy of the hybrid powertrain. The original ETC was therefore scaled to obtain test cycles with lower average torque in order to enable systematic comparison and analysis of powertrain parameters when operating according to the test cycles with different average loads and the same engine speed trace. Positive torque values of the ETC were multiplied by the torque factors (TF) 0.4, 0.5, 0.6 and 0.8 in order to determine new test cycles; these cycles were denoted by ETC_0.4, ETC_0.5, ETC_0.6 and ETC_0.8. This scaling might be considered as operation of the vehicles carrying different loads when real driving conditions are concerned. The dynamometer version of the ETC features relatively low negative torque values as indicated previously. Although the negative torque of the test cycle was determined by the method that assures the largest negative absolute torque values much higher negative torques can occur during the vehicle driving cycles, i.e. rapid deceleration. Therefore, an additional analysis where negative torque values were multiplied by a factor (TFneg) 1.5 and 2 (these cycles were denoted by ETC_N1.5 and ETC_N2) was performed in order to access knowledge of the influences of negative torque magnitudes, i.e. deceleration pattern, on the Δ f . This scaling might mimic more progressive deceleration patterns.

CONTROL STRATEGY HEV consists of many interacting components. Power control strategies are therefore necessary to regulate the power flow to and from different components. Control strategy (CS) has a significant influence on the performance and fuel economy of a vehicle [1,2,4]. Many papers concerning optimum CS for various types of hybrid powertrains have been published (e.g. Refs. [20,21,43,44]). However, it should be noted that minimum fuel consumption of particular hybrid powertrain running under particular operating conditions can only be assessed by global optimization, which can not be used for real-time control [20]. Therefore, the CS used in this study has to be flexible enough to provide equivalence at all hybridization levels. Relatively simple control strategies were applied in order to assure credible

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

130

Tomaž Katrašnik

comparisons of various hybrid powertrain configurations running under different operating conditions. In parallel hybrid powertrain the ICE was considered as the primary source of traction that was supported by the electric motor when needed, thus keeping the battery SOC above a certain minimum value. Batteries are recharged by regenerative braking or by working the ICE at higher torques to replenish the batteries. Control strategy parameters and their actions for parallel hybrid powertrain are summarized in Table 2. Control strategy algorithm executes particular control modes according to their order of appearance in Table 2. Table 2. Parameters of the control strategy for the parallel hybrid powertrain CS 1

Control parameters SOC>SOCmin,

Ptc > PICE ,h ,

FR = 1 2

SOC peff ,dch Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

6

Otherwise

PICE ,h = Ptc

CS 1 represents electrical assistance of the ICE if SOC>SOCmin, however batteries are always of such size that SOC is larger than SOCmin as stated in the section 3.2. According to CS 2 and CS 3 the ICE is operated at higher torque in order to replenish the batteries if they are discharged below the set limits, whereas SOCch2 p eff ,ch1 ; peff ,ch1 and p eff ,ch 2 are shown in Fig. 12. It is discernable from Fig. 12 that effective efficiency of the ICE drops significantly for lower p eff . Therefore, the ICE is operated at p eff ,ch1 when SOC p eff ,dch in order to prevent charging of the batteries above the specified limit, since charging efficiency drops for high SOC as discussed in section 3.2. CS 6 represents normal operation of the ICE. Brake mean effective pressure was applied as the parameter determining engine load, since it is a universal parameter for all engines and is also easily evaluated by the electronic control unit of the ICE.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

131

The following parameter values were applied in the CS: SOCmin=0.4, SOCch1=0.7, SOCch2=0.6, SOCdch=0.75 for the battery, SOCmin=0.25, SOCch1=0.8, SOCch2=0.9, SOCdch=0.91 for the ultracapacitor, and a = 0.75 , whereas p eff ,ch1 and peff ,ch 2 are shown in Fig. 12, and peff ,dch = p eff ,ch1 .

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 12. Parameters of the control strategy for the parallel hybrid powertrain.

Figure 13. Parameters of the OEOL and TH control strategy for the series hybrid powertrain.

A stop&start strategy was not considered in the CS, since a dynamometer cycle was applied, however it is obvious that it is possible to reduce fuel consumption of hybrid powertrains by the application of stop&start strategy when vehicle drive cycles are considered. Two control strategies were applied when analyzing the series hybrid powertrain: optimum engine operation line (Fig. 13-OEOL) [2,4] and “thermostat control” (Fig. 13-TH)

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

132

Tomaž Katrašnik

[44]. In the case of OEOL control strategy optimum operating points of the ICE constitute its optimum operating line, whereas in the case of TH control strategy ICE runs in the specified operating point and turns on and off upon the battery SOC. A compromise considering fuel consumption, exhaust pollutant emissions, power output as well as weight and cost penalty of the components should be made when selecting the operating point of the TH control strategy. In the presented analysis the OEOL control strategy was also based on the battery SOC, i.e. FR∝ SOC: FR is linearly varied between SOC=0.7 ⇒ ICE is turned off and SOC=0.6 ⇒ FR =max., whereas ICE speed is varied according to the FR position that is proportional to the engine mean effective pressure, as shown in Fig. 13. When OEOL control strategy is applied, the ICE is also turned off if SOC>0.7, since power output of the ICE and consequently of the EG would be low resulting in low energy conversion efficiency.

RESULTS The results of the parallel and of the series hybrid powertrains operating according to the ETC drive cycle and its variations are presented in this section. The results of different configurations of parallel and of series hybrid powertrains for different TFs, i.e. positive torque values of the ETC were multiplied by the torque factors (TF), are presented and analyzed in sections 4.1. and 4.2. respectively. Synthesis of the results from sections 4.1. and 4.2. is presented in section 4.3. Section 4.4 presents impact of the electric storage efficiency on the energy conversion efficiency of both hybrid powertrains, whereas section 4.5. gives analysis of the energy conversion efficiency for different negative torque factors (TFneg).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 3. Results of the ICE powertrain TF

0.4

0.5

0.6

0.8

1

m f ,tc [kg]

2.33

2.71

3.1

3.92

4.77

WICE [MJ]

26.8

35

43.2

59.6

75.8

η eff [-]

0.269

0.303

0.326

0.356

0.373

0.387

0.493

peff [MPa] 0.174 0.227 0.28

The results of the conventional, i.e. ICE, powertrain driving according to the analyzed cycles are presented in Table 3. It should be noted that powertrain parameters for ETC_N1.5 and ETC_N2 are equal to those for ETC when the conventional powertrain is concerned, since only negative torque values were modified as introduced in section 3.4. These results form the base for further analysis of both hybrid powertrains. Table 3 shows: m f ,tc , WICE ,

η eff and p eff . It is obvious that WICE and p eff increase linearly with increasing TF due to the scaling procedure of the test cycles introduced in the section 3.4. Contrary, it is discernable from Table 3 that difference between the values of the effective efficiency

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

133

decreases with increasing TF, since it is obvious from Fig. 7 that gradient of

η eff decreases

with increasing p eff , whereas m f ,tc is proportional to WICE

η eff as introduced in eq. (7).

Parallel hybrid powertrain In this section the results of parallel hybrid powertains with HF=0.101, 0.203, 0.304, 0.42 and 0.531, corresponding to RV =0.9, 0.8, 0.7, 0.6 and 0.5, are shown. First, the parallel hybrid powetrain parameters are analyzed as the functions of the HF and afterwards the parallel hybrid powetrain parameters are also briefly analyzed as the functions of the TF in order to interpret the results from the different point of view and to form the basis for comparison with the results of the series hybrid powertrain. In order to support subsequent analysis of the energy conversion phenomena in the parallel hybrid powertrain eq. (16) is rewritten to

m f ,tc , h m f ,tc

⎡ η eff Wbr , h − Wbr ⎤ ⎡ η eff (1 − η EL )W EG ,in ⎤ ⎤ ⎡ η eff −1= ⎢ − 1⎥ + ⎢ ⎥ ⎥ +⎢ W ICE W ICE ⎥⎦ 3 P ⎥⎦ 2 P ⎢⎣η eff , h ⎥⎦ 1P ⎢⎣η eff , h ⎢⎣η eff , h

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

(38) where 1 was subtracted from both sides of the equation. Hence, positive value of particular term indicates increase of the fuel consumption and negative value indicates reduction of the fuel consumption. lhs of the equation represents relative change of the fuel consumption of the hybrid powertrain compared to that of the conventional powertrain. The first term on the rhs of (38) ( rhs1P ) represent the ratio of the baseline ICE efficiency to the efficiency of the ICE of the hybrid powertrain. The second term on the rhs of (38) ( rhs 2 P ) accounts for the energy losses due to electric energy production, storage and consumption multiplied by the ratio of the ICE efficiencies. The third term on the rhs of (38) ( rhs 3 P ) represents the influences of the ICEs efficiencies, energy consumed by the brakes and work produced by the baseline ICE.

Analysis of the powetrain parameters for ETC This section presents the results of the parallel hybrid powertrain applying batteries as electric storage devices and operating according to the ETC. Figure 14 shows: a) relative change of the fuel consumption ( Δ f = m f ,tc ,h m f ,tc − 1 ), relative change of the effective work ( Δ eff = WICE , h WICE − 1 ) and relative change of the indicated

work

( Δ ind = WICE ,ind ,h WICE ,ind − 1 = nc

∫p c

functions of the HF; b) rhs1P = η eff

cyl , h

dV nc ∫ p cyl dV − 1 )

as

c

η eff ,h − 1 = Δη , rhs 2 P , rhs3 P , Δη = η f η f ,h − 1 eff

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

f

134

Tomaž Katrašnik

and Δη mech = η mech pressure

( p eff ),

η mech,h − 1 as functions of the HF; c) cycle averaged mean effective number

of

battery

(

modules

)

( nbat )

and

rhs

of

eq.

(19)

( rhs19 P = 1 + (1 − η EL )WEG ,in WICE + Wbr ,h − Wbr WICE ) as functions of the HF; and d) ratio of braking work to effective work of the baseline ICE ( Rbr = Wbr WICE , where Wbr represents braking work of conventional and hybrid powertrains), dimensionless difference between energies consumed by the brakes ( ΔRbr = − Wbr , h − Wbr WICE = Rbr (HF = 0 ) − Rbr ), ratio of the generator input work to

(

)

effective work of the baseline ICE ( R EG = WEG ,in WICE ) and electric conversion efficiency

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

( η EL ) as functions of the HF.

Figure 14. Parameters of the parallel hybrid powertrain for ETC.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

135

It can be clearly seen in Fig. 14 a) that fuel consumption of parallel hybrid powertrains ( Δ f ) decreases up to HF≈0.3. However, thereafter fuel consumption starts to rise and exceeds that of the conventional powertrain for HF>0.42. It is necessary to analyze particular terms of the eq. (38) in order to adequately interpret this phenomenon. Fig. 14 b) ( rhs1P ) indicates that effective efficiency of ICEs increases with increasing HF, i.e. decreasing RV , since smaller ICEs operate at higher mean effective pressure resulting in higher effective efficiency of the ICEs (Fig. 14 c) -

p eff

and Fig. 8). Increase

of the effective efficiency is due to increase of both, increase of fuel conversion efficiency and increase of mechanical efficiency (Fig. 14 b) - Δη f and Δ η mech ), since both efficiencies increase with increasing

p eff

as discernable from Figs. 5 and 6. This analysis therefore

confirms that downsizing of the ICE clearly contributes to improved energy conversion efficiency of the hybrid powertrains. It is discernable from Fig. 14 a) that Δ ind of the parallel hybrid powertrains exceeds Δ f due to higher fuel conversion efficiency of the downsized ICEs (Fig. 14 b)- Δη f ) and Δ eff further exceeds Δ ind due to higher mechanical efficiency of the downsized ICEs (Fig. 14 b)-

Δη mech ), whereas all differences increase with increasing HF as indicated by Δη f and Δη mech in Fig. 14 b); i.e. W ICE ,h WICE > m f ,tc ,h m f ,tc since η eff , h >η eff . Fig. 14 b) - rhs 2 P obviously indicates that electrical losses significantly deteriorate

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

energy conversion efficiency of hybrid powertrains particularly for high HFs. It is discernable from Fig. 14 b) that rhs 2 P increases with increasing HF, since the amount of the energy needed to support the ICE during the test cycle increases for higher HFs or lower RV s. This is shown by R EG = WEG ,in WICE (Fig. 14 d)) that includes energy gained by regenerative braking and energy produced by operation of the ICE at higher torque output. Electrical losses are proportional to (1 − η EL )WEG ,in and it can be concluded that rhs 2 P → 0 if

η EL → 1 . It is therefore very important to select electrical components with high efficiencies in order to develop hybrid powertrains with high energy conversion efficiency. This is essential particularly for hybrid powertrains with high HF. Fig. 14 b) - rhs 3 P indicates that regenerative braking represents a significant mechanism to improve the energy conversion efficiency of hybrid powertrains as was already indicated in the section 2. It can be observed that rhs 3 P decreases until HF≈0.3 and is approximately constant thereafter. The shape of the rhs 3 P curve could be explained with the shape of Rbr curve. The ratio of the braking work to the effective work of the baseline ICE (Fig. 14 d) Rbr ) decreases with increasing HF, since the amount of electric energy that could be produced by regenerative braking increases due to higher electric generator input power. Less energy is therefore consumed by the brakes and dissipated to heat. At HF≈0.3 all energy

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

136

Tomaž Katrašnik

originating from negative torque values of the test cycle is recuperated by regenerative braking and rhs 3 P is therefore approximately constant for higher HFs. This analysis indicates that energy conversion efficiency of the hybrid powertrain could be further increased by increasing negative torque values of the original ETC. These results are analyzed in section 4.5. A parameter rhs19 P = Δ eff + 1 = WICE ,h WICE (Fig. 14 c)) is introduced in order to analyze influences of electrical losses and of regenerative braking on the energy conversion efficiency of parallel hybrid powertrains. This parameter excludes influences of effective efficiencies of ICEs and therefore represents competing effects of increasing energy conversion efficiency due to regenerative braking and of decreasing energy conversion efficiency due to electrical losses. This parameter also equals to WICE ,h WICE and is thus the measure of the effective work that needs to be produced by the ICE of hybrid powertrains over the test cycle. Analyzing REG and ΔRbr (Fig. 14 d)) it can be concluded that

REG >> ΔRbr for HF>0.2 indicating that considerable amount of the WEG ,in is produced by operating the ICE at higher torque output. Rewriting eq. (16) to dimensionless form considering terms presented in Fig. 14 leads to (1 − η EL )REG − ΔRbr = WICE ,h WICE − 1 = Δ eff = rhs19 P − 1 . It is thus obvious that

WICE ,h > WICE for REG >> ΔRbr and η EL ≈ 0.5 that is observed for HF>0.2. Consequently, rhs19 P (Fig. 14 c)) shows that positive influences of regenerative braking prevail over

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

electric losses up to HF≈0.2, whereas inverse tendency is characteristic for HF>0.2. Lower fuel consumption of the parallel hybrid powertrain compared to the conventional one up to HF≈0.42 (Fig. 14 a) - Δ f ) could therefore be linked to higher effective efficiency of the ICE in the parallel hybrid powertrain (Fig. 14 b) - rhs1P = η eff

η eff ,h − 1 = Δη ), since eff

η eff η eff ,h × rhs19 P = m f ,tc ,h m f ,tc = Δ f + 1 . It can be seen from Fig. 14 c) that number of battery modules ( nbat ) coincides well with

REG (Fig. 14 d)) and rises nearly linearly with increasing HF, since electric consumption/production increases with increasing HF demanding adequate number of battery modules to maintain battery charge-discharge efficiency. Number of battery modules is discrete, therefore an appropriate number of battery modules in best agreement with

η ES = 0.65 was chosen. Fig. 14 d) shows that electric conversion efficiency (η EL ) that includes η EG , η EM and η ES is approximately 0.5. Analysis of the powetrain parameters for ETC_0.6 Cycle averaged mean effective pressure of the original ETC is relatively high for the baseline engine and even higher for ICEs in parallel hybrid powertrains as shown in Fig. 14 c)

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

137

Energy Conversion Efficiency in Hybrid-Electric Powertrains -

p eff . Original ETC was therefore scaled to obtain test cycles with lower average torque

in order to enable systematic comparison and analysis of the powertrain parameters when operating according to test cycles with different average loads and the same engine speed trace as indicated in section 3.4. This section presents the results of the parallel hybrid powertrain applying batteries as electric storage devices and operating according to the ETC_0.6. Figure 15 shows: a) relative change of the fuel consumption ( Δ f = m f ,tc ,h m f ,tc − 1 ), relative change of the effective work ( Δ eff = WICE , h WICE − 1 ) and relative change of the indicated

work

( Δ ind = WICE ,ind ,h WICE ,ind − 1 = nc

∫p

cyl , h

c

functions of the HF; b) rhs1P = η eff and Δη mech = η mech pressure

( p eff ),

dV nc ∫ p cyl dV − 1 )

as

c

η eff ,h − 1 = Δη , rhs 2 P , rhs3 P , Δη = η f η f ,h − 1 eff

f

η mech,h − 1 as functions of the HF; c) cycle averaged mean effective number

of

battery

(

modules

)

( nbat )

and

rhs

of

eq.

(19)

( rhs19 P = 1 + (1 − η EL )WEG ,in WICE + Wbr ,h − Wbr WICE ) as functions of the HF; and d) ratio of braking work to effective work of the baseline ICE ( Rbr = Wbr WICE , where Wbr represents braking work of conventional and hybrid powertrains), dimensionless difference between energies consumed by the brakes ( ΔRbr = − Wbr , h − Wbr WICE = Rbr (HF = 0 ) − Rbr ), ratio of the generator input work to

(

)

effective work of the baseline ICE ( R EG = WEG ,in WICE ) and electric conversion efficiency

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

( η EL ) as functions of the HF. Analyzing Δ f (Fig. 15 a)) it can be concluded that fuel consumption of the hybrid powertrains for the ETC_0.6 decreases with increasing HF, in contrast to the ETC (Fig. 14 a) - Δ f ). Again eq. (38) is applied in order to adequately interpret this phenomenon. Comparing rhs1P for the ETC_0.6 (Fig. 15 b)) and rhs1P for the ETC (Fig. 14 b)) it is discernable that improvement in effective efficiency due to downsizing of the ICE is similar for both TFs. It can also be concluded that fuel conversion efficiency and mechanical efficiency similarly affect improvement of the effective efficiency (Fig. 15 b) - Δη f and

Δη mech , and Fig. 14 b) - Δη f and Δη mech ), since similar relative increase in p eff

similarly

influences fuel conversion and mechanical efficiency and thus effective efficiency (Fig. 15 c), Fig. 14 c) and Fig. 8). Consequently, relative differences between Δ ind and Δ f as well as relative differences between Δ eff and Δ ind for the ETC_0.6 (Fig. 15 a)) feature similar trends as those for the ETC (Fig. 14 a)).

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

138

Tomaž Katrašnik

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 15. Parameters of the parallel hybrid powertrain for ETC_0.6.

Since the trends of rhs1P are similar, very distinct trends of Δ f curves for ETC_0.6 and ETC are therefore the consequence of rhs19 P . It is indeed observed that rhs19 P η eff . It can be seen from Fig. 15 c) that number of battery modules ( nbat ) coincides well with

REG (Fig. 15 d)) and increases up to HF≈0.3. Fig. 15 d) shows that electric conversion Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains efficiency (η EL ) that includes

139

η EG , η EM and η ES is approximately 0.5 for HF≤0.2 and

starts decreasing for larger HFs. This phenomenon is due to lower

η EG and η EM , since

maximum input/output power of electric machines increases with increasing HF resulting in lower relative power rate of electric energy production and consumption for test cycles with lower average torque and thus in lower EM and EG efficiencies.

Comparison of the results for relative change of the fuel consumption Fig. 16 summarizes results of Δ f for parallel hybrid powertrains applying batteries as

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

electric storage devices and operating according to ETC, ETC_0.8 and ETC_0.6. It can be concluded that: 1.) benefit in fuel consumption of parallel hybrid powertrains increases if average torque of the test cycle is decreased and 2.) minimum value of the fuel consumption for particular average torque of the test cycle moves towards higher HF when average torque of the test cycle is decreased.

Figure 16.

Δf

for different test cycles, [15].

Analysis of the parallel hybrid powertrain applying ultracapacitors for ETC It is discernable from the above analysis that electric storage efficiency significantly influences energy conversion efficiency of hybrid powertrain. It is therefore intuitive that ultracapacitors possess the potential to further improve the fuel economy of hybrid powertrains due to their higher charge-discharge efficiency. This section presents the results of the parallel hybrid powertrains applying ultracapacitors as electric storage devices operating according to the ETC.

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

140

Tomaž Katrašnik

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 17. Parameters of the parallel hybrid powertrain applying ultracapacitors for ETC.

Figure 17 shows: a) relative change of the fuel consumption ( Δ f = m f ,tc ,h m f ,tc − 1 ), relative change of the effective work ( Δ eff = WICE ,h WICE − 1 ) and relative change of indicated

work

( Δ ind = WICE ,ind ,h WICE ,ind − 1 = nc

∫p

cyl , h

c

functions of the HF; b) rhs1P = η eff and Δη mech = η mech pressure

( p eff ),

dV nc ∫ p cyl dV − 1 )

as

c

η eff ,h − 1 = Δη , rhs 2 P , rhs3 P , Δη = η f η f ,h − 1 eff

f

η mech ,h − 1 as functions of the HF; c) cycle averaged mean effective electric

conversion

(

efficiency

)

(η EL )

and

rhs

of

eq.

(19)

( rhs19 P = 1 + (1 − η EL )WEG ,in WICE + Wbr ,h − Wbr WICE ) as functions of the HF; and d) ratio of braking work to effective work of the baseline ICE ( Rbr = Wbr WICE , where Wbr represents braking work of conventional and hybrid powertrains), dimensionless difference between energies consumed by the brakes ( ΔRbr = − Wbr , h − Wbr WICE = Rbr (HF = 0 ) − Rbr ), ratio of the generator input work to

(

)

effective work of the baseline ICE ( R EG = WEG ,in WICE ) and number of ultracapacitor modules ( nUC ) as functions of the HF. Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Energy Conversion Efficiency in Hybrid-Electric Powertrains

141

Analyzing Δ f of the powertrains applying ultracapacitors (Fig. 17 a)) and Δ f of the powertrains applying batteries (Fig. 14 a)) it can be concluded that application of ultracapacitors significantly reduces fuel consumption of hybrid powertrains and that fuel cousumption reduction is more significant for high HFs. Equation (38) is applied for the analysis of the energy conversion efficiency. Application of ultracapacitors directly influences the rhs 2 P , whereas rhs1P and rhs 3 P could be influenced by the ratio

η eff η eff ,h . Comparing rhs1P of the powertrains applying ultracapacitors (Fig. 17 b)) and rhs1P of the powertrains applying batteries (Fig. 14 b)) it is discernable that improvement in effective efficiency due to downsizing of the ICE is slightly lower for the hybrid powertrains applying ultracapacitors. Smaller improvement in effective efficiency is due to smaller improvement in fuel conversion efficiency and in mechanical efficiency (Fig. 17 b) - Δη f and Δ η mech , and Fig. 14 b) - Δη f and Δ η mech ), since in slightly lower than

p eff

p eff

of the hybrid powertrain applying ultracapacitors

of the hybrid powertrain applying batteries, whereas difference

increases with increasing HF (Fig. 17 c), Fig. 14 c) and Fig. 8). Consequently, relative differences between Δ ind and Δ f as well as relative differences between Δ eff and Δ ind are also slightly smaller for hybrid powertrain applying ultracapacitors. Improved fuel economy of the powertrain applying ultracapacitors obviously arises from the rhs 2 P , since η EL of the powertrain applying ultracapacitors (Fig. 17 c)) is significantly

η EL of the powertrain applying batteries (Fig. 14 d)). For a particular HF, of the powertrain applying ultracapacitors equals to the WEM ,out of the powertrain

higher than

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

WEM ,out

applying batteries, therefore higher electrical losses, i.e. rhs 2 P = η eff

η EL results in lower WEG ,in (eq. (9)) and thus in lower

η eff ,h × (WEG ,in − WEM ,out ) WICE . This is discernable in

Fig. 17 d) and Fig. 14 d) indicating lower values of R EG for powertrains applying ultracapacitors. It is also obvious that the difference in rhs 2 P between both powertrains increases with increasing HF and therefore results in more significant reduction of the fuel consumption due to application of ultracapacitors. Rbr is equal for the powertrain applying ultracapacitors and for the powertrain applying batteries resulting in very similar rhs 3 P for all HFs. Analyzing ΔRbr and REG - ΔRbr of both powertrains (Fig. 17 d) and Fig. 14 d)) leads to the conclusion that the amount of the WEG ,in that is produced by regenerative braking is equal for both powertrains, whereas the amount of the WEG ,in produced by operating the ICE at higher torque output is lower for the powertrain applying ultracapacitors. Higher

p eff

η EL thus results in lower WICE ,h and lower

and therefore in better fuel economy of the powertrains applying ultracapacitors. This

Electric Vehicles: Technology, Research and Development, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

142

Tomaž Katrašnik

is also reflected in rhs19 P indicating that positive effects of regenerative braking prevail over the

electrical

(1 − η EL )REG − ΔRbr

losses

for

= WICE ,h WICE − 1 = Δ eff = rhs19 P

HF rhs1S > Δηeff . rhs1S > Δηeff , since rhs1S considers efficiencies of particular electric components, i.e.

η EG , η eet , EG → EM and η EM , M , that are smaller than unity, whereas

Δ f > rhs1S , since electric losses, i.e. rhs 2 S and rhs 3 S , prevail over the positive effects due to regenerative braking, i.e. rhs 4 S . Curves of Δ f and rhs1S have similar trend as Δ ηeff ones, and all of them decrease with decreasing TF, since the efficiency of the ICE in series hybrid powertrain remains nearly constant, whereas the efficiency of the baseline ICE decreases with decreasing TF (Table 3). It is discernable form Fig. 20 that the difference between Δ f and rhs1S curves of the OEOL and the TH control strategy is larger than the Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

difference between Δ ηeff curves for TF