12 IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS for leisurelyrides. used mostcommonlyforshorttripstogrocerystoresor poses. IntheUnitedStates,electricbicyclesarecurrently ranger vehicle,orforleisurelyridesandcommutingpur- problem, asaguidebicycleduringraces,park enforcement incitieswhereparkingandtrafficarea purposes, forinstance,asavehiclepoliceorlaw ered .Electricbicyclescanbeusedforavarietyof bicycles,” includingbothfullyandpartiallymotor-pow- tric ”isusedtodescribe“electric-motor-powered elecs” (pedalelectriccycle).Inthisarticle,theterm“elec- that speed.Suchbicyclesarecommonlyknownas“ped- even higherpercentageofhumanpowerisrequiredabove with 50%humanpedalpowerforupto12mi/h,andan electric-motor-powered bicyclesare required tooperate powered byamotor. InothercountriessuchasJapan, E United States,suchbicyclescanbefully U.S. marketforabouttwodecades.Inthe cles havebeenmakingtheirwayintothe LECTRIC-MOTOR-POWERED BICY- BICYCLES ELECTRIC BY ANNETTEMUETZE&YINGC.TAN A performance evaluation A performance 1077-2618/07/$25.00©2007 IEEE © EYEWIRE mentally obtained confirmed byexperi- identified. Theresultsare cal ridingsituationsarealso requirementsindifferenttypi- are supportedandtheoreticallyexpanded.The of professionalbicyclingasastartingpoint,thefindings ments ofelectricbicycles.Usingknowledgefromthefield in variouscountries,anddifferentperformancerequire- less commonlyconsideredtopics,suchasregulatoryissues electric bicycles(e.g.,[1]–[12]).Theoverviewincludes of thestatearttoday’s commerciallyavailable classification ofelectricbicyclesthatincludesanoverview First, thisarticleprovidesasystematic,comprehensive IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS 13 US$0.7/100 km US$4.4/100 km US$4.4/100 = = US$7.1 = US$0.12 100 mi · = for going by car, but only for going for going by . for going by electric 10 h • US$0.12/100 mi • US$7.1/100 • US$7.1/100 mi registration, and parking are registration, and parking needed. of on bicycle paths; avoidance traffic jams. longer-distance commuting 0.1 $/kWh = · gives a comparative overview of the perfor- Friendliness In regard to the United States, electric bicycles are not In regard to the United States, electric able 2 Sample cost calculation for a 100-mi trip on an electric bicycle: see Figure 4) • Power to travel at 10 mi/h: 120 W (experimentally obtained, • Duration: 100 mi/10 mi/h TABLE 1. ASPECTS FAVORING THE USE OF FAVORING TABLE 1. ASPECTS BICYCLES. ELECTRIC Costs Averaging, costs* are Lancer: *Sample cost calculation for a 100-mi trip on a 2002 Mitsubishi • Gas tank capacity: 28 mi/gal [13] • Approximate gas rate as of November 2004: US$2/gal • Costs for 100 mi: US$2/gal/(28 mi/gal) • Energy usage: 1.2 kWh • Madison Gas & Electric rate as of November 2004: US$0.1/kWh • Costs for 100 mi: 1.2 kWh Other Costs license, Generally, no insurance, Traffic Flowvehicle bicycles Most states allow electric Zero-emission Environmental Health Benefit Incorporation of exercise and mance ranges of today’s commercially available electric mance ranges of today’s bicycles according to the authors’ market research. It illustrates how widely the specifications of electric bicycles vary according to the bicycle design and the riding conditions for which the electric bicycle is designed. The influence of several factors and parame- ters on the different criteria and performance require- ments are discussed in the “Investigation of Technical Performance Requirements” section. Criteria for Classification of Electric Bicycles Criteria for classification of electric bicycles have been determined such that they are independent of the country and the purpose of use. These are the bicycle Aspects Favoring the Use of Electric Bicycles Aspects Favoring the Use of Electric Bicycles electric bicycles in A number of aspects favor the use of cost per energy different situations. These include lower when going by distance traveled (1–2% of going by car in other costs electric bicycle) for a single rider; savings parking, such as insurance, licenses, registration, improvement of the traffic flow; environmental friendli- 1). ness; and the health benefit for the rider (Table Performance Range of Commercially Available Electric Bicycles T law to be pedelec-type bicycles. Electric bicycles produced law to be pedelec-type bicycles. Electric are mostly exported to Europe. in Taiwan and most as popular as in the other countries mentioned states, the federal electric bicycles are imported. In some differ. law and the state law for electric bicycles 1 Drive Wheel [1]. ehicle V Motor Accessories Motor Pedals Gears Battery Control Battery Charger Figure 1). Electric Bikes Worldwide 2002 Electric Bikes Worldwide rely on the motor completely pedal and use the motor at the same time pedal only (use as a conventional bicycle). Input Rider ■ ■ ■ oday, China is the largest manufacturer of electric China is the largest oday, The rider of an E-bike can choose to In Europe, most electric bicycles are manufactured in In Japan, most electric bicycles are produced by the T Parallel hybrid schematic diagram [9]. Parallel hybrid schematic Electric bicycles have been gaining increasing attention worldwide, especially in China, Europe, Japan, Taiwan, and the United States. In the following, the most distin- guishing aspects of electric bicycles in these countries are summarized, based on the authors’ own studies and Frank Jamerson’s Overview of Electric Bicycles Worldwide The basic configuration of an electric bicycle drive con- The basic configuration of an electric flow from the sists of a controller that controls the power flow acts in This power battery to the electric motor. the rider via the parallel with the power delivered by pedal of the bike ( Basic Configuration of an Electric Bicycle System Basic Configuration of an Electric Bicycle Evaluation of the State of the Art data that have been collected in the context of real-life data that have been collected in the context key parameters, applications. From the results, the the perfor- needs, and challenges involved in improving The article then mance of electric bicycles are identified. that can serve as gives a summary of the different results This summary a roadmap for such improvements. and techni- includes both market trends and regulations paths of further cal-science-related aspects. Different are outlined in research to build on the presented work the conclusion. Germany and the Netherlands, and pedelec-type electric bicycles are more common. bicycles are required by and electric automotive industry, bicycles, exporting the majority of the electric bicycles while also meeting a strong local demand. According to Qualification Electric Bike General Technical China’s GB17761-1999 [9], Chinese electric bicycles may not exceed 20 km/h and may not be heavier than 40 kg. 14 IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS are shownforeachsubcategory are (Table 3 throttle type,motorplacement,andbatterytype kit type,motorassembly, assisttype, investigation. Electric bicycletestset-upusedfortheexperimental is lessthan20mi/h.” ered solelybysuchamotorwhileridden anoperatorwhoweighs170pounds, less than750W,whosemaximumspeedon apavedlevelsurface,whenpow- two- orthree-wheeledvehiclewithfullyoperable pedalsandanelectricmotorof **Sec. 2085ofthefederallaw[14]definesa “low-speedelectricbicycle”as“a US$800–US$2600 US$250–US$800 Electric bicyclekitand 4.6–22.8kg Electric bicyclekitonly Price range 10–50lbs upto6%slope Electric bicyclekit Weight 12–36V Hill climbingability 100–500Wh 2–6h Torque On-board powersupply 16–80km Power consumption Upto400 Power 10–50mi 32km/h Cycles ofcharge/discharge 19km/h Charging time Batteries 20mi/h 12mi/h Travel range Maximum speed** Average speed Speed AVAILABLE ELECTRICBICYCLES. TABLE 2.PERFORMANCERANGEOFCOMMERCIALLY Pack Battery electric bicycles) bicycle (Custombuilt bicycle weight excluding original (Each fullcharge) (Full charge) ). Theassetsanddrawbacksofthesecriteria Multimeter T Power ap Hub Tap Tap Power Anemometer in T CPU Controller ables 4–10 Brushed HubMotor (bicycle Throttle Thumb 2 at firstglance. paid, bothfull-andhalf-assisttypescanlookthesame the country’s regulation.Unlesscloseattentionis of theassisttypecanbesignificantlyinfluencedby largely ontherider’s personalpreference.Thedesign exist, boththeassistandthrottletypesdepend not beingused.Eventhoughtechnicalaspectsdo know, synchronousmotorsandinductionare turers ofelectricbicycles,but,asfarattheauthors brushed andbrushlessdcmotorsareusedbymanufac- aspects shouldbepointedout:Ingeneral,both placement, andbatterytypes).Inthesetables,several kit, motor, motorassembly, assist,throttle,motor rwak Drawbacks: •Installationneeded • Comparatively Drawback: Assets: • Suitable fordifferent • Little/noinstallation • Safetyfeatures AddOn •Mountingflexibility • Comparatively inexpensive • Goodappearance • High-endbicycles Assets: Custom Built OF DIFFERENTBICYCLEKITTYPES. TABLE 4.ASSETSANDDRAWBACKS atr ye•La cdTable10 Table9 •Leadacid Table7 Table8 Battery Type •Frontwheel Motor Placement Table6 •Thumbthrottle •Full-assist Throttle Type Assist Type •Gear Table4 Table5 Motor Assembly •Brusheddcmachine Motor Type*** •Custom built Bicycle KitType OF ELECTRICBICYCLES. TABLE 3.CRITERIAFORCLASSIFICATION able electricbicyclesand,thus,theyarenotdiscussedhere. induction motorsandsynchronousarerarelyusedincommerciallyavail- bicycle orelectrickitmanufacturers.Totheauthors’knowledge, ***At large,bothbrushedandbrushlessdcmotorsareusedbymostelectric high costs •Bicyclecanberecon- required Bicycle KitType • Others • NiMH • Rearwheel • Pushbutton • Twistthrottle • Half-assist • Friction • Hub • Brushlessdcmachine • Addon • Connections maynotbe bicycle verted toa robust bicycle types conventional Performance Evaluation of Electric Bicycles Even though the technical maturity of electric bicycles Criteria have been defined to evaluate the performance of has been, and is still, improving, still more work needs to electric bicycles. These are technical performance, practi- be done to make electric bicycles competitive with other cability, design, environmental friendliness, and cost and vehicles. This includes more research on the durability economics (Table 11). The subcategories of all criteria, and lifetime of such bicycles, the long charging time of with the exception of the technical performance and cost batteries, and the sparse availability of charging stations. and economics, are commented upon individually in Tables 12–14 (practicability, design, and environmental Investigation of Technical criteria). The technical performance characteristics such as Performance Requirements power, torque, and speed have been investigated both the- oretically and experimentally and are discussed in the Theoretical Background “Investigation of Technical Performance Requirements” The total power Ptotal required to drive the bicycle is section. Cost and economics are discussed in Table 1. given by the sum of the power to overcome the air

TABLE 5. ASSETS AND DRAWBACKS TABLE 7. ASSETS AND DRAWBACKS OF DIFFERENT MOTOR TYPES. OF DIFFERENT ASSIST TYPES. Motor Type Assist Type Brushed dc Motor Brushless dc Motor Full-Assist Half-Assist Asset: Assets: Choices of modes Choices of modes of of operation: operation: • Simple controller • Higher efficiency and brushed motor • Pedaling only • Motor assistance is only available when the • Reduced size when • Motor operation only user is pedaling compared with • Pedaling and motor brushed motor operation in parallel • Level of assistance is determined by the user Drawbacks: Drawback: input • Lower efficiency • More complex controller Asset: Asset: than brushless motor than brushed dc motor • Increased number of • Meets the law • Brushes increase the IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 WWW.IEEE.ORG/IAS choices of modes requirements in more motor size and can of operation countries than the increase the difficulty full-assist type of mounting the motor into the fork of the bicycle Drawback: Drawback: • Not legally allowed in • Rider always has to all countries pedal

TABLE 6. ASSETS AND DRAWBACKS OF DIFFERENT MOTOR ASSEMBLY TYPES. Motor Assembly Type Power Tap Gear Hub Friction Receiver Assets: Assets: Asset: • Provides desired • Motor • Light gear reduction integrated weight ratio in the wheel • Enables easier • Easy mounting torque sensing/ • Minimal adjustment/assists maintenance

Drawbacks: Drawbacks: Drawbacks: • Chain/belt may • Can be • Less get entangled heavy efficient due to friction loss • Chain/belt • Significant shift • Tires wear may need of the center out easily Power Tap Hub maintenance: of gravity lubrication/tension 3 15 Power Tap hub [15] used for the experimental investigation. 16 IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS remarks aresummarized the symbolsforparameters,theirunits,andsome and[8], where show therelationshipsasdiscussedin[6] ing conditions: W power toovercomethefriction P rwak rwak Drawback: Asset: •Needto •Throttle can Drawback: • Maybeless Drawback: •Inexpensive Asset: PushButton •Feelslike • Reducedrisk Asset: TwistThrottle Thumb Throttle OF DIFFERENTTHROTTLETYPES. TABLE 8.ASSETSANDDRAWBACKS drag • Heavy • Cost • • More complexcontrollerthan Drawback: RegenerativeBraking •Recoveredenergy increasesthe Assets: Assetsanddrawbacks depend ontype Others •Cost Drawback: Assets: • Heavy Drawback: •Light NiMH • Inexpensive Asset: Lead-Acid TABLE 10.ASSETSANDDRAWBACKSOFDIFFERENTBATTERY TYPES. ilson’s ye proa formore repetitively accidentally preference) types (personal than other push comfortable be button turned motorcycle moped/ of accidental The threecasesthatcanbedistinguishedaccordingto of thepowerisusedtoovercomeairdrag the rollingresistance faces, themajorityofpowerisusedtoovercome ■ ■ , thepowertoovercomeslope At speedslessthan3m/s( At speedsgreaterthan3m/s( Case 2 Case 1 P → → Bicycle Science friction P P P total drag Flat ground,highspeed: P hill drag = = = = ↑↑ 9 9 C P . . drag , 81 81 d Throttle Type odpromnebicycleperformance • Goodperformance 8 correspondtothefollowingrid- [8] P · hill 2 · · D + in Table G m , · = P · · A hill R v 0 · g c ≈ , + P ( · · P friction 6 mi/h)andatlevelsur- v 15. drag m v P ≈ g g friction , + 6 mi/h),themajority . > . Equations(1)–(4) v w P , control precise ) friction 2 P · hill v g . , andthe , Battery Type (1) (4) (2) (3) Test VehicleDescriptionandInstrumentation life applications: tally evaluateelectricbicycleperformanceduringreal- T Performance ofElectricBicycles Experimental EvaluationoftheTechnical with abrusheddcmotorinstalledinthefronthub, For theexperimentalinvestigation,anelectricbicycle rwak Drawback: • Comparativelycomplex • Frontwheelslides Drawback: odwih includingbicycles Assets: climbing Rear • Suitableforlowland • Goodweight •Bestforlightweight • Comparativelyeasy Assets: Front OF DIFFERENTMOTORPLACEMENTTYPES. TABLE 9.ASSETSANDDRAWBACKS wo typesofmeasurementsweredesignedtoexperimen- 2) The ridingprofilesintermsofpower 2) 1) The requirementsintermsofpower 1) r oedneos installation than rearwheelslides are moredangerous odrasregionsandpoor •Bettertractionforhill vehiclesingeneral, •Suitableformountainous good roads and hillyregionswith distribution installation ground speed mentally determined. ■ speed slope grade leisurely ride,groceryshopping,orcommuting. of differentriders,wherethebicycleisusedforashort with thepowerrequiredtoovercomeslope drag androllingresistanceissmallwhencompared On steephills,thepowerrequiredforovercomingair Case 3 → → → → v g Flat ground,lowspeed: Hilly ground,lowspeed: P P hill friction with respecttotheinfluenceofload Motor PlacementType G ↑↑ v , andheadwindspeed g ↑↑ , nonregenerative type are measuredduringridingintervals P hill , P ,> hill , = P ground conditions drag 0 , , P P friction hill > P P > P , torque v versus ground friction w P drag are experi- . . T , and m , IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS 17 versus total P as a function , small varia- g g v v (Figure 4) illus- (Figure 4) m versus ground speed versus ground (Figure 6). Design Criteria w v Practicability Criteria total as a function of load as a function P g v • no age limit • generally no license required • easy to operate • weather dependence • not winter/wet weather/rain friendly serviced by a conventional bicycle shop. are not well established today. crash tests on electric bicycles traveling at high speed are required. with other bicycles. added for safety purposes. work is needed to be competitive with other work is needed to vehicles. bicycles. lifetime of electric ing can often only be done at home. bicycle performance significantly. are needed. pared with four minutes for a gasoline-fueled vehicle. of wind speed of wind speed • Bicycle size is small. Parking is easy compared • be Honks, headlights, and disk breaks can • • Technical performance is improving, yet more Technical performance • needed on the durability/ More research is • Asset: not as explosive as fuel vehicles (accidents) Drawback: More tests on general road safety and • Long charging time; typically four hours com- • availability of charging stations; recharg- Sparse • • Significant component to increase the electric Significant component to increase the • batteries Lighter-weight, higher-energy-density • Assets: Drawbacks: Asset: Conventional bicycle parts can be Asset: Conventional bicycle parts can Drawback: After-sales service and maintenance In the following, the measurement results are discussed. the measurement In the following, total power of measurements for The series 3) total power power 3) total TABLE 13. ASPECTS OF THE DIFFERENT DESIGN CRITERIA. Ergonomic TABLE 12. ASPECTS OF THE DIFFERENT TABLE 12. ASPECTS PRACTICABILITY CRITERIA. Technical Maturity Safety Battery Charging Battery Operating Condition Service/Maintenance ground speed ground speed trates (3) and (4). For a given ground speed (4). For a given trates (3) and tions in load result in small variations in power tions in load result as a function as a function g g v v . The measurement infor- . The measurement versus ground speed versus ground speed (Figure 5) • Speed • Efficiency • Distance/charge • Battery charging • Operating condition • Service/maintenance • Safety • Battery • Noise G total total Figure 3) [15] P P (Figure 4) m ). This bicycle is a commercially available bicy- is a commercially ). This bicycle of load of slope grade Economics • Other costs Performance • Torque 2) total power 2) total 1) total power 1) total Practicabilitymaturity • Technical Table 7 DesignEnvironment • Ergonomics • PollutionCost and • Unit price Table 13 Table 14 Table 1 Technical • Power TABLE 11. CRITERIA FOR PERFORMANCE EVALUATION OF ELECTRIC BICYCLES. Technical • Text Experimental Investigation of Power Requirement Experimental Investigation of Power Requirement Wind as a Function of Load, Speed, and Head rode the test For these experiments, four different riders different rid- bicycle without using the hub motor under conducted for ing conditions. The experiments were is typical for city speeds up to 12 mi/h (19 km/h), which to be constant. rides. The air density is approximated rolling and Furthermore, based on the theoretical results, constant and drag coefficients are assumed to be almost measurement are not investigated in detail. For each conducted and the point, five to ten measurements were of the the deviation average value was taken. Usually, was in the order individual measurements for one point were car- of less than 20%. Three series of measurements ried out: controller, thumb throttle, and battery pack are used and battery thumb throttle, controller, (Figure 2 experi- All the laboratory. been available in cle that has The this test vehicle. carried out using ments were during wheel is not used motor in the front electric hub set- this bicycle, the actual yet, using the measurements, is represented. For all measure- up of an electric bicycle was kept at 50–60 psi, which is ments, the tire pressure that are used for leisure and com- typical for bicycles commonly not reinflated before muting and that are and speed are directly measured in each ride. The torque wheel of the test bicycle, using a the hub of the rear power tap hub ( mation is transmitted to the Power Tap central process- to the Power Tap mation is transmitted Furthermore, the the receiver. ing unit (CPU) through speed as seen while riding is mea- relative head-wind speed The head-wind anemometer. sured by means of an the difference of anemometer and is then obtained from power tap speed. 18 IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS power difference islessthanafewkilograms. gy neededtopropelthebikeifload not asignificantlylarger amountofener- approximately 10–15W. Thus,thereis systems requiresadditionalpowerof ure 5),visualizesthecorrelationof(3).(Notethatsim- a functionoftheslopegrade and comparing thecurvesfor power isrequired,asillustratedby tion). Fordoubledload,twicethe required powerfor15-kgloadvaria- requirement (20Wdifferenceof and the curvesobtainedfor accounts forthenonlinearincreasefrom ed toovercometheairdrag(2)and area heavier rideralsohasalarger effective G slopes takenasaverage. curve; nowind,weightofrider and bicycle Influence oftheslopepath onthepowerversusspeed on thepowerversusspeedcurve;nowind,constantslope Influence oftheweightriderandbicycleinfluence ≈ Power (W) Power (W) The seriesofmeasurements,total An additionof10kgtothebicycle ( 100 150 200 250 300 350 ( A 1%. 154 100 150 200 250 300 50 154 50 0 which increasesthepowerneed- P 0 5678 total 567891011 + + (Rider 20 versus groundspeed 20 ) ) kg load. kg load.Generally, a + 61 64 75 154 Bicycle) + + + 20 kg 20 kg 20 kg + 20 kg ( Speed (mi/h) ( 64 64 Speed (mi/h) + + G 20 20 (Fig- ) v ) g kg kg .%1.0% 3.9% 2.4% 4.0% as 9 m = 011 10 FOR ABOUTTWO 81 kg,all THE U.S.MARKET THEIR WAY INTO BICYCLES HAVE BEEN MAKING DECADES. POWERED ELECTRIC- MOTOR- 5 4 ground speed u h eut fFgr havebeenselectedastheyare but theresultsofFigure6 been obtainedwithridersotherthantheoneofFigure6, are inlinewith(2).(Notethatsimilarresultshavealso provisions. city ridescanhaveratedpowerbelowthefederallaw a shortperiodoftime.Therefore,motorsdesignedfor hilly terrain,cityridesusuallyneedhightorquesonlyfor speed is8%.Itimportanttonotethatunlessridingon head windspeedstakenasaverage. curve; nowind,weightofrider and bicycle Influence oftheheadwindon the powerversusspeed Noise Pollution ENVIRONMENTAL CRITERIA. TABLE 14.ASPECTSOFTHEDIFFERENT The measurementresultsfortotalpower

Power (W) 100 • 55–60 dBcomparedwithfuel/gasvehiclelevels • No gasemission 10 20 30 40 50 60 70 80 90 0 of 65–70dB 3 4 v g as afunctionofheadwind Environmental Criteria given groundspeed illustrating thisanalysis.)Fora most completesetofmeasurements have beenselectedastheyarethe Figure 5,buttheresultsof5 with otherridersthantheoneof ilar resultshavealsobeenobtained constant, but neglecting the tional totheslopegrade linearly withtheslopegrade reasonable slopeof4%at10mi/h. torque) arerequiredtoclimbupa rider andbicycle,320W(47Nm approximately an80-kgweightof cles canclimbat10mi/hground result, thesteepestslopeelectricbicy- capability at10mi/his110Nm.Asa law of750W, themaximumtorque maximum powerallowedbyfederal 5 W ith electricbicyclesratedatthe 6 Speed (mi/h) 7 P hill P 89 drag is directlypropor- , m P total v = v g w P , 81 kg,all total (Figure 6) G 011 10 P increases 6 mi/h 5 mi/h 3 mi/h 1 mi/h friction . Thus, G versus . For 6 is IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS 19 , 7 8 max T / ave T , torques 15 20 25 15 20 25 max P / are summarized in Table 16. in Table are summarized me (min) ave ime (min) P Ti T max v / ave v Rider 12 Rider Rider 3 Rider 4 [mi/h][km/h] 9.2 20.9 14.8 33.6 18.3 29.4 24.2 39.0 0510 0510 [mi/h][km/h] 5.4 8.7 12.7 20.4 7.6 12.2 13.0 20.9 [W] 204.0 389.1 368.6 857.0 [Nm] 27.9 40.8 26.4 50.2 [W] 35.6 133.9 66.3 179.0 0 0 [Nm] 4.7 8.2 5.9 9.9

av max av max 900 800 700 600 500 400 300 200 100

900 800 700 600 500 400 300 200 100 [kg] [min] , , , , Power (W) Power isconsin,average and maximum 16–26 min. The for (W) Power ave max g g g g Above the speed limit for low-speed electric bicycles according to U.S. federal Above the speed limit for low-speed electric bicycles according ave max TABLE 16. RESULTS OF INTERVAL RIDING ANALYSIS. RESULTS OF INTERVAL TABLE 16. Rider weight 50 75 85 95 law (Table 2). Interval time 18Energy [Wh] 11.9 16 35.7 22 24.31 77.6 25 P v v P T v v T Power versus time profile of Rider 4. Power versus time profile of Rider 1 (same scales as Figure 8 by intention). W requirements total power For illustration, Figures 7 and 8 Figures 7 and For illustration, the power versus show and ground speeds and ground [6]. [3] [6], 2 for a 2 1 m 5 m . 4 = should be 0 . 5 , upright . 0 d , recum- 0 = G 3 77 C . . = d = 0 0 A C A = = d d cyclist [3]. cyclist: effects. For exam- ple, compacted gravel and smooth asphalt paths have different rolling coefficient of 0.004 [3] and 0.014 [6], respectively. bent bicyclist: namic bodies. Typi- cal values are: Passenger car: and area of the mass encountered by the air. Typical values are and rise/run. For steep grades, expressed by arc- tan(rise/run). C C The frontal area is the 3 , the power requirement , the power 2 w v 2 . Yet, at relatively low ground at relatively . Yet, A coefficient friction depends on speed speed coefficient small for aerody- acceleration Density of air kg/m Rolling – The rolling coefficient Ground m/s Head wind m/s Slope grade – The slope grade is Drag – The drag coefficient is Frontal area m Weight**** kg It is important to note that head wind does not seem to It is important to note d c g w 9.81 Gravity m/s **** Rider and bicycle, including accessories TABLE 15. SYMBOL AND PARAMETER DEFINITION. Symbol Parameter Unit Remarks D v v G A C m R Experimental Riding Interval Analysis Experimental Riding of measurements, four different For the second group bicycle around the city of Madison, riders rode the test the most complete set of measurements illustrating this illustrating complete set of measurements the most increasing With analysis.) increases. However, due to the very stochastic nature of very stochastic nature due to the However, increases. of the a rough idea experiment only provides wind, this position or an upright crouching trend. Furthermore, frontal area affects the speed this significantly affects the power requirement, speed this significantly Future work to obtain accurate notably with flat ground. using wind tunnels. results can be done by for city-ride electric bicycles, given be a major criterion city rides. the usual profiles of 20 IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS 17–22. Summarizing,morepublicityisstillneededto pared andcommenteduponindividuallyinTables components andonresearchdevelopment)arecom- tions, special-purposedesign,commentsonoversized categories ofthedifferentareas(markettrends,regula- is needed(Table 5).Inasimilarwayasbefore,thesub- sizing, andidentificationofareaswherefurtherresearch of possiblyoversizedcomponentsandreductionover- cial-purpose-design toattractcustomers,identification and regulations,opportunitiesforimprovementbyspe- of differentkeyparameters.Theseincludemarkettrends bicycle performanceevaluationissummarizedinterms Drawing fromthepreviousdiscussions,electric Requirements Summary ofPerformance common withcellphones. bicycle canberecharged inasimilar wayasistoday could beeasilyputonandtakenoffthebicycle met byalaptop-sizebattery. Suchanenergy source energy requirementoftheridesRiders1–3couldbe with assuminganefficiencyofthedrive50%, energy becauseofthelowerweightrider. Even the rideofRider2,butconsumeslessthan50% 1 isshorterthantherideofRider3andaboutaslong of anaverageride.ItisnoticeablethattherideRider power intheorderof100W, reflects therequirements 400 W, anaveragetorqueof6–8Nm, andanaverage along withamaximumpowerofsomewhatlessthan letic figureofRider4,amaximumtorque30Nm, (Table 2). speed electricbicyclesaccordingtoU.S.lawof20mi/h speed oftherideRider4exceedslimitforlow- by intention).Itshouldbenotedthatthemaximum requirements thatarecovered(samescalesonbothfigures two ridesrepresenttheextremesintermsofpower time profilesofthetworidesRiders1and4.These P eerhad atr Table 22 Table 21 •Battery Research and •Motor Reduction of pca-ups •Ct iyl Table20 Table 19 Table18 •Citybicycle Special-Purpose •On-roadlaw Regulations •Demand Market Trend PERFORMANCE EVALUATION. TABLE 17.SUMMARY:ELECTRICBICYCLE ave iyls•Hillbicycle Bicycles The fourridingprofilescoverabroadspectrumof eeomn •Technical Development •Battery Oversizing Possible , P max , T ave , T max • Speedybicycle • Distancebicycle • Bicycleassembly • Publicity • Regenerativebraking , v ave , and v max . Neglectingtheath- and technicalway. of existingdrivesinasystematical,comprehensive, mance. Furthermore,theycanbeusedforcomparison that willresultinimprovementofthesystemperfor- new drivesystemsaredesignedaroundkeyparameters platform toimproveelectricbicycleperformanceif bicycles. bicycles, hilldistanceandspeedy cient overagivenoperatingcycle.Theseincludecity addressed bycustom-designeddrivesthataremosteffi- The issuesassociatedwithelectricbicyclesmaybe Conclusions and theirusewithelectricbicycles. research bothonthebatteryanddrivetechnology electric bicyclemarketwouldbenefitfromfurther weight ofoversizedcomponents.Inthiscontext,the cles” wouldhelptore-ducetheadditionalcostand cycle, suchascity, hill,anddistance,“speedybicy- bicycles thataremostefficientoveragivenoperating tric bicyclesbeingcompromised.Custom-designed an increaseinpopularityandavoidthequalityofelec- designers/manufacturers ofelectricbicycleswouldfavor from licensing.Auniformstandard/guidelinefor attention needstobepaidreleasingelectricbicycles introduce thepublictoelectricbicycles.Also,more Bicycle Assembly On-Road Law TABLE 19.COMMENTSONREGULATIONS. Publicity Demand TABLE 18.COMMENTSONMARKETTRENDS. The resultsofthestudieslistedherecanserveasa compromised. the qualityofelectricbicycles being manufacturers ofelectricbicycles wouldavoid A uniformstandard/guideline fordesigners/ favor anincreaseinpopularity. Releasing electricbicyclesfromlicensingwould particularly regardingthelicensingaspect. U.S. stateshavedifferentlawsforelectricbicycles, to electricbicycles. More publicityisneededtointroducethepublic years. compared withthesameperiodinprevious bicycles salesinShanghaiincreased99.14%, As aresult,inthefirsthalfof2000,saleselectric vehicles. banned thesaleandoperationoffuel-assisted example, inBeijing,Tianjin,Guangzhoumunicipals increase ifnongreenvehiclesarebanned.For The marketdemandforelectricbicyclesmight Market Trend Regulations IEEE INDUSTRY APPLICATIONS MAGAZINE • JULY|AUG 2007 • WWW.IEEE.ORG/IAS 21 Possibly Oversized Components Possibly Oversized improved battery technology could improve the performance of electric bicycles. influence of battery density and charging time on electric bicycles is needed. efficient over the operating cycle. and drawbacks of different motor types and controllers. Areas of Further Research and Development • Further investigation is needed to examine how • Further investigation on the importance and • The motor should be designed to be most • Further investigation is needed on the assets Regenerative braking will be more useful in hilly areas or when braking is used often, as in city rides. Future work needs to identify the percentage recoverable energy, the impact on efficiency, cost, and the reduction of dependence on battery technology. Maximum power according to federal law: 750 W Maximum power according to federal at 20 mi/h speed required This is much more power than is normally with electric bicycles. do Electric bikes in the current market generally 4 and 5, not exceed 400 W. Drawing from Figures design. this value is a good guideline for general In a similar way as with the motor, careful the heavy selection of the battery could reduce from battery weight. For example, drawing be Table 16, a laptop-size battery that could would be easily put on and taken off the bicycle Then, the sufficient for short rides up to 30 min. in a similar bicycle recharging would be handled way as is today common with cell phones. uscode/#SECTIONS www.cycle-ops.com TABLE 22. COMMENTS ON RESEARCH AND DEVELOPMENT. Battery Drive Regenerative Braking TABLE 21. COMMENTS OF OVERSIZING RISK. TABLE 21. COMMENTS Motor Battery Annette Muetzewith the School is ([email protected]) C. Tan and Ying is of Warwick of Engineering at the University Computer Engineering Department at the with the Electrical and first appeared as This article University of Wisconsin-Madison. of Electric Bicycles” at the 2005 IAS “Performance Evaluation Annual Meeting. [14] States Code United Available: http://www4.law.cornell.edu/ [Online]. [15] WI Products, Inc., Madison, Power Tap, Graber Available: [Online]. , . , Graz, , vol. 1, Proc. 15th Proc. Northcon’94 Bicycling Science , October 13–18, , Bruges, Paper 237, 2002. , Bruges, Paper 237, 2002. , Nov. 7–9, 1995, pp. 555–560. Electric bikes worldwide 2002: With electric scooters & worldwide 2002: With electric Electric bikes Special-Purpose Bicycles ,” Electric Battery Bicycle Co, Naples, FL, 2002. Bicycle Co, Naples, ,” Electric Battery Proc. WESCON’95 Proc. NAECON ’93—National Aerospace and Electronics Conf. ’93—National Aerospace Proc. NAECON Fast acceleration, frequent stops Average power 150 W Average speed 17.6 km/h (11mi/h) High torque capability Fast acceleration capability High-speed capability 29km/h (18mi/h) Average power 200 W Maximum power 300 W at 12.8 km/h (8 mi/h) for a short time corresponding to a (4% slope grade) Designed for traveling at constant and comparatively low speed, but for a longer distance For example average speed 16 km/h (10 mi/h) at average power 100 W For example: guide bicycle in competitions, vehicle for law enforcers in May pp. 428–434. 24–28, 1993, in electrically assisted bicycle,” of an improved Roovers, “Development Applications Soc. Ann. Meeting Proc. 2002 IEEE Industry 2002, pp. 384–389. as a direct drive in an electric bicycle,” in soft magnetic composites Electronics and Applications (EPE) Proc. 9th European Conf. Power (ICEM) Int. Conf. Electrical Machines Oct. 1994, pp. 269–274. 11–13, sion power,” in Cambridge, Mass: MIT Press, 2004. able: http://gltrs.grc.nasa.gov/reports/2001/TM-2001-210972.pdf uscode/#SECTIONS [Online]. Available: http://www.wxetc.gov.cn/article.php/127 Economy Guide [Online]. Available: http://www.fueleconomy.gov/feg/ noframes/17596.shtml neighborhood EVs 2001, Paper 350. soft magnetic composites,” in brushless-dc motors using TABLE 20. SPECIAL-PURPOSE DESIGN OF TABLE 20. SPECIAL-PURPOSE DESIGN OF ELECTRIC BICYCLES. Hill Bicycle Distance Bicycle Speedy Bicycle City Bicycle [3] E.A. A.J.A. Lomonova, J. Vandenput, Rubacek, B. and G. d’Herripon, [4] A. Muetze, A.G. Jack, and B.C.“Brushless-dc motor using Mecrow, [6] W.C. bicycles,” in Morchin, “Battery-powered electric [7] H. Oman, W.C. Morchin, and F.E. Jamerson, “Electric-bicycle propul- [8] D.G. Wilson, J. Papadopoulos, and F.R. Whitt, [9] E-Bike SX [Online]. Avail- NASA, Baseline Testing of the EV Global [10]Code United States [Online]. Available: http://www4.law.cornell.edu/ [11] Extra energy [Online]. Available: http://extraenergy.org/main.php [12] China’s Electric Bike General Technical Qualification [in Chinese] [13] 2002 Mitsubishi Lancer mpg obtained from the United States Fuel [2] B. Kumar and H. battery-electric bicycles,” “Power control for Oman, References [1] F.E. Jamerson, “ [5] A. Muetze, A.G. Jack, and B.C. designs of Mecrow, “Alternate