AN EXTENDED-RANGE HYBRID POWERED SCOOTER

Transcription

1 AN EXTENDED-RANGE HYBRID POWERED SCOOTER Chung-Hsing Chao 1, Jenn-Jong Shieh Department of Electical Engineering Ta-Hwa Institute of Technology Hsinchu, 3743, Taiwan ROC Keywors: hybri power systems, tery, PEM fuel cell, scooter. Abstract As the electric scooter market grows, one of the major problems facing esigners of pure Es is getting sufficient energy from teries that can t cope with long journeys. In such pure electric scooters, electric-only riving range is restricte to just a few miles. By aing a small-powere fuel cell stack to rive a hybri power output or charging the tery when ile or low power perios, an extene-range hybri powere scooter allows its riving istance to be increase further. The alternative with this solution, though, is cost. This paper shows the moelling, simulation an experiments of an extene-range hybri powere scooter. The experimental ata are also compare with the pure tery powere scooter. It is foun that 134 litres of hyrogen fuel can supply the extene-range hybri powere scooter unceasingly running by increasing its range from 6.5 miles to 78.7 miles. Not only can the riving istance be increase further, but also the performance can be more brilliant than that of a pure tery powere scooter. 1 Introuction Electric vehicles (ES) now employ state-of-the-art technology. Engineers are currently integrating electrochemical engineering, mechatronics engineering, an electronic engineering to evelop future vehicles which aress global concerns about the green-house effect an the eventual shortage of petroleum sources. In recent years, esigners of pure ES have foun their own energy management strategies within a wie range of more than 4 kilometres an effective spees by integrating ifferent energy sources. In aition to the primary source of power, the teries, hybri-powere vehicles integrate fuel cells an ultra-capacitors with the merits of high power output, energy ensity, uick recharge capability, long life cycles, an high efficiency [1]. In many countries in Asia, scooters have become the primary vehicles use for transportation. Due to air pollution an other environmental problems that have arisen from scooters with gasoline-powere internal-combustion engines, attempts have been mae to use electric scooters as a zero emission alternative [1, ]. Despite recent evelopments, the rawbacks of limite range, long recharge time, an limite infrastructure still hiner the large scale eployment of electric scooters. Lin [3] propose a conceptual esign an moelling of a pure fuel cell powere scooter, integrating the fuel cell, hyrogen storage, an electric motor/controller, to evelop an overall vehicle esign. Angkee an Nartnarong [4] showe an experimental assessment of fuel cell hybri propulsion system for scooters, base on a moular 1. kw proton exchange membrane fuel cell (PEMFC). Khateeb et al. [5] stuie the thermal management of Li-ion teries for electric scooters an conucte experimental valiation. Sheu [6] escribe a etaile proceure of simulation for analysis of a hybri electric scooter powertrain. Fuel-cell-powere scooters an electric tery-powere scooters are often propose as alternatives, but scooters euippe with fuel cells woul be superior to tery-powere scooters by offering zero emissions because tery-powere scooters can t cope with long journeys. Hybri-powere scooters utilizing fuel cells offer the avantages of extene range an uick refuelling. Hyrogen storage in the form of metal hyries an a PEMFC running at low temperatures are chosen for the reasons of ease of manufacture an operation, low cost, an minimal volume of store hyrogen gas. This paper examines feasibility of fuel cell as a rangeextener for a hybri powere electric scooter as shown in Fig. 1, incluing a hybri power system esign combining a small size of PEMFC with sufficient teries for the most output, which together can hanle all of the transient an average loas. IN+ Fuel Cell 3W 36~53 DC/DC Converter DC+ OUT-/BATT- DC- Rs R Q1 Q Q11 Battery Charger an Battery Discharger Controller OUT BATT+ Battery_ 48 1AH IN- DC Brushless Motor 75W 48 Figure 1: Hybri powere system configurations of an electric scooter with a fuel cell range-extener Moelling an numerical simulation.1 Scooter Moel The scooter is moelle using the ynamics euations, where the total power reuire to rive the scooter is calculate base on aeroynamic rag, rolling resistance,

2 acceleration an power reuirements relating to variable roa graes [8]: 1 P [ ACv mg r cos mg sin ma]( rwheel ) (1) In above euation, the parameters of v, a, an are the velocity(ms -1 ), acceleration (ms - ), an wheel angular velocity (ras -1 ) respectively. The electric motor rive moel is base on experimental ata provie by the manufacturer. The ata is use to construct a table that etermines the motor torue, power an the root-mean-suare current being rawn given a known motor spee.. Fuel Cell Moel A PEMFC consists of three major components: an anoe electroe; a soli polymer electrolyte; an a cathoe electroe. In a PEMFC, hyrogen fuel gas is fe continuously into the anoe, an an oxiant from air is fe continuously into the cathoe compartment. The electrochemical reactions take place at the electroes to prouce an electric current. A number of approaches have been use to moel PEMFC behaviour [9, 1]. The Nernst euation for the open-circuit voltage of the PEMFC can be written in terms of by the values for the changes in the stanar state of entropy [8]: 1/ g f RT ah a O 4 E ln 1.9 (8.5 1 ) ( T 98.15) F F aho 5 1 (4.381 T) (ln ph ln p ) O () where g f is the change in molar Gibbs free energy of formation at stanar temperature (73.15K) an pressure (1 bar), T is the cell temperature (K), R is the gas constant (8.314Jmol-1K-1), p the partial pressure of hyrogen at H the anoe catalyst-gas interface (bar), p O the partial pressure of oxygen at the cathoe catalyst-gas interface (bar), an a 1 the activity of liui water prouce in the H O electrochemical reaction. Owing to using air as an oxiant ( p. O 1 ), the effect of this reuction in partial pressure gives the change in OC cause by a change in the reaction shoul be consiere. It is useful to construct an euation that brings together all these voltage losses. The combine effect of thermoynamics, kinetics, internal resistance, an mass transport etermines the output voltage of a fuel cell as efine by [8]. E act ohmic trans (3) In the present stuy, the experimental ata can be correlate with the following mathematical euation i il E aln( ) ri mexp( ni) i (4) The first term of E of 1. is the reversible stack voltage at about 5 C (35K). The secon term of a can be recognize as the activation overvoltage, which is preominate at low current ensities. The constant a is higher for an electrochemical reaction that is slower. The constant i is higher if the reaction is faster. As a result, at the zero exchange current ensity, we may suppose that there is no activity at the electroes an an electrochemical reaction oes not take place. For the case of a PEMFC using air at normal pressure, the experimental values for the constant a of.5( ), the internal current il of 3( macm ) ue to fuel leak or cross over, an exchange current ensity i of.4( macm ) are reasonable. The thir term of r escribes a linear rop-off, which is the preominant loss in the intermeiate current ensity region, where the internal resistances mainly cause by the membrane of the soli electrolyte an also contact losses where bipolar plates come into contact. The last term becomes preominant in the high current ensity region, an is use to match the rop-off towars the limiting current. A physical interpretation for the parameters m correlates to the electrolyte conuctivity an n to the porosity of the gas iffusion layer. The fuel cell system consists of a stack of N cells connecte in series. Therefore, the total fuel cell voltage is given by stack N (5). The above mathematical moel as escribe in euation (5) inicates that the current rawn, cell temperature, the activation overvoltage, internal resistance, limiting current, an oxygen pressure will affect the fuel cell output voltage. The valiation of the present moel is shown in Fig. an Tab. 1 by comparing the polarization curves i- an i-p of the fuel cell stack, which shows conformity between the experimental an moelle results. It is seen that the experimental ata almost falls into the moelle curves by the values, r = , m =3 1-5, an n =8 1-3, respectively. The maximum erivation between the moelle an experimental results was less than 5%. Figure : Comparison of fuel cell polarization curves between the experimental an preicte results (cell temperature: 5 C, RH% of H/O = 1/1, fuel/oxygen stoichiometric ratio: H=1.5, Air=.5) Table 1: experimental ata of the 3W PEM fuel cell stack Parameters Symbol alue Unit

3 Number of cells Rate performance N P T 7 none 3W 5~3 C Air 5 (35) wattage Air cooling with integrate C cooling fans Fuel cell operating temperature T C (K) Stack weight with fans an w kg cooling units Stack imensions L,W,H 34/1 mm 9/94 Cross section area of the A cm electrolyte Energy Efficiency η 4% none Pt loaing.4 mg/cm Distance between electroes.6 cm Symmetric factor at the anoe.5 none Symmetric factor at the cathoe.5 none Anoe exchange current ensity Cathoe exchange current ensity i.5 macm - i.4 macm -.3 Battery Moel Khateeb et al [11] gives a Li-ion tery moel base on experimental ata. Electrochemical characteristics of the tery are moelle in a similar, simplifie circuit iagram, incluing the internal resistance of the tery. Also, here the voltage-current relationship is non-linear an for the tery it is epenent on the State Of Charge ( SOC ) of the tery. State Of Charge ( SOC ) is the euivalent of a fuel gauge of the tery pack in a tery electric vehicle (BE), hybri electric vehicle (HE) or plug-in hybri electric vehicle (PHE). The units of are percentage points (% = empty; 1% = full). Depth of ischarge (DOD) is an alternate metho to inicate a tery's state of charge ( SOC ). The DOD is the inverse of :as one increases, the other ecreases. While the units are percentages (% = empty; 1% = full), the units for DOD can be Ah (e.g.: = full, 1 Ah = empty) or percentages (1% = empty; % = full). As a tery may actually have higher capacity than its nominal rating, it is possible for the DOD value to excee the full value (e.g.: 11 Ah or 11%), something that is not possible when using a SOC. The open circuit voltage across the tery terminals is moelle as follows. 3 ( SOC) ( (1 5.49(1 (6) it SOC( t) 1 C (7) Where C is the tery capacity (Ah); an i is the tery current (A). The tery voltage of an is calculate by subtracting the voltage rop ue to internal resistance from the open circuit voltage of ( SOC ). ( SOC) ir( SOC) (8) The internal resistance across the tery terminals is moelle as follows. 3 R ( SOC).8 8.4( (1 11.6(1 (9) The tery efficiency in ischarge moe is etermine base on the open circuit voltage shown as follows. (1) ( SOC ) Like all chemical processes, the tery capacity C is temperature epenent, but for the temperature variance is assume to be omitte for simplification in the present stuy. In orer to reflect the fact that only a fraction of the input energy is really store, the average approximation of 9% is use here. The experimental results in the present stuies were fitte to a curve an the Open Circuit oltage (OC) vs. Depth of ischarge (DOD) across the tery terminals was plotte as shown in Fig. 3. Figure 3: Battery OC vs. the epth of ischarge (DOD).4 Electrical Machine Moel The electrical machines can be moelle in various ways. In the first place a simple linear moel can be use to stuy the overall system performance. This also inclues a more etaile moel of the electrical machine in orer to stuy the influence of saturation an other non-linear behaviour. Detaile moels for the permanent magnet synchronous motor (PMSM) an inuction motor can be foun in reference [1]. A PMSM is employe in the rive system for riving the scooter. The following mathematical moel of PMSM is use in this paper. The voltage euations in the synchronous reference frame (-) are expresse as follows. v Ri p e e v Ri p (11) In above euations, the parameters of v, i, R, an p are stator voltage, current, stator resistance, electrical angular velocity an flux linkage, respectively. The flux

4 linkage euations in the synchronous reference frame are expresse as follows. Li Ke L i (1) Where L an K e are stator inuctance an electromotive force constant, respectively. The torue euation is given as follows. P K i ( L L ) i i (13) e Where P is number of pole pairs. The electrical machine moel is base on a - moel that also moels the reluctance torue in the PMSM. Inepenent saturation of L an L is not inclue the influence at high loas in this stuy. In the PMSM rive, a Fiel Oriente Control (FOC) is applie. A position encoer on the shaft is reuire for the transformation from measure phase currents into the - currents neee in the FOC. In this simulation, the PMSM was riven below its base spee. The magnetic fiel of the PMSM epens on the voltage-freuency relationship in the electric motor. Fiel weakening appears once the voltage remains constant an the freuency increases. The torue ecreases in proportion to the increasing spee an the power remaining constant. Since, the PMSM is riven below its base spee, fiel weakening is not applie an the -axis current of i is set to zero [1]. Thus, the commane -axis current of i is irectly etermine from the comman torue,..5 Hybri Power System Control Moe The whole system is implemente as shown in Fig. 1. The tery that maintains a DC voltage level at the main bus an the c/c converter fees a set output voltage of PEMFC. The control moes of the hybri tery an PEMFC system are as follows: (a) System start-up process: when the system starts, a tery is use for start because of fuel cell s output voltage to 48 by c-c converter which is below the open circuit voltage of li-ion tery uner start-up characteristic. While fuel cell can work stably, the controller will switch on the fuel cell to the system, which will take care of the whole system. (b) Range extener process: The c-c converter will convert fuel cell output voltage to 48 to supply the loa. When the PEMFC is power the loa, it can power no more than 3W. If suenly loa is larger than maximum power the tery can affor, the tery be supply the power together with the PEMFC. Once the loa consume excess them, the tery voltage will be reuce, an the PEMFC will supply continuously to the loa when the tery voltage (53.5~48) is less than the PEMFC voltage (48). The concept of fuel cell as an extene-range hybri powere scooter is currently use. The fuel cell has a slow response compare to the tery especially when higher loas are suenly reuire. The tery always respons first when a high loa eman occurs. It is note that ioes shoul be ae to this esign in orer to avoi feeback charge an amage to the fuel cell or c-c converter..6 Numerical Simulation In this hybri power system, a PMSM is use to rive the scooter. An extene-range hybri powere scooter, the tery is use for proviing the normal power for riving in the start-up, braking an provies power at nominal operation spees, while the fuel cell stack assists for proviing the nonnormal power for riving in the climbing, accelerating an extening miles. The hybri powere scooter simulation moels use here focus on the fuel cell moel, tery moel, the moel in control of power electronics, an the electrical machine moel for electrical motors. The moel for the electrical machines can either be a first-orer moel for system behaviour or can be base on FEM calculations. Figure 4: Simulation of the ynamic moel at full-rate spee of 45 kilometres per hour for the electric motor currents i/i, total power, scooter spee v, PMSM torue Te, virtual river torue comman Q an angular position θ from the encoer are inicate from top to bottom, in turn Fig. 4 shows the simulation of the complete electrical an mechanical rive train for a hybri powere scooter. Here, the inverter is moelle using ieal switch moels to achieve a fast simulation. The harmonics ue to the pulse with moulation (PWM) of the inverter appear in the torue prouce by the PMSM. This torue pulsation will affect the ynamics of the entire mechanical rive train. The ynamic moel is simulate for the PMSM an the power electronics with fiel oriente control at full-rate spee. In the figure, power, scooter spee, PMSM torue, virtual river torue comman an angular position from the encoer are inicate from top to bottom, in turn. This animation shows the rive train behaviour of the electric scooter. Only the back wheel is riven via a gearbox by the PMSM. These inputs are the PEM fuel cell stack/tery an the electrical machine, while the output of the torue is connecte to rive the scooter. Scooter characteristics of weight, rag, rolling resistance, an fuel cell polarization curves, are specifie as inputs as shown in Tab.. With a minimum set-up in orer to reuce total weight of electric scooter, the average power is less than 3W. In the present stuy, it is showe that a hybri power esign with a small fuel cell stack as a range extener an two main tery packs coul hanle almost the loa. 3 Experiments

5 Figure 5: Photos of the fuel cell stack, the tery packs, the metal hyrie canister, an the view of the electric scooter Table : Experimental ata of the electric scooter Test items Simulation Dimension mm Weight 63 kg (total) Tires 1 inch aluminum Noise none Limit spee 58km/h Acceleration sec Mileage 1km (fixe spee 45km/h) Inclination Fuel consumption 1.6 g hyrogen/km Lithium Ion Battery Li-ion Battery 8.5kg Two 48/1Ah 55Wh/L W/kg Energy Efficiency > 95% Life Time > Charging Time < hours Motor Brushless DC motor Max Power 135W Rate Power 75W Time of fuel supply none In orer to valiate feasibility of an extene-range hybri powere scooter, a roa test is execute. An electric scooter of e-moving mae by China Motor Corporation was use to test the experiment, which was moifie by replacing the original pure-tery power system with our esign system integrate with a small size of fuel cell stack (see Fig. 5). The framework of the seat was reesigne to make enough space for the fuel cell system installation. The key components of this hybri system inclue two 48 1Ah Li-ion tery packs, a 3W air-blowing PEM fuel cell stack, a c/c converter an a central controller. Table liste the etaile system ata of the main components in the scooter. The fuel cell stack was mae up of 7 cells with a 43. nominal voltage. It coul be operate without external gas humiification at ambient temperature. At the anoe, pure hyrogen was fe into the fuel cell at a controlle pressure aroun ~6kPa. A metal hyrie canister was use for on-boar hyrogen storage. The fuel cell anoe worke in a ea-en moe. An electronically controlle purge valve was fixe at the anoic ownstream line to perioically rain excessive water iffusing from the cathoe to the anoe through the membrane. An another solenoi valve was locate at stack inlet to control the hyrogen feeing. While at the cathoe, air was supplie through three blowers which also acte as cooling fans to keep the fuel cell stack at an appropriate temperature. This operating conition coul be ientifie at an open moe in the storage compartment, as fresh air was supplie through six blowers which keep the air for circulating. Base on the airblowing cathoe, the prouce water cannot be easily expelle from the cathoe by unpressurize airflow, thereby improving the fuel cell s self-humiifying ability. In orer to ensure safe an correct power istribution uring ifferent stages such as start-up, normal operation, an loa variation, a central controller was use to coorinate an comman all the components through efficient management of the control signals, on-off signals, serial port communication, as well as various operating conitions. A programe ata acuisition unit was employe to facilitate status monitoring an experimental ata collection. Moreover, a LCD panel was also fixe on the front of the electric scooter for the convenience of observation. Figure 6: Loa voltage output of monitoring is put into the pure electric scooter compare with a fuel cell range-extener by using 134 litres of hyrogen gas at a fixe spee of 8.1 miles per hour. Before the fuel cell system starte, the central controller was firstly activate by a key-button. Then the fuel cell system was turn-on an transite to the stan-by moe with a LED light. During this stage, sensors an actuators were activate an the control boar began to continuously eliver ata an status signals. Subseuently, a LED light triggere the starting seuence. Then the fuel cell system was on warmup. Once the scooter began to run, the power system woul enter the normal running stage until reaching a new steay state. The sampling ata was kept at 1s through the whole experiment. Real-time experiment of the hybri powere scooter in ynamic riving is of great significance to evaluation the system performance in both laboratory an loa tests. In this stuy, the ynamic performance was the main focus which coul reveal transient mechanisms an the ranges of a puretery scooter in comparison with aing a small size of fuel cell as a range extener. The simulation ata are also compare with the experimental results as showe in Fig. 6. In experimental ata, the result of electric scooter can continuously run by increasing the range from 6.5 miles to

6 78.7 miles at a fixe spee of 8.1 miles per hour with a PEMFC of range extener by using 134 litres of hyrogen fuel. As shown in Fig.6, goo matching results were foun for loa voltage output of monitoring which is to rive the pure tery powere scooter an with a small fuel cell stack as a range-extener. The maximum erivation between the moelle an experimental results is less than 1%. 4 Conclusions This paper shows that, more etaile feasibility on fuel cell range-extener for a hybri powere electric scooter. In experimental ata, the result of electric scooter can continuously run by increasing the range from 6.5 miles to 78.7 miles at a fixe spee of 8.1 miles per hour with a PEMFC of range extener by using 134 litres of hyrogen fuel. The experimental ata are also compare with the pure tery powere scooter. Not only can the range of aing a small size of fuel cell stack be increase further, but also the test performance can be more brilliant than that of a pure tery powere scooter. [8] James Larminie, an Anrew Dick, Electric ehicle Technology Explaine, Secon Eition, Willey Press, pp.54, (3). [9] J. C. Amphlett, R. M. Baumet, R. F. Mann, B. A. Peppley, P. R. Roberge, P.R., an A. Rorigues, A moel preicting transient responses of proton exchange membrane fuel cells, J. Power Sources, ol. 49, pp (1994). [1] Santarelli, M.G., Torchio, M.F., an Cochis, P., Parameters estimation of a PEM fuel cell polarization curve an analysis of their behaviour with temperature, Journal of Power Sources, ol. 159, pp , (6). [11] S. A. Khateeb, M. M. Fari, J. R. Selman, an S. Al- Hallaj, Mechanical electrochemical moelling of Li-ion tery esigne for an electric scooter, Journal of Power Sources, ol.158, pp , (6). [1] P. Bauer, P. Koroni, an P. J. an Duijsen, Integrate Control an circuit Simulation for a Motion Control System, EPE Conference, pp. 1-14, (3). Acknowlegements The author woul like to thank the National Science Council of Taiwan, for financially supporting this research uner the contracts of NSC 1-1-E an NSC E-33-1-MY3. References [1] Y. P. Yang, J. J. Liu, an T. H. Hu, An energy management system for a irectly-riven electric scooter, Energy Conversion an Management, ol.5, pp , (11). [] B. M. Lin, M. H. Yang, an T. Suan, Major activities of light electric scooter evelopment in Taiwan, Worl Electric ehicle Association Journal, pp , (7). [3] B. Lin, Conceptual esign an moelling of a fuel cell scooter for urban Asia, Journal of Power Sources, ol.86, pp-13, (). [4] S.Angkee, an L. Nartnarong, Experimental assessment of fuel cell/supercapacitor hybri system for scooters, International Journal of Hyrogen Energy, ol.34, pp , (9). [5] S. A. Khateeb, S. Amiruin, M. Fari, J. R. Selman, an S. Al-Hallaj, Thermal management of Li-ion tery with phase change material for electric scooters: experimental valiation, Journal of Power Sources, ol.14, pp , (5). [6] K. B. Sheu, Simulation for the analysis of a hybri electric scooter powertrain, Applie Energy, ol.85, pp , (8). [7] W. L. Soong, Fiel-Weakening Performance of Interior Permanent-Magnet Motors, IEEE Transaction on Inustry Applications, ol.38, No.5, pp , ().