international journal of hydrogen energy 34 (2009) 6036–6044 Available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/he Experimental assessment of fuel cell/supercapacitor hybrid system for scooters Angkee Sripakagorna,*, Nartnarong Limwuthigraijiratb a Chemical Energy Conversion Research Unit, Chulalongkorn University, Bangkok 10330, Thailand Department of Mechanical Engineering, College of Engineering, Chulalongkorn University, Bangkok 10330, Thailand b article info abstract Article history: This paper presents an experimental assessment of fuel cell hybrid propulsion systems for Received 6 March 2009 scooters based on a modular 1.2 kW PEM fuel cell. The tests of the hybrid system are Received in revised form carried out using a programmable electronic load. Different configurations of the fuel cell/ 25 April 2009 battery and the fuel cell/supercapacitor hybrid systems are explored. Both systems Accepted 26 April 2009 demonstrate their ability to deliver the requested load satisfactorily. The distributions of Available online 7 July 2009 the fuel cell power delivery, although different between the two systems, are within the region where the fuel cell efficiency is approximately constant. As a result, the rates of fuel Keywords: consumption show no discernable difference between the two systems for all three driving Fuel cell cycles considered. In addition to the fuel consumption, considerations including bus Hybrid propulsion voltage, cost and packaging issues suggest that the supercapacitor has advantages over the Supercapacitor battery for the use as secondary energy storage in fuel cell hybrid propulsion system for Scooter scooters. ª 2009 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved. 1. Introduction In many cities in Asia, the motor scooter is the major mode of transportation. The scooter population accounts for more than half of the vehicle fleet in China and Malaysia and for more than two-thirds in Indonesia, Vietnam, Taiwan and Thailand [1]. Low cost, convenience and flexibility are among the main reasons that make residents in these countries rely heavily on scooters. However, in contrast to cars where pollution prevention measures such as catalytic converters or sophisticated engine management are available, the small size and low cost of scooters make pollution prevention technology nonexistence or difficult to apply. It has been mentioned that even the large sport utility vehicles are 95% cleaner in terms of the pollution compared to conventional motorcycles due to emission control technology not available in motorcycles [2]. Because of environmental problems that have arisen from scooter use, attempts have been made to promote the use of the electric scooter as a zero emission alternative [3,4]. Despite recent developments, the limited range, long recharge time and limited infrastructure still hinder the large scale deployment of electric scooters. The fuel cell vehicle is envisioned as the vehicle of the future in response to environmental, economic and political constraints [5–9]. Most of the recent development of fuel cell vehicles concerns fuel cell hybrids. Benefits of the fuel cell hybrid system include enhancing fuel economy, capturing regenerative braking and reducing the system cost. However, a hybrid system presents a unique challenge in bringing multiple energy sources to work together effectively [10–15]. The battery is primarily used as a secondary energy storage for fuel cell hybrid systems due to its low cost and wide * Corresponding author. Tel.: þ662 2186595; fax: þ662 2522889. E-mail address: angkee.s@gmail.com (A. Sripakagorn). 0360-3199/$ – see front matter ª 2009 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2009.04.059 6037 international journal of hydrogen energy 34 (2009) 6036–6044 availability [16,17]. Recently, supercapacitors are also explored as an alternative to complement the fuel cell under different control strategies [18,19]. In addition, past studies that concern hybrid system applications in small vehicles such as scooters are rather limited [20–23]. There are questions worth exploring including: hybridization degree, consideration of weight and size limits, and comparative advantages of battery and supercapacitor as a secondary energy storage. This paper presents the development of fuel cell hybrid propulsion systems for scooters. Different configurations of the fuel cell/battery and fuel cell/supercapacitor system are explored. This paper explains in detail the installation and testing of the fuel cell hybrid systems. The propulsion systems are tested under different driving cycles using an electronic load. The propulsion systems are compared in terms of driving performance, energy consumption, cost and maintenance requirements. This work also attempts to determine whether the battery or the supercapacitor is the best secondary energy storage for fuel cell hybrid propulsion systems for scooters. Table 1 – Specifications of parameters for the road load modeling. hybrid propulsion. The calculation of the average power following the selected driving cycles yields a maximum of 647 W. The availability of a compact and modular 1.2 kW fuel cell unit was found to fit the need as a main energy source. The secondary energy storage is then selected to fulfill the peak power and peak energy usage over driving cycles. 2.2. 2. Design of the power train 2.1. Traction load modeling The goal of this work is the development of the fuel cell hybrid propulsion system for scooters. The performance target is placed between the performance of a conventional gasoline scooter with 50 cc and 125 cc engine displacement. To design the power train to meet such a performance requirement, the total traction power requirement is determined by: Pm ¼  1 1 fr mgv þ rACd v3 þ mgv sinðqÞ hm 2    nw Iw Im G2 dv þ mþ 2 þ 2 v r r dt (1) The traction power involves rolling resistance, aerodynamic drag, climbing effort and inertia term respectively. The efficiency of the motor is provided by: hm ðT; uÞ ¼ Tu Tu þ kc T2 þ ki u þ ku u3 þ Cl (2) based on the characteristics of brushed DC motors. The parameters adopted for the calculation of the road load is provided in Table 1. For the tests to represent realistic driving conditions, three types of driving cycles specially developed for city driving pattern are employed. The choices include: a) ECE-15 which is based on city driving in European cities and characterized by high velocity and long idling period, b) modified FTP-75 where only the transient phase of the FTP-75 is extracted and characterized by high acceleration and short idling period and c) NYCC which is developed from city driving in New York and characterized by very high acceleration and long idling period. Considering the cost and space constraints of scooter application, the development is directed toward the fuel cell 0.01 0.75 0.6 m2 160 kg 5% of m 0.21 m 1–2 1.5 0.1 105 20 W fr Cd A m nwh Iwh Im G2 þ 2 r2 r r G kc ki Ku Cl Fuel cell system The fuel cell unit employed in this study is a PEM fuel cell system rated at 1.2 kW and supplied by Ballard. The unit comes with a fuel cell stack and auxiliary systems necessary for the power delivery. The unit is lightweight (13 kg in total), air-cooled and utilizes fuel cell product water to humidify the incoming air and, hence, simplify the integration into the propulsion system. In Table 2 the specifications of the fuel cell stack and hydrogen supply are presented. A compact air compressor feeds ambient air into the fuel cell stack. The speed of the compressor and the stoichiometric air ratio are regulated to suit the current drawn from the fuel cell system. From the electrochemical reaction, nitrogen and product water gradually accumulates at the anode. The purge cells periodically flush out inert gases to restore the performance of the unit. The purged volume of hydrogen is negligible (less than 1%) compared to the overall fuel consumption rate. Once diluted into the cooling air stream, the hydrogen content in the purged gas is many times less than the hydrogen flammability limit [24,25]. 2.3. Secondary energy storage For integration into the hybrid system, the battery is first considered as a secondary energy storage. The battery of Table 2 – Specifications of the fuel cell system. Stack Number of cells Active area per cell Rated output power Range of operating stack voltage Stack voltage at rated power 48 115.8 cm2 1200 W 22–50 V 26 V Hydrogen supply Purity Inlet gas pressure 99.99% H2 (Vol) 70–1720 kPa 6038 international journal of hydrogen energy 34 (2009) 6036–6044 choice for the present work is the valve-regulated lead acid (VRLA) battery model CP12180 from Vision. This choice comes from a combination of the low cost, high charge and discharge current ability and long-life advantages of the VRLA battery. The supercapacitor is another option for the secondary energy storage to be explored in this study. Sized to be comparable in volume and operating voltage, a supercapacitor module rated 165 Farad and 48.6 V supplied from Maxwell technologies is used. Table 3 reports the specifications of the battery and the supercapacitor. 2.4. DC/DC converter To enable different energy sources to work together in an electric propulsion system, a simple concept is to match their voltages once they are connected in parallel to the system bus. Considering large variations and different ranges of voltage over the operation of the energy sources, a DC/DC converter is the power electronic device that enables such voltage regulation. This study used a step-up bidirectional DC/DC converter supplied by Zahn electronics Inc. The specifications of the DC/DC converter are provided in Table 4. 3. Experimental setup The schematic diagram of the fuel cell hybrid system is depicted in Fig. 1. Among many configurations to be explored in Section 4, this configuration corresponds to the configurations in Figs. 2 and 3a. The components of the fuel cell hybrid system were installed and connected on a test bench (Fig. 2). The tests of the hybrid system were carried out using a programmable electronic load. The traction load was calculated using the road load modeling. The traction load following a selected driving cycle was then programmed into the electronic load. By using the electronic load, the test can be executed following different realistic driving profiles at minimum time and cost and at maximum safety. The use of electronic load, however, excludes the effect of regenerative braking. In fuel cell vehicles, regenerative braking could save the propulsion energy and extend the range up to 24–28% in heavy buses running in a high-speed driving routine such as FTP-75 [26]. The advantage of the regenerative braking, however, becomes less effective in a passenger car [18] where energy saving is about 17%. The energy saving is even less in scooters where Corbo (2006) [15] arrived at less than 10% with the R40 driving cycle. The present study looked into the different Table 3 – Specifications of the secondary energy storage. Battery Nominal voltage Capacity Weight 12 V  4 units 18 Ah 5.9 kg each Supercapacitor Rated voltage Capacitance Weight 48.6 V 165 F 14.2 kg Table 4 – Specifications of the DC/DC converter. Operating range of input voltage Rated power Input current limit 12–61 V 3.9 kW 120 A modes of road loads (e.g. Eq. (1)) and found that in small vehicles such as scooters, the contribution of air drag to the total road load is more important than other terms especially the inertia term compared to larger vehicles. As a result, although the regenerative energy available per unit mass of the vehicles is comparable in all sizes of vehicle, the ratio of the regenerative energy per traction energy is much less for small vehicles. In addition, efficient use of regenerative braking has proved to be difficult in practice [27]. If the secondary energy storage (such as a battery) is at a high state of charge, the regenerative braking operation is usually tuned down or disengaged to prevent damage to the battery. The braking management system also has to be rather refined such that both regeneration effectiveness and safety are retained [28]. The high complexity and high price of such a system might not be justified for the low cost scooters. For these reasons, this study omits the effect of regenerative braking and proceeds with the use of electronic load for the execution of the tests. For the present study, a 1 kW electronic load connected with two units of 2 kW load booster is employed. The total load rating is 5 kW. The units are supplied by Kikusui. A dedicated data acquisition system supplied by National Instrument was employed to monitor the voltage and current signals between the different components of the hybrid system. This is additional to the embedded data acquisition system and controller board native to the fuel cell system. The controller board of the fuel cell system obtained various system parameters necessary for the fuel cell unit operation including fuel cell stack temperature, stack current, stack voltage, air flow rate, hydrogen pressure, purge cell voltage, hydrogen concentration, cumulative hydrogen consumption, oxygen concentration and ambient temperature. The energy efficiency of the fuel cell system can be determined from the net electrical power output Wnet (in Watts) and _ H2 (in gram H2/sec): the measured hydrogen consumption m hFCS ¼ Wnet _ H2 LHV$m (3) where LHV is the lower heating value of hydrogen in Joule/ gram H2. For a hybrid system, the determination of the fuel consumption is not as straightforward as a nonhybrid fuel cell system. For example, during a test, the state of charge of the battery could change. That refers to not only the hydrogen consumption that contributes to the energy supply but also the changes in the state of charge of the battery. In this study, hydrogen consumption is corrected following the procedure of Ding et al. [29] with some minor modifications. The corrected fuel consumption is HCorr ¼   DESES  Hmeas ðgram H2 Þ 1 EFC (4) international journal of hydrogen energy 34 (2009) 6036–6044 V 6039 V I I Fuel cell system I I Electronic Secondary energy source DC/DC Converter Load V I Signal DAQ LabVIEW Fig. 1 – Schematic diagram of the fuel cell hybrid system. where Hmeas is the measured fuel consumption (gram H2), DESES is the change in the energy level of the secondary energy storage after the test (J), and EFC is the energy delivered by the fuel cell over the test (J). For the present work, the state of charge (SOC) of the battery is correlated with the open-circuit voltage VOC [30–32] and is calculated from VOC ¼ nC ð2:15  0:2$SOCÞ (5) where nC is the number of cells of the battery. The battery is left unconnected for 24 h after the test to ensure an accurate determination of the state of charge. The term DESES is a product of the change in the state of charge and the nominal voltage of the battery. The term EFC is calculated directly from the integration of the product of the operating voltage and current over the time period that the tests were carried out. 4. Selection of the fuel cell hybrid configurations Among many configurations available, this study explores different configurations for fuel cell/battery hybrids and fuel cell/supercapacitor hybrids such that the best configurations in terms of functionality are identified for each case. 4.1. Fuel cell/battery hybrid Fig. 3 shows a configuration for the fuel cell/battery hybrid system. Due to its rather constant voltage over a wide operating range, the battery is typically connected directly to the voltage bus. On the other hand, the fuel cell system generally requires a DC/DC converter to regulate the output voltage to a rather constant value on the voltage bus, which is required for a good propulsive performance of the power train. There are two control strategies for the fuel cell/battery hybrid systems, namely, the load-leveled and the loadfollowing strategies. In the load-leveled strategy, the fuel cell is controlled to deliver a constant power preferably at high efficiency. The battery, in turn, supplies the peak power requirement. For the present setup, the adjustment for the load-leveled operation involves setting the output voltage from the DC/DC converter to 55.2 V which is the charging voltage for the battery set. The maximum input current is set to the maximum efficiency point of the fuel cell unit. This setting, in turn, controls the fuel cell to operate steadily at the best efficiency. For the load-following strategy, the fuel cell is controlled to respond to the requested load instantaneously up to its rated power. Should the peak power requirement go beyond the rating of the fuel cell, the battery steps in to supply the peak Fuel cell Load DC/DC Battery Fig. 2 – The fuel cell hybrid system on test bench. Fig. 3 – A configuration for the fuel cell/battery hybrid system. 6040 international journal of hydrogen energy 34 (2009) 6036–6044 power. For the load-following operation, the DC/DC converter in front of the fuel cell is set such that the output voltage is roughly at the open-circuit voltage of the battery (52.0 V in this case). In addition, for the safety and durability of the fuel cell, the maximum input current from the fuel cell is limited to the maximum power condition of 1.2 kW (45 A). Fig. 4 shows the result following the ECE-15 driving cycle from the load-leveled strategy. It demonstrated that at the time during which the fuel cell power is requested, the fuel cell power delivery is leveled at 400 W where the fuel cell efficiency is at the maximum. There are times when there is a low load request and the battery is rather fully charged; the fuel cell power drops accordingly. Overall, the hybrid system is able to provide the traction load satisfactorily throughout the driving cycle. After continued runs over many cycles, the battery is observed to be slowly depleted of energy under this strategy. That is because the power level of the fuel cell at its best efficiency point is not sufficient for the load averaged over the driving cycle [15]. As a result, the power level is later raised to be slightly higher than the average power over the driving cycle. However, care should be taken not to allow the fuel cell to supply power at a level higher than necessary since that would adversely affect the operating efficiency of the fuel cell. After the adjustment of the fuel cell power delivery, the fuel cell is able to sustain the state of charge of the battery over the prescribed driving cycle. For the load-following strategy, the hybrid system also performs satisfactorily (Fig. 5). The fuel cell is able to supply the traction load with good transient performance up to its rated power. However, under this strategy, the fuel cell barely charges the energy back to the battery over the execution of the driving cycle [33]. The battery is continually depleted over the cycle. For continuous runs of the fuel cell hybrid power train, extra time for battery recharge is required. To conclude, two control strategies for the fuel cell/battery hybrid were investigated. The load-leveled strategy was found to be more appropriate since the strategy allows dedicated battery charging. In addition, the high power delivery from the fuel cell in the load-following operation also means a large charging current to the battery when the traction load is low. That could shorten the life of the lead acid battery in the long run. Battery Power FC Power Net Power Bus Voltage 3000 4.2. Supercapacitors have been applied to propulsion systems to enhance the driving performance and fuel economy [19,34–36]. Nevertheless, details regarding the physical connection of the devices and control strategy are rather limited in the open literature. This study attempted to explore different configurations of fuel cell hybrids for scooter propulsion. In contrast to a battery, the voltage of a supercapacitor changes linearly with its state of charge. This behavior makes the hybrid connection more complicated. Fig. 6 shows three different configurations of the fuel cell/supercapacitor hybrid. With the focus on scooter propulsion, the space and cost constraints need to be taken into account. As a result, this study only examines configuration (6a) and (6b) in Fig. 6 where only one DC/DC converter is employed. The first configuration to be considered is the configuration (6a). In this case the DC/DC converter is installed in front of the fuel cell to regulate the output voltage to 48.6 V (the maximum operating voltage of the supercapacitor module). Installed this way, the fuel cell recharges the supercapacitor whenever the load is low or the system is idle. The input current from the fuel cell system is also limited to 22 A. This current value corresponds to the fuel cell power output of 680 W (the average traction power over the driving cycle is 465 W). Fig. 7 shows that the hybrid configuration (6a) supplies the traction load adequately over the ECE-15 driving cycle. Under the load request, the fuel cell unit responds first with good transient performance. The fuel cell also keeps the power at the same power level to charge the supercapacitor promptly. Under a no-load condition or when the supercapacitor is at its full charge, the fuel cell unit reduces its power delivery. For the configuration (6b), the DC/DC converter regulates the output voltage from the supercapacitor module to match the voltage of the fuel cell at 31 V. This voltage value corresponds to a fuel cell current output of 24 A which is roughly similar to the current limit applied in the configuration (6a). Fig. 8 shows that the configuration (6b) supplies adequate traction power as required. In this case, the supercapacitor responds first to the load request. Only when the state of charge of the supercapacitor reduces its voltage to below the set point voltage at 31 V is the fuel cell power raised to the set 800 -1000 900 10 1000 0 Time (s) Fig. 4 – Power sharing from the fuel cell/battery hybrid system in load-leveled configuration. Power (W) Power (W) 20 0 Bus Voltage 60 40 30 1000 20 0 0 50 100 150 20010 0 -1000 Time (s) Fig. 5 – Power sharing from the fuel cell/battery hybrid system in load-following configuration. Bus Voltage 30 Net Power 2000 Bus Voltage 1000 FC Power 50 50 40 Battery Power 3000 60 2000 Fuel cell/supercapacitor hybrid 6041 international journal of hydrogen energy 34 (2009) 6036–6044 a Fuel cell SC Power Load DC/DC FC Power Net Power 3000 Power (w) 2000 Supercapacitor b 1000 Load Fuel cell 0 800 DC/DC Time (s) Fig. 8 – Power sharing from the fuel cell/supercapacitor hybrid configuration (6b). Supercapacitor c Load DC/DC (6b) cannot satisfactorily function with the motor and motor controller for propulsive purposes. In addition, over the driving cycle, the low voltage translates to the high current drawn from the peak power source (i.e. supercapacitor in this case). This results in a larger current rating of the DC/DC converter and eventually the higher cost of the system [37]. In conclusion, the choice of configurations for the fuel cell/ supercapacitor hybrid is the configuration (6a) where the DC/ DC converter is installed in front of the fuel cell system. The hybrid system handled the traction load adequately over the driving cycle. The bus voltage is also kept at a high and rather constant level. DC/DC Supercapacitor Fig. 6 – Configurations for the fuel cell/supercapacitor hybrid system. point power. Although the system appears to perform well in terms of power delivery, a closer inspection into the voltage variations in Fig. 9 revealed a problem. Instead of a relatively constant value of the voltage in the configuration (6a), the bus voltage varied over a large value in this case. That is because the DC/DC converter is a step-up converter. Under this configuration, whenever the supercapacitor voltage is larger than 31 V, the system operates as if the DC/DC converter does not exist. The voltage of the supercapacitor dictates the bus voltage. Only when the state of charge of the supercapacitor brings the voltage of the supercapacitor below the set point voltage of 31 V does the DC/DC converter step in to maintain the bus voltage at 31 V. Added to this fact, the minimum voltage is set to a low value of 31 V. Hence, the configuration SC Power Net Power 43 60 50 2000 40 30 1000 20 0 1100 10 1200 0 Time (s) Fig. 7 – Power sharing from the fuel cell/supercapacitor hybrid configuration (6a). Bus Voltage Power (W) This work also attempts to identify the best choice of the secondary energy storage between the battery and the supercapacitor in a fuel cell hybrid system for scooters. To reach the goal, comparative tests were performed on the choices of the fuel cell hybrid system identified in Section 4. Considerations include fuel consumption, ability to sustain the traction load over driving cycles, and variations of the bus voltage. The comparison also concerns other issues including cost, expected life and maintenance requirements. FC Power Bus Voltage 3000 1000 -1000 5. The choice of the secondary energy storage Voltage (V) Fuel cell 1000 900 -1000 41 SC Voltage 39 Bus Voltage 37 35 33 31 29 27 25 800 900 1000 Time (s) Fig. 9 – Variations of the bus and supercapacitor voltage from the fuel cell/supercapacitor hybrid configuration (6b). 6042 Comparative tests Battery Power Fuel cell Power Net Power Bus Voltage Bus Voltage 4000 60 3000 50 40 2000 30 1000 20 0 100 -1000 200 10 400 0 300 Time (s) Fig. 11 – Power sharing and bus voltage in the fuel cell/ supercapacitor hybrid system under a modified FTP-75 driving cycle. a result, the average fuel cell efficiencies are almost the same in both cases (see Table 5). Once the corrections for changes in the state of charge of the secondary energy storage are applied, it is not a surprise that both systems arrive at the same fuel consumption. Another issue that is very important in vehicular propulsion is the bus voltage. For the propulsion system to work with the motor and its power electronics, the bus voltage should preferably stay at a relatively constant value [36]. Results shown in Table 6 and Figs. 10 and 11 indicate that the fuel cell/supercapacitor system exhibits a smaller change in the bus voltage compared to the fuel cell/battery system for the driving cycles with low average power such as ECE-15 and NYCC. In contrast, the voltage variation is larger for the fuel cell/supercapacitor system for the driving cycle with high average power e.g. the modified FTP-75. 5.2. Cost and packaging considerations In addition to the performance in terms of propulsive power and fuel consumption, this study also considered other issues including ownership cost and packaging considerations. The ownership cost of the secondary energy storage consists of the initial cost, the operating cost and the replacement cost. For the same operating voltage and about the same volume, in this case, the initial cost of the supercapacitor is approximately four times that of the battery set. The operating cost of energy storage is related to its 50 3000 40 2000 30 1000 20 0 200 300 Bus Voltage Power (W) Fuel cell Power Net Power 60 4000 100 -1000 SC Power Bus Voltage The tests employ the best configurations of the fuel cell hybrid system identified in Section 4. The fuel cell/battery system employed the physical connection depicted in Fig. 3 together with the load-leveled strategy; the fuel cell/supercapacitor system used the hybrid configuration (6a). The tests were executed on each of the hybrid system following the selected pattern of three driving cycles. For all tests, the power delivery from the fuel cell is limited to 750 Watts. This value is derived from the averages of the traction load requested by the three driving cycles described earlier while the fuel cell efficiency is still reasonably high at this power level. Figs. 10 and 11 illustrate that both systems are able to execute the driving cycles adequately and with good response. The figures show the power sharing between the primary and secondary energy sources in cases of the modified FTP-75 driving cycle. For the other two driving cycles (ECE-15 and NYCC), both systems also perform well. The fuel consumption from the two cases is compared in Table 5. Compared to conventional 125 cc 4-stroke gasoline scooters on the same driving cycle, the fuel cell power train provides over 3 times the energy efficiency. The present results for the fuel cell/battery hybrid system compared well with similar studies in scooter applications by Corbo (2006) [15]. Their results yield the corrected fuel consumption at 1.35 and 1.33 g H2/km over the ECE-40 cycle for load-leveled and load-following operations respectively. From Table 5, for all driving cycles investigated, the corrected fuel consumption shows no discernable difference between the two systems. To partially explain this result, the distributions of the fuel cell power delivery over the execution of the three driving cycles are illustrated in Figs. 12 and 13. With the supercapacitor where there is no practical limitation on the charge and discharge current, the fuel cell either delivers the full power (at 750 W) to rapidly charge the supercapacitor or idles at a very low power. The situation differs with the battery. In addition to the delivery at full power for most of the time, the distribution is broader with the delivery of 300–500 W over a large portion of time. Despite such differences, the fuel cell efficiency data displayed in Figs. 12 and 13 suggested a very flat profile over the range of fuel cell power delivery. As Power (W) 5.1. international journal of hydrogen energy 34 (2009) 6036–6044 Table 5 – Test results from the comparison tests of the secondary energy storages. Fuel consumption (gram H2 per km) 10 400 0 Time (s) Fig. 10 – Power sharing and bus voltage in the fuel cell/ battery hybrid system under a modified FTP-75 driving cycle. Average fuel cell efficiency (% for LHV) Supercapacitor Battery Supercapacitor Battery ECE-15 Modified FTP-75 NYCC 1.75 1.87 1.78 1.85 47.1 47.1 47.7 47.5 2.61 2.73 47.5 47.4 international journal of hydrogen energy 34 (2009) 6036–6044 100 100 ECE-15/BATT 80 80 mFTP-75/BATT 60 60 40 40 20 20 0 0 800 400 1200 Efficiency (%LHV) Proportion (%) NYCC/BATT Fig. 12 – The distribution of the fuel cell power delivery and efficiency for the fuel cell/battery hybrid system. maintenance requirement. Although quite rugged in construction, the battery set still needs an equalization charge from time to time from the user. For the battery set to work under high rate partial state of charge conditions in this hybrid propulsion system, a maintenance charge is also required to reduce the accumulation of lead sulfate [38]. In contrast, the supercapacitor requires practically no maintenance over its life time. The replacement cost depends on the life of the energy storage. With the usage pattern as a secondary energy storage, the life of a battery set is determined by calendar years rather than the number of charge/ discharge cycles. On the other hand, the supercapacitor is quoted to last over a million cycles and could last the life time of the vehicle-years [39]. Weight and volume are also important in the packaging of the scooters. In this study, the volumes of the two energy storages are comparable. The battery, however, is almost two times heavier. The weight of the present battery set carries as much as 10 times the amount of usable energy compared to the supercapacitor. Nonetheless, this amp-hour rating of the battery is selected not for its energy but for its power. For the present work, the peak charge/discharge current could reach the maximum of 110 A in the NYCC driving cycle. This value is equal to 6C in the battery terminology. The higher the value of C, the shorter the life span of the battery. Should the battery be a size smaller, the life penalty incurs. In addition, the supercapacitor module currently used is simply a package of 18 individual units internally connected by voltage equalization Proportion (%) 60 60 40 40 20 20 0 0 400 800 1200 % Vbus,min/Vbus,max ECE-15 Modified FTP-75 NYCC Supercapacitor Battery 93 78 91 84 78 72 circuitry. For its inclusion as a commercial product, the supercapacitor offers more flexibility in terms of packaging compared to the battery where the size of the individual unit is larger and the orientation of the unit is restricted. To conclude, in terms of the ownership cost, for the usage expectancy of scooters of 10–12 years, there is no clear advantage of one over another type of energy storage. The batteries, however, carry a hidden cost, namely the required maintenance. In terms of weight and volume, the supercapacitor offers a clear advantage over the battery. The lightweight and the flexibility in packaging are very important in space and weight-conscious scooter design. 6. Conclusions This paper presents the development of fuel cell hybrid propulsion systems for scooters. Among many configurations considered, the choices of the fuel cell/battery and the fuel cell/supercapacitor hybrid are identified. The fuel cell hybrid systems were tested under three different driving cycles characterizing differing aspects of city driving. Both systems demonstrate their ability to deliver the requested load satisfactorily and with good transient response. The distributions of the fuel cell power delivery, although quite different between the two systems, are within the region where the fuel cell efficiency is approximately constant. As a result, the rates of fuel consumption show no discernable difference between the two systems. Other issues including bus voltage, cost and packaging consideration suggest that the supercapacitor has advantages over the battery for use as a secondary energy storage source in fuel cell hybrid propulsion systems for scooters. references Efficiency (%LHV) 100 NYCC/SC ECE-15/SC 80 mFTP-75/SC 80 Table 6 – Voltage variation from the two types of hybrid systems. 0 Power (W) 100 6043 0 Power (W) Fig. 13 – The distribution of the fuel cell power delivery and efficiency for the fuel cell/supercapacitor hybrid system. [1] Policy guidelines for reducing vehicle emissions in Asia, cleaner two and three wheelers, Asian Development Bank; 2003. [2] Technology Review. Lithium-ion motorcycles, http://www. technologyreview.com/energy/19069; 2007 [accessed 15.01.09]. [3] Jonathan Weinert J, Ogden J, Sperling D, Burke A. The future of electric two-wheelers and electric vehicles in China. Energy Policy 2008. [4] Lin BM, Yang MH, Suan T. Major activities of light electric scooter development in Taiwan. WEVA J 2007;1. 6044 international journal of hydrogen energy 34 (2009) 6036–6044 [5] McNicol BD, Rand DAJ, Williams KR. Fuel cells for road transportation purposesdyes or no? J Power Sources 2001; 100:47–59. [6] Van Mierlo J, Maggetto G, Lataire Ph. Which energy source for road transport in the future? A comparison of battery, hybrid and fuel cell vehicles. Energy Convers Manag 2006;47:2748–60. [7] Granovskii M, Dincer I, Ma Rosen. Economic and environmental comparison of conventional, hybrid, electric and hydrogen fuel cell vehicles. J Power Sources 2006;159(2): 1186–93. [8] Richard SP, Whale M, Djilali N. A techno-economic analysis of decentralized electrolytic hydrogen production for fuel cell vehicles. Int J Hydrogen Energy 2005;30(11):1159–79. [9] Schlecht L. Competition and alliances in fuel cell power train development. Int J Hydrogen Energy 2003;28(7):717–23. [10] Lee SH, Jeong KS, Oh BS. An experimental study of controlling strategies and drive forces for hydrogen fuel cell hybrid vehicles. Int J Hydrogen Energy 2003;28:215–22. [11] Pei P, Ouyang M, Lu Q, Huang H, Li X. Testing of an automotive fuel cell system. Int J Hydrogen Energy 2004;29:1001–7. [12] Corbo P, Corcione FE, Migliardini F, Veneri O. Experimental study of a fuel cell power train for road transport application. J Power Sources 2005;145(2):610–9. [13] Corbo P, Corcione FE, Migliardini F, Veneri O. Energy management in fuel cell power trains. Energy Convers Manag 2006;47(18–19):3255–71. [14] Emadi A, Rajashekara K, Williamson SS, Lukic SM. Topological overview of hybrid electric and fuel cell vehicular power system architectures and configurations. IEEE Trans Veh Tech May 2005;54(3):763–70. issn 0018-9545. [15] Corbo P, Corcione FE, Migliardini F, Veneri O. Experimental assessment of energy-management strategies in fuel-cell propulsion systems. J Power Sources 2006;157:799–808. [16] Jiang Z, Gao L, Blackwelder MJ, Dougal RA. Design and experimental tests of control strategies hybrid fuel cell/ battery power sources. J Power Sources 2004;130:163–71. [17] Sapienza C, Andaloro L, Matera FV, Dispenza G, Creti P, Ferraro M, et al. Batteries analysis for FC-hybrid powertrain optimization. Int J Hydrogen Energy 2008;33:3230–4. [18] Rodatz P, Paganelli G, Sciarretta A, Gazzella L. Optimal power management of an experimental fuel cell/ supercapacitor-powered hybrid vehicle. Control Eng Pract 2005;13:41–53. [19] Gao W. Performance comparison of a fuel cell-battery hybrid powertrain and a fuel cell-ultracapacitor hybrid powertrain. IEEE Trans Veh Tech May 2005;54(3):846–55. [20] Tso C, Chang S. A viable niche market – fuel cell scooters in Taiwan. Int J Hydrogen Energy 2003;28:757–62. [21] Bruce Lin. Conceptual design and modeling of a fuel cell scooter for urban Asia. Master thesis Department of Mechanical and Aerospace Engineering School of Engineering and Applied Sciences Princeton University; 1999. [22] Arne LaVen. Development of a prototype fuel cell powered motor scooter. Master thesis University of Nevada; 1999. [23] Colella W. Market prospects, design features, and performance of a fuel cell-powered scooter. J Power Sources 2000:255–60. [24] Reggiani U, Sandrolini L, Giuliattini Burbui GL. Modelling a PEM fuel cell stack with a nonlinear equivalent circuit. J Power Sources 2007;165:224–31. [25] Nexa TM. Power module user’s manual. Ballard Power Systems; June 2003. [26] Folkesson A, Andersson C, Alvfors P, Alakula M, Overgaard L. Real life testing of a hybrid PEM fuel cell bus. J Power Sources 2003;118:349–57. [27] Sovran G, Blaser D. Quantifying the potential impacts of regenerative braking on a vehicle’s tractive-fuel consumption for the U.S., European, and Japanese Driving Schedules, SAE 2006-01-0664. [28] Duoba M, Bohn T, Lohse-Busch H. Investigating possible fuel economy bias due to regenerative braking in testing Hevs on 2wd and 4wd Chassis Dynamometers, SAE 200501-0685. [29] Ding Y, Kulik E, Bradley J, Kochis T, Thomas F, Markel T, et al. Hydrogen fuel cell vehicle fuel economy measurements and calculation SAE 2004-01-1339. [30] Linden D. Handbook of batteries. New York: McGraw-Hill; 1995. [31] Cadirci Y, Ozkazanc Y. Microcontroller-based on-line stateof-charge estimator for sealed lead–acid batteries. J Power Sources 2004;129:330–42. [32] Genesis application manual. 2nd ed. Hawker Energy Products, Inc; 1997. [33] Corbo P, Migliardini F, Veneri O. Experimental analysis and management issues of a hydrogen fuel cell system for stationary and mobile application. Energy Convers Manag 2007;48:2365–74. [34] Schupbach R, Balda JC. The role of ultracapcitors in an energy storage unit for vehicle power management. In: Proceedings of the 58th IEEE Veh Technol Conf, vol. 5; Oct. 2003. p. 3236–40. [35] Cheng Y, Van Mierlo J, Bosschet PV, Lataire Ph. Energy sources control and management in hybrid electric vehicles, EPE-PEMC2006, Portoroz, Slovenia IEEE; 2006. [36] Van Mierlo J, Cheng Y, Timmermans JM, Bosschet PV, Comparison of fuel cell hybrid propulsion topologies with supercapacitor, EPE-PEMC2006, Portoroz, Slovenia IEEE; 2006. [37] Choi W, Enjetic PN, Howze JW. 19th Annual IEEE applied power electronics conference and expositiondAPEC 2004, vol. 1; 2004. p. 385–90. [38] Moseley PT. High rate partial-state-of-charge operation of VRLA batteries. J Power Sources 2004;127:27–32. [39] Bohlen O, Kowal J, Sauer DU. Ageing behavior of electrochemical double layer capacitors part I. Experimental study and ageing model. J Power Sources 2007;172:468–75. Nomenclature Pm,in: traction power requirement, W fr: tire rolling coefficient m: vehicle mass, kg v: vehicle velocity, m/s A: vehicle frontal area, m2 Cd: aerodynamic drag coefficient q: climbing angle, radian nwh Iwh Im G2 þ 2 : rotational inertia, kg r2 r hm: motor efficiency T: motor torque, N.m u: motor rotational speed, rad/s kc: copper loss coefficient ki: iron loss coefficient ku: friction loss coefficient Cl: constant loss, W