Batteries and Ultracapacitors for Electric, Hybrid, and Fuel Cell Vehicles

Transcription

1 INVITED PAPER Batteries and Ultracapacitors for Electric, Hybrid, and Fuel Cell Vehicles Simulations indicate that fuel-efficient hybrid-electric vehicles can be designed using either batteries or ultracapacitors and that the decision between the two technologies is dependent on their cost and useful life. By Andrew F. Burke ABSTRACT The application of batteries and ultracapacitors in electric energy storage units for battery powered (EV) and charge sustaining and plug-in hybrid-electric (HEV and PHEV) vehicles have been studied in detail. The use of IC engines and hydrogen fuel cells as the primary energy converters for the hybrid vehicles was considered. The study focused on the use of lithium-ion batteries and carbon/carbon ultracapacitors as the energy storage technologies most likely to be used in future vehicles. The key findings of the study are as follows. 1) The energy density and power density characteristics of both battery and ultracapacitor technologies are sufficient for the design of attractive EVs, HEVs, and PHEVs. 2) Charge sustaining, engine powered hybrid-electric vehicles (HEVs) can be designed using either batteries or ultracapacitors with fuel economy improvements of 50% and greater. 3) Plug-in hybrids (PHEVs) can be designed with effective all-electric ranges of km using lithium ion batteries that are relatively small. The effective fuel economy of the PHEVs can be very high (greater than 100 mpg) for long daily driving ranges ( km) resulting in a large fraction (greater than 75%) of the energy to power the vehicle being grid electricity. 4) Mild hybrid-electric vehicles (MHEVs) can be designed using ultracapacitors having an energy storage capacity of Wh. The fuel economy improvement with the ultracapacitors is 10% 15% higher than with the same weight of batteries due to the higher efficiency of the ultracapacitors and more efficient engine operation. 5) Hybridelectric vehicles powered by hydrogen fuel cells can use either batteries or ultracapacitors for energy storage. Simulation results indicate the equivalent fuel economy of the fuel cell Manuscript received September 21, 2006; revised November 29, The author is with the Institute of Transportation Studies, University of California- Davis, Davis, CA USA ( afburke@ucdavis.edu). Digital Object Identifier: /JPROC powered vehicles is 2 3 times higher than that of a gasoline fueled IC vehicle of the same weight and road load. Compared to an engine-powered HEV, the equivalent fuel economy of the hydrogen fuel cell vehicle would be times higher. KEYWORDS Batteries; control strategies; fuel cells; hybrid vehicles; improved fuel economy; ultracapacitors I. INTRODUCTION In order to improve driveline efficiency and/or to provide for the use of energy sources other than petroleum for road transportation, engine powered hybrid-electric and fuel cell powered vehicles are being developed by auto companies around the world. The drivelines of these vehicles utilize electric motors and electrical energy storage to supplement the output of the engine or fuel cell during vehicle acceleration and cruise and for energy recovery during braking. The energy storage technologies being utilized are rechargeable batteries and ultracapacitors (electrochemical capacitors). The energy storage units can be recharged from the engine or fuel cell or from the electric grid much like an electric vehicle. In the later cases (often referred to as plug-in hybrids), the vehicles can use both liquid or gaseous fuels and grid electricity. One of the attractive features of the plug-in hybrid vehicle is that it permits the use of grid electricity generated using energy sources other than petroleum. This paper is concerned with the design and performance of electric battery powered, charge sustaining and plug-in hybrid vehicles using engines, and fuel cell powered vehicles using hydrogen. Of particular interest will be how the electrical energy storage units can best be used in the various drivelines (the component configurations and control strategies) and the resultant powertrain 806 Proceedings of the IEEE Vol.95,No.4,April /$25.00 Ó2007 IEEE

2 efficiency and energy use (fuel and grid electricity) for various driving cycles and use patterns of the vehicles. The different design approaches are evaluated using detailed simulation results and when available, vehicle test data. The characteristics of the energy storage and fuel cell components used in the vehicle simulations correspond to the present status of those technologies as well as projected future improvements in their performance. II. BATTERIES AND ULTRACAPACITORS FOR ELECTRIC AND HYBRID VEHICLES A. Energy Storage Requirements for Different Vehicle Designs The electrical energy storage units must be sized so that they store sufficient energy (kwh) and provide adequate peak power (kw) for the vehicle to have a specified acceleration performance and the capability to meet appropriate driving cycles. For those vehicle designs intended to have significant all-electric range, the energy storage unit must store sufficient energy to satisfy the range requirement in real-world driving. In addition, the energy storage unit must meet appropriate cycle and calendar life requirements. These requirements will vary significantly depending on the vehicle driveline (battery or fuel cell powered or engine hybrid-electric) being designed, but they are reasonably straightforward to determine once the vehicle performance targets are established. It is much more difficult and less straightforward to establish storage unit requirements for the weight, volume, and cost of the energy storage units. There are clearly upper limits on these characteristics which would preclude the successful design and sale of the vehicles, but setting practical limits to achieve workable designs is rather arbitrary. The approach taken in this paper will be to note where appropriate the weight and volume of the units for stated performance characteristics (Wh/kg, Wh/L, W/kg, etc.) of the various technologies. Cost issues are not considered in this paper. As noted above, the energy storage units are sized by an energy storage and/or power requirement. In the case of the battery powered electric vehicle, the battery is sized to meet the specified range of the vehicle. The weight and volume of the battery can be easily calculated from the energy consumption (Wh/km) of the vehicle and the energy density (Wh/kg, Wh/L) of the battery discharged over the appropriate test cycle (power versus time). In most cases for the battery powered vehicle, the battery sized by range can easily meet the power (kw) requirement for a specified acceleration performance, gradeability, and top cruising speed of the vehicle. The batteries in this application are regularly deep discharged and recharged using grid electricity. Hence, cycle life for deep discharges is a key consideration and it is essential that the battery meets a specified minimum requirement. In the case of the charge sustaining hybrid-electric vehicle using either an engine or fuel cell as the primary energy converter and a battery for energy storage, the energy storage unit is sized by the peak power from the unit during vehicle acceleration. In most cases for the charge sustaining hybrid vehicle designs, the energy stored in the battery is considerably greater than that needed to permit the vehicle to meet appropriate driving cycles. However, the additional energy stored permits the battery to operated over a relatively narrow state-of-charge range (often 5% 10% at most), which greatly extends the battery cycle and calendar life. In principle, determination of the weight and volume of the battery for a charge sustaining hybrid depends only on the pulse power density (W/kg, W/L) of the battery. However, for a particular battery technology, it is not as simple as it might appear to determine the appropriate power density value, because one should consider efficiency in making this determination. An appropriatevalueofpulsepowerisnotv 2 0 =4R as at that power the efficiency is very low (close to 50%). A more appropriate value of useable peak power of the battery is given by the following expression: P peak ¼ EF ð1 EFÞ V 2 0 =R where EF is the efficiency of the peak power pulse. In this equation, it is assumed the pulse occurs near V 0 and that EF ¼ V pulse =V 0. For an efficiency of 90%, the high efficiency pulse power of the battery is about 1/3 of the V 2 0 =4R value. As will be discussed in the next section of the paper, even using the above expression, advanced batteries designed for use in hybrid vehicles have high power capability making them suitable for use in charge sustaining hybrid vehicles. Ultracapacitors can also be used in charge sustaining hybrid vehicles. In this case, the energy storage unit is sized by the energy storage (Wh) requirement because the energy density (Wh/kg) of ultracapacitors is relatively low (5 10 Wh/kg) and the useable power density is high (1 2 kw/kg). The useable power from the ultracapacitors can be estimated using the following expression: P peak ¼ 9=16 ð1 EFÞV 2 0 =R where EF is the efficiency of the peak power pulse. In this equation, it is assumed that peak power pulse occurs at a voltage of 3/4 V 0 and the efficiency is given by ð1 IR=3=4V 0 Þ,whereI ¼ P peak =3=4V 0. Specification of the energy storage requirement is critical to the design and practicality of powertrain systems using ultracapacitors. As discussed later in the paper, the Wh requirement is highly dependent on the strategy used to control the discharge/charge of the ultracapacitor in the Vol. 95, No. 4, April 2007 Proceedings of the IEEE 807

3 hybrid-electric powertrain. Storage specifications in the range of Wh seem reasonable for mild hybrid vehicles. The corresponding weight of the ultracapacitor units would be kg with peak power between kw. The round-trip efficiency of the units at these powers would be 90% 95%. The ultracapacitors would be periodically deep discharged when required to meet the driving conditions, but would operate at shallower depths of discharge much of the time. The cycle life requirement for the ultracapacitors in the mild hybrids would be in excess of cycles. Sizing the energy storage unit for plug-in hybrids is more complex than for either battery powered or charge sustaining hybrids. This is the case because of the uncertainty regarding the required all-electric range of the vehicles or even what is meant in detail by the term Ball electric range.[ In simplest terms, all-electric range means that the hybrid vehicle can operate as a battery powered vehicle for a specified distance without ever operating the engine or fuel cell. In this case, the power of the electric drivesystemwouldbethesameasthatofthevehicleifit had been a Bpure EV[ and the energy storage requirement (kwh) would be calculated from the energy consumption (Wh/km) and the specified all-electric range. Hence, for large all-electric range, the battery would likely be sized by the energy requirement and for short all-electric range, the battery would be sized by the power requirement. For a particular vehicle design, careful consideration would have to be given to optimizing the battery design (energy and power characteristics) to meet the combination of the energy storage (kwh) and power (kw) requirements. With all battery chemistries there are tradeoffs between the energy density and useable power density of the battery. This will be clear in the next section in which the characteristics of batteries are considered in detail. For relatively short all-electric range of 15 km or less, a combination of batteries and ultracapacitors is a possible design option. To further complicate the issue of battery optimization for plug-in hybrids, the concept of Ball electric range[ can be interpreted to be mean that most of the driving is done using the battery and assist from the engine or fuel cell would occur infrequently only when the power demand is high and/or the vehicle speed exceeds a specified value. The result would be that most of the energy to power the vehicle would be provided by the battery and effective fuel economy could be very high (100 mpg or higher). In this way, the power demand from the electric driveline (electric motor and battery) would be less than that for the vehicle to operate as a Bpure EV.[ The energy consumption (Wh/km) would also likely be reduced. Hence, both the energy and power requirements of the battery would be less demanding resulting in a smaller, less costly battery for the same effective all-electric range. In the case of plug-in hybrids, the battery will be recharged both from the engine or fuel cell and from the wall-plug. The attractiveness of the plug-in hybrid is that a significant fraction of the energy to power the vehicle will be grid electricity generated using energy other than petroleum. Hence, for plug-in hybrids, battery cycle life becomes an important issue. The battery will be recharged from a low state-of-charge (after deep discharges) more often than for the battery powered EV. As a result, the battery cycle life requirement for plug-in hybrids will be more demanding than for the Bpure EV.[ A minimum of cycles will be required. Hence, both in terms of power and cycle life, the plug-in hybrid application is more demanding for the battery than the EV application. B. Status of Battery and Ultracapacitor Technologies In this section, the status of battery and ultracapacitor technology is reviewed. In the case of batteries, the technologies considered are sealed lead-acid, nickel metal hydride, and lithium ion. For ultracapacitors, only carbon/carbon double-layer devices are considered, because to date that technology is the only one that has been commercialized. 1) Batteries: Most of the battery powered and hybrid vehicles tested and marketed to date (2006) have used nickel metal hydride (NMH) batteries. The development of lithium ion batteries has progressed to the state that strong consideration is being given to the use of those batteries in both electric and hybrid vehicles. Much of the recent battery development has been concerned with high power batteries for hybrid-electric vehicles (HEVs) and not high energy density batteries for electric vehicles (EV). As discussed in the previous section, the batteries for HEVs are sized by the power requirement with much less emphasis on energy density. Batteries for plug-in hybrids require both high power capability and high energy density. Much less work has been done to develop batteries for plug-in hybrids, but it is likely their characteristics will be intermediate between those of the EVs and HEVs. A summary of battery characteristics for EV and HEV applications is given in Table 1. The data were gathered from a number of sources [1] [7]. In general, the information for the HEV batteries is more recent than that for the EV batteries because most recent battery development has been directed toward HEV applications and not EV applications. It is apparent from the table that batteries for HEVs are quite different than those for EVs in several ways. First, the cell size (Ah) of the EV batteries is considerably larger than that of the HEV batteries. This is necessary because the voltages of the two systems are comparable, but the energy stored in the HEV storage unit is much smaller than that in the EV unit. It is also apparent from Table 1 that the power capability of the batteries designed for HEVs is much higher than those designed for EVs. This requirement follows directly from the lower weight of the HEV batteries and the need to transfer energy in and out of the HEV batteries at high efficiency. 808 Proceedings of the IEEE Vol.95,No.4,April2007

4 Table 1 Characteristics of Various Technologies/Types of Batteries for Use in Vehicle Applications (EV and HEV) As discussed previously, high power capability requires that the resistance of the battery be low. Hence, knowledge of the resistance of a battery is critical to the ability to assess its power capability. Note also that the energy densities of the HEV batteries are significantly lower than that of the EV batteries of the same chemistry. For example, a lithium ion EV battery would have an energy density of Wh/kg and that of a HEV battery would be Wh/kg. The tradeoff between energy density and power density is a key feature in optimizing batteries for particular vehicle applications. None of the batteries currently available have been designed specifically for plug-in hybrid vehicles (PHEVs). Ideally, such a battery would have an energy density close to that of an EV battery and the power capability close to that a HEV battery. The cell size (Ah) of the PHEV battery will be smaller than for EVs because the energy stored will be less by a factor of 3 4 for most designs. The PHEV batteriesmustbedesignedasdeepdischarge,longcycle life batteries rather than shallow discharge batteries like those in HEVs. Hence, it seems likely that PHEV batteries will have energy density characteristics closer to the EV batteries than HEV batteries, but with higher power capability than the larger EV cells. The key issue will be increasing the power of the EV batteries with a minimum sacrifice in cycle life. This will be particularly an issue for PHEVs with relative short all-electric range. 2) Ultracapacitors: Ultracapacitors for vehicle applications have been under development since about Most of the development has been on double-layer capacitors using microporous carbon in both of the electrodes. From the outset of that work, the twin goals were to achieve an energy density of at least 5 Wh/kg for high power density discharges [8]. The life cycle goal was at least deep discharge cycles. In order to justify the development of ultracapacitors as a distinct technology separate from high power batteries, it is critical that their power and life cycle characteristics be significantly better than the high power batteries because the energy density of the capacitors will be significantly less than that of batteries. Recently, there has been considerable research [9] [11] on ultracpacitors that use pseudocapacitive or battery-like materials in one of the electrodes with microporous carbon in the other electrode. This is being done to increase the energy density of the devices. There are presently commercially available carbon/ carbon ultracapacitor devices (single cells and modules) from several companiesvmaxwell, Ness, EPCOS, Nippon Chem-Con, and Power Systems [12] [14]. All these Vol. 95, No. 4, April 2007 Proceedings of the IEEE 809

5 Table 2 Characteristics of Carbon/Carbon Ultracapacitors companies market large devices with capacitance of F. These devices are suitable for high power vehicle applications. The performance of the various devices is given in Table 2. The energy densities (Wh/kg) shown correspond to the useable energy from the devices based on constant power discharge tests from V 0 to 1/2 V 0. Peak power densities are given for both matched impedance and 95% efficiency pulses. For most applications with ultracapacitors, the high efficiency power density is the appropriate measure of the power capability of the device. For the large devices, the energy density for most of the available devices is between Wh/kg and the 95% power density is between W/kg. In recent years, the energy density of the devices has been gradually increased for the carbon/carbon (double-layer) technology and the cell voltages have increased to 2.7 V/cell using acetonitrile as the electrolyte. The present performance of ultracapacitors is suitable for use in mild HEVs using either engines or fuel cell as the primary energy converter. By mild hybrid is meant designs in which the power rating of the engine or fuel cell is large enough to provide satisfactory vehicle performance even if the energy storage unit is depleted. The ultracapacitor unit would be sized based on the energy storage requirement ( Wh). The power density capability of ultracapacitors is such that the maximum power capability of the Wh unit will exceed the electrical power requirement for the driveline system. Ultracapacitors are not suitable for use in PHEV vehicles as the primary energy storagetechnology,butcouldprovetobevaluablecombined with batteries for PHEV designs with short allelectric ranges. In those cases, the battery pack would be so smallthatitcouldnotaloneprovidetheelectricalpower required to accelerate the vehicle or recover all the available energy during braking. It is unlikely that ultracapacitors would be used in EVs. III. SIMULATION RESULTS FOR ELECTRIC AND HYBRID-ELECTRIC VEHICLES A. Basis of Comparisons of Vehicle Design Options Simulation results are presented in the following sections of the paper for battery powered EVs, charge sustaining hybrid-electric vehicles using batteries and ultracapacitors, plug-in hybrid vehicles using batteries, and hydrogen fuel cell powered vehicles. All the vehicles will be designed to have the same acceleration performance so that the key basis of comparison of the design options will be the total energy consumption per kilometer (efficiency) and what fraction of that energy can be grid electricity for the plug-in hybrid designs. The engine powered vehicles will be compared on the basis of effective fuel economy (mpg). Engine and fuel cell powered vehicles will be compared based on their gasoline equivalent fuel economy usinggasolineandhydrogenasthefuel. B. Simulation Approaches and Tools A number of vehicle types and design options have been simulated using the NREL Advisor program [15] for the hybrid vehicles and the SIMPLEV program [16] developed at INEL for the battery and fuel cell powered vehicles. Severalsimulationresultsforfuelcellpoweredvehicles 810 Proceedings of the IEEE Vol.95,No.4,April2007

6 Table 3 Characteristics of Electric Vehicles Using Lithium Ion Batteries obtained using the UC Davis Fuel Cell Vehicle Modeling program [17] are also presented. When appropriate, the results obtained using the different simulation tools are compared with each other and other simulation results in the literature. In all cases, the comparisons are made for the same vehicle characteristics (test weight, C D A, and rolling resistance coefficient). The results are dependent on the assumptions made concerning the characteristics of the driveline components and the control strategies for their operation. This is especially true of the engine powered hybrid-electric vehicles. The details concerning the modeling are available in the references that are cited and are not repeated here. Key driveline parameters, such as energy stored (kwh) and electric motor and engine/fuel cell power (kw) are cited along with the vehicle simulation results. C. Battery Powered Electric Vehicles and Plug-in Hybrids The first sets of simulation results discussed will be those for battery powered (EV) and plug-in hybrid (PHEV) vehicles. Both of these design options permit the substitution of grid electricity for all or most of the liquid fuel used by conventional ICE powered vehicles. Battery powered vehicles are necessarily range limited and the recharge times for the batteries are long (6 8 h). The range limitation of battery powered vehicles is overcome by the plug-in hybrids. Simulation results for two types of EVs using lithium ion batteries are given in Tables 3 and 4. Both vehicles have ranges of about 240 km, which is the maximum reasonable for EVs even using lithium ion batteries (140 Wh/kg). Shorter range vehicles could be designed with smaller battery packs. Next, consider simulation results for a plug-in hybrid obtained using Advisor. This vehicle can operate as an electric vehicle down to 20% state-of-charge and then as a charge sustaining parallel hybrid. The lithium ion battery pack in the plug-in hybrid is about 1/3 the weight of the pack in the electric vehicle. The simulation results for a compact-size passenger car are summarized in Fig. 1 and Tables 5 and 6. The simulation results for the plug-in hybrid vehicle indicate that their effective fuel economy (gasoline only) can be very high even for long daily driving ranges and that as a result they can use grid electricity for a large fraction of the miles traveled in place of gasoline. This can be accomplished with a relatively small battery pack (98 kg in the case of the compact car simulated). D. Charge Sustaining Hybrid Vehicles Using Batteries and Ultracapacitors Simulations have also been performed for hybrid vehicles that have essentially zero all-electric range. The electric driveline is used to permit more efficient operationoftheengineandtorecoverenergyduringbraking.in these vehicle designs, the engine operates in an on off mode, but it is not off for long periods of time as would be the case in a plug-in hybrid. The electrical energy storage unit is not recharged from the grid, but is maintained in a specified range of state-of-charge by the motor/generator using power from the engine. Hence, these hybrid vehicles use only gasoline (or other fuel). The energy storage unit and the energy stored in it are small and the unit is sized by the power required from it. The energy storage unit can be batteries or ultracapacitors. At the present time, all the hybrids of this type being marketed by the auto companies utilize nickel metal hydride batteries. Proto-type vehicles using lithium ion batteries are being tested and may be marketed in the near future. Consideration is being given to the use of ultracapacitors in charge sustaining hybrids, but only limited work has been done incorporating them in actual hybrid vehicles [18] [20]. Simulation results will be given for Table 4 Simulation Results for Electric Vehicles Using Lithium Ion Batteries Vol. 95, No. 4, April 2007 Proceedings of the IEEE 811

7 Fig. 1. Advisor output for a plug-in hybrid on the FUDS (compact passenger car). charge sustaining hybrids using both batteries and ultracapacitors. A key consideration in designing such vehicles is the maximum power capacity of the electric driveline (motor) and the corresponding size (power) of the engine. If the power of the electric motor is relatively large (50 kw or larger), then the engine can be downsized Table 5 Simulation Results for the Plug-in Hybrid on the FUDS Driving Cycle and the hybrid is referred to as a Bfull[ hybrid. If the poweroftheelectricmotorissmall(lessthan20kw), then the engine is not downsized significantly and the hybrid is referred to as a Bmild or moderate[ hybrid. Batteries can be used in either Bfull[ or Bmild[ hybrids, but ultracapacitors are suitable only for Bmild[ hybrids and only if the energy storage requirement is relatively small ( Wh of useable energy). Simulation results Table 6 Simulation Results for the Plug-in Hybrid Compact Car on the Federal Highway Cycle 812 Proceedings of the IEEE Vol.95,No.4,April2007

8 Table 7 Characteristics of the Hybrid Vehicles of Different Designs will be given comparing the fuel economy improvements that can be expected using the various design approaches. In general, the projected fuel economy improvements are larger using the Bfull[ hybrid approach, but the incremental cost of those hybrid powertrains will be significantly higher than the Bmild[ hybrid systems. Hence, comparing the fuel economy improvements between the two approaches is of special interest. The simulation results presented in this paper have been taken from [21] [24] in which the various design approaches to charge sustaining hybrids have been analyzed in detail. The primary conclusions of those studies have been that: 1) fuel economy improvements of 40% 50% can be achieved using the Bmild[ hybrid approach and that the relatively low incremental cost of the powertrain make the system more cost effective than the Bfull[ hybrid approach and 2) the use of ultracapacitors for energy storage can lead to a 10% 15% larger improvements in fuel economy than with batteries, even high power lithium ion batteries. The economic viability of this second conclusion depends on the continued reduction in the cost of ultracapacitors to the range of cents per Farad [25], [26]. Regardless of the design approach, the simulation results indicate that charge sustaining hybrids offer an attractive means of significantly improving the fuel economy of all types (sizes) of vehicles using both gasoline and diesel engines. The conclusions discussed above can be justified using the extensive simulation results given in [21] [24]. First, consider a comparison of Bfull[ and Bmild[ hybrids for various vehicle sizes and engine types. The simulation results and related economic comparisons are shown in Tables 7 10 taken from [21] and [22]. These results show the large improvements in fuel economy that can be achieved in charge sustaining hybrids of various sizes and engine types. The tradeoffs between fuel economy and incremental vehicle cost for full and mild hybrids are summarized graphically in Fig. 2 for various engines. The tradeoff in percentage terms were found to be essentially independent of vehicle class. Note the change in slope of the curves for the Bmild[ (MHV) and Bfull[ (FHV) cases indicating the Bmild[ hybrid designs are more cost effective than the Bfull[ hybrid designs. Next, the use of ultracapacitors in mild hybrid vehicles will be considered. Such designs have been analyzed in [23] and [24]. A key issue in assessing the viability of using ultracapacitors in hybrid vehicles is establishing the energy storage requirement. If the storage requirement is too large, the weight, volume, and cost of the ultracapacitor unit are too high and ultracapacitors cannot compete with batteries, especially lithium ion batteries. The energy storage requirement is critically dependent on the control strategy assumed for the hybrid vehicle operation. In [23] and [24], it is shown that the energy storage requirement can be minimized using a Bsawtooth[ strategy in which the vehicle is operated sequentially in all-electric and engine-dominated modes similar to that in a series hybrid vehicle. This control strategy permits the engine to operate at high efficiency when it is both powering the vehicle alone and recharging the ultracapacitors. A typical Advisor output for a hybrid vehicle simulation using the Bsawtooth[ strategy is shown in Fig. 3. Note that the stateof-charge of the ultracapacitor is cycled between near full charge and 50% in a sawtooth pattern on the FUDS driving cycle. The corresponding operation of the engine is shown in Fig. 4. Note that the engine operates at high efficiency most of the time (the average energy efficiency for this case is 29%). The ultracapacitor unit should be sized such that the duration of the times that the engine is cycled on Vol. 95, No. 4, April 2007 Proceedings of the IEEE 813

9 Table 8 Baseline Vehicle Characteristics Using the PFI Gasoline Engine and off is not too short. The simulations indicate that for the FUDS and Federal Highway cycles, storage of about 100 Wh useable energy in a midsize car is sufficient to keep the on/off time periods at least 30 s. Engine operation during a demanding portion of the FUDS cycle is showninfig.5. Simulations results for midsize passenger car (test weight 1680 kg) are shown in Table 11 for various combinations of ultracapacitor units and electric motors. Note that the fuel economy is more dependent on the power of the electric motor than the size (kg and thus the Wh stored) of the ultracapacitor unit. Hence for a midsize passenger car, the electric driveline should have a peak power of about 30 kw, but the energy storage requirement of the ultracapacitor need not be greater than 100 Wh. The simulation results using commercially available capacitors from Maxwell indicate an average capacitor efficiency greater than 92% on the FUDS and Highway driving cycles even though the capacitors are deep discharged to close to 50% (one-half of their rated voltage). The Bsawtooth[ strategy was developed to minimize the energy storage requirement and is better suited for use with ultracapacitors than batteries because of their lower resistance and higher efficiency. The strategy does result in higher average engine efficiency than other strategies, but it is more demanding on the electric driveline components than the more conventional control strategies for mild hybrids. This is the case because a greater fraction of the energy to power the vehicle is transferred in and out of storage and losses associated with this transfer can negate Table 9 Attributes of Vehicles Using Mild Hybrid Powertrains 814 Proceedings of the IEEE Vol.95,No.4,April2007

10 Table 10 Attributes of Vehicles Using Full Hybrid Powertrains much of the improvement in average engine efficiency achieved using the Bsawtooth[ control strategy. Attaining the maximum fuel economy improvement in mild hybrid vehicles requires that the engine not be fueled when it is not producing torque (in the off mode) and that theeffectiveenginefrictionbeminimizedduringthose periods. The effect of engine friction on the fuel economy improvement was studied in [24]. The results shown in Table 12 indicate that the effect of engine friction can be significant, but the magnitude of the fuel economy improvements remains large. E. Fuel Cell Powered Vehicles It is generally accepted that the most efficient way to use a fuel, especially hydrogen, on-board a vehicle is to convert the energy in the fuel directly to electricity in a Fig. 2. Tradeoffs between fuel economy improvement and incremental vehicle cost for full and mild hybrids and various engines. Vol. 95, No. 4, April 2007 Proceedings of the IEEE 815

11 Fig. 3. Characteristics of the sawtooth strategy outputs on the FUDS cycle. Fig. 4. Engine operating map with the sawtooth strategy on the FUDS cycle. fuel cell. The driveline of a fuel cell powered vehicle is similar to that of a battery powered vehicle with the battery replaced by the fuel cell and a hydrogen storage unit (see Fig. 6). The driveline shown is for a hybrid design in which the fuel cell can be load leveled using a small battery or ultracapacitor much like in a charge sustaining hybrid engine electric vehicle. The energy stored in the hydrogen is much larger than in a battery. For example, storage of 3 kg of hydrogen is equivalent to 3 gallons of gasoline or about 100 kwh. This is much more energy than can be stored in battery for a passenger car. The operating characteristics of the fuel cell are similar to that of a battery in that they are expressed in terms of voltage, current, and resistance. The fuel cell has an open circuit voltage of about 1.2 V/cell and the cell voltage decreases as the current drawn from the cell increases. The voltage current characteristic (V versus A/cm 2 ) of a proton exchange membrane (PEM) fuel cell [25], [26] operating on hydrogen and air is shown in Fig Proceedings of the IEEE Vol.95,No.4,April2007

12 Fig. 5. On/off engine operation using the sawtooth strategy on the FUDS. Table 11 Summary of Simulation Results Using the BSawtooth[ Strategy and Ultracapacitors in a Mild Hybrid Passenger Car Vol. 95, No. 4, April 2007 Proceedings of the IEEE 817

13 Table 12 Effect of Engine Friction on Fuel Economy Fig. 7. Cell and system characteristics for a PEM fuel cell (V, efficiency versus A/cm 2 ). The cell and system efficiency are also shown on the figure. Note that the cell efficiency is much higher than for IC engines and that it is highest at low power (small currents) rather than at relatively high power as in the case of the engine. It is the high values of cell efficiency that has lead to the expectation that fuel cell powered vehicles will have significantly higher equivalent fuel economy than gasoline fueled vehicles of the same size and performance. A key question concerning fuel cell powered vehicles is estimation of the factor by which their equivalent fuel economy will be higher than that of a baseline gasoline fueled vehicle. Most studies of fuel cell vehicles and related hydrogen demand using those vehicles assume an improvement factor of 2 3. It is of interest to consider whether available test data and simulation results support this assumed improvement. The most relevant test data availableatthepresenttime(2006)areforthehondafcx vehicle [27], which is shown in Table 13. Based on a baseline vehicle fuel economy of 25/34 mpg for the urban/ highway cycles, the improvement factors for the Honda FCX in 2005 are 2.5 on the city cycle and 1.5 on the highway cycle resulting in an average improvement of 2.0. The simulation results for the fuel cell vehicles are reasonably consistent and are only slightly more optimistic than the test data for the Honda FCX vehicle. Hence, it seems likely that fuel economy improvement factors in the range of 2 3 should be achievable. The fuel economy (efficiency) of hydrogen fuel cell vehicles have been simulated in a number of studies [28] [30]. A summary of the results are given in Table 14. It is likely that most fuel cell vehicles in the future will incorporate energy storage (batteries or ultracapacitors) to permit sizing the fuel cell to lower power than needed to meet the peak power of the electric drive system and to recover energy during braking. Energy storage in fuel cell systems is not utilized as in the hybrid engine electric drivelines to improve the average efficiency of primary energy converter (fuel cell or engine), because the fuel cell efficiency is a maximum at a relative small power fraction (see Fig. 7). In the case of fuel cell systems, energy storage is used to recover braking energy and to reduce the size and cost of the fuel cell. IV. SUMMARY AND CONCLUSION The application of batteries and ultracapacitors in electric energy storage units for battery powered (EV) and chargesustaining and plug-in hybrid-electric (HEV and PHEV) vehicles has been studied in detail. The use of both IC engines and hydrogen fuel cells as the primary energy Table 13 EPA Fuel Economy Ratings for the Honda FCV Fig. 6. Schematic of a fuel cell vehicle driveline. 818 Proceedings of the IEEE Vol.95,No.4,April2007

14 Table 14 Fuel Economy Improvement Factors Based on Simulation Results for Hydrogen Fuel Cell Vehicles (Midsize Passenger Cars) converters for the hybrid vehicles was evaluated and compared. The study focused on the use of lithium ion batteries and carbon/carbon ultracapacitors as the energy storage technologies most likely to be used in the future vehicles. The design requirements for energy storage in these vehicle applications have been discussed and compared with the present status of the energy storage technologies. The performance of the EVs, HEVs, and PHEVs was simulated for various vehicle types and driveline designs. Simulation results are presented for energy consumption, fuel economy, and grid electricity usage on the Federal urban and highway driving cycles. The following conclusions can be drawn from the results of the study. 1) The energy density and power density characteristics of both batteries and ultracapacitor technologies are sufficient for the design of attractive EVs, HEVs, and PHEVs. The primary questions concerning these technologies are calendar and cycle life and cost. 2) Battery powered vehicles (EVs) using lithium ion batteries can be designed with ranges up to 240 km with reasonable size battery packs. The acceleration performance of these vehicles would be comparable or better than conventional ICE vehicles. 3) Charge sustaining, engine powered hybrid-electric vehicles (HEVs) can be designed using either batteries or ultracapacitors with fuel economy improvements of 50% and greater. The largest fuel economy improvements can be achieved in Bfull[ hybrids using down-sized engines and relatively large electric motors. These vehicles would use batteries (nickel metal hydride or lithium ion) that are sized by the power demand and are shallow discharged at an intermediate state-of-charge. 4) Plug-in hybrids (PHEVs) can be designed with effective all-electric ranges of km using lithium ion batteries that are relatively small. The effective fuel economy of the PHEVs can be very high (greater than 100 mpg based on gasoline use only) for long daily driving ranges ( km) resulting in a large fraction (greater than 75%) of the energy to power the vehicle being grid electricity. 5) Mild hybrid-electric vehicles (MHEVs) can be designed using ultracapacitors having a energy storage capacity of Wh. Simulation results indicate fuel economy improvements of 40% 50% in city driving using a Bsawtooth[ control strategy. The fuel economy improvement with ultracapacitors is 10% 15% higher than with the same weight of batteries due to the higher efficiency of the ultracapacitors and more efficient engine operation. 6) Hybrid-electric vehicles powered by hydrogen fuel cells (HFCVs) can use either batteries or ultracapacitors for energy storage. Simulation results indicate the equivalent fuel economy of the fuel cellpoweredvehiclesis2 3timeshigherthanthat of gasoline fueled IC vehicles of the same weight and road load. Compared to an engine-powered HEV, the equivalent fuel economy of the hydrogen fuel cell vehicle would be times higher. These fuel economy improvement factors do not include the efficiency of producing the hydrogen from a primary energy source. h REFERENCES [1] A. F. Burke, BCost-effective combinations of ultracapacitors and batteries for vehicle applications,[ presented at the Second Int. Advanced Battery Conf., Las Vegas, NV, Feb. 4 7, [2] K. Ito and M. Ohnishi, BDevelopment of prismatic type nickel/metal-hydride battery for HEV,[ presented at the 20th Electric Vehicle Symp., Long Beach, CA, Nov [3] K. Konecky, BCobasys NiMH energy-storage systems for passenger, commercial, and military vehicles,[ presented at the Sixth Int. Automotive Battery and Ultracapacitor Conf., Baltimore, MD, May [4] N. Fujioka and M. Ikoma, BNickel metalhydride batteries for pure electric vehicles,[ presented at the 15th Electric Vehicle Symp., Brussels, Belgium, Oct [5] J. Kumpers, BLithium ion batteries for hybrid vehicles and new power system supply systems,[ presented at the 18th Electric Vehicle Symp., Berlin, Germany, Oct [6] T. Horiba, BDevelopment of Li-ion batteries with high-power density,[ presented at the 4th Int. Advanced Automobile Battery Conf., San Francisco, CA, Jun [7] K. Nechev, M. Saft, G. Chagnon, and A. Romero, BImprovements in Saft Li-ion technology for HEV and 42V systems,[ presented at the 2nd Int. Advanced Automotive Battery Conf., Las Vegas, NV, Feb [8] A. F. Burke, BElectrochemical capacitors for electric vehicles: A technology update and recent test results from INEL,[ presented at the 36th Power Sources Conf., Cherry Hill, NJ, Jun [9] S. M. Lipka, D. E. Reisner, J. Dai, and R. Cepulis, BAsymmetric-type electrochemical supercapacitor development under the Vol. 95, No. 4, April 2007 Proceedings of the IEEE 819

15 ATPVAn update,[ presented at the 11th Int. Seminar on Double Layer Capacitors, Deerfield Beach, FL, Dec [10] G. G. Amatucci et al., BThe non-aqueous asymetric hybrid technology: Materials, electrochemical properties and performance in plastic cells,[ presented at the 11th Int. Seminar on Double-Layer Capacitors, Deerfield Beach, FL, Dec [11] A. F. Burke, T. Kershaw, and M. Miller, BDevelopment of advanced electrochemical capacitors using carbon and lead-oxide electrodes for hybrid vehicle applications,[ UC Davis Institute of Transportation Studies, Rep. UCD-ITS-RR-03-2, Jun [12] A. F. Burke and M. Miller, BSupercapacitor technology-present and future,[ presented at the Advanced Capacitor World Summit, San Diego, CA, Jul [13] A. F. Burke, BThe present and projected performance and cost of double-layer and pseudo-capacitive ultracapacitors for hybrid vehicle applications,[ presented at the IEEE Vehicle Power and Propulsion System Conf., Chicago, IL, Sep. 8 9, [14] A. F. Burke and M. Miller, BUltracapacitor update: Cell and module performance and cost projections,[ presented at the 15th Int. Seminar on Double-Layer Capacitors and Hybrid Energy Storage Devices, Deerfield Beach, FL, Dec. 5 7, [15] Advisor (Advanced Vehicle Simulator), ver. 2002, National Renewable Energy Laboratory. [16] G. H. Cole, SIMPLEV: A Simple Electric Vehicle Simulation Program-Version 2, EG&G Rep. DOE/ID , Apr [17] R. M. Moore, K. H. Hauer, D. Friedman et al., BA dynamic simulation tool for hydrogen fuel cell vehicles,[ J. Power Sources, vol. 141, pp , [18] R. Knorr, BSupercar-results of a European mild hybrid project,[ presented at the 6th Advanced Automotive and Ultracapacitor Conf., Baltimore, MD, May [19] T. Bartley, BEnergy-storage requirements for hybrid-electric buses,[ presented at the Proc. 6th Advanced Automotive and Ultracapacitor Conf., Baltimore, MD, May [20] R. D. King et al., BUltracapacitor enhanced zero emissions zinc air electric transit busvperformance test results,[ presented at the 20th Int. Electric Vehicle Symp., Long Beach, CA, [21] A. F. Burke, BSaving petroleum with cost-effective hybrids,[ presented at the Powertrain and Fluids Conf., Pittsburgh, PA, Oct. 2003, SAE Paper [22] A. F. Burke and A. Abeles, BFeasible CAFÉ standard increases using emerging diesel and hybrid-electric technologies for light-duty vehicles in the United States,[ World Resource Rev., vol. 16, no. 3, [23] A. F. Burke, BCharacterization of a 25 Wh ultracapacitor module for high-power, mild hybrid applications,[ presented at the Large Capacitor Technology and Applications Symp., Honolulu, HI, Jun , [24] A. F. Burke, M. Miller, and Z. McCaffery, BThe world-wide status and application of ultracapacitors in vehicles: Cell and module performance and cost and system considerations,[ presented at the 22nd Electric Vehicle Symp., Yokahama, Japan, Oct [25] P. Seshadri and Z. Kabir, BSteady state and transient performance capabilities of a PEM fuel cell power plant for transportation ABOUT THE AUTHOR applications,[ in Proc. ASME: 3rd Int. Conf. Fuel Cell Science, Engineering and Technology, May [26] E. J. Carlson, P. Kopf, J. Sinha, S. Sriramulu, and Y. Yang, BCost analysis of PEM fuel cell systems for transportation,[ NREL Rep. SR , Dec [27] EPA Fuel Economy web site, Honda FCV Fuel Economy Data, [28] S. R. Ramaswamy, BUnderstanding fuel cell vehicles: Handbook for FCV workshops,[ ITS-Davis Rep. UCD-ITS-RR-01-08, Dec [29] R. Kirchain and Roth, BTechnical cost analysis for PEM fuel cells,[ J. Power Sources, vol. 109, no. 1, pp , Jun [30] M. Weiss, J. Heywood, E. Drake, A. Schafer, and F. Auyeung, BOn the road to 2020,[ MIT Energy Laboratory, Oct Andrew F. Burke received the B.S. and M.S. degrees in applied mathematics from Carnegie Institute of Technology, Pittsburgh, PA, the M.A. degree in aerospace engineering, and the Ph.D. degree in aerospace and mechanical sciences from Princeton University, Princeton, NJ. Since 1974, his career work has involved many aspects of electric and hybrid vehicle design, analysis, and testing. He was the head systems engineer on the U.S. Department of Energy (DOE)-funded Hybrid Vehicle (HTV) project while working at the General Electric Research and Development Center, Schenectady, NY. While a Professor of Mechanical Engineering at Union College in Schenectady, he continued his work on electric vehicle technology through consulting with the Argonne and Idaho National Engineering Laboratories (INEL) on various DOE electric vehicle and battery programs. He was employed from 1988 to 1994 at INEL as a principal program specialist in the electric and hybrid vehicle programs. His responsibilities at INEL included modeling and testing of batteries and electric vehicles and the technical management of the DOE ultracapacitor program. He joined the Research Faculty of the Institute of Transportation Studies at the University of California-Davis in July He has performed research on and taught graduate courses on advanced electric driveline technologies specializing on batteries, ultracapacitors, fuel cells, and hybrid vehicle design, control, and simulation. He has authored over 100 reports and papers on electric and hybrid vehicles, batteries, and ultracapacitors. 820 Proceedings of the IEEE Vol.95,No.4,April2007