ULTIMATE HYBRID DRIVE SYSTEM

Frank Lev, Tavrima Canada Ltd

December 13, 2000

Environmental concerns and regulatory action combined have stimulated high levels of research toward the development of Hybrid Electric Vehicles (HEV). A hybrid electric vehicle combines an electric drive, an internal combustion engine (or another source of power), and a storage system for electricity with the goal of capitalizing on the best traits of each. Numerous approaches can be taken in HEV design and the technology is currently immature to the point that no clear vision of a “correct” approach has emerged. This paper addresses specific considerations, derived from testing and analyses of the several existing Hybrid Electric Transit Buses (HETB), such as the Orion VI LMCS with batteries, and the NASA HETB with ultra-capacitors. It is intended as an aid for designers of hybrid electric vehicles in further improving fuel efficiency.

A vehicle‘s fuel economy is primarily dependent upon two main operational variables, the average speed of the vehicle and the number of times that vehicle comes to a stop. Acceleration rate, braking distance and the amount of engine idling are also the contributing factors. As such, the drive cycle has a significant effect on measured emission levels and fuel economy. A majority of the substantial fuel economy benefit, derived from today’s HETB, comes from the recovery of kinetic energy through regenerative braking. The amount of regenerative braking energy (RBE) recovered is affected by the cycle on which the vehicle is operating and the design of it’s energy storage system. RBE has the effect of increasing fuel economy by reducing the amount of power to be supplied by the engine within a given time period. One of the main objectives of this paper is to define the ultimate storage capacity, and relying upon the available test results, to identify the most suitable for the RBE technology.

ANALYSES OF THE EXISTING SYSTEMS

The HETB’s, addressed here are of a charge sustaining variety, in that they do not provide an all-electric capability. The storage system of such vehicles is replenished by the engine-generator, to maintain an optimal state of charge. Only an insignificant fraction of RBE is currently being recovered in the existing HETBs. Therefore, the only available source of “free” energy is hardly utilized! Instead, the storage system is replenished primarily by the engine-generator at a loss of potential fuel efficiency. Some fuel savings are gained when the vehicle is at a stop with the engine idling, due to the utilization of the surplus energy produced by the engine and fed back into the energy storage system, via the generator. A transit bus typically dwells (idles) up to 17% of its total cycle time. The best alternative is to turn the engine off, provided there is an alternative way to power the air conditioning and lighting. Ultimately a fuel cell could be used to power such accessories with the engine employed exclusively for the vehicle’s propulsion. The positive impact on the environment, and fuel economy derived from such hybrid technologies, is obvious. The following examples illustrate our conclusions:

Total energy consumption and power requirement for a typical 35,140-lb (16,000 kg) transit bus such as Orion VI LMCS hybrid diesel, designed to comply with the CBD cycle, are as follows:

Power requirement for the same vehicle is as follows:

1. ACCELERATION 0-20 MPH – 170 kW

2. CRUISING @ 20 MPH – 40 kW

3. BRAKING @ RATE OF DECELERATION 6.78 FT/S/S – 280 kW

Parasitic load, as it is obvious from the diagram for CBD cycle would require almost the same amount of energy as acceleration of a 35,140 lb vehicle from 0-20 MPH

KWh/mile ACCELERATION

ROAD LOAD

PARASITIC LOAD

3.0

2.0

1.0

0.0

1.0 VEHICLE ENERGY CONSUMPTION PER MILE

The CBD cycle is typically used to evaluate transit buses and is made up of 14 identical sections containing an acceleration to 20 MPH for 10 sec, a cruising at 20 MPH for 18.5 sec, braking to a stop for 4.5 sec, then idling for 7 sec. The total cycle covers 2 miles over 553 seconds. The total idling (dwell) time is 91 seconds (16.5%).

The Arterial Route cycle is also used for transit buses evaluation and covers 2 miles with 4 stops and a top speed of 40 MPH. It’s four sections contain an acceleration to 40 MPH for 29 sec, a cruise at 40 MPH for 22.5 sec, deceleration to a stop for 9 sec then dwell for 7 sec.

The top speed of the vehicle must be not less than 60 MPH with all accessories operating. The operating duty cycle has also a Commuter Phase cycle covering 4 miles with 1 stop and a maximum speed of 55 MPH. Acceleration to 55 MPH for 90 sec, a cruise at 55 MPH for 188 sec, deceleration to a stop for 12 sec and dwell for 20 sec. All Duty Cycles require the same rate of deceleration 6.78 ft/sec/sec with the grade climbing ability of the bus being 34 MPH on 2.5% grade, 7 MPH on 12% grade with all accessories operating. All the above cycles are based on the Department of Transportation Transit Bus White Book Specification.

The ideal design approach, as we mentioned, would be to use a diesel engine of a sufficient power in tandem with an electric motor. The strategy here is to hold the diesel at the ideal RPM for peak efficiency, whereas transient loads are handled by the electric motor. This hybrid power brings considerable improvements in both fuel efficiency and emissions. Various combinations of engine to motor ratios have been tried with mixed success, allowing us to conclude that the HETB must have a dominating engine presence to comply with the White Book requirements. The motor/storage cannot substitute for the engine power deficiency unless the HETB is designed as a charge depleting variety, which as an “electric vehicle with range extender” is beyond the scope of our analysis.

IDEAL NEW SYSTEMS

An ideal HETB must have a high peak power electric motor and an efficient storage system to receive regenerated kinetic energy with minimum losses. In this respect, supercapacitors are the enabling technology for the storage system, as was proven by NASA in their prototype HETB, which utilized them. The deceleration rate of the CBD cycle requires hard braking in order to fit into the stopping distance of 60 ft. The power required to meet this rate of deceleration is 281 kW at wheels, if road load is ignored. The road load at 20 MPH (32 km/h) will consume approximately 21 kW, therefore the traction motor of the HEV should be capable of 260 kW peak power to meet the CBD cycle’s deceleration rate and to regenerate all kinetic energy. To further illustrate the importance of a properly designed regeneration system let us take a look on the HETB designed by NASA. The vehicle was tested with supercapacitors and, separately, with batteries to determine the regenerating capabilities of each storage technology. Decelerating from the speed of 15.4 MPH the HETB with capacitors covered 103 feet of stopping distance in 10 seconds, and stopped without use of service brakes. With batteries the distance increased to 157 feet and the service brakes were needed to stop the vehicle. The first phase of the test has proven that batteries are not as effective for regeneration as capacitors. Unlike the batteries, the capacitors absorbed almost all the energy regenerated by the motor, providing an electro-dynamic deceleration sufficient to stop the bus. It should be remembered that the CBD cycle allows only 60 ft of stopping distance from 20 MPH. Having a mass of 37600 lb (17 055kg) the NASA vehicle’s traction motor of 150 kW was severely underrated compared to the calculated 300 kW. This lack of motor power manifested itself (in almost direct relationship) in a longer stopping distance and time. This second phase of the test has confirmed that to stop in accordance with the CBD cycle requirements utilizing only regeneration, the motor must have sufficient peak power.

The following calculations confirm these conclusions:

Min deceleration time 32-0 km/h (8.89 m/sec) is specified as: t = 4.5 sec therefore the deceleration (rate) will be (-a) = v/t = 8.89/4.5 = 1.97 m/sec² (-a) = constant. Specific power at wheels required for 32-0 km/h in 4.5 sec deceleration will be as follows:

P_specific = Va² t – a⁴ t² – v²a² = 17.55 W/kg x 16000kg = 281 kW.

Kinetic energy of HETB @ 32 km/h is equal to: ½ m v² = 0.5 x 16000 x 8.89² = 632256 J = 632 kJ

Rolling + Drag power verification

f = f₀ (1 + v²/1500) where f₀ = 0.012, v = 32 km/h

f = 0.012 (1 + 79/1500) = 0.012

P_r = 9.81 f v = 1.1 W/kg. Thus 1.1x16000 = 17.6 kW

P_d = 12.9x10^-6 A v³ = 0.00009675 x 32³ = 3.2 kW (Where A = 2.5m x 3m = 7.5m²⁾

Total roll and drag power = 17.6 + 3.2 = 20.8 kW

At 20 MPH the kinetic energy of a 35,140-lb (16,000-kg) bus is equal to 632 kJ. When energy lost to rolling and wind resistance is considered, the total theoretically recoverable energy is 585 kJ. Therefore, almost 92% of the kinetic energy can be recovered in an ideal hybrid drive on the CBD cycle.

The Arterial Route cycle has 2 times higher than CBD cycle speed, therefore its kinetic energy will be 2528 kJ. The road load losses will be approximately 75 HP (56 kW). To provide the same rate of deceleration as in the CBD cycle without use of service brakes the traction motor has to be 2(2528/9) = 562 kW. The Commuter Phase cycle has maximum speed of 55 MPH. Its kinetic energy will be equal to 4836 kJ. Factoring in substantially higher road load losses of 105 kW and assuming that in theory the traction motor has to be of 2(4836 /12) = 806 kW, then 806 – 105 = 701 kW. The above examples confirm that approximately 90% of kinetic energy is, theoretically, available for regeneration, with subsequent fuel economy improvements. However, the peak power range of the existing traction motors is sufficient only for the CBD cycle. Considered to be one of the best near production HETB, the Orion VI LMCS vehicle has a 186.5 kW maximum power motor. As we mentioned already, this bus theoretically should have had a 260 kW motor to regenerate up to 90% of energy on CBD cycle. With a less powerful motor the theoretical recoverable energy drops from 92% to 66% [0.5(186.5x4.5) = 419.6kJ] with service brakes to be used to dissipate the remaining 26% of the energy. On the Arterial Route cycle the same bus will have to use the service brakes more often for the same rate of deceleration as in the CBD cycle because its motor is much smaller than 562 kW. Therefore the theoretical recoverable energy will be equal to 0.5(186.5x9)=839 kJ with service brakes to be used to dissipate remaining 82% of energy. It appears that the Orion VI could benefit from the CBD cycle the most being able to recover 80% of the kinetic energy, whereas with the AR cycle the recovery goes down to 20%. In order to improve fuel economy of the HEV it is necessary to increase the peak power of its traction motor(s) while reducing the vehicle’s curb weight.

STORAGE SYSTEM CAPACITY

The first line of Table 2 below shows the values of theoretical regenerated energy, without losses in the motor, motor-controller and storage system. The second line shows the values with the motor and controller losses factored in.

TABLE 2

CBD	ARTERIAL ROUTE	COMMUTER PHASE
420 kJ 336 kJ	840 kJ 672 kJ	1119 kJ 895 kJ

It is obvious from Table 2, that to accommodate the regenerated energy of braking the storage system should have maximum usable energy capacity up to 900 kJ. This is many times less than the battery storage system of the Orion VI LMCS. The emphasis should be placed upon efficiency, durability and cost of the storage system. The most enabling technology in this regard is the supercapacitor. Not only it will store almost 96% of the recuperated energy of braking, but it will also return it for use in vehicle acceleration with the same efficiency. It will also endure up to 100 000 complete cycles - sufficient for seven years of continuous usage. A supercapacitor storage system for a bus such as the Orion VI LMCS would be comprised of 14 PSCap-ECOND-300/90 connected in series and parallel. Their total rated capacity would be 1260 kJ to provide the required 900 kJ of usable energy. This system could satisfy either 600V or 900V voltage requirement and will need no forced cooling or power electronics for balancing. The weight of the system should be within 500 kg. (Compared to the existing Orion VI LMCS lead-acid storage system, which weighs about 2200 kg and is capable of accepting only 8% of the 336 kJ [Table 2]). In addition, in this duty cycle, the batteries have a life of less than a year.

ULTIMATE STORAGE SYSTEM CAPACITOR

The most efficient way to connect capacitor cells in series is by using bipolar construction (multiple stacked cells). One major drawback, however, is the resulting length of perimeter seals, which are difficult to make reliably when there are hundreds of large cells stacked in series. PSCap-ECOND supercapacitors are of bipolar construction. Due to years of development, the perimeter sealing methodology has been drastically improved, thus providing products with extremely long service life.

When the single cells are packaged individually, to be further connected in series, there is a cell imbalance problem that has to be overcome. To deal with this problem the cell voltage is reduced to 33%-70% of its theoretical limit, resulting in severe derating of its overall performance. It is very important that the voltage present at an individual cell does not exceed the breakdown potential of the electrolyte. If this occurs, capacitor failure will result. The number of series connected cells and other criteria including desired reliability influences cell voltage derating. The derating is necessary whenever capacitor cells are series connected independent of electrode material, electrolyte or construction. Even for a series-string of theoretically identical cells, voltage derating is required because of variability induced by thermal gradients from dissipated energy.

A supercapacitor storage system for a bus such as the Orion VI LMCS, as we concluded above, would be rated at 1260 kJ. For comparison, let us use a commercial single cell organic electrolyte supercapacitor similar to those manufactured by Maxwell and others. The advertised characteristics of the cell are as follows: 2.3V, 6.5kJ. As we mentioned above the cell’s voltage must be derated for a durable series connection. For a 600V storage system we would need 600/(2.3x0.6)=435 cells. The resulting energy rating will be 435x6.5x0.6² = 1018 kJ with a mass of 410 kg plus supporting hardware. In turn, Tavrima can supply dedicated 600V bipolar supercapacitors ready for parallel installation without the need for external voltage balancing. For instance, Tavrima’s new PSCap-ECOND-600/90 are rated 600V, 90kJ.The total rated capacity of 14 units would be 1260 kJ with a total mass of 490 kg. These capacitors are optimized for the automotive and railway applications and are of a robust construction suitable for installation exposed to the environment. In sharp contrast, the single cell devices would require a protective shroud and forced ventilation, which would add to the systems volume, mass and cost. The reliability of the bipolar capacitors is also a decisive factor, since there are fewer connections compared to series strings of the single cells.

The difference in efficiency between the above capacitors is the most critical factor. Organic electrolyte has a higher breakdown potential than aqueous electrolyte (used in the PSCap-ECOND 600/90 unit). The aqueous electrolyte, however, is much more conductive. The organic electrolyte capacitor is thus more energy dense, but its relatively high internal resistance reduces its efficiency. At high current levels, the difference in efficiency between organic and aqueous electrolyte capacitors is more pronounced. That means more of the rated energy is dissipated as heat, rather than used for the vehicle’s propulsion. For the Commuter Phase the bipolar aqueous electrolyte supercapacitors could store useful energy equal to 1260x0.75x0.96=907kJ, whereas the organic supercapacitors, although having higher capacity, could store just 1018x0.75x0.78=596kJ.

CONCLUSION

One of the main advantages of Electric Vehicles of any type is their ability to regenerate the energy of braking, thus increasing overall efficiency. Speed directly affects the efficiency with which the storage system is able to accept and return the energy flow, and is clearly the most important characteristic of its performance. Due to the short deceleration times, allowed in vehicle stopping, only capacitors are capable of accepting and storing the inrush of energy with minimum losses. The issue of efficiency is so pronounced that only those capacitors that have a RC product less than 1 second can be effectively used. Single cell devices loose their advantage in energy density due to the voltage derating, required for series connection into a balanced storage system with the required characteristics. The difference in energy density between bipolar aqueous capacitors with RC=0.5sec and single cell organic capacitors with RC=1.3sec is significant when it comes to a HETB storage system. The bipolar type also presents more flexibility in the design of storage systems. The storage systems comprised of the bipolar dedicated capacitors have no series connections between the individual units and therefore are free from the inherent weakness of a string of cells, wherein one break causes catastrophic failure of the entire storage system.

Acknowledgement.

The author is very grateful to Mr. David Evans, President of Evans Capacitor Company for his crucial help in writing this paper.

References.

[1] Hybrid-Electric Drive Heavy-Duty Vehicle Testing Project Northeast Advanced Vehicle Consortium, M.J.Bradley & Associates, Inc, West

Virginia University.

[2] SAE-2000 Tavrima supercapacitors prove their worth in Hybrid Electric Vehicle storage systems. Frank Lev, Tavrima Canada Ltd

[3] Hybrid Electrolytic/Electrochemical Capacitors for Electric Vehicles. David Evans, Evans Capacitor Company.