Energy Storage

 

Conceptual Design

There are several commercially viable energy storage systems that are being developed for hybrid electric vehicles (HEV's) on the market today. Major advances are being made on almost a daily basis. The reason for this is the fact that the government is subsidizing a large amount of the alternative fuel research going on by many U.S. as well as foreign manufacturers. The types of devices that hold the most promise to solve the energy storage problems are batteries, flywheels, and ultracapacitors. As shown in the plot below, both gasoline and hydrogen have a higher specific energy than the rest of these electrical storage devices. An advantage of HEV’s is that they can use the high specific energy of liquid or gaseous fuels to provide the vehicle with long-range capabilities. Conversely, the HEV can use the high specific power of electrical energy storage to provide the peak power requirements.

Figure 1: Plot of energy versus power for various energy storage devices.*

Flywheels store energy mechanically by turning a heavy rotor at ever-increasing revolutions per minute to store kinetic energy. Flywheels are good because they store energy very efficiently but their specific energy is very low at this point in their development compared to other available products making them an inappropriate choice for energy storage in our HEV.

The next possible technology is ultracapacitors. These devices work by accumulating and separating unlike charges. The positive abilities lie in the fact that they have no moving parts as well as the number of times that they can be cycled through their charge discharge cycle is very high. Ultracapacitors shortfalls arise from the fact that they have low specific energy in addition to the fact that those commercial versions that are available are in their infant development stages. In the future they may be a very viable solution but just like flywheels they still need to improve in both price and availability.

The final device is one that the team has chosen to store energy in our car. That device is the battery. Batteries are the maturest technology to date that is available right now for implementation into our HEV. Our choice of batteries in itself is a diverse subject. There are hundreds of types of batteries that are available to do the job, but finding the right battery that matches all of our systems requirements is a daunting task. The best method for speeding the process to a final decision on battery choice is to limit that choice as much as possible. There any many battery technologies currently being developed, with Lead-Acid (Pb-acid), nickel cadmium (NiCd), and nickel metal hydride (NiMH) being the most promising technologies. For the purposes of comparing the advantages and disadvantages of the three different types of batteries, the graphs below show the specific energy, specific power, specific cost, and cycle life.  Figure 2 shows the specific energy for current battery types and their potential performance based on estimates from industry experts.

Figure 2: Differences in specific energy for the three main battery types.*

Specific energy is important because it affects the number of batteries, and hence the mass of batteries, that a vehicle needs to carry on-board to attain a certain electric-only range. It is perhaps the most important factor for EV’s because it determines their total range, but not as critical for HEV’s which carry most of their energy in the concentrated form of a liquid or gaseous fuel. As you can see in the above charts, NiMH batteries currently have the most energy per unit mass of the three battery chemistries, and in the future, are projected to maintain a specific energy of over 2 times that of lead-acid and NiCd batteries.

 

Figure 3: Differences in specific power for the three main battery types.*

Specific power, in the units of power per mass, is perhaps the most important parameter measuring electrical energy storage requirements for hybrid electric vehicles. Because hybrids normally depend on the electric energy stored on-board to provide the power for accelerations and hill climbs, the higher the power for less mass, the better. Hence, high specific power batteries are critical to the success of HEV’s. The graph shows how today's lead-acid batteries have a slight edge in specific power, but in the future, it is anticipated that bi-polar lead-acid batteries will have twice the specific power of both NiCd and NiMH batteries.

Figure 4: Differences in specific cost for the three main battery types.*

The definition of specific cost is the cost per unit energy (kWh) of battery, and is related to the economic viability of the battery. Because lead-acid have been around the longest and are the most fully developed battery technology, they have the lowest specific cost at around $125/kWh currently, with projections to decrease that cost to $75 with bi-polar batteries. Currently, NiCd and NiMH batteries are 4 to 5 times more expensive than lead-acid. Once these batteries become standard, their price will decrease. However, due to the raw material cost, they will never be as low as lead-acid.

Figure 5: Differences in cycle-life for the three main battery types.*

The cycle life is a measure of how long the battery will last before it needs to be replaced. Historically lead-acid batteries have had a relatively short life when deep-cycled, while NiCd and NiMH have 3-4 times the cycle life of lead-acid.

Final Design Decision

Limiting our choice of batteries has enabled us to save considerable time. In talking to our team members as well as EV owners and distributors we discovered several limiting factors to our choices. These factors include weight, geometry, availability, and system integration ability. When we started the initial design of this HEV we decided to study currently running vehicles that we could benchmark our own decisions from. After seeing AC Propulsions T-Zero EV we realized that our car was going to be very similar and therefore it would be a good benchmark for us to work from. Collecting a large amount of energy storage data from this car proved to be invaluable. We interviewed the creators of the car extensively and were able to find out the advantages and disadvantages of their energy storage design approach. We found out what they would change if they could start over again and used this powerful information for the foundation of our own design and component choices.

Our first major constraint was the weight of the battery pack. The AC Propulsions car used 28 Optima sealed lead acid Batteries for a combined weight of 1200 pounds. With standard engineering practice in mind the team decided a new battery weight goal of 25% less (900 pounds) would be acceptable for a new upper end goal. In addition to that 900-pound weight we decided that a 700-pound weight would be our ideal number to achieve with 500 pounds or less being the "dream number".

The second major constraint was geometry. Being that the body of our HEV is a production kit car body it confines us to the usable geometry possibilities. Initially we saw that the Riot had two excellent possibilities for battery placement. Those positions were along the side rails of the car on both sides as well as the possibility of a battery storage compartment located under the driver and passenger compartment. We believed that retrofitting these two areas with the proper mounting brackets and wiring distribution capabilities would yield excellent load distribution as well as impact protection. The answer for our battery storage eventually did come with the design of a battery storage box that would fit under the car.

The third constraint to our battery choice was system integration ability. This factor includes things like the battery bus voltage of 336 volts that is necessary to run the AC Propulsion motor and controller unit. It turned out that after extensive dialogue with the engineers at AC Propulsion that we discovered their entire battery charging system and algorithms are based all on lead acid technology. This ended up being an extremely valuable constraint for our project. It limits our choices but simplifies them as well. Researching batteries led us to other important factors to be aware of in order to make a sound choice in battery selection; these being specific energy, specific power, and Ampere-Hour ratings of the batteries. These factors when combined and rated for importance will confine the final battery choice. The cost and cycle life are not in our list of accountable characteristic parameters for battery choice. This is because upon discussion with Dr. Burns early on in the project we decided that due to the fact that this is a prototype vehicle and test bed we would not have to worry about cost of ownership of the vehicle thereby dismissing the cost and cycle life from our decision tree. We did keep in mind the fact that we needed to choose a battery that was not so hyper exotic in its technology that we did not have the knowledge, capability, or time to make our choice successful. Our final battery choice did reflect this fact in making a safer choice to facilitate our success in this project.

The fourth design constraint for the project is component availability. Initially we knew all battery types were available in at least some small production quantities. This ended up being a frivolous belief. We kept finding very attractive choices that met nearly all of our constraints like Lithium Polymer batteries as well as advanced Nickel Cadmium batteries. Unfortunately, we found out they are not commercially available to us and would be impossible to get in accordance with our production timeline.

We found an excellent possibility with the Nickel Metal Hydride Batteries from the Ovonics corporation. At this point we are still investigating their possibility for implementation at a later date. Ovonics offers the only batteries in the industry that were specifically designed and built for Hybrid Electric Vehicles. This is important because HEV batteries due not have to be capable of the same characteristics that purely EV batteries do. There are three major differences in EV and HEV design, which are:

    1. HEV's require higher power capabilities on both the discharge and the charge
    2. HEV's need to be able to withstand many more charge discharge cycles than EV's.
    3. HEV's have a much smaller need of energy storage because of the onboard APU and alternator.

Battery Choices

The culmination of battery research to this point has resulted in two primary battery choices. Although both choices come from the same manufacturer it is important to note that each serves a different purpose in HEV design. The battery that will be used for a preliminary testing in the vehicle will be the Genesis sealed lead acid batteries from Hawker Energy.

Figure 6: The Genesis EP series batteries from Hawker Energy.

Each cell is 12 volts and is rated at 26 amp-hours, which at this point seems adequate but this could change. This possible change in amp-hour size represents the reasoning behind our secondary battery choice. The secondary choice for our vehicle is the Genesis 42Ah cells. This battery offers the same flexibility as its smaller brother except for its increased size of energy storage. It is appropriate at this point to comment on some of the characteristics that make Genesis batteries a superior choice than that of other manufacturers. One of the details that we noticed about the Genesis series batteries is that the internal heat generation within the cells as a function of recharge rate is nearly linear. The data in figure 6 was supplied by the Hawker Corporation shows their test results that prove this fact. In general high heat within and around a cell kills battery life very quickly.

Figure 7: Temperature versus recharge rate for the Hawker batteries.

What that this means is that proper temperature control within the battery compartment needs to be applied. This becomes far easier to control with charts like these so can predict upcoming spikes in battery temperature and begin pre-cooling the compartment. In addition to the easily traceable temperature range of the Genesis batteries there is a long list of reasons to use these batteries in electric and hybrid electric vehicles. The following is a shortened list of the positive reasons to put Hawker Genesis batteries in our vehicle. In an attempt to explain these points as clearly as possible the following was taken from literature provided by Hawker Energy.

Figure 8: Internal resistance versus state of charge.

 

In addition to these incredible characteristics that the Genesis batteries offer we need to look at how well they fulfill the design constraints listed in the beginning of this section. It is important to review how these two battery choices have done in meeting our set requirements.

The first and most important constraint was the weight of the battery pack. Our first choice ended up meeting this goal rather well. With 28 cells at 12V apiece needed to achieve the 336V necessary to run the motor and controller and a single 26Ah battery weighing 22.3 pounds it equated to 624.4 pounds. Our secondary choice of the 42Ah size equated to a battery pack that weighs 938 pounds. Although it is above our acceptable upper limit it must be noted that this choice is in place only to be available in a wworst casescenario situation. What I mean by worst case scenario is that the 26Ah-size battery is already considered to be a conservative choice from initial calculations. Initial testing of the vehicle will hopefully show that we can actually shrink the battery size to the 13Ah range if possible.

The second given constraint was that of geometry. The ability for our battery choices to fit within the battery box under the driving compartment is obviously an important detail. The dimensions of our battery box are roughly 63 x 50 x 7 inches. With seven inches being the critical dimension it is important that at least one dimension of both our battery choices was under that.

 

Specifications of the Hawker Batteries:

Product

Part #

Internal Resistance @ 25° C

Nominal Short Circuit Current

Length, in.

Width, in.

Height, in.

Weight, lb.

G26EP

0765-2001

5.0 mmilli-ohms/b>

2,400 A

6.565

6.920

4.957

22.3

G42EP

0766-2001

4.0 mmilli-ohms/b>

2,600 A

7.775

6.525

6.715

32.9

 

As you can see from the geometry specifications from both the 26 and 42Ah batteries, it will not be a problem getting all of our batteries for both choices inside the battery box. The open ended in housef this battery box allows for future types of batteries that are not available yet to be implemented in our car in the future.

The third constraint was based on availability of the product. Luckily for us, as stated above, both choices are commercially available. We currently have the 26Ah batteries in-house and are awaiting continued fabrication so we can integrate them soon. As a note it took only two weeks to order and receive our batteries at a total cost of less than $2,500. In comparison the Ovonics Nickel-Metal Hydride batteries cost $60,000 that includes a 60% educational discount and a six month lead time.

Lastly came the importance of system integration ability. This means how easily these batteries will interface with the rest of the car and the separate subsystems. Overall they should both work rather well. The 28 batteries will supply the needed 336V to the AC controller at a fairly constant rate. The recharge controller system was designed specifically for lead-acid batteries making that whole area of battery control fairly complete. Aside from the AC motor controller package LabVIEW also has all the necessary current, voltage, and temperature sensors too properly track and adjust the entire energy storage system. The information listed above as well as the fact that they come recommended to us by both AC Propulsion for use on there charging system and other EV manufacturers has proven there potential satisfactory for use in our HEV.

 

Optimization

The optimization section for the vehicle energy storage is difficult in a sense because we have no way of changing a manufactured battery. We can however do our best to make sure that the control system takes care of our batteries as well as possible. This proper care begins with an accurate description of common driving cycles, the power (kW) loads associated with these driving cycles and finally the effect these road loads have on the energy storage system at a real time level. From this information we can build and more importantly test an efficient energy control scheme.

The beginning of this description began with a need for standardized driving cycles so we could accurately represent realistic road loads that our vehicle might see. We thought it would be an easy thing to do, but finding assorted federal driving cycles in a usable numeric array proved to be time consuming. Driving cycles are just standardized descriptions of velocity versus time for an average driving cycle. The two that we use the most being the aggressive driving cycle and the federal urban driving cycle are shown below respectively.

Figure 9: Aggressive Driving Cycle.

 

Figure 10: EPA Urban driving cycle.

Having just these driving cycles as well as many others like them is just the first step in shaping a viable energy storage plan. The next step was to develop a cycle digester that would take the given information from the cycles and supply a point by point description of the loads that the car was experiencing. The software that we used was LabVIEW. It was used for two reasons. The first was that we all had a fairly good understanding of how to use it and, secondly, whatever we developed could be directly integrated with the cars future control system. Developing the necessary virtual instruments in LabVIEW was slow going at first in an effort to collect the most accurate equations to describe road loads. The equations that we ended up using came from the NASA Lewis Research Center hybrid electric bus project. These equations were the only readily available that had been developed for design purposes and then tested for accuracy against the final product. The equations used as well as a brief description is listed below.

Power loss due to aerodynamic drag:

Rolling Resistance Losses:

Road Inclination:

Power for Acceleration:

Parasitic Losses:

Power from Regenerative Braking:

Transmission Efficiency:

Taken to be 96% from the advisement of the transmission design team.

Kilowatt load of given scenario

Gross vehicle mass (kg)

Vehicle frontal area (m^2)

Coefficient of drag of vehicle

Velocity of vehicle (m/s)

Vehicle acceleration (m/s^2)

Density of air (kg/m^3)

Rolling resistance coefficient

Angle of road incline (degrees)

Gravity (m/s^2)

Efficiency of regenerative braking (%)

Inputting the given variables into the road load equations gives a separated description of individual loads. To combine the separated loads we inputted them into the equation below.

The output of this equation gives the kW road load at a given interval of time. When used in conjunction with LabVIEW the completed program that contains these equations combined with the imputed times and velocities of a driving cycle delivers an instantaneous description of the power required to move the vehicle. This piece of information is very valuable because it is the foundation for the entire vehicle dynamics program that we are developing for this project. The vehicle dynamics program also contains the state-of-charge (SOC) calculator for the batteries that was the final piece to the puzzle for accurately describing the power losses and gains within the vehicle.

The SOC calculator makes use of another equation used by the NASA hybrid bus program. We contacted Dale Stalnaker at NASA who was responsible for writing their original program that uses these equations. He said that the equations all worked well giving consistently reasonable results. The SOC equation used in our program is:

The equation makes use of the instantaneous power in kilowatts needed for the t instant in time. These two inputs are directly linked to the cycle analysis program that was described above allowing for accurate comparison between velocity and power demand. The C term in the denominator is the Ampere-hour capacity of the individual batteries (in our case 26) and the V is the system bus voltage 336V. To accurately represent the internal resistance changes and efficiencies within our batteries at any point in their state-of-charge, calibration tables have been integrated. These tables actively account for the difficulty a battery has in accepting a charge as it nears 100% state of charge. This phenomenon occurs mostly in charge because the internal resistance of our batteries during discharge is both low and constant. The value of V in our equation has been reduced by 5% from the true 336V to 322V in order to take into account the efficiency of the pulse width modulator (PWM) and controller that drives the AC motor. The theory behind this is that because the PWM creates a square wave that cannot be railed out to 336V all the time. They have to return to the zero point even when maxed out in order to create the square wave. This instantaneous dip to zero results in roughly a 5% reduction of bus voltage. At this point in time the vehicle dynamics program as well as the state-of-charge calculator are all up and running. Initial results of the program look very promising in supporting our choice in batteries as well as proving our energy storage schemes. For the federal urban driving cycle the Genesis 26Ah batteries are showing a 78.4% SOC upon the completion of the roughly thirty minute cycle. The aggressive driving cycle left the vehicle with a 92% SOC for that ten minute cycle. These initial SOC numbers left us very excited because it proves that the batteries that we chose are very well sized to the energy pool that they need to accept and deliver. If the batteries were completely discharged after three minutes or if the SOC never left 100% over these driving cycles then there would be reason for concern, but fortunately the results look very promising. In fact, as the vehicle dynamics program continues to mature and become more accurate the numbers keep getting better.

* Diagrams provided courtesy of National Renewable Energy Laboratory.

Back to Home

 


HEV Team
Department of Mechanical Engineering
San Diego State University
5500 Campanile Dr.
San Diego, CA 92182-1323
Fax: (619) 594-3599
E-mail: hev@kahuna.sdsu.edu