The power boost ultracapacitor stack.

Manufacturer: Maxwell technologies
Cost: $4,800 (liquidation sale)

The LiIon battery I use proved to be a good energy storage, but when it comes to provide power, it is less than I would like to have at my disposal. The vehicle peformance is still good, but not as impressive as I was hoping for. The energy (Watt-hours) the batters possess represent my range, and the power it can deliver (Watts) represent ability to accelerate quickly. When the battery is connected to the power inverter, the vehicle performs comparable or better than stock one, but due to relatively high internal resistance of the battery, its voltage sag is quite deep when the power for acceleration is demanded. Thanks to the flexible inverter which can be programmed to accept low voltage, even if the pack sag to the half of its OCV (Open Clamp, or no load, Voltage) the power is still enough to get brisk acceleration, but at the expense of increased battery current. This stresses the battery and can be avoided if a hybrid pack is used: the power needed for relatively short acceleration is provided by other source which has small storage capacity but can deliver more power than the traction battery. This can be either high power lead acid battery, or a capacitor capable of storing enough energy for acceleration. In either case, internal resistance of this storage must be low, thus there is no large voltage sag. As a result, the traction battery feels very stiff, providing great power on demand. Another big advantage of this approach is ability to absorb the energy from regenerative braking. I've decided to try capacitors as storage.

This concept is not new, and several test vehicles (mainly buses) are equipped with capacitor banks to supplement the battery during peak power demands. I use capacitors made by Maxwell Technologies. Because of very high specific capacitance compared to traditional capacitors, these are called ultracapacitors.

I decided to try the ultracapacitors in favor of the ectra battery for several reasons:

- Does not need to be charged and equalized as a battery (still needs some equalization though, more on that later);
- Has far greater cycle life (few hundred thousands of cycles), so is permanent part of the car rather than disposable item as the battery;
- As any capacitor, can be discharged to 0V and left there for indefinite time (try it with the battery :-) ;
- Can be charged with the same rate as discharged, i.e. absorbs regen current regardless of "SOC" (up to the max permissible voltage);
- Requires no maintenance.
- In my case I had an opportunity to purchase capacitors at deep discount, as this model was no longer in production.

Disadvantages:
- Complicates charging as the capacitor bank has to be disconnected from the traction battery during charging and connected back for driving; this, however, would be true for the boost PbA pack as well;
- High up front cost. I suppose this kind of disadvantage applies to any component in my EV, so I don't take it as such anymore.

Next question: how many do I need? As few as possible while able to charge up to the max voltage system allow (395V in my case), so at least 395/2.5=158 caps in series. Make it 160 to be on the safe side.

What capacitance? Maxwell currently makes ultracaps with two values: 1.8 kF and 2.7 kF. If course, the larger capacitance, the stiffer the pack and longer it provides power, but it is heavier, bulkier and more expensive. Well, since there are really only two options, let's do some math:

The voltage delta during discharge can be determined as  dV=I*dt/C  (dt = discharge time, I = average current, ESR ignored for the moment).

My acceleration battery current starts at 0 and ends up at about 150A for real hard one, do will assume 75A average.
Acceleration time is 10 sec (I'll be going more than 65 MPH by then). 160 ultracaps in series have total capacitance either 1800/160=11.25 F total (for 1.8 kF caps) or 16.875 F total (for 2.7 kF caps).

So the voltage drop will be: dV = 75*10/11.25 = 66.67V for the smaller caps stack, and 75*10/16.875 = 44 V for the large one. The difference is only 22V and the drive system can easily accommodate that. And, larger caps are heavier and more expensive... Well, suddenly decision was very easy to make:  2.7 kF caps happened to be on liquidating sale, but not the 1.8 kF ones. So, soon I was a happy owner of nearly 17F 395V ultracaps stack.

Maxwell PowerCache ultracapacitor. 2.7kF, 2.5V
Trial fit. All 160 ultracaps will be somewhere here.
An aluminum box was welded and shaped to house whole bank right behind front seats.
Rear view. Lid was added and all the caps installed and padded by a pieces of polystyrene cut in right shapes.
Individual caps are interconnected by flexible braided strips with crimped lugs on the ends.
Fragment of the box. End of chain is connected to a cable.
Zoom in view. A bleeding resistors network is visible. This resistor ladder will equalize the stack.
View from the front with open lid. The hinge along front bottom edge of the box is wisible.
This allow to lift it open to gain acces to the battery box underneath.
Connection on the side of the box (raised position). Gas filled lifters are visible - it is not easy to swing open 160 kg box...

To be continued...