hi Smeee,
> One main problem is that capacitors in series divide the farads by amount of caps to an average. EG. 5X@55farad in series = 11 farads which is a bummer really. A good balance system between each leg of the capacitors may improve the situation.
there's nothing to be improved, capacitance is not energy, it's a different property. in your example, the 11 F capacitor made of 5 series caps when charged to 5x the max voltage per cap would store 5x the energy of a single 55 F cap charged to it's max voltage. that is, every cap would store it's max energy when connected in series. there's no loss of energy capacity, the reduction of the capacitance property due to series connection is of no concern in this regard.
the reason is that energy in a capacitor varies with the square of the voltage. the energy stored in a capacitor is:
E=(1/2)*C*V^2
where C is the capacitance (assumed constant) and V is the voltage. thus, the max energy is:
Emax=(1/2)*C*Vmax^2 (1)
if we put n equal caps in series we could consider the thing as being one single capacitor with these parameters:
Cseries = C/n
Vmaxseries = n*Vmax
since the whole combined thing is in itself a proper capacitor, equation (1) applies to it too, so:
Emaxseries=(1/2)*Cseries*Vmaxseries^2
=(1/2)*(C/n)*(n*Vmax)^2
=(1/2)*(1/n)*C*n^2*Vmax^2
=n*(1/2)*C*Vmax^2
=n*Emax
so the max energy of the series thing (when charge to n*Vmax) is n times the max energy of each cap.
> My guess is that Ultra caps is the best answer to the regen efficiency loses
I'm afraid it's not. in LiFePO4s, the energy efficiency (related to but not exactly the coulombic efficiency) is rather high. if you charge some energy into the batt and then discharge til you reach the original state of charge you'll find that you've recovered most of the energy (more than 95%, maybe a lot more). the efficiency depends on the power used to charge and discharge and the batt quality and power rating. (an A123 would barely heat up when others would strain... if you assume only reversible changes and there's no heat, there really can't be much lost energy.)
so what's up with regen? well, GM's HBS-36 isn't terribly efficient. when hooked to the 36V non-regen it peaks at 80% when outputting relatively low torque, but efficiency falls with higher or lower torque. (check this hub motor for comparison running at peak 98% efficiency:
http://www.csiro.au/files/mediaRelease/mr1999/CsiroPowersFourSolarCars.htm) so when you break and then accel, the cycle going twice though the motor, you start with a peak theoretical efficiency of 64%. but when breaking and accelerating, chances are that you'll be using higher torques that what'd be the optimum, and thus right there half the energy is gone. then keep in mind that, all things equal, regen switching strategies are less efficient. so for starters the regen controller, if of the same quality, can't reach 80%.
and then, people in their minds tend to minimize the effect of losses, particularly aero drag. sure, E=(1/2)*m*v^2, we all know a certain energy is needed to accelerate a mass to 35Km/h; but are you aware of how much power is needed to keep it going at that speed? (there are many calculators on the web, try them out.) so many people don't realize that starts on most trips are a small fraction of the energy expenditure, and only that small fraction is there to be rescued by regen. I think this is the biggest factor of all.
so, raising the efficiency of the energy storage system from 95% to say 99% won't do a thing; in fact it'll cause shorter trips due added weight. also, the limited efficiency of the power converter used to charge and discharge the caps would offset your expected increase in storage system efficiency.
in short, if you want longer trips, pack more batteries, not dead weight (well, almost).
