Fast-charging 6831+ cells

raymond · 06-27-2009, 05:31 AM

Does anybody here have any idea how to fast-charge thousands of Li+ batteries efficiently? All cells having a slightly different capacity do you need to charge each cell individually? Does each cell need its own (tiny) controller?

Fast-charging also needs to work when one cell breaks down (or has a significantly lower capacity because it has partially broken down).

Charging at 1C is easy for one 2.4Ah 1.35V cell. But for thousand of them I think you need some clever trickery. I'm wondering just how much trickery would actually be needed.

WarpedOne · 06-27-2009, 06:01 AM

No trickery. Fast charging is no rocket science, it is dead simple.

Fast charging = high power charging
High power charging = high current
high current = high thermal losses
hot cells = short lifetame

So, all you need for fast charging is:
- fat cables (all the way from power source to the cell)
- high-power power source
- powerful electronics cooling
- powerful battery cooling

Problems are:
- there are no power sources with enough power for fast charging (half hour charge => 200kW, 20 minute charge => 400kW)
- fat cables are heavy and really dificult to handle
- batteries really don't like to charge fast => their lifetime shortens considerably

Whats is wrong with overnight charging?

But you're having problems seeing fast charging is no different from ordinary charging. All the problems you mentioned are there with slow charging.

stopcrazypp · 06-27-2009, 11:44 AM

Quote:

Originally Posted by raymond

Does anybody here have any idea how to fast-charge thousands of Li+ batteries efficiently? All cells having a slightly different capacity do you need to charge each cell individually? Does each cell need its own (tiny) controller?

Fast-charging also needs to work when one cell breaks down (or has a significantly lower capacity because it has partially broken down).

Charging at 1C is easy for one 2.4Ah 1.35V cell. But for thousand of them I think you need some clever trickery. I'm wondering just how much trickery would actually be needed.

The problem you are talking about is cell balancing, which is an issue even in normal charging (but rapid charging makes it worst since it's more stressful on the batteries).

Here's a link with a decent explaination of how they deal with it specifically with li-ion cells:
Cell Balancing and Battery Equalisation

raymond · 06-27-2009, 01:13 PM

Great link, thanks! Indeed the problems for fast charging are no different than for normal charging. (I mistakenly though you could trickle-charge a Li+ pack, but it seems Li+ doesn't like that.)

So it seems there are 6831 tiny controllers somewhere in the PEM or thereabouts. 83 (or so) kW input (assuming a 45 min charge on a 56kWh pack at 90% efficiency) from one cable to be spread out into 6831 12W charge cables does not look like an attractive wiring scheme. I can come up with alternative wiring schemes but am curious about how Tesla does this.

But then again, this may be part of the secret sauce.

PS: 83kW at 400V gives 200+ Amps requiring at least two 12mm (.46") wires between the car and the wall. Two meters of said cable weighs 3951 grams (8.7 lbs) in copper alone! "Hey guys, wanna help me hook up my Tesla Roadster charger cable, please?"

JRP3 · 06-27-2009, 02:34 PM

Quote:

Originally Posted by raymond

So it seems there are 6831 tiny controllers somewhere in the PEM or thereabouts.

I don't think so. Remember a lot of those cells are connected in parallel, which should all self balance. So I think all the paralleled sections can be treated as a single cell when charging, though each individual cell may have a protection circuit on it in case there is a problem that takes it out of spec.