Well, I know that the stock batteries are claimed to have 20AH of capacity.
However, before I found myself a long way from home on a dark night with a yellowing headlight, I thought I'd see...
I took a piece of nichrome heater wire from a dead hair dryer, and clipped a length that had about three ohms of resistance. At the nominal 48V that would be expected to draw 16 Amps or so. That would also mean it was dissipating 768 Watts, about the equivalent of a small toaster, so I submerged the wire in a bucket of water. I did this "downstream" of the shunt for my DrainBrain so that I could monitor the current, voltage, and total power used by the setup.
Now, the batteries are rated as 20AH batteries. This means that they are capable of providing 1A for 20 hours. It would be reasonable to assume that such a battery would either discharge 1A for 20 hours or 20A for one hour, but you would be wrong. As the rate of discharge goes up, the amount of energy that can be extracted from a battery goes down, and at 16 A, I would expect about 15AH of capacity.
So, I hooked up the pack to the load. Sure enough, the DrainBrain showed about 15A discharge. There was a steady fizz of bubbles from the nichrome wire. After 15 minutes, I was surprised that the bucket was steaming! It turned out that the heater was at the top of the bucket, and the water stratified until the surface water was almost boiling.
I carefully watched the voltage, and terminated the test when the battery pack got to 44V. Even assuming a perfectly balanced pack, that would mean each battery was at 11V, very close to their discharge limit. If the batteries weren't balanced, a low capacity battery could already be drawn into reverse voltage.
The AH reading was a gratifying 14.4 AH. The biggest shock was the OEM voltmeter's cheery 85% full reading. I'll certainly keep this in mind on long rides! (Even better, I'll be watching the DrainBrain, and avoiding that part of the discharge cycle.)
