Or lack thereof. Yesterday, when the thermometer showed 0 degrees Celsius (or 32 Fahrenheit), I figured it would be time to go check the cells in my kWsaki electric motorcycle. I took the bike off insurance October 3rd so it's been sitting in the cold garage for about four and a half months. I gave one of the TTL-200C motors a quick spin with the kWsaki traction pack and controller, but I haven't recharged since October.
And when I said cold I meant it. The temperature here has regularly been below zero Fahrenheit and the cells have had to endure it. So I was a little excited when I started the measurements. First cell 3.299 V, second 3.299 V, no make it 3.300 V, next 3.299 V, 3.300 V, 3.300 V and so on. It got really boring really fast. I got a little bit of variance when I couldn't reach the terminals of a couple cells and I had to measure them in pairs. They read 6.61 V. A couple cells showed 3.31 V. This was after I had changed the meter's resolution from 0.000 to 00.00. The whole pack registered at 82.8 volts. For 25 cells that's 3.312 V each.
Considering that the meter is no Fluke and it showed a maximum variance of hundredth of a volt, I don't see evidence of cell drift. I've had nothing connected to the pack for these last months. Note that this is the same pack that I pushed down to 2 volts per cell earlier. If I had damaged any of the cells, they probably wouldn't be happily floating at 3.3 volts after being left to their own devices for months.
Conclusions are that cell drift has not been demonstrated and bottom balancing still seems to work.
If I had more money than I can spend, I could have gotten another 25 cells, put a MiniBMS on them and let them idle for over four months. Alas I do not, but the result is easy to guess. Anyone care to prove me wrong?