NOTE: This is not a condemnation of LiPO; it has some amazing properties and there is a time and place for it. Just a reminder from the real world to be careful.
I have been doing some experiments with a "LiPO" (Lithium colbalt I think is the correct name but it is the bricks from HobbyKing I am talking about) pack that was left connected to a controller too long and which sucked it way down. When I looked at the cell voltages, some were in the 1 v range, some were around 3 v, and some were in the middle. There were three 6s Turnigy packs in series. Two of the three packs had a single cell around a volt and it was the same cell in each pack (don't remember which, but I think one cell from one end or the other. The same end in each case, but I don't remember which end.). The other brick had all cell voltages around 3. I did an experiment with a balance charger, charging each pack first without the balance and cell level checking to get it so the charger would recognize it as 6s and not give a "fault" and then using the balance feature to get to full cell voltage. It took a very long time (like a day) for the two packs with a cell under a volt to balance, but eventually they did. However, upon sitting for several hours, they would get significantly unbalanced again. I rebalanced them multiple times and each time, they seemed to stay balanced better, but any "improvement" was slow. I put them on a bike and ran them until the pack voltage fell to 3.3 volts per cell average voltage (around 60v) and then looked at the cells. Again, the pack was badly unbalanced. Also, the battery capacity was probably a third or less of what it was originally (not surprisingly). I noted with the balance charger that the cells that had the extremely low voltages would very quickly charge up to close that of the "stronger" cells. I placed the pack in a safe location and bulk charged it. It charged to an average cell voltage of 4.2 volts and was surprisingly balanced (there was no BMS) given that it had suffered such an extreme discharge. The pack got warm but not any more so than any pack when charging. I ran it on a bike again and bulk charged it and did this several times. I left it on the bike connected to the controller, outside, after charging. Two days later, I tried to bulk charge the pack again. It had discharged (either through self-discharge or parasitic drain) to an AVERAGE cell voltage of 2.8 volts but I did not check the individual cells; some would have likely been well under this. This was done outside away from flammables. I was working outside near the pack while it charged. I heard a loud "pop" and I went to the pack. It was on fire, and the fire was pretty impressive. The fire put out an impressive amount of acrid smoke. I would stress that charging LiPo inside is pretty dicey in my opinion, even if the pack has been closely monitored and not discharged extremely. I think if this had happened inside, the smoke would have accumulated so fast that putting out the fire would have been difficult. However, that is just my opinion.
This is not "new" information but just a continued warning to CAREFULLY monitor cell level voltages on LiPO. If a brick gets badly out of balance and is bulk-charged, there is probably a significant risk of fire. Anyway, again, this is not a condemnation of LiPO, just a reminder to "be careful out there". The risk of fire is not overstated; it is real and if you are not the kind of person to be careful to a fault all the time and not get lazy, I think a "safer" chemistry might be a wise choice.
Just trying to add to the cumulative experience of this community. Certainly, it was not the fault of the battery, but I think some people will start working with LiPO with good intentions to be careful but will get lazy and complacent over time when everything appears to work well, esp. with a bulk charger. I am pretty sure that a pack always charged with a good balance charger designed for LiPO has a very low probability of having a problem that becomes a safety issue. The potential problem is when bulk charging without insuring that the cell voltages are balanced. I would not think you had to monitor every time you charge but I do think it is important to realize the possible consequence of all the stars lining up against you; bulk charging inside with an unbalanced pack. Currently, you either have to be your own BMS or buy an aftermarket BMS, wire it yourself and either trust it or continually monitor the cells anyway. This is the kind of thing it is all too easy to get sloppy with. I've seen a fire first hand and I have to admit, it was a little on the scary side. Just be careful.
Hope this saves somebody trouble down the road from getting careless.
Jamis Commuter 1.0/Xtracycle Stokemonkey 36V LiFePO4, 15 ahr
Giant Boulder 9C 8x8, 48v, 10 Ahr LiMn from ebikes.ca