What causes pack fires? [unfused dead cells]

I have some oldish double A's that still have a 1.2v charge on them {after acouple years of neglect}. And some that are below 1 volt. I figured those are too dangerous to recharge. I was charging a pair of tripple A's and the got too hot as far as I could tell. I use anitecore intellicharger i4, which is not exactly the best. So I am thingking about a new so called smart charger that has more voltage displays etc. But still they are not great improvemtns. My iMax B6 has an outo turn-off setting but that reads charging voltage not actual cell voltage. But it is good for keeping the end voltage low.Set for 4.1v the cell stops at about 3.9v.....


So what I want to know is just how dangerous it is to charge NiMH cells untill they get hot. I would not do that to Li-Cobalt cells. Except that I did a while back, the cell was down so low it would not take a charge on my "Smart charger" ...so I put it in a peralel pack {plastic tray} with other cells that were up around 3.6-8v and inless than 60 seconds the dead lithium-ion cell started getting HOT.So I ran it outside expecting it to blow. Well I was lucky it did notburst into flames [if I had walked away from it or not felt the cell with my hand]. But it was aslo good to understand how a cell can over heat. If it was in a pack made of old unprotected computer cells there would be no way to detect a cell going below the safe voiltage for recharging.


I need to find fueses (resettable?) for my ebike battery pack; now that I understand how most fires are started. Am I wrong? I don't see howeven an expinsive BMS could stop this, unless each and every cell in the pack had a sensor on it. [And I want to build a pack with 224 cells. Good thing this happened.] This would be a very good demonstration for Youtube. I could blow a dead cell with just one or two live cells. Has anyone made a video about how an unfused deadcell can distroy a whole pack?
I found several on how to revive a dead cells but nothing about what happens if you can't get them out before it's too late.


Are LiFePo4 cells fire proof? I think maybe they are over rated in the saftey factor. I could use 26650 cells if I am able to spend another $500 just for the cells, compared to the Samsung 29E cells. But they would not last much longer than the 29E's. [Unless someone has a much better scource.]

jawnn,

As you've discovered, wiring discharged LiIon cells in parallel with charged cells is a BAD idea.

However, once the cells are in a pack, THIS CANNOT HAPPEN. If cells are in parallel, you cannot discharge one without discharging the others. At least not if the pack has reliable connections. This is why a pack with x cells in parallel and y in series is perfectly fine with a BMS designed for y cells. I'm guessing your pack might be 14S16P. If so, a 14-cell BMS will protect the pack as well as it would a 14S1P pack.

If you have cells that won't take a charge:
NiCad - hit them with high current to burn out the shorts
NiMH - do what you want. The cells are junk, but not dangerous
LiIon - discard the cells. Do not try to recover them. That's a great way to start a fire.

As I've infered above though, overcharging bad cells is NOT how pack fires start. In a pack with a proper BMS, a bad cell will shut down the whole pack if it hasn't already started a fire. Pack fires can start during charging, resting, or discharging, but are rarely due to overcharging. Fires start when a cell self destructs, or shorts and absorbs large amounts of energy from parallel cells. Neither of these is a condition that can be prevented by a BMS. The best defense is quality control during cell manufacturing.

I believe that LiFePO4 cells and batteries are much safer than other LiIon chemistries, especially LiCo, given the way we use them. However, I've read some fairly convincing arguments on CPF that the safety margin is primarily due to the lower energy density of LFP cells. The argument is that if you charge LiCo cells to the same energy density (something like 30-40% of their capacity), they would be just as safe and have similar lifetime as LFP cells. If I had a significant stake in this debate (which I may or may not any time soon), I would put some serious effort into researching this topic.

well thats a relief.....but ai think I want to use fueses any how with the Samsung 29E. the are a lower dinsity cell than the 3400mAh cell every one want to use. 2900mAh and I won't be charging to more than 80%. .......thanks

The reason LiFeP04 cells are claimed to be fireproof is that iron-phosphate (the "FeP04") part of the lithium chemistry does not release oxygen molecules, hence no flame since it is missing the oxygen component to start.

BUT, when abused, they will vent. Dramatically, just like their cousins. However, that is only the electrolyte cooking off.

Basically the difference is between "venting", and "venting-with-flame".

In the real world, when abused like that, it also means that your *connecting infrastructure* is poor and unsafe, THAT is what can set itself on fire with too much current or other problem.

And obviously, counterfeit, fakes and the like which are poorly constructed are another issue - but this was merely the technical consideration, not manufacturer quality issues.