Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Battery Fire (and probable cause)
(01-30-2020, 04:44 PM)rev0 Wrote: Elkooo, where did you buy your cells? Just want to ensure they are real LG MJ1 cells, they're one of the most faked cells along with HG2, 30Q, and a few other "high performance" cells. I assume you capacity tested these prior to assembly anyways though. Fake cells definitely forgo the common safety features like PTC/CID.

This is an interesting possibility. Given that the cell vented from the side and not from the top. Even the fact that it caught fire in the first place. It really should not be possible if the cell is fitted with proper protection.
The cells were genuine (they came from a 100% thrusty source). I capacity testet them all and made matching packs (divergence was only ±1mAh from one 5p block to all others). The BMS was also OK (since the same pack was running well with old cells...) and is able to handle the new cells. 

I think the cell vent might only work for a limited amount of gas. As soon as the cells are too hot (fire around the cell), there may be too much pressure and also particles that block the vent. Possibly the venting gas ignited on a glowing hot piece of metal (all was in a tight package). 
This fire with several hundreds of °C burning around the cells might have caused the vents of the cells to simply not keep up with the amount of pressure building up. And the containments of these cells are also quite fragile (I did pierce cells by accident before, either mechanically or when I didn't behave propperly during spot-welding). 

I think it was just a chain of unlucky events that caused the fire:
1: Short on a + side of one of the packs
2: Ignition of either venting gas or some of the plastic around the cells / in the pack
3: Heat buildup inside the casing and subsequent total loss to fire

The fire would most porobably have been prevented by just one or multiple of the following precautions:
•fusing on cell level (not really possible due to high drain application)
•having extra insulation on + side of cells (paper rings / plastic rings...)
•being more careful assembling the pack overall
•other form-factor of the pack (powerwall-style...)
•no mechanical impact on the pack
This is why I do not use nickel strips. I fuse each cell and use straps to raise the plus busbar away from the cells.

Thank you for sharing your experience with cell failure! I agree that you can't really fuse high drain cells. And most high drain cells do not have PTC protection to limit current during a short circuit, so the best thing to do is to have extra insulation on the positive cap of the cell. A thick plastic ring would be better than the paper rings. Maybe double wrap the cell body too? Couldn't hurt.
Check out my long-term capacity test of 18650s:
I still feel that paper washers would be better for resistance to mechanical wear (e.g. nickel strip cutting into it) due to the fibers, and better against heat as well. And it must be worth something that it's how nearly every commercially built pack is done!
(02-06-2020, 06:26 AM)rev0 Wrote: I still feel that paper washers would be better for resistance to mechanical wear (e.g. nickel strip cutting into it) due to the fibers, and better against heat as well. And it must be worth something that it's how nearly every commercially built pack is done!

Interesting. Most cells have a plastic washer under the PVC wrapper, and then when packs are built, paper washers are used. I guess you get the best of both worlds that way. Good point!
Check out my long-term capacity test of 18650s:
I think the washer underneath the shrink wrap is somewhat smaller than the diameter of the cell - at least in the case of those LG-cells - may it be due to smaller fabrication or due to heat used in shrinking in the cells.

Will redo the 10s5p pack but with 3D printed top covers (which get a layer of paper glued on them for the fibers - I like the idea of having two materials in combination for extra safety).
I will probably post the build in a separate thread and put a link in here...
(01-30-2020, 03:55 PM)OffGridInTheCity Wrote: @elkooo - do you think that cell fusing would have stopped the fire in this case?

It seems to me that the keys are:
1) NEVER USE DAMAGED CELLS - dents, holes etc.
2) BMS - NEVER OVERCHARGE - saw a great video on this. If overcharge (above 4.2v or max for your Lithium Ion chemistry) current continues for very long (minutes even) a fire can start.
3) BMS - Do over discharge - not quite as obviously catastrophic but when in series, one of the cells/packs can actually go < 0v as the overall battery continues to deliver voltage and I'm not sure but I think this can cause fire? It totally destroys the chemistry of the cells for sure.
4) Short protection - battery design/arrangement, fusing (cell and battery level), physical protection from things falling on to the battery and any wear/tear over time.
5) Avoid heating cells - going beyond specs or testing on charge/discharge amps.
6) Physical protection - ambient temps and some physical protection so you don't hit the battery with tools or jab it with piece of pipe or wire while working, or drop anything on it.

I'm sure there are others.... but these come to mind and but by taking care of all the above they seem safe enough to me so far Smile

However, I do wonder about life-span. My bank is perfectly well behaved right now, but what will happen 5 or 10 or 15yrs from now when individual or collective cells fail from too many cycles. Will cells embedded in the packs become dangerous in some way as their capacities go lower and lower in uneven amounts? My plan is to use the packs all the way down to 40-50% of their original (today's) capacity.
I wonder about this myself.. I hope all cells will degrade over time equally.. but that might not be what happens. I purchased ev batteries so all cells were the same to try and avoid this.. but in that case there should be something that tells u that individual cells are losing capacity more than others..
I think single cells loosing capacity in a large bank are not that problemathic (they will be carried through by its neighbours). The larger problem would be cells that self discharge a lot. They may eventually get hot and cause a fire if not fused.
Therefore I think a thermal camera can be a good thing (draw a lot of current from your bank for a while and look for the hot spots... there may be your problems). This all applies for larger banks - smaller high drain packs are a completely different story. There every single cell matters and you better have equal cells!

Forum Jump:

Users browsing this thread: 1 Guest(s)