2013 Li Battery Catastrophic Events - Summary Thread

[bigmoose]

This thread will stay sticky in Battery Technology for the rest of 2013. Ideally there should be only 1 post on this thread per incident. It is a summary thread.

What should be reported here are catastrophic Li Battery events that you have personal, and direct knowledge of.

The Poster should:

1) Put a title in his post that describes the battery and manufacturer, for example: 36V 20 AHr A123 Pack
2) Describe the incident from first hand knowledge: bla bla bla and add pictures. Describe the conditions that lead to the event, perhaps: manufacturing, abuse, or other. Identify who the cell vendor was, and if a battery, who the battery vendor was. If you built the battery, then identify the cells and other pieces parts used.
3) Start a separate discussion thread for the incident and place a link to it in this summary thread. (t would be good to duplicate the information in 1 & 2 in the incident specific thread.) This second thread is where all the discussion and give and take should be.
4) When the Suspected Root Cause is determined, the Original Poster comes back to this thread and edits his post to add the best probable root cause right in the original summary post. Adding the corrective action would be a good thing also.

MattyCiii summarized his view of the process this way:

I would imagine such a sticky thread work as follows:
One post per "event". The poster makes a thread to discuss the incident, throws in a link to that thread, and briefly summarizes what happened. So most of the talk about that one event goes on in its own thread.

We don't want the detailed give and take in this thread. Just summarize the incident, and then come back and summarize the root cause and corrective action.

This thread's intent is to improve the design and manufacture of both commercial, and DIY Li batteries; and to help the community identify reliable and safe Li Battery design techniques, materials, vendors and systems. It was summarized here http://endless-sphere.com/forums/viewtopic.php?f=14&t=46748 and a poll taken to see whether it was a good idea. It seemed to have over 97% support. Hopefully I have added all the suggestions for improvement above.

Note: I had a lot of trouble getting this thread to be a sticky. Something must have gotten fouled up in the data base when I copied it... so I started over.

 

[BikeFanatic]

The link to the detailed discussion thread of this incident: http://endless-sphere.com/forums/viewtopic.php?f=14&t=46921

Lipo fire The senario:

I had a brand new tested Zippy from hobby king Lipo 3 series 5 AMPHR and a pair of Turnigy 4 series hardcase packs on my bike to use to power my headlights, I use vinyl/plastic front tube bags to hold the batteries. One side is mesh the other plastic and you can use a shower cap to cover it futher and protect it From rain. I did not take that extra precaution.

so I was riding to work very early in the morning, on my ebike, it started pouring rain, 20 minutes later i am almost at work and I hear PSSSSST, like a short and something evaporated, say a balance wire. I continue on another minute or two then my3 Series lipo burst into flames. It was seated near handle bar area in a bag

I was able to separate any of my other lipos from contributing to the fire.

I could not find the pack that night I went back in daylight to find the remains that I kicked into the wooded area.That’s when I noticed that I had hooked a fully charged brand new 3 series pack to the fully charged 4 series in parallel .
since I use power poles for everything. I make all my connections the night before, but apparently in my early morning haze I for got and thought I was hooking my light to a battery.

SO you can see I found the wires came off, melted off I guess , no sign of shorting there
I measured the voltages of the four series battery and all cells are at 3.10 volts.

I am concluding that it was an overcharged 3 series that caused the fire. The 4 series overcharged the 3 series and the 3 series had a thermal event.
I suppose there could have been a short , it was raining, but all the evidence is burnt up except the very intact power leads.

 

[bigmoose]

Placeholder for GCinDC's LiPo fire. Thread here:
http://endless-sphere.com/forums/viewtopic.php?p=716486#p716486

A very serious fire in the garage. For some reason it self extinguished, perhaps consumed all oxygen, and saved his garage and home. GCinDC is investigating.

 

[hydro-one]

Sorry so late to the "Party" . Here is a thread of "my" fire. Very serious thermal event of a 15s10ah hk lipo pack . The cause is still unknown but a damaged cell(s) in the pack that was being charged without monitoring is most likely.

http://endless-sphere.com/forums/viewtopic.php?f=14&t=46011

 

[Arlo1]

I have 15 packs I am playing with for many projects 3 spares besides that and 20 on order ATM. I was using the 15 packs for power for the drift trike and never made Y adapters for the ends of the packs I just left methods HVC/LVC boards in place to keep the 3 strings paralleled. On the YSR I made a set of Y adapters for the ends but I can not transfer them to the drift trikes so in my "Testing" on the drift trike I melted the traces on one parallel board at the end of my string and ran a single pack to 0v during hard current draw ~70 amps right down to 0v per cell. It must have happened early on because I Still had 4.2v in the cells on the other two packs that were originally paralleled to it.

THIS IS MY FAULT 100%. I will not make the mistake again. But it was good to see the pack only puffed and that's it! I tried punctureing it and pouring water on it and nothing exciting happened it is at 0v per cell so there is no energy in the pack anymore. Good to know bricking a battery will not be a cause of fire...
As long as you chuck it out and don't try to bring it back that is.

[youtube]vG-nxMPt1hw[/youtube]

 

[raysmith1971]

not had a fire myself luckily but seeing these pictures has been enough to prompt me to fit a fire alarm in the garage where I charge the bike.
I have 4 x 5s batteries connected as a pack .They are connected as 2 packs each in series which give 2x 10s packs. these are then connected in parallel to give 10,000mah. I have fitted in the circuit 2 RCD's (labelled A and B) which disconnect the packs in series and disconnect them from the controller. I switch them off to charge and once I forgot but luckily the RCD tripped and saved me! Left an attachment image so you can see what I mean 9dont know how to put an image up) . I'm sure there is a better circuit to use but this is what my mind could think up. If anyone has an idea for a better circuit I would be interested.

 

[bigmoose]

parajared fire: Thread here http://endless-sphere.com/forums/viewtopic.php?f=3&t=50408

I combined my lipos into 20 ah 6s packs and I was loading them onto my bike and I managed to drop one. Two of the four batteries in the pack were damaged.

 

[leffex]

Three pieces of 4s1p hardcases

So my brother gets into ebikes... second time home alone charging with b6s and chooses nicd or nimh setting... charge for 4 hours... 24 hours later i check his battery and cells and gets confused. I see this: 4.6V

I went for a ebike ride and came back setting them at 3.85v per cell. NO apparent damages on outside of battery, no puffs, almost new from box with 2-3 cycles on them.

Set correct battery type and voltage in the charger as well as informationwise give information to the supposed user, ie, my brother.

Bwhahahah ...after 12 hours another thing happend to me and a FIRE! Well to be short two packs in series and then charged in parrarell via balance wires, when second connection is put it makes a fire and destroyes the charge balance wires in seconds...Sad but true. Not to happend again!!!

2013-23-jul Another update the world burnt but not me. No my brother again go for a ride and draws on pack down hard and a few others to 2.8v, 3.3 and 3.4 per cell. I tried to even out the pack that reached 1.2 but that one just got bigger and hotter and I noticed very little smoke so took all cables out and put it in a lipo safety bag, flying into the oven!

 

[steveo]

Ive been fortunate to never have a lipo fire

I charger to 4.10v a cells ... And never below 3.5v

I also try to keep all my cells in storage voltage of 3.85 to preserve life of them and keep internal resistance from increasing

Ive seen lipo stored during winter fully charges be all puffed up by the time spring hits

Test all your packs to ensure everything is good before you series and parrallel anything

Also.. I run my lipos very conservative.. I always base my max discharge based the continous rating of the lipos and not the burst rating

I respect my lipos, i hope to never have some of the problems post here.. All look to be human error here

-Steveo

 

[bigmoose]

Floont lost his bike, his house, his car and his truck to fire...

RIP Monzilla...

The battery charger caught fire and burned the bike beyond recognition. It also burned my whole house down including my car and my truck.

I will never go back to ebiking again.

FA

http://endless-sphere.com/forums/viewtopic.php?p=781444#p781444

The fire is an absolute fact, not a sick joke.

My thought is that the charger overcharged my pack and it caught them on fire and then caught my garage on fire.

Beware of LiPo and cheap Chinese chargers. They'll ruin your life.

Goodbye to my ebike brethren. I shall not be returning to the hobby. The total cost is just too great.

FA

PS My cat just got out of the animal hospital with tongue and throat burns and injuries due to smoke inhalation. This hobby is not just worth it.

http://endless-sphere.com/forums/viewtopic.php?p=781944#p781944

 

[bigmoose]

Wannesd had a bad case of KFF when a voltmeter lead slipped and the connector errupted into an out of control plasma ball. It self extinguished as all the plastic melted. Details in this thread: http://endless-sphere.com/forums/viewtopic.php?p=784172#p784172

 

[major]

This thread looks like a good place to post this. I was assisting a client with testing when this failure occurred. Here are a couple of photos.

Only a single cell lit up, fortunately. These are A123 from tool packs configured in 10S8P subpacks in a larger battery. This happened during a test in the vehicle with the wheels spinning freely so the cell current should have been only a few amps.

These cells were several years old but with relatively few cycles. They were charged with balancing chargers on each 10S subpack.

I think it was obvious that the cell had developed an internal short. Fortunately it was on the end of the subpack which directed the jetted heat away from adjacent cells. The red hot glow of the failing cell was visible through the Lexan cover and the test was immediately terminated. All we could do was watch as it glowed and smoked, holding the fire extinguisher. After a few minutes the smoke ceased and we started to disassemble the battery.

That is about all I know and all the details I am comfortable giving about this particular application seeing as how the client may not want that. I just thought the failure mode would be of interest or help to you guys.

 

[Wishes]

I had my first Lipo fire this summer after using Lipo's for years. I went riding in the dark without any lights but I was riding paved bike paths and street so I figured I'd be ok. I was going along fine on a bike path at a relatively slow 20 kph. I had to transition from the bike path onto the street, the city is changing out all the piping for many of its streets in residential areas this summer, so seeing a street on gravel was not surprising. What did catch me by surprise, was that the street was 3 feet lower than the bike path.

My bike is a DH Kona stinky deluxe full suspension, but is set up to ride on the street and bike paths only. I sometimes put the Lipo's in a bag that I fix inside the frame's triangle. You can see the pics on my build thread http://endless-sphere.com/forums/viewtopic.php?f=28&t=52029. Well that sudden drop actually compressed my back suspension enough to come squeeze my bag with 3 x 6 cell lipo's in it. It pierced one of the cells on one of the batteries.

As soon as I landed, i heard what sounded like air hissing, my initial thought was that I had caught a flat. I looked down to check my tires and I see smoke coming out of my bag of batteries. I quickly pulled the battery wire to the ESC, zipped open the bag and yanked out all 3 batteries and threw them on the ground.

The battery had only started to smoke, I was fast enough, and mostly lucky enough get them out before any significant heat build up. As soon as the battery hit the ground it caught fire. It was in the middle of the street making an impressive show of fire and smoke. There was no wind that night to move the smoke, it made an pretty impressive site under the dim street lights. The whole streat got smoked, enough to attract the attention of a local resident or 2. It was burning literally in the middle of the street, would have blocked traffic had a car come along. with the residents watching in concerned, I decided to kick burning battery to the side of the curb and out from the middle of the street. Doing so loosened the various cells in the battery, and 4 of the 6 cells separated from the now burning 2 cells. That limited the fire somewhat, I have seen smaller Lipos catch fire in electric planes and helicopters. But those had 1 or 2 AH. It was impressive to see the amount of heat, smoke and fire generated by 2 cells of 5000 mah.

After spending 20 minutes convincing the residents that lived in the houses in front of where I had my little fire, not to call the police and fire department, and after making sure the fire was out, I hopped on my bike and pedaled home. I slapped in some freshly charged batteries, grabbed an empty bag and went back to pick up the remains. Here are some pics of the remains.

 

[CroDriver]

Very informative and good thread so I will share a bit of our experiences too.

We had an accident in our battery-lab recently.

We were cycling two 20S3P modules with A123 cells. This is a common thing we do all the time with a bench developed and produced by us in-house. The bench has a 500V, 70A charger integrated. A PC communicates with the module's built-in BMS to read voltages and temperatures. It is also communicating with the charger and other systems. A lab view program developed by us runs on the PC and controls the charge/discharge process and adjusts current, voltage, turns the charger on and off etc.

During charging, the test-bench PC froze without anyone noticing it so the charger continued to charge after the cells reached 100% SoC. Fortunately, it was during working hours (sometimes they cycle over night) so we noticed the noise of venting. There was just a little smoke and the venting noise, everything else looked normal. The voltage of all cells has been over 4,5 V when we restarted the PC (and after the charger was shout-down of course). So I guess that the voltages have been over 5V before we turned the charger off.

We left the modules sitting a couple of weeks so that no gases remain in the module before disassembling. The foils on the top of the cells, between the terminals have separated allowing the pressure to vent. Apart of that - they look normal. They are soft but not swollen - so the pressure system did its job. All cells have 3,3V.

I'm really glad that we don't use LiPo. And we'll finally finish that battery test bench to prevent future accidents. Devices with an operating system should never be used to control stuff like that. Over $7.000 worth of cells gone because of a frozen PC. But could have been much worse.

 

[wb9k]

Very informative and good thread so I will share a bit of our experiences too.

We had an accident in our battery-lab recently.

We were cycling two 20S3P modules with A123 cells. This is a common thing we do all the time with a bench developed and produced by us in-house. The bench has a 500V, 70A charger integrated. A PC communicates with the module's built-in BMS to read voltages and temperatures. It is also communicating with the charger and other systems. A lab view program developed by us runs on the PC and controls the charge/discharge process and adjusts current, voltage, turns the charger on and off etc.

During charging, the test-bench PC froze without anyone noticing it so the charger continued to charge after the cells reached 100% SoC. Fortunately, it was during working hours (sometimes they cycle over night) so we noticed the noise of venting. There was just a little smoke and the venting noise, everything else looked normal. The voltage of all cells has been over 4,5 V when we restarted the PC (and after the charger was shout-down of course). So I guess that the voltages have been over 5V before we turned the charger off.

We left the modules sitting a couple of weeks so that no gases remain in the module before disassembling. The foils on the top of the cells, between the terminals have separated allowing the pressure to vent. Apart of that - they look normal. They are soft but not swollen - so the pressure system did its job. All cells have 3,3V.

I'm really glad that we don't use LiPo. And we'll finally finish that battery test bench to prevent future accidents. Devices with an operating system should never be used to control stuff like that. Over $7.000 worth of cells gone because of a frozen PC. But could have been much worse.

Mate,

Sorry to hear about the incident. We had a couple similar incidents on (house-built) cyclers in our service center in the past that were also running on Labview. The fix is a secondary OVP circuit that is hardware-based which will override the PC's control of the cycler relays in the case of a lockup. Labview seems to be particularly good at this kind of "glitch". We don't use it in production at all anymore.