MSP & garage fire!

[Dzlchef]

Well, I had one of the worst imaginable events happen with my wheel other than injury, which I've already conquered many times in the past.   

5 days ago...my MSP C30 wouldn't charge after a 10 mile ride and I was working with eWheels to determine if it was the charger or a bad battery pack.   I found out that a cell was burnt on one of the packs as you can see in the photo and working to replace the entire pack.   The bad pack was nearly empty but still attached to the wheel, sitting in the garage, with the wheel being off, and not on the charger.   This afternoon, I took my other wheel out for a ride and as a fire truck blew by me, my wife called and said that my wheel had caught fire and burned in the garage.   If it wasn't for the quick action of my father who happened to be close to the garage and heard the "pops", the entire house would have been toast.   I didn't expect the pack to suddenly ignite without being charged or turned on but it happened.   I will now never charge my wheels overnight or when I'm not home given my recent experience.   My advice to everyone is to not mess with a damaged pack.   Get the pack out of the house and keep it in a metal container used for hot coals or like until properly it can be properly disposed.   

[Marty Backe]

This is the worst nightmare for all of us. I struggle to think where I could possibly store my wheels to be safe. Nothing. I'm stuck playing the odds.

At least we know that your battery pack was known to be bad, so there's that.

Perhaps if you had disconnected it from the other battery pack this may not have happened? Don't know of course.

[Dzlchef]

Yes, worst fears imagined.   I did have the other battery pack disconnected.   When determining whether it was the charger or battery pack, I disconnected each to test to see if it would charge.   I started on the side with the good pack but when I opened the other side and noticed the bad cell, it was obvious.   I disconnected that pack and the other side charged as normal and I brought it to full.   I then disconnected the full pack and reconnected the bad pack to reduce the overall voltage before returning via the mail.   It was reconnected to the "bad" pack only for a couple days and I was working on reducing the overall voltage to the very minimum.   The damaged pack was down to around 4% when sitting in the garage prior to the fire.   However, the full pack was still in the wheel but not connected.   And yes, I covered the disconnected wires with electrical tape so nothing was exposed and the wires secured.

[Dzlchef]

I'm now wondering if it was the type of cell, i.e. 21700 LG M50T, or the poor construction or both that contributed to the battery pack's demise?   Do I replace with another Gotway with the LG cells or opt for another wheel like the Veteran which uses the older and better tested 18650 cells?

[Marty Backe]

13 minutes ago, Dzlchef said:

Yes, worst fears imagined.   I did have the other battery pack disconnected.   When determining whether it was the charger or battery pack, I disconnected each to test to see if it would charge.   I started on the side with the good pack but when I opened the other side and noticed the bad cell, it was obvious.   I disconnected that pack and the other side charged as normal and I brought it to full.   I then disconnected the full pack and reconnected the bad pack to reduce the overall voltage before returning via the mail.   It was reconnected to the "bad" pack only for a couple days and I was working on reducing the overall voltage to the very minimum.   The damaged pack was down to around 4% when sitting in the garage prior to the fire.   However, the full pack was still in the wheel but not connected.   And yes, I covered the disconnected wires with electrical tape so nothing was exposed and the wires secured.

That's good clarifying information. So the battery pack was connected to reduce the charge. But Gotway wheels don't draw power when they are off? Was the wheel on continuously?

I'm thinking that if the battery pack was truly dormant (no current being drawn from it) nothing would have happened?

[Marty Backe]

13 minutes ago, Dzlchef said:

I'm now wondering if it was the type of cell, i.e. 21700 LG M50T, or the poor construction or both that contributed to the battery pack's demise?   Do I replace with another Gotway with the LG cells or opt for another wheel like the Veteran which uses the older and better tested 18650 cells?

Although heavier, the Sherman still rides somewhat like a MSP - meaning it's still agile enough. Except for situations where I need the extra pedal clearance, I'm grabbing the Sherman.

[Dzlchef]

1 minute ago, Marty Backe said:

That's good clarifying information. So the battery pack was connected to reduce the charge. But Gotway wheels don't draw power when they are off? Was the wheel on continuously?

I'm thinking that if the battery pack was truly dormant (no current being drawn from it) nothing would have happened?

I would have thought exactly the same.   The wheel was off, no current being drawn.   I did have the wheel on and was using power to reduce the voltage on the "bad" pack but had the wheel completely off for about 90 minutes before the fire started.  

[Marty Backe]

2 minutes ago, Dzlchef said:

I would have thought exactly the same.   The wheel was off, no current being drawn.   I did have the wheel on and was using power to reduce the voltage on the "bad" pack but had the wheel completely off for about 90 minutes before the fire started.  

I'm going to go out on a limb and say "stressing" the bad battery pack led to the ultimate failure. Even an idling wheel on a bad battery pack is probably not good. Nothing like hindsight though.

Two takeaways for me:

• A wheel that charges and maintains the charge is at extremely low risk of spontaneously going up in flames. This is good for my peace of mind.
• A wheel that is not charging properly and/or exhibits a clear failure (like your burnt cell) should be isolated and the battery packs disconnected.

[Dzlchef]

25 minutes ago, Marty Backe said:

I'm going to go out on a limb and say "stressing" the bad battery pack led to the ultimate failure. Even an idling wheel on a bad battery pack is probably not good. Nothing like hindsight though.

Two takeaways for me:

• A wheel that charges and maintains the charge is at extremely low risk of spontaneously going up in flames. This is good for my peace of mind.
• A wheel that is not charging properly and/or exhibits a clear failure (like your burnt cell) should be isolated and the battery packs disconnected.

I'd say your takeaways seem very logical and valid.   I'm also interested in why the packs fail and how to identify the problem as quickly as possible.   Failing during charging had always been my biggest concern but they can also fail just as easily while riding, and you might not even know it.

[Tawpie]

Personally, I wouldn’t have entertained shipping the bad pack—discharged or not. When a cell fails such that you can visually see evidence, the safety circuit built into that cell has also certainly been killed and the cell can continue to degrade in an uncontrolled fashion. Plus its companions in the pack may have been compromised by the failure event. You wouldn’t be out of line to treat it as if the whole thing has become an immediate fire hazard. Maybe dig a hole, toss in the pack and purposely incinerate it. After you take pictures for warranty purposes of course.

[.....]

I recall reading up on the 'new' lipo battery tech, years ago. I've since forgotten much for details, but back then there was a real concern of packs catching fire from being severly discharged. Not WHILE being discharged, just if they were left to sit at a very low state for long. Usually a pack would bulge, but there were MANY reports of them simply catching fire. Mind you, this was small 2&3 Cell packs and at least a decade ago. Then came along LiFe or some shit. Another variance that was supposed to be safer. It too bulged when too low, but i guess didnt catch fire? Basically I'm viewing this current fire as the same shit we've been rolling the dice on, for years. It must not happen all of the time, but its also not a problem that has gone away completely. Having a VERY discharged lithium battery, is a bad idea imho. I could be working on dated info, but...maybe not? I store my wheels next to a plane thats next to a motorcycle, near a paint closet by the propane tanks. Hell, there's only a few hundred lipos sitting there in those euc's. To be safe, I think we need treat these batteries like dynamite. They COULD be a hazard and we MUST treat and maintain them right. I'm a gambler, so I treat mine right and store them wrong....

OR!!! That scooter was pissed and started this whole damn thing!  Sorry to see the damaged euc and wall, but man were you lucky it didnt take the house!

Edited November 24, 2020 by ShanesPlanet

[Zombie Batman]

That is my worst nightmare.  I been thinking of getting some sort of large metal box to store my wheel in.  

 

https://www.amazon.com/Buyers-Products-Aluminum-Underbody-24x24x24/dp/B01ASXUSD2/

 

[mrelwood]

I think what makes this case special is that an individual cell was visibly burned first. When a single cell dies due to bad balancing or other degradation, it slowly drags the parallel cell group with itself to 0V.

Now that a single cell burned, the change was fast, essentially short circuiting the other healthy and powerful cells of the same parallel group.

I thought that the protection circuit in the cell would disengage the battery completely, but I guess a cell can still short circuit if it goes bad uncontrollably.

 Either way, whenever an EUC stops charging to 100%, something is wrong and requiring immediate attention.

Edited November 24, 2020 by mrelwood

[Zombie Batman]

Is there an EUC on the market that allow riders to monitor their battery health?  The Onewheel BMS allows riders to view the voltages of each of the cells.    (Actually its a bit more complicated than that because Future Motion tries to screw their customers over and prevent them from seeing this information, but that's another story all together)

 

[mrelwood]

35 minutes ago, Zombie Batman said:

Is there an EUC on the market that allow riders to monitor their battery health?  The Onewheel BMS allows riders to view the voltages of each of the cells.

I think Z10 is the only one so far that allows the user to see the individual cell group voltages. Ironically the Z10 has been plagued with battery/BMS issues pointing to faults in the voltage monitoring mechanism.

[Chriull]

6 hours ago, Marty Backe said:

I'm going to go out on a limb and say "stressing" the bad battery pack led to the ultimate failure. Even an idling wheel on a bad battery pack is probably not good. Nothing like hindsight though.

Stressing a pack with a bad cell by charging or discharging should not be done in any case!

Never ever! 

From the reports here there seem to be three stages of battery degradation - the first with the bad cell not getting fully charged anymore. One should/could notice this by the voltage after charging getting steadtily lower after each charge.

Second stage seems to be the cell holding no voltage anymore - it stays at 0V. This stage seems to stay some time - according to the reports of packs with ~4.2V missing.

After this the chance increases by further using this pack, that the polarity of the bad cell is "forcefully" changed.

The ultimate reason for a Li Ion going up (thermal runaway) is an internal short curcuit of the cell - the very thin seperator getting conpromised.

Beside direct physical impact, vibration, overheating etc low voltage can be a reason for compromised seperators: "There are reasons why Li-ion is put to sleep when discharging below 2.50V/cell. Copper dendrites grow if the cell is allowed to dwell in a low-voltage state for longer than a week. This results in elevated self-discharge, which could compromise safety." (1)

During the abovementioned "first stage" of a pack going bad, were the bad cell holds less voltage it actually gets repeatedly "beaten" to low voltages during use. Chances are high that if one uses the pack up to low charge percentages, that the cell goes below this 2.5V.

Quote

Two takeaways for me:

• A wheel that charges and maintains the charge is at extremely low risk of spontaneously going up in flames. This is good for my peace of mind.

Imho the best and easiest way to keep an eye on ones packs health - that it reaches after each charge about the same voltage, and can hold about this voltage!

According to (1) healthy li ion cells loose up to ~5% in the first 24 hours and then ~1-2% per week.

Quote

• A wheel that is not charging properly and/or exhibits a clear failure (like your burnt cell) should be isolated and the battery packs disconnected.

+1!

6 hours ago, Dzlchef said:

   Failing during charging had always been my biggest concern but they can also fail just as easily while riding, and you might not even know it.

Just by "extreme" overburden li ion cells could start such a thermal runaway quite "instantly".

But under "normal" conditions li ion cells  degrade over time noticable before. Unfortionately also new wheels can have bad and/or misbalanced cells!

With the failing during riding - this could be something GW specific? From the pics and videos shown the mosfets, capacitors and battery packs are very near together. So also the mosfets could burn the capacitors which ignite the li ion cells? Or the capacitors ignite themseves by overburden and then the li ion cells?

Or it's just some (still) some undersized wire loosing the insulation and shorting the batteries - as GW BMS have (had?) no overcurrent protection they deliver and heat up...

5 hours ago, Tawpie said:

Personally, I wouldn’t have entertained shipping the bad pack—discharged or not. When a cell fails such that you can visually see evidence, the safety circuit built into that cell has also certainly been killed and the cell can continue to degrade in an uncontrolled fashion.

The used cells in EUCs have no built in safety curcuit.

But yes, every compromised li ion cell can "degrade in an uncontrolled" way anytime. Once the seperator is damaged each small internal short curcuit == elevated self discharge can grow.

3 hours ago, ShanesPlanet said:

I recall reading up on the 'new' lipo battery tech, years ago. I've since forgotten much for details, but back then there was a real concern of packs catching fire from being severly discharged. Not WHILE being discharged, just if they were left to sit at a very low state for long. Usually a pack would bulge, but there were MANY reports of them simply catching fire.

This should still be true. Maybe/presumably the chances for catching fire decreased. From some reports here li ion cells take quite some abuse!

Maybe just the "new" 21700 take a bit less again? Or they are just burdened more? Did not count in detail, but afair reports increased since they were uses? Or just much more wheels came on the market since then?

2 hours ago, mrelwood said:

I think what makes this case special is that an individual cell was visibly burned first. When a single cell dies due to bad balancing or other degradation, it slowly drags the parallel cell group with itself to 0V.

Now that a single cell burned, the change was fast, essentially short circuiting the other healthy and powerful cells of the same parallel group.

Could be some badly welded nickel strips along the cells? As reported from one who took apart the battery of a smoking custom GW from aliexpress with the higher battery capacity.

Possibilities for missed QC are unfortionately quite endless...

Quote

I thought that the protection circuit in the cell would disengage the battery completely, but I guess a cell can still short circuit if it goes bad uncontrollably.

The cells have protection circuit inside!

And the single cell undervoltage protection from BMS was canceled after the first EUC generations, since this lead to many faceplants once the battery got weaker...

You remember @hobby16s topic on BMS shunting?

 

1 hour ago, Zombie Batman said:

Is there an EUC on the market that allow riders to monitor their battery health?  The Onewheel BMS allows riders to view the voltages of each of the cells.    (Actually its a bit more complicated than that because Future Motion tries to screw their customers over and prevent them from seeing this information, but that's another story all together)

Beside the already mentioned Z10 the Veteran Sherman could have such a monitoring. It does not report all single cell voltages, but here is one report of a sherman with quite full batteries showing a low voltage alarm.

So it could be the BMS signaling that one cell got undervoltage? ...or just some firmware/hardware malfunction..? No details are known about Veterans BMS..

(1) from https://batteryuniversity.com/index.php/learn/article/elevating_self_discharge

Edited November 24, 2020 by Chriull

[Simply_Striking]

3 hours ago, Zombie Batman said:

Is there an EUC on the market that allow riders to monitor their battery health?  The Onewheel BMS allows riders to view the voltages of each of the cells.    (Actually its a bit more complicated than that because Future Motion tries to screw their customers over and prevent them from seeing this information, but that's another story all together)

 

This was is true for the older onewheels. Future Motion has since then disabled this feature. Only way to check now is the manual way.

[.....]

25 minutes ago, Simply_Striking said:

This was is true for the older onewheels. Future Motion has since then disabled this feature. Only way to check now is the manual way.

How odd. Derail for sure, but I'm just curious if they quit having that feature enabled, or quit using a bms that monitors it? Was the bms needed to check so many locations, causing more failures, or was it simply cheaper to not incorporate. Maybe it still monitors it, but they simply don't want the user to access it? Lots of questions, but I just find it odd when a seemingly 'safety' feature gets changed/removed/omitted.

[redsnapper]

capturing the charge state of individual cells should be mandatory design feature of PEVs. I think the issue (at least for current EUC) is that there is no interface or intelligence on BMS that allows for this data getting to the controller (as batt state is simply computed off the total voltage).  BMS have gotten more robust over time but this is a crucial next step for signalling cell state problems. @Jack King Song (and other brand reps) should be listening as wheels that self-immolate are not good for business!

[bcarp]

7 hours ago, ShanesPlanet said:

I've since forgotten much for details, but back then there was a real concern of packs catching fire from being severly discharged. Not WHILE being discharged, just if they were left to sit at a very low state for long

I’ve never heard of this. Common advice in the FPV quadcopter world is to discharge battery packs to 0V then dispose on regular trash. At 0V, there is no risk of fire.

The issue I have with our wheels is that what’s happening within the batteries is being masked by the BMS. I have no way of knowing what individual cell packs are doing, what their voltages are, what their internal resistances are. I have no way of individually charging or discharging individual cells. This is fine for regular use, I suppose, but doesn’t help troubleshoot scenarios such as this one.

Discharging the battery pack using the wheel (as per OP) was never going to reduce voltage to the safest level for transport (0V), but short of disassembling the battery and putting each cell through a tungsten bulb equivalent, there’s no other way of doing it. 

[Coffee guy]

Although I’ve read about these runaways in the past and all the associated damage they (could) have caused, and it’s sad for sure.. this has scared the hell out of me, because for months I’ve stored a non working wheel with my other 4 working ones. Minutes ago I separated it and put it outside in the middle of the yard. I’ll either autopsy it and dispose of the battery packs, or the whole thing gets recycled.

I can’t say enough how much I appreciate this community (since 2015/6). 

[null]

Really scary, thanks for sharing. Hope you get past the anguish.
It might be worth having thread about what to do in case hell is loose, other than (on top of) call the fire department.

For example:
whether to stay around trying to control the fire, or if the fumes are too toxic.
Other than using an extinguisher, could spraying it with water help cooling it.
If there is any point in trying to isolate it with a fire blanket.
How long to expect it to last..etc..

Of course prevention is best, but in case shit hits the fan it's better to know what to expect and how to act. Especially for people who live in flats and dont have separated areas like garages this is scary.

edit: not to derail anything here, maybe I should move this to separate thread..

 

Edited November 24, 2020 by null

[Marty Backe]

14 hours ago, Zombie Batman said:

That is my worst nightmare.  I been thinking of getting some sort of large metal box to store my wheel in.  

 

https://www.amazon.com/Buyers-Products-Aluminum-Underbody-24x24x24/dp/B01ASXUSD2/

 

If I had one or two wheels I would store them in such a cabinet. What to do when you have a dozen 

[.....]

11 minutes ago, Marty Backe said:

If I had one or two wheels I would store them in such a cabinet. What to do when you have a dozen 

Bulk up on insurance, and play the odds. You know, same thing we do when we ride. Think happy thoughts!

[AtlasP]

On 11/23/2020 at 6:53 PM, Marty Backe said:

I struggle to think where I could possibly store my wheels to be safe.

For those who own property with sufficient land and in a very temperate climate, I think a cheap tiny pre-built shed (really a 'detached closet' ;-) ) with decent insulation away from everything could be an option. (Detached garages are better than the house, but many people still keep a lot of valuable stuff in garages, from general storage to actual automobiles.) This wouldn't work for those in more variable climates however as temperature swings are also extremely bad for such batteries, unless you were going to try to climate control the shed (I don't know if they make tiny thermostat-controlled h/v/a/c units for something like a shed, or maybe one could repurpose something from an RV or similar).

On the easy side, it would be super easy to mount a smoke detector immediately above where we store wheels (like a foot above them on the wall)--I imagine this would go off quicker than one even in the same room but mounted on the ceiling on the far side of the room or something. Edit: also I assume they probably make 'smart' smoke detectors these days that could give you phone notifications in addition to loud beeping.

Edited November 25, 2020 by AtlasP