Kingsong s20 Cut out FIRE (New York)

[Paul A]

Isn't there a thick, metallic heat sink between/shielding the control board and battery packs?

 

It was approximately twenty seconds between initial popping noises and smoke, to the flamethrower fire.

Could a burning control board generate and transfer enough heat, in twenty seconds, to cause thermal runaway in the battery packs?  

The cells are encased in the aluminium boxes.

[Mayhem]

14 minutes ago, rolis said:

these seem to by effective:

https://firesak.com

Own two, good construction. hope to never need them for it’s intended purpose. 

[Tawpie]

40 minutes ago, Roadpower said:

This year I'm going to build something on my brick & concrete terrace to isolate the wheels in.

If nothing else, this is something every owner of a large Li Ion pack should be considering. Pretty much every Li Ion battery system has had at least 1 or 2 units catch fire—laptops, phones, hoverboards, one wheels, wheels, cars, commercial airplanes, bikes, scooters. While it's a pretty safe energy storage medium, large Li Ion packs store so much energy that they are very very dangerous when enough things go wrong.

We want (demand?) our wheel manufacturers to do better and avoid these kinds of incidents, but we have responsibilities as owners too. Especially when we know from experience that as a group the manufacturers of EUCs don't have a stellar track record.

Try to make certain that it's possible to get people out of danger before things get worse. Get all the people out and away, then and only then worry about the fire.

[FlyboyEUC]

1 hour ago, Tasku said:

Made me wonder if metallic big bin would be good enough to dump device into with help of fireblankets. Then fill it (the trashbin) with water like the firemen did.

When I called the local fire Marshal, he said to put the batteries in the parking lot with sand on top and then cover it with a metal bin. He told me that thermal runaway can only be slowed down. So best thing to do is protect the area from the flames and projectile cells. Putting a wheel in a trash bin with water or sand will help contain the fire but not necessarily extinguish it.

I had two 320wh KS packs that were submerged for about 3 minutes in fresh water. 3 hours later, one pack started to smoke so I threw it in a bucket of water. Called the fire Marshal and followed the directions above. When I went back to the parking lot hours later, someone had uncovered my "burial" either out of curiosity or they saw it on fire and tried to douse it with water. I noticed one pack was still intact and left them both there. Early in the morning the next day, both packs were burned. The second pack must have burst later that night. 

[chanman]

40 minutes ago, Paul A said:

Might be wrongly assuming that the control board fire has spread to and caused the battery pack fires.

 

What we see in the video is a failed board, some smoke, and quite quickly battery smoke and fire. What's not happening is the board on fire heating up the batteries by transferring heat causing failure, that would be ridiculous and would take much longer to occur if it was possible. What is happening is the board has failed in a way that huge amounts of current are being drawn from the battery, causing thermal runaway, with the various safety features on the battery pack failing to prevent this for some reason.

Edited March 28, 2022 by chanman
elaboration

[FlyboyEUC]

31 minutes ago, Paul A said:

Could a burning control board generate and transfer enough heat, in twenty seconds, to cause thermal runaway in the battery packs?  

The cells are encased in the aluminium boxes.

Aluminum is a great thermal conductor (used as heat sinks). But the fact that the two metal boxes are separated from each other buy a couple inches (like 5-7) with the suspension fully extended seems like decent enough separation. Maybe the trolley handle? Heat transfer seems unlikely. Probably more of a current transfer.

I haven't studied the tear downs of the S20, but considering most wheels that have had a mother board failure have not caused the packs to burn, I think there might be something unique to the S20 layout that helped transmit either the heat (unlikely) or the current to the packs.

Edited March 28, 2022 by FlyboyEUC

[Tawpie]

37 minutes ago, Paul A said:

It was approximately twenty seconds between initial popping noises and smoke, to the flamethrower fire.

Could a burning control board generate and transfer enough heat, in twenty seconds, to cause thermal runaway in the battery packs?

<speculation>

I pondered that too. The long duration between popcorn (and light smoke) and cell venting hints that perhaps the battery system was shorted by a failure in the control board but the short circuit protection of the BMS did not work before one or more of the cells started down the path to thermal runaway. It doesn't seem like there was enough 'fire' from just the control board to have heated the battery pack to failure.

I think the top pack was cooked by the fire in the bottom pack, just from the way the video played out.

The post-fire picture of the control board is missing 1 of the 4 160V caps... Hsaing said UStride said (paraphrasing) he thought it popped, like a cap blowing up. It looks like that might have happened?

I don't know, but I think KS has two things they need to look at carefully: if the cap popped, why did it pop, and perhaps more important… if the control board failure shorted the battery leads why didn't the BMS disconnect the battery pack?

</speculation>

The battery pack containment is a good thing and a bad thing... fiery missiles didn't come forth, but because they had nowhere to go they stayed packed nice and tightly where they could spread their joy to other cells and ensure complete involvement of the entire pack.

Edited March 28, 2022 by Tawpie

[BKW]

Was finishing my last delivery at @1:20am on 3/28 when I took a picture of the location where it happened. Street sweepers haven't even got to it yet

[Richardo]

First generation wheels are actually roulette wheels. And the first generation lasts about 18 months. 

[ESB]

Unacceptable and extremely disappointing. For fuck's sake.

[mhpr262]

4 hours ago, Jason H said:

Tragic for the whole community!

 

1 hour ago, ESB said:

Unacceptable and extremely disappointing. For fuck's sake.

I couldnt agree more. This is becoming genuinely problematic. With every such fire a blanket ban of EUC is coming a step closer. This community is not nearly big or influential enough to give any politician pause who feels the need to be seen "doing something"... like giving EUCs the Hoverboard treatment.

[SamSuffit]

20 minutes ago, fbhb said:

King Song's Official announcement re: S20 fire

 

 

Software issue only? Seems like it is convenient for KingSong that the issue seems very easy and cost efficient to fix...
I am not an engineer, but it sounds quite fishy...according to reviewers the hardware also needs security upgrade.

[supercurio]

Thoughts after @Hsiang video, Kingsong response to the public and @Rawnei and I S20 dealers:

Controller looks in very good shape

The controller was not ravaged by flames to begin with, which likely rules out fire of the controller propagating to the battery packs as root cause for the fire escalation.
Then it is possible that the the fault originated from the mainboard, with a component failing for an unknown reason but we can't tell from this.

Credits @Hsiang

 

 

 

Cells shoot upwards

Would be preferable if the cells are ejected downwards instead - if that must happen in case the wheel is upright.
Credits @Hsiang

 

All the safety system redundancy failed at once

Mainboard

The board is supposed to protect itself via hardware and software. In this incident board failure may or may not have been the root cause.
In any case, the firmware is supposed to enforce limits which prevent the controller to self-destruct.
If board failure is the root cause: that didn't work

BMS MOS

The BMS has a bunch of MOS which are here to disconnect the output in case of problem: bad cell health, under-voltage, and short.
If that works as expected and you short the output, essentially nothing will happen besides an initial spark.

Cause, either:

• BMS firmware bug, non fonctional short detection or handling
• BMS hardware design / QC issue, invalidating short detection or handling
• both

Note: I don't know for sure these are power delivery MOS that can disconnect the output. That's my best guess.

Credits Wrong Way

 

BMS Fuse

Can this SMD 60A fuse handle a 126V battery pack, aka up to 126V DC?
I don't think so when looking at these specs!

60V DC for the 60A model, not higher than that. It means that after the fuse blows, current would just flow over the plasma created by a permanent arc bridging over the fuse.

It means that if a small fuse capable of handling high voltages even exist, then the S20 BMS hardware must be upgraded, since this fuse has most likely be proven to be incapable for the one job it has, protecting from short-circuit in case the BMS MOS don't disconnect the output.

Note that its very common to use fuses way beyond their voltage rating on EUCs. For instance, the 100.8V Sherman has fuses rated up to 32 V DC instead.
I'm not even sure any Sherman mainboard fire even triggered these fuse BTW.

Credits Wong Way

 

 

 

Edited March 28, 2022 by supercurio

[Rawnei]

1 hour ago, fbhb said:

King Song's Official announcement re: S20 fire

 

 

Interesting, but even if the root cause is started in software it also shines light on hardware failsafes (fuses?) not working as intended.

[supercurio]

More thoughts following these on the fire itself and failure of all safety mechanism

Kingsong response to the public

Immediate acknowledgement and response: great.
Saying that it's solved with a firmware update: yes and no.

Is it credible that a firmware bug led to this?

Yes.
Inadequate limits set in the firmware, bug in the motor control algorithm, added to a firmware bug in the BMS preventing the BMS MOS to disconnecting the output are all causes that can explain together the fire

Is a mainboard + BMS firmware enough?

No.

1. The fire likely demonstrated that the BMS fuse, rated for 60DC indeed cannot fulfill its duties at the S20 operating voltage range: 96V-126V.
   Requires hardware change, current fuse configuration likely entirely useless.
2. Fire of one pack can easily propagate to the other pack via the channel connecting them
3. Moisture and water can get in the the battery box and corrode the unprotected electronics and raw cells

In their response, Kingsong essentially implies that a software fix is enough, despite this incident violently showed that the hardware safeties supposed to prevent escalation in case of a software issue don't work.

Kingsong response to the dealers

Here's what Kingsong replied to eucshop.lt inquiry, and probably many others.

Quote

Kingsong confirms that upgraded S20 will be send to us in early May. They will make changes regarding the last S20 case in NY for more safety.

I don't know about you, but I read it as all red flags.

• Kingsong hasn't started investigating the hardware yet, but already gives an ETA
• Promises "more safety" despite not knowing root cause
• Promises "more safety" before testing unspecified fixes to the unknown problems
• Doesn't give enough time for hardware fixes, only software and hardware hacks
• Kingsong just cares about convincing customers to not cancel their (pre)order

I don't like this at all.

 

Edited March 28, 2022 by RagingGrandpa
(fix url)

[Wolverine]

As we can see clearly from the video, the mainboard has failed. The fault of the mainboard failing is in two places: hardware and software. Hardware and software play a very important role in the whole system, and they do work in harmony. If one of them fails, the other will protect the rest of the system. In this case, both have failed, and they must be redesigned.

As for the battery defense systems and the smart BMS, it also needs to be redesigned because it didn't work at all, otherwise this would not have happened.

This is the second consecutive problem we have seen with motherboards. These have happened to companies we would not have expected. First the Inmotion V12 HS failed and now the Kingsong S20. As a community, we must stand up for the safety of our riders. I urge you to withdraw from S20 pre-orders. Manufacturers need to understand that we want safe eucs, not life-threatening devices.

[techyiam]

I am hoping Kingsong will go beyond analyzing the failure just from a purely technical perspective. They need to see the bigger picture, including the higher owner/rider safety point of view. 

For example, merely coming up with a fix that will avoid future occurrences of the cutout under the same condition that led to the spectacular failure is not enough. Not only do they have to eliminate random cutouts (not including over-torque or over-speed cutouts), but also if cutouts, or other types of malfunctions do occur, the failure of the controller board(s) will not lead to sparks, flames and smoke.

Edited March 28, 2022 by techyiam

[RagingGrandpa]

10 hours ago, Paul A said:

There are two threads currently discussing this fire.

All merged here now  

[supercurio]

U-Stride, the rider who was riding the S20 published his video.
The wheel over-powered and cut-out on a hard breaking as he was testing the acceleration/braking torque limits.

What I can see in the video:

• Wheel is over-powered, dips backwards and stabilizes backwards at a fixed angle for an instant
• You can hear a sound, like something popping. I initially thought it was the kickstand against the ground but it might not.
• Wheel and rider fall after that, no recovery

 

U-Strides describes later how he heard a pop just earlier which allowed him to kind of jump off the wheel, but you can't hear it in the video.

 

With timestamp for the crash:

 

Edited March 28, 2022 by supercurio

[Cress]

I think the @supercurio post, above has the correct perspective.  It's likely the battery manufacturer will direct attention to battery specifications including peak charge/discharge current and battery case protection.  EUC manufacturers sell EUCs based on EUC performance; when current is limited to battery specifications EUC performance is limited compared to instances that violate battery specifications.

Inmotion sent the V12 to production without an electrical engineering review that would have caught the MOSFET specifications on the motor controller.  Kingsong sent the S20 to production without an engineering review of the batteries that would have caught every point mentioned in the @supercurio post, above.

Manufacturers and consumers each have roles in technology development.  I'll support a manufacturer based on positive contributions to the process but tell the world about stupidity that causes chemical fires, motor-controller failures and whatever disaster happens next.

I'm waiting for each word Kingsong says next.  Accurate, complete and transparent explanations are on-track but NOTHING changes the reality that Kingsong moved the S20 to production without doing a technical review.

For clarity - Lithium battery fires don't spread in the same way as oxygen-based fires.  Battery Manufacturers will confirm that it's O.K. to light a match - but god help you if you violate the specifications.

[KNZ]

Mostly been lurking the past few months but I feel the same as Hsiang, disappointed. Mostly all I'm seeing here lately from not only Kingsong, but other manufacturers, is disappointment. Things like this just make me think staying with the oldies but goodies is the best course of action.

Even if it feels outdated, it's still fun and gets me where I need/want to go confidently. I do hope manufacturers can pull through and produce safer first batch wheels so no one gets hurt that actually wants to spend their hard earned money on first dibs of the latest and greatest...

[Roadpower]

12 minutes ago, KNZ said:

Mostly been lurking the past few months but I feel the same as Hsiang, disappointed. Mostly all I'm seeing here lately from not only Kingsong, but other manufacturers, is disappointment. Things like this just make me think staying with the oldies but goodies is the best course of action.

Even if it feels outdated, it's still fun and gets me where I need/want to go confidently. I do hope manufacturers can pull through and produce safer first batch wheels so no one gets hurt that actually wants to spend their hard earned money on first dibs of the latest and greatest...

This is exactly why I have long not been an earlier adopter. Let a few iterations pass and pick from what has proven itself.

[Richardo]

This is the advice we should be giving newbies who want help choosing a wheel

6 minutes ago, Roadpower said:

This is exactly why I have long not been an earlier adopter. Let a few iterations pass and pick from what has proven itself.

 

[Seba]

6 minutes ago, Roadpower said:

This is exactly why I have long not been an earlier adopter. Let a few iterations pass and pick from what has proven itself.

...and this is why I recently bought an 18XL to be a successor of my 18XL. And in light of all recent incidents and my own experiences with S20, I'm happy that once again I choose a well known, matured, tested and absolutely reliable wheel.

[UniVehje]

Just now, Seba said:

...and this is why I recently bought an 18XL to be a successor of my 18XL. And in light of all recent incidents and my own experiences with S20, I'm happy that once again I choose a well known, matured, tested and absolutely reliable wheel.

I've been actually pondering the same. I had to buy a 16X as a spare wheel when my V11 was broken all the time and spare parts took months to arrive. V11 is now sold as I was planning to get the S20. Now we need a longer range wheel for my spouse and I really regret selling the trusty 18XL. I'm planning to buy one again for this season. 

This fire did not cause distrust for S20 but makes it certain a decent version of it will not be widely available for this summer.