A Model S caught fire while supercharging in Norway (link in Norwegian)

[Ingineer]

Yes, I called it!

Let's say one of the terminal bolts on the SC contactors was over or under tightened at the factory, or during a recent service. That terminal could have enough resistance to heat up the contactor to the point of ignition of it's plastic housing. Some of those contactors are hydrogen filled as well (not sure about Teslas though). There is plenty of plastic in the JB for fuel. All that has to happen is the box get hot enough to ignite the foam seat cushion which is right on top the box.

Over-tightening the terminals on one of those Tyco contactors could distort the internal contact enough to cause appreciable resistance gain. Let's posit that that was a CPO car, and maybe Tesla serviced the car before sale. They could have worked in the HVJB for several reasons, that's all it would take.

My money is on the HVJB. 300 amps is definitely enough to generate a lot of heat! A 3 volt drop over a small area might not be sensed and that's 900 watts, easily enough to light the contactor plastic on fire.

 

[ecarfan]

Tesla has (almost) found the reason for the fire:
Google Oversetter

Is "the distribution box" something that was installed by the local electric utility company, or by Tesla?

 

[Yggdrasill]

Is "the distribution box" something that was installed by the local electric utility company, or by Tesla?

Every Model S has a High Voltage Junction Box. It sits between the chargers underneath the rear seats, and distributes power between the inverter, battery, chargers and charge port. Obviously this falls entirely into the domain of Tesla.

 

[Model 3]

Is "the distribution box" something that was installed by the local electric utility company, or by Tesla?

It is "the distribution box" in the car - or HVJB as it is called.

Edit: Yggdrasill beat me on it...

 

[Ingineer]

Yes, the HVJB takes the power from the charge port and decides where to send it. If it's "normal" AC charging, it sends it to the on-board chargers which then convert the output to the proper voltage DC, which is then sent to the battery pack via the HVJB. In the case of supercharging, the HVJB has several large contactors (basically motorized switches) that send the DC supercharging power directly to the battery pack. When this occurs there is hundreds of amps flowing through there. Any slight added resistance can add up to a LOT of heat in no time. The original (gen 1) HVJB does not have any temperature sensing, whereas the gen2 units introduced in 2014 do.

 

[apacheguy]

So it sounds like they'll modify the voltage monitor which reduces the charge current at home to apply to supercharging as well. Voltage drops caused by increased resistance would be sensed and charge current would be reduced.

 

[Skotty]

eh....so when in 2014 was the gen2 HVJB introduced? I'm picking up a 2014 CPO tomorrow and this would be good to know.

 

[Ingineer]

I can't say for sure. Seems like sometime around VIN 30000. It's easy to check with a peek under the rear seat. I think all D cars have them as well.

Just lift up the rear seat cushion (pull up on each side with a sharp yank), then if the silver box in the middle has a lid with a large raised section on the front it is G1. If it has a relatively flat cover, it's G2.

You can also tell by the large charger module on the passenger side. If it has a series of screws visible on the top of it's lid, it's G1 style. The G2 has it's screws on the side.

 

[Skotty]

I can't say for sure. Seems like sometime around VIN 30000. It's easy to check with a peek under the rear seat. I think all D cars have them as well.

Just lift up the rear seat cushion (pull up on each side with a sharp yank), then if the silver box in the middle has a lid with a large raised section on the front it is G1. If it has a relatively flat cover, it's G2.

You can also tell by the large charger module on the passenger side. If it has a series of screws visible on the top of it's lid, it's G1 style. The G2 has it's screws on the side.

Thanks. I'll be supercharging rather frequently, so if it's a G1 I might want to keep a closer eye on it. I'll check it out. Sounds like it will probably be a G2 though, as mine is P41819.

 

[Ingineer]

Thanks. I'll be supercharging rather frequently, so if it's a G1 I might want to keep a closer eye on it. I'll check it out. Sounds like it will probably be a G2 though, as mine is P41819.

Yeah, I think you're OK.

Even with a G1, I wouldn't worry about it. It's highly unlikely. You are many times more likely to die in a fatal car accident on the way to/from the SpC!

 

[DNAinaGoodWay]

Tesla will update the Model S software for safer charging following a Supercharger fire - Electrek https://apple.news/AgPvs4culM8O6Cff8nJhdEw

 

[Johan]

Tesla will update the Model S software for safer charging following a Supercharger fire - Electrek https://apple.news/AgPvs4culM8O6Cff8nJhdEw

My S which is a very early EU build had to change charge port due to a bad connection and I saw from the work order they'd changed my HVJB to gen 2.

 

[wcalvin]

For both on-board and supercharger charging sequences, the system ramps up the current. I had assumed that that was to check voltage drops at places like the HVJB, just in case of bad contacts and the ensuing heat issues.

No so? If so, what's new about this--are they just going slower up the ramp?

 

[Johan]

For both on-board and supercharger charging sequences, the system ramps up the current. I had assumed that that was to check voltage drops at places like the HVJB, just in case of bad contacts and the ensuing heat issues.

No so? If so, what's new about this--are they just going slower up the ramp?

Yes no apparent reason why they would have to limit charging, just ramp it more carefully or setting tighter tolerances for what's accepted by the firmware as normal. In my work order it says the testing they did was: "test driven the car, checked that the insulation values have been within normal the whole time".

 

[tom66]

I also suspected the HVJB (although not the battery bolts at fault of course), interesting to see Tesla came to the same conclusion. It would be good to know what the root cause was but I'm guessing the car was too badly destroyed to tell or Tesla don't want to say.

My prediction: Early 2013 P85, second owner, Norway. The drive unit will have been replaced due to high failure rate. As part of the service the HVJB may have been serviced. During service, tech disconnected battery bolts. Later they were reconnected but inadequately. As a result, overheating occurred during supercharging. (Could equally apply to HVJB connecting to HVAC battery heater/AC chiller)

 

[Lump]

Yeah, I think you're OK.

Even with a G1, I wouldn't worry about it. It's highly unlikely. You are many times more likely to die in a fatal car accident on the way to/from the SpC!

Sounds like those of us with the G1 won't see much benefit from the upcoming OTA since we lack the sensors, based on the way Tesla handled the 1 seatbelt discovery I assume a recall would be issued?

 

[Johan]

Sounds like those of us with the G1 won't see much benefit from the upcoming OTA since we lack the sensors, based on the way Tesla handled the 1 seatbelt discovery I assume a recall would be issued?

I think we lack the temp sensors but not the ability to detect failing insulation through analyzing the voltage and current flow. Thus it becomes possible to target it by making the software less tolerant to deviations. And then the boxes may get changed over time when in for other things.

 

[Ingineer]

The only way I can think of detecting problems in G1 is to compare the voltage at the pack to the voltage exiting the supercharger and if that voltage drop is too great then either reduce current or shut down the session with an error (or both).

 

[Johan]

The only way I can think of detecting problems in G1 is to compare the voltage at the pack to the voltage exiting the supercharger and if that voltage drop is too great then either reduce current or shut down the session with an error (or both).

Like I said, I'm sure they've been doing this all along and that the software update would be to tighten up the tolerances.

 

[wdolson]

I believe the car was a CPO. I know it was purchased only a couple of days before the fire. Maybe the technician who went over the car accidentally left something in there that caused a short the first time it was plugged into a supercharger? That would be my first guess. I'd look at the service logs from when the car was prepped for CPO sale and see if anything was opened up in the charging circuitry.