Tesla Model S Garage Fire – Official Incident Report

DEC 23 2013 BY ERIC LOVEDAY 39

Official Incident Report of Tesla Model S Garage Fire

Official Incident Report of Tesla Model S Garage Fire

Below you’ll find the “narrative” portion of the official incident report—filed by the Orange County Fire Authority—on the Tesla Model S garage fire that grabbed headlines last week.

UPDATE:  Tesla has since made some helpful changes at their end to help prevent situations similar to this from occurring   Check out our post entitled “Tesla Model S Software Update 5.8.4 Reduces Charging Current by 25% If Input Power Fluctuations Are Detectedhere.

Page 1 of Narrative Section of Incident Report

Page 1 of Narrative Section of Incident Report

Page 2

Page 2

Page 3

Page 3

Page 4

Page 4

Page 5

Page 5

Page 6

Page 6

Follow this link to view the incident report in its entirety.

Categories: Tesla

Tags: , , ,

Leave a Reply

39 Comments on "Tesla Model S Garage Fire – Official Incident Report"

newest oldest most voted

I seems Tesla has updated its Model S software in response to this garage fire. Tesla rolled out the 5.8.4 over-the-air software update for Model S this past weekend. With this update, if the car detects a fluctuation in the current, it will automatically lower electrical current by 25% as a safety measure. Thus, if the Model S is charging at 40 amps when a fluctuation in current is detected, it will lower the charging rate down to 30 amps.

http://www.teslamotors.com/forum/forums/new-tesla-software-update-newer-58

But all that does is make installations that have very long runs with large pressure (voltage) drops safer. A proper installation with relatively large voltage drop is no more dangerous or safer than one with no voltage drop. The voltage will drop when the car goes to charge, and then it will revert to 30 amps, making charging safer, which could be done on other cars (making them safer) at any time. I know the purpose of the software change is to hopefully detect a connection getting worse. This might work some of the time, but it is rather like Tesla’s fix of the model S battery to have “Tesla Air Suspension – Now with more Air!”. Installations with very short runs from a very stout feeder remain unprotected, since a drop at the outlet/attachment plug can still generate alot of localized heat without tripping the point in the software to reduce the current to 30 amps. To me its a bit nonsensical, although granted it may catch a minority of bad connection problems. I think it is more likely that whenever something else in the house kicks on, the Tesla will go down to 30 amps. One thing for… Read more »

I think you mean a fluxuation in voltage. The current is regulated by the car itself to be either 30 or 40 amps no matter what the incoming voltage. But if it is truly a fluxuation in current of a spikey nature that is the tripping point, then that would imply some arcing is going on, there being no evidence of arcing in this particular fire. Strictly overheating. So if the thing trips by step-changes in current, then this fire would have started with the Tesla happily drawing 40 amps as long as it could.

The wording is slightly ambiguous, but the fire dept seems to be saying officially that the cause of the fire was the “high resistance connection at the wall socket or the UMC of the Tesla charging system plugged into the… wall socket”.

The fact that a licensed electrician installed an additional 100 amp sub panel and then ran a circuit to the wall outlet does not at first blush seem like a rickety installation. No wire sizes or types were mentioned, but there was no mention of arcing so it seems to lead to only one conclusion. The part I’m unclear about is there are several feet between the attachment plug and the UMC body itself. I thought all the problem historically was strictly at the attachment plug, I wasn’t aware that the UMC itself had ever melted.

“V” shaped burn patterns were observed at the “vent”, coming from within the wall, and the “wall outlet”. Insullation may have melted, before eventually arching. How are wires, that close together, going to melt insulation without eventually arcing?

The report said from the car to the “control box”, everything was fine. I took that as the UMC controller, not the dongle between it and the wall socket, FWIW.

How? Easily…. The wires are separated by the insulation. As long as there is no tension on the wires, there is nothing to make them touch.. In fact once they did touch, there would be no ‘arc’ but the breaker would immediately trip.

To form an arc, there must be an initial contact that was later seperated. That apparently was not the case here.

I’ve seen Nema 50 amp outlets running on 12 kw of load (50 amps) and they don’t seem very warm at all. So my basic question here is, what is different?

The recepticle is Underwriter’s Laboratories Listed. Is Tesla’s Male Attachment Plug?

Here’s another melted UMC adapter (adapter socket side melted). The cable and adapter was replaced and then turned out to be a receptacle side problem (receptacle side melted a week later, charging at the same socket).
http://www.teslamotorsclub.com/showthread.php/19576-Burned-220V-Adapter

Then another receptacle one in the same thread:
http://www.teslamotorsclub.com/showthread.php/19576-Burned-220V-Adapter/page2?p=472370&viewfull=1#post472370

So 4-5 UMC adapter melt (not sure if the one I posted should be counted) and 3 receptacle melts.

The OTA software update seems ok but also the UMC connection design seems undersized for the 40A current that is flowing through them for hours. I suspect a UMC re-engineering is in order.

The Tesla software update is just a Band-Aid. There still is the underlying issue of bad Tesla hardware/electrical connections.

Just because the a few wall oulets failed, doesn’t mean it was the fault of the wall outlet. The way Tesla has designed its charge connectors, is likely to put undue stress and possibly cause wall outlets to fail. Their adapter acts as a lever to increase the force on an outlet, so someone pulling on the cord, the weight of the module, and anyone tripping or tuging on the cord, the force is multiplied. The typical configuration as many connections close together which concentrates the heat and increases the odds of a bad connection and increases the odds of a fire. Most of the Tesla charger connection failures that I have seen, have not been wall outlet failures. Most have been the fault of Tesla charger connections.

It surprises me that such forensics can be performed without a statement of adequacy with respect to the wiring, its length and gauge.

The fact that a licensed electrician installed the above plus installed a 100 amp subpanel (200% of what is required), tends to indicate those things were much more than adequate.

Even pros make mistakes, and one cannot assume these things. I installed my NEMA 14-50 for my Model S myself. On the back of this outlet, there are usually screws at the wire terminals. If these screws aren’t tight (or if they work themselves loose over time), then resistance at those connections goes up significantly, generating heat. No discussion of this in the report?

Either way, this is pretty conclusive that the car itself was not the direct cause of the fire.

Wishful thinking. Far from conclusive. And this is not the conclusion of the fire dept. So you are putting yourself as judge and Jury over: 1). The Licensed Electrician (He had to have a certain level of proficiency to get his license in the first place). 2). The Conservativeness of the design (100 amp subpanel is 200% of what is required and does not seem to be a guy trying to cut corners). 3). The inspector who would have had to pass the installation (routine). 4). The fire dept who has already given their conclusion. So you want to say you are automatically superior to this troop of experts? What conclusive evidence do you have that automatically invalidates something at least 4 other professional people have signed off on? There is a very slim chance you still may be right, but again, Tesla’s actions seem to indicate they realize there’s a lack of foresight on their part. This is the second time they have made changes without accepting culpability, again there not being iron-clad proof that they are in the wrong, but it doesn’t look good. I assume this licensed electrician has also installed baseboard heating systems and electric water… Read more »

You don’t think it’s possible for a licensed contractor to make a mistake? He could be very competent and doing it right 99% of the time, but just missed something here, or used a part/wire that was substandard without him knowing.

Go watch Holmes on Homes to see what licensed contractors can mess up.

That’s not the argument here. It would have to be the Licensed Electrician AND the Inspector made a “mistake.”

You might have a point, but I’ve seen a Model
S “in action” at an authorized Tesla Service Center. and the plug seemed too warm for my liking.

As far as making a mistake, several bloggers chewed me out for being concerned about such a trivial thing.

Maybe, but then again, I’m not likely to have an electrical fire around me.

1. Even licensed electricians make mistakes. I have heard of many.

2. 100A subpanel doesn’t have anything to do with the 50A circuit installed for the NEMA 14-50 outlet.

3. Most inspectors I have dealt with don’t do anything but a very brief glance over the work.

4. All the fire department does is determine an approximate cause, which they did and again, isn’t relevant here.

Well, Sorry Dave R, I’m calling ’em as I see ’em. And I wont have anything in my house that generates unnecessary heat.

Having said that, the fix is easy. A good nema 14-50 p is under $20.00.

I totally agree. The description of the exterior walkaround of the house was great, but I didn’t once read about the gauge of the wiring in the circuit or the tightness of any electrical connections to the NEMA 14-50 outlet. Those are both significant possible sources for the fire.

The problem is that once the outlet burned up, that makes determining whether or not the set screws were sufficiently torqued before the incident very difficult to determine.

The way Tesla has designed its charge connectors, is likely to put undue stress and possibly cause wall outlets to fail. Their adapter acts as a lever to increase the force on an outlet, so someone pulling on the cord, the weight of the module, and anyone tripping or tugging on the cord, the force is multiplied. The typical configuration has many connections close together which concentrates the heat and increases the odds of a bad connection and increases the odds of a fire.

I’ve noticed in passing several bloggers are concerned about the length of the wire. As long as a given section of wire is not overloaded, it doesn’t matter whether the wire is 10 feet long or 500 feet long. A 500 foot run provides much lower voltage to the car, but that in no way makes things either safer or more dangerous. A 40 amp draw is still straining the connection points equally in either case.

One might have a strand of truth saying an installation with very large voltage drop will take longer to charge up, which is true enough. But these installations are supposed to be installed robust enough for 24 hour a day usage. That is why 50 amp construction is required for a 40 amp loading.

Actually, a larger gauge of wire is required by code when the run is longer. So yes, the gauge of wire is very significant.

That is incorrect. A larger gauge wire is only required on a run 100 feet long, and then that’s only to the first outlet, and besides this it has nothing to do with the heating density of the wire.
To make the statement that you have indicates you have no knowledge of the basic physics of resistance. A 10 foot length of wire will not generate more heat per foot then a 500 foot length of wire will generate heat per foot. The wire has no idea how long it is! It just reacts to the 40 amps going through it. I’ve never as a a matter of fact, seen an installation fail due to the length of its run.
It is much more correct to say that a larger wire size is recommended in very long runs to keep the voltage drop to a reasonable percentage.

I’m not familiar with Orange County Local codes, so please indicate section and verse of the requirement of the pertaining “code”, where you say this is required.

While I am disputing nothing you wrote, Bill, if a run of wire is coiled (a specific case of “longer”) (which should not be done for in-use wiring) then adjacent wire can lead to increased heat (i.e. multiple contributions to an area). Also, the generated heat must be allowed to dissipate, so any wiring should be installed such that there is enough air around to act as a sufficient heat reservoir.

Yeah Gene, but that is not the case here, and I think you knew that before you typed what you said.

I’ve had to replace many coiled up spring trouble lights blobs of wire where someone plugged in a 1500 watt heater to stay warm without reeling out the trouble light first.

I would guess they would need a licensed electrician to determine if the installation was up to code or not, which I am guessing is what the insurance company will do.

O

They’ve already had a licensed electrician install it.

I was instructed by a master electrician to use metal outlet boxes. They don’t melt easily. The fire might have been contained inside the box until the breaker tripped from the wires melting together and/or contacting the grounded box.

From what I understand, the Tesla UMC (Universal Mobile Connector) was completely fine from the “control box” (inclusive) all the way to the car end (which means the issue was not there). What melted was all the insulation inside the junction box all the way to the “control box”. That seems to isolate the issue to between the wall socket and the UMC control box cord (AKA “high resistance connection”). That sounds a lot like a badly tightened or inserted connection, although after the fire it’ll be hard to tell.

Reads more like the Tesla adapter plug which comes with every car was the high resistance point.

Not really. Everything in the junction box was melted, and so was insulation on the cord all the way to UMC box (so we can eliminate the UMC box and car as the cause), so it could have been any of these:
A) something in the junction box (including improperly torqued screws for the connections)
B) loose connection between wall socket and adapter plug
C) loose connection between adapter plug and UMC cord

The investigator said there was too much damage to tell what’s the point (it’s really tough to tell the difference between B & C after some damage, as the heat point is close to each other; and in this case the investigator can’t even eliminate A).

This is what Tesla’s Legal Dept is banking on. There may be a model S in my future, but if things remain the same, the attachment plug is getting chopped off and a spec grade nema 14-50p will be put in its place for $20.

That seems a bit extreme (you lose a lot of utility without the adapter ends). A lot of the problems seem to stem from a loosely connected adapter plug (gap in between from user not fully connecting it, which rev B was supposed to address with a metal clip) or from a worn down receptacle (in which case the receptacle side would melt). As long as you ensure there’s a tight connection, it shouldn’t be a problem (shooting a laser thermometer at the plug end after a long charging session should tell if there’s an issue).

What bothers me about all the “My Tesla Right or Wrong!” guys is that, they don’t seem to have a natural ‘feel’ for electricity (and surprisingly, that also applies to all the designers of the UMCS – they are on their 3rd major try, and with several minor revisions on each of the majors). Just take a look at that crappy connector they supply free of charge with Model S. I took one look at the thing and said, ” You guys are trying to shoehorn 40 amps for 10 hours every day on this piece of crap, what with its dinky button pressure contact ‘connector’;? Compare that with a plain old UL listed 50 amp attachment plug, as well as the surface mating areas between the attachment prong and the female outlet double – pronged surface, which have proven themselves by being in existence basically unchanged for the past 70 years. As far as the decrease in utility, the only thing I loose is 110 volt operation. I could also splice in a nema 5-15p with cap to avoid losing the functionality. And then when a garage finally is toasted, everyone blames the Licensed Electrician, Fire dept, inspector, and… Read more »

Resistance is not futile, its voltage/current … arc arc arc
Happy holidays one and all!