We've had a 928 fire, and here's what we learned...
 

This is a longish post. But you are going to learn about a bad 928 fire, what caused the fire, and what we’ve made to help insure it doesn’t happen to other 928’s.

We have had a customer’s 928 start on fire in our garage. The fire started in the middle of the night, and no one was here. Fortunately, no one was hurt. The 1988 928 was totaled.

We only discovered that we’d had a fire at 7:30 AM the next morning, when one of my employees came in to work. The whole place was filled with black acrid smoke, as the RF tire and the urethane bumper had gone up in smoke during the fire.
The only reason my building is still standing, they tell me, is the result of the maintenance and upkeep we had recently done. I had just replaced the old garage doors and the service door with new ones that had new weather stripping, of course. I’d also sealed all the pass-through’s in the garage ceiling for the pipes and conduit with foam-in-place caulk so the tenant upstairs didn’t have to smell automobile exhaust at night.

The Fire Chief told me that the fire likely self-extinguished because it suffocated – not able to get enough clean air to continue to burn. A review of our security camera footage showed the fire started at about 2:18 AM, and was out by 2:30 AM. He was exactly right. I am so incredibly lucky that this 106 year old building of ours is still here! We sustained $28,600 of damage to our facility – but it could have been much, much worse.

Now to the customer’s 928: we had just finished the work on it that the customer requested, and he was coming the next day to pick it up. I’d taken it for a test drive, all was well, but because it was going to rain, I decided to park it inside for the night.

This fire was investigated by three men: a fire origination specialist from each of two insurance companies, and the Fire Chief. All three experts agreed on the exact location of the start of the fire: the wiring harness just in front of the 14-pin connector had arced to ground, starting the fire. An arcing battery cable with 800 amps or more behind it can start damn near anything on fire. There is actually a hole in the steel inner-fender where the arcing of that unfused positive battery cable to ground literally burned a hole through to the tire-side of the inner fender. From there, the nylon inner fender was an easy thing to start on fire, then the urethane front bumper, and the tire.

Here are some pics including the exact location of the start of this fire:

https://cimg6.ibsrv.net/gimg/rennlis...1c329dfd6c.jpg
How the car was when we found it
https://cimg7.ibsrv.net/gimg/rennlis...5e521efde3.jpg
How the car was when we found it
https://cimg0.ibsrv.net/gimg/rennlis...4616f25473.jpg
How the car was when we found it
https://cimg3.ibsrv.net/gimg/rennlis...0885d2f510.jpg
The mess it left in the shop is too big to even photograph
https://cimg4.ibsrv.net/gimg/rennlis...55be00cb3f.jpg
The electricity shorted out at 2:28 AM
https://cimg0.ibsrv.net/gimg/rennlis...d8783fefaa.jpg
The car outside
https://cimg2.ibsrv.net/gimg/rennlis...3e8301d605.jpg
The car outside
https://cimg5.ibsrv.net/gimg/rennlis...b644c3d2e7.jpg
The specific origin of the fire
https://cimg6.ibsrv.net/gimg/rennlis...f9543b913c.jpg
The specific origin of the fire

How lucky were we to not have lost the whole building?

The previous HVAC person had plumbed the natural gas line to the furnace with flexible gas line designed just for this purpose. When the fire inspector saw that, he told me about how often those are compromised even in a little fire… A further inspection of the natural gas line showed it was within the last mm of being burned through in spots. Had the fire gotten a supply of fresh natural gas – it’d been all over.

Under the hood of the car, the plastic covers over the fuel rails were melted on the right side of the engine, and the OEM rubber fuel lines were visibly cracked. Had the fire gotten to the gasoline – well, you can imagine the rest.

We had to throw out so much inventory as a result of the corrosive nature of the urethane smoke! All the tools, even those in the tool box with the drawers closed, had and still have a coating of rust on them. We have tried to clean them, and lubed them up, but it’s insidious. Our lathe and mill got it bad. A new telephone in the shop, less than a month old, was destroyed because of the corrosive smoke alone. It wasn’t near the heat, it was just in the same room.

I had to throw out camshaft cores, brake rotors, calipers, used clutches, and all kinds of parts that I’d accumulated in 19 years of 928 work. We filled a steel and metal recycling dumpster.

https://cimg5.ibsrv.net/gimg/rennlis...3b14531cb6.jpg
flourescent fixtures melted and hanging from their cords
https://cimg8.ibsrv.net/gimg/rennlis...338772e7a2.jpg
The natural gas line that was very nearly breached by the fire.
https://cimg0.ibsrv.net/gimg/rennlis...79004cd4d9.jpg
https://cimg1.ibsrv.net/gimg/rennlis...527b0563ef.jpg
The corrosion ruined parts
https://cimg7.ibsrv.net/gimg/rennlis...98a889fc13.jpg
...and coated every tool with rust

Here are a few of the things that I learned as a result of this:

- Fires start on unattended cars all the time, the Fire investigators told me.

- The most common are cars with batteries that are remotely located, and have lengths of unfused primary battery cable running around the car. They specifically cited frequent claims for the Mercedes-Benz Sprinter vans, skid-steers, and cars like this 928 with batteries in the trunk.

- That almost always, when the car fire starts inside the garage, it takes the whole building with it. They told me I was extremely fortunate that our building was still here. Once the tire starts on fire, it’s usually SOL for everything after that.

- We learned that the gas and soot given off by the urethane bumper burning is toxic, carcinogenic, and corrosive. It’s really bad. Until we got the right VOC respirators, we could not been in the building for more than 10 minutes before our lungs would hurt.

- I learned I could GOOGLE “gases created by burning polyurethane” and find information everywhere. Try it yourself.

- I learned that starting around 2004 or so, many manufacturers with remote mounted batteries were installing master fuses near the batteries to prevent car fires when the lengthy battery cables would ground. Pic of the Master Fuse in my sons BMW shown below.

- …and I have started inspecting the primary wiring on these 928’s much more carefully! Note the pics of the 1995 GTS Engine Wiring Harness below – and what we found on the UNDERSIDE of it.

https://cimg7.ibsrv.net/gimg/rennlis...6c6a4c52df.jpg
This engine wiring harness looks pretty good...
https://cimg2.ibsrv.net/gimg/rennlis...b73875df1e.jpg
Until you turn it over! This one is from a 1995 GTS.
https://cimg0.ibsrv.net/gimg/rennlis...760380ab61.jpg
One of the BMW master fuse setups on a late-model Bimmer
https://cimg2.ibsrv.net/gimg/rennlis...21b81f768d.jpg
...and a close-up of it

I also noted two very distinct types of burned-out 928 pics on the internet.



There seem to be those that that likely were caused by a gasoline leak, often at the rear of the engine, like this one:

…and those that show no evidence of being near a gas leak, and I now feel are much more likely the positive battery cable arcing to ground, like these:


https://cimg5.ibsrv.net/gimg/rennlis...d94b3ed7e0.jpg
When we see fire damage like this at the back of the motor - we think "gas leak" and I would agree.
https://cimg0.ibsrv.net/gimg/rennlis...a241b40496.jpg
But now I have also learned that when I see fire locations in the fender - that do not involve the engine...
https://cimg2.ibsrv.net/gimg/rennlis...2f99bf6040.jpg
... I now know its likely an electrical fire caused by a grounded primary wire

Wow. Glad your building is ok. I found the same broken insulation on Daniels 89.

There is going to be a big market for wire harness in the not so distant future. Every car I have seen over the last year is in need of new wiring.

WOW! Thankfully no one was hurt and minimal damage.

…that led me to conclude that our 928 wasn’t than unique. And that we’ve lost more 928’s to the grounding of the primary cables then we might have previously thought! It might even be a larger number than we have lost to fuel leaks.

So, I developed this: a Master Fuse kit for the 928.

I have had this kit installed and on my own 1991 928 GT for over a month now with no issues. Specifically, I waited to let it get damn good and cold outside (we hit -20 F), the engine tight and the oil thick, before testing it by cranking over the 5.0 liter to make sure the highest current draw did not blow the fuse unnecessarily. There have been no issues.

See here: https://928motorsports.com/parts/master-fuse-kit.php

https://cimg4.ibsrv.net/gimg/rennlis...c8b969ed3e.jpg
https://cimg6.ibsrv.net/gimg/rennlis...215ca49e03.jpg
https://cimg3.ibsrv.net/gimg/rennlis...d32f470f3c.jpg

Glad nobody was hurt!
Glad it was an easily found source..
GLAD its an easily managed issue to prevent (Worst case fusing best case fusing and new cable)

Glad..everyone will eventually be made whole.

Carl,

Follow the current flow from the battery through the alternator connection to the jump post, then to ground where the primary feeders pass through the fenderwall. How did the fender liner catch? The heat from the shorted section is still physically/thermally isolated from the liner sections by the metal fenderwall.

That section of the body harness KS-1 connects the jump post terminal to the 30 bus bar in the CE panel. Carefully inspect the top of the CE panel where those red feeders connect, and follow the feeders forward through the firewall to see where the insulation damage ends. That may give you a better clue to where the fault actually occurred.

----

You are very lucky that the fire didn't get further back into the carbon canister and the vapor vent line from the fuel tank. I'd also offer a different opinion on why the fire eventually went out, as the total air mass in your shop is quite large compared with the mass of 928 consumed in the fire.


Glad everyone is OK, no animals were harmed during the filming of this episode, etc.


----

For those playing along at home, the amount of fault current available at that point in the harness is limited only by the battery capacity and the size of the conductors between the battery and the fault point. The smallest conductor in that path will offer the most resistance and therefore the highest temperature. The circuit is not protected by a fuse or circuit breaker, so any fault to ground will continue until a conductor fails or the battery is depleted.

Over the years, Alan and I have casually discussed this section of the primary current flow path, especially looking at cars that have additional electrical loads fed from the CE panel. Further discussion on this should continue in a separate thread.

Dr. Bob: The arcing of the short actually burned a hole THROUGH the inner fender well. I think my first pic was too small for you to see the hole. Try this one. That's how it was able to jump to the nylon fender liner on the other side.

Exactly right. If or when you accidentally ground your positive battery cable, you get it ALL. All the output the battery can produce, from 800 to 1200 amps (maybe more), all at once. All that will stop it is the wire melting though (breaking the circuit) or the battery going to zero volts.

One of the fire inspection agents asked me: "have you checked the battery voltage?" to which I said "I assume its dead." He told me thats what most would think, but no. He described how, once any one of the involved wires burns through, the circuit is open and the drain on the battery stops. Unfortunately, by that time, other combustibles are able to keep the fire going.

https://cimg7.ibsrv.net/gimg/rennlis...a6d696bf4e.jpg

Wow. I just checked mine, and the way the positive battery cable is routed at that location it looks "designed" to literally rub up against the edge of that mount for the connector, until it saws through the insulation!
You can even see in the pic where the paint has rubbed off!

https://cimg9.ibsrv.net/gimg/rennlis...883899841d.jpg
https://cimg5.ibsrv.net/gimg/rennlis...867dd121d6.jpg

Good catch, Jason! I hadn't seen that one before. I've always been concerned about the primary cable where it routes through the motor mount support, but Porsche added a heavy rubber cover there. Then there's where it crosses by the steering rack on the way to the starter - always hot, always "ON". Then the engine wiring harness and - the lead that our car burned - the large primary wire after the 14-pin connector.

My pics is EXACTLY on the opposite side of the hole you see in your fender liner metal!

Yikes...those pictures hurt to look at. Thank god it wasn't worse...