Still no news on the official cause of the fire. I requested the incident report, and they said the investigation is still ongoing. They said to give them 6-8 weeks. Tesla finally sent a couple of technicians to our house. They ran a diagnostic check on our system that they said was good, but they wanted to turn the system off and lock us out to prevent us from turning it back on. After speaking to their supervisor on the phone, our system is back on with the verbal acknowledgment that we accept the potential risk.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
House Fire Very near to my installed Powerwalls
- Thread starter Wahiawahi
- Start date
What potential risk did you accept? The unit had passed their tests, so what would have been the next steps that Tesla would have taken if they had insisted that the unit was powered off until those steps were taken?
A scenario like this is one of my fears. It certainly isn't obvious if your PW lag screws penetrated an existing wire. Who knows if that is what happened here, but it could be.
@wwhitney what would be the best way to ensure that a long lag screw wouldn't hit existing wiring that you didn't know where was in the wall? Often enough the Powerwalls are located near the service, so the chances are that eventually someone hits a wire.
I wonder if the guy driving the screw could have noticed anything? Did a breaker trip while he was installing the units? Was there a moment where the lag screw got easier to drive because he found the hole through the stud? Is this something totally unrelated?
@wwhitney what would be the best way to ensure that a long lag screw wouldn't hit existing wiring that you didn't know where was in the wall? Often enough the Powerwalls are located near the service, so the chances are that eventually someone hits a wire.
I wonder if the guy driving the screw could have noticed anything? Did a breaker trip while he was installing the units? Was there a moment where the lag screw got easier to drive because he found the hole through the stud? Is this something totally unrelated?
I had some racks put into my garage. After they left, one of my circuits was dead. They sent out an electrician who found yep, a lag screw was put into wiring. They had to rewire around the issue, costing the company a grand from what they told me.
Another time I woke up and water was driving down the wall all over my new hardwood floors. Turns our a few months back, my contractor
had put in some shelving and put a nail into the copper water line. Took months to finally leak
At the first look of the pics I thought the fire started and was mostly contained in the one wall cavity defined by the corner of the structure and the first stud which is not behind the PWs. On a second look, there is a small section of Romex that is melted in the second cavity and also a pretty clear char line following the Romex in that second cavity. I believe the second cavity is behind the PWs. Could the short have started there and caused the wire in the first cavity to melt and start the larger fire? Regardless I would pull the breakers from those to circuits and have them checked for functionality.
I'm spent more time reading the NEC than installing lag screws, so I'm probably not the best one to answer this question. Obviously there is a risk that you can't eliminate completely.
Knowing the existing construction details and having a good estimate of the thickness of build up from the face of the stud to the face of the finish will let you right size the fastener length to avoid excessive embedment. But even so, you often want to get more embedment than the 1-1/4" that any cables through the studs are supposed to be recessed from the face of the stud.
Maybe there are wall scanners that can detect live electrical wires, not sure about that. They might not pick up a circuit that happens to have no load on it, though. Looking at the reverse side of the wall for any clues as to how circuits might be routed could help.
The only other thing I can think of is to run the lag screw most of the way in, and then back it out and look at the chips for anything metallic.. Not sure if that's wise from a structural point of view, and wouldn't be certain.
It's certainly the case that if a breaker trips while/after driving a lag screw, and the installer ignores it, that's incompetence at best. The issue in the OP's case is that under this fastener hit a wire theory, it presumably did so in a way that didn't trip the breaker, didn't kill power to the load, and if there was an AFCI breaker didn't cause any of the arcing patterns it was programmed to detect. Instead it just caused a high resistance point in the circuit, which under sufficient load got hot enough to start a fire. Again, just a theory.
Cheers, Wayne
Agreed
There are live circuit detectors built into stud finders these days, no load needed.
If lag is predrilled, the chips could indicate problems. If directly driven, I think it would just displace the wire/insulation unless there was a borescope handy.
Beyond that, the questionably useful things I can think of are:
Check lag bolt to ground resistance (ground and neutral check)
Check lag to ground voltage (assuming circuit is on)
IR camera before (find runs)/ after
FWIW, I looked at the manual for one of those (Bosch GMS120) and it says that the maximum scanning depth for live wires is 2". Even if that's conservative and it's 3" or 4", that would not be enough for longer lag screws. Do you know if there are detectors that can reliably do, say, 6" scanning depth for live wires?
Cheers, Wayne
Edit: PS Actual lag screws always require predrilling, but engineered lag screw alternatives that don't require predrilling (Simpson SDWS, some GRK fasteners, etc) are quite popular. But the ability to look at the chips from predrilling is perhaps an additional plus to predrilling for those as well.
(Edit: miswrote ultrasound. I meant radar, thanks @Vines)
Personally, I think that the newer, and, yes, more expensive, wall scanners are the way to go. They can find studs, wires, even PEX and PVC pipe. I'm thinking of the radar (RF) based ones, not AC based.
I do wonder if Tesla actually checked the torque on the wall fasteners, or backed out any of them to check for arcing, near the origin site.
Over the years, I have heard and seen too many accidents from nails and screws penetrating pipes and wires, most often with a delay, like @h2ofun's problem. I don't really fathom the "well I have never had a problem before" attitude from folks putting fasteners into walls. Problems happen. I don't put fasteners in walls without checking anymore.
I do understand why some AHJs require conduit, which lowers the odds, but not entirely, especially with nailguns and impact drivers. Nail plates over romex and pipe locations always seem like a good idea to me.
All the best,
BG
Personally, I think that the newer, and, yes, more expensive, wall scanners are the way to go. They can find studs, wires, even PEX and PVC pipe. I'm thinking of the radar (RF) based ones, not AC based.
I do wonder if Tesla actually checked the torque on the wall fasteners, or backed out any of them to check for arcing, near the origin site.
Over the years, I have heard and seen too many accidents from nails and screws penetrating pipes and wires, most often with a delay, like @h2ofun's problem. I don't really fathom the "well I have never had a problem before" attitude from folks putting fasteners into walls. Problems happen. I don't put fasteners in walls without checking anymore.
I do understand why some AHJs require conduit, which lowers the odds, but not entirely, especially with nailguns and impact drivers. Nail plates over romex and pipe locations always seem like a good idea to me.
All the best,
BG
Last edited:
This d-tect version is more expensive version with radar. There is a 120 and 200 versions with different scan depths up to 7.9" in dry concrete. For the cheaper version, the maximum detection depth is 2.36" in drywall.
12V Max Lithium-Ion Wall/Floor Scanner with Radar Kit 2.0AHDefault Title
This Bosch 12V Max lithium-ion wall and floor scanner (D-TECT200C) features radar technology, which is accurate and reliable, indicating the correct material. The Bosch D-TECT200C 12V Max Wall/Floor Scanner delivers advanced precision.
www.toolbarn.com
Even for the more expensive version, the maximum detection depth is 3.15" in "drywall" (by which they mean any light frame construction).
Cheers, Wayne
I don't own the Bosch, but I can say that the one I own has had no trouble finding studs and wires. I haven't tried it through stucco, nor tried to find PVC pipes, though my underground pipe locator can find those. (As an aside at least some of the underground utility locators units can search walls as well. I'm curious now, and I'm going to see if I can find rebar with it.)
Seriously though, an installer could buy a $60 stud finder, buy a decent horoscope, drill a "safe" hole between the studs, and have a look see, if it didn't trust the stud finder. They are going to put a Powerwall mounting plate up that would cover the inspection hole patch.
I have lived through roofers using the wrong length nails and solar installers missing beams on an exposed ceiling. Those expenses add up over time for the contractors. Penny wise, pound foolish?
All the best,
BG
Seriously though, an installer could buy a $60 stud finder, buy a decent horoscope, drill a "safe" hole between the studs, and have a look see, if it didn't trust the stud finder. They are going to put a Powerwall mounting plate up that would cover the inspection hole patch.
I have lived through roofers using the wrong length nails and solar installers missing beams on an exposed ceiling. Those expenses add up over time for the contractors. Penny wise, pound foolish?
All the best,
BG
Agreed, which at least would be long enough to scan for a 2.5" embedment requirement.
I don't follow you. The detection depth would be from the wall finish. The embedment depth would be from the face of the stud. So unless you know the wall finish over the stud is 0.65" or less (very unlikely if installed on the outside of a wall that has sheathing) then a 3.15" detection depth would not suffice for a 2.5" embedment depth.
Cheers, Wayne
You could be right, hard to say for sure without using it on some walls. Thicker walls you are correct might not detect anything or only detect limited depths. If there was 1" combined of sheathing and siding, then you are missing 3/8" of detection depth from the manual.
Yeah, I should have said if "screw" is predrilled. Technically, lag screws need two drill bits, one for threads and one for the shank. Given the PW installation manual only calls for 1/4 inch wood screws, I'm guessing engineed versions are preferred. 3/16 hole is probably too small for a borescope.
Interesting on circuit detection depth, didn't realize they were that shallow.
