That’s true. A frayed cable would be far more likely to short with possible sparks, or, under really unlucky circumstances, resistively heat, maybe even incandesce in the final moments. A sustained arc might happen as a consequence of the fire but probably didn’t start it to begin with.
I’m not saying it couldn’t happen but the dangerous, accidental arcs seems so rare and the buggy detectors so unreliable, I just wonder if they’re worth the trouble in a typical household.
My own household, by the way, is not typical and I do have an above average number of high voltage goodies that might make for some interesting arc detection reports. I doubt I could make an arc with the nails and wet wood, though. I’m willing to try if there’s interest, though, I have an isolated variac and other tools. And wood.
So my house was built in 2005, and doesn’t have any AFCI breakers anywhere on my electric panel that I can see. Is this something I should address? If so, it looks like a fairly pricey fix, since each breaker appears to cost around $50 or so, and I have something like 35 breakers.
In the meantime, maybe I could install the Ting electrical monitor mentioned upthread? Would that be a good idea?
There’s a lot of wisdom there! And as a former Electrical Officer on a submarine, it brings back a lot of memories for me. The last commandment even mentions the Straight Dope (sort of).
Yea, a frayed cable can cause parallel arcing. These arcs tend to be very energetic. The best case is for the conductors to weld together very soon after the arcing occurs, thereby causing the circuit breaker to trip.
Most research into arcing on branch circuits is not for parallel arcing, though. It is for series arcing. These occur when there is a loose/bad connection, like a loose wire under a screw head, or a poor connection between the plug and receptacle. Though they tend to be less energetic than parallel arcs, they are more insidious because they will no trip a standard circuit breaker (unless, of course, they cause a parallel arc).
A more honest question: How many needless walks to the fuse box before I yank the silly thing? Answer: probably about half a dozen, less if it cuts off at a really bad time. A smoke detector that makes noise every time you preheat the oven won’t last long, either. It simply isn’t tolerable.
When I was living in a college dormitory, we had a few weeks of hysterical latenite fire alarm pulling pranks. After the second false alarm, I stopped evacuating. Most people did, I think. It’s like 10 degrees out, there’s never a fire, and I’ll take my chances.
At the old house, the washing machine was running one day, and the area it was in filled with smoke. Rubbery smelling smoke, so I thought it was a belt. When I pulled the unit away from the wall to look at it, I noticed the outlet was black and charred. There was too much insulation left on the hot wire, and the insulation was partially inserted between the screw and the backing, so that the copper was not making a good connection.
I had never touched that outlet and had been in the house for years at that point. Don’t know why it decided that was it’s day, but I was lucky to be home.
And, I learned that modern washing machines don’t have belts.
Here’s the problem with AFCIs. One of two things often happens when an AFCI trips all the time:
Example 1:
There’s no bad/abnormal electrical arcing going on, but the AFCI trips anyway.
An inspection does not reveal any problems.
AFCI is reset.
Branch circuit works for a while.
Go to 1.
After a while, the homeowner and/or electrician concludes these are nuisance trips caused by normal arcing (motor, etc.). Replacing the AFCI does not fix the problem. AFCI is removed and replaced with a standard circuit breaker. Problem solved.
Or…
Example 2:
There is some bad/abnormal electrical arcing going on somewhere in the branch circuit.
The AFCI detects the arcing.
The AFCI trips.
An inspection does not reveal any problems.
AFCI is reset.
Branch circuit works for a while.
Go to 1.
After a while, the homeowner and/or electrician assumes these are nuisance trips. Replacing the AFCI does not fix the problem. AFCI is removed and replaced with a standard circuit breaker. Problem solved… for a while. Best case: the arcing leads to an over-current situation and the circuit breaker trips, and then (hopefully) the problem will be easier to find. Worse case: arcing causes a fire.
So that’s the problem with AFCIs… when they trip, you often can’t find the problem. And eventually you’ll get frustrated and replace it with a regular circuit breaker. This might work (example 1 above) or might be a bad thing to do (worse case in example 2 above).
I really have no good advice other than to try and identify the load that the AFCI thinks is arcing. Of course, it could be somewhere in the wiring. Or it could just be a poorly-designed AFCI. I dunno…
That is a great summary of my questions. Further, I don’t want the insurance company gets the wrong idea about all my false positives.
A better device would sound a beeper so the homeowner makes the connection between arcing and the garage door opener or the washer kicking on or something.
A better device would stop the current immediately, to prevent a fire, but not have many false positives. And i expect AFCI devices will get there.
The Ting theoretically sends an alert to my phone if there’s a problem. So far, the only problem it’s detected was “power was out for a few seconds”, which i didn’t need an extra device to be aware of. But i bought it to downgrade from “protects you from danger” to “warns you of danger” because i got too many false alarms with the AFCI.
I have one data point about circuit breakers. The ones in the house where I grew up did NOT have a tripped position. If it tripped, the lever flipped to off, and you just flipped it back to on. My bedroom was on the biggest circuit in the house and I tripped it fairly often so I was pretty familiar with the panel. I think it was upgraded from fuses in the 1970s when an addition was built on the home. When I later encountered a breaker with an actual tripped middle position, I remember trying to flip it to on and it just went back to the middle. Then I flipped it off and on and thought “huh, that’s neat.”
As I mentioned upthread, this was my situation. And with the breaker in a bedroom, there were times this was extremely inconvenient. Do I wait until tomorrow, or do I wake him up?
In my opinion, the tech is not yet ready for prime time. I’m all for it. Once it works.
I got a Ting device from my insurance company shortly after my full electrical panel upgrade. I had arcing in one of the connections from the main panel to the sub-panel which was causing flickering lights and brownouts. I’m sure Ting would have blown up my phone. So far no arcing alerts.
The most useful thing about the blackout alerts from Ting is that it lets me know how big the problem is. When it says me and 600 of neighbors are out, I don’t need to do anything or worry about it. When it says just me, or just me and a very few neighbors, I know to contact the power company.
The Ting device is only in a 120 volt outlet, how does it know what is happening on the other leg?
As for AFCI, by code, when I got the panel replaced in my condo all of the breakers are AFCI. In two years I’ve never received a complaint from the tenant about nuisance trips, and I specifically asked for that type of information while the work was still under warranty. I’ve also never had nuisance GFCI trips, so maybe I’m just lucky.
With the advent of remote-reading “smart” meters, ISTM the power company could, in theory, pretty instantly map any outage by pinging a bunch of meters and seeing who reports back and who doesn’t.
Other than a failure of my own branch circuit, ISTM they ought to learn about any failure in near real time.
Does anyone know how much that’s true, or does the power company pretty much rely on crowd-sourced complaints to at least draw their attention to an outage?
Obviously there are as many different answers as there are local power companies and local infrastructures, but I wonder if there’s any industry standard or consensus behavior?
I can’t speak to the industry standard, but the last two times I had a power outage, I got a text from the power company within 5 minutes of my power going out. I doubt that there was much manual processing happening in that 5 minutes.
I know nothing about the Ting device or how it works. But I think that it probably detects issues on the other leg by monitoring voltages on the neutral.
I will suggest that arcs are short-duration but wide-spectrum events. See also lightning & primitive spark-gap radio transmitters. The Fourier transform is a little energy spread across a lot of frequencies.
Such that an arc on the, e.g. A phase will still induce some RF noise into the B phase of a typical residential split-phase power installation.
Getting the thresholds right to detect opposite phase events without too many false alarms is probably pretty challenging. Ting alleges they’ve done so. If a user was concerned enough about opposite phase detections they could put a Ting unit on each of their phases. Which itself would be an interesting experiment. If one unit alarms significantly more often than the other, we could conclude that a) you have a problem on one phase, not the other, and b) Ting hasn’t quite gotten opposite phase detection down as well as they’d like to claim.
As mentioned upthread, the bitch about Tings or AFCIs is they provide little to no trouble location information, such that any false alarm rate above zero is counterproductive for most users who will simply ignore or disable the alerts since they’re not practically actionable.
I wonder if a TDR (Time-domain reflectometer - Wikipedia) would be of any value in locating power-side arc fault sites? With sooo many connections being part & parcel of an houseful of wiring, I sorta doubt it, but I’m not in that biz and never have been.
There’s a lot of wisdom there! And as a former Electrical Officer on a submarine, it brings back a lot of memories for me. The last commandment even mentions the Straight Dope (sort of).