New Home: daily, random, multiple AFCI breakers tripping

afci

We have a brand new home in Puget Sound which we moved into 2 months ago. We are plagued with constant, random, tripping of multiple AFCI breakers on a daily basis.

Some days, just two or three breakers will trip. Other days, we’ve had as many as 7, 8, or 9 of the breakers trip throughout the course of the day. Every one of those 19 breakers will trip at one time or another.

The load panel is Square D, and has 19 AFCI breakers on the left side. The only correlation we can make, as far as what activity will trip breakers, is that it only happens when the American Standard TAM9 air handler/heat pump combination runs. If the air handler/heat strip runs, and the heat pump is offline, there will be no tripping.

The builder has put responsibility for resolving this problem at the feet of the electrical contractor and the HVAC contractor.

The electrical contractor, so far, has replaced every AFCI circuit breaker in the panel twice. He has run a new 240v load line from the service panel to the heat pump. The electric utility company has placed recording meters on our outside meter, and the local service distribution point for the past month and has recorded no anomalies.

The electrical contractor has claimed his lines are all good, clean, and solid and says “It must be something with the HVAC installation.”

The HVAC contractor has replaced the entire system, piecemeal over the past month. First it was the heat strip, then it was the heat pump, and this past week, they put in a new air handler, thermostat wire, and thermostat.

The daily, frequent, random tripping of breakers continues, and the HVAC contractor says “It must be something with the electrical installation.”

The SquareD AFCI breakers have a test button that will report the cause of a tripped breaker by pushing in the test button and switching the breaker back to ON. If the breaker switches off after 3 seconds, it’s an arc fault. If it switches off immediately, it’s a ground fault.

Every time I’ve gone through this test process, the breaker responds with an immediate ground fault response. Every time…ground fault.

Last Friday we had two tech’s from American Standard come out, accompanied by the electrical contractor, the HVAC contractor, and the builder. They spent 3 ½ hours testing, looking, and testing some more. At the end, they were all standing outside in a circle, scratching their heads and saying “I’ve never seen this before.”

Meanwhile, my wife and I are at our wits end. My understanding of electrical matters can fit in a thimble, but ground fault, to me, must mean something in the electrical system.
I am told that our entire home’s concrete, rebar reinforced foundation IS the earth ground for the home’s electrical system. The footprint the foundation is 70’ X 40’.

Is there such a thing as “too much” earth ground? Could it be that the complexity of all that rebar is creating some kind of ground loop that feeds back to the panel. Could it be that the ultimate path to earth ground is defective?

I don’t know, and the electrical contractor tells me that, as far as his tests can show, there is absolutely no problem with the electrical system.

I would really appreciate any knowledgeable insight on this because, all I have thus far, is watching a circular gunfight between the builder, electrical, HVAC, and American Standard.

So far, they are all shooting blanks.

Thank you, in advance, for any thoughts!

Best Answer

Your grounds are certainly not the problem. More grounding is better, and an Ufer ground tied into the foundation is the best there is.

19 AFCI breakers, eh? And people wonder why we say "Go BIG" when buying service panels. They're all on the left side, but that doesn't indicate anything special - for instance they're not all on the same "leg". If every other row of breakers was tripping and not the others, that would be interesting, but I'm sure you would have mentioned that.

A marvel of compact tech

A semi-modern "AFCI" breaker has 3 detection modes of interest to us today.

  • Overcurrent -- you are pulling significantly more than 20 amps (which itself has 2 modes, but that's irrelevant here).
  • Arc Fault - a small computer with a digital signal processor is listening to the electric line, for that characteristic sound of arcing - that "crinkle crunch" sound you get when plugging in headphones, or headphones with a bad connection, or hooking up speakers with the amplifier turned on. This is listening for series arc faults (inline with the load on hot or neutral), or a parallel (shorting) arc fault between hot and neutral.
  • Ground fault - this is a weakened form of "GFCI", closer to a "GFPE" or a European "RCD". Not good enough for human safety, but an effective way to detect parallel (shorting) arc faults hot-ground or neutral-ground, e.g. a hot wire sparking to a ground wire. Those are also ground faults.

Square D's "TIME SAVER Diagnostics", available on breakers with a white or purple TEST button, will store the cause of the last trip: immediate trip means a ground-fault (meaning the AFCI's weak GFCI-section detected a hot-ground or neutral-ground fault). A 2-second delayed trip means the computer detected the "sounds on the wire" of an arc fault. A 5-second delay means, well, the ground- and arc-fault sections don't know why it tripped, so it must be the dumb old overcurrent breaker.

In the panel, everyone can hear you scream

The nature of the overcurrent and ground-fault tripping means that they can only trip for problems on the "Load" side of the device.

However, the "computer, listening" has a side-effect: It's listening to the "Load" side of the breaker, but the circuit breaker has very low impedance across it, and no noise filtering. As such, the computer can "hear" arc faults on the "Line" side of the breaker, i.e. the panel's main bus. An AFCI trip can occur from other circuits having an arc fault.

It's probably one thing

And I think that's exactly what's happening to you. One of your circuits is arc-faulting, and the other circuits are "hearing it" sufficiently to trip them.

It sounds like you can guess which circuit that is.

So where do we look for arc faults on a particular circuit? Start at terminations of wires - breaker lugs, terminal screws, wire nuts - any end of any wire can be the source of the problem. I would say replacing all the heat pump components is an expensive way to re-torque connections... but suffice it to say, any connections disturbed by this overhaul can be excluded since they've been re-torqued.

So I would focus on the connections which were not disturbed by that recent work, e.g. those on the breaker or neutral bar, or any intermediate splices.

It could be in the machine, but if the machine has been replaced, that would be surprising.

The last and remotest possibility is a broken wire - again typically that is near a termination, but it's not common at all. It could also be a nail driven into a cable.