Here's a weird one

[Carultch]

But why would turning off the breaker from the PV stop the nuisance tripping of unrelated AFCI breakers, if we can trust that our customers are giving us reliable information?

AFCI's work by tripping upon detecting waveform features that are characteristic to an arc fault. If another circuit on the same grid generates harmonics, power-line-carrier signals or other waveform features, it can cause the AFCI to "think" there is an arc fault, when there isn't one.

 

[tortuga]

We have 4 customers all within a couple miles of each other who are claiming that their PV systems are causing AFCI load breakers unrelated to the PV system in their houses to trip. One of the customers has told us that his breakers are tripping at night when the PV system is (obviously) not producing. What the installations have in common besides geographic proximity is that they are on the same utility, the same electrician wired all these homes, all the panels and breakers are Square D, and in all cases the PV is interconnected in the MDP while the breakers that are tripping are in subpanels. And of course we built all the PV systems.

The systems are arrays of Sunpower modules with microinverters using their proprietary powerline communication for monitoring. We have several hundred systems in the field built on this same basic design, and these four customers are the only ones reporting this problem, if there is a problem.

The obvious first thing to do was to turn off the PV systems for a couple of weeks and see if the breakers stop tripping; all four customers did that and reported that they had no trips during that time but that the trips started up again when they turned their PV systems back on. Is something in the powerline communication that could indicate to the AFCI breakers that there is an arc fault, even though it's on another breaker in another panel? If so, why is it only happening in only these particular systems out of the hundreds of the same ilk we have built?

I've got nothing. You?

I would run not walk away from any X10 or power-line communication sends a pulse at the zero crossing of an AC system.
That would be a deal breaker for me.
Cat5 cable is pretty economical, you can even get shielded for a decent price and its 2020 permanently installed 'smart' products should use tcp/ip over wired ethernet, for security and reliability.

 

[ggunn]

AFCI's work by tripping upon detecting waveform features that are characteristic to an arc fault. If another circuit on the same grid generates harmonics, power-line-carrier signals or other waveform features, it can cause the AFCI to "think" there is an arc fault, when there isn't one.

I get that, but why only these four systems out of the hundreds we have in the field designed the same way, and why are they all within a mile of one another?

 

[kwired]

I get that, but why only these four systems out of the hundreds we have in the field designed the same way, and why are they all within a mile of one another?

If all in closer proximity to one another seems that might be good indication of something external but in that area to be the main cause. Radio/wireless communications maybe? Plant in the area?

Maybe even something going on in POCO distribution, bad but not yet failed connection or component?

 

[CodeQuandary]

Maybe the PV cabling was acting as an antenna and picking up the RF interference, and the breaker was effectively an antenna disconnect?
If so maybe a common-mode choke on the PV lines might help.
It's all speculation but that's all I can do from this end.

Whoa, you just reminded me: I once installed a solar system at a police station: they had a some other common-mode things (intercoms I think) and the only way I could get it to work was put a choke on the entire PV system. Worked perfectly. I don't know whether others have used this trick. These were Enphase microinverters, maybe 10 years ago, that couldn't find eachother/the Enlighten box until I added the choke. I don't know why that would come into play here, it would be interesting to know if the trips happened at the same time each day (i.e. 30m after sunset or something).

 

[CodeQuandary]

I would also agree/expand on an earlier point: buy a new-generation Square-D AFCI, and replace it in one house, and see if that solves the problem compared to the other houses. Do all the houses trip every night, btw, or just sometimes?

 

[tortuga]

I get that, but why only these four systems out of the hundreds we have in the field designed the same way, and why are they all within a mile of one another?

The first step for me is:

• Visit the site and gather as many facts as you can about the trips and who has done what wiring.
• Create a photo copy of the panel schedule and or create one.
• Visually inspect the service for obvious problems.
• Note which breakers were tripping.
• Provide the system user a notebook to log the time of a trip and what appliances on those circuits were on.
  

 

[synchro]

We have 4 customers all within a couple miles of each other who are claiming that their PV systems are causing AFCI load breakers unrelated to the PV system in their houses to trip. ... What the installations have in common besides geographic proximity is that they are on the same utility, the same electrician wired all these homes, all the panels and breakers are Square D, and in all cases the PV is interconnected in the MDP while the breakers that are tripping are in subpanels. And of course we built all the PV systems.

The systems are arrays of Sunpower modules with microinverters using their proprietary powerline communication for monitoring.
... Is something in the powerline communication that could indicate to the AFCI breakers that there is an arc fault, even though it's on another breaker in another panel?

Is the monitor device powered by the main panel that has the solar interconnect, or is it powered by a subpanel where the AFCIs are tripping? If the monitor is connected thru the main panel then you could put ferrite chokes on the feeder lines to the subpanel to block the powerline carrier signal from getting to the AFCIs there.

 

[tortuga]

Single point grounding Test:
Switch the service off, disconnect the service neutral, MBJ and or GEC from the neutral bar, test the neutral bar for continuity to the GEC with a 250V megger. See if there are parallel paths to the grounding electrode system from the neutral bar.

 

[jaggedben]

Single point grounding Test:
Switch the service off, disconnect the service neutral, MBJ and or GEC from the neutral bar, test the neutral bar for continuity to the GEC with a 250V megger. See if there are parallel paths to the grounding electrode system from the neutral bar.

I don't see what that would have to do with the problem.

...

I think the suggestion to put chokes on the feeders is reasonable.

 

[ggunn]

Is the monitor device powered by the main panel that has the solar interconnect, or is it powered by a subpanel where the AFCIs are tripping? If the monitor is connected thru the main panel then you could put ferrite chokes on the feeder lines to the subpanel to block the powerline carrier signal from getting to the AFCIs there.

Neither. It is powered from the solar load center where the branch circuits from the microinverters are combined for a single point of interconnection in the MDP, which is in essence another subpanel.

 

[tortuga]

I don't see what that would have to do with the problem.

Yeah that was just an idea.
The RFI winnie was mentioning:

...discussions of AFCI devices which are susceptible to RFI (tripped by the radios). Perhaps the same issues are in play here.

made me wonder ..
Could a current circuiting on the ECG could induce current on the phase conductors and trip an AFCI?
I see lots of improper N-G bonds.

 

[ee4fire]

Do these PV systems have one large inverter or the small inverters at each panel? I am assuming they are tied into the grounding system. Since solar panels only provide DC power which has to be inverted to AC power, are the inverters putting transient noise on the neutrals or grounding system? A regular 60 Hz Volt- amp meter wont detect the noise or transients. I realize this is at night, but any light hitting the solar panels may cause a low output causing noise.

You may need to have a testing company or the utility check for transient or harmonic issues causing the AFCI read a fault that doesn't exist.

 

[sthomsen]

Curious problem. How many breakers trip and are they the same ones all the time? Is it consistent every night or random?

 

[sketchy]

I just posted regarding a similar situation here. I don't think it matters about the utility, we're experiencing these issues at sites miles apart. I'd guess it's the wave form theory, it's the common denominator.

 

[tortuga]

@ggunn any update on this? Currious if you ever got to the bottom of it?

 

[tallgirl]

You might try turning off just the monitor's breaker. The microinverters should go to sleep and stop communicating at night. Assuming it's an Equinox system, it might be trying to chat at night, waiting for the sun to come up. I've never worked on Enphase equipment, but I know the people who make them, so more information might be helpful.

 

[tallgirl]

I would run not walk away from any X10 or power-line communication sends a pulse at the zero crossing of an AC system.
That would be a deal breaker for me.
Cat5 cable is pretty economical, you can even get shielded for a decent price and its 2020 permanently installed 'smart' products should use tcp/ip over wired ethernet, for security and reliability.

It's not like an X10 setup. The PLC signal rides on top of the AC wave form continuously.

 

[tortuga]

It's not like an X10 setup. The PLC signal rides on top of the AC wave form continuously.

Thats very interesting do you have more info on this?

 

[tallgirl]

Thats very interesting do you have more info on this?

It's not particularly new or exciting. Internet over power line has been a thing for quite some time. The power conversion equipment that does it on the AC lines just does it at a lower bit rate than standard network speeds.

Power-line communication - Wikipedia

en.wikipedia.org