Hi guys, haven't been on in a long time. Have a weird problem I'm on tonight. 1000 GE breaker serving a gym. At night when they turn off lights to the pool area. 4 circuits controlled by contactor. Controlled by one single pole switch. The 1000 main in the mdp trips. Controls and lighting circuits are in a 225 amp sub fed from mdp. It doesn't trip when turning lights on only when turning lights off

Any thoughts, I'm a little stumped

Defying all logic .....

My dad's GFCI (early 80's) for the pool filter kept tripping for some unknown reason. It turned out that it tripped whenever someone rang the doorbell. I guess the little arc in the doorbell contacts leaked just enough current to trip it.

If that was really the case - removing the doorbell from the pump circuit solved the issue back then - perhaps your problem is something as simple as worn switch contacts, ane replacing the contactor points will solve the problem.

Pool lights = humidity = water condensation = wet insulation

Quick check to isolate the problem, does it still happen when the four circuits feeding the lights are turned off but you still operate the lighting contactor?

If so, I would suspect the inductive kick generated by the lighting contactor is punching thru insulation and find its way to ground. Is there any sort of RC snubber across the contactor coil to reduce the magnitude of the inductive kick? If not, try adding one to see if that cures the problem.

If it does not happen when the lights are off, try the individual lighting circuits one at a time to isolate the trouble making circuit.

I had a bad 277 volt, F40 T-12 ballast self destruct and trip a 2000 amp main w/GF. Never tripped the branch CB on it's way.

Try what Larry suggested.

Thank you so much guys, nice to know there still is a good resource of knowledgable people.
Little follow up after last night , contactor is mechanically held, GE open face 25+ years old in a very humid environment with constant chlorine vapors present. Needless to say you can imagine its condition. I did try what Larry said last night, problem is def in control circuitry. Lights deenergized still tripped, going to replace contractors tonight and test a few more things that came to me while sleeping lol. I believe open relay coil is leaking on power to open contactor. Going to amp ground tonight and see what's on it. I need to make a correction as well it is NOT a GFCI main it is a GFP main. Let go thresholds are 100 times greater on these that's why I'm gonna amp grounds tonight. Again thanks guys stay safe and will update later and try to visit more.

And Reno , how's all the cats bud, prob a few new additions since we talked last Kids picked up new one recently named him Satan , it fits!!!!!!!!!!

Glad to help. Many other people here are much more qualified than I am.

Quick test. I assume this is a two coil contactor. One coil closes contactor, the second coil opens contactor.

____If it is safe to do so_____ isolate the neutral wire on the close coil. If the coil shorted to frame, it should still trip the main. If it still trips with the coil neutral wire disconnected, isolate the line side wire of the coil. If it still trips, then it sounds like the switch wiring is grounded.

good Luck.

Thanks Larry, that is actually one of the other things I was going to do tonight, everything is in a trough underneath contactors, there are more there than these but this is the troublemaker. It hit me last night I saw a big blue wirenut in there with a whole lot of neutrals in it. I'm curious also if some has put lv and hv neutrals together. And you are correct sir, mechanically held open and close coils, but old GE style reverse wound so it's on the same core

Any updates?

GFCI current sensors are apparently easily triggered by radio waves.

In an apartment I used to live in, the bathroom GFCI was apparently wired backwards. When wired backwards, it can still detect leakage current since that is a common mode AC current, so direction is not relevant. The problem with backward wiring is that opening the circuit does not disconnect power to the solenoid that opens the circuit. If for some reason the triggering current remains, a backwards wired device will keep operating the solenoid.

I found that 1 watt of RF in the amateur 2m band on a 1/4 wave or 5/8 wave antenna within 3 meters of the wiring leading to the backward GFCI would cause it to trigger, and it would keep buzzing loudly for as long as the RF was present.

RF can trigger these things. When the power wiring is acting as an antenna for nearby RF, it will be in common mode (both conductors conducting in the same direction at the same time), which will appear like leakage current. A simple rectification circuit to detect a voltage from the current transformer will operate over a large RF frequency range (e.g. from a few Hz to a GHz or more).

The arcing from the doorbell switch could be a long enough RF pulse to do this. Or the bell itself may be doing it.

RF filters that block common mode currents in the RF range (but not in power frequency range) could be used to stop such issues. Just be sure the filter used does not block 50-60 Hz common mode, or it could defeat the proper ground fault detection.

First, GFCI is a specific term from both UL and the NEC, which is used to designate a Ground Fault protective device that is used to protect human life (i.e. 5mA trip).

Your 1000A main has a simple ground fault trip (GF or GFI or GFP). On service entrance equipment these devices typically have a minimum trip point of 200A. At the minimum setting these devices can easily trip before a 20A branch breaker will operate.

Most likely your lighting contactor is a mechanically held device, which means it has a separate circuit for turning Off. I would investigate this circuit for a GF. As pointed out by others, this may only occur after enough moisture has built up.

John,

I too saw a ballast knock out a 1600 amp main GFPE breaker and take out the whole building. Now I make sure the EC's set the GFPE setting correctly.

Hello again,
Well I found a bad fixture that was causing it to trip. But it has once again started it's shenanigans. As I had stated in an earlier post it is not a GFCI breaker but a GFP. After doing some research and talking with some engineers these breakers do not have a set trip value but rather rides inside a bubble if you will that move up and down with the current passing through it, if the ground fault moves outside this bubble it trips. So my theory is that they are on the verge or riding the line of the bubble all the time so it is easy to trip. Also distance plays a huge role in this as well, GFP breakers max trip is with conductors leaving MDP about 60 ft away, the longer the conductors are out to sub panels the more this "bubble" shrinks. This pool panel has prob 250 ft of feeder conductor back to MDP. Just thought I would give a quick update, going back tonight to see if I can another bad fixture.

BTW, these fixtures really aren't bad, as in not working , but probably starting to break down or turning off collapses the magnetic field and induces enough current to trip the main

You stated in your second post that when you turned off the fixtures but operated the lighting contactor, the main breaker still tripped. Did you find and fix the contactor related issue?

Hey Larry,
Thank you, I did replace the contactors but still had a problem in one the lighting circuits , just left site and here is what I found:

Going back to MDP I took off all bottom covers and started looking for anything unusual on grounding side of the system. I found a ground wire coming from a bank of power factor correcting capacitors grounded on the neutral bus bar and not the ground bar, this ground was attached on the load side of the neutral sensor on the GFP. Shut power off and moved it to ground bar, current on neutral bus immediately dropped by 200 milliamps. I then cycled lighting 5 times, waiting in between to let them achieve full brightness and GFP never tripped. My theory is that one or more of the capacitors is starting to fail and was leaking current onto the ground wire, which was tied to the neutral bus, causing an imbalance and was being sensed by the ground fault circuitry and causing the breaker to trip. Does that sound plausible to you

Again thanks for your input, stay safe

Yes it does sound plausible.

So you think you had three sources of ground current leakage? Lighting contactor, failing fixture, and the miswired PFC capacitors ground lead.

Only a theory, but I think there is a cumulative effect going, lots of little things, each small but adding together they are pushing the envelope of the trip settings. So I changed my thinking to looking for many small things instead one large problem. Funny thing is I haven't fixed the one light I took down yet, just about tempted to go hang it back up just to see what happens now

Any idea what the trip point is, when everything is running?

I'll start with this is a GE High Pressure Contact Ground Fault Protected Switch, that's how GE refers to it.
As was explained to me from GE the trip setting is a variable number , really shouldnt go over 1000 milliamps but can be as low as 100 milliamps. So that makes it really clear this is in no way intended or designed to protect personnel, it is to protect equipment only. It is a product of voltage at a given time, current flowing at a given time, voltage drop and resistance. He said if I can give him these numbers at a given time then the trip can be calculated at that instance. I can only assume the window has to be calculated at a 5% voltage drop which is required not to exceed by the NEC. Of course resistance on the system and the voltage and current will play into that, and since the current will fluctuate I can see that number being a variable. He did say maximum trip is calculated with feeder conductors being no longer than 60 ft from source, little confused why the continuation of branch circuits wouldn't affect it. I guess to answer your question would be a no, lol
All in all I'm kind of impressed this is being done real time on a 30+ year old distribution breaker.