One of my clients has a rental house with a primitive water heater set up. It's in a slightly enlarged part of the crawl space - it has a door off the back yard and you step down into an area about 3x3 in front of the heater, which is on a slightly lower level, about 6 inches above a crushed rock floor, with a 2' diameter sump, maybe 2' deep below it. It's on a bit of a hill - seems like water flows to the spot easily, it rises pretty reliably during heavy rain. I first worked on this when I installed a new float switch. The previous switch would work after jostling it which is a sign of leakage and corrosion in my experience (turns out it was wet inside).

I spliced the cord with some 3/4 " dia. clear tubing filled with sikaflex over the joint, hoping to have ultimate waterproofness. Perhaps not necessary as the splice was positioned well above the likely high water mark but who knows. Seemed to work pretty well - then I get a call from the lady yesterday telling me it's not working and that the water rose enough to put out the pilot light again, this 3 months after the original repair.

Turns out it runs well enough, but pops the GFCI whenever the float goes down and shuts off the pump. I discovered this after re-setting the GFCI and watching it all pump out (thank God, I didn't have an aux. pump with me). Makes no sense, except that perhaps the sudden change in current is misinterpreted. I put in a new GFCI as sometimes they get over sensitive when old, but no change.

I then hardwired the pump into the box, bypassing the GFCI. This is not like a hairdryer in a bathroom, people don't hang onto the pump. Works reliably now, I reached down and quickly touched the wet stones, checking for any shock. Couldn't find any. Not sure how else to check for the sort of low current that would be passing through water. Stepping in it for a test would not be wise. I'm wondering about the definition and function of grounding. The water in that sump is in contact with water over a wide area. Would the charge be dissipated so widely that you wouldn't get shocked? I've read that you can get shocked in a flooded basement and often that to will be connected to ground water but maybe not as thoroughly. Jeez, the stuff I don't know.

I found some backup on the web for bypassing a GFCI for a sump pump but I don't feel I've got a complete grasp of this. I know I don't want to replace the pump - it's in a very awkward spot to work in - replacing the switch was bad enough.

__________________
1986 300SDL, 362K
1984 300D, 138K

Used to be you could just use a single gang outlet and bypass the GFCI requirement, then the NEC revised code in 2008 and closed that loophole.

You could try a GFCI breaker, as they are less prone to nuisance tripping.
240v sump is another loophole, go figure...

Not entirely sure, but hardwiring the pump may also provide a sidestep around the issue, as there would be no 'receptacle'

I would put in a single gang outlet and claim it was done prior to 2008, unless the homeowner wants to pony up to bring it into compliance with current code

Inductive loads tend to cause nuisance tripping of GFCI's, but it wouldn't hurt to measure current flow on the ground wire when the pump is running, the GFCI might just be telling you that there really is a problem.

__________________
On some nights I still believe that a car with the fuel gauge on empty can run about fifty more miles if you have the right music very loud on the radio. - HST

1983 300SD - 305000
1984 Toyota Landcruiser - 190000
1994 GMC Jimmy - 203000

https://media.giphy.com/media/X3nnss8PAj5aU/giphy.gif

Cmac- If the pump works fine without gfci protection, but trips the gfci even after replacing the receptacle you have a ground fault in the pump, or the cord going to the pump. It may not be enough to trip the circuit breaker, but can still pose a hazard to human life. The first thing I would rule out is where you spliced the cord,(not something I would have done for a customer!) even if it was never underwater moisture inside can cause a gfi to trip. If thats not the problem it's the pump. A sump pump has GFCI protection for a very good reason- the whole mixing electricity with water thing- so don't bypass the protection and try to claim compliance with a long ago superceded code. Your customer and your conscience will appreciate it. Do what has to be done to make it safe and code compliant, and if they don't want to "pony up" let them find someone else willing to risk their reputation so they can save money. good luck!

__________________
1984 300TD 1981 240D

The splice is a bit dodgy but are you familiar with Sikaflex? I put about A 4" length over one end before I soldered them, then I taped them generously and slid the tubing over it, and shoved Sikaflex in one end til it squeezed out the other. I'd be real surprised if moisture got through even if submerged 24/7 but just in the air, that may as well be a factory seal.

I will do a current test on the ground wire though, that could be instructive.

__________________
1986 300SDL, 362K
1984 300D, 138K

Not sure if its related, but my 52" rear projection tv kept tripping the gfci in my apartment until I bypassed it. (why the hell is there one in the living room in the first place??) Couldn't for the life of me figure out why. It still did it even after I cut the ground wire off the the TV, which shouldn't even be possible...

__________________
Ich liebe meine Autos!

1991 Mercedes-Benz 560SEL | Megasquirt MS3-Pro | 722.6 transmission w/ AMG paddles | Feind Motorsports Sway Bar | Stinger VIP Radar | AntiLaser Priority | PLX Wideband O2 | 150A Alternator | Cat Delete
1981 Mercedes-Benz 300SD | Blown engine, rebuilding someday...
1981 Mercedes-Benz 300SD | Rear ended, retired in garage.
2009 Yamaha AR230HO | Das Boot

Excessive speeding? It ain't excessive till I redline!

I don't want to be some hillbilly who bypasses safety equipment but I'm telling you, GFCIs have been a pain in my ass more than once. I encountered a similar problem - GFCI in a bedroom - removed it and everything worked fine, no shocks coming from anything.

I'm gathering on the web that some older style motors have a tendency to trip GFCIs. This pump moves water really well, I don't know how long the electric seals last or how old the pump is.

The best route would be to replace it but damn, it's in an awkward spot.

I'm curious about grounding though. I know the green or bare grounding wire goes on to the neutral bar in the box but isn't that attached to a long rod driven into the ground? It seems like groundwater would dissipate the charge deeper than one of those rods. During the rainy season, those rods are in steady contact with ground water. So if there was a ground fault in the pump while it's submerged in groundwater, it would almost be the same as if you touched hot to ground, which would trip the breaker. But I don't really think it works that way, I'm not sure why but I think it's because water is not an ideal conductor.

__________________
1986 300SDL, 362K
1984 300D, 138K

Since the ground wire is not sensed, it is certainly possible. All you need is a path to ground. Any path to ground.

No chance that you have other wires attached to the set, is there? Say, an antenna lead?

No need to have gfci protection in a bedroom, however the code requires afci ( Arc fault) protection in bedrooms as of 2008, and in all 15 and 20 amp branch circuits (except those that are gfci protected) as of 2011. The main reason for tripping of gfci or afci protected circuits in my experience is a ground touching a neutral on a receptacle somewhere in the circuit. A neutral to ground connection on the LOAD side of a gfci or afci protected circuit will cause tripping just like a hot to ground connection.

The ground rod at your main service is bonded to the neutral that comes from your utility transformer, but that is the only place they should be connected, otherwise you can be returning current on your ground and that can be bad under certain conditions. Many a wannabe electrician will tell you that ground and neutral are exactly the same, but this is absolutely untrue. I don't have the time or the typing skills to explain all the codes on grounding, but have a look at a forum for professional electricians called mikeholt.com and be ready for some heavy reading.

As far as your pump problem you really only have one choice to repair it properly- you have to replace the pump! While I agree that it is a pain to replace an otherwise functioning pump, if it won't work with gfci protection it is no longer servicable. What seems to happen as the motor wears you can get a small amount of leakage current on the windings, especially at startup, and that can be enough to trip the gfci even though it works fine otherwise and nobody is getting shocked.

__________________
1984 300TD 1981 240D

Question. Have you tried ANOTHER unit? If you try 2 different brands of GFCI, the problem has to be within the pump or wiring. I had a bad GFCI the other day. Bought another one and it kept tripping too. Got yet another one and it has been fine. Guess it was new and bad out of the box.

Could there be something bad in the wiring? Any way to buy another pump and test the circuit?

I would really hate to bypass the GFCI simply because the pump handles water and it is critical that it works at the right time

__________________
01 Ford Excursion Powerstroke
99 E300 Turbodiesel
91 Vette with 383 motor
05 Polaris Sportsman 800 EFI
06 Polaris Sportsman 500 EFI
03 SeaDoo GTX SC Red
03 SeaDoo GTX SC Yellow
04 Tailgator 21 ft Toy Hauler
11 Harley Davidson 883 SuperLow

In the breaker box, makes sure the neutral wire is connected to the neutral buss and the ground wire is connected to the ground buss; and verify the ground buss is connected to a proper ground stake.

I'll second that. Removing that ground just maybe made things more dangerous. Need to find that ground fault! Google how a GFI works and understand it before trying to troubleshoot.

The ground buss and neutral buss are only separated in a subpanel, in which case there should not be a ground rod as the grounding comes from the main panel on a separate wire from the neutral. Installing a ground rod at a subpanel instead of running a fourth wire from the main not only doesn't give you a proper ground as it is not referenced to the grounded conductor from the utility transformer, it is dangerous and can cause a lot of damage from surges etc.

__________________
1984 300TD 1981 240D

Who said anything about a subpanel?

What brand of gfci are you buying? If it is Leviton from the home depot that might explain it. Otherwise I have replaced very few that were tripping randomly without a ground fault present. What did you have plugged in to the recep? I would be concerned that the one that didn't trip might be the bad one!

__________________
1984 300TD 1981 240D

If you have a separate buss for the ground and neutral it is a subpanel. In the main panel they are one and the same. No need to separate grounds from neutrals in a main panel as they are bonded together.

__________________
1984 300TD 1981 240D