GFCI problems on submersable pump
Hey folks. I’m scratching my head over this one…
I have a small pond (1,000 sqr’) that I fill via a submersable pump from a nearby stream. Replaced the weatherproof box next to the stream (that wasn’t so weatherproof anymore) that the pump plugs into, with a new weatherproof box and standard outlet.
This is in turn is fed from a weatherproof box 50′ away that has a toggle switch and GFCI outlet so I can flip the pump on and off easily. Toggle switch and streamside outlet was downline from GFCI, so as to be protected.
GFCI tested out fine (Cheapo pug-in type tyester, with GFCI tripping button). Plug the pump in, and it trips the GFCI. Dragged the pump out of the water (unplugged obviously) and checked it over. No evidence of deterioration on insulation, or anything noticable on exterior.
Have since wired in a regular outlet in place of the GFCI, and all is working fine (as it has for years).
I’m wondering if anything about a submursible pump would cause problems with a GFCI (being underwater and all). I’ve heard of GFCI’s neusence tripping on circuits with flourecent lights, or other types of interference. I’d like to have the GFCI on the circuit for obvious reasons. I’d also like to know if I should throw the pump out ASAP.
Any thoughts appreciated!
-Erik
PS: I have successfullly installed dozens of GFCI’s (interior) without problem, and I’m pretty confident I had this one in correctly.
Replies
First things fitst, I am NOT an electrician, but spend a lot of time watching one. According to him, GFIs can sort of "wear out" over time causing an increase in nuisance tripping. Issue two is that GFIs don't play well with high initial amperage loads. While your pump probably does not draw that much power consistently, the start up draw can sometimes trip a GFI, especially an older one that may be getting weak. (This is the explanation given to me when I asked why everything in the basement of my house is GFI protected EXCEPT for the sump pump).
I would try a new GFI and see if all goes back to normal. But like I said, I'm just a guy . . . . . . . Any electricians might chime in with some more/better info.
Frontiercc, Thanks for the response.
The GFCI was a brand new, out of the box, 20amp GFCI that I installed the same time I did the other oultet.
Your point about startup voltage is interesting though. In fact, the pump ain't no small little guppy-pond pump, it's more the sump pump variety (discharges a real strong flow out of 2" CPVC pipe, up 20' of head).
Not sure what the amperage or GMP is off hand, but I could find out if people think the startup amerage could be the culprit.
another point might be that if you unplug the pump everytime you work on it, why does it need GF protection anyway? My understanding is that the GF protection is primarily to protect the USER from becoming a better ground than the EGC (or lack thereof) or neutral in case of a short in the chassis of a handheld power tool that is not double insulated...especially where the user would be located in or near water.
Brings to mind the time I was shutting off the spigot to a watering trough with a pliers, leaning over an electric fence. Elbow grazed the barb wire. Never did find those pliers. GF protection wouldn't work there either...every time the fence would arc the least little bit, the circuit would shut down. High voltage low amps.
Or a friends grandfather who was electrocuted (died) when he turned on a faulty circular saw while he was standing in water on a driveway. That was before they had GF protection. (and double insulated tools)
Chime in here, electricians...
"another point might be that if you unplug the pump everytime you work on it, why does it need GF protection anyway?"
Well, the pump is pretty much left plugged in all the time. The pond doesn't hold water real well, plus it pumps it up to a nice little rock "waterfall" that the wife like to look at. Hence the easy toggle switch on and off instead of crawling through the bushes to plug/unplug it. I suppose it should be hard wired in, but maybe next week....
The GFCI is also up near the pond, and convenient for hedge trimmers, etc... to be plugged into, which is also right next to the pond. If the trimmers fell in the pond, I'd fry all those nice little frogs that were keeping me awake last night. Hmmm that gives me an idea.... ;~)
I suppose you could leave the GFCI in and wire the toggle from the non-protected side of the recepticle..then you would still be protected from the hedge trimmers.
"I suppose you could leave the GFCI in and wire the toggle from the non-protected side of the recepticle."
I'm with you on that...
I was thinking about doing the same thing. The outlet by the creek wouldn't be protected, but at least the outlet by the pond (where the 2 year old plays) would be.
I once had the same problem: new pump, new GFCI outlet, and the pump tripped it every time.
Al Mollitor, Sharon MA
Get the pump tested by some source or just buy another if you are suspicious. They are cheap enough.
I'd say the weakest link is the toggle switch. That 2 dollar POS is not waterproof. That could be your leak to ground. Waterproof toggles are dearly expensive. Real waterproof switches like those on military vehicles and other outdoor equipment start at 20 clams each in large quantities. Crack the old one open and see how waterproof it is.
GFI's measure power to and from the appliance. If there is a path the current can leak, it will show conductivity. Set your VOM on the finest setting you have and ring out the black conductor to the pump to the case of the pump. Do the same for the other components. Since this is outdoors, the wind can drive rain into the body of the components. Possibly the Gfci is compromised by the moisture inside the body of the GFCI. After all there are printed circuit boards inside it.
BTW is this GFCI of the 3 dollar specials on sale now? If so, it is last years model. The new ones, made this year, are lots more expensive because thay have more features, (I'm stretching to remember but I think conformal coating of the PCB is part of the change) The new ones have some environmental ratings the old style did not have.
"Get the pump tested by some source"
What kind of outfit to look for? Local electrician?
"I'd say the weakest link is the toggle switch. That 2 dollar POS is not waterproof."
Toggle switch was brand new out of the box as well...
"Set your VOM on the finest setting you have and ring out the black conductor to the pump to the case of the pump. Do the same for the other components. Since this is outdoors, the wind can drive rain into the body of the components. Possibly the Gfci is compromised by the moisture inside the body of the GFCI."
Don't have a VOM (But always wanted an excuse to get one). Wind driven rain isn't an issue fo the pump, since it's sitting at the bottom of the creek. Not an issue for the GFCI, Toggle, or regualr outlet, since they were all new out of the box. Plus all tested fine before the pump was plugged into the circuit.
"BTW is this GFCI of the 3 dollar specials on sale now?"
Not on sale, and actually I paid extra for the 20 amp model (instead of the 15amp rated one you'd typically put in a bathroom).
I'm leaning to the startup amperage being the culprit, but if the GFCI is rated for 20 amps...is it conceivable that the sump pump could really be exceeding that on startup?
"Don't have a VOM But always wanted an excuse to get one"
Heck, you can get a chinese VOM off the internet (e.g. Harbor freight) for < $5) which I often do as I forget and leave a bunch (one at a time of course) lying around someplace.
BTW, I have zero GFIs in own house (except have a bunch in parts bins in basement), plus have 'heard' even the space station disabled the GFI equivalents due to false/nuisnace trips. Worked in 70's/80's with a guy (Tom Sussman) that was on the NEC panel for GFI, heard from him that the big push was primarily due to mfgs wanting to sell new stuff. Technical literature is split (as one would expect) on the long term benefits of those decisions.
Signed: A not really so old boy who still tests 120 Vac circuits with fingers, and 9 V batteries with tongue; but, hope to never get across 3kV ever again though!!!!
Early 70s GFIs didn't have much of a track record. Some of the designs didn't work all that well. By the late 70s, I got these points from a local inspector who has tracked these figures for decades, statistical evidence started to mount. By the 80s all of the studies, many by insurance companies, of numbers of people shocked or electrocuted clearly showed that at a minimum several thousand people have had their lives saved by GFIs.
Bumping the radio into the tub with an unwanted wife was a popular plot theme in the 50s. Do it now and your loving wife will likely step out and beat you down with a loofa.
As with any technology there are limitations. Some things you just don't want to trip off. Freezers and fire pumps for example. Most people who have problems with them are using worn out equipment or do not know how to install them. The only power cords that have problems have nicks or tears in their insulation. These should be retired.
I was part of a safety program on a large job site. Every week I would inspect all the cords used. At that time they allowed me to shorten, cut out bad sections, and put on new cord caps. I also checked all the tools for proper grounding, faults and intact double insulation.
In my experience most carpenters stop using a cord only when it cannot be made to run a tool any more. I have seen copper exposed a half dozen places on a 100' cord and a female cord cap that would only run a saw if you twisted it a bit. The carpenter said "It's fine".
Never mind that it makes the hair stand up in the rain. And might kill you one day. LOL. I used to have to cut them into little pieces to keep them from wandering out of the trash. Carpenters were much happier after I talked the boss into letting me buy a case of 12/2 100' models. As soon as the new cords went out the nuisance tripping of the GFIs stopped. For the price of one serious accident you could buy crates full of cords and GFIs. Cheap insurance.
The most probable problem is that the pump, being well used and old, has a slight leak in one or more seals. This combined with normal aging of the insulation of the motor windings can allow current to leak to ground. When the leaking current exceeds something close to 15 milliamperes the GFI trips.
This can be tested, use great care, by disconnecting the ground connections to the circuit and spanning the gap with a meter. For current. This would confirm the functionality of the GFI and the leakage from the pump. If you are not familiar with the tools and techniques needed this is better carried out by an electrician.
An alternative is to replace the existing GFI and see what happens. You could then isolate portions of the circuit and confirm what piece/s are causing the known good GFI to trip.
I suspect that the pump is the problem. If it is a model worthy of messing with replacing the seals or rebuilding may be justified. A few times I have had luck opening up the pump, drying it out and resealing it with a rebuild kit or, if needed, silicon sealant, Form-a-gasket or similar.
Shortening the cord a few inches allows a less worn portion of the cord to make a better seal. Sometimes it works. Other times not. Often you can squeeze out another year or so out of the pump. Seldom is it worth the trouble on anything less than a 1HP industrial model.
The most common spot to leak is around where the cord goes into the pump body. Particularly if the cord is used to move the pump. Always use a rope rigged to keep some slack in the cord near the pump.
Having a GFI in this situation is a good idea. Kids and animals like water. A shock with wet hands and standing in water is very dangerous. Be careful. Remember that having an electrician work on it is still cheaper than an ambulance ride.
Another thought.
He said that he tested the GFCI with one of the $3.95 testers. Those use neon light to show that power is on the right conbination of pins and also has a small load (from hot to ground) to test the tripping feature.
A miswired GFCI can trip with any "load". But I supspect that the test light do draw enough current to trip a miswired GFCI.
One way of isolating this problem would be to put the pump in the bath tub and power it off the bath GFCI. If that one trips then it is the pump.
Bill,
Good idea, I can certainly try that.
If the pump trips it in the bathroom however, I'm still not sure if it's faulty wireing or if the pump's startup amps trip the GFCI. Do you buy into the theory that this could be the cause?
Thanks,
Erik
"I'm still not sure if it's faulty wireing or if the pump's startup amps trip the GFCI. Do you buy into the theory that this could be the cause?"
Not exactly.
A GFCI is NOT AN OVERLOAD DEVICE (unless of course it is part of a breaker). Many people think that high current will trip them. That is not true.
They measure unbalanced currents. And when the motor is starting it draws lot of current so the initial surge might cause some problems with unbalanced currents. But I don't think so and have not really tried to study it or have much experience with operating motors through them.
I have seen refigerators easly trip the GFCI, but I don't think that has anything to do with the fact that it has motor, but rather the design of the insulation around the heaters and all of the moisture from defrost cycles.
I don't think the tripping of the gfi is initiated with overcurrent. I run my Tablesaw in the garage off of a 15 A two pole GFI breaker (QO) never a problem.
You don't have a ground circuit on this pump setup separate from the ground from the GFI do you? The GFI is looking for all current to travel thru it. Any downstream errant ground path will cause it to trip.Jack of all trades and master of none - you got a problem with that?
"You don't have a ground circuit on this pump setup separate from the ground from the GFI do you?"
Nope, no bootleg grounds here... Just the regular ground wire from panel...thru GFCI...to downline outlet...to pump's grounded plug.
Thanks,
Erik
I use a wiggy for testing. It is handy and while it does not calibrate the leakage needed to trip the GFI it does test the basic function. I used to use a resister, calculated to provide a 10 ma at 120v, soldered into a set of leads as a go/no go test. I would have to dig to find it.
Some contractors use a tester that is recalibrated regularly, about $600, to test GFIs. They look really sophisticated and the calibration certificate aids in credibility but many of these still use a simple resistor to do what I'm doing with my home made tester, $10 for the leads and resistor, does. A few are a bit more sophisticated.
The use of the GFI in the bathroom, or kitchen, should work. I wouldn't have thought of that. Carrying a groady pump through the house and messing up a nice white tub usually doesn't go over with most HOs. You might try using the GFI that protects the exterior receptacle, assuming it's not the one your having problems with, and use a bucket.
I was not concerned about testing the GFCI in the normal operating mode. Other then measuring the time a simple resistor to draw 6(?) ma should be OK.
While I don't think that this is the case there are number of times wheere GFCI wire "appears ok", but as soon as a load is pluged in the GFCI trips. For example is a pair of GFCI recpts are used to try to protect downstream recpts on a multiwire circuit.
That is where the little $3.95 tester will show that it is wired correctly and will trip when the test button is pressed. But the current drawn by the neon indicators might not be enough to show that the GFCI is miswired so that it will trip on any load.
But pluging in a 100 watt light should do the trick. If it trips then the problem is in the GFCI wiring.
Stray,
I think I'd be suprised if the pump DIDN'T trip a GFCI receptacle.Whenever you have wet cords they're very likely to trip ground fault circuits.When we're doing a commercial building and the roof deck is on ,but the rubber roof isn't on yet,a rainy day becomes wetter inside than outside.All the extension cords laying on wet concrete constantly trip the GFCI receptacles.If we don't dick around tying them up and eliminating the weakest insulated cords,we don't get anything done,because we're going back and forth to the temporary board every 5 minutes.
I think the only exception for not protecting outdoor receptacles is for snow melting equipment in residential applications.There they want you to use single receptacles that no one can plug anything else into.We do elevator shaft sump pump receptacles that way.And there it CAN'T be GFCI protected so, go figure...
Suspect the GFI recept. They give up the ghost just because the can. those cheap testers are natorious for not wrking as advertised.