Hi-
In this part of the woods a lot of builders have been adding custom backyard fountains to make the small yards more appealing.
I've been setting redwood 4x4 posts and piping up into bellboxes with GFI's and bubble covers to provide power to the 1/4hp cord and plug pumps the landscapers typically install in these fountains.
I've noticed about 1 in 5 of these pumps will start tripping the GFI within a few days to a few months of 24/7 operation.
I usually tell the owner or builder to confirm that the pump is the problem by using an extension cord to plug the pump into another GFI, like the kitchen or bathroom.
When it trips that one I'll send them to another GFI for final confirmation that the fault is in the pump and they need to call the landscaper for warranty.
Now I'm assuming the pump is developing a ground fault or high leakage current and is defective. Sometimes these faults take an hour to trip and sometimes it trips as soon as you plug it in.
I'm operating under the belief that ALL outdoor receptacles MUST be GFCI protected
AND
that it violates the UL listing for me to cut the cord cap off of the SJO cord and, using the correct connector, splice it directly to power in the bell box and apply a WP blank.
I bring this up because I've just received another call on yet another pump and while I haven't had any arguments with any of the landscapers yet I want to know that I have no choice in how I power these pumps up should that argument arise.
Any sharing of experience in this area would be appreciated.
Thanks
Vince

Cutting the cap off of a device and 'hard wiring' it is a NO-NO.

The GFI is tripping, and doing it's purpose more than likely.

Have you found any 'bad' GFI's doing the ext cord thing, or is it always the pump??

John

That's what I figured.
The GFI's always test good and will supply other test loads just fine and the pumps in question will trip other good GFI's as well.
I am wondering if the landscape suppliers are selling a lot of cheap failure-prone pumps in this area.

I have no problem with 'hard wiring,' assuming that there is a disconnecting means within sight, AND it it STILL on a GFI.

I've seen a number of home fountains, and the only one that is even close to being code compliant is the one at my house - and I have NO nuisance GFI tripping.

The most common error in fountain installs involve the power cord. The corn is buried in cement, then travels underground ... only to pop up at the last instant near the receptacle. While this might hide that ugly cord, it is a BAD idea.

The cord material is not very resistant to moist dirt or cement. The rubber jacker deteriorates very quickly ... even if the same material lasts for years on your outdoor extension cords.

Likewise, "gardeners" and "fountain guys" seem to have no qualms about splicing sundry cords together - often creating a wiring error - and using duct tape to hold the twisted wires together.

The GFI just might be doing its' job.

Here are a few tests you might perform, to diagnose the problem:

My favorite is to use a megger on the wires; this will find any bad insulation.

I have a test cord whose outer jacket has been removed, to allow me access to the individual conductors. On a non-GFI circuit, I can also check the circuit with my ammeter. Not every ammeter is sensitive, or accurate, enough to measure such a small imbalance. Simply put, you ought to have the same amps going 'in' on the black wire as come 'out' on the white.

And, let's not forget the obvious: the receptacle itself. Plug something else in, and see if it trips. Your problem may be something as simple as an unused screw not tightened down.

I figure the GFI IS just doing it's job.
Could be the cord I suppose.
They usually just hide them under the decorative bark around here.
When I'm actually informed that a fountain is happening in time to give my input, I try my darndest to get the landscapers to bury a 1 1/4" pipe w/ 90's to accomodate the cord but that's only happened once.
The builder's usually throw these things into the project as an afterthought and the guys doing the work almost never seem to understand our national languge, making communication spotty at best.
I wouldn't really want to hard wire it because I rely on the fact that I only provided a legal outdoor receptacle for the owner to plug "their appliance" into and my liability stops there.
Reno- could you give a detailed description of your almost code compliant installation?
Thanks
Vince

My "almost" compliant installation is an indoor pond with fish, plants, and waterfall.

The tub / basin is plastic, so there was never the issue of bonding the rebar. In any event, it seems the 'pool' requirements kick in at a depth of 18", which mine does not exceed.

There are lights, filters, pumps, a heater and an aerator. The one place where I fail to comply (oh my!) is that I have two power strips daisy-chained for this; this is because I several receptacles are blocked off by the two timers in use. Likewise, the recessed lights have site-made trim, as no appropriate factory trim exists.

This pond is very much a work in progress, and I expect that future work will eliminate these issues.

The pumps, etc., are (for the most part) housed in a compartment made for that purpose. Cords and hoses are run under a 'lip' in the supporting frame, only coming into view as they enter the pond. They enter by going over the lip of the pond: I didn't want to create any leakage issues by piercing the basin.

What NEC section allows this suggestion?

I agree with Hotline, hardwiring this pump is not an NEC compliant option.

Changing the question slightly, has anyone had problems with a GFI holding on the little barrel septic pumps you see on raised bed septic systems? The GFI never trips in my presence, but somewhere in 24 hours it does, this despite running tons of water through it to cycle it while there. Replacing the GFI made no difference.

422.33 and 430.109 both say a cord and plug MAY serve as the disconnecting means. The code does not say it HAS to.

When it comes to the listing, as a general matter, the UL listing for the cord is completely separate from that of the appliance. Indeed, one of UL's constant headaches is unscrupulous manufacturers putting a UL listed cord on a non-listed appliance ... and letting the customer think the whole thing is UL listed.

More to the point, there are a variety of UL listed cord grips, evaluated for use on exactly this sort of cord.

This is a free society; something is assumed to be permitted UNLESS it is prohibited. Not the other way around.

Granted, the NEC may say somewhere in it's hundreds of pages that you can't hard-wire something that came with a plug .... but I'd like to see that citation.

Reno, I wonder if your installation will pass the 682.33 equipotential plane test

This same question of cutting the plug off came up at an IAEI Conference last year.

The issue was cutting the plug off a pump instead of installing the hard-wired version (which cost a bit more).

The Code Panel rep from UL went ballistic over that philosophy; when he calmed down he explained that even though the units looked absolutely identical from the outside, there were construction differences on the interior of the pump/motor. In the case of that particular pump/motor assembly, the materials of some of the internal seals keeping the motor dry were even different.

In addition, the tests used to UL List the plug-in and hard-wired units were different.

The real issue was that cutting off the plug and hardwiring was in violation of the UL Listing and Manufacturer's instructions.

That still finds its way back to being a violation of 110.3B, since it is being installed at variance with its Listing requirements and limitations.

I see the same problem with irrigation controllers in golf courses and HOAs. The landscaper always blames the GFCI outlet or breaker, but it always ends up being leakage or other faults in the irrigation controller (often a tansformer that's heading south).

I think you are handling the problem well by demonstrating to the customer that more than one GFCI trips with the same device. It narrows the problem down to the device, and involves the customer in the troubleshooting process.

Fountain pumps are becoming an off-the-shelf product that are commonly available to your average Home Depot/Lowes shopper, so competitive pricing leads fountain installers and manufacturers to supply cheaper units. We're bound to see abundant failures.

Ghost, it sounds as if the 'cut the plug' question would be a very good question for "UL Corner" in the IAEI magazine. I'd love to have an "official" statement on the subject.

Greg, the issue of "equipotential plane" does not come into the picture, because the "Pool And Spa" code (yes, there is one, separate from the NEC) does not apply itself to bodies of water less than 18" deep.

Even so, I'm a bit at a loss as to what such a 'plane' could possible accomplish. Construction is entirely non-conductive; absolutely nothing is 'grounded' (as the place has but two-prong receptacles), but all is on a GFCI. Even for the pumps, no metal is in contact with water - the impellers are magnetically driven.


Now ... for the 'sales pitch' .... ... drop by the chat room! I'll be happy to share pics, and discuss my specific water garden. I'm there most every night, roughly 6PM-9PM, Pacific time.


BTW ... I wonder if electric eels trip GFCI's?

Great question for Mythbusters!

For Alan B: Mythbusters is an American TV show, put on by some movie special effects people, where they try to prove or disprove myths, urban legends, etc. Try www.discovery.com.

Quick! I need an electric eel and a very long extension cord!

Ian A.

And if the electric eel doesn't trip the GFCI, be sure to repeat the experiment with a backwards-connected utility transformer (oh, with about a 12kV primary shoud do the trick) and have your test dummy "Buster" wired with an explosive charge that will go off if he gets more than 6ma of current.

Hey, there's a pole transformer and 7KV primary in my backyard...

I need a (working) video camera for when I play Mythbusters.

Ian A.