Need to GFCI flourescent

I have a situation in a new facility where we need to GFCI protect some flourescent fixtures due to my concern over safety issues. The fixtures are located above therapy pools with lifts, that could feasiby be touched by someone standing on the lift in it's fully extended position.

The problem is that my electrician can't figure out how to stop the GFCI breaker (or outlet for that matter) from tripping. He knows flourescent fixtures are challenging with GFCI. Is there any way to correct this problem or are we forced to go the route of incandescent fixtures?

Am I correct in assuming that the risk of shock with a flourescent fixture is the same as any other fixture?

I'm hoping to avoid replacing the fixtures but to date we've had no luck in stopping the tripping.

 

What type of flourescent fixtures are they?Are they totally enclosed?What i s the height of fixtures?IS the pool indoors?

 

They are totally enclosed fixtures, with 4 bulbs (the thin ones). They are 9'6" from the floor, but only approx 7'6" from the water height. If a person stands on the pool lift fully extended, they can touch the light, which is my concern, since the lift may be wet. Yes the pool is indoors.

 

Is there anything besides the lights on this circuit ? Is the neutral for the lights shared with any other circuits?Does the GFCI trip before the light switch is turned on or after ?

 

It could be a local code, or your concern for liability. All thou NEC does not require a totally enclosed lighting fixture installed over a Therapy pool to be GFCI protected. But I can understand the concern to have them protected.
The supply to the pool should as well be GFCI protected.
Giving some consideration to safety. It would be an advantage to be sure there is no metal part exposed from the fixtures that can be touched when standing on the platform. If this is the case then the GFCI protection would really only protect the person re lamping the fixture. And then the power should be off.
I have protected both electronic (T8 bulbs) and older style ballast (T12 bulbs) from a GFCI outlet, and never had a problem with doing such.
What Texsparky is insinuating is correct, if the electrician has wired a multi circuit using the neutral for more than one circuit (A/B), or (A/B/C) phases then this will cause the gfci too trip as soon as power is turned on to any of the other (one, or two) phases. If the gfci is tripping as soon as the light switch is turned on then the problem is in the wiring, or the lights them self.
Let us know how things turned out, Someone else may benefit from it.

 

Aphares.,
680.22(B)(2) Indoor Clearances reads as follows; For installations in indoor pool areas,the clearances shall be the same as for outdoor areas unless modified by as provided in this paragraph. IF the branch circuit supplying the equipment IS PROTECTED BY A GROUND FAULT INTERRUPTER, the following equipment shall be permitted at a height not less than 2.3 m (7 ft 6 in)above the maximum pool water level;
(1) Totally enclosed luminaires (fixtures)
(2) Ceiling-suspended ( paddle ) fans identified for use beneath ceiling structures such as provided on porches or patios

The fixtures need to be GFCI protected,7 ft 6 in or higher,and totally enclosed to meet code.

 

Yes Iam Aware of that Tex but look at 680.61 and let me know what you think.


680.61 Permanently Installed Therapeutic Pools.

Therapeutic pools that are constructed in the ground, on the ground, or in a building in such a manner that the pool cannot be readily disassembled shall comply with Parts I and II of this article.

Exception: The limitations of 680.22(B)(1) through (B)(4) shall not apply where all luminaires (lighting fixtures) are of the totally enclosed type.

 

aphares,
Sorry about that.Thanks for pointing this out to me.
Just a thought.........What do you feel is the reasoning behind the exception?
With this exception applied the way it is written, there is neither a gfci requirement or a minimum heigth requirement..I suppose the fixtures could be mounted 1/8 inch above water and meet code as long as they were fully enclosed.


The fixtures sound like 2 x 4 lay in.and probably do have exposed metal(lens frame)

 

Cannot be Readily disassembled, is the key here. In most of the articles about pools there is no restriction to the type of fixture, just restriction to height, and protection. Except Therapeutic pools, then it can only be a totally enclosed type, 680.62(D)(F).

Unless the pool can be dismantled, Then 680.21 (B) 1-4 apply. And then there is the posibality the pool can be set up in any general area where the lighting might not be totally enclosed.

 

I could be wrong here, but If I think of totally enclosed, I think of a fixture that has a plastic housing if exposed, Were there would not be any non-current carring metal parts exposed. And with out looking it up I bet chapter three would apply, as far as requiring these fixtures to be rated for moisture, and If a manufacture was to sell you something then most likely you would end up with a fixture that has a plastic housing. In the description from the poster, I sort of take it, they are fluorescent lighting like you would find in a tunnel. Mounted flush to the ceiling. Your thinking reseced, Even at that there would be exposed metal. Maybe Tom can fill us in on the type installed.

 

Thanks for all the comments. I'm not an electrician, so I've read them and tried to understand what they all say. I'm still confused as to what is meant by totally enclosed. Here's some info:

1. The fixtures are as you thought 2 x 4 lay-in.
2. I'm not there now, but believe the frames are metal.
3. Don't know if anything else is connected to neutral or on the circuit. I need to ask the electrician.
4. FYI, the pools and the lift device are all GFCI protected as you mentioned.
5. There were 2 different GFCI breakers. One flipped immediately when it was turned on. One flipped anywhere from 1 minute to 1 hour after it was turned on.
6. The pool can not be readily disassembled. It is a permanent structure.
7. Moisture isn't an issue as the room is controlled at 55% humidity.

I've seen a number of postings that suggest that there may be problems with GFCI protecting flourescents due to the nature of flourescents having some minimal leakage. What are your thoughts / experience with this?

 

Tom,
The jury is still out on what is meant by totally enclosed.I'm trying to find out more info.on this as well as flourescents tripping gfci.I do know that the neutral can not share with any other circuit or the gfci will trip.Your electrician needs to verify this.Post back and let us know what he finds.

 

I did some research on the term (Totally Enclosed). What I have come up with is the term is used several times in the NEC with out any real definition to it's meaning.
When it is used it pertains to an enclosure that either keeps something from leaking out, or something from leaking in. Pertaining to fixtures: it is used several times, and when it is used, usually it means: The fixture is designed to keep hot metal, and or sparks from escaping.
The Term Totally enclosed I believe would be a listing and labeling from the manufacture. A lot like a totally enclosed Motor is.
It would then be up to the AHJ to determine if these lights meet these criteria. Being mounted above the ceiling, and out of the area no hot metal, or sparks could escape into the area of the pool. Again up to the AHJ, and the kind of day he is having.
I will have to retract my last post (Above this one) of my interpretation of the term totally enclosed.
But the NEC is not and does not require the lighting to be GFCI if the metal can or cannot be touched. But the only type of lighting allowed is of the totally enclosed type. (Listed and labeled.)
Now GFCI and Fluorescent lighting, Lets forget the experience we all have, and go back to some basic thinking for a minute.
GFCI are installed in Kitchens, Bathrooms, Garages, Auto Repair Garages, Workshops, to just name a few. Now Fluorescent lights are installed where? Kitchens, Bathrooms, Garages, Auto Repair shops, So on so on. All around the world they are sold with plugs on them to be bought by us, and plugged into GFCI outlets without tripping the outlet, and or breaker. If you think about it you most likely have one above your work area in the garage plugged into your GFCI outlet. Sorry. I don’t believe the problem is the fluorescent lighting being on a GFCI.

 

OK...thanks. Now keep in mind that I'm not an electrician. But check out this post: http://groups.google.com/groups?q=gf...tel.com&rnum=4 . It mentions some concern over GFCIs and flourescents. It appears, anyway, that there is some debate on this matter.

 

Tom,
How's your electrician doing on tracing out the neutrals of these two circuits?Each gfci protected circuit will need its own neutral.I still feel as though the neutrals may be serving other circuits.Neither the neutral or the hot for each lighting circuit should be feeding anything else.

 

I understand your not an electrician, and I'll try to keep this basic, but if I don’t jump in and let me know.

The basic principle of a GFCI is to sense (current leakage to ground.) "Amps going where they should not be going" and then cause an interconnected overcurrent device to trip open the "FAULTED CIRCUIT".

Fluorescents use to be considered inductive loads.
Inductive loads cause a phenomenon where an EMF "Electrical magnetic force" is induced "jumps in air"in a circuit. Basically "volts" is an EMF. That jumps through the air into the wires. Before this phenomenon can a cure the Current "Amps" has to change.

What happens in an inductive load is volts are imposed “jumps in air”on the other wires "ground" when the current changes. The current "volts" that jumped on to the ground now have to go someplace, and they end up going to the ground. This is picked up by the GFCI as not suppose to be happening, and trips.
So far backing up what your thinking.

The older style ballast was very good for induction phenomenons, they would store electricity "current" then release it all at once, This was called a power factor correcting capacitor. (BOOM "EMF")

As you can see we have in this type of ballast what it would take for this phenomenon to happen as described.

Now in you’re above post you said the electrician installed T8 bulbs. T8 bulbs work from a new and radical ballast, called "Electronic" This type of ballast does not use the same design of the older ballast. The manufactures intergraded electronic circuit boards into the ballast. Basically eliminating the phenomenon know as "Induction"

And true the older style ballast was known to cause this phenomenon, this is what caused Electronic Ballast to be introduced to the market.

Hope this helped.

 

Not to impose or offend your electrician, but if he has a E-mail, then e-mail this page link, and lets see if he can clear a few questions up for us.
Thanks.

 

Aphares asked me to chime in about GFCI's and fluorescents.

I'm not an electrician, but I wrote the electrical wiring FAQ... ;-)

While I have said that there are suspicions/rumors/suggestions that inductive devices can sometimes false trip a GFCI, this (a) seems way too consistent for that (especially on something as "minor" as a fluorescent fixture, when contrasted to a motor) (b) I like the idea of GFCI'ing anything near a tub, even if some interpretations of the code may say it's not necessary and (c) I've run all sorts of things off GFCI's, and never seen false trips more than once-in-a-blue-moon (like, a couple of times per _year_, not every time you turn it on).

In other words, GFCI's are almost always telling you something _is_ wrong.

A misrouted/shared neutral is one possibility. But I rather doubt it.

I _suspect_ a leak in the ballast, but it could be a number of things.

Move the wiring around in the fixture, and see if it affects the GFCI.

If not, have the electrician disconnect the wiring from the fixture, and haywire in an outlet. Run something "largish" on it (like a small heater) and see if the GFCI goes. If it does trip, something's wrong with the GFCI or the circuit. If it doesn't trip, it's the fixture. If the fixture, check the wiring for nicks, cuts and wire chafing (mostly on the ballast primary). Try a different ballast.

IIRC, they tried another circuit, and it trips too. That's suggestive of the fixture.

 

To clarify this a little more, "wirefaq" is the author of the web site Tom posted above, About "fluorescent lights and GFCI" I asked him to read this post, then if needed fill free to post his response about fluorescent lighting and GFCI.

Thank you "wirefaq" for doing such. Aphares.

 

Unfortunately, the electrician is out on other jobs now as our project is mainly finished (except this one nagging problem). I've left him a message this evening. Thank you all so much for your help. What I'm hearing is:

1. It almost certainly isn't just a problem with the "nature" of flourescents and GFCIs.
2. It may be a shared neutral or hot (checking into this).
3. It may be a faulty ballast (actually more than one since more than one circuit is affected).
4. It may be a wire nick, etc.

FYI, my electrician doesn't use email. So I'm having to take the ball and run with it.

I'll keep this updated for everyone's info. Thanks again for all the info. It has been very helpful.

 

Tom,

Right. If a GFI is repeatedly and consistently tripping, it is either defective (but you've had the problems with two GFIs, thence unlikely), OR, there's _really_ something wrong.

Thinking a little more on this, I'm going to suggest that either a fifth possibility (below), OR a faulty ballast are the most likely.

5. Ground-neutral short or reversal in one of the fixtures and/or in or downstream of the light switch.

Sometimes when people are lazy, extending circuits ends up getting confused about neutral versus grounds, and in some cases, I've even seen the bare wire used as the neutral, with the other two wires used for hot and switched hot (especially ceiling fan/lamp retrofits for example). An electrician is unlikely to do this, but you never know - everyone makes mistakes occasionally.

Since you're having to run with this yourself, the first thing you need to do is isolate the fixtures from the equation, to find out whether the problem is in the fixtures, or in the wiring.

Rather than I suggested before (about haywiring in an outlet - you don't need to attach a load to test this), do this:

find where the first fixture in the string is (one electrically closest to the switch). Kill the power and turn off the light switch. Disconnect the feed wires (all _three_) completely, right at the fixture, leaving them hanging in mid-air (wirenut the insulated wires and make sure the bare wire doesn't touch anything). Re-enable the power - does the GFI trip? If so, the problem is either the feed side of the switch, or in the circuit supplying it. If not, turn on the switch - does the GFI trip? If so, the problem is in the "output" side of the switch, or in the wiring to the fixture.

Turn off the power again...

If none of the above identifies the problem, repeat the above process, reconnecting the fixtures one at a time - once you get the GFI to trip, where the problem is should be obvious. Likely either a neutral-ground short in a fixture, or a defective ballast (internal leakage to casing of fixture and thence to ground).

[It's unlikely to be a hot-neutral or hot-ground short, because the lights probably wouldn't go on at all.]

One additional possibility to keep in mind is defective cable. An electrician and I once spent over an hour diagnosing a fault that turned out to be a short _inside_ a piece of brand new, freshly laid and undamaged romex. Neither he nor I had ever seen that before. A ohmmeter (with the power off! and the feed end disconnected) is one way to look for these. But this'll be the least likely possibility. If this turns out to be the case, replacing the romex is your only option.

 

I've checked with my electrician and he says there are not shared neutrals or hots. I've printed out the post and asked him to review. Some very good ideas and info from all that I'm hoping will help him isolate the problem. Will continue to post info as I find out. Thanks again!