I am putting in a pool and I didn't really understand all I knew about pools. I know what the code says and I can pass the tests but inspecting commercial buildings for the last 10 years I really haven't looked at any pools.
My wife builds high end houses so I have the opportunity to see what passes so I looked ... and I am underwealmed.
Is it really the intent of the code that the J box is simply a thermocraft sticking out of the ground on 3-4 RMNC "legs" and *maybe* tywrapped to a 1x4 they found from when the framers bucked the windows? It sure looks like something that is going to "suffer the slings and arrows of the outragous yard men" who ply their trade in this hamlet. <sorry>
I can't believe the buyer of a $750,000 house is happy with that, even if it is "hold your nose" legal.
Don't they make a J box that attaches to a wall? I would certainly like all of this strapped to something more solid than a board stuck in the ground. I haven't seen one in my poking around on the web.
I really want something more substantial at my house.

I believe you are correct and that the jurisdiction needs to "tighten up" on that issue. It is a problem in many counties. I would not accept tie wraps, I would only accept a 4x4 pressure treated post (lumber, not landscape timber).

I will post a better answer soon. Today and tomorrow I am attending the Building Code Administrators and Inspectors Board meeting in Jacksonville.

(Don't have my Code books handy now)

Hopefully Mike will get this and run an answer today or tomorrow.

Just to make a quick reply, there is a bracket made and designed for use to support listed swimming pool junction boxes. They may be attached to the building structure, or post.

Article 680 doesn’t address the support of the listed swimming pool junction boxes or other enclosures, so one must go to article 314.23. That code section 314.23(e) permits enclosures without devices, fixtures, etc.. for boxes with threaded hubs, or hubs identified for the purpose, 100 in. in size, or less, to be supported by two or more conduits threaded wrenchtight into the hub.

Each conduit must be secured within 3 ft of the enclosure if the conduits enter the box from more than one direction.

If all the conduits enter the box from the same direction than the conduit must be supported within 18 in of the box.

The box is not permitted to be supported by only one conduit.

If 314.23(e) cannot be met than you must look to one of the other provisions of 314.23 for support.

The best way is to run the conduit back to the transformer at the pool pack, but most pool companies won't supply the longer corded pool lights. The cost of the longer cord surely couldn't cost more than a pool junction box could it?

Mike, I agree with your response. Also, I would like to add that if I felt it was a "hold your nose" legal installation (love that description) and that would fall apart or be subject to damage from the lawn care guy, I would probably red-tag it citing a violation of 110.3(2).

Do you think I would be correct?

Also, Electricman, I'm not sure but I think that some transformers may prohibit the direct connection that you are talking about. I have to check it out a little bit more.

Yes, you are correct on both items. Actually most transformer enclosures are not listed for direct connection, as they don't conform to the four construction requirements of 680.24(b)(1).

Mike, I believe that article 680.24 (b)(1) pertains to no niche fixtures only as per the most commonly used wet niche. We have been running pvc to transformer enclosures for years.

This section 680.24(a) & (b) applies to both wet-niche and no-niche fixtures. Follow along as I set the stage. We’ll look at part B as it is similar to part A:

680.24(B) Other Enclosures;
An enclosure for a transformer, ground-fault circuit interrupter, or a similar device connected to a conduit that extends directly to a forming shell or mounting bracket of a no-niche fixture shall meet the requirements of this section. You will notice that this section is addressing enclosures for transformers etc..
where the conduit extends to the forming shell of the fixture. That fixture that they are talking about is a wet-niche fixture . Or this conduit may also connect to the mounting bracket of a no-niche fixture .

Next look at 680.23(d);
(D) No-Niche Fixtures. A no-niche fixture shall meet the construction requirements of 680.23(B)(3) and be installed in accordance with the requirements of 680.23(B). Where connection to a forming shell is specified, the connection shall be to the mounting bracket. As you can see the requirements that apply to a no-niche fixture are the same as for wet-niche, as this section sends you to 680.23(b)

680.23(b) (2) Wiring Extending Directly to the Forming Shell.
Conduit shall be installed from the forming shell to a suitable junction box or other enclosure located as provided in 680.24.

As you can see, 680.23(d) takes you to 680.23(b), now 680.23(b)(2) now tells you to install the conduit to a junction box or other enclosure (transformer) as noted in 680.24.

To sum up, the conduit shall be installed from the forming shell of a wet-niche fixture, or to a mounting bracket of a no-niche fixture, to a suitable junction box or other enclosure as provided in 680.24. The transformer enclosure must be constructed according to 680.24(b) if it is to connect directly to a wet-niche fixture or to the mounting bracket of a no-niche fixture

I am becoming more educated about pools ;-)
The answer for why you don't see longer cords on 12v lights comes from the manufacturers. They say voltage drop is the problem and limit the cord at 50'. A 300w 12v bulb will be pulling 25a so that is a reasonable assumption.

Now another question; Can you have too much bonding? Specifically is it wrong to bond your pool all the way back to the service grounding electrode? You already have that path via the equipment grounding conductors in the light and pump but is there any reason why you wouldn't connect the ground electrode to the pool bonding grid with #8?

I say it's wrong, it's not the intent, even though codewise we may not be able to prevent it if they do it that way. See the previous commentary on the iaei.org website (but then come back to this forum because it's much better, less confusing, and easier to use):

Go to iaei.org

Click on Global Forum

Click on Discussion Forum

Go a few posts down to "Pool Bonding"

Thanks Nick I read those. It appears that the Canadian code is really similar to ours in that we also require grounding conductor "without joint or splice" from the junction (deck?) box back to the panelboard, which is usually also where the main bonding jumper and GEC resides. 680.23(F)(2)
Now we have two connections to the bonding grid via different paths (motor and light). If you toss in a cover motor and a heater we are really adding up the amps that could flow in your fault.

The reason this occurred to me in the first place was I was getting ready to run the conductor from my J box to the panel and I wondered what size wire I would use. Cost aside, why wouldn't I run the #8 insulated since I have a roll? That would become what we are talking about since the pool steel is bonded to that same ground bus in the J box.

This whole thing sure sounds like stretch pants with a belt and suspenders anyway since it is all GFCI protected but better safe than sorry I guess.

I saw another interesting thing in the swimming pool code. 424.1.64E.9.006.2(c)(3) in the draft seems to say you can only use 12v lights in a commercial pool but the residential section 424.2 does not have that restriction.
What lights can you use in a residential pool?

1. There are longer cords on wet niche fixtures, the manufacturers provide fixtures with cords up to 100 ft.

2. There is no requirement in article 680 that will prohibit one from running the bonding conductor back and connecting it to the grounding electrode system. Although there is no benefit in doing so there is no harm either. Lets face it, dig a big hole, put in a bunch of steel rebar connected together, cover it in concrete, and what do you have? A great big ufer ground! (concrete-encased electrode).

3. As a mater of fact, if a pool is installed within a specific distance to a lightning protection system, NFPA 780 requires the pool shell and equipment be bonded to the lighting protection system ground. The lightning protections system is now required to be bonded to the grounding electrode system at the service.

4. There is no requirement that the bonding conductor be installed continuos (without a splice or joint). That requirement is only for EGC (equipment grounding conductors).

5. The requirement for bonding is to eliminate voltage gradients, not to provide a path for fault current to return to the source to facilitate the operation of the overcurrent protective device. Thats what the EGC's do.

6. When all the metal parts and electrical equipment associated with the pool as prescribed in 680 are bonded together, that eliminates unequal voltage potentials, therefore there is no hazard.

OK Mike here I go again:

1. Agree. You can special order it.
2. Don't agree that it's harmless. And no one
swims inside a ufer ground. Concrete and
tile is considered a grounded surface,
you know this.
3. Rare.
4. Agree.
5. Agree in part. What you say is true,
however my point was that a small amount of
current may attempt to return to the source
via the pool bonding route. We can't be sure,
so why change the code if we can't be sure.
6. Agree.

We have a difference of opinion on this subject.

Nick you are correct, and we do agree, if there is a fault on any circuit in the electrical system current flows on all grounding and bonding conductors. Some have thought that current flows on the path of least resistance. That is 100% incorrect.

We bond all metal parts and equipment so there is no potential at all between parts, from a fault that may occur in any circuit, cracked lens in the wet-niche fixture, motor winding going to ground, whatever.

The primary purpose of bonding is to ensure that voltage gradients in the pool area are eliminated. The fine print note explains that the 8 AWG conductor's only function is to eliminate the voltage gradient in the pool area. It is not required to provide a path for fault current that may occur as a result of electrical equipment failure.

If there is no potential difference there is no hazard! 20 volts to ground at the motor, 20 volts at the handrail, 20 volts at the fixture...etc... equal potential, no hazard . That’s just what an equipotential plane does, eliminate voltage gradients just like 680.26(a) says.

Oh I forgot, you swim in a concrete-encased electrode all the time. I use that as a joke but the fact is the bonding conductors are connected to the motor EGC, the EGC to the panelboard feeder EGC, than to the service panel where the neutrals and grounds are connected to the grounding electrodes. So the bottom line is the bonding grid is already connected to the grounding electrode system, indirectly, but connected never the less.

If lightning hits and is dissipated its going to that pool....its a better ground than the ground rods.

And there are many pools that are connected to the lightning protection system by those contractors. Happens all the time, and NFPA 780 requires it.

Mike, you have to agree with me that in your example you are not taking into consideration the length of the run of the 12 equipment grounding conductor, or the resistance of that wire. This is the same example everyone uses to bypass the real issue. I've heard it a million times. I don't agree.

We are talking about this in the wrong post. This is supposed to be about junction boxes.

So let me try to understand your point better. What you are saying is that the resistance of the #12 EGC conductor, and its length is a problem? Lets think about it, if that were true, wouldn’t there be too much resistance to allow the current to return to the source and open the breaker in a fault condition? If what you say is correct than no circuit in a fault would ever work because of to much resistance!

Lets get back to the point. If the code panel thought it was unsafe to connect the pool bonding grid to a grounding electrode they wouldn’t say that its just not the intent to require it, but rather our code would say that it shall not be connected to any grounding electrode.

Again lets go back to quote the code:
680.26(a)
FPN:
Now here is a quote from the handbook:
By code if we do that there is no hazard. I agree that if it were a hazard as you believe, than the code should come out a specifically say not to connect it to any electrode. Maybe we do need a code change.

I'm not letting you off that easily....

My first comment - A number 12 wire as you know can handle a fault for a 20A circuit. Since the pool grid is now connected to the electrode system, in theory there can be a larger fault applied to the #8. You were saying how the #8 and the #12 are connected, though not intentionally. So that was what I was getting at, but lets get back to the main point:

Here is the NEC definition of Ground:
A conducting connection, whether intentional or accidental, between an electrical circuit or equipment and the earth or to some conducting body that serves in place of the earth.

Now one step further, here is the NEC definition of Effectively Grounded:
INTENTIONALLY connected to earth through a ground connection or connections of sufficiently low impedance and having sufficient current-caryring capacity to prevent the building of voltages that may result in undue hazards to connected equipment or to persons.

Now, you quoted that the purpose of pool bonding was to eliminate voltage gradients in the pool area. I agree. You stated previously that the purpose of pool bonding was NOT to provide a path for fault current. I agree. If someone connects the #8 bare back to the main service panel or main service electrode, he/she is using the POOL BONDING GRID AS A MADE ELECTRODE. It is serving a totally different purpose other than the intent of pool bonding as we know it. You are using the bonding grid to clear fault current irregardless of how much or how little fault current may flow on it! It is WRONG!

:p

Lets break it down and deal with one part at a time, and talk about your first paragraph. The short circuit withstand rating of a conductor (how much fault current a conductor can carry) is:
For every 42.3% of the circular mill area of a conductor that potion can withstand 1 ampere for 5 sec. Based on I2t (how fast the breaker or fuse will open and clear a fault) the withstand current rating of a #12 copper conductor at ½ cycles (standard circuit breaker operation) is 3,800 amperes. For #8 it is 9,600 amps. If there was a 1/8 cycle overcurrent protective device the withstand rating of that same copper #12 conductor is now 7,600 amps. For a #8 it is 19,200 amps. Since the fault has to return to the source, the amount of fault current (worse case) available is based on the source transformer SCA, and a point–to-point fault current calculation.

The basic purpose of overcurrent protection is to open the circuit before conductors or conductor insulation is damaged when an overcurrent condition occurs. An overcurrent condition can be the result of an overload, a ground fault, or a short circuit and must be eliminated before the conductor insulation damage point is reached. System components include wire, bus structures, switching, protection and disconnect devices, and distribution equipment, all of which have limited short-circuit ratings and would be damaged or destroyed if those short-circuit ratings were exceeded.

Since article 110.9 and 110.10 requires the overcurrent protective devices shall be selected and coordinated to permit the circuit-protective devices used to clear a fault to do so without extensive damage to the electrical components of the circuit.

Every component, conductor, etc..must be able to withstand any fault likely to be imposed upon it .

Maybe its just me but I don’t understand how the size of the conductors, or their resistance comes into the picture, since the code requires that the system be installed to withstand current at all fault levels.

Anyone else care to give us an opinion?
What say y'all

Mike,

You are talking circles around me. Let me see if I can simplify what I was trying to say. Table 250.122 requires a #8 wire for a 100 Amp circuit and #12 wire for a 20 Amp circuit. That was my point. I agree with your fault calculations.

What about my second comment.......that's the real topic here......

Ok lets get back to basics. What is the purpose of the grounding electrode system? In other words what is its function? What job does it perform in the electrical system?

The grounding electrode system is designed to:
1. Limit, or assist to dissipate lightning into the earth.
2. Limit voltages imposed by lines surges, or unintentional contact with higher-voltage lines.
3. Stabilize the voltage to earth durning normal operation.

The earth shall not be used as the sole equipment grounding conductor, or, effective ground-fault current path.

In a nutshell, the grounding electrode system is not used to provide a ground fault path. The earth impedeance is to high. It won't provide a path to the source to open the breaker.

When you consider that the biggest lightning antenna at your home is going to be the screen cage over the pool (which is bonded to pool steel), why wouldn't you want that bonded to the ground electrode system?
Maybe it is just my computer room background but we always tried to bond everything to everything in the "ground plane" with multiple paths. It is only when you get out to the signal lines that we worried about ground loops.
The only scenario I can see where having the pool as part of the grounding electrode system could be more hazardous is when your 15kv primary falls on the service drop. You will still have at least one 12ga or larger egc wire in there anyway, two if you have a light.
If you have a heat pump or toaster you will have that 8ga or larger path back home anyway. I have looked at some installations and when they put the Jandy Aqualink sub out there it is bonded to the pool which in turn bonds back to the panel via the EGC.

I think we have probably beat this one to death. Thanks for the insight, I do have a lot better idea of the issues involved.

The earth shall not be used as an effective ground-fault current path, however this does happen in an emergency. That is why I didn't quote the technical parts and I said that it has to clear a fault. I know, I know, I shouldn't have said it that way. I have seen entire buildings and homes where the neutral has been cut or became unattached where the structure was running entirely on the grounding electrode, and the earth was the return path.

While this is not allowed, if this didn't happen someone could get seriously injured or killed. The electrode provides this safety in the event of an emergency. That's why it's important to have good service grounding.

Consider the situation: A contractor has ran his #8 bare pool bond wire and attached it to the grounding electrode. What happens in the unlikely event that a neutral lug becomes unattached or burns up, and someone is in the pool? Do you know for sure?

Isn't it better to leave bonding alone, and not confuse it with grounding the pool? Why does everyone disagree with me? There is no reason to mix bonding and grounding. None whatsoever. Otherwise rewrite the Code to say "the pool steel shall be effectively GROUNDED." Otherwise leave it alone if no one is willing to change the language.

All that pool bonding does is mechanically bond the associated metal parts of the pool together, so that in the event that some metal part accidentally gets energized, and if someone is in the pool, everything stays at the same potential. That's it in a nutshell. The purpose of pool bonding is not to ground everything.

This issue is becoming more and more confusing as current code language and intent gets ignored time and time again. The intent of the Code is clear. If everyone wants to ground the pool bonding grid, then change the code language to say so.

As long as we are bonding all the things with EGCs in them like lights, motors and heaters, we are bonding the pool to the panel. That copper doesn't know if it is bonding or grounding.
Since we don't have a huge body count piling up I doubt you are going to see much in the way of a language change.

BTW I bet that house with the open neutral that was working fine had city water and copper pipes between the houses.

As for the people in the pool, they are birds on a wire. As long as they stay on the equipotential plane they don't really care what voltage that is.
In that respect I would like my pool bonded to the slab in my house or that first step into the dining room could be a doozy!

You are probably not going to believe this but I disagree. The number 8 bonding conductor serves an entirely different purpose and intent than the number 12 equipment grounding conductor. They are not the same, and you can't have it both ways. Remember even Mike agrees that this connection is unintentional.

BTW, The house running on the ground rod did not have copper pipes, and they had a well. Some soil conditions here in Florida provide an excellent return path as the water table in some places is so high.

As for the people in the pool, I agree with the birds on a wire - provided that pool bonding remain what the intent of the code is - pool bonding. They may no longer "birds on a wire" if you ground the pool with the #8, at least that's my take. Time will tell.

:rolleyes:

Well, the biggest lesson that I have learned today is that the main purpose of the grounding electrode system is NOT to clear the fault. To quote Mike,

The grounding electrode system is designed to:
1. Limit, or assist to dissipate lightning into the earth.
2. Limit voltages imposed by lines surges, or unintentional contact with higher-voltage lines.
3. Stabilize the voltage to earth durning normal operation.

For years (I'm going back to about 95) I have been taught that the main purpose of grounding was to clear the fault. I went to classes where that was taught. I know the people who taught this. I'll try to get over this misconception. Thanks everyone.

Yep that’s correct, but remember, even though the grounding electrode system is not intended to be used as a fault path for current to return to the source, current will flow on all paths. Its just that because earth impedance is so high, more current will take the path of least resistance, to open the breaker, but again current will flow on all paths.

And Nick I agree with your statement and I quote:
That’s exactly what I've been saying, if there was a problem for example, lets say the water has a potential of 20 volts, we bond all the parts per 680.26, that means the ladder has 20 volts the deck has 20 volts. See with all parts at the same electrical potential there is no shock hazard.

But if things were not bonded or bonded properly, if the water was 20 volts, and the handrail was 10 volts, and the deck was 5 volts, now we have a hazard because of the gradient of electrical potential differences between parts.

So you see even though personally I wouldn't connect the bonding grid to the electrodes myself, if it were done, and there was a fault, overvoltage, stray voltages, whatever, the bonding grid potential would be raised equal to the potential of the fault or problem. There should be no gradients, all electrical potential should be equal and there should be no hazard.

Mike I agree with everything you say except one thing. I agree about the mechanics of pool bonding, bird on a wire theory, etc., but I don't necessarily agree that there is no hazard if the #8 goes back to the grounding electrode, and I use a fault as an example.

There are too many variables. And what about pavers? Eveyone is ignoring this. That to me is a grounded surface. Lets say that the #8 is connected to the main service electrode and there is a fault, some part of the fault current travels to the pool bonding grid, person is in pool but touching the paver deck surface...??? What happens - I don't know. Maybe a serious shock.

Maybe I think too much and it gets me in trouble. I know I do. There are stray voltages all over the place here in Florida. The power companies do it on purpose. The electrical community as a whole tends to ignore this problem and the power companies don't want to raise the issue. Now we are going to bond the pool grid into the electrode system? Who knows what that can do. And now, household equipment is getting more technical and sensitive, a computer for example. Who knows what we will create?

And then there's the handrail that was added after the inspector already did his bonding inspection...happens quite a bit. Maybe something to consider.

Nick your 100% correct. Stray voltage is a problem. The utility uses the earth as a return path and this has created a stray voltage problem for pool decks that do not have rebar/wire mesh. Now when you use pavers as you say there is a gradient potential difference with the deck, and other parts of the pool, and equipment associated with the pool.

I had thought there was a proposal to require mesh under pavers so the mesh could be bonded to eliminate the very problem you describe. I haven’t kept up with that, but I'll check the draft and see if there were changes on that.

You know with a deck made of materials that are not or can't be bonded, or just bare ground around the pool I would in fact make them remove the bonding grid from the electrode system. Now as you said you do have a problem with voltage gradients that can't be corrected by bonding, and that is exactly the hazard you have been talking about.

Great point, I had forgotten all about that.

If we are going to take this gradient thing to the extreme lets go back to the interface between the house floor and the pool deck. The house floor is part of the Ufer grounding system and the pool deck is part of the bonding grid. Wouldn't you want them to be bonded? (or do you just hop into the house?)

This is still a distinction without a difference since there is already a solid copper path (call it what you like but the same current flows) between the pool and the GEC via the pump, light and heater circuits. If someone has toaster heat in the spa the EGC may be bigger than the pool bonding.

I would still have to say that the Code's intent is not to do it, and when we consider that not all houses are built with monopour slab construction adjacent to pool deck, it seems to make more sense.

I see your point though.

I think the entire pool bonding articles need to be revised:

1) Pump - require double insulated, no exceptions.
2) Heater - 680.26(D)
3) SPL J Box - design the installation so as not to require the bonding.

See, problem is solved (yeah, right!)