The article "CEC Methods of Grounding" in the latest edition of IAEI magazine is a perfect example of why the NEC should change the way it calls and defines aspects of grounding and bonding.

A special task committee was established after the 2005 NEC was released to investigate several proposals that were submitted to change the term equipment grounding conductors to equipment bonding conductors and perhaps provide a definition for "neutral".

While reading this article, it is clear that some of the functions that are being described as grounding are actually functions of bonding.

The purpose of grounding should be completely distinguished from the purpose of bonding. It has led to significant misconceptions on what is acheived when connecting non-current carrying metallic parts together verses connecting electrical equipment to Earth.

It is my understanding this task group has struck down the proposal and most if not all the terms and definitions regarding grouding and bonding will remain the same for the 2008 NEC. I think this is a mistake, and the article in the IAEI magazine proves it.

As an instructor and inspector, one of my greatest burdens is the explanation of the purpose and use of grounding and bonding. As long as the code and articles like this one continue to mix the two fundementals and describe the two aspects as one process and purpose, electricians will remain confused and misconceptrions will continue to flourish.

I'm not implying the article has incorrect information, however it does not provide a clear understand that the connection to earth is not the major function that the "code defined" grounding serves but is really a function of bonding.

I realize this article is referring to the CEC, however this same issue applies to nearly every article, paper, or standard written on grounding and bonding issues.

If you are a purist, there is really just one "grounding" conductor in most buildings. It goes to the ground electrode. Every other grounded conductor is either bonding equipment or carrying unbalanced circuit current.

You're the man Greg!

That sums it up nicely!

Of course this kicks the "grounding vs bonding" tar baby in a swimming pool.

The 680.26 "bonding" grid will certainly be the best "ground" in the system but we can't call it a ground electrode. Someone (Mike?) says he is not sure it should even be connected to the panel GES but that is unavoidable if you have a bonded pump or an underwater light. (via the EGCs)
Generally speaking we say the 8ga in 680.26 is "bonding" and the EGCs in the various equipment sections is "grounding". That actually seems backward if you have that big "Ufer" we call a concrete swimming pool and deck.
Personally I take some comfort that my screen cage is "grounded" to that big electrode, in case lightning does hit it. I went as far as to bring up a bunch of extra #8s to bond it several times around the perimeter. I figure it has become a Faraday cage of sorts.

NO don't connect the pool grid back to the main panel, please don't get me started on that one again...

The code sets up this problem. It tells you to avoid objectionable current yet permits and/or requires installaton that practically guarantee it. All normal operating current is supposed to return on the service grounded (neutral) conductor.

However, the code permits a significant reduction in the size of the grounded conductor per 220.61 - 2005 NEC. At the same time, the NEC has reqirements that are intended to reduce the grounding electrode system impedance to as low as possible. This creates a situation where more current will flow on grounded parts and through earth as opposed to the insulated system where it should be. Ohm's Law clearly proves this. The pool equipotential grid can produce the same result. Normal operating current will find its way through the bonding grid if it is of sufficently low imdeance and connected to the service equipment.

The only true way in which the NEC can get rid of this inconsistancy and provide a way in which systems are completely free of objectionable currents is to require equipment grounds (bonds) to be installed with the service entrance and all equipment be grounded through that. the only connection to earth would be at the (isolation) transformer.

Nick, I have 4 pumps, 2 heaters and 2 lights that are bonded together (680.26) and also are required to be "grounded" back to the panel. It is hard to say, with a sraight face, that the pool is not connected to the ground electrode system by the 65670 circular mils of copper in the EGCs. In fact the 4ga ground electrode conductor is the bottleneck.

I think you are all confused, except Nick.

Mike it is too soon to say if I'm confused or not (because I didn't read the article yet). Who authored it?

I was holding back on this topic because if we start on grounding vs. bonding it is going to go on forever, but maybe we need to do that. It has been a while I think.

Mike if you are out on your bike this weekend, tonight I will be at the Orlando Harley Davidson for the show there.

What am I confused about? Do you have some way to explain why current won't flow from the pool (680.26) bonding grid to the ground electrode system via the EGCs
In fact the current will be flowing the other way in most cases since a concrete pool is several orders of magnitude more effective than any 250 chapter III electrode. You may hate the idea but you can't stop it from happening without extensively rewriting 680. The slam dunk is the underwater light junction box. I would also be very troubled if an 11 kw immersion heater was not solidly bonded to the service and part of the pool bonding grid.

Who isn't confused when it comes to the grounding and bonding? I would bet this is the least understood aspect of the code and electrical installations in general.

What's confusing? Bonding assures everything bonded together is at one potential, grounding insures it is at some close approximation to the potential of the surrounding dirt.
In Florida that level may be many volts different than the dirt under building across the street when either is affected by a fault or a transient.

In the computer business we would "bond" interconnected equipment if it was served by different services and ground electrodes to prevent lightning damage. We used a conductor in a color other than white or green and called it a "drain" to stay away from article 250 and all that "confusion".

I'm just saying that based on the way the NEC defines terms and establishes requirements, it creates confusion.

I will agree the NEC does mix the terms bonding and grounding indiscriminately.
The terms are used somewhat interchangeably several hundred times in the code.
I think 100 years of tradition is going to be hard to overcome.
I have a hard time not saying "MCM" when I talk about fat conductors.

The earth shall not be used to provide an effective ground-fault current path to return fault-current to the source to open an overcurrent protective device. 250.4(a)(5).

Bonding at a pool is not to return fault-current to the source to open an overcurrent protective device.

The earth, (grounding electrode system) is used to: (250.4(a)(1)
1. limit voltage imposed be lightning.
2. line surges.
3. unintentional contact with high-voltage lines.
4. stabilize voltage to earth.

So when that string of xmas lights, hanging in the window frame, shorts out, you don't want the O/C device to operate?

Bryan is this article online at iaei.org? I looked for it and could not find it. My mag is on the other side of the state at the moment.

I doubt it is online. The magazine just came out. It seems the online articles are from an edition or two previous.

Everything Mike stated is correct (per the code -but not in reality), and everything in the article is technically correct. What makes it confusing is that the author describes both 250.4(A)(1) and (A)(5) as being acomplished by grounding.

I believe this leads to people believing the connection to earth is what establishes effective fault paths. I have been posting comments and topics on the Mike Holt Forum for over three years (4,000+ posts) and at least one post per week is by some electrician or inspector not understanding the purpose of connecting equipment to earth verses connecting all the equipment together.

Yes, but...

...(I am playing the advocate here) even if people are mistaken, can you cite any specific examples of where a system grounding installation is made worse for their lack of knowledge?

Wouldn't people tend to overground?

Is it even possible to overground?

I can remember back when I knew this one fellow, he actually used to hot check a dwelling unit by hooking up 220V, and using the ground rod as the neutral return path (PLEASE don't try this at home)! The lights ran but ran dim. So if current CAN be sent back to the source via the grounding electrode system, what harm does it do if an electrician THINKS this is the purpose of grounding - WHEN in an emergency situation it may perform as such anyway, depending upon soil conditions, etc.?

I agree with Bryan.

That's why we bond the neutrals and grounds at the service, to provide a fault path to the source to open (trip) the breaker and clear the fault.

Mike, yes, that is why we bond the neutral and ground, why don't you think we should be bonding the pool equipotential grid to the service ground/neutral bus?
As I pointed out in my example, if you have a fault to a pool bonded chunk of metal (window frame, ladder, rail, diving tower etc), don't you want to operate the O/C device?

Like the grounding electrode system, the bonding equipotential grid has a specific job or function. Clearing a fault, returning current to the source, to open an O/C devise is not its job. There are EGC's for that purpose.

Oh no....

pool bonding tied into the ground topic....

IT HAS BEGUN

(I am not responsible for what I say from this point on...)

Mike the EGC will only clear a fault to the case of that equipment. If the fault is to a "non-electrical" chunk of metal like a window frame or diving board stand the ONLY way to clear the fault is to get back to the panel. Fortunately one of those paths we don't like to talk about (like the "equipotential" to pump connection) will do it. What happens if there isn't a light and the pump is double insulated. I am not sure how you clear that fault. It would certainly scare some worms up from around the pool deck until it did.

How would a window frame or diving board stand become energized?

I agree with Mike. The equipotential bonding serves a specific purpose. Equipment bonding serves another. The service grounding still another. In most cases, they don't work together.

Rope lights, radios, fans or any other equipment that might get dragged out there. If there was no chance of being energized we could eliminate 680.26.

680.26 is not for the purpose of clearing faults from the premise wiring. If a window frame or other metallic object is likely to be energized by a circuit, it should be bonded to the egc supplied with that circuit and not by the equipotential plane created at the pool. The equipotential plane is established to prevent step and touch potentials created by voltage gradients in and around the pool.

There is a huge difference between these functions. There doesn't even need to be power supplied to a structure or pool for voltage gradients to be present. This is why it is important to understand the difference in the purpose and the function of all the grounding and bonding methods we employ.

If things on the bonding grid couldn't be energized there wouldn't be a gradient to deal with.
I understand that clearing faults is not the intent but it is a good byproduct.

As for how these things might get energized, If homewowners didn't do dumb stuff we could carry the NEC around on a wallet card.
It will not be fixed in place equipment that causes most electrical hazards in a pool. It is the crap they buy at Walmart and drag out there.

Sure you can. If you have metal or conductive parts that are separated by an inusulating or semi-conducting material, a capacitive effect is created. Effectively, a battery is formed. Current will flow from one "plate electrode" to the other "plate electrode" through a person or animal. By connecting these two electrodes togther, you eliminate this potential between them.

So as you can see, the source of voltage is not from the premise wiring or the service. It is created by voltage gradients in and around the pool. Bonding all the metal parts together places everything at the same potential. This system should not be relied on to clear ground faults.

As long as you have a light the bonding grid will be connected to the service, as will the connection to most pumps, pool covers, heaters and other electrical equipment. You will have to get over the idea this is an isolated system. I don't know why you think it wouldn't clear a fault or why you wouldn't want it to.
I still say most gradients are going to be from utility power ... somewhere. Capacitive or galvanic currents will be too low to ever be noticed without some very sensitive lab gear.

I am not saying it is an isolated system, I'm saying it should be. The connection to the pool equipment is in my opinion a bad practice though required by the code. Once connected to the premise wiring system, it becomes a path for normal operating current which creates even more problems.

"Capacitive or galvanic currents will be too low to ever be noticed without some very sensitive lab gear."

Now if that were true, 680.26 could be deleted.

From what I read most stray voltage is ultimately traced back to the grid. (leaking underground feeders etc) We don't hear much about cattle out on the range or Amish farmers having stray voltage issues.
If we followed the advice of those who say the 680.26 system should be isolated from the system ground how would you ever clear the fault from the window fan that energized the <bonded> window frame hence the rest of the pool?
I can come up with plenty of scenarios where you could energize the grid, doing the simple things end users do.
Personally I think everything should be grounded/bonded (steel framing, aluminum siding, gutters or whatever). All I would have to do to justify it is to drive around on Christmas eve and look at how the decorations are installed.
That doesn't even address the use of power tools.

I agree. Many times stray voltage is the result of excess voltage drop on the utility (distribution) neutral.

Whatever circuit is likely to energize a non-current carrying metallic part or material should be used to bond that part or material. This circuit should be used to clear the fault. If the window frame is within 5' of the pool's edge, it should be bonded to the equipotential grid to prevent anyone touch potential from being established.

Two separate issues.

I agree with Bryan - two separate issues.

The pool pump motor can be a double insulated pump motor and that would eliminate the apparent interconnection at that point. BTW, I had an old Soares grounding book (don't have it anymore or can't find it) where it stated that the connection at the pool pump motor was not "deliberate" and that it should not be considered as "interconnected."


Two distinctly different things, different purposes. If you run the pool bond back to the main GEC (discussed in this forum in the past that nothing in the code to prevent this) it is making things worse, not better.


:p

How do you reconcile the fact that we connect the grid and the EGC in the light J box, no matter what kind of light they have? I assume you could
rewrite the code to outlaw 99% of the pool pumps, lights, heaters and pool covers that do have an EGC but that may take 30 years to finally get them all out of the system. They you have simply made the system LESS SAFE.

I still don't see how this is a "separate issue" If the circuit that energizes the nonelectrical pool metal like that window or the screen cage is not grounded equipment you don't have that path to clear the fault. It could be Christmas lights with a 2 wire cord, (radio, TV or any other small appliance with a 2 wire cord). Where is the ground path to clear the fault? Are you just going to let the pool cruise at 120v, sanguine that it is "equipotentially safe"?
Pool safety is based on many redundant systems to insure safety. When you remove the connection to ground you lose one. When the last connection between the grid and ground dissapears on any grand scale, I bet you will see proposals to bond it to the panel.

I promised at the beginning that I wasn't going to get into this again...but,

simply put, the pool bonding serves to keep all metal at the same potential. It's the "bird on a wire" thing. Consider the person in the pool to be the bird. It doesn't matter what the voltage is.

That's it in a nutshell!

I understand the concept but why is there a problem if you make sure that "wire" is at the same potential as all the other things we bond for safety?
You still haven't said why bonding the pool to the grounding system is a bad idea. All we are doing is making that "wire" the whole house and not just the pool. Your concept still stands. It just makes the safety zone larger.

When you connect the pool equipotential grid to the premise wiring system, the pool effectively becomes a grounding electrode. This electrode may have sufficiently low enough resistance to earth to allow objectionable current to flow back to the utility transformer via the #8 grid instead of back on the insulated service grounded conductor where it should be.

Also, anywhere improper neutral-to-case bonding is made at either the home with the pool or even a neighbor’s home will potentially lead to current flowing on pool parts back to the source.

Imagine the utility transformer serving two homes is one the lot line behind the two homes. Both homes have a pool. The #8 grid from both pools connects back to the service equipment at each house. It is reasonable to believe that if the service neutral were to open on one of the houses, the normal operating current could conceivably flow back through the main bonding jumper, into the #8, into the earth, and back to the source transformer right next to the pool which happens to also be grounded right next to the pool.

It is connected via the EGC connected to the motor, light fixture, heater etc..

Bryan your petard is showing.
If it is an equipotential grid and you are really the bird on the wire, why do you care if you are flowing the next door neighbor's neutral current?
You didn't care when it was the "twinkle lights" duct taped to the handrail and shorted to the 120v phase.
In fact you didn't even want me to operate the branch circuit O/C device. We were just floating safely at 120v above ground.

Personally I want fault current, anywhere near the pool, to flow to the MBJ and open the O/C device.

There is no good reason to knowingly ground the occupant(s) of a pool.


goodnight

I never said any of that. I do care if metal parts become energized by a circuit. However I feel it shoulf be the circuit that energizes the part that clears the fualt, not the pool grid.

I do believe all parts should be bonded together at the pool to eliminate step and touch potentials.

But as soon as grid is connected to the premise grounding system, it becomes apart of normal operatiing current.

I keep telling you that these are different issues, and you keep making them all the same.

Voltage gradients at the pool are to be minimized by the pool equipotential grid. Ground faults should be cleared by EGC's of the supplying circuit. And grounding should be made at the service and only the service and it would be advisable to limit how much normal current will flow through the pool system.

Again, three different issues.

You all keep assuming the equipment that energized the pool will have an EGC. Look around your house and note the number of consumer lighting units, TVs, radios, stereos or other things that might ge dragged out to the pool deck that have a 2 wire cord.
I see that window frame as the most likely place where a non-GFCI might energize the grid. I can see a kid setting their bedroom radio on the window sill and later closing the window on the cord. The same could happen with holiday lighting.
I know the user is doing something potentially hazardous but that is what users do.
Didn't you just write a proposal for CO detectors based on the number of people who run generators in the house?

This is not as much about intentionally bonding the pool to the grid, that already happens. It is about your contention that we should intentionally ISOLATE them. That is virgin territory. Pool equipment has historically made this connection in several ways. I think if/when that connection goes away we will see tombstones.