250.104(A) vs. low impedance path

[dana1028]

Scenario: Large commercial building going in [foundation stage]. The 2000A service is at one end of the structure, plastic water service is at the far end of the structure [300+ ft. away]; all interior water piping will be metallic [copper].
250.104(A)(1) says interior metal water piping systems shall be bonded and this can be achieved by bonding to an electrode.
The electrical contractor wants to use a 3/0 copper conductor and bond to the rebar at the far end of the building [i.e. 300+ ft. away from service location]....at the service location he will also use a 3/0 copper conductor to bond from the service equipment to the rebar.
As I understand the code, this is permissible and code compliant.
However, the inspector believes the purpose of interior water piping system bonding is to clear faults and does not feel 300+ ft. of rebar will provide the 'low-impedance path' necessary to clear a fault; thus he thinks a 3/0 conductor should be run from the interior metal water piping system directly back to the service to achieve this objective.
I understand the inspectors concern about the low-impedance path, but I also believe the contractor has met the code requirements with his bonding method.
Could I get some feedback please; both about whether the contractor's method is in fact code compliant and also addressing the low-impedance concern of the inspector.
[I also am an electrical inspector, but an observer in this issue].

 

[bob]

250.104(A)(1) says interior metal water piping systems shall be bonded and this can be achieved by bonding to an electrode.
The electrical contractor wants to use a 3/0 copper conductor and bond to the re bar at the far end of the building [i.e. 300+ ft. away from service location]....at the service location he will also use a 3/0 copper conductor to bond from the service equipment to the re bar.
However, the inspector believes the purpose of interior water piping system bonding is to clear faults and does not feel 300+ ft. of re bar will provide the 'low-impedance path' necessary to clear a fault; thus he thinks a 3/0 conductor should be run from the interior metal water piping system directly back to the service to achieve this objective.

The primary purpose of the grounding electrode is to connect the electrical system to earth. The GE system serves to ground the electrical system and to connect non current carrying equipment to the earth. This connection maintains the electrical system at or near zero potential with the earth during lightning storms and other high voltage surges. I would say that the primary return path for fault currents would be over the equipment grounding conductor and not thru the concrete and re bars or water pipes. Another point is that all available electrodes should be used that are available and not just the re bars.

 

[charlie b]

However, the inspector believes the purpose of interior water piping system bonding is to clear faults and does not feel 300+ ft. of rebar will provide the 'low-impedance path' necessary to clear a fault. . . .

I agree with Bob. The Inspector is wrong here. The Grounding Electrode System, including any and all electrodes and all conductors that connect them, has no role in the clearing of a fault.

There are several different types of faults, but only two that need be discussed here.

One is a fault from an energized metal part that is internal to a component to some external part that is not supposed to ever get energized. The fault current path is from the source, to the component, through the fault point to the metal case, to the EGC, back to the ground bar at the main panel, via the N-G bond, and thus back to the source. Neither planet Earth nor the GES is part of that path.

The other fault is from an energized metal part that is internal to a component to planet Earth. A simple example is using a damaged extension cord outside, perhaps over wet grass. The fault current path is from the source, to the component, through the fault point to dirt, along the dirt to a ground rod, up the GEC to the ground bar at the main panel, via the N-G bond, and thus back to the source. Although the GEC is part of this path, so too is dirt. The resistance of the dirt, and the resistance of the connection between the ground rod and the surrounding dirt, are so high that the fault current will not be high enough to trip the breaker. You could have an extremely low resistance GEC, even a superconducting GEC, and the fault path is still going to include dirt. So it doesn't matter what you choose as a set of grounding electrodes and as the related conductors, no aspect of the GES is going to help clear a fault.

 

[dana1028]

Thank you Bob and Charlie.

I believe we all agree (including the inspector) that the electrode system will not clear faults.

The issue is:
250.90 - Bonding shall be provided where necessary to ensure electrical continuity and the capacity to conduct safely any fault current likely to be imposed. [am I in error in thinking this language indicates the purpose of bonding is to clear faults?].

250.104(A) - "The metal water piping system shall be bonded..."

Yet 250.104(A)(1) says it is OK to bond the interior metal water piping system to an electrode.

My inspector friend sees that the contractor has met the code requirement of bonding the metal interior water piping system [by bonding to the Ufer]...yet there seems to be a performance conflict with the purpose of bonding ["the capacity to conduct fault current"].

If the purpose of bonding the interior metal water piping system is to clear any fault current likely to be imposed, how is this achieved by connecting to the electrode system? [as permitted in 250.104(A)(1)].

Again, thank you for your responses - your expertise in this area will be of great benefit in answering this quandry.

 

[bob]

Thank you Bob and Charlie.

The issue is:
250.90 - Bonding shall be provided where necessary to ensure electrical continuity and the capacity to conduct safely any fault current likely to be imposed. [am I in error in thinking this language indicates the purpose of bonding is to clear faults?]..

I do not think you are in error. The statement says "likely to be imposed".
If an energized conductor could, in some way, make contact with the water pipe then it must be bonded and present a low impedance path. If the pipe
is isolated by position in the plant, then its not going to make contact.
250.4(A).5 says the same thing.

250.104(A) - "The metal water piping system shall be bonded..."
Yet 250.104(A)(1) says it is OK to bond the interior metal water piping system to an electrode.

250.102(A)(1) says the the water pipe shall be bonded to the service encloser because the pipe is one of the grounding electrodes.

My inspector friend sees that the contractor has met the code requirement of bonding the metal interior water piping system [by bonding to the Ufer]...yet there seems to be a performance conflict with the purpose of bonding ["the capacity to conduct fault current"]..

Again, this comes back to the statement "likely to be energized".

If the purpose of bonding the interior metal water piping system is to clear any fault current likely to be imposed, how is this achieved by connecting to the electrode system? [as permitted in 250.104(A)(1)].

The water pipe will never see any fault as long as the conductors do not make contact. Most faults are line to ground when conductors make contact with conduit.

 

[George Stolz]

250.104(A)(1) says the the water pipe shall be bonded to the service encloser because the pipe is one of the grounding electrodes.

I made the correction to your reference. I disagree with your read on the section.

If a water pipe qualifies as a grounding electrode, then it will be (accidentally) bonded by virtue of following the requirements for connecting to it for grounding purposes, as an electrode. (250.50).

250.104 exists (IMO) to cover the other circumstance, where the water pipe does not comply with 250.52(A)(1).

If the purpose of bonding the interior metal water piping system is to clear any fault current likely to be imposed, how is this achieved by connecting to the electrode system?

IMO, because all the electrodes are electrically continuous and connected together by bonding jumpers and GEC's.

My question to your friend would be, 250.64(F) explicitly allows us to daisy-chain electrodes. If it is acceptable for a water pipe used as an electrode, then why would it not be so for a water-piping system not being used as an electrode? The requirements are (in general) more restrictive for electrodes than other equipment.

My other question would be: You can't have a fault without a source. If there's no predictable source (as in this case), how can you determine that 250.4 hasn't been satisfied? If a 4/0 feeder happens to energize the pipe for some reason, that would invite more fault current than a #14 would. How can he accurately gauge when the electrode would be effective enough?

It seems as though a gut distrust of the tying of the rebar is taking a front seat to the words of the code (250.52(A)(3)), in this case. The usual steel tie wires are good enough, as far as use as an electrode goes.

The fault current path is from the source, to the component, through the fault point to the metal case, to the EGC, back to the ground bar at the main panel, via the N-G bond, and thus back to the source.

That would be an argument to get rid of the requirement to bond non-earthing water systems, IMO. Since the EGC is supposed to be doing it's job, then what purpose is bonding the water pipe?

I am going to speculate: I think it is likely that the concern that originally put this requirement into the NEC is that an EGC could fail. Water pipes can have a high incidence of people touching them, especially people immersed in water. I'd like to imagine that people bathe fairly frequently.

Therefore, since low-resistance soaking wet people will be touching a conductive object, the writers of the NEC likely said, "Let's hedge the bets that it never becomes energized."

I believe that is what sets the water pipe apart, and forces us into bonding them - when gas pipes, air ducts, and structural metal may go unbonded if they're deemed "unlikely to become energized."

Anyway, that's just my opinion.

 

[W6SJK]

If you happen to have an electric water heater the EGC serves as the required bond. Likewise for the compressed air piping from the compressor. If you happen to have Mike Holts "Understanding the NEC" Vol 1, see fig. 250-164.

 

[George Stolz]

If you happen to have an electric water heater the EGC serves as the required bond.

Have you looked at 250.104(A) and (A)(1)? They require the bonding conductor for the water piping system to be sized according to Table 250.66. Most (if not all) EGCs for an electric water heater would be too small to comply with 250.104.

Likewise for the compressed air piping from the compressor. If you happen to have Mike Holts "Understanding the NEC" Vol 1, see fig. 250-164.

You can find many of these diagrams online, too.
http://www.mikeholt.com/freegraphics.php?id=2005

I think this is the one you are referring to:

 

[George Stolz]

Here is a picture to demonstrate that using rebar to connect electrodes is okay, in case he was unclear on that aspect. This is from Mike Holt, the NEC handbook shows a similar diagram.

 

[winnie]

I've been thinking about this one.

Can a piece of rebar going into a foundation, and another piece of rebar, in the _same_ foundation but 300 feet away, be considered reliably part of the _same_ electrode?

Clearly they are both concrete encased electrodes. And the normal tie-wires are considered suitable for joining various pieces of rebar to form a CEE...but a CEE is only required to be about 20 feet long.

-Jon

 

[hornetd]

Why bond at the far end?

Why bond at the far end?

Winnie
How is the length of the concrete encased electrode a problem. If that rebar is properly installed and tied it will behave as a single electrode. Twenty feet is a minimum length for that type of electrode but there is no maximum. I see no reason to make the bonding connection at the far end of the building. I would bond the nearest point on the interior metal water piping system to the service equipments bonded neutral bar and let it eat. Nothing in the code requires you to run a bonding conductor to within five feet of the point of entry of the non metallic underground water pipe. In this case you are clearly bonding rather than grounding. Run your bonding conductor to the nearest convenient point to the service equipment on the interior metal water piping.

 

[George Stolz]

Can a piece of rebar going into a foundation, and another piece of rebar, in the _same_ foundation but 300 feet away, be considered reliably part of the _same_ electrode?

Clearly they are both concrete encased electrodes. And the normal tie-wires are considered suitable for joining various pieces of rebar to form a CEE...but a CEE is only required to be about 20 feet long.

Jon, if they have not been made electrically continuous, then they would not be one electrode. Bonding one CEE to a ground rod and bonding a second CEE to a water pipe, does not make the water pipe bonded to the ground rod.

I don't think the description given in 250.52(A)(3) supports the idea that the concrete is the electrode - the concrete is the covering for the electrode, the electrode is still metal.

 

[winnie]

Tom,

You make a very good point. Since the water pipes are _not_ going to be a grounding electrode, the bond connection should be made where convenient, presumably much closer to the service panel.

The problem that I have with the length of this electrode is that it is being used as both a grounding electrode and as a bonding conductor.

A grounding electrode is not expected to carry significant fault current. A bonding conductor is expected to carry fault current.

It is pretty clear that tie-wires and re-bar provide sufficient continuity for use as an electrode. But I have a hard time with the idea that these joints could safely carry the fault current that a 2000A service could source.

However 250.104(A) seems to allow just this connection. Unless one can make a strong argument that the two separate pieces of rebar are not reliably part of the same electrode, which is why I asked the question. Perhaps the CEE is a suitable _bonding conductor_, but only if all of the tie wires are inspected and the contact points cleared of any rust prior to the concrete pour

This is a situation that IMHO the code allows but which I am not comfortable with in my gut

George,

I agree that it is the steel and tie wire is what makes the electrode electrically continuous. I am not at all attempting to suggest that one tie the service to one electrode and bond the water piping to a separate electrode; that would be a recipe for disaster The original question is: can the entire CEE be treated as a bonding conductor. Frankly I think Tom's suggestion of bonding the pipe at a more convenient location makes far more sense than trying to use the CEE in this fashion.

-Jon