Bonding neutral and ground at service end box in NYC

[infinity]

Why do no code-knowledgeable professionals go to the trouble of always using bonding lucknuts to avoid parallel paths on service raceways and enclosures?

Can you explain with some further detail how a bonding locknut avoids parallel paths for the neutral current.

 

[Strathead]

If you bond sooner, you have to run separate EGCs and wire everything as you would feeders and sub-panels.

Unless this 'service end box' is in/on a separate structure.

It is actually now, a "supply side bonding jumper" sized per 250.102, not an equipment grounding conductor sized per 250.122.

 

[wwhitney]

Wayne that was rather a lot of words to point out a tiny exception that makes what I said 99% instead of 100% true.

If that was too verbose, the short version:

250.92(B)(1) (bond the enclosure to the grounded service conductor) is not mandatory, it's one option among several. I'm unclear on why the industry standard practice seems to be OK with parallel paths and always bonds enclosures upstream of the service disconnect to the grounded service conductor. Sometimes that just creates a parallel path and the required bonding can be done other ways.

Cheers, Wayne

 

[wwhitney]

Can you explain with some further detail how a bonding locknut avoids parallel paths for the neutral current.

Sorry, I thought I had a lot of detail already.

But basically if bonding the grounded service conductor directly to an enclosure upstream of the service disconnect means that there are two parallel fault clearing paths, then 250.92 doesn't require you to do make that bond. It allows you to ensure that the other path meets the standards of 250.92(B) instead. Which means any normal RGC/KO connections would need a bonding locknut or the like, per 250.92(B)(4).

Cheers, Wayne

 

[jaggedben]

Can you explain with some further detail how a bonding locknut avoids parallel paths for the neutral current.

It wouldn't necessarily, but at least hypothetically it could. Suppose my setup is weatherhead > RMC > CT cabinet with isolated neutral bar > more RMC > service equipment. If I use bonding locknuts on the CT cabinet and service equipment, and I don't have concentrics or eccentrics, then the requirements of 250.92 are met without bonding the CT cabinet directly to neutral. So then I don't have a parallel path on the RMC.

And I only said parallel paths on service enclosures and raceways. Ground is still a parallel path, however theoretical or not.

 

[infinity]

It wouldn't necessarily, but at least hypothetically it could. Suppose my setup is weatherhead > RMC > CT cabinet with isolated neutral bar > more RMC > service equipment. If I use bonding locknuts on the CT cabinet and service equipment, and I don't have concentrics or eccentrics, then the requirements of 250.92 are met without bonding the CT cabinet directly to neutral. So then I don't have a parallel path on the RMC.

And I only said parallel paths on service enclosures and raceways. Ground is still a parallel path, however theoretical or not.

Theoretical yes but I don't think that any POCO would allow only the raceway to bond the CT cabinet on the line side of the service disconnect.

 

[jaggedben]

Theoretical yes but I don't think that any POCO would allow only the raceway to bond the CT cabinet on the line side of the service disconnect.

Exactly. Bonding with the grounded conductor is how it's done.

 

[wwhitney]

Exactly. Bonding with the grounded conductor is how it's done.

OK, but then it's a POCO requirement, not an NEC requirement. If the AHJ were to consider the grounded conductor current on exposed metal parts of the service to be objectionable current, then 250.6 would require you to take countermeasures.

I see the upside to requiring bonding to the grounded service conductor, as it is something that can be checked locally, without requiring a review of the entire service.

Cheers, Wayne

 

[wwhitney]

OK, but then it's a POCO requirement, not an NEC requirement.

FWIW, PG&E's (California's largest utility) Green Book allows the arrangement I described, although it calls it a "nonpreferred method," so I could imagine one could get some pushback from staff not familiar with the option. It says:

3. Equipment Bonding: Install an equipment grounding conductor (EGC) between the PG&E service termination enclosure (e.g., meter panel) and the service disconnect enclosure. The EGC is required in this layout. Run the EGC (i.e., bonding wire) through metallic conduit and attach it to the inside of each enclosure. Use Myers hub fittings to connect the conduit with the enclosures. These fittings are Underwriters Laboratories (UL) certified for bonding and are required to connect the conduit with the enclosures. PG&E and the local authority having jurisdiction must approve the conduit and fittings. The customer bond wire is allowed to terminate in a manufactured bond lug installed at the bottom of meter panel.

I guess they are requiring a redundant wire-type SSBJ in addition to metallic conduit as the wire-type SSBJ is easier to field verify than checking every conduit connection. Not too surprising they misidentify the SSBJ as an EGC.

Cheers, Wayne

 

[gene6]

Gentleman in NYC your not allowed to bond in the con-ed box as its tagged and sealed, the bond must be accessible in the service gear. As others have said the neutral bus is bonded to the con-ed box.

 

[electrofelon]

Gentleman in NYC your not allowed to bond in the con-ed box as its tagged and sealed, the bond must be accessible in the service gear. As others have said the neutral bus is bonded to the con-ed box.

This doesn't make any sense, I think it's poorly worded. Do you mean you are not allowed to connect the grounding electrode conductor in the ConEd box?

 

[electrofelon]

It wouldn't necessarily, but at least hypothetically it could. Suppose my setup is weatherhead > RMC > CT cabinet with isolated neutral bar > more RMC > service equipment. If I use bonding locknuts on the CT cabinet and service equipment, and I don't have concentrics or eccentrics, then the requirements of 250.92 are met without bonding the CT cabinet directly to neutral. So then I don't have a parallel path on the RMC.

And I only said parallel paths on service enclosures and raceways. Ground is still a parallel path, however theoretical or not.

Theoretical yes but I don't think that any POCO would allow only the raceway to bond the CT cabinet on the line side of the service disconnect.

I frequently use bonding locknuts on both ends to make a raceway an acceptable SSBJ. I find doing this is cheaper and easier at times than bonding a service box or wireway directly to the neutral conductor. I dont recall ever having the required layout to bond a CT cabinet this way, but I wouldnt hesitate to unless the POCO had some specific requirements.

 

[gene6]

This doesn't make any sense, I think it's poorly worded. Do you mean you are not allowed to connect the grounding electrode conductor in the ConEd box?

Yes main bonding jumper and grounding electrode conductor cannot terminate in the con-ed box. They typically terminate in the service disconnect. It is very typical to install a common grounding bar for the installation the OP describes or terminate all GEC's all in one service disconnect and run GEC taps to the others. The con-ed box is bonded to the neutral, like a meter can typically is.

 

[electrofelon]

Yes main bonding jumper and grounding electrode conductor cannot terminate in the con-ed box.

A MBJ would NEVER go in an endbox regardless of utility requirements. An MBJ only goes in a service disconnect. The jumper thingy in the end box would be a SSBJ.

 

[wwhitney]

The con-ed box is bonded to the neutral, like a meter can typically is.

Is that a published Con Edison mandate, for a direct bond inside the service termination box, or just common practice? Because for the NEC, it's not the only option.

Just curious about various utility requirements; so far all I know is PG&E encourages it but doesn't make it mandatory.

Cheers, Wayne

 

[electrofelon]

Is that a published Con Edison mandate, for a direct bond inside the service termination box, or just common practice? Because for the NEC, it's not the only option.

Just curious about various utility requirements; so far all I know is PG&E encourages it but doesn't make it mandatory.

Cheers, Wayne

Seattle City Light specifically disallows bonding of a CT cabinet directly to the neutral/grounded conductor:

Connection to the grounded service conductor (neutral) must not be used to bond current transformer enclosures. Bonding must be derived from the service main grounding point.

 

[Tainted]

The con-ed box is bonded to the neutral, like a meter can typically is.

Wouldn't that be the main bonding jumper? You just said that main bonding jumper cannot terminate to the service end box

 

[Tainted]

Gentleman in NYC your not allowed to bond in the con-ed box as its tagged and sealed, the bond must be accessible in the service gear. As others have said the neutral bus is bonded to the con-ed box.

I can't think of a time where Con-Ed sealed their service end box. Or are you talking about a tagged and sealed CT cabinet?

 

[electrofelon]

Wouldn't that be the main bonding jumper? You just said that main bonding jumper cannot terminate to the service end box

See post #54. I cant think of any reason there needs to be a distinction between MBJ's and SSBJ's, seems like the code could have just lumped them all together and called them one thing.

 

[jaggedben]

Wouldn't that be the main bonding jumper? You just said that main bonding jumper cannot terminate to the service end box

No, a main bonding jumper by definition is in a service disconnect enclosure.