Supply-Side Connection Bonding

[c_picard]

Understood. I'm viewing this enclosure throught the lens of available fault current on the utility side.

That's the road block with the AHJ. We would like to install using the safety standards set in 230, the state of NJ would like to apply the safety standards of the "inverter output circuit", much less strict. Using that logic, we don't even need a service rated disconnect.

For practical purposes, even a slight variation on the AC side of the inverter, and that baby is shutting down.

 

[jaggedben]

Code issues aside, I am a bit confused as to why it would be a safety issue if the ground and neutral were not bonded in the PV disco. The neutral in a supply side tap has continuity with the neutral at the meter and therefore with the neutral in the building MDP. The same is true for the ground.

Your assuming that there is a ground. But if an installer follows Art 230 then there isn't, or at least may not be, or it may not be adequately sized. The installer brings only the grounded conductor to the disconnecting means, and from there fault current returns through the N-G bond. See texie's comment above about what happens if a non-metallic raceway is used. If an installer has done it this way, it's just wrong to ask him to lift the N-G bond.

The neutral and the ground are connected at the service, so anywhere you check, there is continuity between neutral and ground, and neutral is never switched. Where is the hazard?

I think the hazard would be in having the full fault current available from the utility transformer flowing back (in a roundabout way) through either an EGC or a solar GEC that is undersized with respect to 250.66. With AC modules that could be a #12 EGC.

If this isn't a service enclosure, then I don't need the GEC, which means I'm asking the 10AWG equipment ground to carry the fault back to the main panel. I also can install a 30A pullout A/C disconnect from home depot, tap using 10AWG romex, and call it a day.

And that just doesn't seem too right.

That's the road block with the AHJ. We would like to install using the safety standards set in 230, the state of NJ would like to apply the safety standards of the "inverter output circuit", much less strict. Using that logic, we don't even need a service rated disconnect.

I'd say they should at least let you do it whichever way you want. Declare it a new set of service entrance conductors per 230.40 Exception 5, and follow 230. Or don't, and follow all the other rules, with an EGC. But it's just not right to not let you install a new set of service entrance conductors, if that's what you want to do and how you want to designate them. 230.40 Exception 5 clearly says you can do so.

230.40 Exception 5
230.40 Exception 5
230.40 Exception 5
230.40 Exception 5

(And, once they are so designated, then 230.70 clearly designates it a service disconnecting means, and 250.24(B) and (C) clearly apply.)

 

[texie]

As I said in this post earlier and the other thread on this subject, IMO this is not an option or debatable to not bond the neutral and fully comply with 250.24. I'm not alone on this and even John Wiles points this out in his materials. There can be serious safety issues by not bonding under the right conditions. As a license holder I would never even think about not following 250.24 and would take the AHJ to the mat over this if he saw otherwise.
This is not an option. Period.

 

[jaggedben]

As I said in this post earlier and the other thread on this subject, IMO this is not an option or debatable to not bond the neutral and fully comply with 250.24. I'm not alone on this and even John Wiles points this out in his materials. There can be serious safety issues by not bonding under the right conditions. As a license holder I would never even think about not following 250.24 and would take the AHJ to the mat over this if he saw otherwise.
This is not an option. Period.

The more I think about this, the more I agree and wish to backtrack from my previous comments suggesting it might be possible by code to bond a supply-side PV disconnect with an EGC instead of a grounded service conductor. In addition to what's already been mentioned, I'll add the following: There are no rules besides those in 230 that can be applied to tapping service conductors. A conductor tapped off a service conductor does not meet the 240.2 definition of a tap conductor, because there is no "overcurrent protection ahead of its point of supply". And the rules in 240.21(B) cannot be applied for the same reason. So regardless of the grounding and bonding issue, this really can not be done.

 

[ggunn]

The more I think about this, the more I agree and wish to backtrack from my previous comments suggesting it might be possible by code to bond a supply-side PV disconnect with an EGC instead of a grounded service conductor.

Bonding what with the EGC are you speaking of? What part of the disco? The case? The neutral?

 

[jaggedben]

Bonding what with the EGC are you speaking of? What part of the disco? The case? The neutral?

The case. As you would for a load side disco.

The only reason I brought it up is that I assume that something such as this must be what is being asked of c_picard by the AHJs he is dealing with.

But I really agree with texie that that this shouldn't be open to debate. Art 230, and the parts of 250 dealing with services, are the only rules in the code that one can reasonably look to when tapping service conductors.

 

[texie]

The case. As you would for a load side disco.

The only reason I brought it up is that I assume that something such as this must be what is being asked of c_picard by the AHJs he is dealing with.

But I really agree with texie that that this shouldn't be open to debate. Art 230, and the parts of 250 dealing with services, are the only rules in the code that one can reasonably look to when tapping service conductors.

Glad we could come to the same conclusion. To me, it is an interesting subject and question. Now if I could just get Don and Infinity to come around on the other thread about "special" bonding for GEC raceways....

 

[GoldDigger]

Now if I could just get Don and Infinity to come around on the other thread about "special" bonding for GEC raceways....

I doubt that you will get them to change on that. For the case of high fault current or high frequency lightning-induced current, the choke effect of the unbonded ferrous shell will significantly increase the impedance of the ground wire. As mentioned, it is similar to running the wire through a ferrite bead.
The reason that any ferrous enclosure needs to be bonded it that the wire is effectively making one turn through an iron choke core, and there is no way that you can offset the resulting inductance. For situations where the inductance is significant, you must allow the current to flow through the enclosure itself as an alternate path with no added inductance.
In this thread you have physics on your side. In that thread you have physics against you.

 

[c_picard]

As I said in this post earlier and the other thread on this subject, IMO this is not an option or debatable to not bond the neutral and fully comply with 250.24. I'm not alone on this and even John Wiles points this out in his materials. There can be serious safety issues by not bonding under the right conditions. As a license holder I would never even think about not following 250.24 and would take the AHJ to the mat over this if he saw otherwise.
This is not an option. Period.

UPDATE from the mat:

The State of NJ's rep at DCS has informally advised me that the state's view is that the Grounded conductor shall not be bonded at the PV disconnect. Only the bonding requirements of 690 apply, and Art 230 does not apply to PV at all. No more green screws in NJ, I guess.

 

[texie]

UPDATE from the mat:

The State of NJ's rep at DCS has informally advised me that the state's view is that the Grounded conductor shall not be bonded at the PV disconnect. Only the bonding requirements of 690 apply, and Art 230 does not apply to PV at all. No more green screws in NJ, I guess.

This is an interesting situation. As I said in my earlier post, what if the installer chose to use a PVC nipple to the solar disconnect? Now if a conductor from the service side were to short against the metallic enclosure the fault return path would be the long way back VIA the EGC to the source. This seems to me that this could have catastrophic results. Also, as I mentioned, the "experts" such as John Wiles and others seem to agree that bond is required. Maybe this is something that needs clarification in a future NEC issue.
If anybody can make a case that says otherwise, I'm all ears.

 

[c_picard]

Yes, interesting indeed. Mr. Wiles tried his best to get a proposal in the 2014 that would explicitly reference 230 in 690. The CMP rejected it, and the reasoning was that it didn't need to be said. Evidently, it does need to be said in NJ. The real culprit here is that ridiculous graphic, that someone thought was helping, in the beginning of 690. In NJ, everything between the inverter AC terminals and the supply side connection is considered the "inverter output circuit", including the pv system disconnect, and the additional set of service entrance conductors, which I guess aren't service entrance conductors at all. So what the heck are they?

I don't even think a request for formal interpretation will work here. How would I phrase the question?

"Does 250.24(C) apply to the grounded conductor of the set of service entrance conductors permitted by 230.40, exception 5, at the photovoltaic system disconnect enclosure installed at a supply side connection allowed by 230.82(6)?"

Yes, or No?

Even if they answer yes, NJ will reject this because it doesn't say 690 anywhere in there! ***BANG HEAD HERE***

 

[Zee]

Well, here is a very simple thought. Is it possible that they think the Q is whether PV itself - whether load side or supply side tied to the grid - is a new service or not?
As in separately derived power source.
I assume you made it clear that you are only talking about SUPPLY side connected PV AC discos - not AC discos in 95% of the cases that are load side connected (especially residential).

 

[GoldDigger]

I don't even think a request for formal interpretation will work here. How would I phrase the question?

"Does 250.24(C) apply to the grounded conductor of the set of service entrance conductors permitted by 230.40, exception 5, at the photovoltaic system disconnect enclosure installed at a supply side connection allowed by 230.82(6)?"

Yes, or No?
...
***BANG HEAD HERE***

***Pat you on the shoulder and say "There, there now...."***

The way to phrase your question would be, IMHO:

"Does 250.24(C) apply to the grounded conductor of the set of service entrance conductors permitted by 230.40, exception 5, at the photovoltaic system disconnect enclosure installed at a supply side connection allowed by 230.82(6), notwithstanding any provisions of 690?"

Yes, or No?

Another way of looking at it would be:
"Does 250.24 apply to connections and disconnects between a PV-sourced AC system and a utility service?"
That is important since the heading of 250.24 is "250.24 Grounding Service-Supplied Alternating-Current Systems." and that leads to the question of whether "Service-Supplied" limits it to systems whose power source is the service or applies to any system that is connected to Service wiring (including therefore inverter output circuits.)

On the other hand, the question I would like never to see come up is:

"Does the term "photovoltaic output circuit" refer only to the DC circuit(s) which lead to the DC input of the inverter or is it intended to apply to the inverter output circuits as well?

Even though the 2011 definition is very explicit in this regard,

Photovoltaic Output Circuit. Circuit conductors between the photovoltaic source circuit(s) and the inverter or dc utilization equipment.

the references to "photovoltaic output circuit" are still misinterpreted by many inspectors to include "inverter output circuits."

I suspect that any request for a formal interpretation on that one would get the response: "It means exactly what it says and nothing more."

 

[c_picard]

Well, here is a very simple thought. Is it possible that they think the Q is whether PV itself - whether load side or supply side tied to the grid - is a new service or not?
As in separately derived power source.
I assume you made it clear that you are only talking about SUPPLY side connected PV AC discos - not AC discos in 95% of the cases that are load side connected (especially residential).

We did get past this hurdle in round one of the bout. I am not proposing that PV is an additional "service". The local AHJ that started this mess did cite Article 100 as his justification for failing the inspection.

 

[c_picard]

I like that "nowithstanding" bit!

 

[GoldDigger]

The local AHJ that started this mess did cite Article 100 as his justification for failing the inspection.

That sure helps, doesn't it.

Perhaps this specific part of Article 100?

Authority Having Jurisdiction (AHJ). An organization, office, or individual responsible for enforcing the requirements of a code or standard, or for approving equipment, materials, an installation, or a procedure.

That means that the AHJ always behaves responsibly, yes?

 

[c_picard]

Maybe I should propose a change to the definition of service to remove the words "delivering electrical energy"

Service. The conductors and equipment for interconnecting the utility to the wiring system(s) of
the premises.

Aaahh, now that's better! Welcome to the 21st century!

 

[GoldDigger]

Maybe I should propose a change to the definition of service to remove the words "delivering electrical energy"

Service. The conductors and equipment for interconnecting the utility to the wiring system(s) of
the premises.

Aaahh, now that's better! Welcome to the 21st century!

With a small change ("from ... to"to "between ... and") you get the same result.
Keeping the "supplying to" in the definition does not settle the matter either to the extent that the conductors to which you connect the tap are still service conductors unless this is purely a PV farm which never draws power from the service.
Given that, the tap wires are now part of a service-connected wiring system (although not service-supplied) even though they may not be actual service wires beyond that point.

Oh well. There has to be a better way of clarifying this that will bring about the result that most people so far think is right and safe.

 

[shortcircuit2]

Maybe I should propose a change to the definition of service to remove the words "delivering electrical energy"

Service. The conductors and equipment for interconnecting the utility to the wiring system(s) of
the premises.

Aaahh, now that's better! Welcome to the 21st century!

A supply side connection does deliver energy to the PV system. Inverters use a small amount of power.

The specs on this PVI 6000 show a Stand-by and Night time energy consumption...

http://www.power-one.com/sites/powe...rgy/datasheet/pvi-5000-6000-tl_data_sheet.pdf