Equipment grounding conductor question

[ggunn]

Here's the scenario: A client wants to build a ground mounted PV system that is ~1200' from the point of interconnection and for Vd concerns the CCCs are to be sized up considerably. 250.122 says that the ECG needs to be sized up at the same ratio as the CCCs unless the upsizing is due to ambient temp or conduit fill ampacity derating, but 690.45 says that for PV system circuits "increases in equipment grounding conductor size to address voltage drop considerations shall not be required". Is an inverter output circuit not a "PV system circuit"? The previous articles 690.41 and 690.42 explicitly refer to DC circuits, but 690.43 does not; it is not a subset of either of these and there is nothing ahead of these articles that says that they are all governing DC circuits - Part V just says "Grounding and Bonding". Am I missing something?

To be clear, I can't see any reason why clearing a fault on a PV inverter output circuit would be any different from clearing one on a load circuit, and I certainly would not want to take advantage of a loophole or an oversight in the NEC that could make a system unsafe, but what does the code actually say?

 

[jaggedben]

690.45 is not intended to apply to inverter output circuits, whatever it may say. It is intended only to apply to PV circuits where groundfault detection is not done with a circuit overcurrent device.

It would be rather dangerous to apply 690.45 to a 1200' AC circuit so I'm not even going to go there as far as the 'what does the code actually say' part.

 

[ggunn]

690.45 is not intended to apply to inverter output circuits, whatever it may say. It is intended only to apply to PV circuits where groundfault detection is not done with a circuit overcurrent device.

It would be rather dangerous to apply 690.45 to a 1200' AC circuit so I'm not even going to go there as far as the 'what does the code actually say' part.

Again, to be clear, I am only asking what the code actually says. As I said, I can't see where clearing a fault on an inverter output circuit would be any different from doing so on a load circuit. But, as I also said, it only refers to "PV system circuits" which seems to me could be applied to inverter output circuits because it doesn't say otherwise.

For the record, I will be increasing the EGC size proportionately to the CCC upsize when I give the numbers to my client.

 

[wwhitney]

Yes, per the definition the PV system extends from the PV equipment up to (towards the utility/loads) the PV system disconnecting means. So an inverter output circuit between an inverter and the PV system disconnecting means is a PV system circuit.

But this appears to be an error in 690.45. As per the previous comments, the last sentence should not apply to all PV system circuits, just PV system DC circuits.

Cheers, Wayne

 

[jaggedben]

Yes, per the definition the PV system extends from the PV equipment up to (towards the utility/loads) the PV system disconnecting means. So an inverter output circuit between an inverter and the PV system disconnecting means is a PV system circuit.

But this appears to be an error in 690.45. As per the previous comments, the last sentence should not apply to all PV system circuits, just PV system DC circuits.

Cheers, Wayne

Actually there are definitions of "PV System" and "PV DC Circuit" and the latter is 110% what 690.45 is supposed to say. But there is no definition of "PV system circuit" which are the words 690.45 has contained since the 2020 cycle.

2017 said "PV source and output circuits" which were clearly defined as DC. So it is simply another copy editing mistake by the NFPA.

 

[wwhitney]

Actually there are definitions of "PV System" and "PV DC Circuit" and the latter is 110% what 690.45 is supposed to say. But there is no definition of "PV system circuit"

Agreed. That makes the obvious meaning of "PV System Circuit" simply "circuit of the PV System." Which is what my previous post was based on.

"PV System Circuit" is a term used 8 times in 2023 NEC Article 690. For example as the heading of 690.8(A)(1), which has subheadings (a) Photovoltaic Source Circuit Currents, (b) PV DC-to-DC Converter Circuit Current, and (c) Inverter Output Circuit Current. Which is accordance with the idea that the inverter output circuit is part of the PV system and hence is a PV system circuit.

So the error in the NEC is just using the wrong term in 690.45, not that the term is ambiguous.

Cheers, Wayne

 

[electrofelon]

We had a similar discussion about what a pv system circuit is and whether inverter output circuit needs RSD.

 

[jaggedben]

Sigh.

Maybe I will put in some PIs on this for next time (to go into effect in my area 7 years from now?!). Things used to be clear. It's wild how such basic stuff gets messed up.

Generally speaking there is no reason that anything in 690 should apply to inverter outputs IMO. It was a long road to get most of those rules to refer to other articles and now with these ambiguities it's slipping back the other way.

I think at least the RSD requirements are clear because they explicitly apply to inverter output circuits where necessary. These other instances are a different issue.

 

[pv_n00b]

NEC 690.45 was intentionally changed in the 2020 NEC to apply to both AC and DC circuits, the way it reads is not a typo or mistake. It's been this way in the 2020, 2023, and will be the same thing in the 2026.
"PV System" is defined in 100 to be, "The total components, circuits, and equipment up to and including the PV system disconnecting means that, in combination, convert solar energy into electric energy." So the PV System Circuit will be all the circuits in the PV System.

 

[ggunn]

NEC 690.45 was intentionally changed in the 2020 NEC to apply to both AC and DC circuits, the way it reads is not a typo or mistake. It's been this way in the 2020, 2023, and will be the same thing in the 2026.
"PV System" is defined in 100 to be, "The total components, circuits, and equipment up to and including the PV system disconnecting means that, in combination, convert solar energy into electric energy." So the PV System Circuit will be all the circuits in the PV System.

Well, then, I will say again that I see no reason why fault clearing on a PV inverter output circuit could be done with a smaller EGC than would a load circuit when the CCCs are increased for Vd. Can you enlighten me? My client would be happy to hear it.

 

[pv_n00b]

Well, then, I will say again that I see no reason why fault clearing on a PV inverter output circuit could be done with a smaller EGC than would a load circuit when the CCCs are increased for Vd. Can you enlighten me? My client would be happy to hear it.

You are expecting the NEC to make sense? ;-) I had this same question and I posed it to a CMP 4 member. Strap in.

• There are already several exemptions to the wire EGC size increase in 250.122(B), namely for increases to current carrying (CC) conductors due to conditions of use. Adding increases to the CC conductors for voltage drop does not seem out of line to the conditions of use exemptions.
• There is a blanket exception to 250.122(B) allowing that the wire EGC can be sized by a qualified person to any size that provides an effective fault current path. With this exception, in theory, all the sizing of the wire type EGC is permissive and not prescriptive.
• The upsizing only applies to wire type EGCs. EGCs can be a lot of things that are not wire that are not sized based on the CC conductor size, so does size really matter all that much once you cross over a minimum cross section of area?

One of my questions was, if this applies to PV systems shouldn't it apply to all AC systems? Should we not get this change added to 250.122 as an exemption? I was told that while it should apply to all AC circuits getting it through the CMP for 250 would be much more difficult than adding it to 690. That makes sense to me, so it's in 690 and only for PV. But it could just as well apply to all AC circuits.

 

[ggunn]

You are expecting the NEC to make sense? ;-) I had this same question and I posed it to a CMP 4 member. Strap in.

• There are already several exemptions to the wire EGC size increase in 250.122(B), namely for increases to current carrying (CC) conductors due to conditions of use. Adding increases to the CC conductors for voltage drop does not seem out of line to the conditions of use exemptions.
• There is a blanket exception to 250.122(B) allowing that the wire EGC can be sized by a qualified person to any size that provides an effective fault current path. With this exception, in theory, all the sizing of the wire type EGC is permissive and not prescriptive.
• The upsizing only applies to wire type EGCs. EGCs can be a lot of things that are not wire that are not sized based on the CC conductor size, so does size really matter all that much once you cross over a minimum cross section of area?

One of my questions was, if this applies to PV systems shouldn't it apply to all AC systems? Should we not get this change added to 250.122 as an exemption? I was told that while it should apply to all AC circuits getting it through the CMP for 250 would be much more difficult than adding it to 690. That makes sense to me, so it's in 690 and only for PV. But it could just as well apply to all AC circuits.

In my case in point, it's a ~1200 UG run in PVC conduit with (of course) a wired ECG and the CCCs upsized from #2/0 to 1000kcmil aluminum (equivalent) for Vd, and I will be the PE of record. I am not comfortable with keeping the EGC at #6 copper. Would you be, and if so, can you show me why?

 

[Pinnie]

I’m not at all a code expert but doesn’t chapters 5-7 modify 1-4. So if the pv code articles (of which I know nothing) don’t modify the voluntary upsizing of CCC’s rule you referenced then it should still apply I would think. You can use the gemi software to calculate how many amps a certain size egc would draw with different variables (conduit, wire type, distance, etc)

 

[jaggedben]

NEC 690.45 was intentionally changed in the 2020 NEC to apply to both AC and DC circuits, the way it reads is not a typo or mistake. It's been this way in the 2020, 2023, and will be the same thing in the 2026.
"PV System" is defined in 100 to be, "The total components, circuits, and equipment up to and including the PV system disconnecting means that, in combination, convert solar energy into electric energy." So the PV System Circuit will be all the circuits in the PV System.

That makes zero sense to me. Fault carrying on PV circuits is categorically different than on grid-connected AC circuits. There's no reason for a long inverter output circuit to get an exemption for voltage drop when other AC circuits don't. In fact, as I've been arguing in threads about 250.122(B), voltage drop is the only justification for upsizing the EGC and the code should say so instead of having a blanket requirement.

 

[jaggedben]

I’m not at all a code expert but doesn’t chapters 5-7 modify 1-4. So if the pv code articles (of which I know nothing) don’t modify the voluntary upsizing of CCC’s rule you referenced then it should still apply I would think. ...

That is correct. While we may argue about whether 690.45 applies to inverter output circuits, there's no debate that it modifies the rules in 250 for the circuits it applies to.

 

[wwhitney]

That makes zero sense to me.

Yeah, basically the CMP behind Article 690 doesn't like the conclusion that the CMP behind Article 250 reached, so they decided to eliminate that requirement within their purview.

While we may argue about whether 690.45 applies to inverter output circuits

It does, as per the wording, and as per the committee statement behind the 2020 NEC change. The statement accompanying the adopted Second Draft change was "This revision simplifies the sizing of grounding conductors and allows ac equipment grounding conductors to remain the same based on the overcurrent device rating regardless of size increases for performance related reasons.

Cheers, Wayne

 

[wwhitney]

In my case in point, it's a ~1200 UG run in PVC conduit with (of course) a wired ECG and the CCCs upsized from #2/0 to 1000kcmil aluminum (equivalent) for Vd, and I will be the PE of record. I am not comfortable with keeping the EGC at #6 copper.

1000 kcmil Al has a 75C DC resistance (chosen for simplicity) from Chapter 9 Table 8 of 21.2 milliohms per kft, so 1200' is 25 milliohms. While #6 Cu uncoated has a resistance of 491 milliohms per kft, so 1200' is 590 milliohms. A fault at the end of line would have a conductor resistance of 615 milliohms. Since you mentioned #2/0 Al as the base size, the OCPD is going to be 125A or 150A (the latter would require a junction box at the utility end to be able to use the 90C ampacity).

So best case, if the utility end of the run has an infinite AFC, and the fault at the end of line is bolted with 0 resistance, and the voltage is say 277V to ground, you'd get a current of 277 / 0.615 = 450A, or 3.6x a 125A OCPD rating. A random breaker trip curve I checked shows a trip time at 40C ambient of 30-90 seconds; at 25C ambient, it could be up to 175 seconds. So if it's cold outside, possibly much longer.

How long is acceptable to clear a fault? Is there a rule of thumb for a minimum ratio of AFC to OCPD rating? Is/should such a rule of thumb be temperature dependent?

Cheers, Wayne

 

[ggunn]

1000 kcmil Al has a 75C DC resistance (chosen for simplicity) from Chapter 9 Table 8 of 21.2 milliohms per kft, so 1200' is 25 milliohms. While #6 Cu uncoated has a resistance of 491 milliohms per kft, so 1200' is 590 milliohms. A fault at the end of line would have a conductor resistance of 615 milliohms. Since you mentioned #2/0 Al as the base size, the OCPD is going to be 125A or 150A (the latter would require a junction box at the utility end to be able to use the 90C ampacity).

So best case, if the utility end of the run has an infinite AFC, and the fault at the end of line is bolted with 0 resistance, and the voltage is say 277V to ground, you'd get a current of 277 / 0.615 = 450A, or 3.6x a 125A OCPD rating. A random breaker trip curve I checked shows a trip time at 40C ambient of 30-90 seconds; at 25C ambient, it could be up to 175 seconds. So if it's cold outside, possibly much longer.

How long is acceptable to clear a fault? Is there a rule of thumb for a minimum ratio of AFC to OCPD rating? Is/should such a rule of thumb be temperature dependent?

Cheers, Wayne

So accepting all this, with an OCPD of 175A and an AL wire size of 4/0 (sorry, I mistyped earlier), would you be comfortable sealing the design with a #6 copper EGC?

 

[wwhitney]

So accepting all this, with an OCPD of 175A and an AL wire size of 4/0 (sorry, I mistyped earlier), would you be comfortable sealing the design with a #6 copper EGC?

I don't have the experience to answer that. With a 175A OCPD, the best case fault current is now only 2.5x the OCPD rating (and that assumes 277V L-G).

Seems low. Which is why I was asking if there's a rule of thumb. E.g. if 5x is enough, you could bump up the EGC to #2 Cu (based on 3 AWG sizes = twice the area = half the DC resistance).

Cheers, Wayne

 

[Pinnie]

How long is acceptable to clear a fault? Is there a rule of thumb for a minimum ratio of AFC to OCPD rating? Is/should such a rule of thumb be temperature dependent?

Cheers, Wayne

Tom baker said this in a related thread

250.122(b) sizing ground

Its only a issue with very long runs, say large parking lot lights or long path lighting If you play around with the numbers you could see situations where the available fault current is not enough to trip a breaker or it would take hours for either the breaker to trip or the weakest link in the...

forums.mikeholt.com

“Trip times of 5X are based on the insulated cable mfgs assn data, to limit overheating of ungrounded conductors. This is explained in the IEEE green book.”

So I’m assuming it’s conductor manufacturer specs?