Separately Derived Systems and 250.25 & 250.30(C)

[pvgreeze]

Hi all, two questions:

1. Is the working interpretation of 250.25 that the utility side transformer characteristics for a supply-side interconnected PV system must match the characteristics of the existing service?
   • Example - a building has a 2000A 208/120V 3Ph 4W service from the utility. If the PV system on the roof utilizes 480/277V inverters, could a 480/277V - 208V step down transformer be used; that is, the utility side of the transformer is 208V 3Ph 3W delta and does not match the existing service
   • 250.25 does not specifically state that the exact phase and neutral characteristics must match the service, but that they "comply with the requirements of 250.24(A) through (D)." This includes 250.24(C)(3), which includes provisions for a delta-connected service
2. Using the above example, suppose a transformer is installed outside of a building. The POI is inside the building, the PV system is on top of the building. 250.30(C) states that "if the source of the separately derived system is located outside the building or structure supplied, a grounding electrode connection shall be made at the source location to one or more grounding electrodes in compliance with 250.50." Does the equipment grounding conductor (EGC) between the supply-side service disconnect, which is connected to the existing building grounding electrode system (GES) via a properly sized grounding electrode conductor (GEC), satisfy this requirement?
   • This assumes that the primary side (utility side) and secondary side (PV side) grounding pads are electrically connected inside of the transformer by the manufacturer

Be curious to know thoughts. Yes, I know AHJ rules supercede any NEC discussion. Typically I default to utility new service guidelines for any supply side (line side) interconnected PV systems, especially if there is a step down transformer involved.

 

[ggunn]

...a building has a 2000A 208/120V 3Ph 4W service from the utility. If the PV system on the roof utilizes 480/277V inverters, could a 480/277V - 208V step down transformer be used; that is, the utility side of the transformer is 208V 3Ph 3W delta and does not match the existing service

Yes, I have done this many times. The PV transformer should be 208V delta (no neutral) on the utility side and 480/277V wye with a grounded neutral on the inverter side with a kVA rating less than or equal to the maximum kW output of the inverter(s). Such transformers are common off the shelf items for most manufacturers.

 

[wwhitney]

In considering these questions, it is helpful to ignore the fact that the PV system is a source. The transformers involved will all be energized by the utility.

Thus for question (1), 250.25 is only commenting on what you do at the service disconnect and on the utility side of it. Then from the service disconnect, you are free to connect whatever load or interactive source you like. So use a 208V delta : 480Y/277V transformer is fine.

And for question (2), the "source of the SDS" would be the transformer and has nothing to do with the PV location.

Cheers, Wayne

 

[Elect117]

Hi all, two questions:

1. Is the working interpretation of 250.25 that the utility side transformer characteristics for a supply-side interconnected PV system must match the characteristics of the existing service?
   • Example - a building has a 2000A 208/120V 3Ph 4W service from the utility. If the PV system on the roof utilizes 480/277V inverters, could a 480/277V - 208V step down transformer be used; that is, the utility side of the transformer is 208V 3Ph 3W delta and does not match the existing service
   • 250.25 does not specifically state that the exact phase and neutral characteristics must match the service, but that they "comply with the requirements of 250.24(A) through (D)." This includes 250.24(C)(3), which includes provisions for a delta-connected service
2. Using the above example, suppose a transformer is installed outside of a building. The POI is inside the building, the PV system is on top of the building. 250.30(C) states that "if the source of the separately derived system is located outside the building or structure supplied, a grounding electrode connection shall be made at the source location to one or more grounding electrodes in compliance with 250.50." Does the equipment grounding conductor (EGC) between the supply-side service disconnect, which is connected to the existing building grounding electrode system (GES) via a properly sized grounding electrode conductor (GEC), satisfy this requirement?
   • This assumes that the primary side (utility side) and secondary side (PV side) grounding pads are electrically connected inside of the transformer by the manufacturer

Be curious to know thoughts. Yes, I know AHJ rules supercede any NEC discussion. Typically I default to utility new service guidelines for any supply side (line side) interconnected PV systems, especially if there is a step down transformer involved.

For (1), 208 is a wye voltage. 240V is a delta. You might need to confirm the service type. I personally haven't see a utility feed 208V delta. That being said, installing a transformer between them is how your would make the interconnection.

For (2), I will let others respond more thoroughly, but no. You need to ground the transformer effectively. You can probably use a wire sized as a GEC and double the use as both the GEC and EGC but that could be too difficult depending on construction limitations. Like the type of splices or conduit being used. GECs have different rules than EGCs. More than likely you will need to run a separate GEC to the GES to this transformer. And possibly the fused disconnect right before supply side interconnection. You will also need to size your grounding and bonding differently with SSBJs or SBJs rather than just EGCs.

 

[ggunn]

For (1), 208 is a wye voltage. 240V is a delta. You might need to confirm the service type. I personally haven't see a utility feed 208V delta. That being said, installing a transformer between them is how your would make the interconnection.

For a 208/120V wye service he would use a 208V delta primary to 480/277V wye secondary transformer. There should be no connection between the transformer primary and the service neutral.

 

[Elect117]

For a 208/120V wye service he would use a 208V delta primary to 480/277V wye secondary transformer. There should be no connection between the transformer primary and the service neutral.

I thought he was saying the utility was 208V delta. I misunderstood then.

 

[ggunn]

I thought he was saying the utility was 208V delta. I misunderstood then.

He said: a building has a 2000A 208/120V 3Ph 4W service from the utility, i.e. 208/120V wye service. Even if it were 208V delta, it wouldn't change anything with the transformer and inverter connections.

 

[Elect117]

He said: a building has a 2000A 208/120V 3Ph 4W service from the utility, i.e. 208/120V wye service. Even if it were 208V delta, it wouldn't change anything with the transformer and inverter connections.

He also said " This includes 250.24(C)(3), which includes provisions for a delta-connected service". Which is typically used with corner grounded deltas.

I didn't say that it would change anything. Only that it sounded odd to me.

 

[ggunn]

He also said " This includes 250.24(C)(3), which includes provisions for a delta-connected service". Which is typically used with corner grounded deltas.

I didn't say that it would change anything. Only that it sounded odd to me.

Well, the transformer would be delta connected irrespective of the grounding of the service and whether or not there is a neutral; maybe that is what he was referring to.

 

[jim dungar]

Well, the transformer would be delta connected irrespective of the grounding of the service and whether or not there is a neutral; maybe that is what he was referring to.

I agree, he probably meant a delta feeder not a POCO service.

Another case of industry slang getting in the way.

 

[pvgreeze]

Thanl you all for the comments and feedback, appreciate it!

 

[pvgreeze]

I agree, he probably meant a delta feeder not a POCO service.

Another case of industry slang getting in the way.

Just to clarify, what I was trying to get across was that the delta-connected PV disconnect is for all intents and purposes considered a "service disconnect," and not that it is a feeder. I do understand that difference.

Assuming the delta PV disconnect is a "service disconnect" per the rules of the NEC, I was trying to figure out the prevailing theory on whether that also meant that the PV disconnect needs to match the phase, ground, AND neutral characteristics of the existing service. 250.25(A) states that a system on the supply-side of the service disconnect states that such a connection is permitted provided it "comply with the requirements of 250.24(A) through (D)." One of those requirements is 250.24(C)(3), which states that "the grounded conductor of a 3-phase, 3-wire delta service shall have an ampacity not less than that of the ungrounded conductors." Basically, the NEC says we can connect a grounding system on the supply side of a disconnect, and one of those ways that it can be connected in a compliant way (as stated in 250.24(A) through (D)) is via a "Delta-Connected Service" as found in 250.24(C)(3).

 

[jim dungar]

Assuming the delta PV disconnect is a "service disconnect" per the rules of the NEC.

A service can only come from a Utility.
The PV can never be anything other than separately derived.
For all intents these are now treated almost identically, but that has not been true for ever.there still a few holdover locations in the NEC that appear to misuse the term " service" just as there is leftover industry slang.

Yes, 3 wire delta sources, grounded or ungrounded are allowed.

 

[pvgreeze]

A service can only come from a Utility.
The PV can never be anything other than separately derived.
For all intents these are now treated almost identically, but that has not been true for ever.there still a few holdover locations in the NEC that appear to misuse the term " service" just as there is leftover industry slang.

Yes, 3 wire delta sources, grounded or ungrounded are allowed.

2023 NEC 705.11 states differently.

705.11(D) specifically states that "A disconnecting means in accordance with Parts VI through VII of Article 230 shall be provided to disconnect all ungrounded conductors of a power production source form the conductors of other systems." As I'm sure you're aware, Article 230 governs services per the letter of the NEC, and Part VI specifically governs disconnect, which is particularly relevant here. Therefore, if the letter of the NEC calls my PV disconnect a "service disconnect" per 705.11(D) through Part VI of Article 230, then it stands to reason that 250.24 - Grounding of Service-Supplied Alternating-Current Systems is the relevant section of Article 250 of the NEC to invoke for design and review purposes.

I don't disagree that there is an "industry slang" aspect to consider, but when service disconnect per the NEC require special treatment (like connection to the GES via a GEC) then that specific definition becomes very important.

Just to clarify. I'm not trying to debate semantics, but to figure out the intent of the NEC and the broader consensus around some of these fuzzier aspects.

 

[jim dungar]

Therefore, if the letter of the NEC calls my PV disconnect a "service disconnect" per 705.11(D) through Part VI of Article 230....

It is totally acceptable for 705 to require compliance with a few parts of 230, instead of reprinting them, without saying the entireity of 230 must be complied with, which would be the case if your PV was truly a "POCO service".

 

[pv_n00b]

I still think it's funny that for the 20+ years I've been doing PV people have been arguing over if the supply side PV disconnect is/is not a service disconnecting means, and if not should it be installed under the same requirements anyway. The closest the NEC has been to defining this is the 2023 705.11(D) which calls the PV disconnect a service disconnecting means and requires that it be installed as a service disconnecting means. And that was only put in the 2023 NEC to stave off a code war with the CMP that covers 230 who wanted the disconnect to be considered a service disconnecting means and was willing to act on it by adding language to 230. But now that war has been averted the 2026 705.11 will remove the language. Massive changes in the code from cycle to cycle, always good for safety.
Grounding of supply side PV disconnects was defined in 2020 by the addition of 250.25. That was undefined for way too long.

 

[jaggedben]

...
The PV can never be anything other than separately derived.
...

That is incorrect. Almost all PV systems these days are not separately derived systems, any more than a VFD is a separately derived system. (If a PV system is connected through a separate transformer, then the transformer creates an SDS, but not the PV system.)

 

[don_resqcapt19]

But now that war has been averted the 2026 705.11 will remove the language.

If you read the panel statement for FR 8568 there is no real change. The reference to service disconnect was removed because that is already covered in Article 230 and repetition of requirements from Chapters 1-4 is not permitted by the style manual. It is still a service disconnect and rightly so, as the main hazard is from the connection to the electric utility and not from the local power production source.
However the changes do provide some relief from the "nearest the point of entry" restriction for the location of the service disconnect, permitting more unprotected service conductors within in a building than the rules in 230 permit.
Panel statement:

... 705.11(D) – The language has been removed as the requirements of Chapter 2 are applicable unless modified in accordance with 90.3. The requirements of and for a service disconnecting means are already referenced through 230.2(A)(5) and 230.82(6).

 

[jaggedben]

...
Assuming the delta PV disconnect is a "service disconnect" per the rules of the NEC, I was trying to figure out the prevailing theory on whether that also meant that the PV disconnect needs to match the phase, ground, AND neutral characteristics of the existing service. 250.25(A) states that a system on the supply-side of the service disconnect states that such a connection is permitted provided it "comply with the requirements of 250.24(A) through (D)." One of those requirements is 250.24(C)(3), which states that "the grounded conductor of a 3-phase, 3-wire delta service shall have an ampacity not less than that of the ungrounded conductors."

In my opinion this has always been a correct way to think about it. You might be able to find old threads on this forum with knock-down-drag-out fights about whether there is another way to do it, like by bringing an SSBJ from another spot. See pv_noob's post for more on that. But treating it as a service disconnect always had code backing.

You keep asking if multiple service disconnects need to match phase ground and neutral. I think the answer is definitely yes for grounded conductor and neutral, because that is a physical characteristic of the utility source supplying the service. As you identified, 250.24(C) requires bringing the grounded conductor to each service disconnect. There may be a partial exception for phase, but I can only think of one example where it would be reasonable, which is that on a high leg delta you could have a service disconnect that includes the high leg and another one that does not. (With respect to PV, get utility approval before interconnecting to a high-leg).

You cannot change characteristics like phase angle and voltage that are determined by physics: I hope that's clear to you. Similarly I hope it's obvious that you cannot ground different conductors at different service disconnects without creating a direct short circuit through the grounding system or equipment (or at least an unwanted current through earth).

Basically, the NEC says we can connect a grounding system on the supply side of a disconnect, and one of those ways that it can be connected in a compliant way (as stated in 250.24(A) through (D)) is via a "Delta-Connected Service" as found in 250.24(C)(3).

"Grounding system" conflates with other issues. See 250.64(D) for the options of grounding multiple service disconnects. See above comment regarding how you must always ground the same conductor on a given service. See installing a transformer if all that doesn't give you what you want.

 

[ggunn]

That is incorrect. Almost all PV systems these days are not separately derived systems, any more than a VFD is a separately derived system. (If a PV system is connected through a separate transformer, then the transformer creates an SDS, but not the PV system.)

I had that thought as well, but I think he is specifically talking about the system the OP was discussing, which I believe is connected through a transformer. If he was speaking generally, then yes, he is mistaken.