Separately Derived Neutral Required?

[beanland]

Is there an article in the NEC that requires the installation of a grounded conductor (neutral) on the secondary of a grounded-wye transformer? If I install a step-down transformer with a grounded-wye secondary I am required to provide a equipment grounding conductor and grounding electrode conductor at the transformer neutral. Is there an NEC article that requires that I also install a grounded conductor (neutral)? There is an NEC article (230.42(C)) that requires a 250.66 sized grounded conductor for a service-derived source but I have not found the citation for the separately derived source. 250.30(A)(3) says "If a grounded conductor is installed..." which implies that it is optional.

Here is the dilemma: a PV system with grounded-wye transformer is connected to a service-derived grounded-wye system with phase-neutral load. The PV installer says that no neutral is required at their transformer. However, when operating in parallel, the PV system could supply unbalanced loads in the distribution system, leading to "neutral" current flow on the EGC because of the multiple neutral-ground bonds.

However, a neutral-ground bond at the transformer would pose a problem because there will now be multiple neutral-ground bonds in the system.

705.95 is not much help either because it does not cover 3-phase systems.

Is a neutral required? Is a neutral-ground bond required?

 

[david luchini]

Is there an article in the NEC that requires the installation of a grounded conductor (neutral) on the secondary of a grounded-wye transformer?

I don't believe that there is any code requirement be install a grounded conductor in a separately derived system.

If I install a step-down transformer with a grounded-wye secondary I am required to provide a equipment grounding conductor and grounding electrode conductor at the transformer neutral.

If the separately derived system is a grounded system, then you will need a grounding electrode conductor and a system bonding jumper, and possibly an equipment bonding jumper. These should be installed per 250.30(A).

 

[skeshesh]

If the separately derived system is a grounded system, then you will need a grounding electrode conductor and a system bonding jumper, and possibly an equipment bonding jumper. These should be installed per 250.30(A).

David, I think the question he's asking is if this is in fact a seperately derived system. In his case he has a PV that will work parallel with the utility service and he's being cautious about creating a parallel ground path. I think further detail about the distribution system is necessary. If the OP can post a shorthand diagram maybe we can look into it in more detail.

 

[beanland]

No Pix

No Pix

Sorry, I sketched a diagram but am unable to upload.

To an extent, you have it correct. This is two grounded sources of power running in parallel. The output of the PV system (480Y/277V) is grounded but the PV installer has said that no neutral is required. I am concerned that this source is no different from a "service derived" source and requires a neutral. However, it is not separate because the phase conductors are continuous from service to PV transformer. If we think of the transformer (wye-wye) as a load, a neutral is required because there can be unbalanced loading. But, if we consider this a source, then what?

 

[Smart $]

Sorry, I sketched a diagram but am unable to upload.

To an extent, you have it correct. This is two grounded sources of power running in parallel. The output of the PV system (480Y/277V) is grounded but the PV installer has said that no neutral is required. I am concerned that this source is no different from a "service derived" source and requires a neutral. However, it is not separate because the phase conductors are continuous from service to PV transformer. If we think of the transformer (wye-wye) as a load, a neutral is required because there can be unbalanced loading. But, if we consider this a source, then what?

The secondary of the PV system is an alternate power source, not a separately derived source. There is (or should be) no neutral-ground bond at the PV transformer when directly connected to service-supplied system.

 

[skeshesh]

The secondary of the PV system is an alternate power source, not a separately derived source. There is (or should be) no neutral-ground bond at the PV transformer when directly connected to service-supplied system.

I agree, but is that to say that there should be no neutral either? I'm thinking along the lines of both normal and alternative power source having a neutral but having the NGB at one point (service entrance of primary source?) in the system.

 

[augie47]

In direct response to: "Is there an article in the NEC that requires the installation of a grounded conductor (neutral) on the secondary of a grounded-wye transformer? ", I think 250.20 would be the Article {in your case 250.20(B)}

 

[beanland]

Multiple Sources

Multiple Sources

I agree that the PV transformer is not a separately-derived system thus it is an interconnected electric power production source.

And, based on this I should not bond the transformer neutral (if used) to ground. But, what if a transformer neutral is NOT installed?

If the manufacturer says that all PV inverter output is balanced so there is never any neutral current and it complies with 705.100(B). In this case, where no neutral is required, do I ground the wye point? It is not a bond because there is no neutral. In which case any triplen harmonics will flow on the ground.

I would feel better if the neutral is installed (even if it never intentionally carries current) and the enclosure is bonded for fault return but no neutral-ground bond is installed.

Thoughts?

 

[augie47]

My opinion is, partly based on 250.20 but mostly based on the fact that the PV system transformer could see neutral to phase loads, that you should have a neutral connection to the transformer and the transformer should have a EGC connected to the system, but there should be no neutral bond at the transformer as it is not a SDS and such connection would be prohibited by 250.24(A)(5)

 

[Smart $]

I agree that the PV transformer is not a separately-derived system thus it is an interconnected electric power production source.

And, based on this I should not bond the transformer neutral (if used) to ground. But, what if a transformer neutral is NOT installed?

If the manufacturer says that all PV inverter output is balanced so there is never any neutral current and it complies with 705.100(B). In this case, where no neutral is required, do I ground the wye point? It is not a bond because there is no neutral. In which case any triplen harmonics will flow on the ground.

Doesn't matter whether a neutral conductor is installed or not, the PV transformer wye point should not be bonded to ground. The transformer secondary is part of the service-supplied system, which is to have only one main bonding jumper.

I would feel better if the neutral is installed (even if it never intentionally carries current) and the enclosure is bonded for fault return but no neutral-ground bond is installed.

I don't see how a manufacturer can claim the service system side of the transformer will never see neutral current. The PV system itself may not see any. It and the transformer primary are probably delta configured... no neutral. Can't have neutral current if there isn't any neutral conductor. But that's on the PV side...

I would run a neutral to the secondary (service side).

 

[david luchini]

I agree that the PV transformer is not a separately-derived system thus it is an interconnected electric power production source.

I don't see that "separately derived system" and "interconnected electric power production source" are mutually exclusive. If you have an isolating transformer between the PV supply and the service supply then by definition, you have a separately derived system.

And, based on this I should not bond the transformer neutral (if used) to ground. But, what if a transformer neutral is NOT installed?

If the manufacturer says that all PV inverter output is balanced so there is never any neutral current and it complies with 705.100(B). In this case, where no neutral is required, do I ground the wye point? It is not a bond because there is no neutral. In which case any triplen harmonics will flow on the ground.

I would feel better if the neutral is installed (even if it never intentionally carries current) and the enclosure is bonded for fault return but no neutral-ground bond is installed.

Thoughts?

Your question seems to be whether you would need a neutral to ground bound at the transformer since you have a separately derived system. The answer is no. 250.30 requires a system bonding jumper for an SDS anywhere between the source and the first disconnecting means. The source in your case is NOT the transformer, but the PV equipment. And you have already said that the PV system is grounded, so your SDS is grounded per 250.30 without a neutral-ground bond at the transformer.