Three phase inverters and counting the neutral as a current carrying conductor

[SunFish]

I'm using a Fronius Primo 15 kW inverter that is a true three-phase 208/120 inverter, three hots and a neutral. I have a document from Fronius that says their inverters only use the neutral for voltage and phase detection and therefore you can apply the provisions of 705.95(B) and use a neutral conductor sized no smaller than the equipment grounding conductor.

My question is since this neutral conductor is only used for voltage and phase detection can I also apply 310.15(B)(5) which says that "A neutral conductor that carries only the unbalanced current from other conductors of the same circuit shall not be required to be counted when applying the provisions of 310.15(B)(3)(a)" (adjustments for total number of current carrying conductors).

I'm trying to run my inverter output conductors through an existing piece of 1" RMC that already has (6) #12 conductors. If I count the neutral conductor as a current carrying conductor this would be ten conductors in a raceway for a 0.5 adjustment factor and therefore I would have to upsize the Inverter output conductors to a #4 which won't fit in the 1" RMC raceway with the (6) #12 conductors (over max fill allowance). The #12 wires are for three lighting and outlet circuits on 15 A breakers so these would still be

If I avoid counting the neutral conductor when applying adjustments in Table 310.15(B)(3)(a) this brings the adjustment factor up to a 0.7 factor and I my #6 wire for my inverter output calcs out and everything is hunky dory. Also, the (6) #12 wires are for three lighting and outlet circuits each on 15 A breakers so these would still be protected with the new adjustment factor for adding the inverter circuit to the raceway.

I feel like 310.15(B)(5) seems mostly clear that I do NOT need to count this neutral in my conductor bundling adjustments. The neutral conductor only carries the unbalanced current from other conductors of the same circuit so it meets the requirements 310.15(B)(5)(a). Its a true three phase output (three lines and a neutral) so 310.15(B)(5)(b) doesn't apply.

The requirements of 310.15(B)(5)(c) is what I'm not totally clear on but I'm pretty sure inverter outputs are linear outputs (not Non-linear) and there shouldn't be any harmonic currents in the neutral.

Can anyone tell me if I have to count my neutral conductor as a current carrying conductor for my conductor bundling adjustments per 310.15(B)(3)(a)?

 

[ggunn]

I'm using a Fronius Primo 15 kW inverter that is a true three-phase 208/120 inverter, three hots and a neutral. I have a document from Fronius that says their inverters only use the neutral for voltage and phase detection and therefore you can apply the provisions of 705.95(B) and use a neutral conductor sized no smaller than the equipment grounding conductor.

My question is since this neutral conductor is only used for voltage and phase detection can I also apply 310.15(B)(5) which says that "A neutral conductor that carries only the unbalanced current from other conductors of the same circuit shall not be required to be counted when applying the provisions of 310.15(B)(3)(a)" (adjustments for total number of current carrying conductors).

I'm trying to run my inverter output conductors through an existing piece of 1" RMC that already has (6) #12 conductors. If I count the neutral conductor as a current carrying conductor this would be ten conductors in a raceway for a 0.5 adjustment factor and therefore I would have to upsize the Inverter output conductors to a #4 which won't fit in the 1" RMC raceway with the (6) #12 conductors (over max fill allowance). The #12 wires are for three lighting and outlet circuits on 15 A breakers so these would still be

If I avoid counting the neutral conductor when applying adjustments in Table 310.15(B)(3)(a) this brings the adjustment factor up to a 0.7 factor and I my #6 wire for my inverter output calcs out and everything is hunky dory. Also, the (6) #12 wires are for three lighting and outlet circuits each on 15 A breakers so these would still be protected with the new adjustment factor for adding the inverter circuit to the raceway.

I feel like 310.15(B)(5) seems mostly clear that I do NOT need to count this neutral in my conductor bundling adjustments. The neutral conductor only carries the unbalanced current from other conductors of the same circuit so it meets the requirements 310.15(B)(5)(a). Its a true three phase output (three lines and a neutral) so 310.15(B)(5)(b) doesn't apply.

The requirements of 310.15(B)(5)(c) is what I'm not totally clear on but I'm pretty sure inverter outputs are linear outputs (not Non-linear) and there shouldn't be any harmonic currents in the neutral.

Can anyone tell me if I have to count my neutral conductor as a current carrying conductor for my conductor bundling adjustments per 310.15(B)(3)(a)?

No, you do not count the neutral.

 

[Carultch]

Can anyone tell me if I have to count my neutral conductor as a current carrying conductor for my conductor bundling adjustments per 310.15(B)(3)(a)?

It is a very rare scenario that you would ever have to count the neutral as a CCC on an ordinary three phase inverter. The total harmonic distortion is usually no greater than 3%, which means that it is not a majority of the load current that is nonlinear. The reason that is important, is that harmonics that are multiples of 3, such as 180 Hz, accumulate onto the neutral, instead of cancel.

Fronius IG Plus three phase inverters had a mix-mode of operation, where the inverter has three power stages to form the phases, which might output dissimilar amounts of current. For low amounts of power, the inverter would selectively turn off power stages, and therefore only use some of the phases. Even for the 277/480V inverters that connected the power stages line-to-neutral, you didn't need to count the neutral as a CCC. You still did need to run a full size neutral, as opposed to an EGC sized neutral, due to this mix-mode.

Check with the manufacturer, if you have your doubts about this. Ideally, they should have a section in the manual that discusses circuit requirements for the neutral, but few (if any) do.

 

[electrofelon]

I just recently was ranting in a previous thread about how (in my experience) everyone uses full size neutrals on three phase inverters. Good to see someone doing the homework and not wasting time and money.

 

[jaggedben]

Not to hijack the thread (too much) I hope...

Suppose you have a neutral that doesn't exactly qualify for 705.95(B) - mere voltage and phase sensing - but is exempt from being a CCC under 310.15(B)(5). E.G. the inverter might put a little unbalanced current on the neutral, but maybe only because the phase voltages are bit off from each other. Is there code that allows a smaller neutral for such a situation? In general, is there code that allows a less than full size neutral for qualifying mutli-wire circuits? What is the point of 215.2(A)(2)? And if you use a common neutral (215.4) how are you required to size it?

Occasionally I've seen a smaller service or feeder neutral and I've wondered what governs the sizing there, or if it's a violation.

 

[electrofelon]

Not to hijack the thread (too much) I hope...

Suppose you have a neutral that doesn't exactly qualify for 705.95(B) - mere voltage and phase sensing - but is exempt from being a CCC under 310.15(B)(5). E.G. the inverter might put a little unbalanced current on the neutral, but maybe only because the phase voltages are bit off from each other. Is there code that allows a smaller neutral for such a situation? In general, is there code that allows a less than full size neutral for qualifying mutli-wire circuits? What is the point of 215.2(A)(2)? And if you use a common neutral (215.4) how are you required to size it?

Occasionally I've seen a smaller service or feeder neutral and I've wondered what governs the sizing there, or if it's a violation.

The neutral is theoretically sized by a load calc, with a minimum (usually EGC or GEC). Of course this is not usually followed much in practice, i.e dwelling service and feeder neutrals are almost always 2 sizes smaller, and three phase services are usually defaulted to full size.

 

[Smart $]

Not to hijack the thread (too much) I hope...

In short, the inverter(s) maximum calculated unbalanced neutral current is equal to the maximum phase current... assuming we are talking grid-tied inverter(s). No allowance for the neutral to be smaller.

 

[electrofelon]

In short, the inverter(s) maximum calculated unbalanced neutral current is equal to the maximum phase current... assuming we are talking grid-tied inverter(s). No allowance for the neutral to be smaller.

Can you elaborate on and justify that position?

 

[Smart $]

Can you elaborate on and justify that position?

Well if this hypothetical inverter's neutral does not qualify for 705.95(B) "instrumentation circuit", then the neutral must be considered an "output circuit" conductor. Do you see anywhere else in Article 705 which permits a reduction in the size of a neutral conductor?

With nothing else in Article 705, we can take the stance that Chapters 1 through 4 apply and resort to 215.2, which in short says conductors must have an ampacity per Article 220. We thus go to 220.61.

 

[electrofelon]

Well if this hypothetical inverter's neutral does not qualify for 705.95(B) "instrumentation circuit", then the neutral must be considered an "output circuit" conductor. Do you see anywhere else in Article 705 which permits a reduction in the size of a neutral conductor?

No, but I dont see anything that prohibits it. I do find 705.95 rather odd in that it states, "......Shall comply with either (A) or (B)." A and B hardly cover all scenarios.

With nothing else in Article 705, we can take the stance that Chapters 1 through 4 apply and resort to 215.2, which in short says conductors must have an ampacity per Article 220. We thus go to 220.61.

I do not think 220.61 would apply to the inverter output circuit.

 

[jaggedben]

...We thus go to 220.61.

Thank you, that's an answer to my question above about code that generally speaks to neutral sizing. Now I have some studying to do. Notably it does not apply to branch circuits.

..

I do not think 220.61 would apply to the inverter output circuit.

I would like to find a basis to disagree with you. Not that I can find anything solid. :lol:

It seems to me that if you have a three-phase inverter with minimal harmonic distortion, as described by Carultch, there ought to be some allowance for the neutral even if it carries some current. However so far as I've seen it does not appear that anything in the code addresses the question.

 

[ggunn]

Thank you, that's an answer to my question above about code that generally speaks to neutral sizing. Now I have some studying to do. Notably it does not apply to branch circuits.



I would like to find a basis to disagree with you. Not that I can find anything solid. :lol:

It seems to me that if you have a three-phase inverter with minimal harmonic distortion, as described by Carultch, there ought to be some allowance for the neutral even if it carries some current. However so far as I've seen it does not appear that anything in the code addresses the question.

Even if there is current on the neutral, it means that there is less current on the phase conductors. Don't count the neutral.

 

[Smart $]

No, but I dont see anything that prohibits it. I do find 705.95 rather odd in that it states, "......Shall comply with either (A) or (B)." A and B hardly cover all scenarios.

I do not think 220.61 would apply to the inverter output circuit.

It is entirely possible to interpret Code as nothing outside of Article 705 applies to interconnect conductors [705.3].

That said, if 705.95 does not apply, then you have to resort to 705.12...

...and you certainly won't get anything there.

 

[jaggedben]

It is entirely possible to interpret Code as nothing outside of Article 705 applies to interconnect conductors [705.3].

That said, if 705.95 does not apply, then you have to resort to 705.12...

...and you certainly won't get anything there.

I'd say look at 705.60, but the point stands. It does seem like we'd need a new section (705.95(C)) to address it. The problem with 220.61 is that everything refers to 'loads'.

 

[Carultch]

It seems to me that if you have a three-phase inverter with minimal harmonic distortion, as described by Carultch, there ought to be some allowance for the neutral even if it carries some current. However so far as I've seen it does not appear that anything in the code addresses the question.

If you had an inverter that could operate with any serious amount of current on the neutral, while the full nominal current would be on all the phases, chances are, it wouldn't even pass UL standards. Very few (if any) POCOS would allow it to be connected to the grid. We're talking a hobbyist's home built inverter, with insufficient filtering to even resemble a sine wave.

As I mentioned, the inverters that do contribute an imbalance, despite being three phase inverters, will drop in current on one of the phases when the neutral current increases. For pure sine wave currents, no matter what the imbalance of currents, the total heat dissipated among the four conductors will never exceed what it is for just three of the conductors with one conductor dormant. So for inverter output circuit applications, you don't count the neutral as a CCC for ampacity adjustment purposes.

 

[electrofelon]

We have installed some larger systems for two different solar developers, and they both used the solectria PVI series and did not reduce the neutral, despite a statement in the manual clearly meeting the spirit of 705.95(B) (one was a 1 meg with copper conductors - ouch). A third developer used the sunny tripowers, and I have not dug through the manual to see what they state.

 

[electrofelon]

Here is the statement in the Solectrica PVI36TL manual:

When interfacing with a Wye grounded transformer winding, a neutral is required.
Since the neutral is used by the inverter for voltage sensing, the neutral does not carry
full load amps. The size of the neutral may be reduced to a conductor no smaller
than
the EGC.

 

[ggunn]

We have installed some larger systems for two different solar developers, and they both used the solectria PVI series and did not reduce the neutral, despite a statement in the manual clearly meeting the spirit of 705.95(B) (one was a 1 meg with copper conductors - ouch). A third developer used the sunny tripowers, and I have not dug through the manual to see what they state.

SMA won't give it to you in writing, although their tech support says it is so, or at least that was what happened when I asked them a year ago or so.

 

[jaggedben]

If you had an inverter that could operate with any serious amount of current on the neutral, while the full nominal current would be on all the phases, chances are, it wouldn't even pass UL standards. Very few (if any) POCOS would allow it to be connected to the grid. We're talking a hobbyist's home built inverter, with insufficient filtering to even resemble a sine wave.

As I mentioned, the inverters that do contribute an imbalance, despite being three phase inverters, will drop in current on one of the phases when the neutral current increases. For pure sine wave currents, no matter what the imbalance of currents, the total heat dissipated among the four conductors will never exceed what it is for just three of the conductors with one conductor dormant. So for inverter output circuit applications, you don't count the neutral as a CCC for ampacity adjustment purposes.

It's almost as if you're saying, contrary to electrofelon's comment above, that 705.95(A) and (B) DO cover all cases. And thus we could invoke (B) on any inverter that doesn't fit under (A).

 

[Smart $]

I'd say look at 705.60, but the point stands. It does seem like we'd need a new section (705.95(C)) to address it. The problem with 220.61 is that everything refers to 'loads'.

Don't need a new section and 'loads' is not a problem in 220.61. You're looking for permission to reduce the size of a neutral which carries a more than insignificant amount of current and does not qualify for 705.95(B). I'm telling you it is not going to happen no matter where (i.e. section), or when (i.e. Code edition), you look.

The reason I say this is because if the inverter is capable of pushing more than insignificant current, then from a physics perspective, it would have to be capable of handling the max current pushed by any one leg of the inverter, same as if it was comprised of three one-phase inverters connected in a wye configuration... and you see what 705.95(A) says about two-wire-connected inverters.

I pointed to 220.61 because it is the only [other] place that I'm aware of in the Code that one can reduce the size of a feeder neutral. However, note the neutral must be sized to handle the maximum calculated unbalanced neutral current. In the case of wye configured inverters, that maximum is the same as the output of any one of the three inverter legs.