Neutral Calculation for Balanced Inverter per NEC2023

M

MarcialMS

Member
Location
Costa Rica
Occupation
Electrical Engineer
Hi.

How can I design the neutral conductor for a PV inverter according to NEC 2023?

The inverter is a SMA Core 1, 62.5 kW-US. The manufacturer indicates that this inverter has less than 3% of THD.
The phase conductors are going to be 1/0 AWG (because of voltage drop)

Several colleagues indicate that balanced systems do not need the neutral, at all. Others than non lineal systems must be sized 100% of the phase conductors and a few told me to size the neutral as the EGC per 250.122.

This should be ok for 1 inverter, but in my case, I must install 40 Core 1 in a single project. This means 40 x 100%-sized conductors represent a big part of the budget.
I just need the correct Code reference to justify this to the inspector.

Thanks in advance.
 
The installation manual for SUNNY TRIPOWER CORE1-US is here: https://files.sma.de/downloads/STP33-62-US-41-IA-xx-17.pdf

It says that "The inverter is approved for connection to a utility grid with 277/480 V wye connection." (page 19). It further says "The inverter is delivered with a bridge between N and the enclosure as standard. The bridge is absolutely essential if the connection to a utility grid is established without a neutral conductor. If the connection to a utility grid is established with a neutral conductor, the bridge must always be removed as described in the next step." (page 20).

So that makes it sound to me as if there is no need to provide a neutral conductor to the inverter, at least as far as the inverter is concerned. Is there some other reason you have to provide a neutral conductor to each inverter?

Cheers, Wayne

P.S. I would assume based on the above that internally the inverter can be thought of as (3) 480V inverters in a delta configuration. But the technical data (page 38) says that the "rated grid voltage" is 277V. Which makes it sound like (3) 277V inverters in a wye configuration. Which I would think would require a neutral connection, but the inverter clearly doesn't. So I'm slightly confused about this; perhaps viewing it as internally 3 single phase inverters is simply inaccurate.
 
wwhitney said:
I would assume based on the above that internally the inverter can be thought of as (3) 480V inverters in a delta configuration. But the technical data (page 38) says that the "rated grid voltage" is 277V. Which makes it sound like (3) 277V inverters in a wye configuration. Which I would think would require a neutral connection, but the inverter clearly doesn't. So I'm slightly confused about this; perhaps viewing it as internally 3 single phase inverters is simply inaccurate.
Click to expand...
It is balanced, so the neutral current is zero.
 
wwhitney said:
The installation manual for SUNNY TRIPOWER CORE1-US is here: https://files.sma.de/downloads/STP33-62-US-41-IA-xx-17.pdf

It says that "The inverter is approved for connection to a utility grid with 277/480 V wye connection." (page 19). It further says "The inverter is delivered with a bridge between N and the enclosure as standard. The bridge is absolutely essential if the connection to a utility grid is established without a neutral conductor. If the connection to a utility grid is established with a neutral conductor, the bridge must always be removed as described in the next step." (page 20).

So that makes it sound to me as if there is no need to provide a neutral conductor to the inverter, at least as far as the inverter is concerned. Is there some other reason you have to provide a neutral conductor to each inverter?

Cheers, Wayne

P.S. I would assume based on the above that internally the inverter can be thought of as (3) 480V inverters in a delta configuration. But the technical data (page 38) says that the "rated grid voltage" is 277V. Which makes it sound like (3) 277V inverters in a wye configuration. Which I would think would require a neutral connection, but the inverter clearly doesn't. So I'm slightly confused about this; perhaps viewing it as internally 3 single phase inverters is simply inaccurate.
Click to expand...
Yes, I "can" install the neutral.
Yes, the inverter is balanced, therefore 0 Amps on the neutral
Yes, it can be installed as delta.

But, the AHJ is asking us for the neutral. No negotiation on that. My doubt is: how do I size it? at 100%? at 70%? as the EGC??? and the NEC section to it.
 
MarcialMS said:
Yes, I "can" install the neutral.
Yes, the inverter is balanced, therefore 0 Amps on the neutral
Yes, it can be installed as delta.

But, the AHJ is asking us for the neutral. No negotiation on that. My doubt is: how do I size it? at 100%? at 70%? as the EGC??? and the NEC section to it.
Click to expand...
Even though it won't carry current, I would size the neutral to be the same as the EGC, so if using a 100A breaker, it would be 8AWG copper or 6AWG AL.
 
MarcialMS said:
But, the AHJ is asking us for the neutral. No negotiation on that. My doubt is: how do I size it? at 100%? at 70%? as the EGC??? and the NEC section to it.
Click to expand...
Well, the idealized load on the neutral is zero. If you want to consider the harmonic content, you calculate the resulting non-zero current and use that as a more realistic value.

For a feeder, which you have if there is OCPD between the utility and these inverters, the minimum size grounded conductor is given by 215.2(A)(2), which references 250.122 for EGC sizing. So if this size is sufficient for calculated load (as I would expect for any reasonable harmonic content), then that is the minimum size you may install.

For service conductors, the minimum size grounded conductor is specified by 250.24(C)(1).

Cheers, Wayne
 
ggunn said:
It is balanced, so the neutral current is zero.
Click to expand...
For the idealized case, sure. But if there is any imbalance between the L1-N, L2-N, and L3-N voltages, and the internal topology were simply 3 single phase inverters that put out the same current in phase with their line voltage, the currents would not sum to zero. So if the inverter doesn't require a neutral, it must be doing something different, as it can't put any unbalance on its EGC.

But it occurs to me that these hypothetical single phase inverter subunits could communicate with each other and put out the correct currents L-N to ensure that the N currents do add to zero. Conceptually, choose the L1/L2 current ratio so that the sum of the two currents is in phase with L3-N, and then chose the L3 current so they all sum to zero in the neutral.

Anyway, this is all just speculation, the 3 phase inverter topology is probably more optimized than 3 individual single phase inverter subunits.

Cheers, Wayne
 
wwhitney said:
For the idealized case, sure. But if there is any imbalance between the L1-N, L2-N, and L3-N voltages, and the internal topology were simply 3 single phase inverters that put out the same current in phase with their line voltage, the currents would not sum to zero. So if the inverter doesn't require a neutral, it must be doing something different, as it can't put any unbalance on its EGC.

But it occurs to me that these hypothetical single phase inverter subunits could communicate with each other and put out the correct currents L-N to ensure that the N currents do add to zero. Conceptually, choose the L1/L2 current ratio so that the sum of the two currents is in phase with L3-N, and then chose the L3 current so they all sum to zero in the neutral.

Anyway, this is all just speculation, the 3 phase inverter topology is probably more optimized than 3 individual single phase inverter subunits.

Cheers, Wayne
Click to expand...
Leaving out the neutral is following manufacturer's instructions, so I do it and I have for several years with no incidents related to it.
 
solarken said:
Even though it won't carry current, I would size the neutral to be the same as the EGC, so if using a 100A breaker, it would be 8AWG copper or 6AWG AL.
Click to expand...
That's what I would do, following the NEC article for reducing the neutral when it is only used for a voltage reference.
 
ggunn said:
Leaving out the neutral is following manufacturer's instructions, so I do it and I have for several years with no incidents related to it.
Click to expand...
Sure, but the OP indicated that the AHJ wouldn't go for that, for whatever reason.

Anyway, I was just wondering out loud how such an inverter arranges to have zero current on the ground reference when the description as 277V inverter(s) implies it is working L-G rather than L-L.

Cheers, Wayne
 
The 2023 code permits a neutral used "solely for Instrumentation, Voltage Detection, or Phase Detection" to be sized to 250.102. See 705.28(C)(2). Earlier code cycles referenced 250.122. You may need a letter from SMA to get that past this AHJ. I would ask the AHJ to cite chapter and verse on requiring a neutral when the manufacturer says it's not required.
 
wwhitney said:
Sure, but the OP indicated that the AHJ wouldn't go for that, for whatever reason.

Anyway, I was just wondering out loud how such an inverter arranges to have zero current on the ground reference when the description as 277V inverter(s) implies it is working L-G rather than L-L.

Cheers, Wayne
Click to expand...
There is a jumper between the neutral and ground on the inverter as shipped from the factory, that is removed when the neutral is wired to the inverter.
 
electrofelon said:
I would avoid even saying that is an option . Just don't put a neutral on the plans.
Click to expand...
If you are saying he should make the installed system not match the approved planset, in a lot of places that will cause your system to fail its inspection. As it should, I might add.
 
You must log in or register to reply here.
Top