Strategy for sizing the neutral wire in a PV system

[carultch]

When you use a 120V appliance in north america ( "split phase" household system. )

Maybe I should be more specific. I'm not talking about the obvious case of a 120V appliance.

When does the neutral in a circuit dedicated to a 240V split phase PV inverter ever carry current?
Specifically, current that is significant compared to the ungrounded wires.

[inetdog]

Single phase inverters on a 277/480V wye system are connected line to neutral. The aggergate of a group of them might not perform uniformly. It could be that all the A-phase inverters are at full power, while the B and C phase inverters are sourced by a shaded array. And thus the neutral should be expected to carry a significant amount of current.

I respectfully disagree. There are string inverters and microinverters which are designed to be connected line to line.

For a typical commercial or less commonly residential situation (called 208Y/120) the 3-phase configuration of an inverter can use a 208V output instead of a 240V output. Not a big change and easy to accommodate in a single circuit package.

Enphase micro inverters, I believe, can be configured for 240V line to line output or 208V line to line output just by connecting the correct trunk wiring harness and possibly setting an option on the inverter.

[foo1bar]

Maybe I should be more specific. I'm not talking about the obvious case of a 120V appliance.

When does the neutral in a circuit dedicated to a 240V split phase PV inverter ever carry current?
Specifically, current that is significant compared to the ungrounded wires.

My thought is that with a grid-tie system the neutral never really has current - because the grid-tie makes it such a large current sink that whether there's 200A or 20A on each leg, the neutral will have no current. And any failure mechanism that I can think of would result in the inverter shutting down before/as-soon-as there's any real current on the neutral.

So it seems reasonable that you could have it sized smaller.
BUT I'd still
1> ask the manufacturer (because a> warranty. b> they should have the expertise to give a good answer c> their answer should persuade the AHJ)
2> make sure the AHJ was OK with it.


And for most of us (who aren't dealing with large 3/0 wire) it's easier to just say "screw it - I'm going with same size for all 3."

[foo1bar]

I respectfully disagree. There are string inverters and microinverters which are designed to be connected line to line.

For a typical commercial or less commonly residential situation (called 208Y/120) the 3-phase configuration of an inverter can use a 208V output instead of a 240V output. Not a big change and easy to accommodate in a single circuit package.

Enphase micro inverters, I believe, can be configured for 240V line to line output or 208V line to line output just by connecting the correct trunk wiring harness and possibly setting an option on the inverter.

I don't think you're actually disagreeing with him.

[Sunking]

Remember these two Golden Rules.

1. The Inspector is always right.
2. When Inspector is wrong refer to rule 1

If all 3 conductors are the same size makes Inpector happy and will not ask questions. If the grounded circuit conductor is smaller than phase conductors Inspector will ask questions and will make you prove it. Which route do you want to go? Th eone with least resistance, or your way?

[inetdog]

I don't think you're actually disagreeing with him.

I did not look hard enough at the 480Y/277 qualification.
But as long as PV arrays with voltages up to 1000V are allowed, there is no reason not to make an inverter for 480V line to line operation. Maybe that design is not currently popular.

[Sunking]

I did not look hard enough at the 480Y/277 qualification.
But as long as PV arrays with voltages up to 1000V are allowed, there is no reason not to make an inverter for 480V line to line operation. Maybe that design is not currently popular.

They do for commercial operations. The one Wally World system I worked on 5 years ago uses 3 150Kw inverters configured 277/Wye.

[carultch]

I did not look hard enough at the 480Y/277 qualification.
But as long as PV arrays with voltages up to 1000V are allowed, there is no reason not to make an inverter for 480V line to line operation. Maybe that design is not currently popular.

That is just what they do, in order to use the same transformer in a string inverter. There is a jumper that you configure, in order to land on the output taps of the transformer for your grid type and voltage. It either connects you to 208V phase to phase, 240V phase-to-phase, or 277V phase-to-neutral. They also have 230V phase-to-neutral for rest of world (ROW) applications.

If they used phase-to-phase voltage, that would mean that the 480V system inverters would have significantly different internal transformers, than their 208V counterparts.

I like working with the 277V phase-to-neutral applications. A lot fewer wires, an obvious purpose for that neutral, and you can fit a lot more inverter breakers in your dedicated panelboard.

Plus, you can combine three of these inverters, just by using a fused disconnect. Need to service one inverter for only an hour, open the switch and shut off all three. Need to service one for an extended time, you simply pop its fuse and power the others.

Only the neutrals really get combined, in this simplified AC combiner. And they add up to zero, when all perform the same.

[carultch]

Remember these two Golden Rules.

1. The Inspector is always right.
2. When Inspector is wrong refer to rule 1

If all 3 conductors are the same size makes Inpector happy and will not ask questions. If the grounded circuit conductor is smaller than phase conductors Inspector will ask questions and will make you prove it. Which route do you want to go? Th eone with least resistance, or your way?

I'd like to understand the rules, have confidence that my way is correct, and be prepared with exactly the strategy I'd need to prove it (if challenged).

A beginner engineer rounds up out of ignorance to specify an overbuilt system, to avoid being challenged on confusing rules.
A good engineer knows and understands the rules, designs to meet the rules, and has confidence in specifying a code-compliant functional system.

Putting all that copper or aluminum in the neutral conductor, where it isn't ever going to experience even a tenth or even a hundredth of its capacity, sounds a lot like like rounding up out of ignorance. There's gotta be a procedure somewhere for figuring real worst-case scenario neutral currents on PV systems.

[inetdog]

I'd like to understand the rules, have confidence that my way is correct, and be prepared with exactly the strategy I'd need to prove it (if challenged).

A beginner engineer rounds up out of ignorance to specify an overbuilt system, to avoid being challenged on confusing rules.
A good engineer knows and understands the rules, designs to meet the rules, and has confidence in specifying a code-compliant functional system.

Putting all that copper or aluminum in the neutral conductor, where it isn't ever going to experience even a tenth or even a hundredth of its capacity, sounds a lot like like rounding up out of ignorance. There's gotta be a procedure somewhere for figuring real worst-case scenario neutral currents on PV systems.

There is. Figure at most an amp for a neutral which is only used for voltage comparison.
Figure the unbalanced current in the neutral from a worst case imbalance in the inverter outputs, namely zero.
Now that you know the worst case neutral current, try to convince the inspector that you can size for that. You may be able to get him as low as 15A or 20A.
And you cannot really use smaller than #14 anyway.

[carultch]

There is. Figure at most an amp for a neutral which is only used for voltage comparison.
Figure the unbalanced current in the neutral from a worst case imbalance in the inverter outputs, namely zero.
Now that you know the worst case neutral current, try to convince the inspector that you can size for that. You may be able to get him as low as 15A or 20A.
And you cannot really use smaller than #14 anyway.

Hahaha! Try to convince the inspector that this wire at most sees a single ampere, and convince him that a #14 can carry it. This sounds like trying to convince a building inspector that a house's living room doesn't need seatbelts or airbags.

The solution I was thinking I might be able to justify would be a 200A circuit, with 500kcmil phase wires for curtailing voltage drop, a #3/0 neutral, and a #1 ground. That way if for some obscure reason, the neutral does see 160 Amperes continuous, it is still safe. It might trip the inverter internal relays, but I would want it to do that anyway if it got in to this situation.

I think if you figure just a single ampere as the maximum neutral current, the "not smaller than the equipment grounding conductor" rule would govern it. And that rule makes sense to me. You wouldn't want a phase-to-neutral fault to burn up the neutral wire, before it clears the breaker.

I've seen in Bipolar systems, the "combined DC neutral" once the negative ground and positive ground are joined, can in some cases be as small as #16 wire. I don't work with bipolar systems that much, so I don't know its real term.

[adoublee]

I'm surprised this thread had no discussion of NEC 705.95(A) and NEC 705.95(B). Article 705 is for interconnected electric power production sources (grid-tie PV inverters apply). These sections basically says if the neutral is only used for instrumentation, voltage detection, or phase detection (isn't a natural path for all phase current to return) that it can be sized equal to a EGC (NEC Table 250.122).

If a grid-tie inverter is 240V and neutral that would normally be expected to provide service to 120V connection is actually only a voltage detection component to meet IEEE standards (a high impedance path back to the inverter), almost all neutral current will actually be returning to the utility transformer.

[JFinch57]

I installed a dedicated 200A subpanel to my ground array and sized it for future expansion with the largest wire by code that I could get through an existing 1 1/2" PVC conduit that I installed when my geothermal went in. In hindsight I should have installed 2", but with much lube it pulled OK. It turned out that I got 2 #3/0, 1 #4 neutral (smallest size allowed on the lug), and 1#6 ground, which allowed 38KW AC continuous. Submitted the permit application with line diagram to the county, noting that there was a balanced load with essentially no neutral current, and they accepted it.

[SunEagle]

I installed a dedicated 200A subpanel to my ground array and sized it for future expansion with the largest wire by code that I could get through an existing 1 1/2" PVC conduit that I installed when my geothermal went in. In hindsight I should have installed 2", but with much lube it pulled OK. It turned out that I got 2 #3/0, 1 #4 neutral (smallest size allowed on the lug), and 1#6 ground, which allowed 38KW AC continuous. Submitted the permit application with line diagram to the county, noting that there was a balanced load with essentially no neutral current, and they accepted it.

Usually on a 120/240v system the Neutral wire should be at least the same size as the phase wires. IMO that #4 neutral is not big enough based on the #3/0 phase wires.

Remember the Neutral wire can carries the same current as the phase wires on the trip back from the load if there is ever an unbalance load.

[JFinch57]

Usually on a 120/240v system the Neutral wire should be at least the same size as the phase wires. IMO that #4 neutral is not big enough based on the #3/0 phase wires.

Remember the Neutral wire can carries the same current as the phase wires on the trip back from the load if there is ever an unbalance load.

Yes, but it's not a 120/240V system, it's a balanced 240V generator. It only caries the same current if one leg has 0 current, then the neutral matches the one leg that's carrying 100%. The correct code related answer is posted by adoublee above:

I'm surprised this thread had no discussion of NEC 705.95(A) and NEC 705.95(B). Article 705 is for interconnected electric power production sources (grid-tie PV inverters apply). These sections basically says if the neutral is only used for instrumentation, voltage detection, or phase detection (isn't a natural path for all phase current to return) that it can be sized equal to a EGC (NEC Table 250.122).

If a grid-tie inverter is 240V and neutral that would normally be expected to provide service to 120V connection is actually only a voltage detection component to meet IEEE standards (a high impedance path back to the inverter), almost all neutral current will actually be returning to the utility transformer.