Multi-branch common neutral on split phase EG4 inverter

[JCSchwarb]

All-

I reposted this as separate topic to focus on common neutral used across two split phase inverters.
Appreciate the community experts help on my EG4 6400 6 Inverter Installation. My inverters are wired to Grid Input, PV array, and EG4 batteries. I used a three wire 4 Ga with ground to supply 3 sets of split phase, total of 6 inverters. A common neutral has been split between each matched inverter pair L1/black, L2/red with shared 4ga neutral for grid input and a common neutral matched pair with L1/black and L2/red with shared neutral split between the output. While multi branch is permissible in this type of installation, is there anything about the EG4’s that may cause ill effects? Each black/red pair terminate in a single 60amp double pole, per NEC. Signature pointed me to the manual, but do not see multi-branch common neutral as an issue on a split phase configuration. While this was not my preference in wiring, I followed this approach due to no single strand THHN in my area for that gauge wiring. In retrospect, I learned a hard lesson in complexity and possible future faults with so many wiring connections and added cost of Polaris connectors.

Thanks in advance to help me understand if EG4 or other inverters in split-phase can use common multi branch neutral sharing assuming NEC compliance.

-Jay

 

[JCSchwarb]

No takers on this question? Appreciate your help!

 

[BentleyJ]

It sounds like the system is wired up as described above. Its not clear if you've tested the system and had some kind of problem or are asking if its OK to turn on the inverters and use them as is. Is there a specific question?

A 240/120V split phase inverter when wired into a standard single phase breaker panel with 2 separate Hot bus bars, 1 Neutral and 1 ground bus bar should operate the same way as it would if normal utility power were feeding it. A common neutral, multi-branch circuit is code compliant if the following conditions are met.
1) The Hot wires are on opposite bus bars. (you mentioned a standard 2 pole breaker so that condition should be satisfied)
2) There is a handle tie to connect the breaker handles together. (A 2 pole breaker usually has the handle tie installed already)
3) Not sure if there is any additional NEC labeling requirements to specify which circuits are multi-branch.

Hesitate to open this "can of worms", as it usually generates hundreds of replies. There should be ONE and only ONE Neutral-Ground bond in the system. If its in the inverter then the one in the breaker panel, if present, should be eliminated. If the N-G bond is not present in the inverter then it should be in the breaker panel. Most commercially available main panels already have the N-G bond built in while most subpanels do not.

 

[JCSchwarb]

Thank you Bentley! Your answer is very precise and thorough. Really appreciate your confirmation. I am not understanding why the sales/manufacturing company is unable to articulate either NG bond or multibranch topics. I assume, without having more information, these inverters are setup from the factory assuming normal operation as stand alone units not tied to existing panels. I will look up if any labeling is required by NEC, but I have never seen one. The fact it is apparent from wire ties and wire colors should be self evident. Thanks again Bentley!!

Best - Jay

 

[timselectric]

If a single multi wire branch circuit enters the panel, either as a cable or in conduit. No extra identification is required. If more than one enters the panel in a single raceway. Each neutral must be grouped together with its ungrounded conductors. This can easily be done with a wire tie.

 

[JCSchwarb]

Thanks Tim!

 

[JCSchwarb]

Thank you Bentley! Your answer is very precise and thorough. Really appreciate your confirmation. I am not understanding why the sales/manufacturing company is unable to articulate either NG bond or multibranch topics. I assume, without having more information, these inverters are setup from the factory assuming normal operation as stand alone units not tied to existing panels. I will look up if any labeling is required by NEC, but I have never seen one. The fact it is apparent from wire ties and wire colors should be self evident. Thanks again Bentley!!

Best - Jay

Appreciate an answer to the following common question:
if two inverters are outputting 240V split phase, such that black to center tap "neutral" is 120V and center to red is 120V, what amperage max load is expected on the shared neutral?

 

[Hedges]

If split phase, neutral will never carry more than a single line (red or black); it carries |L1 - L2|. So can be same gauge.
If the two inverters can run independently, not synchronized, then neutral could end up carrying the sum, L1 + L2. Need to ensure that can never happen.

 

[JCSchwarb]

If split phase, neutral will never carry more than a single line (red or black); it carries |L1 - L2|. So can be same gauge.
If the two inverters can run independently, not synchronized, then neutral could end up carrying the sum, L1 + L2. Need to ensure that can never happen.

Thanks Hedges - Surely split-phase-designed inverters must communicate the in-phase sine wave in order to deliver a synchronized split-phase load distribution without causing deliterious effects down stream to load equipment. The vendor (Signature Solar) is having a hard time providing support for this question. The EG4 manual does not cover it either. I would like to know how the inverters synchronize with incoming AC voltage and outputting a synchronized 180 deg in-phase sine wave.

 

[Hedges]

Middle of page 18 gives communications connections for 120/240V split-phase.
Page 55 gives commissioning of split-phase.

https://eg4electronics.com/wp-content/uploads/2022/10/EG4-6500EX-48-Manual.pdf

Unlike 2 units in parallel for 120V, 2 units stacked for 120/240V split phase does not have "current sharing cables" in addition to communication cables.
But I think it remains under software configuration capability to operate the inverters in-phase zero degrees, which would overload a neutral of same size. That is, software can start a fire.

It might be most prudent to design/build your connections so operation in parallel does no harm. Putting L1/N/L2 through ganged 3-pole breaker might accomplish that (don't want fuse or un-ganged breakers which could open neutral by itself.)
Or, give each inverter a separate neutral connection to breaker panel, same size as line.

Also, do not use Romex 3 + ground (red/white/black/bare) wires for two 120V circuits sharing a neutral. Use separate 2 + ground for each 120V circuit. That ensures if L1 and L2 are in phase, neutral is not overloaded.

 

[JCSchwarb]

If this is the case, the manual should be very explicit about not sharing neutrals. I suspect that the inverters must be synced otherwise the load side would have catastrophic damage to equipment.

 

[Hedges]

The manual could only be explicit about that if the designers/tech writers thought of it.
Paralleled, each would have its own neutral. Likely stacked for split-phase would too.
In my system I have a shared neutral due to conduit fill constraints. I may not have though through how it could be configured wrong at the time, was careful to make it right when configured as intended.
Since you posed a question, I was in a position to see the forest despite the trees.

If L1 and L2 are either at 0 degrees phase or 180 degrees phase, about all equipment (except a center-tapped autotransformer) should be fine. 120V loads always see correct voltage, and 240V loads either see 240V or zero. 120 degrees (3-phase) is OK too because most 240V loads also work at 208V.
Unsynced with random angles, 240V loads would experience brownout.
One line off and the other live, 240V loads plus 120V loads on "off" phase would experience brownout.

 

[JCSchwarb]

Thanks Hedges. Since I designed as a multi branch split phase installation, I am a bit stuck, as the wiring is in the walls behind durarock. I will test the neutrals to verify what the current load is, to be on the safe side. My assumption has been that the inverter syncs with AC input and then outputs phase and angle accordingly. Wish the manual or vendor had a theory of operations section to explain this. There are lots of dangers lurking in this system. Another biggie is no mention on current sharing cabling necessity (why this is needed) or NG bonding requirements for different installations.

 

[Hedges]

It would sync to input. They would do that even if two completely separate systems.
With grid down, their synchronization to each other is important.

 

[GooseF16]

If split phase, neutral will never carry more than a single line (red or black); it carries |L1 - L2|. So can be same gauge.
If the two inverters can run independently, not synchronized, then neutral could end up carrying the sum, L1 + L2. Need to ensure that can never happen.

Yes correct..because ( and if) they are out of phase in any moment in time.. all 3 wires same size. you must make sure the inverters are synced. NGB ( neutral ground bonding), must be a single source, and coded to be nearest its source or at the first panel available off the split.. zillions of years ago, it was meant to be done in the meter back. but later done in main panels when we went to alternating bus bars.. Make darn sure your sub panels have that bond removed.. test it all the way thru. if you put a generator on it, make darn sure you dont have a double up too.. Many inverters hold the bond when off or on, but pass the bond when in "AC/charge" mode.. Victron gets around this hybrid.. I had the privilege of surviving a 240 volt ground loop getting me whilst standing on a large AC, and trying to pull a stray live wire (with all the appropriate glove, tools, and rubber shoes.. but it was a little damp, and I was the conduit it needed.. It hit me in the hand, and foot.. thank the lord it was my clothing that carried the current. ground loop.. make sure please.

 

[GooseF16]

Thanks Hedges. Since I designed as a multi branch split phase installation, I am a bit stuck, as the wiring is in the walls behind durarock. I will test the neutrals to verify what the current load is, to be on the safe side. My assumption has been that the inverter syncs with AC input and then outputs phase and angle accordingly. Wish the manual or vendor had a theory of operations section to explain this. There are lots of dangers lurking in this system. Another biggie is no mention on current sharing cabling necessity (why this is needed) or NG bonding requirements for different installations.

many inverter chargers pass the AC bond from the AC IN when in bypass mode, but hold the bond in OFF or Inverter mode.. Oh. and if any external transfer units are used they have to be checked too. such as an Transfer relay in an RV.. The onboard generator is usually bonded, and the ATR has to release that when on shore power for example.. Good stuff.. great questions..

 

[JCSchwarb]

Yes correct..because ( and if) they are out of phase in any moment in time.. all 3 wires same size. you must make sure the inverters are synced. NGB ( neutral ground bonding), must be a single source, and coded to be nearest its source or at the first panel available off the split.. zillions of years ago, it was meant to be done in the meter back. but later done in main panels when we went to alternating bus bars.. Make darn sure your sub panels have that bond removed.. test it all the way thru. if you put a generator on it, make darn sure you dont have a double up too.. Many inverters hold the bond when off or on, but pass the bond when in "AC/charge" mode.. Victron gets around this hybrid.. I had the privilege of surviving a 240 volt ground loop getting me whilst standing on a large AC, and trying to pull a stray live wire (with all the appropriate glove, tools, and rubber shoes.. but it was a little damp, and I was the conduit it needed.. It hit me in the hand, and foot.. thank the lord it was my clothing that carried the current. ground loop.. make sure please.

GooseF16 - appreciate the amplifying information. In my neck of the woods, the main meter has the ground and neutral bonded. My first means of disconnect is also bonded. This passed code 2 years ago and has been like that since originally installed prior to my PV system. Since the panels are back to back, perhaps it doesn’t matter, but has caused me a lot of anxiety as it seems to violate NEC, Mike Holt guidance, etc. All my load side panels are not NGB, but ran as 3-wire plus ground that tie back through to the main NGB panel.

 

[Hedges]

I don't think I would be anxious about it.
If you have neutral and ground in parallel, bonded at meter and bonded at disconnect, both wires will carry neutral current.
But so long as all loads branch off from disconnect, none split off at meter to a different breaker panel, I don't see an issue.

Where is ground rod connected, meter or disconnect?
If at meter, perhaps there can be a voltage rise of ground at disconnect and branch circuits where you touch chassis.
But if at disconnect, then no voltage rise.

I don't like the quality of some connections I see, little screws holding aluminum neutral lugs to an aluminum strip. Like one 10 awg flat head screw being the only connection of 200A neutral to utility in a meter box. I prefer something wrapped around the bus and held more solidly.

 

[timselectric]

GooseF16 - appreciate the amplifying information. In my neck of the woods, the main meter has the ground and neutral bonded. My first means of disconnect is also bonded. This passed code 2 years ago and has been like that since originally installed prior to my PV system. Since the panels are back to back, perhaps it doesn’t matter, but has caused me a lot of anxiety as it seems to violate NEC, Mike Holt guidance, etc. All my load side panels are not NGB, but ran as 3-wire plus ground that tie back through to the main NGB panel.

This is standard practice. All metal enclosures and raceways are bonded to the neutral, until you reach the first means of disconnect. This is where all grounding comes together. And also the location of the N/G bond. Everything after is kept separate.
There's no parallel path before it, because there's only one conductor. (You have to ignore the fact that the raceway could be considered a conductor)
This is acceptable because it benefits the power company. (They don't want to provide a ground conductor for every service) so the neutral pulls double duty, until it reaches the first means of disconnect.

 

[JCSchwarb]

This is standard practice. All metal enclosures and raceways are bonded to the neutral, until you reach the first means of disconnect. This is where all grounding comes together. And also the location of the N/G bond. Everything after is kept separate.
There's no parallel path before it, because there's only one conductor. (You have to ignore the fact that the raceway could be considered a conductor)
This is acceptable because it benefits the power company. (They don't want to provide a ground conductor for every service) so the neutral pulls double duty, until it reaches the first means of disconnect.

Thanks Tim! I had an electrician tell me that my main NGB needed to be removed sinve the meter can was bonded.