diy solar

diy solar

Definitive answer on Growatt inverter neutral bonding?

PEFarmer

New Member
Joined
Oct 24, 2021
Messages
8
Hi all,

I'm planning to install a grid assisted solar system on my home (not back-feeding) Like most, my home electrical system has a neutral-ground bond at the first point of entry (meter base). For my install, I believe that I want an inverter that does not bond the neutral to ground *at any time* whether passing grid power through, or running off of battery power via inverter. I've seen conflicting information (including in filterguy's data sheet) about whether the Growatt units bond neutral when making power from battery. There seems to have been a change in firmware removing access to the dry contact. This is a very important thing for me and may lead me to select an entirely different setup if it can't be confirmed. I'd love if someone with a growatt (particularly the 6k or 12k low freq. models) would put a clamp meter on the ground wire and neutral wire when operating in battery mode and under load to verify that there is no current on the ground conductor and there is current on the neutral appropriate to the loads. For reference, I plan to bring inverter power into the house (into a sub-panel) via an interlocked breaker (intended for generator) that does not switch neutrals (neutral bond remains in place for entire system under any condition)

Thanks!
 
If your main panel is bonded, NO OTHER PLACE IN THE SYSTEM CAN BE.
Its that simple. Bonded together in one place only.
And...
The output from the inverter needs to feed THE BONDED PANEL.
Fully agree. That's why I'm strugging so much finding an inverter that NEVER bonds the neutral to ground. I'm seeing lots that do so when running under battery power.
 
Fully agree. That's why I'm strugging so much finding an inverter that NEVER bonds the neutral to ground. I'm seeing lots that do so when running under battery power.
It is fine for the inverter to bond the neutral, you just have to unbond the neutral in the panel.

It matters that the first powered neutral be bonded.
And all downstream neutrals are unbonded.
 
It is fine for the inverter to bond the neutral, you just have to unbond the neutral in the panel.

It matters that the first powered neutral be bonded.
And all downstream neutrals are unbonded.

When you say unbond the neutral in the panel, are you meaning the main neutral-ground bond in the service, or in the critical loads sub-panel served by the inverter? My strong preference would be that the main neutral bond be carried through on the neutral conductor to every sub-panel, and that the inverter not to bond the neutral. 24/7's suggested diagram in their FAQ seems to agree with this, but maybe I'm missing something?
 
When you say unbond the neutral in the panel, are you meaning the main neutral-ground bond in the service, or in the critical loads sub-panel served by the inverter? My strong preference would be that the main neutral bond be carried through on the neutral conductor to every sub-panel, and that the inverter not to bond the neutral. 24/7's suggested diagram in their FAQ seems to agree with this, but maybe I'm missing something?
All sub panels need to have the neutrals and grounds on separate buss bars. And the neutral needs to be unbounded from the ground.
If you have a separate panel fed only by the inverter, having the bonding in the inverter, and the sub panel has its own ground separate from the neutrals is correct.
 
All sub panels need to have the neutrals and grounds on separate buss bars. And the neutral needs to be unbounded from the ground.
If you have a separate panel fed only by the inverter, having the bonding in the inverter, and the sub panel has its own ground separate from the neutrals is correct.
But then what happens when the inverter is required to go in bypass mode, and you have the bonding at the inverter, and the main panel?
 
But then what happens when the inverter is required to go in bypass mode, and you have the bonding at the inverter, and the main panel?
It should not be an issue. The meter panel on homes is always ground bonded, as is the pole transformer. As long as no separate circuits exist between the bonded locations it shouldn’t matter.
 
It should not be an issue. The meter panel on homes is always ground bonded, as is the pole transformer. As long as no separate circuits exist between the bonded locations it shouldn’t matter.
If the main panel is bonded, and the inverter is bonded, then neutral current will be passed along the ground between the two, which will necessarily elevate the ground on the inverter side, unless a ground rod is also placed at the inverter.

If they are right next to each other, with large conductors coupling them then it doesn't matter, but if the inverter is in an outbuilding then it will make a significant difference. Even if the outbuilding is using no significant current, a split phase imbalance in the main building could place a large potential on the ground at the outbuilding. This will electrify the inverter enclosure, possibly the solar panel frames, and every other grounded object in the shed.

I'd love if someone with a growatt (particularly the 6k or 12k low freq. models) would put a clamp meter on the ground wire and neutral wire when operating in battery mode and under load to verify that there is no current on the ground conductor and there is current on the neutral appropriate to the loads. For reference, I plan to bring inverter power into the house (into a sub-panel) via an interlocked breaker (intended for generator) that does not switch neutrals (neutral bond remains in place for entire system under any condition)

I'll do that when I can (over the next few days) with my SPF 12000T DVM (growatt 12k, LF). It would surprise me if it did do a ground bond, though - that's a bad, bad thing. I could understand if a smaller stand-alone unit offered a ground bond for simplicity, such as would be needed in an RV where there's no expectation of a separate electrical panel, but a 12kW unit should always be connected to a circuit breaker panel.
 
Yes, this is certainly an issue if there are any circuits in between.
Under Aus regs (I'm from Australia), every installation can only have one MEN (neutral/ground bond) and this is more what I was thinking of.

Because in the case of bypass mode, the installation will have 2 neutral ground bonds (and be non-compliant in Aus).

I assume by your responses, this isn't the case with the NEC, so all is good?

The issue with having 2 bonds like you suggest, is that the Neutral and ground will be in parallel, and the likelihood is that the ground cable will be undersized, while taking close to half the current that is being supplied to the sub board.

Again, this is only an issue in bypass mode, with a permanent neutral/ground bond at the inverter. Once it flicks to battery mode, all is fine.

Am I wrong with the above?
 
Last edited:
If the main panel is bonded, and the inverter is bonded, then neutral current will be passed along the ground between the two, which will necessarily elevate the ground on the inverter side, unless a ground rod is also placed at the inverter.

If they are right next to each other, with large conductors coupling them then it doesn't matter, but if the inverter is in an outbuilding then it will make a significant difference. Even if the outbuilding is using no significant current, a split phase imbalance in the main building could place a large potential on the ground at the outbuilding. This will electrify the inverter enclosure, possibly the solar panel frames, and every other grounded object in the shed.



I'll do that when I can (over the next few days) with my SPF 12000T DVM (growatt 12k, LF). It would surprise me if it did do a ground bond, though - that's a bad, bad thing. I could understand if a smaller stand-alone unit offered a ground bond for simplicity, such as would be needed in an RV where there's no expectation of a separate electrical panel, but a 12kW unit should always be connected to a circuit breaker panel.
The inverters that I've seen handle this, ONLY place the neutral/ground bond in battery mode. And opens the bond in bypass mode. Which is a safe way of handling it. I don't know how this specific inverter handles it though.
 
@stienman did you ever get run the test?

You reminder is timely - I'm fiddling with the system on a cloudy day, so I was able to fit a test in without difficulty.

Here's my process, in case questions arise in the future about something I might have missed:

1. Turn off the inverter
2. Disconnect all loads at panel via breaker
3. Disconnect all grounds at the inverter's terminals
4. Disconnect the neutral at the inverter's terminals (these two steps because the neutral and ground are bonded at the panel)
5. Turn the inverter on - using both solar and battery, not in bypass mode (though AC is connected and available)
6. Using a multimeter set to resistance, test the resistance between the inverter's ground terminal and inverter enclosure (verify meter is reading correctly, and the ground is bonded to the enclosure) --> Result: ~0.6ohm, therefore ground terminal is bonded to enclosure internally
7. Test the resistance of the inverter's ground terminal to the inverter's neutral terminal --> Result: open circuit, therefore the ground terminal is NOT bonded to the neutral terminal.

So as far as I can tell from external measurements, at no point during operation - whether bypass or inverter mode - is the ground connected to the neutral inside the inverter. For me this is conclusive, and I have no concerns, but clearly it's not a comprehensive test - I didn't test the case where it's running off batteries without solar, or without AC input, but I can't see a reason why they would bond ground and neutral in those cases and not in the case I tested above (AC available, but in inverter mode running off solar and battery)

This is with the Growatt SPF 12000T-DVM (not the newer MPV version)
 
You reminder is timely - I'm fiddling with the system on a cloudy day, so I was able to fit a test in without difficulty.

Here's my process, in case questions arise in the future about something I might have missed:

1. Turn off the inverter
2. Disconnect all loads at panel via breaker
3. Disconnect all grounds at the inverter's terminals
4. Disconnect the neutral at the inverter's terminals (these two steps because the neutral and ground are bonded at the panel)
5. Turn the inverter on - using both solar and battery, not in bypass mode (though AC is connected and available)
6. Using a multimeter set to resistance, test the resistance between the inverter's ground terminal and inverter enclosure (verify meter is reading correctly, and the ground is bonded to the enclosure) --> Result: ~0.6ohm, therefore ground terminal is bonded to enclosure internally
7. Test the resistance of the inverter's ground terminal to the inverter's neutral terminal --> Result: open circuit, therefore the ground terminal is NOT bonded to the neutral terminal.

So as far as I can tell from external measurements, at no point during operation - whether bypass or inverter mode - is the ground connected to the neutral inside the inverter. For me this is conclusive, and I have no concerns, but clearly it's not a comprehensive test - I didn't test the case where it's running off batteries without solar, or without AC input, but I can't see a reason why they would bond ground and neutral in those cases and not in the case I tested above (AC available, but in inverter mode running off solar and battery)

This is with the Growatt SPF 12000T-DVM (not the newer MPV version)
So on this instance, if I am using a Growatt SPF 12000T-DVM exclusively off grid into an off grid main panel, with the only alternative power coming into the Growatt being a generator coming into the Growatt's AC input, I WOULD need to bond the ground and neutral for this off grid system in the main panel? And that would be the one and only location of the ground neutral bond?
 
So on this instance, if I am using a Growatt SPF 12000T-DVM exclusively off grid into an off grid main panel, with the only alternative power coming into the Growatt being a generator coming into the Growatt's AC input, I WOULD need to bond the ground and neutral for this off grid system in the main panel? And that would be the one and only location of the ground neutral bond?
Yes, and ensure the generator does not have a Neutral-Ground bond. Some portable generators have this for safety reasons.

 
Agreed. The ground/neutral bond should happen at one location exactly - no more, no less. So with this inverter, you should have this happen at the main panel.
 
I figured it would be easiest to ask here than start a new thread. I'm looking at a Growatt LF inverter and I have been reading until cross eyed for weeks.

On the Growatts with a LF transformer , there is only a 240v L1 and L2 input, no neutral input. There is a L1, L2 , and N for 240/120 output. I understand the NEC code on ground/neutral bonding but in battery mode this inverter is a separately derived system according to them as it has a transformer. Watts247 shows the neutral output being tied to the main utility neutral. NEC says separately derived is no connection with any current carrying conductor.

In my state the utility company says if you are a separate system you cannot connect electronically or operate in parallel. If you tie neutral to theirs and neutral carries current it is now not off grid and now requires an interconnect agreement with the utility. Am I wrong on this neutral carrying current?

Also, would this neutral out of the transformer tying to utility have the same balancing attempt problem as the other stand alone Growatt transformer?

If inverter output neutral was not tied to utility and bonded to a separate ground in a sub panel, it would be a separate main panel, not a sub panel and the inverter would be the main supply, am I wrong? I'm trying to avoid an interconnect violation and problem with the transformer neutral output.
 

Attachments

  • PanelWiring2of2.jpg
    PanelWiring2of2.jpg
    264.7 KB · Views: 50
I figured it would be easiest to ask here than start a new thread. I'm looking at a Growatt LF inverter and I have been reading until cross eyed for weeks.

On the Growatts with a LF transformer , there is only a 240v L1 and L2 input, no neutral input. There is a L1, L2 , and N for 240/120 output. I understand the NEC code on ground/neutral bonding but in battery mode this inverter is a separately derived system according to them as it has a transformer. Watts247 shows the neutral output being tied to the main utility neutral. NEC says separately derived is no connection with any current carrying conductor.

In my state the utility company says if you are a separate system you cannot connect electronically or operate in parallel. If you tie neutral to theirs and neutral carries current it is now not off grid and now requires an interconnect agreement with the utility. Am I wrong on this neutral carrying current?

Also, would this neutral out of the transformer tying to utility have the same balancing attempt problem as the other stand alone Growatt transformer?

If inverter output neutral was not tied to utility and bonded to a separate ground in a sub panel, it would be a separate main panel, not a sub panel and the inverter would be the main supply, am I wrong? I'm trying to avoid an interconnect violation and problem with the transformer neutral output.
I agree.
The output panel would be a separate panel.
The grid feed ONLY uses 240V, no neutral bond. So the only bond is the ground to the inverter.
As long as you do not EVER tie the output to the grid, it should be fine.
 
Hi all,

I'm planning to install a grid assisted solar system on my home (not back-feeding) Like most, my home electrical system has a neutral-ground bond at the first point of entry (meter base). For my install, I believe that I want an inverter that does not bond the neutral to ground *at any time* whether passing grid power through, or running off of battery power via inverter. I've seen conflicting information (including in filterguy's data sheet) about whether the Growatt units bond neutral when making power from battery. There seems to have been a change in firmware removing access to the dry contact. This is a very important thing for me and may lead me to select an entirely different setup if it can't be confirmed. I'd love if someone with a growatt (particularly the 6k or 12k low freq. models) would put a clamp meter on the ground wire and neutral wire when operating in battery mode and under load to verify that there is no current on the ground conductor and there is current on the neutral appropriate to the loads. For reference, I plan to bring inverter power into the house (into a sub-panel) via an interlocked breaker (intended for generator) that does not switch neutrals (neutral bond remains in place for entire system under any condition)

Thanks!
My setup: 12k Growatt under load. NG bonded in the main panel. No subpanels.

H1 - current
H2 - current
Neutral - current
Ground - no current
 
Back
Top