My 2 x Growatt SPF 5000 ES trigger the upstream RCBO when working in parallel

[GeorgeH]

Hello,

I've installed 2 x Growatt SPF 5000 ES (I live in Europe) in parallel and a very annoying problem I'm now facing is that they occasionally trigger the upstream RCBO (the residual current component not the over current one) fitted by the utility company.

I am confident it's not because of the downstream consumers because if I simply bypass the inverters the RCBO would not trigger.

Even though this is a DIY project I don't think there are problems with wiring because I can (most of the time) power the entire house and works just fine.

Some have suggested that the RCBO is not calibrated correctly but since it does not trigger if I bypass the inverters I think probability is low.

The RCBO would almost certainly trigger when left running for more than a day or two but also it does trigger almost immediately in the following conditions:
- units switched off from power button
- output of units not connected together
- battery not connected
I am aware that for a correct parallel setup the above should not happen but it's up to the inverters to fault rather than trigger the RCBO, right?

Also when the units are not wired to work in parallel the RCBO does not trigger.

What do you guys suggest I do at this point? The system is mostly unusable in this state.

Attaching the wiring diagram for your reference.

Kind regards,
George

 

[timselectric]

Does it do it in all modes?
Or only in battery mode?

 

[timselectric]

These units create a N/G bond in battery mode. And when turned off.

 

[GeorgeH]

Thank you for trying to help @timselectric. I've checked today and the mode is SBU.
Also likely relevant: in this SBU mode if I only connect either 1 of them while turned off from the button it won't trigger the RCBO.
As soon as I connect both in parallel while both turned off they trigger the RCBO.

 

[GeorgeH]

I'm thinking about the N/G bond you've mentioned...
If I understand things correctly it means around half of the return current goes through the N and half through G.
I've just checked: the RCBO is 300mA rather than 30mA.
Standby load for one inverter is around 60W as far as I know which is ~ 0.27A at 220V and divided by 2 (G & N) we get 0.135A which should not trigger. As soon as there are 2 inverters we get much closer to 0.3A.
That could explain why it doesn't trigger when either 1 is connected but does trigger when both are connected.

Am I talking nonsense?

I guess I could try fitting a 30mA RCBO (rather than 300mAone ) upstream of the inverters to see if it triggers with a single inverter connected... ?‍

 

[timselectric]

I'm thinking about the N/G bond you've mentioned...
If I understand things correctly it means around half of the return current goes through the N and half through G.
I've just checked: the RCBO is 300mA rather than 30mA.
Standby load for one inverter is around 60W as far as I know which is ~ 0.27A at 220V and divided by 2 (G & N) we get 0.135A which should not trigger. As soon as there are 2 inverters we get much closer to 0.3A.
That could explain why it doesn't trigger when either 1 is connected but does trigger when both are connected.

Am I talking nonsense?

I guess I could try fitting a 30mA RCBO (rather than 300mAone ) upstream of the inverters to see if it triggers with a single inverter connected... ?‍

I think that you are on the right track.

 

[timselectric]

In SUB mode, these units shouldn't be activating their N/G bond. But when turned off, it will be activated. Because it's a normally closed relay. So for a split second it could have the N/G bonding relay closed before it releases the grid bypass relay. Which could explain why it happens when you turn them off.

 

[Hedges]

I think neutral/ground bonding in the inverter is a feature needed if grid could be disconnected, like mobile RV/boat use.
I would defeat the switchable bonding and simply tie output neutral to input neutral, without it connecting to ground. Always relaying on neutral/ground bond of the utility connection.

In the US, we normally make neutral/ground bond in the box with first circuit breaker after utility meter. I know it varies elsewhere, so not certain yours is bonded.

 

[timselectric]

I think neutral/ground bonding in the inverter is a feature needed if grid could be disconnected, like mobile RV/boat use.
I would defeat the switchable bonding and simply tie output neutral to input neutral, without it connecting to ground. Always relaying on neutral/ground bond of the utility connection.

In the US, we normally make neutral/ground bond in the box with first circuit breaker after utility meter. I know it varies elsewhere, so not certain yours is bonded.

Yup, I feel the same.
Dynamic bonding is only good for portable systems.

 

[GeorgeH]

I think neutral/ground bonding in the inverter is a feature needed if grid could be disconnected, like mobile RV/boat use.
I would defeat the switchable bonding and simply tie output neutral to input neutral, without it connecting to ground. Always relaying on neutral/ground bond of the utility connection.

In the US, we normally make neutral/ground bond in the box with first circuit breaker after utility meter. I know it varies elsewhere, so not certain yours is bonded.

We also have neutral/ground bond in the utility box where the meter is.
Let me clarify what you mean:
* disconnect ground wires from both input and output, or just input?
* wire together neutral from input and output
is that right?
Let me voice my worries here:
* will this not fry the inverters or make them malfunction in some way?
* will this be fine 2 inverters working in parallel?
* will this setup not void the inverters warranty?
* since grind can technically be disconnected by opening the input circuit breakers do I need to have another NG bond for the output?
* is there any use for program 24 "Allow neutral and grounding of AC output is connected together"?
Thank you for your help so far

 

[Hedges]

We also have neutral/ground bond in the utility box where the meter is.
Let me clarify what you mean:
* disconnect ground wires from both input and output, or just input?

RV or boat can have inverter input unplugged from grid.

* wire together neutral from input and output
is that right?

That is what I mean. Some inverters (mobile) have a relay, connecting output neutral to input neutral when on grid, disconnecting that and connecting output neutral to ground when off grid.

(My Sunny Island are not meant to be mobile, rather always connected to grid. They require neutral/ground bond to be made externally. They have neutral internally connected straight through, input and output are a single "net".)

Let me voice my worries here:
* will this not fry the inverters or make them malfunction in some way?

I wouldn't think so.
I don't think they even measure neutral current.
Input neutral is expected to be bonded to ground. When internal relay connects neutrals together, output neutral is bonded to ground.
Internal relay switches, and for a couple milliseconds output neutral is isolated, then it is bonded to ground

* will this be fine 2 inverters working in parallel?

I think one neutral ground bond somewhere is fine.
Whether neither/one/both switch, neutral is still neutral is still grounded in one place.
(Line) would be a problem if one switched but the other didn't; then output of a free-running inverter is connected to grid.)

* will this setup not void the inverters warranty?

That is the issue. Some brands say not to connect neutral input to output. We can't figure out why.
(Except, if allowed to bond output neutral to ground through a relay, you would then have two bonds in the system and "objectionable current" would run through the ground wire that is now in parallel with neutral wire.)

* since grind can technically be disconnected by opening the input circuit breakers do I need to have another NG bond for the output?

Why can ground be disconnected by opening a circuit breaker? We don't put a breaker on ground where I come from.

Does your breaker open neutral? That may be done in Europe. It may be done by RCB. This would be a problem in that it leaves you without neutral being grounded. In that case may want the internal bonding relay, treat the grid like shore power. But then you have the problem of RCB tripping.

* is there any use for program 24 "Allow neutral and grounding of AC output is connected together"?

That may mean you have software control of neutral-ground bonding, don't have to remove a screw or otherwise do hardware modification. I'm not familiar with the settings.


If RCB due in fact open neutral (something I didn't realize but have been told), then I think you need to keep neutral-ground bonding of the inverter. In that case I'm not sure how to solve this. Maybe allow master inverter to perform neutral-ground bond (single inverter setup was OK?) and defeat it for the slave.

 

[GeorgeH]

RV or boat can have inverter input unplugged from grid.



That is what I mean. Some inverters (mobile) have a relay, connecting output neutral to input neutral when on grid, disconnecting that and connecting output neutral to ground when off grid.

(My Sunny Island are not meant to be mobile, rather always connected to grid. They require neutral/ground bond to be made externally. They have neutral internally connected straight through, input and output are a single "net".)



I wouldn't think so.
I don't think they even measure neutral current.
Input neutral is expected to be bonded to ground. When internal relay connects neutrals together, output neutral is bonded to ground.
Internal relay switches, and for a couple milliseconds output neutral is isolated, then it is bonded to ground



I think one neutral ground bond somewhere is fine.
Whether neither/one/both switch, neutral is still neutral is still grounded in one place.
(Line) would be a problem if one switched but the other didn't; then output of a free-running inverter is connected to grid.)



That is the issue. Some brands say not to connect neutral input to output. We can't figure out why.
(Except, if allowed to bond output neutral to ground through a relay, you would then have two bonds in the system and "objectionable current" would run through the ground wire that is now in parallel with neutral wire.)



Why can ground be disconnected by opening a circuit breaker? We don't put a breaker on ground where I come from.

Does your breaker open neutral? That may be done in Europe. It may be done by RCB. This would be a problem in that it leaves you without neutral being grounded. In that case may want the internal bonding relay, treat the grid like shore power. But then you have the problem of RCB tripping.



That may mean you have software control of neutral-ground bonding, don't have to remove a screw or otherwise do hardware modification. I'm not familiar with the settings.


If RCB due in fact open neutral (something I didn't realize but have been told), then I think you need to keep neutral-ground bonding of the inverter. In that case I'm not sure how to solve this. Maybe allow master inverter to perform neutral-ground bond (single inverter setup was OK?) and defeat it for the slave.

By "since grind can technically be disconnected by opening the input circuit breakers do I need to have another NG bond for the output?"
I meant grid not ground, I'm sorry for the typo. There is no circuit breaker for ground. Feel free to have a look at the wiring in the original question.
As far as I can see RCCB / RCBO does indeed open both L and N which normally is not a problem because once you open L there is no need for the return path, right?
When I started this DIY project I didn't imagine grounding is such a complex issue

 

[Hedges]

Correct, no need for N return path. Return path is local to inverter, load, the L and N between them.

But since N is opened by RCCB, your inverter and loads no longer have N bonded to ground, so you do need a neutral/ground bonding relay.
See if you can defeat (or just disable in configuration) the second bonding relay in the second inverter.

Yes, so many different configurations around the world makes it complex.
If this was just a portable Bluetti, only one configuration.

A brute force way to simplify it is isolation transformer between grid and your system. Heavy, expensive, has some losses.