How does phase synchronization of off-grid inverters (AIO's) with grid charging feature work?

[fmeili1]

I wonder what happens in detail just at the moment when a typical off-grid inverter AIO device like EG4-6500EX will be connected to the grid via AC-in (at the moment when the AC-in breakers are switched on - all at the exact same time - while the off-grid inverter is still running in inverter "Solar/Battery" mode)?

When the AIO was started without AC-in connected to the grid, the phase position of the off-grid inverter could not be related in any way to the phase position of the grid. At the moment when the grid will be connected to the AC-in there will be a phase shift between both inverter output and grid. It could be by coincident, that they are 180 degree phase shifted which could result in a maximum of 2*120V*SQRT(2)=340V volt difference per phase (and 680V when using L1 and L2) if one source would have it's maximum at +170V and the other it's minimum at -170V at that moment.
But, if the grid would not be used by the AIO at this moment (if it runs in "Solar/Battery" mode with SBU), this should not hurt.

BUT

The reason for my questions is the following:
To prepare grid-charging for the EG4-6500EX I connect the AC-in's of the AIO's to the grid only if required (some minutes before program 12 setting "back to grid" will be triggered). From time to time (not very often), it happens that some of the inverters show Fault F60 (backfeed protection) just at the moment when the grid get's connected to the AC-in's. Today I had such a situation (F60) again but this time additionally it forced all my GFI's in my house sub-panel to switch off (this was the first time the GFI's gets triggered).

Because of this behavior I wonder if the problem may have something to do with phase shift differences (pure coincident).

I don't know if and how an AIO with grid charging and bypass feature handles this situation.

Will the AIO try to synchronize it's off-grid inverter output to get in sync with the grid phase as soon as the grid will be to connected to the AC-in to be prepared when it must switch to bypass (to not run in a possible high voltage difference while switching to bypass)?

Because the problem occur at the moment when the AC-in get's connected to the grid and NOT at the moment when the inverter changes to bypass, something happens inside the AIO just at the moment when connecting it to the grid - but what could it be?

I know that the EG4-6500EX is based on Voltronic Axpert Max and these AIO's are basically able to feed energy back to the grid (as a typical grid tied inverter), but this feature is disabled in the firmware of EG4 and by most of the other clones (e.g. MPP Solar 6548). But there are clones which allows it (e.g. Phocos PSW-H-6.5KW-120/48V).
So in principle these inverter needs to be able to get in sync with a grid (at least when export is possible). I know that this is not really comparable because off-grid inverters are voltage-sources while grid-tied inverters are current-sources but maybe it's related?

I just try to understand how these devices will handle these situation at all.
Would a zero-crossing switch at the AC-in side help?

Theoretically a grid sense-input at the AIO which is connected to the grid would help the AIO's to sync it's off-grid inverter phase with the grid phase just to be prepared if the grid get's connected... would be nice...

 

[timselectric]

Will the AIO try to synchronize it's off-grid inverter output to get in sync with the grid phase as soon as the grid will be to connected to the AC-in to be prepared when it must switch to bypass

Yes

 

[fmeili1]

Yes

Do you have some more details how this works in detail?

Will the AIO try to synchronize it's off-grid phase to the grid phase as soon as the grid gets connected?
Will the AIO shift it's phase "slowly" as soon as the grid get's connected to get a match with the phase of the grid?

Do you have a possible explanation why the AIO's sometimes throw fault F60 just at the moment when the grid connects to the AC-in (while inverter still running in "Solar/Battery" mode in SBU setting)?

 

[timselectric]

Will the AIO try to synchronize it's off-grid phase to the grid phase as soon as the grid gets connected?
Will the AIO shift it's phase "slowly" as soon as the grid get's connected to get a match with the phase of the grid?

Yes, it will shift as soon as it has a reference.
I'm not sure exactly how fast. But it's pretty quick. And it won't transfer until it's synchronized.

Do you have a possible explanation why the AIO's sometimes throw fault F60 just at the moment when the grid connects to the AC-in (while inverter still running in "Solar/Battery" mode in SBU setting)?

I can only guess. But it could be the voltage sag from switching under load.
As far as the GFCI breakers. That's probably due to switching lag.

 

[fmeili1]

Yes, it will shift as soon as it has a reference.

This was my guess

I'm not sure exactly how fast. But it's pretty quick. And it won't transfer until it's synchronized.

That's good to know that the transfer would not happen after already synchronized! It was my guess also, but I wasn't sure...

When I'll find time I'll try to do some oscilloscope measures to understand how and how fast it'll work (in cases where it don't get the F60) to get a better understanding.

I can only guess. But it could be the voltage sag from switching under load.

Interestingly, I had more often this F60 fault (at the moment when AC-in gets connected) when the AIO don't have any load at all (these were my fist tests). If I have at least a bit of load (my house idle load is about 500W with 6 inverters in parallel, so about 85W per AIO) on the AC-out the problem occur less often - about 10 times less compared to the situation without a load!

But maybe if the load is very high the F60 may happen again more often because of the voltage sag while switching which you've mentioned (the central heat pump was running when it happens the last time).

Because I already control the AC-in's connection to the grid of all six AIO's via two 3-pole AC contactors with smart home rules, I will try to delay the AC-in contactor closing to a time slot where the load will be "low"... maybe this helps. But because this F60 is relatively rare, it will take some time to find out...

As far as the GFCI breakers. That's probably due to switching lag.

ok... nothing I can do here...


Thanks!

 

[AZ Solar Junkie]

I'm just curious why you only connect it to the grid when it's about to need it. Why not just leave it connected all the time? I know there's a small idle draw from the grid when it's connected - is that your reasoning or something else?

 

[fmeili1]

I'm just curious why you only connect it to the grid when it's about to need it. Why not just leave it connected all the time? I know there's a small idle draw from the grid when it's connected - is that your reasoning or something else?

Yes.

It was my first idea to let the grid connected all the time (I have grid anyway because city code does not allow to be disconnected) just as a backup for the very rare case when my batteries will get empty. But I realized that more than the half of the 93 Watt idle consumption of the EG4-6500EX will be drawn from the grid just when connected to it. Because I have 6 of the EG4-6500EX in parallel split phase setup, these 50W per AIO grid usage will add up 300W. I don't want to pay for these 2600 kWh per year just for maybe 1-2 days where I really need it here in south of Mohave County (I have usually enough battery capacity). Also I don't need the grid for an overload situation because the 39kW inverter power will never be the limit, so I've disabled the over load bypass feature.

Because I need the grid very rarely , the problem is not really important but it would be great if it would work!

 

[AZ Solar Junkie]

Yes.

It was my first idea to let the grid connected all the time (I have grid anyway because city code does not allow to be disconnected) just as a backup for the very rare case when my batteries will get empty. But I realized that more than the half of the 93 Watt idle consumption of the EG4-6500EX will be drawn from the grid just when connected to it. Because I have 6 of the EG4-6500EX in parallel split phase setup, these 50W per AIO grid usage will add up 300W. I don't want to pay for these 2600 kWh per year just for maybe 1-2 days where I really need it here in south of Mohave County (I have usually enough battery capacity). Also I don't need the grid for an overload situation because the 39kW inverter power will never be the limit, so I've disabled the over load bypass feature.

Because I need the grid very rarely , the problem is not really important but it would be great if it would work!

Makes sense - especially since you have six of them. With one it would be less of a concern ?