Home transfer switch from ups to grid, what about fase?

[PieterE]

I have a a battery powered ups system.
I Also have solarpanels and an invertor.
The ups I able to controll the output of the invertor if a connect it to the ups line (by frequency modulation). So no problem there.
This all works fine.

But:
I am planning to install a home transfer switch that switches my home from grid to ups.
This switch is so fast that electronics keep running.
When I would switch from grid to ups, there is no problem I think. The invertor and the ups are in sync with the grid. So also in sinc with each other.
Then there is a switch.
The ups will then provide power to the solar-invertor, power that is in phase, so the invertor will just keep running.
But: after few hours running on the ups it is likely that there will be a (big) difference between the phase of the ups and the grid when it comes back.
When the grid come back the transfer switch switches the house and also the invertor back to the grid in a very short time. But the grid will be likely out of phase with the invertor.
Normally if you would switch by hand via a switch that has a 'zero' position. (Not connected) the invertor will lose power and then start to resync, no problem. But if i would use this home transfer switch it is suddenly connected to a phase-out-of-sync powersource without detecting a powerloss.
Question is: can an invertor handle this? Or will there be smoke and fire?

 

[Ramussons]

This is similar to connecting 2 AC power sources in Parallel without Synchronizing.
Yes, there will be smoke and fire.

 

[Tony Stewart]

It might make more sense to always stay in sync and run off the inverter all the time then control the charger source. This assumes you have a reliable design with low ripple current stress and voltage stress on Caps, low heat rise otherwise short MTBF then you need a PLL.

Make sure you have lots of protection, fault sensors and insurance.

All switching requires dead time and the stored energy in the load (reactive energy) that flies back ( back EMF ) must be absorbed even if in phase and voltage sync. Appropriate TVS parts must be considered with passive snubbers.

 

[schmitt trigger]

Back in the late 1980s I worked for a company which made static transfer switches, for sensitive mainframe computers.

The key was that the UPS was always kept running (although unloaded) and it’s frequency and phase were locked to the mains via a CMOS PLL and logic gates.
It wasn’t a trivial circuit, quite the opposite, very challenging. A phase transfer error would result in a large BANG!

Nowadays it is done with FPGAs. Same principle, though. Hardware is way simpler, but do you know VHDL or Verilog programming?

 

[aliarifat794]

Some transfer switches are programmable and can delay the reconnection to the grid, allowing the inverter to resynchronize with the grid before fully switching back. This delay can be a few seconds to ensure synchronization.