SteveCook I am interested to see that this thread is still going. I suppose that's not surprising.
Faced with a high cost for an ATS, my installer has recommended a manual switch, with or without a simple relay contactor to automate the change over. He has suggested that, if I want the relay, it should be set to operate from the grid and not the EPS output, by default. This would mean that, at the point of grid failure, the relay would transfer the EPS circuits to the EPS output, and that at the point of grid recovery, the load would transfer away from the EPS output back to the default.
I have suggested one amendment, to add a broyce delay relay on the control input to the contactor so that, when the grid voltage restores, there is a delay before switching. I have suggested 15 seconds to allow for those failures of the grid which don't restore cleanly. I think such a precaution is common on many electronic devices to protect them from grid errors and I would hope that the Givenergy inverter is itself so protected. My solar inverter (separate pv installation) messes about for minutes checking the grid supply voltage and frequency before reconnecting.
If I understand you correctly, you have suggested that the EPS load should by default run through the inverter. Why?
The EPS load is going to be varied, although in my case I don't expect it to ever exceed 1Kw. But I would expect the load to go up and down, and also that I could go to the designated EPS CU and switch it off or on at the switch disconnector at any time whether operating on the EPS output from the grid or the battery. And also that I could operate the manual changeover switch at any time without causing a problem. The load includes a fridge, a freezer, and several electronic devices, and this is not clean like a resistive load. If the EPS output is sensitive and fussy about the load and might break, then maybe I should reconsider the whole idea.
What do you think? Do you think the EPS output is fussy about load variation?