Hi All. I am hoping someone can offer some insights.
I am running a LiFePO4 20KWH battery wall and a Victron Energy - Quattro 48V/10000 VA Inverter/Charger at my premises in a grid-with-battery-back-up for essential circuits configuration (no solar or generator). The system functions as a UPS with pure sine wave output and a transfer time of less than 20 milliseconds to battery during a grid outage according Quattro datasheets.
On one of the essential circuits I have sensitive data communication electronics which were previously fed via a line interactive UPS for further protection. Everything was working fine in terms of no RCD/RCBO trippings etc. However, the 20ms switchover time and the fact that the UPS was line-interactive with estimated sine-wave output was enough to cause data syncing problems on the electronics equipment after a power outage. I therefore purchased an Eaton 9E2000i online double conversion UPS to replace the secondary line-interactive UPS and ensure absolutely no interruption to the comms electronics during grid power outage.
The odd thing is now when grid power is restored after a power outage and the automatic transfer switch on the inverter switches back to mains the RCBO to the Eaton UPS trips after a few seconds. It only happens once and only when grid power is restored not when grid power is lost. Directly afterward when I flick the RCBO switch up it stays on again. So it is definitely a transient phenomenon. It is an isolated circuit and only the UPS is connected to this RCBO. I have tried another RCBO with the same result. There are no wiring issues and it is literally only one cable connected straight from the RCBO to the UPS. Earth is good and bonded to Neutral at the incoming utility distribution box when not via the bonding relay in the inverter. Grid is in TN-CS configuration. Take note the RCBO is downstream from the inverter so there should be no PE-N bonds after the RCBO.
This is obviously a bit of a problem since the whole idea behind upgrading to the Eaton UPS was to establish improved continuity and stability. The line interactive UPS did not trip the RCBO. From what I can gather, when the grid power is restored, the Victron inverter first phase synchronizes its internal inverter AC with the incoming grid AC and then about a second before it transfers to mains it opens the PE-N bonding relay which was activated during grid outage/inverter operation.
So now my question is, for this moment where there is no PE-N bond before the transfer to mains (after which the PE-N bond at the utility distribution box is visible again), how would the Eaton UPS behave? Does it see this as a fault and temporarily create its own PE-N bond which causes the RCBO to see this as an apparent earth leakage and trip? If indeed so, is there a workaround for this other than removing the protection of the RCBO device or downgrading back to the line interactive UPS which was causing data syncing issues?
