RCCB Tripping After Change of Supply

The only reason for tripping is that the vector sum of the currents passing through the device exceeds its trip setting.

The way the installation is wired cannot be seen from here. That's the trouble with <...from the internet...>. So, engage a qualified local Electrician to sort it out, perhaps?

It would help if we knew where you are, only then can we decide if your installation falls under IEC (European) or NEC (American) regulations /codes.

Your reference to RCCD would indicate IEC.

We? You can do that on your own, mate.

Yes .This will trip.RCCB works on the principle of Ia+Ib+Ic+In=0.

During this period it is not 0, therefore RCCB will trip.

Solution would be install 4 pole changeover switch instead of 2 pole change over switch.

This means during power failure , no power will be available from mains and only DG supply will be available in phase and neutral and no voltage will be available in other two phases.

I'm going to take a wild guess, but I might not even be in the right ball-park on this.

Assumptions: 1.You're British and by "flat", you mean apartment/home. 2. "only one phase circuit works on DG" means that only 1-out-of-three works, meaning that the diesel is also 3-phase, but is using only one. 3. That your flat actually uses 3-phase power, but when it goes to diesel power, it goes to single-phase power/loads. 4. When the main power returns, there are no 3-phase loads restarting, and the diesel still powers the single phase loads as when the main was off. And it has to be manually turned off, and is not auto-off. That means that if there was no intervention, the diesel would still be carrying it's load until it either runs out of fuel or someone turns it off. I make this assumption based on the fact that if there were any 3-phase loads restarting and one of the phases is now on the diesel, then you would have an out-of-phase 3rd-phase (get it?). The only other way to avoid that would be to disconnect the diesel before the main was restored, and you never mentioned anything to support THAT assumption. 5. The RCCB 40A is now sending 2 phases to the flat and the 3rd phase to that xfer box, which means that you now have only single phase loads running on two different phased circuits (since the diesel is now shut down). What the diesel was powering, is now dead, and anything that is running are different loads on different phases.

From these assumptions, I'm guessing that at least one phase of that RCCB 40A breaker is carrying a significant load. And when you re-energize that 3rd phase, it puts an additional "starting current" on one of those other lines sending it over the trip threshold. Now when you go back to reset the breaker, you are now energizing all 3 phases at the same time, and the starting current stays below the threshold.

My guess is India or nearby!!

I'm thinking the RCCB in question, is a 3Φ unit. And effectively you have two supplies in your DB. 1. Being essential supply (i.e. lighting and low power appliances), that is feed either from 1Φ of the 3Φ RCCB or the DG, via the manual change over switch (in the flat/apartment). & 2. The non essential supply for the higher power appliances (i.e. Air-Con), that is feed from the other 2Φ of the RCCB, but doesn't go via the manual change over switch.

If this is the case, when the mains supply is restored, the non essential supply is available in DB. So 2Φ of the RCCB are energized and feed current to the connected appliances via the DB. Now when you switch the manual change over switch from DG to mains, you are allowing current to flow in the third phase(Φ) of the RCCB. Now my feeling is that this supply uses the neutral (where as the other supply does not) and, the sudden surge could very well cause a momentary imbalance on the neutral (30mA probably) to cause it to trip. The RCCB probably only monitors 1Φ & Neutral, that is your essential supply. Can you identify an appliance(s) that is(are) connect to this supply, which is(are) not suitable for connection to a 30mA RCCB protected feed (i.e. something with high earth leakage currents)

This installation is not ideal, as the RCCB/RCD should be in your DB. In this case you could discriminate the supplies to the appliances with high earth leakage currents.

<...identify an appliance(s) that is(are) connect to this supply, which is(are) not suitable for connection to a 30mA RCCB protected feed (i.e. something with high earth leakage currents)...>

Were one such appliance to be connected, the RCCB would trip on whatever supply it was taking the power from, and not just at the instant of changeover. Anyway, such an appliance is known as "faulty" and "dangerous", and is an ideal candidate for condemnation, removal and recycling under the WEEE Directive.

Your load is a transformer, which is inductive. You might try isolating all loads from the transformer before switch over....

I am assuming that you are not switching over "live", which will definitely cause a trip.....as the DG and the mains are most unlikely to ever be "in phase" with one another....

There is that pesky "phase" bug again, I will be glad when we are all powered by light. At least that has only about 4 kinds of polarization, that I can remember.

User could also try and keep his same equipment, but make sure the DG is shut down, and all load from circuits disconnected before starting over on the mains. There appears to be a ground fault, since ΣI =0 is not a wish or a prerogative. It is a mandate for that RCCB system.