Redundant feeder for 220VAC network 1

Please, provide more info:
1. Load size in kW/HP
2. One phase versus three-phase
3. Average length of feeders
4. Feeder cable type/structure
5. Grounding

Dear Mr.jbartos

In response to your queries :
1) Total load connected 5470 watts, 31.1 A, 0.80 PF for PLC, DCS, telecoms console/multiplexer cabinet , etc..)
2) Single phase 220V, 2 wire, 50Hz
3) Approx. 140 meters.
4) Cable construction is 2 cores- copper / XLPE/PVC/ Double galvanized steel armor/PVC, rated 0.6/1 kV.
5) IT system (Isolated neutral), Two incoming feeders ( one normal cable and one redundant cable) to sub. Distr. Panel are powered from the same source ( UPS distr. Board).
UPS stands for Uninterruptible Power supply

Best regards, omime

Suggestions:
1. Verify that the phase A conductor in cable 1 is connected to the phase A conductor of cable 2 at its destination terminal.
2. If one is correct, then 300mA may be increased to higher values untill the circuit breaker does not trip, since the common mode currents may increase significantly in the double galvanized steel armor/PVC cable construction; tacitly assumed that the cable armor is grounded to the equipment ground.
3. The load may be checked for total and single harmonic voltage and current distortions. If excessive, above IEEE 519 and IEC 555 limits, they may be mitigated, e.g. by filtering.
4. Interchange the circuit breakers to make certain that they work properly.

Dear Mr.jbartos

Each feeder from the same power source (UPS distr.) consists of 2-conductor( phase/neutral), armor type cable supplying to sub.dis pl. Phase and Neutral conductors of cable 1 connected together with phase and Neutral conductors of cable 2 at the busbar mounted in this panel.

Is it possible cause of unbalance current occurred when closing both feeder breakers (equipped with 300mA RCD), load current drawn from phase 1 return to neutral 2 in closed loop?

Could you give me more explanation " common mode current" in detail.

Thanks & regards. omime

It appears that the most accurate way to determine the current and voltage distributions in those circuits will be by modeling them, using a suitable software. Yes, the unbalanced nonlinear load could marginally cause the current to return through neutral 2; however, that may or may not be so significant to cause the circuit breaker to trip.
Common mode currents are electrical currents, fundamental and higher frequencies (harmonics, interharmonics), which flow in the same direction to the load (apart from the differential mode currents that flow in the phase conductors to the load and in the neutral to the source). It means they flow in the phase conductor and the neutral conductor to the load. Especially, high frequency currents contribute to common mode currents. This is due to the capacitances, including distributed capacitances along cables, that become more conductive for higher frequency currents (higher harmonics). These common mode currents may cause the circuit breaker ground fault protection to trip since they may cause the ground fault current (residual current leakage) exceed 300mA way up to amperes. A clamp-on amperemeter shall be able to measure those, if you pass the phase and neutral conductors through the clamp-on amperemeter opening. Assuming that your cables are of reasonably good quality, the amperemeter reading shall be the minimal total RMS current value, the circuit breaker ground protection shall be set for. It may be prudent to add some design margin, depending how much of load can be added to the fully rated circuits.

It sounds as though your problem may indeed be an unbalanced return in one of the cables. Are both cables the same and the same length? Differences in circuit impedance could contribute to the unbalance trip.
One point to think about is that by installing the cables in the way that you describe, you are probably not going to achieve the desired redundancy. The reason for saying this is that if one cable develops a fault, the MCB on that cable will trip, but the fault will still be connected to the other circuit through the parallel cable, so the second MCB will trip as well.
I suggest that you think about installing a transfer switch at the sub-distribution panel (high speed 2-4 mS static transfer switches are available), with each cable feeding one side of the switch. This way, faults on one feeder cable are isolated from the second cable and you will get the redudancy required. Depending on the criticality of the loads, you may want to think about a redundant UPS to supply the second cable.
Also, check your loads to see whether they can ride through the voltage sag that will occur during a cable fault - the static transfer switch can help in this situation if it is supplied from an independent source that isn't subject to the sag.