Problem in substation DC system

[Nayeem1993]

Hi everyone.
We have 2 separate 110V DC system with separate FCBC charger and Ni-Cd battery.
system-01 powers 400kV main-01 protection with control(Close coil and Trip coil-01 of CB). System-02 powers main-02 protection(operates Trip coil-02 of same CB).
Recently we are facing a problem where if both chargers are running, then one of them automatically increases it's voltage at battery terminals(upto 154V)- even if the charger is kept in float mode. If only one charger is running then this problem does not happen. To check for any kind of unwanted interconnection between them, we disconnected one charger from system and connected a dummy load. The problem remained same.

What might be the reason? Did anyone face similiar problem?
Both battery and chargers are from HBL, India.

 

[crshears]

Likely an inadvertent parallel between systems 1 and 2, causing the two chargers to fight each other; one way to track down where the tie path is is by the sectionalizing method.

If memory serves, the way forward is to create a switching and checking sequence that progresses through all elements of the combined system, beginning with the least impactful and ending with the most impactful. Having [ presumably ] fully duplicate schemes means that these checks can be performed with a loss of redundancy only, hopefully not having to fully interrupt any single protection scheme at all.

Find a way to place one of the chargers on manual to temporarily preclude the tripping off problem, then place the two systems in parallel.

Interrupt the supply to one protection at a time, and confirm there is no DC present beyond the supplying breaker, then restore and try the next path. Presence of DC even with that breaker open will verify that back potting of that scheme is occurring from the companion supply.

Once the area has been determined, develop a procedure to sectionalize within that branch; this may well involve having to remove a wire, confirm no back potting, and re-terminate wires one at a time until the paralleling path has been found.