Setting transfomer over current for busbar backup

[Mbrooke]

Ok, the perfected: In the event local bus bar differential is out of service; a fault on a bus bar section is cleared via a 10 cycle bi-directional delayed zone 1 on the bus coupler set at 10% impedance of the shortest line, and transmission lines attached to the faulted bus clears via either a 35 cycle remote zone 2 or a 15 cycle reverse local zone 3 (reaching into the bus). This works well and achieves selective coordination in a last resort effort- or older stations that do not yet have dedicated bus bar protection but are being retrofitted with numerical relays.


However, does anyone have any tips on setting the transformer over current? Would you use impedance reach or directional overcurrent and to what magnitude? Or just let everything clear as it likes?


If possible, I'd like to have transformers back-feeding the faulted 230kv bus section clear, while transformers pulling current off the unfaulted bus(es) remain in service even though they feed the same 69kv bus thats reverse feeding the transformers still attached to the faulted buss section.

 

[davidbeach]

Directional overcurrent on transformer relays is fairly recent. Because it's the transformer relay, the directional overcurrent will tend to indicate forward as toward the transformer, so you want to use reverse directional overcurrents. Distance elements would be even nicer, but they're less common than directional overcurrent. Ideally you'd have the transformer trip before any of the lines on the unfaulted side; let each voltage level sort out its own problems.

 

[Mbrooke]

Got it, and I would set the reverse (backword) lower than the forward in order to allow the backfeeding units to open instead of the foward feeding units.


Regarding the voltages needed to do this, would you advise deriving the required voltage from the bus side?

 

[mgtrp]

I'm having to make a few suppositions as to what a "typical" double bus substation looks like in your part of the world, but a couple of questions:
- Are you comfortable with only 5 cycles coordination time between the bus coupler and the reverse-looking zone 3 elements for lines on the healthy bus?
- Are there additional sources of fault current on the 69 kV bus other than the transformers on the healthy bus? If not, you will need to take some care in coordinating between the forward elements on transformers connected to the healthy bus and the reverse elements on transformers connected to the faulted bus, especially as fault current through this path may be very small until your bus-coupler opens.

Cheers,
mgtrp

 

[Mbrooke]

- Are you comfortable with only 5 cycles coordination time between the bus coupler and the reverse-looking zone 3 elements for lines on the healthy bus?

With 3 cycle SF6 breakers and high speed numerical relays IMHO the extra time once reserved for coordination is not needed. Granted I might add 2 or 3 more cycles to those 5 to take care of a worst case breaker opening. Generally a 2 cycle breaker is assumed to have a 3.5 cycle worst case, and a 3 cycle breaker having a 5 cycle worst case.

- Are there additional sources of fault current on the 69 kV bus other than the transformers on the healthy bus? If not, you will need to take some care in coordinating between the forward elements on transformers connected to the healthy bus and the reverse elements on transformers connected to the faulted bus, especially as fault current through this path may be very small until your bus-coupler opens.

Most likely not, I am glad you brought this up.

My plan is having the reverse protection faster than the forward protection. This will allow the reverse transformer to clear first as that is most likely to be the faulted bus its attached to.

..............

One special case I am particularly evaluating in that regard are faults on a 69kv bus section with load stations tied together at the ends. The 69kv coupler will clear, but in theory the faulted bus will be fed directly via TX1 and indirectly via TX2 through the secondary 69kv system. Remote substations near this station in question will probably be operated split (normally open) so enough impedance is present on the secondary system to allow more fault current to pass via TX1 clearing that first.