Differential protection operation on through faults

[Bung]

We have had a number of spurious operations of line differential protection at 33kV on through-faults. The general situation is as follows. This has happened at a number of different points in the system.
System voltage: 33kV

Protection: HO2, HO4 "Translay" schemes. Setting minimum (0.4A), on 500/1 or 1000/1 CTs (and yes, I know it could be considered a bit low). Backup / #2 protection is IDMT overcurrent and earthfault. The CTs are a higher class than minimum suggested by relay manufacturer.

Pilot wire: Overhead "hardex", mixture of 4 and 5 core. Hardex is a steel earth wire laid around an internal pilot wire. The cores are not twisted, they are laid straight with PVC insulation. In appearance it is like a very heavily steel wire armoured PVC cable. The earth wire is not brought into the substation at the ends, but is earthed at each pole, because the fault current rating of the wire is below the fault levels.

Primary system: Two feeders from the source substation to the load substation, each with Translay protection. Feeder length typically 5-10km, loads 500-700A maximum per feeder. The feeders have lengths of 1-2km running in parallel, usually either side of the road (say 20m separation). Fault levels typically 6-10kA.

Scenario: Fault on one of the feeders results in (correct) trip at both ends by the diff protection. Second feeder trips at source end.

Anybody got any ideas? I don't even have any suggestions intelligent enough to use to "seed" any discussion!

Bung
Life is non-linear...

 

[electricpete]

I can tell from your problem description you're much more familiar with this stuff from me. But since you ask I'll throw in my uninformed two cents.

As you mention the 2nd feeder experiences a through fault when feeding a fault on the first feeder.

I have seen on a pilot wire scheme on generator stepup transformer to switchyard line where the system operated fine at 100% power, but tripped on a close-in fault in the transmission system fed by the generator (through fault for this zone). The cause was eventually found to be a ground in the ct. Since the ct circuit was already grounded at another point this effectively shorted out some turns and changed the ratio slightly. This caused a differential current sensed during normal operation but not enough to trip. Only during unbalanced through-fault did it trip (relay was more sensitive to unbalanced differential currents I0 component than to I1 component).

I'm sure there are a number of other things that could go wrong with the CT's, so it may be worth doing a megger test and secondary excitation test.

Also of course the settings and wiring of the relays should be checked carefully. I remember one engineer used to swear by the in-service load-angle tests to verify expected ct secondary current magnitude and phase angle at various points.

 

[busbar]

Stuff you’ve probably already covered:
· Is the pilot wire shared by only two relays? Some Alstom pilot-wire devices seem to be applicable to multiple terminals.
· Is the "hardex" cable used for more than the pilot circuit? [as one pair?]
· Are unused pilot-cable conductors floating, terminated or grounded?
· Has the armored cable's insulation resistance been tested? {The potential difference from station grid to shield ground may at times be significant.}
· Have the relay pairs been bench tested?
· Are the CT characteristics suitably matched ?

Regardless of their lack of technical comprehension, such events can make office-bound administrators unduly excited. The advantage of this may be some loosening of purse strings when it comes to overtime, test equipment enhancements or spares inventory.

 

[Bung]

The problems have only been on two-ended feeders, as far as I know. All the HOA4 (three ended) schemes seem okay (I hope).

The unused cores are brought into the pilot cubicle, but are left floating. We use 15kV isolation transformers on the cores used. No other services of any type run on the other cores in the pilot.

We do a complete in-service check on all protection after any 'unexplained' or unusual operation, and the schemes appeared to work normally. But a full 15kV pressure test may not have been done in this instance - some (not all) previous faults have been traced to breakdowns in pilot junction boxes on the pole top.

Equipment enhancement would probably be easy to justify, but until I know which bit (pilots or the relays) I'm a bit reluctant - been there, done that, and found I'd replaced the wrong thing!

BTW, some have an HO2 at one end and an HO4 at the other, which Alstom assured my predecessor would work fine.
Bung
Life is non-linear...