coordination ques: CTs & instantaneous overcurrent relays

[GreyGoose]

Quick question:
when determining the setpoint for a instantaneous overcurrent relay for a feeder, what value of fault current is appropriate:
Should the setting be determined based on symmetrical rms, asymmetrical, or peak short circuit (three phase bolted) fault value?
Also, this value, is it the same for all relays or does it depend for each relay?
Any help would be appreciated...

 

[dpc]

This is an interesting question. The best answer, I think, is that it depends on the relay.

Old electro-mechanical instantaneous units respond to asymmetrical current, so to be safe, you need to coordinate these based on the peak asymmetrical rms current that can occur. The ratio of symmetrical to asymmetrical current will be function of the X/R ratio of your system.

For new digital relays, it depends on the relay. For GE/Multilin relays (SR and UR), my understanding, based on conversations with their tech support engineers, is that these relays use software filtering. The dc component (and harmonics) are filtered out and the Overcurrent units respond only to fundamental symmetrical current, rms. I believe the harmonic filtering is user-selectable in the UR relays, but dc offset is always filtered.

Schweitzer (SEL) relays, in the past, have applied hardware harmonic filtering, but NOT dc offset filtering, so these relays do respond to the peak asymmetrical (rms) current. I have seen some promotional literature from SEL regarding new software filtering algorithm, but haven't read it yet.

If in doubt, the best assumption is to assume you need to coordinate at the peak asymmetrical current. I would probably use rms, but some low voltage trip units do respond to peak current. Best bet is to check with the relay manufacturer.

You should also allow a little safety factor for CT error and saturation. In the old days, we assumed 1.6 x symmetrical current (rms) plus a 10% safety factor. DC offset filtering can cut this down quite a bit.

 

[GreyGoose]

thank you. This was the type of response i needed. I suppose that my question is not exactly relay-related but more CT-related, esp with regards to CT saturation. The real problem is, do I need to take into account the peak asymmetrical rms value into account when determining CT saturation (time to & magnitude of), OR can I ignore the DC offset (since it is DC, ir should "pass" through the CT) & just use the symmetical rms value?
I agree with you on the 1.6 x symm +10% for CT error (gives ~1.8, a good number), but I am more worried about actual CT behavior during high short circuit fault duties. Once again, thanks...

 

[dpc]

dc current will saturate any CT, so it must be accounted for. You need to consider saturation due to both ac and dc components.

J. Lewis Blackburn's "Protective Relaying" book has a good discussion this.

Sorry if I answered the wrong question!

 

[SphincterBoy]

Blackburn's last book viewed instantaneous relays as "lightning nuisance" devices, which is what we in thunderstorm country use them for.

Compute the L-L fault current at the end of the longest line, or the smallest amount of L-L fault current available on any of the feeders.

Set the instantaneous trip value for this value, ONLY IF THIS VALUE IS GREATER THAN MINIMUM TRIP. If "end-of-line" fault current falls below MIN. TRIP, you'll have to install a Distance Relay, in leu of the Inst. Trip for long lines.
Distance relays "see" and react only to changes in impendance (line characterisitic) values, not current or voltage.

The instantaneous relays should be accompanied or associated with type 79 (recloser) relays (i.e.--reclose after inst. trip), and the pair should be used on feeders only.

If there is a Main breaker involved, disable the Instantaneous and Reclose relays or function for the Main breaker only. Allow only the feeder breakers to take care of this.