OPERATION OF GENERATOR DIFFERENTIAL DURING LIGHTNING

[appunni]

SIRS,
I HAVE EXPERIENCED OFTEN THE MALOPERATION OF GENERATOR DIFFERENTIAL RELAYS USED FOR PROTECTING 130MW GENERATORS DURING EXTERNAL LIGHTNING FAULTS ON THE HV SIDE OF THE GENERATOR TRANSFORMER , THAT IS , CLOSE LIGHTNING FAULTS IN THE OUTGOING FEEDERS. THE RELAYS ARE OF ABB MAKE AND RADSG TYPE.
WE HAVE CONDUCTED ALL POSSIBLE TESTS ON CTS ,AND RELAY. BUT THESE RELAYS SHOWED PERFECT OPERATION BOTH DURING TESTING AND DURING ON ALL OTHER EXTERNAL FAULTS. THIS MALOPERATION HAPPENS ONLY DURING CLOSE LIGHTING EXTERNAL FAULTS. I REQUEST YOUR VALUABLE REMARKS. THANKS...

 

[busbar]

Would poster please end repeated typing of Eng-Tips postings in all caps?

What tests/measurements have been conducted to determine that subject relays are not operated outside of; e.g., IEEE C37.90x levels? It seems unrealistic to expect any degree of shielding and surge suppression to increase security/prevent false trips during close-by lightning.

 

[peterb]

Confirm that the tests performed include the following -
1. The CT secondary wiring is earthed at one location only (lift the single intended earth connection and megger wiring)
2. CT excitation curves have been checked to ensure that the relays are connected to the correct relaying CT cores and not to metering cores

 

[electricpete]

I like Peterb's comment. We had a pilot wire differential trip on a close-in fault outside the zone once. It turns out that we had a ground in a transformer ct junction box... created a 2nd ground in ct circuit (first one at ct neutral) which effictively changed the ct ratio. The error current created was insufficient to cause any problem during normal operation but increased to trip levels during a close-in fault. CT ratio did not find it.

If your circuit is grounded at the neutral then lift that one neutral ground and megger the entire circuit (three ct's relays) at 250VDC in a single test. (250vdc shouldn't hurt even electronic relays). Expect more than 1 megaohms. See ANSI/IEEE standard on instrument transformer testing.

There is another aspect. During a through fault the ct's may go into saturation and if they saturate differently due to the fault current including dc offset, that causes a problem. One thing to check would be a saturation test on the ct's.

 

[pwrengrds]

Check to see if the shield grounding goes through the CT's. During lightning there are ground currents, if the shield is grounded at both ends and the shielding goes through the CT it will cause a fault. Look at the shielding, if the shielding is after the CT it must be routed back through the CT to prevent faults. I have seen this a couple of times before with simlar results.

 

[appunni]

Sirs,
Please excuse me for typing caps all times. I do not do it now on.
I did not understand what Busbar meant by IEEE C37.90X level. I had checked the CTs for verifying ratio and magnetising characterestics and found OK. I conducted stability test of CT circuits and relay by primary injection. That also was OK. I had checked for one point earthing in CT circuits and found OK.
I did not understand what Pwrengrds meant by ground currents due to lightning.
Please note that the relay does not operate for external faults other than lightning. Even for lightning, this happens but rarely.

 

[pwrengrds]

During lightning storms currents are traveling through the ground (earth) to disipate the lightning strike. If your shields for the high voltage cable are grounded (which is typical) the lightning can use the shield as a path for currents. If the ends of the shield go through the CT and then are grounded, the differential relay will see the shield currents as a differential fault and take out the system. It is very easy to have the shield going through the CT. The stress cone of the cable may be made up after the CT, in which case the shield must be routed back through the CT before it is grounded to stop the CT from seeing the shield currents.

 

[appunni]

sir,
Thanks for your valuable replies.

 

[SMB1]

Does the Diff relay operate during the fault or immeditly after the fault is cleared?.
Does it operate 3ph or 1ph only?
Do you have any disturbance recorder output for such an incident?

 

[busbar]

apunni, apologies for slow response to your question to me of 10 March.

Do the misoperating relays have surge-withstand capability defined in ANSI/IEEE C37.90 Relays and Relay Systems Associated with Electric Power Apparatus, and IEEE C37.90.1 Surge Withstand Capability (SWC) Tests for Protective Relays and Relay Systems?

 

[jbartos]

Suggestion: Please, expand the original posting in terms of lightning protection, e.g. arresters, etc.
In case of the direct lightning strike, the differential relay might be the one that trips. Also, there are various differential relay schemes possible to protect the generator. Which one is it? 87G?

 

[appunni]

To Mr. SMB1:
It is not sure which happened first. The feeder and the generator tripped almost simultaneously. The indication in Generator Differential relay was on only one phase. We have no sequence events recorder now.
To Mr.BUSBAR:
We have not conducted such tests. Is it possible for doing the same in the field by us?

To Mr.jbartos:
The generator differential relay 87G was found tripping on lightning. But the overall differential relay which covers generator and generator transformer was not tripping. The gen.diff.relay was of ABB make and RADSG type and the overall diff.relay was of ABB make and RADSB type.

 

[busbar]

On testing...C37.90 is typically a series of manufacturer’s type test, and claimed as such by the manufacturer. It's essentially standard boilerplate for many relays.