Spurious differential trips of power transformer on energization

[bradh01]

Although very infrequent, I have experienced spurious tripping of power tranformers upon energization via their differential relays (restrained trip). These are generator step-up transformers (80-90MVA, 1998-2000 vintage) at natural gas fired generation facilities. The differential relays are generally only "looking at" the transformers themselves. In each case, after appropriate checks and analysis, the transformers were successfully re-energized upon a second energization. Some of the transformers have microprocessor relays affording oscillography, some don't. I am aware that the energizing current can vary widely based on when the transformer is energized and de-energized (during voltage cycle).

Has anyone else experienced this? After some investigation, I am thinking that the problem may be that the relay setting that inhibits tripping during energization (triggered by % 2nd harmonic current as compared to fundamental) may be set too high. Unfortunately some of the differential relays in question have this value factory set to 18% with no adjustment. Others are set at 15%. From what I have heard, some of the newer tranformers have much smaller 2nd harmonic content on energization, than older transformers. Any comments or suggestions would be greatly appreciated.

 

[RalphChristie]

Inrush currents.

Do your relays use harmonic restrain or actually harmonic blocking? And have you checked your other harmonic currents? (4th harmonics or 5th harmonics (overexcitation))

[red]Failure seldom stops us, it is the fear for failure that stops us - Jack Lemmon[/red]

Make the best use of Eng-Tips.com
Read the Site Policies at FAQ731-376

 

[dpc]

It does happen, but the harmonic restraint and/or blocking is intended to make this much less likely.

What type of relays are involved?

On some relays, you can just flat block the differential function for a programmable time delay after the breaker is closed. This obviously degrades the protection somewhat, but it would solve the problem. 5 or 6 cycles is generally adequate.