3V0 via broken delta connection or numerically derived within a protection relay

[crumble]

I was reading through this old thread trying to find an answer to my question.

http://www.eng-tips.com/viewthread.cfm?qid=216248

I work a protection engineer and I'm doing a bit of reading into the merits of Neutral Voltage Displacement (NVD) or Residual Voltage Protection being based on either numerically derived 3V0 or via a Broken Delta connection.

The system to which I'm applying NVD is a normally resistively earthed network but with the possibility of being back-livened via a delta winding of transformer in contingent situations. The NVD is there to detect and trip.

Ferroresonance can be an issue so loading the 'broken-delta' connection helps. But equally an auxiliary VT with a broken delta can assist here.

Any comment to why a 'broken delta' connection could be superior to numerically derived 3V0?

Regards,

Chris

 

[davidbeach]

With what's available today, I'd be loathe to go with a broken delta when I can calculate it in the relay and know what's going on on a phase-by-phase basis at the same time. The broken delta had its time and place; was a very useful connection in its day, but that was then and this is now. If you want to provide stability against ferroresonance you'd be much better off with a much larger transformer providing that ground reference than you are with a small instrument transformer.

 

[HamburgerHelper]

Wouldn't it be cheaper if ferroresonance was a worry to load down the PT during ferroresonant conditions? IEEE 242 "The Buff book" has recommendations for PT loading to prevent resonance. You can have the load inserted and removed depending on system conditions so you don't have something sitting there all the time consuming 700-1000 watts. The load will help prevent the PT from being saturated due to it being overexcited. 3V0 tripping I believe has more to do with the system not being insulated to line to line voltages. Your lightning arrestors will hopefully conduct before your equipment does if your 3V0 scheme doesn't operate fast enough. The load is moreso to resolve the issue of the PT being put into saturation due to seeing high and or unbalanced voltage. I am with David though, if we have microprocessor relays, why bother with a broken delta. I think you can trust a relay to calculate V0 from the 3 phases.

This is a good paper on unexpected delta sources:

Protection for Unexpected Delta Sources, by Ken Behrendt

 

[crumble]

Thanks for the replies. I work as a consultant for a variety of utilities and the 'broken delta' connection is quite common on legacy systems. Given many numerical relays have four VT inputs is often possible to connect the 'broken delta' to the fourth VT(PT) input.

I hadn't come across the idea of switching the VT loading resistor before, I've normally applied burden resistors in the past as a permanent connection. An advantage with broken delta is the loading only generate heat when there is a voltage imbalance or the frequency on the secondary is not cancelled by the other phases. 99.99% of the time, the voltages will be balanced. That 0.01% of the time is when the remote end CB opens disconnecting the earth reference.

Thanks very much for the paper reference. I had a quick read and it looks worth rereading in detail.

Regards,

Chris