Sensitive ground fault

[vandax]

Hi
Does anyone has experience with application of sensitive ground fault relay in the distribution line.

We just have an experienced that the neutral relay is not detect the fault (broken of 1 phase wire and lay down on the ground) in 13.8 kV distribution line (20 ohm grounded system).

Disturbance recorder indicate that the fault current was not adequate to force the relay to open the breaker.

After checking the relay, we know that the relay has been equipped with sensitive ground fault element, but we don't enough experience to apply it. I would apreciate your input, plus delta of this function.

Regards

 

[RRaghunath]

sensitive ground fault element is useful only if you have sensitive or core balance or zero sequence CT.
The setting can be low as 4A with a long time delay of 5 seconds.
These are widely used in OHL distribution circuits.

 

[johnspark]

Vandex, Faulty is right. With a good sensitive neutral CT, sensitive setting and long time delay, lines on ground can be detected.
Years ago I was in a utility that did tests, including placing 11kV conductors onto bitumen, concrete and rubble to see what fault currents would occur.
They also found currents to be very low.

There have been discussions of this subject in the IEEE power journals.

Also, ~9 years ago I saw a Korean attempt at detecting these types of faults. Can't remember the relay manufacturer, or whether they kept pursuing for a guaranteed method of detecting these faults more reliably.

I understand though, that GE Multilin have a feeder relay that has a special algorithm for detecting conductor on ground type faults.

Probably other relay manufacturers have explored this challenge before also.

cheers and hope these thoughts help.
Johnspark

 

[Bung]

You can do it with CT in residual connection. We do it all the time, at typically 2% (of 400/1 CT) (ie 8A primary pickup), with a 10 second definite time. A delayed reset on the SEF timer is useful, as such low faults tend to be arcing faults - the current starts and stops randomly. An instantaneous reset relay will sometimes never get to time out.

If you run a 4-wire system with phase-ground loads, you may be limited in just how sensitive you can go - specific knowledge of your system and a few calcs would be required.
Bung
Life is non-linear...

 

[busbar]

There was a good suggestion by bigamp on negative-sequence detection. Bigamp, you made a posting in a duplicate thread238-52180 that was removed. Can you repost it?

Downed-line protection has always been a difficult case, for zero-sequence currents may vary widely with seasons and weather. A negative-sequence element can be more sensitive, but don’t overlook load imbalance as another source of negative-sequence current.

It is a bit unusual to see resistance grounding combined with poleline distribution. Of course, accommodate the limited ground-fault current in your calculations.

 

[bigamp]

Hello Busbar,

Ah yes, the vanished answer to a deleted thread. I thought someone found it objectionable and red flagged it because I mentioned a specific make of relay (Alstom). I do not work for Alstom but use their relays a lot and was not intending to "push" their product. I do not have a copy of the post so here goes for a memory job:

You can use negative sequence to detect a broken conductor fault if your feeder is reasonably loaded, because you will get negative sequence current if you have a broken conductor. The Alstom P141 relay provides broken conductor protection by looking at the ratio of negative sequence to positive sequence current. This protection only operates once a threshold current has been exceeded. As you say, broken conductor can be very difficult to detect.

My other suggestion was to connect the earth fault relay to a core balance CT on the feeder, but if the feeders do not have core balance CT's fitted, retrofitting them can be a real mission, easier said than done. You get much nore sensitive and reliable earth fault settings by using a core balance CT than you will ever get with a residual connection.

It would also be interesting to know what relay(s) are currently fitted to the feeder(s) in question and what level of earth fault they are set to, line CT ratios etc.

 

[vandax]

Thanks for kind response and great input.

According to bigamp' input,
Most of our feeders are using the bushing CT, and GE's neutral relay set at 90-120 A as a minimum pickup setting with inverse curve. The relay is also coordinated with the recloser/sectionalizer under respective line.

My further questions, is there a different/special algorithm for sensitive ground relay, instead of "common"neutral relay. If not, it means that we just reduce the tap setting of exsitng relay (min 6 A), and perhaps will disturb our operation with its sensitive fault caused by eg. tree, leak of insulator. it compromises between safety and production loss/down time.

Rgds