Using zero Sequcne CT's on solidly grounded system

[rockman7892]

Are there any issues with using 50/5 zero sequence CT's for ground fault protection of a solidly grounded 4.16kV system? I thought I had heard once that there could be a potential saturation issue with using this CT on a solidly grounded system with lots of ground fault current available?

I know a residual connection will work fine on a solidly grounded system and zero sequence CT's are typically reserved for resistance grounded systems with minimal fault current however I'm looking at a particular situation where zero sequence CT's are already existing in a solidly grounded system.

If there are no issues with using these CT's is there a rule of thumb for setting these on transformer feeder circuits? I believe I have seen a reference which recommended setting the relay for an INST pickup of 10A with a 20-30ms delay when using zero sequence? With residual CT's on transformer circuit I've always used minimum tap setting with INST set to 4x transformer FLA as 0.2 sec.

 

[Kiribanda]

rockman7892,
The answer is already in your question. The main reason is the saturation. The CT 50/5 will get saturated at 20 times=1000A.
Normally on a solidly grounded network the ground fault current is much higher than 1000A. Therefore, the correct operation
of the GF relay during a ground fault is doubtful.

 

[rockman7892]

Kiribanda

I guess I was always under the impression that the magnetic field from the three cables passing through the zsct netted to zero and thus the ct only saw minimum fault current for even high ground faults. But it sounds like the netting of the fields actually occur in the ct iron and thus will saturate ct.

Since new relays are Sel 751 relays I believe I can just use residual ground fault setting with phase cts howver there are existing electromechanical ground relays that that are only looking at
Zsct.

Can residual ground cts work effectively on resistance grounded systems with low ground fault current or must you use zsct on these systems for sensitivity?

 

[rockman7892]

If setting a transformer feeder ground relay with an INST setting, it there a minimum time delay for the ground fault relay for the Main breaker in the same switchgear weather the relay is an electromechanical using 51N only or digitial using 51 and 50 element with a delay?

 

[dpc]

There are a lot of flux summation CTs installed on solidly-grounded systems. While core saturation is a real concern, in most cases, the relay will trip because it is generally a fairly low pickup instantaneous trip. Even when the CT saturates, enough energy is produced to operate the relay. If you are using these CTs on a ground fault relay with a slow inverse time characteristic, then you will likely have a problem.

 

[LionelHutz]

You're all over the place here.

The magnetic currents don't sum to zero by themselves or in the core during a ground fault. With a ground fault, all current for the faulted cable/component does not pass through the CT window since the ground current is returning via ground outside the window.

Using a residual ground connection or using a higher ratio ZSCT can both be used on a solidly grounded system.

The residual ground connection can work on high resistance circuits if the CT ratio is relatively low. If the CT can detect the current fairly accurately then the residual connection has a good chance of working well.

Your last question appears to be about co-ordination. The main basically has to have a time delay longer then the detection + clearing time of a downstream feeder if you want to give the feeder a chance to clear the fault first.

 

[rockman7892]

dpc

Interesting point about a flux summation likely working accurately when an Instantaneous setting is used. What would you consider a "fairly low" pickup for an Instantaneous?

I have seen systems that use flux summation CT's throughout the plant including main breaker which only have an inverse time characteristic and no Instantaneous. It sounds like these mains will be a problem but most feeder circuits with instantaneous settings will likely still be able to detect the ground fault?

 

[dpc]

I didn't say it would be highly accurate - but it will still likely trip if instantaneous. For time overcurrent, using residual CT connection would be a better solution (for E-M relays). For digital relays, let the relay calculate the ground current from the phase currents.