Using residual CT connection on LRG system for ground fault detection 2

[rockman7892]

Can a residual CT connection or calculation in a relay be used to successfully detect and coordinate ground fault current on a MV LRG system? System is 2.4kV with resistance grounded secondary at 400A with a CT on the transformer neutral connected to relay in secondary switchgear main.

The feeders in this secondary 2.4kV Switchgear has 1200:5 phase CT's and does not have zero sequence CT's. I have always understood that zero sequence CT's were required for feeders on LRG systems in order for sensitive ground fault sensing (20A or so) Can the same level of sensitivity be achieved with the phase CT wired for a residual measurement or calculation?

With older electromechanical relays I know that you could not set the pickup low enough where high ratios were used for a residual connection (I believe .1 was lowest pickup settings for most electromechanical relays)however with the advent of electronic relays is this an issue anymore since you can set pickup as low as .02 in some relays?

On the contrary I know that you do not want to use zero sequence CT's on LRG systems due to saturation concerns.

Does the size of the CT on the transformer neutral typically match LRG rating i.e. 400A resistor uses 400:5 CT?

 

[krisys]

It all depends on what is the limiting earth fault current and what is the feeder CT ratio.

In your case, when the feeder CT ratio is 1200/1A, with the a typical relay (new or old) the minimum setting possible for the earth fault with the residual connection is 0.1 * 1200 = 120A. May be slightly higher side for the feeders.

If you have numerical relay, with the 0.05xIn setting is possible, the minimum possible setting for the earth fault is 0.05 * 1200 = 60A. You can live with. But not a sensitive setting.

If you have a smaller CT ratio like 400/1A with a numerical relay, 0.05 * 400 = 20A. You get a sensitive setting. So you have a room for having a good selectivity with the upstream relay.

Hope you got the gist of the above analysis.

 

[rockman7892]

Krisys

Thank you. Your explanation makes sense. What are you calling a “typical” relay vs a “numerical relay”?

Is .05 typically the lowest setting avaliable? I thought I have seen some that allow setting down to .02.

What is a typical pickup setting on a feeder when coordinating with main on LRG? On 400A LRG I usually see main set at 400A and feeders set at 10-20A pickup with .2a delay? Is the 60A setting you mentioned with 1200:5 CT not sensitive enough to coordinate with main?

 

[LionelHutz]

Do you need to try and detect 20A? With the ground fault current limited to 400A, a ground fault will likely cause 400A of current to flow. I'd expect you will co-ordinate via time delays, not via staggered pickup levels.

I'd also hope that the main is set to trip at a ground fault current which is less than 400A when the ground fault current is limited to 400A. Otherwise, how do you know that tolerances won't cause a failure to trip?

 

[dpc]

CT rating on neutral of LRG system is generally sized at 50% of the maximum ground fault current. So for 400 A NGR, I'd specify a 200/5 CT.

Check relay for lowest available pickup settings. This varies by manufacturer even for digital relays. Settings for the feeders depend on what is downstream. Since it is often feeding the delta primary windings of transformers, the feeder ground pickup can often be set quite low. The flux summation CTs are more sensitive than a residual calculation. But no reason the residual approach won't work OK. Combine a sensitive inverse time element with an instantaneous trip set higher.

 

[rockman7892]

LionelHutz - No specific reason or requirement to detect 20A. I guess that's just what I've seen in the past and automatically defaulted to that. Are you suggesting that the feeders can have the same pickups as the mains and just coordinate using appropriate time delays? For example...both main and feeders set to 100A with different time delays? I was not suggesting setting pickup at 400A for 400A system but rather was using that as a reference. Is here a rule of thumb for setting pickup value for main on an LRG system?

dpc - Your response may answer my last question hat I posed to Lionel Hutz above. So using a 200/5 CT on a 400A NGR would you typically set the main with a pickup of 200A? Or would you set lower than CT value? So it sounds like if I have a low enough available setting in the relay to give me the desired pickup than a residual connection is ok? I usually se the flux summation CT's using a definite pickup and time delay but you are suggesting an inverse with an instantaneous pickup? Does this pickup usually have a delay? I've seen some suggestions for transformers to set INST at 4x the transformer FLA rating?

So if the residual connection will work in many cases is the flux summation CT becoming a relic of older electromechanical relays? Are there certain applications that necessitate the use of flux summation CT's?

 

[dpc]

The instantaneous trip has a pickup setting higher than the pickup of the inverse time element. Maybe 100 A. This assumes there is nothing downstream you need to coordinate with. You can include a 1 cycle TD on the IT if you're nervous about nuisance trip.

The goal is always the same: set the protection to be as sensitive as possible while still coordinating and not creating misoperations.

 

[LionelHutz]

Yes, you could set both your main and feeders to 100A as long as the delays co-ordinate. The main delay has to be long enough for the feeder breaker to open. Vacuum breakers often spec 3 cycles to open, which means >3 cycles of time delay is required. I think you already know all of this.

I frequently see attempts at using instantaneous levels only in smaller low voltage systems where the main and feeders are all just typical thermal magnetic breaker. But, a fault trips every breaker in between the incoming and the fault. "Saving" the extra money on an electronic trip main breaker means the whole site goes down for every fault. The mistaken belief is that the instantaneous setting is a current limit.

I also frequently see customers wanting to detect very low levels of ground fault current on solidly grounded systems medium voltage systems. But, there's not much point in having a CT to detect 5A of ground fault current when 1000's of amps of ground fault will actually flow.