Ground Fault setting guidelines using 50/5 CT's

[rockman7892]

Does anyone have any good guidelines or reference for setting ground fault protection settings in an electronic relay where the relay uses 50/5 zero sequence CT's for ground fault sensing?

The application in particualr I am looking at is a system with a NGR limiting ground fault to 400A. I am looking for a stating point with the feeder breakers feeding downstream transformers since these ground fault settings on the feeder breaker dont have to coordinate with anything downstream. I saw in another post someone recommending a pickup of .2?

Once I have the feeder breakers set I can coordinate the ground fault upstream, but its the starting point with the settings for the feeders that I'm not sure about.

Thanks for the help!

 

[LionelHutz]

If the NGR allows 400A then the ground fault current will be 400A. You set the trip level to some value below 400A so the relay detects it but not so rediculously low you get nuisance trips. Then set the trip times differently between relays to co-ordinate the tripping.

Setting a rediculously low value like 0.2A won't stop 400A of current from flowing when a ground fault occurs.

 

[rockman7892]

Thanks Lionelhutz

I was referring to .2 as a pickup
Setting not a current setting so .2 pickup
With a 50A ct would be 10A pickup.

I understand that 400A will flow regardless
Of setting so I was looking for a setting
Below 400a to detect ground fault

I was thinking of using a pickup of 10A
With a delay of .3s for feeders and then
Coordinate upstream.

 

[zer0sequence]

There are more things to consider before choosing an "arbitrary" pickup level. For starters, hopefully this transformer does not feed any single phase load or load that is unbalanced in any way?
Your NGR is rated for 400 Amps, but only momentarily-- ask the manufacturer for a damage curve of the resistor. You will find that the actual continuous current rating of your resistor is much much much lower than 400 amps. This is a good place to start. Ideally, you can come up with the damage curve of the resistor, then you can be positive that you are setting your relay correctly.

 

[vandal06]

I've used settings down to 10A before on a 50:5 CT with success. As mentioned you want to make sure this is above any expected loading imbalance.

 

[rockman7892]

Thanks for all of the responses.

As zer0sequence recommended it sounds like the starting point is to set the CT/Relay on the NGR below the NGR damage point. This damage point for most resistors is 400A for 10s and 40A continuous operation. So it seems like the NGR resistor relay should be set with a pickup of 40A and the time dial set to be below the damge point at 10s. After setting the NGR relay the main breaker relay should be set somewhere below the NGR resistor.

Once you have the above 2 relays set I can then go about setting my feeder relays to make sure they coordinate with the above relays. It looks like they will have a pickup somewhere below 40A based on the NGR pickup. From some other references I've seen it was recommended to set the feeder relays a 10% of the NGR value which in this case would be 40A however for coordination it would need to be set somewhere below 40A such as maybe 25 or 30A.

Does my above approach sound correct for setting ground relays for this NGR application.

A couple of other questions:

1) It was mentioned above tht relay should be set above expected loading imbalance. I didn't think a zero sequence CT would detect any loading imbalance since the CT would cancal out any impalance with all three phases passing through it.

2) If this wansnt an application with an NGR how does the approach for settings these relays change? Are there recommended min's and max's for main and feeder ground fault relays in a case where the system is solidly grounded?

3) For feeder relays I've always understood that the time dial should be set to make sure the relay curve is above .1s for a moderatly inversive curve to avoid nusiance tripping from transformer inrush. Is this a typical rule of thumb?


Are there any references out there that talk about ground fault relay settings and coordination.


Thanks for the help!

 

[LionelHutz]

You are posting 40A for the main and something lower for the branch to ensure co-ordination. Once again, the resistor allows 400A of current to flow regardless of the levels set on each relay or breaker. Setting different trip levels in the system does nothing to provide co-ordination. You need to co-ordinate the time settings to co-ordinate the system.

 

[norcalnewb]

I think it is more proper to say the grounding resistor allows a maximum of 400A of ground current to flow. Any impedance in the fault will lower the current to some level below 400A.

Rockman, your basic approach sounds like it is on track. I would say that in general, you need to make sure that coordination is a very application specific task and rules of thumb don't apply to all situations. As Lionel points out, you definitely need to pay attention to your time dial settings in addition to your pick up values for the feeders to make sure you have sufficient gap between your curves to allow downstream devices to clear before the main. Also, do the main relay and NGR relay trip the same device? If so, then you can stack these curves on top of each other to help allow more margin for coordination downstream devices.

For your other quesitons:
1) Load imbalance will flow back through the NGR, so you will see some currrent if you have significant imbalance.
2) The basic approach would not necessarily change, but you would typically allow for higher values on trip points. This is one of those cases where it is application specific, so you would have to run a study to find your max and min ground currents and then look at the capabilities of the equipment you have to protect from damage.
3) Once again, this is a bit of case by case application. You can get around transformer inrush by adjusting time delay or pick up values. An instantaneous trip will not trip on inrush if it is set high enough.

Jim

 

[vandal06]

Rockman, for bullet point #1 - is the neutral of the feeder cable brought back separately or is it concentric?