Earth fault protection high impedance grounding 1

[charlierod]

Hello Dear Colleagues,

I have a 4,16 kV industrial switchgear with high impedance grounding through a ZigZag transformer limiting fault current to 50 A.

Unfortunately the MV metalclad switchgear came without zero sequence CTs and some phase CTs have a high ratio 2000/5 limiting both sensitivity and speed as fast trip times cannot be achieved without compromising security under transients such as transformer energizing and motor starting.

I have opted for maintining a good sensitivity (about 10 A primary) at the expense of speed with delays of 1 -2 seconds in order to deal with the errors of residual measurement.

The fault current itself is not dangerous and I think we are still tripping in a reasonable time to avoid a single phase fault evolving into multiphase although as far as I know this is less likely for a 4,16 kV system than for higher voltages.

I appreciatte your comments on the issue and if there is a guide with recommendations for earth fault protection settings in this kind of systems.

Regards,

Charlie

 

[GroovyGuy]

Hi Charlie,
Any chance of retro-fitting zero-sequence CTs into the switchgear? I don't see a retro-fit as a big-deal, plus I really think that with a 50A resistor you might not be able to achieve the desired sensitivity and coordination with only residual protection. Do the existing PRs have zero-sequence inputs, if so all you need to install (per cell) is one CT and two wires. Piece of cake.
I guess that the resistor is non-continuous rated (ie 15 to 30s). That implies that you will have a towering inferno if your residual protection fails, correct?
GG

"I have not failed. I've just found 10,000 ways that won't work." Thomas Alva Edison (1847-1931)

 

[charlierod]

Hi GroovyGuy,

Thanks for your reply. Indeed retrofitting would be the best solution but we are dealing with Oil&Gas industry and managers will not be willing to invest and lose production to improve GF protection. We have to find a solution with what we have in hands.

All protection devices do have a ground CT input. My main concern is not sensitivy, I could go to 10% or 20% of maximum GF current even with 2000 A CTs (the PR allow that sensitivity). My main concern is tripping time. I have to use long delays to cope with large CT errors during transients.

Regards,

Charlie

 

[GroovyGuy]

Hi Charlie,
I too work in the O&G industry, so I realize how tough it is to find money lying around. I think that, as you have suggested, that a trip point of 10A at 1-2s should be adequate.
However, I wouldn't give up on trying to convince management that the zero-sequence CTs are a good thing. What you are trying to achieve may be doable, but I have never seen it done before. Better to take your lumps now, rather that later should it hit the fan.
GG



"I have not failed. I've just found 10,000 ways that won't work." Thomas Alva Edison (1847-1931)