Sensitive earth fault detection for motors

[AMBMI]

I should set in a industrial plant the 50N elements protecting many motors ranging from 300 kW to 4000 MW.
The system voltage is 11 kV and the ground current is limited to 600 A. The customer decided to use 50/1 cl 1 measurement (!)toroidal CT to feed the relay ground elemnts.
How sensitive can be set the 50N threshold?
What I would propose is 10% of the tore CT ratio (so 5 A) with a 100 ms delay to avoid spurious trips due to the cable capacitive currents. I wonder if a smaller threshold could be used. What's the limiting factor?
With a 5 Amps threshold I think we should already protect 99% of the motor winding.

 

[slavag]

Hey. What's going???
WE do now project with 11kV motors up to 6MW and designer used ZS CT 50/1A and E/F fault is limited to 300A ( OK not 600A).
Is not correct, you take our project.
We ask now same Q's, for avoid unwanted trips during motor start-ups, cable capasitives, etc, I decide add function 67N for small settings and 50N for higher setting value.
In all cses ( as you remember) I request change ZS CT to 100/1A , BTW cl 10P10 is good too.
Regards.
Slava

 

[AMBMI]

Hi Slava!
You have been lucky, I asked the customer but no way for me to change the toroidal CT type... so we have measurement CT which probably are saturating at around 65 A.
Anyway with few meters of cables between the CT and the motor I don't think that the capacitive current feeding a fault on another branch could be too big (100 m of cables I = 11/5 * 0.1 = 0.22 A). 67N is a good idea but it requires a VT, no way for me in this project.

 

[slavag]

Hi AMBMI.
It's crazy, I see all customers are same, around the world.
But we also learn, what possible do against customers.
In this case, I sended to customer graph with toroid CT performance and second, price of 50/1A meas. core CT vs 100/1A 10P10 and explained to customer, that relay internal CT's are protection class and no needed meas class of toroid CT.
100m isn't long cable, we have about 400-500m.

"67N is a good idea but it requires a VT, no way for me in this project."
You meant broken-delta connection VT or busbar VT?
for big motors we coiced relays with option of internal broken delta calculation and phase-neutral VT we need for the undervoltage 27 function.
Regards.
Slava

 

[AMBMI]

with 400-500 m you should have about 1 A max of capacitive current. So in this case too a 5 A primary current 50N threshold should be fine.

VT: I have the VT only on the incomings, no 27 element for each motor, so no phase-neutral VT to detect the I0x3 direction.
Please could you put on the forum the toroid CT performance graph? I remember you had something similar also for the measurement CT. Am I wrong?

 

[slavag]

HI AMBMI.
It's same graph for toroid CT performance. Attached before few days.
"with 400-500 m you should have about 1 A max of capacitive current. So in this case too a 5 A primary current 50N threshold should be fine."
You are right, but I know the designer, main contractor and customer of this project and choiced type of motor protection relay with all possible functionality.
Is more expensive on first time, but I don't worry about commissioning, erection of system and future troubles.
Regards.
Slava