Ground Fault 50N Nuisance Trip Dring Starting 1

[CoolB00Yz]

Hey Guys.
We are commissioning 2 4000HP motors via 4KV controlgear. The issue we are facing is that both motors keep tripping during commissioning performing motor solo run. They are both tripping on 50N Ground Fault via SEL-710 relay. Ground fault current detected via CBCT 50:5. The setting was originally 10 Amps at 50ms and changed to 10 Amps at 100ms but still tripped. FYI, both motors FAT was done successfully. Power cables and motors were meggered, and winding balance.
What can we do further to proceed and perform motor solo run? It has been tiring 5 days.
Appreciate your input.

 

[davidbeach]

Assuming the CBCT is around shielded cable. Do the shields come back through the CT after they're separated from the conductor at the termination?

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[CoolB00Yz]

Hi davidbeach - NO as shields were removed prior going into CBCT. Only the conductors passed through CBCT.

 

[dpc]

How much fault current does the relay have in its fault data? And do you see that same amount of ground current if you sum the phase currents in the fault report, if available?
Are the trips occurring on starting or while running?

Some things you've probably checked, but just in case.....

CT ratio programmed correctly in the relay?

SEL terminology re "ground" vs "neutral" is not consistent with other relay manufacturers.

 

[CoolB00Yz]

Hi dpc, current varies from 10A to 30A in CBCT.
If sum all phases, it doesn't add up to be the same. We suspect the CBCT and will try to change today.
Trip is while starting - 10ms, 20ms and 30ms during starting depends on the time delay set.
Yes, CT programmed correctly 50:5.
SEL terminalogy - there is In and Ig. In using CBCT and Ig using the phase CT.

Will let you know after changing CBCT.

A typical electrical engineer

 

[wcaseyharman]

If I’m not mis-remembering an SEL will have both In (neutral CT input) and Io (calculated from the phase CTs). I would think that you could plot both in your oscillography display and compare the two. If I0 is low and In is higher then that would point to a CBCT problem.
I vaguely recall we had problems with motor starting imbalances during starting and had to push the time delay out, but I think we were using the CT return current from the phase CTs. The original CBCT was undersized for the system ground fault current so we had to convert to using the CT return current into the neutral input.

 

[CoolB00Yz]

Thank you wcaseyharman. will try to change the CBCT and see if we can run the motors.

A typical electrical engineer

 

[davidbeach]

Post a trip event file.

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[RRaghunath]

This kind of trip is common associated with CBCT when used with single core cables. The problem arises due to the residual flux that arises from asymmetry in placing the power cables within the CBCT window. If number of cable runs is more than one per phase, it increases the problem further.
Disabling the 50N protection during motor starting (and depending on the residually connected Earthfault protection) is one simple solution.
Use of one CBCT for each trefoil and paralleling the output of CBCTs could help.
There are CBCTs that come with a shield on the inner circumference of the CBCT which will help in flux balancing and thus help prevent maltripping.
Increasing the protection pickup threshold setting to beyond the measured zero sequence current during starting can be another option. But, this will make the protection less sensitive (reducing the winding coverage).

 

[wcaseyharman]

A further variant on the above post would be to temporarily desensitize the 50N element. The 710 should have at least two 50N elements (50N1 and 50N2). Use 50N1 as the more sensitive element. Using SElogic timers and the torque control (if applied) could disable 50N1 for several seconds after the contactor closes (assuming the relay has an aux contact from the contactor, if not a low overcurrent element could be used)
Just some ideas in case it’s helpful.

 

[CoolB00Yz]

SOLVED!! Thank you for your input.
We did tried couple of times to increase the TD and trip setting.
Below shows our trial and error.
10A 50ms - trip
20A 50ms - trip
20A 100ms - trip
30A 100ms - trip
40A 100ms - OK both motors running.

We found the CBCT IN goes to max of 34A which explains the tripping. Since we figure if real ground fault happens, the motor is still protected we will not change the CBCT and keep the same.

I do agree with RRaghunath theory - but we just dont have time, space and available CBCT to perform the changes at site. As the motor is far away, there are 3 conductors per phase connecting to one busbar. Only the 3 busbar for each phase goes through the CBCT. Yes they are not central in side CBCT since they are busbar.

In short to solve this problem:
1. Increase TD (know what you are doing)
2. Increase pickup (know the max value)
3. Disable IN trip during starting
4. Rearrange cables in CBCT
5. Change CBCT

Thank you for reading and I hope this note will help other in future.

A typical electrical engineer

 

[RRaghunath]

If busbars are flat, CBCT with rectangular window should have been selected and not a circular window.

 

[che12345]

Mr. CoolB00Yz (Electrical)(OP)8 Aug 23 06:13
"....SOLVED!! We did tried couple of times to increase the TD and trip setting. Below shows our trial and error. 10A 50ms - trip.....40A 100ms - OK both motors running...."
Thank you very much for your valuable feedback (informing TD=100ms and current 40A) that the issue is solved. We benefitted from your hard work.
It was unfortunate that no one had pointed out/suggested to record the GF current level and the duration with a simple recorder; with the 50N tripping signal removed, for a short test run. With the data, it would had avoided numerous blind tests.
BTW: Although 40A seems to be on the high side but only for 100ms. It is fine.
Che Kuan Yau (Singapore)