Ground Fault tripping - Parallel Motors

[nauas]

have recently come across a nuisance tripping issue that has me baffled. We have a 34.5kV switchgear feeding two Motors (call them A and D). Both motors are protected by GE 469 MMR and are exactly identical. Recently both motors tripped on Ground fault at the same time. I was called to the plant during the middle of the night and by the time I reached, the operators had already attempted to restart the motors since they are process critical. the sequence of events is as follows:

1:30 am both A and D tripped
1:50 am I get called
2:00 am motor D is started by operations crew and the pump is loaded up with no issues
2:15 am motor A is started and immediately both motors trip on GF. MMR of Motor A also had I> pickup alarm.
2:20 am, I reach the site and check the Insulation resistance of both circuits.

It was found that the cable connected to motor A had failed as the IR value to ground on Phase "C" cable was a few Kilo ohms, both motors and the cable connected to motor D was found OK.

3:15 am Motor D is started again without problems and has been running smoothly so far (more than 4 days in operation).

Later it was found that the cable failed due to problems with a joint approximately half way through the cable run. My question is How can a ground fault on one outgoing circuit affect a parallel feeder. I was initially speculating that since this is a single line to ground fault, Motor D's contribution to the short circuit current is seen by its protection relay and since the protection is non directional and probably too sensitive , it initiated a false trip. However looking at the protection settings this is unlikely.

The GF is implemented through CTs with CTR 50:5 (the usual) and the relay is set for 0.37xCT and 100ms delay. Motor contribution to the fault should be zero or near zero beyond 3-5 cycles (83ms or less for a 60Hz system) under this condition I do not think that the motor contribution caused the trip since at 100ms it should have definitely gone to zero.

Has anyone faced similar problem in the past? Also what could be the possible cause of the trip?

 

[davidbeach]

How many times do the cable shields go through that 50:5 CT? If it's an odd number, like 1, there's your problem. Needs to be even could be 0, could be 2.

 

[waross]

One possibility.-
34.5 kV = 19.9 kV voltage to ground.
19.9 kV / 2000 Ohms = about 10 Amps.
I suspect that part of the ground fault current may have followed the ground conductor to the equipment skid and returned on the grounding conductor of the good motor.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[electricpete]

Good suggestions so far.

I'll throw out one scenario for consideration that you can probably shoot down:
IF the ground relay is fed by three single-phase CT's (rather than one window CT) and the current is high enough to saturate one of the CT's, then the secondary of the CT's can create a non-zero sum even though the primary currents are (ideally close to) zero.

I have no concept to visualize how much current flows here for how long without understanding the system better. I know the motor will not directly feed the ground fault but will interact with the power unbalanced voltages that result from the ground fault.

So some more description of the system might help a little bit, even though I'm probably not the one in this group that will be able to figure it out.....To be honest I'm totally confused about what you described. You mentioned 34.5kv switchgear but surely the motors are not rated 34.5kv? So most likel there's a transformer... if so please tell us about its ratings, winding configuration, low-side voltage along with the motor horsepower rating.

relay is set for 0.37xCT

can you explain what is meant by this.


=====================================
(2B)+(2B)' ?