Effect of installing surge caps on motor ground protection parameters? 1

[Shark96]

Does the addition of surge caps on motor terminals demand the need to adjust ground fault (trip) settings in motor protective relay?

 

[GroovyGuy]

Hi Sahrk,
Yes, a very short time delay will likely be required. I have used 0.1s before, and this seems to work just fine. BTW that was a 16,000hp 13.8kV synch motor supplied from a system that has a 50A NGR.
GG


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

 

[Shark96]

Thanks GroovyGuy,
But what is the reason for this delay?

 

[GroovyGuy]

Hi Shark,
If I recall correctly, that w/o the TD the 50G element (on the MultiLin MPR) would pick up on instantaneous GF on every attempt to energize the motor. When the TD was introduced, the tripping stopped. It was obviously the initial charging of the surge-caps that was the issue.
GG

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

 

[Shark96]

Thank you GroovyGuy!

 

[electricpete]

We have five different applications of 13.2kv motors with surge capacitors, instantaneous ground relay ("50G", electromechanical relay fed from window CT enclosing all three phases). No intentional delay. Fed from HRG system. Been operating many decades without spurious trips.

I hadn't stopped to think about this aspect before (at least not recently enough to recall).

During across the line start, how to analyse capacitor currents. As a first simple model, I'd be inclined to say that all three phase to ground voltages ramp up to their prescribed value in the same delta-T. Assuming the three phase-to-ground voltages sum to zero, the computed dv/dt's under above assumption would also sum to zero. But didn't consider things like LC ringing (rather than smooth ramp), CT saturation, non-simultaneous closing, etc which could throw a wrench in the works.

Any thoughts to further describe this? Does anyone have links to literature describing adjustment of instantaneous ground setting for capacitor "inrush" currents.

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

 

[waross]

Hi Pete.
If it works, don't fix it.
I suspect that the CT arrangement may make the difference.
One CT will probably not have saturation issues.
Three individual CTs may have saturation issues and require a delay.
Comments?

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

 

[electricpete]

Good point. I agree, three individual CT's connected residually would be susceptible to creating non-zero sum from saturation at high phase current levels, whereas a single window encompassing all three phases would not. It's hard for me to gage whether that capacitor inrush current would create saturation (high dv/dt driving the current but only a very tiny capacitance.... hard to believe it could be large enough to drive ct into saturation assuming that ct does not saturate under LRC). But if it did reach saturation levels, that would certainly explain why we don't have problems and others do.

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

 

[GroovyGuy]

Good point Pete. The OP did not indicate whether the 13.8kV system was solidly grounded or NGR grounded. If NGR grounded, it can be quite difficult to pick up a GF with only a set of residually connected CTs. You really need a zero-seq CT if you have a NGR grounded system.

I would think that a GF relay would have a tough time to pick up any surge-cap energization issues with residually connected CTs. I had assumed that he had a zero-seq CT in this situation.
Perhaps the OP could provide some additional information?

GG

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