Ground fault in control wiring of 480 volt motor circuit 1

[Tom04]

I am trying to better understand how a ground fault in the 120 volt control portion of a 480 volt, 3 phase motor starter circuit can cause a ground fault alarm in our high resistance ground fault detection relay on our MCC.

We incurred a ground fault alarm on our 480 volt, 3 phase MCC. I looked at the ground fault relay, and it was showing 270 volts at about 3.3 amps. Our electrician determined that the start-stop station of was full of water, and was causing the ground fault. This is a size 5 starter, and the start-stop controls are 120 VAC via a 480/120 .100 KVA CPT (one side grounded) in the starter. The control is a basic 3-wire field start-stop station, operating a 120 volt control relay in the motor starter bucket, which in turn operates the 480 volt coil on the size 5 starter.

I know it happened, because when we cleaned the water out of the start-stop station, the ground fault went away. But I don't see how the ground fault went through (or around) the control relay and caused the high current (3.3 amps @ 270 volts) in the ground fault relay, without blowing the primary or secondary control fuses around the CPT.

 

[jraef]

1) Assuming that the primary of the CPT was fed from the 480V MCC bus, and it is a 4wire system (must be because you show 277V).
2) Assuming that it is a Zero Sequence GF protection scheme (as opposed to Residual Current method which is typically not used in resistance grounded systems).
3) Assuming that 3.3A is enough current to activate the Alarm level setting of the GFR.

Simple answer based upon these assumptions:
One side of your 3 wire start/stop relay wiring is hot (120V) off of the CPT secondary. This was then flowing to ground in the water, a Ground Fault pulling current which ultimately came from the 480V side of the CPT. That GF current was obviously enough to trip the Alarm in the GF relay AND the relay trip time settings were faster than the thermal curve of the CPT fuses.


"Venditori de oleum-vipera non vigere excordis populi"

 

[busbar]

NEMA size 5 would usually indicate an interposing relay, and line-votage coils(s) on the magnetic starter. If you have not yet, megging control wiring is in order.

It almost sounds like someone may have mistakenly wired the wrong side of the interposing relay into the stop-start station. Else, if it is truly grounded 120V, water should not effect the 480V-ground detection, unless also the CPT had its secondary-side ground lifted, and had a primary-to-secondary fault.

 

[rbulsara]

I tend to agree with busbar. You may want to physically check the voltage at the start/stop station (between the control wire and ground). It appears that start/stop station is a 480V circuit in series with 480V coil.

or there is a 480V system wire present in the push-button station.

Besides I am curious what type of GF relay and scheme is this, because you have voltage reading there.

 

[aolalde]

I assume the in/out fuses of the CPT are too large, allowing operation of the GFCI. review the CPT fuses.

 

[jraef]

busbar,
I should have had another ASSumption in there. IF the CPT primary leads are connected to the 480V down stream from the ZSCT, a ground fault in the 120V of the CPT secondary will be seen by the ZSCT. I have witnessed this several times, sometimes the GFR trips just on the CPT inrush current. The solution is to tap off the CPT primary power from the UPSTREAM side of the ZSCT. It is a common mistake.

aolalde may have a point, especially since people often will oversize non-time delay CPT fuses to get around the nuisance blowing on inrush. But again, moving the CPT feed to the other side of the ZSCT will solve it.

"Venditori de oleum-vipera non vigere excordis populi"

 

[busbar]

jraef — Shouldn’t a CPT-secondary fault be reflected into the primary as phase-to-pahse overcurrent? I wouldn’t think that would trip 480V ground-fault protection.

CPT-fuse types and sizes should be verified. Sometimes, though, "correct" fuses don't always coordinate well with ground-fault protection. There may be non-ideal tradeoff given the need to limit ground-fault damage in controls or motors.

 

[advidana]

usually the sec xfmr pri is factory wired at 10 amps so 3.3 amp should be not problem to get.

 

[peebee]

The only way I can think of that a GF on the 120 side would show up also as GF on the primary side is if the CPT was referenced to a primary phase instead of earth ground -- which would have been a code violation (and if I'm not mistaken, it would have been rather difficult to achieve that connection. . . ). . . . An autotransformer would fit that bill, but that would also be very unlikely (do they even make 480:120v autotransformers?)

jraaf -- I too fail to see any other way that a 120v GF would show up as anything but overcurrent on the primary side, other than this unlikely scenario of a primary-secondary bond. Please explain.

I think busbar's probably got the right idea. In fact, I won't even wait til I find out for sure -- red star for busbar.

It sure seems to me

 

[jraef]

It comes from the issue of using a core balanced Zero-Sequence CT for GF detection. Excess current on 2 legs (the primary of the CPT) that is unbalanced with respect to the 3rd leg will create a current flow in the ZSCT, possibly triggering the GFR. Granted it takes a lot of ifs as I pointed out, but it is possible and I have witnessed it.

"Venditori de oleum-vipera non vigere excordis populi"

 

[peebee]

I disagree -- say the motor's not even running. But the CT's on, drawing perhaps 10 amps.

Well, then we get 10 amps in A, 10 in B, and 0 in C. But the net current through the 3-phase circuit will still be zero, no matter if you're detecting this with zero-sequence or residual detection.

Matter of fact, on a motor circuit like this, the currents will generally be very well balanced, when compared to many switchboards or panelboards that are powering extremely unbalanced loads but have no problems with their GF detection.

I still don't buy it jraef. . . .

 

[jraef]

Remember we were not speaking of normal operation, but a ground short on the secondary. I agree, there is always a level of unbalance in a 3 phase circuit, but in the case of distribution gear you run the neutral through the ZSCT as well, which you typically do not do in a motor controller. I have had Multilin 469 relays trip on GF from the inrush of an improperly connected oversized control power transformer when used with a ZSCT. Move the wires to the other side of the ZSCT and the problem goes away.

We can agree to disagree I suppose.

"Venditori de oleum-vipera non vigere excordis populi"

 

[PWR]

How about an improperly wired zero sequence CT? If the grounding conductor was run through the CT along with the phase conductors this problem could readily occur.

The 120V ground current might then find a lower impedance path back to the CPT by flowing on the 480V grounding conductor.

Happens quite regularly on medium voltage systems.