THD and ground fault trip 2

[bam55]

I have a problem with a ground fault circuit protector in a HID light circuit about 300m long and 15KW, 30 lights. The ground fault trips out when the nieghbouring crane comes on which utilizes around a 200-400kw VFD. The VFD has its line and load reactors like it is supposed to and the whole system is properly and soundly grounded (to seawater). The system is a 4 wire hi resistance ground and the lighting circuit is a 3 wire 600V system. I want to check the harmonics myself to see whats there at the time of the crane operating, but I'm not sure what can be reasonably done on the lighting circuit side to eliminate the problem.
The feeder for the crane and the feeder for the light service main transformer are separate. Any knowledgeable advice?

Thanks,

Bam55

 

[jraef]

EMI/RFI from the VFD output? Sometimes on crane systems the output cables are unshielded flexible cords. This turns the VFD into a powerfull local FM radio transmitter, and some of that EMI or RFI might be affecting your GFR, either by scrambling it's internal electronics or by superimposing an imbalance on the lines feeding your lighting circuit. You said 600V 3 wire so I suppose it is a residual current GFR. They can be sensitive to imbalances.

I have seen some very squirrely things happen from this issue. Just a wild stab though.

"Venditori de oleum-vipera non vigere excordis populi"

 

[rbulsara]

What type of GFCI you have? 3-6mA? or 30ma?

Or is there a ground fault protectin relay?

 

[rbulsara]

Harmonics by themselves do not cause a false GF trip. For a GF trip you need have current in the ground circuit.

What is more likely is that the leakage currents due to electronic circuits or capacitors on HID fixtures will cause a sensitive GFCI to operate. What is the sensitivity or setting of your GFCI?

 

[bam55]

They are low settings. 2A for the NGR trip relay on the 1000KVA main breaker, 200mA on the 15A, 600V feeder breaker for the lights and I/m not sure but I think the setting on the second ground fault relay in series is 6ma or 20ma but I'm not sure. I can check this though. I did a calculation according to IPC's application guide and the return current due to what you were talking about (capacitive impedance effect) is about 6mA.
However the Main breaker trips sometimes too due to this problem.
I don't think its EMI because the service building is about 50m or so away from the crane.

Regards,

 

[bam55]

Further to above I should take back what I said about return current being 6mA. What I was referring to was the charging current for the NGR calculated from the various cable impedances.

Regards,

 

[rbulsara]

I am not sure what the crane motor start is contribuitng, may be causing just enuff disturbance to trick your too sensitive and probably misapplied GF protectors.

Did you mean 150A feeder or 15A feeder?

The whole GFP setup sounds little too sensitive or even questionable. Why do you even have GF on HID lighting circuits? It is a bad idea to begin with.

Are you sure the GFCI are current sensing and not voltage sensing (59G type), a VFD motor start could distort voltage just enuff to set off a too sensitive GF.

Also it is little surprisiing that on a high resitance grounded system you have a instantaneous GFP? Since the current is limite to 2A, all you want is a GF alarm not a trip!!! Or have a long time delay.

 

[bam55]

It is a 15A feeder.
I agree that the second ground fault trip unit in series which was applied by a contractor before my time to stop the main breaker for the switchboard from tripping all the time is abit missapplied.
The reason for this still is that the switch feeding this 1000KVA sub doesn't want to see any ground faults returning to it as it is the generating station and has senitive plc's they don't want affected. So the ground fault automatically trips out when it sees more than 2A or whatever.

 

[RalphChristie]

bam55:

"The ground fault trips out when the nieghbouring crane comes on"
"The system is a 4 wire hi resistance ground"

I assume your resistors are connected between the star-point of the trsf and earth. If you are saying you use a 4 wire system, where are your neutral connected?

 

[busbar]

 
Twenty-milliampere ground-fault protection at 600V corresponds roughly to 4mA at 120V, making the safety objective somewhere near personnel protection on the lighting circuit. I think that a very conservative setting has very limited safety improvement, and opens the lighted area to nuisance trips and loss of ambient lighting at potentially critical times. Lighting continuity in a work area should probably take precedence to some degree of electric-shock personnel protection.

Charging current from zero-sequence capacitive reactance on a 300m discharge-lighting circuit may be close to the of the lighting-breaker trip level {and subject to variability from simple weather change.} A ‘forced’ test to verify current through the grounding resistor {and as-found breaker-trip level} may give better understanding and confirmation of the right approach and acceptable tradeoffs in raising the lighting-breaker GF-trip level.

If low-level GF protection is still desired for the lighting, it may be possible to split it into three 2-pole circuits to minimize total lighting outage.

 

[bam55]

Thanks for the tips,

I'll go to work on it.

Regards,