Pin holes from phantom 15kHz pulse

[rckstr]

480V 3PH 70A ckt feeds six MOV’s (Motor Operated Valves) located in tank farm of petroleum storage facility. The 70A breaker is tapped from an 800A fused disconnect fed directly from local utility (pole mount transformers). #1 THHN runs between MOV disconnects (mostly underground conduit). Each disconnect has a 20A breaker close to the MOV. At the farthest two MOVs (800ft) we get 2 pin holes blown through the insulation about 1 foot apart. This is every few weeks for the last six months. System ran for one year prior with no problems. Recording meter captures a 15,000hz frequency surge with no significant change in voltage or current at the time of each event. Each MOV has a control transformer and forward/reversing starter, but this happens even when none of the MOV’s are in use. We have swabbed the conduits several times with wire ease and rags. We removed the #12 from the load side of the disconnects and pulled #6 to the MOV’s with the same result. Have not used oscilloscope, but imagine we are getting such an ugly sine wave it appears as 15kHz to the recording meter. There are no VFDs within 2000ft. I see no switching power supplies. All the area lighting appear to be functioning properly. The only thing resembling witchcraft close by is an ancient cathodic grounding system. As an electrician the only formulas I deal with involving frequency are about reactance: but we never see a voltage or current anomaly.......Considering abandoning search for the source and pursuing mitigation. Thinking: isolation transformers, TVSS with filtering, and/or XHHW wire….What say ye?

 

[synchro]

With the long line there might be some type of resonance near 15 KHz and that's what you're seeing on the recording meter. If that's the case then the question is whether a resonance was excited after the pinhole shorts had already occurred, or did a resonance increase the voltage during some transient condition and that caused the pinholes. In other words, was it a cause or an effect? In either case some event like a surge must have initiated it.
Another cause of the 15 KHz is that it's just from the limited bandwidth of the recording meter itself, and the actual surge could've had fast risetimes that the meter could not track.

Are the control transformers near the MOVs connected all the time? If so have these survived these events? I'm assuming that they have since you didn't mention it.

... At the farthest two MOVs (800ft) we get 2 pin holes blown through the insulation about 1 foot apart. This is every few weeks for the last six months.

So are you saying there are two pinholes in total, or are there two pinholes near each MOV? And where are the pinholes located, for example inside the disconnects, near the conduit entry, the MOV's, etc.? Is their location relatively consistent each time it blows through?

 

[GoldDigger]

The fact that the holes are spaced a foot or so apart and at the end of what can look like a transmission line, makes me suspect that some sort of high voltage transient at a well defined frequency traveled down the line, hit what looked to it like an open circuit and developed double or more standing wave voltage. The peaks of this reflected transient then caused the pinhole discharge.
All sheer theoretical speculation, of course.
But regardless of the cause and the mechanism of the damage, chances are pretty good that a proper surge suppressor with very short leads mounted at the end of the last run will prevent wire damage,
I would still be concerned about identifying the source.
Changing location of the meter from near (input) end of the line to the far end may help narrow down the source. I would be a little concerned that it might just blow the meter.

 

[rckstr]

With the long line there might be some type of resonance near 15 KHz and that's what you're seeing on the recording meter. If that's the case then the question is whether a resonance was excited after the pinhole shorts had already occurred, or did a resonance increase the voltage during some transient condition and that caused the pinholes. In other words, was it a cause or an effect? In either case some event like a surge must have initiated it.
Another cause of the 15 KHz is that it's just from the limited bandwidth of the recording meter itself, and the actual surge could've had fast risetimes that the meter could not track.

Are the control transformers near the MOVs connected all the time? If so have these survived these events? I'm assuming that they have since you didn't mention it.

So are you saying there are two pinholes in total, or are there two pinholes near each MOV? And where are the pinholes located, for example inside the disconnects, near the conduit entry, the MOV's, etc.? Is their location relatively consistent each time it blows through?

The control transformers remain hot all the time, but are never damaged.
Two pin holes will be in the same conductor, one foot apart and twinxt MOV and disconnect. I can't tell if every incident is always the same phase. One MOV is 10ft from disconnect; the other is 150ft.

 

[synchro]

The control transformers remain hot all the time, but are never damaged.
Two pin holes will be in the same conductor, one foot apart and twinxt MOV and disconnect. I can't tell if every incident is always the same phase. One MOV is 10ft from disconnect; the other is 150ft.

Perhaps the loading of the control transformer has been enough to prevent excessive voltage from resonance or reflections on two of the phase conductors, therefore preventing damage to those conductors as well as to the transformer itself. And so the unloaded phase is suffering the damage. Just a guess at what could be happening.
I agree with GoldDigger that a surge suppressor at the disconnect or MOV, whichever is closer to the insulation failures, would be appropriate.

 

[rckstr]

Much appreciated. Y'all have motivated me to read further about resonance and standing waves. I will go ahead and thank you for surge suppressor advice now as we may not realize success or failure for several weeks.

Before walking away I would like to pick bigger brains a little further: Do you think a torroid ring or looping the circuit conductors would negate the effects of the increased frequency? I would have no confidence sizing a coil myself. I'm really just wondering if I'm grasping the theory correctly. The surge suppressors will definitely make more sense to those in the room where I'll be standing.

 

[petersonra]

At this point I think your best bet is to install a surge protector on the 480 volts inside the disconnect.

Get one with a fault indicator though.