Transformer HV EARTH

[Eskom]

Good day.
I came across a fault yesterday afternoon, its a 22kv 50kva pole mounted transformer with a down running earth from the transformer tank(HV side) into its own crow foot earth and on the LV side it has a down running earth from the LV bushing to its own crow foot so on the transformer it self the HV and LV earth is seperated. Now the fault that occurred is that the HV earth at the bottom of the structure was cut off and when you test voltage between the piece that runs down from the transformer and actual earth it measures 670V and the metal meter box is attached to the pole and made contact with the earth which caused it to be alive. I would understand if it happened on the LV side but on the HV? the fault is sorted out but i would like to know the reason that would cause this voltage on the HV earth.
Thanks

[Justloadit]

How did you know there was a fault in the first place?
Did you test the site , or was it a customer who complained?

[Eskom]

Customer complained,said he saw sparks at the structure.

[AndyD]

The 670v between the disconnected (floating) side of the earth and the actual earth doesn't mean a great deal if it was taken using a digital voltmeter with no LoZ range on it. Seeing 670v with a high input impedance tester doesn't necessarily mean any or much current would flow if the circuit is closed again.

The visible sparking could be caused by a current as low as a fraction an amp but I don't see any reason that the Tx should have leakage even on the 22KV side. A more interesting figure would have been from a clamp meter around the HV side earth wire after the repairs were completed showing the leakage current..

[MullerR]

The HV earth (or MV earth in fact) is connected to the MV side of trfr which is also connected to the surge arrestors. First:the arrestors can also have a leakage or creepage like insulators and secondly, in my experience from the past regarding trfr earths etc. that voltage is nothing more that induction with no or little amps. But like Andy mentioned, the amp reading would have been interested to know if there was in fact a fault current.

[Justloadit]

With wires in the air, and a dry day with wind, then static electricity will build up in the HV wires, as they are isolated by the transformers, and this static must discharge to earth. Also just before a storm when there is high winds and clouds moving, static builds up in these wires.

Radio hams experience this static electricity when they put up long wire antennas. I've experienced arcing on antennae wires as long as 10cm just before a thunder storm, when the antennas are disconnected from radios and left lying unearthed. So the sparking seen on the earth wire, could be the static electricity build up in the HV system and discharging to ground. Typical arcing is 10KV per cm in dry air. So if the arc is 2 cm long, then the voltage is approximately 20KV, and so on.