Sorry... I couldn't find any particular forum on generator maintenance so I posted this in the motor forum, since I guess the context would be quite similar.

At my locations there are several 10.5kV generators. Part of our maintenance is to carry out a 5kV megger and PI test. This is done with all PT, surge capacitor and neutral points removed. And for simplicity, I usually do it without disconnecting the generator terminals. Therefore, the test actually includes the terminal busbars and a step-up transformer. The generator is Y-connected and the transformer is delta with no ground on the generator side.

For 2 of these generators, we could obtain megger values. However on 1 of them, it always shows a zero-value indicating that there is possibly some earth leakage path. If I disconnect the generator and megger the generator alone, the readings would be fine.

The configuration of the 3 generators, i.e. its connection to the transformers are identical. The only difference with the problematic one is that the HV-side of the transformer is connected to an oil-insulated cable of a 150m length. Whereas for the other 2 generators, they are connected to bare overhead lines. Since this is the only difference, I'm wondering whether this is the cause of the megger problem.

Would be glad if anyone can shed some lights.

P.S.: The generators are all healthy and operating well.

It seems to me as a very serious problem. If I understand the description correctly, there are three generators and every one of them has a step up transformer. There are 3 wires going from each generator to the step up transformer. Am I right?
There seems to be a problem with the connection of the generator (the one that you cannot megger) to the step up transformer. One and only one phase must be grounded somewhere between the generator terminals and the terminals of the step up transformer. Since it is the only one phase that is grounded, there is no closed circuit and there is no indication of any problem. Everything looks fine, and even with one phase grounded it can work for ever.
If the high side of the transformer is connected to the cable or the power line does not matter (IF IT IS REALLY A TRANSFORMER). If it is an AUTOTRANSFORMER it would be different story (I believe it IS the transformer).
If it is really a transformer (not autotransformer), you have to locate the ground as soon as possible. It may happen, that another ground develops and then the sparks will fly!
jank

One thing to mention first: the IEEE 43 limits for insulation resistance and PI are geared toward machines tested alone. Long runs of cable and certainly transformers can affect the results. Apparenlty not the transformer in your case. Seems likely the cable.

Two possibilities that come to my mind:
1 - The cable has a very high capacitance which your tester is not capable of charging (the voltage is drooping).
2 - The cable is degrading. Jacket or termination degradation can lead to oil leaking out or water leaking in.

Testing a 10 kv system with a 5 kv megger will only show you gross problems, isn't that right?

Shouldn't Calvin isolate the cable and perform a megger test, then a hipot test, and also megger the generator and transformer primary windings? (Order and process of tests depending upon results of course.)

I'm sure it's okay to test in his current method (which only provides a limited amount of information as to equipment health) but if the current method indicates problems further action should be taken.

It seems to me that a zero megger reading on the system should trip a relay if the generator is grounded properly and zero is a true megger reading. Perhaps the megger doesn't have enough supply to charge the system considering the relatively minor leakage, which isn't causing problems for the generator.

Not that this is the case on this system, but I saw one 3 phase delta transformer system which fed a single phase printing machine so the installer referenced a phase to ground which I think eliminated a high leg.

This message has been edited. Last edited by: Wally gator,

Another thought; could your megger be backfeeding though the generator ground or ground resistor bank? Perhaps though your generator bond or transformer bond? Maybe your cable shield is providing a backfeed path.

Where do you disconnect the neutral (ground point)?

This message has been edited. Last edited by: Wally gator,