Meggering a load bank, different values when bonding frame

[majesus]

All,

Unusual problem that I can't explain, looking for some help:

We have a 480V, 3phase, 2MW resistor load bank for testing generators. If we megger the phases with respect to the frame, with the unit's frame not bonded, we get high MegaOhms on all phases. If we then bond the frame, and repeat the test, phase B reads ~500kohms. Phase A and C still read high MegaOhms.

I can't explain it, because I haven't moved the test leads between Phase B and the frame. I only bonded the frame to the ground grid with an insulated conductor and then phase B resistance dramatically drops. No other cables are connected to the load bank. It is just a load bank, on a skid frame, unconnected and sitting on graded dirt.

If there was an insulation resistance problem with phase B, without moving the leads from phase B and equipment frame, why does it not show until the frame is bonded to the ground grid? The test leads are still in parallel with phase B and the frame. How does grounding the frame, now introduce an insulation leak? Also worth mentioning that the Megger unit is a portable, battery operated tester.

 

[davidbeach]

Have you meggered the phases to ground or only to the frame? Sounds to me like you've got a B-Ground problem that doesn't show if you test between the phase and the floating frame.

 

[majesus]

Just across the frame.

We didn't have the opportunity to investigate further. Just a weird but interesting problem. Vendor took the unit back today. Wanted to post it here if anyone seen something similar.