Arc Furnace Transformer

assuming the oil in the tapchanger is not the problem? it may need changed because its insulation properties are to low ie its dirty? also check the tap changer contacts in case one is damaged and causing poor readings it may be shorting somehwere.the contacts in the manual changing kiosk, assuming you can chnge it maunally as well not just electrically may need examined.its usually something simple.

The tansformer oil is a hygroscopic substance and must have absorbed water vapour during 6 weeks remaining open while work was continuing on it. The effected oil can be in the transformer main tank, tapchanger mechanism and inter connections thereof. It is obvious that phase "C" was isolated from all other apparatus when testing for Insl. Res. It is not known if the winding was tested after oil filling and after dehyderating it. I am sure that the work done on the winding was handled by an expert who reassembled the components taking due care of re-assembly procedures. Testing with a 5000 volt device had to be carried out after testing with a low volage (500V megger) in case the the winding was tested before drying/dehyderating the oil/winding. High volage application on a low insulation resistance condition some times damages the insulation permanently. I would suggest you to go through the whole process and components one by one looking for overlooked procedures. I am sure the fault will be found and properly removed.

This could be one of many different problems & from your details it is very difficult to tell what it may be. There may well have been a problem in the OLTC before the other work was done. You may have a barrier board (between the OLTC & main Tank) that has absorbed moisture, there may be a faulty part eg transition resitors etc, there may be high moisture in oil as suggeted by others.

Can I suggest some of the following test be done, Insulation Resistance (G-ohms), Winding resistance on all taps, at 11kv you could try a power frequency withstand at up to 75% (21kV) this will tell if there is an insulation breakdown & where it is, If you have not filtered the oil then start with that - get it clean & dry first.

If you have the test gear do a Sweep Frequency Response Analysis (SFRA) this looks at the winding & compares each phase against the other. If the wave shape for 1 phase is very different then you know where to start looking - It takes the OLTC into the equation as well so you can pin down which phase to look at there.

You need to monitor the gas in the oil when it is all back together as this will tell you if you have a problem starting to appear & what it is.

Give us some test results & we can look at the possible scenarios for you.

check for "mica" the insulation between bus/surge cable and furnace-arm.