Megger Oil Filled Transformer

[bumblebee52]

thread238-219928 In megger's "A guide to Diagnostic Insulation Testing above 1kV" The paragraph before the Step Voltage Test section states that PI testing is not appropriate for oil filled transformers. I went to the AVO Megger Substation Maintenance Class which does instruct performing the Polarization Index test on oil filled transformers with the proviso that the results will be masked by the oil. In megger's "A stitch in time" in the Transformers sectin they mention that in the absence of more reliable data the minimum 1 minute 500 volt DC insulation resistance in megohms from winding to ground with other windings or windings guarded, or from winding to winding with core guarded is equal to C a constant for 20 degree C measurements times E voltage rating of winding under test divided by the square root of kVA rated capacity of winding under test where R is based on dry, acid-free, sludge-free oil, and bushings and terminal boards are in good condition. PRC mentions that for oil filling C-1.5 with guard and 0.8 w/o guard. Are these values for C taken at 20 degrees C for temperature of oil? What if temperature of oil is at 30 degrees C? Where do I find these values for C at different temperatures? Is there a table somewhere that gives this?
Additionlly, if PI test for oil filled transformers is not appropriate and oil masks results, can you megger an oil filled transformer with confidence or not? It seems this formula is highly dependent upon the condition of the fluid in the transformer being fairly pristine. What if the oil is not so pristine and what is the line of demarcation that determines oil is good, okay use the formula vs. not? Do we go by the FIST manual rating table for transformers?
Can we just PI the oil filled transformer and go with it?
How critically important is this masking effect anyway? Does it really skew the PI results that badly? We have about a 500 gallon transformer and the results are all in the G ohms with PI of 2 for the HV side and 1.67 and 1.91 for LV side for brand new.
If the masking is consistent, as long as I take the initial reads and track them over time to plot, that should be good enough maybe?
When we perform PI we usually use the same 5 steps as set forth in the online free AEMC instruments manaul for understanding insulation testing.

 

[Elmir]

When performing maintenance testing on oil filled equipment normally dissipation or power factor testing is performed. I have seen a correlation between between lower insulation resistance and greater dissipation factor readings on smalled oil filled equipment below 25kV and less than 2MVA which had oil at the end of its life requiring replacement.

On large oil filled equipment I have seen specifications written that only required the PI to be greater than 1, this is also the case in the NETA 2009 acceptance testing specification when manufacture data is not available. I believe best practice is to compare to manufacture results if available. The DC insulation resistance test on large equipment is only good for detecting gross faults which is why AC testing is the preferred method. Power factor testing is not affected by the mass of the insulation.

I believe at the end of the day the for any testing the best practice is to trend results and use more than one test alone to make an assessement.

 

[prc]

As mentioned in earlier referred thread,PI cannot be depended on checking the dryness of transformer.A better method is power fcator of winding.Still due to its simplicity,meggering is the easiest way to check the soudness of insulation and hence followed widely,benchmarking with earlier values rather than going by absolute values.

With temperature change, C changes.It is halved for every 10C rise above 20C and doubles for every 10C drop in temperature.

Generally you can say PI goes up when oil quality deteriorates with IR values coming down.

 

[Zogzog]

I never understood the concept of doing a PI on liquid filled transformers. A PI test is really measuring the ability or ease of the domains to align in a solid material, in a liquid everything is in motion so this measurement is not going to have much meaning. Hence the PF/DF test, which gives a true measurement of the properties of the oil.

 

[ausphil]

IR on an oil filled transformer is a greater indicator....

It lets you know whether or not you want to risk damage to your expensive test transformer when doing a high voltage test.

ausphil