Periodic Testing of Dry Type Transformers

[FreddyNurk]

I'm aware that different scopes of transformer testing have been covered here previously, and am also aware that there can be quite a number of factors involved in choosing a testing regime.

I'm interested in the feasibility of testing such as tan delta / dielectric loss angle for smaller dry type LV isolation transformers. The transformers are installed in marine vessels, and range between 35 to 70kVA. The current regime is insulation testing yearly, with no load loss testing and winding resistance testing every two years.

My understanding is that tan delta is only useful if periodic testing and recording of test results is done (in much the same way that dissolved gas analysis is), but its unclear to me as to whether such a measurement is likely to indicate a problem for an LV only isolation transformer, and also what time interval would be needed in order to make it viable.

I'm well aware that its well below the scale of size for anyone to normally bother with for testing such equipment, but does anyone have any considerations on whether it'd be useful to add it to an existing testing regime?

 

[prc]

Normal test for such dry type transformers are IR value and polarization index ( The polarization index is the ratio of the insulation resistance at the end of a 10 min test to that at the end of a 1min test at a constant voltage)
You can also monitor tan delta. As per Doble 1989 conference paper, typical PF for ventilated /open dry type transformers is 2 % (for HV & HV-LV) and 5 % for LV. For encapsulated units, overall PF is 1 %. Modern day cast coil transformers have a PF 0.8-1 %.

 

[unclebob]

Tan delta test is like a sophisticated megger test using AC voltage instead of DC. IMO, a simple megger test should be sufficient to your needs, since you're dealing with small and low voltage transformers (correct me if I'm wrong). The only downside is that a megger test is highly affected by humidity and temperature, a bad reading does not mean that there's an issue with some transformer. Ambient temperature and humidity should be recorded before megger test. This would add meanings to the trending of the results.

Tan delta is used on larger transformers and equipment rated for more than 600 V (2.4, 4,160, 13.8 kV and more...) where insulation resistance is more exposed to corona and other discharge current problems. Given the importance of those larger equipment, being able to detect a problem before something bad happens is critical. Tan delta test is just more accurate than a megger test. That being said, if you already own a tan delta tester, there is nothing wrong in using it on your transformers.

 

[FreddyNurk]

Thanks for the responses folks. I'm leaning towards just IR and Polarisation index, I'm not convinced that the effort to procure and maintain the test equipment is justified for the equipment for anything beyond the above tests.