During my review of the Megger insulation tester in Blog 9, I tested a 120MW Generator rotor winding that actually failed the insulation resistance test. Following this issue, the bedplate heaters for the generator were reconnected and placed back into service. Around 4 weeks later, I returned to the generator to carry out a further insulation test with the expectation that the heat would have driven out the moisture from the winding and a healthy insulation reading would be obtained.
With the heaters being disconnected during the last test, the winding temperature was taken as being the same as the ambient temperature. With the heaters in service, a higher winding temperature was expected and this was verified by measuring the resistance of the embedded winding temperature sensors. These are PT100 devices and tables are easily obtainable to cross reference the resistance obtained to a temperature value.
The resistance measured cross referenced to 21 degrees Centigrade which is circa 10 degrees above the ambient conditions. To maintain the winding in preservation, the recommendations are to maintain the winding temperature 5 degrees above the ambient and the humidity below 30% RH.
Initially the resistance was measured with the resistance function on the Megger However it was found that the resistance value in excess of 100 ohms was set to a scale without any decimal places This introduces a slight inaccuracy when using the insulation tester for this test For the purposes of an insulation test this isn't an issue but when measuring the winding resistance the corrected value can be affected due to the inaccurate temperature correction as detailed in the video below(lasts 1m 45s
A comparison of the rotor winding resistances obtained can also be seen in the table below.
For this reason the winding temperature sensor resistance was measured with the Keysight to obtain a more accurate value In fairness to the Megger I don't think it was ever designed to replace the resistance function on a multi-meter Even the Keysight only displayed one decimal place for the scale so a good multi-meter would generally offer an extra digit over this as well
The 1 minute insulation resistance test was conducted at 250V and then 365V using the variable voltage function of the Megger This setting was used to match the working voltage rating of the rotor winding A final test was then recorded at 500V
When carrying out multiple insulation tests at different voltages, the expectation would be to see a drop in insulation resistance as the voltage is increased. The results obtained didn't display this trait and this can sometime happen with large inductive apparatus when carrying out a succession of tests, as stored charges can sometimes remain and affect the next test result. This is resolved by allowing for longer discharge times between tests.
However, the values obtained during this set of tests were much better than the previous tests as shown in the comparison table below.
The DAR values were found to be above the minimum value of 1.2 for both the 365V and 500V tests, where as during the previous test they were around 1.00, which would be considered too low. The video of the instrument set-up on the variable voltage and the full DAR test is below and lasts circa 3 minutes.
As such a good 1 minute insulation test value was obtained, the polarisation index test was conducted at 365V and then again at 500V to see if an increase in voltage would have an adverse effect on the insulation reading obtained.
A good polarisation index value was obtained for both test voltages, 3.05 for the 365V test and 3.03 for the 500V test, well above the minimum value of 2.00. The insulation resistance plot seen above shows the classic response for a polarisation index test. This time the 500V test produces a slightly lower reading than the 365V test as expected. The drop however, is not significant enough to cause concerns.
The video below is the polarisation index test at 365V. I have sped up the test to reduce it down to around 4 minutes.
Overall the continues to perform very well and I am happy to continue using it It has done around 300 tests to date of various types and the battery voltage drops down to 50 during the heavier current tests so appears to have a reasonable life