ET online
July 2009
Q and A - Transformer testing

Q and A - Transformer testing

09 July 2009

In this issue, the Megger Technical Support group focuses its attention on transformer testing.

Q: When I use a standard low-resistance ohmmeter to measure the resistance of transformer windings, the results look wrong and I never seem to get the same result twice. What’s the problem, and what can I do about it?

A: This problem is very common! The reason for the unexpected and inconsistent resistance measurements is that it takes a long time for the test current to stabilise. This is because of the high inductance of the transformer windings. The inductance has to be charged before the test current reaches a stable value and this can easily take five minutes or even longer with large transformers. Measurements taken before the current stabilises are virtually meaningless.

There are two solutions to this problem. The first is to be patient, and to watch the display on the low resistance ohmmeter until the reading has been steady for, say, a minute before recording the result. This is fine if transformer tests only have to be carried out occasionally, but for regular testing, a much better solution is to use a low resistance ohmmeter that is specifically designed for transformer testing, such as the Megger MTO210. These instruments use a variety of techniques to speed up testing, such as applying test currents to the transformer primary and secondary simultaneously in such a way that the magnetic effects in the core tend to cancel. This allows testing time to be reduced by as much as a factor of ten compared with using a standard low-resistance ohmmeter.


Q: Recently, when one of our power transformers was being moved, we had a problem with the hoist, and the transformer was dropped. It only fell a few centimetres, but I’m concerned that it might have suffered internal damage. How can I find out whether it’s OK to put back into service?

A: Standard insulation and resistance tests may provide an indication of some types of fault, but they don’t give a complete picture. By far the best approach is to carry out a Sweep Frequency Response Analysis (SFRA) test.

This injects a variable frequency AC voltage into each winding, and plots the resulting current as a function of frequency. Even small changes in the windings or core geometry produce a significant change in the transformer’s response curve.

Ideally, it’s best to have a baseline “fingerprint” for the transformer, produced by an SFRA test carried out before it was moved. This allows an accurate comparison to be made of the before and after results. If baseline results are not available, it may be possible to make comparisons with another transformer of the same type.


Q: Insulation tests on one of our power transformers have revealed that its dielectric strength is poor. What do we need to do next?

A: First make sure that the dielectric strength really is bad, and that the poor test result isn’t a result of, for example, dirty bushings. One way to do this is to use the guard terminal provided in all good HV insulation testers to eliminate the effects of surface leakage. The tester’s manual should provide instructions about how to do this.

If the dielectric strength really is poor, the next step is to carry out a Karl Fischer test to determine the moisture content of the transformer oil, but be careful about the test conditions, as the results obtained are temperature dependent. It is also a good idea to perform a Frequency Domain Spectroscopy (FDR) test, which will reveal the amount of moisture and contaminants present in the solid insulation. It’s worth noting that FDR tests are not affected by temperature, an important point for transformers that have already been taken out of service and are, therefore, cold.

The Karl Fischer and FDR tests will provide reliable information for deciding whether the transformer can be reconditioned, or whether it must be scrapped.


If you have any questions that you would like to ask Megger’s TSG, please give them a call on +44 (0) 1304 502102 or email