Q and A on high voltage transformers

A transformer failure in a distribution network is invariably a costly and disruptive event so it’s not surprising that we are often asked questions about testing transformers as aid to minimising the risk of failures. The questions fall into two categories: how best to test a transformer before putting it into service, and how to determine the condition of a transformer that’s already in service.

Q: When a transformer arrives on site, whether it’s brand new or an older unit that’s been moved from another location, how can it be checked for transit damage before it’s put into service?

A: The best and most convenient approach is to use sweep frequency response analysis (SFRA). This type of testing treats the transformer as an electrical circuit made up of multiple capacitances, inductances and resistances. It injects signals over a range of frequencies – SFRA test sets typically cover a frequency range from 0.1 Hz to 25 MHz – and generates a frequency response curve for the transformer. Measurements can usually be completed in about a minute.

Ideally, SFRA reference curves should be captured when the transformer is at the factory or at its original location so that these can be compared with later measurements at its new site. It is, however, possible to use type-based comparisons with transformers of the same design. The comparison between the most recently measured curve and the reference curves will reveal mechanical and electrical problems within the transformer, many of which cannot be detected by other techniques. SFRA testing is also useful for checking transformers that have been subjected to electrical faults that may have caused displacement of windings or other internal components.

Q: Taking transformers out of service for testing is inconvenient and costly, particularly if the tests take a long time to perform. Bearing this in mind, is routine transformer testing really worthwhile?

A: Only you can make that decision, based on your knowledge of your network and the amount of disruption an unforeseen transformer failure is likely to cause. However, in almost all cases, routine testing is definitely worthwhile – particular as modern test techniques minimise the amount of time that the transformers need to be out of service.

Q: What is the best method to use for routine diagnostic testing of transformers?

A: Many alternatives are available, but one of the most useful is frequency domain spectroscopy (FDS). The big benefit of this method is that it provides a lot of dependable information about the condition of the transformer in a relatively short time.

Q: What is FDS testing and why is it so useful?

A: FDS testing is one of the easiest and most reliable ways of determining the level of contaminants – particularly water – in the transformer insulation. This form of testing involves, in effect, performing tan delta tests over a range of frequencies – typically from around 1 mHz up to 1 kHz. Tests usually take less than half an hour.

The latest instruments automatically correct the results for temperature over a wide range, which means there’s no need to wait for the transformer to cool after it’s been taken out of service. The instruments use the results to perform sophisticated modelling of the transformer insulation system and provide accurate information about the moisture content not only of the oil but also of the solid insulation. This makes it possible, for example, to distinguish positively between a dry transformer with bad oil and a wet transformer with good oil.

Q: Has the value of the SFRA and FDS test methods been confirmed by practical experience?

A: Yes! SFRA testing is recognised as a dependable test technique by many national and international standards including, for example, IEC 60076-18, IEEE PC57 and the requirements detailed in Cigré Brochure 342, 2008. FDS test techniques have been in use for almost two decades and are widely accepted by leading utility companies as an invaluable way of assessing the condition of power transformers in even the most critical applications.