Transformers - Transformer oil testing: here or there?

Typography
Articles

Most engineers would agree breakdown testing provides the most reliable indicator of the condition of the oil in a power transformer. There is, however, far less agreement about whether the testing should be carried out on site or in the laboratory. Paul Swinerd of Megger looks at the arguments for and against each of these options

The condition of the oil in a power transformer is a major influence on the transformer's reliability, operating life and even safety. A dependable and convenient method of assessing oil condition is, therefore, an essential adjunct to transformer operation and maintenance.
Various options are available to meet this requirement including, for example, the Karl Fischer coulometric titrimitry method that can be used to quickly determine the moisture content of the oil. This test is used frequently as water contamination is the most common cause of oil degradation. The most direct measure of the oil's ability to perform adequately as a dielectric medium is, however, given by breakdown testing.

In breakdown testing, a sample of oil taken from the transformer is transferred to a test vessel, which is then loaded into the breakdown tester. Typically the instrument will then carry out a series of tests in a pre-programmed sequence determined by the oil testing specification appropriate to the application.

In addition to the application of test voltages - usually in the tens of kV range - to electrodes immersed in the oil, the test sequence will also include predetermined stirring and standing times.

Breakdown test sets that operate in the way described are available in laboratory versions, and in portable versions that are designed for convenient use in the field. Some manufacturers, including Megger, also offer instruments that are equally well suited to use on-site and in the laboratory. But which is preferable - laboratory testing or field testing?
In order to understand the arguments for and against each approach, it is first necessary to appreciate contamination of the oil sample has a large effect on the accuracy of the results obtained in a breakdown test, with even a tiny amount of contamination making the results unreliable and, therefore, unusable.

Some engineers argue this means it is best to carry out testing on site. Their rationale is that, for laboratory testing, the oil sample has to be bottled and sent to the laboratory, and there will always be doubt about whether the bottle was adequately cleaned before use, and whether it was sufficiently well sealed to guard against contamination in transit.

There are other engineers, however, who will point out the sample is at most risk of being contaminated while it is being collected, and that the contamination risks associated with bottling and transportation are, by comparison, relatively small. Their conclusion is there is no significant difference between the overall contamination risks for on-site and laboratory testing.

Proponents of laboratory testing will also argue, once the oil sample reaches the laboratory, it will almost certainly be tested by a skilled technician who will fully understand the procedures and precautions involved, and will follow them carefully to ensure accurate results are obtained.

On the other hand, tests on site are frequently performed under less than ideal conditions, and there is often pressure to complete the testing process as quickly as possible. These factors are conducive to error, especially if the person performing the tests carries out breakdown testing only infrequently.

Nevertheless, there is one important issue that most definitely favours testing on site, and that is the speed with which results can be obtained - typically within an hour of the sample being taken, and often much faster.

This almost immediate availability of results has two important benefits. The first is that if an unexpected or obviously incorrect result is obtained, the test can usually be repeated at once. The second benefit is, if the tests confirm the oil is in poor condition, the transformer can be taken out of service straight away, thereby reducing the risk of failure.

While important, however, these benefits should not be interpreted to mean on-site testing is always to be preferred. There are most definitely cases where on-site testing is impractical, or where the certainty of tests being carried out consistently and with a high degree of precision outweighs the advantage of obtaining immediate results.

The best advice for those considering the implementation of breakdown testing for transformer oil is, therefore, to consider both the laboratory and on-site options carefully in relation to the application in hand, before making a decision.

Suppliers of oil test sets will undoubtedly be pleased to provide assistance in making this decision but, to be sure of receiving impartial advice, it is most certainly advisable to choose a supplier, like Megger, that offers both portable and laboratory instruments.

Data Centre news from our sister title

@elecreviewmag