Search

This paper presents verification of frequency-response analysis (FRA) and vibroacoustic-method (VM) test methods complementarity for the assessment of the mechanical condition of a transformer's active part. Experiments performed on real power transformers under field conditions are discussed. The presented results cover three cases of faults which would be difficult to detect or correctly interpret with the application of a single method only. The first case shows the interfering influence of radiators on VM results. In the second case, a loosened winding with unchanged geometry, which cannot be detected by the FRA method, is discussed. The third presented case shows the results of an experiment based on introducing faults into the active part followed by FRA and VM tests. These faults were loosening of the core and loosening of winding with its slight stretching. The first fault could only be detected, as expected by VM results analysis, while the second one was confirmed by both methods; however, FRA assessment quality depends on the method of test data comparison.

This paper presents the concept of complementary measurements with the use of two methods that allow for a more detailed assessment of the mechanical condition of a transformer's active part. These methods are frequency-response analysis (FRA) and vibroacoustic method (VM). The VM was interpreted separately in steady and transient conditions, thus allowing, together with the FRA measurements, the determination of possible source of vibrations. This paper presents a theoretical introduction, experimental measurements performed on a real transformer with introduced controlled deformations and, finally, an example of a practical application.