Improving the numerical indices proposed for the FRA interpretation by including the phase response

• Improving the existing indices by including the phase data in the FRA interpretation.
• Defining a new index to include the phase response in the interpretation.
• Demonstrating the method effectiveness using the experimental data.
• Validating the results with simulation models of three transformers.
• Comparing indices against measurement uncertainties for the first time.

The frequency response analysis (FRA) is an effective technique to diagnose the mechanical deformations in a power transformer. Using the numerical indices is one of the methods proposed for the interpretation of the FRA results. Various indices have been proposed in the literature, but most of them are only based on the magnitude response of the FRA results. By contrast, this paper improves a number of the existing indices and defines a new index to include the phase response in the interpretation. The experimental data corresponding to different extents of the axial displacement and disc space variation show that including the phase response in an index increases its sensitivity. These results are also validated with simulation models of three transformers. Afterward, different uncertainties are introduced in the experiment, and it is shown that the new indices have also better performances against uncertainties in the FRA measurement. Although further studies are needed to settle the new indices, this research shows the early functionality and the potential of the phase information to improve the interpretation of the FRA results.