Application of statistical control charts to discriminate transformer winding defects

A common problem in power transformers is the winding defect which is caused by a variety of internal and external conditions. It is necessary to detect and rectify the winding defect as soon as possible, because they are inherently progressive. One of the effective ways used to diagnose mechanical deformations of power transformers is Frequency Response Analysis (FRA). FRA has been standardized in previous studies but interpretation of the results is still being studied. Recently, results interpretation with the purpose of making the results more objective and quantitative has attracted attentions. To this end, this study presents a novel method for classifying faults which compares transfer function indices called statistical control chart. In order to solve this problem, for the first time, X-bar chart, R-chart, and S-chart with eight tests are used to interpret FRA results to detect different short circuits, axial displacement and radial deformation. The results of the experimental data show that visual fault detection is increased and classification accuracy is improved verifying high performance of this method in detecting and determining various faults.

Automated summary

The winding defect in power transformers needs to be detected and rectified quickly, with Frequency Response Analysis being an effective diagnostic method. This study introduces a novel approach for fault classification by comparing transfer function indices using statistical control charts.