| Abstract [eng] |
This final project examines technical condition monitoring solutions for autotransformers in the Lithuanian electricity transmission network. The aim of the project is to assess the main defects of autotransformers, their diagnostic indicators, and the application possibilities of advanced diagnostic methods to identify deterioration of technical condition at an early stage and reduce the risk of emergency failures. The project analyses the importance of autotransformers for reliability, operational safety, and stable quality of electricity transmission. It was determined that, due to their complex design, high economic value, and strategic importance, autotransformer failures may cause significant technical and economic consequences. Therefore, their condition monitoring is an important part of transmission network operation and asset management. Based on the analysis of scientific literature and diagnostic methods, the main groups of autotransformer defects were identified: ageing of the insulation system, thermal and electrical faults, winding deformations, bushing failures, and on-load tap changer defects. It was found that some of these defects do not show clear external symptoms in their early stages; therefore, periodic visual inspections alone are not sufficient. For this reason, complex diagnostic methods must be applied to assess the technical condition of autotransformers. The focus of the project is placed on dissolved gas analysis (DGA) and partial discharge (PD) diagnostics. DGA is considered one of the most important methods for detecting internal thermal and electrical faults, as gas concentrations, gas ratios, and their changing trends make it possible to assess the type of fault and the intensity of its development. Partial discharge diagnostics allows the condition of insulation to be evaluated, while the combination of electrical and acoustic measurement methods increases the possibilities for defect detection and localization. The project also examines the application of machine learning methods for the interpretation of DGA and PD data. The literature analysis showed that artificial intelligence methods can improve fault classification, reduce the influence of noisy data, and create conditions for a more automated assessment of technical condition. It was determined that the combination of classical diagnostic methods and advanced data analysis tools is a promising direction for autotransformer monitoring. 6 In conclusion, complex monitoring of the technical condition of autotransformers, based on DGA, PD, and data analysis methods, enables earlier identification of possible defects, supports maintenance decision-making, and helps reduce the risk of unplanned failures in the Lithuanian electricity transmission network. |
