Sekundarni povzetek: |
In the following thesis, we address the determination of the condition or electrical wear of circuit breaker contacts based on protection relays records. Nowadays, these allow us to analyse a set of individual events on circuit breakers and, consequently, to assess wear and tear of the circuit breaker contacts through several criteria. In the past, the condition was monitored only visually, but today we have considerable amount of data, which, when properly processed, can be usefully employed to support the decision-making process. Based on the criteria applied to the collected data, we can estimate the date of revision or replacement of a specific asset in case of wear or obsolescence, thus contributing to the reliability and safety of the electricity transmission system, while at the same time automating the circuit breaker maintenance process as much as possible.
The thesis starts with a brief description of the electricity network, focusing mainly on the transmission network and the elements inside of it, and introduces an area of asset management in general and in ELES, the Slovenian TSO, where the thesis was carried out as part of the project.
Thesis also deals with the problem of ageing and electrical wear of circuit breaker contacts, the reasons for its occurrence and mitigation strategies. Some of the phenomena that have been investigated, which are common in practice and are indicative of contact wear, are mentioned, as well as methods of solving these problems and early detection of ageing, based on the application of appropriate criteria to the data collected. This is made possible by methodologies that have been scientifically or empirically proven and can be found in literature. The most appropriate methodologies are based on the I2 and I2t criteria, where we are mainly interested in the current during the arcing period, which is the primary cause of the electrical wear of the contacts.
This is followed by presentation and insight in protection relays records and their formats, which are parsed by an inhouse developed algorithm within the Jupyter Notebook. Algorithm plots oscillogram for each event and at the same time calculates all the necessary criteria and parameters for the later analysis of the circuit breaker contacts condition. These are also automatically aligned to particular transmission line system, substation and field, and then allocated to individual circuit breaker data tables.
Finally, the prediction results of the electrical contacts wear to date and in the future forecast are presented, based on the filtered events recorded in the individual tables mentioned above. Not all the events are useful for the analysis. Therefore, some filter must be applied to sort out the failure events that cause the circuit breaker to operate causing the electrical wear of the contacts. As the thesis is based on real life situations not the lab simulations, some simplifications during analysis are also made due to lack of data and events that cannot be categorized. Nevertheless, the results of the algorithm utilisation are much better than prior its development, when no information was available at all. Furthermore, its development, based on real life situations allows for its fast and smooth transition to production stage. Gained wear results were positive and indicate a good condition of the circuit breakers analysed, but it would be interesting to carry out an analysis on all ELES circuit breakers. |