Reliability Study of Furan Level Analysis for Transformer Health Prediction

التفاصيل البيبلوغرافية
العنوان: Reliability Study of Furan Level Analysis for Transformer Health Prediction
المؤلفون: Christof Sumereder, Maximilian Meissner, Martin Mittelbach, Martin Darmann, Sigurd Schober
المصدر: 2019 IEEE 20th International Conference on Dielectric Liquids (ICDL).
بيانات النشر: IEEE, 2019.
سنة النشر: 2019
مصطلحات موضوعية: 010302 applied physics, Pressboard, Materials science, Solid insulation, 01 natural sciences, Standard deviation, law.invention, chemistry.chemical_compound, chemistry, law, Furan, Reliability study, 0103 physical sciences, medicine, Composite material, Mineral oil, Transformer, Kraft paper, medicine.drug
الوصف: To determine the reliability of furan compound analysis for insulation-system health prediction, the furan level found in oil, paper and pressboard was monitored throughout a large-scale ageing study. Therefore, thermally upgraded and normal Kraft paper as well as transformer pressboard were immersed in 4 different types of insulation liquid (mineral oil, G-t-L oil, synthetic and natural ester) at 3 different ageing temperatures (110, 130 and $150^{\circ}\mathrm{C})$ and studied throughout a 56 day ageing setup, resulting in 100 mixed samples and 300 furan analyses. While the evolution of furan compounds in each system was somehow steady, comparisons between different material mixes and temperatures show significant aberrations. For similar stages of wear, standard deviations of 100% and more are detectable for the total furan compound concentration in oil, while deviations in solid insulation materials can be even higher. Due to the complexity and high number of variables, irregular analysis intervals without supporting information seem to be highly arguable for transformer health prediction.
URL الوصول: https://explore.openaire.eu/search/publication?articleId=doi_________::4ec12607e59e42520175942a7a692723
https://doi.org/10.1109/icdl.2019.8796785
حقوق: CLOSED
رقم الأكسشن: edsair.doi...........4ec12607e59e42520175942a7a692723
قاعدة البيانات: OpenAIRE