Condition assessment of main insulation in transformer by dielectric loss data interpolation method and database building

摘要

Frequency Domain Spectroscopy (FDS) diagnostic technique is becoming more popular in recent years and is more suitable for onsite assessment. However, there are few reports about how to use FDS to assess the ageing condition of the oil-paper insulation conveniently. What's more, the lower frequency testing would be taken three or more hours. It's useful to reduce the testing time without effect the assessment result using new data processing method. In this paper, the oil impregnated insulation pressboard samples were thermally aged at 130°C for different times and then they were prepared for different moisture contents. The FDS measurements were taken on them and the dielectric loss of all the samples were calculated by dielectric loss interpolation. The ageing condition assessment database which contains five ageing levels and five moisture contents was built. When the dielectric loss interpolation curves of the oil impregnated insulation pressboard in a transformer mostly coincide with one curve or locate between two curves in the database, the main insulation ageing condition of the transformer is known. The evaluation method proposed is verified in this paper and the result shows that the dielectric loss interpolation method is dependable.