On-site diagnostic method for the degradation of cast resin trasforrmer using optical reflectance.

Summary

We propose a new on-site diagnostic method for the degradation of mold transformer using optical reflectance. The mold transformers which use the epoxy resin for coil insulation have excellent characteristics of fire retardancy and environment resistance. From the viewpoint of electric equipment diagnosis, the mold transformers are treated as exceptional for diagnosis test in many cases because of its high lifetime reliability. However, since the mold transformer has been using for 40 years or more, the epoxy resin for coil insulation must be affected by aging degradation. Therefore, degradation diagnosis of the epoxy resin is required to ensure the operational continuity. IEC 60216-2 standard defines the heat life of epoxy resin as three characteristics: bending strength, breakdown voltage, and mass decrease ratio. From the results of accelerated aging tests, the mass decrease ratio of Fuji electric's epoxy resin is critical character for the shortest life. Furthermore, the fact that accelerated Aging test without oxygen hardly decreases the mass of epoxy resin suggests that oxidation reaction should be causing mass decrease. Based on FTIR (Fourier transform infra-red spectroscopy), the oxidization depth is approximately 100 µm at end of lifetime. Therefore, we consider that detecting of the mold surface oxidization could diagnose epoxy resin degradation of the mold transformer. From the study of surface diagnosis, we found that the mass decreasing has strong correlation with optical reflectance in the wavelength of 500 to 900 nm on resin surface. By using this relation, we developed portable device measuring optical reflectance and built the database that relation optical reflectance of mass decreasing about all Fuji electric's epoxy resins. These are made into one portable system named MOLMOS (Mold Transformer Multi-optical Diagnostic System), it is possible that we diagnose the mold transformer's degradation level quickly and on-site.

Additional informations

Authors

Wirth Isabell, Reumann Andreas, Zink Markus , Schnitzler Tim