In recent years, increased attention to safety and environmental protection policies has directed grid operators and Utilities toward new “green” solutions of power transformers immersed in insulating oil, ester (natural or synthetic). Power transformers insulated in ester oil can also be driven to operate at higher temperatures, with obvious advantages in terms of overloading.

However, the degradation mechanisms of the solid ester oil-paper insulation system, especially under severe transformer operating conditions, are still poorly studied and there is a need to identify appropriate techniques and devices for monitoring the by-products of paper degradation (moisture and furans) in esters. In medium-voltage networks, where resin transformers are more common, investigation of appropriate diagnostic methods to detect solid insulation degradation is still an open field of research.

The activities described in this report are aimed at a deepening of the issues initiated in the previous project and are aimed, as far as power transformers are concerned, at the development of new diagnostic systems, based on the dielectric method and the optical-chemical method, to study the effects on the ester oil-paper insulation system of accelerated aging under temperature, which will be reproduced in a test circuit set up at the RSE laboratories.

In preparation for the laboratory tests, this report has analyzed the dielectric response at different frequencies of new winding samples with unimpregnated paper insulation as a function of the initial drying process of the paper; it has also analyzed the behavior of the optical response to furans in samples of virgin natural ester with the addition of fatty acids; and finally, it has designed and fabricated the paper samples that will be used as sacrificial units to destructively analyze the degradation of the paper.

Regarding fault diagnostics in resin transformers, further improvements have been achieved in the application of vibration measurements and new machine learning methods for predictive analysis of transformer degradation profile from vibrational data.