Nonparametric approach enhancing ageing assessment of power transformers

Power transformers are of great importance by linking different levels of voltages and transmitting energy in generation, transmission and distribution networks. A good understanding of transformers' ageing condition can help asset managers to conduct effective maintenance actions, based on which transformers can be well maintained and fully used to realize its value in the power network. Hence, many different approaches have been proposed in assessing the ageing of power transformers. Among these approaches, a parametric approach named as analysis of variance-based factorial analysis, has been applied to study the effect of different factors on the oil acidity. In this paper, nonparametric approaches (Kruskal-Wallis and Wilcoxon rank-sum tests) are applied to enhance the analysis of variance-based factorial analysis. To prove the feasibility of the enhancement, a case study is done to analyse the influence of the in-service age, manufacturer and year of manufacture on oil acidities from 241 33/11 (or 6.6) kV transformers manufactured in the 1960s. It is observed that transformers with large in-service age or from some manufacturers tend to have high oil acidities. This observation helps asset managers to understand factors influencing transformer ageing for subsequent planning of maintenance strategies in transformer asset management.