Condition Monitoring and Diagnostic Assessment of Transformers

Condition monitoring and diagnostic assessment of transformers is a useful tool for managing power networks. When managing a large fleet of ageing transformers, asset managers have to be more innovative to integrate the data from a variety of condition monitoring systems into the decision-making process. Collective research efforts on this topic were made by a transformer research consortium in the United Kingdom. This paper provides an overview of the research in the following three aspects: analysis of large oil databases from field transformers, detection of methanol in oil as an advanced paper ageing indicator and transformer thermal modelling using computational fluid dynamic (CFD).
Observations from oil database analysis included a generic early degradation phenomenon, testing frequency of the common parameters, correlation of these parameters with transformer age as well as data interpretation in terms of incorporation of sampling temperature and potential revision to standardised values. Laboratory ageing experiments confirmed the suitability of methanol as a potential early paper ageing indicator in both an inhibited conventional mineral oil and a new inhibited gas-to-liquid hydrocarbon based oil. Measurements of methanol for UK field transformers showed a general increasing trend with ageing. Oil flow distribution in the windings has a direct impact on the cooling performance. CFD simulations and dimensional analysis were performed to study the occurrence of reverse flow in the winding. It was found that a higher oil flow rate does not necessarily mean a better cooling performance.