RECENT ADVANCES IN THE CONDITION MONITORING AND ASSESSMENT OF POWER TRANSFORMERS

With ever increasing pressure on resources, achieving a safe and reliable power supply as efficiently and costeffectively as possible is a major goal for any organisation operating electricity transmission or distribution networks. Power transformers are normally very reliable items of electrical equipment, but when faults occur they can develop catastrophically, resulting in the loss of what is usually the most expensive item in the substation, with the likelihood of high forced outage costs and also possible collateral damage from any fire or explosion. Clearly then, despite' the generally good reliability of transformers, in view of the potentially serious consequences of failures it is important that effective condition monitoring techniques are employed so that faults can be detected at an early stage so as to improve the prospects for repairs and minimise the impact of any failures. A particular problem with transformers is that internal visual inspections are very expensive because of the oil handling required and are also often inconclusive because very little of the active part of the unit is accessible for inspection. Therefore condition monitoring of transformers has to be based on non-invasive techniques, which have to be sophisticated enough to take into account the complicated constructional aspects. In addition to the basic need to detect faults at an early stage, there is increasing pressure for monitoring techniques that enable maintenance to be postponed, or a suspect unit to be left in service. There is also an increasing need for techniques to enable the condition of transformers to be assessed, in particular to determine the aged state of the equipment, to provide information on which to base asset management decisions over the long term.

RECENT ADVANCES IN THE CONDITION MONITORING AND ASSESSMENT OF POWER TRANSFORMERS Keywords:

Transformers , Condition monitoring , Dissolved gas analysis , Furfural analysis , Frequency response analysis