Transformer monitoring-moving forward from monitoring to diagnostics

The technologies employed for monitoring transformers have been evolving over the last 10 or more years to the point where they are now commonly accepted, and have been demonstrated to provide useful data on the key parameters and components of critical power transformers. A major aspect of these technologies has been the accumulation of copious amounts of data, and the associated problem, of what to do with it all. With time and resources in short supply to do a proper analysis of this data, to turn it into useful transformer information, there needs to be a new set of technologies and techniques implemented. The advent of new methods of data modeling and interpretation using statistical analysis, rule based, and artificial intelligence systems is now moving from the research stage to practical field implementation. The industry has a very real need to move from "just monitoring" equipment to the point of being able to have the knowledge of the operating condition of the equipment and when things begin to go wrong, diagnose the problem to provide a recommended course of action.