Current industry statistics show that one in 10 transformers will develop a fault in the next 16 years. If the power system is to be operated with maximum reliability and efficiency, it is clear that these faults must be detected, diagnosed, and rectified before they can adversely affect system performance.

The majority of faults are slow to develop and can be detected by periodic DGA testing [CIGRE TF 15.01.01]. In these cases annual data gathering and equipment evaluation is well-justified. However, some assets are sufficiently valuable that continuous monitoring becomes essential to provide the maximum protection against failure, but this can produce its own problems by virtue of the sheer bulk of data generated.

The best condition assessment can only come from the analysis of data from all available sources, e.g. DGA, visual inspections, fluid quality tests, and electrical tests. This data must be evaluated using criteria and rules in the context of both current and past operations in order to reach a global conclusion for each piece of equipment. A large quantity of information from electrical tests, dissolved gases, fluid quality, etc, is required to evaluate equipment health.

Kelman's diagnostic software Perception can provide can improve the speed and accuracy of the the information available to maintenance managers and provide a tool to keep the system operating at maximum efficiency.

Perception software addresses the following key points:

  • Automatic data collection and processing allows for rapid and early detection of potential equipment failure, maximizing the time to respond and therefore minimizing the costs of the remedial actions required.
  • Many utilities are facing the prospect of losing their experienced, skilled engineers. Not only is this expertise difficult to replace, but new research in diagnostic techniques is constantly being added to the industry as the transformer aging issue becomes more widely recognized. Perception uses the current state-of–the art analysis techniques to ensure reliable information.
  • To the inexperienced, test data can appear indecipherable and the basic knowledge necessary for its evaluation may seem inaccessible. Perception demystifies the routine tasks of data analysis and guarantees correct and uniform application of the standard evaluation criteria.
  • Clear recommendations for the appropriate actions are provided.
  • Computerized management of data can minimize or eliminate transcription errors and lead to more accurate analytical results. Perception will interface directly with all Kelman on-line and portable DGA systems.
  • A typical utility will have hundreds, if not thousands of transformers. Simply managing this amount of data can present real challenges for maintenance managers. Perception makes data management into a routine task.
  • By examining the processes used in the analysis of data users can begin to learn the basics of the diagnostics techniques involved.

Perception can provide assistance in the management of the great majority of transformer operation and maintenance issues common in the utility industry today. On the few occasions where a problem is presented with multiple or conflicting diagnosis it may become necessary to perform additional tests or obtain further expert opinion to achieve the most accurate results.