For decades the transformer industry has used dissolved gas analysis (DGA) to assess the condition of transformers. Whenever a transformer undergoes abnormal thermal and electrical stresses, certain gases are produced due to the decomposition of the transformer oil. By observing the percentages of different gasses present in the oil, the internal condition of the transformer can be statistically assessed. Even though DGA provides evidence that there is something abnormal going on in the transformer, it has its limitations:

• Specialized knowledge of standards such as IEC60599 and IEEE C57.104 are needed interpret the results.
  
  
• DGA is a lagging indicator of the condition of the transformer i.e. DGA will only indicate major faults when the transformer has already started deteriorating to a critical point.
  
  
• DGA cannot pinpoint the area in which the fault is occurring.
  
  
• If the transformer has had an oil replacements, then the results won't reflect previous fault conditions.
  
  
• Test results are highly influenced by the methods of extraction: time of day, temperature of the oil, location of the sample point, and the instruments used all impact the result.
  
  
• Often absolute gas concentrations are insufficient to make any determinations about the health of a transformer. Instead, the historical rate of gas change over time is analyzed. For this to occur, DGA needs to be conducted routinely and under the same conditions to obtain comparable results. Accurate interpretation of these results also requires knowledge of the age of the unit, the loading cycle, and major maintenance such as filtering of the oil. It can be particularly difficult to maintain the required consistency when managing a fleet of transformers over the long lifetime of a transformer.

EcoStruxure Transformer Expert can help overcome some of these limitations:

• Single web-based online platform to concurrently view online and offline data and manage transformer health over transformer life, across your fleet, whether that is just one or hundreds of transformers.
  
  
• EcoStruxure Transformer Expert algorithms automatically analyse and interpret the DGA results according to the IEC standards and displayed in a graphical way so that trends are clearly identified. These can be understood by Asset Engineers and can remove the need for a specialized manager.
  
  
• Concentration of gases and oil quality statistics can be compared over years through graphs, enabling the identification of escalating fault conditions.
  
  
• Insights and Recommendations based on the DGA analysis are provided to each user, with configurable automated notifications of conditions of concern.
  
  
• Health Indices clearly rank the transformers in the fleet for replacement and/or maintenance attention.
  
  
• The analysis model combines online moisture, temperature, partial discharge and vibration data with DGA results to provide a comprehensive overview of transformer health. As offline data capture often does not capture the extremes of the transformer operating environment, the EcoStruxure Transformer Expert dashboard enables the user to see the impact of extreme conditions, overlaid with offline results on the same graph.  Together, the two sets of data provide a more complete picture of transformer health and its degradation over time. Insights and recommendations based on both sets of data are provided to users on how to operate a transformer to maximize efficiency and asset utilization.
  
  
• The use of Furans from offline oil tests alone to determine the Degree of Polymerisation (DP) of the paper is notoriously inaccurate. The analysis uses these multiple sources of offline and online data to provide additional accuracy and Insights not available from standard DGA testing alone, including plotting the estimated aging profile of the paper insulation.
  
  
• Multiple sources of data (DGA, oil test, loading cycle, sensor data) are correlated to increase reliability of results.
  
  
  See attached presentation at the bottom of this article to see some examples.

Released for: Schneider Electric Philippines