transformers

Transformers

transformers
A model is being developed which, on basis of all available (incomplete) data on the transformer, gives a prediction of the expected remaining lifetime.

Transformers

Remaining lifetime of power transformers and transformer population

  • Thema: physical infrastructures
  • Onderwerp: Power electronics components and systems

Goal

Development of a predictive power transformer lifetime model based on knowledge of the degradation processes.

About the transformers research

A study is initiated on remaining lifetime estimation of power transformers and populations of power transformers.

Most transformers were installed several decades ago, and approach their end of life.

In view of the expected replacement wave of power transformers a reliable tool is required to obtain insight in the reliability of these assets.

Model

A model is being developed which, on basis of all available (incomplete) data on the transformer, gives a prediction of the expected remaining lifetime.

In addition, the uncertainties in the prediction are determined.

The key concept is to define proper quality parameters for the relevant degradation mechanisms.

Updating their values can increase the prediction reliability. The model can be applied to individual power transformers as well as to population of transformers

Model features

  • The best possible prediction of future failure probability given all data available
  • Well-substantiated support in reviewing scenarios for operational, tactical and strategic decisions
  • Account for the effects of regeneration (maintenance) and degeneration (degradation)
  • Recognizes causes of uncertainty from a sensitivity analysis and suggests measures to reduce uncertainty
  • Applicable to individual transformers and to transformer populations

Discussion & Conclusion

  • The proposed probabilistic method enables asset managers to optimize their policy on maintenance and replacement
  • Model allows to evaluate a representative population of over 500 units
  • Small deviations in historical date and future scenarios have large effect on the predicted reliability

Future work

  • Degradation mechanisms: extend with degradation of bushings and tap-changers; currently only paper degradation is included
  • Additional options: Define a number of restrictions and the model will create the optimum strategy

Partners

Eindhoven University Electrical Engineering, DNV GL, Essent, E-Netbeheer

Scientists

Reageer op dit artikel

Your email address will not be published. Required fields are marked *