Modelling weather dependence in online reliability assessment of power systems

Samuel Perkin*, Arne Brufladt Svendsen, Trond Tollefsen, Ingrid Honve, Iris Baldursdottir, Hlynur Stefansson, Ragnar Kristjansson, Pall Jensson

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)


Probabilistic reliability assessment of power systems is an ongoing field of research, particularly in the development of tools to model the probability of exogenous threats and their potential consequences. This paper describes the application of a weather-dependent failure rate model to a region of the Icelandic transmission system, using 10 years of weather data and overhead line fault records. The studied failure rate model is compared with a constant failure rate model, in terms of variability and how well the models perform in a blind test over a 2 year period in reflecting the occurrence of outages. The weather-dependent and constant failure rate models are used as input to a state-of-the-art risk assessment tool to determine the sensitivity of such software to weather-dependent threats. The results show the importance of weather-dependent contingency probabilities in risk estimation, and in quantitative assessment of maintenance activities. The results also demonstrate that inclusion of weather dependence in power system reliability assessments affects the overall distribution of risk as a positively skewed distribution, with high-risk periods occurring at low frequency.

Original languageEnglish
Pages (from-to)364-372
Number of pages9
JournalProceedings of the Institution of Mechanical Engineers, Part O: Journal of Risk and Reliability
Issue number4
Publication statusPublished - 1 Aug 2017

Bibliographical note

Publisher Copyright:
© Institution of Mechanical Engineers.

Other keywords

  • Environmental risk
  • failure data analysis
  • network reliability
  • probabilistic methods
  • system simulation


Dive into the research topics of 'Modelling weather dependence in online reliability assessment of power systems'. Together they form a unique fingerprint.

Cite this