Principal Consultant Energy Consulting
Qirion, Alliander,
Netherlands
For a small population of assets traditional statistics are often not sufficient as the law of large numbers does not necessarily apply0 In this case Bayesian methods can help as they do not rely on any assumptions and make the assumptions taken very explicit0 Furthermore, these methods give a detailed view on uncertainties upon modelling0 We will show that for asset populations with little available failure data such as high voltage transformers Bayesian methods can make the difference and allow better risk assessment and more accurate predictive maintenance scheduling/ Due to the lack of available failure data of the components the classical methodology is not sufficient to accurate determine the failure probability0 This Bayesian approach gives better insight in the failure probability of assets and is a valuable input for a solid risk assessment and predictive maintenance scheduling