Vaffel - a tool for forecasting transmission line failures based on Bayesian updating, renalysis data and medium range weather forecasts

This paper presents a method for forecasting the probability of weather-related transmission line failures. The method combines failure statistics, a Bayesian updating scheme and reanalysis weather data to produce fragility curves that model the relationship between weather exposure for a component and the probability of failure. The fragility curves are combined with medium range weather forecasts extracted at each span of the individual transmission lines to obtain a forecast of the hourly probability of failure for each line. A tool that implements the method has been developed and is currently being tested operationally within the Norwegian TSO Statnett. We show the data requirements and demonstrate the use of a dashboard that expose the forecasts to an end-user.