In Data Driven Failure Risk Assessment for Predicting maintenance we are looking to do failure risk assessment of components in the grid.

A data analytics and risk-based approach will be used to scope and prioritize maintenance. We will make a FASIT analysis of faults in Elvia and use the FASIT reports as a source of faults occurred in the grid, and focus on assets that have limited failure modes. E.g. disconnectors, earthing switches or circuit breakers. A machine learning model will be trained to classify the occurrence of failures on assets over the next year.

By being proactive in predicting failures with data already available in different business unit systems, we want to quantify probability of failure for selected assets, relying initially on historical failure, usage and maintenance data. We want to make advanced maintenance processes possible and reduce the number of decisions that are too conservative or too risky.

Publications:

• Pilot "DADFRAP" report (in Norwegian)
• Pilot results

Contact person:

• , Elvia