09 January 2023
Predictive Asset Management to avoid incidents with energy not supplied on aged asset systems
Older grid parts with assets are forced to extend their lifetime due to the reality of capex and workload arbitrage.
Why this project
The goal of this project realized in The Nest, our digital incubator, was to find and explore innovative predictive asset management solutions to increase the level of monitoring of these assets, so we can extend their lifetime in a more controlled way and avoid incidents with energy not supplied (ENS).
Approach
A dedicated multifunctional team was setup with Asset Managers, Experts from the field, Data Scientists, Data Experts, Developers and Agile Coaches. Using the ‘Scrum’ and ‘Kanban’ methodology, the team was able to complete a set of user stories to complete three scoped tracks: New Data, Analysis at System level and Condition Monitoring of Circuit Breakers
-
New Data
Enriched failure curves data based on incidents in the last 10 years and additional data sources (asset features, weather, announced works, Malfunction Reports, etc.).
-
Analysis at system levelDeveloped a model for single asset failures in a system and visualised the probability of a circuit’s system failure.
-
Condition Monitoring of Circuit BreakersImproved condition monitoring for circuit breakers with online monitoring using IoT, Artificial Intelligence and Machine Learning in a partnership with Fingrid.
Results
The proof of concept (PoC) has shown the potential of calculating the system failure probability of a circuit, based on the probability of failure of the single asset and the potential taking announced works by 3rd parties into account to prioritize actions and avoid incidents with ENS. Further developments is are required and will be taken on by another internal project, the so called Enhanced Transversality. On the other hand, the PoC has shown that (some types) of circuit breakers tend to get less and less reliable as they get older. The PoC and the partnership with Fingrid will now need to prove that online monitoring is a valuable solution for accurate condition monitoring of the asset, so that we can take timely action and avoid circuit breaker malfunctioning.