On Tuesday April 27th Statnett and Elhub hosted the sixth annual TechDay, the Norwegian energy sector’s very own IT-conference. Both representatives from the hosts and several other actors in the business were present.
Developers, testers, system- and solution-architects, and many more who work with IT systems gathered to share ideas and present their work at #techday. The overarching theme for 2025 was “Faster, stronger, safer: Digital technology as a key to efficient electrification.”
The Data Science section was also well represented and participated in six different presentations and talks. These are listed below in order of appearance. Luckily, if you were unable to be ther in person, all presentations were recorded! The presentations where contributions were made from Statnett’s Data Science section are linked below in chronological order of appearance:
– Hybrid AI for power transformers
Using neural ODEs representatives from the research center NorwAI demonstrate prediction of transformer oil temperatures. This work produces approaches to better estimating the lifespan of transformers which, in turn, could lead to more cost efficient power grid operation.
– How to increase transfer capacity of powerlines by 10% from your office desk
Som og the work from the team focusing on dynamic line rating, leads towards sensorless methods for estimating line capacities. Fruits of this labour are presented here. How different data sources contribute to the final line rating is only one of the many interesting topics of this presentation.
– Teknologioptimistene
The Data Sience section head, Ingeborg, guested the Teknologioptimistene podcast during the conference. She shared her thoughts on how to better exploit the potential of artificial intelligence technologies while at the same time safeguarding society from digital threats. We visit different topics, from how will AI affect job security, to how it can aid us in operating critical systems.
– Voltage Transformers Monitoring: Data-Driven Automation for Maintenance
One approach towards cutting costs is to properly maintain equipment. Manual inspections are costly and time-consuming, and hence this is a prime target for automated support. This presentation focuses on methods for detecting faults in transformers based on patterns in the voltage readings of the transformer components.
– Maskinlæring og sanntidssimuleringer i balanseringen av kraftssytemet – erfaringer etter 2 måneder i skarp drift
Participants from the “Autobal” project on automating the activation of mFRR (manual frequency restoration reserves) share experiences of the first weeks of operation after Go-Live on March 4th. There is an emphasis on the imbalance prognosis and bottleneck handling modules of the project.
– Tidsseriemodellering av Norges strømforbruk med støyete måledata
The team operational prognosis presents their work on using different statistical models for making bidding zone level consumption forecasts and evaluating which is the best performer when using noisy data. Which model will win?
If any of the above gave you some FOMO, make it a goal to attend a future #techday!
Leave a Reply