Summary
Skagerak Kraft is a Norwegian hydropower producer with 52 wholly and partially owned power plants across the south of Norway, producing an annual average of 5.9 TWh – roughly the consumption of 300.000 households.
Skagerak wanted to make it easier for everyone in the company to find and use data. The company partnered with Cognite to accelerate and add to its DIY data lake solution, providing accessible and actionable insights that drive operational and environmental improvements for Skagerak.
Challenge: Managing Complex Operations and Fluctuating Data
Skagerak water reservoirs are its primary resource. The physical power markets in which it operates incentivize Skagerak to spend the water when the demand is the highest. Skagerak also plays a vital role in preparing the river basins it operates in for seasonal floods to minimize their damage impact on wildlife.
To do all this effectively, Skagerak operators need accurate forecasts of power demand inflow and available production capacities. Skagerak’s power stations must be ready to meet peak demand by monitoring and receiving early alerts for potential critical issues. It’s also vital that Skagerak plans downtime according to market forecasts.
This can be a challenge, as the operational patterns for Skagerak power plants fluctuate, and it is difficult to have reliable access to asset data.
Solution
Breaking Down Data Silos and Empowering Field Workers
Skagerak had long known that even minor improvements in its inflow and price forecasts could significantly impact its revenue. In addition, Skagerak understood that they needed to conduct better water management and calculations of water flows to stay within the water waste restrictions on water flow. Data is key to achieving both of these goals. Skagerak needed to break down data silos and create an easier way to find and interact with mission-critical data.
The company set its strategic direction for data value maximization. It aims to make it easy for everyone in the company to find the data they need and process the data with appropriate tools.
Data value maximization strategy helped develop a good workflow for extracting telemetry data into a data lake and built a proprietary operations portal, which its field workers widely use. However, the data lake quickly became its own silo.
Delivering on needs beyond what the operations portal could provide, such as root cause analysis or troubleshooting, proved challenging. As a result, the data team often received requests to extract sensor data into Excel for more in-depth analysis.
“We saw that there was an unfulfilled potential in liberating the data even more, such that the people needing data would have faster access to this data and support data-driven decision-making.” - Kristian Nymoen, Head of Data Management and Data Science at Skagerak Kraft.
To fully deliver on the strategic direction for data value maximization, Nymoen turned to Cognite to make data more accessible across business domains. Using Cognite Data Fusion® as the data foundation, Skagerak combined its technical and domain knowledge to better manage and monitor its water and assets— uncovering valuable insights and generating the highest impact where it mattered the most.
With reliable, easily accessible data in Cognite Data Fusion®, Skagerak operators could build solutions to meet their specific needs, including a Grafana dashboard for tailored monitoring of a particular power plant. The dashboard was developed by a field worker with unique knowledge of the power plant, making it easier for them to plan activities for that power plant on any given day without needing to make data requests to the data team.
Skagerak also uses Cognite Charts to monitor when the water level difference across a trash rack gets too high. These intakes collect debris floating down the river; if a lot of debris builds up, it's more difficult for the water to get through. Skagerak uses the simple calculation function within Charts to quickly understand the water level on either side of the trash rack and how that changes over time. Skagerak can also set a monitoring trigger to alert operators when it's time to clear the trash rack.
Adding Cognite Functions into Charts, Skagerak can calculate and analyze water flow from time series data on water levels. Using inputs such as “water level time series” and “parameter tables,” Cognite Functions processes the data to produce a time series of water flow, which is then stored in Cognite Data Fusion® and visualized through Cognite Charts.
This solution enables Skagerak's operators to make more proactive and independent decisions, improving inflow model tuning and inflow forecasts for higher precision in power market operations. Additionally, it supports environmental monitoring efforts, helping to mitigate the impact on wildlife and ensuring sustainable practices.
Impact
Improving Operations while Building a Future-Ready Data Foundation
In the early stages of adoption, Skagerak saw incredible value in enabling engineers to efficiently locate the data they need for root cause analysis and monitoring. Cognite Data Fusion® is set to remain a central component of Skagerak’s data infrastructure, serving as a hub for ensuring data quality and integration across business domains.
Moving forward, Skagerak plans to explore other Cognite Dta Fusions’ capabilities. They have started their transition to data models, linking documents, 3D models, images, and other sources to enrich the data further and enhance user accessibility. Skagerak is also eager to leverage Cognite Atlas AI™ to accelerate the adoption of agents to drive even more efficiencies in managing water level time series, detecting sensor failures, and providing more robust decision-making support for production and maintenance planning.