Will make better decisions with AI

The power companies will use machine learning and artificial intelligence to improve the utilisation of hydropower.  

Through the use of innovative software, they also hope to respond faster to operational changes.

“Artificial intelligence and machine learning will be particularly important in the future, since the need of flexible power is increasing and the physical trading will happen more and more continuously," says Stein Danielsen of Powel, who is also leading the Flexploit project.

Increased values of flexible energy

Flexploit will apply machine learning and artificial intelligence to improve market utilisation of available flexibility from hydropower. At the same time, the innovation project will automate experience-based decisions.  

Powel is leading Flexploit, and together with Uniper, Glitre Energy and SINTEF Digital, they hope to find answers to many of the challenges the power industry is struggling with today.

“The green shift is likely to increase the need for, and the value of flexible energy. This is due to an expected increase in solar- and wind power production, which by nature is unpredictable and fluctuating. In combination with changes in electricity consumer patterns, these factors will cause an increased demand for flexibility in power production,” says Danielsen. 

Uniper is one of Europe's largest power producers and the third largest hydropower producer in Sweden. They are already using several of Powel's software solutions to streamline their operations.

“The optimization we do today needs to be further developed, in order to meet future demands in terms of speed and complexity. If we are to succeed with this, we have to automate our processes,” says Edvin Schubert of Uniper.

Powel invited recentlig their partners in Flexploit for a meeting. Boris Tistan was informing about the innovation projects plans for technological development. 

More frequent trading 

Traditionally in the NordPool area, the sale and purchase of power has been done by daily bids to the power exchange, with the option of supplementing with intraday trading with gate closure once an hour. Soon, this will turn into a continuous market with gate closure four times an hour, significantly increasing the need for automation for power companies.

“We believe our applications can increase the earnings of the hydropower producers by three to five percent. To succeed, we must co-create with our partners, which we have a unique opportunity through Flexploit,” says Danielsen.

The Norwegian power company Glitre Energy believes the software developed through Flexploit will improve the decision making of their operators.  

“When the software solutions include better ways of visualising data, the operators will easier understand the results. Thus, they can make the right decisions faster and easier in order to ensure efficient power production, says Arnfinn Brede of Glitre Energy Production.

Erlend Moe from Glitre Energi was listening carefully when Powel talked about the innovative technology they are working with in Flexploit.


The innovation project has three focus areas: 

 • Use machine learning to improve the accuracy of predicting available resources, both regulated and unregulated inflow, as well as time delays in the watercourses.  

• Use artificial intelligence to determine optimal production plans that can be applied in practice. 

• New, innovative ways of visualising results that will increase user confidence and thus enable better decision making.