Identifying flexibility with AI
Consumer flexibility will be important in the coming years, but finding the customers that can offer flexibility in order to address specific challenges in the distribution grid can be a time-consuming and demanding task. Therefore, we will investigate whether artificial intelligence (AI) can be used to automatically find the customers that can contribute with flexibility.
CINELDI has previously established a pilot project with a flexibility market on the Norwegian island of Bremangerlandet in order to investigate whether flexibility can be used to solve bottleneck problems in the grid. NODES’ market platform was used for buying and selling flexibility. This pilot project tested flexibility for larger, industrial customers, but further flexibility was needed from smaller customers.
In order to automatically find the customers that can contribute with flexibility, we will pilot AI methods. Multiple AI and machine-learning tools for analysing data offline will be evaluated in order to find the one(s) that are most suitable.
The tools will use measurement data from both the customers (smart meter data - AMS) and substations, in order to identify where flexible consumption is available. By examining, for example, correlation with high loads in a substation, smart meter data will be used to identify customers’ load and potential to support the grid operation by disconnecting customer load. Data from the substations will be used to identify challenges in grid operation, and to identify loads when it is important to have a higher time resolution than the smart meters can provide.
When the project is finished, the results will constitute the basis of more general methods for identifying flexibility.
- , Linja
- , SINTEF Digital
- , Embriq