Algorithm Aids Grid Stability and Smarter Energy Consumption

By January 1, 2027, all power grid companies must implement a time-differentiated power tariff, which means that the cost will be lower if the customer spreads out or shifts their electricity use to times when the load on the grid is lower, or higher if the customer uses electricity when the load is high. Planning for this change is already in full swing at Göteborg Energi Nät.

“Today, there is very high load on the power grid at certain times of the day and in the winter. At other times, the grid is used much less. By introducing differentiated charges depending on when electricity is used, the goal is to even out consumption and thus use the power grid more efficiently,” says Mihail Chigrichenko, Business Developer at Göteborg Energi Nät.

Requires Behavioral Change

Today there are bottlenecks in the regional and national power grids leading into the city, a situation that can lead to grid capacity constraints in Gothenburg within a few years’ time.  Reinforcements are in progress but will take too long to complete to meet the growing need for power. To level out consumption, electricity users need to change their behaviour. Time-differentiated power tariffs are an important part of this. Communication is another.

“We want to be able to identify and communicate directly with customers with a specific consumption profile. A simple example is targeting information to electric car owners who charge in the middle of the day and informing them that they can save money by charging at night instead. This makes our communication campaigns much more effective.”

Turned to RISE

By helping customers use power at times when the cost is lowest, Göteborg Energi can contribute to making the load on the power grid as optimal as possible. To gather data for this data-driven communication, Göteborg Energi turned to RISE.

“We have developed a data analysis model and algorithms that can analyse data from household electricity meters and identify customer segments based on different consumption patterns. This way, we can find groups that, for example, have electric cars, solar panels, or various types of heating systems,” says Claes Sandels, researcher at RISE.

The project used three years of anonymized data from 50,000 of Göteborg Energi Nät’s single-family homes’ customers. The methods have the potential to be scaled up to all electricity customers in Gothenburg. The results can also be used as a basis for power grid planning.

“By understanding how different customer segments change over time, Göteborg Energi Nät gains insights on where to allocate their resources. If the proportion of electric cars increases in an area, reinforcement might be needed there,” says Claes Sandels.

Continuing Development and Implementation

Mihail Chigrichenko is very satisfied with the collaboration with RISE.

“It has worked very well. In the future, we hope to develop the solutions together with other actors in the energy field. Power grid companies cover all of Sweden but do not compete, so there are good conditions for sharing results and continuing development in new projects,” says Mihail Chigrichenko.

One challenge for Göteborg Energi Nät is to actually use the tools and solutions that have already been developed.

“The big challenge is to integrate data-driven communication into the organization’s processes. It’s one thing to have access to data, another to use it. We need to make it easy to use the information in communication work or as a basis for decisions,” says Mihail Chigrichenko and continues, 

“Another important issue is customer privacy. As a company, we are primarily interested in data at the group level, even though we use the information to create benefits for individual customers. It is important that we are transparent about what data we use and how, so that we process customers' personal data in accordance with the Data Protection Regulation,” says Mihail Chigrichenko.

Other Customer Categories

Claes Sandels sees great potential for scaling up in the future.

“We have now looked at customers in single-family homes. In future research and development projects, we want to continue developing for other customer categories, such as industrial and office buildings. The vision is to be able to model an entire city to optimize its power and energy flows,” says Claes Sandels.