How to advance data analysis and AI in the grid – dena recommendations

The Data4Grid project by German energy agency dena’s Future Energy Lab was aimed to advance the use of data analysis and AI in the distribution grids.

Data analysis and AI can deliver added value for grid operators and the system and is key for the transformation to renewables. However, while there are numerous use cases at the distribution grid level, they have not yet been put to widespread use.

The project brought together distribution system operators (DSOs) and start-ups to develop practical ways to advance specific use cases.

Among the findings were that through the collaborations significant progress could be achieved in the use of databased solutions and that the foundations could be laid for a more widespread rollout within a short period of time.

Have you read?
Horizon Europe: Five next-gen data projects
Data and creativity can help solve the energy transition’s biggest challenges, say DISTRIBUTECH keynote speakers

For example, it took just a few months for the development of practicable concepts for three central DSO challenges – increasing grid transparency, forecasting the development of electromobility and improving consumer forecasts.

The collaborations also highlighted some key issues that must be considered for the use of databased methods and the development of applications. One is that the electricity grid is a critical infrastructure and thus subject to strict requirements and the criticality of processes need to be assessed.

Another is that grid operators will only use innovative technologies if they are secure and deliver added value, which in turn requires sufficient knowledge and competence at various levels to assess these aspects. In particular in the small and medium size enterprises the deployment of digital technologies has been forestalled by inadequate knowledge and resources.

Third, the use of data analysis and AI must always be built on a solid database. At present, data availability and quality are inadequate for the broad application of data-driven solutions, especially in the lower voltage ranges. Moreover, general deficits persist in the collection, storage and processing of data.

Recommendations for grid operators

Based on the findings, the project report offers three recommendations for grid operators and policy makers.

● Improve the database.

The objectives are to ensure the availability of the company’s dataset and to enhance its quality for future grid operations and planning.

A central repository at the company makes the data more visible, simplifies linking and improves potential uses. Standardised interfaces as well as data models and processes at grid connection points also enable information sharing and the coordinated provision of grid and ancillary services.

● Strengthen data competence, share experience, exploit expertise.

DSOs must build data competence, including awareness of current legal and regulatory frameworks, as a basic requirement for improved data quality and the use of new technologies.

Data sharing between grid operators can promote knowledge and experience transfer and create added value for small to medium-sized distribution system operators in particular. Moreover, data pooling can create meaningful synergies for the development of new algorithms.

● Develop the regulatory framework.

Standardisation and certification ease the introduction of AI in the electricity grid, as they engender the necessary trust and security. This must be factored into research and development so that test procedures and risk assessments can be established and integrated simultaneously.

The general data situation can be improved by encouraging an open data mentality and facilitating uniform access to relevant datasets, such as GIS or geodata or building and heat cadastres.