The use of artificial intelligence (AI) in the site selection process for a deep geological repository



In a science-based site selection procedure (StandAV), the Federal Republic of Germany searches for the site that guarantees the best possible safety for the final disposal of high radioactive waste (HAW) over a demonstration period of one million years. In this process, the Bundesgesellschaft für Endlagerung (BGE) determines the most suitable site for a repository in an iterative process. For this purpose, the geological subsurface of Germany must be investigated for potential sites and evaluated. Challenges in this process include the varying quality and level of detail of existing geo-data, as well as in the collection and evaluation of new large amounts of heterogenous geodata that must be processed.

The application of AI-based methods offers opportunities to make the evaluation and thus the StandAV more efficient and safer when


the large data sets and model calculations of complex long-term and coupled, geological processes. Accordingly, research on the application of AI has also increased significantly in the geosciences over the last few years.

Based on a comprehensive international literature review, this study identifies areas of application of artificial intelligence (AI) in the geosciences in general and evaluates them with regard to their use for geoscientific questions in the StandAV. In addition, limitations and necessary conditions arising from the risks of using AI are formulated with respect to key activities in StandAV.

The results of the study show that the AI application areas from the literature are generally only transferable to the geoscientific


in the StandAV with methodological and subject-specific adaptations. The evaluation also shows that AI can only be used as a supporting tool in key activi-ties but may not have any decision-making power when used in StandAV.

Opportunities are offered by AI especially in data management and model calculation of complex and long-term geological processes under


condition that the procedure of AI can be presented and validated transparently. Therefore, the study also identifies further research questions.


More information about the project

Status of project

End of project: 2022

Project manager

Project staff

Funded by

Federal Office for the Safety of Nuclear Waste Management (BASE)

Project partners

Clausthal University of Technology