- Energy transition
- Safe and liveable delta
- Effects of mining
- Digital subsurface
Using artificial intelligence (also known as AI), the Geological Survey of the Netherlands (GDN) can extract useful information about the subsurface from the large amounts of data it receives. We incorporate all this information into our models, which are available on open data platforms. In doing so, we are enabling companies to contribute to the energy transition.
Increased expected supply of data
As fossil fuel exploration and production in the Netherlands come to an end, we expect operators to leave the country in the coming years. Departing operators have the legal obligation to transfer all their data on the Dutch subsurface to GDN. This data is a valuable addition to the data already required by the Mining Act and the National Key Registry of the Subsurface (BRO), which is of great importance to enable new applications related to the energy transition. This new data may, for example, lead to the reopening of a depleted gas field for the storage of energy or carbon dioxide. It can also contribute to research on the application of (deep) geothermal energy.
Expanding data management using machine learning
Another important source generating new information is machine learning. Through machine learning, we teach our models to recognise and interpret patterns. We also offer this option for client data. In such cases, our AI network runs at the user’s site, where it is available for use. GDN does not have access to the data itself, but it does have access to the results of the model calculations. This is how we work together effectively.
Machine learning also allows us to extract additional information from old, scanned documents. For example, handwritten notes on a document about the well closures of old oil and gas fields can be analysed and included in research for other applications. In this way, old information can become relevant again for new developments.
Anomalies in time series
Artificial intelligence also offers interesting possibilities in terms of analysing the monitoring data (i.e. time series) of wells. For years, we have been recording data in the Netherlands about subsurface activities using borehole and surface measurements, such as those from geothermal plants. These series are becoming longer with time. By using artificial intelligence to search for exceptions in the data (so-called anomalies), we can discover where changes are taking place.
Much of the data is processed in subsurface models that map the (mechanical) behaviour of the shallow and deep subsurface. Combining various models from different fields creates model links, also known as ‘digital twins’. They combine models of the structure’s load-bearing capacity and expected traffic load and subsurface so as to give advice. Digital twins can, for example, provide insight into whether a bridge requires maintenance (or even replacement).
Open-source data platforms
GDN places the data and developed models on open data platforms, thus making them publicly accessible. Doing so stimulates the cooperation and innovation desperately needed for a range of social issues, including the energy transition.
Do you have a question? Our Service Desk can help you. Contact us via the blue ‘mail directly’ button below.