SCOPE
Use of hydrogen as an energy source
Maturity Level
TRL 6
Development Level
Software
Protection Level
Interpretable Digital Twin for the Management of Hydrogen Injection into the Natural Gas Grid
Entity:
Our proposal is based on the integration of proprietary artificial intelligence (AI) technology within the operational framework of a digital twin. This technology, called Bayesian Machine Scientist, is an algorithm capable of generating models from real-world data. Compared to most AI algorithms, our technology offers two major advantages: (1) interpretability, allowing us to understand the relationships between variables influencing the processes and identify critical ones, and (2) extrapolation capability, enabling reliable evaluation of previously unseen scenarios and thus more accurate simulation of scaled-up experiments and long-term predictions.
Challenges met
- The main management challenge in injecting hydrogen into the natural gas grid stems from the desynchronization between production, transport capacity, and demand. This desynchronization is complex and involves many variables.
- Our technology generates models that can predict and optimize this intricate system, helping to maximize the amount of hydrogen that can be marketed and minimize system maintenance costs.
Scope of application
Companies operating gas logistics and transport infrastructure.
Main publications
- A Bayesian machine scientist to aid in the solution of challenging scientific problems (2020). https://www.science.org/doi/10.1126/sciadv.aav6971
- Bayesian Symbolic Learning to Build Analytical Correlations from Rigorous Process Simulations: Application to CO2 Capture Technologies (2022). https://pubs.acs.org/doi/10.1021/acsomega.2c04736?ref=pdf#
Related projects
- Green H2 Pipes. Task: creation of a digital twin of hydrogen production and transport systems based on renewable energy sources. https://exolum.com/proyecto/greenh2pipes/
Other related technologies
