AI contribution for complex phenomena predictions in model porous media

In New Energy Technologies (NET) fields, answering numerous matters and managing applications like geothermal, carbon dioxide (CO2) or hydrogen (H2) storage, require a good comprehension of fluid flows in porous media.
This PhD subject proposes to predict the complex behavior of fluids in porous media based on advanced mathematic tools, especially Convolutional Neural Networks (CNN) or tools from Cheminformatics (QSPR – Quantitative Structure Property Relationship) and assess experimental predictions with micromodels employment. In its laboratories, IFPEN have a strong knowledge in micromodel experiments (transparent 2D porous micromodels) as so as in automatic learning methods. One of the objectives of this PhD consist in producing experimental databases for different porous media configurations. Then, the goal is to predict complex phenomena such as CO2 trapping in a given situation (trapping localization, trapping fraction, ganglia size, etc.), based on automatic learning. Then, obtained model inversion methods will be investigated; for example, determining what should be porous medium characteristics (porosity, structure, pore and grain distribution and size, etc.) for optimized CO2 sequestration. This PhD work will couple innovative aspects in both automatic learning and experiment with micromodel application and will also offer numerous perspectives in NET and beyond.

Keywords: porous media, microfluidics, flow, AI

• Academic supervisor    Dr Hugues BODIGUEL, hugues.bodiguel@univ-grenoble-alpes.fr, ORCID : 0000-0003-2348-6966
• Doctoral School    ED510 IMEP2, Université Grenoble Alpes
• IFPEN supervisor    Dr Nicolas PANNACCI, ORCID: 0000-0003-3293-7263
• PhD location    IFPEN, Rueil-Malmaison, France
• Duration and start date    3 years, starting in the fourth quarter 2025 (Novembre 3)
• Employer    IFPEN
• Academic requirements    University Master degree in Physics, Physical-chemistry 
• Language requirements    English level B2 (CEFR) 
• Other requirements    Experimental capabilities, Python knowlegde

To apply, please send your cover letter and CV to the IFPEN supervisor indicated here below.