This study focuses on quantifying the response of pore-scale transport properties in a multiphase system involving water, ice, gas, and granules subjected to dynamic thermal conditions. Permafrost is a multiphase medium in which solid (grain and ice) and liquid (water and greenhouse gas) coexist, and significant thermal gradients originated by warming result in dynamic subsurface processes. These processes taken in total exert complicated multiphysics behavior that is beyond the predictability of traditional models. Therefore, this research work aims to elaborate on the fluid transport models to replicate the non-isothermal phenomena by including temperature-dependent parameters and interfacial tension-driven flow. A bench-scale experimental setup was developed to simulate the fluid flow through porous media under applied thermal gradients. The setup is monitored during freeze-thaw cycles by a camera with an arrangement of light sources. The images are analyzed using machine learning enhanced computer vision to differentiate phases, detect a freezing and thawing front, and quantify ice/water ratios. Preliminary results from the experiments reveal that the tools used in this study can successfully characterize transport parameters in permafrost. Future experiments combined with mathematical transport models will contribute to the tailoring of the setup that can be used for the prediction of greenhouse gas emissions from thawing permafrost.
9th International Congress on Environmental Geotechnics (ICEG2023)
Coupled Processes in Environmental Geotechnics