Abstract Summary
The stability and serviceability of stress sensitive subsurface reservoirs for large scale long term energy storage can be examined using geomechanical reservoir models. In this work, we adopt a machine learning-based, finite element modelling framework to model the evolution of the macroscopic geomechanical behaviour of a subsurface reservoir. In efforts to ensure model robustness and prediction physical validity, we resort to the application of the joint Thermodynamics based Artificial Neural Networks x Finite Element Modelling (TANNxFEM) framework to geomechanical reservoir modelling. The proposed framework presents a computationally efficient alternative to current reservoir modelling schemes.
