Challenges, models, and algorithms for flow and transport simulations in deep reservoirs

Abstract The development of deep reservoirs is an emerging topic in the energy industry. This paper analyzes the challenges in simulations of flow and transport in deep reservoirs and introduces models and algorithms aimed at resolving these challenges. A fast, accurate, and robust phase equilibrium model is developed with the aid of deep learning algorithms to accelerate the thermodynamic analysis of deep reservoir fluids. A pixel‐free search algorithm is developed to generate a pore‐network model that describes pore connectivity and porous media fluidity. A fully conservative Implicit Pressure Explicit Saturation algorithm is developed to simulate the Darcy‐scale two‐phase flow while achieving a reliable result for production evaluation. Numerical examples are presented to validate the performance of the developed models and algorithms. This paper also presents suggestions for future studies on deep reservoirs to achieve both scientific and engineering progress.