Multiphase Computational Fluid Dynamics Modelling of Gas Liquid Separation in an Ebullated Bed Hydroprocessor

Abstract

This research built a multiphase computational fluid dynamics model for the gas-liquid separator in an ebullated bed hydroprocessor. The k-ω Shear Stress Transport turbulence model in conjunction with the Tomiyama slightly contaminated drag model were used to investigate gas-liquid separation efficiency. Strategic model selection ensures stable and accurate results, laying groundwork for future model comparisons incorporating inhomogeneous multiple size group population balance models. Case studies explore impacts of independent variables on separation efficiency, revealing that larger bubble size enhances separation by facilitating disengagement from liquid streamlines. Augmenting inlet gas phase fraction increases overall gas holdup, reduces mean liquid residence time and separation efficiency. A greater liquid momentum's effect on residence time is less pronounced but decreases separation efficiency. Viscosity inversely affects separation efficiency due to decreased relative velocities. Increasing liquid recycle ratio decreases separation efficiency by reducing mean liquid residence time. Mean liquid residence time emerges as a reliable predictor of separation efficiency, underscoring its significance in process optimization. Future inclusion of population balance models could provide insights into the impact of different bubble sizes and breakup/coalescence modeling, with validation challenges anticipated. Additionally, incorporating turbulence models accommodating phase inversion phenomena shows promise in improving simulation robustness, especially in complex scenarios like population balance models, warranting further investigation for accurate phase inversion dynamics capture.