A Physics-Based Reduced-Order Model to Predict Gas Holdup in Bubble Columns

Abstract

Bubble columns are extensively used in the process industry for gas–liquid contact operations. The Euler–Euler CFD model is commonly used to simulate bubble columns. However, the high computational cost of these simulations makes them impractical for rapid design-space exploration. Consequently, bubble column design relies heavily on empirical correlations, which have limited generalizability and predictive accuracy. To address this limitation, a physics-based reduced-order model (ROM) is developed via cross-sectional averaging of the Euler–Euler gas and liquid phase continuity and momentum conservation equations. The developed ROM consists of five algebraic equations without any tuning parameters. The unclosed terms appearing in the ROM are analyzed and closed using 3D Euler–Euler CFD simulations of bubble column experiments. The ROM can predict the overall gas holdup and the cross-sectionally averaged gas and liquid velocities within seconds. The model’s accuracy is evaluated against several experiments from the literature, covering a wide range of designs (e.g., column diameter, aspect ratio, and orifice size) and inlet gas velocity (0.01 to 0.2 m/s). The developed ROM outperforms existing empirical correlations in predicting the gas holdup for Weber number We < 2.