X

The challenge of reducing carbon footprint of many chemical processes and bringing down their development costs can be achieved through Process Intensification (PI). Different PI technologies have been investigated over the years with rotating packed bed (RPB) technology receiving much of the attention for its potential of significant intensification in terms of hardware size, capital expenditure, and operating costs. In this study, we present a complete derivation of the dry pressure drop in RPB that differs from the published models in considering the radial distribution of the gas tangential velocity as well as the viscous shear stress between gas layers. Porous media approach was adopted to model the viscous and inertial packing resistance forces. The inertial resistance coefficient was derived using machine learning (ML) techniques based on part of the published data on RPB dry pressure drop (training set). The data learning step relies on the minimization of the absolute error between the pressure drop evaluated from a one-dimensional mathematical model and experimental data to determine the optimum inertial resistance coefficient. Then, an Artificial Neural Network (ANN) was implemented to relate the inertial resistance coefficient to gas flowrate and rotating speed. Finally, the other part of the published data was used to test and validate the proposed approach based on the total pressure drop. The results show that the error in predicting RPB dry pressure drop using the semi-empirical model can be reduced from 25% to 2% when a machine learning algorithm is used to estimate the