CFD modeling and numerical simulation of an industrial adsorption process

Abstract

This paper investigates the model performance of an industrial box used to remove ammonia from the air using a doped activated carbon. To this end, process models are developed and experimental measurements are carried out. The models are based on momentum and mass balance equations along with thermodynamic, hydrodynamic and adsorption kinetic relations. The measurements consist of pressure drops in the air purification box and breakthrough curves of ammonia on activated carbon. The models are implemented and solved using COMSOL Multiphysics®. Adsorption is investigated by means of a Fickian multi-scale model, which demonstrated to efficiently predict the experimental data. The optimal values of the most estimable unknown parameters involved in each model, as well as their confidence intervals are determined from experimental measurements. The performance of the industrial air purification model is rigorously assessed using statistical criteria, including mean-square error (RMSE = 0.0013) and Pearson correlation coefficient (r = 0.999).