Avaliação do processo de absorção para uma captura industrial de CO2 usando K2CO3 ativado por ácido bórico.

Abstract

The Hot Potassium Carbonate (HPC) process aims to remove the CO2 present in the synthesis gas. This removal is carried out in the absorption process through the reaction of CO2 with the K2CO3 solution, due to the reaction occurring slowly, H3BO3 can be used as a catalyst. Two approaches can be used to simulate this process: the equilibrium model and the rate-based model. In general, the equilibrium model does not correctly predict the behavior of the absorption process, and the use of rate-based is more recommended. This approach uses different correlations to calculate important hydraulic and mass transfer parameters. And to evaluate the performance of these correlations, an automatic procedure was developed that tests a large number of equations, using the MATLAB and Aspen Plus software together. The best set of correlations was found after a comparison with industrial data. The correlations proposed by Rocha et al. (1996) to calculate the mass transfer coefficient and interfacial area and Stichlmair et al. (1989) to calculate the net holdup presented errors smaller than 10% for all operational conditions evaluated. Given the above, it can be said that the rate-based model is much more complex and requires a greater number of adjustment parameters and differential equations. In this context, an alternative to use the equilibrium model and increase its representativeness is to calculate the Murphree efficiency of the components present in the process. A methodology based on Artificial Neural Networks (ANNs) to calculate these efficiencies was proposed using the two commercial software mentioned above and the effect of including Murphree's efficiency calculations in the equilibrium model was analyzed. The simulation results were compared with the plant data and predicted that the simplest models based on equilibrium for the absorber can lead to a deviation of up to 20% in the prediction of the CO2 capture rate, while the corrected model with the efficiency of Murphree, calculated from the proposed networks reduce this error to less than 5% in all operational conditions evaluated