Molecular Simulations for Improved Process Modeling of an Acid Gas Removal Unit
Marianna Yiannourakou, Xavier Rozanska, Benoit Minisini, Frédérick de Meyer
Fluid Phase Equilibria
Volume 560, September 2022, 113478
Reliable process modeling and simulation is required to optimize the design and performance of acid gas removal units, which play a key role in the H2S and CO2 removal from natural gas, and in the CO2 capture from flue gasses. Many physico-chemical properties that are key input parameters to the process models are obtained from time-consuming and sometimes dangerous experimental campaigns (H2S) or are not (easily) accessible experimentally and are currently estimated. In this work, we investigated which of these properties could quickly and reliably be obtained using molecular simulations, to improve the accuracy of the process models and to reduce the need for long or difficult experiments. The absorption properties of H2S, CO2, and CH4 in aqueous MethylDiEthanolAmine (MDEA) were computed using a combination of forcefield-based molecular dynamics, Monte Carlo methods, and quantum mechanical approaches. The results are compared to experimental data and/or currently used estimates in the AspenPlus process model. It was found that the key properties which might reliably and relatively easily be obtained by molecular simulations are physical: density, phase behavior, Henry coefficients, diffusion coefficients, viscosity, and surface tension. Quantum mechanical calculations based on the widely used density functional theory are instrumental in exploring reaction mechanisms and in determining the structure of transition states, but obtaining accurate free energies of reactions remains a challenge, especially when energy differences of less than 1 kJ mol−1 are necessary for quantitative predictions. Such an accuracy can only be reached if the solvent effects, which are detrimental, are correctly included in the model.