| Abstrakt: |
The complexity of the pressure swing adsorption (PSA) mathematical model, the need for its multiple calculations to reach the cyclic steady state and a large number of functional dependencies lead to unstable numerical circuits, physically unrealistic oscillations in adsorption profiles, an increase in the calculation time, and the failure of the solver. The paper proposes an approach to optimizing the calculation process, which consists in finding a reasonable balance between the completeness of the PSA mathematical model and the accuracy of the results obtained. The effectiveness of the approach is demonstrated on the example of air oxygen enrichment and hydrogen recovery from synthesis gas. The gas separation processes were simulated for the two-adsorber PSA unit with a granulated 13X adsorbent. The effect of the changes in the model coefficients on its accuracy in the operating range of input variables is investigated. A distinctive feature of this study is the recommendations for choosing a set of the model equations to calculate the PSA processes which are particularly relevant when solving optimization problems with uncertainty. The productivity, cycle duration, the diameter of the adsorbent particles and the flow rate, at which it is advisable to use the isothermal and external diffusion reduced PSA model in the calculations, are established, which will save at least 24.3 and 47.1% of the CPU time with a small loss in accuracy. The proposed approach can be used to form a set of equations for the PSA, rPSA, ultra rPSA, VSA, VPSA models, separation of various gas mixtures on various adsorbents. [ABSTRACT FROM AUTHOR] |
