| In this paper, the effect of the uncertainties of the thermodynamic data and models on the process design is addressed. Two industrially based cases--"Production of Medium Pressure Nitrogen by Cryogenic Air Separation" and "Production of Dimethyl Ether via the Dehydration of Methanol" are simulated in different thermodynamic models using different commercial software packages (Chemcad and Aspen Plus). Meantime, strategies for quantification of such thermodynamics-parameter-induced uncertainties via Monte Carlo simulation with the Shifted Hammersley Sampling technique, Equal Probability Sampling Technique and regression analysis are described.;The effects of the uncertainties in thermodynamic models, binary parameters and simulators on process simulations are shown in detail. In both cases, the study indicates that the designs developed through use of these models are significantly sensitive to these parameter uncertainties. The most important binary parameters for both cases are addressed from the quantitative analysis. Moreover, the reasons of the significantly different results from the different simulators are indicated. |
