Abstract
During the design of a chemical process engineers typically switch from simple (shortcut) calculations to more detailed rigorous models to perform mass and energy balances around unit operations and to design process equipment involved in that process. The choice of the most appropriate thermodynamic and thermo-physical models is critical to obtain a feasible and operable process design and many guidelines pertaining to this can be found in the literature. But even if appropriate models have been chosen, the user needs to keep in mind
that these models contain uncertainties which may propagate through the calculation steps to such an extent that the final design might not be feasible or lead to poor performance. Therefore it is necessary to evaluate the sensitivity of process design to the uncertainties in property estimates obtained from thermo-physical property models. Uncertainty and sensitivity analysis can be combined to determine which properties are of critical importance from process design point of view and to establish an acceptable level of accuracy for
different thermo-physical property methods employed. This helps the user to determine if additional property measurements in the laboratory are required or to find more accurate values in the literature. A tailor-made and more efficient experimentation schedule is the result. This work discusses a systematic methodology for performing analysis of sensitivity of process design to uncertainties in property estimates. The application of the methodology is illustrated using a case study of extractive distillation in which acetone is separated from
methanol using water as a solvent. Among others, the vapour pressure of acetone and water was found to be the most critical and even small uncertainties from -0.25 % to +0.75 % in vapour pressure data have shown a significant impact on the reflux ratio of the extractive distillation process.
In general, systematic sensitivity analysis should be part of process design efforts and expected to contribute to better-informed and reliable design solutions in chemical industries.
that these models contain uncertainties which may propagate through the calculation steps to such an extent that the final design might not be feasible or lead to poor performance. Therefore it is necessary to evaluate the sensitivity of process design to the uncertainties in property estimates obtained from thermo-physical property models. Uncertainty and sensitivity analysis can be combined to determine which properties are of critical importance from process design point of view and to establish an acceptable level of accuracy for
different thermo-physical property methods employed. This helps the user to determine if additional property measurements in the laboratory are required or to find more accurate values in the literature. A tailor-made and more efficient experimentation schedule is the result. This work discusses a systematic methodology for performing analysis of sensitivity of process design to uncertainties in property estimates. The application of the methodology is illustrated using a case study of extractive distillation in which acetone is separated from
methanol using water as a solvent. Among others, the vapour pressure of acetone and water was found to be the most critical and even small uncertainties from -0.25 % to +0.75 % in vapour pressure data have shown a significant impact on the reflux ratio of the extractive distillation process.
In general, systematic sensitivity analysis should be part of process design efforts and expected to contribute to better-informed and reliable design solutions in chemical industries.
Original language | English |
---|---|
Publication date | 2012 |
Number of pages | 18 |
Publication status | Published - 2012 |
Event | CHISA 2012: 20th International Congress of Chemical and Process Engineering - Prague, Czech Republic Duration: 25 Aug 2012 → 29 Aug 2012 |
Conference
Conference | CHISA 2012 |
---|---|
Country/Territory | Czech Republic |
City | Prague |
Period | 25/08/2012 → 29/08/2012 |