The availability of predictive models for chemical processes is the basic prerequisite for offline process optimization. In cases where a predictive model is missing for a process unit within a larger process flowsheet, measured operating data of the process can be used to set up such models combining physical knowledge and process data. In this contribution, the creation and integration of such gray‐box models within the framework of a flowsheet simulator is presented. Results of optimization using different gray‐box models are shown for a virtual cumene process.

Since experimentation is expensive with increasing computation power the significance of modeling, simulation, and optimization in process development has grown. Quite often in such models some parameters are uncertain, e.g., due to high variance in experimental data used for their estimation. Methods for investigating the impact of uncertain model parameters in the simulation and a new extension of an adaptive multi‐criteria optimization method to account for these uncertainties are described and demonstrated based on a cumene process.

Taking account of uncertain model parameters in simulation-based flowsheet optimization is crucial in order to quantify the reliability of the optimization results. Since chemical process design is a multicriteria optimization (MCO) task, methods to deal with uncertain Pareto boundaries are needed. The simplest of such methods consists of a sensitivity analysis of the Pareto boundary. In this work, it is shown how going beyond sensitivity analysis can yield favorable process designs not seen by sensitivity analysis alone. This is achieved by taking uncertainties into account by worst and best case Pareto boundaries or by considering the robustness of the Pareto boundary with respect to uncertain model parameters as additional objectives. In order to increase computational efficiency, for the first time, an adaptive scalarization approach is used to deal with uncertainties in MCO. The methods are illustrated by the calculation of a NQ curve of a distillation column.

No abstract

No abstract