Propagation of Parametric Uncertainty in a Conceptually Designed Bioethanol Production Process

Abstract Understanding the propagation of parametric uncertainty in model-based computer simulations of conceptually designed bioprocesses and their metamodels is a critical step in improving the utilization and usefulness of such models. Generally, the number of design and operational parameters to be identified, calculated or assumed is very high. However, uncertainty analyses of these parameters usually focus on a selected few. The aim of this paper is to analyse the effect of increasing the number of uncertain parameters on the uncertainty of the metrics of interest acquired from the simulations and metamodels generated to relate these metrics to the uncertain parameters. The results indicate that overall uncertainty in simulated process metrics stabilises after a certain number of uncertain parameters is reached regardless of the characteristics of parametric uncertainty. However, metamodeling options such as sampling method, number of samples used, and type of metamodels used have direct effects on how the overall uncertainty can be represented by these metamodels. A demonstrative analysis is offered on a bioethanol production process model as a case study. In conclusion, the findings in this paper highlight the importance of the workflow followed in generating metamodels of bioprocess simulations under uncertainty.