THE EFFECT OF UNCERTAINTIES IN PHYSICAL PROPERTIES ON CHEMICAL PROCESS DESIGN (FUZZY SET THEORY, RELIABILITY, PROBABILITY)

The effect of physical property uncertainty on chemical process design, within the context of all uncertainties, is described. Before performing such an analysis, methods had to be developed for quantifying uncertainty and measuring its effect on the entire design. Uncertainties in data are quantified by probability distributions, and model uncertainties are represented by fuzzy sets. A reliability theory consistent with both the axioms of probability and fuzzy set theory was developed to study the effects of uncertainty on design.;The analytical methods developed in this study were applied to example processes to elucidate the effect of physical property uncertainty on process design. The examples showed that, while uncertainties in equilibria are usually significant, the exact effect depends on the problem and the nature of the uncertainty. The examples show that model errors are much larger than data uncertainty and that model errors can have a significant impact on the design. Hence, equilibrium data are important for design economy. Other physical property uncertainties were shown to be less important than uncertainties in vapour-liquid equilibrium.