Robust stability and sensitivity of real controlled CSTRs

In this paper we describe a procedure to perform stability, sensitivity and bifurcation analysis of non-ideal controlled CSTRs when process and kineti c parameters are subject to uncertainty. Such procedure consists of the sen sitivity analysis of the first Hopf bifurcation locus, performed applying M onte Carlo methods. This approach allows to quantify the effects of uncerta inty on model predictions without too many calculations. The most probable stability regions are detected so to design a robust control system. Advant ages for real process operation are in terms of reliability of predictions, good design and safety. (C) 1999 Elsevier Science Ltd. All rights reserved .