Collect. Czech. Chem. Commun. 1983, 48, 1588-1596

Simulation of chemical processes with uncertain parameters

Mirko Dohnal and Marie Ulmanová

Department of Mechanical Equipment Chemical and Food Industry, Technical University, 611 00 Brno


An integral part of the simulation problem are numerical values of parameters or constants of industrial units. Either these quantities need not be known accurately or it is not possible to control them at the set up value. The result of simulation cannot be then accurate. It is necessary to determine the region in which the flow rates of components through branches could change when the parameters and constants change. For these purposes is applicable the Monte Carlo method. But it is very time-consuming. Here, algorithms are proposed which with the use of the algebraic norm approximately determine the looked-for set of flow rate uncertainty. These algorithms can be applied for solution of a number of practical problems related with optimum sizing of units, sensitivity studies etc. An example is given of a cooling cascade cycle.