Quantification of the chemical reactor reliability in the presence of uncertainties/errors in input parameters
Résumé
Errors and uncertainties in the input parameters can have a significant effect on the reliability, quality and safety of a chemical process. However, the majority of mathematical models use deterministic operating conditions. Therefore, to have a more realistic mathematical design, it is essential to take them into account in order to analyse how they propagate in a chemical reactor and influence the safety criteria.
In this work, random variations were imposed on the input parameters of a batch chemical reactor and the impact on the system reliability was studied. An exothermic reaction of the synthesis of peroxyformic acid was chosen and the maximum temperature of this reaction was taken as a safety criterion. The case of impurity presence in the reactor was also investigated. This impurity can lead to the reactant decomposition and consequently to thermal runaway, despite the adequate setting of the operating conditions.
The probability density functions (PDFs) for the reaction maximum temperature were used to estimate the system reliability. Theses PDFs were obtained by solving Latin hypercube sampling and genetic algorithm combined with the least square method. Stochastic treatment provides a more significant information concerning the reliability and highlights the nonlinearity of the process.