https://hal-ineris.archives-ouvertes.fr/ineris-01855127Liu, QiQiLiuUTC - Université de Technologie de CompiègneProust, ChristopheChristopheProustINERIS - Institut National de l'Environnement Industriel et des RisquesUTC - Université de Technologie de CompiègneLen, ChristopheChristopheLenUTC - Université de Technologie de CompiègneLuart, DenisDenisLuartUTC - Université de Technologie de CompiègneGomez, FrançoisFrançoisGomezUTC - Université de Technologie de CompiègneApplied numerical chemistry in process engineering R&DHAL CCSD2015[SDE.IE] Environmental Sciences/Environmental EngineeringCivs, Gestionnaire2018-08-07 14:24:402022-06-26 09:56:302018-08-07 14:24:40enConference papers1Numerical modeling can be a great help in research and development activities to guide technical choice and/or to better interpret the experimental data. Significant developments are underway about the numerical simulation of complex chemical reaction as those investigated in laboratories. The resources of quantum chemistry are often used. About the industrial process in which chemical reaction occurs, numerical resources are available to describe flows, heat transfer and mechanical aspects. In between, there is a need to be able to foresee the evolution of the chemical reaction (yield, heat releases,...) as function of the process conditions (temperature, pressure,...). Since the reality of the reaction may be rather complex including not only the reaction with multiple components but phase changes, the modeling may prove difficult. Softwares were developed for this but for limited ranges of applications. In combustion processes for instance, NASA developed in the seventies the code CEA which was reproduced elsewhere. The basic principle is to minimize the Gibbs free energy of a given chemical reaction for which a series of potential final products is prescribed. The equation is completed by the usual conservation laws (mass in any case and depending on the reaction the energy and or pressure/temperature). In practice the method is particularly appealing and known to provide a number of very valuable and practical results. Unfortunately, the Gibbs free equation is "stiff" and to solve the mathematical problem drastic assumption are made to "linearise" the Gibbs free energy equation. This results in very strong limitations among which the quais impossibility to account for multiphase chemical equilibriums. In CIRCE, a software developed by the Technological University of Compiegne we tried to solve this problem replacing the traditional method by a Monte-Carlo technique. In this paper, this method is presented and typical results are given including distillations of multi-components mixtures and pyrogasifications.