(476e) Development of a Methodology to Obtain the Kinetic Mechanism of Steam Methane Reforming Compact Reactor Using CFD Coupled With Hybrid Genetic/Surface Response Optimization Model

The process intensification characteristics of compact reactors (enhance heat and mass transfer properties) represent a breakthrough technology for heat and mass controlled reactions since it allows exploring the full potential of catalytic systems during highly endothermic or exothermic reactions. Concerning this issue, special attention is drawn to the steam methane reforming (SMR) reactor, a widely-used process to produce hydrogen or syngas (mixture of hydrogen and carbon monoxide).  A conventional steam methane reformer (SMR) consists of several hundred fixed-bed reactor tubes filled with supported catalyst particles, which converts a mixture of hydrocarbons and water into hydrogen and carbon monoxide via an endothermic reaction. The SMR reaction is supported by the heat provided by fuel combustion in the process furnace (exothermic reaction). Besides the large use of the conventional technology, efforts have been made to improve process efficiency, reduce production cost and footprint, mainly for small-scale applications (offshore methanol and GTL plants, fuel cells applications, car fuelling stations, etc). Steam methane reforming (SRM) using compact reactor has great potential to be a cost-effective means to achieve such goal.  The reaction mechanism for the SMR and catalytic combustion reaction is a key step to optimize the compact reformer reactor. Despite many kinetic mechanisms for the catalytic combustion and steam methane reforming reaction available, concerning different number of reactions, type of catalyst, inhibition effects by carbon dioxide and water in literature, information available is sometimes controversial and cannot be used in all cases.

This article aims to show the application of CFD coupled to a genetic algorithm optimization in order to obtain a reaction mechanism for steam methane reforming, water gas shift and combustion inside a catalytic reactor. The reactions were simulated considering a catalytic compact mini-channel reactor composed by stacked plates containing parallel arrays of mini-channels. The process channels (steam methane reaction) are filled with precious metal catalyst and the necessary heat for the reaction is provided by the intercalated plates containing parallel arrays of catalytic combustion channels. The optimization methodology makes it possible for the engineering professional to prevent the tedious task of manually executing analyses in a try-and-error approach, introducing a methodology where numerical simulations are executed without any user intervention. Multi-Objective Genetic Algorithm is used in the minimization of error between numerical results and experimental data. Four objectives, based on the summation of the squared relative error in each point, were used and 21 kinetic parameters of reaction rate equations of single step methane reforming, water gas shift and methane combustion were evaluated as optimizing variables. Due to heat transfer effects between the plates, catalyst and the gas mixture, the standard techniques to obtain the reaction rate could not be applied. Instead of that, a CFD simulation of a representative domain was used to calculate the conjugate heat transfer together with reactive flow. The optimization was applied to minimize the relative error between a simplified CFD simulation, which evaluated a two-dimensional reactive flow and the experimental results that could be obtained on the real scale equipment. The methane conversion and temperature for the combustion and reforming ducts were evaluated. The reaction mechanism obtained will feed a more complex CFD study, considering the complete reactor geometry, allowing the prediction  of hot spots and mal distribution in some regions. Also, the model makes it possible to analyze the thermodynamic equilibrium for coke formation and the methane conversion for several flow rates and gas composition, which allows provides a safety operational window for the process.