Using Kriging surrogate model to analyze hydrogen generation with dimethyl ether in partial oxidation catalytic fluidized bed reactor

The study aimed to present a fluidized bed reactor filled with 50% platinum catalytic beads and 50% quartz sand for syngas generation via the partial oxidation of dimethyl ether (DME). For obtaining useful information, experiments were carried out using Taguchi orthogonal array method, which cut test run by 75.3%, with the results being used for constructing the Kriging surrogate model. The model was then modified to reduce the mean square error. It is validated that the model can predict the results reasonably well. The outcome shows that the proposed reactor can self-sustain the syngas reforming reaction without incurring extra heat for reactants preheating. In the operation region, the results show that the start-up temperature is most negligible in hydrogen generation. The yield rate of hydrogen increases in sync with the increase of oxygen's molar ratio and DME, thanks to higher reaction temperature. The results also suggest that the higher the molar ratio of oxygen and DME, the more significant the reverse water shift reaction. Playing a key role in the fluidized bed in the proposed reactor for hydrogen generation, the normalized velocity must be at a proper level, in order to induce bubbling fluidization without causing a lower hydrogen yield, the latter due to the short residence time of reaction caused by excessive velocity. The sensitive analysis of independent variables also proved the inference from the measured results. By comparing with the literature, the performance of the proposed reactor in the present study shows a reasonably well performance. Given focused hydrogen generation, the study's findings can be used as a basis for evaluating DME as a hydrogen carrier and the proposed fluidized bed reaction.