Global kinetic modeling of rapidly pulsed reductants for lean NOx traps: Frequency domain analysis and impact of mass transfer

Abstract

In the Rapidly Pulsed Reductants (RPR) process, also known as Di-Air (Bisaiji et al., 2011 [1]), hydrocarbons are injected in rapid pulses ahead of a lean NOx trap (LNT) in order to improve its performance at higher temperatures and space velocities while maintaining a reasonable fuel penalty associated with injecting reductants.In this study, a 23-step global kinetic model is developed to predict the performance and product selectivity of the RPR process using propylene as the reductant over a wide range of pulsing frequencies, particularly for the temperature range of 450–600 °C. Some of the previous studies, (Bisaiji et al., 2012 [2], Perng et al., 2014 [3]) have suggested a reaction mechanism based on the formation of stable isocyanate (NCO) intermediates at high temperatures which promote the NOx conversion efficiency. However, in this study, we are checking whether a global kinetic model based on the conventional (low frequency) NOx storage and reduction mechanism is able to predict the NOx conversion improvements at higher pulsing frequencies at high temperatures, where the stability and role of previously proposed intermediates for NOx reduction is more questionable.A wide range of low frequency switching experiments were used to fit the kinetic parameters to the low frequency response of the RPR process. This model was then used to explore the high frequency performance of the RPR process. The kinetic model can predict the increase in NOx and propylene conversion as well as the change in selectivity of the reaction products, such as NH3 and CO, when the pulsing frequency is increased from conventional LNT to optimal RPR frequencies near 1 Hz corresponding to maximum NOx conversion (Reihani et al., 2018 [4]). However, in order to predict the behavior of RPR above the optimal frequency, it was required to add a model of axial pulse mixing to the kinetic model.The model results indicate that the main mechanism for improvement in NOx conversion at high frequencies and high temperatures is the more efficient utilization of storage sites. Also, the drop in NOx conversion at frequencies higher than the optimal frequency is caused by axial mixing of the reductant pulses. In addition, the model and experimental results provide insights into the importance of washcoat diffusion resistance. Particularly, the NOx reduction step with propylene becomes increasingly limited by washcoat diffusion at high temperatures and pulsing frequencies. The agreement between the model and the experimental results provides strong support for using this model to describe the RPR process.