Global Kinetic Modelling of a Supplier Barium- and Potassium- Containing Lean NOx Trap
Artikel i vetenskaplig tidskrift, 2006
Kinetic modeling, in combination with flow reactor experiments, was used in this study to simulate a supplier lean NOx trap (LNT). The LNT catalyst used is a commercial catalyst that contains barium and potassium as storage components. The results presented in this paper are a continuation of a previous study, where a global kinetic model for NOx storage was developed for a model Pt/Rh/BaO/Al2O3 catalyst. In this work, a simplified model is used, where NO oxidation, nitrite, and nitrate formation are lumped together into one reaction, and the model predicts the total NOx. In this model, only one reaction step must be tuned for each storage component: that is, in our case, one reaction for the formation of barium nitrate and one reaction for potassium nitrate. A broad range of experimental conditions was used when developing this model; five temperatures (200, 300, 400, 500, and 600 degrees C) and three different inlet NO concentrations (100, 200, and 300 ppm) were used. Water and CO2 were present in all experiments, because these can affect the storage behavior. The reductant used in the regeneration period was CO. Long lean and rich cycles were used to capture the kinetics of the reactions accurately. The model was able to describe all 15 experiments well and could adequately capture the amount of stored NOx during the lean period and the NOx conversion during the rich period, including the NOx breakthrough peak that occurred at the beginning of the rich period. The model was validated with short lean-rich cycling experiments, where the lean period was 30 s and the rich period was 2 s. The model could predict the outlet NOx concentration well, and the error for the average conversion was only 1%-2% in the validation simulations.