This content is not included in your SAE MOBILUS subscription, or you are not logged in.

Rapidly Pulsed Reductants in Diesel NOx Reduction by Lean NOx Traps: Effects of Mixing Uniformity and Reductant Type

Journal Article
ISSN: 1946-3936, e-ISSN: 1946-3944
Published April 05, 2016 by SAE International in United States
Rapidly Pulsed Reductants in Diesel NOx Reduction by Lean NOx Traps: Effects of Mixing Uniformity and Reductant Type
Citation: Reihani, A., Corson, B., Hoard, J., Fisher, G. et al., "Rapidly Pulsed Reductants in Diesel NOx Reduction by Lean NOx Traps: Effects of Mixing Uniformity and Reductant Type," SAE Int. J. Engines 9(3):1630-1641, 2016,
Language: English


Lean NOx Traps (LNTs) are one type of lean NOx reduction technology typically used in smaller diesel passenger cars where urea-based Selective Catalytic Reduction (SCR) systems may be difficult to package . However, the performance of lean NOx traps (LNT) at temperatures above 400 C needs to be improved. The use of Rapidly Pulsed Reductants (RPR) is a process in which hydrocarbons are injected in rapid pulses ahead of a LNT in order to expand its operating window to higher temperatures and space velocities. This approach has also been called Di-Air (diesel NOx aftertreatment by adsorbed intermediate reductants) by Toyota. There is a vast parameter space which could be explored to maximize RPR performance and reduce the fuel penalty associated with injecting hydrocarbons. In this study, the mixing uniformity of the injected pulses, the type of reductant, and the concentration of pulsed reductant in the main flow were investigated. We found that all of these parameters are important for the RPR system performance.
To obtain a uniformity of flow with the injected species to approach that of a plug flow, we developed a design using specific mixers to maximize the performance of RPR. The initial hypothesis for the required mixing process was to uniformly mix the injected reductants with the main flow in the radial direction, while keeping the axial mixing as low as possible. This goal was achieved by incorporating different mass transport processes, i.e. advection in the radial direction, and diffusion in the axial direction. Numerical investigation of the mixing of high frequency pulsed gaseous hydrocarbons into the main exhaust flow was performed to design an effective mixer to satisfy the desired mixing conditions. This mixing process and a fast injection system (down to 1ms pulse duration) was shown to have uniform radial mixing and axially separated pulses of reductants that gave the optimal mixing condition and achieved the highest RPR NOx conversion performance.
Employing the designed mixer, a range of reductants (H2, CO, C2H4, C3H6, and C3H8) were tested under similar operating conditions over a Pt/Rh LNT. The effectiveness of different reductants for NOx conversion in different temperature regimes was found to be as follows: T < 270°C: H2 > CO > C3H6 > C2H4; 270°C < T < 500°C: C3H6 > H2 > CO > C2H4; T > 500°C: C3H6 > C2H4 >H2 ∼ CO. In terms of the selectivity of converted NOx, H2 resulted in significant ammonia formation at low temperatures, but overall, the N2 selectivity was as follows: CO ∼ C3H6 > C2H4 >> H2. Generally, it was concluded that hydrocarbon reductants provided higher NOx conversion in the mid-range and especially higher temperature ranges with relatively high nitrogen selectivity. However, it was observed that the reactivity of hydrocarbons and the availability of oxygen had a significant influence on their performance, especially as the pulsing frequency was increased and reduction reaction time became more limited.
In this study we have shown that the use of rapidly pulsed reductants (RPR) can be studied in the laboratory with the equipment and methods presented here. In studies with a LNT catalyst, the variations in NOx performance with several reductants at reasonably high frequencies were shown. This suggests that this system should be able to provide useful information for optimizing the performance of LNT catalysts at high temperatures.