Predictive Modeling of Impact of ANR Non-Uniformity on Transient SCR System DeNOx Performance 2015-01-1055

Selective catalytic reduction (SCR) is a promising technology for meeting the stringent requirements pertaining to NOx emissions. One of the most important requirements to achieve high DeNOx performance is to have a high uniformity of ammonia to NOx ratio (ANR) at the SCR catalyst inlet. Steady state 3D computational fluid dynamics (CFD) models are frequently used for predicting ANR spatial distribution but are not feasible for running a transient cycle like Federal Test Procedure (FTP). On the other hand, 1D kinetic models run in real time and can predict transient SCR performance but do not typically capture the effect of non-axial non-uniformities. In this work, two 3D to 1D coupling methods have been developed to predict transient SCR system performance, taking the effect of ANR non-uniformity into account. First is a probability density function (PDF) based approach and the second is a geometrical sector based approach. Steady state 3D CFD spray simulations are run under multiple test points and the effect of varying ANR uniformity index (UI) is taken into account adaptively while running transient 1D kinetic simulations. Non-uniform storage on the catalyst due to non-uniform ANR distribution is also taken into account. Model predictions are validated against on-engine data for typical aftertreatment configurations. PDF based approach is found to under predict cycle NOx conversion by up to 3% whereas sector based approach is found to over predict NOx conversion by up to 2%. The concept of weighted UI is introduced for transient cycles and good correlation is observed between weighted UI and NOx conversion efficiency for transient cycles.