(617ce) Experimental and Kinetic Modeling of Lean-Rich Switching Study over a Modified Twc System
Lean-Rich switching study over a modified TWC system
Mengmeng Li, Sam Malamis, Michael P. Harold*,
Department of Chemical
and Biomolecular Engineering, University of Houston, Houston, TX 77204
The most recent CAFE standard for automobile
requires an average fuel economy of 54.5 miles per gallon by 2025 . At the
same time, more stringent emission rules enacted by the EPA for 2017-18 will
require an 80% reduction in non-methane organic gases (NMOG) plus NOx from
current Tier 2 Bin 5 levels . Gasoline vehicles that
operate under both lean-burn and stoichiometric combustion hold promise since lean-burn
combustion is more fuel efficient than conventional stoichiometric combustion. However,
elimination of NOx (NO+NO2) by the traditional three-way catalyst
(TWC) is not possible under lean conditions. Advanced NOx reduction strategies must
be applied to exploit the more efficient lean-burn engine combustion such as
the three-way catalyst with a NOx storage function (TWNSC). The TWNSC stores NOx
storage during lean operation and reduces the trapped NOx during stoichiometric
operation. To date, there have been only a few studies that report TWNSC.
In this study a predictive TWNSC model is developed to predict TWNSC
performance and in so doing provide operational insight and optimization.
A model of a catalytic monolith channel was
developed that incorporates a TWNSC kinetic model into a low-dimensional model
formulation [3Ð5]. The convection-diffusion-reaction
equations are averaged in the transverse direction which replaces the
transverse gradient term with an algebraic term that contains an internal mass
transfer coefficient. Overall heat and mass transfer coefficients are used to
represent the heat and mass transfer in the transverse direction between the
bulk of fluid and solid phase. The TWNSC kinetics consists
of a combination of elementary steps and global reactions. Following our
previous experimental study of NOx reduction by propylene over a multifunctional
NOx trap catalyst , the hybrid model includes the
kinetics of the oxidation of CO, H2, NO and hydrocarbons, of
reduction of NO/NO2 by CO, H2 and representative
hydrocarbons, and of storage of NOx and O2. Fig. 1 shows the
reaction network including 10 main paths that described the chemistry in
1. Simplified reaction network of NOx reduction using propylene as reductant
A step-wise approach was used
to develop kinetics for the oxidations of CO, H2 and propylene.
First, global kinetics are fitted tuned to transient light-off experiments
following Raj et al. . For example, the Langmuir-Hinshelwood
expression is given by
Fig. 2 shows the fitted
light-off curve for CO oxidation and the comparison between experimental and
model data. The model shows a reasonable fit and captures the dynamical feature
of CO TPO (temperature programmed oxidation) experiment.
Figure 2. (a) Fitted light-off curve of
CO oxidation (b) comparison between experimental and model-predicted
conversions. [Conditions: 0.5% CO, 0.5% O; temperature ramp:
rate expressions for the co-oxidation of CO, H2, and C3H6
include coupling effects. These are combined with a dual-site NOx storage
model and incorporated into species balances.
will show how the model is effective in capturing the main trends in the
lean-rich switching data. This study advances the understanding of TWNSC
catalyst and provides guidance for optimizing catalyst formulation and
Johnson, SAE Int. J. Engines. 6 (2014) 699Ð715.
Joshi, M.P. Harold, V. Balakotaiah, AIChE J. 55 (2009) 1771Ð1783.
Bhatia, R.D. Clayton, M.P. Harold, V. Balakotaiah, Catal. Today. 147 (2009)
Raj, M.P. Harold, V. Balakotaiah, Chem. Eng. J. 281 (2015) 322Ð333.
Li, V.G. Easterling, M.P. Harold, Catal. Today. 267 (2016) 177Ð191.