F2010C140
Simplified Reaction Modeling of Automotive Catalytic Converters
Mathematical models to simulate the operation of various catalytic converter platforms are successfully used as tools to support catalyst design and optimization. Modeling the chemical phenomena is the most challenging task of such models, due to the complicated reaction mechanisms that occur on the catalytic surface under the transient conditions of real exhaust gas. Typically, a tradeoff exists between the chemical background of the model and the ease of model parameterization. Models employing reaction schemes with global reactions, instead of mechanistic reactions, have a good chemical background, and require reasonable parameterization effort to be applied for design and optimization purposes of exhaust aftertreatment systems. Further simplifications of the reaction kinetics usually disregard the chemical background and are therefore applicable mainly for catalyst control purposes. The present work proposes an approach to further reduce the parameterization effort associated with global reaction models, without totally disregarding the chemical background. The underlying assumption is that dynamic storage phenomena govern the pollutant conversion under real-world operating conditions. Therefore, global reaction schemes can be simplified by eliminating most of the steady-state reactions, and by focusing only on dynamic storage and release reactions. This assumption results in a compact reaction scheme with a relatively small number of tunable parameters. The proposed reaction scheme is implemented in a 1D solution of the energy and species balance equations in catalytic monoliths, based on well-established approaches from the literature. The application range and limitations of this approach are examined for three different catalyst platforms: Three-Way Catalysts, Diesel Oxidation Catalysts and NOx-Storage Catalysts. For Three-Way Catalysts (TWC), the proposed reaction scheme involves 5 oxygen storage and release reactions, and a reaction for hydrocarbon steam reforming. The reaction scheme accounts both for gasoline and for natural gas engine operation. It is validated against a large database of real-world driving cycles, carried out with different vehicles. The model can capture the main trends in pollutant conversion both in hot-mode and in cold-start operation. Two indicative model applications are presented: prediction of catalyst performance in a fleet of vehicles, an application where the low parameterization effort of the model is of critical importance; and a simple investigation of the effect of gas-phase mass transfer and washcoat diffusion in honeycomb and foam substrates. For Diesel Oxidation Catalysts (DOC), the physical importance of oxygen storage reactions is limited, as no oxygen storage components are typically used in the washcoat. However, the proposed simplified approach is still usable, under the following modifications: a very small value is assigned to the oxygen storage capacity, used as input to the model; and an additional reaction for NO oxidation to NO2 is included. Again, the model is validated against real-world driving cycles, carried out with two light-duty commercial vehicles equipped with DOC, and captures pollutant conversion in hot-mode operation well. For NOx storage catalysts (NSC), an additional 3-reaction scheme for storage of NOx on barium oxide during the lean phases, and regeneration of barium nitrate by CO during stoichiometric or slightly rich phases is added to the 6-reaction TWC scheme. The resulting 9-reaction scheme predicts qualitatively the NOx storage and release performance under constant-speed driving conditions.
This abstract is supplemented by a PDF, which can be viewed here.