Runtime efficient simulation of monolith catalysts with a dual-layer washcoat

• Published in: Catalysis today Vol. 188; no. 1; pp. 70 - 79
• Main Authors: Scheuer, A., Drochner, A., Gieshoff, J., Vogel, H., Votsmeier, M.
• Format: Journal Article Conference Proceeding
• Language: English
• Published: Amsterdam Elsevier B.V 01.07.2012; Elsevier
• Subjects: ammonia; Ammonia oxidation; Catalysis; catalysts; catalytic activity; Chemistry; Dual layer; equations; Exact sciences and technology; General and physical chemistry; oxidation; Spline mapping; temperature; Theory of reactions, general kinetics. Catalysis. Nomenclature, chemical documentation, computer chemistry; Catalysis; Dual layer; Ammonia oxidation; Spline mapping; Air pollution control; Catalytic reaction; Residence time; Equation; Modeling; Steady state; Ammonia; Heterogeneous catalysis; Simulation; Exhaust gas; Models; Oxidation; Reactor; Flue gas purification; Catalyst
• ISSN: 0920-5861; 1873-4308
• DOI: 10.1016/j.cattod.2012.03.017

[Display omitted] ► An ultra fast 1-D spline model of a dual layer NH3 oxidation catalyst was built. ► Spline model is based on a 1-D+1-D model including internal and external diffusion. ► The 1-D+1-D model agrees well with a detailed 2-D model published earlier. ► Spline model is approximately 10000 times faster than the 1-D+1-D model. ► Average error of the spline model compared to the 1-D+1-D model is less than 2.18%. This paper demonstrates that the use of a solution-mapping approach can accelerate the simulation of a dual-layer ammonia oxidation catalyst by several orders of magnitude. In a first step, a 1-D+1-D model of the dual-layer catalyst is set up. The results of this model are compared to the numerical solution of a 2-D model, and it is shown that the 1-D+1-D model provides an accurate approximation of the full 2-D solution. In a second step, spline interpolation functions are constructed that map the solution of the mass balance equations for one axial volume element of the 1-D+1-D model as a function of the inlet concentrations, temperature and residence time. During the runtime of the simulation, concentration profiles can be computed by simple successive calls of the interpolation function without the need for a numerical solver. The 5-dimensional spline functions reproduce the output concentrations of a reactor model with 35 axial volume elements with an accuracy of better than 1.25%. The use of the solution mapping reduces the computation time for a steady-state solution of the monolith model from 97.5s down to 10ms.