Subproject B07
Model reduction for reaction-transport systems for emission control


  Name Contact
Viatcheslav Bykov, KIT
Picture: KIT
PD Dr. Viatcheslav Bykov
+49 721 608-48746
Ulrich Maas TRR 150
Prof. Dr. rer. nat. habil. Ulrich Maas
+49 721 608-43930
Moloud Aghayarzadeh M. Sc.
+49 721 608-48747

At a glance

In the video, Etele Berszány presents project B07: Reduced reaction models – emission control.


Selective catalytic reduction (SCR) using urea-water solution is a process to reduce the emission of nitrogen oxides in combustion engines, whose relevance in the automotive sector has strongly increased during the last years. The underlying physical-chemical processes, especially concerning the formation of solid residuals, are not yet understood in detail. These processes include a multitude of potentially coupled physical processes like diffusion, and convection in both liquid and gas phase, evaporation and chemical reactions. For this reason, detailed simulations of the complete SCR process are very complex and the computational requirements are high. To reduce the complexity and computational costs of such simulations, there is a need for efficient and accurate reduced models which can describe the detailed chemical processes as well as the multiphase flow.


The objective of this subproject is the development of reduced kinetic models for urea and its decomposition products. These are required to reliably and efficiently describe the complex chemical kinetics in large eddy simulations (LES) of the processes in the exhaust gas system. For this purpose, detailed reaction mechanisms for the model system are analysed and reduced by using mathematically rigorous reduction concepts. Reactions occurring in all phases (gas, liquid, solid) and phase-changes will be considered. Furthermore, in the reduced model the coupling with flow and transport-processes will be accounted for. The implementation strategy of the reduced mechanism into the LES of the complete model will also be developed.

Previous Findings

It was found that the evaporation of a droplet of urea-water solution can be separated into two main phases of water evaporation and urea decomposition. The time-scales of the following gas phase chemistry between NOx and the gaseous decomposition products are much larger than the time-scales of the evaporation of a droplet. As a result, the gas-phase chemistry can be treated completely uncoupled from the evaporation. Based on this finding, several reduced models were developed to either describe the slow decoupled gas-phase chemistry or the detailed processes close to the droplet including transport. In addition, a simple model to describe the evaporation rate of droplets using polynomials was developed.

The evaporation rate of urea (left) and comparison of mass flows from detailed simulations and the simple droplet model (right).
The evaporation rate of urea (left) and comparison of mass flows from detailed simulations and the simple droplet model (right).


An evaporating and decomposing drop or film of urea-water solution is used as a model system. The internally developed program package INSFLA is used which allows detailed 1D simulations of evaporation and chemical reactions in the gas and liquid phase. Solid residuals will be considered as part of the liquid phase with adapted properties.

The computed time and space-dependent temperature and concentration profiles are then analyzed based on different reduction concepts. If for the investigated system the transport processes only cause a perturbation of the kinetics, the reduction concepts based on time scale separation local, e.g. ILDM (intrinsic low-dimensional manifolds) or global e.g. GQL (global quasi-linearization), will be used. If they directly couple with the chemical kinetics, which is likely for heterogeneous reactions in a film, the REDIM (reaction-diffusion manifolds) concept will be used. A comparison of the respective time scales allows an estimation to which extent a separation of the time scales exists. To incorporate heterogeneous processes into the REDIM concept, the boundary conditions for the REDIM equations have to be modified and different phases of the same species considered as different species.

Water and ammonia-production in system state space (left), relevant part for chemical reactions reduced and parameterized/tabulated (right)
Water and ammonia-production in system state space (left), relevant part for chemical reactions reduced and parameterized/tabulated (right)

Current Work

The main objective is the implementation of liquid and solid chemistry and the corresponding phase changes as well as the extension for surface films into the detailed simulations and the reduced models. This requires further development of the reduced models and their boundary conditions which currently only handle gas phase reactions and, in the case of REDIM, transport phenomena. Furthermore, the influence of convection is expected to be much higher for a film compared to a small droplet, which will also be considered.


The chemical mechanisms which will be reduced in this subproject are developed in sub-projects B04 and B05, while C04 provides experimental reference data and boundary conditions. Further improvement of the reduced models is performed in cooperation with the corresponding subproject (B06) for model reduction of in-engine combustion processes. The resulting reduced models for chemistry and other physical processes are then used in simulation of the exhaust gas system (C05, B01).

Selected Publications

  • Stein, M., Bykov, V., Kuntz, C., Börnhorst, M., Deutschmann, O., Maas, U.: Modeling the decomposition of urea-water-solution in films and droplets under SCR conditions with chemistry in the liquid phase. International Journal of Heat and Fluid Flow, 10.1016/j.ijheatfluidflow.2022.108936, 2022.
  • Yu, C., Maas, U.: Sensitivity of reaction–diffusion manifolds (REDIM) method with respect to the gradient estimate. Combustion Theory and Modelling, 10.1080/13647830.2022.2030494, 2022.
  • Stein, M., Bykov, V., Maas, U.: Reduced simulation of the evaporation and decomposition of droplets and films of urea-water solution in exhaust gas environment. Proceedings of the Combustion Institute, 10.1016/j.proci.2020.06.032, 2021.
  • Golda, P., Blattmann, A., Neagos, A., Bykov, V., Maas, U.: Implementation problems of manifolds-based model reduction and their generic solution. Combustion Theory and Modelling, 10.1080/13647830.2019.1682198, 2020.
  • Nishad, K., Stein, M., Ries, F., Bykov, V., Maas, U., Deutschmann, O., Janicka, J., Sadiki, A.: Thermal decomposition of a single adBlue® droplet including wall–film formation in turbulent cross-flow in an SCR system. Energies, 10.3390/en12132600, 2019.
  • Koksharov, A., Yu, C., Bykov, V., Maas, U., Pfeifle, M., Olzmann, M.: Quasi-spectral method for the solution of the master equation for unimolecular reaction systems: Solution of the master equation for unimolecular reaction systems. Int. J. Chem. Kinet., 10.1002/kin.21165, 2018.
  • Stein, M., Bykov, V., Bertótiné Abai, A., Janzer, C., Maas, U., Deutschmann, O., Olzmann, M.: A reduced model for the evaporation and decomposition of urea–water solution droplets. International Journal of Heat and Fluid Flow, 10.1016/j.ijheatfluidflow.2018.02.005, 2018.
  • Yu, C., Bykov, V., Maas, U.: Global quasi-linearization (GQL) versus QSSA for a hydrogen–air auto-ignition problem. Phys. Chem. Chem. Phys., 10.1039/C7CP07213A, 2018.
  • Gubernov, V.V., Bykov, V., Maas, U.: Hydrogen/air burner-stabilized flames at elevated pressures. Combustion and Flame, 10.1016/j.combustflame.2017.07.001, 2017.
  • Stein, M., Bykov, V., Maas, U.: The effect of evaporation models on urea decomposition from urea-water-solution droplets in SCR conditions. Emiss. Control Sci. Technol., 10.1007/s40825-017-0075-1, 2017.
  • Gubernov, V.V., Kolobov, A.V., Bykov, V., Maas, U.: Investigation of rich hydrogen–air deflagrations in models with detailed and reduced kinetic mechanisms. Combustion and Flame, 10.1016/j.combustflame.2016.03.017, 2016.
  • Korsakova, A.I., Gubernov, V.V., Bykov, V., Maas, U.: The effect of Soret diffusion on stability of rich premixed hydrogen–air flames. International Journal of Hydrogen Energy, 10.1016/j.ijhydene.2016.07.141, 2016.