TY - JOUR
T1 - Development and application of an efficient chemical reactor network model for oxy-fuel combustion
AU - Trespi, Silvio
AU - Nicolai, Hendrik
AU - Debiagi, Paulo Amaral
AU - Janicka, Johannes
AU - Dreizler, Andreas
AU - Hasse, Christian
AU - Faravelli, Tiziano
N1 - Funding Information:
The authors kindly acknowledge financial support through Deutsche Forschungsgemeinschaft (DFG) Projekt 215035359, through SFB/TRR 129. The authors are grateful for the support provided by Prof. Alberto Cuoci, Prof. Alessio Frassoldati, Prof. Alessandro Stagni, Dr. Maria Angela Agizza, and Henrik Schneider.
Publisher Copyright:
©2020 American Chemical Society.
PY - 2021/5/6
Y1 - 2021/5/6
N2 - The excessive computational cost of computational fluid dynamics (CFD) simulations of complex reactors is still a barrier to investigations using detailed chemical kinetics. Chemical rector network modeling is a promising tool for affordable numerical investigations of novel combustor technologies, such as oxy-fuel combustion, using directly coupled detailed chemical kinetics. In this work, a novel chemical reactor network solver, NetSMOKE, which is part of the OpenSMOKE++ suite, is presented and discussed. The numerical model employed, together with the relative solution method, is explained, also exploiting a combined sequential-modular and equation-oriented approach for solving the global system of equations. The novel solver was employed to build a chemical reactor network that represents the Technische Universität Darmstadt oxy-fuel burner using a reduced number of ideal reactors with directly coupled detailed kinetic models for the first time. On the basis of previously available CFD simulations and measurements, the complex flow is accurately characterized and discretized into macrozones, facilitating the development of a simplified reactor network. Carbon monoxide emissions were analyzed in detail, supported by sensitivity analysis with respect to the reactor temperatures. The sensitivity analysis revealed that the post-flame zone is crucial for the overall CO emission. Therefore, on the basis of the sensitivity analysis, an iterative approach for refining the reactor network model is developed. The increase in the number of ideal reactors in targeted areas of the system allows for significant improvements with respect to CO predictions. The chemical reactor network provides good agreement with the experimental data, requiring only a limited increase in the overall computational cost. The presented tool offers a computationally efficient strategy to investigate and predict the behavior of complex reactors, including the emission of pollutants in combustion devices, allowing for the employment of state-of-the-art, detailed chemical kinetic mechanisms available from the literature.
AB - The excessive computational cost of computational fluid dynamics (CFD) simulations of complex reactors is still a barrier to investigations using detailed chemical kinetics. Chemical rector network modeling is a promising tool for affordable numerical investigations of novel combustor technologies, such as oxy-fuel combustion, using directly coupled detailed chemical kinetics. In this work, a novel chemical reactor network solver, NetSMOKE, which is part of the OpenSMOKE++ suite, is presented and discussed. The numerical model employed, together with the relative solution method, is explained, also exploiting a combined sequential-modular and equation-oriented approach for solving the global system of equations. The novel solver was employed to build a chemical reactor network that represents the Technische Universität Darmstadt oxy-fuel burner using a reduced number of ideal reactors with directly coupled detailed kinetic models for the first time. On the basis of previously available CFD simulations and measurements, the complex flow is accurately characterized and discretized into macrozones, facilitating the development of a simplified reactor network. Carbon monoxide emissions were analyzed in detail, supported by sensitivity analysis with respect to the reactor temperatures. The sensitivity analysis revealed that the post-flame zone is crucial for the overall CO emission. Therefore, on the basis of the sensitivity analysis, an iterative approach for refining the reactor network model is developed. The increase in the number of ideal reactors in targeted areas of the system allows for significant improvements with respect to CO predictions. The chemical reactor network provides good agreement with the experimental data, requiring only a limited increase in the overall computational cost. The presented tool offers a computationally efficient strategy to investigate and predict the behavior of complex reactors, including the emission of pollutants in combustion devices, allowing for the employment of state-of-the-art, detailed chemical kinetic mechanisms available from the literature.
UR - http://www.scopus.com/inward/record.url?scp=85099156230&partnerID=8YFLogxK
U2 - 10.1021/acs.energyfuels.0c03560
DO - 10.1021/acs.energyfuels.0c03560
M3 - Article
AN - SCOPUS:85099156230
SN - 0887-0624
VL - 35
SP - 7121
EP - 7132
JO - Energy and Fuels
JF - Energy and Fuels
IS - 9
ER -