Comparison of methane combustion mechanisms using laminar burning velocity measurements

Large amount of experimental data for laminar burning velocity (LBV) measurements of methane ( + H 2 /CO) - oxygen - diluent mixtures (5500 data points in 646 datasets) covering wide ranges of equivalence ratio, diluent ratio, cold side temperature and pressure were collected from 111 publications. The diluents included N 2 , H 2 O, CO 2 , Ar and He. The data files are available on the ReSpecTh site ( http://respecth.hu ). Performances of 12 methane combustion mechanisms on reproducing these LBV measurements were analyzed according to experiment types and conditions. Most mechanisms could predict well the LBVs for stoichiometric and fuel-lean mixtures and for diluent ratios higher than 60%. The performances of several mechanisms were relatively poor at other conditions. Focusing on the operating conditions of natural gas engines, we recommend the application of mechanisms FFCM-I-2016, SanDiego2014, and NUIG1.1-2021 for engine simulations. Mechanisms Aramco-II-2016, Konnov-20 09, Caltech-2015 and Glarborg-2018 have the lowest average errors for the reproduction of all available methane LBV data. Using local sensitivity analysis on the most accurate mechanisms, we identified 29 important elementary reactions, which, however, were not present in all the 12 mechanisms. We also collected large amount of directly measured and theoretically calculated rate coefficients for these reactions and compared them with the rate coefficients used in the 12 mechanisms. Reactions found important in any of the Aramco-II2016, Konnov-20 09 and Glarborg-2018 mechanisms, but missing from the Aramco-II-2016, Konnov-20 09, Glarborg-2018, Caltech-2015, FFCM-I-2016 and NUIG1.1-2021 mechanisms were added to these six mechanisms to investigate if the extended mechanism performs better than the original one. Some of the extended mechanisms became the best performing mechanisms. (c) 2021 The Author(s). Published by Elsevier Inc. on behalf of The Combustion Institute. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ )