NH₃ vs. CH₄ autoignition: A comparison of chemical dynamics

Abstract

In order to obtain physical insights on ammonia combustion, which is characterised by exceptionally long ignition delays and increased NOx emissions, the autoignition dynamics of an ammonia/air mixture is analysed using the diagnostics tools derived from the Computational Singular Perturbation (CSP) methodology. The results are compared to the autoignition dynamics of a methane/air mixture of same initial conditions. Methane was chosen for comparison because, even though the two molecules have a formal similarity, the ignition delay of methane is more than 10 times shorter than the one of ammonia. By using the CSP diagnostics tools, we identified the dominant chemical pathways that relate to the explosive components that drive the system towards ignition for both cases. Furthermore, the reactions that hinder the ammonia ignition were identified. This led to the determination of an interesting difference in the electronic configuration of the molecules of the two fuels, which is the root of their drastically different oxidation dynamics. In particular, it was shown that the autoignition process starts with the formation of methyl (CH${}_3$) and amine (NH${}_2$) radicals, through dehydrogenation of methane and ammonia, respectively. In the methane case, the methyl-peroxy radical (CH${}_3$–O–O–) then forms, which initiates a chemical runaway that lasts for approximately 2/3 of the ignition delay and leads to the gradual oxidation of carbon to CO${}_2$. In the ammonia case, though, the structure of NH${}_2$ is such that it is not possible to form NH${}_2$–O–O–. As a result, the chemical runaway is suspended.

Keywords:

• ammonia
• methane
• autoignition
• dynamics
• chemical kinetics
• computational singular perturbation

Acknowledgments

We would like to acknowledge support for this work from Khalifa University of Science and Technology under project RC2-2019-007.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The authors would like to gratefully acknowledge the support from RC2-2019-007 grant from the RICH Center as well as CIRA-2019-033 grant from Khalifa University of Science and Technology. Part of the work was performed while DGP was employed in the RICH Center and DMM in the Mechanical Engineering Department, both at Khalifa University of Science and Technology.