Uncertainty Quantification on Blending Degree of Ammonia-Hydrogen Combustion

Historically, combustion has been at the heart of transport and power-generation sectors, owing to its technological maturity, widespread infrastructure, and ability to deliver high efficiency. However, the urgent drive toward decarbonization is motivating a shift from conventional hy- drocarbon fuels to carbon-free alternatives. Ammonia emerged as a promising option due to its high hydrogen density, low production and storage cost, and well-established handling in- frastructure. Yet its low heating value, slow flame speed, and high nitrogen content can lead to elevated NOx emissions unless combustor designs are re-engineered to mitigate fuel-nitrogen oxidation [1]. Recent studies showed that blending ammonia with a small fraction of hydrogen can improve flame stability and facilitate ignition, while retaining much of ammonia’s carbon- free advantage [2, 3]. At the same time, however, the fuel-blend composition becomes a critical source of uncertainty, and small variations in the ammonia/hydrogen blending ratio may prop- agate to produce significant variability in emissions and performance. Moreover, high-fidelity simulations of realistic combustor geometries are often too costly for extensive parametric or probabilistic studies. To assess these problems, in the present work, we employ a chemical reactor network (CRN) model and uncertainty quantification techniques to assess the impact of blend-ratio uncertainty on the performance of an ammonia-hydrogen fueled combustor. By forgoing detailed fluid dynamics in favor of a simple network of perfectly mixed and plug-flow reactors, the CRN approach allows us to retain high-fidelity chemical kinetics while enabling rapid evaluation of a large number of calculations. Mapping how input uncertainty in the NH3/H2 blending ratio translates into uncertainty in NOx emissions and combustion efficiency provides actionable guidance for robust ammonia–hydrogen combustor design.