Article révisé par les pairs
Résumé : Methane/hydrogen combustion represents a concrete solution for the energy scenario to come. Indeed, the addition of hydrogen into the natural gas pipeline is one of the solutions foreseen to reduce CO2 emissions. Nevertheless, the replacement of methane by hydrogen will enhance the reactivity of the system, increasing NOx emissions. To overcome this issue, non-conventional combustion technologies, such as flameless combustion represent an attractive solution. This study aims to improve our understanding of the behaviour of methane/hydrogen blends under flameless conditions by means of experiments and simulations. Several experimental campaigns were conducted to test fuel flexibility for different methane/hydrogen blends, varying the injector geometries, equivalence ratio and dilution degree. It was found that a progressive addition of hydrogen in methane enhanced the combustion features, reducing the ignition delay time and loosing progressively the flameless behaviour of the furnace. Reducing the air injector diameter or increasing the fuel lance length were found to be efficient techniques to reduce the maximum temperature of the system and NOx emissions in the exhausts, reaching values below 30 ppm for pure hydrogen. MILD conditions were achieved up to 75%H2 in molar fraction, with no visible flame structures. Additionally, RANS-based simulations were also conducted to shed further light on the effect of adding hydrogen into the fuel blend. A sensitivity study was conducted for three different fuel blends: pure methane, an equimolar blend and pure hydrogen. The effect of chemistry detail, mixing models, radiation modeling and turbulence models on in-flame temperatures and NOx emissions was also studied. In particular, it was found that the usage of detailed chemistry for NOx, coupled with an adjustment of the PaSR model, filled the gap between experiments and predictions. Finally, a brute-force sensitivity revealed that NNH is the most important route for NOx production.