Eddy dissipation concept (EDC) model and flamelet generated manifolds (FGM) model are developed separately to study the temperature profiles and extinction limits of non-premixed hydrothermal flames. Predictions by the two models are evaluated comparatively by experimental data in literatures. FGM model shows relatively better prediction of temperature than EDC model in the near nozzle field. Extinction temperatures can be predicted by EDC model with deviations of 10–33 K. The extinction flow rates predicted by the FGM model are higher than those by the EDC model. Flow fields and reaction source terms are analysed to identify the inherent mechanism leading different results by the two models. It is illustrated that the positive effect of turbulence on reaction rate near the nozzle by the FGM model is the essential reason causing different flame characteristics from the EDC model by which the turbulence only has negative effect on reaction rate.

Counterflow diffusion flames of methanol hydrothermal combustion are investigated to improve the understanding of hydrothermal flames. It is indicated that the thermodynamic properties by the Peng-Robinson equation of state and the modified transport properties can reduce the flame temperature by about 500 K. The Takahashi correlation for mass diffusivity is found to be appropriate for hydrothermal combustion through comparison with the experimental data of Wellig et al. (J. Supercrit. Fluids, 2009, 49, 1). Compared to the Kolmogorov length scale in the experimental combustor, the thickness of the calculated counterflow flame is ten times larger, which means that the flame would be affected by the turbulence. The flame stable range is also reproduced well by the developed hydrothermal counterflow flame model. In the end, a Flamelet Generated Manifold (FGM) table is generated, promising to provide good closure of the non-equilibrium chemical source term in further turbulent flame simulations.

This work reports an experimental study on supercritical water oxidation of quinoline. Moderate preheat temperature (420 °C–510 °C) and initial concentration (1 wt%–10 wt%) are selected to address the possibility of utilizing the heat released during the reaction, in order to realize high conversion rate at relatively low preheat temperature. The effects of temperature, residence time, oxidation ratio, pressure and concentration are analyzed. The results show that considerable conversion can happen at relatively low preheat temperature, while increase in temperature will significantly promote the complete conversion. The yield of carbon dioxide increases with the residence time but there is an upper limit due to the stronger dependence on oxidizer concentration, for which an estimated reaction order is 1.90. When the quinoline concentration is larger than 8 wt%, clear exothermic peaks with temperature rise about 100 °C are detected. These exothermic peaks can be interpreted as a sign of ignition interrupted by the heat loss to the surrounding salt bath. An analogy is made between the start temperatures of these exothermic peaks and the ignition temperatures reported in methanol and isopropanol hydrothermal flame research. We conclude that quinoline solutions can be ignited without co-fuels, at comparable ignition temperature as methanol and isopropanol around 450 °C.