Melting glass requires a lot of energy, which means that process temperatures of 1600 °C and more are necessary in a glass melting furnace, depending on glass quality. In conventional plants, these high temperatures are usually obtained by the combustion of natural gas with strongly pre-heated air. Air pre-heat temperatures may reach up to 1400 °C in regenerator furnaces which, in combination with near-stoichiometric combustion in the furnace, often lead to high emissions of nitrous oxides (NOx). The glass industry is therefore very much interested in technologies which reduce NOx formation in the furnace itself, as secondary measures to filter NOx from the flue gas are expensive.
Oxy-fuel is one possible approach to achieve this: air is substituted by almost pure oxygen as the oxidizer, thus providing high flame temperatures while eliminating the main source of nitrogen in the system. This is different from oxy-fuel combustion in power plants where the main focus is to facilitate CCS by replacing combustion air with a mixture of recirculated flue gas and oxygen.
There are many oxy-fuel glass melting furnaces already in operation, but most of these are based on conventional air-fired furnace designs. In the course of a recent research project, Gas- und Wärme-Institut Essen e. V. (GWI) investigated the use of oxy-fuel combustion in such furnaces in order to provide improved design criteria for purpose-built oxy-fuel glass melting furnaces by both experimental and numerical means. The comparison of the simulation and measurement results shows, however, that many combustion models commonly used in CFD codes seem to be unable to adequately describe oxy-fuel combustion. Only a numerically expensive EDC approach yields a reasonable agreement with the experiments, though at a significantly increased numerical effort. Further research is necessary to provide a means with which oxy-fuel combustion processes can be modelled accurately at reasonable numerical cost.