| OCR Text |
Show on flame characteristics of small scale burners in test furnaces and (3) theoretical heat transfer considerations. Very conservative approaches are generally taken in converting existing furnaces and designing new furnaces for oxy-fuel firing to avoid costly commercial mistakes. The optimum furnace design with oxy-fuel firing is likely to be different from the existing furnace designs for air firing due to a sharp reduction in the flue gas volume and to the heat transfer rate increase possible with oxy-fuel firing. Numerical modelling of the combustion space of glass furnaces offers potential to optimize the furnace design and burner placement for oxy-fuel firing. Although numerical simulation has been used for many years to understand the mechanism of glass melting and for commercial design of glass furnaces, simplified heat transfer models were often used for the combustion space as the focus of modelling was the analysis of glass flow patterns (Ref. 3). In particular, most of these simplified combustion space heat transfer models, do not account for the effects of flame position and characteristics on glass bath heat transfer. On the other side, advances in numerical methods and computer speed and capacities, have led to the development of complex models, which directly couple comprehensive 3-D finitedifference combustion space analysis, with 3-D melter analysis (Refs. 4,5). However, these latter models, while attractive in the straightforwardness of their approach, require still an enormous effort in time to set up and to validate, and in computational and financial resources to run. Therefore, the current analysis was based on an approach which is complex enough to allow study of the impact of local flame radiation on heat transfer to the glass, but which avoids the complexity of the comprehensive models by using simplified boundary conditions at the glass surface and by decoupling the combustion space gas flow analysis, as well. APPROACH The current approach uses a 3-D heat transfer and combustion zone model, which allows accurate assessment of local radiative heat transfer between gas heat sources, furnace gas volume, refractory walls and heat sinks. Furnace flow and flame heat release pattern are semi-empirically prescribed. However, flame heat release patterns are based on measurements in pilot-scale furnaces and actual flame observations. Thermal boundary conditions at the glass surface are simplified either by assigning effective temperatures and emissivities, or by use of a simple heat conduction model into the glass bath. Use of these simplified boundary conditions is justified by the fact that one of the goals of the current study is to identify 02-Burner configurations which generate similar net heat flux distributions as does conventional air firing. 2 |
