Improving the prediction of convection bank heat transfer.

Many sugar mill boilers with baffled convection banks are susceptible to tube erosion failure. SRI has successfully reduced tube erosion at many sugar factories by redesigning convection banks using the CFD (Computational Fluid Dynamics) code FURNACE. In some cases where relatively small changes have been made to extend the life of a tube bank by a few years, a significant reduction in convection bank heat transfer performance has occurred. This reduces boiler efficiency and increases bagasse consumption. The FURNACE code has been able to predict whether or not modifications to convection bank design will change heat transfer performance but has not been able to accurately predict the magnitude of these changes. This paper describes modifications that have been made to the FURNACE code to improve the prediction of heat transfer performance. Comparisons with boiler operating data are used to assess the upgraded FURNACE code. In this work the ability of the FURNACE code to predict the changes in convection bank heat transfer performance caused by baffle modifications has been improved significantly. The upgraded FURNACE code successfully predicted most of the increase in convection bank exit gas temperature caused by convection bank baffle modifications in two boilers. For one boiler the model predicted a convection bank gas exit temperature rise of 33 o C versus an actual temperature rise of 36 o C. The predicted convection bank gas exit temperature rise for the other boiler was 76 o C and the actual temperature rise was 92 o C. Further testing of the model will be performed using data from other boilers with a wider range of convection bank baffle configurations. The model will be applied in the future to achieve the combined aims of improved heat transfer and reduced wear in convection banks and other boiler components.