Popis: |
This thesis is an investigation into the design of a low volatile low NO„ pulverised coal burner. Low NO, pulverised coal burners are designed to achieve high combustion effi-ciency whilst lowering NO„ emissions, primarily by air staging. Due to particle inertia and interaction the distribution of the pneumatically conveyed fuel is affected by the geometry of the piping system and burner. Optimisation of the fuel feed has so far been limited. By increasing the particle concentration at the outlet, the ignition, flame stability and subsequently NO performance for difficult low volatile coals could be improved. Doosan Boiler R&D Centre have previously investigated the combustion of a range of low volatile coals on a standard burner and as coal volatile content decreased, stability and turndown capability were reduced. The industrially based research project proposed the use of the commercial Computa-tional Fluid Dynamics (CFD) code ANSYS FLUENT as a design tool to assess the particle concentration at the burner outlet. CFD potentially offers advantages over experiments in terms of time and cost, particularly at the industrial scale. The minimum particle concentration required to achieve early ignition and a stable flame has been estimated as 2kg,ai/kgair for Hongai coal (7% volatiles) based on a method which accounts for both volatile content and particle size. This local concentration means the particle flow is col-lision dominated and so the two-phase flow modelling approach has been assessed for a case in a scaled pulverised fuel rig and four validation cases. The scaled pulverised fuel pipe work case highlighted possible transients which could result from the geometry and the importance of coupling between the discrete and con-tinuous phase. This contributed to the decision to use a transient approach. Additional particle force models including wall roughness, inter-particle collision, lift forces, turbu-lence effects and structure dependent drag have been assessed against experimental data for vertical, horizontal and two 90° bend geometries. As understood from the literature, accurate simulation of gas-particle flows requires inclusion of these models. The differ-ences in scale for each of the cases investigated during the course of the work showed clearly that one universal approach was not applicable to all and this lead to the proposal of a structure dependent drag model, improvements to the inter-particle collision model and an understanding of the shortcomings of the particle turbulence model and the coarse grain particle assumption. The final two-phase flow model developed in this work was a significant improvement over standard industry practice. The two-phase flow model is applied to a test case of Doosan's burner firing Hongai coal. However, excessive particle concentration was further exacerbated for this scale and geometry. By utilising one-way momentum coupling a prediction of the path of the discrete phase was made. The prediction of particle concentration at the outlet of the burner is comparable to previous combustion test results and indicates that the burner only partially meets the concentration requirement for Hongai coal. |