Modelling of Biomass Combustion in Furnaces
Sammanfattning: Biomass combustion in grate fired furnaces is an important approach to convert renewable biomass fuel to heat and electricity. It is not only used in the Swedish energy industry but also by a great number of Swedish households for domestic heating. Burning biomass has the advantage of being CO2 neutral. However, it suffers from several difficulties, for example, the poor combustion efficiency ? there are considerable amount of emissions of unburned volatile gases and solid fuel particles. The emission of pollutants such as NOx, CO, dioxin, and also particulate matters, etc, damages our environment and has caused great public concern. To meet the stringent legislation on environment protection, efficient and clean biomass combustion furnaces must be developed. To achieve this goal, understanding of the biomass combustion process must be improved by systematic scientific investigations on the processes, and reliable simulation and design tools must be developed. The aims of this thesis work are to study the fundamental details of biomass combustion in grate-fired boilers by using computational methods, and to develop and validate modelling tools to study biomass combustion. The three major aspects of biomass combustion in grate fired boilers have been studied. They are the combustion process in the fuel bed, the volatile combustion process in the free board and the radiation heat transfer process in the free board and between the flames and the fuel bed. Sub-models for these different processes have been developed and validated against available experimental data. A two-zone and a three-zone bed model for biomass combustion are developed based on the functional group concept and an existing coal combustion model. To account for the spatial inhomogeneity of the fuel bed, a detailed quasi two-dimensional bed model considering detailed transport inside each particle has been developed. The two-zone model (for counter-current beds) and the three-zone model as well as the quasi two-dimensional model (for cross-current beds) are applied to simulate the volatile compositions from the bed. The gas phase combustion process in the free board is modelled using Favre averaged Navier-Stokes equations, together with transport equations for enthalpy and mass fractions of different species. Turbulence is taken into account by the two-equation k-epsilon closure. The chemical reactions are modelled using global mechanisms for both the oxidation of volatiles and for the NOx emissions and re-burning. Different NOx models have been investigated. Several different radiation models for the radiation heat transfer process in the free board have been investigated, including the P1 model and the FVM model. In addition, assumptions of optical thick and optically thin have been examined in different boiler applications. In addition to the model development and validation, an efficient boundary correction method has been developed to remedy the stiffness problem in modelling the various small secondary air jets in large scale boilers. Two large-scale grate-fired district heating boilers, one small-scale household heating boiler and one laboratory scale pellets-fired reactor have been studied within the thesis work. The studies have revealed the important features in grate fired biomass boilers. For example, it was shown that turbulence plays important role in the volatile oxidation process ? the laboratory scale pellet reactor study showed that the turbulence level in the primary combustion zone has a dominant influence on the temperature and species distributions. The study of a large scale 50 MW industry boiler has demonstrated the advantage and potential of the computational methods in future boiler analysis and design. The present thesis work has also shown the complexity of the process and limitations of the models investigated. This lays a solid ground for the future development of reliable tools for the computational analysis of biomass combustion in grate fired furnaces.
KLICKA HÄR FÖR ATT SE AVHANDLINGEN I FULLTEXT. (PDF-format)