Modeling Bio-Dust Combustion in Suspension Firing

Abstract

During the past decades there has been an increasing focus on climate change, and consequently also a demand for energy production methods with a lower net CO2 emission. One possible options is suspension firing of biomass. The focus in this thesis has been to model the biomass devolatilization as happening in pulverized fuel fired boilers. The aim has been to add to the knowledge on suspension firing and consequently allow for an increased scientific basis for development of new biomass boilers and optimization of already exciting plants. The thesis is divided into a literature study and three main parts. The literature review briefly covers the basics of biomass particle devolatilization. Devolatilization is a subprocess happening during combustion of a biomass particle. It is the release of volatiles from the particle after heat up and evaporation of water and before the oxidation of said volatiles and the remaining char. Devolatilization can also happen concurrently with evaporation and oxidation processes, but for simplicity, the phenomena are often considered and modeled separately. The literature study also describes suspension firing equipment and conditions, which include high heating rates (> 103 K/s), high maximum temperatures (> 1300 K), and small particle diameters (< 3 mm). The first model presented in the thesis is based on multivariate data analysis, and estimates the char yield obtained from high heating rate pyrolysis. The char yield differs markedly when performing devolatilization at low and high heating rates, so the heating rate is an important factor to include, if one wants a model, which can adequately predict char yield obtained under suspension firing conditions. The model presented here is simple yet accurate and validated against experimental data. It can be used to define important input parameters for computational fluid dynamics (CFD) or other rigorous devolatilization models. The second model presented in this thesis revolves around the influence of biomass particle morphology on devolatilization under suspension firing conditions. Biomass particles are typically more elongated than the historically combusted coal particles, which have often been modeled as spheres. Spheres are one dimensional in nature and thus easier to implement in models, consequently this has typically also been the approach for biomass particles. The best simple geometry to mimic the shape of elongated biomass particles is, however, the cylinder. Here a model is presented, which accounts for gradients in temperature and mass, and can describe the devolatilization of both spherical and cylindrical particles. Thereby, differences in devolatilization of different particle morphologies can be compared. The model has been compared to relevant experimental data from literature and shows good agreement. The third model presented in this thesis is also based on multivariate data analysis and aims to include the gradient effects in the cylindrical model without adding to the computational costs in a CFD simulation. This is done by including the heat transfer limitations in the parameters of the kinetic scheme. By lumping the kinetic and the heat transfer limitations in one single first order reaction (SFOR) a simple isothermal particle model can describe the devolatilization of cylindrical biomass particles under suspension firing conditions without putting a strain on computational resources. In conclusion, this work contributes to the knowledge on devolatilization and gives an insight into how it can be modeled using multivariate data analysis. The developed simple models can be implemented into CFD without adding to the computational costs. Furthermore, the morphology model is a tool for investigating the effect of particle morphology and for assessing to which degree biomass particle characterization is necessary.