Woody biomass combustion modeling in moving grate furnaces

Abstract

Woody biomass combustion in moving grate furnaces occurs in two distinct regions: the solid bed region and the freeboard region above the fuel layer. This thesis reports about the work related to the development of a model to simulate the biomass bed combustion. The developed model is a 1-D transient model and it has been applied to study a multidimensional process such as the biomass thermal conversion in moving grate furnaces. This is possible since the physical variable gradients in the direction of the bed height are much higher than those in the direction of the bed length, thus, thermal and mass exchanges in the grate direction can be neglected. Considering these observations, the study of the combustion process in moving grate furnaces can be reduced to the observation of a vertical column of fuel that is moving along the grate inside the furnace and a 1-D transient system can be used to model the thermal conversion of the biomass fuel inside the combustion chamber. Biomass combustion modeling is central to defining the right strategies for controlling and optimizing the combustion process. Indeed, the quality of the biomass fuels in terms of physical and chemical properties can vary significantly. Thus, it is readily apparent that a detailed understanding on what happens when biomass is burned and, in particular, what changes occur when the fuel characteristics and the operational conditions vary, enables us to make the correct decisions with respect to the optimal design and operational principles of the biomass combustion applications. The process of biomass combustion involves a series of physical and chemical phenomena strictly coupled with each other and, irrespective of the complexity of the combustion technology and fuel characteristics, they can be divided in: initial heating, moisture evaporation, devolatilization of volatile matter ad char formation, heterogeneous and homogeneous reactions and bed shrinkage. While these phenomena proceed, the gases flowing through the packed bed and the biomass particles exchange heat and mass between each other. Under a mathematical point of view the simulation model is a 1-D system of partial differential equations (PDE’s). To solve this system, the calculation domain and the PDE’s have been discretized to obtain an algebraic system of equations. The calculation domain was divided into a number of non-overlapping control volumes which surround each grid point. In the meshed geometry, obtained with this procedure, all the variables are defined at the grid points. To discretize the diffusion term of the equations a piecewise linear profile for the variables was assumed, whereas for the convective term the upwind scheme was adopted. In this scheme, the internal conditions of the cell in term of concentration, temperature, velocity and density define the output of the calculation node. Once the system has been discretized, the Modified Newton Method was applied to minimize the equation residuals and find the solution vector. To come to the formulation of the simulation program an accurate literature review on biomass combustion modeling and experiments has been done. This part represents a complete outline on the state of the art on biomass combustion modeling and experiments and enabled us to make the right decisions regarding which mathematical formulations to adopt for the simulation of the different phenomena. The last part of the model development was its validation against experimental results. The comparison between the simulation outcomes and the experimental measurements showed good prediction capability of the developed program and allowed its utilization with the aim to study biomass bed combustion on moving grates furnaces. With this purpose a series of six simulations were performed and their results were presented and discussed. The simulations took into account the same kind of woody biomass with two different amounts of moisture and under different operational conditions in terms of inlet air flow rate and temperature. The results have shown that a higher moisture content increases the combustion time since it reduces the bed temperature and slows down the combustion reaction rate; moreover, a reduction of the inlet air flow rate reduces the combustion time in case of reaction limited regime because of the reduction of the cooling effect of the gases that pass through the bed. Finally, the utilization of pre-heated air reduces the duration of the combustion process because the biomass bed at the reaction front is already drier and warmer. This ensures a higher combustion rate. To conclude it has to be considered that, the outcomes of the developed model, in terms of gas temperature, composition and flow rate, are structured to be used as inlet data for a Computational Fluid Dynamics (CFD) model of the gaseous phase combustion in the chamber above the solid bed, in order to have a comprehensive understanding of the biomass combustion dynamics.