Modelling of Biomass Steam Gasification in a Bubbling Fluidized Bed Gasifier
Thesis DisciplineChemical Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
This thesis presents the numerical simulation study of the biomass-steam gasification in a bubbling fluidized bed (BFB) gasifier including the mathematical model development and the experimental validation of the model developed. The study focused on two main areas on developing the mathematical model: a one-dimensional (1D) non-isothermal reaction kinetics model and a two-dimensional (2D) model consisting of the reaction kinetics model for gasification reactions and two-dimensional CFD model for hydrodynamics characteristics.
The biomass gasification with steam as the gasification agent at elevated temperatures can be considered to include two main stages: initial pyrolysis and subsequent gasification reactions. In the first stage, the pyrolysis reactions of the biomass occurred instantaneously for a short duration decomposed into volatile gases, char and tar at the bottom of the gasifier filled with the bed materials. In the subsequent stage of gasification, reactions occurred among the gasification agent (steam), the volatile gases and the char evolved from the initial stage of pyrolysis at high temperatures. The producer gas from biomass-steam gasification mainly consists of CO, H2, CO2, CH4 and H2O and least quantities of higher hydrocarbons at elevated operating temperature. Hence the reactions such as Steam Gasification reaction, Water Gas-Shift reaction, Methanation reaction, Bouduard reaction, and Steam Methane Reforming reaction were considered. The final gas composition of the producer gas was affected by the products of pyrolysis process which were quantified with a product distribution function of temperature developed from experimental results in this work.
The development of the subsequent gasification reaction model was based on two-phase theory of gases and solids consisting of the particle-lean bubble phase and the particle-rich emulsion phase which were distributed homogeneously when the gas velocity through the bed was in excess of the minimum fluidization velocity. In the two-phase theory model, the heat and mass transfer rates were related to the fluidization characteristics of the bed. Therefore, understanding and quantitatively description of the hydrodynamics of the gas-solid within the gasifier were important which were adopted from literature for the development of the 1D model for the gasification process in the BFB.
In the 1D gasification model, the non-linear partial differential equations (PDEs) describing the mass and energy balances (for both phases) with the reactions kinetics based on Arrhenius correlations were numerically solved using a solver function from the PDE modules of Matlab software with properly defined initial and boundary conditions. In the development of the 2D gasification model, the developed reaction kinetics model was integrated into the 2D CFD Eulerian-Eulerian granular kinetic model called Gasification embedded 2D CFD model using CFD ANSYS FLUENT 12.1 package. In this 2D model, the built-in ‘laminar finite-rate’ model was adopted to describe the chemical kinetics using Arrhenius reaction kinetics expressions. The momentum equation considered laminar viscous model for laminar flow at low steam to biomass (S/B) ratio (lower than 0.6) and k-ε turbulence viscous model for transient to turbulent flow regime for high S/B ratio. These were solved using a Phase Coupled SIMPLE solver algorithm based on FVM.
To provide validation data for the developed biomass gasification models, experiments were undertaken on the 100kW DFB gasifier at the University of Canterbury. In the experiments, it had been found that the simulation results from the initial pyrolysis model were in close agreement with the experimental data with discrepancies of ±1.0% (mol/mol) for H2, ±0.8% (mol/mol) for CO, ±0.6% (mol/mol) for CO2 and ±0.3% (mol/mol) for CH4.
After this, the gasification agent steam was introduced for normal gasification operation under various operation conditions (temperature, S/B ratio). The experimentally measured producer gas compositions for the operating conditions of 680-780oC, 1 atmospheric pressure and S/B ratio of 0.53 ranged from 17.9% to 28.3% for H2, from 35.7% to 38.5% for CO, from 23% to 28.8% for CO2 and from 13% to 15% for CH4 (mol/mol on dry basis). Under the above operation conditions, the discrepancies between the experimentally measured producer gas compositions and the predicted results using the 1D model were, respectively, 4.5% for H2, 1.4% for CO, 7.5% for CO2 and 1.2% for CH4 (mol/mol on dry basis). For the 2D model, the discrepancies were, respectively, 2.4% for H2, 2.9% for CO, 4.9% for CO2 and 0.8% for CH4 (mol/mol on dry basis). However under the operating conditions of 780oC and S/B ratio of 0.53, the predicted CO2 and H2 concentrations from the 2D model were, respectively, 8.6% (mol/mol on dry basis) higher and 4.8% (mol/mol on dry basis) lower than the measured value.
The experimentally measured producer gas compositions for the operating conditions of 710oC and S/B ratio of 0.33-0.84 ranged from 24.4% to 32% for H2, from 32.7% to 44.2% for CO, from 15% to 21.8% for CO2 and from 13.6% to 16.4% for CH4 (mol/mol on dry basis). The discrepancies between experimentally measured producer gas compositions and the model predicted results for the above operating conditions were 1.6% for H2, 2.7% for CO, 1.8% for CO2 and 0.6% for CH4 (mol/mol on dry basis) for the 1D model while those for the 2D model were 4% for H2, 1.6% for CO, 1% for CO2 and 1.6% for CH4 (mol/mol on dry basis).
From the model validation, it was found that the 1D model results and 2D model simulation results were closely in agreement and show small discrepancy with the experimental results. In addition, the 1D model uses less computing time than the 2D model; therefore, the 1D model has been used to investigate the effects of operating conditions (temperature and S/B ratio) on the producer gas composition. It was observed that the gas concentration of CO, CO2 and CH4 in the producer gas decreased while the H2 increased with increasing operating temperature in the examined range from 680-780oC. Similarly the gas concentration of H2 and CO2 in the producer gas increased while CO and CH4 decreased with increasing S/B ratio in the examined range from 0.33-0.84.
The 2D model can be used to predict gas distribution within the gasifier thus it can be used to gain better understanding of the gasification process and effect of gasifier configuration and operating conditions on the gasifier performance. Further studies are proposed for improvements on the 2D model.