Masayuki Horio

DEM simulation of fluidized beds for gas-phase olefin polymerization

Abstract The temperature behavior of particles and gas in a fluidized bed reactor for polyolefin (PO, i.e., polyethylene (PE) and polypropylene (PP)) production was numerically analyzed based on the Discrete Element Method (DEM). Simulation was performed using a numerical code (modified SAFIRE code) by modifying SAFIRE ver.1 of Mikami, Kamiya and Horio (1998, Chemical Engineering Science, 53 , 1927–1940) by incorporating the energy balance and the reaction rate. Heat transfer from a particle to gas was estimated using the Ranz–Marshall equation. The reaction rates were calculated by Arrenius type zero-th order kinetic expressions including the effect of catalyst weight. Hot spot formation at the bottom corner immediately above the perforated plate distributor was observed even when an equal amount of monomer gas was fed to all orifices. When the gas was not uniformly supplied, rather stable particle swirls formed in the bed creating a hot spot at the core of each swirl. The degree of mixing was found an effective parameter to predict the hot spot formation.

Three-dimensional flow visualization of dilutely dispersed solids in bubbling and circulating fluidized beds

Three-dimensional flow structures of dilute suspensions in the freeboard of a bubbling fluidized bed as well as in a circulating fluidized bed of 200 mm internal diameter, roughly a (125)-scale model of a commercial boiler, were observed by the laser sheet technique. With this technique the suspension behavior in an arbitrary cross section can be observed. Furthermore, by applying multiple laser sheets the three-dimensional structure of the suspension can be investigated. In the dilute suspension above a bubbling fluidized bed, the so called “ghost bubbles” were clearly visualized. In the dilute phase transportation regime the existence of clusters was confirmed. The average shape of a cluster was a paraboloid heading downward and enclosing a gas wake in the upper side. From a cluster, particles were shed to the dilute phase continuously and these were again absorbed by other clusters. Cluster sizes and their velocity distributions were determined from the video image analysis. The present data would serve for the future construction of mathematical models on the flow structure of circulating fluidized bed reactors.

Effects of gasifying conditions and bed materials on fluidized bed steam gasification of wood biomass

The effect of steam gasification conditions on products properties was investigated in a bubbling fluidized bed reactor, using larch wood as the starting material. For bed material effect, calcined limestone and calcined waste concrete gave high content of H(2) and CO(2), while silica sand provided the high content of CO. At 650 degrees C, calcined limestone proved to be most effective for tar adsorption and showed high ability to adsorb CO(2) in bed. At 750 degrees C it could not capture CO(2) but still gave the highest cold gas efficiency (% LHV) of 79.61%. Steam gasification gave higher amount of gas product and higher H(2)/CO ratio than those obtained with N(2) pyrolysis. The combined use of calcined limestone and calcined waste concrete with equal proportion contributed relatively the same gas composition, gas yield and cold gas efficiency as those of calcined limestone, but showed less attrition, sintering, and agglomeration propensities similar to the use of calcined waste concrete alone.

SOLID DISTRIBUTION AND MOVEMENT IN CIRCULATING FLUIDIZED BEDS

A comprehensive diagnosis system was developed and applied to obtain detailed information on the mechanism of gas-solid contact in a circulating fluidized bed (CFB). In the present system pressure distribution and solid circulation rate can be continuously monitored. Particle concentration and the velocity of clusters were measured by optical fiber probes with a FFT analyzer. The measured velocity distributions showed clearly the formation of annular flow in the riser. The annulus thickness changed little along the column height even though there existed a sigmoidal distribution of particle concentration. The cluster size increased downward in the annulus but in the core it was much smaller and mostly constant. It was found that the internal circulation is as strong as the external circulation. Examination of previous models indicated that the formation of the annular flow is tightly related to the presence of clusters.

Particle and bubble movements around tubes immersed in fluidized beds – a numerical study

A numerical study was conducted based on the discrete element method (DEM) to analyze the behavior of particles and bubbles in fluidized-bed combustors with immersed tubes. Effects of fluidization velocity and tube arrangement were investigated in terms of bed voidage, particle–tube impact velocity, impact angle and the number of impacts around a tube for the case of a two-dimensional bed. Calculated particle and bubble flow patterns around a tube were in good agreement with previous experimental findings. Particle–tube impacts were found to be concentrated mainly during the initial period of bubble wake attack. The maximal impact velocities were observed on those parts of the surface of a single tube corresponding to 60° arc deflection with respect to the bottom of the tube. The maximal tube erosion rates are predicted for those regions. As for staggered and inline tube banks, quite different particle impacts and bubble behaviors were observed. Bubbles were easily elongated and tended to pass vertically through the lane between the inline tube columns. The distributions of average particle impact velocity and impact angle around a tube were generally asymmetrical. This is attributed to local bubble-passage habit. Validations were performed using a two-dimensional fluidized bed. The experimental results compared well with the DEM simulation results.

NUMERICAL SIMULATION OF ENTRAINED-FLOW COAL GASIFIERS PART II: EFFECTS OF OPERATING CONDITIONS ON GASIFIER PERFORMANCE

The comprehensive entrained flow coal gasification model developed in Part I was used for parametric studies to provide a better understanding of two-stage air blown entrained flow gasifiers. A series of numerical tests was performed for the 200 t/d pilot-scale gasifier under various operating conditions (heterogeneous reaction rate, coal type, particle size, and air/coal partitioning to the two stages). The coal conversion, product gas composition, calorific value and gas temperature profiles throughout the gasifier were simulated. The results show that coal devolatilization and char oxidation were responsible for most of the carbon conversion (up to 80%) in the two-stage air blown entrained flow gasifier. The predicted carbon conversion was independent of devolatilization rate, sensitive to the chemical kinetics of heterogeneous reactions on the char surface, and less sensitive to a change in coal particle diameter. It was found that the air ratio had a significant effect on gasifer performance with strong coal type dependence. The effect of air/coal partitioning to the two stages, and the feed rate of recycle char was found to be limited.

Use of numerical modeling in the design and scale-up of entrained flow coal gasifiers

A comprehensive model for entrained flow coal gasifiers was developed. In addition to the numerical methods and sub-models conventionally used in the modeling of pulverized coal combustion, a Multi Solid Progress Variables (MSPV) method was used to simulate the gasification reaction and reactant mixing process. The gas flow fields, gas temperature distributions, extent of burnout, and particle trajectories as well as particle concentrations within a two-stage air blown gasifier and its four variations were predicted by using the simulation model. The simulation results provided reasonable explanation on the experimental observations and demonstrated the ability of the model to sensitively account for subtle changes in design parameters such as the throat diameter ratio and swirl ratios of burner injections. It was concluded that the throat diameter was critical for the gas flow field in the entrained flow gasifier, which might control the behavior of molten ash slag deposition on the gasifier wall. The importance of selecting a proper swirl ratio for different burner sets was demonstrated.

On the nature of turbulent and fast fluidized beds

Abstract Fluidization characteristics have been investigated over the range from bubbling to fast fluidization by using a circulating fluidized bed cold model (riser diameter: 50 mm i.d., riser height: 2 390 mm, powder: fluidized cracking catalyst (FCC) and silica sand). Special focus was placed on the concepts of turbulent and fast fluidization regimes. Based on careful measurements of flow structure and cluster behavior, it is demonstrated that the turbulent fluidization regime is a transition regime between bubbling and fast fluidization. Experimental evidence supports the postulate that in fast fluidization the clusters adjust their size so that the gas drag force acting on them compensates with their gravity.

Prediction of agglomerate sizes in bubbling fluidized beds of group C powders

A mathematical model to predict an equilibrium agglomerate size in agglomerating fluidized beds of Geldart group C powders was derived based on the balance of bed expansion force caused by bubbles and agglomerate-to-agglomerate cohesive rupture force. The bed compaction force caused by bubbles of ordinary sizes was found not sufficient to break agglomerates. Among the two solutions the one corresponding to a stable equilibrium was calculated by a simple iteration. The model was successfully validated by experimental data in the literature and additional new data in the present work. Previous models are also critically reviewed and evaluated.

The mechanism of defluidization of iron particles in a fluidized bed

To develop a qualitative model for the fluidization characteristics of cohesive iron particles, a systematic investigation was performed focusing attention on the mechanism of defluidization and the particle-to-particle neck growth. The neck growth rate determined by a scanning electron microscope (SEM) agreed fairly well with the prediction by the surface diffusion model. Below 1200 K, cohesion force was successfully explained by the neck growth model with the surface diffusion mechanism. The tensile strength of a neck measured by the diametral compression test was 20 MPa. This value agreed with the one obtained from the bed breaking velocity measurement. Defluidization criterion was discussed based on the balance between bubble buoyancy force and neck force. The prediction was validated experimentally.