Hydrodynamic modeling of a circulating fluidized bed

Hydrodynamics plays a crucial role in defining the performance of circulating fluidized beds (CFB). The numerical simulation of CFBs is very important in the prediction of its flow behavior. From this point of view, in the present study a dynamic two dimensional model is developed considering the hydrodynamic behavior of CFB. In the modeling, the CFB riser is analyzed in two regions: The bottom zone in turbulent fluidization regime is modeled in detail as two-phase flow which is subdivided into a solid-free bubble phase and a solid-laden emulsion phase. In the upper zone core-annulus solids flow structure is established. Simulation model takes into account the axial and radial distribution of voidage, velocity and pressure drop for gas and solid phase, and solids volume fraction and particle size distribution for solid phase. The model results are compared with and validated against atmospheric cold bed CFB units' experimental data given in the literature for axial and radial distribution of void fraction, solids volume fraction and particle velocity, total pressure drop along the bed height and radial solids flux. Ranges of experimental data used in comparisons are as follows: bed diameter from 0.05-0.418 m, bed height from 5-18 m, mean particle diameter from 67-520 μm, particle density from 1398 to 2620 kg/m³, mass fluxes from 21.3 to 300 kg/m²s and gas superficial velocities from 2.52-9.1 m/s.As a result of sensitivity analysis, the variation in mean particle diameter and superficial velocity, does affect the pressure especially in the core region and it does not affect considerably the pressure in the annulus region. Radial pressure profile is getting flatter in the core region as the mean particle diameter increases. Similar results can be obtained for lower superficial velocities. It has also been found that the contribution to the total pressure drop by gas and solids friction components is negligibly small when compared to the acceleration and solids hydrodynamic head components. At the bottom of the riser, in the core region the acceleration component of the pressure drop in total pressure drop changes from 0.65% to 0.28% from the riser center to the core-annulus interface, respectively; within the annulus region the acceleration component in total pressure drop changes from 0.22% to 0.11% radially from the core-annulus interface to the riser wall. On the other hand, the acceleration component weakens as it moves upwards in the riser decreasing to 1% in both regions at the top of the riser which is an important indicator of the fact that hydrodynamic head of solids is the most important factor in the total pressure drop.     

Dynamic characteristics of bed collapse in threephase fluidized beds

Transient behavior of the bed collapse after shut-off the gas supply into a three-phase fluidized bed was determined. Experiments were carried out in a 210-mm diameter half-tube acrylic column having a 1.8 m-high test section. The polymer beads (d _p =3.2 mm,ρ _s = 1,280 kg/m³) were fluidized by cocurrent flow of deionized water and air. The transient behavior of the bed collapse after cut-off the gas supply to the bed was monitored by a video camera (30 frames/s). The dense bed surface height was measured from the image of videotape. At lower liquid velocity, the dense bed surface increases with the elapsed time and then reaches a bed height, whereas at higher liquid velocity the dense bed surface increases sharply with the elapsed time, then decreases and reaches the bed height corresponding to the liquid-solid fluidized beds (water-polymer beads).     

The effect of bed materials on the solid/bubble motion in a fluidised bed

Gas fluidisation provides good mixing and contact of the gas and particle phases as well as good heat transfer. These attractive features are achieved by the high degree of bubble-induced particle circulation within the bed. Bubble and particle motion vary with bed materials and operating conditions, as investigated in the present study, by the use of the non-intrusive positron emission particle tracking (PEPT) technique. The selected materials were spherical polyethylene and glass particles.The data obtained by the PEPT technique are used to determine the particle velocities and circulation pattern. Bubble rise velocities and associated sizes can be inferred from the particle velocity data, since particles travel upwards mostly in the bubble wake. The results indicate that the flow structure and gas/solid motion within the fluidised beds were significantly different, even at the same value of the excess gas velocity, U-U_mf. The solid circulation pattern within the beds differ: if for glass beads, a typical UCDW-pattern existed (upwards in the centre of the bed, downwards near the wall), the pattern in the polyethylene bed is more complex combining a small zone of UWDC movement near the distributor and a typical UCDW-pattern higher up the bed. Transformed data demonstrate that at the same value of excess gas velocity, U-U_mf, the air bubbles in the polyethylene fluidised bed were smaller and rose more slowly than in the fluidised bed of glass beads, thus yielding a longer bubble residence time and improved gas/solid contact. For polyethylene beads, the size and rise velocity of air bubbles did not increase monotonically with vertical position in the bed as would be predicted by known empirical correlations, which however provide a fair fit for the glass beads data. Bubble sizes and solid circulation patterns are important parameters in the design of a fluidised bed reactor, and vary with the bed material used.     

The screened waveguide for intrusive acoustic emission detection and its application in circulating fluidized bed

A screened waveguide with a 90° elbow, which had a rubber surface to reduce noise generations and a steel/air interface to block noise propagations, was designed to measure local particle movements in certain directions. In the riser of circulating fluidized bed, the noise energy only accounted for 2.1% of the total energy received by the waveguide. Besides, the radial acoustic emission energy distribution detected via the waveguide was highly consistent with the radial solid flux distribution measured by the extraction probe. Furthermore, the radial flow patterns in the riser with Geldart D particles were measured under dense phase conveying by this new method, which always demonstrated as core-annulus flows and the transition points (r/R) increased with decreasing solid–gas ratios.     

The solids flow in the CFB-riser quantified by single radioactive particle tracking

Single radioactive particle tracking was used to measure the overall solids residence time (and its distribution) in the riser of a CFB, operating at superficial air velocities (U) of 1 to 9 ms^− 1 and solids circulation fluxes (G) between 20 to 600 kg m^− 2s^− 1.The results demonstrate that the particle motion and mixing differ according to the operating mode of the riser, with a fairly constant velocity throughout the riser achieved in the dilute or dense riser flow, but with a significant amount of back-mixing for intermediate values of U and/or G. This back-mixing is due to the core-annulus mode of particle flow. Whereas experimental results and theoretical predictions are in fair agreement for the dilute and dense riser flow, the core-annulus regime needs to account for a U and G dependent slip factor (φ), in excess of the commonly proposed value φ = 2, especially at U-U_TR < 2 ms^− 1.Moreover, the previously published riser operation diagram is confirmed by the experiments, although a further analysis of the core-annulus regime is needed.     

Particle motion in the CFB riser with special emphasis on PEPT-imaging of the bottom section

The riser is the key-part of a circulating fluidized bed (CFB) and its hydrodynamics are determined mainly by the combined operating superficial gas velocity, U, and solids circulation flux, G. The bottom part of the riser contributes to the total pressure drop of the riser and affects the solids residence time in the riser, due to the possible existence of a dense bed and to the presence of an acceleration zone. Positron Emission Particle Tracking (PEPT) is applied to study these phenomena by measuring the real-time particle motion in a riser of 0.09 m diameter, defining (i) the extent of the acceleration zone, including acceleration length and acceleration time; (ii) the occurrence of a bubbling/turbulent bed under specific conditions of U and G; (iii) the establishment of a fully developed flow immediately after the acceleration zone; (iv) the occurrence of core-annulus flow under specific combinations of U and G; and (v) the disappearance of the intermediate core-annulus region at high values of U and G, where riser hydrodynamics will be either dilute or dense solid up-flow.The particle upflow velocity, U_pf, after acceleration was measured and compared with the situation of dilute transport. When the solids circulation flux increases, the dilute transport mode no longer prevails, and U_pf should be calculated using an appropriate slip factor, itself a combined factor of U and G. The acceleration length and time are nearly constant, at an approximate average of 0.26 m and 0.21 s respectively, independent of U and G. The acceleration length can be modelled fairly accurately, using a C_D-factor of approximately 3.2, which is about half the value predicted by empirical equations established for dilute transport.Dense Suspension Upflow (DSU) is achieved when G exceeds ~ 130 kg m⁻ ² s^− 1.     

Flow regime transition identification in three phase co‐current bubble columns

Bubble columns have wide applications in absorption, bio‐reactions, catalytic slurry reactions, coal liquefaction; and are simple to operate, have less operating costs; provide good heat and mass transfer. Experiments have been performed for identifying transition regime in a 15 cm diameter bubble column with liquid phase as water and air as the gas phase. Glass beads of mean diameter 35 µm have been used as solid phase. The superficial gas velocity is in the range 0 ≤ U_g ≤ 16.3 cm/s and superficial liquid velocity in the range of 0 ≤ U_l ≤ 12.26 cm/s. Solid loading up to 9% (w/v) has been used. Pressure signals have been measured using differential pressure transducers (DPTs) at four different axial locations. Classical analysis (Wallis approach and Zuber–Findlay approach), Statistical analysis and Fractal analysis have been used for regime transition identification. Statistical analysis and Fractal analysis have shown almost the same transition points for all the liquid and gas velocities. Effect of solid concentration, liquid velocity and gas velocity over transition regime has also been studied. As the solid concentration is increased it has insignificant effect over transition regime for lower values (<1%), while transition values decrease for higher solid concentration (>1%). © 2012 Canadian Society for Chemical Engineering     

气固环流燃烧器内颗粒流动行为

针对石油焦及气化余焦的燃烧特点和流态化特性,提出了一种采用气固密相环流烧焦与快速床管式烧焦技术相组合的新型燃烧器结构。在不同操作条件(导流筒区表观气速0.772～1.674m.s-1,环隙区表观气速0.223～0.519m.s-1,装置系统的颗粒外循环强度40.8～229.4kg.m-2.s-1)及两类颗粒体系下,采用光纤测量仪对组合燃烧器环流段内颗粒流动特性进行了系统的实验研究。结果表明,两类颗粒体系的固含率和颗粒轴向速度在导流筒区、底部区和颗粒分流区床层内沿径向的分布规律为中心区小、边壁区大的环-核型分布,体现了气固流化床典型聚式流态化的非均一性特征;在环隙区,受环流段结构的影响,两类颗粒体系的固含率和颗粒轴向速度参数沿床层径向的分布相对较均匀;混合颗粒体系的固含率、颗粒轴向速度较单一石英砂颗粒体系的要小,细颗粒的加入在一定程度上能改善气固混合的均匀程度;两类颗粒体系在底部区和颗粒分流区的径向流动具有剪切破碎气泡的作用,有利于环流段内气固的充分混合接触。     

Predictions on bubble and solids movement in laboratory polyethylene fluid beds as visualized by X-ray computer assisted tomography (CAT) scanning

Solid and gas distributions are tomographically quantified as a function of position with high resolution in a series of laboratory fluid beds containing air and polyethylene particles. The resolution used is 0.4 mm by 0.4 mm by 3 mm. The laboratory models are Plexiglas columns of 10 cm in diameter and the settling bed L/D ratios vary between one and three. Large particles (up to 1.5 mm in diameter) of high density polyethylene and linear low density polyethylene are used. The superficial gas velocities vary from the minimum fluidization velocity to 50 cm/s. In this paper, the analysis of fluid bed CAT scanner images is extended to show bubble, emulsion and dense phase distribution. The analysis is also used to determine the bubble diameter and to predict the flow direction of solid particles as well as the velocity of descending solids. The voidage frequency distributions of a bed at different gas flow rates are compared to each other and the voidage threshold values corresponding to gas, emulsion and dense phases are determined. These threshold values are used to prepare ternary images that clearly show the parts of the bed cross-section corresponding to bubble, emulsion and dense phases.     

Experimental Study on Bubble Rising and Descending Velocity Distribution in a Slurry Bubble Column Reactor

To determine bubble rising and descending velocity simultaneously, a BVW‐2 four‐channel conductivity probe bubble parameters apparatus and its analysis are used in gas‐liquid and gas‐liquid‐solid bubble columns. The column is 100 mm in internal diameter and 1500 mm in height. The solid particles used are glass beads with an average diameter of 17.82 μm, representing typical particle size for catalytic slurry reactors. The effects of superficial gas velocity (1.0 cm/s ≤ U_g ≤ 6.4 cm/s), solid holdup (0 % ≤ ?_s ≤ 30 %), and radial location (r/R = 0, 0.4, and 0.7) on bubble velocity distributions are determined. It is found that increasing U_g can increase the velocity of bubbles but do not exert much influence on bubble velocity distribution. Solid holdup mainly affects the distribution of bubble velocity while the radial direction affects bubble velocity distribution only slightly. The ratio of descending bubbles to rising bubbles increases from the bubble column center to the wall. It can be proved experimentally that large bubbles do not always rise faster than small bubbles at higher U_g (for example 6.4 cm/s).     

Heat transfer in three-phase (G/L/S) circulating fluidized beds with low surface tension media

Characteristics of heat transfer were investigated in a three-phase circulating fluidized bed whose diameter and height were 0.102 m (ID) and 2.5 m, respectively. Effects of gas and liquid velocities, particle size (0.5–3.0 mm), solid circulation rate (2.0–6.5 kg/m² s), and surface tension (47.53–72.75×10⁻³ N/m) of liquid phase on the heat transfer coefficient were examined. It was found that the heat transfer coefficient (h) between the immersed vertical heater and the riser proper of the three-phase circulating fluidized bed increased with increase in gas and liquid velocities, but did not change considerably with a further increase in liquid velocity, even in the higher range. The value of heat transfer coefficient increased gradually with increase in the size of fluidized solid particles without exhibiting the local minimum, which represented that there was no bed contraction in three-phase circulating fluidized beds due to the higher liquid velocity. The heat transfer system could attain a stabilized condition more easily with increase in particle size. The value of heat transfer coefficient increased with increase in solid circulation rate in all the cases studied due to the increase of solid holdup in the riser. The value of heat transfer coefficient decreased with increase in surface tension of liquid phase, due to the decrease of bubbling phenomena and bubble holdup. The decrease in liquid surface tension could lead to an increase in elapsed time from which the temperature difference between the heater surface and the riser became an almost constant value. The experimentally obtained values of heat transfer coefficient were well correlated in terms of dimensionless groups as well as operating variables.     

Liquid hold-up and minimum fluidization velocity in a turbulent contactor

Transient-response experiments have been performed in conjunction with bed expansion measurements to determine liquid hold-ups and minimum fluidization velocities in a 12-in. turbulent-bed gas-liquid contactor. The amount of liquid hold-up was found to be independent of gas velocity, but dependent upon both liquid rate and packing diameter in the same manner as reported for conventional fixed-bed absorbers. The data on minimum fluidization velocity, G_mf, which was interpreted in the present study as the maximum gas mass velocity at which the bed maintained its static height, showed a considerable variation with packing diameter, d_p and liquid flow rate, L. A correlation of G_mf with d_p and L was presented.     

Study of Local Gas Holdup and Specific Interfacial Area in a Split-Column Airlift Bioreactor Using Sphosticated 4-Point Optical Probe for Culturing Microlgae/Cyanobacteria

Local gas holdup (?) and interfacial area (a) at different axial locations of the riser and downcomer of a split-column airlift bioreactor were investigated using a sophisticated four-point optical probe. Such a type of a reactor has been found to outperform both bubble-column and draft-tube airlift bioreactors for culturing microalgae. The effect of superficial gas velocity (0.3–2.8 cm/s) on both gas holdup and interfacial area was studied using air–water system. It was found that both gas holdup and interfacial area significantly decrease from the top to the bottom of the downcomer for all superficial gas velocities, while their variation from the bottom to the top for the riser was found to be much less than that of the downcomer at the same superficial gas velocities. It was found that the interfacial area of the riser tends to increase by 35% from the bottom to the upper middle point of the column (6.15 Z/D from the bottom), then declines by 10% at the top location (7.7 Z/D from the bottom). Empirical correlations were obtained relating the gas holdup and specific interfacial area to superficial gas velocity of the riser and the downcomer of the bioreactor. It was found that the riser has to be represented as upper and lower halves to be best correlated, while the only upper half of the downcomer was successfully correlated. Having obtained variable interfacial area (a) at different locations of both the riser and the downcomer of the bioreactor, the local K_La consequently changes as a function of the location of the bioreactor and hence needs to be investigated locally as opposed to the current studies that have only measured and correlated the overall K_La.     

Wall-to-bed heat transfer characteristics of spout-fluid beds

Wall-to-bed heat transfer characteristics have been investigated in a rectangular spout-fluid (S–F) bed segment column (20 cm length, 5 cm width and 50 cm height) utilizing glass beads (D_p = 0.254, 0.388 and 0.461 mm) and air as fluid. Results indicate that h values in the S–F bed increase with increasing air mass velocity and particle diameter, and decrease with increasing bed height. Under identical flow conditions h values in the S–F bed were about 30% more than for the corresponding fluidized bed.     

Radial Liquid Dispersion and Bubble Distribution in Three‐Phase Circulating Fluidized Beds

The liquid dispersion and bubble distribution in the radial direction have been investigated in the riser of a three‐phase circulating fluidized bed whose diameter is 0.102m and 3.5m in height. Effects of gas and liquid velocities and solid circulation rate have been determined. It has been found that the radial distribution of bubbles is related closely to the liquid dispersion in the radial direction. The size and rising velocity of bubbles tend to increase as the radial position approaches to the center of the riser. The bubble size increases with increasing U_G, but it decreases with increasing U_L or G_S in all radial positions. The radial dispersion coefficient of the liquid phase increases with increasing U_G or G_S, however, it tends to decrease with increasing U_L. The value of Dr has been well correlated in terms of dimensionless groups based on the isotropic turbulence model.     

Particle segregation in a spouted bed of binary mixtures of particles

Minimum spouting velocity and segregation behaviour of binary mixtures of particles differing in size have been studied. The experiments were carried out in a bed of 20 cm diameter at superficial gas velocities up to 1.3 U_ms by use of silica sand of four different particle sizes from 0.655 to 2.23 mm. An empirical equation was proposed for the minimum spouting velocity of binary mixtures. The effects of the particle size difference and the superficial gas velocity on segregation were investigated. Results showed that considerable radial segregation as well as axial segregation occurred even for high gas velocity under the condition of large particle size difference.     

Layer inversion and mixing of binary solids in two- and three-phase fluidized beds

Water fluidization in a 210 mm diameter semi-cylindrical acrylic column of a binary solids mixture of 3.2 mm polymer beads (ρ_s=1280 kg/m³) and 0.385 mm glass beads (ρ_s=2500 kg/m³) at superficial liquid velocities from 18.1 to 43.1 mm/s is shown to generate layer inversion at a superficial liquid velocity, U_L, of 33.1 mm/s. Introduction of air with a superficial velocity, U_g, of 1.92 mm/s yielded a layer inversion velocity at U_L=30.4 mm/s. The latter is explainable if it is assumed that the determinant of layer inversion is the interstitial liquid velocity and that therefore the main function of the gas in this respect is to occupy space.Mixing of the binary solids, as quantified by a mixing index applied to measured particle compositions at different levels of the fluidized bed, is shown to be greatest at the layer inversion velocity for liquid fluidization and, in general, to increase as co-current gas flow increases at a fixed value of U_L.     

Particle motion at the wall of a circulating fluidized bed

The motion of alumina particles of mean size 74.9 μm in the region near to the wall of the 305 mm diameter riser of a cold model circulating fluidized bed has been studied using a high-speed video camera employing normal and magnifying lenses. Particles in this region were found to move predominantly downwards, against the main gas flow. High density groups or swarms of particles typically arch-shaped were observed to descend in contact with the wall at velocities in the range 0.3–0.4 m s⁻¹. Tfie distribution of swarm descent velocities was shown to be little affected by changes in superficial gas velocity over the range 3–5 m s⁻¹ and imposed mean solids mass flux over the range 2 to 80 kg m⁻² s⁻¹. A region of steady bulk downflow of solids with a velocity of approximately 1.0 m s⁻¹ was observed to appear a few millimetres from the wall at mean suspension densities greater than 5.6 kg m⁻³. Motion of particles in contact with the riser was analysed by identification of three flow forms; dilute, dense and swarm flow. Results of the analysis are linked with the observations of other workers concerning the onset of the so-called ‘similar profiles’ regime. The relationship between the measured effective particle swarm length and the cross-sectional mean suspension density was established and the implications for modelling suspension-to-wall heat transfer discussed.     

EFFECTS OF SECONDARY AIR INJECTION ON GAS-SOLID FLOW BEHAVIOR IN CIRCULATING FLUIDIZED BEDS

Effects of secondary air injection on the hydrodynamics such as solid holdup and gas-solid flow behavior were investigated in a circulating fluidized bed. The gas velocity in the riser, the ratio of secondary air velocity to that of primary air, and the solid circulating rate were chosen as operating variables. Fluid cracking catalyst(FCC) with a density of 1840 kg/m³ and a mean diameter of 74 um was employed as the solid phase. The secondary air was fed to the riser radially or tangentially at the wall of the column. Pressure drop fluctuations in the riser were measured and analyzed by adopting the stochastic method to characterize the effects of secondary air injection on the gas-solid flow behavior in the bed.

It has been found that the injection of secondary air into the riser can increase the solid holdup in the riser considerably, and that the tangential injection of secondary air is more effective for the increasing the solid holdup than the radial injection. However, the gas-solid flow behavior has been found to become less persistent with the injection of secondary air; the resultant flow behavior is more complex when the air is injected tangentially than radially. The solid holdups in the primary as well as secondary zones of the riser have been well correlated in terms of not only operating variables but also fractal dimension of the pressure fluctuations.