Hydrodynamics and flow regimes in turbulent bed contactor with non‐Newtonian liquids

The flow regimes normally encountered in a turbulent bed contactor (TBC) are static, partially fluidised, completely fluidised and flooding regimes. Experiments were conducted in a TBC operating in Type I mode to identify the flow regimes with non‐Newtonian liquid. Flow regime transition velocities were obtained from the pressure drop and bed expansion measurements at various operating and geometric variables. The variables include apparent viscosity of the liquid, gas and liquid velocities, size and density of the particles, and static bed height. The effect of the above variables on delineation of flow regime transition was studied. Based on the experimental data, correlations were proposed for predicting the transition velocity from one regime to the other. The influence of the variables on regime transition velocities is more or less similar to that observed for Newtonian liquids. © 2011 Canadian Society for Chemical Engineering     

Effect of louver baffles on hydrodynamics and gas mixing in a fluidized bed of FCC particles

The effect of louver baffles on the particle concentration profiles, pressure fluctuations, bed expansion, and gas mixing of a fluidized bed was investigated in a transparent 2-D column of cross-section 500×30 mm and height 6 m over a broad range of operating conditions covering both the bubbling and turbulent flow regimes. Visual observations, pressure fluctuations and steady gas tracer experiments showed that louver baffles can break bubbles, as indicted by the lower amplitudes and higher mean frequencies of differential pressure fluctuations, but they were only effective for superficial gas velocities <∼0.7 m/s for the FCC particles considered in this study. The ability of louver baffles to break bubbles reached a maximum near the onset of the turbulent flow regime. A gas cushion of low particle concentration appeared below the louver baffle, and its height increased with increasing superficial gas velocity, indicating increasing suppression of solids backmixing. Internal emulsion circulation was promoted above the louver baffle, causing an uneven distribution of gas flow. The addition of louver baffles reduced the upstream tracer gas concentrations by 80-90%, indicating a significant decrease in the backmixing fluxes of both gas and solids across the baffle layer. The tracer gas concentrations above the louver baffles increased resulting from the promoted emulsion circulation by louver baffles.     

Fluidization of Group B particles with a rotating distributor

A novel rotating distributor fluidized bed is presented. The distributor is a rotating perforated plate, with 1% open-area ratio. This work evaluates the performance of this new design, considering pressure drop, Δp, and quality of fluidization. Bed fluidization was easily achieved with the proposed device, improving the solid mixing and the quality of fluidization.In order to examine the effect of the rotational speed of the distributor plate on the hydrodynamic behavior of the bed, minimum fluidization velocity, U_mf, and pressure fluctuations were analyzed. Experiments were conducted in the bubbling free regime in a 0.19 m i.d. fluidized bed, operating with Group B particles according to Geldart's classification. The pressure drop across the bed and the standard deviation of pressure fluctuations, σ_p, were used to find the minimum fluidization velocity, U_mf. A decrease in U_mf is observed when the rotational speed increases and a rise in the measured pressure drop was also found. Frequency analysis of pressure fluctuations shows that fluidization can be controlled by the adjustable rotational speed, at several excess gas velocities.Measurements with several initial static bed heights were taken, in order to analyze the influence of the initial bed mass inventory, over the effect of the distributor rotation on the bed hydrodynamics.     

Detailed measurements of flow structure inside a dense gas–solids fluidized bed

Detailed local flow structures are investigated in bubbling and turbulent fluidized bed with FCC particles. The operating conditions ranges from 0.06 to 1.4 m/s. Extensive experiments are carried out using a newly developed optical fiber probe system, which can measure the solids concentration and velocity at multi-points. The results reveal that with increasing U_g, local solids concentrations go through three evolution stages, reflecting a gradual regime transition process. Under all operating conditions, upflowing and descending particles co-exist at all measuring locations. The upflowing particle velocity is strong function of both superficial gas velocity and spatial position. However, the descending particle velocity mainly depends on superficial gas velocity. The bed radial symmetry and the effects of static bed height on the local flow structures are also investigated.     

Hydrodynamic characterisation of liquid–solid two–phase fluidised beds: Vibration signature and pressure fluctuations analyses

Hydrodynamic behaviour of a two–phase liquid–solid fluidised bed was investigated over a wide range of liquid velocities by means of simultaneous vibration and pressure fluctuations analyses. The liquid velocities were set in a way that covered two most important hydrodynamic events in the bed, namely minimum fluidisation and circulating‐solid regime. To prevent solids from being carried out of the bed, the maximum liquid velocity was kept lower than the terminal velocity of solids. Statistical analysis on the vibration signatures of bed shell proved to be a strong representative for minimum fluidisation characterisation and solid regime change. The minimum fluidisation velocity can be obtained from the intersection of two linear parts in the standard deviation of vibration fluctuation signals. Moreover, the kurtosis of vibration signals could predict the minimum fluidisation and approximate solid regime transition successfully. Meanwhile, statistical parameters, such as standard deviation, skewness, and kurtosis as well as newly‐introduced parameters, namely the energy and average cycle frequency of pressure signals, determined both of minimum fluidisation condition and circulating‐solid flow regime. © 2011 Canadian Society for Chemical Engineering     

Identification and characteristics of different flow regimes in a circulating fluidized bed

A series of experiments was conducted in a 0.3-m diameter circulating fluidized bed (CFB) cold model to evaluate the operating flow regimes and their transitions. A single unambiguous experimental method was developed to identify the transitions between CFB operating regimes. Experiments were conducted at riser gas velocities ranging from dense phase turbulent, through fast fluidization (S-shape riser pressure profile), and up to dilute-phase flow regimes. A transient method was applied to a low density, Geldart Type B, cork bed material. Two distinct transition velocities were found by analyzing the time required to empty out all solids from the riser of the CFB after cutting off solids flow. The lowest transition velocity marked the transition between the dense-phase turbulent and the fast fluidization flow regimes, while a higher or second transition represented the transition between the fast fluidization and the dilute-phase flow regimes. Based on the experimental results, the axial pressures and its fluctuations along the riser exhibited markedly distinct profiles in each of the three different operating flow regime regions as defined by these two transport velocities.     

SORPTION DRYING OF SOYBEAN SEEDS WITH SILICA GEL. I. HYDRODYNAMICS OF A FLUIDIZED BED DRYER

ABSTRACT

Fluidization technology has been employed to mix soybean seeds and silica gel particles in aim of sorption drying of particulate agricultural products. The characteristic fluidization velocities, mixing mechanisms and fluidization quality have been studied in a 180 mm I.D. fluidized bed. Two sizes of silica gel particles (SG2 and SG3) were selected, with the mass fractions in the range of 0.33–0.75 to form a static bed with height from 100 to 280 mm (H/D = 0.56–1.56). The results show that the dispersion rate of soybean seeds is increased with addition of either SG2 or SG3, and that the frequency and the span of pressure fluctuations within the bed are increased. It is inferred that the gas–solids contacting is improved with addition of small particles in the bed of large particles. In addition, for practical design and operation of such a fluidized bed of binary mixtures, empirical correlations for characteristic fluidization velocities were developed with a wide application.     

Hydrodynamics of an inverse liquid–solid circulating conventional fluidized bed

An inverse liquid–solid circulating conventional fluidized bed (I-CCFB) is realized by injecting particles from the top of a conventional liquid–solid fluidized bed (0.076 m ID and 5.4 m height) that is operated in a newly developed circulating conventional fluidization regime located between the conventional and circulating fluidization regimes. The I-CCFB can achieve a higher solids holdup compared to both conventional and circulating liquid–solid fluidized beds. A new parameter, the bed intensification factor, is defined to quantify the increased solids holdup observed with external solids circulation. The Richardson–Zaki equation is shown to be applicable to the I-CCFB regime and can be used to correlate the slip velocity and solids holdup, both of which increase with the solids circulation rate. A new flow regime map is presented, including the I-CCFB and a variety of other liquid–solid fluidized beds.     

Transient dynamics of solid concentration in downer fluidized bed

This paper presents new results from experiments performed in a 15 cm diameter down-flow fast-fluidized bed. Tests were conducted at room temperature and near atmospheric pressure, with 125 μm glass beads. Superficial gas velocities range from 0 to 6 m/s. A needle capacitance probe was used to obtain measurement of instantaneous solids concentration (ε_s) in a small sampling volume of approximately 0.1 cm³. With the downer operating at steady-state, the capacitance probe provided a time-trace of the local solid fraction, indicating that ε_s was generally less than 0.01. However the measurements indicated that the local concentration did have excursions to significantly higher concentrations for short durations. This behavior is indicative of clusters in the down-flow fast-fluidized bed. While clusters are known to be a dominant feature of upward fast fluidization in risers, their presence has been questioned for solid/gas down flows. Sample data are presented to indicate the density, duration, and time fraction of clusters observed in this downer fluidized bed. Preliminary comparisons are made to the limited data available for comparable operating conditions in a riser.     

Distributor effects near the bottom region of turbulent fluidized beds

The distributor plate effects on the hydrodynamic characteristics of turbulent fluidized beds are investigated by obtaining measurements of pressure and radial voidage profiles in a column diameter of 0.29 m with Group A particles using bubble bubble-cap or perforated plate distributors. Distributor pressure drop measurements between the two distributors are compared with the theoretical estimations while the influence of the mass inventory is studied. Comparison is established for the transition velocity from bubbling to turbulent regime, U_c, deduced from the pressure fluctuations in the bed using gauge pressure measurements. The effect of the distributor on the flow structure near the bottom region of the bed is studied using differential and gauge pressure transducers located at different axial positions along the bed. The radial voidage profile in the bed is also measured using optical fiber probes, which provide local measurements of the voidage at different heights above the distributor. The distributor plate has a significant effect on the bed hydrodynamics. Owing to the jetting caused by the perforated plate distributor, earlier onset of the transition to the turbulent fluidization flow regime was observed. Moreover, increased carry over for the perforated plate compared with the bubble caps has been confirmed. The results have highlighted the influence of the distributor plate on the fluidized bed hydrodynamics which has consequences in terms of comparing experimental and simulation results between different distributor plates.     

湍动流化床过渡段固含率分布特征的实验及数值模拟

在湍动流化床中,过渡段对于包括甲醇制烯烃在内的气固催化快反应有着重要的作用。采用PV6D反射型光纤探针对内径95 mm的湍动流化床内过渡段的固含率分布和脉动参数进行了测量,分别考察了表观气速和静床高的影响,并采用修正的基于颗粒动力学的三段曳力双流体模型进行模拟。实验表明,湍动流化床过渡段中固含率的轴向分布呈现S型和指数型两种类型,固含率轴向与径向分布都在过渡段内出现最大梯度,表明过渡段中固体浓度分布比稀相段和密相段更不均匀。表观气速和静床高的变化将导致S型和指数型分布的相互转变,并且对过渡段底部与壁面附近的固体高浓度区影响最为显著。局部固含率脉动概率密度分布表明,在静床高较小时,随着气速的增大,床层下部气含率最大值位置将从中心区移动至环隙区,呈现气含率的双峰型分布。本文提出的修正三段曳力模型考虑了颗粒团聚的影响,对过渡段中分布板影响区之外的固含率分布均能较好地模拟。     

Hydrodynamics of three-phase fluidized bed systems examined by statistical, fractal, chaos and wavelet analysis methods

Hydrodynamic studies were conducted in gas-liquid-solid systems (0.1 m ID, 2 m high) of 3.0 mm glass beads and of 2.1 mm polypropylene low-density particles, with particles densities of 2471 and 1290 kg/m³, respectively. Simultaneous measurement of differential pressure and bubble conductivity probe signals sampled at 500 Hz for 60 s enabled the investigation of the change in flow structure in relation to the flow regime transitions. Superficial gas velocities ranged between 0.010 and 0.052 m/s for polypropylene particles, and extended to 0.12 m/s for glass beads, while the superficial liquid velocities covered the ranges of 0.0007-0.045 m/s for polypropylene particles, and ranged up to 0.056 m/s for glass beads.Spectral analysis of the pressure fluctuations revealed a transition from dispersed to coalesced bubbling flow with decreasing liquid velocity for a given superficial gas velocity. The use of a conductivity probe facilitated characterization of the local flow structure in terms of bubble movement. The measurements were extensively analyzed using fractals and chaos, power spectra frequency analysis and wavelet decomposition in addition to the standard statistical analyses. The coefficient of variation of the bubble probe signals was found to be the most effective in deducing the transition velocity between coalesced and dispersed bubbling flow regimes, while wavelet energy confirmed the similarity in the distribution between two axial positions once operated in the dispersed flow regime. Comparison of the flow structure between glass beads and polypropylene particles showed that both the minimum liquid fluidization velocity and the transition velocity between the bubble flow regimes were much higher for the glass beads than for the lighter polypropylene particles. Furthermore, the standard deviations of the decomposed bubble probe signals through wavelet transformation successfully highlighted the difference between the two systems of particles.     

Tomographic imaging of a conical fluidized bed of dry pharmaceutical granule

Hydrodynamics in a conical fluidized bed were studied using electrical capacitance tomography (ECT) for a bimodal and mono-disperse particle size distribution (PSD) of dry pharmaceutical granule. The bimodal PSD exhibited a continuous distribution with modes at 168 and 1288 μm and contained approximately 46% Geldart A, 32% Geldart B and 22% Geldart D particles by mass. The mono-disperse PSD had a mean particle size of 237 μm and contained approximately 71% Geldart A, 27% Geldart B, and 2% Geldart C particles by mass. The granule particle density was 830 kg/m³. Experiments were conducted at a static bed height of 0.16 m for gas superficial velocities ranging from 0.25 to 2.50 m/s for the mono-disperse PSD, and from 0.50 to 3.00 m/s for the bimodal PSD. These gas velocities covered both the bubbling and turbulent fluidization regimes. An ‘M’-shaped time-averaged radial voidage profile appeared upon transition from bubbling to turbulent fluidization. The ‘M’-shaped voidage profile was characterized by a dense region near the wall of the fluidized bed with decreasing solids concentration towards the centre. An increased solids concentration was observed in the middle of the bed. Frame-by-frame analysis of the images showed two predominant bubble types: spherical bubbles with particle penetration in the nose which created a core of particles that extended into, but not through, the bubble; and spherical bubbles. Penetrated bubbles, responsible for the ‘M’ profile, were a precursor to bubble splitting; which became increasingly prevalent in the turbulent regime.     

Dimension measurements of hydrodynamic attractors in circulating fluidized beds

Experimental measurements of the correlation dimension, Kolmogorov entropy, and Lyapunov exponent of circulating fluidized bed (CFB) chaotic attractors were obtained by recording differential pressure and γ-ray porosity time series along the height of a cold experimental CFB operating with 75 μm diameter fluid catalytic cracking (FCC) catalyst. The attractor dimension did not vary with respect to the type of measurement taken. Both differential pressure and localized γ-ray densitometry measurements showed the existence of a low order hydrodynamic attractor, whose dimension varied between 1.5 and 2.0 over the range of gas velocities and solids fluxes studied. Differential pressure measurements indicated that the attractor dimension decreased slightly in the lower section of the CFB and at higher solids fluxes. Localized radial γ-ray bed porosity measurements indicate that the attractor dimension did not significantly vary across the bed cross-section but did show a tendency to be slightly lower near the riser wall.     

EFFECT OF PARTICLE SIZE DISTRIBUTION ON LOCAL VOIDAGE IN A BENCH-SCALE CONICAL FLUIDIZED BED DRYER

ABSTRACT

The effect of particle size distribution (PSD) on local voidage has been investigated in a conical fluidized bed containing dried placebo pharmaceutical granule. For each of the five PSDs examined, the static bed height was varied between 0.12 and 0.17 m and the superficial gas velocity was varied between 0.05 and 0.75 m/s. The local voidage was measured using a twin-plane electrical capacitance tomography (ECT) system. A wide PSD containing 12 wt% solids with a diameter of 2 mm or larger resulted in two different types of gas flow: an annular gas flow up to a gas velocity of 0.50 m/s and a centrally concentrated gas flow above 0.50 m/s. The mixtures containing less coarse material exhibited a centrally concentrated gas flow surrounded by a dense phase at the walls of the bed over the entire range of gas velocities and bed heights examined. Consideration of previous work by other researchers suggests that the behavior of the wide PSD mixture is due to segregation at the lower velocities. The local voidage was sensitive to small changes in static bed height. For the wide PSD mixture at a fixed gas velocity, the gas tended to spread more uniformly over the bed cross-section as static bed height increased. The opposite was true of the other mixtures, i.e., the gas flow became more centralized with increasing bed height.     

CHARACTERIZATION OF AXIAL NONUNIFORMITY IN FLUIDIZED BED BY THE AMPLITUDE OF LOCAL PRESSURE DROP FLUCTUATIONS

The amplitude of local pressure drop fluctuations was used to characterize the axial stability of a freely bubbling fluidized bed. The experiments were carried out in a bed of 0.318 m dia at normal and elevated temperatures. It was found that the transition between a freely bubbling zone, consisting of small bubbles distributed uniformly across the bed, and a zone of incipient slugging occurs if bubble to bed diameter ratio is about 0.15. The amplitude of local pressure drop fluctuations was found to be a linear function of gas velocity in a broad range of temperatures if measured in the freely bubbling zone     

EFFECT OF PARTICLE SIZE DISTRIBUTION ON LOCAL VOIDAGE IN A BENCH-SCALE CONICAL FLUIDIZED BED DRYER

The effect of particle size distribution (PSD) on local voidage has been investigated in a conical fluidized bed containing dried placebo pharmaceutical granule. For each of the five PSDs examined, the static bed height was varied between 0.12 and 0.17 m and the superficial gas velocity was varied between 0.05 and 0.75 m/s. The local voidage was measured using a twin-plane electrical capacitance tomography (ECT) system. A wide PSD containing 12 wt% solids with a diameter of 2 mm or larger resulted in two different types of gas flow: an annular gas flow up to a gas velocity of 0.50 m/s and a centrally concentrated gas flow above 0.50 m/s. The mixtures containing less coarse material exhibited a centrally concentrated gas flow surrounded by a dense phase at the walls of the bed over the entire range of gas velocities and bed heights examined. Consideration of previous work by other researchers suggests that the behavior of the wide PSD mixture is due to segregation at the lower velocities. The local voidage was sensitive to small changes in static bed height. For the wide PSD mixture at a fixed gas velocity, the gas tended to spread more uniformly over the bed cross-section as static bed height increased. The opposite was true of the other mixtures, i.e., the gas flow became more centralized with increasing bed height.     

Hydrodynamics of the Stripper Section of Fluid Cokers

The hydrodynamics of fluid bed cokers were studied by analyzing pressure fluctuations and particle motion in a half‐column cold model, geometrically and dynamically similar (with key dimensionless groups matched) to the stripper section of two commercial fluid cokers. Superficial gas velocity and solids circulation strongly affected the hydrodynamics. The pressure drop over the top section of the stripper decreased at high solids circulation fluxes and high gas velocities due to flooding. Flooding occurred prematurely when fouling was simulated. Steam redistribution did not improve stripper performance for the conditions investigated. However, steep sheds on the top row, aeration behind the solids exit and standpipe aeration all improved solids circulation, leading to reduced fouling in two commercial fluid cokers.     

The axial hydrodynamic behavior in a liquid-solid circulating fluidized bed

Based on experimental results from a 7.6 cm I. D. and 3 m high liquid–solid circulating fluidized bed, the liquid–solid circulating fluidization regime has been separated into two zones: the initial circulating fluidization zone and the fully developed circulating fluidization zone. The distinct hydrodynamic behavior and the influence of particle properties in the two circulating fluidization zones have been studied. The overall flow structure in LSCFB, although still somewhat non–uniform, is much more uniform than that in a gas-solid CFB. Our experimental results also show that the axial flow characteristics and the regime transition can be strongly affected by the particle density. The stable operation range of the circulating fluidization system and the influences of some associated factors, such as the solids inventory and the particle density, were also investigated for the first time.     

STOCHASTIC BEHAVIOR OF FLUIDIZED PARTICLES IN A LIQUID-SOLID FLUIDIZED BED

In a liquid-solid fluidized bed, the apparently irregular or stochastic behavior of particles gives rise to various flow regimes depending on parameters such as the particle size, liquid flow rate, static bed height and axial position in the bed. It is highly plausible that this irregular behavior manifests itself as pressure fluctuations; thus, the effects of these parameters on the particle behavior or the particle flow regime were investigated through measurement and spectral analysis of the pressure fluctuations. The results indicate that the amplitude of pressure fluctuations exhibits a maximum and that the decay constant in the autocorrelation function attains its minimum at the intermediate liquid flow rate where the particle flow regime undergoes transition from the cluster circulation to the individual quasi random motion. The model composed of the periodic and stochastic components of pressure fluctuations, is in good accord with the experimental results in terms of both the autocorrelation and power spectral density functions.