Expanded Bed Height Research Articles

The effect of ring baffles on the hydrodynamics of a gas—solid bubbling fluidized bed using computational fluid dynamics

An Eulerian—Eulerian two-fluid model (TFM) integrating the kinetic theory for emulsion phase was used to simulate gas—solid fluidized beds. Validation of the model was investigated based on hydrodynamic parameters such as bed expansion ratio, H/ H0, gas volume fraction profile, bubble behaviour, and motion of the particles. A good agreement was found between numerical results and experimental values. The model was used to study a bubbling fluidized bed (BFB) including the ring baffles. Predicted results show that the ring baffles have an important role in the flow pattern of the bed. Baffles increase the bed expansion height and particle velocities at axial locations on the top of the highest baffle as well as uniform distribution of gas volume fraction between the baffles area. In spite of increasing the dead zones in the bed, ring baffles cause the improvement of mixing and heat transfer in the bed. The present study provides a useful basis for further works on the effect of baffles in BFBs.

Catalytic Synthesis of Glycerol Monoacetate Using a Continuous Expanded Bed Column Reactor Packed with Cation-Exchange Resin

The selective and continuous production of monoacetin using an expanded-bed column reactor, in which the reactive fluid flows vertically upward and the catalyst is hydrodynamically floated, was studied. Acetic acid and glycerol were used as the substrates and the proton-type cation-exchange resin Amberlyst 16 was used as the catalyst. Batch reaction experiments under various conditions were first conducted to investigate the details of the heterogeneous catalytic reaction mechanism. The esterification of acetic acid with glycerol to produce the target product monoacetin was proved to proceed predominantly over the undesired consecutive esterification reactions to diacetin and triacetin. The intraparticle and extraparticle mass-transfer resistances for the catalytic resin were negligible. The Eley−Rideal-type catalytic reaction model was successfully applied to the experimental data for the various conditions. Next, the hydrodynamic correlation between the flow rate of the reactive fluid and the corresponding expanded-bed height on the continuous reactor was determined experimentally. Furthermore, the reaction and transport model for a continuous expanded-bed column reactor was constructed based on the kinetic model for the batch system. The numerical calculation simulated the experimental data well, and the optimum operating conditions were successfully determined by the theoretical analysis.

Expanded bed adsorption of an alkaline lipase from Pseudomona cepacia

An extracellular lipase was isolated from Pseudomona cepacia by expanded bed adsorption on an Amberlite 410 ion-exchange resin. Enzyme characterization and hydrodynamic study of a chromatography column were done. Enzyme purification was done at three condition of expanded bed height ( H): at one and half (6 cm), at two (8 cm) and at three (12 cm) times the fixed bed height ( H 0 = 4 cm). The results showed that the experimental data was fitted to the Richardson and Zaki equation, and the comparison between the experimental and calculated terminal velocities showed low relative error. In enzyme purification for better condition, a purification factor of about 80 times was found at 6 cm of expanded bed height, or 1.5 times of expansion degree. Purified lipase had an optimal pH and a temperature of 8 and 37 °C, respectively.

Expanded bed height determination in a tapered fluidised bed reactor

Fluidised bed coating is a commonly applied technique to microencapsulate food powder ingredients. Particularly for the food industry, the use of a top-spray configuration in a tapered reactor is considered to be most appropriate. Good particle-droplet contacting can be established by controlling the air flow rate and by adjusting the distance between the nozzle and the particle bed. In this paper, the determination of bed height in a tapered expansion chamber of fluidised bed equipment was investigated using static pressure measurements along the wall. As an alternative to this traditional approach, the use of wall surface temperature measurements was proposed as a non-invasive routine method. The experiments were performed in a Glatt GPCG-1 top-spray fluidised bed unit in which a polycarbonate expansion chamber was mounted. This allowed comparison of the results of the pressure and temperature measurements with visual observations.

Effect of solid hold-up on nitrite accumulation in a biofilm reactor - molecular characterization of nitrifying communities

Biological ammonium oxidation was carried out in two inverse turbulent bed reactors fed with synthetic mineral wastewater containing a high ammonium concentration (100 mg N-NH4+/L). Both reactors were started-up and operated in the same conditions except for the solid carrier concentration: the solid hold-up ratios applied, defined as the ratios of static to expanded bed height, were 0.1 and 0.3 in reactors R10 and R30 respectively. These two solid hold-up ratios generate different particle-to-particle collision frequencies and, therefore, detachment forces. The influence of solid hold-up on biofilm growth and nitrifying performance was studied from a macroscopic (i.e. nitrate and/or nitrite production) and microbiological point of view. After 60 days of operation, both reactors contained the same amount of biomass. However, R10 produced only nitrate while nitrite accumulated in R30. A comparison of microbial populations in the reactors showed that R10 contained both ammonium and nitrite oxidizing populations such as Nitrosomonas and Nitrospira, whereas in R30, ammonium oxidizing populations were much greater than those of nitrite oxidizers. The major ammonium-oxidizing organism was not the same in both reactors.

Modeling of a Bubbling AFBC with Volatiles Release

A comprehensive model for continuous combustion of lignite with wide size distribution burning in its own ash in an atmospheric bubbling fluidized bed combustor (ABFBC) has been presented and used to correlate the data from a 0.3 MW ABFBC test rig. The model consists of submodels for hydrodynamics, volatiles release and combustion, char combustion, particle size distribution, entrainment and elutriation and is based on conservation equations for energy and chemical species. The volatiles release model is based on a particle movement model for estimation of portion of volatiles released in bed and a devolatilization kinetics model described by the distributed activation energy model for determination of time-resolved devolatilization profile of lignite particles. Uniform in-bed volatiles release is assumed. The overall model was applied to a 0.3 MW ABFBC test rig fired with lignite with VCM/FC ratio of 2.16. The fuel was fed 0.22 m above the distributor plate and the expanded bed height was 1 m. Predictions of the model were compared with measured concentration and temperature profiles and good agreement was obtained. Bed/freeboard combustion split predicted by the model was found to be 80/20 as opposed to the value 84/16 determined from measured species profiles.

Particle size and density distributions of two dense matrices in an expanded bed system.

The size and density distributions of two commercial media, that is, Streamline particles and 6% agarose coated steel beads (6AS), in an expanded bed system has been studied with a glass column (26 mm I.D.) modified by side ports. The Streamline particles have a broad size distribution but a relatively uniform density, while the 6AS beads have both broad size and density distributions. The effect of liquid-phase flow velocity, liquid viscosity and settled bed height on the particle size and density distributions is investigated. It is found that the radial mean particle size and density of the two matrices are uniform, while axial classifications are obvious in the expanded beds. For the Streamline, the volume-weighted mean particle size decreases linearly with increasing expanded bed height. For the 6AS beads, however, the mean particle size is even in the axial direction, but the particle density decreases exponentially with the increase of bed height. Moreover, the mean particle size of the Streamline or the density of the 6AS beads is well expressed as a function of the normalized bed height (that is, the ratio of the distance from bed bottom to the expanded bed height). The liquid flow-rate, liquid viscosity and settled bed height influence the mean axial size or density distribution by affecting the expanded bed height.

Behavior of particles and bubbles around immersed tubes in a fluidized bed at high temperature and pressure: a DEM simulation

Numerical simulations were carried out to investigate the behavior of particles and bubbles around immersed tubes in a two-dimensional fluidized bed at elevated pressures. Bed temperature was kept constant at 850°C and bed pressure was varied from 0.1 to 1.2 MPa. The particle–tube impact velocity and impact angle, the number of particle–tube impacts and the bed voidage around tubes in a staggered tube bank were calculated. It was found that the bed expansion height increased with the increasing pressure both at constant u 0/ u mf and u 0− u mf. Under elevated pressure, 1.2 MPa, the bubble frequency was about 4–8 Hz which was slightly higher than 2–4 Hz for the ambient condition, and the time-averaged velocity of particle–tube impacts at tube bottom area was also higher than that under the ambient pressure. The average particle–tube impact velocity showed its maximum at 10–70° from the tube bottom. The time average of bed voidage around a tube increased with the increasing pressure. The distributions of bed voidage and particle–tube impact velocity around a tube were asymmetrical, which may be due to the local bubble habits. The tube erosion rate was predicted to be 7.9–19.4 μm/1000 h and the maximum erosion occurred at ambient pressure because of the large amount of particle–tube impacts.

Estimation of bed expansions in a freely-bubbling three-dimensional gas-fluidized bed

Knowledge of bed expansions is important in the design and operation of gas-solid fluidized beds. This paper presents a study on the estimation of expanded bed height in a large three-dimensional gas-fluidized bed with a square section of 0.61 x 0.61 m 2 . All experiments were performed at the freely bubbling mode and the bed expansions were recorded by a video camera. Bed materials were used 593 μm raw perlite and 1233 μm sand falling within the categories of Geldart's Groups B and D, respectively. The bed height at minimum fluidization ranged from 0.0398 to 0.3176 m, while the excess air velocity from 0.034 ms -1 to 0.7453 m s -1 . Equations related to the bed expansion were given using a modified form of two-phase theory of fluidization. A correlation for the average bed expansion (void fraction) was also presented that has been derived from the principal form found successful in gas-liquid systems as follows: R = 0.5482 d p -0.129 (U o - mf ) 0.111 with an average deviation of less than 1%. The experimental findings were compared with previously reported results and were discussed in the light of available correlations. © 1998 John Wilev & Sons, Ltd.

A Model for Particle Collection in a Turbulent Bed Contactor—New Packings

This work presents an investigation of novel packings for collecting particles in a turbulent bed contactor. Perforated spheres, oblate spheroids and the conventional plain spheres, all made of polypropylene material, were tested for collecting particles in a 0.264 m diameter column. Alumina (aerodynamic diameter from 1.5 to 5.5 μm) and itanium dioxide (aerodynamic diameter = 0.7 μm) were used as test powders. The results show a lower penetration of particles through the fluidized bed of perforated packings in comparison to plain conventional spheres and oblate spheroids. This indicates that perforated packings are suitable as a contacting medium for particle collection in a turbulent bed contactor. A semi-empirical model for predicting particle collection on the gas-liquid interfacial area formed by the packings is presented in this work. The model describes particle penetration through the bed as an exponential function of the expanded bed height (H) and a global coefficient of particle transfer (m), similar to the global molecular mass transfer coefficient. The coefficient (m) was fitted as function of dimensionless numbers, which describe the hydrodynamics of phases contacting, bed structure and the mechanisms involved in particle collection on a TBC. A statistical analysis of the model predictions has confirmed it to be satisfactory for all types of packings investigated.

Segregation phenomena in a shaker

The segregation phenomena of different binary mixtures were studied experimentally in a vertical shaker. The larger particles tend to move upwards in the shaker in a certain range of vibrational accelerations. The segregation effect was more significant for the mixture with a larger size difference. The segregation coefficient was measured and found to increase with vibrational acceleration until reaching a maximum and then to decrease with acceleration. The bed expansion height was also measured. It was found that the greatest segregation effect occurred when the bed transformed from a dense state to a loose state.

A unified approach to the scale-up of gas—solid fluidized bed and gas—liquid bubble column reactors

The aim of this work is to develop a unified approach to the scale-up of gas—solid (G—S) fluid bed and gas—liquid (G—L) bubble column reactors. The unified approach relies on analogies in the hydrodynamic behavior of G—S and G—L systems in both the homogeneous and heterogeneous flow regimes. The homogeneous G—S fluidization regime is to be identified with homogeneous G—L bubbly flow while the heterogeneous G—S fluidization regime is to be identified with the churnturbulent regime for G—L bubble columns. In the heterogeneous flow regime of operation, the classic two-phase theory developed for G—S fluidized beds can be applied with profit to describe also the hydrodynamics of G—L bubble columns provided the “dilute” phase is identified with the fast-rising large bubbles, and the “dense” phase is identified with the liquid phase containing entrained “small” bubbles. Quantitative analogies in the hydrodynamic behavior of G—S and G—L systems are demonstrated by the use of extensive experimental data obtained in five columns of three different diameters (2 × 0.1 m, 1 × 0.19 m and 2 × 0.38 m). The total expanded bed height in the experiments varied in the range 0.5–3.5 m. About 4,000 dynamic gas disengagement experiments were carried out with various systems to determine the total gas hold-up and the hold-ups of the “dilute” and “dense” phases. The gas phases used in the experiments were air, helium, argon and sulfur hexafluoride. Fluidized cracking catalyst (FCC) was used as the solid phase in fluid bed experiments. In bubble column operations the liquid phase used was water, paraffin oil or tetradecane. Sintered plates were used for gas distribution in all the columns. The gas hold-up in the “dense” phase, ɛ df was found to be practically independent of the scale of operation. The hold-up of the fast-rising “dilute” phase, ɛ b , on the other hand was found to be a significant function of the column diameter, D T , and of the total dispersion height, H. The “dilute” phase gas hold-up can be modeled for both G—S and G—L systems using a theory allowing for bubble growth in the region above distributor. The bubbles are assumed to grow in diameter up to a distance h *, at which the bubbles reach their equilibrium size. The equilibration height h * was found to increase with the superficial gas velocity through the dilute phase, U — U df ). For air—FCC, the value of h * varies in the range 0.4–1.2 m. For bubble columns the values of h * are significantly smaller and lie in the range 0–0.5 m. Increasing gas density increases the total gas voidage in both G—S and G—L systems but has no significant effect on the hold-up of the dilute phase. In G—L bubble columns, the liquid properties affect the total gas hold-up but have only a minor influence on the dilute phase hold-up. The unified model to describe the bubble hydrodynamics in G—S fluid beds and G—L bubble columns is a useful tool in scaling-up these two reactor types, because of the possibilities of cross-fertilization of design data.

Applicability of various averaging concepts to predict the average dynamic properties of poly-dispersed liquid-solid fluidized bed systems

To avoid the necessity of having a large number of correlations to predict the dynamic properties of poly-dispersed fluidized bed systems, a statistical approach has been adopted in the present paper to get an expression which can predict such property from a knowledge of similar property of the individual species that constitute such heterogeneous bed. In the present paper, the dynamic property means any property that is associated with, or dependent on a term “velocity”, e.g., beginning elutriation velocity, u be; free settling velocity, u t; onset semifluidization velocity, u osf; expanded bed height, L e, etc. Based on 10,517 computations, embracing 5 different dynamic properties of liquidsolid fluidized bed systems and covering 13 different averaging formulae — similar to those, used to calculate average particle size for a poly-dispersed system — the following expression has been found to give the best fit: P=[ 1 σ ( X i P 3 i )] 1 3

Prediction of axial concentration profiles in catalytic fluidized reactors

AbstractA previous study by Fryer and Potter (1976) of the catalytic decomposition of ozone in a bubbling fluidized‐bed reactor is considered in the light of a recently developed reactor model. It is shown that since in‐bed sampling leads to an averaged concentration that is biased toward the emulsion‐phase gas composition, an increase in the ozone concentration in the sampled gas is predicted as the probe is raised above the bubbling bed. The analysis, which does not have to assume gas backmixing, shows that the minima in the ozone concentration profiles associated with this transition occurs between Hmf and H, the expanded bed height, and may even occur at lower points in the bed at higher excess fluidizing velocities. The model also correctly predicts the overall conversion of ozone as a function of the reaction rate constant and gas velocity without resorting to the fitting of any parameters.

An improved relationship to predict expanded bed height in particulate fluidization

To predict the bed expansion in liquid-solid fluidization, which is typically particulate in nature, the earlier equation, proposed by the author, is a polynomial one of the following form: 1 R = 1 − 1.762U + 0.95U 2 . The proposed equation, unfortunately, suffers a few drawbacks, the most serious of which is that it does not satisfy one of the essential conditions of bed expansion, i.e. at U = 1 (or u f = u t), 1 R = 0 . To eradicate such defect, an alternative method is suggested to predict directly the expanded bed height of poly-dispersed systems as: L e = 1 Σ alli ( x i L ei ) . The equation is found to cause an overall improvement of 16.7% from the previous one, based on 646 bed expansion data of 49 multicomponent fluidized bed systems.

Reduction of NO emissions from a coal-fired fluidized bed combustor by addition of NH 3 through a water-cooled injector

To reduce NO emissions, ammonia was injected into a coal-fired fluidized bed combustor having a diameter and expanded bed height of 30 cm and a freeboard height of 100 cm. The injector consisted of a water-cooled stainless steel tube that could be moved axially to adjust the injection height. Experiments were performed to investigate the optimal position for NH 3 addition, and the influence of coal type and operating conditions, such as the fluidization number ( u u mf ), the NH 3:NO molar ratio β, the excess air level and the temperature. A distinct optimal position in the axial coordinate was always found, and depends on the fluidization number and coal type. The amount of NO reduction increases with β, with maximal reduction levels of 70% at β = 2 and a temperature of 850°C. The amount of excess air was also found to have an important influence on the NO reduction: optimal results were obtained at about 5% O 2 in the flue gas. The rather low temperature level at which the reaction proceeds, and the weak temperature dependence seem to indicate that the NONH 3O 2 reaction is influenced by heterogeneous reaction steps in the environment of the fluidized bed combustor.

Chemical/Bio‐fluidized Bed Treatment of Slaughterhouse Wastewater

A combined chemical coagulation/flocculation-FBBR process was evaluated for the treatment of an equalized slaughterhouse wastewater. Results obtained from jar and tank tests indicated that alum, with supplemental use of ferric chloride and an anionic polymer, yielded the following removal performance: BOD\d5, 24–53%; grease, 66–80%; and SS, 62–73%. Furthermore, the performance of chemical coagulation/flocculation process at constant chemical dosages improved with increasing feed BOD\d5, grease, and SS concentrations. The steady-state performance of the oxic FBBR inn treating the chemically pre-treated slaughterhouse wastewater was evaluated under the following conditions: F/M ratio, 1.45–9.25 kg/kg-day; empty bed hydraulic retention time, 8.8–30.8 minutes; and feed BOD\d5 concentration, 305–602 mg/L. The observed BOD\d5, grease, and NH\d3-N removals were 71–94%, 29–84%, and 20–73%, respectively. More than 70% of feed BOD\d5, grease, and NH\d3-N could be removed up to an F/M ratio of 2.0 kg/kg-day. The FBBR would become anaerobic when the F/M ratio applied exceeded 4.0 kg/kg-day and its performance deteriorated sharply. A relatively high and constant biomass holdup (i.e., 10,000 mg TVS/L) could be maintained in the oxic FBBR over the range of F/M ratios applied. This was achieved through regular bioparticle wasting designed for expanded bed height control. Secondary treatment performance was achieved in this investigation at an F/M ratio of 2.0 kg/kg-day without secondary clarification.

Application of a gamma-radiation density gauge for determining hydrodynamic properties of fluidized beds

A modified, off-the-shelf, gamma-radiation density gauge has been interfaced into a microprocessor and mounted to a movable assembly mechanism which allows fluctuating bed density to be monitored on-line at any axial height in a fluidized bed. The signal processing required to determine expanded bed height L f , dense phase voidage ε D , dense phase superficial gas velocity u Do , cross-sectional centerline bubble phase volume fraction 〈δ B 〉 t c , bubble size d b , frequency f b , is described in detail. Example measurements are presented as made on both a cast-acrylic model column and an industrial, pilot-scale, high-pressure, fluidized-bed reactor. Accuracy of the measurements is discussed.

Characteristics of fluidization at high pressure

Experiments were conducted to investigate fluidization fundamentals at pressures up to 6485 kPa using nitrogen as the fluidizing gas. The particles under study were coal, char and Ballotini. Both a three-dimensional bed (10.16-cm-i.d.) and a two-dimensional bed (1.9x 10.16cm) were used in the experiments. The fundamentals of high pressure fluidization examined in this study include minimum fluidization velocity, bed voidage at minimum fluidization, bed expansion, and bubbling behavior. An empirical correlation was developed for determining minimum fluidization velocity. The effects of pressure upon bed voidage at minimum fluidization and expanded bed height were analyzed for several types of particles. High speed photographs were studied to describe bubbling behavior in a fluidized bed over a range of pressures.

Bed expansion in a mobile‐bed contactor

AbstractThe expanded bed height has been investigated in a 0.14‐m diameter mobile‐bed gas‐liquid contactor. The data obtained by two methods of measurement show that the bed height varies linearly with the gas flow rate and is independent of the packing density. The operation of the bed is characterized by the condition of flooding. Correlations are presented for the expanded bed height and the minimum fluidization velocity in terms of the independent system variables.