• Energy Research

AbstractFlow maldistribution was investigated in an 8.5 m high circulating fluidized bed with a rectangular cross section of 1 × 0.3 m using a 14‐bubble‐cap distributor and a 33‐bubble‐cap distributor. Experimental results indicate that the bubble caps in the center region have a large flow rate and the bubble caps at the wall region have a small flow rate. The extent of the flow maldistribution is quantitatively characterized by the ratio of the maximum flow rate through a bubble cap to the average flow rate through all bubble caps (maldistribution number, Vmax/Vav). A correlation is proposed for predicting flow rate distribution, maldistribution number, in the bubble cap distributor by the ratio of distributor pressure drop to riser pressure drop (pressure‐drop ratio, ΔPd/ΔPr). © 2005 American Institute of Chemical Engineers AIChE J, 2005

Abstract The dynamic flowsheet simulation of solids processes is an area of increasing interest in recent years. Compared to the well-established flowsheet modelling of liquid-gas systems, the modelling of granular materials requires different approaches, strategies and algorithms. Therefore the new dynamic flowsheet simulation framework Dyssol has been developed within the Priority Program SPP 1679 “Dynamic simulation of interconnected solids processes (DYNSIM-FP)” of the German Research Foundation (DFG). In this contribution the architecture of the novel simulation framework and computational methods employed in it are presented. The system is based on the sequential-modular approach supplemented with partitioning and tearing methods. Waveform relaxation method, as well as several convergence methods and data extrapolation algorithms have been implemented to improve system performance and to increase convergence rate. To perform a correct calculation of multidimensional distributed parameters an approach with transformation matrices is used in the Dyssol system. Simulation case studies calculated with new system have shown good stability, convergence rate and agreement of simulation results with test systems and experimental results.

Abstract In the current work, a model of the fluid mechanics in the riser of a circulating fluidized bed (CFB) has been implemented using computational fluid dynamics (CFD). The model developed shall be used in future as the basis of 3D-reactor model for the simulation of large scale CFB combustors. The two-fluid model (TFM) approach is used to represent the fluid mechanics involved in the flow. The computational implementation is accomplished by the commercial software FLUENT. Different closure formulations are tested on a simplified geometry. Two different turbulence formulations, namely the swirl modified RNG k–ɛ model and the Realizable k–ɛ model, are tested in combination with two different approaches to solid phase turbulence, namely the dispersion and per phase approach. One focus of the current work is put on the study of different drag correlations. Besides the drag correlations by Syamlal et al. [Syamlal, M., Rogers, W., & O’Brien, T. J. (1993). MFIX documentation theory guide. Technical Report DOE/METC-94/1004, U.S. Department of Energy (DOE). Morgantown Energy Technology Center: Morgantown, WV] and Gidaspow [Gidaspow, D. (1994). Multiphase flow and fluidization. New York: Academic Press] the EMMS model has been used to determine the momentum exchange between the two phases. The resulting formulation is then used to simulate a 1-m × 0.3-m cold CFB setup and is validated by experimental results [Schlichtharle, P. (2000). Fluid dynamics and mixing of solids and gas in the bottom zone of circulating fluidized beds. Unpublished doctoral dissertation, Technische Universitaet Hamburg-Harburg, Shaker Verlag: Aachen].

Abstract Combustion mechanisms of two types of coffee husks have been studied using single particle combustion techniques as well as combustion in a pilot-scale fluidised bed facility (FBC), 150 mm in diameter and 9 m high. Through measurements of weight-loss and particle temperatures, the processes of drying, devolatilisation and combustion of coffee husks were studied. Axial temperature profiles in the FBC were also measured during stationary combustion conditions to analyse the location of volatile release and combustion as a function of fuel feeding mode. Finally the problems of ash sintering were analysed. The results showed that devolatilisation of coffee husks (65–72% volatile matter, raw mass) starts at a low temperature range of 170–200°C and takes place rapidly. During fuel feeding using a non water-cooled system, pyrolysis of the husks took place in the feeder tube leading to blockage and non-uniform fuel flow. Measurements of axial temperature profiles showed that during under-bed feeding, the bed and freeboard temperatures were more or less the same, whereas for over-bed feeding, freeboard temperatures were much higher, indicating significant combustion of the volatiles in the freeboard. A major problem observed during the combustion of coffee husks was ash sintering and bed agglomeration. This is due to the low melting temperature of the ash, which is attributed to the high contents of K2O (36–38%) of the coffee husks.

Abstract Based on the assumption of a core-annulus structure, radial gas mixing characteristics were experimentally studied in the core of the upper dilute zone of a pilot scale circulating fluidized bed (CFB), 0.4 m in diameter and 9 m in height. C0 2 as a tracer was injected continuously into the center of the bed by a point source. Radial tracer gas concentration profiles were measured in several planes downstream of the injection point. An analytical solution derived by Klinkenberg et al. ( Ind. Eng. Chem., 6 (1953) 1202) [1] for the description of gas mixing in turbulent single phase flow was successfully applied to determine the Peclet number, Pe r,c , for radial gas mixing in the core zone. Pe r,c was found to be independent of the superficial gas velocity u , which is in agreement with the gas mixing behaviour in turbulent single phase flow. A mean value of Pe r,c =465, which is also in agreement with other authors' measurements in single phase flow, was calculated from measurements at different solids rates G s , up to 70 kg m −2 s −1 . The fact that Pe r,c was found to be independent of G s is attributed to the low mean solids concentration in the core zone and to the size of the particles, which is too small to cause a significant influence on the turbulent intensity of the gas phase.

The Zhundong coalfield in Xinjiang, China, is the largest integrated coal basin newly found. The present work concentrates on the application of chemical looping combustion (CLC) with a Zhundong lignite, which is characterized by high sodium content. Some experiments in a laboratory scale fluidized bed facility with an active iron ore oxygen carrier, were performed using the lignite as fuel and CO2 as gasifying agent at a temperature of 900 °C, with the objective of investigating its combustion performance and sodium transfer behavior in CLC. Results indicate that the gasification reactivity of the three coals follows the order of German lignite > Zhundong lignite > American bituminous coal in the current experimental conditions. During reducing stage, the unique product of sodium transfer from coal to the fly ash is albite (NaAlSi3O8) due to the reactions between sodium and other coal ash. The sodium deposition on the oxygen carrier particles was not found. 40 reducing-oxidizing cycles were performed, and sodium accumulation in the bed materials with cycles was found due to some ash staying in the bed. However, the growth of bed particles due to the sodium accumulation was not observed by determining the particle size distributions of bed materials. This indicates that burning the high sodium Zhundong coal in the present conditions have no influence on the particle agglomeration. Finally, a literature survey was made and results indicate that the main sodium in the Xinjiang coal basin of China is water soluble with an average value of 64%. The pure salt of NaCl, as one common water soluble sodium phase in Zhundong coals, was introduced to a bed of iron ore particles at 900 °C with regard to investigate the influence of NaCl on fluidization stability. Based on the measurements of pressure drop, bed temperature and SEM-EDS, it was found that NaCl does not react with the iron ore but in fact only acts as glue between iron ore particles. Further, the sodium transfer routes in CLC of Zhundong coal with iron ore based oxygen carrier are given and some discussions are made with regard to practical operation. The corrosion problems on the heating surface in the air reactor can be significantly reduced compared to a conventional Zhundong coal fueled furnace, since most of sodium will release and be converted in the fuel reactor.

Abstract Mechanical vibration is often applied in industrial scale fluidized bed dryers for food and pharmaceutical powders to overcome operational problems caused by the cohesiveness of the products. However, the understanding of the process regarding detailed modeling, apparatus design, up-scaling and process optimization is still incomplete. Almost all of the experimental research in the field of vibrating fluidized beds is conducted on a lab-scale. Within the current project, experiments are conducted in a pilot plant scale unit with a cross section of 250 × 500 mm2 and a total height of 3 m. The influence of several process parameters, such as gas velocity, vibration intensity, powder moisture content and bed mass, on the fluidization characteristics of whole milk powder is studied. In order to characterize the lower limit of fluidization from an operational point of view, the velocity of complete fluidization (ucf) is used and quantified. This is defined as the lowest superficial gas velocity at which the entire bed is fully fluidized. It could be observed that an increase in moisture content of the powder results in a significant increase of ucf. The introduction of mechanical vibration into the bed results in the reduction of ucf, the expansion of the bed and the reduction of the bubble volume fraction. This expansion of the suspension phase explains the increase in heat and mass transfer in vibrated fluidized beds, reported in the literature. Adding the effect of vibration intensity to well established correlations allows for the accurate prediction of gas hold-up and bed porosity of fluidized beds of cohesive whole milk powder under mechanical vibration at low vibration frequencies.

In the current overview paper, various issues related to the combustion of agricultural residues are discussed. Attention has been given to the problems associated with the properties of the residues such as low bulk density, low ash melting points, high volatile matter contents and the presence of nitrogen, sulfur, chlorine and sometimes high moisture contents. Consequently the issues discussed include densification, the combustion mechanisms of agricultural residues, problems of low melting point of ash such as agglomeration and fouling, emissions and co-combustion. Further, design considerations of facilities for the combustion of agricultural residues are discussed.

AbstractThe performance of a catalytic fluidized‐bed reactor is strongly dependent on the properties of the catalyst of which the particle‐size distribution is one. The main influences on the particle‐size distribution are attrition of the catalyst particles, and the classifying effect of the solids recovery system. In a fluidized‐bed reactor a particle will be subjected to attrition due to different mechanisms in different parts of the system, namely attrition by gas jets near the bottom of the fluidized bed, bubble‐induced attrition in the fluidized bed itself, and attrition during the passage through a cyclone. All these different attrition mechanisms are considered in this work by different mathematical models. It is known that a fresh catalyst is much more fragile and exhibits a much higher attrition rate at the beginning of exposure to mechanical stress than under steady‐state conditions. Depending on the mechanism the particles need different times or in the case of attrition in a cyclone a certain number of passages to reach a constant value of the attrition rate. In the fluidized‐bed system a particle will during its aging experience all the different attrition mechanisms. In order to summarize the effect of these stresses on the particle within the different parts of the fluidized‐bed system, the concept of the “stress history” has been developed, which allows a uniform treatment of the different attrition mechanisms. This concept has been implemented into an existing population balance model. Experiments with a FCC‐catalyst in a fluidized/bed‐cyclone circulation system are well described by this model. © 2007 American Institute of Chemical Engineers AIChE J, 2007