Simulation of heavy metal vaporization dynamics in a fluidized bed
Introduction
A wide range of wastes including municipal solid wastes (MSW), sewage sludge, petroleum waste, and paper industry waste, etc. can be incinerated in fluidized beds. Trace element emission from the waste incineration results primarily from the vaporization of elements during the combustion process. Among trace elements, heavy metals (HMs) are of the most concern because of their toxicity. The targeted metals are not only volatile compounds, such as Hg, but also semi-volatile, such as Sb, As, Cd, Pb, etc. which cause a significant enrichment in fly ash even though a substantial portion may remain in the bottom ash.
The behavior of trace elements in combustion and incineration systems was mainly examined by analysis of solid residues (Linak & Wendt, 1993; Chandler et al., 1997; Wey, Yu, & Jou, 1998). Ho, Lee, Shiao, Hopper, and Bostick (1995) incinerating artificial solid wastes showed that HMs with high saturated vapor pressure enter the gas-phase easily after vaporizing, subsequently enriching in fly ash more than in bottom ash.
The thermodynamic equilibrium method was successfully applied for identifying and quantifying emissions in entrained bed coal gasification and pulverized coal combustion (PCC) (Frandsen, Dam-Johansen, & Rasmussen, 1994; Yan, Gauthier, & Flamant, 2001). However, there is experimental evidence that equilibrium calculations over predict the amount of metal vaporized (Barton, Maly, Clark, & Seeker, 1988). In fluid bed combustors, the temperature is lower and the particle size is much greater than that in PCC systems, which is favorable to kinetic limitations. Therefore, the results of equilibrium calculations may be clearly questionable in some cases.
There exist a few models in literature dealing with HM vaporization during combustion and accounting for other phenomena that chemical equilibrium. Helble (2000) developed a semi-empirical model of HM emission, in which fundamental laboratory results as well as field emission data were included. Ho et al. (1995) have developed a kinetic model. This model was based on the heat and mass transfer law and on reaction kinetics to simulate the metal vaporization process. Experiments were used to identify kinetics in the model, although the latter predicted reasonably well the experimental observations. More recently, our team developed a mathematical model (Abanades, Flamant, & Gauthier, 2001) to predict the fate of metallic species according to the main phenomena controlling the vaporization process: heat and mass transfer (transport phenomena) and chemical reactions involving HM (by a thermodynamical analysis). This model permits to predict, without any fitting parameter, the extent of HM vaporization from a mineral porous matrix as long as its physical properties are known.
In order to simulate the HM release from burning particles in fluidized bed, a model of hydrodynamics is required. Due to the development of computer capacities, discrete particle models, or Lagrangian models, have become very useful and versatile tools to study the hydrodynamic behavior of particulate flows. They offer a more natural way to simulate gas solid flow with particles of different size and/or different density, since each individual particle is tracked in the simulation. Moreover, Lagrangian simulation is a powerful tool to investigate the detailed phenomena at the individual particle scale. Discrete particle models have been combined with a Eulerian fluid model to simulate phenomena such as bubbling, slugging and solid transport in bubbling and circulating fluidized beds (Tsuji, Kawaguchi, & Tanaka, 1993; Hoomans, Kuipers, Briels, & Van Swaaij, 1996; Xu & Yu, 1997; Helland, Occelli, & Tadrist, 2000). These models have been properly validated by comparison with experiments (Van Wachem, Van der Schaff, Schouten, Krishna, & Van den Bleek, 2001; Hoomans, Kuipers, Mohd Salleh, Stein, & Seville, 2001). In addition, Kaneko, Shiojima, and Horio (1999) used the discrete element method (DEM), incorporating the energy balance and the reaction rate, to investigate the temperature effect on particles and gas in a fluidized bed reactor producing polyolefin.
The aim of this paper is to validate the predictions of the Eulerian–Lagrangian simulation of HM vaporization with experimental data, which were obtained by fluidizing a mixture of sand and CdCl2-spiked alumina in a high-temperature fluid bed. The model combines two recent theoretical approaches: on the one hand, a vaporization model accounting for chemical reaction, physical phenomena, and mass transfer at the particle scale (Abanades et al., 2001). On the other hand, a Lagrangian model of fluidized bed, which permits, in particular, to predict the flow structure of a fluid–particle mixture at the particle scale (Hoomans et al., 1996). Solving the mass transport equation in the gas flow combines both models. The model predicts the HM vaporization dynamics from a particle, the particle and gas flow structures and the HM distribution in the fluidized bed. Experiments were carried out in a bench-scale, electrically heated, fluidized bed. Experimental data gave the HM vaporization percentage on the basis of mass balance in the particles.
Section snippets
HM vaporization model
The objective of the model is to predict the vaporization dynamics of a HM contained in a porous particle, and consequently the enrichment of the neighboring gas due to the HM vaporization.
A thermodynamic model is used first, to determine the partitioning and the speciation of the HM at the considered temperature. Then, the possible limiting steps of the overall kinetics are considered: thermal conduction through the particle, diffusion of metallic vapors from the porous matrix to the particle
Experimental setup
A schematic representation of the experimental setup is displayed in Fig. 4. More details are given in Abanades (2001). The AISI 316L S.S. reactor is a and high cylinder. The non-corrosive (air or mixture ) fluidizing gas is heated through an electrical resistance, and the reactor is heated by two half-cylinder radiative shells and temperature-controlled by K-thermocouples. A thick alumino-silica layer insulates the high-temperature fluidized bed. The fluidizing
Simulation results and discussion
Fig. 5 shows the gas and particle flow structure in a fluidized bed with homogeneous inlet condition. There exists a strong impulsive start-up process . Then, a stable fluidization succession of bubble formation and disappearance in the bed is established. It can be seen that the gas flows towards the regions of high porosity. Highly preferential flow leads to a very strong non-uniform drag force in the bed, which in turn affects greatly the particle flow structure. CdCl2 concentration
Conclusion and future works
A Eulerian–Lagrangian simulation of CdCl2 vaporization in two-dimensional fluidized bed is validated by experiments. Simulation was performed by combining a CdCl2 vaporization model and a Lagrangian model developed for bubbling fluidized bed. Models were coupled by including the CdCl2 transport equation in the gas flow. Experimental data were obtained by fluidizing a mixture of sand and CdCl2-spiked alumina at . Theoretical and experimental results concerning the time evolution of CdCl2
Notations
HM concentration in the porous media, HM concentration in the transport gas, effective drag coefficient constant number initial HM concentration in the pore, true molecular diffusion coefficient, pseudo-diffusion coefficient, effective diffusivity, molecular diffusion coefficient of the HM, particle diameter, impulsive force, N drag force, N fluid–particle interaction force, N total vaporization rate of HM in a computational cell,
Acknowledgements
Authors are grateful to AFCRST (Association Franco-Chinoise pour la Recherche Scientifique et Technique), ADEME (Agence de l'Environment et de la Maı̂trise de l'Energie) and to Natural Science Fundation of China, Key Program of China for International Cooperation of Science and Technology No. 5007615 for financial assistance.
References (31)
- et al.
Modeling of heavy metal vaporization from a mineral matrix
Journal of Hazardous Material
(2001) - et al.
Transport phenomena in multi-particle systems—3. Active particle mass transfer in fluidized beds of inert particles
Chemical Engineering Science
(1988) - et al.
Trace elements from combustion and gasification of coal—an equilibrium approach
Progress in Energy and Combustion Science
(1994) The mass transfer coefficient for oxygen reacting with a carbon particle in a fluidized bed or packed bed
Combustion and Flame
(2000)A model for the air emissions of trace metallic elements from coal combustors equipped with electrostatic precipitators
Fuel Processing Technology
(2000)- et al.
Numerical study of cluster formation in a gas-particle circulating fluidized bed
Powder Technology
(2000) - et al.
Metal behavior during fluidized bed thermal treatment of soil
Waste Management
(1995) - et al.
Experimental validation of granular dynamics simulations of gas-fluidized beds with homogeneous in-flow conditions using Positron Emission Particle Tracking
Powder Technology
(2001) - et al.
DEM simulation of fluidized beds for gas-phase olefin polymerization
Chemical Engineering Science
(1999) - et al.
Mass transfer to large particles in fluidized beds of smaller particles
Chemical Engineering Science
(1984)
Toxic metal emissions from incinerationmechanisms and control
Progress of Energy Combustion Science
Fundamentals of turbulent gas-solid flows applied to circulating fluidized bed combustion
Progress in Energy and Combustion Science
Discrete particle simulation of two-dimensional fluidized bed
Powder Technology
Experimental validation of Lagrangian–Eulerian simulations of fluidized beds
Powder Technology
The influence of heavy metals on the formation of organics and HCL during incinerating of PVC-containing waste
Journal of Hazardous Material
Cited by (19)
Behaviour of mercury during Co-incineration of sewage sludge and municipal solid waste
2020, Journal of Cleaner ProductionCoupling scales for modelling heavy metal vaporization from municipal solid waste incineration in a fluid bed by CFD
2015, Waste ManagementCitation Excerpt :It is thus essential to understand their behavior so as to achieve a strict control of their emission. Several authors have studied, both experimentally (Abanades, 2001; Sun et al., 2004; Rio et al., 2007; H. Zhang et al., 2008; Y. Zhang et al., 2008; Yu et al., 2012, 2013) and theoretically – by thermodynamic modelling – (Zhou et al., 2002; Sørum et al., 2003; Ménard et al., 2006; Mazza et al., 2009, 2010;), the heavy metal release from different kinds of particles. The metal speciation can be predicted from the waste and the gas compositions, but the transient system evolution cannot be predicted by thermodynamics calculations (Falcoz et al., 2009; Asthana et al., 2010b).
Modeling trace element partitioning during coal combustion
2014, Fuel Processing TechnologyCitation Excerpt :The diffusion rate is influenced by the physical characteristics of the minerals such as their melting behavior, relative position to other minerals, and mineral transformations such as coalescence, fragmentation, and vaporization. Many existing TE models are based on bulk gas-phase conditions and empirically derived predictors [7–9] while other models are based on thermodynamic equilibrium [7,10]. The latter approach assumes that vaporization and reaction to gas phase speciation are thermodynamically limited when in fact, these processes are expected to be transport limited at the highly dynamic conditions of char pyrolysis (i.e., fast heating under reducing atmospheric conditions).
Local CFD kinetic model of cadmium vaporization during fluid bed incineration of municipal solid waste
2013, Journal of Hazardous MaterialsCitation Excerpt :It is then necessary to be able to predict the evolution of metal speciation and partitioning during this thermal process. Several authors have studied, both experimentally [2,4–7] and theoretically–by thermodynamic modeling – [3,8–10], the heavy metal release from various particles. The metal speciation can be predicted from the waste and the gas compositions, but the transient system cannot be predicted by thermodynamics calculations only.
A method for measuring the kinetics of organically associated inorganic contaminant vaporization during coal combustion
2011, Fuel Processing TechnologyCitation Excerpt :Such models would also assist in trouble-shooting TE-related process issues, such as corrosion and fouling by TE deposition. Previous work associated with trace-level contaminants has not generated data with sufficient definition to allow modeling based on theoretical constructs; although many empirical-based models exist [1–6]. More specifically, data and models of trace element vaporization and subsequent partitioning to fly ash particle surfaces have not been developed at the micro-environmental conditions of the burning char particle.
DEM simulation of particle mixing in flat-bottom spout-fluid bed
2010, Chemical Engineering Research and Design