A new experimental set-up and a new simulated moving bed (SMB) operation are presented in this work. A desktop SMB unit developed as a modification of the commercial AKTA(TM) explorer working platform has been utilized for the separation of different mixtures of nucleosides. Both two fraction and three fraction SMB separations have been carried out, the latter made possible by the adoption of a new SMB configuration and operating mode (three fraction SMB, 3F-SMB, operation). Experiments demonstrate the feasibility of the 3F-SMB operation, and confirm the trends predicted based on considerations about retention of the components to be separated along the unit.
A model for flow-induced collisions of fractal aggregates is developed. The model is based on the analysis of the relative trajectories between a pair of aggregates that takes into account their hydrodynamic and their colloidal interactions. Regarding the former, the aggregates are modeled as permeable spheres where the Brinkman equation is used to describe the flow inside the aggregates. Interparticle forces are incorporated by considering the forces between the primary particles in the two aggregates that are the nearest. The model results in a collision efficiency that depends on the masses of the colliding aggregates, the fractal dimension, and a nondimensional Hamaker constant characterizing the interparticle forces. The collision efficiency model is used to investigate the dynamics of a suspension undergoing coagulation. Significant deviations with respect to existing collision efficiency models are evidenced.
Solid supported evaporation (SSE) is a simple, nonselective method for isolating nonvolatile compounds from a solution. The solution is put in contact with porous polymer beads onto which the compound deposits upon evaporation of the solvent. This brings some advantages over direct evaporation to dryness in terms of safety, thermal decomposition, and solid handling, as the loaded beads form a free flowing granular material that is easily recovered. In this paper, SSE in a semicontinuous operating mode is investigated where the solution is continuously fed to (respectively sprayed over) an agitated bed of dry beads put under vacuum. It is found that under conditions where the solvent evaporation fate is high with respect to the feed rate, high bead loadings can be achieved before extensive sticking of beads and compound to the vessel walls occurs. The type of compound and solvent had little influence on the process performance, and, in cases where this was explored, the bead loading was found to be homogeneous. Based on a balance equation for the solvent fed to the system, a model is developed that results in a simple scale up criterion. The latter was successfully applied for transferring SSE from lab to the kilo lab scale.
The hydrodynamic breakup of small inertial aggregates in homogenous and isotropic turbulence is studied through numerical simulations. Small inertial aggregates are subject to shear stress caused by the local velocity gradient and drag stress caused by the relative velocity of the aggregate and the fluid flow. In our simulations, we follow aggregates moving through the flow and record the total stress acting on them. Breakup is assumed to occur when the total stress overcomes a predefined threshold representing the aggregate strength. By determining how long it takes for an aggregate to reach a stress exceeding its strength for the first time, we are able to derive a breakup rate. It is found that with increasing aggregate inertia, the drag stress rapidly becomes the dominant stress resulting in an increase of the breakup rate with increasing the aggregate inertia.
Breakup of small tracer-like aggregates is studied by means of numerical simulations in four different flows, namely homogeneous isotropic turbulence, smooth stochastic flow, turbulent channel flow, and developing boundary layer flow. Aggregate breakup occurs when the local hydrodynamic stress σ ∼ ε1/2, where ε is the local energy dissipation, overcomes a given threshold value σcr [or equivalently εcr ∼ σcr2 ] characteristic for a given type of aggregates. Following the aggregate trajectory upon release and detecting the first occurrence of local energy dissipation exceeding the predefined threshold allows for estimating the breakup rate as a function of εcr. Results show that the breakup rate decreases with increasing threshold. For small values of the threshold, this decrease assumes consistent scaling among the different flows which is explained by universal small scale flow properties.
Breakup of small aggregates in fully developed turbulence is studied by means of direct numerical simulations in a series of typical bounded and unbounded flow configurations, such as a turbulent channel flow, a developing boundary layer and homogeneous isotropic turbulence. The simplest criterion for breakup is adopted, whereby aggregate breakup occurs when the local hydrodynamic stress sigma similar to epsilon(1/2), with epsilon being the energy dissipation at the position of the aggregate, overcomes a given threshold sigma(cr), which is characteristic for a given type of aggregate. Results show that the breakup rate decreases with increasing threshold. For small thresholds, it develops a scaling behaviour among the different flows. For high thresholds, the breakup rates show strong differences between the different flow configurations, highlighting the importance of non-universal mean-flow properties. To further assess the effects of flow inhomogeneity and turbulent fluctuations, the results are compared with those obtained in a smooth stochastic flow. Furthermore, we discuss the limitations and applicability of a set of independent proxies.
Aggregates grown in mild shear flow are released, one at a time, into homogeneous isotropic turbulence where their breakup is recorded by three-dimensional particle tracking velocimetry (3D-PTV). The aggregates have an open structure with fractal dimension around 2.2, and their size varies from 0.9 to 3.1 mm which is large compared to the Kolmogorov length scale η = 0.15 mm. 3D-PTV allows for the simultaneous measurement of aggregate trajectories and the full velocity gradient tensor along their pathlines which enables us to access the Lagrangian stress history of individual breakup events. The analysis suggests that aggregates are mostly broken due to accumulation of drag stress over a time interval of order Kolmogorov time scale, O(τη). This finding is explained by the fact that the aggregates are large, which gives their motion inertia and which increases the time for stress propagation inside the aggregate.
The breakup and aggregation of small solid particle aggregates in hetero-geneous °ows is modeled through a multizonal population balance model. Breakupand aggregation are described through comprehensive rate expressions that takeinto account the local properties of the turbulent °ow. Two approaches are pursuedto de¯ne the zones: in the ¯rst approach the zones are prede¯ned and computational°uid dynamics is used to compute the zone properties. In the second approach thezone properties are directly ¯tted to experimental data. Model results compare wellwith experiments for the aggregation of a polystyrene latex in a stirred tank evenwhen only two zones are used.
Coagulation of small particles in agitated suspensions is governed by aggregation and breakage. These two processes control the time evolution of the cluster mass distribution (CMD) which is described through a population balance equation (PBE). In this work, a PBE model that includes an aggregation rate function, which is a superposition of Brownian and flow induced aggregation, and a power law breakage rate function is investigated. Both rate functions are formulated assuming the clusters are fractals. Further, two modes of breakage are considered: in the fragmentation mode a particles splits into w ≥ 2 fragments of equal size, and in the erosion mode a particle splits into two fragments of different size. The scaling theory of the aggregation-breakage PBE is revised which leads to the result that under the negligence of Brownian aggregation the steady state CMD is self-similar with respect to a non-dimensional breakage coefficient θ. The self-similarity is confirmed by solving the PBE numerically. The self-similar CMD is found to deviate significantly from a log-normal distribution, and in the case of erosion it exhibits traces of multimodality. The model is compared to experimental data for the coagulationof a polystyrene latex. It is revealed that the model is not flexible enough to describecoagulation over an extended range of operation conditions with a unique set of parameters. In particular, it cannot predict the correct behavior for both a variation in the solid volume fraction of the suspension and in the agitation rate (shear rate).
The breakup of solid aggregates suspended in a turbulent flow is considered. The aggregates are assumed to be small with respect to the Kolmogorov length scale and the flow is assumed to be homogeneous. Further, it is assumed that breakup is caused by hydrodynamic stresses acting on the aggregates, and breakup is therefore assumed to follow a first-order kinetic where K-B(x) is the breakup rate function and x is the aggregate mass. To model K-B(x), it is assumed that an aggregate breaks instantaneously when the surrounding flow is violent enough to create a hydrodynamic stress that exceeds a critical value required to break the aggregate. For aggregates smaller than the Kolmogorov length scale the hydrodynamic stress is determined by the viscosity and local energy dissipation rate whose fluctuations are highly intermittent. Hence, the first-order breakup kinetics are governed by the frequency with which the local energy dissipation rate exceeds a critical value (that corresponds to the critical stress). A multifractal model is adopted to describe the statistical properties of the local energy dissipation rate, and a power-law relation is used to relate the critical energy dissipation rate above which breakup occurs to the aggregate mass. The model leads to an expression for K-B(x) that is zero below a limiting aggregate mass, and diverges for x -> infinity. When simulating the breakup process, the former leads to an asymptotic mean aggregate size whose scaling with the mean energy dissipation rate differs by one third from the scaling expected in a non-fluctuating flow.
Aggregation of rigid colloidal particles leads to fractal-like structures that are characterized by a fractal dimension d(f) which is a key parameter for describing aggregation processes. This is particularly true in shear aggregation where d(f) strongly influences aggregation kinetics. Direct measurement of d(f) in the early stages of shear aggregation is however difficult, as the aggregates are small and few in number. An alternative method for determining d(f) is to use an aggregation model that when fitted to the time evolution of the cluster mass distribution allows for estimating d(f). Here, we explore three such models, two of which are based on an effective collision sphere and one which directly incorporates the permeable structure of the aggregates, and we apply them for interpreting the initial aggregate growth measured experimentally in a turbulent stirred tank reactor. For the latter, three polystyrene latexes were used that differed only in the size of the primary particles (d(p) = 420, 600, and 810 nm). It was found that all three models describe initial aggregation kinetics reasonably well using, however, substantially different values for 4 To discriminate among the models, we therefore also studied the regrowth of preformed aggregates where d(f) was experimentally accessible. It was found that only the model that directly incorporates the permeable structure of the aggregates is able to predict correctly this second type of experiments. Applying this model to the initial aggregation kinetics, we conclude that the actual initial fractal dimension is d(f) = 2.07 +/- 0.04 as found from this model.
The hydrodynamic interaction of two neutrally buoyant porous aggregates is investigated under creeping flow conditions for the case where the undisturbed velocity of the surrounding flow field is a linear function of position. In this framework, the relative velocity between two aggregates is given by the deformation of the undisturbed flow expressed through the rate of strain and the angular velocity of the flow field, and by two flow-independent hydrodynamic functions, typically referred to as A and B, which account for the disturbance of the flow field due to the presence of the particles [G. K. Batchelor and J. T. Green, J. Fluid Mech. 56, 375 (1972)]. In the present paper, the analysis of thehydrodynamic interaction that is known for the case of two impermeable, solid particles is extended to the case of porous aggregates by applying Brinkman's equation to describe the flow within the aggregates. A reflection scheme is applied to calculate A and B and the obtained expressions are applied to interpret the orthokinetic aggregation ofaggregates in diluted suspensions, where the collision frequency is computed using the method of relative trajectories of a pair of aggregates.
The breakup of small solid aggregates in homogeneous and isotropic turbulence is studied theoretically and by using direct numerical simulations at high Reynolds number, Re-lambda similar or equal to 400. We show that turbulent fluctuations of the hydrodynamic stress along the aggregate trajectory play a key role in determining the aggregate mass distribution function. The differences between turbulent and laminar flows are discussed. A definition of the fragmentation rate is proposed in terms of the typical frequency at which the hydrodynamic stress becomes sufficiently high to cause breakup along each Lagrangian path. We also define an Eulerian proxy of the real fragmentation rate, based on the joint statistics of the stress and its time derivative, which should be easier to measure in any experimental setup. Both our Eulerian and Lagrangian formulations define a clear procedure for the computation of the mass distribution function due to fragmentation. Contrary, previous estimates based only on single point statistics of the hydrodynamic stress exhibit some deficiencies. These are discussed by investigating the evolution of an ensemble of aggregates undergoing breakup and aggregation.
Pyrolysis of biomass in a rotary kiln finds application both as an intermediate step in multistage gasification as well as a process on its own for the production of biochar. In this work, a numerical model for pyrolysis of lignocellulosic biomass in a rotary kiln is developed. The model is based on a set of conservation equations for mass and energy, combined with independent submodels for the pyrolysis reaction, heat transfer, and granular flow inside the kiln. The pyrolysis reaction is described by a two-step mechanism where biomass decays into gas, char, and tar that subsequently undergo further reactions; the heat transfer model accounts for conduction, convection and radiation inside the kiln; and the granular flow model is described by the well known Saeman model. The model is compared to experimental data obtained from a pilot scale rotary kiln pyrolyzer. In total 9 pilot plant trials at different feed flow rate and different heat supply were run. For moderate heat supplies we found good agreement between the model and the experiments while deviations were seen at high heat supply. Using the model to simulate various operation conditions reveals a strong interplay between heat transfer and granular flow which both are controlled by the kiln rotation speed. Also, the model indicates the importance of heat losses and lays the foundation for scale up calculations and process optimization.
This work investigates the ternary phase behavior of the two enantiomers of ibuprofen and water. The two enantiomers crystallize as a racemic compound and exhibit a thermodynamically stable liquid-liquid phase separation (LLPS), which extends over the entire enantiomeric composition range. First, the generic phase behavior of racemic compound forming systems exhibiting a stable LLPS is derived by exploiting the consolidated knowledge of conglomerate forming systems obtained in the first part of this series. Below the onset temperature of the LLPS, the system behaves like a typical racemic compound forming system. As for conglomerate forming systems, the onset of the LLPS is found to occur through a ternary monotectic equilibrium, where a new, solute-rich liquid phase emerges inside each solid-solid-liquid phase region. Then, the ternary phase diagram of the ibuprofen/water system in the temperature range from 40 to 82 °C is presented together with the outcome of a thorough experimental investigation. Our theoretical considerations are in excellent agreement with experimental results.
This work investigates the ternary phase diagram of the enantiomers of ethyl-2-ethoxy-3-(4-hydroxyphenyl)propanoate (EEHP) in cyclohexane. The enantiomers of EEHP form aconglomerate, and both the pure enantiomer and the racemic mixture exhibit stable oilingout in cyclohexane. Our analysis shows that the ternary phase diagram of such a system assumes a unique structural evolution around the onset temperature of oiling out; that is, we found that the onset of oiling out strictly occurs through the emergence of a second liquid phase of racemic composition. Furthermore, we found that the further evolution of the ternary phase diagram, i.e., above the onset temperature of oiling out, is dictated by the properties of the phase diagrams pure enantiomer/solvent and racemic mixture/solvent. Our theoretical considerations are in excellent agreement with experimental measurements of EEHP in cyclohexane.
A methodology for the design of cooling crystallization processes for chiral resolution from nonracemic initial solutions is presented. Such processes are encountered when chiral resolution is attained by hybrid processes, where the crystallization step is preceded by a pre-enrichment step accomplished by either asymmetric synthesis or another separation technique. The work focuses on substances that crystallize as conglomerates and accounts for the occurrence of oiling out, i.e., an undesired liquid liquid phase separation during crystallization. The generic ternary phase diagrams for conglomerate-forming systems with and without oiling out are derived. This knowledge is then applied to identify suitable operating conditions for chiral resolution. As crystallization is started from saturated solutions, the crystallization process is characterized by three parameters: the initial enantiomeric excess and the initial temperature, which together implicitly define the position of the operating point in the phase diagram, and the final operating temperature, which defines the composition and the amount of the phases present at the end of crystallization. For any initial enantiomeric excess, the methodology yields distinct areas in the initial versus final temperature plane containing pairs of operating temperatures that are suitable for chiral resolution. Such operating map bears great potential in improving the design and optimization of chiral resolution processes by crystallization.
Initial aggregation kinetics for three particle sizes and broad range of Péclet numbers were investigated under turbulent conditions in stirred tank. This allowed us to observe the transition from diffusion-controlled to purely shear-induced aggregation. The evolution of the root-mean-square radius of gyration, zero-angle intensity of scattered light, and obscuration was obtained by small-angle static light scattering. For a given particle sizethe measured evolution of all integral quantities obtained for various volume averageshear rates (G), scales with a dimensionless time, τexp = αexp × (G) × φ × t. The experimentally obtained aggregation efficiency αexp, follows the power law αexp = Pe-n, where Pe is the primary particle Péclet number. With increasing particle size a decrease in n is observed in accordance with theory and literature data. As previously predicted by population balance equation simulations three aggregationregimes were observed experimentally.
In this paper we study the ductile breakup of tracer aggregates in an incompressible, homogeneous, and isotropic three-dimensional turbulent flow. The flow dynamics is studied by means of a direct numerical simulation, whereas the Lagrangian velocities and stress statistics along trajectories are obtained by particle tracking. We investigate the breakup dynamics under the hypothesis that aggregates are able to deform and accumulate energy. Within this framework, breakup occurs when the energy transferred to the aggregate by the flow exceeds a critical value. We contrast our predictions for ductile breakup with those obtained for brittle breakup. We observe that turbulence intermittency is crucial for the breakup of brittle aggregates, while it becomes less relevant for ductile aggregates. In the limit of highly ductile aggregates the breakup rate is dictated by the mean properties of the flow. We propose a simple model to capture this behaviour.
The breakup of inertial, solid aggregates in an incompressible, homogeneous and isotropic three-dimensional turbulent flow is studied by means of a direct numerical simulation, and by a Lagrangian tracking of the aggregates at varying Stokes number and fluid-to-particle density ratio. Within the point-particle approximation of the Maxey–Riley–Gatignol equations of motion, we analyze the statistics of the time series of shear and drag stresses, which are here both deemed as responsible for aggregate breakup. We observe that, regardless of the Stokes number, the shear stresses produced by the turbulent velocity gradients similarly impact the breakup statistics of inertial and neutrally buoyant aggregates, and dictate the breakup rate of loose aggregates. When the density ratio is different from unity, drag stresses become dominant and are seen to be able to cause to breakup of also the most resistant aggregates. A transition from a shear-dominated to a drag-dominated breakup regime is observed, and a power-law is seen to well describe the breakup rate of loose aggregates regardless of their inertia. The present work assesses the role of shear and drag stresses on aggregate breakup and computes breakup rates to be possibly used in population balance models.
This paper investigates numerically the shear-induced aggregation of mixed populations of colloidal particles leading to the formation of clusters. Suspensions with different amounts of positively and negatively charged colloidal particles are simulated. To resolve the aggregation kinetics and structural properties of the formed clusters, we resort to a mixed deterministic-stochastic simulation method. The method is built on a combination of a Monte Carlo algorithm to sample a statistically expected sequence of encounter events between the suspended particles and a discrete element method built in the framework of Stokesian dynamics to simulate the encounters in a fully predictive manner. Results reveal a strong influence of the composition of the population on both the aggregation kinetics and the aggregate structure. In particular, we observe a size-stabilization phenomenon taking place in the suspension when the relative concentration of the majority particles lies in the range 80-85%; i.e., starting from primary particles, after a short growth period, we observed a cessation of aggregation. Inspection of the aggregate morphology shows that the formed clusters are composed of few minority particles placed in the inner region, while the aggregate surface is covered by majority particles, acting to provide a shielding effect against further growth.
This work reports the results of the life cycle assessment (LCA) of two carbonation processes aimed at permanent CO2 storage, employing Basic Oxygen Furnace (BOF) slag from steel manufacturing as alkalinity source. Specifically, the performance of the slurry phase and wet carbonation routes were compared assuming to store the CO2 emitted from a 10 MW conventional natural gas power plant. The LCA was based on the material and energy requirements for each of the involved process steps, i.e. pre-treatment and transport of raw materials, CO2 compression, carbonation, post treatments and management of the obtained products. The slurry and wet route resulted in a net avoided greenhouse warming potential (GWP) of 473 and 384 kg CO2/MWhel, respectively. Nevertheless, both routes affected the other environmental impact categories. In general, the wet route had approximately two times higher impact than the slurry route, due in particular to the higher material and energy requirements. An exception was the abiotic resource depletion which resulted higher for the slurry route due to greater water requirement with respect to the wet route. The contributions to all mid-point impact categories were mainly due to energy requirements. A sensitivity analysis showed that the environmental impacts are affected by the energy mix and by the transport distance of slags and carbonation products.
Transport and particulate processes are ubiquitous in environmental, industrial and biological applications, often involving complex geometries and porous media. In this work we present a general population balance model for particle transport at the pore-scale, including aggregation, breakage and surface deposition. The various terms in the equations are analysed with a dimensional analysis, including a novel collision-induced breakage mechanism, and split into one- and two-particles processes. While the first are linear processes, they might both depend on local flow properties (e.g. shear). This means that the upscaling (via volume averaging and homogenisation) to a macroscopic (Darcy-scale) description requires closures assumptions. We discuss this problem and derive an effective macroscopic term for the shear-induced events, such as breakage caused by shear forces on the transported particles. We focus on breakage events as prototype for linear shear-induced events and derive upscaled breakage frequencies in periodic geometries, starting from nonlinear power-law dependence on the local fluid shear rate. Results are presented for a two-dimensional channel flow and a three dimensional regular arrangement of spheres, for arbitrarily fast (mixing-limited) events. Implications for linearised shear-induced collisions are also discussed. This work lays the foundations of a new general framework for multiscale modelling of particulate flows.
Particle laden flows with reactive particles are common in industrial applications. Chemical reactions inside the particle or deposition at the surface can generate additional flow phenomena that affect the heat, mass and momentum transfer between the particle and bulk flow. This work aims at investigating the effect of Stefan flow on the drag coefficient of a spherical particle immersed in a uniform flow. Fully resolved 3D simulations were carried out for particle Reynolds numbers based on the free stream velocity ranging from 0.5 to 3. Simulations are carried out in foam-extend CFD software, using the Immersed Boundary(IB) method for treating fluid-solid interactions. The simulations were validated against data for particles without reactive flow, and against the analytical solution for Stefan flow around a particle in a quiescent fluid. We found that in the considered range of Reynolds number the drag coefficient decreases linearly with in increase in Stefan flow velocity.
The aim of the work was to study the effects of neighboring particles with uniform Stefan flow in particle- fluid flows. Particle-resolved numerical simulations were carried out for particles emitting a uniform Stefan flow into the bulk fluid. The bulk fluid was uniform and isothermal. The Stefan flow volume emitted from the two particles is equal, such that it represents idealized conditions of reacting particles. Particles were located in tandem arrangement and particle distances were varied between 1.1 and 10 particle diameters (1.1 <= L/D <= 10). Three particle Reynolds numbers were considered during the simulations (Re = 2.3,7 and 14), which is similar to our previous studies. Three Stefan flow velocities were also considered during simulations to represent inward, outward, and no Stefan flow. The drag coefficient of the particles without Stefan flow showed that the results fit with previous studies on neighbor particle effects. When the particle distance is greater than 2.5 diameters (L/D > 2.5), the effects of Stefan flow and neighboring particles are independent of each other. I.e. an outward Stefan flow decreases the drag coefficient (CD) while an inward Stefan flow increases it and the upstream particle experience a higher CD than the downstream particle. When L/D <= 2.5, the effect of Stefan flow is dominant, such that equal and opposite pressure forces act on the particles, resulting in a repelling force between the two neighboring particles. The pressure force showed a large increase compared to the viscous force at these distances. The effect of Stefan flow is weakened at higher Reynolds numbers. A model was developed for the calculation of the drag coefficient. The model, which reproduce the results from the numerical simulations presented above, is a product of independent models that describe the effects of both neighboring particles and two distinguished effects of the Stefan flow.
A Stefan flow can be generated during a phase change or reactions of a particle immersed in a fluid. This study investigates the effect of Stefan flow on the exchange of momentum (drag coefficient (C-D)) and heat transfer (Nusselt number (Nu)) between the particle and bulk-fluid. Fully resolved simulations were carried out for a flow near a spherical particle immersed in a uniform bulk flow. The immersed boundary method is used for implementing fluid-solid interactions and the particle is considered as a static boundary with fixed boundary conditions. In a non-isothermal flow, the changes in thermophysical properties at the boundary layer played a role in the variation of C-D and Nu by a Stefan flow further. The previously developed model for the drag coefficient of a spherical particle in a uniform isothermal flow was modified for a uniform non-isothermal flow. The model is developed based on physical interpretation. A new model is developed for the Nusselt number for a spherical particle with a uniform Stefan flow combining available models in literature. The models are validated for Stefan Reynolds number - 8 <= Re-sf,Re- p <= 25 and particle Reynolds number of 2 <= Re-f <= 30 in gas flow (i.e. Pr approximate to 0.7).
Particle laden flows with reactive particles are common in industrial applications. Chemical reactions inside the particle can generate a Stefan flow that affects heat, mass and momentum transfer between the particle and the bulk flow. This study aims at investigating the effect of Stefan flow on the drag coefficient of a spherical particle immersed in a uniform flow under isothermal conditions. Fully resolved simulations were carried out for particle Reynolds numbers ranging from 0.2 to 14 and Stefan flow Reynolds numbers from (-1) to 3, using the immersed boundary method for treating fluid-solid interactions. Results showed that the drag coefficient decreased with an increase of the outward Stefan flow. The main reason was the change in viscous force by the expansion of the boundary layer surrounding the particle. A simple model was developed based on this physical interpretation. With only one fitting parameter, the performance of the model to describe the simulation data were comparable to previous empirical models. The Authors.
Lake Katwe is a saline lake within the East African Rift system in Western Uganda, with a rich source of mineral salts. The present work aims at evaluating possibilities of future salt extraction from the lake deposit. An isothermal evaporation experiment was conducted on the lake brines. The precipitated salts were characterized by X-ray diffraction (XRD) and scanning electron microscopy (SEM) methods. Various economic salts such as thenardite, gypsum, mirabilite, burkeite, hanksite, anhydrite, trona, halite, nahcolite, thermonatrite, and soda ash precipitate from the lake brines. The experiments also reveal the sequence of mineral salt precipitation in the order sulfates→chlorides→carbonates.
Lake Katwe brines and evaporites were investigated to determine their chemical, mineralogical and morphological composition. 30 brine samples and 3 solid salt samples (evaporites) were collected from different locations of the lake deposit. Several analytical techniques were used to determine the chemical composition of the samples including Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES), Inductively Coupled Plasma-Sector Field Mass Spectrometry (ICP-SFMS), ion chromatography, and potentiometric titration. The mineralogical composition and morphology of the evaporites was determined using X-ray diffraction (XRD) and scanning electron microscopy (SEM), respectively. Physical parameters of the lake brines such as density, electrical conductivity, pH, and salinity were also studied. The results show that the lake brines are highly alkaline and rich in Na+, Cl-, CO32-, SO42-, and HCO3- with lesser amounts of K+, Mg2+, Ca2+, Br-, and F- ions. The brines show an intermediate transition between Na-Cl and Na-HCO3 water types. Among the trace metals, the lake brines were found to be enriched in B, I, Sr, Fe, Mo, Ba, and Mn. The solid salts are composed of halite mixed with other salts such as hanksite, burkeite and trona. It was also observed that the composition of the salts varies considerably even within the same grades.
The characteristics of Katwe salt lake are briefly discussed. The lake is the largest of the eight saline lakes in the Katwe-Kikorongo volcanic field and is a major source of salt production in Uganda. Today, salt production at the lake is carried out using traditional and artisanal mining methods. Attempts to mechanize the production of domestic and commercial grade salt at the lake were unsuccessful due to the use of a wrong technology. In this paper, the most common available technologies for salt extraction from brine are described. These are divided into four broad categories, namely thermal, membrane, chemical and hybrid processes. A review of the state of the art, previous research and developments in these technologies is presented. A detailed analysis of the processes used was done based on studies reported in the literature. From the analysis, it was observed that thermal salt production processes, especially distillation and solar evaporation have the highest share in installed capacities worldwide. Membrane technologies such as Electro-dialysis, Reverse Osmosis and chemical technologies have not found wide application in the commercial salt industry. Electro-dialysis and Reverse Osmosis have been used mainly as pre-concentration processes for subsequent thermal processes. Prospects for application of hybrid systems for salt production through integration of thermal desalting processes should be investigated for better performance efficiencies and recoveries at the salt lake.
Solid supported evaporation (SSE) is a simple method to isolate dissolved compounds as a solid material. The solution is put in contact with granular porous polymer beads onto which the compounds deposit upon evaporation of the solvent. This brings some advantages over direct evaporation to dryness in terms of safety and handling of the solids. In this paper, SSE in batch mode is explored where the solution is added to the polymer beads at once, i.e. opposite to the semicontinuous mode where the solution is sprayed over a bed of beads. A number of compounds varying widely in their physical and chemical properties is studied. It is found that all compounds could be loaded onto the beads; however, the loading capacity depends on the properties of the compound and in general was lower than in the semicontinuous operating mode studied in an accompanying paper. For highly soluble compounds, higher loadings could be achieved when solvent evaporation was slow. In cases where tested, bead loading was found to be homogeneous within a batch. Recovery of compound from loaded beads was achieved by dispersing the beads in a solvent and washing of the filter cake after filtration. A relatively large amount of solvent is required to achieve full recovery.
Air pollution from volatile organic compounds (VOCs) is one of the most important environmental hazards. Advanced oxidation processes (AOPs) with UV systems have been showing high potential for the abatement of VOCs. This work is aimed at modeling UV reactors for scaling-up AOPs from lab-scale to full-scale. The proposed model has a novel approach coupling the UV fluence rate to the photo-kinetic mechanism, for a robust understanding of the phenomena involved. The results show that the 185 nm wavelength is deeply absorbed within few centimeters by oxygen, while the 254 nm wavelength is weakly absorbed by the ozone generated in the reactor. Based on the fluence rate calculations, the reactions of ozone generation and depletion were modeled. The ozone net concentration was compared to the experimental results, for model verification. The model accurately predicts the effect of the airflow rate and reactor diameter for the tested cases. The acetaldehyde oxidation reaction was modeled using a simplified kinetic mechanism, using the experimental data of VOC conversion for a further model verification. The suggested reactor models accurately predicted the effect of airflow rate, while exhibiting limitations for the effect of different reactor diameters. Therefore, a computational fluid dynamics (CFD) investigation is needed for an accurate modeling of the VOCs oxidation reaction, implementing the developed analytical expression for reducing the computational workload. By combining the developed model with a CFD simulator, it would be possible to simulate several reactors, also at full-scale, for predicting their performance and identifying optimal configurations.
We explore the possibility to derive a completely model-free rate expression using isoconversional methods. The Friedman differential method (Friedman, 1964) and the incremental integral method by Vyazovkin (2001) were both extended to allow for an estimation of not only the apparent activation energy but also the effective kinetic prefactor, defined as the product of the pre-exponential factor and the conversion function. Analyzing experimental thermogravimetric data for the pyrolytic decomposition of microcrystalline cellulose, measured at six different heating rates and three different initial sample masses (1.5-10 mg), revealed the presence of secondary char forming reactions and thermal lag, both increasing with increased sample mass. Conditioning of the temperature function enables extraction of more reliable prefactors and we found that the derived kinetic parameters show weak dependence on initial sample mass. Finally, by successful modeling of quasi-isothermal experimental curves, we show that the discrete rate expression estimated from linear heating rate experiments enables modeling of the thermal decomposition rate without any assumptions regarding the chemical process present. These findings can facilitate the design and optimization of industrial isothermal biomass fed reactors.
Gasification is a key technology for the utilization of biomass as an energy carrier. The WoodRoll process developed by Cortus Energy is a multistage gasification process where drying, pyrolysis and gasification are conducted in separate units. A central role is thereby given to the pyrolysis step which provides the gas to heat the entire process. In the WoodRoll process pyrolysis is run in an indirectly heated rotary kiln. In this work we study pyrolysis in a rotary kiln by means of numerical simulations and by evaluating pilot plant data obtained from a 500 kW pilot. The simulations indicate the importance of the heat transfer to the solid bed and the exothermic pyrolysis reactions that occur in the late stage of the pyrolysis process. The latter can cause an overshoot of the solid bed temperature. Evaluation of the pilot plant data shows the robustness of the process, expressed in good reproducible and stable operation.
Sulfidation of zinc oxide is a viable option for the removal of hydrogen sulfide from raw syngas and biogas. Recent experiments showed that sulfidation of nanoscopic zinc oxide particles leads to an outward growth and the formation of voids inside the particles. In this work, we derive a micro-scale model to describe this phenomenon. The model accounts for nucleation and growth of voids inside the particles, diffusion of Zn and O through the product layer, and deposition of the solid product at the particle surface as a result of the reaction between ZnO and H2S. The model is thus opposite to the well known shrinking core model where an inward growth of the product layer is assumed. To explore the effect of the outward growth on the dynamics of a packed bed adsorber the micro-scale model is combined with a macro-scale model that accounts for intra-pellet diffusion and convection along the packed bed. In the limit of fast nucleation and growth of voids inside the zinc oxide particles, the micro-scale model shows a qualitatively similar conversion profile to the shrinking core model, while when nucleation controls an inflection point in conversion profile is found. On the macro-scale, the outward growth can cause the clogging of pores inside the pellets which prevents the pellets from reaching full conversion. This leads to shorter breakthrough times of the packed bed due to the sealing of unreacted zinc oxide. Our results thus provide a possible explanation of the incomplete conversion of zinc oxide in packed beds.
Design and optimization of biomass gasification faces the challenge of feedstock variation. Specifically, design calculations require kinetic rate expressions for the given feedstock, whose rigorous determination is demanding and often exceeds available recourses in an early development stage. In this work, we model the slow pyrolysis of biomass for the production of biochar. The aim is to predict the conversion of raw biomass to biochar as a function of the process conditions. Here, we will show that TGA data processed with an isoconversional method is enough to obtain an effective rate expression which allows for predicting the behavior of the biomass at an arbitrary temperature evolution. Such rate expressions can then be used in the process model to simulate conversion of raw biomass to biochar. To illustrate the feasibility of this approach we consider four vastly different biomass, namely spruce wood, pulp, lignin and xylan–lignin, undergoing slow pyrolysis in an indirectly heated rotary kiln reactor. The results of our modeling are compared to experimental data obtained from a 500 kW pilot plant pyrolyzer and to a more detailed process model.
Microalgae is considered as a viable feedstock to biomass gasification. After synthesis in water medium, microalgae are separated and dried to a suitable degree to be fed to the gasification process. In order to achieve an efficient separation, a flocculation process is employed, in which microalgae primary particles aggregate and form larger clusters. Although flocculation is a well-established process, there are still some unknown issues related to it, that are worth further research. Experiments show that the mean size of clusters during flocculation goes through a maximum and then decreases with time. We refer to this pattern in the mean size profile as the overshoot. Studying this phenomenon is crucial since the size of clusters has a significant effect on the overall efficiency of the separation of microalgae from water. In this work, we aim at investigating the mechanisms behind the overshoot. The flocculation process is modeled as an aggregation-breakup system by using population balance equations (PBEs). The primary results show that the aggregation and breakup alone cannot lead to the overshoot in the mean size profile. Thus, we suggested three mechanisms that can lead to the overshoot: deposition of large clusters (DLC), restructuring of clusters (RC), and primary particle aggregation (PPA). These mechanisms were examined with numerical simulations and it was revealed that all three lead to the overshoot.
In monitoring the mean size of clusters during an aggregation-breakup process, it has been repeatedlyobserved that the mean size increases with time to a maximum and then decreases, resulting in an overshootin the means size prole. The decrease in the mean size prole has often explained as an eect of breakupof aggregates. In this work, we investigate three possible mechanisms that lead to an overshoot by usingnumerical analysis of population balance equations describing aggregation-breakup systems. The consideredmechanisms are deposition of large clusters, restructuring of clusters, and primary particle aggregation, wheretwo clusters aggregate only if one of them is a primary particle. The results show that all the suggestedmechanisms lead to an overshoot. It is only in the case of primary particle aggregation that the decrease inthe mean size is due to the dominance of breakup. Also, restructuring of fractal aggregates to form morecompact clusters explains the overshoot during aggregation-breakup with restructuring of clusters, while inthe case of deposition of large clusters, mass loss due to deposition of large aggregates leads to an overshoot.
Fuel cells are viable alternatives as power backup systems for mini-grids. In this work a case is considered, where the hydrogen fuel to the fuel cells is supplied from biomass gasification. However, the producer gas obtained from biomass gasification needs to be cleaned of impurities and contaminants. In this work we examined the superiority of the hot producer gas cleaning, which results in a better thermal efficiency since the heat loss from the system is reduced. In order to have a viable hot cleaning process, sulfur should be removed at 800°C and this was shown possible by promising primary data from the experiments where H 2 S was removed down to an acceptable level.
Sulfur, and in particular, H2S removal is of significant importance in gas cleaning processes in different applications, including biogas production and biomass gasification. H2S removal with metal oxides is one of the most viable alternatives to achieve deep desulfurization. This process is usually conducted in a packed bed configuration in order to provide a high solid surface area in contact with the gas stream per unit of volume. The operating temperature of the process could be as low as room temperature, which is the case in biogas production plants or as high as 900 °C suitable for gasification processes. Depending on the operating temperature and the cleaning requirement, different metal oxides can be used including oxides of Ca, Fe, Cu, Mn and Zn. In this review, the criteria for the design and scale-up of a packed bed units are reviewed and simple relations allowing for quick assessment of process designs and experimental data are presented. Furthermore, modeling methods for the numerical simulation of a packed bed adsorber are discussed.
Aggregation and breakup of small particles in stirred suspensions often shows an overshoot in the time evolution of the mean cluster size: Starting from a suspension of primary particles the mean cluster size first increases before going through a maximum beyond which a slow relaxation sets in. Such behavior was observed in various systems, including polymeric latices, inorganic colloids, asphaltenes, proteins, and, as shown by independent experiments in this work, in the flocculation of microalgae. This work aims at investigating possible mechanism to explain this phenomenon using detailed population balance modeling that incorporates refined rate models for aggregation and breakup of small particles in turbulence. Four mechanisms are considered: (1) restructuring, (2) decay of aggregate strength, (3) deposition of large clusters, and (4) primary particle aggregation where only aggregation events between clusters and primary particles are permitted. We show that all four mechanisms can lead to an overshoot in the mean size profile, while in contrast, aggregation and breakup alone lead to a monotonic, "S" shaped size evolution profile. In order to distinguish between the different mechanisms simple protocols based on variations of the shear rate during the aggregation-breakup process are proposed.
Models for gas-solid reactions in porous particles typically consist of a set of mass and energy balances in the form of conservation equations. For spherical or close to spherical particles, these equations are formulated in ID spherical coordinates. In the case where accumulation of gas inside the particle is significant, the balance equations contain convective terms. The present work presents a simple numerical scheme based on flux limited finite volume methods for discretizing conservation equations for convective and diffusive transport of mass and energy in radial direction in a porous sphere. The velocity is governed by Darcy's law coupled to an equation of state. The proposed scheme is applied to a series of test problems that admit full or partial analytical solutions. For the cases where only partial analytical solutions are available, a Comsol model is adopted for comparison. It is found that the scheme is able to resolve step gradients without generating oscillations and that it properly handles changes in the sign of the convective velocity. Applying the scheme for solving a common model for biomass pyrolysis reveals the importance of convective gas transport in the pyrolysis of thermally thick biomass particles.