Bitumen recovery by the water-based extraction process produces waste streams known as tailings. When discharged into the tailing ponds, the coarse solids in the tailings stream settle out quickly, while the fine solids accumulate over years of settling to a solids content of 30-35% by weight. The formed fluid fine solids sludge, known as mature fine tailings (MFT), traps 1-3% by weight hydrocarbons within its stable slurry structure. The remediation of these mature fine tailings is one of the major challenges facing the oil sands industry. This study was intended to investigate the recovery of residual hydrocarbons in the MFT by froth flotation process. Using a laboratory Denver flotation cell operated in a batch mode, the effect of MFT dilution ratio by process water or tap water, the flotation hydrodynamics and aeration rate on hydrocarbon recovery kinetics was studied. It was found that at 1:2 dilution by weight of the MFT with process water, increasing aeration rate has a more favourable effect on recovering more than 85% of the hydrocarbons from the MFT. The hydrocarbon-rich froth produced was treated by naphtha and was found to produce a hydrocarbon product similar to diluted bitumen obtained in bitumen extraction process, suitable for upgrading. Similar approach was applied to the hydrocarbon-rich tailings from the Tailings Solvent Recovery Unit of paraffinic froth treatment. Satisfactory recovery of hydrocarbons from the MFT was obtained using a flotation column operated in a continuous mode, confirmed the results obtained from the batch tests. The tailings produced from the continuous flotation experiments were treated with polymer flocculants such as Magnafloc-1011 and Al-PAM to study the effect of hydrocarbon recovery on the remediation of the MFT. The results from initial tests showed that both flocculants were not as effective on flocculating MFT solids following the recovery of hydrocarbons by froth flotation. / in Chemical Engineering

Esta tese investiga o efeito de dissipação de energia (energy input) sobre a cinética de flotação de apatita em meio básico (pH 10,5) utilizando oleato de sódio (40 mg/L e 80 mg/L) como coletor. Uma célula de flotação com agitação promovida por grades oscilantes (OGC - Oscilating Grid Flotation Cell), desenvolvida especialmente para avaliação do efeito de energia dissipada na flotação, foi adotada para o estudo devido à célula apresentar uma turbulência praticamente isotrópica, enquanto que as células mecânicas convencionais de flotação geram regiões com diferentes níveis de turbulência, visto que, próximo ao rotor a turbulência é muito maior que nas regiões mais afastadas. A agitação gerada pelo impelidor interfere em todos os sub-processos da flotação: colisão entre bolhas e partículas, adesão e destacamento (detachment). A macroturbulência na célula de flotação suspende as partículas no interior da célula, enquanto a microturbulência dispersa as bolhas e promove as colisões entre as bolhas e as partículas. Todos estes fenômenos podem influenciar a cinética do processo. Além disso, pelo fato das partículas muito grossas (+150µm) e finas (-44µm) de apatita apresentarem recuperações mais baixas do que uma faixa de tamanho ótimo, é importante estudar a influência da dissipação de energia na cinética dessas partículas. Os resultados mostraram que a cinética de flotação de apatita é fortemente influenciada pela dissipação de energia, pelo tamanho das partículas e pela concentração de coletor (ângulo de contato). As partículas finas (-44µm) exigem um elevado nível de energia (2 kW.m-3), devido ao aumento das frequências de colisão entre bolhas e partículas, esse aumento de energia não destruiu o agregado bolha/partícula para fração fina. Por outro lado, a cinética de flotação de partículas grossas (+150µm) foi fortemente afetada pelo aumento da energia transferida, os maiores valores da constante cinética para os grossos foram obtidos com o menor nível de dissipação de energia testado (0,1 kW.m-3), qualquer aumento na turbulência apresentou uma redução na taxa de flotação de no mínimo 40%, devido à destruição do agregado de partículas/bolha. O efeito da hidrofobicidade da apatita foi avaliado através de testes de flotação com duas concentrações de coletor (40 e 80g/L). A alta concentração de coletor (80g/L) apresentou resultados mais elevados de constante cinética para todos os níveis de dissipação de energia e para qualquer tamanho de partícula. Estes resultados evidenciam a influência da hidrofobicidade na adesão e na estabilidade do agregado partícula/bolha. / This thesis investigates the effect of energy dissipation (also called energy input) on the flotation kinetics of apatite in basic medium (pH10.5) with sodium oleate (40mg/L and 80mg/L) as collector. A special equipment called Oscillating Grid Flotation Cell (OGC) was adopted because the movement of the grids inside OGC creates a near isotropic turbulence, whereas the conventional mechanical flotation cells generate regions with different levels of turbulence: the intensity of turbulence is strongly higher close to the impeller and is very low far from it. The agitation created by the impeller influences all flotation sub process: particle/bubble collision, adhesion and detachment. Macroturbulence promotes the particle suspension inside the cell, whilst microturbulence disperses the bubbles and promotes collisions between bubbles and particles. All these phenomena influence the kinetics of the process. Moreover, because very coarse (+150µm) and fine (-44µm) particles of apatite show lower recovery than an optimum size range, it is important to study the influence of energy input on the kinetics of these sort of particles. The results showed that flotation kinetics of apatite is strongly influenced by the energy input, particle size and collector concentration (contact angle). Fine particles (-44µm) require high energy input (2 kW.m-3) probably to promote collision and the increase of energy input did not destroy the bubble/particle aggregate. Conversely, the flotation kinetics of coarse particles (+150µm) was strongly affected by energy input: 0,1 kW.m-3 promoted high flotation rate, but increasing the energy input to any higher level the flotation kinetics exhibits a clear decay, probably due to the destruction of the aggregate particle/bubble. The effect of hydrophobicity is evaluated testing two collector concentrations. With high collector concentration the flotation rate increased for all levels of energy input, these results evidenced the influence of hydrophobicity on the attachment and stability of the aggregate particle/bubble.

Flotation is one part of LKAB's refining process where minerals are separated from iron ore in cascade connected tanks. A good level control in the flotation tanks is vital to get a high recovery of the iron. Nominal and robust performance of different level controllers are compared in the study. Robust performance is analyzed for uncertainties in valve contants and for hysteresis in the valves. The results show that LQG control give better nominal performance than PI control. The robust performance is worse for the LQG controller than the PI controller for modelled uncertainties in valve contants. The robustness analysis shows also that LQG control is less sensitive to hysteresis in the valves compared to PI control. Thus, type of uncertainty affects different control strategies differently. / Flotation är en del av LKAB:s förädlingsprocess där mineraler separeras från järnmalmen i seriekopplade tankar. En god nivåhållning i flotationstankarna är viktigt för att erhålla ett högt järnutbyte. Olika reglerstrategiers nominella och robusta prestanda jämförs i studien. Robust prestanda undersöks vid osäkerhet i ventilkonstanter och hysteres i ventiler. Resultaten visar att LQG-reglering ger bättre nominell prestanda än PI-reglering. Den robusta prestandan vid osäkerhet i ventilkonstanter är sämre för LQG-reglering jämfört med PI-reglering. Robusthetsanalysen visar även att LQG-regulatorn är mindre känslig för hysteres i ventiler än PI-regulatorn. Typen av osäkerhet inverkar sålunda olika på olika reglerstrategier.

A first-principle flotation model was derived at Virginia Tech from the basic mechanisms involved in the bubble-particle and bubble-bubble interactions occurring in a flotation cell (Yoon and Mao, 1996; Sherrell and Yoon, 2005; Do, H, 2010). The model consists of a series of analytical equations for bubble generation, bubble-particle collision, attachment, detachment, and froth phase recovery. The process of bubble-particle attachment has been modelled on the premise that bubble-particle attachment occurs when the disjoining pressure of the thin liquid in a wetting films formed between particle and bubble is negative, as was first suggested by Laskowski and Kitchener (1969). These provisions allow for the flotation model to incorporate various chemistry parameters such as zeta-potentials, contact angles, surface tension in addition to the physical and hydrodynamic parameters such as particle size, bubble size, and energy dissipation rate. In the present work, the effects of both hydrodynamic and chemistry parameters have been studied using the model-based computer simulator. A series of laboratory batch flotation experiments carried out on mono-sized glass beads validated the simulation results. The flotation feeds were characterized in terms of particle size, contact angle, and Hamaker constant, and the flotation experiments were conducted at different energy dissipation rates, gas rates, froth heights. The flotation tests were also carried out on mixtures of hydrophobic silica and hydrophilic magnetite particles, so that the grades of the flotation products can be readily determined by magnetic separation. The experimental results are in good agreement with the model predictions both in terms of grade and recovery. / Master of Science

Carla Abarca Ph.D. Thesis / The selective separation of valuable minerals by froth flotation is a critical unit operation in mineral processing. Froth flotation is based on the ability of chemical reagents, called collectors, to selectively lower the surface energy of valuable mineral particles, facilitating attachment of the modified mineral particles to air bubbles in the flotation cell. The mineral laden bubbles rise to the surface forming a froth phase that can be isolated. Novel cationic polystyrene nanoparticle collectors have been developed recently to be used as effective flotation collectors, aiming to recover challenging nickel sulfide ores that respond poorly to conventional molecular flotation collectors. However, optimizing nanoparticle flotation collectors is a challenge. An effective nanoparticle collector candidate should meet three requirements: (1) it should be colloidally stable in the flotation media; (2) it should be hydrophobic enough to change the mineral surface and induce an air bubble-mineral particle attachment; and (3) specifically and strongly bind to metal-rich minerals. Producing nanoparticles that are simultaneously colloidally stable and sufficiently hydrophobic presents a problematic task. Thus, a delicate balance of nanoparticle properties is required for commercially viable nanoparticle collectors. This thesis presents a promising approach for discovering and characterizing novel nanoparticle collectors by using high throughput screening techniques. Developed was a workflow for fast fabrication and testing of nanoparticle candidates, including: (1) parallel production of large nanoparticle libraries covering a range of surface chemistries, (2) a high throughput colloidal stability assay to determine whether a nanoparticle type is stable in flotation conditions; (3) an automated contact angle assay to reject nanoparticles that are not hydrophobic enough to induce efficient bubble-particle attachment, and; (4) a laboratory flotation test in sodium carbonate (pH~10) with the best nanoparticle candidates. The automated colloidal stability assay was based on the optical characterization of diluted nanoparticle dispersions in multiwell plates, yielding critical coagulation concentrations (CCCs) of sodium carbonate. To pass this screening test, the CCC of candidate nanoparticles must be greater than the effective carbonate concentration in commercial flotation cells. Since the nanoparticle size affects the intrinsic light scattering properties of the nanoparticles, two routes were developed. The colloid stability assay was suitable for nanoparticles ranging between 50 nm and 500 nm, since nanoparticle size. The automated contact angle assay used a miniature 16-well plate format where flat glass slides were exposed to 200 μL nanoparticle dispersions. The cationic nanoparticles formed a saturated adsorbed monolayer on the glass, and after rinsing and drying, the water contact angle was automatically measured. Effective nanoparticle candidates had contact angles greater than 50 degrees, a criterion developed with model experiments. During the development of the automated workflow platform, a series of nanoparticles with methyl-ended PEG-methacrylate monomers were prepared. Although the PEG chains greatly enhanced colloidal stability, the particles were too hydrophilic to be effective collectors. Interestingly, nanoparticles with long PEG chains acted as froth modifiers, giving wetter and more robust foams as well as increased entrainment of materials that did not adhere to bubbles. Conventional laboratory scale latex synthesis methodologies are far too inefficient to generate large library of candidate nanoparticles. Instead, we started with a few parent nanoparticle types and then used Click chemistry to generate a large range of surface chemistries. Specifically copper-mediated azide alkyne cycloaddition reaction was used to functionalize the surface of azide nanoparticles with different chemical groups, ranging from hydrophilic amine-terminated PEG chains, to hydrophobic hexane-terminated materials. The Click library exhibited an extensive range of critical coagulation concentrations and contact angle values. For example, for a given parent azide nanoparticle, the contact angles ranged from 62 to 101 degrees, depending upon the density and type of click reagent. A novel paper chromatographic method was developed for the quantitative determination surface azide. This assay was critical for determining the surface density of functional groups from the click reactions. Overall, high throughput screening techniques were designed and applied to the development of nanoparticle collectors for froth flotation. Automated screening assays of critical coagulation concentration and contact angle proved to be effective in obtaining flotation domain maps, and finding the most promising nanoparticle collectors for froth flotation. I believe the work in this thesis is one of the first reported uses of high throughput methodologies for the development of mineral flotation reagents. / Thesis / Doctor of Philosophy (PhD) / Novel cationic polystyrene nanoparticle collectors have been developed to be used as effective flotation collectors, aiming to recover challenging nickel sulfide ores that respond poorly to conventional molecular flotation collectors. However, optimizing nanoparticle flotation collectors is a challenge. This thesis presents a promising approach for discovering and characterizing novel nanoparticle collectors by using high throughput screening techniques and click chemistry. Development of nanoparticle libraries and automated screening assays of critical coagulation concentration and contact angle proved to be effective in obtaining flotation domain maps, and finding the most promising nanoparticle collectors for froth flotation.

Flotation is regarded as the best available separation method for the recovery of valuable minerals such as chalcopyrite (CuFeS2), sphalerite (ZnS), etc., from mined ores. Practically all metals humans use today are produced by flotation. The process relies on controlling the stability of the thin liquid films (TLFs) of water formed between minerals and air bubbles (wetting film), air bubbles (foam film), and mineral particles (colloid films). In flotation, a desired mineral is rendered hydrophobic by surfactant coating as a means to destabilize the TLFs, so that they can be attached to the hydrophobic air bubbles. A TLF ruptures when the disjoining pressure (or surface forces per unit area) of the film becomes negative, i.e., Π < 0. Thermodynamically, a wetting film can rupture when the contact angle (θ) of a mineral surface is larger than zero. It would, therefore, be reasonable to consider the roles of the surface forces to better understand the fundamental mechanisms involved in flotation. The surface forces considered in the present work included the electric double layer (EDL), van der Waals (vdW), and attractive hydrophobic (HP) forces. A flotation model has been developed by using the hydrophobic force as a kinetic parameter, which made it possible to track the fates of mineral particles of different of size, surface liberation, and contact angle to predict both recovery and grades for the first time. The model has been validated against the plant survey data obtained from an operating copper flotation plant. The simulation results obtained using the first principles model have been utilized to address the limitations of current flotation practices. One such limitation is the presence of slow-floating target minerals present in the cleaner-scavenger tails (CST) that are routinely recycled back to the rougher flotation bank as circulating loads (CLs) to allow longer retention times for the slow-floating particles for additional recovery. The simulation results show also that opening a flotation circuit by treating the CST streams separately in an advanced circuit can substantially improve the plant performance. One of the major limitations of flotation is that the coarse particles in a feed stream are difficult to recover due to the low hydrophobicity associated with poor surface liberation. A new flotation model developed in the present work suggests various ways to address the problem. One is to increase the hydrophobicity of the composite (poorly liberated) particles using the Super Collectors that can increase the contact angles to 150 -170o. Simulation results obtained using the model developed in the present work show significant financial benefits of using Super Collectors. Flotation is controlled by surface forces as noted above. As particle size becomes larger than 150 µm, however, the gravitational force comes into the picture and can override the surface forces. A new flotation cell has been developed to mitigate the effects of the extraneous force by decreasing the effective specific gravity (SG) by attaching air bubbles to facilitate levitation and by creating a pulsation to allow particles to move according to SGs independent of particle size, which should help increase the upper particle size limit of flotation. Surface forces in foam and oil-in-water emulsion films have been measured at different temperatures to determine the changes in thermodynamic properties of the thin liquid films (TLFs) of water confined between two bubbles and two oil drops. The results show that the films are destabilized by the attractive hydrophobic forces created during the course of building H-bonded structures in confined spaces, which entails decreases in enthalpy (H < 0) and entropy (TS < 0), the second term representing the thermodynamic cost of building the structures. / Doctor of Philosophy / Flotation is a kinetic process designed to separate valuable minerals from mined ores. This process depends on several hydrodynamic and surface chemistry parameters making it hard to model. A U.S. patent was awarded to Sulman and Picard in 1905 for using air bubbles to selectively collect hydrophobic particles from the aqueous phase, leaving hydrophilic particles behind. Since then, the separation process known as flotation has been used to produce practically all metals humans use. Many investigators developed flotation models using hydrodynamic parameters, e.g., particle size, bubble size, energy dissipation rate, etc., but without a reference to particle hydrophobicity. Therefore, the models were successful in predicting recoveries but not product grades. Derjaguin and Dukhin (1961) were the first to model flotation using surface forces but without due consideration of the role of hydrophobic force in flotation. Therefore, it also failed to predict product grades. In the current work, a new flotation model has been developed using the hydrophobic force as a kinetic parameter. This approach made it possible to predict both recoveries and grades for the first time. The model has been reduced to a simple form mimicking the Arrhenius equation so that it can be used to delineate the different conditions required for optimizing coarse and fine particle flotation. The model has been derived by considering the surface forces in the thin liquid films (TLFs) of water confined between bubbles, and bubbles and particles. It has been found that the hydrophobic force plays a decisive role in destabilizing a wetting film and inducing bubble-particle attachment. The surface forces measured in the present work show that the hydrophobic interactions in macroscopic scales are controlled by enthalpy rather than entropy, which is contrary to the nanoscale hydrophobic interactions. The model has been validated against a full-scale plant operation and demonstrated predictive capabilities. The simulation results have been analyzed to determine the limitations of the current flotation practices. It was found that coarse particle flotation is difficult either due to the presence of composite particles reducing the particle contact angle or due to their poor hydrodynamic properties. Utilizing the insights from the model, various methods of alleviating these limitations have been developed and presented in the current work. References Derjaguin, B.V., Dukhin, S.S., 1961. Theory of flotation of small and medium-size particles. Inst. Min. Metall. 241–267. Sulman, H.L., and Kirkpatrick-Picard (1905). U.S. Patent No. 793,808.

Flotation is a process of separating particulate materials of different surface properties in a hydrodynamic environment, and is used extensively for separating different minerals from each other in the mining industry. In this process, air bubbles are introduced at the bottom of a particulate suspension (pulp), so that bubbles coated with hydrophobic particles rise to the top and form a froth phase while hydrophobic particles stay in suspension. The selectivity of the flotation process is determined by the hydrophobicity of the particulate materials involved, while the kinetics of the process is controlled by the hydrodynamic conditions and the disjoining pressures in the thin aqueous films between air bubbles and particles. In the present work, a mathematical model for the flotation process has been developed by considering both the hydrodynamic and surface chemical parameters. The model can describe the events occurring in both the pulp and froth phases of a mechanically-agitated flotation cell. The pulp-phase model is based on predicting the kinetics of bubble-particle attachment using the DLVO extended to include contributions from hydrophobic force and the theory of turbulent collision. The froth-phase model is based on predicting the rate of bubble-particle detachment by considering bubble coarsening and water recovery. The predictions from the overall flotation model are in general agreement with the results obtained in single-bubble flotation experiments and the flotation test results reported in literature. Since the model has been developed largely from first principles, it has predictive and diagnostic capabilities. / Ph. D.

Froth floatation is an important process used in the mining industry for separating minerals from each other. The separation process is based on rendering a selected mineral hydrophobic using an appropriate hydrophobizing reagent (collector), so that it can selectively attach onto the surfaces of a rising stream of air bubbles. Thus, controlling the hydrophobicity of the minerals to be separated from each other is of critical importance in flotation. If one wishes to separate plastics from each other by flotation, however, it would be necessary to render a selected plastic hydrophilic and leave the others hydrophobic. In the present work, the possibility of separating common plastics from each other by flotation has been explored. While water contact angle is the most widely used measure of the hydrophobicity of a solid, it does not give the information on the kinetics of flotation. Therefore, the forces acting between the surfaces coated with different hydrophobic polymers (or plastics) in water were measured using the Atomic Force Microscope (AFM). The results obtained with polystyrene, polymethylmethacryrate (PMMA), polypropylene (PP), and Teflon showed the existence of long-range attractive forces (or hydrophobic force) that cannot be explained by the classical DLVO theory. The surface force measurements were conducted in pure water and in solutions of surfactant (alkyltrimethylammonium chloride) and a salt (NaCl). In pure water, the attractive forces were much stronger than van der Waals force. In the presence of the surfactant and NaCl, the long-range attraction decreased with increasing concentration and the alkyl chain length. A series of contact angle measurements were conducted to determine the hydrophobicity of polystyrene (PS), polyvinyl chlorite (PVC), and polymethylmethacrylate (PMMA) in the presence of different wetting agents (surfactants). The results show the possibility of separating plastics from each other by flotation, and a series of microflotation tests conducted on PS and PVC showed promising results. / Master of Science

O presente trabalho teve como objetivo o estudo da potencialidade de aplicação da flotação por ar dissolvido (FAD) lamelar (de alta taxa) e convencional como parte final de sistema de tratamento sequencial de esgoto sanitário, constituído de reatores anaeróbios de manta de lodo (UASB) seguidos de tanque de aeração. O trabalho, desenvolvido na Estação de Tratamento de Esgotos Jardim das Flores da cidade de Rio Claro, foi dividido em três etapas. Na primeira etapa, realizada com equipamento Flotateste, foram investigados os valores mais adequados de dosagens de polímero sintético catiônico, amido de araruta, cloreto férrico e associações entre cloreto férrico e polímero ou amido de araruta, de razão ar/sólidos (A/S), bem como de tempo e gradiente de mistura rápida e de floculação. Nessa etapa, os melhores desempenhos da FAD foram obtidos com o emprego de polímero catiônico, tanto isoladamente quanto associado ao cloreto férrico, sendo esta última condição imprescindível para maior remoção de fósforo. A flotação de liquor misto utilizando apenas o cloreto férrico como coagulante, apresentou desempenho extremamente insatisfatório, com a formação de flocos com características ruins de flotabilidade. Além disso, observou-se que a flotação desse tipo de efluente requisitou condições mais amenas de mistura, sendo que, em muitos casos, os melhores resultados foram obtidos mediante a eliminação da mistura rápida ou da floculação. Nas segunda e terceira etapas, realizadas com unidades piloto de flotação lamelar de alta taxa (FADAT) e convencional, respectivamente, e empregando apenas polímero como auxiliar de floculação/flotação (dosagem em torno de 1 g de polímero/kg de SST), foram investigados os valores mais adequados de Taxa de Aplicação Superficial (TAS), Taxa de Aplicação de Sólidos (TS) e quantidade de ar fornecida à flotação, calculada através da razão A/S ou da dosagem de ar (Dar), em g de ar/'M POT.3' de esgoto afluente, dependendo da concentração de sólidos suspensos totais afluentes à flotação (SSTafl). Através dos resultados dos ensaios pôde-se observar que a razão A/S é mais adequada para o cálculo da quantidade de ar na flotação de suspensões concentradas (SSTafl acima de 1100 mg/L no presente estudo), ao passo que Dar é mais representativo no caso da flotação de suspensões diluídas. Empregando a unidade FADAT, alimentada com liquor misto de tanque de aeração em boas condições de biofloculação, foi possível obter resultados satisfatórios para TAS de até 350 'M POT.3'/'M POT.2'.dia e TS entre 120 e 250 kg de SST/'M POT.2'.dia, desde que fornecida quantidade adequada de ar ao processo (Dar na faixa de 15,0 a 18,0 g de ar/'M POT.3' de afluente), para SSTafl entre 800 e 1100 mg/L. Na flotação lamelar de alta taxa, constatou-se que o bom desempenho esteve diretamente relacionado a baixos valores de SSTafl. As recomendações para a flotação com unidade de FAD convencional tiveram que ser mais conservadoras, devido às condições ruins de biofloculação apresentadas pelo liquor misto do tanque de aeração durante a realização da terceira etapa do trabalho. Nesses ensaios, os resultados mais satisfatórios foram obtidos com o emprego de TAS em torno de 200 'M POT.3'/'M POT.2'.dia, TS menor que 200 kg de SST/'M POT.2'.dia e A/S em torno de 0,020 (Dar de 21,0 g de ar/'M POT.3' de afluente), para SSTafl em torno de 1400 mg/L / The aim of this research was the study of the lamella design (high rate) and conventional dissolved air flotation (DAF) process potentiality of application as the final part of a domestic sewage sequential treatment system made up anaerobic reactors (UASB) followed by aeration tank. The work, developed at the Jardim das Flores WWTP of Rio Claro city, was separated in three phases. In the first phase, using the flotatest apparatus, the most suitable values of chemical (synthetic cationic polymer, arrow-root starch, ferric chloride and ferric chloride combined with polymer or arrow-root starch) dosages, air to solids ratio (A/S), as well as rapid mix and flocculation time and mean velocity gradient values, were investigated. In this phase, the best DAF performances were obtained with the application of cationic polymer, alone and associated to ferric chloride. The use of ferric chloride was essential to obtain better phosphorus removal results. The mixed liquor flotation using ferric chloride alone as primary coagulant presented extremely unsatisfactory performance, with the formation of flocks showing bad characteristics of flotability. Moreover, it was observed that flotation of this kind of effluent had requested gentler mixture conditions. In many cases, the best results were obtained by means of rapid mix or flocculation step elimination. In the second and third phases, using the lamella design and conventional flotation units respectively and cationic polymer as the only flocculation/flotation aid (dosage around 1 g of polymer/kg TSS), the most suitable values of Overflow Rate (OR), Solids Application Rate (SAR) and amount of air supplied to the flotation process, expressed as the A/S ratio or air dosage (AD, in g of air/'M POT.3' of influent) depending on influent total suspended solids concentration (TSSinfl), were investigated. The flotation essays results showed that the A/S ratio is more suitable for the air supply estimation in flotation of concentrated suspensions (TSSinfl above 1100 mg/L in this study). Contrarily, the parameter AD is more representative in cases of flotation of diluted suspensions. Using the lamella design unit, fed with aeration tank mixed liquor presenting good bio-flocculation conditions, it was possible to obtain adequate results applying OR as high as 350 'M POT.3'/('M POT.2'.day) combined with SAR values between 120 and 250 kg of TSS/('M POT.2'.day), since the adequate amount of air has been provided for the process (AD values between 15,0 and 18,0 g of air/'M POT.3' of influent), for TSSinfl between 800 and 1100 mg/L. For the lamella flotation, it was noticed that the best performance has been directly associated to lower values of TSSinfl. The recommendations for flotation using the conventional DAF unit had to be more conservative, taking into account the poor bio-flocculation conditions presented by the mixed liquor during the third phase of work. In theses essays, the more adequate results were obtained by applying OR around 200 'M POT.3'/('M POT.2'.day), SAR under 200 kg of TSS/('M POT.2'.day) and A/S ratio around 0,020 (AD of 21,0 g of air/'M POT.3' of influent), for TSSinfl around 1400 mg/L

O presente trabalho teve como objetivo o estudo da potencialidade de aplicação da flotação por ar dissolvido (FAD) lamelar (de alta taxa) e convencional como parte final de sistema de tratamento sequencial de esgoto sanitário, constituído de reatores anaeróbios de manta de lodo (UASB) seguidos de tanque de aeração. O trabalho, desenvolvido na Estação de Tratamento de Esgotos Jardim das Flores da cidade de Rio Claro, foi dividido em três etapas. Na primeira etapa, realizada com equipamento Flotateste, foram investigados os valores mais adequados de dosagens de polímero sintético catiônico, amido de araruta, cloreto férrico e associações entre cloreto férrico e polímero ou amido de araruta, de razão ar/sólidos (A/S), bem como de tempo e gradiente de mistura rápida e de floculação. Nessa etapa, os melhores desempenhos da FAD foram obtidos com o emprego de polímero catiônico, tanto isoladamente quanto associado ao cloreto férrico, sendo esta última condição imprescindível para maior remoção de fósforo. A flotação de liquor misto utilizando apenas o cloreto férrico como coagulante, apresentou desempenho extremamente insatisfatório, com a formação de flocos com características ruins de flotabilidade. Além disso, observou-se que a flotação desse tipo de efluente requisitou condições mais amenas de mistura, sendo que, em muitos casos, os melhores resultados foram obtidos mediante a eliminação da mistura rápida ou da floculação. Nas segunda e terceira etapas, realizadas com unidades piloto de flotação lamelar de alta taxa (FADAT) e convencional, respectivamente, e empregando apenas polímero como auxiliar de floculação/flotação (dosagem em torno de 1 g de polímero/kg de SST), foram investigados os valores mais adequados de Taxa de Aplicação Superficial (TAS), Taxa de Aplicação de Sólidos (TS) e quantidade de ar fornecida à flotação, calculada através da razão A/S ou da dosagem de ar (Dar), em g de ar/'M POT.3' de esgoto afluente, dependendo da concentração de sólidos suspensos totais afluentes à flotação (SSTafl). Através dos resultados dos ensaios pôde-se observar que a razão A/S é mais adequada para o cálculo da quantidade de ar na flotação de suspensões concentradas (SSTafl acima de 1100 mg/L no presente estudo), ao passo que Dar é mais representativo no caso da flotação de suspensões diluídas. Empregando a unidade FADAT, alimentada com liquor misto de tanque de aeração em boas condições de biofloculação, foi possível obter resultados satisfatórios para TAS de até 350 'M POT.3'/'M POT.2'.dia e TS entre 120 e 250 kg de SST/'M POT.2'.dia, desde que fornecida quantidade adequada de ar ao processo (Dar na faixa de 15,0 a 18,0 g de ar/'M POT.3' de afluente), para SSTafl entre 800 e 1100 mg/L. Na flotação lamelar de alta taxa, constatou-se que o bom desempenho esteve diretamente relacionado a baixos valores de SSTafl. As recomendações para a flotação com unidade de FAD convencional tiveram que ser mais conservadoras, devido às condições ruins de biofloculação apresentadas pelo liquor misto do tanque de aeração durante a realização da terceira etapa do trabalho. Nesses ensaios, os resultados mais satisfatórios foram obtidos com o emprego de TAS em torno de 200 'M POT.3'/'M POT.2'.dia, TS menor que 200 kg de SST/'M POT.2'.dia e A/S em torno de 0,020 (Dar de 21,0 g de ar/'M POT.3' de afluente), para SSTafl em torno de 1400 mg/L / The aim of this research was the study of the lamella design (high rate) and conventional dissolved air flotation (DAF) process potentiality of application as the final part of a domestic sewage sequential treatment system made up anaerobic reactors (UASB) followed by aeration tank. The work, developed at the Jardim das Flores WWTP of Rio Claro city, was separated in three phases. In the first phase, using the flotatest apparatus, the most suitable values of chemical (synthetic cationic polymer, arrow-root starch, ferric chloride and ferric chloride combined with polymer or arrow-root starch) dosages, air to solids ratio (A/S), as well as rapid mix and flocculation time and mean velocity gradient values, were investigated. In this phase, the best DAF performances were obtained with the application of cationic polymer, alone and associated to ferric chloride. The use of ferric chloride was essential to obtain better phosphorus removal results. The mixed liquor flotation using ferric chloride alone as primary coagulant presented extremely unsatisfactory performance, with the formation of flocks showing bad characteristics of flotability. Moreover, it was observed that flotation of this kind of effluent had requested gentler mixture conditions. In many cases, the best results were obtained by means of rapid mix or flocculation step elimination. In the second and third phases, using the lamella design and conventional flotation units respectively and cationic polymer as the only flocculation/flotation aid (dosage around 1 g of polymer/kg TSS), the most suitable values of Overflow Rate (OR), Solids Application Rate (SAR) and amount of air supplied to the flotation process, expressed as the A/S ratio or air dosage (AD, in g of air/'M POT.3' of influent) depending on influent total suspended solids concentration (TSSinfl), were investigated. The flotation essays results showed that the A/S ratio is more suitable for the air supply estimation in flotation of concentrated suspensions (TSSinfl above 1100 mg/L in this study). Contrarily, the parameter AD is more representative in cases of flotation of diluted suspensions. Using the lamella design unit, fed with aeration tank mixed liquor presenting good bio-flocculation conditions, it was possible to obtain adequate results applying OR as high as 350 'M POT.3'/('M POT.2'.day) combined with SAR values between 120 and 250 kg of TSS/('M POT.2'.day), since the adequate amount of air has been provided for the process (AD values between 15,0 and 18,0 g of air/'M POT.3' of influent), for TSSinfl between 800 and 1100 mg/L. For the lamella flotation, it was noticed that the best performance has been directly associated to lower values of TSSinfl. The recommendations for flotation using the conventional DAF unit had to be more conservative, taking into account the poor bio-flocculation conditions presented by the mixed liquor during the third phase of work. In theses essays, the more adequate results were obtained by applying OR around 200 'M POT.3'/('M POT.2'.day), SAR under 200 kg of TSS/('M POT.2'.day) and A/S ratio around 0,020 (AD of 21,0 g of air/'M POT.3' of influent), for TSSinfl around 1400 mg/L