Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Bench-scale two-dimensional fluidized bed hydrodynamics and struvite growth studies Qu, Xiaocao 2007

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2008_spring_qu_xiaocao.pdf [ 21.05MB ]
Metadata
JSON: 24-1.0063064.json
JSON-LD: 24-1.0063064-ld.json
RDF/XML (Pretty): 24-1.0063064-rdf.xml
RDF/JSON: 24-1.0063064-rdf.json
Turtle: 24-1.0063064-turtle.txt
N-Triples: 24-1.0063064-rdf-ntriples.txt
Original Record: 24-1.0063064-source.json
Full Text
24-1.0063064-fulltext.txt
Citation
24-1.0063064.ris

Full Text

BENCH-SCALE TWO-DIMENSIONAL FLUIDIZED BED HYDRODYNAMICS AND STRUVITE GROWTH STUDIES  by  Xiaocao Qu B.Sc., Peking University, P.R. China, 2003  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Civil Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA  December 2007  © Xiaocao Qu, 2007  ABSTRACT A bench-scale, two-dimensional multi-compartmentalized fluidized reactor was designed and studies of hydrodynamic behavior of fluidization of struvite pellets were performed. Also size growth distribution tests were analyzed qualitatively. The study validated a previously-proposed theory, concerning the relationship between dynamic pressure drop and upflow velocity as well the experimental protocol to determine the minimum fluidization velocity. Findings indicated that the mixture of twosized particles would behave rather independently of each other, before the bed expansion. It was suggested that bed height measurement could be another promising method to pinpoint minimum fluidization velocity as there is a sharp bed surface "waking episode" during the process of a packed bed being gradually fluidized. Bed expansion equations for the prediction of void fraction as a function of superficial upflow velocity or vice versa, have been established for 4 groups of monosize particles, as well as two mixtures of two-sized particles. The equation constants did not agree well with previously established ones. The two layers of segregated mixture bed had congruent linear relationships between the logarithmic void fraction and logarithmic upflow velocity. It was found that a mixture does not always go through segregation, but only when the size difference is large enough. Size growth distribution tests were performed under different hydrodynamic configurations as well as seeding conditions. Conclusions can be made that a bed with uniformly- distributed particle void fractions and higher mixing energy input (upflow velocity), normally has better performance of struvite growth in size.  ii  TABLE OF CONTENTS ABSTRACT ^  ii  LIST OF TABLES ^ LIST OF FIGURES ^ ACKNOWLEDGEMENTS ^ CHAPTER 1 - INTRODUCTION ^  vi viii 1  1.1 Driving force for phosphorus recovery ^  1  1.2 Motives for a hydrodynamic study ^  2  CHAPTER 2 - RESEARCH OBJECTIVES ^  4  CHAPTER 3 - BACKGROUND AND BRIEF LITERATURE REVIEW ^ 5 3.1 Liquid fluidization ^  5  3.1.1 Hydrodynamic representation ^  6  3.1.2 Minimum fluidization ^  7  3.1.3 Bed expansion ^  7  3.1.4 Segregation ^  8  3.1.5 Fluidized-bed reactors ^  9  3.2 Crystallization ^  10  3.2.1 Struvite growth mechanism ^  10  3.2.2 Growth rate ^  11  CHAPTER 4 - MATERIALS AND METHODS ^  13  4.1 Reactor design ^  13  4.2 Hydrodynamic phase ^  15  4.3 Struvite growth tests ^  18  iii  4.3.1 System configuration ^  18  4.3.2 Sampling methodology ^  18  4.3.3 Flow rate measurement ^  19  CHAPTER 5 - RESULTS AND DISCUSSION ^ 5.1 Hydrodynamics ^  20 21  5.1.1 Minimum fluidization ^  21  5.1.2 Bed expansion ^  30  5.2 Size growth distribution ^  38  5.3 Application of research findings to a struvite production and recovery system: practical considerations ^ CHAPTER 6 - CONCLUSIONS ^  44 45  CHAPTER 7 — RECOMMENDATIONS FOR FURTHER RESEARCH ^ 47 REFERENCES ^  49  APPENDIX A: HYDRODYNAMIC RUNS ^  52  APPENDIX B: SIZE GROWTH TESTS ^  60  iv  LIST OF TABLES Table 4.1 The dimensions of the reactor shown in Figure 4.1 ^ 15 Table 5.1 Specific equation coefficients derived and particle properties ^ 24  v  LIST OF FIGURES Figure 4.1 Two-dimensional reactor and struvite recovery configurations. ^ 14 Figure 5.1 Dynamic pressure drop as a function of upflow velocity for dp=4mm solids 22 Figure 5.2 Dynamic pressure drop as a function of upflow velocity for dp=2.9mm solids ^  22  Figure 5.3 Dynamic pressure drop as a function of upflow velocity for dp=2mm solids 23 Figure 5.4 Dynamic pressure drop as a function of upflow velocity for dp=0.9mm solids 23 Figure 5.5 dynamic pressure drop as a function of the upflow velocity for the mixture of dp=4mm and dp=2mm particles. ^  26  Figure 5.6 dynamic pressure drop as a function of upflow velocity for the mixture of 2mm and 0.9mm. ^  27  Figure 5.7 Bed height change over the increase of upflow velocity for dp=4mm solids . 29 Figure 5.8 Bed height as a function of the upflow velocity for the two separate layers respectively of the binary mixture bed of dp=4mm and dp=2mm solids. ^ 26 Figure 5.9 Bed expansion linear relationship between upflow velocity (log) and void fraction (log) for dp=4mm solids ^  31  Figure 5.10 Bed expansion linear relationship between upflow velocity (log) and void fraction (log) for dp=2.9mm solids. ^  32  Figure 5.11 Bed expansion linear relationship between upflow velocity (log) and void fraction(log) for dp=2mm solids ^  32  Figure 5.12 Bed expansion linear relationship between upflow velocity (log) and void fraction(log) for dp=0.9mm solids. ^  33  Figure 5.13 Void fraction(log) vs. upflow velocity(log) for the top and bottom layers of the segregated bed of mixtures of dp=4mm and dp=2mm solids. ^ 34 vi  Figure 5.14 Void fraction (log) vs. upflow velocity (log) for the completely mixed bed and the segregated bed with the same mixture of dp=4mm and dp=2mm solids. ^ 35 Figure 5.15 Size growth within 12 hours for upflow velocity =0.021m/s, recycle ratio=17.9, seeding 4.71g particles of size < 0.119mm. ^  41  Figure 5.16 Size growth within 12 hours for upflow velocity =0.023m/s, recycle ratio=20, seeded with 9.22g particles of size < 1.19mm. ^  42  Figure 5.17 Size growth within 12 hours for upflow velocity =0.030m/s, recycle ratio=21, seeded with 15g particles of size < 0.149mm. ^  43  vii  ACKNOWLEDGEMENTS I would like to acknowledge the assistance, support and encouragement that have been provided by the following people: •  I would first like to thank my supervisor, Dr. Don Mavinic, for his technical guidance, funding and understanding throughout this research.  •  Dr.Victor Lo, for the funding and all the understanding and encouragement.  •  Fred Koch, for his continuous and irreplaceable knowledge and experience input, his effective technical ideas and his supervision on reactor design as well as experimental set up.  •  Md. Saifur Rahaman, the doctoral student in the struvite research group, for his crucial and original inspiration on the project, for his major contribution on the research blueprint and his encouragements.  •  Susan Harper, for her help and instruction on lab work preparations.  •  Paula Parkinson, for testing the Phosphorus and Ammonium samples for me and all the in-time handy assistance on equipments and chemical regents.  •  Melissa Ying Zhang, for her friendly support and valuable input during a number of discussions.  •  Wayne Lo, the doctoral student on the neighbouring bench, for all his help during experiments crisis.  •  Bobby Chan, for drawing my reactor demonstrations and for being such a supportive friend.  •  My parents, for their unwavering support and love.  viii  INTRODUCTION  CHAPTER 1 - INTRODUCTION 1.1 Driving force for phosphorus recovery Removal of excess phosphorus and nitrogen is required to prevent eutrophication problems in lakes and enclosed water systems. However, the efficiency of biological phosphorus removal processes can be inadequate to meet discharge standards, while chemical phosphate precipitation is costly itself and causes p-content sludge problems. (Seckler et al., 1996; Battistonia et al., 2002) Struvite (MgNH4 PO 4 • 6H20, MAP, commonly known as struvite) is a hard, crystalline mineral that accumulates onto pipe surfaces in contact with struvite-supersaturated process fluids (Ohlinger et al., 2000). Struvite grown on the walls of the pipes in anaerobic digestion systems have been reported (Yoshino et al., 2003) as a costly problem in the wastewater treatment industry. Struvite accumulation can result in diminished pipe flow capacity, blinding and fouling of mechanical equipment, and premature equipment failure. Remediation of struvite encrusted piping and equipment is costly and rather ineffective. The resulting financial consequences of struvite encrustation include increased operations and maintenance costs for energy, labor, and diminished capacity as well as increased capital costs for capacity oversizing and replacement of prematurely failed systems (Ohlinger et al., 2000). The cost of the disposal of additional phosphate sludge, the need to limit phosphate emission to the aquatic environments, as well as the impending resource shortage due to the rapidly depleting phosphorus rock reserves (Steen 1998), have diverted the attention to phosphorus recovery; this has led to the production of a recyclable material in the form of phosphatereleasing pellets. Fluidized-bed reactor (FBR) crystallization of struvite as an efficient approach to recover phosphorus, by producing MAP, has been recently developed in the  1  INTRODUCTION  Netherlands, Italy, Japan and most recently in Canada (Seckler e et al., 1996; Battistoni 1998; Adnan 2002; Yoshino 2003).  1.2 Motives for a hydrodynamic study The struvite recovery reactor design has rarely been based on any theoretical or even empirical knowledge of fluidization, but rather randomly according to the designer's "feelings" (Fred Koch, Research Associate, Department of Civil Engineering, University of British Columbia, pers.com .) A reactor design approach, based on hydrodynamic studies of a fluidized bed, should be developed for future reactor scale-ups, for both research and technology transfer. Ohlinger noted that the hydraulic detention time decided the reactor performance and that FBR geometry in system design was controlled by both the hydraulic detention time and reactor kinetics (Ohlinger et al., 2000). The hydraulic detention time is a function of the reactor height, upflow velocity and the particle void fraction while reactor kinetics depends on the mass transfer and heat transfer rate; these, in turn, are greatly affected by the fluidization hydrodynamic properties(Epstein 2003). Therefore, studies on fluidization processes, as well as on bed expansion phenomena, are crucial to reactor design. Researchers have investigated the operational windows for fluidized-bed reactors of several designs, in particular, those concerning crystallization chemistry conditions. However, there has rarely been any research that examined the hydrodynamic optimization, to achieve optimal struvite production performance. In addition, it has been proposed that the best approach to small-scale, fluidized-bed studies is to separate the tests on chemical reaction kinetics and hydrodynamics (Zhou 2000). In this research, fluidization behaviour, as well as size growth distribution of struvite pellets, was studied in a two-dimensional, multi-compartmentalized vertical column, under continuous fluidization conditions. It is believed that the database will contribute to an understanding of the hydrodynamic characteristics of fluidized-bed reactors, relating to the  2  INTRODUCTION  optimization of reactor design and fluidization operations as it pertains to struvite production and recovery.  3  RESEARCH OBJECTIVES  CHAPTER 2 - RESEARCH OBJECTIVES A bench-scale two-dimensional reactor was designed and set up for observations of the hydrodynamic behaviours of a fluidized bed containing struvite particles. Struvite growth runs and analysis were also performed. The specific research objectives of the project were: 1) To test a protocol for determining the minimum fluidization velocity of different configurations of particles; 2) To establish bed expansion equations for variously sized particles for prediction of void fraction, based on superficial upflow velocity or vice versa; and 3) To examine the feasibility of the two-dimensional reactor, concerning fluidization effect.  4  BACKGROUND AND BRIEF LITERATURE REVIEW  CHAPTER 3 - BACKGROUND AND BRIEF LITERATURE REVIEW 3.1 Liquid fluidization Fluidization is a process that, by running fluid through a bed of solid particles, the interactive forces between the solids and the fluid counteracts the weight of the solids, so that a fluidlike totality of the solids suspension is formed. The swarm of solids that get fluidized are relatively free to flow and deform and is characterized by the homogeneous random motion of particles. This is in contrast to packed beds, wherein the particles are held in a relative rigid structure (Gibilaro 2001). Fluidization provides a significant displacement of particles and, hence, an adequate environment for aggregation; this is believed to be an important mechanism behind the the growth of struvite pellets in fluidized reactors (Seckler et al., 1996). Higher mixing of the fluidized state also leads to higher mass transfer rate and, thus, enhancement of the chemical reaction rate in the contained system. Under fluidization conditions, the force exerted on particles consists of buoyancy and drag force, which is understood to be friction forces. Since steady-state, fluidized particles appear free of gravity, without other support (other than the flow), macroscopically for a representative particle in the fluidized bed the downward gravitation force= upward buoyant force + upward drag force. This can expressed mathematically as Equation 3.1 (Epstein 2003).  Vp p g=^g — F D  ^  Eq. 3.1  So 5  BACKGROUND AND BRIEF LITERATURE REVIEW  F D = V(Ppg — pg)^  Eq. 3.2  The value of the upward drag force is constant for this representative particle in a steadystate fluidized bed, independent of the upflow velocity.  3.1.1 Hydrodynamic representation The total pressure drop across the bed height (L) of an FBR is attributed to both the liquid pressure caused by gravity, as well as the dynamic pressure drop (or unrecoverable pressure loss) due to the motion of the particles, i.e.,  —Ap = —AP + pgL^  Eq. 3.3  The total pressure drop (-Ap) can be measured from right above the distributor to the plane right above the bed surface by gauge or transducer. Assuming the effects of the bed entrance, the distributor, the bed exit and the wall effect can be neglected, the total pressure drop can be calculated by the bed suspension density ps„ since the whole bed is considered as a fluid, i.e.,  —.11/3 = p B gL^  Eq. 3.4  And  Ps = PP( 1— s) + PE  ^  Eq. 3.5  6  BACKGROUND AND BRIEF LITERATURE REVIEW  where g represents the void fraction or the ratio of the volume of liquid in relation to the volume of the whole bed (Epstein 2003). Thus, it can be deduced from the above equations that  —AP = L(1— E)(pp — p)g^  Eq. 3.6  3.1.2 Minimum fluidization  By imparting fluid flow through a fixed bed, the particles are " loosened up" and increasing the upflow velocity incrementally can gradually fluidized the bed; during this process the dynamic pressure drop keeps "rising", until it hits a "blunt" maximum and then levels off (Epstein 2003). This sudden transition of slope is where the minimum fluidization occurs. Finding the minimum fluidization condition for particles of a certain density and size is critical for determining the minimum energy input, in order to fluidize the bed of particles. It has been found that the minimum fluidization (Umf) increased with an increase of the particle diameter and a decrease of solid density and was independent of initial bed height (solid loading) (Lakshmi et al., 2000). 3.1.3 Bed expansion  Before the upflow velocity reaches the minimum fluidization velocity, the bed surface remains the same and there is no change in the bed height or the void fraction. Once the upflow velocity passes the minimum fluidization velocity, the bed begins to expand and the bed height can be observed to rise; also, the distance between particles starts to increase, which decides the magnitude of the void fraction macroscopically. It is worth noting, again, that the bed expansion process does not cost extra energy, since the dynamic pressure drop is  7  BACKGROUND AND BRIEF LITERATURE REVIEW  a constant value for the fluidized state of the bed, independent of the change of upflow velocity (Gibilaro 2001; Epstein 2003). Previous researchers have discovered that there exists an empirical linear relationship between the logarithm upflow velocity and the logarithm void fraction, as shown by Equation 3.7, for a fluidized bed loaded with monodisperse (uni-size and uni-density) non-porous solid spheres; however, there is little agreement on the correlation coefficient in the mathematical expression (Gibilaro 2001; Epstein 2003).  dl og(U) di oz (g)  = n = constant  Eq.3.7  From Equation 3.7, it can shown that  U = Ut  En  Eq.3.8  where Ut stands for the free settling velocity of the representative particle(Richardson 1954a). But some other researchers found that the U < Ut  n  in their experiments and add a  correction coefficient k (k< 1) to the equation, by including a wall effect factor in the definition of k (Epstein 2003). The value of slope n of the plots for dlog(E) against dlog(U) ranges from 2.4 in turbulent flow to 4.8 in viscous flow (Gibilaro 2001; Epstein 2003). 3.1.4 Segregation  Segregation is a process during which a particular size fraction preferentially occupies a certain region of the fluidized bed, so as to render the net forces acting on each particle to be  8  BACKGROUND AND BRIEF LITERATURE REVIEW  equal to zero. In this process, the smaller particles will rise toward the top of the fluidized bed, while the larger ones will tend to settle toward the bottom of the bed. On the other hand, particle dispersion has a homogenizing effect, which tends to uniformly mix the solid particles throughout the bed, irrespective of their size differences. Under the influence of these two counteracting mechanisms of mass transport, the particles of a particular size fraction are generally present throughout the bed in varying concentrations. This nonuniform concentration distribution of different-sized particles, in the bed, leads to a variation in the voidage along the bed height. This behavior is a characteristic feature of the fluidized beds containing different-sized particles, resulting in the presence of density gradients in the bed. Furthermore, since the specific reactivity of the particles is expected to be a strong function of particle size, both the volume fraction and the particle size distribution have pronounced effects on the volumetric production rate, along the length of the reactor (Asif 1994). 3.1.5 Fluidized-bed reactors Fluidized-bed reactors have been recently put into practice for struvite growth and recovery since it has several advantages: a small pressure drop, good heat and mass transfer, the mobility of the particles and reduced risk of blockage (Miura 1997). Among the previous research reported on struvite production in fluidized bed, there have been different designs, such as a vertical cylinder reactor column (Seckler et a/.,1996), a tapered bed (Huang 2000), a tapered bottom with a straight cylinder top design (Yoshino et al., 2003) and a multi-compartmentalized reactor, with increasing widths of the sections from bottom to the top (Britton 2002). Tapered beds or increasing-width-multi-sectioned reactors are better for particles classification and reducing washout of fine particles (Huang 2000; Fred Koch, Research Associate, Department of Civil Engineering, University of British Columbia, pers.com .). However, it has rarely been examined for a potential optimal design of fluidized-bed reactor, for phosphorus recovery or the rationale for this optimization.  9  BACKGROUND AND BRIEF LITERATURE REVIEW  It has been proposed that the best approach to small-scale fluidized-bed studies is to separate the tests on chemical reaction kinetics and hydrodynamics (Zhou 2000). Such was the intention of this study.  3.2 Crystallization 3.2.1 Struvite growth mechanism Struvite precipitation occurs when the combined concentrations of Mg 2+ , NH4 + , and PO 4 3exceed the struvite solubility limit. Availability of the three components is controlled by the system pH, because the ion activity of Mg 2+ , NH 4+ and PO4 3- are highly pH dependent, and the total dissolved concentrations of magnesium, ammonia, and phosphorus species (Ohlinger 1998). Struvite precipitation comprises nucleation and growth, steps that occur sequentially. Nucleation, the formation of crystal embryos, is primarily a reaction-controlled process. It has an inherent lag period, which is a function of the struvite supersaturation level. The supersaturation level determines struvite precipitation potential. The supersaturation level can be represented by supersaturation ratio (SSR); the definition of SSR is the ratio between the ion concentration product of the three constituent ions and the struvite solubility production (Steen 1998). As supersaturation is increased, higher concentrations of constituent ions become available, and the nucleation lag period is reduced. Growth, incorporation of constituent ions into an existing crystal lattice structure, follows nucleation and is primarily a transport-limited process. The reduction of growth rate limitations has been experimentally demonstrated with the input of mixing energy. Energy input disrupts concentration gradients in boundary layers surrounding growing crystals and increases the struvite crystal growth rate (Ohlinger et al., 2000). Thermodynamically, the metastable zone is defined as the critical zone of supersaturation of solution, where crystallization is not governed by nucleation and thus avoids rapid and/or spontaneous precipitation (Ali 2006). 10  BACKGROUND AND BRIEF LITERATURE REVIEW  Preferential accumulation has been observed to occur in locations with high mixing energy (Ohlinger et al., 2000). In the same study, mixing energy was demonstrated to be the primary influence in overcoming the transport limitations to crystal growth and was concluded to be the primary parameter controlling preferential accumulation. By selecting struvite crystals as seed media in a high-energy mixing precipitation reactor, nucleation, with its inherent lag period, can be bypassed and growth can ensue immediately (Ohlinger et al., 2000). 3.2.2 Growth rate  Although the average struvite growth rate was not a limiting parameter in the pilot scale testing, factors affecting growth rate must be given consideration during scale-up design, to assure that a limitation is not induced. The growth rate was calculated per square meter of available media surface area. Because specific surface area varies significantly with media size, a surface area limitation, and hence growth rate limitation, could be induced if a relatively large media was selected. The higher flow velocity required to fluidize a larger media particle bed would require increased pumping energy. Conversely, small media particle size selection would provide a large specific surface area and require a lower flow velocity to fluidize the bed. The limitations on small particle size is the need to prevent media washout. Also, because the media will increase in size over time, it is important to design the flow adjustment and media removal capability into the FBR (Ohlinger et al., 2000). As the crystallization progresses, the seed crystal grows excessively and as a result, the effective reaction surface areas are decreased and the fluidization effect is degraded, causing the recovery ratio to be decreased (Shimamura et al., 2003). The seeding of mother crystals (previously-grown struvite particles as seeds) in struvite crystallization provides a positive influence on struvite growth. Mother crystals provide support of efficient diffusion integration processes, leading to faster growth of struvite during the crystallization (Ali 2006).  11  BACKGROUND AND BRIEF LITERATURE REVIEW  Struvite crystal growth can be a transport limited process depending on environmental mixing energy. The FBR process was selected, in part, because of the potential for the high mixing energy in the reactor, to overcome the limitations for transporting constituent ions to their respective incorporation sites, in the growing crystal lattice structure (Ohlinger et al., 2000).  12  MATERIALS AND METHODS  CHAPTER 4 - MATERIALS AND METHODS 4.1 Reactor design A two-dimensional, multi-compartmentalized, fluidized-bed reactor was designed for the specific project, as shown in Figure 4.1. It has four different areas of cross section, increasing from the bottom to the top. For a given upflow volumetric rate, each section has a different superficial upflow velocity, decreasing from the bottom to the top. Although the cross section is commonly circular, this specific reactor was deliberately designed to be rectangular, for a better view into the reactor, as it provides a flat front. The thickness of each section is 3.6 cm and the width of each section doubles the width of the section beneath it. The specific dimensions are given in Table 4.1. This design builds a two-dimensional reactor, providing a large flat front, made of transparent plexiglass for better visual observation of the particles' hydrodynamic behaviors. The whole experimental process was divided into two stages. During the first stage, the liquid running through the reactor was under controlled equilibrium conditions chemistry wise, in other words, no struvite was supposed to melt or form in the working solution since the concentration product of the three struvite constituent ions (magnesium, ammonia, phosphate) was adjusted to the synthetic solubility of struvite 13.92 (Adnan 2002). The hydrodynamic properties of particles with different sizes or arrangement of mixtures, were tested and prediction equations for the fluidized bed expansion process were established. During the second stage, struvite growth, as well as size distribution tests were performed under several different configurations, in terms of upflow velocity, as well as seeding size and quantity.  13  MATERIALS AND METHODS  U  3.6cm  IA  // SEED HOPPER  40cm  cu  REACTION ZONE  ri 12cm 3.6cm  U  ACTIVE ZONE  Cl  6 3.6cm  E U  ri  HARVEST ZONE  3.6cm 3.6cm  ^0 NaOH 112L E U 0  /^/  r,  /  ^  EFFLUENT (^ 124L  0  FEED 246L  Figure 4.1 Demonstration of the two-dimensional reactor and struvite recovery configurations.  14  MATERIALS AND METHODS  Table 4.1 The dimensions of the reactor shown in Figure 4.1  Width(cm)  Height(cm)  Thickness(cm)  Volume(L)  40  22.5  3.6  3.24  Second  12  43.2  3.6  1.87  Third  6  43.2  3.6  0.93  Bottom  3.6  43.2  3.6  0.56  Top section  4.2 Hydrodynamic phase A two-dimensional, multi-compartmentalized column was designed to perform the experimental studies. The increasing width of the compartments from bottom to top of the reactor was used to classify the particles of various sizes or even densities. As shown in Figure 4.2, a long and transparent tube made of glass, functioning as a monometer, was set up vertically beside the reactor, with the bottom inlet connected to the reactor inflow right before the flow passes the distributor plate; this was placed at the inlet of the FBR reactor and perpendicular to the flow. The distributor was a metal plate of the same size of the inlet of the reactor, with evenly distributed holes of 0.2 mm diameter, to keep the inflow uniform  15  MATERIALS AND METHODS  across the column X-section and to give stability to the bed bottom boundary, when stopping the solids. The water level difference between in the monometer and in the reactor was measured, which is directly the unrecoverable pressure loss(dynamic pressure drop) in order to mobilize the struvite pellets ( Gibilaro 2001). Water level in the monometer is higher than in the reactor and the level difference is in proportion to the energy loss on the momentum of the particles in the flow. Since the water level difference was measured while the flow was running upward, a small visibility error, due to the mobility of the two water levels in the monometer and in the reactor, was likely involved. To develop a more in-situ and automatic method for determining the dynamic pressure drop, pressure gauges, installed at the liquid surface and the reactor inlet, should be employed in further research and the results compared to the "water level difference measurement" protocol, developed in this study, for accuracy validation. By controlling the pump speed, a range of Reynolds numbers was maintained, so as to investigate the unrecoverable pressure loss as a function of fluidization velocities; it also determined the minimum fluidization conditions for various sizes of particles. Bed heights were measured, with a regular ruler, and the corresponding void fractions were calculated, based on the pre-determined particle characteristics SSD (saturated solids density), diameter and dry mass, as well as the reactor geometry. Six groups of solids, were graded out in terms of size, by copper sieves available in the Environmental Lab of UBC. Each group was considered as monodisperse solids, although it actually comprised a size range of particles due to the limited precision of the sieving measures (0.17mm to 1 mm). For further research, it is recommended to employ a more complete set of sieves with finer gradient of the sieve perforation diameter.  16  MATERIALS AND METHODS  3.6cm  40cm  Cu  12cm 3.6  /  cu ri 6cm CU  U  E U 0  f*,  0 SYNTHETIC SOLUTION /  ^246L  Figure 4.2 Demonstration of the monometer set up beside the reactor  17  MATERIALS AND METHODS  4.3 Struvite growth tests 4.3.1 System configuration As demonstrated in Figure 4.1, the reactor was vertically set up and secured onto a vertical wall for stability. The inlet located at the bottom end of the reactor was connected to the NaOH tank first, and then to the recycle tank and the main feed tank separately. The effluent of the reactor, exiting the top clarifier of the reactor, descended through a pipe into the recycle tank. The recycle (effluent) tank functioned as both the outflow storage as well as the dilution source. The recycle flow, exiting the effluent tank, converged with the main feed that was the source of the major constituent ions required for struvite production; the flow then ascended to meet the alkaline addition(NaOH), forming a certain magnitude of SSR (Super Saturation Ratio) by raising the pH (Ohlinger et al., 1998). The concept of SSR is fully discussed elsewhere, but briefly, it is the ratio between the ion concentration product and struvite solubility; SSR is an index of the struvite precipitation potential (Steen 1998). The main feed was a synthetic solution, with a composition of 0.18mol/L MgC12 (Magnesium Chloride) and 0.026mol/L NH 4 H 2 PO 4 (Ammonium Dihydrogen Phosphate), which has a pH of 6.2, calculated by ChemBuddy pH Calculator, a software based on acid and base titration calculation. The NaOH tank contained a NaOH solution of 0.1N. The effluent tank would then be topped up by tap water at the start-up of each run. The main feed inflow and the NaOH flow were set, respectively, at 0.086 L/min and 0.011L/min permanently. The recycle ratio was adjusted to provide various upflow velocities, as well as potentially different SSRs, as the experimental run warranted it.  4.3.2 Sampling methodology A long sampler, made of a metal stick supporter with a tube tied to this supporter, was inserted into the reactor from the top; at every time of sampling, it sucked in about 5m1 to 10m1 of the bed suspension, composed of liquid and fluidized solids with the help of a  18  MATERIALS AND METHODS  syringe at the end of the tube, left outside the reactor. The sample itself, was retained in the tube, until the tube was removed from the reactor and it was injected into a beaker. Drops of sample taken by a dropper would then be put on a glass carrier and examined under a digital microscopic camera, Motic W10X/20, connected to a computer. The computer software Motic Images Plus 2.0 could then measure the diameter of particle images in the microscopic photos taken of the drop sample, composed of a little liquid and a few particles. At each time of sampling, four samples were taken, respectively, from the four sections of the reactor, such that a representative size distribution throughout the bed could be constructed. 4.3.3 Flow rate measurement  A stop watch and a 1000-m1 cylinder were used to measure the outflow (=inflow).The time for the outflow filling up the originally empty cylinder was recorded and, by using the equation Q=V/t, the flow rate could be calculated based on the actual field data. Although it is actually simple in concept, this technique proved quite accurate and reliable in determining actual flow rates in the FBR.  19  RESULTS AND DISCUSSION  CHAPTER 5 - RESULTS AND DISCUSSION Two phases of runs were performed on the two-dimensional column reactor. The first phase tested the hydrodynamic properties of the reactor loaded with the struvite pellets previously grown from a couple of pilot-scale FBR struvite reactors, located at the LULU Island wasterwater treatment plant, in Richmond, B.C., Canada (Md.Iqabal Hossain Bhuiyan, PhD student, Environmental Engineering group, Department of Civil Engineering, UBC, pers.com .). During the first phase of the study (hydrodynamic study), the solution, running through the bed, was maintained at chemical equilibirum concerning struvite formation or resolving. The second phase investigated the qualification for growth, under different fluidization velocities. The basic patterns of plots for dynamic pressure drop and fluidization velocity, as well as for void fraction and fluidization velocity, are in good agreement with previously reported research, with different reactor and solids configurations. However, the specific mathematical relationships established deviated from the classical ones reported in the literature. Binary mixtures of solids were tested on the same protocol of bed expansion experiments, for monosize solids; the hydrodynamic behavior results are not exactly congruent with those unisize solids runs, but show meaningful divergence. The growth tests in the second stage of the project shed light on the effects of seeding size and quantity, as well fluidization velocity, on maintaining phosphorus removal efficiency and struvite recovery.  20  RESULTS AND DISCUSSION  5.1 Hydrodynamics 5.1.1 Minimum fluidization  Dynamic pressure drop By running the flow through a bed loaded with particles of different sizes or combinations and pumping the flow upward with incremental increases of the flow rate from zero flow to start with, the dynamic pressure drop measured with the monometer increased until the fluidization velocity reached the minimum fluidization velocity; it then levelled off throughout the entire fluidized state, regardless of the subsequent increases in fluidization velocity. The slope transition point, near where a blunt maximum of the dynamic pressure drop occurs, is taken as the minimum fluidization velocity. Figure 5.1 to Figure 5.4 show the dynamic pressure drop as a function of upflow velocity for the runs of four monosize particles, for four different representative diameters (dp). The representative diameters are defined as the medium value of the range of diameters for a certain group of "monosize" solids, since the precision of the "grading measure" ranges from 0.17mm to lmm (see Chapter 4 Materials and Methods). The pattern of the curve, plotted between the dynamic pressure drop (log) and the upflow velocity (log), repeats for those four runs and only deviates at the minimum fluidization velocity; this is expected and agrees well with the previously reported research (Gibilaro 2001; Epstein 2003). The minimum fluidization velocity pinpointed on each curve, for each group of unisize solids, is larger for bigger size solids; this is shown from the four figures, as well as in Table 5.1, which confirms the general observation that finer particles require less pumping to be fluidized.  21  RESULTS AND DISCUSSION  dp=4mm  log(AP)  1.200 1.000 0.800 0.600 0.400 0.200  -3.500^-3.000^-2.500^-2.000^-1.500^-1.000^-0.500_0.2067. 00  log(u)^-0.400 -0.600 -0.800  Figure 5.1 Dynamic pressure drop as a function of upflow velocity for dp=4mm solids  dp=2.9mm  0.800 0.600 0.400 0.200  -3.50^-3.00  0  Figure 5.2 Dynamic pressure drop as a function of upflow velocity for dp=2.9mm solids  22  RESULTS AND DISCUSSION  1.000  dp=2 mm log(AP)  0.800  0.600  0.400  0.200  -3.500^.000^-2.500^-2.000^-1.500^-1.000^-0.500^0. 00  log(u)  -0.200  Figure 5.3 Dynamic pressure drop as a function of upflow velocity for dp = 2mm solids  1.000  dp=0.9 mm log(4)  0.900 0.800 0.700 0.600 0.500 0.400 0.300 0.200  log(u  0.100  -4.00^-3.50^-3.00^-2.50^-2.00^-1.50^-1.00^-0.50^0.00  Figure 5.4 Dynamic pressure drop as a function of upflow velocity for dp = 0.9mm solids  23  RESULTS AND DISCUSSION  Table 5.1 Specific equation coefficients derived and particle properties  Group Number  #1  #2  #3  Size range(mm)  3.5 - 4.5  2.5 - 3.3  1.7 - 2.2  dp(mm)  4  2.9  2  SSDp(g/c m3 )  1.445  1.487  U na(m/s)  0.0109  Ut(m/s)  #4  0 83 -  Mixture of group#1 and group#3  Well mixed  Top layer  Bottom layer  0.9  2  4  1.519  1.5  1.445  1.519  0.0087  0.0061  0.0017  0.0055  0.0069  3.88  2.228  1.074  0.228  3.880  1.074  Ue(m/s)  0.769  0.857  0.680  0.734  0.955  0.264  n  3.01  5.46  3.36  3.77  6.098  Re  3.1 246.9  2.2 78.5  1.5 20.0  0.3 - 8.1  E  0.42 0.73  0.59 0.73  0.54 0.65  0.42  0.59  0.54  rilf.  ' 1.0  Mixture of group#3 and group#4  Whole bed segregated  0. 0.0055 0.0069  0.0038  0.259  0.33  0.789  1.383  1.307  1.59  4.000  0.56 34.26  0.5616.72  1.1434.26  0.56 34.26  0.5 - 48.9  0.37 0.58  0.47 0.51  0.49 0.72  0.49 0.60  0.51 - 0.67  0.43 0.70  0.37  0.47  0.49  0.49  0.51  0.43  0.0055  -  As for multidisperse solids, two mixtures were tested. The first binary mixture of solids consisted of 50% weight of dp = 4mm solids and 50% weight of dp = 2mm solids. The second binary mixture of solids tested were composed of 50% weight of dp = 2mm solids and 50% weight of dp = 0.9mm solids. The mixture of dp = 4mm solids and dp = 2mm solids were loaded into the reactor (bed) well mixed and behaved with an identical pattern as the monosize particle beds; and the minimum fluidization episode hit a similar and even slightly smaller value of the upflow velocity, as in the dp = 2mm monosize particle test (Figure 5.5). Thus, it can be concluded that a mixture of distinctively different size particles tends to behave more closely to the smaller size constituent pattern.  24  RESULTS AND DISCUSSION  Employing a surge flow (suddenly a significant increase in upflow), the entire bed became segregated into two distinct layers of solids, with the top layer composed of small particles (dp = 2mm)and the bottom layer of large particles(dp = 4mm). There was a sharp interface between the two layers of solids and the interface did not disappear or melt away with the fluidization process (no intermixing observed). The same fluidization test was performed on the packed bed, after segregation, and the results were different. The lower curve in Figure 5.5, for the segregated bed, has a smoother slope transition as opposed to a blunt one of the completely-mixed bed before segregation. It also has one more transition (the step during the short horizontal-like stretch) further along the line; however, it was not very clearly displayed in Figure 5.5, due to the confusion brought by the second major turning point, as well as by some hysteresis. However, plots of bed height against upflow velocity (Figure 5.6) show more explicitly two "minimum fluidization" episodes, one for the top layer at first and another further down the process, for the bottom layer. The curve of the segregated bed climbs upward again (the second major turning point of the lower curve in Figure 5.5), after a short "horizontal" stretch, when the bed height exceeded the top of the bottom section of the reactor and the flow slowed down in a bigger cross-sectional part. Thus, as the top part of the bed experienced a decelerated upflow, the bed was partially defluidized and the dynamic pressure drop started to increase again. It needs to be noted that only in Figure 5.5, the logarithmic flow rate instead of the upflow velocity, was employed, since the upflow velocity is not consistent throughout the solids due to the flow deceleration over the top of the bottom section.  25  RESULTS AND DISCUSSION  Mixture of dp=4mm and dp=2mm 1.400 1.200^log (AP) 1.000 0.800 0.600 0.400^Completely  log itt)  mixed  0.200^Segregated  -4.000^000^-2.000^-1.000 -0.20 00 -0.400 -0.600  Figure 5.5 Dynamic pressure drop as a function of the flow rate for the mixture of dp = 4mm and dp = 2mm particles. 18^ 16  Fluidization transitions for segregated bed  L(cm)  14 12  10  —Bedheight of top layer  8 6  ^  —Bedheight of bottom layer  4 2  U(m/s  )  0 0.0000 0.0050 0.0100 0,0150 0.0200 0.0250 0.0300 0.0350 0.0400  Figure 5.6 Bed height as a function of the upflow velocity for the two separate layers  respectively of the binary mixture bed of dp = 4mm and dp = 2mm solids.  26  RESULTS AND DISCUSSION  Interestingly, another mixture of solids (Figure 5.7) tested did not repeat the overall results of the first mixture, as discussed above. The second mixture of solids did not go through any segregation even under surge conditions. The reason for this is probably that the two sizes comprising the mixture do not have a significant difference; for each of the "monosize" constituents, the "scattering" of the size of individual particles can be as large as the difference between the two component groups, considering that the group of solids (with a representative diameter of 4mm) ranges from 3.5 to 4.5mm, due to the limited precision of the grading equipment. In a sense, each monosize group is a "mixture" already, but with as small as possible a size variation. It is reasonable to believe that, for segregation to happen, it requires that the size difference between the two constituents has to be above a minimum. When it comes to the representative minimum fluidization velocity pinpointed during the process of gradual fluidization for this mixture, the value falls between the two corresponding minimum fluidizations of the monosize beds of each size component.  1 '00  Mixture of dp=2 mm and dp=0.9mm logrApi  1.000 0.800 0 600 0. -I00  0.200 log( ti -3 500^-3.000^- 1 5011^-2 000^-1 500^-1 000^-0 500^0 ( 00 -0 '00 ,  Figure 5.7 Dynamic pressure drop as a function of upflow velocity for the mixture of 2mm and 0.9mm.  27  RESULTS AND DISCUSSION  Bed height  A plot of bed height versus upflow velocity can be another way to determine the minimum fluidization velocity. As Figure 5.8 shows, during the same experimental process of gradually fluidizing a packed bed loaded with dp=4mm particles, the bed surface remains at the same height until the upflow velocity passes the minimum fluidization velocity and then it rises with a following increase in upflow velocity. The point where the bed height just started to rise can be considered as the minimum fluidization velocity; this is the same minimum fluidization velocity pinpointed on the plot between dynamic pressure drop (log) and upflow velocity (log). In addition, since it is a relatively straight line after the minimum fluidization point, it can be perceived that the bed height increases linearly with an increase in upflow velocity in a perfect cylinder with the same cross-sectional area throughout the height. However, previous research has demonstrated a close to exponential rise of bed height over the increase of fluidization velocity (Epstein 2003). The deviation is probably due to a wider range of void fraction (from 0 up to close to 1), thus a wider range of fluidization velocities that was facilitated in previous research; this may have been due to the use of straight beds, as opposed to a bed with increasing cross-sectional area from bottom to top, as in this study. It could be concluded that the linear trend of the bed height rise, over an increase in upflow velocity, is only valid within a limited range of bed void fraction (c), for monosize particles. For mixtures, the bed height increase for each layer of the mixture of dp = 4mm and dp = 2mm shows more of an exponential trend with more hysteresis (as shown in Figure 5.6), instead of a relatively smooth straight line in Figure 5.8.  28  RESULTS AND DISCUSSION  dp = 4mm 40 35 30 25 20 15 10 5 0  Ulm /s)  0.0000^0.0100^0.0200^0.0300^0.0400^0.0500^0.0600^0.0700  Figure 5.8 Bed height change over the increase in upflow velocity for dp = 4mm solids  Table 5.1 also shows a compilation of the minimum fluidization velocities for different configuration of particles and we can conclude that the magnitude of the particle size can decide the magnitude of the minimum fluidization velocity. To achieve a fluidized state for larger particles, a higher energy input (pumping) is required. The plots of bed height versus upflow velocity, for each layer of the segregated bed of mixture of dp = 4mm and dp = 2mm solids (Figure 5.6), shows clearly that the bed with segregated particles goes through two "minimum fluidization" points; the first is in relation to the top layer, coinciding with the minimum fluidization episode during the fluidization of the completely mixed bed, with the same mixture of solids, while the second one occurs later on. The top layer's minimum fluidization point agrees well with the results from the bed composed purely of just the small particles (dp = 2mm), which constituted the top layer; the value of the top layer's minimum fluidization velocity is only slightly smaller compared to the monosize bed of dp = 2mm particles. However, the bottom layer's apparently independent fluidization process does not show a similar minimum fluidization velocity,  29  RESULTS AND DISCUSSION  compared to the bed composed purely of just the large particles (dp = 4mm). In fact, there is a significant deviation (Table 5.1). In summary, when the bed becomes segregated, the top (small particle) layer gets fluidized first and then the bottom layer, proving that the two layers behave quite independent of each other. The minimum fluidization velocity for the large particles (bottom layer) is significantly lower, compared to the results from the monosize bed composed of the same size particles, as shown in Table 5.1; it is probably due to a relatively denser upper boundary (smaller void fraction) and thus higher slipping velocities between the solids and the liquid moving through the voids (Gibilaro 2001). It is important to note that, both minimum fluidization velocities of the segregated bed are smaller, to a different degree, than the corresponding minimum fluidization velocities of the monosize beds. Therefore, from an operational perspective, it requires less energy input to fluidize a reactor loaded with multidispersive solids in terms of size, compared to a monosize bed composed of the larger size component, for both the segregated bed and completely mixed bed. 5.1.2 Bed expansion After the upflow velocity passes the minimum fluidization point, the bed starts to expand and appears in a homogeneous fluidization state (Gibilaro 2001). An important relationship between upflow velocity and void fraction is normally derived from the fluidization process, beyond the minimum fluidization moment. It is crucial to the reactor design that a quantitative knowledge of the bed expansion, as a function of the liquid upflow velocity to be developed and evaluated (Epstein 2003). Strong linear relationships between the logarithmic liquid superficial velocity and logarithmic void fraction, as previous research has shown (Gibilaro 2001; Epstein 2003), have been established in this work, for all seven fluidization runs(six groups of particles with different arrangements, two processes for the first mixture of solids as discussed above). As shown in Figures 5.9 to 5.12, the experimental results confirmed the previously-suggested, 30  RESULTS AND DISCUSSION  empirical relationship as developed in Equation 3.7. There was not a decreasing tendency of bed expansion rate (indicated by the slope of the linear trend-line) with smaller particles, as it would be perceived, but the equation coefficients appear rather random, instead of in proportion to sizes of the particles (Table 5.1). The linear lines between upflow velocity (log) and void fraction (log) were extrapolated up to c =1[log(c)=0], in order to derive the intercept for the line equation. Even though the actual data points didn't cover the entire range of  £  as from 0 to 1 (due to aforementioned  experimental design specifications), previous research findings confirm that the linear relationship is always valid for the entire fluidization process from minimum fluidization to "wash out"( 8 close to 1) status of the bed (Gibilaro 2001; Epstein 2003).  dp=4mm log( 0.300  log(e) = 0.332Iog(U) +0.263 R 2 = 0.994 -2.500  -2.000^-1.500  -0.500 _0.1000.'00^0.500 -0.200 -0.300 -0.400 -0.500  Figure 5.9 Bed expansion linear relationship between upflow velocity (log) and void fraction (log) for dp = 4mm solids.  31  RESULTS AND DISCUSSION  dp=2.9mm  logy) 0.200 0.150 100  logic) 0.1331014(0 10.154 11 2 0.907  0.050 0.000  -2.50  ^  -2.00^-1.50  l og(11)  -0.50^0.00 -0.050  ^  0.50  -0.100 -0.150 -0.200 -0.250  Figure 5.10 Bed expansion linear relationship between upflow velocity (log) and void fraction (log) for dp = 2.9mm solids. 0.500  dp=2 mm  0.400  lo^ ' )  300 log(c) 0.298Iog(u) 0.385 R 2 0.839  logo]) -2.500^-2.000  0.200 0.100 0.000  -1.000  -0,500^0.000 -0.100  0.500  -0.200 -0.300 -0.400  Figure 5.11 Bed expansion linear relationship between upflow velocity (log) and void fraction(log) for dp = 2mm solids.  32  RESULTS AND DISCUSSION  dp=O.9mm log(e) 0.2651og(u) t 0.309 R 2 0.998  0.400  10g( t)  0.300 0.200 0.100 10g,(111  0.000 -3.00  -2.50  -2.00  -1.5  -1.00  -0.50^0.00 -0.100  0.50  -0.200 -0.300 -0.400 -0.500  Figure 5.12 Bed expansion linear relationship between upflow velocity (log) and void fraction(log) for dp = 0.9mm solids.  When it comes to the two binary mixtures of particles, a linear relationship exists between dynamic pressure drop and upflow velocity, just as in the monodisperse beds. For the mixture of dp = 4mm solids and dp = 2mm solids, not only was the average mathematical correlation for the whole segregated bed established but the specific relationship within each layer was explored. It turned out, strikingly, that although the boundary conditions of the segregated layers are not the same for the well-mixed bed or the monodisperse bed, the hydrodynamic behaviour does not differ very much. As it can be seen in Figure 5.13, for each layer, the Rsquare value is very close to one which indicates a perfectly normal and expected linear trend. Thus, even the top layer has an "unstable" particle bottom boundary instead of a fixed distributor, and the bottom layer has an "unstable" particle top boundary, instead of the liquid-solids interface; each layer of particles behave just as would be expected with the normal boundaries.  33  RESULTS AND DISCUSSION  It is also important to notice that, although each layer expands rather independently according to the bed height plots, the bed expansion, linear relationships ended up very similar and even uniform, except that they started at different minimum fluidization velocities. This indicates that the two, seemingly independent layers actually affect each other to some extent, such that they can be still considered as a whole bed, from a practical point of view.  1.600  Log(C)  1.400  1200 1.000  ;  •  Bottom layer(Large particles)  •  Top layer(small particles)  0.800 Log(t s )= 0.7231og{U) +1.330 R 2 = 0.938  0.600 0.400 Log(c0= 0.7651og{U) +1.350 0.200 R 2 = 0.927  0.000 -2.500^;A*6  -1.500^-1.000  Log(U)  -0.500 -0.20'000^  0.500  -0.400 -0.600  Figure 5.13 Void fraction(log) vs. upflow velocity (log) for the top and bottom layers of the segregated bed of mixtures of dp=4mm and dp=2mm solids. Also comparing well was the well-mixed bed and the bed after segregation, concerning the bed expansion linear relationship between dynamic pressure drop and superficial upflow velocity. This is shown in Figure 5.14, where it can be seen that the segregated bed doesn't appear as strict a straight line as the completely-mixed bed; this is reasonable, because the segregated bed doesn't have that same consistency and uniformity throughout the bed, as in a completely-mixed bed. Therefore, to predict the expansion of a segregated bed, it is recommended to consider the expansion of each layer, respectively. At the same time, the 34  RESULTS AND DISCUSSION  significant difference between the line slopes indicates that a segregated bed could expand faster than a completely-mixed bed. From an operational perspective, it can be considered to pre-segregate the seeding solids, so as to achieve a more efficient fluidization, while maintaining the spatial classification of solids.  1.200^, tog( ,e.'  log(t) 0.629Iog(u) 11.103 R 2 0.895  0.600^+ Completely mixed —Segregated  log( -2.500  -0.500^0.000 -0.200 log(e) - 0.164log,(u) 0.046 R 2 0.993  0.500  -0.400  Figure 5.14 Void fraction (log) vs. upflow velocity (log) for the completely mixed bed and the segregated bed with the same mixture of dp 4mm and dp = 2mm solids. Equations: The classical empirical equation between superficial velocity and void fraction is as follows as discussed in Chapter 3 (Gibilaro 2001):  35  RESULTS AND DISCUSSION  cil og (U)  cll o g (.0  =n^  Eq.3.7  U = Ut • £n,^  Eq.3.8  where  (Pf — PdgdP 2  Ut = ^ Creeping flow regime (RE< 0.2) 1811  Eq.5.1  and  Ut = V3.03gdp(p p — p f)/p f , Inertial flow regime (Re>500)(Epstein 2003)^Eq.5.2  Although there does exist a linear relationship between log(U) and log(c), as confirmed in this study, the equation constants derived are not in agreement with the Richardson and Zaki equation (Richardson 1954a), due to different reactor design and FBR configurations. After  transforming the equation of  dl o g(U) c^ ll og( )  = n to the form as U = Ut • En (Eq.3.8), for the  mathematical expressions derived from the seven runnings, Ue is used instead of Ut (the terminal free settling velocity of a certain particle size and density). Therefore,  U Ue^  Eq.3.8 36  RESULTS AND DISCUSSION  where Ue is the exponential function of the intercept in those linear bed expansion relationships and n is the reciprocal of the slope. For example, for the linear correlation in the dp = 4mm solids run  log(E) = 0.332 log(U) + 0.263^  Eq.5.3  and U.e-0.263E1/0.332=0 709•0.01  ^  .  Eq.5.4  Since the Reynolds numbers are quite small (see Table 5.1), Ut was calculated according to the inertial flow equation; the comparison between Ut and Ue based on this research is also listed in Table 5.1. Since the Reynolds numbers for each test is between the creeping flow condition (Re<0.2) and the inertial flow condition (Re>500), the derived n (based on this study) is expected to fall into the range from 2.4 (inertial flow) to 4.8 (creeping flow), as discussed in Chapter 3. However, n for dp = 2.9 mm solids, as well as n for the mixture of dp = 4mm and dp = 2mm, no matter how well mixed, did not fit the rule proposed from previous research. What is more intriguing is that, Ue deduced based on the run of dp = 0.9mm, is even larger than Ut calculated according to Stoke's Law. Radial dispersion A large-scale, bulk circulation has been observed with upward flow in the central region and downward near the walls, where non-uniformity is accentuated in the top section(s) because of larger deadzones (the flat plane area beside the entrance of each section) in the upper sections with larger widths. This dispersion could negatively affect the fluidization effect and requires more energy input, to maintain the convective stabilities, by setting up more distributors(Epstein 2003). An obvious consequence of such a bulk circulation is particle loss  37  RESULTS AND DISCUSSION  distributors(Epstein 2003). An obvious consequence of such a bulk circulation is particle loss into the deadzone areas. Thus, a round X-section reactor with tapered transitional part(s), is proposed, since a round cross-section (amongst all shapes) guarantees the smallest deadzone for the same cross sectional area and the same entrance size. A tapered transitional part attenuates the flow deceleration and the horizontal flows, at each entrance, from one section to the next one above and thus, has an effect of maintaining a uniform upflow.  5.2 Size growth distribution The experimental stage on the size growth, as well as size distribution throughout the whole bed, produced some interesting and qualitative results. As previously introduced in Chapter 4, a long sampler, made of a metal supporter with a tube tied to the metal supporter was inserted into the reactor from the top; extraction of about 5m1 of the bed suspension took place, composed of liquid and fluidized solids, with the help of a syringe at the end of the tube, left outside the reactor. The sample was retained in the tube until the tube was removed from the reactor and it was injected into a beaker. Drops of sample taken by a dropper were then put on a glass carrier and examined under a microscopic-level camera connected to a computer. The computer software Motic Images Plus 2.0 could measure the diameter of particle images in the microscopic photos taken of the drop sample, composed of a little liquid and a few particles. Different recycle ratios, upflow velocities and seeding arrangements were designed for the similar inflow of synthetic feed. 12-hour growth was observed by sampling twice, at a time interval of 12 hours, for each configuration as shown in Figure 5.15 to 5.17. The vertical axis represents the number of particles in one drop of sample solution and the horizontal axis labels out a range of particle sizes. For the configuration of upflow velocity=0.021m/s, recycle ratio=17.9, seeding mass=4.71g, size<0.119mm, after 12 hours of continuous running, the particles still clustered around 0.2mm, while the number of the smallest particles (0.1mm) increased in all four sections; this indicated that the growth consists mainly of nucleation. The reason for a predominant nucleation growth could be that the seeding mass was not 38  RESULTS AND DISCUSSION  adequate, thus there was not enough contact area for molecular growth to occur. As for aggregation, since the seeding comprised only very small particles (0.119mm), the incidence of collision between fine particles and large particles was also probably a minimum. As for the configuration of upflow velocity=0.023 m/s, recycle ratio-20, seeding mass = 9.22g, size< 1.19mm, after 12 hours' continuous running, although the smaller particles in the top two sections did not change much in terms of quantity, the relatively large particles (>=0.15mm) in all the sections, except the bottom one, were reduced in number but the whole spectrum of particles had an increase in the bottom section; this was true especially for the smallest particles(0.1mm)(see Figure 5.16). It can be suggested that secondary nucleation still dominates the process while aggregation (more in the third and bottom sections) and molecular growth did occur in significant amounts, since there was a larger mass of seeding of much bigger size particles compared to the set up in Figure 5.15. When the upflow velocity was adjusted to a much higher value (= 0.030m/s) with a recycle ratio=21 and 15g of size< 0.149mm seeding(see Figure 5.17), the observation after 12 hours' growth is that the distribution shifted to the right (bigger size) on the horizontal axis, although not by much. In another words, particles in each section grew in size except in the top section, where only the number of the smallest particles (0.1mm) increased (all other larger-sized particles were absent). It is suspected that a significant amount of deadzone in the top section(s), contributed to the significant particle loss. While it is interesting to notice that, even though the second and third configurations provided an even lower SSR (because of same flow rate of feed but larger recycle ratio), there was more convincing growth in several aspects (nucleation, aggregation and molecular growth) for the second and third configurations, with a larger mass of seeding as well as a higher upflow velocity. At the same time, while the third configuration had a smaller seeding size, compared to the second configuration, the particles in the bottom section grew to a larger size over a same period of time; this was probably due to the higher mixing level of a larger upflow velocity and a denser bed (more seeding mass), thus providing a higher 39  RESULTS AND DISCUSSION  possibility of aggregation. The general feeling is that, the more evenly distributed the void fraction throughout the bed, the higher the growth rate.  40  ^  RESULTS AND DISCUSSION  Numbers of particles in a drop^ of sarr4)8%  At the 12th hour  35.0 30.0 25.0  Aliflifip  .  =Second  20.0  le Third  15X)  ■ Bottom  10.0 5.0 0.0 -  T^  cc' C:■)"^,z).  ,  _a,^AL  ,  ^  _a__ ^me__^it,_  cc'^cc`^<(`^<(` tS4'^Q.+"'^C34'^Qi.`"^c:■`"  0^0^0^(■^ 0^4, <0'Z' ^1\C. ''.• ^• 'N, 1,^°))^C.^41  c`^c.(` ,Z,'^,C). ^CbC  xlp range (k" Q,`"  O'Z'  ',.  -  -^— - -^  Numbers of particles in a drop^ of sample 40  12 hours before  35 30  ■ Top  25  ■ Second 20  ■ Third ■ Bottom  15 10 5 0  IL, 1001.1m 1501.1m 200mm 300gm 400um 500unn 600um  dp range 700um 8001Ltm 1000um  Figure 5.15 Size growth within 12 hours for upflow velocity =0.021m/s, recycle ratio=17.9, seeding 4.71g particles of size < 0.119mm.  41  RESULTS AND DISCUSSION  30 ^  At the 12th hour  Number of particles in a drop of sample 25  20  • Top 15  • Second • Third  10  • Bottom  ill  Q.••  dp range  -  ^r-  ----""-r-^--7^-•"---1  Ck's^'6's^tk\^(k.^fC^\^C(‘^.6'^CP .. , ^(:)• ^ps..)"^Q,N4'^Q,N)*^p`4'^oN)"^C) (;,)'^ (,\J C) ps N.^1.. ')>QI  ^eP^ ^(0°^AQ'^co°^N  4,  Number of particles in a drop of sample 30  12 hours before  25  • Top  20  ^  • Second Third  15  • Bottom  10  ' dp range  c` •  Figure 5.16 Size  •,›  caN'  ‹c•^c•^cc•^ c%^•c•^cc‘ (oN' <,p^(00^AO^(60^Q,0  growth within 12 hours for upflow velocity =0.023m/s, recycle ratio=20, seeded with 9.22g particles of size < 1.19mm. 42  RESULTS AND DISCUSSION  Number of particles in a drop of sample 20  At the 12th hour  a op -  _,a..■ econd  _  iv hird '  • I lottom  ■  m--,-  dr range  mi  fi^cc`^fi^fi^fi^fi^<(`^fi^fi^fi  os.'^Q.‘  z,''^c)•.'^o\)*^.z.\'^Q.' ^o'.)'^c)N ^Q.‘'  4.)^0^<1^Ci^0^0^0^0^0^0^0 .^,  .  .  ".^N,^'V^'13^tx^<1^Co^A'b^0 .>  Number of particles in a drop of sample 20.0 ^  12 hours before  18.0 16.0 14.0  ^'Top  12.0  Second  10.0 8.0  ^■ Third  6.0  ^■ Bottom  4.0 2.0  0.0  ^—rdp-range fi^cc's^fi^fi^fi^cc`^fi .0`" 0`'^ 0^0^0^0^0 <-)^<-)^0^0^0^ ^$,^°I,^1,^tx^4)^co^A^,b  Figure 5.17 Size growth within 12 hours for upflow velocity =0.030m/s, recycle ratio=21, seeded with 15g particles of size < 0.149mm.  43  RESULTS AND DISCUSSION  5.3 Application of research findings to a struvite production and recovery system: practical considerations For the optimization of a struvite production and recovery system, it is critical to develop a set of bed expansion equations, as well as to determine the minimum fluidization velocities, for a spectrum of particles in terms of size; this would allow us to achieve desired configurations for optimal void fractions and mixing level. Previously established equations in other research or this research, can be referenced; however, unless it is of the identical design, specific equation coefficients need to be revised according to each specific design based on experiments. In order to maintain an efficient fluidization and a vertical classification of differently- sized solids, it is recommended to pre-segregate seeding particles, since segregation can provide faster bed expansion and doesn't require higher pumping for minimum fluidization, compared to a well-mixed bed. In addition, the seeds need to be comprised of components with distinctive size differences, so as to guarantee the occurrence and maintenance of segregation. As for the reactor design, a circular cross-section, multi-compartmentalized, tapered transition reactor is still recommended, to reduce deadzone areas and to attenuate flow deceleration, for less particle loss and higher convective stability. To achieve higher growth in size, an adequate amount of seeding, with a portion of relatively large particles, is required; this will provide more contact area for molecular growth, as well as a higher possibility of aggregation. Therefore, it is required to know the minimum fluidization conditions and the bed expansion equations of each seeding configurations, in terms of size and composition; this would provide sufficient flow for effective fluidization, as well as a desired void fraction. As the struvite grows bigger in time, a higher flow should be provided, in order to prevent bed "collapse".  44  CONCLUSIONS  CHAPTER 6 - CONCLUSIONS The following conclusions are drawn, based on this research project performed on the twodimensional struvite reactor at UBC. •  Measuring the dynamic pressure drop at two ends of the fluidized bed and plotting the dynamic pressure drop as a function of the upflow velocity, is an experimentallyproven way to determine the minimum fluidization velocity for particles of a certain size and density, and for different mixtures, as well.  •  Dynamic pressure drop increases during the process of fluidizing the packed bed, until the upflow passes the minimum fluidization velocity. The dynamic pressure drop is a constant value for a fluidized bed, irrespective of bed expansion.  •  A seemingly perfect linear relationship exists between dynamic pressure drop (log) and void fraction (log) for fluidized beds, even with binary mixtures of particles of different sizes. However, the mathematical equation derived does not necessarily share similar constants or coefficients as the previously published research has suggested.  •  A binary mixture of particles is likely to go through segregation and once segregated, each of the two distinctive layers of particles, with different sizes, behaves independently like a monodisperse bed alone and goes through different, minimum fluidization episodes. However, the bed expansion equations derived independently for the two layers, eventually show convergence.  •  Overall, the fluidized-bed expands faster when segregated, which is indicated by the bigger slope of the linear line between upflow velocity (log) and void fraction (log), compared to the well-mixed bed expansion equation. Also the packed-bed, after  45  CONCLUSIONS  segregation, fluidizes at a lower minimum fluidization velocity than the one for the smaller size constituent. A segregated bed will not require more energy, but actually less energy, to have the same fluidization effect.  •  A binary mixture of particles, having insignificant size difference between the two constituents, is not likely to go through any segregation, even at surge flow conditions. Such a mixture has a minimum fluidization velocity that falls between the minimum fluidization velocities determined, respectively, with a monodisperse bed composed of each of the constituents.  •  A certain quantity of seeding has to be provided to have enough contact area for aggregation and molecular growth to happen. A fluidized bed, with more seeding and higher flow, can grow particles better even at a lower supersaturation ratio.  •  A large-scale, bulk circulation (due to the sudden deceleration at the transition between two neighbouring sections) was observed; it usually appears as upward flow in the central region and downward flow near the walls. This circulation likely aggravated the particle loss into the deadzones.  46  RECOMMENDATIONS FOR FURTHER RESEARCH  CHAPTER 7 - RECOMMENDATIONS FOR FURTHER RESEARCH The following recommendations are made, based on the knowledge gained from the study on fluidized bed hydrodynamics and struvite growth, as well as size distribution tests. •  In order to derive an empirical equation to predict the minimum fluidization velocity just based on knowledge of size, density and shape of the particles, bed fluidization experiments should be run for a wider spectrum of particles classified, with better grading precisions.  •  It would be useful to investigate the bed expansion phenomenon in detail on more hydrodynamic aspects, for the period after the bed surface exceeds the transition and rises into the next section above, with a larger cross-section area.  •  Various arrangements of mixtures should be tested extensively, to gain a more general understanding of segregation, as well as the hydrodynamic behaviour of a multidisperse bed.  •  Bed expansion should be studied with a wider range of void fraction or Reynolds numbers, so as to determine the predictive equation between upflow velocity and void fraction, in a more general sense.  47  RECOMMENDATIONS FOR FURTHER RESEARCH  •  In an operational context, a tapered transition, round cross-section reactor should be considered, for the optimal design, to achieve more convective stability and to reduce particle loss.  •  The operational window, in terms of seeding mass, size and upflow velocity, should be studied further, in order to develop an optimal operational protocol combined with the chemical knowledge already gained from previous research, by the UBC struvite recovery team.  48  REFERENCES  REFERENCES ADNAN, A. (2002). PILOT-SCALE STUDY OF PHOSPHORUS RECOVERY THROUGH STRUVITE CRYSTALLIZATION. DEPARTMENT OF CIVIL ENGINEERING. VANCOUVER, THE UNIVERSITY OF BRITISH COLUMBIA. MASTER OF APPLIED SCIENCE. ALI, M. I. AND SCHNEIDER, P. A. (2006). "AFED-BATCH DESIGN APPROACH OF STRUVITE SYSTEM IN CONTROLLED SUPERSATURATION." CHEMICAL ENGINEERING SCIENCE 61: 3951-3961. ASIF, M. AND PETERSEN, J. N. (1994). "A DYNAMIC MODEL OF THE HYDRODYNAMICS OF A LIQUID FLUIDIZED BED." IND.ENG.CHEM.RES. 33: 2151-2156. BATTISTONI, P., PAVEN, P. AND CECCHI, F. (1998). "PHOSPHATE REMOVAL IN REAL ANAEROBIC SUPERNATANT: MODELLING AND PERFORMANCE OF A FLUIDIZED-BED REACTOR." WAT.SCI.TCH 38(1): 275-283. BATTISTONIA, P., ANGELISA, A. AND PRISCIANDAROB, M. (2002). "P REMOVAL FROM ANAEROBIC SUPERNATANTS BY STRUVITE CRYSTALLIZATION: LONG TERM VALIDATION AND PROCESS MODELLING." WATER RESEARCH 36: 1927-1938. BRITTON, A. (2002). PILOT SCALE STRUVITE RECOVERY TRIALS FROM A FULL SCALE ANAEROBIC DIGESTER SUPERNATANT AT THE CITY OF PENTICTON ADVANCED WASTEWATER TREATMENT PLANT DEPARTMENT OF CIVIL ENGINEERING. VANCOUVER, BRITISH COLUMBIA, THE UNIVERSITY OF BRITISH COLUMBIA.  MASTER OF APPLIED SCIENCE. EPSTEIN, N. (2003). LIQUID-SOLID FLUIDIZATION. HANDBOOK OF FLUIDIZATION AND FLUID-PARTICLE SYSTEMS. W.-C. YANG. NEW YORK, USA, MARCEL DEKKER TAYLOR & FRANCIS - CRC: 1868. GIBILARO, L. G.. (2001). FLUIDIZATION DYNAMICS. L'AQUILA, BUTTERWORTHHEINEMANN.  HUANG, J.-S., YAN, J.-L. AND WU, C.-S. (2000). "COMPARATIVE BIOPARTICLE AND HYDRODYNAMIC CHARACTERISTICS OF CONVENTIONAL AND TAPERED ANAEROBIC FLUIDIZED-BED BIOREACTORS." JOURNAL OF CHEMICAL TECHNOLOGY AND BIOTECHNOLOGY 75: 269-278.  49  MIURA, H. AND KAWASE, Y. (1997). "HYDRODYNAMICS AND MASS TRANSFER IN THREEPHASE FLUIDIZED BEDS WITH NON-NEWTONIAN FLUIDS." CHEMICAL ENGINEERING SCIENCE 51(21/22): 4095 - 4104. OHLINGER, K. N., YOUNG, T. M. AND SCHROEDER, E. D. (1998). "PREDICTING STRUVITE FORMATION IN DIGESTION." WAT.RES 32(12): 3607-3614. OHLINGER, K. N., YOUNG, T. M. AND SCHROEDER, E. D. (2000). "POSTDIGESTION STRUVITE PRECIPITATION IN A FLUIDIZED BED REACTOR." JOURNAL OF ENVIRONMENTAL ENGINEERING APRIL: 361-368. RICHARDSON, J. F. AND ZAKI, W. N. (1954A). "SEDIMENTATION AND FLUIDIZATION." TRANS.INST.CHEM.ENG. 32. SECKLER, M. M., 0. S. L. B. AND G M. V. R. (1996). "CALCIUM PHOSPHATE PRECIPITATION IN A FLUIDIZED BED IN RELATION TO PROCESS CONDITIONS: A BLACK-BOX APPROACH." WAT.RES 30(7): 1677-1685. SECKLER, M. M., BRUINSMA, 0. S. AND ROSEMALEN, G. M. (1996). "PHOSPHATE REMOVAL IN A FLUIDIZED BED - IDENTIFICATION OF PHYSICAL PROCESSES." WAT.RES 30(7): 1585-1588. SHIMAMURA, K., TANAKA, T. AND MIURA, Y. (2003). "DEVELOPMENT OF A HIGHEFFICIENCY PHOSPHORUS RECOVERY METHOD USING A FLUIDIZED-BED CRYSTALLIZED PHOSPHORUS REMOVAL SYSTEM." WATER SCIENCE AND TECHNOLOGY 48(1): 163-170. STEEN, I. (1998). "PHOSPHORUS AVAILABILITY IN THE 21ST CENTURY: MANAGEMENT OF A NON-RENEWABLE RESOURCE." PHOSPHORUS AND POTASSIUM SEPTEMBEROCTOBER(217): 25-31. YOSHINO, M., YAO, M. AND TSUNO, H. (2003). "REMOVAL AND RECOVERY OF PHOSPHATE AND AMMONIUM AS STRUVITE FROM SUPERNATANT IN ANAEROBIC DIGESTION." WATER SCIENCE AND TECHNOLOGY 48(1): 171-178. ZHOU, J. J. AND LEE, S. M. (2000). "APPLICATION OF BENCH-SCALE TESTS IN UNDERSTANDING A COMMERCIAL FLUIDIZED-BED REACTOR OPERATION." IND.ENG.CHEM.RES. 43: 5460-5465.  50  51  APPENDIX A: HYDRODYNAMIC RUNS  52  Run for dp = 4mm solids Upflow velocity U (m/s) 0.00077 0.00151 0.00242 0.00327 0.00448 0.00509 0.00622 0.00708 0.00742 0.00764 0.00787 0.00880 0.00918 0.00926 0.00930 0.01087 0.01224 0.01484 0.01574 0.01860 0.02031 0.02204 0.02411 0.02530 0.02661 0.02858 0.03026 0.03283 0.03355 0.03507 0.03588 0.03858 0.04061 0.04286 0.04538 0.04823 0.05144 0.05273 0.05511 0.05935 0.06173  log(U)  Flow rate Q(L/s)  dynamic pressure drop (cm of water)  Loa^ (AD 1 •  bed height(cm)  log(L)  Void fraction  Log(E)  -3.113 -2.820 -2.617 -2.485 -2.349 -2.293 -2.206 -2.150 -2.129 -2.117 -2.104 -2.056 -2.037 -2.033 -2.032 -1.964 -1.912 -1.829 -1.803 -1.731 -1.692 -1.657 -1.618 -1.597 -1.575 -1.544 -1.519 -1.484 -1.474 -1.455 -1.445 -1.414 -1.391 -1.368 -1.343 -1.317 -1.289 -1.278 -1.259 -1.227 -1.210  0.00100 0.00196 0.00313 0.00424 0.00580 0.00660 0.00806 0.00917 0.00962 0.00990 0.01020 0.01140 0.01190 0.01200 0.01205 0.01408 0.01587 0.01923 0.02040 0.02410 0.02632 0.02857 0.03125 0.03278 0.03448 0.03703 0.03922 0.04255 0.04348 0.04545 0.04650 0.05000 0.05263 0.05555 0.05882 0.06250 0.06667 0.06833 0.07142 0.07692 0.08000  0.20 0.30 0.80 1.80 2.80 4.70 5.60 6.10 6.25 6.35 7.20 7.70 7.80 7.70 8.26 8.26 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72 5.72  -0.699 -0.523 -0.097 0.255 0.447 0.672 0.748 0.785 0.796 0.803 0.857 0.886 0.892 0.886 0.917 0.917 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757 0.757  17.9 17.9 17.9 17.9 17.9 17.9 17.9 17.9 17.9 18.0 18.0 17.9 18.0 18.0 18.1 18.4 18.8 19.2 19.8 20.6 21.6 22.3 23.2 23.8 24.8 24.9 25.5 26.3 27.0 27.7 28.3 29.2 30.3 31.6 32.8 34.5 35.9 37.8 39.1 40.9 41.9  1.253 1.253 1.253 1.253 1.253 1.253 1.253 1.253 1.253 1.255 1.254 1.253 1.254 1.254 1.258 1.266 1.275 1.284 1.298 1.314 1.334 1.349 1.366 1.376 1.394 1.397 1.406 1.420 1.432 1.442 1.451 1.466 1.482 1.499 1.516 1.538 1.556 1.578 1.592 1.612 1.622  0.407 0.407 0.407 0.407 0.407 0.407 0.407 0.407 0.407 0.409 0.408 0.407 0.408 0.408 0.413 0.422 0.432 0.443 0.458 0.476 0.497 0.512 0.529 0.538 0.556 0.558 0.567 0.580 0.590 0.598 0.606 0.618 0.630 0.644 0.657 0.672 0.684 0.699 0.708 0.721 0.727  -0.391 -0.391 -0.391 -0.391 -0.391 -0.391 -0.391 -0.391 -0.391 -0.389 -0.389 -0.391 -0.389 -0.389 -0.384 -0.375 -0.364 -0.354 -0.339 -0.322 -0.304 -0.290 -0.277 -0.269 -0.255 -0.253 -0.247 -0.237 -0.229 -0.223 -0.217 -0.209 -0.200 -0.191 -0.183 -0.172 -0.165 -0.156 -0.150 -0.142 -0.139  53  Run for dp = 2.9 mm solids Upflow velocity U(m/s)  log(U)  Flow rate Q (L/s)  dynamic pressure drop (cm of water)  Log(Op)  bed height (cm)  Void fraction  Log(E)  0.0008  -3.1126  0.0010  0.2  -0.6990  8.5  0.5630  -0.2495  0.0015  -2.8116  0.0020  0.5  -0.3010  8.5  0.5630  -0.2495  0.0022  -2.6502  0.0029  1  0.0000  8.5  0.5630  -0.2495  0.0034  -2.4692  0.0044  1.6  0.2041  8.5  0.5630  -0.2495  0.0045  -2.3432  0.0059  2.6  0.4150  8.5  0.5630  -0.2495  0.0056  -2.2523  0.0073  4.5  0.6532  8.5  0.5630  -0.2495  0.0065  -2.1883  0.0084  5.3  0.7243  8.6  0.5669  -0.2465  0.0073  -2.1340  0.0095  5.7  0.7559  8.9  0.5782  -0.2380  0.0078  -2.1083  0.0101  6.2  0.7924  9.1  0.5854  -0.2326  0.0084  -2.0752  0.0109  6.5  0.8129  9.3  0.5923  -0.2274  0.0087  -2.0595  0.0113  6.8  0.8325  9.45  0.5974  -0.2237  0.0091  -2.0422  0.0118  6.2  0.7924  9.6  0.6023  -0.2202  0.0094  -2.0262  0.0122  6.2  0.7924  9.7  0.6056  -0.2178  0.0098  -2.0105  0.0127  6.2  0.7924  9.9  0.6119  -0.2133  0.0089  -2.0523  0.0115  6.2  0.7924  9.7  0.6056  -0.2178  0.0112  -1.9515  0.0145  6.2  0.7924  10.25  0.6224  -0.2059  0.0127  -1.8951  0.0165  6.2  0.7924  10.9  0.6406  -0.1934  0.0146  -1.8368  0.0189  6.2  0.7924  11.5  0.6559  -0.1832  0.0167  -1.7761  0.0217  6.2  0.7924  12.3  0.6744  -0.1711  0.0191  -1.7199  0.0247  6.2  0.7924  12.9  0.6870  -0.1631  0.0208  -1.6812  0.0270  6.2  0.7924  13.7  0.7023  -0.1535  0.0224  -1.6505  0.0290  6.2  0.7924  14.3  0.7129  -0.1470  0.0245  -1.6110  0.0317  6.2  0.7924  15.1  0.7259  -0.1391  0.0264 0.0271  -1.5786 -1.5673  0.0342 0.0351  6.3 6.2  0.7993 0.7924  15.4 15.9  0.7304 0.7377  -0.1364 -0.1321  54  Run for dp = 2 mm solids  log(U)  Flow rate(L/s)  dynamic pressure drop (cm of water)  Log(Ap)  bed height (cm)  Void fraction E  Log(c)  0.00077  -3.1126  0.0010  0.8  -0.0969  16.8  0.5233  -0.2812  0.00154  -2.8116  0.0020  1.2  0.0792  16.8  0.5233  -0.2812  0.00224  -2.6502  0.0029  1.7  0.2304  16.8  0.5233  -0.2812  0.00334  -2.4761  0.0043  2.9  0.4624  16.4  0.5133  -0.2896  0.00468  -2.3301  0.0061  4.9  0.6902  16.4  0.5133  -0.2896  0.00540  -2.2675  0.0070  6.6  0.8195  17.0  0.5282  -0.2772  0.00607  -2.2166  0.0079  7.5  0.8751  17.4  0.5376  -0.2695  0.00671  -2.1731  0.0087  7.3  0.8633  17.7  0.5445  -0.2640  0.00725  -2.1395  0.0094  5.4  0.7324  18.4  0.5586  -0.2529  0.00795  -2.0998  0.0103  5.3  0.7243  18.8  0.5679  -0.2457  0.00829  -2.0812  0.0108  5.4  0.7324  19.1  0.5729  -0.2419  0.00895  -2.0481  0.0116  5.4  0.7324  19.6  0.5834  -0.2340  0.00918  -2.0371  0.0119  5.4  0.7324  19.8  0.5872  -0.2312  0.00934  -2.0298  0.0121  5.4  0.7324  19.9  0.5881  -0.2306  0.00980  -2.0088  0.0127  5.4  0.7324  21.0  0.6082  -0.2160  0.01026  -1.9888  0.0133  5.4  0.7324  24.0  0.6523  -0.1855  Upflow velocity U (m/s)  55  Run for dp = 0.9 mm solids Upflow velocity U (m/s)  log(U)  Flow rate(L/s)  dynamic dy pressure drop (cm of water)  Log(4p)  bed height (cm)  Void fraction E  Log(E)  0.00034  -3.4692  0.0004  1.1  0.0414  14.5  0.3462  -0.4607  0.00075  -3.1251  0.0010  3.5  0.5441  14.5  0.3466  -0.4602  0.00126  -2.8986  0.0016  5.3  0.7243  14.6  0.3492  -0.4569  0.00165  -2.7829  0.0021  7.3  0.8633  15.1  0.3658  -0.4367  0.00200  -2.6993  0.0026  6.1  0.7853  15.9  0.3933  -0.4053  0.00247  -2.6068  0.0032  5.6  0.7482  16.5  0.4139  -0.3831  0.00294  -2.5309  0.0038  5.5  0.7404  17.3  0.4361  -0.3604  0.00351  -2.4551  0.0045  5.5  0.7404  18.0  0.4566  -0.3405  0.00482  -2.3167  0.0063  5.6  0.7482  19.7  0.4980  -0.3028  0.00572  -2.2430  0.0074  5.6  0.7482  20.7  0.5185  -0.2852  0.00654  -2.1845  0.0085  5.6  0.7482  21.7  0.5393  -0.2682  0.00721  -2.1420  0.0093  5.6  0.7482  22.5  0.5531  -0.2572  0.00764  -2.1169  0.0099  5.6  0.7482  22.9  0.5601  -0.2517  0.00796  -2.0993  0.0103  5.6  0.7482  23.2  0.5653  -0.2478  0.00848  -2.0717  0.0110  5.6  0.7482  23.4  0.5694  -0.2446  0.00867  -2.0620  0.0112  5.6  0.7482  23.7  0.5743  -0.2408  0.00887  -2.0521  0.0115  5.6  0.7482  23.8  0.5756  -0.2399  56  Run for the mixture of dp=4mm and dp=2mm solids, completely-mixed  Upflow velocity U(m/s)  log(U)  Flow rate(L/s)  0.00028 0.00077 0.00121 0.00154 0.00206 0.00255 0.00301 0.00338 0.00397 0.00451 0.00522 0.00547 0.00607 0.00637 0.00665 0.00692 0.00739 0.00772 0.00810 0.00838 0.00849 0.00856 0.00933 0.01162 0.01206 0.01286 0.01715 0.01975 0.02031 0.02204 0.02411 0.02530 0.02661 0.02858 0.03026 0.03283 0.03355  -3.5456 -3.1126 -2.9167 -2.8116 -2.6861 -2.5941 -2.5215 -2.4711 -2.4008 -2.3462 -2.2827 -2.2620 -2.2166 -2.1956 -2.1771 -2.1598 -2.1312 -2.1126 -2.0914 -2.0768 -2.0712 -2.0673 -2.0302 -1.9348 -1.9186 -1.8907 -1.7659 -1.7044 -1.6924 -1.6567 -1.6178 -1.5970 -1.5750 -1.5440 -1.5191 -1.4837 -1.4743  0.00037 0.00100 0.00157 0.00200 0.00267 0.00330 0.00390 0.00438 0.00515 0.00584 0.00676 0.00709 0.00787 0.00826 0.00862 0.00897 0.00958 0.01000 0.01050 0.01086 0.01100 0.01110 0.01209 0.01506 0.01563 0.01667 0.02222 0.02560 0.02632 0.02857 0.03125 0.03278 0.03448 0.03703 0.03922 0.04255 0.04348  Op(cm)  Log(Ap)  bed height(cm) before  Void fraction c  0.45 0.75 1.25 1.55 2.35 3.15 3.9 5.15 6.75 7.8 9.35 12 10.35 10.35 10.55 10.55 10.6 10.6 10.6 10.6 10.6 10.6  -0.3468 -0.1249 0.0969 0.1903 0.3711 0.4983 0.5911 0.7118 0.8293 0.8921 0.9708 1.0792 1.0149 1.0149 1.0233 1.0233 1.0253 1.0253 1.0253 1.0253 1.0253 1.0253  15.29 15.29 15.29 15.29 15.3 15.3 15.3 15.3 15.34 15.58 15.8 15.95 16.25 16.4 16.54 16.57 16.75 16.95 17.15 17.25 17.31 17.34  0.4522 0.4522 0.4522 0.4522 0.4525 0.4525 0.4525 0.4525 0.4537 0.4609 0.4674 0.4717 0.4801 0.4842 0.4880 0.4888 0.4935 0.4987 0.5038 0.5063 0.5078 0.5085  Log(c)  -0.3447 -0.3447 -0.3447 -0.3447 -0.3444 -0.3444 -0.3444 -0.3444 -0.3432 -0.3364 -0.3303 -0.3263 -0.3187 -0.3150 -0.3116 -0.3109 -0.3067 -0.3022 -0.2978 -0.2956 -0.2943 -0.2937  57  Run for the mixture of dp = 4mm and dp=2mm solids, after segregation  Flow rate (L/s)  Ap (cm  0.0004 0.0010 0.0016 0.0020 0.0027 0.0033 0.0039 0.0044 0.0052 0.0058 0.0068 0.0071 0.0079 0.0083 0.0086 0.0090 0.0096 0.0100 0.0105 0.0109 0.0110 0.0111 0.0121 0.0151 0.0156 0.0167 0.0222 0.0256 0.0263 0.0286 0.0313 0.0328 0.0345 0.0370 0.0392 0.0426  0.35 0.65 1 1.4 1.8 2.6 3.5 4.5 5 6 6.5 6.9 7.4 7.5 8 8 8 7.5 7.5 7.5 8 8.5 8.5 8.5 8.5 9.5 9.5 9.5 9.5 10.5 10.5 11.5 12.5 15.5 16.5 17.5  -0.456 -0.187 0.000 0.146 0.255 0.415 0.544 0.653 0.699 0.778 0.813 0.839 0.869 0.875 0.903 0.903 0.903 0.875 0.875 0.875 0.903 0.929 0.929 0.929 0.929 0.978 0.978 0.978 0.978 1.021 1.021 1.061 1.097 1.190 1.217 1.243  0.0435  17.5  1.243  Lo)(Ap  Bed heigh t (cm)  Void fractio nc  Log(c)  16.6 16.65 16.7 16.7 16.7 16.7 16.7 16.7 16.65 16.75 16.95 17.3 17.5 17.7 18.25 18.6 19.6 20.6 22.5 24.2 25.4 26.6  0.490 0.491 0.492 0.492 0.492 0.492 0.492 0.492 0.491 0.494 0.499 0.508 0.512 0.517 0.530 0.538 0.559 0.578 0.610 0.634 0.650 0.664  -0.310 -0.309 -0.308 -0.308 -0.308 -0.308 -0.308 -0.308 -0.309 -0.307 -0.302 -0.295 -0.290 -0.286 -0.276 -0.269 -0.253 -0.238 -0.215 -0.198 -0.187 -0.178  L of top layer (cm) 9.1 9.15 9.2 9.2 9.1 9.1 9.1 9.1 9.15 9.05 9.15 9.5 9.7 10 10.55 10.8 11.1 12.2 13.6 14.5 15.4 16.4  log( ^  s  s  L of botto  log(EIL  m  layer (cm) 0.492 0.495 0.498 0.498 0.492 0.492 0.492 0.492 0.495 0.489 0.495 0.513 0.524 0.538 0.562 0.572 0.584 0.621 0.660 0.681 0.700 0.718  -0.308 -0.306 -0.303 -0.303 -0.308 -0.308 -0.308 -0.308 -0.306 -0.310 -0.306 -0.289 -0.281 -0.269 -0.250 -0.243 -0.234 -0.207 -0.180 -0.167 -0.155 -0.144  7.5 7.5 7.5 7.5 7.6 7.6 7.6 7.6 7.5 7.7 7.8 7.8 7.8 7.7 7.7 7.8 8.5 8.4 8.9 9.7 10 10.2  0.487 0.487 0.487 0.487 0.492 0.492 0.492 0.492 0.487 0.497 0.502 0.502 0.502 0.497 0.497 0.502 0.533 0.529 0.550 0.579 0.589 0.595  -0.312 -0.312 -0.312 -0.312 -0.308 -0.308 -0.308 -0.308 -0.312 -0.303 -0.299 -0.299 -0.299 -0.303 -0.303 -0.299 -0.273 -0.276 -0.260 -0.237 -0.230 -0.225  58  Run for the mixture of dp = 2mm and dp = 0.9mm  U pflow velocity U (m/s)  log(U)  flowrate (L/s)  dynamic pressure drop(cm of water)  Log(Op)  bed height (cm)  log(L)  Void fraction  Log(E)  0.00049  -3.3133  0.00063  0.8  -0.0969  13.4  1.1271  0.4104  -0.3868  0.00111  -2.9542  0.00144  1.7  0.2304  13.4  1.1271  0.4104  -0.3868  0.00170  -2.7702  0.0022  3.3  0.5185  13.4  1.1271  0.4104  -0.3868  0.00231  -2.6369  0.00299  5  0.6990  13.35  1.1255  0.4086  -0.3887  0.00279  -2.5551  0.00361  6.5  0.8129  13.35  1.1255  0.4086  -0.3887  0.00329  -2.4822  0.00427  8.5  0.9294  13.35  1.1255  0.4086  -0.3887  0.00375  -2.4260  0.00486  9.2  0.9638  14  1.1461  0.4317  -0.3648  0.00445  -2.3514  0.00577  5.5  0.7404  14.5  1.1614  0.4483  -0.3484  0.00515  -2.2885  0.00667  5.5  0.7404  14.9  1.1732  0.4609  -0.3364  0.00600  -2.2216  0.00778  5.5  0.7404  15.5  1.1903  0.4788  -0.3198  0.00701  -2.1540  0.00909  5.5  0.7404  16.1  1.2068  0.4955  -0.3049  0.00772  -2.1126  0.01  5.5  0.7404  16.6  1.2201  0.5087  -0.2936  0.00877  -2.0572  0.01136  5.5  0.7404  17.35  1.2393  0.5271  -0.2781  59  APPENDIX B: SIZE GROWTH TESTS  60  Run for upflow velocity =0.021m/s, recycle ratio=17.9, seeding 4.71g particles of size < 0.119mm. At start-up Top section  Sample # 1 2 3 4 Average  100pm  150pm  200pm  300pm  400pm  500pm  600pm  700pm  800pm  1000pm  20 19  2 3  0 0  0 0  0 0  0 0  0 0  0 0  0 0  0 0  9 15 15.75  3 0 2  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  22 22 12 8 16  1 0 3 3 1.8  0 0 0 0 0  0 0 0 0 0  0 0 0 0 0  0 0 0 0 0  0 0 0 0 0  0 0 0 0 0  0 0 0 0 0  0 0 0 0 0  15 11 6 7 9.8  1 1 2 3 1.8  1 0 1 1 0.8  0 5 0 0 1.3  0 0 2 0 0.5  0 0 0 0 0.0  0 0 0 0 0.0  0 0 0 0 0.0  0 0 0 0 0.0  0 0 0 0 0.0  17 18 14 10 11 14 14  1 1 3 0 4 5 2.3  0 1 2 0 0 0 0.5  0 0 0 0 1 0 0.2  0 0 0 1 0 0 0.2  0 5 3 0 0 0 1.3  0 0 0 0 0 0 0.0  0 0 0 1 0 0 0.2  0 0 0 0 0 0 0.0  0 0 0 0 0 0 0.0  Second section  1 2 3 4 Average Third section  1 2 3 4 Average Bottom section  1 2 3 4 5 6 Average  61  ^ ^  Run for upflow velocity =0.021m/s, recycle ratio=17.9, seeding 4.71g particles of size < 0.119mm. After 12 hours  Top section  Sample 1 2 3 4 Average  100pm 150pm 200pm 300pm 400pm 500pm 600pm 700pm 800pm 1000pm 20^2^0^0^0^0^0^0^0^0 11^1^0^0^0^0^0^0^0^0  14^1^0^0^0^0^0^0^0^0 20^0^0^0^0^0^0^0^0^0 16.3^1^0^0^0^0^0^0^0^0  Second section  1 2 3 4 Average  12^0^0^0^0^0^0^0^0^0 16^2^0^1^0^0^0^0^0^0 15^0^0^0^0^0^0^0^0^0 22^0^0^0^0^0^0^0^0^0 16.3^0.5^0.0^0.3^0.0^0.0^0.0^0.0^0.0^0.0  Third section  1 2 3 4 Average  15^4^0^0^0^0^0^0^0^0 11^1^1^0^0^0^0^0^0^0 15^3^0^0^0^0^0^0^0^0 26^5^0^0^0^0^0^0^0^0 16.8^3.3^0.3^0.0^0.0^0.0^0.0^0.0^0.0^0.0  Bottom section  1 2 3 4 5 6 Average  43^7^5^0^0^4^1^0^0^0 36^11^0^2^2^2^2^0^0^0 46^11^2^2^6^0^0^4^0^0 35^4^2^3^0^0^0^1^0^0 28^6^1^0^0^0^0^0^0^0 40^5^2^2^4^1^1^0^0^0 38.0^7.3^2.0^1.5^2.0^1.2^0.7^0.8^0.0^0.0  62  Run for upflow velocity =0.023m/s, recycle ratio=20, seeded with 9.22g particles of size < 1.19mm. At start-up Top section Sample 1 2 3 4 Average  100pm  150pm  200pm  300pm  400pm  500pm  600pm  700pm  800pm  1000pm  7 10 11  0 1  0 0  0 0  0 0  0 0  0 0  0 0  0 0  0 0  6 8.5  2 2 1.25  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  0 14 9 8 7.8  0 1 1 0 0.5  1 0 0 0 0.3  0 0 0 0 0.0  0 0 0 0 0.0  0 0 0 0 0.0  0 0 0 0 0.0  0 0 0 0 0.0  0 0 0 0 0.0  0 0 0 0 0.0  10 15 9 20 23 15.4  2 1 1 2 1 1.4  0 0 0 0 0 0  0 0 0 0 1 0.2  0 0 0 1 0 0.2  0 0 0 0 0 0  0 0 0 0 0 0  0 0 0 0 0 0  0 0 0 0 0 0  0 0 0 0 0 0  25 22 34 29 20 25 25.8  1 2 3 8 4 0 3.0  1 0 0 4 2 2 1.5  0 1 0 0 0 0 0.2  1 0 0 0 0 0 0.2  0 0 0 0 0 0 0.0  0 0 0 0 0 0 0.0  0 0 0 0 0 0 0.0  0 0 0 0 0 0 0.0  0 0 0 0 0 0 0.0  Second section  1 2 3 4 Average Third section  1 2 3 4 5 Average Bottom section  1 2 3 4 5 6 Average  63  ^  Run for upflow velocity =0.023m/s, recycle ratio=20, seeded with 9.22g particles of size < 1.19mm. After 12 hours Top section  Sample # 1 2 3 4 Average  100pm 150pm 200pm 300pm 400pm 500pm 600pm 700pm 800pm 1000pm  2^3^0^0^0^0^0^0^0^0 6^1^0^0^0^0^0^0^0^0 4^0^0^0^0^0^0^0^0^0 4^3^0^0^0^0^0^0^0^0 4^1.8^0^0^0^0^0^0^0^0  Second section  1 2 3 4 Average  7^1^0^0^0^0^0^0^0^0 15^0^0^0^0^0^0^0^0^0 12^1^0^0^0^0^0^0^0^0 18^2^0^0^0^0^0^0^0^0 13.0^1.0^0.0^0.0^0.0^0.0^0.0^0.0^0.0^0.0  Third section  1 2 3 4 Average  9^2^0^0^0^0^1^0^0^0 18^1^0^0^0^0^0^0^0^0 17^1^0^0^0^0^0^0^0^0 27^2^0^0^1^0^0^0^0^0 17.8^1.5^0.0^0.0^0.3^0.0^0.3^0.0^0.0^0.3  Bottom section  1 2 3 4 5 6 Average  17^1^1^0^0^0^0^1^0^0 16^3^1^0^0^0^0^0^0^0 27^6^0^0^0^0^0^0^0^0 19^5^0^0^0^0^0^0^0^0 12^2^0^0^0^0^0^0^0^0 10^1^2^0^0^0^0^1^1^1 16.8^3.0^0.7^0.0^0.0^0.0^0.0^0.3^0.2^0.2  64  ^  Run for upflow velocity =0.030m/s, recycle ratio=21, seeded with 15g particles of size < 0.149mm. At start-up Top section Sample # 1 2 3 4 Average  50um 100pm 150pm 200pm 300pm 400pm 500pm 600pm 700pm 800pm 1000pm  8^6^2^0^0^0^0^0^0^0^0 6^8^1^0^0^0^0^0^0^0^0 5^2^1^0^0^0^0^0^0^0^0 9^3^0^0^0^0^0^0^0^0^0 7.0^4.8^1^0^0^0^0^0^0^0^0  Second section  1 2 3 4 Average  10^2^2^0^0^0^0^0^0^0^0 12^3^0^0^0^0^0^0^0^0^0 23^10^4^1^0^0^0^0^0^0^0 24^6^0^0^0^0^0^0^0^0^0 17.3^5.3^1.5^0.3^0.0^0.0^0.0^0.0^0.0^0.0^0.0  Third section  1 2 3 4 Average  5^3^1^0^0^0^0^0^0^0^0 6^6^0^0^0^0^0^0^0^0^0 4^0^1^0^0^0^1^0^0^0^0 0^3^0^0^0^0^0^0^0^0^0 3.8^3.0^0.5^0.0^0.0^0.0^0.3^0.0^0.0^0.0^0.0  Bottom section  1 2 3 4 Average  25^3^1^0^0^0^0^0^0^0^0 25^5^3^0^0^0^0^0^0^0^0 0^0^1^0^0^0^0^0^0^0^0 10^2^0^0^1^0^0^0^0^0^0 15.0^2.5^1.3^0.0^0.3^0.0^0.0^0.0^0.0^0.0^0.0  65  Run for upflow velocity =0.030m/s, recycle ratio=21, seeded with 15g particles of size < 0.149mm. After 12 hours Top section Sample # 1 2 3 4 Average  50^100 um^pm  150pm 200pm 300pm 400pm 500pm 600pm 700pm 800pm 1000pm  15^6^0^0^0^0^0^0^0^0^0 20^2^0^0^0^0^0^0^0^0^0 20^11^1^0^0^0^0^0^0^0^0 20^5^0^0^0^0^0^0^0^0^0 18.8^6.0^0.25^0^0^0^0^0^0^0^0  Second section 1 2 3 4 Average  0^22^0^0^0^0^0^0^0^0^0 0^14^3^2^0^0^0^0^0^0^0 0^9^3^0^0^0^0^0^0^0^0 0^17^8^1^0^0^0^0^0^0^0 0.0^15.5^3.5^0.8^0.0^0.0^0.0^0.0^0.0^0.0^0.0  Third section 1 2 3 4 Average  0^6^2^0^0^0^0^0^0^0^0 0^8^1^0^0^0^0^0^0^0^0 0^15^1^0^0^0^0^0^0^0^0 0^7^2^0^0^0^0^0^0^0^0 0.0^9.0^1.5^0.0^0.0^0.0^0.0^0.0^0.0^0.0^0.0  Bottom section 1 2 3 4 Average  0^9^1^0^0^0^0^2^0^2^0 0^13^0^0^0^0^0^0^0^0^0 0^6^1^0^1^0^0^0^0^0^0 0^19^2^0^0^0^0^0^0^0^0 0.0^11.8^1.0^0.0^0.3^0.0^0.0^0.5^0.0^0.5^0.3  66  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0063064/manifest

Comment

Related Items