Modelling of an industrial fluidised bed dryer

1. Introduction 1.1 Historical background of fluidisation Although the technique of fluidisation was in commercial use as early as 1926 for the gasification of coal, it was not until the early 1940s that its widespread use began with the construction of the first fluidised bed catalytic cracker (FCC). Since then, under commercial and wartime pressure, together with compromise between daring innovation and the need to reduce catalyst losses, the design of fluidised bed has been steered away from high velocity (upflow) mode towards low velocity (downflow) operation. More than 350 FCC units have been built and most are still in operation. In the late 1940s, the technique was successfully applied to the roasting of sulphide ores and since that time, virtually all new ore roasters have been fluidised beds. Fluidised bed dryers also made rapid progress and by the mid 1950s, the technique was well established. The Sohio process for making acrylonitrile in a fluidised bed was immensely successful; since 1960 virtually all new acrylonitrile plants have been fluidised beds and 50 large units are in operation worldwide. Undoubtedly, the major success in the late 1970s and 1980s is the Union Carbide polyethylene synthesis process. The alternative high pressure liquid phase reactors were limited in scale of operation, whereas the low pressure gas phase fluidised bed units can be built as large as required; the better quality product and dramatic reduction in costs which are features of this fluidised bed process have ensured the demise of virtually all competitors. Also in the 1970s and 1980s, fluidised bed combustion has attracted much attention largely due to its relatively low temperature operation (800 to 900°C) and its ability to absorb sulphur dioxide (SO2) through the use of limestone or dolomite. These features mean that nitrogen oxides (NOx) and SO2 emissions in the flue gases can be made acceptably low. More than 2,000 fluidised combustors are in use worldwide - on a wide variety of duties including burning of plastic waste, providing hot gases for drying grass, and raising steam for process use (Geldart, 1986). A very comprehensive account of the historical developments since the first commercial FCC fluidised bed in 1940s as well as problems encountered and tackled during these developments is summarised by Grace and Matsen (1980). Fluidised beds are frequently selected as processing tools because of their excellent heat transfer properties and because they permit controlled transfer of solids into, out of, and within the system. There are probably more fluidised beds used as dryers than in any other single application (Geldart, 1986). The first commercial unit was installed in the USA in 1948 to dry dolomite or blast furnace slag (Zahed et al., 1995). Since then, hundreds of fluidised bed dryers have operated worldwide primarily for granular materials that can be easily fluidised, such as sand, grains, chemical crystals and fertilisers. Vanecek et al. (1966) provides an extensive survey of different materials (granular, in solutions, suspensions and pastes) that are dried using fluidised beds. 1.2 Drying of small particles Drying of small particles is an important industrial application to process raw materials, intermediates and finished products. Depending on the particle size and the nature of feed, the preferred mode is usually either spray drying (for particle size between 10 to 500 µm), flash drying (between 10 and 3,000 µm) or fluidised drying (between 50 and 5,000 µm). While the spray dryers and flash dryers are capable of taking liquid feed, fluidised beds are generally operated with wet solid feed (Chandran et al., 1990). The advantages that fluidised beds offer are a large transfer contact area between solids and gas phases, high heat and mass transfer between phases, small surface area required by heat exchangers within the beds, a high degree of mixing of materials, ease in controlled handling and transport of fluidised materials (important for fragile crystals), fluids with negligible temperature and concentration gradient within the beds, lack of moving parts other than feeding and discharge mechanisms leading to high reliability and low maintenance costs, low construction costs and easy operability. These advantages make fluidised beds suitable for large-scale operations (Kunii and Levenspiel, 1991; Geldart, 1986). Some disadvantages of fluidised beds are the absence of temperature and concentration gradients within the bed when desired, high pressure drop, attrition of solids and erosion of the containing surfaces leading to entrainment (Chandran et al., 1990). Another technical drawback is the large number of sequential steps required in order to build and operate an industrial fluidised bed dryer. Although much more is now known about fluidised beds and their widespread commercial applications, it was reported that the only safe basis for such designs still remains pilot-plant drying test work combined with industrial experience (Devahastin, 2000; Mujumdar, 1995). Results from the cold model of a laboratory-scale fluidised dryer need adaptation on a pilot-scale before a commercial unit can be commissioned. It is difficult to describe the flow of gas, as there could be large deviations from the ideal plug flow particularly for bubbling beds of fine particles (Kunii and Levenspiel, 1991; Edwards and Avidan, 1986; Geldart, 1986). In practice, good economic factors of low construction costs and easy operability have been favoured over this technical drawback. Various designs of fluidised beds have been developed to suit many types of wet solids that can be fluidised by hot gas. Several designs of conventional fluidised bed dryers are shown in (Kunii and Levenspiel, 1991). Inorganic materials, such as dolomite are usually dried in single- bed furnace because the residence time characteristics of the particles to be dried are not important. For particles requiring nearly equal drying times, the residence time characteristics of single-stage operations can be narrowed using multi-stage dryers with vertical partition plates. Very sensitive materials, such as some pharmaceuticals, may require identical drying times for all particles. The use of rotating distributors ensures an ideal batch-continuous treatment of the particles. For temperature-sensitive materials, the inlet gas temperature must be kept low. To reduce loss in thermal efficiency, heat is recovered from the exiting dry solid stream in a two-stage dryer. When the feedstock is very wet, particles tend to agglomerate and there is difficulty in fluidising. A back-mix dryer can be employed, followed by plug flow dryer. By increasing heat transfer by adding in heat exchange tubes or plates within the fluidised bed, the volume of fluidising gas needed can be greatly reduced, resulting in smaller pumping cost, less particle attrition and lower construction cost of the exhaust gas cleaning system. Certain materials are not suited for the ordinary fluidised bed dryer and need special treatment, for example, cohesive and sticky solids that agglomerate or stick to dryer surfaces. For such materials, the vibro-fluidised bed provides the solution. The hot air distributor vibrates in such a way that facilitates the conveying of particles across a shallow bed from entrance to exit. Spouted bed dryers are often used for large uniformly sized particles, such as beans, peas, which are often difficult to fluidise. Comparatively small particles of minerals or salts that are only surface wetted, requiring very short drying times, can be effectively dried in lean-phase fluidised beds or in pneumatic transport lines, called flash dryers (Kunii and Levenspiel, 1991). Fluidised beds can also be used for making dry powder from a feed that is a slurry or solution. The feed is sprayed onto the bed, usually with a pneumatic atomiser to give a very fine atomisation. The particles in the bed are continually growing, by coating or agglomeration in the presence of liquid as binder. If a continuous process is applied, the product stream is continually withdrawn and classified into required size ranges. The fines and crushed oversize are returned to the bed for further growth, while the product of the required size is taken off (Geldart, 1986). The use of fluidised bed drying for granular materials is well established throughout the food and chemical processing industry. However, design and scale-up of fluidised bed dryers are seldom found in open literature. Bahu (1994) reported using a combined material and equipment model to address scale-up problems. In contrast with technology development in industry, fundamental research on fluidised bed drying has not made similar progress (Theologos et al., 1997; Mujumdar, 1995). 1.3 Objective The objective of this work is to develop a comprehensive and yet practical model to predict the performance of a typical industrial-scale fluidised bed dryer. The challenge is to build a model that incorporates intensive theoretical gas-solids flow patterns and yet retains a certain degree of simplicity and straightforwardness for practical design as well as operational control and optimisation. Although many researchers have conducted experiments or theoretical studies in the fields of fluidisation and drying, so far, there are few attempts to combine proven theoretical equations describing fluidisation patterns with drying models to model an industrial-scale dryer. They are further discussed in Chapter 2.3. In classical textbooks on unit operations (McCabe et al., 1985), on drying (Keey, 1972 & 1978; Mujumdar, 1995) and on fluidisation (Kunii and Levenspiel, 1991; Geldart, 1986; Davidson et al., 1985), the performance of fluidised bed dryers is predicted using single-phase drying models. In such kinetic models, fluidisation is assumed to be homogeneous (single-phase). This method has a major advantage in that it is straightforward and can be used to estimate the drying rate. However, during visual observation of drying operations, the above assumption often does not hold because gas-solids fluidisation most commonly occurs in a bubbling mode (two- or threephase). It is attempted in this work to develop theoretical criteria to predict when the single-phase drying model no longer holds and a more comprehensive model with twophase fluidisation pattern should be applied. A realistic drying model, incorporating well-established gas-solids two-phase fluidisation patterns, will be developed, and applied to the optimum operating performance of an existing industrial-scale fluidised bed dryer for nylon-6.6 particles. 1.4 Organisation of thesis This thesis is organised into six chapters. Chapter provides an introduction to the historical development of fluidisation and how it is applied to the drying of solids in industries. The objective of this work is defined. Chapter outlines the recent approaches in the modelling of fluidised bed drying. These approaches can be divided into two types of models: single-phase and two- or three-phase fluidisation. Chapter describes the design of a laboratory-scale experimental set-up which is used to study the fluidisation and drying behaviour of two particle systems. It includes experiments to measure the drying of internal moisture in single particles using thermal gravimetric analysis (TGA). Chapter shows the application of theoretical drying models to predict the drying rates in laboratory-scale fluidised bed dryer. Experimental results are compared and discussed with these theoretical calculations. In Chapter 5, specifications of the existing industrial-scale fluidised bed dryer and the operating conditions are stated. The modelling approach incorporating well- established two-phase fluidisation patterns into a drying model is described in detail. A solution strategy to solve two coupled mass transport differential equations iteratively is developed. In Chapter 6, the modelling approach developed earlier is used to simulate the performance of the industrial dryer. To optimise the plant operation, the sensitivity of operating parameters (temperature, weir height, fluidisation velocity, residence time) is studied extensively using the two-phase drying model. Arising from the optimisation studies, some recommendations are proposed. To achieve even higher performance, some design modifications to the dryer are also suggested. In Chapter 7, conclusions are drawn to summarise the modelling results in this work. 2. Literature review 2.1 Introduction Fluidised beds are frequently selected as processing equipment because of their excellent heat transfer properties and the ease in controlled transfer of solids into, out of, and within the process system. In gas-solids fluidisation, more fluidised beds have been used in drying operations than in any other single application (Geldart, 1986). However, in contrast with technology development in industry, the fundamental research on fluidised bed drying has not made similar progress (Theologos et al., 1997; Mujumdar, 1995). Much difficulty arises from the modelling of complicated mass transfer between two or three different phases within the same bed. Today’s process design still requires extensive laboratory and pilot-scale testing and verifications (Vanecek et al., 1966; Davidson et al., 1985; Geldart, 1986; Kunii and Levenspiel, 1991; Mujumdar, 1995; Devahastin, 2000). Batch drying curves obtained from laboratory-scale experiments for at least one set of operating conditions are extrapolated to predict drying performance in large-scale operations and other operating conditions. Neglecting actual fluidisation hydrodynamics, total heat and mass balances are often taken across the whole dryer to estimate the drying rate (Vanecek et al., 1966). 2.2 General drying model In general, the rate of drying of wet solids in flowing air depends on drying conditions in a complicated way. Van Meel (1958) first postulated that, for most engineering purposes, the drying rate can be described with reasonably accuracy by the following general drying equation: − dX = β ⋅ U ⋅ g(X) dt (2.1) The first term on the left hand side (LHS) describes the rate of change in moisture content of solids over time or the rate of drying. The factors affecting the rate of drying are divided into three terms on the right hand side (RHS). They are the mass transfer coefficient, β , which depends on the airflow conditions or contacting pattern between air and solids; the driving force or drying potential, U, which is a function of temperature and/or humidity of the drying air, and finally g(X), a dimensionless function of the average moisture content in solid particles, which takes into account the properties of the material to be dried. When the moisture content is greater than a critical moisture content, g(X) is equal to unity as the material properties not limit the rate of drying. This critical moisture content is dependent on the type of material and can also be dependent on the airflow conditions. In most cases, β and g(X) must be obtained from laboratory drying experiments under conditions as close as possible to the actual conditions. U can be defined as a function of either humidity and/or temperature. In order to describe g(X), Keey and Suzuki (1974) developed a semi-empirical characteristic drying curve method, which enabled material data obtained from one set of experiments to be approximated and extrapolated for other conditions. An extensive overview of theoretical drying models for slabs has been presented by Coumans (2000). Numerous recent theoretical works in fluidised bed drying adopted these drying models (to model the g(X)-term) for spherical particles and coupled them with 10 Drying rate [kg water / s ] 0.0270 0.0250 0.0230 0.0210 0.0190 0.0170 0.0150 1.2 1.5 No of times of minimum fluidisation velocity T65 Height 0.1m T65 Height 0.2m T65 Height 0.3m T65 Height 0.5m T65 all heights (single-phase) Fig. 6.1 Drying rates with air entry temperature at 65°C in Section Drying rate [kg water / s ] 0.0380 0.0360 0.0340 0.0320 0.0300 0.0280 0.0260 0.0240 0.0220 0.0200 1.2 1.5 No of times of minimum fluidisation velocity T85 Height 0.1m T85 Height 0.2m T85 Height 0.3m T85 Height 0.5m T85 all heights (single-phase) Fig. 6.2 Drying rates with air entry temperature at 85°C in Section 111 Drying rate [kg water / s] 0.0170 0.0160 0.0150 0.0140 0.0130 0.0120 0.0110 0.0100 0.0090 0.0080 1.2 1.5 No. of times of minimum fluidisation velocity T25 Height 0.1m T25 Height 0.3m T25 all heights (single-phase) T25 Height 0.2m T25 Height 0.5m Fig. 6.3 Drying rates with air entry temperature at 25°C in Section Drying rate [kg water / s] 0.1200 0.1000 0.0800 0.0600 0.0400 0.0200 0.0000 1.2 1.5 No. of times of minimum fluidisation velocity Section T65 Height 0.2m Section T85 Height 0.2m Section T65 Height 0.2m Section T85 Height 0.2m Section T25 Height 0.2m Fig. 6.4 Comparison of drying rates from two-phase model at bed height of 0.2 m 112 From the above results, some conclusions can be drawn. The drying rate of surface water increases with the velocity of the fluidising air stream. Till the velocity reaches 1.5 times the minimum fluidisation velocity, the increase in drying rate is not hampered by the bubble phase. This can be seen from Fig. 6.1 to 6.3, where the predicted drying rates using the single-phase model (without considering bubble phase) are very close to those using the two-phase bubble model. Above 1.5 times the minimum fluidisation velocity, the two-phase model shows that the bubble phase serves as a bypass and reduces further increases in drying rate. At times the minimum fluidisation velocity, the bubble phase reduces the drying rate by about 10% as compared against the cases where no bubble phase is present. This behaviour was expected from the earlier analysis of the fluidisation conditions using the cross-flow factor. As the air fluidising velocity increases beyond 1.5 times the minimum fluidisation velocity, the cross-flow factor is less than three. The influence of the bubble phase reduces further increase in drying rate due to vapour bypass in the bubble phase. As the air fluidising velocity increases, the bubble size and bubble velocity increases correspondingly. With a larger bubble size, the contact surface area per bubble volume (between the bubble and dense phase) is reduced and the moisture transfer between the two phases decreases. At constant bed height, an increase in bubble velocity also reduces the contact time between the bubble and dense phases. This, in turn, lowers the moisture transfer between the two phases. In general, the bed height has little influence on the drying performance except when temperature is at 85°C and the air fluidising velocity is above 1.5 times minimum fluidisation velocity shown in Fig. 6.2. At 85°C and air fluidisation velocity of 2·umf, 113 the drying rate decreases as the bed height increases from 0.1 m to 0.3 m. However, at bed height of 0.5 m, the drying rate increases again and is higher than that at 0.1 m. The change in drying rate with bed height is governed by two opposing factors, the contact time between the bubble and dense phase and the surface area available for mass transfer between the two phases. As the bed height increases, there is more contact time for moisture transport between the two phases. This enhances the drying process. However, the bubble size also depends on the bed height. An increase in bed height results in larger bubble size, which reduces the surface area between bubble and dense phase. As the bed height is increases from 0.1 m to 0.3 m, the latter factor dominates and reduces the drying rate. However, as the bed height increases beyond 0.3 m, the first factor dominates. These two effects are more significant at 85°C than at 65°C and 25°C because the difference in humidity values between entry and exit air streams is much larger at 85°C. The exit air stream consists of both exit air from the dense and bubble phase. Therefore, the effect of bubble bypassing the fluidised bed will be higher at 85°C than at lower temperatures. From Fig. 6.4, in both sections and of the bed, the drying rate increases when the temperature of the fluidising air stream increases from 65°C to 85°C. This is due to the larger driving force, i.e. the humidity values between entry air and that of adiabatic saturation. The rate increases more in section of the bed than in section 1. This is entirely because section is longer than section and therefore, 3.5 times more fluidising air enters section 2. 114 In drying of internal water, as changes in the hydrodynamic conditions not affect the drying rate, the earlier results in Table 6.5, except the last column showing moisture content removed, are applicable. The moisture content removed will depend on the residence time of particles in the bed. The increase in bed height will lead to an increase in residence time according to eqn (6.6) and the amount of moisture removed will increase correspondingly (eqn (6.5)). At velocity higher than minimum fluidisation velocity, the bed void fraction, ε, is no longer at 0.418 as determined in Section 3.4. The bed void fraction and the ratio of actual bubble rise velocity to interstitial air velocity, α, can be determined using eqns (6.7) and (6.8) (Davidson et al., 1985) for large particles.     u−u ε − ε mf mf  ⋅ = 1− ε  u br + (3 ⋅ α − 1)   u mf  u mf  α = u br ⋅    ε mf u mf (6.7) (6.8) Results showing the drying performance of internal water at 1.5 and times minimum fluidisation velocities are shown in Tables 6.14 and 6.15 respectively. 115 Table 6.14 Drying performance (internal water) in different bed sections at 1.5 times the minimum fluidisation velocity and bed height of 0.5 m Bed Section Section Section Section Temperature of entering air stream, Ti [°C] Average drying rates after initial drying, kg H O dX [ ] dt kg solids ⋅ s Moisture content removed, Xr [kg H2O per kg solids] 65 2.98·10-6 1.42·10-3 85 3.04·10-6 1.44·10-3 65 2.98·10-6 4.95·10-3 85 3.04·10-6 5.05·10-3 25 7.99·10-7 9.47·10-4 Table 6.15 Drying performance (internal water) in different bed sections at times the minimum fluidisation velocity and bed height of 0.5 m Bed Section Section Section Section Temperature of entering air stream, Ti [°C] Average drying rates after initial drying, kg H O dX [ ] dt kg solids ⋅ s Moisture content removed, Xr [kg H2O per kg solids] 65 2.98·10-6 1.17·10-3 85 3.04·10-6 1.20·10-3 65 2.98·10-6 4.11·10-3 85 3.04·10-6 4.19·10-3 25 7.99·10-7 7.86·10-4 In conclusion, after modifying the weir height to 0.5 m and by increasing in the fluidising air stream by 50%, the dryer is capable of removing a total either 0.240 kg H2O per kg solids of surface water (24.0 wt.%) or 0.00744 kg H2O per kg solids of 116 internal water (0.744 wt.%). Compared to the existing dryer capacity, this is equivalent to 188% of surface water and 172% of internal water removed. By replacing the existing gas blower to increase the fluidising air stream by 100%, the new dryer capacity will be a total of either 0.297 kg H2O per kg solids of surface water (29.7 wt.%) or 0.00618 kg H2O per kg solids of internal water (0.618 wt.%). Compared to the existing dryer capacity, this is equivalent to 232% of surface water and 143% of internal water removed. The amount of internal water removed is less than that at 150% fluidising air stream because the solids have a smaller residence time in the dryer due to the higher bed void fraction. With the above modifications, the capacity of the existing dryer can be increased by at least 50% above the original design basis. To further improve the performance of dryer, more capital and time consuming measures can be considered: i. Reduce the operating temperature of cold trap, which will lower the moisture content in the entry air. This increases the driving force of the drying process in terms of humidity gradient. ii. Install heat exchangers inside the fluidised beds, such as heating panels or heating tubes operated on hot oil or air (Chisholm and Howard, 1988; Mujumdar, 1995). The maximum inlet temperature of the heating medium should not exceed 85°C, so as not to cause discolouration of nylon-6.6. The disadvantage is the wearing out of heat exchanger surfaces due to collision by solid particles. 117 iii. Install horizontal baffles inside the fluidised beds to reduce the bubble sizes at higher air fluidising velocities. Smaller bubble sizes lead to an increase in the contact surface area between bubble and dense phases, higher residence time of bubbles inside the bed and therefore, an increase in moisture transfer across the two phases. iv. Install temperature and humidity transmitters to measure both the conditions in fluidised beds and the exit air streams immediately above the bed surfaces. These transmitters provide a check on the actual operating conditions against the predictive model developed. Further optimisation work can be carried out with such operating data. This is in-line with the concept of “smart” dryers. In this concept, online sensors are installed to measure solids properties such as moisture content, temperature, etc. as the wet solids travels through the dryer. Real-time computer control monitors the solids properties and adjust the local drying conditions according to a mathematical model, fuzzy logic or artificial neural network (Mori et al., 2001). 118 Conclusions Theoretical single-phase drying models have been adopted successfully to predict the drying of nylon-6.6 particles in a laboratory-scale fluidised bed. Using one of these models, the performance of an industrial-scale dryer operating at minimum fluidisation velocity was determined. In process optimisation, the dryer needs to be operated at bubbling fluidisation. Two-phase drying models are developed and applied to predict the influence of bubble phase on the drying process and the improved drying performance. Current fluidised bed drying models can be divided into two categories: single- and two/three-phase drying models. To select between the two types of models, theoretical criteria have been developed to predict when the accuracy of the singlephase model is no longer sufficient. Using the flow chart shown in Fig. 5.2, the operating fluidisation regime is first identified. Knowing the regime, the cross-flow factor is used to determine the influence of the bubble phase on the drying process and whether the single- or two/three-phase drying model should be adopted. This algorithm approach to select the suitable type of drying model has not been reported in previous literature. The improved two-phase drying model has two new features. Existing models assume free bubbling fluidisation takes place during fluidisation and the bubble size is uniform throughout the bed. Though free bubbling is generally valid for large- diameter beds, wall effects reduce the bubble velocities in small beds. A uniform bubble size means that the contact surface area between dense and bubble phase is unchanged along the bed height and the bubble velocity is constant. Literature 119 correlations show that such assumptions not hold and therefore, could lead to inaccuracies in predicting inter-phase mass transfer. The improved model accounts for both the influence of wall effects and the varying bubble size along the bed height. During bubbling fluidisation, determining the drying rate typically requires solving coupled differential heat and mass balance equations. In this work, a solution strategy is devised to avoid the need to solve the heat and mass balances simultaneously. By iteratively calculating the moisture profile of the dense and bubble phases, the moisture transport between particle surface, dense- and bubble phases can be determined. This approach has not been previously reported. A theoretical drying model has been developed which incorporates intensive twophase gas-solids flow patterns and yet retains sufficient straightforwardness for process design and optimisation. Arising from this work, practical recommendations are made to improve the performance of the existing industrial fluidised bed dryer. So far, in open literature, similar attempts to incorporate two-phase fluidisation into drying models have been largely limited to laboratory-scale equipment. To complete a robust model for future design engineers, more industrial case-studies and operational data would be helpful to further fine-tune the current model developed and extend the validity to different particle sizes and types. 120 References: Baeyens, J. and D. Geldart. An investigation into slugging fluidized beds. Chem. Eng. Sci., 29, Pp. 255-265. 1974 Bahu, R.E. Fluidised Bed Dryer Scale-up. Drying Technology, 12 (1&2), Pp. 329-339. 1994 Brereton, C.M. and J.R. Grace. Transition to turbulent fluidization. Chem. Eng. Res. Des., 70(A3), Pp. 246-251. 1992 Burgschweiger, J., H. Groenewold, C. Hirschmann and E. Tsotsas. From hygroscopic single particle to batch fluidized bed drying kinetics. The Canadian J. of Chem. Eng., 77, Pp. 333-341. 1999 Cengel, Y.A. and M.A. Boles. (eds). Thermodynamics, An Engineering Approach. 2nd Ed. Pp. 699, McGraw-Hill, Inc. 1994 Chandran, A.N., S.S. Rao and Y.B.G. Varma. Fluidized bed drying of solids. AIChE Journal, 36(1), Pp. 29-38. 1990 Chen, Z. W. Wu and P.K. Agarwal. Steam-drying of coal, Part 1. Modeling the behavior of a single particle. Fuel., 79, Pp. 961-973. 2000 Chen, Z., P.K. Agarwal and J.B. Agnew. Steam-drying of coal, Part 2. Modeling the operation of a fluidized bed drying unit. Fuel., 80, Pp. 209-223. 2001 Cheremisinoff, N.P. and P.N. Cheremisinoff. Hydrodynamics Of Gas-solids Fluidization. Pp. 151 US: Gulf Publishing Co. 1984 Chisholm, D. (Ed) and J.R. Howard, Heat Exchanger Technology. Pp. 96. Elsevier Applied Science. 1988 Ciesielczyk, W. Analogy of heat and mass transfer during constant rate period in fluidized bed drying. Drying Technology, 14(2), Pp. 217-229. 1996 121 Cranfield, R.R. and D. Geldart. Large particle fluidisation. Chem. Eng. Sci., 29, Pp. 935-947. 1974 Coumans, W.J. Models for drying kinetics based on drying curves of slabs. Chem. Eng. Process., 39, Pp. 53-68, 2000 Davidson J.F., R.B. Thorpe, O. Al-Mansoori, H. Kwong, M. Peck, R. Williamson. Evaporation of water from air-fluidized porous particles. Chem. Eng. Sci., 56, Pp. 6089-6097. 2001 Davidson, J.F., R. Clift and D. Harrison. (Eds). Fluidization. 2nd Ed. Pp. 529-561, 695-696, Academic Press. 1985 Dean, J.A. (Ed). Lange’s Handbook Of Chemistry. 14th Ed. Tab. 5.15, McGraw-Hill, Inc. 1992 Delvosalle, C and J. Vanderschuren. Gas-to-particle and particle-to-particle heat transfer in fluidized beds of large particles. Chem. Eng. Sci., 40(5), Pp. 769-779. 1985 Devahastin, S. (ed). Mujumdar’s Practical Guide To Industrial Drying. Pp. 12, 80, 9091, Thailand: Thananuch Business Ltd. Part. 2000 Edwards, M. and A. Avidan. Conversion model aids scale-up of Mobil’s fluid-bed mtg process, Chem. Eng. Sci., 41(4), Pp. 829-835. 1986 Geldart, D. (ed). Gas Fluidization Technology. Pp. 9, 33-54, 17, John Wiley & Sons Ltd. 1986. Geldart, D. Types of gas fluidization. Powder Technology, 7(5), Pp. 285-292. 1973 Gnielinski, V. Waerme- und Stoffuebertragung in Festbetten, Chem.-Ing.-Tech., 52, Pp. 228-236. 1980 Grace, J.R. and J.M. Matsen. History of fluidized solids development at Exxon. Proceedings of the 1980 international fluidization conference August 3-8, Pp. 3-23. 1980 122 Groenewold, H. and E. Tsotsas. A new model for fluid bed drying. Drying Technology, 15(6-8), Pp. 1687-1698. 1997 Groenewold, H. and E. Tsotsas. Predicting Apparent Sherwood numbers in fluidised bed. Drying Technology, 17, Pp. 1557-1570. 1999 Gupta, S.N., R.B. Chaube and S.N. Upadhyay. Fluid-particle heat transfer in fixed and fluidized beds. Chem. Eng. Sci., 29, Pp. 839-843. 1974 Hoebink, J.H. and K. Rietema. Drying granular solids in fluidized beds – I. Chem. Eng. Sci., 35, Pp. 2135-2140. 1980 Hoebink, J.H. and K. Rietema. Drying granular solids in fluidized beds – II. Chem. Eng. Sci., 35, Pp. 1980 Holman, J.P. Heat Transfer. 7th Ed. Tab. A-5, McGraw Hill, Inc. 1992 Hunt, D. Prediction of sorption and diffusion Of water vapor by nylon-6,6. Polymer, 21, Pp. 495-501. 1980. Jackson, R. The Dynamics Of Fluidized Particles. Pp. 153-199. UK: Cambridge University Press. 2000 Jischa, M. Konvektiver Impuls-, Waerme- und Stoffaustausch. Pp. 70, 150. Braunschweig, Wiesbaden: Vieweg. 1982 Keey, R.B. Drying Principles And Practice. Pergamon Press. 1972 Keey, R.B. and M. Suzuki. On the characteristic drying curve. Int. J. Heat Mass Transfer. 17. Pp. 1455-1464. 1974 Keey, R.B. Introduction To Industrial Drying Operations. Pp. 189-195, Pergamon Press. 1978 Kerkhof, P.J. Relations between local and effective air-side mass and heat transfer coefficients in fluidized bed drying. Drying Technology, 12(5), Pp.1191-1210. 1994 123 Kerkhof, P.J. Some modeling aspects of (batch) fluid-bed drying of life-science products. Chem Eng. Process., 39, Pp. 69-80. 2000 Kohan, M.I. (ed.). Nylon Plastics Handbook. Pp. 324, 457-459, 511. USA: Hanser/Gardner Publications. 1995 Kunii, D. and O. Levenspiel. Fluidization Engineering. 2nd Ed. Pp. 12, 270, 313-336, 415-426, 430-431, Toronto: Butterworth-Heinemann. 1991 Levenspiel, O. G/S reactor models – packed beds, bubbling fluidized beds, turbulent fluidized beds and circulating (fast) fluidized beds. Powder Technology, 122(1-9), 2002 Lide, D.R. (ed). Handbook Of Chemistry And Physics. 81st Ed. CRC Press. 2001 Lim, K.S., J.X. Zhu and J.R. Grace. Hydrodynamics of gas-solid fluidization. Int. J. Multiphase Flow, 21, Suppl, Pp. 141-193. 1995 McCabe, W.L., J.C. Smith and P. Harriot. Unit Operations Of Chemical Engineering. Ed., McGraw-Hill Book Co. 1985. McLaughlin, L.J. and M.J. Rhodes. Prediction of fluidized bed behavior in the presence of liquid bridges. Powder Tech., 114, Pp. 213-223, 2001 Mori, Y.H. and K. Ohnishi (eds), A.S. Mujumdar. Energy And Environmental Technological Challenges For The Future. Pp. 112-124. Tokyo: Springer Verlag. 2001 Mori, S. and C.Y. Wen. Estimation of bubble diameter in gaseous fluidized beds. AIChE J., 21, 1, Pp. 109-114, 1975 Mujumdar, A.S. (ed), S. Hovmand. Handbook Of Industrial Drying Vol. 1. Pp. 195248. 2nd Ed. New York: Marcel Dekker, 1995 Mujumdar, A.S. and S. Suvachittanont (eds). Developments In Drying Vol. Food Dehydration. Pp. 59-111, Bangkok, Thailand: Kasetsart University Press, 2000 124 Ranz, W.E. Friction and transfer coefficients for single particles and packed beds, Chem. Eng. Prog., 48, 247-256. 1952 Schaafsma, S.H., N.W. Kossen, M.T. Mos, L. Blauw and A.C. Hoffman. Effects and control of humidity and particle mixing in fluid-bed granulation. AIChE Journal, 45, 6, Pp. 1202-1210. 1999 Sherwood, T.K., R.L. Pigford and C.R. Wike. Mass Transfer. Pp. 245-247, McGrawHill. 1975 Sit, S.P. and J.R. Grace. Effect of bubble interaction on interphase mass transfer in gas fluidized beds. Chem. Eng. Sci., 36, Pp. 327-335. 1981 Snaar, J.E., P. Robyr and R. Bowtell. Strong gradients for spatially resolved diffusion measurements. Mag. Res. Imaging, 16(5-6), Pp. 587-591, 1998 Subramanian, D. and H. Martin, E.U. Schluender, Stoffuebertragung zwischen Gas und Feststoff in Wirbelschichten, Vt-Verfahrenstechnik, 11, Pp. 748-750. 1977 Theologos, K.N., Z.B. Maroulis and N.C. Markatos. Simulation of transport dynamics in fluidized-bed dryers. Drying Technology, 15(5), Pp. 1265-1291. 1997 Thoenes, D. J. and H. Kramers. Mass transfer from spheres in various regular packings to a flowing fluid. Chem. Eng. Sci., 8, Pp. 271-283. 1958 Tsotsas, E. From single particle to fluid bed drying kinetics. Drying Technology, 12(6), Pp. 1401-1426. 1994 Van Meel, D.A. Adiabatic convection batch drying with recirculation of air. Chem. Eng. Sci., 9, Pp. 36-44. 1958 Vanecek, V., M. Markvart and R. Drbohlav. Fluidized bed drying. Pp. 153-165, London: Leonard Hill. 1966 Wen, C.Y. and Y.H. Yu. Mechanics of fluidization. Chem. Eng. Progr. Symp. Ser. 62, Pp. 100-111. 1966 125 Zahed, A.H., J.X. Zhu and J.R. Grace. Modelling and simulation of batch and continuous fluidized bed dryers, Drying Technology, 13(1&2), Pp. 1-28. 1995 Zhao, H.W. and G. Chen. Heat and mass transfer in batch fluidized bed drying of porous particles. Chem. Eng. Sci., 55, Pp. 1857-1869. 2000 Zoglio, M.A., W.H. Streng and J.T. Carstensen. Diffusion model for fluidized bed drying. Pharma. Tech., 64(11), Pp. 1869-1873. 1975 126 [...]... to the difficulties in modelling complicated mass transfer between different phases (dense, bubble and solid phase), today’s process design of fluidised bed dryers still requires cold modelling, laboratory- and pilot-scale tests For similar reasons, to model an industrial fluidised bed dryer, it is useful to study the same gas-solids system using a laboratory-scale fluidised bed The laboratory apparatus... Nusselt and Sherwood Numbers from Ranz (1952), which were developed for fixed beds Kerkhof (2000) adopted the model of Subramanian et al (1977) to calculate the mass and heat exchange taking place only in the dense phase The bubble phase was assumed to act only as an inactive short-cut with no solids present and is described using Kunii and Levenspiel (K-L) model It was concluded that due to uncertainty of. .. Temperature and humidity probe Fig 3.1 Experimental set-up of laboratory-scale batch fluidised bed dryer Table 3.1 Characteristics of fluidised bed dryer and fluidising gas Fluidised bed Column Distributor Internal Diameter (ID) & Height Freeboard ID & Height Temperature of inlet air Relative humidity of inlet air (RH) Drying duration Cylindrical; made of Plexiglas Porous Plate (Sintered Stainless Steel)... (with manufacturer’s guaranteed accuracy of RH ± 2% and ± 0.5°C) Fig 3.2 Photograph of laboratory-scale batch fluidised bed dryer 22 3.3 Materials and methods The two types of solid particles under investigation are industrial nylon-6.6 polymer particles and industrial expanded polystyrene particles The nylon-6.6 particles are almost mono disperse, elliptical in cross-section and cylindrical in length... temperatures of the air stream at inlet and outlet of the fluidised bed and by weighing the particles before and after each experimental run, average drying rates are calculated from a mass balance of humidity values at inlet and outlet air streams This is checked by taking and weighing particle samples during experimental runs The stationary bed height of 0.1 m is chosen to be same as that in the industrial- scale... the industrial process by constructing a laboratoryscale fluidised bed The main objective is to obtain some insights on fluidised bed drying of the same gas-solids system as an input to detailed theoretical modelling 3.2 Experimental equipment The experimental set-up is shown in Fig 3.1 and Fig 3.2 The laboratory-scale fluidised bed consists of a cylindrical perspex column with internal diameter of. .. industrial- scale fluidised bed dryer The different fluidisation regimes can be observed by varying the fluidisation velocity Drying rate, dX , and moisture content, X, are calculated based on experimentally dt measured values of temperatures, relative humidity and volumetric flow rates of the entering and exit air streams across the fluidised bed (Keey, 1978; Devahastin, 2000) A mass balance across the bed over... was developed Mass transport coefficients between bubble and dense phases and gasparticle boundary were obtained from Sit and Grace (1981) and Ranz (1952) respectively Tsotsas and co-workers (Burgschweiger et al., 1999; Groenewold and Tsotsas, 1997; Tsotsas, 1994) adopted a similar approach and accounted for backmixing within the dense phase as well as wall -bed heat transfer Mass transfer coefficient... kg·m-3 (Dean, 1992) 1.8462·105 kg·m-1·s-1 (Holman, 1992) 9.81 m·s-2 (Dean, 1992) 21 The analytical equipment consists of the Brooks thermal mass flow control 5863i (calibrated for 0 to 700 Lmin-1 with manufacturer’s guaranteed accuracy of ± 1 Lmin-1), Flotech Setra differential pressure transducer model C230 and Testo T635 thermo hygrometer for measuring relative humidity and temperature (with manufacturer’s... models often can and do go way beyond measurements and analytical abilities In order for an engineer to make predictions for practical design, the challenge is to use good judgment in the choice of simple yet adequate models The most common form of gas-solids flow pattern in fluidised bed drying is bubbling fluidisation Since the 1990s, attempts have been made to develop a complete fluidised bed drying . designs of fluidised beds have been developed to suit many types of wet solids that can be fluidised by hot gas. Several designs of conventional fluidised bed dryers are shown in (Kunii and Levenspiel,. developed, and applied to the optimum operating performance of an existing industrial- scale fluidised bed dryer for nylon-6.6 particles. 1.4 Organisation of thesis This thesis is organised into. wet solid feed (Chandran et al., 1990). 3 The advantages that fluidised beds offer are a large transfer contact area between solids and gas phases, high heat and mass transfer between phases,