REVERSE FLOW CYCLONES FOR COLLECTION OF TEA DUST M.Sc (Chemical and Process Engineering) I.M.B.M. De Silva IB /3>Otij\03/o3 REVERSE FLOW CYCLONES FOR COLLECTION OF TEA DUST i'L 'f By I.M.B.M. De Silva This Thesis was submitted to the Department of Chemical and Process Engineering of the University of Moratuwa in partial fulfillment of the Degree of Master of Science in Chemical and Process Engineering Department of Chemical and Process Engineering University of Moratuwa Sri Lanka July, 2003 i s C o um Thesis 113,6-2 79262 7 9 2 6 2 DECLARATION "I certify that this thesis does not in corporate without acknowledgement any material previously submitted for a degree or diploma in any University to the best of my knowledge and belief it does not contain any material previously published, written or orally communicated by another person except where due reference is made in the text." Signature of the candidate (I.M.B.M. De Silva) To the best of my knowledge, the above particulars are correct Supervisor (Dr. B.M.W.P.K. Amarasinghe) iii ABSTRACT Reverse flow cyclones are used most extensively in the chemical process industries for gas - solid separation. Cyclones are often employed to collect large particles (>5u.m) that can be used not only as an air pollution control device, but also for recover particulate matter and size separation of particles. Common features found in locally designed cyclones are ineffective and crudely designed. Design of cyclone is more towards realizing a shape of the cyclone than the performance. Customized design approach gives a cyclone with greater collection efficiency, smaller in size or with lower pressure drop that would be found for a conventional standard design. Since the customized design procedure requires trial and error calculations, this research focused on the importance of the development of a computer package: "CycDesign". Using this package, a pilot scale reverse flow cyclone is designed and fabricated. This unit was used to examine the suitability of abating the air pollution caused due to dust generated from the fluidized bed dryers in tea industries. Trials were also done for sawdust, cement, quarry dust, talc powder and silica sand. Inlet and outlet particle size distributions were measured. Above 90% Overall collection efficiencies were attained for all the types of dust tested. For tea dust 99.2% collected experimentally which was predicted as 100% by the computer package. Also the computer package can be used to predict performance and dimensionless parameters for a cyclone design. It predicts that a continual decrease of Stokes number based on cut diameter, with increasing Reynolds number Re, for cyclones having different height to diameter ratio H/D. According to predictions, collection efficiency increases with H/D ratio of the cyclone. The declining patterns of fractional efficiency can be visualized with decreasing pressure drop across the cyclone and particle density. A decrease in fractional efficiency can be observed with the increasing of gas flow rate, gas temperature, and gas density. iv ACKNOWLEDGEMENT I would like to express my sincere gratitude to my supervisor Dr. (Mrs.) B. M. W. P. K. Amarasinghe, Head of the Department of Chemical and Process Engineering, University of Moratuwa for her great encouragement, guidance, dedication, patience paid through out the research. My heartiest gratitude is due to Dr. Ajith de Alwis, Mr. S. A. S. Perera for their expert comments, suggestions, recommendations and motivation for research work other than being the members of progress committee. My special thank goes to National Science Foundation for granting me financial assistance for the research project including stipend of Research Assistant. And of course I should thank Prof. (Mrs.) N. Rathnayaka, Director, Post-Graduate Studies, University of Moratuwa and her staff for great support offered through out the research project. I would like to thank Dr. Ziad Mohammad, Director, Tea Research Institute, for giving me the valuable opportunity to visit and study the process at St. Coombs Tea factory, Thalawakelle, and all the staff at the factory. I also would like to thank Mr. K. Jeyagoban and Mr. G. N. Wijesekara who were very helpful in developing the computer package "CycDesign". My great appreciation is also due to Mr. K. Udayarathne, Engineer, Alloy Fabrication Ltd., Ratmalana for the fabrication of main body of Reverse flow cyclone pilot unit, Mr. Kodithuwakku of S. D. Engineers, Panadura, manufacturer of air blower and Mr. C. U. Fonseka for the manufacturing of all other miscellaneous parts and involving the installation of the cyclone pilot unit. My special Thanks are due to Mr. N. L. Chandrasiri, Unit Operations Laboratory, and all other members of technical staff at Department of Chemical and Process Engineering, University of Moratuwa. I wish to thank Mr. S. Jayathilake and his staff at Industrial Technology Institute for the measurements of Laser beam particle size distribution analysis. Dr. O. K. Dissanayake, Head, Department of Earth Resource engineering, University of Moratuwa for providing necessary equipments for the measuring purposes. And also I am grateful Dr. T. Sugathapala, Department of Mechanical Engineering, University of Moratuwa for his valuable comments on the subject. Also I would like to mention the service rendered by the library staff, University of Moratuwa - Ms. R. Kodikara, Librarian, Ms. Wanigasekara and all the library staff members. I am thankful to the academic staff of Department of Chemical and Process Engineering, University of Moratuwa specially Mr. S. Wijesinghe giving enormous help to complete my literaturary work. As usual, the help at home base has been wonderfully generous and encouraging. Advice and great support given to me by my colleagues are excellent. Finally I offer them my heartiest gratitude. NOMENCLATURE AH Pressure drop expressed as number of inlet velocity heads AP Pressure drop f l Gas viscosity \f/ Cyclone inertia parameter rj Efficiency Captured fraction pa Gas density p v Particle density Stk 50 Stokes number based on cut diameter v Escaped fraction ^ / > , , c , , w Fractional efficiency ^ m , ™ / / Overall efficiency a Gas entry height Ac Interior collecting surface of the cyclone b Gas entry width B Dust outlet diameter C Cyclone geometry factor Ca Allowable outlet dust concentration Ci Inlet dust concentration C0 Outlet dust concentration D Cyclone cylinder diameter Dc Gas outlet diameter d Particle diameter d/no Critical particle diameter dsn Cut particle diameter Fc Centrifugal force acting on particle Fj Drag force acting on particle H Cyclone overall height h Cyclone cylinder height n Vortex exponent Number of turns gas makes within cyclone Vortex exponent at gas temperature at 7/ ( usually 283 K) Vortex exponent at gas temperature at 7^ K Pressure drop factor Volumetric gas flow rate Radial distance from cyclone axis Reynolds number Radial distance from cyclone axis to cyclone wall Gas outlet height Time Gas temperature corresponding to vortex exponent, «/ Gas temperature corresponding to vortex exponent, ri2 Radial component of particle velocity Tangential component of particle velocity Gas velocity Gas inlet velocity Radial component of gas velocity Tangential component of gas velocity Gas tangential velocity at cyclone outer wall Migration velocity of the particle Natural vortex length CONTENTS INTRODUCTION 1 LITERATURE SURVEY 3 2.1. Mechanism 4 2.1.1. Cyclone Types 4 2.1.2. Main Parts of a Reverse Flow Cyclone 6 2.2. Performance 8 2.2.1. Tangential Gas Velocity 8 2.2.2. Vertical And Radial Gas Velocity 10 2.2.3. Pressure Distribution 10 2.2.4. Overall Gas Flow Pattern 10 2.2.5. Pressure Drop 11 2.2.6. Collection Efficiency 13 2.2.7. Fractional Efficiency 15 2.3. Cyclone Design 19 2.3.1. Necessary Design Information 19 2.3.2. Standard Designs 19 2.3.3. Customized Design 22 2.3.4. Theory of Optimization 24 2.4. Novel Modifications 28 2.4.1. Cyclones for liquid droplets 31 2.5. Applications of Cyclone Separators 31 2.5.1. Main Features of Cyclone Separators 31 2.5.2. Application of Cyclones in Sri Lankan Industries 32 A CASE STUDY: DUST PROBLEM IN TEA MANUFACTURING 33 3.1. Introduction 33 3.2. Tea Manufacturing Process 33 3.3. Fluidized Bed Tea Dryers 35 4. SOFTWARE SOLUTION FOR CUSTOM DESIGN OF CYCLONES 40 5. EXPERIMENTAL PROCEDURE 43 5.1. Design and Fabrication 43 5.2. Testing Procedure 46 6. RESULTS AND DISCUSSION 48 6.1. Comparison of performance that predicted by the program and experimental results 48 6.1.1. Fractional efficiency 48 6.1.2. Overall efficiency 51 6.2. Predictions using computer program 53 6.2.1. Comparison of efficiencies with gas flow rate 53 6.2.2. Comparison of efficiencies with gas temperature 54 6.2.3. Comparison of efficiencies with particle density 56 6.2.4. Comparison of efficiencies with pressure drop 57 6.2.5. Comparison of efficiencies with the dimensions of the cyclone 58 6.2.6. Effect of H/D Ratio 59 6.2.7. Effect of Reynolds number - Dimensional analysis 60 6.2.8. Customized Design Vs Standard Design 62 6.2.9. Significance of the Geometry Parameter, C and Inertia Parameter, y/ on Fractional Efficiency 64 6.2.10. Graphical predictions for various dust applications 67 7. CONCLUSION 71 8. SUGGESTIONS 73 ix LIST OF FIGURES Figu re 2.1. Straight-through cyclone 4 Figure 2.2. Reverse flow cyclone with standard dimension notations 5 Figure 2.3. Flow through a reverse flow cyclone separator 6 Figure 2.4. Tangential gas velocity profile for a reverse flow cyclone 9 Figure 2.5. Vertical gas velocity profile for a reverse flow cyclone 9 Figure 2.6. Radial gas velocity profile for a reverse flow cyclone 9 Figure 2.7. Static and total pressure profile for a reverse flow cyclone 9 Figure 2.8. A typical fractional efficiency curve for a cyclone 13 Figure 2.9. Forces acting on a particle in a cyclone 14 Figure 2.10. Cyclone efficiency vs. wr Avf* from Bath (1956) 16 Figure 2.11. Cyclone design parameters vs. pressure drop factor P {H= 4D) 23 Figure 2.12. Cyclone design parameters vs. pressure drop factor P (H= 5D) 23 Figure 2.13. Cyclone design parameters vs. pressure drop factor P (H = 6D) 24 Figure 2.14. Common types of cyclone entry 28 Figure 2.15. Fines eductor 29 Figure 2.16. Common dust discharge systems 30 Figure 2.17. Inlet duct designs 30 Figure 3.1. Tea Making Process at St. Coombs Factory, Thalawakelle, Sri Lanka 34 Figure 3.2. Schematic drawing of the TRI-CCC Fluidized Bed tea dryer 36 Figure 4.1. Algorithm of the computer program 41 Figure 5.1. A photograph of Installed laboratory scale cyclone separator system ..45 Figure 6.1. Comparison of PSD of Feed & Dust Collected (for Cement) 49 Figure 6.2. Comparison of PSD of Feed & Dust Collected (for Quarry dust) 49 Figure 6.3. Comparison of fractional efficiencies obtained by experiments and predicted with the computer program - for Cement 50 Figure 6.4. Comparison of fractional efficiencies obtained by experiments and predicted with the computer program - for Quarry dust 50 Figure 6.5. Comparison of Overall Efficiency - Experimental and Predicted 52 Figure 6.6. Comparison of efficiencies with particle density 56 Figure 6.7. Comparison of efficiencies with pressure drop 57 Figure 6.8. Variation of Geometry Parameter, C with H/D Ratio 59 Figure 6.9. Relationship between Reynolds number and Stokes number 61 Figure 6.10. Fractional efficiency comparison for customized design and Lapple's design 63 Figure 6.11. Behaviour of Inertia parameter as with the Pressure drop 66 Figure 6.12. Predictions for dust having \i = 6.45 and a = 3.08 67 Figure 6.13. Predictions for dust having Li = 11.53 and a = 5.38 68 Figure 6.14. Predictions for dust having u = 13.25 and a = 5.77 68 Figure 6.15. Predictions for dust having [i = 8.83 and a = 3.44 69 Figure 6.16. Predictions for dust having (j. = 25.28 and a = 6.96 69 Figure 6.17. Predictions for dust having u. = 35.02 and a = 16.12 70 LIST OF TABLES Table 2.1. Equations for predicting pressure loss as number of inlet velocity heads, AH...12 Table 2.2. Resultant equations derived and assumptions made for solving Equation (10) . 18 Table 2.3. Comparison of standard designs for reverse flow cyclones 21 Table 3.1. Ambient air quality standards 38 Table 5.1. Operating conditions of the pilot scale cyclone separator - Inputs to "CycDesign" 43 Table 5.2. Design parameters of the pilot scale cyclone separator - output from "CycDesign" 44 Table 5.3. Tested dust samples and collection points 46 Table 6.1. Comparison of overall efficiency - experimental and predicted 52 Tabic 6.2. Comparison of fractional efficiency with the gas flow rate 54 Table 6.3. Comparison of fractional and overall efficiency with the gas temperature 55 Table 6.4. Relationship between gas temperature, gas density and gas viscosity 55 Table 6.5. Effect of dimensions and process design changes on pressure drop, efficiency, and cost 58 Table 6.6. Operating conditions for the designs in Figure 6.10 63 Table 6.7. Increase of fractional efficiency as with H/D ratio 64 Table 6.8. Increase of fractional efficiency as with pressure drop assumed 65 x i