PREDICTION OF BED PRESSURE DROP AND TOP PACKED BED HEIGHT FORMATION IN THREE PHASE SEMI-FLUIDIZED BED WITH REGULAR HOMOGENOUS BINARY MIXTURES D.K. SAMAL* Y.K. MOHANTY** G.K. ROY*** *Dept. of Chemical Engineering, Gandhi Institute of Engineering & Technology (GIET), Gunupur, Rayagada, Odisha, India **Dept. of Chemical Engineering, Gandhi Institute of Engineering & Technology (GIET), Gunupur, Rayagada, Odisha, India ***Dept. of Chemical Engineering, Gandhi Institute of Engineering & Technology (GIET), Gunupur, Rayagada, Odisha, India ABSTRACT Hydrodynamics study have been conducted in a 0.05m internal diameter Perspex column, with water and air (secondary) as fluidizing medium in gas-liquid-solid semifluidized bed with spherical two different sizes of glass beads. The study is relating to bed pressure drop and packed bed formation. Experimental parameters studied include superficial gas and liquid velocities, average particle size and static bed height and the bed expansion ratio. Empirical and semi-empirical models have been developed. The calculated values from predicted models have been compared with the experimental values and fairly good agreement has been obtained. The hydrodynamics study conducted and the correlations developed thereof can find potential applications in chemical processing. KEYWORDS: Gas-Liquid Solid Semi-Fluidization, Regular Homogeneous Particles, Binary Mixtures, Hydrodynamics, Dimensional And Statistical Analyses. 1. INTRODUCTION A semi-fluidized column is defined as the combination of fluidized and packed bed conditions coexisting in one column, which operates in series, was first initiated by Fan et al, in 1959. Hydrodynamic characteristics such as bed pressure drop and height of top packed bed are essential for designing and improving the quality of semi-fluidization. Regular binary homogenous particle mixtures are preferable due to easy formation of semi-fluidized bed at low bed pressure drop compared to that of single sized particles. The regular shaped particles are being used in the catalytic reaction engineering. Many of the catalytic reaction involved in the process industries deals with binary and ternary mixture of catalyst. Several investigators have been attempted to explain the behavior of semi-fluidized bed in gas-solid, liquid-solid or gas-liquid-solid systems with various types of particles and their mixtures. Most of the studies relate to single particle while a few studies on binary homogenous and 202
heterogeneous irregular mixtures have been reported [1-6]. Hydrodynamics of regular single particles of more than 2 mm diameter (glass beads) have been studied by Jena et al.[7] with initial static bed height of 0.171 m to 0.301 m in 2009. Samal et al. [8-9] have studied the liquid-solid semi-fluidized bed hydrodynamics of irregular homogenous ternary mixture and gas-liquid-solid semi-fluidized bed hydrodynamics of irregular homogenous binary mixture in 2013. Practically no work has been conducted on hydrodynamics of homogeneous binary mixtures of regular particles in a gas-liquid-solid semi-fluidized bed. Thus, the hydrodynamics of regular homogenous binary mixture particles have been studied in gas-liquid-solid semifluidized bed. The purposes of use of binary mixture of glass bead are to investigate the effect of particle interaction on the hydrodynamic characteristics. The objective of the present work is to study the effect of superficial liquid velocity (U sfl ), superficial gas velocity (U sfg ) average particle diameter (d pav ) (mixture composition, wt. %), initial static bed height (H s ) and the bed expansion ratio (R) on the bed pressure drop, top packed bed formation. Simultaneously, correlations have been developed for the prediction of the responses in a liquid-solid semifluidized bed using dimensional and statistical analyses. The experimental data have been compared with the values obtained from the developed correlations in this study. The unique feature of the current study is the investigations carried out in 0.05m internal diameter cylindrical columns with binary mixture of spherical glass beads of average particle size of 0.001854 m (-8+10 BSS), 0.001504 m (-10+12 BSS), as in industrial practice 20 μm to 0.025 m diameter spherical catalyst particles are used for fixed and fluidized bed reactors. Various applications in different fields have been reported in literature where semifluidization plays an important role. Dynamics of Fines Deposition [10], Simultaneous Removal of Cyanide and copper Ions [11], Formation of nitrogen oxides and deposits on the heating surfaces of boilers [12], Extractive fermentation of ethanol by immobilized yeast cells [13], Thin-layer drying of high-moisture faba beans in semi-fluidized bed [14] and the treatment of palm oil mill effluent [15] are some of them. 2. Materials and Methods Schematic and pictorial representation of the experimental setup is shown in Fig. 1. The details of the set-up and the method of investigations are described in an earlier paper [9]. 203
The scope of the experiment is presented in Table 1. The procedure was repeated varying the initial static bed height, average particle size, and bed expansion ratio for dimensionless responses like semi-fluidized bed pressure drop (ΔP sf /ΔP mf ) and height of top packed bed (H pa /H s ). The responses ΔP sf /ΔP mf and H pa /H s respectively have been taken to be functions of sets of independent variables H s /D c, d pav /D c, G sfl /G mfl, G sfg /G mfl, R. Table 1:- Scope of the experiment for regular binary mixture (A) Properties of bed material (B) Binary mixture properties Size Particle size Composition Avg. particle (d p x 10 3, m) Mixture (wt%) (d pav x 10 3 m) d p1 1.854 Mixture 1 50 : 50 1.6825 d p2 1.504 Mixture 2 60 : 40 1.7142 Mixture 3 70 : 30 1.7471 Mixture 4 80 : 20 1.7814 Mixture 5 90 : 10 1.8169 (C) Materials Materials Glass Beads ρ s, kg/m 3 2450 (D) Bed expansion ratio R 1.5 2.0 2.5 3.0 3.5 In this work mathematical and statistical model have been used for predicting For statistical analysis, Central Composite Design (CCD) [16-18] has been used to develop correlations for four responses (dependent dimensionless variables). The complete experimental range and the levels of independent variables are given in Table 2. 204
Table 2:- Level of independent variables for regular binary mixture Variables Symbol - α -1 0 +1 +α Aspect Ratio (H s /D c ) X 1 0.65 1.2 1.6 2.0 2.55 Particle diameter (d pav /D c ) X 2 0.0334 0.03431 0.0350 0.0356 0.0366 Liquid mass velocity (G sfl /G mf ) X 3 1.747 2.666 3.333 4.0 4.919 Gas mass velocity (G sfg /G mf ) X 4 0.828 2.666 4.0 5.333 7.171 Bed expansion ratio (R) X 5 1.31 2.0 2.5 3.0 3.69 The design of the experiment is given in Table 3 (dimensional analysis) and Table 4 (statistical analysis). For statistical analysis, a statistical software package Design-Expert- 8.0.7.1, Stat-Ease, Inc., Minneapolis, USA, has been used for regression analysis of the semifluidized bed responses. Table 3:- Experimental design matrix and responses (Dimensional Analysis) for regular binary mixture. H s /D c d pav / D c G sfl /G mfl G sfg /G mfl R Expt. ΔP sf /ΔP mfl Cal. ΔP sf /ΔP mfl (Eq. 1) Expt. H pa /H s Cal. H pa /H s (Eq. 2) 0.8 0.03494 3.333 4 2.5 10 12.51888 0.5 0.504508 1.2 0.03494 3.333 4 2.5 16 17.52187 0.433 0.444665 1.6 0.03494 3.333 4 2.5 22 22.24229 0.4 0.406563 2 0.03494 3.333 4 2.5 26 26.76316 0.35 0.379271 2.4 0.03494 3.333 4 2.5 32 31.1311 0.34 0.358338 1.6 0.03365 3.333 4 2.5 28 29.69274 0.46 0.488548 1.6 0.03428 3.333 4 2.5 26 25.73294 0.44 0.446049 1.6 0.03494 3.333 4 2.5 22 22.24229 0.42 0.406563 1.6 0.03562 3.333 4 2.5 16 19.16163 0.4 0.369793 1.6 0.03633 3.333 4 2.5 14 16.47101 0.35 0.335874 1.6 0.03494 2 4 2.5 4 4.064333 0.0375 0.056132 1.6 0.03494 2.666 4 2.5 10 10.5788 0.14 0.171066 1.6 0.03494 3.333 4 2.5 22 22.24229 0.38 0.406563 1.6 0.03494 4 4 2.5 38 40.81757 0.78 0.824657 1.6 0.03494 3.333 1.333 2.5 12 10.07378 0.17 0.162527 1.6 0.03494 3.333 2.666 2.5 16 16.60255 0.29 0.289805 1.6 0.03494 3.333 4 2.5 22 22.24229 0.37 0.406563 1.6 0.03494 3.333 5.333 2.5 28 27.36628 0.48 0.516837 1.6 0.03494 3.333 6.666 2.5 32 32.14081 0.6 0.622589 1.6 0.03494 3.333 4 2 32 34.21266 0.8 0.824033 1.6 0.03494 3.333 4 2.5 22 22.24229 0.42 0.406563 1.6 0.03494 3.333 4 3 14 15.64529 0.24 0.228265 1.6 0.03494 3.333 4 3.5 10 11.61974 0.1 0.140115 205
Table 4:- Experimental design matrix and responses (Statistical Analysis) for regular binary mixture. H s /D c d pav / D c G sfl /G mfl G sfg /G mfl R ΔP sf /ΔP mfl ΔP sf /ΔP mfl 1.20 0.0343 2.666 2.666 2.00 11.0 0.293 2.00 0.0343 2.666 2.666 2.00 16.9 0.249 1.20 0.0356 2.666 2.666 2.00 8.2 0.243 2.00 0.0356 2.666 2.666 2.00 12.6 0.207 1.20 0.0343 4.000 2.666 2.00 42.6 0.990 2.00 0.0343 4.000 2.666 2.00 65.2 0.988 1.20 0.0356 4.000 2.666 2.00 31.8 0.980 2.00 0.0356 4.000 2.666 2.00 48.5 0.989 1.20 0.0343 2.666 5.333 2.00 18.1 0.522 2.00 0.0343 2.666 5.333 2.00 27.6 0.443 1.20 0.0356 2.666 5.333 2.00 13.5 0.432 2.00 0.0356 2.666 5.333 2.00 20.6 0.367 1.20 0.0343 4.000 5.333 2.00 69.8 0.990 2.00 0.0343 4.000 5.333 2.00 106.6 0.980 1.20 0.0356 4.000 5.333 2.00 51.9 0.980 2.00 0.0356 4.000 5.333 2.00 79.4 0.970 1.20 0.0343 2.666 2.666 3.00 5.0 0.081 2.00 0.0343 2.666 2.666 3.00 7.7 0.069 1.20 0.0356 2.666 2.666 3.00 3.8 0.067 2.00 0.0356 2.666 2.666 3.00 5.7 0.057 1.20 0.0343 4.000 2.666 3.00 19.5 0.388 2.00 0.0343 4.000 2.666 3.00 29.8 0.330 1.20 0.0356 4.000 2.666 3.00 14.5 0.322 2.00 0.0356 4.000 2.666 3.00 22.2 0.273 1.20 0.0343 2.666 5.333 3.00 8.3 0.144 2.00 0.0343 2.666 5.333 3.00 12.6 0.123 1.20 0.0356 2.666 5.333 3.00 6.2 0.120 2.00 0.0356 2.666 5.333 3.00 9.4 0.102 1.20 0.0343 4.000 5.333 3.00 31.9 0.691 2.00 0.0343 4.000 5.333 3.00 48.7 0.587 1.20 0.0356 4.000 5.333 3.00 23.8 0.573 2.00 0.0356 4.000 5.333 3.00 36.3 0.486 0.65 0.0350 3.333 4.000 2.50 10.4 0.531 2.55 0.0350 3.333 4.000 2.50 32.5 0.343 1.60 0.0334 3.333 4.000 2.50 31.6 0.500 1.60 0.0366 3.333 4.000 2.50 15.6 0.320 1.60 0.0350 1.747 4.000 2.50 2.6 0.033 1.60 0.0350 4.919 4.000 2.50 80.6 0.990 1.60 0.0350 3.333 0.828 2.50 7.2 0.108 1.60 0.0350 3.333 7.171 2.50 33.4 0.646 206
1.60 0.0350 3.333 4.000 1.31 76.6 0.990 1.60 0.0350 3.333 4.000 3.69 10.4 0.116 1.60 0.0350 3.333 4.000 2.50 22.0 0.398 1.60 0.0350 3.333 4.000 2.50 22.1 0.399 1.60 0.0350 3.333 4.000 2.50 22.0 0.400 1.60 0.0350 3.333 4.000 2.50 21.9 0.390 1.60 0.0350 3.333 4.000 2.50 21.9 0.399 1.60 0.0350 3.333 4.000 2.50 22.1 0.400 1.60 0.0350 3.333 4.000 2.50 22.1 0.388 1.60 0.0350 3.333 4.000 2.50 22.1 0.390 3. Results and Discussions Hydrodynamic parameters for the gas-liquid-solid semi-fluidization using homogenous regular binary viz. the semi-fluidized bed pressure drop, height of top packed bed have been found to be dependent on material property (average particle size) and system parameters (initial static bed height and bed expansion ratio). 3.1 Development of correlations by Dimensional Analysis: For dimensionless bed pressure drop, the equation is ΔP sf /ΔP mfl = 4 10-12 (H s /D c ) 0.8292 (d pav /D c ) -7.6681 (G sfl /G mfl ) 3.3281 (G sfg /G mfl ) 0.7208 R -1.9297 (1) For dimensionless top packed bed height, the equation is H pa /H s =2 10-9 (H s /D c ) -0.3114 (d pav /D c ) -4.8757 (G sfl /G mfl ) 3.8769 (G sfg /G mfl ) 0.8344 R -3.166 (2) Fig. 2 shows the comparison between the experimental and calculated values of ΔP sf /ΔP mf with a standard deviation of 1.31 and a coefficient of correlation of 0.9895 respectively. Fig. 3 shows the comparison between the experimental and calculated values of H pa /H s with a standard deviation of 0.021 and a coefficient of correlation of 0.9941 respectively. 207
3.2 Development of correlations by statistical analysis approach For dimensionless semi-fluidization pressure drop, the equation is ΔP sf /ΔP mf =21.94+5.60 X 1-3.95 X 2 +16.64 X 3 +6.50 X 4-11.46 X 5-0.87 X 1 X 2 +3.49 X 1 X 3 +1.43 X 1 X 4-2.21 X 1 X 5-2.45 X 2 X 3-1.00 X 2 X 4 +1.55 X 2 X 5 +4.04 X 3 X 4-6.23 X 3 X 5-2.55 X 4 X 5-0.19 X 2 2 1 +0.19 X 2 +3.37 X 2 3-0.39 X 2 2 4 +3.71 X 5 (3) For dimensionless top packed bed height, the equation is H pa /H s =0.40-0.024 X 1-0.026 X 2 +0.24 X 3 +0.075 X 4-0.19 X 5 +2.047 10-3 X 1 X 2-7.465 10-4 X 1 X 3-5.954 10-3 X 1 X 4-3.807 10-3 X 1 X 5-1.296 10-3 X 2 X 3-6.207 10-3 X 2 X 4-3.928 10-3 X 2 X 5 +3.094 10-4 X 3 X 4-0.069 X 3 X 5 +0.015 X 4 X 5 +9.603 10-3 2 X 1 +4.830 10-3 X 2 2 +0.023 X 2 3-1.021 10-3 X 2 2 4 +0.030 X 5 (4) Predicted vs. Actual Predicted vs. Actual 120.00 1.20 100.00 1.00 80.00 0.80 0.60 60.00 0.40 40.00 0.20 20.00 0.00 0.00-0.20 0.00 20.00 40.00 60.00 80.00 100.00 120.00-0.20 0.00 0.20 0.40 0.60 0.80 1.00 1.20 Fig. 4: Comparison of Experimental with Calculated (Eq. 3) values Fig. 5: Comparison of Experimental with Calculated (Eq. 4) values Fig. 4 shows the comparison between the experimental and calculated values of ΔP sf /ΔP mfl with a standard deviation of 3.49 and a coefficient of correlation of 0.9865 respectively. Fig. 5 shows the comparison between the experimental and calculated values of H pa /H s with a standard deviation of 0.071 and a coefficient of correlation of 0.9686 respectively. 4. CONCLUSION Investigations have been carried out to study the behavior of regular homogenous binary mixtures with the superficial liquid and gas velocities in gas-liquid-solid semi-fluidized beds with particles of different size in 0.05 m column diameter. It may be concluded that the bed 208
pressure drop and height of top packed bed are strong functions of superficial liquid and gas velocities and the average size of particle. The initial static bed height is also important for the purpose. Correlations for the calculation of bed pressure drop and height of top packed bed have been proposed. The values calculated from the developed correlations have been compared with the experimental ones. The coefficients of correlations are found to be greater than 0.96 in both the approaches, thus emphasizing the validity of the developed correlations over the range of the operating parameters investigated. The present hydrodynamics study along with the developed correlations can find potential applications in physical and chemical processing. In the present investigation, the wide range in experimental parameters involved especially with respect to mixture particle and bed expansion ratio, makes it amenable to logical scale-up while considering design of gas-liquidsolid semi-fluidization systems. Acknowledgement The authors gratefully acknowledge the support and encouragement received from GIET, Gunupur, Odisha, India-765022 Nomenclature BSS British Standard Sieve D, d Diameter, m G Mass velocity, Kg.m -2.s -1 H, h Height, m P Pressure, N.m -2 R Bed expansion ratio Greek letters Δ Difference Subscripts av average c column f fluid g gas l liquid mf minimum fluidization p particle pa top packed bed s static sf semi-fluidization References 1. Dash. J, Roy. G. K, 1977. Liquid-Solid Semi-fluidization of Homogeneous Mixture-I, Prediction of Onset Semi-fluidization Velocity'. Indian Chemical Journal, February, p 1. 209
2. Dash. J, Roy. G. K, 1976. Liquid-Solid Semi-fluidization of Homogeneous Mixtures II, Prediction of Maximum Semi-fluidization Velocity. Indian Chemical Journal, January, 3. Kurian. J, RajaRao. M, 1970. Hydrodynamics of semi-fluidized bed, Indian Journal of Technology, 8: 275-284. 4. Murthy. J, S, N, Roy. G, K, 1986. Semi-fluidization: a Review, Indian Chemical Engineer, Vol. XXIX, No.2: 9-22. 5. MYDLARZ JERZY, 1987. Prediction of the packed bed height in liquid-solid semifluidization of homogeneous mixtures. The Chemical Engineering Journal, 34: 155 158 6. Roy. G. K, Sarma K. J. R, 1974. Relation between maximum semi-fluidization and minimum fluidization velocity in liquid-solid systems JI. of the Inst,of Engrs., (India), 54: 34 7. Jena, H. M, Roy, G. K, Meikap, B. C, 2009. Hydrodynamics of regular particles in a liquidsolid semi-fluidized bed. Powder Technology, 196: 246-256 8. Samal, D. K, Mohanty, Y. K, Roy, G. K, 2013, Hydrodynamics of liquid-solid semi-fluidized bed with irregular homogenous ternary mixture. Powder Technology, 235: 921-930 9. Samal, D. K, Mohanty, Y. K, Roy, G. K, 2013, Prediction of bed pressure drop and top packed bed formation in gas-liquid-solid semi-fluidized bed with irregular homogenous binary mixtures. Korean J. of Chemical Engineering, 30(6): 1326-1334 10. Dehkissia, S. Baçaoui, A, Iliuta, I, Larachi, F, 2008, Dynamics of Fines Deposition in an Alternating Semi-fluidized Bed, A.I.Ch.E. Journal, 54: 2120-2131. 11. Kim, S. J, Hwang, K. R, Cho, S. Y, Moon, H, 1999, Simultaneous Removal of Cyanide and copper Ions in a Semi-Fluidized Ion Exchanger Bed. Korean Journal of Chemical Engineering, 16: 664-669. 12. Is'emin, R. L, Zaitseva, N. A, Osipov, A. D, Akol'zin, A. P, 1995, Formation of nitrogen oxides and deposits on the heating surfaces of boilers with semi-fluidized bed furnaces fired with low grade coal, Promyshlennaya Energetika, 2: 37-38. 13. Dias, S. M. M, 1991. Extractive fermentation of ethanol by immobilized yeast cells. NTIS. Report, Dept. Engg., Tech. Univ. Lisbon, Lisbon, 218. 14. Chayjan, R. A, Shadidi, B, Modeling high-moisture faba bean drying in fixed and semifluidized bed conditions, Journal of Food Processing and Preservation, DOI: 10.1111/j.1745-4549.2012.00766.x 15. Alade, A. O, Jameel, A. T, Muyibi, S. A, Karim, M. I. A, Alam, Z, Application of semifluidized bed bioreactor as novel bioreactor system for the treatment of palm oil mill effluent (POME), African Journal of Biotechnology, DOI: http://dx.doi.org/10.4314%2fajb.v10i81. 16. Box, G. E. P, Hunter, J. S, 1957, Multi-factor experimental design for exploring response surfaces, Ann Math Stat., 28: 195-241 17. Box, G. E. P, Hunter, W. G, 1961, The 2k-p fractional designs, parts I and II, Journal of Technometr, 3: 311-458 18. Napier-Munn, T. J, 2000. The Central Composite Rotatable Design JKMRC, The University of Queensland, Brisbane, Australia. 210