VIEWS: 82 PAGES: 5 CATEGORY: Technology POSTED ON: 3/3/2010 Public Domain
Two Phase Gas-solid Pressure Drop in the Riser of a Circulating Fluidised Bed D Acharya, Non-member Data on the static pressure distribution for the gas-solid system in the riser of a circulating fluidized bed have been collected to obtain the experimental pressure drop. Theoretical pressure drop equations based on mechanical energy balance as well as expanded packed bed analogy have been derived. Since, the solids friction factor data from literature gave large deviations between experimental and theoretical pressure drop, the solids friction factor values were determined experimentally. Axial voidage distribution correlations were developed for the dense and dilute phases which provided the voidage values required by the theoretical pressure drop equations. The predictions from the pressure drop models agreed well with the experimental data. Keywords: CFB; Riser; Dense and dilute phase; Voidage; Static pressure NOTATION ∆P : pressure drop, N/m2 Dfb , DP : riser and particle diameters, respectively, m ε : gas void fraction fg , fs : gas and solid skin friction factors, respectively µg : gas viscosity, N/m2 Ldl , Ldn : dilute and dense phase heights, respectively, m ρg , ρs : gas and solids densities, respectively, g/cm3 LT : riser height, m INTRODUCTION M g , Ms : gas and solid mass flow rates, respectively, In recent years, it has been observed that bubbling fluidized kg/s bed boilers used in thermal power plants have shown Q g , Qs : volumetric flow rate of gas and solid decreased efficiency in steam generation. This is because of partial combustion of coal particles in the bubbling fluidized respectively, m3/s bed. To remove this inherent deficiencies and also to obtain Ref : gas phase Reynolds number, DfbUg ρ f / µ f enhanced rate of heat transfer from the combustion of coal to water tubes of the boiler, the concept of circulating fluidized Res : solid phase Reynolds number dpUs ρ s / µ f , d i dP V g − U P ρ f / µ f bed (CFB) has come into existence. The CFB consists of two fluidized beds, namely, a fast Ret : Reynolds number for solids-settling fluidized bed (FFB) and a slow fluidized bed (SFB). In the SFB d pU t ρ f / µ f the solids are fluidized using the minimum fluidization Vg , Vs : gas and solids velocities, respectively in riser, velocity from where the solids enter the FFB also known as m/s riser via a solids recirculation line. The solids from the riser are recirculated into the SFB through a cyclone separator and Vdl , Vdn : dilute and dense phase solids velocities, a solids return leg. Therefore, FFB (or riser) is the main respectively, m/s equipment where high rate of combustion of coal is achieved because of recirculation of solids. Ug : superficial velocity of gas, m/s The data on two phase pressure drop in the riser are required UP : particle velocity in riser, m/s to determine the horse power of the blower for its Ut : solids settling velocity, m/s specification. W : solids mass flux, kg/m2s LITERATURE SURVEY Z : riser height, m In order to develop pressure drop models for gas-solid flow in the riser, the flow pattern for the two phase flow must be D Acharya was with the Department of Mechanical Engineering, MGM known. It has been reported1-2 that axial or longitudinal College of Engineering, Nanded 431 602, Maharashtra. voidage distribution in the riser showed an S shaped curve, This paper was received on October 23, 2003. Written discussion on the paper indicating a point of inflection between the dense phase (above will be entertained till March 31, 2005. distributor) and a dilute phase above the dense phase. This Vol 85, January 2005 219 complex flow pattern has made the theoretical prediction of K the pressure drop very difficult. There are few attempts to propose empirical equations3 for static pressure distribution or Legend semi-theoretical equations4-5 for prediction of pressure drop in A Blower P7 G the riser. Most of the reported works used riser (FFB) having V7 C Orifice Meter diameter more than 5.0 cm and height more than 2.0 m using D Distributor F excessive fine particles with high solids circulations rates. E Solids Transfer line V6 P6 Since, the voidage distribution depends upon, gas flow rates in F Fast Bed Column V5 riser, riser height, riser diameter, solids recirculation rates, G Cyclone Separator H Solids Return Leg P5 particle size etc, the correlations on voidage distribution I Slow Bed Column V4 H reported for large CFB may not be applicable for shallow CFB K Air Exit P4 V3 (diameter less than 5 cm and height around 2 m). There is P0 to P7 Pressure Taps scope for use of shallow CFB for thermal captive power V0 to V7 Quick Opening P3 Valves V2 plants, bench scale studies for heat transfer and combustion of W Air Control Valves P2 coal. V1 I P1 In the present work, data on static pressure distribution ob- V0 tained in the riser of a shallow CFB system for low solids E D circulation rate have been analysed and the experimental P0 pressure drop data were compared with those predicted by the C theoretical pressure drop models derived. W EXPERIMENTAL DETAILS C W Set-up A schematic diagram of the experimental set-up is shown in Figure 1. The riser or FFB (F) is made of perspex, 4.6 cm in A diameter and 2.0 m height (measured from the distributor to gas exit). The riser is provided with butterfly valves V0 to V7 and pressure tappings P0 to P7 . The slow bed column (I ) is connected to a cyclone separator (G) through a solids return Figure 1 Schematic diagram of the experimental set-up for voidage and leg (H ). The solids from the slow bed (I ) are re-circulated in static pressure measurement in circulating fluidized bed (CFB) the riser through solids transfer line (E ). The gas flowrates were measured by orificemeters. Solids flowrates in the solids Table 1 Experimental details for study on axial voidage and static transfer line (E ) were measured by segmental orifice meter. pressure distribution in CFB Experimental Procedure Operating Conditions Material Characteristics Material Density, Size, Initially static beds of solids were maintained in both the riser g/cm2 cm and slow bed column. Then the solids in the riser were fluidized using high air velocity and the flow rate of air in the Range of superficial slow bed was kept just above minimum fluidisation velocity. air velocity in riser, The fluidized solids from the riser were recirculated into the 290 cm/s975 cm/s Non-coking coal 1.723 0.0780 same bed through cyclone separator, solids return leg, slow Range of superficial bed column and solids transfer line. air velocity in slow bed, Data on voidage distribution were obtained by simultaneously 8.7 cm/s49 cm/s Bauxite ore 2.650 0.0754 closing the valves V0 to V7 and collecting the solids deposited Mass velocity of recirculated on these. Static pressure distribution data were collected from solids, 10.81 kg/(m2s) manometer readings of tappings P0 to P7 . to 33.25 kg/(m2s) Limestone 2.870 0.0427 Visual observations confirmed that the FFB consisted of a dense phase, a dilute phase in the annulus and a thin film of RESULTS AND DISCUSSION downflowing solids at wall region surrounding the dilute Effect of Riser Height and Air Velocity on Voidage phase. Distribution, ε The solids used were limestone, bauxite ore and non-coking Figures 2 and 3 show the relation between ε and riser height, z, coal. The fluidizing medium was air. Table 1 presents the for the system bauxite-air, and non-coking coal-air. The curves experimental details for the studies on axial voidage and static show a change in slope around a riser height of 120.0 cm. The pressure distribution in CFB. curves are S shaped as reported for large CFB. The change in 220 IE (I) JournalMC Correlation for the Axial Voidage 848 cm/s 930 cm/s, gas vel An attempt was made to calculate ε from the correlations given in literature and it was found that the calculated values 986 cm/s of ε are much higher than those obtained in this work. The Bauxite-air, Particle dia = 0.0754 cm present CFB being shallow, the flowrates of gas and 0.95 Solid rate = 12.02 g/s re-circulated solids were much less. Therefore, using the experimental data on voidage, the correlations proposed are: For the dense phase 0.90 e ε dn = 0.312 ( z )0.119 Re f / Re s j 0.076 (1) 0.85 For the dilute phase Axial voidage distribution 0.80 e ε dl = 0.437 ( z )0.068 Re f / Re s j 0.061 (2) Pressure Drop Models For predicting two phase gas-solid pressure drop in the riser of 0.75 CFB two approaches have been made. Pressure Drop Equation based on Mechanical Energy Balance 0.70 ∆PT = ∆P1 + ∆P2 + ∆P3 + ∆P4 (3) where ∆P1 is potential energy; ∆P2 , skin frictional energy; 0.65 ∆P3 , kinetic energy; and ∆P4 , the momentum of gas-solid 20 40 60 80 100 120 140 160 180 200 mixture. Riser height, cm Figure 2 Axial voidage distribution in riser of CFB for bauxite-air system d i ∆P1 = M g + M s g LT / Q g + Q s d i (4) where M g = V g ε Aρ g ; Ms = Vs ( 1 − ε ) A ρ s ; Q g = Vg ε A ; and 0.95 388 cm/s Qs = Vs ( 1 − ε ) A . 469 cm/s, Gas vel 545 cm/s ∆P2 = ∆P fg + ∆P fs (5) 0.90 non-coking coal-air, Particle dia = 0.078 Solid rate = 12.02 g/s e 2 ∆P fg = 2 f g LT U g ρ g / D fb j (6) 0.85 ∆P fs = ∆P fs ( dense phase ) + ∆P fs ( dilute phase ) Axial voidage distribution 0.80 e 2 j e 2 = 2 f s Ldn V dn ρ s / D fb + 2 f s Ldl V dl ρ s / D fb j (7) ∆P3 = V g2 ρ g / 2 (8) 0.75 ∆P4 = W V sl / 2 (9) 0.70 b g where W = V s 1 − ε av ρ s ; and V sl = mean slip veocity = Vg − Ut . 0.65 Therefore, 20 40 60 80 100 120 140 160 180 200 o b ∆P1 = U g ρ g + 1 − ε av V s ρ s g LT g t / U g + b1 − ε av g V s Riser height, cm (10) Figure 3 Axial voidage distribution in riser of CFB for non-coking coal- where U g = V g ε . air system In equation (6), fg gas friction factor is given by Blasius the slope of the curve is marked by a point of inflection indi- equation, cating a dense phase above grid and a dilute phase above the − 0 .25 dense phase. f g = 0.316 Re f (11) Vol 85, January 2005 221 Equation (11) is valid for Ref < 100 000 Equation (3) 50 Equation (17) For evaluation of ε av : Equation (3) using Exp. Solids friction factor, non-coking coal-air system z z 120 190 ε av = ( 1 / 90 ) ε dn dz + ( 1 / 70 ) ε dl dz (12) 45 30 120 Theoretical pressure drop, g/cm2 where dense phase height starts from 30 cm above grid and extends upto 120 cm and dilute phase height is from 120 cm 40 upto 190 cm (entry level to cyclone separator). The actual solids velocities in the dense and dilute regions (Vdn and Vdl ) can be obtained modifying Yangs equations as given 35 by equations (13) and (14), respectively. For dense phase: 30 eV gn −V dn j / U t e j 2 2 = 1+ f sn V dn / 2 g D fb ε 4.7 dn (13) For dilute phase: 25 eV gl −V dl j / U t e j ε 4dl.7 2 2 = 1 + f sl V dl / 2 g D fb (14) 38 40 42 44 46 48 50 52 54 Experimental pressure drop, g/cm2 For computing the solids friction factor, Yangs equations can Figure 4 Comparison between experimental and theoretical pressure be used. drop in riser b g b gb f s ε 3 / 1 − ε = 0.0126 1 − ε R et / R es g − 0.979 Equation (17) considers the total pressure drop at incipient fluidisation and those due to skin friction of the gas and solids. for Vg / Ut > 1.5 (15) Figure 4 shows the comparison between ∆PT predicted by equation (17) and the experimental values. The deviation is and more than 40%. The solids friction factor, fs required in equa- tion (17) was obtained from Yangs equations. b g b gb f s ε 3 / 1 − ε = 0.0410 1 − ε R et / R es g − 1.021 From analysis of theoretical pressure drop equations it was found that the values of fs applicable for shallow CFB should for Vg / Ut < 1.5 (16) be determined to minimize the percent deviation. Therefore, where fs = fdn or fdl . using the experimental pressure drop data of this work in equation (17), the values of solids friction factor, fs were The voidage ε dn or ε dl were calculated from the equations (1) calculated which were in the range of 3.15 × 10 − 4 to and (2), respectively. 4.82 × 10 − 4 . Using these values of fs , the ∆PT theoretical was The theoretical pressure drop was calculated from equation (3) again calculated from equation (3) and now the deviation and compared with the experimental data as shown in Figure 4. between ∆PT (theoretical) and ∆PT (experimental) was less than 10% as The deviation is above 30%. This is because Yangs equations (15) shown in Figure 4. and (16) for solids friction factor, fs , are valid for large CFB having riser diameter more than 5.0 cm and height more than CONCLUSION 2.0 m. Experimental data on axial voidage and static pressure Pressure Drop Equation for an Expanded Packed Bed distribution in a shallow CFB were obtained for low solids The riser of the CFB may be considered as an expanded circulation rates. To predict voidage distribution, correlations packed bed during the two phase flow. The total pressure have been proposed for dense as well as dilute phase. Using the drop, ∆PT , can be written as solids friction factor data from literature-correlations, the theoretical pressure drop models developed in this work gave e je j 2 e ∆PT = 1 − ε mg ρ s − ρ g Lm + 2 f g U g ρ g LT / D fb j large percentage deviations for ∆PT (theoretical) . Therefore, using the experimental pressure drop data, the values of fs were + e2 f V s 2 dn ρ s Ldn j/ D fb (17) calculated for shallow CFB and these values of fs when used in 222 IE (I) JournalMC equation (3) predicted with a percent deviation below 3. Y Li and M Kwauk. The Dynamics of Fast Fluidisation in Fluidisation. (ed) 10%. J R Grace and P Matsen, Plenum Press, New York, 1980, p 537. ACKNOWLEDGEMENT 4. C Brereton, J R Grace and J Yu. Axial Gas Mixing in a CFB. in Circulating Fluidised Bed Technology II. (ed) P Basu and J F Large, Pergamon Press, The author wishes to thank Shri S S Choudhary of the Oxford, 1988, p 307. Department of Mechanical Engineering, MGM College of Engineering, Nanded for his guidance during this work. 5. W C Yang. A Model for the Dynamics of a Circulating Fluidised Bed Loop REFERENCES in CFB Technology II, (ed) P Basu and J F Large, Pergamon Press, Oxford, 1988, p 181. 1. C Brereton and L Stromberg. Some Aspects of the Fluid Dynamic Behaviour of Fast Fluidised Beds. in Circulating Fluidised Bed Technology. 6. K E Wirth. Steady State Diagram for CFB. in Circulating Fluidised Bed (ed) P Basu, Pergamon Press, Toronto, 1986, p 133. Technology III. (ed) P Basu, M Horio and M Hasatani, Pergamon Press, 2. K Smolders and J Baeyens. Hydrodynamic Modelling of Axial Voidage Oxford, 1991, p 99. Profile in the Riser of a Circulating Fluidised Bed. Can Journal of Chemical Engineering, vol 79, June 2001. 7. W C Yang. Powder and Bulk Solids Technology. 1977, p 89. Submission of Manuscripts for IEI Journals Authors desirous to publish technical papers in the Journal of the Institution under various engineering disciplines are requested to send the manuscript in quadruplicate accompanied by one soft copy in CD or floppy disk (text in MS-Word and figures in JPG or TIFF format) along with original illustrations/photographs. It may be noted that the evaluation process will not be initiated unless these requirements are fulfilled. Vol 85, January 2005 223