Proceedings of the ASME 2011 International Mechanical Engineering Congress & Exposition IMECE2011 November 11-17, 2011, Denver, Colorado, USA IMECE2011-64485 Design and Simulation of a Hybrid Entrained-Flow and Fluidized Bed Mild Gasifier Part 2 Case Study and Analysis A.K.M. Monayem Mazumder, Ting Wang, and Jobaidur R. Khan Energy Conversion & Conservation Center University of New Orleans New Orleans, LA 70148-2220, USA ABSTRACT To help design a mild-gasifier, a reactive multiphase flow computational model has been developed in Part 1 using Eulerian-Eulerian method to investigate the thermal-flow and gasification process inside a conceptual, hybrid entrained-flow and fluidized-bed mild-gasifier. In Part 2, the results of the verifications and the progressive development from simple conditions without particles and reactions to complicated conditions with full reactive multiphase flow are presented. Development of the model starts from simulating single-phase turbulent flow and heat transfer in order to understand the thermal-flow behavior, followed by introducing seven global, homogeneous gasification reactions progressively added one equation at a time. Finally, the particles are introduced, and heterogeneous reactions are added in a granular flow field. The mass-weighted, adiabatic flame temperature is validated through theoretical calculation and the minimum fluidization velocity is found to be close to Ergun s correlation. Furthermore, the predicted exit species composition is consistent with the equilibrium values. INTRODUCTION The simulated mild gasifier is unique in its hybrid characteristics with both entrained-flow and fluidized bed features, so no similar CFD study has been performed before using this type of mild gasfier. The closest simulation will be those numerical analyses performed for a fluidized bed. For example, Yu et al. [1] developed a numerical model based on the two-fluid model (TFM) including the kinetic theory of granular flow (KTGF) and complicated reactions to simulate coal gasification in a bubbling fluidized bed gasifier (BFBG). They determined the coal gasification rates by combining the Arrhenius rate and the diffusion rate for heterogeneous reactions and using the turbulent mixing rate for homogeneous reactions. The flow behaviors of the gas and solid phases in the bed and freeboard were obtained from the analysis. The calculated exit values of gas composition agreed well with the experimental data. They discussed the relationship between gas composition profiles with the height of the gasifier and the distributions of temperature, gas and solid velocities, and solid volume fraction. 1 Wang, Jin, and Zhong [2] developed a three-dimensional numerical model to simulate coal gasification in a fluidized bed gasifier. They considered both gas-solid flow and chemical reactions. They modeled the gas phase with the k-ε turbulent model and the particle phase with the kinetic theory of granular flow. Their analysis considered the coal pyrolysis, homogeneous reactions, and heterogeneous reactions. The reaction rates for homogeneous reactions and heterogeneous reactions employed the Arrhenius-Eddy dissipation reaction rate and the Arrhenius-diffusion reaction rate, respectively. They obtained the flow patterns, gas velocities, particle velocities, composition profiles of gas product, and distribution of reaction rates. The predicted exit gas compositions were in agreement with the experiments. The open-source code MFiX (Multiphase Flow with Interphase exchanges) [3] is a general-purpose computer code developed by the National Energy Technology Laboratory (NETL) of the U.S. Department of Energy for describing the hydrodynamics, heat transfer, and chemical reactions in fluidsolid systems. It has been used for describing bubbling and circulating fluidized beds and spouted beds. MFiX calculations give transient data on the three-dimensional distribution of pressure, velocity, temperature, and species mass fractions. In Part 1, a reacting multi-phase flow model has been developed by using the Eulerian-Eulerian method to solve the multi-phase Navier-Stokes equations, the energy equation, and seven species transport equations with three heterogeneous (gas-solid), two homogeneous (gas-gas) global gasification reactions, and a two-stage volatiles thermal cracking and reforming reaction. Both the finite rate and eddy-breakup reaction models are solved for each homogeneous reaction and the smaller of the two rates is used. The constitutive equations for granular flows are employed to simulate inter-phase interactions between the gas and solid phases including drag, heat and mass transfer, and shear stresses. The effective and bulk properties of the granular flow are obtained via kinetic theory. The objective of Part 2 is to present the results and verification process for building the computational model.
RESULTS AND DISCUSSIONS Since there is no experimental data available for comparison, development of the model starts from simulating single-phase turbulent flow and heat transfer to understand the thermal-flow behavior, followed by employing seven global, homogeneous gasification and thermal cracking reactions by progressively adding one equation at a time. Then, the particles are introduced without reaction to build and verify the nonreactive, multiphase hydrodynamic model of the granular flow. Finally, the heterogeneous reactions are activated between the particles and gases in the entrained-flow and fluidized bed regions. Following this procedure, the computational model can be verified by comparing the calculated mass-weighted, adiabatic flame temperature with the theoretical temperature for exothermic reactions. The endothermic reactions require an energy source to complete, so they must be combined with the exothermic reactions in order to verify the exit temperature. Figure 1 shows the process of building up the simulation model by going through simpler and more fundamental thermalflow phenomena, followed by progressively adding more complex features into the model and eventually achieving the objective of establishing the very complex, multiphase, reactive flow model. Without Particle No Reaction (Case 1) Fluidized Bed Mild Gasifier Homogeneous (Case 2) No Reaction (Case 3) With Particle Heterogeneous (Case 4) Figure 1 A building block diagram showing the process of developing the CFD capability from simple to complex cases. Without going through details of more than 50 cases during the development process, four representative cases are presented here: Case 1: Thermal-flow behavior (no solids and no reactions). Case 2: Homogeneous reaction with volatiles, but no solids. Case 3: Thermal-flow behavior with solid (no reactions). Case 3a: 4 m/s coal particle fed at draft tube's inlet Case 3b: 5 m/s coal particle fed at draft tube's inlet Case 4: Heterogeneous reaction (gas-solid) with volatiles (complete simulation). Validation of Temperature Validation of the reacting flow is conducted first by comparing the mass-weighted, adiabatic flame temperature for the exothermic reactions with the theoretical, adiabatic flame temperature, followed by combining the endothermic and exothermic reactions. The results show the predicted values are all excellent, within 1-4 o C (less than 0.4%) of the theoretical values. Validation of Minimum Fluidization Velocity To verify the gas-solid multiphase interactions, the minimum fluidization velocity is compared with several theoretical and empirical correlations, which are listed in Table 1. To reduce complexity, this validation is conducted with an investigation of thermal-flow behavior with solid particles in a simplified, 2D, preliminary geometry, as shown in Fig. 2. The height and width of the simplified geometry is 1m and 0.3m, respectively. The particle bed is 0.254 m (10 inches) deep. The simplified geometry has seven perforated inlets at the bottom and one outlet at top of the domain. The width of each inlet is 0.02 m. The outlet is 0.1 m wide. The aerodynamic drag force from the air pushes the particles upward and the gravitational force (i.e. the weight of the carbon particles) drives the particles downward. When the inlet air reaches a certain velocity, the upward aerodynamic drag force and downward gravitational force will be equal. This velocity is known as the minimum fluidization velocity, which is an important milestone during simulation iterations because, below this velocity, the particles will remain at the bottom of the bed. If the air velocity is much higher than the minimum fluidization velocity, some portion of the particles could be entrained out through the exit, resulting in a non-ideal condition for operating the gasifier. In the CFD simulation, the minimum fluidization velocity is determined by starting the velocity at a very low speed, 0.2 m/s and a relatively high speed at 5 m/s, which is close to the value (4 m/s) calculated by a correlation given by Kumar and Gupta [4]. Then the bisecting method is used to zig-zag the inlet air velocity values between a high value which shows the particles are moved and a low value which will not move the particles until the critical value (within one decimal value) is achieved. The result of CFD is 2.65 m/s, which is very close to the value calculated by Ergun's correlation [5], as shown in Table 1. An air inlet velocity of 4 m/s (above the CFD determined minimum fluidization velocity) is selected to provide a sustained fluidization condition for all cases. Table 1 Comparison of CFD calculated minimum fluidization velocity values with different correlations with 5 mm diameter carbon solids and 0.5 volume fraction Minimum Correlations fluidization velocity (m/s) Todes and Citovich [6] 1.56 Saxena and Vogel [7] 2.06 Ergun [5] 2.32 Kumar and Gupta [4] 4.00 Miller and Logwinuk [8] 16.32 CFD result 2.65 Figure 2 shows sample snapshots of the dynamic change of volume fraction of carbon particles at regular time intervals between 1.8 and 4.0 seconds. The converged CFD results adequately show the periodic formation of air bubbles rising with expected behavior such as bubble coalescences and breakups. This result verifies the fact that the hydrodynamic model has been adequately set up. 2
circulations at the top of the gasifier has been resolved. The success of this case provides the foundation necessary to move forward and simulate the more complicated cases by using the flow and temperature results of this case as the initial conditions for later cases. K Figure 2 Unsteady variations of volume fraction of carbon particles with 4 m/s air inlet from time intervals between 1.8 and 4.0 seconds in the simplified 2D preliminary geometry Case Studies The 2D mild gasifier used in the simulation domain is shown in Fig. 3. This conceptual design contains the following main components: three inlets, four outlets, a draft tube, perforated plates, and a particle deflector. Discussions of four representative cases are presented below. Case 1: Thermal-flow behavior (no solids and no reactions) The fluidization air enters form the bottom horizontal inlets at 0.42 m/s and 300 K. The velocity 0.42 m/s at the horizontal fluidized gas inlet will become 0.6 m/s at the perforated openings due to the area difference between the fluidized gas inlet and the perforated bed. The air entering the draft tube from the central, bottom inlet is at 0.6 m/s and 1000 K. The inlet conditions are summarized in Table 2. Table 2 Velocity inlet conditions for Case 1 Parameters Case 1 Inlet Position Fluidized Draft tube bed inlet inlet Air inlet velocity at horizontal, m/s 0.42 Air inlet velocity at perforated openings, m/s 0.60 Air inlet velocity at draft tube, m/s 0.60 Inlet temperature, K 300 1000 In Case 1, no particles and reactions are considered, so the result in Fig. 3 shows the thermal-flow mixing of these two air streams at different temperatures. The flow pattern inside the gasifier is complicated, even without the presence of particles. The difficulty in achieving computational convergence due to flow passing through small perforated holes and large 3 Figure 3 Air velocity vector plots colored by temperature (K) distribution without particles and reactions (Case 1) Case 2: Homogeneous reactions (no solids) including volatiles validation against equilibrium results This case investigates the adiabatic flame temperature and the distribution of gas mass-fractions by introducing the five global gasification reactions (Eqs. 1-5) with nitrogen (N 2 ) as an inert gas, together with the two-step volatiles thermal cracking (Eq. 6) and reforming (Eq. 7) in the mild gasifier. Heterogeneous (solid and gas) phase C(s) + ½ O 2 CO (1) C(s) + CO 2 2CO (Gasification, Boudouard reaction) (2) C(s) + H 2 O(g) CO + H 2 (Gasification) (3) Homogenous gas phase CO + ½ O 2 CO 2, (4) CO + H 2 O(g) CO 2 + H 2, (Water-Gas shift) (5) Two-Step Volatiles Cracking and Reforming CH 2.121 O 0.5855 0.5855CO + 0.8532H 2 + 0.069C 6 H 6 (6) C 6 H 6 + 3O 2 6CO +3H 2 (7)
In Case 2, the instantaneous gasification model is used as an intermediate step by assuming the carbon solids change instantaneously to carbon gas. The interphase exchange rates of mass, momentum, and energy are assumed to be infinitely fast. Since the carbon particles are made to gasify instantaneously, the solid-gas reaction processes can be modeled as homogeneous combustion reactions. This approach is based on the locally-homogeneous flow (LHF) model proposed by Faeth [9], implying infinitely-fast interphase transport rates. The instantaneous gasification model can effectively reveal the overall combustion process and results without dealing with the details of the otherwise complicated heterogeneous particle surface reactions, heat transfer, species transport, and particle tracking in turbulent reacting flow. The eddy-dissipation model is used to model the chemical reactions. The eddy-dissipation model assumes that the chemical reactions are faster than the turbulence eddy transport, so the reaction rate is controlled by the flow structures. In other words, the instantaneous gasification model is actually a single phase simulation. All of the above reactions are treated as homogeneous reactions (i.e. Carbon, C(s) treated as "gas" for homogeneous reaction). This is a necessary stepping stone to gain confidence before putting consideration into modeling heterogeneous reactions. The inlet conditions for Case 2 are summarized in Table 3. The results for the distribution of the species mass fractions are shown in Figs. 4 and 5 for reactants and products, respectively, and the velocity vector-plot is shown in Fig. 6, with colors representative of the temperature. Gasification is observed occurring partially in the draft tube and continues in the freeboard region because the productions of CO, CO 2, and H 2 are seen highest at the top of the gasifier. The velocity and temperature profiles in Fig. 6 are different than in Case 1 (Fig. 3) due to the presence of chemical reactions. gasification energy is calculated from the endothermic reactions: Eqs. (2) and (3). Figure 4 Distribution of reactant mass fractions in the mild gasifier with the instantaneous (homogeneous) gasification model including volatiles but without particles (Case 2) Table 3 Inlet, boundary and operating conditions for Case 2 Parameters Case 1 Inlet Position Fluidized Draft tube bed inlet inlet Fuel, oxidant, and gases C gas, H Carbon 2 O vapor, air, gas, air volatiles Gas inlet velocity at horizontal, m/s 0.42 Gas inlet velocity at perforated openings, m/s 0.60 Air inlet velocity at draft tube, m/s 2.00 Inlet temperature, K 300 1000 Mass fraction at inlet: C H 2 O O 2 N 2 CH 2.121 O 0.585 0.188 0.000 0.182 0.630 0.000 0.157 0.015 0.170 0.555 0.103 Operating Pressure (kpa) 101.325 101.325 Wall Temperature (K) Adiabatic Adiabatic The oxygen provided for all simulations is based on the theoretical (stoichiometric) amount needed to produce all gasification reactions necessary for complete carbon conversion. The theoretical energy needed to supply the Figure 5 Distribution of product mass fractions in the mild gasifier with instantaneous (homogeneous) gasification model including volatiles but without particles (Case 2) 4
The equilibrium result of the five global reactions (Eqs. 1 to 5) with nitrogen (N 2 ) and the two volatiles reactions (Eqs. 6 and 7) can be obtained by combining them into one reacting equation. The benzene (C 6 H 6 ) is an intermediate product, so it is not seen in the equilibrium equation. 3C(s) + 2H 2 O (g) +1.207(O 2 +3.76N 2 ) + CH 2.121 O 0.585 3CO+ CO 2 + 3.06H 2 +4.538N 2 (8) The inlet mass fraction from the draft tube inlet consists of: C(s) = 0.157, O 2 = 0.170, H 2 O = 0.015, CH 2.121 O 0.585 = 0.103 (volatiles), and N 2 = 0.555. The fluidized bed inlet consists of: C(s) = 0.188, O 2 = 0.182, and N 2 = 0.630. Based on the equilibrium result of the above reaction equation, the product gas mass fraction is calculated as: CO = 0.321, CO 2 = 0.168, H 2 = 0.023 and N 2 = 0.488. This equilibrium result is to be compared with the CFD result. If the gasifier is large enough and the residence time is long enough, the CFD result should be identical to the equilibrium result. The CFD predicted mass weighted averages of CO, CO 2, H 2, and N 2 compositions at outlet of the domain are 0.322, 0.168, 0.023, and 0.487 respectively, which are almost equal to the equilibrium result of 0.321, 0.168, 0.023, and 0.488, respectively. The mass-weighted average static temperature is 952.35 K at the outlet of the domain. be reached if reactions are considered. Two sub-cases with two different air inlet velocities (4 m/s and 5 m/s, respectively) are studied. Case 3a: 4 m/s solid inlet at draft tube The simulated result of Case 3a with 4 m/s inlet air velocity is shown in Fig. 7. In this case, a small portion of the coal particles reaches the deflector, but most of the coal particles fall over to the fluidized bed at the end of the draft tube. Periodic large-circulation eddies can be seen in the fluidized bed, indicating the selected fluidizing air velocity is adequate and the simulation model runs reasonably well. However, a few of the particles are seen passing straight through the perforated openings and going down to the bottom wall of the domain in a short, transient period in the beginning of the simulation even though the particle diameter (5 mm) is larger than the perforated opening (3.8 mm). Those coal particles are seen to remain at the bottom of the gasifier for the remainder of the simulation. With an in-depth examination, it is concluded that this unrealistic phenomenon is caused by the Eulerian method of the current model which employs the void fraction approach. Different from the discrete particle model (DPM) implemented in a typical Lagrangian particle tracking method, the actual particle size does not actually exist in the present approach of using the void fraction. In the void fraction approach, the particle diameter is used with the particle density, particle velocity, and the total mass flow rate to calculate the solidity (or void fraction which equals 1 - solidity). Therefore, when the gases pass through the perforated plate, the solid portion also passes through the perforated plate. This modeling flaw, although it can't be overcome easily at this point, definitely needs to be resolved in the future. However, because of the following reasons, no further effort is made to resolve this issue in this study: (a) the amount of gases passing downward through the perforated plate is not significant compared to the remainder of the flow, (b) it only occurs in a short, initial transient period, (c) the questionable body of the multiphase flow is trapped and remains stationary at the bottom of the gasifier, (d) this phenomenon does not affect the simulation mechanisms taking place in the other parts of the gasifier, and (e) it is not a trivial task to resolve this problem immediately. The inlet conditions are summarized in Table 4. Table 4 Different inlet velocity values for Cases 3a and 3b Figure 6 Gas velocity vector-plots colored by temperature (K) using the instantaneous (homogeneous) gasification model including volatiles but without particles (Case 2) Case 3: Thermal-flow behavior with solids (no reactions) Initially, the 5 mm diameter carbon particles at room temperature (300 K) are placed side by side in the fluidized bed, with a bed-depth of 10 inches (0.254 m). The air passes up through the perforated interior (holes) coming from the two horizontal fluidization gas inlets with a velocity of 2.8 m/s and at a temperature of 300 K. The carbon, transported by air, enters the draft tube with water vapor through the vertical draft-tube's bottom inlet at 1000 K to reflect the high temperature that can Parameters Case 3a Case 3b Inlet Position Fluidized Draft Fluidized Draft bed inlet tube inlet bed inlet tube inlet Air inlet velocity at horizontal, m/s 2.8 2.8 Air inlet velocity at perforated openings, m/s 4.0 4.0 Air inlet velocity at draft tube, m/s 4.0 5.0 Carbon solid inlet velocity at draft tube, m/s 4.0 5.0 Inlet temperature, K 300 1000 300 1000 5
Case 3b: 5 m/s solid inlet at draft tube The velocity vector-plots with 2.8 m/s fluidized gas inlet for the horizontal, bottom inlets and 5 m/s solid particles at the draft tube inlet (Case 3b) are shown in Figs. 8 and 9. The colors of the vectors correspond to the air temperature. The purpose of this case is to examine the effects of increasing the carbon transport air speed in the draft tube on the particle entrainment rate from the char bed into the draft tube near the bottom of the gasifier and the potential for coal particles to escape from the outlet. It can be seen from the figure of carbon solid volume fraction for Case 3b, a large amount of the carbon particles at the 5 m/s draft tube inlet reaches the deflector and are deflected by the deflector after 0.4 seconds and come back downward to the fluidized bed area after 0.6 seconds. This kind of particle deflection is minimal and almost absent in Case 3a. From the particle velocity vector plot in the enlarged insert in Fig. 8a, Case 3b shows more particles are entrained from the char bed into the draft tube than in Case 3a near the bottom of the draft tube. The particle velocity of Case 3a is not shown here to save space, but it is documented in [11]. Figure 7 Distribution of volume fraction of carbon solid particles with 2.8 m/s at horizontal fluidization inlet and 4 m/s solid inlet at draft tube in different interval (Case 3a) Figure 8 Velocity vectors plot for (a) particles and (b) air colored by temperature (K) distribution with 2.8 m/s fluidization air at the horizontal inlet and 5 m/s solid at the draft tube inlet (Case 3b) 6
Table 5 Boundary and operating conditions for Case 4 Parameters Case 4 Inlet Position Fluidized Draft tube bed inlet inlet Fuel Air C solid, air, volatiles Air inlet velocity at horizontal, m/s 2.8 Air inlet velocity at perforated plate, m/s 4.0 Air inlet velocity at draft tube, m/s 4.0 Solid C inlet velocity at draft tube, m/s 4.0 Inlet temperature, K 300 300 Mass fraction at inlet: O 2 N 2 CH 2.121 O 0.585 0.233 0.767 0.000 0.120 0.395 0.485 Operating pressure (kpa) 101.325 101.325 Operating temperature (K) 288.16 288.16 Operating density (kg/m³) 1.225 1.225 Gravitational acceleration (m/s²) 9.81 9.81 Wall temperature (K) Adiabatic Adiabatic Figure 9 Transient distribution of volume fraction of carbon solid particles between 0.2 and 2.2 seconds for Case 3b. Case 4: Heterogeneous (gas-solid) reaction with volatiles The result of Case 3a are used as initial conditions for Case 4, and both homogeneous (gas-gas) reactions and heterogeneous (gas-solid) reactions are added in this case. The inlet conditions are summarized in Table 5. As shown in Fig. 10, some carbon solid particles are carried away by the syngas at the syngas exits at the top of the gasifier. During an actual gasifier s operation, these particles will be collected through a cyclone unit outside the gasifier. In this case, the finite rate/eddy-dissipation model is used for the three heterogeneous reactions, Eqs. (1) to (3), based on the work of Mann and Kent [10]. A user defined function (UDF) must be written and used in this case for getting the reaction rates effects incorporated into the simulation. The eddy-dissipation model is still used for the two homogeneous reactions (Eqs. 4 and 5) and the two volatiles reactions (Eqs. 6 and 7.) The residence time is not sufficient for all of the reactions to be completed when the flow exits the gasifier. The distribution of various species mass fractions in the mild gasifier is shown in Figs. 10 and 11. Volatiles are seen to be thermally cracked on the way up the draft tube (Fig. 10) with some trace of benzene remaining in the freeboard region (Fig. 11). Both CO and H 2 are produced and collected in the freeboard region (Fig. 11). Fig. 12 shows the particle and air velocity vectors colored by temperature distribution. The oxygen in the fluidization air activates exothermic reactions to bring the temperature of the char particles to above 1000 K which is hotter than the corresponding gas mixture at the same location, while the gas mixture temperature increases to 1150K in the area above the fluidized bed where the char cools down. The temperature in the draft tube is uncharacteristically low. This may be caused by limited air supply in the draft tube, so the exothermic reactions rates are lower in the draft tube. Figure 10 Distribution of reactant gas mass fraction and char particle volume fraction for Case 4 The mass weighted averages of CO, CO 2, H 2, C 6 H 6, (Benzene) and N 2 (as inert gas) at the outlet of the domain are 0.2706, 0.1781, 0.0228, 0.0524, and 0.3687, respectively. The mass-weighted average temperatures for the gas and solid phases at the outlet are 907.04 K and 856.45 K, respectively. 7
the solution will be left for future study when more computational resources are available. Table 6 Grid sensitivity study of Case 4 Parameters Coarse 6,960 cells Medium 30,876 cells Fine 65,355 cells Exit gas temperature (K) 1311.61 938.69 840.64 Exit mass fraction of CO 0.1552 0.1915 0.1771 CONCLUSIONS This study began with the simulation of a single-phase, turbulent flow and heat transfer inside a 2D, mild gasifier. An intermediate step was conducted by employing the instantaneous gasification model, which progressively added one reaction at a time to eventually include all seven reactions. Finally, the particles were introduced with heterogeneous reactions. When no experimental data are available for verification, this progressive building process from simple to complex models allows step-by-step examinations of the effect from each addition of new parameters to ensure that the simulated results are physically reasonable and fundamentally sound. The results are summarized below: Figure 11 Distribution of product gas mass fraction including volatiles for Case 4. (a) (b) Figure 12 Temperature (K) distribution of (a) carbon solid (b) gas mixture, and (c) volume fraction of carbon solid for Case 4. Grid Sensitivity Study for Case 4 A grid sensitivity study has been conducted using three different grids: a coarse grid (6,960 cells), an intermediate grid (30,876 cells), and a fine grid (65,355 cells). Parameters and operating conditions for Case 4 given in Table 6 are used in this grid sensitivity study. Table 6 shows the mass-weighted average temperature and species mass fractions of the exit gas for all grids. In Table 6, the exit temperature and product gas compositions are different for different grid sizes. Therefore, the simulation is very sensitive to grid size. It can be seen that the solutions, although all achieve convergence, have not reached grid-independence. Since the goal of this project of establishing the simulation model has been achieved, further refinement of (c) 1. The CFD predicted species composition of product gas using the instantaneous gasification approach was almost exactly equal to the results obtained from the equilibrium calculation. 2. The minimum fluidization velocity was found to be 2.65 m/s seconds, which is close to the 2.06 m/s calculated from Ergun s correlation. 3. The deflector was found to be successful in deflecting the majority of the particles, but some char particles still escaped and were carried out by the syngas stream. 4. The single-phase flow field prediction (without particles) can provide a quick and relatively meaningful guidance of the multiphase flow field, especially outside the fluidized bed. 5. The developed model of multiphase flow with heterogeneous reactions in this study is successful, but continuous efforts are required to improve this model, particularly in regards to resolving the issue of particles passing through the perforated plate with holes smaller than the particles. In this study, full gasification is achieved due to high draft tube inlet flow and fluidization velocities. The future study will use the developed model to simulate different operating conditions in order to produce different levels of mildlygasified syngas. The potential parameters that can be varied to affect the gasification level include controlling the oxygen content in the fluidization air flow (e.g. by adding nitrogen), or by changing the speeds of the draft tube inlet flow and fluidization air together with controlling the temperature. 8
ACKNOWLEDGMENTS This study was partially supported by the Louisiana Governor's Energy Initiative via the Clean Power and Energy Research Consortium (CPERC) under the auspices of Louisiana Board of Regents and partially supported by the Department of Energy contract NO. DE-FC26-08NT01922. REFERENCES 1. Yu, L., Lu, J., Zhang, X., and Zhang, S., 2007, "Numerical Simulation of the Bubbling Fluidized Bed Coal Gasification by the Kinetic Theory of Granular Flow (KTGF)," Fuel, Vol. 86, 2007, pp 722-734. 2. Wang, X., Jin, B., and Zhong, W., 2009, "Three- Dimensional Simulation of Fluidized Bed Coal Gasification," Chemical Engineering and Processing: Process Intensification, Vol. 48, 2009, pp 695-705. 3. WWW.MFIX.org, July, 2011. 4. Kumar, A. and Sen Gupta, P, 1974, "Prediction of Minimum Fluidization Velocity for Multicomponent Mixtures," Indian Journal of Technology, No. 12, pp 225-227, May 1974. 5. Ergun, Sabri, 1952, "Fluid Flow through Packed Columns," Journal of Chemical Engineering Progress, Vol. 48, No. 2, 1952, pp 89-94. 6. Todes, O.M. and Citovich, O.B., 1981, "Reactors with Coarse Particle Fluidized Beds (in Russian)," Leningrad, Khimiya. 7. Saxena, S.C. and Vogel, G.S., 1977, "The measurements of incipient fluidization velocities in a bed of coarse dolomite at temperature and pressure," Trans. Inst. Chem. Eng., Vol. 3, pp 184 195. 8. Miller, C.O. and Logwinuk, A.K., 1951, "Fluidization studies of particles," Ind. Eng. Chem., No. 43, pp 1220 1226, 1951. 9. Faeth, G.M., 1987, Mixing, Transport and Combustion in Sprays, Progress in Energy Combustion Science, Vol. 13, pp. 293-345. 10. Mann, A.P. and Kent, J.H., 1994, "A Computational Study of Heterogeneous Char Reactions in a Full-scale Furnace," Combustion and Flame, Vol. 99, Issue. 1, October 1994, pp 147 156. 11. Mazumder, A. K. M. and Wang, T., 2010, "Development of a Simulation Model for Fluidized Bed Mild Gasifier," ECCC Report 2010-05, Energy Conversion and Conservation Center, University of New Orleans. 9