Effects of the Distributor and Baffle on the Hydrodynamic Performance of a Bubbling Fluidized Bed

Xuelian Xing^a, Cedric Briens[*]^a, Chao Zhang^b

^aInstitute for Chemicals and Fuels from Alternative Resources(ICFAR),

The University of Western Ontario

1151 Richmond St, London, ON N6A 3K7

^bDepartment of Mechanical and Materials Engineering,

The University of Western Ontario

1151 Richmond St, London, ON N6A 3K7

Abstract:

In this study, the multi-phase Eulerian-Eulerian two-fluid method (TFM) coupled with the kinetic theory of granular flow (KTGF) was used to investigate the hydrodynamics of particle flows (Geldart Group B) in a lab-scale bubbling fluidized bed. The focus of this study was to develop a technology to improve the bubble flow behavior inside the fluidized bed to reduce the formation of wet agglomerates when the liquid was sprayed into the fluidized bed. The effects of different gas distributors and baffles on the gas-particle mixing quality and bubble distribution for various superficial gas velocities were investigated. Different gas-solid drag models used in CFD simulation were tested and the results from those drag models were compared. The predicted voidage profiles under different operating conditions agree well with the experimental data.

Key words: Baffle, Bubbling fluidized bed, Hydrodynamic, Eulerian-Eulerian method, CFD simulation

Introduction

Bubbling fluidized bed has been widely applied in petroleum oil upgrading processes, owing to its inherent benefits, such as high rates of heat and mass transfer and quick mixing of solids. However, particle agglomeration occurs when the liquid is sprayed into a gas-solid fluidized bed. Undesired agglomerates will decrease heat and mass transfer rates, causing operating problems. Variation of the bubble flux across the section over fluidized bed plays an important role in helping the bed to reach an expected flow pattern and reduce agglomeration.

Various experimental methods have been used to determine the radial gas bubble distribution including digital image analysis, optical fiber probe, X-ray computed tomography and radioactive particle tracking¹². However, each technique has its own drawbacks and limitation. Digital image analysis is limited to very dilute systems due to the opaque nature of gas-solid fluidization beds. Multiple bubbles are difficult to be detected by X-rays³. In this study, sturdy triboprobes were used.

Currently, the most commonly used approaches for simulating the gas-solid two-phase flow in a fluidized bed are Eulerian-Lagrange approach [also denoted as the Discrete element method (DEM)] and Eulerian-Eulerian (E-E) approach [or called as the Two-fluid model (TFM)]. In the TFM, both gas and particle phases are treated as interpenetrating continua, each phase has its own governing equations for momentum, continuity, and energy. The kinetic theory of granular flow (KTGF) is introduced to close the governing equations for the solid phase. In the DEM, the trajectory of each particle is tracked by Newton’s second law, which makes the DEM method more computational expensive compared to the TFM⁴. For these reasons, the DEM is not widely used for the simulations of two-phase flows in fluidized beds. In conclusion, the DEM method is more suitable for dilute flows, while the TFM) is more suitable for dense flows.

Therefore, in this present study, the hydrodynamics of particle flows in a bubbling fluidized bed was studied by a 2-D transient two-fluid model using a commercial CFD package (ANSYS Fluent 18.2). Afterwards, the effects of the gas distributor and baffles characteristics (shape and position) on the gas distribution in the bubbling fluidized bed were investigated.

METHODOLOGY

The experiments were carried out in a fluidized bed using a gas distributor with a 45˚ slope. There were 20 tuyeres distributed evenly on the gas distributor, and the gas flow rate supplied to each tuyere was individually controlled to control the lateral distribution of the gas. Three gas distribution schemes were used, each with 10 active tuyeres: the “Western case” used tuyeres on the higher side of the sloped distributor, the “Eastern case” used tuyeres on the lower side and the “Flat case” used tuyeres uniformly distributed over the entire distributor. Air was used as the fluidizing gas with a superficial gas velocity ranging from 0.1 to 1 m/s. Silica sand with a Sauter-mean diameter of 190 μm was used. Baffles inside the fluidized bed were added later to study and compare the bubble distribution profile with those without baffles. The baffle used in this research has a 45° angle, an 18 cm by 18 cm triangular geometry and occupies 36 % of the cross-sectional area of the bed.

The simulations were carried out for 30 seconds for each case, but only the results from the last 10 seconds were used to obtain the time-averaged voidage profile and gas velocity distribution, since stable results were obtained after 20 seconds.

Results and discussion

For the bubbling fluidized bed without the baffle, the sensitivity tests for different mesh sizes were performed. Different grid sizes in the gas distributor and fluidization sections were compared. Four different grid sizes (0.0005 m, 0.001 m, 0.002 m, 0.004 m) for distributor section and four different grid sizes (0.001 m, 0.002 m, 0.004 m, 0.006 m) for the fluidization section were calculated. All the simulations performed at a superficial gas velocity 0.2 m/s with the initial bed height 1.6 m, with further simulation runs planned at higher gas velocities. The time-averaged mean gas velocity and solid volume fraction at different bed heights were checked. It was found that, when the number of nodes is over 83392, the simulation results are independent on the mesh size. For this reason, the mesh with 83392 nodes was used. The effect of different drag models (Gidaspow, Gibilaro, Syamlal-Obrien, Wen-Yu, and Syamlal-Obrien-para models) on the simulations results was studied. It was found that the model from Syamlal-Obrien-para gave a better agreement with the experimental data.

The results indicated that the gas distributor and baffle significantly affected the gas volume fraction and velocity distribution in the fluidized bed. The geometry of the gas distributor played an important role in the solid volume fraction and gas velocity distribution. Furthermore, the presence of baffles is an important factor affecting the direction of the gas bubble flows.

Reference

1. Dubrawski, K. et al. Traveling column for comparison of invasive and non-invasive fl uidization voidage measurement techniques. 235, 203–220 (2013).
2. Tebianian, S. et al. Investigation of particle velocity in FCC gas- fl uidized beds based on different measurement techniques. Chem. Eng. Sci. 127, 310–322 (2015).
3. van Ommen, J. R. & Mudde, R. F. Measuring the Gas-Solids Distribution in Fluidized Beds -- A Review. Int. J. Chem. React. Eng. 6, (2008).

4. ANSYS Fluent Theory Guide. (2017).

[*] Corresponding author.

Email: cbriens@uwo.ca (C. Briens)