CFD-DEM Simulation of Hydrodynamics in Wet Gas-Solid Fluidized Beds

It is common in many industries, including energy, pharmaceuticals and polymer production, for small amounts of liquid to be injected into gas-solid fluidized beds. This liquid can serve as a reactant or facilitate agglomeration or heat transfer. The liquid often covers the surface of particles and proceeds to form cohesive liquid bridges between particles, ultimately leading to the agglomeration. The cohesive force caused by small amounts of liquid can alter hydrodynamics dramatically, since agglomerates, rather than individual particles, need to be fluidized. In extreme cases, liquid bridging and agglomeration can lead to local or device-scale defluidization. Despite this industrial importance, the effect of wetness on hydrodynamics is not fully understood. A limited number of experimental studies show differences in bed fluidity [1,2], particle velocities  [3–6], bed height  [1], bubble size  [6,7] and minimum fluidization velocity  [6] caused by liquid bridging in gas-solid fluidized beds.

Various efforts have been undertaken to model the effect of wetness on fluidization computationally. Most work has involved an Euler-Lagrange approach often referred to as the Computational Fluid Dynamics – Discrete Element Method (CFD-DEM)  [8] in which each individual particle is modeled, gas flow is simulated on a CFD grid and the two phases are linked via a drag law. In this approach, contacts between particles are modeled, often with a coefficient of restitution dictating the level of inelasticity in the collisions. A number of studies have accounted for wetness by lowering this coefficient of restitution  [9,10], since the cohesive force provided by liquid bridges will act to slow the particles from rebounding off of one another after a collision. A fundamental issue with approach is that liquid bridging does not merely make collisions more inelastic; liquid bridges provide cohesion necessary to keep particles together for long periods of time and form agglomerates. To address this issue with a modified coefficient of restitution approach, a few more detailed studies have directly modeled forces arising from liquid bridges  [11,12]. One issue with these studies is that they have assumed liquid bridges to form instantaneously. In fact, the surface tension and viscosity of the liquid will affect the rate at which liquid pours from the surface of a particle into a liquid bridge  [13], and in certain cases, particles can rebound off of one another before a significant liquid bridge has had time to form.

In this study, we model the effects of liquid on fluidization hydrodynamics using a CFD-DEM model in which liquid bridges are modeled directly and liquid bridges form at a finite rate. We simulate laboratory-sized fluidized beds and investigate the effects of liquid content, viscosity and surface tension on bed height, particle velocity, bubble behavior and minimum fluidization velocity for comparison with effects seen in the literature.

References:

[1]    S. L. McDougall, M. Saberian, C. Briens, F. Berruti, and E. W. Chan, Int. J. Chem. React. Eng. 2, (2004).

[2]    S. McDougall, M. Saberian, C. Briens, F. Berruti, and E. Chan, Chem. Eng. Process. Process Intensif. 44, 701 (2005).

[3]    V. S. Sutkar, N. G. Deen, A. V. Patil, E. A. J. F. Peters, J. a. m. Kuipers, V. Salikov, S. Antonyuk, and S. Heinrich, AIChE J. 61, 1146 (2015).

[4]    Q. Zhang, Y. Zhou, J. Wang, B. Jiang, Y. Yang, S. Stapf, C. Mattea, and Q. Gong, Chem. Eng. Technol. n/a (2015).

[5]    M. L. Passos and A. S. Mujumdar, Powder Technol. 110, 222 (2000).

[6]    Y. Zhou, C. Ren, J. Wang, and Y. Yang, AIChE J. 59, 1056 (2013).

[7]    J. Ma, D. Liu, and X. Chen, Ind. Eng. Chem. Res. 55, 624 (2016).

[8]    Y. Tsuji, T. Kawaguchi, and T. Tanaka, Powder Technol. 77, 79 (1993).

[9]    P. Darabi, K. Pougatch, M. Salcudean, and D. Grecov, Powder Technol. 214, 365 (2011).

[10]  V. S. Sutkar, N. G. Deen, J. T. Padding, J. a. m. Kuipers, V. Salikov, B. Crüger, S. Antonyuk, and S. Heinrich, AIChE J. 61, 769 (2015).

[11]  T. Mikami, H. Kamiya, and M. Horio, Chem. Eng. Sci. 53, 1927 (1998).

[12]  M. Girardi, S. Radl, and S. Sundaresan, Chem. Eng. Sci. 144, 224 (2016).

[13]  M. Wu, J. G. Khinast, and S. Radl, in (Barcelona, 2014).