Numerical simulation and experimental study of suspended particulate matter removal for efficient water recovery and reuse in solid-liquid separation

Abstract

In the present study, an empirical investigation was undertaken to evaluate the efficacy of a hydrocyclone in separation processes, followed by its implementation within a biogas plant setting. The laboratory phase employed sawdust as a surrogate material to facilitate hydrocyclone testing, while biogas slurry served as the practical material for subsequent experimentation within the biogas plant. The separation efficiency was approximately 50% for particles in the 50-200-micron range. Computational Fluid Dynamics (CFD) simulations were performed using the Fluent module embedded within ANSYS, employing the Renormalization Group (RNG) k - ε turbulence model to numerically solve the three-dimensional Navier-Stokes equations, thereby facilitating precise predictions of swirl-induced phenomena. However, a notable disparity between the numerical results and experimental data was observed. Further refinement and calibration of the numerical model are required to align it more closely with the acquired experimental insights. In conclusion, the study suggests that approximately 40% of wastewater can be reclaimed and reused using the evaluated method. The overall study demonstrates the potential of the hydrocyclone to improve the efficiency of separation processes within a biogas plant setting.

Keywords: Anaerobic digester; Biogas plant; CFD; Experimental study; Hydrocyclones.

Fig. 1

Design parameters.

Fig. 2

Schematic of the experimental setup.

Fig. 3

Experimental setup in the Laboratory.

Fig. 4

Flow rate Vs collection efficiency plot. Notice due to the effect of the coagulant, the collection efficiency declines.

Fig. 5

Schematic of the BMP setup.

Fig. 6

(a) BMP, (b) Experimental setup, (c) BWR, (d) Dried slurry collected at CP.

Fig. 7

(a) 3D geometry of the separator, (b) Meshed fluid domain, (c) and (d) illustrates the sections of the meshed separator section.

Fig. 7

(a) 3D geometry of the separator, (b) Meshed fluid domain, (c) and (d) illustrates the sections of the meshed separator section.

Fig. 8

Results along the axis for different mesh a. Pressure b. Velocity. Notice the significant overlap between the medium and fine meshes for both velocity as well as pressure plots.

Fig. 9

Static pressure (a, b) and tangential velocity (c, d) contours are illustrated for medium and fine meshes respectively. Notice the minimal change in the flow variables and the flow field.

Fig. 9

Static pressure (a, b) and tangential velocity (c, d) contours are illustrated for medium and fine meshes respectively. Notice the minimal change in the flow variables and the flow field.

Fig. 10

Different contours (for inlet velocity 0.81 m/s): (a) Static pressure, (b) Tangential velocity, (c) Particle tracks- Particle trajectories reveal cyclonic motion, with heavier particles collected at the underflow and lighter particles exiting through the overflow, (d) Velocity streamlines- Swirling flow patterns highlight the interaction between the inner and outer vortices, both rotating counterclockwise.

Fig. 10

Different contours (for inlet velocity 0.81 m/s): (a) Static pressure, (b) Tangential velocity, (c) Particle tracks- Particle trajectories reveal cyclonic motion, with heavier particles collected at the underflow and lighter particles exiting through the overflow, (d) Velocity streamlines- Swirling flow patterns highlight the interaction between the inner and outer vortices, both rotating counterclockwise.

Fig. 11

Flow rate Vs collection efficiency plot. The collection efficiency predicted by the CFD simulation suddenly declines as the RNG k − ε turbulence model, while effective for capturing swirling flows, may not fully account for the complex interactions at high flow rates, especially near the vortex finder or in regions with strong shear layers.