Water. desalination + reuse

RESEARCH | 46 | Desalination & Water Reuse | August-September 2014 References 1.NSJ Fawcett (1997), The hydraulics of flotation tanks; computational modelling- dissolved air flotation, The Chartered Institution of Water and Environmental Management, 1, pp 51-72. 2.M Lundh, L Jonsson and L Dahlquist (2002), The influence of contact zone configuration on the flow structure in a dissolved air flotation pilot plant, Water Research, Vol. 36, pp 1585-1595. 3.M Lundh, L Jonsson and L Dahlquist (2000), Experimental studies of the fluid dynamics in the separation zone in dissolved air floatation, Water Research, Vol. 34, pp 21-30 Figure 3. CFD model simulation The plotted CAD files were exported to binary standard tessellation language format to be used in Matlab. The tanks were subsequently reconstructed in an xyz coordinate system using Matlab. Velocity files were also loaded in the computer with assigned location data. Three-dimensional flow plotting with cones can easily present velocity direction and magnitude at the same time through the size and angle of cones. Moreover, by showing slice planes in designated space with different colour-mapping functions, flow movement and directional change can easily be analyzed. Additional plot view functions, including three-dimensional rotation and zooming, help in the analysis of velocity at a specific point or angle. Flow pRoFilE AnAlySiS Studies of visualized velocity vectors in a pilot-scale DAF enabled the analysis of a velocity profile at high loading rate (figure 4). The fastest flow in the tank was found right after the air-bubble generation nozzle. Due to faster flow from the nozzle at the centre, the side part of the contact zone showed slower velocity flow vectors, and even opposite Z-directional vectors. And the slowest vector was found near the topside of the separation zone. Flow from the contact zone to the separation zone is faster. However, most flow vectors in the separation zone were significantly slow except on the edges (figure 5). Flow vectors in the top part of the separation zone tended to go downwards at first followed by a large mass of flow movement, however, upward directional vectors working as complementary vectors were found at the opposite side of the separation zone. Flow vectors in the separation zone were only about half of those on the top layer. The difference could give the benefit of separating aggregated flocs from the water. Because the outer flow vectors in the separation zone were relatively faster than the inner vectors, flocs may have longer travel pathways in the pilot geometry and have higher possibility to be removed by bubbles. The plotted flow vectors in this study clearly revealed specific patterns in the DAF tank. This indicates that flow profiling using ADV and Matlab can provide efficient visualization of the flow patterns in the DAF tank which could be used to optimize DAF operations. l Figure 2. Axis orientation of side-looking ADV probe (left) and a mounted ADV probe in a DAF tank (right).

• Cover