Cross-flow filtration of fruit juice 6 Scaling6.1 Transparent housing Filtration experiments with other medias have shown difficulties in 1m2 scale. To investigate the problems, a flexible experimental set-up has been established, for filtration of water. 6.1 Transparent housing A transparent filter house has been made, to be able to follow the flow inside, by adding colour to the water. Colour experiments Experiments with colour added to the water used for BackShock, and the BackShock device placed at the inlet end of the filter did not produce any useful information. When the BackShock device was moved to the middle of the filter, more information was revealed. From the middle to the outlet all the water was coloured, but the colour did not move towards the inlet end at all. When the colour was added to the feed water instead, it could be seen what was happening. The feed passes the membrane to the permeate side, here it flows outside the fibres to the outlet end, where it passes the membrane once again. Because of the high water permeability of the filter the resistance is lower going this way, than flowing in the fibres. In a filtration situation with a partly fouled filter, this effect will be reduced. It should however be taken into account. 6.2 Shadow effect It has been proposed, that the fibres closer to the BackShock-inlet would absorb all the introduced liquid, leaving little or nothing for the rest of the bundle. To investigate this, selected fibres of a filter have been supplied with small tubes in one end. The tubes should collect the water leaving that end of the filter. The volumes collected seemed to point to a small shadow effect when using low BackShock pressures. The effect dissapeared when the BackShock pressure was increased. 6.3 Pressure measurements A filter house was modified to have six ports instead of two. This allowed for more detailled pressure measurements to be undertaken. Pressure transducers were placed at the ports and before and after the filter. The transducers were connected to a computer running a data acquisition program. The pressure peaks for 1m2 filters show up much higher than for smaller filters. This indicates an increased resistance against moving the bigger volumes needed. 6.4 The BackShock device The traditional BackShock device consisteded of a rubber tube inside a steel house. The device was connected directly to the filter, in a way that allowed the permeate to enter the inside of the rubber tube. The BackShock is performed by closing the permeate outlet and applying compressed air to the outside of the rubber tube. The tube will collapse, pressing the permeate in it out through the filter. Unfortunately this approach leaves no way of controlling the volume of permeate flushed. If a constant BackShock time and pressure is used, the volume will be high when the filter is still clean and permeable. When the filter fouls the permeability will decrease, which in turn will decrease the volume of permeate flushed. In other words; when the demand is still low, a high volume will be flushed, but when the demand is high, only a small volume is flushed. After the BackShock, the compressed air is removed. This leads to a dramatic pressure drop at the permeate side. This will cause the permeate to rush back into the BackShock device to refill it. The high flow that will exist for a short while after the BackShock, will drag the removed dirt back into the pores, thereby undoing the BackShock. 6.5 Conclusion Colour experiments have shown that the water takes a shortcut through the membrane to the permeate side and back again. There is no reason to worry about the shadow effect. Pressure measurements show significant problems trying to flush with the volumes needed for bigger filters. To be able to scale up the BackShock technique, it is nessecary to solve some problems in controlling the BackShock. Other projects will continue development of the ideas from this project.
|