It has been known that 0.1% to 5% redundancy (or looseness) in fiber length relative to the distance between the two headers greatly reduce membrane fouling tendencies (Henshaw, 1996). But the looseness of >10% is likely to lead to membrane breakage (Cui, 2003). Fig. 1 shows the effect of fiber looseness on membrane fouling rate, where fouling rate significantly decreases when looseness increases from 0% to 1%. Further increase of looseness to 4% hardly decreases membrane fouling further under the experimental condition.
The loosely held fibers randomly move within the range set by the looseness in two phase flow. The range of fiber movement (or amplitude) can be monitored in clean water using a video camera (Wicaksana, 2006). The cycle time of fiber movement decreased as air flow rate increased, but there was a minimum cycle time for a given membrane size and length. At the same time, the fiber amplitude, which meant the range of fiber movement, also increased with air flow rate, but it had a maximum limit, too. These observations agree very well with the prior findings of the existence of the maximum flux achievable by increasing air flows as shown here (Ueda, 1997).
Fig. 1. Effect of fiber looseness on membrane fouling rate. Fiber length=50cm, OD=0.65 mm, ID=0.39mm, pore size = 0.2 micron, polypropylene, flux=30 LMH, 5g/L yeast, scouring air flow=1.5L/min, nozzle size = 1mm (Wicaksana, 2006).
The pseudo-steady-state fluxes at three different conditions are compared in Fig. 2a, where reservoir water was filtered. When single phase flow (water) was used to scour the tightly held membranes, fluxes were the lowest. When air bubbles were added while maintaining identical water flow rates, flux increased significantly. The highest flux was obtained when membranes were held loosely, where 420 mm fiber was held in between the two headers with 410 mm space (tightness = 97.6% or looseness = 2.4%). Fig. 2b compares two TMP profiles of tightly held fibers and loosely held fibers at 30 LMH. No TMP increase was observed for 300 minutes for the fiber with 5 % looseness, but TMP increased from 4 kPa to 50 kPa in the same time period for tightly held fibers.
The minimum fiber length required to take advantage of the fiber looseness appears dependent on the experimental conditions including air flow rate, fiber size, fiber material, air flow rate, etc., but it appears that short fibers with < 200 mm have little beneficial effects from the fiber looseness.
Fig. 2. Effect of looseness of hollow fiber on membrane fouling tendency in two phase flow: a) measured by pseudo-steady-state permeate flux, where GE’s ZW500 fibers were used after cutting fibers at 420 mm length (Bérubé, 2006b), b) measured by TMP (Chang, 2002a).
© Seong Hoon Yoon