It has been well known that internal pressure drop in hollow fiber membrane can cause excessively high flux near the permeate exit, which not only causes quick cake layer formation in the local area, but also causes cake layer compaction. The excess internal pressure drop eventually causes flux decline near the permeate exit. Under a constant flux mode, flux in upstream must increase to maintain average flux constant, which in turn causes higher fouling potential in the upstream. Eventually the fouling spreads from permeate exit to the other end of the fiber.
As fouling proceeds, the intact area available in upstream declines and the lost permeability in downstream must be compensated by the increasing flux in the smaller intact area in upstream. As a result of decreasing intact membrane areas, membrane fouling rate accelerates with time. This might be an additional mechanism of sudden TMP rise specifically for hollow fiber membranes on the top of the cake layer compaction phenomenon that is discussed here.
Above concept has been experimentally proven. When the porosities of the cake layer formed in each segment of the reverse U-shaped hollow fiber bundle was analyzed by confocal laser scanning microscopy (CLSM), it was found that cake layer porosity was the lowest near the permeate exit and tended to increase toward the upstream as shown in Fig. 1 and Fig. 2 (Lee, 2008). The high porosity ranging 60-75% in Fig. 2 is unrealistically high, but it happens when threshold is set to distinguish pores from cake, which is rather subjective process. However, the trends in the graph are still effective. It is also noticeable that such imbalanced filtration along the hollow fiber occurred in a short membrane with 15 cm length. If longer fibers were used, cake porosity changes would be more dramatic
Fig. 2. Local changes in the bio-cake porosity along the membrane fiber length in an MBR (Lee, 2008).
Fig. 3 shows a mass of cake layer attached on five sections of a hollow fiber after 6 hr filtration under the condition employed (Li et al., 2013), where a single hollow fiber membrane with both ends open was used to filter 5 g/L yeast solution with air scouring. In the figure, TD1 to TD5 represent the equally spaced five sections of a hollow fiber from one end to the other end. It is apparent that more foulants deposits in both ends of the hollow fiber than in the middle, which is caused by higher flux in both ends than in the middle.
Fig. 3. Gravimetric analysis of the membrane samples obtained from the corresponding areas detected by TD1-TD5 after 360 min of fouling operation, where TD1-TD5 are the five pieces of hollow fibers with identical lengths from one open end to the other open end (Li et al., 2013).
Advantages of long hollow fiber membranes that can partially or fully compensate their disadvantages
In spite of the internal pressure drop (or loss) and the associated membrane fouling, long fibers provide some benefits that can at least partially offset the disadvantages from the internal pressure drop.
- First, as discussed here, the maximum amplitude of fiber is proportional to the fiber length at a given looseness. While the random lateral fiber motion is a principle antifouling mechanism of hollow fiber, the larger amplitude of longer fibers can partially offset the increased membrane fouling by the internal pressure drop.
- Second, the frequency of random fiber motion is proportional to the fiber length. If one segment of a long hollow fiber is suddenly pulled by turbulence, the sudden motion spreads through the fiber. Longer fibers have more exposure to the turbulence so that they have more chances to have sudden movement.
- Third, scouring air can be used more efficiently with longer fibers. Although more energy is required to generate same scouring air flow at higher static pressure, power consumption is not linearly proportional to the head pressure. Overall, energy efficiency increases with longer fibers at deeper water.
- Fourth, module costs can be reduced since more membrane area is available per permeate header. Fiber potting in headers is a major cost item in module production rather than fiber spinning.
A new way to optimize hollow fiber membrane dimension
Fibers with larger diameter is beneficial for reducing internal pressure drop, but they tend to be more rigid than smaller fibers. Rigidity of fiber makes the membrane more vulnerable to membrane fouling by suppressing the random fiber movement as discussed in here.
To reduce internal pressure drop without loosing fiber flexibility and strength, a new attempt has been made using more advanced membrane materials and fabrication methods (Suk, 2012). In this study, inner diameter was increased from 0.8 mm to 1.0 mm while keeping outer diameter same at 2.1 mm compared to GE’s hollow fiber membranes. In spite of the thinner fiber wall, fiber length could be maintained at 2 meters as same as benchmarked commercial membrane without hampering the physical strength of the fiber. In a pilot test, the new hollow fiber membranes with larger inner fiber diameter showed more stable trans-membrane pressure (TMP) than GE’s hollow fiber membranes as shown in Fig. 4.
Fig. 4. Comparison of two hollow fiber membranes with different inner diameter. Outer diameter and fiber length were identical for both hollow fibers. MLSS=10 g/L. (Suk, 2012)
© Seong Hoon Yoon