As summarized in Table 1, a number of factors are involved in hollow fiber module optimization. The three major goals in hollow fiber optimization are
- Maximization of random fiber movement in two phase flow
- Minimization of internal pressure drop
- Mechanically sustainable module configuration in two phase flow
The challenge is those three goals are conflicting each other in many ways. For example, random fiber movement can be maximized by reducing fiber diameter and increasing fiber length, but these directly causes higher internal pressure drop and fragile fibers. By increasing fiber diameter, internal pressure drop can be minimized and module becomes mechanically more sustainable, but fiber flexibility would decreased with a larger fiber diameter. Therefore, the three major goals must be compromise to obtain the optimum design.
Depending on materials, structure, and the manufacturing process of hollow fiber, optimum combination of inner diameter, outer diameter, and length varies. In one instance, if membrane material has high tensile strength, hollow fiber can be made relatively small to maximize the random fiber movement in the turbulence of air-liquid flow. Although the small fibers suffer more from internal pressure drop and subsequent membrane fouling, enhanced random fiber movement can compensate the drawback at least partially. Small fibers also allow high packing densities, which in turn allows scouring air savings although fiber clogging is another factor to consider. In contrast, if membrane’s tensile strength is low, hollow fibers with larger diameter are more plausible. Although the thick fibers do not move as much as its smaller counterpart, reduced internal pressure drop can help reduce the cake layer formation near permeate exit.
The tensile strength of membrane can be greatly enhanced by using hollow braids as supports, where the skin layer can be either chemically integrated or physically laminated on hollow braid. The enhanced tensile strength allows larger and longer fibers, where the internal pressure drop can be controlled low. Though fiber flexibility and lumen diameter must be compromised somewhat due to the large wall thicknesses, the longer fiber compensates the loss of fiber flexibility at least partly by allowing larger amplitude of movement.
Table 1. Summary of the important physical properties of hollow fiber membrane and their effect on membrane performance
|Large inner diameter||· Reduced internal pressure drop|
allows long fibers
|· Compromised packing density|
|Thin membrane wall||· Increased fiber flexibility|
· Increased random movement
· Increased packing density
|· Increased fiber breakage|
|Extended fiber length||· Reduced specific aeration demand|
· Reduced footprint
· Increased random fiber movement
· Increased scouring effect
· Reduced specific fabrication costs (more membrane area per potting)
|· Increased fiber breakage|
· Increased internal pressure drop that induces membrane fouling
· Reduced handling property
|High flexibility||· Increased random fiber movement|
· Reduced membrane fouling
|· Fiber abrasions from random collision among fibers|
|Dual layer structure|| High tensile strength allows long fibers|
Reduced fiber breakage
| Compromised random fiber movement|
Compromised lumen diameter
Interestingly there are two commercial hollow fiber membranes whose fiber dimensions suggest opposite design concepts. As summarized in Table 2, GE’s ZW500d® module has larger ID and shorter effective length than Asahi’s Microza® module (0.9 mm vs 0.7 mm for ID, 1,000 mm vs 2,000 mm for length). Due to the smaller ID and the longer fibers, internal pressure drop is much higher for Microza® than that for ZW500d®.
If only the internal pressure drop is considered, sustainable flux of Microza® should be much lower than ZW500d® since the unbalanced flux along the fiber can exacerbate the membrane fouling as discussed here. However, the average operating flux of Microza® membranes is reported as high as those of ZW500d® in typical municipal wastewater treatment.
This paradox can be attributed to the differences in the sizes of the membrane fibers and the resulting fiber amplitudes. The fiber ID/OD reported in literature for the two membranes are 0.9/1.9 mm and 0.7/1.2 mm, respectively. Therefore, the cross-sectional area of Microza® based on OD is less than a half of the ZW500d®’s. In addition, the membrane wall thickness is only a half (0.025 mm vs. 0.05 mm). It might be possible that the enhanced fiber amplitude/movement due to the small fiber size fully compensates the disadvantages from internal pressure drop.
Table 2. Comparison of two commercial hollow fiber membranes
(membrane + braids)
|Material||PVDF / Polyester||PVDF|
|Pore size (um)||0.04||0.1|
|Height (mm)||2,000||2,000||Membranes only|
|1,000||2,000||Two sides suction for GE membrane|
|Module dimension (mm)||5||150mm diameter|
x 2,000 mm longth
|Area per module (m2)||32||25|
- High tensile strength that allows thin wall and long fiber length simultaneously
- Thin wall in turn allows larger ID that reduces drawbacks from internal pressure loss
- Thin wall also allows better fiber movement in water, which is crucial for anti-membrane fouling
© Seong Hoon Yoon