Membrane cleaning flux recovery

Based on measurements of water flux data, the polysulfone hollow fibers proved difficult to clean. Flux recovery from run to run was inconsistent, and the membrane occasionally required extra effort to clean. 

Continued decline in performance indicates a membrane cleaning or compatibihty issue. The adequacy of the cleaning step is determined by the recovery of at least 80 percent of the initial normalized water flux. Although some variability in water flux is typical, any consistent dechne reflects an inadequate cleaning procedure. 

Bird MR and Bartlett M. Measuring and modelling flux recovery during the chemical cleaning of MF membranes for the processing of whey protein concentrate. J. Food Eng. 2002 53 143-152. 

FIGURE 22.10 Dynamics profile of the flux recovery in cleaning WPC-fouled PVDF MF membranes using NaOH solution at pFI 13 and 30°C, and at crossflow velocity of 0.45 m s and 350 kPa TMP. (From Mercade-Prieto, R. and Chen, X.D., J. Membr. Sci., 254, 157, 2(X)5. With permission.)... 

MF and UF used in pretreatment to RO are more frequently cleaned by physical cleaning and less frequently by chemical cleaning. Cleaning frequencies reported in ht-erahire varied widely. Physical cleaning frequency is approximately every 40 min with a chemical clean scheduled every 6 mo (98). An air backwash frequency of 15-20 min is sufficient for hollow-fiber MF membranes (77). In a UF evaluation study, backwashing was able to achieve an average flux recovery of 86.5% (99). It was observed in the same study that flux restoration could be achieved even when backwash was reduced from 10 to 1 min. 

Particular attention was devoted to the control of membrane fouling and membrane cleaning. Acid-alkaline washing was tested and low concentration chlorine solutions were also used. The recovery of initial fluxes was generally 50% with the new modulus, and higher then 95% with a used modulus. These results indicate the existence of a certain irreversible fouling of the new membranes, which come to steady state values, and does not increase with membrane reuse. 

Hollow fiber membrane modules have minimum hold-up volumes and can operate without blockage of the fibers if the l.d. of the fiber is substantially larger than the maximum cell aggregate. Until recently, hollow fiber modules were only available with UF membranes. The advent of polysulfone MF hollow fibers makes possible cleaning with acid and base and even autoclaving at 121° in some instances. Since MF membranes are prone to internal pore fouling, membrane cleanability with flux recovery is particularly important. 

Al-Obeidani et al. [93] reported on the development of more effective and economical procedures for cleaning polyethylene hollow fiber MF membranes that have been used for removing oil from contaminated seawater. In their study, alkaline cleaning showed higher recovery of operating cycle time but lower permeate flux recovery than acid cleaning. 

Figure 2.41 Membrane flux recovery with successive cleaning steps. Source Schafer et al.

UF of BSA solutions BSA adsorption on the membranes decreased from 159 2 to 13 2 jg/cm after 10 min of grafting, the fouling degree of the blend membrane with 7 min grafting was reduced by 66%, and the flux recovery after chemical cleaning increased by about 32% (Zhang et al. 2009)... 

Treatment with N2 plasma introduces various chemical groups, such as amine, imine, amide, or nitrile, on the membrane surface, making it more hydrophilic and less liable to fouling (Kull et al. 2005). For example, after continuous operation in an SMBR for about 90 h, flux recoveries for plasma-treated PP membranes for 8 min were 62.9% and 67.8% higher than those for the virgin membrane after water and NaOH cleaning (Yu et al. 2007). 

The hydrophilicity of the NH3 plasma-treated PP membranes increased with the increase in plasma treatment time and decreased with the increase in storage time (Yan et al. 2008). The adsorption of BSA on the modified membranes was lower than that on the unmodified-membrane surface and the flux recoveries after water and caustic cleaning for the NH3 plasma-treated membranes for 1 min were 51.1% and 60.7% higher than those for the unmodified membrane. However, the mechanical properties of the membranes decreased after prolonged plasma treatment, thus the optimal plasma treatment time for membrane modification was taken as 1 min. 

---