Laboratory Efficacy of Locally Available Backwashing Methods at Removing Fouling in Hollow-Fiber Membrane Filters Used for Household Water Treatment

Abstract

Hollow-fiber membrane filters (HFMFs) for household water treatment (HWT) can efficaciously remove disease-causing organisms in laboratory settings. However, lower effectiveness in use in low- and middle-income countries (LMICs) and humanitarian contexts (HCs) has been observed and attributed to membrane fouling and the associated cleaning. In LMICs/HCs, it is not possible to prevent and control fouling using commonly known methods (e.g., testing influent water, maintenance regimes), and the literature on fouling/cleaning of HFMFs distributed in LMICs is scarce. As such, controlled laboratory experiments were conducted to determine the efficacy of locally available (in LMICs/HCs) backwashing solutions at removing fouling using different influent waters and HFMF types. Four commonly distributed HFMFs were selected; fouling layers were developed by filtering three influent water compositions, representing LMIC/HC waters, for 10-days, and bleach, water, or vinegar backwashing solutions were used for daily backwashing. Filter performance indicators included: fiber mechanical properties (strain at break, break force), water quantity performance (flow), water quality performance (turbidity, E. coli), and imaging. The study found fouling developed rapidly and altered mechanical properties and water quantity indicators within 200 h of filtration. Fouling did not decrease water quality indicators. Backwashing improved the filter's mechanical properties and water quantity performance, but it did not fully recover the initial performance. Additionally, recovery differed between backwashing solutions, and no universal cleaning recommendation appropriate for HFMFs in LMICs/HCs was identified. Overall, fouling development and control depended on HFMF type, influent water quality, and backwashing solution type; thus, caution before distributing HFMFs for long-term use in LMICs/HCs is recommended.

Keywords: backwashing; fouling; household water treatment; humanitarian response; membrane filters.

Figure 1

Experimental design. The filtration/backwash cycle was repeated for nine consecutive days. On the tenth day, filtration was performed, but only one module of each pair was backwashed after filtration. The other module from each pair was used as no-backwash control to assess the impact of the final study-end backwash. Finally, all modules were run with demineralized water and proceeded to postmortem analysis.

Figure 2

SEM images of fouling development on F2 for control modules (nonbackwashed) by influent water. Outer membrane surface at 500× and 15,000× (insets) for control: (a) clean fiber, (b) Bacteria, (c) BacChem, and (d) BacSed. Cross-section showing fouling layer thickness on the outer surface (2500× and 10,000× (insets)) for (e) clean fiber, (f) Bacteria, (g) BacChem, and (h) BacSed.

Figure 3

Normalized flow, strain at break, and break force for MF membranes (F2) and UF membranes (F4) by influent water (Bacteria, BacChem, and BacSed), backwashing solutions (colors: controls, water, bleach, and vinegar), and backwashing completion (shade-in, without and with a last backwash). Results for F1 and F3 are similar and presented in the supporting information.

Figure 4

SEM images of fouling control on F2 for modules filtered with BacChem influent by backwashing solutions. Outer membrane surface at 500× and 15,000× (insets), showing fouling layer after BacChem filtration with final backwash using (i) DI water, (j) bleach, and (k) vinegar and without final backwash (l) DI water, (m) bleach, and (n) vinegar.

Figure 5

Relative change of indicators from baseline to endline (%, semiquantitative scale) for (a) control modules (fouling consequences) and backwashed modules (backwashing consequences) and (b) backwashed modules by backwashing solutions (backwashing solution consequences).