Impacts of non-uniform filament feed spacers characteristics on the hydraulic and anti-fouling performances in the spacer-filled membrane channels: experiment and numerical simulation
2020
Abstract Feed spacer is universally used in spiral-wound nanofiltration (NF) and reverse osmosis (RO) membrane modules. It can separate membrane sheets, create flow channels, promote turbulence and enhance mass transfer. However, it also induces increased pressure drop across the flow channel, and generates dead zones for biofilm growth at specific locations. Optimization of feed spacer geometries is highly desirable for energy saving and biofouling control. In this study, four kinds of commercial feed spacers featured with non-uniform filaments were compared in terms of hydraulic and anti-fouling performances. Computational fluid dynamics (CFD) simulations were launched to give insights into the impacts of feed spacer characteristics on the flow field. Results show that the hydraulic performance was substantially affected by the number of filament layers (single or dual layer), the non-uniformity of filament diameter and the width of thinning zones. The design of single layer feed spacer of non-uniform filaments was not recommended due to the high flow resistance and poor anti-fouling performance. The feed spacer structure of alternating filament diameter contributed to reducing dead zones and alleviating membrane fouling. The thinning zones located adjacent to the filament junctions achieved better anti-fouling performance, as it disturbed the dead zones and partially washed away the deposited foulants. This study demonstrates for the first time that the characteristics of non-uniform filament feed spacer had a crucial impact on the hydraulic and anti-fouling performances, and suggests that more emphasis should be laid on number of filament layers, variation of filament diameter and width and positioning of thinning zones for the optimization of feed spacer geometries in the future.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
53
References
8
Citations
NaN
KQI