The commonly used SDI (Silt Density Index) and related indices are inadequate to accurately characterise fouling potential of the feed water. There is a lack of linkage between the fouling F indices of feed water and fouling development in fullscale RO processesRisks of possible failure due to membrane fouling have hindered further development of RO processes for water reclamation. Therefore, fullscale pilot tests, which are costly and time-consuming, have to be conducted for fouling characterisation.

An innovative fouling characterisation method was developed in this study. A new fouling index was proposed to provide a link between membrane fouling in a full-scale plant and the fouling potential of the feed water. A protocol has been developed for a quick and accurate estimate of the fouling index of feed water. The newly proposed fouling index of the feed water is independent of the operating conditions and refl ects solely the fouling property of the feed water. The actual fouling rate in a full-scale RO process can be determined with the fouling index for various operating conditions. One great advantage of the index is that it can be obtained within a few hours with a fi ltration experiment on a small membrane device (Figure 1). In contrast, the conventional way to determine fouling strength using a pilot study requires at least a few months. A schematic of the new and conventional characterisation methods is given in Figure 2.

The new method is expected to provide an accurate characterisation of feed water fouling tendency. A better characterisation of feed water fouling strength will signifi cantly reduce the failure rate of RO processes. This in turn would boost industry and public confi dence in RO processes for water reclamation and seawater desalination. Better designs of RO process and more suitable pretreatment of feed water can be achieved with the new fouling characterisation. This would result in substantial economic benefits. Further exploitation includes development of better and more cost effective pretreatment methods, development of characterisation instruments, and development of software for system design and optimisation of RO processes. This work was performed by Assoc Prof LF Song, Assoc Prof JY Hu, Prof SL Ong, Prof WJ Ng, and graduate student Tay Kwee Guan.

Figure 2: Schematic of the new and conventional characterisation methods.