Water. desalination + reuse

| 28 | Desalination & Water Reuse | May-June 2014 TECHNOLOGY _________ Jan C Schippers, Sergio G Salinas-Rodriguez and Maria D Kennedy (TU Delft), UNESCO-IHE, and Siobhan Boerlage, Boerlage Consulting ___ Editor's Note: The fouling potential for reverse-osmosis and other membrane systems has been a hot topic for many years. D&WR has published several articles comparing the Silt Density Index with the newer Modified Fouling Index. As it is becoming apparent that the latter system of measurement is about to become the preferred method, this article looks at why this is happening. PARTICULATE FOULING has plagued reverse osmosis (RO) systems since their first use in desalination and remains a persistent issue today for RO and other pressure-driven systems such as microfiltration (MF), ultrafiltration (UF) and nanofiltration (NF). In the early 1960s, the Du Pont Company/Permasep Product successfully launched the hollow-fine-fibre (HFF) permeator onto the desalination market, where it dominated for several decades. A well-known weakness of this permeator was its vulnerability to fouling. Initially, this vulnerability was attributed to suspended and colloidal matter in the feed water, ie, particulate fouling. Therefore, Du Pont developed, the Silt Density Index (SDI), initially named the Fouling Index, as a parameter to characterize the fouling potential of the feedwater for permeators. The fouling mechanism turned out to be more complicated than just fouling of the membrane surface as initially assumed. Gradually it became clear that the fouling was initiated by local clogging of the woven or non-woven fabric between the fibres, which is needed to ensure equal flow distribution of the feed water (see Figure 1). This primary fouling mechanism disturbs the flow pattern resulting in localised low flow, causing high concentration polarization and higher recovery rates in that area. This then leads to higher osmotic pressure, deposition of suspended and colloidal particles and scaling, for example, of calcium sulphate, reducing the permeate flow. In the 1980s, it became clear that biofouling also frequently occurred, resulting in the same phenomena and exacerbated fouling. The synergistic effects of these fouling types made the fouling problem even more complicated. In the 1990s, spiral-wound elements were gaining ground in the market, claiming to be less vulnerable to fouling, which was reflected in their less stringent SDI guidelines, ie, a maximum SDI of 5 was allowed in membrane guarantees, with an SDI of <3 preferred. SDI guidelines for Permasep permeators were SDI <3 and preferably SDI of <1. Spiral-wound elements were indeed less vulnerable to clogging than the HFF permeators, attributable to differences in design and wide spacing between spacer and membrane surface. The same holds for the HFF element used today, having cross-wound fibres with wide spacing between these fibres. While the SDI is a useful tool in characterizing the particulate fouling potential of RO feed water, when it comes to clogging of fabric in permeators, spacers in spiral-wound elements and the new type of HFF elements, it may not account for the direct fouling of the membrane surface itself, which results in a permeability decline. This raises the question: Is the SDI a useful tool in predicting this type of fouling as well? This paper examines this question and traces the development of the SDI and the more recent Modified Fouling Index (MFI). Silt DenSity inDex The SDI, standardized by ASTM 1 , is based on filtration of feed water through a 0.45 µm membrane in dead- end mode at constant pressure (207 kPa). The rate of Why MFI is edging SDI as a fouling index Figure 1: Fabric in hollow fine fibre permeator of Du Pont Permasep

• Cover