Water. desalination + reuse

May/June 2014

Water. Desalination + reuse

Issue link: https://fhpublishing.uberflip.com/i/322350

Contents of this Issue

Navigation

Page 30 of 47

TECHNOLOGY May-June 2014 | Desalination & Water Reuse | 29 | plugging is measured and expressed as % flux decline per minute. As the SDI is simple to perform and cheap, it has been universally applied for the last 50 years as a tool to assess the particulate fouling tendency of a feedwater, the effectiveness of pretreatment processes etc, and is often the basis of membrane guarantees and other plant performance contracts. However, increasingly, the value of this test to predict the rate of fouling in RO systems due to particle deposition is being questioned. The limitations of the SDI test are well documented and include 2-6 : • No correction for test water temperature • Result heavily dependent on the test membrane permeability • Not applicable for testing high- fouling feedwater, eg, raw water – ASTM recommends that turbidity should be <1 NTU • Not applicable for testing UF permeate, which is increasingly being used in desalination pretreatment • No linear relation with colloidal/ suspended matter • Fouling potential of particles smaller than 0.45 µm not measured • Not based on any filtration mechanism. The absence of temperature correction results in higher SDI values at higher temperatures. Additionally, the non-linear relation between the measured SDI value with particle concentration means that water appears less fouling than it is, as the test filter becomes progressively plugged. This is demonstrated in Figure 2 for SDI measurements after 5, 10, and 15 minutes filtration for a formazine solution. The net result is that SDI cannot be directly compared when measured at different temperatures or for different filtration intervals and for more fouling conditions. Theoretical prediction of flux decline in RO systems based on SDI results in extremely high fouling rates, eg, SDI = 3, effectively means a flux decline of 3% per minute 7 . Applying a direct correction between the SDI test flux (>1,600 L/m 2 h at the start) to a typical RO flux (about 20 L/m 2 h), predicts a flux decline of 20% per hour. This rate of fouling is far outside the rates observed in practice. To overcome these deficiencies, the Modified Fouling Index (MFI) using the same 0.45 µm membrane filters (MFI-0.45) and MFI using ultrafiltration membranes and performed at constant flux (MFI-UF) were developed. MOdifiEd fOuLiNG iNdEx (Mfi-0.45) The MFI-0.45 test uses the same equipment as the SDI test 2 . It takes into account that, initially, pore blocking occurs, followed by cake/gel filtration and, finally, cake/gel blocking and/or enhanced compression. The MFI value is determined from the stage of cake/gel filtration and is defined as the minimum slope (tg α) in the curve t/V versus V. Where t = total filtration time, and V = total filtered volume (see Figure 3). Figure 2: SDI as func on of formazine concentra on and filtra on me 2 Equa on 1 Figure 3: Filtra on curve t/V versus V 8 A: Blocking filtra on B: Cake filtra on C: Cake filtra on with clogging and/or cake compression Figure 4: Rela on between MFI and formazine concentra on 2 Formazine mg/L t/V (s/L) Filtrate volume (L) 200 150 100 0 0 1 2 3 4 5 6 SDI Formazine (mg/L) 15 10 5 0 0 1 2 3 4 5 6 0 1 2 3 4 5 6 7 100 75 50 25 0 MFI (s/L 2 ) SDI 5 SDI 10 SDI 15

Articles in this issue

Archives of this issue

view archives of Water. desalination + reuse - May/June 2014