Seawater pre-treatment for reverse osmosis system

Publication Type:
Thesis
Issue Date:
2017
Full metadata record
Files in This Item:
Filename Description Size
01front.pdf310.43 kB
Adobe PDF
02whole.pdf1.82 MB
Adobe PDF
Membrane based desalination technology such as reverse osmosis (RO) has rapidly become a viable alternative to conventional treatment for drinking water production from seawater. However, membrane fouling is a major concern in reverse osmosis (RO) based seawater desalination. The fouling on RO membrane deteriorates the performance of RO membranes and increases the energy consumption and even requires more frequent replacement of the membranes. The objective of the study was to assess the different pre-treatment systems to reduce membrane fouling reduction, and remove organic matter in terms of dissolved organic carbon in RO desalination projects. Silt density index (SDI), modified fouling index (MF/UF-MFI) and cross-flow sampler modified fouling index (CFMF-MFI) were used to study the pre-treatment efficiency of different process such as flocculation, deep bed filtration, microfiltration, ultrafiltration and biofiltration. A long term on site biofilter experiment was investigated in terms of removal of particulate matter, different fouling indices and dissolved organic carbon (DOC) from sea water by the use of biofiltration. In this study, three biofilter columns were operated packed with granular activated carbon (GAC), anthracite and sand as a filter media. The experimental results indicated that biofiltration pre-treatment systems reduced organic matter and particulate matter. It was expected that biofilter can lower fouling to a subsequent RO process in desalination plant. In terms of DOC removal efficiency, GAC biofilter showed higher and stable removal efficiency (41-88%), than sand biofilter (7-76%) and anthracite biofilter (3-71%). All biofilters used in this study removed most of hydrophobic organic compounds (around 94%). On the other hand, hydrophilic organic removal varied depending on the media filter. GAC biofilter removed more organic bio-polymers (51%), humic substances (75%) and building blocks (50%) compared with sand and anthracite biofilters. Thus GAC filter was the best medium to provide the lowest fouling potential as it showed the highest removal efficiency of DOC, including hydrophilic, humic, building blocks and biopolymer. The fouling potential of treated seawater (filtrate) was evaluated using three different fouling MF-MFI, UF-MFI, and CFMF-MFI. GAC biofilter had lower fouling potential compared to sand and anthracite biofilters. The in-line flocculation and spiral-flocculation followed by media filtration (sand or anthracite) have been investigated as a pre-treatment of seawater to reverse osmosis (SWRO). In the case of in-line flocculation filtration system, the seawater was passed through the media filter just after rapid mixing of raw seawater with flocculants for 10 seconds. In the case of spiral-flocculation filtration, after the rapid mixing of seawater with flocculants, it was then passed through the spiral-flocculation. Both filtrations showed good turbidity removal efficiency (up to 71%). In-line flocculation filtration showed 2-3 times higher headloss than the spiral-flocculation filtration. The UF-MFI reduction was 63-70% for sand as medium in the presence of the flocculant whereas it was 65-76% for anthracite. Both filtration systems in the presence of flocculant (3 mg/L Fe³⁺) led to 50-65% removal of hydrophobic organics. The hydrophilic organic removal was around 30-38%. The predominant portion of hydrophilic was humic substances which had a poor removal. In general sand filter gave a higher removal than anthracite filter. The performance of TiCl₄ and Ti(SO₄)₂ was compared to FeCl₃ at different coagulant concentrations (1-30 mg/l) of Ti salts and FeCl₃ and at different pH of 5 to 9. Coagulation was conducted using conventional jar test. For each jar test, six 1 litre beakers were filled with raw seawater. The pH was adjusted with 0.1 N solution of hydrochloride acid and sodium hydroxide prior to coagulant addition. The solution was subjected to rapid mixing (100 rpm) for 2 min followed by slow mixing (20 rpm) for 30 min. It was then stopped to allow the aggregated flocs to settle down for 30 min. The supernatant samples were drawn for the measurements of turbidity, UV-254 absorbance and DOC, zeta potential and particle size distribution. The results showed that at pH of 8.0 (similar to seawater pH), TiCl₄ had advantages over FeCl₃ and Ti(SO₄)₂ at the same coagulant dose of 20 mg/L. Under this condition, TiCl₄ achieved ~70% DOC and UV-254 removal. This was approximately two times higher than FeCl₃ and Ti(SO₄)₂. Nevertheless, FeCl₃ and Ti(SO₄)₂ showed better turbidity removal. At higher coagulant dose (30 mg/L), the turbidity removal of TiCl₄, was especially compromised. The differences in the performance of the coagulants were associated with the coagulant mechanisms based on the floc zeta potential evaluation. The coagulant mechanisms of Ti-salts could be associated to charge neutralization while FeCl₃was inclined towards adsorption mechanism. The study found that biofiltration, in-line flocculation and spiral-flocculation followed by media filtration, coagulation and flocculation are appropriate pre-treatment before RO. In particular, Biofilter showed to a consistent removal of organic matter over a long period of time.
Please use this identifier to cite or link to this item: