Subsurface intake systems are known to improve the feed water quality for SWRO plants. However, a little is known about the feasibility of implementation in coastal settings, the degree of water quality improvements provided by these systems, and the internal mechanisms of potential fouling compounds removal within subsurface intake systems. A new method was developed to assess the feasibility of using different subsurface intake systems in coastal areas and was applied to Red Sea coastline of Saudi Arabia. The methodology demonstrated that five specific coastal environments could support well intake systems use for small-capacity SWRO plants, whereas large-capacity SWRO facilities could use seabed gallery intake systems. It was also found that seabed intake system could run with no operational constraints based on the high evaporation rates and associated diurnal salinity changes along the coast line. Performance of well intake systems in several SWRO facilities along the Red Sea coast showed that the concentrations of organic compounds were reduced in the feed water, similar or better than traditional pretreatment methodologies. Nearly all algae, up to 99% of bacteria, between 84 and 100% of the biopolymer fraction of NOM, and a high percentage of TEP were removed during transport through the aquifer. These organics cause membrane biofouling and using well intakes showed a 50-75% lower need to clean the SWRO membranes compared to conventional open-ocean intakes. An assessment of the effectiveness of seabed gallery intake systems was conducted through a long-term bench-scale column experiment. The simulation of the active layer (upper 1 m) showed that it is highly effective at producing feed water quality improvements and acts totally different compared to slow sand filtration systems treating freshwater. No development of a “schmutzdecke” layer occurred and treatment was not limited to the top 10 cm, but throughout the full column thickness. Algae and bacteria were removed in a manner similar to slow sand filtration, but it took many months to produce consistent reductions in NOM fractions and TEP. The data suggested that a thicker active layer (2m) is needed to facilitate a more rapid reduction in the main potential fouling organics.
|Date made available
|KAUST Research Repository