Water Journal : Water Journal April 2011
refereed paper technical features 92 APRIL 2011 water membranes & desalination of bacteriological research (Couto and Hogg, 1999, Ericsson et al., 2000 and Auty et al., 2001), but has been used specifically for the direct enumeration of physiologically active bacteria from drinking water (Boulos et al., 1999; Hoefel et al., 2003) and bottled mineral water (Ramalho et al., 2001). Limitations to this technique include inaccuracies in comparison with culture techniques with no consistent proportionality comparing concentrations of active and culturable bacteria (Hoefel et al., 2003); however, the rapid nature of FCM has the potential to replace time- consuming culture-based techniques for the bacteriological assessment of water, especially reporting active counts of nuisance microorganisms that are difficult to culture. Such rapid assessment would allow appropriate management practices to be implemented sooner, resulting in operational cost savings (Hoefel et al., 2003). The application of FCM to RO membrane biofouling assessment may be undertaken while the membrane elements remain in service by analysing the bacterial content of the feed, concentrate and permeate streams. To gain further information about the nature of the bacteria attached to the RO membrane surface, DNA profiling (via 16S RNA methods) can be employed. When preparing a 16S RNA analysis, each cultured sample undergoes a polymerase chain reaction (PCR) which is applied using universal 16S RNA gene primers. Clone libraries of 16S RNA analysis provide information on the different species of isolated bacteria. This technique is somewhat novel for analysis of RO biofouling; however, one study at the Carlsbad Desalination Pilot Plant found that Proteobacteria dominated the biofilm (61.2%), which included Roseobacter, Ruegeria, Rhodobacteracea and Donghicola. Other species included Sphingobacteria, Flavobacteria, Planctomycetacia and Chloroflexi groups (Zhang et al., in press). This study also ascertained that the bacteria found on the RO membrane closely matched that in the raw sea water. Thus the 16S RNA method may also be a useful tool for identifying potential problem bacteria in the feedwaters of RO plants. In the current study, FCM was used in conjunction with the fluorescent dyes from the LIVE/DEAD® BacLightTM bacterial viability kit for detection of live and dead cells at the Adelaide Desalination Pilot Plant (ADPP). In addition, 16S RNA profiling was used to identify the bacterial species on the RO membrane surface. Experimental Process description In February 2008, the SA Government approved $9.5 million for the design, construction, operation and maintenance of the Adelaide Desalination Pilot Plant (ADPP). The plant would facilitate testing of various technologies, gather data on raw and process water quality and generate samples for environmental testing. Water Infrastructure Group (WIG) won the contract and was granted full site possession in June 2008. All systems, except the RO and post-treatment, were operational by August 2008. The plant was fully operational in December 2008. In April 2009, AdelaideAqua took over the operation of the plant. The plant was modified to better replicate the UF pre-treatment and hybrid design of the First Pass RO unit. In April 2010, a disk filtration system (ArkalTM) was installed upstream of the UF systems. For the purpose of this study, information was collected from 2009 onwards. Figure 1 shows a photograph of the ADPP showing the RO, UF and chemical dosing containers. The ADPP is located 1km south of the full-scale plant. The ADPP intake pipe is 1,620m long with seawater abstracted from a depth of 15m. It is situated within the intake corridor of the full-scale plant. The intake is capable of delivering 570kL/ day at a maximum inlet velocity of 0.15 ms-1. The intake structure incorporates a ½ inch 90/10 copper/nickel screen to minimise biogrowth. Figure 2 shows a process flow diagram of the plant and a description of the treatment process follows. The 100 μm disk filtration system was used for initial screening of the raw seawater prior to entering the raw water tank. The plant was able to simultaneously operate two pre-treatment systems, a Norit SeaguardTM pressurised UF system and a Siemens-Memcor continuous submerged UF system (4 x S10V modules). Filtered seawater from each UF unit was directed to a feed tank. Cartridge filters supplying the RO system were employed as security filters in the event of turbidity breakthrough from either UF unit. The RO system is capable of producing 100kL/day of permeate through a partial two-pass system. The First Pass (RO1) used 8 x 8" elements and was designed to be operated at 50% recovery, with 50% of the First Pass permeate polished through a two-stage Second Pass unit operating at 80-90% recovery. The First Pass incorporated an ERI PX-30S energy recovery device and used an internally staged design to more effectively balance flux across the elements and reduce energy consumption. The internally staged design of the First Pass employed Hydranautics 'SWC5 Max' and 'SWC6' elements. The permeate tube was blocked between elements 3 and 4, which is a process patented by Acciona Agua and referred to as a 'blind split'. Permeate from the 'SWC5 Max' elements (3 of) is referred to as 'RO1 Front Permeate' while permeate from the SWC6 elements (5 of) is referred to as 'RO1 rear permeate'. For the purpose of this study, only the performance of the First Pass was assessed. The final permeate stream was subject to post-remineralisation. Carbon dioxide and hydrated lime were added to the permeate to increase alkalinity, adjust the pH and, simultaneously, its corrosivity. The target alkalinity range was 60-70mg/L as CaCO3. A dechlorination system employing sodium metabisulphite (SMBS) solution was used to neutralise acidified and chlorinated waste streams generated from Figure 2: Process flow diagram of the Adelaide Desalination Pilot Plant.
Water Journal March 2011
Water Journal May 2011