Water Journal : Water Journal April 2011
refereed paper membranes & desalination water APRIL 2011 103 JAA Plant During commissioning it was found that the feed pressure and membrane differential pressure increased and salt rejection decreased. The caustic flush was set up to three-hourly to mitigate iron fouling. However, the fouling developed to a level over time that the caustic flushing and frequent (weekly) CIPs were unable to keep operations sustainable. It was proven that the water quality was out of specifications. The percentage ratio of ferrous to ferric iron was lower (<50% vs >85%), the pH was lower <4.5 and the conductivity was >60,000 μS/cm. This water had the highest iron and TDS and the lowest pH that Osmoflo has treated. It was determined that ferrous had been oxidised through the pipeline delivery system, including through the turkey nest dam and raw water tanks which are exposed to air. It was impractical from the client's operations for the feed water to be kept anoxic to minimise ferrous oxidation, as recommended by the temporary trial plant and designed for in the permanent plant. The membranes were visually checked and found to be covered with a bright orange/brown colour foulant, as shown in Figure 7. An autopsy was performed on a lead and tail membrane and concluded that iron oxide was the main foulant and that the lead membrane had a higher concentration of iron than the tail membrane. The Fujiwara test indicated oxidation damage had not occurred, therefore, the lower salt rejection was likely caused by the iron fouling layer increasing the concentration polarisation. The severe fouling meant it was more practical to replace the 12 membranes than to clean them, given continued water production was important. The fouled membranes underwent cleaning tests in Osmoflo's workshop, but it was proven the performance recovery was not significant so they were discarded. When the first-pass RO salt rejection decreased there was a cascading effect on the second-pass RO, where the membranes fouled/scaled rapidly as this pass operated at high recoveries with concentrate recirculation and there was no antiscalant dosing. These membranes were able to be mostly recovered via CIPs. Osmoflo worked with Iluka to arrive at an alternative pre-treatment solution. The appropriate process identified included pH adjustment to 7.5 to precipitate the iron, settling in a 1.8ML open dam, sodium hypochlorite dosing to assist with oxidation, iron removal filtration (IRF) with DMI-65 media and dechlorination with SMBS. New dosing systems with instrumentation were installed for caustic, sodium hypochlorite (including for IRF backwashing) and SMBS dosing. The media inside all the MMFs were removed and replaced with DMI-65. The modified pre-treatment system was commissioned in February 2010. There was a stable operating period of approximately one month, and then the instrumentation including feed water pH, feed water free chlorine and RO feed ORP began to foul due to lack of maintenance. The feed caustic and sodium hypochlorite dosing were not functioning correctly and caused the first-pass RO to foul, resulting in higher differential pressures, lower flux and lower salt rejection. However, the performance was able to be recovered through CIPs. Once again, there was also a cascading effect where the second-pass RO membranes flux rapidly decreased. Given the second-pass RO had residual fouling from when the first-pass RO membranes severely fouled with iron, it was practical to replace these seven membranes. Iluka recognised the challenges associated with the feed water and considered a permanent contract to support the RO plant. A two-year operations (remote control and monitoring) and maintenance (monthly and optional emergency visits) contract was finalised in June 2010 with Osmoflo. Conclusion Osmoflo has experienced a number of feed water challenges for the two RO plants treating hypersaline groundwater. The challenges included deviations from the contract feed water quality: • High conductivity of up to 70,000 μS/cm; • Colloidal iron from ferrous oxidation; • Very high total iron of up to 60mg/L; • High ferrous iron of up to 40mg/L; • A rare case of iodine fouling; • pH < 5 but adequate ORP for ferrous oxidation. There was a lot of effort invested into analytical investigation to find the root causes of the water quality deviations. Once the root causes were identified the pre-treatment processes were modified in co-operation with the clients to manage the deviations and minimise downtime associated with rapid fouling and scaling of the filter cartridges and RO membranes. The experience reinforced the importance of regularly monitoring feed water quality and management of any deviations promptly via process changes to ensure service delivery to clients. The Authors Hiep Le, a Chemical Engineer, is a Senior Process Engineer and is involved with membrane applications for some of the most difficult waters to treat in the world. Tim Pritchard is a highly experienced automation/electrical specialist with particular skills in the commissioning of large water treatment projects. Neil Palmer was, at the time, General Manager Technical Services and has since become the Chief Executive Officer of the National Centre of Excellence in Desalination. Osmoflo is a company which specialises in providing desalination systems for any type of water in any location. Those which operate in remote regions are remotely monitored and controlled from the head office in Adelaide. References National Groundwater Commission, 2008, Groundwater Position Statement, Plazinska A, 2007: Science for Decision Makers -- Understanding Groundwater, Australia Government Bureau of Rural Sciences, Sydney, NSW, Australia. Figure 7: Severe iron fouling.
Water Journal March 2011
Water Journal May 2011