Water Journal : Water Journal May 2011
refereed paper membranes & desalination water MAY 2011 107 Filtered water was directed back to the primary RO skids, but the secondary skids remained bypassed. Primary reject was diverted to sewer. Chloramine dosing was started at 2ppm, and was gradually increased to determine the optimum dose. A measurable impact was not achieved until 3ppm--4ppm, with further improvement up to 5ppm--6ppm. Doses above 6ppm did not yield further improvement. Figure 4 illustrates the effect of increasing chloramine dose on the total pressure drop and normalised permeate flow of the primary RO skid operating with the Toray membranes. Two primary RO skids were operated with 5ppm--6ppm chloramine in the RO feed. Run times were still limited to two to four weeks. Operation of the secondary skids resumed, but run times were only four to seven days. A run was terminated and the skid cleaned when the variable speed high pressure feed pump reached full speed, causing permeate flow to decline by 10%. This represented a decline in normalised permeate flow of 30%, well in excess of the 10--15% accepted decline to initiate cleaning. Membrane Fouling Inspection of the primary RO vessels revealed sulphur deposition on the end caps (Figure 5). The sulphur originated from sulphate in the feed water, which was being reduced to sulphide in the GAC contactors, as dissolved oxygen was consumed by the biological activity. Subsequent dosing with chloramine oxidised the sulphide back to sulphur, and the colloidal particles passed through the media filters and 40 micron cartridge filters. Inspection of the secondary RO vessels revealed severe membrane fouling (Figure 6) comprised of biomass and iron sulphide. The iron arose from incomplete iron removal in upstream processes and sulphide from further biological reduction of sulphate. The high level of chloramine required to manage biofouling was not rejected by the membranes. Chloramine levels in the RO permeate levels were similar to that in the feed and reject. This required significant dechlorination to enable disposal of excess water to Botany Bay, and together with the elevated levels of total organic carbon (TOC) passing through the membrane, prevented the use of the water in demineraliser feed applications. Biological Pre-Treatment Based on the opportunistic biological activity occurring in the GAC contactors, the first stage (lead units) of contactors were retrofitted with aeration to actively promote biological uptake of the RBOC. The GAC media was replaced with zeolite, to allow vigorous backwashing of the media. Although converting only the lead units would provide only half the expected required capacity, it was unclear whether non-volatile CHCs, which were being degraded in the GAC contactors under anoxic conditions, would be removed under aerobic conditions. The RBOC fraction of TOC was suspected as the major contributor to downstream biofouling. This fraction would be taken up first, leaving the less biodegradable fractions which would not support biofouling to the same extent. Figure 7 shows modifications to the GTP process, including chloramine dosing and reconfiguration of the dual media filters. The GAC contactors were converted to biologically aerated filters (BAFs) one at a time. As each filter was commissioned and ripened, TOC levels in the feedstream to the RO skids were gradually reduced. Figure 8 shows the falling TOC concentrations over the duration of BAF conversion, and the corresponding reduction in chloramine dose required to maintain operation. The BAFs performed in line with expectations, removing approximately 50% of the influent TOC. The GAC contactors removed a further 20% TOC, and continued to remove the non-volatile CHCs to the required levels for discharge. However, recovery of the backwash waste had to be abandoned as the process could not effectively remove the waste biological sludge. Figure 4: Pressure drop and normalised permeate flow with the introduction of chloramine dosing. Figure 5: Sulphur coating on RO end caps. Figure 6: Fouling in secondary RO elements. Figure 7: Botany GTP -- modified treatment process. Figure 8: TOC reduction with BAF conversion.
Water Journal April 2011
Water Journal July 2011