Water. desalination + reuse

TECHNOLOGY The regime designed to address the constraints was to operate the RO plant twice weekly, supplying 25,000 m3 on each occasion to turn over the network pipeline, and producing enough permeate to flush all RO membranes in the plant. Risk assessments were carried out prior to trials in late 2010 and included separate exercises focussed on plant based risks and water supply based risks involving a wider stakeholder base. Trials commenced in December 2010, but were interrupted by the critical water supply requirement associated with the January 2011 Brisbane flood, which disrupted other water sources. This event meant that further trials could not be conducted until April 2011 (Figure 3). The trials consisted of ten cycles (five weeks) comprising start-up, run, flush and shutdown. The trials were successful in that the regime was proven feasible and was able to meet start-time objectives without creating unmanageable or unacceptable risks. Naturally this conclusion was on the basis of the limited duration of the trials. After a brief period of continuous operation to supplement supply during maintenance of other surface-water treatment plants in mid-2011, Hot Standby was implemented and the regime has been in place ever since. DisCussiON AND REsuLT ANALYsis Energy and Chemical Consumption Power consumption per unit water production increases significantly under Hot Standby, given the increased ratio Figure 3: Hot Standby trial and investigation period of intake and pretreatment operating-time relative to RO operating-time. An indicative figure for specific energy consumption during Hot Standby has been calculated as 6.9 kWhm3 assuming pretreatment plant operation for approximately 24 h prior to RO plant operation while in continuous operation mode the figure has been as low as 3.2 kWh/m3. This is particularly the case if intake seawater flows are maintained in between Hot Standby run days and if a conservative approach is taken in stabilising pretreatment silt density indices (SDI15) prior to running the RO plant. Similarly chemical consumption costs are reduced under Hot Standby but are subject to the same semi-fixed characteristic as energy depending on pretreatment plant operation. Water Quality and Membrane Performance Customer complaint frequencies in the Figure 2: SEQ water supply system schematic illustrating connectivity of Gold Coast Desalination Plant to Brisbane City and beyond. distribution system were monitored and found to be unaffected during Hot Standby trials. Equally water quality in the network pipeline has been stable in regard to disinfection residual and pH. Recent experience has suggested that production volumes are conservative to such an extent that optimisation is being investigated. An issue that became apparent during the April trials was the increased average normalised salt passage due to the 'bed-in' duration of membranes being longer than the Hot Standby run time. Figure 4 (on page 35) illustrates this phenomenon. The figure shows that because each run in Hot Standby never progresses to the period where salt passage stabilises, the average salt passage will be elevated in Hot Standby. On this basis it is expected that the product water quality limits, which govern membrane replacement, may be reached earlier in Hot Standby mode. Despite the 'bed-in' issue described above, membrane performance has generally surpassed expectations in terms of permeate water quality and membrane life. August-September 2013 | Desalination & Water Reuse | 33 |

• Cover