Towards indirect potable reuse in South East Queensland

Transcription

1 153 Q IWA Publishing 2008 Water Science & Technology WST Towards indirect potable reuse in South East Queensland W. H. Traves, E. A. Gardner, B. Dennien and D. Spiller ABSTRACT Faced with limited water supply options in the longer term and the worst drought on record in the short term, the Queensland Government is constructing the Western Corridor Recycled Water Project which will supply up to 182 ML/day of purified recycled water for industrial and potable purposes. The project is one of a suite of capital works projects in progress which in the longer term will supply up to 10% of the region s potable water supply. Key words advanced water treatment, indirect potable reuse, water recycling W. H. Traves GHD Pty Ltd, 201 Charlotte Street, Brisbane, Qld, Australia warren.traves@ghd.com.au E. A. Gardner Queensland Department of Natural Resources & Water, 80 Meiers Road, Indooroopilly, Qld, Australia ted.gardner@nrm.qld.gov.au B. Dennien D. Spiller Queensland Water Commission, 80 George Street, Brisbane, Qld, Australia INTRODUCTION South East Queensland (SEQ), Australia, is experiencing significant population growth, with a projected increase of 1.2 million residents to a total of 4.0 million by 2026 (PIFU 2007). Combined with industry growth, this is leading to significant increases in demand for water. Assessments of yield from existing major sources are also declining as service standards are reconsidered and the impacts of the current drought and potential for long-term climate change are taken into account. As a result, there is a shortfall in overall supply availability in the next ten years as well as a more acute short-term requirement to address security of supply in current drought conditions the worst in over 100 years of records. The Queensland Government has responded with a major capital works program to secure the future water supply for the region. This includes the 125 ML/day Tugun Desalination Plant on the Gold Coastand the 232 ML/day Western Corridor Recycled Water Project (WCRWP). Part of the strategy is to augment surface water supplies in the major storage, Wivenhoe Dam, with purified recycled water from the WCRWP. This paper considers the implementation of potable reuse from the WCRWP which is planned to commence in October It includes discussion of the factors leading to the decision to implement potable reuse and the risk management frameworks that are being developed. REGIONAL WATER SUPPLY ISSUES Supply demand balance The SEQ Water Supply Strategy Stage 1 Report (Department of Natural Resources and Mines 2004) indicated that there was sufficient supply available to meet projected demand for water until It was not known that SEQ had, by then, started its worst drought on record. Three years later, the region is facing severe drought conditions and many capital works projects are being fasttracked to augment supply. Concurrently, demand has been significantly reduced. Provided that demand continues to be managed effectively and the capital works projects are delivered on time, projections indicate that a shortfall of supply can be avoided. doi: /wst

2 154 W. H. Traves et al. Indirect potable reuse in South East Queensland Water Science & Technology WST Demand management As a short-term drought response, the Queensland Water Commission has established a demand management program incorporating restrictions and efficiency measures (Queensland Water Commission 2007). Level 1 water restrictions started in May 2005 and progressively increased to Level 6 by November Residential demand has been successfully reduced from around 300 L/person/day to around 120 L/person/day. Demand management measures have included use restrictions, a residential water efficiency program (including rebates for replacement of shower heads, water-efficient washing machines and rainwater tank systems), and a program targeting larger commercial water users. Depending on usage, businesses and other users such as sporting organisations have been required to develop and implement Water Efficiency Management Plans (WEMPs). This has been carried out against a background of intensive advertising, culminating in the very successful Target 140 campaign to reduce residential use to less than 140 L/person/day on average. Yield assessments The critical period identified in previous assessments of historical no-failure yield was the drought between 1899 and The drought experienced over the last few years, however, has been worse from a hydrological perspective, and this has led to a reassessment of water availability from all of the dams in South East Queensland. Climate change and a revised level of service have also been taken into account, resulting in a decrease in water available from the five major storages from 542 GL/year (Department of Natural Resources and Mines 2004) to 359 GL/year (Queensland Water Commission 2008) a reduction of some 34%. Source augmentation Additional dams are currently planned in the northern and southern parts of the region. The Traveston and Wyaralong Dams are proposed to provide an additional 250 ML/day of water for the region (Queensland Water Infrastructure 2007a,b). While these will contribute to long-term supply, they will take too long to develop and fill to have any impact during the current drought. The Tugun Desalination Plant is also under construction on the Gold Coast, to provide an additional 125 ML/day by December Further desalination plants are under consideration. Role of recycled water Part of the suite of augmentation projects is the Western Corridor Recycled Water Project. The initial stages of the project are directed at reducing use of raw (dam) water by power stations, while providing security of supply for electricity generation. As at September 2007, 15 ML/day of recycled water is being supplied to the Swanbank Power Station, and water will be supplied to the Tarong Power Station by June The remaining water will be used for augmentation of surface water supplies in Wivenhoe Dam and possibly for agricultural use. WESTERN CORRIDOR RECYCLED WATER PROJECT Overview The US$2.0 billion Western Corridor Recycled Water Project (WCRWP) is currently under construction, as shown in Figure 1. Treated effluent will be collected from wastewater treatment plants and further treated at three advanced water treatment (AWT) plants incorporating micro-filtration, reverse osmosis, advanced oxidation and residual disinfection. The initial treatment capacity will be 232 ML/day. The project also includes approximately 190 km of largediameter pipelines and various pump stations. In February 2007, the Queensland Government committed to an indirect potable reuse (IPR) scheme using the purified recycled water produced by the AWT plants. Up to that time, the project had been progressing on the basis that it should neither include nor preclude the use of the water for IPR. The February 2007 decision committed to IPR supply by December Project description Treated wastewater is collected from six wastewater treatment plants operated by two local governments, Brisbane City Council and Ipswich City Council. In total, the

3 155 W. H. Traves et al. Indirect potable reuse in South East Queensland Water Science & Technology WST Figure 1 Map showing extent of western corridor recycled water project. wastewater treatment plants service a population of around one million people with an average dry weather flow (ADWF) of 292 ML/day. The three AWT plants have been located adjacent to the major wastewater treatment plants at Bundamba, Gibson Island and Luggage Point. Table 1 shows the flows available from the six wastewater treatment plants and the sizing of the AWT plants. History of project development A large-scale reuse scheme has been under investigation since Prior to 2005, proposals were predicated on the use of recycled water for irrigation and possibly cooling purposes, but not for potable application. None of the previous proposals has been workable from an economic perspective, Table 1 Wastewater and advanced water treatment plants AWT pand capacity by stage Site Average dry weather flow (ML/d) Available effluent (ML/d) Stage 1A capacity (ML/d) August 2007 Stage 1B capacity (ML/d) June 2008 Stage 2A capacity (ML/d) October 2008 Stage 2B capacity (ML/d) December 2008 Luggage Point Gibson Island Bundamba Oxley Wacol 7 7 Goodna

4 156 W. H. Traves et al. Indirect potable reuse in South East Queensland Water Science & Technology WST because full capital and operating costs were expected to be recovered directly from end users in water charges. Several studies showed that these proposals were not financially viable even if only the operating costs were considered. It was only when the region s broader water resource requirements were highlighted by the unprecedented drought that the project not only became viable, but a necessary part of the region s future water resource portfolio. Accordingly, the project in its current form commenced in late A concept report and preliminary business case were completed in March 2006, and a concept for implementation was approved for commencement in July Procurement, detailed design and delivery commenced in August 2006, and the project is on track to be completed by the end of Water treatment strategy The water treatment strategy is shown in Figure 2. Key constraints that have been taken into account in the development of the strategy include: Treatment must deliver water that is suitable for indirect potable reuse. As far as possible, salts should be removed and managed at their source. Water quality in the Brisbane River and Moreton Bay should be improved if possible. The treatment process is to be considered as part of the overall management of the water cycle. Wastewater treatment The first step in the process of water recycling is the waste water treatment plant (WWTP) that reduces the suspended solids, organic content and nutrients present in the wastewater. Currently, effluent is discharged from these plants after treatment to a standard suitable for release to local waterways and Moreton Bay. The WWTPs generally deliver water of a secondary standard. All of the plants include biological nutrient removal and achieve typical total nitrogen levels of 5 mg/l or lower. Phosphorus removal varies between the plants, with total phosphorus averaging around 3 mg/l. Salinity (TDS) is generally of the order of 500 mg/l but is significantly higher at Luggage Point (typically more than 1,000 mg/l and up to 2,000 mg/l) because of tidal ingress that is apparent within the catchment. The WWTPs generally deliver water of a high secondary standard. All of the plants include biological nutrient removal and achieve typical total nitrogen levels of Figure 2 Overall treatment process.

5 157 W. H. Traves et al. Indirect potable reuse in South East Queensland Water Science & Technology WST mg/l or lower. Phosphorus removal varies between the plants, with total phosphorus averaging around 3 mg/l. Salinity (TDS) is generally of the order of 500 mg/l but is significantly higher at Luggage Point (typically more than 1,000 mg/l and up to 2,000 mg/l) because of tidal ingress that is apparent within the catchment. Extensive testing carried out at WWTPs in Queensland has shown that this treatment step also achieves significant removal of many of the synthetic chemicals present in our wastewater, such as oestrogens and pharmaceuticals. There is substantial on-going work on the characterisation of the effluent, particularly in respect of chemicals of concern such as pesticides, herbicides, endocrine disrupting compounds (EDCs), volatile organic carbons (VOCs), N-Nitrosodimethylamine (NDMA) and radionuclides that are not typically tested for environmental release. This work is necessary in respect of the overall risk management framework, to understand the high-risk components that require ongoing monitoring. All of the WWTPs connected to the WCRWP use an activated sludge biological nutrient removal (BNR) process, with a sludge age of up to 20 days required for biological nitrogen and phosphorus removal. The relatively long sludge age in South East Queensland WWTPs results in the increased diversity of the activated sludge microflora, increasing the efficiency of micropollutant removal by biodegradation and hydrolysis of complex molecular bonds, and the adsorption of hydrophobic chemicals onto the activated sludge. Consequently, many of the synthetic organic chemicals and heavy metals that are of concern in PRW are effectively removed in the wastewater treatment process (Gardner et al. 2007). Membrane filtration The core of the treatment process includes microfiltration (MF) and reverse osmosis (RO). The choice of RO is based on achieving the organic chemical and TDS targets of the product water. Many of the treatment steps upstream were selected to support the RO process. To prevent excessive RO fouling and related high costs due to frequent cleaning and short membrane life, the feed water must be pre-treated to a high quality (e.g. turbidity,0.1 nephelometric turbidity units (NTU) and Silt Density Index (SDI),3 SDI units). MF membrane filtration has become the accepted costeffective pre-treatment technology for RO on water reclamation plants. In addition, the MF provides a microbial barrier. Based on testing conducted for the US EPA and California Department of Health Services, MF has been shown to provide at least 4-log (99.99%) removal of protozoan Giardia cysts and Cryptosporidium oocysts. Therefore, the MF provides microbial removal as well as pre-treatment for the RO. The next separation process is the RO membrane which removes dissolved solutes including nutrients, inorganic salts, organic molecules as well as viruses. The nominal pore size ( mm range) is one to two orders of magnitude smaller than virus particles. Removal of organic molecules is considered to be of the utmost importance as it is not feasible to measure or identify all the organic chemicals that can pass through a WWTP. Hence the conservative approach is to drastically reduce the concentration of these chemicals as measured by the total organic carbon (TOC) concentration in the AWTP product water, to values less than 100 mg/l TOC. Experimental observations at commercial scale operations (. 10 ML/d) using RO membranes report removal of pharmaceuticals (Drewes et al. 2005) and volatile organic compounds, nonvolatile organic compounds and disinfection by-products (Daugherty et al. 2005) to non detect levels ( mg/l) giving confidence that the separation process is very effective in removing organic and inorganic compounds. As viruses are one to two orders of magnitude larger than aqueous salt molecules, they too are very well rejected by RO membranes. Confidence in the mechanical integrity of the membranes is of the utmost importance and real time testing with direct (e.g. pressure decay) and indirect (e.g. turbidity and particle counts) tests will be implemented. In both examples the real time test parameter will be converted into a critical control point to prevent out of specification water from moving further down the treatment processing train. Advanced oxidation In addition to TDS removal, MF and RO provide removal of many other constituents; however, there is limited rejection

6 158 W. H. Traves et al. Indirect potable reuse in South East Queensland Water Science & Technology WST of some low molecular weight organic chemicals such as NDMA. To provide surety of removal, the proposed treatment approach includes advanced oxidation with ultraviolet light combined with hydrogen peroxide (UV/H 2 O 2 ) to form hydroxyl free radicals which are a very strong non-selective oxidant. Advanced oxidation is proven to be very effective at removing EDCs and NDMA. Advanced oxidation also provides a multiple barrier approach for inactivation of microbes (protozoa, bacteria and viruses). Perozone (ozone/peroxide) was an alternative advanced oxidation process, but UV/H 2 O 2 was selected since it is less likely to form disinfection by-products, is less costly and is proven to remove NDMA. There is some argument that the advanced oxidation step is unnecessary, even for indirect potable reuse applications. To some extent, the inclusion of advanced oxidation is the best available technology and meets a more practical need to develop further confidence by the end users. As with MF and RO, real time monitoring configured as critical control points is an important part of the operation of the advanced oxidation process. This will be achieved by measuring the UV intensity, the UV transmissivity and peroxide concentration (approximately 5 mg/l) of the water. Stabilisation and disinfection Due to ph adjustment and TDS removal, the RO product water will be very aggressive. Post stabilisation will be provided as lime dosing following by carbon dioxide injection to control ph. This is followed by chlorination to manage any residual ammonia and reduce biological growth in the pipelines. It may have been preferable not to stabilise the product water for cooling at power stations A lower TDS would potentially reduce water consumption by increasing the number of cycles through the cooling water circuit before the environmental release limit for salts is reached. Stabilisation is required, however, to protect the project s infrastructure (concrete tanks, cement lined pipes, etc.) and to maximise the flexibility of the system to deliver purified recycled water to other users. Chlorine levels will also need to be managed for end users. REVERSE OSMOSIS CONCENTRATE MANAGEMENT The introduction of the AWT plants provides an opportunity to reduce nutrient loads to the Brisbane River and Moreton Bay. Phosphorus removal is being achieved through chemical precipitation, using ferric chloride dosing at the AWT plant. The Luggage Point and Bundamba AWT plants are using lamellar plate separators, while the Gibson Island AWTP is using an ActifloY system. It is currently projected that up to 90% of the phosphorus from the various WWTPs will be removed from the waste stream, reducing the phosphorus load into Moreton Bay by a similar amount. Nitrogen removal strategies are still under consideration. Originally, the intention was to provide denitrifying filters on the reverse osmosis concentrate (ROC), but there are concerns about the operability and stability of filter operation in an environment with relatively high salinity (TDS up to 8,000 mg/l in the ROC at Luggage Point). Alternatives, including ion exchange, denitrification as pretreatment and the use of wetlands, remain under consideration. Potable reuse framework The use of treated wastewater effluent, even using RO treatment, to supply industrial reuse is not unique in Australia. However, the planned use of recycled wastewater to augment potable water supplies is relatively unusual in the world and unprecedented in Australia. Consequently, Queensland is a pathfinder in this area in the Australian context and has leaned heavily on overseas precedent and experience to inform the various government agencies on the relevant design, operation and regulatory processes. There is a general agreement that purified recycled water requires a multi-barrier philosophy and a seven-barrier approach has been adopted. The concept of the multi barrier approach is shown in Figure 3. Taken overall, the MF/RO/AOP treatment train of the AWT plants in the WCRWP is confidently expected to consistently and reliably produce pathogen, organic and (essentially) salt free water which is suitable for drinking without further treatment. The technology has been well tested in Orange County, Scottsdale and Singapore (see, for example, Gardner et al. 2007).

7 159 W. H. Traves et al. Indirect potable reuse in South East Queensland Water Science & Technology WST Figure 3 Schematic of multi-barrier approach (Queensland Water Commission). As part of the overall risk framework, however, additional consideration is required of source control, the environmental barrier and water treatment. Source control In most planned IPR schemes, considerable effort has been invested to separate domestic sewage from industrial sewage examples range from Windhoek in southern Africa to Montebello in Los Angeles (Gardner et al. 2007). In contrast, the existing WWTPs in the Brisbane and Ipswich areas that will supply water to the WCRWP receive effluent from both domestic and a wide range of industrial sources. An extensive range of organic and inorganic contaminants can therefore be expected to be received at the WWTPs. Trade waste controls are already implemented by the Brisbane and Ipswich City Councils to minimise disposal of synthetic organic chemicals, heavy metals, salts and excessive organic material that can affect the functioning of the WWTPs, but may need to be upgraded for the WCRWP based on detailed measurement of sewer composition and associated risk assessment (QNRM&E & EPA 2004). Pharmaceutical waste from hospitals is of concern to the general public, particularly the potential disposal of cytotoxic and radioactive drugs into the sewer system. Current regulations prohibit the disposal of pharmaceuticals and cancer treatment drugs into the sewer system but a significant export pathway is non-metabolised drugs in the patient s urine. It can be expected that these chemicals will exist in sewage influent entering the WWTPs (Carballa et al. 2004).

8 160 W. H. Traves et al. Indirect potable reuse in South East Queensland Water Science & Technology WST Environmental barrier The buffering storage is another important component of the multi barrier philosophy in IPR schemes (Figure 3). The buffer provides dilution and travel time and allows further natural treatment of the purified recycled water (PRW). A review of overseas PRW schemes (Gardner et al. 2007) identified only Windhoek in Namibia where PRW is directly connected into the potable water supply. Other schemes either inject into or recharge an aquifer system, pump into a reservoir, or discharge into a river system. In California, the minimum required transit time in the aquifer before reuse is six months (DHS 2004) with monitoring wells to track water quality between the injection well and the potable extraction well. In the Upper Occoquan (near Washington DC) the recycled water is discharged into a long (32 km) narrow reservoir and contributes about 10% of the total flow entering the reservoir but over 30% of the safe yield extracted from reservoir for potable use. In comparison, there is a 40 km reach of the Brisbane River between the Wivenhoe Dam release point and the intake at Mt Crosby Water Treatment Plant which is sunlit and well mixed, and unpublished data (SEQ Water) has reported a substantial improvement in nutrient concentrations and turbidity. The key to managing an environmental buffer is quantitative information on the travel time from the point of release into the reservoir to the point of extraction, and the amount of dilution of PRW with catchment runoff water in the reservoir. Preliminary studies suggest a detention period of more than 6 months in Lake Wivenhoe, although this is the subject of current limnological and hydrodynamic study. However, if one accepts the simplistic assumption of perfect mixing, the contributions from PRW could vary from 10% during normal supply conditions to around 37% when the storage is around 10% capacity. Buffering storage also provides assimilative capacity for extra nutrients imported in the PRW. Target nutrient contents are 0.8 mg/l for total nitrogen and 0.1 mg/l for total phosphorus with provision for reductions in membrane performance. There is some concern that the bioavailable phosphorus concentration may contribute to an algal outbreak in Wivenhoe Dam, which would be of particular concern if blue-green species (cyanobacteria) dominated. A limnological study has been commissioned to assess the nutrient assimilative capacity of Lake Wivenhoe, and this is expected to continue for 5 years after purified recycled water is introduced into the lake. This information will then inform the scheme operators and regulators if additional treatment is warranted to manage the (then) quantified risk of algal outbreaks. Water treatment plant The Mt Crosby water treatment plant produces the majority of the potable water consumed in SEQ. The treatment process at Mt Crosby is quite conventional in its use of flocculation/ sedimentation, granular media filtration and chlorine disinfection to produce potable water. Removing suspended sediments, and therefore any attached pathogens and chemicals, is a primary treatment goal. However this treatment will not remove dissolved contaminants including iron, manganese and synthetic organic chemicals (Gardner et al. 2007). A small proportion of the water from Lake Wivenhoe is treated and used much closer to the dam to supply the small rural towns of Esk, Lowood, Gatton and Laidley. These smaller supplies are likely to involve a higher risk because of proximity to the dam and the scale of operations. Consideration will be given to the overall treatment strategies at these sites through a detailed risk analysis. Upgrades to these plants may be required, perhaps using an ozone/bac process. Risk management framework A whole of cycle risk management approach is being adopted for water quality management. Not surprisingly, this is focussed on water quality at the tap and requires consideration of the whole cycle, as shown in Figure 3. One of the complications is that there are multiple entities currently involved in the whole cycle. A framework has been established between the various entities, although this is complicated by a major restructure of the water industry that is currently in progress. The regulatory framework under which the WCWRP will operate is being developed by the Queensland Government and will integrate with the new national recycled water guidelines (NRMMC et al. 2007) which were released in draft form for public comment in July

9 161 W. H. Traves et al. Indirect potable reuse in South East Queensland Water Science & Technology WST The framework requires the operator to develop a Recycled Water Management Plan that will include an audit of the activities that occur in each wastewater catchment (i.e. trade waste control), a comprehensive audit of influent composition and source control, and an evaluation of the capacity of the wastewater treatment process to reduce the concentration of chemicals of concern. These data will then be complemented by water quality monitoring. Real time monitoring data will be required to ensure the continuous operation of the various treatment processes and their commissioning as critical control points to ensure that out of specification water is not released to the next treatment step (in keeping with the HACCP philosophy). An independent Expert Advisory Panel incorporating international expertise has been formed by the Queensland Water Commission to review water quality standards, the regulatory guidelines, the validation protocols for the AWT plants, the operating procedures for the AWT plants and water quality results as the AWT plants come on line. CONCLUSIONS Indirect potable reuse of purified recycled water has been adopted as part of the long-term water supply strategy for South East Queensland. Early availability of this supply has become critical due to the current drought, and the Western Corridor Recycled Water Project is being delivered on a fast-track basis as part of a suite of supply and demandside measures to ensure ongoing supply of water to the region. The first supply of water was achieved in August 2007, with indirect potable reuse scheduled to commence in October In the first instance, recycled water is being directed to major industrial consumers of water power stations to release existing raw water supplies for urban consumption. The balance of available water will be released into Wivenhoe Dam to augment surface water supplies. It is anticipated that the region s water supply, on average, will include approximately 10% recycled water. A treatment process including microfiltration, reverse osmosis and advanced oxidation has been selected as the core of a seven-barrier whole-of-cycle approach for indirect potable reuse in South East Queensland. A risk management framework is being implemented to comply with evolving national guidelines and a new State-based regulatory regime. REFERENCES Carballa, M., Omil, F., Lema, J., Llomport, M., Carcia-Jones, C., Rodriguez, I., Gomez, M. & Ternes, T Behaviour of pharmaceuticals, cosmetics and hormones in a sewage treatment plant. Water Res. 38, Daugherty, J., Deshmukh, S., Patel, M. & Markus, M Employing Advanced oxidation for water reuse in Orange County, Watereuse Association, California Section Conference, San Diego, California. Department of Natural Resources and Mines 2004 South East Queensland Water Supply Strategy Stage 1 Report, Queensland Government and Brisbane City Council, Brisbane, Queensland, Australia. DHS 2004 Title 22 California Code of Regulations, Division 4 Environmental Health, Chapter 3 Recycling Criteria, Department of Health Services, Sacramento, California. Drewes, J. E., Bellona, C., Oedekoven, M., Xu, P., Kim, T. U. & Amy, G Rejection of wastewater-derived micropollutants in high-pressure membrane applications leading to indirect potable reuse. Environ. Prog. 24, Gardner, E., Yeates, C. & Shaw, R. (eds) 2007 Purified Recycled Water for Drinking: The Technical Issues. Queensland Water Commission, Brisbane, Australia. NRMMC, EPHC & AHMC 2007 Australian Guidelines for Water Recycling: Managing Health and Environmental Risks (Phase 2), draft, Natural Resource Management and Ministerial Council, Canberra, Australia. PIFU 2007 Projection of Demographic Indicators, Council of Mayors (South East Queensland) & Office of Urban Management, Planning Information and Forecasting Unit, Department of Local Government, Planning, Sport and Recreation, Brisbane, Queensland, Australia. QNRM&E and EPA 2004 Trade Waste Management Plan. Department of Natural Resources, Brisbane, Queensland. Queensland Water Commission 2007 The Framework for a South East Queensland Regional Demand Management Program (Consultation Draft). Queensland Water Commission, Brisbane, Queensland, Australia. Queensland Water Commission 2008 Water for Today, Water for Tomorrow South East Queensland Water Strategy (Draft). Queensland Water Commission, Brisbane, Queensland, Australia. Queensland Water Infrastructure 2007a Traveston Project Overview. Queensland Water Infrastructure, Brisbane, Queensland, Australia. Queensland Water Infrastructure 2007b Wyaralong Project Overview. Queensland Water Infrastructure, Brisbane, Queensland, Australia.