AWA .::r =...-.. 11,,,,_
-
•
AUSTRALIAN WATER ASSOCIATION ABN 78 096 035 773
Volume 34 No 3 May 2007
Journal of the Australian Water Association
OPINION AND INDUSTRY NEWS OPINION The Way Forward DBarnes, Presidenf, AWA Energy' s the Game CDavis, CEO, AWA My Point of View David Mower, CEO, Wesfernporf Water AWA NEWS Includes: WaterAid Australia, A Place for Indirect Potable Reuse in Australia's Urban Water Management AWA EDUCATION Includes: Water Education Network (WEN); Young Water Professionals (YWP) OZWATER 2007 Reporl by EA (Bob) Swinlon, TLoos The Carbon Challenge for the Victorian Water Industry The VicWator Board recommendations PMaddy CROSSCURRENT National Issues and Policy, States, New Reports & Papers, Programs & Awards, People in the News AWA MEMBERSHIP NEWS New Members
4
5 6 8 14 22 30
38 42
PROFESSIONAL DEVELOPMENT NATIONAL EVENT CALENDAR
44
EVENTS
46
TECHNICAL FEATURES ( ·,
indicates the paper has been refereed)
INDIRECT RECYCLE ISSUES [i] Indirect Potable Reuse: Managing the Risks The formal management of risk associated with the supply of drinking water MEMBRANE TECHNOLOGY Membrane Technology Conference Development and Future of Membrane Technology Seawater desalination is here to stay, so long as people prefer to live in coastal communities. MBRs: How Much do they Cost? They cost more but have significant advantages in certain situations Membrane Bioreactors in Australia - Forecast is for Growth Asurvey of MBR pro;ects, mainly municipal, either operating or pro;ected in the near future
HJackson
50
Reporf by EA (Bob) Swinfon
56
SAdham
62
SJudd
65
SChapman
68
WATER SUPPLY Interaction Between Treatment and Distribution Systems: the Key to Delivering Quality Interested utilities can be still be involved in this pro;ect. J QJ CVerberk, MDrikas [i] Trihalomethane Management In Drinking Water: A Novel Approach Aeration dissipates THMs once formed in the bulk water. RTrolio, RWalker, SMcNeil, THuynh, NJismi [i] Evaluation of Two Water Toxicity Testing Kits The IQ-Tox™was more sensitive than the Edox™ test kit to the contaminants tested CCook, DWren, GRose
72
77
82
WATER BUSINESS NEW PRODUCTS AND BUSINESS INFORMATION • SPECIAL FEATURES:
INSTRUMENTATION, SLUDGE MANAGEMENT, CORROSION
89
ADVERTISERS' INDEX
100
OUR COVER "Western Australia was hit earlier and harder by the current climate seq11ence" said Dr Jim Gill, CEO ofthe Water Corporation, at Ozwater 2007, as he explained the shift in ocean currents which may be the cause of the reduction in Perth '.r rainfall. Their 125 ML/d desalination plant is only one ofthe initiatives to guarantee an adequate water supply into the future. See Ozwater 2007 report on page 22.
Journal of the Australian Water Association
Water
MAY 2007 1
~ AWA CONTACT DETAILS
liiiflll' 'Promoting the,I'sustainable , ...~ management o1 water o\llUUUA..
POSTAL ADDRESS PO Box 388, ARTARMON NSW 1570
Journal of the Australian Water Association
EMAIL inlo@owo.osn.au WEBSITE http://www.owo.osn.au PRESIDENT
ISSN 0310-0367
David Barnes - president@owo.osn.au
CHIEF EXECUTIVE OFFICER Chris Davis - cdovis@owo .osn.au
CHIEF OPERATIONSOFFICER Ion Jarmon - ijormon@owo.osn.au
EVENTS Lindo Phillips - 61 2 9495 9914 lphillips@owo.osn.au
MEMBERSHIP INFORMATION AND INQUIRIES Michael Seller - 02 658 1 3483 mseller@owo.osn.au
MEMBERSHIP RENEWALS AND CHANGES Membership Team - 1300 361 426 info@owo .osn.a u
MEDIA AND MARKETING Jennifer Sage - jsoge@owo.osn.au
SCIENTIFIC AND TECHNICAL INFORMATION Dione Wiesner PhD - 61 2 9495 9906 dwiesner@owo.osn.au
WATER EDUCATION NETWORK Corinne Cheeseman - 6 1 2 9495 9907 ccheesmon@owa.osn.au
NATIONAL SPECIALIST NETWORK Louro Evanson - 6 1 2 9495 9917 levonson@owo.osn.au
AWA BRANCHES: AUSTRALIAN CAPITAL TERRITORY and NEW SOUTH WALES Errin Dryden - 61 2 9495 9908 edryden@awo.asn.au NORTHERN TERRITORY c/o Ion Jarmon - 61 2 9495 9911 ijarman@owo.osn.au SOUTH AUSTRALIA Sarah Co rey - 61 8 8267 1783 sobronch@owo.osn.au QUEENSLAND Kathy Bourbon - 6 1 7 3397 5644 owoq@owo.osn.au TASMANIA & VICTORIA BRANCH c/o Rochel-onn Martin - 6 1 3 9235 14 16 tosbronch@owo.osn.au vicbronch@owo.osn.au W ESTERN AUSTRALIA Coth Miller - 04 16 289 075 cm iller@owo.osn.au INTERNATIONAL WATER ASSOCIATION, AUST. (IWAA) c/o Ch ris Davis - cdovis@owa.osn.au
DISCLAIMER Australian Water Association assumes no responsibility for opinion or statements of focts expressed by contributors or advertisers.
COPYRIGHT AWA Water Journal is subject to copyrig ht and may not be reproduced in any format without written permission of AWA. To seek permission to reproduce Water Journal material email your request to: jsoge@owo.osn.au
2 MAY 2007
Water
Volume 34 No 3 May 2007
AWA WATER JOURNAL MISSION STATEMENT 'To provide aprint ;ournal that interests and informs on water matters, Australian and international, covering technological, environmental, economicand social aspects, and to provide arepository of useful refereed papers.' PUBLISH DATES Water Journal is published eight times per year: February, March, May, June, August, September, November and December EDITORIAL BOARD Chairman: FR Bishop BNAnderson, TAnderson, CDiaper, GFinlayson, AGibson, GAHolder, BLabza, MMuntisov, CPorter, DPower, FRoddick EDITORIAL SUBMISSIONS Water Journal inviles edilorial submissions for: Technical Papers and topical articles, Opinion, News, New Products and Business Information. Acceptance of editorial submissions is subject to editorial board discretion. Email your submissions to one of the following lhree categories: 1. TECHNICAL PAPERS AND FEATURES Bob Swinton, Technical Editor, Water Journal: bswinton@bigpond.net .au AND journal@owo.osn.au Papers of 3000-4000 words (allowing for graphics); or topical stories of up to 2,000 words. relating lo all areas of the water cycle and water business. Submissions ore tabled at monthly editorial board meetings and where appropriate are assigned lo referees. Referee comments will be forwarded to the principal author for further action. See box on page 10 for more details. 2. OPINION, INDUSTRY HEWS, PROFESSIONAL DEVELOPMENT Jennifer Sage, jsage@awa.osn.au Articles of 1000 words or less 3. WATER BUSINESS Brian Rauh, National Soles & Advertising Manager, Hallmark Editions brion.roult@halledit.com.au Water Business updates readers on newproducts and associated business news within the water sector. ADVERTISING Brion Rauh, National Soles & Advertising Manager, Hallmark Editions Tel: 61 3 8534 5014 (direct), 61 3 8534 5000 (switch), brion.rault@halledit.com.au Advertisements are included as an information service to readers and are reviewed before publication to ensure relevance to the water environment and objeclives of AWA. PURCHASING WATER JOURNAL Single issues available@ $12.50 plus postage and handling; email dwiesner@awa.asn.au BACK ISSUES Water Journal back issues ore available to AWAmembers at www.awo.asn.au PUBLISHER Hallmark Editions, PO BOX84, HAMPTON, VICTORIA 3188 Tel: 61 3 8534 5000 Fax: 61 3 9530 8911 Email: hallmark.editions@halledil.com.au
Journal of the Australian Water Association
s
INDIRECT POTABLE REUSE: MANAGING THE RISKS H Jackson Abstract T his paper considers the management of risk associated with indirect potable re use. lcs foc us is on che managemen t of legal risk w ith a particular emphasis on che use o f risk m an agement plans.
Introduction T he reuse of created wastewater (recycled water) fo r va rious benefi cial uses has rapidly gain ed accep tance as an appropriate response co the water shortages fa cing many pares of Australia. Pro longed d rought and che impacts of climate change have resulted in w h at would , until qu ite recently, have been co nsidered quite radical proposals gaining accep tance in relatively sho re periods of time. Recycled water is co mmonly used in industry and agriculcu re and is also being
increasingly introd uced into d om estic contexts via third or "pu rple" pipe systems where recycled water is used for coilec fl ush ing and garden use. However, the use of recycled water for p otab le uses has, where it has been che subject of o pen public consul catio n or d ebate, received less favou rable responses fro m the general pu blic. No netheless, indirect potab le reuse is already a reality in many dri nking water catchments and, in many cases, has been so fo r many years o r d ecades . In both new and existing scenarios, che m anagem ent of risk will be a key issue. Cer tainly, che recent Toowoomba referend um experience d em o nstrates char managing co mmun ity expectations and perceptio n w ill be of co nsiderable importance in ch e case of fucure poten tial
indirect potable reuse schemes. I t is suggested chat careful management of community expectations and percep tions will signi fic antly impact o n (and potentially red uce) ocher risks including legal risks. T his paper considers the management of risk associated with indirect potable reuse. I cs foc us is on che m anagem ent of legal risk with a particular emphasis o n the use o f risk management plans. The Victorian Safe Drinking Water Act 2003 (SDWA) requires the suppliers of drinking water and managers of dams that supply drinkin g water co prepare and implement r isk managem ent plans. T he paper b riefly co nsiders the strengths and weaknesses of risk management p lans and suggests that a regime similar to chat in Victoria should be implemented in ocher jurisdictions.
Indirect potable reuse
Specify and use Link-SealÂŽ seals to seal pipe penetration
T he ter m " indirect p otable reuse" is sim p ly o n e of seve ral terms used co d escribe che situatio n wh ere" .. . recycled water rep lenish es the so u rce of drinki n g water fro m eithe r gro undwater bas ins or surface water reservoirs." (Di m itri adis 2 00 5) . I t is impo rtan t co st ress tha t ind irect p otab le reuse can be both planned , as in t he case of the T oowoomba p roposal o r u nin te ntio nal. I ndeed, t he m edia attent io n chat focused on th e Toowoo mba p roposa l brou ght co ligh t several exa mpl es of lo ng term unin tend ed ind irect pota bl e reuse, including in relat ion co ch e water supplies of bot h M elbourne and Br isbane.
Slide the assembly into the space between the pipe and wall opening
PR •jEX
When the bolts are tightened, Link-Seal modular seals expand to create a gas and water tight seal. A vailable in Australia and New Zealand
Projex Group Pty Limited Telephone: 02 8336 1666, Facsimile: 02 8336 1670 www.projex.com.au , email : mail@projex.com.au
50
MAY 2007
Water
l e is also important co note chat so me considerab le work has already been done in relatio n co che managem ent o f the technical risks asso ciated with indirect potable reuse including the repo rt of che W A EPA "S trategic Advice on M anaged Aquifer Recharge Using Treated W astewater on che Swan Coastal Plain" (Bulletin 1 199 O ctober 200 5).
Journal of the Australian Water Association
The fonnal management ofrisk associated with the supply of drinking water.
What is /WES th
Now in its 28 year, /WES is the largest and most successful continuing education program in Australia for professionals in the water and related industries. /WES courses are designed to keep busy professionals abreast of the latest trends and developments.
AWA ~
•
AWA support IWES courses
/WES courses will run in Sydney, Gold Coast, Perth, Kuala Lumpur and Manchester in 2007.
AUSUALIAN
WAHl
ASSOCIATION
The Courses
Principles of Wastewater Treatment The aim of this course is to teach the key enabling scientific and process fundamentals which underpin wastewater treatment. Th is is the most popular wastewater treatment course in Australia. 5 days
Trade Waste and Industrial Wastewater Treatment Learn how to characterise trade waste and industrial wastewater, how to specify appropriate technologies for treatment and re-use, and how to trouble-shoot industrial wastewater treatment problems. 5 days
Design of Biological & Advanced Wastewater Treatment Plants This course focuses on the design of leading edge systems for both municipal and industrial settings. You will become familiar 5 days with the issues that are driving modern treatment plant designs.
Introduction to Potable Water Treatment Th e aim is to introduce the issues relating to water quality, identify the need for water treatment and describe the water treatment processes in use today. This is a practical course, and case 3 days studies are used extensively in teaching.
Introduction to Membrane Technology For people with an interest in the use of membrane processes to treat drinking water and wastewater in Australia. Learn how to select which membranes to use for different applications, size plants 1 day and understand fouling problems.
Membrane Plant Design Learn how to design membrane systems for water and wastewater treatment, water recycling and desalination. The course is designed for water professionals with a basic understanding of membrane 2 days systems, water chemist ry and the design process.
Water Re-use for Urban and Industrial Applications Learn how to plan, design and implement reclamation and re-use schemes. The course covers public health goals, non-potable recycli ng in irrigation, grey water and storm water recycling and indirect potable re-use applications. 2 days
Environmental Toxicology and Risk Assessment for Water Re-use The aim of this course is to introduce the fundamentals of environmental toxicology and how the information can be used in health 2 days risk assessment.
Water Recycling for Agricultural and Horticultural Applications Learn about the benefits and hazards of irrigating with recycled water, and how to determine if the water quality is appropriate for the intended use. This is a practical course aimed at developing real ski lls in water recycling for agricultural and horticultural applications. 2 days
Principles of Air Quality Learn the principles of air quality monitoring, modelling, assessment and em issions control.
2 days
Odour Assessment and Management This course provides a practical grounding in the principles of odour monitoring, assessment and control. Regulatory frameworks for odour assessment and management are examined. 2 days
Integrated Urban Water Management Learn about the ph ilosophies, principles and techniques of Integrated Urban Water Management. Attendees will have the opportunity to develop skills in techniques such as technology selection, end-use ana lysis, life cycle costs and triple-bottom-line analysis. 2 days
Water Trading and Water Markets Learn practical tools and methodologies to enable you to participate in water markets and trading in Australia . This course is designed to up-skill water and environment industry professiona ls and managers. 2 days
Ecological Risk Assessment Learn how to assess and quantify ecological risks to water resources, and use this risk analysis to develop risk management plans and 2 days monitoring programs.
Principles of Wastewater Microbiology The aim of this course is to teach the microbiological principles t hat underpin biological wastewater treatment processes.
2 days
For f urther informat ion phone 1800 000 404, Ema il info@iwes.com.au or visit our web site www.iwes.com .au where you can view the schedule and download the brochure and regist ration form .
indirect rec ~ Planned indirect potable reuse is said to provide many of the same benefits chat are provided by non-potable recycled water projects including the ability co offset demand for raw water, che retention and restoration of environmental flows (in che surface water context), reducing the need for the creation of new water harvesting infrastructure and an increase in security of supply. Of course, it is also important to note that in some cases it might be argued chat indirect potable reuse rep resents one of several options, each of which may be necessary to ensure the security of long term d rinking water supplies. In such a case, it is suggested char community consulcacion and management of public outrage may well differ. However, in all ocher senses, man agement of risk should be co mparable.
campaigns as well as the real risk of public health impacts, both of which require careful management.
be che case char for a large planned ind irect potable reuse scheme legislative reform is required.
2. Financial or economic risks include the very significant coses associated with establishment of the necessary infrastructure (pipes and pumps, treatment works, etc) . Financial or economic coses also include che coses of managing any public outrage and also include che coses of compliance with any relevant regulatory requirements. In relation to chis last point, it seems reasonable to suggest char these coses will rise.
Ocher legal risks include che risk of consumer action, whether chat is for a breach of statutory duty, for example, an action under the Trade Practices Act 1974 (C'ch) for supply of goods char are nor "fie fo r purpose" as well as a more general claim in negligence. The litigation followi ng the contamination of oysters by faecal matter at Wallis Lakes in NSW is a good examp le o f the legal risk to wh ich suppliers of drinking water as well as regulators are exposed - see Ryan v Great Lakes Council (1999) 102 LGERA 123 appealed under different names ro the Full Court of the Federal Court and, on limited grounds ro che High Court.
3 . Environmental risks include energy impacts and associated greenhouse gas emissions associated with the treatment of wastewater to a standard sufficient to allow it co access d rinking water supplies (these should be assessed utilising the emerging science of life cycle analysis), the risk of pollution and any impacts on che water resource itself.
Risks associated with indirect potable reuse The risks associated with indirect potab le reuse can be summarised under three headings that broadly correspond with a Triple Bottom Line approach:
Another factor, which is h inted at in various of the social, financial and environmental risks associated with indirect potable reuse is legal risk. Legal risks include whether the legislative or regularory regime allows indirect potable reuse. It may
1. Social risks include both che impacts of large-scale or well-resourced public outrage
A
CLEXAN-PEAK l'lllSON Nil A range of Permanent & Contract Water Industry vacancies are available: Water Services Engineer. Contract 3mths. Narrabri NSW. Local Govt. Excellent rate - Neg. Ref: W1467CD. Water/Water Treatment. Consulting Engineer. Long-term - Contract/ Permanent. Sydney. Ref: W1547CD. Urban Drainage Officer/Engineer. Perm. Dubbo. Govt. Org. Sound knowledge & exp. in civil eng. field particularly design & construction of stormwater facilities. Ref: W1378LV Water/Wastewater Engineer. Perm. Sydney CBD. Relevant tertiary quals. with extensive exp. in water/ wastewater industry. Ref: W1508LV. Design Engineer - Water/W'water. Perm. Sydney CBD. Consulting background with relevant technical exp. water industry. Ref: W1508LV Mech. Engineer - Water Supply. Contract 3-6 months. Sydney CBD. Exp/Knowledge water t/f reticulation infrastructure. Ref: W1 625RM. Civi l Engineer - Water Supply. Contract 3-6 months. Sydney CBD. Exp. in hydraulic analysis - water infrastructure. Ref: W1626RM. Electrical Engineer - Water Supply. Contract 36 months. Sydney CBD. Exp. in water treatment, design & commissioning. Ref: W1631 RM.
Project Manager - Water Supply. Contract up to 6 months. A.C.T. Water recycling/ dams knowledge. Roller compacted concrete. Ref: W1646RM. Project Engineer - Water Supply. Contract 6 12 months. A.C.T. Knowledge of Membrane Filtration Plants. Ref: W1648RM. General Manager (NSW) Water/ Wastewater Treatment. Permanent. Sydney. Strong Business development skills. Extensive water/ wastewater treatment experience. Ref: W1425RL. Project Manager - Water/Wastewater. Permanent. Brisbane. Expanding Contracting Company. Ref: W1427RL. Site Engineers - Water Industry Perm. SE OLD. Various Roles Major Infrastructure projects. Ref:W1428RL. Sile/Project Engineer - Water/ Wastewater. Permanent. Brisbane. Alliance projects. 2 yrs minimum water/wastewater exp. Ref:W1543RL. Senior Design Engineer - Membrane Specialist. Perm. Brisbane. MF/UF/RO Membrane Plant exp. Ref: W1659RL. Senior Mechanical Engineer. Perm. Brisbane. 5-1 Dyears exp. water/ wastewater industry. Pumps & pipes design. Ref: W166DRL.
For a confidential discussion call (02) 9876 8888
52 MAY 2007 Water
Journal of the Australian Water Association
Finally, there is a risk of legal action brought by a relevant regulator whether that is under the environmental regime estab lished in che relevant jurisdiction or the public health regulator. In Victoria there is also a risk of enforcement action for breach of che Safe Drinking Water Act 2003 (SDWA).
Risk management The requi rement, imposed by che SDWA for suppliers of potable drinking water to d evelop and implement risk management plans (SDWA Sections 7 and 8) represents a "third wave" of regulation (Cunningham 1996). First generation regulation was often referred to as "specification" regulation which sough t to achieve a certai n goal by p rescribing che manner in which a regulated body was required to undertake an activity. The second wave of regulation imposed performance standards, often drafted very broadly, which regulated entities were required to achieve in a manner of their choosing. In comparison, "systems based regulations" specify procedures to be followed in managing particular hazards. Proponents of such an approach argue char organ isational factors (structures, responsib ilities, practices, procedures, processes and resources) are incorporated into che process or system; something which does nor happen in specification or performance based regulation. Such an app roach therefore "spans che entire organisation by relating che organisation to its environment, setting the goals, developing comprehensive, strategic and operational plans, designing che structure and establishing co ntrol processes" (Koor and Rosenweig,1974). T h is in turn is said to bring abou t systemic organisational change across the organisation as a whole, resulting in a movement "beyond
Increased flow capacity Long term hydrostatic strength Extraordinary resistance to impact Half the weight- reduces OH&S risks Size range 100 - 375mm Series 2, PN 12.5 to PN 20 Half the embodied energy Half the raw materials 100% recyclable No heavy metal stabilisers Energy savings in pumping costs Manufactured to comply with AS 4441 (lnt):2003 Minimum lifespan of 100 years Australian innovation
For more information on Supermain or other great products - Contact Vinidex TODAY!
HOTLINE E-MAIL WEB
13 11 69 info@vinidex.com.au www.vinidex.com.au
Quality
Vi~
Company
Systems & Solutions
SAIGlobol
*comparisons have been based on Vinidex Vinyl Iron pipe and Supermain pipe.
Endorsed ISO 8001 iJ0 670
__!_ESR, Allstraliil
Answer all of your "where" questions with ESRI Australia's GIS solutions ESRI Australia's GIS solutions allow you to not only organise and manage this location-based information from various sources, but also enables distribution of this information via the web and out in the field.
With so much of a water utility's business information containing a geographic reference, being able to effectively access this where-related information can make a big difference when impacted by events such as main breaks and stoppages.
Be better positioned to answer all of your "where" questions, and improve incident management, deliver better customer service, and increase your operating efficiencies.
Find out more at our free GIS Solutions for Utilities event, being held in Brisbane on 9th August (following GITA). Contact us for more details.
wheresolutions@esnaustralia.com.au www.esriaustralia.com.au Freecall 1800 447 111
Where Advantage Begins
technical features
t
indirect recycle issues compliance" chat ocher forms of regulation do not or cannot achieve.
NRMMC Australian Drinking Water Guidelines).
Of course, systems-based regulation is a m uch broader concept than simply requiring the creation of risk management p lans. Nonetheless, risk management plans as a regulatory tool are a good example of this "third wave" of regulation .
By imposing minimum standards, the SDWA avoids two of the most criticised aspects of systems based regulatio n: the danger of "shelf systems" and the asso ciated danger of d ifficu lty of enforcement.
Australian Standard 4360/ 1999 requires chat risk management occur in a five-s tep process .
1. T he first seep is to establish a context including the d etermination of criteria against which certain risks will be assessed. 2 . The second seep requires the identification of risk - what can happen and how? le is viral chat all risks are identified, including chose beyond the control of che parry in question. 3 . The third seep is co undertake an analysis of the risk, a process which requires establishing both the likelihood of ari even t occurring as well as considering what consequences might flow from such an event. 4. The fourth seep requires an evaluation of che risks identified. This requires an assessment of the results of the analysis against the criteria established in the contextual phase. I mportantly, it also requires establishing criteria around "acceptable risk" risks. 5 . F inally, the fifth seep requires creacmenc of chose risks chat are determined ro be "unaccep table". T he Australian Standard also requires consultation with all relevant stakeholders at each of the five stages as well as o ngoing review and revision.
Risk management plans The SDWA requires the suppliers o f drinking water ro prepare, implement, review and revise a risk management plan in relation co the supply of drinking water. Ir is prescriptive in char it requires a risk management plan ro:
1. D escribe the systems (a process which might otherwise be described as establishing the con text); 2. Identify risks both co the system and risks posed by the system; 3. Assess che risks (which might otherwise be seen as including both che analysis and the evaluation phase); and Sec out steps co be taken co manage chose risks (SDWA Section 9). The SDWA also requires chat drinking water supplied by a water supply authority must meet minimum standards (SD W A Section I 7) (generally the N HMRC &
54
MAY 2007
Water
By imposing minimum standards ch at muse be met, ic is less likely chat risk management p lans will be created b ut not implemented. The obligatio n co meet m inimum standards also provides the regulator with definitive outcomes against wh ich the regulated entity can be assessed . Notwithstanding the positive aspects of the regularory regime established by the SDWA cwo criticisms can be made. The first is that whereas AS4360 requires consultation with relevant stakeholders at each stage of the risk management process, the SDWA imposes no obligation to consul t excep t where minimum standards cannot be m et and ic is necessary co seek authorisation co vary the aesthetic standards or an exemption fro m stated water quality standards (SDWA Sections l 9(3)(d) and 20(2)(e)). While the complexities and d ifficulties o f extensive public consultation should not be under-estimated, it seems reasonable ro suggest chat the pu blic (or relevant segments of it) have a right co be consul red on the m anner in which their drinking water supplies are being managed . The other aspect of the SDWA that is open co criticism, is in the area of "acceptable risk". Acceptable risk is defined as a risk which either has a very low likelihood o f occurring or one where the cost of creating the risk outweighs the benefits char accrue from char treatment. AS4360 states that acceptable risks need not be treated. (Standards Australia 1999). This is to be compared with the Safe Drinking Water Act which requires chat all risks be treated but is silent as co how char treatment should occur (SDWA Section 9(l)(d )). l e is at chis point where che lack of consultation required by the Ace in preparing a risk management plan and the concept of acceptable risk collide. The concept of a risk management plan depends on che "rational" treatment of risk in a scientific manner. The To owoomba referendum scenario should be sufficient co remind us chat the scientific or ratio nal assessment of risk and its treatment is not necessarily suffi cient when dealing with the supply of potable water co the public. There is a well recognised distinction between what science determines co be an acceptab le risk and what the p ublic
Journal of the Australian Water Association
determines co be an acceptable risk with many factors impacting on chat distinction including the identity o f the parry m aking the decision , the identity of the party chat controls the risk, whether the community trusts the entity in control and whether the party chat bears the burden of the risk also gees any benefits associated with, for example, cost savings.
It is suggested chat the safe drinking water regime in Victoria might benefit from additional guidance, much as the regime fo r managing Legionella risk under the Building Act 1994 provides for che creation of formal gu idelines, developed by the relevant department.
Conclusion Victo ria has establish ed che first legal regime in Australia co require the formal management of risk associated with the supply of drinking water. T har regime, while not perfect, provides an admirable attem p t co ensure chat the supplier of drinking water undertakes a thorough review o f che d rinking water supply system, identifies all risks chat might be available to that system as well as risks proposed by chat system and undertakes the treatment of chose risks . Indirect potable reuse p oses vario us risks, primarily associated with public health, some of which manifest themselves in public outrage and some of which manifest themselves as the risk of legal action following public health impacts. A well managed process of development and implementation of a risk management plan can, it is suggested, adequately manage the risks associated with indirect potable reuse. Such a risk management process should include well considered consultation with the public in order to avoid a conflict between different versions of "acceptable" risk.
The Author Henry Jackson is a partner with Deacons, specialisi ng in water and environmen tal matters, henry.jackson@deacons.com.au
References Cunningham N (1996) "From Compliance co Best Practice in OHS: the Roles of Specification Performance and SystemsBased Standards" 9 Australian journal of Labour law p 22 I Dimitriadis S, 2005 . Issues encountered in advancing Australia's water recycling schemes, Parliament of Australia, Parliamentary Library, 16 August Koor and Rosenweig (I 974) Organisation and Management: A Systems Approach, p 113 Standards Australia ( 1999) A Basic Introduction to Managing Risk: ASINZS4360/1999
technical features
MEMBRANE TECHNOLOGY CONFERENCE Melbourne, February 2007 Report by EA (Bob) Swinton Another excellent Specialist Conference, where practitioners meet and discuss freely and novices learn from the legends. Diane W iesner, in collaboration with Tony Fane and Greg Leslie, had organised a wide spectrum in che program, interspersing news on che R& D front with presentations from some of the practitioners in the workplace.
The Keynote Speakers Two from overseas were Dr Sam er Adham from Los Angeles and Prof Simon Judd from Cranfield University, UK, (both their presentations are summarised in chis issue) and Gary Crisp, whose keynote address was a review of desalination throughout inland WA and then a photo tour of the newly commissioned 125 ML/d SWRO plant at Kwinana. W ith the newly developed ceramic rotor energy recovery modules, the energy demand is the lowest co dace . A 225 GWh/a contract between the Water Corporation and the energy company has fac ilitated che building of a wind farm north of Perch. In one slide he co ntrasted che public furore about 'bottled energy' with two facets of public life in WA. One was the beer fridge in che garage, which consumes enough electricity co supply 300400 kL of desalinated sea water per year, the other the typically badly designed MacMansions which need air conditioners running full blast throughout summer, enough co supply desalinated water co eight or nine homes. His paper, outlining the reasons for the decision co install desalination, che site options, and the plane itself will be published in a future issue of Water. H e m ade a fi nal point. .. Australian towns and cities on the coastal fr inge should
look 50 years into che future and reserve sites for possible desalination plants now, before che inexorable spread of development makes chem unavailable, because if climate change proceeds as forecasced, desalination will certainly be necessary co service an increased population.
Research and Development Research and d evelopment is proceeding apace world-wide. Roger Ben Aim of Toulouse University introduced MEDINA, a coalition attacking che problems and challenges of desalination. Tor-Ove Leiknes (Trondheim), Anja Drews (Berlin) and Matthew Braddock (UNSW), introduced the European Union's "MBRNetwork". le is a coalition of membrane experts, fo cusing on municipal applications, which will build on current European expertise in order co foster technology advances, competitiveness, acceptance and application in the EU. le comprises AMEDEUS (in which UNSW is a partner), EUROMBRA (which involves
Table 1. Percentage Exergy Destruction: An industrial waste RO plant. TDS 3000 mg/L. Minimum separation work 1.7 kW Actual 9.8 kW.
Percentage 'wasted' UF pretreatment HP pump RO, stage 1 RO, stage 2 Permeate throttle Brine throttle Mixer
33.7 1B.3 29.4 2.4 2.1 13.9 0.2
UTS), MBR-TRAIN and PURATREAT, each partner specialising in a separate area of research and development and allocated particular 'work packages'. They collaborate via a common website (www.mbr-network. eu). Workshops and conferences are planned, one being che I MSTEC conference in Sydney in November.
Greg Leslie (UNSW) listed three current energy reduction projects.
Exergy is the difference between minimum theoretical work and actual work, and has been recalculated by graduate Robert Kempton. The theoretical work is far more complex than che osmotic pressure demand. Table 1 analyses the exergy destroyed by each component of an RO process, including the pumps and energy recovery, so chat R&D can focus on priorities.
Mixing efficiency. In Singapore two bioreaccors, side by side, are being assessed , one flat place, one hollow fibre. Tracer studies identify stagnant zones, and so reduce aim co the energy required for antifouli ng. These techniques are also being applied at North H ead and Victor H arbour, and will be extended co Aquafin in Belgium and a pilot plant at Anjou.
Nutrient recovery. Nitrogen fertiliser uses 3 -4 kWh/kg in manufacture, plus 6-8 for removal from wastewater by BNR, i.e. 9-12 in coca!. Victoria University is studying use of nanofiltracion co recover divalents and phosphates before che RO scage, then ammonia recovery from che RO brine. The question is what will be the energy demand for chis process? In discussion ic was noted chat supply of phosphate fertil iser could be limiting agriculture in che next century. Discussion also focused on waterless toilets
Quality water from any source Prime Water Technologies delivers novel water solutions across a range of industries: • Conservation • Enhance Regulatory Compliance • Harvesting • Storage and Treatment • Performance Improvement • Sewer Mining • Process Troubleshooting and Analysis
56 MAY 2007
water
Journal of the Australian Water Association
PRIME
Qi WATER
TECHNOLOGIES (02) 8811 2070 www.primewater.com.au enquiry@primewater.com.au
WATE RTE CH NO LOG I ES Leader in Liquid Solid Separation
ReAquaTM MBR Eimco Water Technologies ReAqua MBR's have been successfully used to treat municipal, commercial and industrial waters for discharge and reuse applications. Today, with thousands of installations operating world wide, MBR technology is shaping the way we view wastewater treatment and water conservation in Australia and around the world.
MBR Specialists: - Commercial building - Transportable - Municipal - Industrial
A$ ~ COMPANY
Nww.eimcowater.com.au
~ 02 9645 8000 ...., 02 9645 8001 4 damon.good@glv.com
technical features
and use o f che u ndiluted urine for conservation of N and P.
Tony Fane (UNSW) gave a review of ocher research directions, with UNSW collaborating with Nanyang Technical U niversity and rhe Singapore M embrane Technical Centre. T he role of research is not just optimisation bm innovatio n, leading hopefully to q uantum changes. (Nore, Singapore never refers to 'wastewater' b ut to ' used water').
Modified UF membrane. Hydrophilicity is obtained by UV-grafting active polymer molecules onto the surface of the substrate, but by using electro phoresis a far better produ ce is obtained, with both a more homogeneous pore strucrure and vastly d ecreased contact angle. T he next phase is co embody inorganics such as colloidal silica withi n che polymers.
Hybrid processes. Advanced oxidation by TiO 2 /UV removes trace organics. Newly developed membranes wh ich resist such oxidation may now b e app lied co that reacror. Desalination. A yield of up co 90% can be achieved by second stage of RO fo llowed by recovery of water from the concentrated brine by membrane distillation. Membrane distillation gives go od fl ux even with eigh t ti mes seawater concentration, and che inorganic salts can be crystallised as a solid product. T his is only feasible if was re heat is available to run chat stage at 50°C.
Fouling control. In rhe laboratory better hydrodynam ics have allowed the critical flux point co be increased , and in actual planes the GrahamT ek AC field has increased critical fl ux by 25-50% . The detailed mode of action of the coarse bub bles used co scour membranes has been observed in the labo ratory. The average shear has been correlated with fl ux maintenance and the shear stress vectors in rhe boundary layers has been visualised by PJV. Special spacers have been shown to improve flat plate systems.
A novel MER. Retention of non-degraded organics can be achieved by using a righ ter membrane (NF) but the flux is too low. A provisional patent has been claimed for the use of membrane distillation ro recover the water. As above, this is only feasible if waste hear is availab le and the biological process itself must use thermophilic organisms in ord er to operate at a temperature higher than 50 °C. The prod uct water is exceptionally pure and one can envisage a single-stage waste (used) water reclamation plant for potable reuse.
Integrity. For path ogen removal the integrity of a membrane module is vical, particularly for decentralised plants with no
58
MAY 2007
Water
operators. A provisional paten t has been obtained for a simple reliable moniror based on a novel particle counter.
Stuart Khan and Long Nghiem (Wollongong) are investigating ulcrafiltration of waste (used) water for the retention of three commonly used (and excreted) pharmaceu ticals, carbamazepine, ibuprofen and su lfamechoxazole. They used th ree grades of ultrafilrrarion membranes, both clean and fouled, and found char borh size excl usion and electrostatic interaction are active mechanisms. For rhe loose membrane the retention was significantly affected by pH and ionic composition, wh ich co ntrolled both the zeta potential of the membrane and the polarity of two of rhe compou nds. Stephen Gray et al at Victoria University have been investigating rhe effec t of various constituents of NO M , h ydrophobic and hydrophilic, on membrane fou ling, using surface water from various sources. They fo und chat colloidal NOM leads co rapid fo uling but its removal during backwashing is affected by other components because of the residual adhesion of fatty acids and proteins . Hydrophobic h umic acids p rovide adhesion between the cake bu ild up and hydrophobic membranes. Vigi Vigneswaran at UTS described their research projects linked into E UROMBRA. A recent d evelopment is a semi-pilot MBR operated wirh 10% by volume of sponge cubes in rhe MLSS. The anoxic pores facilitate de-nitrification, bu t the main advantage is the gentle 'sponging' of the membrane combined with che coarse bubble scouring, which has almost doubled th e critical fl ux . Roger Ben Aim (T oulouse) outlined lines of research in France on SWRO, main ly on p retreatment systems co control bio-fouling, caused by marine organisms and their excretion produces. Fufang Zha (Memcor) reviewed advances in fabr ication of che low pressure membranes for microfiltration and ultrafi lrracion, co suit the varied applications. These ranged from clarifi cation of drinking water, where the emphasis was on integrity co en sure pathogen removal and also a small foo tprint. Filtratio n of tertiary wastewater requires low foul ing and tolerance to high suspended solids. For MBR, not only tolerance for the high concenrracion of MLSS but also fibre strength co cope with the oscillatio ns of rhe fibre bundle during coarse bubble cleaning. For prefilrrarion o f seawater for RO, an unexpected factor was the need for abrasion resistance, co cope with rhe silica d iatoms. For industrial
Journal of the Australian Water Association
processes, resistan ce co a range of chemicals. New d evelopments are hydrophobic membranes for membrane d istillation and solvent tolerant membranes for nonaqueous applications . T he aim of rhe manufacturer is co achieve the narrowest range of the required pore size, allied co rhe optimal properties for rhe particular application. H e listed a bewildering array of polymers which have been used fo r membranes, noting chat since 1990, PVD F has the best combination of properties. ECTFE can be used ar high temperatures and can be made hydrophilic or hydrophobic, and there are other polymers being t rialled. Over rhe past ten years there have b een sign ificant advances leading to the cost of installed systems dropping more rhan ren-fold, with further advances in sight.
Michelle Wittholz (Adelaide) summarised their ini tial investigations on Forward Osmosis (see Bolco et al, Water, March 2007} using polypropylene glycol as a draw solution, since on warming by 5°C it spl its into two phases so char it may be recycled. D avid Cook (AWQC) is aiming to recycle the brine used for regeneration of MIEX. UF or NF could concentrare the NOM removed into a much smaller volume fo r disposal if it has co be transported.
Applications Keith Craig, Veolia, gave us more details of the huge Ashkelon SWRO plant in Israel (see Water, J une 2006). Energy consumption has been less than 4 kWh/kl and the p lant has met its design production and quality levels. I nitial problems with organics fo uling have been overcome. Jeff Peeters, GE Zenon, reviewed the steady advances in both chemistry and engineering of the Zeeweed M BR system. Energy demand for scouring the h ollow fibres has been reduced from 1.0 kWh/m 3 in 1994 co 0.1 in 2005, the latter due co rhe evolution of'Eco-aeracion', controlled by varying the aeration frequency in response to solids loading. T he MBR 100 ML/d plant at Pooler, GA, operates on 10 secs on, 30 secs off for about 90% of che rime, then swi cches co 10 sec on, l O msec off as required, e.g. during peak events. Anthony Davey, Earth Tech, described Melbou rne's Eastern Irrigation Scheme. After commissioning, it is now p roducing 26 ML/d at 75% yield. Pressure for UF is provided from the main 350 ML/d outfall pipeline pumps. Polysaccharides are a nuisance, bu t algae in the holdi ng basin have nor presented problems. They achieve 4.8 log removal of coliphage, 6 log for E-coli, with a furth er 0.8 log by chlorine.
tuQQed lntelllCAL™ Probes handle the harshest environment
>luQ and Play with DIQltal Probes without re-calibration
v\ultl-Parameter Flexlblllty cables, with more probes to come ...
>ata Transfer through
use
technical features
Their validation for C lass A accreditation is by turbidity, particle count, water displacement test, 1 bar air pressu re, which is sensitive to 3 u breach (i.e. crypto). The water quality is TDS 480-800, SAR 56.5 which are both suitable for irrigation, bur they are monitoring effect of nutrients on the soil.
John Poon, Melbourne Water, oudined the trials of sale reduction for the treated wastewater (ca 1000 mg/L) at the Western Treatment Plant, for wider recycling to horticulture, industry and third-pipe develo pments. Two pilot scale membrane planes are being assessed u nder contract with their product being blended with undesalced water to provide ca. 500 mg/L for d istribution and he quoted preliminary results.
Russell Taylor, CSIRO , is operating a second-hand EDR plane alo ngside and he quoted similar results for ED R, which is simple, robust and cost-competitive. On the same project David Stork (RMIT) reported thorough investigations on the degree of fouli ng of rhe MF or UF membranes which would be installed ahead of the salt reduction plants, caking water samples fro m three places in the treatment sequence. N umerous parameters were analysed , including fractionation of the NOM. It was fou nd chat after removal of the suspended solids (algae, etc) the final effluent was more suitable than the earlier clari fier overflow d ue to the slow degradation in the lagoons o f foul ants such as polysaccharides and p roteins.
Scott Caothien (Pall Corporation) scanned their experience over the past eight years in using hollow fibre PVDF membranes for a wide range of applications, mainly in the USA. Pall MF membranes are bei ng use at Brisbane's Luggage Point 15 ML/d recycle plant, and also for sewer mining. The efficacy o f their EFM (Enhanced Flux Maintenance) system was demonstrated when the pump tripped. Within rwo days the trans-membrane pressure climbed from a steady 70 kPa to 250 kPa. but rapidly returned to normal when the fault was remedied. Und er EFM protocol, 80% higher flux can be maintained.
Gary Crisp and Mario Slujinsky described the HERO process which has wide application for inland groundwaters where high yield and problems with brine d isposal dominate. In this patented process, hardness ions, which cause scaling, are first removed by ion exchange, CO 2 is degassed, then the pH raised by NaOH so that silica foul ing is obviated. Yields of 90-95% can be attained. There are options for the ionexchange pretreatment system and the
60 MAY 2007
Water
process is being investigated at Yalgoo, in outback WA.
Franz Heindl (HUBER), described their novel tech nologies, the 'Clearbox" fo r decentralised applications (50 PE to 500 PE) and for larger installations they mount the UF membranes on a rotating cylindrical rack which passes a concentrated aeration cleaner zone, thus reducing energy demand.
Con Pelekani (AWQC), with a collaborative ream from a number of agencies, assessed the Graham Tek system which applies an AC electro-magnetic field (EMF) to the membrane module. They worked on a small BWRO plane in Nhill, in drought-stricken rural Victoria. The feed was 1200 mg/L TDS, 350 mg/1 hardness, 30-40 mg/L dissolved silica. Although the current results are inconclusive it would seem that the EMF has a definite effect on silica fouling, bur lirrle or no effect on calcium sulfate. The best runs were a hybrid of the EMF with a reduced dose of chemical anriscalanr. However, he noted char EMF operates very successfully in Singapore and so may be specifi c for a certain range of water quality.
Jean Schrotter (Veolia) extolled the two decades of experience char Veolia has developed in designing and building, and sometimes operating, several hundreds of membrane planes world-wide. T o back up the ream they set up ARAMIS in 2003, the Anj ou Research Advanced Membrane Investigation Service. H e gave examples of the various techniques they are using, ranging from characterisation of membranes, e.g. by v irus challenge, to chromatographic identification of potential organic fo ulanrs and to autopsy of a fou led membrane from an operating plane, in order to specify, for specific cases, the required d egree of pretreatment and chemical cleaning regime.
Carolyn Madden (South East Water). A demonstratio n MBR plant, being operated on greywater from a 237 u nit apartment cluster to provide toilet flush and garden irrigation, is being used to monitor the reliability of pathogen reduction. Since it uses Kubota flat place membranes, integrity cannot be reseed by air pressure, and a combination of turbidity and particle counters has been checked against microbial testing. Validation for National Guidelines must hinge on worst case, nor average values, and a turbidity limit of 0. 1 NTU has been correlated with 2-log removal. The investigation has gathered useful data o n greywater co nscimcion.
on the biological plant upsets caused by lack of sufficient alkalinity and varying salinity, both of which can enhance foaming microorganisms which increase the fo uling race of the membranes. This can be corrected by dosing alkali but Eimco has perfected a method of increasing denitrification controlled by a DO p robe in a separated membrane zone. In ocher words, there's more to MBR than the membrane technology.
Julian Briggs (CH 2MHill), reported on performance at Sydney's North Head MBR, now operating smoothly with about o ne hou r per week attendance. Estimated opex cost fo r 2 ML/dis 16c/m3, wh ich does nor yet allow for membrane replacement.
Brenton Gibbs (Tenix) compared the two largest M BR's in Australia to dace, Joe White M altings in Perth and Victor Harbour in SA. Boch planes have featured in recent issues of Water. Craig Timms, EcoNova, described their fu lly automated MBR systems which they are offering fo r either individual or cluster domestic installation. As above, maintenance visits are limited co one per week fo r installations such as Noosa North Shore (200 kL/d).
Upcoming Plants Andrew Hodgkinson (Parsons Brinckerhof) reviewed the very derailed d esign optio ns for the G ippsland Water Factory which will recycle 20-30 ML/d of puri fied domestic and industrial sewage to Australian Paper, thus releasing potable water but also reducing the load, both hydraulic and aesthetic, on the open sewer. Afrer anaerobic and aerobic MB R the effluents will be desal inated by RO. Because of the high recovery percentage there is a danger that the final RO stage will permanently fou l, so a 'canary' membrane will be operated.
John Messenger (SKM) detailed the rational design features for the 18 M L/d MBR being installed in North Cairns, with emphasis on optimal energy use.
A WORKSHOP was run, mainly
Stephen Kennedy (Eimco Water
discussing a new parameter for p redicting fouli ng. T he current SDI system is practically useless since it is based on inorganic cake formation whereas it is the blocking by organics which is posing rhe problem. Discussion raged on the differe nce berween dead-end fi ltration and crossflow instru ments. The outcome was beyond the competence of this reporter to report with any d egree of accuracy!
T echnologies) have 80 Kubota membrane planes in UK and Ireland and 140 worldwide. His presentation concentrated mainly
Membrane tech nology confe rences in 2007 are being run in H arrogate, U K, in May and I MSTEC, Sydney in November.
Journal of the Australian Water Association
V
0() 0~ QC/ .,
~ b'
"
0"'~
p
.,
(:,
0 u
technical features
membrane technology
DEVELOPMENT AND FUTURE OF MEMBRANE TECHNOLOGY S Adham MF/UF - Why Choose them over Conventional Filtration? • Regulations for product water and effluents are increasingly more stringent • UV disinfection is increasingly popular, requiring better transmissivity • Alternative water sources require better treatment Bue the main advance is chat as the technology has matured, wich both technological advancements and competition in the market place, che costs have dramatically reduced. Membranes were first used for microbiological analysis in the 1950s, then in brewing industry in the 1960s, but the first interest for water purification originated in the 1980s, only after backwashing techniques had been developed by Memcor in Australia and Aquasource in France. Since chen there has been dramatic development. O ut of a total of 657 installations in 2006 there were 321 in America, 189 in Europe, 97 in Asia (mainly Japan) and 47 in Australia. Capacity has increased accord ingly as shown in Figure 1. As shown in Figure 2, installation of che low pressure submerged system has increased faster than the pressurised system, particularly for larger installations. A survey by MWH of 40 owner/operators has shown that their main reasons stated fo r memb rane replacement were: integrity loss, 25: productivity loss, 12: age, 4.
3500
§
3000
:E
;!' 2500
·c:; co
g-
(.)
2000
C:
·~
1500
II)
C ~
1000
'i
:i
E
500
:I
(.)
0
Figure 1. Cumulative capacity of microfiltration plants world-wide.
The AWWA Research Foundation has a program aiming to optimise pretreatment. Membrane flux and TOC removal by MF is greatly enhanced by pre-coagulation, with up to 25% removal possible of TOC using a dose of 5 mg/L of polyacrylic, and can be enhanced even further by a dose of powdered activated carbon. Nonetheless, at this stage of development, a pilot investigation is a wise investment. MF/UF produces water wh ich is suitable fo r industry, etc, but fo llowed by RO and advanced oxidation (AOT) to destroy low
molecular weight organics. Advanced Water Treatment (AWT), it is being considered for indirect potable reuse. One of the options being explored for San D iego is to pass a proportion of the reclaimed water through AWT, then to the San Vicente
Seawater desalination is here to stay, so long as people prefer to live in coastal communities.
They also asked if the membranes had been integrated into an existing WTP. Out of 80 respondents, 33 said yes and 47 were standalone. Use of membrane fi ltration for removal of pathogens is a function of che pore size. For example, MF, with pore size of0.l to 0.2 micron, gives about 1 to 2 log removal of virus, whereas UF, with pore size of 300 kD can achieve 4 log and a tighter membrane, with 100 kD pores, can achieve 6-log. This is an edited version of the keynote address to the AWA Specialist Conference on Membrane Technology, February, 2007.
62
MAY 2007
Water
Figure 2. System Configuration. Design capa city. (1 M GD = ca 4 M L/d] .
Journal of the Australian Water Association
technical features
Reservoir with two years detention rime, where ic is blended with the normal supply (of lower quality) before being w ithdrawn through the Alvarado Filtration Plant. A pilot facility is in operatio n, as sh own in Figure 3 . His presen tation includes a table of the degree of removal of some 29 ED Cs and PPPCs. After rh e RO stage, only nine are detectable in ca 1 nanogram/ L concentratio n, and after AOT, all are lower than their detection limit. This is consistent with results from the 2 50 ML/d Orange County plant, W est Basin and Singapore.
Estimated Costs Fo r a USA installation, with electricity at I 0c/ kWh , membrane replacement at 7 years, and 6 % on capital over 20 years, an estimate(+/- 30%) for co sts of such water is given in T able I.
Advanced Water Treatment research facility at the North City Water Reclamation Plant (NCWRP) Figure 3. San Diego Indirect Potable Reuse pilot testing.
But What of the Future? There will undoubtedly be new membranes, innovative p rocess streams and the benefits of large scale applications. The challenges to manufacturers of polymeric membranes are to reduce the pore size range, increase permeability, increase resistance to cleaning chemi cals and increase membrane life.
Wetor colloctlon cell
T here is an opt ion already in service, the ceramic membrane. Figure 4 is a cutaway diagram o f su ch a membrane. M ade from accurately graded granules, pore size distribution is much narrower. Table 2 is a co mparison of the likely advantages of such membranes . The disadvantage is their current capital cost, but a large plant is already in operatio n in Japan , as shown in Figure 5. H e drew attention to a Mitsubishi innovation, M BfR, which uses a sp ecially developed hydrophobic tubular memb rane. H ydrogen gas permeates from the inside, through the pores, and sustains an anaerobic biofilm which can reduce nitrate, p erchlorate and other chemicals often present in groundwater. (Readers are referred to a similar but contrasting Australian development where air or oxygen sustains an oxidising biofi lm. Taylor, ANSTO, 2 006).
Figure 4. Design of ceramic membrane. Adham finally disp layed photographs of the rwo 400 ML/d plants now being co mmissio ned, the T win O aks plant in California and the Lake View plant in Ontario, both by Zeno n, dealing w ith
Table 2. Comparison of anticipated performance of ceramic membranes. Polymeric
Ceramic
30-60 gfd {50 - 100 lmh)
100-150 gfd (165 - 250 lmh)
85-95%
98%
15-30 min
1-4 hr
Low-Moderate
High
Clean-In-Place
1 per month
1 per year
Maintenance Clean
1 per week
1 per month
Life
5-10 years
20-25 years
Flux Recovery
Table 1. Estimate of cost of 3-stage AWT. 20 ML/ d 100 ML/ d Capital, total (M. USO) Membrane system (M. USO) 0 and M (c/ kl) Water cost (c/ kl)
8.6 3.0 6.9 18
28.2 8.0 4.4 12
co ntaminated surface waters. So the future of MF/UF is certain bur even so , greater effons are n eeded to reduce energy and chemical usage in response to enviro nmental drivers.
Backwash Interval Chemical Resistance
Journal of the Australian Water Association
Water
MAY 2007 63
Seawater Desalination Desalinat ion by Reverse Osmosis has been operating for 40 years but fo r every proposed new installation there are significant challenges. From the outset there will be regulatory and permitting issues, and che d ifficulty of find ing a suitable sire adjacent to the coast, where development of residential as well as industrial properties has reduced the availability and raised the cost of land. Technical issues relate to che intake and outfall structures (no small component of che capita l cost) and to the choice of pretreatment to reduce fo uling of the membranes. The required energy is significant, bur nor in relat ion to ocher forms of energy consumption in a modern community. H e reviewed overall coses, which have declined dramatically from USD 2.00/kL in 1990, to USD 0.50/kL in 2006 . However, he cautioned chat this latter figure cannot be taken as a firm guide to che future. Membrane costs may be fa lling, but every other cost is soaring. Oi l prices have risen but so have the p rices of steel, ski lled labour and overall construction costs. Capital cost estimates are heavily dependent on external factors, but a survey by MWH shows a steady decline from USD 2 .00 per kL design capacity in 1990 to USD 0.50 in 200 3, but then a steady increase to about USD 0.9 in 2006.
Figure 5. 4 0 ML/d ceramic membrane pla nt in Japan (2006). Repair & Maintenance (2%) Membrane Replacement (5%) Chemicals (8%)
Labor (10%)
Construction Cost (52%)
Figure 6 is a pie- chart of no rmal coses, showing the dom inance of capex on the cost of water production . Figure 7 is perhaps a bette r estimate of current cost. Although in any coastal situation ocher opt ions for supply of water may be less expensive at p resent, there will be a limit to their availability, so char a diverse portfolio is prudential. H e noted char despite the success of Singapore's NEWater recycl ing planes, they have still built a 120 ML/d SWRO plane at Tuas to further guarantee supply. One of the advantages of a SWRO plant is chat, provided the site is established, the plane can be built, commissioned and in production within cwo years, as evidenced bot h in Singapore and Perch.
100,000 m3 /day Plant Figure 6. Impact of co nstruction cost o n water cost fo r a l 00 ML/d pla nt.
For SWRO, pretreatment is absolutely essential. The most common system currencly in service is coagulation and mixed media filter, fo llowed by cartridge filters as a precautionary measure to guard the membrane stacks. The alternative of MF/UF is being cons idered, since it has some advantages in prolonging the life of the RO membrane. H e noted that seawater is by no means an easy source to deal with. I r is alive with organisms, both microscopic and visible. Red tides (algae) are sporadic, silt levels can escalate dramatically after storms and the attack of barnacles and clam shells on submerged structures is well known, bur they can also proliferate on intake pipes. (In one plane it has been found chat clams can learn to 'clam down' to survive a periodic dose of chlorine. However, they can be fooled by erratic dosing).
What of the Future? There will be higher performance membranes, particularly with che application o f nanotechnology, energy recovery from the reject scream is already efficient bur alternative energy systems may be developed. He quoted Forward Osmosis as a possibility, with the Water Reuse Fo undatio n investigating using as the draw solution such high tech developments as dendrimers, which because of their charge structure have the requisite high osmotic pressure despite their high molecular weight. They can
64
MAY 2007
Water
Journal of the Australian Water Association
Figure 7. Red uction of cost with size: Capacity m3 / d.
be separated from the produce water by cheaper UF and recycled . Seawater desalination is here to stay, so long as people prefer to live in coastal communities.
The Author
Dr Somer Adham is Vice Presi dent of MWH, California, and Manager of the Applied Research Department. He has been involved with San Diego's Water Reuse P rogram since 1994.
technical features
membrane technology
MBRS: HOW MUCH DO THEY COST? S Judd Abstract Membrane bio-reactors are more expensive to operate than conventional activated sludge plants primarily because of the coarse bubble aeration which is required to flus h the membranes. Membrane maintenance is a facto r, bur in rhe competitive market which has developed, this is not as critical as in previous years, since membrane prices have fallen from 89 c/m2 in 1992 to ca. 7-8 c/m 2 in 2006. Even so, their application to municipal waste treatment is steadily increasing, fo r a number of practical reasons.
Systems Available Membrane bioreactors were initially developed some 20 years ago to deal with concentrated ind ustrial wastes, often where 'footprint' in a restricted ind ustrial site was an important issue. In the past 10 years, there has been an increas ing momentum to their application to municipal wastes, fo r a number of practical reasons, even though opex is typically 50-80% higher than a conventional ASP. T here are currently two forma ts of immersed MBR offered, one where the banks of membranes are immersed in the actual biological aeration tank, the other where, mainly for ease of maintenance, a sidemeam of MLSS is cycled to a separate membrane tank. There is little difference between the two in terms of the major O PEX which is the coarse bubble aeration required to minimise fouling of the membranes. The actual membranes can be either Flat Sheet (FS) which are considered to be simpler to design and operate, or H ollow Fibre (H F) bundles, which are usually cheaper. A recent development by one manufacturer is the multi-rube system (MT) . There is no doubt that the technology has developed rapidly. Figure 1 provided by one manufacturer, shows graphically how operation and maintenance costs have been progressively reduced over thirteen years, mainly d ue to the decreased price and increased reliability of the membranes themselves. Currently rhe major cost is the power required for aeration, which is split T his is an edited version of the keynote address to the AWA Specialist Conference on Membrane Technology, February, 2007.
CJ PDYi-er Millnlenance
C Chemicals
C Membrane replacemert • Screenings c SIJdge disposal
• Rent and rates
8640 m3/d plant at FFT
Figure 1. Decline in costs since 19 9 2. (Kennedy & Churchill, 2005). OlherEqup. 1.11%
between the fi ne bubble aerators which supply the oxygen to the biomass, which cannot be reduced, and the coarse bubble aeration which is needed to clean the membrane surfaces. The coarse bubbles do supply a proportion of the oxygen demand, but only at low efficiency.
Process Pumps 5.37%
Process Air Blowers Membrane
34.68%
Aeration Blowers 38.17%
Mixers_,,
Power Demand
4.42%
Recirruationj
Figure 2 summarises another manufacturer's estimates for rhe power demand of a typical plant. Compared to the 35% used for aeration, the coarse bubble air consumes some 38% of the power demand, since the focus must be on a high rise velocity to accomplish the necessary scouring of the membrane surface..
Pumps 15.84%
Figure 2. Estimate of Power Demand. Courtesy of GE Zeno n International.
They cost more but have sign,ificant advantages in certain situations.
Design of MBRs is therefore, to a large extent, focused on reducing the coarse bubble power demand, without losing too much membrane flux between clea ns, thus optimisi ng between OPEX and CAPEX. 45.00
•
40.00
y = 544.17x + 12.928 2 R = 0.9994
,
-- 35.00
i
30.00
0
I • •
N
s
--, 25.00
y = 391.71x + 11.783 2
R
=
0.8987 • Three-monthly chemical clean, FS • Monthly chemical clean, FS
20.00
• HF,
15.00 0.020
0.030
0.040
0.050
0.060
0.070
U..,,
0.080
0.090
0.100
0. 110
0.120
U (m s·')
Figure 3. Correlation of aeration intensity w ith critical flux (G ug lielm i, 2007). Journal of the Australian Water Association
Water
MAY 2007 65
technical features
Gugliemi (private comm.) at the MBR plan t in Brescia, Italy, h as studied the effe ct of aeration intensity on the critical membrane flux, as shown in Figure 3. Aeration intensity (U, mis) is defi ned as the in-module air fl ow velocity and it co rrelates roughly with sh ear. Jcrit is the maximum flux , litres per m 2 per h, before che membrane scares co block (normalised co 20°C). T he difference between the effect of a monthly chemical clean and a three month ly clean is obvious, bu r there is a maximum U beyond wh ich there is no further improvement, and che plane at Brescia was optimised for chis point. The coloured points are for the Flat Sheers used at Brescia, bur the same effect applies co Hollow Fib res if o ne uses a net flow velocity for U. Considering the geometry only the Specifi c Aeration Demand (SAD) fo r fib res and flat sheet respectively will be:
(i-
SAD = 3.6xl06U P 4LJ ¢
d ) r
6
SAD = 3.6xl0 U( hw) P JL y
packing density (m 2/ m3)
fibre diameter w, y = module, sheet width (m)
.,,C
"'
D
70
...
•
E Q) -c 50 C 0 ·;:,
~
<> Hube r • Puron o Toray
D
60
...
...
40
D
"'u
30
~
20
-= Ti
•
~
(/)
10
:o
300 m-1
y = 0 .95 w h = 0 .008 m m
then the correlation coefficients fo r HF and FS modules are 0 .0018 and 0.0088 respectively. If J relates primarily co U then
0
•
D
B
0
<> •• ••
• •
0 0
5
10
25
20
15
30
35
40
Flux, LMH
Figure 4. Data on specific aeration demand from some manufacturers. (Hollow fibre supplied by Zenon, M Rayon, Puron . Flat sheet supplied by Toray, Kubota, Huber). FS can never b e economically competitive with H F o n basis of specific aeration demand. H owever, full-scale data refutes the theory. Figure 4 compares data from some of the manufacturers for various planes, showing char there is no significant difference between the two types.
• Aeration is nor solely used for surface sco uring bur also co promote membrane agitation for fibres and co suppress channel clogging.
L = 0.95m (FS), 1.8m (HF)
D
Q)
• Stacking of flat sh eers one above the ocher reduces SAD by 20-25% .
Taking appropriate values:
df= 0 .002
(/)
• J crir is not unambiguously d efi ned
h = sheet separation
=
d
<t
• Membrane aeration in practice is maintained at the same rate regardless of flux , which accounts for the high SAD at low fluxes.
dr=
0
• Zenon o Kubota A M Rayon
80
The reasons are probably char:
where L = module length (m) 0 =
90
Obviously, when it comes co keeping the membranes clean, spacing between adjacent membranes is critical. Too close and there is no room for material co be sloughed off, coo wide, and the effect of the bubbles is
d issipated. Figure 5 reports data from examples of each type, showing char the separation of membrane from membrane is less in the HF plants, enabling rhe modules co be smaller. The majority of HF plants have a spacing of about 3 mm.
Recent Advances
If we could use finer b ubbles we could improve oxygen transfer co do the biological work., bur the result would b e poor scouring efficien cy. The recent concept of 'jet aeratio n' from Memcor seems promising, but we need two years of trial co confirm its long-term efficacy. As noted above, FS modules can be stacked two-high so char aeration is more efficient, but at the cost of extra height, blower pressure and CAPEX. Nore chat the aerato r manifo ld at the base is subject co clogging by detritus so consideration muse be given co cleaning it. Cleaning frequency by backflush is normally 10% downtime, bur chemical cleaning is only about 2% downtime, so not of great significance, bur dealing with
r=-:l
w~ c..=.J
PCM
~
Scum, Sludge & other Unmentionables
Tel. (02) 6581 0744
66
MAY 2007
Water
Fax. (02) 6581 0790
Journal of the Australian Water Association
PUMP!Ji@D!1DVW I
www.pumpability.com
mail@pumpability.com
0.012
+ HF
a FS
0.01
• MT
f g>
0.006
"
:2
t....
0.004
.
0.002
0 0
•
•
0.008
•
• •
• ••
• •• 2
4
6
8
10
12
Membrane separation, mm
Figure 5. Relation between packing density and membrane separatio n. (Hollow Fibre, Flat Sheet, Multitube) (Judd, 2006). rhe chemical waste may be a problem. Solids residence rime (SRT) is normally 15-30 days, with low F/ .M ratio and maintaining 8-10 MLSS fo r improved aeratio n efficiency. In the US there is a trend to lower SRT of 5-10 days to reduce plant footprint and increase aeration efficiency further, bur rhis requires on-sire WAS processing since more sludge is generated.
Summary: Overall Costs In general, OPEX is 40-80% more than fo r conventional ASP on the basis of process costs alone, since MBR requires so me 50% more power and requires membrane replacement, despite claims for 10 year life. The waste sludge from MBR is less due to longer SRT bur evidence suggests rhar ir is also more di ffic ul t to dewarer. Intermittent aeration is proving advantageous bur involves increased
Water Advertising To reach the decision-makers in the water field, you should consider advertising in Water Journal, the official journal of Australian Water Association. For information on advertising rates, please contact Brian Rault at Hallmark Editions, Tel (03) 8534 5000 or email brault.rault@halledit.com.au
freq uency of valve operation. For example a cycle of 10 sec on, 10 sec off requires 3 million operations a year, which must increase risk of process failure - if only marginally. However, there are distinct benefits in applying MBR for particular situations. For example, there are no problems with bulking sludge, the CAPEX and footpri nt of conventional clarifiers can be significant, and the ability ro deal with shock loads is useful. However, rhe dominant factor is the better quality of rhe permeate, and rhis is of the utmost value when re-use is being considered, particularly if rhe oxidation plant is to be followed by advanced systems.
Acknowledgments For the data in Figure 3 I am indebted ro Dr Giuseppe Guglielmi, a Research O fficer in rhe Department of Civil and Environmental Engineering at the University of Trento.
References Guglielmi, G (2007) Pers. Comm . Kennedy, S and Churchouse, S.J (2005) Progress in membrane bioreacrors: new advances. Proc. of Water and Wastewater Europe Conference, Milan, June 2005. Judd S (2006) . The MBR Book (Elsevier).
The Author Professor Simon Judd is Head of the Centre for Water Sciences and occupies the Chair in Membrane Technology at Cranfield University. Email S.J.Judd@cranfield.ac.uk.
technical features
MEMBRANE BIOREACTORS IN AUSTRALIA FORECAST IS FOR GROWTH S Chapman Abstract Membrane Bioreactor (MBR) Technology is no longer viewed as an emerging technology in the Australian marker place. Client organisations are now prepared to view MBR as an established candidate tech nology for comparison with ocher options to meet their needs. Of the (approximately) 20 Australian MBR reference projects, 7 of the domestic wastewater examples have been selected for discussion as brief case studies within chis paper. The paper covers some "lessons learned" as well as addressing the question of "why was MBR selected for this particular sire". Although not a panacea for all wastewater challenges, M BR does have its place - and the recent worldwide proliferatio n of MBR is testimony to the fact that it is cost competitive and worthy of consideration, when there are requirements for high levels of treatment.
Figure 1. Various MBR Membrane Modules.
Introduction Three years ago C hapman, Leslie and Law (2004) discussed the suitability of MBR for reuse applications in Australia, highlighting char local considerations may play a part in the degree of uptake of chis technology. For some rime now this innovative technology has been considered well established in places such as North America and pans of Europe, but there had been relatively few Australian reference sires. Now, it seems, there is critical mass in terms of local reference sires and local knowledge and there is a sense that MBR is here to stay.
operational data. Ocher available variants (which to date have been less prevalent) also include Mitsubishi, Norit, H uber, Asahi and Toray to name bur a few.
Information sharing amongst Australian water authorities, international consultan ts and suppliers of the technology during the past few years has now reached a point
where most people in the industry have a working knowledge of M BR technology. Client organisations are now prepared to view MBR as an established candidate technology for compariso n with other options to meet their needs. Meanwhile, during this same rime period, the water industry overseas has continued to see a proliferation of MBR reference sites redefini ng what constitutes a "large scale MBR". For example, rhe Bei Xiaohe facility in China, with an ADWF of 60 ML/d, and rhe Brightwarer facil ity in the USA, with an ADWF of 148 ML/d, are indicative of the scale of rhe most recent MBR projects awarded overseas.
A survey ofMBR projects, mainly municipal, either operating or projected in the near future.
In the period 2005/2006 there were many new MBR faci lities commissioned in Australia, which has helped build local trust in rhe technology. I n fact Memcor, Kubota, Zenon and Puron are all represented in this new sui te of Australian MBR reference sires, and now have a significant amount of
By rhe end of2006 the collective treatment capacity (in terms of ADWF) of all MBR facilities installed in Australia totalled approximately 10 ML/d. As shown by Figure 2 and derailed in Table 1 to follow, the collective treatment capacity will be an order of magnitude higher by 2009.
68
MAY 2007
Water
Journal of the Australian Water Association
Register of Australian MBR Facilities The early installations such as Picnic Bay ST P, North Head RWP, Victor Harbor STP, Cobargo STP and Brooklyn STP , have been viewed as the projects responsible for establishing the technology in Australia. However, compared with these early installations, the next suite of MBR projects will be remembered as the ones that were able to demonstrate the technology on a "large scale" .
technical features
After accounting for rhe commissioning of the Gippsland Water Factory (35 ML/d), Cleveland Bay (29 ML/d), the Northern Treatment Plant in Cairns (19 ML/d) and several others, the collective treatment capacity (in terms of ADWF) will be approximately I 00 ML/d. Table l was derived from vendor info rmati on, updated in early 2007. The data included in rhe analysis is nor a complete picture of rhe MBR marker in Australia, due to the exclusion of some very small facilities and due to rhis paper's foc us being on domestic wasrewarer (as opposed to industrial applications) .
100 90
..,
80
ML/d
70
~
60
.....
50 40
.,,
30
1
20 10 0
.. ..
2003
2004
~ ~
2005
2006
Case Studies O f rhe Australian MBR reference sires, seven of the domestic wasrewarer examples have been selected for inclusion here as brief case studies. For rhe operational plants, this sectio n of the paper covers some "lessons learn ed" as well as address ing rhe question of "why was MBR selected fo r rhis particular sire".
Picnic Bay (Magnetic Island), Queensland Commissioned in October 2002, Picnic Bay was the first ' full-scale' or permanent membrane bioreactor facility in Australia. (De Haas et al, 2004, Henle et al 2006, Young D , 2006) . Although expandable, the plant was buil t with an initial capacity of 0.55 ML/d. Aquatec-Maxcon was awarded the prime contract, utilising Kubota flat sheet membranes. T he membranes and associated process expertise were provided via the then Aquator MBR T echnology, UK. Some of the p roject drivers or facto rs leading co rhe selection of this technology incl uded rhe benefits of modular expansion and the need to produce a very high quality of rreared water fo r re-use and for discharge to the environmentally sensitive area in the world heri tage protected Great Barrier Reef).
Lessons learned The treatment plant experienced low influent BOD and COD concentrations compared with design values, which was attributed to rhe warm temperatures and long residence time in rhe reticulation system. This has lead co an increased
2007
2008
Figure 2. Collective MBR Treatment Capacity (in terms of ADWF).
requirement for supplemental carbon dosing. T he issue is one of T N removal within the biological system and is not MBR specific. Highly intermittent flows have mean t that the balance rank has been viewed as an invaluable inclusion.
Cobargo STP (Bega), NSW
Operational summary T he plant has been in continuous operation since Ocrober 2002 wi thout membrane replacement, and with excellent water quality perfo rmance.
Commissioned in May 2006, Cobargo STP was the first of 4 MBR plants for rhe Bega Valley Sewerage Program Alliance.
Table 1. Register of Australian MBR Facilities. Name of Plant, Location
Picnic Bay, Qld Couron Cove, Qld North Head RWP, NSW Victor Harbor, SA Joe White, WA Cobarga STP, NSW Durham Green, NSW The Urban Workshop, Vic Southern Cross, Vic Horseshoe Bay, Qld Kolaru STP, NSW Wolumla STP, NSW Candelo STP, NSW Brooklyn STP, NSW Bangalow, NSW Cleveland Bay, Qld CBW, Vic Gippsland Water Factory, Vic Burpengary East, Qld Cairns (Northern Plant), Qld
Type
Start-up Date
Average Flow (ML/ d)
Peak Inst. Flow (L/s)
Kubota Zenon Memcor Kubota Puron Puron Zenon Memcor Memcor Kubota Puron Puron Puron Zenon Zenon Zenon Memcor Memcor Memcor Zenon
Oct-02 Jun-05 Aug-05 Dec-05 Feb-06 May-06 Jun-06 Jul-06 Sep-06 Sep-06 Mar-07 Apr-07 May-07 Jul-07 Sep-07
0.6 0.3 2 5.2 1.4 0.18 0. 1 0.1 0. 1 0.7 0.18 0.18 0. 18 0.6 0.85 29 0.1 35 3 19.4
19 6 23 130 25 3 2 9 8 40 3 3 3 19 30 870 7 463 70 674
Jun-08 Dec-08 Dec-08 Nov-09
â&#x20AC;˘ Information gratefully received from Zenon, Memcor, Aquatec Maxcon, Koch and Tenix.
Journal of the Australian Water Association
Water
MAY 2007 69
technical features
The Cobargo STP design incorporates Puron membranes, supplied by Koch membrane systems. T he plant has been designed with an average flow of 0 .18 ML/d and is connected to a low pressure sewer system, which removes the usual impact of wet weather peaks. Some of the project drivers o r factors leading to the selection of chis technology included the afore-mentioned synergy with low pressure sewers, and the need to produce a very high quality of treated water for re-use, including irrigation of recreational areas.
North Head Recycled Water Plant (Sydney), NSW
bioreaccor. Overflow baffles were also provided between compartments in the bioreactor to allow movement of foam co the WAS withdrawal point. However, during the early months there were several occasions where a chick layer of foam developed. When chis occurred it tended to hinder the flow of WAS th rough the bell mouth, which in cum prevented the required wasting volume from being withdrawn . This caused the foam to worsen, magnifying the problem. In order to overcome this problem a recycled water spray was installed directly above the WAS bellmouch. The p roblem has not reoccurred si nce the installation of the spray system.
Operational summary The plant has been in continuous operation since August 2005 without membrane replacement, and with excellent water quality performance. The plane has proven itself to be robust and requi res only minimal operations and maintenance attention. Commissioned in August 2005, North Head RWP was the first 'full-scale' membrane bioreactor facility in Australia to use the locally manufactured Memcor membranes. (Chapman and Gabriel, 2006) With an initial capacity of 2 ML/d (expandable to 9 ML/d), the plane was the largest MBR in Australia at the time of commissioning. CHJY Freshwater JV was awarded che prime contract, utilising Memcor membranes. Some of che project drivers or factors leading to the selection of chis technology included the need to produce a very high quality of created water for re-use and the need for modular expansion. H owever, of critical importance was the need to minimise plant footprint to reduce the flora/fauna impact on che adjacent National Park.
Lessons learned - seeding Despite che careful seed selection process, the imported seed sludge contained large quantities of coarse material which quickly fouled the 3 mm fixed basket strainers installed on the mixed liquor recycle scream, just upstream of the membrane zone . Consideration should be given to screeni ng of the seed sludge as it is introduced to che M BR.
Victor Harbor STP, South Australia Victor Harbor MBR wastewater treatment plant was tendered as a 20 year BOOT contract. Tenix Alliance designed, constructed and commissioned the plant for United Utilities Australia (UUA) who own the facility and will operate it for 20 years under a contract with SA Water. T he MBR was commissioned in late 2005. The plant design is for an average flow of 5.2 ML/day (although, until sewage flows increase in future years, initial membrane installation was for 3.4 ML/d) and a design peak flow of 11 .2 ML/day. The plant utilises the Kubota flat sheet membrane.(Newland, 2006) Treated wastewater from the plant is p umped to the recently refurbished Hindmarsh Valley Reservoir and from there is reticulated for irrigation re-use by neighbouring agriculturalists. However, if the reservoir is full the treated wastewater will be d ischarged to the sensitive receiving waters of the Inman River.
Brooklyn STP (Sydney), NSW
Lessons learned - foam The WAS withdrawal system is often designed such chat filamentous bacteria associated with foaming are preferentially removed. This was considered in the design of the North Head RWP and wasting was set up to be drawn from a bellmouth just below the liquid level at the end of the
70
MAY 2007
Water
To be commissioned in mid-2007, Brooklyn STP will be a furt her development in Sydney Water's application
Journal of the Australian Water Association
of MBR technology. (Kitching and Bruzzone, 2006) Following pilot trials at Rouse Hill and a sidestream "constant flow" system at North H ead (utilising a small subset of North Head STP flows to meet recycled water demand), the Brooklyn STP will be Sydney Water's fi rst MBR to have a live connected STP treating 100% of incoming flows. Th e combination of using a four-s tage Bardenpho configuration coupled with a Zenon membrane system and a low pressure sewer system, is presently unique in A ustralia. The project is being delivered under an alliance contract involving Sydney Water, MWH Australia, John Holland, UGI and Manidis Roberts. Some of the project drivers, or facto rs leading to the selection of this technology, include the cost reductions associated with an MBR's small footp rint (because piling is required, and there's a requiremen t for odour covers for every square meter of the plant), as well as a need for created water q uali cy suitable for discharge to sensi rive waters and potential fo r re-use if/where appropriate. The combination of a low pressure sewer system and MBR has an advantage in terms of cost-effectiveness of treatmen t (d ue to avoidance of high wet weather peak flows).
Cleveland Bay STP,Townsville To be commissioned within 2007, the Cleveland Bay STP will be che largest M BR in Australia in terms of peak flows being treated. The plant capacity is 29ML/d (ADWF) with peak flow capacity of 75ML/d (870 Lis). This fast moving project for Ciciwacer in Townsville is being delivered under the Water Matters Alliance (Citiwater, Abigroup, GHD, UGI and Aquatec-Maxcon) . Zenon was selected as th e supplier of membranes for this project. Some of the project drivers or facto rs leading to the selectio n of this technology included the need for small footprint (because piling is required on chis site) and the need for created water quality suitable for discharge and re-use (industry, parks, gardens and open space re-use).
Gippsland Water Factory (Latrobe Valley), Victoria To be commissioned in 2008, the Gippsland Water Factory is a project of particular significance to the state of Victoria. In terms of average flow rate, it will be the largest MBR in Australia (and one of the larger examples anywhere in the world) at 35 ML/d. The Gippsland Water Factory incorporates treatment of domestic wastes and wastes from a pulp and paper mill. The project is being delivered under an alliance contract involving Parsons
Brinckerhoff, CH2MHILL, Transfield Services, and with Gippsland Water as the principal spo nsor. Membranes were procured via a competitive render process wich Memcor being selected as che successful vendor. Some of che project drivers, or factors leading co che selection of chis technology, included che need co produce a very high quality of created water for reuse and there were also co ncerns regarding the seccleabilicy of a mixed liquor of industrial origins. Value analysis of che smaller civil scrucrures generally required for MBR planes and che associated civil construction coses also contributed co che selection of MBR at chis sire.
Discussion and Conclusions As rhe case study information indicates, M BR is a strong candidate technology when there are requ irements for a high quality effluent, coupled with ocher drivers. The ocher drivers may include a need for minimal plant footp rint (due co there being a space constrained site or si mply reaping benefits from reduced pili ng coses, civil coses etc). Each rime MBR is being considered for selection, however, there needs co be an identified means of dealing with the challenge of peak wee weather flows (via balancing, bypassi ng or avo iding such peaks by utilising low pressure sewer systems). Although not a panacea for all wastewater challenges, MBR does have its place - and the recent worldwide proliferati on of MBR is testimony co rhe face rhac it is cost competitive and worthy of co nsideration, when there are requi rements for high levels of treatment. With refe rence co Figure 2, there will be an impressive ten-fo ld increase in Australia's collective MBR treatment capacity over the co urse of rhe next cwo years alone. Based on the increasing awareness in Australia for the need for re-use and based on the continued support for ch is tech nology overseas, ic is predicted char there wi ll be co ntinued growth in the Australian MBR market.
References Chapman S, Leslie G, Law I (2004) Membrane Bioreaccors (MBR) for Municipal Wastewater Treatment - An Australian Perspective, Water 31, No 2. Chapman S, Gabriel I (2006) A Report on Membrane Bioreactor (M BR) Performance - North H ead Recycled Water Plane, Enviro 06 Con ference. De Haas D , Hertle C, Turi P (2004), Magnetic Island Water Reclamation Plant - Membrane Bioreactor Nutrient Removal Technology One Year On, Enviro 04 C onference. H ertle C, Lowther R, Hartley K, Fracchia H, T uri P, Young D (2006), Membrane Bioreactors Big and Small, Brownfield and Greenfield - Cose Competitive in Far North Queensland, Enviro 06 Conference. Kitching S, Bruzzone M (2006) Applying MBR T echnology for Nutrient Removal, the Issues and Solutions, Enviro 06 Conference. Newland M (2006) Australia's Two Largest MBR Water Recycling Planes, Water 33 No 2. Young D , H ertle C (2006) Magnetic Island Water Reclamation Plane, Water 33 No 2.
The Author Stephen Chapman is a principal process engineer wich MWH, specialising in membrane technologies and municipal wastewater treatment processes. The Water Environment Federation (WEF) invited Stephen to co-author che MBR chapeer fo r the textbook Membrane Applications for Wastewater Treatment, which was published in the United Scates by WEF in 2006. Email: scephen.chapman@mwhglobal.co m
I
WATER
Reed Constructions Australia Pty Ltd Level 3 , 41 McLaren Street North Sydney NSW 2060 Ph: 02 9965 0399 Fax: 02 9955 8812
www.reedg roup.com .au
technical features .fereed paper
INTERACTION BETWEEN TREATMENT AND DISTRIBUTION SYSTEMS: THE KEY TO DELIVERING QUALITY J Q J C Verberk, M Drikas Abstract A new collaborative project on the impact of distribution systems on water quality at the cap has just commenced at the CRC for Water Quality and Treatment. Seven organisations are already participating and ocher interested utilities can still be involved . Construction of che crearmenc installations and pilot distribution systems will begin early 2007.
II
D
I D
Introduction D elivering safe water of high qualiry ro the customers is the key objective of water utilities worldwide. This is achieved based on the Total System App roach (T SA) principle (refer Figu re 1). The TSA is a risk management approach from source to tap and originated after the Milwaukee cryprosporidiosis outbreak in 1993. T he microbiological safety of drinking water should be secured th rough a multiple barrier concept. TSA consists of three pillars: • know your source water • know your treatment • protect your distribution system. Many o rganisations which have established drinking water guidelines, including the National H ealth and Medical Research Council (in Australia) and the European Union and World Health Organization, have adopted this Total System Approach. In Australia, TSA is better known as the "source to cap approach". Imp rovement in water quality con tinues to be che subject of extensive research into optimisation and development of new drinking water treatment processes in both laboratory experiments and pilot installations. This can be termed proactive research. However for the distribution system most work is focused on water
This project is funded by the CRC for Water Quality and T reat ment, SA Water Corporation, Grampians Wimmera Mallee Water, United Water International, \Xlater Corporation, Orica Watercare and Delft University of Technology in the Netherlands.
72
MAY 2007
Water
II
I
Figure 1. Total system approach as used by water utilities . quality deterioration caused by external effects, and is usually a reactive approach generally focussed on operational means of maintaining water quality such as impact of hydraulics, frequency of cleaning and predictive modelling of water quality events. Despite the d istribution system being known as a vase and uncontrollable biological reactor, only limited research on improving water quality with in th e distribution system is undertaken, as there are no operational parameters that can be easily changed. Water quality research in distribution systems is fu rther complicated due to the high cost and the limited application of the resultant case-specifi c information (Eisnor and Gagnon, 2003) . H owever, as it has become clear in recent years that che distribution systems can have an enormous effect on the water quality at the customers tap many experimental or model reactors have been designed for modelling of d istribution systems. Well known examples are che simulated distribution systems for chemical reactions and annular reactors for determin ing microbial processes (Eisnor and Gagnon, 2003; Angles et al., 2003; Camper et al., 1996). However, no overall studies caking in to account ch e combined influence of particulates, organic material and chemical interactions are known. In addition to
Journal of the Australian Water Association
experimental research on water quality in distribution systems, models can be used to simulate water quality ch anges, typically add-on programs to EPANET or H2ONET. These add -ons are mainly focused on disinfection residual prediction (Disinfection Management, 20 05). The Particle Sediment Model Qayaracne et al, 2004) is an example of an add-on module calculating che sediment deposition in a d istribution network. Add-on modules chat calculate non conservative water quality parameters as a resul c of the creacmen c and distribution are not available. These models would need to account for pose flocculation, corrosion formation, leach ing of piping material, resuspension of deposits and biofilm growth as a fu nction of rime and place in che distribution network.
In water utilities there is often no direct interaction between the two separate groups managing the treatment processes and che distri bution system. Similarly researchers tend to work in either the field of treatment or distribution research. This project aims to bridge this gap to imp rove che water qualiry supplied at che customer cap.
Interested utilities can be still be involved in this project.
technical features
a - - --.... -
water supply Interaction Between Treatment and Distribution System Coagulation I FIitration
Australian water industry. The latter three treatment process trains have been chosen as advanced treatmen c processes and have been targeted for two reaso ns: • the availability of an established treatment plane (Mc Pleasant in South Australia) with pilot plane facilities on site using a range of treatment processes (Figure 3 shows rhe conventional pilot plant installation at Mc Pleasant). The MIEX® and microfil tracion inscallarions are pre-existing at the Mt Pleasant treatment plane.
It has become clear over the lase Mt. Pl••ant WTP decade char rhe distribution system is one of the largest uncertainties in ~ - - - PrOCHS .:heme 2 rhe provision of drinking water. MIEX" This is parcly due co aging infrastructure and parcly because of the poorly understood interaction between che distriburion system PrOCHS .:heme :, and the water. The water quality entering the distribution system is ' MIEX" ' :'I critical ro chis internal reaction • within che distribution system. -............ : The importance of che acute and long-term impact of treatment ProceH .:heme 4 strategies on water quality in ... • minimal chemical interaction distribution systems was identified • with the treated water in ac a Global Water Research keeping with the trend coward Coalition wo rkshop held in Sydney Figure 2. Proposed treatment trains p roduci ng hig h quality less chemical treatment. in 2002 . Two countries interested drinking water. in developing this topic into a The different treacmen c trains research project were Australia and will produce differing water Treatment Processes the Netherlands. Representatives from the qualities and identify the extent of Four different treatment trains (Figure 2) Cooperative Research Centre for Water treatment necessary co produce stable water will feed the distribu tion test rigs. The fou r Quality and T reacmenc and Delft withi n a distri bution system. Whilst ocher proposed treatment process schemes co be advanced treatment processes could also Un iversity of T echnology in the assessed are: Netherlands volunteered co develop a new have been chosen the selected treatment research project. processes were those readily available, in • Conventional surface water treatment keeping with the philosophy of minimising consisting of coagulation, sedimentation This project is aimed at identifying che chemical rreacmenc, within che budget of and high race filt ration for removal of most differences that water treatment, and hence the project an d of interest co the fundi ng particulate material and some natural water quality, can have on che disrribucion utili ties. However, the comparison of che organic matter (NOM); system in terms of parciculares, chlorine impact of the chosen treatment trains on demand and biodegradable natural organic • MIEX® treatment fo llowed by the water quality and che distribution matter as a nutrient source for biofi lms. By microfil cracion fo r removal of NOM and system will potentially be transferable co usi ng fou r idenrical distribution test rigs particulate material; ocher water qualities. run in parallel, and comparing che water • MlEX® treatment followed by quality ac the end and at selected places The raw water source will be che River microfilcracion and granular activated Murray supply from Mann um, South within the distribution test rigs, differences carbon fo r removal of particulate material caused by the varied water quality entering Australia which supplies the Mc Pleasant and enhanced removal ofNOM; che test rigs will be observed, as Water Treatment Plant (WTP) via the • Micro fil cracion followed by nanofilcracion theoretically any differences will be caused Man num-Adelaide pipeli ne. This water has co provide ulrra pure water. by the water quality, not the test rigs. high salinity and alkalini ty with 5 year T he firs t treatment process train has been average turbidity of 50 NTU, medium The objectives of chis project are: selected as the 'control', as conventional HU) and medium dissolved colour (15 • To characterise the optimal water quality treatment is still the major surface water organic carbon (DOC) (3-1 0 mg/L) required ro minimise water quality treatment process used within the (Drikas et al., 2003a). deterioration within distribution systems. This will be achieved using four different treacmenc schemes co provide a range of water quality. Each creacmenc scheme will feed a separate idencical test rig co simulate distribution systems with decencion time of three days.
ll
ll
··----·- ------------
• To assess water quali ty parameters and analytical rools used co moniror water quality within distribution systems and identi fy chose chat can be used co predict water quality deterioration. • To use info rmation gained from chis project ro validate and modify established models co evaluate and predict che water quality within che distribu tion system.
Figure 3. Conventional water treatment pilot installation at Mt Pleasant. Journal of the Australian Water Association
Water
MAY 2007 73
technical features
m - - --••.• -
water supply Mc Pleasant WTP is an o perating treatment plant supplying high quality treated water to the local community (Overview of Mt P leasant WTP is shown in Figure 4). H owever, the plant is more complex than required as it has been designed co enable the use o f M IEX®, a con tinuous io n exchange process for NOM removal (Drikas et al, 2002), to be assessed fo llowed by a com parison of rwo processes for particulate removal. The plant has been d ivided into rwo screams, each of 1.25 ML/d design capacity, each incorporating the MIEX® DOC process. Scream 1 is designed to structurally simulate the retrofit of MIEX® into a conventional plant without any major infrastructure modificat ions or additio ns, followed by coagulation, flocculation, sedimentation and fil tration. Scream 2 comprises the greenfield process utilising MIEX® treated water directly o nto Veolia CMF-S (sub merged) polyvinylidene fl uoride microfiltration membranes. In addition, an extensive pilo t plant facil ity has been estab lished on site to assess the impact of MIEX® on subsequent treatment p rocesses such as granu lar activated carbon.
Pilot Distribution System Four identical distribution test rigs will be used to assess the impact of the various treated water qualities on distribution system performance. The systems will be designed at such flow to achieve approximately three days (72 hours) deten tion at the end of the system, inclusive of overnight stagn ation periods. T his detentio n rime is considered an app ropriate representatio n for metropolitan areas and sim ilar detention rimes have been used in ocher studies (H ole et al., 1994; Maier, 1999) . The length of each d istribution rest rig will be approximately 1 km. Each treated water source will feed a separate rest rig. As the diameter of the test rigs will need to be small to min imise water use (diameter ranging from 150 to 50 mm) they will be constructed from practical materials (PVC and PE) which are con sidered less likely co contribu te to corrosion or b iological growth than chose used in larger mains such as ductile iron or concrete. However these materials are currently specified fo r the construction of mains of similar sizes and their use will become more co mmon in new d istrib utio n systems. T he effect of age and type of distribution infrastructure on water quality deterioration are also key parameters but will not be addressed within chis project which is focussed on the impact of differing water q uality. T he velocity to be ap plied in the p ipes will be less than 0 .0 l m/s which is sufficiently low to enable both adhesion to
7 4 MAY 2007
Water
Figure 4. Overview of the Mt Pleasa nt WTP.
pipe walls and sedimentation due to gravity (Wu et al., 2003). It is intended to sim ulate diurnal flows by stopping the flow overnigh t for a set period. A schematic which outlines the proposed experimental design is given in Figure 5. The impact o f the treated water on th e d istribution system w ill be assessed over a perio d of 12 - 18 month s. T h e simplest, preferred approach is to supply the test rigs with unchlorinaced treated water however to simulate d isinfection p ractice in Australia, chlorination is being considered. T h e chlori ne dose to be app lied to the pilot systems will differ for each system and will be selected based on achieving <0 .1mg/L residual after app roximately 48 hours, that is, no resid ual for approximately 24 hours at the end of the distribution system. Whilst chis is not Australian water industry practice, it will simulate the loss of chlorine residual at the end of longer d istrib ution systems an d enable fo rmatio n of biofilms while still incorporatin g the impact of chlorination on distrib ution system perfo rmance. T he use of chlorination will
Ac the co m pletion of the trial, the pilot systems will be individ ually fl ushed at high velocity and the water will be collected and analysed for sedi ment and microbial content. This will en able an ad d itional q uantitative comparison o f th e impact of the various treatments on the d istribution system.
Water Quality System Monitoring The distribution test rigs will be monitored regularly for change in water quality, sediment deposition and biofilm growth over the period of the operation of the d istribution systems. R egular samples will b e taken d irectly after treatment, prior to entering th e distrib ution systems (sp 1 in
lsp3
spl Flow (1/ h) 250 Flow 5amp (1/ h) 50
complicate the implementation, increase the cost and affect interp retation of the impact of the varying water q u ality arisi ng out of the fou r treatment processes, however it is more d irectly relevan t to Australian supplies. It is also likely chat chlorination will actually accelerate the biofilm growth d ue to the known increase in bacterial regrowth potential (D rikas et al., 20036).
200
150 Pipe Diam (mm ) Velocity (m / s) 0 .0031 Pipe length (m) 150 Timecum (h) 13 .3 T im ecum included stagnation periods (h)
l
:so 150 i 0.0024! 150 l 31.2
i
Figure 5. Schematic design of distribution test rig system .
Journal of the Australian Water Association
100
ispS !50
50
0.014 300 37 .1
i
50 0.007! 450 l 54.8 i 76 .5 ,
technical features
Figure 5), and at the end of the distribution systems. However, ic is also proposed co cake samples less freque ntly at locations including where the pipe diameter changes (sp2, sp3 sp4 and sp5 in Figure 5). Sampling freque ncy will vary depending on che parameters measured. Parameters used co mon itor treatment operation such as rurbidity and DOC will be measured more frequently whilst ocher parameters which are more constant (such as iron) will be monitored less. Water quality mo nitori ng will initially include alumi nium, calcium, manganese, iron, silica, turbidity, particle counts, co lou r, conductivity, absorbance at 254 nm, DOC, organic character, chlori ne resi duals, hecerorrophic place counts, disinfection by-product fo rmation and potential for biological regrowth. Furthermore advanced on-line monitoring cools wi ll be used co fo llow the water quality change over ri me. An S::CAN onli ne UV monitoring instrument (Langergraber et al. 2002) will be moved throughout the distribution systems to allow measurement of changes in organic character at various intervals through the project for 2-3 week periods.
Furthermore TILVS (Time Integrated Large Volume Sampling) will be used to determine sediment deposition and the natu re of particles in the system. TILVS is an on-line filtration unit developed by che Delfr University of Technology which enables collection of particulate material over extended time periods which can then be fu rther characterised using traditional analytical techn iques, such as suspended solids measurement and inorganic analysis (Verberk et al., 2006). Monitoring of biofilm growth will be achieved by sampling at the end of the distribmion system for bacterial regrowth potential and biodegradable dissolved organic carbon. Furthermore bio fi lm forma tion monitors will be installed ac the begi nni ng of the fou r distribution systems (prior co chlorination) to monitor potential fo r biofilm growth. To enable water quality comparison of the pilot distribution systems with water supplied co customers, che water quality in the distribution system sourced from the Mc Pleasant WTP will also be sampled . W hilst the treated water serving the Mt Pleasant distribution system is a combination of wate r treated with microfiltration and conventional treatment following MIEXÂŽ, it will be useful co
undertake a comparison of the test rigs with a similar water quality in a real system.
Anticipated Outcomes T his project will identify che impact of water quality on the potential for dirty water events and microbial deterioration. Direct comparison of rhe varying water quality arising from rhe fo ur treatments will allow rhe key characteristics which cause water quality deterioration in the network to be identified. Th is should enable the establishment of guideli nes for water quality entering the distribution system which will minimise the likelihood of water quality deterioration. A range of analytical parameters and on-line cools (incl uding on-line techniques like S::CAN, which can be used co simultaneously monitor multiple parameters in water) will be used and new cools (like the on-line TILVS collectors in combination with particle counting) wi ll be evaluated for their usefu lness in monitoring water quality in rhe distribution system . Th is wi ll identify water quality parameters and analytical cools which are most suited for mon itoring water quality within distribution systems and can be used co predict water quality deterioration.
Corrosion Protection Lining AKS (Anchor Knob Sheet) PE (polyethylene) Li ning, as exclusively supplied in Australia by Humes, is an im pervious corrosion protection lining system for both precast and insitu applications. Firmly anchored to the structure, the lining forms a superior protective layer able to withstand both abrasion and harsh chemical environments.
Journal of the Australian Water Association
Water
MAY 2007 75
technical features
8
water supply C u rrent understanding of distribution system performance will be improved and enhance the use of current computer models in calculating the water qualiry at the customer tap based on water quality entering the system and characteristics of the ·distribution system.
How to Become Involved This project is curren tly supported by SA Water Corporation, Grampians Wimmera Mallee Water, Water Corporation, United Water I nternational, Orica Watercare, Delft University of Technology, and the Co-operative Research Centre for Water Qualiry and Treatment. These organisations have all contributed both resources and funding to enable this project to proceed over the proposed 2 1 months. Ir is proposed to hold a biannual workshop where the results of the project will be presented and discussions with project participants will take place. The first workshop was held in November 2006. Interested utilities can be still be involved in this project as the project is in its starting phase and co nstruction of the treatment installations and pilot d istribution systems will begin early 2007.
Laborat01y and Pilot-Scale Experiments.
Contact Details
AWWA Research Foundation, Denver, CO.
For more information contact Mary Drikas, ph 08 8259 0291, fax 08 8259 0228 or email: mary.drikas@sawater.com.au
The Authors Jasper Verberk is an Assistant Professor in drinking water supply at Delft University of Technology and worked on a secondment for eighteen months at the Co-operative Research Centre for Water Quality and Treatment in Adelaide. Mary Drikas is Program Leader for Water Treatment Technology Program within the Cooperative Research Centre for Water Qualiry and Treatment and Principal Research Chemist at the Australian Water Qualiry Centre in Adelaide.
References Angles M.L., Kastl G ., Chandy J.P., Payyappat S., Sathasivan A., Fisher I., Stevens M. (2003). Biofilm Devices to Facilitate Water Quality Invest igations in Distribution Systems. Proceedings ofthe !WA leading Edge
Conference: Drinking Water and Wastewater Treatment Technologies, Amsterdam, the Netherlands. C amper, A. K. (1996) Factors limiting
Filtration Media ... . ..
We are pleased to offer you:
• An extensive product range at competitive prices
. . .
.... .. . ..
..
So, tap into our extensive experience and helpful service when next you require fllter media.
683 Beenleigh-Redland Bay Road, Carbrook Queensland 4130 International +61 7 3287 6444 Freecall 1800 077 744 Facsmile 07 3287 6445
76 MAY
2007
Water
Journol of the Australian Water Association
{0
QUALITY MANAGEMENT
SYS T EM
ISO 9001 NATA CERTIFIED
Drikas M., Chow C. and Cook D ., (20036) The Impact of Recalcitrant Organic C haracter on Disinfection Stability, T rihalomethane Formation and Bacterial Regrowth : An Evaluation of Magnetic Ion Exchange Resin (MIEX®) and Alum Coagulation, journal of Water Supply: Research and Technology -Aqua, 52, 7, 475-487 Eisnor J.D., Gagnon G.A. (2003). A Framework for the Implementation and Design of Pilotscale Distribution Systems. journal of Water Supply: Research and Technology -Aqua, 52 , 7, 501-5 19 .
Langergraber G., Fleischmann N., Hofstaedter F (2002) A Multivariate Calibrat ion Procedure for UV/VIS Spectrometric Quantification of Organic Matter and Nitrate in Wastewater, Fleischmann N. et al. (Eds.): Proceedings of
240.
• Manufacturing in accordance with Quality System AS/NZS ISO 900 1:2000
• Prompt delivery Australia-wide and Overseas
Proceedings ofthe American Water Works Association Water Quality Technology Conference, Philadelphia, USA
the International !WA Conference on Automation in Water Quality Monitoring AutMoNet 2002, Vienna, Austria, pp . 233-
• Media produced to the AWWA B100-89 Standard
• Proficient technical assistance and support
Drikas M. , Morran J.Y., Cook D. and Bursill D.B. , (2003a) Operating the MIEX· Process with Microfilcration or Coagulation,
Jayarame, A., Ryan, G., Graigner, C., Wu, J., and Noui-Mehidi, M. N . (2004). Modelling of Particles in Water Supply Systems, AWA Water, 31, 8, 30-36.
• Decades of mineral processing experience
• Packaging alternatives to suit any requi rement
Drikas M ., Morran J.Y., Pelekani C., Hepplewhite C. and Bursill D.B., (2002), Removal of Narnral Organic Marrer - a Fresh Approach, Water Science & Technology: Water Supply, 2 , I, 71-79
Water Works Association Water Quality Technology Conference, San Francisco, USA.
Sand, Gravel, Coal, Garnet, Manganese Greensand
Invest in accurately graded, durable media from your complete fllter media professionals.
Disinfection Management, Implementing Tools for Optimising Disinfection (2005). Cooperative Research Centre for Water Quality and Treatment, Adelaide, Australia
Holt D.M., Woodward C.A., Rachwal T., D elanaoue A., Colbourne J.S . (1994). A Pipe Based Experimental System to Srndy Factors Influencing Bacterial Growth in London's Water Mains. In: Proceedings of the American
Microbial Growth in Distribution Systems:
---~--RIVER SANDS ~~
. . . . . . ...
Maier S.H. (1999). Modelling Water Quality for Water Discriburion Systems. PhD thesis, Brunel University, Uxbridge. NHMRC/NRMMC (2004), Australian Drinking Water Quality Guidelines, http://www.nhmrc.gov.au/publicat ions/ synopses/eh 19syn.htm VerberkJ.Q.J.C., H amilton L.A., O 'H alloran, K. ]., Horst W. van der, VreeburgJ. (2006) Analysis of Particle Numbers, Size and Composition in Drinking Water Transportation Pipelines: Results of On-line Measurements, Water Science and Technology: Water Supply, 6 , 4, 35-43.
... ... ... ..... .. . ..
Wu J., Noui-Mehidi N., Grainger C., Nguyen B.V., Ryan G ., Jayarame A., Mathes P . (2003) . Particle Sediment Modelling: PSM Software. Particles in Water Distribution Systems - 6 th Progress Report Project 4 .3 .6, CRC for Warer Quality and Treatment.
techniral features
water supply
fereed paper
TRIHALOMETHANE MANAGEMENT IN DRINKING WATER: A NOVEL APPROACH R Trolio, R Walker, S McNeil, T Huynh, N Jismi Abstract Western Australia has numerous so u rces of water char are high in natural organic matter (NOM ), which causes high chlorine demand and may produce u ndesirable chlorination by-produces, such as cri halomethanes (T HMs) . Bench-scale laboratory experiments have enabled T H M fo rmation to be pred icted through che development of empirical models for individual sources. O perators are now able to optimise the chlorine dose rates. Some distribution systems requi re further reduction ofTH M concentrations. T he installation of a recirculatio n and aeration device in service reservoirs/ranks provides a novel solution to volatilise THMs. In two case studies a 40 to 70 % red uction in concentrat ion was ach ieved. This novel concept w ill provide water utilities with a new option in maintai ning THMs well below the Australian D rinking Water Guideli ne (ADWG) health related guideline value, while longer term treatment options are consid ered.
Key Words Trihalomethanes (THMs), Aeration, Natural Organic Matter (NOM), Drin king Water, T H M Formation Potential (TH MF P)
Introduction The Water Corporation (WCWA) is the main supp lier of drinking water in Western Australia, servicing a population of over 2 mill ion people spread over 2 m illion square kilometres. It operates 245 water distribution schemes in urban , rural and remote communities, making use of local surface o r grou ndwater sources. The southern part of Western Australia comprises of sparsely located towns with local surface and groundwater sources. Many of these sources are high in nacural o rganic matter (NOM). The interaction of NOM w ith chlorine during d isinfection creates challenges fo r operators as they m ust achieve effective d isin fection at the same time as minimising disin fect io n by-products such as trihalomethanes (THMs). In most
300 - , - - - - - - - - - - - - - - - - - - - - - - - - - - - - , llc:ilth Guideline Value
250 -
~ -:>
~
200
THMFP
= 50.8 x Chlorine Consumed
R2 = 0.991
ISO
E-- 100
50
0
2
3
4
5
6
Chlorine Consumed (mg/L)
Figure l. THMs formed versus ch lorine consumed after a seven day reaction period for Lake King Da m waters. schemes it is also importan t to maintain chlorine residuals to the extremity of the d istrib ution system to manage Naegleria and bacterial regrowth. An inadequate chlo rine dose race may result in CT (contact time) value that will not achieve the desired pathogen removal. Co nversely, an excessive chlorine dose rate can lead to che formation of THMs which exceed the health guideline value of 0.25 mg/L (equivalent to 250 Âľg/L), as specifi ed in the Australian Drin king Water Guideline (ADWG). In the past, the WCWA has used che Standard Trihalomechane Formation Potential (TH MF P) rest to assist operators in determining the optimum chlorine dose race. This method involves determining the chlorine d emand of che sample water, then buffering an aliquot of the sample water to pH 7, chlorination to provide an excess of free chlorine of 3 to 5 mg/L after 7 days and then measurement ofTHM concentration (Clesceri et al., 1998). The method however only provides the operator with a T H M concentration at a single specifi c chlorine dose. Chlorine dose race and reaction rime are the most important factors chat influence the format io n ofTHMs, with temperature and p H secondary facto rs. THMFP was
d eveloped by performing bench scale rests char vary the chlorine dose race and reaction rime. Understanding che formation of THMs under various conditions will assist operators in controlling THM co ncentrations in the distributio n systems. In some distribution systems effective di si nfection canno t be achieved at che same rime as maintain ing T H Ms below che health guideline value. Initially, aeration was em ployed by the WCWA fo r several years to reduce the T H M co ncentrations within the system . Aeration was achieved in small ranks by installing a ran k inlet splash place and a tank roof ventilacor (Henderson and Koska, 2003). More recently new techniques h ave b een developed for larger ranks and reservoirs. T his study investigated THM mitigation at three localities in southern Western Australia: Lake King, Den mark and Boddington. Initially THMFP tests were conducted and a new aeration technique trialled as an option for reducing THM concentrations.
Aeration dissipates THMs once formed in the bulk water.
Journal of the Australian Weter Association
Water
MAY 2007 77
technical features
e .. --.....
water supply Discussion
Table 1. THMs formed, free chlo rine resi dual and calcu lated chlorine consumed at
Lake King Distribution System The town of Lake King is located 462 km south east of Perth. Water for the town is supplied from an excavated earth dam (Lake King Dam with a nominal capacity of 12 ML) which is fed from an artificially constructed catchment area designed to capture rainfall fo r runoff. The catchment area is cleared of vegetation and surfaced with a thin layer of bitumen. The local source is the only water available to Lake King. Failure of the source would require water to be carted from Lake Grace, a distance of 11 6 km. Raw water from the dam is disinfected using sodium hypochlorite at the outlet of a transfer pump station near the dam. The disinfected water is then transferred to a service tank, which supplies the town reticulation system. In recent years, due to the low runoff into Lake King Dam the concentration ofNOM has steadily increased. As a consequence, both the chlorine demand and THMFP of the so urce water has increased, resu lting in a challenge for operators to effectively manage disinfection. Operators are required to increase the chlorine dose rate to accommodate changes in chlorine demand, however, they had no knowledge of the impact on IHM formation in the reticulation.
day 7 measured at various chlorine doses for Lake King Dam waters.
Lake King THM Formation Potential Tests To understand and manage the T HMFP of Lake King Dam waters, bench scale tests were co nducted to observe the effects of chlorine dose race and reaction time. Waters were measured into 1 L brown glass bottles and dosed with known amounts of sodium hypochlorice (NaOCI), ranging from 2.1 - 8.4 mg/L. T he samples were then allowed to react for seven days (168 hours) at a temperature of 20°C. Sub-samples were then analysed for THMs. Dissolved organic carbon (DOC) and ultraviolet absorbance (UV(254)) were measured on unchlorinaced waters as these are surrogate measurements for NOM. Free chlorine residuals were measured on day seven and the resulting T H Ms formed during this time frame are derailed in Table 1 and shown in Figure 1. Table 1 shows T HMs form above the health guideline value (255 µg/L) when Lake King Dam waters co nsume 5.24 mg/L of chlorine. The formation ofTHMs above the health guideline value is well supported by the UV(25 4> and DOC values of 0.068 and 3.72 mg/L respectively, which indicate a high NOM co ncentration present in Lake King Dam waters. 78 MAY 2007
Water
Cl2 dose lmg/L)
Free Cl2 Residual Meosured at Day 7 lmg/L)
Free Cl2 Consumed Measured at Day 7 lmg/ L)
THMs lµg/ LI
0 2.1 4.2 6.3 8.4
0.00 0.0 1 0.10 l.92 3.1 6
0.00 2.09 4.10 4.38 5.24
0 116 206 234 255
Lake King Chlorine Decay Tests
Figure 1 shows chat a strong linear relationship (correlation coefficient of 0.991) was apparent between free chlorine consumed and THMs. This relationship had previously been observed (Boccelli et al., 2003). T he following THMFP regression model is proposed fo r Lake King Dam waters based on chlorine consumed:
(THMFP in µg/LJ = (50.8) x (chlorine consumed in mg!LJ Eqn (1) Equation 1 above concluded char if the chlorine consumed was greater than 4.92 mg/L, the THMs concentration would exceed the health guideline value of 250 µg/L. The model was validated by profile sampling within the Lake King reticulation system measuring the chlorine co nsumed and T HMs. For example, at the assessable reticulation sample point in Lake King township, the T HM concentration was 182 µg/L (laboratory result) with a chlorine consumed of 3.56 mg/L. This compares to the derived model calculated value of 181 µg/L, verifying the applicability of the model. W hen adjusting chlorine dose races operators were able to use chis model to predict the T HMs formed in the distribution sys tem and ensure the health guideline value was not exceeded.
The primary purpose of disinfecting drinking water is to eliminate pathogenic organisms responsible for waterborne diseases. In addition co maintaining THMs below the health guideline value, the WCWA has adopted principles to control microbial contamination. Firstly, the minimum dose rate should be based on maintaining a minimum CT value of 15 before the water reaches the firs t customer i.e. 1.0 mg/L residual for 15 minutes co ntact. Secondly, a free chlorine residual of higher than 0.2 mg/L is required in the reticulation system to manage Naegleria and prevent bacterial regrowth. To attain the above requirements it was imperative to determine the chlorine decay characteristics of the water. The bench scale chlorine decay tests involved measuring waters into 1 L brown glass bottles and dosing with known amounts of sodium hypochlorice (NaOCl), ranging from 2.1 8.4 mg/L. The chlorine residuals of the samples were measured at timed intervals and at a temperature of 20°C. The resulting chlori ne residuals formed during chis cimeframe are listed in Table 2 and shown in Figure 2. The detention rime in the Lake King service tank can be up to three days in winter. T he travel rime to the extremity of the reticulation is estimated to be about
Table 2. Measured free ch lor ine residua ls for Lake King Dam waters after se lect ti me interva ls. Chlorine Dose Rates lmg/L)
2.1 mg/ L
4.2 mg/ L
6.3 mg/ L
8.4 mg/ L
Time lhoursl
Free Chlorine Residual lmg/L)
Free Chlorine Residual lmg/ L)
Free Chlorine Residual lmg/L)
Free Chlorine Residual lmg/ L)
0 0.5 4.5 23 49 72 144 168
2.10 0.79 0.34 0.05 0.03 0.02 0.02 0.01
4.20 2.76 l.98 1.02 0.74 0.40 0. 17 0. 10
6.30 5.80 5. 16 3.88 3.32 2.90 2.38 l.92
8.40 7.00 6.42 5.56 5.12 4.84 4. 60 4.38
Journal of the Australian Water Association
technical features
1B ....••.•.
water supply Discussion
Table 1. THMs formed, free chlorine residual and calculated chlorine consumed at
Lake King Distribution System
day 7 mea sured at various chlorine doses for Lake King Dam waters.
The town of Lake King is located 462 km south east of Perth. Water for rhe town is supplied from an excavated earth dam (Lake King Dam with a nom inal capacity of 12 ML) which is fed from an artificially constructed catchment area designed co capture rainfall for runoff. The catchment area is cleared of vegetation and surfaced with a chin layer of bitumen. The local source is rhe only water available co Lake King. Failure of rhe source would require water co be carted from Lake G race, a distance of 116 km. Raw water from the dam is disinfected using sodium hypochlorire at the outlet of a transfer pump station near the dam. The disinfected water is then rransferred co a service tank, which supplies the town reticulation system. In recent years, due co the low runoff into Lake King Dam the concentration ofNOM has steadily increased. As a consequence, both the chlorine demand and THMFP of rhe source water has increased, resulting in a challenge for operators co effectively manage disinfection. Operators are requi red co increase the chlorine dose rare to accommodate changes in chlori ne demand, however, they had no knowledge of the impact on THM fo rmation in rhe reticulation.
Lake King THM Formation Potential Tests To understand and manage the THMFP of Lake King Dam waters, bench scale rests were conducted co observe the effects of chlorine dose rate and reaction rime. Waters were measured into 1 L brown glass bottles and dosed with known amounts of sodium hypochlorite (NaOCI), ranging from 2.1 - 8.4 mg/L. The samples were then allowed co react for seven days (168 hours) at a temperature of 20°C. Sub-samples were then analysed for THMs. Dissolved organic carbon (DOC) and ultraviolet absorbance (UY(ZS4)) were measured on unchlorinated waters as these are surrogate measurements for NOM. Free chlorine residuals were measured on day seven and the resulting THMs formed during chis rime frame are derailed in Table 1 and shown in Figure 1. Table I shows TH Ms form above rhe health gu ideline value (255 ~1g/L) when Lake King Dam waters consume 5.24 mg/L of chlorine. The fo rmation of THMs above the health guideli ne value is well supported by the UY(ZS4) and DOC values of 0.068 an d 3.72 mg/L respectively, which indicate a high NOM concentration present in Lake King Dam waters. 78 MAY 2007
Water
Cl2 dose (mg/L)
Free C'2 Residual Measured at Day 7 (mg/L)
Free Cl2 Consumed Measured at Day 7 (mg/L)
THMs (119/L)
0 2.1 4.2 6.3 8.4
0.00 0.01 0. 10 l.92 3.16
0.00 2.09 4.10 4.38 5.24
0 116 206 234 255
Lake King Chlorine Decay Tests
Figure 1 shows that a strong linear relations hip (correlation coefficient of 0.991) was apparent between free chlo rine consumed and THMs. This relationship had previously been observed (Boccelli et al., 2003). The fo llowing T HMFP regression model is proposed for Lake King Dam waters based on chlorine consumed:
Equation 1 above concluded that if the chlorine consumed was greater than 4.92 mg/L, the THMs co ncentration would exceed the health guideline value of 250 µg/L.
The primary purpose of disinfecting drinking water is to eliminate pathogenic organisms responsible for waterborne diseases. In addition to maintaining THMs below the health guideline value, the WCWA has adopted principles to control microbial contamination. Firstly, the minimum dose rate should be based on maintaining a minimum CT value of 15 before the water reaches the first custo mer i.e. 1.0 mg/L residual for 15 minutes contact. Secondly, a free chlorine residual of higher than 0.2 mg/Lis required in the rericularion system to manage Naegleria and prevent bacterial regrowth.
The model was validated by profile sampling withi n rhe Lake King reticulation system measuring the chlorine consumed and THMs. For example, at the assessable reticulation sample point in Lake King township, the T HM concentration was 182 µg/L (laboratory result) with a chlorine consumed of 3.56 mg/L. This compares co rhe derived model calculated val ue of 181 µg/L, verifying the applicability of the model. When adjusting chlorine dose rares operators were able co use this model co predict the T HMs formed in rhe distribution system and ensure the health guideline value was not exceeded.
To attain rhe above requirements it was imperative co determine rhe chlorine decay cha racteristics of the water. The bench scale chlorine decay tests involved measuring waters into I L brown glass bottles and dosing with known amounts of sodium hypochlorite (NaOCI), ranging from 2.1 8.4 mg/L. The chlorine residuals of the samples were measured at timed intervals and at a temperature of 20°C. The resulting chlorine residuals formed during this timeframe are listed in T able 2 and shown in Figure 2. The detention rime in the Lake King service tank can be up to three days in winter. The travel time to the extremity of the reticulation is estimated to be about
[THMFP in µg!L} = (50.8) x [chlorine consumed in mg!L} Eq n (1)
Table 2. Measured free chlorine res idua ls for Lake King Dam waters after select time intervals. Chlorine Dose Rates (mg/L)
2.1 mg/ L
4.2 mg/ L
6.3 mg/ L
8.4 mg/L
Time (hours)
Free Chlorine Residual (mg/ L)
Free Chlorine Residual (mg/L)
Free Chlorine Residual (mg/ L)
Free Chlorine Residual (mg/L)
0 0.5 4.5 23 49 72 144 168
2.10 0.79 0.34 0.05 0.03 0.02 0.02 0.0 1
4.20 2.76 l.98 1.02 0.74 0.40 0.1 7 0.10
6.30 5.80 5.16 3.88 3.32 2.90 2.38 l.92
8.40 7.00 6.42 5.56 5.12 4.84 4.60 4.38
Journal of the Australian Water Association
technical features
water supply seven days in winter (incl uding rhe three days in the service rank). From Figure 2 a dose rare of 4.2 mg/L gives a chlorine residual of 0. 10 mg/L after seven days . Therefore a dose rare of approximately 4.3 mg/L achieved the target of 0.20 mg/L at rhe extremity of che reticulation.
refereed paper
Table 3. Typ ical Quick Reference Sheet, given to operators, for Lake King Dam waters.
Quick Reference Sheet Interpreting rhe information gathered from the T HMFP and chlorine decay experiments enabled a better understanding of the im pact of ch lorine dose rare on the T HMFP and chlorine decay behaviour. T his information was disseminated to the operators in rhe fo rm of an easy to read 'Quick R eference Sheer' shown in Tab le 3. The THM concencrarion values in Table 3 assume the chlorine is completely consumed, to provide rhe operators with a worst case scenario. The optimum dose race for Lake King Dam waters was therefore between 4.0 to 4.5 mg/L, with a target of 4.3 mg/L. Since che implementation of the operatio nal strategy as detailed in che Quick Reference Sheet, operators have been able ro achieve residual in the reticulation system an d maintai n THMs below the health gu ideli ne value.
Denmark Distribution System Den mark is located in che G reat Southern region of Western Australia, 54 km west of che regional centre of Albany (4 10 km sou ch of Perth). The town water supply is drawn fro m Quickup Dam located 6.5 km north ease of che town. Quickup Dam water is created by sedimentation, clarificat ion and filtratio n. T he created water is disinfected by chlorine gas as it leaves the Water Treatment Plan t, pumped into Horsley Road Reservoir and then pumped into Horsley Road Summ it Tank. From rhe Su mmit Tank water is gravity fed inco che reticulation system or pumped co various service tanks. In March 2002, chlorination dose rare was increased to overcome a detection of coca! coliforms in the reticulation. As a consequence, che TH M concencracions exceeded the heal ch guideline value of 250 Âľg/L on several occasio ns. W hile no t an immediate health co ncern, chis trend concerned both rhe WCWA and the Department of Health. This concern prompted an investigation inco che reaso n for che exception, and co determine shore term measures chat wou ld manage che T HM issue, while awaiti ng enhanced treatment for NOM removal.
;::i'
--
7
7 E 6 __, -; 5 = t>l)
.
"O
¡i;;
4
~
3
0
2
....
-...- 2. Jmg/L
- . . - 4.2 mg/L
-
-
6.3 mg/L
8.4mg/L
-
J
I
0 0
20
40
60
80
100
120
140
160
180
Time (hours)
Figure 2. C hlorine residual versus time for Lake Ki ng Dam waters.
(. lOWARA
Hydrovar is the world 's fi rst pump mounted microprocessor based pump controller. It does much more than just change pump speed, it truly manages your pump to match a wide range of system conditions. Energy savings of up to 70% can be achieved.
Contact us now about your specific application. Melbourne
03 9793 9999 Sydney
02 9671 3666
Web: www.brownbros.com.au Email: info@brownbros.com.au
~nn \ii
Bro"'.Vn Brothers 1 !'.,lf)'jf II Engineers Australia Pty Ltd
Journal of the Australian Water Association
Water
MAY 2007 79
technical features
~
water supply
- - --
.... -
Denmark THM Formation Potential and Chlorine Decoy Tests At the outset, THMFP and chlorine decay tests were undertaken on waters sampled post treatment bur prior to chlorination following the same m ethodology used for Lake King. The results of these tescs would help develop the Quick Reference Sheet fo r D enmark Treated waters. The UV(254) and DOC values for the treated waters were 0.128 and 6.8 mg/L respectively, so inevitably this wou ld produce T H M concentrations above che health guideline value. Equation 2 shows che T H MFP model proposed based on chlorine consumed.
[THMFP in Âľg!L] = (66) x [chlorine consumed in mg!LJ Eqn (2) The above equatio n concluded that if che chlorine consumed was greater chan 3.8 mg/L, the T H M concen tration wou ld exceed the health gu ideline value. Validation of the model also occurred by measu ring the ch lo rine consumed and T HMs throughout che D enmark distrib ution system . For example, on the inlet to Horsley Road Reservoir che T HM concentration was 247 pg/L (laboratory result) with a chlorine consumed of3.7 mg/ L. This compared to the d erived model calculated value of 244 pg/L. From the chlorine decay rests, a chlorine dose of 6.3 mg/L p roduced a chlorine residual of 0.45 mg/L after seven days . T he detention time in D enm ark during winter m onth s is in excess of seven days. The Quick Reference Sheer illustrated chat for this warer, ir was di fficu lt to maintain THMs below the health guideline value ar che sam e time as achieving a chlorine residual of 0.2 mg/L in the extremity of che reticulation
Denmark Reservoir Aeration Trials Ar che time of the invescigacion, the chlorine dose race was 7 mg/L ac the outlet of the WTP . At chis dose rate, THMs could fo rm as high as 462 pg/L, as calculated usin g the derived model. Therefore, ir was pertinent to reduce concentrations to below the health guideline value. H enderson and Koska (2003) showed chat bench scale aeration tests i.e. cascading water from one 500 mL beaker to another between 5 to 10 rimes, res ulted in THM concentrations being reduced by 35%. Following on from this work, the WCWA installed splash plates at a number of sm all tanks to aerate chlorinated water as it entered che tank. le was found splash plates successful ly reduced T HMs where rhe rank was less than 0.5 ML in volume. Effectiveness declined wirh increasing volume and rhere were doubts
80 MAY 2007
Water
Figure 3. Aeration at Denmark Ho rsley Road Reservoir.
Boddington System Investigation
chat this could be used ac Horsley Road Reservo ir, a reservoir wirh a volume of 4.5 ML. Additionally, the splash plate can only be utilised when inflow occurred.
The town of Boddington is supplied from Harris River Dam which is located on che Harris River 192 km sou ch of Perth . The water ac Boddington complies with che ADWG aesth etic guidelines, however it contains a high concentration of NOM.
It was nored chat inflow to H orsley Road Rese rvo ir only occurred for cwo to fou r hours per day. H owever, if a recirculation system cou ld be installed in che reservo ir, the effe ctive rime for aeration could be increased by an o rder of m agnitude. As a resul t, ic was decided to trial conci nuous aeratio n (recirculating the reservoir co ntents th rough an aerator) ac che Horsley Road Reservoir. The system co nsisted of a sub mersible pump located on the reservoir floor, fl exib le discharge hose and a section of slotted uPVC pipe (aerator) , as shown in Figure 3.
Chlorination for disinfection occurs at the dam outlet and water is re-chlorinated on t he inlet to the 4 .5 ML Boddington storage rank. Disinfected water then gravitates into the Boddington reticulation system. Boddington town water supply has o n occasion exceeded the h ealth guid eline for THMs as shown in Figure 4. This corresponded with UV(2 s4) readings for Harris River Dam being greater than 0 .07 during November to January, indicating high concentrations ofNOM. THMFP tests were conducted co create the model
Insrallacion of rhe continuous aeration system in Horsley Road Reservoir, resulted in a su ccessful reduction in the average THM concentrations value from 2 18 pg/L to 73 pg/L. Free chlorine residual was also measured pre- and post-aerarion and chis showed no change.
d epicted in Equation 3 .
[THMFP in Âľg!Lj = (75) x [chlorine consumed in mg/LJ Eqn (3)
350 300 Health Guideline Vnl11c
250
_ _"-,( , 247
;3' ch 200
-5 ,,,
i
E--
150 LOO
so 0
"'
Q
~
"' "'
Q Q
::!:
"'
Q Q
V,
::!:
"'
Q ..... Q
::!:
"' ~ Q ....
"'
Q
~
.... :S Q
::!:
....
Q "' Q
....
....
Q Q
V,
::!:
....
Q ..... Q
....
.... "'
Q Q
::!:
....
V,
V\
~
Q Q
....
Q Q
Q
Date
Figure 4. Historical THM co ncentrati ons in Beddington distribution system.
Journal of the Australian Water Association
"' ::!:
technical features
The success of continuous aeration at Denmark to dissipate THM concentrations, prompted a similar approach to be trialled ac Boddingcon Tank.
Boddington Tank Aeration Trials The trial recirculation and aeration system at Boddingcon Tan k slightly differed to chat of Denmark Horsley Road Reservoir. Jc comprised of cwo DN 150 uPVC Class 18 aerator headers mounted to che inside of the tank wall, each with three aerator sp ray nozzles at 1800 mm centres spraying slightly below horizontal. The aerator headers as shown in Figure 5 were located in che vicinity of the rank inlet to maximise recirculation of the rank contents, particularly during periods of low water demand . The amount of water aerated per day was approximately 1.2 ML. Figure 6 shows the THM concentrations pre- and posr-i nsrallarion of the recirculatio n and aeratio n system at Boddingto n. The effectiveness of the system is shown as the T H M co ncentration at the outlet has continuously decreased since insrallarion. Following system implementation, T H Ms decreased from 195 pg/L (31/08/04) to 85 pg/L (2 1/06/05). T his reached a plateau of76 pg/L fo r two weeks before a further reduction to 40 pg/ L (12/07/05). The most recent THM sample (05/ 12/06) taken in Boddington's distribution system returned a result of 72 pg/L, well below the health guideline value. T he reduction in THMs at Boddingto n can be attributed solely to the aeration system as the quality of water supplied co Boddingron from Harris River Dam has remained constant.
Conclusion Potable source waters with high NOM concentrations have caused difficulties fo r operators to manage effective disinfection at the same rime as maintai ning THM concentrations below che health guideline value. Empirical models, based on benchscale laboratory experiments have been developed which predict the formation of T HMs for a particu lar water, based on chlorine co nsumed. T hese enable operators to optimise the chlorine dose rares. In some distribution systems it is difficult to maintain T HMs below the health guideline value. Fo r example, ac Denmark and Boddingco n, recirculation and aeration systems were installed in a reservoir and rank respectively, to continuously aerate rhe waters to vo latilise the THMs. T his device was highly effective in reducing THM concentrations which have already formed in the bulk water to well below the ADWG health guideline value.
Tank Outlet Valve Pit 2 Reclrculatlon Pumps
Figure 5. Dia gra mmatic representation of aeration at Bodd ingto n tank. Acknowledgments Much support was provided by the Water Corporation's Grear Southern Regional staff in particular Graham Wright, Colin Collier and Danny Burkett who organised the sampling and made all the operational and asset modificatio ns described in this paper.
The Authors
Dr Rino Trolio is the Manager Operations Support for Drinking Water Qual ity Branch, Water Corporation , Leederville, Western Australia 6007, rino.rrolio@warercorporarion.com.au. Richard Walker is rhe Manager Drinking Water Quality at the Water Corporation. Sharon McNeil is a Chemistry Graduate at rhe Water Corporation. Tran Huynh is rhe Catchment Coo rdi nator at the Water
Corporation and Norqurshiah Jismi is a Scientific Officer at the Water Corporation.
References ADWG (2004) Australian Drinking Water Guidelines. Nacional H ealch and Medical Research Council and che Natural Resource Management Ministerial Council. Commonwealth of Aust ralia, Canberra. Boccelli D. L. , T ryby M. E., Uber J. G . and Summers R. S. (2003) A reactive species model fo r chlorine decay and T HM formation under rechlorinacion conditions. Water Res. 37, 2654 - 2666. Clesceri, LS., Greenberg, A.E. and ÂŁaeon, A.O. (Eds). ( 1998) Standard Mechods for che Examination of \'v'acer and Wascewacer 20ch Edition. Washington DC: American Public Health Association. Henderson R. and Koska L. (2003) Innovative low cost measures co reduce THM concentrations in distribution systems. OZWATER Convention and Exhibition, Perth, Australia, 6- 10 April, CD-Rom.
250 ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~
J
~
Aeration Installed
195
200
150
,5
::;"' a: [-
100
50
0 27/07/04
31 /08/04
21 /06/05
28/06/05
05/07/05
12/07/05
Date â&#x20AC;˘ Inlet Boddington Tank II Outlet Boddington Tank
Figure 6. THM co ncentration s at the inlet and outlet of Bodd ington Tan k pre- a nd post-aeration.
Journal of the Australian Water Association Water MAY 2007 81
technical features .fereed paper
EVALUATION OF TWO WATER TOXICITY TESTING KITS C Cook, D Wren, G Rose Abstract T he potential for threats to public infrasrrucrure has alerted Government agencies (health, security, water) to the need to assess the capabilities of field rest kits fo r the rapid analysis (within a few minutes) of potable water quality without the need for derailed laboratory analysis. Two commercially available cesc kits were evaluated for their ability to rapidly screen water for chemical contaminants. The Eclox™ kit uses chemiluminescence and specific rest strips while the IQ-Tox™ kit uses a species of zoo plankton , Daphnia magna, both reseed in USA fo r detection of lethal doses. Our interest was to determ ine if they could detect toxic chemicals at levels below the lethal dose, but still likely to adversely affect human health. Two hundred potable water samples were reseed using the Eclox™ kit to provide baseline data. Resu lts were compared with routine test parameters provided by Yarra Valley Water to determine if any correlation could be established. Water was also spiked with arsenic, cyanide, chlorpyrifos, glyphosare or paraquat at several concentrations and reseed using both kits to compare functionality and limits of detection . Generally, the IQ-Tox™ kit was more sensitive than the Eclox™ kit, but borh kits failed to provide a su itable response fo r some chemicals at levels of concern for human health.
Introduction In the current enviro nment of potential threat, the Victorian Department of Human Services (DHS) is seeking to provide rapid responses to threatened or deliberate contamination of drinking water supplies with toxic chemicals or other agents. Rapid rest kits are suitable for demonstrating due diligence in a cost
The IQ-Tox TM was more sensitive than the Ee/ox TM test kit to the contaminants tested. 82 MAY 2007
Water
C
•
80
Data points - - - - - - Regression
••
0
:;:
:0 .c
.5 ~ 0
...
Cl) Cl)
60
I------t·'-I!!:· •* ;•·~-.•
40 20
0
C
0
...J
-20
·e::,
•
1-.! ___ -------I
E
6
7
8
9
10
pH (field) Figure 1. Relationship between field pH and luminometer % inhibition. effective manner and can be readily mobilised if the water source an d distribution system is threatened or intentional or accidental conta mination is suspected. Considering chat these kits will be providing critical information during a contamination incident, it is important chat they be rigorously validated to demonstrate chat they are fie for purpose. Test kits can also be utilised for routine quality evaluation through the distribution network to pinpoint problem areas or issues. T he Environmental Health & Chemistry laboratory at DPI Werribee, with the cooperation of Yarra Valley Water, evaluated cwo commercially available rapid water testing kits. They were the Eclox ™ Rapid Response Water Test Kit and the IQ-Tox™ Threat Detection Kie. Boch of these kirs have been assessed by the US EPA Environmental Technology Verification (ETV) Program, but the emphasis was on detecting lethal dose concentrations (reports available from Imp:/ /www.epa.gov/ ecv/verificacions/ verification-index.html). Our interest in these kits was to determine if they could detect toxic chemicals at levels below the lethal dose, but still likely to adversely affect human health. The objectives of chis project were to: 1. Test duplicate samples from Yarra Valley Water's routine testing program using the EcloxTM kit. This would provide valuable
Journal of the Australian Water Association
baseline data and also allow correlation of the Eel ox™ reading against a comprehensive suite of water quality parameters. 2. Test water samples spiked with a range of readily available toxic chemicals using the Eclox™ and IQ-Tox™ kits. This would determine if the kits could detect a contaminant at levels likely to adversely affect human health and also provide a comparison of the two kits.
Eclox™ kit The Eclox™ kit, supplied by Severn Trent Services, UK, is a broad screening kit designed for use in the field, with mu ltiple tests in the package. The main test is the chemiluminescence test i.e. the reaction of luminol and an oxidant in the presence of a catalyse enzyme. The addition of an enhancer allows a steady output of light to be measured by the Eclox™ Lumi nomecer. Contaminants in a water sample will suppress the reaction and reduce the light output relative to a clean reference sample, usually distilled water. Interpretation of the reading requires some experience of the natural variation found in potable waters. A reading of 50% inhibition or more would warrant further investigation of the water quality. The kit is battery operated and readily deployable to the field. The required reagents are claimed to be stable at room temperature fo r up to two years (see Arsenic below).
technical features
Table 1. Correlations between va rious Eclox™ luminometer, Eclox™ pH meter and water laboratory test results. Test
pH (field) Lum inometer % inhibition Luminometer % inhibition Luminometer % inh ibition luminometer % inhibition Luminometer % inhibition Luminometer % inhibition
Where
x
n
182 182 182 182 182 182 182
7.4 18 18 18 18 18 18
RSD
Correlation test
n
5.3 65 65 65
pH (EcloxTM)
182 182 182 182 182
65 65 65
pH (Eclox™) pH !field) Total chlorine Free chlorine TDS HPC22
182 182
x 7.4 7.4 7.4 0.12 0.08 48 5
RSD
R2
p
5.3
<0.001 0.003 0.031 0.004 0.005 0.21 0.008
0.98 0.47
5.3 5.3 62 83 58 530
0.02 0.42 0.36
See Figure 1
<0.001 0.23
See Figure 2
n = number of samples tested X = mean of sample results RSD = relative standard deviation of sample results P = Probab;l;ty of correlation, where <0.05 would be considered s;gnificant R2 = linear correlation factor contribution to overall variation, where 0.5 =50% contribution
Ocher tests available in the kit, along with the manufacturer's claimed test ranges, are: • Arsenic test strips (Range 0-4 mg/L, LOO = 0.1 mg/L) • Pesticide/ nerve agents test scri ps (presence/absence indication only) • C hlorine test kit {Range 0-8 mg/L free ch lorine) • Colour comparison chart (Range 0-90 Hazen) • Total dissolved solids meter {Range 02000 mg/L) • pH meter • Mustard gas rest kit (LOO= 1 mg/L) (Note: chis kit was withheld in the country of origin and was not available for assessment). Further in formation can be fou nd on the company website: http://www. severn tren tservices. com/i nstru men cation prod uccs/potable_warer_assessmen c/i ndex.jsp
IQ-Tox™ kit The IQ-Tox™ kit supplied by Kingwood Diagnostics, US, measures rhe uptake of a fl uorescencly labelled sugar by a live crustacean, Daphnia magna, to measure water toxicity. Such Daphnia exposed to contaminated water exhibit reduced or no fluorescence when exposed to UV light. Each rest uses 18 individual organisms; if fo ur or more organisms are adversely affected then the water sample may be toxic to humans.
The IQ-Tox™ kit, has greater sens itivity to some contaminants relative to Eel ox TM, according to the manufacturer's specifications and also co nfirmed by results from the US EPA ETV Program. This kit uses exotic crustaceans and is not deployable in the field, due to Australian quarantine requirements. However, we were able to eval uate it in our quarantine approved laboratory. Further information can be found on the company website: http:/ /www.kingwooddiagnostics.com/
Yean water sources, were tested in the DPI Environmental Health & Chemistry laboratory {not in the field) using the EcloxTM kit for pH, total dissolved solids (TDS) and luminometer % inhibition. Test repeatability was determined by conducting measu rements on seven samples, recalibrati ng rhe test equ ipment and rhen repeating the measurements. Reproducibility was determined by turning off the instruments and recalibrating and repeating the measurements the following day.
Experimental
Spiked water samples Spiked water samples were produced by addition to de-ionised water. The pesticides selected were technical chlorpyrifos {organophosphace insecticide), formu lated glyphosace (systemic herbicide of low human toxicity) and formulated paraquat {contact herbicide of high human toxicity). These pesticides are in common use and regarded as sentinel co mpounds for a range of herbicides and pesticides. The inorganic toxicants were arsenic {as arsenate) and cyan ide (as potassium cyanide) . These inorganic toxicants are regarded as high risk and potentially available fo r contamination of drinking water. The toxicant additions were made at several concentrations, including Suggested No Adverse Response Levels (SNARL) and estimated human lethal dose.
Drinking water samples As part ofYarra Valley Water's routine mon itoring program, samples are taken from various points throughout their distribution system and tested for a range of water quality parameters by their contract laboratory {WSL Consultants). Routine tests included total coliforms, E. coli, free and total chlorine, pH and turbidity. Total coliforms, place count and E. coli were determined using WSL internal laboratory methods. Free and total ch lorine were determined using the method APHA 4500-CI G and pH using method APHA 45 00-H• B. Over a five week period during October November 2004, approximately 200 duplicate samples, from a wide range of northern and eastern suburbs outlets supplied from che Silvan, Winneke and Yan
Three concentrations of each spiking agent were prepared, covering a concentration
Journal of the Australian Water Association
Water
MAY 2007 83
/Jakr CtJJdrtJIS,p/ulhw
•
s:: 0
80
••
:;;
:0 .r;
.E ~
...
0
Q) Q)
E
0
60
~-1Il.-•----1-1--.,~-.. ____,__ -..
40
- -
20
.E
0
::s
-20
..E.
E
..J
Data points ------ Regression
...
• 0
-----;c: 200
150
100
50
. . __
T otal Dissolved Solids (mg/L)
Figure 2. Relationship between total dissolved solids and luminometer % inhibition. range from the SNARL to several hundred or several thousand times the SNARL concentration. Each concentration was tested in triplicate using the Eclox TM luminometer and arsenic and pesticide scrips. Arsenic and pesticide scrips were used for all chemicals tested to determine if there was any cross reactivity of the cesc scrips to non-target chemicals. With the IQTox™ test kit, tests were only conducted in duplicate due to the high numbers of Daphnia required. If a positive response was not obtained at che highest concentration tested, add itional spiking solutions were prepared in an attempt to gauge che level at which a positive response could be obtained.
Results
Drinking water samples Laboratory results for pH, total and free chlorine, TDS and m icrobial hecerotrophic place count (H PC 22 ) were compared against the Eclox™ luminomecer % inhibition results for the majo rity of samples. Statistical regression analysis (linear and ANOVA) was performed to
0
.:;
1i .r;
s::
:::!:! 0
...
Q)
Q)
E .5 E 0
::s
..I
80 70 60 50 40 30 20 10 0 -1 0 -20
There was no correlation between field pH and EcloxTM pH. Although che average results for all samples are identical , the comparison between fie ld and Eclox™ pH for individual samples did not indicate any correlation. However, there was a variable elapsed time between field sampling and measurement in the laboratory, during which time che samples were stored at 4°C. le is possible chat samples could have adsorbed or desorbed carbon dioxide affecting the pH of these low TDS, unbuffered, samples. T he EcloxTM luminomecer measures the inhibition of light production caused by a sample, which can be influenced by che TDS and field pH and other parameters not measured, as shown in Figures 1 and 2. le is unlikely chat che Eclox™ luminomecer % inhibition would be strongly influenced by any single water impurity unless it was at a level well above normal drinking water values. Figure 3 shows luminomecer % inhibition plotted against sampling dace for che three
& Yan Yean
• Winneke
• Silvan
s::
determine if there were any correlations. Summary results are presented in Table 1.
•
• ••
• ;
.!,
u
0
0 N
; t5 ~ N
; .!,
u
0
0
"'
• ; > 0 z
...
9> 0
z
d,
'iz 0
d,
9> 0
...z N
Figure 3. Luminometer data and trend lines over time for different source waters. 84 MAY 2007
Water
Journal of the Australian Water Association
'i 0
z
d,
N
technical features refereed paper
source supplies. le shows a slight upward trend in readings as sampling progresses from spring cowards summer. This trend also shows a slight minimum during early November for all three sources. In general, Silvan shows higher luminomerer % inhibition during rhe sampling period indicating char the rest is detecting a difference in water quality between sources. N everrheless, ch is is nor reflected in rhe range of rest results available from the contract laboratory.
Table 2. Relevant human health reference levels and Dophnio mogno LC50 levels for tested chemicals and other potentially toxic contaminants. Compound
Arsenic Cyanide Chlarpyrifos Cypermethrin Glyphosate MCPA Paraquat Permethrin Simazine Strychnine
Tested toxic level (mg/ L)
>0.06 - <0.6 <0.2 >0.1 - <1 Not tested <9 Not tested >103 Not tested Not tested Not tested
Daphnia LCso (mg/L)
7.4 0.0017 0.00015 780 >190 6. 1 0.0006 >100 10
ADWG Health based (mg/ L)
SNARL (mg/ L)
0.007 0.08 0.01
0.06 0.2 0. 1 1.5 9 0.045 0.12 0.2 0.0 15 0.15
0.03
Estimated human
lethal dose (mg/ L) 30 86 7500 90000 2.8
From rhe range of readings obtained for rhe various water sources during the period of resting, ir would be reasonable co sec a value of 50% fo r rhe Eclox ™ luminomerer percentage inhibition as a trigger fo r further investigation. From Figure 3, it can be seen char 2 samples exceeded chis value during the rest period. One sample (sampled 28/10/04, source Silvan) was found co have one of rhe highest coral and free chlorine levels (0.44 mg/Land 0.36 mg/L). H owever, another sample with similar chlorine levels (0.48 mg/L and 0.36 mg/L, sampled 12/ 11/04, source Yan Yean) gave an Eclox™ luminomerer % inhibition of only 16%.
Arsenic
Cyanide
Arsenic has a SNARL of 0.06 mg/Land an estimated human lethal dose of 30 mg/L. For the rest range 0.06 - 30 mg/L, arsenic gave positive readings at all concentrations using the arsenic rest scrips but was nor detectable using che Eclox™ luminomecer. A toxic effect was readily noted using rhe IQ-Tox™ kit at all concentrations except the SNARL.
Cyanide has a SNARL of 0. 2 mg/L and an estimated human lethal dose of 86 mg/L. For che rest range 0.04-86 mg/L, pesticide rest scrips gave negative readings bur ir was readily detectable at all concentrations using the Eclox ™ luminomerer, and was readily detectable at the SNARL concentration using the IQ-Tox™ kir.
The ocher sample (sampled 19/11/04, source Winneke) did nor have any marked water quality parameters which would have led co such a reading, although it did have rhe highest soluble aluminium resu lt (0.099 mg/L, 6 samples rested) an d rhe highest zinc result (0. 18 mg/L, 12 samples reseed).
(Nore re arsenic rest strips. In order co achieve che required sensitivity a new batch of arsenic reagent was required, even though rhe initial batch was still within expiry dare. Based on rhe physical appearance, ir was suspected char it had been degraded by water absorption).
Spiked water samples The SNARL is rhe suggested maximum concentration of a substance in water that is considered to represent no significant risk to human health over a shore period. The respective human lethal dose (UKWIR Toxicity Database), SNARL (UKWIR Toxicity Database) and, for comparison, Australian Drinking Water Guidelines (2004) are shown in Table 2. T he Australian D rinking Water Guideli ne Val ues were not selected as dilutions for this experiment as it was felt these values would be below the sensitivity of rhe EcloxTM ki r. Eclox™ luminomecer % inhibition, arsenic/pesticide scrip readings and IQT ox™ results for the spiked water samples are shown in Table 3.
Glyphosote Glyphosace has a SNARL of 9 mg/Land an estimated human lethal dose of 90,000 mg/L. The pesticide rest scrips gave positive readings at 9000 mg/L. For the rest range l.8-9000 mg/L it was readily detectable by rhe Eclox ™ luminomerer assay an d was readily identified by the IQ-Tox™ kit at and below the SNARL. Ir is important co note char ocher components in rhe glyphosare formu lation , such as surfactants, may contribute co rhe toxicity reading. Diffe rent glyphosare produces, especially chose specifically formu lated fo r use in or near waterways, may produce different rest results.
Chlorpyrifos Chlorpyrifos has a SNARL of 0.1 mg/Land an estimated human lethal dose of 7500 mg/L. For the rest range 0.1-100 mg/L, ch lorpyrifos gave negative readings at all concentrations using rhe pesticide test scrips and was nor readily dececrable using rhe Eclox™ luminomerer, but was detectable using rhe IQ-ToxTM kit. The IQT ox™ result at the SNARL co ncentration was ambiguous, due co one of rhe replicate rests indicating toxicity (8 Daphnia affected) and the ocher indicating low toxicity (2 Daphnia affected). The EcloxTM lum inometer did nor identify toxicity in any chlorpyrifos spiked water sample, bur because ch lorpyrifos has a low solubility in water (1.4 mg/L), rhe higher concentration spikes were probably saturated solutions and the effective concentration in water was probably not much above 2 mg/L. In order co determine the likely sensitivity of rhe pesticide rest strips, a formulation of another organophosphorous insecticide, malathion, was reseed. Even at concentrations of 5000 mg/L, rhe rest scrip
Journal of the Australian Water Association
Water
MAY 2007 85
technical features
ra . - -. . . ...
water supply Table 3. Spiked water sa mple test resu lts. CONTAMINANT
CONC. mg/ L
#1 Arsenic
G lyphosate
KCN
C hlorpyrifos
Paraquat
,_
0.06 0.6 3 6 30 1.8 9 9000 0.04 0.2 86 0.1 1 10 100 0. 12 2.8 28 102.5 1025
positive positive positive -
positive not tested negative negative positive negative negative negative negative negative negative negative negative
-
negative negative positive positive
IQ-TOX Number of active Daphnia2
LUMINOMETER % INHIBITION 1
ARSENIC/ PEST STRIP
27 2 19 46 24 77 84 100 94 95 100 -39 23 33 8 -37 -4 20 21 72
#2
39 6 5 9 29 63 85 99 97 92 100 -18 55 50 15 -30 -4 19 29 72
#3
43 37 17 8 27 76 92 100 98 99 99 -29 51 46 11 -43 -12 45 38 74
Mean± %RSD
36±23 15±128 14±55 21±103 27±9 72± 11 87±5 100±1 96±2 95±4 100±1 -29±37 43±41 43±2 1 11±3 1 -37± 18 -7±69 28±53 29±29 73±2
Toxic N N N N N
y y y y y y N N N
#1
#2
Toxic•
15 0 0 0 0
17
N
0 0
y y y y
nt
nt
nt
0 0 13 0 0 10 0 0
0 0 17 2 0 16 0 0
y y ? y y ? y y
N
nt
nt
nt
N
18 17 18 18 13
17 16 17
N N N
nt
N
nt
y
N N N
y
SNARL Suggested No Adverse Response level lethal dose Estimated human Lethal Dose (Glyphosate lethal dose is 90,000 mg/l; chlorpyrifos lethal dose is 7500 mg/L) 2
nf
>50% indicates potential toxicity, based on our baseline data obtained from water system samples. If the number of Dophnio adversely affected is 4 or more (i.e. < 15 organisms out of 18 fluorescing), then the sample is regarded as potentially toxic. = not tested
result was ambiguous. Malathion has a health based SNARL of 9 mg/L.
Discussion EcloxTM kit
Paraquat
T he EcloxTM kit proved simple to learn and to operate, both in the laboratory and in the field.
Paraquat has a SNARL of and an est imated human lethal dose of 2.8 mg/L. For the rest range 0.12 - 1025 mg/L, paraquat gave positive readings with the pesticide test strips at the two highest concentrations, was only detectable using the EeloxTM luminometer at the highest concentration rested, and was only detectable at the two highest concentrations using the IQ-Tox™ kit. No test indicated toxicity at the SNARL or lethal dose concentrat ions. The inability of any rest to readily detect paraquat contamination is of concern. Positive readings using the pesticide rest strips were o nly ob tained at concentrations 50 rimes higher char the estimated human lethal dose. Since the paraquat used was in the form of paraquat dichloride, a sample was treated with chlorine removal agent to determine if there was any effect. There was no significant change in Eclox luminomerer percentage inhibition from untreated and treated paraquat spiked water samples.
86 MAY 2007
Water
The arsenic rest strips were qui re capable of detecting arsenic contamination, although the stability of the arsenic reagent could be of concern. The pesticide rest scrips are obviously only suitable for detecting high levels of some types of pesticide contamination or identifying concentrated products. These strips are designed specifically to detect organophosphate insecticides and chemical nerve agents. One cause of concern was the kit instructions required the strip to be incubated under the armpit fo r 3-4 minutes. Because of the toxicity of the chemicals being tested, this was not done for OH&S reasons and the strips were incubated at room temperature. However conducting the test at 40°C using an incubation oven gave no difference in test result. Repeatability and reproducibili ty data for Eclox™ rest pH, TDS and luminometer in hibition are reported in Tables 4 and 5.
Journal of the Australian Water Association
The reproducibility and repeatibility data and the replicate spike analysis data in Table 3 indicate that the Eclox ™ luminometer percentage inhibition has an inherent variability that must be considered when interpreting luminometer results. The luminometer repeatability for 30th December 2004 shown in Tab le 4 (range mean=4 l) is greater than normally co nsidered accep table. H owever the reprod ucibility fo r the luminomete r percentage inhibition for 4th and 5th Jan uary 2005 shown in Table 5 (range mean =-0.1 ) is well within the accep table range for a scientific instrumen t. There is no apparent reason fo r the poor repeatability of results on the 30th December 2004 and this is evident for TDS and pH in addition to the luminometer percentage inh ibition. The concern is that the three electronic meter based tests can all be subject to such apparently rando m variation . T he rep roducibilities for pH , TDS an d luminometer fo r the samples rested on 15th November are within an acceptable range. I t would be sui table to check th e repeatabili ry on a daily basis to define the accep table range fo r each of the pH, TDS
technical features refereed paper
Table 4. Repeatability
(reca li brate a nd re-read same day) of Eclox™ meter pH , total dissolved solids and lumi nometer in hibition (range mean = (X 1-X2J/n) .
pH Sample Date
Source Water
Date Tested
Reading l
Reading 2
TDS (mg/ L) Reading l Reading 2
%
Luminometer % Inhibition Reading l Reading 2
26/ 10/ 2004
Silvan
15-Nav-04
7.9
7.5
30
30
24
12
26/10/2004
Silvan
15-Nov-04
7.4
7.6
40
30
11
21
26/10/2004
Silvan
15-Nov-04
7.5
7.6
30
30
10
12
26/ 10/2004
Si lvan
15-Nov-04
7.5
7.4
30
30
21
15
27/ 10/2004
Si lvan
15-Nov-04
7.5
7.4
30
30
20
6
27/10/2004
Silvan
15-Nov-04
7.5
7.6
30
30
12
16
27/10/2004
Silvan
15-Nov-04
7.5
7.6 0.01
30
30
9
8 2.4
Range mean
1.4
28/10/2004
Silvan
30-Dec-04
7.3
7. 1
60
40
28
11
28/10/2004
Silvan
30-Dec-04
8.0
7.2
40
40
16
28/10/2004
Silvan
30-Dec-04
8.4
7.1
60
40
16
16 .39
28/10/2004
Silvan
30-Dec-04
8.2
7.1
40
40
73
.57
28/10/2004
Silvan
30-Dec-04
8. 1
7.5
50
50
19
3
28/ 10/2004
Silvan
30-Dec-04
8.4
7.3
40
40
16
-48
28/10/2004
Silvan
30-Dec-04
7.8
7.2
40
40
27
22
Range mean
0.8
Table 5. Reprod ucib ility (instruments stored , recalibrated lumi no meter % inhibition (rang e mean = (X 1-X 2)/n).
5.7
a nd re-read next day) of Eclox™ meter p H, total dissolved solids and
pH Sample Date
Source Water
Date Tested
41
Reading l
Reading 2
TDS (mg/ L) Reading l Reading 2
Luminometer % Inhibition Reading l Reading 2
9/11/2004
Silvan
4/5-Jan-05
7.1
7.2
30
40
11
4
9/11/2004
Silvan
4/5-Jan-05
7.2
7.1
30
40
7
15
9/ 11/2004
Silvan
4/5-Jan-05
7.2
7.1
30
30
13
11
9/11/2004
Si lvan
4/5-Jan-05
7.2
7.0
30
40
38
39
9/11/2004
Silvan
4/5-Jan-05
7.1
7. 1
40
40
13
12
9/1 1/2004
Silvan
4/5-Jan-05
7.1
7.0
30
40
9
16
9/11/2004
Silvan
4/5-Jan-05
7.1
7.1
30
40 ·7.1
19
Range mean
and luminomecer meters and proceed with diagnostic resting if the repeatability criteria are met.
0.06
The data produced in chis study are comparable with chose obtained by ocher researchers (Scares et al, 2004; van der
14
-0.l
Schalie et al, 2006), who also found variable sensiriviry between different contaminants. In our srudy, paraquat was nor detected except ar a concentration of I 025 mg/L, while van der Schalie et al (2006) calculated a minimum derecrable concentration using the EcloxTM of 744 mg/L. Eel ox™ luminomecer percentage inhibition gives an overall indicator of water quality, bur a major issue is what level of reading is a cause of co ncern. Readings will vary depending on the source water and seasonal influences. Use of the luminomerer as a rhrear assessment cool will require awareness of che "normal" readings obtained from the water distribution system in question. Scares et al (2004) also concl uded char "In order co assess the significance of an individual screening result from a water system sample, it is necessary co first establish the baseline value for chat site."
Figure 4. Effect of chlorpyrifos o n Daphnia magna . Control cell on left, showing hea lthy Daphnia magna after uptake of fl uorescent sugar. Cell on rig ht with water spiked with c hlorpyrifos at l mg/L, with all Daphnia magna dead.
IQ-Tox™ kit Figure 4 shows rhe sensitivity of Daphnia magna co a chlorpyrifos test solution. The
Journal of the Australian Water Association
water
MAY 2007 87
technical features
IG!I ••..••••.
water supply Daphnia magna used in the IQ-Tox™ kit are known to exhibit a w ide range of sensitivities to various chemicals. Some li terature values o f lethal concentrations (LC50) (The Pesticide Manual, 2003) of various chemicals are shown in Table 2, with their corresponding human SNARL and estimated lethal d ose, to illustrate this point. Note chat most of these LC 50 values will have been obtained using p ure standards of the chemicals being cesced, nor formulated produces, and most measure lechalicy over a 48 hour period, no t che 15 min ute exposure specified in the IQ-Tox™ kit. From a comparison of Daphnia LC50 and human SNARL concentratio ns in Table 2 and as we have demonst rated in this smdy, Daphnia may not necessarily provide a response to contaminants, such as p araquat, at levels which would be of serious concern to human health . Conversely, a significant response may be shown by Daphnia to contaminants to which they are particularly sensitive at levels which are not particularly dangerous to human health. For example, based on the data in Table 2 , Daphnia wo uld be expected to show a response to the synthetic pyrerhroid cypermerh rin at a level l/ 10,000ch of the human SNARL.
Conclusions W ith the exception of paraquat, the I Q-
T oxTM was more sensi rive than the Eel ox™ test kit to the conraminanrs tested . However, the IQ-Tox™ is not suitable for field rest deployment as it uses exotic organisms which must be conrained in a quarantine facility. T he use of native Daphnia species could be invesrigared for this kit if field deployment was considered desirable. (During the peer-review process the authors were advised char a IQ-ToxTM kit using native Australian Daphnia has recencly been made available) . Both kits provide a positive/negative assessmenr of the water sample. H owever, there is the risk of false negative results being generated for contaminated water samples which would be of concern to human healch. Therefore, a negative result does n ot necessarily guarantee that the water is fi r fo r consumption. A positive result would indicate the need for furthe r resring, bur does nor necessarily prove that the water is unfit fo r consumptio n. Similar conclusions have been reached by other research ers. The precaution ary principle would apply in circumstances where a
C&SBRAND AUSTRALIAN FILTER COAL FOR DEEP BED COARSE DUAL MEDIA FILTRATION "More UFRVs for your money, and better quality water"
C&S BRAND GRANULAR & POWDERED ACTNATED CARBONS JAMES CUMMING & SONS PTY LTD
Email:
QUALITY ENDORSED COMPANY AS/NZS ISO 9001 STANDARDS AUSTRALIA
jamescumming@jamescumming.com.au
Licence no: 1628
319 Parramatta Rd AUBURN NSW 2144 Phone: (02) 9748 2309 Fax: (02) 9648 4887
88
MAY 2007
Water
Journal of the Australian Water Association
positive result was obtained, and rl1reat assessments should consider the weight of evid ence available from all sources.
Acknowledgments T he assistance and support of Ian Filby and Konrad Gill from Yarra Valley Water for arranging the water samples and WSL rest resulcs is gratefully acknowledged. T he Victorian D epartment of Human Services provided the fu nding which made this work possible. Robert McKenzie from the Department of H uman Services provided valuable feedback on the written report.
The Authors
Colin Cook (email colin.cook@dpi.vic. gov.au), Damien Wren and Gavin Rose are scientists at the Victorian Department of Primary Industries W erribee Cenrre, Victoria.
References Ausrralian Drinking Warer G uideli nes, National Health and Medical Research Council, 2004 Van der Schalie, W . H ., James, R. R. and Gargan II, T. P.; Selection of a barrery of rapid toxicity sensors for drinking warer evaluarion; Biosensors and Biolectronics 22, p 18-27 (2006) States, S., Newberry, J. , Wichterman, J., Kuchta, J ., Scheuring, M and Casson, L; Rapid analytical techniques fo r drinking warer securiry investigations; journal American Water Works Association 96(1), p 52-64 (2004) The Pesticide Manual, British Crop Protection Council, 13 th ed, 2003 Unired Kingdom Warer I ndustry Research (UK\'if!R) T oxiciry Database
A Local Test Kit To overcome restrictions limiting the use of the IQ-Tox™ assay to quarantine approved p remises, Dr Rick Krassoi of Ecotox Services Australasia, adapted it using a locally collected variant of the water flea, Ceriodaphnia dubia, found in many parts of Australia
C. dubia was first used by the Ecotoxicology Laboratory of the then NSW Seate Pollution Control Commission in the mid- l 980s, and is now used widely in laboratory ecotoxicity tests for the assessment of effiuenrs, ground waters, receiving waters, leachates, chemicals and for use in Toxicity Identification Evaluation studies aimed at identifying the causes o f any observed toxicity. Adaptation of the local species to the fluorimetric test kit to now allows for unrestricted use of IQ-Tox™ in the field, anywhere in Australia.