Water Journal June 2001

Page 1







www.grundfos.com contact-au@grundfos.com


up to 600kW for flows up to 3500Usec .

Grundfos Sarlin submersible pumps -

sewage, raw water and industrial effluent.

complete solutions for pumping of

Through the acquisition of Sarlin Pumps, the Grundfos Group is now able to provide

Grundfos is one of the world's leading manufacturers of pumps and pumping systems.

Volume 28 No 4 June 2001




AWA ;:e-t


and access news, calendars, bookshop and over 100 es of Information at

Visit the

Water is sent to all AW A members eight times a year. It is also available via subscription.


Australian Water Association (AWA) assumes no responsibility for opinio ns or statements of fa cts expressed by contributors or advertisers. Edit0rials do no t necessarily represent ofTicial A WA po licy. Advertisements are included as an information service to read ers and are reviewed before publication to ensure relevance to the water environment and objectives o f AW A. All material in Warer is copyright and should not be reproduced wholly or in part without the written permission of the General Editor.

C hris Davis

Executive Director

Barry Norinan

Federal President

ABN 74 054 253 066

Australian Water Association Inc

is pu b lished in January, March, April, June, July, Septemb er, O ctober and Decembe r.

Water (ISSN 0310 - 0367)

Graphic design: Mitzi Mann

Hallmark Editio ns PO J3ox 84, Hampton, Vic 3 188 Level I, 99 Bay Street , Brighton, Vic 3 186 T el (03) 9530 8900 Fax (03) 9530 89 11 Email: hallmark@ halledit.com.au

Water Advertising & Production

PO Box 388, Artarm on, NSW 1570 Tel +61 2 941 3 1288 Fax: (02) 941 3 1047 Email: in fo@awa.asn.au

AWA Head Office

Email: brecs@awa.asn.au

W (Bill) Rees PO Box 388, Artannon, NSW 1570 T el +61 2 9413 1288 Fax: (02) 9413 1047

Crosscurrent Editor

EA (Bob) Swinton 4 Pleasant View C rcs, Wheelers H ill Vic 3150 T el/ Fax (03) 9560 4752 Email: bswinton@ bigpo nd.net .au

Features Editor

Peter Stirling PO Box 84, Hampton Vic 3188 T d (03) 9530 8900 Fax (03) 9530 8911

Managing Editor

Submissions sho uld be made to E A (Bo b) Swinton, Features Editor (see below for derails) .


•, Water is a refe reed journal. This symbol indicates t hat a pap er has b een refereed.

F R Bish op, C hairman B N A n d erson , R Con sidine, W J Dulfer, G Finke, G Fin layson , G A H o lder, B Labza, M Muntisov, P Nadebaum , J D Parker , J Rissman , F R o ddi ck , G 11..yan , E A Swinto n ,

Editorial Board

Journal of t he Australian Water Association

IWA Conference, Sydney: Odour and VOCs, G Finke


Repor ts by B Swinton, B Rees, M M etz



Magnetic ion exchange removes NOM, alum removes the turbidity

C Pelekani, M Drik as, D B Bursill


The history and structure of the Australian Drinking Water Guidelines Framework

B M cRae, P Ca llan, D Cun liffe, S Rizak, D Bursil l, A Neller


Reviews the current frameworks in the States and Territories

C Piccinin


if water management. Tlte 19tlt A WA Co11veutio11 i11 C1111berra fornssed more 0 11 policies for future 1na11agement of rvater resources tlta11 on tech110/ogy. R eports i11 this issue cover rite keynote speakers and the Oz water Exhibition, and we publish the wil111ers of the Michael Fly1111 Award. Photo: Irene Lorbergs P hotograp hy, Canberra .



Applying the Victorian EPA protocols for assessing impact of effluents

D Vertessy


Explaining the statistics involved in the new Guidelines

DR Fox

OUR COVER: 2001 Water Odyssey. vVe euter i11to rite next milleniwn













CROSSCURRENT: Water News Around the Nation

IWA Conference, Taipei: Sludge Management, D R Dixon, D J Lee



IWA Report

WEF Report




MY POINT OF VIEW: Setting Standards? D




2 3




One of my first substantive tasks as the new AW A President was to take part in a strategy m eeting among the Australian Water Industry Forum (AW I F) partners, [AA, ANC ID and AWA. T he three associations are all national and, between them, cover both urban and rural water issues. The W ater Services Association, WSAA, is indirectly involved through its sustaining m embership of AW A. Although AWIF has been nominally in existence since last year, progress has been slow to the usual resource and time constraints in volunteer associations. The day-long session which we h eld in Melbourne was very useful and set us on the path to a meaningful fu ture. First up, we re-affirmed our attributes: an informal alliance among national water associations, aimed at addressing high-level issues with relevance to all the partners. Equally importantly, AWIF is not another association, and exists only to provide a single point of contact between the broad water indust1y, government and the m edia. We agreed to rotate the Chair and the secretariat function among the partner

Australian Water Industry Forum

For the first time, the Canberra Ozwater convention coincided with the changeover of Presidents, from Allen Gale to m e. It was rewarding to be able to celebrate the changing of the gavel with many of our members, rather than as part of a Council (now a Board) m eeting. Apart fro m the ritual change-over, I was also ve1y pleased with the Convention as a whole. The Canberra organising Committee, led by Prof Ian White, and Q uitz Even ts Management, led by M argaret Bates, teamed up ve1y effectively to deliver a very stylish event, with good papers (poster and platform) and flawless logistics. My personal thanks and congratulations to th em and to al l who participated. For Perth in 2003, our next Ozwater event, Barry Sanders w ill achieve the unique distinction of being the first person ever to chair three biennial conventions. We will be holding the convention and exh ibition in Burswood C asino which is well equipped for all aspects of an excellent func tion. I will be liaising with Bany and his committee, not only to ensure the normal high standard, but also to can vass ideas for some innovative activities, Eke 'open space technology' (free-wheeling group discussio ns) and more interactive workshops. Your suggestions would be most welcome




associations, on an annual basis, and AW A is first cab off the rank in that respect. Thus I will act as spokesm an for AW IF for the next twelve months, along with AWA Executive Direc to r Ch ris Davis. In general, AWIF aims to be proactive and setting the agenda, rather than reactive to past events. In our planning meeting, our original intention was to address three key areas of national water policy: trading; the N ational Action Plan for Water Quality and Salinity; and the pending fo rmation of a Natural Resou rce Managem ent Cou ncil. It did not take long for us to realise that an over-arching unanswered question is, what constitutes sustainable management of Australian water resources? After some discussion we generally agreed that the best operational definition of sustainability is managing resources nowto ensure a good quality of life, without compromising the ability of our descendants to enj oy the same natural resources in perpetui ty . Actually applying that principle to water resource management is much harder to understand than we had expected, so that has become a "work-inprogress", fro m which we hope will flow intelligent comment. A reasonably simple issue for AWIF, though, was the need for an independent, expert advisory panel to service the proposed N atural R esource Management Council. The Council, which seems likely to take over much of what the current ARMCANZ and ANZECC have been doing, is slated to include two ministers from each jurisdiction. A WIF is keen to see that a completely independent panel of experts advises the new NRM C,

Barry Norman

The Water jo urnal welcomes the submission of papers equivalent to 3,000-5,000 wo rd s (a llowing for graphics) relating to all areas of the water cycle and water business to be published in the journal. T opical stories of up to 2,000 words may also be accepted. In the fi rst instance, email a draft copy to the Features Editor, Bob Swinton (email bswinton@bigpond.net.au). Following his assessment of suitability, he will table the paper at a monthly J ourna l Committee meeting where, if appropriate, it will be assigned to referees. T heir comments will be passed back to the principal author. If accepted and after any conm1ents have been dealt with , the final paper can be emailed with the text in MS Word but with high resolu tion graphics (300 dpi tiff, jpg or eps files - Z ip disks or CDROMs can be accepted) in separate files (N.B. graphics embedded in Word files can not be used) or h,trd copy photos and graphics suitable fo r scanning by the publisher can be mailed to 4 Pleasant View Cres, Wheelers Hill, Vic 3150.

Contributions Wanted


Barry Norman

Branch Presidents to Meet One of the ongoing tensions that arise in AW A, as for most national associations with regional branches, is relationships between the regions and the central operation. Although I am not as steeped in the culture of AWA leadership as some of my colleagues, it did not take long for m e to perceive that this can be counterproductive. To ensure that we m inimise the amount of energy devoted to internal fri ctions, I have convened a series of regular teleconferences between all the Branch Presidents, me, and senior staff from our Arta1mon office. I hope that will enable eve1yone to be kept across current developments and to air any concerns early, rather than allowing them to fester and emerge later in a waste of otherwise useful en ergy. Assoc iations like ours are chronically under-resourced to carry out the long list of good works to which we aspire. It's therefore essen tial to have everyone 'singing from the same sheet of 111us1c'.

including a blend of academics and practitioners from the relevant disciplines.





Flow Control




national convention centre

Water & Rivers Commission, WA

South East Water Limited

Egis Consulting Australia Pty Ltd

Brown & Root Services, Asia Pacific Pty Ltd




H,IJllol C• ••nlrlu Hd,1 • 1, ,4. ,1,•II•

Natural Heritage T111st



Environment " - Australia












The author was awarded the Michael Flynn Award at the 19th Convention for the best platform presentation. This version has been edited and updated.

Australia's fede ral system of government has given rise to diffe rent forms of economic regulation for the urban water industty. N evertheless, there is a clear trend toward independent price regulators. There is also an emerging trend for the same economic regulator to have the role of monitoring customer service standards. M ost State and T erritory jurisdictions have an economic regulator with oversight fo r urban water. The most common form for that regulator is to have oversight spanning a number of industries with m onopolistic features. Western Australia is the only Australian jurisdiction w ith a water specific regulator and South Australia is also an exceptio n in having no independ ent regulato r. Q uite sensibly, small local government owned water service providers are generally exempt from econo mic regulation in all jurisdictions, the exception being Victoria where such water service


l n Au stra li a's federal system o f govern ment responsibility for oversight of water and sewerage services rests primarily with State and Territory governments. Given the diversity of demograp hic, socioeconomic and geographic profiles among the various jurisdictions, it should com e as no surprise that the regulato1y frameworks for water services in Australia va1y significantly between States and T erritories. Given the complexity of the regulato1y task and the disparity of the various water markets both within and between the various States, differences in approaches are to be expected and a degree of flexibility should be welcomed. In addition, this diversity of regulatory approach has provided a virtual laboratory experiment of regulation across the nation. This experiment offers a unique opportunity to observe and learn from the experiences of the various approaches taken in different jurisdictions. Key words: Economic regulation, refonn, pricing, customer service standards and cost benefit analysis.


While economic regulation of the Australian water indust1y is State, or Te rrito ry, based, the Nationa l Competition Policy agreem ent reached between t he Commonwealth, the States and the Territories ensures t h at there is a commo n an d consistent aim for all the regulatory regi m es . This co mmo n aim is based o n th e natio nal competition principles and

The COAG Ref orms



Nasier of En,ironmenia ana9emen1 (NEN)

Tel: 02 938S 4973 Fax: 02 9663 IO I Email: r.harding@unsw.edu.au

For further information please contact: Institute of Environmental Studies, UNSW

Combine a solid foundation in environmental management with a choice from a wide range of el allowing a tailor-made program to suit your needs.

- Masters, GradDip and GradCert in Environmental Management.

The University of New South Wales is offering 3 pr grams

Master of Environmenta Mana9emen1 {MEM)

Century. D evelop, build and enhance your care ~ o meet t his challenge w ith a

their application in respect of the water providers have long been amalgamated into industty is perhaps best articulated in the sizeable regional water service providers National Competition Council's (NCC) which are expected to come under similar Water R esource Policy attachment to its regulat01y a1Tangements as the metropolitan Com.pendium of Agreements. water busin esses. A number of particular developments point the way to the fu ture. There is a need to stress that regulation The first is for the economic regulator to of the urban water industry extends set customer service standards as well as to beyond economic regulation to span monitor them. The second development both public health and environmental is the proposal to set customer service areas. While this paper does not deal with standards with in a cost benefit analysis how these areas are regulated, one sho uld framework after seeking the actual views not lose sight of the fact that these forms of custom ers. T he last interesting develof regulations impact significantly on the opment is the contracting out of the industty's cost structure and , accordpractice o f economic regu lation w hile ingly, eco nomic regulation must be maintaining the jurisdictional right to mindful of these cost pressures . determin e th e shape of that regulation . In The Water R esource Policy was shaped terms of over-the-horizon aspects of by a number of agreem ents reached at economic regulation there is a clear need different m eetin gs of the Council of for the explicit integration of a w hole-ofAustralian Governments (CO AG) and the government approach to regulation . That Agricultural and Resource Managem ent means standards set by governments, or othe r r egulators , which translate into higher costs must be taken into cons ideration b y the price setting mechanism (be it an independent Environmental Management for sustainabilit regu l ator or greatest challenge to all profess ionals as we e ter the 2 1st government).

C Piccinin



Fits in a 50mm Well Pumps up to 18m Inexpensive Complete with leads & alligator clips

Push on 12 mm


- High & Med Capacity, up to 700cm2 - 0.45 micron element - No intermediary vessels requi red - Fully disposable

Disposable Groundwater Filters

Grab a Purger!

Part Number - SP6012J


Portable 12 V Pump for Purging & Sampling

major cities now charge a two-part tariff for their residential customers. A recent development in pticing has been the introductio n of cost reflective pricing by H unter Water w ith its location based prici ng fo r large industrial customers. W hile pricing regulation in Queensland is still in its infancy, the Queensland Competi t ion Au t hority (Q CA) ha s indicated that it would allow the introduction of cost reflective pricing system s. Full cost recove1y explicitly includes externalities - that is the inclusion o f all relevant costs and benefits associated with the service pro visio n that were not the prima1y intent of th e original tra nsaction. Examples of externalities include e nviron mental damage and social benefi ts accruing due to recreational facilities associated with , say, parks around water transport or catchme nt areas. W hile it is the intent of the COAG reforms that externali ties should be included in full cost recovery, ch is has yet to take place. T he refo rms furthe r requ ire that pricing for water services lead to earning of real rates of retu rn and be within a band that avoids earn ing mo nopo ly rents and ensures that the water business is at th e vety least finan-


Part Number - QMPSPJ


- The Ultimate groundwater portable sample pump - Micropu rge sampling gives high quality samples - Easy, rapid disassembly ( 1/4 turn , no tools) - Disposable bladder, low cost

Portable M lcroPurge• Pumps

Part Number - S101 J

- Very Rugged Meter - Long-Life Poly Tape with guide - Buzzer & Lamp - Tapes up to 900m - 2 Probe Options

Sol inst 101 Water Level Meter

provider, the principles explicitly leave it to the discretion of each State and T erritory govern ment to c hoose from contracting out, corporatisation or p1ivatisation. The reform pri nciples call for the institutional separation of water resource management, standard setting, regulatoty enforcement and service provision. T here is also a req uirem en t for inte r-agency performance compa1ison to ensure service providers see k to ach ieve international best practice (the NCC has adopted the annual WSAAfacts as the ind ustry publication designed to meet this requirement) . T hese reforms have been carried out for the urban water industry. The pri cin g reform principles further reinforce the commercial foc us of the CO AG reforms. The reforms req uire the adoptio n of a volumetric charge fo r bulk water and a two- part consumptio n based tariff for retail wa ter, the elim ination of cross subsidies, th e identifi cation of re main ing subsidies through a transparen t re porting of co mmunity serv ice o bligations (CSOs) and full cost recove1y. T he reforms introduced by COAG have had considerable impact on the urban water Industry's pricing practices. All

Whale Super Purger

Council of Australia and New Zealand (ARM C ANZ) . T he agreed consistent framework of reform applicable to the urban water industry covers a number of aspects aimed at promoting a more efficie nt , cus t o m er-dri ven serv ice provision . Th e reform principles target: • Natural resource management, • Pri cing, • T rading in water e ntitlem en ts (specific and immediate references to rural and irri gati o n, but future applications may be w id e r), • 1nscitutional reform, and • Improved publi c consultation. T he reform prin ciples for th e urban wa ter industry go to some degree of detail in resp ec t of institutional , pri cing and pu b lic consultation reforms and these are summarised in the follow ing paragraphs. T h e CO AG in sti tu ti o na l reform principles aim to promote a co mmercial fo c us fo r water service providers and also call for an integrated approach to natura l resource manage ment , i n clud ing in resp ect of water catc hment. In terms of th e in stitutiona l requireme n t fo r a co111m ercially fo c ussed water servi ce




Economic regulation in NSW is the responsibility of the Independent Price and Regulation Trib unal (l PART).

Economic Regulation in the Various Jurisdictions

cially and socially (inclusive of externalities) viable. Most urban water service providers achieve real rates of return. Finally, the fina ncial principles include the payment of tax, or tax equivalent payments, to ensure competitive neutrality be tween private and public sector providers. T he Commonwealth has put into place arrangements to ensure that tax equivalent payments from State owned enterprises are retained by the State. This has resulted in considerable progress in this area for State and Territory owned water businesses. However, no similar considera t ion has b een made regardi n g enterprises owned by local government, although the Comm.onwealth Treasurer has indicated that th is will be addressed in the near future. On the issue of public consultation, the reform principles include requireme n ts for co nsultation an d education of the public regarding the refo rms but also specifica lly in respect of changes to pricing and standards.

!PART determines pricing and return o n capital matters after public hearings during which interested observers can make submissions. It is important to note that IPART oversees other utility industries such as gas and electricity. To date IP ART has also monitored customer service standards set out in operating licences and customer contracts . In 200 1 IP ART will begin a series of reviews to determine customer service standards. It also seems that IP ART will seek to determine these service standards within a cost benefit analysis framework in which custom ers' preferences w ill be assessed (An approach which has WSAA's support). In V ictoria t h e Office of t h e Regulator-General (ORG) has responsibility for m onitoring perform ance in respect of customer service standards as set out in the operating licences of the three Melbourne retailers. The ORG has price oversight for other utilities (eg gas and electricity) bu t not for water. T he previous Government had stated its intenti on to eventually hand regulation of water p rices to the ORG but never did so . T he State T reasu ry previously detern1.ined prices after receipt of information by the fo ur Melbourne w ater businesses.


The process was not subject to public scrutiny. T he price setting role has been passed to the Department of Natural Resources and the Environment for the current price determination (prices had been frozen until J uly 2001). T he Victorian Government has an nounced it w ill establish the Essential Services Commission and that the Commission will become the independent water price regu lator as well as having responsibility for monitoring customer service standards. The Quee n sland Compe t ition Auth01ity (QCA) has the responsibility for economic regulation for utility industries such as water. T he QCA has an obligation to ensure adherence to competitive neutrality principles and responsibility regarding any third party access regime. The provision of wa ter services in Quee nsland is u ndertaken by local government and, to date, pricing is a matter for the individual Councils through their budgetary processes. H owever, a recent am endment to the QCA Act has given the QCA su rveillance powers over water prices. The QCA has recom m endatory powers in respect of publi cly owned enterprises but deterministic powers regarding private companies. Its find ings, however, would be pu blic. To date fo ur water businesses have been declared for such price oversight. T he State Government can declare w hich water businesses would come under the price oversight of the QCA under its own initiative, at the request of the local government or at that of the QCA (which is developing the criteria for determining which businesses ought to be declared for price surveillance). The Depa rtment of Natural R esources is the customer service regulator and has also respo n si b ility for strateg i c asset management (including the power to appo int third party or spot audits) . T he Q ueenslan d Water Refo rm U nit is proposing that any change to regulated sta ndards (except for public hea lth standards which are required to have a Regulatory Impact Statement) be subject to a cost benefit analysis to ensure the new m easure is cost effective. In Western Australia the Office of Wa ter Regulatio n (OWR) and the Minister for Water Resources share the responsibility for economic regulation of the statewide water service provider (the Water Corporation). The OWR licenses the Water Co rporatio n , setting the customer service standards and monitoring performa nce. The OWR can chastise, fi ne and, in extrem e cases, cancel the licence of the service provider. T he OWR. advises the Mi nister rega rding the water service provider's performance. Wh ile th is paper

Providing effective water inf rastruot11,, e management

only dea ls with eco nomic regulatio n, it service provider reports its performance pricing to the discreti o n of the water is worth noti ng th at all maj or environou tco mes in its annual reports. The shi ft authorities is T asmania. H owever, the m e n tal, resource and public health from AC TEW to the joint venture w holesaler fo r the city of H obart faces a incidents require separate reporting to the between ACTEW and AGL took place recommendati on by an independent O WR. However, the M inister (the on 1 O ctober 2000. With the new regi me price regulator and a determination by a Treasurer is consulted) determines th e there w ill be a new eco nomic regulator Minister). financial elements of economic regulations (the IC R C) w hich is still expected to O verall there is a definite trend (eg pricing, return on ca pital and CSO s) . contract out the pricing and rate of return towards independent price regulation It is also worth no ting that C SOs con side ratio ns. H o weve r, the joint w ith onl y W estern and So uth Australia account for almost a quarter of Water venture service provider will report to the giving no indi catio n of m o ving in that C orporation's revenu e and that an agreed regulator regarding perfoimance standards directio n. The m onitoring of custom er n1.e thodology d etermines the amo unt of and be subject to external audits and service standards has a clear trend towards CSO s and that this is not a subj ec t o f finan cial penalties fo r any breaches. independent regula to rs with N SW , dispute between the government and the In th e Northern Territo ry econ omi c Victoria, W estern Australia and, shortly, service provider. It sho uld be noted that regulatio n is also in the process of the ACT. Queensland has chosen departthe new W estern Australian Governm ent change. Price setting is being shifted from m e ntal oversig ht. Tasmani a's wa te r ca m e into power w ith a policy to ministerial oversight to the same TeLTitoiy retailers are quite small but South introd uce a multi-utili ty regulator. regul ator which regulates elec tricity Australi a, wi th a statew ide se rvice Sou th Australia also has a statewide prices. Th e C ommo nwealth Government provider, has opted fo r self- reportin g. w ate r serv ice provid er, SA W ater. provid es C SO payments to the water The interestin g development in this area C ustomer service standards are set by th e service provider. is that N SW will m ove to W estern Minister fo r W ater R esources in SA Australia's m odel o f an ind ependent The Regulatory Experiment W ater's perfo rmance state ment. The regulator setting (as we ll as monito rin g) do cumen t covers all other performance From the above snapshot it is clear that custom er se rvice standards. Another asp ects (including financial perfo rman ce, Australian regul atory models span th e interestin g develo pment is th e intentio n w h ich is negotiated with the Treasury) . entire gamut fr om co mpletely arms by N SW (and also a pro posal in In terms of eco nomic regul atio n, the length regulatio n of prices, return o n Queensland) to use cos t benefit analysis M inister for W ater R esources determines capital and monitorin g o f customer to evaluate customer service standards. In prices, sho rt-term return on assets and service standards to se lf determi nation of t e r m s of struc t ural form fo r th e C SOs after C abinet consul tatio n and prices and customer servi ce standards. ind e p e nd e nt reg ulators, th e o nl y su bmissions fro m SA W ater. There is no T here is a middle ground of Min isterial exceptio n is W estern Australia with a ind ependent oversight of customer service or departmental determ inatio n o f pri ces utility sp ecific regulator and this m ay be standards. In 1996 SA Water had been and customer service standards which about to chan ge. All other econo mic d eclared fo r p rice ove rsight by the suffe rs from the obvio us C ompetition C ommissioner. H oweve r, conflicts of interest o f after the C ommissioner issued a report on the shareholder wanting pricing principles, th e State Go vernment hi g h e r divid e n d s, a rej ected th e recommendations and the government having an de claratio n lapsed in 1999. o bliga ti o n r ega rdin g consum er pro tecti o n As in Quee nsland, local go ve rnments fr om m o n opo li st i c in T asmania have kept the responsibility behaviour and a political for th e provision of water services. a ve rsi o n to h i g h e r Th ere is no regu lar monitoring by pri ces . Th e re i s, reg ulators but water authorities ha ve however, a clea r trend inte rnal reports fo r custom er serv ice away from self-deterstandards and also provide reports to their mination towards either o w ner and customers. Local governments Mini s t e ri a l/ d e partset prices for custo mers o n th eir own mental declaration or to ini t iative . T h e Governm en t Pric es O versight C o mm.issio n sets target rates of arms length regulators (t he onl y exce p tion return for assets and recommends pricing being So uth Australi a principles and maximum revenu e levels which started with piice for H o bart Water, th e city's water ove r s i g h t b y a n wholesaler. Cabinet discusses these and • water supply independent regu lator the Minister announces the outcome The services only to shift to minisind ividu al owners o f H o bart Water terial determinati o n). (eigh t C ouncils in th e city of H o bart) • operations and receive dividend and tax equivalent Indeed, although all maintenance paym ents to satisfy competitive neutrality j urisdi ctions w ith small principles T he ACT has a regulator that l oc al g o ve rnm en t sets the scene for economic regulation but own ed water service the administration of those arrangem ents providers exempt th em is contracted out to the NSW economi c from eco nomic regulator in respect o f pricing and rate of re gul ation , t he only ret urn on capital. Customer service jurisdiction which seems standards are not regulated but the w ater content with leavin g


It seems clear that State and Territory responsibili ty for economic regulation of water has resulted in differe nt regulatory models. Nevertheless, there is a trend emerging towards independent regulators for pricing and customer service standards. The area where the experi ments are still at the developmental stage is in standard setting. T he most promising development is the intentio n to develop customer service standards by: • seeking the actual views of customers (rathe r than those of consulta t ive committees), • setting standards with in a cost benefit analysis framework. WSAA supports such approaches to standard setting. Given the importance of the Industry's underground infrastructure to its total costs and the potential of higher standards leading to premature renewal of such infrastructu re, it is im portant that proposed standards lead to net benefits for the community.

Concluding Comments

regulators span a n umber of industries w ith monopolistic features.

To promote disc ussion in this area, WSAA is undertaking a significant research project focussing on continuity of water service supply. T he study will develop and test a methodology for setting standards in this area of customer service by using cost benefi t analysis. Th e views of customers will be sought to determine which aspects of customer service are of greatest importance and how they can best be measured. T h e be nefits to the customers will be quantified and offset against the costs. Lt is hoped that the results of the research wi!J be completed by th e end of calendar 2001. An area that remains undeveloped in economic regulation is a w hole-ofgovernment approach between standard setting by all regulators and price setting. lt must be recognised that different regulators set standards which have the potential to drive up the industry's cost structu re. T hese regulators include those with environmental and pu blic health responsibilities as well those responsible fo r setting customer service standards . In setting standards it would also be appropria te for the regulators to be foc ussing


Claude Piccinin is the D epu ty Executive Officer of WSAA servicing the Strategic Committee on Industry and R egulatory Reform. H e has worked in the Plastics and Chemicals Industries Association, the Business Council of A ustralia, and as a Senior Adviser on industry issu es in the Department of the Prime Minister and C abin e t. H e graduated B.Ec. (1st Class H ons) from Monash University and has a M .Ec. from the AN U .

The Aut hor

on desired outcomes and how these could best be ach ieved within a broader perspective than by assuming that increasing past standards will necessarily lead to an improvem ent . Leaving aside, for the purposes of this paper, how these standards are set, it is also imp erative that whoever has the responsibil ity for price setting explicitly takes the impact of the additional costs caused by these standards into account and translates these into higher prices. In addition, best practice regu lation calls for this process to be both transparent and subject to appeal.

011 5 May 200 1, the N ational Health and J\lledical R esea rch Council released th e "Frarn ework f or Ma11agement of Drinking Water Quality " (Tl,e Framework) f or public comn1e11t. The Framework proposes a ne111 paradig111 f or e11m ri11g tl,e safety and quality of p11blic dri11ki11g water supplies. It emphasises proactive risk 111a11agemen.t, from catcl,ment to cons11mer, rather than the more traditional fo cus on monitoring of the end product . The approacl, is not 11011el in the context of public safety; indeed, this is tl,e basisf or much ef the quality ass urance undertaken fo r f oo d production. However, the Fra,nework is unique in that it is the.first document pronrnlgated by a national go11ernme11t that provides

Management Cou ncil of Australia and N ew Z ealand (ARMCANZ) and the A u s tralian a nd New Z e a l a nd Environment and Conservation C ouncil (ANZECC ) . W ith i n th e Fed e r a l Government, a number of age ncies are the focal point for specifi c guidelines of the National Water Quality Management Strategy, including: Agriculture Fisheries and Fo restry Australia , Environment Australia and the N ational H ealth and Medical R esearch C ou ncil (NHMR C) . The NHMRC and the ARMCANZ have responsibility for the A ustralian Drinking Water Guidelines (ADWG). As th e primary reference on drinking water





Climate Monitoring

Flood Warning Systems

Radio Telemetry Systems

Stormwater & Wastewater Studies

Hydrologic / Hydrographic Services

Lake , River & Wetland Monitoring

Coastal & Estuarine Data Collection

Th e N a tional Wat e r Qualit y Managem en t Strategy is the prin cipal po licy document governing water quality in Australia. The Strategy consists of a series of g uidelines addressing specific issues and is overseen at the highest level by th e Ag ri cu lture and R eso urces

Regulation of Water Quality in Australia

a methodology of this sort specifically desig 11ed for comprehensive 111anagernent ef its public drinking water supply. Tl,is article provides a discussion of /,ow this dornment was brougl,t to this stage, and wl,ere it proposes to take Australia .

B McRae, P Callan, D Cunliffe, 5 Rizak, D Bursill, A Neller



fl04l • • • • • •


FAX: (07) 3287 6445

683 Beenlelgh-Redland Bay Road, Carbrook Qld 4130 INTERNATIONAL: +6173287 6444



So, tap into our extensive experience and helpful seNice when next you require filter media.

Media produced to the AWNA B100-89 Standard. An extensive product range at competitive prices. Manufacturing In accordance with Quality System AS 3902/lSO 9002 Packaging alternatives to suit any requirement. Proficient technical assistance and support. Prompt delivery Australia-wide and Overseas.

• Decades of mineral processing experience.

'U'e au fdea4d t6 ~

Invest in accurately graded, durable media from your complete filter media professionals.


.::EE CALL: 008 on 744


requirements (from the 1984 W HO Guidelines to the 1996 ADWG). Compliance monitoring has been the primary mec h anism for managi ng drinking water quality, notionally for the protectio n of public health, both within Australia and in general across the world. Recent discussions have h ighlighted a number oflimitations associated with th.is approach, including: the limitations of sampling an d ana lytical techniques; inability to m onitor continuously for most parameters; failure to compensate for u neven distribution of 1nicro-organisrn.s and other contaminants within the supply; inadequate consideration of the range of events that impact on dri nking water quality; and failure to provide an effective response to contaminants without a prescribed numerical gu ideline value or established method of analysis. In addition, it must be considered that sampling and analysis is not always possible, either due to remoteness from analytical facilities or inadequate fi nancial and/ or staff resources. While the 1996 ADWG provide guidance on a range of holistic issues beyond compliance-based end point


quality in Australia, the ADWG p rovides guidance for the p rovision of a safe and high quality drinkin g water supply, which protects public health and meets the needs and expectations of consum ers. The ADWG provides information and advice on the multiple barrier approach, understanding the entire water supply system, quality system s, d isinfection protocols, community consult ation, performance mo nitoring and system evaluation and reporting. In addition th e ADWG includes health and aesthetic guideline values and supporting Fact Sheets for a range of microbiological, chemical, physical and radiological parameters. The term "guideline" is significant the Federal Government role with respect to drin king water supply is to provide guidance to the State Governments, where regulatory power is vested. States use a range of 1nechanisms to set service requirements for drinking water suppliers (utilities), the primary mechanism being licence conditions. Because licence conditions are imposed at different times (typ ically w hen a li cence expires), Australian utilities operate under differe nt


In response to a desire from the peak Australian health and water councils for a strategy that emphasises preventative quality management from catchment to consumer, the NHMRC D rinking Water Review Coordinating Group (DWRCG), ini t iated the development of the Framework in collaboration with the Coop erative Research Centre for Water Quality and Treatment. T he goal was to pro duce a quality managem ent approach that is specifically designed for the water industry . The resulting Framework is flex ible, providing guidance that facilitates development by an indi vidual water utility of a comprehensive and verifiable Drinking Water Quality Management System that is specific to it's drinking water supply syste m . It promotes proactive risk management by the utility, in concert w ith their affiliated stakeholders. It adopts a holistic approach to drinking water quality management that all ows m onitorin g to be placed in proper perspective, as a to ol that assists in assuring the performance of operational processes and verification of product. Monitoring becorn.es just one aspect of an overall approach for assuring the quality of drinking water and protection of public health. T his shift in emphasis necessarily requires an expanded focus, one that exten ds from catchment to consu mer.

Evolution of the Framework

monitoring, it has been observed that the practice of implementing the ADWG has resulted in an emphasis on numerical limits, with other advice often underemphasised or overlooked entire ly. Australia has a wide variety of utility circumstances (some 300 utilities with the largest 20 serving 60% of the population and the 200 smallest serving only 20% of the population). Regulation by numerical limits has been noted as a relatively inflexible approach, not readily accommodating the diversity of utilities, and generally promoting reactive, rather that proactive, management. Because of the time lag frequently associated with data analysis and reporting, there is a high risk that consumers will be exposed before monitoring results indicate a need to react; proactive risk managem ent emphasises preventi o n an d is perceived to be preferable for ensuring protection of public health. The Framework is a response to the above observations regarding the limitations of current practices. It is designed to promote best practic e throu ghout the industry ; reinforce rather than replace or revolutionise current good practices.

The use of quality management is now well established in business as an effective means of assuring product and service quality. A basic model of quality management embodies the philosophy of co ntinuous improvement and consists of generic elements such as: • Policy development, • Assessment, • Planning, • Implementation and operation, • Monitoring, P reventative and corrective action , and • Evaluatio n Quality management has become inc reasingly common in the Australian water industry, with many of the larger water suppliers implementing quality m anagement systems in their busi ness. This take up in the utility sector has been complemented by signs of cultural changes in the regulatory sector, both in Australia and internationally, favo uring a more proactive approach to the protection of p u blic h ealth , greate r att enci on t o preventive system management, and acknowledgemen t of catchments and consumers. In preparation for an initial draft of this document, established quality and risk m anagement systems and progra ms were reviewed against the requirements fo r drinking water quality management. This was undertaken co ensure that the Framework could be aligned and integrated w ith systems already implemented in the water industry and to help identify an appropriate scope and stru cture for the Framework. Systems examined included: the ADWG , ISO 900 1 (Quality Systems) an d I SO 14001 (Env i ro n m e ntal Management Systems), Australian and New Zealand Standard 4360 (Risk M a nagement) and the Hazard Analysis and C ritical Control Point (HACCP) System . R eview of the existing sys te ms provided valuable guidance . H owever, because these systems are generic, rather than industry specific, several significant gaps and limitations with respect to drinking water quality management were revealed. For example, ISO 9001 fails to address the preventative requirements of system analysis, hazard identification and control and risk assessment - all critical for effectively managing dri nking water qu ality. HACCP addresses these aspects and offers guidance on fundamental elen1ents associated with process control for consumer safety. H owever, it does not address non-safety issues of quality or other important areas of e mployee awareness, traini ng and community involvement, nor does it readily support the multiple barrier approach ou tlined in the 1996 ADWG. Several other areas considered important were not addressed within any of th e considered systems: stake holder involvement (other than customers), research and development, management of large-scale emergencies, and communication and procedu res for reporting. In add ition to th ese observations regarding " gaps" against categories that were considered necessary, the assessment revealed that areas that were covered did not always provide a good fi t to the specifics of d ri nking wate r quality management, further supporting the need to develop a custo m- tailored system. The unique challenges of drinking water quality management are particularly notable in the areas of the catchment and distributio n sys tem (the treatme n t c omponent, whi le s till possess ing challenges, provides better alignment to traditional manufacturing environments such as food processing). A drinking water utility receives suppli es (raw water) w hose quality is subject to a variety of activities that are often difficult for the utility to influence, are frequently complex and are difficult to understand or manage. Compared to other industries, water utilities have less flexibility to select alternative suppliers. Also, the concept of critical control points embodied within HAC CP does not lend itself readily to

Drinking Water Quality Monitoring Consumer Satisfaction Short-term Evaluation of Results Corrective Action

Verification of Drinking Water Quality

Operational Procedures Equipment Capability Materials and Chemicals Operational Monitoring Operational Corrective Action

Implementation · Operational Procedures and Process Control

Multiple Barriers Critical Control Points

Planning · Preventive Strategies for Drinking Water Quality Management

Water Supply System Analysis Review of Water Qual ity Data Hazard Identification and Risk Assessment

Assessment of the Drinking Water Supply System

Drinking Water Quality Policy Requirements Partnership Agencies

Commitment to Drinking Water Quality Management



catchments and distributio n systems. T hese aspects require a far more flexible definition and a range of'non- traditional' monitoring and co ntrol mechanisms. Finally, stakeholder considerations and the need for partnering have a primacy for water supply operations that is atypical of private sector manufacturing. This assessment led to a draft concept for the Frameu1ork that was presented and further developed at a national workshop held in October 1999. Desktop pilot trials were then initiated using four utilities, selected to represent a diverse range of ci rcumstances: Melbourne Wate r Corporation /South East Water (surface water, protected ca t c h ment), th e Katherine system of the No rt he rn Territory Power and Water Authority (small water supply) , Sydney Water Co rporation /Sy dney Ca t c h ment Authority (surface water, mixed use catchment) and a small plant operated by the Water Corporation of Western Australia (groundwater supply). Comments from the workshop, pilot trials and additional targeted consultation chat followed the pilot trials were used to refine the Framework, resulting in the current draft that has been released for public consultation. Throughout these varied consultations, there has been consistent and strong indust1y support fo r the need for the Framework approach.

Senior Executive Review Drinking Water Quality Improvement Plan

Review and Continual Improvement

Long-term Evaluation of Results Drinking Water Quality Management Audit

Evaluation and Audit

Documentation and Records Management Reporting

Documentation and Reporting

Investigative Studies and Research Monitoring Validation of Processes Design of Equipment

Research and Development

Community Consultation Communication

Community Involvement and Awareness

Employee Awareness and Involvement Employee Training

Employee Awareness and Training

Communication Incident and Emergency Response Protocols

Incident and Emergency Response

Table 1 . Framework for Management of Drinking Water Quality




The Framework includes guidance on the wide range of issues that need to be considered in managing drinking w ater quality from catchment to consumer. It addresses technical and operational issues together with other aspects such as corporate commitment and relationships with key stakeholders and with consu111ers. A central feature of th e Framework approach is that a utility will undertake a systematic and comprehensive analysis to provide understanding of the characteristics of their drinking water supply. The analysis includes physical mapping of the system, identification of potential hazards, assessment of the risks that might arise from th e hazards and the identification of pro cesses and practices that can affect drinking water qu ali ty. The assessm ent is used as a basis for planning and prioritising strategies to prevent or control the hazards, based on their likelihood of causing problems (i.e. their risk). Preventative strategies are based on the applicatio n of the multiple barrier approach and the identification of critical processes, including the m ec hanisms that w ill provide operational contro.l at th ese points. The scope and stru cture of the Frarnework is shown in T able 1. The Framewo rk includes 12 el e m ents considered good practice for system managem ent of drinking water supplies. Although listed as discrete entiti es, the 12 elem ents are interrelated and the relationship of each to achieving the goal of a safe and reliable drinking water supply shou ld be und erstood. Each element is important and each supports th e effectiveness of the others. Training, communication and documentation for example are an integral requirement of most elements. Th e need to address the eleme nts as a package ca n be demonstrated by the fact that multiple facto rs contribute to mo st water qua lity problems. T h e conce pt of c ontinu a l improv em ent is e mb o di ed in th e Fran·,ework. T he Frarnework does not contain static requirements w hich once m et never change . Rather implem.entation is an ongoing and iterative process of improvem ent achieved by continually evaluating and reviewin g performance. The Framework is intended to in.corporate sufficient flexibility for use by all water suppliers regardless of size and syste m complexity (i .e. both small su ppli es and large rangin g from minimal

Structure and Approach of the Framework

The Framework encompasses activities that will to differing degrees already be part of most utilities operations. [ndividual organisations can choose to implement the Framework in a manner that suits their own circumstances, e.g. som e parts may not be applicable to their system. H ow implementation is achieved will be dependent upon the needs of the organisation . This applies to the order in which the elem ents are taken up as well. It is up to each individual organisation to use judgement and decide how best to implem ent the elements. Factors involved in this decision may inclu de consideration o f the range of w ater supply systems involve d and managem ent systems already in place. As many of the principles and elements specified in the Framework are likely to be existing practises for the utility, all that m ay be needed would be to review, document and forma lise these practices an d address an y areas where improvem ents are required. In this sense, the Framework provides fo cus, drawing attention to the fact that alJ of these activities und ertaken by the utility are intended to support the overall goal of ensuring a consistently high quality water supply - o ne that is both safe and satisfying for consumers. The time and resources required to implem ent a Drinking Water Quality M anagem ent System (DWQMS), or integrate the Fra111ework with a utility's ex.isting management system, will depend on what is already being done within the orga nisation and how advanced it is with its management systems. Some elements of the Framework will require more work than others and improvem ents m ay need to be prioritised and implem ented sequentially. For example, with regard to implementation of preventive strategies a utility should first ensure that critical processes have been identified and that asso ciated process controls and operational procedures are establish ed. Controls and procedu res can then gradualJy be developed for remaining activities. The most important step would be getting started. D ocumenting current

Implementation of the Framework

treatment to full treatment). In addition, to reflect the wide range of differences between water supply systems across Australia and the varying institutional arrangements (corporations, local authorities, contractors, wholesale/retail, etc), the Framework is necessarily broad-based and the elem ents listed are not intended to be exclusive or prescriptive.


Drinking w ater is a universal good; everyone is a co nsu mer. Utilities are ultimately accountable fo r drinking water quality, bu t can 't hope to do this effectively in isolation. The Framework is a tool intended to promote ownership - across all parts of the system and inclusive of all parties. The Framework is a work in progress under the auspices o f the National H ealth and Medical R esearch Cou ncil and th e Dri nking Wa t er R eview Coordinating Group. The initial model was prepared b y t he Co-operative Research Centre for Water Quality and Treatment and supported by its Director

Further Developments

practice is often the most effec tive place to start. In doing this, and assessing the drinking water supply system, it is important to avoid getting involved in so much detail that making progress on implementing the Framework is inhibited. A DWQMS requires a long-tenn and ongoing commitm ent to ensu re effectiveness and continual improvement. One option may be to establish a Water Quality Committee (or organisationa.l unit) with responsibility and authority for the overall system. Commitment of senior executive management and communicating with all employees from the beginn ing would be essential fo r creating awareness within the o rga nisation. Employees should be encouraged to participate in the development of the DWQMS wherever possible. Their involvement and understanding contributes to an increased commitment and a sense of responsibility for the implementation of the system. Drinking water suppliers will need to show leadersh ip in implementing the Framework. However, in m ost jurisdictio ns this will require consultation and colJaboration with other agencies. H ealth authorities usually provide guidan ce on potential health impacts of drinking water parameters, while implementation of catc hment managem ent strategies is normally the responsibility of water resource and en vironment agenc ies, in collaboration w ith the community and ca tchmen t groups. In each case implementation of the Fra111ework will require d efinit ion of the variou s agencies' responsibilities. All relevant agencies need to be actively involved and encouraged to recognise their roles and responsibilities in regard to safeguarding drinking water quality. Partnership arrangements should be established .

The list below provides citations for son1e primary reference documents as well as some recent articles of relevance to th e subject, either the Fra111e111ork specifica lly or concep ts related to the document. T he list below is by no m eans exhaustive; a number of articles over the

References and Selected Sources

Professor Don Bursill. Samantha R izak has undertaken a substantial amoun t of the writing and research with assistance from Phil Callan, Dr David C unliffe, Dr John H oward, Professor Steve Hrudey, Dr Martha Sinclair and Roslyn Vulcano. Numerous other individuals and organisations have contributed to the present document, including those participating in workshops and the pilot trials. The Frr1111ework is currently available fo r pu blic consu ltation and can be obtained from : www.health .gov.au/nhmrc/ advice/waterbkd. htm. Comments on the Framerr10rk wi U be accepted u ntil 6 Jul y 2001. The au th ors s trongly encourage all readers of W ater to take this opportuni ty co contribute to this national initiative. T he DWRCG w ill oversee further development of the document based on comments received du ring the public c on s u lt a ti on. Upon finalisation, NHMRC and ARMCANZ are expected to endorse the docu ment. The DWR.CG is also undertaking a number of other initiatives, including a consumer guide to drin king water and revisio n of th e ex i sting 1996 ADWG chap ters to integrate the Fra111e111ork. The enti re ADWG are subject to a rolling review and the current work program includes consideration of a number of specific Fact Sh eets and other initiatives such as con sideration of chemicals associated with drinking water treatmen t. Interested readers are invited to contact Phil Callan (philip.callan@health .gov.au) regarding the ADWG or tl1e.DWRCG . As discussed above, the inform ation from existin g quality man agement and risk managem ent systems and the 1996 ADWG was used as a basis for developi ng this water industry specific approach. It sho uld also be acknowledged that the World Health O rganization, under the guidance of Dr Jamie Bartram, has been considering many of th e concepts inh e rent to the Fra111e111ork as part of the revi sion of the WHO Drinking Water Gui delines. Discussions with colJeagues involved in that initiative have been invaluable.

Brian McRae (A ustra lian Water Associatio n), Phil Callan (N ational Health an d M edical Research Cou ncil), David Cunliffe (Department of Human Services, SA), Don Bursill (CRC fo r Water Quality and Treatment) and Anne Neller (University of the Sunshine Coast) are all members of the Drinking Water Review Coordinati ng Group. Samantha Rizak (CR C for Water Q uality and T reatment) has had a principle role in drafting (and revising) the Frame111ork. The au thors w ish to acknowledge their colleagues on the DWRCG who were not involved with this paper: John Langford (Water Services Association of Australia), Alec Percival (Consumers' Health Forum) and Peter Scott (Melbourne Water).

The Authors

fsf/Codex/ lntroducingHACCP. pdf

World Health Orga111sation. 1997. HACC P /11trod11ci11g tl,e Hazard A 11alysis a11d Critical Co11trol Poi111 System (W/-10/FSF/FOS/ 97.2). Avai lable at: www.w ho.in t /

World Health Organisation. 1993. C11ideli11esfor Dri11ki11g Water Q11ality. W H O.

National H ealth and Medical Research C oun c il / Agriculture and R. eso urce Management Council of Australia and New Zealand. 1996. A11stm/ia11 Dri11ki11g Water C11ideli11es. NHM R. C/AR. M CANZ . Available at: www.hcalth.gov.au/ nhmrc/ advice/waterbkd.htm

Nadcbaum, et al. 2000. " Improved Management of Drinking Water Quality". Water. 27(4): 12-16.

AS / NZS ISO 1400 I: 1996. E1111iro11111e11ta/ Ma 11agement Systems - Speciflcatio11 with C11ida11ce for Use. Standards Australia / Standards New Zealand. Hrudey, S.E. 2001. "A R.jsk Management Approach: A Canadian puts forward h is p ragmatic view". Water. 28(1) : 29-32.

AS / NZS IS O 9001 ( Int) :2 000. Quality Mmiage111e11t Systems - Req11ire111wts. Standards Australia/ Standards New Zealand.

AS/ NZS ISO 9001: l 994. Quality Systems: Model for Quality Asrnm11ce i11 Desig11, De11elop111e11t, Prod11ctio11, l11stallatio11 a11d Seivici11g. Standards Australia/Standards N ew Zealand.

AS/NZS 4360:1999. Risk Ma11age111e11t. Standards Australia/Standards New Z ealand.

AJJen,JJ.; Clancy, J.L.; Rice, E.W. 2000. "The Plain Hard Truth About Pa t hogen Monitoring". Joumal of the American Water Works Associatio11. 92(9): 64-76.

Agriculture and R.eso urces Management Council of Australia and New Zealand. 1994. Natio11al Water Quality Ma11ageme11f Strategy: Water Quality Ma11ageme11t - A11 Outline of Policies. Available at: http: / / www.affa. gov.au/ docs/nrm/water/water_ reform/ nwqms/ nwqms_toc.htmJ

last decade and before, in th is journal alone, have addressed various aspects of this issue.





This paper was judged the best poster paper at rhe 19th AW A Convention.

The MIEX®DOC process (a registered trademark of O rica Australia Pty. Ltd.) has the ability to remove a significant fraction of the natural organic m atter (NOM) present in drinking water supplies. Some








0.5 1.1




Deep Monomedium

## Uniformity Coefficient (UC) : ratio of the 60% passing size to the ES (measure of size variation).

Effective Size (ES): size at which 10 percent (by mass) of the particles are smaller.


Anthracite UC #

1.1 1.45

Sand UC##


Sand ES# (mm) Anthracite ES (mm)



280 450

2 Du al Media

1 Dual Media

Gravel (mm)

Anthracite Depth (mm)

Sand Depth (mm)

Filter Type


Table 1 . Filter media configurations used in MIEX® direct filtration study

of the benefits of this technology that have been identified include lower disinfection by-product formation and reduced coagulant demand. For example, at the Hope Valley Treatment Plant (Adelaide, SA), the MIEX®DOC pil ot pla n t achieved a 70 percent reduction in the downstream alum dose required for turbidity control compared with alum dosing of the raw water in conventional treatment operation (Morran et al., 1996). Similar resu lts were obtained for a pilot plant lo cate d at t h e Wanneroo Groundwater Treatment Plant (Perth ,

WA) (Bourke et al., 1999). The advantage of this is a reduction in treatment chemical costs and a subsequent reduction in waste slud ge hand ling and disposal costs. The MIEX® resin removed a significant frac tion of the dissolved organic carbon (DOC) that would react with the alum coagulant, thereby reducing the organ ic coagulant demand. With the MIEX® DOC process, the p1ima1y function of the d ownstream coagu lan t addi t ion is turbid ity control. In contrast to conventional treatment, the practice of direct filtration (DF) has primarily been limited to low turbidity waters. Key advantages of DF include reduced capital and operating costs. Based on pilot-scale and full-scale studies, Janssen et al. (1986) found that DF was suitab le for waters w ith turbidity less than 10 NTU. Vigneswaran et al. (1983) proposed that DF not be used when peak turbidity exceeded 25 N TU. Hutchison (1976) suggested an upper li mit of 12-15 mg/ L alum for direct filtration. Logsdon et al. (1993) performed DF pilot-scale testing

C Pelekani, M Drikas, D B Bursill T he MIEX® DOC (Magnetic Io n Exchange Resin) process, jointly developed by SA Water, CS I RO and Orica Watercare h as been shown to be efficient for the continuous removal of NOM from wate r supplies. By applying this technology upstream of the conve nti onal treatn1ent train, nun1erous benefits can be achieved, including a reduction in the coagulant demand of up to 75 percent. With th e M I EX ® DOC p rocess, the primary funct i on of the downstream coagulation process is tu rbidity control. Extensive laboratory studies by the Australian Water Quality Centre (AWQC) have demonstrated that with MIEX® treated water from a wide variety of sources, alum doses ofless than 20 mg/L are sufficient to obtain an acceptable treated water tu rbidity. At these relatively low doses, direct filtration may become a viable option. This paper summarises successful laboratory and pilot-scale studies on turbid Rive r Murray water where product water turbidities of less then 0 .3 NTU were achieved The same MIEX®DOC pilot plant was also used to assess the performance of downstream membrane filtration for turbidity removal. This work was also presented at the AWA conven tion.






Resin dose = 6 ml / L

L aboratory jar tests were perform ed with M IEX® treated water fro m the pilot plane in 2 L p lastic jars. E ither alum (Al 2(S0 4)J. 18H 2 0) or polyDADMAC LT -35 (a cationic polymer) were dosed and rapid mixed fo r 1 mi nute, fo llowed by e ither fil tration through Whatman No. l filter paper (to simulate direct filtration), or fi ve minutes of floccu lation at 20 rpm followed by filt ration (to simulate contact filtr ation). Jar tests simulating conventional treatmen t (coagulation, flocc ulatio n, sedimentatio n and filtration) of raw water and laboratory batch MIEX® treat ed water were also conducted.

Materials and Methods

on a tu rbid water sou rce (20- 28 NTU) that experienced episodes of naturally high turbidity (up to 59 NTU). Optimisation o f the coagulant dose and to a greater extent the polymer dose was found to be critical in producing low turbidity product wa t er. T he alum dose was 7 mg/ L, with 0.26 mg/ L cation ic polymer and 0.01 mg/ L non-ionic polymer. The system was fo u nd to operate successfully when a deep sand filter was used; a shall ower dualmedia filte r was not acceptable. SA Wa ter commissioned a small M IEX® pilot pla nt to obtain detailed design param eters for constructio n of a fu ll-scale plant at Moun t P leasant in the Ad e laide H ills. In particular, assess the feasib ility o f co upling M IEX® treatment with D F and identify the necessary filter med ia configuration. Membrane filtration following M IEX® treatm ent was also exam ined, but is not discussed in thi s paper. T he plant was located at M urray Bridge, 90 kilometres east of Adelaide, and operated with raw Ri ver Mu rray water. Based on th e reduced alu m coagulant demand associated with MIEX® pre treatment, a study was initiated to assess the feasibi lity of using direct filt ration with MIEX® treated R iver M urray water.


0.2 0.1


1 .3 1.3 1.3 1 .3 1.3 1 .3

DOC (mg/ L)

54 1.3 1.4 1.5 1.4 1 .6


Results and Discussion

drop profiles were also recorded . The filter feed water turbidity was consistently 55-60 NTU, the pH was 7.8 and the DOC varied from 2.6-3.8 mg/ L du ring the filtration trials, corresponding to 40-50 percent DOC remova l with t h e M I EX ® pretreatment stage.


0 0


0 0 5 10



5 10 0

0 .5 1 .0

0 0 .2

0.5 1.0 0

0 0 .2



Flocculation time (min)

Dose (mg/ L)

0 .30 0 .28 0.13


0 .28 0.18 0.10 2. 71

3.37 0.23






Turbidity (NTU)

T~ble 3: Jar _test resu lt s with MIEX® treated wat er (MIEX® resin dose = 8 ml / L) (direct f 1ltrat1on and contact filtration simulations)


T able 2 summarises the laboratoiy jar testing results on raw and batch M IEX® treated river Murray water. These jar tests simula ted con ve ntio nal water treatment. A 74 percent reduction in DOC was obtained with a resin dose of 6 mL/ L and 30 minutes con tact time . Altho ugh o nly 20 mg/ L alu m was required to obta in the desired filtered water turbidity of 0.3 NTU with the raw water, less than 10 mg/L was necessary for the MIEX® treated water; a reduction in alum dose of at least 50 percen t. T he low alum require ment for the raw.water is possibly associated with the nature of the organ ic material at the time of sampling 0 anuaiy 2000) . To m o re closely examine direct filtratio n, jar tests were performed o n MIEX® treated water from the pilot plant, using a floccu lation time of either O or 5 minutes (T able 3). In both cases, an alum dose of approximately 1O m g/L was req uired to meet the 0.3 NTU turbidity goal, with 0.2-0.5 m g/L LT-35 providing even better turbidity removals. T hus, the use of low coagulant doses with turbi d River M urray water appeared feas ible . Figure 1 shows the eillu ent turbidity profiles for various alum doses with column 3 (deep sand fil ter) . A low


0. 15 0 .13 0. 11 0.09


Turbidity (NTU)

MIEX9Treated Water#

Three filter colum n media configurations were selected for pilot plant testing and are summarised in Table 1. Columns 1 and 3 are representative of those used at H ope Valley and Ha ppy V all ey wa t e r t r eatme n t plan ts in metropoli tan Adelaide, while colum n 2 is represen ta tive of that used at Morgan water treatment plant in the R iverland. T hese co nfigu rations we re selected because of anticipation of retro-fitting the MIEX® technology in to existing plants, w hich wo uld be cheaper tha n building new fac ilities. The co lum n diam eter was 20 cm. Two separate dosi ng units were set up for alu m and polym er. Polymers tested incl uded polyDADM AC LT-35 (Ciba Special ity C hem icals Australia, NSW), LT - 22 (an anion ic po lymer used as a coagulant a.id at several Adelaide water treatment plants) and polyalu minium chloride (H ardman Australia, NSW). Coagulant m etering was controlled with two solenoid dosing pumps (Prominen t® beta BT 4a, All Pumps, SA). Fil tratio n rates of 4- 9 m / hr were tested. The turbidity of column filtrate samples was analysed using a ratio tu rb idimeter (Model 18900-700, H ach C hemical Company, USA) . Transient pressure

3.8 2. 2

3.3 2.8




20 30 40



Turbidity (NTU) Unfiltered

5 .0

DOC (mg/ L)

Raw River Murray Water

0 10

Alum Dose (mg/ L)

Table 2. Turbidity and DOC removal resu lts bef ore and after MIEX® treatment (conventional treatment simulation)







hours was recorded. The filtration rate of 4.5-5 m / hr was 100 --· -- --------production time is defin ed as initially used to identify suitable th e tim e that water of coagulants and doses. Visible turbidity less th an 0.3 NTU pinpoint floe formation was ._13.1 mg/l.Alum was produ ced prior to observed in the column water 10 ..... 29. 9 mglL Alum turbidity breakthrough. LThead. For a wide range of alum ...,.. 38.S mo,'LAlom 35 was fou nd to yield short doses (13-85 mg/ L), th e ....,. 51.3 mg/l.Alom IH 59.7 mg/L Alom 5 rip en i ng times (30- 4 0 effluent turbidity goal of 0.3 ~ -+- 68.2 mg/l Alum minutes) and excellent filter NTU could not be attained. -+- 85.2 motL AJum 1 produ ction times at both low Increasing the dose yielded a (4.5 m / hr) and moderate (9 visibly higher concentration of m / hr) filtration rates with the floe in the water head above dual - m e d ia fi lters, wi t h the filter column and resulted column 2 displaying the best 0.1 ' - - - -- - - -~ - - - - - - - - - - - - - - in a lower initial turbidity and ,o 60 60 100 120 0 20 Fillet Run Time (mlnu1es) perfo rmance du e to the faster rip ening (defined as the d ee p e r a nthra c it e bed. time period between start-up Figure 1. Effluent turbidity profiles for the deep sand filter Although not shown , the and w hen the filter effluent (column 3) filter pressure profiles showed goal is achieved). However, excellent penetration of the with alum doses above 38.5 10 - - - - - - - - - - - - - - - - - - - - floe through the anthracite m g/L , th e time that low layer; good utilisation of the turbidity could be sustained void volume in the anthracite decreased with increasing alum bed. Testing of alum with dose. A possible explanation is LT-22 as a coagulant aid did that with.increasing alum dose yield production times in the the coagulation reaction rate order of 4-6 hours, but high increases, resulting in larger floe LT-22 doses w ere required that may be more easily sheared (typical doses in practice are within the column , yielding less than 0.1 mg/ L) and very small p articles amenabl e to --- 2 mg/l LT-35 -+- 5.1 m LLT-35 long ripen ing t im es (2-4 filter penetration through the hours) were required. relatively large voids in the 20 40 60 80 100 120 140 coarse sand filter. In the jar tests, 10 mg/ L Filter Run Time (minutes) alum yielded good flo e The filter performance at two different LT-35 polym.er Figure 2. Performance of LT-35 with the deep coarse sand formation. However, in the doses is shown in Figure 2. At filter (column 3) pilot plant, no floe formation a dose of 2 mg/ L a higher was visible in the filter column 100 c oncentration of flo e was water head, even at higher observed in the column water alum doses of 20- 30 mg/L. head than with alum, indicating A possible explanation is the improved floe formation with relatively cold water temperLT-35. However, th e 0 .3 a tu r e at th e p il o t plant NTU turbidity goal was not (13-15°C) versu s 23-25°C reac hed after 100 minutes and used in the jar tests. The rate the test was halted. It was o f alum hydrolysis to fonn the concluded that the deep coarse required polynuclear hydroxo sand filter would not be suitable complexes for effective coagufor direct filtration. The higher lation may have been too polymer dose of 5 m g/ L slow. deleteriously impacted filter To test this hypothesis, 100 150 200 50 performance. Very little floe FIiter Run Time (minutes) polyalumin ium c hloride formation was observed in the (PAC]) was also tested at an filter, indicative of particle Figure 3. Effluent turbidity profiles with dual media filter equivalent alum dose of 21 (column 1) restabilisation. mg/ L. In Table 4, PAC] Figure 3 illustrates the results c oagulation wa s more Sudden increase in filtration obtained with column 1 (dual- media effective than with alum. A ripening time rate filter). An alum dose of 65 m g/L was of 1 hour and a production time of 3 .5 required to reach th e 0.3 NTU goal, hours was achieved, but was still less T able 4 summarises the results of which co uld not be achieved w ith the effective than LT -35.Other than temperadditional filter runs performed with alum deep sand filter. H owever, this could only ature, pH is also an important parameter. and various polymers. Ripening time is be sustained for approximately 90 minutes p H controls both the kinetics and nature defined as the time required from the start prior to turbidity breakthrough. No of the coagulation mechanism. The pH of a run to achi eve the filtered water significant in crease in pressure drop of the MIEX® treated water after alum turbidity goal (0 .3 NTU). G enerally, if across the filter was observed, indicating dosing was in the range of 7.3-7 .6. this was not achieved within 2 hours of Labora tory jar tests showed faste r flo e that turbidity breakthrough was the commencing a run then the test was formation, as well as lower filtered limiting factor. halted and a production time of zero






21. 7

4.6 0.35


decreased to less than 0.2 NTU with increasing run time. Although the fi lter run was halted after 390 minu tes it was anticipated that sustai n e d wate r production welJ in excess of 10 hours could be achieved. Direct filtration with p H adj u stment confirmed chat

5 4




Professor Don Bursill is Director, Mary Drikas is Prin cipal R esearc h C hemi se (Water Treatment) and Dr. Con Pelekani is a Research Officer, all at the Australian W ater Q uality Centre. Private Bag 3, Salisbu1y SA 5108. Email: con. pelekani @sawater.sa.gov.au

The Authors

Thanks co GeoffKilrnore (SA W ater) for organisi ng the filter media and Jim M orran (A WQC) fo r pro viding useful co mm ents in the editi n g of thi s manuscript.


~~·· ·~·:==:·


Figure 4. Effluent tu rbidity and pH profiles for col umn 2 (alum coagu lati on with pH adjustment)




Direct filtration of M IEX® treated river Murray water with an average tu rbidity of 55 NTU is feasible. A dualmedia filte r consisting of 280 mm sand (0.5 m m ES) and 700 111111 anthracite (1 .1 mm ES) was identified as th e best filter medium. T he low water temperatu res (<15°C) during the pilot trials clearly slowed down the kinetics of alum floe fo rmation, resulting in the fil tered


pH can n egate th e adverse effect oflow water temperature on alum coagulati on kinetics. T he test also con firmed the feasibi lity of di rect fi ltration of M !EX® treated Ri ver Murray water of high turbidity (55-60 NTU) with lo w alum doses. It is impo rtant to emp hasise that becau se of the limited duration of th is study other media con figurations an d longer co ntact tim es before fi ltration ma y also be a so lution, rather than pH adjustment.



> 6#






Production Time (hrs)

wa ter turbidity goal of0.3 NTU not being achieved. Adjusting the pH to 6.0 w as fo und co greatly in crease the rate of floe formation and resulted in excellent filtration performance (ie. turbidity <0 .3 NTU and goo d fl oe pene t rati on into th e anthraci te bed). Unlike a l um , the c a tionic polymer polyDADMAC LT -35 was not affected by the cold water tem pera cu re and was ve ry effective at low doses (2 m g/ L ) with o u t pH adjustm ent.


No. 12., Bangkok.


E11viro11111e11tal S@itatio11 R e,,iellls,

.Bourke M, Slunj ski M, O' Leary B and Smit h P. ( 1999), Scale-up o f the MIEX® DOC process for ful l-scale water treatment plants, Proc. 18 1!, Fed. Co1w. A 11s1. Water & Wastewater A ssoc., Adelaide, CD-R.OM. Hutchison W R. (1976), High-rate direct fi ltrat ion, Amer. Wat. Works Assoc ., 68(6), 292. Janssens J G, Ceulemans J. and Dirickx J. ( 1986), Experiences wit h directfiltration: plant-scale evaluation and pilot-scale investigations, Wat. 8 S11pply, 4, 347-366. --- Turbidity Logsdon G S, N ede n D G , Alum dose increased to 22 mg/L .. .,.. H Ferguson M D and La.Bonde S D (1993), Testing direct fi ltration for Alum dose decreased back to 18 mg/L the treat ment of high- turbidity 7 water, A111er. 1,1/at. l,1lorks Assoc., StarteddosingLT·35(<0,1 mg/1.) 85( 12), 39-46. I a. Morran J Y, Bursill D B, Drikas M . and Nguyen H. (1996), A new .-0 ····· 6· · ·· 6 ·· · · •.O.·· · ·· 6 ······· 6 ... .o.·· technique for the remova l of 6 ·" natural organic matter, Proc. A11s1. Water & Wa stewater Assoc. WaterTECH Con{, Sydney, 428···················· ····~ 7 432 . Vigneswaran S, Tam D M and 0 50 100 150 200 250 300 400 350 Visvanathan C ( 1983), Water Filter Run Time (minutes) Filtration T ech nol ogies for D eve loping Count rie s.

tu rbidi ty at low alu m doses (20 n1g/L), w h en the pH was adjusted to 6.0 after alum addition. Based on these res ults, a filter run with co lumn 2 (dual- media) was performed using an alu m dose of approximately 20 mg/L. T he al um stock solution was spiked w ith concentrated hydrochloric acid su ch chat at an alum dose of 20 mg/ L the target pH of 6 .0 would be attai ned. T he product water turbidity and pH profi les are shown in Figu re 4. Th e water tem peratu re was 13°C and a filtration rate of 4.3 m/ hr was used. Greatly improved filter performance was attain ed compared with ea rli er tests. A ripen ing tim e of 80 m inutes was req u ired to m eet the 0 .3 NTU fin ished water turbidity goal. At 45 minutes the alum dose was in creased slightly because the pH was not near the target p H of 6.0 . Th e pH stabilised at 6.1, correspo nding to an alum dose of 18 mg/ L. Pol y DADMAC LT-35 was dosed at less than 0 .1 mg/L as a coagulant aid after three 100 hours to assist maintenance of low turbidity. Residual alun1inium was less than S0.06 mg/ L, well below the Iz ;;:10 Australian drin ki ng water :§ gu ideline level of0.2 mg/ L. -e i= The product water turbidity

240 120 60


1.0 0.9

21 .3

Ripening PACI Dose Time (mg/ Las alum) (min)



LT-22 Dose (mg/L)


6.0 2.2

5.1 2.3


LT-35 Dose (mg/ L)

#Run stopped because of no available night operation staff.






Alum Dose (mg/ L)

9 4 .7

4.6 4.3 4.1 4.3 4.6 8.9



13.5 13.6 13.7 13.1



3 1 1 1 1 1

Flltratlon Rate (m/ hr)

Water Temp. ("C)


Table 4. Summary of direct filtration trial results





The new ANZECC/ ARMCANZ Water Quality Guidelines (ANZECC & ARMCANZ 2001a} and the companion document, the Monitoring and Reporting Guidelines (ANZECC & ARMCANZ 20016) pave the way for a quan tum leap forward in the way in which the environmental condition of water bodies in Australia and New Zealand will be monitored and assessed. The development of new 'trigger' values and associated monitoring has been underpinned by two overriding objectives: 1. Risk equity; and 2. Simplicity. The riskbased fram ework represents a significant and welcome departure from the old 'command and control' mode whereby 'compl iance' was assessed by a prosaic comparison of sample data with a fi xed criterion. Both the industry and its regulators have been aware of deficiencies in this approach for some time, the most significant perhaps being the inherent lack of recognition of natural variation in water quality parameters over time and space even for relatively undisturbed systems . Thus the challenge for the ANZECC/ ARMCAN Z team was to devise statistical m ethods that were: • Fit for purpose • Easy to implement and understand • Robust under a w ide range of operating conditions • Easy to interpret With respect to the derivation of 'trigger values' for toxicants, the second dot point has been the most difficult to satisfy. This is largely a consequence of ANZECC/ARMCANZ adopting the now well-established statistical method of Aldenberg and Slob (1993). The compa ni o n Mo ni toring and Reporting Guidelines document (ANZECC & ARMCANZ 2001 6) , provides a review of statistical inference and related concepts such as power, level of significance, and sample size calculations . While the tools of statistica l inference provide us with a consistent and

Introduction Population

scientifically credible way of making decisions under uncertainty, they are not foolproof and errors invariably occu r. In the context of statisti cal hypothes is testing, two types of error are possible and statisticians refer to these (somewhat unimaginatively) as Type I and Type II errors. A Type I error arises when a true hypothesis is incorrectly rejected while a Type II error will have been committed if the statistical test leads us to incorrectly accepting a false hypothesis. The ability to 'get it right' in this situation is an important feature of any statistical test procedure and it is called the test's poi11er. It is intuitively obvious that statistical power increases as the sample size n increases. Curves that depict this relationship for a specific statistical test are often consulted at the planning stage of an investigation to help balance effort (cost} and Type II error. The calculations underpinning these curves are complex and best consigned to computer software. The Monitoring and Reporti ng Guidelin es docu m ent (ANZECC & ARMCANZ 200 16) gives more information about power and sample size calculations as well as references to freewa re. CS IRO 's own product, PowerPlant® and accompanying documentation can be freely downloaded from the following site: ftp://ftp.per.its.csiro .a u/ csiro-wa/biometrics

A toxicant is defined as a chemical capable of producing an adverse response in a biological system, seriously injuring structure or function or producing death. Considerable effort has been expended in refining the manner in w hich Australian and New Zealand guidelines for toxic chemicals are establish ed. The aim of the previous ANZECC (1992) guidelines was to protect all forms of aquatic life and all aspects of the aquati c life cycle and this was done using the 'assessment factor' method. The seq u ence of steps is as follows: 1. Define an appropriate end-point (eg. mortality) for the (chronic) toxicity test; 2. Establish the 'no observable effects concentration ' or NOEC for a number of test sp ecies. The sample NOEC is the highest concentration tested which is not significantly different to the control; 3. Scale the NOEC by an arbitrary facto r (the 'assessment factor') in an attempt to introduce an added level of protection so as to protect the most sensitive species. There are difficulties with each of the steps outlined above. The identification of an appropriate end-point has been problematical, although there is agreement that these ought to relate to functions of life (mortality, reproduction, growth)

Establishing trigger values for toxicants

Figure 1 . Relationship between sample statistics and population parameters



Statistics X


DR Fox



rather than behavioural or biochemical characteristics (Holdway 1996, McCarty and Munkittrick 1996). Determination of NOECs is usually statistically based (eg. regression or analysis of variance methods) and is also fraught with difficulties (Fox 1999, C hapman et al. 1996) while the arbitrariness of the magnitude of the assessment facto r itself (H art 1974, N i cholson 1984, OECD 1992, 1995, Rand et al. 1995) and lack of scientific underpinning (Warne 1998) has attracted criti cism. A number of alternati ve methods have been developed over the last decade (Stephan et al. 1985, Kooijman 1987, van Straalen and Denneman 1989, Wagner and Lokke 1991, Aldenberg and Slob 1993) in an attempt to alleviate these concerns and provide a more rational and scientifically defensible method of developing water quality criteria. T he extent to which these objectives have been met is d ifficu lt, if not impossible to assess. What is certain is that these new methods have relied on increasingly more sophisticated statistical methods, although as I have argued elsewhere (Fox 1999), this does not necessarily guarantee a superior outcome. The method by which the new ANZECC/ ARMCANZ guidelines for toxicants have been derived is an adaptation of the Aldenberg and Slob (1993) approach. T he objective is to determine the concentration in either fresh or marine waters fo r a toxicant such that some lugh percentage (typically 95%) of all species in that environment will be protected. An extra statistical dimension is introduced by attaching a level of confidence to the stated concentrati on. This is intended to acknowledge th e inherent uncertainty in the estimated concentration that arises from using only a small sam.ple of NOECs fo r the analysis. T he end resu lt is the somewhat awkward concept of, for example a 95:95 trigger value. The interpretation of this number is that it is a concentration for which it is claimed that 95% of aU species will be protected - with 95% confidence. Clearly there are an infinite number of possibilities - 95:50, 50:50 etc . although the ANZECC/ ARMCANZ guidelines have used 95:50 as the basis fo r setting triggers for 'slightly' to 'moderately' disturbed ecosystems. The justification for these choices rests on the foUowing arguments: (i) a 95% level of protection is thought to be sufficient to protect the ecosystem; and (ii) high levels of confidence applied to high protection levels are difficult to defend and tend to produce nonsensical results (eg. extremely low metal concentrations whic h were ofte n below background concentrations).

Website: www.egisconsult.com.au

Offices throughout Australia and South East Asia


f£eg i5 consulting

Emissions Testing & Air Quality Monitoring

Environmental Management Planning & System Development

Environmental Impact Statements

Solid Waste Management

Hazardous/Industrial Waste Management

Contamination Assessment & Remediation

Environmental AudiVSite Investigations

Water Resources Development, Hydrology, Irrigation & Drainage Wastewater Collection, Treatment & Reuse

Management Systems & Compliance Auditing


Water Quality Management


SA Paul Lindon (08) 8271 2322

Paul Turner (08) 9220 9300


Don Vincent, Dr Jeff Bazelmans, Mike Juleff, Dr Wayne Drew (03) 9272 6666


Stephen Trainor, George Khouri & Brad Steele (07) 3233 1611


David Russell, Dr Ian Garrard (02) 9412 9999


Ian Gregson (02) 4929 3255


Neville Ward, Ron Edwards & Dr Peter Nadebaum (03) 9272 6666

ENVIRONMENTAL GROUP Water Supply & Treatment



methodology is that this sample of NOECs is from a larger population (see Figure 1) whose distribution is characterised by a theoretical model called the log-logistic distribution. The next step in the A&S methodology is to use the imputed log- logistic distribution to estimate the concentration that is exceeded by 95% of all NOECs. This concentration is the basis of the ANZECC/ ARMCANZ trigger value. T he imposition of a confidence level on the estimated trigger value is more complex and will not be described here except to say that it has the effect of reducing the in itial 95% value. In an attempt to overcome some of the limitations of the A&S methodology identified in Fox (1999), Shao (2000) used a more flexible funiily of probability distributions of which the log-logistic is a member. T his modification has been adopted by ANZECC / ARMCANZ (Warne, 2001) and embodied in the companion software package BurrliOz® which is available fo r download from CSIRO at: http:/ /www.cmis.csiro.au/ produ cts. html


In the remainder of this section I shall attempt to provide a simple explanation of the A&S technique. First, consider the derivation of NOECs for the test species. There are a number of ways in which this can be done, although perhaps the most common is a statistical analysis of the data obtained from a series of dilution experiments. For example, organisms might be exposed to five concentrations of a chemical toxicant as well as a 'control' in w hich the toxicant is absent. The NOEC is then the rughest concentration for w hich the result at that concentration is statistically indistinguishable from the result for the control. A more rigorous, although more ti me consunling approach is to ch aracterise the dose-response relationship for each species. This is simply a plot of %mortality (say) against concentration. T he NOEC can then be estimated by extrapolating these curves back to the horizontal axis - that is, the concentration at which zero mortality is observed (Eg Mayer et al., 1994; Lee et al., 1995). By repeating either approach for a (usually small) number of species, a sample of estimated NOECs is obtained. A critical assumption of the A&S


3 -




Monitoring physical· chemical stressors next level Investigation trlggered

un ce rtain ty. Th e n ew ANZECC/ ARMC ANZ guidelines have avoided being • no action required Stressors are defined prescriptive on this issue and wamlng - investigation may be necessary as physical, chemical or inst ead ha ve provid ed an biological factors that o pportunity for the water can cause an adverse quality manager to decide on effect in an aquatic C: 0 a level of sampling that balances ecosystem as measured ~ risk w ith cost . c by a condition ~ Mindful of th e difficulties C: indicator. 8 1 enco untered with th e appliA signifi ca nt shortca tio n of 'classical' statistical corning of any omnibus tests to environmencal data Reference site PBO guidelin e o r environ(Fox 200 1), the ANZECC/ m ental criterion is the ARMCANZ team developed 0 lack of recognition of sitea more robust monitoring tool specific conditions and/or 10 11 12 2 3 4 5 6 7 8 9 known as the ' P80:P50 ' anomalies. The revised Month co mpari son pr oce dure . ANZECC/ ARM CA NZ T hough its genesis is a little Figure 2. Hypothetical control chart for water quality monitoring guidelines have a number sketchy, the team concluded (after ANZECC/ ARMCANZ 2001) of fea tures built into them that the approach t h at atte mpt to • Is intuitively appealing wetlands; estuaries and m arine. R egional acknowledge regional differences, distur• Is simple to implem.ent groupings are: south-east Australia (VIC, bance category, and ecosystem type. N SW, ACT, south-east Q LD , and TAS); • Requires no assumptions to be made Ecosystem s are classified as: upland and south-west Australia (southern WA); lowland rivers; lakes, reservoirs, and about the underlying statistical distributropical Australia (northern WA , NT, tions northern QLD); south central Australia • Is robust under a wide range of condilow rainfall area (SA) and N ew Z ealand. tions and environments Disturbance categories are: high conser• H as acceptable statisti cal performance vation/ ecological valu e (condition 1 characteristics ec osys tem s); sligh tly or mod erat ely • Is flexible THE UNIVERSITY OF NEW SOUTH WALES disturb ed (condition 2 ecosystems), and Assuming a suitable reference location highly disturbed (condition 3 ecosystems), School of Civil and each having an associated leve l of has been identified (the guidelines provide advice on this issue) it is suggested that protection. Natural resource managers are Environmental Engineering further at least 24 readings from regular sampling encouraged to adopt site-specific (eg. monthly sa mpling for two years) be monitoring w ith acceptable water quality Study while working obtained from the reference location. The criteria judged relative to local reference 80'" percentile of the sample of 24 site conditions. Th e use of reference sites ExternaVDistance Learning readings (P 80) from th e reference location is a key feature of the new gu idelines . becom es the trigger valu e for the current Before outlining the approach it wilJ be Postgraduate Coursework comparison at the test location. T he instructive to make explicit the competing sample median (i. e. the 50'" percentile, P 50 ) obj ectives of any monitoring program. Certificate, Graduate Diploma from a series of readings from. the test Monitoring is a key requ irement of & Master of Engineering Science location is used for this comparison. The environmental protection, although the sample size (i. e. number of readings) at objectives are som ewhat different for Master of Environmental the test location used to compute this regulators and operators. This subtle, Engineering Science median is determined by the analyst and although nonetheless important difference could be as sm alJ as n= 1, a single Enhance career prospects is exem.plified by the different emphasis reading . T his is where the risk trade-offs Improve professional placed on th e components of the 'triple occur. W hen n= 1, the Type I error performance bottom line'. For the regulatory agencies, (probability of a false positive) is 20% the primary concern is the environment. Specialisations available in: although this can be halved (for example) The ec onomi cs of environmental if three samples are used rather than one. protection is to some extent a secondary Project Management Thus the issue for the water quality issue. Their interest is in minimising the Environmental Engineering Typ e II error, that is that a detrimental manager is to balance the consequences of incorrectly triggering furth er action impact goes undetected. For industry , Waste Management w ith the extra cost of sampling. This is environmental performance cannot be dea matter of individual (or corporate) utility Water and Wastewater coupled from economic considerations. and cannot be mandated by any regulatmy An operator seeks to have low T ype I Management agency. Although the use of the sot11 error, that is that a low/ no im.pact For information and application forms please contact: percentil e is somewhat arbitra1y, the situation is declared detrimental. Statistical External Studies Program, School of Civil and Environmental ANZECC/ AR C MANZ team felt that a Engineering, UNSW, Sydney 2052. infere nce is inextricably linked with Ph: (02) 9385 5080 Fax: (02) 9385 6139 median at the test locatio n that was environmental monitoring since this is an Email: m.oconnell@unsw.edu.au nurn.erically equal to the 80th percentil e application of decision-making under


It is ev ident th at th ese quantitative approac hes to water quality mo nitoring an d assessm ent demand a higher level of technical sophistication. However, uptake of these n1.ethods can only be assured if ma nagers are pro vid ed with easy-to- use cools that faci litate the decision- making process. To this end, the new guideli nes ad voca te the use o f co ntro l charts w hereve r appropriate. Control charts are no t new, they were developed under the u mbrella of Statistical Pro cess Control (SP C) for the manufacturing industries back in the 1930's. Unfortunately th eir migration from the industrial setting co the environmenta l setting has been slow to occur and it is only more recently that th e utility of these tools for natural resou rce m a na ge m e nt has b ee n recognise d. Chapter Six of the new M o nitoring and Reporting Guid eli nes (ANZECC & ARMCANZ 20016) provides more detail and examples of the use of control charts . One important development has been the linking of the reference site-test site com parisons w ith the control chart. Ass u 1ning sampling is conducted on a monthly basis, the initial reference valu e is obtained as the 80th percentile from the first two years data (as described in the previous section). A rolli11g 80th percentile is u sed for all subsequent comparisons by dropping th e oldest reading and adding the reading for the cu rren t month and recomputing the 80th percentile from this set of twenty-four observations. Clearly, this revised estimate w ill incorporate a significant am ount of'hiscory' w hich will make it resili ent to rapid fluctuations cau sed by in termittent 'spikes'. It will nevertheless respond t o a constant upwards or downwards trend over time. These were seen to be desirable features of a local reference value. The control

The control chart - a management tool for water quality managers

at the reference location represented a shift worthy of further investigation. This links in w ith the notion of ecological significance bu t avoids quantification of this problem atic concept. The advantage of the percentile comparison is that it is based on a relative change and not an absolute. We have referred to the magnitude of the Ps0 - P 50 shift as a 'measurable perturbati on'. W hether this shift is ecologically significant is another matter entirely, which requires an informed judgement of the consequences of such a shift co the ecosystem . Judging ecological importance is a vexed issue, which is explored further in backgro und material provided in the ANZECC/ ARC MANZ guidelines.




Email: jamescumming@jamescumming.com.au Licence no: 1628

319 Parramatta Road AUBURN NSW 2144 Phone: (02) 9748 2309 Fax: (02) 9648 4887


Phone or fax for sound technical advice from experienced personnel.


Dr. David Fox is a Program Leader in CSIRO's Division of Land and Water. H e is a C hartered Statistician and Fellow of the R oyal Statistical Society, London as well as a member of the Ameri can Statistical Association. He was joint recipient of CS IRO 's C hairman's Gold Meda] in 1996 fo r his contributions to the Pore Philip Bay Environmental Study and recipient of the prestigious Deming Award fro m the University ofWyoming. H e has held researc h and teac hing positi o ns in statistics departments in Austra lia , United Kingdom, and the United States. Da vid is also Di rector of the new Adelaide Coastal W aters Study and Directo r of the Boa gs R ocks O utfall M onitoring Projec t. Co ntact: (08) 93336625. david.fox@per.clw. csiro.au

The Author

In particular, Dr. Graeme Batley (CSIRO Energy Techno l ogy), Dr. C h ris Humphrey (ERISS), D r. Bany H art (M onash University), Dr. John Chapman (NSW EPA), D r. Bill M aher (Canberra University), Mr. Kevin M cAlpine (WA DEP), Dr. Leon Barmuta (University of T asm ania), and D r. Mi chael W arne (NSW EPA). Thanks are also due co Bob Swinton for his editorial assistance.


The au thor is grateful for th e helpful advice and feedback from a number of colleagues in the preparatio n o f this article.


The new ANZECC/ARMCANZ guidelines represent a significant shift in chinking abo ut how water quality in Australia an d New Zealand is monitored and assessed. Wh ile the revised Water Quality Gu idelines have und ergone substantial develop ment and review it is recognised that the real test of the strategy's efficacy will only occur through implementatio n and experience. Th is is abo u t to happ en. P rac titione rs are encouraged to adopt the new methods and report their experiences (both positive and nega ti ve) back to ANZE CC/ ARMCANZ. T his will help ensure that the guidelines remain relevant, robust, and achieve their aims of balancing the competing risks identified in this article .

Concluding Remarks

chart which reflects the monthly comparisons is nothing more than a plot of the rolling 80th percen tile as the reference value and the monthly median at the test site. An example of such a plot appears in Figure 2






A Halliburton Company <l'.



Mono Pumps (Australia) Pty Ltd Melbourne 03 9580 5211 Sydney 02 9521 5611 Brisbane Perth 08 9479 0444 Adelaide 08 8447 8333 Darwin Hobart 03 6249 8704 Auckland +64 9829 0333

d~Mng and su9P'nsion of soli<'ls In municipal and industrial ,,_1ment plants.

07 3350 4582 08 8984 3099

Fox DR (2001 ) Environmental Power Analysis - A N ew Perspective, E11viro1w1etrics, Vo l 12 no 4 (in press). H art B T (1974) A compilation of Australian water quality criteria. AWRC Technical Paper 77, Australian Government Publishing Service, Canberra. H oldway D.A (1996) The role ofbiomarkers in risk assessment. H1m,a11 Ecol. Risk A ssess., 2, 236-267 . Koojim an S A L M ( 1987) A safety facto r for LC 50 value allo wing for differences in sensitivity among species. Water Research, 21, 269-276 . O ECD (Organisation for Economic Cooperation and D evelopment) (1992). R eport o f the OEC D workshop on extrapo lation of labo rato1y aquatic tox icity data to the real e n vironm e nt. OE C D En v iro nm enta l Monographs 59, O ECD , Paris. OECD (O rganisation for Econo mic C ooperatio n and D evelopm en t) ( 1995) . G uid eline docum ent for aquatic effects assessment. OECD E nvironmental M o nographs 92 , O E CD , Paris. M ayer F L, Krause G F, Buckler D R , Ellersieck MR and Lee G (1994) Predicting chronic lethality of chemicals to fishes from acute toxicity tests data: concepts and linear regression analysis. E11viro11 . Toxicol. Chem., 13, 671-678 . Lee G , Elkrsicck M R , Mayer FL and Kra use G F (1995) Predi cting chronic lethality o f che micals to fishes from acute toxicity tests

Subl!Mllllble mixers lo< blendlng,


AN Z ECC (1992) N ational w ater q uality management strategy: Australian water quality guidelines for fresh and marine waters. Australia and New Zealand Environment and C onservation Council , Canberra. ANZ ECC and ARMCANZ 2001a . Australian and N ew Zealand guidelines for fresh and m arine waters. N atio nal Water Quality Management Strategy Paper N o 4, Australian a nd N ew Zealand En vironmen t and Conse rvation Coun cil & Agriculture R esource Management Council of Australia and N ew Zealand, Canberra . ANZECC and ARMCANZ 200 16. Australian guidelines for water quality m onitoring and reportin g . N a ti ona l Water Quali ty Management Strategy Paper N o 7, Australian a nd N ew Z e aland E nvironm e nt and C o nser vati o n Co un c il & Agri cultu re R esource Managem ent C ouncil o f Australia and N ew Zealand, Canberra. Aldenberg T . and Slob W. (1992) Confidence limits for hazardous concentrations based on logistically distributed N OEC tox icity data. Ecotoxirology a11d E1wiro11111e11tal Sefety, 25 , 4863 . C hapman P M, Cardwell R. S and C hapman P F (.I 996) A warning: NOECS are inappropriate for regulatory use . E1111iro11. Toxiwl. Clie111., 15, 77-79. Fox D R. (1999) Setting water quality guidelines: a statistician's perspective. SETA C Ne111s, May 1999 , 17-18.



Membrane Bioreactors for Wastewater Treatment by Stephenson T., Brindle K., J udd S. and Jefferso11 B. ISBN 190 022 2078 IWA P11blicatio11s. ] 11 n e 2 0 0 0 . Av a il a b l e A WA bookshop@aiva.asn.att or tel (02) 9413 1288. RRP $161. 73 plus postage & handli11g. As the demand for wastewater treatm e nt , recycling and reuse technologies becomes more urgent, membrane separation bioreactors are being turned to increasingly as providing the most effective and focused solutions which minimize the need for chemicals. This book targets the maj or users of the future and the increasing demands being placed on industrial and municipal operations for on-site reuse and in relation to discharges. Chapters cover membrane fundamentals, biological processes, biorea ct ors fo r biomass separatio n by membranes, membran e aeration and extractive bioreactors. Also included are chapters on commercial membrane bioreactor systems, the range of membrane bioreactor applications - plus case studies . A good quality publication to m eet a growing market demand.


data: Multifactor probit analysis. E11viro11. Toxicol. Chem. , 14, 345- 349. M cCarty L .S a n d Munki m ick (1996) En vi ronm e ntal bi o m arkc rs i n aq uatic toxicology: Fictio n, fantasy or fi ctional? H11111a11 Ecol. Risk Assess. , 2 , 268-274. R and G M , Wells P G , and M cCarty L S (1995) Introduction to aq uatic toxicology. In Fundamentals of Aquatic T oxicology, Effects, Environmental Fate and Risk Assessm ent. 2Editn. Rand, G .M . Ed. T aylor and Francis, W ashingto n D. C. USA pp 3-67. Shao Q (2000) Estimatio n fo r hazardous concentrations based on NOEC toxicity data: an alternative approach. E1111iro11111etrics, 11(5), 583-595. Stephan CE, Mount D 1, H ansen DJ, G entile J H , C hapman G A and Brungs W A (1989) G uidelines for deriving numerical national water quality criteria for the protection of aquatic o rganisms and their uses. U SEPA PB 85-227049. Van Straalan N M and Denneman C A.J ( l 989) Ecotoxicological evaluation of soil quality criteria. Ecotoxicol. E1111iro11 . Saf 18, 241 - 251. W agner C and Lo kke H ( 1991) Estimation of eco toxico logical protec ti on levels from N OEC toxicity data. Water Research, 25, 1237- [242. Warne M StJ (I 998) Critical review of metho ds to d erive water quality guidelines for toxicant, and a proposal fo r a new framework. Supervising Scientist R.eport 135, Supervising Scientist, Canberra. Warne M St). (200 1) D e ri va ti o n of the ANZECC and AR.M CANZ water quality g ui delin es fo r tox ica n ts . A11Stralas. ). Erotoxirol. , In press.


amou nt), c hara c te ristics o f st r ea m morph o logy a n d w hether there are existi ng or potential threats to the aquatic ecosystem (EP A Vic. 19986). In m ost cases t h e biological indicator used wou ld either be macroinvertebrates, dia toms or a comb ination of the two 1• B iological sa mpling in these studies conforms to the R.ap id Bioa ssess m e nt Meth odo logy (RBA) outl ined in EPA Vic. P ublication 604 (E PA Vic. 19986). Th e protocol has been adopted by many States and p rovides metho d s fo r th e col.lection of q11alitatir1e biological data, w here the purpose is to sample the w idest va riety of communi ty taxa poss ibl e. fn many circum stances there is a need to modify th e R.BA methodology to invoke a q11m1titati 11e appro ach that relates ab undance to other measures su ch as density or biomass. Coilection of qu antitative data enables the use of paran1etric statistics to id en ti fy significant effects on river health due to treatme nt plant operation . Modifications to the R.BA meth odology are made by consul tation between the researcher, EPA Vic. and the treatm ent plant operator (water authoiity). Where possible a variety of habitats should be sampled (e.g. riffies, pools etc.) as comm un ity composition, and thus sensiti vity, may vary between habitats . Sim ul taneous phys ico-chemical sampling at ail sites is also requ ired by the protocols. Physico-chem.ical water samples and i11.-si111 measurements need to be undertaken monthly (when discharging) . Stream and effiu ent discharge volumes must be measured daily . Envi ronmental paramete rs o f importance includ e nutrients, suspended solids, Escheriscl,ia coli co ncentrations, disso lved oxygen, pH , flow velocity, electrical conductivity and



For the purpose of this paper the discussion will be restricted to focus on macroinvertebratcs only. Diatoms, however, are particularly suited to assessing n utrient effects as their species distributions react directly to nutrient concentrations (EPA Vic. 1998a).


The Environment Protection Autho1ity of V ictotia (E PA Vic.) has developed clear protocols for in-scream mon itoring of fresh water environments that receive wastewater eillu ent from treatment plants. T h ese protocols include directions on the design and impl emen tation of in-stream


T he protocols are designed to assist water autho riti es and their environm ental consultants to design a mo nitoring program that satisfies E PA Vic. requirem ents. To summarise bri efly, it requ ires a minimum of 2 to 4 sites immediately upstream (control effect) and a minimun1 of 3-6 sites do w nstream of the treatm ent plant outfall (treatment effect) in the receiving waterway (EPA Vic. 1998a). These sites represent " rep li cate" u nits, although , as will be discussed, there are statistical limitati ons on such an expe rimental design. The appropriate biological indicator shou ld be selected for the study, w hich is dependent on facto rs such as individual features of the discharge (e.g. ti ming and

The In-Stream Monitoring Protocols

monitoring programs (E PA V IC. 1998a) and standard field and laboratory methods for stream sa mpling (E PA Vic. 1995a, 19986) (Newall et al, 200 1) .

D Vertessy

MONITORING IN-STREAM IMPACTS USING MACRO-INVERTEBRATES Monitoring o f aq u atic macroinvertebrates has become a r e quir ement of the Environment Protectio n Authority of Victoria (E PA Vi c.) treatmen t p lant lice nces w here point so urce discha rges are released to streams, rivers and lakes. Macroinvertebrates such as mayflies and mudeyes are well known to anglers, are vis ible to the naked eye and are com mo nly found in rive rs and stream s. Th ey are considered to be an excellent and cost-e ffi cient indicator of river health. Stri ct EPA Vic. protocols exist rega rdi ng the m in imum effo rt required by treatment plant operators fo r in-stream monitoring and these are bri efly in troduced . These mon itoring pro tocols outline appropriate sampling meth odologies, site selection criteria and methods of data analysis such as S IGNAL and AUSRIVAS. Despite these overridin g directives it remains critical that each study be individually designed to determine differences between macroi nvertebrate com munities upstream and downstream of the selected treatment plant. Examples of such studies are given. Th e issues associated with the introduction of this new protocol fro m the project design to implementation and final fee dback stage, are outlined. T he importan ce of baseline m on itoring prior to treatment plant upgrade is also discussed.



0 .823


1.345 0.413



1 77.918




0 .822

0 .000

0 .006




0 .290

0 .183

0 .000

0 .000





0 .657

8 .677



The analysis is completed at several stages. Firstly there is an analysis of the variation between seasons, t hen between treatments and t hen between seasons and treatment. Therefore season * treatment refers to the differences in variation between season and treatment.

A high F statistic usually results in a lower probability and therefore a significant impact occurring.





0 .004

0 .001

Log 10




Where F = F Statistic, a ratio of the mean sum of squares/ mean error and P=Probability, Values in bold are significant.



* Treatment

Treatment Season


0 .199 Season

Goulburn River

* Treatment

13.762 Season




Log 10



Hughes Creek



Total P

Total N


• they have short life cycles, and therefore many life stages (e.g. larvae, pupae, adults) may be studied in a short period of time, • they are relatively immobile, and are therefore unable to escape the effects of instream pollution stresses, • they are relatively easily sampled, and • they usually occur in great diversity and numbers (Davey, 1980). Perhaps the greatest advantage in u sin g fres h water rna croinvertebrate communities in the assessment of river health is that these fa una are extremely sensitive to changes in water quality, and ecological responses to environmen tal stresses may be seen over a very short time-scale (Cairns & Dickson 1971). Macro invertebrate commu nities also provide a continuous record of environmen tal degradation, as opposed to snapshot physico-chemical sampling of surface waters. Macroinvertebrates are also biologically important in that they are a major component of freshwa ter ecosystems and are important food resources for fish, amp hibians and waterfowl (Cairns & Dickso n 1971) . M acro inverteb rate comm u ni t ies also have important funct ional roles with in streams with respect to nutrient cycling (Vannote et al. 1980) and organic matter processing (Wallace et al. 1977 , Ward 1989). T he following discussion will ou tline a number of the methods used by qualified aquatic ecologists to identify the effects of wastewater from treatment plants

Table 1 . ANOVA Results for Hypothetical Data from Hughes Creek and the Goulburn River

Freshwa ter m acroinvertebrates are a group of anim als visible to the naked eye that reside in aquatic environ ments and i n clude i n sects, worms, l eeches, crustaceans and snails. Macroinvertebrates are a cost-efficient indicator of river health, and an ideal biological monitoring tool in that: • they have differential susceptibility to different pollutants,

Freshwater Macroinvertebrates

biological oxygen demand (BOD). The State Environment Protection Policy (SEPP), Waters if Victoria (Government ofVictoria 1988) sets water quality objectives fo r all general surface waters in the State, an d provides a Statewide policy framewo rk fo r several catchment-specific policies. Where numerical nutrient objectives are not provided in a SEPP, guideline concentrations from Preliminary Nutrient Guidelines for Victorian Inland Streams (EPA Vic. 19956) will be used. To conform to RBA methods, a minimum of two seasons of biological sampling is reconunended (usually autumn and spring) to accou nt fo r temporal variation in community composition, life history and stream discharge (EPA Vic. 19986). T he p ro tocols req uire in-stream monitoring to commence two years prior to the implementation of a new treatment plant. W here upgrades are made to existing treatment plants, in-stream monitoring should be undertaken at least one year prior to the upgrade and continue for at least one year thereafter (EPA Vic. 19986).


AUSRIVAS Model AUSRIVAS is a conununity modelling tool that predicts which macroinvertebrate fa milies should be present in specific stream habitats under reference conditions.

The SIGNAL score is a biotic index, which uses aU conun unity data to produce a score between O and lO that reflects the degree of water pollution determined by the presence of macroinvertebrate families of a known tolerance or intolerance to pollutants. This is a read ily understood tool for water managers. SIGNAL scores b el ow 4 indicate p robabl e severe pollution, scores of 4-5 indicate probable moderate poll ution, scores of 5-6 indicate doubtfu l quality, possible mild pollu tion and scores grea ter than 6 suggest clean water status (Chessman 1995). Specific SIGNAL scores have been developed for the Yarra R iver catchment (EPA VIC. 1999) and, in draft form, for the Western Port catchment (EPA Vic. 2000) .


The pu rpose of discharge impact studies is to compare macroinvertebrate community structure and water quality between u ps t ream (control) and downstream (impacted) sites to evaluate a treatmwt effect. All analyses take habitat variations between sites into consideration and also seasonal variation , including flows. T herefore selection of sites must be made to redu ce confounding variables such as incoming storm water drains or stream tributaries. Similarly, site selection should be made to minimise differe nces in stream morphology such as stream substrate composition and riparian stream cover. Interpretation of the data is assisted by numerical analyses, including param etric univariate analyses such as Analysis of Variance and BACI designs, explo ratory multivariate statistics such as classificatio n and ordination, indices including SIGNAL (Stream Invertebrate Grade Nu mber Average Level, C hessman 1995) and by pr edic t ive mode l s, for example, AUS R l VAS (AUStral i an R I Ver Assessment System, Simpson et al. 1997). T hese are discussed briefly with som.e examples below. Interpretatio n must assess results against ecological (e.g. waterways in the Yarra River Catchment, EPA Vic. 1999) and water quality objectives (SEPP guidelines, Government of Victoria 1988).

Design and Analysis

on river health, including the use of biological indicators and univariate/multivariate statistical tools. Examples of results will be presented where appropriate.

The ANOV A results include statistical tes ts co nduc ted o n total nitroge n , to tal p h osph o rou s an d dissol ved o xyge n co n centrations. On H ughes Creek th ere wer e significa nt effec ts (at a con fi de nce li mi t of 95%) for all param eters w ith

Univariate Statistics A nalysis of Varian ce (AN OVA) models may be used to evaluate treatme nt effects o f wastewater d ischarge. T ypi cally all e nvironm enta l parame ters and sumn1.ary b io logica l param e te rs su c h as tota l numbe rs of indi vidu als, SIGNAL score, tota l nu mbe rs of fam il ies and numbers of key mac roinve rte brate fam ilies are used as ci1e inde pe nde nt va riabl e in th ese analyses. T he statistical test is condu c ted to test wh e ther any o f these parameters is sig nificantly different between upstream and downstream sites, and given an ad eq uate experimental des ign, w he th er these sign ifi can t diffe rences may be attrib uted to the waste water discharge . T he example below (T able 1) includes so me hypo theti cal e nvironmenta l data from the G o ulburn R iver catchment. The hypothetical data includes nu tri ent and dissolved ox')'gen concen trations for a tribu ta ry, whic h receives wastew ater (Hu ghes C reek), and the mai n wate rway thi s tri b uta ry di sc h a rges i n to (th e G o ulb u rn Ri ve r) . Mo n ito ring o f a m ainstrea m do w nstream of a tri butary rece iving a wastewa te r discharge may be recommended in the experi mental design if t he tribu tary is small or in close prox imity to a maj or wa terway. In this example multiple sites were placed on Hug hes C ree k, upstream and downstream of the discharge and also on the G o ulburn Rive r upstream and downstrea m of its co nflu e n ce w i t h H u g h es C r eek. Co n sequently, effects of the trea tment plant on th e Goulburn Ri ver can also be teste d .

It does so b y comparing test data (in this case from upstream and downstream sites of the trea tme nt plant) with a group of reference sites w hich are as free as possible of enviro nmental impacts bu t with similar morphological characteristics. A ratio is calculated which expresses the observed number of fa milies fou nd at a test site against the expec ted number o f fam.ilies found under reference conditions, and th is is known as th e O / E Index . O / E sc ores m ay be co mpared to bands represen ting different levels o f biologi cal condition, as recom m ended under the N RHP (Simpson et al. 1997) . G ive n th e confound i ng e n v iron m e nta l e ffe c ts a ss o c iated w i t h u rba n s t r e am s, AUSR IV A S is best su ited to assessin g impacts o n regional waterways .

i 0


0 up







atter location and time

before down

after down




atter location and ti me

bef0<e down


~ IC







aner location and time

bef0<e down

after down

Change between upstream and downstream dlllars between before and alter: Impact occurring


Change between upstream and downstream: no Impact


aner down


location and time

bef0<e down








No evidence of any affect: no Impact


Altho ugh the design presented abo ve will prod uce resu lts th at can be used by the ecologist, th ere is a ve ry important experim ental design violation inherent in this study and in alJ studies, w hich use this des ign. T he violation is tha t the design is " pseudo- re pli cated " (Hu rlbu rt 1984) because we only have one real experimental unit (the creek) . C onsequently we have no real way of knowing w hether this result is m erely the effect of simpl e upstream/ downstream variations along the wate rway rathe r than being du e to the disc harge . Id eally, the design should in corp orate a numbe r of morphologica lly similar creeks in the imm ediate geographic area, each with treatm ent plants. Such a design

BAGI Designs

respect to treatment (contro l vs. impacted sites) and season (spring vs. summer), with no significant inte ractio n effect be tween seaso n and treatm ent. T o su m marise this result, we cou ld say total N , total P and dissolved 0>-')'gen we re: • signifi cantl y affected by the treatment plant, • sig nifi ca n tly affec ted b y se ason al variation, and • effects re mained consistent between seasons.



Multivariate Analysis Exploratory data techniqu es such as clusteri ng and ordination may be used to fi nd spatial pattern s between biological community data and e n vironmental data. Whilst th ese tech niques are not as statis-

is not usually possible, so this violatio n is o ft e n ig nored. T here are, ho we ve r , ANOV A me thods that may be invoked to overcom e pseudo- repli catio n (lack of spati al ind epe nde nce) fo r studi es o f treatme nt p lants prio r to their commi ssio ning. BACI (Before-After- ControlIrnpac t) designs (Underwood 1993) use mon ito ring data from upstrea m and dow nstream sites collec ted pri o r to tr e atm e n t plant c o n st ru c t io n and discharge, and then compare these d ifferences with upstream and downstream sites m o nit o r e d aft e r d isc harg es hav e co mme nced. In this example th e interaction term (time * treatme n t) is th e important test statistic. If this term is signifi cant it may be inferred that spatial effects between upstream and dow nstream only occurred o nce discharge had bee n initiated, and thus treatm ent effec ts are a ttribute d to th e trea tm e nt pla nt. H yp oth etical data is prese nted in Figure 1, which ill ustrates th e possible o utcom es o f a BAC I design on river health befo re and aft er discharge has comme nced.

Figure 1. A set of hypot hetical results from a BACI design. Note: Only in Graph D is there evidence that an impact is occ urring.








Change between before and alter: no Impact





• Biological mon itoring is required for all streams receiving waste wate r d ischarges from treatment plants in V ictoria, as stipulated by the EP A Vic. (1995a) • The E PA Vic . (199 5a) has identified a strict protocol for the mi nim um sampling effort, m ethodology and analysis requ ired w here treatment plants are discharging.


Proced u r e base d on Habitat Specific Sampling, Fa111ily Level Identification and a Biotic Index. A11srralia11Jo1m1nl of Ecology 20, 122-129. Clarke, K.R. and Warwick, R.M. (1994) Changes in Mari ne Communities: An A p proach to Statistica l Analysis an d Interpretation. Ply111outh Marine Laborato1y, United Kingdom. Davey, G . W. (I 980) The use o f biological

K.W. , Sedcll , J .R. and C u shing, C. E. ( 1980) The R iver Continuum Concept. Ca11adia11 jolfr11al qf Fisheries mid Aquatic Scie11ces 33 , 130-137. WaUace, J.B., W ebster, J.R . and Woodall, W.R. ( 1977) The role of filter feeders in fl owing w aters. Arc/1i11 flfr I-Jydrobiologie 79, 506-532. Ward,J.V. ( 1989) T he fo ur-dimensional nature of lotic ecosystems. jo11mal of rhe Norri, A111erirn11 Bwrhological Society 8 , 2-8.

parameters for the assessment of tically reliable and capable of water quality - a literature producing d iscrete results review. EPA VIC. Publication as univariate techni ques, 99. Victorian Governm e nt t hey are capable of Printer, Melbourne . summarising large sets of EPA Vic. (1 995a) A Guide to the spring data and are u sefu l for Sam pling and Analysis of Water autumn a nd W astewate r. EPA catchmen t level interpreP ubli cation 44 ·1. Victorian tati o n in particular. Results Government Print er, from an o rdinati on are upstream TP Melbourne. presented in Fig ure 2. The 101 E PA Vic. (1995b). Preliminary ordination technique calcu102 Nuttient Guidelines for Victotian 103 lates what is kn own as a Streams. EPA Publication Inland 104 d issimila rity matri x that 4 78 . Victorian Governm ent downstream TP sun1marises multivariate data Printer, M elbourne. 105 EPA Vic. (1998a) Point Source as distances between samples. 106 Discharges to Streams: Protocols These distances may then be AXIS1 107 for In-Stream Monitoring and plo tted as axis scores on an 108 Assessment. EPA Publication ordination graph and this is 596, Febrnary 1998 . Victorian Figure 2. An example of biologica l ordination results using the e quivalent of expressing Government P rinter, hypothetical comm unity dat a. these distances as wo uld be M elbourne. done on a spatial m ap. EPA Vic. (1998b) R ap id In many cases these m ay be varied to Spatial and temporal patterns may then be Bioasscssmenr of Victorian Streams. The account for local factors such as proximity Approach and Methods of the Environment discern ed from t he ordi nation. Figure 2 Protection Authority. EPA Publication 604, to major wate rways or if treatmen t provides a h ypothetical output from an Febrn ary 1998 . Vi ctorian Government plants are situated in estuarin e or intero rdination analysis . In the example given Printer, Melbourne. mittent stream syste m s. I n all cases it is the axes reflect d iffe rent patterns, wit h EPA Vi c . (1999) Variation o f the State critical that studies are specifically designed Axis 1 d iscriminating seasonal d ifferences Enviro11111ent Protection Policy (Waters of to evaluate treated eilluent effects. betwee n sa1nples while Axis 2 d iscrim iVictoria) Insertion Schedule F7 of the Yarra nates spatial differences between upstream • T he most conu11on bio logical ind icators Catchment. Victorian Govern ment Printer, and downstream sites. Although not Melbo urne. used in such studies are ma croinve rteEPA Vic . (2000) Protecting t he Waters of presented in the example in Figure 2, brates and d iatoms, w hich are relatively W estern Port and Ca tchment. A summary environmental data may also be fit ted to easy and inexpensive to sample. of the Western Port and its catchment. the ordination space, and through statis• A w ide range of ecological and statisSc hedul e (F8) to State Environ111ent tical simulations it is p oss ib le to identify t ical tools are available to identify where Protection Policy (Waters o f Victoria) . EPA significant environ menta l parameters, significan t e ffects of treat m ent plant Pub li cation 706, M ay 2000. Victorian which are responsible for the spatial Governme nt Printer, Melbourne. operations occu r. Where these occur patterns in the biological data . These may Government of Victoria ( 1988) . State m anagement decisions must be adopted Enviro nment Protection Policy (Waters of be fi tted as vectors to the ordination to minimise environmental st ress, w hich Victoria). Victorian Go vern111ent Printer, graph . This m u lt ivariate appro ac h may include d isposal to drainage basins o r Melbourne. becomes a more interpretative m ethod plantations, or alterna te tim ing of e illuent Hurlburt, S. H . ( 1984) Pseudoreplication and the than univariate analysis and is depen dent disposal to take into account biological design of ecological fie ld experimen ts. on the ecologist's ability to nuke sense commu nity factors and flow discha rge Ecological Mo11ogmpl,s 54, 187-211. of environmental and biological pa tterns patterns. N ewall P, T iller D , Rissm an], Hansen P (200 1) with respect to the o perations of the Environmental regulatory ch ange - a Author treatment plant. Vic torian example. AW A ·1 9th Convention Proceedings, Paper 207. Other tech niques such as Analysis of Dan Vertessy is Principal Biologist at Simpson, J., Norris, R ., Barmuta, L. and Sim ilarity (ANOS IM) are multivariate WSL Consu ltants P ty Ltd, Rich mond , Bla ckman, P. ( 1997). Australian River equivalents of the univariate ANOVA Vic . Email wsl@ozema il.com .au Assessment System. National l~iver H ealth models that may identify treatm e nt Program P redictive Model M anual (1st References effects. Key macroinvertebrate indicator dra f t) . F ro m the web site Cairns,J.Jr. and Dickson , K. L. (1971) A simple familie s may also be identified by http://ausrivas.canberra. edu. au/ausrivas/ method fo r the biological assessment of t he S IMPER analysis that dete rmin es w hich Underwood, A.J. (1993) T h e mechanics of e ffects of waste discharges on aquatic taxa are responsible for observed differspatially replicated sampling programmes to bottom-dwelling organisms. Jo1m1nl of rlie ences between upstream and downstream detect environmental impacts in a variable Wafer Pol/111io11 Co11rrol Fedemrio11 43 , 755-772. world. A11srmlia11 Jo,m,a/ of Ecology 18, 99- 11 6. sites (Clarke & Warw ick 1994). Chessman , B .C. (1995) R apid Assessment of Vannote, R.L. , Minshall, G . W. , C ummins, Rivers Using Macro inve rte b rates : A