Water Journal September 2001 by australianwater

Volume 28 No 6 September 2001 Journal of the Australian Water Association

Editorial Board F R Bishop , Chairman B N Anderson , R. Co nsid in e, W J Dulfe r, G Finke, G Finlayson, G A H o lder, B Labza, M Muntisov, P Nadebaum,J D Parker, ] Rissman , F Roddi ck, G R ya n, S Gray

[II Water

is a refe re ed journal. Th is symbol indi cates chat a pap er has bee n refereed.

CONTENTS 4

FROM THE FEDERAL PRESIDENT: AWA Activities

FROM THE EXECUTIVE DIRECTOR: The Great Divide

Submissions

INTERNATIONAL AFFILIATES:

Submissions should be made co E A (Bob) Swinton, T echnical Editor (see below for details).

WEF Report

MY POINT OF VIEW: Bringing the Community Along Prof.

CONFERENCE REPORT: ENVIRO 2002 & IWA Congress shaping up nicely

CROSSCURRENT: Water News Around the Nation

Managing Editor P eter Stirling PO Box 84, Hampton Vic 3188 Tel (03) 9530 8900 Fax (03) 9530 8911

Technical Editor .EA (B ob) Swinton 4 Pleasanc View Cres, Wheelers Hill Vic 3150 Tel/Fax (03) 9560 4752

FEATURES:

Email: bswincon@bigpond. net.au

PROJECT IMPLEMENTATION

Crosscurrent Editor

W (Bill) R ees PO Box 388, Arcarmon, SW 1570 T el +61 294131288 Fax: (02) 9413 1047 Email: brees@awa.asn.au

ancy Millis

CONSTRUCTING THE NORTHSIDE STORAGE TUNNEL R Cuttler & A Henderson

A multi-million dollar project achieved by alliance contracting

WATER

AWA Head Office PO Box 388, Artarmon, NSW 1570 T el +61 2 9413 1288 Fax : (02) 9413 1047 Email: info@awa .asn. au

Water Advertising & Production H all mark Editions PO Box 84, H ampton, Vic 3188 Level 1, 99 Bay Street, Brighto n, Vic 3186 T el (03) 9530 8900 Fax (03) 9530 891 1 Email : hallmark@ hall edit. com.au

E A Swin ton & C Walsh

50 million people at risk: Can Australia help? 34

ARSENIC REMOVAL USING UV ASSISTED OXIDATION

A technology applicable to groundwater and mine waters

POLYALUMINIUM COAGULANTS IN WATER TREATMENT

P Ge bbie

Graphic design: Mitzi Mann

Australian Water Association

M Zaw, M Emert & P Prasad

Comparative costings: some Victorian applications

Water (ISSN 0310 â&#x20AC;¢ 0367) is published eight tim es a yea r in ch e months of J anu ary, March, April , J une, Ju ly, Septembe r, O ctober and D ece mbe r.

ARSENIC: THE AUSAID PROGRAM IN ASIA

EFFECT OF MONITORING TECHNOLOGY ON REGULATORY COMPLIANCE B Labza, J Rissman & P Donlon

Better analytical techniques: The Victorian regulatory strategy

MICROBIAL ACTION: A KEY TO CUPROSOLVENCY IN PLUMBING?

M M Critch ley, N Crornar, N McClure, H Fallowfield

ABN 78 096 035 773

Strong evidence that it is not just water composition

Federal President

WASTEWATER

Barry Norman

Executive Director Chris D avis

AUSTRALIAN WATER ASSOCIATION

Ausrralian Water Association (AWA) assum es no responsibili ty for opinions or statements of fa ces expressed by conrribucors or advertisers. Editorials do not necessarily represenc official AW A policy. Advertisements are included as an infom1ation service co readers and are reviewed befo re publication co ensure relevance to the wa ter environment and objec-

tives of AWA. All material in Water is copyright and should not be reproduced wholly or in part without the written permission of the General Eclicor.

P Thomas & R Croom e

Capital, monitoring costs and legal risks can make it difficult

r 10n

and access news, calendars, bookshop and over 100 pages of infonnatlon at

[El NON-POTABLE RE-USE OF TREATED EFFLUENT N Cromar and D Cunliffe

Pragmatic questions and answers at a workshop on management of health risks

ENVIRONMENT

~ SEDIMENT CONTAMINANTS IN URBAN STREAMS N O'Connor & S Moore

Focus your monitoring of stream health on heavy metals and stream biota

Subscriptions Water is sent co all AW A members eight times a year. It is also available via subscri ption.

Im THE USE OF RECLAIMED WATER FOR IRRIGATION: SOME ISSUES

63 64

MEMBERSHIP MEETINGS

OUR COVER: Th e cover image shows Sydney Water's Northside Storage Tunnel under construction with the timnel boring machine tunnelling through sandstone towards lvliddle Harbour. Th.e tunnel, which is up to 6. 5 metres in diam eter, has the capacity to prevent hundreds of millions of litres of dilute sewage from entering Sydney Harbour. See article comm encing page 23. Photo: courtesy of Trans.field P ty Ltd.

PROJECT

IMPLEMENTATION

CONSTRUCTING THE NORTHSIDE STORAGE TUNNEL R Cuttler and A Henderson Abstract

T he Northside Storage T unnel has bee n constru cted along the northern side of Sydney Harbour to intercept storm overflows from the main northern sewer. T he proj ect, with a capital cost of approximately $460M, w as conceived, planned and delive red in less than fo ur years. This paper briefly describes the bac kgro und of the proj ec t, and th e alliance contract system w hich enabled fast tracking of design and constru ction and the attainment of the multiple obj ectives of cost, time, environment, community liaison and safety .

Key <===> Storagetunnel route Spoil tunnel route

•

Loca tions with

RYOE

Introduction

The Northside Storage T unnel is one component of W aterPlan 21 , Sydney W ater's long term strategy fo r ecologically sustainable wastewater ma nagem ent in its ar ea of op e ra tion s. Wat e rPl an 2 1 comprises a series of proj ects to be under:taken over the n ext 20 years to upgrade Sy dn ey W a te r's sewe rage sys te m s. Specifically W aterPlan 21 is a three billion dollar progranune of w orks in the Sydney area compnsm g: • protection of rive r eco-system s; • upgrading of the major coastal sewage treatment plants; • improved treatment and development of reuse opportunities; • and finally, a major initiative to prevent 80 - 90% of wet wea ther sewage overflow events. The Northside Storage T unnel is part of the we t w eather ove rflo w redu ctio n programme.

KIIGmetres

Tunnel route and locations

• T hirdly, environmental practices and procedures on the proj ect we re to be w orld class . • Similarly, social practices and procedures were to be rated as w orld class with genuine sensitivity sho wn to community m embers and groups. • Finally, the proJ ect was to be characterised by w o rld class safety, industrial relations and quality standards. To ensure total alignment of all alliance participants with these obj ectives, commercial dri ve rs we re incorporated in

th e Proj ect Alliance Agreem ent in the form of pre-agreed risk/ reward formu lae. T hese formu lae were selected to apply to the Alliance's p erformance in the obj ecti ves of co st , tim e, e n vironm e nt , community and safety . The indu strial relations obj ective was closely linked w ith performance in cost and time, and therefor e w as not required to be the subj ect of a separate risk/ reward form ula. Similarly adoption of a proj ect sp ecifi c quality system incorporating all fi ve proj ect obj ectives w as considered to

Project Objectives

Apart from the prime obj ective of redu cing pollution of Sydney's wa terw ays from we t wea ther sewage ove rflows, this proj ect had the fo llow ing delive ry obj ectives : • Firstly, the proj ect was to be completed at a cost regarded as high value for money by industry standards. • Secondly, the tunnel w as to be available for use by the time of the 2000 O lympics.

Thi s paper is an edited version of th e paper publi shed at the Internation al Tunn elli ng C on fe rence, M ilan , Jun e 2001.

North Head Underground Works WATER SEPTEMBER 2001

IMPLEMENTATION

PROJECT

Lane Cove River and extends som e 16 km easterly to North H ead Sewage Treatment P l an t in Manly. Approximately half way along the m ain tunn el, a branch tunnel extends 3.5 km northerl y to Scotts Creek. As part of the project, a 1.5 km spoil conveyor tunnel was constru cted between North H ead and Littl e Manly Point o n Sydney H arbo ur. The major controls for the vertical ali gnment we re the deeply in cised channels at the Lane Cove Ri ver, Middle H arbour and M anly. The main aim in selecting the vertical alignm.ent was to ensure that the tunnel was located in sound rock (sandstone) w herever possible to minimise the constru ction time.

Pictured is the 6.56 metre tunnel boring machine which excavated the North Head to Middle Harbour section of the Northside Storage Tunnel.

be the most approp riate means of integrating quality managem ent across the project. The performance incentives in the measured obj ectives implied a requirement for zero re-work, and therefore also provide appropriate drivers for quality. The Overflow Problem

Overflow events occur in wet weather throughout the metropolitan area but it is those on the north side of the harbo ur that have the greatest impact on harbo ur wa ter quality. T he four major contributors are those overflows located at the Lane Cove River, Tunks Park and Quakers Hat Bay on the main northern subu rbs ocean outfall sewer (NSOOS) , and at Scotts Creek on one of the major trunk sewers. Only two of the top twe nty-seve n overflows that impact directly on Sydney H arbour are located on the south side. T he locations of the fo ur maj or overflows o n Sydney's north sid e suggested that effective capture and storage could be provided by a tunnel system. T he excess storm flows could then be transported by the tunnel to the major coastal treatment plant at North H ead, near the entrance to Sydney H arbour. The major benefits of such a system were that: • it wo uld provide a backbone for furth er overflow reduc tion in the future; • it was the most cost effective option; and • having minimal surface facilities, it was of least disruption to local residents.

WATER SEPTEMB ER 2001

The largest overflow, at Lane Cove, discharges into the Lane Cove River between th e suburbs of Hunters Hill on the west and Lane Cove on the east. The magnitude of this overflow is such that in major storms the rate of discharge into the Lane Cove River is more than twice the normal dry weather sewage flow through North H ead Sewage Treatment Plant. T h e second largest overflow, at Quakers Hat in Mosman, has been the cause of many complaints by residents over the yea rs because of pollution of the bay and noticeable odo urs during its operation. Design and Construction

T h e route of the main tunnel commences on the weste rn side of the

At N orth H ead separate decli nes were constru cted to the existing underground treatment plant , to a n ew tunnel pumping station located approximately 100 m below sea level, and to the main storage tunnel. The declines and caverns we re excavated by three large-lO0Kw roadheader tunnelling mac hines . T he declines were on a slope of 1 :7. At the storage tunnel level a 6.6 m diame ter tunnel boring machine commenced th e main tunnel drive towards Middle H arbo ur. A similar an-angement was adopted for the Tunks Park site w here a furt her two 1 00Kw roadheaders comp l ete d excavation of the access decline and several undergro und caverns . Three tunnel boring machines operated from this site. One 3.8 m diameter machine excavated th e tunnel to th e Lane Cove Ri ve r. A seco nd m ac hine, 6.3 m diameter, excava ted the drive towards Middle Harbour, w hile a third, 6.0 m

Lane Cove 25

Quakers Hat Bay

«i Q) >,

0 C Q)

E :::,

Tunks Park VJ

Scotts Creek

Q) E .!'2 15

Largest south side overflows to harbour

.0 :::,

VJ C

;;:: .Q

't: Q)

> 0

E 5

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Overflow ranking by volum e to harbour Ranking of Overflows

PROJECT

diam eter, constructed the branch tunnel to Scotts C reek . T he storage tunnel has the pumping capacity of up to 350 million litres per day. T hese pumps we re designed and manufactu red by Weir Engineering, to the specification of the Alliance Team. The design of the storage tunnel is based on the results of sophisticated system modelling using the DHI Water and E nvironment MOUSE sewer modelling software. Two high performance, mixed flo w pumps, each w ith a minimum capacity of 175 million litres per day, are used to evacuate the stored sewage overflow from the tunnel. A small capacity high head pump deals w ith seepage water inflow. Overflows from the NSOOS are transferred to the tunnel via drop shafts, w ith access sh afts for maintena nce purposes at the Lane Cove River and Scotts C reek ends of the tunnel system. Access is also provided at the North H ead STP. Air is removed from the tunnel w hile it fills at a rate equal to the liquid filling rate plus the rate of air induced through the drop shafts . There is constant ve nti-

IMPLEMENTATION

Pictured is the 6.3 metre Wirth tunnel boring machine looking forward to the boring head.

lation of the empty tunnel to prevent any build up of gases and limit corrosion. T he main exhaust vent is at the North H ead STP. Additional vents are located at Lane Cove, Scotts Creek and Quakers H at. Odour control meas ures are installed at each overflow drop structure and at the North Head STP.

Construction Innovation

R emoval of spoil from the tunnelling operations was a m ajor issue on this project. Because of the location of the tunnelling works in highly developed residential areas, an alternative to spoil removal by road transport had to be found. The fin al arra?gement entailed

United KG, as electrical sub-alliance partner on the Northside Storage Tunnel, is proud of its contribution to this outstanding project. As part of United Group Limited, United KG combines more than 100 years of diversified engineering experience. United KG is one of Australia's leading integrated engineering, construction and maintenance businesses. Utilising the skills and experience of over 1,700 employees throughout Australia, United KG is committed to developing partnerships with our customers in the water industry. We invite you to take advantage of our capabilities and be part of our future success. United KG Pty Limited

~ lNITEDl<C Building partnerships with industry.

77 Parra matta Road Silverwater NSW 2141 Telephone (02) 9364 1444 Facsimile (02) 9364 1599 www. unitedgroup.com .au A United Group Limited Company

WATER SEPTEMBER 2001

PROJECT

IMPLEMENTATION

spoil removal from the tunnel boring ma chines by continuous conveyo rs, and removal from the underground wo rks by a combination of inclined, vertical and horizontal co nveyo rs to barge loading points o n the harbourside. All up there was over 25km of conveyors employed on the project. Barges then transported all of the 1,800,000 tonn e of excavate d sandstone (spoil) son1.e 18 km across Sydney H arbour co a commercial railhead in White Bay. From there spoil was transported predominantly by rail to points on the weste rn o utskirts of Sydney w here it was used for industrial development earthworks. One barge loading point was lo ca ted at Little Manly Po int Reserve in Spring Cove, North Harbour. Spring Cove is the habitat for Sydney H arbour's only colony of Little Penguin. Problems to be overcom e at this site we re the prevention of disturbance to the Interior view of the Northside Storage Tunnel. penguin colony and the avoidan ce of damage to sea grasses in the Point, and the spoil conveyor from North dedicated marine park area . H ead was installed in the tunnel. The Little Manly Point Reserve was A si milar spoil handling facility was also the location of a forme r gas wo rks, installed at the second main constru ction w hich had been landscap ed and restored as a community park. A certain amount site at T unks Park, Cammeray, within an of contaminated soil resulting from the environmentally sensitive residential area original gas wo rks remained on the site and adj ace nt to parkland and a public but was successfully stabilised in the recreational reserve. Stringent noise and restoration of the reserve. Constru ction space constraints had to be overcome to of the spoil loading facility required allow the proj ect to proceed. removal of som e of the contaminated Again assessm ent of options against all material. Careful environmental planning objectives led to an innova tive solution was necessary to minimise disturbance of being reached. Two underground caverns the material and avoid the risk of we re constru cted for storage of spoil contaminating the highly sensitive aquatic during night-time and allow tunnelling environment surrounding the reserve . An to continue 24 hours per day. Offices and environmental specialist was put in am enities buildings we re constructed on ·charge of the works at Little M anly Point, piles over th e b ay and all surface directing the constru ction, environwo rkshops and barge loading facilities m ental and community liaison ac tivities. were totally enclosed w ithin buildings After completion of the spoil con veyor lined w ith aco ustic suppression cladding designed to minimise noise transmission tunnel, the w harf and barge loading faciland disturbance to the local community. ities we re constructed at Little Manly

Alliance Project Delivery Method

There we re three main reasons fo r selection of the alliance m ethod for delivery of this project. The first was the required proj ect schedule. In J anu ary 1997, the NSW Government requ es ted Sydney Water to assess w hether a m ajor component ofWacerPlan 21 co uld be brou ght_for':'7a rd and completed prior to the Sy ney 2000 O lympics. A team was set up to assess the feasibility of constru cting the sto rage tunnel in the required time period and it became clear that under conventional contrac ting m ethods, proj ec t delive ry co uld not be ac hi eve d w ithin th e d es ir ed timefram e. Fast cracking of design and construction wo uld be necessary and the most appropriate method to ac hieve chis was to fo rm a stra tegic alliance with priva te industry gro ups fo r delive ry of the proj ect. Und er an alliance, both the designer and constructor co uld be brou ght on board at the earliest possibl e time. Alliancing also provided a m eans of introducing flexibility in the design and constru ction methodologies thus allo wing optimisation to be achieved thro ugh o ut the delivery process. Finally the alliance environment provided an opportuni ty to meet m ultiple o bj ectives and to achieve outstanding performance in the project o utcom es . While alliancing had been successfully used in the oil and gas industry in Australia in recent yea rs, the Northside Storage Tunnel was the first public sector capital works proj ect to be delivered under an alliance. Alliance Agreement

The cornerstone of any alliance is th e Alliance Agreement, w hich provides: • Principles to be followed in such a cooperative working relationship ; • a regime for sharing gain or pain; • a framework fo r establishing a target cost;

NORTHSIDE STORAGE TUNNEL Sub-Alliance Partners - Survey Management Whatever your project, contact us first. HARD

+ FORESTER 26

Hard and Forester, 52 Frederick Street, PO Box 175 Rockdale NSW 2216 Australia Phone: 612 9599 4077 Fax: 612 9599 2146 email: haf@ozemail.com .au

WATER SEPTEMBER 2001

Tunnel Surveying

Guidance · GIS

Barges were used to remove 1.8 million tonnes of spoil.

• an agreed process fo r reimbursem ent of cost; • ·an agreed level of profit for the Alliance participants; • commercial incentives to encourage the achievement of outstanding results; and • clea rl y defin ed consequ ences of poor performance . The Northside Storage Tunnel Alliance

T he Northside Storage Tunnel Alliance comprised : • Own er and C lient - Sydney Water Corporation; • Constructor - Transfield Pty Ltd; and • Engineering Consultants - ConneU Wagner Pty Ltd and Montgomery W atson Australi a Pty Ltd. Three aUiance partnershi ps were established fo r the sp ecialist ac tivities of: • Surveying - Hard & Forester Consulting Surveyo rs; • Barging - Adsteam; • Pumpsta tion electrical fit out - Kilpatrick Green. T hese specialists were integrated into the Alliance team to support and ·work alongside the other company employees . The organisational structure of the alliance was m ade up of two separate team s: • A Proj ect Alliance Leadership T ea m (PALT), w hich effecti vely was a Proj ect Board, made up of two representatives from each of the alliance participants; • An Integrated Proj ect T eam , headed by the Alliance Proj ect M anager, w hich comprised all functions requi red to deliver the proj ect and undertake the works at the va rio us constru ction sites. Some fun cti ons were centralised and others were decentralised to th e constru ction sites . The Integrated Proj ect Team was made up of personnel drawn from all four participants on th e basis of 'best person for the job' . In practice, these individuals saw th emselves more as m embers of the aUiance tea m , rather than emplo yees of their home o rganisations. Some of the requirements of the Norchside Storage Tunnel Alliance Agreement we re as follows: • All acco unting was to be totally open book . • Variations on the proj ect were to be limited to changes in proj ect scope. • The parties we re committed to resolving aU co nflicts within the alliance without any form of claims or litigious action . • A Target or 'Business as Usual' Cost Estimate for the proj ect was to be agreed and independently validated. • The alliance participants were to place aU profit margin and overheads at risk in the event of any cost overrun,

To solve your fluid handling problems, you may need more than just a pump. At Weir, we design and manufacture pumps of all types and sizes for every application you can think of. We have the pumps, the technology, experience and the people to give you the complete turnkey solution.

--·~ ~

weir

INNOVATIVE ENGINEERING SOLUTIONS

Weir Engineering Pty Ltd NSW

VIC

OLD

Ph: (02) 4349 2999 Fax: (02) 4349 2900

Ph: (03) 9866 2459 Fax: (03) 9866 3322

Ph: (07) 33471400 Fax: (07) 33471499

Ph: (08) 9479 9670 Fax: (08) 9479 9671

Ph: (08) 8346 9511 Fax: (08) 8346 7731

WATER SEPTEMBER 2001

PROJECT

w hich would be shared w ith the Client, Sydney W ater.

IMPLEMENTATION

Table 1

Risk/Reward Arrangements

Th e C o st Risk / Re w ard Model adopted for the project provided for the sharing of fin al savings or overrun between the alliance participants and the client, measured against an agreed and indep endently verified targe t estimate of project direct cost know n as the 'Business as U sual' Direct C ost Estimate. Part of the client's share of any savings was to be used to provide additional reward incentive for outstanding performance in the non-cost . obj ectives . The m aximum contribution of the alliance participants to cost overrun w as capp ed at to tal profit and ove rheads. Sydn ey W ater, as C lient, w ould always m eet the proj ect direct costs. A separate Time Risk / R eward Model provided reward or penalty relating to the time of Practical Completion of the project. A standard model was adopted fo r reward or penalty relating to environm ental, community relations and safety performance.

Best Shift (12 hr) Best Day (24 hr) Best Week (7 day)

6.6 me

3.8 me

47 m 70 m

80 m

328 m

377 m

57 m

A fee modifier, applying to overall gains, was incorporated in the paym ent m echanism to ensure that poor performance in any one objective was not traded off against good p erformance in another. Advantages of the Alliance Contract

The alliance contrac tual arran gem ent allow ed the Northside Storage Tunnel to be delivered as a true 'fast- track' proj ect. At a capital cost of approximately $460 million , th e proj ect was con ceived , planned and delivered in a total peri o d of less than four years. During the tw o-year constru ction peri o d some 2 1 km of drive n tunnel, 2 km of declines and caverns and 1.05 km of vertical shafts we re constru cted. Two underground pumping stations, o n e

located 100 m below sea level and each of over 16 MW installed pump capacity, were constructed at the North H ead end of the tunnel involving the placem ent of over 7000 1113 of concrete. 1,860, 000 tonnes of spoil were removed fr o m the tunnelling w orks and transported fo r benefi cial re-use, involving ove r 2,300 barge move m ents and 1,075 train loads. The four tunnel boring machines achieved a combined best week's adva nce of 1,093 m. Individually, the range of performance fr om the largest machine (6.6 m diam eter Wirth T BM) to the smallest m achine (3.8 m diam eter Wirth TBM) is sho w n in T able 1.

Th e a d va nta ge of th e allian ce contrac tual arran ge m ent w as clea rl y evident w hen extremely difficult ground conditio ns we re encountered in sectio ns of the tunnel crossing under Sydney H arbour. This area n ecessitated a preexcava tio n grouting program to be implem ented to redu ce the ground wa ter inflows from a m aximum of 200 litres p er second down to 1 to 2 litres per second. The alliance ;m ange ment allowed a to tal fo cus to be m aintained by all parties on overcoming the constru ction problems. Under a conve ntional contract serious disputes could have arisen over the allocation o f responsibility for such 'latent' conditions, leading to inevitable delays and asso ciated cost increases. The tunnel was available for use during the Sydney 2000 Olympics, had rain been sufficient to cause overflows into th e Harbo ur. Since wet commissio ning commenced in late J anuary the tunnel has now prevented some 13 ove rflow events totalling 3, 140 ML o f dilute sewage from entering the H arb our. Already it appears to be having an improvem ent to the wa ter quali ty of the harb our. D espite m any constraints and technical difficulties, the Northside Storage Tunnel successfully m et all of its proj ect obj ectives. This outcom e can be attributed to a great degree to the choice of an alliance contract as the m echanism for p roj ect delivery and the alliance partners. The Authors

Allan Henderson is Proj ect Director N orthside Storage Tunnel Alliance Sydney W ater and Russell Cuttler is Operations M anager,Transfield Pty Ltd, Tunnelling Division. 28

WATER SEPTEMBER 2001

WATER

The Problem According to the W odd Health Organization, increased risk of lung and bladder cancer and of arsenic-associated skin lesions have been observed at drinking water arsenic concentratio ns of less than S0ppb. The WHO established l 0ppb as a provisional guideline for arse nic in 1993. In Australia, the guideline is 7 ppb. In the USA a level of less than 50 ppb is mandatory and negotiations are proceeding on the feasibility of 5 ppb standard.

The Contaminated Wells of Bengal At the 19th AWA Convention, Rahman, Gurung and Shanableh (fro1'n QUT) presented a paper entitled "Arsenic Pollution of the Groundwater in Bengal Basin: Any Solution?" . The purpose of their paper was to review the situation and invite practical ideas and contributions from delegates . The fo llowing paragrap hs are edited extracts from that paper. In Bangladesh and W est Bengal (India), 90% of the population now get their drinking wa ter from hand-pumped tubewells. Since water quality from the tubewells was found to be accepta ble in terms of microbiological pollution compared to the highly polluted surface supplies, this significantly reduced water-borne diseases and in the 1970s was viewed as a success story in the public health sector in the region. Unfortunately, w hile the installation of tubewells substantially reduced wa ter-borne diseases in th e region, it caused the largest mass poisoning with arsenic in the wo rld . Arsenic pollution of groundwater was first noticed in the early 1980s in W est Bengal. Despite the potential risks, tub ewell installation programs continued without proper assessment of the risks posed to human health . In the 1990s, w h en hundreds of patients

Sewage Treatment and Disposal Expressions of Interest Expressions of interest are sought from organisations that may be eligible for a su bsidy through th e Advanced Wastewater Treatment Technologies Scheme (AWIT), which encourages th e introduction of innovative and efficient sewage treatment technologies into the State. The $7 million scheme aims to foster more cost-effective and environmentally friendly solutions to sewage treatment and enhance the options ava il able to local governments by progressing resea rch projects. To date, $4.3 million has been allocated to AWlT resea rch projects and implementation works. Organisations interested for co nsid eration of an AWIT subsidy are invited to apply in writing. Appli cants must address the selection criteria below and provide a brief explanation and cost analysis of th e project. Selection Criteria To be eli gible for these researc h/pilot plant grants, projects must: - be innovative: - be a pilot/demonstration or research (with proof of concept) project; - prov ide enhan ced va lue fo r money in terms of cost efficiencies and environmental and social impacts; - be feasibl e and have a so und conceptual basis; - have the capacity to be widely used within Queensland and within 10 years; and - involve Queensland local government participation. Detailed gu id elines and an application form can be obtained by contacting Rebecca Joyce on (07) 3225 8667 or Denise Sea rle on (07) 3225 8678 or email us at rebeccaJoyce@dlgp.qld.gov.au or visit www. dlgp.ql d. gov.au/local__govt/gra nts_funding.honl Express ions of inte rest will be reviewed by a technical panel. Candidates who appear to have a reaso nabl e chance of obtaining funding, will be invited to supply a more detailed proposal fo r rev iew. Express ions of interest should be marked "Private and ConfidClltial" and addressed to : The Manager, Local Government Funding Division Queensland Department of Local Government and Planning PO Box 31

BRISBANE ALBERT STREET QLD 4002

Queensland Government Department of

Expressions of Interest close on 31 January 2002

Local Governm ent and Pl anning

WATER SEPTEMBER 2001

WATER

presented themselves fo r treatment of arse ni c p o isoning (a rse ni cosis), the awa reness of the community of the seriousness of the problem was highlighted and gove rnment authorities started large scale m o nito ring of groundw ater. A number of surveys involving interna tio nal agencies have highlighted the pro blem . It is estimated that am o ng the fo ur million tubewells in Bangladesh, 40% were contaminated w ith arsenic well above the 50 ppb level. In 1998, for example, the British Geological Survey tested 2000 samples in one area at depths of 10 to 100 m . They reported that 35% were above 50 ppb, 25% were above 100 ppb and 8% were above 300 ppb. E ngineering and geological scientists es timate that in B an gladesh ,w ith a populatio n of 129 m.illion, the number of people now consuming arsenic contam inated wa ter is 28 to 35 millio n (using a drinki ng wa ter standard of50 ppb ( 0.05 mg/ L) currently used in Bangladesh and elsewhere) and 46 to 57 million people if a mo re conserva tive health guideline of 10ppb is used. Altho ugh there is adequ ate surface wa ter available in the rainy season (JuneSeptember) to store th.is fo r safe use in the

d1y season requires infrastru cture and trea tment w hich is currently beyond the financial and institutional resources of the 68000 villages w hich comprise the bulk of the population .

Occurrence of Arsenic in Groundwater T he arsenic contaminatio n in Bengal delta appears to be of geologic origin. T he wo rst affected aquifers are the alluvial deposits benea th the recent fl ood plains. The older sediments in th e central, western and south-eastern parts are not significantly contaminated. T he arsenic content appears to be a maximum at a depth of 10 to 100 m eters and appears to have been released from arseno- pyrite alluvi um by ingress of oxygen caused by draw-down of the wa ter table, fo llowed by microbiological reductio n to the trivale nt form. Gro undwater fro m depths of m o re than 150-200 me ters appears to be arsenic free. Arsenic is commo nly fo und as As(III) or As(V) in ground wa ter. As(III) can be co nve rted to As(V) under oxidizing conditi ons and As(V) can be converted to As(III) under redu cing conditions (e .g. anaero bic groundwater). T he co nversion

Th e National Ce ntre fo r Groundwate r Man age ment IN CGM) at UTS is recogni se d by the Fed eral Gove rnm ent through LWRRD C as a National Ce ntre for Trai ning , Research and Consultancy in Groundwater and Environm ental Appli ca tio ns. A co mprehensive rang e of HECS- ba sed post gradu ate pro grams is cur rently offe red: • Master of Eng inee ring and Maste r of Sc ience !Coursewor k - full or par t-ti m e) • Graduate Diplomas !En gi neering or Sc ience !full or part -t im e) • Mast er of Eng inee ring and Mast er of Science !R esearch - full or pa r t- time) • PhD !Full or part -tim e) • Gr aduate Cours e I15 w ee ks) Areas of study in th e coursework subje cts in clude: • Groundwater Contaminant Transport Modelling

WATER SEPTEMBER 2001

• Quality and Qu ant ity Optimi sa ti on Strategi es for Wat er Resource Deve lopment • Waste Managem ent and Groundwa ter • Co ntamin ated Land Evalu ation and Rehabili ta ti on • Bo re Fouling an d Mainte nance • Practi ca l Areas of Hydro geology • Land an d Groundw at er Sa lin ity Courses are flexi ble, w ith options in full -tim e !block model, part-time !block model, by di sta nce or in short course forma t. Li ke to know more? For information and appli ca tion s co ntac t Professor Michael J Knight, Director, Nati onal Ce ntre for Groundwater Man agement, UTS, PO Box 123, Broadway NSW 2007 Australia . think.change.do. 02 9514 1984 Fax: 02 9514 1985 Emai l : groundwater.management@uts.edu.au Website: http ://groundwater.ncgm.uts.edu.au/ncgm/

WATER

of arsenic between the tw o oxidation states is quite slow; hence th e reduced species can be found in oxidized portion of t h e aquife r s a nd v i ce ve r sa . Microorganism s, plants and animals can convert the inorganic arsenic compounds into organic species that are less toxic. T he As(III) form of arsenic is abo ut sixty times more toxic than As(V) (BCAS, 1997). In addition, As(III) is more mobile than AS (V) in groundw ater.

Removal of Arsenic In developed co untries , redu cing the concentration of arsenic in groundw aters to safe levels for drinking or from industrial or mining efflu ents presents only economic or minor technical problems. The arsenic is usually oxidised to the As(V) form by chlorine, hypochlorite, ozone, etc. and th en co-precipitated onto ferric, alum or lime flo es, w hich are removed in conventional settling or filtration systems. (Chapman, Wa ter, Sept. 1998). However, these powe rful oxidants are neither available nor safe for use in the small village environment which typify the Bangladesh situation. Som e mor e appropriat e technology is required.

Wa ste M anag e m e nt and Pollution C ontrol. T his proj ect w ill also establish a w ater q uali ty testing laboratory at D haka Community Hospital and do some extended research into th e risk to Bangladeshi communities of arsenic poi sonin g v i a th e food c hain. Organisations/ compa ni es involved in this proj ect include gis Consulting Australia in collaboration w ith D haka C ommunity Hospital, CS IRO Land & W ater, Adelaide and the CRC for W as te Managem ent and Pollution Control. This activity is already underway . Research to determine and demonstrate the most effective and appropriate technologies to remove arsenic and to dispose of the resultant arsenic contaminated sludge

If these qu estions are not answered Organisations/ companies involved in satisfa ctorily lives and money w ill be at this proj ect include Murdo ch University stake. Vast sums of money (US $40 in collaboration w ith R egional R esearch million from Wo rld Bank alone) have Laboratory, Orissa, India and researchers been committed to intervening in arsenic in Banglad esh w ithout an adequate from CSIRO and AquaR es . T his activity knowledge of the h ealth consequences of has also commenced. intervention and of the possible, indeed Can Australia Help? • Charlotte Walsh is th'e AusAID officer likely, consequences of ill designed interAs the arsenic phenomenon has been coordinating the program. Email: ve ntions. Organi sa tion s w h ic h ar e describ ed as an international p ublic charlotte_walsh@a11said.gov.au currently involved in contracting for this health catastrop he, the impoverished proj ec t includ e t h e South Asia region urgently needs interA u stralian National national m edical, financial, technical and University's National scientific support to better edu cate the atC entre for Epidemiology risk pop ulation , provide safe w ater . an d Publi c H e al th alternatives and prepare for the health and (NCEPH) in collaboeconomic consequ ences exp ected in ration w ith the National years to come from the long- term R esea rch C e ntre for consumption of arseni c by a large Environmental Toxicproportion of the pop ulation. ology, the (Bangladesh) AusAID fund ed some preliminary NGO For u m for trials in 1998/ 9 conducted by the CRC Drinking Water Supply W as te M anage m e nt and Pollution and Sanit a tion , and Control, and subsequently agreed with the Mon ash Uni ve r sit y Government of Bangladesh, and the D e partm e nt of Government of India in 2000 , to use aid Epid e m i o l o gy and money to address the issue of arsenic Preventative M edicine. contaminatio n in Bengal and Bangladesh A review and trial of ($4 million ove r three years starting July different treatment 2001) . T h e first three proj ects that methods for providing • water supply Au sAID is u nd ertaking und er the clean water with safe services umbrella Australian Arsenic Mitigation arsenic levels Program fo r Bengal and Bangladesh • operations and Th e s e in c lud e include: maintenance methods based on simple Research into the costs and benefits of absorption like CSIRO's arsenic mitigation programs; specifically • engineering 'T ea Bag' technology the research questions will be: services and other methods based • W hat are the health and social benefits on UV oxidation and of arsenic mitigation programs; and adsorption, such as those • Do the benefits outweigh the potential developed by ANSTO , a health and social risks of arsenic mitigation m ember of the CRC for programs?

Providing effective water infrastructu,e management

WATER

ARSENIC REMOVAL USING UV ASSISTED OXIDATION M Zaw, M Emett and P Prasad Abstract

of iron salts under illumination using ultra-violet light, either Arsenic (III) in water needs to from natural sunlight or from be oxidised before it can be effecblack-light UV lamps. T he t i ve l y r e mo ve d u s in g UV dramatically accelerated coag ul a tion / se ttling , ionthe process . (Em ect & Kho e, exc h a ng e, reverse osmosis, 2001). In 1996 the system was Âˇ microfiltration or lime softening demonstrated in Montana , methods. Since the oxidation rate USA, on acid water draining of dissolved arsenic(III) by oxygen from an abandoned mine . is extremely slow, oxidants such Th e B a nglad es hi Hi g h as chlorine, ozone or permanC ommi ss ion in Canberra ganate n eed to be used. Instead enquired w hether it could be of chemical oxidants , a n ew used co treat well water and oxidation process using dissolved with funding from AusAID sulfur(IV) (s ulfite) and ultraviolet the system was trialled in (UV) light has been developed A Bangladeshi visitor adds a small amount of iron to a B angladesh and developed and patented to accelerate - the days water supply to start its solar oxidation. into a demonstration proj ect, arsenic oxidation reaction with w here a few gra ms of iron air. The use of UV light accelsatisfacto1y disposal options. The majority sulfate were added to a 40 L tray of arsenic erates the oxidation of arse nic(III) to of these arsenic-bearing was tes come from polluted well-water ' which was then arsenic(V) in ne utral or alkaline conditions industrial activities. exposed to sunlight for a few hours, at by more than several orders of magnitude Arsenic occurs naturally in groundne utral pH. The precipitated ferri c in comparison to the commonly know n wa ter/ surface wa ter resources in two hydroxide, wi th co- precipitated arsenic, sulfite-o},,7gen auto-oxidation process. primary chemical forms : arsenite [As(III)] was allowed to settle o vernight and the This n ew process can be applied to and arsenate [As(V)]. T he predominant supernatant draw n off (Khoe et al 1999a). drinking water (surface waters or groundspecies depends on the pH , water quality The CRC are demonstrating this pro cess wa ters), industrial wastewater or process profile, oxygen content and other fac tors. in collaboration with CS IRO Land and liquors. Waste sulfur dioxide gas from Arsenic(III) in water needs to be oxidised Water and Egis Consulting as part of the sm elters may be used instead of sodium before it can be effecti vely removed using AusAID program. sulfite for non-drinking water applications. coagul ation/ se ttling, ion-exchange , Subsequently, we have developed and This process has been demonstrated at a reverse osmosis, microfiltration or lime p atented a pro cess using disso lve d pilot scale in a Water Treatment Plant in softening m ethods. Since the oxidation sulfur(IV) (s ulfite) as the photo-absorber. California. More commercial demonrate of dissolved arsenic(III) by oxyge n is (Khoe et al, 19996). T he use of UV light strations will be carried out using this extremely slow, oxidants such as chlorine, accelerates the oxidation of arsenic(III) to technology in the future. ozone or p ermanganate need to be used. arsenic(V) in neutral or alkaline conditions Key words: Arsenic, oxidatio n, ultraIn previous work, o ur team, as a by more than several orders of magnitude violet light, sulfite, geothermal water, m e mb e r of th e C R C for Wa st e in comparison to the commonly know n gro undwa ter. Mana gement and Pollution Control, sulfite-oxyge n au to-o xidation pro cess developed and patented a system of Introduction eve n in th e prese n ce of disso lve d oxidation by dissolved air in the presence carbonate w hich is common in gro undArsenic occurs widely in the earth' s Consequ ently, the applicability of water. crust (th e twentieth most abundant the sulfite-UV technology for oxidising element). As well as occurring in som e arsenic in surfa ce water/ gro undw ater gro undwa ters it is an unwanted byand geothermal wa ter w as evaluated. product of mining or/ and industrial activities such as extrac tion of the metals copper, cobalt, gold, lead, nickel and urani um. In Australia , as elsewhere in the wo rld, partially treated or untreated arsenic-bearing wastes are often scored in temporary storage loca tions p ending Thi s paper was presented at th e 19th AW A Convention and has sin ce been ed ited.

WATER SEPTEMBER 2001

Experimental Reagents and apparatus

Pilot scale open channel UV reactor for treatment of geothermal waters.

All solutions were ptepared from analytical reagent grade chemicals unless otherwise stated. Purified water (Milli-Q system) and acid-cleaned bottles were used for storage of samples. This pH was controlled using M etrohm Impulsomat (No. 614) and Dosimat (No. 725) .

WATER

The arsen i c- c ontaining sample (1 .7 litre) was placed in 2 L reaction vessel. Air was bubbled at a rate of 2.5 L p er minute and the solution was adjusted to pH 9 using sodium ca rbonate. Sodium sulfite was usually added as a single dose. However, for some experiments (especia lly for hi g h arse ni c concentrations), sodium sulfite was added continuously to avoid the risk of th e dissolved 0>-.')'gen b ecomi n g d e pl ete d. The solutio n co ntaining arsenic(III) and sulfite was illuminated w ith UV light from a 15 W lowpressure m ercury (non-ozone producing) lamp. The corresponding changes in arse nic(V) concentrations in the present and absence of UV li ght were measured by the molybdenum blue spectrophotometric method Qohnson and Pilson 1972) . Drinking water

"O (l) (/)

100

"O X

-> 0

(/)

80 Arsenic (light)

-+-

Sulfite (light)

--e-

Arsenic (no light)

( /)

~ 0

Results and Discussion 0

0.5

1.5

2.5

Elapsed time (minute) Figure 1. Percent of arsenic( II I) and su lfur( IV) oxidised in an aerated so lution with and without 254 nm ill umination (11 mW/ cm 2 ) (Initial As(II I) 0.47 mg/ Land S( IV) 4 mg/ L: pH 9: air sparged rate 2.5 l / min) 100 "O (l) (/)

:2 X 0

....

-+2.4 mg/ L ~

(/)

4 mg/L 3.2 mg/L

-~ (l)

The geo thermal wate r was illum..inated with UV light from a 15 W low-pressure mercury lamp. The con-esponding changes in arsenic(V) concentrations in the present and absence of UV li ght were m eas ured by the molybdenum blu e sp ectrophotometric m ethod m ention ed above.

1.6 mg/L

-----

T he sulfite/ UV process has been deve l o p ed for th e oxidation oflow arsenic concentrations fo und in groundwater and for the higher concentrations found in geothermal wa ter. In the first application sodium sulfite is added as a single dose at the start of the experiment. This is possible, as the small amo unt s required d o n ot consume all of the dissolved oxygen and create redu cing conditions . In the second ap plica tio n , w here th e sulfur (IV) dem and is higher, the so dium sulfite is added continuously at a set rate so that the consumed oxygen can be replenished. T his later method better simulates the sparging of a waste w ith a sulfur dioxide and oxygen gas mixture, which is possible for geothermal wa ters .

0 mg/L Several litres of ground wa ter <( were obtained from a small bore, ~ 0 20 w hich supplies drinking water for a town in Western Victoria. 4 8 12 16 20 The water was clear an d Illumination time (minute) colo urless , and pH of the water was 8 .1. Major cations and Figure 2. The effect of sulfite conce ntration on the an i ons were ana l ysed b y percent of arsen ic(II I) oxidised (Initial As( III ) 0.47 mg/ L: Indu c ti vely Co upl e P lasma S(IV) doses varied from Oto 4 mg/ L: pH 9: air sparged Atomic Emission Spectroscopy rate 2.5 l / min). (I CP-AES) (Perkin E lm e r , Application to drinking water Optima 3000DV), Indu ctively Effect of UV light 100 Co upl e Pl as m a M ass "O A soluti on co ntaining 0.47 Spectroscopy ( ICP-MS ) (l) (/) mg/ L of As (III) was adjusted to 80 (Aglient, HP 4500) and Ion '6 pH 9 with sodium carbonate . Âˇx C hrom.atogra phy. Before each 0 Sodium sulfite was then added 60 experiment, 1. 7 litres of ground-~ to give a rnlfur(IV) conce nC wate r was placed in the reactor (l) tration of 4 m g/ L and the vessel and aera ted. A continu ous ~ 40 <( solution vJas aera ted w ith the air sa mple was w ithdraw n at 8 mL ~ flo w rate of 2 .5 L p er minute. 0 per minute and analysed fo r 20 The sulfite co ncentration did As(III ) by co ntinuous flow not change for at least 3 minutes hydride vapour generation. T he 0 and as sh own in Fi gure 1 4 2 3 5 6 7 0 arsine vapour was then analysed neither did the concentration of Elapsed Time (m in) using an ICP-AES. When a arse nic(V). When the expersteady analysis was ac hieved, the Âˇ Figure 3. The effect of pH on the percentage of iment was repeated with UV sodium sulfite aliquot was added arsen ic( III) oxidised with 254 nm illumination. (Initial li g ht , the arsenic(III) was and the clock started. After 30 As(III) 0.47 mg/ L: S(IV) dose rate 0 .8 mg/ l / min: air co mpl e tely oxidised in 1. 5 seconds of mixing the light was sparged rate 2.5 l / min). minutes and the sulfite was also turned o n . Then the samples geothermal wa ter was placed into the oxidised in 1.8 minutes after a sh ort were w ithdraw n and an alyse d for reaction vessel and aera ted. The pH was indu ction period . Arsenate and sulfate arsenic(III) as mentioned above . then adjusted to the starting point of 8.4 were confirmed to be the onl,y products. Geothermal Water with either dilute sodium hydroxide or Clearly, the oxidation of arsenic(III) is initiated by UV light. T h e light initiated lime. Sodium sulfite was then added to Several litres of geothermal water from give a sulfur(IV) concentration of 10 this pro cess involv ed the primary the South America were received for m g/L per minute. Where large doses of production of the sulfur(V) radical bench scale tests . Major cations and anions sulfite would depl ete the dissolved (S O 3 -) from su lfu r(IV) (SO /- ) were analysed using ICP-AES / ICP-MS (Backstrom, 1927) and fo llowed by a and Ion C hromatography. One litre of oxygen, sulfite was added continuously.

WATER SEPTEMBER 2001

WATER

300 50

Spiked

250

- - â&#x20AC;˘ - Âˇ Non e-S( IV) 5 ppm

.E 200 :::i Cl 2, 150

> VJ c,::

100

<i, c,::

None, S(IV) 2.5 ppm Iron , S( IV) 2.5 ppm ---+-- Org ., S( IV) 2.5 ppm

50 0

0 0

9 2

Light Intensity (mW /cm

)

Figure 4. The effect of light intensity (254 nm) on the rate of arsenic( III ) oxidation (Initial As( II I) 0 .5 mg/ L: S( IV) doses 3.2 mg/ L and pH 8.65)

Effect of light intensity A simplified synthetic gro undwater was prepared w hich contained 10 mM of Effect of sulfite dose sodium bicarbona te and 0 .5 m g/ L of The effect of sulfur(IV) concentration arsenic(III). The pH of this solution was on the rate of arsenic(III) oxidation is 8.65. Sodium sulfite was added to the shown in Figure 2. In the absence of reaction mixture to give a sulfur(IV) sulfur(IV) no arse nic oxidation occurred concentration of3 .2 mg/ L. The solution which confirms that arsenic(III) is not was aerated. T he effect of the intensity of the light using this synthetic groundwa ter being directly excited by the UV light. is sh own in Figure 4. The rate of arsenic(III) oxidation increased The rate of the arsenic(III) oxidation steadily" w ith additive of sulfite until the is proportional to the intensity of the light sulfite concentration reached optimum in the range produced by comm ercial dose [ie. 3.2 mg/ L of sulfur(IV)]. lo w -press ure merc ur y lamp s. This Effect of pH indi cates that the rate of light absorption controls the observed ra te. Free radical The effect of pH on the rate of arse nic propagation and termination reactions oxidation is sho w n in Figure 3. T his pH generally have very fas t rate constants and range was chosen as it corresponds to that it is concluded that results indicate that of natural gro undwa ters. The rate of the initiation step is rate controlling. arsenic(III) oxidation was observed to Further indi cation of this initiati o n increase as the pH was raised across the limited kinetics was found in a separate range. T h e oxidation of arsenic(III) was experiment in w hich the light was turned completed in 5 minutes at pH 8 and 1 off w hen only one third of the arsenic(III) minute at pH 9.5 in the presence ofUVwas oxidised and the oxidatio n rate light. immediately was reduced to that of a dark reaction. Furthermore, the lo w absorbance of th e so luti o n results in the Table 1. Compositi on of t he Vi ctorian groun dwater majority of the light being and geot hermal water from South Am eri ca transmitted to the o uter Element Victorian water Geothermal water PVC reactor wall. As the ( mg/ L) (mg/ L) light is available the excited sulfur species are evenly Arse nic(I II ) 0.043 25 distributed and no diffusion Arsen ic(V) 0.015 4 .7 limitations can occur. Ca lcium 244 19.9 Potassium

3.95

140

Effects of Iron and organic

Sod ium

150

3300

compounds In o rder to determine the effe ctive n ess of the process in real gro undwater, several tests were conducted using a W estern Victorian groundwater with the composition show n in

7.3

Silicon

6.9

35.9

Chloride

170

6210

Alkalinity as HC0 3

174

TOC pH

8.1 WATER SEPTEMBER 2001

7.1

Figure 5. The effect of iron hydroxide and fulv ic ac id on th e rate of arsenic( III ) oxidation for Victorian groundwater (in it ial As( III ) 0 .05 mg/ L: pH 8.1)

subsequ ent radical chain reaction to oxidise d arsenic(III ) to arsenic(V) .

Su lfur

40 60 Time (second)

Table 1. Seve ral litres of gro undwater were obtained from the small bore, w hich supplies drinking wa ter for a tow n in Western Vi ctoria . T h e arsenic( III ) concentration of this wa ter was determined in the field to be 0.043 mg/ L (total arsenic 0.058 m g/ L) , however som e oxidation of the arsenic occurred during storage and arsenic trioxide was used to spike the arsenic(III) concentration to 0.05 mg/ L before each experiment. As th e arse ni c ,co n ce ntration in Victorian gro undwater was an order of magnitude lower than in the previous tests selective vapour generation was used for the analysis. T he effects of pH and sulfite dose were similar to the earlier experim ents. When 5 mg/ L of sulfite was added to the wa ter at pH 8. 1 and illuminated the arsenic(III) was oxidised in less than 10 seconds (as shown in Figure 5) . Groundwaters often contain iro n (II) w hich rapidly oxidises w hen it is aerated afte r surfacing to form a colloidal iro n hydroxide precipitate . In Figure 5, the sulfur(IV) concentration was redu ced to 2.5 mg/ L to allow the effects of interference to be m ore easily studied . When 2 m g/ L of iron (iron as ferric chloride) was added, the alkalinity of the wa ter caused the aqu eous iron(III) to rapidly hydrolyse and a precipitate of amorphous iron hydroxide rapidly formed . This increased the absorbance of the solutio ns at 254 nm from 0.025 to 0.12. The presence of iron impurities did not affect the overall rate of arsenic oxidation. Howeve r the iron did make it difficult to oxidise more than 95% of the arsenic(III). It is possible the unoxidised arsenic has been occluded in the irbn hydro xide precipitate and is therefore inaccessible to the light . The unoxidised arse nic wo uld be removed along w ith the arsenic(V) during subsequent filtration. The effect of organic compo und w as tested by adding fulvic acid (10 m g/L) to

WATER

rate. Normally organics in groundwater are present at an order of (b) no UV (a) UV light m agni tu de lower concentration so 240 :J' this did not raise any concerns. After c, 30 30 200 :J' .§. the arsenic oxidation is complete the ~ C) 160 c, E > residual sulfite rapidly oxidises with .§. ::: 20 ~ 20 120 ~ th e UV ligh t. Other dissolved - + - As (V) c:. 0 C/1 - - S(IV) in contaminants such as manganese(II), < 10 80 ~ 10 iron (II) etc can also be rapidly < 40 oxidized . 0 0 The process is n~ w the subject of 20 0 10 30 40 40 10 20 0 30 a patent jointly owned by ANSTO Time (min) T ime (m in) and the CRC WMPC. ANSTO have previously demonstrated the Figure 6. The effect of (a) UV light and (b) without UV light on t he oxid ation of t he South concep t of this process for drinking Am eri can geothermal water (Initial As( III ) 25 mg/ L: pH 8.4 : S(IV) dose 10 mg/L/ min ) water t rea tmen t i n t h e wa t er treatment plant of a small C alifornian town. W e are now preparing for a th e Vi c to rian gro u ndwater. T h e and no residual sulfite in the solution high- profile demonstration in the eastern absorbance of the water at 254 nm was (Figure 6a). Complete arsenic oxidation USA in collaboration with an American increased to 0.244 and produced a can be achieved about half an hour engineering company and a university . without ultraviolet light (Figure 66). T he definite brown colo u r. T he initial sulfite concen tration in creased to a sulfur(IV) concentration in this experAcknowledgements m aximum at 30 minutes and then iment was 2.5 mg/ Land the pH was 8.1 . This work was partly fu nded by the As can be s·een in Figure 5, this dram atically dropped to zero within one CRC for W as te Managem ent and increased the time required fo r oxidation ho ur. T hese experiments were repeated Pollution Control Limi ted , a centre and the solutions pHs were adjusted with of arsenic(III). T he fulvic acid app ears established and supported under the sodium hydroxide. R esults were very to slow the rate process . It is infe rred that Australian Government's Cooperative similar using either sodium hydroxide or the fu lvic acid is interfering w ith the li me. H ence inexpensive lime can be Research Centres Program. T he authors propagati ng free radical reactions . T he used in place of sodi um hydroxide. The w ish to express their appreciation to Dr TOC level used for this experiment is results also show that UV light not only ten times grea ter than the level whic h G. Kho e for his technical advice during w as fo und naturally in the groundwater accelerates the oxidation of arsenic(III) this project. so no noticea ble effec t would b e to arsenic(V) by more than four orders The Authors e21.-pected in typical gro undwaters. More of m agnitude but also reduces the than 95% of the arsenic(III) was oxidised requi rement amo unt of sulfite salt by Dr Myint Zaw, Dr Maree Emett w ithin one and half minutes . eight times . and Ms Patricia Prasad are R esearch Officers of the Environment Division , Application to geothermal water Conclusions Au stra lian N u cl ea r Science and The composition of the geotherm al T h e arsenic level in drinking water Technology Organisation wa ter from South America is also shown needs to be reduced to WHO and the in Table 1. T he solu tion pH was References Australian drinking wate r limit because adj usted with lime to 8.4. Sodium sulfite Backstrom, H. L. J. (1927) The chai11-reaction of the concerns for cancer risk. A new was added continuo usly so that the total theory of negative catalysis. J. Am. C hem. Soc., adva nced oxidation process using sulfite sulfur concentration increased by 10 mg 49, 1460-1472. as th e photo-absorber h as b een / L per minute. As depicted in the Figure Johnson D. L. and Pilso n M. E. (1972) developed to oxidise Arsenic(III) in 6, the experim ents were conducted with Spectrophoto111etric Deter111i11atio11 of Arswite, wa ter before removing it by coagulation / and w itho ut UV. Arswate, and Phosphate i11 Nat ura l Waters. settling, ion-exchange, reverse osmosis , A11alytical Chimica Acta, 58, 289-299. The use of ultraviolet light accelerated microfiltration or lim e so ft e nin g Emett M T and Khoe G H (2001) Water Res the rate of arsenic oxidati on wi th methods. In all cases the use of UV light, 35 ,3 : pp 649-656 completion occurring within 8 minutes Khoe G H, Emett MT, Zaw Mand Prasad P esp ecially in neutral and alkaline condi(1999a) R eport to AusA ID . Australian tions, significantly improves the rate of ...-- --=,---------. "OI ucl ear Scie n ce a nd Te c hn o logy N arsenic oxidation in compariso n to the Organ isation § commonly known sulfite-oxygen autoKhoe, G. H ., Zaw, M. Prasad, P. S. and Eme tt, ~ oxidation process . Generally, the rate of :0 M . T. (19996) Photo -assisted oxidation of "'"" arse nic(III) oxidatio n either surface i11orga 11ic species in aq11eous so !H tio11s, "'s: wa ter/ gro und wa te r or geo th erm al lllternational Publica tion Number: WO 9 99 / 05065, International Publication date: 4 increases w ith the intensity of the UV Februa1y 1999. light, sulfite concentratio n and solution HMRC (1996) Australian Drinking Water N pH. Where large doses of sulfite wo uld Guidelines. NHMRC and Agric. a11d R esource deplete the dissolved oxygen it is better Mi11i11g Co1111cil of Australia al!d New Zealand. added continuo usly. Iron impurities up Co111111011wea lth of Australia, p FS93. to 2 m g/L were fo und not to affec t the W HO (1993) Guide/in.es fo r drin ki11g-1vnter oxidation rate. H owever, the presence Single chamber bench q11ality. World Health Organization Geneva, scale UV reactor. of dissolved organics reduces the reaction p 41. 40 . . - - - - - - - - - - - - - ,

280

WATER SEPTEMBER 2001

WATER

POLYALUMINIUM COA IN WATER TREATM SOME VICJD

Completed by Fisher Stewart under an EPC contract to US Filter/ Vivendi Water, the 8 ML/ d in-filter/ OAF Daylesford Filtration Plant uses PACI as the primary coagulant.

Abstract

Polyaluminium Coagulants

Polyaluminium coagulants are finding increasing use in potable water treatment plants througho ut Australia, with polyaluminium chloride (PACI) in particular now havi ng wide application . This paper reviews the properties and adva nta ges of using these chemicals with particular reference to experience at Daylesford , the Grampians region and Swa n Hill , in Victoria .

A number of alternative aluminiumbased coagulants have been developed fo r water treatment applica tions. T hese compo unds have the general for mula (Al 11 (OH)111 C l(3n-m))x and have a polym eric stru cture, totally soluble in wa ter. T he length of the polymerised chain, molecular weight and number of ionic charges is determined by the degree of polym erisa tion. These highly polymerised coagulants include the following: • polyaluminium chloride (PACI, n=2 and 111=3), • aluminium chloro hydrate (ACH , n =2 and m=5), and • polyaluminium chlorohydra te (PACH ): similar to AC H . In practice, there is li ttl e difference between the performance of ACH and PAC I in water treatment applications, even though ACH is more hydrated. Polyaluminium silicate sulphate (PASS) is another polyaluminium coagulant that has been used with some success in Canada and the UK but to date, not in Australia. An important property of polyaluminium coagulants is their basicity. T his is the ra tio of hydroxyl to aluminium ions in the hydrated complex and in general the higher the basicity, the lower w ill be the consumption of alkalinity in the treatment process and hence impac t on pH.

Key word s: Polyaluminium chloride (PAC I), Aluminium c hl oro h ydrate (ACH), Coagulants, Water Treatment

Introduct ion Alum (aluminium sulphate) is the m ost commo nly used coagulant in Australian wa ter treatment plants- low cost being its major attrac tion . Alum however, has a number of disadva ntages : • limited effective pH range : typically between 5.5 to 6.5, • supplemental addition of alkalinity to the raw water is often required to ac hi eve an optimum coagulatio n pH, particularly for soft , coloured surface wa ters that are common in Australia, • residual aluminium levels in the treated wa ter can often exceed acceptable limits, and • alum flo e produced is particularly fragile, often necessitating the need for polyelectrolyte addition.

WATER SEPTEMBER 2001

For exa mple, aluminium chlorohydrate has a basicity of 2.5, .PAC I 1.5 and alum (Al2(SO 4h) 0.20 (Lu, et al 1999) . T he mass percentage of aluminium monomers de creases linea rly as th e basicity of the coagulant increases Qiang and Gra ham,1997). Various suppli ers of AC H and PAC I in Australia express the basicity of their produ ct as a percentage e.g. Orn.ega M EGAPAC-23 (40.2% w / w aluminium chloro hydrate) has a basicity of 83% (Omega C hem..icals, 2000). Table 1 summarises the characteristics of commercially available polyaluminiu m coagulants. Details for alum and sodium alum..inate are also included for comparison. Unl ess othe rwise no ted, alum is considered to have eighteen molecules of water of hydration throughout this paper.

Advantages of Polyaluminium Coagulants These highly hydrolysed coagulants are effective over a broader pH range compared to alum (AWWA Coagulation Committee, 1989). Experience shows that PAC! works satisfactorily in a pH range of 5.0 to 8 .0. A very important advantage of using polyaluminium coagulan ts in wa te r treatment processes is the reduced concentration ofSO 4 added to the treated wa ter. T his directly affects SO 4 levels in domestic was tewater. A raw wa ter w ith a sulphate level of 3 to 5 mg/L w ill

WATER

Table 1: Typical aluminium-based coagulants used in water treatment (Typical prices are on the basis of 15-20 kl load ,

ex-works Melbourne) CHEMICAL SYMBOL

SUPPLIER

NAME

COMPOSITION

FORMULA

STRENGTH, % w/w

WT.

Al 2 0 3

ASIS

SG at 20°c

OTHER

TYPICAL PRICE EX WORKS MEL, $/kg ASIS

100%

Aluminium Chlorohydrate

ACH

OMEGA

MEGAPAC 23

Al 2( 0H)5.CI

1 74.45

23.5

12. 4

40 .2

1.33

Basicity 8 3%, Chloride 8.5%, pH 3 .5

1.00

2.49

8.04

Alu minium Chlorohydrate

ACH

ALUMINATES

PAC 23 (APAC 3 33)

Al 2 (0Hl5- CI

174.45

23.5

12 .4

40 .2

1.331 .35

Basicity 83-85%, Chl oride 8 .0-8 .5%, pH 3 .5-4. 5

1.00

2.49

8 .04

Polyaluminium Chloride

PACI

OMEGA

MEGAPAC 10

Al 2(0H l3.Cl 3

211. 33

10.5

5 .6

21 .8

1.18

Bas icity 55%, Chlorid e 10. 5%, pH 3 .0

0 .50

2.30

9 .00

Pol yaluminium Chl oride

PACI

DELTREX

AC100S

Al 2(0Hl3.Cl 3

211.33

10.0

5.3

20 .7

1 .20

Sulphate 3 .5%, pH 3 .0

0 .57

2 .75 10.7 7

Pol yaluminium Chloride

PACI

DELTREX

SAB18

Al 2(0Hl3.Cl 3

211.33

18.0

9 .5

37 .3

1.38

Sulphate nil , pH 3 .0

0 .87

2 .33

9 .13

Aluminium Sulphate

ALUM

ALUMINATES, OMEGA

LIQUID ALUM

7.5

4 .0

49 .0

1 .30

pH 2. 5

0 .19

0 .39

4 .79

17 .8

9.4

28 .6

1 .50

12% NaOH 20 .1% Na20 pH 14

0.68

2 .37

7.19

Sodium Alumin ate

ALUMINATES

typically have a concentration of 15 to 25 m g/ L following treatment with alum. At a water treatment plant in the Otway region of Victoria , polyaluminium chloride replaced alu m and in so doing SO 4 levels in the treated water were reduced from 27 to 4 to 5 mg/ L. Previously, alum was dosed at 45 to 55 mg/ L at this plant. The change to PACl had a major impact on SO 4 levels in the sewage, with reduced odour problems evident at several pump stations in the sewerage system. Ocher advantages of polyaluminium coagulants include the fo llowing: • the high levels of hydration assis ts in achieving very low levels of residu al alumini um in treated water, typically 0.01-0.05 mg/ L, • PA C l and ACH work extremely well at low raw water temperatures . Floes formed from alum at low temperatures settle very slowly, whereas floes formed from polyaluminium coagulants tend to settle equally well at low and at normal water temperatures (Dempsey, et al, 1985; Matsui , et al, 1998), • less slu dge is produced compared to alum at an equivalent dose, • flo e formation and destabilization of colloidal particles (tu rbidity) occurs at a 40

WATER SEPTEMBER 2001

Al 2(S04l3 .18H 20 666.43 NaAI0 2 (Na 20 .Al 20 3)

81 .97

mu ch faster rate than alum (Matsui, et al, 1998) , • destabilization of finely divided particles is more effective than alum, • lower doses are required to give equivalent res ults co alum (Omega Chemicals, 2000). For example, a dose of 12 mg/ L PACI (3 .1 mg/ L Al) was required for treatment of a coloured, low turbidity water (Otway region, Victoria) compared to similar performance obtained when using an alum dose of55 mg/ L (4.5 mg/ L Al), and • the increase in chloride in the treated water is much lower than the sulphate increase from alum, resulting in lower overall increases in the TDS of the treated water. From Table 1, polyal uminium coagulants are typically twice the price of liquid alum on per kilogram al uminium basis. However lower doses of the polyal uminium coagulant and lower pre- and post-treatment alkali doses can still make its use economical. Pofyalu minium chloride solu tion (10% Al2 O 3 ) is stable for 4 to 5 months w h en stored at less than 50°C and so is ideal for bulk storage and dosing installations. One possible disadvantage in using ACH / P AC! relates to the removal of

dissolved organic carbon (DOC) from water. It is well documented that effective DOC removal is . possible with alum, particularly when coagulating at lower pH values using so-called "enhanced coagu"lacion". Alum is often a superior coagulant as far as removal of hu mic and fulvic colour constituents are concerned. A higher coagulation pH is adopted with polyaluminium coagulants and it possible that removal ofTHM precursors may not be as complete as with alum.

Examples of use in Victoria The fo llowing examples illustrate that the advantages of using polyaluminium coagulants will depend on the particular raw water and location in question. For each case, equivalent doses as "mg/ L Al" are also shown in bracke ts to enable a more convenient comparison of coagulants. Daylesford

The Daylesford Water Filtration Plant is a new 8 ML/ d in-filter/ dissolved air flotation plant constructed by Vivendi Water/ US Filter, with pro cess and deta il ed des i gn, e n gi ne e ri ng an d procurement provided by Fisher Stewart. The plant treats highly coloured raw water from either the Wombat or Bullarco R eservoirs. Raw water charac-

WATER

Table 2. Raw Water Analysi s, Wombat Reservoir at Daylesford ASION mg/ L

AS CaC0 3 mg/L

EC µS/ cm

ASION mg/L

AS CaC03 mg/ L

CALCIUM

1.8

4.5

ALKALINITY

19.9

1 6.3

MAGN ESIUM

2 .3

9.5

CH LOR IDE

11.7

16.5

SODIU M

9.0

19.6

6.7

SU LPHATE

1.5

1.6

POTASSIUM

0.7

0.9

te ristics for the supply fro m Wombat R eservoir are de tailed in Ta ble 2. T he wate r is also ve ry corrosive, w ith a pH w hich ca n be as low as 6.4 and a calcium carbo nate precipitatio n po tenti al (CCPP) valu e of - 17 .5 mg/ L CaC O 3 (at 20°C) . Typically this water supply has a true colour of30 Pt/ C o units and a· turbidity of 2.6 NTU, altho ugh on occasio ns the colour can approach 100 Pt/ Co units. T he temperatu re of the raw water entering the treatment plant can be as low as 5°C in winter months. For ra w w ater at 20°C w ith a tru e colo ur of 60 Pt/ Co units, turbidity 3 .0 NTU and pH 6.7 , Wa terQ ual (a wa ter treatment and quality assessment m odel developed by th e auth or at Fisher Stewart) was used to compare the predicted performance of alum versus PAC I. A chlorine dose of 1. 5 mg/ L for disinfection and a target trea ted wa ter pH of 7.5 were adopted fo r each case. For an al um dose of 45 mg/ L (3 .7 mg/ L Al) and a coagulatio n pH of 6.9, pre- and p ost- treatment doses of ca ustic soda of 17 .9 and 4.7 mg/ L were required (to tal 22 .6 m g/ L). These proj ecti ons co mpare well w ith the res ults of j ar-test inves tigatio ns carried o ut to determine the treatability of raw wa ter supplies at D aylesford using alum (GHD 1996) . An equivalent PACl dose of 12 m g/ L (3. 1 m g/ L Al) was adopted and used in the Water Qual m odel. Addition of caustic soda at 3.4 mg/ L was necessary to achi eve a coagulation pH of 6. 9. A post- treatment alkali dose of 4. 7 mg/ L was required in this instance (total dose 8. 1 mg/ L) . C haracteristics of the treated wa ter for each coagulant optio n determi ned fro m Wate rQual are given in Table 3 . T he advantages of using PAC! in regard to the effect on treated wa ter TDS and sulphate levels are apparent. Note also a slight improvement in the CCPP value of the treated wa ter when using PACl. The chemical doses predicted from Water Qu al w hen using PA Cl compare very well w ith ac tual requirem ents at the D aylesford W ater Filtration Plant. Experience at D aylesford has also confirmed the suitability of PA Cl, with very effective flo cculation observed at the low water temperatures noted during plant start-up in July 2000 (5- 10°C). T he operational advantages of using PA Cl, particularly in cold m o nths, make it an attractive choice at D aylesford .

Table 3. Treated Water Quality Predicted Using WaterQua/, Wombat Reservoir, Daylesford , at 20°C COAGULANT

ccPP 2

TDS 3

ALUM

-2 .2

-8.2

2 1 .0

PACI

-2 .2

-7.9

1.5

1. LSI: LANGELIER SATURATION INDEX 2. CCPP: CALCIUM CARBONATE PRECIPITATION POTENTIAL, mg/ L CaC0 3 3. TDS: TOTAL DISSOLVED SOLIDS , mg/ L 4. S0 4 : SULPHATE, mg/ LAS ION

Swan Hill

R aw w ater for Swa n Hill is abstracted from the Murray Ri ver and has typical qu ali ty charac teristi cs give n in Tabl e 4. Alum doses in the range 30 to 60 m g/ L (2.4-4. 9 m g/ L Al) have been used to trea t this wate r, w hi ch has a tru e colo ur of 20-30 Pt/ Co units and a turbidity of20 to 40 NTU under normal rive r fl ow conditi ons. Pre-treatment dosing w ith lime is o nly required when alum doses grea ter than 30 mg/ L are required . T he coagulatio n pH is usually 6.3 to 6.5 and a post-treatment dose of 10- 15 mg/ L lime is req uired to give a treated wa ter pH of 7. 1 to 7 .3. Lowe r Murray W ater Authority has recently changed over fr om using liquid alu m to AC H (MEGAPAC 23) at the Swan Hill W ater Treatment Plant. As a consequ ence, the prac tice of pre- and post- treatment dosing with lime to adjust pH and alkalinity has now been discontinu ed. Typi cally, AC H doses of 6 to 12 m g/ L (1. 9-3.7 m g/ L Al) are required .

•

ALUMINATES CHEMICAL INDUSTRIES Quality & Reliability

Suppliers of PAC 23 , PAC 10, PACS, PASS and all new aluminium coagulants.

Suppliers of PFS, FS4 , ferric su lfate and al l new iron coagulants . As well as suppliers of a lum , sodium aluminate and sodium silicate for over 43 years . Ask us to help you identify the most economical coagulant for your water treatment.

Contact: Aluminates (Morwell) Pty Ltd Telephone : Email :

61 3 5134 5262 morwell@aluminates.net

Aluminates (Port Melbourne) Pty Ltd Telephone : Emai l:

LSl 1

THM levels in the trea ted wa ter are generally 40-5 0 ~Lg/ L w hen treating raw wa ter with a true colo ur of 90-100 Pt/ Co units. This is well below the current A WD G recommendation of 250 µg/ L. In this instance DO C rem oval appea rs to be satisfactory .

61 3 9646 2115 ptmelb@a luminates .net

Aluminates (Tas) Pty Ltd Telephone: Email:

61 3 6431 5499 tassie@aluminates .net

Aluminates (N.S.W.) Pty Ltd Telephone: Emai l:

61 2 4353 3388 nsw@a luminates.net

WATER SEPTEMBER 2001

WATER

Table 4. Typical Raw Water Analysis at Swan Hill ASION mg/L

AS CaC03 mg/L

EC µS/cm

CALCIUM

5.6

14.0

210

ALKALINITY

36.6

30.0

MAGNESIUM

5.4

22.2

CHLORIDE

35.0

49.4

22.6

49 .2

7.4

SULPHATE

8.0

8.3

1.9

2 .4

SODIUM POTASSIUM

The raw wate r has a pH of 7.4 to 7.7 and following ACH addition , the coagulation pH is 7.2-7 .5, reducing to 7.0 to 7.2 following chlorination (gas, 1.5 mg/ L dose) . T he treated water using ACH is slightly more aggressive : CCPP - 10.9 to - 13.3 mg/ L compared to -1 0.7 mg/ L with alum, at 20°C. By adding a small dose of lime (0.5-1.5 mg/ L) at the inlet of the plant and coagulating at a slightly higher pH (7 .6), the co rrosivity of the treated wa ter following chlorination could be improved to give a CCPP of -9.8 mg/ Lat 20°C, i. e. making the water only mildly corrosive. T he changeover to ACH has allowed Lower Murray Water to defer planned capital expenditure in upgrading the lime storage and dosing fac ilities at the Swan Hill Water Treatment Plant.

ASION mg/L

AS CaC0 3 mg/L

T here has been no noticeable change in the level ofTHM's in the treated water at Swan Hill since changing over to ACH. TypicalJy the total co ncentration of THM's in the treated water is 35-40 µg / L, even wh en trea ting raw wa ter w ith relati vely high colour levels . Further, there are no discernable levels of taste- and odour-causing compo unds in the treated water (Neaves 2000) . Grampians Region

ACH is currently being used at three of six recently constructed in-filter/ DAF wa ter treatment plants in the Grampians region of Victoria; at Birchip, C harlton and Rainbow. Raw water to these three townships principally comes from the WimmeraMalJee C hannel system. Typically, the raw

Super solutions f0r the water industry Join many of Australia's lead ing uti lities already enjoying •

top level hassle free service

•

excellent investment perform ance

•

competitive fees (no commissions)

•

defined benefit & accumulation divisions

•

a fu ll range of products and service s

Cal l Brian Towers on

WATER SEPTEMBER 2001

(03) 9248 5911

wate r has the fo llowing average characteristics: pH 8.5 - 8.7, TDS 500 to 600 mg/ L, true colour 10 to 15 Pt/ Co units and turbidi ty 1.5 to 2.5 NTU. The high pH and TDS levels make treatment of this particular water difficult and if alum is used, doses are higher than wo uld be normally expected; often 60 to 100 mg/ L (4 .9-8. 1 m g/ L AJ). For the C harlton raw water supply- pH 8.5, true colour "t3 Pt/ Co uni ts and turbidity 2.5 NTU, an alum dose of 110 mg/ L (8 .9 mg/ L AJ) was required for effective treatment in laboratory j ar- tests compared to 40 mg/ L using PAC I (10 .2 rng/ L Al). T he coagulation pH was 6.4 with alum and 7.2 with PACI, illustrating the reduced impact PA Cl has on pH. Lower resid ual AJ levels in the treated water were also ac hi eved using PA Cl (GHD 1998). Treatment of the wa ter supply at Murtoa (also in the Grampians region) using PAC! was also found to be effective. This wa ter supply is also largely derived from the Wimmera-Mallee C hannel and has a TDS of aro und 550 m g/ Land a pH of 8.6. Alum doses typicalJy in the range of SO to 70 rng/ L (4.1-5 .7 mg/ L AJ) we re fo und to be required, with supplemental addition of sulphuric acid needed to achieve a desirable pH and so avoid excessive coagulant doses . By contrast, PAC I doses required to give equivalent treatment were 16 to 22 mg/ L (4.1-5.6 mg/ L), w ithout the need for pretreatment pH correction (USF 1998). T he above results are consistent with actual plant operating experience at the treatment plants at Birchup, Charlton and Rainbow, w here typicalJy ACH is dosed at approximately 40 mg/ L (12.4 mg/ L AJ). Sulphuric acid is used at these three sites to red uce the pH of the raw water to approximately 7.8. However, no posttreatment addition of alkali is required to correct the alkali nity of the treated water follow ing disinfectio n . T he raw wate r quality at Rainbow is very similar to that at Birchip . T he results of jar-tests (GHD 1998) and projections from WaterQua l comparing the p erformance of alum and ACH at R ainbow are summarised in Table 5 for raw wa ter with a TDS of720 mg/ Land pH 8.4, at 25°C. T he coagulation pH adopted for alum was 6.7 and for ACH 7.6, w hilst the target treated water pH was 7.5 1 in each case. A chlorine dose of 1.5 mg/ L was ass umed in each case for disinfection. Using alum, the sulphate level in the treated water increases from 17.9 to 56 .8 mg/ L, whilst for ACH the corresponding increase is only 2.3 mg/ L.

WATER

And finall y, a wo rd of advice: w hen changing over from alum to PACl/ ACH, it is important to make sure sto rage tanks, do sing p u mp s and p i ping are all thoro ughly flu shed o ut w ith clean water, to avoid forming an " aluminium j elly" !

TABLE 5. Predicted Performance of Alum v ACH at Rainbow COAGULANT

DOSE, mg/L

PRE-TREAT & DOSE 1 , mg/L

POST-TREAT & DOSE 2 , mg/L

TDS, mg/L

SO4, mg/L

CCPP, mg/L CaCO3

800 750

56.8

-6 .2

20. 2

-6 .0

ALUM

NaOH , 2 7 .7

ACH

H2S04, 2.4

References

1. FOR PRE-TREATMENT pH ADJU STMENT

2 . FOR PO ST-TREATM ENT pH ADJ USTM ENT TO 7 .5

T h e fo ll ow in g c h e mi ca l cos t s (delivered to site) we re used to calculate the operating cost of each trea tment option: • Alum, 20 kL: liquid (47% w/w), $200/ t • AC H , 20 kL: liquid (40 .2% w/w), $1100/t • caustic soda , 1000 L IB C: liquid (46% w / w) $600/ t • sulphuric acid , 1000 L IBC: liquid (34% w/w) 900/ t, and • chlorin e (gas, 980 kg drums) : $1600/ t For the al um option, total chemical costs amount to $76.8 / ML, whilst for AC H $63.5/ML; a saving of approximately 17%. This example illustrates how lower operating costs can sometim es be realized when using AC H coagulant compared to alum. Superior quality treated wa ter is also produced w ith resp ect to TDS, SO 4 and sho ws a marginal improveme nt in the CCPP value.

Conclusions Polyaluminium coagulants - AC H and PACl - can often give significant

hydrology hydraulics

water treatment

water supply

advantages over alum, including: • reduced chemical costs, • lower residual aluminium levels in the treated water, • improv e d t r ea t e d water q u ali ty including lower TDS and sulphate levels and possibly higher CCPP valu es, and • lower slu dge production. In many cases, post-treatment pH adjustment using an alkali is not required, reducing the overall capital cost of the plant and adding to operator am eni ty . It should be pointed out these alternative coagulants need to be evaluated for each specific raw water in question, taking into acco unt proj ected chemical costs, trea ted water quali ty and operations and m ainten an ce iss ues. Howeve r, with increased competition in the marketplace, the unit cost of polyaluminium coagulants will no doubt decrease in the future, m akin g convers ion from al um to A C H / PACl mor e a ttr ac ti ve and, widespread in A ustralia. Limited information suggests that THM formatio n will not be compromised w hen using ACH / PACl bu t this sho uld be first confim1ed in the laborato1y dur:ingjar- tests with the raw water in question.

wastewater collection

wastewater treatment

A WW A Coagul ation Committee (1989), ] A WWA, 81, 10, 75 D empsey, B.A; et al, (1985)., JAWvVA, 77, 3, 74 G HD (1996), Dayleiford Water S11pply, R eport 011 Water Q 11ality l111prove111 ertt Works: Cemral H ighlands Water Authority , M elb o urn e, Austral ia GHD (1998), Design a11d Co11struciio11 Specifications for Birchip, Charlto11 a11 d Rain.bow W ater Trea t111ent Plai ,ts: C ra111pians R egional Water A uthority, Melbourn e, Australia Jiang, Ji a-Qian and Gra ham, N.J.D. (1997), Chemistry a11d In d11st1y, M ay, 45- 50 Lu, G; Q u, J; and Tang, H ; (1999), Water R esearch, 33 , 3, 807-8 13 Matsui, Y; et al, (1998) ]A WWA, 90, 10, 96 Neaves, K. (2000) Lower Murray R egional Water Authority, Private Com ,mm.ication, Mildura Omega C hem.icals (2000), Megapac 23 Iiifor111atio11 Leaflet, Melbourn e, Victoria US Fil ter (1998), Private Co 1111111,111icat io11, Melbourn e

The Author Peter Gebbie is a Senior Engineer in the W ater Industry Gro up at Fisher Stewart, Melbourne, with over 25 yea rs local and international experience in the water industry. H e is responsible for process design and detailed engineering tasks associated with the delivery of water and wastewater treatment proj ects. Tel: (03) 8517 9268 Fax: (03) 8517 9422 Email : peterg@fisherstewarc.com.a u

urban drainage

water quality

catchment management

WATER SEPTEMBER 2001

WATER

EFFECT OF MONITORING TECHNOLOGY ON REGULATORY COMPLIANCE FOR DRINKING WATER B Labza, J Rissman, P Donlon Abstract Should regulatory agencies permit compliance w ith drinking water quality standards to fall if a new analytical technology is introduced, but the underlying quality of water is unchanged? This is a major problem for regu lators, as new technologies are developed fas ter than the regulatory systems w ithin which they are used. This paper examines how Victoria responded to such a challenge, when faced with a new analytical techniqu e that was able to m easure the microbiological quality of potable wa ter supplies more accurately than before.

Key words: D efi n e d Substrate Technolo gy, Colilertâ&#x201E;˘ , Laboratory,

Drinking Water R egulation, E . coli, Coliforms, Compliance Monitoring

Introduction Microbiological Monitoring of Drinking Water

Enzyme substrate technology, also known as D efined Substrate T echnology (DST) , is an analy tical technology now employed worldwide for th e analysis of water samples for the indicator bacteria E. coli and total colifoims.CoWertâ&#x201E;˘ the DST method most commonly in use, has significant benefits compared to the m embrane filtration m ethods traditionally used for potable wa ter analysis. In particular, it permits wa ter authorities to respond to confirm ed results within 18-22 hours

rather than presumptive results. This method also simultaneo usly detects and enumerates total coliforms and E. coli directly from water samples wi tho ut requiring a further confirmatio n step. Compared to th e m embrane filtration method , D efined Substrate Technology is capable of detecting a greater numb er of coli forms. It is one of a number of newer te chnologies involving genetic-base d techniqu es for detecting bacteria rath er than the older method- based criteria. It was developed during the 1980s and was a pproved for u se b y th e US Environmental Protection Agency m 1989 (H ackett & O'Toole, 1999). Regulatory Background In Victoria

The method and m onitoring programs for compliance monitoring of potable water suppli es in Victoria are determined by the Victorian D epartme nt of Human Services . Under the H ealth (Quality of Drinking Water) Regulations 1991 , the Victorian water industry is obliged to use analytical laboratories acc redited pursuant to specific techni cal and manage rial requirem ents set by the D epartment. The acc reditation is carri ed out by the Nationa l A ssoc i atio n of Te stin g Authorities (NATA), in accordance with a Memorandum of Und e rstanding between the Department and NATA. Under this M emorandum, any new analytical method must be validated aga inst the traditionally used m ethods before it can be introduced across Victoria. In the mid- 1990s, the M emorandum referred to the 1984 edition of British R eport 71 to establish the technical requirements. T he M emorandum was subsequently updated to incorporate the 1994 edition of the British R eport (Report 71, 1994) . The p1incipal difference between the two editions of Report 71 is that the definition of coliforms had changed (Section 7.7 of the R eport). This rectified scientific anomalies with respect to the definition of colifo rms that were appearing with the development of the new technologies, w hilst also permitting the use of D efin ed Substrate Technology . This paper was presented at the 19th AW A Conve nti on.

WATER

Strategic Process D uring the mid- 1990s, it became increasingly appare nt to the Victorian Department of Human Services that new an alytica l m e th o d s a nd a c h ange d definition of coliform bac teria we re being developed w hich had not been considered by the regulatory m echanisms established in V icto ria. T hese developments challen ged the traditional pre-eminence of m embrane filtration and may have had significant longterm consequ ences on the way compliance data was assessed fo r wa ter supply systems in Victoria. Acco rdingly, the D epartment e111barked, in 1997, on a long-term strategic p rocess to address these issues. T he first step was to determine the significance of the new definiti ons of colifo rm bacteria and th e new methods, under the conditions commo nly enco un te red in Vi cto ri a. T h e Departmen t commissio ned a validatio n study in 1997, w hich ran in 1997 / 98. It involved comparati ve analysis, under controlled conditio ns, of samples taken fro m wa ter supplies w ith a ra nge of trea tment and disinfec tion regimes . T he stu dy area ranged fr om the M elbo urne metropolitan system to a va riety of sm aller surface and gro und wate r supplies across Victoria . T he validation study concluded that, alth ough there may be som e significant localised change in res ults for particular wa ter types, there was a high degree of correlation across Victoria for the methods tes ted. That is, the complian ce res ults fo r total colifo nns under the new tes t were usually w ithin ab out 10% of the fi gure obtained by m embrane filtratio n and the p revio us colifo rm definiti on , within the statistical co nfidence attributable to th e sam pling program. T h e co mpara tive compliance res ults for E. coli were closer still (H ackett & O'Toole, 1999). Interes tingly, this particular study fo und that most of the va1iation in the total colifo rms data arose fr o m a grea ter recove ry efficiency from D ST of colifo rm species that we re accepted under the old defi nitio n, ra ther than from detecting a w ide r range of organisms that we re . newly labelled as 'colifo rms' . T he results of the validation study were w idely publish ed by the D epartmen t in early 1999 (H ackett & O 'Toole, 1999) . Copies of the study repo1t were disni buted to all water authorities and acc redi ted laboratories in Victo ria early that year, along with a reminder that the new edition of R ep ort 71, with its updated defi nitio n of coliforms, had recently been approve d for use by the D epartment. Water authorities in Victoria could now choose from a variety of validated m ethods, such as the traditional membra ne filtration and M ost

Probable N umber (M PN), as well as Defin ed Substra te Technology. T he D epartment of Human Services stro ngly encouraged all wa ter authorities in Vic toria to consider the long-term benefits of im proved m ethods, witho ut m akin g them m andatory at this poin t. T his posed an interesting problem . It was necessary for wa ter au tho rities ac ross the state to use compara ble m eth ods, but the degree to w hich individual wa ter autho ri ties embraced the newly validated m ethods varie d w idely . Individual wa ter authorities w ho were keen to m ove to the new m eth ods ge nerally preferred their co mpetitors or p eer wa ter authorities to likewise mo ve, o r to have already do ne so . Very few wanted to m ove first, du e mainly to uncertainty over how regulators wo uld int e rpr et mi cro bi o l og i ca l compliance data possibly affec ted by the more rigorous detection capability of D ST. Consequ ently, very few Victorian wa ter authorities adopted DST during 1999. T hose w ho did so embraced it enthusias ticall y, usually fo r operational reasons or because the compliance data did not deteriorate fo r their particular supplies . T he wa ter authorities w ho adopted D ST, therefore, did no t wish the D epartment to ' turn the clock back' by enshrining the

previo us m etho ds. It is interesting to no te that D efined Substrate Technology had alrea dy been implem ented in So uth Australia by this time. Issues For Water Authorities

W hen the agreem ents and licences establishing microbi ological compliance requirements were w ritten, the traditional m ethods of m embrane filtr ati on and MPN were the mos~ commo nly used analytical m etho ds. W hile a number of Victorian wa ter autho rities we re unconcerned about potential method changes fo r E. coli and to tal colifo rms, so m e held the view that significant cost increases may result simply du e to a m ethod chan ge, w hile the quality of wa ter suppli ed to consumers was essentially unchanged . If the intro du ction of D ST signifi cantly redu ced compliance fo r to tal colifo rms, particularly fo r non- m etrop olitan supplies that were disinfected only, wa ter authorities may have been face d w ith additional op erational expenditure or expenditure associated w ith installation of additional wa ter treatment fac ilities . Consequently, wa ter authorities needed to think carefully about strategies to handle this issue and also ensure that public confidence in wa ter supplies was no t adverse'ly affected. W ater

POSITIVE DISPLACEMENtli Âˇ""

â&#x20AC;˘

Corro Pump

1:~r@:1 Project Pumps Ph: (02) 9792 6644 Fax: (02) 9792 6655 Web: www.issproject.com.au

4/ !32 Bryant Sreet, PADSTOW NSW 2211

Email: sa les@1sspro1ect.com.au

WATER SEPTEMBER 2001

WATER

authorities also wished to ensure that the Âˇco nsultation proc ess es tablish ed b y regulatory agencies considered their views on any prop osed change . Issues For Laboratories

T he imminent introduction of D efin ed Substrate T echnology raised issues for accredited analytical laboratories . D ST had been used extensively in the USA, Europe, Asia, N ew Zealand and South Australia befo re its introductio n into Victoria. Consequently, validation data was available co mparing th e results of tests using Colilertâ&#x201E;˘ to those using m embrane filtration, altho ugh the membrane filtration m ethods varied widely between states and countries (Adcock & Saint, 1997; Clark & El-Shaarawi, 1993; Cowburn et al., 1994; Edberg et al. , 1988; Gale & Broberg, 1993; Whyte & Finlay, 1995). The va lidation stud y (H acke tt & O 'Toole, 1999) sp ecifi cally targeted comparison of DST technology with the R eport 71 membrane filtration method used in Victoria . T herefore in- ho use validation for laboratories required only the demonstration of repeatability and reprodu cibility of results using Colilertâ&#x201E;˘ . As with any new m eth od introdu ction, laboratory staff were required to demon-

strate proficiency in its use . T his was achieved by participating in internal and external proficiency programs. Laboratory validation results and staff proficiency were assesse d by NATA , for eac h accredited laboratory, at routine audits condu cted prior to the proposed introduction of DST. Fo r labo ratories, DST can generate labour savings, as m edia preparation is not required and positive results do not need to be confirmed. Consequently, a number of wa ter auth orities requ ested price negotiation based on expected labour savings alo ne. H owever as there was an increase in materials costs, significant price red uctions were not usually possible in commercial laboratories. T he ptice laboratories negotiated with water authorities for their services was recognised by all parties as a contractual matter between those parti es, rather than a regulatory matter to be managed by Government. Laborato ri es were co ncerned that, unlike with membrane filtration methods, it was not possible to detect nuisance o rga nism s su c h as A e romona s and Pseudomonas using DST. Therefore many laboratories recommended specific tests for nuisance organisms in conjunction with the introduction of DST.

C&SBRAND AUSTRALIAN FILTER COAL

& C&SBRAND GRANULAR& POWDERED ACTIVATED CARBONS Phone or fax for sound technical advice from experienced personnel.

JAMES CUMMING & SONS PTY LTD 319 Parramatta Road AUBURN NSW 2144 Phone: (0 2) 9748 2309 Fax: (02) 9648 4887 Email: jamescumming@jamescumming.com.au Licence no : 1628 46

WATER SEPTEM BER 2001

QUALITY ENDORSED COMPANY AS / ISO 9002 STANDARDS AUSTRALIA

Consultation Process And Outcomes

The D epartment of Human Services needed to minimise future potential fo r variation in data across Victoria and ensure that data was as acc urate, reliable and comparable as possible. In early 2000, it considered w hether routine microbiological compliance mo nitoring sho uld so l ely e mp l oy D e fin e d Sub str ate T echnology, rather than the range of previously permitted m ethods. If implem ented , t his wo uld m ean that th e traditi onal membrane filtration and MPN methods wo uld be phased out for ro utine microbiological compliance monitating in Victoria . T he D epartment wrote to each stakeholder, water authority and accredited labora tory ac ross Vi ctoria , to determine the level of broad industry support fo r mandatory change and address any issues that were raised, particularly in terms of the potential impact o n water authori ty and laborataty business and the operational lead times required. Stakeholders were asked to clearly enunciate their position in w riting, providing any necessa ry supporting evidence. Although opposition to this change was encountered, the majority of water au~horities supported the proposal. T his eliminated mu ch hea ted verbal debate from the issue and paved the way for the n ext step. Th e Vi c torian Wat e r Indu str y Association, representing the water authorities, convened a Working Gro up of regulato1y agencies and water autho rity represe ntati ves , to d e b ate th e k ey compliance issues. Particular care was taken at the Working Gro up meetings to ensure that all individual participants realised that identical requirem ents were being established statewide and that standards for to tal colifo rms were not linked to particular methods. T he regulatory age ncies also conceded on some issues, w here doing so wo uld not generate risks to public health or compromise their prima1y duty of care to the community generally . T hese clear p erceptions were instrun1ental in the success of the outcom es . In order to fac ilitate the transition, and .to overcome the shortcomings of assessing water q uality on n um erical indicato rs alo ne, the D epartment inco rporated risk assess m ent w hen assessing total coliforms compliance. T his provided flexibility for those individual supplies where the introdu ction of a new analy tic;al method may have generated a significant change in to tal coliforms complian ce data, whilst not necessa1iJy indicating an underlying change in the microbiological risk arising from that supply . No change was permitted for E. coli compliance data, du e to the greater public health significance of this

WATER

indicator and the similarity between the results for the va rious methods in the validation study . All stakeholders agreed on a process w hereby there wo uld be no anticipated regulatory impact in th e event that a water authori ty's total coliforms compliance data changed in the first yea r of the mo ve, when compared to data based on m embrane filtration. A limit of 5% variation was selected, as this acco rded w ith the results of the validation study. Essentially, a simpl e straight pass/fail system w as replaced by a pass/conditional pass/fail system. This gave wa ter authorities th e opportunity to dem onstrate that suitable steps were being taken to manage risks in th e eve nt that num er ica l complian ce data fell by no more than 5%. In order to be considered, water authorities were required to demonstrate that: • E. coli data satisfi ed the established standards and; • any numerical change in total coliforms compliance data for the supply in question was simply attributable to an improved, more sensitive m ethod and; • the wate r authority had adopted a risk management approach to investigating any individual unsatisfactory microbiological results and; • there was no underlying change to the inherent microbiological risk in the supply. Water auth oriti es w ith nume ri cal compliance data for total colifor ms that had redu ced by 5%, w hen compared to their m embrane filtra tio n data from the previous year, needed to satisfy these four criteria in order for regulato1y penalties to be waived. With standards for total coliforms typically set at 95% annual compliance (for non-metropolitan urban supplies), this meant that to tal coliform results from 90 to 95% under DST may b e acce ptabl e in Victoria ( 11 011metropolitan) in 2000/0 1, under the restricted circumstances outlined above. In practice, redu ctions slightly greater than 5% for total colifo rms may be permitted in exceptional circumstan ces, in the unlikely event of an unforeseen statistical effect o n data for a particular town. T h e onus was clearly placed on water authorities to dem onstrate that they complied w ith these requirem ents if compliance reduced by the permitted 5%. Water authorities were therefore obliged to prepare and retain auditable records, to assist in this process. These records and activities were no more onerous than those that most w ater authorities wo uld keep to satisfy du e diligence requirem ents or as part of their op erational activities . D etailed instructions to this effect were set out as attachments to correspondence from the D epartment to the water industry, labora tories and key stake-

h o ld e rs in April and Ma y 2000 . Accordingly, DST was established in Vi ctoria on 1 July 2000 as the mandated method of monitoring for E. coli and coliform bacteria at consumer supplies. The first full yea r data w ill be available sho rtly after 1 July 2001. Although th e above pro cess has been established, total coliform compliance results for the majority of supplies around Victoria for 00/ 01 are nevertheless not expected to redu ce significantly . For wa t e r a uthor iti es, t h e ri sk assessm ent approac h that was developed significantly alleviated the major con cern of additional capital expenditure due to the change to DST m ethods. A number of w ater authoriti es and regulatory agencies also chose to include information about th e imp ending change in their 1999 / 00 annual reports (the final yea r of membrane filtration), which were published in 2000. Positive disclosure of information such as this has ensured that public confidence in wa ter supplies in Victoria is maintained.

Conclusions T echnology often leaps in front of the regulato1y systems designed to stabilise particular activities . The key to keeping reg ul atory pro cesses in st ep w ith technology is to recognise that uncertainty arisi ng from change is universal, and to plan accordingly. As a long-term strategy, it is never possible to keep tec hnological change at bay in a globalised world, nor is it possible to wait until all possible variation or uncertainty is eliminated. The only strategy open therefor e is to develop regulato ry pro cesses that can handle uncertainty and are flexible enough to accommodate futur e events, w ithout having to predict the exact nature of the future events in adva nce. In the case of a new analytical m ethod fo r microbiological analysis of potable water samples, the Victorian strategy of validation study, soft launch , detailed indusny consultation and flexible processes ensured that stakeholders were successfully involved in the process and that decisions we re soundly based on fact rath er than uncertainty. The formal adop tion of risk managem ent and the concept of a conditional pass were instrumental in this success . T he principles of risk managem ent and consultation demonstrated here are being incorporated in more detail in the wider regulato1y framework being developed for drinking wa ter quality in Victoria. These principles can also be easily adapted by other jurisdictions considerin g such a move . T he pro cess described can be applied to any new m ethod as it is indep~ndent of the technical features of a particular scientific m ethod.

Acknowledgements T he autho rs w ish to acknowledge the co ntributi ons of all age n cies and individuals involved in the change process described in this paper. In particular, we wish to thank Allen Gale of Goulburn Valley Water, Graeme Jackson of the Victorian Water Indust1y Association, Gra h am Pool ey of th e Vi c torian D epartment of N aturt l Reso urces and Environment, and Marcus Crudden, of the Victorian Office of the Regulator-General, for their invaluable assistance in the process described and for their comments and advice on this pap er. T he authors also w ish to thank D ea n Comrie of WSL Consultants Pty Ltd fo r reviewing th e mi crobiologi cal information in this paper.

The Authors Brian Labza is a policy adviser with th e D epartment of Human Servi ces (Victoria) , Julie Rissman , a microbiologist, is a Director w ith WSL Consultants Pty Ltd, Richmond , Victoria. Peter Donlon is Manager , E nvironmental Services, Goulburn Valley Water

References Adcock, P.W. & Saine, C '. P. (1997). Trials of Colilerc Sys tem. Water, March/ April. pp .2225 Clark, ]. and El-Shaarawi, A.H. (1993). Evaluation of Co mmercial Presence-Absence Kits fo r Detection of Total Coli fo rms, Escherichia coli, and other Indicator bacteria. Applied and E,111iron111ental Microbiology, Feb, pp. 380-388. Cowburn, J.K. , Goodall, T. Fricker, E.J. , Walter, K.S and Fricker, C. R. (1994). A prebrninary study of the use of Colilert™ fo r water quali ty mo nitorin g. Letters in Applied Microbiology 19: 50-52. Edberg, S.C., Allen, M. J., Sm.ith, D.B. and The Na tional Collaborative Study (1988). Natio nal Field Eva luation of a D efin ed Substrate Method for the Sim ultaneous Enumeration or Total Coliforms and Escherichia coli from Drink.ing Water: comparison with the Standard Multiple Tube Fermentation M ethod. Applied & Enviro11111ental J\llicrobiology, June, pp. 15951601 Gale, P. , and Broberg, P.J. (1993). Eva luation of a rapid , defined substrate tec hnology method for enumeration of total coliforms and Escherich ia coli in chlorinated drinking wa ter. Letters in Applied J\llicrobiology 17: 200-203 W hyte, R. and Finlay, R. (1995) . Monitoring the Q uality of Drinking Waters: Compariso n of a T raditional Method with Colilert™ and Colisure™. Water & J1Vastes in New Zealand . November, pp 43-46 H ackett,]. & O'Toole,J. (1999) . Trial ofColilertl 8TM / Q uanti-Trayn1 on Metropoli tan and Rural Drink.ing Water in Victoria to Evalu ate the Impact on Colifo rm Compliance R ates. D epartment of Human Services (Victoria), 1999 Report No . 216/ 98 R eport 71, 1994. The Microbiology of Water Part 1 - Drinking Water, R eport on Public H ealth and M edical Subj ects No. 71, Methods for the Examination ofWaters and Associated Materi als, HMSO , London

WATER SEPTEMBER 2001

WATER

MICROBIAL ACTION: A KEY TO CUPROSOLVENCY IN PLUMBING SYSTEMS? " M M Critchley, N Cromar, N McClure, H Fallowfield

Abstract In recent years, many areas of Australia Âˇ have experienced sporadic problems with copper contami nation of drinking w ater. T his is primarily due to the generalised corrosion of copper plumbing tubes, a pr ocess kno w n as c upro sol ve n cy . C uprosolvency causes increases in free copper concentrations in potable wa ter supplies, which has aesthetic and public health conce rns. The m ec hanism is complex and frequently attributed to water chemical composition. However, th e results of our research provide strong eviden ce for m icrobially influen ced corrosio n (M IC) .

Introduction Copper has been extensively used as a m ate rial in d o m esti c h o u se h o ld pl umbin g sys te m s. It is n o rmally considered corrosion resistant, altho ugh there are m any instances of in-service fa ilures . The corrosio n of copper can be both localised and gen eralised w ith copper released as bo th insoluble and soluble compounds. Three distinct types

of copper corrosion have been identified in drinking wa ter sys tem s. Pitting corrosion , characterised by localised pits or holes, and " blue w ater" conditions , characterised by the release of voluminous blue-green corrosion particulates into wa ter, occur commonly thro ugho ut the eastern states of Australia receiving soft, weakly buffered wa ters (Nich olas and Carlyle 1999) . C uprosolve ncy, characterised by the release of soluble copper ions, occurs mainly in areas receiving hard wa ter w ith high buffering capacities, including Brisbane and Adelaide drinking wa ter supplies (Cox and Dillon 1980) . Unlike pitting corrosion, cuprosolve ncy is rarely a concern for tube integrity, rather an issue for water contamination. Cuprosolve ncy is usually only evident th rou g h customer taste complaints, staining of materials in contact w ith wa ter or water discolouration . Concentrations of soluble copper can typically increase by several milligrams (per litre) due to

WATER SEPTEMBER 2001

Cuprosolvency and corrosion by-product release

c up roso l ve n cy du r in g over ni g ht stagnatio n. T he occurren ce of cuprosolve ncy in po table water supplies has implications for public health . T he Australian N ational H ealth and M edical R esearch Council

have all been impli cated in increasing aqueo us copp er concentrations (Cox and Dill on 1980). T he nature of copp er corrosio n problems in wa ter distrib ution systems tends to be sporadic or rand om , with only certain areas of systems affected.

health guideline value for copper in

This occurs despite affected and unaffected

drinking wate r is 2 m g/ L (NHMRC 1996) . The co nsum ptio n of high er con centrations causes stom ach cramps, nausea and vo miting (Pizarro et al. 2000) . There are also populatio ns w ith an in crease d su sce ptib ility t o co pp er , incl udi ng ind ivid u als w ith ge n e tic disorders such as W ilson's disease. C opper in drinking wa ter causes a characteristic metallic tas te, w ith detectio n thresholds ranging fr om 1 - 3 m g/ L (Z aca rius et al. 200 1). Con centrations of copper above 1 m g/ L can cause aes thetic probl em s, mainly by staining dom estic bathroom fittings as well as human hair during showering. Howeve r, exposure to high co n ce ntra tion s o f co pp e r ca n b e temporarily m anaged by flushing the plumbing for several minutes befor e use. C uprosolvency is generally related to conditions within distribution system s, especially water chemistry . W ater pH , hardness, oxygen content and temperature

areas receiving the sa me wa ter supply. Corrosion problems are also so metimes restricted to the extremities of distribu tion systems w here chlorine residuals are low and microb ial regrowth can occur (Arens et al.1 995).

Biofilms and Microbiologically Influenced Corrosion The conditions in distribution systems do not favo ur the survival of planktonic (ie. free-swi mming) microorga nisms. They are exposed to disinfectants and have low retention times w ithin systems. Microo rganisms survive by attac hing and colonising surfaces formin g biofilms, acc umulating o rga ni c and inorga ni c compounds (Le C hevallier et al. 1988) . T he stru cture of biofilms provides som e pro tection from disinfection . M icrobial extracellular polym eric substan ces can n eutralise and prevent the penetrati on of chlo rine and chloramines to cells (Le

WATER

C h evallier et al. 1988) . Orga nism s contained within biofilms, including bacteria, fun gi, protozoa and viruses, can provide a source of microbial innoculum for the bulk wa ter phase with their detachment. The potential ability of biofilms to h arbour p athogens also presents serious risks for public health (Rusin et al. 1997) . Microorganisms have been implicated as a maj or contributor in the corrosion of metals in both marine and freshwater environments. Microbiologically influ en ce d corro sion (MI C) h as b ee n implicated in the initiation of pitting, crevice and generalised corrosion in some nietals. H owever, MIC is not in itself considered a separate type of corrosion. T raditionally, MIC is considered an an ae robi c ph en o m e non (H amilton 1995) . Paradoxically, the high dissolved oxygen levels in drinking wa ter do not favour the survival of anaerobes, although it is known that aerobic iron oxidising bacteria can corrode iron materials in drinking wa ter distribution systems, thereby fa cilitating anae robic conditions under the corrosion tubercles (H amilton 1995) . C uprosolvency activity by aerobic heterotrophic bac teria found in drinking water biofilms has also been demo ns trated in lab oratory studies (Dutkiewicz and Fallow fi eld 1998) . Thus, a more complex picture may be emerging. Biofilms have been implicated in copper corrosion in hot water plumbing systems in many institutional buildings (Arens et al. 1995; W alker and Keevil 1990). Corrosion usually occurs at water

temperatures below 50Â°C and is generally acc omp ani e d b y co pi o u s biofilm formation and corrosion products of microbial origin . Problems are more co mm o n in buildin gs w ith lo n g horizontal runs of copper tube and stagnant "dead legs" of pipe work, which result in low chlorine residuals. The accumulation of sediments can also enhance biofilm formation by increasing the surface area available for microbial growth. This can also ca use the development of anaerobic conditions, allowing the growth of sulphate reducing bacteria and other corrosive anaero bes (W alker and Ke evil 1990) . Although requi red as a trace elem ent fo r microorganisms, copper in its free fo rm is considered antimicrobial and causes pro tein oxidation, cleavage of D N A and RNA and impairment of respiratory chain ac tivity (Domek et al. 198 4). There are several mechanisms by w hich microorganisms can survive in co pp e r e n v ir o nm e nt s . Or g ani c compounds in drinking water can chelate dissolved copper ions, counteracting their antimicrobial properties. Microbial aggregation by planktonic and biofilm cells m ay prevent exposure to copper , allowing the growth of 1101mally sensitive o rganism s (Ga rdun o et al. 1993) . Microbial genetic resistance to copper has been identified and may be inducible in the presence of copper (Gilotra and Srivastava 1997) . Copper can also induce physiological changes in microorganisms. Th ese includ e c h a n ges in ce ll adhesiveness and membrane composition, as well as the promotion of

increased exopolysaccharide production, which can shield bacteria from copper exposure (H arwood and Gordon 1994). H ot and cold water plumbing systems exp erien cing copper corrosion have responded to antimicrobial treatments. Water chlorination has b een show n to prevent the release of " blu e water" corrosion products from tube surfaces into the bulk water phase (Taylor 2001 ). Arens et al. (199 5) demonstrated both heat treatment at 65Â°C and antibiotic dosing in corroding copper plumbing decreased cuprosolve ncy, all of which is consistent with a microbial role in copper co rros10n.

Research objectives C urrent research in the D epartment of Environmental H ealth at Flinders University is direc ted towards investiga tin g th e r o l e o f biofilm s i n cuprosolve ncy in Adelaide drinking wa ter. T he specific aims of the resea rch are: 1. to determine biofilm accumulation in copp er plumbing systems receiving filtered and unfiltered drinking water supplies; 2. to examine the infl~en ce of biofilm composition on corrosion ac tivity; 3. to assess the effect of water quality on cuprosolvency by biofilms.

Biofilm Accumulation in Copper Plumbing Systems Biofilms we re id entifi ed w ithin domestic cold water copper plumbing systems in Adelaide Metropolitan and

PERMASTORE TANKS & SILOS v Glass fused to steel (ceramic) sheeting v All bolted (relocatable) construction v'

v v v

Low maintenance Suitable for water and a wide range of liquids Proven with over 100,000 units in use Competitively priced

Permastore Tanks & Silos (Australasia) 4-8 Leighton Place, PO Box 240, Hornsby NSW 2077 Ph:(02) 9477 7944 Fax: (02) 9477 7363 Email: clientservice@ harvestore.com.au

WATER SEPTEMBER 2001

WATER

sampled from one locatio n in Adelaide Hills areas receiving Table 1. Detection of biofilm in copper plumbing tubes the Adelaide Hills significantly filtered and unfiltered drinking receiving Adelaide filtered and unfiltered drinking water increased cupro solve ncy in wa te r supplies respec ti vely supplies unfilt e r ed drinkin g wa te r (Ta ble 1, C ritc hl ey et al. (Fig ure 2 ; C ritchl ey et al. Parameter Filtered water Unfiltered water 2 0 0 1a) . Th ese plumbin g 200 1a) . A species of Acido11orax systems we re no t undergoing Mean (n=S) Mean (n=7) was identified as a predonunant significant cuprosolvency based 2) 656.84 Total Biomass (µg TOC/ cm 828 .33 m ember of this biofilm. This o n wa ter stagnation assessspecies was also present in the m ents. Biofilms we re detected Viable Counts R2A (cfu/ cm 2 ) 9.00x10 7 1.19x10 5 2 9 5 other biofilms, suggesting spatial at levels comparable to biofilm 2 .7 5x10 Viable Counts R2A+Cu (cfu/cm ) 2.51x10 relatio nships of cells w ithin acc umul a ti o n on o th e r 2 Respiration (µg 0 2 min / cm ) 3.76 1.32 materials exposed to drinking biofilms may be impo rtant in initiating these effects. water (N iqu ette et al. 2000). Influence of Biofilm Composition Biofilm biomass was similar in both filtered T he results of experiments w ith pure on MIC of Copper a nd unfilt ere d wa t e r sys t e m s. I n culture biofilms of bac teria isolated from comparison, significantly higher biofilm Biofilms isolated from the copper the copper plumbing pip es demonstrated heterotrophic counts were detected in plumbing systems were used in laboratory that the m aj ority of viable bacteria (83% copper pipes receiving filtered w ater, w ith coupon exp e rime nts assess ing th eir of isolates) had no sig1uficant influ ence on obvio us sediment and particulates present po tential corrosion ac ti vity (Critchley et free copper concentrations (Critchley et in these tubes . al. 2001a). T hese included experiments al. 200 16). H owever, certain bac terial with mixed microbial population biofilms The levels o f biofilm detected we re sp ec ies (14% o f isolates) including from the in-situ copper pipes and pure influ enced by va riatio ns in the chemi cal Acido11o rax delafieldii and Co ryn.ebacteria sp. culture biofilms of isolated bac teria. In composition of the potable wa ter supplied were show n to increase cup rosolve ncy in most instances, the mixed population to the plu mbing syste ms. Diss olved comparison to corrosion occurring in the bi ofi lms isolated from copp er pip es organic carbo n and soluble phosphate sterile controls (Critchley et al. 200 16). were shown to increase the amount of receiving filtered wa ter were shown to The species Acido11orax has previously been decrease the release of copper into the biomass detected (Critchley et al. 2001a). isolated from ac tively corroding copp er aqu eous phase in co mparison to sterile [n comparison, increasing wa ter io nic plu mbing systems (D avidson et al. , 1996) . controls (Figure 1; C ritchley et al. 2001a) . content and alkalinity decreased to tal Biofilm bac teria were also identified In contras t, mixed po pulation biofilm biom ass (C rit c hl ey et al. 20 01 a) . w hi.ch decreased cuproBi carb o n a te d os in g is solve ncy (3% of isolates) . commonly used for the Single isolate biofi lms of trea tment of corroding a species of X anthomonas Cl 6 copper plumbing systems and Rhodococws decreased ..§.. and m ay have immediate the rate of cuprosolvency ~ 5 advantages if corrosio n is in comparison co sterile ~C: 4 suspected to be rnicrobially control copper coupons influ enced, by potentially (Critchley et al. 20016). ~ 3 0 redu cing the amount of Biofilms of Rhodococcus sp. <.> biofilm present. Increasing h ave prev io usly bee n 2 C. wa ter pH usually decreases C. s h o w n to indu ce 0 (.) cuprosolvency but can also phosphating on unalloyed decrease the effi ciency of pro v idin g s t ee l , 0 disinfec ti o n. Increasing protection from cotTosion E A B F Cont rol D G C Biofilm sa m pie wa ter pH was show n to (V olkland et al. 2000) . ch an ge th e n ature of Th ese bac teria m ay have Figure 1. Cuprosolvency activity of mixed popu lation biofilms extracted biofilm , r es ultin g in future potential in the from copper plumbing tubes conveying filtered drinking water bi o fi l m w ith a hi g h bior e m e di a ti o n of polysaccharide to cellular corroding systems. content (Critchley et al. Th e m ec h a ni s m s C) E 6 2000). This m ay h ave throu gh w hich biofilms C impli c at i on s for th e can initi ate corro sive .g 5 chlorine resistance of the reactions are varied, and ~ biofilm. Increasing turbuit is unlikely that o ne C: 4 lence within the plumbing m ec h ani sm is so l ely g 3 tubes, shown as Reynolds responsible (Figure 3). 0 (J numbers, also decreased Biofi lm s ca n direc tl y ... 2 Cl) the amount of vi a bl e indu ce c9rrosion through a. a. bac teria detected within 0 microbial m etabolism or (.) biofilms (Critchley et al. influ e n ce corro sion 0 2001a). This suggests the through their adherence A B C D E Contro l h y drod y n a mi cs of to surfaces . Acido11orax Bioiilm samp le domestic plum.b ing systems del.afi.eldii has previously Figure 2. Cupro$olvency activity of mixed population biofilms extracted may also influence the been sho w n to increase from copper plumbing tubes conveying unfiltered drinking water amount ofbiofilm present. copp e r corro si o n b y

7-...--------------------------,

Cl)

7-r--------------------~----

Cl)

WATER SEPTEMBER 2001

WATER

BIOFILM CUPROSOLVENCY MECHANISMS

.Jf

CORROSIVE tv1 IC RO BIA L METABOLITES

BINDING OF COPPER IONS TO EXOPOLYSACCHARIDES

--- """,._____

ELECTROCHEMICAL â&#x20AC;˘ ,,.,,, ... IN TE RAC TION S

ENHANCED SOLUBILITY OF CORROSION PRODUCTS Figure 3. Cuprosolvency mechanisms by biofilm s

produ ci n g acids p ro m o ting co pp er ionisatio n (Davidson et al. 1996). T his is pt'obably du e to the produ ction of weak organic acids, such as ace tic or citric acid at the biofilm-metal interface . T his rarely res ul ts in a bulk wa ter pH change, m ore so in Adelaide drinking wa ter du e to its high buffe ri ng capacity. T he hardness of Adelaide water also favo urs the fo rmation of a pro tective scale on copper tube, providing som e protection from corrosive elem ents in water. H owever, biofil m fo rmatio n ca n change the nature of protective copper oxide films which fo rm on exposure to drinking wate r. T he resulti ng film is m ore poro us and less adh erent, w ith microorganisms directly inco rpora ted in to the oxide m atrix (Webster et al. 2000) . Biofilms have also been shown to bind m etal ions wi thin extracellular polymeric substances (Geesey et al. 1996) . T he abili ty of biofilms to acc um ulate copper io ns may additionally increase aqu eo us copper con centra tio ns through biofilm detachment into drinking wa ter.

Influence of Water Quality on Cuprosolvency by Biofilms T he role of wate r quality in the microbial corrosion of copper is complex and poorly understood. Inte rac tio ns b e t wee n bi ofi lm s a nd th e wa t er enviro nment m ay elu cidate w h ether MI C is favo ured by certain conditions. Pure cultures of h etero trophic bac teria isolated from the copper plumbing tubes we re used in labo ratory co upon experiments to inves tigate their cup rosolvency ac tivity in a range of differing wa ter compositions. The wa ter samples used in the experiments represented a range of drinking wa ter supplies from around Australia and varied in hardness, alkalinity, ionic and nutrient co ncentrations. Cuprosolvency activity in the presence of biofilms was indexed against cuprosolve ncy by the sterile control co upons in the experimental w aters, to adju st fo r corrosio n no t m ediated by biofilms. Corrosion ac tivity by these biofilms

generally increased with increasing biofilm bio mass (Critchley et al. 2001c) . In m ost insta nces, bio mass was directly rela ted to the concentratio n of dissolved nu trients in th e experimental wate r, includi ng soluble phosphate and organic carbon (Cri tchley et al. 2001c) . Dissolved organic ca rbo n comp ounds, w hich are high in Adelaide drinking water, have previously been shown to decrease copper corrosion (Broo et al. 1999) . H owever, organics in wa ter can also provide a nutrient so urce for microorganisms, prom oting regrowth and biofilm form ation (Ellis et al. 2000). T he role of nutri ents in biofilm accumu lati on is unclear, and is likely to be both nutrient and biofilm community specific. Increases in water conductivity also increased corrosion ac tivity. Increasing ionic concentrations increased the polysaccharide to protein ratio present w ithin the biofilms (Critchley et al. 2001 c). In this instance, polysaccharides may be produced b y bi ofi lm b acteria as a sur v ival m echanism in response to the increasing ionic strengths. T h e involve m ent of polysaccharides in initiating copper dissolution has previously been established and m ay explain the results observed here. In conclusion, the results suggest that the chemical comp ositio n of wa ter may potentially influence both the acc umulation of biofilms as well as their potential to ca use corrosion.

Conclusion Understanding the causes of changes in copper concentrations in drinking water is the first stage in a risk assessm ent m odel. Th e ability of biofilms to acc umulate and potentially increase cuprosolvency presents serio us implications for both publi c h ea lth a nd di s tributi o n sys t e m manage m ent. Our research continues to inves tiga te b o th of these imp o rtant aspects.

Acknowledgements T his research was supported by the International Copp er Association . T he authors wish to thank Russell Taylor and R oger O' H a ll o r a n , CS IR O M anufacturing Science & Technology, for editorial comments.

The Authors T his research is being pe1formed in the D epartment of Environmental H ealth, Flinders University . Michelle Critchley is a postgraduate student, Dr Nancy Cromar is Senior Lecturer and Ass Prof Howard Fallowfield is head of the D epartme nt. Dr Nick McClure is senior lec tur e r in B iotec hn ology in th e D ep artme nt o f B iologica l Scien ces, Flind e r s Un ivers it y. Co nt ac t : howard. fallowfi eld@flinders.edu. au

References Arens P , Tschewitski G J , Wollman M , Foll ner H and Jacobi H (1995) Indicato rs fo r microbiologically induced con osion of copper pipes in a cold water plumbing system . Zen.tralblatt ju ,Âˇ H ygiene und Umwelt111edizin 196 (5), 444454. Brno A E , Berghult B and H edb erg T (1999) D rinking water distribution - th e effect of natu ral orga nic ma tter (NOM) on the co rrosio n of iron and copper. Water Scieuce and Technology 40 (9) , 17-24. Cox S and Dillon B (1980) C uprosolve ncy in Adelaide water. Co rrosion Australasia, 4-7 . C ritchley MM, Cro mar N J , McClure N and Fallowfield HJ (2000) The influence of water chem.istry on biofilm accumulation and

Generalised corrosion of copper tubes extracted from filtered drinking water plumbing systems.

WATER SEPTEMBER 2001

WATER

Biofilm accumulation in a hard drawn copper plumbing tube

WATER TREATMENT SPECIALISTS EPG has one of Australia 's broadest ranges of brands, products and services, to provide the water industy with systems and solutions .

I~ I I 1,,7r,-y~c,,,

',, CORP.

•Non Metallic Sludge Scraper Systems

~ --

"l!t,._,,

;

~,......;._.

..-

·•c-·

:;.;.;

~-.

GORMAN-RUPP •Self Priming Water and Wastewater Pumps

TSURUMI •Submersible Grinder Cutter and Solids Handling Pumps Also available Harben Sewer Jetting and Aquaturbo Surface Aerators

For further information call 1800 351 755

Engineered Products Group Pty Ltd Email:gateway@epg.com.au Website : www.epg .com.au

WATER SEPTEMBER 2 001

cuprosolvency activity in drinking water. Corrosion and Prevention 2000, 40th A 11uual Conference of the A ustralasia11 Corrosio11 Association. Auckland , N ew Z ealand , N ovember 19-22. Critchley M M, Cromar NJ , M cClure N and Fallowfi eld HJ (2001a) Biofilms and rn.icrobiall y influenced cuprosolvency in domestic copper plumbi ng systems. journal of Applied Nlicrobiology (In Press) Critchley M M , Cromar NJ, M cClure N and Fallowfi eld HJ (20016) The effect of distribution system biofilm bacteria on cuprosolve ncy in Adelaide drinkin g water. Water Science n11d Technology: Water S1-1pply. 1(4): (In Press) Critchley M M , Cromar NJ , M cClure N and Fallowfield HJ (2001c) T he signifi cance of water chemistty on cuprosolven cy activity by drinking water biofilms. Water R esearch (In Preparation) Davidson D , Beheshti B and Mittelman M (1996) Effects of A rthrobacter sp. , Acidovorax delafieldii and Bacillus 111egateri11111 o n copper solvency in a laboratory reactor. Biofouli11g 9(4), 279-284. Domek M J, LeChevallier MW, Cameron S C and M cFeters GA. (1984) Evidence for the role of copper in the inju ry process of coliform bactetia in drinking water. Applied and Environm ental Microbiology 48(2), 289293 . Dutkiewicz C and Fallowfi eld H (1998) Assessment of mi crobi al involvement in the elevation of copper levels in drinking wa ter. ]011r11al of A pplied Microbiology 85 (3), 597-560. Ellis B D , Butterfield P, Jones W L, M cFeters GA and Camper AK (2000) Effects of carbon source, carbon concentration and chlorination on growth related parameters of heterotrophi c biofilm bacteria. Microbial Ecology 38, 330-347. Garduno RA , Hall Band R obinson MG (1993) Bacterial consorcs whi ch enhance th e copper tolerance of Amphora co.ffeaeformis. Biefouli11g 7, 285297 . Geesey G G, Mittelman MW, Iwa oka T and Griffiths PR (1986) Role of bacterial exopolymers in the deterioration of metallic copper surfaces. Materials Pe,formance 25 , 37-40. Gilotra U and Stivastava S (1997) Plasmid-encoded sequestration of copper by Pseudomonas pickettii Strain US321. Current Microbiology 34, 378-381. H amilton W A (1 995). Biofilms and Microbially Influenced Corrosion. In Microbial Bio.films. H. M. Lappin-S cott and J. W. Costerton (Eds) . Cambridge University Press, Great Britain. Harwood VJ and Gordon AS (1994) R egulation of extracellular copperbinding proteins in copper-resistant and copper-sensitive mutants of Vibrio alginolytiws. Applied a11d E11 viro11111ental Microbiology 60(6), 1749- 1753. LeChevalli er MW, Cawthon C D and Lee R G (1988) Factors promoting survival of bacteria in chlorinated wa ter supplies . A pplied an.d Environmental Microbiology 54 (3), 649-654. Nation al H ea lth and M edical R esearch Co un cil (1996) Guidelines for chinking water quality in Australia. Australian Water R esources Council, Canberra. N icholas D and Carlyle G .(1999) M anagement of "blu e water" issues the Evans Head experi ence. Corrosion a11d Prevention 1999, 39th Annual Conference ef the A ustralasian Corrosion Association.. Sydney, Australia, November 21-24. Niqu ette P, Servais P and Savoir R (2000) Impacts of tube materi als on densities of fixed bacterial biomass in a drinking water distributi on system. Wa ter R esearch 34 (6), 1952- 1956. Pizarro F, Olivares M , Uauy R , Contreras P, R ebelo A and Gidi V (1999). Acute gastrointestinal effects of graded levels of copper in drinking water. Environmental Health Perspectives 107 (2), 11 7- 121. Rusin P A, Rose J B , H aas C N and Gerba C P (1997) Risk assessment of opportunistic bacteri al pathoge ns in drinking water. R e11iews of Environmental Conta111ination a11d Toxicology 152 , 57-83. Taylor R . (2001) Manufacturing Science and T echnology, CSIRO . Personal communication. Volkland HP, H arms H , Mii ller B, R epphun G, Wann er O and Z ehnder A J B (2000) Bacterial phosphating of mild (un alloyed) steel. Applied and Environmental Microbiology 66 (10), 4389-4395. Walker J T and Keevil C W (1990) The influence of water chemistry and environmental conditions on the microbial colonisation of copper tubing leading to pitting corrosion espec ially in institu'tion al buildings. foternational Copper Association Project No . 40 7. W ebster B J , Werner S E, W ells D B and Bremer P J (2000) Microbiologically influ enced corrosion of copper in potable water systems - pH effe cts. Corrosion 56(9), 942-950. Zacarias I, Yanez C G, Araya M , O raka C , Olivares M and Uauy R (2001 ). D etermination of the taste threshold of copper in water. C hemi cal Sense 26, 85-89 .

WASTEWATER

THE USE OF RECLAIMED WATER FOR IRRIGATION: SOME ISSUES P Thomas and R Croome Abstract

Introduction

R ecla imed water irrigation is an environmentally accepted alternative to wa ter-based disposal of treated efiluent. However, a number of issues must be dealt with successfully to achieve the full benefits of using reclaimed water in this way . This paper summarises the basic processes associated with the use of reclaimed wa ter in Victoria, and explores in particular som e of the economic and legal issues imp acting on the ready ado pti on of small reclaimed wa ter irrigation sch emes.

Although Australia 's scarce surface water resources are valued for purposes of wa ter supply, recreation and the maintenance of aquatic life , many still receive treated dom es ti c reclaimed wa t er discharges. T here is now a growing appreciatio n, however, that such discharges are a valuable resource available for re use, rather than a waste requiring disposal. Federal and State Authorities have both recognised the imp ortance of minimising nutrient inputs to surface waters as well as the beneficial use of nutrients (nitrogen

PACIFIC

and phosphorus), and reclaimed wa ter irrigation in particular is identified as one of the alternatives to water based disposal. The practice of reclaimed water irrigation is not n ew (M elb o urne's Western Treatment Plant h as b ee n treating wastewater using landbased treatment systems since the 1890's), and a scientifically rigorous approach has been adopted in Australia for many years eg. a study at Wodonga, Victoria in the early 1980's evaluated issues such as tree growth and nutrient acc umulatio n, and changes in soil chemical properties associated with a landbased discharge (Stewart et al, 1988). But

DYNAMICS

10 years experience in the Water Industry Exclusive suppliers of:

S e ~

TECH NO LOG

Ill

GUTERMANN

PARTICLE COUNTING SYSTEMS

WATER NETWORK CONTROL SYSTEMS

LEAK AND PIPE LOCATION SYSTEMS

We install, train the operators and support all systems. For further information, contact: PACIFIC DYNAMICS PTY LTD,

PO Box 24, Artarmon, NSW 1570 ACN 063 234 373

(02) 9460 9099 FAX (02) 9460 9019 EMAIL: pacificdynamics@compuserve.com WE B PAGE: www.pacificdynamics.com.a_ u PHONE:

Phone: (021 9894 4511 Fax: (02) 9894 4522 Email: sales@aqualab.com.au Internet: www.aqualab.com.au

Leaders in Water Quality Monitoring! WATER SEPTEMBER 2001

WASTEWATER

while many treated wastewa ter irrigation schem es are in place across the country (Thomas and Sathiamoorthy, 1999; Donlon and Surapaneni, 2000; Gardner and Gibso n, 2000) , only some 4.6% (16 880 ML) of the estimated 367 000 ML of effiu ent produced in Victoria in 1999 was re-used, the balance being discharged to either rivers or oceans (Dillon, 2000). A marked increase in reuse is expected, however, as W ater Authoriti es and C iry Councils are urged to move to landbased beneficial uses to protect surface water resources from the effects of treated eilluent discharges in particular.

Legal aspects and guidelines In Victoria , the current Acts, policies and re gulations releva nt to reclaimed wa ter irrigation are administered by th e Environment Protection Authority and other Go ve rnment age ncies eg. Department of Human Services, and Department of Natural R esources and Environment. Acts particularl y relevant to reclaimed water irriga tion are the E1wiromue11t Protection Act 1970, H ealth Act 1958 and Livestock Disease Control Act 1994. Pollution of gro undwater and / or surface water, changes in soil characteristics, and risks to public and/ o r sto ck health may occur from poorly designed and managed reclaimed wa ter irriga tion schemes. A number of guidelines are available to minimise these threats and in Vi ctoria at present, the Water Authorities and other agencies fo llow the local gu idelines for was tewater reuse (EPA Victoria, 1996), was tewater irriga tion (EPA Victoria , 1991) and the use of reclaimed wa ter (EPA Victoria, 2001). The latter, w hich is

delft software australia exclusive Australian agents for WL delft hydraulics of the Netherlands providing World Class Software for:

a revised version of the wastewater reuse guidelines (EPA Vi ctoria, 1996) is at the date of writing still not available to the public. These guidelines together w ith the guidelines from other States, and the Au stralian National G uid elines (ANZECC 2000) , are more rigorous than those follow ed in many other countries . Failure to comply w ith th e guidelines may result in severe penalties under the Environment Protection Act 19 70 for the misuse of was tewater. The pro cess required to establish and manage a reclaimed water irrigation sch eme involves public cons ultation, fo llowed .by a legal co ntract between the reclaimed water provider and the purchaser. There have been extended delays in establishing reclaimed water irriga tion schem es in many areas du e to legal issues, public obj ection or additional costs associated with the requirements of environme ntal manage ment plans. The capital and ongoing monitoring costs can also be high depending on the nature of the irrigation scheme and the potential risks attached to it. Other factors, which ma y ca use delays or inhibit reclain1ed wa ter schem es are health and planning issues, enviro nmental issues and h eritage issues, and the availability of other eco nomical water so urces eg. river diversion o r gro undwater.

Community opinion and consultation Communiry acceptance is a key factor in the success of any reclaimed water irrigation proposal. Important characteristics here are the difference in cost of potable water and reclaim ed water, the extent of publi c conta ct w ith the reclaimed water, and the qualiry of treatment given to the wastewater. In Victoria the cost of potable water varies between 0.35 - 0 .90 cents per kilolitre while reclaimed water is offered almost fr ee by so m e authoriti es, and the major objections to reclaimed water irri gatio n ha ve come mainly from residents of rural co mmunities due to perceived aesthetic (odour), environmental (saliniry) or health effects.

Risks associated with reclaimed water irrigation

Hydrodynamics of oceans, bays , rivers , estuaries and reservoirs. Coastal and marine environmental impacts. Sediment transport (cohesive and non-cohesive sediment) and erosion investigations. Water quality and transport dispersion investigations. Stormwater and effluent outfall , cooling water and recirculation studies. Full three dimensional flow, quality, stratification , mixing and dispersion modelling . Wastewater system modelling and control. Stormwater trunk drainage system design. GIS interfacing. Dam failure and tsunami investigations. Port/marina investigations. Wind/wave short and longwave modelling. Oil spill modelling . We offer sales and full technical support on a range of software products to cover all the above fields including : DELFT 3D Âą Full 2D & 3D hydrodynamic and coastal/ocean processes . DELFT FLS Âą Full 2D Floodplain modelling. SOB EK Âą Integrated 1D & 2D simulation of rivers , canals , pipes, sewers & estuaries. Web Address : www.delftsoftware.com.au E-mail : info @delftsoftware.com.au

Brisbane:

Pre 07 3855 3809 Fax: 07 3855 1904 Melbourne: Ph : 03 9558 9231 Fax: 03 9558 9956

Sydney: Ph: 02 9983 1267 Fax : 02 9983 1055

WATER SEPTEMBER 2001

It is essential that the suppliers and users of reclaimed wate r irrigation schemes are aware of the potential risks associated with the activiry, and that precautions are taken to minimise or eliminate them . The major risk facto rs are generally identified as environmental, human and stock health, produce safery, and legal issues (EPA Victoria , 200 1). The relative degree of th ese risks will depend on the nature of the irrigation scheme, whether it is urban or rural, its size, the degree of treatment give n to the reclaimed water, and the efficacy of the overall environmental management plan. Reclaimed water irrigation schemes that are appropriately designed and managed w ill prevent/ minimise water, soi l and air contamination, the risks to public and agric ultural animal health , unacceptable levels of microbial or chemical contamination of food produ ce, and legal risks associated with the use of reclaimed water.

Roles and responsibilities of suppliers and users The National Guidelines for Sewerage Sys tems - U se of R e claim e d Wat e r (ANZECC , 2000) and th e Draft Environmental Guidelines for the U se of Redaimed W ater (EPA Victoria , 2001) specify the roles and responsibilities of the suppliers and users of reclaimed water. If there is mo re than one party involved in the reuse scheme the roles and responsibilities are generally subj ect to a collective agreem ent betv.,reen the parties. Some of the important issues considered 111 an agreement between a supplier and user include:

WASTEWATER

• definition of rol es and responsibilities; • contract duration - term, conditions for termination; • cost of reclaimed water; • ownership of facilities;

Table 1. Classes of recla imed water use, showing quality and monitoring requirements (Source: EPA Victoria, 2001) Class Range of uses

• reclaimed wa ter characteristics; • commen cem ent of use; • responsibility for operation, mainten a n ce, monitorin g a nd a uditin g processes; • nature of the reclaimed wate r use;

Urban (non-potable , no restri ctions on public access), Agricu ltural (raw human food crops exposed to rec laimed water), Industrial (potential for worker exposure), All Class B, C and D uses.

• reliability of supply; • environmental improve m ent plan; and

• liabilities.

Treated wastewater quality for irrigation

Water quality

Monitoring requirements

pH 6.0 - 9.0 BOD < 10 mg/ l SS < 10 mg/ l Turbidity < 2 NTU E. Coli < 10 org/100ml Viruses < 1 org/25 l Parasites < 1 viable helminth Egg/1 0 l Cl 2 residua l > lmg/ l Nutrient, salinity and toxicant controls

pH - week ly BOD - week ly SS - weekly Turbidity - continuously E.Coli - week ly/monthly Viruses - twice yearly Parasites - twice yearly Cl 2 residual - continuous ly

Agricu ltural (human food crops not exposed to reclaimed water, dairy shed washdown, stock drinking except pigs), Industrial (saleyard washdown) , All Class C and D uses .

pH 6.0 - 9.0 pH - weekly BOD - weekly BOD < 20 mg/ l SS < 30 mg/l SS - weekly E.Coli < 100 org/100 ml E. Coli - week ly Cl 2 residua l - daily Cl 2 residual > 1 mg/l Nutrient, salinity and Nutrient, salinity and toxicant controls toxicant - regu larly (weekly)

Urban (non-potable, contro lled public access), Agricu ltural (crops not exposed to rec laimed water, processed human food crops, grazing/

pH 6.0 - 9.0 pH - month ly BOD < 20 mg/ l BOD - month ly SS < 30 mg/l SS - monthly E.Coli < 1000 org/100 ml E.Coli - week ly Nutrient, salinity and Cl2 residua l - dai ly

fodder for catt!e, sheep and

toxicant contro!s

In Victoria , untreated and primary treated reclaimed waters are not generally r eco mm e nd e d for irri gat ion , the minimum treatm ent required being secon dary tr ea tm e nt. Additional treatment processes beyond secondary treatm ent level are required w here the risk to human or lives to ck exposure is high. T he four classes of reclaimed wa ter specified in the latest draft environmental guidelines for the use of reclaimed wa ter (EPA Victoria, 200 1) are summarised in Table 1. The monitoring requirements listed in Table 1 are for a scheme using more than 1 ML of reclaimed wa ter p er da y o n any day.

Environmental Improvement Plan

Class A quality reclaimed wate r is normally recommended for urban usage such as ga rden wa tering, toilet flu shing, irrigation of open spaces, parks and gardens, sports gro unds and golf courses; for agricultural use where the wa ter could com e in contact wi th food crops; and for industrial use w ith potential for human exposure. Class B qu ality water is advised for use in pasture irrigation or som e agricultural uses, w hereas C lass C quality wa ter is for som e municipal uses w ith restricted public access, and pasture irrigation schem es with no direct contact wi th milking animals or pigs . Class D quali ty reclaimed water is restricted to horticulture, woodlots or som e agricultural use. The degree of treatment required of the reclaimed water w ill depend on the locality of the irrigated site: a protected site ,¥.ith restricted access and controls may be managed \¥.ith lower levels of treatment than a site exposed to the public or lives tock.

An Environmental Improvement Plan (E IP) needs to be prepared and included w ithin any reclaimed wa ter re use agreement. The plan should cover all aspects of the schem e which pose a risk to the enviro11111ent, human and livestock health, and provide a framework to assess th e long-term sustainability of the scheme. T he E IP should address issues such as reclaimed water quality and qu antity, winter storage, site controls, buffer distances, warning signs , applica tion rates and timing, irriga tion m e thod s, g round wa t er qualit y monitoring, salinity controls and soil testing. A good E IP wo uld also address water and nutrient balances, withholding periods, staff training and perhaps a risk analysis, independent auditing and the need for an annual report. Although the monitoring requirem ents for the quality of reclaimed wa ter, groundwater and soil are usually based on the Guidelines for Wastewa ter Reuse (EPA Victoria , 1996) and/ o'r Environmental guidelines for the

horses), Industria l (no potential for worker exposure) , All Class D uses. D

Nutrient, salinity and toxicant - weekly

Agricu ltural (non-human food crops including instant turf, woodlots and flowers)

Nutrient, sa!inity and toxicant - monthly

pH 6.0 - 9.0 BOD < 20 mg/ l SS < 30 mg/ l E.Coli < 10 000 org/ 100 ml

pH - month ly BOD - month ly SS - month ly E. Coli - weekly

U se ofReclaimed Water (EPA Victoria, 200 1) there are instan ces w here slight modifica tions have been made with the approval of the EPA to suit local site specific conditions. M any urban users of reclaimed wa ter view th e regular monitoring costs as a sizea ble and ongoing financial burden. Although the supplier is normally responsible for monitoring the quality of reclaimed w ater at the point of delivery, EIPs require many urban users to independently monitor the reclaimed wa ter quality to avoid legal risks. This can add around $15,000 per annum to the cost of other wa ter quality monitoring and soil testing required by the user. For an average irrigation site with concern for impa cts on gro undwa ter quality may have about six groundwa ter bores, w hich will cost around $3,00Q in the first yea r for monitoring every 4 months, and $2,400 p er annum th e rea ft e r for monitoring eve ry 6 months (E PA Victoria , 2001). The cost of testing soil for degradation or contamination will be around $2,000 every 2-3 yea rs. In WATER SEPTEMBER 2001

WASTEWATER

many sit uations there is no direc t economic incentive for either the supplier or the user to adopt reclaimed wa ter irriga tion, but th ey may be compelJ ed to do so because of the lack of any other environmentalJy accep table solution. It is clearly evident that establishment and operation of a small wastewater irri ga tion sch eme in an urban and sensitive area must pay heed to the exp en se of th e regular monitorin g requirements .

Some areas of concern The guidelines related to the use of reclaim ed water are stringent and commit the supplier and user of reclaimed wa ter to a number of requirements . For m any small intending suppliers and users, satisfying these requirements w ill m ean the commitment of substantial additional funding towards both new infrastructure and yea rly monitoring costs. T h ese costs are far in excess of that required for systems using potable wa ter supply. Although the major obj ective of the guidelines is to protect public health and

environment, they may becom e the primary hurdle to the adoption of a new concept in reclaimed water management. M any intended suppliers and users are co n ce rn ed about lega l actions by regulatory authorities, or by third parties as a result of sickness or injury. T he legal responsibility for contamination currently lies with the user (when the latter is not also the supplier). If this responsibility was shared w ith the supplier, even though there would have to be an extra charge there would be m ore potential users of reclaimed water, less time sp ent in the preparation of agreem ents, and possibly a reduction in monitoring costs. Users could be responsible for th eir activities, w ith th e suppliers being required to ensure that the reclai m ed wa ter delivered was of a required, agreed quality and quantity. It is furth er anticipated that revising the existing guidelines from knowledge gained after a period w ill alleviate some of the issues related to the cost of establishing and managing a reclaimed wa ter irrigation scheme .

FILTRATION MEDII

SAND,GRAVEL,COAL,GARNET MANGANESEGREENSAND Invest in accurately graded, durable media from your complete filter media professionals.

U'eau~tl>o//e't,~: • Decades of mineral processing experience. • • • • • •

Media produced to the AWNA 8100-89 Standard. An extensive product range at competitive prices. Manufacturing In accordance with Quality System AS 3902/ISO 9002 Packaging alternatives to suit any requirement. Proficient technical assistance and support. Prompt delivery Australia-wide and Overseas.

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

---~-RIVER SANosr-;g

683 Beenlelgh-Redland Bay Road, Carbrook Qld 4130 INTERNATIONAL: +6173287 6444 FREE CALL: 008 077 744

WATER SEPTEMBER 2001

FAX: (07) 3287 6445

Conclusions 1. Reclaimed water is a viable alternati ve

source of water for irrigation w here th ere is shortage of wa ter from other sources or restrictions on water based disposal. 2 . Reclaimed wa ter is often offered cheaply w hen compared to the supply of public wate r, bu t capital and monitoring cos t' can m ake sm all reclaimed water irrigation schem es relatively expensive operations. 3. Guidelines and regulatory frameworks wo uld need to be modified in order to minimise the capital and operating costs 4. Legal risk is considered to be a maj or issue by many organisations embarking on a program of reclaimed wa ter irrigation.

References ANZECC (2000) Guid elin es for Sewerage Systems - Use of reclaimed water. Australian and New Zea land Environm e nt and Conservation Council, Australia. Dillon P. (2000) Water reuse in Australia: current status, proj ections and research. Water recycling Australia, Proceedi11gs of the First Symposium, 19-20 October, 99- 104. Donlon P. and Surapaneni A. (2000) Ma11aging effluent irrigation.: Th e sodi.fication threat. Water, Vol. 27, No.5, 45-48. EPA Victoria (1991) C11idelines for Wastewater Irrigat ion. EPA Pub lica ti on No. 168, M elbourne, Australia. EPA Victoria (1996) Guidelines for Wastewater Reuse, Best Practice Environmental Management series. EPA Publication No. 464, Melbourne, Australia EPA Victoria (200 1) Draft Environm en tal Guidelines for the Use of R eclaimed Water, Best Practice Manageme,,t series, M elbourn e, Australia . Gardner T. and Gibson H. (2000) Effl uent reuse by crops in Queensland. Water, Vol.27 , No.6, p41. Stewart H. T.L., Hopkins P. , Flinn D.W. , Hillman T.J. , and Collopy J. (1988) Evaluation of irrigated tree crops for land disposal of municipal effluent at Wodo 11ga, T echni cal Report No. 7, A l bury-Wodonga Develop ment Corpora ti on, Australia. T homas P.R. and Sathiamoorthy A. (1999) Effect of long-term wastewater irriga tion on. soil characteristics. Proc eedings of Interna ti onal Conference on Urban Pollution Co ntrol Technology, 13-15 O ctober, H ong Kong, 603-608.

The Authors Percival Thomas (Civil Engineer : Senior Lecturer) and Ro'ger Croome (Aquatic Biologist : H ead of Department) are both w ith th e D epartm ent of Environmental M anagement and Ecology at the Albury/ Wodonga Campus of La Trob e University . Email: pthomas@ aw.latrobe.edu.au

WASTEWATER

NON-POTABLE REUSE OF TREATED EFFLUENT: MANAGEMENT OF HEALTH RISKS Report by N Cromar and D Cunliffe Durin g t h e Water R ecycl i n g Conference in Adelaide, October, 2000 (reported in "Water" January and April 2001) a half-day workshop was run. It commenced with a largely theoretical sessio n b y Dr Cro m a r (Flind e r s University) on using the 'risk' framework to assess health risks associated w ith 11011pota ble recycled water. A seco nd presentation by Dr C unliffe (SA D ept of H uman Services) focussed on the nature of the microbiological and chemical risks. These were fo llowed by case studies exploring the similarities and differences

in two reuse schemes of vas tly different scale and operation. The first feat ured the Virginia Pipeline Sch em e, supplying reclaimed water from Bolivar W astewater Treatment Plant through a dedicated pipeli ne for unrestricted irrigation of all types of food crops in the Virginia Triangle (SA). This case study was co ntras ted w ith th e seco nd w hich considered a prestigious metropolitan golf course in Adelaide, w hich was supplem enting a gro undwater supply for spray irrigation of the course w ith reclaimed water from a Wastewater Treatment Plant managed by the local water authority. The

TUBE SETTLERS

+ + + + +

increase existing plant capacity improve settled water quality improve clarifier utilisation

reclaim ed water was treated to a lesser quali ty and greater restrictions were placed on its usage and subsequent potential for exposure. Both schemes were chosen because they included some nonstandard features that needed assessment . Finally an 'open fomm' was conducted w ith questions and co mments from the audience to the workshop coordinators. T he written record w hich was made of this session fo llows.

Q : Have concerns been expressed by residents n ear golf courses?

111111 ~~~::,:NICS P~!u~!~ An Australian company with the technology and expertise to extend your clarification plant capacity and performance limitations.

increase downstream filter run times

TELEPHONE:

(02) 9757 1922

reduce sludge draw-off and washwater volumes

FACSIMILE:

(02) 9757 1238

E-MAIL:

aquapac@msn.com.au WATER SEPTEMBER 2 001

WASTEWATER

A: W e have no record of any concerns being expressed. Care has been taken to minimise risks to nearby residents, mu ch of w hich was outlined in the second case stud y.

Q: Are there time lim.its / exemptions set on irrigation (in SA)? such as the restrictions in Victoria relating to timing of irri gation and only irrigating dry grass A: The principal control applied is a restriction to night time watering. In SA's climate the surface of grassed areas tends to dry qui ckly

Q: Can we minimi se ex t e n si ve monitoring if sa feguards are in pla ce for schem es such as the Virginia pipeline' A: Monitoring at high level is usually essential to m.in.imise risks to public health, parti cularly in the ea rly stages of a scheme development. Even if sa feguards are in place, it is important to continu e an extensive monitoring program to assure users and the general public that the scheme is 'sa fe' .

Q: Is there a limitation on disposal of grass clippings from greens irrigated with reclaimed water ? A: Not in SA.

Q: What is the nature of public consultation before using reclaimed water for irrigation?

A: For golf courses - consultation is usually through local newspapers. The Virginia Triangle irrigation scheme was ve1y widely publicised before introdu ction . Th.is included a formal process associated with the Bolivar WWTP upgrade required through a legislated environment improve m ent program. There were also education/ information sessions targeted at the growers and buyers of produ ce . Q: Are there w ritten agreem ents with users/ controllers of such schemes? A: Y es . Th ere are two regulato ry approvals from Department of Human Services (DHS) and a license from the EPA. DHS approvals have been issued to SA Water for the supply of reclaimed water and to W ater Reticulation Sys tems (Virginia) for the use of the w ater. In addition all individual users have written contracts with WRS(Virginia). Comment: In WA all licensed schem es require monitoring at point of delive1y (sprinkler) rather than at the distribution so urce as in SA. 58

WATER SEPTEMBER 2001

Response: In SA, the decision to monitor at the point of ent1y to the irrigation system is based on the logic that regrowth of enteric pathogens is ve1y unlikely to occur. International gui delines su ch as th e USEPAs also support monitoring at this point.

Q: If reclaimed wa ter is stored in a dam , does the birdlife w hich may be present pose a potential hazard? A: W e see litcle difference betwee n storing rive r wa ter or reclaimed wa ter in dams. Birds cany a mu ch narro wer range of human enteric pathogens than do humans. For example viruses and to a lesser extent protozoa tend to ha ve relatively specific host ranges, so bird faecal matter is unlikely to pose a significant risk to human health through irrigation wa ter. Of course, the dam must be protected from human and li ves tock was te.

Q: Have there been studies on the quality of plants (vegetables) at Virginia, in terms of nutrient and pathogen co ntent etc? A: Studies are underway at present by gro ups such as Primary Industries and R esources (So uth Australia). In initial studies of microbiological quality numbers of E. coli were less on pl ants grow n using reclaimed water than on comparable plants available at the markets. A large body of evidence has been amassed for the quality of produ ce internationally from systems such as the Monterey schem e.

Q: Was subsurface irrigation considered for the golf course gree ns and other schemes? A: Subsu1face irrigation is very expensive and research to date appea rs to show that there is minimal requirem ent for it, based on health risks to the public .

Q: How do you explain 'risks' to the public? Comment: It is important to give the public a feeling of control over risks Response: Explanation of relative risk to the public is not always easy . Th e key to success may be attempting to understand the public's perception of risks Comment: There is a possibility that we may lose the confidence of the public if we erect signs saying 'water not drinkable' Response: Signage is an issue at Virginia, but there is a legal requirem ent that it be present. The accep tance of signage by

users at Virginia has been something of a problem that may need to be monitored. Information p amphl e ts ha ve b ee n produ ced detailing the quality of the reclaimed wa ter, which in is sup erior in (mi crobial) quality to surface wa ters used for irri gation waters. It is important to rem ember that river water (no matter how polluted) is co mmonly used for irrigatio n and usual] without testin g. Comment: There has been discussion that th e Murr ay D a rlin g B as in Commission should install signs on the Murray ri ve r declaring that it is ' not fit to drink' DHS has issued m edia advi ce on three occasions that untrea ted surface wa ter is unfit to drink.

Q: If wa ter quality of reclaimed water is higher than ocher 'naturally produced' so urces, th en w hy allow people to irrigate with ' natural' water?

A: ANZE CC / ARMCANZ National W ater Quality Manage m ent Strategy Guidelines cover this aspect of irrigating with 'natu ral' wa ter. Comment: The issue is again one of risk perception . ImportaRt considerations as to w hether or not people will accept health risks are related to factors such as 'trust', 'knowl edge' and 'control'. Other factors w hich may be important include the tim.ing and duration of the potential effect (being hit by a golf ball is a more immediate danger than getting sick from inhaling aerosols from reclaimed wa ter irri gation); w h e th e r th e hazard is known/ unknown and w hat level of 'dread' is associated w ith the effect e.g. cancer. A grea t deal of work has been done is this area by researchers such as Sand e m an (RISK = HA ZARD + OUTRAGE); Slovic and others (see references). Comment: If vegetables are irrigated with water that does not comply with US Tide 22 standards - ho w might this be viewed internationally? What are the advertising implications (locally, nationally and internationally) .

A: The microbiological and chemical quality criteria are the important issue. W e have retain ed th e same standards even though the Virginia Pipeljne Scheme is using a modified treatment train. On a national scale we have been addressing th.is issue through an education campaign by talking to buye rs, including from the sup ermarket chains . In terms of export products Australian Quarantine Inspection

WASTEWATER

Service has been kept informed and they have indicated that the quality of the w ater used is the important iss ue.

m ust be set using scientifically rigorous m ethods such as risk assessm ent and then used sensibly .

Alternative view: Given the 'quality' of the reclaimed water in comparison to other sources used for irrigation, perhaps growers should consider the potential marketing opportunities in promoting this 'enviro nmentally fri endly, ye t safe' usage .

Comment: Many reuse schemes (eg. H ervey Bay) have been in o peration for a considerable length of time w ith no reported morbidity.

Comment: Informing the publi c is the key . We must tell the public about the quality of reclaimed water in an open yet considered way.

Q: Qu antitative microbiological risk assessm ent is a relatively new science, w ith lots of un know ns. In the event of a legal case, ho w should the risk asse ss or/ legislator respond to su ch criticism?

A: Quantitative risk assessment is best practice and attempts to provide a systematic approach to the assessm ent of risks w hich other single disciplines, such as epidemiology or toxicology in isolation cannot answer adequately . U sing Virginia as an example - the guidelines that have been set are deliberately conservative and monitoring of treatment effective ness is conducted according to best practice protocols.

Q: How can we m easure wa ter quality in lagoons? Are there concerns with using faecal coliforms / E. coli as indicators ? What about risk of helminths? Do other standards need to be develop ed? A: Indicator o rga nisms such as faecal coliforms/ E. coli are only one method used to establish water quality in lagoons. The risk from helminths is relatively low in the southern states but greater consideration may be required in northern/ tropical latitudes. Research has established that a detention period of20-25 days in a lagoon is sufficient to reduce helminth numbers to levels below concern to public health, at least in the so uthern regions of Australia. In filtered water a number of physical and chemical parameters such as turbidi ty and chlorine residual are used as surrogate indica tors of microbial qu ality Comment: Reclaimed water sch emes need to be developed in consultation w ith a number of auth orities including h ealth and environment. Guidelines

Comment: How can we fac to r 'public behaviours' into th e risk equation? There is a security risk of golf co urses being used out of hours w hile actually being irrigated. In addition, m any golfers exhibit strange behavio ur patterns such as licking golfballs. Response: Metropolitan golf courses using reclaimed water (in SA) are fenced to prevent unauthorised access. How can we regulate for human behaviours?

Q: C hildren may also play under sprinklers and 'paddle' in puddled irrigation waters. Is C lass 'A' water suitable for primary contact recreation?

been retained in the current ve rsion of the SA Guidelines . It will possibly be deleted from later versions because it is unlikely that anyone w ill wa nt to use reclaimed wa ter for this purpose du e to public p erception issues . The open forum closed w ith som e summary remarks from the workshop coordinators on the requirem ent for 'risk trade- offs' eg. risks of chemical treatment contamination compared w ith microbiological contamination of reclaimed wa ter, and the importance of informing the public and balancing the maintenance of public health w ith environmental considerations.

References Johnson & Slovic (1998) Lay views on uncertainty in e nvironmenta l h ea lth ri sk assessment. Journal of risk resea rch 1 (4): 26 1-279 Krewski , Slovic, Bartlett, Flynn & M ertz (1995)

A: In the 1996 draft of the N ational Guideli nes C lass A reclaimed wa ter was classed as fit for release into primary contac t impoundments and this has

H ealth risk perception in Ca nada 1: Rating hazards, so urces of information and responsibility for health pro'tecti on. Human & Ecological R isk Assessment 1 (2) : 117-132

_ยงJK l@dtiik

www.seca.com.au

WinCan and SECA are proud to announce our latest breakthrough in technology. It is now possible to record directly to the hard disk of your PC. No more VHS tape/recorders in the CCTV van!!!! WinCan is a modular design that grows with your business Easy to learn and use Start using straight = -= "!!" away with minimal training Operates with Windows me, Windows 98 or NT, 2000 . Can be used with any make of CCTV system. Includes full set of ACCEM, SIRCA, WRC descriptions. Colour coded reports and extensive graphics to aid interpretation . Efficient asset management and data analysis with DDE Link to GIS systems available

Sewer E9uipment Co (Aust) 7-9 Carrington Rd Marrickville, NSW, 2204

With our inspection systems a user can cover a number of diverse areas. The investigation of drain damage is just as possible as the final inspection of newly built laterals . iPEK colour TV systems are used for the investigation of air conditioning systems and cable pits, power station lines and filtering units as well as in the reconstruction of buildings and the monitoring of cavities. Experience the benefits and costeffectiveness of our portable colour TV system . The heart of the system is a high resolution solid state colour-TV-camera, which may be continuously focussed from '2 cm to 1m. It is equipped with an integrated light ring, ensuring absolute colour stability. In combination with an iPEK crawler, you have an unbeatable team.

Ph: 61 2 9559 5622 Fax: 61 2 9559 5076

Email: seca@seca .com.au Web: www.seca .com .au

ENVIRONMENT

Iii

-SEDIMENT CONTAMINANTS IN URBAN STREAMS: THEIR EFFECT ON STREAM HEALTH N O'Connor and S Moore Abstract In order to m eet increasingly stringent guidelines and public expectations for the condition of urban streams, urban waterway managers have been seeking a clearer understanding of the causes of urban stream degradation. A range of N factors such as the degree of urbanization , direct anthropogenic stream alterations (e.g. c hann elisa tion , stream-bed paving) and contamination from point and non- point sources are known to adversely affect streams. However, to date the relative contribution of these fa ctors to a stream 's eco logical h ealth , and in particular to their macroin0 5 10 15 20 verte brate co mmunities has KILOMETRES been unknow n. Map of the study area - numbers indicate sampling In recognizing the importance of understanding the m e ntal and co ntaminant variabl es . contribution of these factors to stream However, the only group of contaminants h ealth, th e National Ri ver H ealth that seemed to have any relationship w ith Program in association w ith M elbourne m ac roinve rtebrate communi ty stru cture Water fund ed the present study. The study was heavy m etals. was designed to explore the relationship In general, the results of the study b etwee n b enthi c ma croinve rtebrate suggest that the site-specific nature of communities (as an indicator of stream sediment contamination m akes it almost h ealth) , and common urban stream polluimpossible to predict sites of contamitants (e .g. hydrocarbons, heavy m etals) by nation with any useful degree of accuracy . sampling 44 sites throughout m etropolitan Waterway managers should foc us on M elb ourne. id entifying and co ntrolling existing The patterns of sediment contaminant contamination so urces as a priority distribution within urban streams were follo wed by rem ediation of historically complex . Howeve r, not surprisingly the contaminated areas. Routine monitoring most important factor determining the of macroinvertebrate conununities should degree of sediment contamination was the be sufficient to indicate many elem ents of proximity to the pollution so urce . overall urban stream health. However, if Significantly, landscape features, such as results indicate possible sediment contamthe stream gradient, patterns of urbanination then confirmation efforts sho uld ization and direct anthropogenic stream focus initially on heavy metals, give n that alterations also appeared to play important this group of contaminants, in contrast to roles in determining the contaminant load organics, were found to be correlated with and stream health . m acroinvertebrate community structure. Analysis of the ma croinve rtebrate Introduction conununity data revealed some broad scale Urban development in catchments has relationships between particular environ-

WATER SEPTEMBER 2001

a m aj or imp act on th e structure of waterways and on wa ter qu ality . Widespread building and paving increases the imperviou sness of the catchment surface, hence the proportion of rainfall that runs off as stormflow, and the speed at w hich that runoff occ urs in creases . Int e n se sh ort-term, or flashy, runoff events have greater erosive powe r and in natural waterco urses thi s ma y ca u se ex t e n si ve bank and b ed erosion. Commonly, urban streams h~ve been subjected to engin eering w orks to allow flood peaks to pass quickly, including channelisa tion and strea m-b ed p avi ng . These changes have a profound effec t on stream ecological co mmunities that must cope sites. w ith more fr equ e nt and intensely scouring high flows, lower base flows since steady groundwater seepage is diminish ed by catchment paving, and decreased habitat availability. In addition to physical changes, urban streams are commonly exposed to a range of contaminants from diffuse and point so urce discharges, such as industrial eilluent and stormwa ter runoff. Urban stormwater can be contaminated with a wide range of material including dust and soil, litter and garden rubbish , animal wastes, fertilisers and pesticides, oil and grease, sewage overflows, and tip and landfill leachate (O 'Loughlin et al., 1992). In addition, contaminants m ay also be washed out of the atmosphere. M aterial that builds up in the catchment over dry periods is generally considered to acco unt for the poor quality of " first flush " storm wa t e r. M a n y o f th e to x i c constituents of stormwater (e .g. h eavy metals, pesticides, solve nts, petroleum produ c t s, and num ero u s organ i c chemicals) are hydrophobic and bind w ith sediment particles . These contaminated particles can concentrate in the deposi-

ENVIRONMENT

tional areas of a stream and can have adve rse environmental effects . At the time of the study no extensive studi es of stream toxicants and their effect on stream health as indicated by macroinvertebrate conununities had previously been undertaken , although there have been m any small-scale site-specific studies including studies conducted in M elbourne (e.g. Bagg et al., 198 1, Pouliot, 1993 , R eed, 1992) and an earlier study on heavy metals condu cted by the Victorian Environment Protection Authority (EPA 198 1). Melbourne Water also conducted to xi cant monitoring as part of its Stream watch program however the range of contaminants was limited, covering only heavy m etals. In recognition of the importance of increasing our understanding of urban streams in order to improve the managem ent of urban steams, the current study was fund ed by Environment Australia's National Rive r H ealth Program and by Melbourne Water.

Measuring macroinvertebrate, sediment contaminants and physico-chemical variables in urban streams T he study involved sampling forty-four sites in autumn and spring across metropolitan Melbourne for macroinvertebrates and a range of physico-chemical parameters. Sites were chosen to give a range of different land uses, while the physico-chemical parameters were chosen to charac terise the habitat of a site and the level of common urban contaminants in the sediment. Contaminants w ithin sediments measured were heavy m etals, polycylic aromatic hydrocarbons, benzene, toluene, ethylbenzene, xylene, phenols, organochlorine p esticides, polychlorinated biphenyls, petroleum hydrocarbons, oil and grease, phthalates and volatile halogenated organics. The toxicants were chosen on the basis of their likely occurrence above detection limits in Melbourne m etropolitan stream s, and their known toxicity. At each site m acroinvertebrates were sampled according to the protocols developed as part of the Monitoring Rive r H ealth Initiative (MRHI, 1994). T his is a qualitative method of sampling designed to collect the greatest diversity of macroinve rtebrates possible in a limited time . Sediment for the determination of contaminants was collected from depositional areas at each site, with sampling techniques designed to avoid contamination from sampling equipment . Sediment for the determination of h eavy m etals was sieved through a 65 Âľm nylon m esh net and approximately 100 g of the sieved material was retained in plastic or glass jars for later laboratory analysis. Sediment fo r the organics analyses was scooped, using a stainless steel scoop, into a stainless steel container. The sediment was then passed through a 2 mm sieve to achieve a uniform particle size and the slurry placed into glass j ars that were kept chilled until analysed. Habitat assessments w ere made at each site and in situ physico- chemical measurements were also taken. Data was subj ected to exploratory multivariate analyses to: 1) identify relationships between benthic macroinvertebrate community structure and sediment contaminants; and 2) provide recommendations for the design of effective urban stream ecosystem health assessment tools.

Patterns of distributions of toxicants covered in this study were complex and several factors appeared to be involved in determining toxicant concentrations at any particular site. The clumped distribution of particular contaminants indicates that the most important of these factors is the source of the contaminant, w ith most contaminants appearing to be a res ult of local point source discharges. However, landscape features (e.g. stream gradient) and patterns of urbanisation and direct anthropogenic stream alteration also appear to play a key role. T he interplay of these factors m eans that m any unique combinations of stream alteration and contamination m ay exist ~ nd that even with 44 stream sites located ac ross Melbourne only a few sites could be classified as similar on the basis of sediment contaminants.

Contaminant sources The results of the study indicated that no toxic param eters were correlated with the m easured geomorphic variables (e.g. percentage substrate type etc.). Therefore local stream geomorphology was not an important determinant of contamination. Contaminants appeared to be distributed more in relation to degree of urbanisation, with som e differentiation between classes of contaminant and the locations at which they were recorded. For example, highest concentrations of cadmium, copper, chromium, lead and zinc were found in highly industrialised parts of western Melbourne. Multivariate statistical analysis showed these variables to be highly correlated with each other, indicating a common type of source; most probably electroplating industries . Electroplating industries commonly produce wastewaters containing zinc and chrome (Nalco, 1988). Copper, tin, silver,

The distribution of toxicants in urban streams With the exception of volatile halogenated organic compounds, all contaminants were recorded above analytical detection limits at one or more sites. Complex effiu ents from m ajor industries and runoff from highly built up areas appear to be a major source of sediment contamination in the Melbourne urban area . WATER SEPTEMBER 2001

ENVIRONMENT

cadmium and lead are also commonly involved in . electroplating (Cole et al., 1981) and are also likely to be released with efflu ents. T his probably accounts for the correlation between lead and other heavy metals in urban stream sediments. In ge neral , indu stries discharging was tes to waterways in Melbourne have a license to do so from the Victorian Environment Protection Authority. Other industries may discharge to sewer thro ugh a trade waste agreement with Melbourne Water or tru ck their waste to dedicated disposal sites. It is possible that the observed high levels of h eavy m etals at some sites may be due to illegal discharges, poor site managem ent prac tices or past prac tices prior to connection to sewer, as heavy metals are known to be very persistent.

nities. Similar views w ere expressed by Pitt et al. (199 5), in a stud y of urban stormwater toxicity . They concluded that effects of stormwa ter runoff on receiving w aters depended on a wide range of factors including the magnitude of the dry and wet wea ther discharges; the transport and fate mechanism s of the toxicants; and the effects from other discharges and receiving water conditions. T hey considered that these factors and the site-specific relationships betwee n them make the prediction of receiving wa ter effec ts difficu lt , if not imposs ibl e . Co n se qu e ntl y, if urb an waterway managers are to understand the sediment condition of their waterway assets they sho uld foc us their monitoring efforts on heavy m etals and stream biota in the first instance . In addition, they should also establish if any ongoing contamination is occurring and the so urce of this co ntamination so that it can be either stopp ed at the source or intercepted. Remediation efforts can then fo cus on histo rical contamination. Du e to th e p ersistence of heavy m etals, once the spatial extent of contamination is established, sediment monitoring need only be conducted every few years unless more compreh ensive rem ediation has been undertaken (eg, de siltin g, excava tion of contaminated soils etc) . In such situations, monitoring of sediments post rem ediation is important to establish if rem ediation efforts have been effective . Finally, routine monitoring of ma croin-

Adams, W .J and Saeger, V.W . (1993). Uti li ty of Laboratory M icro cosms for Predicting the Environ me ntal Face of C hemi cals: A Comparison of T wo Mi cro cosm D es igns with Bucy! Benzyl Phchalace. ASTM Special Tech11ical Publication No. 1216 1993. ASTM, Philadelphia, USA , p 103- 11 9. Anon. (198 1) Na tional Tox icology Program, report NTP 8037, N IH 82-1773 . Anon. (1992) . Standard Methodsfor the Exa111i1wtion of Water and Wastewater. America n Publi c H ealth Association: Washington , D C. ANZECC (1997) Draft ANZECC interim sediment quality guidelines. Prepared fo r Environmental R esearch Institute of the Supervising Scienti st August, 1997 Bagg, J. , Smith , J D. , and Ma her, W.A. (1981). Distribu tion of polycycl ic aromati c hydroca rbons in sediments from estuari es of south eastern Australia. Australian. J ou rnal of Nlarine a11d Freshwa ter R esearch 32:65-73. Cole, A. R. H. , Wates, D. W. , and Buca c, R. B. (1981) . Chemical prope11ies and reactions. Cole, Cole, Watts, and Bucac, 61 H aldane Street, Mc Claremont 6010, W estern Australia. EPA (1981). The use of sediments and bi ota to assess heavy metal pollution in streams Interim report: results from phase 1. August 1981 (unpublished). Environment Protection Au th ority, Vi ctoria.

vertebrate communities, although there

vertebrate communiti es should be

Melbourne Water Corporation. (1997). The state

we re some broad scale correlatio ns between particular environmental and contaminant variables . Some organic co mpounds did marginally excee d th e ma x imum AN ZECC interim sediment quality guideline levels (ANZECC 1997) at som e locations (eg. ant hracene , py r en e, benzo(a)pyrene, total PCB's, lindane, dieldrin, endrin and DDT) , h owever, no relationship between these parameters and macro.invertebrate community structure w as detected. R easons for this could be that there truly was no relationship beca use the contaminants were not present at toxic concentrations, or that any toxic effe cts we re minor and below the limit of detection of the m ethods used in this study.

sufficient to indica te many elem ents of overall urban stream health . If signs of impacts on ma croinve rtebrate communities are evident, and expected to be from sediment contamination, confirmation efforts should fo cus initially on heavy m etals as the present study has established that these are the m ost likely gro up of contaminants to have effects on stream health.

of grea ter 1'1elbourne's waterways: Strea ,nwatc/1 Annual R eport 1996. M elbourne Water Corporation , East Ri chm ond , Victo ria. MRHI (1994). Ri ver Bioassessment Manual. Version 1. Nati onal Ri ver Processes and Management Program. Monito ring Ri ver Health Ini tiati ve. Program Coo rdinator P.E. Da vies. D epartment of Environment, Sport and T erritori es, Land and Wa ter R eso urces R esearch and Development Corpo ration and Co mmonwealth Environment Pro tection Agency O'Loughlin, E., Young, W .J. and Mo ll oy, J.D. (1992) . Urba11 stonn water: impacts 011 the environ111ent. CS IRO cons ultancy report 92/29 Novemb er 1992. Nalco (1988). The Nalco water handboo k. 2nd Edn. Frank N. Kremmer (ed.). McGraw-Hill, New York. Pouliot, A. (1993). Th e effect of an urban drain 011 the aq uatic 111acroim1ertebrates in. Darebin Creek. Victorian Environment Protection Auch01~cy. R eport No. SRS91 / 016. R eed, J. (1992) . A biological assessment of lower K ororoit Creek. Victorian E nvironm ent Prote c ti o n A uth o rit y, R eport No . SR S90/ 012. Pitt, R ., Field, R. , Lalor, M. and Brown, M. (1995). Urban stormwacer toxic pollucan~ : assessment, so urces, and creatabilicy. Water Environ,nen.t R esearch 67 :260-275 .

Macroinvertebrate-sediment toxicant relationships One of the m ajor findings of this study is that o nly heavy m etals and macroinvertebrate co mmunity stru cture were found to correlate. Although , the results of th.is exploratory study do not constitute proof of this relationship they do indicate that the most profitable direction for future exp erimental, or co nfirmatory, studi es in this area sh o uld b e the relationship between heavy m etals in sediments and macroinvertebrate comnmnities . Other co ntaminants, w hi ch were all organic compounds , did not exhibit any significant correlation w ith macroin-

Conclusions Due to the wide variety of potential contam in ants and th e rang e of geomorphic alterations possible in urban stream s, it was difficult to elucidate clear trends b etween types of effects and particular macroinve rtebrate commu62

WATER SEPTEMBER 2001

Acknowledgments The proj ect was fund ed as part of the Land and Water R esources R esearch and Development Corporation Urban SubPro g ram in association with th e M elb ourne Water Corporation and was undertaken at A WT. The authors thank Kylie Lewin, Emma Bradshaw and Vin Pettigrove for their contribution to the proj ect. This paper forms an updated and abridged ve rsion of Sedim ent Chemistry Macroinvertebrate Fauna Relationships in Urban Streams. O 'Connor, N. , Lewin, K. , Moore , S. and Bradshaw, E. (1999) National River H ealth Program, Urban

Sub Program, Report No. 8, LWRRDC Occasional Paper 15/ 99.

The Authors Dr Nick O'Connor and Dr Suzanne Moore are environmental scientists specialising in aquatic ecology at ECOS Environmental Cons ulting. Tel. (03) 9859 6734 . Email ecos@bigpond. com

References