Water Journal August 2008

Page 1


Journal of lhe Ausrralian Water Associa~on

Volume 35 No 5 August 2008


OPINION Water After Gornout An Engaging Business CROSSCURRENT Notional, States, Industry, International, People in the News NEWS YOUNG WATER PROFESSIONALS

DBarnes, President, AWA TMollenkopf, CE, AWA


s 6 12




25 26


28 30 31

TECHNICAL FEATURES !~ indicates the paper has been refereed) WATER SENSITIVE URBAN DESIGN ~ Modelling of Greywoter Reuse for Kuching, Malaysia Decentralised greywater recycle has potential in high-density estates DYSMah, FJ Putuhena, SSalim and SHLai MEMBRANE TECHNOLOGIES A Carbon Action Pion for o Californian Desalination Pion 77%credits obtained by replacing import from an 800 km distant source, I0%by onsite, 13%by offsite initiatives NVoutchkov PUMPING & PIPELINES Upgrading Eroring's Cooling Water Capacity: Hydraulic Challenges An innovative design thoroughly analysed by three dimensional CFO modelling SScaturro, S Randhawa WATER EFFICIENCY [i] Water Usage in Multi-Unit Developments Conservation affects reticulation design parameters TCartwright, SKrishnan, CMoore Appliance Stock and Usage Patterns Survey 2007 Asubstantial decrease in water use since the 2003 survey AAthuraliyo, KGan, PRoberts COMMUNITY CONSULTATION Yorro Valley Water Survey: The Methodology Innovative and effective use of engineering students AAthuraliyo, J Hoye, MRoberts SLUDGE TREATMENT ~ Viscous Bulking: o Full-scale Investigation at Three Activated Sludge Plants Excessive polysaccharide production in response to high carbon concentrations Bvan den Akker, HMartyn, UKaeding The Solar Sludge Dryer: Moroochydore' s First Year Halves the weight of biosolids MThomas, GBick, PJensen, KHartley, RWalpole Dual Polymer Sludge Conditioning for High Speed Centrifuge The aerobically digested sludge had only a weak negative charge J Young WATER SUPPLY New Water Quality Regulation in Queensland: Resolution or Revolution? HGibson, PChier The South-East Australia Drought: Possible Causes Further declines in inflow across the MOB are likely WCai, TCowan Microfungal Contamination of Municipal Water Supplies - A Review Acritical review of the literature NBSammon, KMHarrower




66 72


79 83 89 92 95 98


103 116

OUR COVER Eastern Pipeline Alliance in an engineeringfeat laying three parallel large diameter pipelines as part ofthe Western Corridor Recycled Water Project. The project recently received honours in the ! WA Project i nnovation Awards. Photo credit: Chris Lees Photography. Journal of the Australian Water Association


AUGUST 2008 l


essure Pipe to

PE100 polyethylene pipe is ideal for infrastructure water supply. Its flexibility and strength makes it easy and cheap to install. Trenchless installation techniques reduce time on site and the costs associated with disruption to existing services and utilities. The fully welded joint system reduces the need for anchorage blocks and prevents leaks from developing over time. PE100's high tensile strength ensures it can easily withstand the water pressures required. PE100's corrosion resistance and slow crack growth resistance ensure long life. This provides substantial longterm savings by eliminating the need for costly repairs. PE100 will not lose carrying capacity due to internal encrustation and is unaffected by ground movement. PE100 pipe being installed Into old pipe.


Polyethylene Metric Pressure Pipe is manufactured in accordance with AS/ NZS4130 "Polyethylene (PE) pipe for pressure applications". Furthermore PPl's pipe has WaterMark Product Certification which entails an independent audit of material selection, production and testing. It signifies an added comm itment to quality and is requ ired by all plu mbing and water authorities. WaterMark branding is standard on blue or blue stripe pipe.

Material: The basic polyethylene (PE) material classifications used for pressure pipes are PESO and PE100. The number classification refers to the strength of the material with PE1 00 being stronger than PESO. Pipe made from PE100 can withstand a greater pressure than pipe with the same diameter and wa ll thickness made from PESO material. PE100 was first introduced into larger diameters but it is now bei ng adopted as t he standard material for all PPl's Metric Pressure Pi pe. This change is in line with worldwide trends and it ensures that our cust omers are able to take advantage of the latest developments in plastics.

Pipe Life: The 50-year stress regress ion data used for classifying PE pipe material has incorrectly lead to the assumption that PE pipe systems have a life expectancy of 50 years. In reality, such systems can reasonab ly be expected to last much longer than 50 years. Systems that are manufactu red and installed correctly ca n be reasonably expected to last in excess of 100 years before requiring rehabilitation . However a firm prediction can not be made, as there are many u nforseeable factors operat ing in each system.

Pressure Rating & SDR: The pressure rating (PN -expressed in bar) is quot ed at a temperature of 20°C for the conveyance of water. It provides a base guide to the pipes performance but it needs to be adjusted for other fluids and temperatures. The Standard Dimension Ratio (S OR) is the nominal ratio of the pipe's outside diameter to its wa ll thickness. . SOR i.e.

= min . outside diameter min. wall thickness

The higher the SOR, the thi nner the pipe and the lower the pressure rating. The term SOR is more widely accepted by engi neers as it accurately represents t he pipe di mensions and allows the pressure rating to be determined after assessing the installation's conditions.

111'!!11!1~ •Html PN 4

PN 6.3

PN a

PN 8 PN 10

PN 10


PN 12.5

PN 16

PN 20

Water at 20°c


PN 20

Water at 40°C


PN 25

Compressed air at 40°


Colour - stripes or jacket: PE pipe is normally black due to the carbon black added for UV stabi lity. In cases whe re colour is required to hel p identify the contents, then it can be added as stripes or a jacket. PPI uses S stripes on all diameters. Stripes usually provide adequate identification, but in small diameters in confined spaces, jacketing may be preferred. PESO Blue

Water or compressed air





PE100 Blue

Water or compressed air


Fire Services




Recycled/Reclaimed Water

PE100 Yellow




PP I is always prepared to help develop pipes to suit individual projects, so please contact our sales office to discuss your special requirem ents. Our extra capability includes other PE or PP materials, custom sizing, speci al colours, jackets & stripes.

PPI Corporation Pty Ltd

300 Email: sales@ppi.com.au

With regard to pricing iss ues, it has been noted that in March 2008, subsequent to the completion of the VCEC report, the Victorian Essential Services Comm ission (ESC) released a draft Price Decision for the regional water corporations, which adopted a higher cost of capital (in recognition of changed fi nancial market conditions) and flagged more flexible arrangements for cost pass-throughs during the regulatory period. The VCEC itself has found that prices may not increase as much as fores hadowed by the metropolitan retailers in their draft water plans if adjustments to operating expenditures and a review of deman d projections are allowed for, consistent with those identified by ESC. With regard to potentially significa nt price divergences, the VCEC recommends that the Govern ment's concerns would be best addressed by a reall ocation of Melbourne Water's wholesale water and wastewater costs between the retailers. If further measures are required to deliver greater pricing parity (after a full cost review by the ESC), the VCEC suggests that Government consider a deferral in retailers' regulatory depreciation. The Government supports-in-principle these approaches proposed by the VCEC, and has asked the ESC to advise on options to achieve its pricing expectations beyond those identified in the fi nal VCEC report. T he Government has also made several other interesting decisions incl uding: • to develop a state based access regime for water and wastewater infrastructu re services; • that the metropolitan water businesses should develop consistent internal guideli nes to facilitate sewer mining; • that rhe Mi nister fo r Water is respo nsible fo r strategic planning for the metropolitan water sector. [The O ffice of Water in DSE will coord inate and review the metropolitan water supply-demand strategy and work closely with the water businesses in chis process]; • chat the ESC should develop a methodology for ring-fencing costs to better understand the cost of providing water distribution, wastewater collection and retail services and to inform stakeholders about the allocatio n of system-wide costs; • chat the retailers remove any unnecessary differences between chem in the standards chat they impose on developers and plumbers; and • chat for each pricing pe riod, the Government specify in the Statement of Obligations the quantifiable outcomes chat it expects the retailers to achieve.

LETTER TO THE EDITOR Congratulations to Tom Mollenkop f on his article in the June Water. Although what Tom says should be obvious, much of what he has pointed out escapes the comprehension of many of our decision makers. Last year when I paddled my kayak down the lower Murray River I was proud to have come from Brisbane. While caravan parks on the Murray banks were watering their grass, no such activity would have occurred in Brisbane. The education program by the Queensland Water Commission has been very effective. The SEQ Water Strategy follows on from some of chis great work, as outlined by Elizabeth Nosworthy in the May issue of Water. Unfortunately I believe rhe strategy has a fa tal flaw. A major component is predicated on the construction of the Traveston Crossing Dam on the Mary River. The decision to construct the dam was made under duress with dwind ling supplies in the S.E. Queensland dams. It was a hasty decision. But the Water Commission, along with the government, has dug its heels in and insisted that the dam is essential infrastructure and will proceed. I can find no independent review that supporcs this claim. Chris Davis made sense with his comments to the senate enqui ry, one of which was, 'The single point all AWA members would agree upon is that any dam proposal, together with all other options, must be assessed comp rehensively and fairly across all economic, environmental and social aspects." There is no evidence that this has been done. A research centre fo r freshwater ecology is to be established for work after the dam is finished. The Climate Change Centre of Excellence is to investigate the possible climate changes after the dam is finished. At the Senate enquiry Senators MacDonald, T rood, Joyce, and Boswell said "At the completion of the Inqui ry we have been left with the very firm co ncl usion that the Traveston Crossing T he Government has also decided further reform over the longer term should address such matters as: • clarifying roles and responsibilities in an augmented, diversified and interconnected water supply system, as is being implemented, including long-term, statewide planning responsibilities;

Dam, in particular, is a political response to a serious problem, but is not one which will solve the problem." A report by Cardno and the Institute for Sustainable Futures at UTS said, "The proposed dam at Traveston Crossing on the Mary River represents a high total cost, high unit cost, high risk and high environ mental and social impact option. Hence using key decision-making criteria the Traveston Crossing scheme should not be considered for implementation and human and fi nancial resources currently allocated to this project should be reallocated to dealing with the respo nse to the current drought." O n my recent kayak journey up the Brisbane River, over the Conondale Ranges and down the Mary River it was difficult liscen ing to the heartbreak of some of those affected. T his is the price we pay for such infrastructure and most of th is damage is already done. It is the future that worries me. My hope, my dream, is that the QWC would apply holistic analysis to all water supply options including proposed dam sites. Do we not owe this to future generations? Steve Posselt www.kayak4earth.com

SUBMITTING LETTERS The AWA Water Journal Editorial Committee encourage your feedback on articles featured in the Water j ournal, or on issues affecti ng the water secto r. Letters should be less than 300 words, and if referring to a specific article, provide the issue number, page and article ti tle.

Please include your full contact details with letters, and email to journal@awa.asn.au or fax (02) 9436 0155.

• developing water markets to enable water to move to its highest value use including exploring the feasibility of a large user market in Melbourne; and • pricing reforms to signal efficient future investment and use. All in all, a good, thoughtful, progressive outcome.

Journal of the Australian Water Association


AUGUST 2008 29

water sector reform

AWA CONFERENCE ON URBAN WATER REFORM 'GOVERNANCE AND STRUCTURES IN A CHALLENGING CLIMATE' Hilton Hotel, Brisbane, 19 June 2008 Terry Loos, Policy Officer, Awa The continued interest in water reform around rhe nation provided the impetus for chis specialise one day conference. The agenda, speakers and delegates all reflected the importance placed on water sector governance and we were delighted co have so many eminent presenters. In all, a most stimulating mix of thought-provokers, visionaries and pracririoners and a coral of approximately 120 delegates. AWA's CEO Tom Mollenkopf opened proceedings wich rhe observation char Australian jurisdictions have taken a 'world of approaches' co urban water structures and regu latory reforms, but often with commo n themes. He suggested chat with the emergence of the 'rrifecra' of increasing demands fo r water, past over-allocation and decreasing yield due co climate change, the community and poli tical spotlight was focused on rhe water seccor remained expectations for fu rther reform.

coo much reliance is placed on restrictions. There are huge hidden coses of restrictions. Water is the only service where public education on the efficient use of che service and the pricing of che service work in opposite directions. The suggestions chat raising water charges would disadvantage low income people is not accurate because even in low income households the coca! expenditure on water is small relative co ocher household expenditures. Also, it is suggested char pricing is ineffectual because water use is largely price inelastic, but chis ca n be cackled by becrer billing practices chat make consumers more immed iately aware of rhe cost of their water usage. Dr Byron co ntemplated a future where private sector companies could enter the

water supply marker and, through their retailers, offer supply contracts co householders (using rhe existing network). He envisaged a futu re where householders had a choice in the security of their supply, particularly in relation co their discretionary water uses. In shore, a full menu of reforms (from light-handed co heavy) co uld involve: PRIC ING - Improved block tariffs >Scarcity pricing > Competition in Supply. TRADI NG - Rural to Urban> Major Users and Utilities > Household Level. STRUCTURES - Stronger Corporatisation/Commercialisation >Contracting/Outsourcing> Structural separation > Private ownership and competition.

Producriviry Commissioner Dr Neil Byron discussed rhe PC's recent Research Paper 'Towards Urban Water Reform' and argued chat while the urban water seccor has been performing well, and while considerable urban water reforms have been introduced since the 1994 CoAG water refo rm agreement, there is still scope fo r further reforms in the structu ral, governance and pricing areas. T he face chat 80% of urban Australia is on water restrictions, and char infrastructure solutions co chis situation are emerging in an ad hoc and sometimes political manner, suggests chat the system is still imperfect. Rainfall reductions are only pare of the problem. Inadequate institu tional arrangements are also a key pare. We muse be able co 'rweak' some improvements inco these arrangements. Dr Byron argued chat water services should be provided like ocher services such as electricity, gas, telephony, transport, heal ch. People wouldn 't accept restrictions co these services the way water services are restricted e.g. would you accept a restriction on phon ing your fami ly only between 4 and 6p m on Tuesdays and Thursdays? Ic is difficult co accept chat water services provision is efficient wirhouc a water pricing approach char is linked co scarcity of supply. Water services are coo cheap, and

Orica Watercare can help with: • Water treatment plants • Water reuse/recycling issues • UV disinfection • Asset protection and odour control • Sludge treatment/conditioning • Brackish water treatment For more infomation contact Orica Watercare: Ph. (03) 9665 7111 Fax. (03) 9665 7908

Journal of the Australian Water Association


AUGUST 2008 31

Australia's National Water Conference and Exhibition Registration opens October 2008 Look out for your copy of the registrat ion brochure in Water Journal November Edition 3ponsorship and Exh ibition Opportunites are Available


Organised by


uitable for a multitude of sewerage and drainage applications, Vinidex is roud to offer an advance d piping sys tem, which is environme ntally se n sitive, os t e ffective and easy to install. Ran gin g from 150mm to 900mm, th e pipes re comple m e nte d by a full rang e of fitting s available in all sizes.

13 11 69


E info@vinidex . com.au


Vi ~ W

www.vinidex.com . au

Systems & Solutio ns



With over 40 years of Hach product experience ... we are ready when you are!


Environmental & Process Technologies division of BIOLAB

technical features

water sensitive urban design 181

refereed paper

MODELLING OF GREYWATER REUSE FOR KUCHING, MALAYSIA D Y S Mah, F J Putuhena, S Salim, S H Lai Abstract This paper explores the potential for greywater reclamation by the Ecological Sanitation (ECOSAN) approach in the unsewered city of Kuching, Sarawak, Malaysia. Blackwater is dealt with separately by septic tanks. By recycling purified greywater for toilet flushing, outside use and irrigation, the pollution caused by current discharge of greywater to street drains is eliminated. A nine-dwelling pilot trial was successful. Modelling for blocks of up to 200 dwellings has shown potential water saving of about 40% with attendant cost savings. The Wallingford Software model, Info Works Water Supply (WS) was employed to carry out a mathematical modelling of a hypothetical greywater recycling system as an integrated part of the Kuching urban water supply.

Introduction The philosophy of Ecological Sanitation , also known as ECOSAN, is an alternative sanitation technique based on the concep t of human excreta and wastewater as a valuable resource being developed by a number of European countries (Langergraber and Muelleger 2005, Werner 2006) . Nutrients from human faeces and urine are recovered fo r the benefits of agriculture. Such systems also ensure that water is used economically and is recycled in a safe way to the greatest possible extent for purposes such as irrigation or groundwater recharge.

Pilot Study of Ecosan in Sarawak The Sarawak State Government in Malaysia has collaborated with the Danish Cooperation on Environment and Development (DANCED) to establ ish an ECOSAN pilot project fo r treatment of greywater, in its capital, Kuching City, since 2003. The city is located northwestern part of Borneo Island with a population of 579, 900 (2006 census). Kuching is not sewered and, as in most Malaysian houses, the greywater is released untreated to the stormwater drains. Blackwater undergoes a partial treatment in septic tanks before the overflow is discharged to the stormwater drains. Therefore the water quality of drains,




Figure 1. Detached Houses . streams and rivers in Kuching are heavily polluted. However, it is possible to collect and reuse greywater as it can be treated to a less health-hazardous standard. With the expertise from DANCED, a structure of constructed wetlands with integrated aerobic filte r using the Norwegian technology and sizing CTenssen et al. 2005) was adopted as pilot facilities in the Hui Sing housing estate in Kuching. The pilot system connects to nine households of single-storey detached houses with an average of five persons per household (see Figure 1). Greywater from the ki tchen, shower and washing machine is channelled to the ECOSAN faci lities and treated through an Oil and Grease Tank, then pumped to the Vertical Biofilter, and lastly flows through the Horizontal Biofilter. The ECOSAN treatment plant covers a rather small area of about 2.5 m x 4 m. A 3D depiction of such structure is shown in Figure 2.

Water Quality Improvement Baseline monitoring and sampling results indicated a high efficiency of the pilot system. A summary of the average inlet and

Decentralised greywater recycle has potential in high-density estates.

Journal of the Australian Water Association

outlet concentrations is tabulated in Table 1. The project is under constant supervision and observation by the Sarawak Natural Resources and Environmental Board (NREB) to check any irregularities in the operation. The average values for treated greywater show more than 90% removal of BOD 5, COD, TSS and ammonia. The effluen t meets the WHO (2006) greywater management requirements with respect to BOD 5 (<l0mg/L). Currently, after treatment, the water is d ischarged as surface water to the storm water drains.

Potential For Reuse Since greywater is generated directly in every household in just about equal volumes every day independent of the weather, it presents a constant resource and it is quite a waste to throw away. Such treatment facilit ies as ECOSAN open up the potential for reuse for non-consumptive purposes. This lowers water supply costs, as potable water is expensive in treatment co sts and the needs of storage facilities management, even though Malaysia has no shortage of water. This paper explores the potential of ECOSAN greywater recycling in Kuch ing C ity on a larger scale. Non-consumptive water reuse has an excellent and welldocumented performance track records, which to date have featured no documented health problems, strong public acceptance

technical features refereed paper

and good regulatory compliance (Radcliffe 2004). Network simulations were evaluated before time, money and materials are invested in a real-world project. The modelling approach was rested as a medium size sys tem in Tabuan Jaya housing estate in Kuching City. The model can be used co adapt the methodology to local conditions and to allow for replication of the med ium system to a larger system elsewhere. T he Tabuan township is a suburban satellite of Kuching City, located about 5 km away from the city centre with a standin g establishment of over 20 years. Ir covers an area of 0.85 km 2 with a dense population close to seven thousand. By mapping the urban land uses, breaking down to domestic, commercial, industry and public institutions in different layers to provide a full picture in one integrated system, it is easy to view and interrogate. By using the ESRI ArcView software, the model calculated a coral of 1335 units of domestic houses in the Tabuan Jaya housing estate, where the residential area contain s 372 units (27.9%) of single storey terrace, 704 units (52.7%) of dou ble storey terrace, 127 units (9 .5%) of si ngle storey

Pretreatment biofilter

Pump sump Oil and grease tank

Level control and effluent chamber

Horizontal subsurface flow constructed wetland

Figure 2. Graphical of Ecological Sanitation Design (Jenssen et al. 2005). semi-detached and 132 units (9 .9%) of double storey semi-detached houses. The current scudy concentrates on the domesric users, thereby the commercial users are not included for the case study. The major principles ofECOSAN are that the system is on-site local treatment of greywater and decen tralised in nacure. Taking the Hui Sing housing estate pilot

project as reference, the treatment plane is expected to be small and sui table for urban environment. In chis work, the number of houses is hypothesised as a target of about 150 - 200 houses for each treatment system, so about seven ECOSAN systems are needed, and the estate was divided into seven stand-alone close-looped recycling systems.

Journal of the Australian Water Association


AUGUST 2008 43

tech n ica I features refereed paper

The Wallingford Software model, InfoWorks Water Supply (WS), was used. The existing pipelines were modelled according to the network modelling principles as the first scenario. T he second scenario represents the hypothetical greywater system settings, in addition to the conventional water supply system, in which the treated effluent from the ECOSAN faci lities is purified, reclaimed and reused in the domestic environment. In foWorks WS is not custom -made to model ECOSAN fac ilities. Thereby, the elements are simplified and modified in the model. T he Oil and Grease Tank and the Biofilters (combining both the vertical and horizontal) are modelled as rwo wet wells. T he pumps are modelled in parallel, wh ich is a normal practice. These pumps are modelled to produce a low fl ow rate to allow a retention time of about three hours (for treatment) but maintaining a high pumping head to allow fl ow downstream. T he greywater is to be transported to the treatment plant by gravity. 125 mm polyethylene (PE) pipes would be expected to be laid beside the storm water drains at the backyard of the consumers' houses. The treated water would be recycled back by a series of 125 mm pressurised PE pipes, with 12 mm PE pipe co nnections to households.

quality water for appropriate purposes. It is d ifficult to change the behaviours of local residents for their general perception is that water is cheap and plentiful. H owever, once the ECOSAN system is in place, water is saved with little or no inconvenience or need to change their lifestyles. Thus the

EC O SAN system presents potentially an ideal solutio n both to improve the river water quality, and at the sam e time, reduce the cost of potable water supply. The advantages of the integrated system are too great to ignore in an expanding city like Kuching.

Table 1. Average Inlet and Outlet Conce ntrations of ECOSAN System . Wenssen et al. 2005) Parameter

Inlet (n = 131

After Biofilter (n = 61 (n = 6)

Outlet (n = 131

% Removal

BOD5 COD TSS Total N • Ammon ia Nitrate Total P Faecal Coli. mpn/100ml E. Coli. mpn/l00ml

12B,75 212,42 75,95 37, 15 12,62 2,05 2,39

<2 11,93 6,01 14,38 2, 13 5,44 nd 5596 576

<2 10,6 1 2,66 9,24 0,83 5,27 0,33 646 389

>98,4 95,0 96,5 75, 1 93,5 -157,5 86,2

• Kjeldal + ammonia + nitrate Malaysia is a multi-cultural society. Some cultures practice wet toilet (no paper is used) and washing is necessary for them. We ore aware that the source of F. coli comes from the bathroom.

For modelling purposes in this project, the water uses are not modelled in detail but instead they are grouped into three mai n uses, namely Consumption, Recyclable and One-Time Use as in Figure 3. There was a lack of specific data, so for modelling purposes it was assumed that Recyclable greywater production is about 50% of water supplied, while h uman Consumption is about 10%. The remaining 40% is assumed as the One-Time Uses, e.g. toilet fl ushing, cleaning and irrigation. These assumptions are in close agreement to the Australian guidelines that greywater generation is about 55-65% (NSW D epartment of Energy, Utilities and Sustainability 2007).

Consu mption - drinking - cooking

I Recyclable 2· Consu~ption 3 One-Tune Use


To cosan/


/ Recycle Back .


Nehvo rk with ECOSAN

Water Consumption Table 2 presents the average water consumption for the four types of dwellings in Tabuan Jaya. Further consumer survey on detailed water use d ata would be needed. W hen these figures were inputted to Info Works, with inclusion of a hypothetical ECOSAN system in each of the seven zones as a looped system, the simulation predicted that water use would drop at three tapping points by 41 %, 44% and 36%, as summarised in Table 3. At present in Sarawak, little regard is paid to either minimising or utilising approp riate




, In: 12mm

PE ~

J .,.,..

M ;;--

Figure 3. Dua I Pola ble-Greywater System Model Ii ng.

44 AUGUST 2008 Water Journal of the Australian Water Association

Lrl<: System Type -


Best Practice Guidelines • Conceptual Design of Water Sensitive Developments (New! ) • WSUD Technica l Design Guidelines for South East Queensland • WSUD Construction and Establishment Guidelines (New! ) • WSUD Asset Management Guidelines (early 2009)

Professional Development Training courses currently on offer in clude: • Introduction to Water Sensitive Urban Design (WSUD) • Conceptual Design of Water Sensitive Developments (New! ) • Fundamentals of Bioretention Systems • Fundamentals of Constructed Wetland Systems • Fundamentals of Swales and Sediment Basins • Stormwater Harvesting (early 2009) • Advanced Class in Bioretention System Design (New! ) • Advanced Class in Stormwater Wetland Design (New! ) • Construction and Establishment of Vegetated Stormwater Treatment Systems (New! ) • MUSIC Assessors Training (New!) • Maintenance of Vegetated Stormwater Treatment Systems (early 2009)


Part of a National Approach to WSUD Capacity Building Water by Design is a member of the National Consortium of Capacity Building Programs supporting sustainable urban water management. Our parter programs are:

• WSUD in the Sydney Region (NSW) • Clearwater (Vic) • HCCREMS (NSW) • New WAier Ways (WA) • WSUD in the Adelaide Region (SA) • Derwent Estuary Program (Tas)

~,111""""""""11,,.llm., .i!iiiiiiiiiiiiiiiiii!iiiiiiiiiiiiiiiiiiiiii=


=:iiiiii!!!i!!!iii= .: =1si=

technical features

water sensitive urban design El Table 2. Estimation from Field Measurement and Billing Water Use Data . Estimated Parameters (// household!

Single Storey Terrace

Double Storey Terrace

Single Storey Semi-Detached

Double Storey Semi-Detached

Average Daily Use "

1050 1360

950 1590

1020 1590

1040 1590

MWA Standard of Daily Use•

" Estimated from Kuching Water Board billing records of 2006 (12 monthly data per household!

refereed paper

users is reduced by an average of 40%. H owever, more work is still required such as che energy consump tion of implementing such system, the associated coses in che attendant infrastructure and the toxicity risks associated with the use of greywacer before the eventual acceptance of ch is strategy.

• Malaysia Water Association (MWAI Design Guidelines for Water Supply Systems (1 9941

Acknowledgment Table 3. W ater Use Data at Tapping Points to Supply Tabuan Jayo . Network with Ecological Sanitation

Conventional Network

Daily Withdrawal (ML) Highest Peak Values /1/s)

0.52 11.39 0 .29 6.49

0.30 6.72 0. 16 3.64

Daily Withdrawal (ML) Highest Peak Values /1/s)

0.52 11.56

0.33 7.34

Tappi ng Poi nt 1

Daily Withdrawal (MLI Highest Peak Values /1/s)

Tappi ng Point 2 Tappi ng Point 3

Conclusion The Sarawak Scace Government is consid ering adopting che system; however, the responsible Ministry is facing some d iffic ulties in deciding the management and maintenance of such facilities. The current research work p resents an opportunity to incorporate the ECOSAN facil ities as pare of the urban water su pply system , rather

than solely as a sanitation system. Among the greywater treatment methods, T H E ECOSAN approach exhibits simple characteristics with low maintenance. T he mathematical modelling efforts of utilising Info Works WS to investigate the potential of ECO SAN greywater recycling system have shown positive results. The system-wide water demand of domestic

MICRON HORIZONTAL FIBREGLASS FILTERS Waterco's Micron horizontal fibreglass filters embocy the latest in fibreglass winding technology. Waterco's digitally controlled filament winding machine faultlessly winds continuous strands of fibreglass filament to create a filter vessel with refined consistency and superior quality. The Micron Horizontal filter's spherical ends are designed to give uniform flow from both inlet and outlet collection assemblies. The filter's hydraulic design provides adequate distribution to evenly expand the filter bed during backwash and the entire media bee is utilised during the filter cycle.

The authors acknowledge financial support by Universici Malaysia Sarawak through the UNIMAS Fu ndamental Research Grant No. 02(75)/576/2006(09). T he authors are thankful to Sarawak Stace Planning U nit (SPU), Sarawak ICT Unit (ICTU), Sarawak Land and Survey Department (L&S), Sarawak Natural Resources and Enviro nmental Board (NREB), Sarawak Department of Irrigation and D rainage (DID) and Kuching Water Board (KWB) for their support.

References Oeparrmenr of Energy, Utilities and Sustainabiliry (2007) New South Wales Guidelines for Greywater Reuse in Sewered, Single Household Residenrial Premises. ISBN 0975751689 . Jenssen, P.O., Lau, S., Chong, B., Huang, T. H. , Fevang, Y. and Skadberg, I. (2005) An Urban Ecological Sanitation Pilot Study in Humid Tropical Climate. Proceedings ofthe

3rd International Conference on Ecological Sanitation, May 23-26, Durban, South Africa. Langergraber, G. and Muelleger, E. (2005) Ecological Sanitation - A Way T o Solve Global Sanitation Problem. Environmental International Volume 3 1, Issue 3, pp. 433 444. Radcliffe, J.C. (2004) Water Recycling in Australia. © Australian Academy of T echnological Sciences and Engineering, ISBN 1875618 80 5, BPA Prine Group, Victoria. Werner, C. (2006) ECOSAN Project Experiences in German Development Cooperation - Examples, Obstacles and Opportunities. DWA-BMZ-GTZ ECOSANSymposium - New Sanitation Concepts lnrernational Project Experiences and Dissemination Strategies - Eschborn, Germany. WHO (2006) O verview of Greywater Managemenr: Health Considerations. Document WHO-EM/CEH/125/E.

The Authors


3210 3820 4570



MD5500 MD6500

21 73 2173 2173 2173

2040 2040 2040 2040

1200 1200 1200 1200

5.50 6.50 8.00 11 .00

200 200 200 250

198 234 288 396

6356 7548 9795 13329

SA/NT WA NZ OLD VIC'/TAS 0298988686 07 32999900 03 9764 1211 08 8244 6000 08 92731900 095257570

NSW \Head Office)

WATEi.CC ~ Water, the

liquid of hfe

www.waterco.com •

46 AUGUST 2008

© . .. '- • WAT ER ~ auAu1v , ~ 0 . . ~ ... ..,. . . :a,.!!::


Journal of the Australian Water Association

Darrien Mah Yau Seng is a Ph D candidate, email: darrien_ mah@ yahoo.co.uk; Prof Frederik Josep Putuhena and Prof Salim Said are in che Department of Civil Engineering of the Faculry of Engineering, U niversiti Malaysia Sarawak; Dr Lai Soi Hin is in the River Engineering and U rban Drainage Research Centre (REDAC), Universici Sains Malaysia.

www.abbaustralia.com .au

Power and productivity for a better world TM

jl 1111 ,.,1,1,

technical features

membrane technologies

A CARBON ACTION PLAN FOR A CALIFORNIAN DESALINATION PLANT N Voutchkov Abstract The Carlsbad desalination project is targeted to be in operation by mid-2011 and co supply 190 ML/d to San D iego County. In 2006 California legislation introduced the AB 32 Global Warming Solutions Act which aims to reduce the greenhouse gas (GHG) emissions of the state to 1990 levels by year 2020. In proactive response to chis legislation, Poseidon Resources has voluntarily taken upon the commitment to completely offset the carbon footprint associated with desalination plant operations. The mix of GH G reduction alternatives will be p rioritised and implemented under a Carbon Action Plan (CAP) which is a roadmap for carbon-neutral seawater desalination.


Figure 1. Carl sbad Seawater Desalination Project.

Over the past five years desali nation has gained a significant momentum in Californ ia. With more than ten projects in various stages of environmental review, design and construction, desalination is planned to provide 1,5 00 to 2,000 ML/d of new fresh drinking water supplies for the scare by year 2015. One of the largest and most advanced projects under development today is rhe 189 ML/d (5 0 MGD) Carlsbad seawater desalination plane (Figure 1). This project is co-located with rhe Encina coastal power generation station wh ich currently uses seawater for once-through cooling. The Carlsbad seawater desalination project is developed as a public-private partnersh ip between Poseidon Resources and eight local utilities and municipalities. The Carlsbad project is targeted co be in operation by mid-20 11. When completed, chis project would supply 6 to 8% of the d rinking water in San D iego County and would be the largest seawater desalination plane in the USA. In 2006 California legislation inrroduced the AB 32 G lobal Warmi ng Solutions Act which aims to reduce the greenhouse gas (GH G) emissions of the scare to 1990 levels by year 2020. The project proponent, Poseidon Resources, has developed a




Climate Action Plan which outlines a portfolio of operational and design technologies and measures; green energy supply alternatives and carbon emission offset initiatives .

Assessing Project Gross Carbon Footprint The total plant carbon footprint is dependent on two key factors: (1) how much elecrrici ty is used by the desalination plant; and (2) what sources (fossil fuels, wind, sunlight, etc.) are used to generate the electricity supplied to the plant. Both of these factors could be variable over time and therefore, the C limate Action Plan has co have the flexibility to incorporate such changes. The Carlsbad seawater desalination plant is planned to be operated continuously, 24 hours a day and 365 days per year, and to produce an average drinking water flow of 189 ML/d. When the plant was originally conceived in 2002 the coral baseline power use for this plant was projected at 3 1.3 megawam (MW) or 3.96 KWh/ m 3 of drinking water. This power use incorporates both production (2.96 KWh/m 3) and delivery (1 .0 KWh/m 3 ) of water co the

Journal of the Austra lian Water Association

utilities and municipalities served by the plant. However, seawater desalination technology has evolved since p roject inception, which allowed the total plant power use to be decreased to 3.57 KWh/ m 3 (2.68 KWh/m 3 of production). As a result, the total ann ual energy consumption for the Carlsbad seawater desalinatio n project used to determine the plant carbon footprint is 246,156 MWh/yr for 189 ML/d x 365d. To convert the desalination plant annual energy use into carbon footprint, chis is multiplied by the electric g rid emission factor. The power supplier for the desali nation plant operations (SDG&E) determines their Emission Factor based on a standard protocol developed by the California C limate Action Registry. The most recent SDG&E emission factor is

77% credits obtained by replacing import from an 800 km distant source, 10% by onsite, 13% by offsite initiatives.

www. m econ nel I dowel I. com CIVIL • ELECTRICAL •


Head Office: Tally Ho Business Park, 16 Lakeside Drive, Burwood East, Vic 3151 , AUSTRALIA Phone: 03 8805 5200 Fax: 03 8805 5376 Email: macdow_corp@macdow.com.au Sydney 02 8440 6300 • Perth 08 9315 4440 • Brisbane 07 3219 0456 • Adelaide 08 8346 8796


technical features

Kwi nana seawater desalination plant in Perch, rather than Pelton wheels. In addition to the state-of-the-art p ressure exchanger energy recovery technology, the Carlsbad desalination plan t design incorporates variable frequency d rives on seawater intake pumps, filter effiuenc transfer p u mps, and produce water p umps as well as p remium efficiency motors for all pumps in contin uous o peratio n chat use electricity of 500 hp or more. T hese efficiencies are, however, already accounted for in the above estimates of the desalination plane carbon footprint.

Warm water source

Figure 2. Solar Panel Rooftop System. 248.4 kg of CO2 per MWh. (www.climaceregiscry.org/CARROT / Public/Reporcs.aspx) T hus the coral carbon footprint for the Carlsbad seawater desalination project is calculated at 61,00 4 metric cons CO 2/yr. T he emission facto rs of all Californ ia power suppliers are expected co decrease in the future. For example, at present approximately 10% ofSG D &f' ~ retail electricity is generated from renewable sources (solar irradiation, wind, geothermal heat, etc.) and chis power supplier is committed to increase their share by 1% each year, reaching 20% by year 20 17.

Carbon Footprint Offset By Reduced Water Imports Currencly, San Diego County imporcs 90 % of its water from che distant (over 800 km) Sacramento Bay Delea and Colorado River via the complex system of che Scace Water Project. The amou nt of electricity needed co deliver and treat chis water is 2.748 KWh/m3, (2.589 KWh/m 3 for delivery; and 0.0 79 KWh/m 3 for evaporation losses fro m reservoirs and water treatment). Since the desalination project will offset the import of 189 ML/d of water chis will save 189,800 MWh/yr, or 47,240 metric cons CO 2/yr. Thus the Carlsbad desalination plane's net carbon footprint is estimated co be reduced from 61,000 co 13,764 metric cons COzfyr.

Further Actions To Reduce Carbon Footprint The main purpose of the Climate Action Plan for the Carlsbad seawater desalination




project is to eliminate che plane's nee carbon footprint altogether by: energy efficient design and operations; green building design; use of carbon dioxide for water pro duction; on-site solar power generation; carbon dioxide sequestratio n by creation of coastal weclands and reforestation; fund ing renewable power generation projects; and acquisition of renewable energy cred its. Carbon neutrali ty would be achieved by a balanced combination of these measures. The size and priority of the ind ivid ual projects included in the Plan will be determined based on a life-cycle costbenefit analysis and overall benefit for the local community. Implemen tation of energy efficiency measures for water production, green building design, and carbon d ioxide sequestration p rojects in the vicinity of the project site will be given the highest p riority. Periodic assessment and re-prioritisation of activities is a very essential compo nent o f the Plan because both desalination technology and green power generation (i.e. solar, wind and bio-fuel-based power) are expected co undergo accelerated develop ment over the next decade as they evolve from marginal co mainstream sou rces of warer sup ply and power su pply, respectively. T he specifi c carbon footp rint reduct ion measures incorporated in the Carlsbad C limate Action Plan, and their key benefits and co nstraints are d iscussed below.

Energy efficient design and operations The Carlsbad plant will recover energy from the exhaust brine by an energy recovery system similar co that used for the

Journal of the Australian Water Association

An average of 5% of add itional energy savings and respective carbon footp rint reduction (12,308 MWh/yr and 3,057 cons/CO2 per year) are p roj ected to be achieved by using warm cooling water from the Encina Power Generation Station as source seawater since osmotic pressure decreases with the increase of temperature. Over 80% of che desalination plane piping would be made of low-frictio n fi breglass reinforced plastic (F RP) and highdensity polyethylene (HDPE) materials, which in turn would yield add it ional energy savings for seawater conveyance. The desalinat ion plane operatio ns will be fully auto mated which would allow reduction of staff req uirements and associated GHG emissions for staff cransporcacion and services.

Green building design The d esalination plane will be located on a site curren tly occupied by a dilapidated fuel oil storage tank. Reclaim ing the land will reduce project imprint on the environment as compared co using an undisturbed site fo r the desalinatio n plane. A key "green" featu re of the plane design is ics compactness by shari ng co mmon walls, roofs and equipment wh ich would allow signi ficant reduction of its physical footpri n t. The total area occupied by che desalinatio n plane faci lities would be less than 5 acres. When built, ch is would be the smallest footpri nt desalination plane in the world per unit prod uction capacity (5 acres per 18 9 ML/d). For comparison, the 95 ML/d T ampa Bay seawater d esalinatio n plane occup ies 8 acres; che 273 M L/d Orange County G roundwater Recharge Project, which also uses reverse osmosis system, occupies approximately 40 acres; and the 325 ML/d Ashkelon seawater desalination plane, which currently is che largest seawater reverse osm osis faci lity in the world, occupies 24 acres. A plane with a smaller physical footprint would yield a smaller carbon footprint during

construction. Reduced construction site footprint also generates less dust emissio ns and requires less water fo r dust control. Building design will fo llow the principles of the Leadership in Energy and Environmental Design (LEED) program chat aims to reduce che overall impact on the en vironment of building construction and functions by: (1) sustainable site selection and development; (2) energy efficiency; (3) materials selection; (4) indoor envi ronmental quality, and (5) water savings. The building design team will include professional engineers who have achieved the LEED Accredited Professional certification. M inimising energy use for lighting, air conditioning and ventilation will be achieved by, for example, translucent panels co maximise daylight use and views co the outside, automatic light switching, monitored ventilatio n in che ind ivid ual working areas and water conserving fi xtures for service faci lities and landscape irrigatio n. The additional coses associated with the implementation of che green building design as compared co the coses for a standard building are estimated at US$5 million and che potential energy savings are in a range of 300 MWh/yr co 500 MWh/yr. The potential carbon footprint red uction associated with chis design is berween 75 and 124 cons of CO 2 per year (0.5 co 0.9% of che net power plane footprint). The unit cost is estimated at US$3,0 00 co US$5,000/ con of CO 2. The coca! actual energy red uction that would result from green building design will be reviewed by an independent third party consultant.

On-site solar power generation O ne enhancement of the green building d esign is che installation of rooftop photovoltaic (PV) system for solar power generation (see Figure 2). The main desalination plan t b uilding would have a roof surface of approximately 50,000 square feet, which would be adeq uace co house a solar panel system chat could generate ap proximately 777 MWh/yr of electricity and reduce the nee carbon footprint of che d esalination plant by 193 metric cons of CO2 per year, w hich is approximately 1.4% of che nee desalination plane carbon footprint. The construction cost o f the rooftop solar power system is estimated at US$4. l MM. T he net p resent worth cost of power generation using chis alternative is US$366,700 /yr which corresponds co 47.2 cents/kWh, about five times h igher than power supplied fro m the grid. T he unit cost of carbon footprint reduction for chis alternative is US$ l ,900/con of CO 2•


Use of carbon dioxide for water production Approximately 2,100 cons of CO 2 per year are planned co be used at the desalination plane for pose-treatment of che fresh water. At a pH of 8 .3 co 8.5, ch is CO 2 would remain sequestered until consumed with the drinking water. The desalination plant is planned co use on ly carbo n dioxide produced in a CO2 Recovery Plane. A certificate of origin requirement will encourage che suppliers to recover chis gas from industrial waste streams, e.g. breweries, rather than generating new gas by combustion. By sequestering 2, 100 cons of CO 2 per year in the desalination plant posc-creacmenc process, che net carbon foo tprint of the plane would be reduced by 15.3%. Ac annual expenditure for carbon dioxide supply of approximately US$ l 47,000/yr, chis carbon foo tprint red uction alternative is very cost-competitive (US$70/con CO 2).

Reducing energy for water reclamation The Carlsbad Municipal Water District owns and operates a 15 ML/d water reclamation plant which consists of advanced tertiary treatment facilities for the entire fl ow and a 3.75 ML/d reverse osmosis brackish water desalination system, which at present uses

www.acrodyne.com.au __, K::., BECK



J J .,EL


II':: fA..__ Lim/torque

fzc •

technical features

1,950 MWh of electricity per year to reduce the salinity of rhe treated effiuenr from 1,400 mg/L to below 1,000 mg/I in order to make rhe effiu ent suitable for irrigation (the city's drinkin g water can reach 1,000 mg/l ar rimes). The replacement of rhe existing city high-salinity imported water supply with desalinated water would eliminate the need for operation of the brackish water desalination p lant and would reduce the carbon footprint of rhe Reclamation Facility by 484 tons of COzfyr. Since this GHG reduction is direcrly credited to the seawater desalination plant operations, the Carlsbad desalination plant's carbon foo tprint could be reduced by 3.5%.

Offsets Re-vegetation of wildfire zones Almost every year pares of San Diego County experience large wildfires. For example, in 2007 San D iego wildfires burned over 35,000 acres, including forests, tree farms and urban forestry. In resp onse to an ongoing wildfire zone revegecacion program spearheaded by the California Coastal Comm ission , Poseidon has co mmitted to investi ng US$ LO million in reforestation activities. Ac an average tree planting cost of US$26.7/tree and maintenance costs for 25 years @ US$5.5/tree, ch e total cost would be US$ 164.2/tree. For US$ 1.0 M the total amount of trees planned to be replanted is 6,090. At an annual tree sequestration rare of 60 lbs/tree over rhe 25-year period, the total annual carbon footprint reductio n associated with the tree sequestration project is escimaced ar 365,400 lbs (166 metric tons) of CO 2 per year. This is approximately 1.2% reduction of che net desalination plant footprint. At an nual expenditure of US$33,500/yr, the unit carbon footp rint reductio n cost for this alternative would be US$202/ton of CO 2 .

Carbon dioxide sequestration in coastal wetlands As a part of the Carlsbad seawater desalination project, Poseidon Resources is planni ng ro d evelop 37 acres of new

coastal werlands in San Diego County. These werlands will be designed to create habitat for marine species similar co these found in the Agua H edionda Lagoon (see Figure 1), from wh ich source seawater is collected for che power plant and for desalination plant operations. Once the werlands are fully developed, they wi ll be maintained and monitored over the life of the desalination p lant operations. The cost of the werland restoration project is estimated ar US$ 3 .0 M. In add ition to the benefir of marine habitat restoration and enhancement, coastal werlands also act as a "sink" of carbon dioxid e. Tidal werlands are very productive h abitats that remove significant amounts of carbon from rhe atmosphere, a large portion of which is scored in the werland soils. W hile freshwater werlands also sequester CO 2, rhey are often a measurable source of methane emissions . For comparison, coastal werlands and salt marshes release negligible amounts of greenhouse gases and therefore, their carbon sequestration capacity is not measurably reduced by methane p roduction. Based o n a detailed study completed in a coastal lagoon in Southern California (www.slc.ca.gov/Reporcs/Carlsbad_Desali nization_ Plan t_Response/Attachment_4. p df) the average annual rare of seq uestration of carbon in coastal werland so ils is estimated at 0.03 kg ofC/m 2.yr. Another source (www.sfbayjv.org/tools/climare/CarbonWt landsS ummary_07_ Trulio.pdf) indicates that in addition to accumulating CO 2 in the so ils, central and southern California tidal marsh es co uld also sequester 0.45 kg of C/m 2.yr in the macrophytes growing in the marshes and 0.34 to 0.63 kg of C/m 2 .yr in the algal biomass. For the 37 acres (149,739 m 2) the maximum sequestration capacity of the coastal werlands could b e 1.11 kg of C/m 2.yr, or up co 83 cons of C/m 2.yr. With a conversion factor from carbon to carbon dioxide of 3.664 the estimated offset of the desalination plane carbo n foo tprint by the werland project is estimated at 304 tons of COif year (a 2.2% reductio n of the

net carbon footprint). At total present worth costs for werland development and maintenance of approximately US$120,000/yr, rhe unit carbon footprint reduction cost for chis alcernacive would be US$395/ton of CO 2) . Typically it takes three co five years for a coastal werland p roject to be fu lly functional and co begin co yield en hanced habitat. Site-specific research is then p lanned to quantify the GHG sequestration benefits.

Investing in regional renewable energy projects Poseidon plans co invest in a number of solar power proj ects with three of ics public partners. Table 1 presents the various solar power projects and associated GHG offsets. Based on rhe projects described in T able 1, the total carbon footprint offset fo r the desalination plant is projected ac 561 tons of CO 21 year (4.1 %) . (Cost of US$ l ,900/ ton of CO 2).

Other carbon offset projects and renewable credits Fo r the remainder of the Project's carbon emissions, Poseidon will purchase a combination of carbon offset projects and Renewable Energy C redits (RECs). Contracts for offset projects p rovide more price stability and are typically established fo r longer terms (10-20 years) then RECs (1-3 years) . About one-and-half to rwo years before operations begin, Poseidon will develop and issue a request for proposal (RFP) for carbon offset p rojects and renewable energy credits. Poseidon w ill rhen select the best mix from the resp onses and contract for their acquisitio n or development. T he exact n ature and cost of the offset projects and RECs will be known once rhe RFP process is complete. If Poseidon were to purchase local offset projects, each with 10 year contract terms, rhe initial solicitatio n would cost an estimated US$ 1.2 million in offset proj ects for the first decade and over US$5 million over the Project's 30 year life. Offsets or RECs will be used as the

Table 1. Reg ional Carbon Footprint O ffset Projects. Desalination Project Public Partner (Proponent of Green Power Generation Facilityl

Green Power Project Description

City of Encinitas

95 KW Solar Panel System Installed on City Hall Roof


Valley Center Municipal Water District

1,000 KW Solar Panel System


250 KW Solar Panel System

420 2,260

Rainbow Municipal Water District

Total Green Power Generation Capacity (MWh/ yr)

52 AUGUST 2008


Journal of the Australian Water Association

Annual Capacity of Green Energy Projected to be Generated by the Project (MWh/yr)

technical features

swing mitigation option to "true-up" annual changes to the Project's net carbo n foo tp rint.

Estimated Annual Carbon Emission Balance Table 2 presents the total and net carbo n footpr int estimates of the Carlsbad seawater desalination project and q uantifies GH G emissio n reduction and mitigatio n options which are planned to be implemented in order to reduce the plant net carbon emission footprint to zero. Analysis of data presented in Table 2 indicates that up to 43% of the GHG emissions associated with seawater desalination and drinki ng water delivery will be reduced by on-site reduction measures and the remainder will be mitigated by off-site mitigation projects and by purchase of renewable energy credits. It should be noted that the contribution of on-site GHG reduction activities is expected to increase over the useful life (i.e. in the next 30 years) of the project because of the fo llowing key reasons: • T he power supplier (SDG&E) is plan ning to increase significantly the percentage of green power sources in its electricity supply portfolio, which in turns will reduce its Emission Factor and the net desalination plant carbon foo tprint.

• Advances in seawater desalination technology are expected to yield further energy savings and carbon footp rint reductions. Over the last 20 years the use of power for production of fresh water by seawater desalination has halved. Such a trend is projected to continue in the fu ture.

wildlife, and wi ll have added recreational benefits for county residents and visitors.

Summary and Conclusions Greenhouse gas emissions associated with prod uction of desalinated seawater at the Carlsbad seawater desalination project in Southern California are planned be mitigated by state-of-the-arc energy recovery technology and by a portfolio of alternative tech nologies and offset measures:

The lowest unit cost of carbon footp rint reduction can be achieved by using carbon dioxide for post-treatment of the desalinated water (US$70/ton CO 2). T he most costly carbon footpr int reduction o ptions are green building design (US$3,000 co 5,00 0/ton CO 2) and installation of rooftop solar power generation system (US$1,900/ton CO 2) . Development of new coastal wetlands is a very promisi ng carbon footpri nt reduct io n option (US$395/ron CO 2), which could be several times less costly than the construction of solar panel generation system of the same carbon footprint reduction capacity. Similarly, reforestation could also be a cost-competitive GHG reduction alcernative (US$202/to n CO 2) . As compared to green power generatio n (solar and wind power) reforestatio n and wetland mitigation have added environmental benefits.

The most significant is the credit from reduced water imports from the existing energy intensive Seate Water Project. Additional credits are obtained from the use of carbon dioxide for water production, green building design, the development of on-site and off-site green energy projects and carbon dioxide sequestration by reforestation and new coastal wetlands. The mix of GHG reduction alternatives will be prioritised and implemented under a Carbo n Action Plan which defines a roadmap for carbon-neutral seawater desalinatio n.

The Author

Nikolay Voutchkov is Senior Vice

Fo r example, the new coastal wetlands will be des igned to create habitat fo r a variety of marine organisms, plants and

President - Technical Services, Poseidon Reso urces Co rporation, email: nvoucchkov@poseidonl .com

Table 2. Desalination Project Net GHG Emission Zero Balance. Carbon Dioxide Emission Generation Source

1. Seawater Desalination and Product Water Delivery - High Energy Efficiency Design 2. Carbon Emission Reduction Due to Reduced Water Imports 3. Total Net Power Use & Carbon Emissions (Item 1 - Item 2)

Total Annual Power Use (MWh/ year)

Total Annual Emissions (tons CO2/year)

246,156 189,800 56,356

61,004 47,240

Accounted for in Item 1 (12,308)

Accounted for in Item 1


On-site Carbon Dioxide Emission Reductions

4. Energy Efficient Plant Design 5. Use of Warm Cooling Water 6. Green Building Design 7. On-site Solar Power Generation 8. Use of CO2 for Water Production 9. Reduced Energy for Water Reclamation 10. Subtotal On-site Power/GHG Emission Reduction (Sum of Items 4 through 9)

(300 to 500) (777)

(NA) (1,950) (15,335- 15,535)

(3,05 7) (75 to 124) (193) (2, 100) (484) (5,909-5,958)

Off-site Carbon Dioxide Emission Mitigation

11. CO2 Sequestration by Re-vegetation of Wildfire Zones 12. CO 2 Sequestration in Coastal Wetlands 13. Regional Green Power Generation Projects (see Ta ble l ) 14. Other Carbon Offset Projects and Purchase of Renewable Credits

15. Subtotal Off-site Power/GHG Mitigation Reduction (Sum of Items 11 through 14) Total Net CHG Emission Balance (Item 3 - Item 9 - Item 14)

Notes: NA - Not Applicable. 54 AUGUST 2008


Journal of the Australian Water Association

(NA) (NA)


(2,260) 38,561-to 38, 761) 40,821 to 41,021

(304) (561) (6,775 to 6,824) (7,806 to 7,855)



technical features

pumping & pipelines



Abstract Innovative techniques were used in che design of the Eraring Transfer Pumping Station co minimise its foo tp rint and reduce excavation. Some innovations created complex hyd raulic conditions and the need fo r parts of the system to accommodate flow in two directions based on the station's mode of operation. These challenges had to be addressed during che detailed design stage co ensure chat the system would meet performance requirements. This paper examines the differences between the initial concept design and the fina l design, identifies the resulting hydraulic challenges, and describes the design solutions. The fina l design reduced construction costs by over 15%.

Background Eraring Energy operates a series of wind, hydro, and coal-fired electricity generation plants, supplying approximately 9% of energy required on rhe National Electricity Markee (An nual Report 2007). Located on the western shore of Lake Macquarie near Dora Creek in New South Wales, Eraring Power Station is the largest of the corporation's generation planes with a capacity of 2640 MW. Seawater pumped from Lake Macquarie flows through a high level canal into the power station to cool its generators. After circulating through a series of condensers, the seawater is discharged into an outlet canal char flows back to Lake Macquarie. A weir is provided o n the high level canal to transfer additional water to the outlet canal when necessary. This transfer cools the discharge water to a temperature acceptable for release into Lake Macquarie. Because recirculation of che seawater raises the temperature of the lake, Eraring Power Station is required to operate under

An innovative design thoroughly analysed by three dimensional CPD modelling. 56








, ... w...

Figure 1. Locality plan . stringent environmental co ntrols. During the hottest days of summer when energy demand tends to be greatest, rhe temperature of d ischarged water rises to a level chat forces che power station to reduce generation. T o meet che increased energy demand while maintaining an acceptable temperature of the d ischarged water, Eraring Energy called for renders to design and construct an 800 M L reservoir and transfer pumping statio n based on the initial concept design shown in Figures 1 and 2. Tenderers were also given the option to submit alternative designs, provided that all performance requirements were met. The pumping station will fill the reservoir from rhe high level canal during off-peak hours and contro l the flow from the reservoir back into the canal as needed for additional cooling during peak energy generation periods. The main performance requirements of the system are that reservoir fi lling time muse not exceed 16 hours, che reservoir draining time must no t exceed 8 hours, and a maximum flow rate of 40 m 3 /s must be maintained for at least the fi rst 3 hours of draining.

Journal of the Australian Water Association

Haslin Constructions engaged Parsons Brinckerhoff as che design consultant and submitted an alternative design which was accepted by Eraring Energy. T he $50+ mill ion design and construct contract was awarded to Haslin Constructions in midFebruary 2008. Ar the time of writing, the fi nal design was nearing completion with construction expected to commence in July 20 08.

Key Dates • Concept and hydraulic design Completed • Detailed structural design - 85% Complete • Start of construction - July 2008 • Commissioning - August 2009

Initial Concept Design In the initial concept design, the pumping station works are positioned around a triangular wet well chat has o ne side parallel to the high level canal as shown in Figure 2. Ten 18-m-long, DN1800 pipes configured in parallel across the canal bank carry seawater either into or out of the wee well.

Orrcon·s pipe and steel products are used in major projects throughout Australia.

[ urobodall;i Pipe: ire. NSW

Orrc'ln W'lter Pip,

Melbourne' Cr rke• GrOL r-d

Mi>lbourne Sport~ & Aquatic Cert ,

We offer an alternative lined piping system. Orrcon·s p ipeline system is not cement li ned, it's epoxy lined to AS 4020. Ou r epoxy l ining is smooth , resists bacteria , has no r esidue leac hing, less chemical dosing, is cheaper to r un and the resul ts are clean , clear w at er.

Member Of:




Stop logs can be installed on each of the DNl 800 pipes co isolate the pumping station from che canal so chat ic can be d rained for maintenance purposes. Because rhe high level canal muse be kept in operation during construction of the new works, a temporary sheet pile cofferdam approximately 70 m long is required along the ban k of the canal. T he pumping works, consisting of nine duty pumps and one standby pump, are adjacent co one side of the wee well. Each pump discharge line is installed with a non-return valve and fl ow control valve. Ar che interface between the wee well and pumping works are crash screens with a mechanical crash raking system. The reservo ir d ischarge works are adjacent co the o cher side of rhe wee well and consist of 12 DN900 pipes, each with a flow control valve.

HIGH LEVEL CANAL •••••••••••••••••- •••••••••••••

.........·•·· ................

, Temporary sheet-- i pile cofferdam j


3-m-diameter·-J....·a .1 pipeline to reservoir

Figure 2. Initial concept design .

Boch the pumping works and reservoir discharge works are con nected co a common step ped-diameter mani fo ld. The valves on the reservoir d ischarge lines are closed while the reservoir is filling, and rhe valves on the pumping d ischarge lines are closed while the reservoir is d raining. Thus, flow through the manifold is bi-directional based on the system's mode of operation. The manifold connects co a single 3-mdiamerer pipel ine via fou r DN2100 pipes, each with a DN2 l 00 flow control valve. T he 3- m-diamerer pipeline carries flow co and from rhe 800 ML reservoir located approximately 130 m from the pumping station. Ac the reservoir, the pipeline connects co a m orni ng glory type spillway which acts as a combined inlec/ouclec structure. The structure is roofed, approximately 30 m in diameter and 2 m in height. With a fl oor level approximately 10 m below the existing ground su rface, the p u mping station dep th is governed by the h igh level canal flow depth and nee positive suction head requirements of the pumps. Ac the reservoir, the depth of the 3-md iameter pipeline is dictated by the morning glory structure, resulting in an invert level ranging from 15 m co 20 m below existing ground level. Tunnelling was chosen as the method of installation due co this large depth and the presence of rock close to the ground surface.

58 AUGUST 2008


reduced by approximately 35% co 990 m 2 . In addition , the footprint became rectangular in shape and directly in line with the flow entering from (or exiting into) rhe canal as shown in Figures 3 and 4 .

Final Design T he most significant changes from che ini tial co ncept design included reducing the overall fo otprint of the pumping station and raising the 3-m-diamecer pipe invert level. Reducing the station's footp rint involved housing both the pumping works and discharge works in a single enclosure. By stacking che pumping works on cop of the reservoir discharge works, the pumping statio n footp rint was

The new pumping station layout includes a wee well and a d ry well. The lower level of the dry well contains 12 DN900 reservoir discharge lines which connect in pairs co six 1.5 m x 4.9 m box culverts in the lower level of the wee well. Flow from these culverts combines in the wee well and exits into the canal. The DN1800 pipes linking the canal and wee well in che initial concept design were removed, and twin box culverts with crash screens are instead used in the fin al design. Due co the increase in cross sectional flow area through these culverts, the wid th of the connection co che canal was reduced by half co approximately 16 m. Minimising the width of the culvert translated into minimising che span of che vehicular bridge above and reducing che length of the temporary coffe rdam by half. N ine submersible pumps sic above che six reservoir discharge culverts in the wee well and pump seawater into DN900 pump d ischarge pipelines chat pass

Reservoir Inlet/outlet structure Reservoir embankment

Wet well

Concrete encased 3-m-dlameter pipe 4.3 m x 3.75 m box culvert

1.5 m x 3.0 m box culverts

Figure 3. Pumping station and reservoir.

Journal of the Australian Water Association

ENERPAC.~ ~ Enerpac offers complete families of advanced bolting technologies to handle the toughest jobs dished out by the water and wastewater industries, increasing safety and efficiency in the fabrication and maintenance of pipelines and pumping machinery and plant. Backed by 50 years of experience in Australasia, our tools offer world-class power, precision, safety and efficiency for jobs like pipelines, topside and subsea construction, fabrication, plant servicing, heavy vehicle and dozer maintenance and all the jobs where you can't afford tools to let you down.

technical features

pumping & pipelines through the upper level of rhe dry well. The pumps are grouped into th ree bays wh ich can be isolated from each ocher and the rest of the wee well for maintenance p urposes. T h is innovation allows for only part of rhe wet well to be taken offline ar any given rime, thereby reducing uplift on the structure. Boch rhe pumping discharge lines and reservoir discharge lines are connected to a common set o f three 1. 5 m x 3 m box culverts at rhe back of the dry well. These box culverts lead to a 10-m-high drop structure which in turn connects to rhe main pipeline leading to the reservoir. The initial concept design placed chis 3m-diamerer pipeline below the bottom water level of the reservoir such char it flowed fully pressurised for all reservoir water levels. To change the construction method from tunnell ing to open-cue, chis pipeline was raised by approximately seven metres in the final design which resulted in significant construction cost savings. Because the 3-m-diamerer pipe is at a h igher elevation than in rhe ini rial concept d esign, it is not p ractical to have

Figure 4. Pumping stati on. a morning glo ry type spillway as a reservo ir inler/ouclec structure. The structure in the fi nal d esign has a 20 0 m 2 trapezoidal fo otprint with a roof and openi ngs on three sides. The

3-m-diamecer pipe connects to the side of the inlec/ourler structure and slopes down coward the drop structu re at 0.5%.

Hydraulic Analysis To co nfirm chat rhe fi nal design would meet the contract performance requ irements, detailed hydraul ic analyses were carried ou t for both the reservoir fill ing and reservoir draining fl ow scenarios. These hydraulic analyses were performed in conjunctio n with a sensitivity analysis of flow surface roughness. The rwo extremes between which the system is designed to operate are a new/clean system (low surface roughness) and a system with controlled marine growth (high surface rough ness and effective reductions in pipe diameters and culvert dimensions) . During reservoir fill ing, the pumps are required to o perate with a static head ranging from 2.3 m when the reservoir is empty to 17.1 m when ful l. This variability requires rhrorrling the pump flow co ntrol valves at low static head s. Fo r low surface roughness values, the reservoir filli ng time is 14.6 hou rs, increasing co 15 .2 hours for a system with con trolled marine growth . An initial hydraulic analysis of the fi nal design showed char the system impacted by marine growth co uld only maintain rhe required 40 m 3/s d rainage flow race fo r ap proximately 2.5 hours. Complete drainage of the reservoir rook j use under the required eight hours, leaving licrle margin . Several design modifications were implemented to reduce head losses, thereby allowing the system to meet

60 AUGUST 2008


Journal of the Australian Water Association

performance requirements. Because of the relatively high velocities through the 3-m-diamecer pipe, che greatest head loss was occurring in the transition from che 3-m-diamecer pipe, through che drop structure, into che three 1.5 m x 3 m box culvercs. Therefore, a 4.3-m-wide, 3.75-m-high, 30-mlong box culvert was introduced to connect the drop structure and 3-m-diamecer pipe. Lower velocity through the transition reduces head loss and improves overall system performance. In addition, che floor of the box culvert is chamfered at the connection with the drop structure. As a result, flow through the drop structure is less turbulent which creates more uniform flow conditions across che three 1.5 m x 3 m box culvercs that lead to the DN900 discharge pipes. Bell-mouths were added to the entrances of these pipes to further reduce head loss and provide more stable flow condi tions across the flow control valves. Subsequent hydraulic analyses with these design improvements showed chat the system impacted by marine growth can sustain the required 40 m3/s for 3.3 hours and completely drain che reservoir in 7.4 hours. For a new or clean system, 40 m3/s can be sustained for 4.5 hours, with complete drainage occurring in 6.7 hours. When che reservoir is full and all flow control valves are fully open, flow races range from 53 m3/s co 67 m3 /s depending on the degree of marine growth. Initially, each of the reservoir discharge li nes were fitted with a si ngle flow control valve to limit flows to 40 m3 /s. Upon reviewing the expected system performance, the valve manufacturer concluded char the valves would cavitace and produce high noise levels because of che extent co which chey would need co be chroccled. To overcome chis challenge, two valves are provided on each reservoir discharge line, each of which will be chroccled less chan in the case of a single valve. The valves will be gradually opened as che reservoir water level drops and flow race decreases. When che reservoir water level drops co roughly the same elevation as che obvert of the 3-m-diamecer pipe at che reservoir inlec/ouclec structure, flow through the pipe and box culvert is wholly free-s urface. The intermediate stage between this condition and a fully pressurised system produces complex flow regimes and required a more derailed hydraulic analysis. The free-surface flow, combined with a change in cross sectional area when che flow enters the 4.3 m x 3.75 m box culvert, creates the potential fo r a hydraulic jump. Further investigation of the flow regimes revealed that a hydraulic jump initially forms near che interface between che 3-m-diamecer pipe and che 4.3 m x 3.75 m culvert and moves along che box culvert cowards che drop structure as che flow race drops. The location of che hydraulic jump is also highly sensitive co surface roughness. Low surface roughness values cause the jump to occur closer to che drop structure, while higher surface roughness values cause che jump to occur closer to the 3-m-diamecer pipe. Overall system performance and reservoir drainage times are not negatively impacted by che travelling jump; the challenge was determining a suitable location for a flow measurement device.

Flow Measurement


FLOWSERVE ...............

Our long experience enables us to offer the highest levels of ingenuity, advanced technical ability and efficient after sales service, making us leaders in water, wastewater, industrial water and process water pumps. Flowserve's TKL Super-Titan split-case pumps are a typical case in point, offering: • Awide range, making it easy to choose the right model for highest efficiency and reliability • Pumps designed for continuous operation • Space saving horizontal or vertical configurations • Hard wearing pump materials to minimise downtime • The ability to pump water or low viscosity clear liquids rapidly, economically and reliably Compare! You'll find TKL Super-Titan pumps lead the rest in helping your installations stay productive and efficient all day, every day. For further information, please call, fax or visit our web site now.

Flowserve Pump Division 26 Faigh Street Mu/grave Victoria 3170 Phone +61 3 9562 0744 Fax +61 3 9562 9562

Early in che detailed design stage, it was determined char a single flow measurement device would be placed between the drop structure and che reservoir to avoid using a device on each of che twelve DN900 reservoir discharge pipes. W ithout Journal of the Australian Water Association

Water AUGUST 2008 61

technical features

pumping & pipelines substantially enlarging the footprint of the pumping station, the DN900 pipes do not provide the straight length required for reasonably accurate flow measurement. Prior to the detailed hydraulic analysis, the device was placed in the centre of the 4 .3 m x 3.75 m box culvert since access to this location would make it convenient for calibration and maintenance. The selected flow measurement device measures depth using a pressure sensor at a single location and averages velocity over a given distance using ultrasonic technology. Thus, placing the flow measurement device in the box culvert would lead to inaccurate readings due to turbulence resulting from the hydraulic jump. Placing the device near the brink of the pipe is also not recommended because critical depth occurs a short d istance upstream of the brink as the flow changes from subcritical flow in the pipe to supercritical flow at the beginning of the culvert; flow near critical depth varies rapidly and therefore would result in inaccurate readings. Upstream of critical depth, the flow transitions fro m pressurised to free-surface at a point when the fluid separates from the crown of the pipe. This separation point moves further upstream as the reservoir level and flow rate decrease, until the point when the flow through the pipeline is wholly free-surface. Based on the analysis of flow conditions in the box culvert and a further study of the flow behaviour within the 3-m-diameter pipe, it was decided to place the flow measurement device midway along the


Advertising To reach the decision-makers in the water field, you should consider advertising in Water Journal, the official journal of Australian Water Association. For information on advertising rates, please contact Brian Rault at Hallmark Editions, Tel [03) 8534 5000 or email brian.rault@halledit.com.au




Figure 5. CFD analysis of reservoir in let/ outlet structure during reservoir drain ing .

pipe. This location is least impacted by the transient hydraulic conditions.

Inlet/Outlet Structure - Reservoir Draining Theoretical calculations of flow through the reservoir inlet/outlet structure during reservoir drainage result in an average velocity of 0.4 mis through the trash screens for the maximum design flow rate of 40 m 3/s. Due to the trapezoidal footprint of the structure, higher velocities were expected near the vertices of the walls at the ends of the trash screens. While small variations in velocity thro ugh the trash screens are not a concern, a substantially higher velocity has the potential to cause blockage if debris is present near the structure. In addition, it can also present a safety issue in the unlikely - though possible - event that a person or animal fa lls in the vicinity of the structure. W ithout a convenient analytical solution to determine the magnitude of these higher velocities or the region over which they extend, CFD analysis was used to model the reservoir draining scenario. With a draining flow rate of 40 m 3 /s, the CFD analysis indicates that average velocities are approximately 0.4 mis through the trash screens as expected and maximum velocities reach about 1.2 mis over an area confined to the predicted locations. In addition, the CFD outputs

Journal of the Austrolian Water Association

show that this maximum velocity drops off rapidly; at a distance of 1. 5 m to 2 m into the reservoir away from the structure, velocity is negligible. A challenge associated with having the 3-m-pipeline at a higher invert level and connected to the side of the inlet/outlet structure is that the system will draw in air at relatively high reservoir levels. If large volumes of air are drawn into the pipeline, adverse flow conditions might lengthen the time required to drain the reservoir beyond the maximum allowable duration of eight hours. CFD analysis shows that surface depressions begin to form at a reservoir level approximately 0.5 m above the roof of the inlet/outlet structure as shown in Figure 5. It is expected that vortices will form and air will be drawn into the pipe when the reservoir levels falls below this point. At these reservoir levels, approximately 85% of the reservoir is drained. Air that gets drawn in to the system d uring this time does not have a significant impact on the overall drainage time as a portion of the system is already experiencing free-surface flow. Air valves are provided at critical locations on the 4 .3 m x 3.75 m box culvert to allow air movement into or out of the structure due to changes in flow regime.

technical features

pumping & pipelines o Energy Dissipation During normal operation of the pumping station, some minimum volume of water will always be present in the reservoir. However, certain maintenance activities within the pumping station will requi re char rhe reservoir and pipeline between the wet well and reservoir be completely drained. After these periods of maintenance, there will be no rail water to dissipate energy at the commencement of reservoir filling. Without any other energy dissipation measures, init ial fi lling flow rares up to 18 m 3 /s would require extensive areas of rip-rap around the reservoir inlet/outlet structure to reduce the high exit velocities and prevent scouring of the reservoir's clay liner. To reduce exit velocities when there is no tail water, several stilling basin designs were investigated. Typical designs are based on the Froude number of flow entering the basin and make use of dissipation blocks or baffle walls. Most of rhe basin configurations are based on empirical data collected from physical

modelling. An inherent limitation in predicting flow behaviours within the basins thus arises if rhe design deviates from rhe modelled basins. While it is possible to interpolate between the various modelled basins to approximate flow behaviour, it becomes increasingly difficult to do so the more the baseline design is modified. This limitation is further compounded under transient flow conditions, such as when the rail water level is nor constant. To suit specific site conditions and meet performance requirements, the design selected for the reservoir inlet/outlet structure is a hybrid between the CSU rigid boundary basin and the USBR Type III Basin (USDOT 2006). The CSU rigid boundary basin is suitable for use at the end of culverts or pipelines and includes baffle blocks to dissipate energy from the incoming flows. Though no rail water is required for the effective use of this basin, rhe design is classified as a srreambed level dissipater and is typically used to re-establish incoming flows to meet downstream

conditions. The basin is usually integrated into the downstream channel withour any further flow obstructions downstream of the baffle blocks. T o help prevent scour of the reservoir lining and prevent excessive sedimentation within rhe structure during reservoir draining, rhe end sill in the USBR Type III Basin was incorporated into rhe final structure. Because the reservoir inler/ourler structure is used to control flows into a reservoir rather than a downstream channel, one of rhe major differences between rhe designed inlet/outlet structure and standard stilling basins is the provision to allow flow through the side walls in addition to the front wall. Such a provision helps to achieve lower average velocities without requiring an excessively large structure. The final design of the structure was validated using CFD analysis which shows that exit velocities fall within an acceptable range.

Conclusion The innovations described in this paper ultimately led to a more optimised design. Ir was therefore imperative to undertake derailed investigations to manage rhe potential risks. Analyses involved an iterative and integrated approach combined with rhe use of emerging technologies such as three dimensional CFD modelling. The need to minimise rhe overall footprint of the pumping station and reduce complexity of excavation was the d riving fo rce behind chis approach which was justified by rhe resulting cost savings to rhe clien r.

Acknowledgments The authors acknowledge the support and feedback from John H astings (Director) and Chris Janerzki (Project Manager) of Haslin Constructions. Both John and Chris provided valuable consrrucrability inp ut and advice during development of the design.

Authors Sal Scaturro (sscarurro@pb.com.au) is a Warer/Wasrewarer Engineer, and Satinder Randhawa (srandhawa@pb.com.au) is a Principal Engineer at Parsons Brinckerhoff, Sydney.

References Eraring Energy, Annual Report 2007. <http://www.eraringenergy.com .au/ docs/ EE_ 2007. pdf> US D eparrmenr ofTransporcarion (2006), Publicarion FHWA-NHI-06-086




Journal of the Australian Water Association

For information on our Airlift MBR system please contact Australia- 02 9918 5088

leading in purification

Singapore- +65 6338 9919


WATER USAGE IN MULTI-UNIT DEVELOPMENTS T Cartwright, S Krishnan, C Moore Abstract

Table 1. Current Design Parameters.

In 2005, the introduction by the NSW Department of Planning of the Building Sustainability Index (BASIX) for multi-unit developments meant that developers were required to p rovide a wide range o f water effi cient fittings and appliances to ensure compliance with the BASIX requirements.

Average Day (kL/day/NetHa) Maximum Day (kL/day/NetHa) Maximum Hour (kl/day/NetHa)

This paper d iscusses current water demands and the issues in deriving new design parameters for pipe networks supplying multi-unit developments. It emphasises char more data are required co confidently develop these new design parameters.

Introduction Revision of Sydney Water's current design parameters for multi-unit developments has nor occurred fo r 10 years. The application of the Building Sustainability Index (BASIX) to multi-unit developments will change demands char are ap plied to Sydney Water's pipe networks, both for new areas and in-fill development. Without modelling and actual data, it is difficult to estimate rhe effect of BASIX on any reduction in multi-units' average day, maximum day and maximum hour design parameters. The design parameters in dual reticulation areas are particularly difficult to d evelop due to variable demands on the recycled water system. Sydney Water intends to collect more data over the next few years and would welcome liaiso n between interested parties on sharing any available data.

The Impact of BASIX on Multi-Units BASIX is an online program, developed by the NSW Department of Planning, char ensures residential dwellings are designed to use less drinking water and produce less greenhouse gases. BASIX secs water reduction targets based on the NSW average consumption of drinking water. The benchmark is the average NSW annual drinking water consumption from rhe residential sector measured on a per capita basis and is equal co 90,340 litres of water per person per year (247.5 litres per person per day). The BASIX target for d rinking water usage in Sydney is a 40% reduction from rhe benchmark.




Density (Dwellings/ NetHa)

<30 21.6 41 82

BASIX models a dwelling's maximum acceptable water consumption in the fo llowing way. Using geographic location and number of bedrooms, BASIX predicts the likely numbers of residents using Australian Bureau of Statistics (ABS) occupancy data. BASIX then multiplies the water per capita benchmark by the occupancy rare and then applies the 40% reduction target. T o assist in satisfying rhe BASIX target, developers install water efficient products co reduce overall water demand. Multi-unit developers target in ternal efficiencies such as shower heads, raps, toilers and washing machines. Alternative water sources, such as rainwater or recycled water, may be required to meet the target in multi-unit developments with high predicted water consumption. Those multi-unit d evelopments char are in recycled water areas will generally have toilers and irrigation areas supplied with the recycled water. BASIX calculates potential water consumption based on water design features. With reduced external water demand per capita compared co single dwellings, multi-units are intrinsically more water efficient. BASIX reco gnises this when predicting water consump tion. Most multiunits achieve the target reduction with water efficient fittings and appliances (the exceptions co this can be dwellings such as town houses and villas, which usually require an alternative source of water, e.g. rainwater) . The savings from washing machines and dishwashers are available for determining the potential water consump tion for multi-units, whereas they

Conservation affects reticulation design parameters.

Journal of the Australian Water Association

30¡60 31.6 60 120

61.100 42

101 - 140 52.6

80 160

100 200

Table 2. Current Design Parameters. Density¡ (Dwellings/NetHa) Average Day (kl/day/dwelli ng) Maximum Day (kl/day/dwelling) Maximum Hour (KL/day/ dwelling)

> 140 0.42 0.8 1.6

are nor included for the single dwelling situation. BASIX fo r multi-uni ts will therefore not necessarily result in a 40% reduction from current multi-unit consumption. H owever, it is designed to achieve a 40% reductio n from the benchmark. Without modelling and actual data, it is difficult to estimate the effect of BASIX on any reduction in mulriunirs' average day, maximum day and maximu m hour design parameters.

Current Design Parameters The current design parameters were developed in 1996. For densities less than 140 dwellings/NetHa (NetH a = net hectare = gross hectare less reserves/parks but including roads) the design parameters are shown in T able 1. These design parameters are presented in fo ur ranges of dwelling density. For d ensities greater than 140 dwellings/NerHa, the parameters are provided on a per dwelling (unit) basis as shown in Table 2. For a flat, in an area with greater than 140 dwellings per net hectare with rwo occupants, the design parameter co rresponds to 210 litres per person per day compared to the benchmark (pre-BASIX) of 247.5 lirres per person per day. Note char the bench mark applies to all housing types, e.g. single dwellings and flats. Shortcomings of the current design parameters are: (i) they were developed prior to the introduction of BASIX and therefore are expected to be conservative; (ii) currently there are no official design parameters for multi-unit developments for

technical features refereed paper

water efficiency the individ ual p ipes (drinking and recycled water) in dual reticulation areas; and (iii) having o ne set of parameters covering all areas across the Sydney region means that some areas may have pipes designed with a conservative set of design parameters.

Table 3. Historical Demands - Sydney Wide.

Deriving New Design Parameters With BASIX Compliance

Table 4. Historical Dema nds - Example of Municipalities.




171 0.47

177 0.48

180 0.49

Average Yearly Demand (KL/year/dwl) Average Daily Demand (KL/day/dwl)





Lane Cove









The followi n g sections review some limited general data and also review specific data on developments. T he data provide b ackground information in deriving new design parameters and raise several issues.

2001 /02 Average Yearly Demand (KL/year/dwelling) 2001 /02 Average Daily Demand (KL/day/dwelling)

Historical Average Day Results

Table 5. Single Reticulation Data.

Some h istorical resu lts for multi-unit developments, includ ing town houses and villas, in Sydney are shown in Table 3. The results predate the impact of the current drough t restrictions.

Description (including building date)

Table 3 results show little variance on a Sydney wide basis. H owever, there is wid e variance around the results from area to area, as shown in Table 4 (results for 200 1/2 year) . T h is wide variance would most probably be caused by differences such as the occupancy ratio an d in the number of occupants who sray at home during the day. This variance o n a dwelli ng basis from area to area is an other difficulty in estab lishing


Block 1 - 198 1 Block 2 - 198 1 Block 3 - 200 l

No. of Units

39 42 60


2293 2512 3010

Dwl/ NetHa

170 167 199

No. of years of data

Average Day Drinking KL/ day/ dwl

Drinking KL/day/ NetHa

0.38 0.60 0.32

64 100 64

5 5 5

Table 6. Sub metering information. Block of Units






Cooling Tower Usage (litres/day/ dwl) Boseflow (litres/day/dwl)

5 36

8 218

70 38





e ft>UMPS

technical features refereed paper

uniform design parameters on a Sydney wide basis.

3 .00

Selected Multi-Units - Average Day Results

2 .~


...,.., 2.00

Developments built since 1980 were studied to assess the effect of water efficient fitt ings and appliances. These would have some of the water efficient fitti ngs and appliances as chose required by BASIX. T he results were derived fro m quarterly meter readings.








The results of average demand (kL/day/dwelling) fo r the individual blocks of units show that block 2 has a high demand (0.60 kL/ day/dwelling) relative to the ochers. Note that the current average day d esign parameter for this type of d evelopment is 0.42 kl/day/dwelling. This implies that there may be some issues with the results for block 2. Sub metering (with dataloggers) was placed on the blocks of units above (as well as some others) for a period of about two months in 2004. From th e results that occurred from 2 am to 5 am, an estimated baseflow for Block 2 of 0.04 kl/day/ dwelling was derived. This baseflow implied a degree of leakage that was not detected for the other two blocks o f units. The baseflow result would partially exp lain the relatively high demand for block 2 .

Cooling towers and baseflows Sub meteri ng results, specifically fo r cooling tower usage and b aseflows, for six blocks of units designated A to F are shown in Table 6. The baseflows, which are generally a combination of actual usage and leakage, are derived from information obtained from dataloggers berween 2 am to 5 am. Although the sample size is small, the resul ts seem to show that some cooling cowers are not being operated efficiently (see especially F). Audits by Sydney Water of co oling cowers in commercial buildings provided similar results. The document called 'Best P ractice G uidelines For Water Conservation in Commercial Office Buildings and Sh opping Centres (2007)' was produced from the audi ts and chis is available fro m our website www.sydneywater.com.au/SavingWater/ In Your Business/FaccSheets.cfm.

68 AUGUST 2008








n 11








Single pipe reticulation areas Three blocks of units in single pipe reticulation (drin king water only) areas in C h atswood and Rozelle were selected. Usage data of these selected units, as shown in Table 5, were obtained from quarterly billed data.



._. 1.50


V\J l'V\ ~

~I • /W v-...,"..,.,,

0.00 00:00


















Figure 1. Diurnal Flow Example. Current Design Parameters vs Dept of Commerce Modelling 240 220

- r - - - - - - - - - - - - - - --:-r~- - - - - 1



Mu Hour {Current Dtllgn)

• - Mu Hour - (Model)




Max Day (C urrent Oulgn)

• •Max D1y (Model)

lS160 ,- - - - ()

~ 140 :




~ .:,,:

. . . - .. ... - .

BO 60

--.... . .-

40 20












.. -


Dwellings/Net Hectare

Figure 2. Current Design Parameters vs Modelling - Multi-units. Table 7. Rouse Hill Dual Reticulation Data


No. of Blocks

No. of Owl







Ave. kL/day/dwl Drinking Recycled Total 0.28


0 .39

Ave. kL/day/NetHa Drinking Recycled Total 24.3



*One block of units (12 dwellings) had one exceptionally high quarterly reading

Table 8. Newington Dual Reticulation Data. No. of Blocks

No. of









Ave. kL/day/dwl Drinking Recycled Total 0.39

The high baseflow for the block of unit B seems to indicate a severe leak, numerous small leaks or a tank permanently overflowing.

Maximum Day and Maximum Hour Demands The information for the current maximum day and maximum hour design parameters was obtained many years ago from various flowmete rs in the Sydney area. BASIX, with its requirement for water efficient fitti ngs

Journal of the Australian Water Association

0 .20


Ave. kL/ day/ NetHa Drinking Recycled Total 28.6



and appliances, is expected to reduce demands . Determining the new BASIX driven maximum day and maximum h our d esign parameters for multi-unit developments will require substantial monitoring over a few years. The most important design parameter for pipe network design is the maximum h our figure. The sub metering study (with dataloggers) mentioned in th e previous section only occurred for a period of a few

technical features refereed paper

monrhs. A sample of one of the d iurnal cu rves for one block of units on 24 April 2004 is shown in Figure 1. T he morning peak can be seen as well as the significant baseflow in rhe early morning. Developing maximum day and maximum hour demands fo r this type of multi-unit developmenr was nor possible from this study due co the short rime frame.

Modelling of Maximum Hour and Maximum Day Demands


Newington (•U unital - Drinking

Newington (•41 unita) - Recycled





+ - - - - - - -- -- - ~

0.80 + - - - - - -- -- - - - - - - - - i

'5 0.60

+ - - - - - - - - -- -- - - -- --!

060 +-- - - - -- -- - - - - - - - - 1

J o.40

+ - - - - = . - - - - -- - -- - - - - - !


r-- - - - - - - - - - - - - - - ,

§ 020








2006/07 2001102 2002/03








'-------- - - - - -- - - - - ------' ~ • 02

, - - - - - - - - - -- - -- - - - - - - - - , ~ Route Hill (42 unital - Ofinklng 1.00


I rao +----- - - -- - - - - - - - - l ==-0.60

1 040

+ - - - - -- - -- - - - - - - - - - - - l

+ - - - - - - - --=..,



Rouse HIii (42 unita) - Recycled 1.00



+--- - -- - - - - - - -- -- --i

§ 0.20 + - - - - - - - - - - - - -- ---! Sydney Water : < commiss io ned rhe < 2001/02 2002/03 2003/04 2004/05 2005106 2001/02 2002/03 2003/04 2004/05 2005106 2006/07 Departmenr of Year Year Commerce co undertake modelling to provide Figure 3. N ewington versus Rouse Hill - Quarterly Data . estimates of the maximum day and maximum hour demands fo r mulri-unirs. T he study cakes into account rhe reduction in water use resulting fro m BASIX. The results fo r coral demands for single pipe systems are shown in Figure 2.

! 020


___..., rl,., ,.__,..,

The fo llowing observatio ns can be made: (i) the modelled maximum day demands are sign ifican rly lower than the current design parameters; and (ii) the modelled maximum hour demands are generally lower than the cu rrent design parameters, however the results seem to align closely after the 140 kL/day/NerHa density ban d . The model uses, as input data, estimated Sydney-wide maximum day demands for various en d uses char were facto red from average day demands. Maximum hour d emands are then derived using the Monce Carlo simulatio n model. The model does not specifically make allowances fo r rh ose blocks of u nits char have items such as swimming pools and cooling cowers. A factor of safety co accounr fo r these issu es is required for fo rmulating design parameters, however determining rhis is d ifficult ar chis stage.

Motor Driven Metering Pumps

Dual Reticulation Areas Mulci-unirs selected were a range of blocks o f units in Rouse Hill (see Table 7) and Newington (see Table 8). The Rouse Hill dual reticulation area currenrly serves abo ut 17, 000 dwellings in che north west area of Sydney, ulcimacely growing co about 35,000 dwellings. Recycled water production scarred in 2001. The recycled water is suitable for irrigation use and railer

WaterGuard Analysers 602 - as above and 702 model not shown


Journal of the Australian Water Association


AUGUST 2008 69

technical features refereed paper

water efficiency flushing as well as car washing. Newington is a smaller development contiguous to che Olympic Park with the recycled water supplied by the Sydney Olympic Park Authority. The results of average total demand (0.39 and 0.59 kl/day/dwelling for Rouse Hill and Newington respectively) highlight chat there is a significant difference between the two areas, with both d rinking and recycled demands higher at Newington. Issues are: (i) occupancy races o f the blocks of uni cs are not accurately known; (i) about one in three blocks of units in Newingto n had swimmi ng pools, while it appeared chat the blocks at Rouse Hill did not; (iii) it was not known whether there were cooling towers present in the units; and (iv) che Newington units are known to have irrigation systems and the landscaping appears very well maintained compared co the Rouse Hill units. Another way of analysing che results is by q uarterly data. As an example, Figure 3 shows the results for one block of u nits at Rouse Hill (42 units) and one at Newington (4 1 units) . T he results show chat the Newington block of units has considerably more recycled water usage as compared to the Rouse Hill block of units, similar to the results in Tables 7 and 8. The Rouse Hill quarterly data suggests chat the recycled water usage for chis block of uni cs would mostly be fo r toilet flu shing. Figure 4 presents the results for blocks of un its, rather than che summatio n of individual dwellings in one block. Drinking, recycled and total average day demands at Newington are presented to show the sp read of results for blocks of units even in the same area. Regression lines are plotted through the data to provide an indicative trend. As density increases, there appears to be an increasing demand for drinking water and total demands while a slight reduction in demand fo r recycled water. The results for each block of units can be quite variable, e.g. che recycled water demands range from about 7 to 30 kL/day/NetHa. There is little data on maximum d ay and maximum hour demands in dual reticulation areas .

Discussion Sydney Water is reviewing its design parameters for mult i-unit developmen ts due to the requirements of BASIX. Average demands for recent develop ments are obtai nable from quarterly-read water meters (mostly a master meter for che whole development). However, it is difficult co obtain an

70 AUGUST 2008


AVERAGE DAY Demand Newington Dual Water Data (kUday/Net Ha)

80 70

60 ~







40 30

20 10


0 20.00







Dwelling /Not Ha


• •

Drinking Ave Day (New'ton) Total Ave Day (New'lon) Linear (Recycled Ave Oay (New'ton))


Recycled Ave Day (New'ton) Linear (Qrinking Ave Day (New'ton)} Linear (Total Ave Day (Newton))

Figure 4. Newington Actual Average Day Demand - Hectare Basis. accurate breakdown of average demands in existing blocks of units. Issues include the lack of individual meters on units; unknown leakage races; and che usage (where relevant) of cooling towers, swi mming pools and irrigation. The current average d emand data analysed shows wide variance in the results, even in che recently built areas. There is a notably wide variance in recycled water usage between Newington and Rouse Hill. There is currently little data available for BASIX compliant development, especially for maximum day and maximum hou r demands. From a pipe network design perspective, understanding maximum hour demand fo r multi-unit develop ments is vital. Trying co assess che design parameters in recycled water areas is further complicated as two pipe systems (drinking and recycled water) need to be designed. This is especially so for the recycled water pipes, as demand seems to vary widely.

Where To Now? BASIX will affect che demands in multiunits and hence the design parameters. Modelling of demands has been undertaken incorporating the BASIX requirements. H owever, as many assumptions were used , fi eld verification of the modelled demands will be required. To do chis, monitoring of numerous mul ti-unit developments must occu r. Sydney Water intends ro increase its monitoring in the next few years. The additional field data will make it possible co develop the revised d esign parameters with a degree of confidence chat currently does not exist. The moniroring of demands has the additional benefit for che building/un it

Journal of the Australian Water Association

owners as it facilitates the detection of any problems. The problems already discovered by Sydney Water, while monitoring, indicate chat dacaloggers would be advisable in any large d evelopment. Eventually, the development of benchmarks of usage for certain developments will be a welcome byp roduct of the monitoring. We would welcome liaison between interested parties who share an interest in d esign parameters.

Qualification The actual water usage d ata for multi-unit developments presented is provided co interested parties for general information and co water authorities for assistance with developing their design parameters. However, care is required in using the p resented water usage data as it is expected chat demands would vary from one development co another due co items such as the overall number of bed rooms in the developments, differe nt climates, lot sizes and demographics as well as the extent of efficient appliances installed.

Acknowledgment The authors acknowledge the assistance of Andrzej Liscowski of che Sydney Olympic Park Authority.

The Authors All authors work in che Water Configuration team in Strategic Asset Management at Sydney Water. Tony Cartwright is Product Strategy Leader, cony.cartwright@syd neywacer.com.au; Sheila Krishnan is Experienced Produce Strategy Planner; Chris Moore is Product Strategy Planner.




Hydrofilter'" Performance 0




99 %


95 %


96 %


99 %


90 %


cJ1 0



91 %


. .....


technical feat• featHrl!l-; rl!l-;

water efficiency

APPLIANCE STOCK AND USAGE PATTERNS SURVEY 2007 A Athuraliya, K Gan, P Roberts Abstract

Shower Duratio n (In mlutes)

The 2007 Appliance Srock and Usage Patterns Survey (ASUPS 2007) study involved household visits co 850 homes in Yarra Valley Water's area and was undertaken from August co October 2007. le gathered extensive data on both the stock profile of water-using appliances and the way in wh ich they are used.


• To enable rhe establishment and maintenance of en d- use models used co p redict demand, • To assist the des ign and evaluation of demand management programs, and • To inform commu nications with customers regarding rypical and efficient levels of water use. ASUPS 2007 is the third large sample research scudy of chis nature undertaken by Yarra Valley Water, the p rior studies being the Residential Forecasting Study 1999 and the Appliance Srock and Usage Patterns Survey 2003. All these studies primarily involved respondents estimating their usage patterns on the basis of recall as opposed co actual measurement over a period of time. Yarra Valley Water has also un dertaken two complementary end-use measurement studies namely the Residen tial End Use Measurement Study 2004 and the 2005 Evaporative Air Con ditioner Study. Whilst provid ing more accu rate data che sample sizes for such studies were necessarily small given ch e high cost of collecting end-use measurement data. Ac the rime of the study, Melbourne was in Stage 3a restrictions and usage both in and outside the home has decreased substantially. This is a summary version of rhe full report 'Appliance Stock and Usage Patterns Survey 2007' which is on Yarra Valley Water web site.

72 AUGUST 2008



ASUP 2007 Avoraga Duration: 5.7 minutes St d Deviation : 3.2 M edian Dura tion : 5 minutes


.g° l. i :

In order co better manage water resources it is essential co understand both the nature of water use in che residential sector (behaviours) and the on going changes in che types of appliances in use with in households (scock p rofile). This knowledge is important co Yarra Valley Water for a number of reasons, in particular:



. l











• ASUP 2007 -

• AS UP 2003






- •-





11 7

9 Duraton (minutes )



1] 13




Figure 1. Shower Duratio n - shorter than previous survey. W hilst che reduction in outside use is primarily che result of drought restrictions the reduction inside th e home (as shown by a 19% d rop in the Sep tember quarters, 2007 versus 2003) muse resu lt from voluntary behavioural and appliance change.

ind icator of usage. However, since che study sample had an average househ old size of 3.0 versus Yarra Valley Water's overall average of 2.73 persons the survey fi nd ings probably understate some or all of the true population parameters.


The frequency of showering has nor changed significancly remaining around an average of 0.9 showers per person per day.

The house-co-house survey was performed by che in novative use of th ird-year civil and environmental engineering scudents and is summarised in Achuraliya A, Hoye J, Robercs M. Appliance and Usage Survey: The Methodology (20 08).

Results: Comparisons Between 2003 and 2007 Relationship between household size and usage The survey resulcs show the same posi rive correlation between household size and usage as was foun d in 2003 albeit with a substantial downward shift in volumes. The variation in usage around the mean fo r each household size is as large as it was in 2003 reaffirming chat whilst household size is a major determinant of usage it is nor a good

A substantial decrease in water use since the 2003 survey.

Journal of the Australian Water Association


Conversely there is a modest bur statistically significant reduction in the average d uration of showering with an average of 6.2 falling co 5.7 minutes (Figure 1) . Perhaps che biggest change is the incidence of 3 scar (previously AAA) showers, which have increased their share of scock from 14% co 23%. This outcome reflects the success of showerhead exchange programs and regulations char mandate 3 scar showers in all new dwelli ngs. Typically householders run their shower at between one half and three quarters of the capacity fl ow race. The typical flow rare reduced from I 0.6 litres per minute (Lpm) co an average of9.5 Lpm. (F igure 2). T he typical flow race for all showers with less than 3 scar racing was measured at 10.2 Lpm which is well below their average capacity bur still substantially higher than the 6 .8 Lpm. for 3 star showers.

technical features

water efficiency Toilets The proportion of homes with at least one dual flush toiler has increased only marginally from 82% to 83%. However the share of toilets that are dual flush increased from 77 .5% of all toilers to 81.2%. Further classification of single and dual fl ush toilers into their various categories (6/3 litre, 9/4.5 litre etc) is problematic with some 15% of toilets unable to be accurately identified. Of chose chat could be identified, however, 56% were found to be something other than 6/3 litre or 4.5/3 litre dual flush toilers indicating substantial potential remains for efficiency gains. The average estimate of fl ush frequency was found to have decreased substantially to an average of 2.7 flushes per cap ita per day compared to an average of3.7. The extent of this decrease is surprisi ng and may be questionable given the degree of difficulty of estimating th is parameter. Average full and half flus h volumes were generally fo und to be close to specificatio n. The one outstanding exception to th is was for the half flush volume of 6/3 dual flush toilets where the measured average flush volume was 3.7 litres or 23% above specification.

Clotheswashers The penetration of front loading washing machines increased from 2 1% to 29.5% in 2007, an average increase of 2.1% per an num. This rate of increase mirrors chat fo und between the 1999 and 2003 surveys (Table 1). The average number of loads of washing per week was fou nd to be 4.3, substantially lower chan the average of 5.0 loads found in 2003. (Figure 3). This drop is plausible given the voluntary response to drought although che three surveys have demonstrated ch is parameter to be particularly vo latile varying between 3.8 in 1999 and 5.0 in 2003. There is a very strong correlation between the number of people in che household and che number of loads washed per week enabli ng a simple model to be fo rmulated to estimate average loads for each household size. Using chis model with Yarra Valley Water's average household size of 2. 73 persons gives an average of 4.1 loads per week.

Distribution of Shower Typical Flow Rate 30% ~ - - - - -- - - - - - - - - -- - - - - - - - - - - - - - ~


Average Flow Rate: 9.51 pm Std Deviation : 5.5 Median Flow Rate: 9 lpm

>- 20%

.. u









• ASUP 2007 • ASUP 2003

LltrH per mlnut,

Figure 2. Typical Shower Flow Rate - lower than previous survey. Oiatrlbutlon of Cloth•• WHMr Loads per WN k



----Average Freq: 4 .3 (5.0 In 2003) Std Deviation: 3.0 (3.3 In 2003) Median Freq 4 .0 (4.0 In 2003)


~ 25%

.. g..:li .. C






• ASUPS 2007 lllASUPS 2003



o•t. 1-2


7-$ Loada per WHk

Top loading Front Loading Have Washing Mach ine



Figure 3. Clothes was her loads per week - households washing less loads. The use of automatic water level selection co ntinues to increase and now is the commonly selected water level in 37% of households up from 28% in 2003. As was the case in 2003 very little clothes washing was found to be done by hand with 68% of households doing no hand washing at all and the rest doing 90% or more of their washing by machine.

Dishwashers Fifty eight per cent of households were found to have a dishwasher down from 62% in 2003. H owever chis drop is not statistically significant and intuirively it is

Table 1. Type of Clothes washer - front loader ownership continues to grow. Washing Machine Type




69.8% 29.5% 99.3%

78.0% 21.0% 99.0%

unlikely char the household penetration of dishwashers wou ld have declined. Of chose households with a dishwasher 6% indicated they never used it so in fact only 55% of all households actually use a dishwasher to wash their dishes. The average use of a dishwasher is 3.8 times per week, significantly lower than the average 4.4 rimes found in 2003. Most households still wash some of their dishes by hand bur in houses without automatic dishwashers hand washing is undertaken an average of 10 rimes per week compared to only 5.6 ti mes per week in houses with automatic dishwashers. Eighty fou r per cent of households do their hand washi ng of dishes with the plug in the sink down from 92% in 2003.

Indoor tap use Indoor cap use occurs at hand basins in the bathroom or ensuice and at the kitchen and laundry si nks.

Journal of the Australian Water Association


AUGUST 2008 73

technical features

water efficiency The average typical or usual flow rate for hand basins was measured at 8 Lpm compared co just 4.9 Lpm in 2003. However the high end range of flow rates suggests that in many cases it was the "capacity" flow rate rather than the "typical" flow rate that was in fact measured. Therefore the 2003 survey data are considered more reliable for this particular parameter. There was a significant decrease in the frequency of use of the hand basin with average uses per person decreasing from 5. 5 times per day in 2003 co just 3.5. Over 70% of households have either an aerator, flow controller or both on their kitchen sink rap. The use of aerators is far more common than flow controllers but these have only a relatively small reducing effect on the capacity flow rare compared co flow control devices. The average capacity flow rare of kitchen raps was found to have decreased from 19 Lpm to 15 Lpm. Ar just over 20.2 Lpm the laundry trough raps had the highest average measured flow rare of the indoor raps al though chis has decreased significantly from the average 25.7 Lpm found in 2003. Incidence of small leaks at raps was generally found co be low bur nevertheless higher than was found in 2003. Nearly 5% of laundry raps were found co have a small leak whilst 4.4% of kitchen raps had a leak.

Baths Bach use by adults remains minimal with 78% of households having no adult use of the bath at all. Average usage per adult is less than 0.2 baths per week. On average when rhe bath is used by an adult it is just half fi lled. Bath use by children is far more frequent although still nor widespread with baths ..., being taken in just 21 % of households. For households char use the bath with children less than 12 years of age the average use is 2.5 baths per child per week.

Table 2. Air Conditioner Ownersh ip Evaporative conti nues to grow. Type of Air Conditioner

% of Households

Reverse Cycle

26% 19%

Relrigeroted Evaporative

27% 73%

All Types

pote~cial co investigate current behaviours than in previous surveys. Ir also limits comparison with the 2003 survey in some cases. For example in 2007 irrigation with a bucker or watering can is the most common methodology used co irrigate garden beds but this has co be viewed in the context char manual and automatic sprinkler systems are currencly prohibited. However respondents were asked how they anticipated they would irrigate their lawns and gardens when restrictions were lifted and these provided some useful insights into how behaviours may differ in the future. Ninety one per cent of surveyed properties had some sort of garden. However under Stage 3a restrictions 32% of homes with gardens (as distinct from lawns) do not water them at all. Even when restrictions are lifted 26% of homes still said they would not be watering their garden (Figure 4). In rhe case of lawns 65% of homes with gardens indicated they would not irrigate chem when restrictions are removed. These findings suggest that a permanent behavioural change in outdoor use may have occurred as a result of extended drought restrictions.

Perhaps influenced by restrictions garden watering events are of fairly short duration with most garden irrigation being undertaken for only 15 minutes co half an hour. When restrictions are lifted the dominant garden irrigation methodology would return to being the hand-held hose with relatively litcle use of both manual and aucomaric sprinkler systems. H owever for lawn irrigation slighcly more irrigation would be undertaken with sprinkler systems than with the hose. Householders generally considered char their garden had survived well during the drought bur were less positive about their lawns where 22% felt their lawn had died. Use of rain sensors is rare with just 5% of households using one whilst soil moisture sensors are extremely rare with only I % of households using them. H owever rhe ownership of unused rain and moisture sensors is considerably higher probably as a result of Stage 3a restrictions. Questions about car washing frequency and method were abandoned since there was a coral ban on chis activity for a lengthy period prior co and during the survey.

Rainwater tanks The peneuarion of rainwater ranks has increased co 19% of households compared co 7.4% in 2003. One in five rainwater ranks are connected to toi ler moscly in addition to the garden although a very small percenrage of ranks are just connected to the toiler (Figure 5). One third of homes built in the last five years have a rainwater tank and more than one third of these are connected co toiler as well as garden .

Frequency of Garden Irrigation • All Properties Under No Water Restriction

30.0% ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~ 26.6%


Over ninety per cent of households where children use the bath only fill co half capacity or less.

Evaporative air conditioners The percentage of households with an air condi tioner increased from 66% co 73%. This increase has been almost solely in evaporative units which are now in 27% of households (Table 2). On hot days the average estimated duration of use fo r evaporative coolers is 6.2 hours which is around an hour longer than the estimated 5.3 hours in 2003.

Gardens Because the survey was undertaken during severe drought restrictions there was less 74 AUGUST 2008


3 4 5 6 Garden Watering ¡ Nuni>erofTlmes per Week




Figure 4. Estimated frequency of garden watering when restricti ons are lifted.

Journal of the Australian Water Association

technical features

water efficiency Greywater


Whilst ownership of permanent greywater systems is fairly rare (3%) reuse of greywater is widespread with around 70% of homes using it ro some extent. The predominant forms of greywater reuse are from the shower and clothes washer with 67% of households using either or both of these sources.

Swimming pools Although ownership of swimming pools at 8% was found to be slightly lower than in 2003 this is not a statistically significant reduction. Forty five per cent of homes with swimming pools now use rainwater to top up their pools either solely (26%) or in conjunction with potable water (19%).

Conclusion The 2007 survey findi ngs for many usage parameters were generally indicarive of a substantial decrease in water use since the 2003 survey. Ir is Yarra Valley Water's intention to undertake additional end use measurement to complement the AS UPS 2007 in the 2008/09 financial year.


Rainwater Ta nk Ownership - ASUP2007









Don'I IWWl bl..c'#OUkt*-one

_ ..,..,..


,,. ,,. W '" D TrilO~andTOIM

Figure 5. Ra inwater Tank Ownership. Rainwater tanks rea lly appeal, with great opportunity to p lumb to toilet.

The Authors


Asoka Athuraliya is Forecasting Analyst (aathural@yvw.com.au), Kein Gan is Manager Water Supply Demand and Peter Roberts is Demand Forecasting Manager, all with Yarra Valley Water, Melbourne.

Appliance Scock and Usage Patterns Survey 2007 - Yarra Valley Water. Athuraliya A, Hoye J, Robcrrs M. Appliance and Usage Survey: The Methodology (2008) Water 35 No. 5.

Specialist Environmental Solutions Ecowise Environmental improves the management of our environment by delivering quality outcomes to our clients. Ou r expert staff provide environmental data, technology and consulting services for clients across government and industry. Ecowise Environme nta l has experience in catchments, mining, industrial and urban applications. We use the latest advancements in technology to provide innovative, efficient and cost-effective solutions that address regulatory and operational objectives in diverse environments.

Ph: 1300 32 69 47 I www.ecowise.com.au Consulting

I ''""

'"""'" I Analytical I GIS

61~ 1~ 1~

Ecowise Environmental

Journal of the Australian Water Association


AUGUST 2008 75

technical features

community consultation

YARRA VALLEY WATER SURVEY: THE METHODOLOGY A Athuraliya, J Hoye, M Roberts and dwelling type per suburb . In addition, it was also important to have a good m ix of ho usehold size (number of occupants), so overall target q uotas were also set fo r household size, based on Australian Bureau of Statistics' (ABS) Census 2006 population d ata fo r the area.

Abstract W hen Yarra Valley Water undertook a fu rther survey of usage patterns in 2007 they outsourced the selection of participant householders to the Ipsos-Eureka Social Research Institute (Ipsos), and the organisation of householder aud its to Coomes Consulting. T he project was successfully completed using engineering students as auditors and valuable data obtained.

Introduction In 200 7, YVW undertook a repeat Figure l . O verview of Project Mana gement Roles. of its Appliance and Usage Patterns Survey (ASUP), with assistance that has worked with YVW on a number of from the Ipsos-Eureka Social Research assignmen ts involving recruitment of Inst itute (Ipsos), Coomes Consulting and customers to participate in research. The Victoria University. team at Ipsos designed the recruitment The results are reported in the metho dology and questionnaire, and accompanying paper by Athuraliya et al. assisted YVW with the int rod uctory letter, T h is article su mmarises the methodology while I-view programmed the online and fo r a) recruiting a representative sample of telephone recru itment interview households, by Ipsos, and b) o rganising the questionnaires, conducted mail -outs of an team of house-to-house assessors from third introductory letter about the survey, and and fou rth year engineering students, by undertook and managed the recruitment Coombes Consulting. fie ldwork. An overview of the project management In itially, a letter was sent to all of the structu re is provided in Figure 1. customers asking them to register their

PARTICIPANT RECRUITMENT: IPSOS Recruitment of participating households began on 27th July 2007, and the actual survey started on the 11th of August and was completed in mid October. The survey covered 850 households of which 119 were rental p roperties. To achieve this sample size, 1042 peo ple were recruited to participate, allowing for some who may not have been available at the time of the actual household audit. The task of recruiting 850 of the customers fro m 15,000 was outsourced to Ipsos and its fi eldwork company I-view. Ipsos is a social and market research agency This arcicle comprises edited exrracts from the full paper by Hoye J, Athuraliya A, Robens P (WA I 5) in the Enviro08 Conference Proceedings and from Robercs M , Ath uraliya A (EFF 32) presented at the Water Efficiency Conference.




interest, using a unique identification number provid ed in the letter. Customers were offered a $40 MYER gift voucher as an incentive to participate, as well as their own sum mary report to keep; bo th to be provided o nce the audit was com plete. Participants were assured that there wo uld be no additio nal follow- up with their household in associatio n with the aud it. In the letter, customers were given three o ptions by which to register their interest 1. by completing a brief online survey on the website 2. by call ing a 1800 number, or 3. simply by responding to an invitation to participate th rough a d irect call. To ensure that as representative a mix of households as possible, YVW prepared a derailed target quota matrix. The matrix covered home owners vs tenants; standard customers, pensio ners and concessionaires,

Journal of the Australian Water Association

Only some of the information required was available o n the customer database. Therefore, to qualify for participation, additio nal data needed to be collected from those wishing to participate hence the need for a brief recruitment survey. It was also considered important to offer customers the option of speaking with a person, to answer any q uestions potential participan ts may have. Target quotas were programmed in to the survey scripts for both the online and telephone options, and monitored closely. A Compu ter Assisted Telepho ne Interview (CATI) Questio nnaire was provided to the call cen tre staff. The online q uestionnaire was essentially the same, although worded in self -completion format. Following the initial recruitment period, customers who had agreed to participate were contacted by the assessor team to confirm their participat ion and arrange a time for the audit. Almost 82% of those who were originally recruited ended up participating in the actual aud it. From the 15,000 customers to whom the letter was sent, 1042 were recruited to participate, which rep resents just under 7% of the total target population. The final sample closely matched the target quotas set by YVW, in terms of suburb, owner/tenant mix, stand ard/ pensioner/ concessionaire customers, as well as household size. Of the 1042 customers recruited to the research, a significant 60 % (or 620 households) were recruited v ia the online, self-co mpletion registration method. In addition, a further 196 customers registered their interest online, but quotas were

Innovative and effective use of engineering students.


No-one delivers like KSB. KSB Australia Pty Ltd VIC. 03 9314 0611 NSW. 02 9584 2099 QLD. 07 3725 8200 WA. 08 9455 7900 SA. 08 8234 0066 NZ. 09 476 4047 24 HOUR SERVICE. 1300 301 356 SPARE PARTS HOTLINE . 1300 KSB SPARES


Kse 6J

technical features

already fu ll. Roughly another 180 customers registered interest in the survey (either online or via the telephone) but either chose not to proceed, were unavailable for the duration of the research, or could not speak any English.

To reduce travel for studen ts, where possible students were allocated to one or two suburbs (neighbouring) only. Some students were not available every weekend or only on certain days and times and some preferred to work as a team and self selected their partners.

The online response in particular was well and truly above and beyond expectations, and has important implications in terms of understanding the ways in which customers can and w ish to interact with YVW.

Conclusions Specifically, chis survey has shown chat online interactions and surveys can be regarded as a valid, and popular, method for engaging YVW customers, and should be considered for future customer research. After 2007 the ASUP remains a relevant and effective means of involving customers in YVW's process for planning water resources, and helping to ensure that its water conservation programs are effective.

THE AUDIT TEAM: COOMES CONSULTING Coomes Consulting approached Dr Peter Lechte at Victoria University on the possibility of incorporating such a house-tohouse survey into problem-based learning assignments for third and fourth year swdents in civil and environmental engineering. The proposal was then submitted to YVW and details developed.

Risk Management A workshop was conducted between YVW, Coomes and University staff to identify major risks to the project such as swdents not wrning up at appointed times, car troubles, hard night etc, being accused of theft, OH&S issues, vehicle accident, safety within households, data collection issues, identification of appliances, and measurement. All these issues were d iscussed and mitigation strategies and appropriate controls established and agreed.

Strategies and Controls A roster and audit program was established by suburb, household type, auditor, date and time of audit (essentially this became a logistics and scheduling exercise) . Trai ning of the students was outlined customer service, preparation, checklist, conduct of the audit, entering dara and reporting. Weekend audits were coordinated e.g. emergency contact no's, appointment rimes with residents, rosters.




At the kitchen sink: measuring and recording flowrates. C hecklists for swdents were developed e.g. preparation (what you should have) from a spare pen to extra forms.

Organisation Compulsory training was essential students would go through the entire process from knocking at the door, introductions, measurement and recording to providing feedback. Uniforms would be issued together with measuring equipment, memory sticks, YVW software and printers. Method of payment, transfer of information and level of expectations were drafted.

Selection of Students 40 swdents applied, and 29 were selected, based on: • communication skills;

For each weekend the spreadsheet had to keep a record for each student to ensure that they were not doubled booked, continually update the total number of surveys to ensure chat targets in each of the categories were achieved.

Scheduling of Households Residents were provided with a brief outline of the su rvey process and asked to record appliance make and model. For chose with access to the internet,the software p rogram "Survey Monkey.com" was utilised so chat residents could register and provide responses to questions which were then provided to the scuden ts. Students were to make contact between 7:00 -9:00pm on week nights to schedule their weekend appointments. Shortly into the auditing p rocess students realised that 1.5hrs was too long and requested more households to survey so chat they were able to audit 5 houses per day and some (in pairs) were able to achieve 6 and 7. The students also started caking the initiative and calling residents co confirm appointments the night before and this increased the rate of residents being home for their ap pointment.

• own vehicle, mobile phone and laptop;


• willing to work weekends.

Apart from rhe gathering of valuable data on an increase in the uptake of water efficient products and appliances (see Achuraliya et al, this issue) there was valuable publicity for YVW, Victoria University and Coomes Consulting, and it successfully involved students and the university with a commercial process. All partners are confident that given the right coordination and training similar partnerships can prove successful in the future.

Training Volunteer staff households were selected (23 from YVW employees and 5 from Coomes staff) and training organised within rhese actual households. The trainers provided feedback and suggestions for improvement.

The Auditing Process Students were required to visit 4 households per student per day (1.5hrs per audit and 30mins travel time) at weekends to conduct a survey, measure flows from kitchen, bathroom taps, shower, laundry and toilets . They were also to inspect for leaking taps and toilet washers and finally to provide feedback and p resent a gift voucher to the resident for participating in the survey. An Excel spreadsheer was designed to import the data to YVW and divide the data into suburb, whether the household was standard, owner/tenant, pensioner, or nonpensioner/ concession.

Journal of the Australian Water Association

The Authors Asoka Athuraliya is Forecasting Analyse with Yarra Valley Water, email aachural@yvw.com.au; Jasmine Hoye is the Director, Sustainable Communities and Environment, at the Ipsos-Eu reka Social Research Institute, Melbourne, email: jasmine.hoye@ipsos.com.au; Mark Roberts is Principal Consultant - Sustainability and Environment, for Coomes Consulting Group Pcy Ltd, Melbourne, email Mark.Roberts@coomes.com. au.

technical features refereed paper

sludge treatment



The purpose of this study was to analyse the protein and carbohydrate composition of mixed liquor (ML) sampled from three Adelaide full-scale activated sludge (AS) plants (G lenelg, Christies Beach and Bolivar) and to investigate their possible affects on sludge secdeabiliry. This research has fo und char bulking sludge and variations in sl udge seeding performance, identi fied at Glenelg and Christies Beach, were most likely due co high mixed liquor polysaccharide concentrations. Boch Glenelg and Christies Beach are supplemented with carbon for improved denicrificacion and che results are indicative of excess available carbon; mosr likely caused by over-dosing of molasses during low flows. In co ntrast, Bolivar AS ML was typically lower in carbohydrate, and was significa ntly higher in protein composition. This was indicative of a lower and more stable carbon loading, and was coupled with better sludge settling performance. Introduction

Sludge settling is crucial to the successful operation of biological nutrient removal (BNR) processes (Verstraete and van Vaerenbergh, 1986). Sludge bulki ng is a universal problem encountered in the operarion of AS plants which can cause poor treatment performa nce and is undesirable because it adversely affects rhe separation of microbial floes into the concentrated sludge fraction and rhe clarified effluent. This can lead to high levels of suspended solids in the discharged created effluent, which has impacts both for reuse and the environment. In addition, sludge bulking may lead to a 'washout' of the resident microbial community responsible for nutrient removal. Two major causes of sludge bulking in rhe AS process have been identified. The most common is filamentous bulking, caused by filamentous organisms char develop under adverse conditions such as


ยง m


Centri fugal decanters for sludge thickening at Christies Beach STP.

low dissolved oxygen and low food to microorganisms ratio (F:M) (Sezgi n, 1982; W ilen and Balmer, 1999; Metcalf and Eddy, 2003). The other cause for bulking has been termed viscous bulking, and has nor been invesrigared to the same extent as filamentous bulking problems. Viscous bulking is caused by an excess of extra-cell ular polysaccharides (EPS), which produces a floe with sl ime-like characteristics (Metcalf and Eddy, 2003), often low in microbial density. Research conducted by Shin et al. (200 l) on sequencing batch reactors (SBRs) confirmed char large amounts of EPS had a negative effect on the compressibility of sludge. The carbon to nitrogen ratio (C:N) has been recognised as a sign ifi cant factor

Excessive polysaccharide production in response to high carbon concentrations.

responsible for influencing polysaccharide composition of microbial floes (Stover, 1980; Veiga et al. 1997; Durmaz and Sanin, 200 I; Liu and Herbert, 2003) . When the C:N ratio is high, carbon utilisation by the microbial population can shift to excessive EPS production. This was co nfirmed by Bengtsson (1991) who found char high yields of EPS in pure culture systems were most evident when carbon was present in excess, and/or when nitrogen was limited. Durmaz and Sanin (200 I) reported similar observations when varying C:N ratio in a laboratory-scale activated sludge reactor. Preliminary Investigation

Bolivar, Glenelg, and Christies Beach municipal wastewater treatment plants (W\XITPs) are located across metropolitan Adelaide, South Australia. Unlike Bolivar, the AS plants located at Glenelg and Christies Beach WWTPs are integrated fixed film activated sludge (!FAS) processes. Despite these differences, all plants were operated in a similar manner. Each was operated in a pre-denirrificarion-

Journal of the Australian Water Association


AUGUST 2008 79

technical features refereed paper

nitrificatio n configuration, with recycle (100-300%) ro feed nitrate back ro the front of the plant. Sludge age was typically maintained between 4-8 days and mixed liquor suspended solids (MLSS) was maintained between 2500 and 3000 mg/L. A desk-cop study of che seccleability of mixed liquor sampled from these AS planes showed that bulking sludge, as defined by a sludge volume index (SVI) greater chan 150 Qenkins et al. 1993), was a regular problem ac both Glenelg and Christies Beach, buc not ac Bolivar (Table 1). Routine microscopic examinations also revealed char filamenrous growth was not always prevalent and therefore was not always the causative facror for the sludge bulki ng. This conclusion highlighted che need for the cause for these sludge bulking episodes ro be fu rther examined. The feed wascewacers for Glenelg and Christies beach AS plant's are carbonlimited and therefore rely on the addition of molasses ro drive denitrification. Unfortunately there are no ' feed-back' systems (i.e. flow-paced dosing or on-line nitrate monirors) co regulate molasses dosing, making the plants susceptible ro over-dosing and high C:N ratios during low flow (see Figure I). In contrast, the feed wascewacer co Bolivar WWTP contains adequate readily available organic carbon, and therefore does not require an external carbon source. Sludge bulking at Bolivar has rarely been observed. It was hypothesised that periodically high C:N during low flows at Glenelg and Christies Beach AS planes may be a causative facror for their sludge bulking. Preliminary investigations at Glenelg WWTP

Table 1. Comparison of sludge settleability between the three WWTP's. (± 1 SD) .


Mean settlers

Max settlers

Mean SVI




188 ± 40 413 ± 214 509 ± 214

366 950 920

79 ± 16 200 ± 113 213 ± 56

123 524

366 366 315

Glenelg Christies Beach

confi rmed the link between carbon dose and sludge setcleability. Data obtained over 107 days of operation showed a significant correlation between SVI and the volume of molasses dosed (Spearman's r coefficient (r,) = 0.71, p<0.000 I; Figure These preliminary results suggest chat the variable seeder performance observed at Glenelg and Christies Beach AS planes may be attributed ro excessive polysaccharide production, owing co high and variable C:N.

Most of che available literature on the effects of carbon and nitrogen on rhe distribution of protein and carbohydrate in ML, and its impact on sludge setcleability, has come from research involving laboratory-scale reactors. Therefore che purpose of this further study was co survey ML coca! carbohydrate and protein concentrations from each of the three full scale WWTP's, and co investigate their affects on sludge sectleabiliry.

Methodology Total carbohydrate and total protein analysis

ML was sampled from the aeration zones of each plane, chilled during transport and analysed the same day. 100 µL aliquots of the mixed 30 270 liquor were analysed in triplicate 240 for total carbohydrate and total 25 0 210 0 protein. Tora! carbohydrate was 0 vi' 180 measured by using the anchrone 20 .. ~ 150 z method, using glucose as the 0 ~ 120 0 15 "l standard (Hanson and Phillips, z u::: 90 1981). Protein composition was 60 measured by using the Bradford 30 method, using Bovine Serum 0+ - - - ~ - - ~ - - ~ - - + 5 Albumin as the standard 750 1000 1250 500 1500 (Bradford, 1976; Hanson and lime (24h) Phillips, 1981) . Results are represented as mg of carbohydrate Figure 1. Relationship between hydra ulic flow (,,I or pro rein per g of MLSS. and sCOD: NO3-N ratio (o) observed i n the Cl)

denitrification zone of Glenelg's AS plant. Shaded region represe nts period of molasses over-dosi ng. 1Recommended maximum COD: NO3-N limit for pre-denitrification-nitrification configured biological filters (Rother a nd Cornel,


80 AUGUST 2008 Water

measuring rhe depth of seeded sludge in mm after 30 minutes. SVI was used co characterise che effectiveness of sludge settling. SVI is the volume in millilitres occupied by 1 gram of sludge after 30 minutes of settling in a I litre cylinder (Liu and Herbert, 2003) and was calculated using the following equation:

2) .

Full-scale Study

Sludge settling and sludge volume index (SVI)

Sludge settling was determined by pouring 1000 mL of mixed liquor into a 1-litre glass cylinder, and

Journal of the Australian Water Association


SVI = Sludge seeding (mm)*l000/MLSS (mg/L) Real-time polymerase chain reaction (PCR)

Quantification of total EUB bacteria (members of the domain Bacteria) was undertaken, using Real-time PCR, by the Australian Water Quality Centre, Bolivar, South Australia.

Results and Discussion Comparison of ML carbohydrate and protein concentration between WWTPs

Figure 3 compares the mean carbohydrate and pro rein concentration of ML sampled from Bolivar, Glenelg, and Christies Beach WWTP, monitored over a period of three months. Mean MLSS carbohydrate concentrations at Glenelg and Christies Beach WWTPs were similar (AN OVA p>0.05), however they were significantly greater chan observed at Bolivar WWTP (p<0.001) . Similarly to Liao et al. (2001) and Sponza (2003), chis study has found that the carbohydrate concentration of Bolivar, Glenelg and Christies Beach ML reflected available carbon concentrations. The higher 700 600 500

5 Cl)



400 300



200 100


0+----~-~--~--~ 40

50 60 70 80 Molasses dose (UML of influent)

Figure 2. Relations hip between SVI and molasses dose, observed at Glenelg WWTP (Dashed lines represent 95% Cl).

technical features refereed paper

carbohydrate concentrations measured at Glenelg and Christies Beach were indicative of excess available carbon, which was most likely che result of over-dosing molasses during low flows. In contrast, rhe lower ML carbohyd rate concentrations identified at Bolivar were indicative of a lower and stable C:N, which complemented rhe better sludge seeding performance.

7.5•10 11

750 600

6.0•10 11

m C CD -I

4.5•10 11



3.0•10 11

1.5•10 11


~ !!!.




0 <ii



~ El·


forces ofEPS (Liao et al. 2001). In addition, a negative correlation between: (i) ML protein concentration and SVI (r, = -0.67, p<0.0001); and (ii) protein co carbohydrate ratio and SVI was identified (r, = -0.59, p<0.0001). These results confirmed char sludge with better seeding characceriscics typically had higher protein and lower carbohydrate concentrations. In addition to poor seeding and bulking, EPS can contribute to foaming. T oral carbohydrate measured in rhe foam ac Glenelg WWTP during a foaming event represented 40% by weight of foam mass.

0 ..L...1.£.al__JJ__L__.J£L.JI..__JL_..J._....J..L..£11....-'L-L.J.. 0.0 From Figure 3, it can be seen Glenelg Christies Beach Bolivar char rhe MLSS protein concentration at Bolivar was WWTP significantly greater than identified Total EUB bacteria in Glenelg (ANOVAp<0. 001) and Total protein Total carbohydrate Christies Beach ML (p<0.00 1). Ir can also be seen that the Figure 3. Comparison of MLSS total carbohydrate, protein Ir is important co note char the employment of real-rime PCR and EUB bacterial concentrations between WWTP's. Error average daily dose of molasses at analysis co quantify coral EUB bars represent mean values ± 1 SD. Christies Beach and Glenelg was bacterial numbers across WWTPs accurately calculated to provide were comparable co che quantities 0.6 0 the optimum C:N ratio for :;:; of protein in rhe ML (Figure 3). ~ in turn, nitrogen discharge denirrificarion; Similar co rhe protein concenrracions, ~ 0.5 limits have been consiscenrly met. T he results showed Glenelg and Christies "''O>, .c 0.4 inference was char excess carbon was Beach ML was lower in coral EUB .8 being dosed at night, during low flows. bacterial numbers, in comparison with 0.3 ~ These periodic rime intervals of feast and Bolivar. In addition, rhe ratio of protei n C '§ 0.2 famine can also stimulate bacteria co to carbohydrate was highest at Bolivar, in ea. produce more extracell ular co mparison co Glenelg (p<0.001) and (/) 0.1 (/) polysaccharides (Liu et al. 2003). Christies Beach (p<0.001; Figure 4). ...J ::;; 0.0 Therefore, the higher carbohydrate and These fin dings are in parallel to Sponza Glenelg Christies Beach Bolivar lower protein composition of ML at (2003), who reported low protein WWTP Glenelg and Christies Beach !FAS planes concentrations fro m activated sludge floes may also be accributed to fl uctuations in with high carbohydrate content. Figure 4. Comparison of protein : C:N ratio caused by the diurnal flow The differences in ML protein and ca rbohydrate ratio between W WTP' s. patterns. carbohydrate co mposition between each Erro r bars represent mea n values ± 1 SD.

= = =

WWTP appeared to correspond co C:N ratio. For rhe Bolivar AS plant, which operated u nder a steady and lower C:N , rhe higher protein to carbohydrate ratio suggests char carbon urilisation by rhe microbial communiry was driven more cowards biomass synthesis than EPS production. In contrast, fo r rhe Glenelg and Christies Beach !FAS planes, the distributio n of carbohydrate and protein suggest char carbo n utilisation was driven more cowards che copious production of EPS; an indication that readily available carbon was in excess. These results support the findings of Durmaz and Sanin (2001) and Wu eta!. (1982), who reported char a shift from low co high C:N ratio caused a significant increase in carbohydrate and a decrease in protein concentration of AS floes. Relationship between ML carbohydrate and protein composition with sludge settling characteristics

Results from chis scudy confirmed char ML total carbohydrate were highest in

Conclusions wastewater containing high concentrations of available ca rbo n. Since EPS concentration has been reported to infl uence sludge secrleabili cy (Shin et al. 2001; Liu and Herbert, 2003), the infl uence of ML carbohydrate concentration on SVI was also investigated. This study fo und a sign ificant relationship between ML coral carbohydrate concentratio n and sludge serdeabiliry. A positive correlation between SVI and carbohydrate concentration was identified (r, = 0.75, p<0.0001 ), where carbohydrate concentrations increased with SVI (Figure 5.a). A similar correlation was observed between sl udge settling (mm) and ML carbohydrate concentration (r, = 0.69, p<0.01; Figure 5.6). These correlations indicate the importance of EPS on sludge secdeabi liry, where high carbohydrate concentrations had a negative effect on che compressibiliry of the ML. T he poor seed ing could be explained by che sreric

Results from chis scudy have confi rmed char observations co ncerning sludge properties and sludge serrleabi licy previously identified in well-controlled laboratory studies also occur in fu ll-scale AS planes. Specific conclusions from this study are as follows: • ML exposed co high and fluctuating C:N was typically low in protein and high in carbohydrate concentrations, which suggested chat bacterial carbon urilisarion shifted from synthesis of cellular biomass, co EPS production when carbon was in excess. • Bulking sludge identified at Glenelg and Christies Beach IFAS planes was attributed co high ML carbohydrate concencracions. Mean ML carbohydrate concencracions of Glenelg and Christies Beach plants were significantly greater than char observed at Bolivar, with che likely cause being the over-dosing of molasses ac these sires during periods of

Journal of the Australian Water Association


AUGUST 2008 81

technical features refereed paper

SY! was most likely caused by excessive polysaccharide production in response to unnecessarily high carbo n concentrations.



(J) ~ 700


• Results from this study suggest that better sludge setrleabili ry could be achieved by improving rhe control of molasses dosing, e.g. by flow or load pacing.

-2' 600 Cl






~ .c 300

.8 rou

The Authors



At rhe rime of this investigation Ben van den Akker was a Process and T ech nical Advisor with United Water Internacional, and is now a research fellow with Flinders University's Research Centre for Coastal and Catchment Environments. Helen Martyn is a Process and Technical Advisor, Uwe Kaeding is a Technical Support Manager, all with United Water Inrernarional, which is the operator of the plants described in this study. For correspondence, emai l ben.vandenakker@uwi.com.au



ro 100


o------~~-~---~ January - April 2007


50 100 150 200 250 300 350 400 Sludge volume index

B) (J)

500 0

~ :a 450

~ 400 _§_ ~



-c >,


.c 300 0










,0' /





___ .,,......,,..... -··





March - May 2008

200-t---r---r---r---r---r---i 200 300 400 500 600 700 800 Sludge settling (mm)

Figure 5. Relationship between ML total carbohydrate concentration and sludge settleability measured as: (a) sludge volume index; and (b) slu dge settli ng. Symbols denote: (o) Bolivar WWTP; (o) Glenelg WWTP; and (,:,) Christies Beach WWTP. (Error bars represent mean values ± 1 SD of 3 determinations; Dashed li nes represent 95% Cl).

Bengtsson, G . ( 199 I). Bacterial exopolymer and PHB production in flucmat ing ground-water habitats. FEMS Microbial. Ecology., 86, 15-24. Bradford, M.M. (1976). A rapid and sensitive method for the quantification of microorganisms quantities of protein utilising the principle of protein-dye binding. Analytical Biochemistry., 72, 248-254. Durmaz, B. and Sanin, F. D. (200 1). Effect of carbon ro nitrogen ratio on the composition of microbial ext racellular polymers in activated sludge. Water Science and Technology. , 44(10), 221-229. H anson, RS and Phillips. JA. (I 981) . C hemical composition. In Manual of

low flows. Low ML carbohydrate concentrations at Bolivar were indicative of a lower, and more srable C:N ratio. This observatio n reflected better sludge settling characteristics at Bolivar. • The concentration of ML protein and carbohydrates influ enced the compressibility of sludge measured as SY!. SYI correlated well with ML carbohydrate, and protein concentrations. Results confirmed that ML high in carbohydrate and low in protein concentrations had the most negative effect on the compressibility of sludge. • Statistical analysis of historical plant performance data identified a significant relationship between molasses dose and SY!. This study confirmed that the high

82 AUGUST 2008


Methods .from General Bacteriology. P. Gerhardt, R. Murray, R. Costlow, E. Nester, W. Wa, N. Kneg and G. Briggs-Phillips (Eds). Academic Press, London, United Kingdom: 333. Jenkins, D. , Richard, M. G., and Daigger, G. T., Eds. ( 1993). Manual on the Causes and

Control ofActivated Sludge Bulking and Foaming, 2nd ed., Lewis Publishers, Boca Raton, London, Tokyo, pp. 11- 13. Liao, B. Q. , Allen, D . G ., Drappo, I. G., Leppard, G . G., and Liss, S. N . (2001). Surface properries of sludge and their role in bioflocculation and setrleability. Water Research, 35(2), 339-350. Liu, Y. F., Herbert, H. P. (2003). Influences of extracellular polymeric substances (EPS) on Aocculation, seeding, and dewatering of activated sludge. Critical Reviews in

Environmental Science and Technology., 33(3), 237-258.

Journal of the Australian Water Association

Metcalf and Eddy. (2003). Wastewater engineering: treatment and reuse, 4th ed., McGraw-Hill, New York, USA. Rother, E., and Corne!, P . (2006). Potentials and limits of a pre-denitrification/ nitrification biofilrer configuration for advanced municipal wastewater treatment. International Water Association, Biofilm Systems VI conference proceedings, Amsterdam 24 - 27 September., 179-188. Sezgin, M. (1982). Variation of sludge volume index with activated sludge characteriscics. Water Research., 16, 83-88. Shin, H.S., Kang, S.T., Nam, S.W . (2001) . Effect of carbohyd rate and protein in the EPS on sludge setrling characteristics. Water Science and Technology 43(6), 193-

196. Sponza, D.T. (2003). Investigation of extracellular polymer substances (EPS) and physicochemical properties of different act ivated sludge floes under steady-state condit ions. Enzyme and Microbial Technology., 32, 375-385. Srover, E. L. (1980) . Starr-up problems at a plant treating food-p rocessing wastewater.

journal of Water Pollution and Control Federation., 52(2), 249-256. Veiga, M.C., Jain, M.K. , Wu, W .M., Hollingsworth, R.l., Zeikus, J.G. ( 1997). Composition and role of extracellular polymers in met hanogenic granules.

Applied and Environmental Microbiology., 63, 403-407. Verstraete, W., and van Yaerenbergh, E., Aerobic activated sludge. In Biotechnology, Vol. 8, Rehm, H.J. , and Reed, G ., Eds., VCH Yerlagsgescllschafr, Weinheim, Germany, pp. 44-102, I 986. Wilen, B.M ., and Balmer, P. ( 1999). The effect of dissolved oxygen concentration on che strucrure, size and size distribution of act ivated sludge floes. Water Research., 33(2), 39 1-400. Wu, Y. C., Sm ith, E. D., and Novak, R.(1982) . Filterability of activaced sludge in response to growth conditions. journal

of Water Pollution and Control Federation., 54(5), 444-456. W uertz, S., Bishop, P. and Wilderer, P. (2003). Biofilms in Wastewater Treatment, An Interdisciplinary Approach., !WA Publishing, Cornwall, UK.

Water Advertising To reach the decision-makers in the water field, you should consider advertising in Water Journal, the official journal of Australian Water Association. For information on advertising rates, please contact Brian Rault at Hallmark Editions, Tel (03) 8534 5000 or email brian.rault@halledit.com.au

THE SOLAR SLUDGE DRYER: MAROOCHYDORE'S FIRST YEAR M Thomas, G Bick, P Jensen, K Hartley, R Walpole Abstract A newly designed AquaSol™ model Solar Dryer for sludge was commissioned in February 2007. The results from irs firs t fu ll year in operation are presented along with che outcomes of an optimisation program for both the Solar Dryer, and for rhe preceding sludge treatment processes. The typical evaporation rares were 5.4 mm/d during summer and 3.6 mm/d during winter, which were similar ro the ambient environmental pan evaporation races. As a result of the additional drying achieved by rhe Solar Dryer, che annual biosolids weight transported from the plant was approximately halved. Even better outcomes are expected in future years following implementation of rhe optimisation program.

Introduction Maroochydore Sewage Treatment Plant (STP) is located on the Sunshine Coast about 100 km north of Brisbane, Q ueensland. The plant currenrly serves an equivalent population of about 95,000 people and rrears an average dry weather flow of 24 ML/d. T he plant has recenrly completed a capital works project ro expand its capacity and enhance the treatment quality, including rhe insrallarion of rhe Solar Dryer. Biosolids transport and disposal costs are a major component of rhe operating cost for any treatment plant, and Maroochydore STP is no exception; prior ro commissioning of rhe Solar Dryer they amounted ro more than $500,000 p.a. Consequently, improvements in rhe dryness of biosolids have rhe potential ro provide significant savings in operating costs. The

Dewatered sludge ca ke is spread across the width of the bed using an automatic slewing conveyo r.

Australian climate is ideal in many locations ro provide cosr-effecrive biosolids drying, and several Solar Dryers are now in operation: e.g. Caboolture (Nathan et al, 2004) and T oowoomba in Q ueensland and the Morni ngron Peninsula, Vicroria.

Description of Maroochydore STP The quantity and characteristics of the biosolids generated by a wastewater treatment plant depend on the wastewater characteristics and the treatment process. Consequently rhis sectio n provides a brief description of rhe liquid treatment process, a more derailed description of the solids treatment process, and a summary of the wastewater load treated. Maroochydore STP utilises a biological nutrient removal (BNR) process that is supplemented with al um and methanol dosing ro achieve effluent rota! N < 3 mg/L and rota! P < 1 mg/L. Secondary effluent is treated by IO µm disk fil rers and ultraviolet

Statistical Criteria

Flow COD

50%ile 50%ile

BODs Suspended solids Total nitrogen Total phosphorus

50%ile 50%ile 50%ile 50%ile

The average sludge production processed th rough the Solar Dryer is 5 T/d of dry so lids, which comprises anaerobic digested sludge derived from the combined chemical-biological nutrient removal process. The sludge consists of a mixture of fe rmented primary sludge, waste activated sludge and alum sludge. The solids treatment process stream at Maroochydore STP includes: • Fermentation of primary sludge. • Thicken ing of waste fermented sludge using rotary screen thickeners (RST) . • Th ickening of waste activated sludge (WAS) using dissolved air flotation (DAF). • Anaerobic digestion of thickened primary and seco ndary sludge. • Dewatering of digested sludge using a centrifuge, with a belt press providing stand-by dewatering capacity.

Table 1. Maroochydore STP Influent Characteristics - M ar 07 to Feb 08 . Parameter

(UV) disinfection. Approximately 17% of dry weather flows are furthe r disinfected with hypochlorite ro produce Class A water for reuse, and the remaining effl uent is discharged ro the Maroochy River.


Mass Load


24.8 ML/d

380 mg/L 140 mg/ L 200 mg/ L 48 mg/L

10.3 T/ d 3.5 T/d 5.1 T/d l. 14T/d

7 mg/L

0.16 T/d

• Further drying of dewatered sludge cake using the Solar Dryer. • Reuse of biosolids via application ro agricultural land.

Halves the weight of biosolids.

Journal of the Australian Water Association


AUGUST 2008 83

technical features

The sludge was generated from wastewater of predominantly domestic origin, and a summary of the key influent characteristics and loads are shown in Table 1.

The Solar Dryer T he Solar Dryer was designed and fabricated in Australia by Aquarec Maxcon Pry Led and was the fi rst fu ll-scale installation of rhe AquaSol™ model. The Solar Dryer combines some aspects of a traditional sludge drying bed with certain aspects of a greenhouse in order to maximise evaporative drying of biosolids. Evaporation is maximised by rhe combination of rhe translucent roof and walls char exclude rain without blocking solar radiation or air flow, and by a mechanical sludge turner char repeatedly turns over rhe sludge thereby exposing rhe bu ried layers to the air and steadily advances rhe cake from entry end ro the exit. Consequently rhe su rface area required to reach a certain level of d ryness is red uced, and rhe process footprint is more compact than a traditional drying bed. The inclusion of mechanical turners, and in-load and om-load conveyors results in a mechanised , automated process wirh reduced labour requirements. Finally, the Solar Dryer operates as a mechanised, continuous "plug flow" process, as opposed to a "batch" p rocess in typical sludge drying beds. The dimensio ns of rhe Solar Dryer beds ar Maroochydore STP are: • Number of beds = 2 • Length o f each bed = 106 m (effective length in use)

The rotating blades turn the biosolids over and steadily move it onwards.

Out-loading from the beds is carried our by the turners pushing dried biosolids onto a conveyor which loads directly into a truck. T here is no hopper for storage of the dried biosolids, and instead they are allowed to stock-pile to a depth of about 500 mm within the final 10 m length of the beds. Unfortunately, co mposting occasionally occurs within these stock-piles and this can generate some odours during truck loading. The nearest neighbours are only 30 m away and we have received some odour complaints. So now we sched ule any truck loading activities for the afternoon when the on-shore winds have typically scarred blowing.

• Width of each bed= 13 m


• Height of side walls = 0.6 m

All analyses were performed by Maroochy Water Services laboratory using Standard

• T oral Solar Dryer bed area = 2756 m2

75 70 65 60


~ 50

!. 45

I: 30


25 20


.-..._.......- -

- -• ...-


./° ... _


15 10

~ 0

I tli 0







~ 0

1l ! I I t I 0

Ceroifuge Cake



0 -







Final Biosoids

Figure 1. Solar Dryer input and output solids dryness.




Journal of the Australian Water Association

i J~ 0



Methods (2005) with some minor in-house variations. The laboratory is accredited under ISO 9001 and follows the proced ures of AS/ISO 170 25, although it is not presently accredited by NATA. The derails of the procedures and calculations adopted during the optimisation programme and for evaluation of the data are discussed in the relevant sections of this paper.

Solar Dryer Results T he total solids dryness (%TS) results fo r the dewatered sludge cake and the dried biosolids that were inpllt and output from the Solar Dryer are shown in Figure 1. Prior to commissioning of the centrifuge, sludge dewarering was carried out usi ng a belt p ress, and it continues to be used as the stand-by dewatering unit. T ypical dewarering performance of the belt press was 13-1 4%, compared to 18-21 % for the centrifuge. The "ultimate" drying characteristics of rhe Maroochydore sludge would appear to be 70% d ryness, and 60% dryness was readily achieved during summer. The typical solids dry ness profiles along the length of the Solar Dryer beds are shown in Figure 2 for "summer" (Oct to Feb) and "winter" (May to Aug) conditions. The sample locations were at approximately equal intervals of 20 m along the length of the bed, except that Location 1 was 5-10 m from rhe input from rhe centrifuge cake conveyor. The results plotted are rhe average values obtained from weekly samples collected over a period o f several months. The seasonal influence on drying was apparent and a more detailed assessment of evaporation rates was carried o ur. Two

technical features

methods were used to evaluate th e evaporation rares from the Solar D ryer:

70 ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~


1. Diffe rence in dryness concentratio n (%T S) between 2 locations alo ng rhe bed.



2. Mass balance of biosolids wet weight between inlet and ou tlet. T he calculation fo rmulae, assumptio ns and limitations of these two methods are summ arised in Table 2, and rhe season al evaporatio n results char were determined are shown in Figure 3 and Figure 4, and summarised in Table 3. T he variability in evaporation rares along the length of the bed was minimal fo r the summer period (Figure 3), ind icating ch ar evaporatio n occurs uniformly from rhe entire bed area. The apparent variability in rhe win ter evaporation data, e.g. between

25 20


. . .-----



Location 1

Location 2

Location 3

Location 5

Location 4

Slrnple Location •

- -Vllrner


Figure 2. Typical solids dryness profiles along the length of the Solar Dryer bed s.

Table 2. Solar Dryer Evaporation Rate Methods. Method 1: Difference in % Dry Solids Concentrations Calculation

Evap. = DSmass 1100 / %TS; - 100 / %TS;.1) / (N x W x d) x 1000

Where: • Evap. = average evaporation rate (mm/ d; i.e. m3/ m2/ d x 1000) between sample locations " i" and " i+ l " • DSmoss = mass of sludge dry solids processed daily (T/ d) • %TS 1 = total (dry) solids concentration at sample location " i" (%TS) • N = number of Solar Dryers in service; normally = 2 • W = width of each Solar Dryer bed = 13 m • d = distance (m) along length of bed between locations " i" and "i+ l " Assumptions

• DSmoss was assumed to be constant and equal to the long-term average value of 5 T/ d. • Process was assumed to be at pseudo-steady-state. Biosolids at Location 1 on a given day have a typical transit time of 2 days to get to Location 2. However, samples from multiple locations along the bed are all collected on one day and are assumed to be representative of a given "batch" of biosolids as it moves along the bed . • Grab samples were collected weekly and results from a period of several months were averaged to dampen affects of daily variabi lity. Grab samples were assumed to be representative.


• Mass of dry solids processed each day is variable, and if it systematically trends above or below the average for an extended period of time then this will result in errors with this method. • Variability in sample results indicated that the biosolids within the bed were not homogenous, and it was difficult to collect a representative 10 g sample from a drying bed containing tonnes of biosolids which were not homogenous.

Method 2: Mass Balance Between Inlet and Outlet of Solar Dryers Calculation


=(Mw.1 dewatered - Mwet trucked) /

(N X W X L) / dmonth X 1000; Mwet dewatered

=Vdewatered X %TSdigester / %TS,entriluge

Where: • Evap. = average evaporation rate (mm/ d; i.e. m3/ m2/ d x 1000) from the entire bed over the period of the month • Mwel dewotered = total wet weight of sludge that wos dewatered during the month (wet T/ month) • Mwet trucked = total wet weight of biosolids that were transported from site during the month (wet T/ month) • Vdewotered = total volume of digested sludge that was dewatered during the month (m 3/ month) • %TS; = total (dry) solids concentration at sample location "i" (%TS) • N = number of Solar Dryers in service; normally = 2 • W = width of each Solar Dryer bed = 13 m • L = length of each Solar Dryer bed = 106 m • dmonth = number of days in the month Assumptions

• Process was assumed to be at pseudo-steady-state, such that the inventory of biosolids in the beds remains constant and biosolids storage was not accumulated or depleted. This was not always true, for example at the end of December 2007 when the centrifuge was out-of-service and no sludge was input to the Solar Dryer, biosolids continued to be out-loaded and the inventory was depleted.


• Wet weig ht of dewatered sludge was calculated from ( 1) the volume, which was accurately measured by flowmeter; (2) the digester solids concentration , which was not very variable because the digesters were well mixed; and (3) the centrifuge cake dryness. • Wet weight of dried biosolids transported from site was determined from truck weigh-bridge dockets, which were expected to be accurate.

Journal of the Australian Water Association


AUG UST 2008


technical features

Locatio n 3 and Location 4, appears random and seems co be an artefact arising from che difficulty in collecting representative samples as discussed in Table 2 with the assumptions and limitations. Nevertheless, the measured evaporation races were in reasonably close agreement with the longterm climatic data for pan evaporation races from nearby weather stations (Bureau of Meteorology, 2008) (Figure 4). However, ir m ay be expected chat with a roof co keep out che rain then che evaporation races should be higher than environmen tal evaporation. Indeed che typical evaporation races of 8 mm/d reporce<l for Caboolrure (Nathan et al, 2004) were higher than at Maroochydore and are consistent with chis con cep t. Comparison of local Bureau of Meteorology wind data with chat fro m a weather station installed on the Solar D ryer indicated that che local topography appears co be significancly decreasing wind speeds, and consequencly also decreasing evaporation. T he wet weigh t of biosolids transported from che plane has roughly halved since the Solar Dryer was comm issioned (Table 4), which corresponded co a saving in biosolids transport coses of $270,000 p.a . T he decrease can parcly be acrribuced co che better dewatering performance of the centrifuge relative co the b elt press, but moscly due co the additional drying obtained in the Solar Dryers. The results in Table 4 also illustrate several points with regard co data quality: • The dry weight o f sludge dewacered and biosolids trucked agreed within ± 10%, and chis is co nsidered co indicate go od data quality given the difficulties in obtaining mass b alance closure on com plex fu ll-scale planes subject co normal variability. • Prior co commissioning of che Solar Dryers the wet weight of sludge dewatered and biosolids transported should have been eq ual and they agreed withi n ±10%, which validates the results' accuracy. • The apparent increase in specific sludge prod uction from lower than normal values in 200 5 and 2006, co more typical values in 2007-08 is likely co be an error. Similarly che apparent decrease in influent CO D and solids loads between 2005-06 and 2007-08 is also likely co b e an error. Prior co February 2007 che influent samples were grab samples chat included recycles, which appear co have been non-representative samples with "high" results.

Operational Observations on Solar Dryers This Solar D ryer was che first full-scale inscallacion of the AquaSol™ model. 86 AUGUST 2008


9 8




.!! 5 ~


= = =



[ i=



3 2


Location 1

Location 2

location 3




Location 5





Figure 3. Solar Dryer evaporation rotes calculated by Method 1. 9 ,----,,,=.=,=,===,===.------ - - - - - - - -- - - -- - -- - -------i




ISolds 1--ay

acc...-ed in bed


_. --, \


·-- ·-- :: :v - - :-;f _::.~~

-: -~=;·y ·· depleted from bed



. depleted from bed


QL - - - - - - - - -' - - - - - -L-- - - - - -======="--- ---' _, § z ~ •


B<isbaneAirport 1951-1986

- Nimlour DPI 1965-2007


Brisbane BoM 1986-2000

Figure 4. Solar Dryer evaporation rates calcu lated by Method 2. Consequencly ic was som ewhat like an advanced prototype and during ics initial months of operat ion it was susceptible co various m echanical, electrical and controls problems. The operations staff provid ed valuable input co the supplier in rectifying many issues, and ic is now m ore rel iable. The followi ng fe edback from che operators may be useful fo r improvem ents co future installations: • The control program was revised with input from the operators' observation s. Software and control logic should not cry co be "coo smart", b ecause simple solutions are sometimes th e most robust. • D ewacered slud ge with small granule sizes and crumbly appearance seemed co produce the best drying, presumably due co its higher specific surface area. T herefore, centri fuge operation should consider this in add ition co targeting m aximum dryness.

Journal of the Australian Water Association

Table 3. Seasonal Solar Dryer Evaporation Rates. Parameter

Method l

Method 2

Summer Winter

5.6 mm/d 3.7 mm/d

5.2 mm/d 3.6 mm/d

• Bel t conveyors should be preferred over screw conveyors co maintain sm all granule sizes, and inclined conveyors should be at gradien ts of about 20 ° to prevent sl udge roll ing back down ch e conveyor and " balling-u p" . • The co nveyors were installed with sheermetal guards around chem, which made access for cleaning and maintenance difficult, and allowed any spillage of biosolids off the belts co accumulate inside the guards around the rollers and belt. Consequencly the guards were replaced with a coarse, open mesh.

technical features

sludge treatment • If sludge wetter than about 17-18% T S was input to the Solar Dryer then rhe sludge seemed to fo rm large, gluggy, pasty lumps char didn't dry well, similar to what was observed at Caboolrure (Nathan et al, 2004). • Ar the ocher end of the bed, if rhe biosolids dried to more than about 60% T S then dust was an issue, particularly during ou r-loading of biosolids into a truck. • The Maroochydore Solar Dryer did nor incorporate a hopper or bin fo r storage of the final biosolids, instead it was accumulated in a stockpile at rhe outlet end of the bed. On occasions the temperatu re in the stockpile increased to 60°c and an ammonia-type odour was generated. Therefo re it appeared that composting was occurring in the bed. • Further in vestigations of this composti ng process are required to understand whether the hear generates any significant benefits in terms of additional drying and/or pathogen die-off, and how best to manage the risk associated with odour generation. • The weather station installed at the Solar Dryer indi cates that th e local topography partly shelters the installation from the prevailing winds, and chis aspect of sire layout should be taken into co nsideration to maximise evaporation.

Table 4. Maroochydore STP Annual Sludge Production and Disposal. Annual Period

Dewatered sludge - dryness Dewatered sludge - dry weight Dewatered sludge - wet weight Final biosolids - dryness Final biosolids - dry weight Final biosolids - wet weight Influent COD load Specific sludge prod'n vs COD Influent TSS load Specific sludge prod' n vs TSS

Jan 05 to Dec 05

Jan 06 to Dec 06

Mar 07 to Feb 08

13.5% TS 4.5 dry T/ d 33.2 wet T/ d 13.5% TS 4.1 dryT/d 30.4 wet T/ d 11. l T/d 0.41 kg SS/kg COD 6.3 T/d 0.71 kg SS/kg SS

13.5% TS 4.6 dry T/ d 34.2 wet T/ d 13.5% TS 4.5 dry T/d 33 .2 wet T/d 11. lT/d 0.41 kg SS/kg COD 6.6 T/ d 0.70 kg SS/kg SS

19.5% TS 5.0 dry T/ d 26.2 wet T/ d 38.7% TS 5.5 dry T/d 15.2 wet T/d l0.3T/d 0.49 kg SS/kg COD 5.1 T/d 0.98 kg SS/kg SS

Conclusion The key concl usions and important lessons learnt from the fi rst year of operation were: • T he Solar Dryer was capable of drying the centrifuge cake from 20% T S to a final biosolids dryness of 60% TS in summer when operated with a residence time of I 0 days. • Solar Dryer pe rformance was seaso nal. T he typical evaporation rates were 5.4 mm/d du ring summer and 3.6 mm/d during win ter, which were similar to the

ambient environmental pan evapo ration rates. • Solar Dryer perfo rmance was strongly dependent on climatic cond itions and insensitive to a range of operational controls i.e. the mass of water removed by the Solar Dryer was largely outside the operator's control and for given climatic conditions the water removed by the Solar Dryer was essentially constant. Consequently, the final biosolids dryness was higher and the overall system dewatering perfo rmance was better if rhe

Optimisation of Upstream Solids Stream Processes Evaporation depends on seasonal climatic conditions and cannot be controlled by the operator. Furthermore, modelling indicated that Solar Dryer performance was relatively insensitive to variations of the available operational controls withi n a practical range (Gloag, 2007). T herefore, optimisation of the Solar Dryer cannot be done in isolation and must inherently go hand-in-hand with optimisation of the entire sludge treatment process, particularly the dewatering centrifuge. Consequently, a systematic programme was implemented to address the performance of each unit process within the solids stream , and the outcomes are summarised in T able 5.

The Advanced Utility Solutions (AUS) team has over SO years experience In the utility location business AUS • a relawely , _ ~ and has been fonned to team up with one of the world leaden in udlty locauon.

SebaKMT Is a German company who have been suppfying the utility locating market since 1951. The teaming of the SebaKMT companies and AUS will offer a vory compeutM! range of products to the uulity locating market here In Austraf,a, In the range we have water leak detect1011 equ,pmont from Seba. Pipe & Cable locatorS from Metrotech & YIYax and Inspection Camera~

from Ywax. This ~ a comprehensNe and dynamic range which Is second to none for quality and value for money.


Easyloc ~ a wekorne econom,cal and technk:al solution to uubty locating Incorporating a rugged housing, automatic dighal display and high qualrty features for greater productMty at lower operaoonal costs

Solar Dryer Cost Estimates The estimated capital and operati ng coses fo r the Solar Dryer are: Capex = $2.25 million. Opex: • Electrical power = 125 kW h/d, which corresponds to about $6,000 p.a. • Labour = 2 to 3 man-hours/d, which corresponds to about $20,000 p.a. • Maintenance= typical long-term coses have yet to be determined.

HLSOOO The MW HLS Is a pmfe<slonal and vory flexible mini water leak detection unit ,.;th both visual and audible ,nd,cation of the leak and with wireless uansmfsslon of the leak noise. It belongs In the jacket pocket of every water spedabst and fitter for an Instant. on the spot leak monrtonng check



t.w1 (VI\I\X )

The VCam Digital b the third generation of CCTV inspect,on

camera designed and devoloped by Vwax. As With all Y,vax products - the design has been heav,ly influenced by our customers and the expenence of extensive field use.


~ r


This modern device from the Hydrolox fam,1y for professionab Is your fi~t choice. With the HLSOOO and HLSOO. you can find leaks quickly, eas,ly and rehably. The equipment uses modern digital signal process,ng technology (DSP) to dearly recognise the leak sound, even when there is a lot of noise in the enwonment. The sounds are not just perceived audibly, but are also displayed graph,cally USlng ultra-modern dual segment analysis (OSA) technology. for the user thts means more rehabu1ty when locating water losses.

I ~-



vl ocPro


The v\.ocPro represents a new generation of pope & cable locator ubltz,ng the latest technology and raises the bar in underg<Ourld locating, while stiff pmv,d,ng a simple to operate Instrument.

Email salesOsebakmt-aus.com

Sydney 02 8005 0006 Melbourne 03 9016 9211


Journal of the Australian Water Association

Porth 08 9467 5744

Brisbane 07 3103 2844 Advanced U1~ty Soluoons Pty Ltd


AUGUST 2008 87

technical features

sludge treatment centrifuge cake dryness was as high as possible. Therefore, optimisation of the centrifuge operation was essential for optimum Solar D ryer performance. • As a result of the additional d rying achieved by the Solar Dryer, the annual biosolids weight transported from the plane was approximately halved and the correspond ing saving in biosolids transport coses was $270,000 p.a. Beerer outcomes are expected in future years because the initial year included periods of sub-op timal operation, which was a necessary pare of che "trial and error" learning process during opt imisation.

The Authors

Michael Thomas is Process Engineer (email: mike.chomas@sunshinecoasc. qld.gov.au), Glen Bick is Assistant Operaror, Paul Jensen is Technical Supervisor, and Ron Walpole is Treatment Planes Manager, all with Sunshine Coast Regional Council, Maroochydore STP, Maroochydore, Qld. Ken Hartley is Consulting Engineer and Adj unct Professor, Chemical Engineering, University of Queensland, Sc Lucia, Q ld.

Th e Maroochydore solar dryer.

References Bureau of Mereorology websire (2008), Climate scaciscics for Ausrralian locarions, web address - http://www.bom.gov.au/climare/averages/ rables/cw_040282_All.shcml.

S. Nat han, B. C larke & N. Byres (2004), "Solarmix - co ntinuous biosolids drying", Water: journal ofAWA, Yol.31, No.6, pp.5660, Sepe 2004.

G . Gloag (2007), Maroochydore STP: Process Operations Manual, pare of the O&M Manuals prepared by JWP Pry Led for rhe Maroochydore STP Upgrade Alliance, Feb 2007.

Srandard Merhods (2005), Standard Methods for the examination ofwater and wastewater, 21 sr Edition, prepared and published joinrly by APHA, AWWA and WEF, Washington DC.

Table 5. Solids Stream Process O ptimisation Results. Process Unit

Control Voriable

Optimisotion Result


Scroll torque

Coke dryness increased from 17% to 18% as the torque setting w as increased from 40% to

Feed sludge flow

Coke dryness increased from 18% to 19.5% os feed flow was decreased from the nominal capacity of 3.5 L/s down to 1.5 L/s. The centri fuge feed flow can be minimised by achieving maximum sludge thickening performa nce from the RST and DAF.

Poly dose

C ake dryness increased from 17.5% to 18% as poly dose was increased from 8 to 11 kg/T, with no improvement ot higher doses.

Pool depth

Cake dryness increased from 17.7% to 19.2% os pool depth was increased from 60 mm (minimum setting) up to 70 mm.

Bowl speed

Cake dryness increased from 19.5% to 20.5% as bowl speed was increased from 2400 rpm to 3000 rpm, with no improvement ot the maximum speed of 3200 rpm.

Poly dilution water flow

Coke dryness increased from 19.0% to 19.4% os poly dilution water flow was decreased from 2.2 L/s to 0.3 L/s.

Anaerobic Digestion

Residence time (HRT)

Centrifuge dewotering performance generally improves os the VS fraction of the feed sludge decreases, and maximum VS reduction in the d igesters can be achieved by maximising the HRT. Therefore, optimisation of the digesters focused on achieving the maximum HRT by means of maximum sludge thickening performance for the RST and DAF.

Rotary Screen Thickener (RST)

Poly dilution water flow

Thickened sludge increased from 5.5% to 7.5% as the poly dilution water flow was increased from 1.1 L/s to 2.2 L/s.


Drum speed

Thickened sludge increased from 5 .0% to 6.5% os the drum speed was increased from 25% to 90%.

Poly dose

Optimum poly dose was 1.5 kg/T.

Dissolved Air

Skimmer speed

Thickened sludge increased from 3.5% to 4.3% os the skimmer speed was decreased from l 00% to 50%.

Flotation (DAF)

Solids load

WAS flow control logic was re-programmed to minimise hour-to-hour variability in solids load to the DAF, and to effectively ensure that the air-to-solids ratio was maintai ned above 0.02 kg/kg.

Recycle o i r moss

Recycle flow and soturotor pressure cannot be controlled independently in th is DAF, and an increase in recycle fl ow could only be ach ieved ot the expense of o decrease in soturotor pressure and vice verso. Therefore, the recycle air moss couldn't be readily increased unless both duty and standby recycle pumps were run in parallel.

Poly dose

Poly dosing was not effective, causing o significant deterioration in solids capture and sludge thickening.

88 AUGUST 2008


Journal of the Australian Water Association

DUAL POLYMER SLUDGE CONDITIONING FOR HIGH SPEED CENTRIFUGE J Young Abstract Penrith STP has been operating a high-speed centrifuge since 2003 in order to dewater aerobically digested sludge; it was the firs t high speed centrifuge Sydney Water had in operation. High speed (also called high G) centrifuges are mod ified solid-bowl centrifuges that are designed to produce a drier cake solids. T he in itial results achieved were well below expected sludge cake TSR of 21 %. As pare of the optimisation of the high G centrifuge, the use of a dual polymer conditioning was trialled.

Pre Dose (anionic charge) Polymer Injection Point Waste Activated Sludge I--+ from Bioreactor

Sludge Thickening H Process OAF

Transport 100%

Beneficial Reuse


Aerobic Digestion

Dewatering Centrifuge I+-~ Operation

Th is paper provides an overview of the successful develop ment of the dual polymer dose system.


Sludge Pumping

Cationic charge Polymer Injection Point

Introduction Figure 1. Solids Stream Flow Path. Penrith STP is located at the foot of the Blue Mountains, 45km West of Sydney. It receives The Challenge approximately 23 M L/ day of sewage and this will increase to 38ML/day in 2021. The initial performance of the dewatering The biosolids from the dewatering was poor and well below the expected process are trucked about 200 km west over specification of producing bio-solids sludge cake of 21% TSR using 7kg/dt of cationic the Blue Mountains into the Central West region of NSW and are used 100% for polymer. The centrate was also very dirty. agricul tural and horticultural purposes. The challenge was to create a stronger sludge floe, which could withstand the Stage 8 Upgrade was commissioned torque generated in the high speed between June and December 2003 to centrifuge. double the capacity. It invo lved newly constructed IDALS to treat l 9ML/d ADWF, increased fi nal effi uent filtration capacity and a dewatering process upgrade with the introduction of a high-speed centrifuge, a Westfalia 501 series runn ing at 3500rpm.

The sludge treatment system is illustrated in Figure 1. Sludge thicken ing involves dissolved air flotation (DAF) (Figure 2) and aerobic digesters (Figure 3). The OAF plant is designed to thicken waste from the secondary bioreactor, with the clarified effi uent recycled to the plant and the thickened sludge pumped across to the aerobic digester. The benefits of thickening the sludge for dewatering include: • Allows fo r more storage of solids in the digesters • Longer detention time in digester resulting in improved sl udge stability • Thicker feed sludge to centrifuge for improved dewatering - dryer cake% TS R T he major advantage of aerobic digestion is that it produces a biologically stable end product suitable for subsequent treatment in a variety of processes. The digested sl udge is pre-cond itioned by polymer dosing and pumped to the high G centrifuge (Figure 4).

Initial Results of the High Speed Centrifuge Dewatering Performance

T he dewatering plant was designed to produce 21 % cake at a cationic polymer dose rate of ?kg/dry ton with a solids capture rate of>95%. This was awarded Best Paper at the NSW Operators and En ginee rs Conference.

T he biological trearment at Penrith SI P is activated sludge with biological phosphorous and nurrient removal. The waste sludge from the IDALS is transferred across to the Stage 7 BNR bioreactor. T he extended aeration process has been adopted to allow for natural destruction of solids and a homogenous blend of sludge to be sent fo r thickening

Advances in centrifuge technology have led to the development of the high-speed centrifuges which are

Figure 2. Sludge DAFF thickener.

The aerobically digested sludge had only a weak negative charge.

Journal of the Australian Water Association


AUGUST 2008 89

technical features

sludge treatment the increased strength of the sludge floe and its ability to withstand the high shear

Pre-dose pilot plant In September 2004, the team installed a temporary pre-dose poly system to test the dose rates and compare the impact of dosing the anionic polymer as well as the regular cationic polymer.

Figure 3. Aerobic Digester.

The result was a dramatically improved centrate quality returning to the plant and ultimately producing dryer bio-solids sludge cake. To achieve good cake TSR % and clear centrate, the total polymer dose rate needed to be increased from the expected 7 kg/dt, to around 18kg/dt. W ith continued optimisation the polymer usage has dropped to a current dose rate of 1.5 kg/dt for the anionic polymer and 10.5 kg/dt fo r the cationic polymer, producing >20% TSR and crystal clear centrace recycled to the plant.

Figure 4. High Speed centrifuge.

The Production Team has developed a better understanding of the sludge conditioning process.

designed to produce drier cake solids, but the increased operating torque levels apply h igher shear forces to the feed sludge.

Dual Polymer Sludge Conditioning

The high shear at Penrith STP led to destruction of the sludge floes, Initial results from the centrifuge were poor, with cake 14- 15% TSR and d irty centrate at the specified dose rate (7kg/dts of cationic polymer).

Dual polymers are usually used when the sludge particles have a net electrical surface charge near zero. The sludge at Penrith STP has a very weak negative charge which has led to the need of predose polymer.

The Plant Team worked through the operational parameters that are available for the optimisation of the centrifuge performance for the best actual combi nation.

The anionic polymer is added in an amount sufficient to impart a stronger negative charge to the small floe so that the cationic polymer will be able to form a larger and more stable floe, as illustrated in Figure 5.

• The internal weir heights were varied to test the centrate pool depth; separated liquid is d ischarged contin uously over adjustable weirs. • Torque control set points were tested • A range of di fferential speeds were trialled • Different polymer inj ecting points were used • Slowing the bowl speed of centrifuge was carried out

Small areas of positive charge on the floe enable the anionic polymer to attach. The polymer ratios are crucial to the sludge conditioning process.

cake (Higgi ns, Chen and M u rthy, 200 6) . Figure 6 indicates the difference in the centrate quality.

Bio-Solids Management Efficiency Gains Polymers are expensive and their cost can represent a significant fraction of the operational costs at Penrith STP. The dual polymer sludge conditioning was using 17-18 kg/de when the trials were firs t commenced. The dual polymers were working but optimisation was needed. The Plant Team has steadily reduced to che current dual polymer dose race of 12kg/dt, while achieving 21 % cake TSR and clear centrate, as shown in Figure 7. Achieving higher biosolids cake TSR results in less tonnes of water transported in the sludge cake thus saving money on

Anionic Poly Dosed to

Cationic Poly Dosed to bind with the sludge particle

increase charge

-- -o- -

--0- -----0--

--~ _-_: ~- :Q: Anionic Sludge particles

+ & Cationic Polymer chain

Figure 5. Sl udge Flocculation/Conditioned Sludge - more robust stronger Floes.

To increase the polymer reaction an inline static mixer was installed after the cationic injection point. This along with the bend s and the distance prior to entering che centrifuge allows for thorough mixing and sludge floe formation.

• Varied polymer dose rates were tested • Finally, introduction of a predose polymer. After varied unsuccessful mechanical attempts the result of dosing anionic polymer (Pre-dose polymer) before the addition of regular cationic polymer, improved the dewatering process substantially, the improvement being due to




Centrate Quality The designed requirements for the dewatering included a solids capture rate of >95%. Optimised operation of the dewatering process can be measured by the quality of the cenrrate as well as the dryness (TS R %) of the sludge

Journal of the Australian Water Association

Figure 6. The sample on the left is aerobically digested thickened sludge w ith no polymer added, th e middle sample is a centrote sample, note the high amou nt of solids returning to the plant. The d ewatering process is not optimised. The centrote sam ple on the right is relatively clear with > 95% solid s capture rote, i.e.optimised operation.

technical features

Because the aerobically digested sludge at Penrith STP had only a weak negative charge it was found chat a dual polymer system was required.

20 15

With dual polymer system and the optimisation of the dewatering process the Penrith Plant Team has been able to achieve bio-solids sludge cake of 21 % TSR with > 95% solids capture rate usi ng J 2kg/dc.






ID Polym er Kg/dt •



Biosolids Cake TSR


Figure 7. The graph indicates the rise in cake TSR above 20% and the decrease in total polymer consumption.

transport costs prior to agricultural use (Figure 8).

T he Pen ri th STP Production Team has tailored the operation of the dual poly dose system to ensure that the Bio-Solids produced are 100% su itable for reuse in the agricultural/horticulcural industries. W hile there is co ntinued demand for efficient operations in the management of biosolids, che team has achieved an annual saving in excess of$ 134,000.



The high-speed cenrrifuge applies more G forces in order to squeeze our water from the feed sl udge but a stro nger floe is needed.

Our thanks go to che Centrifuge Ma nufactures, Wescfalia Decanters Pty Led, Ciba Chemicals Pty Led, fo r thei r continued interest and support in

Figure 8: Agricultu ral reuse, pasture imp rovement.

optimising che dewacering process and also the Sydney Water Penrith STP Plant Team for che coordination of che sampling and data analysis.

The Author John Young is Production Officer, Busi ness Services Penrith STP, Sydney Water Corporation, email: john.yo ung@sydneywacer.com.au

References Mathew J. Higgins, Ph.D. and Yen-Chih Chen, Ph.D. Department of Civil and Environmental Engineering Bucknell University. Sudhir N. Murhry, Ph .D., P.E. District of Colu mbia water and Sewer Authority. 2006. Understanding Factors Affecting Polymer Demand far Conditioning and Dewatering. ISBN: 1-84339-726-9

Australian Company - Global Expertise Protube Ultra Dewatering tubes use state of the art Filament Textile technology for rapid dewatering and consolidation of: • Industrial Sludges • Waste Water• Paper and Pulp Sludges. NSW (02) 9631 0744 • WA, SA, NT (08) 9258 4311 • QLD (07) 3865 7000 • VIC, TAS 0404 897 979

Global Synthetics

NEW WATER QUALITY REGULATION IN QUEENSLAND: RESOLUTION OR REVOLUTION? H Gibson, P Chier Abstract In May 2008, the Queensland Parl iament passed the Water Supply (Safety and Reliability) Act 2008 that add ressed the water quality issues for potable and recycled water in two ways. This paper discusses che operatio n of chis Act, and is based on the various publications produced during the parliamentary process and d iscussions w ith officers from Q ueen sland Health and Natural Resources and Water.

Introduction Over the lase cwo years the Queensland Government has announced major infrastructure investments for water supply in South Ease Queensland associated with the rapid population growth, reduced yields from existing dams associated with global warming, and a d eficit in past water infrastructure expenditure by all p revious governments . Linked with these major infrastructure expenditures is the introduction of a complex system of instimtional arrangements for the plann ing, construction and operation of the existing and proposed assets. In essence, these new arrangements are based on the overall control of the planning and policy management of water supplies to be with the Queensland Water Co mm ission with the operational aspects to be controlled by a number of new state and local governments and private organisations. T hese arrangements were explai ned in the document "O ur Water - Urban Water Supply in South East Queensland" 1, although minor ch anges in these responsib ilities have occurred since chis document was anno unced. The changes in the administrative responsibilities has generally meant that som e of the government instrumentalities who were previously more involved in the water supply planning aspects, will in the fu ture, be m ore involved in the setting of operational standards and




monitor ing the performance of the water su pplies.

Outline of the Changes One of the areas that the govern ment has decided to regulate is water quality for both potable and recycled water. While the government has fo r many years been pondering the introduction of regulation to control the standard of potable water it was the adoption of the Western Corridor Recycled Water Project that gave emphasis to the introduction of this regulatory regime. Ir will be substantially completed by the end of 2008, and will deliver 84,?00ML/a of Purified Recycled Water into the headwaters of Wivenhoe D am, where it will be mixed with the existing water in rhe dam, prior to being released into the Brisbane River. W hen required , it will be released to Aow downstream and be extracted and treated at Mr Crosby as potable water to South East Queensland residents, i.e. indirect potable re-use. Previously, in 2005, the Queensland E nvironmental Protection Agency had p roduced water recycling guidelines which included advisory standards for recycled water up to C lass A+ standard, used for do mestic non-potable use. It reco mmended that any water used for indirect potable use sho uld be investigated on a case by case basis, based on a HACCP risk based approach. The new Water Supply (Safety and Reliability) Act, 2008, add ressed the issue of the introduction of water quality standards, and secondly it introduced the concep t of the production of water quality and water recycl ing management plans. These two initiatives come under the responsibility of Queensland H ealth and Nawral Resources and Water respectively.

Queensland Health In so far as water quality standards are concerned, Q ueensland H ealth is the

Journal of the Australian Water Association

main player under the changes. While it has always had the power to regulate standards fo r water q ualiry fo r potable water, it has never exercised chis and gen erally has only acted reactively to water qualiry issues for p otable supply. Under the changes, as well as regulating water quality for drinking water, it can now also regulate standards in the area of recycled water, greywater and water used for recreational activities. It intends to do chis and is curren tly drafting new regu lations under its Ace for water quality standards for drinking water and recycled water. In the case of drinking water, the main criteria used will be for fluoride , wcbidity and E. coli, which will be in accordance with the Australian Drinking Water Guidelines. O cher parameters that need to be measured will be determi ned on a case by case basis. In the case of recycled water standards, it intends to cover the areas of purified recycled water, such as ind irect potable quality, and C lass A+ water, the latter being used for dual reticulation systems and irrigation of crops eaten raw or with only m inimal processing. It believes chat it is nor possible to produce standards for lower class levels of recycled water, as the water quality required is more dictated by th e operational procedures of the user. The important new power chat it has been given is in the area of proactive policing of water qual ity. It can now, through the provision of Improvement Notices, require service providers to make ch anges to cheir management practices when a drinking water is unsafe, or where recycled water is not fie for use. I n the case where a service provider does provide unsafe water, it may be fined up to $225,000 o r two years imprisonment, and in the case of p roviding recycled water th at is not fie for use may be fined $100,0 00 or two years 1mpnsonment.

echnical features

Natural Resources and Water

• Preparing Audie Reports; and

Natural Resources and Water (NRW) also has che power to make standards for water quality through the provision of mandatory water quality guidelines. Although such guidelines cannot be inco nsistent with Health regulations, it will be interesting to see what a seemingly double- up in responsibilities produces .

• Preparing Ann ual Plans.

Drinking water T he new Act requires all service providers who provide drinking water to p repare a

Water Quality Management Plan (WQMP) with che purpose of protecting p ublic h ealth. In essence chis is a risk assessm ent of all possible hazards and hazardous events that may occur, togethe r with how rhe service provider proposes to manage these even ts. T he plan has to be developed in accordance with an y guidelines produced under the new Act chat cover: • Prep aring Drinking Water Quality Managem ent Plans; • Water Quali ty of Drinking Water;

The guid elines o n drinking water quality, however, must be consistent wi th any water quality standards prod uced by Queensland H ealth. Ir is expected that rhe WQMPs will be similar to rhe Recycled Water Man agement Plans as described in the Queensland Recycled Water G uidelines. The p lans need to be approved by the NRW, who may also require rhe plans to be reviewed or amended under certain circumstances. Ar app roximately rwo yearly intervals, each service provider must produce an Audit Report, which is to verify rhe accuracy of che monitoring program carried o ur u nd er the plan . T o ensure che quality of rhe audit, ir must be carried o ur by someone who is certifi ed under the RABQSA's D rinking Water Q uality Management System Aud itor Certificatio n Schem e, developed in collaboratio n with che D epartment of Human Services (Vic), the Victorian Water In dustry Association and rhe

Water Services Association (WSM, or an equivalent process. T he NRW may also have a spot audit carried our if it believes rhar rhe Audit Repo rt carried our is inadequate. All water service providers will be required to produce Annual Reports, which wi ll have to re port on the water quality m onitoring and aud its carried our and be in a format determined by rhe N RW guid el ines. The date when rhe service water providers m ust p roduce their first WQMP is: • l J uly 2011 for large service providers; • 1 Ju ly 2012 for m edium water service providers; and • 1 J uly 20 13 fo r small water service provid ers. NRW is however likely to requ ire all service water p roviders to start an interim monitoring program, and report on such, starring in January 2009. While che Act does nor require validatio n m on itoring ro be carried our specifi cally for WQMP, N RW proposed to have a sect ion in its

NOBLE Water·

Your water treatment solution providing complete turn key projects for: •

Sewage treatment plants using the latest technology to provide

re-use in small local applications or large townships. •

Potable water treatment plants for surface, bore and seawater,

built on site or in a transportable design. •

Industrial wastewater treatment and recycling systems that comply w ith effluent discharge or higher quality requirements for re-use a p plications.

Domestic water treatment, reverse osmosis, softeners and filtration.

In addition, components and small packages that adapt to existing infrastructure, enhancing performance. Whichever your need; Noblewater Is your solution provider.

Noblewater Industries Pty Ltd T: 07 3888 5222

Web: www.noblew ater.com.au

delivering clean water Journal of the Australian Water Association


AUGUST 2008 93

technical features

guidelines chat may require a water service provider to carry out validation monitoring if a new untried treatment process is proposed co be used.

Recycled water In the case of recycled water, NRW has now been given specific responsibility co regulate the use of recycled water through the introduction of Recycled Water Management Plans (RWMPs) . While the Environmental Protection Authority in 2005 produced the Q ueensland Water Recycling Guidelines, these were for guidance only, and had no regulatory backing. The government announced in 2006 chat they intended co call up this document co make many of the sections in it, specifically in relation co RWPMs, obligatory. It is interesting co note chat chis has now been done, however, only in so fa r as co protect public health. I c is assumed chat changes will be made co the environmental legislation in respect co the environment issues. In general terms, all recycled water providers will be required co produce RWMPs. The Ace accually states just what needs co be covered in the plans in very general terms. T he pl ans will need co produce a risk-based management system based on a HACCP approach , where all hazards and hazardous events are identified, che risks evaluated and management systems identified. The plan also needs co include details of the operational and verification monitoring program , and how the parameters measured will be used co indicate compliance. The NRW will produce guidelines co explain chis in more derail, and co do chis chey will produce a document similar co the Queensland Recycled Water Guideli nes water guideli nes. The timing for the commencement of the need co produce RWMPs , carry out che audits and report to NRW is as below: • For existi ng dual reticulation systems or schemes used to irrigate minimally processed crops, by 1 July 2009 . • For new (constructed between J ulyDecember 2008) dual reticulation schemes or schemes used co irrigate minimally processed crops, with in six months of water being first supplied from the planes. • For new (constructed between JanuaryJ uly 2009) dual reticulation schemes or schemes used co irrigate minimally processed cro ps, by July 2009. • For any critical recycled scheme, at any time the NRW co nsiders necessary, e.g.




the Western Corridor Recycled Water Scheme, will be required co produce a RWMP immediately. (The NRW may declare any recycling scheme chat augments a drinking water supply, a dual reticulation scheme of size> 500kL/d, or a scheme sup plying> 5ML/d co a electric power station, co be a critical scheme). • For any ocher recycled water scheme, by 1 J uly 2013. The new Ace also adopts the concept of Advisory Counci ls used in che previous Water Act to provide guidance co NRW in che administration of both the management plans. The actual ways that these advisory councils will operate is still unknown, although rhe Act is quire open in the functions and responsibilities of such cou ncils. NRW believes that such councils will only be fo rmed on an as needs basis.

Comparisons with Victoria and

NSW As the water supply srruccure in Victoria is very different to that of Queensland, legislation dealing with water quali ty and supply is dealt with in a number of Acts and Regulations by a number of authorities. Victoria has a number of water corporations char supply water and sewerage services co customers in their respective regions. Victorian Water Authorities muse prepare, implement and review risk management plans which are similar co the drinking water quality management plans men tioned in the new Bill. The NSW Department of Health supports the reuse of created effluent, as long as it has been appropriately treated, since chis can reduce the demand for potable drinking water and help protect the environment. NSW Health recommends the use of the Australian Guidelines for Water Recycling: Managing Health and Environmental Risks (AGWR phase 1) (2006) for che development of recycled water schemes. Hence, che AGWR have replaced the 1993 NSW Recycled Water Coordination Committee's NSW Guides for Urban and Residential Use of Reclaimed Water for use with dual reticulatio n schemes. T he Department of Water and Energy has also released a guideline document; Management of Private Recycled Water Schemes. This guideline aligns the principles outlined in the AGWR co the approvals process fo r private recycled water schemes (requiring section 68 approval) in NSW. T he

Journal of the Australian Water Association

guideline replaces the NSW Heal ch Interim Guidance fo r Greywater and Sewage Recycling for Multi-Unit Dwellings and Commercial Premises (GL 2005/051 previously C ircular 2004/7 1).

Summary While most of the co mponents in ch is new Ace have been talked about for a number of years, rhe exact form of the legislation has not been through a consulcacion stage, which is unfortunate. The general thrust of the legislation should be welcomed by the water industry and the community as a whole as it will provide a much more secure water supply and give more confidence co the community. T he Act itself gives details of the overall scheme, while the operational matters, which are extremely important, are still co be released in che regulations, a few months off. From the recycl ing side, the legislation is still relatively narrow, as it does not incl ude industrial or agricultural recycling, nor covers most on-sire systems. Queensland Health has scared char these areas are likely co be covered in futu re amendments. Another area of concern is in the level of monitoring char may be required in the verification monitoring for high levels of recycl ing. Ir is important rhac sufficient validation monitoring is carried out early in the process co prove the rreacmenc processes so char the risk-based HACCP approach can be used co determine what parameters need ro be measured co verify the quality of water produced. These parameters may incl ude both water quality and process parameters. It is believed chat NRW can use the Advisory Councils very effectively, not only co provide more technical information co the Department, bur also co provide rhe community with more co nfidence with the quality of water received from the water supply and water recycling schemes.

The Authors Boch che authors work for Ecowise Environmental P/L, Howard Gibson is a Principal Consultant in Brisbane and Patty Chier a Senior Environmental Engineer in Melbourne, email h.gibso n@ecowise.com.au

Reference I. Web Reference: //hccp:www.qwc.qld.gov.au/ myfiles/uploads/institutional%20 arrangements/Urban_ Wacer_Supply_

THE SOUTH-EAST AUSTRALIA DROUGHT: POSSIBLE CAUSES W Cai, T Cowan The Drought As we settle into winter the drought that has gripped much of southeast Australi a (SEA) for the past decade shows no signs of breakin g. The au tumn months of 2008 brough t licrle relief through much of southern Australia (excluding pares of Western Australia), with most rainfall tota ls between 40-60% of the long-term average. Across the Murray Darling Basin (MDB), arguably one of Austral ia's most viral econom ic resources, rhe 2008 aurumn was che fourth driest on record . The combination of higher than normal temperatures and reduced rainfall since the middle of last century has meant char water avai lab ility in throughout the catchments of SEA has reached record lows. What facto rs are co ntrolling chis late autum n rainfall decline across SEA, and is cl imate change a major factor, or is it just longterm natural variabil ity of the climate system? And what role does late aurumn rainfall play in the cu rrent drought co nditions, and how important is tempera ture/

Declining Autumn Rainfall Across Victoria Since the 1950s much of southern and eas tern Australia has exhibited stro ng rainfall reductions on a season al and regional basis. Winter rainfall across southwest Western Australia has decreased by about 12%, wh ile summer rainfa ll in southern Q ueensland has also been trending downwards. Down in Victoria one can see a considerable decline in autumn rainfall , with 50% of the autumn reduction occu rring in the month of May (Figure 1). Si nce 1990 only 5 out ofrhe 19 May months have been above the 19611990 rainfall average. Th is is in contrast ro the 13 years since I 990 that May rainfall has been well below average (between 2575%). In rhe preceding forty years May rainfall showed a roughly even distribut ion.

This is a condensed version of two recent p eerreviewed papers by the CSI R O auth ors published by the American Geophysical Union journal, Geophysical Research l mm.

Victoria May rainfall 100




> . 50


E 0


co co



"cB -50



L---'----'-- - ' - - - - ' - - - ' - - - - ' - - - - ' - - - ' ----''-----'--- - ' - - - '







Time (years) Figure 1. Percentage change in Victoria May rainfa ll from 1950-2008 (relative to the 1961-1990 average). Red (blue) bars indicate drier (wetter) than average, with an l 1-yr running mean shown as a black line.

By using rainfall data from the Bureau of Meteorology, and climate data such as sea surface temperatures, sea level pressure and surface winds fro m a combination of observations and state-of-the-arr climate models, we are able to study how the changing dynamics of the oceans surround ing Australia and the atmospheric conditions in the subtrop ics and midlacirude regions are impacting on the au tumn rainfall in Victoria. In terms of May rainfall we identified two sources of vari ab il ity in th e oceans chat are cond ucive to good rainfall, one being the region north of Australia called the Indonesian Throughflow (which co nnects che Pacific to the Indian Ocean), and the other being the subtropical and mid-latitude Indian Ocean.

Western Pacific Ocean Influences Higher sea surface temperatures in rhe Indones ian Throughflow regio n are associated with the fami liar northwest cloud bands char help generate rain fall for

Further declines in inflow across the MDB are likely.

much of central and sou th-eastern Australia. T hey typically occur during the so uthern rainy seaso n from April September. The temperature of this Through flow surface water is strongly linked to an ocean-atmosphere feedback cycle known as the El N ino-Southern Oscillation (ENSO). A transition from an El N ifio (which typically brings dry cond itions to much of eastern Australia) to a La Nifia (wetter conditions across eastern Australia) tends to be conducive to late autu mn rainfall across northern Victoria, south-west New So uth Wales and eastern South Australia. From rhe mid- l 970s onwards there have tended to be more El Nifio events than La Nifias, wh ich have also been more protracted in terms of duratio n. W hat chis means is char the western Pacific Ocean, which is linked to lace autumn rainfall across SEA, has tended be more El N ifio-like, and spending less time transitioning to La Ni fia-like condi tions. If one were to look at a rime series of the Southern Osci llation Index (SOI) which is a measure of che surface pressure difference between Tahiti and Darwin (negative values indicate El N ifi olike condi tions in the western Pacific

Journal of the Australian Water Association


AUGUST 2008 95

technical features


~~~-a"'"")_S~O_l....,_,(M_a___,y"""",_19~5_0_-2_0_0..,..6)_ _

a) Ann. MOB inflow vs. Ann. rainfall






:;::: C









~ 10000


Time (years)


40000 30000

~ 20000





.. ~ .. ,'









_.__..J 800


----~ 40000

180 120W 60W .3

slope= 55.1 :t 14.8 correl. = 0.45

~ 50000

...i ..

I I I I .2I -.3 -.2 -.1 .1


b) Ann. MOB Inflow vs. MAM rainfall








Annual rainfall (mm)


;jl u1·. )




b) MSLP wavetrains (as correlation) .' ',

slope = 42.2 :t 5. 7 correl. = 0.71

~ 50000


:;::: C





::::, 20000



~ 10000 0 L.....=~~~--1-----L-----L----L--L---'





MAM rainfall (mm)


-1 1950








Time (years) Figure 2. (a) A time-series of the SOI in May. The SOI is a measure of the surface pressure difference between Tahiti and Darwin. Negative (positive) va lues indicated El Nino (La Nifia)-like conditions in the western Pacific Ocean. An l 1-yr running mean is shown as a black line. (b) Correlation between an Indian Ocean sea surface temperature index and mean sea level pressure for May during 1950-2006 (red indicates high pressure associated with a heating in the midlatitude Indian Ocean, whilst blue indicates lower pressure). Areas confined by dashed lines indicate correlations that are significant at the 95% confidence level. (c) Time series of 500 mb geopotential height empirical orthogo nal function number 2. More negative values indicate a weakening of the pressure wave-train s emanating from the Indian Ocean.

O cean), we can see chat during May the SOI has been in decline since the 1980s (Figure 2a). A shift coward a more El N ino-like state is consistent with other studies showing a global warming-induced weakening of the easewest atmospheric circulation pattern of the tro pical Pacifi c Ocean, which is called the W al ker C irculatio n.

Indian Ocean Influences Across the ocher sid e of Aust ralia th e Indian Ocean is important in provid ing the means for rainfall generatio n for the south ern regio ns of Victoria. H earing in t he mid-lati tude Ind ian Ocean 96 AUGUST 2008


Journal of the Australian Water Association

Figure 3. (a) Scatter diagram of annual-total rainfa ll versus annual-total inflow in the MDB. (b) Scatter diagram of autumn (March, April, May or MAM) rainfall versus annual-total inflow to the MDB . All data are linearly detrended, and the analysis is carried out without 1956 data (circled), which is considered an outlier. A correlati on greater than 0 .27 is significant at the 95% confidence level.

generates d ownst ream pressure wave-trains, or spatially al te rnating high and low pressu re system s. These wave-t rains show a strong coherence with low mean sea level p ress ure over south ern Victo ria in late aut um n , rh us p roviding th e m eans fo r gen eratin g rai nfa ll (F ig ure 26 , blue colour). Due to a bas in-wide warming of the Ind ian O cean , the sea surface temp erature gradi ent between the sub tropics and the m id-lat itudes has red uced leadi ng co a weaken ing of these pressure wave-trains (F igure 2c). This has led co a rise in sea level pressu re over sou thern Victoria in May, and a subsequent reduction in rai n fa ll. T h rou gh the co m parison of observatio ns with glob al climate m odels stud ies we have been able co show char the Indian O cean warming is in part d ue co global warming. In fact the m idlatitudes of rhe Indian Ocean are so me of t he fastest warming oceans in the wo rld (they have wa rmed by almost l °C since 195 0). T herefore , ch e face char rhe weakeni ng in the pressure wave-trains are lin ked co the long-te rm Indian Ocean warming suggests a component of cl imate change is active in th e rainfall reduction over sou thern V icto ria in late autumn .

Reduction of Inflow to the Murray-Darling Basin Nor only has rain fa ll across much of SEA been in decline since the 1950s, but inAows into Australia's longest river system, the

technical features

a) JJA: Inflow & Tmax




b) SON: Inflow & Tmax



40S 130E


.5 .4 .3 .2 .1 0 -.1 -.2 -.3 -.4 -.5


Figure 4. Maps of correlations between MDB residual tota l inflow and residual mean maximum temperature for (a) w inter (June, July, August, or JJA) and (b) spring (September, October, November, or SON). All data are li nearly detrended, and the a nalysis is carried out without 1956 data.

emphasise chis fearnre because MDB rainfall experiences a greatest reduction in autumn, in the same vein as Victoria. H owever, chis seasonality of rainfall reduction is not the only catalyst fo r the unprecedented inflow decline; it is a case of higher rem peratures.

Impact of Rising Temperatures O ver che MDB variations in temperature and rainfall are not independenc, as rainfall events rend to lower daily maximum temperatures. We can establish a relationship between fluctuations of inflow and maximum temperatures chat are not related to rainfall {which we call residual Tmax). This gives us a clear picture of how temperatu re relates to inflow without the cooling influence of rainfall. In winter and spring there is a strong relationship between MDB inflow and residual Tmax (Figure 4). In these seasons an increase in maximum remperarnre is associated with a lower inflow across much of the MDB (shown in red in Figure 4). We are able to come up with a sensi tivity factor of maximum temperature with inflow for these two seasons (in summer and aurn mn there is vi rtually no relationship between inflow and temperature, mainly due to low inflow). T he sensitivi ty of resid ual inflow to residual Tmax in winrer an d spring is statistically signi ficant, at 319 GL C- 1 per month and 306 GL C- 1 per month, respectively, totall ing to 1875 GL 0 C- 1 for che two seasons. These results suggest chat a rise of 1°C in maxi mum temperature red uces long-term average inflow by about 15%. The annual average temperature of the MDB has increased by about 0.9°C since 1950. This analysis is cons istent with previous stud ies chat suggest rising temperatures are exacerbating che impact of the present dry period. 0


The economic impacts of these declines in inflows also need to be considered. For example, with water at a price of 50¢ per I 000 litres, che direct econo mic loss is about $900M, however chis does not include flow-on impacts.

The Future Murray-Darli ng, have also seen strong reductions. D uring the 1892-1902 Federation drought the average annual inflow into th e MDB was 5400 GL yr· 1 (1 GL = 10 9 litres). D uring the present drought (2000-2007) the average annual inflow was 4150 GL yr· 1, with a 12-month minimum of770 GL yr· 1 recorded for the 2006-2007 period (ending in March). Co nsidering thar rainfall during these two droughrs was comparable, it raises so me questions as to why our current drough t seems to be more devastati ng than previous droughts, including the Federation drought. One difference is that the recent drought is hotter, with maximum temperarnres in the order of 1°C higher than they were at the sra rt of 20th Century. Using dara provided by the Murray-Darl ing Basi n Commission we are able to look ar how rainfall and temperatures are linked to inflow variability and long-term trends. Firstly, we were able to come up with a relationship between rainfall and inflow across the MDB. We fou nd char on an annual basis the "conversion rate" between rainfall and inflow is 42 GL mm· 1 (Figure 3a), whereas aurnmn rainfall has a co nversion race to annual inflow of 55 GL mm · 1 (Figure 36) . What this means is that autumn rainfall across the basin is important fo r the annual-inflow total. By co nerase, che aucum n conversion rate to aurnm n inflow is low (6 GL mm· 1) indicating a soil "wetti ng mechanism" by autumn rain with delayed impacts on che proceeding seasons. We

If the relationship between temperature and inflow in winter and spri ng were to persist under a fu ture climate, a 2°C increase by 2060 (based on cu rrent climate model projections) would reduce che inflow across the MDB to about 30% of the present day levels. The current batch of climate models also project a med ian annual rainfall reduction of between 5-15% by 2060 over che MDB. Taking che worse case scenario, a 15% rainfall decrease (or about 75 mm) would contribute to a 3150 GL reduction by 2060, or about 25% of the long-term average level. So it seems likely char we will experience further declines in inflow across the MDB, as surface temperatures continue to climb and rai nfall events dwindle. Further understandi ng of how climate change will impact future SEA aucumn and winter rai nfall awaits investigation, buc for now the fu ture is nor bright.

The Authors Dr Wenju Cai is a Senior Principal Research Scientist at CSIRO Marine and Atmospheric Research, and leader of the Ocean Cl imate Characterisation and Predictions stream in the CSlRO Natio nal Flagship: Wealth from Oceans (Email: wenju.cai@csiro.au) . Tim Cowan is a member of the Climate Change group, in the Centre for Australian Weather and Climate Research, a partnership between CSIRO and the Bureau of Meteorology (Email: cim.cowan@csiro.au) Journal of the Australian Water Association


AUGUST 2008 97

MICROFUNGAL CONTAMINATION OF MUNICIPAL WATER SUPPLIES A REVIEW N B Sammon, K M Harrower Keith Harrower is a mycologist who has interests in the fungal ecology of air, water and plants. He is currently Editorin-Chief of Australasian Plant Pathology. At the 2006 International Congress of Mycology in Cairns he noticed two papers - both from overseas - on the mycology of drinking water. A survey showed that no such study had been reported from Australia and so funding was established, and a competent post-graduate student appointed. This critical literature review is the first of several anticipated publications from this study - still the first of its type in Australia - as far as is known. (In March 2007 the Centre for Plant and Water Science received funding from Fitzroy River Water to undertake a study on microfungi in the town's reticulated water supply). Abstract Very little work has been carried our on the mycology of municipal water supplies even though it is known char some fungal species are in vo lved in the production of off-tastes and odours, and others produce allergens and mycotoxins. Unti l recently, attention has been directed mainly to pathogenic bacterial, viral and protistan contaminants, and the occurrence of opportunistic microfungal pathogens in mun icipal water supplies has generally been ignored. The work that has been done is difficult to compare because of the lack of sta ndard methodologies. However, the possibili ty char che source of invasive opportun istic funga l infections, such as aspergillosis, in immuno-compromised patients may be water-bo rne spores is gaining acceptance.

Key words: Opportunistic fun gal pathogens, nosocomial, immunoco mpromised, aspergillosis, water-borne spores, municipal water supplies. Background The literature on potable water quality in municipal water supply systems is notable for the paucity of work which has been carried out in relation to the microfun gal populations of those systems. T he microbiological quality of municipal drinking water is currently determined by the presence of rhermo tolerant coliform bacteria as indicators of faecal contamination, and numerous scientific papers have been published on water-bo rne pathogen ic bacterial, viral and procistan

A critical review of the literature. 98



contami nants (Craun 1979; Szewzyk et al. 2000; Leclerc et al. 2002) . In contrast, little attention has been directed to microfungal contami nants and only a handfu l of papers on chis subject, predominately Eu ropean and American, have been published. Raw water for mun icipal water distribution systems is typically drawn either from surface water impoundments or fro m underground sources. T hese waters are created by various methods, principally chemical (e.g. alum) Aocculacion, sedimentation, and sand fil tra tion to remove suspended solids followed by san itisation to in activate micro-organisms. In some cases no treatment is ap plied. The World Health Organisation's (WH O) Guideli nes for D rinking Water Quality set the international standard for drinki ng water quality regulation bur do nor include microfungal co ntaminants. Sweden is che only country known to include specific criteria for maximum numbers of microfungi in drinking water; the Swedish Drinking Water Guidelines specify 100 colony forming units (CFUs) of microfungi per 100 mL of water as char criterion (Hageskal et al. 2007). The Australian Drinking Water Guidelines (2004) (www.nhmrc.gov.au/ pub Iications/ synopses/ _files/adwg_ l 1_06.pdn , provides the Australian water supp ly ind ustry with a framework for the management of drinking water quality, and guidance as to what constitutes drinking water that is safe to use. These guidelines do nor include any standards for the mycobiora (microfungal populations) of potab le water supplies. This lack of an official standard for fungal contamination of drinking water is no doubt one of the reasons fo r the paucity

Journal of the Australian Water Association

of work undertaken in this field. However, awareness of the potential problems posed by water-bo rne fungi which are pathogenic to humans is starting to generate interest within the scientific co mmunity. T he lack of official interest possibly stems from the insidious nature of opportu nistic human fungal infections compared to the acute natu re of water-borne epidemic disease caused by pathogenic bacteri a, viruses and proriscs. The recognition that water-borne fu ngi contribute co both aesthetic and human health prob lems has resul ted in an increasing interest in microfu ngal contamination of water distribution systems. T hese problems include taste, colour and odour in household water, and chronic human health effects caused by fu ngal allergens and mycotoxins. Opportunistic fungal pathogens are qu ick to rake advantage of co mpromised immune systems occasioned by diseases such as H IV/AIDS or by medical therapies associated with, for example, organ transplants and cancer treatment. Ir was clearly demonstrated in rhe studies included in this review chat microfungi are com mon contaminants of mun icipal water supp lies. Ic is notable chat microfungi were isolated from all pares of every municipal water supply system in vestigated, despite the geographical distribution of those systems, che varied sources of raw water, and the treatment processes applied in chose systems. The possibility that reticul ated water supplies contaminated by fu ngal spo res may be the source of nosocomial (hospitalacquired) mycoses, particularly invasive aspergillosis, in immuno-compromised patients has scarred to command scientific

accenrion wichin the lase few years. The long-held belief chat nosocomial mycoses were caused only by airborne fungal spores is now being questioned. le has been posculaced chat hospital shower facilities and associated inhalation of contaminated water vapour may be an alternative route of infection, bur invesrigarion of chis potential source has been limited (Arvaniridou et al. 1999; Anaissie and Cosca 200 I ; Warris et al. 200 I a, b; Anaissie et al. 2002; Hapcioglu et al. 2005).

Identification of Microfungi Identificatio n of microfungi is established by microscopic examination of the spores and the reproductive structures which produce chose spores. In order to achieve this, fungal propagules from water samples have to be isolated and cultured on suitable artificial media designed to encourage germination, growth, and sporularion. Figures I and 2, for example, show the markedly different reproductive structures and spores of rwo common genera of microfungi, Paecilornyces and Curvularia, which have been isolaced from municipal water supplies.

Figure 1. Reproductive structure and spores of Paecilomyces lilacinus Bar=lOf.Jm . workers included I ml (Goerlich et al. 2002), 20 ml (Nagy and Olso n 1982), 50 ml (Rosenzweig et al. 1986; West 1986), I 00 ml (Alvarez 1993; Arvanicidou et al. 1999; Hageskal et al. 2006, 2007; Yamaguchi et al. 2007; Kanzler et al. 2007), 500 mL (N iemi ettd. 1982; Warris et al. 2001 a), and I l (Hinzelli n and Block 1985; Anaissie and Cosca 2001 ; Anaissie et al. 2002; Gon~lves et al. 2005). In one case, sample size was not even scared (Hapcioglu et al. 2005).

Early Investigations O ne of rhe earliest invesrigario ns into the occurrence of microfungi in municipal water systems was conducted by Mackenzie (1938), cited in Burman (1965), who found large numbers, and a va riety of species, of microfungal contaminants in British cap water after public complaints about rasce and odour problems. Burman ( 1965) also reported on rhe enumeration of microfungi in British municipal water supply systems, as pare of a study conducted into rhe role of microfungi and acrinomyceres in caste and odour problems.

Very small samples are appropriate only when rhe water is highly contaminated and

Figure 2: Reproductive structure and spores of Curvulario eragrostidus Bar= l O t.Jm .

the mycobiora is evenly dispersed. Conversely, low co ntamination levels necessitate rhe processing of higher volume samples to ensure recovery of representative numbers and species. A sta ndard sample volume, such as is used in bacteriological analysis, should be used if co mparison is be meaningful since pro-raring rhe results of different sample sizes to a standard volume does nor necessa rily overcome the probl em. For exa mple, Arvaniridou et al. ( 1999), who analysed 126 water samples of 100 ml each, recovered filamentous microfungi from 82.5% of chose samples whereas Goerlich et al. (2002), using l mL samples,

BIO MESH TUBING The ideal filter medium

Standard Method There is no standard method for quantitative and qualitative research in to microfungal co ntamination of municipal water supplies. T he diversity of experimental protocols used in the few mycological studies reported to dare is of concern since comparison of resul ts is consequently made difficult and possibly misleading.

Sample Volume A good example of che diversity of meth odology used is rhe volume of water sample analysed. Sample sizes used by various

100% Australia owned and manufactured • • • • • • • •

high surface area for bacterial production encourages excellent biological film growth modular design for easy maintenance and handling cost effective, efficient, reliable chemically inert used in trickle filters, aeration and degassing systems, submerged filtration and treatment systems perfect for all aqua cultural environments ideal as gutter guard, cooling tower fill pack and in oil separation units

Acacia Filtration Systems Pty Ltd

Tel: 02 9756 6077 Fax: 02 9756 0173 sales@acaciaproducts.com.au www.acaciaproducts.com.au

Journal of the Australian Water Association


AUGUST 2008 99

technical features

Table 1. Genera most commonly recovered and % of total Genus % Genus %

Authors Genus

% 28

West (1986)

Penicillium Penicillium Aspergillus Cladosporium


Alvorez (1993)



Gottlich el al. (2002) "

Phialophora Penicillium 41 Penicillium Cladosporium 75

Nagy & Olsen ( 1982) Hinzellin & Block (1985) Rosenzweig el al. (1986)

Goncalves el al. (2005) Hogeskol el al. (2006) " Kanzler el al. (2007)

23 18

Sporocybe Aspergillus Allernaria Phoma Hormodendron (Cladosporium) Acremonium Acremonium Trichoderma

28 16 17 19 16


Bosidiomycetes 56



Acremonium 10-15 Rhizopus 14 Cladosporium 17 Allernaria 7 Fusarium 12

Paecilomyces Trichoderma Penicillium Exophiala Penicillium


Exophiola Phialophora Aspergillus Penicillium

Penicillium Cladosporium Phialophora Acremonium

4 49

12 10 7 10

4 26

(Unspecified) Arvonitidou el al. (1999) Worris el al. [2001) Anoissie el al. (2002) " Hopcioglu el a/. (2005) * Note: *

Penicillium Aspergillus Penicillium Penicillium

33 40

= Number of positive samples only reported,

reported chat only 7.5% of 2657 water samples tested positive for microfu ngi. These results highlight the need for use of a standard sample size in enumerating fungal popu lations in water. T hey may have been representative of the waters reseed, bur on the ocher hand the large variance in the number of positive samples may have been attributable co the sample size.

Isolation Method With che exception of Goerlich et al. (2002) who used the spread place method, all workers were consistent in using the membrane filtration method co isolate fun gal propagules. This was followed by incubation of the spore-laden filters on various agar-based culture med ia. The main problem associated with the spread place method when used for enumeration studies, is the very small volume of sample involved which may give rise co possibly misleading results.

Media Most of the researchers used Sabouraud Dextrose agar or Mair Extract agar as germination media, with or without the addition of various bactericides such as aureomycin, chloram phenicol, strepcomycin, gencamicin, or penicillin co suppress bacterial growth. Niemi et al. (1982) and Rosenzweig et al. ( 1986) incorporated the dye rose bengal inco their media. Rose bengal is commonly added co media used co isolate molds from food and soil since it has antibacterial properties, inhibits spreading fungi such as Rhizopus spp. and Mucor spp., and limits the colony l 00 AUGUST 2008


Aspergillus Trichoderma Aspergillus Aspergillus

28 29

Acremonium Paecilomyces Paecilomyces Acremonium

5 24

Chrysosporium 5 Cladosporium 20 Trichoderma

- ---

percentages not stated.

diameters of ocher genera. However, it has been suggested that the dye may completely inhibit some yeast and mold strains Qarvis, 1973). Goerlich et al. (2002) used blood agar, a standard bacteriological medium bur an inappropriate medium for fungi, for germinatio n followed by sub-culturing onco Poraco Dextrose agar and Mair Extract agar. This choice of an inappropriate fungal medium, together with the small sa mple size of 1 mL, may explain the fact that Goerlich et al. (2002) reported a much lower recovery of CFUs than any of the ocher researchers mentioned in chis review.

Incubation Temperature Most researchers incubated their isolation media within the temperature range of 20°C t0 25°C but there were [ WO exceptions. Warris et al. (200 I a) used 35°C, and Hapcioglu et al. (2005) used both 26°C and 37°C bur did not differentiate when reporting results. In two cases, incubation temperature was seated as 'room temperature' (Alvarez 1993; Arvanicidou et al. 1999), and in one case rhe incubation remperarure was nor given at all (Rosenzweig et al. 1986).

Replicate Samples Testing of replicate samples was reported in on ly t:wo of the studies reviewed. Arvanicidou et al. ( 1999) used cwo replicates and Nagy and Olson (1982) used five replicate samp les. Ir is recognised chat collection and processing of replicates can be difficult logistically bur duplicate samples, at least, are necessary ro provide srariscical rigour co rhe results.

Journal of the Australian Water Association

Periodicity of Sampling In several insrances the srnd ies were based on samples collected once only (Arvanicidou et al. 1999, Kanzler et al. 2007, Yamaguchi et al. 2007) while most sampl ing was incermiccenc. With the exception of Goerl ich et al. (2002), Anaissie et al. (2002), and Gon<;alves et al. (2005), most of the studies were restricted co periods of less than twelve months and hence did not encompass all seasons. Samples collected at one po int of rime cannot be considered as representative of the system under investigation since rhey may be affected by seasonal variations in physico/chemical properties, or by one-off even ts such as treatment plant malfunctions, breaches of distribution mains allowing ingress of contaminated water, or soil run-off into raw water sources during rainy seasons.

Seasonality and Physico/ Chemical Properties Seasonal icy can have a significant impact on the physico/chem ical properties of municipal warer suppl ies, particularly temperature, as well as turbidity and pH in the case of chose sourced from surface waters. Six of the researchers, Arvanicidou etal. (1999), Warris eta!. (2001a), Hapcioglu et al. (2005), Hageskal et al. (2006, 2007), Yamaguchi et al. (2007), and Kanzler et al. (2007), either did nor record any physico/chemical parameters or did nor report chem. Ir is interesting that only the early researchers (Niemi et al. 1982; Nagy and Olson 1982; H inzellin and Block 1985; Rosenzweig et al. 1986; West 1986)

technical features

recorded three or more physico/chemical properties of the waters rhey investigated.

chemical coagulation was lower chan in water treated by sand filtration alone.

Reporting of Treatment Protocols


Despite rhe fact char treatment of public water suppl ies is intended, ar lease in part, to remove or inactivate harmful microorganisms, ir is surprising that over half of the researchers whose studies are reviewed, either failed to elucidate the treatment protocols, including sanitisatio n, and the physico/chem ical properties of rhe systems in their investigation, or else recorded incomplete information in this respect. Of particu lar importance is che concentration of free chlo rine in treated water leavi ng che treatment plant as well as rhe levels of resid ual free chlorine in the distribution system. Knowledge of the treatme nt protocols employed, togecher with kn owledge of rhe source raw water and its fungal load, is essential to the understanding and interpretation of the significance of fu ngal contaminatio n of a reticu lated treated water supply. For instance, as a result of their studies, Niem i et al. ( 1982) were able to show char the incidence of fungi in water treated by

Several researchers showed that correlations existed between physico/chemical parameters and rhe number of microfungi recovered from municipal water. West ( 1986) found that recovery of fi lamento us fungi was positively correlated with low temperature and high heterorroph ic bacterial place counts (HPC), and that recovery of yeasts was correlated with high temperatures and low HPC. Nagy and Olson ( 1982) found char turbidity and pH were pos itively and significan tly co rrelated with the frequency of fungal isolations from chlorinated water, and according to Alvarez (I 993) recovery of yeasts was pos itively correlated wich turbidity. Ocher researchers demonstrated a positive correlation between yeast colony numbers and total co li for ms (Arvanitidou et al. 1999; Yamaguchi et al. 2007), while Arvanitidou et al. (I 999) fou nd that the number of CFUs of filamentous microfungi was pos itively correlated with the number of heterorrophic bacterial colonies. On the other hand

Specify and use Link-Seal® seals to seal pipe penetration

Slide the assembly into the space between the pipe and wall ope ning

PR • jEX

When the bolts are tightened, Link-Seal modular seals expand to create a gas and water tight seal. Available in Australia and New Zealand

Projex Group Pty Limited Telephone: 02 8336 1666, Facsimile: 02 8336 1670 www.projex.com.au, email : mail @projex.com.au

Gons;alves et al. (2005) found a negative correlation between rhe numbers of bacterial/yeast colonies and fi lamentous microfungal colonies. It is curious as to why they used combined bacterial/yeast numbers since yeasts are fungi and have no relationship with bacteria.

Identification of Fungal Contaminants ldentification, at least to generic level, of microfungi recovered was reported by all researchers except Niem i et al. (1982) and Yamaguchi et al. (2007). The latter identified yeasts only. Fungi were identified to species level in very few of rhe reports. Table 1 sets our the genera of filamentous fungi most commonly recovered by each of the ocher researchers. The genus Aspergillus is prominent in char list and is well-known as rhe aetiological agent of invasive aspergi llosis wh ich can infect most parts of the human body (Rippon, 1988). Some Aspergillus spp. produce al lergens and mycotoxi ns. Penicillium spp. and F11sari11m spp. also produce mycoroxins (Rippon, 1988). Many of rhe other genera listed in Table 1, for

-= -8 awwws



Designer a nd ma nufacturer of hig h effic iency, low speed floating a nd fixed surface aerators fro m 3kW to 2 20 kW wi th a n unmatche d 5 year, unlim ited hou rs g uarantee. By-Jas offers fle xible fina nci ng and delivery solutions includ ing renta l, purchase a nd fully mainta ined operating leases. Ring now for a current stock list. Other prod ucts in our range include settling tanks (12 designs), packaged sewage and water treatment pla nts, reuse filters a nd clarifiers to Cla ss B a nd Class A sta nda rd.

For more information, contact: By-Jas Engineering Pty Ltd PO BOX 424, HASTINGS VIC 3915 Tel: (03) 5979 1096 Fax: (03) 5979 1524

Journal of the Australian Water Association


AUGUST 2008 101

technical features

example Cladosporittm, Acremonium, Exophiala, Phialophora, and Paecilomyces, contain species which are either superficial or systemic opporcuniscic human pathogens (Rippon , 1988). While all of these genera have been recovered from municipal water supplies there is no definitive evidence chat water is a human infection route; however the possibility is a compelling theory, particularly in relation to chose who are immuno-comprom ised.

humans through municipal water supplies wi ll, if confirmed, generate considerable official interest in the monitoring and control of microfun gi in potable water supplies.

Acknowledgments The authors are grateful to Professor R Reed and A/ Prof L Fabbro for a critical reading of the manuscript.

The Authors Hospital Waters Investigations by Arvanitidou et al. (1999), Wa rris et al. (200 l a,b), Anaissie and Cosca 2001 , Anaissie et al. (2002), and Hapcioglu et al. (2005) strongly suggest chat hospital water, particularly shower aerosols, may be a possible route of nosocomial infection by Aspergillus and other fungal genera. These mycological studies were valuable additions to the scien tific knowledge. However, since che hospital waters investigated received no secondary treatment or sanitisation , apart from storage in on-site tanks, they were liccle more than proxies for the municipa l water systems from which the hospitals derived chei r water.

Conclusion All of che studies included in chis review showed chat microfu ngi were present in all of rhe municipal drinking water systems which were investigated. Mose of che genera recovered were common to each of chose water systems, and many of chose genera contain species which are known to be opportunistic human pathogens, either superficial or systemic. However, chis work has been concentrated in Europe and America. Mycological analysis of water supply systems in ocher areas, particularly in Austral ia's tropical and temperate regions, is needed. Extension of the work pioneered by Arvaniridou et al. (1999) , Warris et al. (200l a,b), Anaiss ie et al. (200 l , 2002), and Hapcioglu et al. (2005) on rhe possible role of hospital shower and rap waters as a route for nosoco mial fungal infections is urgently needed. If waterborne infection can be shown to occur in hospitals then water-borne infect ion in the home is also highly likely since hospital water is invariably drawn from municipal supplies. T he recent trend to minimise the duration of hospital in-patient care, with subsequent on-goi ng therapies administered after discharge to immunoco mpromised patients living at home would support chat view. The theory chat invasive mycoses, particularly aspergillosis, may be transm itted to immuno-compromised



Noel Sammon is a pose-graduate researcher and A/Prof Keith Harrower is a staff member in the Centre for Plant and W ater Science, Central Queensland University, Rockhampton, Australia, email k.harrower@cqu.edu.au

References Alvarez C lV ( I 993) Caracrerizaci6n m icol6gica de aquas "crudas" y filcradas en la Planca de T racamiento de Tres Rios, Cosca Rica. Revista de Biologia Tropical 41 , 4 17-422. Anaissie EJ , Costa SF (2001 ) Nosocomial aspergillosis is water-borne. Clinical lnfectio11s Diseases 33, I 546- 1548 . Anaissie EJ , Stratton SL, Dignani M C, Lee C, Summerbell RC, Rex JH (2002) Pathogenic molds (including Aspergillus species) in hospi tal water distribution systems: a 3-year prospective study and clinical implications for patients with hematologic malignancies. Blood 10 1, 2542-2546 . Arvanitidou M, Kanellou K, Consranrinides TC, Kacsouyannopoulos V ( 1999) The occurrence of fungi in hospital and co mmunity potable waters. l etters in Applied Microbiology 29, 8 1-84. Burman NP ( 1965) Taste and odour due to stagnation and local warming in long lengths of piping. Proceedings ofthe Society for Water Trtatmem Examination 14 , 12513 1. Craun GF (1979) W ater-borne Giardiasis in the United Scares: A Review. American journal of Public Htalth 69, 8 17-8 19 Gorn;:alves AB, Russell R, Paterson M, Lima N (2005) Survey and significance of filam entous fungi from cap water.

International journal of Hygiene and Environmental Health, 257-264 . Gocclich E, van der Lubbe W, Lange B, Fiedler S, Melchert I, Reifenrath M, Flemmi ng HC, de Hoog S (2002) Fungal flora in groundwater-derived public drinking water.

International joumal of Hygiene and Environmental Health 200, 269-279. Hageskal G, Gaustad P, Heier BT, Skaar I (2007) Occurrence of moulds in drinking water. joumal ofApplied Microbiology 102, 774-780. Hageskal G, Knutsen AK, Gaustad P, de Hoog GS, Skaar I (2006) D iversity and significance of mold species in Norwegian drinking water. Applied and Environmental Microbiology 72, 7586-7593. Hapcioglu B, Yegenoglu Y, Erturan Z, Nakipoglu Y, lssever H (2005) Hererorrophic bacteria and filamentous

Water Journal of the Austra lian Water Association

fu ngi isolated from a hospital water distribution system. Indoor Built Environment 14, 487-493 . Hinzelin F, Block JC (1985) Yeasts and filamentous fu ngi in drinking water. Environmental Technology l etters 6, IO 1106. Jarvis B (1973) Comparison of an improved rose bengal chlortetracycline agar with other media fo r the select ive isolation and enumeration of moulds and yeasts in foods. journal ofApplied Bacteriology 36, 723-727. Kanzler D, Buzina W, Paulicsch A, H aas D, Platzer S, March E, Mascher F (2007) Occurrence and hygien ic relevance of fungi in drin king water. Mycoses (OnlineEarly

Articles). Leclerc H , Schwarczbrod L, Dei-Cas E (2002 Microbial Agenrs Associated with Waterborne D iseases. Critical Reviews in Microbiology 28 (4) , 37 1-409. Mackenzie EFW (1 938) Thirty-third Annual Report on T he results of the Bacteriological, Chemical and Biological Examinatio n of che London Waters, Twelve months ended 31st December, 1938. In. (Ed. MW Board) pp. 27. (P.S. King and Son Led.: London) Nagy LA, O lson BH (I 982) The occurrence of fila mentous fungi in drinking water disrriburion systems. Canadian journal of Microbiology 28, 667-67 1. N iem i RM, Knuth S, Lunclsrrom K (1982) Acrinomyceres and fun gi in su rface waters and in porable water. Applied and Environmental Microbiology 43 , 378-388. Rippon JW (1988) 'Med ical Mycology - T he Pat hogenic Fungi and che Pathogenic Acrinomyceres' 3'd Edn . (W.B. Saunders Company: London) Rosenzweig WO, M innigh H, Pipes WO ( 1986) Fungi in potable water distribution systems. journal ofthe American Water Works Association 78, 53-55 . Szewzyk U, Szewzyk R, Manz W, Schleifer KH (2000) Microbiological safety of drinking water. Amwal Review of Microbiology 54, 81- 127. Warris A, Gaustad P, Meis JFGM, Voss A, Verweij PE (2001a) Recovery of filamentous fungi from water in a paediat ric bone marrow transplantation unit. Journal of Hospital Infection 47, 143-1 48. Warris A, Voss A, Verweij PE (2001b) Hospital sources of Aspergillus species: New routes of t ransmission? Revista lberoamericana de micologia 18, 156- 162. W est PR (1986) Isolation races and characterisation of fungi in drinking water systems. In ' in Proceedings of the water quality technology conference'. D enver, C O. pp. 457-473. (AWWA research fou ndation and the American Water Works Association, Denver, CO.) Yamaguchi MU, Rampazzo RC P, YarnadaOgacca SF, Nakamura C V, Ueda-Nakamura T , Filho BPD (2007) Yeasts and filamentous fun gi in borcled mineral water and cap water from municipal supplies. Brazilian Archives of Biology and Technology 50, 1-9.