This article was downloaded by: [95.156.182.177]On: 28 February 2015, At: 01:52Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Urban Water JournalPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/nurw20

Assessing integrated water management options forurban developments – Canberra case studyA. K. Sharma a , S. Gray b , C. Diaper a , P. Liston c & C. Howe aa CSIRO Land and Water , PO Box 56, Highett, Vic, Australiab Victoria University , PO Box 14428, Melbourne, Victoria, 8001, Australiac Environment ACT , PO Box 144, Lyneham, ACT, 2602, AustraliaPublished online: 25 Jun 2008.

To cite this article: A. K. Sharma , S. Gray , C. Diaper , P. Liston & C. Howe (2008) Assessing integrated water managementoptions for urban developments – Canberra case study, Urban Water Journal, 5:2, 147-159, DOI: 10.1080/15730620701736829

To link to this article: http://dx.doi.org/10.1080/15730620701736829

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Assessing integrated water management options for urban developments – Canberra case study

A.K. Sharmaa*, S. Grayb, C. Diapera, P. Listonc and C. Howea

aCSIRO Land and Water, PO Box 56, Highett, Vic, Australia; bVictoria University, PO Box 14428, Melbourne, Victoria 8001,Australia; cEnvironment ACT, PO Box 144, Lyneham, ACT, 2602, Australia

Urban water services in the Australian Capital Territory (ACT) are currently provided through conventionalcentralised systems, involving large-scale water distribution, wastewater collection, water and wastewater treatment.A study was conducted to assist Environment ACT in setting broad policies for future water services in Canberra.The current paper presents the outcomes of a study examining the effects of various water servicing options on waterresources and the environment, for two townships in Canberra, one existing and one greenfield site. Three modellingtools were used to predict the effects of various alternative water servicing scenarios, including demand managementoptions, rainwater tanks, greywater use, on-site detention tanks, gross pollutant traps, swales and ponds. The resultsshow that potable water reductions are best achieved by demand management tools or a combination of greywaterand rainwater use for existing suburbs, while third pipe systems are preferred for greenfield sites. For this specificclimatic region and end use demands, modelling predicted increased water savings from raintanks compared togreywater systems alone, with raintanks providing the additional benefit of reduced peak stormwater flows at theallotment scale. Rainwater and stormwater reuse from stormwater ponds within the catchments was found toprovide the highest reduction in nutrient discharge from the case study areas. Environment ACT amended planningcontrols to facilitate installation of raintanks and greywater systems, and commenced a Government funded rebatescheme for raintanks as a result of this study.

Keywords: integrated urban water management; water sensitive urban design; raintanks; water; wastewater;stormwater; water balance modelling

1. Introduction

Integrated urban water management (IUWM) is basedon a multi-dimensional approach to water manage-ment, where water resources are optimally utilisedbased on the fit-for-use concept. The aim of IUWM isto include all aspects of the water cycle, wastewater,stormwater, drinking water and evaporation to opti-mise operation and management solutions. Water-sensitive urban design (WSUD) is an aspect of IUWM,which focuses on the planning and configuration ofdevelopments to minimise impacts and move towardsmore sustainable systems. The protection of naturalecosystems, integration of water resources for provi-sion of water services and reduction of peak storm-water flows are some of the objectives of WSUD andplanning. A number of structural and non-structuraltools can be used to achieve IUWM and WSUDobjectives, the selection of which will be dependent ona large number of factors including; the type ofdevelopment, catchment conditions, climate, customeracceptance and allocation of financial resources. Someexamples of structural tools are rainwater tanks,

greywater treatment and reuse, wastewater reuse,stormwater use, on-site detention tanks, buffers,swales, bioretention devices and ponds.

Both IUWM and WSUD encompass concepts ofsustainability as they incorporate economic, social andenvironmental dimensions. In terms of the water cycle,the fundamental objectives of IUWM are based onboth water quantity and quality and can be simplifiedas reducing contaminant discharge to the environmentand maintaining environmental flows. These objectivesare combined with economic, social and other envir-onmental objectives to provide more sustainableoptions for water servicing. Moreover, the basicIUWM process consists of number of tasks for itsimplementation.

The current work does not encompass all aspects ofthe IUWM process but focuses on the water balanceanalysis, stormwater quality management and allot-ment scale stormwater peak flow management with anaim to efficient use of water resources and to maintaindevelopment conditions as far as possible close topredevelopment stage.

*Corresponding author. Email: ashok.sharma@CSIRO.au

Urban Water Journal

Vol. 5, No. 2, June 2008, 147–159

ISSN 1573-062X print/ISSN 1744-9006 online

� 2008 Taylor & Francis

DOI: 10.1080/15730620701736829

http://www.informaworld.com

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

Numerous studies have been conducted around theglobe on the assessment of alternative water servicingoptions for infill and greenfield developments, but fewof these studies have outcomes available in publishedliterature. Dixon et al. (1999) indicated that the societymust move toward the goal of efficient and appropriatewater use for a sustainable urban future and describedthat the reuse of domestic greywater and rainwater wasa step in this direction. Similarly, Manios and Tsanis(2006) also highlighted the importance of wastewaterreuse in water resources management. Hardy et al.(2005) highlighted that the process of urbanisation anddevelopment can significantly change the input andoutput flows and quality from an area and can create ahighly inefficient system in terms of water resources.The inputs and outputs can be minimised through theefficient use of water resources.

Mitchell et al. (2001) highlighted the use ofstormwater and wastewater as a potential substitutefor a portion of the fresh water from reticulated supplysystem and also presented a water balance model(Aquacycle) representing water flows through theurban water supply, wastewater and stormwatersystems. Grimmond et al. (1986) presented a modelto calculate the water balance components for anurbanised catchment. They divided the total area inthree discrete surface types, namely (a) impervious area(roads, parking lots, buildings), (b) pervious unirri-gated (lawns, other greenspace and open land notartificially watered) and (c) pervious irrigated (lawnsparks and other areas watered). The Aquacycle modelused in this study is also based on the similar concept.

Parkinson et al. (2005) examined the impact ofretrofitting source control technologies e. g. rainwaterand grey water usage and reduced flush toilets. Theyindicated that the introduction of these technologiescould have implications on the performance ofdownstream infrastructures and treatment processes.However, these technologies would reduce domesticwater demand and peak load on water supplydistribution systems. Rueedi et al. (2005) also studiedthe impact of greywater reuse and rainwater diversionfrom roof runoff for irrigation along with otherretrofitting scenarios for a suburb with a populationof 15 000. It was reported that the household wateruse decreased by about 12% due to grey water usagefor toilet flushing and rain water usage for irrigationdecreased stormwater flows up to 16% and mainssupply by 6%. Similarly Villarreal and Dixon (2005)analysed retrofitting rainwater collection systems inexisting developments by modelling rainwater fortoilet flushing, laundry, garden irrigation and carwashing. Low water consumption appliances werealso considered in evaluating water saving efficiencyand water conservation. It was concluded that the use

of rainwater contributed to important savings indrinking water.

The aim of the current paper is to present amethodology for the assessment of alternative waterservicing options and stormwater quality and quantitymanagement for both greenfield and existing develop-ments and to provide commentary on the appropriate-ness of the assessment tools used. More importantly,the study assisted Environment Australian CapitalTerritory (ACT) to develop policies and directions forfuture water services in Canberra. Several modellingapproaches were applied in this study to examine theimpact of various water servicing options against setobjectives. The township of Woden was the case studyfor an existing suburb and Gungaderra as a case studyfor a greenfield site.

2. Analysis methodology and modelling approach

The methodology adopted for the water resourcesquantity and quality modelling and assessment ofservicing options is depicted in Figure 1. This includesthe identification and formulation of study objectives,selection of various options for achieving objectives,development of water servicing scenarios combiningvarious options, data collection for study area char-acteristics including climate and water consumptionfor various end users, analysis of scenarios for water,wastewater and stormwater, stormwater quality andquantity including allotment scale stormwater peakflows, interpretation of analysis and finally recommen-dation of options to meet objectives. Here a scenario isdefined as ‘a set of options that fully describes theprovision of water, wastewater and stormwater op-tions in a given area’.

To compare various integrated water servicingcombinations, information on the variation in totalwater demand, wastewater and stormwater flows andcontaminant loads between selected water servicingoptions is essential. Three computer models were usedto estimate the water and contaminants flows forvarious composite alternative water serving scenariosas no single model was capable to model water,wastewater and stormwater quantity and quality. Thetotal urban water balance was modelled using Aqua-cycle, stormwater flows, contaminants and treatmentdevices were modelled using MUSIC (Model forUrban Stormwater Improvement Conceptualisation)and peak stormwater flows from allotments weremodelled using PURRS (Probabilistic Urban Rain-water and Wastewater Reuse Simulator). The briefdescription of these models has been described below:

AquaCycle incorporates potable water, wastewaterand stormwater flows into a single framework,allowing the effect of the water sensitive design

148 A.K. Sharma et al.

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

measures on the total water cycle to be assessed.AquaCycle uses a daily time step and so the localclimate variability is included in the model. A varietyof different water servicing options can be modelled atallotment, neighbourhood and catchment scale such asrainwater tanks, grey water irrigation, on-site waste-water treatment, cluster scale stormwater and waste-water storage for reuse, aquifer storage and recovery,catchment scale stormwater and wastewater storages(Mitchell et al. 2001).

The MUSIC model has been used to estimatestormwater quality and simulate stormwater qualityimprovement measures such as gross pollutant traps,wetlands, buffers, swales, bioretention systems, pondsand sedimentation basins. MUSIC can simulate theperformance of a group of stormwater managementmeasures configured in series or in parallel to form atreatment train. MUSIC uses a six minute time step andso can simulate the mean flows, annual flows and totaldaily loads/concentrations of total suspended solids(TSS), total nitrogen (TN) and total phosphorous (TP)

for specified treatment train scenarios (Wong et al.2002).

The PURRS model was used for event basedstormwater peak discharge calculations (Coombes2003). The model uses pluvio data and the descriptionof the site to calculate peak flows through an on-sitedetention (OSD) tank. As the model is not a designtool, the sizing of an OSD tank for a given frequencyof storm event (eg. 1 in 5 year storm event) was carriedout by a trial and error method.

3. Application of methodology

3.1 Development of study objective

Following discussions between CSIRO and Environ-ment ACT the following objectives were developed toincorporate both population growth and environmen-tal issues for the case study sites:

(1) A 25% reduction in household potable waterdemand.

Figure 1. Water resources quantity and quality analysis modelling and assessment methodology.

Urban Water Journal 149

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

(2) Peak stormwater flows for individual catch-ments no greater than pre-development.

(3) Phosphorus loads associated with stormwaterflows at pre-development levels.

(4) Suspended solids loads associated with storm-water flows at pre-development levels.

(5) Nitrogen loads associated with stormwaterflows at pre-development levels.

(6) A 20% reduction in wastewater discharged perhousehold.

The first objective of a 25% reduction in potablehousehold water demand was identified as the criticalobjective, with all others being of secondaryimportance.

3.2 Selection of options

Various alternative water-servicing options were as-sessed against these objectives using computer modelsto predict water flows and stormwater quality andquantity. The following alternative techniques (op-tions) of integrated water service provisions wereinvestigated:

. Water demand management,

. Raintanks,

. Greywater irrigation of gardens,

. Both raintanks and greywater irrigation ofgardens,

. A local wastewater treatment plant for thesupply of non-potable water for toilet flushingand garden irrigation,

. Gross pollutant traps (GPTs) and ponds forstormwater treatment and irrigation of publicopen space (POS) with water from stormwaterponds, and

. OSD at allotment scale to lower peak stormwaterflows.

3.3 Development of servicing scenarios

3.3.1 Existing development (Woden) scenarios

The scenarios for the existing development contained alimited number of alternative water system measures.Dual pipe systems were considered too expensive toinstall in existing suburbs as previous work hasindicated that the economics of water reuse schemesfavours application to new developments rather thanthe retrofit projects (Alegre et al. 2004). In this workthe advantages and disadvantages of recycling, rain-water and conservation options were reviewed andsuggested that while the recycling would providecontinuous supply, systems could be expensive

paticularly if dual reticulation was required. Sharmaet al. (2005) conducted a study for a 3062 ha greenfielddevelopment and estimated the cost of a dual pipesystem, proposed to supply recycled water for toiletand garden irrigation. The cost of the dual pipe systemwas 73% of the water supply system designed for theremaining supply. Considering the constraints, limita-tions and additional cost of reinstatement worksincluding the cost associated with the management ofconstraints in retrofitting dual pipe systems in existingdevelopments, it can be reasonably considered thatsuch an option will be expensive.

The following on-site measures were considered inexisting development:

. rainwater tanks for garden irrigation and toiletsupply,

. irrigation of gardens using greywater from thelaundry and bathroom,

. both rainwater tanks and greywater irrigation ofgardens.

If both raintanks and greywater system were incorpo-rated for garden irrigation on a single allotment, theraintanks were used as a back up supply for gardenirrigation as well as providing water for toilet flushing.Stormwater in this catchment is currently dischargedto a creek via stormwater drains and open channels.Three stormwater ponds were proposed for storage ofwater for the irrigation of 166 ha of public open space.These ponds would also operate as a device forstormwater treatment. Two ponds 3 ha 6 1.5 mdeep and a third pond 4 ha 6 1.5 m deep wereconsidered.

To conceptualise the improvement in the catch-ment stormwater runoff quality, a number of regionalstormwater treatment measures such as gross pollutanttraps and bioretention systems were also considered inorder to achieve stormwater quality objectives. Inaddition to OSD tanks at allotment scale the followingtreatment trains were also considered:

. GPTs and ponds in three suburbs

. GPTs, ponds and bioretention systems

. Buffers only (grass strips)

. Buffers and swales

. Buffers and bioretention systems

3.3.2 Greenfield development (Gungaderra) scenarios

Planned water services for the greenfield developmentare typical for Australia, with a conventional reticu-lated water system and wastewater collection andtreatment at a remote location. A series of tenstormwater ponds, with a total surface area of

150 A.K. Sharma et al.

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

11.85 ha and capacity of 159 ML, are planned toreduce the peak stormwater flows. This description ofplanned water services is referred to as the base case,and is used as a reference point for comparison ofalternate scenarios. In the greenfield development agreater range of alternative water measures wereconsidered, compared to those for the existing devel-opment. Allotment scale measures considered wereidentical to those for the existing development withrainwater tanks, greywater irrigation of gardens,rainwater tanks and greywater irrigation of gardenson the same block and on-site detention tanks. Anumber of off-block measures were also considered,such as swales and a local wastewater treatment plantwith reclaimed water returning to residential propertiesfor garden irrigation and toilet flushing.

To conceptualise the improvement in the catch-ment stormwater runoff quality, the treatment trainsconsidered in addition to the existing 10 stormwaterponds were:

. GPTs

. GPTs and swales

4. Data collection

4.1 Study area description

4.1.1 Existing development (Woden area)

The township of Woden was developed in the early1960s and 1970s and is located in the South East regionof Canberra. The total population at the time of thestudy was estimated at 32 611 people, living in 13 890dwellings over an area of 2990 ha with an averageoccupancy rate of 2.35 per dwelling. For modellingpurposes the residential area was divided into 13suburbs, with an overall average household block sizeof 1013 m2 (377 m2 min to 5228 m2 max). The area iscategorised in to the following land uses: 49%residential, 4% commercial, 37% public open spacesand 10% roads (Mitchell 2003). One of the suburbs,O’Malley, area 257 ha is a diplomatic sector having

large lots of 5228 m2 with only 10% as built up areaand remaining 90% as garden area. Urban waterservices are provided by conventional systems forwater supply, wastewater collection and disposal andstormwater drainage. The stormwater from the town-ship flows into Yarralumla Creek via lined openchannels.

4.1.2 Greenfield development (Gungaderra area)

The township of Gungaderra is in the outer NorthWest of Canberra, and is currently in the developmentphase. No building had commenced at the start of thisstudy. The total area of Gungaderra is 630 ha, ofwhich 465 ha will be developed for detached andmedium density dwellings. The detached dwellings willhave an area of 379 ha consisting 5200 dwellings withlot size of 450 m2 for 13 520 population. Similarlymedium density development in 77 ha area will have1300 dwellings with a lot size of 300 m2 for 3380populations. The population estimation is based on theoccupancy rate of 2.6 persons per dwelling. Thistownship also has an area of 9 ha allocated forcommercial development. The remainder of the sitewill provide roads, commercial premises, open spaceand stormwater ponds.

4.2 Water usage and current demand managementpractices

Water use data for residential allotments wereobtained from average household water consumptiondata for Canberra (Table 1). The consumption figureswere then adjusted to account for increased uptake ofwater efficient appliances. All data on water consump-tion patterns, household occupancies and water sav-ings from increased use of low water using householdappliances was provided by Environment ACT. Thewater consumption data were based on the currentdemand management practices that were considered aslow demand management water consumption figuresin this study.

Table 1. Typical in-house water usage.

Water demand

Typical water usage per person

Low rate ofdemand management

High rate of demandmanagement – Woden

High rate of demandmanagement – Gungaderra

L/c/day L/c/day L/c/day

Kitchen 23 22 23Bathroom 77 54 47Laundry 50 43 38Toilet 70 53 49Total 220 172 157

Urban Water Journal 151

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

The low demand management water use data wasbased on uptake rates of 57% dual flush toilets, 32%low flow shower heads, 15% front loading washingmachines, 47% dishwashers, 10% reduced lawn areaand 10% irrigation system. It means that 57% ofhouses would have dual flush toilets and so on. ForWoden high rate of demand management water usedata uptake rates were increased to 90% water efficientdishwashers, 77% water efficient showerheads, 82%dual flush toilets and 65% water efficient washingmachines. In Gungaderra, 100% provision of waterefficient appliances was considered in high rate ofdemand management water use figures.

The average garden irrigation figures (Table 2)were derived from water balance modelling andaverage annual outdoor water use data supplied byACT Department of Urban Services. The modelassumes that garden irrigation was undertaken whenthe soil moisture content fell below a predefined limit.This limiting value (trigger to start garden irrigation)was selected such that the estimated average annualirrigation demand was equal to the annual outdoorwater usage. Water efficient garden irrigation practiceswere not investigated as part of this study for largeblocks in O’Malley suburb.

4.3 Climate data

Historical daily climate data from 1984 to 2003 wereused for modelling purposes. The minimum andmaximum mean monthly temperatures vary from 08Cto 278C and the long-term average annual rainfall is640 mm.

5. Water balance modelling

5.1 Optimisation of raintanks

The optimisation of raintank sizes was carried outusing Aquacycle. Based upon the raintank size andvolumetric reliability relationship due to differentwater demands, the raintank sizes for various lot sizeswere obtained (Table 3). The volumetric reliability is

the proportion of the total demand on the tank supplythat actually comes from the tank. Raintank reliabilityincreases with size, as increasing the size of the tankincreases the amount of water available for use. Theoptimum size was assumed to correspond to that tanksize where further increases in size produced only asmall increase in reliability. Thus a volumetric relia-bility of 40% was considered to correspond to theoptimum tank size for 350 m2, 700 m2 and 1000 m2

allotments. For large allotment size, 5200 m2, nooptimum size was observed so a 20 KL tank wasassumed. The raintanks were considered for gardenirrigation and toilet flushing.

5.2 On-site measures –existing development (Wodenarea)

Water balance modelling (with low rate of demandmanagement) was conducted for the following options;raintanks, grey water for garden irrigation and alsocombination of both as onsite measures. The impacts ofon site measures on the long term average potable wateruse, wastewater discharge and stormwater dischargewere found to provide a reduction of up to 34% forpotable water demand, 20% reduction in wastewaterproduced and 23% for reduction in stormwater runoff(Table 4). The optimum reduction on stormwater flowsoccurred when only raintanks were simulated and theoptimum reduction in potable water use was observedwhen both raintanks and greywater were simulated.

The reductions in water, wastewater and storm-water flows were linearised to calculate the uptake raterequired for a 25% reduction in potable waterdemand. This analysis showed that a 25% reductionin potable water usage could be achieved with thefollowing combinations of water servicing options:

. 100% of the houses with raintanks.

. 75% of houses with both raintanks and grey-water reuse.

. 60% of houses with both raintanks and grey-water reuse and remaining 40% houses withgreywater reuse only.

Table 2. Average garden irrigation values for the variousblock sizes considered.

SuburbBlock size

m2

Garden irrigation

L/hh/d* L/c/d**

Woden 1,022 382 162Woden 712 311 133Woden 5,228 3,020 1,285Woden 367 152 65Gungaderra 450 97 37Gungaderra 300 86 33

*Litres/household/day and **Litres/capita (person)/day.

Table 3. Optimal rain tank sizes for irrigation and toiletflushing.

Allotment size(m2)

Optimal tanksize in Woden area

(kL)

Optimal tank size inGungaderra area

(kL)

350 3.5700 71000 145200 20All 5

152 A.K. Sharma et al.

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

Table 4. Effect of water saving measures in Woden township.

Water/Wastewater/Stormwater

Base caseNo raintanks

or greywater irrigation Raintanks in useGreywater for garden

irrigationRaintanks andgreywater in use

ML/yr ML/yr % reduction ML/yr % reduction ML/yr % reduction

Water 4765 3649 24% 4166 13% 3160 34%Wastewater 2836 2836 0% 2256 20% 2258 20%Stormwater 4875 3758 23% 4858 0% 3850 21%

Figure 2. Effect of on-site water saving measures combined with demand management (DM) in Woden.

Greywater usage for garden irrigation in all the housesalone cannot achieve a 25% reduction in potable waterdemand without additional provision of rainwatertanks. The final selection of the option to achieve 25%reduction in potable water would depend upon thetotal cost, reliability and environmental considera-tions. The reliability of greywater availability is higherthan the rainwater. The cost of a greywater systemdepends on the degree of treatment required, based onhealth and environmental guidelines. For known unitcost of raintanks and greywater systems, the costfunction can be solved to estimate the optimalcombination of raintanks and greywater systemssubject to 25% potable water reduction constraint.The optimisation of the total cost was not conducted inthis study to estimate the right mix of raintanks andgreywater systems.

The effect of introducing numerous demandmanagement tools on potable water demand was alsoestimated. To model this behaviour, per capita rates ofwater demand were modified to mimic those obtainedif a high uptake rate of demand management measureswere achievable. The modifications reduced per capita

in-house consumption from 220L/c/day to 172 L/c/dayfor the existing development and to 157 L/c/day for theproposed development area in Gungaderra. Thedemand management techniques represented in thismodelling included the use of water efficient dish-washers, showerheads, dual flush toilets and washingmachines, which resulted in further saving of 0.6 KL,26 KL, 18 KL and 10 KL per household, respectively,in annual water usage.

The water balance modelling with revised waterusage (high demand management) was conducted forWoden township to re-examine the impact of raintanksand greywater reuse alone and in combination withhigh demand management techniques. The waterdemand reduced by 12% with the implementation ofhigh demand management techniques only (Figure 2).The combination of demand management and rain-tanks reduced the water demand to 33%, thecombination with greywater to 22%, and the combina-tion of all three tools reduced the water demand by41%.

Comparing the alternative water servicing optionsrequired to achieve 25% water reduction objectives

Urban Water Journal 153

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

with high water management and usual water manage-ment techniques, the number of households required toadopt such options decreases. With high demandmanagement the uptakes rates for the various optionsare reduced to:

. �60% of the houses with raintanks

. �45% of houses with both raintanks and grey-water reuse

. �30% houses both raintanks and greywaterreuse for garden irrigation and �40% houseswith greywater reuse.

The OSD tanks were modelled at allotment scale toreduce peak stormwater flows of 5 year averagerecurrence interval (ARI) to pre-development level.The modelling indicated that peak stormwater flowswould be reduced to pre-development levels in allallotment types except one, if rainwater tanks wereprovided as per Table 3. The exception to this was forthe allotments with large lot size (5200 m2) as the lowusage from these tanks led to a high water level in theraintanks before a rain event. In the absence ofraintanks, the OSD sizes modelled to reduce 5 yearARI flows to pre-development stage for allotments ofsizes 5200 m2, 1000 m2, 700 m2 and 350 m2 were15 KL, 5 KL, 3.5 KL and 2.5 KL respectively. ThePURRS model was used for this analysis.

5.3 On-site measures – greenfield development(Gungaderra area)

The water balance was performed for each allotmentsize, firstly with no water saving measures and thenwith 100% uptake of 5 kL raintanks, greywater fromboth the bathroom and laundry for irrigation and100% of the raintank and greywater irrigationcombination.

A maximum reduction in potable water demand of34% was observed for allotments having both rain-tanks and greywater for garden irrigation, which alsoproduced a 20% reduction in stormwater discharge.Greywater use alone showed a 13% reduction inpotable water use and a 14% reduction in wastewater

flows (Table 5) and was less effective than raintanks forreducing potable water usage.

Similar to Woden area analysis, the reductions inwater, wastewater and stormwater flows were line-arised to calculate the uptake rate required for a 25%reduction in potable water demand. As described forWoden area development no cost optimisation wasconducted to estimate this optimal combination ofrainwater tanks and greywater systems for 25%potable water reduction. The potable water demandreduction of 25% could be achieved with some of thefollowing combinations of water servicing options:

. 85% of the houses with raintanks and 15% ofhouses with raintanks and greywater reuse. Thiswould also lead to 20% reduction in stormwaterflows and 2% reduction in wastewater flows.

. 58% of houses with raintanks and greywaterreuse and 42% of houses also with greywaterreuse. This option would result in 13% reductionin stormwater flows and 9% reduction in waste-water flows.

. 67% houses with both raintanks and greywaterreuse for garden irrigation.provides the 25%potable reduction and 11% reduction in storm-water flows and 14% reduction in wastewaterflows.

Again high uptake rates of raintanks and greywaterreuse systems are required for significant reductions inhousehold potable water demand to be achieved. Theuptake rates required are generally higher than forthe existing suburbs, as the smaller household blocks inthe greenfield site were modelled with lower outdoorwater use. Hence, less of the total household waterdemand is associated with garden irrigation, and thislowers the potable water reductions that can beachieved by substitution of potable water with rain-tank water or greywater for garden irrigation.

A water balance analysis for Gungaderra townshipwas also conducted with high rates of demandmanagement techniques, which resulted in waterdemand reduction by 23% (Figure 3). Smaller allot-ment sizes considered for Gungaderra mean that the

Table 5. Effect of on-site water saving measures on potable water demand-Gungaderra.

Water/Wastewater/Stormwater

Base case Raintanks in useGreywater for garden

irrigationRaintanks andgreywater in use

ML/yr ML/yr % reduction ML/yr % reduction ML/yr % reduction

Potable water 1588 1213 23% 1396 13% 1041 34%Wastewater 1362 1362 0% 1175 14% 1176 14%Stormwater 1771 1406 21% 1770 0% 1424 20%

154 A.K. Sharma et al.

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

amount of water used in the garden is small, andtherefore the high rate of demand management leadsto proportionally larger reductions in potable waterdemand than was the case for Woden. Demandmanagement in combination with rainwater usereduces potable demand by 44%; in combinationwith greywater by 34%; in combination with bothrainwater and greywater by 52%. With raintanks andgreywater reuse, 38% reduction in wastewater flowsand 16% reduction in stormwater flows can also beachieved. The 25% water reduction objective with highdemand management option can be achieved with thefollowing alternative servicing options:

. �5% houses with raintanks and greywater usage

. �10% houses with raintanks

. �10% houses with greywater usage

As was found with calculations for Gungaderra, therequired uptake rate of onsite systems reduceddrastically to achieve the water demand reductionobjective, when implemented with high uptake rates ofdemand management techniques.

Allotment scale measures such as OSDs similar toWoden development were considered for Gungaderaarea to lower the peak stormwater flows to pre-development levels with or without raintanks. For the450 m2 and 300 m2 allotments, OSDs of 4 kL and 2 kLcapacities without raintanks and 3 kL and 1 kL withraintanks were required to reduce 5 year ARI peak flowsto predevelopment level. Raintanks alone could notreduce 5 year ARI peak flows to predevelopment level inGungadera area due to reduced outdoor water usage.The PURRS model was also used for this analysis.

It can be concluded from water balance modelling inWoden and Gungaderra areas that the reduction inpotable water demand due to on-site measures isdirectly proportional to lot sizes, pervious area ratio,roof sizes and occupancy rate irrespective of existing orgreenfield developments. Raintanks alone were effectivetools in reducing peak stormwater (5 year ARI) flows topredevelopment level if the impervious area ratio wasless than 0.35. In the case of lots with higher imperviousratios (Gungaderra area), raintanks alone could notreduce peak stormwater flows to predevelopment level.

5.4 Catchment (development area) scale measures

Catchment scale modelling was conducted for Wodento estimate the combined impact of raintanks as on-sitemeasures and stormwater use from local ponds forpublic open space irrigation. Both, the base case wherepotable water is used for irrigation of POS, andraintanks along with stormwater for POS irrigationfrom local stormwater ponds were modelled usingAquacycle.

Raintanks and POS irrigation from stormwaterponds reduced stormwater flows by 27% (Table 6).This compares to a 21% reduction in stormwater if theeffect of raintanks alone were considered. Therefore,irrigation of the POS from the stormwater pondsprovides a further reduction in stormwater dischargeof 6% and a potable water reduction of 25%.

For Gungaderra development, in addition toallotment raintanks and greywater, POS irrigationfrom local stormwater ponds and the use of treatedwastewater for garden irrigation and toilet flushingwas also modelled. Treated wastewater for garden

Figure 3. Effect of on-site water saving measures combined with demand management (DM) in Gungaderr.

Urban Water Journal 155

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

watering and toilet flushing provided a slightly greaterreduction in potable water demand at 36% (Table 7)compared to the 34% reduction observed for grey-water reuse and raintanks (Table 5).

The reductions up to 44% in wastewater dischargewere predicted when using treated wastewater for bothgarden irrigation and toilet flushing. Maximum storm-water reduction of 24% was achieved by the use ofrainwater tanks and stormwater irrigation of POS(Table 7).

5.5 Stormwater quantity and quality modelling

5.5.1 Existing development (Woden area)

The conceptual layouts of stormwater flows withvarious combinations of treatment trains were devel-oped for modelling purposes. It was found that acombination of ponds, GPTs and bioretention systems

provided the maximum improvement in stormwaterquality (Table 8) with reductions in TSS, TN and TPloads of 71%, 32% and 42%, respectively, in compar-ison to urban development with no stormwater treat-ment provisions. However, GPT and ponds alone gavepredicted TSS, TN and TP load reductions of 68, 26 and38% respectively. This was considered the preferredoption, as the additional provision of bioretentionsystems did not significantly reduce contaminant loads.

None of the treatment trains alone were predictedto provide stormwater quality equivalent to thepredevelopment level. Thus, the combined impact ofraintanks and stormwater use from ponds for POSirrigation GPT and ponds was modelled, and found topredict contaminant loads of less than the predevelop-ment case. The provision for raintanks and stormwaterreuse for irrigation also resulted in 30% reduction inmean daily stormwater flows to 0.131 m3/s in compar-ison to typical urban development flows, but the mean

Table 6. Water balance on catchment scale scenarios: Woden area.

Option

Potable water Wastewater Stormwater

ML/yr % reduction ML/yr % reduction ML/yr % reduction

Base case 5524 0 3340 0 4880 0Raintanks and stormwater usefrom ponds for POS irrigation

4167 25 3340 0 3565 27

Table 7. Water balance on catchment scale scenarios – Gungaderra area.

Scenarios

Potable Wastewater Stormwater

ML/yr%

reduction ML/yr%

reduction ML/yr%

reduction

Stormwater and rainwater 1267 20 1355 0 1419 24Wastewater and stormwater 1022 36 759 44 1764 5

Table 8. Reduction* of stormwater contaminant loads and concentrations: Woden area.

ParametersMean flow

(m3/s) TSS (kg/day) TN (kg/day) TP (kg/day) TSS (mg/L) TN (mg/L) TP (mg/L)

Total area as greenfield 0.086 962 18.7 2.34 24.6 2.13 0.168Developed area withno treatment measures

0.186 2390 41.6 5.48 42.2 2.2 0.193

Flow %reduction

TSS %reduction

TN %reduction

TP %reduction

TSS %reduction

TN %reduction

TP %reduction

GPT þ Ponds(4 ha þ 3 ha þ 3 ha)

0 68 26 38 59 26 25

GPT þ Ponds(4 ha þ 3 ha þ 3 ha) þBioretention systems

0 71 32 42 61 33 26

Buffers 0 37 18 25 28 16 24Buffers þ Swales 0 41 18 27 26 21 20Buffers þ Bioretention 0 45 29 34 57 70 67

*Reductions in comparison to development with no treatment measures.

156 A.K. Sharma et al.

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

daily flows were still higher than predevelopmentconditions. The MUSIC model was applied forstormwater quantity and quality assessment.

5.5.2 Existing development (Gungaderra area)

To conceptualise the improvement in the catchmentstormwater runoff quality, two treatment trains wereconsidered, firstly GPTs and ponds and also GPTs,ponds and swales. The dimensions of swales adoptedfor modelling were top width 4 m, bottom width 1 mand depth 0.5 m with a vegetation of 0.25 m.

Significant reductions were predicted in nutrientloading with GPTs and ponds as treatment measureswith TSS, TN and TP loads reduced by 82%, 33%and 50%, respectively, in comparison to the base case(Table 9). However, the predicted TN and TP loadswere still higher than the pre-development conditions.Slightly greater reductions were observed with theaddition of swales, with TSS, TN and TP loads reducedby 85%, 36% and 52%, respectively, but this improve-ment was only marginal and it was concluded that theprovision of ponds and GPTs was an effective option.

Similar to Woden area, none of the stormwatertreatment measures were predicted to reduce thecontaminant loads to predevelopment levels in Gun-gaderra area. Thus the impact of the rainwater tanksalone and in combination with stormwater reuse forPOS irrigation and GPTs and ponds, on stormwaterquantity and quality were also modelled. The con-taminant loads reduced by 25% with raintanks alone.With the combination of raintanks and stormwaterreuse for POS from ponds the contaminant loads werereduced below predevelopment levels and the storm-water flows reduced by 28% in comparison to typicalurban development. However, the mean daily flows

were still higher than the predevelopment flows. Thus,the raintanks alone are effective in contaminant loadreduction.

6. Discussion

Raintanks are not only effective in reducing potablewater demand but also in reducing low intensity peakstormwater flows and stormwater contaminant loadsin smaller allotments to predevelopment levels.

The demand management techniques are veryeffective means of achieving water demand reductionobjectives. High uptake rates for raintanks and grey-water reuse systems are required for significantreductions in household potable water demand inexisting developments. In response to this and otherinformation the ACT Government is actively promot-ing the use of rainwater and greywater. Planningcontrols have been amended to facilitate installation ofraintanks and greywater systems, and a Governmentfunded rebate scheme for raintanks has commenced.

Similarly, high uptake rates of raintanks and grey-water reuse systems are required for significant reduc-tions in household potable water demand in greenfielddevelopment. The uptake rates required are generallyhigher than for the existing suburbs, as the smallerhousehold blocks in the greenfield site were modelledwith lower outdoor water use. Hence, less of the totalhousehold water demand is associated with gardenirrigation, and this lowers the potable water reductionsthat can be achieved by substitution of potable waterwith rainwater or greywater for garden irrigation.

In case of greenfield developments, the combina-tion of wastewater reuse and stormwater use fromcatchment (development) scale ponds is very effectivetool in potable water reduction in comparison to

Table 9. Reduction* of stormwater contaminant loads and concentrations – Gungaderra area.

Treatment optionMean flow

(m3/s) TSS (kg/day) TN (kg/day) TP (kg/day) TSS (mg/L) TN (mg/L) TP (mg/L)

Total area as grasslandwith no treatment

0.017 173 3.27 0.4 24.6 1.89 0.15

Urban developmentwith no treatment

0.057 730 12.2 1.61 41.3 1.7 0.15

Flow %reduction

TSS %reduction

TN %reduction

TP %reduction

TSS %reduction

TN %reduction

TP %reduction

Total area as grasslandwith GPT þ Ponds

0 86 45 52 50 33 12

Urban developmentwith GPT þ Ponds

0 82 33 50 66 14 9

Urban developmentwith GPT þ Ponds þSwales

0 85 36 52 67 16 10

*Reduction in comparison to values for urban development with no stormwater treatment.

Urban Water Journal 157

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

on-site rainwater and greywater usage. In this studydevelopment scale wastewater reuse reduced waste-water flows by 44% in comparison to only 14% whenon-site rainwater and greywater systems are used. Thecombined application of stormwater use and waste-water reuse provides improved ecological benefits incomparison to on-site rainwater and greywater use.

From a policy perspective for Environmental ACT,the value of household scale measures needs to bebalanced with relying on individual householders tooperate such measures effectively. Should householdscale measures not perform, retrofitting of measures inpublic spaces may be required. In consequence thegeneral approach applied in Canberra has been toencourage rather than mandate individual householdmeasures, on the basis that measures adopted volun-tarily are more likely to be maintained.

The reduction in potable water demand due to on-site measures is directly proportional to lot sizes,pervious area ratio, roof sizes and occupancy rateirrespective of existing or greenfield developments.Raintanks alone were predicted to be effective tools inreducing peak stormwater (5 year ARI) flows topredevelopment level if the impervious area ratio wasless than 0.35. In the case of lots with higher imperviousratios raintanks alone are not capable of reducing peakstormwater flows to predevelopment levels.

No single modelling tool is available to analyse thetotal water cycle for quantity and quality in an urbandevelopment. A combination of modelling tools isrequired to cover all aspects of water, wastewater andstormwater quantity and quality. In this study Aqua-cycle was applied to model water, wastewater andstormwater flows for various servicing options andalso rainwater tanks optimal sizes for various lots.MUSIC model was used to conceptualise stormwatertreatment units for stormwater quantity and qualityand PURRS model was used to conceptualise on-sitedetention tanks and also to analyse impact of rain-water tanks on stormwater peak reduction. Theraintank sizes and stormwater/rainwater reuse quan-tity estimated using Aquacycle was used as an input toMUSIC model to conceptualise treatment trains forstormwater quality and quantity. The information onraintanks was also used as an input to PURRS modelto conceptualise OSD tanks and to estimate allotmentscale peak stormwater flows at predefined ARI. Themodels were validated and parameters adjusted forstormwater flows.

7. Conclusions

The purpose of the current study was to assistEnvironment ACT in setting broad policies for futurewater services in Canberra. Environment ACT

amended planning controls to facilitate installation ofraintanks and greywater systems, and commenced aGovernment funded rebate scheme for raintanks as aresult of this study. The investigation provided anassessment of potential preferences of alternative waterservicing options. A number of allotment and catch-ment scale options were investigated to achieve setobjectives, and the extent to which these techniquesmight need to be implemented. The effect of differentlevels of demand management on the changes broughtabout by alternate water measures was also examined.A combination of three models were used in this studyto model complete water cycle for quantity and qualityas no single model was available to analyse all aspectsof water cycle. The computer models used in this studywere found suitable for water balance analysis, storm-water quality and quantity estimations and allotmentscale stormwater flow estimation including on-sitedetention tank sizing. The output from one modelwas used as an input to other models.

High demand management is a very effective toolfor potable water reduction at allotment scale, but itrequires the use of highly water efficient appliances.Such an option may be possible in greenfield develop-ments by financial incentives and/or regulatory ar-rangements. High uptake rates are required fromraintanks to achieve the 25% potable water reductionobjectives. Use of on-site greywater reuse systems andraintanks was more efficient than either raintanks oron-site grey water reuse systems alone, while greywatersystems alone could not achieve the potable waterreduction objective. Raintanks are more efficient forCanberra than on-site greywater reuse for gardenirrigation. Coupling the use of raintanks with highuptake rates of demand management significantlylowered the uptake rates of raintanks required tomeet the 25% reduction in potable water demand. Thepotable water reduction due to on-site measures wasfound to be proportional to lot size, if it is assumedthat the irrigation demand increases with increased lotsizes.

The raintanks not only lead to lower potable wateruse, but also significantly reduced peak stormwaterflows at the allotment scale. Moreover, on-site deten-tion systems are an effective means of reducing peakflows to predevelopment levels. The provision ofadditional treatment trains such as bioretention systemor swales, ponds and GPTs did not result in anysignificant additional reduction in nutrient loadsbeyond the implementation of ponds and GTPs alone.However, the rainwater/stormwater reuse along withthe treatment trains significantly decreases nutrientfrom stormwater runoff.

For greenfield sites, wastewater recycling for non-potable use was the most effective and reliable means

158 A.K. Sharma et al.

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5

for reducing potable water demands and wastewaterflows. This option is generally considered too expen-sive for implementation in existing suburbs, although ithas not been costed.

Following this work the ACT Government estab-lished an ambitious target for potable water reductionand wastewater reuse. The demand managementmeasures modelled in this study are being implementedthrough an initiative, which will continue over numberof years. In addition, this work is informing infra-structure requirements for both allotment and publicopen space measures currently being formalised inguidelines.

Acknowledgement

The authors wish to thank Ross Knee from EcowiseEnvironmental, John Neal and Allan Caulley from theACT Planning and Land Authority, Jacqui Goonery fromActewAGL, and Steve Bresnan from Roads ACT for theirco-operation for making this study possible.

References

Alegre, N., McIntosh, B., Thomas, J.S., Hardwick, I., andRiley, S., 2004. Strategic options for sustainable watermanagement at new developments: the application of asimulation model to explore potential water savings.Water Science and Technology, 50 (2), 9–15.

Coombes, P., On-site water balance modelling usingPURRS. In: Workshop on Rainwater Tank and Waste-water Reuse Modelling – Key Elements in Water CycleAnalysis and Sustainable Development, 19–20 June 2003,Melbourne, Australia.

Dixon, A., Butler, D., and Fewkes, A., 1999. Water savingpotential of domestic water reuse systems using grey-water and rainwater in combination. Water Science andTechnology, 39 (5), 25–32.

Grimmond, C.S.B., Oke, T.R., and Steyn, D.G., 1986.Urban water balance 1. a model for daily totals. WaterResources Research, 22 (10), 1397–1403.

Hardy, M.J., Kuczera, G., and Coombes, P.J., 2005.Integrated urban water cycle management: the Urban-Cycle model. Water Science and Technology, 52 (9), 1–9.

Manios, T. and Tsanis, I.K., 2006. Evaluating waterresources availability and wastewater reuse importancein the water resources management of small Mediterra-nean municipal districts. Resources Conservation andRecycling, 47 (3), 245–259.

Mitchell, V.G., Mein, R.G., and McMahon, T.A., 2001.Modelling the urban water cycle. Environmental Model-ling and Software, 16 (7), 615–629.

Mitchell, V.G., McMahon, T.A., and Mein, R.G., 2003.Components of the total water balance of an urbancatchment. Environmental Management, 32 (6), 735–746.

Parkinson, J., Schutze, M., and Butler, D., 2005. Modellingthe impacts of domestic water conservation on thesustainability of the urban sewerage system. Water andEnvironment Journal, 19 (1), 49–56.

Rueedi, J., Cronin, A.A., Moon, B., Wolf, L., and Hoetl, H.,2005. Effect of different water management strategies onwater and contaminant fluxes in Doncaster, UnitedKingdom. Water Science and Technology, 52 (9), 115–123.

Sharma, A., Grant, A., Gray, S., and Mitchell, G., 2005.Sustainability of alternative sewerage and water servicingoptions-Yarra Valley Water-Kalkallo and Box Hill PACdevelopments. CSIRO, Australia.

Villarreal, E.L. and Dixon, A., 2005. Analysis of a rainwatercollection system for domestic water supply in Ring-dansen, Norrkoping, Sweden. Building and Environment,40 (9), 1174–1184.

Wong, T., Coleman, J., Duncan, H., Fletcher, T., Siriwardena,L., and Wootton, R., Music-model for urban stormwaterimprovement conceptualisation, version 1.0, user manual2002 (CRC for Catchment Hydrology, Urban Storm-water Quality Program, Melbourne).

Urban Water Journal 159

Dow

nloa

ded

by [

95.1

56.1

82.1

77]

at 0

1:52

28

Febr

uary

201

5