Marine Policy ] (]]]]) ]]]–]]]

Contents lists available at SciVerse ScienceDirect

Marine Policy

0308-59

http://d

n Corr

E-m

Pleasthe s

journal homepage: www.elsevier.com/locate/marpol

How robust is the environmental impact assessment process in SouthAustralia? Behind the scenes of the Adelaide seawater desalination project

Jochen Kampf n, Beverley Clarke

School of the Environment, Flinders University, PO Box 2100, Adelaide, SA 5001, Australia

a r t i c l e i n f o

Article history:

Received 26 April 2012

Received in revised form

17 August 2012

Accepted 17 August 2012

Keywords:

Environmental Impact Assessment

Seawater desalination

Marine conservation

Operating licence

South Australia

7X/$ - see front matter Crown Copyright & 2

x.doi.org/10.1016/j.marpol.2012.08.005

esponding author. Tel.: þ61 8 8201 2214; fax

ail address: jochen.kaempf@flinders.edu.au (J

e cite this article as: Kampf J, Clarkecenes of the Adelaide seawater desa

a b s t r a c t

This work tests the robustness of policies and procedures designed to protect South Australia’s marine

environment through a case study of the Adelaide Desalination Plant—the most expensive (�A$1.8

billion) infrastructure project in South Australia’s history. Although this project has been subject to an

Environmental Impact Assessment (EIA)—the highest level of assessment in Australia—on inspection it

appears that the current operating licence for the desalination brine discharge breaches Government

approval conditions and ignores the collective expert scientific advice of the project’s Environmental

Impact Statement (EIS). Hence, the EIA process in South Australia for this project is flawed.

Improvements could be made to the South Australian system by including the requirements for

operating licences as an integral part of the EIA.

Crown Copyright & 2012 Published by Elsevier Ltd. All rights reserved.

1. Introduction

Following a period of prolonged drought in Australia (2003–2009), and with predictions of a drying climate, state govern-ments have actively pursued alternatives to natural water sources(rain and river) to augment and guarantee potable supply forseveral Australian capital cities. One of the most popular butcontroversial alternative supply sources is desalination. MostAustralian capital cities have recently constructed or are in theprocess of constructing seawater desalination plants. Thisresponse is in keeping with worldwide trends; water scarcityhas promoted development of alternative water sources withdesalination a common option [1]. The Arabian Gulf, the Medi-terranean and Red Seas, coastal China, California and Australia arereferred to as centres of desalination ‘activity’ [2].

Due to the scale of building construction associated withdesalination plants and the potential for their environmentalimpact, such major developments tend to undergo the moststringent of pre-approval assessment. There are legally consti-tuted approvals processes for major developments in Australia:the Environmental Impact Assessment (EIA). There is an inter-nationally recognised process for EIA but the legislative andadministrative nuances in a particular place affect how thisprocess translates into practice. One of the weaknesses of EIA isfollow-up or post decision monitoring [3]. This paper uses a

012 Published by Elsevier Ltd. All

: þ61 8 8201 2676.

. Kampf).

B. How robust is the envirlination project. Mar. Polic

recently approved South Australian major development, theAdelaide Desalination Plant (designed to reduce reliance on theRiver Murray and augment the city of Adelaide’s water supply), toassess the adequacy of the EIA process towards protecting thelocal coastal environment.

EIA in South Australia is legislated under Section 46 of theDevelopment Act 1993. Depending on the scale of a proposeddevelopment there are three possible levels of assessment, theEnvironmental Impact Statement (EIS) the Public Environment

Report (PER) and the Development Report (DR) [3]. The AdelaideDesalination Plant was assessed at the highest level triggering therequirement for an EIS. The South Australian EIA system requiresthat at least one public meeting accompany an EIS to explain thedevelopment proposal and an invitation for public comment onthe report (the EIS). In South Australia, the Planning Ministermakes the declaration that a development proposal is of ‘majorenvironmental, social or economic importance’, thus triggeringthe EIA process. Once such a declaration is made, the decision-making process follows the standard steps of data gathering,reporting and assessment. Much of the data gathering andreporting is undertaken by the proponent. After public commentshave been acknowledged and addressed, the proponent submitsthe appropriately amended final EIS to the Planning Ministerfor assessment. The quality of proponent reports is assessed bythe relevant government agency administering the EIA process(currently the State Department of Planning and Local Government).The Minister is provided with an Assessment Report for authorisa-tion [3]. This phase of the process does not allow for further externalinput; there may continue to be behind-the-scenes negotiations,

rights reserved.

onmental impact assessment process in South Australia? Behindy (2012), http://dx.doi.org/10.1016/j.marpol.2012.08.005

J. Kampf, B. Clarke / Marine Policy ] (]]]]) ]]]–]]]2

such as condition setting, between the government and the pro-ponent. In South Australia the Governor (effectively the StateCabinet) makes the final decision. A nuance of the South Australiansystem is that the Governor’s decision is final; there are no rightsof appeal [3]. The Federal Government can intervene in the SouthAustralian EIA process if a development proposal triggers theAustralian Government Environmental Protection and Biodiversity Act

(EPBC 1999) over matters of ‘national environmental significance’.After governmental approval, and hence after the pre-decision phaseof EIA, certain activities, such as the functioning of seawaterdesalination plants, are made subject to an operating licence, whichis negotiated between the plant’s operators and the state’s Environ-mental Protection Authority.

Using the Adelaide desalination project as a case study thispaper addresses the important question as to whether or not thefinal operating licence of the desalination plant reflects therecommendations made in the EIS. Discrepancies between EISrecommendations and licence conditions would shed doubt intothe proper functioning and robustness of South Australia’s EIAprocess, especially for post decision follow-up.

Fig. 1. The Gulf St. Vincent region indicating the location of the Adelaide

Desalination Plant near Port Stanvac.

2. The Adelaide desalination project: background andapproval process

2.1. Desalination plant details

The Adelaide Desalination Plant became operational in late2011. In 2009, a government announcement was made that theoriginally proposed 50 GL annual output would be doubled; at fullcapacity the plant can produce 100 GL of drinking water per year,corresponding to approximately 270 ML per day. The Adelaideplant has been constructed at Port Stanvac, Lonsdale, SouthAustralia, about 20 km south of the city. Adelaide is situated onthe eastern shore of Gulf St. Vincent (Fig. 1). AdelaideAqua, theplant’s operator, will complete the commissioning of the 50 GLplant (135 ML/d) mid-year 2012. The upgrade to 100 GL will notbe complete until the end of the 2012.

Similar to other major Australian seawater desalination plants,the Adelaide plant uses reverse osmosis technology (a filteringtechnique) to attain fresh water from seawater. Seawater ispushed under high pressure through membranes. This enableswater molecules to pass but blocks molecules of salt. Hence, thisprocess separates seawater into two new solutions: freshwaterand hypersaline brine, commonly referred to as ‘desalinationbrine’. Of the seawater that feeds the process, a plant recoveryefficiency of 50% implies that half the seawater is converted intofreshwater, whereas the other half attains twice the salt concen-tration, or salinity. Typical recovery efficiencies range between25% and 50%. The desalination brine is typically discharged backinto the sea, together with other chemical bi-products of thedesalination process such as antiscalents. Marine discharge ofdesalination brine is the most common and ‘‘cheapest’’ option forhandling the unwanted desalination brine.

2.2. Marine impacts and environmental legislation

Desalination brine is heavier than seawater. The most immedi-ate environmental hazard associated with desalination brinedischarges is the formation of so-called ‘brine underflows’ [4].This flow is associated is the formation of a thin layer of hypersa-line water spreading along the seafloor and becoming depleted indissolved oxygen. For example, marked reductions in dissolvedoxygen levels have been observed in vicinity of the outlet ofAustralia’s first desalination plant in Cockburn Sound, WesternAustralia [5].

Please cite this article as: Kampf J, Clarke B. How robust is the envirthe scenes of the Adelaide seawater desalination project. Mar. Polic

Modern discharge technology consists of multiple portslocated above the seabed through which the desalination brineis injected into the ambient water column. This process, beingsimilar to that of fountains, facilitates the initial mixing ofdesalination brine with ambient seawater. The mixing product,still being denser than the ambient water, tends to fall back to theseabed at some distance from the discharge ports. Dilution is ameasure of how much ambient seawater is mixed with the brineconcentrate and the chemical contaminants it contains. A higherdilution generally implies less adverse marine impacts.

The Environmental Protection (Water Quality) Policy 2003 issubordinate legislation supporting the Environment Protection Act

1993. The policy provides for the development of environmentalvalues and water quality objectives for South Australian waters.The policy outlines additional regulations for point source anddiffuse pollution to ensure achievement of water quality objec-tives. According to this legislation, marine pollution from a point-source discharge is permitted within a horizontal distance of100 m, being referred to as the ‘mixing zone’. This implies thatthere should be no or only little adverse marine impacts at theedge of this mixing zone. For desalination discharges, this definesa certain minimum dilution requirement at the edge of themixing zone, also called ‘safe dilution value’ or ‘species protectiontrigger value’ (SPTV). For point source pollution, such as desalination

onmental impact assessment process in South Australia? Behindy (2012), http://dx.doi.org/10.1016/j.marpol.2012.08.005

Table 1Chronology of the EIA process for the Adelaide desalination plant.

Event Date

Declaration as major development 17 April 2008

Release of guidelines by Development

Assessment Commission with level

of assessment determined – EIS

required

18 September 2008

Release of EIS document by

proponent for public comment

12 November 2008–24 December

2008Public meeting was held 17

November 2008

Response by proponent to public and

agency submissions on EIS

document

26 February 2009

Assessment report by Minister 26 February 2009

Decision by Governor 26 February 2009

Delegate of Governor Decision 12 March 200911 June 2009

J. Kampf, B. Clarke / Marine Policy ] (]]]]) ]]]–]]] 3

brine, the determination of this SPTV plays a central part in an EIS.Typically both hydrodynamic modelling and laboratory-based eco-toxicology testing are used to derive this threshold value. Operatinglicences are often based on certain trigger values.

2.3. Marine environment and choice of location

The characteristics of the gulf into which the Adelaide Desa-lination Plant releases its brine are worthy of close consideration.South Australian gulfs (Gulf St. Vincent at its neighbour SpencerGulf) are large inverse estuaries of a salinity exceeding that in theadjacent ocean [6,7]. Water exchanges with the ambient ocean,also called ‘‘flushing’’, are remarkably slow. For instance, it takesabout 1 year to flush the upper reaches of Gulf St. Vincent. Theflushing time of Adelaide Metropolitan waters is around 3–5months [8]. In comparison, a stretch of water body in the openocean the same length as Gulf St. Vincent (around 145 km) wouldhave a flushing time of only 10–20 days. Hence, in terms offlushing, Gulf St. Vincent is substantially sheltered from oceanflows as well as from vigorous swell waves of the Southern Oceanthat would otherwise stir the water column.

The tides in South Australian gulfs display distinctive features.Semidiurnal lunar and solar tidal constituents are of a similarmagnitude. Owing to interplay between these constituents, tidalcurrents significantly weaken on a roughly fortnightly basis [7].Locally, this feature is known as ‘‘dodge tides’’, first reported byexplorer Matthew Flinders in 1802. Scientifically, dodge tides arepronounced examples of so-called ‘‘neap tides’’. Periods ofenhanced tidal flows between neap tides are termed ‘‘springtides’’. Tidal flows provide a means of vertical stirring of point-source discharges, such as desalination brine, with ambientwater. Dodge tide situations are therefore most critical periodsduring which the absence of tidal stirring may promote thecreation of brine underflows.

South Australian gulfs accommodate a rich variety of endemic(region-specific) marine flora and fauna. Gulf St. Vincent has adiverse range of habitats and is a globally significant region fortemperate biodiversity. South Australian gulfs have a high level ofendemism—or uniqueness of species—of over 85%, compared toonly 15% in tropical areas such as the Great Barrier Reef [9]. GulfSt. Vincent contains some of the most extensive areas of tempe-rate mangrove forests and seagrass meadows in Australia—

habitats of considerable ecological and economic importance [10].South Australia also has a strong, viable commercial fishing

industry. In 2007–08 the State’s commercial wild fisheries wereworth A$207.5 million. Commercial wild-catch fisheries in SouthAustralia include marine-based fisheries for species such asabalone, garfish, King George whiting, mullet, pilchards, prawns,rock lobster, snapper, tuna, tommy ruff and sharks and variousriverine and freshwater fisheries. The industry is an importantsource of employment, both directly and via support industries,and as an earner of valuable export dollars. Unfortunately,urbanisation of the Adelaide region has resulted in an overalldegradation of near-shore coastal water quality that has causedwidespread losses of seagrass and mangrove forests in Gulf St.Vincent [11]. Physical factors (slow flushing, pronounced neaptides) and ecological factors (high biodiversity, many endemicmarine species) fuelled concerns from the community and marinescientists about potential adverse marine impacts of the AdelaideDesalination Plant on the fragile and ecologically importantenvironment of the gulf.

International experts [2] make the following recommenda-tions regarding the choice of location for desalination inlets andoutlets. First, ecosystems or habitats should be avoided, if they areunique within a region or worth protecting on a global scale,inhabited by protected, endangered or rare species, important in

Please cite this article as: Kampf J, Clarke B. How robust is the envirthe scenes of the Adelaide seawater desalination project. Mar. Polic

terms of their productivity or biodiversity, or if they play animportant role as feeding or reproductive areas in the region.Second, the site should furthermore provide sufficient capacity todilute and disperse the salt concentrate and to dilute, disperseand degrade any residual chemicals. The load and transportcapacity of a site will primarily depend on water circulation andexchange rate as a function of currents, tides, surf, water depthand shoreline morphology. In general, exposed rocky or sandyshorelines with strong currents and surf may be preferred overshallow, sheltered sites with little water exchange. According tothese principles, the positioning of the Adelaide DesalinationPlant is less than ideal.

2.4. The EIA process for the Adelaide desalination plant

The EIA process for the Adelaide desalination plant culminatedwith governmental approval in early 2009, subject to a number ofenvironmental performance conditions and monitoring require-ments published in the Government Gazette [12].

Table 1 provides an overview of the time taken for theAdelaide desalination plant to progress through the South Aus-tralian EIA system. In total, from triggering to decision, theprocess was completed within a year; this is one of the mostrapid EIAs on record in South Australia. It took less than twomonths from the release of the government’s guidelines docu-ment (which directs the content of the EIS) to the publication ofthe EIS. A study [13] of all EIAs undertaken in South Australiabetween 1994 and 2005 showed that the average time from theissue of the Guidelines to release of the EIS was 67 weeks. For theAdelaide desalination Plant the Governor’s final decision wasgiven on the same day the government’s assessment report wasreleased.

After approval, the then Minister for Urban Development andPlanning Paul Holloway (now resigned—and the agency has sincebeen restructured and renamed) said that the Adelaide Desalina-tion Plant, approved by the State Governor, was subjected to the‘‘most transparent and robust assessment process available underSouth Australian development laws’’.

2.5. Dilution requirements according to the EIS

In numerous places, the EIS of the Adelaide desalination plantrefers to a minimum dilution ratio of 50:1 [14]. This implies thateach litre of desalination brine is mixed with at least 50 l ofambient seawater at the edge of the mixing zone (at 100 mdistance from the discharge). In particular, the documentationstates that ‘‘The target initial dilution is 50:1 at the seabed under

onmental impact assessment process in South Australia? Behindy (2012), http://dx.doi.org/10.1016/j.marpol.2012.08.005

Table 2Relationship between salinity levels above ambient and dilution ratios for an

ambient salinity of 37 PSU after Eq. (A7).

Salinity level above

ambient (PPT)

Dilution ratio for a

recovery efficiency of 50%

Dilution ratio for a

recovery efficiency of 25%

0.3 122:1 40:1

0.6 61:1 20:1

0.9 40:1 13:1

1.2 30:1 9:1

1.3a 27:1 8:1

1.5 24:1 7:1

1.8 20:1 6:1

2.1 17:1 5:1

2.4 14:1 4:1

2.7 13:1 o4:1

J. Kampf, B. Clarke / Marine Policy ] (]]]]) ]]]–]]]4

all operating and tidal current conditions and the final designwill be required to comply with this specification’’ ([14], p. 27)and ‘‘Based on all ecotoxicological tests to date, the dilution of50:1 is sufficient to avoid any toxic effects on local marine species’’([14], p. 77).

Confusingly, the EIS also contains some more flexible statementsallowing for negotiations with the Environmental ProtectionAuthority (EPA) such as: ‘‘The outfall must achieve the requiredinitial dilution of 50:1 at the seabed, or as otherwise agreed withthe EPA, under all current scenarios for the full range of operatingconditions/flows’’ ([14], p. 12).

Minister Holloway’s and his department’s assessment of theEIS ([15], p. 40) specified that: ‘‘In conclusion, should the projectbe approved it is recommended that conditions ensure the finaldiffuser location and design achieve the Environmental Objectivesand Performance Criteria outlined in the EIS, including a mini-mum dilution of 50:1 at the seabed, which includes the subtidalreefs.’’

The governmental approval conditions ([12], p. 7) specifyunder the section Outfall Structure (c) that: ‘‘The outfall mustachieve the required initial dilution of 50:1 (or the dilution rateidentified by ecotoxicity assessments, if higher) at the seabed,under all current scenarios for the full range of operating condi-tions/flows and the plume height must not reach the watersurface at any time.’’

In addition to this, the proponent’s subsequent report ([16], p.38) confirms this requirement stating that: ‘‘The design achievesthe required initial dilution target of at least 50:1 with anequivalent 58:1 due to higher recovery rates into the localambient water column under all current scenarios for the fullrange of operating conditions and/or flows. A bypass system hasbeen incorporated into the plant design as an assurance that the58:1 criterion is met during lower plant production rates.’’ Notethat the above dilution thresholds refer to the first point ofcontact of the brine plume with the seabed. This distance ispredicted to be within 30 m from the discharge outlets, which iswell within the regulatory mixing zone. Consequently, the dilu-tion ratio at the edge of the mixing zone (100 m from thedischarge) should be even higher than 58:1.

a The operating licence uses this level as an upper threshold value.

Fig. 2. Relationship between salinity levels above ambient and dilution values for

different recovery efficiencies. Values calculated from Eq. (A7) for an ambient

salinity of 37 PPT. The vertical line denotes the dilution range corresponding to the

1.3 PPT salinity-anomaly threshold used in the operating licence. The horizontal

line denotes the range of salinity anomalies that would correspond to a dilution

ratio of 50:1.

3. The Adelaide desalination project: Reality check

3.1. Dilution requirements according to the operating licence

After these strong documented commitments by the propo-nent and the State Government, it is now interesting to verify thedilution criterion in the actual operating licence [17] of theAdelaide Desalination Plant. Surprisingly, the licence does notrefer to any dilution criterion. Instead, we find a condition formaximum salinity levels above ambient. In detail, the licence(valid from 1 December 2010 to 30 November 2030) reads underpoint 2: ‘‘If average salinity at any point 100 m from the diffuserstructure exceeds 1.3 PPT (parts per thousand) above ambientsalinity for a six hour period, then the Licencee must (1) notify theEPA within six hours, and (2) take appropriate corrective action tomanage salinity in the receiving environment.’’ If this situationlasts for 24 h or longer, point 3 in the licence requests that thedesalination discharge be stopped within six hours.

Dilution ratios can be inferred from salinity measurements. Ifthe discharge salinity is unknown, it can be estimated from theplant’s recovery efficiency (see Appendix). Hence, in principle, thelicence’s salinity trigger value can be converted to dilution ratios.

Before inspecting salinity monitoring data, we determinedilution ratios that correspond to the licence’s salinity triggervalue. Owing to missing information, the salinity of the discharge

Please cite this article as: Kampf J, Clarke B. How robust is the envirthe scenes of the Adelaide seawater desalination project. Mar. Polic

has to be estimated from a plant’s recovery efficiency in a range of25–50% and an ambient (intake) salinity that seasonally naturallyvaries between 36.0 and 38.0 PPT [7]. To this end, we chose anambient salinity of 37.0 PPT in all estimates, noting that varia-tions in ambient salinity have little (o3%) influence on theresults.

Table 2 shows dilution ratios as a function of salinity levelsabove ambient for recovery efficiencies of 25% and 50%. Itbecomes apparent that the 1.3 PPT salinity-anomaly thresholdof the operating licence of the Adelaide Desalination Plant permitsdilutions as low as 8:1 for a recovery efficiency of 25%, or 27:1 fora recovery efficiency of 50%. These dilution ratios are significantlylower (2 to 6 times) than originally assured by the project’sproponent and the State Government.

The curves in Fig. 2 are continuous versions of Table 2’s data.Two features become apparent from these curves. First, a salinity-based licence condition corresponds to a range of dilution valuesand not a single value. In particular, reduction of the dischargesalinity meets the salinity trigger value at lower dilutions. Thisfeature also illustrates that blending of desalination brine withlower salinity water (such as stormwater or sewage) may beemployed to overcome licence conditions that are based on asalinity trigger value [18]. Second, the entire possible range of

onmental impact assessment process in South Australia? Behindy (2012), http://dx.doi.org/10.1016/j.marpol.2012.08.005

J. Kampf, B. Clarke / Marine Policy ] (]]]]) ]]]–]]] 5

permitted dilution values is markedly lower than the 50:1dilution. The latter would rather correspond to salinity levelsabove ambient of 0.24–0.72 PPT (depends on recovery efficiency)that are much smaller than the licence’s salinity-anomaly thresh-old of 1.3 PPT.

3.2. Analysis of first monitoring results

In South Australia, the proponent undertakes much of the datagathering and reporting as part of the regulatory monitoring.Currently AdelaideAqua undertakes salinity monitoring requiredfor commission of the 50 GL plant. According to SA Water, whooversee the monitoring process, discharge rates will vary until theplant has been fully commissioned. As part of the monitoringprocess an array of marine mooring stations were installed atdistances of 100 m and 200 m from the desalination outlet.According to AdelaideAqua, instruments provide real-time infor-mation on salinity levels measured at a distance of 50 cm abovethe seabed with an update every 10 min. Graphical displaysof salinity data, presented as 24-hour rolling average, togetherwith information on mooring locations are available from theirWebsite.

Fig. 3 shows an example of salinity data data for the month ofFebruary 2012. Scientific interpretation of data is challenginggiven the lack of other relevant information such as dischargecharacteristics and tidal ranges. Nevertheless, it is still possible toidentify four individual short-lived events of locally elevatedsalinities as clear signatures of discharge activity. The datasuggest that each of the four discharge tests was limited to aday in duration. If this is so, then higher salinity levels are likely toappear for prolonged (or continuous) discharges. During thedischarge test on 16 February, salinity levels increased markedlyto a maximum of 0.7 PPT above ambient at the edge of the mixingzone. Apparently, the observed salinity levels meet the licenceconditions, but the important question remains whether this alsoholds for a continuous discharge at full 100 GL per annumoperation.

In the absence of information on discharge characteristics, wecan still estimate dilution ratios for the measured salinity levelsshown in Fig. 3. To this end, we use a background salinity of36 PPT and estimate discharge salinities from recovery efficien-cies in a range of 25–50%. After Eq. (A7), we find that the observedsalinity level of 0.7 PPT above ambient would correspond to adilution ratio of 16:1 for a recovery efficiency of 25%, or a dilution

Fig. 3. Example of salinity data reported by AdelaideAqua for Feb

Please cite this article as: Kampf J, Clarke B. How robust is the envirthe scenes of the Adelaide seawater desalination project. Mar. Polic

ratio of 50:1 for a recovery efficiency of 50%. These results are afirst indication that the discharge design may technically notachieve a target dilution exceeding 50:1 under all current scenar-ios for the full range of operating conditions/flows, despite earlierassurances given by the proponent.

3.3. Outcomes of a recent parliamentary enquiry

A parliamentary enquiry has been made into the operatinglicence of the Adelaide Desalination Plant [19] based on a draftversion of this work. In response to the enquiry, the Minister forIndustrial Relations, Minister for State/Local Government Rela-tions (Hon. R.P. Wortley) replied that the Minister for Sustain-ability, Environment and Conservation (Hon. Paul Caica) has beenadvised that [see [19]:

‘‘Development Approval required an outfall design to meet adilution ratio of 58:1, which was achieved and demonstrated byAdelaideAqua. The Environment Protection Authority (EPA) oper-ating licence criteria of 1.3 parts per thousand (PPT) was set aftereco-toxicology studies, required by the EPA to be undertaken byAdelaideAqua, determined that the safe level for relevant localspecies was approximately 2.6–2.7 PPT above background levels.The EPA operating licence conditions are best practice and themost stringent in Australia. The limits are very conservative andset well below the level at which environmental impacts wouldbe expected.’’

It should be noted that the cited ‘‘ecotoxicology report’’ byAdelaideAqua is not publicly available. The claimed ‘‘safe salinitylevels’’ of 2.6–2.7 PPT above background levels are thereforescientifically doubtful and questionable. After Eq. (A7), thesesalinity anomalies relate to extremely low dilution ratios ofo4:1 for a recovery efficiency of 25%, or 13:1 for a recoveryefficiency of 50%. Such dilution ratios, being the lowest amongall Australian desalination projects, are in stark contrast with EISrecommendations. This unverified Government response illustratesthat the post-decision EIA process and Government approvalprocedures are far from robust in South Australia.

4. Summary and conclusions

This investigation reveals that conditions of the operatinglicence of the Adelaide Desalination Plant are much softer thanassurances made earlier by the project’s proponent (in the EIS and

ruary 2012. Available at: /http://www.adelaideaqua.com/S.

onmental impact assessment process in South Australia? Behindy (2012), http://dx.doi.org/10.1016/j.marpol.2012.08.005

J. Kampf, B. Clarke / Marine Policy ] (]]]]) ]]]–]]]6

an additional report) and approval conditions imposed by theState Government. With reference to expert advice of scientistsinvolved in the EIS, this implies a higher risk of damage to themarine environment of Gulf St. Vincent. There are a number ofimportant improvements that could be made to the EIA proce-dures in South Australia. These are:

1.

Pth

The operating licence application should be made an integralpart of the EIS (currently the licence application and the EISare separate processes);

2.

Fig. A1. Illustration of negatively buoyant desalination brine discharges and the

Monitoring should be led and undertaken by truly indepen-dent scientific experts (currently operators undertake most ofthe monitoring);

associated salinity field.

3. All relevant monitoring data must be made publicly available(current monitoring information is incomplete and not pro-vided with the required scientific rigour; only monitoringreports submitted to the EPA will be publicly available); and

4.

Fig. A2. Illustration of the mixing fraction e..

The State Government could develop a process that ensuresthat others adhere to its approval conditions.

The implementation of these changes is critical for a betterprotection of important natural resources and habitat in SouthAustralia. While this work exclusively focussed on the AdelaideDesalination Project, it would be interesting to verify the pre (EIS)and post (operating licence conditions) construction monitoringrequirements of other major developments to further test therobustness of the EIA process in Australia and other countries.

While there are lessons here to be learned for other national orinternational developments, the use of standardized methodolo-gies should be approached with some caution given the distinc-tive circumstances (e.g. legislative, environmental, social, andeconomic) of any given project. According to Arts and Saunders([20], p. 287) ‘there is no single recipe for success’ and that ‘EIAfollow-up will have to be tailored to specific project circum-stances’. It is well recognised that post-decision monitoring in EIAis an area in need of increased effort. The lessons from this SouthAustralian case study reinforce the message that monitoringshould be closely linked to the EIA process, not separate to it. Italso suggests the need for a best-practice approach [20] thatincludes appropriate legislative requirements, commitment fromregulators, and independent review of follow-up programs andresults.

Acknowledgements

This research did not receive financial support from externalsources. We are grateful to Rob Keane who produced the graphicsfor Fig. 1.

Appendix A. Link between salinity anomalies anddilution values

A desalination outlet discharges brine concentrate at a salinityof Sbrine¼SoþDSn, where DSn is the excess salinity with referenceto the ambient salinity So. At some distance from the dischargepoint, the mixing between brine concentrate and ambient sea-water results in a salinity of SoþDS, where DS is the salinity levelabove ambient. Fig. A1 displays a sketch of this situation.

The product is a mixture of the brine concentrate with acertain volume of ambient seawater. The salinity of the mixturefollows from the mixing of the two source waters, which can beexpressed by the linear mixing equation:

SoþDS¼ ð1�eÞSoþeSbrine ¼ ð1�eÞSoþeðSoþDS*Þ ðA1Þ

lease cite this article as: Kampf J, Clarke B. How robust is the envire scenes of the Adelaide seawater desalination project. Mar. Polic

where e is the mixing fraction ranging between zero and one(Fig. A2).

The mixing fraction should not be confused with the dilutionfactor. Rearrangement of (A1) gives:

DS¼ eDSn) e¼ DS

DS*ðA2Þ

For example, a salinity excess of the discharge of DSn¼10 PPTand an observed salinity anomaly of DS¼1 PPT corresponds to amixing fraction of e¼0.1; i.e. desalination brine contributes 10%to the mixing product. Ambient water provides the other 90%.

Dilution, on the other hand, is the ratio of the volumes of brineand ambient water that correspond to a certain mixing fraction.This volume ratio is related to the mixing fraction by:

Dilution¼ð1�eÞe � n : 1 ðA3Þ

where the number n indicates the number of volume units ofambient seawater mixed with one unit of brine concentrate.Using (A2), relation (A3) can be written as:

n¼DS*

DS�1 ðA4Þ

Hence, determination of the dilution factor (A4) requiresmeasurements of an undisturbed ambient salinity together withsalinities directly at the outlet and at the location of interest (suchas the edge of the mixing zone). For example, a salinity excess ofDS*¼10 PPT together with a salinity anomaly of DS¼1 PPTcorresponds to a dilution factor of 9:1.

For a known salinity of the discharge Sbrine, relation (A4) can bewritten as:

n¼Sbrine�So

DS�1 ðA5Þ

Further assumptions are required if the discharge salinity isunknown. In this case, one can take advantage of the fact that thedischarge salinity is a multiple of the ambient (intake) salinity So.The recovery efficiency E is defined as the volume percentage of

onmental impact assessment process in South Australia? Behindy (2012), http://dx.doi.org/10.1016/j.marpol.2012.08.005

J. Kampf, B. Clarke / Marine Policy ] (]]]]) ]]]–]]] 7

intake water being converted into freshwater. Hence, the volumeof desalination brine returned to the sea per time unit is (1�E/100%) times the intake volume. The salt mass contained in theintake becomes concentrated in a smaller volume, and salinity ofthe return water increases. If we ignore small density changes(o1%) that occur during this process, the salinity of the returnwater can be estimated at:

Sbrine ¼So

1�E=100%ðA6Þ

The discharge salinity is twice that of the receiving water forE¼50%, whereas the salinity excess is reduced to 1/3 So forE¼25%. Then the dilution factor can be estimated from:

n¼So

DS

1

1�E=100%�1

� ��1 ðA7Þ

whereby a is a number between 0.5 and 1. BHP Billiton proposes asalinity threshold of 10% above ambient for the proposed sea-water desalination plant in Upper Spencer Gulf at Point Lowly[21]. For this instance, i.e. DS¼0.1 So, relation (A7) can beformulated as:

n¼ 101

1�E=100%�1

� ��1 ðA8Þ

For E¼50%, relation (A8) gives a dilution threshold of 9:1. ForE¼25%, the dilution threshold drops to around 2:1. The aboveformulae can also be used to derive salinity-anomaly thresholdsfor a given target dilution value.

References

[1] Fritzmann C, Lowenberg J, Wintgens T, Melin T. State-of the-art of reverseosmosis desalination. Desalination 2007;216:1–76.

[2] Lattemann S, Hopner T. Environmental impact and impact assessment ofseawater desalination. Desalination 2008;220:1–15.

[3] Harvey N, Clarke B. Environmental impact assessment in practice. Melbourne:Oxford University Press; 2012.

[4] Hodges BR, Furnans JE, Kulis PS. Thin-layer gravity current with implicationsfor desalination brine disposal. Journal of Hydraulic Engineering 2001;137:356–372.

[5] Okely P, Antenucci JP, Imberger J, Marti CL. Field investigations into theimpact of the Perth Seawater desalination plant discharge on CockburnSound. Final Report. Centre for Water Research. University of WesternAustralia; 2007; 102 pp.

Please cite this article as: Kampf J, Clarke B. How robust is the envirthe scenes of the Adelaide seawater desalination project. Mar. Polic

[6] Nunes RA, Lennon GW. Physical property distributions and seasonal trends inSpencer Gulf, South Australia: an inverse estuary. Mar Freshwater Res1986;37:39–53.

[7] Kampf J, Brokensha C, Bolton T. Hindcasts of the fate of desalination brine inlarge inverse estuaries: Spencer Gulf and Gulf St. Vincent, South Australia.Desalination and Water Treatment 2009;2:325–333.

[8] Kampf J, Payne N, Malthouse P. Marine connectivity in a large inverseestuary. J Coastal Res 2010;26(6):1047–1056.

[9] Edyvane KS. Conserving marine biodiversity in South Australia—Part1—Status review and framework for marine biodiversity conservation inSouth Australia. Technical Report, South Australian Research and Develop-ment Institute; 1999; 173 pp.

[10] Edyvane KS. Conserving marine biodiversity in South Australia—Part2—Identification of areas of high conservation value in South Australia.Technical Report, South Australian Research and Development Institute;1999; 328 pp.

[11] Fox DR, Batley GE, Blackburn D, Bone Y, Bryars S, Cheshire A., et al. Adelaidecoastal waters study, Final Report—Volume 1, June 2007, Summary of StudyFindings, CSIRO; 2007; /http://www.clw.csiro.au/acws/documents/ACWSFinalReportVol1.pdf S viewed 13 April 2012.

[12] Government of South Austalia. DEVELOPMENT ACT 1993: SECTION 48,Decision by the Governor. The South Australian Government Gazette, 2009;12:714–9. /http://www.governmentgazette.sa.gov.au/2009/february/2009_012.pdfS viewed 11 April 2012.

[13] Clarke B, Harvey N. The changing role of environmental impact assessmentunder the South Australian Development Act 1994 to 2005. Adelaide, TheRoyal Geographical Society of Australasia (SA Branch) Inc, South AustralianGeographical Papers, No 7; 2006.

[14] Government of South Australia & SA Water. Environmental Impact Statementfor Proposed Adelaide Desalination Plant 2008; /http://www.planning.sa.gov.au/index.cfm?objectid=5087096B-F203-0D46-AC6D2892342F3D00Sviewed 13 April 2012.

[15] Government of South Australia. Assessment Report for the EnvironmentalImpact Statement—Port Stanvac (Adelaide) Desalination Project Develop-ment Proposal, 2009;/http://dataserver.planning.sa.gov.au/publications/1346p.pdfS viewed 13 April 2012.

[16] AdelaideAqua. Report—Intake and Outfall Systems Environmental Perfor-mance Summary. Document no: H332401-1000-05-124-0090 2009; /http://www.epa.sa.gov.au/xstd_files/Water/Public%20consultation/desal_report.pdfS viewed 13 April 2012.

[17] Environmental Protection Authority (EPA). Licence EPA 26902 2010; /www.epa.sa.gov.au/xstd_files/Water/Other/desal_licence.pdfS viewed 13 April2012.

[18] Kampf J. The impact of blending on dilution of desalination brine in a tidalshallow sea. Mar Pollut Bull 2009: 58(7):1032-8, /http://dx.doi.org/10.1016/j.marpolbul.2009.02.009S.

[19] Parnell M. ‘QUESTION WITHOUT NOTICE: Port Stanvac Desalination Plant’Mark Parnell MLC, /http://markparnell.org.au/speech.php?speech=1062Sviewed 18 August 2012.

[20] Arts J, Morrison-Saunders A. Lessons from EIA follow-up. In: Morrison-Saunders A, Arts J, editors. Assessment Impact: Handbook of EIA and SEAFollow-up. James & James, London: Earthscan; 2004. p. 286–314.

[21] BHP Billiton. Supplementary Environmental Impact Statement for OlympicDam Expansion 2011; /http://www.bhpbilliton.com/bb/odxEis.jspS viewed13 April 2012.

onmental impact assessment process in South Australia? Behindy (2012), http://dx.doi.org/10.1016/j.marpol.2012.08.005