Water Journal November 2010

Volume 37 No 7

NOVEMBER 2010

AWA JOURNAL OF THE AUSTRALIAN WATER ASSOCIATION

-

"

0

::, ::, ::,

~

~ C L

-

'\',

1

_

I

-

'

1

'•'

~°tJAj,11,i'llfi•

·--..ii..,;;···

'

WV . .. . . - - ~ •

' .U

~

:

I;:"lll:J'l,1,1-n , •

~~"9!t;;i..---

--------------.u.~~

,.·, •

--e'·.-...,',J.,;

!)""'

~

___.nl.\,i

•ldUII,'

'

w at er

Journal of the Aust,allan Water AssoclaUo, ISSN 0310-0367

Olfactory Characterisation of NMVOC Emissions from WWTP Inlet Works see page 82

TECHNICAL FEATURES( ~

Volume 37 No 7 November 2010

contents

Sustainable Odour Control at Perth's Wastewater Treatment Plants - see page 87

INDICATES THE PAPER HAS BEEN REFEREED)

MEMBRANE PROCESSES

[i]

Does Polymer Addition Negatively Affect Microfiltration and Reverse Osmosis?

L Schimmoller, J Lozier, T O'Neill

50

B Soito, M Hoang, T Tran

58

C Owens, C Snape

61

J Barclay

65

J Cesca, R Forbes, C Martinson, H Bustamante, T Nguyen, WWang

72

Pilot testing demonstrates polymer might even improve performance

[i]

Dual Nanofiltration/Reverse Osmosis Systems

Desalination with low energy and lower fouling, but higher capex CUSTOMER SERVICE

[I]

SEQ Water Grid Demonstrates Confidence In Water Quality

An interactive web-based monthly report, which aims to educate consumers ODOUR MANAGEMENT

Dispersion Modelling: Tracing and Anticipating that Odour Modelling can conduct experiments that would be cost prohibitive in the real world

liJ

The WERF Biosolids Odour-Reduction Roadmap

First Full-scale Application at North Head, Sydney

[ii

Sewer Odour Abatement Practices - An Australian Survey E Sivret, R Stuetz

77

X Wang, G Parcsi, E Sivret, R Stuetz, J Cesca

82

C Nichols, S McNeil, GP Van Durme, K Cadee, I Wallis, A Adams

87

Identifying key areas of improvement for the design and operation of processes

[i]

Olfactory Characterisation of NMVOC Emissions from WWTP Inlet Works

Combining chemical and sensory assessment to gain understanding of the significant odorants

Ii]

Sustainable Odour Control at Perth's Wastewater Treatment Plants

The treatment systems represent international best practice INTERNATIONAL PROJECTS

[ii

Sustainability Assessment Tool for Water Treatment Technologies in Rural Communities

Developed during a project in Tenganan, Indonesia

N Cocks, R Kurup

97

WATER BUSINESS

New Products and Business Information. Special Features: Odour Control; Pumps

103

Advertisers' Index

112

2 NOVEMBER 2010 water

w at er

Journal oflhe A"51raUan Water Association ISSN 0310-0367

Volume 37 No 7 November 2010

contents REGULAR FEATURES From the AWA Chief Executive The Price is Right?

T Mollenkopf 4 T Fricke

My Point of View

5

Crosscurrent

8

Industry News

16

AWA News

32

Events Calendar

36

Conference Reviews

38

New Bluetongue Brewery Targets World's Best Practice in Water Usage - see page 16

FEATURE REPORT Challenges in Adapting Australian Water Resources and Infrastructure to Climate Change MDShort, ERocheta, GM Peters, MSchulz, WL Peirson, RM Stuetz, RJ Cox

42

Pacific Water Conference - see page 38

AWA CONTACT DETAILS Australian Water Association ABN 78 096 035 773 Level 6, 655 Pacific Hwy, PO Box 222, St Leonards NSW 1590 Tel: +61 2 9436 0055 Fax: +61 2 9436 0155 Email: info@awa.asn.au Web: www.awa.asn.au DISCLAIMER Australian Water Association assumes no responsibility for opinion or statements of facts expressed by contributors or advertisers.

COPYRIGHT AWA Water Journal is subject to copyright and may not be reproduced in any format without written permission of the AWA. To seek permission to reproduce Water Journal materials, send your request to media@awa.asn.au WATER JOURNAL MISSION STATEMENT 'To provide a journal that interests and informs on water matters, Australian and international, covering technological, environmental, economic and social aspects, and to provide a repository of useful refereed papers. ' PUBLISH DATES Water Journal is published eight times per year: February, April , May, June, August, September, November and December.

EDITORIAL BOARD Chair: Frank R Bishop; Dr Bruce Anderson, AECOM; Dr Terry Anderson, Consultant SEWL; Michael Chapman, GHD; Robert Ford, Central Highlands Water (rtd}; Anthony Gibson, Ecowise; Dr Brian Labza, Vic Health; Dr Robbert van Oorschot, GHD; John Poon, CH2M Hill; David Power, BEGA Consultants; Professor Felicity Roddick, RMIT University; Dr Ashok Sharma, CSIRO; and E A (Bob} Swinton, Technical Editor.

AWA

EDITORIAL SUBMISSIONS Water Journal welcomes editorial submissions for technical and topical articles, news, opinion pieces, business

information and letters to the editor. Acceptance of editorial submissions is at the discretion of the editor and editorial board.

• Technical Papers and Features Bob Swinton, Technical Editor, Water Journal- bswinton@bigpond.net.au AND journal@awa.asn.au Papers 3,000-4,000 words and graphics; or topical articles of up to 2,000 words relating to all areas of the water cycle and water business.

Submissions are tabled at monthly editorial board meetings and where appropriate are assigned referees. Referee comments will be forwarded to the principal author for further action. Authors should be mindful that Water Journal is published in a 3 column 'magazine' format rather than the full-page format of Word documents. Graphics should be set up so that they will still be clearly legible when reduced to two-column size (about 12cm wide}. Tables and figures need to be numbered with the appropriate reference in the text e.g. see Figure 1, not just placed in the text with a (see below} reference as they may end up anywhere on the page when typeset.

• Industry News, Opinion pieces and Media Releases Helen Kelton, Editor, Water Journal- journal@awa.asn.au • Water Business and Product News Brian Raul!, National Sales and Advertising Manager, Hallmark Editions - brian.rault@halledit.com.au

ADVERTISING Advertisements are included as an information service to readers and are reviewed before publication to ensure relevance to the water sector and objectives of the AWA. Brian Rault, National Sales and Advertising Manager, Hallmark Editions - brian.rault@halledit.com.au Tel: +61 3 8534 5014 AWA BOOKSHOP Copies of Water Journal, including back issues, are available from the AWA Bookshop for $12.50 plus postage and handling.

Email: bookshop@awa.asn.au PUBLISHER Hallmark Editions, PO Box 84, Hampton, Vic 31 88 Tel: 61 3 8534 5000 Fax: 61 3 9530 8911 Email: hallmark.editions@halledit.com.au

The microfiltration area at Bundamba Advanced Water Treatment Plant, part of the Queensland's Government's $2.4 billion Western Corridor Recycled Water Project. The Project was recently named the Global Grand Prize Winner for Superior Achievement at the International Water Association's Global Project Innovation Awards. See story on page 28.

water

NOVEMBER 2010 1

Innovation in action

Pipe and Civil prides itself on doing things differently and we understand that no two jobs are exact ly alike.

As such, our team of highly skilled engineers and technicians consider each project as a blank sheet, taking a fresh look at how we can improve on what was done before.

We work with our clients to come up with solutions aimed at getting the job done ahead of both time and budget. This approach has resulted in our people designing a number of unique pieces of equipment that give us the edge when it comes to safety, the environment, quality and production.

Located in Brisbane, we also have offices throughout Austra lia.

Brisbane - Head Office Tel: 07 3262 4600

Melbourne Office Tel: 03 9024 5420

Adelaide Office Tel: 08 8233 5941

Perth Office Tel: 08 9288 1805

www.pipeandcivil.com.au

feature article

Challenges in Adapting Australian Water Resources and Infrastructure to Climate Change MD Short*, E Rocheta, GM Peters, M Schulz, W L Peirson, RM Stuetz, R J Cox UNSW Water Research Centre, University of New South Wales, Sydney, NSW 2052, Australia (*E-mail: m.short@unsw.edu.au; tel: +61 2 93855947; fax: +61 2 93138624) There is widespread agreement amongst the international scientific community that climate change is occurring, that it is largely manmade and that it will have significant implications for humanity. Climate change poses major problems for many sectors, such that there is now growing awareness of the need to adapt in order to minimise the negative impacts and exploit possible future benefits under a changed climate. To assist in defining Australia's climate char:,ge adaptation research agenda for the water sector, a comprehensive review was undertaken as part of the Australian Climate Change Adaptation Research Network for Settlements and Infrastructure (ACCARNSI)- one of the eight thematic research networks under the National Climate Change Adaptation Research Facility (NCCARF) (www.nccarf.edu.au).

This review seeks to outline a range of critical knowledge gaps and future research priority areas for the water sector. It is suggested that future research effort in these key areas wou ld assist the water industry in formulating effective water infrastructure and water security adaptation responses. Given the extensive range of potential climate change impacts on water resources, the research potential in this field is known to be almost limitless (USEPA, 2008). Consequently, this review provides a high-level overview of suggested ' priority' research needs and by no means represents an exhaustive list of all current challenges and future research needs for water sector adaptation.

Data and information A limiting factor and major challenge for the water industryand others-is having sufficient access t o the right data. Data is crucial for informing infrastructure decision-making, where proper access to the right kind of data can often have implications for the long-term valid ity of such decisions.

1. Data collection Research progress is often hindered by a lack of data which subsequently affects good decision making. More data is needed to reduce uncertainty in climate model outputs and to detect and monitor the effects of climate change as they occur. Increased monitoring efforts in relation to ground-based weather stations are needed to reverse the current trend of shrinking observation networks (Kundzewicz et al., 2008) and to collect critical hydro-meteorological data necessary to accurately characterise the weather record . Similarly, improved observations of atmospheric conditions are needed to better define and understand mechanisms underlying atmospheric processes which are essential to assist climate modellers project future climatic conditions.

Observational climate data should be made broadly available to all stakeholders, since access to this data is a prerequisite for improving understanding of ongoing climate

42 NOVEMBER 2010 water

...

\ \

\

.....-········

/

!

,,//'

___ ,../ '

......····•·•··

.I

.,.-·····

'--

------

Figure 1. A schematic diagram describing the statistical downscaling approach (BoM 2010) http://www.bom.gov.au/info/climate/change/gallery/65.shtml change. There is a need for a central repository to facilitate access to water and climate data, surpassing the existing fragmented and inconsistent system. This need wi ll ideally be addressed via the ongoing implementation of the Australian Water Resources Information System (http://www.bom.gov.au/water/).

2. Climate model downscaling and c limate scenarios Current general circulation models (GCMs) are of limited value to water managers because of their low spatial resolution.

These models project circu lation processes on grid areas that are too coarse to be useful for catchment-level water resource planning or assessing local or even regional impacts. Spatial downscaling is used to translate climate projections from coarse resolution GCMs to finer spatial scale models (see Figure 1) that can then be used by catchment-level water managers to make predictions of local supply impacts.

Furthermore, hydrological impacts on temporal scales from seasonal to inter-annual and longer-term need to be distinguished and more research is req uired to achieve this.

Research is also needed to enable accurate estimates to be made of relative certainty or uncertainty regarding future climate change predictions based on local and global climate model outputs.

A viable alternative to quantitative GCM modelling is to base adaptation planning and research on plausible cl imate scenarios and associated qualitative or quantitative impact scenarios which assess the vulnerability of water resources

feature articles

Parsons Brinckerhoff (PB) is a global leader when it comes to design ing and delivering strategic infrastructure for water. Right now we are working on major water projects around Australia - and more are in the pipeline. That means we're looking for innovative professionals to take up roles that matter within our organisation including Design Managers, Project Managers and Design Engineers.

GLOBAL. INNOVATIVE. LEADING. FLEXIBLE. MAJOR PROJECTS. CAREER BUILDING. Are you ALL OF THE ABOVE? Visit www.pb.com.au

-

=-

r~

25

Yl!Alr!I

PARSONS BRINCKERHOFF

feature article and infrastructure (Pittock, 2008). The development of adequate simulation methodologies remains a principal research challenge in the presence of high level uncertainty, mainly due to the scope of the localised analyses needed to address the variety of impacts in different regions. In addition to the need for new regional climate change scenarios, new scenarios are needed for use by water planners and managers in relation to: future water use patterns, changes in water demand, population and demographic changes, changing use of water, and water resource and infrastructure impact scenarios to be used as a basis for vulnerability assessments.

The relevance of existing climate and planning scenarios based on "stationarity"-the idea that natural systems fluctuate within an unchanging envelope of variability-has been recently brought into question by the apparent demise of climate stationarity (Milly et al., 2008). Water managers, therefore, must develop novel approaches to overcome this lack of climate certainty, as it is no longer appropriate to assume past hydrological conditions wi ll continue in perpetuity.

3. Water quantity and quality In general, a broad range of detailed projections of future water quality and quantity are needed in order to facilitate better water planning and management. For example, increased variability of water quality entering storages will require increased upstream catchment monitoring efforts as well as state-of -the-art monitoring systems for water treatment plant influents. Research is also needed to understand the impacts of climate change in relation to: • Variable catchment water quality implications for downstream water treatment processes; • Long-term , spatially-integrated research at regional, transregional, or continental scales to address the impacts of extreme climate variability on ecosystems and water supplies; • Uncertainties associated with diminished water quantity/quality; • Changing temperatures and runoff effects on aquatic ecosystems; • Sediment transport and contaminant deposition rates in catchments; • Chemical and biological process effects including contaminant ecotoxicity; • Increasing prevalence of Cyanobacteria in water supplies (e.g. Wiedner et al., 2007; Paerl and Huisman, 2009); • The downstream effects of both toxic and non- toxic algal blooms in terms of resulting toxi ns and taste and odour compounds; and • The intensity and frequency of water-borne disease in Australia (Pittock, 2003). For example, climat e change effects on the geographical distribution and regional transmission of some pathogens such as Ross River Virus (e.g. Tong et al., 2008; Woodruff and Bambrick, 2008).

4. Water and energy linkages Water and energy are intrinsically linked: energy generation uses water in a variety of ways and similarly, energy is used in the provision of water through extraction, treatment and distribution. In Australia, the Water Services Association of Australia (WSAA) has recognised that an improved understanding of the 'water- energy nexus' is critical in order to

44 NOVEMBER 2010 water

Figure 2. Dry Soil.

target investment to achieve maximum greenhouse gas (GHG) red uctions (WSAA, 2009a). Internationally, the International Water Association (IWA) has also recognised that the energy and water sectors must work toward joint solutions rather than continuing to operate individually (IWA, 2009). This joint 'holistic' approach would encourage creative solutions to cooptimise the efficiency of power and water provision in a more synergistic manner. The possibility and validity of expanding the water- energy nexus to 'water-energy-nutrients' should also be investigated. Research into these interconnections, as well as integrated models that can assist policy makers in these areas, are needed.

5. Water infrastructure performance

i. Urban water infrastructure Adequate preparedness of urban water infrastructure for climate change is critical and more research is needed to provide policy and planning guidance for adapting urban growth policy to global environmental change (Bobylev, 2009).

There remains considerable uncertainty and gaps in our current understanding of the effects of climate change and urbanisation on urban water infrastructure, particularly underground infrastructure. The emerging chal lenge is to understand how climate change will impact on assets and drive future maintenance and replacement decisions.

Research is currently being conducted in order to better understand how assets fail, which degradation processes will be accelerated as well as methods for maximising infrastructure lifespan (WSAA, 2008).

Determinations of critical infrastructure that may be susceptible to compou nding or 'cascading' impacts during extreme climate events are also needed (e.g. implications of multiple compounding hazardous events such as storm surge , wave action and extreme rainfall on coastal infrastructure) and plans must then be formul ated regarding how these impacts may best be avoided. At the same time, there is a parallel need to research alternative infrastructure options to minimise exposure to cascad ing impact scenarios. ii. Flood management

There is currently very high-level confidence that cl imate change will adversely affect the capacity of c urrent water management practices to reduce flood risk (Parry et al., 2007).

There is a need for improved water infrastructure performance in relation to effective flood management protocols. Improved coordi nation and cooperation between the relevant stakeholders (i.e. local, state and federal agencies) will be

feature articles

A New Force in Water is Emerging

feature article central to the success of such future actions. In the area of flood management, there is also a need for:

needed in order to ensure relevance with future climate change.

• Modifications to existing infrastructure design standards and re-evaluation of probable maximum precipitation and design floods for major infrastructure to account for expected future extreme events;

Leakage reduction and pressure management efforts in urban water distribution networks also offers the potential for large water savings, and although many Australian utilities are already engaged in aggressive initiatives in these areas, there remains scope for additional effort and research into improved approaches. These are especially relevant in relation to irrigation practices in the agricultural sector where, in some instances, large volumes of water are lost through leaky distribution and irrigation networks and open aqueducts during on-farm distribution. Research into changes in irrigation technology, drought resistant crops and cropping systems as well as establishing soil-moisture conservation practices are among the more frequently discussed research agendas in th is area. Improved metering also has a vital role to play in adaptive water resource management and efficient end use by assisting with the uptake of water efficiency practices, ensuring that water use will be consistent with water sharing plans and new licence conditions, aiding leak detection, and facilitating a water trading market.

• Developing up-to-date methodologies, such as continuous simulation methods and determination of antecedent moisture conditions, used for predicting design flood intensities; • Developing methods for estimating flood risk under the new paradigm of climate 'nonstationarity' (Milly et al., 2008; Wagener et al., 2008); and • Determining how to accommodate climate change uncertainties into the design and management of water infrastructure. New design and management protocols need to be developed that account for future extremes in weather events.

To this end, Engineers Australia is currently revising the guideline document Australian Rainfall and Runoff used for assessment of rainfall, runoff, water resources and flooding so as to incorporate such issues related to climate change.

8. Water pricing and markets

Policy and Regulatory Reform 6. Policy and regulatory reform The Australian water industry has undergone over a decade of reform since the introduction of the Water Reform Framework in 1994, resulting in changes to the structure, ownership and regulatory arrangements of the industry.

More recently, the 2004 intergovernmental National Water Initiative was signed in an attempt to refresh the reform program, with the overall aim to develop a nationallycompatible, market-, regulatory- and planning-based system for managing surface and groundwater resources (AGO, 2004). Furthermore, the current round of revisions to the Australian Drinking Water Guidelines is nearing completion, with a number of draft items already open for public consultation.

According to WSAA, new approaches to environmental regulation of the urban water sector are required (WSAA, 2009b). At present, most conventional environmental regulations are 'single issue' based, such as point-source regulations on effluent water quality standards. Additionally, the adoption of sustainability principles into policy and regulation has not yet been achieved and requires a more holistic and integrated regulatory approach. It is suggested that new regulatory approaches should recognise the need to balance competing environmental, social and economic outcomes in partnership with the water industry's objectives, possibly through the adoption of 'life cycle' approaches.

Furthermore, all future legislation needs to provide water managers with the flexibility and capacity to adjust and adapt to the impacts of climate change (Flett et al., 2008).

7. Water end use efficiency and demand management Increasing water end use efficiency and reduced user demand for water has already demonstrated and continues to offer excellent potential for combined water and GHG emissions reductions, with such measures often considered as 'low-hanging fruit' abatement alternatives. Although much has been done in this area already, more research into the technological, public awareness and implementation sides of water use efficiency and demand management measures are

46 NOVEMBER 2010 water

There is an ongoing need to review and refine water rate structures so that they reflect the 'true cost' of water and water service provision as changes in future patterns of water supply and consumption emerge. Other measures such as the expansion of water trading, economic incentives (e.g. new metering and pricing approaches) to encourage water conservation, water allocation reforms and reallocation of water to higher value uses also need further exploration.

In the context of adaptation via water pricing and policy/legislation reform, there is also a parallel need to better understand the environmental justice and social equity consequences of such adaptation measures (Mitchell et al., 2007; Wilbanks et al. , 2007).

9. Adaptive management and adaptation under uncertainty Traditional strategic planning approaches for addressing climate change risks do not have adequate flexibility, promptness and responsiveness to enable adaptive changes as new information and emerging realities arise. To enable government institutions to become more responsive to change, there is an identified need for strategic decision support systems or decision-making frameworks that allows new scientific findings to be converted into policies in a short period of time (Mirfenderesk and Corkill, 2009).

Furthermore, it is important that the strategic plans or decision making frameworks themselves are equally adaptive and non-static in order to complement adaptive management practices.

Given the importance of water supply security, it is essential that new water management strategies are developed to be sufficiently diverse, flexible and adaptable such that current high-level uncertainties associated with future climate change projections can be accounted for.

Despite recent improvements in the characterisation of climate change 'uncertainty', more research to further characterise and reduce such uncertainties will ultimately help water managers in their efforts to adapt to uncertain future hydrological changes (Kundzewicz et al., 2008).

feature articles

10. Sustainable adaptation The fields of climate change adaptation and mitigation require more close integration. Research in the area of adaptation-mitigation interconnections should reveal future adaptation pathways that can best reduce the likelihood of maladaptation. For example, desalination as an adaptation strategy has a high energy consumption and GHG emissions intensity such that it is at odds with mitigation objectives. There is a need for an improved understanding of the complex inter-relationships between adaptation and mitigation strategies in order to minimise potential negative impacts of adaptation actions whilst at the same time exploiting any potential positive climate change impacts.

Leading the world in the modernisation of Irrigation Infrastructure

The application of sustainability assessment tools such as life cycle assessment will be important in helping water utilities identify suitable process improvements and adaptation responses. Additionally, and given the recent interest in decentralised treatment systems, there is a need for detailed assessments of the economic, environmental and social sustainability of these types of systems relative to the conventional centralised water supply paradigm (WSAA, 2009b).

Rubicon is leading the world in t he modernisation of irrigation systems. Total Channel Control~ technology has set a new International benchmark of 90% distribution efficiency.

The FlumeGate'" is the cornerstone of TCC~ and provides accurate flow measurement and control.

Rubicon has worked with the University of Melbourne for over 10 years pioneering innovative approaches to water delivery technology.

Human adaptation 11. Collaboration, communication and education There is a general need for more cross-disciplinary interaction between researchers as well as improved communication between scientists and water managers/policy makers. For example, water managers need to have climate model outputs in useful and interpretable formats amenable to incorporation into resource management models and available at scales (regional or catchment) useful for resource management activities. The feasibility of establishing an information broker to act as the intermediary between climate scientists and water practitioners, and of involving practitioners in the development of scientific research agendas, should also be investigated. Greater knowledge transfer and capacity-building within the water industry will also empower water managers and other stakeholders with the necessary skills to enable: better management of water resource allocations; planning for and monitoring of floods and droughts; development of more accurate long-term supply-demand forecasts; and more appropriate evaluation and integration of alternative future climate scenarios into future water management activities.

Rubicon is Motorola's Distribution Partner, offering MOSCAD and ACE RTU solutions for management and control of Water and Wastewater for urban and rural Utilities.

• • • • •

SCADAConnect'" FlumeGate'" Sli pGate'" SolarDrive'" Total Channel Control~ (TCC)* • FarmConnect'" • Australian Premier Distributor for Motorola MOSCAD and ACE.

A parallel need exists for climate change adaptation education and awareness building in the wider community.

These should aim to communicate information about climate change impacts and risk reduction strategies so as to improve general understanding of who and what systems are at greatest risk and the actions necessary t o reduce these risks, with the aim of changing community behaviour (e.g. Sampei and Aoyagi-Usui , 2009).

Importantly, t his information needs to be presented in a meaningful and relevant form so that community groups can easily comprehend its context and meaning.

~

~

Conclusion

RUBICON

Many sectors and systems (e.g. water supply and sanitation, agriculture, energy, human health, settlements and infrastructure, industry, transportation, tourism, insurance and financial services) are dependent on secure water resources. Any adverse future changes to

www.rubicon.com.au

p

Lf J

Rubicon Systems Australia Pty Ltd 1 Cato Street Hawthorn East VIC 3123 Australia Telephone +6139832 3000

1A Wheeler Street Shepparton VIC 3630 Australia Telephone +61 3 5820 8800 www.rubicon.com.au

feature article hydrological regimes and water quality as a result of cl imate change will, therefore, have significant socioeconomic implications (Kundzewicz et al., 2008). More work is needed to enable us to better manage the risks and uncertainties relating to climate change impacts for water resources and water infrastruct ure and to further the development and implement of appropriate adaptation strategies. We believe that the concepts discussed here will provide a basis for t he development of further ideas and activities in this critically important area.

References AGO (Australian Government) (2004). Intergovernmental Agreement on a National Water Initiative between the Commonwealth of Australia and the Governments of New South Wales, Victoria, Queensland, South Australia, the Australian Capital Territory and the Northern Territory, June 2004.

Available at: htt p://www.nwc.gov.au/resources/documents/lntergovern mental-Agreement-on-a-national-water-initiative.pdf (accessed 11/05/09).

Bobylev, N. (2009). Urban Underground Infrastructure and Climate Change: Opportunities and Threats. Cities and Climate Change: Responding to an Urgent Agenda, Fifth Urban Research Symposium. Marseille, France, J une 28-30, 2009.

BoM (bureau of Meteorology) (2010). Images from "The Greenhouse Effect and Climate Change". Australian Government . Available at: http://www. bom.gov.au/info/climate/change/gallery/65.shtml (accessed 11/08/10).

Flett, D., I. Varley, N. Schofield and T. Ladson (2008). New water entitlement products for Australia-A Thinkpiece. Waterlines Report Series No. 12, November 2008. Published by the National Water Commission, Canberra ACT. ISBN 978-1-921107-66-5.

Hennessy, K., B. Fitzharris, B.C. Bates, N. Harvey, S.M. Howden, L. Hughes, J. Salinger and R. Warrick (2007). Australia and New Zealand. In: Clima te Change 2007: Impacts, Adaptation and Vulnerability. Contribution of

Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. M.L. Parry, O.F. Canziani, J.P. Palut ikor, P.J . van der Linden and C.E. Hanson (Eds.). Cambridge University Press, U.K. pp: 507-540.

IWA (International Water Association) (2009). Submission by IWA International Water Association relevant for the preparation of the 6th session of the AWG LCA. International Water Association, London, 24 April 2009. Available at: http://unfccc.inVresource/docs/2009/smsn/ ngo/146.pdf (accessed 11/05/ 09).

Kundzewicz, Z.W. , L.J. Mata, N.W. Arnell, P. Doll, B. Jimenez, K. Miller, T. Oki, Z. fien and I. Shiklomanov (2008). The implications of projected climate change for freshwater resources and their management.

Hydrological Sciences Journal 53(1): 3-10.

Milly, P.C.D., J. Betancourt, M. Falkenmark, R.M. Hirsch, Z.W. Kundzewicz, D.P. Lettenmaier and R.J. Stouffer (2008). Stationarity is dead: whither water management? Science 319(5863): 573- 574.

Mirfenderesk, H. and D. Corkill (2009). The need for adaptive strategic planning: Sustainable management of risks associated with climate change. International Journal of Climate Change Strategies and Management 1(2): 146-159.

Mitchell, C. , S. Fane, J. Willetts, R. Plant and A. Kazaglis (2007). Costing for Sustainable Outcomes in Urban Water Systems - A Guidebook.

Research Report No 35, Published by The CRC for Water Quality and Treatment. ISBN: 1 8766 16 601.

Paerl, H.W. and J . Huisman (2009). Climate change: a catalyst for global expansion of harmful cyanobacterial blooms. Environmental Microbiology Reports 1(1): 27-37.

Parry, M.L., O.F. Canziani and J.P. Palutikof and Co-authors (2007).

Technical Summary. In: Climate Change 2007: Impacts, Adaptation and Vulnerability. M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and C.E. Hanson (Eds.). Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge UK. pp: 23-78.

Pittock, B. (Eds.). (2003). Climate Change: An Australian Guide to the Science and Potential Impacts, Published by the Australian Greenhouse Office. ISBN 1 920840 12 5.

Pittock, J. (2008). Climate Change and Water Supply. Water21 (magazine of the International Water Association) June issue: 12-14.

Specify and use Link-SealÂŽ seals to seal pipe penetration

Sampei, Y. and M. Aoyagi-Usui (2009). Mass-media coverage, its influence on public awareness of climate-change issues, and implications for Japan's national campaign to reduce greenhouse gas emissions. Global Environmental Change 19(2): 203-212.

Tong, S., P. Dale, N. Nicholls, J.S. Mackenzie, R. Wolff and A J. McMichael (2008). Climate Variability, Social and Environmental Factors, and Ross River Virus Transmission: Research Development and Future Research Needs. Environmental Health Perspectives 116(12): 1591-1597.

USEPA (2008). National Water Program Strategy: Response to Climate Change, Published by the U.S. Environmental Protection Agency, Office of Water (4101M), Report No. EPA 800-R-08-001.

Wagener, T., M. Sivapalan and B. McGlynn (2008). Catchment classification and services-toward a new paradigm for catchment hydrology driven by societal needs. In: Encyclopedia of Hydrological Sciences. M.G. Anderson (Eds.), John Wiley & Sons, Ltd. 12 p.

Wiedner, C., J. Rucker, A. Bruggemann and B. Nixdorf. (2007). Climate change affects timing and size of populations of an invasive cyanobacterium in temperate regions. Oecologia 152(3): 473-484.

Wilbanks, T. J., P. R. Lankao, M. Bao, F. Berkhout, S. Cairncross, J.-P. Ceron, M. Kapshe, R. Muir-Wood and R. Zapata-Marti (2007). Industry, settlement and society. In : Clima te Change 2007: Impacts, Adaptation and Vulnerability. M . L. Parry, 0. F. Canziani, J.P. Palutikof, P. J . van der Linden and C. E. Hanson (Eds.). Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge UK, pp: 357-390.

Slide the assembly into the space between the pipe and wall opening

PR •jEX

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

Available in Australia and New Zealand

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

48 NOVEMBER 2010 water

Woodruff, R. and H. Bambrick (2008). Climate change impacts on the burden of Ross River Virus disease, Report commissioned by the Garnaut Climate Change Review, June 2008. Available at: http://www.garnautreview.org.au/ (accessed 07 September 2009).

WSAA (the Water Services Association of Australia) (2008). WSAA Report Card 2007/2008: Performance of the Australian Urban Water Industry and projections for the future. Published by the Water Services Association of Australia Ltd. ISBN 1 920760 32 6.

WSAA (the Water Services Association of Australia) (2009a). WSAA Report Card 2008/ 2009: Performance of the Australian Urban Water Industry and projections for the future. Published by the Water Services Association of Australia Ltd. ISBN 1 920760 46 6.

WSAA (the Water Services Association of Australia) (2009b). Vision for a sustainable urban water future: Position Paper No. 3. Published by the Water Services Association of Australia Ltd. ISBN 1 920760 34 2.

feature articles

Water Quality Specialists providing analysers and sensors for the followi measurments:

pH Conductivity Dissolved Oxygen ORP Chlorine and more

To contact your local Rosemount Ana lytical sales representat ives, please ca ll us on 1300 -55-30 51

ROSEMOUNT. Analytical

EMERSON_ Process Management

.

membranes processes

~

refereed paper

DOES POLYMER ADDITION NEGATIVELY AFFECT MICROFILTRATION AND REVERSE OSMOSIS? L Schimmoller, J Lozier, T O'Neill Abstract CH2M HILL completed an assessment of the impact of polymer on microfiltration (MF) and reverse osmosis (RO) membrane performance using pilot testing. Inclined plate clarification of tertiary wastewater with ferric chloride and polymer addition was used upstream of the membranes for solids, organics, and phosphorus removal. An anionic polymer was selected to mitigate the potential adverse impact of the polymer on the downstream membrane processes. Comparison with nonpolymer operations determined that anionic polymer addition did not negatively impact MF and RO performance. In fact, MF performance may have improved as evidenced by lower Transmembrane Pressure (TMP) and higher permeability during polymer addition.

Determining the Value of Polymers in Membrane Performance Chemical c larification of tertiary wastewater can significantly improve the performance of microfiltration (MF) and reverse osmosis (RO) by reducing organic matter and orthophosphate levels that can contribute to membrane fouling

water Future Features DECEMBER - Trenchless technology, pressure sewerage, water policy FEBRUARY - Sewer processes, smart systems, metering APRIL - Membranes and desalination , reu se

50 NOVEMBER 201 0 water

MICROFILIRAJION SOLIDS PATHOGENS

REVERSE OSMOSIS

NUTRIENTS - N&P ORGANICS TOS PATHOGENS MICROCONSTITUENTS

VYIAPYANCEP O!lQAIIQN

NOMA PATHOGENS MICROCONSTITUENTS FINISHED WATER

Exhibit 1. LPAWTP Process Flow Diagram. and RO scaling. The addition of polymer to chemical c larification technologies (such as lamella plates, sand ballasted clarification, and solids contact clarifiers) is often necessary for good solids settling performance. However, fou ling may occur due to polymer attachment to the membrane surface or an adverse reaction with the RO antiscalant chemical. For this reason membrane suppliers are often reluctant to endorse polymer use.

A three month pilot study was conducted at the Brisbane, Queensland, Luggage Point Advanced Water Treatment Plant (LPAWTP) to assess the impact of polymer addition to the MF and RO membranes. The LPAWTP is a 70 MUd indirect potable reuse plant, which uses dual membranes (MF/ RO) as part of the treatment process, as shown in Exhibit 1. Chemical precipitation is provided upstream of MF to remove phosphorus, solids, and organics.

Phosphorus is specifically targeted to reduce the potential of phosphate scaling on the RO membranes. Polymer addition can assist chemical clarification during poor raw water quality events, which periodically occur at LPAWTP.

Pilot testing demonstrates polymer might even improve performance.

Establishing Polymer Selection and Pilot Testing Protocols Key to Evaluation Success In determining polymer selection for the pilot testing there was a perceived risk of adverse membrane fouling from the use of even a low polymer dose (0.2 mg/L) to improve clarifier performance. However, to further mitigate th is risk an anionic polymer was selected, offering the following benefits: • Both the LPAWTP MF membrane (Pall Microza) and RO membrane element (Toray TML 20-400) have a negative charge, which reduces the likelihood of polymer attachment to the MF or RO membrane surfaces • The RO antiscalant (Spect raGuard as manufactured by PWT) has a negative charge • Most literature from RO manufacturers indicates potential problems with cationic polymers, not anionic polymers Jar testing was conducted on the two anionic polymers that were already in use at LPAWTP- dry powder polymer (Flopam AN905 SH) used with the centrifuges and the emulsion polymer (Flopam E532) formerly used with the gravity thickener (Flopam AN905 SH is now used). Both polymers produced good settling floe, but the dry powder polymer was selected for clarification for the following reasons:

technical features

ABB Au stralia Pty Limited For enquiries phone: 1300 782 527 www.abbaustralia.com .au

Power and productivity for a better world™

jl 1111 ,.,1,1,

~

membranes processes • Unlike the emulsion polymer, the dry powder polymer does not contain oil that may cause problems with downstream processes

r efere e d pape r

MF Transmembrane Pressure (TMP) 150 140

• The existing powder polymer storage and feed system at the main plant is more suitable for expansion

130

,,

120

..

According to the manufacturer (SNF), the dry polymer has low charge density and a high molecular weight (11 t o 13 million Daltons).

... .. .. . . . .

110

No l)O¥ner additlon with high plale

100

loadw,g

.. ..

•

90

D..

~ 80

D..

70

I-

60

:::e

" ._ ,f··

50

Jar tests showed that for almost all operating conditions, a 0.12 mg/L polymer dose would produce a clarified turbidity of less than 5 NTU after three to five minutes of settling, simulating the plant design of about 7 minutes hydraulic retention time (HAT). Because jar tests can represent optimum conditions, a polymer dose of 0.2 mg/L was selected for application at the pilot.

•o 30 20

10 · 0

I :(:

Date

Exhibit 2. MF TMP During Polymer Pilot Test.

Pilot Plant Testing Periods Determine if Polymers are Causing Irreversible Fouling

polymer addition. A vertical paddle flocculator was used followed by an inclined plate clarifier. Monochloramine addition was added upstream of the MF system to maintain a monochloramine residual at the RO feed of approximately 1.5 mg/L.

The pilot plant used previously to confirm and refine full-scale plant design criteria and operating conditions, respectively, was recommissioned and used for the polymer evaluation. Ferric chloride was fed at a dose of 50 mg/ L to 100 mg/L to maintain the RO feed orthophosphat e concentration below 0.6 mg P/ L. Polymer was fed at a rate of 0.2 mg/ L for most of the test, although it was increased to 0.9 mg/L for one week at the end of the pilot to simulate a massive overdosing event. The polymer was fed into the rapid mix chamber to provide good mixing and to allow the ferric chloride to hydrolyse prior to

The pressurised MF membranes were operated at a flux of 65 litres per square metre per hour (LMH). MF backwash was conducted every 20 minutes; backwash water was returned to the rapid mix tank. The schedule of extended flux maintenance (EFM) was: • Day 3: Citric acid and sulphuric acid EFM • Day 6: Citric acid and sulphuric acid EFM • TCPWithoutPolymef TCPW<ha AodEFM

MF Temperature Corrected Permeability (TCP) 1,,4 0

.

1 30 1 20

ci'

1 00

D..

~ 090 ~

I

··~

r

~

1. 10

080

....

"' ~

.

0.60

D.. 0 50 (.)

t-

o.•o

~

0 30

f - :-

(

l

High pfote tood!ng 1011discussion

(.) 0 70

-:·,

IP

~v1

---r

.

··..

• Day 10: Reset to Day 1 and repeat EFM sequence The RO system included three stages in a 4:2:1 array. It was operated at a flux of approximately 17 LMH and 85 per cent recovery. Antiscalant was dosed at 5 mg/ L. The pilot plant was first operated without polymer addition for nine days to establish "baseline" operating conditions for the MF and RO units and to conduct a complete EFM cycle (acid/ acid/hypo). The pilot was then operated with polymer addition for approximately 9 weeks. Polymer dose was set at 0.2 mg/L for most of this period, but was increased for about 1week to simulate a massive overdosin g event.

A clean-in-place (GIP) was cond ucted on the MF system immediat ely after the 9-day baseline operating period and after the 9-week polymer addition period to restore permeability and minimise any impact "pre-fouling" might have on polymer use. A comparison of post-GIP permeability following each period helped determine if polymer use was causing irreversible fouling.

Hypo EFM O

• Day 9: Sodium hypochlorite EFM

Hlghpoly,!Wdo&e tesbng

Pilot Testing Results Reveal Importance of Clean-In-Place Procedures in M embrane Performance

-·

020

0 10 000

~

f

:(:

i

"

~

~d,

i'e. i ::::

~

~ ~

i1

Exhibit 3. MF TCP During Polymer Pilot Test.

5 2 NOVEMBER 2010

water

!

Date

~

~

I ~

g l,

,!

"'

i

~

i ~

g

j

"'

Microfiltration me mbra nes

~ i ~

The primary metrics for evaluating the ongoing performan ce of the MF membranes with polymer addition are Transmembrane Pressure (TMP) and

technical features

GISTER YOUR INTEREST The 18th International Botanical Congress is calling for expressions of interest for registration, sponsorship and exhibition.

This international Congress is focusing on the latest developments and best practice in Botanical Science Research. Over 3000 delegates from around the world are expected to attend. This event is held only every 6 years.

Register your interest in attending, sponsoring, exhibiting or presenting at the congress on-line at www.ibc2011.com The Congress website also features up-to-date information on the Scientific Program and Keynote Symposia.

To discuss sponsorship and exhibition opportunities Tel: (02) 9254-5000 and ask for the Sponsorship and Exhibition Manager of the IBC2011 Congress or visit the website for more details.

membranes processes Temperature Corrected Permeability (TCP). In addition, effective restoration of membrane permeability by clean-in-place (CIP) following polymer addition is important so that polymer (or polymer-associated solids) do not accumulate in or on the membrane causing irreversible fouling.

~

RO Normalised Permeate Flow

4.25._~--- - - -- ~-- - - -- -- ----------,

·-

4.00

nt

3.75

• Decreased solids loading to the MF system because of lower clarifier effluent (MF feed) turbidity and solids levels and/or

The CIP performed at the conclusion of the polymer addition period completely restored permeability (Exhibit 2). The post-GIP temperature corrected permeability was 1.3 LMH/kPa for operation directly following the CIPs conducted before and after polymer addition.

Reverse osmosis membranes The primary metrics for evaluating the impact of polymer addition on RO performance are normalised permeate flow (NPF) and normalised differential pressure (NDP), by stage.

A decline in NPF, or an increase in NDP, beyond that observed during 2007-2008 pilot testing (no polymer

54 NOVEMBER 2010 water

~9111 w/ PolyrM, X Stagt,2 w(Polymer XStlgt,3w/oPolYfMI'

• Si.ge3w/Pofyrnef

Date

Exhibit 4. RO Normalised Permeate Flow.

RO Normalised Differential Pressure

110

-----------------0-.----------

100

..

80

D..

~ 70

D..

.,,C 60 •

..

~ 50

E 40 0

•St•

z

-

30

20

• A lower level of pore fouling because colloidal-size particles were more effectively settled Furthermore, the trendline is made more positive by the high TMP values that occurred during the middle of the run when polymer was not added.

Omitting these values from the regression would likely produce a flat or negative TMP slope and an indefinite cleaning frequency.

• Suige 1 w/o Pdymw

• Stage 2 wlo PolYff*

The pilot plant was operated from 13 December 2008 through 13 February 2009 (9 weeks) with approximately 0.2 mg/L polymer addition. There was a four-day shutdown January 4 to 8 due to a raw water pump failure. The TMP during this period was typically between 60 and 70 kPa, which matches previous piloting experienced at the LPAWTP during 2007 and 2008.

Exhibits 2 and 3, which present TMP and TCP values as a function of operating date, indicate that polymer did not negat ively impact MF performance. In fact, membrane performance may have improved as evidenced by the lower TMP and higher TCP during polymer addition.

This effect was most likely the result of either:

refere ed paper

1 w/o Polymer

Sl.;,e1w/ Pofyrner • St•2w/opdYI"* )( Sl1t1ge2w/ potrme< x Sl•ge: 3 w/o polymtr • Stage3w/poi'fm11r

10 0

! ~

8

8

.,

d,

'*

1 0

'*

~

Exhibit 5. RO Normalised Differential Pressure.

-

RO Stage 1 Normalised Permeate Flows Before and After Polymer Addition 4.00 - - - - - - - - - - - - - - - - - - - - - - - - - - ,

3.95

f

3.90

S

3.85

~

3.80

I

3.75

I

C")

ii: ~ 3.70

l, 3.65

-MaximJm

j

3.60 +--~ -- - - 4 - - - - - - - - - - . - - _ _ , . 99% Upper Cont. LIMII .!!l -Mean ~ 3.55

0 z

3.50

-------Minimum

+------------3 .40 - - - - - - - - - - - - -- -- - - - - - - - -----< 3.45

S1 Begiming

S1 End

Exhibit 6. RO Stage 1 Normalised Permeate Flows at the Start and End of the 0.2 mg/L Polymer Addition Period.

technical features

oO

Advertorial

AquaArmourO

Saving money & Australia's most precious resource In a world first, an innovative fully tested and cost effective solution for combating evaporation on major dams and water storages is being launched in Australia and is being manufactured in Victoria.

This Australian product wi ll significantly reduce the impact of drought on farmers and regional communities. It will do this at considerably less cost in comparison to water recycling and desalination.

The product known as AquaArmour™ is proven to save 88% of water lost due to evaporation.

AquaArmour™ is a genuine and simple answer to what has been a continued challenge in water management - preventing the huge loss of water through evaporation.

On average, we lose four times the total annual rainfall to evaporation.

That is a huge amount of water and a major contributor to low water levels in our dams.

AquaArmour™ will complement desalination and provide water authorities in catchment areas not serviced by desalination w ith an answer to combat evaporation to preserve precious water resources.

Water authorities, miners and those in the agricultural sector can now protect their water resources saving shareholders, rate payers and ultimately farmers money.

While it's been a wetter year, many catchments are still well below optimum levels. The introduction of this new solution provides an opportunity to preserve recent rainfall and maintain catchment levels. Climate change predictions indicate that this year will be an anomaly, which means we must protect our rainfall from the wetter years so we are better prepared for what's ahead.

It's far easier and more cost effective to manage water than it is to make it. Protecting water in

existing dams and catchments is more energy efficient and cost effective than building new ones, not to mention the cost of additional water purification plants and the environmental issues relating to these applications.

Another benefit of AquaArmour™ is that it prevents 90% of visible & UV light penetrating the water body, inhibiting algal growth .

The first major deployment of AquaArmour™ will take place for Grampians Wimmera Mallee Water at Ouyen in Victoria where evaporation can reach up to three metres per annum.

This initial two hectare deployment placed on the township's dam will save enough water to supply three quarters of the entire town's population of 1, 100 people with their annual water requirements under the Victorian Government's 155 Water Initiative, year after year.

This world first deployment will give long term water security to a small township and will showcase this great Australian innovation, AquaArmour™, to an international audience.

For further information please contact Aqua Guardian Group on: P: +61 3 8530 2000 E: sales@aquaguardiangroup.com W: www.aquaguardiangroup.com

membranes processes use) may indicate polymer fouling.

Exhibits 4 and 5 show NPF and NDP, respectively, for the duration of the pilot test.

The change in normalised permeate flows over the 0.2 mg/L polymer addition period are shown using "box and whisker" plots in Exhibits 6 to 8 and summarised in Table 1. Note that the pilot plant was shut down from 4 January th rough 8 January d ue to a failed raw water pump. The shutdown had a sig nificant beneficial impact on normalised permeate flows and normalised differential pressures. The normalised permeate flow after this shutdown increased significantly in Stages 1 and 2, stabilising at higher values, most likely due to either the removal of a foulant from t he RO membrane that was present before the pilot was restarted on 2 December 2008 or a "relaxation" of the membrane, which can occu r after prolonged shutdowns. A similar, but opposite effect was observed for differential pressure; pressures in Stages 1 and 2 decreased after the shutdown and stabilised at lower values, consistent with the removal of fou lant. Also note that the normalised permeate flow in all stages consist ently declined from 3 December 2008 through 20 December 2008, after wh ich time the rate of decline decreased significantly. A small , initial permeate f low decline is not atypical for operation on tertiary effluent due to organic fouling.

Inspection of the figures and tables suggest that polymer did not negatively impact RO performance, as demonstrated by the following information:

• Normalised permeate flow: The mean NPFs in Stages 1 and 2 and the total system increased during this period, most likely due to the cleaning effect observed during the four-day shutdown. The mean normalised permeate flow decreased in Stage 3; however, the difference in means between polymer and non-polymer periods is not statistically significant at the 99 per cent confidence level (Table 1). In addition, no gradual decreases in normalised permeate flow over the operating period were observed, which indicates that polymer accumu lation and fouling did not occu r in any of the stages.

• Normalised differential pressure: The NDP for Stages 1 and 2 and for the system decreased d uring t he period of polymer add ition, but

56 NOVEMBER 2010 water

~

refereed paper

RO Stage 2 Normalised Penneate Flows Before and After Polymer Addition 1.55 1.50

~ ;;; 1.45 E

i

1.40

£_

1.35

Cl)

oi 1.30 Cl)

e 1.25 Cl)

a.

"Cl

1.20

-~cu

1. 15

----

l

I- - - - -

__ J

e 1.10 0

z

1.05 1.00 S2 Beginning

S2 End

Exhibit 7. RO Stage 2 Normalised Permeate Flows at the Start and End of the 0.2 mg/L Polymer Addition Period.

RO Stage 3 Normalised Penneate Flows Before and After Polymer Addition

-c..;

.c

§..

~

i.i: CII

j

e

l..

"Cl

0.40 ~ - - - - - - - - - - - - - - - - - - - - - - - - 0.38 + - - - - 0.36 0.34 0.32 0.30 -· 0.28 0.26 0.24 0.22 0.20

-~ 0.18 ~ C 0

0.16

- --- -------

Z 0.14

0.12 - - - - - - - - -- - 0.10 + - - - - - - - - - -- -- - , , - - - - - - - - - - - - - - 1 S3 Beginning

S3 End

Exhibit 8. RO Stage 3 Normalised Permeate Flows at the Start and End of 0.2 mg/L Polymer Addition Period.

Table 1. Change in Normalised Permeate Flows Over 0.2 mg/L Polymer Operating Period.

Period

Normalised Permeate Flow (m 3/hr)* - Mean Values Stage 1

Stage 2

Stage 3

Total

Pre-Polymer Addition (20 Dec 2008 after initial decline)

3.58

1.26

0.30

5.13

Post-Polymer Addition (4 Feb 2009)

3.81

1.39

0.27

5.48

Is the pre-polymer mean statistically different than the post-polymer mean (99% Confidence Level)?

No Yes N/A; the means 6.8% are not statistically increase different because of data scatter • Normalised permeate flows shown represent the mean value of twenty-four hourly data points

Per cent change

Yes

Yes

6.4% increase

10.3% increase

increased for Stage 3. These changes suggest that foulants that had accumulated in Stages 1 and 2 during previous operation of the RO pilot

unit (before 2 December 2008) were mobilised during the four-day shutdown in January and subseq uently transferred to Stage 3

technical fea ures

~

refereed paper

upon restart . In addition, no gradual increases in differential pressure during the polymer addition period were observed in any of the stages. These changes (or lack of) correlate wel l with those of normalised permeate flow and suggest that any free polymer addition that might have passed through the MF unit was not accumulating on the membrane surface or w ithin the feed-bri ne spacer of the RO elements.

Conclusion Based on the observed increase in normalised permeate flows in the MF system as well as RO Stages 1 and 2, and the "potential" nominal decline in RO Stage 3, along with the lack of any observed gradual change in NPF or NOP, it can be concluded t hat polymer addition at 0.2 mg/ L did not negatively affect MF or RO performance nor was there any evidence of polymer accumulation or fouling.

Acknowledgments WaterSecure is the owner of LPAWTP and funded the work described in this paper. WaterSecure is the Queensland Government authority responsible for the Western Corridor Recycled Water Scheme, which includes t he LPAWTP. This paper has been adapted from one originally presented by the authors at t he Water Environment Federation's Membrane Applications Conference in Anaheim, California, USA in June 2010.

membrane processes The Authors

Larry Schimmoller is a Princi pal Technologist and Senior Project Manager with CH2M HILL, and is based in Denver, Colorado, USA.

Jim Lozier is a Technology Fellow and Membrane Technology Leader with CH2M HILL, and is based in Phoenix, Arizona, USA. Larry and Jim were both involved with the commissioning of the Luggage Point Advanced Wat er Treatment Plant.

Tony O'N eill is a certified water treatment operator. He leads the operations of CH2M Hill 's Luggage Point Advanced WTP and the Gippsland Water Factory.

Just like our planet, we live for water. ALS Water Resources Group offers a leading range of chemical, biological and microbiological testing solutions to the greater water industry, including: Drinking/Potable Water Catchment Management/Water Resource â&#x20AC;˘ Desalination Waste Water Treatment/Industrial Water

Recycled Water General Water Quality Monitoring Trade Waste Sampling of Catchment/Drinking Water

Drink it in. Visit www.alsglobal.com or call 1300 32 69 47 to get started. ~ All ALS WRG laboratories are ISO 17025 (NATA) accredited to ensure laboratory results meet the highest standards.

V

ALS also employs multiple layers of data verification to ensure its accuracy.

membranes processes

~

refereed paper

DUAL NANOFILTRATION/ REVERSE OSMOSIS SYSTEMS B Bolto, M Hoang, T Tran Abstract In seawater desalination it makes sense to have a low pressure nanofiltration (NF) stage before reverse osmosis (RO), taking out the multivalent ions, plus some sodium chloride and organics, leaving a feed for the following RO system that is of much lower ionic strength than the original raw water. The consequence is that there is a smaller osmotic pressure effect and the applied pressure requirement and hence energy need overall is much less. Also , the yield of product water is higher. Unfortunately, the total costs for an NF/RO system are usually about 10% more than for an RO only system. However, there are certain situations where the approach is justified on economic grounds, because of the organics removal by NF resulting in a marked decrease in RO membrane fouling and a significantly enhanced membrane life. A ful l scale plant has been operating on this basis in Saudi Arabia which lowers the pressure used in the RO plant by 17% and gives a 30% increase in water yield. Other examples show a 25-30% lower energy need and a 50100% higher yield.

Introduction Nanofiltration (NF) membranes are used for softening and the removal of organic compounds. The pore size for such membranes is 0.5-1.5 nm, rangi ng between that of ultrafiltration (UF) and RO membranes (Schafer et al., 2005). Their most distinctive characteristics are the lower pressure need of 0.3-3 MPa versus 4-7 MPa for RO, and their unique rejection properties, in that they can take out: • virtually all multivalent anions such as su lphate and phosphate

58 NOVEMBER 2010 water

Table 1. Dual NF/RO systems with seawater feed.

Energy Saving,%

17

Water Recovery Increase,%

Reference

100

Hassan et al. , 2000

50 30

Eriksson et at., 2005

-20

Tanninen et al., 2005 Le Gouellec et al. , 2006

25-30 20

Drioli et al., 2002

• 0-70% of sodium chloride, plus • uncharged and positively charged species, depending on their size and shape.

NF/RO Combinations Sea water desalination Pre-treatment of seawater with NF provides an excellent feed water for RO, allowing operation at much higher fluxes and recovery rates, and also at lower pressure as the osmotic pressure to be overcome in the RO stage is lower because of the reduced salts content (Hassan et al. , 2000). In a seawater desalination demonstration plant of capacity 72 kUday it was found that NF removed material that had MW >200 Da, reduced the CaSO 4 content by about 90%, and lowered the NaCl level by about 50%. There was a doubling of the RO product water yield following this pretreatment, plus an improvement in the fi nal wat er quality to <200 mg/L of total dissolved solids (TDS). Another study has shown an increase in the water recovery of 50% (Drioli et al., 2002).

Desalination with low energy and lower fouling, but higher capex.

A ful l scale plant has now been operated along these lines in Umm Lujj, Saudia Arabia, following experiments to determine conditions under which economic operation was possible (Eriksson et al., 2005). The plant treats 8.6 MUday, with the NF component: • reducing hardness from 7,500 to 220 mg/L • lowering TDS from 45,460 to 28,260 mg/L • rejecting sulphate to >99% • reducing divalent cations by 80-95% • lowering the pressure used in t he RO plant by 17 %.

The NF stage operated at a 65% conversion rate and the RO stage at a 56% conversion rate, giving an overall conversion of 36%. This compared favourably with a parallel RO-on ly plant which had a conversion rate of 28%, thus showing a 30% increase in overall recovery for the NF/RO system.

Other estimates of the energy saving are 25-30% (Tanninen et al., 2005).

A system designated as a two-pass NF process, known as NF2 , has been tested by the Long Beach Water Department (Leung and Rohe, 2006; Le Gouellec et al., 2006). It makes use of an initial loose membrane, followed by a tighter second membrane. Pilot testing has shown that the pre-treat ed seawater of TDS 37,500 mg/L can be lowered to 3,250 mg/L after the first pass, the residue being mainly NaCl, and then to 218 mg/L in the final product wat er. It has been demonstrated on the pi lot scale that a water quality equivalent to that from RO treatment alone can be achieved at a -20% lower operating pressure (1.7-3.6 MPa versus 5.9-8.3 MPa). A 568 kUday

technical features

~

membrane processes

refereed paper

prototype plant is planned, together wit h a parallel RO plant of similar size.

However, the recovery was only 40 %, versus 50 % for a typical seawater RO system, wh ich needs further explanation (Gouellec et al. , 2006). A summary of the advantages of the dual systems is given in Table 1.

The cost advantage of pre-t reatment with NF is not widely achieved (Fritzmann et al., 2007). Higher recoveries can in special circu mstances make up for the additional investment cost, but these are stated to be unlikely to occur in the practical operation of a desalting plant.

The operating costs for an NF/ RO syst em are usually about 10% more than for a two-pass RO system.

An important finding at the Umm Lujj plant was t hat t he NF membranes were less prone to fouling than polyamide RO membranes (Eriksson et al., 2005). This was partly ascribed to t heir lower salt rejection, as with the better salt rejection of t he RO membranes t here is a higher ion ic strength at the membrane/feed interface, which reduces t he repulsion force between the charged col loidal particles so t hat it is easier for them to aggregat e and form larger agglomerates.

lat.-

\

= d s.com

Table 2. Rejection of ions in seawater treatment by NF (Kyburz and Meindersma, 2005).

Ion

Rejection, %

SO/¡

99.9

Mg2+

98

Ca2+ HCOr TDS

91

56 37.7

These are more prone to deposit on the membrane surface. The higher costs for dual systems are based on the assumption t hat normally there would be minimal membrane foul ing. However, this was not the case at Umm Lujj. Here the NF pre-treatment for RO makes good economic sense because of the marked reduct ion in the RO fou ling rat e. It has been noted t hat biofouling is a serious operational problem in NF and RO installations, with membrane autopsies revealing biofou ling in 12 of 13 pilot plants investigated (Vrouwenvelder and van der Kooij, 2001 ). Autopsy of t he Umm Lujj NF membranes found t hat there was no biofouling after 9000 h of operation , with t he membrane deposits found t o be

mainly organic matter, iron, c hromium and fungus (AI-Amoudi and Farooque, 2005). Fouling of NF membranes has been reviewed recent ly (AI-Amoudi, 2010).

The applicat ion of NF before RO has been investigated also as a means of reducing the concentration of divalent ions t hat are responsible for membrane scaling. Another viewpoint however considers that the approach merely transfers the scaling problem further upstream (EI-Maharawy and Hafez, 2000).

Brackish water desalination Similar improved water recoveries have been obtained in brackish water applications when NF precedes RO (EIZanati and El-Khatib, 2007). The overal l cost with the more dilute feed waters can be less t han half that for seawat er desalination.

Scaling by silica can be prevented if treatment is carried out under alkaline condit ions, when the silica is present mainly in the anionic form, with water yields of 98% being cl aimed from river water of TDS 2000-3000 mg/ L when the operation is at pH 10.8 on raw water containing 10-30 mg/ L of silica (Mavrov

GeoVISIONâ&#x201E;˘ Borehole Camera

water NOVEMBER 2010 59

membranes processes et al., 1999). At that pH level the silica is 94% anionic (Albert and Serjeant, 1962).

However, precipitation of carbonates may become an issue (Her et al., 2000). The usual practice is t o operate at pH 8.5 or higher (AWWA, 2007). 14% of the silica is then anionic.

Two distinct mechanisms of scale formation, surface and bulk crystallisation, have been identified and explored (Lee and Lee, 2005). Attempts to decrease scale formation include: • increasing the fl uid velocity • adding antiscalants • chemical pre-treatment

where brine is isolated as a raw material for the inorganic chemical industry (Kyburz and Meindersma, 2005). NF prior to RO or multi-stage distillation is used to improve the recovery from a 45,000 mg/L feed by decreasing its scaling intensity.

Also, treatment of the RO concentrate containing mainly NaCl produces a purer form for chlori ne production in the ch loralkali industry. Another option is for the NF concentrate, mainly salts like MgSO4 , to be used as a source material for the magnesium metal industry. Typical result for various ion rejections by spiral wound units of an Osmonics DS5 DK membrane at a large NF plant in Bahrain are given in Table 2.

• ion exchange as a pre-treatment • combi ning a crystalliser with NF/ RO. Silica can form an amorphous inorganic deposit on membranes (AWWA, 2007).

Because of its often polymeric form, scale inhibitors have not been very effective in stabilising supersaturated si lica solutions.

There are, however, dispersants which are successful in controlling si lica deposition. Silica has a changeable nature, being generally considered to exist as undissociated silicic acid in most naturally occurring waters having a pH level of up to 8. The structure of precipitated particles is dependent on the silica concentration, the solution pH, the presence of ions such as calcium and especially magnesium, and t he temperature (Koo et al., 2001).

The reverse coupling of RO and NF for the desalination of brackish water is reported, with the emphasis on the effects of di- and monovale nt cations on operation (M'nif et al., 2007). At a pressure of only 0.6 MPa when run in sequence with RO before NF, the brine reject from the RO unit can be desalted with respect to divalent ions, so that the recovery of 40% for RO is increased to 80% with the cou pled system.

Conclusions In seawater desalination the use of NF to remove multivalent ions, plus organics and some sodium chloride before RO wil l result in a lower pressure requirement for the RO stage and a 17-30% lower energy need overall. Also, the yield of product water is 30-100% higher. However, this is only economic for seawater desalination where serious fouling of the membranes occurs in an RO-only system. Then the organics removal by NF results in a marked decrease in RO membrane fouling and a significantly enhanced membrane life. Similar improvements are possible in brackish water applications where the overall cost of desalting the more dilute feed waters can be less than half that for seawater desalination, and water yields can be doubled. Fouling and scaling are particular issues. Silica can be tolerated by operating at high pH levels of <8.5.

References

• Electroplating (Mohammad et al. , 2004).

Albert, A. and Serjeant, E. P. (1962). Ionization Constants of Acids and Bases. Methuen, New York, pp. 174-175.

AI-Amoudi, A. S. (2010) . Fact ors affecting natural organic matter (NOM) and scaling fouling in NF membranes: A review. Desalination 259, 110.

AI-Amoudi, A. S. and Farooque, A. M. (2005).

Performance restoration and autopsy of NF membranes used in seawater pretreatment.

Desalination 178, 261 -271.

Ashaghi, K. S. , Ebrahimi, M. and Czermak, P. (2007). Ceramic ultra- and nanofiltration membranes for oilfield produced water treatment: a mini review. The Open Environmental Journal 1, 1-8 .

AWWA (2007). Reverse osmosis and nanofilt ration. Manual of Water Supply Practices M46, 2nd Edition, American Water Works Association, Denver.

Drioli, E., Criscuoli, A. and Curcioa, E. (2002) .

Integrated membrane operations for seawater desalinat ion. Desalination 147, 77-81.

EI-Maharawy, S. and Hafez, A. (2000). Technical management of RO system. Desalination 131 , 173-188.

Of the various industries, that of most interest is NF/RO seawater desalination

EI-Zanati, E. and El-Khatib, K. M. (2007).

Integrated membrane-based desalination system. Desalination 205, 15-25.

Industrial water desalination

Wastewaters from a wide range of industries have been subjected to NF treatment. The industries include (Schafer et al., 2005) • Food production • Chemical processing • Pulp and paper • Textiles • Metal separation • Petroleum (Ashaghi et al., 2007)

60 NOVEMBER 2010 water

refereed paper

Eriksson, P., Kyburz, M. and Pergande, W. (2005).

NF membrane characteristics and evaluation for sea water processing applications.

Desalination 184, 281-294.

Fritzmann, C., Uiwenberg, J., Wintgens, T. and Melin, T. (2007). State-of-the-art of reverse osmosis desalinat ion. Desalination 216, 1-76.

Hassan, A. M., Farooque, A. M., Jamaluddin, A. T. M., AI-Amoudi, A. S., AI-Sofi, M. A. K., AIRubaian, A. F., Kit her, N.M., AI-Tisan, I. A. R. and Rowaili, A. (2000). A demonstration plant based on the new NF-SWRO process.

Desalination 131 , 157-171.

Her, N., Amy, G., Jarusutthirak, C. (2000).

Seasonal variations of nanofiltration foulants: identification and control. Desalination 132, 143-160.

Koo, T., Lee, Y. J. and Sheikholeslami, R. (2001).

Silica fouling and cleaning of reverse osmosis membranes. Desalination 139, 43-56.

Kyburz, M. and Meindersma, G. W. (2005).

Nanofiltration in the chemical processing industry. In: Nanofiltration - Principles and Applications, A. I. Schafer, A. G. Fane and T. D. Waite (Eds.), Elsevier, Oxford, pp. 331-332.

Le Gouellec, Y. A., Cornwell, D. A., Cheng, R. C., Tseng, T. J. , Vuong, D. X., Wattier, K. L., Harrison, C. J. and Childress, A. E. (2006). A novel approach to seawater desalination using dual-staged nanofiltration. Awwa Research Foundation, American Water Works Association, Denver.

Lee, S. and Lee, S. H. (2005). Scale formation in NF/RO: mechanism and control. Water Sci. Technol. 51 (6-7), 267-275.

Leung, E. and Rohe, D. L. (2006). Prot otype t esting facility for two-pass nanofiltration membrane seawat er desalination process. In: Membrane treatment for drinking water and reuse applications: A compendium of peerreviewed papers, K. J. Howe (Ed.), American Water Works Association, Denver, pp. 45-87.

Mohammad, A. W., Othaman, and Hila!, N. (2004).

Potential use of nanofiltration membranes in treatment of industrial wast ewater from Ni-P elect roless plating. Desalination 168, 241-252 .

Mavrov, V., Chmiel, H., Heitele, B. and R6gener, F. (1999). Desalination of surface wat er to industrial wat er with lower impact on t he environment. Part 3. Desalination under alkaline conditions. Desalination 123, 33-43.

M' nif , A., Bouguecha, S., Hamrouni, B. and Dhahbi , M. (2007). Coupling of membrane processes for brackish water desalination.

Desalination 203, 331-336.

Schafer, A. I., Fane, A. G. and Waite, T. D. (2005) .

Nanofiltration - Principles and Applications, A. I. Schafer, A. G. Fane and T. D. Waite (Eds.), Elsevier, Oxford, Chapt er 1.

Tanninen, J. , Kamppinen , L. and Nystrom, M. (2005). Pretreatment and hybrid processes.

Nanofiltration - Principles and Applications, A. I. Schafer, A. G. Fane and T. D. Wait e (Eds.), Elsevier, Oxford, pp. 253-254.

Vrouwenvelder, J. S. and van der Kooij, D. (2001).

Diagnosis, prediction and prevent ion of biofouling in NF and RO membranes.

Desalination 139, 65-71.

The Authors Dr Brian Bolto, Dr Manh Hoang and Dr Thuy Tran (email: brian. bo lto@csi ro .au; manh.hoang@csiro.au; thuy.tran@csiro.au) work for CSIRO Materials Science and Engineering, Clayton, Victoria.

technical features

[El

customer service

refereed paper

SEQ WATER GRID DEMONSTRATES CONFIDENCE IN WATER QUALITY C Owens, C Snape Abstract The introduction of a regional approach to water management in South East Queensland has allowed for the amalgamation and development of tech nical skill sets, and has provided the opportunity to have a coordinated , consistent service to customers. The Customer Confidence Report (Bulk Water) is a vital tool in achieving these efficiencies, providing members of the South East Queensland public with access to reg ular updates on the quality of t he SEQ Water Grid's bulk drinking water supply.

Published monthly, the SEQ Water Grid Manager prepares the Customer Confidence Report (Bulk Water) in partnership with t he three Grid Service Providers; Seqwater, WaterSecure , and LinkWater. Further, it is formulated as an interactive web-based tool , and aims to educate consumers on the structure and funct ion of the SEQ Water Grid while increasing public underst anding of, and confidence in, the region 's bu lk water supply.

Introduction Over the past two years, South East Queensland has undergone a significant amount of change in the way its drinking water is supplied and managed. The SEQ Water Grid began operation in July 2008 under new management arrangements, including the amalgamation of 22 separate entities that previously owned and operated the bu lk water t reatment, storage and transport infrastructure.

These responsibilities were transferred to fou r bulk water and three distribution retail entities. The establishment of the SEQ Water Grid has lead to an improvement in t he ability to supply high quality water from a variety of sources and to extensively monitor water to ensure safety and reliability (Owens, 2010).

The new arrangements have also led to the ability for whole-of-government responses to water quality incidents.

Following the North Pine f luoride incident, and as recommended in the Investigation into the Fluoride Dosing Incident - North Pine Water Treatment

Figure 1. Initial interface of the Customer Confidence Report (Bulk Water), showing sites reported and major transmission pipelines; 'water droplet' icon for zone 2 also shown.

Works - April 2009 (Pascoe, 2009), the whole-of-government responded by chargi ng the SEQ Water Grid Manager wit h providing greater t ransparency and accountability for the quality of South East Queensland's drinking water supplies, by way of monthly public reporting .

The prospect of a monthly report provided the SEQ Wat er Grid Manager with an opportunity to address two key

An interactive web-based monthly report, which aims to educate consumers.

public education concerns. Firstly, the general misunderstanding amongst the broader commu nity regarding the structure and function of the SEQ Water Grid, and secondly, the technical nature of water quality monitoring of which there is limited understanding among key audiences. To address t hese issues, the SEQ Water Grid Manager developed an online, interactive report t hat wil l help to educate users on both the structure of the SEQ Water Grid, and the quality of the bu lk drinking water supplied.

With water quality results updated monthly, the Customer Confidence Report (Bulk Water) is the most frequently updated drinking water quality report available in Aust ralia show ing

water NOVEMBER 2010 61

customer service monthly and 12-monthly compliance with the Australian Drinking Water Guidelines (2004).

~

Table 1. Sampling specification for the Customer Confidence Report (Bulk Water).

Parameter

Audience The Customer Confidence Report (Bulk Wat er) endeavours to cater to technical and non-technical audiences. To achieve this, the report includes three levels of detail: 1. A whole-of-Grid report assessing results against the Australian Drinking Water Guidelines (2004) long-term compliance measures for three categories of parameters; microbial, chemical, and aesthetic. Each category is depict ed as a 'water droplet'; easily interpreted by way of a fuller drop meaning more parameters in compliance.

2. Regional results for four zones within the SEQ Water Grid district, showing water quality performance in terms of monthly and 12-monthly compl iance with the Australian Drinking Water Guidelines (2004), as well as a comparison of each zone's water quality performance juxtaposed with that of the SEQ Water Grid as a whole.

3. Site-based reports showing the month's compliance to the Australian Drinking Water Guidelines (2004) for 19 sites across the region representative of the SEQ Water Grid.

Method Report overview The Customer Confidence Report (Bulk Water) is presented in a unique, interactive manner (Screen shot Figure 1).

When loading the report's web page, users are able to navigate to reports based on geographic location; either by selecting the sampling point nearest to the desired location, or by selecting a 'zone', indicated by a water droplet icon.

In order to increase the report's educational value, the interface also includes the ability to toggle on and off important SEQ Water Grid assets such as major bulk transmission pipelines and the reported sampling locations. It also shows major bodies of wat er, and the location of cities and towns.

The report is published within 22 business days of the end of every calendar month. This timeline is achieved through a purpose-built, extensible Markup Language (XML) data commun ication process, established between the SEQ Water Grid Manager and the three Grid Service Providers; Seqwater, WaterSecure and LinkWater.

62 NOVEMBER 2010 water

refere ed paper

Sampling frequency per site

Escherichia coli

2: 1 per week

Fluoride

2: 1 per week

Total trihalomethanes

2: 1 per month

Lead

2: 1 per month

Copper

2: 1 per month

Manganese

2: 1 per week

Hardness

2: 1 per month

pH Turbidity

2: 1 per week

Total dissolved solids

2: 1 per week

True colour

2: 1 per week

Alkalinity

2: 1 per month

Aluminium (acid soluble)

2: 1 per week

Iron

2: 1 per week

Health microbial

•

Category Health chemical

• • • • •

2: 1 per week

Analytes reported The report is based on the Grid Service Providers' laboratory results for 14 analytes relevant to the South East Queensland region. Samples are analysed according to a monitoring program, and are grouped under three categories (Table 1).

Each of the Grid Service Providers contract their water quality t esting to different laboratories, therefore results are often returned with differing Li mits of Detection, dependent on the analysis method used. For the purpose of the Customer Confidence Report (Bulk Water), agreement across Grid Service Providers was gained regarding the handling of data. Specifically, the Grid Service Providers agreed to a single reportable Limit of Detection for each parameter. Confirmation regarding the handling of results reported as less than the Limit of Detection was also needed; these results are replaced with zero when used in calculations. The Grid Service Providers also agreed to the analytes, sites, and frequencies to monitor. The monitoring program was determined in the context of the empirical water quality risk profile of the SEQ Water Grid.

Included sites Nineteen representative sites within the SEQ Water Grid are included in the Customer Confidence Report (Bulk Water). These consist of eight water treatment plants, 10 bulk water blending and transmission points, and the Gold Coast Desalination Plant.

Individual sites are summarised into four geographic zones supplied by distinct components of the SEQ Water

Aesthetic

• • • • • • • • • •

Grid. By selecting the relevant zone, users are able to see a summary of results representative of t he majority of bulk water supplied to their area.

Statistics Throughout the report, monthly statistics show whether samples have met the values listed within section 10.8 of the Australian Drinking Water Guidelines (2004) for the given month.

Twelve-monthly compliance is calculated using the appropriate longterm compl iance measure listed within section 10.7 of the Australian Drinking Water Guidelines (2004). That is, applying the appropriate 98 per cent pass rule for Escherichia coli, the 95th percentile rule for health-related chemical parameters, and the median pass rule for aesthetic parameters. Where a parameter has both a health and an aesthetic value, the respective guideline value is assessed using the relevant rule.

Whole-of-Grid report The purpose of t he whole-of-Grid report is to provide a summary of the SEQ Water Grid's 12-monthly compliance with Australian Drinking Water Guidelines (2004) long-term measures for important health and aesthet ic parameters.

The whole-of-Grid report is shown when users first click on a reporting zone's water droplet icon. It provides a high-level representation of the quality of bulk water supplied to South East Queensland.

The report assesses analyte performance for the three parameter categories: 1. ' Health - microbial'.

technical features

Q

customer service

refereed paper

2. 'Health - c hemical' .

3. 'Aesthetic'.

By using information from al l 19 reported locations, the whole-of-Grid report is representative of approximately 80 per cent of the populat ion servic ed by the SEQ Water Grid.

Data for water reporting zone 2 for

Annual performance for water reporting zone 2 Pnmmeter name

Zone reports f&J/1

The purpose of the zone reports is to show an assessment of major functional segments of the SEQ Water Grid. The reports detail monthly and 12-monthly compliance to the Aust ralian Drinking Water Guidelines (2004).

There are four zones reported; for each zone, four reports are prepared. These include: 1. A comparison report, showi ng the percentage of samples in compliance to the Australian Drinking Water Guidelines (2004) for the month, in comparison to t hat of all sites reported. The report is broken down into the th ree parameter categories described above.

2. The zone's monthly performance, showing t he number of samples performed in the reported month for each parameter, and indicating whether t hey have passed or failed relevant guideline values.

3. The zone's annual performance, against long-term compliance measures (Figure 2), updat ed each month using a 12-month rolling window.

4. The zone' s deta iled monthly performance, indicating for each reported parameter: a. The number of samples analysed for each parameter in the zone t hat month.

).if 2010

.l.mlll

.tStm;

lal

~

HeillUISDrnples Number

Number

thntpnssed

thntfllled

1503 452 117 194 194 451

0 0 0 0 0

Aesthetic samples Number Number thntpnssed thntf'llled

l

llimll=i:i glJ

.IWlKttt 11:!t.Jlri<solwdgl:I ~

compiant to ADWC

98"'

9S"'

pen:ent

pera,ntlle

Average

Yes Yes Yes Yes Yes Yes

0 0 0 0 0

194 451 194 452 452 371 452

J 0

Yes Yes Yes Yes Yes Yes Yes

illlilDlÂĽ .ilLIIDIIJ

416

0

Ill!]

449

l

Yes Yes

Figure 2. A zone's annual performance report, showing complian ce to the relevant Australian Drinking Water Guidelines (2004) long-term measures; updated each month based on the 12-monthly rolling window.

Site reports

Detailed results for the 19 reported sampling sites are also available.

The p urpose of the site reports is to show the water quality performance

(e LOWARA

of indi vidual SEQ Water Grid assets. Acceptabi lity c an be inferred by t he month's complianc e to the Australian Drinking Water Guidelines (2004).

Hydrovar is the world 's first pump mounted microprocessor based pump controller. It does much more than just change pump speed, it truly manages your pump to match a wide range of system conditions.

Energy savings of up to 70% can be achieved.

b. The minimum, average, and maximum value for each analyte that month.

Contact us now about your specific application.

c. The number of samples analysed in t he 12-month rolling window.

Melbourne

d. The annual resu lt for percent and percentile based compliance measures.

03 9793 9999 Sydney

e. The number of sample exceedences per month under the Australian Drinking Water Guidelines (2004).

For the purpose of transparency, all previously published monthly zone reports are available for comparison. Any exceedences reported within monthly and annual statistics are linked to a popup wi ndow which describes the exceeding sample's result, explanatory notes regarding the exceedence, and if app licable, relevant reg ulatory com mentary and the incident's regulatory status.

02 9671 3666 Brisbane

07 3200 6488

Web: www.brownbro5.com.au Email: info@brownbro5.com.au

~n n Iii 1

Bro'-'."n Brothers

!'..lfll1 Engineers

Australia Pty Ltd

water

NOVEMBER 2010 63

G

customer service For each parameter reported, the site reports include monthly figures for: 1. The number of samples for each parameter analysed at the site.

2. The minimum, average, and maximum sample value returned for each parameter.

3. The number of exceedences for each parameter under the Australian Drinking Water Guidelines (2004).

4. Explanatory notes for any exceedences.

Discussion The three levels of complexity included in the report allow users of all technical levels the opportunity to understand the results present ed. The who le-of-Grid report is aimed at basic users; its purpose is to answer the fundamental question of whether South East Queensland's bulk water supply is of good quality. In addition , the zones' monthly performance, annual performance, and comparison reports are aimed at intermediat e users. The detailed zone and site reports then cater to the needs of users with experience in the wat er quality realm; f or example, scientists and industry personnel.

Recently much of the public attention on the water industry has focused on lack of supply, a consequence of the 'millennium drought'. Several utilities provide innovative delivery methods for reporting dam levels, such as Melbourne Water's iPhone App (Melbourne Water, 2010), and Seqwater's interactive map (Seqwater, 2010). The technical nature of drinking water quality often results in information being disseminated in a strictly static format; for example, a table of information for each water treatment plant or area serviced. Alternatively, the Customer Confidence Report (Bulk Water) allows users to navigate toward wat er quality reports for an area of interest, while also learning about the basics of the SEQ Water Grid's infrastructure. This method of delivery enhances the level of engagement and educational value for the user. The Customer Confidence Report (Bulk Water) represents a preliminary step forward in innovative delivery methods for the public reporting of water quality information .

Another factor of the report's value is the frequency by wh ich it is updated.

The report is published within 22 business days of the end of every calendar month. By updating the report

64 NOVEMBER 2010 water

frequently, the SEQ Water Grid Manager is able to provide timely information to residents and businesses. The report is arguably the most frequently updated comprehensive drinking water quality report in Australia. In comparison, nearly all major Australian drinking water utilities prepare a static report once per year, if at all. Sydney Water and ActewAGL are notable exceptions; Sydney Water releases daily Cryptosporidium and Giardia updates, with all other parameters reported once per quarter (Sydney Water, 201 0); ActewAGL reports monthly against health and aesthetic targets, with compliance under the Australian Drinking Water Guidelines (2004) reported once per year (ActewAGL Retail, 2010).

As the Customer Confidence Report (Bulk Water) sources its data from three Grid Service Providers, data must be handled in a consistent and appropriate manner. New technologies were developed by the Grid Service Providers and SEQ Water Grid Manager in order to streamline the report's generation. The report's XML data provision standard , data collection tools, and report generation tools were designed with the scalability of sites and parameters in mind. The parameters chosen for the report were based on either their common use in water quality reporting, or the result of an assessment of risk in the South East Queensland context. The parameters and sites included in the report can be extended as the reporting process matures, and if the risk profile for South East Queensland's drinking water supply changes. Automation of Grid Service Providers' provision of data from Laboratory Information Management Systems is an imminent achievement.

Conclusions By preparing a comprehensive report monthly, the SEQ Water Grid Manager, in partnership with Seqwater, Wat erSecure, and LinkWater, is able to assure the confidence of the South East Queensland community in its bulk drinking water supply. The interactive nature of the report represents a technological step forward in the way government utilities engage with the public in techn ical areas.

The Customer Confidence Report (Bulk Water) can be accessed at http://seqwgm.com.au/Customerconfidence-report.aspx.

refereed paper

Acknowledgments The authors would like to thank all personnel involved in the development of the Customer Confidence Report (Bulk Water). The authors also gratefully acknowledge Seqwater, WaterSecure, and LinkWater for their continual provision of water quality data, as well as Queensland Health and the Office of the Water Supply Regulator, Department of Environment and Resource Management for their advice.

The Authors

Chris Owens is the Water Quality Officer at SEQ Water Grid Manager, and is responsible for the monthly publication of the Customer Confidence Report (Bulk Water). His background is in molecular microbiology and software architecture.

Chris Snape is a Micro-Electronics Engineer (Telecommunications), working within the SEQ Water Grid Manager as portfolio manager. He managed the delivery of the Customer Confidence Report (Bulk Water). Contact via med ia@seqwgm.com .au.

References ActewAGL Retail. (2010). ActewAGL: Water Quality: Water quality. Retrieved September 2010 from http://w ww.actewagl.com.au/wat er/quality/.

Melbourne Water. (2010). Melbourne Water iPhone app. Retrieved September 2010 from http://w ww.melbournewater.com.au/content/ new s_and_events/whats_new/whats_new/ 20100503.asp.

Owens, C.E., Crossin, R., Burrell, P., Davis, N. (2010). A regional approach to water quality monitoring and reporting in the South East Queensland Water Grid. Proceedings of the Ozwater ' 10 conference: Brisbane, Australia.

8-1 O March, 2010.

Pascoe, M. (2009). Investigation into the Fluoride Dosing Incident - North Pine Water Treatment Works - April 2009. Retrieved August 2009 from http://www.derm.qld.gov.au/ compliance/wic/pdf/fluoride_pascoe.pdf.

Seqwater. (2010). Interactive Map. Retrieved September 2010 from http://seqwater.com.au/ public/dam-levels/interactive-map.

Sydney Water. (2010). Sydney Water - Water quality. Retrieved September 2010 from http://ww w.sydneywater.com.au/ WaterQuality.

technical features

odour management

DISPERSION MODELLING: TRACING AND ANTICIPATING THAT ODOUR J Barclay Overview This review explores the theme of dispersion modelling, based on the Keynote Presentation to the AWA Odours Speciality Conference, Sydney, August 2010. A clear understanding, proper implementation and valid interpret ation of results are crucial for effective and realistic odour management.

Meteorology - surface - upper air - overvr.1ter buoy (optional ) - precipitation (optional) - gridded 30 data (optional)

ISource

IEmissions

Characteri stics - type of source. e.g. area - buildings - coordinates. Datum - terrain. LU

- collection method - detection threshold

Dispersion Model

The following topics are broadly discussed: • The difficulties faced with modelling odours Ground level Concentrations

• Why conduct dispersion modelling • Model inputs • Model reliability • Important role of meteorology and 3D Numerical prognostic data • Comparison of steady-state vs. nonsteady-state models • Model applicability (terrain, sea breeze, calm/stagnation) • Model outputs and interpretation of results

Introduction Odour perception is highly subjective and our individual responses to it can be quite different. Some people will be highly sensitive to short bursts of high peaks of odour, others will be very unhappy with constant low levels of odour over a prolonged period. Further, one odour can disguise or mask the presence of a second odour and, a cross over effect can occur where the weaker odour can become the dominant effect.

The management and control of 'Odours' is the single most important environmental issue in siting and implementing wastewater treatment and biosolids management faci lities in the world today. Expanding residential areas along with a less tolerant public are steadily placing significant pressure on these once isolated facilities. Current day odour regulations now insist that these industries be made accountable for their own discharges regardless of their length of prior occupancy or history of usage.

For regu latory purposes much of the

Figure 1. The three main dispersion model inputs are emissions, source characteristics and meteorological data of which the most important is the meteorology. focus of attention in the last couple of decades has been in trying to establish odour guidelines in the hope of bringing a degree of consistency to the control and regulation of odours. Less effort has been spent assessing the best tools suited to compute odour impacts with respect to accurate emission rates , source characterisation, and the important role of local meteorology, interpretation of modelled results, or, the suitability and applicability of one dispersion model over another. This paper aims to address several of these key issues central to the theme of effective management and odour regulation .

The annoyance level of any odour is typically a fu nction of the local meteorology, emission rate and source characteristics. The best tool available to compute odour impacts for regulation purposes is the dispersion model, which is able to predict downwind odour concentrations and explicitly account for the simultaneous combined effects of emissions, topography, land use and , meteorological data. Resulting isopleths

Modelling can conduct experiments that would be cost prohibitive in the real world.

maps of OU provide a spatial area wherein odour annoyance is expected to occur.

Modelling Odours The mechanisms of odorant dispersion in the atmosphere are the same as the dispersion of other pollutants. However, there are some special problems that must be considered when attempting to quantify a sources odour impact with dispersion modelling. Among them are; determining the emission rates of the pollutant, the high degree of subjectivity in the perception and intensity of odours, the short time period over which odours are observed, and, the enhancing or masking of odours by the combinations of different odours. In light of all this uncertainty, predicted maximum odour concentrations are at best a guideline or indicat or of the maximum odour concentration likely to be expected .

Despite the uncertainty with regards measuring and the accuracy of the predicted maximum odour concentration, dispersion models as tools to regulate odours are an important part of the process. Even if the emission rates are not accurate a good dispersion model sympathetic to three-dimensional meteorological data can give very accurate results especially in so far as predicting frequency of odour exceedances past a certain limit and,

water NOVEMBER 2010 65

odour management identifying worst case dispersion conditions. Advanced meteorological models have the ability to provide dynamically consistent three-dimensional flow such that a complex event like a sea breeze from its conception through to its collapse and the re-circulating of pollutants within the system are captured . The key factor to successful odour modelling is not so much accurate emission rates but good meteorological input data and a model that understands this and is able to interpret it well. In all cases, worst case odour conditions are a direct reflection on the local meteorology.

Figure 1 Shows the three main groups of model inputs, the meteorology which includes surface and upper air data, optional precipitation data, overwater buoy data and numerical gridded 3D data. The second input component is the emission rates, usually in g/s of each pollutant to be modelled and the third is the specific source characteristics such as type of source, the coordinates of the facility, the exit temperature (K), exit velocity (m/s) and diameter (m) of each source, terrain and Land use properties.

In many odour applications correct interpretation of a dispersion model

OCS are committed to providing products and services that demonstrate sound environmental management practices to suit a wide range of applications.

We specialise in design, installation and operation of odour control systems, including : Chemical dosing units

(sympathetic to complex meteorology) in so far as analysing frequencies of exceedances and identifying worst-case dispersion conditions along with a welldone complaint analysis is often sufficient to bring about the required change to an existing facility.

Why Conduct Dispersion Modelling There are several reasons why we would want to conduct air quality modelling.

Dispersion modelling provides quantitative estimates of impacts at many points over a wide geographical area.

Receptors are inexpensive and unlike monitors are not restricted to just a few measurement points. Dispersion modelling can serve as a ' numerical laboratory' and is able to conduct experiments that would otherwise be cost prohibitive in the real world.

Modelling is able to evaluate rare events for planning purposes and is well suited for accidental releases and evaluating worst case scenarios. Models are frequently used to evaluate impacts of proposed and future sources and are being used more often in real-time forecast modes as an operational emissions control tool.

We offer world c lass technology and innovative solutions backed by over 20 years experience.

Solve your odour problems ...call the experts today!

A wide range of extraction and treatment systems Investigation & monitoring

66 NOVEMBER 2010 water

Odour Control

Sy.!1..!!W!

Regulatory Models used in Australasia There are four types of regulatory air pollution models used in Australasia, of which three types are more commonly used. First, there are the steady state Gaussian plume models such as AUSPLUME, AERMOD and ISCST3, then the non-steady state Lagrangian Gaussian puff model, CALPUFF, the chemical Eulerian models such as CMAC and CALGRID and, lastly CSIRO's DAR Lagrangian Part Puff Particle model, 'The Air Pollution Model', (TAPM). The Steady state plume models of AUSPLUME and AERMOD have several basic assumptions. They assume straight line trajectories, and steady state meteorological conditions, they have spatially uniform meteorological fields, have no memory of t he previous hour's emissions and assume a non-zero wind speed. They are ideally suited for screening cases and near-field, flat terrain applications that are away from the coast, where conditions are expected to be steady-state. The second type of model, the Lagrangian approach solves a set of equations that mathematically follows the release of many pollution parcels as the plume moves through the atmosphere. The CALPUFF model is such a system which allows for variable and curved trajectories, the 3D meteorology has ful l spatial variability in the winds and turbulence fields.

CALPUFF retains information from previous hours of emissions and is well suited for modelling stagnation, fumigation and recirculation events. The Eulerian models, CALGRID and CMAQ are sim ilar to the Lagrangian model in that they also track the movement of a large number of pollution plume parcels as they move from their initial location.

The most important difference between the two types of models is that the Eulerian model uses a fixed 3 D Cartesian grid as a frame of reference rather than a moving frame of reference. These models are meant to be used explicitly for chemistry computations in particular ozone modelling. Th ese models are not frequently used as simple NOx and S0 2 chemistry effects can be easily simulated using another modelling system such as CALPUFF. The last type of model, the particle model is the most accurate in so far as being able to compute mean concentrations and fluctuations about the mean. The Air Pollution Model (TAPM) is a particle-puff model which means a Gaussian representation in the horizontal direction and particles in the vertical.

Particle models allow highly sheared

technical features

odour management environments and are desirable for computing concentrations above the ground. Particle models are usually more complex to run and less computationally efficient than the other models, although TAPM 's part-puff approach has meant that it is more efficient than other particle models. TAPM is best suited to regional scale modelling in gentle terrai n.

Kincaid SF 6 (Development)0rdered by Rank (Q1=3) Arc-Max 1000

1 :1 Line

Model Reliability and Accuracy Accuracy of model predictions is often a source of debate. Measurements and model predictions can be compared in a variety of ways, each providing a different perspective on model performance. A model may show good com petency in certain predictions (e.g. maximum concentrations) but poor in others (e.g. the frequency of concentrations above a certai n threshold). Usually the reasons for poor model performance are due to uncertainties in the input val ues for example, poor quality or unrepresentative meteorological, geophysical and source emission dat a, or, lack of modelling expertise, or, incorrect and unsuitable model depending on the application.

+ CALPUFF '15.8 CTDM CALturb Psv?sw

*

1

°io

CALP\Jff v5.8 CTDM PG

a

AERMOD v07026 NoSVW PsvPsw

•

ISCSCT3 ISCMIT

1000

100 Observed Chl/Q (ns/~

Figure 2. Results of various regulatory models when evaluated with the Kincaid Power Station Data set.

In general: • Models are more reliable for estimating longer time averaged concentrations than for estimating short-term c oncentrations at specific locations • Estimates of concentrations that occur at a specific time and site are poorly correlated with actual observed concentrations (paired in space and time) and are less reliable (mostly due to red ucible uncertainty such as error in pl ume location due to a wind direction error).

• Models are reasonably reliable in estimating the highest concentrations occurring sometime, somewhere in an area. Model certainty is expected to be in the range of a factor of 2.

It is common to apply a range of statistical procedures to evaluate model accuracy, these include graphical displays such as wind and pollution roses and quantile-quantile plots. Statistical performance measures include scalar and vect or mean wind speeds, standard deviations in measured and observed winds, RMSE errors (total plus systematic and unsystematic components). two model skill measures, the Index of Agreement , as wel l as the mean and standard deviations in modelled and observed wind speeds.

Figure 2 shows the results of various models when their performance was

Stable Onshore

Wel1M1xed Surface Layer

Flow

Land

Figure 3. Thermal Internal Boundary layer. Fumigation occurs in morning hours when mixing heights over land increase faster than over the sea. evaluated against the Kincaid Power Plant data set. Release of (Sulfur hexafluoride), SF6 from a tal l 187m st ack in Illinois, US was evaluated downwind at arcs of recept ors, stretching from 500 m to 50 km. The plot shows that of the models evaluated (CALPUFF with PG curves, CALPUFF with turbu lence based coefficients, AERMOD and ISCST3 (which is the same as AUS PLUME) that all models lie well within the line of a factor of 2 rel iability with some models clearly doing better than others.

Three-Dimensional Meteorology Sea breezes, Thermal Internal Boundary Layer fum igation, inversion break up fumigation , terrai n channelling effects, calms and stagnation events and,

horizontal and vertical wind shear effects are all complicated 3-dimensional features that requi re sophisticated meteorological models in order to realistically simulate these events. These phenomena are significant everyday occurrences th at affect all source types from ground level based wastewater treatment plants to odours being produced from pulp and paper mill factories, which traditionally have tall stacks.

Coastal fumigation is a potential concern for facil ities emitting odours from stacks close t o the coastl ine. Figure 3 graphically demonstrates coastal fumigation, where a pl ume released near a coastline produces a narrow plume in

water NOVEMBER 2010 6 7

odour management Table 1. Comparison between steady state plume models and a non-steady state model.

Feature

AUSPLUME, AERMOD, ISCST3, CTDM

CALPUFF

Causality effects considered?

No - plumes extend to infinity

Yes

Spatial variability of surface characteristics

AERM0D - Land use variability allowed in wind sectors centred at met. Station Others - No.

Full variability

Policy on treating surface variability

AERM0D looks upwind of met. station in a radius of 1-km to characterise turbulence up to 50-km downwind of source

CALPUFF determines surface conditions in each grid cell and applies it to a puff as it passes that cell.

Downwind conditions of each and every source are evaluated on a puff-by-puff basis

Horizontal wind variability

None. Single station wind is applied over entire modelling domain

Full variability considered when CALMET meteorological dataset used

Calm winds

Not treated - removed from the analysis

Calm winds treated

Mass accumulation during stagnation. Memory?

Not treated. Models have no memory of pollutants emitted during previous hours

CALPUFF retains previous hours emissions within domain and evaluates impacts from them

Coastal effects, fumigation

No coastal TIBL or fumigation algorithm

0verwater turbulence module and TIBL algorithm to treat coastal fumigation

stable conditions over the coast. The plume q uickly becomes unstable and fumigates to t he ground when the plume intercepts the growing boundary layer during the morning hours . Inversion break up fumigation can occur readily for tall odorous sources in valleys w here the plume is initially released above the stable boundary layer and wi ll t ravel some distance without penetrating back through the inversion layer, until some instabil ity is set up, usually caused by heating of the ground.

The only way to capture these phenomena is to interface soph isticated

.,.,

12.,w

112.5W

3D numerical meteorological models to a fine resolution diagnost ic. Interfacing of gridded 3D wind fields from t raditional weat her-type models with a fine resolution diagnost ic meteorological model allows regional flows t o be captured with the added benefit of including observation stations.

The procedure of combining soph isticated numerical 3D gridded data into a diagnostic meteorological mod el perm its the prognostic model to be run with a sign ificant ly larger horizontal grid spacing and different vertical grid resolution th an that used in the

111:~

: :-

Model Applicability- Steady State vs Non Steady State

[1414.S ~ I I ~

..,.. ;

"" .,,. "'

.,,. "'

"'

"'

diagnostic model which can t hen be run at a much finer resolution (< 250m) incorporating fine scale terrain and Land Use data. This option allows certain features of the flow f ield such as the sea breeze circulation with return flow aloft, w hich may not b e captured in the s urface observational data, to be introduced into the diag nostic w ind field results. In many instances gridded 3D numeri cal data may be more useful than observations which on ly represent cond itions in their immediate vicinity, frequently suffer from missi ng loss of data and are limited to just the surface. Observations on their own are not able to capture the 3D signal in t he atmosphere. Precipitation, gridded c loud cover and detailed sea surface t emperatures are a significant advantage of using numerical meteorological data.

,..

UTM ÂŁ-.. (kt!IJ

"'

..,

...

"'

-----.----

.,,,---.---.---m m * * "'

"'

..,

~

N

Figure 4. Cumulative impacts and terrain channelling effects from the 1D simple plume models (left) to the more advanced 3D non-steady state model (right). Terrain channelled wind are shown on the rhs, v.s. a uniform wind field produced by AUSPLUME and AERMOD on the lhs. The red circles and blue diamond represent surface stations used in the 3D model v.s. 1D model respectively while the solid blue circle represents the upper air station.

68 NOVEMBER 2010 water

Two key factors that should be considered in evaluating whether to use a conventional stead y-state plume model such as A USPLUME or a more sophisticated approach are, whether the stead y-state assump tion is val id, and, whether the t echnical parameterisations in the plume model adequately treat t he situation to be modelled. The effects of buildings, terrai n features, coastal effects and other flow ob structions, as well as other factors such as source height and receptor distance f rom the source can all affect the flo w and it can be argued that conditions are never st eady state.

Table 1 shows a comparison between the most commo nly used regulatory models, the steady- state plume models vs. the non-steady state approach of CALPUFF. A major d rawback of the steady-state assumption of AUSPLUME and AERMOD

technical features

Leaders in the design, engineering, manufacturing and installation of GRP covers and ducting systems for the containment, control and management of odour in wastewater treatment plants

Composite design and engineering Project management Manufacturing Site installation Turnkey Service:

• 3D modelling and FEA analysis; • FEA ducting system analysis; • ducting structural supports; • flow control GRP dampers; • GRP special fittings; • odour control GRP roof covers; • GRP and aluminium trafficable covers; • inspection and access hatches; • GRP stacks and vent shafts

RPC Technologies Pty Limited ABN 29 100 903124 24 Powers Road, Seven Hills NSW 2147

Contact: Simon Karpeles

. . . . . . .. ... . . .

simon.karpeles@rpctechnologies.com

(+612) 9624 9800

www.rpctechnologies.com

r1~ RPC Technologies

odour management is that the models assume that the meteorological data is horizontally uniform. Usually, the winds are derived from a single point measurement from a nearby site, such as an airport which does not necessarily reflect the flow in the valleys. St eady-state models do not adjust the winds to reflect the terrain effects and the net effect is that the steady-stat e flow field does not reproduce the terrain-induced spatial variability in the wind fields. In addition, the straight-line trajectory assumption of the plume model is unable to handle the curved flow associated with terraininduced deflection of chan nelling.

Figure 4 shows the resu lts of 3 individual sources and their related plumes from steady state models, AUSPLUME, AERMOD and ISC vs. that from CALPUFF. In a com plex terrain simulation, the plumes from the simple plume model blow directly across the valley, regardless of terrain. In this scenario the plume models produce no cumulative impacts whereas the nonsteady state model clearly shows overlapping plumes as the plumes are each deflected by the terrain and are channelled through the valley. The winds at 1Om are shown in the lower figure.

Calm and Stagnant Conditions Most wastewater treatment plants t end to be located at low points in the community and are often located close to water. These facilities lend themselves to the most adverse dispersion conditions which occur typically at night due to increased stability associated with very light wind, low turbulence and persistent radiation inversions which act to restrict the vertical dispersion of odours released near the ground.

Stagnation conditions are usually associated with rapid nighttime cooling of air near the val ley surface. Increased temperatu res above the nocturnally cooled surface create the inversion that prohibits mixing. Odours emitted into this st able nocturnal environment will either accumulate if calm conditions prevail or else wil l flow downwind with the drainage flow. Build-up or retention of odours over several hours is common and will only disperse with the onset of increased turbulence usually at sunrise. Steady state plume models such as AUSPLUME or AERMOD are not able to simulate the stagnation of odours. For one the models have no prior knowledge of the previous hours' meteorological or emission conditions as each hour is treated independent of the next. Also, the models have no causality effects so

70 NOVEMBER 2010 water

.,

.... l

..01,

~~

..:Âť

o0'2 5

<IO:U

~

U'l'"')(c:.,,dli,u(luoiw:::.M,l

30

1

I"

10

..,

S.O

,.,

"

10

20

"

so

.,

Figure 5. Comparison of ground level concentration contour plot using hourly meteorological data and computed turbulence coefficients (left), vs. sub-hourly meteorological data and real measured turbulence parameters (right). material is carried instantaneously and without changing to the receptors many miles away. Further the inverse wind speed dependence of the steady-state plume equation causes plume models to break down during low wi nd speed or calm conditions. Steady stat e models either set the concentration t o zero for the hour of calm condition , or, it forces the wind to a minimum speed, usually 1 m/s or greater, which means that the plume will travel to infinity even within the first hour. A more realistic approach of dealing with calms is to use a model that treats the transport and dispersion separately. This means in a true cal m condition the mean position of the plume material can stay unchanged even though the puff experiences growth due to dispersion.

There are two ways to improve a model 's behaviour in calm conditions, the first is to use sub-hourly meteorological data and the second is to use true measured turbulence parameters, sigma y and sigma z. Figure 5 shows the difference in the concentration contours around a single st ack depending on the meteorological data used. The plots to the left use 1 hour meteorology, model defaults and computed dispersion parameters. The figure to the right uses 10 minute meteorology and real measured turbulence parameters. The sub hourly fluctuations in the wind field are clearly evident in the concentration contour plot.

Technical Considerations computing < 1 hour ground level concentrations, using 1 hour meteorology Because odour fluctuations occur on the order of seconds the peak 1-hour average concentration delivered by ISC3, AERMOD, AUSPLUME and CALPUFF when using 1 hour met eorology is not an accurate representation of odour behaviour which is characterised by intermittent short-period concentrations.

There are a couple of ways t o account for the short-term peaks. AUSPLUME adjusts the 60-minute horizontal dispersion parameter, a y using the following unrobust power law equation; a y = (averaging time/60) 0 -2 -. This same option was included in CALPUFF in case a user wishes to run the model in a manner consistent with Victoria EPA. This approach is not recommended as the ay adjustment occurs for all hourly meteorological data and at all receptors regard less. The consensus now is to scale the predict ed concentrations by multiplying by a correction factor called the "peak-to-mean ratio" which is the peak concentratio n divided by the mean concentration (assumed constant over an hour). Peak-to-mean ratios were developed to estimate the difference between the mean and the peak fluctuation for various source-receptor configurations and they differ according to source type, near or far-field and

technical features

odour management stability class. Another method of estimating the short-term peak concentration is t o apply a power law equation to each and every receptor for each hour of the day:

CP = Cm(tpltmJ- P Here Cm is the measure of mean concentration over a 1-hour averaging time scale tm for which meteorological conditions are persistent, wh ile Cp is the expected peak concentration , averaged over time tp, (5 minutes) during the time tm. The choice of a value for the exponent "p" depends on source t ype and height, distance downwind and terrai n. Peak concentrations Cp tend to occur at night in patches of poorly diluted odorous source material sensed by a 'nose' downwind of the source. The value of tP can vary from molecular diffusion timescales (sub-seconds) to seconds and minutes; However, none of these schemes is ideal and there are many unresolved questions, such as, whether the plume is wake affected each hour or not. Further the currently preferred method of using peak-to-mean rat ios is complicated with unrealistic concentration step changes between near field and far field. The high

peak-to-mean ratio for plumes that are not wake affected also means that increasing the stack height may lead to higher concentrations being predicted even though dispersion is improved. The effect of this would be to encourage odours to be released from short stacks into the building wake rather than from taller stacks. Other complications occur when considering multiple source types, differing stack heights, complex terrain and coastal effects are considered.

Conclusion In this paper we have highlighted the important role of dispersion models in the control and regulation of odours.

Dispersion models bring a degree of objectivity to a subjective process and if interpreted, analysed and used correctly can give very realistic answers. The quality of the output of the dispersion model depends not so much on accurate odour emission rates but the model's ability to interpret and use the very best available meteorological data. Advanced meteorological models can provide dynamically consistent three dimensional flows which when coupled to a diagnostic meteorological with fine scale terrai n and landuse effects as well as

observations, can accurately simulate odour behaviour. Some of the limitations of applying steady-state Gaussian plume models such as AUSPLUME and AERMOD to odour applications and examples in both complex terrain , c oastal and calm conditions have been discussed. The difficulties in converting one hour odour concentrations to 1second concentrations for compliance were highlighted with the inclusion of sub hourly meteorological dat a offered as the most accurate solution.

The Author

Jenny Barclay, based in New Zealand, is a Senior Scientist with TRC, a US based engineering consultancy, in their Atmospheric Studies Group (ASG). She has over 16 year's experience in applied meteorology and atmospheric sciences.

The ASG are the developers of numerous dispersion models including PRIME, CALGRID, BLP and CALPUFF. Email JBarclay@trcsolutions.com.

When Tyco Water w as awarded the contract to manufacture all of over 5,000 SINTAKOTE9 steel pipes required for the Sugarloaf project, we dedicated one of our people full time onto the site. Just one of the ways Tyco Water works w ith customers to promote successful project s.

wawv

-tqca

your pipeline partner

GJ

odour management

refereed paper

THE WERF BIOSOLIDS ODOUR-REDUCTION ROADMAP J Cesca, R Forbes, C Martinson, H Bustamante, T Nguyen, W Wang Abstract Over the past several years, the Water Environment Research Federation (WERF) has sponsored a 4-phase research project tackling biosolids odours issues.

The resultant Biosolids OdourReduction Roadmap encompasses solutions beyond use of odour scrubbing or masking agents, as it incorporates a cradle-to-grave biosolids odour management approach - from a treatment facility's headworks to the biosolids cake storage, disposal, or beneficial use site.

The Biosolids Odour-Reduction Roadmap has been applied to a realworld biosolids odour investigation at the North Head Sewage Treatment Plant in Sydney. The Roadmap was used to identify potential short-, mid-, and longterm approaches to reduce the potential for biosolids odour at the plant. These approaches include operational improvements and recommendations for futu re design modifications that will result in the reduction of biosolids odours, along with related benefits such as improved volatile solids red uction and increased digester gas production.

Introduction and Background Over the past decade, the Water Environment Research Federation (WERF) has sponsored multi-phased research project tackling the biosolids odours issues. The focus of the research has been on identification and red uction of odours from anaerobically digested and dewatered biosolids, since those unit processes represent the large majority of biosolids generated and processed at North American WWTPs. The project has been underway since 1999 when Phase 1 began. Three phases have now been completed and a fourth phase is underway, which is investigating the relationship between odours in biosolids and the sudden increase or regrowth of pathogen indicators.

Phase 1 of the project included an extensive literature search of the current state of knowledge regarding odour

72 NOVEMBER 2010 water

characterisations, odour measurements, and odour control for wastewater treatment and collection . Findings of the Phase 1 Research were published by WERF in 2003 under the title, Identifying and Controlling Odour in the Municipal Wastewater Environment Phase I: Literature Search and Review.

At a WERF subscribers' meeting during Phase 1 of the research, odours from biosolids were voted as the largest odour problem facing the wastewater industry by the meeting participants, which led to Phase 2 of the research on odours following anaerobic digestion and dewatering of biosolids. Phase 2 of the research was based on sampling and collection of field data from wastewater treatment plants (WWTPs) across North America, focusing on how sludge characteristics and existing treatment processes affect odour emissions from biosolids. Many process parameters across the liquids and solids treatment processes were documented and evaluated to determine effects they had on biosolids odour emissions.

Phase 2 of the WERF biosolids odour research identified volatile organic su lfur compounds (VOSCs) as the primary constituents of cake odours and bioavailable protein as the primary "food source" of odour-causing bacteria in anaerobically digested biosolids. Phase 2 find ings also showed that high-solids centrifuge dewatering tend to produce higher biosolids cake odours than other dewatering methods, while dosing of iron or aluminum to digested or dewatered biosolids can help bind bioavailable protein, reducing the cake odour production.

Findings of the Phase 2 Research were published by WERF in 2003 under the title, Identifying and Controlling Odour in the Municipal Wastewater Environment Phase II: Impacts of In-Plant Parameters on Biosolids Odour Quality.

First full-scale application at North Head, Sydney.

Phase 3 of the WERF biosolids odour research focused on proving or disproving hypotheses related to the findings of previous phases. It included laboratory-scale and full-scale evaluations of cations in sludge, advanced anaerobic digestion , postdigestion chemical addition, and specific dewatering parameters (such as centrifuge bowl speed and torque) on biosolids odours.

Phase 3 research findings identified the effects of anaerobic digestion and enhancements to anaerobic digestion processes such as phased digestion, sludge homogenisation, or acid phase digestion; the effects of dewatering practices such as centrifuge, rotary press and belt filter press dewatering using various types of polymers, and the effects of chemical addition on biosolids cake odours. Results of the Phase 3 Research were published by WERF in 2008 under the title, Identifying and Controlling Odour in the Municipal Wastewater Environment Phase Ill: Biosolids Processing Modifications for Cake Odour Reduction.

WERF research on biosolids odours subsequent to Phase 3 has also been conducted, primarily on a laboratory scale, led by the civil and environment al engineering departments of Bucknell University and Virginia Polytech nic Institute and State University (Virginia Tech). This phase of the research was cond ucted largely to quantify the effects of aluminum and iron when present in wastewater sludges, added prior to digestion, or added as part of the biosolids dewatering process. The WERF final reports resu lting from those studies are: 1. WERF Phase /Va: Evaluation of Aluminum and Iron Addition during Conditioning and Dewatering for Odour Control, by Matthew J. Higgins, Ph.D., Principal Investigator.

2. WERF Phase /Vb: Effect of Aluminum and Iron on Odours, Digestion Efficiency, and Oewatering Properties, by John T. Novak, Ph.D., Principal Investigator.

In view of the large amount of research cond ucted and findi ngs published on the

technical features

odour management

refereed paper

identification and red uction of odours from biosolids over the past decade, WERF separately authorised CH2M HILL to develop a web-based tool for practitioners (engi neers, operators, and treatment managers) to use to address t heir own specific odour issues related to biosolids processing and management. The remainder of t his paper discusses t he features of t he tool and its potential applications.

The objectives of this paper are threefold : • Describe the components behi nd t he odour-reduction roadmap for biosolids odours, and • Identify factors in various unit processes that may contribute to odours in biosolids from North Head sewage treatment, • Identify possible short, medium and longer term options to minimise biosolids odours using the odour reduction roadmap .

Essential Features of the Biosolids Odour-Reduction Roadmap The WERF Biosolids Odours Reduction Roadmap is a web-based tool for WERF subscribers. The Roadmap tool started with a process-based schematic diagram of potential measures for identifying and reducing odours from biosolids at different points in a typical wastewater t reatment p lant. The schematic roadmap for biosolids odour reduction was used as a quick reference to identify the probable causes of odou rs from biosolids and then used to recommend measures for reducing t he odou rs. The schematic diagram is shown in Figure 1.

WERF su bscribers expressed a desire t o be able to drill into the schematic diagram with a decision-based tool that could guide t hem to specific unit processes and the specific approaches for reducing odours associated with those processes. The resu lt of this decision-based approach is a webbased tool t hat has been developed based on t he collection of questions and pot ential responses from the decision tree for reducing biosolids odours. The basic decision tree is shown in Figure 2.

The tool can be navigated by providing "yes/ no" answers to the appropriate series of questions shown in Figure 2 above. Once the user of the web-based tool has answered enough questions to reach the bottom of t he

--

I

Potential Reduction Measures for Biosolids Cake Odors*

I

I

I

l

I Process I Equipment I

I

~

.. ..

Anaerobic Digestion Digester pretreatment Improve mixing Increase VS reduction O perate in series Multiple phases Increase temperature

..

I

I

Operational Changes

Design Changes

I

I

.. ..

Pre-Di gestion Aluminum Iron Other cations pH control

.. .. .. .

Anaerobic Digestion Digester pretreatment Better mixing Longer SRTs Egg-shaped Acid-Gas (multi-phase) Thermophilic Temperature-phased

r

l

I Dewai ering & ~ ak!l Convei1ance/ Storage Decrease centrifuge bowl speed Decrease centrifug e torque Increase cake moisture content Consider BFPs Belt Conveyors Cake storage/aging

. . . .. .

1

I Chemical Addi tion

~

Dewatering & Cake Storage Belt Filter Press Rotary Press Variable-speed centrifuges Adjust cake solids Minimize conveyance Belt conveyors Enclose cake storage areas

.. .

Pre - or PostDewatering Aluminum Iron Other ca tions pH con trol Polymer dose and type

.. .. .

.

. ..

* Note: Non e of these options sh ould b e considered ind ependently of the o thers, and odor reducti ons in one area may Impact treatment pro cesses and odors in o ther areas. Therefore, an integrate d and c ustomiz ed appro ach is required for each WWTP.

Figure 1. Schematic Diagram of Biosolids Odour-Reduction Roadmap.

•Ci;Jesiia1µ&tre;llret • Ae- a pat-<twmrg Fe'pj a:l:itm - ~cbeal'.1!'}1:e a;jWrnrt •Qjinilllcar.esdcl; catert

•Ci;Jesiia1P'fflEl1Tei

· Ci;Jesiia1~

·npo,e~ni*!J

•ln']roiedglSi!'qlraicn · MJ~~q:e;aicn • MmizectYmirgshls'

· R3:ire~kmrg • Oo'Eeleartrtl.g!spmlaq..e • Onid9' EfFs a rday fill$ • Ae- a pat-<twmrg Fe'pj

• Qjinill!car.esdcl; a:rmt • Onid9'ecmrg slmgetire

• Ae-a~Fe'pj attic,)

• Qmill! ai<e s:jj, catert • Casi:13' ecmrg slmgetire

• Q:iirile chm:al axitia, •lnpo.ecigs8'q:aai01 • MJti.µ'eledgs8'qlraicn • A'o.«JeJj-lcatrd • Mriri2e ci!,>,Eisi'g slur • Q:iirile car.e s:j(k ami •Omar eden:lrgsblg!tirre

Figure 2. Decision Tree for Reducing Biosolids Odours.

water NOVEMBER 2010 73

odour management flowchart shown in Figure 2, a set of briefly described potential solutions wi ll be presented in the tool, defined as " Reports 1 through 9," along with options to view research publications further describing those solutions and available findi ngs from WERF research. The locations of Reports 1-9 in the tool are denoted by the cells in Table 1.

The primary goals of the Biosolids Odour Reduction Roadmap tool were established as follows: 1. Make the tool a web-based platform for posting by WERF 2. Provide a layered structure for the tool to allow user navigate between the: a. Decision Tree (Roadmap) consisting of a decision matrix guidance for odour management options based on a Q&A input by the user b. Brief Descriptions of Potential Biosolids Odour Reduction Solutions pertaining to each subject area in the flowchart, potentially allowing space for user comments on experience with solutions c. WEF/WERF publications Presenting More Detail on the Biosolids Odour Reduction Solutions 3. Provide key word links to the WERF Biosolids Odours project report sections (i.e. Phases 1 through 4 WERF reports) 4. Provide key word links to a published literature database 5. In a future phase, potentially provide a Bulletin Board to allow WERF subscribers to share their biosolids odour reduction experience and solutions.

The Roadmap tool is structured to provide access to all of the content generated by the WERF research teams as well as to supplementary material available to WERF subscribers. The structure is interactive to make the content readi ly accessible to users of varying backgrounds and expertise, as described in the following subsections.

Internet Platform The Roadmap tool has been constructed as a WERF website, and is currently being beta-tested by the WERF Project Subcommittee. The current plan is for the tool to reside behind the currently available subscriber log-in access security protocol. This will provide the greatest flexibi lity in terms of maintenance, interlinking resources, and user access. Several levels of password protection may be used: non-protected access for the general public to certain

74 NOVEMBER 201 0

water

~

refereed paper

Table 1. User and Response Report Matrix.

Uw RllpOnse and Report Matrix b lie oooraricem at a plmed onr1

exstrg-fd/1 11Sflepl00Ullpo,wn11inrfalhll1al benefcii use sle?

Nolt.ere "!fodalcilemi:ameeswnen iesuholla!a mlrlJe? Askfordata(H2S,m ~. DMSI mead As, .,.u,, oaia. l9COld and roo;e krnext

Reportl

Report2

Report3

Report4

Repono

1<epo1t•

1<epo1t1

1<epon,

Kepolt~

N

y

y

y

y

y

y

y

y

y

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

y

y

y

N

N

N

y

y

y

N

N

N

N

y

y

N

y

y

N

N

y

N

N

y

N

y

y

N

y

y

a lschemi:aacl:iirlk>lhecolecti)1s)Stem

Jmi:ed?

Is chemi:aacl:iirl at lhe pin hea:t,,al<s, p;na.yc.secon<lryb9'imen Iris. at lhe li},stels, alerci;jestm Jmi:ed?

Ask for SRT d lhe secon<lry beelnuri, mcadaidmC¥ek>lhenextQ

Ask forci;jesttrfeed m(l;m d PSWAS, %VSl."""'1aidmC¥eklhnextO Anaeoobi: ci;jeslilo pe,fmmro, is ~po:Nirg aVSR>50%?

Do lhe d!\oa~ and c.i<e caM!'JIID! IAl!e '!V1 slwlecJn:tgies? lsc:11<lec.i<eslorageavaW!?

tool features, full access to all Roadmap tool features accessible by a WERF member password, and secure access by an administrator for updating site content. The administrator may be a WERF staff member or contractor retained by WERF. The potential for the tool to be maintained under subsequent phases of WERF research is also being considered.

Maintenance A basic plan for updating the Roadmap tool content has been established so it can remain relevant as new information is developed. WERF project SRSK4T08, Wastewater Treatment Plant Design and Operation Modifications to Improve Management of Biosolids Odours and Sudden Increases in Indicator Organisms, which is in progress, is expected to identify additional causes of biosolids odours and potential control measures.

The tool will be updated with information from this project after it becomes available.

Updating the Roadmap tool content is planned to help the team and WERF staff ensure relevance and quality of new content to be added. It is anticipated that a WERF staff member or contractor would be the administrator of the tool and responsible for minor updates and repairs. Major updates that can impact the decision logic of the tool and/ or require screening of tech nical information pertinent to biosolids odours management wil l be implemented in the future if needs dictate and funding is available.

Functionality Upon entering the Roadmap tool website, the user selects from a number of features, including a decision matrix, biosolids odours management guidance, study reports, and the literature database. Some features are structured to serve as a look-up reference; others are designed more as guidance to identify potential solutions for biosolids odour problems. Each feature is searchable by keyword and linked internally to other tool features.

Essentially, the Roadmap tool is designed with two functional objectives: first, to quickly connect the user seeking specific information with his target information; second, to present information orderly to the user seeking guidance or general information on a subject.

Content Tool content comprises problem-solving and guidance aids and relevant resources and information. Three main features are being built into the tool , with a potential fourth feature to be added in the future, as described below.

1. Study Reports: The tool contains links to the full WERF Phase I, II Ill , and IV reports on biosolids odour identification and reduction. Each chapt er from the report is linked to the table of contents , and a bibliography will be linked to a literature database. Each chapter of the study reports is intended to be provided to WERF subscribers as a dow nloadable PDF file. The tool wil l reside in a password protected area on the WERF server/website, and would only allow the

technical features

u;J

odour management

refereed paper

WERF subscribers access to the tool and the Phases I-IV reports.

2. Literature Database: The literature database feature would contain an indexing function for searching and sorting each field of bibliographic data.

Additionally, paper abstracts and WERF team paper reviews would be keyword searchable.

3. Decision Matrix: A key feature of the Roadmap tool would be a decision tree matrix structured to provide guidance for design, problem solving , and technology selection. Th is feature uses a "berry finder"-type logical structure to help users focus on approaches and issues pertinent to their specific problem. The decision matrix does not provide definitive answers or detailed design findings; rather, it directs users to the appropriate methods and technologies that should be considered .

The decision matrix also flags points in the decision-making process where additional data might be needed. The decision matrix also flags common pitfalls corresponding to the subject being queried and cautions against counterproductive actions. Each level of the matrix is linked to more-detailed information in the WERF reports and other tool resources.

4. User Forum/ Bulletin Board (Future Phase): A "discussion/ user forum" is being considered as a future component of this roadmap tool. This wou ld enable users to ask questions of other users, seek clarification on methods and information presented in the tool , and alert other users to new research and resources. The tool administrator on WERF staff could be in charge of screening the postings in accordance with the WERF communications guidelines. The content and information from the postings can be used to update content or add to the literature database as appropriate in a later phase of th is project. Depending on available WER F resources, this feature may be provided under a future upgrade of the tool.

Case Study Using the Biosolids Odour Roadmap Schematic North Head Sewage Treatment Plant (STP), situated on the North Head Peninsula of Sydney Harbour, is one of the largest STPs owned and operated

by Sydney Water Corporation (SWC).

The STP provides primary treatment to an average dry weather wastewater flow of 300 MUd. Recently, sludge digestion facilities including sludge thickening,

anaerobic digestion, dewatering, and storage were constructed as part of a plant upgrade. Digested and dewatered cake is stored on-site in hoppers prior to transported off-site for beneficial use.

provides greater flexibility to deal with process upset cond itions or peak load conditions without compromising the stability of the biosolids and thus not risking having higher odour levels.

The sludge digestion project was implemented to substantially reduce truck movements by digesting the sludge in order to reduce the volume of biosolids and to generate methane for on site energy use. The project has been largely successful and truck movements have been reduced by around 70% and the digester gas has provided subst antial energy and carbon offsets via the new power cogeneration system.

3. The use of high speed (high shear) centrifuges for dewatering:

Al l biosolids from North Head STP meet the grade B stabilisation standards and are beneficially used in agricultural and horticultural applications. The biosolids are transported in covered trucks to a facility in south western Sydney for co-composting with green waste.

The WERF study showed bio-available protein (SAP) to be a primary food source for RBA that leads to biosolids odours. Thus, any process or unit operation that facil itates protein availability can be a potential cause for odours. North Head STP operates high speed , high G centrifuges for the dewatering of biosolids to increase the solids content in the dewatered cake thus reducing the volume req uired for truck transport and disposal. North Head STP also uses screw conveyors to transport dewatered biosolids. The WERF study showed that high speed operation of the dewatering centrifuges increased odour potential.

4. Long and under-sized screw conveyors after dewatering.

The roadmap was used to identify contributing factors to North Head STP biosolids odours that could be targeted for further improvement.

This process also increases the shear imparted on bacterial cel ls and can free up protein making it bioavailable.

Main Contributing Factors to Odours from Biosolids

5. Dewatered biosolids storage of more than 2 days but less than 15 days.

The following factors were identified: 1. A relatively high amount of volatile solids remain ing in the digest ed biosolids: The average volatile solids destruction (VSD) rate of 65% at North Head STP is typical for anaerobic digestion process w ith designed solids retention time of 20 days. However, the average residual volatile solids, VSb, in digested sludge is relatively high at 72%. VSb can be seen as an indicator of the amount of residual biological activity (RBA) remaining in biosolids after digestion and research has shown that this value is highly correlated with odours. A real istic target VSb of 65% should be achievable at North Head STP. 2. A digestion solids retention time of less than 20 days: Although the design SRT of 20 days provides adequate VSD to produce Class B stabilised biosolids, longer SRTs or phased digestion can provide higher VSD and result in lower VSb and thus lower odour in digested solids.

Research indicated that biosolids cake odour could decrease substantially if SRT was increased to a range of 25 to 30 days. Having the ability to operate at an SRT of longer than 20 days also

Short, Medium, and Long Term Options Identified using the Roadmap Schematic A staged implementation (short, medium, and long term) was recommended for reducing the biosolids odours. A review is recommended upon implementation of each stage to assess the performance and improvement, and the need for further action.

Short-term options: The short-term options recommended targeted operational improvements to the existing biosolids treatment processes wh ich cou ld reduce the potential for odourous biosolids. They are expected to require minimal capital costs and could be implemented with relative ease. These include improving the thickeni ng of digester feed , decreasing the duration of dewatered cake storage, reducing the centrifuge bowl speed, and optimising polymer dose and type.

The results of the implementation of these options are as follows: â&#x20AC;˘ Improve sludge digester feed: The high VSb at North Head STP was due to the unusually high volati le solids in North Head STP raw sludge and the

water NOVEMBER 2010

75

odour management teethi ng problems during the commissioning of the sludge digester feed and overall digestion process. Since then, the process has been optimised and is more stable resulting in t he reduction of VSb to between 64% - 72%.

• Decrease the durat ion of dewatered cake storage: The storage duration of the biosolids in normal o peration is one day or less. However, due to regulatory requirements not to transport biosol ids during weekends, the storage duration is two or t hree days during these periods.

• Reduce the centrifuge speed: A trial was carried out at North Head STP to assess t he impact of centrifuge speed (range 2500 rpm t o 3000 rpm) on biosolids odours. It was found that w hile the solids content in dewatered cake was not affected, lower centrifuge speeds resulted in higher concentration of hyd rogen sulphide, ammonia and amine compounds emission from biosolids.

Therefore th is o ption was not imp lemented.

• Optimise polymer dose: A trial was carried out at North Head STP to assess the impact of polymer dosage (range 7 to 8.5 kg/d ry t on solids) on biosolids odours. It was found that biosolids odours were not affected but centrate concentration reduced with higher polymer dosage.

Medium-term options: The medium-term options target improvements to the reliab ility of t he biosolids treatment processes w hich could resu lt in a lower frequency and reduced potential of odourous biosolids associated with process variability. These include improvements to the exist ing digester feed thickening process, recuperative thickeni ng, operating all t hree digesters in series, and replacing the cake conveyance system with belt conveyors .

The results of consideration of these options are as foll ows: • Digester feed thickening process: An optimisation project is proposed to improve t he reliabil ity and performance of t he Rotary Drum Thickeners of the digester feed.

• Recuperative thickening digestion: This process aimed to increase the d igester sol ids retention time by increasing the solids content in the

76 NOVEMBER 2010 water

digesters. A similar project was implemented at Bondi STP with good results. The implementation of t his process at North Head is in the planning phase.

• Replace the cake conveying system with belt conveyors: A belt conveyor system is not possible due to space constraints. However, the existing conveyer system is being replaced with a higher capacity system to improve reliabi lity.

A monitoring program will be implemented to assess the impact of these improvements on biosolids odours.

Long-term options: The long-term options target improvements to t he long term efficiency, reliability, and redu ndancy of the sludge digestion process. These would provide the most substantial reduction in the frequency and potential of odourous biosolids. These include bui lding an add itional mesophilic digester, and/or converting to temperature-phased or two- phased (acid-gas) anaerobic digestion.

These options are not considered at this stage.

Conclusions The Biosolids Odour-Red uction Roadmap is a useful tool to identify the contributing factors to be targeted for improvement. The results of t he application of this tool to North Head STP wi ll provide useful real life data for the WERF team to further improve the too l.

The road map was successful in identifying possible sources of odours at North head wastewater t reatment plant. This was also achieved by conducting a focused monitoring of TSVOC at various stages of the plant.

This resulted in the design of range of short, medium and long terms measures than can be pot entially applied in stages depending on t he level of odour reduction needed

The Authors

Josef Cesca is the ANZ Reg ional Techno logy Manager with CH2M HILL,

GJ

refereed paper

Sydney specialising in biosol ids treatment and odour control (email Josef.cesca@ch2m.com.au) and Carol Martinson is a process engineer and the assistant regional technology leader for biosol ids with CH2M HILL Sydney.

Robert Forbes is a principal member of CH2M Hil l's Global Residuals Technical Group with CH2M Hill, Carolina. From 2002 t hroug h the present, he has led a large and diverse technical t eam on research under t he Water Environment Research Foundation (WERF) to investigate the causes of odours from biosolids and develop odour- reduction met hods, producing two major WERF reports and a number of techn ical papers in the process.

Heriberto Bustamante (HERI.BUSTAMANTE@Sydneywater. com.au) is Project Manager, Science & Technology (Sydney Water).

Tung Nguyen is Technical Services Manager and Wanxin W ang is Production Officer, both with Sydney Water .

References Adams, G.M., J.R. Witherspoon, Z.K Erdal, R.H. Forbes, R.H. , J.R. Hargreaves, M.J. Higgins, D.W. McEwen and J.T. Novak (2007). Ident ifying and Controlling the Municipal Wastewater Odour Environment Phase 3: Biosolids Processing Modifi cations for Cake Odour Reduction.

Water Environment Research Foundation Report No.03-CTS-9T, Alexandria, VA. Adams, G.M., J. Witherspoon, T. Card , Z. Erdal, R. Forbes, J. Geselbracht, D Glindemann, R. Hargreaves, L. Hentz, M.J. Higgins, D. McEwen , and S. Murthy (2004).

Ident ifying and Controlling Odour in the Municipal Wastewater Environment Phase 2: Impacts of In-Plant Operational Parameters on Biosolids Odour Quality.

Water Environment Research Foundation Report No. 00HHEST, Alexandria, VA. Higgins, M.J. (2010). WERF Phase IVa: Evaluation of Aluminum and Iron Addition During Conditioning and Dewatering for Odour Control. Water Environment Research Foundation, Report No .03-CTS9T, Alexandria, VA. Novak, J.T., and Park C.M. (2010). WERF Phase IVb: Effect of Aluminum and Iron on Odours, Digestion Efficiency and Dewatering Properties. Water Environment Research Foundation , Report No.03-CTS9T, Alexandria, VA.

technical features

~

odour management

refereed paper

SEWER ODOUR ABATEMENT PRACTICES - AN AUSTRALIAN SURVEY E Sivret, R Stuetz Abstract A survey of nine Australian wastewater utilities was undertaken t o assess current sewer network odour abatement practices. The survey results indicat ed that activated carbon based systems are the most common fol lowed by biofilters, and that most processes are treating passive emissions from the sewers. Most abatement processes are not being routinely maintained and are monitored solely through community complaints. Activated carbon media is generally replaced in a reactive manner. H2S is the dominant on line and offline monitoring parameter with limited use of non-H2S odorant analysis.

Th ere were sign ificant limitat ions in the industry's ability to provide fundamental process data.

a)

Activated Alumina Filler

3.2 %

Activated Carbon Fi ller 44.9 % Zeoli le Filler

51 .9 %

b)

4.7% CJ Biofiltcr

networks, representing a major portion of the Australian wastewater industry.

Understanding the existing industry practices with regards to odour assessment and abatement wi ll enable the identification of key areas of potential improvement to provide a focus for ongoing research effort within the SCORe subproject 3, Odour measurement and assessment, evaluation of odour treatment technologies.

J0. 2 % D Biofilter + Activated

The survey was composed of two sub-surveys: one on odour GJ Soil Bed Filter abatement processes employed 27.9% by indust ry partners and the â&#x20AC;˘ Biotrickling Filter + second on odour monitoring Activated Carbon practices used to assess â&#x20AC;˘ Biotrickling Filter + Soil Bed Filter abatement performance, and D Bioscrubber + was conducted between June 18.6% Biofilter and October 2009. Data from the surveys were collected by Figure 1. Odour abatement process types: a) adsorption; the UNSW Wat er Research b) biological.

Introduction Centre's Atmospheric Emissions and Odour Laboratory and Th e importance of managing and monitoring of sewer odour merged into a database with the odorous emissions from sewers has abatement processes, a survey of nine terminology used by the respondents become more significant in the past Australian wastewater utilities (Barwon being standardised to facil itate analysis.

10-20 years as popu lations have grown, Regional Wat er Corporation, Gold Coast Analysis focused on extracting trends sewage concentrations and temperatures Water, Hunter Water Corporation, from the data, although the depth of the have increased, buffer zones have been Melbou rne Water Corporation , South analysis was somet imes limited by gaps eroded by development, and complaints Australian Water Corporation , South East within t he dataset. It should be noted that have been exacerbated by the increasing Water Limited, Sydney Water these surveys on ly included odour sophistication of the populace.

Corporation, United Wat er International, abatement processes treating emissions and Water Corporation Western Aust ralia) Odorants that are present in the liquid from sewer networks, and not those was undertaken as part of t he Australian p hase of a sewer system are emitted into installed at wastewater treatment plants.

Research Council (ARC) Sewer Corrosion ambient air wherever there is a liquidand Odour Research (SCORe) linkage gaseous interface (Hvitved-Jacobsen, Odour Abatement Processes project.

2002). The odou r emission rate is mainly The first sub-survey indicated that 204 dependent on two factors: (i) physicalodour abatement processes were being Methodology chemical circumstances such as operated by t he indust ry partners at 187 Industry partners participating in the turbu lence, interface size and pH, and (ii) sites (some sites had multiple processes survey serve over 8.4 million people and the amount of odorants present in the operating in parallel). The odour operate over 59,000 km of sewer liquid phase due to microbial abatement processes (Figu re 1) were transformation (sulfate red uction and dominated by adsorpt ion based fermentation) of the wastewater into Identifying key areas of processes (76.5%), followed by biological hydrogen sulfide (H2S) and volatile processes (21.1 %). In general, there was improvement for the organic compounds (VOCs) during a similar distribution of odou r abatement sewage transport. processes across all of the industry design and operation To understand current industry partners, although some used more of processes. practices with regards to the operation biological processes t han others. The Carbon

water NOVEMBER 2010 77

odour management

~

large number of adsorption based processes is most likely due to the distributed nature of odour abatement units in sewer networks and the desire for these units to be low maintenance.

Forced

39.2%

Passive 58.8%

Figure 2. Odour abatement process feed supply type.

Generally, adsorption processes are supplied via passive venti lation, along with a few of the lower maintenance types of biological processes (biofilters), although some activated carbon filters are supplied by forced venti lation. More technologically complex and controlled biological processes (such as biotrickling filters and bioscrubbers) are supplied through forced ventilation. The greater use of passive ventilation for adsorption processes and biofi lters again most probably reflects their better suitability for application at small , distributed sites.

The air flow rate to be treated is a fundamental sizi ng parameter and is required to enable effective design of odour abatement processes,yet very little f low data was reported: 11 .6% of the passively ventilated sites, and 39% of the forced ventilation sites. The low availability of flow data for the forced ventilat ed sites was unexpected given fans are being used at these sites to drive the airflow and sizing data should be readily available. This lack of fundamental flow data suggests limitations in either data storage or data availability by the industry partners. The

Air supply and flowrate is a fundamental parameter for odour abatement process design and can have a significant influence on process selection and operation. A summary of process feed types is provided in Figure 2.

Most proc esses are treating a passive sewer air flow with a minority of the processes being supplied by forced (fan driven) flow. A comparison of feed type (forced or passive ventilation) versus process type is shown in Figure 3.

---------- -- -

110 100 90 ~

70

c,;: 0

50

0

.,, ~

ยง z

I

I I I

81

60

40 30 20 10 0

,_~ ------

!

80

u~

flow data for the 45 sites where it was reported is summarised are summarised in Figure 4.

No Response 2.0%

Adsorption based processes were dominated by two process types, activated carbon filters and zeolite filters.

Act ivated carbon filters were employed by most of the indust ry partners, except for one industry partner that only used zeol ite filters. For b iological based processes, biofilters were t he dominant process at 44.2% of total biolog ical processes, or 62.8% if soil bed filters are included in this number. This probably reflects the less active nature of these processes and the preference for smaller, contained processes which are more suitable for distribution to multiple sites and require less ongoing supervision and maintenance than more complex biological odour abatement processes such as biotrickling filters and bioscrubbers, which were 32.6% and 4.7% of biological processes, respectively. All of the reported biotrickling filters and b ioscrubbers were backed up by polishing processes (mostly activated carbon) to ensure treatment integrity.

~

I~ I

..,

i

I 1..1..

r,-+-

58

0=2=

i_

14

G

7

.r n--+--o -+D__i.+

5

,o

,.

I

,2..,

:~

1

D Passive

Cl No Response

Figure 3. Odour abatement process type by feed type (forced or passive ventilation).

78 NOVEMBER 2010 water

Types of Sewage Information was requested with regards to the dominant sewage type (domestic, industrial, or mixed). The surveyed odour abatement processes were operating on predominately (60.8%) mixed sewers (containing both domestic and industrial wastewater), with the remai nder of the processes operating on domestic sewers. Only one site was reported as being an industrial sewer.

On domestic sewers, activated carbon and b iological based processes were dominant, while only adsorption based processes were applied on mixed sewers. Given the lack of design information it is likely that the source sewage type is not being considered in odour abatement process selection and design. This is most likely a direct result of the current lack of understanding of specific odorants in sewer emissions.

Performance Criteria The use of performance criteria to set the basis for odour abatement process design and performance evaluation was very limited. Overall 75% of respondents did not respond, and of the remainder, only 5.5% specified some type of H2 S measurement or removal.

Additionally, the use of performance guarantees for odour abatement processes installed on sewer networks was limited (reported for only 4.9% of the processes). The overal l results once again demonstrated sign ificant gaps wit hin the industry record keeping and/or availabi lity of fundamental process data.

Only 5.4% of the odour abatement processes were designed using data supplied by the industry partners. An analysis of the data supplied for the activated carbon based processes is summarised in Figure 6. There appears to be a fairly even distribution of the types of data being provided (H 2 S, odour, dimethyl sulfide (OMS) and total volatile organic compound concentrations (TVOC)), however, it should be noted that design data were only provided for 10% of the activated carbon based processes, all of which were reported by two of the industry partners.

Maintenance

~ D Forccd

refereed paper

A good response rate (84.8%) was obtained with regards to maintenance information. However, routine

technical features

[]

odour management

r e f eree d paper

maintenance programs were reported for only 17% of the abatement processes (Figure 6), and mostly for the more complex and maintenance demanding process types (biotrickling filters and bioscrubbers). A "set and forget" approach was more commonly applied to many adsorption based processes, which are oft en only maintained in a reactive manner when there is evidence of a process failure (often by complaints).

data or any other performance data was provided by any of the respondents.

13.3%

Process Flow Rate

(rn

3

/ hl

D 125. 250

Online monitoring was reported for 20.1% of existing odour abatement processes, offline monitoring for 26% of processes, and 5.9% of the processes were completely unmonitored. 59.3% of the odour abatement processes were reported as being solely monitored through complaints from the surrounding community (Figure 9).

D 251 - 500 D 501 · 1 000 • 1 001 • 5000 C 5001 - 10000

D 10001 . 14400

15.6%

Figure 4. Comparison of odour abatement process size and process flow rate for 45 processes only.

The industry partners provided a significant amount of information with regards to media replacement frequency for activated carbon based processes (Figure 7). In general, many (40%) activated carbon based processes are operated in a purely reactive manner with media being replaced following evidence of odour breakthrough. A slightly larger number (47.1 %) are operat ed in a more proactive manner with media replacement frequencies rang ing from three months up to ten years. The industry partners also provided information on the identified sources of odour abatement failure events (Figure 8), with media saturation and moisture impacts being the dominant reported failure sources for adsorption based processes.

process control is purely focused on maintaining specific operating parameters (such as pH or water levels) at the desired settings.

Odour Monitoring Odour abatement process performance data was requested in the second subsurvey. No inlet/outlet H2 S monitoring

Further analysis of monitoring conducted on the adsorption based processes showed that these systems are predominately monitored using odour complaints. The use of online and offline process monitoring is common for biological based odour abatement processes, which is probably a resu lt of

1 00 90 80

!: ~

70

s

60

ct 0

50

" 0

F"a;

40

34.3

30

u

~

C.

20

Process Control

10

Application of process control to odour abatement processes was limited (3.4 % of processes), with process control only being applied for biotrickl ing filters and bioscrubbers. No data was provided with regards to the specific control methodologies applied, i.e. whether off gas H2 S data is being used as an odour surrogate and process operating parameters are being adjusted to maintain a desired level of removal, or instead (and most likely) if existing

0 ~

0

•

c.. ~

C

"'0

~ 8C

z

0

u

Figure 5. Data provided to designers for activated carbon odour abatement processes (70 processes).

50 -

40 - , - - - - - - - - - - - - - - - - - - - - - - - - - - ,,. .:.4:0::;.0~- - ~

~

16.7%

C

C." -=" :I:

~

Yes

~

C

._g

e

30

+ - - - - - - - - - - - - - - - - - - - - - - - ----1

C.

0

"

~ 20 + - - - - - - - - - - - - - - - - - - - - - - ---'

~

"

12.9

12.9

c..

10.0

10 + - - - - - - - --1 5.7

No 70.1 %

Figure 6. Odour abatement process routine maintenance program.

u

9

12

15

5.7

18

24

48

Media Replacement Frc<1uency (Monlhs)

120

Reaclive

No Response

Figure 7. Media replacement frequency for activated carbon odour abatement processes only.

water NOVEMBER 2010 79

odour management

refere ed paper

the relative complexity of some of the biological processes and the associated monitoring demands. Process monitoring for biofilters, on the ot her hand, was simi lar to adsorption based processes, primarily through complaints only.

While a range of online monitoring parameters was reported for odour abatement monitoring, H2 S was dominant, used for 87.8% of online monitored processes (Figure 10), and the on ly parameter directly relat ed to odour.

Industry partners indicated that the dominant use of the online data was for decision making and diagnosis purposes (70.7%), with 24.4% not indicating how t he information was being used, and online process monitoring data being leveraged for process control in on ly 14.6% of the processes. Of the 35 sites that reported using online H2S monitoring, 16 sites (45.7%) could not report t he inst rument manufacturer. Four of industry partners reported using a range of H2S instrumentation, with variabil ity in instrument selection both between and within the industry partners.

Unlike online monitoring, offline process monitoring was dominated by community c omplai nt information (68.8 %). While not an instrumentation based (or objective) monitoring approach, it was the primary reported process monitoring technique for two of t he indust ry partners. Of offline process monitoring conducted using instrumentation, H2 S was once again the dominant monitoring parameter, although t here was some indication of voe monitoring at a few sites. Overall there was very little response with regards to the use of offline monitoring data with only 29% of sites having an indicated

120 ~

Q.

~ 100 ~

2

90

.]

80

~

.

5

~

~ ~ ~

~

'o ~ §

z

70 60

54

50 40 30 20 10

11

0

Fans

Pumps

D Zeohte Filter 0 SOtl Bed Filter

Spra y No z.zles

a Bfotrickhng Filter

data use. The dominant uses of offline monitoring data was once again decision making and diagnosis. Typical ly, complaint data was also used in a decision making capacity to initiate maintenance such as to init iate a change in media for adsorption based odour abatement processes.

• omine _ % 12 3

0

n

-+D_, ID w _,

Figure 10. Online monitoring parameters.

80 NOVEMBER 2010 water

73

D No Response

Figure 9. Overview of odour abatement process

2

rQ.

I C:::::, l

Offline and Sensory

Cl None

~n -• n I

•

IJ Complainl s

-= -

No

Response 0 81ofUter

ul Onlinc

monitoring.

The survey also requested information on non-process odour/odorant monitoring conducted by the industry partners (in particular sampli ng and analysis methodologies and instrumentation). The industry partners reported very little assessment of non-H2 S odorants supporting t he conclusion that H2 S monitoring was the primary form of monitoring conducted for design and diagnostic purposes. Sampling was

19.5

2.:!...,

I No Failures

I

[] Online and Offline

I

I -I bz-11- I

W aler

Supply

D Acllvated Alumina Filter D Bbscrubber

.L.egro_d

4 1.5

10

Mois lure

Figure 8. Reported sources of odour abatement process failure.

,2::...

7;t-

Media Saluralion

O Actlvaled Carbon Filter

100 90

r

11 5

-- -

110

I

generally done from a single point (assumed to be representative of the general flow) ; although no significant sampling and analysis methodologies (or methodologies specified by external providers) were indicated, with only one industry partner reporting standard operating procedures for sample collection and analysis.

Industry partners reported the use of instruments from a single instrument manufacture for non-process H2S monitoring, and most indicated the instruments had a high level of rel iability when maintained in accordance wit h manufacturer specificat ions (including 6 monthly factory cal ibration). Given t he high level of use of t his instrument type (both for odour abatement process monitoring and non-process monitoring purposes) and that H2S monitors are known to be impacted by other su lfur compounds in terms of sensor cross sensitivities and interferences (Stuetz and Frechen, 2001), it could be beneficial to the industry to further evaluate H2 S instrument performance with regards to sensor interferences and cross sensitivities.

technical features

5]

odour management

re fe ree d p a per

Three industry partners also reported the use of external service providers to conduct testing (primarily odour and H2S sampling/analysis, w ith one partner also and red uced sulfur conducting compound analysis). No details were provided with regards to sample storage/ transport protocols (standard operating practices) used for odour or process monitoring and limited analytical cost data was provided by industry partners.

voe

Conclusion The surveys indicated several trends within the industry: (i) activat ed carbon based odour abat ement processes are well distributed across nearly all of the industry partners; (ii) a mixture of domestic and mixed sewage odour sources with minimal industrial only sites; (iii) most processes are treating a passive flow; (iv) most sites are maintained in some form , but most are not being routi nely maintained; (v) activated carbon media is generally replaced in a reactive manner following fail ure and the receipt of complaints; (vi) media saturation and moisture impacts are the dominant sources of odour abatement process failure, with other reported failures primarily being mechanical in nature; (vi i) many processes are being monitored solely through community complaints; (viii) H2S is the dominant online and offline odour abatement proc ess monitoring parameter, and the on ly monit oring parameter directly related to odours; (ix) significant variability in online H2 S monitoring instruments used across the industry; (x) low level of process control being used in the industry, but given the dominant types of processes this is to be expected; (xi) H2 S is the dominant non-process monitoring parameter and (xi i) limited reports of non-H 2S odorant analysis (VOCs and reduced sulfur compounds).

appear to be centrally stored in a readily accessible manner for use in decision making, planning, or abatement process design and selection . Enhancing the management and availability of this data within wast ewat er utilities would offer significant opportunities to improve the selection, planning and design of odour abatement processes.

Understanding the existing industry practices with regards to odour assessment and abatement will enable the identification of key areas of potential improvement for the design and operation of sewer odour abatement processes.

Acknowledgment This work was supported by the Australian Research Council Linkage Project LP0882016 (w ith industry support from Barwon Regional Water Corporation , Gold Coast Water, Hunter Water Corporation, Melbourne Water Corporation, South Australian Water Corporation, South East Water Limited, Sydney Water Corporation, United Water International , Water Quality Research Australia, and Water Corporation Western Australia).

The Authors Dr Eric Sivret (email: e.sivret@unsw.edu.au) is a Senior Research Associate and the odour program manager.

Prof Richard Stuetz (email:

r.st uetz@unsw.edu.au) is Co-Director of the UNSW Water Research Centre (www.water.unsw. edu.au) at The University of N ew South Wales in Sydney.

References Hvitved-Jacobsen T. (2002) Sewer processes: Microbial and chemical process engineering of sewer networks. CRC Press, Boc a Raton, 237p, Sivret E.C. and Stuetz R.M . (201 0} S ewer odour monitoring and abatement: a survey of the Australian industry. UNSW Water Research Centre, University of New South Wales, Sydney . WRC report 2010/ 8, 56p .

Stuetz R. M and Frechen F-8 (2001 ) Odours in Wastewater Treatment. IWA Publishing , London, 437p.

There were significant limitations in the industry's ability to provide fundamental proc ess data such as process age and flow for processes supplied by forced ventilation, in addition to process design data (design criteria and input characterisation), operating costs data, and proc ess monitoring/complaints data.

Furthermore, the industry partners were unable to identify the selection/ design rationale for existing odour abatement proc esses, nor provide abatement proc ess monitoring data. Wh ile it is most likely that much of this data exists within t he industry partners, it does not

water NOVEMBER 2010 81

Q

odour management

refereed paper

OLFACTORY CHARACTERISATION OF NMVOC EMISSIONS FROM WWTP INLET WORKS X Wang, G Parcsi, E Sivret, R Stuetz, J Cesca Abstract The emission of odour from wastewater treatment plants (WWTPs) can cause annoyance to a local population. Odour measurement is usually assessed either as odour concentrations (di lution olfactometry) or by the chemical assessment of t he odorous compounds. These techniq ues either provide information on the perceived effect of the emission (olfactometry) or chemically speciate the odours (analytical), however provide limited information on t he relationsh ip between odour impact and the chemical composition.

The integration of chemical and olfactory techniques using GC-MS/0 allows for the correlat ion of chemical and sensory measurements. An olfactory detection port (ODP) provides odorant characterisation, whilst t he instrumental detector (MSD, FID, etcetera) provides chemical speciation. GC-MS/0 analysis of WWTP inlet works emissions has shown that samples vary in both chemical species present, and their respective concentrations, reflecting potential for nuisance from a range of different odorants. It also demonstrates the potential benefits that GC-MS/0 analysis can offer in identifying key chemical markers for odour abatement and mitigation system design.

Introduction Complaints due to odour annoyance have become a major issue for wastewater treatment operators as the repeated release of unpleasant odours from these facil ities can constitute a nuisance to a local population (Gostelow et al., 2003). The affect on local populations has become more significant wit h the expansion of suburb ia and the associated encroachment, resulting in closer proximities between wastewater treatment plants (WWTPs) and residential receptors. Traditionally, odour abatement for WWTPs has been managed by buffer distances between t he industry and local receptors. More recently odour management strategies have become more proactive with the installation of odour abatement systems that either collect and disperse the emission or treat the emission to acceptable level to limit the impact on local populations. Often these systems do not deliver the expected reduction in odour emissions and/or meet thei r orig inal design specifications for removal efficiency, resu lt ing in the emission of odorous compounds which reach local receptors and lead to odour complai nts. The cause of these process failures is often due to inadequate chemical and odorant characterisat ion and misinterpretation of potential

Combining chemical and sensory assessment to gain understanding of the significant odorants.

82 NOVEMBER 2010 water

Figure 1. Olfactory analysis of odour sample. mass loadings. A secondary effect of inadequate odour composition information is the ineffective evaluation of odour abatement performance during commission ing and operation.

The assessment of odour management and abatement systems is based on an understanding of the emissions present in the faci lities with backg round environmental conditions. Typical odours emitted from WWTPs consist of a wide range of odorants; the essential components being hydrogen su lphide (H2S), methanethiol (C H3 SH), dimethyl sulfide ((CH 3hS), aldehydes and ketones. However, t he design of most odour abatement system designs is based on the use of one key odorant, hydrogen sulfide, to determine the loading capacity for the system. To limited extent some secondary odorants such as volatil e organic compounds (VOCs) and/or volat ile organo-sulphur compounds (VOSCs) have been used during the design/ assessment process.

The implication of only using one or two odour markers to design an abatement/mitigation system is the potential for failure due to significant variation in source composition and more specifically odorant composition. Failure to adequately remove the malodourous compounds from an emission may result in odour nuisance to local receptors.

Odour Measurement The measurement of odorous emissions is usually assessed either as odo ur concentrations (OU) by dilution olfactometry (using a national standard e.g . AS/NZ 4323:2001 or EN 13725:2003 for dilution olfactometry) or by the chemical analysis of t he odorous compounds using analytical instrumentation such as gas chromatography coupled with mass spectrometry (GC-MS) or the use of surrogate chemical

t

5]

refereed paper

odour management

.¡¡.

Figure 2. Field monitoring of H2S.

Figure 3. Continuous monitoring of H2S.

markers, like H2 S (Gostelow et al. , 2001). Sensory measurements use a panel of human sniffers (Figure 1) to characterise odours in t erms of their perceived effect but give no information regarding the chemical composition , whereas analytical measurements characterise odours in terms of their chemical composition but give little information as to t heir perceived effect of the odou r on a receptor.

The end of GC column is split into two st reams by way of a column splitter (Figure 5) that directs col umn effluent simultaneously to t he MSD and ODP though heated transfer lines. The ODP operators are able to describe the odorant character and intensity level using dedicated software that integrates with the chromatograph ic software.

Chemical methods for odour monitoring can include field and laboratory analysis of H2S emissions (Figure 2) utilising electrochemical sensors, titrimetric methods etc, and the contin uous in-situ monitoring of H2 S and VOCs (Figure 3) using electrochemical sensors, chemical impregnated tapes etc.

More recently, the integration of olfactory and chemical analysis techniques (GC-MS/0) has been applied to the analysis of odours and off-flavours. The coupling of an olfactory detection port to a GC-MS and splitting the GC column effluent between the two detectors allows sim ultaneous chemical and odorant speciation. GC-MS/0 allows the odorants to be separat ed, individually identified and described in terms of their perceived odour character.

The ODP (Figure 4) consists of a nose cone where panellists detect t he separated odorous compounds by continuously sniffing the GC colum n effluent and characterises it for perceived intensity and an odou r description.

Figure 4. ODP operator sniffs GC column effluent for odorants.

GC-MS/0 analysis has had limited applications to the assessment of nuisance and malodorous emissions but has been extensively applied to the analysis of taste and odour (or off-flavours and taints) in drinking water and moreover for aroma analysis of foods and beverages. In drinking water monitoring, GC-MS/0 analysis has been applied to the characterisation of common off-flavours such as geosmin and MIB (Hochereau and Bruchet, 2004) the results of which have yielded the drinking water odour wheel (Figure 6), which relates odour descript ors to specific odorants (Suffet et al., 1999).

The application of GC-MS/0 for the assessment of environmental odours has mainly focused on characterisi ng the changes in composition of odorous emissions from various agricultural and waste management operat ions such as swine finishing barns, poultry houses and dairy facil ities (Kai and Schafer, 2004; Wright et al. , 2005).

Simultaneous chemical and olfactory analysis (GC-MS/0) frequently characterises an emission matrix as a complex

Figure 5. Column splitter directing column effluent to the MSD and ODP.

water NOVEMBER 2010 83

odour management

~

refereed paper

adequate separation of the eluting compounds. The mass selective detector was operati ng in contin uous scan mode (35 - 500 m/z) with the obtained mass spectra recorded using the Agilent ChemStation software and analysed offline using the Enhanced Data Analysis package (Ag il ent Technologies) . The identification of the volatile organic compounds relied upon the matchi ng of the acquired mass spectra with the ChemSt ation databases (NIST02 and Wiley275).

Identification of the compounds present withi n the matrix yielded a large number of different classes of compounds including aromatics, su lphur containing organic species, nitrogen containing species, aldehydes, ketones, alcohols and ot her general hydrocarbons.

Figure 6. Example of odour wheels for off-flavours in drinking water (Suffet et al., 1999). mixture of odorous and non-odorous chemicals; often a multitude of chemical peaks will correspond to only a limited number of olfactory responses. Furthermore, the olfactory detector (human nose) is more sensitive than that of the analytical detector (MSD) for specific odorants, resulting in olfactory responses from the ODP yet an absence of a corresponding instrumental detector peak.

This paper will describe the application of simultaneous sensory and analytical assessment (GC-MS/ O) for the chemical and odorant characterisation of the emissions from inlet works of several waste water treatment plants. It will also address the broader application of GC-MS/O to the design and assessment of abatement and mitigation system performance.

GC-MS/O analysis involved splitting the gas-ch romatograph effluent between the mass selective detector (MSD) and the ODP. The temperature and flow programmes for the GC were identical to those given in t he preceding text. The olfactory stimulus chromatograms (or odour chromatograms) were recorded using the Gerst el ODP Recorder software. Offline data analysis was performed using the Agilent ChemStation Data Analysis software.

To enhance the potential for olfactory detection the split between the MSD and ODP was set to 2:3 (MSD:ODP), these split ratios were calculated using the Gerstal Column Calculator (Gerstel GmbH & Co. , Germany.) These c alculations were based on a co lumn f low of 1.6ml.min-1 for t he carrier gas

Table 1. A summary of NMVOCs detected at > 3 WWTP inlet works.

WWTP

Compound

A

B

C

D

trichloromethane nonane

Materials and Methods The results that are presented here focus on the NMVOC analysis of odorous samples collected from t he inlet works of 4 WWTPs in Northern Queensland. NMVOC samples were preconcentrated using sorbent tubes containing either Tenax TA (inert, hydrophobic sorbent for n-C7 to n-C30) or Carbotrap 300 (a blend of Carbopack C, Carbopack Band Carbosieve SIii for ethane to n-C20) (Markes International, UK). Calibrated sam pli ng pumps were used to draw known volumes of sample onto t he sorbent, to allow for subsequent relative quantification. The use of different sorbents ensures t hat the compounds identified in subsequent analysis accurately represent t he suite of compounds that are being emitted from t he source. The analytes were t hermally desorbed from t he sorbents and refocused within the general purpose graphitised carbon cold trap of the thermal desorber (Markes Unity, Markes International, UK).

Analytes were separat ed and identified using a GC-MSD (Agi lent 6890N GC, 5973NMSD, Agilent Technologies), odorants within the sample matrices were assessed using an olfactory detection port (ODP2, Gerstel GmbH & Co., Germany) (Figure 4). The chromatographic column was a polar (HPINNOWax, 30m x 0.25mm x 0.25µm , J&W Scientific Agi lent Technolog ies) capillary column. The helium carrier gas flow rate was maintained at a constant flow rate of 1 .6mUmin during t he chromatographic run . The oven was temperature programmed for a total run time of 26.50min (50°C for 2 min, 5.00°C/min to 125°C, 10°C/min to 200°c hold for 2 min) this provided

84 NOVEMBER 2010 water

decane undecane dodecane tridecane toluene o-xylene p-xylene 1,4-dichlorobenzene 1,2,3-trimethylbenzene 1,3,5-trimethylbenzene naphthalene 2-methyl-naphthalene d-limonene eucalyptol 1-ethyl-2-methyl-benzene 1-ethyl-3-methyl-benzene 2,6-dimethyl-naphthalene 2,?-dimethyl-naphthalene 4-methyl-nonane dimethyl disulfide ethylbenzene a- inene ocatane 1-methyl-naphthalene

tecnn1cat

eatures

Q

odour management

refer ee d paper

.

",lL.JJ..lJJJJ~),uh ,,1.L,,_l~--

..._ . ... , .. HI, .. . .. , .. Ot u, ...... .... ,, .. ,u,1u,,• •.•• .• • . , ............ 11 .. it tt2'N:• • » • .i-•

Figure 7. GC-MS/0 analysis of emissions from the inlet works of WWTPs. Helium with an initial temperature of 50°C with the flow prog rammed to be constant flow as the temperature increases.

The analysis of NMVOCs was supplemented by the collect ion and analysis of odour bags by dilution olfactometry (to AS/NZ 4323:2001 ), this allows for the comparison to be drawn between the NMVOC emissions and the odour concentrations.

Results GC-MS Analysis A summary of the NMVOC detected at t he 4 WWTP inlet works is shown in Table 1. The GC-MS analysis identified 123 different NMVOCs in total from all the WWTPs. The chemical

THE HYDROSENSE• Portable woter content system Instant measurements 12 cm or 20 cm probe option

7

ET107 Weather Station

No wiring required; Uses sealed connectors for sensor hook-up.

The composition of t he detected NMVOCs in t he 4 WWTPs was composed of C1-C19 compounds with nearly 95% of t hem being C5-C15 compounds, mainly simple hydrocarbons.

The other groups of compounds consisted of seven ch loride compounds (trichloromethane, 1,4-dichlorobenzene, 1,2dichlorobenzene, 1,3-dichlorobenzene, methylenechloride, methyl-cyclohexane, tet rachloroethylene) and three sulfu r compounds (dimethyl disulfide, dimethyl sulfide, and aminomethanesulfonic acid). There was on ly one nitrogen containing compound detected at the 4 WWTPs, aminomethanesulfonic acid, which is also a sulfur containing compound .

It shou ld be noted with caution that the methodology that was used for this st udy does not favour the collection and assessment of sulfur and nitrogen containing compounds, reflecting the limited number of nitrogen and su lfur compounds detected in the samples.

(i;;)CAMPBELL ~ SCIENTIFIC ~ Australia, Pty Ltd

• •••••

Full Featured Support Software.

analysis also highlighted t he significant variability in c hemical composition and abundance between. It is plausible that th is variation reflects the nature of the wastewater sources entering the sewer and subsequently reaching the WWTP. From the 123 NMVOC detected, 16 compounds (trichloromethane, tol uene, o-xylene, p-xylene, nonane, 1,3,5-trimethyl-benzene, decane, 1,2,3-trimet hyl-benzene, 1,4-dichloro-benzene, dlimonene, eucalyptol, undecane, dodecane, naphthalene, t ridecane , and 2-methyl-naphthalene) were found in all 4 WWTPs whereas the remaining 107 compounds, only 10 were found in 3 of the WWTPs and 33 were only found in 2 of the WWTPs. More t han half of 123 compounds detected were only found in one of the four WWTPs .

• • •• •• ••

MAXON MODMAX · Next G'" Module Robus1, compact solution to provide JG connectivity.

data loggers sensors weather stations monitoring systems interfaces software

OBS·3A Monitors suspended sed1me111, turbidity, solinay, 1emperowre, depth & wove height CAMPBELL SCIENTIFIC DATA LOGGERS & period Versatile. stand-alone measurement & control.

Rugged, reliable data acquisition systems.

water

NOVEMBER 2010 85

odour management Table 2. A summary of the seven major odorants identified at the 4 WWTPs inlet works.

RI (min)

Odour description

11.01 - 11.87

Alcohol, solvent, seafood, rubber

12.50 13.30 13.83 14.14 14.27 15.52 -

Vegemite, yeast, greeny, meaty worm, rosemary, burning Solvent, metallic, burning Earthy, earthy metallic, solvent, yeast Metallic, earthy, solvent, curry, onion Solvent, plastic, rubber, tan, pungent metallic Burning, smoky earthy, solvent

12.63 13.49 14.13 14.24 14.38 15.66

It was observed that less NMVOCs were detected in the emissions of WWTPs A and B than those found in t he emissions of WWTPs C and D. This difference can be attributed to the discharges entering t he sewers that feed into the WWTPs. The discharges feeding in to WWTP C and D were more industrial in origin compared to WWTPs A and B. The high industrial based composition of wastewater t herefore aids in t he generation of a large number and variety of NMVOCs at the WWTP inlet works.

A comparison of the composition of the detected NMVOCs with published odorous compounds (Gost elow et al., 2001) shows that only 3 NMVOC (dimethyl disulfide, dimethyl su lfide, and acetic acid) detected in the inlet works are reported to known odorous compounds. An addit ional 38 of the NMVOCs detected at the 4 WWTP inlet works have also been reported in gaseous emissions from WWTPs (Bianchi et al., 1997; Langenhove et al., 1985; Wu et al., 2006) or in wastewater (Escalas et al., 2003; Nikolaou et al., 2002).

ODP Ana lysis

GC-MS/O analysis allows the simu ltaneous col lection of olfactory and mass spectral data from GC analysis. Figure 7 shows a typical total ion chromatogram with t he odour chromatogram overlayed to identify t he odorants within the matrix for two emissions from the inlet works of WWTP. The results show that only a limited number of NMVOC present in the emission can by identified by the ODP operator as an odorous compound and t herefore cou ld be potentially responsible for the odorous emissions from the WWTP inlet works.

Figure 7 also shows that t he intensity of odorous compounds can be scaled from 0-3 thereby identifying the compounds that have been ident ified as having a high odour impact on the assessor, and pot entially most likely to cause odour nuisance to local receptors.

The GC-MS/O analysis of the samples col lected at the waste water treatment plants showed that between 2 to 14 different odorants were detected by the ODP operators. Seven of t hese odorants detected were present in more than 50% of the ODP analyses. Table 2 lists the retention time and odour character as perceived by the ODP operators.

The characters of the odorants perceived by the ODP operators reflect the potential of these emissions to cause odour nuisance if they were to reach local receptors. The design of an odour abatement system would need to aim at mitigating the compounds that will have the highest potential to cause odour nuisance if they are to reach a local receptor.

refereed paper

collection of olfactory and mass spectral dat a by way of GCMS/O analysis demonstrates that only a small number of t he NMVOC's present within the entire emission matrix are responsi ble for the result ing odour. A total of 123 distinct NMVOC's were detected across all the WWTP inlet work samples.

However only 16 of these were common between sites.

These compounds reflected a variety of long chain alkanes (nonane (C9) to t ridecane (Cd) substituted aromatics (o- & pxylene, 1,2,3- & 1,3,5-trimethylbenzene, 1,4-dichlorobenzene, toluene, naphthalene, 2-methyl -naphthalene) terpenes (dlimonene, eucalyptol, a-pinene) and trichloromethane. Other compounds that were present wit hin selected samples included halohydrocarbons, volati le organo-sulphur compounds (disulfides). The sensory analysis of the the samples yielded fourteen unique odorants, seven of these being common to all samples.

Th is study illustrates t he potential for combined chemical and sensory assessment utilising GC-MS/O to gain critical understanding as to the chemical composition and more significantly the odorants t hat are present within a sample matrix.

Acknowledgment The authors acknowledge the fi nancial support from the Australian Poultry Cooperative Research Centre for Gavin Parcsi 's PhD Scholarship, and CH2MHill, Australia.

The Authors Xinguang Wang is a Research Engineer, Gavin Parcsi* is a Postdoctoral Fellow, Eric Sivret is Senior Research Associate and odour program leader, Ri chard Stuetz is a Professor and Co-Director at the UNSW Water Research Centre, The University of New South Wales (www.water.unsw.ed u.au). Josef Cesca is the Regional Technology Manager at CH2MHILL, Aust ralia. Email: g.parcsi@unsw.edu.au.

Refe rences Bianchi, A. P. & Varney M. S., (1997) Volatisation process in w astewater treatment plants as a source of potential exposure to VOCs. Annals of Occupational Hygiene 41(4): 437-454 Escalas, A., Guadayal, J. M., Cortina, M., Rivera, J. and Caixach, J., (2003) Time and space patterns of volatile organic compounds in a sewage treatment p lant. Water Research 37: 3913-3920 Gostelow, P., Parsons, S. A. and St uetz, R. M., (2001). Odour measurement in sewage treatment - a review. Water Research. 35(3): 579-597 Gostelow, P., Longhurst P. J. , Parsons, S. A. and Stuetz, R. M. (2003).

Sampling for Measurement of Odours. IWA Scientific and Technical Report No.17, IWA Publishing, London Hochereau, C. and Bruchet, A. (2004). Design and application of a GCSniff/MS system for solving taste and odour episodes in drinking water.

Water Science and Technology. 49(9): 81-87 Kai, P. and A. Schafer (2004) Identification of key odour components in Pig House Air using hyphenated gas chromatography olfactometry.

Agricultural Engineering Int ernational: the CIGR Journal of Scientific Research and Development, VI (04 006) Langenhove, H. van , Roelstraete, K., Schamp, N. and Houtmeyers, J., (1985) GC-MS identification of odorous volatiles in wastewater. Water Research 19(5): 597-603.

Nikolaou, A. D. , Golfinopoulos, S. K., Kostopoulou , M. N., Kolokythas, G. A. and Lekkas, T. D. , (2002) Determination of volatile organic compounds in surface waters and treated wastewater in Greece. Water Research 36: 2883-2890.

Conclusion

Suffet, I. H., Khiari, D. and Bruchet, A. (1999}. The drinking water and odour wheel for the millennium: beyond geosmin and 2-methylisoborneol. Water Science and Technology. 40(6): 1-13

GC-MS/O analysis of samples from WWTP inlet works revealed that that there is a complex mat rix of NMVOCs that form the emissions from these unit processes. The simultaneous

Wright, D.W. , et al. (2005) Multidimensional gas chromatography-olfactometry for the identification and prioritization of malodors from confined animal feeding operations. Journal of Agriculture and Food Chemistry. 53: 86638672

86 NOVEMBER 2010 water

I featur s

~

odour management

refereed paper

SUSTAINABLE ODOUR CONTROL AT PERTH'S WASTEWATER TREATMENT PLANTS C Nichols, S McNeil, G P Van Durme, K Cadee, I Wallis, A Adams development of an improved odou r management system for the upgraded plant including containment, conveyance and treatment.

Abstract The Water Corporation of Western Australia formed t he W2W Alliance, including Black & Veatch, Theiss and SKM, to deliver a program of upgrades to Perth's three largest metropolitan wastewater treatment plants (WwTPs). Thi s paper covers the recent upgrades of odour management at the Beenyup and Woodman Point WwTPs, which were implemented t hrough the Alliance.

Previous Odour Control System

Primary, Secondary and Sludge Areas Discharge Stack

The odour control system at t he Beenyup WwTP had been established and upgraded t hrough three phases between 2003 and 2005. The primary, prelim inary, Preliminary secondary (biolog ical Treatment Area process) and sludge Introduction treatment areas had been Stage 1 Chemical covered and were extracted The W2W Al liance was Scrubbers at relatively high rates to the formed by Western odour treatment system Australia's Water Corporation Figure 1. Previous odour control system process flow diagram. consisting of t hree trains of to deliver a five-year, $352 two-stage chemical million program of upgrades Part 1.

Beenyup WWTP two smaller treat ment scrubbers.

The to Perth 's metropolitan wastewater trains (termed Stage 1) received a high treatment p lants (WwTP) at Beenyup, Odour Control System concentration, low flow airst ream from Subiaco and Woodman Point. As part of Upgrade t he preliminary treatment area, and t he t he wastewater upgrades, a significant remaining larger train (termed Stage 2) upgrade to the existing odour control The Beeny up Wastewater Treatment received a low concentration, high flow systems at the Beenyup and Woodman Plant (WwTP) is the second largest airstream from the primary, secondary Point WwTPs was undertaken. wast ewater treatment plant in Western and sludge treatment areas as shown in Australia.

The plant serves domestic, Part 1 of this paper covers the odour Figure 1. commercial and industrial customers control system upgrades undertaken at Perth' s northern suburbs with a form A comprehensive odour monitoring the Beenyup WwTP to increase the rapidly growing equivalent populat ion in and modelling program confirmed that system's nominal capacity from 120 MLD excess of 600,000. the previous upgrades had significantly to 135 MLD, with consideration for the plant's ultimate capacity of 150 MLD. The existing odour management system was upgraded to restrict odour emissions to the levels currently experienced.

Part 2 of the paper deals with Woodman Point, which had undergone considerable upgrades to secondary treatment since it was first commissioned in an industrial zone in 1966. To ensure that local residents are not unduly impacted by odour, a substantial upgrade to the Woodman Point WwTP odour control system was implemented to reduce odour emissions by 50%.

This paper is a combined summary of two papers presented at the AWA Odour Specialty Conference, Sydney, August 2010.

In order to meet t he upgraded treatment capacity of 135 MLD, significant upgrades were required to the primary and sludge treatment systems including refu rbishment of three primary sedimentation tanks, new sludge thicken ing facil ities, additional digestion capacity and a new dewatering centrifuge.

Regulatory requirements restricted odour emissions to the levels currently experienced. These drove the need for

The treatment systems represent international best practice.

reduced the plants odour emissions to an estimated 37,500 OU/s in ground level fugitive emissions. Therefore, t he Beenyup WwTP upgrade from 120 MLD to 135 MLD had a target maximum ground level odour emission of 37,500 OU/s or less to maintain or reduce the odour levels around t he plant.

Increased Air Flows and Odour Loads The extraction rates throughout the treatment plant were evaluated and increased to bring them in line with previously established guidelines (Cadee and Wallis, 2007) for ext raction rates to achieve a negat ive pressure under covered areas. In addition, a truck

water NOVEMBER 2010

87

odour management loadout enclosure controls t he high level of odour emissions experienced during biosolids loading. Altogether this resulted in the ground level odour emissions being maintained at current levels.

Th e extraction rates established to achieve the reduced odour emissions from the expanded treatment plant resulted in an increase in the total airflow rate from 140,000 m3/hr for t he current 120 MLD capacity to 260,000 m3/h r for the 135 MLD upgrade. A further increase to 320,000 m3/hr is expected in t he future to maintain the existing odour emissions from the plant at t he 150MLD ult imate capacity.

The odorous contaminant concentrations in the ai rstream were assessed to ensure the most su itable treatment system was implemented.

Concentrations were evaluated using a year of hydrogen sulphide (H 2S) testing throughout the existing treatment p lant, detailed analysis of high risk, high impact ai rstreams and comparison with the comprehen sive database of odorous containment concentrations which had been established as part of the Wood man Point WwTP odour control system upgrade (see Part 2). Figure 2 presents the flows and co ntaminant concentrations from different odour sou rces after completion of the 135 MLD upgrade. The equivalent H2S concentrations are based on hydrogen sulphide, volatile organic carbon (VOC), mercaptan, dimethyl su lphide and ammonia concentrations.

Several distinct airstreams began to emerge with differing characteristics.

There were localised areas of high odour levels f rom the preliminary, primary and sludge treatment areas, but there were

~

140,000

350

120,000

300

100,000

250

.~-

80,000

200

V,

E C

.c ..... "' !.

~

::z:

0

'::

:..

60,000

150

....... m

0

'E

<

'o 100 '0 3 .5.

20,000

50

0

0 Preliminary Treatment Area

• Airflow (m3/hr)

High Odour

Primary Treatment Areas

Areas

Areas

Chemical Scrubber Capacity Assessment The existing chemical scrubber capacity was evaluated, which concluded that t he Stage 1 scrubbers had significantly less load capacity than the current operation based on a discharge of 100ppb H2 S. The Stage 2 scrubbers cou ld cope with a significant increase in both flow and load

200

m

.,:c· ~ ::,

...

150

100,000

:I: N VI

n 0

80,000

::,

100 fil

::,

60,000

50

20,000

Secondary Treatme nt Area

• Peak Equivalent H2S Concentration

0 Stage 1Stage 1 - ReStage 2 Stage 2 - ReExisting Duty evaluated Existing Duty evaluated Point Capacity Point Capacity

and still rel iably meet t he this discharge criteria. Th is substantial increase in the scrubber's capacity was in part due t o removal of the high CO2 , secondary treatment area airstream Figure 3 presents a summary of the original and revised capacity of the Stage 1 and 2 chemical scrubber systems.

Sustainability Assessment of Potential Treatment Systems A primary consideration was to improve the sust ainabi lity of the system while increasing the treatment capacity. The existing chemical scrubbers have high chemical and potable water consumption; therefore, alternat ive technologies including biotrickling filters were considered .

Assessment were mad e against 17 sustainability criteria tailored to WwTPs and the odou r control system upgrade, but each criterion was identified as either a prim ary driver (an essenti al objective or outcome) or a secondary consequence (impacts or effects t hat need to be managed). The criteria identified as primary drivers were: • Chemical use

~

• Potable water use

o· ::, ,;-

• Delivering the project for the best long term value for money in capital and operating costs

!!!

'tl

0

water

Treatment

Areas

A key to optimising the odour control system was recognising the three distinct airstreams an d routing each to t he most appropriate treatment process.

120,000

NOVEMBER 2010

Treatme nt

· also areas with relatively low odour levels from primary and sludge treatment processes. In addition, the secondary treatment area had distinct attributes with very low odour levels, but a high flow rate and high carbon dioxide (CO 2) concentration.

40,000

88

Primary

Low Odour Sludge

Figure 2. Airflows and equivalent H2S concentrations for similar air streams for the 135 ML/d plant.

.&l C

3

~

• Peak Equivalent H2S Concentration Capacity (ppmv)

Figure 3. Existing chemi cal scrubber capacity.

lowOdour

High Odour Sludge Treatment

• Average Equivalent H2S Concentra tion

250

• Airflow Capacity (m3/hr)

~

io·

:,

40,000

140,000

.s3

n

@

160,000

...

N

g

3

180,000

.:::;

refereed pap e r

• Consistency with a longer term plan ned program of upgrades The sustainability assessment indicated a strong preference for t he biotrickling filters due to reduced chemicals, potable water and energy

technical features

[sl

odour management

refereed paper

5

5

4 3

4 3

2

2

e1

e1

go

o 0 ~-1

u,_1 -2

-2 -3

-3 -0 C

(U

...J

2

Q) U)

Q) t/)

?

>,

u

~ E ~

::,

(1)

Q)

0

t/)

Q)

u:::

Q)

w

.c

0:: 0

c

X

X

Q)

Q)

Q)

u (U

-0

?

C

(U

...J

social

consumption together with a substantial reduction in operating costs. The assessment resu lts are summarised in Figures 4 and 5.

Treatment of Secondary Area Odour Sources As discussed above, it was decided to t reat the secondary treatment area airstream separately due to its distinct attributes. A number of technologies were considered and compared using a net present val ue (N PV) assessment as shown in Figure 6.

An activated carbon system was implemented due to the lower capital cost and NPV when compared with the other technologies as shown in Figure 6.

The high capital cost of a biotrickling filter system that would meet stakeholder expectations was prohibitive for the treatment of low odour level airstream.

However, its operating costs, including reclaimed effluent and power, were significantly lower than the other alternatives evaluated.

The elevated CO2 concentration in the airstream wou ld resu lt in high chemical consumption rat es (and operating costs) for a Chemical Scrubber when compared with other technologies.

S20,000,000

2

~

c

U) 'X X ::, ti Q) Q) ?

(U .!,? ?

0 a. a. 1i -0 a.

Q) :c 0) E (U Q) 0 ~ 0 .5 .c '6 e! 0 !:: (1) Q) Q) iii ::, ::, .E a. e! a. ~ Q) u u::: ii= 0 w 0:: 0) u <( Q) C t/) 0 C ::, C: Q) .c ·1: a..

0 0:: 0 E C (U E ~ ::, 0 a:: a.

0

Q) t/)

::,

Q) t/)

?

~

C rou :c ·x ·u

::,

Q)

?

1i .!!!

E ~

(1)

e

e

e

Figure 4. Biotrickling filter sustainability assessment.

Q)

criteria

0::

Denvironmental Deconomic I

social

(1)

0::

Figure 5. Chemical scrubber sustainability assessment.

An independent NPV evaluation undertaken by a supplier of odour control systems confirmed that an activated carbon system would provide the lowest NPV for the secondary area airstream.

The supplier highlighted that at a CO 2 concentration of 400ppm, the chemical scrubber and activated carbon systems have roughly equivalent NPVs . However, at CO2 concentrations of 2000ppm or greater, the activated carbon system has a substantially lower NPV. A preliminary investigation into the use of a biofilter (also referred to as a compost or bark filter) showed that while there was sufficient space and a low NPV were anticipated, it was not considered a viable solution due to the high maintenance requirements, inability to discharge via an elevated stack and potential issues with media replacement.

req uirements for the future 150 MLD upgrade, without modification of the ductwork, fans or platforms. Figure 7 is a photo of the installed system.

The odorous airstream goes through a mist eliminator before entering the dual bed activated carbon unit and passing through either the upper or lower activated carbon media bed. Each media bed is 1000mm deep, and consists of a mixture of 40% virgin acti vated carbon and 60% caustic impregnated carbon to target H2S, volati le organic compou nds (VOC) and other odorous com pounds.

The activated carbon media removes the odorous compounds within the foul air stream by physical and chemical adsorption prior to discharge through the stack.

The empty bed residence time (EBRT} of the activated carbon system is 3.5

The impact of discharging the secondary treatment airstream directly through a stack with no prior odour treatment was also evaluat ed. Odour modelling showed that direct discharge of airstream throug h a 50m stack resulted in a peak odour of 25 OU at community receptors. As the target peak odour concentration from the stack is 0.3 OU in areas zoned as urban, a direct discharge was not acceptable.

Activated carbon system

$16,000,000 s,2,000,000 S8,000,00'J $4,000,000

S·

U)

.!,?

.s

e

Denvironmental Deconomic I

criteria

'::,

a. a. ? 0 Q) 1i 0) E (U 0 1i -0 a. .c '6 ~ 0 0 e!::, (1) iii e! a. ~ Q) u a.

0:: 0) u ·1: <( 0 C ::, C: ·1: a..

0 E C i (U E 0 :i a:: a.

0

:c C ~ ~ rou ·x Ii :c .!!! ::, .E :t: Q) ::,

• Activated Carbon System

•capital

Blotrickling FIiter

Chemical Scrubber D()peraling

Figure 6. NPV assessment for secondary area odorous airstream treatment.

A turnkey activated carbon and extraction fan system was supplied by Aromatrix Pty. Ltd. consisting of 5 duty/ 1 standby dual bed activated carbon units and 2 duty/ 1 standby extraction fans. The system allows for the easy incorporation of one additional activated carbo n unit to meet the odour control

Figure 7. Activated carbon system.

water NOVEMBER 2010 89

odour management

~

seconds when 5 units are in operation (i.e. standby unit is offline).

The treatment system is performance guaranteed to meet a discharge of < 50 ppb H2S and < 500 OU for a minimum media life of 12 months.

Treatment of Preliminary, Primary & Sludge Odour Sources While removing the f low and load from the secondary treatment area significantly reduced the demands on the existing chemical scrubbers, there was still insufficient load capacity withi n t he exist ing chemical scrubbers to meet t he discharge criteria of 0.1 ppm H2S as shown in purple in Figure 8.

In reviewing the characteristics of the different airstreams from the preliminary, primary and sl udge treatment area, it was identified that some sources had very high odour levels while others were relatively low sources of odour. This led to consideration of incorporating a biotrickling filter upstream of the existing chemical scrubber systems to provide bu lk removal of H2 S and other odorous compou nds at a relatively low cost.

The chemical scrubbers would then act as a polishing system to ensure the odour targets of a 100 ppb H2S and 1000 OU were met.

Treating only the high odour sources with the biotrickling fi lter was com pared with pretreating the entire airstream to minimise t he load on the downstream chemical scrubbers.

Figure 9 shows that with t he incorporation of a biotrickling fi lter system, t he St age 2 chemical scrubbers alone had sufficient capacity whether either all preliminary, primary and sludge airstreams were treated by the biotrickling fi lt er or only those airstreams with a high odou r levels.

An NPV assessment identified that two additional biotrickling filter towers would be req uired to treat the airflow from these moderate odou r sou rces, whi le only a minimal amount of additional ductwork, dampers and instrumentation would be required to treat the moderate odour sources at the chemical scrubber.

It identified approximately $1.5 million in cap ital investment savi ngs by ducting the moderate odour sources directly to the chemical scrubbers, wh ile t he additional operating costs were estimated at only $30,000 per year. This was a significant reduction in the upfront capital investment whi le maintaining lower operating cost s over the 40 year design life of the scrubber system. These advantages were magnified when the futu re upgrade to a plant capacity of 150 MLD was considered, providing an NPV of over $2 million in saving when the moderate odour sources were ducted

90

NOVEMBER 2010

water

refereed paper

180,000

250

m

.0

160,000

C

...

.i:: ...... m .§.

3: 0 c.: ... ~

~-

200

140,000

;r

... ::,

120,000 150

100,000

:c

"'"'n 0 ::,

80,000

100

60,000 40,000

50

20,000

lil

::,

...5·.::::,

,;'ti

0

0

3 .5.

Stage 1- ReStage 1 Stage 2 Stage 2 - ReExisting Duty evaluated Existing Duty evaluated Point Capacity Capacity Point • Air fl ow Capacity (m3/hr)

• Peak Equivalent H2S Concentration Capacity (ppmv}

Figure 8. Chemical scrubber capacity vs. required treatment capacity.

.c ...... m .§.

3:

0

't

:,i

200,000 180,000 160,000 140,000 120,000 100,000 80,000 60,000 40,000 20,000 0

60.0

m

.0

50.0

.c· C

;r 40.0

... ::,

::c

"' 30.0 "' n 0 ::,

20.0 lil ::, 10.0

.....::-

5·

:,

0.0 Stage 1 & 2 Chemical Scrubber Capacity

'ti

3

High Odour BTF .5.

BTF Treatment of Treatment + Prelim, Moderate Primary& Odour Sludge Odorous • Peak Equivalent H2S Concentration Capacity (ppmv)

Stage 2 Chemical Scrub be r Capacity

• Airflow Capacity (m3/hr)

,;-

Figure 9. Comparison of chemical scrubber capacity vs. required treatment capacity for different pre-treatment configuration. directly to the chemical scrubber system.

By identifying the differences and similarities in different odou r sou rces, they cou ld be appropriately grouped to minimise t he NPV cost of the odour control system while taki ng advantage of existing assets and maintaining the high standard for the discharged treated air. Figure 10 presents a process flow diagram summarising the system config uration.

The St age 1 chemical scrubbers are no longer required and will be decommissioned.

Biotrickling filter system A turnkey biotrickling filter and extraction fan system was supplied by Cleanteq Ltd. consisting of 5 duty/ 1 standby biotrickling filters and 2 duty/1 standby extract ion fans. The system

allows for the easy incorporation of two additional biotrickling filters to meet t he odour control requirements for the fut ure 150 MLD upgrade, without modification of the w ater supply syst em, nutrient dosing system, ductwork, fans or platforms. The installed system is pictured in Figure 11.

The biotrickling filters consist of two distinct treatment phases contained within a single vessel, where the first phase operates at about pH 2 and primarily removes hydrogen sulphide (H2 S) and the second phase operates at a neutral pH and removes other odorous compounds including mercaptans, dimethylsulphide and volatile organic compounds (VOCs).

The biotrickling filter media, composed of polyurethane foam and

technical features

[sl

odour management

refereed paper

polypropylene balls, is designed to maximise the surface area for contact between the gas and liquid phases, promoting mass transfer of gaseous contaminants into the liquid phase where they are available to the attached microbes on the media surface. The mixed bacter ial culture then oxidises the odorous cont aminants, reducing the odour levels by a minimum of 99.5% H2 S and 95% OU.

High Odour Level Sources IPrimary, Preliminary & Sludge Treatment Areas)

79,000 m3/ hr@ 131 ppm 11 avg. equiv. H2S

Moderate Odour

EJEJEJ

5 Duty, 1 Standby Biotrickling Filters

99.5% H25 removal 95% OU removal

Level Sources (Preliminary & Sludge Treatment Areas)

An EBRT of 17.5 seconds is achieved when six vessels are in operation and 14.6 seconds when five vessels are in operation (i.e. standby unit is offline).

The water supply system is a critical component of the biotrickl ing filters as it provides hydration and nutrients for the micro-organisms degrading odorous compounds. A fu lly redundant water supply system was inst alled, with rec laimed effluent supplied during normal operating conditions and potable water supplied when reclaimed effluent is unavailable. Nutrient is dosed into the biotrickling filter water supply, regardless of the water supply source, to ensure sufficient micronutrients are present for microbial metabolism.

Stage 2 chemical scrubbers Airstreams with a moderate odour level from the primary and sludge treatment areas were directed to the Stage 2 chemical scrubbers in conjunction with the biotrickling filter discharge.

The odorous airstream is extracted through the fans to the primary scrubber, where the gas travels upwards throug h a

48,000 m 3/hr@ 12 ppm, avg. equiv. H2S Secondary Treatment Area 132,000 m 3/hr@ 4 ppm, avg. equiv. H2S

127,000 m 3/hr @ 8 ppm, avg. equiv. H2S & 16 ppm, peak equiv. H2S

DDD

Existing Discharge Stack <100 ppb, H2S <lOOOOU

- - - - - - - -......â&#x20AC;˘

DOD

5 Duty, 1 Standby Activated Carbon Scrubbers

Discharge Stack < SO ppb,H2S <500 OU

Figure 10. The 135 ML/d odour management system at Beenyup WwTP. 3300mm deep random pack polypropylene media bed. NaOH is dosed into the primary scrubber and NaOH and NaOCI dosed into the secondary scrubber recirculating streams to maintain the pH and oxidation reduction potential (OAP) setpoints, respectively. Demisters at the top of each scrubbing vessel remove any liquid droplet s that may be entrained with the gas before discharge to the 50m high stack.

The primary scrubber targets approximately 50% equivalent H2 S removal, while the secondary scrubber targets removal of the remaining H2 S, mercaptans, dimethylsulphide (OMS) and other odorous contam inants to meet the target discharge of 100 ppb H2S and 1000 OU.

Commissioning & Performance Testing One of the chal lenges of the new odour control system was the need to minimise odour emissions throughout construction and commissioning. This led to the system being installed in parallel with the existing odour control system with a series of tie-ins to complete the installation. A comprehensive Construction and Commissioning Odour Management Plan was developed, which detailed these critical tie-ins, with attention being given to the need to develop sufficient biomass in the new bio-trickling filters.

Treatment of the secondary treatment area airstream through the activated carbon system commenced in March 2010. The inlet H2 S concentration has

100 90 80 70

s

2.

60

.; >

!

5 0 "Cl

0

50 40 30 20

10 0

03/06/10

06/06/10

08/06/10

09/06/10

Date

Figure 11. Biotrickling filters.

Figure 12. Activated carbon system discharge odour levels.

water NOVEMBER 2010

91

~

odour management been somewhat lower t han pred icted, with an average H2 S concentration of 0.5 ppm , wh ich may be due to the inc reased extraction rates and cooler weather.

refereed paper

50 ..------------------------------, 0.99

0.98

The activated carbon system has cont inuously met the performance criteria for H2 S since commissioning, with no H2 S detected on the system discharge.

The odour levels from the activated carbon system are presented in Figure 12 and comply with the performance criteria for t he system of < 500 OU. The biotrickling filter system was comm issioned in May 2010 with odorous air from t he primary and sludge areas.

Commissioning of t he syst em wi ll be com pleted by t he end of this year, with preliminary H2 S removal results presented in Figure 13.

0.97

0 .96

'*-

:::c N

V,

0.9S

;:,c ~

0.94

0 .93

3 0 < !!!..

0 .92

0 .91

Average Discharge H2S Concentration 0.52 ppmv 0 .9

20/07/10 21/07/ 10 22/07/10 23/07/10 24/07/10 25/07/10 26/07/10 • Biot rickling Filter Inlet H2S Concentration ppmv • Biotrickling fi lter % H2S Removal

• Biotrickling Filter Outlet H2S Concentration ppmv

Figure 13. Biotrickling filter performance.

Conclusion, Part 1 • Understanding the odour source attributes

By identifying three distinct airstreams and optimising the odour treatment process for each airstream , a step change in the sustainability of t he Beenyup WwTP odour control system was accomplished.

The upgrading program was based on the following drivers

• Using the right treatment process for each airstream

• Establishing extraction rates to achieve objectives

• Maximising the benefit from existing assets

Part 2. Woodman Point Odour Control System Upgrade Introduction

• Separat ing dissimilar airstreams (duct ing can be less costly than addit ional treatment systems)

400,000

l

300,000 250,000

0

()

,

l

~

200,000

Recent urbanisatio n of the area adjacent to Woodman Point has led to increased demand for odour control from the community, cu lminating in the Water Corporation's commitment to red uce odour emissions by 50%. Th e W2W Alliance undertook a program of work to upgrade the Woodman Point WwTP and inst all best practice odou r control for the first of three stages.

The Stage 1 odour control improvements include covering the anoxic zones of the Sequencing Batch Reactor (SBR), upgrading existing odour containment structures, increasing extraction rates throug hout the treatment plant, decommissioning the existing scrubbers, constructing a new odour control system and installing a 50 m high discharge stack. The Stage 1 improvements also incorporated a sludge amplification component w hich included a new sludge loadout enclosure, ferric chloride dosing of sl udge and hightemperature waste gas flares to red uce odour emissions. The Stage 1 odour

92 NOVEMBER 2010 water

The Stage 1 upgrade wi ll provide over 50% reduction from current odour levels at t he current plant inflow rate of 120 MUd. Further odour reductions wi ll be achieved with the Stage 2 and 3 upgrades. Emissions may increase slightly when the plant throughput is increased to 240 MUd, but the overal l odour levels should be 73% lower than current odour emissions.

450,000

350,000

i

1SO,OOO

",

;;_

100,000

0

i

50,000

The recently installed system is currently meeting its specified performance targets as the system commissioning continues. A detailed emissions measurement, odour modelling, and ground-truthing program w ill be undertaken is t he summer of 2011 to co nfirm t he system has meet its goal of maintaining odour emissions to the su rrounding environment.

Exls1Ing

Stago 1

Stago 2

Stago 3

Ew:pandod

120 MUd

160 MUd

160 MUd

160 MUd

240 MUd

Figure 1. Odour reduction with staged improvement. control improvements cost $52.5 million and were commissioned from 2008 to 2009.

The planned Stage 2 works will further improve the effectiveness of odou r control at the plant's preliminary and primary works by increasing the air extraction rates under the covers. Further odou r control measures have been considered for increasing the plant capacity in t he f uture, including coveri ng the aerated sections of the secondary treatment system (SBR) as part of t he Stage 3 upgrades if required.

As shown in Figure 1, each st age of the odour control upgrades wil l red uce t he tot al odour emissions from the plant.

Flows and Loads A comprehensive assessment of odour sources and emissions rat es was conducted throughout the t reatment plant in 2005, establishing the basis for design of t he odour control system upgrade (Nichols, 2010). Additional t est ing in 2007 improved the accuracy of the predicted odour levels from the SBR bioselectors when covered, which are a significant portion of the total load reaching the scrubbers, by taki ng measurements at a more representative location (Van Durme et al., 2009).

Odour Control System Design Chemical scrubbers had been installed at Woodman Point in 2002 and in the other Perth metropolitan WwTPs as t he Water Corporation's preferred t ec hnology for odour removal. They provide effective

technical features

~

odour management

refereed paper

and reliable treatment and operations staff are familiar with the operation and maintenance requirements.

However, with the high H2 S loads predicted at the Woodman Point WwTP, chemical scrubbers would consume a large quantity of chemicals.

• Preliminary Area • Primary Area • Sludge Area .._---+--+---.----i Total= 75,300 m3/hr

SBR Bioselector 65,900 m3/hr Stage 1 Total 14 1,200 m3 /hr

Based on comprehensive odour modelling, a 50 m discharge stack was required to achieve the target 5 OU (1hr average, 99.9% frequency) at the nearest odour sensitive premises for a discharge of 1,000 OU through the stack plus fugitive ground level emissions.

It was decided to employ a Figure 2 presents a combination of biotrickling schematic of the Stage 1 filters to pre-treat the odorous odour control system for the 50 m Stack Discharge air from most of the treatment Woodman Point WwTP. plant, combined with Figure 2. Woodman Point WwTP odour control system schematic.

Odour Control System downstream chemical Description scrubbers, to polish the significantly reduced by the upstream biotrickling filter discharge and treat the Biotrickling filters biotrickling filters. low odour SBR bioselector airstream, The biotrickling filter system was (Van Durme et al., 2009). This Two stages of chemical scrubbers supplied by Bioway and consists of six combination of treatment syst ems were used to meet the target discharge duty vessels operating in parallel, as minimised the number of biotrickli ng odour levels, minimise chemical costs shown in Figure 2. fi lters requi red without compromising the and allow removal of one scrubber for syst em discharge quality and took maintenance when required. The The bottom section of the tower advantage of the low operating costs of scrubbers were designed to allow for removes H2 S and operates around pH 2, the biotrickling filter system for conversion to single stage operation in whi le the top section operates at a airstreams with high odour levels. The the future to double the flow capacity, neutral pH allowing other odorous chemical scrubbers provide the required provided that long-term performance contaminants to be removed. Water is additional treatment at a very low confirms that single stage treatment is introduced at the top of the biotrickl ing operating cost because the odour load is accept able. fi lter intermittently to wet the microbes

A new name for 100 years of experience in energy and water

Since 1914, Hydro Tasmania has forged a reputation as a leader in renewable energy and water management. And more recently as its consulting arm, we've used that experience and knowledge to solve complex issues for other businesses. So while we have changed our name from Hydro Tasmania Consulting to Entura, when it comes to delivering the smartest solutions in energy and water - it's business as usual. To learn more about how we've changed, go to www.entura.com.au

odour management

refereed paper

and assist in the removal of su lfuric acid.

The water supply system is fully redundant, with reclaimed effluent supplied during normal operating conditions and potable water dosed w ith artificial nutrients supplied when reclaimed effluent is unavailable.

A minimum gas residence time of 14 seconds is provided in each scrubber to achieve a minimum of 99.5% H2 S removal and 95% odour unit removal.

Chemical scrubbers The Environmental Group Limited (EGL) supplied the chemical scrubber, chemical dosing system and extraction fans. As shown in Figure 2, the plant consists of four scrubbers operated as two trains of two scrubbers in series. The chemical scrubbers are designed to achieve less than 100 ppb H2 S and less than 1,000 OU. NaOCI and/ or NaOH is dosed into the recirculation flu id, which is then introduced at the top of each scrubber where it reacts with contaminants in the gas stream. The primary scrubber is operated with NaOH only at a pH of 12.0 to achieve a minimum of 70% H2S removal and 50% removal of mercapt ans.

The secondary scrubber uses both NaOH and NaOCI to treat the remaining odorous contaminants. The secondary scrubber is operated at a pH of 9.35 and an oxidation reduction potential (ORP) of 690mv, This chemical dosing configuration provides effective, low cost treatment of the odorous airstream by minimising the use of costly NaOCI. All the chemical scrubbers have the capability to use both NaOH and NaOCI to al low each unit to operate as a single stage as required.

For the greatest overall economy of construction , the chemical scrubbers, chemical dosing systems and extraction fans were sized to treat the Stage 2 upgrade airflow of 195,900 m 3/hr. Key interconnecting ductwork are sized for the 240 MUd design, so future scrubbers and extraction fans can easily be added with flanged connections to existing ductwork. The scrubber ductwork configuration and isolation dampers provide flexibility so that the influent airflow can be treated by either the primary scrubber, or secondary scrubber or primary and secondary scrubbers (which is the normal mode of operation).

Additional blanked flanges are provided for connection of additional ductwork in the future to allow the system to operate as four single stage chemical scrubbers.

A fish-eye photo of the odour control system is shown in Figure 3.

94 NOVEMBER 2010 water

Figure 3. Biotrickling filter and chemical scrubber odour control system with 50m stack at Woodman Point.

50 -

45

Ea. 40

.

-

lnlet (ppm)

~

-¡

r'

T T outlet (ppm)

-

5 E Q. 4 ~

~ C:

~

C:

35

0

.:: 3 E

E 30

'E., (J

'E.,

25

(J

C:

C:

0 (.)

20

2

~ 15

-

::c:

::c:

., -;

]i 10

-=

0

(.) (/) N

0

5 0

0

"' 0

0

~

"'

0 0

a:

;:

0 "' 0

;;: ;:

0 "' 0

~

;:

"'00 5:

0 "' 0

5: ;: ;:

0 "' 0

"' :;:

0 "' 0

;:

;:

;:

;:

0

.;

~

"' 0 0

..a:

!:?

Date

Figure 4. H2 S performance of biotrickling filters.

0.8

40 Treatment lapie due to 1xmcr failure 35

s 6

-

S crubber Inlet A2S

-

Scrubber Outlet H2S

0.7

s Q.

30

0.6 8,

25

0.5 -.c

.

20

""'C

.

en

IS

0.3 en

.S!

10

0.2 .!I

g

¡=

.

C

I;

"

l:: C

""'C u

C

u

ii

-

ii

:,

-=

0 0.1 0.0

Si

"' Si

'Date

Figure 5. H2S Performance of chemical scrubbers.

Commissioning A priority during construction and commissioning of the odour control system was to achieve low odour emissions to the surrounding commu nity during incorporation of the new system.

One of the risks highlighted in these plans was the time required to develop

sufficient microbial biomass in the biotrickling filter, which typically requires several weeks.

The advanced planning and conscientious execution resu lted in a highly successful commissioning with no major odour excursions (Nichols, 2010).

During the entire commissioning period there were only a few occasions of short

technical features

Gs]

odour management

refereed paper

200,000

2

Gi >

., _, ;;

.,, 0

0

'\.

/

140,000 120,000

I

100,000

r

1,800

r

A

160,000

s-

2,000

100.0%

180,000

\

\

&

Jf'

"'"

99.0%

I'

98.0%

Q.

0

~

97.0%

'/

80,000

1,600 0

96.0%

60,000

95.0%

!:;

"'"'3

.....,

~

"C

93.0%

0.

., m., "'., "'., "'., r-., ., "'., ....., 0. 0. 0. 0. 0. 0. 0. 0.

E

E

E

E

E

E

E

.11!

.11!

.11!

.11!

.11!

.11!

.11!

.11!

LabA

-

E

00

N

'ii

_,.,> 1,000

< !!!.

94.0%

-

1,400

ÂŁ 1,200

0

40,000 20,000

s...

:, 0

800

0

600 400 200

0

E

~

0.

E ~

.....,

N QJ

QJ

C.

C.

C.

&'l

&'l

&'l

E

-

LabB

....., Removal Effici ency

Figure 6. Odour removal performance of biotrickling filters. duration (less than 3 hours) where the H2S concentration exiting the stack was above 1.5 ppm , which were due to mechanical problems or planned shutdowns.

Performance Testing Airflows The design airflows through the biotrickling filter and chemical scrubbers were simulated during performance testing as the actual airflows were not available due to the the staged construction and commissioning process. One biotrickling filter was partially isolated during performance testing to achieve an airflow of 14,500 m3/ hr t hrough five of the biotrickling filter units and 5,000 m 3/ hr through the sixth unit.

E

M

E

"'

QJ

C.

E

&'l

"' QJ

C.

E

&'l

"'

r--

00

QJ

QJ

QJ

C.

C.

C.

.

E

"'

E

"'"'

E

&'l

Figure 7 . Odour removal performance of chemical scrubber system.

H2 S concentration was less than 50 ppb (the minimum detection limit on the H2 S analyser). During commissioning, peaks of up to 37 ppm H2 S were recorded on the inlet analyser, wh ich was significantly greater than the design peak concentration of 17. However, the chemical scrubbers still consist ently achieved an outlet H2 S level below 50 ppb despite the increased inlet odour levels.

Eight additional grab samples were col lected over four days at the chemical scrubber outlet and analysed at a NATA certified laboratory. All test results had an H2 S concentration

The Stage 2 chem ical scrubber airflow co uld not be safely drawn throug h the odour control system as not all Stage 2 sources are connected. Th erefore, the Stage 2 airflows were simulat ed by balancing the airflows to bias Train 1 for Days 1 to 6 and then bias Train 2 for Days 7 to 10 to achieve the desired design airflow of 96,500 m3/ hr through one trai n with a tot al airflow of 125,000 m 3/ hr th rough the system .

H 2 S removal The biotrickling filter consistently achieved the target H2 S concentration discharge of less than 0.5 ppm for an average inlet H2S concentration of 21 ppm and up to a maximum of 44 ppm, as shown in Figure 4. Eight additional grab samples were co llected over four days at the biotrickling filter inlet and outlet and analysed at a NATA certified laboratory in accordance with Standard Method 4500-S2 D. The total sulfide concentration at the biotrickling fi lter discharge was below the minimum detection limit of 25 ppb for all samples, resulting in greater than 99 .8% H 2S removal when compared with the observed inlet concentrations ranging from 15 to 29 ppm H2 S. In terms of surge load handling , the biotrickling filters performed well and outlet H2 S remai ned below 0.5 ppm during the initial rampi ng up period when odour loads were rapidly increased from less than 30 ppm H2S to over 80 ppm. The observed inlet H2S loads have been significantly lower than the design basis established from the comprehensive odour sampling program.

As shown in Figure 5, the chemical scrubbers have consistently achieved the design target of less than 100 ppb, except during planned shutdowns for commissioning activities or mechanical interruptions. The majority of the ti me, the outlet

A ,x water NOVEMBER 2010 95

odour management less than 60 ppb, with the exception of two samples that were considered nonrepresentative due to sample contamination. This confi rmed the chemical scrubber system met the performance guaranteed discharge of < 100 ppb H2S. Odour removal The odour removal from the biotrickling filters was determined from 10 grab samples, at both the inlet and outlet , over a four day period. The inlet odour concentration averaged 104,400 OU with a maximum of 185,400 OU and a minimum of 61,200 OU. The outlet odour level was always below the performance guaranteed level of 5,000 OU and ranged between 1,000 OU and 3,200 OU. This resulted in an odour removal efficiency above the performance guaranteed 95%, with the system achieving 95.6% to 99.4% odour removal during the test period. Figure 6 presents the biotrickling filter odour removal performance.

It should be noted that two laboratories were used for the biotrickling filter odour performance testing. While both labs had similar outlet odour levels, the inlet odour levels were consistently higher as measured by Lab B. This highlights the variability in odour measurement, and its potential impact on contractual guarantees.

Good odour reduction was also observed during the chemical scrubber performance tests as shown in Figure ?,with an average discharge odour of 921 OU. While three of the eight odour samples exceeded the discharge odour target of less than 1,000 OU, given the subjective nature of odour assessment and associated degree of uncertainty the system was considered to achieve a satisfactory level of odour reduction.

The NATA accredited laboratory that analysed the samples provided the 95% confidence level for the samples they analysed under method AS/NZS 4323.3:2001 at +21/ -55% as shown on Figure 7. This confidence level results in a minimum odour level of less than 1000 OU for all samples.

It is not unusual to have high variability with Olfactometry analysis due to the method relying on 'calibrated' human noses. Some of the outlet odour samples were charact erised as a chlorine odour while others were described as a sewage smell.

A 28 day proof running period was completed for both the biotrickling filter and chem ical scrubber systems to

96 NOVEMBER 2010 water

confirm that the systems provided reliable operation and performance during normal conditions. Both systems operated in a stable manner during the period, with no significant upsets affecting the systems performance.

Performance One Year Later The biotrickl ing filters have generally reduced H2S levels from 30 ppm to less than 0.2 ppm, although there have been events with higher discharge odour levels during mechanical or power supply failures.

The chemical scrubbers have generally reduced the H2S levels from 2 to 10 ppm to less than 50 ppb. The target discharge odour levels of < 100 ppb have been exceeded for short periods of time on three occasions during the last year due to mechanical failures, but the syst em's flexibility and attentive operations and maintenance teams have minimised the duration of these excursions.

The system reliability over the last year has been 96% as calculated by the up time of the extraction fans and recircu lation pumps. While most months have achieved 99% availability, the system uptime has been impacted by the connection of new infrastructure in the sludge treatment area and incorporation of a standby generator system. The odour control system is expected to met the rel iability target of 99.9% in future, once the current program of upgrade works in completed.

The new covers on the SBR bioseletor, grit tanks and digester scum boxes have been highly effective, with high negative pressures consistently achieved .

However, the improved sealing of the existing primary sedimentation tank covers have not effectively contained odours due to the frequent need to remove the covers for maintenance activities.

Conclusion Part 2 A combination of biotrickling filters (treating high odour loads from the preliminary, primary and sludge treatment areas) and chem ical scrubbers (polishing the biotrickling filter discharge and treating the low odour loads from the SBR) were implemented to provide a sustai nable solution for treating the multiple odorous air streams. Both systems reliably met their performance requirements for odour treatment, achieving the key system goal of < 100 ppb H2S and < 1000 OU for the discharged treated airstream.

~

refereed paper

Generally the community has been satisfied with the reduction in odour emissions, with no odour complaints received in the last four months (to June 2010). Reports of odour events declined by more than 90 per cent as a result of the odour upgrade, and the residual odour is considered less annoying.

Overall Conclusion These t wo projects have shown that control of odour emissions from large wastewater treatment plants, situated close t o residential areas, can be minimised by using appropriate ventilation rates, containment systems and odour treatment processes. The treatment systems installed at Perth's metropolitan WwTPs represent international best practice, where financial, environmental and social impacts have been minimised for the life of the odour control system.

The Authors Cora Nichols (cxnichols@ skm.com.au) is a Senior Project Manager and Process En gineer in SKM's Perth office. Sharon McNeil is a Process Commissioning Engineer in the Water Corporation's Process Expertise Group. Gayle Van Durme is an Odour Management Consultant with Black & Veatch in Kansas City, USA. Keith Cadee is the General Manager of the Water Corporation's Acquisition Group.

Dr Ian Wallis is the Director of Consulting Environmental Engineers (CEE) in Melbourne. Angus Adams is a Process Mechanical Engineer with SKM in their Perth office.

References Cadee, K. and Wallis, I. 2007. Odour Containment and Ventilation at Perth' s Major WwTPs. Water, Journal of the Australian Water Association (Mar 07) 54-60.

Nichols, C.; Meyer, J.; Van Durme, G.; Cadee, K. 2009. Building Confidence in Large Scale Activated Carbon Systems. OzWater, Australian Water Association, Melbourne, VIC, Australia.

Nichols, C.; Wallis, I.; McNeil, S.; Adams, A.; Van Durme, G.; Cadee, K. 2010. Woodman Point WwTP Odour Control System Upgrade Complies with Rigorous Odour Emission Reduction Requirements. Odour Specialty Conference, Australian Water Association, Sydney, NSW, Australia.

Van Durme, G.; Nichols, C.; Cadee, K. 2009.

Biotrickling Filters Reduce Odour Control Costs at Woodman Point WwTP. OzWater, Australian Water Association, Melbourne, VIC, Australia.

Van Durme, G.; Shaw, A.; Nichols, C.; Cadee, K.; McNeil, S. 2009. Biotrickling Filter and Chemical Scrubber Combination Meets Perth's Stringent Odor Control Requirements.

WEFTEC, WEF, Orlando, FL, USA.

tee h n i ca I features

~

international projects

refereed paper

SUSTAINABILITY ASSESSMENT TOOL FOR WATER TREATMENT TECHNOLOGIES IN RURAL COMMUNITIES N Cocks, R Kurup Abstract Many water and sanitation projects, aided by donor organisations, in developing countries and rural regions in developed nations have failed withi n a few months of commissioning. It may be attributed as a result of decision makers' failure to address the characteristics and needs of receiving commun ities.

Sustainability assessment tools can assist stakeholders in selecting appropriat e t echnologies by addressing the receiving community's features and requirements, and thus assisting in ensuring a potential project's long-term sustainability.

This paper deals with a sustainabilit y assessment tool developed for t he selection of appropriate combinations of pre-treatment and filtration technologies for water treatment. The tool was applied to the Tenganan Wat er Supply Project, which was undertaken by Engineers wit hout Borders to improve the Tenganan community's water supply.

Applying relevant criteria resulted in a short list of appropriate treatments, wit h a recommendat ion that the combination of pre-treatment and slow sand filter wou ld be a better solution to add ress the fluctuations in water quality which are expected to occur between t he wet and dry seasons in Tenganan.

Introduction The selection of appropriate t echnologies for rural communities is a difficult task (Loetscher & Keller 2002), as t he sustainabi lity of water treatment technologies in these regions is affected by a range of social , economic, environmental and technical factors (Brikke & Bredero 2003; Parr & Shaw 1996; Smet & Wijk 2002). Failure to recognise t he req uirements and characteristics of receivi ng communities, as well as their abi lity to operate and

pasar

50 km

~..,.,

Figure 1. Map of Bali illustrating the location of Tenganan (Bowen 2006). maintain technologies, has seen the fai lure of many water supply project s.

Add itionally, many projects have failed as a result of a lack of consultation with t he receiving community, an insufficient amount of technical assistance being provided to the community, or the community not taking ownership of the implemented technology (Brikke & Bredero 2003; Mukherjee 1999).

One way to assist in reducing the pot ential fai lure of water t reatment projects within rural commu nities is th rough the utilisation of a sustainability assessment tool to assist stakeholders in t heir decision making processes (Loetscher & Kel ler 2002; Palaniappan, Lang & Gleick 2008). Such tools enhance and support the judgement of decision makers and hence, improve the effectiveness of decision making processes as well as t he likelihood of a correct decision (Loetscher & Kel ler 2002; Struss et al. 2004).

Developed during a project in Tenganan, Indonesia.

Not only do t hese tools assist in t he selection of appropriate technologies based on a commun ity's requirements and characterist ics, they also facilitate the participatory involvement of the receiving commun ity (Loetscher & Keller 2002). This assists in identifying the receiving community's ability to operate and maintain technolog ies and capacity building requirements, as well as giving t hem a sense of ownership of the project (Bri kke & Bredero 2003). All of these benefits aid in ensuring t he sustai nability of implemented technolog ies in rural comm unities.

Tenganan Water Supply Project A sustainability assessment tool was developed to assist in the selection of an appropriate combination of pre-t reatment and filtration technologies for water treatment for the Tenganan Water Supply Project undertaken by Engineers without Borders. The Tenganan community is locat ed in one of the poorest and driest regions in Bali, Indonesia (Figure 1); and as t he local community depends on rain driven surface water bodies for water supply, during extended drier periods,

water NOVEM BER 2010 97

~

international projects the community does not have enough water of a sufficient water quality to meet basic human requirements (Third et al.

2008). The Tenganan Water Supply Project, initiated by Engineers without Borders aimed to improve t he wat er supply of the Tenganan community.

Table 1. Comparison of water quality test results and Indonesian drinking water guidelines.

Parameter

Value

Indonesian Drinking Water Guideline Limits

7.1

6.5 - 8.5

8

N/A

pH Chemical Oxygen Demand (mg/I)

r e f e r eed p ap e r

combinat ion of pre-treatment and filtration technologies for water treatment) as wel l as through household scale t reatment, mainly disinfection by boili ng (Micenko et al. 2008).

Household scale treat ment water processes such as N/A boiling (which is the method 50 currently adopted by the 1.904 1.5 It was envisioned that community for the treatment 0.3 Iron (mg/I) 0.193 of potable water) are the quantity of water would Manganese (mg/I) 0.01 0.1 be increased by extracting generally less expensive and Nitrite (mg/I) 0.017 3 water from a more reliable easier to operate and 124.84 500 Hardness (mg/I) source, Babi Tunu dam in maintain than communitythe Buhu River (Figure 2). scaled alternatives, though Turbidity (NTU) 2.93 5 poor hygiene during The quality of the water Colour (TCU) 17.1 15 collection, handling and collected during dry season Thermotolerant Coliforms (MPN/100ml) 94 0 storage can result in t he reand t he Indonesian Drinking Fluoride (mg/I) 0.01 1.5 Water Guidelines Li mits are contamination of the treat ed provided in Table 1. wat er (Montgomery & from contaminated catchments and Elimelech 2007). This aside, an additional Although most of the parameters are increased suspended solids. within the limits, ammonium nitrogen and st udy was also undertaken as part of the Thermotolerant Coliforms exceeded the Tenganan Water Supply Project to The quality of the community's water limits. Due to limitations of funding, only investigate the appropriateness of a supply would be enhanced throug h treatment of the source water at a a single sample was collected and number of other water t reatment commu nity-level t reatment facility (which technologies which cou ld be analysed. The quality is expected to implemented in the household level worsen during rainy season due to runoff was to be comprised of an appropriate (Borovac, 2009).

Dissolved Oxygen (mg/I) Total Phosphate (mg/I) Nitrate (mg/I) Ammonium - N (mg/I)

6.2 3.164 2.847

N/A

Methodology The development of the sustainability assessment tool consisted of three main processes: • Initial identificat ion of appropriate wat er treatment technologies for assessment; • the development of sustainability assessment criteria; and • the evaluation and final selection of water t reatment technologies using t he developed assessment criteria.

The technologies which were init ially identified for assessment were those wh ich met t he comm unity constraints of the Tenganan Water Supply Project.

These constraints included the following parameters: • t he use of no mechanical equipment; • no use of chemicals;

- .__--,g,,_.,..._ ~ .......~ ,.,.,._ .., (~ l ......,.0... .....

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

C -....... T,._ S.

•

low capital, operating and maintenance costs; and

• minimal maintenance requirements.

100m Surlac• Contours

eon....-

...'",.,""

0

500

1,000

-_ ,,..

The identified tech nologies which met the community constraints of the project included both pre-treatment processes and main filtration processes. The sustainability assessment tool was then used to further analyse the different combi nations of the pre-treatment and main filtrat ion processes to select t he

Figure 2. Schematic of the existing and proposed water supply system.

98 NOVEMBER 2010 water

technical features

international projects

refereed paper

most sustainable water treatment solution. Analysis was also carried out for the sustainability of each of the selected filtration technologies without a pre-treatment st ep as well as the option of having no community-scale water treatment (i.e. the tool evaluated if the raw source water would meet the relevant standards without any treatment other than disinfection in the household.

applicable, what conditions could be imposed on the treatment system to make it appropriate for implementation.

selection Crl~ria

Tec-hno~rv Selr ctlon

Results and Discussion

Pre-Treatment

Te chnoloav

Technologies selection

In ue:ncln& F1ctor s

calculatk>n In order to adequately utilise the lnouts developed sustainability assessment tool , relevant data relating to the site characteristics and the selected pre-treatment and filtration technologies were col lected. Data collection methods CO lnatton Co Jnatlon Analysis tnapprocriate ACIIUOOfllte included water quality sampling and laboratory testi ng; soil sieve testing; stakeholder consultation; Imposed COn dJUon(sJ an array of engineering design calculations including hydraulic, Figure 3. The flow diagram of the sustainability process and cost calculations; and assessment tool. hydraulic modelling. The collected data, sustainability assessment pre-treatment and no water treatment) for criteria and selected water treatment evaluation. Upon selection of a technologies were then integrated combination of pre-treatment and together to shape the sustainability filtration technologies, the tool utilises the assessment tool. The flow diagram of the relevant data collected and a number of sustainability assessment tool is shown in appropriate algorithms to assess the Figure 3. selected water treatment system 's ability

to f ulfil each assessment criterion, with a quantitative and/or qualitative output for each criterion created and displayed on the interface. If any criterion was not met, the combination of technologies was deemed inappropriate for implementation, with the tool generating a prompt to inform the user of this conclusion or if

The sustainability assessment tool was developed using a visual basic enhanced excel spreadsheet to facilitate a simple and interactive user interface (Figure 4) which allowed users to compare a combination of pre-treatment and filtration technologies (as well as the options of a fi ltration technology without

,......... _.......... .._,_4,.......

Stttcc• Prf(tHtl'MtltT~

5-«•F-.~T~

I

a...

°"""'"

I

C rltedoA to ~ form CWaM 1asb AY~-..oltook"rtouirt-dfOf()ldl,1

c-.

Avalllbilt•olnwtffllb:r <df«OOM r-su1t....,1«dtdrtd dtt'Mnd a~......z .., I« lnnutnt W.Ctf CMfac<triJtics • - - . c , f0tclmM•lnd-..,.,_,.,oh¥acttflsUCs

MinlmtJmHIOl~Ht..d--fflW'f'll:s{m} MlnlrrunLWMIAl••-•tmH11:s rm2r

-.-.,Cost, rsus1 ....____,.•Mwlt~Costs

.

1:1

Pre•TreM.-..t TM_ _.__

Flltr.wkMI T H ~ V at" Treat....t Scst . . C1ke1ioll MM? Y••. VithC.1atc«111n....--.... Y H , I J i t h ~ d • ~ YH YH YH , Vl t h C . - d ~ Yts.V1th C - d 1 L » Y•r YH YH,VllhC-..-,o.wtoonwnr. Yts,VJth~K«,~ YH YH Yu YH Yu Y• r YH Y•r YH Y•• YH YH YH Yu

...

1

... ~

"

n40•m& 22 - 31

)4571 . 311570 2'1 - 531

115711-3*'

.......

...

7.1

7.1- 7.I

oo•-•

u

U - '-2

EffklitntWMffrlUMitt

•

coo•~

, ,...,.,.1

Totlil----........ ..

"""· -·-· ,.,,,-

·-.,_..,rnn111

Nukt'~ HM-dnt-ssftsC.co311ffiM11 Tw>ldkdNTU1

coeowtTl"'tn

-·-·

ThHmocol«wColfo,,ns,---..n

.....

2.t47

..uo,...

0.01 0.017

,

2.147

1

313 . 670

YH YH

7,1- 7,I ,. 8 U - S2 i .., 2.t47

Yu NIA NIA NIA Yu

1.$04

L904

.ott -.m

,ot, • . DS

.001 • .007

.001 - .007

0.017

0.017 1204

12, ...

1.2$-2.315

12U4 ."4 - lt51

12.995 - IUS . ... 59.22 0.01

7.798 - 11.&42

7,7$8 • tl542

O-U'1

0 - 2.961

001

0.01

0.00,

.OOOOS • .0007

.00005 . .0007

."4-

use

Alstnlo 1- - .

Lt.td ' --..1

YH Y•r

,.,

Yu Yu Y• r Y•• Yu YH

'"'

Yu NIA YH

Figure 4. The user interface of the sustainability assessment tool with an example output.

Table 2 highlights the technologies which were reviewed and evaluated against the com munity constraints of the Tenganan Water Supply Project.

It was determined that primary sedimentation as well as the gravel filtration tech nologies (dynamic, upflow in series and upflow in layers) were appropriate pre-treatment technologies. Additionally, the combination of coagulation, flocculation and sedimentation was concl uded to be an appropriate pretreatment technology if natural coagu lants (i.e. Moringa olefiera seeds} were utilised and the system consisted of hydrau lic mixing devices (provided there is natural hydraulic head; without the use of any mechanical pumps) to facilitate rapid mixing and gentle stirring for coagulation and flocculation respectively Furthermore, downflow and horizontal rough ing filtration were reviewed and not selected for evaluation as upflow filters are generally technical ly and economically preferred over these configurations (Smet & Wijk 2002; Visscher 2006; Wegelin 1996). A schematic diagram of each gravel filtration technology is il lustrated in Figure 5.

The filtration technologies for water treatment w hich met the com munity constraints criteria were slow sand filtration and reed bed treatment system s.

On the other hand, the filtration technolog ies wh ich did not meet the commu nity constraints of the project were rapid sand, diatomaceous earth and membrane filtration and ion exchange due to the high costs and maintenan ce requ irements associated with a community scaled treatment unit, as wel l as the need for chemicals and mechanical equipment for the technology's operation.

These technologies require backwash in order for cleaning. This wou ld result in intensive maintenance requirements and the need for a pumping system to obtain the high pressure required for the backwash. Such requi rements would also inherently add to the capital, operating and maintenance costs of the project.

Sustainability assessment criteria As previously mentioned, the sustainability assessment criteria were developed from a range of social ,

water

NOVEMBER 2010 99

international projects economic, technical and environmental factors which are shown in Table 3.

These c riteria were used to assess each wat er treatment system's suitabi lity for implementation in Tenganan , w ith a water treatment system deemed inappropriate if any of these criteria was not met. On t he contrary, if the criterion relating to the evaluat ed water treatment system 's ability to produce an appropriate treated water quality was t he only criterion which was not met, the evaluated wat er treat ment system was inferred to be appropriate so long as the water cou ld be disinfected at the household scale to meet the required standard .

Application of the Tool for the Tenganan Water Supply Project The developed sustainabil ity model was applied for t he Tengangan water supply project. The model provided a number of viable alternatives and identified the technologies that did not meet with t he technical, geophysical, environmental, economic and social constraints of the project.

Based on the field survey of commun ity's human capital , and assessment of t he operation and maintenance (0 & M) requirements of t he identified processes/technologies , the ability of the community to perform the tasks associated with O & M of each technology has been estimated and presented in Figure 6. A rank ing of 1 signifies that the tasks required for the 0 & M of the tec hnology can be performed by the commu nity; 2 signifies that t hat the commun ity would be able to perform the tasks required for the O & M of the technology if capacity bui lding initiatives were undertaken to teach the community how to undertake the tasks; wh il st 3 denotes that the comm unity would not at all be able to perform the tasks required for the O & M of the technology. A technology was considered to be appropriate so long as a ranking of 3 was not awarded. The rankings assigned to each of the technologies were confirmed by the key stakeholders of the project.

The estimated capital cost of each assessed technology is presented in Figure 7 (a) and the estimated O & M costs for each evaluated technology are illustrated in Figure 7 (b). The capital, operation and maintenance costs for each technology were based on specific cost fig ures (usually expressed as per m 2 or m 3) from water supply projects undertaken throughout the developing world and were corrected to reflect current day prices.

100 NOVEMBER 2010 water

refereed paper

OY<,r_

L 1nn....1 -

------t. .

Un....t

"cf>'

Onolu&• Dynamic gmd filter

Vpflow 9f'QWI fil1v i11 lo.yvs

L

lnfu.nt - ~

=e> -

I -.

•

0

~ I

rniu.n,

OfllNigr

Downllow gnY<I filter In SffitS

0.•irvgc

· · ~ C>ownflow 9f'CIWI f flfv in lo.yvs

OrA.na&c

Horizontol !low gr•vd fiker

Figure 5. Schematic diagrams of each gravel filtration technology (Smet & Wijk 2002; Wegelin 1996).

Table 2. Selection of water treatment technologies using the criteria of the community constraints.

Technology

Low No Mechanical Low Capital, No Maintenance Equipment Operation and Chemical Requirements Required Maintenance Costs Additives

Primary Sedimentation (PS)

Yes

Yes

Yes

Dynamic Gravel Filtration (DyGF)

Yes

Yes

Yes

Yes

Upflow Gravel Filtration (in series) (UFGS)

Yes

Yes

Yes

Yes

Upflow Gravel Filtration (in layers) (UFGL)

Yes

Yes

Yes

Yes

Yes 1· 2 Yes

Yes 1

Yes 1, 2

Yes2

Yes

Yes

Yes

Reed Bed Treatment Systems (RBTS)

Yes

Yes

Yes

Yes

Rapid Sand Filtration

No

Yes

No

Yes

Diatomaceous Earth Filtration

No

No

No

Yes

Coagulation, Flocculation and Sedimentation (CFS) Slow Sand Filtration (SSF)

Yes

Notes: 1 - Using hydraulic mixing devices tor rapid mixing and gentle stirring tor coagulation and flocculation respectively.

2- Using natural coagulants (i.e. the seeds from the Moringa oleitera plant).

Table 3. The sustainability assessment criteria.

No

Factor

Criterion

Social

Is the community able to perform the tasks required for the operation and maintenance of the system?

2

Technical

Is the system feasible given the community's water demand?

3

Technical

Are tools available in the community to operate & maintain the system?

4

Technical

Are materials available in the community to operate & maintain the system?

5

Environmental

Is the system appropriate for the source water characteristics?

6

Technical

Is the system capable of producing an appropriate water quality? 1

7

Financial

Is the community capable of paying for the capital costs of the system?

8 9

Financial

Is the community capable of paying for the operation and maintenance costs?

Environmental

Is the technology appropriate for the climate?

10

Technical

Is land available for the treatment facility? 2

11

Technical

Is the hydraulic head available in the distribution line?2

Notes: 1 - Using the water quality parameters of turbidity, pH, colour, dissolved oxygen, chemical oxygen demand, fluoride, lead, arsenic, iron, manganese, thermotolerant coliforms, hardness, ammonium, nitrate, nitrite and phosphate as a gauge.

2 - Given that a location tor the water treatment facility has been already been proposed.

The reed bed treatment technology was considered inappropriate as the capital costs of the technology exceed

the community's financial capacity and the land area required exceeds the space available at the proposed location. In

technical features

~

international projects

refereed paper

addition, there are not many case stud ies of application of reed bed technology for treatment of drinking water. Additional ly, the combinat ion of upf low gravel filtration in series and slow sand filtration was determined to be inappropriate as the capital, operation and maintenance costs associated with the combi nation may be beyond t he commun ity's financial capacity.

T he Ab ility of the C o mmunity to Pe rform t he 0 & M T a sks for e a ch T e chnology Gravel Filtration (Dynamic, Upflow in Layers & Upflow in Series) 3

The remaining pre-treatment and filtration technologies were considered to be viable options, w ith t he viable pretreatment and filtration technology combinations including:

Reed Bed Treatm ent Syst ems

Primary Sedim entation

• Dynamic gravel filtration and slow sand filtration;

-

• Primary sedimentation and slow sand filtration.

• Slow sand filtration without pret reatment.

Incidentally , the model also predicted that t he raw water could be provided so long as the elevated levels of ammon ium N, colour an d thermotolerant coliforms wit hin the source water can be removed by household-scale treat ment processes.

It essentially indicates that no centralised pre-treatment is required. The study had on ly one set of water quality dat a collected in t he drier period as an input to

Assign ed Ranking

Figure 6. The community's ability to perform the t asks associated with the operation and maintenance (0 & M) of each technology.

Figure 8 shows the number of criterion met by each of assessed water treatment technologies. The treatment options that have satisfied 10 out of the 11 criteria presented in Table 3 are considered to be appropriate for implementation in Tenganan.

The o nly shortcomings of these options are their ability to produce an effluent water q uality with acceptable concentrations of ammonium N and thermotolerant coliforms. However, t hough the slow sand filtrat ion technology's ability to remove ammonium N was not quantified in t his study, based on documentat ion that slow sand filtration is effective at reducing nutrient levels (Clapham 2004). it is likely that this technology would have the abi lity to reduce the ammonium N content in the source water. Although any removal of microbiological contami nants by any of t he evaluated tec hnologies is thought t o be beneficial, these technolog ies are not designed to completely remove microbiological contaminants; this is the pu rpose of disinfection technologies wh ich wi ll be adopted o n a household scale.

Coagulation, Flocc ulation & Sedimentation

Slow Sand Filtration

• Upf low gravel filtration in layers and slow sand fi ltration and;

Capital Costs of Each Technology 100000

iii ::i

80000

80000

60000

60000

40000

40000

20000

20000

-

Minimum Estimated Capital Costs ($US)

-

Maximum Estimated Capital Costs ($US)

-

Estimated Funds Available Towards the Capita l Costs ($US)

~

..."'

0

u

0

0 UGFL UGFS DyGF PS

(a)

CFS SSF RBTS

Technology

O&M Costs of Each Technology 1200 1000

...

1200

I

1000

"'CV

800

800

::i

600

600

0

400

400

200

200

> .......

V, ~

..."'

u

0

0 UGFLUGFS DyGF PS

(b)

-

Minimum Estimated O & M Costs (SUS/year)

-

Maximum EstimateO& M Costs (SUS/year) -Funds Available TowardsO &M Costs (SUS/year)

CFS SSF RBTS

Technology

Figure 7- (a) The estimated capital costs of each assessed technology and (b) The estimated O& M costs for each evaluated technology (abbreviations are listed in Table 2).

water

NOVEMBER 2010 101

international projects the model. Due to the paucity of the water quality data, the temporal variation of the source water quality was not quantified. Additional ly, since this set of water quality data was taken in the dry season, it is also unlikely that this data is representative of the source's worst water quality, which is hypothesised to occur during the wet season . In the wet season, it is likely that increases in rainfall would increase the amount of particles and microbiological contam inants which are washed int o the source water, increasing the level of contamination in the water body (LeChevallier & Au 2004).

Consequently, it is believed that any of the viable pre-treatment and slow sand filtration combinations would be more appropriate than slow sand filtration alone or no centralised water treatment, as such options would be better suited to meet the fluctuations in raw water quality (Galvis, Visscher & Lloyd 1992).

Conclusion Simple, low cost water treatment technologies are effective means for improving water quality in rural communities and inherently, improving public health and economic development in such regions. The use of sustainability assessment tools for the selection of water treatment technolog ies can assist in achieving these benefits by addressing a commun ity's attributes and facilitating participatory involvement. For these tools to be successful however, it is important that they assess the technical components of potential technologies as well as the social, economic and environmental aspects of receiving comm unities.

Although the tool developed in this project has been tailored specifically for Tenganan, the t ool may be useful for future projects of a similar nature in developing nations as well as remote regions of developed countries by donor agencies including AusAID (AusAID 2005), the Asian Development Bank and the World Bank and volunteer organisations such as Engineers without Borders.

It should be stressed that although the use of sustainable water treatment tech nologies can improve the quality of drinking water, the sanitation and hygiene in rural communities also needs to be addressed for improved human health.

The integration of improved water supplies, sanitation and hygiene greatly improves human health in rural communities and inherently, the quality of life in such regions.

102 NOVEMBER 201 0 water

refereed paper

Number of Criterion Met for Each Water Treatment Syste m

DyGF

11 10 No PreTrea tmen t

UFGFL

CFS

UFGFS

-+- RBTS

-

SSF

PS - . - No Filtration Treatment

Figure 8. The number of criterion satisfied for assessed water treatment system.

References AusAID 2005, Safe water guidelines for the Australian aid program 2005: a framework and guidance for managing water quality, Commonwealth of Australia, Canberra.

Borvac, C. 2009, Appropriate water treatment technology t hat can be applied on the household scale t o the developing community of Tenganan, Bali, Engineering Thesis, The University of Western Australia.

Bowen, J 2006, A multidisciplinary look at capacity development and its application through the characterisation of water resources in Tenganan, Indonesia., Engineering Thesis, The University of Western Australia.

Smet, J & Wijk, C (eds} 2002, Small community water supplies: technology, people and partnerships, IRC, Delft.

Struss, P, Roque, W, Salles, P & Heller, U 2004, 'Model-based Decision Support Systems - An Application to Water Treatment' , e-Environment: Progress and Challenge, pp. 47- 58.

Third, K, Fun, O.M., Bowen, J., Micenko, A., Grey, V. and Prohasky T. 2008, Rural Water Supply for Bali: Off-shore Design by Young Engineers, Water, 38 (4), pp 87-90.

Visscher, JT 2006, Facilitating Community Water Supply: From transferring filtration technology to multi-stakeholder learning, IRC, Delft.

Brikke, F & Bredero, M 2003, Linking technology choice with operation and maintenance in the context of community water supply and sanitation, WHO & IRC, Geneva.

Wegelin, M 1996, Surface water treatment by roughing filters: a design, construction and operation manual, SKAT, St. Gallen.

Clapham, D 2004, Small Water Supplies: A Practical Guide, Spon Press, New York.

The Authors

Galvis, G, Visscher, JT & Lloyd, B 1992, 'Multistage surface water treatment for community water supply in Colombia' , Waterlines, vol. 1 0, no. 3, pp. 26-29.

LeChevallier, MW & Au, K-K 2004, Water treatment and pathogen control: Process efficiency in achieving safe drinking water, WHO & IWA Publishing, London.

Loetscher, T & Keller, J 2002 , 'A decision support system for selecting sanitation systems in developing countries', Socio-Economic Planning Sciences, vol. 36, no. 4, pp. 267-290.

Nathan Cocks is a recent graduate from the University of Western Australia with a Bachelor of Engineering with First Class Honours.

Micenko, A, Hewett, J, Third, K, Hindle, C , Fun, M , Bowen, J & Karunaratna, K 2008, Tenganan Water Supply Upgrade, EWB, Australia.

Montgomery, MA & Elimelech, M 2007, 'Water and Sanitation in Developing Countries: Including Health in the Equation', Environmental Science & Technology, vol. 41 , no. 1, pp. 16-24.

Mukherjee, N 1999, 'Measuring sustainability recent lessons from Indonesia' , Waterlines, vol.

18, no. 1, pp. 13-16.

Palaniappan, M , Lang, M & Gleick, PH 2008, A Review of Decision-Making Support Tools in the Water, Sanitation, and Hygiene Sector, Pacific Institute, Oakland.

Parr, J & Shaw, R 1996, 'Choosing an appropriate technology', Waterlines, vol. 15, no. 1, pp. 15-18.

Dr Raj Kurup is Director of Environmental Engi neers International, a Perth based consultancy and Adjunct Professor at the School of Environmental Systems Engineering, The University of Western Australia. Email Raj.kurup@enviroeng ineers.com.au

technical features