Water Journal August 1990 by australianwater

Official Journal

water

ISSN 0310-0367

Volume 17, No.4, August 1990

AUSTRALIAN WATER AND WASTEWATER ASSOC IATION

CONTENTS My Point of View .. .. ... .... ............. .. .. ................................ Association News President's Message ... .............. ...... ...... .. ............. It Seems to Me .......... .. .. ...... .......... ............. ...... .. .... Guy Parker Award .. .......... ... .............. ..... .... .. .......... IAWPRC News .............................. .... ..... .. ... .. .... ............. .. Personalities .... ... ..................... .... ..... .......... ... ...... .. ..... ...... Industry News .. ..... ................ .. .. .. .. .. .... ... .... .. .. .. .. .............. AWRC - Water Technology Centre ..... .. ...... ........ Setting Site Clean-up Standards The Use of Risk Assessment T.L. Bulman .......................... ...... .. ....... ....... .......... ..... The Triumphs and Tribulations of Artifical Mixing in Australian Water Bodies T.F. McAuliffe and R.S. Rosich .... .................... Monitoring Filamentous Bacteria in Activated Sludge Plants E.M. Seviour and J.A. Sodde ll ......................... A Comparison of Upflow and Downflow Tertiary Filters for Sewerage Treatment M. Laginestra .. ........ .. ... ....... .. .. ....... ................... .. ..... BRN-1 - First Australian Conference on Biological Nutrient Removal ......... ............ .... ............ The Microbiology of Bulking and Foaming in Activated Sludge L.L. Blackall .. .......... ... .. ............... ...................... ... ..... Chlorine Dioxide- - The American Experience ....... ..... ..... ... .... .. ... ......... .... ... .. .......... .... ....... .....

FEDERAL SECRETARIAT PO Box 460, Chatswood NSW 2057 Facsimile (02) 410 9652 Te lephone (02) 413 1288 Office Manager - Margaret Bates

FEDERAL PRESIDENT Peter Norman Telep hone (08) 226 2249

EXECUTIVE DIRECTOR Peter Hughes Telephone (02) 410 9654

FEDERAL SECRETARY Greg Cawston Telephone (042) 29 0236

FEDERAL TREASURER John Molloy Telephone (03) 6 15 599 1

BRANCH SECRETARIES Canberra, ACT P. Cox , PO Box 306. Woden 2606 (062) 498 522

New South Wales Mr David Hope, PO Box 460. Chatswood 2057 (02) 269 5212

Victoria J. Park C I - Water Training Centre. PO Box 409. Werribee 3030 (03) 7 41 5844

Queensland D. Mackay. PO Box 412 , West End 4101 (07) 840 4844

5 6 6 6 12 12

Experience with Nocardia Scum in Activated Sludge J.A. Sealzi .......... ..................... .... ....................... ... ... .. 49 Determining Digester Active Volume - A Cheap Method D. Gregory .. ............... ..... .. ... .......... .................... .. ...... 50 Book Reviews ....... ...... .... ..... ..... .... ..... ..... .............. ............ · 54 Seminar Report .... ........ .... .. ............... ... ... .. ... ................... 55 Plant, Products, Equipment .. ... ....... ... ...... ................. 56

14 16

OUR COVER 18

The Septage Treatment Plant in Perth - Re-use of an old S.T.P.

The old sewerage treatment plant in Westfield has been decommissioned and converted for re-use in treating Perth's 130 ML per ann·um production of septage and grease trap waste. The refit was designed by Sinclair Knight ~ _: and Partners and is being managed onsite by Cleana..yay (Enviroguard). Process components include intial screening, screenings dewatering, grit and scum removal, lim e stabilization, sludge dewatering and effluent clarification. The final effluent must meet criteria of 3000 mg!L 800 5, 1400 mg!L suspended solids and 100 mg I L oil and grease for discharge to sewer. A pilot hybrid anaerobic bioreactor, operated by Campbell Environmenta l, is currentl y in place on the site to treat a portion of effluent for greater reduction of BOD levels and subsequent reduction of charges for sewer discharge. A second area of the site will handle non-hazardous industrial wastes through lim e stabilisation. (Further deta ils in June Issue, page 8)

30 34

44 48

South Aust rali a R. Townsend. Cl- State Water Laboratories , E&WS Private Mail Bag , Salisbury 5108 (08) 259 0244

EDITORIAL CORRESPONDENCE E.A. Swinton, 4 Pleasant View Crescent, Glen Waverley 3150 Office Phone and Autofax (03) 560 4752 Home (03) 560 9306

Western Austra lia A. Gale. PO Box 356. West Perth 6005 (09) 242 4677

Tasmania AB. Denne. PO Box 78A, Hobart 700 1 (002) 30 5562

Northern Territory D. Hardiment PO Box 37283 , Winnellie 0821 (089) 41 0 144

ADVERTISING Ann Sykes,

Applta, 191 Royal Parade, Parkville 3052 (03) 347 2377 Fax (03) 348 1206

PRODUCTION EDITOR J. Grainger, Applta, 191 Royal Parade, Parkville 3052 (03) 347 2377 Fax (03) 348 1206

WATER August 1990

SETTING SITE CLEAN UP STANDARDS THE USE OF RISK ASSESSMENT by Terri L. Bulman SUMMARY Cleanup standards may be set through the use of government guidelines, background concentrations, lowest levels achievable by technology and site specific risk assessment. Risk assessment offers some specific advantages which contribute to the selection and justification of cleanup standards for industrial sites. These include the evaluation of site specific factors, identification of critical information to be obtained during site investigation, assessment of the relationship between soi l concentrations, land use patterns and human health, estimation of the risk level associated with soil concentrations, cost/ benefit analysis of cleanup scenarios and estimation of uncertainty in the protection offered by cleanup standards. The ideal approach wou ld combine these factors with an understanding of what the public perceives as desirable cleanup standards, what criteria have been used and accepted in similar situations in the past , and what levels can be physically achieved by available cleanup methods. AERIS modelling provides a cost effective method of performing site specific risk assessment with substantial evaluation of the effects of uncertainty in risk assessment (such as human toxicology, site characteristics and human land use patterns). AERIS results have been used successfully in Australia to develop health-based soil cleanup criteria and identify critical exposure pathways for use in assessing soil cleanup criteria. Parameters such as soil porosity, soil saturated water content and contaminant distribution coefficients have been found to have a significant effect on soil criteria and should be determined as part of the site investigation process for sites which require extensive remediation .

INTRODUCTION Spills, leaks and the improper disposal of raw materials and waste products frequently contaminate soil and groundwater during operations at an industrial site. Once operations cease, the site must be decommissioned in a manner that will render it suitable for future use. Site decommissioning involves a number of steps, including: • defining the cleanup problem through a detailed assessment of contamination on the site, • determination of the cleanup goal, ie appropriate contaminant concentrations which can remain on site, • development of a strategy to achieve the cleanup goal, • implementation of the cleanup strategy and • assessment of the success of the cleanup through long term monitoring. Decommissioning is a complex problem because cleanup activities are specific to industry type, products and byproducts, the age of the plant, its location (geography, geology, hydrogeology and climate of the site), waste management practices and the proposed 'future use of the site. A critical step in cleanup of a contaminated site is the process of setting the cleanup goal - the cleanup criteria. Wherever industrial activities have caused on-site soils and groundwater to become contaminated and redevelopment to another use is being considered, there is the potential for future users or those located near the site to be exposed to substances in the soil and groundwater. Exposure can occur through various pathways, such as the inhalation of vapours, direct ingestion of soil and produce, dermal absorption, or the ingestion of local groundwater. The establishment of cleanup criteria must address those concerns in some way and must be justified to the local community. Criteria may be developed which consider exposure on a site specific basis. Alternatively, regulatory agencies can establish specific guidelines or regulations that identify concentrations of substances in soil and groundwater that are sufficiently low to avoid exposures that exceed acceptable levels.

FOOTNOTE· This paper is an edited and updated version of a paper presented to the 3rd National Conference on Hazardous Waste Management, November 1989. 18

WATER August 1990

Terri Bulmam is Sen ior Soil Scientist with Campbell Environmental, based in Perth. She previously spent seven years as a Research Scientist with Environment Canada, specialising in the fate and transport of environmental contaminants in soil and remediation of soil contaminated with petroleum products. She chaired the Technical Working Group which was responsible for the development of the AERIS risk assessment model. Terri Bulman

METHODS FOR SELECTION OF CLEANUP CRITERIA Methods for establishing cleanup criteria are summarized in Figure I and include exposure assessment, the use of background concentrations, previously established or published guidelines, and standards achievable by technology. At the present time, few juris- · dictions have established guidelines for acceptable concentrations of residual contaminants in soil and groundwater. Several jurisdictions, under pressure to clean sites to some level, have identified interim guidelines. The most well known example of interim guidelines are the 'A, B, C' criteria identified by the Netherlands National Institute of Public Health and Environmental Hygiene (RIVM, 1988) and adopted, in some form, by many countries. Because these criteria were intended to be applied to a wide variety of site characteristics, they were intended to be very conservative. They are based on concentrations found ip pristine environments or on analytical detection limits. Other criteria that are sometimes used are concentrations which are locally considered pristine or background. While established criteria can prove very useful in some situations, they provide an unide9tified and inconsistent level of risk over a variety of sites . The scientific basis for these criteria and the protection they offer for human health cannot be easily identified. OUTPUT

INPUT

~RANSPORT AND SITE CHARACTERISTICS, ATE PATHWAYS PROPERTIES OF POLLUTANT

I I

LAND USE/RECEPTOR

ACCEPTABLE EXPOS URE (ADI , VSD)

CONC. IN SOIL, WATER AIR, PLANTS

EXPOSURE ~ RISK ASSESSMENT ASSESSMENT

SO IL CRITERIA

SOIL CONC. VS EXPOSURE SOIL CONC. VS PREDETERM INED LEVELS PROPORTION OF EACH EXPOSURE ROUTE

Figure 1: Methods for Selecting Cleanup Criteria A Systematic Approach

A selection of established guidelines is presented in Table 1. Acceptance criteria are generally set as concentrations below which soil can be considered clean ('A' level). Investigation thresholds are generally set as concentrations above which soil is considered potentially contaminated and requiring further investigation ('B' level). This is often referred to as a 'trigger' or 'action' level for further investigation . A third level may be set ('C') above which soil is considered severely contaminated and site cleanup is required. In the Netherlands concept of the 'A, B, C' criteria, the 'A' criteria are used as standards for cleanup. Other jurisdictions use either 'A' or 'B' criteria for cleanup standards, or develop criteria on a site specific basis. In practical use, these established guidelines are generally used as screening tools, to establish whether serious contamination exists, rather than as the sole basis for cleanup standards.

Another possible approach is to select cleanup criteria based on what levels are achievable by the best available technology. This would result in cleanup to the lowest possible levels and in some cases may limit the extent of cleanup that is performed. With this · approach, however, the level of risk would still be unidentified and inconsistent for a variety of sites and, furthermore, the level of risk in many cases will not be proportional to the cost involved or the practicality of implementing the technology. An alternative approach to setting cleanup criteria is the use of exposure/ risk assessment. This approach is based on determining the likelihood of an adverse effect occurring in a receiving population due to the presence of contaminants which remain in soil after cleanup. Use of this approach assumes (1) knowledge of the risk level to a population for specific doses of the contaminant and (2) knowledge of the dose resulting from concentrations of the contaminant in soil through numerous and interconnected pathways. Use of the risk assessment approach may be limited by lack of knowledge in those areas. Although such information can be severely limiting for some chemicals and some complex geological terrains, sufficient information generally exists from the fields of toxicology and from detailed site assessment to use the risk assessment approach, at least as a tool in assessing cleanup criteria.

Table 1: Selected soil quality criteria current established in Australia, Europe and North America Netherlands 1 B A (Clean) (invest) mg/ kg

arsenic • cadmium• lead • cyanide I (free) benzene 0.05 PCB phenol 0.05 hydrocarbon

. .

mg/ kg

30 5 150 10

Canada 2 Ontario (Cleanup std) mg / kg

Australia NHMRc2 Quebec EPAV; (invest) EPAV B A (clean) (invest) mg/ kg mg/ kg mg/ kg mg/ kg

20- 50. 3-8· 375-1000.

10 1.5 50 I

0.5 I I

1000b

10000c

0.1 0.1 0.1 d 100b 100

30 5 200 10

0.5

0.01 0.05

I I

• - background concentrations or site specific • - depends on land use b - mineral oi l c - total oil and grease d - petrol

o.o;. 20

150d 1000b

50 5 500 10 I I I

1000b

100b

2 3 4

RIVM (1989) Monenco (l 989) EPAV (1989) NHMRC (1989)

CURRENT USE OF RISK ASSESSMENT The use of risk assessment involves an evaluation of the fate and transport of contaminants on the site and their interaction with site features such as climate, soil, groundwater, man-made structures and human behaviour patterns on the site. Routes by which human receptors may be exposed to contaminants are identified and cumulative doses from all exposure routes are estimated. Cleanup standards which will not exceed 'acceptable' doses (as defined by current toxicological information) may be developed through a series of manual worksheets or a computer program. The risk assessment approach is often used, but to date has several disadvantages. It involves extensive professional time to manually go through worksheets and results in voluminous documents that require individual review and acceptance by regulators. Evaluation of site variability and 'what if scenarios are generally not performed due to the level of effort required. Ideally the selection process for setting cleanup criteria would incorporate the best aspects of all the approaches discussed previously. A computer model, Aid for Evaluating the Redevelopment of Industrial Sites (AERIS) was recently developed for the Canadian Government to assist in performing site-specific risk assessment and compare the results to known background concentrations and published guidelines which reside in a resident database. The AERIS model was one component in the development of recently issued Canadian National Guidelines for Decommissioning Industrial Sites (Monenco 1989). This document specifies a twotiered approach for the development of cleanup standards which combines existing standards, background concentrations, analytical detection limits and results of previous cleanup experience (Tier 1) with development of site-specific cleanup criteria using risk assessment (Tier 2) . Tier 1 criteria are considered to be the most conservative and allow for the most sensitive (or unrestricted) land use. The option of the Tier 2 criteria development is generally applicable where contaminants are identified for which Tier 1

guidelines do not exist or where background levels exceeding Tier 1 guidelines occur. The AERIS model is a tool used by the federal and provincial governments of Canada to assist in developing Tier 2 criteria. It is also used by industry as part of internal environment audits to assist in evaluating liability due to contamination on industrial sites. In Australia, national cleanup guidelines are at an earlier stage of development. Notification criteria have been established by the Environment Protection Authority of Victoria (EPAV, 1989) for use in screening sites which must be submitted to EPAV review. Actual cleanup standards are developed on a site specific basis or using the Netherlands 'A, B, C' criteria. Draft guidelines for investigation threshold levels were developed by the National Health and Medical Research Council (NH&MRC, 1989) and released for public comment in 1989. The Australian and New Zealand Environmental Council (ANZEC) are currently in the process of developing unified national guidelines for cleanup of contaminated sites. The NH&MRC document has oeen merged with, that of ANZEC and released for public comment in July of 1990. These guidelines include soil quality objectives for a range of compounds. Specific levels for investigation threshold and cleanup standards are still being developed . Under these guidelines, it is likely that the risk assessment approach will also play a part in developing criteria where standards do not exist or site specific concerns must be addressed. The routine use of risk assessment as an approach for determining cleanup goals is gaining acceptance with several other nations as well. Risk assessment procedures are an integral part of guidance for conducting investigations of superfund sites in the US (USEPA 1988a, USEPA 1988b). Current activities in development of guidelines in the Netherlands include a two-pronged risk assessment approach for human health risk and environmental (biota) risk (R. van den Berg, pers. comm.) .

RISK ASSESSMENT USING THE AERIS MODEL The AERIS model was developed by the Canadian consulting firm, SENES Consultants Ltd, based on previous work which developed site-specific cleanup guidelines for two decommissioned oil refinery sites in Ontario, Canada (Ibbotson et al, 1987). Much of the information in the AERIS model was compiled from other sources. Examples include (I) typical chemical and toxicological characteristics of the chemical, (2) characteristics and behaviour patterns of the receptor and (3) equations which predict transport of the chemical and estimate the dose from specific exposure pathways. The AERIS model is unique in that this information is integrated into an 'expert system' framework to provide a consistent approach to establishing soil cleanup criteria . The acronym AERIS was chosen to help users remember its limitations and intended use, that of an aid for evaluating industrial sites. As an aid, AERIS is suited to identifying the factors that are likely to be major contributors to potential exposures and concerns at sites, or those aspects of a redevelopment scenario with the greatest need for site-specific information. Rather than recommending a single cleanup criteria, based on risk assessment, the AERIS model has been designed to provide a range of criteria according to site variability and proposed land use patterns, as well as available information on background concentrations and criteria from other jurisdictions. The intent is that all of these criteria be compared to assist in selection of the appropriate cleanup level. Levels which are achievable by vario us techno logies are available from other sources and would be used in designing a final cleanup strategy. Details of the algorithms, assumptions and data flow processing in the AERIS model have been well documented by the model developers (SENES, 1988 and 1989) and have been summarised previously (Bulman 1989). Figure 2 illustrates a simplified flow diagram of processes in the AERIS model. Site specific data are entered which relate to the properties of the contaminant, characteristics of the receptor (adult or child), the proposed land use (residential, agricultural, commercial/industrial, or recreational) and characteristics of the site's climate, soil and underlying geology. Additional input includes exposure levels which are considered 'acceptable' (result in no adverse effect or an 'acceptable' level of risk) according to toxicological data. Fate and transport pathway models are used to predict the concentration of the contaminant which will accrue in soil, water, air and plants. The dose of contaminant to which the receptor is exposed is calculated using the predicted concentrations of the contaminant in soil, water, air and garden produce and the intake by the receptor of each of these environmental media. The selected land use and receptor provide values for ingestion or inhalation of environmental WATER August 1990

media (for example, volume of air breathed, volume of water consumed, proportion of produce consumed which was grown on the site). The values can be manipulated, if desired, to create a.sitespecific scenario. Routes of exposure which are assessed include ingestion of dirt and dust, ingestion of garden produce, ingestion of drinking water and inhalation of vapours and particulate matter. These exposure pathways include distinction between active and passive behaviour, indoors and outdoors, in winter and summer months. Doses from individual exposure pathways are added together to predict a total dose in units of mg of contaminant per kg of receptor body weight per day. The total dose to the receptor is then compared to the user specified 'acceptable' dose. The concentration of the contaminant in soil is calculated which will result in human exposure not exceeding the 'acceptable' dose. Risk assessment in the AERIS model is based solely on human health effects. Other factors which may influence selection of the appropriate cleanup criteria, such as smell, taste, staining, tainting and toxic effects to wildlife are not considered. For many cases, the acceptable soil criteria based on human health risk will be higher than that which meet other environmental criteria. An additional feature of the AERIS model allows the user to estimate soil criteria that will meet established criteria for air, water or plant quality. The model output includes a recommended concentration of the contaminant in soil. A comparison of the recommended soil concentration with preclean-up and background concentrations and current regulatory or guideline criteria for soil is also provided. Water, air and plant concentrations which correspond to the recommended soil concentration for site cleanup are provided as well as current regulatory or guideline criteria for each of those media. The proportion that each exposure pathway contributes to the overall exposure is also provided. Each general path category can be further broken down according to types of human activities. This information is useful for developing a cleanup strategy which minimises risk by reducing the most significant exposure pathway.

While the primary objective of usiBg the AERIS model is to develop site-specific cleanup criteria, AERIS also identifies the factors that are major contributors to potential exposures and problems at a site and those aspects of a site redevelopment scenario with the greatest need for better information. This provides an indication of the extent to which remedial action is needed at a site. The AERIS model is intended to be used as one of several tools for developing appropriate cleanup criteria on a site-specific basis. Model results are to be considered in combination with other published criteria, local background concentrations and levels achievable using available technologies. In addition, results should be considered from several model runs which demonstrate the effects of sensitive model parameters. Model limitations Limitations to the use of the AERIS model are similar to those associated with other risk assessment methods. For many chemicals the exposure pathways are not completely understood. Acceptable risk is defined by toxicological data which may vary in the level of confidence that can be attributed to it. Most toxicological criteria provide a high level of conservatism through incorporation of high 'safety' factors. Actual validation (comparison of human exposure to human risk) for either toxicological criteria or risk assessment in general has not been rigorous due to ethical considerations and the complexity of an epidemiological approach. This limitation is likely to be overcome in the future as more information on health effects become known.

APPLICATION OF AERIS IN AUSTRALIA Campbell Environmental is participating with the Canadian Government and SENES in the further development and refinement of the AERIS model. In the course of that work, AERIS has been used for a variety of risk assessment applications in Australia ranging from evaluation of previously established state guidelines, evaluation of costs for cleanup and redevelopment of former industrial property prior to purchase, and comparison with riskbased criteria developed by other means (providing a 'second opinion'). A contract with ANZEC is currently underway to perform sensitivity analysis on soil cleanup criteria for JO compounds over a variety of Australian and New Zealand soil and climatic conditions and land uses. Recommendations will also be made regarding cleanup criteria for C3 to C >9 alkane hydrocarbons for use in evaluating cleanup of fuel spills.' Much has been learned through the use of AERIS in Australia. In particular, specific soil characteristics have an important influence on the acceptable level of contaminants in soil. Characteristics such as soil porosity, saturated moisture content and the adsorption coefficient of contaminants to soil are critical to understanding the fate of contaminants and should be determined routinely during site investigations.

CONCLUSIONS Figure 2: Flow Diagram of the AERIS Model

·Use of model results AERIS considers only those exposures that are experienced on site. Off-site exposure, such as those that might be experienced by people whose drinking water supply is down gradient of a site, are considered indirectly by comparing concentrations in air, water and produce with existing environmental criteria such as point-of1mpingement criteria for air quality and drinking water objectives. In many cases, air, water or plant quality will drive the criteria selection process, rather than human health effects. The AERIS model may be run to estimate criteria which will meet desired environmental quality guidelines. The AERIS model is intended to be run a number of times for each contaminant of concern, with differing input values. A major benefit to be derived from the use of the AERIS model is the ability to evaluate the site for many different scenarios, including different land uses and a variety of receptor characteristics. Thus the use of the AERIS model will allow for the assessment of cleanup criterion values in soil and groundwater associated with various levels of risk and various levels of confidence in input parameters. This capability should provide useful information in considering the approach to cleanup and re-use of the site and is highly relevant to the discussions that will take place with site developer and site vendor subsequent to the completion of this phase of the project. 20

WATER August 1990

Risk assessment offers some specific advantages which contribute to the selection and justification of cleanup standards for industrial sites. These include: • Evaluation of site specific factors • Identification of critical information to be obtained during site investigation • Assessment of the relationship between soil concentrations, land use patterns and human health • Estimation of the risk level associated with soil concentrations • Cost/benefit analysis of cleanup scenarios • Estimation of uncertainty in the protection offered by cleanup standards These factors must be combined, however, with an understanding of what the public perceives as desirable cleanup standards, what criteria have been used and accepted in similar situations in the past, and what levels can be physically achieved by available cleanup methods. Risk assessment therefore makes a valuable contribution to a criteria selection process which includes consideration of factors other than risk. Risk assessment must also be undertaken with caution, as many important considerations (such as human toxicology, site characteristics and future human use) are not known with certainty. Risk assessment therefore should always be performed with an assessment of the effects of uncertainty on the cleanup standard.

AERIS modelling provides a cost effective method for performing site specific risk assessment with substantial uncertainty analysis. AERIS results have been used successfully to develop health-based soil cleanup criteria and identify critical exposure pathways for use in assessing soil cleanup criteria. Appropriate use of the model, however requires a careful assessment of sources of uncertainty in input parameters. Parameters such as soil porosity, soil saturated water content and contaminant distribution coefficients have been found to have a significant effect on soil criteria and should be determined as part of the site investigation process for sites which require extensive remediation. It is anticipated that the AERIS model will be beneficial not only as an aid in setting cleanup criteria, but also in the site assessment process and in development of a cleanup strategy. A major benefit to be derived from the use of the AERIS model is the ability to evaluate the site for many different scenarios, including different land uses and a variety of receptor characteristics. This capability should provide useful information in considering the approach to cleanup and reuse of the site. Sensitivity analysis of model input parameters will also identify the most important site characteristics which should be evaluated as part of the site investigation process.

HOW IS YOUR

EFFWEMT TREATMENT!

ACKNOWLEDGMENTS The author wishes to thank SENES Consultants Ltd and the Government of Canada for the opportunity to participate in further development and use of the model and for provision of technical support.

REFERENCES BULMAN, T.L. (1989). Industrial site decommissioning - the AERIS model. Proc. 3rd National Conference on the Management of Hazardous Waste. (Australian Water and Wastewater Association, Melbourne, Victoria). EPAV (1989). Publication 269 - Interim Advice on Approaches to Determine Whether a Site is Potentially Contaminated. (Environment Protection Authority of Victoria, Melbourne). IBBOTSON, B.G. , GORBER, D.M., READES, D.W., SMYTH, D., MUNRO, I., WILLES, R.F., JONES, M.G., GRANVILLE, G.C., CARTER, H .J. and HAILES, C.E. (1987). A Site-Specific Approach for the Development of Soil Cleanup Guidelines for Trace Organic Compounds. Proc. 2nd Conf. on Environmental and Public Health Effects of Soils Contaminated with Petroleum Products. (University of Massachussets, Amherst, Massachusetts, USA). MONENCO (1989). National Guidelines for Decommissioning Industrial Sites (Environment Canada, Ottawa, Ontario) NH&MRC (1989). Australian Guide lin es for the Rehabilitation of Contaminated Land - 1989 Draft (National Health and Medical Research Council, Canberra). RIVM (1988). Leidraad bodemsanering, afl.4, november '88 (Rij ksinstituut voor Volksgezondheid en Milieuhygiene, the Netherlands) SENES (1988). Contaminated Soil Cleanup in Canada Volume 2 Interim Report on the 'Demonstration' Version of the AERIS Model' (Environment Canada Decommissioning Steering Committee, Ottawa, Ontario) SENES (1989). Contaminated Soi l Cleanup in Canada Volume 5 Development of the AERIS Model, Final Report (Environment Canada Decommissioning Steering Committee, Ottawa, Ontario) USEPA (1988a). Guidance for Conducting Remedial Investigations and Feasibility Studies Under CERCLA - Interim Final (United States Environmental Protection Agency, Washington, D.C .). USEPA (1988b). Superfund Exposure Assessment Manual (United States Environmental Protection Agency, Washington, D.C.).

Are you having difficulty meeting the latest effluent discharge standards? Supafto, the effluent treatment specialists can assist you in overcoming these problems. We carry out surveys, characterise the waste stream and design an effective treatment plant. Supaflo can also upgrade your existing plant. For further information about effluent treatment plant and equipment call us in Sydney or Perth.

Supaflo, meeting new standards.

~ :.

Don't delay! Call us now... we can help you.

ADVERTISING DISCOUNT Members of AWWA are entitled to a 10% discount for advertising placed direct through Appita Headquarters.

YOUR AWWA MEMBERSHIP WORKING FOR YOU!

.,,,,.JPAFI.D Supaflo Pty. Ltd. 1/14 Roseberry Street,

Balgowtah N.S.W. 2093 AUSTRALIA Telephone (02) 949 3011 FacslmNe (02) 905 5856 Contact: Mitchell Gordon Jonathan Trope

15/7 The Esplanade Mount Pleasant W.A. 6153

AUSTRALIA Telephone (09) 316 1966 FacslmUe (09) 3161952 Contact: Ian Arbuthnot Lara 89Âˇ ,37

WATER August 1990

The Triumphs and Tribulations of Artificial Mixing in Australian Water Bodies T F McAULIFFE AND R S ROSICH

ABSTRACT This paper reviews the current status of reservoir destratification in Australia, focusing on the effectiveness of artificial destratification in controlling biological and chemical aspects of water quality. It is concluded that, for the most part, Australian destratification applications have been of limited success. The applications succeeded typically in attempts to alleviate problems associated with colour and turbidity, and to a lesser degree problems associated with iron and manganese. It is in the area of algae control that destrati fication applications have been least successful.

INTRODUCTION Artificial mixing (or artificial destratification) is the process of adding energy to a water body to break down existing stratification, or more usually to prevent the onset of stratification. This mixing has been provided primarily by artificial aeration and to a lesser extent by mechanical mixers. The first attempt at artificial destratification in an Australian water body was in 1966 in Little Nerang Creek Dam in Queensland (Queensland Department of Local Government and Gold Coast City Council). This "air gun" system was first replaced by a form of variable offtake and then converted back to aeration in 1983. In 1974-75 experiments with destratification continued, using water jet and artificial aeration in a section of Eildon Reser voir in Victoria. The water column was successfully destratified in three hours, with the data obtained allowing the development of design criteria for future applications (Brown 1986, Burns 1977). These latter experiments were carried out under the guidance of the Rural Water Commission of Victoria (formerly State Rivers and Water Supply Commission of Victoria) and it was in Victoria where the first full applications of destratification were initiated. The late 1970s saw destratification systems installed in Tarago Reservoir and Lake Eppalock. From this stage destratification applications have spread throughout Australia. To date, destratification has been tested or implemented in about 60 different water bodies in Australia, with varying degrees of success.

WHY DESTRATIFY? Stratification may lead to reduced water quality in a variety of ways, with the two main impacts being increased algae growth and / or an increase in the release of substances from the sediment. Algae populations may increase with stratification usually as a result of either being able to trap more light, by staying at or near Âˇthe water surface, or by having access to an increased amount of nutrients as a result of stratification or overturn. Particular types of algae may adversely affect water quality. This is of special concern in water supply reservoirs due to the release of toxic as well as odorous substances. Stratification may provide these algae with a competitive advantage over other types. The second general way in which stratification reduces water quality is through enhancing the release of undesirable substances from the sediment. Such release is usually associated with a decrease in the concentration of dissolved oxygen in the bottom water layer (hypolimnion). During mixed conditions oxygen is brought to the sediment by water that has previously been aerated at the surface. Stratification prevents this from occurring, mixed and aerated water is confined to the uppermost layer (epilimnion), while the hypolimnion has a limited amount of dissolved oxygen trapped in it at the time of stratification. Without light this oxygen is steadily depleted by the natural biological processes in the water and upper sediment. Bacterial respiration at the sediment-water interface is a major consumer of this limited supply of oxygen. Even in the absence of dissolved oxygen, bacterial respiration continues in an anaerobic 22

WATER A ugust 1990

Tim McAuliffe graduated B. App. Sc. from the University of WA, and worked as Research Officer with their Aquatic &ology Group. For this study on destratification he workedfor W.A.W.A. He is currently Senior Environmental Scientist with Halpern Glick Maunse/1.

Mr T. McAuliffe

Dr Ron Rosich is a graduate of the Âˇ University of Western Australia and returned to Perth to take up his present position with WAWA in June 1983 after 23 years research, consulting and university lecturing overseas (Canada, USA and Switzerland) and in Australia; with the last 14 years being in the field of resources management. Ron presently leads the Investigations Sections, Scientific Services Branch, of the Authority.

Dr R Rosich

state causing compounds other than oxygen to be reduced. In the upper sediment reduction of elements such as iron and manganese releases these elements and others bound to them (such as phosphorus) to the overlying water. The release of such substances may lead to a variety of water quality difficulties. Although the control of algae numbers and sediment release are the two main areas of concern in Australia, the complete list of problems associated with stratification is much broader. More specifically, solutions through destratification have been sought to the problems of: unpleasant tastes and odours; high numbers of algae; undesirable species of algae; algae toxicity; high iron levels; high manganese levels; high colour and turbidity; highly variable water quality; release of low temperature water; high zooplankton numbers; high numbers of aquatic insects and fish kills

HAS DESTRATIFICATION SUCCEEDED IN AUSTRALIA? As part of this review of destratification more than fifty cases were examined in detail. The degree to which water quality improved due to destratification was assessed both individually and as groups. Results are Summarised in Table I Table I - Degree of effectiveness of artificial destratification in maintaining high water quality in Australian water bodies. Degree of Success

Number of Cases

Success

12 (240/o)

Limited

23 (460/o)

Failure

14 (280/o)

Uncertain

The Method of Assessment Performance has been judged as either successful, of limited success, a failure, or uncertain depending on the degree to which

destratification solved the water quality problems for which it was implemented. In the case of chemical water quality variables, "successful" instances have been judged as those where the variables in question (eg. iron concentrations) have been reduced to levels set as acceptable by the National Health and Medical Research Council (1987). Cases displaying marked reductions but not to these guideline levels have been categorized as of "limited success". For biological water quality problems, applications were judged successful if they resulted in the elimination of the problem, usually due to a major reduction in the numbers of the problem-causing organism(s). Such an evaluation is partly subjective, but should still serve to give insight into the success of destratification as a water quality management tool in Australia . It must also be remembered that this review has not had to access to all water quality data for each system, nor information on all systems destratified throughout Australia. Both considerations may reduce the accuracy of such an assessment. Specific water quality objectives often were not set for the destratification applications and therefore it was not possible to assess their success against case-specific objectives. Furthermore, it was often not possible to assess the adequacy of the system design or the way it was operated. Consequently we could not be sure that lack of success in improving water quality was due to the destratification technique itself or deficiencies in its implementation. It must also be emphasised that the criterion for "successful" is very stringent, particularly where the objective of the destratification was for an improvement in water quality but not necessarily to comply with the National Health and Medial Research Council (1987) guideline.

The Results of Assessment The majority of cases evaluated in this review (700Jo) have had at least limited success in maintaining high water quality through the implementation of an artificial destratification program (Table !). Only 240Jo of applications, however, have been seen to more completely succeed in maintaining high water quality. Evaluation of the degree of success achieved in response to particular water quality difficulties provides a better insight into why such a high proportion of cases have enjoyed only limited success. As may be seen from Table 2, artificial destratification has been most successful in raising dissolved oxygen levels, reducing colour and turbidity and in reducing levels of iron, manganese and sulfides in the water column. This degree of success has not, however, carried over into the field of algae control.

Table 2 -

The degree of success achieved over specific water quality problems in Australian water bodies.

interface, ensuring high redox potentials and mm1m1smg the reduction of iron, manganese and sulfate oompounds. In cases where destratification has failed to lower the release of iron, manganese etc from the sediment, the failure can usually be traced to an inability to maintain high dissolved oxygen levels throughout the entire water column. In a number of cases, water bodies have experienced both sediment release and algae difficulties. In some of these cases even though considerable control over sediment release appears to have been achieved, algae levels have remained high. Examples exist where, prior to artificial aeration, surface water during summer was poor in quality due to algae blooms, while bottom water was unusable for potable supply due to the accumulation of reduced substances such as iron and manganese. In some of these cases artificial aeration now successfully maintains high quality bottom water (ie low in Fe, Mn) yet surface waters still often experience high algae levels giving rise to problems. In such instances although the supply of water is now assured on a year-round basis, overall the destratification application has achieved only limited success. In cases where control over algae levels has not been achieved the cause is difficult to identify. Where destratification has reduced iron and manganese release (and presumably nutrient release) yet not controlled algae levels, it seems likely that some other factor is controlling a lgae numbers. The responses of algae to mixing are less well understood than the responses of sediment to dissolved oxygen levels and thus require more detailed investigation.

AREAS REQUIRING FURTHER ATTENTION As shown by Table 2, it is in the field of algae control that destratification has had least success and it is in this area that the role of destratification as a management tool appears to be least well understood. The degree of mixing experienced by the water column and the bearing this has on the growth of an indi vidual algae cell needs to be well understood before one may accurately predict the response of the a lgae community to artificia l destratification. Where algae blooms occur in response to increased light rather than to increased nutrients, mixing may be used to remove the algae cells away from the upper, well-lit zone and into the darker zones for periods long enough to retard their growth. To achieve this retardation we need a better understanding of the ' relationships between: • energy input and the movement of water, • the movement of water and the movement of algae cells, • the movement of cells and the amount of light they receive, and • the amount of light received and the effect on growth. A better understanding of these principles, as well as a better understanding of the sediments' response to mixing in a particular water body, will greatly improve the abi lity to predict the effectiveness of artificial destratification.

Number of Cases

Water Quality Problems

Success

Limited Success

Unpleasant tastes & odours 3 (25 % ) High chlorine demand I High algae levels 4 (16 % ) 4 (16 % ) Undesirab le alge species 4 (27%) 4 (27%) High colour and turb idity 4 6 High iron or magnese or sulfides 12 (38%) IO (31 %) Variable water quality 7 (78%) 1 Low DO levels /release 23 (74%) Low temperature High zooplankton / insect numbers Fish kills Dirty wa ter complaints

Failure 8 (67% ) I 17 (68%) 7 (46 % )

2 10 (31%) I 8 (26%) 1

ACKNOWLEDGMENTS This project forms part of a larger investigation into artificial destratification by the Urban Water Research Association of Australia and the Water Authority of Western Australia. A great many people unselfishly contributed information to this review. They are far too numerous to be listed here but are acknowledged in the McAuliffe and Rosich (1989) report on Stage I of this project. The contributions of many people within the Water Authority of Western Australia, especially Scientific Services Branch and North West Region, are greatly acknowledged.

REFERENCES WHERE ARTIFICIAL DESTRATIFICATION HAS FAILED In many cases artificial destratification has successfully reduced the release of iron, manganese, sulfides and other reduced ions from the sediment. In such cases the destratification program has maintained high levels of dissolved oxygen at the sediment-water

Brown , l K (1986) Review of the a pplicatio n of ae ratio n/ destratification techniques in Australian surface wa ter storages. Dept of Loca l Government Queen sla nd. Burns, FL (I 977) - Localised destratificatio n o f la rge reservoirs to contro l di sc ha rge temperatures. Progress in Water Technology 9 pp 53-63. McAu liffe, T F and Rosich, R S (1989) - A Rev iew of the Effects o f Artificia l Destratification on Water Qualit y in Australian Water Storages: Urba n Wa ter Research Associa tion Project No WS 2. 233pp. In Press . Na tional Health a nd Medical Researc h Co uncil (1987). G uidelines fo r d rinking wa ter qualit y in Australia. Aust Govt Pub, Canberra , 33pp.

WATER August 1990

Monitoring Filamentous Bacteria in ActivatedÂˇ Sludge Plants E.M. Seviour, J.A. Soddell, R.J. Seviour, G.C Knight, KC Lindrea and W.C Murdoch SUMMARY The paper discusses the advantages for operators of actwater sludge plants of identification of problem bacteria. It also outlines some of the difficulties which can be encountered in trying to do so.

INTRODUCTION The successful separation of sludge from the mixed liquor, and clarification of the effluent, essential in the activated sludge process, can be disrupted by the presence of large numbers of filamentous bacteria which are associated with the operational disorders of BULKING and FOAMING. With BULKING, extensive growth of these filamentous organisms leads to the development of low density, poorly compacting floes which settle slowly. These can be observed by carrying out simple settlability tests. Those present at the BNRl Conference at Bendigo will remember George Ekama's comment that the most appropriate and useful test of settling character is the Diluted Sludge Volume Index (DSVI) test. The rationale for the use of this procedure and the method have been described in the literature (Ekama and Marais 1984). However, monitoring should not stop there. It is a relatively simple and quick procedure to take a small sample and examine it under the microscope. At lO0X magnification, the viewer can readily see the general size, shape and appearance of the floes, whether they are small, large, dense and compact, or diffuse and irregular. Also, some of the larger filamentous bacteria can be seen, and to a certain extent, their influence on floe structure assessed. Such a microscopic examination will also reveal the presence of higher organisms like the protozoa, and enable the viewer to differentiate between the sessile, stalked ciliate protozoa like Vortice/la, Epistylis and Opercularia, and the creeping ciliates like Aspidisca. It is a generally held view that the presence of large numbers of these ciliates indicates a healthy biomass and a plant performing well . FOAMING, involving the development of a stable viscous foam on the surface of the aeration tanks is a relatively recently recognized problem . Again a sample of the foam can be examined under the microscope and viewed at 400X or lO00X magnification, and such a simple procedure will enable the filamentous bacteria present to be seen. Proficiency in identification using morphological features only comes with practice and experience, and is often fraught with dangers (see later). However, there are probably only 6-10 of these filaments which occur commonly and which appear to be associated with 99% of cases of Bulking and Foaming (see later) . As well as looking at wet preparations, a number of simple and cheap staining procedures can also be used to pick out certain diagnostic features of these filaments to aid in their identification . For example, the Gram stain separates bacteria into two major groups, the Gram positive and Gram negative bacteria. These microscopic examination techniques have been described by Eikelboom & van Buisen (1983) and Jenkins et al (1984) . The immediate benefit of microscopic monitoring may not be obvious, but regular collection of such data, together with the usual chemical and plant operational data, will build up a detailed picture of the microbial community associated with the plant. However, it is important to remember that each plant is unique and what may apply to one plant may not apply to another. Two plants may both have light brown foams, but until these foams are examined under the microscope, it cannot be determined if they are associated with the same organisms. In addition, the type of filament may change in the same plant with time, and consequently it is important that regular monitoring be carried out. Even if the plant is performing well, microscopic examination is still useful, so that any changes might be used to predict early the onset of possible operational problems. For example, a significant increase in the number of nocardioform filaments suggest that a foaming episode may occur in the near future. However, interpretation of what you see under the microscope is not without its problems, and some of these will be discussed later.

Principal Authors: Beth Seviour. Graduated in 1968 from University of Bath with an Honours Degree in Microbiology. Worked as a Research Assistant for two years in Microbiology of Eggs before coming to Australia where she worked at the ANU on Yeast Mitochondrial Genetics. More recently has become interested in identification methods for filamentous bacteria while employed by the Bendigo Water Board and the Bendigo CAE.

Jacques Sodde/1, Graduated in 1973 from UNSW with a major in Microbiology. Worked for 6 years in Microbiology Laboratory at Tooheys Brewery in Sydney before coming to Bendigo as Technical Officer in Microbiology. Currently researching filaments causing foaming, particularly Nocardia pinensis, for this Master of Applied Science.

MORPHOWGY OF FILAMENTOUS BACTERIA It is now recognized that up to 30 different morphological forms of filamentous bacteria can be found in actiyated sludge plants, although often the distinction between these is not as clear as the literature suggests. There are, however, a number of important points to make. The first is that Sphaeroti/us natans, once widely proposed as the cause of bulking, is probably noJ very important, and plants which send samples of sludge and foam away for independent analysis should be very suspicious if their filaments are always identified as this organism. A number of different filamentous bacteria have similar morphology (Wanner & Grau, 1989) and differences may not be detected if the microbiologist carrying out the identification procedure is not very experienced with these filaments. Secondly, morphology can be sometimes misleading as an identification criterion, and later in this article we discuss examples of this. Thirdly, and related to the above, is the recognition that only when these organisms have been cultured and their taxonomic position clarified can any rational (ie scientifically sound) approach to controlling bulking and foaming be undertaken. We believe that the industry needs to support serious studies in this area using techniques of direct isolation based on micromanipulation (Blackall, personal communication) where microscopic appearance of the organism and isolated colony can be related unequivocally.

FILAMENTS IN BULKING PLANTS The filaments normally associated with bulking activated sludge plants are generally considered to be Gram negative filaments and these include Eikelboom types 021N, 0041, 0803, 0092 and 0914, and the Gram positive unbranched filament Microthrix parvice/la. Some of these are shown in Figs 1-6. Microscopic examination of mixed liquor and DSVI measurements all provide advance warning of bulking and both should be part of regular plant monitoring and their use can assist operations staff to be aware of the nature of bulking. Our experience of their perceptions of bulking is that it is often limited to the sight of sludge flowing over clarifier weirs. Many circumstances can cause sludge to leave the clarifier, including floating sludge from denitrification and hydraulic overload, which are not associated with bulking. Bulking should always be defined in terms of the DSVI test and if a bulking episode occurs then the filaments responsible should be identified (Wanner & Grau , 1989) WATER August 1990

FILAMENTS IN FOAMING PLANTS

CONTROL OF BULKING A~D FOAMING

Most of the work carried out at Bendigo has been involved with the microbiology of foaming activated sludge plants. A recent paper (Soddell & Seviour, 1990) critically appraises the often confusing and contradictory literature on this poorly understood problem. Foams were originally thought to be caused by Nocardia amarae (Lechevalier & Lechevalier, 1974), but microscopic examination of foam samples from plants all over the world have revealed the presence of a much wider range of filaments. These are usually Gram positive Actinomycetes or nocardioforms, but Microthrix parvice/la and several Eikelboom morphological types have also been reported (Blackbeard et al 1986, 1988). This is highlighted in a recently published survey of the filamentous populations from foams in plants in Australia (Seviour et al 1990) . M. parvicella, N. amarae-like organisms and Nocardia pinensis were the most commonly seen by microscopic examination, but Eikelboom types 0092, 0914 and 0041/0675 were also observed as dominant in some foams. Their relative frequencies varied from state to state, and comparisons with the study of Blackall (1986) supported the view that foaming is often an intermittent problem and the filaments present can change dramatically over time. Some of these foaming filaments are shown in Figs 6-8.

Filamentous bulking has been reported to be affected by an number of factors including the composition of the waste, nutritional (nitrogen and phosphorus) deficiency, sludge loading rate, sludge oxygen levels, septic influent and low pH (Chudoba, 1985, Jenkins et al 1984) . Knowledge of the organism causing the foam may help in determining the mode of treatment. For example, Microthrixparvice/la is favoured by low DO (Slijkhuis & Dienema, 1988) and filaments like Type 021N seem to be favoured in completely-mixed low F:M systems (van Niekerk et al, 1987), so adjustment of these parameters can help eliminate bulking caused by these organisms. Similarly, high content of sulphides may encourage Thiothrix and Beggiatoa species to grow and these can be controlled by removal of sulphide through pre-aeration or precipitation (Chudoba 1985). There is no guaranteed method of control of foaming because it is caused by a number of different organisms, each with different growth requirements. Until we are able to rapidly distinguish between the different organisms with similar morphology, and learn more about their metabolism and ecology, we will not be able to apply scientifically sound methods of control. Reduction of sludge age has sometimes been successful (Jenkins 1984), but as already pointed out, its success may depend on the organism that is causing the problem . Chlorination has also been widely used (Blackall 1987), but it can also cause serious problems by killing useful organisms, particularly if Microthrix parvicella is the cause of the foam, as this organism is known to be more resistant than most to chlorine (Neethling et al 1985). In all cases, physical removal of the foam is essential and the ability to do this should be built into every plant. A simple floating pontoon system was developed in Bendigo by Ron Bergmeier (Solidsep Pty Ltd) and has been successfully trialled in a number of plants. It must be emphasised that foam recovered from a plant MUST NOT be recycled with the sludge as this will reseed the plant with the organism causing the problem.

WORDS OF WARNING Most studies have relied solely on microscopic examination to identify the filaments present in foam and in bulking sludge. However, a number of questions need to be addressed. (a) Are all Nocardia filaments identical? As already mentioned, three main types of filaments have been observed in foam Nocardia amarae-like organisms (NALO) which typically branch at right angles, Nocardia pinensis which branches at angles less than 90 degrees and exhibits a "pine tree" like morphology, and the unbranched Microthrix parvicella, which is not related to the nocardioform filaments (see Figs 6-8). However, attempts to isolate and grow nocardioform filaments have yielded not only Nocardia amarae, but also other Nocardia species and Rhodococcus species, and these often predominated (Lechevalier et al, 1976; Lemmer and Kroppenstedt, 1984; Sezgin et al, 1988; Blackall et al, 1988, 1989). Our experiences with isolation techniques using micromanipulation of Nocardia-like filaments in Australian foams confirm that filaments which look similar under the microscope, upon isolation give rise to a number of different Nocardia and Rhodococcus species, and possibly Mycobacterium as well. Since the growth rate of the different nocardioform filaments may vary from relatively rapid (Rhodococcus sp) to very slow (Nocardia pinensis), one would expect that control measures affecting growth rate (eg sludge age manipulation) would have variable success rates. Therefore, all the contradictory reports in the literature are not surprising (eg Pipes, 1978; Dhaliwal, 1979; Fleissner & Foess, 1980; Hartley, 1982) and these might be explained by the presence of physiologically quite different, but morphologically similar, organisms. There are suggestions that some organisms involved in foaming may vary their morphology in response to different environmental constraints. This has already been shown for rhodococci, which can exhibit coccal morphology at low growth rates (Williams and Wellington, 1980). The predominance of rod shaped bacteria with minimal branching was first described by Sezgin and Karr (1986) and we have recently found organisms with similar morphology in some Australian foams. Without further study, including isolation and identification, it is impossible to tell whether these are new foam formers or possibly different growth stages of Nocardia or Rhodococcus, both of which are known to fragment. (b) It is ironic that many of the filaments seen in bulking and foaming plants grow so well in the environment of activated sludge but have never been grown successfully in pure culture in the laboratory. As a result their precise taxonomic status is unknown, as is our understanding of their physiology and ecology. Consequently, there is the danger that microscopic examination may not present the full story, and attempts to control bulking and foaming only on the basis of microscopic examination may fail. Our unsuccessful attempts to isolate and grow Microthrix parvice//a using the methods of Slijkhuis (1983) also raise the question as to whether Australian strains of this organism are physiologically the same as those reported overseas. The same could be true for other filaments. More work obviously needs to be done on the basic microbiology of the filaments found in Australian activated sludge plants to minimise the danger of relying on overseas experiences and data in attempts to control these problem.s. 26

WATER August 1990

CONCLUSIONS The Australian wastewater industry needs to become aware of the need to carry out basic microbiological studies on these filaments causing the problems of bulking and foaming. It would be interesting to see what percentage of the present costs entailed in coping with these problems is spent on these activities. The figure would not be very flattering.

REFERi:NCES Blackall, L.L. (1987) Actinomycete scum problems in activated sludge plants. Thesis, Microbiology, University of Queensland, Australia. Blackall, L.L., Harbers, A.E., Greenfield, P.F. & Hayward, A.C. (1988) Actinomycete scum problems in Australian activated sludge plants. Water Science and Technology 20, 493-495. Blackbeard, J.R. , Ekama, G.A. & Marais, G.v.R. (1986) A survey of filam entous bulking and foaming in activated sl udge plants in South Africa. Water Pollution Control 85, 90-100. Blackbeard, J.R., Gabb, D.M.D., Ekama, G.A . & Marais, G.v.R. (1988) Identification of filamentous organisms in nutrient removal activated sludge plants in South Africa. Water SA 14, 29-33. Chuboda, J. (1985) Control of activated sludge filamentous bulking - VI: Formulation of basic principles. Water Research 19, 1017-1022. Dhaliwal, B.S. (1979) Nocardia amarae and activated sludge foaming. Journal Water Pollution Control Federation 51, 344-350. Eikelboom , D.H . & van Buijsen, H.J.J. (1983) Microscopic Sludge Investigation Manual. Second Edition. Delft: TNO Research Institute of Environmental Hygiene. Ekama, G.A. and Marais, G.v. R. (1984) Two improved sludge settlabi lity parameters. IMIESA Vol 9 No 3 20-25. Fleissner, J.T. & Foess, G.W. (1980) Discussion: Nocardia amarae and activated sludge foaming. Journal Water Pollution Control Federation 52, 2594 . Hartley, K.J. (1982) Nocardia in activated sludge plants. Queensland Local Authority Engineers Conference Proceedings 109-132. Jenkins, D., Richard, M.G. & Dai gger, G.T. (1984) Manual on the causes and control of activated sludge bulking and foaming. Pretoria: Water Research Commission. Lechevalier, H.A. Lechevalier, M .P., Wyszkowski, P.E. & Mariat, F. (1976) Acunomycetes found m sewage-treatment plants of the activated sludge type. In Actinomycetes: The Boundary Microorganisms. ed. Arai, T. pp. 227-248. Tokyo: Toppan Co. Ltd. Lechevalier, M.P. & Lechevalier, H.A. (1974) Nocardia amaraesp. nov., an actinomycete common in foaming activated sludge. International Journal of Systematic Bacteriology 24, 278-288. Lemmer, H. & Kroppenstedt, R.M. (1984) Chemotaxonomy and physiology of some actinomycetes isolated from scumming activated sludge. Systematic and Applied Microbiology 5, 124-135 . Neethling, J.B., Johnson, K.M. & Jenkins, D. (1985) Using ATP to determine the ch lorine resistance of filamentous bacteria associated with activated sludge bulking. Journal Water Pollution Control Federation 57, 890-894. Pipes, W.O. (1978) Actinomycete scum production in activated sludge processes. Journal Water Pollution Control Federation 50, 628-643. Seviour, E.M., Williams, C. J., Seviour, R.J., Soddell , J.A. & Lindrea, K.C. (1990) A survey of filamentous bacterial populations from foaming activated sludge plants in eastern states of Australia. Water Research 24, 493-498.

Sezgin, M. & Karr, P.R. (1986) Control of actinomycete scum o n ae ration basins and clarifiers. Journot Woter Pollution Control Federation 58, 972-977. Sezgin, M., Lechevalier, M.P. & Karr, P.R . (1988) Isolation and identification of actinomycetes present in activated sludge scÂľ m. Water Science Technology 2Q, 257-263. Slijkhuis, H . (1983) Microthrix parvicella, a fil a mentous bacterium isolated from activated sludge: cultivation in a chemically defined medium. Applied and Environmental Microbiology 46, 832-839 . Soddell J.A. & Seviour R.J. (1990) Microbiology of foaming in ac ti vated sludge plants: A review. J Appl. Bacteriol. In Press.

va n Niekerk , A.M., J en kin, D. & Richard, M. (1 987) the competitive growth of Zoogloea ramigera and Type 021N in activated sludge and pure cultu re - A model for low F:M bulking. Journal WPCF 59, 262-273. Wanner, J. & Grau, P. (1988) Filamentous bulking in nutrient removal activated sludge systems. Water Science and Technology 20, 1-8. Wanner, J. & Grau, P. (1989) Identi fication o f fil amentous microorganisms from act ivated sludge: A compromise between wishes, needs and possibilities . Water Research 23, 883-891. Williams, S.T. & Wellington, E.M.H. (1980) Micromorphology and fine structure of ac tin omycetes . In Microbiological C lassification and Identification ed. Goodfellow, M. & Board, R.G. pp. 139-165 . London: Academic Press.

MICROPHOIOGRAPHS OF TYPICAL FILAMENTOUS BACTERIA

Figure 1- Type 021N. Tapering filaments, unsheathed, septate, staining Gram negative. No attached growth.

Figure S. Type 0092. No sheathed or attached growth, but are Neisser positive, and are often missed unless this staining procedure is carried out.

-Âˇ

Figure 2. Type 0803. Straight, even filaments, often extending from floes, but can be free. Septate and unsheathed. Gram negative.

Figure 3. Type 0041. Sheathed, usually covered with attached growth. Gram variable.

Figure 6. Microthrix parvicella. Unbranched filaments, which are often coiled, so they can appear in tangles. Not s'1eathed, and no attached growth. Gram positive.

Figure 7. Nocardia amarae-like organisms. Short filaments with right angled branching. No sheath or attached growth_ Gram positive.

Figure 4. Type 0914. Often seen extending from floes, sometimes with attached growth. No sheath, but septate. Gram negative.

Figure 8. Nocardia pinensis. Similar to Nocardia amarae-like organisms, except that the branches have angles less than 90 degrees and all align in the same direction, giving the appearance of pine tree. Branching frequency, branch angles, and branch lengths may vary. WATER- August 1990

A Comparison of Upflow and Downflow :rertiary Filters for Sewage Treatment by MITCHELL LAGINESTRA ABSTRACT The Water Board (Sydney) has two types of filters used in tertiary treatment of sewage; upflow (sand) and downflow (dual media) type. The aim of this study was to look at the different features of each, including advantages and disadvantages in operation, and to determine which type was more effective in removing suspended solids (or nonfilterable residue, NFR). It is concluded that upflow filters are more effective in removing NFR and have many other advantages over downflow filters.

INTRODUCTION Secondary sewage treatment plants (STPs) often cannot produce effluents that will comply to standard discharge requirements. Current SPCC licences for inland plants involve 50% percentile limits of 10/ 15 mg/L Biochemical Oxygen Demand (BOD)/ NFR. Hence a tertiary treatment stage must be applied to reduce BOD and NFR concentrations in the effluent to an acceptable level. The main methods of tertiary treatment are oxidation ponds or tertiary filters . Filters reportedly achieve 70% NFR reduction with a design loading rate of 120-450 m' Im' day (Ellis, 1980), while oxidation ponds achieve approximate 900Jo NFR Reduction at 0.1-0.2 m'/m' day (Joint Committee, 1978). Hence, filters are generally regarded as the most viable option for the tertiary stage of sewage treatment plants where land space is limited (which is common in Sydney). Overall, the purpose of tertiary filters at STPs is to remove NFR, and hence reduce BOD from secondary clarifier effluents. Particle removal reduces the bacterial content and so decreases the amount of chlorination required for the plant effluent (Ellis, 1980). Effluent appearance is greatly improved after filtration . This paper summarises the results of a survey of the filters operating at a number of sewage treatment plants (STPs) in the Sydney area.

wash, the coarse grains naturally remain at the base, ie the feed end. In order to prevent the bed from fluidising during filtration, a steel grid is installed just below the surface of the bed and the sand forms inverted arches which bridge the gap in the grid. Before backwashing, air scour is applied, which breaks up the arch formation allowing the bed to expand in the normal way for backwashing. Typical media size ranges are: anthracite (S.G. 1.6) 1.19 - 1.68 mm sand (S.G. 2.6) 1.18 - 0.60 mm Filters are operated in either, or a combination of two modes. • Declining rate, where the rate of flow is allowed to decrease. • Constant rate, where a control valve, which is nearly closed at the beginning of a run, gradually opens to compensate for the clogging of the voids of the bed. At all filtration plants the filtered effluent passes to a retaining tank (washwater holding tank) which acts as a source of water for backwashing the filters. All filters are backwashed in the upward mode after air scouring. Backwashing is generally carried out once per day, although the frequency necessarily increases during high filter loadings, eg, during wet weather flows . The dirty backwash water goes to a foulwater holding tank before being pumped back to the start of the works, except at Quakers Hill where the dirty backwash water goes straight to the incoming sewer to the plant. This can be a problem where flow surges are transmitted through the plant (particularly during wet weather flow) .

NFR REMOVAL OVER THE FILTERS Monitoring The best measure of tertiary filter performance is NFR removal (Tomlinson, Jago, 1980), since that is the primary func-

M. Laginestra

Mitchell Laginestra, BE (Sydney) is a Chemical Engineer with the Water Board at North Head Sewage Treatment Plant. He is currently doing a Masters of Environmental Studies at Macquarie University and has over 8 years experience in operation and investigation of sewage treatment processes.

tion. Hence, for this survey laboratory results for clarifier effluent NFR (ie filter influent) and filter effluent NFR (grab and composite samples) were obtained. In addition, continuous monitoring of influent and effluent was implemented to provide a more complete picture of filter performance. Continuous monitoring was carried out for most of the filtration plant using NFR sensors/indicators with continuous chart recorders. The period of monitoring at each plant varied, with Quakers Hill and Castle Hill being the most intensive (lasting for 2 and 4 month~ respectively). Laboratory analyses were used to cross-check those of the monitors. Grossly inconsistent results were disregarded.

RESULTS AND DISCUSSION The results for the STPs surveyed are summarised in Table 2. These show that upflow filters are significantly more effective in removing NFR than downflow, with Castle Hill having the best performance (70% removal) . Results for upflow and downflow filters were grouped to determine the relationship

Table I Water Board STP Filter Plants Filter Type

STP (Year of Commissioning

ADWF ML/Day

Number

of FIiters

!liters)

FILTERS AT THE STPs The Water Board (Sydney) operates both downflow and upflow filters. Details are shown in Table 1. The downflow filters are dual-media type, incorporating an upper layer of anthracite with a larger particle size than the finer sand at the base. After backwash, the lower specific gravity of the anthracite ensures that it remains on the top of the bed, ie the feed end, where the larger voids provide extra capacity, leaving the polishing to be performed by the finer sand layer below. The upflow filters contain both coarse sand grains and finer ·sand grains. After back30

WATER A ugust 1990

p F L 0

D 0

N F L 0

Castle Hill (1974) Glenbrook (1975) Mt Riverview• (1973) Hornsby Heights (1984) Quakers Hill (1980) Riverstone (1986) St Marys (1985) West Camden (1985) West Hornsby (1980)

Total Surface Area (m')

Surface Rate Design Capacity

L/ s/m' Actual

Average

48.

3.1

0.9

2.8

77.6

2.04

0.4

1.1

19.8

2.04

0.6

3.6

2.8

3.3

0.6

26.5

4 (Double) 2

300

3.25

1.0

3.1

0.1

450

3.13

0.9

3.0

6 (Double) 2

76.5

3.1

0.5

7.5

99.2

3.25

0.9

1.2 33.5

• All STPs use the activated sludge secondary process except Mt Riverview which employs trickling filters . A third filter has just been commissioned at Castle Hill STP.

Loadlng

Table 2 Average NFR Reduction by Tertiary Filter P'lant

of NFR removal to clarifier effluent NFR concentration. These are shown in Figures 1 and 2 respectively. The figures indicate that downflow filters have a relatively steady NFR removal (50-570Jo), over a wide influent quality range. However, with upflow filters the relationship appears as a curve - they reach their peak performance (72% removal) at influent NFR of 25-40 mg/ L. The graphs show that upflow filters have higher NFR removal than downflows for clarifier effluent NFR concentrations of about 13-60 mg/ L. Below 13 and above 60 mg/ L NFR, downflow filters perform better (although little data was available in these ranges). Hence, it appears that upflow filters are much better suited to tertiary treatment of sewage than downflow type, since secondary effluents typically contain 20-30 mg/ L NFR. One possible explanation for the lower removal efficiencies of downflow filters might be the stratification within each of the two layers, which probably results in high capture of solids at the top of each layer with decreasing capture through the subsequent layer depths. Continuous monitoring results for Castle Hill and West Hornsby are shown in Figures 3 and 4 respectively, and show average influent and effluent NFR concentrations over 24 hours. These plots were typical of the responses found at the other filter plants, ie, filter effluent was dependent on influent changes during the day. STP flow appears to be a major factor in diurnal variations of secondary effluent quality and the figures show that the filters have a definite dampening .effect.

Filter Type

p F L 0

Castle Hill

700Jo

670Jo

Glenbrook

580Jo

640Jo

580Jo

Mt Riverview

690Jo

Average

660Jo

Hornsby Heights

580Jo

630Jo 410Jo

Quakers Hill

58%

56%

F L

St Marys

50%

53%

West Camden

53%

West Hornsby

22 Average

52%

53%

Upflow filters do not require filter control. The throughput is limited by gravity

•

. . ,, "' . . . ' ,. ' ... .' .,,.,,. " '

iii

f> 60

4-0

KEY -

":astle HIii STP 10

•

'rf!.

...., , •• ff•,

£ CC

•

4-0

KEY

en E

::,

Upflow filters: NFR removal

40 ~ ~

4-0

cij,

z z

West Hornsby STP •

4-0

..J

:S LL

~c(

::,

1~ 4-0

en w cc

-!I!

..J LL 100 ~

~c(

Average Se ondary Effluent

CD w

~ .J ii:

West Camden STP

Downflow filters: NFR removal vs clarifier effluent NFR .

..J

•

Fig. 2 -

..J

St. Marys STP •

Clarifier Effluent NFR, mg/L

w cc

Clarifier effluent NFR.

•

Quakers HIii STP

Hornsby Heights STP 60

•

:.,

Clarifier Effluent NFR, mg/L. Fig. 1 -

.. . .. '

: '), LS-.J'f.1 , . r .',- '

Mt. Riverview STP • ~

• flow measurement from each individual filter; • an effluent rate control valve for each filter; and • a level transmitter in the influent channel. This acts as a 'Master' controller in cascade with the flow controllers (known as 'slaves'). The cascade loop utilises two feedback controllers with the output signal of the master controller going to the set point of the slave controller. The output of the slave controller goes to the control valve. The filtration system is dependent on the effective operation of the effluent rate control valves. Previously the system was prone to failure due to some problems with the hardware, especially the valve actuators, and poor signals from the flowmeter tube - these have now been largely overcome. The design control system for the downflow filters is know as the proportional level, equal rate system (or variable controlled coistant rate) and is operational at West Hornsby and Quakers Hill STPs. The control system maintains equal flow from

'' ·' t ..., . '' ..... : \ rf• • ·~-. ~ • '•

Glenbrook STP •

100

' ' f'"','~

'rf!. 80

51%

clarifier efflu ent

:;:;:

100

54%

as well as by the filter bed (and steel grid above the bed), negating the need for a control valve. Thus filtering control is unnecessary, and the only control required is the decision to initiate, and operate, the backwash. The upflow filters currently in use at the board's plants are all much older than the downflow type. At Castle Hill, the auto-backwash is run entirely by cam timers which operate the appropriate equipment (Glenbrook and Mt Riverview are both manually backwashed). The electrical controls at Castle Hill could easily be upgraded for a totally automatic backwash initiation. Additional backwashing could be initiated once a predetermined head loss has been reached (similar to existing manual systems). The downflow filter systems rely on a Programmable Logic controller (PLC) for filter control and backwashing initiation. The basis of automatic operation is control of the flow through the filters. This system was designed as a double loop system to utilise the following parameters:

FEATURES OF UPFWW AND DOWNFWW FILTERS Flow Control

cc cc

C.E. 'NFR,mg/ L

w D

Continuous Monitor Results NFR Removal C.E. 'NFR,mg/ L

Laboratory Results NFR Removal

. STP

Average Filter Effluent

0000

Fig. 3 -

0600

0900

1800

2100

Average- Filter Effluent

1-, 10

10i - - - - - - -

ii:

2400

TIME

Castle Hill STP: NFR variations over 24 hr.

z z

0000

Fig. 4 -

0600

1800

2100

24-00

TIME

West Hornsby STP: NFR variations over 24 hr. WATER August 1990

all filters while keeping the water level within a narrow range. The rate will gradually decline as the filter clogs, and . to compensate, the valve is increasingly opened. The flow passing through each filter is independently controlled and is eq~al to other on-line filters. Blockage vanes between the filters and hence head l?ss will _v_ary. Backwashing is generally t1me-spec1fied but additional cleaning is initiated when the control valve approaches complete opening. . A new type of control system has been implemented at St Marys. This basically consists of a mixture of the equal rate and ~eclining rate systems, and is called proporuonal. level, declining rate (PLDR) . It involves varying the inlet water level over a defined band. All the filter effluent valves are set. in the same position, in direct proportion to the position of the water level within the band. (The rate control valves close when the water level is at the predetermi~e? bottom level, and open for the top pos1t1on). Cleaner filters take a higher flow the share declines as other filters are back: washed. (Browning, 1987). Only one controller is needed for the whole filtration plant, simplifying the system, but the individuality of the filters is eliminated. The inlet level transmitter is used to modulate the rate control valves to the same position for all filters. The outpu_t signal from the level transmitter directly dnves the effluent valve (and is used for backwash initiation). The only control is the inlet level, but it does not receive feedback from the output.

Backwashing . Backwash is applied in the upflow direction for both upflow and downflow filters . Each STP has a similar backwash procedure. This entails shutting off the inlet penstock (or valve), opening backwash outlets, and allowing drain down time for the surface of the water in the filter box to reach approximately 0.4 m below the top of the troughs. Once this level is reached th~ bed is agitated by air scouring for mmutes, (which dislodges solids from the media) when air is stopped the bed settles and the backwash pump is started. (With upflow filters there is no settling period). _Anthracite loss is an enormous problem with the Board's downflow filters and has also been documented overseas (Fitzpatrick & Swanson, 1980). Hence, a settlement period is required for the media. Gradual backwash pump speed build-up is needed to avoid air entrainment of anthracite media (initiation of washwater brings air bubbles trapped in the bed to the surface). The upflow sand_ filters do not suffer from large scale media loss, and require no settlement

3-5

'a, E

"':::,C ii>

"'a:

,,.,

""'

"'ID .,, ..J

< a:

"'!J

·,.

z z Fig. 5 -

Characteristic/

Upflow

Factors

Filters

Downflow Filters

S'~spended Solids (NFR) Removal Filter Box Construction Cost Cost of Electrical & Mechanical Equipment Backwashing Efficiency Media Losses Maintenance Record Ability to Handle Peak Loads - Hydraulic - Solid Energy Usage Provision for Automatic Control Logic

Very Good High Lower Better Very Little Good

Satisfactory Slightly Lower High Good High Poor

Very Good Better Lower Relatively Simple

Satisfactory Good High Complex

period or backwash flow control. Backwashing will affect the performance of a filter since it is necessary to clean out the accumulated solids thoroughly from the beds. On the other hand extended back:,vashing and the subsequent dirty water will 1mpos_e a considerable hydraulic (and orgamc) load on the STP. Hence, it is impo~tant to optimise the backwashing penod. Backwash (dirty) water from several !ilter plants was sampled (at regular time mtervals) and analysed for NFR. Backwash pump running times were also measure. These were used to determine the extent of solids and hydraulic load placed on the STP b~ recycling the dirty backwash water. Figure 5 shows the changing NFR concentration in the dirty water at different times during a backwash at Castle Hill and St Marys. Time taken for back wash water fl~w varied for each STP (usually 6-10 mmutes depending on the pump flow rate which is generally between 150-450 Lis. Th~ ~lower flow rates necessitate longer cleanmg). Generally the filter plants followed the same trend as depicted in Figure 7 - an excellent efficiency in removing NFR was demonstrated and by the end of the backwash, a low level of NFR had been attained. At most of the plants (upflow and downflow) several deadspots (pockets of almost stat!onary liquid with high NFR) were noticeable for a relatively large proportion of the backwash. These were particularly noticeable away from the backwash troughs, along the filter box wall. This was due to the insufficient number of troughs (usually one or two only). An increased number of ~roughs (or reduced surface area) could possibly reduce backwashing time, and consequently the load on the STP. This has been rectified at the most recent filtration plant at Riverstone (4 troughs) . The backwashing trial results confirmed that backwashing imposed a considerable load on the STP, with hydraulic and solids loading being increased by some 5-150Jo at all plants. This result should stress the importance of aiming to recycle dirty backwash flows to the head of the STP only during periods of low flow. I~ is also important to backwash filters dunng low flows (if possible) to avoid hav!ng filters off line during flow peaks. This 1s generally done at the Board's plants.

leakage problems with square shaped troughs at Quakers Hill caused by the impact of the washwater flow. J?ownflow filters have suffered from a vanety of failures since their commissioning leading to costly repairs/replacement and extended downtime. There has been no filter equipment update at Castle Hill STP since commission!ng. Also there has been virtually no repair necessary, apart from occasional routine replacement eg, seal for a solenoid valve. However, Glenbrook STP has had a histo~y of problems with the air lines. Despite bemg heavy duty plastic some back pressure has caused failure in the pipes. It is now accepted that plastic pipes are not suitable for this application. Castle Hill and Mt Riverview filters do not suffer from this problem (they have metal pipes on the air discharge line). Glenbrook has also suffered from ~he filter inlet nozzle blockages by screemngs. However, the nozzles have only 4 mm openings. The problem has been overcome by installing fine, manuallycleaned screens in the filter inlet. This pr?blem has not been experienced at Castle ~ill (1~ mm nozzle openings) or Mt Riverview (fme screening upstream of trickling filters) . Generally speaking, Castle Hill is the best designed filtration plant within the Board's STPs. Advantages include: •. simple head measurement, (level in the differential inlet pipe chamber); • no level/flow control system required to maintain head; • no modulating rate control valves required; • fixed speed backwashing, hence backwash flow control system is unnecessary; ~ release (to atmosphere) on air delivery !me, to allow escape of trapped air bubbles fro_m underneath the bed to prevent air entramment of media particles. . One inadequacy of the Castle Hill filtration sy~tem is the shape and subsequent hydraulics of the header tank. There is no flow splitting device, or barrier, so the secondary effluent preferentially feeds filter No. 1. This problem is not restricted to Castle Hill. It is important that secondary e!fluent flow be distributed evenly between filters to maximise NFR removal.

General

SUMMARY AND CONCLUSIONS

It is important that there be more than

;;---....._

U: 0

Filter Backwash

CASTLE HILL STP ST. MARYS STP ,.__

Table 3 Comparison of Filter Characteristics

----.. s

TIME minutes

NFR concentration in Dirty Backwash water.

WATER August 1990

two filters at any filtration plant. Otherwise ?ac~washing will result in a 500Jo reductio~ m filter plant operational capacity. The more recent design of U-shaped b~ckwash troughs at Hornsby Heights, R1verstone etc, has overcome warpage and

. Upflow type filters are better in removmg su~pended solids (for NFR) than downflow filters . Upflow filters achieve an overall average of 640Jo NFR reduction, compared to 540Jo for downflow type, a significant Continued on page 51

BNR-1 A Report by Bob Swinton

First Australian Conference oil

Biological Nutrient Removal Bendigo, 9-12 July 1990 Eutrophication of streams, rivers, lakes and storages by man made interference with the natural environment is increasing. As the Australian community becomes more environmentally conscious, and the results of eutrophication become more obvious, the removal of the nutrients nitrogen and phosphorus from wastewater effluents will become increasingly necessary. Nitrification-denitrification is already well-established in some treatment plants and chemical precipitation of phosphorus is a tool already used in a few critical locations. However, the discovery, almost by accident, in South Africa some 15 years ago that modifications to the activated sludge process could attain effluents with low levels of phosphorus as well as. of nitrogen gave promise of a significantly cheaper technique for protection of the receiving waters. Variations of the process have been applied in a number of countries, but significant differences, even fai lures, have emerged. It would seem that local conditions have a significant effect, and laboratory investigations and pilot plants are still necessary before the process can be guaranteed in specific applications in Australia. A number of authorities and institutions have been conducting investigations and at least three plants are in operation, with others planned. The need for a full-scale Australian conference on biological nutrient removal was evident and planning began after the wellattended AWRC workshop in Ballarat, 1988 (see Water, February, 1989). The untiring efforts of the Organising Committee and the support group resulted in a highly successful conference, a salient point being the ready exchange of views between delegates, speakers and overseas guests. The Bendigo College of Advanced Education had been suggested as a venue and it proved to be ideal. The involvement of the College in the research and development leading to the installation of BNR for the Bendigo Water Board's new sewage treatment plant is well documented, and the offer by the Board and SKP for an inspection of the new construction as it neared completion was an added bonus. However, the most appreciated point was the comfort and convenience of the relatively new campus, combined with the typical warm country welcome extended to delegates, not to mention the efficiency of the organisation . The distance of 150 km from city of Melbourne and its airport proved to be of no consequence, with interstaters filling a special bus, or taking the opportunity to drive through delightful hilly country, far removed from the Hume or Western Freeways.

The Conference Executive, left to right: Bill Raper, Warren Wealands, Neil Burns

Post-conference inspections of both the Ballarat phosphorus removal plant, and the huge complex of land disposal systems, lagoons and 'artificial wetlands' at the Melbourne Board's Werribee Farm were well attended. A total of 180 persons registered, drawn from a wide spectrum of Federal and State Government Agencies, both regulatory and investigatory, Water Authorities, (some foreseeing a not-so-distant requirement for nutrient removal), academics and CSIRO, consultants, construction companies, chemical companies and equipment suppliers. Their specific interests and knowledge of the various aspects of biological nutrient removal led to the ready exchange of views mentioned above, both 34

WATER August 1990

during the technical sessions and over refreshments. A highlight was the open forum which was not only instructive but also entertaining. (Why is it that sewage raises so much hilarity, even among the practitioners?) The technical program encompassed a total of 27 papers presented by eminent speakers from Australia and overseas, together with 18 poster presentations, all very relevant to the subject. The international significance of the work being performed in Australia was demonstrated by the ready agreement of the top men in the international field to present reviews to the conference. The names James Barnard and George Ekama of South Africa and David Jenkins of California seem to be synonymous with BNR. Prominent Australian speakers were also invited, and their contributions are recorded in this report. This report on the technical sessions and discussions can only briefly summarise the main points brought out in the papers, and can be no real substitute for the bound papers, a number of which are available for purchase (see page 52). The reviews by the invited speakers are themselves virtually a textbook on the subject.

The Conference Committee, left ta right: Norm Pilkington, Ken Lindrea, Bob Seviour, Warren Wealands, Jonathon Crockett, Harry Crynberg (John Jennings, not in photo)

This reporter gained an overall impression of the technical discussions. The need for nutrient removal from STPs is by no means clear-cut. Each case must be studied carefully to ensure that the most cost-effective decisions are made, particularly taking into account the contributions of non point sources. The poor reaction of some water bodies to recent actions may not be the best guide, since mobilisation of phosphorus from sediments laid down in the past can continue for many years. The philosophy of 'less is better' seems reasonable, but costs must be considered. The use of chemical precipitation for phosphorus removal is expensive, and adds salinity to the receiving water. It was these facts which led the South Africans to devote so much effort to development of combined biological nitrogen and phosphorus removal. They have succeeded, to the extent that there are currently some 50 such plants operating in South Africa, ranging from 5 to 150 ML/ d, and the process is stable enough to be left to normal operators with some extra training. Design can now be based on computer models. Nonetheless it is wise to have back-up chemical facilities in case of system hiccups. However, operation of various pilot plants in Australia to date indicates that the characteristics of the sewage are sufficiently different for there to be a need for further research. Although designs for larger scale plants are based on many variations of the original 'Bardenpho' system, it seems that for smaller systems, modifications to the intermittent extended aeration plants may be quite successful. Although engineers have gone ahead with designs on a pragmatic basis, research into understanding the microbiology and biochemistry of the processes is justified, because it is only such understanding which will enable process upsets to be diagnosed, and it is also to be hoped that the conditions for the development of more effective microbiological populations may eventuate.

THE OPENING At the opening, delegates were welcomed to Bendigo by Mrs Elaine McNamara, Chairman of the Bendigo Water Board, and addressed by Peter Norman, President of AWWA. Peter d·rew attention to the four issues which the water industry must now address: • Effective long-term protection of resources, • Planned renewal of infrastructure • New-found parameters of public health • Community expectations for reduced impact on the environment. To cope with all of these, whilst being denied increased funding, the industry has to develop 'smart' solutions. The BNR process is typical, in that it has a philosophy of low energy input, and minimal use of chemicals. He closed by saying that the joint interests of AWWA and the IAWPRC in communication and technology transfer were reflected in their joint sponsorship of the conference, and complimented Bendigo CAE on the pleasant and appropriate venue.

KEYNOTE ADDRESS:

Bendigo CAE Support Group, left to right: Beth Seviour, Sue Scott (Conference Secretary(, Karen Hamilton, Jenny McQueen, Stuart Pilson, Maree Kelly (CSIRO), Wendy Mellington

by Dr DAVID MITCHELL, Director of the Murray-Darling Freshwater Research Centre

THE FWSH OF SUCCESS: INNOVATION IN THE MANAGEMENT OF WASTEWATER Dr Mitchell's address was directed at the philosophy of innovation in the management of wastewater, and his purpose was to emphasise that gains in technology were unlikely to succeed in the long run unless there was a change in the environmental ethic of western society. He commenced with an ironic review of the history of sanitary engineering, pointing out that very few men had achieved a knighthood in its practice, the exception being the well-known Sir Thomas Crapper, whose design of the Valveless Water Waste Preventer, in other words, the siphon flusher, considerably reduced inflow into the early sewers of 19th century London. The fact that his name led, by a closed loop through America, to a new word in the English language was an added commendation. However, contrary to popular opinion, he was not the inventor of the water closet. The first record belongs to another knight, Sir John Harington, who described in 1596 a system which would "keep the privy as fair, sweet and savourie as the bedchamber". (Sir John, however, gained his knighthood by marriage to royalty). It was the invention in the late 18th century of the D-trap and the S-trap by men such as Cummings, Cosser and Bramah which really led to success of the water closet, but their systems only fed cesspits which had to be emptied at night, and, by law, not into the street drains. However, as the systems gained in popularity, 'capacity augmentation' became vital, and in 1847, the City of London reversed its policy and enforced discharge of cesspit waste · into the stormwater drains, and thus to the River Thames. The results in terms of environmental degradation and public health disasters are well known and have led gradually to the modern profession of sanitary engineering, with all its innovations. The history of innovation up to, say, 1960, has been a battle between the driving forces of waste disposal, private hygiene, aesthetics, then water-borne disease, and the opposing forces of ignorance, prejudice and parsimony. In recent years, public concern for environmental welfare has been added to the driving forces, but the opposing forces have been reinforced by public health apprehension and ill informed, unbalanced environmental concerns, leading to uncertainty. Uncertainty reinforces the conservative bureaucrat's desire to do NOTHING which could be construed as a risk . The setting of regulatory standards always errs on the side of ignorance. The conservatism of bureaucracy is assisted by the fact that there are too many departments to deal with, that they don't agree with each other, and that preservation of State Rights seems more important than preservation of a water resource. He averred that this would be the single most difficult hurdle for innovation in this field .

David laid down these principles: • Innovations are never adopted unless they happen to be convenient. • The likelihood of success is less important in the public sphere than the perceived risk of failure. David also argued that we need to explore trade-offs between quality and reliability, we need flexibility, even a trade-off between standards and costs for different situations, because higher standards lead to exponentially increasing costs, beyond the capacity of many small communities to su;port. Above au; we need better ecological understanding. The public in general seems more concerned about specific elements in the environment (eg the green and the furry) rather than a whole ecosystem. In his opinion as a biologist, the resilience of natural systems was frequently overlooked. He restated his view that decisions on matters such as waste management must be based on sound sustainable environmental ETHIC. There is a vital need for this .. .an unselfish, long term ethic, which can be called upon for everyday guidance, at all levels of society. Rules and regulations, however carefully laid down, can and will be avoided, or modified, to suit human selfishness wherever it is called to account. The ethic he referred to is exemplified by the famous statement made in North America by Chief Seattle, in 1854. He concluded by quoting from the Aboriginal ethic as dictated by old Bill Neidjie, the Kakadu Man, to Stephen Davis.• I belong to this earth and earth stay with us ... forever. This earth ... I never damage, I look after. I know I come back to my country. When I die I come back to (become) earth. I love this country and this earth. You'll be part of this earth when you die. You responsible now. This story for all people. Everybody should be listening. Same story for everyone, Just different language. • extrac ted from Australia's Kakadu Man, Bill Neidjie by Bill Neidjie, Stephen Davis and Alan Fox. Published by Resource Managers Pty Ltd. Darwin, 1986.

. WATER August 1990

TECHNICAL SESSIONS These were divided into Jive themes: - Microbiology The Need Process Research Low Cost Options - Full-scale plants

THEME 1: THE NEED FOR NUTRIENT REMOVAL The question, why bother? was explored by a number of invited speakers concerned with the aquatic environment, and it became clear that there was no simple answer. Ian Smalls, of the Department of Water Resources, NSW, concerned himself mainly with water supply reservoirs. Changing public perceptions have led to a demand for improved water quality. A survey conducted in 1989 by SKP and DWR of 'perceived' water quality problems in NSW showed thateutrophication shared the lead with the physical parameters of colour and turbidity. However, eutrophication is not solely controlled by the presence of adequate nutrients. Physical and biological factors have to be considered, as well as the chemical inputs, whether from point sources, diffuse sources from agriculture, or even natural inputs of nitrogen and phosphorus. In attempting to set criteria for concentrations of nitrogen and phosphorus, the unique nature of Australian water bodies must be taken into account. Data on algal growth have shown that the response to phosphorus is far less than in Europe and USA, due presumably to the effects of turbidity, highly variable flows and stratification, despite our warmer climate, with longer growth seasons. Ian discussed the OECD (1982) criteria, with their admitted ambiguities, and led to the concept of the 'nutrient sensitivity index' being developed by Bek, Cattell & Smalls for AWRC to take account of the nonchemical factors (to be published in 1990). As examples, a 'tolerant' water body might sustain a phosphorus concentration of 30 microgram/ L, whereas a 'sensitive' body might react to 10 microgram/L. Peter Cullen of the Water Research Centre at University of Canberra has been investigating the limnology of inland lakes and storages for some years, and in 1987-8 worked with Forsberg of Uppsala on a survey of the effects of reducing point sources of phosphorus in a number of lakes throughout the world. He made the point that the knee-jerk response to algal blooms of installing nutrient removal at the local STP, though perhaps necessary, was not always effective. For example, whereas the accepted wisdom in southern Australia is that phosphorus is the controlling factor, at certain seasons of the year nitrogen can be the limiting nutrient. The role of turbidity is complex. Light limitation is commonly cited as the controlling factor, but adsorption of phosphorus on the particulate matter and the effect of cations on coagulation of particulates may be more significant.

The inter-relationships between all the biological components of the eco-system are highly significant. Peter introduced the very different model of biomass production in lakes adopted by the fisheries biologists, ie the top-down, rather than the nutrientdriving, or bottom-up, approach. Stocking of fish in itself can solve a nuisance eutrophication situation, but on occasion may actually precipitate an algal bloom by reducing the crop of zooplankton which normally graze on the phytoplankton. He then discussed the vexed question of point sources versus non-point sources, the need to work out a nutrient budget for the receiving water. This is easy enough for the point sources, but estimation of inputs from diffuse sources is not at all precise. Erratic storm events are far more significant than normal run-off. For example, in a study on Lake Burley Griffin, 69% of phosphorus transport occurred in only 9% of the study time. In relation to sediment release, although the overall annual flux to a lake following STP upgrade may decrease, seasonal cyclical release and recapture by the benthic layers may yield adequate phosphorus for nuisance blooms at critical periods for many years after such action. The biological consequences of advanced wastewater treatment are not easy to predict, but is necessary if we are to make sensible and cost-effective decisions about whether to reduce a particular nutrient input, and by how much, in order to achieve acceptable biological consequences. Barry Hart, of The Centre for Stream Ecology, Chisholm Institute (now part of Monash University) has specialised in the chemistry of nutrients and trace elements in freshwater streams. His paper introduced the audience to the concept of nutrient spiralling in flowing streams, where the cycling from sediment to the water column and the biota occurs, as in lakes, but is stretched out by the dimension of flow. He differentiated between upland and lowland streams. For the former, the major effects of human activity are alterations to the riparian zone, sediments from forestry, and occasionally (as in the Alpine resorts) some nutrient inputs. In contrast most lowland streams and rivers have suffered serious disturbances from degradation of riparian vegetation, erosion, agricultural run-off, stock faeces and many also receive sewage effluents. In faster streams, attached algae predominate; in the larger slow flowing rivers, backwaters and impoundments, phytoplankton may also be increased. There are also examples where macrophytes such as Phragmites have practically blocked a stream channel, presumably due to nutrient enrichment.

"It's a lot more difficult to get the P out of sewage than to put it in." Professor David Jenkins. 36

WATER August 1990

There have been few studies of sewageenriched streams undertaken within the nutrient spiralling concept. The Centre for Stream Ecology has developed a relatively simple technique for determining the spiral length for phosphorus over a 50m reach, and is applying it to some Victorian streams, using flow injection analysis for simultaneous determinations of phosphorus and a tracer such as bromide. Barry suggested an interim approach to assist in making a decision on the need for nutrient removal. After a detailed study of the chemistry and biology of the stream, a series of small channels in stream would be proportionally dosed with the proposed effluent (or a simulation) and would be studied for changes in their community composition, as a guide to the first level of standards. Follow-up of the effects of full-scale application would be vital to accumulate evidence for future use. The theme was rounded off by papers from two 'regulators'. Brian Robinson, Chairman of the Victorian EPA, presented an overview of existing legislation in the Australian States, and their underlying philosophies. Far more cooperation towards a national approach to our problems is emerging. For nutrients in receiving waters, the objectives recommended by the 'Australian and New Zealand Environment Council are listed in Table 1. State legislative systems differ, but all agree basically with the above, although modified by tl1eir particular conditions. For example, WA have found it necessary to limit the application of agricultural fertilisers in critical areas to control the effects of run-off. Table 1 ANZEC Guidelines Potable supplies mg/ L

Recreation use mg/ L

Total Phosphorus (as P) Total Nitrogen (as N) Chlorophyll a

0.020 0.500 0.005

0.050 0.500 0.020

Effluent Water Qual ity Indicator

Level

mg/ L

Percentile achievement

I 15

90 90

Receiving \Valer Quality

Indicator

Total Phosphorus (as P) Total Nitrogen (as N)

The Victorian EPA has a policy of waste minimisation as a first option. For point sources of nutrients, discharge of effluent to land wherever practicable or environmentally beneficial is the next stage. Only when this is not possible will the EPA consider discharge to water, and only then if the discharge will not seriously affect the 'beneficial use' or 'environmental value' of the receiving water body. For an STP with a capacity greater than 500 ep, the basic requirement is secondary treatment to 20/30, or in cases of low river flow, to 10/ 10. For existing STPs, if the waterway shows signs of nutrient enrichment, ie excessive

"Can we make an estimate of the total P exported out of the country?" - Lance Bowen on one possible beneficial effect of P.

Table 2 Nutrient Reduction Facilities in Victoria Population

algal growth, or depressed DO at night, the plant must be upgraded either by re-use of the effluent on land, or by addition of nutrient reduction facilities. New plants must either use land disposal or reuse, OR reduce nutrients by 'commonly available technology'. Limits of 10 mg/ L N and 2 mg/ L P are commonly applied. With no mandatory, only recommended, criteria such limits and their attainment are open to negotiation. The EPA recognises the difficulty of scientifically determining discharge limits, but follows the philosophy that 'less is better'. The argument that diffuse sources are the major source of nutrients is countered by the fact that at base water flows, the point source usually dominates the receiving water. Victorian STPs with nutrient removal facilities are listed in Table 2. The use of biological techniques for reduction of phosphorus, with chemical precipitation for back-up only, is being encouraged to reduce both salt loads and costs. With regard to diffuse sources, the strategy of Integrated Catchment Management is being drafted. Three Catchment Coordinating Groups are already fully established, and several more are being formed, their aims being to 'optimise the sustainable beneficial uses of the physical environment'. The paper then outlined the experiences following nutrient control in two inland streams, receiving effluents from Ballarat South and Sunbury. Comparison of the flows in the first summer after nutrient reduction did not confirm the role of phosphorus in plant growth, but demonstrated the importance of numerous other factors. However, since the creeks had been nutrient enriched for some time, conclusive shortterm improvement could not realistically be expected. Three other case histories were discussed; Lake Colongulac, Lake Colac,

Localion

(equivalent)

Sunbur y

30,000

Ba llarat North Lilydale (MMBW) Ballara t South Brushy Creek (MMBW) Wodonga Bendigo Werribee (MMBW) H ealesville (MMBW)

27,000 20,000 75,000

Total

Limil

mg/ L

32,000 45,000 70,000 2,700,000 4,000

â&#x20AC;˘ - No enhanced reduction B - Biological removal technique

and the Wimmera River downstream of the Horsham STP. In each case, nutrient reduction strategies have been applied in the past few years, phosphorus levels in the water body have been dramatically reduced, but the biological results so far have been disappointing, illustrating once again the complexity of the problem and the need for further investigations. In his opinion, it is futile to seek nutrient criteria which will have general application. He therefore recommends the following approach where land disposal is not possible: a. Follow the recommended objectives of the ANZEC 1987. b. Put the onus on the discharger to establish that the proposed effluent does not lead to eutrophication, and if evidence be found, that the discharge cease. He concluded by stating that the prognosis for our waterways is not good, and there will be further degradation. Given the time-lag for recovery it is imperative that we maintain the impetus to reduce nutrient inputs. Douglas Nicolaisen, of the State Pollution Control Commission, NSW, presented a poster paper, (written up in the conference proceedings), which discussed the role of the regulatory organisation. In the field of nutrient removal particularly, the role has to go far beyond that of environmental policeman, but must encompass that of education both of industry and the com-

Total P Limil mg/ L

Comments

Operating since mid - 1970s; about to reduce P to 1 mg/ L Operating since 1984 Operating since 1985 Operating since 1989 Operating since 1989 Operating since 1989 Under construction experimental To be operational 1991 C - Chemical removal technique t - Annual TN objective = 4,300 t

munity. The organisation is also obliged to provide feedback to the regulatory system to ensure that regulations accurately reflect the standards which can be met in practice. (His paper also contains some fascinating examples of pollution control regulations in our early history).

Discussion on Theme 1 Ekama, U. Cape Town: In one study in

South Africa, 80% of the phosphorus input took place in 3% of the time. Bowen , UNSW: The bulk of Australia's population lives on the coastal fringe, so what work is going on about nutrient pollution of seawaters, such as estuaries? Smalls: Estuaries are more tolerant than freshwater, but the red tide phenomenon is not fully explained. Swinton, AWWA: Can biological control be applied to storages, eg,artificial stocking of zooplankton? Cullen: Some attempts have been made, but not successfully. The choice of species may be too limited. Smalls: One can sometimes manipulate a storage to favour graziers. Mitchell, MDFWRC: The question of public perceptions should be addressed, eg what is too green? Cullen: Most certainly, it is a value judgement. A few areas of Lake Burley Griffin are quite green, but this does not worry anyone. The culture of Sweden demands almost perfect clarity, but greenness does not worry a Northern Territorian.

THEME 2: WW COST OPTIONS The frequent mention of 'land disposal' of effluent to protect water bodies implies that this technique is simple and poses no problems. This is not the case, for the impacts of an effluent on the soil itself, on the vegetation growing on it, and on the groundwater must all be considered. However, the low cost options of land disposal, lagoons and artificial wetlands should be suitable for a fair range of Australian conditions, in order to _protect receiving water bodies, whether impounded or flowing, from the effects of nutrients as well as of basic pollutants. This concept was addressed in Theme 2 of the conference by a number of authors. Kathleen Bowmer, of CSIRO Division of Water Resources, Griffith Laboratory, 38

WATER August 1990

presented a paper on the disposal of rural industry effluents and wastes, most of which are highly concentrated and charged with nutrients which have the potential to be reused. The rapidly growing beef feedlot industry was a case in point. Irrigation of such wastewaters is not only cheaper than sewage treatment, but it does return the nutrients to the land, and perhaps more significantly, it can return valuable organic matter to the soil to maintain its structure and fertility. However, in harnessing the nutrients we have to manage all the other components and preserve a long-term balance in the soil. The paper reviewed the potential for problems for the major components; nitrogen, phosphorus, BOD, salt and toxic chemicals. The goal of 'no discharge' is admirable, to protect surface waters, but the dilemma is that this policy

may lead to land degradation due to accumulation of salt; or other recalcitrant compounds, in the soil or groundwater. The difficulty with the concept of irrigation of crops or trees by effluents is that very few effluents have the cprrect balance of nutrients and also the uptake of nutrients and water by vegetation is by no means constant throughout the year, or even from year to year. Kathleen referred to a number of models which are being used to balance nutrient and water supply, and her colleague Peter Laut described the models which are being used to predict the longterm effects of such industries on the water courses and the aquifers, particularly taking account of the cumulative effects of a number of developments, to guide the proper siting of such industries.

The emphasis was shifted to urban wastewater in the paper by Peter Fisher, of the Sydney Water Board, who reviewed the nutrient removal mechanisms and dynamics of artificial wetlands. To date, the performance of such systems has been very variable, which reflects the need for further work to optimise design and operating procedures. However, they have the potential to remove organics, nitrogen and coliforms for small communities, at low cost, but at present they do not effectively remove phosphorus. They require large areas of suitable land, typically 3 ha for 1 ML/ d of primary effluent. However, as well as research projects, some larger scale demonstration projects are now in operation in NSW, (Blue Mountains, Byron Bay, Coffs Harbour and Minmi) and their long term results will be a valuable addition to our knowledge. Tom Davies, of the Chisholm Institute of Technology, updated the conference on the preliminary results from the experimental reed beds at Frankston, which were commissioned in December 1988. They are based on the Kickuth concept of no harvesting of the reeds, which die down each season and regrow. Consequently, the only way phosphorus can be removed from the wastewater is by adsorption or precipitation in the substratum or root zone. The necessity for maintenance of a reasonable degree of hydraulic conductivity demands the use of sand/ gravel rather than clay soils so that phosphorus capacity is rapidly reached. The aim therefore is limited to nitrification/ denitrification to reduce ammonia and total nitrogen in the effluent. Preliminary results indicated that nitrification was the rate controlling step, so to assist the natural diffusion of oxygen from the emergent plants down to the submerged roots, intermediate aeration bays are being operated to assist nitrification of ammonia. The planted beds, once established,

David Jenkins, Professor of Environmental Engineering, Berkeley, CA. presented a review prepared together with V. Tandoi of the Italian Water Research Institute, entitled "The Applied Microbiology of EBPR, Accomplishments and Needs". The printed review, over 5000 words long with over 50 references, covers empirical observations and suggested mechanisms, and discusses the mathematical modelling predictions. Âˇ Models based on pilot trials on a particular wastewater can be used for plant design, but he was critical of most of the laboratory experiments, conducted with pure cultures, which aimed to establish the rate coefficients for the individual processes. In his opinion, these did not reflect the very varied cultures which develop in continuous plants, and the thesis that polyphosphate storage is a lowgain survival strategy by stressed microorganisms. He concluded by listing suggestions for chemostat experiments which could test this hypothesis.

removed BOD by 70-90%, somewhat better than the unplanted controls (65-85 % ) Ammonia removal was good in summer (80%), but practically zero in winter. Initial results on the aerated systems show an improvement, and the experiments are being continued. A poster paper summarised one of such experiments. A dye tracer was used to determine the flow patterns, and a technique employed for taking samples along the bed showed that for removal of suspended solids and BOD, a length of 15m rather than 30m was adequate, equivalent to a residence time of 2 days. In subsequent discussions, and in a poster paper by Sorrell et al, the performance of the CSIRO upflow reed bed system was highlighted. As reported by Breen in Ballarat, 1988, forcing the water to pass through the root system, instead of it tending to short-circuit beneath the roots, achieved very good removal of both nitrogen and phosphorus. On the laboratory scale COD was reduced from 209 to 52 mg/ L, TKN from 40 to 2 mg/ L, and TP from 33 to 0.3 mg/ L. Detailed research has shown that most of the nutrient removal is performed in a zone some 5-20 mm from the root tips. A patented design for a large scale adaptation is being developed by CASSIRO, a joint venture company, and is being installed at Coffs Harbour. CSIRO have started a three-year project on the phosphorus transport mechanism. The huge lagoon systems operated at the Melbourne Board's Werribee Farm are wellknown. Efforts are being made to better understand the dynamics of nitrification in order to improve the poor removal of ammonia, particularly in winter. Peter Gross presented a paper on the further application of Constable's 1988 study on 'standard instantaneous residence

THEME 3: MICROBIOLOGY OF NUTRIENT REMOVAL Nancy Millis (Melbourne Uni) presented a review of the biochemistry of nitrification and denitrification. She stressed the low growth rate (e.g. doubling times of 11-18 days in continuing cultures), the slow reaction rate of the nitrifiers, and their response to different forms of nitrogen in the incoming substrates. In contrast, the denitrifiers are very common in nature, and although they prefer oxygen will rapidly utilise nitrate in anaerobic conditions. She drew attention to the success of the AAA (alternating aerobic/ anaerobic) process developed by CSIRO (Ip and Bridger) which has been proved for five years in the Brushy Creek plant, and is to be employed by the Melbourne Board for the new plants at Craigieburn and Keilor, not only to achieve nitrogen removal but also to economise on aeration energy (However, this system does not reduce phosphorus).

time'. Since nitrifying bacteria are slow growing, there is a danger that if the population is washed out of a lagoon, it may take some weeks to re-establish itself. The SIRT formula seeks to correlate growth rates with temperature and pH in order to establish the maximum flowrate under any conditions. Thirty days was initially necessary, but activity would then be maintained with an SIRT of no less than 10 days. Analysis of the results on the full-scale lagoons were far from conclusive, so three experimental lagoons, each 23 m wide and 260 m long, were built. EPl is 0.6 m, EP2 and EP4 are 1.2 m and EP3 is 2.4 m deep. Results are not yet complete, but the relationship between nitrification and high algal activity, with consequent high pH and DO, seems clear. The deeper ponds do not perform well. The SIRT concept seems only partially successful, and modification of the temperature coefficient will probably be necessary. A poster paper by Baskaran, Scott and Connor described the experiments being conducted to enhance nitrification by incorporating biofilm supports within lagoons. Early work utilising cheap materials such as old tyres had shown that a mixed population of algae and bacteria could be established, and that these films had a high nitrifyihg capability, presumably because of the local oxygen supplied by the algae. However, the data was inadequate for design pu1;1:oses. In the current experiments the effect of light became very apparent and this may lead to a complete reappraisal of the system.

The work of the Bendigo College, by Beacham, Seviour, Lindrea and Soddell, on the taxonomy of Acinetobacter isolates from the Bendigo BNR pilot plant, had been complimented by Professor Jenkins in his presentation. Their data showed a much greater range of diversity than was obtained by Duncan from the Lower Plenty pilot plant. It would seem that no single taxonomic group controls phosphorus removal, and it is unlikely that plant performance could be monitored by attempts to follow population trends. Papers from the team at Monash University and CSIRO, led by Ron Bayly, delved deeply into physiology of various pure cultures of strains of Acinetobacter, and some mutants, in an endeavour to determine the reasons for their marked differences in ability to accumulate polyphosphate. Experiments with various enzyme inhibitors are in progress to try to elucidate the biochemistry of polyphosphate synthesis.

"There's no such thing as a stupid micro-organism." Professor David Jenkins on bacterial behaviour.

"Denitrifiers are like the poor - they're always with you." Professor Nancy Millis.

"It's a real pity these organisms don't understand mathematics." Professor Jenkins again.

"Engineered bugs are really a load of old rhubarb." Professor Millis again. WATER August 1990

A second paper, presented by Annabelle Duncan, postulated that phosphorus metabolism is encoded in the plasmids present in all Acinetobacter cells, rather than in the main chromosomes, since this property can sometimes be lost spontaneously without the cell losing its viability. Gel electrophoresis is being used to compare the profiles of the plasmids in good and poor strains, and the loss of two large chunks of DNA was noted. However, more work is necessary to try to isolate the gene. Papers on microscopic examination of the active populations were presented by a number of authors. A paper by Robyn Tuft, of CSF, described the observations made at the Castle Hill pilot plant, under various configurations. For good phosphorus removal, significant levels of clusters of Acinetobacter are required, but they develop gradually, and other microorganisms have been shown to perform the same function. Because of the long sludge ages of BNR plants, and the presence of low oxygen zones, filamentous bacteria are encouraged, and some may even contribute to phosphorus removal. Bulking and foaming problems result, but good plant design can cope with them. A paper by Linda Blackall and members of the Italian Water Research Institute described data obtained from a bench-scale

The flagship of this session was undoubtedly the keynote review by Ekama, Wentzel and Marais of the University of Cape Town. The printed version contains 9000 words of erudite discussion of the development of understanding of Âˇthe twin processes of nitrogen removal and phosphorus removal (backed up by 90 references). George Ekama holds the view that biological nutrient removal is now firmly established and the mechanisms now so well understood that design of ful l-scale systems by mathematical models to suit particular inflow situations is feasible. Its future now lies not so much on better understanding of the mechanisms but on the ability to deal with the practical problems of operation, such as sludge bulking, pre-fermentation, back-up systems to cope with plant hiccups, and the problem of sludge disposal. His review dealt mainly with the work, mostly South African, on development of kinetic models. Only when a successful model had been developed in the early 1980s for nitrification/ denitrification was it possible to design to ensure the elimination of nitrate return to the anaerobic zone, by the UCT and the MUCT systems. For the first time, phosphorus removal could be studied without this complication. The influences of the short chain fatty acids, and of the cations Mg, K, and to a lesser extent Ca, were studied. 40

WATER August 1990

BNR system, operated for four months alongside a conventional AS system, under strict steady-state -laboratory conditions, using a synthetic feed of sodium acetate and pepone. The peak s of phosphorus concentration in the biomass in the reactors were clearly demonstrated, but it was noted that they comprised some 300Jo of orthophosphate as well as polyphosphate. Material balances across each of the stages indicate that protein synthesis and uptake into the biomass of the first anaerobic reactor was probably occurring. The phosphorus-accumulating microorganisms were by no means all Acinetobacter. A poster by Bob Seviour and Wayne Murdoch explained the expertise developed in BCAE in microscopic monitoring of the Bendigo BNR pilot plant over four years. This has enabled them to develop a range of methods including SEM for the identification of both bacteria and higher organisms in sludges, both BNR and conventional. The diversity of grazing organisms has been shown to be an indicator of the health of the treating system. The relationship between filamentous organisms such as Nocardia spp and Microthrix parvicella and the SVI may be a useful tool for operators, and the College offers an identification service to assist all operators of activated sludge plants experiencing problems.

Discussion on Theme 3 Jenkins, USA: In conventional AS, the

presence of filamentous organisms always seems to indicate anaerobic conditions at the head of the plant. Bayly, Monash: In experiments wi th

THEME 4: PROCESS RESEARCH AND DEVEWPMENT In a laboratory UCT system, an enhanced culture of phosphate accumulating organisms, mainly Acinetobacter spp., was developed. In the anaerobic zone, 253 mg/ L P was released, in the aerobic zone, 314 mg/ L was taken up, ie net uptake of 61 mg/ L, which was 0.12 mg P removed per mg of COD in the influent, some five times the norm. The biomass contained massive amounts of P, some 0.38 mg P/ mg MLVSS (In comparison, a mixed culture running on municipal sewage returned figures of 45 release, 57 uptake, ie 12 mg/ LP removed). Study of the enhanced culture enabled Wentzel et al in 1989 to formu late the kinetics of the P-removal process and to achieve excellent correlation between predictions and experimental results. Some simplifications (1990) enabled a steady state model to be developed for constant flow and load and the model has checked well with the results of 30 laboratory scale systems, of various configurations, which had been operated in the previous six years. However, there now remains an inconsistency in the nitrogen models. Application of the early N/ DN model would indicate that denitrification in the anoxic reactor should be too slow because there would be insufficient RBCOD available. In practice, this does not occur. For

activated primary tan~s, there may be a need to differentiate between the chemistry and the microbiology, since anaerobes will develop in such conditions. There is a danger that if we aim for maximum P removal we may make it hard for the nitrifiers. Kaye, U. NSW: Is there a relation between phosphate accumulation and SVI? Bayly: All BNR plants have settling problems, due to the filamentous bacteria. However, in pure laboratory cultures, clumped Acinetobacter settle so fast that it can be a nuisance. Greenfield, U. Qld: What is the likely stability of a phosphorus-accumulating plasmid? In other fermentation reactions, large recombinant plasmids are less stable than natural ones, but by repeated cycling in the fermenters and imposing stresses, one can enhance their stability. Bayly: The loss of activity noted in pilot plants may be due to loss of the plasmid, but it could be the loss of an element in the plasmid which is regulating the real mechanism. Millis: What keeps the plasmid in the cell? Bayly: We do not know. In plants there can be a loss of phosphate storage very suddenly, and it is all released in a surge. Greenfield: We also have noted a sudden shudder in operation, after which the plant recovers slowly. Jenkins: It is essential to keep a chemical precipitant handy! Barnard, South Africa: After one system collapse, the population shifted to Pseudomonas spp., which worked, though not as effectively as Acinetobacter. Daigger, USA: Physiological conditioning is as important , as genetic makeup. reasons not yet understood, the anaerobic zone must increase the rate of hydrolysis of SBCOD. For the most common systems, the design equation , fortuitously, still works, but in designs where N removal is not critical, but protection of the anaerobic reactor from nitrate is still essential, the model over predicts the nitrate reduction in the anoxic stage. Ekama then discussed the difficulties which can arise in practice, notably the sporadic bulking incidents. In general, filamentous organisms develop in low F/ M conditions, accentuated by the anoxic/ aerobic alternation. Although it is possible to apply temporary cures by chlorine or ozone on the RAS, when dosing ceases the filaments, inexorably, regrow. He discussed in detail the experiences of several BNR plants in South Africa, with regard to the anoxic mass fractions, retention times, nitrate and DO concentrations. He concluded that no specific answers were available as yet, and that if strict N and P standards are required, the plant design must incorporate large clarifiers. The need for back-up systems is still vital. Every mg of P removed requires the release of 7 mg and uptake of 8 mg. Any breakdown in aeration could result in a massive release of P from the sludge, equivalent to several days of previous hard work. Alw, as with all biological systems, hiccups do occur.

Back-up aerators and chemical precipitation are therefore necessary. Fe and Al salts can be dosed into the mixed liquor, but it has been noted that continuous additions lead to a gradual deterioration in biological removal. Bill Raper, CSIRO, did not accept this optimistic view of the simplicity of design. He said that in the Australian situation results so far are not so clear as in South Africa. This may be due to differences in the composition of the sewage, but he proposed that there may even be new mechanisms involved. He reviewed the experiences of all the pilot trials and full-scale plants which have operated in Australia, and stressed the necessity for adequate pre-fermentation, (which sometimes may take place in a septic sewer). In particular, the flexible Bardenpho pilot plant operating at the CSIRO field station at Lower Plenty yielded excellent results (2 mg/ L oxN, 0.4 mg/LP) only when the activated primary tank (APT) was operated with a mean sludge age of 30 days. This was well beyond the time required for optimum formation of acetate. However, it was noted that both soluble COD and TKN were significantly increased. The role of nitrate also requires better definition. The established correlation between nitrate and increasing leakage of phosphorus may represent a symptom rather than a basic cause. Bill referred to the work by Bayly et al, Monash, which investigated the lack of correlation between P removal and acetate in the APT effluent. It was observed that, contrary to the South African results, acetate additions to pure cultures of Acinetobacter did not stimulate phosphate release in the anaerobic zone. The absence of significant numbers of Acinetobacter in operating systems has frequently been noted, whilst other organisms, even grampositive species, have been found. The particular substrate may determine the dominant organism, and it is reported that Acinetobacter is formed only with acetate. A poster paper by Norman Pilkington, CSIRO, outlined a current project to analyse the chemistry of the action of the activated primary tank, operated at Lower Plenty on fresh sewage. HPLC and GC are being used. The common assumption that the significant factor is the generation of VFA is at variance with the Australian experience to date. This project will be accompanied by a parallel investigation of the microbiology, conducted at Monash University. The experience of the Hunter D W B pilot plant at Marmong Point was summarised in a poster paper by Cooksey & Cheng. The plant is supposed to be a replica of the Lower Plenty plant but it seems likely that the mechanisms are not the same, even though a long sludge age APT was operated. Differences in sewage characteristics would seem to be the cause, particularly the concentration of NFR. In conclusion, it would seem that much has yet to be learned, particularly for local conditions. John Bridger, CSIRO, described the test system he has evolved to assess the performance of the activated primary tank. At the "Models are no replacement for consultants." Dr Ekama.

Lower Plenty field station the raw sewage is sampled from a trunk sewer serving a short domestic catchment. The profile of both flow and COD is erratic and dramatically steep. Since the sewage is so fresh, an APT with a long sludge age has been proved to be essential, but its performance in relation to P-removal is not yet fully understood. Current techniques for estimation of RBCOD are time-consuming. The Lower Plenty technique is capable of unattended semi continuous data collection. In essentials it comprises a completely mixed reactor in a thermostat bath, fed continuously with incoming sewage. Air is bubbled in for 15 minutes, then air-off for 15 minutes. The oxygen uptake rate during the air-off period is plotted by a DO probe connected to a recorder. The apparatus was calibrated against the conventional 2-hour batch reactor using acetate. Two systems were used in parallel to follow the profiles of RBCOD in the incoming sewage and the output from the APT. The paper discussed the relationships between flows, turbidity, COD, TOC, RBCOD, and VFA. A paper presented by Grant Lockwood, of the Bendigo team, investigated the changes in the polyphosphates within the active cells, along with the transport of the cations, magnesium, calcium and potassium. They defined the shifts from high molecular weight polyphosphates to low molecular weight, and the subsequent transport of both magnesium and phosphate through the cell wall to the liquid phase during anaerobic treatment. In the course of their investigations, the pilot plant from which they drew their samples was modified from Modified UCT to Modified Phoredox. The latter system entails more recycle of nitrate to the initial anaerobic zone, and the team found a significant difference in the transport pathways. In the absence of nitrate there was an abundance of polyhydroxybutyrate, and a low concentration of low molecular weight polyphosphate in the cells. In the presence of nitrate, this situation was reversed. The inference was that more needed to be known about the mechanisms. The pilot plant evaluation trials conducted at the Bendigo CAE for the proposed BNR plant for Bendigo were summarised by Bruce Farnell. They encompassed the systems labelled Modified Phoredox, Johannesburg and Modified UCT, together with a further modification . This was termed the 'Farnell' process, and was proposed in case the incoming sewage had a high TKN/COD ratio. A portion of the feed to the JBG system was split to the secondary anoxic zone instead of all to the anaerobic zone, to compensate for the limited RBCOD, to improve the denitrification potential for the same anoxic mass fraction, and to reduce the nitrate returned to the anaerobic zone. All the processes were compared using the TKN/ COD ratio as a common reference. During the investigations, the incoming sewage characteristics were relatively constant, and TKN was supplemented to cover the range required. The UCT models (Wentzel, 1990) were used to correlate the results.

It was found that for nitrogen removal there was good correlation between predictions and experimental results, and all the processes denitrify similar quantities of nitrate for the same TKN/COD ratio. However, for phosphorus removal all processes removed more than the predicted quantity, and almost half the P-uptake occurred in the anoxic zone, which the current model does not consider. (UCT are working on this) . For high TKN/COD ratios, there was little to choose between the MUCT and Farnell processes. The others, though capable of good operation, were restricted to a smaller range of variation in influent TKN.

The poster presentations by the team from the Biotechnology Research Centre at BCAE outlined the procedures and strict attention to detail which were developed during their long term operation of the pilot plants to ensure that the data obtained were meaningful. A vital part of the investigation was the detailed characterisation of the Bendigo sewage. In order to place the full scale plant design on a firm basis, long-term and seasonal variations were modelled and sensitivity analyses produced. Further poster presentations explained their use of the UCT steady state models. A spreadsheet version was developed which enabled staff to rapidly compare different configurations against changing loads. The work relating to the design of the Bendigo Water Board plant finished in 1989, but a new 1200 L pilot planl is being installed in order to provide both on-going research into such processes, and hands-on training at undergraduate, post-graduate and operator levels. Computer models will provide a rational basis, both for education and for consultant applications in plant design and operation. However, it was emphasised that the assumptions built into such models must be understood. Peter Griffiths, SKP, who is involved in the chemical engineering design of the fullscale plant for Bendigo, used the UCT models to test another variation in the process system, to optimise the performance of the full-scale plant. When the diurnal variations in TKN are imposed on the system, he showed that in order to ensure complete denitrification at peak periods, the primary anoxic zone would be over-designed for the remainder of the day. He proposed a separate anoxic compartment or compartments in the recycle line. Using an adjusted version of the UCT ASP 87 program he ran simulations which demonstrated the point. In practice, all this involves in the full-scale plant is retention of the recycle to the first tank, but the provision of options for the influent sewage to be fed to either the first, second or third anaerobic tanks. The full-scale plant design enables a wide variety of plant configurations to be set up relatively simply, so that it should be possible to optimise its performance for a wide range of influent loads over the lifetime of the plant. WATER August 1990

THEME 5: FULL SCALE PLANTS

James Barnard, who must be regarded as the progenitor of the process, presented a 5000 word history of the development of the process. Unlike many developments, the observations started on full-scale plants, work then concentrated on pilot plants, and only then was attention paid to the microbiology and the mathematics. The eutrophication situation in the Vaal valley of South Africa in the early 1970s was critical. One receiving-water and storage, Haartbeestpoort Dam, was almost completely overgrown with macrophytes. Lime precipitation was being explored for control of phosphorus, and NIWR was actively studying nitrification/ denitrification. Âˇ During the operation of a four-stage pilot plant for the N / DN process, Barnard observed that excellent phosphorus removal was also occurring. The reason remained a mystery, until eventually it was discovered that there was a dead space in the second anoxic reactor which functioned as an anaerobic zone, with limited interconnection to the main flow. The rationale that for P removal there had to be P release from the sludge, in the absence of nitrates, followed by a greater P uptake in the aerobic reactor, was formulated, and work proceeded apace on the 'Bardenpho' system. More evidence stemmed from the fullscale N/ DN plant being operated at Johannesburg, when an electrical storm tripped some aerators. When the plant was restarted, it was noted that P was absent from the effluent. Deliberate experiments at the Alexandra plant confirmed the effect, and within two years, seven plants were operating or under construction. Whilst microbiologists and biochemists worked to explain the phenomenon, the engineers concentrated on analysing the reasons for the variations between the performance of the plants, since although N/ DN was generally good, P-removal fluctuated wildly. Gradually, a number of operational causes were identified. Barnard discussed the learning curve of the past fifteen years. In general, whenever a theory or a model has appeared to provide all the answers, experience at a full-scale plant poses more questions. This applies in particular to the condition of the incoming sewage, and the necessity or otherwise to adopt pre-fermentation. His conclusion is that theory and models, such as those developed at UCT, are most valuable, but only in the hands of an expert. Some American experience was presented by Glen Daigger, of CH2M Hill. The VIP flowsheet seems to be yet another variation on the Bardenpho sequence, but is a 'high rate' system, with an HRT of 3.7 hours in summer and 7 hours in winter, encompassing plug flow characteristics by designing each zone as at least two reactors in series. Recycle RAS and recycle nitrified solution are returned to the anoxic stages, and there is a recycle of anoxic sludge back to the initial anaerobic stage. 42

WATER August 1990

One full-scale and two pilot scale systems are in operation. The first system was investigated in a flexible retrofit installation to an existing AS plant at York River, Va., the others at Phoenix and Rock Creek. The septicity of the sewage to the first two plants was high, and good phosphorus removal was obtained, with complete nitrification. However, at Rock Creek, the fresh sewage did not perform so well. The performance of the Albury plant was summarised by Darryl McGregor. In 1987, the existing plant was augmented, and the NSW SPCC imposed draft licence conditions, to achieve TP of less than 4 mg/ L, TKN less than 15 mg/ L with NH 3 less than 2 mg/ L by December 1989. Augmentation was achieved by modifying an existing lagoon to a continuous activated sludge system. At the same time, it was decided to take advantage of the principles of BNR by the positioning of baffles and mixers to form a primary anaerobic zone. The floating and submerged aerator systems were operated under differing regimes to form an anoxic zone. A major improvement was the feeding of waste whey from a local dairy, transported by tanker, directly into the anaerobic zone to supplement the RBCOD. There is a strong inverse correlation between phosphorus removal and nitrification. The conclusions drawn from experience of plant operation were listed. Dale Kretser of West Wodonga summarised the performance of the oxidation ditch with appended anaerobic and anoxic tanks which was designed to achieve the Victorian EPA licence of TP less than 2 mg/ L and NH less than 5 mg/ L. This design was substantially cheaper than similar size plants. During the commissioning period generally good removal of both N and P was achieved, although a sludge bulking episode caused a major upset. The difficulties of running .a sophisticated plant on a highly variable industrial load were illustrated when in the latter months of 1990, a significant reduction in the input of RBCOD from a major factory interfered with the operation of the anaerobic zone, and TP exceeded the licence limits. The Ballarat South plant came on-line in 1988, designed to achieve TP of 2 mg/ L, 10 mg/ L NH 3 , but no limit on NO 3 â&#x20AC;˘ Bruce Price summarised the layout, which concentrated on phosphorus removal in the activated sludge system and relied on an upgrade of the existing bank of trickling filters to achieve nitrification. Experience demonstrated that although the plant achieved satisfactory P removal in average flow conditions, there was insufficient aeration to cope with peak flows . Samples aerated in the laboratory fell to zero TP, but at varying rates. Addition of acetate, magnesium and potassium had little or no effect on uptake rates. Another blower has been added, and a-2.3 ML balance tank has been installed to reduce the diurnal peak inflow from 26 ML/ d to the average of 16 ML/d.

Through 1989, the average TP in the effluent was 1.65 mg/ L, and reasons can be ascribed to most of the occasional peaks of up to 4 mg/ L. The modular Intermittent Extended Aeration process developed by the NSW PWD has been shown to be capable of producing low TP, less than 1 mg/ L, and low TKN, less than 5 mg/ L, without specific control of any process parameter. This surprising performance was observed at Bathurst STP in April 1987, following cessation of sludge lagoon supernatant return to the inflow. Fred Cozma detailed the investigations which followed this observation, which have demonstrated that the whole process can be achieved in the one tank, the fill, aeration, settling and decant cycles being controlled by timers. The operator adjusts the process by measuring the depth of the sludge blanket before the decanter operates, and by changing the aeration fraction. It was noted that 250Jo of the sewage was derived from a rising main with an average detention of 12 hours. It is still believed that the sludge supernatant has to be treated separately. One observation made during the investigation was that there was an increase in the TP in the water layer during the decant period. This could have been due either to release from the sludge or to short circuiting since sewage inflow was not stopped during this period, but merely baffled . Mervyn Goronszy, who left the PWD and set up operations in USA, described the system whiclY he had developed from this early observation. Termed the 'Fed-batch Reactor' it comprises twin tanks, operating in a somewhat similar system, but with alternating filling cycles to eliminate shortcircuiting. Positive recycling of sludge to the inlets is claimed to select for the required microorganisms and against the undesirable sludge bulking organisms. A.full scale plant in Dundee, Michigan was commissioned in September 1989 to retreat 3 ML/d AWDF, and has averaged 5 mg/ L COD, 5 mg/ L SS, 0.2 mg/ L NH 3 , 0.9 mg/ L TP. A further variation on the IEA system for small communities was described in a poster paper by Anderson of Aeration & Allied Technology, but aimed solely at nitrogen removal. A twin tank system is used, but operating in series, with RAS being pumped back to the lead tank. The company has also developed a gas-locked syphon decanter to reduce sludge pick up.

Bendigo Plant The full-scale plant at Bendigo is currently under construction, and was visited by delegates at the close of the conference. It features a fully compartmentalised layout, with channels allowing recirculation to any of a number of compartments so that it will be a relatively simple matter to modify the system. A full description of the plant will be published in a future issue of this Journal.

TOPIC 1. RELIABILITY Jonathon Crockett asked if BNR alone could attain and maintain TP <\t I mg/ L.

Jenkins: It would be a very brave operator not to have chemical back-up, even for a 2 mg/ L limit. Daigger: It depends on the BOD/ P ratio, but that can be modified. Barnard: In South Africa, one operator achieved 0.5 mg/ L for two years (with tertiary filters). Another had a limit of 0.15 mg/ L, and reached it for eight months, but not in the rainy season. In Canada, a plant attains 0.3 mg/ L by using 10 mg/ L alum as a tertiary polish. He summarised a survey he had made of plant performances, not as reported in the literature, but by personal query to the operators, and commented on the differences. Jenkins: For limits below I mg/ L tertiary fi ltration is essential, since any sludge suspended in the effluent carries its high polyphosphate loading. In Blue Plains, chemical additions are made to both the first and second stage AS, to achieve 0.12 mg/ L. Smalls: Why bother with concentration limits? A receiving water manager would be happy with a monthly mass load, rather than concentration limits. Mitchell: The economic cost of such low limits must be enormous. Discharges should be based on the ecology, and particularly if there is seasonal flooding, a single limit is nonsense. Jenkins: In USA, particularly in California, the regulators have the philosophy that if the engineers can do it, they damn well will have to. Price: For Ballarat, the Victorian EPA has estimated that an annual load of 2.2 tonnes would be tolerated by the receiving water. This is equivalent to an average of 1.3 mg/ L.

TOPIC 2. FWWSHEETS Bill Raper asked is there such a thing as an all-purpose flow sheet? For example, in 1988, he had put up a slide with 15 flow sheets, now there are about 30 with names attached! The simplest is a German system where RAS is simply returned to the primary settler instead of to the aeration tanks. At Lower Plenty the relatively simple 3-stage Bardenpho, coupled with an APT, attains 0.3 - 0.4 mg/ L TP for months at a time. Jenkins: If nitrogen removal is not critical, a sophisticated plant is not necessary. Ekama: For combined N and P removal, variations on the basic Bardenpho have proved to be necessary because of site specific factors such as hydrology of the sewers and ratio of domestic to industrial loads . It is vital to think it through , and do R&D if required, before you spend $20M on concrete. In USA, the planning process comprises 3-steps, the second one of preliminary design must explore all the options before the final design is approved. Daigger: There is an alternative to BNR, and that is chemical precipitation. Therefore, keep the BNR system as simple as possib le, and rely on chemical polishing. It is essential to forecast the effects of inflow changes (eg the effects of a specific industrial discharge).

THE WORKSHOP As mentioned in the preamble, one of the characteristics of this conference was the ready exchange of views. The Workshop itself was unstructured but fascinating, with the panel of experts being challenged by each other as well as by the audience. The session was chaired by James Barnard. Griffiths: The Bendigo plant design takes account of future inflow changes by the flexibility of the compartment design. Kaye: What is the panel's opinion of the IEA plant systems? Barnard: In South Africa we steered away from these because of our flash storm flows. In our simplest plants, such storm flows are diverted to the final clarifier. They are not suitable for 0.2 mg/ L limits. Cozma: In NSW we have 50 IEA plants on automatic control, they can deal with 10 times AWDF, and achieve 5 mg/ L TKN . Law: The development from the Pasveer system through the Bathurst Box has led to the installation at Port MacQuarie, which achieves a 10/ 10 effluent. We have found that the TKN is less than 5 mg/ L, and recently TP is down to 2-5 mg/ L. It can easily deal with storm flows. For lower limits, chemical polishing would be cheaper than a redesign. Ekama: I consider that the two controls of airflow and cycle time are insufficient for combined N and P removal. For South African inland waters, chemical addition is not appropriate because of the extra salinity. Trestrail: It is only a matter of time before the River Murray will come into that category. Nicolaisen: Note that ANZEC suggests a limit of 0.05 mg/ LP in receiving waters. A plant discharging 1.0 mg/ L requires 20:1 dilution, but in Australia, this is not always the case. (He mentioned that in a recent survey of urban run off, the second biggest contributor to TP was dog droppings .)

TOPIC 3. SLUDGE DISPOSAL Since all the phosphorus removed is contained in the sludge, how do we handle its ultimate disposal? This topic was addressed by Bruce Price who related the experiences at Ballarat South. The belt-pressed sludge was supposed to be stored on site for two years, then contracted out for agriculture. However, the Victorian Health Department have recently insisted on seven year storage, so a large acreage of pine plantation has had to be cut down to provide storage space. He contrasted that to the UK approach of seven day storage. He related the tale of the ultimate in ultimate disposal. In Baltimore in the summer of 1989, the sludge composting machine broke down. At 200 tpd, insufficiently stabilised sludge was loaded into a train, totalling 3000 tonnes, to be used for agriculture down south in Louisiana. The train broke down in a siding and it wasn't long before it hit the national headlines as the Poo Poo Choo Choo. Eventually it had to be returned to Baltimore, by which time a cold snap had frozen it solid. At the end of January it was finally bulldozed out to landfill.

The Japanese are developing a process for manufacture of a fertiliser from sludge. It is dewatered to 15%, both extra ammonia and phosphoric acid are added and the temperature rises sufficiently for pathogen kill.

TOPIC 4. PHOSPHORUS BANS The effect of phosphate bans in domestic detergents was discussed, following from an exhibit by the Australian Chemical Specialities Manufacturers Association which argued that the alternatives have their own disad van tag es. Jenkins: In USA there has been a steady decline in P input, from a range of 6- 8 mg/ L to 3-4 mg/ L. This has operated in the Great Lakes area for some years. Barnard: In South Africa, despite some restrictions leading to 20-30% reduction from an average discharge of 0.25 kg/ c/ annum, the typical input is 6-8 mg/ L. Detergent manufacturers are themselves limiting phosphate softeners because of the price. Jenkins: In USA squeeze liquids are becoming more popular; they have less phosphate but more surfactant. In Japan, input is now only 2 mg/ L.

TOPIC 5. BULKING AND FOAMING The tendency of activated sludge plants to devi::lop troublesome foaming and

bulking sludge is accentuated by the conditions necessary for BNR. Jenkins: The major cause is recycling of trapped foam. Installation of scum boards probably initiated the phenomenon; in the old days scum went out with the effluent. Barnard: Recent plant designs incorporate successive slight drops between the zones designed to take the foam out of the system. Cadee: WAWA have successfully eliminated foam from normal AS plants by a series of measures . (Control of sludge age, elimination of dead pockets, MLSS to lowest possible, avoiding over-aeration, and sometimes chlorinating the RAS.) Barnard: In South Africa it is a winter problem, but in USA it occurs in summer. Jenkins: In USA it is probably Nocardia, in S.A . it may be Microthrix. Note, there are many Nocardia spp and others which look similar. If serious foams get into the sludge digester, this can be really bad. Price: In Ballarat we have had to reduce digester sludge concentration from 3.5% to 2.8%, and now only 2.5%. Jenkins: The cause is probably Microthrix.

Editor: The interest aroused was so evident that the Journal commissioned the two papers, published in this issue, relating to conventional AS systems, including the problem of bulking sludge. The Technical Note by Scalzi may prompt other operators to pass on their experiences. WATER August 1990 43

The Microbiology of Bulking and Foaming in Activated Sludge by Linda L. Blackall INTRODUCTION An activated sludge sewage treatment plant is really an enrichment culture of microorganisms. The type of waste coming to the plant as well as the mode of operation selects the microorganisms that are capable of degrading the organic material in the influent. The biomass that does the degrading must settle in the secondary sedimentation tank making sedimentation a necessary attribute of the microorganisms and one that must be selected for. Bacteria play a primary role in the biodegradation process, although protozoa and rotifers also contribute to the purity of the plant effluent. The microbial composition of the biomass and its activity are thus the result of many complex factors: kinetic constants of every species (maximum growth rate and substrate saturation constants, mainly o.f organic substrate and oxygen), availability and nature of substrates, physical condition (pH, temperature, dissolved oxygen, mode of agitation , configuration of the plant) and substrate diffusional resistances of the floes. Furthermore, interactions between the various components of the microcommunities should be considered, such as predation, commensalism and competition. In recent years, much attention has been given to some specific bacterial populations which are important because of their useful physiological properties. The nitrifiers and denitrifiers are examples and more recently, phosphorus - accumulating bacteria have been studied. Bacteria that are undesirable to the activated sludge process have also attracted attention. Filamentous bacteria that cause separation problems such as bulking and foaming are examples of the latter. The types of solids separation problems that can be encountered in activated sludge are (van Niekerk, 1985): • Dispersed growth, where floe formation does not occur resulting in poor gravity settling of the activated sludge. • Deflocculation, where the floe disintegrates. This can be caused by operational conditions such as low dissolved oxygen (DO), decrease in pH and toxic shock loads. • When microorganisms produce a capsular material, generally polysaccharide in nature, it binds large amounts of water and prevents rapid settling. Capsule production is sometimes observed in the biological treatment of industrial waste water that is deficient in an essential nutrient such as nitrogen or phosphorus. The incorrect term, "zoogloeal bulking", is attached to this category. Pinpoint floes are small, compact, weak, roughly spherical floes, the larger of which settle rapidly but the smaller ones give rise to turbid effluents. Plants operating at a low food / microorganism (F/ M) ratio ( < 0.2 g Chemical Oxygen Demand (COD)/ g MLVSS/ day) are prone to this. When filamentous microorganisms proliferate in the activated slu_dge, bulking occurs. The filaments interfere with the settling and compaction of activated sludge either by physical bridging between floes or they allow the development of open, diffuse floes. A sludge volume index (SVI) of greater than 150 mg/ g constitutes a bulking sludge. When denitrification occurs in the sludge blanket of the clarifier, molecular nitrogen is released into the floes which are buoyed to the surface giving rise to bulking sludge. This problem is common in nitrifying activated sludge plants during hot weather and is exacerbated by long retention of sludge in the clarifier. The selective accumulation of filamentous bacteria, notably actinomycetes, on the surface of the aeration tank and the clarifier is termed scumming or foaming activated sludge. This can be a persistent problem and a difficult one to control. This paper will concentrate on bulking and foaming .

BULKING Activated sludge bulking occurs when the biomass settles slowly and compacts poorly in the Secondary Sedimentation Tank (SST). Predominantly this is because of high numbers of filamentous microorganisms in the mixed liquor (Pipes, 1987a; Pa)m et al., 1980; 44

WATER August 1990

Linda Blackall is a microbial ecologist with a PhD degree from the University of Queensland (1987); thesis entitled '54.ctinomycete Scum Problems in Activated Sludge Plants''. A CSIRO Postdoctoral Award allowed her to spend 18 months at the University of California, Berkeley, with Professor David Jenkins and 12 months at the University of NSW with Professor Kevin Marshall. A briefperiod was spent on phosphorus removal microbiology at the Water Research Institute, Rome, Italy in 1989. She is currently working at the Annual Research Institute, D.P.l, Qld.

Strom & Jenkins, 1984; Jenkins et al., 1986). The result usually is a highly turbid effluent which can escalate into an uncontrollable situation. Failure of the activated sludge to settle and compact in the SST will produce effects to the following aspects of wastewater treatment: • Uncontrolled loss of solids from the activated sludge process can produce an increased sewage substrate loading rate per unit mass of sludge in the system. Potential loss of nitrification due to the increased plant loading will result in a further deterioration in effluent quality. In extreme cases of continued loss of solids, a reduction in carbonaceous oxidation can occur. • Operational problems can also be expected in solids handling processes receiving a thin sludge which is difficult to concentrate and compact. • Tertiary treatment processes receiving a poor quality clarifier effluent from the activated sludge process will be affected. A high suspended solids concentration in the influent may require higher dosages for flocculation and disinfecting chemicals and result in accelerated clogging of filter beds. Although filamentous rnicroorganisll}s are present during bulking episodes, they are also required for effective sewage treatment as it is postulated that they form a network or backbone within the floe and onto which the floe forming bacteria attach (Jenkins et al., 1986). The filamentous microorganisms provide the floe with strength and allow it to grow into an irregular shape rather than spherical. However, the overgrowth of these filaments is deleterious. Many different filaments have been reported as causing sludge bulking, however, up until recent years, Sphaerotilus natans was thought to be the sole culprit. Eikelboom & van Buijsen (1981) and Jenkins et al. (1986) have developed an in situ identification key for filamentous bacteria in activated sludge. The filaments are differentiated on the basis of morphological characteristics, two staining reactions (Gram and Neisser stains) and a sulfur storage test. It must be noted that very few of these filaments have been isolated and identified in the laboratory and of those that have been identified, only a couple possess nomenclaturally valid names. The identification keys developed (vide supra) are therefore to morphological type, not to genus or species. Many surveys have been completed recently to establish the relative abundance of various filamentous microorganisms in bulking sludge. These surveys demonstrate that the same filamentous organism types are observed frequently and that approximately 10 types account for at least 90% of all bulking episodes. The relative frequency of occurrence of dominant individual filamentous organisms in bulking sludges varies significantly between geographical areas (see Table 1). These differences can be explained in terms of differences in waste type and strength and in plant operating conditions (Jenkins et al., 1986). Tuft (1985) has noted specific filaments that cause bulking in the Sydney region but which are not significantly reported in other parts of the world. This also highlights the regional specificity and the need for each plant to monitor the filaments present. It has also been determined that an individual filamentous microorganism may proliferate under specific growth conditions. Thus, although the situation is far from complete understanding, the presence of certain specific filamentous microorganisms can be correlated with operational and / or environmental parameters at

Table I: Comparison of dominant filamentous organisms in bulking sludge from several geographical areas. Ranking in Order of Prevalence

South

Hlamenwus Organism Nocardia sp. Type 1701 Type 021N Type 0041 Thiothrix sp.

USA' 1 2 3 4

Sphaerotilus natans Microthrix parvicella

6 7 8 9 10

Type 0092

Haliscomenobacter hydrossis Type 0675 Type 0803

Nostocoida limicola Type 1851 Type 0961 Type 0581

Beggiatoa sp. Fungi lype 09 14

Holland'

FRG 3 Africa 4 Australi~

8 I 3

2 6 19 7 I 4 3

12 13 14 15 16 17 18

3 8 9 6

12 10 8 18 15

2 9 1 3

4 2

4 8

5 2 7

10 7

Richard et al., 1984; Stro m & Jenki ns, 1984; 525 samples fro m 270 treatment plant s. Eikelboom, 1977; 1100 samples rrom 200 treatmen t plants. Wagner, 1982; 3500 samples from 315 treatment plants. 4 Blackbeard et at., 1986; 60 samples from 50 treat ment plants. ' Tuft , 1985; 6 treatment plants.

the activated sludge plant. A specific cause and effect relationship has been established for a few filaments and these include Type 1701, S. natans, "Microthrix parvicel!a·'' and Type 021N (Jenkins et al., 1986, Table 2). Because there is apparently a correlation between filament type and operating mode, identification of the filament is of prime importance followed by scrutinisation of previously correlated parameters. The original activated sludge process (Ardern & Lockett, 1914) was based on a batch fill-and-draw process, however, in practise many plants were built for completely mixed, continuous feed operation. Bulking activated sludge was a problem from the earliest time. Many hypotheses to explain the overgrowth of bulking filaments have been postulated, however, the scientific study of these theories is one area that needs attention at present. As described above, certain conditions have been correlated with specific filament types indicating physiological attributes that these filaments may have. Table 2: Dominant filament types as indicators of conditions causing activated sludge bulking (Jenkins et al., 1986) Suggested Causative Co ndition

Indicative Filament Types

Low Dissolved Oxygen

Type 1701 , Sphaerotilus na/ans,

low F/ M

Ha/iscomenobacter hydrossis Microthrix parvicella, Haliscomenobacter hydrossis, Nocardia

Septic Wastewater/ Sulfide Nutrient Deficiency

sp., Type 021N, Type 0041, Type 0675 , Type 0092, lype 0581, Type 0961, Type 0803. Thiothrix sp., Beggiatoa sp., lype 021N. Thiothrix sp., Sphaerotilus natans, Type 021N, Possib ly Haliscomenobacter hydrossis, Type 0041 and lype 0675. Fungi.

Low H

In 1973, Chudoba et al. (1973) presented a kinetic selection theory in mixed cultures. The theory is based on the Monod equation presuming different growth constants Ks and µmax for different species and, consequently, their different specific growth rate relationships to substrate concentration. The theory has proven to be accurate for bulking due to S. natans (Van den Eynde et al., 1982) and Haliscomenobacter hydrossis (Van Veen et al., 1982). According to the theory, microorganisms with low Ks and µmax values should prevail in aeration systems with low substrate concentrations. In completely mixed activated sludge systems, the soluble COD concentration is similar throughout the aeration basin and is low at all times due to dilution of influent with Return Activated Sludge (RAS). Based upon the theory of Chudoba et al. (1973), organisms with low Ks values will be favoured over those with higher Ks values. This is because of the high substrate affinity that organisms with low Ks values have. It has been postulated, and proven in a few cases, that bulking filamentous organisms have low Ks values for carbon sources and hence possess a selective advantage in situations where the ambient COD concentration is low. When a substrate gradient is introduced, by the incorporation of intermittent feeding of influent or by the inclusion of a selector, the high SVI of the bulking plant can be reduced with eradication of the bulking incident (Jenkins et al., 1986; van Niekerk, 1985). A selector is a small basin located upstream of the aeration tank where the_sewage

and RAS are mixed (van Niekerk et al., 1987b). Substrate gradients allow the RAS to be presented with a higli COD concentration for a short period of time. The dilution of sewage into the large aeration basin does not occur in this situation thus obviating the advantage that the low Ks bulking filaments have, which allows organisms that have high COD uptake and storage rates to be selected for. Ideally, most of the soluble COD should be removed within a short time after RAS and sewage meet. The substrate that is rapidly taken up is not all metabolised instantly since a large fraction of it is stored by the organisms. The RAS sewage mixture then passes into the aeration tank where metabolism of the stored substrate occurs. Control of bulking in low F/ M plants has been obtained consistently in laboratory scale plants and in full scale activated sludge plants by empirical measures such as the introduction of batch or plug flow characteristics to the aeration basin, intermittent feeding of wastes, compartmentalisation of the .aeration basin, use of selectors or fed-batch operation. All of the above measures produce a carbonaceous substrate gradient in the aeration tank or a high substrate concentration at the point where RAS and influent waste enter the aeration tank (Jenkins et al., 1986). The use of selectors to produce a substrate gradient also controlled the growth of filamentous organisms when the F/ M ratio was close to l g COD/g MLVSS/day (Chudoba et al., 1985), hence this potential control strategy is not limited to low F/ M plants. Design of selectors is still largely empirical and Jenkins et al. (1986), who have been working vigorously on the use of selectors to control filamentous bulking, have the following important comments to make: • The uptake of soluble (0.45 µm filterable) organic matter in the selector appears to be of importance in determining whether the system will control bulking. A significant fraction of this component must be removed before passing to the aeration tank. • Substrate uptake rates and DO uptake rates in selector activated sludge plants is higher than in completely mixed plants. This can be used as an indication that the selector mode has been reached and that bulking may be alleviated. • Because batch fill-and-draw systems reveal the same uptake information as continuously fed systems do, then in the laboratory scale testing to determine the uptake rates; the former system can be set up in preference to the latter which is more difficult to operate. • The energy and electron acceptor requirements in selectors are poorly understood, however, it atipears that the rate of soluble COD removal is slower under anoxic conditions than under aerobic conditions. Anaerobic selectors which are exploited in enhanced biological phosphorus removal processes can also be contemplated as an option for a selector. As well as the kinetic selection theory proposed by Chudoba et al. (1973) to explain filamentous bulking, others have also been postulated. These include: • The surface area:volume ratio hypothesis as described by Pipes (1967) which argues that filamentous microorganisms possess a larger surface area than floe-forming microorganisms which are aggregated into floes. Direct contact between filamentous organisms and growth-limiting substrate is said to be occurring whereas diffusion into activated floes must occur for the floe formers to access the substrate. • The accumulation-regeneration theory recognises the ability of some microorganisms to rapidly accumulate and store substrate when exposed to high substrate concentrations in the growth environment (Grau et al., 1982) Van den Eynde et al., 1982). This of course is considered in the kinetic selection theory of Chudoba et al. (1973) and was discussed as the mode of microbial selection after a substrate gradient was imposed on the sewage feed (vide supra). It is hypothesised and proven in at least a few cases (van Niekerk, 1985) that flocculant microorganisms have higher rates of excess substrate accumulation and storage than filamentous microorganisms. The competitive advantage provided to flocculant microorganisms can, however, only be maintained if these organisms are allowed a sufficiently long period of endogenous respiration to regain their full substrate accumulating capacity. Grau et al (1982) suggest that mixed cultures dominated by flocculant microorganisms can only be obtained in systems where at least 50% of the stored substrate is oxidised before the next substrate presentation occurs. The starvation-resistance hypothesis again is an extension of the kinetic selection theory of Chudoba et al. (1973). The hypothesis suggests that the entire activated sludge microbial population is comprised of three classes of model microorganisms: WATER August 1990

• A fast growing floe-forming microorganisms with moderate starvation resistance. • A fast growing microaerophilic filamentous microorganism with limited starvation resistance. • A slow growing filamentous microorganism with high starvation resistance. The physiological and growth characteristics of these organisms are detailed in Table 3. Table 3: Physiological characteristics of proposed model activated sludge micro-organisms) Chiesa & Irvine, 1982) Floe-forming

Physiological Property

I. Subsrrare Affinit y 2. Maximum Growth Rate 3. Endogenous Respiration Rate 4. Starvation Resisrance 5. Ability to Accumulare Storage Prods 6. Rate of Storage Product Accumulation

Microorganls

Filamentous Microorganisms Associated with

High Orga nic Loading

Low Organic Loading

Low High Moderate Moderate

High Moderate High Low

High Low Low High

V. High

Moderare

Low

V. High

Moderate

Low

Control of bulking can be specific, after the cause and specific filament types are known, or non-specific. Non-specific control is generally via dosing chlorine or hydrogen peroxide at the point where maximum killing effect will be elicited (Jenkins et al., 1986). Target SVI values are decided and remedial procedures are initiated once these targets are not met. For this, regular microscopic monitoring is of enormous benefit as not only can the specific filaments be identified and filament type changes observed, but also this monitoring can be diagnostic tool for bulking such that early remedial measures can be implemented (Jenkins et al., 1986; Tuft, 1985). Specific control strategies require more knowledge and also microscopic monitoring. Jenkins et al. (1986) cite case studies for control.

FOAMING Foaming in activated sludge was first reported in the literature in 1969 (Anon, 1969). Since that time it has escalated in severity and incidence. The biological foam is thick, viscous, grey to brown cream in colour, quite heavy in consistency, up to 30 cm deep or deeper and water sprays applied to the surface have little effect upon its collapse. It generally initiates on the aeration tank surface in a quiescent zone but escalates to become thicker and spreads in its coverage. The secondary sedimentation tank inlet well can overflow and complete coverage of the SST is possible. Obviously this floating material can leave the plant with the effluent and thus give rise to high levels of Biochemical Oxygen Demand (BOD) and Suspended Solids (SS) in this component. The foam can totally cover the aeration tank and SST and may even overflow the sides of these tanks (Lechevalier, 1975; Hartley, 1982; Pipes, 1976b; Sezgin et al., 1988). Some specific problems that have been attributed to scum are: • Extra housekeeping needed because of overflows, wind blown foam and accumulations, • Blockage of scum removal systems, • Odour generated by drying scum, • Health risks associated with pathogens in wind blown scum, • Reduction in oxygen transfer at the surfaces of mechanically aerated basins, • Increased BOD and SS in the plant if scum reaches the effluent, • Increased operating costs associated with scum control and • Associated anaerobic digester foaming problems (van Niekerk et al., 1987a) When this biological foam or scum is examined microscopically, many Gram-positive, branching, filamentous bacteria are commonly observed (Lechevalier, 1975). The organisms associated with foaming in activated sludge plants are Nocardia amarae (Lechevalier & Lechevalier, 1974; Jenkins et al., 1986); "Microthrix parvicella" (Hart, 1985; Black beard et al., 1986; Jenkins et al., 1986; Goddard & Forster, 1986; Baumann et al., 1986; Blackbeard et al., 1988); Rhodococcus rhodochrous (Lechevalier, 1975; Lemmer & Kroppenstedt, 1984); 1ype 0092 (Hart, 1985; Blackbeard et al., 1986; Blackbeard et al., 1988) "Gordona aurantiaca" - now Tsukamurella paurometabolum (Lemmer & Kroppenstedt, 1984;_Collins et al., 46 WATER August 1990

1988); Type 0041, Type 0803 and Type 0675 (Hart, 1985), Type 1851 (Baumann et al., 1986); Nocardia pinensi§ (Blackall, 1987) and "Nostocoida limicola" III (Tuft, 1986). In Australia, Nocardia amarae and Nocardia pinensis are the predominant foaming organisms (Blackall, 1987; Seviour et al.,1990) followed by "Microthrix parvicella'; Type 0092, Type 0914 and Type 0041/0675 (Seviour et al., 1990). Of the above organisms 4 are actinomycetes whilst the remaining 8 are filaments that are more traditionally associated with bulking activated sludge. Selective accumulation in the foam phase over the mixed liquor phase is an important feature in the foaming. Blackbeard et al. (1988) report the incidence of many filaments in the foam of nutrient removal plants, however, only Nocardia and "Microthrix parvicella" were selectively accumulated in the upper floating phase. Surveys into foaming activated sludge have not been able to unequivocally identify any operational feature, design configuration, influent constituent or weather pat em that precipitates foaming (Blackall, 1987). However, operation at long sludge ages or with high MLSS levels does tend to exacerbate the problem (Pipes, 1978b; Fleissner & Foes, 1980). The problem is widespread and has been reported in Europe (Lemmer & Kroppenstedt, 1984) the United States (Jenkins et al., 1986, Pitt & Jenkins, 1990), Japan (Hiraoka & Tsumura, 1984), the United Kingdom (Goddard & Forster, 1986), South Africa (Blackbeard et al., 1988) and Australia (Blackall, 1987; Seviour et al., 1990). Unfortunately, little definitive information is known about the physiology of Nocardia amarae, the most predominant foaming organism. It is known that this organism can use a range of long chain carbon compounds as a sole carbon source (l.echevalier & Lechevalier, 1974; Sezgin et al., 1988; Kampfer, 1988). Their growth rate and substrate affinities are not known with these substrates. Lemmer & Baumann (1988) reported that the addition of hydrophobic substrates to laboratory scale activated sludge experiments enhanced the competitive success of N. amarae, particularly in low F/ M plants. This is in line with observations by wastewater personnel who note that, generally, high levels of grease and oil enhance foam production . Blackall (1987) found that N. amarae has a J,Lmax of 0.1 hr - 1 when acetate was the sole carbon source and that the yield was 0.5 g cells/ g acetate. This is in the range bf reported values for bacteria from completely mixed activated sludge plants (van Niekerk, 1985). Baumann et al. (1988) reported a remarkable value for Ks of 675 ± 124 mg/ I when fructose in a complex medium (peptone and yeast extract) was usell for growth with N. amarae. The J,Lmax in this system was 0.087 ± 0.008 hr - 1• Chemostat methodology was used in the latter experiments, however, BOD was used for substrate detection and steady-state was assessed only by optical density. The exceptionally high Ks value reported is 4 orders of magnitude higher than values commonly reported (van Niekerk, 1985). In another study, the Ks value for N. amarae was found to be 0.5 mg acetate/ I (one strain ASF3) and 2.5 mg acetate/ I (one strain ASACl)(Blackall et al., 1989). Table 4 details some further physiological data and shows some comparisons with other activated sludge organisms (Blackall et al., 1989). Blackall & Marshall (1989) reported that cells of N. amarae are more than 900Jo hydrophobic when the bacterial adhesion to hydrocarbon test was used. Furthermore, this latter publication reports that N. amarae produces a released surfactant which is required for foam initiation in a laboratory scale foaming protocol. The importance of cell hydrophobicity was highlighted as when the cells were made hydrophilic, transport of the cells to the bubble phase upon aeration did not occur. Other authors (eg Hao et al., 1988) rely upon conjecture when trying to explain the phenomenon of nocardial foaming. They discuss properties that other nocardioform actinomycetes may possess and translate that information to N. amarae. However, when the information that is known about N. amarae is examined, it appears that the hydrophobicity (Mori et al., 1988; Blackall & Marshall, 1989) of the organisms is important to foam stabilisation and that surfactants (Jenkins & Ho, 1988; Khan & Forster, 1988; Koizumi et al., 1989; Blackall & Marshall, 1989) are required to initiate the foam. These latter observations are in agreement with adsorptive bubble separation theory for foam stability. N. amarae can grow on acetate, a model soluble COD substrate, as rapidly as can bacteria from completely mixed plants (Blackall et al., 1989). The role that the insoluble substrates take in the competitive growth of N. amarae may be very important (Lemmer & Baumann, 1988) and more knowledge is required in this area.

Table 4 Physiological test results for some activated sludge organisms taken from va n Niekerk (1985) and Shao & Jenkins (1988') and for Nocardia amarae (ASF3 and ASACI) from Blackall et al. (1989)

O rga nism

Half Sa tu rati on Co nstant for: Oxygen (mg L - 1)

Maxim um Aceta te

Maximum

U ptake (mg g - 1 min - 1)

Respiralion Rate (mg g - 1 min - 0)

Nitrate Reduction Rate (mg g - 1 min·')

Acelate (mg L- 1)

4.9±0.9 1. 8+1.4 5 .3 4.2

NDd

0 .5 1 0.38

13.2 ±4 .2 4.4±3 .0 15.7 12 .6

ND 0.33 8 .3x10 -•

ND 0.30 0,07

ND 0.12 0.06

0.42 0.41

8.8 5.9

3.1 1.8

3 .8x10 -• 2.6x !0 - 3

0.50 2.60

0.13 ND

Yield (g g - •J

From Shao & J e nkin s (1988) a n d van Niekerk (1985)' ASl • b

AS2, Zoogloea ramigera Type 021N

5.4 2.5 5.5 3.8

0.39

o.36

Nocardia amarae from Blacka ll et al. ( 1989): ASF3 ASAC I

2.8 3. 1

itrate reduction rate from Shao & Jenkin s (1988) , remainder from van Niekerk (1985). b AS!, Average of bacteria isolated from an aerobic selector activated sludge plan1. '" AS2, Average of bacteria isolated from a co mpletely mixed activated sludge plant. 0 D, Not done.

•

The control of foaming, as with bulking, is largely by non specific means including chlorination of the foam (Blackall, 1987), reduction in sludge age and MLSS levels (Sezgin & Karr, 1984; Sezgin & Karr, 1986), return of anaerobic digester supernatant to the aeration tank (Lechevalier, 1975), dosing of influent with ferric chlo ride (Dhaliwal, 1979), modification of plant configuration (Hiraoka & Tsumura, 1984), physical removal (Nelson, 1981; Hoffmaster, 1981 ; Ludwig, 1981), exploitation of flotation (Pretorius & Laubscher, 1987), dosing of the plant with pickle liquor (Heath & Chan, 1985) and exposure of mixed liquor to periodic anoxia (Gasser, 1987). Pitt (1988) found that 660Jo of activated sludge plants in the US were affected by foaming incidents. He discovered that N. amarae growth was MCRf and temperature dependent and that bench scale testing showed that anaerobic selectors were success ful in reducing Nocardia levels. Aerobic selectors for the control of Nocardia foaming have been the foc us of attention by Cha (1989) and physiological data indicate that either anaerobic or anoxic selectors could help in controlling nocardial foaming (Blackall et al., submitted).

KEY REFERENCES This is a short list of recent review articles that may be of most use to the beginning reader. Cha mbers, B. & Tomlinson, E.J. Bulking of Activated Sludge: Preventative and Remedial Methods. E llis Horwood Ltd, Chichester. 1982. Lechevalier, H.A. 1975. Actinomycetes of Sewage Treatment Plants. U.S. Dept of Commerce National Technical Information Service Report No PB 245 914 Jenkins, D., Richard, M. & Daigger, G.T. Manual on the Causes and Control of Activated Sludge Bulking and Foaming. Water Research Commission of So uth Africa, Pretoria. 1986.

REFERENCES Anon. 1969. Milwaukee mystery: unusua l operating problems develop. War Sewage Works 116, 213. Ardern, E. and Lockett, W.T. 1914. Experiments on the ox idation of SC\vage without the aid of filters, J Soc Chem Ind 33, 523-539. Baumann, M., Lemmer, H. and Popp, H. 1986. Abstract s of the Fourth International Symposium on Microbial Ecology, Yugoslavia. Baumann, M. , Lemmer, H . and Rei s, H . 1988 Scum act inomycetes in sewage treatment plants Part I. Growth kinetics of Nocardia amarae in chemostat culture. Wat Res. 22, 755-760. Blackall, LL 1987 Actinomycete Scum Problems in Activated Sludge Pla nts. Ph.D. Thesis, Univ of Qld., Aust. Blackall, L.L. and Marshall, K.C . 1989. The mechanism of stabilisat ion of actinomycete roams a nd the preve ntion of foa mi ng under laboratory co ndition s. J Ind Microbial 4, 181- 188. Blackall , L.L, Tandoi, V. and Jenkins, D. 1989. The physiology of Nocardia amarae isolated From foaming activated sludge. In Proceedings of the 13th Federal Convention of the Australian Water and Wastewater Association, March 6-10. 1989, pp 461-464. Blackall , LL., Tando i, V. and Jenkins, D. submitted. Continuous cu lture studies with Nocardia amarae from act ivated slud ge and their implicat io ns for Nocardia foa ming control. J Wat. Poll. Com. Fed. Blackbeard, J.R. , Ekama, G.A. and Marais, G.vR. 1986. A survey of bulking and foaming in activated sludge plants in South Africa. Wat. Poll. Co11t. 85, 90-100. Blackbeard, J.R., Gabb, D.M.D., Ekama, G.A. and Marais G.vR. 1988. Identification of filamentous orga ni sms in nutrient remova l activated sludge plants in So uth Africa. Water SA 14, 29-33. Cha, D. 1989. Department of Civi l Engineering, University or California, Berkeley, 94720, California, U.S.A. PersonaJ Co mmunication. Ch iesa, S.C. and Irvine, R.L. 1982 Growth and co ntrol of fi lamentous microbes in act ivated sludge - an integrated hypothesis. 55th Annual Conj. Wat. Poll. Corot. Fed., St. Louis, MO. C hu doba, J .• G rau. P. and Ottova, V. 1973. Co ntrol of ac tivated slud ge filamentous bulking. II . Selection of microorganis ms by means of a selector. Wat. Res. 1, 1389- 1406. Chudoba, J., Chech, J.S. Farkac, J. and Grau, P. 1985. Control of activated sludge filamentous bulking - exper im ental verification of a kinetic selection theory. Wat. Res 19, 191-197. Collin s, M.D., Smida, J., Dorsch, M. and Stackebrandt, E. 1988. Tsukamurella gen nov. harbouring Corynebacterium paurometabolum and Rhodococcusaurantiacus. Int. J. System. Bacterial. 38, 385-391. Dhaliwal, B.S. 1979. Nocardia amarae and activated sludge foam ing. J Wat. Poll. Cont. Fed 5 1, 344-350. Eikelboom, D.H. 1977. Id entifi cation of filamentou s organisms in bulking activated slud ge. Prag i11 Wat Tech 8, 153- 162. Eikelboo m. D.H. and van Buijsen, H.J.J. 1981. Mi croscop ic Sludge Investigat ion Manual. IMGTNO Report A94A, Delft, Netherlands. Fleissner, J.T. and Foes, G.W. 1980. Discussion: Nocardia amarae and activated sludge foaming . J Wat. Poll. Cont. Fed. S2, 2594. Gasser, J.A. 1987. Control of Nocardia scu m in act ivated sludge by periodic a noxia. J. Wat. Poll. Cont. Fed. 59, 914.

Gerardi , M.H. 1986. Control of actinomycete foam and scum production. Public Works January, 82-83. Goddard, A.J. and Forster, C.F. 1986. Surface tension of activated sludges in relat ion to the fo rm ation of stab le foams. Microbios 46, 29-43. Grau, P., Chudoba, J. and Dohanyos, M. In Bulking in Activated Sludge: Preventative and Remedial Methods. B. Chambers & E.J. Tomlinson, Eds., Ellis Horwood Ltd., Chichester, 1982, pp 111-127. Hao, O.J., Strom P.F. and Wu, Y.C. 1988. A review oft he role of Nocardia-like filaments in activated sludge foam ing. Water SA 14, I05- l1 0. Ha rt , M.A. 1985. Scum formation in a nutrient removing activated sludge plant. Water SA 11 , 171-178. Hartley, K.J. 1982 Nocardia in activated sludge plants. Paper presented to 1982 Queensland local Authority Engineers Conference. 23 pp. Heath, C.W. and Chan, J. 1985. Laboratory and plant sca le trials usi ng pickle liquor for Nocardia control in activated slud ge plants. In Proceedings of the /Ith Federal Convention of the Australian Water and Wastewater Association, Apri l 28- May 3. 1985, pp 346-353. Hi raoka, K. and Tsumura, K. 1984. Suppression ofactinomycete scum production - a case study at Senboku wastewater treatment plant, Japan. Wat Sc Tech 16, 83- 90. Hoffmaster, W.W. 1981. Foam contro l in aerat ion tanks Pan II. Water/ Engineering and Management May, 34 & 39. J enkin s, D. and Ho, H. 1988. The effect of surfacta nt s on Nocardia foaming in activated sludge. Sanitary and Environmental Engineering, Departme nt of Civil Engineering, UC. Berkeley, Report to Victor Valley Wastewater Reclamation Authority. Jenkins, D., Richard, M. and Daigger, G.T. Manua l on the Ca uses and Contro l of Activated Sludge Bulking and Foaming. Water Research Co mm ission of Sout h Africa, Pretoria. 1986. Kampfer, P. 1988. Department of Hygiene, Technica l University, Amrumerstr. 32, Berlin, D-1000, FRG. Personal Communication. Khan, A.R. and Forster, C.F. 1988. Biosurfactant production by Rhodococcusrubra. Environ. Tech. Lett. 9, 1349-1360. Koizumi, J. , Takeda, M. and Mori , T. 1989. Extrace llular product of Nocardia amarae induces bacterial cell noccu lation. FEMS Microbiot. Lett. 57, 61 - 64. l.echevalier, H .A. 1975. Actinomycetes of Sewage Treatment Plants. U.S. Dept of Commerce National Technica l Information Se rvice Repon No PB 245914. Lechevalier, M.P. and Lechevalier, H.A. 1974. Nocardia amaraesp. nov., an ac1 inomyce1e co mm on in foam ing activated sludge. Int. J System. Bacterial 24, 278-288. Lemmer, H. and Baumann, M. 1988. Scum actinomycetes in sewage treatment plants - Part 2. The effect of hydrophobic substrate. Wat. Res. 22, 761-764. Lemmer, H. and Kroppenstedt, R.M. 1984. Chemotaxonomy a nd physiology of so me actinomycetes iso lated from scumm ing activated sludge. System and App Microbial S, 124-135 . Ludwig, K.L. 1981. Foam co ntrol in aerat io n ta nks Part III. Water/ Engineering and Management June, 34 & 39. ' Mori, T., Sakai, Y., Honda, K., Yano, I. Hashimoto, S. 1988. Stable abnormal form in act ivated sludge process produced by Rhodococcus sp. with strong hydrophobic properties. Environ, Tech. Lett. 9, I041 - 1048. Nelson, E.D. 1981. Foam control in aeration· ta nks Pan I. Water/Engineering and Management A pril, 63-64. Palm, J.C. Jenkins, D. and Parker, D.S. 1980. Relationship between organic loading dissolved oxygen concentration and sludge sett labilit y in the co mpletely mixed activated sludge process J Wat. Poll. Co11t. Fed 52, 2484-2506. Pipes, W.O. 1967. Bulking or activated sludge. Advance in App Microbio/ 9 185-191. Pipes, W.O. 1978a. Microbiology of activated sludge bu lking . Advance in App Microbio/24 85-127. Pipes, W.O. 1978b. Actinomycete scum production in activated sludge processes. J Wat. Poll. Cont. Fed. 50 628- 634. P itt, P. 1988. Causes and co ntrol of Nocardia foa ming in activated sludge. Ph.D. Thesi s, Univ. of Californ ia, Berkeley, U.S.A. Pitt, P. and Jenkins, D. 1990. Causes and control of Nocardia in activated sludge. J Wat. Poll. Com Fed 62 143-150. P retorius, W.A. and Laubsche r, C. J.P. 1987. Contro l of bio logical scu m in act ivated sludge pla nt s by means of selective flotation. Wat Sc Tech 19 , 1003- 1011. Richa rd, M.G. Shimizu , G.P. and Jenk ins, D. 1984. The growth physiology of the filamentous organism Type 021N and its significance to activated slud ge bulking. J Wat. Poll. Cont Fed 57, 1152- 1157. Seviour, E.M., Williams, C.J .• Seviour, R.J. Soddell, J.A. and Lin dea, K.C . 1990. A survey of fil amentous bacte ri al popul ati ons and foaming activated sludge plants in eastern states of Australia. War. Res. 24, 493-498. Sezgin, M. and Karr, P. 1984. Controlling scum by lowering sludge age. Op Forum August , 28-29. Sezgin, M. and Karr, P. 1986. Control of actinomycete sc um on aerat io n basi ns a nd clarifiers. J Wat. Poll. Con t. Fed 58 972-977. Sezgi n, M. and Karr, P. and Lecheva lier, M.P. 1988. lsolat ion a nd identification of actino mycetes present in activated sludge scum. p. 41.1-41.7 Proceedings of the (AWPRC Conference on the Microbiology of Water and Wastewater, Newpon Beach, CA. February, 1988. Shao, Y.-J. and Jenkins, D. 1988 . The use of anoxic se lectors fo r the contro l of low F/ M act ivated sludge bulking. P roceed ings of the IAWPRC Conference. Brighton, Engla nd, July, 1988. Strom, P.F. and Jenkins, D. 1984. Ide ntifi cation and signifi cance of filamentous microorga ni sms in activated sludge. J Wat. Poll. Com. Fed 56 449-459. Tuft, R.G. 1985. Application of monitoring for filamentous organisms for control of activated sludge plants. In Proceedings of th e //th Federal Convention of the Australian Water and Wastewater Association, Apri l 28-May 3, 1985, pp 354-360. Tuft , R .G. 1986. Gutterid ge Hask ins and Davey Pt y Ltd ., 39 Regent St., Railway Square, N.S.W., 2000. Personal Comm unication. Yan den Eynde, E., Houtm eyers, J. and Yerachtert, J. 1982. In Bulking in Activated Sludge: Preventative and Remed ial Methods, 8. Chambers & E.J. Tom lin son , Eds., Ellis H orwood Ltd., Ch ichester, 1982, pp 128-146. van Niekerk, A.M . 1985. Com petiti ve Growth of Flocculant and Filamentous Microo rganisms in Activated Sludge Systems. Ph.D. Thesis, Univ of Cal., Berkeley, U.S.A . van Niekerk, A.M., Kawahigashi, J., Rei chlin, D. , Malea, A. and Jenkins, D. 1987a. Foaming in anaerobic digesters-a survey and laboratory invest igation. J. Wat. Poll. Cont Fed S9, 249-253. va n Niekerk. A.M. Jenkins, D. and Richard, M.G. 1987b. The competitive growth or Zoogloea ramigera and Type 021N in activated sludge and pure culture - A model for low F:M bu lking. J Wat. Poll. Cont Fed 59, 262. Van Veen, W.L., Kru l, J.M. and Bu lder, C. J.E.A. 1982. Some growth parameters of Haliscomenobacter hydrossis, a bacterium occurring in activated sludge. Wat Res 16, 53 1-534. Wagner, F. In Bul ki ng in Act ivated Slud ge: Preve ntat ive and Remed iaJ Methods, B. Chambers & E.J. Tom li nso n, Eds., Ellis Horwood Ltd., Chichester, 1982, pp 29-46.

WATER August 1990

Chlorine Dioxide -

The American Experience

Dr Ronald Gates, of the Rio Linda Chemical Co. of Sacramento, USA addressed a series of seminars in February and March, sponsored by Albright & Wilson Pty Ltd, in conjunction with Branches of AWWA. In view of the increasing interest in chlorine dioxide as a disinfectant for potable water, as well as for other applications (see Water, April, p.43) the following summary of his talk has been abstracted from a report by Mark Pascoe and Terry Loos, in the Queensland Branch Newsletter.

CHLORINE DIOXIDE The compound is a greenish-yellow gas at normal generating concentrations and is reasonably soluble in water (0.3-1 O'/o ). It does not react with water in the same manner as chlorine, but remains in true solutions. Being volatile, it can be stripped out by aggressive aeration. Because it is sensitive to temperature and pressure, chlorine dioxide cannot be shipped in bulk and must be generated on site. One method of generation is through the reaction of sodium chlorite ·with chlorine solution. Chlorine dioxide disappears from water within six hours of its initial formation, and chlorite ion is formed as a major reaction product (480'/o) with 220'/o reduced to chlorate, and 280'/o further reduced to chloride. Inefficient utilisation of the sodium chlorite in the initial generation will obviously lead to higher concentrations of chlorite in the treated water.

ADVANTAGES OF CHLORINE DIOXIDE • It is a very effective disinfectant. In terms of CT ratings (ie concentration x time to eliminate target microbes) it is second only to ozone; ozone -chlorine dioxide -chlorine -chloramine • It is effective against cysts, such as Giardia. • It does not form THMs (trihalomethanes) and forms far less TOX (total organic halides) than chlorine. • It reduces formation of THMs by any subsequent chlorination, since it oxidises the precursor compounds. • It oxidises taste and odour compounds better than chlorine. (Chlorine may accentuate some tastes and odours). • It produces less aldehydes and ketones than ozone. • It does not react with ammonia. • It is effective over a wide range of pH, and can therefore be used at more points of the treatment proces&. • It oxidises sulfides all the way to sulfates. • It aids coagulation. Finally, it is currently about one quarter of the cost of ozone. Previous misconceptions regarding high cost could be attributed to the efficiency of the generators. As the efficiency of conversation of NaCIO 2 to CIO 2 has improved from ca. 80 to 970'/o, the cost of generation has come down.

DISADVANTAGES OF CHLORINE DIOXIDE • It can explode if mishandled. • There is no established way to measure chlorine dioxide production or purity. • An excessive level of chlorine ion in potable water has an adverse health effect (similar to that of nitrate). • The measurement of the residual ions is difficult. There are as yet no light-weight test kits for chlorine dioxide, chlorite and chlorate. In consequence, chlorine dioxide has suffered less than perfect image in the USA over recent years. There are currently some 500 WTPs using chlorine dioxide (contrasted to 40-50 using ozone) but a large percentage of them have admitted to having problems. Certain trials have appeared not to work. Dr Gates and others have investigated these cases and are satisfied that there are straightforward explanations. These relate mainly to

WATER August 1990

the analytical techniques which had been used. These difficulties have now been overcome and over the last six months the technique has received a cleaner bill of health. This point was made strongly at a major international symposium* held in Denver in November 1989, which had a significant input from experienced European operators, including Dr Masschelien.

EXPLOSIVE TENDENCY Chlorine dioxide vapour at concentrations greater than 100'/o will explode under the slightest shock (eg water hammer). There have been some bad experiences at pulp mills using chlorine dioxide for bleaching, but these involved generation systems using out-dated technology. However, there is no chance that such concentrations could build up in water treatment situations. The gas is extremely soluble and the dose rate is very low (of the order of 5mg/ L). Obviously, chlorine dioxide cannot be transported in cylinders like chlorine, and it has to be generated on-site and dissolved immediately into the water stream. Modern generators are explosion-proof. However, there is a danger from the dry sodium chlorite, and correct clean-up measures must be adopted for accidental spills.

DOSING POINTS In order to maximise the effectiveness of the dose, the point of addition should be as far as practicable from: lime dosing; PAC/ GAC processes; air-stripping; permanganate dosing. It is not seen as a complete substitute for chlorine or chloramine disinfection, although some authorities in Europe adopt it as their sole disinfectant, and in fact re-dose it in their distribution systems.

RESIDUALS The USEPA stipulates that the total of chlqrine dioxide, chlorite and chlorate present in drinking water shall not exceed 1.0 mg/ L. The decision to combine the three parameters probably reflects the difficulty in monitoring them separately. Sulfur dioxide or activated carbol)can be used to reduce excess chlorite. Some WTPs use a high chlorine dioxide dose at an early stage of treatment to remove THM precursors, Fe, Mn or tastes and odours, and then inject sulfur dioxide or pass the water through an activated carbon bed to reduce the chlorite to an acceptable level. It is said that taste testing has shown that people can detect the presence of chlorite at concentrations of 0.4 mg/ L, which is lower than the adverse health effects level. Dr Gates advises that these were European taste tests. In USA, the taste threshold was 1.8 mg/ L, since Americans expect their water to have a chlorinous taste. The first complaints re "bleachiness" arose at 1.8 mg/ L. Similar tests with free chlorine produced no complaints until 2.0 mg/ L was reached. The initial impact of chlorine dioxide on a distribution system is to strip organic material from the pipes and produce excessive tastes and odours. Consequently the system requires 3 to 4 days to stabilise, after which the problems disappear. One of the risks with chloramination is that nitrification can occur in certain circumstances in the distribution system. If used in conjunction with chloramination, chlorine dioxide will kill off the nitrifying bacteria.

CONCLUSION There are now a number of companies in Australia offering reliable equipment for dosing of chlorine dioxide. A balance of its advantages and disadvantages indicates that there will be an increasing application for the material.

*The International Chlorine Dioxide Symposium held in Denver, Colorado on 1-2 November 1989, published a Workbook. Copies are available from Dr Robert R. Romano or Ms Lisa Hogan, Chemical Manufacturers Association, 2501 M Street NW, Washington DC 20037, USA. (Fax 202 887 123 7).

TECHNICAL NOIE:

Experience with Nocardia Scum in Activated Sludge By J A SCALZI Supervising Technical Officer E & WS, S.A. SUMMARY This paper discusses an experience with Actinomycete Nocardia (brown viscous) in the Port Adelaide Treatment Works (activated sludge plant) in South Australia. The different control methods applied and their effectiveness are described. THE PLANT The Port Adelaide Sewage Treatment Works is located at Frederick Road, Royal Park, South Australia. The works consist of two activated sludge plants. The first plant (Plant A) was commissioned in 1935 and the second (Plant B) was completed in 1954. The treatment process consists of screenings removal through coarse bar screens, primary sedimentation tanks, aeration tanks (fine bubble type dome diffusers), final separating tanks and anaerobic digestion. The raw sewage is made up of domestic and industrial wastes. The current connected population is approximately 120,000 persons, however, due to its industrial component it has an equivalent domestic population (EP) of 150,000 persons. The average raw sewage flow is 36 ML/day with a peak wet weather flow of 90 ML/day. The composition of the sewage is summarised as follows: Average

Flow (ML) daily pH Units BOD mg/ L Suspended Solids mg/ L Ammonia Nitrogen mg/ L Total Kjeldahl Nitrogen mg/ L Total Phosphorus mg/ L Total Dissolved Solids mg/ L

Range

35.6 25-50 7.4 6.7-7.9 226 140-340 234 120-334 30.6 25.6-41.6 50.2 33.3-82.3 9.2 4.6-14.4 4500 4200-4900

THE PROBLEM Since February 1980, Nocardia scum has been experienced at the Port Adelaide Sewage Treatment Works on a few occasions. It was during the period May to July 1982 when it appeared on the surface of Plant B aeration tanks that an effective control method was devised. Microphotographs of the scum organisms are shown in Figures 1, 2, 3. During that period, Plant A which was receiving the same substrate, was free of any brown scum on the surface of the aeration ·tanks. However, a fine filamentous bluegreen alga, probably Oscillatoria, was present and this caused sludge bulking in the final separating tanks (S.V. = 300). (See Figure 4). Problems caused by nocardial foaming can be: - Extra maintenance due to overflows and windblown foam. - Slight increase in BOD and SS in plant effluent caused by occasional escape of scum from the secondary clarifiers. This was not as a result of poor settling sludge (SVI = 100). Increased operating costs associated with foam control using anti-foaming compounds. Odour generated by the drying scum. Health risks associated with pathogens in airborne foam.

A comparison of Plant A and B operating data at that time is shown below: Plant A

Fine filamentous alga Osci/latoria

Plant B Nocardia

Loadings F:M Ratio 0.26 0.25 Sludge age 5 days 5 days Mixed liquor 2800 mg/ L suspended solids 4000 mg/ L Dissolved oxygen 2.0 mg/ L 3.0 mg/ L Sludge volume index 300 100 Average detention time in aeration tank 4 hours 4 hours Dome diffusers Dome diffusers Type aeration fine bubble type fine bubble system along one side of type across the tank (spiral flow) floor of entire tank 21 mg/ L 9 mg/ L Ammonia - N 3 mg/ L 10 mg/ L Oxidised nitrogen

CONTROL MEASURES TAKEN • Foam Removal. The brown scum was physically removed from the aeration tanks. This was successful in the short term but the scum soon reappeared. • Foam Breakers. These had little effect. • Chlorination. Chlorine dosages of up to 10 mg/L had little effect. • Dissolved Oxygen and Mixing Intensity. The blowers were reduced for Plant B and dissolved oxygen concentration thereby lowered to 1.5 mg/ L. This had little effect;

the brown scum was reduced on the surface due to the reduction in mixing, but microscopic analysis revealed that Nocardia was still present in the mixed liquor. The air was further reduced, thereby reducing the mixing intensity and reducing the dissolved oxygen to 0.5 mg/L. The brown scum disappeared and after a period of7 days the Nocardia were eliminated from the mixed liquor. The conditions, however, were then favourable for filamentous blue-green algae which appeared as the actinomycete disappeared. • It should be noted that both Plant A and Plant B were receiving the same substrate and had the same sludge age, yet were behaving quite differently. The effect of the difference in MLSS is uncertain. CONCLUSIONS On this one occasion it seems that control by reducing mixing intensity and dissolved oxygen concentration was the best form of control of Nocardia scum. This form of control has been applied successfully on the few occasions that the Nocardia has reappeared. ACKNOWLEDGMENTS The author thanks Mr Peter Baker, Scientific Officer, <ind Mrs Christine Riley, Technical Officer Grade 2, State Water Laboratory, for the photography and identification of the specimens.

Fig. I Nocardia Mag. 180x

Fig. 3 Nocardia and Epistylis (Stalked Ciliate) Mag. 720x

Fig. 2 Nocardia and Epistylis (Stalked Ci!iate) Mag, 180x

Fig. 4 Fine Filamentous Blue Green Algae Mag. 180x WATER August 1990

TECHNICAL NOTE

Determining Digester Active Volume -

A Cheap Method

by David Gregory; Chemist, SEPP, Melbourne Board of Works SUMMARY A novel method of determining the active volume of anaerobic digesters is described. The method is based on the measurement of non volatile solids within the digester. Digester contents are diluted by the addition of water at a known flowrate. From the rate of decrease of non volatile solids content the active volume of the digester may be calculated. The method is much quicker and cheaper than the current commonly-used Lithium Chloride tracer method.

INTRODUCTION Most sewage treatment plants which operate anaerobic digesters experience loss of digester capacity due to the accumulation of grit and other inert material over a period of time. The effect of such accumulations is to reduce the proportion of the geometric volume which is well mixed and therefore available for digestion of feed sludges. Taking a digester out of service for cleaning is a lengthy and costly Âˇexercise and the ability to do so may be limited by available digester capacity. A number of techniques is in use for the determination of active digester volume, based on the addition of a tracer compound to the digester. Digester feeding, mixing and withdrawal continue as normal. The decay in concentration of the tracer is determined over a period of time. This together with the liquid flow rate is used to calculate the actual hydraulic detention time which may then be compared with the geometric volume of an empty clean digester. The tracer must fulfil several requirements. It must be unreactive, and unaffected by the digestion processes. Compounds which are naturally present in sewage at fluctuating concentrations are obviously unsuitable. It must be innocuous, not toxic and should not present any environmental problems with subsequent sludge disposal. It must be readily and accurately determined by available analytical methods. It must not adsorb onto solids.

TYPES OF TRACERS The following types of tracers have been proposed or used at various times. Radioactive isotopes Use of a radioactive tracer has been investigated. The use of a short half-life tracer is obviously preferable from the point of view of subsequent sludge reuse, however this complicates the calculations as significant decay takes place within the duration of the experiment. There are extreme hazards in the initial dosing with the large quantity of highly radioactive tracer required. Specialised staff are required to perform the initial dosing and additional pipework and shielding of hot areas are required. Specialised and non routine equipment is required for the subsequent radioactivity determinations on sludge. Sodium fluoride The use of sodium fluoride as a tracer was attempted at the SEPP in 1982. Sodium fluoride was added to give a starting concentration of 11 mg/ ! in the digester. However the highest concentration measured was 4 mg/ !. It is likely that the fluoride concentration was reduced by adsorption on to sludge particles and reaction with other components of the sludge. Since the tracer was shown not to be associated only with the liquid phase, this approach was abandoned. Lithium chloride This is the most widely used tracer. Lithium chloride is readily water soluble and does not adsorb to any great extent onto the sludge. Lithium is readily determined by flame photometry or atomic emission spectroscopy. Background concentrations of lithium are very low, effectively around or below limits of detection, and the compound has low toxicity with no subsequent problems in sludge disposal. The main disadvantages are the availability of lithium chloride, usually several weeks delivery and at a cost of $5500 per digester (1990 costs), based on 300 kg per each 8 Ml digester. Non volatile solids Latterly, digester detention time studies have been performed using the non volatile solids fraction of the digester as the tracer. Normally, total solids and volatile solids contents of the digesters are determined as part of the routine testing. Non volatile solids represent the difference between total and volatile solids. The non 50

WATER August 1990

volatile content of sludge is composed of salts and mineral material which are not affected by digestion processes. Water is added to the digester at a constant flow rate and non volatile solids determined at known time intervals. The advantages of this technique are that it is cheap, rapid and that it may be performed using relatively simple procedures and equipment. The main disadvantage is that the digester must be taken out of the feeding cycle for the test to be performed. However on most plants with multiple digesters, it is possible to take one digester out for a few days and reduce sludge wasting and subsequently catch up after conclusion of the test. Non volatile solids can be determined with great accuracy i.e. 0.76% relative standard deviation, compared with, for example lithium with approx. IOOJo relative standard deviation With good sampling of the digester, a more precise result is available in a shorter time for non volatile so lids compared with lithium tracer determinations. THEORY The digester is a mixed reactor of volume V whose contents of non volatile solids at concentration are diluted by the incoming flow q of water. Mass flow of non volatile solids into the reactor = 0 Mass flow of non volatile solids out of the reactor = qc Rate of change of non volatile solids in the reactor d (Ve) dt Q = _! (Ve) + qc dt V .s!ÂŁ = - qc dt The solution to this equation for the initial conditions c = Co at t = 0 is: In c/ Co = -Q/ V Where Co is the concentration of no/1 volatile solids when dilution commenced and Q is the total volume of water added to time t when the non volatile solids concentration is c. This expression indicates that a plot of In c against Q will have a slope of -1/V.

METHOD The digester was taken out of the feed cycle while mixing continued. Water was added via a hydrant and the flow was measured using a nozzle and pi tot tube. Liquid was withdrawn from the digester at approximately 4-hour intervals so that the liquid level was maintained to within 5% of the initial value. At approximately 8 hour intervals a sludge sample was taken and a flow reading of water fed to the digester was recorded. The exact time was noted . The flow of water varied, presumably because of pressure variations in the supply line. The non volatile solids content of the samples was calculated from Total Volatile and Fixed Residue values determined by APHA method 209E . From the time and flow rate information the cumulative total volume of water added was calculated for each sample. The natural log of non volatile solids concentration In c was plotted against total water volume fed (Q). From the slope of the regression line (- 1 / V) the active digester volume was calculated, as shown in the table and the graph .

DISCUSSION The slope of the attached data indicates an active digester volume of 5.6 ML compared with the 8 ML theoretical capacity of a clean digester. Another digester was subjected to both Lithium Tracer and Dilution Tests. This gave a loss of active volume of 25% by the Lithium Test and 21 OJo by the Dilution Test. These values are consistent with the observed loss in process capacity of around 30%.

It is probable dilution during the test will vary the hydrodynamic properties of the digester contents and this may have the effect of altering the well-mixed volume. There is little evidence of such a

DILUTION TEST OF DIGESTER E2 N~ ~IA~~ kl Wt CJJo (c) In c · (Q) 0.665 -0.409 0 0.654 -0.424 232 0.621 -0.477 347 0.640 -0.447 490 0.548 -0.601 645 0.594 -0.521 785 0.568 -0.566 929 0.549 -0.599 1041 0.546 -0.605 1145 0.538 -0.620 1254 0.527 -0.640 1374 0.511 -0.672 1472 0.500 -0.695 1588 Slope

0.0001 84 Therefore active volume

M. LAGINESTRA Co ntinued from page 32 difference. Castle Hill, which is the best equipped upflow filtration system, achieved a much higher NFR removal than Quakers Hill, (the best downflow system) . Loadings on the filtration plants are very similar. Upflow filters appear better suited to tertiary treatment of sewage. Tertiary filters provide a good buffer between the secondary treatment plant and effluent discharge. However, the filters cannot be expected to remove everything. Like all sewage treatment plant processes, effluent quality will be affected by the influent. Hence, it is important to maintain a good clarifier effluent. The response of filter effluent quality to the clarifier effluent NFR and STP flow should dictate that backwashing be carried out only during low flow periods. Downflow filters entail a complex system . Upflow are simpler - there are no inlet penstocks, effluent rate control valves or backwash rate control valves. Downflow

trend in tests to date but it would appear Qrudent to limit the extent of dilution during tests of this type so thahhis effect is minimised.

FURTHER POSSIBILITIES The dilution technique could be speeded up by using higher flow rates of water and could also be used in studies of the speed and efficiency of mixing. It is likely that plastics accumulation will become more of a problem in digester operation and this technique may be useful to plant operators as a means of determining loss of capacity in this area.

ACKNOWLEDGEMENTS To the Plant operational staff who did most of the work.

5545 kl.

filters have experienced many problems with the hardware and the electrical controls, which led to breakdowns and subsequent loss of filter service. As for most designed systems, the simpler the better, and hence the fewer operational and maintenance problems. A comparison of some upflow and downflow filter characteristics is presented in Table 3, which shows that upflow filters have a considerable advantage over downflow filters; the only disadvantage is cost of construction (estimated from the deeper bed requirement for upflow filters). However this would easily be offset by the capital and maintenance costs of the hardware/control equipment for downflow filters.

ACKNOWLEDGEMENTS I thank many people within the Water Board who contributed in various ways, including personnel at Castle Hill, Qua.kers Hill, West Hornsby, Glenbrook, Mt Riverview, St Marys, Hornsby Heights and West Camden STPs and John Macindoe,

Elizabeth Seidle, Greg Mumford, Dick Browning, Graeme Dyer and Denise Prescott.

REFERENCES BROWNING, R. C. (1987) - New Method and Revised Nomenclature for Granular Filter Flow Control. A.W.W.A . 12th Federal Convention, Adelaide. 202-209. ELLIS, K. V. (1980) - The Tertiary Treatment of Sewages. Effluent and Water Treatment Journal, September, 422-430. FITZPATICK, J. A. and SWANSON, C. L. (1980) - Evaluation of Full-Scale Tertiary Wastewater Filters. United States Environmental Protection Agency Report, May (Cincinnatti, Ohio). Joint Committee of the Water Pollution Control Federation and the ,American Society of Civil Engi neers (1977) - Wastewater Treatment Plant Design - a Manual of Practice. (Lancaster Press Inc, USA). 10MLINSON, E. J. and JAGO, P.H. (1980) A compatison of Upflow and Downflow Full-Scale Sand Filters for Tertiary Treatment of Sewage Effluent. Trib. Cebedeau, 444, 463-467.

CONFERENCES

HYDRANAUTICS

WATER AND ENERGY CONSERVATION IN COMMERCIAL BUILDINGS

WORLD'S LARGEST ZERO DISCHARGE WASTEWATER TREATMENT PLANT

Brisbane, Australia 17, 18, 19 & 20 October 1990

Commercial buildings, especially in the central business district, have the potential for enormous waste of water and energy. The conference will address these issues in three mai n streams: • Water conservation in commercial buildings. • Energy efficiency in commercial bu ildings. • Energy effic ient design and construction. The conference will provide for airing of the issues by variou s experts in their fie lds, discuss ion, debate and the subsequent deve lopment of a strategy for promotin g the use of water and e nergy efficient app lia nces and methods of bui lding construct ion in Aus tralia.

The conference is expected to attract more than 400 international and local delegates, including building owners, architects and e ng ineers, and will attempt to draft new Australi an standards and bui ldin g by-laws re lat ing to water and e nergy conservat ion . It will , for example, address the effects of recycl ing of grey water for toilet flushing, and of dual flush systems. Information and registration brochures are availab le from the Conference Secretariat, lntermedia Conven ti ons a nd Eve nts Management. Enquiries: Marian Hudson (07) 369 0477, Fax (07) 369 1512

A new six-page illustrated application report from Hydranautics describes the wor ld 's largest ze ro discharge wastewater t reatme nt p lant located at the 4640 MW Bayswater/Lidde l I power stat ion comp lex in New South Wales. Process details, RO pretreatment , RO plant des ign, automatic membrane cleaning, and monitoring and control system are in c luded in the desc ri ptio n . F low diagrams, performance curves and equ ipment photographs g ive the reader an insight into the operating features of the plant. Informat ion : Jackie Kirk at Hydranautics, 8444 Mi ra lan i Dr, San Diego, CA 92126 . Fax : (619) 536 578

WATER Augus1 1990

Groundwater Contamination Assessment of works . Sites by R.J. Parker and R.D.A. Wolfe.

SALE AWWA BOOKSHOP All prices include postage

In our June issue, the above paper was printed without our usual author biodata. We apologise to the authors and our readers.

While stocks last

P.O. Box 460, Chatswood NSW 2057 Facsimile (02) 410 9652 Telephone (02) 413 1288 Water Pollution Control Federation Publications Joint Treatment of Industrial & Municipal Wastewater (1976) Regulation of Sewer Use (1975) Uniform System of Accounts for Wastewater Utilities (1970) Design & Construction of Sanitary & Storm Sewers (1976) Water Treatment Plant Design 1st Edition (1971) Basic Course for Wastewater Treatment Plant Operators includes - administrators handbook, student workbook, 280 slides, 10 audio cassettes, carry case American Water Works Association Publications Basic Management Princtples for Small Water Systems Glossary: Water and Wastewater Control Engineering Before the Well Runs Dry - Vol I Before the Well Runs Dry - Vol II Microcomputers for Water Utilities Water Rates (Ml) 1983) Water Meters - Selection, Installation, Testing & Maintenance (M6) (1986) Installation, Field Testing & Maintenance of Fire Hydrants (Ml 7) (1980) Precoat Filtration (M30) (1988) Guidelines for the Selection of Laboratory Instruments (MIS) (1979) Groundwater (M21) (1973) Ozonation: Recent Advances & Research Needs (1986) Standard Set (in binder - 6"x9"), AlO0-EIO0's • Standard Set (in binder - 8.5"xll") AI00-C300's • Standard Set (in binder - 8.5"xll") C400-ElO0's • Guidance Manual - Procedures Manual for Selection of Coagulant, Filtration & Sludge Conditioning Aids in Water Treatment Cross-Connection Control Training Package, includes 100 slides with script, student handout, teachers planning guide Water Supply Operation Series Instructor's Guide and Solution Manuals: Volume II Volume IV Basics of Well Construction Instructor's Guide Package, 25 colour slide presentation and sample student text • Does not contain all current standards

Richard Wolfe is a civil engineer with a post graduate degree in environmental engineering. He is an Associate of Camp Scott Furphy and is currently project manager for the site characterisation and clean up of a farmer gasworks site and rail and dock areas in Melbourne.

$5 $5 $5 $10 $7 $150 $20 $15 $15 $20 $25 $20 $20 $10 $15 $5 $5 $10 $250 $200 $200 $15 $50 $50 $30

Roger Parker graduated B.E. and MEng.Sc. from Monash and has worked for Golder Associates Pty Ltd since 1976. He is now responsible for their Australian environmental engineering activities, which involve the assessment of soil and groundwater contamination and the hydrogeological assessment of new and existing landfills.

FILAMENTOUS BACTERIA A SHORT COURSE IN THEIR IDENTIFICATION While Professor David Jenkins was in Bendigo for the BNR-1 conference in July, he discussed with staff from the BCAE the possibility of running short courses in the identification of Filamentous Bacteria in Activated Sludge Plants, possibly during July/August 1991, with Beth Seviour, Ken Lindrea, Jacques Soddell and Bob Seviour from BCAE together with Linda Blackall from Brisbane. If enough interest was shown, these courses could be run in a number of capital cities over a period of a couple of weeks to minimise travel costs for attendees. ' People who would be interested in attending such a course are asked to contact Ken Lindrea at Bendigo CAE, PO Box 199 Bendigo Victoria 3550 (telepl}one 054 443 702) by 31st October, to see if the idea is worth pursuing. Their names will be put on a mailing list to keep them informed.

$50

BNR-1 First Australian Conference Biological Nutrient Removal from Wastewater PROCEEDINGS 330 Pages $50

Hon Sec. Victorian Branch AWWA PO Box 409 Werribee 3030 (03) 741 5844

WATER August 1990

Dr. R. Seviour Bendigo CAE PO Box 199 Bendigo 3550 (054) 44 7459

Bendigo College of Advanced Education

FILAMENTOUS BACTERIA IN WASTEWATER TREATMENT PLANTS AN IDENTIFICATION SERVICE We can identify the filaments present in samples of mixed liquor and foam sent to us, and we will supply you with results within 48 hours of us receiving your samples. We will also recommend, on the bases of our identifications, possible courses of action you might try to eliminate your problem. Our service will then monitor these remedial treatments to allow a better evaluation of their effects on the whole biomass. With time, we should build up an extensive data base on plant performance and occurrence of filament types, which will be available as a reference service to everyone in the wastewater industry in Australia. Contact: K. C. Lindrea or R. J. Seviour (054) 44 7459.

THE UNREASONABLE MAN by Richard Raxworthy A NSW Publlic Works Department history project. Hale & Ironmonger, $29.95 (cloth) 153 pages, 1989. The remarkable "unreasonable" amount of work carried out by J.J.C. Bradfield permeates this book. Though the Sydney Harbour Bridge is accepted as his most remarkable achievment many other aspects of his background and life are covered. The family history, his early childhood from his birth in 1867 in Ipswich, Queensland, then onto study at Sydney University are described in some detail. His entry into the NSW Public Works Department as a drafts).Tlan in 1891 shows him engaged on a variety of projects. The covered roads, bridgeds and the city railway. Though he was not personally involved in the bubonic plague in Sydney, it had a marked effect on his career. From the earliest years the project of a bridge to connect Sydney with North Sydney occupied not only his professional duties but also affected his private life as he lived in North Sydney after his marriage in 1891. His extensive involvement with the development of this important project to its completion are vividly portrayed. The obstacles were numerous - the severe economic depression of the 1890s with all its political and financial implications, unemployment and reduced public works. Bradfield's extensive involvement with transport, especially the railways of NSW is developed in some detail. Similarly his work in tramways, city railway, town planning, Cataract Dam and Burrinjuck Dam are shown to have occupied his extensive engineering expertise for many years. When Bradfield was in 1912 appointed Chief Engineer for Metropolitan Railway Construction and Sydney Harbour Bridge no concepts for either of these major projects had been laid down . Constant demands during world war 1914-1918 were made from the man whom the then Premier of NSW, W.A. Holman described as " ... gifted with the closest approach to complete genius in engineering matters". Staff was short and a considerable amount of outside war work was carried out under his supervision. Some space is devoted to the extensive wrangle in the attempts to initiate calling tenders for a bridge design in USA and in Britain . Campaigns for much more costly and inefficient tunnels were constantly obstructing Bradfield's bridge proposals. Finally we are introduced to the demanding attempts to have a satisfactory design selected from the many tenders put forward . Also analysed are the tests of special steel and riveting equipment to be used. Other concerns of Bradfield at the time are described. They included city railway progress, the structure and bells for the War Memorial Carillon at the University of Sydney, the bridge approaches from the City and from North Sydney, the ANZAC Memorial in Martin Place, Sydney and many others. A section is devoted to the serious 54

WATER August 1990

debates which raged about the claim that Mr Ralph Freeman, the chief consulting engineer of the contractor Dorman, Long and Co, not Bradfield was the sole designer of the bridge. Also carefully presented is Bradfield's enlightened attitude to the workmen which he showed in his court evidence supporting adequate remuneration for the workers on the bridge. An exciting chapter deals with details of the material assembly and the growth of the extending arches from each end, finally closing with great precision on 9 September 1930 to the satisfaction of all the engineers and gangs of men involved. Further detail is devoted to the final opening on 19 March 1932. The final pages are devoted to Bradfield's activities after his retirement from the PWD

and establishment with his son Bill of the firm of consulting t ngineers. The innumerable excellent illustrations and photographs of personalities and structures at various stages, engineering drawings and sketches and cartoons obtained from numerous sources make this a splendid record. The additional presentation of the family tree of the Bradfield family, carefull references, glossary of technical terms, bibliography, index and overall layout make this an extremely valuable index and overall layout make this an extremely valuable book. It not only honestly portrays this outstanding "unreasonable" man and his time but it gives an excellent indication of the immense contribution made during his lifetime by J.J.C. Bradfield. H. Bandier

SEWAGE SLUDGE TREATMENT AND USE

Research Institute described investigations carried out between 1959-1980 on organic substances in soils and plants after intensive application of sewage sludge. It was found that chlorinated hydrocarbons did not accumulate in the soils and did not transfer to crop plants. Polychlorinated biphenyls were found to have increased up to 17-fold and polycyclic compounds up to IO-fold in the soils. However, no regular transfer to crop plants was detectable where the order of magnitudie was in micrograms/ kg. The study concluded that these organic substances do not need to be limiting factors for the use of sewage sludge in agriculture.

EDITED BY A.H. DIRKZWAGER AND P. L'HERMITE - (ISBN 1-85166 418-1) Elsevier Science Publishers, U.K. ÂŁ65. Proceedings of the Amsterdam COST conference, September 1988, organised jointly by the Commission of European Communities, the European Water Pollution Control Association and the Netherlands Association of Wastewater Treatment and Water Quality Control. The proceedings cover: sludge characterisation; sludge management options for treatment and disposal; agricultural use of sewage sludge: quality aspects; sludge dewatering and biological conditioning; sludge incineration; other thermal processes; landfilling: poster sessions. In 1994, compared with 1984: Denmark and Italy expect greater utilisation of incineration and less landfilling; Greece expects greater use of sludge recycling and less landfilling; The Federal Republic of Germany predicts a substantial increase in incineration with less sludge recycling; The UK expects greater use of recycling and incineration with decreases in ocean disposal and landfilling. There are increasing constraints on the EEC member countries for the use of sewage sludge in agriculture particularly relating to contaminants. Long wet winters also necessitate alternative disposal/utilisation routes during winter. A useful paper by C.D. Bayes et al details four options for utilising sewage on land as a resource other than for agriculture: forest fertiliser; soil conditioner for restoration of disturbed soils; soil forming material for reclaiming derelict land; producing soil for the urban environment. W. Kampe of the Speyer Agricultural

The sessions on sludge dewatering described three relatively new developments in dewatering technology: in order to achieve a minimum of 35% dry solids. CHP high pressure filter press, Centripress, and the Hi Compact Press The session on sludge incineration reported that incineration utilising fluidised bed (F.B.) furnaces was appropriate for densely populated and industrialised areas. Auto-thermal operation of a F.B. furnace can be achieved with efficient dewatering and / or pre-drying. Emissions of dust , sulphur dioxide, nitrogenous oxides and volatile heavy metals can be controlled successfully with modern flue gas treatment technology. The risk for dioxin emissions has not yet been fully evaluated but seems never-the-less considerable smaller than for the incineration of municipal waste. Other thermal processes covered in the text are: Oil From Sludge lechnology and Thermophilic Aerobic Stabilisation The publication is a very useful reference for keeping abreast of recent developments in Europe m the field of sludge management . IAN PITTAWAY Camp Scott Furphy Pty Ltd

SEMINAR REPORT

Commercialisation of Research A Report by Bill Rees Considerable interest was aroused by the April 1990 issue of Water devoted to research and development for urban water in Australia. As a follow-up to this, the NSW Branch held a seminar on 1 August at which nine speakers discussed various aspects of the commercialisation of research. Executive Director, Peter Hughes, opened the proceedings with a call for us to change the "Lucky Country" into the "Clever Country" by the appropriate development of good ideas in a commercial manner. The first speaker was Jeff Belton, Director of BHP's Central Research Laboratories, who spoke of current trends. In recent years his company has been increasing its research expenditure by $!OM/ annum and in 1990/ 91 will spend about $90M, representing 0.62% of annual turnover. On any research project the pattern of expenditure is much the same; comparatively small amounts in the research and development phase followed by massive capital input as a new product is developed, leading hopefully to a profit in due course. This process could take between five and ten years to achieve. At present the Australian financial situation is not conducive to the successful development of new products. The prime lending rate is much higher than for our competitors - Australia 16.5% compared with Great Britain, German and the United States at 10% and Japan at 5% - and this has a big effect on our cumulative cash flow. We have to work twice as hard to match Japan's cash flow but there are some ways in which we can improve the situation. It is possible to absorb early costs in the company's other related activities and certainly there is a much greater chance of success if it is possible to exploit a unique resource. Potential customers should be involved in the development of a research project but it is important not to press on to a new development when the improvement in technology is only marginal and it may be difficult to persuade customers that the proposed change is worthwhile. Boyd Keogh of Boyd Keogh Industries Pty Ltd, spoke of the work that has to be done in preparing a patent. Firstly it is important to establish priority for a new discovery by filing a provisional patent application prior to publication. Then it is necessary to protect the discovery by foreseeing all possible developments for a product. St_art with a wide ambit claim, filing with all relevant countries, with the purpose of achieving maximum protection with greatest flexibility. It is necessary to prosecute the patent application by meeting all target dates as advised by the patent attorney and to persevere in this work by fighting until you succeed. Finally it is important to develop a portfolio of projects by the use of lateral thinking to establish all possible uses for a discovery and so, by the use of a growing portfolio of good patents and technical know-how, to develop an industry. The valuation of intellectual property was the subject addressed by Ian Plater of Arthur Andersen and Co. There are a number of key issues that have to be considered in valuing intellectual property Âˇ- how exclusive is it? for how long will it remain so? and how much will someone pay to make use of it? Such questions have led to the development of valuation methods which need to consider the value of comparable market transactions, the replacement cost, the economic benefits to be obtained and so on. However many methods have distinct disadvantages. Often there is a lack of markets, the unique nature of some intellectual property makes it difficult to value, the profits attributed to it may be difficult to dissect and the replacement cost does not value the future economic benefit. Nevertheless the whole valuation exercise is considered to be worthwhile but it requires good judgement and skill in the industry. Tony Hooley, the Managing Director of lnsearch Ltd, the commercial arm of the University of Technology, Sydney, spoke about the protection of intellectual property using patents but, at the same time, opened up the discussion to cover alternative forms of prote~tion. These might include the registration of designs, use of copynght etc. and can be cheaper and involve much less work. There are many ways to get around patent rights and it may be necessary to spend a great deal to protect such rights particularly against large companies with wide resources. He suggested that the

first step in defending a patent where infringement has occurred is to send a solicitor's letter demanding that the infringement cease and claiming damages. This is preferable to going to court which can be very time-consuming and expensive. In cases where an infringement appears to have occurred, the defence might be that the patent is invalid because of prior publication or because it is not sufficiently explained or not achievable and the holders of the patent must ensure that such claims cannot be sustained. Another ~pect of project development, tl)e licensing of research, was dealt with by Roger Kearin of James Hardie Industries a firm which has licensed the technology for the manufacture ~f GRP pipes. Many factors have to be taken into account in such an arrangement for the licensee can make or break a product and cannot always deliver what is promised. On the other hand the licenser needs to be familiar with the end product that the lic:nsee is going to market and be in a position to provide follow-on technology so that the aims of the project are achieved. In many ~ases training must be provided in manufacturing techniques and, m the case of GRP pipes, pipe laying techniques are an important consideration. Many different aspects have to be considered in drawing up the licence arrangements - financial, protection, long term performance, training, development, future knowledge ' options for use, legal etc. David Hogg, Institute of Manufacturing Management and Technology, Unisearch, dealt with ways intellectual property could be disseminated - by licensing, joint venture or by sale or assignment. He emphasised a point that recurred on many occasions throughout the Seminar - all methods employed in the passing on of information or technology involve the expenditure of a lot of hard work on both sides, with a lot of iterative effort, if the transaction is to be successful in terms of the development of a successful product, monetary gain and so on. If it is decided to sell intellectual property, an important consideration is the type of sale; should it be a one-off sale or should it be an upfront payment plus regular royalty payments? The later part of the Seminar wa~ devoted to a number of case studies aimed at illustrating the more general points made by previous speakers. John Johnstone of Sirotech Ltd. described how CSIRO was commercialising the results of its research and development activities. This was originally done through its Corporate Commercial Group and more lately through Sirotech Ltd., a technology transfer company established in October 1984. A flexible attitude is taken towards the commercialisation of research with th_e ~ick vaccine project a collaborative venture, the zirconia project a J_o~nt venture, while the gene shears project is being managed by a JOmt venture company. Norman Longworth, General Manager of ASEA (Brown Boveri) discussed his company's Coarse Coal Dense Phase Hydro Transport project which had already involved the expenditure of $25M in ten ~ears. _He emphasised the large scale of a project of this type, mvolvmg many steps in the investigation process, and vigorously presented the point that Australia will get nowhere without successful research and development, requiring major investment often without a guaranteed outcome. The scene on the government side was presented by John Schlafrig of the NSW Electricity Commission who reported on the changes of attitude to research work in recent years. Previously research and development has had low priority, a small budget, the need for direct application and often a loss of intellectual property. There is a need in _t~e publi~ sector _to stretch asset life, to improve operating efficiency while reducmg resources. There is now a greater emphasis on_ joint ventures and budgeting over a longer time scale. He spoke bnefly on current projects including microwave smelting, a bulk supply study, electric vehicles and plasma ignition. This Seminar represented a new venture for the NSW Branch. Although matters relating to water supply and sewerage were barely mentioned, the opportunity was given for members to learn about the commercialisation of research from experts in other fields and this resulted in a very productive afternoon. WATER August 1990