Water Journal March 1988

Page 1

EXECUTIVE DIRECTOR P. Hu ghes Box A232 Sydney Sout h 2000 (02) 269 6814

FEDERAL PRESIDENT M. Du rea u, Ke nt Instrum ents P/L P.O. Box 333, Caringbah 2229 (02) 525 2811 .


ISSN 0310-0367

Official Journal


Vol. 15, No. 1, Marc}l 1988

FEDERAL SECRETARY G. Cawston Box A232 P.O. Sydney Sth ., 2000. (02) 269 6157

FEDERAL TREASURER J . D. Mol loy , C/· M.M.B.W. 625 Lt . Colli ns St., Melbourne, 3000. (03) 6 15 5991

CONTENTS Viewpoint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .



Association News, Views and Comments ........ . ...... . . . . , . . . . .


M. Sharpin, • Wi lli ng & Part ., P.O. Box 170, Curt in, A.C. T. 2605. (062) 815 81 1

IA WPRC News . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .


New So uth Wales

Index-Water Volume 14, 1987 . ... .. . . .......... . . .. .. ..... . . .. .


Development of Water Quality Models for the Control of Eutrophication in Lake Burley Griffin -L. A. Nagy and D. C. Butters ... ... .. ... . . .. . . . ... . ....... .


National Hazardous Waste Management Conference Report-Errol Samuel .. .. . ... . .. . . . .... . .. ........... ... . .


Joint Taks Force on Intractable Wastes Report-Errol Samuel . .. . . . .. . , ..... . ... .. ...... . .. ..... . . .


Letters to Editor . . .. .... .. . . .. . .. . . . . . .. . .. ..... ............. ,


Calendar . . . .... . ..... . .. ....... . . .. .. . . .. . . . ... t . . ......... .


Phosphorus Movement Through Sands Modified by Red Mud -N. Kayaalp, G. Ho, K. Mathew and P. Newman


Reed Bed Treatment of Wastewaters: A European Perspective -T. H. Davies . . . ... ... . .. .. .. . ..... .. .... ... .. . .. . ..... . .


The ACT Approach to Improving Urban Stormwater Quality -B. C. Phillips, A. I. Lawrence and A. G. Goyen . .... . .... . .... .


Mrs S. Tonkin -Hi ll , Sinclair Knight & Part. 1 Chandus St., St . Leo nard s, 2065, (02) 436 7166

Vi cto ri a J . Park , Wa ter Training Centre, P.O. Box 409, Werribee, 3030. (03) 74 1 584 4

Qu ee nsland D. Mackay, P.O. Box 412, West End 4101. (07) 844 3766)

South Au stral ia A. Townsend , State Wate r Laboratories, E. & W.S. Private Mai l Bag , Sal is bu ry , 5108 . (08) 259 0316

Western Au stralia Mr K. Cadee, Water Aut h. of W.A. , P.O. Box 100, Leedervil le 6007 (09) 420 2457

Tasmania G. No lan, G.P.O. Box 78A, Hobart 7001 (002) 44 0600

North ern Territory P. Abbey , P.O . Box 37283 Win ne llie,

N.T. 5789. (089) 89 7290

EDITORIAL & SUBSCRIPTION CORRESPONDENC E G. A. Golli n, 7 Mossman Dr., Eag lemo nt 3084 03 459 4346

ADVERTIS ING M iss Ann Sykes

App ita 191 Royal Parade, Parkv i lle 3052 03 347 2377

The Biocarbone Process: A New Development in Wastewater Processing

-K. G. Barr . . ... . . . ......... . ... . . ... ... , . .. . ... .. ... . . . .


Conferences • Courses • Symposia . ........ .. ........... , ..... .


Plant • Products • Equipment . ... .......... . . . . .. ... . . . ........ .


COVER PICTURE The Water Pollution Control Pond in the foreground is representative of pollution control structures being utilised in the ACT to intercept urban stormwater pollutants prior to discharge to the Murrumbidgee River (in background). The Lower Stranger WPCP, a 90 ML impoundment, was designed by Willing and Partners Ltd and constructed by Roche Bros Ltd for the National Capita l Development Commission. It has been designed to intercept 98.5% of bacteria and 60 % of suspended solids and phosphorus . Photo, courtesy of Roche Bros Ltd and spon sored by Willing and Partners Ltd.

Th e sta temen ts made or opin io ns expressed in ' Wa ter' do no t necessa rily reflec t the views of the Aus tralian Water and Was tewa ter Assoc ia tio n, its Cou ncil or comm itt ees.


Development of Water Quality Models for the Control of Eutrophication in Lake Burley Griffin L. A. Nagy and D. C. Butters SUMMARY Lake Burley Griffin, located in the Australian Capital Territory, has exhibited signs of eutrophication, as exemplified by high phosphorus concentrations and summer blooms of bluegreen algae. Based on the results of a long term monitoring program, two water quality models were developed to help predict the response of the lake to eutrophication control measures. The models indicated that the planned reductions in phosphorus at the major point -source Queanbeyan Sewage Treatment Works .(STW), would result in significantly lower levels of blue-green algae in the lake. The models also indicated that apart from phosphorus, algal numbers in the lake were also influenced by water temperature and water flow volume. The importance of these findings to Australian conditions is described, and the applicability of the overall approach to other water quality management problems is indicated.

INTRODUCTION Lake Burley Griffin, since its establishment in 1964, has provided Canberra with a recreational and architectural focal point. However, due to a variety of factors, the recreational amenities of the lake have not always been realised. One of the factors responsible for this has been the eutrophic nature of the lake, exemplified by high phosphorus levels and frequent summer blooms of blue-green algae. Phosphorus is one of the critical nutrients used by algal cells, and consequently algal biomass in rivers and lakes is strongly influenced by the concentrations of phosphorus present (Ganf et al., 1982; Gloterman, 1975; Kalff and Knoechel, 1978; Reynolds and Walsby, 1975; Schindler, 1974 and 1975). This influence of phosphorus on algal biomass , and therefore on eutr@phication, has also been noted for Australian rivers and lakes (Bayly and Williams, 1973; Cullen and Smalls, 1981 ; Ganf, 1980; Hart et al., 1983; Williams, 1980; Wood, 1975) . as well as for water bodies in the Canberra region (Cullen et al., 1978; Department of Construction/Binnie International/Maunsell and Partners, 1978; Department of the Capital Territory, 1981; Humphries and Imberger, 1982; National Capital Development Commission, 1981a and 1981b; Philp, 1982 and 1985 ; Rosich and Cullen, 1981). Phosphorus enters the waters of Lake Burley Griffin from a variety of point and non-point sources (Cullen et al., 1978). Point source inputs of phosphorus originate from the Queanbeyan STW, which is located on the Molonglo River, upstream of the lake . Non-point source inputs include lake sediments, macrophytes, as well as agricultural and urban runoff. In 1981 a long-term monitoring program was undertaken to determine the relative importance of these different sources of phosphorus (Department of Housing and Construction, 1982 and 1984) ; The Office of ACT Administration, 1987). The results of this program indicated that the main phophorus source was the Queanbeyan STW. Consequently, steps have been taken to reduce phosphorus inputs from this source and thus reduce eutrophication problems in the lake . The aim of this study was to develop water quality models to identify the impact of phosphorus on algal growths and provide tools for management decisions concerning the control of algal blooms in Lake Burley Griffin.

MATERIALS AND METHODS Sites and Sample Collection

Sampling sites selected included six locations in the Molonglo River upstream of Lake Burley Griffin and two locations within the upstream part of the lake (Figure 1) . The locations in the Molonglo River included one site upstream, one S'ite at, and four 14

WATER March, 1988

Laslo N. Nagy, B.Sc. (Hons) and Ph .D., is a Director of Aquatech Pty • Ltd. He has published several papers in the fields of microbiology, statistics and water management. For the past three years he has worked on various aspects of ACT water quality, including the rehabilitation of Lake Burley Griffin .

David Butters, B.Sc. is the Manager Scientific Services, Water Branch with the Office of ACT Administration. He Joined the Department in 1968 and since has been involved in a wide variety of water quality issues associated with the ACT Water Supply, Sewage Treatment and the environment.




sites downstream of the Queanbeyan STW. The two upstream basins of the lake were selected because of their proximity to the , point source phosphorus input. At the river sites surface samples were collected, while in the lake , surface and tube (3 m depth at location 1 and 5 m depth at location 2) samples were obtained. The above sites were sampled ea~h month between November 1983 and June 1986. S [A LE





Figure 1. Location of sampling sites used in the study. Physical-Chemical, Nutrient and Algal Parameters

A variety of physical-chemical, nutrient and algal parameters were determined during the monitoring program, and these were used in the , construction of the water quality models. The physical-chemical parameters included·temperature, dissolved oxygen, specific conductance, euphotic depth, turbidity, pH and alkalinity. The nutrient parameters were phosphorus (total and filterable ortho) and nitrogen (Total oxidised, ammonia, total kjeldahl, total inorganic and total). Algal parameters were total algae, blue-green algae and chlorophyll-a. All parameters were determined according to documented procedures used by the Water Quality and Investigations Laboratory, The Office of ACT Administration.

Hydrological and Climatic Parameters

A number of hydrological and climatic parameters were calculated from archival data collected from the Hydrology and Water Resources Unit, The Office of ACT Administration, or from the Canberra City Meteorological Station. These were: (i) rainfall (average rainfall per day received in 14 day period prior to each sample time), (ii) wind (average of maximum daily wind speeds recorded in 7 day period prior to each sample time), (iii) river flow (average river flow per day in the Molonglo River at site 1, in 14 day period prior to each sample time), and (iv) effluent flow from the Queanbeyan STW (in the absence of a flow meter on the effluent, average flow into the plant was used as a surrogate: this flow was estimated to be a relatively constant 7 .0 ML. day- 1 or 0.08 m3 .sec- 1 , G. Pike, Queanbeyan City Council Engineer, personal communication , 1986). The rainfall, wind and flow determinations were added to the data to evaluate their influence on the other parameters. Flow was also added to help estimate nutrient loadings from the Molonglo River as well as from the Queanbeyan STW. Statistical Analyses

The available information was used to create a data set, which contained the nutrient parameters from the Molonglo River sites, and the corresponding physical-chemical, nutrient and algal parameters for Lake Burley Griffin. The lake values used for each parameter were an arithmetic mean of values at locations 1 and 2, surface and tube results. By utilising the average of four determinations, a more representative value was obtained for each parameter. The algal information was lagged 1, 2, 3 and 4 months in order to ascertain if nutrient levels in the Molonglo River (originating from the Queanbeyan STW) had any impact on algal levels in the lake 1, 2, 3 or 4 months later. Results were analysed using correlation (Pearson' s r), and regression techniques. Construction of Water Quality Models

The above monitoring program and statistical analyses provided an overall approach for quantitatively describing and predicting water quality parameters in Lake Burley Griffin. Blue-green algae represented an important water quality parameter, as they were the main eutrophication problem faced by lake managers. Consequently two water quality models (regression eq uations) were constructed to explain the response of Lake Burley Griffin blue-green algae to variations in environmental conditions, and changes in phosphorus inputs from the Queanbeyan STW. The regression models used parameters that provided the maximum level of statistical explanation, but at the same time were based on a plausible causal explanation in terms of the accepted theories and hypotheses within the aquatic sciences. Although blue-green algae were used in this study, the same overall approach could be used for any other parameter of importance.


TABLE 1: VALUES OF VARIOUS MEASURED PARAMETERS DURING MONITORING PROGRAM (FROM NOVEMBER 1983 TO JUNE 1986, n = 29) Molonglo River Location I (upstream of the Queanbeyan STW) Total Phosphorus (mg P.V ') Filterable Ortho Phosphate (mg P .V') Tota l Oxidised Nitrogen (mg N.L-') Ammonia (mg N.V') Total Kjeldahl Nitrogen (mg N.V') Total Inorganic Nitrogen (mg N.V') Total Nitrogen (mg N.L-') Molonglo River Location 2 (Queanbeyan STW Effluent) Total Phosphorus (mg P.L-') Filterable Ortho Phosphate (mg P.L-') Total Oxidised Nitrogen (mg N.V') Ammonia (mg N.V') Total Kjeldahl Nitrogen (mg N.L-') Tota l Inorganic Nitrogen (mg N.L-') Total Nitrogen (mg N.L-') Molonglo River Location 3 (downstream of the Queanbeyan STW) Total Phosphorus (mg P.L-') Filterable Ort ho Phosphate (mg P .V') Total Oxidised Nitrogen (mg N.V') Ammonia (mg N .L-') Total Kjeldahl Nitrogen (mg N.V') Total Inorganic Nitrogen (mg N .V") Total Nitrogen (mg N.V') Lake Burley Griffin (Locations 1 + 2, surface + tube) Temperature ( 0 C) Dissolved Oxygen (mg O, .L-') Specific Conductance (uS) Euphotic Depth (m) Turbidity (NTU) pH

Total Phosphorus (mg P .L-') Filterable Ortho Phosphate (mg P .L-') Total Oxidised Nitrogen (mg N.L-') Ammonia (mg N.L-') Total Kjeldahl Nitrogen (mg N.L-') Total Inorganic Nitrogen (mg N.V ') Total Nitrogen (mg N.V') Total Algae (cells.mL- ') Blue-Green Algae (cells.mL-') Chlorophyll-a (ug.L-') Log Total Algae (log cells.mL-') Log Blue-Green Algae (log cells.mL-')

Hydrological and Climatic Rainfall (mm.day-•) Wind (Km.hour-') River Flow (m'.sec-') Log River Flow (log m' .sec· ') Effluent Flow Queanbeyan STW (m'.sec-')

Mean 0.038 0.011 0.088 0.036 0.499 0, 123 0.587

5.73 5. 16 2.12 6.93 9.81 9.05 11.93

S.D. 0.021 0.010 0.064 0.025 0. 14 1 0.072 0.150

1.90 1.78 1.29 4.11 4.63 4.63 5.08

0.698 0.642 0.325 0.917 1.522 1.221 1,847

0.920 0.888 0.1 78 1.059 1.311 1.174 1.419

16.6 8.2 199 •1.1 28 7.86 0.111 0.056 0.204 0.061 0.755 0.265 0.959 7300 6000 14 3.23 1.61

4.9 1.5 49 0.3 10 0.23 0.030 0.020 0.094 0.029 0.127 0.102 0. 134 17400 17500 12 0.68 1.66

2.16 31.0 6.4 0.31 0.08

1.94 6.8 10.8 0.75 NA

Physical-Chemical, Nutrient and Algal Parameters

Hydrological and Climatic Parameters

Physical-chemical parameters were relatively similar at the Molonglo River sites, however , the nutrient and the algal parameters showed considerable variation. The concentrations of phosphorus and nitrogen were relatively low in the Molonglo River upstream of the Queanbeyan STW, however, downstream of this site concentrations were generally one order of magnitude higher, as indicated for the main Molonglo River sites in Table 1. The concentrations of the various nutrients in the Queanbeyan effluent over the study period are indicated in Table 1. Most nutrients showed considerable variation with total phosphorus ranging from 1.23 to 8.84 mg P.L- 1 and total nitrogen ranging from 3.32 to 22.71 mg N.L- 1 • The levels of the various nutrient and algal parameters in Lake Burley Griffin are also shown in Table 1. These parameters were relatively constant, except for the algal measures, which once again showed considerable variation. Because of this variation, the algal parameters were expressed in a log form.

The hydrological and climatic parameters also showed considerable variation . This was particularly evident in the case of average river flow, which like the algal parameters, was also•expressed in a log form. Statistical Analyses

Algae in Lake Burley Griffin correlated with phosphorus concentrations downstream of the Queanbeyan STW. This correlation improved when algal numbers were lagged 1, 2 or 3 months to allow some time for biological growth as well as for physical transport of materials from the river into the lake . Figure 2 shows the scattergram for total algae in Lake Burley Griffin and total phosphorus concentrations at the first site downstream of the Queanbeyan STW. The relationship, with an r value of 0.52, was positive, indicating that as phosphorus increased, algal levels in the lake also increased. The scattergrams for the WATER March, 1988





:-cON( fp--;~;- 7


t QU[ANBEYAN STW JI L ___ ____


, '' ,,, ''










"' "'> UJ


1- - - - - - - : ,



:l wAHR TfHPI:


L-- y -,J


' ',


-1.2 lagged 0 month


..J (.!)

lagged 1 month






0 ..J 0 :Ii

. ,_. ,_ ,_













~------~ l

1 - - LAKE BUR LEY I [ : GRIFF IN 1 SIHS1 AN02 I c: ______ j


0.6- C










-12+-----,---~--~ 2 3 4 5


lagged 2 months


4 tagged 3 months



r- ---,-- - - :-,






1 SIHI L- - -



- - - - -


1 LOG TOTAL ALGAE ( log cells .mL )

Figure A-D. Scattergrams for log total algae in Lake Burley Griffin (Lagged 0, 1, 2 and 3 months), and log total phosphorus in the Molonglo River at site 3 (downstream of Queanbeyan STW).

lagged correlations are shown in Figure 2 b-d . With each month's lag, the correlation value increased from 0.58 to 0.63 to 0.72. A similar relationship was observed between blue-green algae in Lake Burley Griffin and total phophorus levels in the Molonglo River at site 3, where the unlagged correlation coefficient was 0.46. When log blue-green algal numbers were lagged by 1 and then by 2 months, the correlation value once again increased (to 0.52 and 0.58 respectively) . However, when blue-green algal numbers were lagged by three months the correlation value decreased. It appears that for total algae the most appropriate lag is 2-3 months , whereas for the smaller and faster gr"owing bluegreen algae, the most appropriate lag is 1-2 months. Construction of Water Quality Models

Two relatively similar models were constructed to describe the response of lake algae to variations in environmental conditions and changes in phosphorus inputs from the Queanbeyan STW. Both models were based on the observed influence of: (i) water temperature, (ii) water flow volume, and (iii) total phosphorus load ·upon Lake Burley Griffin blue-green algae . The expected casual sequence on which the models are based is indicated in Figure 3. The dependent variable for Model I was log blue-green algae (average of Lake Burley Griffin sites 1 and 2), lagged by 2 months. The independent variables were: (i) water temperature, average of Lake Burley Griffin sites I and 2 (oC), (ii) log of (mean depth / hydraulic residence time) calculated from lake mean depth, lake volume and water flow volume (log m.year- 1) , (iii) log total phosphorus load from the Queanbeyan STW (log Kg P.day-'), (iv) log total phosphorus load from the Molonglo River location 1, upstream of the Queanbeyan STW (log Kg P.day- 1). The statistical information on which Model 1 is based is indicated in Table 2. Model 2 was a simplified version of Model I above. The dependent variable for Model 2 was also log blue-green algae (average of Lake Burley Griffin sites I and 2), lagged by 2 months. The independent variables were: (i) water temperature, average of Lake Burley Griffin sites 1 and 2 (oC), 16

WATER March, 1988


• Figure 3. Casual sequence of predictive models showing response of blue-green algae in Lake Burley Griffin.

(ii) log river flow at Molonglo River site 1 (log m 3 .sec- 1), and (iii) total phosphorus concentration in Queanbeyan STW effluent (mg P .L- 1 ). In this model, log (mean depth / hydra1>lic residence time) was substituted by log river flow, because the latter is readily and regularly determined, and is one of the main factors influencing the former parameter.


= Independent


= .2WT - 1.45 LOG(M/ HRT) +2 .6 LOG

n multiple R multiple R' standard error of est imate

= 29 = 0.82 = 0.67




= =


= Water Temperature in Lake Burley Griffin



Log of ratio of Mean Depth/ Hydraulic Residence Time (log m.year-')



Log of Total Phosphorus Load from the Queanbeyan STW (log kg P.day-•)


= Log of Total Phosphorus Load from the

Constant 4.43

1.05 Log Blue-Green Algae in Lake Burley Griffin sites I and 2 with 2 month lag (log cells.mV') sites I and 2 (°C)

Molonglo River site I (log kg P .day-')

The two log total phosphorus loading parameters (Queanbeyan STW site and Molonglo River site 1) were substituted by total phosphorus con~entration at the Queanbeyan STW. This was possible because with the high phosphorus load from the STW site , the impact of the relatively low load from the Molonglo River was negligible. Furthermore, log total phosphorus load from the Queanbeyan STW was substituted by total phosphorus concentration at that site, because once again, the latter is readily and regularly determined and is one of the main factors influencing the former parameter.

The statistical information on which Model 2 is based is indicated in Table 3.


= Independent = 0. 190 WT -



n multiple R multiple R' standard error of estimate


1.04 LRF + 0.300 TP 3. 10.

= 29 = 0.84 = 0.70


= 0.98 = Log Blue-Green


= Water


= Log


= Total P hosphorus concentration at the

Algae in Lake Burley Griffin sites I and 2 with 2 month lag (log cells.mL'')

Temperature in Lake Burley Griffin sites I and 2 ( 0 C)

River Flow in the Molonglo Ri ver at site I (log m' .sec- ') Queanbeyan STW effluent site (mg P . L'')

Each of the three independent variables contributed a significant amo unt of explanation to Model 2, and together they explained 70% of the variance in blue-green algal levels which was a slight improvement on Model 1. Under typical summer conditions, different levels of total phosphorus from the Queanbeyan STW would result in different levels of blue-green algae in Lake Burley Griffin. These summer conditions , and the levels of blue-green algae as predicted by the two models are shown in Table 4.

DISCUSSION Resu lts of the study indicate that phosphorus inputs from the Queanbeyan STW strongly influence the concentration of algae in Lake Burley Griffin. This influence is particularly important under conditions of high temperature and low flow. Because such conditions are relatively common under Australian summers, the control of phosphorus has important implications for water management in Australia. Two water quality models were produced to indicate the relationship between Lake Burley Griffin algae (dependent parameter), and water temperature, water flow and phosphorus (independent parameters). Model 1 was conceptually similar to the Vollenweider approach (1969, 1975 and 1976), which has been used in previous descriptions of Lake Burley Griffin eutrophication (Cullen et al., 1978; Gutteridge Haskins and Davey Pty Ltd, 1983), whereas Model 2 was a simpli fied version of Model I. Chlorophyll-a the dependent parameter in the Vollenweider model, was replaced by log bluegreen algae (lagged 2 months) in both Model I and 2 because algal counts were regarded as a more accurate and reliable measure of algal biomass than chlorophyll-a concentrations. Although several chlorophyll-a models were constructed during the study, their predictive powers were low when compared to Model I and

2. Several management options are suggested by these two water quality models, not just for Lake Burley Griffin, but for other Australian lakes as well. Because of the relationship between algae, temperature, flow and phosphorus, the most critical times for phosphorus control are the summer months . Therefore, all care must be taken to optimise phosphorus removal processes during these critical summer months. At facilities with land irrigation of effluent, or occasional phosphorus removal, care should be taken to utilise these management options as much as possible during periods of high temperature and low flow. The two regression models provided similar, although slightly different predictions of blue-green algal levels in Lake Burley Griffin, fo llowing phosphorus reduction at the Queanbeyan STW (Table 4) . However, both models indicated that as long as the concentration of total phosphorus at the Queanbeyan STW was

maintained below 1.0 mg P .L- 1 the levels of blue-green algae in the lake would be under I ,000 cells.mV' during average summer conditions . The anticipated changes at the Queanbeyan STW, aim to reduce phosphorus under the 1.0 mg P.L- 1 level. Consequently, based on the results of the abovementioned algal response models, the anticipated reductions at the Queanbeyan STW would significantly reduce the phosphorus load as well as the level of blue-green algae in Lake Burley Griffin .

TABLE 4: THE RESPONSE OF BLUE-GREEN ALGAE IN LAKE BURLEY GRIFFIN TO VARIATIONS IN TOTAL PHOSPHORUS INPUTS FROM THE QUEANBEYAN STW EFFLUENT, UNDER TYPICAL SUMMER CONDITIONS USING MODEL 1 AND MODEL 2 Typical summer conditions: temperatu re o f water in Lake Burley Griffin sites I and 2 = 22( C) ri ver now in Molonglo River site I = 0. 1 (m' .sec· •) log ri ver flow in Molonglo River site I = - 1.0 log (m' .sec- ') concentration of total phosphorus at Molonglo River site I = 0.030 (mgP .L'') load of total phosphorus at Molonglo River site I = 0. 26 (kgP.day· •) log load of total phosphorus at Molo nglo River site I = - 0.59 (log kgP.day- ') fl ow from Queanbeyan STW = 0.081 (m' .sec-') mean depth/ hydrauli c residence time = 1.08 (m .year· ') log (m ean depth / hydraulic residence time) = 0.03 log m. year· • 0

Total Phosphorus Concentrations Queanbeyan ST W (mg P.L·')

Log Total Phosphorus Load Queanbeyan STW (log kg P.L· ')

8.0 6.0 4.0 2.0 1.0 0.5 0.2

1.75 1.62 1.45 1.1 5 0.84 0. 54 0.1 5

Log Blue-Green Algae 2 months later in Lake Burley Griffin Model I (log ce/1.mL· ')

Model2 (log cell. mL · •j

4.4 4.0 3.6 2.8 2.0 1. 2 0.2

4.4 3.8 3.2 2.6 2.3 2.2 2. 1

The steps used in this study provided an' approach for quantitatively describing and predicting blue-green algae, which represent the major water quality management issue for Lake Burley Griffin. The models provided a means for evaluating various management options in relation to irilproving water quality in the lake. If required, the same general approach could be used on any one of the other parameters measured during the 3 year monitoring program. Furthermore, provided sufficient representative monitoring information is available, the same general approach could be used in other aquatic environments.

ACKNOWLEDGEMENTS The authors thank the staff of the Water Quality and Investigations Laboratory, The Office of ACT Administration and officers of the National Capital Development Commission for their assistance in the project.

REFERENCES BAYLY , I. A. E. and WILLIAM S, W. D. ( 1973). Inland Waters and Their Eco logy, Lo ngman , Melbo urn e. CU LLEN , P., ROSIC H , R. S. and BEK , P . (1978). A Phosphorus Budget for Lake Burley G riffin and Management Implicatio ns for Urban Lakes . Technical Paper No . 31 , Australian Water Resources Council, Canberra. CULLE N , P. and SMALLS , I. C. (198 1). Eutrophication in semi-arid'areas th e Australian experience. WHO Water Quality Bulletin 6, 79-9 1. DEPARTMENT OF CON STRUCT ION/ BINNIE INTERNATIONAL/MAUNSELL AN D P ARTNERS (1978). ACT Region Water Qualit y Study. Report to th e National Capital Development Commission . DEP A RTME NT OF HOU SIN G AND CONSTRUCTION ( 1982). Water Quality Monito rin g - La ke Burley Griffin and Lake Ginninderra , December 198 1 to June 1982. Report to the National Development Commission. DEPARTMENT OF HOUS ING AND CONSTRU CTION (1984). Wat er Quality Monitoring - Lake Burley Griffin and Lake Ginninderra , Jul y I 982 to June 1983. Report to th e national Capital Development Commission . DEPARTMENT OF T HE CAPITAL TE RRITORY (1981). Lake Burley Griffin: a situ ati on report Commo nwealth Government Printer, Canberra .

Continued on page 43 WATE R March, /988


National Hazardous Waste Management Conference Report by Errol Samuel The Second National Hazardous Waste Management Conference, organised by the Institution of Engineers, Environmental Panel and the Australian Water and Wastewater Association , was held in Sydney, November 15-18, and attracted an attendance of 231 delegates from all States of Australia and from New Zealand. Highlights included: the Opening Address, Keynote Address, papers by three overseas experts and a workshop on siting of hazardous waste facilities - all of which are summarised below. In addition, there was an afLer dinner address by Daryl George, Chief Executive of the Confederation of Austral ian Industry and a site visit to the Metropolitan Waste Disposal Authority's aqueous waste treatment plant which is due to commence operations in July , 1988. The Conference was addressed by a total of 28 speakers and included sessions on waste minimisation, emerging technology, aqueo us waste management, incineration, transportation, landfill management and siting. A trade exhibition formed part of the Conference.

OPENING ADDRESS Senator Graham Richardson, Federal Minister of Enviroment and the Arts

Senator Richardson drew the attention of the Conference to the enormous cost of the improper disposal of industrial wastes and cited clean-up costs of $20 to $100 billion in the USA and $10 billion in the Federal Republic of Germany. The Minister announced the formation of a joint Victorian, NSW and Commonwealth Government taskforce (details on page ??) to investigate the problem of disposal of intractable wastes, which mainly comprise stab le organochlorine compounds such as PCBs. At present intractable wastes not stored, are shipped overseas (at great cost) for high temperature incineration and A ustralia is poorly placed to deal with unexpected hazardous waste problems when they arise. The recent banning, from agriculturaJ use and recall of, organochlorine pesticides was cited as an example. The Minister stressed the importance of control of hazardous wastes and of the adoption by the Australian Environment Coun cil of 'National Guidelines for Managing Hazardous Wastes'.

KEYNOTE ADDRESS Peter Horsley, Director of Metropolitan Waste Disposal Authority, Sydney

Mr Horsley pointed out that Sydney, a majo r manufacturing and processing centre, produced a relatively small amount (67 000 tonnes/ year) of industrial liquid wastes not suitable for discharge to sewer. These wastes are only 0.04% of the effluent from industrial premises and they are transported by road for disposal at the MWDAs secure landfill - the Castlereagh Depot. The MWDA has always regarded landfill as an interim method of disposal of liquid wastes and has, over a number of years, endeavoured to establish a plant for treatment of these wastes. With the long lead time necessary for development approval for a treatment faci lity, a number of extensions of the depot has been -necessary, the latest extension bringing the total land area to over 300 hectares. Towards the end of 1985 the MWDA obtained development approval for the establishment of an aqueous waste treatment at Hill Road, Lidcombe. This plant will cost about $23 million and commissioning is scheduled to commence in April 1988. Mr Horsley concluded by indicating that the 1990s are likely to see increasing standards in all areas (including the transportation) of hazardous waste management and a greater public awareness of the activities of the industry.

Errol Samuel is a Federal Councillor of the A WWA representing the New South Wales Branch. 20

WATER March, 1988

VISITING SPEAKERS Marcia Williams Marcia Williams, Director of the US, EPA, Office of Solid Waste and Emergency Response presented a paper titled International Perspectives on the Management of Hazardous Wastes. Ms Williams considered that establishment of a national regulatory control program with appropriate legislation , regulations and permits, is the single most important step in protecting the environment and public health from the mismanagement of hazardous waste. The common elements of a control program include: definitions of hazardous waste, responsibilities of waste generators and transporters, design and operating requirements for management facilities, permitting of faci lities, clean-up requirements for past and future releases, delegation of regulatory control, strategies for enforcement, and public participation in the regulatory and permitting process. Common elements of a regulatory control program were discussed from an international viewpoint along with the control framework estab lished in the United States. Observations based on the US experience in developing and implementing each element of the framework were outlined. Highlights of concerns and initiatives of some internll,tional organisations were summarised. Finally, the paper concluded with a discussion of important factors to be taken into consideration in the development and implementation of a control program based, on experience over the last decade, in the USA. In this most informative paper Ms Williams made the following comments upon various aspects of hazardous waste disposal in the United States. , • On the Resource Conservation & Recovery Act 1976 - 'One highlight of this amendment was to completely prohibit the land disposal of untreated hazardous waste by the end of this decade. Another highlight was to require ~ermitted hazardo us waste facilities to clean up their past and future releases'. • On the Toxic Substances Control Act - 'TSCA goes even further, allowing EPA to regulate the manufacture and disposal of substances that threaten human health and the environment' . • On the stigma of calling a waste 'hazardous' - 'At times we can better protect the environment by regulating the waste without labeling it as "hazardous".' • On liability - 'All parties involved in the generation , transport, or disposal of the released hazardous substances are liable unde the law. For example , a generator who disposed of a waste commercially is a responsible party even if subsequent handlers of the waste are negligent.' • On the Public and Risk - 'We have learned that the public must be listened to. The public is very uncomfortable with discussions of risk and analysis of cost and benefits. They tend to want zero risk which is not a very realistic goal in all situations.' Ken Simpson

Ken Simpson of the Alberta Special Waste Management Corporation presented a paper discussing 'Obtaining Approval for and Operating a High Temperature Incinerator Facility'. An integrated hazardous waste treatment plant which includes a high temperature incineration, physical/chemical treatment plant and secure landfill was commissioned in late 1987 in the Province of Alberta in Canada. Mr Simpson indicated that the successful siting of this treatment centre is a first for North America where the usual reaction from local citizens has been 'not in my backyard'. By using open dialogue with the public and establishing trust and confidence, several Alberta communities responded to the call to be host of the special waste treatment centre. Each proposal met the necessary environmental criteria and for each a plebiscite indicated that the majority of citizens were in favour of hosting the facility. From the sites recommended by the local communities, the Alberta government chose Swan Hills, a site about 200 km from Edmonton.

The treatment centre cost about $55 million and will initially have a design capacity of 15 000 tonnes/ yea r. A n interesting featu re o f the plant is that the inci nerator used is o f the rocking kiln design and is the first commercial incinerator of this ty pe. The plant will emp loy about 46 people. Monitoring o f air, sur face and groundwater, soil and vegetation is an ongoing requirement of the facility.

Kathryn Kelly Kathr yn Kelly o f Environmental Toxicology in th e USA gave the confe rence an overview o f the health risk assessment process as it is applied to the evaluation of hazard ous was te incineration fac ilities. Dr Kell y's paper discussed the sources, path ways a nd recepto rs of ex posure to fa cility emissions; available models for estimating environmental fate, tra nsport and health risk; assessing risk o f emissions once the o ffsite concentrations are known; and some o f the limitatio ns and ass ump tions inherent in the risk assessment process .

WORKSHOP A panel of local and overseas experts and representati ves of enviro nmental gro ups, under th e Chairmanship o f Richard Connolly, discussed fac tors of importance in siting the process. The consensus of opinion seemed to be as follo ws:

• Go public early in the siting process . Do not choose a site on tec hnical grounds and then try and j usti f~, it sociall y. • As k the community to propose a site, but do not ex pect this to be a fi rm commitment. • Wor k in small groups in order to aid the communication process. • Respond honestly to citize ns' concerns. • Be prepared to over design and budget about 25% above an adequate design. • State of the a rt monitoring of the environment should be undertake n befo re and after esta blishment of the plant. • Set up a local ad visory committee and hire consultants to represent the communi ties interests . • Listen to the public as they can often make a valuable contribution. • T he faci lity's name should have positive connotations. At the end of the siting wo rkshop, a representative of the Melbourne and Metropolita n Board of Wo rks advised that the Victorian Governmen t had announced the proposed location for its liquid was te treatment fa cility - on land off Holden Road located within the Shire of Melton . The proposal is presently open to submissions by individuals and groups and will be assessed by an independent panel appointed by the Minister of Planning and Environment. Copies of the con fe rence papers are obtainable fr om David Russell , Cl - Camp Scott Furph y, PO Box 994 , Chatswood, NSW 2067. Price $35 (A ust. ) including mailing.

Joint Taskforce on Intractable Waste Errol Samuel At the second National Hazardous Was te Management Confe rence Senator Graha m Richardson a nnounced that the Commonwealth, New South Wales and Victo rian Go vernments had jointly established an independent Taskforce to examine and advise on the minimisation and management of intractable hazardous waste and the development of fac ilities in south-eastern Australia for its disposal. The more commonly known o f such wastes in A ustralia includes those containing polychlorinated biphenyls (P CBs), hexac hlorobenzene a nd organochlorine pesticides such as DDT . The Taskfo rce will report to the Ministers of th e various gove rnments responsible for its format ion. An advisory committee with members representing the sponsoring government age ncies , industry, trade union and conservation movements and local government , will ac t as ad visers to the Taskforce .

TERMS OF REFERENCE T he Taskforc e is to undertake its wor k in three phases and the mo re significant as pects of th e terms o f reference for each phase ar.e as follow s: Phase 1 - to produce a preliminary report to Ministers within 4 month s which: I . provides a genera l background on the nature of the prob lems o f intrac ta ble waste management ; 2. assesses the current and estimated future demand for disposal fac ilities tak ing account of stocks and generation rates of intractable was tes; 3. reviews existing and emerging technologies for intractable waste destru ction ; 4. recommends regulatory mechanisms for operation of a disposal facility; 5. evaluates the economic viability of a facility ta king into account its likely size and type; 6. discusses means of ensuring that the operation of a fac ility does not provide an incentive for th e production o f was tes. 7. outlines options for the management , ownership and monitoring of a faci li ty and community involvement in these activities;

Errol Samuel is a Federal Co uncillor of the A W WA representing the New So uth Wales Branch.

8. discusses in general terms operational, transport infr astru cture, a nd occupational health and safety aspects, environmental standards, and associated insurance and decommissioning procedures; 9. reviews and recommends siting criteria for south-eastern Australia ; ~ 10. ge nerally considers quantita ti ve risk assessment of a site; 11 . provides preliminary advi ce on a strategy for local community consultations durin g Phase 3; 12. proposes a pac kage of measures, and identifies costs and benefits, for local communities arising from a facility; 13. makes recommendations on a preferred type and size of faci lity, and all necessary associated arrangements; and 14. analyses public input on items 1- 13 followin g adve rtisements seeking such input and consultations with identified interest groups in the community. Phase 2 - fo llowing consideration by Ministers o f the Phase 1 re port , and assuming a decision by government to proceed with the identi fication of possible locations for a facility, the tas kforce, wi thin 3 months, is to: 1. prepare the phase 1 report fo r release for public comment with revisions as appropriate; 2. provide a report which • identifies two or three sites each in Victoria and NSW which meet the selection critiera; • recommends community consultation processes fo r the four to six sites chosen; and • provides analysis of and ma kes recommendations for eac./1 site, with regard to require ments relating to environmental , health and safety considerations. Phase 3 - with the ap proval of Ministers, over a 2 month period ass ist in consultations with communities in the localities of the sites identified in Phase 2.

ISSUES PAPER In December 1987 the Taskforce published an 'Issues Paper' on which it sought public comment. The Taskforce hoped that this paper could assist interested individu als a nd groups to prepare submissions which the Tas kforce would take into account in preparing its preliminary report. The Iss ues Paper can be obtained fr om the Task force and is summarised below. WATE R March, /988


Background - The governments making up the Australian Environment Council have agreed on common objectives for the management and disposal of hazardous industrial wastes, including the introduction of the polluter pays principle and high priority for waste reduction, recycling and exchange. However, kP.owledge of the hazardous wastes stream is still fragmentary and the approaches in the Commonwealth , Sta tes and Territories are often inconsistent, overlapping and incomplete. Following the failure of several proposals for integrated waste plants and high-temperature incineration faci lities and subsequent representations, particularly from major environmental groups on the need for cooperative action, the Commonwealth, New South Wales and Victorian Governments established this Taskforce to stud y and make recommendations on the management , minimisation and disposal of intractable wastes. Intractable Wastes - Intractable wastes have been defined as wastes for which environmentally acceptable means of disposal are not presently available. In Australia it has come to be applied, in particular , to a fair ly narrow range of very stable organochlorine wastes including PCBs , hexachlorobenzene, dioxins, organochlorine pesticides such as DDT and residues such as those from the plastics industry . In the past, some of these wastes were disposed of to land fi ll , a practice which is no longer regarded as environmentally responsible . In many overseas co untries they are destroyed by high-temperature incineration at l 100-1200°C. This technology is very expensive, but there are many high temperature incinerators operating overseas at very high standards of safety and destruction efficiency. As there is no high-temperature incinerator or other disposal facility in Australia, the commonly identified intractable wastes are being stored pending an acceptable disposal method becoming available. In recent years a small proportion of Australia's identified intractable waste stockpile has been shipped overseas each year at great expense ($5000 or more per tonne) to be incinerated in the U .K. An important question is whether there are wastes other than the commonly listed intractable wastes which are not now being disposed of in an environmentally acceptable manner. There are also other wastes (eg, some sludges , tars, spent organic solvents) which are currently being landfilled in Australia because no more environmentally acceptable disposal method is available. Whether landfilling is environmentally acceptable in these circumstances and whether such wastes might be managed better if a suitable facility were available, is a matter for consideration. The Taskforce will therefore examine the requirements for the management and disposal of some wastes in addition to the commonly identified stable organochlorines. Public input on the range and types of wastes which should be regarded as intractable will assist the Taskforce's work. Proposals for Intractable Waste Destruction - Since 1980 there have been at least eight proposals in Australia to build facilities for the destruction of intractable wastes. All these proposals have been based on high temperature incinerators and all have led to public opposition to their siting. As a consequence the Victorian and NSW proposals have been abandoned. Despite the siting difficulties there is still general agreement that there is a need to find some environmentally acceptable means of disposing of intractable wastes. There are many disposal facilities overseas all of which are based on high temperature incineration which currently appears to be unique in its ability to handle all types of intractable waste regardless of physical form, chemical composition and concentration. There are, however, several emerging disposal processes which the Taskforce will examine. The Taskforce will need public comment on the feasibility and efficacy of various disposal processes. International Comparlsons - It is reasonable to expect that the per capita generation of hazardous industrial waste would be greater in the larger and more industrially-developed countries such as the USA , West Germany and the U.K. than in Australia. However, it is of significance that in many European countries where the population is of the same order as the population of Australia and where industry is probably no more developed or complex, highly developed industrial waste management systems have been installed. Sweden, with about the same population , is estimated to produce 150 000 to 200 000 tonnes ¡ of hazardous 22

WATER March, 1988

waste annually; much of this is disposed of onsite but 30 000 to 40 000 tonnes are incinerated at a high temperature plant of mixed government/ private ownership. Much of the debate in Australia has been focussed on high temperature incineration for organochlorine wastes. Existi ng estimates indicate that there are about 9000 tonnes of these materials in stock with a bout 1000 tons being generated annually. In considering whether high temperature incineration is the best option and recognising that substantial amounts of auxiliary fue l must be supplied to incinerate organochlorine wastes, the Taskforce must take into acco unt the case for the co-incineration of non-organochlorine hazardous wastes of high calorific value (which are currently disposed of by other means) in order to reduce the amount of premium fuel , which would otherwise have to be used , and to ensure proper p.rocess ao ntrol. Hazardous Waste Management - In most OECD coun tries, policies establish the following o rder of priority for dealing with chemical waste: prevention , minimisation, recycling, treatment, incineration, landfill. Management effort is directed along these lines by legislation covering the whole spectrum of waste management from generation through transportation to ultimate disposal, with appropriate controls on emissions associated with the disposal process. In Australia, an approach to a common national strategy has been in progress since 1979. The Australian Environmental Council adopted national guidelines for the management of hazardous wastes in 1986. Facility Ownership and Management Ownership and management of centralised disposal fac ilities in other co untries cover a wide range of options involving public and/ or private entities. The owner may operate the facility or may engage a contractor for this purpose. The Taskforce must consider the best overall solution for an Australian faci lity and welcomes contributions from the public on this problem.

THE T ASKFORCE , The Taskforce has four members, the Convenor is Dr. Gavan McDonell and it has offices in Melbourne and Sydney, addresses and contact numbers being: Melbourne: 6th Fir, 601 Little Collins St. , Melbourne, Vic . 3000. Tel. (03) '6 15 4425. Sydney: 7th Floor, 157 Liverpool St., Sydney, NSW 2000. Tel. (02) 265 8967. •

LETTER The Editor Sir, We wo uld like to clarify comments relating to the Ballarat South Treatment Plant made by Graham and Peck in their paper 'Water Quality Management in Victoria' in the December issue of 'Water'. The upgraded Ballarat South Treatment Plant will incorporate a biological phosphorus remo val step which was modelled on the extensive pilot work carried out by CSIRO . However this plant is not designed to remove nitrogen; it only provides a nitrification step to reduce ammonia concentration below 10 mg/ L. Selection of the final process configuration was carried out in consultation between the Board, CSIRO and a consultant team of Caldwell Connell Engineers Pty Ltd and Binnie & Partners Pty Ltd . Detail design and documentation was undertaken by the consultants.

Dr G. J . Sewards Binnie & Partners Pty Ltd

J. C . Co ulson Caldwell Connell Engineers Pty Ltd

PHOSPHORUS MOVEMENT THROUGH SANDS MODIFIED BY RED MUD M. Kayaalp, G. Ho, K. Mathew and P. Newman ABSTRACT Field and laboratory column experiments show that red mud (waste from bauxite refining) neutralised with gypsum increases Phosphorus (P) sorption capacity of sandy soils. Some 1.68 g P / kg RMG (red mud neutralised by 50Jo gypsum) was sorbed from 9.5 mg/ LP solution during a continuous flow of 50 cm/ d for 750 pore volumes. Batch tests indicate almost no desorption of sorbed P, however during continuous flow leaching, a total of 140Jo of the sorbed P was desorbed. During flooding - drying cycles of laboratory columns, with secondary effluent, 91 OJo P removal occurred through 300Jo RMG (300Jo RMG, 700Jo Bassendean sand mixture) , and 630Jo removal through 200Jo RMG , both )Vith some more capacity to sorb P; whereas lOOJo RMG continued to sorb P at 500Jo efficiency even after the calculated sorption capacity was exhausted. Initially previously sorbed P in soil would be leached out by the alkaline leachate from red mud incorporated above it. In general the potential for P removal by renovated soils seems high.

INTRODUCTION At Murdoch University, a three year project was conducted to examine the potential of red mud (a waste product from bauxite refining) as a sandy soil modifier for waste management purposes. The project covered characterisation of the properties of red mud neutralised with gypsum, copperas, and other materials,


Carnac I.,, Garden


Pt Peron





Figure 1. Geomorphic elements of the Swan Coastal Plain. Ridge Hill Shelf (R), Pinjarra Plain (P), Bassendean Dunes (B), Spearwood Dunes (S), Quindalup Dunes (Q). Source: McArthur and Bettenay, 1974. 26

WATER March, 1988

M. Kayaalp

Dr K. Mathew

Dr P. Newman

Mehlika Kayaalp is a Civil Engineer with a Post-graduate Diploma in Sanitary Engineering (Dip.SE) from Delft University, The Netherlands and is currently completing a Masters Degree in Environmental science at Murdoch University. She has ¡ been working in the areas of waste management, wastewater treatment and disposal, drinking water treatment, and environmental aspects of engineering projects. Dr Kuruvilla Mathew trained in India and Holland in Civil and Environmental G.Ho Engineering, completed his doctorate at Murdoch and has since been researching wastewater techniques using soil. He has extensive experience in low cost water supply and waste systems in remote areas through work in India and Nepal. ' Dr Peter Newman is Senior Lecturer in Environmental Science, Murdoch University. He has a B.Sc.(Hons.) and Ph.D. from University of W.A. in chemistry and a Post-graduate diploma in Environmental Science and Technology (Dip.ES&T) from Delft University, the Netherlands. His research interests are in urban transport, soil-waste interactions and remote area technology. He is currently a part-time member of the Environmental Protection Authority in W.A. Dr Goen Ho is a Senior Lecturer in Environmental Engineering, Murdoch University. He has a B.E. and M.Eng. & Sc. from the University of Sydney and Ph.D. from the University of Queensland. His research interest is in the treatment and utilization of industrial, mining and municipal wastes and wastewaters. and investigations by laboratory and field trials of the ability of red mud to renovate waste water through soil infiltration . In this paper, an aspect of this research project, the phosphorus sorption capacity of red mud by laboratory columns is described . The Swan Coastal Plain, W.A., extends from Geraldton in the north to Dunsborough in the south, and is bounded in the east by the Darling Fault. The Swan Coastal Plain is formed almost entirely of depositional material (McArthur and Bettenay, 1974). The water retention capacity of the Swan Coastal Plain's sands is very low, resulting in limited crop productivity of the soil. A further consequence is the leaching to the receiving waters of nutrients from the fertilizers used to improve soil productivity, a problem reaching major proportions in several large estuaries (Hodgkin, 1984). The sorption capacity of the soil for phosphorus (P) has been found to be very low (Cameron and Ho , 1984), and nutrients in wastewater applied to the sands for groundwater recharge have not been removed before the water reaches the water table (Mathew, Newman and Ho 1982). The application of neutralised red mud has been suggested as a possible solution for P retention (Barrow, 1982; Ho, Mathew and Newman, 1983; Ho, Newman, Mathew and de Potter, 1985; Glenister and Thornber, 1985). Figure 1 shows the Swan Coastal Plain about the Perth Metropolitan region. Bassendean sand and Spearwood sand occupy most of the area; both are characterised by high saturated

hydraulic conductivity, and low cation exchange capacity (Table 1), particularly Bassendean sand. The aim of this study was to investigate P movement through the sandy soils of the Swan Coastal Plain modified with neutralised red mud. Consequently the process involved in P movement in these soils was investigated , and an evaluation was undertaken to ascertain whether and to what degree soils modified with neutralised red mud retain P .




1. 2. 3. 4. 5. 6. 7.

0.15 2.4 2.61 1. 55 30

± o.oi ± 0.2 ± 0.04 ± 0.01 ± 2 4.6

0.1 8 ± 0.01 1.9 ± 0. 1 2.65 ± 0.01

Effective size, mm Uniformity coefficient Particle density, g/cm' Bulk density, g/ cm' Hydraulic coriductivity, m/ day pH Cation exchange capacity (Ammonium saturarion method) , meq/ 100 g 8, Adsorption equilibrium distribution coefficient for ammonium (mg/ kg)/ (mg/ L)

16 ± 1 5.6

0.5 6 ± 0.05

1.5 ± 0.2


0.70 ± 0.04

Source: Mathew et al. , 1982.

Characteristics of Red Mud Red mud for this research was obtained from the Kwinana refinery of Aloca Australia. The chemical and physical properties of red mud are tabulated in Table 2. Because of its high pH, red mud should be neutralised before mixing with sand. Several neutralising agents were tested for this purpose, the content per 1 kg of dry red mud being gypsum 50Jo, copperas 70Jo, 0.59 L Laporte effluent (acidic effluent from a titanium dioxide manufacturing plant containing waste sulphuric acid and ferrous sulphate), also 0.95 L of 1 M HNOJ, or leaching with 0.01 M CaCI, solution, and 0.03 M NaCl solution. Of all these, neutralisations of red mud by gypsum and copperas seemed to be the most promising (Ho, Newman, Mathew and Parker , 1984). During this study red mud used in the experiments was neutralised with 50Jo gypsum (CaSO •. 2H,O waste gypsum from the processing of phosphate rock into superphosphate).

various agents have indicated that the mud had a high sorption capacity for P (Ho, Newman , Mathew, and Parker, 1984) . Accordingly, a column sorption experiment was conducted to test the removal under conditions closer to field conditions. Two perspex columns, each with 4.3 cm internal diameter, were packed to a 5 cm height with a mixture of 700Jo Bassendean sand and 300Jo gypsum neutralised red mud (red mud-gypsum, RMG) . The RMG contained 950Jo red mud and 50Jo gypsum. The pore volume of the soil after packing was estimated to be 400Jo . Before the application of P solution the soil columns were backwashed with 0.01 M CaCI, solution in order to saturate the soil. AP solution of around 9.5 mg/ L concentration was prepared and its pH was adjusted to 7 .0. For every 10 litres of P solution 1 mL of chloroform was added to prevent P removal by bacterial action. The columns were fed with the solution using the Mariotte bottle arrangement, and leachate flow was regulated using a peristaltic pump. The superficial velocity of flow was maintained at 50 cm/ d. Effluent samples from the columns were collected with a fraction collector, each sample being almost 15 cc in volume, half the pore volume of the packed soil. One of the columns was stopped when the effluent from that column reached about 600Jo of the initial P concentration , the other column was kept running for a month or so until breakthrough was reached. Outlet samples were collected daily and pH was measured immediately. The samples were kept at - 10°C if P concentration was not immediately determined. Good agreement was obtained between the results of both columns . For up to 20 pore volumes there was no P leaching (Figure 2). Sixty per cent of the initial P concentration was reached at 120 pore volumes. Approximately 990Jo of initial P concentration was reached after 750 pore volumes had drained through the column, and a total of 54 mg of P was sorbed by that time, corresponding to 1.68 g P/ kg RMG. : COLU MN 1 (S T OPPE D AT 8 0% INITIAL P) : COLU MN 2



0.9 0.8 0 .7

0.8 0 .5


0.3 0 .2


0.1 0

1020 50

100120 150




C · Conc•n l ratlon of phosphorus ol e lllue n t sam ple

TABLE 2. PROPERTIES OF RED MUD H ydraulic Conductivity Total Alkalinity

0.0003 mi d 5.0 g/ L as Na, CO,

Alkalinity (on dry weight basis)

In entrained liquor In desilication product CEC

2. 7 g as Na, CO, 39.2 gas Na, CO, 20.4 meq/ 100 g

Source: (Ho et al., 1984) .

MATERIALS AND METHODS Either PVC or perspex columns were employed for the four experiments described below. The soil was packed to a bulk density of 1.5 gl ee. To ensure this, the columns were packed with 1 cm or 2 cm increments of the soil wetted with 50Jo 0.01 M CaCI, solution and compacting it to the desired depth increment. The solution was fed to the columns using the Mariette Bottle principle or syphon arrangement. The laboratory experiments were carried out with at least two replicates. Phosphate concentration as phosphorus was determin. ed by the Murphy and Riley Method (1962). Colorimetric determination was performed on a Shimadzu UV-210A double-beam spectrophotometer at a wavelength of 882 nm and slit width of 15 nm using a 4 cm cell. The P solutions were prepared from potassium dihydrogen orthophosphate (KH,PO.). The samples were filtered through 0.45 um pore size membrane filter before being analysed and if not immediately analysed after collection were stored below - 10°C.





no . ol po re vol ume,

Co : Concentration ol pl'loapl'lorua ol Inlet (9 . 5 mg/ L)

Figure 2. Phosphorus adsorption by red mud-gypsum.

From the batch experiments P sorbed was found to be: • with lh contact: 0.86 g/ kg • with Id contact: 2.6 g/ kg For 5 cm RMG in the columns, the corresponding values should be: • with lh contact: 0.0282 g • with Id contact: 0.0852 g With a superficial velocity of 50 cm/ d, the residence time of the solution in the 5 cm columns was close to 1 hour. During this period the soil in the column was in contact with solutions of different concentrations, as the sorption capacity of the soil gradually became saturated from the top downwards. Due to continuous flow in the column, gypsum was expected to leach from the soil and hence the retention capacity for P to be reduced. The results indicate that phosphorus sorption by the RMG was much better than the calculated amount from the batch experiment results for a lh contact time, but less than for a contact time of ld. The pH of the samples varied between 7.5 and 8.7 (Figure 3), and did not show any rising trend over the 30 day period . The pattern of the daily measured pH values showed the fresher samples to have a pH about 8. 7, higher than samples taken earlier in the day (pH around 7.5). This may have been due to the dissolution of atmospheric CO, into the samples before pH was measured .

2. Phosphorus Desorption Experiment EXPERIMENTS AND RESULTS 1. Phosphorus Sorption Column Experiment Batch tests upon several samples of red mud, neutralised by

The desorption of the sorbed P in soils has been studied widely (Chakravarti and Talibudeen, 1962; Comargo et al, 1978; Barrow , 1980, 1983; Stuaney and Enfield , 1984) . If neutralised red mud is incorporated into the sandy soils in the field for the purpose of P removal from wastewater or as a WATER March , 1988




10 .0



8 .5






,. l,.X





l- { X·x l l\.;, 1/ lt lli,. lX l xllxltd·a l,.x \x'< u,.n~1xfx1 xlx ).XX/ _.xx x x ).x 1



i l x/ll


l 11



- ,ro--r,o-o--,'oo---,o',o----,,oro--',b-,o--,'b,0---,,00-,-...,0

•-•'. , 0

no. o l p o r e Yolume1

Figure 3. pH values -

phosphorus adsorption by red mudgypsum.

storage for P from fertilizer, the above experiment indicates that P will be sorbed, but it is also expected that some of this sorbed P may eventually be leached by rain. A desorption experiment was carried out to investigate how much of the sorbed P would be desorbed. The P saturated column of the sorption experiment was leached by distilled water. Initially, three pore volumes of the outlet solu. tion had P concentration equal to the original P concentration (Figure 4). This gradually dropped and at 55 pore volumes became less than 50Jo of the initial concentration; 7 .5 mg of P, (23.4 mg P / kg RMG) was desorbed in total. This amounts to 140Jo of the sorbed P value, though the results of the batch tests indicated almost none would be leached . During the desorption batch test, RMG was dried following saturation with P, therefore a stronger degree of bonding between P and RMG had taken place. On the other hand, during the desorption column experiment, drying of the soil at 65 °C was not carried out. Therefore the desorption of P was easier when the leaching started. The pH of the leachates was between 7 .60 and 7 .85, which indicate that after leaching of an additional 55 pore volumes, the red mud remained neutral.

laboratory columns . The following aspects were investigated: • the optimum percentage of RMG needed to be added to Bassendean sand; • the adsorption and leaching of P, NH., NO 3 ; • hydraulic conductivity of the columns. The NH. and NO 3 monitoring will be reported separately. In this paper the emphasis is on P sorption . Three columns were packed with different combinations of RMG and Bassendean sand. In the first column the RMG was 300Jo; in the second 200Jo; and in the third 1OOJo. The columns were made of PVC pipes having an internal diameter of 10.45 cm and height of 4.0 m. The lower 0.5 m of the columns were packed with gravel, then 1.5 m with Bassendean sand and above that, 1.0 m of the RMG Bassendean sand mixture. An overhead tank of 26 L capacity was provided for each column Secondary sewage effluent from the Westfield Wastewater Treatment Plant was pumped to these tanks from a storage tank at ground level. The columns were fed from the overhead tanks through a syphon arrangement. Cycles of 10 days flooding and 18 days drying were adopted (after Mathew, Newman and Ho, 1982), each cycle taking just under a month. Inlet and outlet samples from the columns were collected every day. After 9 cycles of operation of Column I, and 9 cycles of operation of Columns II and III the following results were obtained: column I removed 3845 mg of phosphorus; at each cycle the removal efficiency was high and the average of 9 cycles was 91 OJo: column II removed 3525 mg of phosphorus and the average removal was 630Jo: column III removed 4335 mg with a 500Jo average removal efficiency (Table 3).

TABLE 3. SUMMARY OF PHOSPHORUS BALANCE (Column Experiment Simulating P Removal from Wastewater) Cycle No .

P Conce ntr ation mg/L

10 .0

pH X

9 .0


O : pH values


8 .0

444 352 243



5 .0

0 Oo


8 . 25

l 2 3 4

8 .00

5 6 7 8


3 .0

00 2.0





o o

xx ~ · Xxx~x 0 C C C0


0 ~ -- , - ---r--10 20

7 . 50


7.25 X XX

- - , - - - , - - - ~ -- ~ - - - ' 7 .00 30 40 50 60

Figure 4. Phosphorus desorption experiment gypsum.

red mud-

3. Column Experiment to Simulate P Removal from Wastewater The aim of this experiment was to simulate groundwater recharge in Bassendean sand with secondary. wastewater in WATER March, 1988

% Removed

99 98 98 95 81 77 87 91 98


COLUMN II 101 1130 274 365 302 358 345 535 381 283 134 290 148 344 403 7


43 54 39 57 68 70 90


I) C

no . of por e volum es


1231 639 660 880 664 424 492 4 10

Amount Removed mg P

COLUM!" I 594 3 11 6 18 362 9 505 29 425 10'2 394 118 386 58 322 30 4 293

Total quantity removed

7 . 75 X

Effluent Quantity mg P

Total quantity removed



6 .0

4 .0


8 9




x : P concentrations

7 .0


597 629 371 534 527 512



Influent Quantity mg P

2 3 4 5 6 7 8

2998 714 937 968 11 23 733 786 512

COLUMN Ill 1291 1707 111 603 447 490 348 620 574 549 458 275 324 462 37 1 141

Total quantity removed

57 15 48 36 52 62 41 72


Removal efficiencies were in accordance with the amount of red mud-gypsum mixture in the columns. With Columns I and II, P sorption capacities were not yet exhausted as compared to batch test results (3845 mg sorption compared to 10036 mg expected capacity for Column I and 3525 mg sorption to 6690 mg expected capacity for Column II) . However, Column III sorbed beyond its capacity (4335 mg sorption compared to 3344 mg expected capacity).

A similar situation occurred with the small column sorption experiment (experiment I above). In time more phosphorus was¡ sorbed - most probably due to the release of new sites for sorption by way of precipitation during the long experimental period. This- is consistent with the findings of Barrow (1982).

4. Experiments with Spearwood Sand A mixture of red mud-gypsum was applied to the soil at the Kwinana Experimental Recharge Basin. The soil in this area is Spearwood sand. The wastewater app lied after the addition of the red mud-gypsum consisted of one third primary and two thirds secondary sewage effl uent from the Wastewater Treatment Plant. Initial samples from the bores and the pans beneath the basin contained similar or higher P concentrations than samples taken before the application of red mud-gypsum - contrary to the expectation that red mud-gypsum would sorb P as the previous batch and column tests indicated . An experiment was then devised to determine if the occurrence of P in the leachate , after the application of red mud-gypsum at Kiwnana Experimental Recharge Basin, was due to alkaline leaching of previously sorbed P from the sand beneath the basin. Three perspex columns, 4.5 cm internal diameter, were packed , as follows : Column A contained 20 cm of 30% red mud-gypsum and 70% Bassendean sand. Column B contained 30 cm of Spearwood sand from the spare recharge basin at Kwinana, which had received secondary sewage effluent for a number of years. Column C contained 20 cm of red mud-gypsum (30%) and Bassendean san d mixture over 30 cm Spearwood sand. The columns were leached by distilled water and leachate samples were analysed for P. Negligible amounts of P (less than 0.1 mg/ L) were found in the leachate from Column A as expected from the batch and the large column tests with red mud-gypsum previously discussed. Column B, which was packed with Spearwood sand alone, contained up to 5 .4 mg/ L of P which, after 11 litres of water leaching , decreased to a value of 0.4 mg/ L. This P would be due to the previous contact of the sand with the sewage . After contact with water containing no P it released these high P concentrations. It can therefore be further concluded that the sand was probably saturated with P at a level around the same as the concentration of P in the effluent (about 10 mg/ L). Column C represented the situation in the field and leached high amounts of phosphorus. The average P concentration was 7 .4 mg/ L after 1.25 L of leachate had been collected . Since no significant amount of P came from red mud-gypsum (Column A) , it must have derived from the Spearwood sand . This amount was higher than the P leaching from the Spearwood sand (Column B). The extra P was probably due to the high pH of the red mudgyps um leachate (around 8.3-8.5) which shifts the equilibrium towards P desorption from the sand (Figure 4). This experiment adequately explains the high P concentration in the monitoring bores at the Kwinana Experimental Recharge Basin. It can be expected that the P concentration in the effluent will. decline to a very low level in time.

In the case of agricultural application, since the input of P is not continuous and P uptake by plants takes place, replacement of the RMG layer will not be necessary. The results of a simulated rainfall application test showed that only 1OJo of the applied P (super phosphate as 80 kg P / ha) leached out from RMG in a oneyear period (Ho et al., 1984) . During the 8 to 9 months of operation of the column recharge experiment, Column I (30% RMG) was only 38% saturated with P and had a removal capacity of more than 90% and, it continued to be above 90% during a further 11 cycles or 11 months operation . (Ho et al, 1986); Column II was 52% saturated with 630Jo removal efficiency, and Column III (l0OJo RMG) was oversaturated (1 .29 times) in relation to amounts expected from the batch tests, with a removal efficiency of 50%. According to the results of the P sorption experiment mentione above (experiment 1), during the Kwinana Recharge Basin trials with 1 m of RMG, breakthrough of P should not occur during the first 8 m of infiltration or 16 days of flooding at an infiltration rate of 50 cm/ d ; furthermore, the RMG in the basin should remove Pup to 300 m of infiltration, although the percentage removed would decrease with time. The experiments with Spearwood sand showed that P leaching from the RMG app lied layer at the Kwinana experimental recharge basin was from the sand beneath that layer, which had been contaminated with P by a previous effluent discharge to the basin. T his leads to the conclusion that RMG application onto sands already containing sorbed P will release P as result of the alkaline leachate from red mud during the initial application of water, however this will soon stop. Parallel studies at Murdoch showed that RMG modified sand has twice the availability of P to plants, compared to sands alone, i.e . 48 ug/ gP compared to 24 ug/ gP (Yeates et al, 1984) . Another result was the lack of significant difference between the bacterial populations of soils modified with red mud and that of soils in the field without any modification. The only adverse result was the amo unt of salts added to the groundwater from RMG application which was 20 times that of sand without modification. The amount of salts produced can be reduced by minimising gypsum addition and relying on plants to neutralise it. Other results indicated that concentrations in leachate were only F, 1 mg/ L; Cd, 0.02 mg/ L; and Fe, 148 mg/ L. These are acceptable concentrations for the beneficial uses of the groundwater (Ho et. al. 1987). A ll these studies lead to the coÂĽclusion that red mud in neutralised form such as RMG can be used as a soil modifier for agricultural, as well as secondary sewage effluent recharge purposes. The only technical problem remaining is the ultimate disposal of RMG layers in the field when their sorption capacities are exhausted and / or their alkalinity is leached out. However, bringing together two wastes, namely gyps um and red mud appears to resu lt in an environmentally acceptable and useful product.

ACKNOWLEDGEMENTS The authors wish to acknowledge the financial support of ALCOA of Australia and the Western Australian Water Authority.

REFERENCES CONCLUSIONS The results of sorption and leaching experiments indicate that RMG greatly improve the P holding capacity of the Swan Coastal Plain sands. Sorption experiments showed that RMG sorbed Pup to 750 pore volumes. During that time a total of 54 mgP or 1.68 . gP / kg RMG was sorbed. Moreover, during the application of the first 20 pore volumes no P leached out, with a total of 5.4 mg (0.17 g P / kg RMG) sorption. If secondary effluent from a sewage treatment plant (serving 20 000 to 25 000 people, with a daily discharge of 8000 m 3 ) is to be recharged to aquifers through 1 metre depth of RMG modified soil, the area required would be about 17 ha for 1 year continuous flow. During the first 170 days, no P would be expected to leach from the soil beneath the RMG layer. If a leachate of 50% initial concentration is considered acceptable , then with one application of RMG, secondary sewage effluent could be treated for 2.32 years . If on the other hand RMG layer is replaced every year, only l0OJo of initial P concentration will leach during the second half of the year (after 170 days).

BARROW , N. J. (1980). Differences amongst a wide-ranging collection of soils in the rate of reaction wit h phosphate. Aust. J. Soil Res., 18, 215-214 . BARROW, N. J. (1982). Possibili ty of using caustic residue from bauxite for improving the chemical and physical properties of sandy soils. Aust. J. Agric. Res., 33 , 275-285. BARROW , N . J. (1983). On the reversibility of phosphate sorption by soils. Journal of Soil Science, Vol. 34(4) , pp . 751 -758 . BOLT, G. H. and BRUGGENWERT, M. G. M. (1976). Soil Chemistry A. Basic Elements. E lsevier Scientific Publishing Company, Amsterdam . CAMERON, I. a nd HO , G . E. (1984). Phosphorus movement through sandy soils and groundwater in the Peel-Harvey Catchment Area . In Hodgkin, E . P. (ed.) 'Potential for Management of the Peel-Harvey Estuary' . Bulletin No. 160, p. 249-253, Department of Conservation a nd Environment, W.A. CHAKRAVARTI , S. N . and TALJBUDEEN (1962). Phosphate equilibria in acid soils . J. Soil. Sci. 13 , 231-240 . DE COMARGO , A. 0. , BIGGER, J. N. and NEILSON, D.R . (1979). Transport of inorganic Phosphorus in an Alfisol Soil. Soil Soc. Am. 43, 888-890 . G ILLMAN, G. P . (1979) . A proposed method for the measurement of exchange properties of highly weathered soils. Aust. J. Soil Res., 17, 129-139.

Continued on page 45 WATER March, 1988


REED BED TREATMENT OF WASTEWATERS A EUROPEAN PERSPECTIVE T. H. Davies ABSTRACT The Root Zone Method of wastewater treatment commonly called 'Reed Bed ' treatment, is well established in Europe as a treatment method for small towns and villages and more recently in the United Kingdom. In a recent visit to Denmark, West Germany, Austria and the United Kingdom, 18 installations were visited and observations of their effectiveness, management and of their problems are covered in this paper together with comments upon possible application in Australia .

Thomas H. Davies, M.Sc., ARACJ, Lecturer in Analytical Chemistry and Water Studies at Chisho lm Institute of Technology for 19 years, after 20 years in industry in food chemistry and management. He is a team member of the Water Studies Centre at Chisholm Institute. T. H. Davies

OBJECTIVES As a preliminary to the establishment of an experimental reed bed facility for wastewater treatment in Melbourne, Australia, for the Mornington Peninsula and District Water Board, it was felt advantageous to visit operating reed bed systems in Europe to obtain first hand information upon the design, establishment, management and operation of the beds. The visit was made in June/ July I 987. This paper gives an overview of European reed bed experience and discusses its possible application in Australia.

INTRODUCTION Reed bed technology for the treatment of wastewaters was developed in Germany in the seventies, the main scientific studies · being by Professor R. Kickuth (Kickuth 1984) . The basic design features and concepts of the Kickuth approach are: (i) A bed of porous soil is placed on a waterproof membrane (Figure I) ; (ii) Suitable macrophytes such as Phragmites australis are planted in the soil and an expanded 'root zone' is developed by lowering the water table at the appropriate times to encourage downward root growth and so increase the hyc;!raulic conductivity. (iii) The wastewaters are to flow horizontally underground through the soil in the expanded ' root zone' with retention times of 3 to 5 days . (iv) The plants are not to be harvested but any dead leaves, etc, falling onto the bed surface are allowed to compost. (v) The reed bed area required approximates 3 m 2 per person. (vi) Development of full treatment capacity requires about three years . Further, the hollow root system (rhizomes) are capable of carrying oxygen into the soil (Lawson 1985) so that both aerobic and anaerobic zones are establi shed with a consequent high population of aerobic and anaerobic bacteria capable of breaking down organic matter (Hansen & Anderson 1981). A lso, the dosing of the soil with phosphorus absorbing compounds such as calcium, iron, or aluminium salts makes possible the removal of phosphorus from the waters where the soil does not naturally have this capacity. The expectations advanced by Kickuth include: • rapid and substantial BOD and SS removal; • nitrogen removal by nitrification and dentrification; • phosphorus removal. Kickuth's reports on a large reed bed facility at Othfresen in West Germany substantiated these expectations and showed the effluent quality to be as good as that of the groundwater in the area. 'In fact the claimed effluent sampling points were really sampling gro undwater and not the true effluent from the beds. Investigations by Brix (1987) using tracers to identify the actual effluent showed that substantial treatment was taking place, but not to the extent of Kick uth's claims. Kickuth has been involved in the commercial application of his process in several countries but his continually changing ideas have resulted in some dissatisfaction on the part of some of his clients . Nevertheless, reports and information received were such 32

WATER March, 1988

·60-l00mm stones

Figure 1. Reed bed longitudinal section.

as to warrant consideration of an experimental installation. In view however, of the discrepancies and un certainties encountered, it was felt desirable to make a fact finding visit to Europe to ascertain the true situation as a preliminary to embarking on the project.

THE EUROPEAN SCENE Denmark More than 100 reed bed systems are operating in Denmark, most of them for small vi llages, wit~ many more in the process of construction. There are two or three companies specialising in reed bed construction as a complete package, at least one of these having German connections . The installations visited were small, generally with little evidence of management or maintenance. Surface runoff was very common, an indication of unsatisfactory hydraulic conductivity of the soils used . When the water flow is predominantly on the surface, it is obvious that phosphorus and nitrogen removal, which depends on intimate contact with the soil, will not be achieved. Often Phragmites and Typha were planted together with Typha being placed nearer the inlet because of its claimed ab ility to clump faster and better withstand strong wastewaters. Most beds were heavily infested with weeds which competed strongly with the p lanted reed species . Results of a study of 14 Danish installations (Brix & Schierup 1986) showed wide variation in performance ranging from excellent removal of all three BOD, nitrogen and phosphorus, to poor removal of a ll three, with a wide range in between. For example, BOD removal ranged from 48 to 960Jo, average 7 I OJo ; nitrogen removal 13 to 840Jo , average 420Jo ; phosphorus removal 80Jo to 830Jo, average 380Jo . Phosphorus and nitrogen removal seemed to be the most erratic and unpredictable , indicating the need for more detai led study of the parameters involved in their removal.

West Germany Although the reed bed technology originated in Germany, it is not officially recognised as a standard method of treatment and, as Government funding is available only for approved technology, re latively few beds have been installed, most of them privately. The non-recognition could reflect Kickuth's reluctance to co-operate in any independent monitored trials of the system. Another factor has been the widespread success of lagoon treatment for small village situations which has limited interest in reed bed application .

The one reed bed installation visited in this country was an experiment set up at great expense ·but showing little success in operation. This was possibly due to the use of several types of macrophytes , none of which had thrived under the conditions existing. Surface flow was evident with consequent short circuiting.

Austria There are several reed bed treatment systems in Austria and continuing and detailed research is being carried out at Mannersdorf by the University of Soil Science, Vienna, which was visited.

reed bed projects. An accompanying and significant feature has been a deliberate change in terminology in the United Kingdom where the 'Root Zone Method' descriptor has been replaced by 'Reed Bed Treatment'. As the United Kingdom beds are of quite recent origin, none are fully developed, the oldest being only in its second full growing period. Treatment data are accordingly very limited and the results are rather variable. Substantial BOD and SS removal does occur but phosphorus and nitrogen removal is most erratic. Little improvement is evident in hydraulic conductivity but this may depend on the full development of the rhizome system. At some reed bed sites where the workers were also involved in other treatment methods, there seemed to be a distinct lack of involvement and communication between management and workers regarding reed bed operations. ,This lack was resented and probably prejudiced contribution by the workers to resolving the management and maintenance problems.


Figure 2. Reed beds at Mannersdorf, Austria.

The development of the reeds was excellent (Figure 2) and had virtually eliminated any unwanted species. There was a little surface flow for part of the bed but the treatment obtained is very effective. Results are well documented in the Annual Report on the project (Harbel 1986). Despite the cold winter conditions in this area, the beds continue to perform satisfactorily even when subjected to heavy snow .

United Kingdom In the UK there is great interest in reed bed treatment of wastewater following upon a visit in July 1985 to installations by Prof. Kickuth in West Germany. The visit was made by a group of Water Authority and WRC staff (Boon 1985). The first UK installation was at Acle in October I 985 (Phillips et al. 1987) and now there are more than 40 beds of commercial size and a number of smaller experimental installations. The earlier beds were Kickuth designed, with fairly steep slopes (2-50'/o) and despite the desirable criteria of initial conductivity for the soil of 10-• mi s, heavier soil from sugar beet washing was incorporated at Kickuth's insistance, which substantially lowered the hydraulic conductivity and resulted in major problems as follows: • Surface flow of up to 900'/o of the influent due to the unsuitable hydraulic conductivity. This does not seem to be improving significantly even after two years of rhizome development. • Massive weed growth originating from the large number of seeds in the sugar beet soils, resulting in poor stands of reeds and poor performance, with the steep slopes making control of weeds by periodic flooding impossible. The reaction to these problems has been to change to gravel beds with their high initial hydraulic conductivity, but the high cost of this media may limit further use. Early installations were generally over-engineered with much concrete construction, particularly in the inlet areas, but more recent installations have adopted simple systems having pipes with multiple outlets rather than expensive concrete creations. Water level controls have also been simplified using swivelling outlet pipes or a vertical control pipe which can be built up or lowered by the addition or removal of socketed sections . As the number of Kickuth designed installations increased so did dissatisfaction as changes in design parameters were made in an ad hoc manner resulting in loss of confidence. As a consequence Kickuth has almost no personal involvement in current

This low technology method developed in the colder European environment should have an even greater application in Australia, with its temperate climate and the wider availability of suitable land, where many small towns and communities could use reed beds as a low cost, unobtrusive treatment facility. Capital costs could range from 25 to 750'/o of conventional treatment works depending upon, whether soil or gravel is used and their on-site availability, the need for and quality of a liner, and the sophistication of the inlet and outlet facilities adopted. Operating costs are likely to be 10 to 250'/o of conventional treatment costs. The potential for substantial nitrogen and phosphorus removal in addition to BOD reduction, adds further to the attraction of the method and could solve many potential eutrophication problems caused by inadequately treated effluent discharges.

CONCLUSION • The Reed Bed Treatment of wastewater in Europe is a well established technique with rapidly increasing utilisation, particularly in the United Kingdom. Application is generally for village and small towns with muftiples of individual bed units capable of servicing up to 500 persons. The usual influent is screened or primary settled sewage and BOD reduction of up to 950'/o is not uncommon, despite the frequent very cold conditions. Phosphorus and nitrogen removal can be very high, but the wide variation found indicates a lack of understanding of the processes and of factors necessary to successful removal. The management of the beds seen was frequently poor leadjng to poor performance, exacerbated by the fierce competition of weeds preventing full development of the reeds. Odours were not encountered at any site despite the higher temperatures of the summer season prevailing at the time of the visit. The slow development of the rhizome system is somewhat of a disadvantage (up to three years for full development) but nevertheless, partial treatment can be achieved after the first year, and will increase steadily with time. With a better understanding of the mechanisms of treatment and removal of nutrients in the process the method could have very successful application in Australia.

REFERENCES BOON, A . G . (1985). Report of a visit by members and staff of WRC to Germany (FRG) to investigate the Root Zone method for treatment of wastewaters. WRC Water Research Processes. BRIX, H. ( 1987). The applicability of the wastewater treatment plant in Othfresen as a scientific documentation of the Root Zone Method . Water Science and Technology (in press). BRIX, H. and Schierup, H. H. (1986). Root Zone Systems, operational experience of 14 Danish systems. Report to the Environmental Protection Board. HANSEN, J. I. and Andersen, F. 0. (1981). Effects of Phragmites australis roots and rhizomes on redox potentials, nitrification and bacterial numbers in the sediment. In: 9th Nordic Symposium on Sedimen(s 72-88. Uppsala University.

Continued on page 39 WATER March, 1988


THE ACT APPROACH TO IMPROVING URBAN STORMWATER QUALITY B. C. Phillips, A. I. Lawrence and A. G. Goyen ABSTRACT New urban development policies within the Australian Capital Territory have led to an integrated approach being adopted to the management of urban runoff to protect the quality of local streams, lakes, rivers. The development of a range of water pollution control structures which respond to these policies is described. Water pollution control structures including water pollution control ponds and major and minor gross pollutant traps are described. The economics of the ACT strategy is also briefly discussed. It is concluded that the ACT strategy, which is successfully achieving its water quality objectives , provides one model for the development of urban water quality strategies elsewhere in Australia.

INTRODUCTION The Murrumbidgee River, whose headwaters rise in the Snowy Mountains to the south of the Australian Capital Territory, provides Canberra with many riverside recreational venues and is an important habitat for fish and fauna. It is also the receiving water at greatest risk from urban development in the ACT because all stormwater from Canberra drains into the Murrumbidgee River. Consequently, the protection of the quality of the waters of the Murrumbidgee River and local lakes from polluted stormwater runoff is of vital interest to the Canberra community. Problems which have been encountered in ACT receiving waters include eutrophication, toxic algal blooms and increased faecal coliform levels, accelerated rates of sediment aggradation in lakes and streams and the rapid degradation of stormwater systems. The response to these problems has been the formulation and adoption of a comprehensive water quality management strategy to improve the quality of urban runoff and to protect the quality of local streams, lakes, rivers.

THE ACT WATER QUALITY STRATEGY In the Australian Capital Territory, non-point source pollution of stormwater is primarily associated with the continuing development of Canberra. The response to this problem , namely the formulation of the ACT Water Quality Strategy has been substantially based on the performance_ of Lake Burley Griffin (Cullen, Rosich and Beck, 1978 and Rosich and Cullen, 1981) and on the performance of natural streams and wetlands (Dept. of Territories , 1986). The primary aim of the ACT Water Quality Strategy is to maintain the quality of urban stormwater runoff during and after urbanization (Lawrence, 1986; Goyen, 1987 and Lawrence and Goyen, 1987) to achieve the adopted receiving water quality objectives (Lawrence, 1985). The strategy encompasses the construction and operation of a range of innovative water quality control structures (Goyen, Phillips & Neal, 1987, 1988 and Phillips and Goyen, 1987). These physical measures include the: • Establishment of urban lakes , primarily as biological treatment systems; • Construction of water pollution control ponds and wetlands to act as physical and biological treatment systems; • Construction of major and minor gross pollutant traps on storrpwater channels to intercept trash, debris and coarse sediments; • Construction of temporary 'off-line' sediment retention ponds as a part of land development works to intercept and treat stormwater from development sites before it is discharged into the stormwater system; • Retention of natural creeks augmented by retardation basins in preference to the construction of trunk stormwater pipe systems and concrete-lined drains. 36

WATER March, 1988

B. C. Phillips

A. I. Lawrence

A.G. Goyen

Brett Phillips, BE (Civil), MEngSc, PhD is a senior engineer with Willing and Partners Pty Ltd. He has been directly involved in the design of a range of urban water quality control structures and in the development of gross pollutant traps guidelines. Ian Lawrence, BE (Civil) , DipCE, MA(Admin) is the head of the Water Resources Planning Section of NCDC. He is responsible for the development of urban runoff pollution control strategies in the ACT. Allan Goyen, ME(Civi/) is a director of Willing & Partners Pty Ltd. He has been involved in both the research and design phases of urban runoff quantity and quality projects in Australia. In the ACT a high priority is given to the urban environment. This priority is reflected in the multi-purpose nature of the adopted pollution control facilities . These facilities not only provide environmental protection, they also serve a range of secondary uses including passive and active recreation, landscape features and flood management. Additionally, the promulgation and subsequent enforcement of the AC'_f Water Quality Ordinance a'lid Regulations (ACT 1984a, 1984b) m 1984 has led to the construction of temporary pollution control measures during construction activities including land development. The scope of major water quality control structures currently established or under construction within Canberra are presented in Figure 1.

URBAN LAKES T~e monitored performance of Lake Burley Griffin (Cullen, Ro~1ch and Beck, 1978 and Cullen and Rosich, 1981) , a lake which was constructed in 1963 essentially as a landscape feature governed to a great extent the adoption of urban lakes as a component of the strategy. The establishment of lakes in urban areas, however, is potentially conducive to eutrophication, odour and bacterial pollution problems. The strategy addresses these concerns in five ways (Lawrence and Goyen, 1987 and NCDC, 1987a) . • urban areas are sewered and nutrient removal is undertaken in sewage treatment plants; • where possible lakes are sized to limit loadings to acceptable levels ; • water pollution control ponds are constructed upstream of urban lakes to limit the loading of pollutants on the lakes; • the stratification potential of lakes is minimised by limiting lake depths; • macrophyte areas are protected by zoning lake usage to maximise the quality of the deeper lower reaches. The urban lakes in operation or under construction in the ACT are listed in Table 1.

WATER POLLUTION CONTROL PONDS The primary aim of water pollution control ponds is to intercept polluted stormwater runoff and by a combination of the

Lower Stranger Creek Water Pollution Control Pond The Lower Stranger Creek Water Pollution Control Pond is a recently completed facility which is representative of water quality control ponds being presently constructed in Canberra. The primary purpose of the pond is to intercept polluted stormwater runoff from the lower Stranger Creek catchment thereby protecting the Murrumbidgee River and Pine Island recreation reserve from runoff from urban development in the catchment. The secondary purpose of the pond is to act as a landscape feature and as a venue for passive recreation. This facility is presented in Plate 1.








G.P.T. 2






Plate 1. Aerial View of Lower Stranger Creek Water Pollution Control Pond.

Figure 1. Major ACT water quality control structures.

physical process of sedimentation and biological processes centred on the action of macrophytes to reduce significantly the concentrations of fine sediment, nutrients and E. coli discharged into urban lakes and the Murrumbidgee River. The secondary aim of water quality. control ponds is to form landscape water features suited to passive recreation (Goyen, Phillips and Neal, 1988).


Year of Catchmen t Ultimate Volume Area Commission (ha) Area Percentage (GI) (km') Urbanised

(%) Lake Burley Griffin Lake Ginninderra Lake Tuggeranong

1963 1975 1989

I 865 92 64

3 80 70

33.0 3.7 2.6

640 105 70

The water pollution control ponds in operation or under construction in the ACT are summarised in Table 2.


Year of Catchment Ultimate Volume Area (ha) Comm ission Area Percentage (GI) (km') Urbanised

Hydrological investigations were conducted to determine the spillway dimensions and the crest height of the zoned earthen embankment which impounds the flows in Stqrnger Creek. These investigations also highlighted a secondary benefit; the pond was found to¡ act as an efficient retarding basin which enabled the spillway dimensions to be minimized (Willing & Partners, 1986) . Aesthetic requirements also grea~y influenced the design of the pond spillway. In the case of the Lower Stranger Ck WPCP a step-pool cascading spillway has been constructed (Willing & Partners, 1987). Water quality simulations were undertaken using the POLLUTE model (Phillips, 1987a). This model, which has been previously described by Phillips and Goyen, 1986 and Phillips, 1987 is described briefly below. The water quality simulations predict that when the macrophyte regime becomes established and the land development phase is completed that the Lower Stranger Ck WPCP will markedly improve the quality of the runoff from the 1.2 km 2 urbanized catchment by retaining not less than 96% of Suspended Solids, 81 % of Total Phosphorous and 99% of E. co li during an average year. A variety of pond edge treatments were utilized to create visual diversity . The adopted edge treatments inclu,ded: 'rock jetties' to provide screened vantage points around the pond, mulched planting beds to provide protection to waterbird habitats and stabilized grassing of the pond edges . Once the pollutant loadings from the catchment have stablised selective planting of aquatic vegetation will be undertaken .

(%) Tugg~ranong WPCP Isabella WPCP Stranger Ck WPCP L. Stranger Ck WPCP Point Hut WPCP Ginninderra WPCP Gungaderra WPCP

1978 1987 1987 1987 1987 1989 1989




70 70 60 60 80


12 40 15

0. 11 0 . 11 0.06 0.08 0 .25 0.30 0.20

4 6 3 4 12 24 18

Guidelines on the analysis and sizing of water pollution control ponds has been previously provided by Phillips and Goyen, 1987 . Similarity, details on the biological aspects of water pollution control ponds has been previously provided by Cullen, Lambert and Sanders, I 988. A brief description of the analysis and sizing of a representative water pollution control pond is presented below.

The POLLUTE Model The POLLUTE model was originally developed in the ACT to simulate the pollutant loadings generated from catchments within the ACT and to simulate the performance of water pollution control ponds . More recently it has been used with success outside the ACT to determine the effect of urban development on receiving waters and to simulate the effect of water pollution control measures. The first component of the POLLUTE model simulates daily runoff using the rainfall-runoff model developed by Lawrence and Lansdown, 1977. This rainfall-runoff model comprises urban and rural sub-models which are based on Boughton's model. A garden watering component is also incorporated in the urban submode!. WATER March, 1988


The daily runoff calculated by the rainfall-runoff sub-models is in turn used to calculate the daily export of a range of pollutants using the pollutant export relations reported by Lawrence, 1986. The pollutants examined by Lawrence and their correlation with runoff are presented in Table 3.


Coarse Sediment (kg / km')

Suspended Solids (kg / km')

Total Phosphorus (kg/ km')

Total Nitrogen (kg / km')

E.coli (Count / km')

Urban Rural

1000 R"' 400 R"'

200 R 20 R

0.39 R0 · " 0.115 R0 · "

3.0 R0 · " 3.0 R"' 0

30000 R0 ·' 500 R0 ·'

The second component of the model is a daily reservoir water balance model which has been expanded to calculate the percentage retention of various pollutants on a monthly basis using relations fitted to the retention curves previously reported by Lawrence, 1986 and Phillips and Goyen, 1987. Retention . algorithms for both sedimentation and macrophyte regimes are incorporated in the model.

on the category of the receiving water. The guidelines recommended that the following criteria be adopted for the following categories of receiving waters: • small on-line ponds of volume < 10 ML landscaped creeks and drains, the designated annual retention be 700'/o of grain sizes ~0.08 mm, and • lakes and on-line ponds of volume > 10 ML, the designated annual retention be 700'/o of grain sizes ~0.04 mm. The G.P. TRAP model (Phillips, 1987b) was used to develop the sediment retention curves presented in the guidelines. This model uses the daily rainfall-runoff sub-models and coarse sediment export algorithm implemented in the POLLUTE model to predict the daily coarse sediment export from a catchment. The daily coarse sediment export is in turn partitioned on the basis of hourly recorded rainfall and processed by an algorithm which calculates the deposition of sediments within a gross pollutant trap. The guidelines also take into account the expected number of cleaning operations per year.

Major Gross Pollutant Traps The Stranger Creek Gross Pollutant Trap (designed by Willing & Partners Pty Ltd) is a representative example of major gross pollutant traps in the ACT. This structure is presented in Plate 2.

GROSS POLLUTANT TRAPS In the ACT, gross pollutant traps are installed in situations where there is a need to (NCDC, 1987b and Phillips and Goyen, 1987) : • Protect the aesthetic and environmental quality of small on-line ponds and landscaped drains by limiting the rate of sediment aggradation and intercepting debris and trash to maintain a high visual quality, and • To protect the macrophytes and fauna habitats at the upper end of water pollution control ponds and urban lakes by limiting the rate of coarse sediment aggradation within the ponds and lakes. A major/ minor approach to the interception and trapping of gross pollutants has also evolved in the ACT . The objective of this approach is to optimize the number of structures to be maintained and thereby minimize maintenance costs. Major, open gross pollutant traps are located in major floodways and drains to intercept medium to high stormwater flows from large catchments. Minor, enclosed gross pollutant traps are located within the stormwater drainage system, including at (NCDC , 1987b) : • The head of major floodways, • Locations where stormwater pipes discharge laterally into flood ways, and • On the shores of ponds and lakes where stormwater discharges directly into these water bodies . The major gross pollutant traps currently in operation or under construction in Canberra are summarised in Table 4.


Sullivans Ck G.P.T. Giralang G.P.T. Wanniassa G.P.T. Kambah G.P .T. Stranger G .P.T. Tuggeranong Ck G.P .T. Wanniassa Sth G.P.T.

Year of Catchment I Yr ARI Commission Area Flow (km') (m' l s) 1979 1986 1986 1986 1987 1988 1988

50. 0 12.2 6.0 11.0 3.9 36. 3.7

50.0 16.1 11.0 2 1.0 7.3 43.4 7. 9

Volume (m')

Area (m')

1770 130 ~ 120 ~ 120 270 810 270

2250 250 300 300 430 1320 430

Guidelines on the sizing and design of gross pollutant traps has been previously provided by NCDC, 1987b and Phillips and Goyen, 1987. A brief description of the guidelines is presented below.

Plate 2. Stranger Gross Pollutant Trap.

In accordance with current practice , this trap consists of a single wet basin with a vertical trash rack at the downstream outlet to the basin and a base flow by-pass at the head of the trap. In the ACT, vertical trash racks have been found to be extremely effective in trapping both floating debris and trash and causing fine sediments to deposit in an upstream basin . The trash rack is designed to be overtopped by flows of magnitude greater than a 1 Yr ARI (Average Recurrence Interval) flow . The rack itself is typically constructed from 50 mm wide galvanised rectangular hollow sections at 100 mm centres. Cleaning of major gross pollutant traps consists of frequent (5-10 times per year) surface cleans to remove litter and debris from the trash rack and the complete removal of trapped sediment approximately once a year. During a major cleaning operation base flows are diverted around a major trap. Typically, front end loaders are used to remove material from the trap for transportation to a landfill site. In the past, water within the traps was returned to the stormwater system , however new traps will incorporate facilities to discharge the liquid contents of a trap to the sewer. This practice is expected to increase the overall capture of a range of constituents from the stormwater system . Evolving maintenance equipment capabilities and improved trap design are also gradually overcoming past maintenance problems including the draining, removal and transportation of the sludge-like material typically collected in the traps and traction difficulties on access ramps .

Gross Pollutant Trap Guidelines The ACT gross pollutant trap guidelines provide curves of the average annual retention of sediments in a gross pollutant trap and the expected average annual export of coarse sediment from Canberra catchments. The sizing of gross pollutant traps is based 38

WATER March, 1988

Minor Gross Pollutant Traps Minor gross pollutant traps are typically used to intercept sediment and trash from small urban catchments.

Minor traps are typically 2.0 m wide and range in length from 4.0 m to 12.0 min lengths which are. multiples of the standard ;2.0 m long trash rack panel. The adopted design flow through the vertical trash rack is the 1 Yr ARI flow. The standard trash rack heights are 300 mm, 500 mm and 700 mm and standard pool depths are 600 mm, 1000 mm and 1400 mm . Design guidelines and standard arrangements for minor gross pollutant traps have also been formulated (NCDC , 1987b) . A representative minor gross pollutant trap, designed by Maunsells Pty Ltd with reference to the draft guidelines, is presented in Plate 3.

Plate 3. A typical Minor Gross Pollutant Trap.

Cleaning of minor gross pollutant traps, although still evolving, is primarily by way of a truck equipped with a suction pump which educts water together with sediment and debris. Water which has been filtered through a fine screen is then returned to the stormwater system or diverted to the sewer system. The material removed from the trap is disposed to a landfill site.

ECONOMICS OF THE ACT WATER QUALITY STRATEGY The cost of the ACT Water Quality Strategy can be gauged by the Tuggeranong Valley scheme currently under construction in the south of Canberra. This scheme alone represents a capital costs of $16 million and a further on-going maintenance cost of $0.15 million per year. However, an economic analysis of the strategy indicates that the pollution control strategy is economically viable. In the ACT, separable costs attributable to pollution control are substantially reduced by developing multi-purpose facilities. Benefits include (NCDC, 1987a): • enhancement of land values and the property rating base; and • the protection of substantial recreation benefits associated with the use of Canberra's lakes and rivers. Hence, the economic analysis of the pollution control strategy indicates a Benefit/Cost ratio of 6.

ACKNOWLEDGEMENT The permission of the National Capital Development Commission to publish this paper is gratefully ack nowledged.


No. 65. AUSTRALIAN CAP ITAL TERRITORY (1984b). Water Pollu tion Regula tions.

No. 25. CULLEN, P. , ROSICH, R . and BECK, P. (1978). A phosphorous budget for Lake Bu rley Griffin and management implications for urban lakes. AGPS, Can berra. CULLEN, P ., LAMBERT, 0. and SANDERS, N. (1988). Design a nd management considerations for water pollutio n control ponds. Proceedings of Hydrology and Water Resources Symposium, 1988, Canberra, 1-3 February. DEPARTMENT OF TERRITORIES ( 1986). P oll uta nt removal in an u rban wetland. Canberra. GOYEN, A. G . (1 987). Current trends in A ustralia n stor mwa ter manage ment. Proceedings of Conference on Storm water and Water Quality Management. US EPA, Denver, March. GOYEN, A . G., P h illi ps, B. C. and NEAL, J. F . (1988). Urban stormwater qua li ty - Australian control stru ctures . Proceedings of 4th International Conference on Urban Drainage, Lausanne, August-September. GOYEN, A. G ., PH ILLIPS, B. C. and NEAL, J. F. ( 1987) . Urba n storm water quality control structures in the ACT. Proceedings of Hydrology and Water Resources Symposium, 1988, Canberra, 1-3 Febru ary. LAWRENCE, I. (1985) . Water resources plan ni ng in the ACT . Water planning in Australia. Eds W . R . D. Sewell , J, W . Handmer and 0. I. Smith , CRES, ANU, Canberra. LAWRENCE, I. (1986). Source and fate of urban ru noff constituents a nd their management. Proceedings of 12th Symposium on Stormwater Quality in Urban Areas. Wa ter Research Foundation of A ustralia, University of Wollongo ng, July. LAWRENCE, A. I. and GOYEN, A. G. (1987) . Im proving urba n stormwater q uality - An Australia n strategy, Proceedings of 4th International Conference on Urban Drainage. Lausanne, August-September. LAWRENCE, A. I. and LANSOOWN, P. B. (1977). Development of a rainfallrunoff model for Lake G inni nderra. NCDC Internal Report, Canberra. NATIONAL CAP ITAL DEVELOPMENT COMty!ISS lON . (1987a) . Water Quality - lmproving Canberra's Lakes and Rivers, Canberra. NATIONAL CAPITAL DEVELOPMENT COMMISSION. ( 1987b). Gross pollutant traps. Stormwater Section 11.0 Guidelines on Engineering a nd Environmental Practices - Hyd raulics , Canberra. PHILLIPS, B. C. (1987a) . POLLUTE Usefs G uide, W illing & Partners Pty Ltd, Can berra . PHILLIPS, B. C. (1987b) . G.P. TRAP Use rs G uide. Wi lling & Partners Pty Ltd, Canberra. PHILLIPS , B. C . and GOYEN, A. G. ( 1987). Guidelines fo r Improving Urban Stormwater Q uality Using Water Q uality Control Stru ctures. Proceedings of Seminar on Urban Runoff Water Quality. IEAust/ AWWA . Sydney, 2 1 July. ROSICH , R . and CULLEN, P . ( 198 1). Water quality in Canberra's urba n lakes. Report for NCDC, Canberra. WILLING & PARTNERS PTY LTD. (1 986). Lower Stranger Creek Water Quality Control Pond. Final Design Report. Canberra. W ILLING & PARTNERS PTY LTD. (1987a). Lower Stranger Creek Water Quality Control Pond. Construction Report. Canberra.


CONCLUSIONS The ACT Water Quality Strategy has been formulated in response to the crucial need to protect the quality of the waters of the Murrumbidgee River and urban lakes in Canberra. The aim of the strategy is to maintain the quality of urban stormwater runoff during and after urbanization by intercepting and treating stormwater runoff. A feature of the strategy has been the construction of a range of noteworthy water quality control structures, namely water pollution control ponds and major and minor gross pollutant traps. Several water quality models have also been implemented to aid the analysis and design of these control structures. It is concluded that the ACT strategy, which is successfully achieving its water quality objectives, represents a significant commitment to the maintenance of the quality of ACT lakes and rivers. It also provides one model for the development of urban water quality strategies elsewhere in Australia. .

T. H. DAVIES Continued from page 33 HARBEL , R. (1986). Wet land p lant sewage treatme nt system, Ma nnersdorf/ L. University of Soil Science, Vien na. KICKUTH, R. (1984). Das wurzelra um verfah ren in der Praxis. Landsch. Stadt., 16, 145- 153. LAWSON , G. J. (1985). Cultivating reeds fo r Root Zone Treatment of sewage . Contract report to the Water Research Centre by the Institute of Terrestrial Ecology, 20-30. PH ILLIPS, G. L. , Ayli ng, B. , Clark, C. and Thomas, C. (1987). Sewage treatment using Phragmites, the Root Zone Method , initia l observations at Acle STW. Paper presented to lWPC East Anglican Branch , March 1987.


WATER March, 1988


THE .'BIOCARBONE' PROCESS A New Development in Wastewater Processing K. G. Barr INTRODUCTION The Biocarbone Filter is a new approach in wastewater treatment technology that shows considerable promise for lowering treatment costs and land area requirements. The process is best described as an aerated biological filter. The filter is similar in construction to a rapid gravity sand filter , but uses a coarse porous media which is continu ously aerated. High concentrations of biomass grow in the media, allowing high loading rates to be achieved (hydraulic detention times I to 2 hours). The filter cell is backwashed periodically to remove surplus solids. The process is essentially a combination of several established processes for the treatment of water and wastewater, namely, ac. tivated sludge, fixed film biological systems and solids separation by gravity filtration . Primary settling normally precedes the Biocarbone filter, however clarification of the effluent is achieved within the filter and fina l settling tanks are not required. Sludge digesters would normally be required . The process was developed in Europe by the French Company, OTV and several full scale plants are operating successfull y in France. The process is licensed for use in Australia through Aquatec-Maxcon . Following an interest in the process by the Brisbane City Council, an agreement was made between OTV and Aquatec-Maxcon to allow the Council to undertake pilot studies to establish the design and performance parameters for local conditions.

Keith Barr, B. Eng(Honours) Civil, Grad.Dip.Environ.Eng., QIT, is an operations engineer in the Department of Water Supply and Sewerage of the Brisbane City Council. He has been involved with the investigation and operation of wastewater treatment plants for many years. He is curren tly with the Water Operations Section which controls the distribution of water for Brisbane City.

K. G. Barr

PROCESS DESCRIPTION A filter bed depth of 2 to 2.5 mis usually adopted in full scale plants. The filter media is an expanded schist or shale with an average grain size of 4 mm . It is a porous material with a high internal void area which provides an ideal environment for the adsorption and growth of biomass . The filter bed is always submerged as the influent percolates downwards in the opposite direction to the process air supply (see Figure I). The oxygen required for biological growth is supplied via a diffuser system located at 300 mm above the filter floor. A particularly high oxygen transfer rate is achieved by the passage, impact and breaking-up of air bubbles on the media grains in conjunction with the counter directional flow of the sewage.



goooooooo..,__ _ _ _ __ -<

BCC Pilot Plant the fixed biomass. Most of the influent suspended solids and biologically produced solids are trapped with in the aerated portion of the media bed . The layer of the filter media beneath the aeration level has a polishing effect which ensures a satisfactory clarification of the effluent which then exits from a nozzle underdrain system. The filter clogs gradually due to the growth of biomass and the penetration of suspended solids. Regular backwashing is required at intervals va rying with the applied loads. The filter wash run consists of a reverse flow and air scour sequence similar to the methods used for washing sand filters. An initial air scour loosens the materials, a second air + wash-water stage draws the loosened particles to the surface of the bed and a third stage, using water only discharges the waste sludge by siphoning. A sufficient quantity of biomass continues to adhere to the media grains to ensure that the purifying action in the biological zone is instantly resumed after a washing run.


PILOT TESTING The pilot plant used by the Brisbane City Council employed the same media , filter bed depth, air injection and underdrain details as would be used in a full scale plant. The capacity of the pilot plant was equivalent to 1000 ep. for BOD 5 removal only with one hour detention and 500 ep . for nitrification with two hours detention. The process has proven to be very reliable in removing the traditional pollutants of BOD. and NFR under all conditions and loading rates. Overall process performance is presented in Table 1.







0 0 0

0 0


0000000000000 0 0 WASHING WATER EFFLUENT


Figure 1. Flow diagram for the Biocarbone Process. The aeration process loosens the media bed continuously allowing the induction of the suspended solids contained in the infl uent. At the same time, the filter media serves as a support for 42

WATER March , /988


BOD, mg/ L NFR mg/ L TKN mg/ L

Se//led Sewage

150 100 40

Effluent I hour detention

2 hour detention (nitrification)

20 20 30

10 15 10

Process air requ irements for BOD, removal and nitrification are marginally less than the air req¡uirements for conventional activated sludge plants. The biomass produced in the Brisbane pilot plant had a different species distribution to typical biomass produced in E urope. The biomass was more strongly attac hed to the media stones and considerable difficulties were experienced initially in removing the attached biomass to achieve a satisfactory holding capacity and filter run time. A more vigoro us backwashing system and the periodic pulsing of the media bed with backwash air were employed to achieve filter run times of greater than 24 hours. The backwash sludge produced was less stabilized than a typical humus sludge but more stablized than activated slud ge from conventionally loaded plants. Settleability was ge nerally good with SSV < 90. Capilla ry suction time tests indicate th at Biocarbone sludge and activated sludge (from conventional plants) have similar dewatering properties.

DESIGN ASPECTS The Brisbane pilot study has established the local design parameters for the Biocarbone Process to produce a 20:20 effl uent without nitrification with one hour detention . In the nitrification mode with two hours detention the process was able to remove 75% of Total Kjeldah l Nitrogen in producing an effl uent of 10 mg/L TKN. The Brisbane pilot plant was not designed for denitrification and up to 15 mg/ L of oxidized nitrogen was present in the effluent. Biocarbone sludge produced in the nitrification mode wi ll denitrify rapidly under normal conditions. Consequently, the return of Biocarbone sludge direct to primary tanks is not recommended and the use of a separate gravity thickener in batch operation would be preferred . Surface skimming equipment would also be required to remove any float sludge . The French company, which holds the patent for the process,

L. A. NAGY and D. C. BUTTERS Co ntinued from page 17

GANF, G. G. ( 1980). Factors Contro lli ng the Growth of Ph ytoplankton in Mount Bold Rese rvoi r, So uth Australia. Technical Pape r No. 48, Australian Water Resources Counci l, Canberra. GANF, G. G., OLJVER, R. L. a nd STON E, S. J . L. ( 1982). Phytoplankton growt h . In ' Prediction in Water Qualit y' (O ' Loughlin , E. M. and Cullen, P. Editors), Australia n Academy of Scie nce , Canberra. GOL TERMAN, H . L. ( 1975). 'Physiological Limnology . An Approach to the Ph ysio logy o f Lake Ecosystems ' . Elsevier , A msterdam . GUTTER IDGE, HASKINS a nd DA VEY PTY . LTD. ( 1982). Review of Water Qualit y Issues fo r Lake Burley Griffi n . Report to the National Ca pit al Development Commission. GU TT E RIDGE , HAS KINS and DAVEY PTY . LTD . (1983). Lake Burley Griffin - A Body of Still Water. Report to the Nat ional Capital Development Com-

mission. HA RT , B. T., BECKETT , R., SINCLA IR , P. , WESTON, B., SMALLS, I. C. and S HAW , S. (1 983) . The nature and bio logical avai lab ility of particulate associated phosphorus . Proc. Fed. Conv. Australian Water and Wastewater Assoc. 10(34), 1- 18. HUMP HRI ES, S. E . and IMBERGER, J . ( 1982). The Innue nce and the Int ernal Structure and Dyna mics of Burrinjuck Reservoir on Phytoplankton Blooms . Centre for Water Research, Uni versity of Western Austral ia. KA LFF, J . and KNOECHEL, R. (I 978). Ph ytoplankto n and their dynamics in oligotrophic and eutrophic lakes. A nn. Rev. Ecol. Syst. 9, 475-495. National Cap ita l Development Com mission ( 198 1a) Mu rru mbidgee River Ecology Study. Techn ical Paper 33, Nat ional Capital Development Comm ission, Canberra.

has deveioped a dual cell aerobic/ anoxic Biocarbone filter which is suitable for den itrification duty . T he Water Board of Sydney is currently conducting pilot studies into nitrification/ denitrification using a Biocarbone filter .

CONCLUSION The principal advantage of the process is the performance of the entire secondary phase of treatment in one reac tor, occupying as little as 20% of the land required for conventional secondary treatment processes. Since the solids separation is achieved by physical clogging of the media bed rather than gravity sedimentation, another significant process advantage is that the settling characteristics of the biomass become leS;S important and do not effect effluent quality . The importa nce of pilot testing for novel processes to determine design parameters for local conditions cannot be over emphasized . Different climatic conditions and sewage characteristics res ulted in a species distrib ution of the biomass differing from the typical biomass produced in Europe. A modified backwashing system was developed to remove the more resilient biomass and achieve satisfactory fi lter 'run times in the test plant. The Brisbane pilot study has confirmed the potential of this compact and reliable secondary treatment process. Further research aimed at improving nitrification and denitrification are required to fina lize the design parameters for the process suitable for Australian application .

ACKNOWLEDGEMENTS Appreciation is extended to Mr Alan Gi nn, Manager of the Department of Water Supply and Sewerage who initiated the project and to the officers of the Sewerage Operations and Scientific Services Branches who took part in the project. •

NATI ONAL CAP ITAL DEVELO PMENT COMM ISS ION (1981b) . Wa ters of the Can berra Region , Metro po litan P lann ing Issues. Technical Paper 30, National Capital Development Commissio n , Canberra. PH ILP, 0 . M. (1982) . Waste disposal to Au1!ralian inland waters. In ' Prediction in Water Qualit y' (O'Loughlin, E. M. a nd Cu llen , P . Editors), Australian Academy of Science, Canberra. PHILP, 0. M. (1985). Phosphorus removal at the Lower Molonglo Water Qualit y Control Centre. Journal Water Polllllion Con trol Federation 57, 84 1-846. REY NO LDS, C . S. and WALSBY , A. E. (1975). Water-blooms. Biological Reviews 50, 437-481. ROS IC H, R. S. a nd CU LLEN, P. (1981). Wate r Quality in Canberra's Urban Lakes. Repo rt to the National Capital Development Commiss ion . SCHINDLE R, 0. W. (1974) . E utrophicat ion a nd recovery in experiment al lakes : implications for lake managements . Science 184, 897-899. SCH INDLER , D. W . (1975). Who le- lake eutrophication experiment s with phosphoru s, nitrogen a nd carbo n. Verh . Internal. Verein . Limnol. 19, 3221-3231. THE OFFICE OF ACT ADM INISTRATION (1987). Wate r Qualit y Mon itoring Lake Burley G riffin a nd Lake Ginninderra , December 1981 to June 1986. Report to the National Capit al Development Commissio n. VOLLENWEIDER, R. A. ( 1969). Possibilities and limits of elementary models concernin g the budget of substances in la kes. Archs Hydrobiot. 66, 1-36. VOLLENW E ID E R, R. A. (1975). Input-output mo dels with special references to the phosphoru s loading concept in limnology. Schweiz. Z. Hydro/. 37( 1), 53-84. VOLLENWEIDER , R. A. (1976) . Advances in defining critical loading levels for phosphorus in lake eutrop hication Mem. 1st. Ital. ldrobiol. 33, 53-83. WJLLJAMS , W. 0 . (I 980). A n Ecological Basis for Water Resource Management. Au stralian National Uni versit y Press, Canberra. WOOD , G. (1975) . An Assess ment of E utrophicat ion in Australian Inland Waters . Technical P aper No. 15, Australian Water Resources Co uncil, Canberra .


WATER March, 1988


(______ C_O_~_F._f:R_E._N_C_E_s_ · •_S_Y._M_P_O_S_I_:Jl_•_C_O_U_R_S_'E_S_ · __ INTERNATIONAL SYMPOSIUM


June 12-15, '90

CALL FOR PAPERS Sponsors of the Symposium in clude: UNESCO , UNEP, UNIDO , Mani toba Hydro, Env ironment Canada, CSCE, IAH S, IWRA, IAHR , HSERC and NSF. The Symposium will deal with the application of systems analysis in water resources management , design , planning and policy. Original papers are invited for presentation. Abstracts to be submit ted by August 31 '88, fu ll papers by June 30 '89. Information: S. P. Simonovic, Civil Engineering Dept., University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2.

Australian Water Research Advisory Council




Proposals must fall within the following programs

NATIONAL PRIORITIES Project funding for a maximum of three years for research within water research priorities identified by AWRAC .

PARTNERSHIP RESEARCH To assist short term research , fund ing wi ll be provided to projects involving a third party .

SCIENTIFIC MERIT Limited funding will be prov ided for high quality research of fundamental nature with signif icant potential for breakthrough in theoretical understand ing .




Over the last two years a data base of all principal long -term effects of phosphorus control measures has been collected and analysed . Fina l results will constitute the main subject at the symposium. Information: Institute for Environment and Systems Analysis , Lake Management Foundation, Emmastraat 16, 1075 HT Amsterdam. The Netherlands.

RESEARCH FELLOWSHIPS For persons holding a Ph.D or equivalent, with less than five years doctoral experience . Preference to Australian citi zens or residents . Salary $28 456 ris ing to $30 894 pa. Host institutions to submit app li cations .

APPLICATION FORMS AND PROPOSALS (5 COPIES) MUST BE RECEIVED BY APRIL 27, 1988. Applications and information: The Secretary, AWRC Advisory Council , GPO Box 858, Canberra, ACT 2601 . *

M.KAY N ALP , G. HO , K. MATH EW and P. NEWMA N Continued from page 29 GLENISTER, D. J. and THORNBER, M. R. (1 985) . Th e Alka linity of Red Mud and its ap plication fo r the Manage ment of Acid Wastes. P roceedings of Chemica 85, Institution of Chemical Engineers , Perth , 25-28 August, p. 109-114. HART , B. T . (1 982) . Australian water quali ty criteria for heavy metals. Australian Water Resources Counci l, Canberra, A. G .P .S. HO , G. E., MATH EW, K. and NEWM AN, P . W . G. (1 983) . Water Quality Improvement of Treated Wastewater by Soil Percolation . Proceedings of Water Quality Seminar, Water Research Foundation o f Australia, Perth, 13- 14th Octo ber, p . 124- 129. HO , G . E . , NEWM AN, P . W. G ., MATH EW K. a nd P ARKE R (1 984). Red Mud Research Report to ALCOA a nd Western Australian Water Authorit y. Environmenta l Science , Murdoch University. HO , G. E., NEWMAN, P . W. G. , MATH EW, K. and DE POTTER, (1985). Neutra lisation of Bauxite Processing Residues with Copperas. Proceedings of Chemeca 85, Institution of Chemical Engineers, Perth , 25-28 August, p. 103- 108. HO , G. E . , MATH EW, K. and NEWMAN, P. W. G. (1986j. Laboratory Recharge Column Experiment. Environmental Science, Murd och Uni versity.



Sydney, June 1-3, '88 Emergence of Pacific Regi, Econom ic Linkages and Technolo Urban i:ievelopment, Metropolit, Strategies Central City Strategie Transport , Communication Housin Emp loym ent & Industry , Urbc Design , Heritage & Conservation. The NSW Department of Enviro, ment and Planning invites delegate to this Bicentenn ial event. Information: Conference Seen tariat 1988, International Plannin Conference, PO Box 468, Paddingtor NSW 2021. *


QIT Faculty of Information Technology

THE UNIX SYSTEM THREE SHORT COURSES The cour:,es will be held on the Queensland Institute of Technology Campus in June, July, and November 1988. copRSE 1 - An Introduction to UNIX. Aim is in troduction to UNIX as an operating system . No previous computer operating system is assumed. COURSE 2 - Shell programming. In troduces the Bourne Shell and the faci liti es avai lable to write programs which manipulate Shell variables and use common filters . COURSE 3 - Advanced UNIX. Provides further experience with She ll programm in g with emphasis on problem so lvin g and training of system managers. Information: Course Secretariat, Q SEARCH, QUIT, GPO Box 2434, Brisbane 4001. Ph. (07) 223 2196.

HO , G . E., MATHEW, K. and NEWM AN , P . W . G . (1987). Leachate Quali ty fro m Gy psum Neutralised Red Mud Applied to Sand y Soils. Submitted for publicati on to Water, Air and Soil Pollution . HODKIN , E. P. (ed) (1984). Potential for Ma nagement of Paul Harvey Estuary. Bulletin No . 160, Department of Conservation and Enviro nment, W.A . MATH EW, K. , NEWMA N, P . W . G . and HO, G. E. (1982). Groundwater recharge with secondary sewage efflu ent. Australian Water Resources Council , Technical paper no. 71. McARTH UR, W. M. , and BETTENAY , E. (1 974) . The development and distribution of soils of t he Swan Coastal Plain . W. A. Soil Publication No. 6, CS IRO , Melbourne. MURPH Y, J . and RILEY, J . P . (1962) . A modified single solution method for the Determin ation o f P hosphorus in Nat ural Waters. A nalytical Chemica, Acta, vol. 27, pp. 31-3 6. Standard Methods (1975). APHA-AWWA-WPCF. 13th Ed. STUANES , A. 0 . and ENFIELD , C . G. (1 984). Prediction of Phosphate Movement th ro ugh Selected Soils. J. Environ. Qua/. Vol. 13, No. 22 . YEATES , J . S., DEE LEY, D. M. and ALLEN, D. (1984) . Fertilizer Management agro nomic as pects. In Hodgkin , E. P. (ed .). P otential for management of the Peel-Harvey Estuary. Bulletin No . 160, pp . 43-57, Department of Conservation and Enviro nment , W .A . • WAT ER M nrr h

I OllD