Pertanika 3(2), 133-141 (1980)

Application of Ponding Systems in the Treatment ofPalm Oil Mill and Rubber Mill Effluents

K. K. WONGDepartment of Environmental Sciences, Faculty of Science and Environmental Studies,

Universiti Pertanian Malaysia, Serdang, Selangor.

Key words: Ponding; Lagoons; Palm oil mill effluent; Rubber mill effluent; rational design models;completely mixed flow models; plug flow models; kinetic coefficients; costs data; economicsof ponding systems.

RINGKASAN

Kolam telah dicadangkan sebagai satu cara yang murah untuk memproseskan efluen-efluen dari kilangkelapa sawit dan kilang getah. Pertimbangan-pertimbangan tioretikal telah diulas dan dibincangkan ber-sangkutan dengan rekabentuk kolam. Rekabentuk untuk model-model empirikal dan rasional telahpun di-bentangkan. Model-model tersebut dianggap sebagai pemilihan yang paling tepat untuk rekabentuk kolam.Kegunaan-kegunaan sistem-sistem kolam untuk efluen-efluen kelapa sawit, getah dan betongan tempatan diMalaysia telah dibincangkan. Angkali kinetik tingkat pertama untuk efluen-efluen kilang getah telah di-perolehi berdasarkan pada data-data sistem kolam yang telah wujud. Angkatap-angkatap ini bernilai dari0.065 hingga 0.27 hari~x. Angkali kinetik untuk efluen kilang kelapa sawit bagi sistem campuran sebati dansistem aliran "plug" juga dibentangkan. Angkali untuk kolam yang menunjukkan aliran "plug" telah di-dapati bernilai 0.068 hari~*. Kekurangan data telah menghalang usaha terbitan angkatap kinetik bagikolam-kolam oksidasyen betongan tempatan. Data kos untuk perbandingan diantara sistem kolam dan sistem-sistem lain telah diperolehi dari sistem-sistem memproseskan efluen secara anaerobik di kilang kelapa sawityang telah wujud.

SUMMARY

Ponding is proposed as a viable low-cost treatment method for palm oil and rubber mill effluents. Theore-tical considerations were reviewed and discussed in relation to the design of ponds. Both empirical and rationaldesign models were presented. Adoption of such models were deemed closest to a way of designing ponds.Applications of ponding systems for palm oil mill effluent, rubber effluent and domestic sewage in Malaysia werediscussed. Some first order kinetic coefficients were calculated for rubber mill effluents based on data for existingponding systems. These constants ranged from 0,065 to 0.27 day1. Kinetic coefficients for palm oil milleffluent for both completely mixed and plug flow systems were also presented. The coefficient for a pond exhibitingplug flow was found to be 0.068 day~x. Insufficient data prevented the derivation of kinetic constants fordomestic sewage oxidation ponds. Costs data were derived from existing palm oil mill effluent anaerobic treat-ment systems for comparison between ponding and other systems.

INTRODUCTION

Ponding is a general term which includeswaste stabilization lagoons (ponds) and oxidationponds. The term oxidation pond has also beenloosely used and can mean aerobic, facultative,maturation or sometimes it may even be usedfor anaerobic ponds. Ponding essentially employsa biological method of treatment for wastewaters.It is also used for settling of sludge or suspendedsolids (sedimentation ponds).

Ponding can be used where land space isavailable. It can achieve a reasonable degree of

treatment, is low in construction and operatingcosts and is easily maintained as the technologyrequired is relatively unsophisticated. Pondshave been used extensively in several othercountries for the treatment of industrial waste-waters amenable to biological treatment. Pond-ing has been used in oil refineries, slaughter-houses, dairies, piggeries, poultry-processingplants and even chemical works. Wherevernecessary, nutrients are added to help raise theBOD: N:P ratio to 100:5:1 in order for pondsto function efficiently in industrial installations.

133

K. K. WONG

Ponding has been applied to a considerableextent in Malaysia. Palm oil mill effluent, rubberfactory effluent and domestic sewage effluenthave all been treated in this manner (Wong andSpringer 1980; John and Ong 1979; Asairinachan1979).

This paper deals with the theoretical con-siderations for ponding systems, design criteria,local applications and cost estimates.

THEORETICAL CONSIDERATIONS

Biological treatment processes can basicallybe divided into:

(a) aerobic

(b) facultative

(c) anaerobic

Each one of the above processes involvesbiological cells which require a chemical substrateand some optimal physical conditions for theirgrowth and survival before they ultimately decay.

Scientists have attempted to rationalise thegrowth of bacterial cells (including those involvedin the treatment of wastewaters), into simplifiedgrowth models with accompanying involvementof growth kinetics (Monod 1949; McCarty 1966;Pearson 1968; Thirumurthi 1974; Marais 1970,1974). These will be discussed in this section.

Design EquationsPond designs can be derived from equations

which are in turn arrived at from either:(a) empirical data from actual field experi-

enceor (b) a rational design model employing an

understanding of biological kinetics.

BOD is usually used as the pollution para-meter to describe the performance and designof a pond.

Areal or surface BOD loading ks is theweight of BOD applied per unit area of pondper day and is given as:

l O L i Q

0)

Volumetric loading rates are sometimes usedparticularly with anaerobic ponds. If \ v is theweight of BOD applied per unit volume per daythen

Xv = (2)AD

where D = pond depth

ADand = retention time = t

Q

LiThen Xv = —

In facultative ponds, Av is usually in therange of 15-30 g/m3 day whilst anaerobic pondsdevelop odour at \ v > 100 g/m3-day (Mara1976).

Empirical models for industrial wastewatertreatment using ponds are still seriously lacking.Most empirical models have been based upondomestic sewage and are only for facultativeponds. Industrial waste ponds represent themost complex of the ponding systems ofteninvolving aerobic, anaerobic and settling pro-cesses. The discussion of empirical modelsbelow is only limited to domestic sewage.

McGarry and Pescod (1970) found thatareal BODS removal, L r, may be estimatedthrough knowledge of areal BOD5 loading, L o by

L r = 9.23 +0.725 L o (3)

Larsen (1974) proposed an empirical designmodel for pond surface area as follows:

MOT = (2.468RED + 2.468TTC +23.9/TEMPR +150.0/DRY) . 106 (4)

where the dimensionless products:

where A = Area of pond in hectaresQ = Flow rate of wastewater in

m3/dayand Li = Influent BOD in kg/m3

surface area (solar radiation)1/3

MOT =influent flow rate

(influent BOD5)1/3

RED =influent BOD5 - effluent BOD5

influent BOD5

134

PONDING SYSTEMS FOR PALM OIL AND RUBBER MILL EFFLUENTS

wind speed (influent BOD5)1'3

TCC =(solar radiation)1 '3

lagoon liquid temperatureTEMPR =

air temperature

and,DRY — relative humidity

Larsen thus related solar radiation, windspeed and temperature to the pond design.

Gloyna (1976) proposed the followingequation for a facultative pond design:-

V = 3.5 x 10-5 QL a (a<3 5-T)ff . . -(5)

where,V =Q =L a =

a =TV =f =

pond volume, m3

influent flow rate, 1/dultimate influent BODU or COD,mg/1temperature coefficientpond temperature, °Calgal toxicity factor, andsulphide oxygen demand

All these empirical equations for wastestabilization pond design appear to be peculiarand particular to the cases that they are derivedfrom and the design of ponds in Malaysia neednot necessarily follow these equations.

Rational design equations may be of someassistance to pond designers. Marais (1970)proposed first-order kinetic equations fordescribing completely mixed ponds as follows:

Le 1

L, 1 + k, t(6)

whereLe = BOD in effluent (mg/1)Li == BOD in influent (mg/1)k! — first order kinetic constantt = retention time

This model was adopted by Wong andSpringer (1980) in their evaluation of anaerobicponding systems for palm oil mill effluent(POME). The model could justify the designof functioning and completely mixed pondingsystems for POME. This equation has alsobeen applied on aerated lagoons.

To take into consideration temperatureMarais suggested that

kT - kTo a<T"2<» (7)where,

kr = kinetic constant at temperatureT

kT0 — kinetic constant at temperatureT o

anda = temperature coefficienta was reported to be 1.047 by Bishop

and Grenney (1976) for domestic sewage.

In a plug flow ponding system (more like along ditch) Mara (1976) reported the followingdesign equation:

Le = Li (8)

Limitations of first-order kinetics have beenconsidered by studies on composite and retardedexponential models (Gameson et al. 1958) andsecond order kinetics (Young and Clarke 1965;Tucek and Chudoba 1968). Wehner andWilhelnVs (1958) model for first-order bioxida-tion reaction in dispersed flow is complex forsimple pond designs, but nevertheless can beused. The justification for using a more complexmodel was supported by Thirumurthi (1974)who refuted the use of a completely mixed flowmodel in the rational design of stabilizationponds. He found that facultative ponds exhibitnonideal flow patterns and recommended the useof Wehner-Wilhelm's chemical reactor equationfor pond design:

where,

Le

Li

Le

Lik

td

and dwhere,

DuL

+ a)2 _ ( l _ a ) 2 e -a/2d... (9)

effluent BODS, mg/linfluent BOD5, mg/11st order BOD5 removal coeffi-cient, day"1

1 + ktdmean detention timedimensionless dispersion numberD/uL

axial-dispersion coefficient, m2/hfluid velocity, m/hcharacteristic length, m

Thirumurthi's definition of k is

k - KSC1C2/C3

135

K. K. WONG

where,Ks

C,c2

and, C3

= standard BOD5 removal coefficient= correction factor for temperature= correction factor for organic load= correction factor for toxic

chemicals

More recently Finney and Middlebrooks(1980) disputed the applicability of both empiri-cal and rational design models. They were notsatisfied because the empirical and rationalmodels could not predict their published pondperformance data. They believed that thehydraulic retention time used in many of thedesign equations had very little research doneon them.

DESIGN CRITERIA AND PERFORMANCEOF PONDS

As both the empirical and rational equationscan be subjected to questioning for particulartypes of pond, with a particular climate and aparticular effluent, it would be more reasonableto base designs on a range of performance.Metcalf and Eddy (1979) provided informationfor design of all the different types of ponds witha range of design criteria.

A number of different approaches havebeen used to design aerobic ponds (Gloyna 1971;Marais and Capri 1970; McKinney 1971).Generally first-order kinetics can be applied andthe kinetic constants can be determined throughactual operating data - something hard to comeby for industrial wastewaters.

Depending on the type of flow regime thedesign of both facultative and anaerobic pondsmay also employ first order kinetic equations.The design can be more accurate if the dispersionfactor, d, is known in the Wehner-Wilhelmequation. The value of k can be calculated fromFigure 1 (after Thirumurthi 1969). For facul-tative ponds with aerators, the aerators shouldhave a capacity adequate to satisfy from 175 to225% of the incoming BOD5. Such ponds arein existence in Malaysia but are not consideredhere.

Accumulation of sludge can be a seriousproblem for wastewaters with high solids contentfor this would reduce the performance of ponds.Very often operators think that, because pondingrequires minimal maintenance, they can thereforebe left on their own. This has actually happenedfor a lot of anaerobic and facultative pondstreating palm oil mill effluent. They are eithersour, overloaded or all clogged up because ofpoor mixing. Ponds do need maintenance andoccasional desludging.

SOME PONDING SYSTEMSIN MALAYSIA

Ponding has been used considerably inMalaysia in the treatment of the two main agro-industrial effluents: rubber and palm oil mil]effluent.

There are several types of rubber effluents:block rubber, sheet rubber, crepe rubber andlatex concentrate effluent. Ponding systems havebeen investigated by several workers (Muthu-rajah et al. 1973; John et al. 1974; Ponniah et al.1975; Ponniah et al. 1976; Ahmad et al. 1979;John and Ong 1979; John 1979).

Ponding systems in palm oil mill effluenthave also been reported (Seow 1977; Wong andSpringer 1980). Anaerobic ponds are usuallythe first stage in the treatment of POME.

In domestic sewage wastewater treatment,there is ponding, activated sludge, extendedaeration and diffused air in use in Malaysia(Asairinachan 1979). The efficiency of BODremoval in the oxidation ponds ranged between74 to 81%. As influent BOD was around 190 mg/1an effluent. BOD of less than 50 mg/1 couldbe achieved. For a 2-stage oxidation pond systema 450 kg/ha/day BOD loading rate with anexpected BOD discharge of less than 50 mg/1can be realised.

CALCULATIONS OF DESIGN CRITERIAFOR RUBBER AND PALM OIL MILL

EFFLUENT PONDING SYSTEMS

The models presented in section 2 providea basis for the choices of rational models usedin the calculations of design criteria for rubberand palm oil mill effluent ponding systems.

The empirical models outlined are modelsderived for specific wastewaters treated underoperating conditions which are applicable onlyto those particular climate and pond configura-tions. As these models were arrived at in coun-tries which are climatically very different fromMalaysia and as the wastewater considered ismainly domestic sewage, the empirical modelspresented are not acceptable.

Rational design models were thus selectedbased not only on the process of elimination butalso upon the immediate need for guidelines todesign ponds of the right configuration andtreatment capability to process rubber and palmoil mill effluents.

136

PONDING SYSTEMS FOR PALM OIL AND RUBBER MTLL EFFLUENTS

As discussed in the Section on "DesignCriteria and Performance of Ponds" the flowregime is important for selecting the type ofrational model used for designing ponds. Twolimiting cases are available, i.e. completely mixedflow and plug flow. Most ponds operate in aflow regime somewhere in between, dependingupon the pond configuration, and, operating andmaintenance procedures.

The model which accounts for the flowregime is the Wehner-Wilhelm model. As thedispersion factor, d, varies from 0.1 to 2.0 forwaste stabilization ponds depending upon theconfiguration, certain assumptions and choiceshave to be made. Based upon the experience ofMetcalf and Eddy (1979) on ponds of differentconfiguration, d was taken to be 0.25 for theblock rubber, sheet rubber and crepe rubbereffluent as the ponds used in each respectivecase were actually two ponds in series. Thelatex concentrate effluent h?.d four ponds inseries and thus a value of d = 0.1 was used.This closely approximates plug flow. Figure 1was used to derive the kinetic coefficients forthese rubber effluents.

6 8 JO 20

PERCENT REMAINING, S /S

Figure 1: Values ofkt in the Wehner and WilhehnEquation versus percent remaining forvarious dispersion factors.

The data on palm oil mill effluent pondingsystems fitted the two limiting systems: a com-pletely mixed flow case (Wong and Springer,1980) and a completely plug flow case. A plugflow was assumed here because the pond usedwas actually a 1609.8 m long ditch with a width

of about 3 to 4.6 metres and a depth of 1.3 to2 metres.

Thus equation (6) was used for derivingthe kinetic coefficients by Wong and Springer(1980) and equation (8) was used to derive thecoefficient of the plug flow ponding system forPOME considered here.

a. First Order Kinetic Coefficients and Perform-ance of Ponds ConsideredUs ing data on rubber effluent ponding

systems, the values of the first order kineticcoefficients derived on the basis discussed abovefor block rubber, sheet rubber, crepe rubberand latex concentrate were found to be 0.27,0.23, 0.16 and 0.065 day"1 respectively. Thesevalues are, therefore, very much a function ofthe pond's shape or configuration. The designermust thus be cognizant of the type of pond he isdesigning before he uses these values.

The first order kinetic coefficient for thecompletely plug flow model calculated fromdata of the long ditch ponding system in thetreatment of POME was found to be 0.068 day"1.This differed quite significantly from the valueof 0.36 day"1 in the case of the completely mixedflow ponding system. It, however, comparesvery closely with the almost plug flow con-figuration for the latex concentrate pond.

A summary of the wastewater characteristics,performance of ponding and the calculatedvalues of the first order kinetic coefficients andretention times are shown in Table 1.

In all these calculations BOD was used asthe basis with the exception of the plug flowmodel for POME where COD was used. Figure2 plots the expected value of the effluent toinfluent BOD (or COD) ratio at different treat-ment retention times for the various ponds dis-cussed. The curves were plotted by applyingthe first-order kinetic coefficients derived on thevarious rational models using Figure 1 andequations (6) and (8).

b. CostsThe cost of operating ponds depends a

great deal on the price of land at a specific time*If land has to be purchased to construct pondsfor waste treatment the capital costs can be high.Land prices vary from M$10,000/ha in the ruralareas to $1 million/ha in urban Kuala Lumpur(1979 figures).

As ponding construction costs are similarfor all treatment systems, reliable data from

137

oo

Summary of Wastewater Characteristics,

Parameter

Wastewater

Block Rubber6

Sheet Rubber6

Crepe Rubber6

Latex Concentrate0

Palm Oil Milld

POMEe

Domestic Effluentf

BOD

RawEffluent

1769

1322

305

3524

25000

-

176

(mg/1)

removal

96.7

95.5

89.5

96.0

94.0

_

77.0

COD

RawEffluent

2899

2477

846

4849

12549

-

(mg/1)

removal

92.1

85.8

73.3

89.0

86.7

-

TABLE 1

Ponding Performance and First-order kinetic coefficients for Rubber, Palm Oil Milland Domestic Sewerage Effluents

Total Solids(mg/1)

RawEffluent

1961

1976

546

_

13046

-

o/0

removal

63.2

69.6

18.5

-

64.7

-

Suspended Solids(mg/1)

Raw %Effluent removal

322 61.2

7

818 56.1

5508 91.3

176 50.8

Total nitrogen(mg/1)

RawEffluent

141

143

75

602

252

-

f'O

removal

61

58.0

30.7

66.4

61.9

-

Amm. nitrogen(mg/1)

Raw %Effluent removal

68 38.2

73

6.4

466 71.2

12.3 55.1

-

Firstorder

Coefficient(k)a (day"1)

0.27

0.23

0.16

0.065

0.36

0.068

-

RetentionTime

(days)

22

22

20 >

64

50

31

-

*

Q

zo

a. The first-order kinetic coefficient (k) was calculated from Figure 1. Values of the dispersion factor (d) range from 0.1 to 2.0 for waste stabilization ponds.b. These were ponds in series. Value of d = 025c. There were 4 ponds in series and this was assumed to be close to plug flow. Value of d was taken to be 0.1.d. Taken from Wong and Springer (1980). As POME volume is higher and ponds larger, completely mixed flow was adopted for selected well maintained

and operated ponds.e. Data in this study was taken from a long ditch (5720 feet long) which assumes a plug-flow configuration. The first-order kinetic coefficient is based on

the COD and may be slightly on the low side if converted to BOD.f. Insufficient data from Asairinachan (1979).

PONDING SYSTEMS FOR PALM OIL AND RUBBER MILL EFFLUENTS

(1) Block rubbe(2) Sh<*<?t rubbe< 3) Crcp« rubbe

15) Palm Oi ! Hi

•emi mi »ed flow, d * 0.25wnri mixed f lm/ , d = 0.25iemi -nixed flow, d • 0.25

Figure 2: Performance of ponds for rubber andpaint oil mill effluents.

some treatment systems for domestic waste-waters were used to analyze the financial viabilityof ponding systems. Table 2 shows the con-trasting differences between three different treat-ment systems for domestic wastewaters for apopulation of 20,000 in terms of capital costsand land used (after Asairinachan 1979).

Although the land area required in oxidationponds is much higher, the operating and main-tenance costs are definitely lower at aroundM$5000 per year. Since the price of land isthe constraint, Figure 3 reveals that if land ispriced above M$700,000/ha (1979 figures) then aconventional activated sludge process will becheaper than a two stage oxidation pond in termsof initial capital investment. However, we willstill have to account for annual operating andmaintenance costs which are definitely higherfor the conventional activated sludge process.It appears that installing aerated lagoons isgenerally cheaper than installing oxidation ponds.With high energy prices and substantial wearand tear of mechanical parts, operating andmaintenance costs in the activated sludge process

TABLE 2

Construction Capital Costs and Land Requirement ofTreatment Systems Sized for 20,000 people

(in 1979 figures)

Construction Land AreaCapital Required

Treatment System Cost (ha/10,000(M$ per person) persons)

Conventional Activated 275.00Sludge Plant

Two-stage Oxidation 27.50Ponds

Aerated Lagoons 30.00

0.76

4.25

1.45

can be very prohibitive. Aerated lagoons alsorequire mechanical aeration, hence higher opera-ting and maintenance costs. At the time ofwriting, local data for operating and maintenanceexpenditures in the conventional activated sludgeand aerated lagoons for domestic wastewatertreatment were not available.

Nevertheless, some indications for runningcosts in the anaerobic treatment of palm oil milleffluent are shown in Figure 4. The Figureshows the comparative annual costs of threesystems currently used by palm oil mills inMalaysia: anaerobic ponds, unstirred openanaerobic tanks, and continuous mixed flowclosed anaerobic tanks (with stirrers). Thesecosts are for a design meant to produce an effluentBOD3 of about 2000 ppm. The running costs(in 1979 Malaysian ringgit) include capital costsamortized at a rate of 10% over 10 years, opera-ting costs (mainly electrical energy and man-power) and some maintenance costs. Electricalenergy constitutes the bulk of the running costs.

From the examination of these costs itappears that ponding can be indeed low-costfor the industrial plant if space is available.

ACKNOWLEDGEMENTS

The author wishes to thank Miss Sariatun,Miss Pauline Yeow and Ismail b. Nordin forservices rendered; United Plantations for pro-viding the data on the long ditch treatment ofPOME; and Universiti Pertanian Malaysia forthe time and facilities provided in the preparationof this paper.

139

K. K. WONG

( t ) Convent iona l A c t i v a t e d Sludge P l a n t (2) Two Stage O x i d a t i o n Ponds (3) Aera ted Lagoons

I n i t i a t

C a p i t a l

Costs

( i n lCT Ringgi t )

uu

80

, p

20

1 098

76

(1) /

/t /

<**

, ^ -. — - ^ -1

^ ^

•

'

y

/

——-/

y

/.

/

10 20 60 80 100 200

Land Costs Per Hectare

( i n 103 Ringgi t )

600 800 1000

Figure 3: Initial capital costs for treatment system as a function of land price (1979 Figures)

M

TREATMENT SYSTEM TO PRODUCE EFFLUENT OF 2OO0ppm B O D 3

Figure 4: Running cost estimates for anaerobictreatment systems in Malaysia.

REFERENCES

AHMAD, L, SETHU, S., MOHAMMED, Z.K. and ZAID, I.(1979): Anaerobic/ facultative ponding systemfor latex concentrate effluent - commercial plantstudies. Proc. RRIM Planters' Conf., KualaLumpur.

ASAIRINACHAN, K. (1979): Sewage Treatment -Practical System. Symp. on Recent Developmentin Effluent Treatment Technology. Universiti SainsMalaysia, Penang, Malaysia. Dec 10-11.

BISHOP, A.B. and GRENNEY, WJ. (1976): CoupledOptimization - Simulation Water Quality ModelJ. Env. Eng. Div. Proc. A mer. Soc. Civil Engr12: 1071.

FINNEY, B.A. and EJ . MIDDLEBROOKS (1980): Facul-tative Waste Stabilization Pond Design. J. WaterPollution Control Federation. Six 1 (134-147).

GLOYNA, E.F. (1971): Waste Stabilization Ponds.WHO Monograph Series No. 60. Geneva.

GLOYNA, E.F. (1976): Facultative Waste StabilizationPond Design in "Ponds as a Wastewater Treat-ment Alternative", Water Resources SymposiumNo. 9. E.F. Gloyna, J.F. Malina Jr., and E.MDavis (Eds.). Center for Research in WaterResources. Univ. of Texas. Austin. 143.

140

PONDING SYSTEMS FOR PALM OIL AND RUBBER MILL EFFLUENTS

GAMESON, A.L.H. and WHEATLAND, A.B. (1958):Ultimate Oxygen demand and course of oxidationof Sewage Effluent. J. and Proc. of the Instituteof Sewage Purification. 2: 106.

JOHN, C.K., PONNIAH, CD. , LEE, H. and AHMAD, I.(1974): Treatment of effluent from block rubberfactories. Proc. RRIM Planters Conf. LualaLumpur, p. 229.

JOHN, C.K. and ONG, C T . (1979): Ponding as aTreatment System for Effluent From RubberFactories in Malaysia. 2nd. Symp. on OurEnvironment. Institute of Natural Sciences.College of Graduate Studies, Nanyang Univer-sity, Singapore, Nov. 14-16.

JOHN, C.K. (1979): Treatment Technology of RubberEffluent. Symp. on Recent Dev. in Eff. TreatmentTech. Universiti Sains Malaysia. Penang.

LARSEN, T\B. (1974): A Dimensionless DesignEquation for Sewage Lagoons. Dissertation.Univ. of New Mexico. Albuquerque.

MARAIS, G.R. and M.J. CAPRI (1970): A SimplifiedKinetic Theory for Aerated Lagoons. 2nd. Int.Symp. for Waste Treatment Lagoons. KansasCity, Mo. U.S.A.

MARAIS, G.V.R. (1970): Dynamic Behaviour ofOxidation Ponds. 2nd. Int. Symp. for WasteTreatment Lagoons. Missouri Basin Eng. HealthCouncil. Kansas City, Mo. U.S.A.

MARAIS, G.V.R. (1974):Stabilization Ponds.ASCE. 100: 119.

Faecal bacterial kinetics inJ. of the Env. Eng. Div.

MARA, D. (1976): "Sewage Treatment in Hot Climates'*.ELBS and John Wiley & Sons. New York. p. 81.

MCCARTHY, P.P. (1966): Kinetics of Waste Assimi-lation in Anaerobic Treatment. In ''Develop-ments in Industrial Microbiology". Am ericanInstitute of Biological Sciences, Washington D.C-7: 144-5.

MCGARRY, M.C and PESCOD, M.B. (1970): Stabili-zation Pond Design Criteria for Tropical Asia.2nd. Int. Symp. for Waste Treatment Lagoons.Missouri Basin Eng. Health Council, KansasCity, Mo. 114(1970).

MCKINNEY, R.E. (1971): Waste Lagoons-State ofthe Art. U.S. Env. Protection Agency ProjectNo. 17090 EXH.

METCALF AND EDDY INC. (1979): "WastewaterEngineering: Treatment Disposal Reuse" revisedby G. Tchobanoglous. (2nd ed.). Tata-McGraw-Hill Publishing Co. Ltd. New Delhi.

MONOD, J. (1949): The Growth of Bacterial Cultures.Annual Rev. Microbiol. Vol. III. CE . Clifton(Ed.). Stanford, California. Annual ReviewsInc., pp. 371-394.

MUTHURAJAH, R.N., JOHN, CK. and LEE, H. (1979):Development on the treatment of effluents fromnew process SMR factories. Proc, RRIM PlantersConf., Kuala Lumpur.

PEARSON, E.A. (1968): Kinetics of Biological Treat-ment. In "Advances in Water Improvement".Gloyna, E.F. and W.W. Eckenfelder, Jr. (Eds.).University of Texas Press. Austin. Texas, pp.381-399.

PONNIAH, CD. , CHICK, W H . and SEOW, C M . (1975):Treatment of effluents from rubber processingfactories. Proc. Int. Rubber Conf. Kuala Lumpurp. 367.

PONNIAH, CD. , JOHN, C.K. and LEE, H. (1976):Treatment of effluent from latex concentratefactories - Preliminary report. Proc. RRIMPlanters Conf. Kuala Lumpur, p. 310.

SEOW, C M . (1977): The Role of Chemistry in theAnaerobic Digestion of Palm Oil Effluent. Proc.Symp. on Improving the Quality of Life in Malaysia,Inst. of Chem. Malaysia. Kuala Lumpur, Mar.24-26.

THIRUMURTHI, D. (1969): Design Principles of WasteStabilisation Ponds. J. Sank. Eng. Div. ASCE.95 No. SA2.

THIRUMURTHI, D. (1974): Design Criteria for WasteStabilisation Ponds. J. Water Pollution ControlFederation. 46(9): 2094.

TUCEK, F. and CHUDOBA, J. (1968): PurificationEfficiency in Aeration Tanks with Mixing andPiston Flow. Water Research. 3: 559-570.

WEHNER, J.F. and R.F. WILHELM (1958): BoundaryConditions of Flow Reactor. Chemical Engineer-ing Science. 6: 89.

WHITE, R.K. (1977): Lagoon Systems for AnimalWastes. In "Animal Wastes". E-P. Taiganides(Ed.). Applied Science Publishers. Londonp. 215.

WONG, K.K. and A.M. SPRINGER (1980): A First-order Kinetic model for Designing AnaerobicPonds. In "The Treatment of Paln\ Oil MillEffluent" in Agricultural Wastes. Applied SciencePublishers. London (in press).

YOUNG, J .C and CLARKE, J.W. (1965): Second OrderEquation for BOD. J. of the San. Eng. Div.Amer. Soc of Civil Eng. 91: 43.

(Received 26 September 1980)

141