sustainable water conservation and wastewater reuse in a palm...

Water Qual. Res. J. Canada, 2002Volume 37, No. 4, 711–728Copyright © 2002, CAWQ

* Corresponding author; [email protected]

Sustainable Water Conservation and WastewaterReuse in a Palm Oil Mill:

A Case Study in Southern Thailand

SHING TET LEONG,1* SAMORN MUTTAMARA1 AND PREECHA LAORTANAKUL2

1Urban Environmental Engineering and Management Program, School of Environment,Resources and Development, Asian Institute of Technology, G.P.O. Box 4, Klong Luang,Pathumthani 12120, Thailand2216 Soi Satsana 5, Phya Thai, Bangkok 10400, Thailand

The palm oil industry is one of the major agro-industries in Southern Thailand.It requires a large amount of water for its operation and discharges consider-able quantities of wastewater. This creates a serious threat to the environmentand sources of potable water. This study proposes recommendations for waterconservation and reuse and improvement of wastewater treatment facilities toovercome these problems.

In order to attain the highest reduction of all problems, waste minimizationis introduced as the most effective solution. Changing behaviour in housekeep-ing can reduce water usage. An upflow anaerobic sludge blanket (UASB) processcoupled with an activated sludge plant is recommended to upgrade the waste-water treatment system. For water reuse purposes, a rock bed filtration unit is rec-ommended to treat effluent of the treatment plant.

The overall water balance of the palm oil mill suggests that water reuse of322 m3/d will reduce raw water consumption by 27.66% and achieve a 23%reduction in the water discharged to the river.

Key words: activated sludge plant, palm oil mill, rock bed filtration unit, UASBprocess, water conservation and reuse

Introduction

General

Water problems have been experienced in certain parts of Thailandfor many years. The National Statistical Office, Thailand (NSO 2000),reported that the estimated breakdown of water usage was greatest forthe agricultural sector (475 m3 p-1 yr-1), followed by the industrial sector(32 m3 p-1 yr-1) and domestic sector (21 m3 p-1 yr-1). In the past 15 years,water use in the agro-industrial sector has increased at an extremely rapidrate, especially in the southern part of Thailand. This region is where mostof the palm oil and rubber industries are located and both require a largeamount of water for operation. In Southern Thailand, the dry season isfrom December to April and the wet season is from May to November.

712 LEONG ET AL.

During the dry season, water resources in these areas are depleted due toover-use and poor pollution control. This will be a potential threat to boththe palm oil extraction process and effluent treatment plant performance.However, the severity of the problems varies from year to year, depend-ing upon the rainfall.

There are 49 registered palm oil processing mills in SouthernThailand with an annual production capacity of 411,480 metric tonnes ofcrude palm oil (NSO 2000). Of these, 17 mills use a wet process and rep-resent about 50% of the processed palm oil in local market share. Kuttikun(2000) reported that the extraction of palm oil generates a significantamount of wastewater which is acidic and has high oil/grease and solidcontents. When discharged without treatment, the wastewater is detri-mental to flora and fauna of streams and potable water supplies.Problems are magnified if the receiving stream affords little or no dilutionduring the dry season, giving rise to obnoxious odours adjacent to facto-ries. In addition there can be run-off problems during the rainy season(Thanh et al. 1980). The rapid expansion of the palm oil industry inThailand will continue to be a serious threat to the environment and qual-ity of life in rural areas unless the concepts of a sustainable water policyand wastewater management receive wide enough attention by factoriesto initiate water recycling and efficient wastewater treatment methods.

To protect depletion of natural resources, minimize energy con-sumption and upgrade environmental quality, waste minimization wasfound to be the most promising strategy (U.S. EPA 1988). Wastewaterminimization programs emphasize systematic reduction of wastewatergeneration with optimum utilization of resources (Huisingh 1989).Wastewater minimization results in improved product quality, reducedoperating and maintenance costs and maximization of the competitiveposition in the business (Mendicino and Thomas 1995).

Outline of the Production Process

The selected factory is situated at Krabi, Southern Thailand. The loca-tion of the mill is shown in Fig. 1. The total land area of the mill is131,200 m2 including 51,200 m2 of available land area for the wastewatertreatment system. The mill has an operational capacity of 30 metric tonnesh-1 of fresh fruit bunch (FFB). The factory works on a two-shift day(16 h d-1) and employs 75 workers. The flow diagram for the completemill process is shown in Fig. 2.

The fresh palm fruit bunches (FFB) are sterilized by steam at 90ºC for2 hours before they are subjected to stripping or threshing from the bunchstalks. The loose sterilized fruit is then sent to the digester unit for cook-ing with live steam for 30 minutes. The usual method of extracting crudepalm oil from the digested fruit is by pressing. The oil from the vibratingscreen is pumped to a sand cyclone to remove sand from the oil. The set-tled oily sludge from the oil clarifier is then discharged into a three-phasedecanter system where the sludge is centrifuged to recover the oil. The

WASTEWATER REUSE IN A PALM OIL MILL 713

Fig. 1. Location of the palm oil mill.

714 LEONG ET AL.

liquid phase is recycled back to the oil clarifier to make up dilution water,while the solid is carried away for disposal. The wastewater coming fromthe decanter and the purifier is sent to the sludge tank to further recoveroil from wastewater before it is pumped to the effluent treatment plant. Inaddition, a scrubbing system such as hydrocyclone (wet process) andmulticyclone are used to separate kernels from cracked shells and partic-ulate concentration from the boiler flue gas, respectively, resulting in acontinuous discharge of wastewater.

Currently some of the mills have already adopted a dry process inseparating kernels from cracked shells (wind silo), where air is blown

Fig. 2. Flow diagram for the complete process in the palm oil mill.


from the bottom and kernels are separated from cracked shells by gravi-ty. Wastewater is thus eliminated from this process.

Methodology

The total amount of water usage in the main processing areas can bequantified by directly reading the meters installed at the two pipelines con-nected to the raw water tank. The daily average amount of water used bythe mill is recorded by the Water Supply Utility Section. Water usage in var-ious unit processes was measured either in the form of steam or hot water.Some losses of water were also estimated in this area. Composite watersamples were collected from general floor washing, equipment cleaningand the treatment plant to determine the wastewater characteristics.Wastewater in the open channel was measured by using a container andtimer or current flow meter. A portable ultrasonic flowmeter equipped witha Portaflow sensor was also used for easily measuring the flow rate fromthe outside of a processing pipe. Water and wastewater analytical methodsare strictly followed in the “Standard Methods” (APHA 1992). All analyti-cal measurements are performed in triplicate to give mean values. Resultsof replicated analyses are given in the tables with standard deviation.

Laboratory-Scale Experiment for Wastewater Treatment System

Due to long retention time and obnoxious odour, the existing waste-water treatment system of anaerobic, facultative and aerobic ponds wasfound not suitable for the palm oil mill. Therefore, a laboratory-scaleexperiment was conducted to find an alternative wastewater treatmentfacility for the system. The experiments for laboratory-scale upflow anaer-obic sludge blanket (UASB) process, activated sludge and rock bed filtra-tion units were conducted in the Research Station of the EnvironmentalEngineering Laboratory, Asian Institute of Technology (AIT).

Upflow anaerobic sludge blanket process The laboratory-scale UASB unit has a working volume of 21.5 L and

a gas collection assembly. The arrangement of the laboratory-scale UASBis shown in Fig. 3. All experiments were performed at ambient room tem-perature, 30 to 35°C. Initially, about 50% reactor volume was seeded withactive sludge from the starch wastewater anaerobic lagoon. A COD:N:Pratio of 300:5:1 was maintained during start-up using urea and KH2PO4

and was neutralized by NaOH to a pH of 7. The process takes about44 days to reach steady-state operation. Sufficient upflow velocity wasmaintained to achieve proper fluidization and mixing of sludge granulesinside the reactor. Effluent was collected in a settling tank and the super-natant was partly recirculated back into the reactor. An average of 84% ofthe digested COD was converted into methane. Biogas generated fromthe reactor was stored in a gas collection assembly that worked on thewater displacement principle.

716 LEONG ET AL.

The hydraulic retention time (HRT) of the laboratory-scale UASBunits was operated at 2, 3 and 4 days, respectively, and the organic load-ing rate was fixed at 7 kg BOD m-3 d-1. The pH of the wastewater was inthe range 6.8 to 7.2. The experimental results with the effects of HRT areshown in Table 1. The highest percentage removal efficiencies of SS, BODand COD were 73, 94 and 89%, respectively, when being operated at the3-day HRT.

Activated sludge processThree laboratory-scale units were used. Each has the same dimen-

sion of 35 × 35 × 20 cm. The activated sludge unit was operated athydraulic retention time (HRT) of 1, 2 and 3 days, respectively.Treatability study on activated sludge showed that organic loadingrate was at 1.2 kg BOD m-3 d-1, food to microorganism (F/M) ratiomaintained at 0.33 d-1 and optimum aeration time varied between 17 to23 hours. Regular sludge wasting at recirculation ratio (Qr/Q) of 0.5 iscompulsory and this necessitates competent operation. The activatedsludge treatment performances with the effects of HRT are shown inTable 2. The highest percentages of removal efficiencies of SS, BOD andCOD were 97, 96 and 92, respectively, for the experimental unit operat-ing at the 2-day HRT.

Fig. 3. Schematic diagram of upflow anaerobic sludge blanket unit.


Rock bed filtration unitsThe pilot-scale rock bed filtration units, made of reinforced concrete,

were built at the Environmental Research Station, Asian Institute ofTechnology (Fig. 4), each with a dimension of 0.5 × 4.0 × 0.5 m (W×L×D)and a bed slope of 1%. A 50-cm layer of 2.5-cm diameter rock with aporosity of 0.5 was placed in each rock bed filtration unit. Hydraulicretention time (HRT) was varied at 0.5, 1.0, 2.0 and 4.0 days and organicloading rates were in the range of 0.002 to 0.04 kg BOD m-3 d-1. The exper-imental results with the effects of HRT are shown in Table 3. The treat-ment efficiencies with respect to SS, BOD and COD removal were foundto increase with increasing HRT, the highest percentage being 85, 77 and51%, respectively, for the experiment conducted at the 4-day HRT.Although, the removal efficiency at 4-day HRT was higher compared tothe removal efficiency at 1-day HRT, the 1-day HRT was chosen as theoptimal HRT to economize the size of the rock bed filtration unit.

Results and Discussion

Water Source and Usage

Raw water supplied to the mill is pumped from a nearby canal andis treated by a series of unit processes such as coagulation, flocculation,sedimentation and filtration. Before distribution, clear water is further

Table 1. Mean concentration and % removal of SS, BOD and COD for the UASB unit

HRT (days)

2 3 4

SS(inf),a mg L-1d 10,845 (92.10)e 13,100 (105.00) 8202 (90.00)SS(eff), mg L-1 3370 (37.00) 3520 (38.30) 1722 (23.00)% SS removed 68 73 79

BOD(inf),b mg L-1 9017 (125.00) 19,347 (197.00) 15,044 (175.00)BOD(eff), mg L-1 3139 (37.60) 1069 (13.90) 6274 (20.00)% BOD removed 65 94 58

COD(inf),c mg L-1 8949 (121.00) 33,248 (279.00) 22,435 (243.00)COD(eff), mg L-1 2416 (24.30) 3428 (35.00) 6730 (65.73)% COD removed 73 89 70

a SS; Suspended solid.b BOD; Biological oxygen demand.c COD; Chemical oxygen demand.d All concentrations = mean ± S.D., n = 3. e Values in parentheses are standard deviation.

718 LEONG ET AL.

treated by a mixed-bed filter (sand and activated charcoal) and is storedin a water tower. Water from the raw water tank is supplied to the wholefactory, to the processing plant, and to other areas by two main pipelines.Results of in-plant evaluation studies revealed that the average dailywater consumption of the mill is about 1164 m3d-1. The principal waterusage within the mill is process water (431 m3d-1), equipment cleaning(322 m3d-1), boiler feed water (220 m3d-1), domestic (149 m3d-1) and back-wash (42 m3d-1). There is no control over the water extraction from thecanal and water conservation plans.

Water Conservation and Reuse Strategies in the Mill

During the dry season water from nearby canals is withdrawn fasterthan it can be replenished by rain. Consequently, the decline in canalwater level negatively affects raw water supply. This affects plant opera-tion performance. Thus, a sustainable approach to water management isneeded to conserve water resources. In order to combat limited watersupplies, the owner of the palm oil mill is encouraged to change his tra-ditional water consumption habits and adopt long-term water conserva-tion and reuse strategies. In addition, a water reuse program should beencouraged to create a sustainable water supply.

Water balance was conducted based on actual measurements follow-ing an intensive water usage monitoring program for each key unit oper-

Table 2. Mean concentration and % removal of SS, BOD and COD for the activated sludge unit

HRT (days)

1 2 3

SS(inf),a mg L-1d 4850 (41.70)e 3535 (17.80) 6396 (55.00)SS(eff), mg L-1 1650 (10.98) 110 (1.35) 2894 (20.45)% SS removed 66 97 55

BOD(inf),b mg L-1 891 (6.50) 1086 (10.98) 885 (5.55)BOD(eff), mg L-1 196 (1.84) 43 (0.37) 103 (1.08)% BOD removed 78 96 88

COD(inf),c mg L-1 3419 (37.50) 3242 (35.30) 3077 (33.58)COD(eff), mg L-1 283 (2.97) 236 (1.60) 316 (3.97)% COD removed 91 92 89



ation and wastewater generated in the different processes. Figure 5 delin-eates the overall water balance of the mill indicating the average dailyrates of water usage and wastewater generation from one process toanother. Monitoring of process evaluation of water use and wastewatergeneration also allows some in-plant changes that can be made to mini-mize water usage and recover oil loss. It is possible to recycle someprocess water several times in certain processing areas and treat it at theend of its period of usefulness such as sewage wastewater, storm waterrunoff and steam condensate from the boiler house. Effective implemen-tation of these procedures would conscientiously reduce wastewater vol-ume and pollutant levels in the plant.

Identification of Possible Abatement of Water Usage

There are signs of excessive water use and concerns about control-ling waste strength. There is also evidence that amounts of oil lost in thewastewater are abnormally high due to inadequate oil recycling meth-ods. Palm oil mill effluent (POME) is a mixture of high polluted effluent(from sterilizing station, oil clarification room, press station and hydro-cyclones) and low polluted effluent (from steam condensate, coolingwater, backwash water and sanitary effluent). Normally, low-strengthwastewater is discharged directly into the final oil trap. Instead of beingwastefully discarded, a balancing tank or pond which operates under the

Fig. 4. Schematic diagram of rock bed filtration unit.

720 LEONG ET AL.

same principle as an oil trap should be installed for this wastewater andreused for various purposes like floor and equipment cleaning.Furthermore, it is possible to remove oil from some process water bygravity-type oil separators and recycle it into the process at digester,screw press or settling tank. By segregating wastewater, water quantityto be treated can be reduced, thus, considerably lowering both cost andburden for the wastewater treatment.

Frequently, hand-washing taps in processing areas were foundturned on without being put to any use and generated a large quantity ofunpolluted water. Installation of automatic shut-off taps are recom-mended to reduce the water consumption and wastewater generationfrom hand and body washing. Floor washing is conducted very fre-quently on each shift and whenever there is a spill. The high frequencyof floor and equipment washing is identified as a cause of high waterconsumption and wastewater generation in the mill. Reduction in waterand detergent use can be achieved by using more efficient cleaning tech-niques such as high-pressure, spring-loaded hose nozzles for floor andequipment cleaning purpose.

Leakage from pipelines and clogged gutters were found in manyplaces. All old components of steam and water piping systems should bereplaced whenever leakage is detected. Excessive water overflows fromoverhead tanks were detected in both the press station and oil clarifica-

Table 3. Treatment performances of the rock bed filtration unit at varied HRT values

HRT (days)

0.5 1 2 4

SS(inf),a mg L-1d 57 (0.61)e 128 (1.08) 94 (0.86) 85 (0.71)SS(eff), mg L-1 17 (0.22) 27 (0.24) 15 (0.15) 13 (0.11)% SS removed 70 79 84 85

BOD(inf),b mg L-1 24 (0.21) 48 (0.47) 42 (0.61) 31 (0.32)BOD(eff), mg L-1 10 (0.73) 19 (0.34) 10 (0.86) 7 (0.11)% BOD removed 58 60 76 77

COD(inf),c mg L-1 272 (2.97) 260 (2.55) 283 (3.30) 332 (6.44)COD(eff), mg L-1 217 (1.78) 190 (1.91) 177 (1.62) 163 (1.08)% COD removed 20 27 37 51



Fig.

5.

Wat

er b

alan

ce o

f th

e pa

lm o

il m

ill w

ith

prop

osed

wat

er r

euse

.

722 LEONG ET AL.

tion rooms, which can be eliminated by installing a floating valve. If theseoptions are properly implemented, it can save an estimated 20 to 30% oftreated water per year and can reduce the organic loading to the waste-water treatment plant (WWTP) by the same amount.

Wastewater Generation and Characteristics

The major amount of wastewater was generated from the sterilizingand extraction processes and there was a potential for water reuse in themill. A minimum amount of water was added in the screw press to mini-mize oil loss in the pressed cake. Hot water was added in the sand trap andvibrating screen to remove coarse solids from the crude oil. Hot water wasalso used as make-up water in the decanter and purifier, and in somecases, it was used for equipment cleaning. From the investigation, thewastewater generated was greatest for the processing area (570 m3d-1), fol-lowed by the service area (428 m3d-1) and unaccounted losses (43 m3d-1).The combined wastewater from palm oil extraction is mainly composed ofsterilizing condensate, oil clarification room discharge, hydrocyclone andboiler house effluent, and floor drains. Wastewater from the sterilizer andoil clarification room always has high values of organic, solid and oil con-tents; while hydrocyclone and boiler house wastewater has relativelylower organic, solids and oil contents. Table 4 shows the typical waste-water characteristic from various waste-generating points in the mill.

In general, raw palm oil wastewater is highly organic but has relative-ly low nitrogen and phosphorus concentrations (BOD:N:P ratio of100:1.63:0.95). Aerobic biological treatment requires a BOD:N:P ratio of100:5:1 for proper microbial growth of waste degradation (Thanh et al.1980). This results in operating problems needing nutrient and pH adjust-ment. Despite the fact that the combined wastewater is deficient in nitrogenwith respect to BOD, anaerobic biological treatment appears to be the mostfeasible because of the high proportion of BOD and COD concentrations inthe wastewater (Chin et al. 1996). The BOD to COD ratio was in the rangeof 0.50 to 0.70, indicating the biodegradable organic nature of palm oilwastewater. Total solids were higher than suspended solids, indicating thatmost of the solids were in the dissolved form, thus requiring sedimentationof solids. The pH of the combined wastewater was slightly acidic and var-ied between 3 to 4. Oil and grease content was in the range of 338 to18,500 mg L-1. However, high contents of oil and grease in wastewater sam-ples have caused a great concern in biological treatment. Thus, there is aneed for oil and grease removal before biological treatment. Generally, theamount of oil loss in any palm oil mill can effectively be overcome by mak-ing an improvement on performance of the existing oil trap.

Existing Wastewater Treatment System

The initial wastewater treatment plant designed by the consultants is apond system. The wastewater from the process and the scrubber are collect-


Tab

le 4

.Pa

lm o

il w

aste

wat

er c

hara

cter

isti

cs a

t dif

fere

nt p

roce

ssin

g un

its

Com

bine

d

Para

met

erSt

erili

zer

Dec

ante

rPu

rifi

erH

ydro

cycl

one

Scru

bber

Boi

ler

proc

ess

was

tew

ater

Tem

pera

ture

(ºC

)83

.4

80.3

54

.565

.765

.778

.074

.0

pH4.

85.

05.

85.

87.

86.

43.

8

BO

Da

(mg

L-1

)d73

,815

(54

)e58

,848

(51

)37

,540

(34

)19

50 (

11)

4240

(34

)84

(7.

8)36

,246

(37

)

CO

Db

(mg

L-1

)13

4,21

0 (1

25)

84,8

60 (

81)

52,6

37 (

49)

3600

(37

)17

,465

(16

)18

8 (4

.2)

98,4

84 (

86)

Tota

l sol

ids

(mg

L-1

)63

,240

(65

)42

,160

(44

)18

,500

(21

)26

00 (

22)

24,6

00 (

22)

210

(5.9

)82

,580

(77

)

Susp

. sol

idc

(mg/

L)

38,4

00 (

36)

28,0

00 (

41)

9000

(15

)14

96 (

9.4)

14,9

60 (

10)

54 (

4)47

,940

(49

)

Tota

l nit

roge

n (m

g L

-1)

1306

(8.

4)93

6 (8

.8)

680

(6.7

)47

(3.

2)15

4 (9

.2)

2 (0

.01)

1720

(16

.2)

Tota

l pho

spho

rus

(mg

L-1

)93

0 (9

.3)

150

(7.2

)36

3 (6

.2)

24 (

0.3)

40.2

(3.

4)1

(0.0

1)12

7 (0

.71)

Oil/

grea

se (

mg

L-1

)60

0 (5

.5)

18,5

00 (

17)

6100

(14

)16

00 (

13)

338

(6.5

)94

50 (

14.7

)

a B

OD

; bio

logi

cal o

xyge

n d

eman

d.

bC

OD

; che

mic

al o

xyge

n d

eman

d.

cSu

sp. s

olid

; sus

pend

ed s

olid

s.d

All

conc

entr

atio

ns =

mea

n +

S.D

., n

= 3

. e

Val

ues

in p

aren

thes

es a

re s

tand

ard

dev

iati

on.

724 LEONG ET AL.

ed separately in two ash ponds and thereafter, they are treated by two anaer-obic ponds followed by three facultative waste stabilization ponds in series.The treated effluent is discharged into an earthen polishing pond before it isdisposed into a nearby river. The water from the last pond is recycled backto the scrubber system to remove particulate from the flue gas. In-plant sur-veys show that there are operating problems exhibited in the existing treat-ment plant. Solids and oil/grease removal are not provided in the treatmentand the anaerobic pond has been acting as a basin for solid and oil/greaseseparation. Over a long period of operation, a thick layer of scum covers thetwo anaerobic ponds. This normally results in an acidifying process andodour-generating substrates released causing bad smells. This odour has notbeen a serious problem in the past due to the remoteness of the mill’s loca-tion. However, the areas near to the mill have presently been transformedinto residential areas, and odour is now an air quality problem. Similarly, thethree facultative waste stabilization ponds are found to be operating as anindividual unit with a blackish effluent, which prevents light from penetrat-ing into the algae culture system. The overall BOD, COD and SS removalefficiencies of the existing wastewater treatment system are 55%, 55% and63%, respectively. The effluent BOD from the treatment system is 570 mgL-1 which does not meet the Industrial Effluent Standard of 20 mg L-1 as setby the Pollution Control Department, Thailand.

Modified Wastewater Treatment System

Recommendations are put forward to upgrade the wastewater treat-ment plant. The modified wastewater treatment system consists of anupflow anaerobic sludge blanket process, UASB (Borja-Padilla et al. 1996)followed by an activated sludge plant and polishing pond in series. A set-tling pond with depth of 1.5 m and hydraulic retention time of 1 day is pro-vided as a pretreatment unit for the purpose of oil and grease removal andeasy de-scumming. A pre-settling tank is also used for the purpose of solidremoval from untreated multicyclone scrubber wastewater. Appropriatenutrients are sometimes added to maintain bacterial growth when needed.The neutralized wastewater which has initially been diluted by treated efflu-ent (19,347 mg BOD L-1) is treated in a UASB process for primary stabiliza-tion. The effluent from the UASB process is further treated by an activatedsludge plant which is equipped with three 19-kw aerators. The influent BODis in the range of 2207 mg L-1. The mixed liquor suspended solids in activat-ed sludge are maintained at 2500 to 4000 mg L-1 corresponding to food tomicroorganism (F/M) of 0.33 d-1 and retention time of 2 days. The effluentBOD is brought down to as low as 50 mg L-1. Thereafter, the wastewater issent to a polishing pond before it is discharged into a nearby river. Some por-tions of the treated effluent in the polishing pond are recycled at a ratio of 1:2to the inlet of the settling pond, as this will provide pH and temperature cor-rections, and reduction in BOD concentration of raw wastewater. Averageoverall removal efficiencies of the modified treatment system are 94.5% forCOD, 99.6% for BOD and 90% for total solids (10 days overall detention


time). The effluent BOD from the treatment system is 38 mg L-1, which canalmost meet the Industrial Effluent Standard set by the Pollution ControlDepartment (1997), Thailand. Upgrading the existing wastewater treatmentplant benefits land reclamation by improved performance of the wastewatertreatment plant. The detailed flow diagram and characteristics for theupgraded treatment plant are presented in Fig. 6 and Table 5, respectively.

Tertiary Treatment and Wastewater Reuse

The daily volume of wastewater discharged into the wastewater treat-ment plant (WWTP) is 1006 m3d-1 of which 115 m3d-1 is lost due to evapo-ration and soil seepage and 682 m3d-1 is discharged into the river. Theremaining volume of 209 m3d-1 together with backwash (42 m3d-1) and con-densate (79 m3d-1) is recommended to be reused and undergoes tertiarytreatment to remove suspended and dissolved substances remaining afterconventional secondary treatment. Research conducted by Hosokawa et al.(1992) revealed that effluent of treatment plants can be effectively treated bya rock bed filtration unit. This method is considered as low-cost water treat-ment based on natural purification processes. In adequate time, suspendedsolids are removed by biodegradation, sedimentation and filtration in therock media or void space. The proposed rock bed unit is made of 8 rein-forced concrete channels, each with a dimension of 2 × 2 × 20 m (D×W×L)and the average diameter of rocks are 15 cm for the first 5-m length, 10 cmfor next 5-m length and 5 cm for the last 10-m length. The porosity of rockis 50%. The selection of rock media having high specific surface area isimportant to enhance more biomass growth. The rock media should also below in cost, readily available at the site, easy to install and easy to take outfrom the reactor. The hydraulic retention time of the bed is one day with an

Fig. 6. Schematic diagram of upgraded wastewater treatment plant.

726 LEONG ET AL.

organic loading rate of 0.01 kg BOD m-3d-1. It is also required to flush therock bed filtration unit periodically (8 m3d-1 of water) to overcome mosqui-to breeding and heavy algae growth inside the rock media.

The major concerns for wastewater reuse application are health riskscaused by pathogens and by aesthetics related to suspended solids andturbidity. To attain effective disinfection in the water-reuse system, theeffluent must be low in suspended solids and turibidity, which can inter-fere with disinfectant dose and contact time. Rock bed filtration can beused to improve the turbidity quality of the reclaimed wastewater and tomeet the stringent disinfection requirement (Tchobanoglous and Burton1991). About 89.6% of the total suspended matter, 73.7% of the BOD, and45.6% of the COD can be removed by the rock bed filters to obtain a cleareffluent, substantially free from matter in suspension or in the colloidalstate (Hosokawa et al. 1992). Table 6 shows effluent characteristics of finaldischarge from the treatment systems. The water which is free of sus-pended solids, pathogen and colloidal matter is feasible for reuse inprocesses which do not require high-quality water such as cleaning, flush-ing, washing and wet scrubbing.

The reuse of 322 m3d-1 water will reduce raw water consumption by27.7%. This reuse strategy will reduce the water discharged to the riverfrom the original 883 m3d-1 to 682 m3d-1, achieving 23% reduction. It isalso cheaper to reuse water, as the cost of raw water treatment is0.2 US$ m-3 compared with 0.1 US$ m-3 for treating reused water. Table 7shows a comparison between no water-reuse and water-reuse systems.The attractive water reuse incentives have given encouragement to thefactory owner to adopt a long-term economic implementation inincreased water conservation and wastewater reuse plans.

Conclusion

In the study, technology modification, good housekeeping and wastesegregation, and water conservation and reuse are found to be the most

Table 5. Characteristics of the proposed upgrading wastewater treatment plant

UASB Activated Polishing process sludge pond

Area (m2) 256 700 5000

Depth (m) 12 4 1.5

Flow rate (m3d-1) 1090 1497 1497

Retention time (d) 3 2 5

BOD loading (kg BOD m-3d-1) 7 1.2


probable options for waste minimization. The pond system has originallybeen adopted by the palm oil mill for wastewater treatment and is foundto be inefficient, and this is mainly due to a lack of knowledge in manag-ing the pond system. In addition, anaerobic conditions generate odourproblems in the anaerobic pond which is unacceptable to the local com-munity. Recommendations are proposed to upgrade the treatment systemusing upflow anaerobic sludge blanket process in combination with anactivated sludge plant. Although the initial investment will be high, theUASB process proves to be an economical possibility. It operates at lowerproduction costs, offers a benefit for biogas production and yields lessodour. For water reuse, a rock bed tertiary treatment system is recom-mended to polish WWTP effluent. Under the recommendations of theconsultants, the factory is now implementing the various proposed strate-

Table 6. Effluent characteristics of final discharge from treatment systems

Before After Rock bed Parameter upgraded upgraded unit

Temperature (ºC) 29 28 27

pH 8 7.8 7

BODa (mg L-1)d 570 (5.6)e 38 (2.7) 10 (0.9)

CODb (mg L-1) 1543 (15.4) 230 (4.5) 125 (3.6)

Susp. solidc (mg L-1) 1300 (9.7) 164 (2.8) 17 (1.2)

Total nitrogen (mg L-1) 85.6 (4.8) 34 (2.7) 27 (1.9)

Total phosphorus (mg L-1) 14.33 (1.5)

a BOD; biological oxygen demand.b COD; chemical oxygen demand.c Susp. solid; suspended solids.d All concentrations = mean + S.D., n = 3. e Values in parentheses are standard deviation.

Table 7. Comparison of the no water-reuse and water-reuse systems

No water Water Parameter reuse system reuse system

Raw water consumption (m3d-1) 1164 842Water discharge to river (m3d-1) 883 682Annual water treatment cost (US$) 83,808 72,216

728 LEONG ET AL.

gies. Upon completion of the project, an annual maximum conservationof about 116 thousand cubic metres of water is expected.

References

American Public Health Association (APHA). 1992. Standard methods for theexamination of water and wastewater. 18th Edition. American Public HealthAssociation. Washington D.C.

Borja-Padilla R, Banks CJ, Sanchez E. 1996. Anaerobic treatment of palm oileffluent in a two stage anaerobic sludge blanket (UASB) System. J. Bio-Technol. 45:125–135.

Chin KK, Lee SW, Mohammad HH. 1996. A study of palm oil effluent treatmentusing a pond system. Water Sci. Tech. 34(11):119–123.

Hosokawa T, Ootsuki T, Niwa C. 1992. Channel experiments on water purifica-tion by porous bed using crushed stones. Water Sci. Tech. 26(9-11):2007–2010.

Huisingh D. 1989. Cleaner technologies through process modifications, materialsubstitution and ecologically based ethical values. Ind. Environ. 12(1):4–8.

Kittikun A, Prasertan P, Srisuwan G, Krause A. 2000. Environmental manage-ment for palm oil mill. Material flow analysis of integrated bio-systems.(http://www.ias.unu.edu/proceedings/icibs/ic-fa/kittikun/paper.html).

Mendicino L, Thomas D. 1995. Developing and maintaining a pollution preven-tion program. Industrial pollution prevention handbook, Freeman HM (ed.).McGraw Hill Inc. 99-119.

National Statisatical Office (NSO). 2000. Statistical reports of region. Office of thePrime Minister, Statistical Data Bank and Information DisseminationDivision, Larn Luang Road, Bangkok 10100, Thailand.

Pollution Control Department. 1997. Laws and standards on pollution control inThailand.

Tchobanoglous G, Burton FL. 1991. Wastewater engineering: treatment, disposaland reuse. Third edition. Series in water resource and environmental engi-neering. Metcalf and Eddy, Inc. McGraw-Hill Book Co. Singapore.

Thanh NC, Muttamara S, Lohani BN. 1980. Palm oil wastewater treatment studyin Malaysia and Thailand. International Development Research Centre. FinalReport No. 114. Asian Institute of Technology, Thailand.

United States Environmental Protection Agency (U.S. EPA). 1988. Waste mini-mization opportunity assessment manual, Hazardous Waste EngineeringResearch Laboratory. United States Environmental Protection Agency,Cincinnati, Ohio.

sustainable water conservation and wastewater reuse in a palm...

Documents